1commit 86a72cb1927dd91132d231bb7920b651704601ef 2Author: Matt Turner <mattst88@gmail.com> 3Date: Thu Jul 29 11:23:15 2021 -0700 4 5 xserver 1.20.13 6 7 Signed-off-by: Matt Turner <mattst88@gmail.com> 8 9commit f85e4edba489316950fc3422f39e3c2ef8fb1ac5 10Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 11Date: Thu Jul 22 13:12:05 2021 +0200 12 13 modesetting: unflip not possible when glamor is not set 14 15 This is fixing crashes of xfce when running under qemu 16 17 (cherry picked from commit 8836b9d243444031b6396d39d345f2f83b5fa6a9) 18 19commit b03d7184b7906121a432b05473675937ab6c0f1e 20Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 21Date: Mon Jul 19 14:25:28 2021 +0200 22 23 present: fallback get_crtc to return crtc belonging to screen with present extension 24 25 Since crtc can belong to secondary output that may not have present 26 extension enabled we should fallback to first enabled crtc or fake crtc. 27 28 Fix for issue xorg/xserver#1195 29 30 (cherry picked from commit d6c02ffd9c910637f6b3b7249507998e9e45f93c) 31 32commit b88ad0b342c954109bfd4b1d32c04ff55545cb0f 33Author: Matt Turner <mattst88@gmail.com> 34Date: Thu Jul 8 12:15:21 2021 -0700 35 36 xserver 1.20.12 37 38 Signed-off-by: Matt Turner <mattst88@gmail.com> 39 40commit 8dea50223b9b9495aa2e2f4e6aef2a842c0e0a2f 41Author: Matt Turner <mattst88@gmail.com> 42Date: Thu Jul 8 12:15:13 2021 -0700 43 44 Build xz tarballs instead of bzip2 45 46 Signed-off-by: Matt Turner <mattst88@gmail.com> 47 48commit b7adbac7d14bca3b0790fd7d1a706be28d954af5 49Author: orbea <orbea@riseup.net> 50Date: Wed Jun 23 05:50:21 2021 -0700 51 52 hw/dmx/config: Link directly with libdmxconfig.a 53 54 When building xserver with slibtool (https://dev.midipix.org/cross/slibtool) 55 the build will fail. 56 57 /usr/bin/ld: cannot find -ldmxconfig 58 59 This is because xserver creates libdmxconfig.a internally and then links with 60 the linker flag -ldmxconfig. However according to automake documentation the 61 -lfoo linker flags should only be used for external dependencies and all 62 internal libraries should be linked with the libtool archive file (.la) or 63 the static archive (.a) when the former is not available. 64 65 GNU libtool is far more permissive and happens to silently obscure this issue 66 while slibtool fails because it instead sees '-L./.libs -ldmxconfig'. 67 68 Signed-off-by: Matt Turner <mattst88@gmail.com> 69 (cherry picked from commit 2531ee02450e3be3549daa09407557c3addeba02) 70 71commit e49738f0c9d4370c7a7e6e86c2c74151aa2e33fa 72Author: Zoltán Böszörményi <zboszor@pr.hu> 73Date: Mon Jun 21 12:12:41 2021 +0200 74 75 xfree86: Fix NULL pointer dereference crash 76 77 screenp->displays[count] (passed to configDisplay() in 78 configScreen()) is NULL if there is no Virtual setting 79 in the configuration. 80 81 Fixes: f8a6be04d0c7e6a99824ff888ad6c010960c5c21 ("xfree86: Change 82 displays array to pointers array to fix invalid pointer issues 83 after table reallocation") 84 85 Signed-off-by: Zoltán Böszörményi <zboszor@gmail.com> 86 (cherry picked from commit ef89b6648e2a806237a6d2fa598e1b9c83f128b4) 87 88commit 9a59631a490876eae57f813e3a274c46d35d5287 89Author: Olivier Fourdan <ofourdan@redhat.com> 90Date: Fri Jun 18 14:52:55 2021 +0200 91 92 glx: Fix use after free in MakeCurrent 93 94 The fix from commit c468d34c7 - "glx: Set ContextTag for all contexts" 95 is actually incomplete, it correctly sets the context tag for direct 96 contexts as well, but would fail to mark the context's currentClient. 97 98 As a result, when the context is destroyed, it would be freed 99 immediately rather than being just scheduled for deletion, even though 100 it is still current for some client. leading to a use-after-free. 101 102 Make sure to also set the context's currentClient for direct contexts as 103 well, not just indirect ones. 104 105 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 106 Fixes: c468d34c7 - "glx: Set ContextTag for all contexts" 107 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1186 108 Reviewed-by: Adam Jackson <ajax@redhat.com> 109 (cherry picked from commit aad61e8e03311eb8bae4f7db59e65634733eadc2) 110 111commit 1fc0ca6e680fbc50b7a4b60b8797d0fce69cf6fb 112Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 113Date: Wed Jun 23 15:26:23 2021 +0200 114 115 xfree86: Fix out of array bound access to xf86Entities 116 117 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 118 (cherry picked from commit d66b7ec12986efd4f32fcc84cdaacbd04ec98ccb) 119 120commit a4210fe07a8bd5fd4851ffa202ecffca1cc8062c 121Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 122Date: Tue Jun 8 20:02:21 2021 +0200 123 124 xfree86: Change displays array to pointers array to fix invalid pointer issues after table reallocation 125 126 There are rare cases when xf86SetDepthBpp is resizing displays array in confScreen. 127 As that array is shared between set of ScrnInfoRec's then realloc might invalidate chached DispPtr display values in 128 otheres ScrnInfoRec objects. 129 130 If we will change displays array as an array of pointers to DispRec then cached DispRec pointers in ScrnInfoRec 131 won't be invalid after reallocation of displays array. 132 133 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 134 (cherry picked from commit f8a6be04d0c7e6a99824ff888ad6c010960c5c21) 135 136commit 22449f6300aad600cc49024640651a46a128e3ad 137Author: Jose Maria Casanova Crespo <jmcasanova@igalia.com> 138Date: Tue Jun 8 14:52:17 2021 +0200 139 140 glamor: Avoid using GL_QUADS on V3D 141 142 Like in 0e3f1252dacdc3194 ("glamor: Avoid using GL_QUADS on VC4") 143 this will avoid mesa to fallback doing conversion for QUADS primitives. 144 145 Signed-off-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com> 146 Reviewed-by: Emma Anholt <emma@anholt.net> 147 (cherry picked from commit 9adb13e296baead53112a268ddfd291195742c71) 148 149commit d16f6403153596b62c555fb511881289f59f9e81 150Author: Ville Syrjälä <ville.syrjala@linux.intel.com> 151Date: Tue Sep 3 16:44:56 2019 +0300 152 153 modesetting: Fix possible_crtcs 154 155 Populate outout possible_crtcs as the union of possible_crtcs from 156 the encoders rather than the intersection. Otherwise we're easily left 157 with possible_crtcs==0 when all the possible encoders have 158 non-overlapping possible_crtcs. 159 160 No idea what the magic 0x7f is about, but keep it around in case 161 it matters. 162 163 Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> 164 (cherry picked from commit 0e4bd71d02d09c320e76280ae8f14df169387507) 165 166commit f1e76731a02e77c73521134933e424a17df40a1c 167Author: Ville Syrjälä <ville.syrjala@linux.intel.com> 168Date: Tue Aug 20 17:46:09 2019 +0300 169 170 modesetting: Update props for dynamically added outputs 171 172 Dynamically added outputs should have their properties 173 properly updated as well. Otherwise we're left with an output 174 with many of its propeties not exposed. 175 176 Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> 177 Reviewed-by: Michel Dänzer <michel@daenzer.net> 178 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1143 179 (cherry picked from commit 0c5179c280a193a1dbbe29160906087f06eae6f2) 180 181commit 7bf4777116b9b6d416914561015662f494bb4184 182Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 183Date: Mon Mar 8 21:02:40 2021 -0800 184 185 xquartz: Ensure the mouse pointer is shown when switching to a native macOS alert or window 186 187 Fixes: https://github.com/XQuartz/XQuartz/issues/100 188 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 189 (cherry picked from commit 4ae36f418ed683532b0b6991e4a1e7d56740b845) 190 191commit 3ef6e7b766a9f20b208b50203ad6b45f65fbc29e 192Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 193Date: Fri Mar 5 22:57:24 2021 -0800 194 195 xquartz: Fix appFlags build failure with macOS 10.15+ SDKs 196 197 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 198 (cherry picked from commit e59848548a462f1a0ac751bb8bb6afe56258cf4e) 199 200commit 2a83c840c8f38281b7f93b5522e20a71f6723b08 201Author: Olivier Fourdan <ofourdan@redhat.com> 202Date: Thu Jun 3 17:51:01 2021 +0200 203 204 glx: Set ContextTag for all contexts 205 206 Currently, xorgGlxMakeCurrent() would set the context tag only for 207 indirect GLX contexts. 208 209 However, several other places expect to find a context for the tag or 210 they would raise a GLXBadContextTag error, such as WaitGL() or WaitX(). 211 212 Set the context tag for direct contexts as well, to avoid raising an 213 error and possibly killing the client. 214 215 Thanks to Erik Kurzinger <ekurzinger@nvidia.com> for spotting the issue. 216 217 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 218 Reviewed-by: Adam Jackson <ajax@redhat.com> 219 (cherry picked from commit c468d34c7208c9041f9c077b54a00ae9cccad6a3) 220 221commit 394b6cc1cf614ba7b13db7976b0739035da86d28 222Author: Erik Kurzinger <ekurzinger@nvidia.com> 223Date: Thu Dec 10 14:24:32 2020 -0800 224 225 glx: don't create implicit GLXWindow if one already exists 226 227 If a GLXMakeCurrent request specifies an X window as its drawable, 228 __glXGetDrawable will implicitly create a GLXWindow for it. However, 229 the client may have already explicitly created a GLXWindow for that X 230 window. If that happens, two __glXDrawableRes resources will be added 231 to the window. 232 233 If the explicitly-created GLXWindow is later destroyed by the client, 234 DrawableGone will call FreeResourceByType on the X window, but this 235 will actually free the resource for the implicitly-created GLXWindow, 236 since that one would be at the head of the list. 237 238 Then if the X window is destroyed after that, the resource for the 239 explicitly-created GLXWindow will be freed. But that GLXWindow was 240 already destroyed above. This crashes the server when it tries to call 241 the destroyed GLXWindow's destructor. It also means the 242 implicitly-created GLXWindow would have been leaked since the 243 FreeResourceByType call mentioned above skips calling the destructor. 244 245 To fix this, if __glXGetDrawable is given an X window, it should check 246 if there is already a GLXWindow associated with it, and only create an 247 implicit one if there is not. 248 249 Signed-off-by: Erik Kurzinger <ekurzinger@nvidia.com> 250 Reviewed-by: Adam Jackson <ajax@redhat.com> 251 (cherry picked from commit b7a85e44da91d1663d5b4eabac06327c92a80f91) 252 253commit 0086535e7a331f81823a0f4e578bc3346ee1c312 254Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 255Date: Fri Sep 18 14:32:01 2020 +0200 256 257 modesetting: Remove few common functions from ms namespace 258 259 A lot of that code is the same as in xf86-amdgpu and xf86-nouveau drivers. By removing that functions from 260 ms namespace we can move that code to common implementation. 261 262 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 263 (cherry picked from commit 5be3b80b8d084ca5721be8791910d5827d1b6014) 264 265commit 321964443be6b6daae0f7c1097a34106782ae338 266Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 267Date: Fri Sep 18 14:31:53 2020 +0200 268 269 modesetting: remove unnecessary ms_covering_xf86_crtc dup of ms_covering_randr_crtc 270 271 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 272 (cherry picked from commit c282be503e267afdd050df837d4cf61e0e7287b7) 273 274commit 52eb801d048f63952ca48487d86ba089dfcbe131 275Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 276Date: Fri Sep 18 14:31:47 2020 +0200 277 278 modesetting: Find crtc on slave outputs as fallback instead of returning primary crtc 279 280 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 281 (cherry picked from commit b923364c5ec8251e81b4aa5fd451d9d7bb038009) 282 283commit c7a2da7b994cdf829d805944c80f9fefd9246109 284Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 285Date: Fri Sep 18 14:31:34 2020 +0200 286 287 present: fix msc offset calculation in screen mode 288 289 Instead of getting the current msc value from the window crtc, 290 which not exist take the last saved msc value saved in 291 the window_priv struct 292 293 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 294 (cherry picked from commit d2dce22b6775c772282ac2c0e224ef9685ab34fa) 295 296commit 31544e68e605f0595e700bf4ce468b16e7b195ac 297Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 298Date: Fri Sep 18 14:31:25 2020 +0200 299 300 present: Use crtc's screen present operation for syncing 301 302 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 303 (cherry picked from commit 908deb0895f9fe345d70be933d251ea10511d30a) 304 305commit 464cbee1c625d60ac9e18b7f4e3d53bc83b0d585 306Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 307Date: Fri Sep 18 14:29:25 2020 +0200 308 309 modesetting: Initialize present extension despite glamor is disabled 310 311 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 312 (cherry picked from commit d03c0de77ba36bb5221d00ac84dcdd6aae0df453) 313 314commit 4e11bd390a28144fac03dc21a6caf0800596e2fa 315Author: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 316Date: Tue Jun 8 18:53:07 2021 +0200 317 318 modesetting: Disable reverse prime offload mode for displays running on evdi,udl 319 320 This mode for displays running on evdi/udl as side effect of failed glamor_egl_init 321 reverse_prime_offload_mode was initialized to FALSE 322 323 After Mesa upgrade to 21.0.0 GL_RENDERER is not llvmpipe that results in successful glamor_egl_init 324 and reverse_prime_offload_mode enabled. 325 326 This commit is explicitly disabling reverse_prime_offload_mode for evdi and udl drivers 327 328 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@synaptics.com> 329 (cherry picked from commit 7e7c147105ac704a7b33bdee0f85dc9ccdd8f747) 330 331commit 6272529339f975777f22bb0e6162ac43480ae257 332Author: Olivier Fourdan <ofourdan@redhat.com> 333Date: Mon Feb 15 11:33:56 2021 +0100 334 335 dix: Guard against non-existing PtrFeedbackPtr 336 337 Trying to change the pointer control settings on a device without 338 PtrFeedbackPtr would be a bug and a crash in the Xserver. 339 340 Guard against that case by returning early with a BadImplementation 341 error, that might kill the X11 client but the Xserver would survive. 342 343 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 344 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 345 Related: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1137 346 (cherry picked from commit dee2bb033e78e2f4aae7f5708c3d4fc74a9957de) 347 348commit d41b43345cde2330648ad1890fb95bdbd5362f7f 349Author: Olivier Fourdan <ofourdan@redhat.com> 350Date: Mon Feb 15 11:23:59 2021 +0100 351 352 xwayland: Add PtrFeedback to the touch device 353 354 Trying to change the acceleration/threshold on Xwayland cannot work, and 355 the corresponding handler xwl_pointer_control() is a no-op. 356 357 Yet, an X11 client trying to change those on the touch device may 358 possibly cause a crash because the touch device in Xwayland doesn't set 359 that. 360 361 Initialize the touch device's PtrFeedback to make sure that just cannot 362 happen. 363 364 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 365 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 366 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1137 367 (cherry picked from commit ab76272a7d2bff997985893c89147412a7360310) 368 369commit 23a53f0d5460eb45f05b0b370dfcec712652598c 370Author: Vasily Khoruzhick <anarsoul@gmail.com> 371Date: Fri Mar 20 20:36:25 2020 -0700 372 373 glx: fixup symbol name for get_extensions function 374 375 glxProbeDriver() concatenates __DRI_DRIVER_GET_EXTENSIONS with driver name 376 to get symbol name for get_extension function. Unfortunately that doesn't 377 work for drivers that have hyphen in their name, e.g. sun4i-drm -- 378 get_extensions() for these uses underscore instead. 379 380 As result dlsym() doesn't find get_extension() function and AIGLX 381 initialization fails resulting in following message in Xorg.0.log: 382 383 (EE) AIGLX error: sun4i-drm does not export required DRI extension 384 385 Replace all non-alpha-numeric characters with underscore to fix the issue. 386 387 Signed-off-by: Vasily Khoruzhick <anarsoul@gmail.com> 388 (cherry picked from commit b56e501092169a9c0a60663d832ee71898a8bc4b) 389 390commit 6b767cdf6574081164aa6c6a7b617cda579b84c6 391Author: Matt Turner <mattst88@gmail.com> 392Date: Tue Apr 13 10:01:34 2021 -0400 393 394 xserver 1.20.11 395 396 Signed-off-by: Matt Turner <mattst88@gmail.com> 397 398commit a1a1aa2c14636284669b28a956d756d5705dcf15 399Author: Matthieu Herrb <matthieu@herrb.eu> 400Date: Sun Mar 21 18:38:57 2021 +0100 401 402 Fix XChangeFeedbackControl() request underflow 403 404 CVE-2021-3472 / ZDI-CAN-1259 405 406 This vulnerability was discovered by: 407 Jan-Niklas Sohn working with Trend Micro Zero Day Initiative 408 409 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 410 411commit 8890c44a75304097667ac7d42e83e2d78b105cb5 412Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 413Date: Sun Feb 21 21:49:58 2021 -0800 414 415 xquartz: Remove a check for NSAppKitVersionNumber >= NSAppKitVersionNumber10_7 416 417 This check is always true on our supported systems. 418 419 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 420 (cherry picked from commit 5ae47a9d579e8cb0fbe938455faea53ed75d7098) 421 422commit 3c3680c366f78168e828e4a35d41a048c6e2bb54 423Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 424Date: Sun Feb 21 21:49:14 2021 -0800 425 426 xquartz: Don't include strndup.c any more since we no longer support 10.8 and older 427 428 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 429 (cherry picked from commit b960675257d301605def6008bddcccb0980a3c4c) 430 431commit e1fdc856aedfcb4788011415930a0c6861df5123 432Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 433Date: Sun Feb 21 20:58:42 2021 -0800 434 435 xquartz: Add a launch trampoline to better integrate with modern versions of macOS 436 437 Fixes: https://github.com/XQuartz/XQuartz/issues/6 438 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 439 (cherry picked from commit 694724e42c4a3eadb32891220084b03504f9586b) 440 441commit 8f8e9c53e71ca58e8c23482d8f37687840371cda 442Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 443Date: Fri Feb 19 23:16:56 2021 -0800 444 445 xquartz: Don't process AppKit events if we haven't finished initializing 446 447 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 448 (cherry picked from commit 25035229b73742f9e6a96ac3e535b30b5c6196a8) 449 450commit aa6f84021aaae145f54ebf98787e363e1c2022c6 451Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 452Date: Sat Feb 20 17:24:14 2021 -0800 453 454 xquartz: Allocate each fbconfig separately 455 456 A change during the 1.20 development cycle resulted in fbconfigs being walked 457 and deallocated individually during __glXScreenDestroy. This change 458 now avoids a use-after-free caused by that change. 459 460 ==50859==ERROR: AddressSanitizer: heap-use-after-free on address 0x00010d3819c8 at pc 0x0001009d4230 bp 0x00016feca7a0 sp 0x00016feca798 461 READ of size 8 at 0x00010d3819c8 thread T5 462 #0 0x1009d422c in __glXScreenDestroy glxscreens.c:448 463 #1 0x10091cc98 in __glXAquaScreenDestroy indirect.c:510 464 #2 0x1009d2734 in glxCloseScreen glxscreens.c:169 465 #3 0x100740a24 in dix_main main.c:325 466 #4 0x10023ed50 in server_thread quartzStartup.c:65 467 #5 0x199ae7fd0 in _pthread_start+0x13c (libsystem_pthread.dylib:arm64e+0x6fd0) 468 #6 0x199ae2d38 in thread_start+0x4 (libsystem_pthread.dylib:arm64e+0x1d38) 469 470 0x00010d3819c8 is located 200 bytes inside of 12800-byte region [0x00010d381900,0x00010d384b00) 471 freed by thread T5 here: 472 #0 0x101477ba8 in wrap_free+0x98 (libclang_rt.asan_osx_dynamic.dylib:arm64e+0x3fba8) 473 #1 0x1009d4240 in __glXScreenDestroy glxscreens.c:449 474 #2 0x10091cc98 in __glXAquaScreenDestroy indirect.c:510 475 #3 0x1009d2734 in glxCloseScreen glxscreens.c:169 476 #4 0x100740a24 in dix_main main.c:325 477 #5 0x10023ed50 in server_thread quartzStartup.c:65 478 #6 0x199ae7fd0 in _pthread_start+0x13c (libsystem_pthread.dylib:arm64e+0x6fd0) 479 #7 0x199ae2d38 in thread_start+0x4 (libsystem_pthread.dylib:arm64e+0x1d38) 480 481 previously allocated by thread T5 here: 482 #0 0x101477e38 in wrap_calloc+0x9c (libclang_rt.asan_osx_dynamic.dylib:arm64e+0x3fe38) 483 #1 0x100925a40 in __glXAquaCreateVisualConfigs visualConfigs.c:116 484 #2 0x10091cb24 in __glXAquaScreenProbe+0x224 (X11.bin:arm64+0x100730b24) 485 #3 0x1009cd840 in xorgGlxServerInit glxext.c:528 486 #4 0x10074539c in _CallCallbacks dixutils.c:743 487 #5 0x100932a70 in CallCallbacks callback.h:83 488 #6 0x100932478 in GlxExtensionInit vndext.c:244 489 #7 0x10020a364 in InitExtensions miinitext.c:267 490 #8 0x10073fe7c in dix_main main.c:197 491 #9 0x10023ed50 in server_thread quartzStartup.c:65 492 #10 0x199ae7fd0 in _pthread_start+0x13c (libsystem_pthread.dylib:arm64e+0x6fd0) 493 #11 0x199ae2d38 in thread_start+0x4 (libsystem_pthread.dylib:arm64e+0x1d38) 494 495 Regressed-in: 4b0a3cbab131eb453e2b3fc0337121969258a7be 496 CC: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 497 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 498 (cherry picked from commit 487286d47260782d331229af10df17711cbca1ea) 499 500commit 7aa51bb5728f001bdffdca7f669db1678d903160 501Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 502Date: Sat Feb 20 15:17:18 2021 -0800 503 504 xquartz: Fix a compiler warning about const incompatible pointer assignment 505 506 driWrap.c:541:30: error: assigning to 'GCOps *' (aka 'struct _GCOps *') from 'const GCOps *' (aka 'const struct _GCOps *') discards 507 qualifiers [-Werror,-Wincompatible-pointer-types-discards-qualifiers] 508 pGCPriv->originalOps = pGC->ops; 509 ^ ~~~~~~~~ 510 511 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 512 (cherry picked from commit 6a83fb51b7a8b2e167e7d6380229b69e5452f91f) 513 514commit d751c46bd7ce1f350df04cfc9ec9f3ba73058ccd 515Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 516Date: Thu Feb 18 23:16:15 2021 -0800 517 518 xquartz: Fix build with sparkle enabled 519 520 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 521 (cherry picked from commit a3ddcdd56c246e2226c7cdf372c2a1294eb6d888) 522 523commit 03c2e12a82e772461e069addbafdecdfff029ca5 524Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 525Date: Thu Feb 18 12:43:53 2021 -0800 526 527 xquartz: Silence a compiler warning about missing internal methods on NSApplication 528 529 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 530 (cherry picked from commit 279bcbd9cf9e557a6789d6e4cede8e8799c3788e) 531 532commit fcbd57367ef2daa125d7acbbf00953477f9dc5cc 533Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 534Date: Thu Feb 18 22:19:21 2021 -0800 535 536 xquartz: Rewrite Window menu handling to not depend on X11App.windowsMenu.numberOfItems being correct in -awakeFromNib 537 538 Fixes: https://github.com/XQuartz/XQuartz/issues/56 539 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 540 (cherry picked from commit fe89c70e472a9da0541b798eea60c5362b49a99d) 541 542commit b27c6602b39e02aa7ef25ea03b59ef499fc3dc1e 543Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 544Date: Thu Feb 18 15:41:29 2021 -0800 545 546 xquartz: Convert X11Controller ivars into @properties 547 548 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 549 (cherry picked from commit 41aed8f69634ec61ea0e40fff1cfdaf868be843e) 550 551commit 625c7e4deced7e57a724b05750a92586a2785ff0 552Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 553Date: Thu Feb 18 14:39:46 2021 -0800 554 555 xquartz: Convert X11Application ivars into @properties 556 557 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 558 (cherry picked from commit c2750e1fab774c8e6675ecf284124ff55b5be9cf) 559 560commit 3017fec60581c1ab32ad3d6768baf04179fda5e2 561Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 562Date: Thu Feb 18 14:47:50 2021 -0800 563 564 xquartz: Fold quartzCommon.h into quartz.h 565 566 Everything declared in it comes from quartz.c, so match reality. 567 568 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 569 (cherry picked from commit f51b97b0de2e562e341f2d72c5f00a74c71a159f) 570 571commit bdaff44f98587fd0112f3d86a461ee3ce190277c 572Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 573Date: Thu Feb 18 13:00:43 2021 -0800 574 575 xquartz: Fold away some unnecessary hops to X11Controller through X11Application 576 577 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 578 (cherry picked from commit 4b4500c48f06e7ef41cd94f417e49b3f4f1412ae) 579 580commit 7d22031a6bd949f9357904310422901515a5ace7 581Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 582Date: Thu Feb 18 09:41:30 2021 -0800 583 584 xquartz: Fold away array_with_strings_and_numbers and simplify with more modern Objective-C 585 586 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 587 (cherry picked from commit 39c0e1c0ab6a0a89a71f26446973c779ca7fd927) 588 589commit 937b63ff44edecc58ef020a0bfa453d98332482c 590Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 591Date: Thu Feb 18 09:33:56 2021 -0800 592 593 xqaurtz: Remove message_kit_thread() and use dispatch instead 594 595 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 596 (cherry picked from commit 87f8fe1f74f10faf0ffc84f03539799ad4c2465e) 597 598commit e531d3a4cbca1a689189fe512fed7d7aeb84de34 599Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 600Date: Wed Feb 17 23:58:53 2021 -0800 601 602 xquartz: Use objc_autoreleasePoolPush / objc_autoreleasePoolPop directly in QuartzBlockHandler 603 604 It violates @autoreleasepool best practices, and this helps collapse quartzCocoa.m into quartz.c 605 606 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 607 (cherry picked from commit 94e4e173486c2a94ddcfd2d0515e1ee6731f6656) 608 609commit 08cf6c90fdc31a563843dceeae3caeb8840951da 610Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 611Date: Wed Feb 17 22:04:55 2021 -0800 612 613 xquartz: Minor code modernization -- @autoreleasepool adoption 614 615 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 616 (cherry picked from commit fba421f700498fa382089df47942df36a2d75ce6) 617 618commit 26e0c59a970229b343d00e6343436cbde2ce6354 619Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 620Date: Wed Feb 17 20:55:06 2021 -0800 621 622 xquartz: Remove some dead code for compatibility with older nibs 623 624 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 625 (cherry picked from commit 318f8a4a8a47a0ce4bbbf4290469e933602c9b30) 626 627commit 2853f38965f521cafb8991737db1c655960acc72 628Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 629Date: Thu Feb 18 16:05:34 2021 -0800 630 631 xquartz: Remove a workaround for AppKit versions older than Lion 632 633 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 634 (cherry picked from commit 72a39dccf99191fbfbb4b399c446fd017d55f24e) 635 636commit 1edc9b980b49e6f9c79f90b7045f9f94c5553247 637Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 638Date: Thu Feb 18 22:26:08 2021 -0800 639 640 xquartz: Apply spell check fixes from master for easier cherry-picking of changes in xquartz 641 642 See also: 23e83724df4809fd7857cc609c33ce7e8d3021a4 643 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 644 645commit 67f25cc1870a587f0e687da9f1281031b6543533 646Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 647Date: Wed Feb 17 15:45:45 2021 -0800 648 649 xquartz: Fix applications menu table background color for dark mode 650 651 Fixes: https://github.com/XQuartz/XQuartz/issues/32 652 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 653 (cherry picked from commit 7e2875035800887f3f41f75cba4299088daf939a) 654 655commit 4028c2ad14a71e6b0c8a892a154f1a50736f2dd5 656Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 657Date: Wed Feb 17 15:16:26 2021 -0800 658 659 xquartz: Apply Xcode 12.4 automatic updates to nibs 660 661 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 662 (cherry picked from commit 85beee9885a8e65960fbbde2de9aa28598b4d6ae) 663 664commit ff1c8e2f72a7dc978187a46e089577b09c009c27 665Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 666Date: Wed Feb 17 13:38:31 2021 -0800 667 668 xquartz: Update the about box copyright to 2021 669 670 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 671 (cherry picked from commit 4e892aa6e132447e43d7bae7e3aca4f1fb172f93) 672 673commit a16df6028458597c1ab722c7fe3464e6dd315b43 674Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 675Date: Mon Feb 1 16:27:38 2021 -0800 676 677 xquartz: Ensure we call into TIS on the main thread 678 679 There is a place where this code was called on the main thread. 680 681 We're using a rather nasty anti-pattern to just call a block inline rather 682 than synchonously calling it on the main thread if we're already on the main 683 thread. This code could use a good overhaul, but I don't have time to rip 684 it apart right now. This will address the immediate issue. 685 686 Fixes: https://github.com/XQuartz/XQuartz/issues/40 687 Fixes: https://github.com/XQuartz/XQuartz/issues/48 688 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 689 (cherry picked from commit c9a3b14c1472632afaff340f73a77a2b961f195a) 690 691commit 2087b7782cde6e6a6c5b4786c2c1f136ee18bb51 692Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 693Date: Wed Jan 27 15:03:46 2021 -0800 694 695 xquartz: Ensure that NSRunAlertPanel() is run on the main thread 696 697 Fixes: https://github.com/XQuartz/XQuartz/issues/30 698 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 699 (cherry picked from commit 520e7a1310ddc25b30bcaa0ea3eeaa6c4d137c8c) 700 701commit 2fe5bf4badb21479511479914def04bf29312e4d 702Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 703Date: Wed Jan 27 13:46:02 2021 -0800 704 705 xquartz: Remove support for older versions of libXplugin 706 707 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 708 (cherry picked from commit 7d0bb7ed061458698ff27856976939b261e73b23) 709 710commit bc1a2a0d830c76b03c5210b4059aea0f0470f095 711Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 712Date: Wed Jan 27 13:43:01 2021 -0800 713 714 xquartz: Remove unused include of AvailabilityMacros.h from various sources 715 716 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 717 (cherry picked from commit 5e7c0762e6fdaefc9b8257c0f875432b1877c49e) 718 719commit 393da8b4390f63e6b634a51f733f1b129626ee1a 720Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 721Date: Wed Jan 27 13:40:48 2021 -0800 722 723 xquartz: Remove support for building for i386 724 725 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 726 (cherry picked from commit 59f22341a8b4cd468d6f37fb17dd7fde347e430b) 727 728commit 43aaa10931b4d3a43c2c575cfd120bea442581ea 729Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 730Date: Wed Jan 27 13:35:17 2021 -0800 731 732 xquartz: Remove support for Mountain Lion and earlier versions of macOS 733 734 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 735 (cherry picked from commit aea15a76593f98205e44f20632178dc384c02d57) 736 737commit fb492686d70fc3166c005c273a587dc7dd4da651 738Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 739Date: Wed Jan 27 13:33:22 2021 -0800 740 741 xquartz: Remove support for Lion and earlier versions of macOS 742 743 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 744 (cherry picked from commit c0b2d3e099a60f55b18ca9c30373000f12d40fe2) 745 746commit 34784415ad7c369b93cee1deeea98e8610f133f2 747Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 748Date: Wed Jan 27 13:29:52 2021 -0800 749 750 xquartz: Remove support for SnowLeopard and earlier versions of macOS 751 752 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 753 (cherry picked from commit cc9cf6f085be6e8264f925a11d67a12ad47a042b) 754 755commit d3f81ecaf99c804728b206155875c52e41f0723b 756Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 757Date: Wed Jan 27 13:09:50 2021 -0800 758 759 xquartz: Remove check for libdispatch now that we don't support pre-SnowLeopard 760 761 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 762 (cherry picked from commit f699aac2ea2cf6a3e2d3bdcb2d8179e103de1d4e) 763 764commit 739c5bd32f5683c1e03c96700f5bb499517b3eaf 765Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 766Date: Wed Jan 27 13:23:18 2021 -0800 767 768 xquartz: Remove support for Leopard and earlier versions of macOS 769 770 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 771 (cherry picked from commit 5ad49102722274f53b9b011082d9e0f202fcd9a4) 772 773commit 2d7eb824996e51875a5bc8b76c159bb52249690c 774Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 775Date: Wed Jan 27 13:18:35 2021 -0800 776 777 xquartz: Remove support for Tiger and earlier versions of macOS 778 779 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 780 (cherry picked from commit 6e6db055f8b517ae9d63351d4c00fd480cb6b54a) 781 782commit 080f9eb76a13252f0fc132192ab99232df0e0d92 783Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 784Date: Wed Jan 27 13:12:34 2021 -0800 785 786 os: Remove support for Tiger and earlier versions of macOS 787 788 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 789 (cherry picked from commit 20b86c4060c1d4fbc099675d5100fe17e4ac0147) 790 791commit be9d2fd871139ec491806c0843af59cbac65fc84 792Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 793Date: Wed Jan 27 13:25:54 2021 -0800 794 795 xquartz: Remove support for Panther and earlier versions of macOS 796 797 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 798 (cherry picked from commit 74aef85bd88a1fdb22d445ad14975232a7bebdd7) 799 800commit d39eb58409a4452353c73d362d847f9262181e10 801Author: Jim DeLaHunt <from.github@jdlh.com> 802Date: Mon Dec 11 00:08:52 2017 -0800 803 804 Fix typo "XQaurtz" in Xquartz.man 805 806 (cherry picked from commit 0e272ac458fbd530787273073473e566b0ac306e) 807 808commit 1f2b73176f25567e28ac227faffd9311d6b65dff 809Author: Christopher Chavez <chrischavez@gmx.us> 810Date: Sat Apr 25 05:09:51 2020 +0000 811 812 XQuartz: recognize F16-F20 and Menu keys 813 814 Signed-off-by: Christopher Chavez <chrischavez@gmx.us> 815 (cherry picked from commit 462beb5338a44390e2fff03096942b035b509830) 816 817commit ecc4ebf5355503268fc04923115b15b8d2e5d338 818Author: Jon Turney <jon.turney@dronecode.org.uk> 819Date: Thu Apr 18 11:37:32 2019 +0100 820 821 xquartz: Add stub ddxInputThread() 822 823 Omitted from 4ad21c32 824 825 (cherry picked from commit f013979507da96377fad0e58b0699d9de051bb39) 826 827commit f5df31c76118816124be323a3caed22f42cdb50c 828Author: Fabrice Fontaine <fontaine.fabrice@gmail.com> 829Date: Thu Dec 10 22:32:59 2020 +0100 830 831 meson.build: KMS support also depends on dri2 832 833 Kernel modesettings support also depends on dri2, see 834 ./hw/xfree86/drivers/modesetting/meson.build 835 836 So update meson.build to reflect the changes made in configure.ac by 837 commit 9c81b8f5b5d7bc987f73e8ef01a81e61205e58ee 838 839 Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com> 840 (cherry picked from commit 5d73a8b59e0d01d1ee4cb484e4b43563cec2eee9) 841 842commit b09f5f42dcc58cf73ae5ca85a65018aaee72e06f 843Author: Mariusz Ceier <mceier+freedesktop@gmail.com> 844Date: Wed Dec 2 12:29:15 2020 +0100 845 846 xwayland: Replace LogMessage with LogMessageVerb 847 848 LogMessage logs only when the XLOG_VERBOSITY is >= 1, but by default 849 XLOG_VERBOSITY is 0. 850 851 Signed-off-by: Mariusz Ceier <mceier+freedesktop@gmail.com> 852 (cherry picked from commit 95539ab37baa1292d1f2daaa552f9741b0364cad) 853 854commit c17872d50215938b1a5d899d50907ced01b28eb9 855Author: Michal Srb <msrb@suse.com> 856Date: Thu Jun 21 13:44:04 2018 +0200 857 858 xkb: Fix heap overflow caused by optimized away min. 859 860 Calling strlen on char[4] that does not need to contain '\0' is wrong and X 861 server may end up running into uninitialized memory. 862 863 In addition GCC 8 is clever enough that it knows that strlen on char[4] can 864 return 0, 1, 2, 3 or cause undefined behavior. With this knowledge it can 865 optimize away the min(..., 4). In reality it can cause the memcpy to be called 866 with bigger size than 4 and overflow the destination buffer. 867 868 Fixes: 83913de25d35 (xkb: Silence some compiler warnings) 869 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/288 870 Signed-off-by: Matt Turner <mattst88@gmail.com> 871 (cherry picked from commit 74627d13c75cab7aa73c0e713feae0934e780ba0) 872 873commit bc111a2e67e16d4e6d4f3196ab86c22c1e278c45 874Author: Matt Turner <mattst88@gmail.com> 875Date: Tue Dec 1 10:55:00 2020 -0500 876 877 xserver 1.20.10 878 879 Signed-off-by: Matt Turner <mattst88@gmail.com> 880 881commit 06d1a032ee491547f7037c3ff042065dc2aeaa99 882Author: Matthieu Herrb <matthieu@herrb.eu> 883Date: Thu Nov 12 19:15:07 2020 +0100 884 885 Check SetMap request length carefully. 886 887 Avoid out of bounds memory accesses on too short request. 888 889 ZDI-CAN 11572 / CVE-2020-14360 890 891 This vulnerability was discovered by: 892 Jan-Niklas Sohn working with Trend Micro Zero Day Initiative 893 894 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 895 (cherry picked from commit 446ff2d3177087b8173fa779fa5b77a2a128988b) 896 897commit 7ccb3b0eabb4658daf0ecb2c78a53609ae2c263b 898Author: Matthieu Herrb <matthieu@herrb.eu> 899Date: Sun Oct 11 17:05:09 2020 +0200 900 901 Fix XkbSetDeviceInfo() and SetDeviceIndicators() heap overflows 902 903 ZDI-CAN 11389 / CVE-2020-25712 904 905 This vulnerability was discovered by: 906 Jan-Niklas Sohn working with Trend Micro Zero Day Initiative 907 908 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 909 (cherry picked from commit 87c64fc5b0db9f62f4e361444f4b60501ebf67b9) 910 911commit 440ed5948ba5818abf5ea5fdc5a9d98514658fd3 912Author: Michel Dänzer <mdaenzer@redhat.com> 913Date: Tue Dec 1 11:25:31 2020 +0100 914 915 present/wnmd: Translate update region to screen space 916 917 The region as passed in is in the source pixmap's coordinate space, so 918 intersecting it with the clipList (which is in screen space) resulted in 919 disappointment. 920 921 Fixes Firefox popups such as the hamburger menu when using the EGL 922 backend. 923 924 v2: 925 * Drop vblank->x/y_off from RegionTranslate call, since they're always 926 0 here (present_wnmd_check_flip rejects flips for x/y_off != 0). 927 928 Reported-by: Robert Mader <robert.mader@posteo.de> 929 Tested-by: Robert Mader <robert.mader@posteo.de> 930 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 931 Tested-by: Joakim Tjernlund <joakim.tjernlund@infinera.com> # v1 932 (cherry picked from commit 466b8b43fb355c6040cee45406860b8b8c04e948) 933 934commit 54f9af1c61bd10e2e65cbb17069e0c9ec1f9a2c5 935Author: Kishore Kadiyala <kishore.kadiyala@intel.com> 936Date: Tue Dec 1 11:13:51 2020 +0100 937 938 modesetting: keep going if a modeset fails on EnterVT 939 940 There was a time when setting a mode on a CRTC would not depend on the 941 associated connector's state. If a mode had been set successfully once, 942 it would mean it would work later on. 943 944 This changed with the introduction of new connectors type that now 945 require a link training sequence (DP, HDMI 2.0), and that means that 946 some events may have happened while the X server was not master that 947 would then prevent the mode from successfully be restored to its 948 previous state. 949 950 This patch relaxes the requirement that all modes should be restored on 951 EnterVT, or the entire X-Server would go down by allowing modesets to 952 fail (with some warnings). If a modeset fails, the CRTC will be 953 disabled, and a RandR event will be sent for the desktop environment to 954 fix the situation as well as possible. 955 956 Additional patches might be needed to make sure that the user would 957 never be left with all screens black in some scenarios. 958 959 v2 (Martin Peres): 960 - whitespace fixes 961 - remove the uevent handling (it is done in a previous patch) 962 - improve the commit message 963 - reduce the size of the patch by not changing lines needlessly 964 - return FALSE if one modeset fails in ignore mode 965 - add comments/todos to explain why we do things 966 - disable the CRTCs that failed the modeset 967 968 Signed-off-by: Kishore Kadiyala <kishore.kadiyala@intel.com> 969 Signed-off-by: Martin Peres <martin.peres@linux.intel.com> 970 Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> 971 Tested-by: Kishore Kadiyala <kishore.kadiyala@intel.com> 972 Closes: #1010 973 (cherry picked from commit efb3abddd49fb75bd6d0e31046fed43d258c93da) 974 975commit bd0f53725b581e4698f5f3ec366a9507bd2556e2 976Author: Martin Peres <martin.peres@linux.intel.com> 977Date: Tue Dec 1 11:10:34 2020 +0100 978 979 modesetting: check the kms state on EnterVT 980 981 Normally, we would receive a uevent coming from Linux's DRM subsystem, 982 which would trigger the check for disappearing/appearing resources. 983 However, this event is not received when X is not master (another VT 984 is selected), and so the userspace / desktop environment would not be 985 notified about the changes that happened while X wasn't master. 986 987 To fix the issue, this patch forces a refresh on EnterVT by splitting 988 the kms-checking code from the uevent handling into its own (exported) 989 function called drmmode_update_kms_state. This function is then called 990 from both the uevent-handling function, and on EnterVT right before 991 restoring the modes. 992 993 Signed-off-by: Martin Peres <martin.peres@linux.intel.com> 994 Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> 995 Acked-by: Kishore Kadiyala <kishore.kadiyala@intel.com> 996 Tested-by: Kishore Kadiyala <kishore.kadiyala@intel.com> 997 (cherry picked from commit 293cf660c95d7ba36510bcc4114d7fd5c5f3801c) 998 999commit 5c400cae1f9817045378966effa6bca91e45aead 1000Author: Olivier Fourdan <ofourdan@redhat.com> 1001Date: Tue Oct 27 16:33:55 2020 +0100 1002 1003 configure: Build hashtable for Xres and glvnd 1004 1005 With autoconf, hashtable support is built along with Xres support. 1006 1007 Yet, glvnd also use it, so when disabling Xres from configure, the 1008 build will fail at link time because hashtable functions are not 1009 available. 1010 1011 Untie the build of hashtable from Xres support, just like meson build 1012 does. 1013 1014 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 1015 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1091 1016 (cherry picked from commit 899cebb76ab7754fea49f7babcd64a7e94052cc8) 1017 1018commit 253569a3d281996d7c84041020e14313756900da 1019Author: Olivier Fourdan <ofourdan@redhat.com> 1020Date: Thu Nov 5 18:35:54 2020 +0100 1021 1022 xwayland: Create an xwl_window for toplevel only 1023 1024 One general assumption in Xwayland is that the xwl_window remains the 1025 same for all the child windows of the toplevel window. 1026 1027 When mapping a new X11 window, ensure_surface_for_window() checks for an 1028 existing xwl_window by using xwl_window_get() which will just check for 1029 the registered xwl_window for the window. 1030 1031 That means that a client mapping a child window of an existing window 1032 with a xwl_window will get another different xwl_window. 1033 1034 If an X11 client issues a Present request on the parent window, hence 1035 placed underneath its child window of the same size, the Wayland 1036 compositor may not send the frame callback event for the parent's 1037 Wayland surface which is reckoned to be not visible, obscured behind 1038 the other Wayland surface for the child X11 window. 1039 1040 That bug affects some games running in wine which may get 1 fps because 1041 the repaint occurs only on timeout with a long interval (as with, e.g. 1042 https://bugs.winehq.org/show_bug.cgi?id=47066) 1043 1044 Fix ensure_surface_for_window() by using xwl_window_from_window() which 1045 will walk the window tree, so that a child window won't get another 1046 xwl_window than its parent. 1047 1048 https://gitlab.freedesktop.org/xorg/xserver/-/issues/1099 1049 See-also: https://bugs.winehq.org/show_bug.cgi?id=47066 1050 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 1051 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 1052 (cherry picked from commit 606ba7fc51e5420646f75c4e672fbe61eb7c7e6e) 1053 1054commit 0811a9ff783c252b3e4d558bf31c6b066b0633e1 1055Author: Olivier Fourdan <ofourdan@redhat.com> 1056Date: Mon Nov 2 15:44:31 2020 +0100 1057 1058 xwayland: non-rootless requires the wl_shell protocol 1059 1060 When running non-rootless, Xwayland requires that the Wayland compositor 1061 supports the wl_shell protocol. 1062 1063 Check for wl_shell protocol support at startup and exit cleanly if 1064 missing rather than segfaulting later in ensure_surface_for_window() 1065 while trying to use wl_shell_get_shell_surface(). 1066 1067 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 1068 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 1069 Reviewed-by: Simon Ser <contact@emersion.fr> 1070 (cherry picked from commit ffd02d9b26bd560849c407a6dd4f5c4d7d2c1736) 1071 1072commit b3ae038c32870a4a1bb42b6e1cf2ebedc8b7bc8d 1073Author: Alex Goins <agoins@nvidia.com> 1074Date: Mon Oct 5 18:19:52 2020 -0500 1075 1076 glamor: Update pixmap's devKind when making it exportable 1077 1078 When making a pixmap exportable, glamor will currently create a temporary 1079 exported pixmap backed by a GBM bo, with the devKind updated to the stride of 1080 the bo. However, when the backing of the exported pixmap is swapped into the 1081 original, the devKind of the original is not updated. 1082 1083 Some GBM bos may get implicitly padded, in which case the devKind of the pixmap 1084 will not match the stride of the backing bo. For example, an 800x600 pixmap will 1085 have a devKind of 3200, but the bo's stride will be 3328. This can cause 1086 corruption with PRIME, when the sink uses the wrong stride to display the shared 1087 pixmap. 1088 1089 This commit changes glamor_make_pixmap_exportable() to update the devKind of the 1090 original pixmap after it swaps exported pixmap's backing into it, keeping 1091 everything consistent. 1092 1093 Fixes issue #1018. 1094 1095 Signed-off-by: Alex Goins <agoins@nvidia.com> 1096 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 1097 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 1098 (cherry picked from commit 7a7e55c5c1d6461a9f4d9a4e0129c1c6e1fd3d66) 1099 1100commit d6c389cb87b220f7005cebf483708267b5f4a1c3 1101Author: Bernhard Übelacker <bernhardu@mailbox.org> 1102Date: Sun Sep 27 18:03:48 2020 +0200 1103 1104 os: Fix instruction pointer written in xorg_backtrace 1105 1106 The address retrieved in "pip.start_ip" is not necessarily the same 1107 address as unw_get_proc_name finds as nearest symbol and returns in "off". 1108 Therefore using "pip.start_ip + off" is not reliable, at least 1109 visible in the binaries from the Debian repository. 1110 1111 Bug-Debian: https://bugs.debian.org/971088 1112 1113 Signed-off-by: Bernhard Übelacker <bernhardu@mailbox.org> 1114 (cherry picked from commit c15dd0ba4893f79f7181e783cb1ba404edca917a) 1115 1116commit c3e4c1a0fd5d4d6015e9e6317b758018317e56d1 1117Author: Michel Dänzer <mdaenzer@redhat.com> 1118Date: Tue Jul 28 18:40:47 2020 +0200 1119 1120 present/wnmd: Execute copies at target_msc-1 already 1121 1122 It always takes one update cycle for the copy to become visible on the 1123 host windowing system, so waiting for the target MSC resulted in 1 cycle 1124 delay. 1125 1126 We re-use the idle list for copies which were executed but need their 1127 completion event sent. 1128 1129 Fixes black seams when resizing the "Builder" sub-window of 1130 1131 GDK_BACKEND=x11 gtk4-demo 1132 1133 on Xwayland (see 1134 https://gitlab.gnome.org/GNOME/mutter/-/issues/1290#note_873557). 1135 1136 Unfortunately, this cannot completely fix the seams with apps which 1137 queue up multiple frames in advance, since there's always at least one 1138 queued frame corresponding to the old window size. But it should at 1139 least help a little in that case as well. 1140 1141 v2: 1142 * Bug fix: Don't update exec_msc in present_wnmd_check_flip_window. 1143 (Roman Gilg) 1144 * Use exec_msc = target_msc - 1 instead of exec_msc--, and add a 1145 comment, for clarity. 1146 v3: 1147 * Drop exec_msc = target_msc again in present_wnmd_execute. 1148 * present_execute_copy should never set vblank->queued in 1149 present_wnmd_execute now, so replace that branch with an assertion. 1150 (Roman Gilg) 1151 1152 Reviewed-by: Roman Gilg <subdiff@gmail.com> 1153 Tested-by: Roman Gilg <subdiff@gmail.com> 1154 (cherry picked from commit 1cccb486d48a5d2e7649836b993805bb65dc09e3) 1155 [Since present_wnmd_event_notify hasn't been split up on the 1.20 1156 branch, it needs to check vblank->flip. Doing the same in 1157 present_wnmd_free_idle_vblanks to be safe, though I'm not sure it's 1158 actually possible to hit non-flips there.] 1159 1160commit 96ef31e0f20fcffb2edfe4cb9510f994c188785f 1161Author: Michel Dänzer <mdaenzer@redhat.com> 1162Date: Tue Jul 28 18:53:45 2020 +0200 1163 1164 present/wnmd: Move up present_wnmd_queue_vblank 1165 1166 Allowing it to be called from more functions than before. No functional 1167 change. 1168 1169 Reviewed-by: Roman Gilg <subdiff@gmail.com> 1170 Tested-by: Roman Gilg <subdiff@gmail.com> 1171 (cherry picked from commit d14ea667feccf085c7d66a7c63f380975e07af66) 1172 1173commit 669e40390c3679b649db33f0aa4ae4cfdd17e2a9 1174Author: Michel Dänzer <mdaenzer@redhat.com> 1175Date: Fri Sep 11 17:00:35 2020 +0200 1176 1177 present: Add present_vblank::exec_msc field 1178 1179 For tracking the MSC when the present can be executed separately from 1180 the target MSC. 1181 1182 Allows removing the requeue field instead, plus more later. 1183 1184 v2: 1185 * Rename wait_msc → exec_msc (Roman Gilg) 1186 * Use exec_msc = target_msc instead of exec_msc++, for clarity. 1187 * Bug fix: Set exec_msc = target_msc also if present_flip returned 1188 false in present_execute. 1189 v3: 1190 * Set exec_msc = target_msc also if present_wnmd_flip returned 1191 false in present_wnmd_execute, for consistency. 1192 v4: 1193 * Specifically check for exec_msc == crtc_msc + 1 in 1194 present_execute_wait/copy, to avoid re-introducing 1195 https://bugs.freedesktop.org/show_bug.cgi?id=94596 . 1196 1197 Reviewed-by: Roman Gilg <subdiff@gmail.com> 1198 Tested-by: Roman Gilg <subdiff@gmail.com> 1199 (cherry picked from commit b0b3159abd8001fa3f6dfc44a288a95a62aa5cf6) 1200 1201commit dae234efdb7bba75744aa1697386df3c0db5348a 1202Author: Michel Dänzer <mdaenzer@redhat.com> 1203Date: Fri Sep 11 17:00:18 2020 +0200 1204 1205 present: Move flip target_msc adjustment out of present_vblank_create 1206 1207 Preparation for different handling between SCMD & WNMD. No functional 1208 change intended. 1209 1210 Reviewed-by: Roman Gilg <subdiff@gmail.com> 1211 Tested-by: Roman Gilg <subdiff@gmail.com> 1212 (cherry picked from commit 4c92dea952f7fed19857904f0f552900257ef4b9) 1213 1214commit 1930ed233fdec5d22e4fc192769a0126faabb3ea 1215Author: Olivier Fourdan <ofourdan@redhat.com> 1216Date: Mon Sep 14 15:39:10 2020 +0200 1217 1218 xwayland: Remove pending stream reference when freeing 1219 1220 The EGLStream backend keeps a queue of pending streams for each Xwayland 1221 window. 1222 1223 However, when this pending queue is freed, the corresponding private 1224 data may not be cleared (typically if the pixmap for this window has 1225 changed before the compositor finished attaching the consumer for the 1226 window's pixmap's original eglstream), leading to a use-after-free and a 1227 crash when trying to use that data as the window pixmap. 1228 1229 Make sure to clear the private data when the pending stream is freed. 1230 1231 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1055 1232 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 1233 Tested-by: Karol Szuster <karolsz9898@gmail.com> 1234 Reviewed-by: Adam Jackson <ajax@redhat.com> 1235 (cherry picked from commit a5f439dcd21b4fda093cb382bb1a758b434a1444) 1236 1237commit 1ac389dda89b2882c80767c91bbe88e01818491c 1238Author: Greg V <greg@unrelenting.technology> 1239Date: Tue Sep 15 17:41:04 2020 +0300 1240 1241 xwayland: use drmGetNodeTypeFromFd for checking if a node is a render one 1242 1243 Major/minor numbers are a.. major (ha) source of pain in FreeBSD porting. 1244 In this case, Xwayland was thinking that /dev/dri/card0 is already a render node, 1245 because the st_rdev on FreeBSD was passing the Linux-style check, 1246 and because of the assumption, acceleration would fail because 1247 various ioctls like AMDGPU_INFO would be denied on the non-render node. 1248 1249 Switch to libdrm's function that already works correctly on all platforms. 1250 1251 Signed-off-by: Greg V <greg@unrelenting.technology> 1252 Reviewed-by: Emmanuel Vadot <manu@FreeBSD.org> 1253 (cherry picked from commit 239ebdc9e447d4f836d0c2aa6068c6064fffb46c) 1254 1255commit d108c2c82cba242ea5998d7c91254d90d3a5db71 1256Author: Olivier Fourdan <ofourdan@redhat.com> 1257Date: Fri Sep 11 11:30:18 2020 +0200 1258 1259 xwayland: Do not discard frame callbacks on allow commits 1260 1261 Currently, when a X11 client (usually the X11 window manager from a 1262 Wayland compositor) changes the value of the X11 property 1263 `_XWAYLAND_ALLOW_COMMITS` from `false` to `true`, all pending frame 1264 callbacks on the window are discarded so that the commit occurs 1265 immediately. 1266 1267 Weston uses that mechanism to prevent the content of the window from 1268 showing before it's ready when mapping the window initially, but 1269 discarding the pending frame callbacks has no effect on the initial 1270 mapping of the X11 window since at that point there cannot be any frame 1271 callback on a surface which hasn't been committed yet anyway. 1272 1273 However, discarding pending frame callbacks can be problematic if we 1274 were to use the same `_XWAYLAND_ALLOW_COMMITS` mechanism to prevent 1275 damages to be posted before the X11 toplevel is updated completely 1276 (including the window decorations from the X11 window manager). 1277 1278 Remove the portion of code discarding the pending frame callback, 1279 Xwayland should always wait for a pending frame callback if there's one 1280 before posting new damages. 1281 1282 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 1283 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.com> 1284 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 1285 https://gitlab.freedesktop.org/xorg/xserver/merge_requests/333 1286 (cherry picked from commit 66da95a172c4941b75ffedcdaa0138c0a48f11fb) 1287 1288commit 174cb91d82d643f78d8ba2b9999312b9d7dea98c 1289Author: Michel Dänzer <mdaenzer@redhat.com> 1290Date: Mon Sep 7 18:40:34 2020 +0200 1291 1292 present/wnmd: Remove dead check from present_wnmd_check_flip 1293 1294 present_wnmd_toplvl_pixmap_window returns a window with the same window 1295 pixmap, so the check could never fail. 1296 1297 Reviewed-by: Roman Gilg <subdiff@gmail.com> 1298 (cherry picked from commit b6b1161fd7ac409156fb69439897bcabdeacf393) 1299 1300commit 51ee6e5ceb0cacc8e2fa225ad5391ffb159e36a0 1301Author: Michel Dänzer <mdaenzer@redhat.com> 1302Date: Mon Sep 7 18:39:17 2020 +0200 1303 1304 xwayland: Check window pixmap in xwl_present_check_flip2 1305 1306 We can only flip if the window pixmap matches that of the toplevel 1307 window. Doing so regardless could cause the toplevel window pixmap to 1308 get destroyed while it was still referenced by the window, resulting in 1309 use-after-free and likely a crash. 1310 1311 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1033 1312 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 1313 Reviewed-by: Roman Gilg <subdiff@gmail.com> 1314 (cherry picked from commit 4c25356d6cd908c5030c70e712076dff318ac00d) 1315 1316commit f4006d795cb199bf7deeb00f542c592db6a5becb 1317Author: Michel Dänzer <mdaenzer@redhat.com> 1318Date: Mon Sep 7 18:35:10 2020 +0200 1319 1320 present/wnmd: Can't use page flipping for windows clipped by children 1321 1322 Noticed this was missing while working on the following fix. 1323 1324 v2: 1325 * Dropped present_wnmd_can_window_flip hunk (that function is never 1326 called, will be cleaned up in a follow-up MR). 1327 1328 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> # v1 1329 Reviewed-by: Roman Gilg <subdiff@gmail.com> 1330 (cherry picked from commit 7ac303c7b1e3b1be79ba3648e217798683e65a99) 1331 1332commit 1e84fda2020b12a83c6b9c6d31ab84f993c6176c 1333Author: Michel Dänzer <mdaenzer@redhat.com> 1334Date: Tue Sep 1 11:24:57 2020 +0200 1335 1336 xfree86: Take second reference for SavedCursor in xf86CursorSetCursor 1337 1338 The same pointer is kept in CurrentCursor as well, therefore two 1339 RefCursor calls are needed. 1340 1341 Fixes use-after-free after switching VTs. 1342 1343 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1067 1344 (cherry picked from commit 919f1f46fc67dae93b2b3f278fcbfc77af34ec58) 1345 1346commit 8c3c8bda2c44fb3d62b954b02b08e3b1771ef5bc 1347Author: Michel Dänzer <mdaenzer@redhat.com> 1348Date: Tue Aug 25 17:26:56 2020 +0200 1349 1350 glamor: Fix glamor_poly_fill_rect_gl xRectangle::width/height handling 1351 1352 (Using GLSL 1.30 or newer) 1353 1354 The width/height members of xRectangle are unsigned, but they were 1355 being interpreted as signed when converting to floating point for the 1356 vertex shader, producing incorrect drawing for values > 32767. 1357 1358 v2: 1359 * Use separate GL_UNSIGNED_SHORT vertex attribute for width/height. 1360 (Eric Anholt) 1361 1362 Reviewed-by: Eric Anholt <eric@anholt.net> 1363 (cherry picked from commit 032af35657aa95c6bbdb74ff8c72e535b9b56cfa) 1364 1365commit b28c882889cbea9be3748d3dee1b21ba6eb90b66 1366Author: Arthur Williams <taaparthur@gmail.com> 1367Date: Fri Sep 25 04:52:57 2020 +0000 1368 1369 include: Increase the number of max. input devices to 256. 1370 1371 Extending the decade old f0124ed93, to increase the number of input 1372 devices from 40 to 256. 40 translates at most 9 MD, while 256 will allow 1373 63 MD. It is an arbitrary number, but people are hitting the current 1374 limit under reasonable conditions. 1375 1376 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=64793 1377 Signed-off-by: Arthur Williams <taaparthur@gmail.com> 1378 1379 (cherry picked from commit fe439596b99db65bbae35eff1ea0b79db167f8d6) 1380 1381commit af4c84ce8855e84c0ad89b929bc972e884f0b8e3 1382Author: Olivier Fourdan <ofourdan@redhat.com> 1383Date: Tue Sep 8 10:03:33 2020 +0200 1384 1385 Revert "linux: Make platform device probe less fragile" 1386 1387 This reverts commit 74b7427c41b4e4104af7abf70a996c086d3d7628. 1388 1389 https://gitlab.freedesktop.org/xorg/xserver/-/issues/1068 1390 1391commit 39cb95e959fab97a7e255dda1a1599b096fb0f7e 1392Author: Olivier Fourdan <ofourdan@redhat.com> 1393Date: Tue Sep 8 10:03:11 2020 +0200 1394 1395 Revert "linux: Fix platform device PCI detection for complex bus topologies" 1396 1397 This reverts commit 5c96eb5f44e62a4cfe835023cde304eb5795b8fd. 1398 1399 https://gitlab.freedesktop.org/xorg/xserver/-/issues/1068 1400 1401commit 4b6fce5975c2f931a0478cf4deeec97529b05eb6 1402Author: Olivier Fourdan <ofourdan@redhat.com> 1403Date: Tue Sep 8 10:01:55 2020 +0200 1404 1405 Revert "linux: Fix platform device probe for DT-based PCI" 1406 1407 This reverts commit 249a12c54a9316b089bd22683c011519348496df. 1408 1409 https://gitlab.freedesktop.org/xorg/xserver/-/issues/1068 1410 1411commit afb77415e1fb862c322754230f63bb70fd596943 1412Author: Matt Turner <mattst88@gmail.com> 1413Date: Tue Aug 25 08:23:42 2020 -0700 1414 1415 xserver 1.20.9 1416 1417 Signed-off-by: Matt Turner <mattst88@gmail.com> 1418 1419commit 705d7213935820d9f56563ee9e17aa9beb365c1e 1420Author: Matthieu Herrb <matthieu@herrb.eu> 1421Date: Tue Aug 18 14:55:01 2020 +0200 1422 1423 Fix XRecordRegisterClients() Integer underflow 1424 1425 CVE-2020-14362 ZDI-CAN-11574 1426 1427 This vulnerability was discovered by: 1428 Jan-Niklas Sohn working with Trend Micro Zero Day Initiative 1429 1430 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 1431 (cherry picked from commit 24acad216aa0fc2ac451c67b2b86db057a032050) 1432 1433commit 5b384e7678c5a155dd8752f018c8292153c1295e 1434Author: Matthieu Herrb <matthieu@herrb.eu> 1435Date: Tue Aug 18 14:52:29 2020 +0200 1436 1437 Fix XkbSelectEvents() integer underflow 1438 1439 CVE-2020-14361 ZDI-CAN 11573 1440 1441 This vulnerability was discovered by: 1442 Jan-Niklas Sohn working with Trend Micro Zero Day Initiative 1443 1444 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 1445 (cherry picked from commit 90304b3c2018a6b8f4a79de86364d2af15cb9ad8) 1446 1447commit eff3f6cdd398bfac040351e99e64baf3bf64fa2e 1448Author: Matthieu Herrb <matthieu@herrb.eu> 1449Date: Tue Aug 18 14:49:04 2020 +0200 1450 1451 Fix XIChangeHierarchy() integer underflow 1452 1453 CVE-2020-14346 / ZDI-CAN-11429 1454 1455 This vulnerability was discovered by: 1456 Jan-Niklas Sohn working with Trend Micro Zero Day Initiative 1457 1458 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 1459 (cherry picked from commit 1e3392b07923987c6c9d09cf75b24f397b59bd5e) 1460 1461commit 1d3a1092c30af660b1366fcd344af745590aa29f 1462Author: Matthieu Herrb <matthieu@herrb.eu> 1463Date: Tue Aug 18 14:46:32 2020 +0200 1464 1465 Correct bounds checking in XkbSetNames() 1466 1467 CVE-2020-14345 / ZDI 11428 1468 1469 This vulnerability was discovered by: 1470 Jan-Niklas Sohn working with Trend Micro Zero Day Initiative 1471 1472 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 1473 (cherry picked from commit 11f22a3bf694d7061d552c99898d843bcdaf0cf1) 1474 1475commit 249a12c54a9316b089bd22683c011519348496df 1476Author: Huacai Chen <chenhc@lemote.com> 1477Date: Sun Jul 5 05:59:58 2020 -0400 1478 1479 linux: Fix platform device probe for DT-based PCI 1480 1481 On a DT-base PCI platform, the sysfs path of vga device is like this: 1482 /sys/devices/platform/bus@10000000/1a000000.pci/pci0000:00/0000:00:11.0/0000:04:00.0. 1483 1484 Then the ID_PATH from udev is platform-1a000000.pci-pci-0000:04:00.0 and 1485 the BusID will be pci-0000:04:00.0, which causes Xorg start fail. This 1486 is because config_udev_odev_setup_attribs() use strstr() to search the 1487 first "pci-" in ID_PATH. To fix this, we implement a strrstr() function 1488 and use it to search the last "pci-" in ID_PATH, which can get a correct 1489 BusID. 1490 1491 (backported from commit 9fbd3e43dd9e13700df96b508c3d97f77e2b9f7e) 1492 1493 Reviewed-by: Dave Airlie <airlied@redhat.com> 1494 Signed-off-by: Huacai Chen <chenhc@lemote.com> 1495 1496commit 5c96eb5f44e62a4cfe835023cde304eb5795b8fd 1497Author: Adam Jackson <ajax@redhat.com> 1498Date: Wed Jun 19 14:23:56 2019 -0400 1499 1500 linux: Fix platform device PCI detection for complex bus topologies 1501 1502 Suppose you're in a Hyper-V guest and are trying to use PCI passthrough. 1503 The ID_PATH that udev will construct for that looks something like 1504 "acpi-VMBUS:00-pci-b8c8:00:00.0", and obviously looking for "pci-" in 1505 the first four characters of that is going to not work. 1506 1507 Instead, strstr. I suppose it's possible you could have _multiple_ PCI 1508 buses in the path, in which case you'd want strrstr, if that were a 1509 thing. 1510 1511 (backported from commit 9acff309434a8029bcce1b22530043459bb71791) 1512 1513 Signed-off-by: Adam Jackson <ajax@redhat.com> 1514 Signed-off-by: Huacai Chen <chenhc@lemote.com> 1515 1516commit 74b7427c41b4e4104af7abf70a996c086d3d7628 1517Author: Adam Jackson <ajax@redhat.com> 1518Date: Tue Sep 18 14:37:51 2018 -0400 1519 1520 linux: Make platform device probe less fragile 1521 1522 At the point where xf86BusProbe runs we haven't yet taken our own VT, 1523 which means we can't perform drm "master" operations on the device. This 1524 is tragic, because we need master to fish the bus id string out of the 1525 kernel, which we can only do after drmSetInterfaceVersion, which for 1526 some reason stores that string on the device not the file handle and 1527 thus needs master access. 1528 1529 Fortunately we know the format of the busid string, and it happens to 1530 almost be the same as the ID_PATH variable from udev. Use that instead 1531 and stop calling drmSetInterfaceVersion. 1532 1533 (backported from commit 0816e8fca6194dfb4cc94c3a7fcb2c7f2a921386) 1534 1535 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 1536 Signed-off-by: Adam Jackson <ajax@redhat.com> 1537 Signed-off-by: Huacai Chen <chenhc@lemote.com> 1538 1539commit 4979ac8f0be6fa2c4a1edd8a527f7d2134d8586a 1540Author: Matthieu Herrb <matthieu@herrb.eu> 1541Date: Sat Jul 25 19:33:50 2020 +0200 1542 1543 fix for ZDI-11426 1544 1545 Avoid leaking un-initalized memory to clients by zeroing the 1546 whole pixmap on initial allocation. 1547 1548 This vulnerability was discovered by: 1549 Jan-Niklas Sohn working with Trend Micro Zero Day Initiative 1550 1551 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 1552 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 1553 (cherry picked from commit a6b2cbe91793ae4967cd21a7103d889248029553) 1554 1555commit 2720b871575504349d9f4dffbc73539f1626bd78 1556Author: Aaron Ma <aaron.ma@canonical.com> 1557Date: Thu Jul 30 11:02:39 2020 +0200 1558 1559 xfree86: add drm modes on non-GTF panels 1560 1561 EDID1.4 replaced GTF Bit with Continuous or Non-Continuous Frequency Display. 1562 1563 Check the "Display Range Limits Descriptor" for GTF support. 1564 If panel doesn't support GTF, then add gtf modes. 1565 1566 Otherwise X will only show the modes in "Detailed Timing Descriptor". 1567 1568 V2: Coding style changes. 1569 V3: Coding style changes, remove unused variate. 1570 V4: remove unused variate. 1571 1572 BugLink: https://gitlab.freedesktop.org/drm/intel/issues/313 1573 Signed-off-by: Aaron Ma <aaron.ma@canonical.com> 1574 Reviewed-by: Adam Jackson <ajax@redhat.com> 1575 (cherry picked from commit 6a79a737e2c0bc730ee693b4ea4a1530c108be4e) 1576 1577commit 7da8e7babee16f7d518cd9ee2a71c950fe2c3c3f 1578Author: Roman Gilg <subdiff@gmail.com> 1579Date: Fri Jul 24 12:21:37 2020 +0200 1580 1581 present: Check valid region in window mode flips 1582 1583 For Pixmap flips to have well defined outcomes the window must be contained by 1584 the valid region if such region was specified. 1585 1586 The valid region is inserted as an argument to the check in window mode. 1587 Setting this argument is missing in screen mode as well but we ignore it for now 1588 and only add it to window mode. 1589 1590 It seems there are none or only very few clients actually making use of valid 1591 regions at the moment. For simplicity we therefore just check if a valid region 1592 was set by the client and in this case do never flip, independently of the 1593 window being contained by the region or not. 1594 1595 Signed-off-by: Roman Gilg <subdiff@gmail.com> 1596 (cherry picked from commit 591916ea9e7a77f68f436b4a541402d9deadfe64) 1597 1598commit 4a65b6617ecc43b754885894f6575fb7dc4bb74d 1599Author: Michel Dänzer <mdaenzer@redhat.com> 1600Date: Thu Jul 23 19:14:28 2020 +0200 1601 1602 xwayland: Handle NULL xwl_seat in xwl_seat_can_emulate_pointer_warp 1603 1604 This can happen e.g. with weston's headless backend. 1605 1606 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 1607 (cherry picked from commit e33453f9111b21e4814d628e6ae00bc7b200f404) 1608 1609commit 10cabe0b978677cfac4bfb405295a16ee8eedb34 1610Author: Michel Dänzer <mdaenzer@redhat.com> 1611Date: Tue Mar 17 12:58:12 2020 +0100 1612 1613 xwayland: Propagate damage x1/y1 coordinates in xwl_present_flip 1614 1615 This couldn't have worked correctly for non-0 x1/y1. 1616 1617 Noticed by inspection. 1618 1619 Reviewed-by: Simon Ser <contact@emersion.fr> 1620 (cherry picked from commits 9141196d3104ab37385c3e385deaa70c002dd184) 1621 (cherry picked fixup from commit 85a6fd11c723888ca093785a3df43066fdca9c33) 1622 1623commit 3b51978b9ca8cdc71508f6db2411255ca6406c3a 1624Author: Alan Coopersmith <alan.coopersmith@oracle.com> 1625Date: Tue Jul 21 18:34:28 2020 +0200 1626 1627 doc: Update URLs in Xserver-DTrace.xml 1628 1629 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 1630 (cherry picked from commit 0006aecba097b437f96a462075494d68bdad24c1) 1631 1632commit 6cbd6a09b9037c6e6074127324978a51442ca375 1633Author: Olivier Fourdan <ofourdan@redhat.com> 1634Date: Tue Jun 2 11:23:46 2020 +0200 1635 1636 xwayland: Use a fixed DPI value for core protocol 1637 1638 The way Xwayland works (like all Wayland clients), it first queries the 1639 Wayland registry, set up all relevant protocols and then initializes its 1640 own structures. 1641 1642 That means Xwayland will get the Wayland outputs from the Wayland 1643 compositor, compute the physical size of the combined outputs and set 1644 the corresponding Xwayland screen properties accordingly. 1645 1646 Then it creates the X11 screen using fbScreenInit() but does so by using 1647 a default DPI value of 96. That value is used to set the physical size 1648 of the X11 screen, hence overriding the value computed from the actual 1649 physical size provided by the Wayland compositor. 1650 1651 As a result, the DPI computed by tools such as xdpyinfo will always be 1652 96 regardless of the actual screen size and resolution. 1653 1654 However, if the Wayland outputs get reconfigured, or new outputs added, 1655 or existing outputs removed, Xwayland will recompute and update the 1656 physical size of the screen, leading to an unexpected change of DPI. 1657 1658 To avoid that discrepancy, use a fixed size DPI (defaults to 96, and can 1659 be set using the standard command lime option "-dpi") and compute a 1660 physical screen size to match that DPI setting. 1661 1662 Note that only affects legacy core protocols, X11 clients can still get 1663 the actual physical output size as reported by the Wayland compositor 1664 using the RandR protocol, which also allows for the size to be 0 if the 1665 size is unknown or meaningless. 1666 1667 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 1668 Reviewed-by: Simon Ser <contact@emersion.fr> 1669 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/731 1670 (cherry picked from commit b0413b6e99c6b5fbc04229ce64ddf1f41b08e63e) 1671 1672commit d4e8c4622890b0cdcfe5f4b9c5608d15ce976901 1673Author: Simon Ser <contact@emersion.fr> 1674Date: Tue Jul 21 18:28:01 2020 +0200 1675 1676 xwayland: only use linux-dmabuf if format/modifier was advertised 1677 1678 Previously, linux-dmabuf was used unconditionally if the buffer had a 1679 modifier. However creating a linux-dmabuf buffer with a format/modifier 1680 which hasn't been advertised will fail. 1681 1682 Change xwl_glamor_gbm_get_wl_buffer_for_pixmap to use linux-dmabuf when 1683 the format/modifier has been advertised only. 1684 1685 Signed-off-by: Simon Ser <contact@emersion.fr> 1686 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1035 1687 Tested-by: Emmanuel Gil Peyrot <linkmauve@linkmauve.fr> 1688 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 1689 (cherry picked from commit c0e13cbf5a56e1fdd1e4ce58ebdefb6d2904e4b3) 1690 1691commit c726ceacc1a39c56d2b054ac5f35798d0c3640d7 1692Author: Martin Weber <martin.weber@secunet.com> 1693Date: Tue Jul 21 18:24:41 2020 +0200 1694 1695 hw/xfree86: Avoid cursor use after free 1696 1697 During a VT-Switch a raw pointer to the shared cursor object 1698 is saved which is then freed (in case of low refcount) by a call to 1699 xf86CursorSetCursor with argument pCurs = NullCursor. 1700 This leads to a dangling pointer which can follow in a use after free. 1701 1702 This fix ensures that there is a shared handle saved for the VT-Switch cycle. 1703 1704 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 1705 (cherry picked from commit 7ae221ad5774756766dc78a73d71f4163ac7b1c6) 1706 1707commit 0679d4660579d0f399b5a9b8140d0c0d3483fa9f 1708Author: Alan Coopersmith <alan.coopersmith@oracle.com> 1709Date: Tue Jul 21 18:14:51 2020 +0200 1710 1711 Update URL's in man pages 1712 1713 Mostly http->https conversions, but also replaces gitweb.fd.o 1714 with gitlab.fd.o, and xquartz.macosforge.org with xquartz.org. 1715 1716 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 1717 (cherry picked from commit a5151f58cf98d1696d60a3577dc50851f159da8a) 1718 1719commit 3059a2e62ae7ef665a537ee0fc2ec24beccf17ae 1720Author: Olivier Fourdan <ofourdan@redhat.com> 1721Date: Wed Jun 3 10:17:13 2020 +0200 1722 1723 xwayland: Disable the MIT-SCREEN-SAVER extension when rootless 1724 1725 Xwayland is just a Wayland client, no X11 screensaver should be 1726 expected to work reliably on Xwayland when running rootless because 1727 Xwayland cannot grab the input devices so it has no way to actually 1728 lock the screen managed by the Wayland compositor. 1729 1730 Turn off the screensaver on Xwayland when running rootless by setting 1731 the screensaver timeout and interval and their default values to zero 1732 and disable the MIT-SCREEN-SAVER extension. 1733 1734 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1051 1735 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 1736 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 1737 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 1738 (cherry picked from commit 5c20e4b834145f590c68dbc98e33c7d3d710001a) 1739 1740commit 23c55ec32973e0a75d723e3f37769dd711c9c59c 1741Author: Michel Dänzer <mdaenzer@redhat.com> 1742Date: Wed Jul 22 18:20:14 2020 +0200 1743 1744 xwayland: Hold a pixmap reference in struct xwl_present_event 1745 1746 In the log of the commit below, I claimed this wasn't necessary on the 1747 1.20 branch, but this turned out to be wrong: It meant that 1748 event->buffer could already be destroyed in xwl_present_free_event, 1749 resulting in use-after-free and likely a crash. 1750 1751 Fixes: 22c0808ac88f "xwayland: Free all remaining events in 1752 xwl_present_cleanup" 1753 1754commit 1179938c179a6e96170275e24adac3325539be65 1755Author: Alex Goins <agoins@nvidia.com> 1756Date: Thu Jul 2 20:12:43 2020 -0500 1757 1758 randr: Check rrPrivKey in RRHasScanoutPixmap() 1759 1760 RRHasScanoutPixmap() is called from xf86CheckHWCursor(), regardless of whether 1761 or not RandR has been initialized. 1762 1763 As mentioned in commit 4226c6d, it's possible that RandR has not been 1764 initialized if the server is configured with Xinerama and there is more than one 1765 X screen. Calling rrGetScrPriv when RandR isn't initialized causes an assertion 1766 failure that aborts the server: 1767 1768 Xorg: ../include/privates.h:121: dixGetPrivateAddr: Assertion 1769 key->initialized' failed. 1770 1771 Just as in commit 4226c6d, fix the problem by checking 1772 dixPrivateKeyRegistered(rrPrivKey) before calling rrGetScrPriv. 1773 1774 Signed-off-by: Alex Goins <agoins@nvidia.com> 1775 Acked-by: Olivier Fourdan <ofourdan@redhat.com> 1776 (cherry picked from commit 8eeff5d7880c6885ee6f206355599f13d739afa7) 1777 1778commit 4912f693e8e3abbd090ee02b4bfe14703c3bad0f 1779Author: Jose Maria Casanova Crespo <jmcasanova@igalia.com> 1780Date: Fri May 15 18:33:11 2020 +0200 1781 1782 modesetting: Fix front_bo leak at drmmode_xf86crtc_resize on XRandR rotation 1783 1784 Since the introduction of "modesetting: Remove unnecessary fb addition from 1785 drmmode_xf86crtc_resize" the fb_id isn't initialited at 1786 drmmode_xf86crtc_resize. 1787 1788 Rotate operation of XRandR uses rotate_bo. So in this case the fb_id 1789 associated to the front_bo is not initialized at drmmode_set_mode_major. 1790 So fd_id remains 0. 1791 1792 As every call to drmmode_xf86crtc_resize allocates a new front_bo we should 1793 destroy unconditionally the old_front_bo if operation success. So we free 1794 the allocated GBM handles. 1795 1796 This avoids crashing xserver with a OOM in the RPI4 1Gb at 4k resolution 1797 after 3 series xrandr rotations from normal to left and vice versa reported at 1798 https://github.com/raspberrypi/firmware/issues/1345 1799 1800 Signed-off-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com> 1801 Reviewed-by: Keith Packard <keithp@keithp.com> 1802 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1024 1803 Fixes: 8774532121 "modesetting: Remove unnecessary fb addition from 1804 drmmode_xf86crtc_resize" 1805 (cherry picked from commit 73480f172aeced074dd9301ae4d97f7d2f3a9a45) 1806 1807commit ccbcf083d5c676311aab77cc837a2539a7278a62 1808Author: Lyude Paul <lyude@redhat.com> 1809Date: Tue Jul 14 18:32:39 2020 -0400 1810 1811 xwayland: Store xwl_tablet_pad in its own private key 1812 1813 When a slave device causes the master virtual pointer device to change 1814 device types, the device's private data pointer 1815 (device->public.devicePrivate) is also changed to match the type of the 1816 slave device. This can be a problem though, as tablet pad devices will 1817 set the device's private data pointer to their own xwl_tablet_pad 1818 struct. This can cause us to dereference the pointer as the wrong type, 1819 and result in a segfault: 1820 1821 Thread 1 "Xwayland" received signal SIGSEGV, Segmentation fault. 1822 wl_proxy_marshal (proxy=0x51, opcode=opcode@entry=0) at src/wayland-client.c:792 1823 792 va_start(ap, opcode); 1824 (gdb) bt 1825 0 wl_proxy_marshal (proxy=0x51, opcode=opcode@entry=0) at 1826 src/wayland-client.c:792 1827 1 0x00005610b27b6c55 in wl_pointer_set_cursor (hotspot_y=0, 1828 hotspot_x=0, surface=0x0, serial=<optimized out>, wl_pointer=<optimized 1829 out>) at /usr/include/wayland-client-protocol.h:4610 1830 2 xwl_seat_set_cursor (xwl_seat=xwl_seat@entry=0x5610b46d5d10) at 1831 xwayland-cursor.c:137 1832 3 0x00005610b27b6ecd in xwl_set_cursor (device=<optimized out>, 1833 screen=<optimized out>, cursor=<optimized out>, x=<optimized out>, 1834 y=<optimized out>) at xwayland-cursor.c:249 1835 4 0x00005610b2800b46 in miPointerUpdateSprite (pDev=0x5610b4501a30) at 1836 mipointer.c:468 1837 5 miPointerUpdateSprite (pDev=0x5610b4501a30) at mipointer.c:410 1838 6 0x00005610b2800e56 in miPointerDisplayCursor (pCursor=0x5610b4b35740, 1839 pScreen=0x5610b3d54410, pDev=0x5610b4501a30) at mipointer.c:206 1840 7 miPointerDisplayCursor (pDev=0x5610b4501a30, pScreen=0x5610b3d54410, 1841 pCursor=0x5610b4b35740) at mipointer.c:194 1842 8 0x00005610b27ed62b in CursorDisplayCursor (pDev=<optimized out>, 1843 pScreen=0x5610b3d54410, pCursor=0x5610b4b35740) at cursor.c:168 1844 9 0x00005610b28773ee in AnimCurDisplayCursor (pDev=0x5610b4501a30, 1845 pScreen=0x5610b3d54410, pCursor=0x5610b4b35740) at animcur.c:197 1846 10 0x00005610b28eb4ca in ChangeToCursor (pDev=0x5610b4501a30, 1847 cursor=0x5610b4b35740) at events.c:938 1848 11 0x00005610b28ec99f in WindowHasNewCursor 1849 (pWin=pWin@entry=0x5610b4b2e0c0) at events.c:3362 1850 12 0x00005610b291102d in ChangeWindowAttributes (pWin=0x5610b4b2e0c0, 1851 vmask=<optimized out>, vlist=vlist@entry=0x5610b4c41dcc, 1852 client=client@entry=0x5610b4b2c900) at window.c:1561 1853 13 0x00005610b28db8e3 in ProcChangeWindowAttributes (client=0x5610b4b2c900) 1854 at dispatch.c:746 1855 14 0x00005610b28e1e5b in Dispatch () at dispatch.c:497 1856 15 0x00005610b28e5f34 in dix_main (argc=16, argv=0x7ffc7a601b68, 1857 envp=<optimized out>) at main.c:276 1858 16 0x00007f8828cde042 in __libc_start_main (main=0x5610b27ae930 <main>, 1859 argc=16, argv=0x7ffc7a601b68, init=<optimized out>, fini=<optimized 1860 out>, rtld_fini=<optimized out>, stack_end=0x7ffc7a601b58) at 1861 ../csu/libc-start.c:308 1862 17 0x00005610b27ae96e in _start () at cursor.c:1064 1863 1864 Simple reproducer in gnome-shell: open up an Xwayland window, press some 1865 tablet buttons, lock and unlock the screen. Repeat if it doesn't crash 1866 the first time. 1867 1868 So, let's fix this by registering our own device-specific private key 1869 for storing a backpointer to xwl_tablet_pad, so that all input devices 1870 have their private data pointers set to their respective xwl_seat. 1871 1872 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 1873 Signed-off-by: Lyude Paul <lyude@redhat.com> 1874 (cherry picked from commit ba0e789b912671c724a21b3a30291247718bcf7d) 1875 1876commit cc36135595fad70ce81697d7c98eb8a26cec9a72 1877Author: SimonP <simonp.git@gmail.com> 1878Date: Tue Jun 9 13:26:48 2020 +0200 1879 1880 xwayland: Initialise values in xwlVidModeGetGamma() 1881 1882 ProcVidModeGetGamma() relies on GetGamma() to initialise values if it 1883 returns TRUE. Without this, we're sending uninitialised values to 1884 clients. 1885 1886 Fixes: xorg/xserver#1040 1887 (cherry picked from commit 6748a4094158d2bde1630b915a5318f9f22c8e0a) 1888 1889commit 533cc6ca046a9e59503bd0763ab28f66284f7dc7 1890Author: Sjoerd Simons <sjoerd@collabora.com> 1891Date: Fri Apr 10 16:34:06 2020 +0200 1892 1893 xwayland: Fix crashes when there is no pointer 1894 1895 When running with a weston session without a pointer device (thus with 1896 the wl_seat not having a pointer) xwayland pointer warping and pointer 1897 confining should simply be ignored to avoid crashes. 1898 1899 Signed-off-by: Sjoerd Simons <sjoerd@collabora.com> 1900 (cherry picked from commit d35f68336b0a462dc660797d1779581f348af04e) 1901 1902commit 3aa31823dbc3dd026ce08958ca5324d7af94a86a 1903Author: Olivier Fourdan <ofourdan@redhat.com> 1904Date: Tue May 5 15:40:43 2020 +0200 1905 1906 xwayland: Clear private on device removal 1907 1908 Xwayland uses the device private to point to the `xwl_seat`. 1909 1910 Device may be removed at any time, including on suspend. 1911 1912 On resume, if the DIX code ends up calling a function that requires the 1913 `xwl_seat` such as `xwl_set_cursor()` we may end up pointing at random 1914 data. 1915 1916 Make sure the clear the device private data on removal so that we don't 1917 try to use it and crash later. 1918 1919 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 1920 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 1921 https://gitlab.freedesktop.org/xorg/xserver/issues/709 1922 (cherry picked from commit 4195e8035645007be313ade79032b8d561ceec6c) 1923 1924commit 22c0808ac88f917541fe40ab30e85364adac82bc 1925Author: Michel Dänzer <mdaenzer@redhat.com> 1926Date: Fri Jun 19 18:10:18 2020 +0200 1927 1928 xwayland: Free all remaining events in xwl_present_cleanup 1929 1930 At the end of xwl_present_cleanup, these events aren't reachable 1931 anymore, so if we don't free them first, they're leaked. 1932 1933 (cherry picked from commit 64565ea344fef0171497952ef75f019cb420fe3b) 1934 1935 v2: 1936 * Simpler backport, no need to keep a reference to the pixmap on the 1937 1.20 branch. 1938 1939commit 37779d7f4028e94c8383a937eab57a6ffb15c38b 1940Author: Michel Dänzer <mdaenzer@redhat.com> 1941Date: Fri Jun 19 18:14:35 2020 +0200 1942 1943 xwayland: Always use xwl_present_free_event for freeing Present events 1944 1945 Minor cleanup, and will make the next change simpler. No functional 1946 change intended. 1947 1948 Reviewed-by: Dave Airlie <airlied@redhat.com> 1949 (cherry picked from commit 1beffba699e2cc3f23039d2177c025bc127966de) 1950 1951commit ba52e5eb0e3b30aa63a8b51e8899eca88d9e0a30 1952Author: Michel Dänzer <mdaenzer@redhat.com> 1953Date: Thu Jun 25 18:11:31 2020 +0200 1954 1955 present/wnmd: Free flip_queue entries in present_wnmd_clear_window_flip 1956 1957 When present_wnmd_clear_window_flip is done, present_destroy_window 1958 frees struct present_window_priv, and the events in the flip queue 1959 become unreachable. So if we don't free them first, they're leaked. 1960 1961 Also drop the call to present_wnmd_set_abort_flip, which just sets a 1962 flag in struct present_window_priv and thus can't have any observable 1963 effect after present_destroy_window. 1964 1965 Closes: https://gitlab.freedesktop.org/xorg/xserver/-/issues/1042 1966 Reviewed-by: Dave Airlie <airlied@redhat.com> 1967 (cherry picked from commit 1bdedc8dbb9d035b85444c2558a137470ff52113) 1968 1969commit b3310ed5036b098509e40b3642c2265132445e6a 1970Author: Michel Dänzer <mdaenzer@redhat.com> 1971Date: Thu Jun 25 18:09:27 2020 +0200 1972 1973 present/wnmd: Keep pixmap pointer in present_wnmd_clear_window_flip 1974 1975 The comment was incorrect: Any reference held by the window (see 1976 present_wnmd_execute) is in addition to the one in struct present_vblank 1977 (see present_vblank_create). So if we don't drop the latter, the pixmap 1978 will be leaked. 1979 1980 Reviewed-by: Dave Airlie <airlied@redhat.com> 1981 (cherry picked from commit bc9dd1c71c3722284ffaa7183f4119151b25a44f) 1982 1983commit fc297c87d6755c11380a44e3510689cc76eb1ee1 1984Author: Simon Ser <contact@emersion.fr> 1985Date: Mon May 4 18:20:17 2020 +0200 1986 1987 xwayland: import DMA-BUFs with GBM_BO_USE_RENDERING only 1988 1989 Drop GBM_BO_USE_SCANOUT from the GBM_BO_IMPORT_FD import, add 1990 GBM_BO_USE_RENDERING to the GBM_BO_IMPORT_FD_MODIFIER import. 1991 1992 If the DMA-BUF cannot be scanned out, gbm_bo_import with 1993 GBM_BO_USE_SCANOUT will fail. However Xwayland doesn't need to scan-out 1994 the buffer and can work fine without scanout. Glamor only needs 1995 GBM_BO_USE_RENDERING. 1996 1997 Signed-off-by: Simon Ser <contact@emersion.fr> 1998 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 1999 Reviewed-by: Daniel Stone <daniels@collabora.com> 2000 (cherry picked from commit 421ce458f1d295015c108eb32f9611e527649cf8) 2001 2002commit 0430d13c1ed8ddbb1d5a57d7e507771f8079d9af 2003Author: Olivier Fourdan <ofourdan@redhat.com> 2004Date: Fri Apr 24 17:45:49 2020 +0200 2005 2006 xwayland: Fix infinite loop at startup 2007 2008 Mutter recently added headless tests, and when running those tests the 2009 Wayland compositor runs for a very short time. 2010 2011 Xwayland is spawned by the Wayland compositor and upon startup will 2012 query the various Wayland protocol supported by the compositor. 2013 2014 To do so, it will do a roundtrip to the Wayland server waiting for 2015 events it expects. 2016 2017 If the Wayland compositor terminates before Xwayland has got the replies 2018 it expects, it will loop indefinitely calling `wl_display_roundtrip()` 2019 continuously. 2020 2021 To avoid that issue, add a new `xwl_screen_roundtrip()` that checks for 2022 the returned value from `wl_display_roundtrip()` and fails if it is 2023 negative. 2024 2025 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 2026 Reviewed-by: Roman Gilg <subdiff@gmail.com> 2027 Reviewed-by: Jonas Ådahl <jadahl@gmail.com> 2028 2029 (cherry picked from commit 785e59060c00129e47da6c0877604a56d7e0e32f) 2030 2031commit b8b10e29302d94832be8f0d8aa7aa1b54e42762c 2032Author: Hans de Goede <hdegoede@redhat.com> 2033Date: Mon Jun 24 21:46:26 2019 +0200 2034 2035 modesetting: Disable pageflipping when using a swcursor 2036 2037 The miPointerSpriteFunc swcursor code expects there to only be a single 2038 framebuffer and when the cursor moves it will undo the damage of the 2039 previous draw, potentially overwriting what ever is there in a new 2040 framebuffer installed after a flip. 2041 2042 This leads to all kind of artifacts, so we need to disable pageflipping 2043 when a swcursor is used. 2044 2045 The code for this has shamelessly been copied from the xf86-video-amdgpu 2046 code. 2047 2048 Fixes: https://gitlab.freedesktop.org/xorg/xserver/issues/828 2049 2050 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 2051 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 2052 (cherry picked from commit 0aaac8d783e78c040a70a55ba8d67809abd7e625) 2053 Signed-off-by: Łukasz Spintzyk <lukasz.spintzyk@displaylink.com> 2054 2055commit 271934db9f3b1297754b29855646e0a4ee01db59 2056Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 2057Date: Tue Oct 30 18:43:51 2018 +0100 2058 2059 dix: do not send focus event when grab actually does not change 2060 2061 c67f2eac5651 ("dix: always send focus event on grab change") made dix 2062 always sent events when it's a NotifyGrab or NotifyUngrab, even if 2063 from == to, because 'from' can just come from a previous XSetInputFocus 2064 call. 2065 2066 However, when an application calls XGrabKeyboard several times on 2067 the same window, we are now sending spurious FocusOut+FocusIn with 2068 NotifyGrab, even if the grab does not actually change. This makes screen 2069 readers for blind people spuriously emit activity events which disturb 2070 screen reading workflow when e.g. switching between menus. 2071 2072 This commit avoids calling DoFocusEvents in that precise case, i.e. when 2073 oldWin is a previous grab and the new grab is the same window. 2074 2075 Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 2076 Reviewed-by: Adam Jackson <ajax@redhat.com> 2077 (cherry picked from commit 364d64981549544213e2bca8de6ff8a5b2b5a69e) 2078 2079commit f84ad082557f9cde6b8faa373eca6a0a89ba7d56 2080Author: Matt Turner <mattst88@gmail.com> 2081Date: Sun Mar 29 13:02:03 2020 -0700 2082 2083 xserver 1.20.8 2084 2085 Signed-off-by: Matt Turner <mattst88@gmail.com> 2086 2087commit 8837279869309317c110afb6f2f3c24484c77657 2088Author: Jon Turney <jon.turney@dronecode.org.uk> 2089Date: Wed Apr 17 11:37:11 2019 +0100 2090 2091 Fix old-style definition warning for xf86OSInputThreadInit() 2092 2093 ../hw/xfree86/os-support/stub/stub_init.c: In function ‘xf86OSInputThreadInit’: 2094 ../hw/xfree86/os-support/stub/stub_init.c:29:1: warning: old-style function definition [-Wold-style-definition] 2095 2096 (cherry picked from commit 7c266cafed14b38c039091651069ae9888c3a8ae) 2097 2098commit 0c012f968b4e02a2bc892ce71f7bea9bd3f7fb22 2099Author: Jon Turney <jon.turney@dronecode.org.uk> 2100Date: Wed Mar 13 14:57:14 2019 +0000 2101 2102 Add xf86OSInputThreadInit to stub os-support as well 2103 2104 stub os support also needs to provide xf86OSInputThreadInit, omitted in 2105 ea1527a8 2106 2107 (cherry picked from commit c020769dbfb965740c8441d8242b738ef572a7c9) 2108 2109commit b259485975078087fe6bde2b9e1eccf4ae14120c 2110Author: Michel Dänzer <mdaenzer@redhat.com> 2111Date: Tue Mar 17 11:45:22 2020 +0100 2112 2113 xwayland: Delete all frame_callback_list nodes in xwl_unrealize_window 2114 2115 We were only calling xwl_present_unrealize_window for the toplevel 2116 window, but the list can contain entries from child windows as well, 2117 in which case we were leaving dangling pointers to freed memory. 2118 2119 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/1000 2120 Fixes: c5067feaeea1 "xwayland: Use single frame callback for Present 2121 flips and normal updates" 2122 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 2123 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 2124 (cherry picked from commit 5e91587302e85fd6f0e8d5ffbe30182e18c6913f) 2125 2126commit a033571644d277dc49a489f7ae32c4ad92856543 2127Author: Jonas Ådahl <jadahl@gmail.com> 2128Date: Fri Sep 13 17:11:27 2019 +0200 2129 2130 xwayland/glamor-gbm: Handle DRM_FORMAT_MOD_INVALID gracefully 2131 2132 The compositor may send DRM_FORMAT_MOD_INVALID instead of a list of 2133 modifiers for various reasons. Handle this gracefully by ignoring it. 2134 2135 Without this, if a compositor would send DRM_FORMAT_MOD_INVALID, it'd 2136 result in empty windows provided by Xwayland. 2137 2138 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 2139 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 2140 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 2141 (cherry picked from commit edf964434eac10ffbe27cc883e3ab95505669aee) 2142 2143commit 3c48bd50ad33f2a533ac76afa38d6e3906ebc28a 2144Author: Arthur Williams <taaparthur@gmail.com> 2145Date: Sun Oct 6 18:55:35 2019 +0000 2146 2147 dix: Check for NULL spriteInfo in GetPairedDevice 2148 2149 There is a race when reseting the XServer that causes spriteInfo to be 2150 NULL in GetPairedDevice resulting a segfault and subsequent crash. The 2151 problem was noticed when opening a connection, creating master devices, 2152 destroying master devices and closing the connection during testing. 2153 2154 Signed-off-by: Arthur Williams <taaparthur@gmail.com> 2155 2156 2157 (cherry picked from commit e693c9657f98c334e9921ca2f8ebf710497c0c6a) 2158 2159commit 1610ef1d6b5ba99da9d1a639f3b65b2e61514a7d 2160Author: David Seifert <soap@gentoo.org> 2161Date: Fri Jan 24 12:49:44 2020 +0100 2162 2163 Fix building with `-fno-common` 2164 2165 * GCC 10 will switch the default to `-fno-common`. 2166 https://gcc.gnu.org/PR85678 2167 2168 Bug: https://bugs.gentoo.org/705880 2169 Signed-off-by: Matt Turner <mattst88@gmail.com> 2170 2171commit 2a185dd22ddb5b0d7d2ef5948591028766bb9530 2172Author: Michel Dänzer <mdaenzer@redhat.com> 2173Date: Mon Mar 2 18:09:31 2020 +0100 2174 2175 xwayland: Use frame callbacks for Present vblank events 2176 2177 Instead of only the fallback timer. 2178 2179 Fixes https://gitlab.freedesktop.org/xorg/xserver/issues/854 2180 2181 v2: 2182 * Drop unused frame_callback member of struct xwl_present_window 2183 (Olivier Fourdan) 2184 2185 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 2186 (cherry picked from commit 9b31358c52e951883bf7c01c953a9da080542244) 2187 2188commit 99a6d6b15e0757a4652a569a1b2070c76a00b567 2189Author: Michel Dänzer <mdaenzer@redhat.com> 2190Date: Wed Nov 27 18:04:06 2019 +0100 2191 2192 xwayland: Use single frame callback for Present flips and normal updates 2193 2194 Using a list of Present windows that need to be called back. 2195 2196 This prepares for the following change, there should be no change in 2197 observed behaviour. 2198 2199 v2: 2200 * Use xwl_window_create_frame_callback instead of making the 2201 frame_listener struct non-static (Olivier Fourdan) 2202 2203 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 2204 (cherry picked from commit c5067feaeea115761f0a72f37407c6e5e943d1a1) 2205 2206commit 915cc107767624bd7914c962347bab9c2e21cdff 2207Author: Michel Dänzer <mdaenzer@redhat.com> 2208Date: Fri Dec 13 18:26:35 2019 +0100 2209 2210 xwayland: Add xwl_window_create_frame_callback helper 2211 2212 This will be used by the following changes. No functional change 2213 intended. 2214 2215 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 2216 (cherry picked from commit f80eea0529b2cfb805a9c7d4994a4235451131e3) 2217 2218commit 94dad4f05133171805ee94095bbcd20ece754eba 2219Author: Dor Askayo <dor.askayo@gmail.com> 2220Date: Wed Feb 19 17:22:11 2020 +0100 2221 2222 xwayland: clear pixmaps after creation in rootless mode 2223 2224 When a pixmap is created with a backing FBO, the FBO should be cleared 2225 to avoid rendering uninitialized memory. This could happen when the 2226 pixmap is rendered without being filled in its entirety. 2227 2228 One example is when a top-level window without a background is 2229 resized. The pixmap would be reallocated to prepare for more pixels, 2230 but uninitialized memory would be rendered in the resize offset until 2231 the client sends a frame that fills these additional pixels. 2232 2233 Another example is when a new top-level window is created without a 2234 background. Uninitialized memory would be rendered after the pixmap is 2235 allocated and before the client sends its first frame. 2236 2237 This issue is only apparent in OpenGL implementations that don't zero 2238 the VRAM of allocated buffers by default, such as RadeonSI. 2239 2240 Signed-off-by: Dor Askayo <dor.askayo@gmail.com> 2241 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/636 2242 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 2243 (cherry picked from commit 0e9a0c203c2ae4eae12bdbb95428f398211c7bee) 2244 2245 [ Michel Dänzer: 2246 * Squashed in commit ebf549db2d9341d99e0d0847b948dd798d98f7dc 2247 * Dropped code related to glamor_format, which only exists on master ] 2248 2249commit 0238359bced17f9db0e266111897d154ab117d68 2250Author: Michel Dänzer <mdaenzer@redhat.com> 2251Date: Fri Feb 7 12:15:07 2020 +0100 2252 2253 xwayland: Call glamor_block_handler from xwl_screen_post_damage 2254 2255 In between the two phases introduced by the previous change. This makes 2256 sure all pending drawing to the new buffers is flushed before they're 2257 committed to the Wayland server. 2258 (cherry picked from commit a542224ea28e2e8ccaf5e0df85bf6c603e97599a) 2259 2260commit a93bce6bfc6c610676a7fbc76639854c5553cb2c 2261Author: Michel Dänzer <mdaenzer@redhat.com> 2262Date: Fri Feb 7 12:06:39 2020 +0100 2263 2264 xwayland: Split up xwl_screen_post_damage into two phases 2265 2266 The first phase sets the new surface properties for all damaged 2267 windows, then the second phase commits all surface updates. 2268 2269 This is preparatory for the next change, there should be no observable 2270 change in behaviour (other than the order of Wayland protocol 2271 requests). 2272 2273 Reviewed-by: Adam Jackson <ajax@redhat.com> 2274 (cherry picked from commit f88d9b1f779835302e02e255fcd45989db7f488d) 2275 2276commit 1ba5e528d52ed9d7d67eb45c5d3e04b6f5d22b05 2277Author: Maarten Lankhorst <dev@mblankhorst.nl> 2278Date: Thu Apr 25 11:46:41 2019 +0200 2279 2280 modesetting: Disable atomic support by default 2281 2282 The atomic driver has issues with modesetting when stealing 2283 connectors from a different crtc, a black screen when doing rotation 2284 on a different crtc, and in general is just a mapping of the legacy 2285 helpers to atomic. This is already done in the kernel, so just 2286 fallback to legacy by default until this is fixed. 2287 2288 Please backport to 1.20, as we don't want to enable it for everyone 2289 there. It breaks for existing users. 2290 2291 The fixes to make the xserver more atomic have been pending on the 2292 mailing list for ages. 2293 2294 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110375 2295 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=110030 2296 References: https://gitlab.freedesktop.org/xorg/xserver/merge_requests/36/commits 2297 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> 2298 (cherry picked from commit f0d78b47ac49977a6007f5fe081f00c6eb19a12e) 2299 2300commit d44bbb4710961651dcf10701bc562f1f01509010 2301Author: Eric Anholt <eric@anholt.net> 2302Date: Mon Feb 10 10:22:34 2020 +0100 2303 2304 glamor: Fix a compiler warning since the recent OOM fixes. 2305 2306 Signed-off-by: Eric Anholt <eric@anholt.net> 2307 (cherry picked from commit 3b26b90cb787a14fa5f8bb2033eab8ab6562a9a5) 2308 2309commit d2a6c8708ca4f27c8d9aade6db2c4e7f2d8c624f 2310Author: Paul Kocialkowski <paul.kocialkowski@bootlin.com> 2311Date: Mon Feb 10 10:21:02 2020 +0100 2312 2313 glamor: Fallback to system memory for RW PBO buffer allocation 2314 2315 We currently support two modes of operation for RW PBO buffers: one 2316 that allocates a pack buffer with GL memory and one that uses system 2317 memory when the former is not supported. 2318 2319 Since allocation with system memory is less likely to fail, add a 2320 fallback to system memory when GL memory failed instead of bailing 2321 out. 2322 2323 Signed-off-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com> 2324 (cherry picked from commit 8c4e8d9eff03cefc987f13c900b0a47403946127) 2325 2326commit ca034c2f2cfff8e49b816b8ecbaa96215b796e36 2327Author: Paul Kocialkowski <paul.kocialkowski@bootlin.com> 2328Date: Mon Feb 10 10:20:30 2020 +0100 2329 2330 glamor: Propagate glamor_prepare_access failures in copy helpers 2331 2332 glamor_prepare_access can fail for a few reasons, especially when 2333 failing to allocate a PBO buffer. Take this in account and bail in 2334 the copy helpers that call the helper when a failure happens. 2335 2336 Signed-off-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com> 2337 (cherry picked from commit de6b3fac1f26075ce915006c914c4a4755617715) 2338 2339commit a7b165d994d74131778a5a9bcffec957f1d1cacb 2340Author: Paul Kocialkowski <paul.kocialkowski@bootlin.com> 2341Date: Mon Feb 10 10:20:04 2020 +0100 2342 2343 glamor: Error out on out-of-memory when allocating PBO for FBO access 2344 2345 Packed buffer allocation (which happens at glBufferData time with the 2346 buffer bound) can fail when there is no GL memory left. 2347 2348 Pick up the error when it happens, print a proper error message, do 2349 some cleanup and bail. 2350 2351 Signed-off-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com> 2352 (cherry picked from commit bc2e12239f86e5a4acd220744f42eb83ba55d328) 2353 2354commit 428b5ce4da99a43bfa0c1933ec447f7feb3639a1 2355Author: Paul Kocialkowski <paul.kocialkowski@bootlin.com> 2356Date: Mon Feb 10 10:19:18 2020 +0100 2357 2358 glamor: Propagate FBO allocation failure for picture to texture upload 2359 2360 When uploading a picture to a texture, glamor_upload_picture_to_texture 2361 calls glamor_pixmap_ensure_fbo to ensure that there is backing FBO. 2362 The FBO will be allocated if the picture's drawable pixmap does not have 2363 one already, which can fail when there is no GL memory left. 2364 2365 glamor_upload_picture_to_texture checks that the call succeeded and will 2366 enter the failure path if it did not. However, unlike many other 2367 functions in glamor, this one has ret set to TRUE initially, so it needs 2368 to be set to FALSE when a failure happens. 2369 2370 Otherwise, the error is not propagated and the failure path return TRUE. 2371 This leads to a fault when trying to access the FBO pointer later on. 2372 2373 Signed-off-by: Paul Kocialkowski <paul.kocialkowski@bootlin.com> 2374 (cherry picked from commit c98c7709c67d8ed6b7455ec700a49b58c396ec2c) 2375 2376commit 948afd768398955f043fef8e14d7d154cea25f85 2377Author: Dave Airlie <airlied@redhat.com> 2378Date: Thu Feb 6 17:59:08 2020 +0100 2379 2380 modesetting: remove unnecessary error message, fix zaphod leases 2381 2382 I introduced this error with the MST hotplug code, but it can trigger 2383 on zaphod setups, and is perfectly fine. There is no support for 2384 MST/hotplug on zaphod setups currently, so we can just skip over 2385 the dynamic connector handling here. However we shouldn't skip 2386 over the lease handling so move it into the codepath. 2387 2388 Fixes: 9257b1252da9 ("modesetting: add dynamic connector hotplug support (MST) (v3)") 2389 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 2390 Signed-off-by: Dave Airlie <airlied@redhat.com> 2391 (cherry picked from commit 1cfdd1a96580733df3625bcea3384ffee3dc92df) 2392 2393commit 1c3e51dabadbf65e7fdedbebbdcd19a85fb03e34 2394Author: Michel Dänzer <mdaenzer@redhat.com> 2395Date: Thu Feb 6 17:57:16 2020 +0100 2396 2397 xfree86/modes: Bail from xf86RotateRedisplay if pScreen->root is NULL 2398 2399 Avoids a crash in xf86RotatePrepare -> DamageRegister during 2400 CreateScreenResources if rotation or another transform is configured for 2401 any connected RandR output in xorg.conf. The generic rotation/transform 2402 code generally can't work without the root window currently. 2403 2404 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/969 2405 Fixes: 094f42cdfe5d "xfree86/modes: Call xf86RotateRedisplay from 2406 xf86CrtcRotate" 2407 Acked-by: Olivier Fourdan <ofourdan@redhat.com> 2408 Reviewed-by: Adam Jackson <ajax@redhat.com> 2409 (cherry picked from commit 6a5e47c57d16de8b6a6a2636f3cbad1aebec32e2) 2410 2411commit af2fd88b1019f63fe3ce871f9e99b3e1e4608b25 2412Author: Daniel Llewellyn <daniel@bowlhat.net> 2413Date: Thu Feb 6 17:56:12 2020 +0100 2414 2415 os: Ignore dying client in ResetCurrentRequest 2416 2417 You might as well, it's harmless. Better, some cleanup code (like DRI2 2418 swap wait) needs to run both normally and at client exit, so it 2419 simplifies the callers to not need to check first. See 4308f5d3 for a 2420 similar example. 2421 2422 Props: @ajax (Adam Jackson) 2423 2424 Fixes: xorg/xserver#211 2425 2426 Signed-off-by: Daniel Llewellyn <diddledan@ubuntu.com> 2427 (cherry picked from commit 578371616e09364318c9fb2371a693d438b31b29) 2428 2429commit e5293f1c5d7b20d98ed4975dc29a6f88c8bc6a0d 2430Author: Adam Jackson <ajax@redhat.com> 2431Date: Thu Feb 6 17:55:25 2020 +0100 2432 2433 Revert "dri2: Don't make reference to noClientException" 2434 2435 It's true that the value would always be -1, if it's not zero, but it's 2436 usually zero is the problem. As a result we return failure from 2437 otherwise successful indirect GLX paths, which isn't very nice of us. 2438 2439 This reverts commit 7d33ab0f8c7958b205076f71e4b47c24aace77fd. 2440 2441 Fixes: https://gitlab.freedesktop.org/xorg/xserver/issues/211 2442 (cherry picked from commit e1fa3beb2fe2519e69f859f0acdc68e5a770de27) 2443 2444commit d845ceae53bb425695e6a185b51ae1b432dd4672 2445Author: George Matsumura <gmmatsumura01@bvsd.org> 2446Date: Thu Feb 6 17:54:36 2020 +0100 2447 2448 Restrict 1x1 pixmap filling optimization to GXcopy 2449 2450 This restricts an optimization whereby the filling of 1x1 pixmaps 2451 went around the driver-provided function to cases where the 2452 source color is meant to be directly copied to the destination, 2453 as opposed to other operations which should produce different 2454 destination values than just the foreground color. 2455 2456 Signed-off-by: George Matsumura <gmmatsumura01@bvsd.org> 2457 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 2458 (cherry picked from commit 83826075e59c0393c16d2a2482dc5c9f2fdf4564) 2459 2460commit c2ef88c4d3a551ff7646bfb86550cae32b02a510 2461Author: Michel Dänzer <mdaenzer@redhat.com> 2462Date: Tue Jan 14 09:23:34 2020 +0100 2463 2464 modesetting: Explicitly #include "mi.h" 2465 2466 For the miClearDrawable prototype. Apparently it doesn't get pulled in 2467 for some build configurations, breaking the build. 2468 2469 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 2470 (cherry picked from commit a24a786fc8490fda08b15c3dab6fa6750f008ecb) 2471 2472commit ad7364d8d7f936b9b08195e47d2f6ee9329ff687 2473Author: Kenneth Graunke <kenneth@whitecape.org> 2474Date: Mon Jan 13 23:34:49 2020 -0800 2475 2476 configure: Define GLAMOR_HAS_EGL_QUERY_DRIVER when available 2477 2478 Commit 195c2ef8f9f07b9bdabc0f554a9033b7857b99c7 added this to the Meson 2479 build but neglected to add it to autotools. 2480 2481 v2: Also update dix-config.h.in 2482 2483 Fixes: 195c2ef8f ("glamor: Add a function to get the driver name via EGL_MESA_query_driver") 2484 2485 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> [v1] 2486 Reviewed-by: Eric Engestrom <eric@engestrom.ch> [v1] 2487 Reviewed-by: Matt Turner <mattst88@gmail.com> 2488 (cherry picked from commit 25ca99df38a2c28c25ab20a917e68442285f2353) 2489 2490commit 489f4191f3c881c6c8acce97ec612167a4ae0f33 2491Author: Matt Turner <mattst88@gmail.com> 2492Date: Mon Jan 13 14:47:48 2020 -0800 2493 2494 xserver 1.20.7 2495 2496 Signed-off-by: Matt Turner <mattst88@gmail.com> 2497 2498commit 279789183ed377127073955d21d44ee3b01ac763 2499Author: Alan Coopersmith <alan.coopersmith@oracle.com> 2500Date: Mon Sep 23 15:12:01 2019 -0700 2501 2502 ospoll: Fix Solaris ports implementation to build on Solaris 11.4 2503 2504 Wrong version got committed, but wasn't noticed since it only builds 2505 with meson, not autoconf. 2506 2507 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 2508 (cherry picked from commit 0e8c0d2f238e5d50daaf4672bd80ad519673b5e3) 2509 2510commit cc503031c32496efb28ed81d32a547ded46a0815 2511Author: Alan Coopersmith <alan.coopersmith@oracle.com> 2512Date: Thu Feb 21 15:38:07 2019 -0800 2513 2514 os-support/solaris: Set IOPL for input thread too 2515 2516 Since the Solaris kernel tracks IOPL per thread, and doesn't inherit 2517 raised IOPL levels when creating a new thread, we need to turn it on 2518 in the input thread for input drivers like vmmouse that need register 2519 access to work correctly. 2520 2521 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 2522 (cherry picked from commit 12769c516d9356bd92f90e2f53a4853dbfdc4aed) 2523 2524commit f778e76eb4bc6b9219a8b1a903a2fb1a30c2c92a 2525Author: Alan Coopersmith <alan.coopersmith@oracle.com> 2526Date: Thu Feb 21 15:35:38 2019 -0800 2527 2528 Add xf86OSInputThreadInit call from common layer into os-support layer 2529 2530 Allows os backends to run additional code as necessary to set up the 2531 input thread. 2532 2533 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 2534 (cherry picked from commit ea1527a8a662dcc5ac3ed49135740aa5f24f74bc) 2535 2536commit e3f26605d85d987da434640f52646d728f1fe919 2537Author: Alan Coopersmith <alan.coopersmith@oracle.com> 2538Date: Thu Feb 21 15:22:57 2019 -0800 2539 2540 Add ddxInputThread call from os layer into ddx layer 2541 2542 Allows ddx's to run additional code as necessary to set up the 2543 input thread. 2544 2545 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 2546 (cherry picked from commit 4ad21c3247d98ac6c5ad71fa36be60ed04f7c92c) 2547 2548commit bb405cdc85b6e31c0beef60a07a2cfe5b87dcde6 2549Author: Alan Coopersmith <alan.coopersmith@oracle.com> 2550Date: Thu Feb 21 14:51:22 2019 -0800 2551 2552 os-support/solaris: Drop ExtendedEnabled global variable 2553 2554 Keeping track of kernel state in user space doesn't buy us anything, 2555 and introduces bugs, as we were keeping global state but the Solaris 2556 kernel tracks IOPL per thread. 2557 2558 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 2559 (cherry picked from commit 7533fa9bd5a4a0f7743d553be186514d684308c8) 2560 2561commit 977f3acfceb91f1dd10c8bf84b9d7515b2a59457 2562Author: Michel Dänzer <mdaenzer@redhat.com> 2563Date: Fri Dec 20 15:51:00 2019 +0100 2564 2565 glamor: Only use dual blending with GLSL >= 1.30 2566 2567 It can't be used with older GLSL. Fixes a crash when attempting to 2568 anyway. 2569 2570 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/97 2571 Fixes: e7308b6c7756 "glamor: Add support for CA rendering in a single pass." 2572 Reviewed-by: Dave Airlie <airlied@redhat.com> 2573 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 2574 (Cherry picked from commit 5bfca0038e92d61e58d4dc1a54748faef8273023) 2575 2576commit cfc5e5040c93486323165196550dbe18aec46402 2577Author: Aaron Plattner <aplattner@nvidia.com> 2578Date: Thu Dec 26 13:40:17 2019 -0800 2579 2580 modesetting: Check whether RandR was initialized before calling rrGetScrPriv 2581 2582 Calling rrGetScrPriv when RandR isn't initialized causes an assertion 2583 failure that aborts the server: 2584 2585 Xorg: ../include/privates.h:121: dixGetPrivateAddr: Assertion `key->initialized' failed. 2586 2587 Thread 1 "Xorg" received signal SIGABRT, Aborted. 2588 0x00007ffff78a8f25 in raise () from /usr/lib/libc.so.6 2589 (gdb) bt 2590 #0 0x00007ffff78a8f25 in raise () from /usr/lib/libc.so.6 2591 #1 0x00007ffff7892897 in abort () from /usr/lib/libc.so.6 2592 #2 0x00007ffff7892767 in __assert_fail_base.cold () from /usr/lib/libc.so.6 2593 #3 0x00007ffff78a1526 in __assert_fail () from /usr/lib/libc.so.6 2594 #4 0x00007ffff7fb57c1 in dixGetPrivateAddr (privates=0x555555ab1b60, key=0x555555855720 <rrPrivKeyRec>) at ../include/privates.h:121 2595 #5 0x00007ffff7fb5822 in dixGetPrivate (privates=0x555555ab1b60, key=0x555555855720 <rrPrivKeyRec>) at ../include/privates.h:136 2596 #6 0x00007ffff7fb586a in dixLookupPrivate (privates=0x555555ab1b60, key=0x555555855720 <rrPrivKeyRec>) at ../include/privates.h:166 2597 #7 0x00007ffff7fb8445 in CreateScreenResources (pScreen=0x555555ab1790) at ../hw/xfree86/drivers/modesetting/driver.c:1335 2598 #8 0x000055555576c5e4 in xf86CrtcCreateScreenResources (screen=0x555555ab1790) at ../hw/xfree86/modes/xf86Crtc.c:744 2599 #9 0x00005555555d8bb6 in dix_main (argc=4, argv=0x7fffffffead8, envp=0x7fffffffeb00) at ../dix/main.c:214 2600 #10 0x00005555557a4f0b in main (argc=4, argv=0x7fffffffead8, envp=0x7fffffffeb00) at ../dix/stubmain.c:34 2601 2602 This can happen, for example, if the server is configured with Xinerama 2603 and there is more than one X screen: 2604 2605 Section "ServerLayout" 2606 Identifier "crash" 2607 Screen 0 "modesetting" 2608 Screen 1 "dummy" RightOf "modesetting" 2609 Option "Xinerama" 2610 EndSection 2611 2612 Section "Device" 2613 Identifier "modesetting" 2614 Driver "modesetting" 2615 EndSection 2616 2617 Section "Screen" 2618 Identifier "modesetting" 2619 Device "modesetting" 2620 EndSection 2621 2622 Section "Device" 2623 Identifier "dummy" 2624 Driver "dummy" 2625 EndSection 2626 2627 Section "Screen" 2628 Identifier "dummy" 2629 Device "dummy" 2630 EndSection 2631 2632 The problem does not reproduce if there is only one X screen because of 2633 this code in xf86RandR12Init: 2634 2635 #ifdef PANORAMIX 2636 /* XXX disable RandR when using Xinerama */ 2637 if (!noPanoramiXExtension) { 2638 if (xf86NumScreens == 1) 2639 noPanoramiXExtension = TRUE; 2640 else 2641 return TRUE; 2642 } 2643 #endif 2644 2645 Fix the problem by checking dixPrivateKeyRegistered(rrPrivKey) before 2646 calling rrGetScrPriv. This is similar to what the xf86-video-amdgpu 2647 driver does: 2648 https://gitlab.freedesktop.org/xorg/driver/xf86-video-amdgpu/blob/fd66f5c0bea2b7c22a47bfd5eb1f22d32d166d9c/src/amdgpu_kms.c#L388 2649 2650 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 2651 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 2652 (cherry picked from commit 4226c6d0329df440551b7b91ae573a82c64a1ac9) 2653 2654commit fb2540648a293def46bf0afa183e0884fa4496be 2655Author: Peter Hutterer <peter.hutterer@who-t.net> 2656Date: Wed May 29 16:19:55 2019 +1000 2657 2658 Xi: return AlreadyGrabbed for key grabs > 255 2659 2660 We can't have high keycodes because everything in XKB relies on 8 bits. XI2's 2661 API allows for 32-bit keycodes so we have to take those but nothing in the 2662 server is really ready for this. The effect of this right now is that any high 2663 keycode grab is clipped to 255 and thus ends up grabbing a different key 2664 instead. 2665 2666 https://bugzilla.redhat.com/show_bug.cgi?id=1697804 2667 2668 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 2669 (cherry picked from commit f4cdbf640b6440df4af784ca35e1b1340965cc10) 2670 2671commit 87ca1bdf6935c4b11b41ca071f1f4b8d8347fee9 2672Author: Michel Dänzer <mdaenzer@redhat.com> 2673Date: Tue Dec 3 18:39:40 2019 +0100 2674 2675 xwayland: Do flush GPU work in xwl_present_flush 2676 2677 The Present code sends the idle notification event to the client after 2678 xwl_present_flush returns. If we don't flush our GPU work here, the 2679 client may race to draw another frame to the same buffer, so we may end 2680 up copying (parts of) that new frame instead of the one we meant to. 2681 2682 Fixes https://gitlab.freedesktop.org/xorg/xserver/issues/835 2683 2684 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 2685 (Cherry picked from commit 2a2234ad1a0fe88400c1511fea67741e4ad09f7f) 2686 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 2687 2688commit 8aad09dd79f0504715edac56e21439809a5d529e 2689Author: Michel Dänzer <mdaenzer@redhat.com> 2690Date: Tue Nov 26 17:17:12 2019 +0100 2691 2692 modesetting: Clear new screen pixmap storage on RandR resize 2693 2694 Fixes random garbage being visible intermittently. 2695 2696 Reviewed-by: Adam Jackson <ajax@redhat.com> 2697 (Cherry picked from commit 9ba13bac9dd076f166ff0d063fc144b904a40d12) 2698 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 2699 2700commit 094f42cdfe5d4c0b8e329445c529ee1e59278999 2701Author: Michel Dänzer <mdaenzer@redhat.com> 2702Date: Tue Nov 26 17:16:37 2019 +0100 2703 2704 xfree86/modes: Call xf86RotateRedisplay from xf86CrtcRotate 2705 2706 If a new rotate buffer was allocated. This makes sure the new buffer 2707 has valid transformed contents when it starts being displayed. 2708 2709 Reviewed-by: Adam Jackson <ajax@redhat.com> 2710 (Cherry picked from commit 327df450ffcf5bda5b4254db0208d355860d1010) 2711 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 2712 2713commit b50175fe867ae067888049318264824524c0616a 2714Author: Michel Dänzer <mdaenzer@redhat.com> 2715Date: Tue Nov 26 17:14:46 2019 +0100 2716 2717 modesetting: Call glamor_finish from drmmode_crtc_set_mode 2718 2719 This makes sure any pending drawing to a new scanout buffer will be 2720 visible from the start. 2721 2722 This makes the finish call in drmmode_copy_fb superfluous, so remove it. 2723 2724 Reviewed-by: Adam Jackson <ajax@redhat.com> 2725 (Cherry picked from commit c66c548eabf06835cb0cb906598fb87c7bb30cf4) 2726 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 2727 2728commit 255d8c3c36a9d52b3acc5e016efa0ef0f2371f79 2729Author: Kenneth Graunke <kenneth@whitecape.org> 2730Date: Thu Nov 21 23:03:50 2019 -0800 2731 2732 modesetting: Use EGL_MESA_query_driver to select DRI driver if possible 2733 2734 We now ask Glamor to use EGL_MESA_query_driver to obtain the DRI driver 2735 name; if successful, we use that as the DRI driver name. Following the 2736 existing dri2.c logic, we also use the same name for the VDPAU driver, 2737 except for i965 (and now iris), where we switch to the "va_gl" fallback. 2738 2739 This allows us to bypass the PCI ID lists in xserver and centralize the 2740 driver selection mechanism inside Mesa. The hope is that we no longer 2741 have to update these lists for any future hardware. 2742 2743 (backported from commit 8d4be7f6c4f7c673d7ec1a6bfdef944907a3916e) 2744 2745 Acked-by: Michel Dänzer <mdaenzer@redhat.com> 2746 2747commit 2a1a96d956f4023796737d26a32354e940c8e6cf 2748Author: Kenneth Graunke <kenneth@whitecape.org> 2749Date: Thu Nov 21 23:01:28 2019 -0800 2750 2751 glamor: Add a function to get the driver name via EGL_MESA_query_driver 2752 2753 This maps to eglGetDisplayDriverName if EGL_MESA_query_render is 2754 supported, otherwise it returns NULL. 2755 2756 (cherry picked from commit 195c2ef8f9f07b9bdabc0f554a9033b7857b99c7) 2757 2758commit 6b3fafa9bfa94b9b04a1a44dc52afb7c4bc250ce 2759Author: Matt Turner <mattst88@gmail.com> 2760Date: Fri Nov 22 17:52:04 2019 -0500 2761 2762 xserver 1.20.6 2763 2764 Signed-off-by: Matt Turner <mattst88@gmail.com> 2765 2766commit 88f12aa74bf4fea25d5b8d8002b3088432feb405 2767Author: Matt Turner <mattst88@gmail.com> 2768Date: Thu Nov 21 11:23:18 2019 -0500 2769 2770 xfree86: Test presence of isastream() 2771 2772 isastream() was never more than a stub in glibc, and was removed in 2773 glibc-2.30 by commit a0a0dc83173c ("Remove obsolete, never-implemented 2774 XSI STREAMS declarations"). 2775 2776 Bug: https://bugs.gentoo.org/700838 2777 Reviewed-by: Julien Cristau <jcristau@debian.org> 2778 Signed-off-by: Matt Turner <mattst88@gmail.com> 2779 (cherry picked from commit e6ab7f9f342f463092c45226f3294074351fdd5e) 2780 2781commit 0e60139064b84b856c59e5a456e26c60710b1b69 2782Author: Olivier Fourdan <ofourdan@redhat.com> 2783Date: Mon Nov 18 17:28:45 2019 +0100 2784 2785 present/wnmd: Relax assertion on CRTC on abort_vblank() 2786 2787 Currently, the function `present_wnmd_abort_vblank()` would fail if the 2788 given `crtc` is NULL. 2789 2790 However, `xwl_present_get_crtc()` can return `NULL` under some 2791 circumstances, which would cause an unexpected termination of Xwayland 2792 in such a case, caused by the assertion failure being triggered. 2793 2794 Remove the assertion, considering that the `crtc` isn't actually used in 2795 neither `present_wnmd_abort_vblank()` nor `xwl_present_abort_vblank()`. 2796 2797 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 2798 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 2799 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/937 2800 (cherry picked from commit 4f984fc06bd57cabfa38f6191f10714878dc8969) 2801 2802commit 2edadf26f1f8deddbe171115fa502337ac62df02 2803Author: Aaron Plattner <aplattner@nvidia.com> 2804Date: Tue Nov 19 10:08:51 2019 -0800 2805 2806 os: Don't crash in AttendClient if the client is gone 2807 2808 If a client is in the process of being closed down, then its client->osPrivate 2809 pointer will be set to NULL by CloseDownConnection. This can cause a crash if 2810 freeing the client's resources results in a call to AttendClient. For example, 2811 if the client has a pending sync fence: 2812 2813 Thread 1 "X" received signal SIGSEGV, Segmentation fault. 2814 AttendClient (client=0x5571c4aed9a0) at ../os/connection.c:942 2815 (gdb) bt 2816 #0 AttendClient (client=0x5571c4aed9a0) at ../os/connection.c:942 2817 #1 0x00005571c3dbb865 in SyncAwaitTriggerFired (pTrigger=<optimized out>) at ../Xext/sync.c:694 2818 #2 0x00005571c3dd5749 in miSyncDestroyFence (pFence=0x5571c5063980) at ../miext/sync/misync.c:120 2819 #3 0x00005571c3dbbc69 in FreeFence (obj=<optimized out>, id=<optimized out>) at ../Xext/sync.c:1909 2820 #4 0x00005571c3d7a01d in doFreeResource (res=0x5571c506e3d0, skip=skip@entry=0) at ../dix/resource.c:880 2821 #5 0x00005571c3d7b1dc in FreeClientResources (client=0x5571c4aed9a0) at ../dix/resource.c:1146 2822 #6 FreeClientResources (client=0x5571c4aed9a0) at ../dix/resource.c:1109 2823 #7 0x00005571c3d5525f in CloseDownClient (client=0x5571c4aed9a0) at ../dix/dispatch.c:3473 2824 #8 0x00005571c3d55eeb in Dispatch () at ../dix/dispatch.c:492 2825 #9 0x00005571c3d59e96 in dix_main (argc=3, argv=0x7ffe7854bc28, envp=<optimized out>) at ../dix/main.c:276 2826 #10 0x00007fea4837cb6b in __libc_start_main (main=0x5571c3d1d060 <main>, argc=3, argv=0x7ffe7854bc28, init=<optimized out>, fini=<optimized out>, rtld_fini=<optimized out>, stack_end=0x7ffe7854bc18) at ../csu/libc-start.c:308 2827 #11 0x00005571c3d1d09a in _start () at ../Xext/sync.c:2378 2828 (gdb) print client->osPrivate 2829 $1 = (void *) 0x0 2830 2831 Since the client is about to be freed, its ignore count doesn't matter and 2832 AttendClient can simply be a no-op. Check for client->clientGone in AttendClient 2833 and remove similar checks from two callers that had them. 2834 2835 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 2836 (cherry picked from commit 4308f5d3d1fbd0f5dce81e22c0c3f08c65a7c9d8) 2837 2838commit 68cfee97bc59580724d594c82f5ee55a980dadf0 2839Author: Adam Jackson <ajax@redhat.com> 2840Date: Wed Oct 9 11:57:18 2019 -0400 2841 2842 dix: Call SourceValidate before GetImage 2843 2844 This ensures that any prep work for the drawable we're about to read 2845 from is already done before we call down to GetImage. This should be no 2846 functional change as most of the callers with a non-trivial 2847 SourceValidate are already wrapping GetImage and doing the equivalent 2848 thing, but we'll be simplifying that shortly. 2849 2850 More importantly this ensures that if any of that prep work would 2851 generate events - like automatic compositing flushing rendering to a 2852 parent pixmap which then triggers damage - then it happens entirely 2853 before we start writing the GetImage reply header. 2854 2855 Note that we do not do the same for GetSpans, but that's okay. The only 2856 way to get to GetSpans is through miCopyArea or miCopyPlane - where the 2857 callers must already call SourceValidate - or miGetImage - which this 2858 commit now protects with SourceValidate. 2859 2860 Fixes: xorg/xserver#902 2861 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 2862 (cherry picked from commit 516e75dbb6bbf97c7af9c2226242ffd9259fec11) 2863 2864commit 2320fd4b52a60411714170d516fc58e6ec516353 2865Author: Adam Jackson <ajax@redhat.com> 2866Date: Wed Oct 9 11:27:12 2019 -0400 2867 2868 mi: Add a default no-op miSourceValidate 2869 2870 Slightly simplifies the callers since they don't need to check for 2871 non-NULL anymore. 2872 2873 I do extremely hate the workarounds here to suppress misprite taking the 2874 cursor down though. Surely there's a better way. 2875 2876 [1.20: Do not in fact simplify the callers as above, since it would 2877 change the ABI. - ajax] 2878 2879 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 2880 (cherry picked from commit ff310903f34ac34c916ad74e919d1ac12aea4715) 2881 2882commit 07efd81b819f21046afc0bb7e19a6cdf7f3c7ea9 2883Author: Olivier Fourdan <ofourdan@redhat.com> 2884Date: Mon Jul 1 13:20:39 2019 +0200 2885 2886 compiler.h: Do not include sys/io.h on ARM with glibc 2887 2888 <sys/io.h> on ARM hasn't worked for a long, long time, so it was removed 2889 it from glibc upstream. 2890 2891 Remove the include to avoid a compilation failure on ARM with glibc. 2892 2893 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 2894 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/840 2895 (cherry picked from commit fe4cd0e7f5c58fa94db36326aadc1bd4e6d73eba) 2896 2897commit 03b91e17ce72649b2d6946213f5a7d152bfe7098 2898Author: Aaron Plattner <aplattner@nvidia.com> 2899Date: Wed Nov 6 10:45:43 2019 -0800 2900 2901 xfree86: Call ScreenInit for protocol screens before GPU screens 2902 2903 During startup, the xfree86 DDX's InitOutput() calls PreInit for 2904 protocol screens first, and then GPU screens. On teardown, dix_main() 2905 calls CloseScreen in the reverse order: GPU screens first starting with 2906 the last one and then working backwards, and then protocol screens also 2907 in reverse order. 2908 2909 InitOutput() calls ScreenInit in the wrong order: for GPU screens first and then 2910 for protocol screens. This causes a problem for drivers that have global state 2911 that is tied to the first screen that calls ScreenInit. 2912 2913 Fix this by simply re-ordering the for loops to call PreInit for 2914 protocol screens first and then for GPU screens second. 2915 2916 (cherry picked from commit e5e9a8ca91d1c087edd8788a110251ad940dbc94) 2917 2918commit 325f694e97a117aec47aa2e1a9b2b28c3c47994e 2919Author: Alex Goins <agoins@nvidia.com> 2920Date: Tue Sep 17 18:36:35 2019 -0500 2921 2922 modesetting: Implement ms_covering_randr_crtc() for ms_present_get_crtc() 2923 2924 ms_present_get_crtc() returns an RRCrtcPtr, but derives it from a xf86CrtcPtr 2925 found via ms_dri2_crtc_covering_drawable()=>ms_covering_crtc(). As a result, it 2926 depends on all associated DIX ScreenRecs having an xf86CrtcConfigPtr DDX 2927 private. 2928 2929 Some DIX ScreenRecs don't have an xf86CrtcConfigPtr DDX private, but do have an 2930 rrScrPrivPtr DDX private. Given that we can derive all of the information we 2931 need from RandR, we can support these screens by avoiding the use of xf86Crtc. 2932 This change implements an RandR-based path for ms_present_get_crtc(), allowing 2933 drawables to successfully fall back to syncing to the primary output, even if 2934 the slave doesn't have an xf86CrtcConfigPtr DDX private. 2935 2936 Without this change, if a slave doesn't have an xf86CrtcConfigPtr DDX private, 2937 drawables will fall back to 1 FPS if they overlap an output on that slave. 2938 2939 Signed-off-by: Alex Goins <agoins@nvidia.com> 2940 (cherry picked from commit 562c7888be538c4d043ec1f374a9d9afa0b305a4) 2941 2942commit aa43dce15f69624ce939bb69bafc3df5a9b78dbb 2943Author: Alex Goins <agoins@nvidia.com> 2944Date: Tue Sep 17 18:03:34 2019 -0500 2945 2946 modesetting: Fix ms_covering_crtc() segfault with non-xf86Crtc slave 2947 2948 DIX ScreenRecs don't necessarily have an xf86CrtcConfigPtr DDX private. 2949 ms_covering_crtc() assumes that they do, which can result in a segfault. 2950 2951 Update ms_covering_crtc() to check the XF86_CRTC_CONFIG_PTR() returned pointer 2952 before dereferencing it. This will still mean that ms_covering_crtc() can't fall 2953 back to the primary output when a drawable overlaps a slave output (going to the 2954 1 FPS default instead), but it won't segfault. 2955 2956 Signed-off-by: Alex Goins <agoins@nvidia.com> 2957 (cherry picked from commit 797e7a0ceb673fe2870c673e248835526a84d000) 2958 2959commit 341a4787157c68597951727d5d11a2813a31ec7e 2960Author: Alex Goins <agoins@nvidia.com> 2961Date: Wed Aug 28 18:24:16 2019 -0500 2962 2963 modesetting: Fix ms_covering_crtc() segfault with non-modesetting slave primary 2964 2965 ms_covering_crtc() uses RRFirstOutput() to determine a primary output to fall 2966 back to if a drawable is overlapping a slave output. 2967 2968 If the primary output is a slave output, RRFirstOutput() will return a slave 2969 output even if passed a master ScreenPtr. ms_covering_crtc() dereferences the 2970 output's devPrivate, which is invalid for non-modesetting outputs, and can 2971 crash. 2972 2973 Changing RRFirstOutput() could have unintended side effects for other callers, 2974 so this change replaces the call to RRFirstOutput() with ms_first_output(). 2975 ms_first_output() ignores the primary output if it doesn't match the given 2976 ScreenPtr, choosing the first connected output instead. 2977 2978 Signed-off-by: Alex Goins <agoins@nvidia.com> 2979 (cherry picked from commit 3ef9029ace4245d9f8929aa71e22bc6a6f40b7b3) 2980 2981commit 2dd9dfc8d9a5dc906ee06d667a756ab267f55e64 2982Author: Alex Goins <agoins@nvidia.com> 2983Date: Wed Sep 4 12:25:31 2019 -0500 2984 2985 randr: Fix RRCrtcDetachScanoutPixmap() segfault during server teardown 2986 2987 During server teardown, mrootdraw is NULL, which can cause segfaults if 2988 master->Stop{,Flipping}PixmapTracking() don't do NULL checking. In this case we 2989 shouldn't need to do master->Stop{,Flipping}PixmapTracking() anyway, so just 2990 skip it. 2991 2992 Signed-off-by: Alex Goins <agoins@nvidia.com> 2993 (cherry picked from commit c82f814313a813d7e1a2d7d3b1f7561810446b34) 2994 2995commit 642c1f83b9ed2452ceb20457a4423a8f52a7a8c4 2996Author: Hans de Goede <hdegoede@redhat.com> 2997Date: Mon Nov 4 20:03:17 2019 +0100 2998 2999 glamor/xwayland: Define EGL_NO_X11 3000 3001 Define EGL_NO_X11 everywhere were we also define MESA_EGL_NO_X11_HEADERS, 3002 EGL_NO_X11 is the MESA_EGL_NO_X11_HEADERS equivalent for the egl headers 3003 shipped with libglvnd. 3004 3005 This fixes the xserver not building with the libglvnd-1.2.0 headers: 3006 3007 In file included from /usr/include/EGL/eglplatform.h:128, 3008 from /usr/include/epoxy/egl_generated.h:11, 3009 from /usr/include/epoxy/egl.h:46, 3010 from glamor_priv.h:43, 3011 from glamor_composite_glyphs.c:25: 3012 /usr/include/X11/Xlib.h:222:2: error: conflicting types for 'GC' 3013 222 | *GC; 3014 | ^~ 3015 In file included from glamor.h:34, 3016 from glamor_priv.h:32, 3017 from glamor_composite_glyphs.c:25: 3018 ../include/gcstruct.h:282:3: note: previous declaration of 'GC' was here 3019 282 | } GC; 3020 | ^~ 3021 3022 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 3023 (cherry picked from commit 741bd73429e337071f49509ddcc5fb392e20b0f6) 3024 3025commit 2f55e8dc270fb65e8ff013c5e846a1a510c761e0 3026Author: Adam Jackson <ajax@redhat.com> 3027Date: Wed Oct 9 16:03:17 2019 -0400 3028 3029 meson: Apparently 1.2 is < 1.2.0 3030 3031 Siiiigh. 3032 3033 (cherry picked from commit 3340ddf37743fdf197dfc16a97107435ff53c6e4) 3034 3035commit e16eb696f2238796827794d2d3b23f268cf4daf0 3036Author: Adam Jackson <ajax@redhat.com> 3037Date: Wed Oct 9 15:17:47 2019 -0400 3038 3039 meson: Fix another reference to "gl" 9.2.0 3040 3041 Previous commit e6ef2b12 missed a spot. 3042 3043 (cherry picked from commit c7486613b0dc586f93f6ba293d9b55c69995b05c) 3044 3045commit a57729d3180ee4bc3ae012cfd03fe5188d858a83 3046Author: Marvin Schmidt <marv@exherbo.org> 3047Date: Tue Sep 24 09:58:43 2019 +0200 3048 3049 build: glx: Lower gl version to work with libglvnd 3050 3051 When using mesa with libglvnd support, mesa will no longer install the 3052 gl, glx, egl pkg-config files but instead let libglvnd provide them. 3053 libglvnd maintainers decided to change the versioning as it was 3054 mesa-specific previously. Now the libraries have versions of the API 3055 they expose[1]. 3056 This causes problems when building the X server: 3057 3058 checking for glproto >= 1.4.17 gl >= 9.2.0... no 3059 configure: error: Package requirements (glproto >= 1.4.17 gl >= 9.2.0) were not met: 3060 3061 Requested 'gl >= 9.2.0' but version of gl is 1.2 3062 3063 Lower the version requirement to 1.2 to allow building against libglvnd 3064 provided libraries 3065 3066 [1] https://github.com/NVIDIA/libglvnd/commit/0dfaea2bcb7cdcc785f95e244223bd004a2d7fba 3067 3068 (cherry picked from commit e6ef2b12404dfec7f23592a3524d2a63d9d25802) 3069 3070commit 3491f1dc5c701656e2f69f74b459dfcaef9451d6 3071Author: Sven Joachim <svenjoac@gmx.de> 3072Date: Wed Oct 2 18:26:07 2019 +0200 3073 3074 modesetting: Fix broken manpage in autoconf build 3075 3076 The autoconf build for the modesetting driver still relied on 3077 xorg-macros.m4 for string replacements and did not include the 3078 top-level manpages.am. As a result, no substitutions took place after 3079 commit 2e497bf887aca832dc0dd30d071c5288ab5c1e15. 3080 3081 This should be a candidate for the 1.20 branch. 3082 3083 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 3084 (cherry picked from commit de0d39f825cce7422986011ee9d5db91f15e2987) 3085 3086commit df1f8f691d7550a18d2076a44c109b1cb6da05ae 3087Author: Alexander Tsoy <alexander@tsoy.me> 3088Date: Mon Sep 23 18:23:40 2019 +0300 3089 3090 configure: Set libdrm flags correctly if only XORG is enabled 3091 3092 This fixes modesetting driver build failure which can be triggered with 3093 the following configure options: 3094 3095 $ ./configure --disable-dri --disable-dri2 --disable-dri3 3096 --disable-config-udev --enable-xorg 3097 3098 Bugzilla: https://bugs.gentoo.org/689768 3099 Signed-off-by: Alexander Tsoy <alexander@tsoy.me> 3100 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 3101 (cherry picked from commit 334f1107146a2e6ebf6f7743b77aec2ad5c46f75) 3102 3103commit fada61e89dd1f20fe8e2e0fb9e4baefb45b4b848 3104Author: Matt Turner <mattst88@gmail.com> 3105Date: Fri Aug 9 20:22:29 2019 -0700 3106 3107 dix: Assert noPanoramiXExtension is false in PanoramiX code 3108 3109 When compiling with link time optimization, GCC thinks it's discovered 3110 undefined behavior: 3111 3112 events.c: In function 'XineramaConfineCursorToWindow': 3113 events.c:609:13: warning: iteration 2147483647 invokes undefined behavior [-Waggressive-loop-optimizations] 3114 events.c:609:11: note: within this loop 3115 events.c:605:49: warning: array subscript -1 is below array bounds of 'struct _Window *[16]' [-Warray-bounds] 3116 events.c:606:31: warning: array subscript -1 is below array bounds of 'struct _Screen *[16]' [-Warray-bounds] 3117 events.c:610:39: warning: array subscript -2 is below array bounds of 'struct _Screen *[16]' [-Warray-bounds] 3118 events.c:617:38: warning: array subscript -2 is below array bounds of 'struct _Window *[16]' [-Warray-bounds] 3119 events.c:619:35: warning: array subscript -2 is below array bounds of 'struct _Screen *[16]' [-Warray-bounds] 3120 3121 This results from 3122 3123 i = PanoramiXNumScreens - 1; 3124 3125 RegionCopy(&pSprite->Reg1, &pSprite->windows[i]->borderSize); 3126 off_x = screenInfo.screens[i]->x; 3127 off_y = screenInfo.screens[i]->y; 3128 3129 where GCC believes that PanoramiXNumScreens might be 0. Unfortunately 3130 GCC is just smart enough to be an annoyance because this case is not 3131 actually possible: XineramaConfineCursorToWindow() is only called when 3132 noPanoramiXExtension is false, and if noPanoramiXExtension is false then 3133 PanoramiXNumScreens must be >1 (see PanoramiXExtensionInit()). 3134 3135 So, add an assert(!noPanoramiXExtension), which to my surprise provides 3136 GCC with information even in release builds and lets GCC understand that 3137 the code is not doing anything that is undefined behavior. 3138 3139 I chose this solution instead of the proposed assert(i >= 0) because the 3140 same pattern occurs in CheckVirtualMotion() but is inside an 3141 'if (!noPanoramiXExtension)' and does not generate any warnings. 3142 3143 Fixes: xorg/xserver#590 3144 Signed-off-by: Matt Turner <mattst88@gmail.com> 3145 (cherry picked from commit 61aa40aeb3d4efefda47f245ed4b83a1a19b1d4c) 3146 3147commit e815db543450c38d1d4ec03184b8e82fa0a37498 3148Author: Alexander Volkov <a.volkov@rusbitech.ru> 3149Date: Mon Feb 11 18:54:10 2019 +0300 3150 3151 shm: Use memfd_create when possible 3152 3153 It doesn't require shared memory dir and thus allows 3154 to avoid cases when this dir is detected incorrectly, 3155 as in https://bugreports.qt.io/browse/QTBUG-71440 3156 3157 Signed-off-by: Alexander Volkov <a.volkov@rusbitech.ru> 3158 (cherry picked from commit f6753c117ef0f83499d5e2d6dda226fec9ddf803) 3159 3160commit 05c5b970136f0a11fffa16692744bdf5a23963e8 3161Author: Eric Anholt <eric@anholt.net> 3162Date: Wed Sep 19 13:28:06 2018 -0700 3163 3164 shm: Pick the shm dir at run time, not build time. 3165 3166 Prodding the builder's filesystem for tmp dirs doesn't necessarily 3167 tell you anything about what the actual host's filesystem is going to 3168 look like, so we should just try the dirs at runtime. 3169 3170 Signed-off-by: Eric Anholt <eric@anholt.net> 3171 (cherry picked from commit 19f6cb570becbc4e355807199c6e251fc7935132) 3172 3173commit 9ee4e35bfe084fa942c2b2ce46405e048eb335a8 3174Author: Eric Anholt <eric@anholt.net> 3175Date: Wed Sep 19 13:20:12 2018 -0700 3176 3177 shm: reindent shm_tmpfile to follow our standards. 3178 3179 Signed-off-by: Eric Anholt <eric@anholt.net> 3180 (cherry picked from commit 804a9b4f57107fa2d0ed1ae0becda5bebaffe6e1) 3181 3182commit b6ee04e774af1fc2e2560583c95dfac688494c2a 3183Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 3184Date: Mon Jul 1 02:31:02 2019 +0200 3185 3186 Fix crash on XkbSetMap 3187 3188 Some devices may not have keyboard information. 3189 3190 Fixes #574 3191 3192 (cherry picked from commit 8469bfead9515ab3644f1769a1ff51466ba8ffee) 3193 3194commit 616b3e1dd19d51688e4cba380394bd93c62b50d4 3195Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 3196Date: Mon Jul 1 02:33:26 2019 +0200 3197 3198 Fix crash on XkbSetMap 3199 3200 Since group_info and width are used for the key actions allocations, 3201 when modifying them we need to take care of reallocation key actions if 3202 needed. 3203 3204 (cherry picked from commit fabc4219622f3c0b41b1cb897c46e092377059e3) 3205 3206commit c747dbb250c11c6a02641b253e7322145dc9b913 3207Author: Michel Dänzer <mdaenzer@redhat.com> 3208Date: Mon Sep 23 18:35:41 2019 +0200 3209 3210 miext/sync: Make struct _SyncObject::initialized fully ABI compatible 3211 3212 With a 32-bit build, putting the initialized field at the end of the 3213 struct bumped the struct size from 20 bytes to 24, changing the layout 3214 of other structs embedding struct _SyncObject. While this would be 3215 acceptable on master, it caused crashes with 1.20. 3216 3217 Making the initialized field a char and putting it in the hole before 3218 the beingDestroyed field restores the 32-bit ABI as well. 3219 3220 Fixes https://gitlab.freedesktop.org/xorg/xserver/issues/892 3221 3222 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 3223 Reviewed-by: Alex Goins <agoins@nvidia.com> 3224 (cherry picked from commit 0d8cc7bb180468fee72330f40d87f6e9db7b42c0) 3225 3226commit 5aadaac9499c71ebf88c0e5fc658d9d56c8a7e16 3227Author: Andres Rodriguez <andresx7@gmail.com> 3228Date: Fri Sep 6 18:32:44 2019 -0400 3229 3230 xf86: Disable unused crtc functions when a lease is revoked 3231 3232 This fixes 'non-desktop' displays staying powered on after their lease 3233 has been revoked. 3234 3235 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111620 3236 Cc: Keith Packard <keithp@keithp.com> 3237 Signed-off-by: Andres Rodriguez <andresx7@gmail.com> 3238 3239commit 63346c74393e1df4555f84367529802a67578ef6 3240Author: Carlos Garnacho <carlosg@gnome.org> 3241Date: Mon Jan 7 15:33:35 2019 +0100 3242 3243 xwayland: Handle the case of windows being realized before redirection 3244 3245 If Xwayland gets to realize a window meant for composition before the 3246 compositor redirected windows (i.e. redirect mode is not RedirectDrawManual 3247 yet), the window would stay "invisible" as we wouldn't create a 3248 wl_surface/wl_shell_surface for it at any later point. 3249 3250 This scenario may happen if the wayland compositor sets up a X11 socket 3251 upfront, but waits to raise Xwayland until there are X11 clients. In this 3252 case the first data on the socket is the client's, the compositor can hardly 3253 beat that in order to redirect subwindows before the client realizes a 3254 Window. 3255 3256 In order to jump across this hurdle, allow the late creation of a matching 3257 (shell) surface for the WindowPtr on SetWindowPixmapProc, so it is ensured 3258 to be created after the compositor set up redirection. 3259 3260 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 3261 Reviewed-by: Michel Dänzer <mdaenzer@redhat.com> 3262 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 3263 (cherry picked from commit 78cc8b6f9613fc71f6ecc7e8848d54364a250634) 3264 3265commit 12db645c7fc0539752a881df7ac2bcd09e3cb17b 3266Author: Carlos Garnacho <carlosg@gnome.org> 3267Date: Mon Jan 7 15:33:31 2019 +0100 3268 3269 xwayland: Refactor surface creation into a separate function 3270 3271 This is just called from xwl_window_realize() ATM, but will be useful in 3272 future commits. 3273 3274 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 3275 (cherry picked from commit c2e8ae964052944312c5023ca7ea5c41a92990e5) 3276 3277commit e0af09061f9e8397ca564ec3bbedea51974455d4 3278Author: Carlos Garnacho <carlosg@gnome.org> 3279Date: Mon Jan 7 15:20:05 2019 +0100 3280 3281 xwayland: Separate DamagePtr into separate window data 3282 3283 This will be dissociated in future commits to handle the cases 3284 where windows are being realized before there is a compositor 3285 handling redirection. 3286 3287 In that case, we still want the DamagePtr to be registered upfront 3288 on RealizeWindowProc before a corresponding xwl_window might be 3289 created. Most notably, it cannot be lazily created on 3290 SetWindowPixmapProc as damage accounting gets broken. 3291 3292 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 3293 (cherry picked from commit 4e50440ae20c537d6a4edf356cda67dd33d4e5a8) 3294 3295commit 045add84927051a33569ed701097e1fd514bf0ca 3296Author: Adam Jackson <ajax@redhat.com> 3297Date: Fri Aug 16 14:56:19 2019 -0400 3298 3299 glx: Fix previous context validation in xorgGlxMakeCurrent 3300 3301 vnd has already verified that the context tag is valid before this gets 3302 called, and we only set the context tag private data to non-null for 3303 indirect clients. Mesa happens to be buggy and doesn't send MakeCurrent 3304 requests nearly as much as it should for direct contexts, but if you fix 3305 that, then unbinding a direct context would fail here with 3306 GLXBadContextTag. 3307 3308 Sadly Mesa will still need to carry a workaround here for broken 3309 servers, but we should still fix the server. 3310 3311 (cherry picked from commit 95dcc81cb122e5a4c5b38e84ef46eb872b2e1431) 3312 3313commit 8449c8623d6208b77e76596e8b10250817d3e8ab 3314Author: Adam Jackson <ajax@redhat.com> 3315Date: Mon Aug 19 14:27:54 2019 -0400 3316 3317 miext/sync: Fix needless ABI change 3318 3319 The initialized field was added in: 3320 3321 commit 82f01ad7869e3f2be51e41a8246dab5982bbc36a 3322 Author: Alex Goins <agoins@nvidia.com> 3323 Date: Wed Apr 10 13:48:02 2019 -0500 3324 3325 xsync: Add resource inside of SyncCreate, export SyncCreate 3326 3327 But it added this field not at the end of SyncObject. It may not have 3328 been _usefully_ possible to create those from another extension prior to 3329 that commit, but that's still an ABI-incompatible change. 3330 3331 (cherry picked from commit 194ba38728b34a76885275dae153057c8afaf72e) 3332 3333commit 39b3005c329bc63676df72c43529d641bf305bcd 3334Author: Aaron Plattner <aplattner@nvidia.com> 3335Date: Tue May 21 10:50:42 2019 -0700 3336 3337 GLX: Set GlxServerExports::{major,minor}Version 3338 3339 Commit 56c0a71fdd94a008e5d746261f70a713c4767f93 incremented the 3340 GLXSERVER_VENDOR_ABI_MINOR_VERSION define, but this define was not actually 3341 being used to set glxServer.minorVersion. 3342 3343 Update the initializer for glxServer to use the correct version numbers. 3344 3345 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 3346 (cherry picked from commit b4231d69028adc8123801a7552b40a15ea928d1b) 3347 3348commit d3034ef2f5121d85ae766a73fda4e523399043a9 3349Author: Kyle Brenneman <kbrenneman@nvidia.com> 3350Date: Thu May 2 07:17:21 2019 -0600 3351 3352 GLX: Add a function to change a clients vendor list. 3353 3354 Add a new function, GlxServerExports::setClientScreenVendor, which will change 3355 the vendor that handles GLX requests for a screen, but only for requests from 3356 a specific client. 3357 3358 v2: Increment the GLXVND minor version number. 3359 v3: Note the GLXVND version requirement for setClientScreenVendor. 3360 3361 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 3362 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 3363 Reviewed-by: Adam Jackson <ajax@redhat.com> 3364 (cherry picked from commit 56c0a71fdd94a008e5d746261f70a713c4767f93) 3365 3366commit abeae4a6d356653d50026707ecc2afceac83631e 3367Author: Kyle Brenneman <kbrenneman@nvidia.com> 3368Date: Wed May 8 08:44:54 2019 -0600 3369 3370 GLX: Use the sending client for looking up XID's 3371 3372 When GlxGetXIDMap looks up an unknown XID, it will now look up a vendor based 3373 on the screen number for the XID and the client that sent the current request. 3374 3375 In GlxGetXIDMap, if the XID is for a regular X window, then it won't be in the 3376 (XID -> vendor) mapping, so we have to look up a vendor by screen number. 3377 3378 With this change, GlxGetXIDMap will use the (screen -> vendor) map for 3379 whichever client sent the current request, instead of using the global 3380 (screen -> vendor) map. 3381 3382 Since GlxGetXIDMap doesn't take a ClientPtr argument, GlxDispatchRequest will 3383 store the client for the current request in a global variable. That way, the 3384 ABI for GLXVND doesn't need to change. 3385 3386 v2: Fix an error check in GlxDispatchRequest. 3387 3388 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 3389 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 3390 Reviewed-by: Adam Jackson <ajax@redhat.com> 3391 (cherry picked from commit 8b67ec7cc6fda243480a5a8ca118b66242f3eb2c) 3392 3393commit 1fdb7cbce538f0b37304a3cfc9fae4ff2fe9ece9 3394Author: Kyle Brenneman <kbrenneman@nvidia.com> 3395Date: Thu Oct 19 15:14:51 2017 -0600 3396 3397 GLX: Add a per-client vendor mapping. 3398 3399 Each client now has its own (screen, vendor) mapping. 3400 3401 Currently, it's just a copy of the global mapping, but later changes will allow 3402 it to change. 3403 3404 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 3405 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 3406 Reviewed-by: Adam Jackson <ajax@redhat.com> 3407 (cherry picked from commit 37a36a6b5b887d5c5a17a6931ceba8ad5d1bb6d5) 3408 3409commit 82f01ad7869e3f2be51e41a8246dab5982bbc36a 3410Author: Alex Goins <agoins@nvidia.com> 3411Date: Wed Apr 10 13:48:02 2019 -0500 3412 3413 xsync: Add resource inside of SyncCreate, export SyncCreate 3414 3415 As shown by DRI3 adding the SyncCreateFenceFromFD() function, extensions may 3416 want to create a fence, then initialize it in their own way. This currently 3417 can't be done without adding a function directly to Xext/sync.c due to the fact 3418 that the RTFence resource type is private and there is no external interface to 3419 add to it. 3420 3421 To facilitate other X extensions creating fences and initializing them, this 3422 change exports SyncCreate() and adds the resource directly within it. Callers no 3423 longer need to call AddResource() after SyncCreate(), they only need to 3424 initialize the SyncObject. 3425 3426 To prevent FreeFence() and FreeCounter() from segfaulting if the call to 3427 AddResource() fails before the sync object is initialized, this adds a new 3428 'initialized' parameter to SyncObject that, when FALSE, causes FreeFence() and 3429 FreeCounter() to skip de-initialization and simply free the object. 3430 Initialization after adding the resource shouldn't otherwise be a problem due to 3431 the single-threaded nature of X. 3432 3433 Signed-off-by: Alex Goins <agoins@nvidia.com> 3434 Reviewed-by: James Jones <jajones@nvidia.com> 3435 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 3436 (cherry picked from commit 7f962c70b6d9c346477f23f6c15211e749110078) 3437 3438commit dfd51be3ca2a244bbca27a95310b60e0c14940df 3439Author: Olivier Fourdan <ofourdan@redhat.com> 3440Date: Tue Jul 23 11:01:47 2019 +0200 3441 3442 xwayland: Do not free a NULL GBM bo 3443 3444 Both `gbm_bo_create()` and `gbm_bo_create_with_modifiers()` can fail and 3445 return `NULL`. 3446 3447 If that occurs, `xwl_glamor_gbm_create_pixmap()` will not create a 3448 pixmap for the (NULL) GBM bo, but would still try to free the bo which 3449 leads to a crash in mesa: 3450 3451 [...] 3452 #7 <signal handler called> 3453 #8 in gbm_bo_destroy (bo=0x0) at ../src/gbm/main/gbm.c:439 3454 #9 in xwl_glamor_gbm_create_pixmap () at xwayland-glamor-gbm.c:245 3455 #10 in ProcCreatePixmap () at dispatch.c:1440 3456 #11 in Dispatch () at dispatch.c:478 3457 #12 in dix_main () at main.c:276 3458 3459 To avoid the crash, only free the GBM bo if not `NULL`. 3460 3461 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 3462 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 3463 Bugzilla: https://bugzilla.redhat.com/1729925 3464 (cherry picked from commit d9ec525059dbe96fc893c73c0362be2a6dd73e85) 3465 3466commit df7ee10d9812897b297c67fde29f0b134ffba3ec 3467Author: Adam Jackson <ajax@redhat.com> 3468Date: Mon Jul 15 11:38:44 2019 -0400 3469 3470 xwayland: Expand the RANDR screen size limits 3471 3472 There's not really a good way to query this from the wayland server, so 3473 just set the maximum to the X11 protocol limits. While we're at it, 3474 lower the minimum screen size to something implausibly small too, just 3475 in case. 3476 3477 Fixes: xorg/xserver#850 3478 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 3479 (cherry picked from commit d0850241c6218f61127c45c2f95d6e791c3fea44) 3480 3481commit 44c693f45d6abd6f7f3bd2f756d35811db143af7 3482Author: Olivier Fourdan <ofourdan@redhat.com> 3483Date: Fri Jun 28 16:55:11 2019 +0200 3484 3485 xwayland: Update screen pixmap on output resize 3486 3487 Running Xwayland non-rootless and resizing the output would lead to a 3488 crash while trying to update the larger areas of the root window. 3489 3490 Make sure we resize the backing pixmap according to the new output size 3491 to avoid the crash. 3492 3493 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 3494 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/834 3495 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 3496 (cherry picked from commit ce9455b5ee389b100a9b7da76b79690d97211b7a) 3497 3498commit 1804e7327113d128ac9a759da55fd3de0b9f8c35 3499Author: Matt Roper <matthew.d.roper@intel.com> 3500Date: Wed Jun 26 09:34:46 2019 -0700 3501 3502 dri2: Sync i965_pci_ids.h from mesa 3503 3504 Copied from Mesa with no modifications. 3505 3506 This update brings in a significant number of new platform ID's. 3507 3508 Syncs with mesa up to commit e334a595e ("intel/icl: Add new ICL 3509 PCI-IDs"). 3510 3511 Signed-off-by: Matt Roper <matthew.d.roper@intel.com> 3512 (cherry picked from commit a8d9ebeb43583e8d8ef182ac15adbdce256d4867) 3513 3514commit 98e3db2ac43d4a3f13475cb160c8ce0155ac7d61 3515Author: Marco Trevisan (Treviño) <mail@3v1n0.net> 3516Date: Fri Oct 26 19:52:49 2018 +0200 3517 3518 Xi: Use current device active grab to deliver touch events if any 3519 3520 When Retrieving touch delivery data we need to check if we have an active 3521 grab on such device, and in that case use it to delivery events. 3522 If we don't do this, when rejecting the touch events in DeactivatePointerGrab, 3523 we will end-up in creating an implicit grab that will change the device 3524 deviceGrab's state, causing a recursion during TouchEndTouch. 3525 3526 Fixes #7 3527 3528 https://bugs.freedesktop.org/show_bug.cgi?id=96536 3529 (cherry picked from commit 35e5a76cc1d02801fadd49d12e60664b02e4bebc) 3530 3531commit 26b6ba5347953ed5f60b7fb9454b45754b145c6f 3532Author: Michel Dänzer <michel.daenzer@amd.com> 3533Date: Wed Jul 3 11:53:56 2019 +0200 3534 3535 Revert "present/scmd: Check that the flip and screen pixmap pitches match" 3536 3537 This reverts commit ef91da2757050652c724f6e674e8b1acf5d0cb31. 3538 3539 I thought this couldn't work under any circumstances, but I was wrong, 3540 and drivers are already checking for cases that really cannot work. 3541 3542 Fixes issue #839. 3543 3544 (cherry picked from commit 2f0dd1816d556eef8c4b691330d39f13e1ed93f0) 3545 3546commit c86222d4bd94892f3bf3c5947c19793ca18bd9e2 3547Author: Carlos Garnacho <carlosg@gnome.org> 3548Date: Wed May 22 17:51:04 2019 +0200 3549 3550 xwayland: Reset scheduled frames after hiding tablet cursor 3551 3552 Hiding the tablet tool cursor results in it being hidden forever after. 3553 This is due to the stale frame callback that will neither be disposed 3554 or replaced. This can be reproduced in krita (X11) as the pointer 3555 cursor is hidden while over the canvas. 3556 3557 Clearing the frame callback ensures the correct behavior in future 3558 xwl_tablet_tool_set_cursor() calls (i.e. a new cursor surface being 3559 displayed, and a new frame callback created), and is 1:1 3560 with xwl_seat_set_cursor() for pointers. 3561 3562 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 3563 (cherry picked from commit dea4a74621294391ce5901bb3339e1b8e7151efc) 3564 3565commit 6e199e4613e149cd5d5ce69cdd6a259744b6cb44 3566Author: Olivier Fourdan <ofourdan@redhat.com> 3567Date: Fri Apr 26 13:37:09 2019 +0200 3568 3569 glamor: Make pixmap exportable from `gbm_bo_from_pixmap()` 3570 3571 If a pixmap is not exportable, `glamor_gbm_bo_from_pixmap()` would fail 3572 and the modesettings driver would consequently fail to do its page flip, 3573 which both prevents Present from working and also fill up the logs with 3574 error messages such as: 3575 3576 (EE) modeset(0): Failed to get GBM bo for flip to new front. 3577 (EE) modeset(0): present flip failed 3578 3579 Refactor the code so that `glamor_gbm_bo_from_pixmap()` takes care of 3580 making the pixmap exportable. 3581 3582 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 3583 Signed-off-by: Yuxuan Shui yshui@hadean.com 3584 See-also: https://gitlab.freedesktop.org/xorg/xserver/merge_requests/131 3585 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/68 3586 Fixes: 86b2d8740a "glamor: Reallocate pixmap storage without modifiers 3587 if necessary" 3588 (cherry picked from commit 26fe29f4fa53cbb7d51892e2cf397c084093812f) 3589 3590commit 34ad57e570f96dfe4bc493f14726b7a0ae6d45f9 3591Author: Olivier Fourdan <ofourdan@redhat.com> 3592Date: Tue Mar 12 15:38:03 2019 +0100 3593 3594 xwayland: Check status in GBM pixmap creation 3595 3596 The current code in `xwl_glamor_gbm_create_pixmap_for_bo()` may fail in 3597 several cases that are not checked for: 3598 3599 - `eglCreateImageKHR()` may have failed to create the image, 3600 - `glEGLImageTargetTexture2DOES()` may fail and set an error, 3601 - `glamor_set_pixmap_texture()` may fail for very large pixmaps 3602 because the corresponding FBO could not be created. 3603 3604 Trying to upload content to a pixmap with no texture will crash Mesa, 3605 glamor and Xwayland, e.g.: 3606 3607 XXX fail to create fbo. 3608 (EE) 3609 (EE) Backtrace: 3610 (EE) 0: Xwayland (OsSigHandler+0x29) 3611 (EE) 1: libpthread.so.0 (funlockfile+0x50) 3612 (EE) 2: libc.so.6 (__memmove_avx_unaligned_erms+0x215) 3613 (EE) 3: dri/i965_dri.so (_mesa_format_convert+0xab3) 3614 (EE) 4: dri/i965_dri.so (_mesa_texstore+0x205) 3615 (EE) 5: dri/i965_dri.so (store_texsubimage+0x28c) 3616 (EE) 6: dri/i965_dri.so (intel_upload_tex+0x13b) 3617 (EE) 7: dri/i965_dri.so (texture_sub_image+0x134) 3618 (EE) 8: dri/i965_dri.so (texsubimage_err+0x150) 3619 (EE) 9: dri/i965_dri.so (_mesa_TexSubImage2D+0x48) 3620 (EE) 10: Xwayland (glamor_upload_boxes+0x246) 3621 (EE) 11: Xwayland (glamor_copy+0x4d1) 3622 (EE) 12: Xwayland (miCopyRegion+0x96) 3623 (EE) 13: Xwayland (miDoCopy+0x43c) 3624 (EE) 14: Xwayland (glamor_copy_area+0x24) 3625 (EE) 15: Xwayland (damageCopyArea+0xba) 3626 (EE) 16: Xwayland (compCopyWindow+0x31c) 3627 (EE) 17: Xwayland (damageCopyWindow+0xd3) 3628 (EE) 18: Xwayland (miResizeWindow+0x7b7) 3629 (EE) 19: Xwayland (compResizeWindow+0x3a) 3630 (EE) 20: Xwayland (ConfigureWindow+0xa96) 3631 (EE) 21: Xwayland (ProcConfigureWindow+0x7d) 3632 (EE) 22: Xwayland (Dispatch+0x320) 3633 (EE) 23: Xwayland (dix_main+0x366) 3634 (EE) 24: libc.so.6 (__libc_start_main+0xf3) 3635 (EE) 25: Xwayland (_start+0x2e) 3636 (EE) 3637 Fatal server error: 3638 (EE) Caught signal 11 (Segmentation fault). Server aborting 3639 (EE) 3640 3641 Check for the possible cases of failure above and fallback to the 3642 regular glamor pixmap creation when an error is detected. 3643 3644 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 3645 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/661 3646 (cherry picked from commit fc6380a11be4c6202ed72f241dd9ee8c7c24671d) 3647 3648commit 6711b5c6fdf0581c77150306fff909d7bc63b8a4 3649Author: Olivier Fourdan <ofourdan@redhat.com> 3650Date: Thu May 9 10:36:19 2019 +0200 3651 3652 xwayland: Avoid a crash on pointer enter with a grab 3653 3654 On pointer enter notification, Xwayland checks for an existing pointer 3655 warp with a `NULL` sprite. 3656 3657 In turn, `xwl_pointer_warp_emulator_maybe_lock()` checks for an existing 3658 grab and the destination window using `XYToWindow()` which does not 3659 check for the actual sprite not being `NULL`. 3660 3661 So, in some cases, when the pointer enters the surface and there is an 3662 existing X11 grab which is not an ownerEvents grab, Xwayland would crash 3663 trying to dereference the `NULL` sprite pointer: 3664 3665 #0 __GI_raise () 3666 #1 __GI_abort () at abort.c:79 3667 #2 OsAbort () at utils.c:1351 3668 #3 AbortServer () at log.c:879 3669 #4 FatalError () at log.c:1017 3670 #5 OsSigHandler () at osinit.c:156 3671 #6 OsSigHandler () at osinit.c:110 3672 #7 <signal handler called> 3673 #8 XYToWindow (pSprite=0x0, x=0, y=0) at events.c:2880 3674 #9 xwl_pointer_warp_emulator_maybe_lock () at xwayland-input.c:2673 3675 #10 pointer_handle_enter () at xwayland-input.c:434 3676 3677 Avoid the crash by simply checking for the sprite being not `NULL` in 3678 `xwl_pointer_warp_emulator_maybe_lock()` 3679 3680 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 3681 Bugzilla: https://bugzilla.redhat.com/1708119 3682 (cherry picked from commit 0a07446318f248b65fcbc8ab8a73ead51153f09e) 3683 3684commit fbbdd5613122f0f7d74471f240c28e4c3b796065 3685Author: Adam Jackson <ajax@redhat.com> 3686Date: Thu May 30 14:07:06 2019 -0400 3687 3688 xserver 1.20.5 3689 3690 Signed-off-by: Adam Jackson <ajax@redhat.com> 3691 3692commit 240868527bb11beb9a03ad0d3f24506f1b983729 3693Author: Peter Hutterer <peter.hutterer@who-t.net> 3694Date: Mon Mar 25 13:19:41 2019 +1000 3695 3696 dix: leave last.valuators alone on slave switch 3697 3698 Terms: 3699 dev->last.valuator[] is the last value given to us by the driver 3700 dev->valuator.axisVal[] is the last value sent to the client 3701 dev->last.scroll[] is the abs value of the scroll axis as given by the driver, 3702 used for button emulation calculation (and the remainder) 3703 3704 This function updates the device's last.valuator state based on the current 3705 master axis state. This way, relative motion continues fluidly when switching 3706 between devices. Before mouse 2 comes into effect, it's valuator state is 3707 updated to wherever the pointer currently is so the relative event applies on 3708 top of that. 3709 3710 This can only work for x/y axes, all other axes aren't guaranteed to have the 3711 same meaning and/or may not be present: 3712 - xtest device: no valuator 2 3713 - mouse: valuator 2 is horizontal scroll axis 3714 - tablet: valuator 2 is pressure 3715 3716 Scaling the current value from the pressure range into the range for 3717 horizontal scrolling makes no sense. And it causes scroll jumps: 3718 3719 - scroll down, last.valuator == axisVal == 20 3720 - xdotool click 1, the XTest device doesn't have that valuator 3721 - scroll up 3722 - updateSlaveDeviceCoords reset last.valuator to 0 (axisVal == 20) 3723 - DeviceClassesChangedEvent includes value 20 for the axis 3724 - event is processed, last.value changes from 0 to -1 3725 - axisVal is updated to -1, causing a jump of -21 3726 3727 The same applies when we switch from tablet to mouse wheel if the pressure 3728 value is 0 on proximity out (basically guaranteed). So let's drop this code 3729 altogether and only leave the scaling for the relative x/y motion. 3730 3731 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 3732 (cherry picked from commit d7b1753d446ecde3ff58e3de39a634c3137473c8) 3733 3734commit 5bc29a67b8ab251048e0737eedcceb29bba25f5f 3735Author: Olivier Fourdan <ofourdan@redhat.com> 3736Date: Fri Mar 8 18:19:16 2019 +0100 3737 3738 glamor: pixmap FBO may not be allocated 3739 3740 If `_glamor_create_tex()` fails to allocate the FBO because of 3741 GL_OUT_OF_MEMORY error, the `pixmap_priv->fbo` is NULL. 3742 3743 However, `glamor_get_pixmap_texture()` doesn't actually check whether 3744 the `pixmap_priv->fbo` is NULL and will segfault with a NULL pointer 3745 dereference trying to access the `pixmap_priv->fbo->tex`. 3746 3747 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 3748 Closes: https://gitlab.freedesktop.org/xorg/xserver/issues/647 3749 (Cherry picked from commit 74479a99ecf18fdfdafd033b5efd8d785dd12c28) 3750 3751commit d0c7483b3edadf3ff9c12b4b024abd7b11b34ef1 3752Author: Peter Hutterer <peter.hutterer@who-t.net> 3753Date: Fri Apr 5 12:39:00 2019 +0200 3754 3755 xwayland: fix a realloc OOM error case 3756 3757 Found by coverity 3758 3759 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 3760 (Cherry picked from commit 7c25439f0db4c20a4eab4ed44262eb396305f58a) 3761 3762commit bb74db6b38f7ff75ed52a900fa62a995e77aede4 3763Author: Olivier Fourdan <ofourdan@redhat.com> 3764Date: Fri Apr 5 12:33:28 2019 +0200 3765 3766 xwayland: search for a render node to use 3767 3768 wl_drm's protocol "device" event provides the path to the DRM device, 3769 which may not be a render node, thus causing Xwayland to fall back to 3770 DRM authentication which may fail if the user has switched to another 3771 VT while Xwayland is starting. 3772 3773 Search for a render node corresponding to the given DRM device and try 3774 to use it instead, as render nodes do not need DRM authentication and 3775 Xwayland can make use of them if it can find one. 3776 3777 Closes: https://bugs.freedesktop.org/108038 3778 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 3779 (Cherry picked from commit 361894497c6802b62c2da4a3dc7e98939fb24404) 3780 3781commit 87491ced98705a14be2bb5a3ce7e41cb5ffc11d0 3782Author: Michel Dänzer <michel.daenzer@amd.com> 3783Date: Mon Mar 25 12:32:48 2019 +0100 3784 3785 present/scmd: Check that the flip and screen pixmap pitches match 3786 3787 If they don't, flipping will result in corrupted display. 3788 3789 Test case: 3790 3791 * Run Xorg at 1920x1080 with no window manager 3792 * glxgears -geometry 2048x1080 3793 (Cherry picked from commit ef91da2757050652c724f6e674e8b1acf5d0cb31) 3794 3795commit 578156337176bf759dace40fb5a81919f65562f1 3796Author: Michel Dänzer <michel.daenzer@amd.com> 3797Date: Mon Mar 25 12:32:04 2019 +0100 3798 3799 xwayland/present: Destroy sync_callback in xwl_present_cleanup 3800 3801 xwl_present_cleanup frees the struct xwl_present_window memory, 3802 so if there's a pending callback, we have to destroy it to prevent 3803 use-after-free in xwl_present_sync_callback. 3804 3805 Should fix issue #645. 3806 3807 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 3808 (Cherry picked from commit 2e18eec6f03cc4d9d5bf62853c65d341b363f4b3) 3809 3810commit 8e2f5410cec5b052fb4da9d1b774a5858903dc35 3811Author: Michel Dänzer <michel.daenzer@amd.com> 3812Date: Mon Mar 25 12:29:23 2019 +0100 3813 3814 Revert "gitlab-ci: Only run docker-image stage if relevant source files change" 3815 3816 This reverts commit 8694395fcfd407190db74f31064960e484747377. 3817 3818 Some scenarios have come to light where this failed to ensure the docker 3819 image exists: 3820 3821 * If the master branch of a forked repository is used for an MR which 3822 doesn't modify .gitlab-ci.yml, the docker-image job may not run. 3823 * If the docker-image job of the first pipeline in a forked repository 3824 is cancelled or fails for any reason, and .gitlab-ci.yml isn't 3825 modified for the next pipeline run. 3826 (Cherry picked from commit 0ec9a13c2a207e98e2dd6b352552c0e069af0727) 3827 3828commit 34553f502650641aec4f20ab43180804f17032c6 3829Author: Ray Strode <rstrode@redhat.com> 3830Date: Fri Nov 16 14:36:55 2018 -0500 3831 3832 dix: ensure work queues are cleared on reset 3833 3834 If the server resets, most client workqueues are cleaned up as the 3835 clients are killed. 3836 3837 The one exception is the server's client, which is exempt from 3838 the killing spree. 3839 3840 If that client has a queued work procedure active, it won't get 3841 cleared on reset. 3842 3843 This commit ensures it gets cleared too. 3844 3845 (cherry picked from commit 8738ce85df535bdfdfecfce1c0d64e209cc6e508) 3846 3847 Fixes: xorg/xserver#670 3848 3849commit 94f036d412f4f426abe950b57b9ec99b78068c39 3850Author: Adam Jackson <ajax@redhat.com> 3851Date: Mon Feb 25 13:31:13 2019 -0500 3852 3853 xserver 1.20.4 3854 3855 Signed-off-by: Adam Jackson <ajax@redhat.com> 3856 3857commit 9850d20c2aa2d8380f2e9c4521ef3c46a6c8fb00 3858Author: Michel Dänzer <michel.daenzer@amd.com> 3859Date: Fri Feb 22 10:22:49 2019 +0100 3860 3861 gitlab-ci: Run make distcheck in autotools build & test job 3862 3863 We don't set the PIGLIT_DIR and XTS_DIR environment variables for make 3864 distcheck for now, otherwise it complains about files left behind by 3865 piglit. 3866 3867 (cherry picked from commit 432fad04e7aaad1f10d732a51665ecbe934246c6) 3868 3869commit fcac9b10c90c8a06c60f563f5ad8da1d4e49b7fc 3870Author: Michel Dänzer <michel.daenzer@amd.com> 3871Date: Fri Feb 22 10:21:27 2019 +0100 3872 3873 glx,xquartz: Fix make distcheck 3874 3875 Guard BUILT_SOURCES and CLEANFILES by XWIN_GLX_WINDOWS/XQUARTZ. 3876 3877 (cherry picked from commit f9bbc9d5ead5cf298f0e3e0b7ac5229af830f6e7) 3878 3879commit 01a03a476a5f963846fdb7702217be17ac1ecd6c 3880Author: Michel Dänzer <michel.daenzer@amd.com> 3881Date: Fri Feb 22 11:16:36 2019 +0100 3882 3883 present/wnmd: Allow flipping if the window pixmap matches the toplevel's 3884 3885 Instead of testing window->redirectDraw. 3886 3887 With Xwayland, the toplevel window is always redirected, so this would 3888 unnecessarily preclude flipping there in some cases, e.g. with wlroots 3889 based Wayland compositors or with fullscreen X11 windows in weston. 3890 3891 Fixes issue #631. 3892 (Cherry picked from commit a093a88531599832811a1ba5cb85101af91a545e) 3893 3894commit e5984241ab8d7ca75eb3e5e893d63adb017ed0f5 3895Author: Michel Dänzer <michel.daenzer@amd.com> 3896Date: Fri Feb 22 11:56:23 2019 +0100 3897 3898 gitlab-ci: Add autotools build & test job 3899 (Cherry picked from commit 2f12c8017508f23195db92503435fc3ef183da4b) 3900 3901commit de6d87b61959cd3997feed3efe25fb2a36e543fb 3902Author: Michel Dänzer <michel.daenzer@amd.com> 3903Date: Fri Feb 22 11:56:01 2019 +0100 3904 3905 gitlab-ci: Use ccache 3906 3907 Meson picks it up automatically. 3908 3909 Based on: 3910 3911 * https://gitlab.freedesktop.org/mesa/mesa/merge_requests/240 3912 * https://gould.cx/ted/blog/2017/06/10/ccache-for-Gitlab-CI/ 3913 * https://stackoverflow.com/questions/53659419/ccache-no-hits-in-gitlab-ci 3914 3915 v2 based on the corresponding Mesa change: 3916 * Quote CCACHE_(BASE)DIR environment variables. 3917 * Clear ccache stats in before_script. 3918 * Move cache stanza to the build-and-test job, the cache isn't used in 3919 the docker-image job. 3920 3921 Reviewed-by: Adam Jackson <ajax@redhat.com> # v1 3922 (Cherry picked from commit b577df77451d34e28cfef3bda5031a43caeba90a) 3923 3924commit 37ad8dd978e8d35233e489e771f7dcf3e09c3b7a 3925Author: Michel Dänzer <michel.daenzer@amd.com> 3926Date: Fri Feb 22 11:55:31 2019 +0100 3927 3928 gitlab-ci: Add ccache to docker image, and leave in autotools 3929 3930 We're going to make use of these in build & test jobs. 3931 (Cherry picked from commit 537f06e21ee2c4c88ee062c26555338257e043ef) 3932 3933commit 61eccbcec930efeb1a124cc8edff70e6f65c927e 3934Author: Michel Dänzer <michel.daenzer@amd.com> 3935Date: Fri Feb 22 11:54:57 2019 +0100 3936 3937 gitlab-ci: Don't rely on $CI_PROJECT_NAME 3938 3939 The name of a forked repository can be changed later, in which case this 3940 would fail to refer to the main repository. 3941 3942 Pointed out by Eric Engestrom in 3943 https://gitlab.freedesktop.org/mesa/mesa/merge_requests/224 . 3944 (Cherry picked from commit fede384962edd175b1a2ce6edd8d2ef9b0fd74f0) 3945 3946commit e10cfd8fcc395360ce9acd93bfd7503da8e35381 3947Author: Michel Dänzer <michel.daenzer@amd.com> 3948Date: Fri Feb 22 11:54:41 2019 +0100 3949 3950 gitlab-ci: Only run docker-image stage if relevant source files change 3951 3952 Otherwise there's normally no need to run it. It will also run when a 3953 new branch is created, which ensures that the docker image always exists 3954 (e.g. in a newly forked repository). 3955 3956 Inspired by https://gitlab.freedesktop.org/mesa/mesa/merge_requests/143 3957 (Cherry picked from commit 8694395fcfd407190db74f31064960e484747377) 3958 3959commit 7246730edfb4f9079eb0ef1c41c43c47854a7432 3960Author: Michel Dänzer <michel.daenzer@amd.com> 3961Date: Fri Feb 22 11:53:55 2019 +0100 3962 3963 gitlab-ci: Set LC_ALL=C.UTF-8 3964 3965 Although piglit could now handle non-ASCII characters in the 3966 environment, meson was still failing without this (even though it's 3967 using Python 3). 3968 3969 Reviewed-by: Eric Anholt <eric@anholt.net> 3970 (Cherry picked from commit bc6998b728749015bfd93a85fbae48af777c9a34) 3971 3972commit 8db5a711da6a0b8f28c885b968b1e6bf2d45ce62 3973Author: Michel Dänzer <michel.daenzer@amd.com> 3974Date: Fri Feb 22 11:53:37 2019 +0100 3975 3976 test: Use .../piglit instead of .../piglit-*.py 3977 3978 The latter use Python 2 and break with any non-ASCII characters in the 3979 environment, the former uses Python 3 and works fine in that case. 3980 3981 Reviewed-by: Eric Anholt <eric@anholt.net> 3982 (Cherry picked from commit 4aaaf69229be574d6d11084225ecd91446a15170) 3983 3984commit 3314af2d15085118f2ca0d4ce65ae6ec1f38fbad 3985Author: Michel Dänzer <michel.daenzer@amd.com> 3986Date: Fri Feb 22 11:53:12 2019 +0100 3987 3988 gitlab-ci: Docker image can be generated as part of pipeline 3989 3990 This removes the dependency on an externally generated docker image, and 3991 should make it easier to update the docker image or make other changes 3992 related to it. 3993 3994 This is based on Debian testing, because I'm most familiar with Debian. 3995 But it should be easy to base it on another distro. 3996 3997 v2: 3998 * Use kaniko instead of docker-in-docker for image generation, so it can 3999 also work in unprivileged runners. 4000 * Drop piglit.conf & tetexec.cfg overrides, just make sure the files in 4001 the image work. 4002 (Cherry picked from commit f56d8e2282a3d1344aa1300c78faede0071a3bfa) 4003 4004commit e8b4e94d5192f72be3f768ac11a5e27890e6bc8d 4005Author: Michel Dänzer <michel.daenzer@amd.com> 4006Date: Fri Feb 22 11:52:42 2019 +0100 4007 4008 Drop Travis Linux build in favour of GitLab CI 4009 4010 Fold build-travis-deps.sh into .gitlab-ci.yml. 4011 4012 Preparation for the next change, which would break the Travis Linux 4013 build. 4014 4015 Reviewed-by: Eric Anholt <eric@anholt.net> 4016 (Cherry picked from commit ed44f9cd6a4ccf6de7dda70432764b4da10f3f44) 4017 4018commit 4925f38dc0559a8cbe01ec947edeb5eeb78a4df6 4019Author: Michel Daenzer <michel@daenzer.net> 4020Date: Fri Feb 22 11:51:57 2019 +0100 4021 4022 Make artifacts of piglit results if job fails 4023 4024 Can be useful for figuring out what caused the failure. 4025 (Cherry picked from commit a7472da94136ef977b8d6af956a15767736236ea) 4026 4027commit b6ef90d83b4fd9230ca334116c61fd52f1d68209 4028Author: Michel Daenzer <michel@daenzer.net> 4029Date: Fri Feb 22 11:51:35 2019 +0100 4030 4031 travis: Use a single meson invocation 4032 4033 The prefix setting didn't take for some reason. 4034 (Cherry picked from commit c1bb392b1df4023e9dec489facfa221a53499ddc) 4035 4036commit 7753fd9989e280783daeee1994d019155735dee5 4037Author: Adam Jackson <ajax@redhat.com> 4038Date: Fri Feb 22 11:50:05 2019 +0100 4039 4040 gitlab: Skip the docker-in-docker step 4041 4042 No idea which cult's cargo I was looking at there. Cuts about a minute 4043 off the build time. 4044 4045 Signed-off-by: Adam Jackson <ajax@redhat.com> 4046 (Cherry picked from commit a41ccaa0857146e986f11360d4f2aeb3913af9a6) 4047 4048commit 524104e15c7169114e4fa3dd832990c58ee563a4 4049Author: Peter Hutterer <peter.hutterer@who-t.net> 4050Date: Fri Feb 8 13:29:14 2019 +1000 4051 4052 Xi: lock the input thread for any pointer barrier list manipulation 4053 4054 The input thread checks the barriers for pointer positioning, swapping the 4055 list out from underneath is considered impolite. 4056 4057 Reported-by: Michel Dänzer <michel.daenzer@amd.com> 4058 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 4059 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4060 (cherry picked from commit 678d64aa2e929368b6d6f2b83bbf5540c4fa292d) 4061 4062commit ae9dda1e2620f402b434f10df581b0fdf0495ee8 4063Author: Peter Harris <pharris@opentext.com> 4064Date: Wed Jan 30 14:51:07 2019 -0500 4065 4066 os: Fix GetTimeInMicros resolution 4067 4068 GetTimeInMillis is called first, which sets clockid to 4069 CLOCK_MONOTONIC_COARSE, which is typically much lower resolution than 4070 the callers of GetTimeInMicros want. 4071 4072 Prior to a779fda224bee0c4d27636503367e55ae93b33c2, GetTimeInMillis and 4073 GetTimeInMicros did not share a clockid. 4074 4075 Restore the clockid split to fix the granularity of GetTimeInMicros. 4076 4077 Signed-off-by: Peter Harris <pharris@opentext.com> 4078 (cherry picked from commit 937a5b78a2f6ea771132ff0f9ece708a23c1bdad) 4079 4080commit 013c28a122a61beaf3a4ee7cc92b7ca2c92de7ab 4081Author: A. Wilcox <AWilcox@Wilcox-Tech.com> 4082Date: Sat Jan 26 15:37:56 2019 -0600 4083 4084 DRI2: Add another Coffeelake PCI ID 4085 4086 A user of Adélie Linux reported that modesetting wasn't working properly on 4087 their Intel i7-9700K-integrated UHD 630 GPU. Xorg.0.log showed: 4088 4089 [ 131.902] (EE) modeset(0): [DRI2] No driver mapping found for PCI device 0x8086 / 0x3e98 4090 [ 131.902] (EE) modeset(0): Failed to initialize the DRI2 extension. 4091 4092 Indeed, that PCI ID is missing from i965_pci_ids. Adding it fixed the issue 4093 and allowed the system to work with i965_dri under modesetting. 4094 4095 (cherry picked from commit d3a26bbf618507e1ca05b2bc99a880075b77db77) 4096 4097commit a51d7a730cfd82ab44ea5510eafdbe9d7043e706 4098Author: Adam Jackson <ajax@redhat.com> 4099Date: Tue Oct 16 12:58:25 2018 -0400 4100 4101 vnd: Fix a silly memory leak 4102 4103 'disp' was already allocated by LookupVendorPrivDispatch above, 4104 clobbering it will do no good. 4105 4106 Signed-off-by: Adam Jackson <ajax@redhat.com> 4107 (cherry picked from commit 709c6562975c3bea10dd0571527a4aac79a6bf6f) 4108 4109commit 544d0e961cfa96043ed3e5a718a768444c4a4c4d 4110Author: Adam Jackson <ajax@redhat.com> 4111Date: Wed Jan 16 14:20:17 2019 -0500 4112 4113 mi: When {en,dis}abling extensions, match names case-insensitively 4114 4115 Both because extension names are inconsistently capitalized on the wire, 4116 and because the table we're walking spells it COMPOSITE not Composite. 4117 The latter is certainly also a bug, but there's no reason for us to be 4118 that strict. 4119 4120 Signed-off-by: Adam Jackson <ajax@redhat.com> 4121 (cherry picked from commit bf991a5f989c5e6e726a3731f468b7b7d65d9f4a) 4122 4123commit 2215e8c7cf06f46f9995c8a7817c7ede9774a551 4124Author: Adam Jackson <ajax@redhat.com> 4125Date: Mon Jan 14 12:42:00 2019 -0500 4126 4127 dri3: Fix XACE access mode for open and get_supported_modifiers 4128 4129 Neither opening a screen nor querying its modifiers confers the right to 4130 attach the buffer for any particular pixmap. GetAttr seems more correct. 4131 4132 Fixes: xorg/xserver#550 4133 (cherry picked from commit 086c2e3de55bbf0cbc1d97f7dc2db70a7f5e69e3) 4134 4135commit 59e0b5f041eee92d0a51303082005ad8f053d8ca 4136Author: Michel Dänzer <michel.daenzer@amd.com> 4137Date: Wed Dec 5 18:31:27 2018 +0100 4138 4139 xfree86/modes: Don't clobber gamma LUT of compatibility output's CRTC 4140 4141 If the driver calls xf86HandleColormaps, CMapChangeGamma updates the HW 4142 gamma LUT of all CRTCs via xf86RandR12LoadPalette. However, 4143 xf86RandR12ChangeGamma was then clobbering the gamma LUT of the RandR 4144 1.2 compatibility output's CRTC with the gamma curves computed from the 4145 screen's global gamma values. 4146 4147 Fix this by bailing if xf86RandR12LoadPalette is installed. 4148 4149 Fixes: 02ff0a5d7e32 "xf86RandR12: Fix XF86VidModeSetGamma triggering a 4150 BadImplementation error" 4151 (cherry picked from commit 30044b2253c2dd51e1aedb2f897159c657ca8f0d) 4152 4153commit 712d0e86aaee96ffe1ad6a708b9134889e47d1c6 4154Author: Maya Rashish <maya@NetBSD.org> 4155Date: Thu Jan 10 14:55:17 2019 +0200 4156 4157 xfree86: Try nouveau on NetBSD as well. 4158 4159 (cherry picked from commit e3fb178617a579c98877a3baae14c4dfe4d55db8) 4160 4161commit ff1d1692e73b1729a6ced84bf58bcbe17f1c9d94 4162Author: Maya Rashish <maya@NetBSD.org> 4163Date: Thu Jan 10 18:39:33 2019 +0200 4164 4165 Fix typo in error message 4166 4167 (cherry picked from commit bf2a7bb4ffbd199882fe4bd183561469833a9e6b) 4168 4169commit c091ea5e38b5c41eaf3b10c43dc043401e8251c5 4170Author: Alan Coopersmith <alan.coopersmith@oracle.com> 4171Date: Tue Jan 1 16:49:32 2019 -0800 4172 4173 os: Report errors opening authorization file (#469) 4174 4175 Fixes: xorg/xserver#469 4176 4177 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 4178 (cherry picked from commit 7fb6338c68e158053295cb448faa5c559aa9990c) 4179 4180commit 7b0f6102df3b18a048b791fe2304679b1eb2c9e7 4181Author: Michel Dänzer <michel.daenzer@amd.com> 4182Date: Wed Dec 19 10:06:23 2018 +0100 4183 4184 glamor: Check that storage format is compatible with RENDER format 4185 4186 Fixes x2r10g10b10 related rendercheck failures. 4187 4188 Reviewed-by: Eric Anholt <eric@anholt.net> 4189 (cherry picked from commit 7e6faa5b3c05e0b7149ee840403885b0b40b5827) 4190 4191commit 356cf07b8a8fe6faedf701ae259f8f077a234dd5 4192Author: Peter Hutterer <peter.hutterer@who-t.net> 4193Date: Wed Nov 28 10:05:49 2018 +1000 4194 4195 test: fix failing tests 4196 4197 Broken since 69d8ea4a49793a94f821d1a328856901a1c02a5a because our fake screen 4198 didn't have a root window and writing the XKB rules prop would happily 4199 segfault. Fix this by setting up the required bits. 4200 4201 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 4202 Tested-by: Michel Dänzer michel.daenzer@amd.com 4203 (cherry picked from commit fde27b9b4814b18aca6ec587bd3cfe9ab04b0c72) 4204 4205commit c44eee243e6f4e488b586ad02e4fd00f587ebb49 4206Author: Ilia Mirkin <imirkin@alum.mit.edu> 4207Date: Mon Dec 10 23:34:11 2018 -0500 4208 4209 modesetting: fix conn_id termination and potential overrun by 1 byte 4210 4211 Noticed when porting this logic to xf86-video-nouveau, and valgrind 4212 complained about conditional jump based on uninitialized data. 4213 4214 Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> 4215 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.com> 4216 (cherry picked from commit 48b1af2718ab81c66f565438553415c05f1faa5c) 4217 4218commit 10609630e753b475566be27b186af4b2e290fdc6 4219Author: Adam Jackson <ajax@redhat.com> 4220Date: Thu Nov 29 14:48:11 2018 -0500 4221 4222 automake: Distribute meson's configure header templates 4223 4224 Fixes: xorg/xserver#17 4225 (cherry picked from commit 82ed89c0f8b18d8214430580dc80c8d3e37bef33) 4226 4227commit f5a77233b983eb06eb727d25148b68d1bc410022 4228Author: Lionel Landwerlin <lionel.g.landwerlin@intel.com> 4229Date: Fri Nov 9 12:34:59 2018 +0000 4230 4231 present: fix compile warning with debug traces 4232 4233 Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> 4234 (cherry picked from commit a425eee6dce3b0cfd18b591907e8302a91b648c6) 4235 4236commit 4e12cba65682e97b056d8a8207189d4cf9c31862 4237Author: Lyude Paul <lyude@redhat.com> 4238Date: Tue Nov 13 20:14:10 2018 -0500 4239 4240 modesetting: Actually disable CRTCs in legacy mode 4241 4242 Believe it or not, somehow we've never done this in legacy mode! We 4243 currently simply change the DPMS property on the CRTC's output's 4244 respective DRM connector, but this means that we're just setting the 4245 CRTC as inactive-not disabled. From the perspective of the kernel, this 4246 means that any shared resources used by the CRTC are still in use. 4247 4248 This can cause problems for drivers that are not yet fully atomic, 4249 despite using the atomic helpers internally. For instance: if CRTC-1 and 4250 CRTC-2 are still enabled and use shared resources within the kernel (an 4251 MST topology, for example), and then userspace tries to go enable CRTC-3 4252 on the same topology this might suddenly fail if CRTC-3 needs the shared 4253 resources CRTC-1 and CRTC-2 are using. While I don't know of any 4254 situations in the mainline kernel that actually trigger this, future 4255 plans for reworking the atomic check of MST drivers are absolutely 4256 going to make this into a real issue (they already are in my WIP 4257 branches for the kernel). 4258 4259 So: actually do the right thing here and disable CRTCs when they're not 4260 going to be used anymore, even in legacy mode. 4261 4262 Signed-off-by: Lyude Paul <lyude@redhat.com> 4263 (cherry picked from commit 7a44e8d4007b9c3ca55a5cc3f5e98601565311c7) 4264 4265commit 652918e736bcc577e221184415dcf61c05ac7bfb 4266Author: Alan Coopersmith <alan.coopersmith@oracle.com> 4267Date: Sun Nov 25 12:56:29 2018 -0800 4268 4269 Update configure.ac bug URL for gitlab migration 4270 4271 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 4272 (cherry picked from commit 17a22ad948009badbc79bbcd9a067004c98f5744) 4273 4274commit 40b22a0571d422473a7a324daa65f6a768181644 4275Author: Alan Coopersmith <alan.coopersmith@oracle.com> 4276Date: Sun Nov 18 15:49:27 2018 -0800 4277 4278 Update README for gitlab migration 4279 4280 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 4281 (cherry picked from commit 5d097c2a20fce44cdb9d5c302d46bc7fa16edfec) 4282 4283commit a352f979545723054b0a74862a56dc53b1be93fb 4284Author: Olivier Fourdan <ofourdan@redhat.com> 4285Date: Tue Jan 8 12:48:53 2019 +0100 4286 4287 xwayland: handle case without any crtc 4288 4289 Xwayland creates and destroys the CRTC along with the Wayland outputs, 4290 so there is possibly a case where the number of CRTC drops to 0. 4291 4292 However, `xwl_present_get_crtc()` always return `crtcs[0]` which is 4293 invalid when `numCrtcs` is 0. 4294 4295 That leads to crash if a client queries the Present capabilities when 4296 there is no CRTC, the backtrace looks like: 4297 4298 #0 raise() from libc.so 4299 #1 abort() from libc.so 4300 #2 OsAbort() at utils.c:1350 4301 #3 AbortServer() at log.c:879 4302 #4 FatalError() at log.c:1017 4303 #5 OsSigHandler() at osinit.c:156 4304 #6 OsSigHandler() at osinit.c:110 4305 #7 <signal handler called> 4306 #8 main_arena() from libc.so 4307 #9 proc_present_query_capabilities() at present_request.c:236 4308 #10 Dispatch() at dispatch.c:478 4309 #11 dix_main() at main.c:276 4310 4311 To avoid returning an invalid pointer (`crtcs[0]`) in that case, simply 4312 check for `numCrtcs` being 0 and return `NULL` in that case. 4313 4314 Thanks to Michel Dänzer <michel.daenzer@amd.com> for pointing this as a 4315 possible cause of the crash. 4316 4317 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 4318 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 4319 Bugzilla: https://bugzilla.redhat.com/1609181 4320 (cherry picked from commit e8295c50209f2963fa2823e8de7e8363a38cd2d1) 4321 4322commit 210cd529064348de7d4f9a9050b0cf68f8fd326c 4323Author: Michel Dänzer <michel.daenzer@amd.com> 4324Date: Thu Nov 15 17:16:59 2018 +0100 4325 4326 xwayland: Don't take buffer release queue into account for frame timer 4327 4328 The buffer release queue has two kinds of entries: 4329 4330 * Pending async flips. 4331 * Completed flips waiting for their buffer to be released by the Wayland 4332 compositor. 4333 4334 xwl_present_timer_callback neither completes async flips nor releases 4335 buffers, so the timer isn't needed for the buffer release queue. 4336 4337 (cherry picked from commit e6cd1c9bdefe83e7d99b703a68d26eebb451f889) 4338 4339commit 7c28b0e34ecbe9842193733dfd86097c06921406 4340Author: Michel Dänzer <michel.daenzer@amd.com> 4341Date: Fri Nov 9 17:18:53 2018 +0100 4342 4343 xwayland: Don't need xwl_window anymore in xwl_present_queue_vblank 4344 4345 Fixes issue #12. Presumably the problem was that Present operations on 4346 unmapped windows were executed immediately instead of only when reaching 4347 the target MSC. 4348 4349 (cherry picked from commit f541615342ce6bfb0e6d4e68deb3a924a87e8ba9) 4350 4351commit 46135957095ec954e21107d1001452e9533ef2ee 4352Author: Michel Dänzer <michel.daenzer@amd.com> 4353Date: Thu Nov 1 18:24:28 2018 +0100 4354 4355 xwayland: Add xwl_present_unrealize_window 4356 4357 When a window is unrealized, a pending frame callback may never be 4358 called, which could result in repeatedly freezing until the frame timer 4359 fires after a second. 4360 4361 Fixes these symptoms when switching from fullscreen to windowed mode in 4362 sauerbraten. 4363 4364 (cherry picked from commit 8c9538573cb9a342897eb3fb4b0c1e4ed917bd0e) 4365 4366commit 98f41563e6599eb762e6a3ec12f99ba6b5388039 4367Author: Michel Dänzer <michel.daenzer@amd.com> 4368Date: Thu Nov 1 18:44:24 2018 +0100 4369 4370 xwayland: Replace xwl_window::present_window with ::present_flipped 4371 4372 There's no need to keep track of the window which last performed a 4373 Present flip. This fixes crashes due to the assertion in 4374 xwl_present_flips_stop failing. Fixes issue #10. 4375 4376 The damage generated by a flip only needs to be ignored once, then 4377 xwl_window::present_flipped can be cleared. This may fix freezing in 4378 the (hypothetical) scenario where Present flips are performed on a 4379 window, followed by other drawing requests using the window as the 4380 destination, but nothing triggering xwl_present_flips_stop. The damage 4381 from the latter drawing requests would continue being ignored. 4382 4383 (cherry picked from commit 6b016d58d23d16eaae9908a92ed90547d1926317) 4384 4385commit f393801dbbe89bce716a8ceeb2b5c8440b9021ce 4386Author: Michel Dänzer <michel.daenzer@amd.com> 4387Date: Thu Oct 18 17:42:01 2018 +0200 4388 4389 xwayland: Complete "synchronous" Present flips from xwl_present_msc_bump 4390 4391 Completing them from xwl_present_sync_callback had at least two issues: 4392 4393 * It was before the MSC was incremented in xwl_present_frame_callback, 4394 so the MSC value in the completion event could be lower than the 4395 target specified by the client. This could cause hangs with the Mesa 4396 Vulkan drivers. 4397 * It allowed clients to run at a frame-rate higher than the Wayland 4398 compositor's frame-rate, wasting energy on generating frames which 4399 were never displayed. This isn't expected to happen unless the client 4400 specified PresentOptionAsync (in which case flips are still completed 4401 from xwl_present_sync_callback, allowing higher frame-rates). 4402 4403 v2: 4404 * Make xwl_present_has_events return true when there's a pending 4405 "synchronous" flip, so those complete after at most ~1 second even if 4406 the Wayland server doesn't send a frame event. 4407 4408 Bugzilla: https://bugs.freedesktop.org/106713 4409 (cherry picked from commit ace551d8a2603e37b18237a52f62d627c75d9e2a) 4410 4411commit e646e3054a3e1dbe8ff3906a546897246bcc78f0 4412Author: Michel Dänzer <michel.daenzer@amd.com> 4413Date: Thu Oct 18 17:36:24 2018 +0200 4414 4415 xwayland: Rename xwl_present_events_notify to xwl_present_msc_bump 4416 4417 And consolidate more code from xwl_present_timer_callback and 4418 xwl_present_frame_callback in it. 4419 4420 (cherry picked from commit 2bfc46d4147dc0bec4cdbb80431a0f4cc1d3b030) 4421 4422commit 47aed554b7c12c0c7f496c86a435dddaa51ae9bf 4423Author: Michel Dänzer <michel.daenzer@amd.com> 4424Date: Wed Oct 24 11:23:05 2018 +0200 4425 4426 xwayland: Use xwl_present_reset_timer in xwl_present_timer_callback 4427 4428 Apart from simplifying the code, this should also prevent a condition 4429 (which might only be possible with the following fix) reported in 4430 https://gitlab.freedesktop.org/wayland/weston/issues/115#note_52467: 4431 4432 1. xwl_present_timer_callback indirectly calls xwl_present_reset_timer 4433 -> xwl_present_free_timer 4434 2. xwl_present_timer_callback then returns a non-0 value, so DoTimer 4435 calls TimerSet with the old xwl_present_window->frame_timer pointer 4436 which was freed in step 1 => use after free 4437 4438 Calling xwl_present_reset_timer explicitly passes NULL to TimerSet if 4439 step 1 freed xwl_present_window->frame_timer, and it will allocate a new 4440 one. 4441 4442 (cherry picked from commit 5e8b9a3a563047e3998d45e761f7a50e4b0f6cb3) 4443 4444commit cf8e064ec0bed45b8cda9ae390c7af78d8ede50f 4445Author: Olivier Fourdan <ofourdan@redhat.com> 4446Date: Fri Oct 19 16:04:32 2018 +0200 4447 4448 xwayland: do not crash if `gbm_bo_create()` fails 4449 4450 The function `xwl_glamor_gbm_create_pixmap()` first creates a buffer 4451 objects and then creates the xwl_pixmap from it. 4452 4453 However, `xwl_glamor_gbm_create_pixmap_for_bo()` is not called if the 4454 buffer object creation fails, and `xwl_glamor_gbm_create_pixmap()` 4455 simply returns `glamor_create_pixmap()`. 4456 4457 The problem with this is that if `xwl_glamor_gbm_create_pixmap_for_bo()` 4458 is not called then neither is `xwl_pixmap_set_private()` and further 4459 calls to `xwl_pixmap_get()` will return NULL and cause a NULL pointer 4460 dereference if the return value is not checked: 4461 4462 #0 xwl_glamor_gbm_get_wl_buffer_for_pixmap () 4463 at hw/xwayland/xwayland-glamor-gbm.c:248 4464 #1 xwl_window_post_damage () at hw/xwayland/xwayland.c:697 4465 #2 xwl_display_post_damage () at hw/xwayland/xwayland.c:759 4466 #3 block_handler () at hw/xwayland/xwayland.c:890 4467 #4 BlockHandler () at dix/dixutils.c:388 4468 #5 WaitForSomething () at os/WaitFor.c:201 4469 #6 Dispatch () at dix/dispatch.c:421 4470 #7 dix_main () at dix/main.c:276 4471 #8 __libc_start_main () at ../csu/libc-start.c:308 4472 #9 _start () 4473 4474 (gdb) print xwl_pixmap 4475 $1 = (struct xwl_pixmap *) 0x0 4476 4477 Make sure we check for `xwl_pixmap_get()` returned value where relevant 4478 and fail gracefully if this is the case. 4479 4480 See also: https://gitlab.gnome.org/GNOME/mutter/issues/340 4481 4482 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 4483 Reviewed-by: Marco Trevisan <mail@3v1n0.net> 4484 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 4485 (cherry picked from commit 036794bebce72a3fa2f95996d2e537ff568e0ff1) 4486 4487commit f89518e17f7d507734af212785e0b3e47954f603 4488Author: Olivier Fourdan <ofourdan@redhat.com> 4489Date: Mon Oct 22 11:48:25 2018 +0200 4490 4491 present/wnmd: Fix use after free on CRTC removal 4492 4493 Xwayland will add and remove CRTCs as Wayland outputs are added or 4494 removed. 4495 4496 If there is a pending flip when this occurs, the 4497 `xwl_present_sync_callback()` will be triggered after the Xwayland 4498 output's RRCtrcPtr has been destroyed, hence causing a crash in Xwayland 4499 while trying to use freed memory: 4500 4501 #1 abort () 4502 #2 OsAbort () at utils.c:1350 4503 #3 AbortServer () at log.c:877 4504 #4 FatalError () at log.c:1015 4505 #5 OsSigHandler () at osinit.c:156 4506 #6 <signal handler called> 4507 #7 dixGetPrivate () at ../include/privates.h:122 4508 #8 dixLookupPrivate () at ../include/privates.h:166 4509 #9 present_screen_priv () at present_priv.h:198 4510 #10 present_wnmd_flip () at present_wnmd.c:358 4511 #11 present_wnmd_execute () at present_wnmd.c:466 4512 #12 present_wnmd_re_execute () at present_wnmd.c:80 4513 #13 xwl_present_sync_callback () at xwayland-present.c:287 4514 #14 ffi_call_unix64 () from /lib64/libffi.so.6 4515 #15 ffi_call () from /lib64/libffi.so.6 4516 #16 wl_closure_invoke () at src/connection.c:1006 4517 #17 dispatch_event () at src/wayland-client.c:1427 4518 #18 dispatch_queue () at src/wayland-client.c:1573 4519 #19 wl_display_dispatch_queue_pending () at src/wayland-client.c:1815 4520 #20 wl_display_dispatch_pending () at src/wayland-client.c:1878 4521 #21 xwl_read_events () at xwayland.c:814 4522 #22 ospoll_wait () at ospoll.c:651 4523 #23 WaitForSomething () at WaitFor.c:208 4524 #24 Dispatch () at ../include/list.h:220 4525 #25 dix_main () at main.c:276 4526 4527 To avoid the issue, get the `ScreenPtr` from the window instead of the 4528 CRTC that might have been just freed, `xwl_present_flip()` has no use 4529 for the CRTC anyway. 4530 4531 Bugzilla: https://bugs.freedesktop.org/108249 4532 Suggested-by: Michel Daenzer <michel.daenzer@amd.com> 4533 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 4534 Reviewed-by: Michel Daenzer <michel.daenzer@amd.com> 4535 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 4536 (cherry picked from commit b768b7d6cec41b8b320c468ec41aab5a8b49b27b) 4537 4538commit 64f5e6ec2d297f90e9b9785a1cb7285d609a1877 4539Author: Michel Dänzer <michel.daenzer@amd.com> 4540Date: Fri Oct 19 18:27:37 2018 +0200 4541 4542 xwayland: Plug leaks in xwl_present_sync_callback 4543 4544 xwl_present_window->sync_callback was leaked. 4545 4546 The event memory was leaked if the corresponding buffer had already been 4547 released. 4548 4549 (cherry picked from commit cb0de153bf0c486da7e968ab0f258c9c0c9ed34a) 4550 4551commit 971d418113740cae2d7d393850bad4926d1a7e86 4552Author: Adam Jackson <ajax@redhat.com> 4553Date: Thu Oct 25 09:03:18 2018 -0400 4554 4555 xserver 1.20.3 4556 4557 Signed-off-by: Adam Jackson <ajax@redhat.com> 4558 4559commit da15c7413916f754708c62c2089265528cd661e2 4560Author: Matthieu Herrb <matthieu@herrb.eu> 4561Date: Tue Oct 23 21:29:09 2018 +0200 4562 4563 LogFilePrep: add a comment to the unsafe format string. 4564 4565 CVE-2018-14665 also made it possible to exploit this to access 4566 memory. With -logfile forbidden when running with elevated privileges 4567 this is no longer an issue. 4568 4569 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 4570 Reviewed-by: Adam Jackson <ajax@redhat.com> 4571 (cherry picked from commit 248d164eae27f1f310266d78e52f13f64362f81e) 4572 4573commit 8a59e3b7dbb30532a7c3769c555e00d7c4301170 4574Author: Matthieu Herrb <matthieu@herrb.eu> 4575Date: Tue Oct 23 21:29:08 2018 +0200 4576 4577 Disable -logfile and -modulepath when running with elevated privileges 4578 4579 Could cause privilege elevation and/or arbitrary files overwrite, when 4580 the X server is running with elevated privileges (ie when Xorg is 4581 installed with the setuid bit set and started by a non-root user). 4582 4583 CVE-2018-14665 4584 4585 Issue reported by Narendra Shinde and Red Hat. 4586 4587 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 4588 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 4589 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 4590 Reviewed-by: Adam Jackson <ajax@redhat.com> 4591 (cherry picked from commit 50c0cf885a6e91c0ea71fb49fa8f1b7c86fe330e) 4592 4593commit cfc3dec09e1a56bb050ba31bde0bbf499596063a 4594Author: Peter Hutterer <peter.hutterer@who-t.net> 4595Date: Tue Oct 16 09:42:51 2018 +1000 4596 4597 xfree86: fix readlink call 4598 4599 Misplaced parenthesis caused us to compare the sizeof, not the readlink return 4600 value. 4601 4602 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 4603 (cherry picked from commit bd5fe7593fd0df236f3b2be1f062166ddba7d67c) 4604 4605commit 2a0c6c15c35cd262e7cdb86dcc43cb1aeb714c8e 4606Author: Adam Jackson <ajax@redhat.com> 4607Date: Mon Oct 15 11:17:35 2018 -0400 4608 4609 xserver 1.20.2 4610 4611 Signed-off-by: Adam Jackson <ajax@redhat.com> 4612 4613commit f5dc787fc9bb102e4eea477d1817e35370084ea2 4614Author: Adam Jackson <ajax@redhat.com> 4615Date: Wed Oct 10 14:09:11 2018 -0400 4616 4617 fbdevhw: Refuse to touch PCI devices on the fallback probe path 4618 4619 Fixes: https://gitlab.freedesktop.org/xorg/driver/xf86-video-fbdev/issues/9 4620 Signed-off-by: Adam Jackson <ajax@redhat.com> 4621 (cherry picked from commit fc78bcca21e767697de6ad4d8e03b6728856f613) 4622 4623commit 4795c069a503144ea31f01de0c039f32d9880292 4624Author: Adam Jackson <ajax@redhat.com> 4625Date: Fri Oct 5 14:50:20 2018 -0400 4626 4627 glamor/egl: Avoid crashing on broken configurations 4628 4629 0a9415cf apparently can tickle bugs in the GL stack where glGetString 4630 returns NULL, presumably because the eglMakeCurrent() didn't manage to 4631 actually install a dispatch table and you're hitting a stub function. 4632 That's clearly not our bug, but if it happens we should at least not 4633 crash. Notice this case and fail gently. 4634 4635 Signed-off-by: Adam Jackson <ajax@redhat.com> 4636 (cherry picked from commit af151895f3cb1755a7a5631f2398a3d3b219cbef) 4637 4638commit 1e3c5d614ee33d9eac1d2cf6366feeb8341fc0f4 4639Author: Adam Jackson <ajax@redhat.com> 4640Date: Fri Sep 14 11:33:43 2018 -0400 4641 4642 glamor_egl: Don't initialize on llvmpipe 4643 4644 Mesa started supporting GL_OES_EGL_image on llvmpipe in 17.3, after this 4645 commit: 4646 4647 commit bbdeddd5fd0b797e1e281f058338b3da4d98029d 4648 Author: Gurchetan Singh <gurchetansingh@chromium.org> 4649 Date: Tue Aug 1 14:49:33 2017 -0700 4650 4651 st/dri: add drisw image extension 4652 4653 That's pretty cool, but it means glamor now thinks it can initialize on 4654 llvmpipe. This is almost certainly not what anyone wants, as glamor on 4655 llvmpipe is pretty much uniformly slower than fb. 4656 4657 This fixes both Xorg and Xwayland to refuse glamor in such a setup. 4658 Xephyr is left alone, both because glamor is not the default there and 4659 because Xephyr+glamor+llvmpipe is one of the easier ways to get xts to 4660 exercise glamor. 4661 4662 The (very small) downside of this change is that you lose DRI3 support. 4663 This wouldn't have helped you very much (since an lp glamor blit is 4664 slower than a pixman blit), but it would eliminate the PutImage overhead 4665 for llvmpipe's glXSwapBuffers. A future change should add DRI3 support 4666 for the fb-only case. 4667 4668 Reviewed-by: Eric Anholt <eric@anholt.net> 4669 Signed-off-by: Adam Jackson <ajax@redhat.com> 4670 (cherry picked from commit 0a9415cf793babed1f28c61f8047d51de04f1528) 4671 4672commit b58aa8ed9b0a052d615d34b103aa54ce357df12e 4673Author: Alexander Volkov <a.volkov@rusbitech.ru> 4674Date: Mon Feb 26 19:41:18 2018 +0300 4675 4676 os/xdmcp: Don't create a new socket in XdmcpReset() 4677 4678 xdmcpSocket survives during the reset, there is no 4679 need to create a new one. 4680 4681 This commit restores logic that was broken by 4682 49c0f2413d32fdfe36e45861fcb32aaeab633094 in Xorg 1.19. 4683 4684 Signed-off-by: Alexander Volkov <a.volkov@rusbitech.ru> 4685 (cherry picked from commit 32677ce03d793a1f2aa8871112eb3d19b3cb762f) 4686 4687commit b1215fb075b0166d5742c72041f58d512273bb50 4688Author: Pierre Ossman <ossman@cendio.se> 4689Date: Wed Oct 3 10:28:52 2018 +0200 4690 4691 Switch automatic composite update to WorkQueue 4692 4693 It is currently (ab)using the screen BlockHandler callback to do 4694 this. But this can cause problems with other extension as their 4695 block handlers might have executed before Composite's. And the 4696 operations Composite does might result in them wanting to change 4697 timeouts. 4698 4699 Practically this caused problems for TigerVNC's VNC extension which 4700 failed to send out updates for Composite's screen updates. 4701 4702 (cherry picked from commit 1bd5d0a53c5ff4169c5a6704c1c4b276f998b938) 4703 4704commit a41b6ef2244f1a0003dada0962a9d785e74d7981 4705Author: Adam Jackson <ajax@redhat.com> 4706Date: Mon Oct 1 11:42:37 2018 -0400 4707 4708 modesetting: Don't free(dst) in drmmode_prop_info_copy 4709 4710 The destination is always either on the stack or in the middle of some 4711 struct. 4712 4713 Signed-off-by: Adam Jackson <ajax@redhat.com> 4714 (cherry picked from commit 43a0f9a5dbb469f4f403a8530f33be67618933b8) 4715 4716commit 9403335910f4e6b862af5c5ef835074be2c52edb 4717Author: Cedric Roux <sed@free.fr> 4718Date: Wed Sep 12 19:14:18 2018 +0200 4719 4720 miext/damage: take care of the coordinate mode in damagePolyPoint 4721 4722 The mode (CoordModeOrigin or CoordModePrevious) was not taken into 4723 account when computing the box. The result was a bad drawing of 4724 points in some situations (on my hardware/software configuration, 4725 calling XDrawString followed by XDrawPoints in the mode 4726 CoordModePrevious). 4727 4728 Signed-off-by: Cedric Roux <sed@free.fr> 4729 Signed-off-by: Eric Anholt <eric@anholt.net> 4730 (cherry picked from commit bc36594e0eb8bfa5a673bcfd8c8168f70994a1df) 4731 4732commit c26a47b4f2e170e283ca1683d5b6ca2d04823e80 4733Author: Olivier Fourdan <ofourdan@redhat.com> 4734Date: Thu Sep 20 16:32:29 2018 +0200 4735 4736 xwayland: Use `double` for `xwl_tablet_tool` 4737 4738 So we do not lose subpixel precision in Xwayland. 4739 4740 Suggested-by: Peter Hutterer <peter.hutterer@who-t.net> 4741 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 4742 Closes: https://gitlab.freedesktop.org/libinput/libinput/issues/138 4743 (cherry picked from commit 734b2d6907f730571a2805cbc53fe7056190f19e) 4744 4745commit fb01b238c610e1955457dd9878d75b2c34c9a0c6 4746Author: Adam Jackson <ajax@redhat.com> 4747Date: Wed Aug 29 15:42:20 2018 -0400 4748 4749 xfree86: Fix Option "MaxClients" validation 4750 4751 The old code would not in fact validate the option value, though it 4752 might complain about it in the log. It also didn't let you set some 4753 legal values that the -maxclients command line option would. 4754 4755 Signed-off-by: Adam Jackson <ajax@redhat.com> 4756 (cherry picked from commit 7d689f049c3cc16b8e0cb0103a384a2ceb84ea33) 4757 4758commit cc4051ad6a3712c4407ffb608c8d23f2b66952d6 4759Author: Dave Airlie <airlied@redhat.com> 4760Date: Wed Sep 12 11:39:32 2018 +1000 4761 4762 devices: break after finding and removing device from lists 4763 4764 Coverity complains about a use after free in here after the 4765 freeing, I can't follow the linked list so well, but whot 4766 says the device can only be on one list once, so break should 4767 fix it. 4768 4769 Signed-off-by: Dave Airlie <airlied@redhat.com> 4770 (cherry picked from commit ba0f5d854f4db52974ab5cd09191303b01b075ff) 4771 4772commit 64a7aac257ddbb000759d37386454866d8810545 4773Author: Dave Airlie <airlied@redhat.com> 4774Date: Wed Sep 12 11:12:51 2018 +1000 4775 4776 mibltblt: free prgnSrcClip on error path. 4777 4778 Pointed out by coverity. 4779 4780 Signed-off-by: Dave Airlie <airlied@redhat.com> 4781 (cherry picked from commit c4591ea17110b5c318a5ef4b0f17a4eea306ea71) 4782 4783commit fcbdb7c8b0d7bcc6a85e0beb2ce1dba76803486b 4784Author: Dave Airlie <airlied@redhat.com> 4785Date: Wed Sep 12 11:09:40 2018 +1000 4786 4787 xkb: fix what looks to be a copy-paste error with first vs firstMM 4788 4789 Pointed out by coverity. 4790 4791 Signed-off-by: Dave Airlie <airlied@redhat.com> 4792 (cherry picked from commit 758393951233d1b2520cf4cefd33ec4288a3880a) 4793 4794commit 83ef02839ce8982c2e841bc400abddbeb620481d 4795Author: Dave Airlie <airlied@redhat.com> 4796Date: Wed Sep 12 11:05:45 2018 +1000 4797 4798 posix_tty: free leak of xf86SetStrOption return value. 4799 4800 Pointed out by coverity. 4801 4802 Signed-off-by: Dave Airlie <airlied@redhat.com> 4803 (cherry picked from commit cad3a1a82da3c8421b5cc98af27a779a38b5c709) 4804 4805commit 66d36010a324480ac656e702e53f9c1f6fb60d5c 4806Author: Dave Airlie <airlied@redhat.com> 4807Date: Wed Sep 12 10:56:40 2018 +1000 4808 4809 modesetting: get pEnt after error checks 4810 4811 This saves us having to make sure we clean it up. 4812 4813 Pointed out by coverity. 4814 4815 Signed-off-by: Dave Airlie <airlied@redhat.com> 4816 (cherry picked from commit b6c29a881eb49300fddfd0187cb10891f5257d3b) 4817 4818commit 795c58a1febb3deb9066d981b178d9cf14fdad03 4819Author: Dave Airlie <airlied@redhat.com> 4820Date: Wed Sep 12 10:52:25 2018 +1000 4821 4822 glamor: fix leak of fs_getcolor_source. 4823 4824 This is created using XNFstrdup, so it needs to be freed. 4825 4826 Pointed out by coverity. 4827 Signed-off-by: Dave Airlie <airlied@redhat.com> 4828 (cherry picked from commit f0a5c0d1fdaeee3cd701215f4f57b7eacaf783c2) 4829 4830commit bb384d0b110a99f2d935ed73d6589af46f818823 4831Author: Dave Airlie <airlied@redhat.com> 4832Date: Wed Sep 12 10:50:21 2018 +1000 4833 4834 fboverlay: move bpp checks above malloc 4835 4836 Avoids having to free the malloced object. 4837 4838 Pointed out by coverity. 4839 4840 Signed-off-by: Dave Airlie <airlied@redhat.com> 4841 (cherry picked from commit c7fa6a0a0d4ad28ee21b8dd9ad4fbe33f462bc71) 4842 4843commit 26a83f9833054f5cf7c7d73b0f5aa080d400c26b 4844Author: Dave Airlie <airlied@redhat.com> 4845Date: Wed Sep 12 10:33:04 2018 +1000 4846 4847 xi: free modifiers_failed on error path. (v2) 4848 4849 Pointed out by coverity. 4850 4851 v2: set modifies_failed to NULL at start (whot) 4852 4853 Signed-off-by: Dave Airlie <airlied@redhat.com> 4854 (cherry picked from commit 51ae6126dcf3e234d0f678f02934bc9515abb8ae) 4855 4856commit bdeab7863eb7abe2d9d05eed259542695935559c 4857Author: Dave Airlie <airlied@redhat.com> 4858Date: Wed Sep 12 10:30:13 2018 +1000 4859 4860 shm: move shmsize verify before allocating the drawable. 4861 4862 Otherwise if the VERIFY_SHMSIZE macro fails we leak the drawables 4863 we allocated earlier. 4864 4865 Noticed by coverity scan. 4866 4867 Signed-off-by: Dave Airlie <airlied@redhat.com> 4868 (cherry picked from commit 3439929c510501929c6ef1d90477c33bf8838632) 4869 4870commit 8e646ca9850c071994354b503b5d2b62f99ef2a8 4871Author: Adam Jackson <ajax@redhat.com> 4872Date: Wed Aug 29 16:07:59 2018 -0400 4873 4874 modesetting: Document Option "DoubleShadow" in the man page 4875 4876 Signed-off-by: Adam Jackson <ajax@redhat.com> 4877 (cherry picked from commit c4f3e42fe316c90a73908a838ce64abbb28fe0a2) 4878 4879commit fdb80a327cfe213a206bc76d3cb43cbb33855547 4880Author: Adam Jackson <ajax@redhat.com> 4881Date: Wed Aug 29 15:57:46 2018 -0400 4882 4883 modesetting: Lie less in the man page 4884 4885 We don't support 8bpp, and we do have acceleration. 4886 4887 Signed-off-by: Adam Jackson <ajax@redhat.com> 4888 (cherry picked from commit 0dc2c419e11cf28b8fa1e607d4cf2d15f2eb2306) 4889 4890commit cffac815b957fd1296d61cc5c20ba3709a77ee4e 4891Author: Olivier Fourdan <ofourdan@redhat.com> 4892Date: Wed Sep 5 10:49:27 2018 +0200 4893 4894 xwayland: Remove xwl_present_window from privates on cleanup 4895 4896 Xwayland's `xwl_destroy_window()` invokes `xwl_present_cleanup()` 4897 before the common `DestroyWindow()`. 4898 4899 But then `DestroyWindow()` calls `present_destroy_window()` which will 4900 possibly end up in `xwl_present_abort_vblank()` which will try to access 4901 data that was previously freed by `xwl_present_cleanup()`: 4902 4903 Invalid read of size 8 4904 at 0x434184: xwl_present_abort_vblank (xwayland-present.c:378) 4905 by 0x53785B: present_wnmd_abort_vblank (present_wnmd.c:651) 4906 by 0x53695A: present_free_window_vblank (present_screen.c:87) 4907 by 0x53695A: present_destroy_window (present_screen.c:152) 4908 by 0x42A90D: xwl_destroy_window (xwayland.c:653) 4909 by 0x584298: compDestroyWindow (compwindow.c:613) 4910 by 0x53CEE3: damageDestroyWindow (damage.c:1570) 4911 by 0x4F1BB8: DbeDestroyWindow (dbe.c:1326) 4912 by 0x46F7F6: FreeWindowResources (window.c:1031) 4913 by 0x472847: DeleteWindow (window.c:1099) 4914 by 0x46B54C: doFreeResource (resource.c:880) 4915 by 0x46C706: FreeClientResources (resource.c:1146) 4916 by 0x446ADE: CloseDownClient (dispatch.c:3473) 4917 Address 0x182abde0 is 80 bytes inside a block of size 112 free'd 4918 at 0x4C2FDAC: free (vg_replace_malloc.c:530) 4919 by 0x42A937: xwl_destroy_window (xwayland.c:647) 4920 by 0x584298: compDestroyWindow (compwindow.c:613) 4921 by 0x53CEE3: damageDestroyWindow (damage.c:1570) 4922 by 0x4F1BB8: DbeDestroyWindow (dbe.c:1326) 4923 by 0x46F7F6: FreeWindowResources (window.c:1031) 4924 by 0x472847: DeleteWindow (window.c:1099) 4925 by 0x46B54C: doFreeResource (resource.c:880) 4926 by 0x46C706: FreeClientResources (resource.c:1146) 4927 by 0x446ADE: CloseDownClient (dispatch.c:3473) 4928 by 0x446DA5: ProcKillClient (dispatch.c:3279) 4929 by 0x4476AF: Dispatch (dispatch.c:479) 4930 Block was alloc'd at 4931 at 0x4C30B06: calloc (vg_replace_malloc.c:711) 4932 by 0x433F46: xwl_present_window_get_priv (xwayland-present.c:54) 4933 by 0x434228: xwl_present_get_crtc (xwayland-present.c:302) 4934 by 0x539728: proc_present_query_capabilities (present_request.c:227) 4935 by 0x4476AF: Dispatch (dispatch.c:479) 4936 by 0x44B5B5: dix_main (main.c:276) 4937 by 0x75F611A: (below main) (libc-start.c:308) 4938 4939 This is because `xwl_present_cleanup()` frees the memory but does not 4940 remove it from the window's privates, and `xwl_present_abort_vblank()` 4941 will still find it and hence try to access that freed memory... 4942 4943 Remove `xwl_present_window` from window's privates on cleanup so that no 4944 other function can find and reuse that data once it's freed. 4945 4946 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1616269 4947 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 4948 (cherry picked from commit 3f31f56929e80001970e3821ed8b10c6075df8e6) 4949 4950commit 8dd7173eeba08f1ecfb414915625c609ad4b3297 4951Author: Lionel Landwerlin <lionel.g.landwerlin@intel.com> 4952Date: Tue Aug 28 21:30:05 2018 +0100 4953 4954 xwayland: fix access to invalid pointer 4955 4956 xwl_output->randr_crtc is used in the update_screen_size() function : 4957 4958 ==5331== Invalid read of size 4 4959 ==5331== at 0x15263D: update_screen_size (xwayland-output.c:190) 4960 ==5331== by 0x152C48: xwl_output_remove (xwayland-output.c:413) 4961 ==5331== by 0x6570FCD: ffi_call_unix64 (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 4962 ==5331== by 0x657093E: ffi_call (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 4963 ==5331== by 0x4DDB183: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 4964 ==5331== by 0x4DD79D8: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 4965 ==5331== by 0x4DD8EA3: wl_display_dispatch_queue_pending (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 4966 ==5331== by 0x14BCCA: xwl_read_events (xwayland.c:814) 4967 ==5331== by 0x2AC0D0: ospoll_wait (ospoll.c:651) 4968 ==5331== by 0x2A5322: WaitForSomething (WaitFor.c:208) 4969 ==5331== by 0x27574B: Dispatch (dispatch.c:421) 4970 ==5331== by 0x279945: dix_main (main.c:276) 4971 ==5331== Address 0x1aacb5f4 is 36 bytes inside a block of size 154 free'd 4972 ==5331== at 0x48369EB: free (vg_replace_malloc.c:530) 4973 ==5331== by 0x1F8AE8: RROutputDestroyResource (rroutput.c:421) 4974 ==5331== by 0x29A2AC: doFreeResource (resource.c:880) 4975 ==5331== by 0x29AE5B: FreeResource (resource.c:910) 4976 ==5331== by 0x152BE0: xwl_output_remove (xwayland-output.c:408) 4977 ==5331== by 0x6570FCD: ffi_call_unix64 (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 4978 ==5331== by 0x657093E: ffi_call (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 4979 ==5331== by 0x4DDB183: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 4980 ==5331== by 0x4DD79D8: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 4981 ==5331== by 0x4DD8EA3: wl_display_dispatch_queue_pending (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 4982 ==5331== by 0x14BCCA: xwl_read_events (xwayland.c:814) 4983 ==5331== by 0x2AC0D0: ospoll_wait (ospoll.c:651) 4984 ==5331== Block was alloc'd at 4985 ==5331== at 0x48357BF: malloc (vg_replace_malloc.c:299) 4986 ==5331== by 0x1F93E0: RROutputCreate (rroutput.c:83) 4987 ==5331== by 0x152A75: xwl_output_create (xwayland-output.c:361) 4988 ==5331== by 0x14BE59: registry_global (xwayland.c:764) 4989 ==5331== by 0x6570FCD: ffi_call_unix64 (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 4990 ==5331== by 0x657093E: ffi_call (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 4991 ==5331== by 0x4DDB183: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 4992 ==5331== by 0x4DD79D8: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 4993 ==5331== by 0x4DD8EA3: wl_display_dispatch_queue_pending (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 4994 ==5331== by 0x14BCCA: xwl_read_events (xwayland.c:814) 4995 ==5331== by 0x2AC0D0: ospoll_wait (ospoll.c:651) 4996 ==5331== by 0x2A5322: WaitForSomething (WaitFor.c:208) 4997 4998 Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> 4999 Reviewed-by: Daniel Stone <daniels@collabora.com> 5000 (cherry picked from commit 53ce2ba0a19af9c549f47a4cc678afcebeb6087e) 5001 5002commit 1191b23f942cace785234f5d7fac972251c8f1db 5003Author: Olivier Fourdan <ofourdan@redhat.com> 5004Date: Wed Sep 5 15:20:17 2018 +0200 5005 5006 glx: check for indirect context in CreateContextAttribsARB() 5007 5008 Commit 99f0365b "Add a command line argument for disabling indirect GLX" 5009 added a test to check if indirect context are enabled in 5010 `DoCreateContext()` but `__glXDisp_CreateContextAttribsARB()` doesn't 5011 use `DoCreateContext()` and doesn't check if indirect context is 5012 enabled. 5013 5014 As a result, clients can still manage to create indirect contexts using 5015 `glXCreateContextAttribsARB()` even if indirect contexts are disabled, 5016 which can possibly crash Xservers such as Xwayland or Xephyr when the 5017 context is destroyed. 5018 5019 To avoid the issue, check for `enableIndirectGLX` in 5020 `__glXDisp_CreateContextAttribsARB()` as well. 5021 5022 Fixes: 99f0365b "Add a command line argument for disabling indirect GLX" 5023 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107508 5024 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5025 (cherry picked from commit 75448671abe2c6ae3745eb5d2ed2e76df2de9c41) 5026 5027commit 1b0db2c74258d20e3f99bd69c2914fd445abe920 5028Author: Lionel Landwerlin <lionel.g.landwerlin@intel.com> 5029Date: Mon Aug 27 01:04:45 2018 +0100 5030 5031 present: fix freed pointer access 5032 5033 When a vblank has been marked as aborted, it's going to be free in the 5034 flip_notify function when stopped. We can't notify it after it's 5035 stopped because the pointer is invalid. 5036 5037 Valgrind backtrace: 5038 5039 ==5331== Invalid read of size 8 5040 ==5331== at 0x212B4D: present_vblank_notify (present_vblank.c:34) 5041 ==5331== by 0x21439B: present_wnmd_flip_notify (present_wnmd.c:194) 5042 ==5331== by 0x21439B: present_wnmd_event_notify (present_wnmd.c:228) 5043 ==5331== by 0x156216: xwl_present_sync_callback (xwayland-present.c:282) 5044 ==5331== by 0x6570FCD: ffi_call_unix64 (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 5045 ==5331== by 0x657093E: ffi_call (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 5046 ==5331== by 0x4DDB183: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 5047 ==5331== by 0x4DD79D8: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 5048 ==5331== by 0x4DD8EA3: wl_display_dispatch_queue_pending (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 5049 ==5331== by 0x14BCCA: xwl_read_events (xwayland.c:814) 5050 ==5331== by 0x2AC0D0: ospoll_wait (ospoll.c:651) 5051 ==5331== by 0x2A5322: WaitForSomething (WaitFor.c:208) 5052 ==5331== by 0x27574B: Dispatch (dispatch.c:421) 5053 ==5331== Address 0x1b44dc98 is 40 bytes inside a block of size 184 free'd 5054 ==5331== at 0x48369EB: free (vg_replace_malloc.c:530) 5055 ==5331== by 0x213B0A: present_wnmd_free_idle_vblanks (present_wnmd.c:118) 5056 ==5331== by 0x213B0A: present_wnmd_flips_stop (present_wnmd.c:161) 5057 ==5331== by 0x2143EF: present_wnmd_flip_notify (present_wnmd.c:192) 5058 ==5331== by 0x2143EF: present_wnmd_event_notify (present_wnmd.c:228) 5059 ==5331== by 0x156216: xwl_present_sync_callback (xwayland-present.c:282) 5060 ==5331== by 0x6570FCD: ffi_call_unix64 (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 5061 ==5331== by 0x657093E: ffi_call (in /usr/lib/x86_64-linux-gnu/libffi.so.6.0.4) 5062 ==5331== by 0x4DDB183: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 5063 ==5331== by 0x4DD79D8: ??? (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 5064 ==5331== by 0x4DD8EA3: wl_display_dispatch_queue_pending (in /usr/lib/x86_64-linux-gnu/libwayland-client.so.0.3.0) 5065 ==5331== by 0x14BCCA: xwl_read_events (xwayland.c:814) 5066 ==5331== by 0x2AC0D0: ospoll_wait (ospoll.c:651) 5067 ==5331== by 0x2A5322: WaitForSomething (WaitFor.c:208) 5068 ==5331== Block was alloc'd at 5069 ==5331== at 0x48377D5: calloc (vg_replace_malloc.c:711) 5070 ==5331== by 0x212D9F: present_vblank_create (present_vblank.c:69) 5071 ==5331== by 0x214014: present_wnmd_pixmap (present_wnmd.c:610) 5072 ==5331== by 0x21576C: proc_present_pixmap (present_request.c:150) 5073 ==5331== by 0x27599D: Dispatch (dispatch.c:479) 5074 ==5331== by 0x279945: dix_main (main.c:276) 5075 ==5331== by 0x633AB16: (below main) (libc-start.c:310) 5076 5077 v2: Still notify aborted flips (Roman) 5078 5079 Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> 5080 Reviewed-by: Daniel Stone <daniels@collabora.com> 5081 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107314 5082 Reviewed-by: Roman Gilg <subdiff@gmail.com> 5083 Tested-by: Roman Gilg <subdiff@gmail.com> 5084 (cherry picked from commit ce271535adb6974e0a43bb64c8ed7a5dcaff67a2) 5085 5086commit 48300a7775b4ff3200de200e8c5502d99bd99104 5087Author: Scott Anderson <scott@anderso.nz> 5088Date: Mon Aug 6 18:09:26 2018 +1200 5089 5090 xwayland: use wayland axis_discrete event 5091 5092 This prevents multiple scroll events happening for wayland compositors 5093 which send axis values other than 10. For example, libinput will 5094 typically return 15 for each scroll wheel step, and if a wayland 5095 compositor sends those to xwayland without normalising them, 2 scroll 5096 wheel steps will end up as 3 xorg scroll events. By listening for the 5097 discrete_axis event, this will now correctly send only 2 xorg scroll 5098 events. 5099 5100 The wayland protocol gurantees that there will always be an axis event 5101 following an axis_discrete event. However, it does not gurantee that 5102 other events (including other axis_discrete+axis pairs) will not happen 5103 in between them. So we must keep a list of outstanding axis_discrete 5104 events. 5105 5106 Signed-off-by: Scott Anderson <scott@anderso.nz> 5107 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 5108 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 5109 (cherry picked from commit cd285922cdec966825e47220b1182a57abc1ff90) 5110 5111commit cd19a752f8bbc6caae7bf2457c53c850121fe9b7 5112Author: Jim Qu <Jim.Qu@amd.com> 5113Date: Mon Aug 27 13:37:38 2018 +0800 5114 5115 modesetting: code refactor for PRIME sync 5116 5117 The X will be crashed on the system with other DDX driver, 5118 such as amdgpu. 5119 5120 show the log like: 5121 5122 randr: falling back to unsynchronized pixmap sharing 5123 (EE) 5124 (EE) Backtrace: 5125 (EE) 0: /usr/lib/xorg/Xorg (xorg_backtrace+0x4e) 5126 (EE) 1: /usr/lib/xorg/Xorg (0x55cb0151a000+0x1b5ce9) 5127 (EE) 2: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7f1587a1d000+0x11390) 5128 (EE) 5129 (EE) Segmentation fault at address 0x0 5130 (EE) 5131 5132 The issue is that modesetting as the master, and amdgpu as the slave. 5133 Thus, when the master attempts to access pSlavePixPriv in ms_dirty_update(), 5134 problems result due to the fact that it's accessing AMD's 'ppriv' using the 5135 modesetting structure definition. 5136 5137 Apart from fixing crash issue, the patch fix other issue in master interface 5138 in which driver should refer to master pixmap. 5139 5140 Signed-off-by: Jim Qu <Jim.Qu@amd.com> 5141 Reviewed-by: Alex Goins <agoins@nvidia.com> 5142 (cherry picked from commit f79e5368512b72bb463925983d265b070261b7aa) 5143 5144commit 5396a4aa084853805a73a2cfbbce156c47573916 5145Author: Alex Goins <agoins@nvidia.com> 5146Date: Tue Aug 14 15:05:46 2018 -0500 5147 5148 randr: rrCheckPixmapBounding should only increase screen size 5149 5150 The purpose of rrCheckPixmapBounding() is to make sure that the fb is large 5151 enough to accommodate the region scanned out by a GPU screen. Currently, however, 5152 it will actually shrink the fb if it's larger than it needs to be. 5153 5154 This is a problem when combining PRIME output slaving with arbitrary transforms 5155 with xrandr. 5156 5157 Although arbitrary transforms are not supposed to constrain the size of the fb 5158 (https://lists.freedesktop.org/archives/xorg-devel/2018-January/055563.html), 5159 xrandr will use RRSetScreenSize to resize the desktop to accommodate scaling 5160 transforms, e.g. scaling a 1920x1080 display to 3840x2160 will result in a 5161 desktop size of 3840x2160. 5162 5163 In the case of PRIME, rrCheckPixmapBounding() will be called after 5164 RRSetScreenSize() and it will resize the fb back down to what it would be 5165 without the scaling transform, e.g. 1920x1080. This represents divergence in 5166 behavior between PRIME and non-PRIME outputs. 5167 5168 I had originally made rrCheckPixmapBounding() account for arbitrary transforms, 5169 but realized that the fb being large enough to accommodate arbitrary transforms 5170 is not a hard requirement enforced in the server. Instead, this change simply 5171 makes it so that rrCheckPixmapBounding() will only resize the fb to be larger 5172 than it already is, preventing it from stepping on prior requests to increase 5173 the size of the fb. 5174 5175 Signed-off-by: Alex Goins <agoins@nvidia.com> 5176 Reviewed-by: Keith Packard <keithp@keithp.com> 5177 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 5178 (cherry picked from commit a90f33721eba7f2dbde4a7278f1a213d696c85e9) 5179 5180commit 051a0efc5c28fb4be1199d73d59e0e4d329b1144 5181Author: Peter Hutterer <peter.hutterer@who-t.net> 5182Date: Wed Aug 8 15:22:41 2018 +1000 5183 5184 dix: check_modmap_change() returns Success, not true 5185 5186 Not sure what if anything calls XSetDeviceModifierMapping() but this would've 5187 failed all the time. check_modmap_change() returns Success but we were 5188 treating it like a boolean. Fix this. 5189 5190 Reported-by: Adam Jackson <ajax@redhat.com> 5191 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 5192 Reviewed-by: Keith Packard <keithp@keithp.com> 5193 (cherry picked from commit 4fe02b8da3ca58500f7e6e017e388907cf4a8b73) 5194 5195commit 1508ea68065b7976dbf56039f9b20dcaf0394431 5196Author: Adam Jackson <ajax@redhat.com> 5197Date: Tue Aug 7 12:26:05 2018 -0400 5198 5199 xserver 1.20.1 5200 5201 Signed-off-by: Adam Jackson <ajax@redhat.com> 5202 5203commit 101d15c7634b26eafb3c5f2edb7cafde102ce208 5204Author: Keith Packard <keithp@keithp.com> 5205Date: Tue Jun 26 09:20:00 2018 -0700 5206 5207 During reset/shutdown, clean up leases in DIX instead of each driver 5208 5209 Instead of having every video driver loop over any pending leases to 5210 free them during CloseScreen, do this up in the DIX layer by 5211 terminating leases when a leased CRTC or Output is destroyed and 5212 (just to make sure), also terminating leases in RRCloseScreen. The 5213 latter should "never" get invoked as any lease should be associated 5214 with a resource which was destroyed. 5215 5216 This is required as by the time the driver's CloseScreen function is 5217 invoked, we've already freed all of the DIX randr structures and no 5218 longer have any way to reference the leases 5219 5220 Signed-off-by: Keith Packard <keithp@keithp.com> 5221 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=106960 5222 Cc: Thomas Hellstrom <thellstrom@vmware.com> 5223 (cherry picked from commit 1ef7aed3e2bb2af32330f19b1e7560000512ddfe) 5224 5225commit 9347326d28fffc7534cad0b084539e936aacfd45 5226Author: Peter Hutterer <peter.hutterer@who-t.net> 5227Date: Wed Jul 18 13:22:43 2018 +1000 5228 5229 Xext: dynamically allocate the PanoramiXDepths[j].vids array 5230 5231 Control flow is: 5232 PanoramiXMaybeAddDepth() allocates an array size 240 (pDepth->numVisuals) 5233 PanoramiXMaybeAddVisual() finds up to 270 matches (pScreen->numVisuals) 5234 and writes those into the previously allocated array. 5235 5236 This caused invalid reads/writes followed by eventually a double-free abort. 5237 5238 Reproduced with xorg-integration-tests server test 5239 XineramaTest.ScreenCrossing/* (and a bunch of others). 5240 5241 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 5242 Reviewed-by: Keith Packard <keithp@keithp.com> 5243 (cherry picked from commit 93cafb0828d2e24bd14616df1aa9883fb843dd6c) 5244 5245commit cbf1ca2dba7bc3561cf1a8023e5e18706adbdba6 5246Author: emersion <contact@emersion.fr> 5247Date: Fri Jul 13 15:51:26 2018 -0400 5248 5249 xwayland: rotate logical size for RRMode 5250 5251 The logical size is the size of the output in the global compositor 5252 space. The mode width/height should be scaled as in the logical 5253 size, but shouldn't be transformed. Thus we need to rotate back 5254 the logical size to be able to use it as the mode width/height. 5255 5256 This fixes issues with pointer input on transformed outputs. 5257 5258 Signed-Off-By: Simon Ser <contact@emersion.fr> 5259 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 5260 (cherry picked from commit ce2dde9ed0243a18ae18af0879134f7c1afbd700) 5261 5262commit 18a52a8e16046d8cabea6cf45913bf0b5be07709 5263Author: Olivier Fourdan <ofourdan@redhat.com> 5264Date: Wed Jul 25 16:33:23 2018 +0200 5265 5266 xwayland: Enable DRI3 for glamor 5267 5268 glamor_fds_from_pixmap() will bail out early if DRI3 is not enabled, 5269 unfortunately Xwayland's glamor code would not set it as enabled which 5270 would lead to blank pixmaps when using texture from pixmap. 5271 5272 Make sure to mark DRI3 as enabled from glamor_egl_screen_init() in 5273 Xwayland. 5274 5275 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107287 5276 Fixes: c8c276c956 ("glamor: Implement PixmapFromBuffers and BuffersFromPixmap") 5277 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5278 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 5279 (cherry picked from commit cdec2b3c195d1d080207ef01c55ff14b45370010) 5280 5281commit c256e31a9eba20da259f31ee70d1c8e1870993f1 5282Author: Takashi Iwai <tiwai@suse.de> 5283Date: Thu Jul 19 14:38:19 2018 +0200 5284 5285 modesetting: Fix cirrus 24bpp breakage 5286 5287 The recent rewrite of modesetting driver broke the 24bpp support. 5288 As typically found on cirrus KMS, it leads to a blank screen, spewing 5289 the error like: 5290 failed to add fb -22 5291 (EE) modeset(0): failed to set mode: Invalid argument 5292 5293 The culript is that the wrong bpp value of the front buffer is passed 5294 to drmModeAddFB(). Fix it by replacing with the back buffer bpp, 5295 drmmode->kbpp. 5296 5297 Signed-off-by: Takashi Iwai <tiwai@suse.de> 5298 Tested-by: Stefan Dirsch <sndirsch@suse.de> 5299 Reviewed-by: Adam Jackson <ajax@redhat.com> 5300 (cherry picked from commit d625e16918ef9104863709eb108346464767c444) 5301 5302commit 2da0bde4498cf4ee422e5a2cb2c53cbbb73df937 5303Author: Matt Turner <mattst88@gmail.com> 5304Date: Tue May 22 21:10:55 2018 -0700 5305 5306 xfree86: Inline xf86{Read,Write}Mmio{8,16,32} on alpha 5307 5308 In commit 9db2af6f757e (xfree86: Remove xf86{Map,Unmap}VidMem) we 5309 somehow stopped exporting xf86{Read,Write}Mmio{8,16,32}. Since the 5310 function pointer indirection was intended to support dense vs sparse and 5311 sparse support is now gone, we can just make the functions static inline 5312 in compiler.h and avoid all of this. 5313 5314 Bugzilla: https://bugs.gentoo.org/548906 5315 Tested-by: Christopher May-Townsend <chris@maytownsend.co.uk> 5316 Reviewed-by: Adam Jackson <ajax@redhat.com> 5317 Signed-off-by: Matt Turner <mattst88@gmail.com> 5318 (cherry picked from commit 166ac294aefe21a6cfcc0e7a4c9525eef49c2f0c) 5319 5320commit b9e9eda08bd9198a6d089acd5d50677cd8713b33 5321Author: Stefan Agner <stefan@agner.ch> 5322Date: Thu Dec 22 15:41:06 2016 +0100 5323 5324 modesetting: Fix 16 bit depth/bpp mode 5325 5326 When setting DefaultDepth to 16 in the Screen section, the current 5327 code requests a 32 bpp framebuffer, however the X-Server seems to 5328 assumes 16 bpp. 5329 5330 Fixes commit 21217d02168d ("modesetting: Implement 32->24 bpp 5331 conversion in shadow update") 5332 5333 Reviewed-by: Adam Jackson <ajax@redhat.com> 5334 Signed-off-by: Stefan Agner <stefan@agner.ch> 5335 (cherry picked from commit 1c7f34e99ff9750979a03ae20c6be1f2b42c284c) 5336 5337commit 5c0662d44852363fe258c045250710ed881e56b7 5338Author: Lyude Paul <lyude@redhat.com> 5339Date: Fri Jun 22 12:49:47 2018 -0400 5340 5341 meson: ensure the libc has RPC functions when secure-rpc is enabled 5342 5343 Currently our meson.build just makes the assumption that the libc is 5344 going to provide RPC functions. This doesn't actually seem to be the 5345 case on Fedora, which causes compilation to fail unexpectedly: 5346 5347 ../../Projects/xserver/os/rpcauth.c:47:10: fatal error: rpc/rpc.h: No such file or directory 5348 #include <rpc/rpc.h> 5349 ^~~~~~~~~~~ 5350 compilation terminated. 5351 5352 So, in the event that we can't use libtirpc ensure that we actually 5353 check whether or not the libc provides rpc/rpc.h. If it doesn't, raise 5354 an error. 5355 5356 Reviewed-by: Adam Jackson <ajax@redhat.com> 5357 Signed-off-by: Lyude Paul <lyude@redhat.com> 5358 (cherry picked from commit d95a1310ef8e08a93a28f9766d1b4093f7891404) 5359 5360commit 10285bc36b411edd923ee5d369244647ba2b714b 5361Author: Keith Packard <keithp@keithp.com> 5362Date: Tue Jun 26 16:34:29 2018 -0700 5363 5364 xf86-video-modesetting: Lease planes as well if using atomic 5365 5366 If we're using atomic modesetting, then we're also using universal 5367 planes, and so the lease we create needs to include the plane. 5368 5369 Signed-off-by: Keith Packard <keithp@keithp.com> 5370 Reviewed-by: Daniel Stone <daniels@collabora.com> 5371 (cherry picked from commit d83efc47b7a524b4f8d4a993c27a3e402a98fa7c) 5372 5373commit a530198ac0ffdc4c268d13e59079f7583930a284 5374Author: Keith Packard <keithp@keithp.com> 5375Date: Tue Jun 26 14:05:31 2018 -0700 5376 5377 xf86-video-modesetting: Don't enable UNIVERSAL_PLANES separately 5378 5379 We don't want universal_planes unless we're using atomic APIs for 5380 modesetting, and the kernel already enables universal_planes 5381 automatically when atomic is enabled. 5382 5383 If we enable universal_planes when we're not using atomic, then we 5384 won't have selected a plane for each crtc, and this will break lease 5385 creation which requires planes for each output when universal_planes 5386 is enabled. 5387 5388 Signed-off-by: Keith Packard <keithp@keithp.com> 5389 Reviewed-by: Daniel Stone <daniels@collabora.com> 5390 (cherry picked from commit 4a11f66e4690f9e94dc61cb264f5ef78dbdb255a) 5391 5392commit cd7680adccdf8f6905a83ae08cbde7ac76693fe5 5393Author: Keith Packard <keithp@keithp.com> 5394Date: Thu Jun 28 11:45:16 2018 -0700 5395 5396 xfree86: Wrap RRCrtcIsLeased and RROutputIsLeased to check for DIX structures 5397 5398 Before DIX structures are allocated for crtcs and outputs, we don't 5399 want to call DIX randr code with NULL pointers. This can happen if the 5400 driver sets video modes early in server initialization, which Nouveau 5401 does in zaphod mode. 5402 5403 Cc: thellstrom@vmware.com 5404 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=106772 5405 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=106960 5406 Reviewed-by: Adam Jackson <ajax@redhat.com> 5407 Signed-off-by: Keith Packard <keithp@keithp.com> 5408 (cherry picked from commit 2faf4cef8bcf9bb2034a27219a656ea7221afc6c) 5409 5410commit ebd4cd71f49151fb699cf45e835f3719e6b31bf5 5411Author: Keith Packard <keithp@keithp.com> 5412Date: Thu Jun 28 11:45:15 2018 -0700 5413 5414 xfree86: Reset randr_crtc and randr_output early in xf86CrtcCloseScreen 5415 5416 The DIX crtc and output structures are freed when their resources are 5417 destroyed, which happens before CloseScreen is called. As a result, we 5418 know these pointers are invalid and referencing them during any of the 5419 remaining CloseScreen sequence will be bad. 5420 5421 Reviewed-by: Adam Jackson <ajax@redhat.com> 5422 Signed-off-by: Keith Packard <keithp@keithp.com> 5423 Cc: thellstrom@vmware.com 5424 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=106960 5425 (cherry picked from commit c55a44a9a86aaece17c1a2e73c77e3e665c4888e) 5426 5427commit 1f169d5b38d93d725029d5f7dc237b558e8bb137 5428Author: Jon Turney <jon.turney@dronecode.org.uk> 5429Date: Sat Jun 30 12:53:35 2018 +0100 5430 5431 meson: use absolute paths in manpage substitutions 5432 5433 paths returned by get_option('foodir') are potentially relative to prefix 5434 5435 Noticed when comparing manpages generated by a meson build with those 5436 generated by an autotools build 5437 5438 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 5439 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 5440 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 5441 (cherry picked from commit 49283e238a0ba6051034ae635e5970891f17f9df) 5442 5443commit bc1882aa23a129c4b72b0bfc132cfd04db801a90 5444Author: Keith Packard <keithp@keithp.com> 5445Date: Thu Jan 18 18:07:29 2018 -0800 5446 5447 modesetting: Allow a DRM fd to be passed on command line with -masterfd [v2] 5448 5449 This lets an application open a suitable DRM device and pass the file 5450 descriptor to the mode setting driver through an X server command line 5451 option, '-masterfd'. 5452 5453 There's a companion application, xlease, which creates a DRM master by 5454 leasing an output from another X server. That is available at 5455 5456 git clone git://people.freedesktop.org/~keithp/xlease 5457 5458 v2: 5459 Always print usage, but note that it can't be used if 5460 setuid/gid 5461 5462 Suggested-by: Lyude Paul <lyude@redhat.com> 5463 5464 Signed-off-by: Keith Packard <keithp@keithp.com> 5465 Reviewed-by: Lyude Paul <lyude@redhat.com> 5466 (cherry picked from commit 38ff29ec8ead0a293f50ac52146199babe4205ca) 5467 5468commit d60ce5b01f5988abc9957473d7175872c8845b08 5469Author: Laurent Carlier <lordheavym@gmail.com> 5470Date: Sat Jun 16 13:00:01 2018 +0200 5471 5472 meson: Add configuration of listening on tcp, unix and local 5473 5474 bugzilla: https://bugs.kde.org/show_bug.cgi?id=395419 5475 bugzilla: https://bugs.archlinux.org/task/59025 5476 5477 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 5478 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 5479 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 5480 (cherry picked from commit 2f39b2a07805194fdd7def30d941666963d6e670) 5481 5482commit 79795bf9df395001c8d659ac8d28090ea9cefc8d 5483Author: Lyude Paul <lyude@redhat.com> 5484Date: Wed Jun 27 20:29:42 2018 -0400 5485 5486 modesetting: Fix uninitialized memory usage in drmmode_crtc_get_fb_id() 5487 5488 This really sucked to find out :( 5489 5490 Signed-off-by: Lyude Paul <lyude@redhat.com> 5491 Reviewed-by: Karol Herbst <kherbst@redhat.com> 5492 (cherry picked from commit c41d4ff48f72aa964afd131b59e1538295d062dc) 5493 5494commit 820ce7cb8bb07018de89e13aa50ae6090f2d84eb 5495Author: Thomas Hellstrom <thellstrom@vmware.com> 5496Date: Wed Jun 20 19:23:48 2018 +0200 5497 5498 glamor: Work around GEM usage v2 5499 5500 KMS drivers are not required to support GEM. In particular, vmwgfx 5501 doesn't support flink and handles and names are identical. 5502 Getting a bo name should really be part of a lower level API, if needed, 5503 but in the mean time work around this by setting the name identical to 5504 the handle if GEM isn't supported. 5505 5506 This fixes modesetting driver dri2 on vmwgfx. 5507 5508 Reviewed-by: Deepak Rawat <drawat@vmware.com> 5509 Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> 5510 (cherry picked from commit 9f02855e7a1b7a3c1e2ee7bfbc73e87c29126920) 5511 5512commit ba6a9283817254dca90fd8f9ee695504e7c20908 5513Author: Lyude Paul <lyude@redhat.com> 5514Date: Wed Jun 20 19:12:32 2018 -0400 5515 5516 randr: Scream when creating a shared pixmap fails 5517 5518 This seems like a problem worth screaming about. 5519 5520 Signed-off-by: Lyude Paul <lyude@redhat.com> 5521 Reviewed-by: Dave Airlie <airlied@redhat.com> 5522 (cherry picked from commit dc90b1c3c328f1d0b022a234b69ef32bda7ccb01) 5523 5524commit 91ec6245d6269a88a3f3ecd10fa7a7b326b0a616 5525Author: Lyude Paul <lyude@redhat.com> 5526Date: Wed Jun 20 19:12:31 2018 -0400 5527 5528 glamor: Unbreak glamor_fd_from_pixmap() 5529 5530 When support for allocating GBM BOs with modifiers was added, 5531 glamor_fd_from_pixmap() was changed so that it would return an error if 5532 it got a bo with modifiers set from glamor_fds_from_pixmap(). The 5533 problem is that on systems that support BOs with modifiers, 5534 glamor_fds_from_pixmap() will always return BOs with modifiers. 5535 5536 This means that glamor_fd_from_pixmap() was broken entirely, which broke 5537 a number of other things including glamor_shareable_fd_from_pixmap(), 5538 which meant that modesetting using multiple GPUs with the modesetting 5539 DDX was also broken. Easy reproducer: 5540 5541 - Find a laptop with DRI prime that has outputs connected to the 5542 dedicated GPU and integrated GPU 5543 - Try to enable one display on each using the modesetting DDX 5544 - Fail 5545 5546 Since there isn't a way to ask for no modifiers from 5547 glamor_fds_from_pixmap, we create a shared _glamor_fds_from_pixmap() 5548 function used by both glamor_fds_from_pixmap() and 5549 glamor_fd_from_pixmap() that calls down to the appropriate 5550 glamor_egl_fd*_from_pixmap() function. 5551 5552 Signed-off-by: Lyude Paul <lyude@redhat.com> 5553 Reviewed-by: Dave Airlie <airlied@redhat.com> 5554 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 5555 Fixes: c8c276c956 ("glamor: Implement PixmapFromBuffers and BuffersFromPixmap") 5556 (cherry picked from commit 186a21c4bac744ffe645c8d1a6dda2d41c6d33d8) 5557 5558commit 821f38fa56087fcb11d2d2483366307fc88bf365 5559Author: Lyude Paul <lyude@redhat.com> 5560Date: Thu Jun 7 20:30:34 2018 -0400 5561 5562 modesetting: Also disable CRTC in drmmode_output_disable() 5563 5564 So, this did actually work on older kernels at one point in time, 5565 however it seems that this working was a result of some of the Linux 5566 kernel's atomic modesetting helpers not preserving the CRTC's enabled 5567 state in the right spots. This was fixed in: 5568 5569 846c7dfc1193 ("drm/atomic: Try to preserve the crtc enabled state in drm_atomic_remove_fb, v2") 5570 5571 As a result, atomic commits which simply disassociate a DRM connector 5572 with it's CRTC while leaving the CRTC in an enabled state aren't enough 5573 to disable the CRTC, and result in the atomic commit failing. This 5574 currently can cause issues with MST hotplugging where X will end up 5575 failing to disable the MST outputs after they've left the system. A 5576 simple reproducer: 5577 5578 - Start up Xorg 5579 - Connect an MST hub with displays connected to it 5580 - Remove the hub 5581 - Now there should be CRTCs stuck on the orphaned MST connectors, and X 5582 won't be able to reclaim them. 5583 5584 Signed-off-by: Lyude Paul <lyude@redhat.com> 5585 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 5586 Reviewed-by: Dave Airlie <airlied@redhat.com> 5587 (cherry picked from commit c12f1bd4b76088ea66e3bec9ab9721a52b20cdf2) 5588 5589commit 2f4d0d84266b82d8838b8b9b7972f45d66f0e2d1 5590Author: Olivier Fourdan <ofourdan@redhat.com> 5591Date: Fri Jun 15 08:57:12 2018 +0200 5592 5593 modesetting: use drmmode_bo_import() for rotate_fb 5594 5595 drmmode_shadow_allocate() still uses drmModeAddFB() which may fail if 5596 the format is not as expected, preventing from using a rotated output. 5597 5598 Change it to use the new function drmmode_bo_import() which takes care 5599 of calling the drmModeAddFB2() API. 5600 5601 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106715 5602 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5603 Tested-by: Tomas Pelka <tpelka@redhat.com> 5604 Reviewed-by: Lyude Paul <lyude@redhat.com> 5605 (cherry picked from commit a85e94a50c94b07574c8701a3ff3c1243f4257f4) 5606 5607commit 394ed02f8761c7599fa08628a4d28d34337028f6 5608Author: John Lumby <johnlumby@hotmail.com> 5609Date: Tue Jun 26 17:14:34 2018 -0700 5610 5611 Change the DPMS initialization to be conditional on not set from config 5612 5613 Any DPMS timeout values set in ServerFlags section of the xorg.conf 5614 are being overwritten by DPMS extension initialization. Therefore 5615 change the DPMS initialization of timeout values to be conditional on 5616 not set from config. 5617 5618 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106963 5619 Signed-off-by: John Lumby <johnlumby@hotmail.com> 5620 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5621 Signed-off-by: Keith Packard <keithp@keithp.com> 5622 (cherry picked from commit f5aace7a27d6d61068dcae46c9756b669ab51d50) 5623 5624commit 8c2f0f8cbdba04f3d2eb65e3ea98f809545f8e87 5625Author: Damien Leone <dleone@nvidia.com> 5626Date: Mon Jun 18 16:24:28 2018 -0700 5627 5628 os: Recompute whether any clients are ready after check_timers() 5629 5630 If a driver calls AttendClient() from within a timer callback we 5631 need to re-compute the local 'are_ready' to prevent the attended 5632 client from waiting until WaitForSomething() times out. 5633 5634 This is a fix similar to commit 9ed5b263. 5635 5636 Signed-off-by: Damien Leone <dleone@nvidia.com> 5637 Reviewed-by: Keith Packard <keithp@keithp.com> 5638 (cherry picked from commit f33cb4264387ed14a586ba080885b4d21e4aa48b) 5639 5640commit 180ab06d4564579cf05cb44c2533ac914b68be77 5641Author: Olivier Fourdan <ofourdan@redhat.com> 5642Date: Mon Jun 11 09:21:08 2018 +0200 5643 5644 xwayland: mandatory EGL backend API 5645 5646 The API init_wl_registry() and has_wl_interfaces() are marked as being 5647 optional, but both GBM And EGLStream backends implement them so there is 5648 point in keeping those optional. 5649 5650 Suggested-by: Emil Velikov <emil.velikov@collabora.com> 5651 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5652 (cherry picked from commit 92daeb31fa3235dc791e0444b072ec4bbc6e35ab) 5653 5654commit c641d10ef04475f2898cc40536bd5a03371f2761 5655Author: Olivier Fourdan <ofourdan@redhat.com> 5656Date: Mon Jun 11 09:13:30 2018 +0200 5657 5658 xwayland: simplify xwl_glamor_pixmap_get_wl_buffer() 5659 5660 When retrieving the Wayland buffer from a pixmap, if the buffer already 5661 exists, the GBM backend will return that existing buffer. 5662 5663 However, as seen with the Present issues, if the call had previously 5664 passed a wrong size, that buffer will remain at the wrong size for as 5665 long as the buffer exists, which is error prone. 5666 5667 Considering that the width/height passed to get_wl_buffer() is always the 5668 actual pixmap drawable size, and considering that the EGLStream backend 5669 makes no use of the size either, there is really no point in passing the 5670 width/height around. 5671 5672 Simplify the xwl_glamor_pixmap_get_wl_buffer() and EGL backends API by 5673 removing the pixmap size, and use the drawable size instead. 5674 5675 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5676 (cherry picked from commit 792359057bd54548555674d2d309c0cfeebac12d) 5677 5678commit 79ebd7f689492314067fc863510ad44683b67e64 5679Author: Olivier Fourdan <ofourdan@redhat.com> 5680Date: Tue Jun 5 19:38:45 2018 +0200 5681 5682 xwayland: EGL_IMG_context_priority required by EGLStream 5683 5684 xwl_glamor_eglstream_init_egl() uses "EGL_IMG_context_priority" 5685 extension, make sure it's actually available before using it. 5686 5687 Suggested-by: Emil Velikov <emil.velikov@collabora.com> 5688 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5689 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5690 (cherry picked from commit bdadaa25f5c1f62d30d8e76b4ebfcef414ed9c90) 5691 5692commit 81969ab773ddf7bae04eb7d1cca3fe01344dcae9 5693Author: Olivier Fourdan <ofourdan@redhat.com> 5694Date: Tue Jun 5 19:38:44 2018 +0200 5695 5696 xwayland: check for EGLStream backend explicitly 5697 5698 Now that we have separate backends for EGLStream and GBM, we can 5699 explicitly check for the EGLStream backend to disable present support 5700 in that case. 5701 5702 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5703 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5704 (cherry picked from commit 5d843f6947538dabde258584a5795e0b25ea8779) 5705 5706commit 60020989b9759a05153f832a160e6f573d53e266 5707Author: Olivier Fourdan <ofourdan@redhat.com> 5708Date: Tue Jun 5 19:38:43 2018 +0200 5709 5710 xwayland: refactor EGL backends for wayland registry 5711 5712 To be able to check for availability of the Wayland interfaces required 5713 to run a given EGL backend (either GBM or EGLStream for now), we need 5714 to have each backend structures and vfuncs in place before we enter the 5715 Wayland registry dance. 5716 5717 That basically means that we should init all backends at first, connect 5718 to the Wayland compositor and query the available interfaces and then 5719 decide which backend is available and should be used (or none if either 5720 the Wayland interfaces or the EGL extensions are not available). 5721 5722 For this purpose, hold an egl_backend struct for each backend we are to 5723 consider prior to connect to the Wayland display so that, when we get to 5724 query the Wayland interfaces, everything is in place for each backend to 5725 handle the various Wayland interfaces. 5726 5727 Eventually, when we need to chose which EGL backend to use for glamor, 5728 the available Wayland interfaces and EGL extensions available are all 5729 known to Xwayland. 5730 5731 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5732 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5733 (cherry picked from commit d7185a84b60ed03aaa84eb522dcff365218e7211) 5734 5735commit cb698ec2ba11b827c4e6785693a83a27eec3c635 5736Author: Olivier Fourdan <ofourdan@redhat.com> 5737Date: Tue Jun 5 19:38:42 2018 +0200 5738 5739 xwayland: move EGL backend init to glamor 5740 5741 Move EGL backends initialization to its own function in 5742 xwayland-glamor.c 5743 5744 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5745 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5746 (cherry picked from commit 48f037a27c45b571c9750ac812977ac0a33ab12b) 5747 5748commit 8ffee3a6bd8901b4c87890087b8bd2ccb4a30a5c 5749Author: Olivier Fourdan <ofourdan@redhat.com> 5750Date: Tue Jun 5 19:38:41 2018 +0200 5751 5752 xwayland: Add Wayland interfaces check 5753 5754 Introduces a new egl_backend function to let the EGL backend check for 5755 the presence of the required Wayland interfaces. 5756 5757 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5758 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5759 (cherry picked from commit f2fcb4877e976d078b0eb4755177170467341484) 5760 5761commit aad1525180c0352ab9ea22c35405226b3afa69cc 5762Author: Olivier Fourdan <ofourdan@redhat.com> 5763Date: Tue Jun 5 19:38:40 2018 +0200 5764 5765 xwayland: move egl_backend to its own struct 5766 5767 EGL backend availability requires both EGL extensions and Wayland 5768 interfaces to be present, so we will need to consider multiple backends 5769 during initialization. 5770 5771 As a preliminary work, move the egl_backend to its own struct so that we 5772 can have more than one backend at any given time. 5773 5774 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5775 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5776 (cherry picked from commit b74b0f18b8d3032317e38453ee63ae9efd33a098) 5777 5778commit 831f7194b7006da14424da219661373e7d49114a 5779Author: Olivier Fourdan <ofourdan@redhat.com> 5780Date: Tue Jun 5 19:38:39 2018 +0200 5781 5782 xwayland: skip drm authentication with render node 5783 5784 If using a render node, we can skip DRM authentication. 5785 5786 Suggested-by: Emil Velikov <emil.velikov@collabora.com> 5787 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5788 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5789 (cherry picked from commit de004eefc60ec595f5d6d81658468e082f8df930) 5790 5791commit 60eda2af0acd54e05ab794947f106f8f2d421e90 5792Author: Olivier Fourdan <ofourdan@redhat.com> 5793Date: Tue Jun 5 19:38:38 2018 +0200 5794 5795 xwayland: GBM should fail w/out "GL_OES_EGL_image" 5796 5797 Surely, we should fail to init GBM backend if "GL_OES_EGL_image" is 5798 missing. 5799 5800 This seems to have been lost with commit 1545e2dba ("xwayland: Decouple 5801 GBM from glamor"). 5802 5803 Suggested-by: Emil Velikov <emil.velikov@collabora.com> 5804 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5805 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5806 (cherry picked from commit b823b43dca143810146f563d09e8996058b9d09e) 5807 5808commit f8e96b22bfac0733328fb9902c151003db5cc724 5809Author: Olivier Fourdan <ofourdan@redhat.com> 5810Date: Tue Jun 5 19:38:37 2018 +0200 5811 5812 xwayland: swap "name" and "id" in init_wl_registry() 5813 5814 Both xwl_glamor_init_wl_registry() and the Wayland global registry 5815 handler use the interface id/name in that order, using name/id in the 5816 egl_backend vfunc makes things confusing and error prone. 5817 5818 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5819 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5820 (cherry picked from commit 78ce4aa979ff3f5870fbc12a7e5c53547084a61a) 5821 5822commit 443e3348bb3c561160a2542365507179bd8dd825 5823Author: Olivier Fourdan <ofourdan@redhat.com> 5824Date: Tue Jun 5 19:38:36 2018 +0200 5825 5826 xwayland: move glamor specific routines 5827 5828 Functions such as: 5829 5830 xwl_glamor_egl_supports_device_probing() 5831 xwl_glamor_egl_get_devices() 5832 xwl_glamor_egl_device_has_egl_extensions() 5833 5834 Are of no use outside of EGLStream support, move them to the relevant 5835 source file. 5836 5837 Similarly, the other glamor functions such as: 5838 5839 xwl_glamor_init() 5840 xwl_screen_set_drm_interface() 5841 xwl_screen_set_dmabuf_interface() 5842 xwl_glamor_pixmap_get_wl_buffer() 5843 xwl_glamor_init_wl_registry() 5844 xwl_glamor_post_damage() 5845 xwl_glamor_allow_commits() 5846 xwl_glamor_egl_make_current() 5847 5848 Are useless without glamor support enabled, move those within a 5849 a "#ifdef XWL_HAS_GLAMOR" in xwayland.h 5850 5851 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5852 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5853 (cherry picked from commit f6b2109c1b49484af772df518314ec8e8432febe) 5854 5855commit dea40be93577dff80cf25e9b39c2aaea030b9c9e 5856Author: Olivier Fourdan <ofourdan@redhat.com> 5857Date: Tue Jun 5 19:37:58 2018 +0200 5858 5859 xwayland: make xwl_output_get_xdg_output() static 5860 5861 Make xwl_output_get_xdg_output() private, it doesn't need to be 5862 available elsewhere. 5863 5864 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5865 Reviewed-by: Lyude Paul <lyude@redhat.com> 5866 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5867 (cherry picked from commit d31a7be15e259275599a9f67e0d921471ae64913) 5868 5869commit 04a19291c93aaa4feeaf7a87663eaf1ef7c1f4b8 5870Author: Olivier Fourdan <ofourdan@redhat.com> 5871Date: Tue Jun 5 19:37:57 2018 +0200 5872 5873 xwayland: do not disable glamor if EGLStream failed 5874 5875 EGLStream requires glamor, but the opposite is not true. So if someone 5876 passes "-eglstream" with a GPU which does not support EGLStream, we 5877 could maybe still try GBM and be lucky. 5878 5879 That allows Wayland compositors to pass "-eglstream" regardless of the 5880 actual hardware, if they want to enable EGLStream on GPU which support 5881 it. 5882 5883 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5884 Reviewed-by: Lyude Paul <lyude@redhat.com> 5885 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5886 (cherry picked from commit e16a6da79dea793a335be70ba07d5e1c1295b5eb) 5887 5888commit de40a55235614d6114e97bd8c6f687bb555bdba8 5889Author: Olivier Fourdan <ofourdan@redhat.com> 5890Date: Tue Jun 5 19:37:56 2018 +0200 5891 5892 xwayland: process Wayland events after adding screen 5893 5894 When we're done adding a new screen, we need to process any pending 5895 Wayland events again. 5896 5897 Hence we don't end up processing xdg_output events unexpectedly when 5898 glamor is disabled. Be that because "-shm" was passed or "-eglstream" 5899 has failed. 5900 5901 Failing to do that could lead to a crash at startup: 5902 5903 Xwayland: dixGetPrivateAddr: Assertion `key->initialized' failed. 5904 (EE) 5905 (EE) Backtrace: 5906 (EE) 0: Xwayland (OsSigHandler) 5907 (EE) 1: libpthread.so.0 (funlockfile) 5908 (EE) 2: libc.so.6 (gsignal) 5909 (EE) 3: libc.so.6 (abort) 5910 (EE) 4: libc.so.6 (?+0x0) 5911 (EE) 5: libc.so.6 (__assert_fail) 5912 (EE) 6: Xwayland (dixGetPrivateAddr) 5913 (EE) 7: Xwayland (_fbGetWindowPixmap) 5914 (EE) 8: Xwayland (getDrawableDamageRef) 5915 (EE) 9: Xwayland (damageRegionProcessPending) 5916 (EE) 10: Xwayland (damagePolyFillRect) 5917 (EE) 11: Xwayland (miPaintWindow) 5918 (EE) 12: Xwayland (miWindowExposures) 5919 (EE) 13: Xwayland (miHandleValidateExposures) 5920 (EE) 14: Xwayland (SetRootClip) 5921 (EE) 15: Xwayland (update_screen_size) 5922 (EE) 16: Xwayland (apply_output_change) 5923 (EE) 17: libffi.so.6 (ffi_call_unix64) 5924 (EE) 18: libffi.so.6 (ffi_call) 5925 (EE) 19: libwayland-client.so.0 (wl_log_set_handler_client) 5926 (EE) 20: libwayland-client.so.0 (_init) 5927 (EE) 21: libwayland-client.so.0 (wl_display_dispatch_queue_pending) 5928 (EE) 22: libwayland-client.so.0 (wl_display_roundtrip_queue) 5929 (EE) 23: Xwayland (InitInput) 5930 (EE) 24: Xwayland (dix_main) 5931 (EE) 25: libc.so.6 (__libc_start_main) 5932 (EE) 26: Xwayland (_start) 5933 (EE) 5934 (EE) 5935 Fatal server error: 5936 (EE) Caught signal 6 (Aborted). Server aborting 5937 (EE) 5938 Aborted (core dumped) 5939 5940 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5941 Reviewed-by: Lyude Paul <lyude@redhat.com> 5942 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5943 (cherry picked from commit 44560af02823239e1b5b236e8cb365b5bdf24b6a) 5944 5945commit 2d31a40db5ab3dc93863dc5ed9d9865490508575 5946Author: Olivier Fourdan <ofourdan@redhat.com> 5947Date: Tue Jun 5 19:37:55 2018 +0200 5948 5949 xwayland: "EGL_EXT_device_base" required for EGLStream 5950 5951 eglQueryDevicesEXT() would abort if the required extensions are not 5952 available, meaning that enabling “-eglstream” on a non-EGLStream 5953 capable hardware would lead to an abort(). 5954 5955 Check that "EGL_EXT_device_base" extension is available and bail out 5956 early if not, so we don't abort() later in eglQueryDevicesEXT(). 5957 5958 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5959 Reviewed-by: Lyude Paul <lyude@redhat.com> 5960 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5961 (cherry picked from commit dbde3fec32641ae23389cffa12a04d3608936795) 5962 5963commit 65d46b2dfd638f6f6c6e52f3494c64716fc6687e 5964Author: Olivier Fourdan <ofourdan@redhat.com> 5965Date: Tue Jun 5 19:37:54 2018 +0200 5966 5967 xwayland: allow "-eglstream" option 5968 5969 The command line option "-eglstream" used to enable EGLStream support 5970 for NVidia GPU was made available only when Xwayland was built with 5971 EGLStream support enabled. 5972 5973 Wayland compositors who spawn Xwayland have no easy way to tell whether 5974 or not Xwayland was built with EGLStream support enabled, and adding 5975 "-eglstream" command line option to Xwayland when it wasn't built with 5976 EGLStream support would prevent Xwayland from starting (“Unrecognized 5977 option” error). 5978 5979 Make sure we support the command line option "-eglstream" regardless of 5980 EGLStream support in Xwayland. Obviously, if Xwayland was built without 5981 EGLStream support, this has no effect. 5982 5983 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 5984 Reviewed-by: Lyude Paul <lyude@redhat.com> 5985 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 5986 (cherry picked from commit 06c31e782e360363238cf20e7afd3e4990a2c304) 5987 5988commit 38835d1d8f613c3553149fc05e61a8bb9b0aef48 5989Author: Vladimir Panteleev <git@thecybershadow.net> 5990Date: Wed Jun 20 13:37:45 2018 +0000 5991 5992 glx/vndcmds: Fix vendor hash table key size 5993 5994 The keySize parameter of the hashing/comparison functions was 5995 incorrectly specified to be sizeof(void*), even though the keys of 5996 this hashtable are CARD32. 5997 5998 Fixes address sanitizer failure on 64-bit builds. 5999 6000 Reviewed-by: Adam Jackson <ajax@redhat.com> 6001 (cherry picked from commit 707d0f912b916e7546c4f7e7a5f7023a53e74615) 6002 6003commit d7220428daee26acecb6ae9771e255c801c83572 6004Author: Alexander Volkov <a.volkov@rusbitech.ru> 6005Date: Tue Jun 5 13:05:39 2018 +0300 6006 6007 Xext/shm: Refuse to work for remote clients 6008 6009 Avoid access to System V shared memory segment on the X server side 6010 for clients forwarded via SSH. Also prevent them from hanging while 6011 waiting for the reply from the ShmCreateSegment request. 6012 6013 v2: Allow ShmQueryVersion request even for remote clients 6014 6015 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=11080 6016 Signed-off-by: Alexander Volkov <a.volkov@rusbitech.ru> 6017 Reviewed-by: Adam Jackson <ajax@redhat.com> 6018 (cherry picked from commit ec7e2b54c5b4a34b2a077082967bc3ead30e227e) 6019 6020commit 9e417072bcde15f32dbf083327b6fe0202bb4cf4 6021Author: Michał Górny <gentoo@mgorny.alt.pl> 6022Date: Mon Jun 11 17:17:31 2018 -0400 6023 6024 xfree86: Makefile shouldn't rely on superuser being named 'root' 6025 6026 Change the 'chown' statement in Makefile.am to use the numeric UID 6027 of superuser instead of relying on the name 'root'. 6028 6029 Bugzilla: https://bugs.freedesktop.org/27726 6030 Signed-off-by: Adam Jackson <ajax@redhat.com> 6031 Signed-off-by: Michał Górny <gentoo@mgorny.alt.pl> 6032 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 6033 (cherry picked from commit 5c95be38e5af28108d8c6c08dfe08d1dfcb7ed42) 6034 6035commit 2e66ed066e4a9b52bcebf2ca3762f7e1ad141cf0 6036Author: Olivier Fourdan <ofourdan@redhat.com> 6037Date: Fri Jun 8 16:23:44 2018 +0200 6038 6039 xwayland: use pixmap size on present flip 6040 6041 If the pixmap size does not match the present box size, flickering 6042 occurs. 6043 6044 This can happen when the client changes its size (e.g. switching to 6045 fullscreen), and since the buffer is kept as long as the pixmap is 6046 valid, once the buffer is created, it remains at the wrong (old) size 6047 and causes continuous flickering. 6048 6049 Use the actual pixmap's drawable size instead of the present box to 6050 create the buffer so that it's sized appropriately. 6051 6052 Bugzilla: https://bugs.freedesktop.org/106841 6053 Fixes: 0fb2cca193e6 "xwayland: Preliminary support for Present's new 6054 window flip mode" 6055 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 6056 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6057 Reviewed-by: Roman Gilg <subdiff@gmail.com> 6058 (cherry picked from commit 1993f147d08170f07a72e43f0a0f27687e16967b) 6059 6060commit 883d4d00b46f18902ad402a156d56b1bfcc06ebe 6061Author: Michel Dänzer <michel.daenzer@amd.com> 6062Date: Thu Jun 7 17:55:21 2018 +0200 6063 6064 present/wnmd: Preserve window pixmap's screen_x/y on flip 6065 6066 The incorrect values could result in the new pixmap's contents 6067 getting corrupted down the line. 6068 6069 v2: 6070 * Guard screen_x/y lines by #ifdef COMPOSITE 6071 6072 Bugzilla: https://bugs.freedesktop.org/106841 6073 Fixes: 029608dd8020 "present: Add window flip mode" 6074 Reviewed-by: Adam Jackson <ajax@redhat.com> # v1 6075 Reviewed-by: Keith Packard <keithp@keithp.com> # v1 6076 Reviewed-by: Roman Gilg <subdiff@gmail.com> 6077 Tested-by: Olivier Fourdan <ofourdan@redhat.com> # v1 6078 (cherry picked from commit 10eec2ccb11701fe29ab246acd6c0bdc2991b775) 6079 6080commit 55171d738a0c2dd8a37ba2c880ea4658fd8001a2 6081Author: Dave Airlie <airlied@redhat.com> 6082Date: Tue May 29 10:39:22 2018 +1000 6083 6084 xwayland: fix typo in non-modifier fallback path 6085 6086 Pointed out on irc by q66. 6087 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 6088 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 6089 6090 (cherry picked from commit 6300049a9a4b84789a2e8141f5cad17adf22185c) 6091 6092commit 64bf28572946da1db84f6e61b02b5e6432f13d10 6093Author: Roman Kapl <code@rkapl.cz> 6094Date: Thu Jun 7 03:31:42 2018 +0200 6095 6096 Xi: add forgotten byte-swaps for Valuator fields 6097 6098 This has caused nonsensical values in xinput output. 6099 6100 Signed-off-by: Roman Kapl <code@rkapl.cz> 6101 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 6102 (cherry picked from commit 40586cc4f8faa271d62fd84fe98c41debb878984) 6103 6104commit 273115a01285ff2ce4837687f9bfc9d745a519a3 6105Author: Roman Kapl <code@rkapl.cz> 6106Date: Thu Jun 7 03:31:41 2018 +0200 6107 6108 Xi: fix byte-swapping of button labels 6109 6110 The byte-swapping code forgot that the xXIButtonInfo is followed by a 6111 button mask, not directly by the button labels. This resulted in client 6112 crashes in cross-endian setups, for example in `xinput list --long`, 6113 since the client got an invalid atom. 6114 6115 A new function was introduced to get the right positions for the label 6116 and mask data. 6117 6118 Signed-off-by: Roman Kapl <code@rkapl.cz> 6119 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 6120 (cherry picked from commit cefbc6a9356e5c6cf935b61557efa897762defae) 6121 6122commit 69980e4e5a0d8ff3b36acf09e3c1e8eb76710726 6123Author: Michel Dänzer <michel.daenzer@amd.com> 6124Date: Fri Jun 1 11:57:15 2018 +0200 6125 6126 exa: Use PictureMatchFormat for source-only picture format description 6127 6128 Their pFormat member is NULL, which resulted in a crash in 6129 miRenderColorToPixel. 6130 6131 Fixes: 8171d4c2d67b "render: Store and use all 16bpc of precision for 6132 solid pixels (v2.1)" 6133 Reviewed-by: Adam Jackson <ajax@redhat.com> 6134 (cherry picked from commit 3ebef6ab850675e1091df2125ebdfbb147436e6c) 6135 6136commit 33f20c38fc15e6664635ad22a581d7860e4b185d 6137Author: Michel Dänzer <michel.daenzer@amd.com> 6138Date: Wed May 23 11:18:02 2018 +0200 6139 6140 glamor: Propagate glamor_fds_from_pixmap error in glamor_fd_from_pixmap 6141 6142 glamor_fds_from_pixmap returns 0 on error, but we were treating that as 6143 success, continuing with uninitialized stride and fd values. 6144 6145 Also bail if the offset isn't 0, same as in dri3_fd_from_pixmap. 6146 6147 v2: 6148 * Reduce to a simple one-liner fix (Emil Velikov) 6149 6150 Fixes: c8c276c9569b "glamor: Implement PixmapFromBuffers and 6151 BuffersFromPixmap" 6152 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 6153 (cherry picked from commit 4d5950ce14676f970d9de97380929a93948b98f2) 6154 6155commit 544caffae1b0dc786383b117c036284feb579a66 6156Author: Michel Dänzer <michel.daenzer@amd.com> 6157Date: Wed May 23 11:38:13 2018 +0200 6158 6159 glamor: Always return 0 from glamor_fds_from_pixmap on error 6160 6161 This matches what glamor_egl_fds_from_pixmap and dri3_fds_from_pixmap do 6162 and what proc_dri3_buffers_from_pixmap expects. 6163 6164 Fixes: c8c276c9569b "glamor: Implement PixmapFromBuffers and 6165 BuffersFromPixmap" 6166 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 6167 (cherry picked from commit 3da999a0390407d512bf784e4faf01482bfff9ea) 6168 6169commit a9a17581ce73a49f0a8d3b9dbb2db083fa44cb69 6170Author: Michel Dänzer <michel.daenzer@amd.com> 6171Date: Fri May 18 11:23:01 2018 +0200 6172 6173 modesetting: Pass O_CLOEXEC when opening a DRM device 6174 6175 We don't want DRM file descriptors to leak to child processes. 6176 6177 Reviewed-by: Keith Packard <keithp@keithp.com> 6178 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 6179 (cherry picked from commit 315c63c41d833700c505e4f21ffac4c59e1e3b2c) 6180 6181commit c3a06e330b7e55e1daeefbb5c16a075c80233bb0 6182Author: Michel Dänzer <michel.daenzer@amd.com> 6183Date: Fri May 18 11:23:00 2018 +0200 6184 6185 xfree86: Fix O_CLOEXEC usage in lnx_platform 6186 6187 It was passing O_CLOEXEC as permission bits instead of as a flag. 6188 6189 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 6190 Reviewed-by: Keith Packard <keithp@keithp.com> 6191 (cherry picked from commit ab53e2859facecc0486344679dc01ea31fb427e9) 6192 6193commit 662273397c537d031e1aee7c607c925ec6bb2d10 6194Author: Lukas F. Hartmann <lukas@mntmn.com> 6195Date: Tue May 15 21:51:56 2018 +0200 6196 6197 glamor_init: clamp GLSL to 120 if platform doesn't have instanced arrays 6198 6199 Hi, 6200 6201 I upgraded Xwayland and the assorted libraries from git masters today, 6202 and noticed that glamor wouldn't work anymore on i.MX6/etnaviv. The 6203 error was: 6204 6205 No provider of glVertexAttribDivisor found. Requires one of: 6206 Desktop OpenGL 3.3 6207 OpenGL ES 3.0 6208 GL extension "GL_ANGLE_instanced_arrays" 6209 GL extension "GL_ARB_instanced_arrays" 6210 GL extension "GL_EXT_instanced_arrays" 6211 GL extension "GL_NV_instanced_arrays" 6212 6213 The problem is that etnaviv offers GLSL 140 on GL 2.1 and glamor 6214 rendering assumes that glVertexAttribDivisor() is always available on 6215 GLSL>=130, which is not the case here. Forcing GLSL 120 makes glamor 6216 work fine again on this platform. After chatting with ajax in 6217 #xorg-devel, the following solution was proposed. 6218 6219 This is my first time of submitting a patch, so please excuse me and 6220 advise if I'm doing it wrong ;) 6221 6222 Cheers 6223 Lukas (mntmn) 6224 Reviewed-by: Eric Anholt <eric@anholt.net> 6225 6226 (cherry picked from commit 7437b6dbdee050f8ebb3a79b9077d051c91880c3) 6227 6228commit 590374bd9213091b51679887d473d1a3067c99b5 6229Author: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl> 6230Date: Sun May 13 15:31:37 2018 +0200 6231 6232 DRI2: Sync radeonsi_pci_ids.h from Mesa 6233 6234 Fixes DRI2 client driver name mapping for newer AMD GPUs with the 6235 modesetting driver, allowing the DRI2 extension to initialize. 6236 6237 Fixes using GL with the modesetting driver for me. 6238 6239 Seems we were way behind on this one, time to look into something 6240 more scalable? 6241 6242 Signed-off-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl> 6243 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 6244 (cherry picked from commit 3ab32a537840c6e6d6228b4ba62f98fbf5224f8a) 6245 6246commit 5407be2c516fa23c505e7ba254367b5739399960 6247Author: Jon Turney <jon.turney@dronecode.org.uk> 6248Date: Sun May 13 19:36:05 2018 +0100 6249 6250 meson: don't put literal 'PACKAGE_STRING' and 'XORG_MAN_PAGE' in man pages 6251 6252 Instead, substitute the same values as autotools does 6253 6254 Reviewed-by: Adam Jackson <ajax@redhat.com> 6255 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 6256 (cherry picked from commit 27eff10bfa1a02da7338a82e2312d2080e3e5876) 6257 6258commit acd881e271419a2fc08cfc34ed9845346cee9e84 6259Author: Jon Turney <jon.turney@dronecode.org.uk> 6260Date: Sun May 13 19:36:04 2018 +0100 6261 6262 meson: don't install xorg wrapper manpages if suid-wrapper isn't being used 6263 6264 Reviewed-by: Adam Jackson <ajax@redhat.com> 6265 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 6266 (cherry picked from commit 47321bb455b1226c04711241496e62667eea34ea) 6267 6268commit c09c78298cdb4cdd8383e252d8fe68adae1f3c54 6269Author: Jon Turney <jon.turney@dronecode.org.uk> 6270Date: Sun May 13 19:36:03 2018 +0100 6271 6272 meson: install xwinclip and Xwinrc man pages 6273 6274 Omitted from a1e8dc05 6275 6276 Reviewed-by: Adam Jackson <ajax@redhat.com> 6277 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 6278 (cherry picked from commit 08a3583b5b2f812cc7f039aebd74ff9b1cb0dbcd) 6279 6280commit 82f8fc5ffc0065537896ec43ae60d67d9a164c75 6281Author: Adam Jackson <ajax@redhat.com> 6282Date: Thu May 10 12:24:11 2018 -0400 6283 6284 xserver 1.20 6285 6286 Signed-off-by: Adam Jackson <ajax@redhat.com> 6287 6288commit b6bf68b8416ec4b717eb78703fb63789c68e509a 6289Author: Aaron Plattner <aplattner@nvidia.com> 6290Date: Wed May 2 12:54:26 2018 -0700 6291 6292 meson: Fix module_dir configuration (v2) 6293 6294 meson.build has code to set the module_dir variable to 6295 ${libdir}/xorg/modules if the module_dir option string is empty. 6296 However, this has several problems: 6297 6298 1. The variable is only used for an unused @moduledir@ substitution in 6299 the man page. The rule for xorg-server.pc uses option('module_dir') 6300 directly instead. 6301 2. The 'module_dir' option has a default value of 'xorg/modules' so the 6302 above rule doesn't do anything by default. 6303 3. The xorg-server.pc rule uses ${exec_prefix}/option('module_dir'), so 6304 the effect of #2 is that the default moduledir is different between 6305 autoconf and meson. E.g. if ${prefix} is /X, then you get 6306 6307 autoconf: moduledir=/X/lib/xorg/modules 6308 meson: moduledir=/X/xorg/modules 6309 6310 Fix this by using the module_dir variable when generating xorg-server.pc, and by 6311 using join_paths() to assign module_dir unconditionally. 6312 6313 v2: Keep the 'xorg/modules' default path, but use join_paths() unconditionally (Thierry Reding) 6314 6315 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 6316 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 6317 6318commit 9d628ee5facf6318368d5b8ca181e083adcffe8c 6319Author: Roman Gilg <subdiff@gmail.com> 6320Date: Wed May 2 21:32:48 2018 +0200 6321 6322 modesetting: set gbm as dependency in meson build 6323 6324 Modifiers support needs gbm as a dependency. Without setting the dependency 6325 included headers are not found reliably and the build might fail if the 6326 headers are not placed in the default system include paths. 6327 6328 Signed-off-by: Roman Gilg <subdiff@gmail.com> 6329 Reviewed-by: Adam Jackson <ajax@redhat.com> 6330 6331commit 0445705a8bbf76a5532997823ea8110daff26c6d 6332Author: Adam Jackson <ajax@redhat.com> 6333Date: Mon May 7 17:21:20 2018 -0400 6334 6335 man: Fix automake seddery 6336 6337 Because this is an automakefile, things inside @@ get expanded, which 6338 means your sed ends up saying s|/var/log|/var/log| and your manual pages 6339 still have @logdir@ in them. Fix this by hiding the @s inside a trivial 6340 character range, which keeps the pattern preserved all the way into the 6341 Makefile. 6342 6343 Signed-off-by: Adam Jackson <ajax@redhat.com> 6344 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 6345 6346commit 9f21872ad8591c1f6f8ea1e466f212c3582d3c33 6347Author: Adam Jackson <ajax@redhat.com> 6348Date: Mon May 7 17:21:19 2018 -0400 6349 6350 glx: Be sure to set an error for ghost contexts 6351 6352 Otherwise the caller is going to return garbage memory for the error 6353 value. 6354 6355 Signed-off-by: Adam Jackson <ajax@redhat.com> 6356 6357commit b9f415cbad47412bfb218cf7375d0c2856a27d1b 6358Author: Adam Jackson <ajax@redhat.com> 6359Date: Mon May 7 17:21:18 2018 -0400 6360 6361 present: Fix swapping of PresentCompleteNotify events 6362 6363 The code would fall through to the PresentIdleNotify case, and nothing 6364 good would come of it. 6365 6366 Signed-off-by: Adam Jackson <ajax@redhat.com> 6367 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 6368 6369commit cc66777d85f3509b0f9dfc9210d0a0415a2a388d 6370Author: Adam Jackson <ajax@redhat.com> 6371Date: Mon May 7 17:21:17 2018 -0400 6372 6373 xwayland: Don't create a "fake" crtc for Present 6374 6375 We probably don't want a fake crtc to be visible to clients, and we 6376 definitely don't want to generate events every time we create such a 6377 fake (which would happen as a side effect from RRCrtcCreate hitting 6378 RRTellChanged). As it happens we're not actually using that crtc for 6379 anything because xwayland doesn't store any state on the crtc object, 6380 so it suffices to use the real crtc for the screen. 6381 6382 Signed-off-by: Adam Jackson <ajax@redhat.com> 6383 Tested-by: Roman Gilg <subdiff@gmail.com> 6384 Reviewed-by: Roman Gilg <subdiff@gmail.com> 6385 6386commit b23a0e4ded62500f2c248f23962ff5ac718467e6 6387Author: Eric Anholt <eric@anholt.net> 6388Date: Mon May 7 16:46:20 2018 -0700 6389 6390 xwayland: Fix a 32-bit build warning. 6391 6392 Signed-off-by: Eric Anholt <eric@anholt.net> 6393 Reviewed-by: Adam Jackson <ajax@redhat.com> 6394 6395commit ef95331603ff31d3643360c399b3865db5b0b97d 6396Author: Eric Anholt <eric@anholt.net> 6397Date: Mon May 7 16:46:19 2018 -0700 6398 6399 dri3: Switch fds_from_pixmap to stdint types. 6400 6401 Again, this was causing 32-bit build warnings due to mixing CARD* and 6402 stdint. 6403 6404 Signed-off-by: Eric Anholt <eric@anholt.net> 6405 Reviewed-by: Adam Jackson <ajax@redhat.com> 6406 6407commit 4c754b01fafc4b042d9918c05a71157f6fa6392a 6408Author: Eric Anholt <eric@anholt.net> 6409Date: Mon May 7 16:46:18 2018 -0700 6410 6411 dri3: Switch get_modifiers to using stdint. 6412 6413 We were mixing stdint and CARD* types, causing compiler warnings on 6414 32-bit. Just switch over to stdint, which is what we'd like the server 6415 to be using long term, anyway. 6416 6417 Reviewed-by: Adam Jackson <ajax@redhat.com> 6418 6419commit 5e86484a183f051d7a8a67ea13c23a3d6e69e76b 6420Author: Eric Anholt <eric@anholt.net> 6421Date: Mon May 7 16:46:17 2018 -0700 6422 6423 dri3: Switch get_drawable_modifiers to using stdint. 6424 6425 We were mixing stdint and CARD* types, causing compiler warnings on 6426 32-bit. Just switch over to stdint, which is what we'd like the server 6427 to be using long term, anyway. 6428 6429 Signed-off-by: Eric Anholt <eric@anholt.net> 6430 Reviewed-by: Adam Jackson <ajax@redhat.com> 6431 6432commit 4ec02b573ef2424965d7ce2d33d150ddb92ec544 6433Author: Eric Anholt <eric@anholt.net> 6434Date: Mon May 7 16:46:16 2018 -0700 6435 6436 randr: Fix a compiler warning on 32-bit. 6437 6438 Signed-off-by: Eric Anholt <eric@anholt.net> 6439 Reviewed-by: Adam Jackson <ajax@redhat.com> 6440 6441commit e1ccd0fa0e1081edf8a2c69ce6e8e3f67a4aecba 6442Author: Eric Anholt <eric@anholt.net> 6443Date: Mon May 7 16:46:15 2018 -0700 6444 6445 dix: Fix a warning about GetTimeInMillis return value in XFont2. 6446 6447 Signed-off-by: Eric Anholt <eric@anholt.net> 6448 Reviewed-by: Adam Jackson <ajax@redhat.com> 6449 6450commit cf838f5ca81e0c967902b74fb1971bc6fac5c601 6451Author: Roman Gilg <subdiff@gmail.com> 6452Date: Fri May 4 03:07:31 2018 +0200 6453 6454 xwayland: persistent window struct on present 6455 6456 Instead of reusing xwl_window introduce a persistent window struct for every 6457 window, that asks for Present flips. 6458 6459 This struct saves all relevant data and is only freed on window destroy. 6460 6461 Signed-off-by: Roman Gilg <subdiff@gmail.com> 6462 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 6463 6464commit c9afd8cb5ec975b189ab7b678e1f997d6a2ba5ee 6465Author: Mario Kleiner <mario.kleiner.de@gmail.com> 6466Date: Sun May 6 07:34:31 2018 +0200 6467 6468 modesetting: Fix and improve ms_kernel_msc_to_crtc_msc() 6469 6470 The old 32-Bit wraparound handling didn't actually work, due to some 6471 integer casting bug, and the mapping was ill equipped to deal with input 6472 from the new true 64-bit GetCrtcSequence/QueueCrtcSequence api's 6473 introduced in Linux 4.15. 6474 6475 For 32-Bit truncated input from pageflip events and old vblank events 6476 and old drmWaitVblank ioctl, implement new wraparound handling, which 6477 also allows to deal with wraparound in the other direction, e.g., if a 6478 32-Bit truncated sequence value is passed in, whose true 64-Bit 6479 in-kernel hw value is within 2^30 counts of the previous processed 6480 value, but whose 32-bit truncated sequence value happens to lie just 6481 above or below a 2^32 boundary, iow. one of the two values 'sequence' 6482 vs. 'msc_prev' lies above a 2^32 border, the other one below it. 6483 6484 The method is directly translated from Mesa's proven implementation of 6485 the INTEL_swap_events extension, where a true underlying 64-Bit wide 6486 swapbuffers count (SBC) needs to get reconstructed from a 32-Bit LSB 6487 truncated SBC transported over the X11 protocol wire. Same conditions 6488 apply, ie. successive true 64-Bit SBC values are close to each other, 6489 but don't always get received in strictly monotonically increasing 6490 order. See Mesa commit cc5ddd584d17abd422ae4d8e83805969485740d9 ("glx: 6491 Handle out-of-sequence swap completion events correctly. (v2)") for 6492 explanation. 6493 6494 Additionally add a separate path for true 64-bit msc input originating 6495 from Linux 4.15+ drmCrtcGetSequence/QueueSequence ioctl's and 6496 corresponding 64-bit vblank events. True 64-bit msc's don't need 6497 remapping and must be passed through. 6498 6499 As a reliability bonus, they are also used here to update the tracking 6500 values msc_prev and ms_high with perfect 64-Bit ground truth as baseline 6501 for mapping msc from pageflip completion events, because pageflip events 6502 are always 32-bit wide, even when the new kernel api's are used. Because 6503 each pageflip(-event) is always preceeded close in time (and vblank 6504 count) by a drmCrtcQueueSequence queued event or drmCrtcGetSequence 6505 query as part of DRI2 or DRI3+Present swap scheduling, we can be certain 6506 that each pageflip event will get its truncated 32-bit msc remapped 6507 reliably to the true 64-bit msc of flip completion whenever the sequence 6508 api is available, ie. on Linux 4.15 or later. 6509 6510 Note: In principle at least the 32-bit mapping path could also be 6511 backported to earlier server branches, as this seems to be broken for at 6512 least server 1.16 to 1.19. 6513 6514 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 6515 Reviewed-by: Adam Jackson <ajax@redhat.com> 6516 Cc: Keith Packard <keithp@keithp.com> 6517 Cc: Michel Dänzer <michel.daenzer@amd.com> 6518 6519commit 73f0ed2d928afc692ed057eb3d7627328a6e5b12 6520Author: Mario Kleiner <mario.kleiner.de@gmail.com> 6521Date: Fri May 4 14:14:10 2018 +0200 6522 6523 modesetting: Remove ms_crtc_msc_to_kernel_msc(). 6524 6525 The function is ported from intel-ddx uxa backend around 2013, where its 6526 stated purpose was to apply a vblank_offset to msc values to correct for 6527 problems with those kernel provided msc values. Some (somewhat magic and 6528 puzzling to myself) heuristic tried to guess if provided values were 6529 unreasonable and tried to adapt the corrective vblank_offset to account 6530 for that. 6531 6532 Except: It wasn't applied to kernel provided msc values, but the values 6533 delivered by clients via DRI2 or Present, so valid client targetmsc 6534 values, e.g., requesting a vblank event > 1000 vblanks in the future, 6535 triggered the offset correction in arbitrarily wrong ways, leading to 6536 wrong msc values being returned and thereby vblank events queued to the 6537 kernel for the wrong time. This causes glXSwapBuffersMscOML and 6538 glXWaitForMscOML to swap / return immediately whenever a swap/wait in > 6539 1000 vblanks is requested. 6540 6541 The original code was also written to only deal with 32 bit mscs, but 6542 server 1.20 modesetting ddx can now use new Linux 4.15+ kernel vblank 6543 api to process true 64 bit msc's, which may confuse the heuristic even 6544 more due to 32 bit integer truncation/wrapping. 6545 6546 This code caused various problems in the intel-ddx in the past since 6547 year 2013, and was removed there in 2015 by Chris Wilson in commit 6548 42ebe2ef9646be5c4586868cf332b4cd79bb4618: 6549 6550 " uxa: Remove the filtering of bogus Present MSC values 6551 6552 If the intention was to filter the return values from the kernel, the 6553 filtering would have been applied to the kernel values and not to the 6554 incoming values from Present. This filtering introduces crazy integer 6555 promotion and truncation bugs all because Present feeds garbage into its 6556 vblank requests. 6557 6558 " 6559 6560 Indeed, i found a Mesa bug yesterday which can cause Mesa's 6561 PresentPixmap request to spuriously feed garbage targetMSC's into the 6562 driver under some conditions. However, while other video drivers seem to 6563 cope relatively well with that, modesetting ddx causes KDE-5's 6564 plasmashell to lock up badly quite frequently, and my suspicion is that 6565 the code removed in this commit is one major source of the extra 6566 fragility. 6567 6568 Also my own tests fail for any swap scheduled more than 1000 vblanks 6569 into the future, which is not uncommon for some scientific applications. 6570 6571 Iow. modesetting's swap scheduling seems to be more robust without this 6572 function afaics. 6573 6574 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 6575 Reviewed-by: Adam Jackson <ajax@redhat.com> 6576 Cc: Chris Wilson <chris@chris-wilson.co.uk> 6577 Cc: Keith Packard <keithp@keithp.com> 6578 Tested-by: Mike Lothian <mike@fireburn.co.uk> 6579 6580commit f5ded22e14e2e15390eff8e01ce32de496ae0e86 6581Author: Aaron Plattner <aplattner@nvidia.com> 6582Date: Fri May 4 10:48:17 2018 -0700 6583 6584 meson: Set XCONFIGFILE to 'xorg.conf' instead of '/etc/xorg.conf' 6585 6586 The autoconf build hard-codes XCONFIGFILE to just 'xorg.conf': 6587 6588 XF86CONFIGFILE="xorg.conf" 6589 AC_DEFINE_DIR(XCONFIGFILE, XF86CONFIGFILE, [Name of configuration file]) 6590 6591 Later, the X server passes that into DoSubstitution() which expands the path: 6592 6593 DoSubstitution(template="/etc/X11/%X", ..., XConfigFile="xorg.conf") 6594 6595 This returns "/etc/X11/xorg.conf". 6596 6597 The Meson build, on the other hand, sets XCONFIGFILE to 6598 join_paths(get_option('sysconfdir'), 'xorg.conf'). If sysconfdir is /etc, this 6599 results in '/etc/xorg.conf', resulting in DoSubstitution returning 6600 '/etc/X11/etc/xorg.conf'. 6601 6602 Fix this by just hard-coding XCONFIGFILE to 'xorg.conf'. 6603 6604 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 6605 Reviewed-by: Thierry Reding <treding@nvidia.com> 6606 6607commit 1a3e4a2f6722048c5c7c4c1a9d6748e68a895a3e 6608Author: Aaron Plattner <aplattner@nvidia.com> 6609Date: Fri May 4 15:09:22 2018 -0700 6610 6611 meson: Define DEFAULT_LIBRARY_PATH as join_paths(get_option('prefix'), get_option('libdir')) 6612 6613 'libdir' defaults to 'lib', so running X -showDefaultLibPath just prints 'lib' 6614 instead of '/usr/lib' or '/usr/local/lib'. Use joint_paths() to get the correct 6615 full path. 6616 6617 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 6618 Reviewed-by: Thierry Reding <treding@nvidia.com> 6619 6620commit 1dcd784a677c58e77c368a676fd9dbd3159db94f 6621Author: Anuj Phogat <anuj.phogat@gmail.com> 6622Date: Thu May 3 11:29:11 2018 -0700 6623 6624 dri2: Sync i965_pci_ids.h from mesa 6625 6626 Copied from Mesa with no modifications. 6627 6628 Gives us Cofeelake platform names updates and sync on Kaby Lake, 6629 Ice Lake PCI IDs. 6630 6631 Acked-by: Kenneth Graunke <kenneth@whitecape.org> 6632 Signed-off-by: Anuj Phogat <anuj.phogat@gmail.com> 6633 6634commit 4191b59bd5458ea3a8b0d5272bdcd250fee1e129 6635Author: Adam Jackson <ajax@redhat.com> 6636Date: Wed May 2 15:09:52 2018 -0400 6637 6638 meson: Fix build with three-component version numbers 6639 6640 Otherwise: 6641 6642 include/meson.build:5:0: ERROR: Index 3 out of bounds of array of size 3. 6643 6644 Signed-off-by: Adam Jackson <ajax@redhat.com> 6645 6646commit 531e1648fc0da03816ecc57481479e6fc6a45ac3 6647Author: Adam Jackson <ajax@redhat.com> 6648Date: Wed Apr 25 13:39:18 2018 -0400 6649 6650 gitlab-ci: Add for gitlab.freedesktop.org 6651 6652 Looks quite a bit like the travis path, doesn't it? Still, nice to not 6653 rely on an external service if we don't have to. 6654 6655 Signed-off-by: Adam Jackson <ajax@redhat.com> 6656 6657commit 22285a6f1c5fa6ba22f4214c4ccc09e27d1925f6 6658Author: Roman Gilg <subdiff@gmail.com> 6659Date: Fri Apr 20 17:10:06 2018 +0200 6660 6661 present: fix msc offset calculation in window mode 6662 6663 Instead of getting the current msc value from the window, which might be 6664 different to old one directly take the last saved msc value saved in 6665 the window_priv struct. 6666 6667 Signed-off-by: Roman Gilg <subdiff@gmail.com> 6668 Reviewed-by: Adam Jackson <ajax@redhat.com> 6669 6670commit d7297b00444b0e2cd936fbfb08206a575ab8c29d 6671Author: Chris Wilson <chris@chris-wilson.co.uk> 6672Date: Mon Apr 30 15:19:53 2018 -0700 6673 6674 randr: Account for panning and transforms when constraining the cursor 6675 6676 commit 56c90e29f04727c903bd0f084d23bf44eb1a0a11 [1.10.99.901] 6677 Author: Adam Jackson <ajax@redhat.com> 6678 Date: Mon Nov 15 14:29:14 2010 -0500 6679 6680 randr: Add RRConstrainCursorHarder 6681 6682 introduced a regression as it ignored the effect of panning and 6683 transforms upon the crtc bounds. The result was that the cursor would be 6684 constrained to the visible area even though the panning arena was much 6685 bigger, or the cursor was constrained to a region that did not even 6686 match the visible area when the output was transformed or reflected. 6687 6688 This supercedes the hack introduced by 6689 commit 1bf81af4a6be1113bcc3b940ab264d5c9e0f0c5d [1.12.99.904] 6690 Author: Rui Matos <tiagomatos@gmail.com> 6691 Date: Mon Jul 30 14:32:12 2012 -0400 6692 6693 xf86RandR12: Don't call ConstrainCursorHarder() if panning is enabled 6694 which disabled the cursor constraints if a panning mode was active, but 6695 did not fix the regression with arbitrary output transforms. 6696 6697 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 6698 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=39949 6699 Reviewed-by: Adam Jackson <ajax@redhat.com> 6700 Cc: Rui Matos <tiagomatos@gmail.com> 6701 6702commit a843c61456b2046fb1780415f17162b503243af7 6703Author: Roman Gilg <subdiff@gmail.com> 6704Date: Fri Apr 20 17:07:42 2018 +0200 6705 6706 xwayland: restrict present cleanup to presenting and top parent window 6707 6708 Clean up only if the request points to the presenting window or its top 6709 parent window. 6710 6711 Since in this case all events are removed unconditionally, always stop 6712 the timer. 6713 6714 Reviewed-by: Adam Jackson <ajax@redhat.com> 6715 Signed-off-by: Roman Gilg <subdiff@gmail.com> 6716 6717commit 975d3a509667579b81d88616b91491472059354d 6718Author: Adam Jackson <ajax@redhat.com> 6719Date: Mon Apr 30 14:10:26 2018 -0400 6720 6721 xwayland: Avoid using epoxy_has_egl() 6722 6723 There's no real point - if we don't have EGL then the extension check is 6724 also going to fail - and the entrypoint is new in 1.5.0, which we don't 6725 need to require yet. 6726 6727 Signed-off-by: Adam Jackson <ajax@redhat.com> 6728 Reviewed-by: Mario Kleiner <mario.kleiner.de@gmail.com> 6729 6730commit d6f2272f44ce1a7a96ae04fdbde8708514c60771 6731Author: Adam Jackson <ajax@redhat.com> 6732Date: Mon Apr 30 14:07:44 2018 -0400 6733 6734 meson: Bump version number here too 6735 6736 Signed-off-by: Adam Jackson <ajax@redhat.com> 6737 6738commit 6cace4990abc2386b6ea68536b321994d264c295 6739Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 6740Date: Thu Apr 26 11:04:15 2018 -0400 6741 6742 modesetting: Fix GBM objects leak when checking for flip 6743 6744 GBM objects were never destroyed after looking for format and 6745 modifier compatibility when deciding whether flipping or copying 6746 a presented pixmap. 6747 6748 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106106 6749 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 6750 6751commit e00ada9fbbddf4570fe38c3ceeb922af7afcfd1a 6752Author: Mario Kleiner <mario.kleiner.de@gmail.com> 6753Date: Mon Apr 30 09:06:09 2018 +0200 6754 6755 glamor: Don't fail in glamor_get_formats if not dmabuf_capable. 6756 6757 If dmabuf_capable is false, because the server "dmabuf_capable" 6758 debug flag isn't set, treat it as successfull query with zero 6759 returned formats, instead of failure. 6760 6761 This allows the servers cache_formats_and_modifiers() function 6762 to cache the fact that formats are not supported during the 6763 current server generation, instead of pointless retesting at 6764 every invocation. 6765 6766 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 6767 Reviewed-by: Adam Jackson <ajax@redhat.com> 6768 6769commit 55db3c9cfc1bc92a53b75b99e8fa49a32d7efe63 6770Author: Mario Kleiner <mario.kleiner.de@gmail.com> 6771Date: Mon Apr 30 09:06:08 2018 +0200 6772 6773 dri3: Robustly clamp to 1.0 if not all screens support 1.2 6774 6775 Checking for dri3_screen_info_rec.version >= 2 is insufficient, 6776 as some shipping drivers, e.g., intel-ddx, nouveau-ddx, set the 6777 version to DRI3_SCREEN_INFO_VERSION, ie. to whatever version the 6778 installed servers headers define. On server 1.20 that would 6779 be version 2, but the drivers still don't support the v1.2 6780 hooks. Make sure all hooks are defined before reporting v1.2. 6781 6782 Also make clamping of reported version to minimum of client 6783 or server robust against possible future clients with possible 6784 majorVersion >= 2.0. 6785 6786 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 6787 Reviewed-by: Adam Jackson <ajax@redhat.com> 6788 Cc: Daniel Stone <daniels@collabora.com> 6789 6790commit fbc5c5cd532962d2df96a04e5c8cc0d21b47b08a 6791Author: Mario Kleiner <mario.kleiner.de@gmail.com> 6792Date: Mon Apr 30 09:06:07 2018 +0200 6793 6794 dri3: Fix error handling in dri3_buffer_from_pixmap request. 6795 6796 The old info->fd_from_pixmap() driver hook, which is 6797 preferentially used in dri3_fd_from_pixmap(), can return 6798 error codes other than -1, e.g., -EINVAL (-22) on nouveau-ddx. 6799 6800 Not handling the error causes a broken/corrupted X-Connection 6801 resulting from a failed request. 6802 6803 This fixes failure of sddm-greeter to start up under nouveau-ddx 6804 with DRI3 enabled and DRI3 protocol version properly clamped 6805 to 1.0 by the server (see followup patch). 6806 6807 Fixes: 75bba3aedcb0 ("dri3: Use single-FD screen call for single-FD request") 6808 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 6809 Reviewed-by: Adam Jackson <ajax@redhat.com> 6810 Cc: Daniel Stone <daniels@collabora.com> 6811 6812commit 19d006ee3db4b4635ef9ef5c92562f3ffeddb305 6813Author: Adam Jackson <ajax@redhat.com> 6814Date: Tue Apr 24 16:08:26 2018 -0400 6815 6816 dri3: Clamp to 1.0 if not all screens support 1.2 6817 6818 Signed-off-by: Adam Jackson <ajax@redhat.com> 6819 Reviewed-by: Daniel Stone <daniels@collabora.com> 6820 6821commit 352a5ac87fd344936b759a5766eb74271e7d295d 6822Author: Mario Kleiner <mario.kleiner.de@gmail.com> 6823Date: Tue Apr 24 10:17:26 2018 +0200 6824 6825 dri3: Fix DRI3.2 support for drivers other than modesetting-ddx. 6826 6827 Both xf86-video-intel and xf86-video-nouveau cause OpenGL clients to 6828 fail when used with DRI3 on server 1.20 with Mesa 18.1. 6829 6830 Reason is that the servers DRI3 version is now unconditionally reported 6831 as DRI3 1.2 to 1.2 capable clients. This causes clients using Mesa 18.1 6832 to use the new DRI 3.2 requests DRI3GetSupportedModifiers, 6833 DRI3PixmapFromBuffers, etc. Drivers other than modesetting-ddx do not 6834 support the needed hooks like info->pixmap_from_fds or 6835 info->get_formats, info->get_modifiers. Unfortunately we can't simply 6836 report the servers DRI3 version as 1.0 in this case, as the reported 6837 version can not be specific to a X-Screen, and different screens may 6838 have drivers with different capabilities. 6839 6840 Luckily the server has fallbacks to ->pixmap_from_fd, ->fd_from_pixmap, 6841 and simply reporting an empty set of supported modifiers for the 6842 DRI3GetSupportedModifiers request if the ddx doesn't support DRI 3.2. 6843 6844 Clients like Mesa 18.1's dri3 loader respond to the empty set of 6845 reported modifiers by falling back to a dri driver selected buffer 6846 format (image->createImageWithModifiers responds to a NULL modifier_list 6847 by acting like ->createImage()). This works, but Mesa 18.1 will still 6848 try to use the DRI3PixmapFromBuffers request to create the corresponding 6849 pixmap, just passing in a modifier that corresponds to whatever tiling 6850 the dri driver selected by default. To prevent this request - and 6851 thereby the client - from failing with a BadImplementation error, remove 6852 the check for modifier == DRM_MOD_FORMAT_INVALID in the pixmap_from_fd 6853 fallback path of dri3_pixmap_from_fds() and trust that if we hit the 6854 fallback path then the client will have passed a buffer with some driver 6855 specific default tiling that can be handled by pixmap_from_fd. 6856 6857 Another approach would be for Mesa's dri3 loader to keep track how a 6858 buffer was created (with explicit modifiers or not), and then call 6859 DRI3PixmapFromBuffers or DRI3PixmapFromBuffer, but then any future DRI3 6860 client implementation would need to be fixed, so the server side is 6861 probably the better place for this. 6862 6863 Tested on Intel Ivybridge and NVidia Pascal. 6864 6865 Fixes: 6e7c40f62db6 ("dri3: Add multi-planar/modifier buffer requests") 6866 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 6867 Cc: Daniel Stone <daniels@collabora.com> 6868 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 6869 Tested-by: Mike Lothian <mike@fireburn.co.uk> 6870 Reviewed-by: Adam Jackson <ajax@redhat.com> 6871 6872commit c6ab21022ce876f9c1409db4bb1967134f9f6dbe 6873Author: Adam Jackson <ajax@redhat.com> 6874Date: Tue Apr 24 17:03:07 2018 -0400 6875 6876 xserver 1.20 RC5 6877 6878 Signed-off-by: Adam Jackson <ajax@redhat.com> 6879 6880commit c593d843f6305dd8bc5fa7762273d319a223abc4 6881Author: Daniel Stone <daniels@collabora.com> 6882Date: Tue Apr 24 21:27:47 2018 +0100 6883 6884 dri3: Don't call vfuncs on old DRI3 screens 6885 6886 Only call the get_supported_modifiers vfunc if the DRI3 screen struct is 6887 sufficiently new. 6888 6889 Signed-off-by: Daniel Stone <daniels@collabora.com> 6890 Reviewed-by: Adam Jackson <ajax@redhat.com> 6891 6892commit 54ac09717cd8c49259f53a4a227d903ebe8e0a32 6893Author: Lyude Paul <lyude@redhat.com> 6894Date: Fri Apr 20 14:38:05 2018 -0400 6895 6896 xwayland: Add glamor egl_backend for EGLStreams 6897 6898 This adds initial support for displaying Xwayland applications through 6899 the use of EGLStreams and nvidia's custom wayland protocol by adding 6900 another egl_backend driver. This also adds some additional egl_backend 6901 hooks that are required to make things work properly. 6902 6903 EGLStreams work a lot differently then the traditional way of handling 6904 buffers with wayland. Unfortunately, there are also a LOT of various 6905 pitfalls baked into it's design that need to be explained. 6906 6907 This has a very large and unfortunate implication: direct rendering is, 6908 for the time being at least, impossible to do through EGLStreams. The 6909 main reason being that the EGLStream spec mandates that we lose the 6910 entire color buffer contents with each eglSwapBuffers(), which goes 6911 against X's requirement of not losing data with pixmaps. no way to use 6912 an allocated EGLSurface as the storage for glamor rendering like we do 6913 with GBM, we have to rely on blitting each pixmap to it's respective 6914 EGLSurface producer each frame. In order to pull this off, we add two 6915 different additional egl_backend hooks that GBM opts out of 6916 implementing: 6917 6918 - egl_backend.allow_commits for holding off displaying any EGLStream 6919 backed pixmaps until the point where it's stream is completely 6920 initialized and ready for use 6921 - egl_backend.post_damage for blitting the content of the EGLStream 6922 surface producer before Xwayland actually damages and commits the 6923 wl_surface to the screen. 6924 6925 The other big pitfall here is that using nvidia's wayland-eglstreams 6926 helper library is also not possible for the most part. All of it's API 6927 for creating and destroying streams rely on being able to perform a 6928 roundtrip in order to bring each stream to completion since the wayland 6929 compositor must perform it's job of connecting a consumer to each 6930 EGLstream. Because Xwayland has to potentially handle both responding to 6931 the wayland compositor and it's own X clients, the situation of the 6932 wayland compositor being one of our X clients must be considered. If we 6933 perform a roundtrip with the Wayland compositor, it's possible that the 6934 wayland compositor might currently be connected to us as an X client and 6935 thus hang while both Xwayland and the wayland compositor await responses 6936 from eachother. To avoid this, we work directly with the wayland 6937 protocol and use wl_display_sync() events along with release() events to 6938 set up and destroy EGLStreams asynchronously alongside handling X 6939 clients. 6940 6941 Additionally, since setting up EGLStreams is not an atomic operation we 6942 have to take into consideration the fact that an EGLStream can 6943 potentially be created in response to a window resize, then immediately 6944 deleted due to another pending window resize in the same X client's 6945 pending reqests before Xwayland hits the part of it's event loop where 6946 we read from the wayland compositor. To make this even more painful, we 6947 also have to take into consideration that since EGLStreams are not 6948 atomic that it's possible we could delete wayland resources for an 6949 EGLStream before the compositor even finishes using them and thus run 6950 into errors. So, we use quite a bit of tracking logic to keep EGLStream 6951 objects alive until we know the compositor isn't using them (even if 6952 this means the stream outlives the pixmap it backed). 6953 6954 While the default backend for glamor remains GBM, this patch exists for 6955 users who have had to deal with the reprecussion of their GPU 6956 manufacturers ignoring the advice of upstream and the standardization of 6957 GBM across most major GPU manufacturers. It is not intended to be a 6958 final solution to the GBM debate, but merely a baindaid so our users 6959 don't have to suffer from the consequences of companies avoiding working 6960 upstream. New drivers are strongly encouraged not to use this as a 6961 backend, and use GBM like everyone else. We even spit this out as an 6962 error from Xwayland when using the eglstream backend. 6963 6964 Signed-off-by: Lyude Paul <lyude@redhat.com> 6965 Acked-by: Daniel Stone <daniels@collabora.com> 6966 Reviewed-by: Adam Jackson <ajax@redhat.com> 6967 6968commit 994f781007079176e0247ee64af5696d34debcde 6969Author: Lyude Paul <lyude@redhat.com> 6970Date: Fri Apr 20 14:38:04 2018 -0400 6971 6972 xwayland: Add xwayland-config.h 6973 6974 Just a small autogenerated header that will soon contain more then just 6975 one macro. 6976 6977 Signed-off-by: Lyude Paul <lyude@redhat.com> 6978 Reviewed-by: Daniel Stone <daniels@collabora.com> 6979 Reviewed-by: Adam Jackson <ajax@redhat.com> 6980 6981commit 1545e2dbadcb147d7d52b546d053149de866a031 6982Author: Lyude Paul <lyude@redhat.com> 6983Date: Fri Apr 20 14:38:03 2018 -0400 6984 6985 xwayland: Decouple GBM from glamor 6986 6987 This takes all of the gbm related code in wayland-glamor.c and moves it 6988 into it's own EGL backend for Xwayland, xwayland-glamor-gbm.c. 6989 Additionally, we add the egl_backend struct into xwl_screen in order to 6990 provide hooks for alternative EGL backends such as nvidia's EGLStreams. 6991 6992 Signed-off-by: Lyude Paul <lyude@redhat.com> 6993 Reviewed-by: Lyude Paul <lyude@redhat.com> 6994 Reviewed-by: Daniel Stone <daniels@collabora.com> 6995 Reviewed-by: Adam Jackson <ajax@redhat.com> 6996 6997commit d2d664df974ac5a55d5819f0379fcdac05d22fa3 6998Author: Adam Jackson <ajax@redhat.com> 6999Date: Mon Apr 23 17:21:09 2018 -0400 7000 7001 vfb: Fix man page in re depth 7002 7003 32 is not a valid depth, and the default is now 24 not 8. 7004 7005 Signed-off-by: Adam Jackson <ajax@redhat.com> 7006 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 7007 7008commit 79a7137557d33d4e92713ae8cabe838c44fea488 7009Author: Adam Jackson <ajax@redhat.com> 7010Date: Mon Apr 23 16:19:15 2018 -0400 7011 7012 glx: Require depth > 12 for GLX visuals 7013 7014 fb is happy to do TrueColor to 8bpp drawables, but mesa is not. Depth 12 7015 is the biggest pseudocolor anyone ever really did, and 15 is the least 7016 truecolor. 7017 7018 Without this Xvfb at depth 8 would "have" GLX, but no vendors 7019 would actually back any of the screens. libGL will attempt to call 7020 GLXQueryServerString to figure out the GLX version, and vnd will throw 7021 an error because there's no vendor to dispatch that to, and then clients 7022 crash. 7023 7024 Signed-off-by: Adam Jackson <ajax@redhat.com> 7025 7026commit 818885e6198cf2883155cb3d2e22c8c7bc4239fb 7027Author: Adam Jackson <ajax@redhat.com> 7028Date: Mon Apr 23 16:19:14 2018 -0400 7029 7030 vnd: Disable GLX if no vendors successfully initialized 7031 7032 Signed-off-by: Adam Jackson <ajax@redhat.com> 7033 7034commit fc25bceb515e5c18eecdebf5933c3e05cdac1a5a 7035Author: Adam Jackson <ajax@redhat.com> 7036Date: Mon Apr 23 16:19:13 2018 -0400 7037 7038 dix: Allow an extension to disable itself 7039 7040 GLX registers an extension before we know if there are any screens that 7041 can actually do it. It's inconvenient to shrink the extension list, so 7042 instead allow the extension to simply zero out its base opcode to 7043 indicate that it needed to panic and disable itself. 7044 7045 Signed-off-by: Adam Jackson <ajax@redhat.com> 7046 7047commit 73a1cb9c92c936c2c1ae3d69fed743e21916d687 7048Author: Adam Jackson <ajax@redhat.com> 7049Date: Mon Apr 23 16:19:12 2018 -0400 7050 7051 dix: Factor out extension availability check 7052 7053 Signed-off-by: Adam Jackson <ajax@redhat.com> 7054 7055commit 9d5af632fde0373babfa32e66a59cfbf26ed7e5d 7056Author: Adam Jackson <ajax@redhat.com> 7057Date: Mon Apr 23 15:21:14 2018 -0400 7058 7059 animcur: Fix crash when removing a master device 7060 7061 Reproducer: 7062 7063 $ Xvfb -ac -noreset :1 & 7064 $ DISPLAY=:1 xinput create-master touch1 7065 $ DISPLAY=:1 xinput remove-master "touch1 pointer" 7066 7067 Bugzilla: https://bugs.freedesktop.org/105761 7068 Signed-off-by: Adam Jackson <ajax@redhat.com> 7069 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 7070 7071commit 8275903956d7876519fa55fb805669dc301bf243 7072Author: Thomas Klausner <wiz@NetBSD.org> 7073Date: Mon Apr 23 14:56:17 2018 -0400 7074 7075 sdksyms: Cope with __pid_t and __uint32_t 7076 7077 Kludge sdksyms.c generator to not fail on GetClientPid. 7078 It returns pid_t which on NetBSD is #define pid_t __pid_t 7079 This slightly alters the GCC preprocessor output which this fragile 7080 code could not deal with when using GCC 5+ 7081 7082 Signed-off-by: Adam Jackson <ajax@redhat.com> 7083 7084commit e29d783278608544a208f211231a6a3ad2f8fa09 7085Author: Mario Kleiner <mario.kleiner.de@gmail.com> 7086Date: Fri Apr 20 19:59:40 2018 +0200 7087 7088 modesetting: Only use modifiers on kms drivers which do support them. 7089 7090 Use the DRM_CAP_ADDFB2_MODIFIERS query to make sure the kms 7091 driver supports modifiers in the addfb2 ioctl, and fall back 7092 to addfb ioctl without modifiers if modifiers are unsupported. 7093 7094 E.g., as of Linux 4.17, nouveau-kms so far does not suppport 7095 modifiers and gets angry if drmModeAddFB2WithModifiers() is 7096 called (-> failure to set a video mode -> blank screen), but 7097 Mesa's nvc0+ gallium driver causes gbm_bo_get_modifier() to 7098 return a valid modifier by translating the default tiling of 7099 bo's created via gbm_bo_create() into a modifier other than 7100 DRM_FORMAT_MOD_INVALID (see Mesa's nvc0_miptree_get_modifier()). 7101 7102 Testing for != DRM_FORMAT_MOD_INVALID is apparently not 7103 sufficient for safe use of drmModeAddFB2WithModifiers. 7104 7105 Bonus: Handle potential failure of populate_format_modifiers(). 7106 7107 The required DRM_CAP is defined since libdrm v2.4.65, and we 7108 require v2.4.89+ for the server, so we can use it unconditionally. 7109 7110 Tested on intel-kms, radeon-kms, nouveau-kms. Fixes failure on 7111 NVidia Pascal. 7112 7113 Fixes: 2f807c2324b4 ("modesetting: Add support for multi-plane pixmaps when page-flipping") 7114 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 7115 Cc: Daniel Stone <daniels@collabora.com> 7116 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7117 Reviewed-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7118 7119commit fe4d1876b4f01c0b0e1916d548c398789f196164 7120Author: Lyude Paul <lyude@redhat.com> 7121Date: Wed Apr 18 17:57:54 2018 -0400 7122 7123 meson: Fix indenting in glx/meson.build 7124 7125 No functional changes, just fixing a tabs vs. space error I noticed 7126 7127 Signed-off-by: Lyude Paul <lyude@redhat.com> 7128 Reviewed-by: Adam Jackson <ajax@redhat.com> 7129 7130commit 4e28a6a223c4f9d0f5defe0313a94e22e0416787 7131Author: Lyude Paul <lyude@redhat.com> 7132Date: Wed Apr 18 18:09:15 2018 -0400 7133 7134 meson: Ensure we always build Xext/hashtable.c for glx 7135 7136 Seems that while glxvnd relies on some of the hashtable functions in 7137 Xext, we only build hashtable support for Xext if we're also building 7138 the res extension. This leads to some errors if you try to build glx 7139 without res enabled: 7140 7141 glx/liblibglxvnd.a(vndcmds.c.o): In function `LookupVendorPrivDispatch': 7142 /home/lyudess/Projects/xserver/glx/vndcmds.c:65: undefined reference to `ht_find' 7143 /home/lyudess/Projects/xserver/glx/vndcmds.c:67: undefined reference to `ht_add' 7144 glx/liblibglxvnd.a(vndcmds.c.o): In function `GlxDispatchInit': 7145 /home/lyudess/Projects/xserver/glx/vndcmds.c:405: undefined reference to `ht_generic_compare' 7146 /home/lyudess/Projects/xserver/glx/vndcmds.c:405: undefined reference to `ht_generic_hash' 7147 /home/lyudess/Projects/xserver/glx/vndcmds.c:405: undefined reference to `ht_create' 7148 glx/liblibglxvnd.a(vndcmds.c.o): In function `GlxDispatchReset': 7149 /home/lyudess/Projects/xserver/glx/vndcmds.c:468: undefined reference to `ht_destroy' 7150 collect2: error: ld returned 1 exit status 7151 ninja: build stopped: subcommand failed. 7152 7153 So, make sure that hashtable.c gets both for both glx and res 7154 7155 Reviewed-by: Adam Jackson <ajax@redhat.com> 7156 Signed-off-by: Lyude Paul <lyude@redhat.com> 7157 7158commit 3b4671f9e9c85f23e7593652e1482b11dc3ad4af 7159Author: Olivier Fourdan <ofourdan@redhat.com> 7160Date: Thu Apr 19 09:13:23 2018 +0200 7161 7162 xwayland: Clean up all frame callbacks 7163 7164 Regardless of the order we un-realize windows. 7165 7166 Suggested-by: Roman Gilg <subdiff@gmail.com> 7167 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 7168 Reviewed-by: Roman Gilg <subdiff@gmail.com> 7169 7170commit 8b8f9007cc56fdcee5479876e88e06b7f8b6de35 7171Author: Olivier Fourdan <ofourdan@redhat.com> 7172Date: Wed Apr 18 16:02:02 2018 +0200 7173 7174 xwayland: avoid using freed xwl_window on unrealize 7175 7176 xwl_unrealize_window() would use freed xwl_window which can lead to 7177 various memory corruption and crashes, as reported by valgrind: 7178 7179 Invalid read of size 8 7180 at 0x42C802: xwl_present_cleanup (xwayland-present.c:84) 7181 by 0x42BA67: xwl_unrealize_window (xwayland.c:601) 7182 by 0x541EE9: compUnrealizeWindow (compwindow.c:285) 7183 by 0x57E1FA: UnrealizeTree (window.c:2816) 7184 by 0x581189: UnmapWindow (window.c:2874) 7185 by 0x54EB26: ProcUnmapWindow (dispatch.c:879) 7186 by 0x554B7D: Dispatch (dispatch.c:479) 7187 by 0x558BE5: dix_main (main.c:276) 7188 by 0x7C4B1BA: (below main) (libc-start.c:308) 7189 Address 0xf520f60 is 96 bytes inside a block of size 184 free'd 7190 at 0x4C2EDAC: free (vg_replace_malloc.c:530) 7191 by 0x42B9FB: xwl_unrealize_window (xwayland.c:624) 7192 by 0x541EE9: compUnrealizeWindow (compwindow.c:285) 7193 by 0x57E1FA: UnrealizeTree (window.c:2816) 7194 by 0x581189: UnmapWindow (window.c:2874) 7195 by 0x54EB26: ProcUnmapWindow (dispatch.c:879) 7196 by 0x554B7D: Dispatch (dispatch.c:479) 7197 by 0x558BE5: dix_main (main.c:276) 7198 by 0x7C4B1BA: (below main) (libc-start.c:308) 7199 Block was alloc'd at 7200 at 0x4C2FB06: calloc (vg_replace_malloc.c:711) 7201 by 0x42B307: xwl_realize_window (xwayland.c:488) 7202 by 0x541E59: compRealizeWindow (compwindow.c:268) 7203 by 0x57DA40: RealizeTree (window.c:2617) 7204 by 0x580B28: MapWindow (window.c:2694) 7205 by 0x54EA2A: ProcMapWindow (dispatch.c:845) 7206 by 0x554B7D: Dispatch (dispatch.c:479) 7207 by 0x558BE5: dix_main (main.c:276) 7208 by 0x7C4B1BA: (below main) (libc-start.c:308) 7209 7210 This is because UnrealizeTree() traverses the tree from top to bottom, 7211 which invalidates the assumption that if the Window doesn't feature an 7212 xwl_window on its own, it's the xwl_window of its first ancestor with 7213 one. 7214 7215 This reverts commit 82df2ce3 7216 7217 Reviewed-by: Adam Jackson <ajax@redhat.com> 7218 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 7219 7220commit 12a6b189fb17894d2c3851b70a396bbf41f444c6 7221Author: David Woodhouse <dwmw2@infradead.org> 7222Date: Tue Apr 17 22:22:21 2018 +0100 7223 7224 glamor: fix glamor_xv_query_image_attributes() for odd-width images 7225 7226 Images which are one pixel wider than a multiple of 8 are being handled 7227 incorrectly. Other drivers round up the width to a multiple of two 7228 before they start calculating. Do the same. 7229 7230 https://bugzilla.gnome.org/show_bug.cgi?id=795235 7231 7232 Reviewed-by: Adam Jackson <ajax@redhat.com> 7233 Signed-off-by: David Woodhouse <dwmw2@infradead.org> 7234 7235commit ac7a4bf44c68c5f323375974b208d4530fb5b60f 7236Author: Chris Wilson <chris@chris-wilson.co.uk> 7237Date: Sun Apr 15 15:40:03 2018 +0100 7238 7239 os/WaitFor: Check timers on every iteration 7240 7241 Currently we only check timer expiry if there are no client fd (or 7242 other input) waiting to be serviced. This makes it very easy to starve 7243 the timers with long request queues, and so miss critical timestamps. 7244 7245 The timer subsystem is just another input waiting to be serviced, so 7246 evaluate it on every loop like all the others, at the cost of calling 7247 GetTimeInMillis() slightly more frequently. (A more invasive and likely 7248 OS specific alternative would be to move the timer wheel to the local 7249 equivalent of timerfd, and treat it as an input fd to the event loop 7250 exactly equivalent to all the others, and so also serviced on every 7251 pass. The trade-off being that the kernel timer wheel is likely more 7252 efficiently integrated with epoll, but individual updates to each timer 7253 would then require syscalls.) 7254 7255 Reviewed-by: Peter Harris <pharris@opentext.com> 7256 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 7257 7258commit 78b6f940217c127f0f345b7710aa5994c6ded99c 7259Author: Adam Jackson <ajax@redhat.com> 7260Date: Mon Apr 16 14:08:27 2018 -0400 7261 7262 modesetting: Fix inverted check in dri2 WaitMSC 7263 7264 ms_queue_vblank() returns false on failure. 7265 7266 Reported-by: Chris Wilson <chris@chris-wilson.co.uk> 7267 Signed-off-by: Adam Jackson <ajax@redhat.com> 7268 Reviewed-by: Frank Binns <frank.binns@imgtec.com> 7269 Tested-by: Mike Lothian <mike@fireburn.co.uk> 7270 7271commit 4e92c51ce4d6ab50c1507a23a01c6be0d1954d79 7272Author: Mario Kleiner <mario.kleiner.de@gmail.com> 7273Date: Mon Apr 16 08:14:47 2018 +0200 7274 7275 dri3: Fix dri3_buffers_from_pixmap request. 7276 7277 Sending pixmap depth and bpp was omitted, so the Mesa 7278 X11 + EGL + DRI3 side of things always failed to 7279 dri3_create_image_khr_pixmap_from_buffers(), which led 7280 to failure of X11 + EGL compositing under DRI3 under, 7281 e.g., KDE Plasma 5. 7282 7283 Fixes: 6e7c40f62db6 ("dri3: Add multi-planar/modifier buffer requests") 7284 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 7285 Cc: Daniel Stone <daniels@collabora.com> 7286 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7287 Reviewed-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7288 Tested-by: Mike Lothian <mike@fireburn.co.uk> 7289 7290commit a98a95b798a0062783ae6fd2135ad488af5efcfd 7291Author: Matt Turner <mattst88@gmail.com> 7292Date: Sun Apr 15 23:37:45 2018 -0700 7293 7294 modesetting: Move GBM code inside #ifdef GLAMOR_HAS_GBM 7295 7296 Fixes a compilation error without Glamor. 7297 7298 Bugzilla: https://bugs.gentoo.org/653288 7299 Signed-off-by: Matt Turner <mattst88@gmail.com> 7300 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 7301 7302commit c3ae963a7b19e05b8296aae8b6354587756d48fa 7303Author: Olivier Fourdan <ofourdan@redhat.com> 7304Date: Mon Apr 16 09:39:09 2018 +0200 7305 7306 xwayland: Fix build without glamor 7307 7308 Present support in Xwayland relies on glamor, make sure Xwayland can 7309 be built without glamor by moving references to Present code inside 7310 the conditional GLAMOR_HAS_GBM. 7311 7312 Reported-by: Matt Turner <mattst88@gmail.com> 7313 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 7314 Reviewed-by: Matt Turner <mattst88@gmail.com> 7315 Reviewed-by: Roman Gilg <subdiff@gmail.com> 7316 7317commit bf147f67b2b7170fcc5cca07192f6b195dce85e5 7318Author: Adam Jackson <ajax@redhat.com> 7319Date: Thu Apr 12 17:24:08 2018 -0400 7320 7321 xwayland: Don't crash on WarpPointer(dest_w = None) 7322 7323 Turns out that's legal, and xts exercises it, and we crash: 7324 7325 Thread 1 "Xwayland" received signal SIGSEGV, Segmentation fault. 7326 dixGetPrivate (key=0x813660 <xwl_window_private_key>, privates=0x20) at ../../include/privates.h:122 7327 122 return (char *) (*privates) + key->offset; 7328 (gdb) bt 7329 #0 dixGetPrivate (key=0x813660 <xwl_window_private_key>, privates=0x20) at ../../include/privates.h:122 7330 #1 dixLookupPrivate (key=0x813660 <xwl_window_private_key>, privates=0x20) at ../../include/privates.h:166 7331 #2 xwl_window_of_top (window=0x0) at xwayland.c:128 7332 #3 xwl_cursor_warped_to (device=<optimized out>, screen=0x268b6e0, client=<optimized out>, window=0x0, sprite=0x300bb30, 7333 x=2400, y=1350) at xwayland.c:292 7334 #4 0x00000000005622ec in ProcWarpPointer (client=0x32755d0) at events.c:3618 7335 7336 In this case, x/y are the screen-space coordinates where the pointer 7337 ends up, and we need to look up the (X) window there. 7338 7339 Signed-off-by: Adam Jackson <ajax@redhat.com> 7340 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 7341 7342commit d7269b49c3408f93a6cb8e28ddc3dda88224620a 7343Author: Emil Velikov <emil.velikov@collabora.com> 7344Date: Wed Apr 11 18:51:16 2018 +0100 7345 7346 configure.ac: make use of wayland-scanner.pc 7347 7348 Replace the current (incorrect) assumption that wayland-scanner is 7349 located in the wayland-client prefix. Make use of the wayland_scanner 7350 variable in wayland-scanner.pc 7351 7352 It was introduced back in 2013 and we already require newer wayland bits 7353 7354 Reviewed-by: Adam Jackson <ajax@redhat.com> 7355 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7356 7357commit 0031bbad8400d35699664d18c0d8fc2e9e82d8a7 7358Author: Adam Jackson <ajax@redhat.com> 7359Date: Wed Apr 11 15:39:33 2018 -0400 7360 7361 sdksyms: Skip empty symbols 7362 7363 Apparently on NetBSD we can hit failures like this: 7364 7365 sdksyms.c:1773:15: error: expected expression before ',' token 7366 (void *) &, /* ../../dri3/dri3.h:110 */ 7367 7368 I've been unable to reproduce that locally (even in a NetBSD vm), but 7369 an obvious workaround might be to just notice empty symbol names and 7370 ignore them rather than emit invalid C code. 7371 7372 Tested-by: Thomas Klausner <wiz@netbsd.org> 7373 Signed-off-by: Adam Jackson <ajax@redhat.com> 7374 7375commit d61e516c847b9ff7e89e6be269050032070c1ad3 7376Author: Adam Jackson <ajax@redhat.com> 7377Date: Tue Apr 10 15:45:02 2018 -0400 7378 7379 xserver 1.20 RC4 7380 7381 Signed-off-by: Adam Jackson <ajax@redhat.com> 7382 7383commit a42992a4cca49cedd3930f5694c7a16e4f614b36 7384Author: Emil Velikov <emil.velikov@collabora.com> 7385Date: Mon Apr 2 16:41:24 2018 +0100 7386 7387 dri3: rework format/modifier caching 7388 7389 Cut down the unnecessary malloc/memcpy/free by utilising the explicit 7390 copy provided by the client. 7391 7392 But above all: do so, after ensuring we get valid data from the 7393 implementation. 7394 7395 Fixes: cef12efc15c ("glamor: Implement GetSupportedModifiers") 7396 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7397 Cc: Daniel Stone <daniels@collabora.com> 7398 Reviewed-by: Adam Jackson <ajax@redhat.com> 7399 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7400 7401commit 71a069fd7fbe815d386fc1b3c44cda732cff7af0 7402Author: Emil Velikov <emil.velikov@collabora.com> 7403Date: Mon Apr 2 16:41:23 2018 +0100 7404 7405 glamor: zero num_formats from the start 7406 7407 The caller may ignore the return value (will be addressed with later 7408 commit) so simply zero the count from the get-go. We're pretty much do 7409 so, in all cases but one :-\ 7410 7411 Fixes: cef12efc15c ("glamor: Implement GetSupportedModifiers") 7412 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7413 Cc: Daniel Stone <daniels@collabora.com> 7414 Reviewed-by: Adam Jackson <ajax@redhat.com> 7415 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7416 7417commit ac48724639e0a6a9e421b3b4e545d8506fd6bf5d 7418Author: Emil Velikov <emil.velikov@collabora.com> 7419Date: Mon Apr 2 16:41:22 2018 +0100 7420 7421 xwayland: zero num_formats from the start 7422 7423 The caller may ignore the return value (will be addressed with later 7424 commit) so simply zero the count from the get-go. We're pretty much do 7425 so, in all cases but one :-\ 7426 7427 Fixes: cef12efc15c ("glamor: Implement GetSupportedModifiers") 7428 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7429 Cc: Daniel Stone <daniels@collabora.com> 7430 7431 Reviewed-by: Adam Jackson <ajax@redhat.com> 7432 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7433 7434commit e2f45002fc3c53c6196935447b8fe77d0850175b 7435Author: Emil Velikov <emil.velikov@collabora.com> 7436Date: Mon Apr 2 16:41:20 2018 +0100 7437 7438 dri3: check for ::get_drawable_modifiers failure 7439 7440 Currently if the function fails, we'll fall into two false assumptions: 7441 - the the count is zero 7442 - that the storage pointer is safe for free() 7443 7444 I've just fixed the former (in glamor + xwayland) and have no 7445 plans on adding yet another workaround for the latter. 7446 7447 Simply zero both variables. Regardless if the implementation is missing 7448 the callback or it foobars with output variables (normally a bad idea). 7449 7450 Bonus points - this fixes a bug where we feed garbage to free() further 7451 down ;-) 7452 7453 Fixes: cef12efc15c ("glamor: Implement GetSupportedModifiers") 7454 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7455 Cc: Daniel Stone <daniels@collabora.com> 7456 Reviewed-by: Adam Jackson <ajax@redhat.com> 7457 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7458 7459commit 150e4b12ad160b093899107ed586aa0cb258879e 7460Author: Emil Velikov <emil.velikov@collabora.com> 7461Date: Mon Apr 2 16:41:19 2018 +0100 7462 7463 glamor: zero num_modifiers from the start 7464 7465 The caller may ignore the return value (will be addressed with later 7466 commit) so simply zero the count from the get-go. We're pretty much do 7467 so, in all cases but one :-\ 7468 7469 Fixes: cef12efc15c ("glamor: Implement GetSupportedModifiers") 7470 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7471 Cc: Daniel Stone <daniels@collabora.com> 7472 Reviewed-by: Adam Jackson <ajax@redhat.com> 7473 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7474 7475commit b36a14c0b0e7e38406622eb5ff0666a8b8bc50f4 7476Author: Emil Velikov <emil.velikov@collabora.com> 7477Date: Mon Apr 2 16:41:18 2018 +0100 7478 7479 xwayland: zero num_modifiers from the start 7480 7481 The caller may ignore the return value (will be addressed with later 7482 commit) so simply zero the count from the get-go. We're pretty much do 7483 so, in all cases but one :-\ 7484 7485 Fixes: cef12efc15c ("glamor: Implement GetSupportedModifiers") 7486 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7487 Cc: Daniel Stone <daniels@collabora.com> 7488 Reviewed-by: Adam Jackson <ajax@redhat.com> 7489 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7490 7491commit a83ceec868a6d544bc7775a753b67aa40d0d0efc 7492Author: Emil Velikov <emil.velikov@collabora.com> 7493Date: Mon Apr 2 16:41:15 2018 +0100 7494 7495 dri3: simplify dri3_open() implementation 7496 7497 Reviewed-by: Adam Jackson <ajax@redhat.com> 7498 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7499 7500commit 9a159f37e00ed47ec8cbff7c57d8787b8f5685f5 7501Author: Emil Velikov <emil.velikov@collabora.com> 7502Date: Mon Apr 2 16:41:14 2018 +0100 7503 7504 dri3: annotate fds/strides/offsets arrays as const 7505 7506 It makes it perfectly clear that we should not be modifying them. 7507 Should help highlight issues like the one fixed with previous commit. 7508 7509 Fixes: cef12efc15c ("glamor: Implement GetSupportedModifiers") 7510 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7511 Cc: Daniel Stone <daniels@collabora.com> 7512 Reviewed-by: Adam Jackson <ajax@redhat.com> 7513 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7514 7515commit 877fa0c66469628748dbd01506f15ddc4f11b849 7516Author: Emil Velikov <emil.velikov@collabora.com> 7517Date: Mon Apr 2 16:41:13 2018 +0100 7518 7519 xwayland: don't close() fds we don't own 7520 7521 The glamor_pixmap_from_fds error path erroneously closes the fds. 7522 We don't own them, plus the caller closes them after the function in 7523 called. 7524 7525 Fixes: cef12efc15c ("glamor: Implement GetSupportedModifiers") 7526 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7527 Cc: Daniel Stone <daniels@collabora.com> 7528 Reviewed-by: Adam Jackson <ajax@redhat.com> 7529 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7530 7531commit 66b632bb068672f507212b00bd313b5040bf1a39 7532Author: Emil Velikov <emil.velikov@collabora.com> 7533Date: Mon Apr 2 16:41:12 2018 +0100 7534 7535 dri3: annotate the dri3_screen_info data as const 7536 7537 dri3_screen_info is the user provide dispatch. Something that we do 7538 not and should not change. 7539 7540 When using the _ptr typecast + const the compiler barfs at us 7541 (rightfully so), so use the _rec one. 7542 7543 [Silence a new const mismatch warning too - ajax] 7544 7545 Fixes: 56313829886 ("dri3: Add DRI3 extension") 7546 Cc: Keith Packard <keithp@keithp.com> 7547 Reviewed-by: Adam Jackson <ajax@redhat.com> 7548 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 7549 7550commit c67f2eac56518163981af59f5accb7c79bc00f6a 7551Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 7552Date: Mon Apr 9 14:35:30 2018 +0200 7553 7554 dix: always send focus event on grab change 7555 7556 Focus events are useless when 'from' and 'to' are the same. But when 7557 this is the result of a (Un)GrabKeyboard request, we should always send 7558 them, including when the window manager had previously used XSetInputFocus 7559 to specify the focus on a window which happens to be now taking a grab. 7560 7561 This is notably needed for window manager using XI to always get keyboard 7562 events even during grabs, so they can determine exactly when grabbing is 7563 active. 7564 7565 Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 7566 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 7567 7568commit 14be894b3f7976c133fc186e0e3c475606bab241 7569Author: Michal Srb <msrb@suse.com> 7570Date: Tue Aug 1 10:29:03 2017 +0200 7571 7572 Xext: Fix memory leaks in hashtable. 7573 7574 Do not forget to delete key and value of every item and the hashtable itself. 7575 7576commit 1326ee0bc5eb858c3c00847b3ba65134e4ca2e2d 7577Author: Michal Srb <msrb@suse.com> 7578Date: Tue Feb 13 10:33:30 2018 +0100 7579 7580 glx: Do not call into Composite if it is disabled. 7581 7582 Otherwise X server crashes if GLX is enabled and Composite disabled. For 7583 example the compIsAlternateVisual function will try to lookup CompScreenPtr 7584 using the CompScreenPrivateKey, but that was never initialized if Composite is 7585 disabled. 7586 7587 Fixes: f84e59a4f4. ("glx: Duplicate relevant fbconfigs for compositing visuals") 7588 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104993 7589 Signed-off-by: Michal Srb <msrb@suse.com> 7590 7591commit b9764b8489cabd15b50c360cfbd799fdab0883fd 7592Author: Jon Turney <jon.turney@dronecode.org.uk> 7593Date: Tue Apr 3 16:54:00 2018 +0100 7594 7595 hw/xwin/glx: Allocate fbconfigs correctly 7596 7597 4b0a3cba fixed leaking of GLX fbconfigs, so now xwin needs to allocate them 7598 correctly (individually, rather than all at once), so they can be freed 7599 successfully. 7600 7601 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 7602 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 7603 7604commit e0a137ce5d653063604fa8d16c8498b8ac3ab3a7 7605Author: Kyle Brenneman <kbrenneman@nvidia.com> 7606Date: Fri Apr 6 12:42:33 2018 -0600 7607 7608 GLX: Fix a use after free error with the GLVND vendor handle. 7609 7610 The GLVND layer will destroy all of the vendor handles at the end of each 7611 server generation, but the GLX module then tries to re-use the same (now-freed) 7612 handle in xorgGlxServerInit at the start of the next generation. 7613 7614 In xorgGlxCloseExtension, explicitly destroy the vendor handle and set it to 7615 NULL so that the next call to xorgGlxServerInit will recreate it. 7616 7617 Reviewed-by: Adam Jackson <ajax@redhat.com> 7618 7619commit 31c1489eeb8c5391cd978303989de167819f0041 7620Author: Peter Hutterer <peter.hutterer@who-t.net> 7621Date: Fri Apr 6 12:28:56 2018 +1000 7622 7623 xfree86: drop KDSKBMUTE handling 7624 7625 This was never merged upstream. It was a Fedora kernel patch but dropped from 7626 Fedora in 2013 with kernel 3.12. 7627 7628 The reason for the KDSKBMUTE proposal has been fixed in systemd in Feb 2013, 7629 systemd 198. 7630 https://lists.freedesktop.org/archives/systemd-devel/2013-February/008795.html 7631 7632 Reviewed-by: Adam Jackson <ajax@redhat.com> 7633 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 7634 7635commit 74aef564a7faea4410e92fa606216ab877b0d452 7636Author: Adam Jackson <ajax@redhat.com> 7637Date: Thu Apr 5 13:24:13 2018 -0400 7638 7639 xwayland: Silence a build warning if we can 7640 7641 [735/786] Generating 'hw/xwayland/Xwayland@exe/relative-pointer-unstable-v1-protocol.c'. 7642 Using "code" is deprecated - use private-code or public-code. 7643 See the help page for details. 7644 7645 Use private-code if wayland-scanner is new enough. 7646 7647 Signed-off-by: Adam Jackson <ajax@redhat.com> 7648 7649commit 6f0903ddc905f44272b85942323a467d82fef644 7650Author: Adam Jackson <ajax@redhat.com> 7651Date: Thu Apr 5 13:09:38 2018 -0400 7652 7653 dix: Hush an almost certainly bogus warning 7654 7655 ../dix/getevents.c: In function ‘transformAbsolute’: 7656 ../dix/getevents.c:1195:28: warning: ‘oy’ may be used uninitialized in this function [-Wmaybe-uninitialized] 7657 struct pixman_f_vector p = {.v = {*x, *y, 1} }; 7658 ^ 7659 ../dix/getevents.c:1234:22: note: ‘oy’ was declared here 7660 double x, y, ox, oy; 7661 ^~ 7662 7663 This one is truly special. Even though both ox and oy are set and read 7664 along the same paths, only oy is marked for this warning! Initializing 7665 just oy = 0.0 fixes it entirely, but let's not make a weird thing 7666 weirder. 7667 7668 Signed-off-by: Adam Jackson <ajax@redhat.com> 7669 Acked-by: Keith Packard <keithp@keithp.com> 7670 7671commit 57e872301f5e836be2efb8f952f9c9711650b447 7672Author: Adam Jackson <ajax@redhat.com> 7673Date: Thu Apr 5 13:07:09 2018 -0400 7674 7675 mi: Hush an almost certainly bogus warning 7676 7677 In file included from ../mi/miexpose.c:83: 7678 ../mi/miexpose.c: In function ‘miHandleExposures’: 7679 ../include/regionstr.h:174:22: warning: ‘expBox.y2’ may be used uninitialized in this function [-Wmaybe-uninitialized] 7680 (_pReg)->extents = *(_pBox); 7681 ~~~~~~~~~~~~~~~~~^~~~~~~~~~ 7682 ../mi/miexpose.c:139:12: note: ‘expBox.y2’ was declared here 7683 BoxRec expBox; 7684 ^~~~~~ 7685 7686 etc. It's initialized if (extents), and then only read if (extents), 7687 but gcc doesn't seem to figure that out. Whatever, bzero it to be 7688 explicit. 7689 7690 Signed-off-by: Adam Jackson <ajax@redhat.com> 7691 Acked-by: Keith Packard <keithp@keithp.com> 7692 7693commit 83913de25d35709b3ab7b0ab124b73924145d2dd 7694Author: Adam Jackson <ajax@redhat.com> 7695Date: Thu Apr 5 12:59:11 2018 -0400 7696 7697 xkb: Silence some compiler warnings 7698 7699 Of the form: 7700 7701 ../xkb/XKBGAlloc.c: In function ‘SrvXkbAddGeomKeyAlias’: 7702 ../xkb/XKBGAlloc.c:591:13: warning: ‘strncpy’ specified bound 4 equals destination size [-Wstringop-truncation] 7703 strncpy(alias->real, realStr, XkbKeyNameLength); 7704 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 7705 7706 This is intentional; the code that reads from these fields never reads 7707 more than 4 bytes anyway. Rephrase things in terms of memcpy so that's 7708 clear. Obviously this is awful but in XKB awful is par. 7709 7710 Signed-off-by: Adam Jackson <ajax@redhat.com> 7711 Acked-by: Keith Packard <keithp@keithp.com> 7712 7713commit d13cd3862e9ccd35c91a06680d02f2fc8fd03420 7714Author: Adam Jackson <ajax@redhat.com> 7715Date: Thu Apr 5 12:48:26 2018 -0400 7716 7717 dmx: Silence a string truncation warning. 7718 7719 ../hw/dmx/config/dmxparse.c: In function ‘dmxConfigCreateOption’: 7720 ../hw/dmx/config/dmxparse.c:385:13: warning: ‘strncpy’ output truncated before terminating nul copying as many bytes from a string as its length [-Wstringop-truncation] 7721 strncpy(option->string + offset, p->string, len); 7722 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 7723 ../hw/dmx/config/dmxparse.c:383:23: note: length computed here 7724 int len = strlen(p->string); 7725 ^~~~~~~~~~~~~~~~~ 7726 7727 The thing it's warning about is intentional, the surrounding code does 7728 its own nul-termination. Make that obvious by using memcpy instead. 7729 7730 Signed-off-by: Adam Jackson <ajax@redhat.com> 7731 Acked-by: Keith Packard <keithp@keithp.com> 7732 7733commit 176f26e96ab9958c84c98c88f31729d0240c420e 7734Author: Adam Jackson <ajax@redhat.com> 7735Date: Thu Apr 5 12:45:51 2018 -0400 7736 7737 dmx: Clean up some argument parsing code 7738 7739 This threw: 7740 7741 ../hw/dmx/input/dmxarg.c: In function ‘dmxArgParse’: 7742 ../hw/dmx/input/dmxarg.c:128:5: warning: ‘strncpy’ specified bound depends on the length of the source argument [-Wstringop-overflow=] 7743 strncpy(tmp, string, len); 7744 ^~~~~~~~~~~~~~~~~~~~~~~~~ 7745 ../hw/dmx/input/dmxarg.c:126:11: note: length computed here 7746 len = strlen(string) + 2; 7747 ^~~~~~~~~~~~~~ 7748 7749 This code predates xstrtokenize, but that's no excuse. 7750 7751 Signed-off-by: Adam Jackson <ajax@redhat.com> 7752 Acked-by: Keith Packard <keithp@keithp.com> 7753 7754commit be99072a1a20af44d2457b8c86bd9041f61efa79 7755Author: Adam Jackson <ajax@redhat.com> 7756Date: Thu Apr 5 12:35:59 2018 -0400 7757 7758 dmx: Fix a read-from-uninitialized warning 7759 7760 ../hw/dmx/dmxpixmap.c: In function ‘dmxBitmapToRegion’: 7761 ../include/regionstr.h:174:22: warning: ‘Box.x1’ may be used uninitialized in this function [-Wmaybe-uninitialized] 7762 (_pReg)->extents = *(_pBox); 7763 ~~~~~~~~~~~~~~~~~^~~~~~~~~~ 7764 ../hw/dmx/dmxpixmap.c:208:12: note: ‘Box.x1’ was declared here 7765 BoxRec Box; 7766 ^~~ 7767 7768 Signed-off-by: Adam Jackson <ajax@redhat.com> 7769 Acked-by: Keith Packard <keithp@keithp.com> 7770 7771commit c3b190f9da3a8cd6f98c127220683dd20aed0f9b 7772Author: Adam Jackson <ajax@redhat.com> 7773Date: Thu Apr 5 12:31:04 2018 -0400 7774 7775 dmx: Fix some snprintf warnings. 7776 7777 snprintf doesn't terminate the string if it truncates, so things like 7778 this are lurking crashers: 7779 7780 ../hw/dmx/dmxprop.c: In function ‘dmxPropertyIdentifier.part.0’: 7781 ../hw/dmx/dmxprop.c:94:36: warning: ‘%s’ directive output may be truncated writing up to 255 bytes into a region of size 123 [-Wformat-truncation=] 7782 snprintf(buf, sizeof(buf), "%s:%s:%s", DMX_IDENT, hostname, display); 7783 ^~ ~~~~~~~~ 7784 ../hw/dmx/dmxprop.c:94:5: note: ‘snprintf’ output 7 or more bytes (assuming 262) into a destination of size 128 7785 snprintf(buf, sizeof(buf), "%s:%s:%s", DMX_IDENT, hostname, display); 7786 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 7787 ../hw/dmx/dmxprop.c: In function ‘dmxPropertyWindow’: 7788 ../hw/dmx/dmxprop.c:372:36: warning: ‘%d’ directive output may be truncated writing between 1 and 11 bytes into a region of size between 0 and 127 [-Wformat-truncation=] 7789 snprintf(buf, sizeof(buf), "%s,%d", id, dmxScreen->index); 7790 ^~ 7791 ../hw/dmx/dmxprop.c:372:5: note: ‘snprintf’ output between 3 and 140 bytes into a destination of size 128 7792 snprintf(buf, sizeof(buf), "%s,%d", id, dmxScreen->index); 7793 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 7794 7795 We could be more precise about termination, but meh. 7796 7797 Signed-off-by: Adam Jackson <ajax@redhat.com> 7798 Acked-by: Keith Packard <keithp@keithp.com> 7799 7800commit 4c1453393feaebd688571ed1ba16c21703119ced 7801Author: Adam Jackson <ajax@redhat.com> 7802Date: Thu Apr 5 12:21:53 2018 -0400 7803 7804 gtf: Warning fix 7805 7806 ../hw/xfree86/utils/gtf/gtf.c: In function ‘print_fb_mode’: 7807 ../hw/xfree86/utils/gtf/gtf.c:241:50: warning: cast from function call of type ‘double’ to non-matching type ‘int’ [-Wbad-function-cast] 7808 printf(" timings %d %d %d %d %d %d %d\n", (int) rint(1000000.0 / m->pclk), /* pixclock in picoseconds */ 7809 7810 That's pretty nitpicky of you, gcc, but at least it's easy to fix. 7811 7812 Signed-off-by: Adam Jackson <ajax@redhat.com> 7813 Acked-by: Keith Packard <keithp@keithp.com> 7814 7815commit 99f9b077c62e14ba955b9c1f7afda47f7799d317 7816Author: Daniel Stone <daniels@collabora.com> 7817Date: Thu Apr 5 16:47:39 2018 +0100 7818 7819 modesetting: Actually get framebuffer ID 7820 7821 We would fail to get the FB ID if it wasn't already imported, since we 7822 were checking to see if the pointer was NULL (it never was) rather than 7823 if the content of the pointer was 0. 7824 7825 Signed-off-by: Daniel Stone <daniels@collabora.com> 7826 Reported-by: Olivier Fourdan <ofourdan@redhat.com> 7827 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 7828 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 7829 7830commit 8ff1cdb2bff72a37e75004f562737dd0af6daf14 7831Author: Daniel Stone <daniels@collabora.com> 7832Date: Thu Apr 5 15:00:38 2018 +0100 7833 7834 dri3: Set stride and size for old clients 7835 7836 For old clients using the fd_from_pixmap entrypoint, make sure we set 7837 stride and size correctly. 7838 7839 Noticed by inspection. 7840 7841 Signed-off-by: Daniel Stone <daniels@collabora.com> 7842 7843commit 78574a66b5b286e26839877640592980de089d64 7844Author: Daniel Stone <daniels@collabora.com> 7845Date: Thu Apr 5 14:58:40 2018 +0100 7846 7847 modesetting: Don't reuse iterator in nested loop 7848 7849 drmmode_crtc_set_mode has a loop nested inside another loop, where both 7850 of them were using 'i' as the loop iterator. Rename it to avoid an 7851 infinite loop. 7852 7853 Signed-off-by: Daniel Stone <daniels@collabora.com> 7854 Reported-by: Michel Dänzer <michel.daenzer@amd.com> 7855 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 7856 Reviewed-by: Adam Jackson <ajax@redhat.com> 7857 7858commit ce2a4313dd31084f7766af59b8477cabe029bf44 7859Author: Mario Kleiner <mario.kleiner.de@gmail.com> 7860Date: Wed Apr 4 03:49:02 2018 +0200 7861 7862 modesetting: Fix page flipping harder under DRI 3.2. 7863 7864 Non-atomic kms drivers like radeon-kms (or nouveau-kms with 7865 default setting of "atomic ioctl disabled") don't export 7866 any formats, so num_formats == 0. 7867 7868 Some atomic drivers (nouveau-kms with boot param nouveau.atomic=1, 7869 or intel-kms on, e.g., Linux 4.13) expose num_formats == 0, or 7870 don't expose any modifiers, so num_modifiers == 0. 7871 7872 Let the drmmode_is_format_supported() check pass in these cases 7873 to allow page flipping, as it works just fine. 7874 7875 Tested on NV-96 for nouveau, HD-5770 for radeon, Intel Ivybridge 7876 with Linux 4.13 and drm-next to fix page flipping. 7877 7878 Fixes: 9d147305b404 ("modesetting: Check if buffer format is supported when flipping") 7879 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 7880 Reviewed-by: Daniel Stone <daniels@collabora.com> 7881 7882commit 44e7098367b87c79470d6760753e42014be7ca01 7883Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7884Date: Wed Apr 4 00:01:15 2018 -0400 7885 7886 modesetting: Have consistent state when using atomic modesetting 7887 7888 We need to make sure that the atomic commit are consistent 7889 or else the kernel will reject it. For example, when moving 7890 a CRTC from one output to another one, the first output CRTC_ID 7891 property needs to be reset. Also if the second output was using 7892 another CRTC beforehands, it needs to be disabled to avoid an 7893 inconsistent state. 7894 7895 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7896 Tested-by: Daniel Stone <daniels@collabora.com> 7897 Reviewed-by: Adam Jackson <ajax@redhat.com> 7898 7899commit bc4d278132956ec3c43695f1bd34083ef5fe7f22 7900Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7901Date: Wed Apr 4 00:01:14 2018 -0400 7902 7903 modesetting: Use atomic modesetting to set DPMS mode 7904 7905 CRTCs and outputs needs to be enabled/disabled when the current 7906 DPMS mode is changed. We also try to do it in an atomic commit 7907 when possible. 7908 7909 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 7910 Tested-by: Daniel Stone <daniels@collabora.com> 7911 Reviewed-by: Adam Jackson <ajax@redhat.com> 7912 7913commit 23c67987a337beb91292f8e318d566941453baa3 7914Author: Daniel Stone <daniels@collabora.com> 7915Date: Wed Apr 4 16:16:38 2018 +0100 7916 7917 glamor: Add fd_from_pixmap hook 7918 7919 Add a fd_from_pixmap (singular) hook to go with fds_from_pixmap, which 7920 will ensure that the pixmap is allocated without modifiers and is thus 7921 exportable to non-modifier-aware clients. 7922 7923 This makes it possible to run a compositing manager on an old GLX/EGL 7924 stack on top of an X server which allocates internal buffer storage 7925 using exotic modifiers from modifier-aware GBM/EGL/KMS. 7926 7927 Signed-off-by: Daniel Stone <daniels@collabora.com> 7928 Reported-by: Adam Jackson <ajax@redhat.com> 7929 Reviewed-by: Adam Jackson <ajax@redhat.com> 7930 7931commit 9c407f0a1b40128fc65b19b6a499f1d4dae6f702 7932Author: Daniel Stone <daniels@collabora.com> 7933Date: Wed Apr 4 16:16:37 2018 +0100 7934 7935 glamor: Fall back to non-modifier allocations 7936 7937 If we try to allocate with particular modifiers but it fails, try to 7938 fall back to non-modifier allocations. 7939 7940 Signed-off-by: Daniel Stone <daniels@collabora.com> 7941 Reported-by: Adam Jackson <ajax@redhat.com> 7942 Reviewed-by: Adam Jackson <ajax@redhat.com> 7943 7944commit 86b2d8740a330deafe8a9bbf0402705a43efbb42 7945Author: Daniel Stone <daniels@collabora.com> 7946Date: Wed Apr 4 16:16:36 2018 +0100 7947 7948 glamor: Reallocate pixmap storage without modifiers if necessary 7949 7950 If we need a pixmap's storage to be exported to a context in which we 7951 aren't aware of modifiers, reallocate the buffer again without 7952 modifiers. 7953 7954 This makes it possible to run a compositing manager on an old GLX/EGL 7955 stack on top of an X server which allocates internal buffer storage 7956 using exotic modifiers from modifier-aware GBM/EGL/KMS. 7957 7958 Signed-off-by: Daniel Stone <daniels@collabora.com> 7959 Reported-by: Adam Jackson <ajax@redhat.com> 7960 Reviewed-by: Adam Jackson <ajax@redhat.com> 7961 7962commit aab5c46ccbe769830cae383330fd62c074a0d2f7 7963Author: Daniel Stone <daniels@collabora.com> 7964Date: Wed Apr 4 16:16:35 2018 +0100 7965 7966 glamor: Push make_exportable into callers 7967 7968 Rather than calling make_exportable from the get_bo entrypoint, make 7969 sure that someone has already explicitly requested the pixmap be 7970 exportable. 7971 7972 This is technically an ABI break in that it changes observable 7973 behaviour, but no driver other than modesetting has ever used get_bo. 7974 7975 Signed-off-by: Daniel Stone <daniels@collabora.com> 7976 Reported-by: Adam Jackson <ajax@redhat.com> 7977 Reviewed-by: Adam Jackson <ajax@redhat.com> 7978 7979commit 1b9fa3b64ca420eb54b5e5f28074c326e1fbe825 7980Author: Daniel Stone <daniels@collabora.com> 7981Date: Wed Apr 4 16:16:34 2018 +0100 7982 7983 glamor: Track if BO allocation used modifiers 7984 7985 Keep track of whether or not we fed modifiers into GBM when we allocated 7986 a BO. We'll use this later inside Glamor, to reallocate buffer storage 7987 if we allocate buffer storage using modifiers, and a non-modifier-aware 7988 client requests an export of that pixmap. 7989 7990 This makes it possible to run a compositing manager on an old GLX/EGL 7991 stack on top of an X server which allocates internal buffer storage 7992 using exotic modifiers from modifier-aware GBM/EGL/KMS. 7993 7994 Signed-off-by: Daniel Stone <daniels@collabora.com> 7995 Reported-by: Adam Jackson <ajax@redhat.com> 7996 Reviewed-by: Adam Jackson <ajax@redhat.com> 7997 7998commit 0e9504e10c4363e24a83f1a82e6a4b9f5fd8f846 7999Author: Daniel Stone <daniels@collabora.com> 8000Date: Wed Apr 4 16:16:33 2018 +0100 8001 8002 drmmode: Track if BO allocation used modifiers 8003 8004 Keep track of whether or not we fed modifiers into GBM when we allocated 8005 a BO. We'll use this later inside Glamor, to reallocate buffer storage 8006 if we allocate buffer storage using modifiers, and a non-modifier-aware 8007 client requests an export of that pixmap. 8008 8009 This makes it possible to run a compositing manager on an old GLX/EGL 8010 stack on top of an X server which allocates internal buffer storage 8011 using exotic modifiers from modifier-aware GBM/EGL/KMS. 8012 8013 Signed-off-by: Daniel Stone <daniels@collabora.com> 8014 Reported-by: Adam Jackson <ajax@redhat.com> 8015 Reviewed-by: Adam Jackson <ajax@redhat.com> 8016 8017commit 75bba3aedcb04132a4fe2806680cd38a8f2bd206 8018Author: Daniel Stone <daniels@collabora.com> 8019Date: Wed Apr 4 16:16:32 2018 +0100 8020 8021 dri3: Use single-FD screen call for single-FD request 8022 8023 When importing client buffers into Pixmaps, we can use the fds_to_pixmap 8024 hook for both single-FD and multi-FD client requests without any harm. 8025 8026 For the other direction of exporting Pixmap buffers to client FDs, 8027 create a new helper which calls the old pixmap_to_fd hook if available. 8028 This allows the implementation to ensure that the Pixmap storage is 8029 accessible to clients not aware of multiple planes or modifiers, e.g. by 8030 reallocating and copying. 8031 8032 This makes it possible to run a compositing manager on an old GLX/EGL 8033 stack on top of an X server which allocates internal buffer storage 8034 using exotic modifiers from modifier-aware GBM/EGL/KMS. 8035 8036 Signed-off-by: Daniel Stone <daniels@collabora.com> 8037 Reported-by: Adam Jackson <ajax@redhat.com> 8038 Reviewed-by: Adam Jackson <ajax@redhat.com> 8039 8040commit 574069c2912893a9b9fd3d701eb3ad4784e31553 8041Author: Aaron Plattner <aplattner@nvidia.com> 8042Date: Mon Apr 2 15:04:08 2018 -0700 8043 8044 xfree86: Restore newline before "X Protocol Version" string 8045 8046 The newline before the protocl version got lost in commit 8047 6cbefc3e0a33b380c147c533914437c7798d9b93. Prior to that commit, the 8048 release date printed a newline at the end: 8049 8050 X.Org X Server 1.19.6 8051 Release Date: 2017-12-20 8052 X Protocol Version 11, Revision 0 8053 Build Operating System: Linux 4.14.12-1-ARCH x86_64 8054 8055 Now, that string gets run together with the version: 8056 8057 X.Org X Server 1.19.99.903 (1.20.0 RC 3)X Protocol Version 11, Revision 0 8058 Build Operating System: Linux 8059 8060 Since the version string printing has a variety of #ifdefs in it, just 8061 add the newline to the begining of the protocol version string. 8062 8063 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 8064 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 8065 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 8066 8067commit df6cbf7a2b3b45be6910afb34459d30ca7783c16 8068Author: Adam Jackson <ajax@redhat.com> 8069Date: Mon Apr 2 15:44:50 2018 -0400 8070 8071 xserver 1.20 RC3 8072 8073 Signed-off-by: Adam Jackson <ajax@redhat.com> 8074 8075commit f580116f3c89b3c086655cbd441f84e50115ea78 8076Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 8077Date: Thu Mar 29 01:07:26 2018 -0400 8078 8079 modesetting: Fix reported size when using atomic modesetting 8080 8081 The framebuffer can include multiple CRTCs in multi-monitors 8082 setup. So we shouldn't use the buffer size but the CRTC size 8083 instead. Rotated displays are shadowed, so we don't need to 8084 worry about it there. 8085 8086 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 8087 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 8088 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 8089 Reviewed-by: Daniel Stone <daniels@collabora.com> 8090 8091commit ce7d5087cfe1d2df756d93b7f567b3c0b0a7f6a8 8092Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 8093Date: Thu Mar 29 01:07:25 2018 -0400 8094 8095 modesetting: Ignore alpha channel when importing BOs for modesetting 8096 8097 Fixes a regression caused by modifiers support. For some hw to 8098 continue working even if not supporting ARGB8888 and ARGB2101010 8099 formats, we assume that all imported BOs are opaque. 8100 8101 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 8102 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 8103 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 8104 Reviewed-by: Daniel Stone <daniels@collabora.com> 8105 8106commit 85b3fc18606e56d01734defb0e9fa04182f9fc1c 8107Author: Olivier Fourdan <ofourdan@redhat.com> 8108Date: Thu Mar 29 09:42:08 2018 +0200 8109 8110 modesetting: Use actual crtc position for pageflip 8111 8112 Otherwise the same content is shown on all outputs. 8113 8114 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 8115 Reviewed-by: Daniel Stone <daniels@collabora.comM> 8116 8117commit a10f1c9e08b982029186028b144d9243a2ebdde3 8118Author: Emil Velikov <emil.velikov@collabora.com> 8119Date: Mon Apr 2 18:51:21 2018 +0100 8120 8121 docs: remove resource management references 8122 8123 The code referenced was removed back in 2009. 8124 8125 Reviewed-by: Adam Jackson <ajax@redhat.com> 8126 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 8127 8128commit 32c07e6b83fdb5b0ed99fc6b1eaa18064f6c3527 8129Author: Emil Velikov <emil.velikov@collabora.com> 8130Date: Mon Apr 2 18:51:20 2018 +0100 8131 8132 docs: purge some ISA references 8133 8134 The respective ISA functions were dropped back in 2008 8135 8136 Reviewed-by: Adam Jackson <ajax@redhat.com> 8137 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 8138 8139commit 80d4098411ad96b662e830db06e59a7ab7f5c9d3 8140Author: Thierry Reding <treding@nvidia.com> 8141Date: Thu Mar 29 13:07:56 2018 +0200 8142 8143 meson: Add pixman-1 to required modules in xorg-server.pc 8144 8145 pixman headers will be included for builds of external modules against 8146 the xorg-server SDK. Make sure pixman is listed as a required module so 8147 that the correct CFLAGS will be added. 8148 8149 Note that the xorg-server.pc generated by the autotools-based build has 8150 many more modules listed, but this seems to be enough to build at least 8151 some of the external drivers against an X server built with Meson (I've 8152 tested with xf86-input-libinput, xf86-video-nouveau and xf86-video-ati). 8153 8154 Reviewed-by: Adam Jackson <ajax@redhat.com> 8155 Signed-off-by: Thierry Reding <treding@nvidia.com> 8156 8157commit 88c7b8bf4bae080ecb879fe7e3f8be2bede6f0a6 8158Author: Thierry Reding <treding@nvidia.com> 8159Date: Thu Mar 29 13:07:55 2018 +0200 8160 8161 meson: Distribute more SDK headers 8162 8163 Install missing headers to the SDK directory to allow external modules 8164 to properly build against the SDK. After this commit, the list of files 8165 installed in the SDK include directory is the same as the list of files 8166 installed by the autotools-based build. 8167 8168 Reviewed-by: Adam Jackson <ajax@redhat.com> 8169 Signed-off-by: Thierry Reding <treding@nvidia.com> 8170 8171commit f3b0a2aee21f12f3332bf598d3ba0fcbd2832f95 8172Author: Thierry Reding <treding@nvidia.com> 8173Date: Thu Mar 29 13:07:54 2018 +0200 8174 8175 meson: Remove usage of pkg-config --variable=includedir 8176 8177 Querying a pkg-config variable using the --variable option produces the 8178 value of the given variable as stored in the pkg-config file and should 8179 not be used to add directories to the include search path. 8180 8181 The reason for this is that it breaks cross-compilation, because header 8182 files are installed relative to the host sysroot. pkg-config supports a 8183 PKG_CONFIG_SYSROOT_DIR environment variable that points to this sysroot 8184 and will prepend that to the path of directories in -I or -L options in 8185 pkg-config's Cflags, Libs or Libs.private keywords. However, because no 8186 context can be inferred from variable names, as opposed to the keywords 8187 with fixed meaning, the sysroot path will not be prepended to them. The 8188 build system is responsible for doing so if necessary since it is aware 8189 of the context in which the variable is used. 8190 8191 Adding the include directory returned by pkg-config to the include path 8192 leaks build system information into the cross-build and break with very 8193 confusing errors such as this: 8194 8195 In file included from include/misc.h:82:0, 8196 from dix/atom.c:55: 8197 /usr/include/pthread.h:682:6: warning: '__regparm__' attribute directive ignored [-Wattributes] 8198 __cleanup_fct_attribute; 8199 ^~~~~~~~~~~~~~~~~~~~~~~ 8200 8201 or this: 8202 8203 In file included from include/misc.h:139:0, 8204 from dix/atom.c:55: 8205 /usr/include/stdlib.h:133:8: error: '_Float128' is not supported on this target 8206 extern _Float128 strtof128 (const char *__restrict __nptr, 8207 ^~~~~~~~~ 8208 8209 Fix this by replacing the include directory with the appropriate xproto 8210 dependency required to add the correct include directory to the compile 8211 command for subdirectories that are missing the dependency. As detailed 8212 above, this gives pkg-config the opportunity to prepend the sysroot for 8213 all paths in -I compiler options. 8214 8215 Reviewed-by: Adam Jackson <ajax@redhat.com> 8216 Signed-off-by: Thierry Reding <treding@nvidia.com> 8217 8218commit 1c002bc43472063cf8599abb0d6d7367e30456e2 8219Author: Tobias Klausmann <tobias.johannes.klausmann@mni.thm.de> 8220Date: Fri Mar 30 22:32:13 2018 +0200 8221 8222 modesetting/drmmode: add NULL pointer check in drmmode_output_dpms 8223 8224 drmmode_output_dpms is called especially with !output->crtc found in 8225 xf86DisableUnusedFunctions so we have to guard for it, else the server 8226 segfaults: 8227 8228 0 0x00007fdc1706054b in drmmode_output_dpms (output=0x55e15243c210, mode=3) at 8229 drmmode_display.c:2243 8230 1 0x000055e1500b6873 in xf86DisableUnusedFunctions (pScrn=0x55e152133f00) at 8231 xf86Crtc.c:3021 8232 2 0x000055e1500be940 in xf86RandR12CrtcSet (pScreen=<optimized out>, 8233 randr_crtc=0x55e1524b2b90, randr_mode=0x0, x=0, y=0, rotation=<optimized out>, 8234 num_randr_outputs=0, randr_outputs=0x0) at xf86RandR12.c:1244 8235 3 0x000055e1500fa1c2 in RRCrtcSet (crtc=<optimized out>, mode=0x0, x=0, y=0, 8236 rotation=rotation@entry=1, numOutputs=numOutputs@entry=0, outputs=0x0) at 8237 rrcrtc.c:763 8238 4 0x000055e1500fba9e in ProcRRSetCrtcConfig (client=0x55e152bfae50) at 8239 rrcrtc.c:1390 8240 5 0x000055e150044008 in Dispatch () at dispatch.c:478 8241 6 0x000055e150047ff8 in dix_main (argc=13, argv=0x7ffc68561038, 8242 envp=<optimized out>) at main.c:276 8243 7 0x00007fdc1a0c6a87 in __libc_start_main () at /lib64/libc.so.6 8244 8 0x000055e150031d0a in _start () at ../sysdeps/x86_64/start.S:120 8245 8246 Fixes: ba0c75177 ("modesetting: Fix up some XXX from removing GLAMOR_HAS_DRM_*") 8247 Signed-off-by: Tobias Klausmann <tobias.johannes.klausmann@mni.thm.de> 8248 Reviewed-by: Adam Jackson <ajax@redhat.com> 8249 8250commit 258081759fa70202a2e6134eeee4fe25ca54bedb 8251Author: Tobias Klausmann <tobias.johannes.klausmann@mni.thm.de> 8252Date: Fri Mar 30 22:31:46 2018 +0200 8253 8254 glamor: sanitze handling of "Debug" Option for "dmabuf_capable" 8255 8256 With the appearance of the this new Option the X Server would crash if the 8257 option is not set. 8258 8259 Default dmabuf_capable to off/FALSE for now - a user without knowledge about 8260 this option does not want to enable it until its save. 8261 8262 Fixes: d11d5bb80 ("glamor: Hide new DRI behind Option "Debug" "dmabuf_capable") 8263 Signed-off-by: Tobias Klausmann <tobias.johannes.klausmann@mni.thm.de> 8264 Reviewed-by: Adam Jackson <ajax@redhat.com> 8265 8266commit debf75bd0b0ddfec82dad9580efa7d19479b55df 8267Author: Adam Jackson <ajax@redhat.com> 8268Date: Wed Mar 28 15:47:33 2018 -0400 8269 8270 xserver 1.20 RC2 8271 8272 Signed-off-by: Adam Jackson <ajax@redhat.com> 8273 8274commit 726839459cb887a60361cb4779f9871c03f372b1 8275Author: Adam Jackson <ajax@redhat.com> 8276Date: Wed Mar 28 16:11:18 2018 -0400 8277 8278 autotools: Derive xkb configuration from xkbcomp.pc 8279 8280 ... if available, falling back to the current heuristics otherwise. This 8281 _finally_ gets me to being able to run util/modular/release.sh without 8282 overriding $prefix. 8283 8284 Signed-off-by: Adam Jackson <ajax@redhat.com> 8285 8286commit d11d5bb80d21f2aaa54ddf2f26a50aa6ef31d2fa 8287Author: Adam Jackson <ajax@redhat.com> 8288Date: Wed Mar 28 12:46:01 2018 -0400 8289 8290 glamor: Hide new DRI behind Option "Debug" "dmabuf_capable" 8291 8292 ... for xfree86, at least for now. Things appear to work for Xwayland 8293 but not yet for modesetting. Hopefully we can fix that before 1.20 but 8294 in the meantime this makes testing both paths easier than a rebuild. 8295 8296 Signed-off-by: Adam Jackson <ajax@redhat.com> 8297 8298commit df13ea7da9174d2e1c732985d9b7a9a35a6916bc 8299Author: Adam Jackson <ajax@redhat.com> 8300Date: Wed Mar 28 12:46:00 2018 -0400 8301 8302 xfree86: Add Option "Debug" to ServerFlags 8303 8304 This provides a generic way to control obscure runtime behavior knobs 8305 without making interface promises. 8306 8307 Signed-off-by: Adam Jackson <ajax@redhat.com> 8308 8309commit c95361465ee98dd97a99778d8a31be88fe9e4c4f 8310Author: Antoine Martin <antoine@nagafix.co.uk> 8311Date: Wed Mar 28 15:16:02 2018 -0400 8312 8313 os: Fix -logfile when used with -displayfd 8314 8315 Trivial way to reproduce the bug: 8316 8317 $ Xorg -logfile /tmp/mylog -config /etc/xpra/xorg.conf -displayfd 2 8318 8319 The server then moans: 8320 8321 Failed to rename log file "/tmp/mylog" to "/tmp/mylog": No such file or directory 8322 8323 And the log file is created but immediately renamed to "/tmp/mylog.old". 8324 This is caused by the changes to the log file handling introduced by 8325 this commit: 8326 8327 https://cgit.freedesktop.org/xorg/xserver/commit/?id=edcb6426f20c3be5dd5f50b76a686754aef2f64e 8328 8329 To fix this, only rename the logfile if the log filename contains the 8330 magic substitution string "%s". 8331 8332 Signed-off-by: Antoine Martin <antoine@nagafix.co.uk> 8333 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 8334 8335commit b355e55f500c18d55d0ac241e5f7e826ed62f555 8336Author: Olivier Fourdan <ofourdan@redhat.com> 8337Date: Wed Mar 28 12:17:48 2018 +0200 8338 8339 automake: Fix 'make dist' 8340 8341 Automake would skip a few meson files that would prevent to build with 8342 meson a dist file previously generated by automake. 8343 8344 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 8345 Reviewed-by: Thierry Reding <treding@nvidia.com> 8346 8347commit 16639ab77d14cecb7d03fd2d5c796c2fb1714c18 8348Author: Kyle Brenneman <kbrenneman@nvidia.com> 8349Date: Fri Mar 2 17:58:07 2018 -0700 8350 8351 Don't delete GLX's extensionInitCallback list during a reset. 8352 8353 When a callback list is initialized using CreateCallbackList via AddCallback, 8354 the list gets added to the listsToCleanup array, and as a result the list gets 8355 deleted at the end of the server generation. 8356 8357 But, vendor libraries add themselves to that callback list only once, not once 8358 per generation, so if you delete the list, then no vendor will register itself 8359 on the next generation, and GLX breaks. 8360 8361 Instead, use a static CallbackListRec for the extensionInitCallback list. That 8362 way, it doesn't get added to listsToCleanup, and doesn't get deleted during a 8363 reset. 8364 8365 Reviewed-by: Adam Jackson <ajax@redhat.com> 8366 8367commit 7004a7c3c983b4453756b92341bf22790b1eb98d 8368Author: Daniel Stone <daniels@collabora.com> 8369Date: Fri Mar 23 13:50:23 2018 +0000 8370 8371 modesetting/drmmode: Remove unused flink call 8372 8373 We don't use flink in the GetFB import path anymore, as we do an 8374 FD-based import instead. 8375 8376 Signed-off-by: Daniel Stone <daniels@collabora.com> 8377 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 8378 8379commit be087778a0eae3093ffdbba3ff7c9f3863d8e1d4 8380Author: Roman Gilg <subdiff@gmail.com> 8381Date: Tue Mar 13 16:00:57 2018 +0100 8382 8383 xwayland: Activate Present flips in rootless mode with Glamor 8384 8385 Link the newly introduced support for Present flips. For now flips can only 8386 be used in rootless mode together with Glamor. 8387 8388 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8389 Reviewed-by: Adam Jackson <ajax@redhat.com> 8390 8391commit 07750ff3c084c6549a5612d1f935a9a3ab3df67c 8392Author: Roman Gilg <subdiff@gmail.com> 8393Date: Tue Mar 13 16:00:55 2018 +0100 8394 8395 xwayland: Implement queuing present vblanks 8396 8397 Queue present events to msc values. Fake msc events with a refresh rate of 8398 about 60fps when flips are not possible. When flips are executed rely on 8399 frame callbacks with a slow updating timer as fallback. 8400 8401 This is important for applications, that want to limit their framerate. 8402 8403 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8404 Reviewed-by: Adam Jackson <ajax@redhat.com> 8405 8406commit 86df366973de1c10da5fbdc57d1ff12b681c321f 8407Author: Roman Gilg <subdiff@gmail.com> 8408Date: Tue Mar 13 16:00:54 2018 +0100 8409 8410 xwayland: Add fallback timer for msc counting 8411 8412 When the compositor is not sending frame callbacks while we still wait 8413 on buffer release events fake a continuous msc counter with a timer. 8414 8415 Having this timer is a prerequisite for queuing events. 8416 8417 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8418 Reviewed-by: Adam Jackson <ajax@redhat.com> 8419 8420commit 0fb2cca193e60b731c8e75a2a7e795477fb5fd8f 8421Author: Roman Gilg <subdiff@gmail.com> 8422Date: Tue Mar 13 16:00:53 2018 +0100 8423 8424 xwayland: Preliminary support for Present's new window flip mode 8425 8426 Introduce support for Present's window flip mode. The support is not yet 8427 complete, but works reasonable well for the most important use case, that 8428 is fullscreen applications. 8429 8430 We take a Present flip and if the xwl_window->window has the same dimensions 8431 as the presenting window, the flip is represented by a wl_buffer and attached 8432 to the main wl_surface of the xwl_window. 8433 8434 After commit we are listening for the sync callback in order to tell Present, 8435 that the pixmap flip is not longer pending, for the frame callback in order 8436 to update the msc counter and for the buffer release callback in order to tell 8437 Present that the pixmap is idle again. 8438 8439 The following functionality is missing from this patch: 8440 * (slowed down) flips in case the compositor is not sending frame callbacks, 8441 * queuing events to MSC times, 8442 * per window flips for child windows with smaller size than the xwl_window. 8443 8444 To make use of this functionality Xwayland must run rootless and with 8445 Glamor/GBM. 8446 8447 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8448 Reviewed-by: Adam Jackson <ajax@redhat.com> 8449 8450commit 8fba2a03f1410f3bc7504e218ac1e5c964279ea2 8451Author: Roman Gilg <subdiff@gmail.com> 8452Date: Tue Mar 13 16:00:52 2018 +0100 8453 8454 xwayland: Add arguments to glamor_pixmap_get_wl_buffer 8455 8456 Add arguments to give the caller more information and control 8457 over the creation of a wl_buffer with GBM, in particular let 8458 the caller determine the size of the buffer. 8459 8460 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8461 Reviewed-by: Adam Jackson <ajax@redhat.com> 8462 8463commit 902429f077325b98e30ede2710bd7a88440d2937 8464Author: Roman Gilg <subdiff@gmail.com> 8465Date: Tue Mar 13 16:00:51 2018 +0100 8466 8467 present: Add exported init function of window flip mode 8468 8469 Allow drivers now to initialize window flip mode. 8470 8471 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8472 Reviewed-by: Adam Jackson <ajax@redhat.com> 8473 8474commit a337949f99bc473ea0ae0af64736eae3d5b39399 8475Author: Roman Gilg <subdiff@gmail.com> 8476Date: Tue Mar 13 16:00:50 2018 +0100 8477 8478 present: Add cleanups for window flip mode 8479 8480 Make sure that vblanks and windows get cleaned up correctly 8481 in window flip mode. 8482 8483 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8484 Reviewed-by: Adam Jackson <ajax@redhat.com> 8485 8486commit 66a5c0bccb222ad8b9b57b10490c3041e1b3f05e 8487Author: Roman Gilg <subdiff@gmail.com> 8488Date: Tue Mar 13 16:00:49 2018 +0100 8489 8490 present: In window flip mode report damage on flip to driver 8491 8492 Calculate damage before trying to flip and report it to the driver. 8493 This allows drivers to optimize their rendering. 8494 8495 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8496 Reviewed-by: Adam Jackson <ajax@redhat.com> 8497 8498commit 029608dd80204ac96423ef79ec46c1a18bbdd5ff 8499Author: Roman Gilg <subdiff@gmail.com> 8500Date: Tue Mar 13 16:00:48 2018 +0100 8501 8502 present: Add window flip mode 8503 8504 In contrast to screen flip mode this mode: 8505 * supports flips per windows (these windows currently need to have the same 8506 size as their parent windows with the same pixmap), 8507 * sends pixmap idle signals to the client only after the driver has given 8508 an additional event notify. 8509 8510 This patch only introduces the new mode as a stub. It additionally needs a 8511 driver hook, such that it can get initialized and appropriate cleanup 8512 functions. 8513 8514 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8515 Reviewed-by: Adam Jackson <ajax@redhat.com> 8516 8517commit 8d370fcdcaed210d9f4afc1650aa8b161c7fbb44 8518Author: Roman Gilg <subdiff@gmail.com> 8519Date: Tue Mar 13 16:00:47 2018 +0100 8520 8521 present: Add driver facing window flip mode hooks 8522 8523 To enable special functionality of window flips introduce for window flips 8524 a separate set of driver facing function hooks. 8525 8526 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8527 Reviewed-by: Adam Jackson <ajax@redhat.com> 8528 8529commit 7b071b4e440313254398f06eb59b1596a6d3e8fe 8530Author: Roman Gilg <subdiff@gmail.com> 8531Date: Tue Mar 13 16:00:46 2018 +0100 8532 8533 present: Adapt flip mode API hooks for window flip mode 8534 8535 Flipping pixmaps per window needs additional arguments in the 8536 flip mode API. Add these as preperation for window flip mode. 8537 8538 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8539 Reviewed-by: Adam Jackson <ajax@redhat.com> 8540 8541commit 92b91b8cf34a38de39281044d8441b6cabe87a85 8542Author: Roman Gilg <subdiff@gmail.com> 8543Date: Tue Mar 13 16:00:45 2018 +0100 8544 8545 present: Add flip_idler vblank property 8546 8547 Introduce vblank property for flip modes, that demand explicite 8548 allowance by the driver for vblanks to become idle. 8549 8550 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8551 Reviewed-by: Adam Jackson <ajax@redhat.com> 8552 8553commit 84e47f3fe68f05f7b0b762e96acd4c95fa8000ca 8554Author: Roman Gilg <subdiff@gmail.com> 8555Date: Tue Mar 13 16:00:44 2018 +0100 8556 8557 present: Add present_window_priv properties for window flip mode 8558 8559 For window flip mode data about flips needs to be stored per window. 8560 Add properties to 'present_window_priv' and initialize them on creation. 8561 8562 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8563 Reviewed-by: Adam Jackson <ajax@redhat.com> 8564 8565commit 3aaaac0be573fb09a206966075d81ebe0510ca23 8566Author: Roman Gilg <subdiff@gmail.com> 8567Date: Tue Mar 13 16:00:43 2018 +0100 8568 8569 present: Refactor present_screen_init 8570 8571 To initialize easily different flip modes, refactor 8572 'present_screen_init'. 8573 8574 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8575 Reviewed-by: Adam Jackson <ajax@redhat.com> 8576 8577commit 6d813bbd5ea0fc38a8114c08368a7954eeb2ef37 8578Author: Roman Gilg <subdiff@gmail.com> 8579Date: Tue Mar 13 16:00:42 2018 +0100 8580 8581 present: Add more hooks to internal flip mode API 8582 8583 Add hooks to query caps, get crtcs, abort vblanks and destroy 8584 a flip. 8585 8586 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8587 Reviewed-by: Adam Jackson <ajax@redhat.com> 8588 8589commit 6a338b5959ca5a9e5260d71b6a739a5c672d77e7 8590Author: Roman Gilg <subdiff@gmail.com> 8591Date: Tue Mar 13 16:00:41 2018 +0100 8592 8593 present: Move timings adjustment in common part of flip mode API 8594 8595 To reduce future code duplication refactor timings adjustment out 8596 as a separate function. 8597 8598 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8599 Reviewed-by: Adam Jackson <ajax@redhat.com> 8600 8601commit 84112a1d0b221c00d7d3c23fd5b97687e6e3749a 8602Author: Roman Gilg <subdiff@gmail.com> 8603Date: Tue Mar 13 16:00:40 2018 +0100 8604 8605 present: Add flip mode API hook for present_can_window_flip 8606 8607 Flip modes can now have different implementations of 8608 present_can_window_flip. 8609 8610 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8611 Reviewed-by: Adam Jackson <ajax@redhat.com> 8612 8613commit 1db7cf0429eabf33f8e2b55a15db4d1f87e1fb95 8614Author: Roman Gilg <subdiff@gmail.com> 8615Date: Tue Mar 13 16:00:39 2018 +0100 8616 8617 present: Add flip mode API hook for present_pixmap 8618 8619 Make present_pixmap a common function callable by any 8620 flip mode. 8621 8622 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8623 Reviewed-by: Adam Jackson <ajax@redhat.com> 8624 8625commit 679ffbf5f39822ea508e50f1b7c92a2a9e79f7bb 8626Author: Roman Gilg <subdiff@gmail.com> 8627Date: Tue Mar 13 16:00:38 2018 +0100 8628 8629 present: Refactor execute in separate file 8630 8631 To be shared by multiple flip modes, refactor execute functionality, 8632 such that logical chunks can go in new separate file. 8633 8634 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8635 Reviewed-by: Adam Jackson <ajax@redhat.com> 8636 8637commit 1e7d8902bfe7cfb79c41b14fc6b50bcbe4f7c800 8638Author: Roman Gilg <subdiff@gmail.com> 8639Date: Tue Mar 13 16:00:37 2018 +0100 8640 8641 present: Add flip mode API hooks for several functions 8642 8643 Add 'queue_vblank', 'flush' and 're_execute' hooks, that 8644 are supposed to be shared with other flip modes. 8645 8646 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8647 Reviewed-by: Adam Jackson <ajax@redhat.com> 8648 8649commit 5365ece70a75a05df3d6351767d19c3edcf0305d 8650Author: Roman Gilg <subdiff@gmail.com> 8651Date: Tue Mar 13 16:00:36 2018 +0100 8652 8653 present: Move vblank functionality in seperate file 8654 8655 With the new internal flip mode API move vblank creation 8656 and so on into a seperate file, such that it can be shared 8657 between flip modes. 8658 8659 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8660 Reviewed-by: Adam Jackson <ajax@redhat.com> 8661 8662commit c5c50c6db1e71e976596750277b1a618704c04aa 8663Author: Roman Gilg <subdiff@gmail.com> 8664Date: Tue Mar 13 16:00:35 2018 +0100 8665 8666 present: Preliminary internal flip mode API 8667 8668 Add some basic function hooks to our future present-internal flip mode API, 8669 that will allow us to share functionality in between modes and move more code 8670 in separate files. 8671 8672 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8673 Reviewed-by: Adam Jackson <ajax@redhat.com> 8674 8675commit dda7efec36b495e100e670e057d7ef5022ef3976 8676Author: Roman Gilg <subdiff@gmail.com> 8677Date: Tue Mar 13 16:00:34 2018 +0100 8678 8679 present: Move screen flip functionality in separate file 8680 8681 As a preperation for future flip mode alternatives move most of the 8682 functionality from 'present.c' into a separate file. 8683 8684 Leave some functions needed by future other flip modes in 'present.c'. 8685 8686 Signed-off-by: Roman Gilg <subdiff@gmail.com> 8687 Reviewed-by: Adam Jackson <ajax@redhat.com> 8688 8689commit 4303deae78c4e2a60856be87bf87cd1643c268f5 8690Author: Thierry Reding <treding@nvidia.com> 8691Date: Wed Mar 28 14:04:38 2018 +0200 8692 8693 meson: Properly extract ABI versions for xorg-server.pc 8694 8695 The newline in the middle of the awk expression confuses Meson and 8696 causes it to pass only the string before the newline to awk, which will 8697 subsequently fail because it encounters an unterminated string. 8698 8699 One fix would be to escape the newlines ('\\n'), but that causes the 8700 newline to end up in the pkg-config file and separate the ABI version 8701 lines by blank lines. 8702 8703 Instead, simply drop the newlines to make the generated pkg-config file 8704 look more like the one generated as part of the autotools-based build. 8705 8706 Signed-off-by: Thierry Reding <treding@nvidia.com> 8707 Reviewed-by: Adam Jackson <ajax@redhat.com> 8708 8709commit 465b38cadd573ce50eca959f4b56bf4a6a638d91 8710Author: Thierry Reding <treding@nvidia.com> 8711Date: Wed Mar 28 14:04:37 2018 +0200 8712 8713 meson: Fix generation of xorg-server.pc 8714 8715 Meson stores relative paths for includedir, libdir and friends. These 8716 have to be concatenated with the ${prefix} or ${exec_prefix} variables 8717 to create a working pkg-config file. 8718 8719 While at it, set a default value for the module_dir option so that it 8720 points to the same location as used in the autotools-based build. 8721 8722 Signed-off-by: Thierry Reding <treding@nvidia.com> 8723 Reviewed-by: Adam Jackson <ajax@redhat.com> 8724 8725commit d905e3330fd4f9b15a0277b22d608705787cb8bc 8726Author: Thierry Reding <treding@nvidia.com> 8727Date: Wed Mar 28 14:04:36 2018 +0200 8728 8729 meson: Fill in missing data for xorg-server.h 8730 8731 Provide more values for the definitions in xorg-server.h for Meson 8732 builds. The only missing defines left after this are _BSD_SOURCE, 8733 _POSIX_SOURCE and _XOPEN_SOURCE, but Meson seems to already define 8734 these via the command-line if necessary. 8735 8736 Signed-off-by: Thierry Reding <treding@nvidia.com> 8737 Reviewed-by: Adam Jackson <ajax@redhat.com> 8738 8739commit 0babe78ed5afec8a06a517346f39b5ebd003773c 8740Author: Thierry Reding <treding@nvidia.com> 8741Date: Wed Mar 28 14:04:35 2018 +0200 8742 8743 meson: Generate xorg-server.h 8744 8745 Some drivers (such as xf86-input-libinput) require the xorg-server.h to 8746 build. Generate it and install it so that it can be used by users. 8747 8748 Signed-off-by: Thierry Reding <treding@nvidia.com> 8749 Reviewed-by: Adam Jackson <ajax@redhat.com> 8750 8751commit b04ff31b9d471822fe96ed63eda93fc4cd8c22d0 8752Author: Thierry Reding <treding@nvidia.com> 8753Date: Wed Mar 28 14:04:34 2018 +0200 8754 8755 dix: Remove unused definitions from dix-config.h.in 8756 8757 There is no code around that will ever define values for the 8758 WORKING_UNALIGNED_INT or XORG_RELEASE symbols, so they will always end 8759 up commented out and are therefore completely useless. 8760 8761 Signed-off-by: Thierry Reding <treding@nvidia.com> 8762 Reviewed-by: Adam Jackson <ajax@redhat.com> 8763 8764commit c62a8b60b9a07092ebebf9aeb67a0c39c583aabe 8765Author: Thierry Reding <treding@nvidia.com> 8766Date: Wed Mar 28 14:04:33 2018 +0200 8767 8768 xorg: Remove unused definitions from xorg-server.h.in 8769 8770 There is no code around that will ever define values for the XLOADABLE, 8771 WORKING_UNALIGNED_INT or XORG_RELEASE symbols, so they will always end 8772 up commented out and are therefore completely useless. 8773 8774 Signed-off-by: Thierry Reding <treding@nvidia.com> 8775 Reviewed-by: Adam Jackson <ajax@redhat.com> 8776 8777commit 66402fcd8280b393d3992121c0bcff9c646c97ed 8778Author: Thierry Reding <treding@nvidia.com> 8779Date: Wed Mar 28 14:04:32 2018 +0200 8780 8781 meson: Fix build if Xdmcp is missing 8782 8783 Xdmcp is an optional dependency, so make sure the build succeeds if it 8784 is missing. 8785 8786 Signed-off-by: Thierry Reding <treding@nvidia.com> 8787 Reviewed-by: Adam Jackson <ajax@redhat.com> 8788 8789commit 0a7dabc86c1a7b8d91a5d22ad12fbf03d859f43e 8790Author: Adam Jackson <ajax@redhat.com> 8791Date: Mon Mar 26 14:57:07 2018 -0400 8792 8793 meson: Add option to set default font path (v2) 8794 8795 The autotools build gets this from some macros in fontutil, but they're 8796 just wrappers around pkgconfig. 8797 8798 v2: Use same default as autotools (Keith Packard) 8799 8800 Signed-off-by: Adam Jackson <ajax@redhat.com> 8801 Reviewed-by: Keith Packard <keithp@keithp.com> 8802 8803commit a26195d5c1eba86ada8f43e775c4190683e0f874 8804Author: Adam Jackson <ajax@redhat.com> 8805Date: Tue Mar 27 12:58:15 2018 -0400 8806 8807 Revert "suid touchup" 8808 8809 Pushed the wrong thing, sigh. 8810 8811 This reverts commit 73a0562615aa1adfb934b953e23b1e69b126db4f. 8812 8813commit 73a0562615aa1adfb934b953e23b1e69b126db4f 8814Author: Adam Jackson <ajax@redhat.com> 8815Date: Tue Mar 27 12:06:38 2018 -0400 8816 8817 suid touchup 8818 8819commit 64b3227374f7f161e610df735f376fcc5bb0fd58 8820Author: Adam Jackson <ajax@redhat.com> 8821Date: Tue Mar 27 10:47:05 2018 -0400 8822 8823 autotools: Fix excessive \-escaping that broke the build 8824 8825 Signed-off-by: Adam Jackson <ajax@redhat.com> 8826 8827commit d3e8af99dc1139fac3307bbcb0046500b3feff11 8828Author: Adam Jackson <ajax@redhat.com> 8829Date: Mon Mar 26 18:43:02 2018 -0400 8830 8831 meson: Fix install path for 10-quirks.conf 8832 8833 Signed-off-by: Adam Jackson <ajax@redhat.com> 8834 Reviewed-by: Keith Packard <keithp@keithp.com> 8835 8836commit 84507507ea03db2f6197b94c160bbee9768c4baa 8837Author: Adam Jackson <ajax@redhat.com> 8838Date: Mon Mar 26 18:43:01 2018 -0400 8839 8840 meson: Install xorg-server.m4 8841 8842 Signed-off-by: Adam Jackson <ajax@redhat.com> 8843 Reviewed-by: Keith Packard <keithp@keithp.com> 8844 8845commit 8c684f59d517db0812fbad5fc6bcd90c7c810493 8846Author: Adam Jackson <ajax@redhat.com> 8847Date: Mon Mar 26 18:43:00 2018 -0400 8848 8849 meson: Generate xorg-server.pc 8850 8851 Otherwise external drivers can't build against us. 8852 8853 Signed-off-by: Adam Jackson <ajax@redhat.com> 8854 Acked-by: Keith Packard <keithp@keithp.com> 8855 8856commit 9b67303a52f5380bc96862624778b101f6379527 8857Author: Adam Jackson <ajax@redhat.com> 8858Date: Mon Mar 26 18:42:59 2018 -0400 8859 8860 meson: Install the dmx utilities 8861 8862 And add the forgotten dmxrminput to the list. 8863 8864 Signed-off-by: Adam Jackson <ajax@redhat.com> 8865 Reviewed-by: Keith Packard <keithp@keithp.com> 8866 8867commit f9ef293cf612e50830ff316d67066ea544905dc9 8868Author: Adam Jackson <ajax@redhat.com> 8869Date: Mon Mar 26 18:42:58 2018 -0400 8870 8871 meson: Build cvt and gtf 8872 8873 Signed-off-by: Adam Jackson <ajax@redhat.com> 8874 Reviewed-by: Keith Packard <keithp@keithp.com> 8875 8876commit a1e8dc05163956de9ab73e567f86b91cb8790c0f 8877Author: Adam Jackson <ajax@redhat.com> 8878Date: Mon Mar 26 18:42:57 2018 -0400 8879 8880 meson: Install man pages 8881 8882 Signed-off-by: Adam Jackson <ajax@redhat.com> 8883 Acked-by: Keith Packard <keithp@keithp.com> 8884 8885commit 1e0df64738d6924151adbbd4b040df80178cfa6b 8886Author: Adam Jackson <ajax@redhat.com> 8887Date: Mon Mar 26 18:42:55 2018 -0400 8888 8889 meson: Build Xorg suid wrapper 8890 8891 Signed-off-by: Adam Jackson <ajax@redhat.com> 8892 Reviewed-by: Keith Packard <keithp@keithp.com> 8893 8894commit ab063cf96779c521618c022d4f669f69e3d9cc8f 8895Author: Adam Jackson <ajax@redhat.com> 8896Date: Mon Mar 26 18:42:54 2018 -0400 8897 8898 meson: Fix installing protocol.txt 8899 8900 One fix the constructed path, two actually install it. 8901 8902 Signed-off-by: Adam Jackson <ajax@redhat.com> 8903 Reviewed-by: Keith Packard <keithp@keithp.com> 8904 8905commit e3baacacda29afe524b3e4cab78ca057241c0551 8906Author: Adam Jackson <ajax@redhat.com> 8907Date: Mon Mar 26 18:42:53 2018 -0400 8908 8909 meson: Add libdrm to hw/xfree86/common's dependencies 8910 8911 Signed-off-by: Adam Jackson <ajax@redhat.com> 8912 Acked-by: Keith Packard <keithp@keithp.com> 8913 8914commit 2e497bf887aca832dc0dd30d071c5288ab5c1e15 8915Author: Adam Jackson <ajax@redhat.com> 8916Date: Mon Mar 26 18:42:52 2018 -0400 8917 8918 man: s/__/@/g 8919 8920 A cosmetic change for automake (though we have to replicate some of 8921 xorg-macros.m4 in manpages.am now), but meson's configure_file() wants 8922 @-delimited strings. 8923 8924 Signed-off-by: Adam Jackson <ajax@redhat.com> 8925 Reviewed-by: Keith Packard <keithp@keithp.com> 8926 8927commit 6cbefc3e0a33b380c147c533914437c7798d9b93 8928Author: Adam Jackson <ajax@redhat.com> 8929Date: Mon Mar 26 18:42:51 2018 -0400 8930 8931 autotools: Stop caring about XORG_DATE 8932 8933 Signed-off-by: Adam Jackson <ajax@redhat.com> 8934 Acked-by: Keith Packard <keithp@keithp.com> 8935 8936commit ba0c75177239ced7037d74cf61590377f6d075bb 8937Author: Adam Jackson <ajax@redhat.com> 8938Date: Thu Mar 22 14:47:21 2018 -0400 8939 8940 modesetting: Fix up some XXX from removing GLAMOR_HAS_DRM_* 8941 8942 Signed-off-by: Adam Jackson <ajax@redhat.com> 8943 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 8944 8945commit 757e0ee1884f2a34a287d632462e3bc689eccb3d 8946Author: Mario Kleiner <mario.kleiner.de@gmail.com> 8947Date: Mon Mar 12 17:57:50 2018 +0100 8948 8949 modesetting: Fix page flipping under DRI 3.2. 8950 8951 The check for "no modifier specified" in drmmode_is_format_supported() 8952 should check for DRM_FORMAT_MOD_INVALID, not for zero, as zero actually 8953 means DRM_FORMAT_MOD_LINEAR. 8954 8955 This allows page-flipping again when appropriate, as 8956 tested under nouveau and ati drivers. 8957 8958 Fixes: 9d147305b404 ("modesetting: Check if buffer format is supported when flipping") 8959 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 8960 Reviewed-by: Daniel Stone <daniels@collabora.com> 8961 8962commit 136b45e0c0f5b96b5ce014681d77756dc47f0768 8963Author: Adam Jackson <ajax@redhat.com> 8964Date: Fri Mar 16 18:03:18 2018 -0400 8965 8966 travis: Drop Xcode 8.2 8967 8968 Travis has only so many OSX builders, presumably due to the EULA 8969 requiring that it be run on Apple-branded hardware. They're also really 8970 slow to provision, presumably because they're virtual machines and not 8971 containers. In any case it's not something I'm willing to be punished 8972 for. 8973 8974 Signed-off-by: Adam Jackson <ajax@redhat.com> 8975 8976commit 243a34e170c4a7cc5d38b78f7409704adf703e9b 8977Author: Adam Jackson <ajax@redhat.com> 8978Date: Fri Mar 16 18:28:19 2018 -0400 8979 8980 travis: Bump rawhide image 8981 8982 The image now has enough mojo to run xts, hooray! 8983 8984 Note there's a couple of ugly bits in here. The piglit config fragment 8985 and stock tetexec.cfg reflect missing features from the test suites 8986 (using an installed copy of xts, figuring out the xserver test matrix at 8987 runtime, etc.), I'll be happy to remove them once the tests are fixed. 8988 8989 Signed-off-by: Adam Jackson <ajax@redhat.com> 8990 8991commit 610055809f4030bd0e7312c6b0c561fdfe6e0183 8992Author: Eric Engestrom <eric.engestrom@imgtec.com> 8993Date: Tue Mar 13 10:55:22 2018 +0000 8994 8995 exa: promise not to touch the data when swapping pointers 8996 8997 exa/exa.c:525:10: warning: initialization discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers] 8998 swap(pExaGC, pGC, funcs); 8999 ^ 9000 9001 Signed-off-by: Eric Engestrom <eric.engestrom@imgtec.com> 9002 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 9003 9004commit d36128a72acac4d54813c52c93efefad2dc9af41 9005Author: Alexander Volkov <a.volkov@rusbitech.ru> 9006Date: Mon Mar 12 15:12:10 2018 +0300 9007 9008 Xext/shm: Downgrade from error to debug log output about success 9009 9010 Signed-off-by: Alexander Volkov <a.volkov@rusbitech.ru> 9011 Reviewed-by: Adam Jackson <ajax@redhat.com> 9012 9013commit 319af6f471912160ab3eb6395ef50f9950063d43 9014Author: Nicolai Hähnle <nicolai.haehnle@amd.com> 9015Date: Tue Mar 13 17:46:37 2018 -0400 9016 9017 glx: honor LIBGL_DRIVERS_PATH when loading DRI drivers 9018 9019 Allow switching to another driver build without a full installation. 9020 9021 Glamor already takes LIBGL_DRIVERS_PATH into account, so this change 9022 makes sure that the same driver is used in both parts of the server. 9023 9024 Signed-off-by: Nicolai Hähnle <nicolai.haehnle@amd.com> 9025 Reviewed-by: Ben Crocker <bcrocker@redhat.com> 9026 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> 9027 Tested-by: Ben Crocker <bcrocker@redhat.com> 9028 9029commit 75a869a4e7c06072380931b714ac83b1037d3bbe 9030Author: Nicolai Hähnle <nicolai.haehnle@amd.com> 9031Date: Tue Mar 13 17:46:36 2018 -0400 9032 9033 xfree86: replace all uses of xf86PrivsElevated with PrivsElevated 9034 9035 [... but leave it defined and exported, since we're ABI-frozen - ajax] 9036 9037 Signed-off-by: Nicolai Hähnle <nicolai.haehnle@amd.com> 9038 Reviewed-by: Ben Crocker <bcrocker@redhat.com> 9039 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> 9040 Tested-by: Ben Crocker <bcrocker@redhat.com> 9041 9042 restore abi 9043 9044commit 1b6910af12cfefcc293be2f4144986fe3b56a36b 9045Author: Nicolai Hähnle <nicolai.haehnle@amd.com> 9046Date: Tue Mar 13 17:46:35 2018 -0400 9047 9048 os: use PrivsElevated instead of a manual check 9049 9050 Signed-off-by: Nicolai Hähnle <nicolai.haehnle@amd.com> 9051 Reviewed-by: Ben Crocker <bcrocker@redhat.com> 9052 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> 9053 Tested-by: Ben Crocker <bcrocker@redhat.com> 9054 9055commit 9ef602de46de2beae1d6231dc72a1a783a26122f 9056Author: Nicolai Hähnle <nicolai.haehnle@amd.com> 9057Date: Tue Mar 13 17:46:34 2018 -0400 9058 9059 os: move xf86PrivsElevated here 9060 9061 Having different types of code all trying to check for elevated privileges 9062 is a bad idea. This implementation is the most thorough one. 9063 9064 Signed-off-by: Nicolai Hähnle <nicolai.haehnle@amd.com> 9065 Reviewed-by: Ben Crocker <bcrocker@redhat.com> 9066 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> 9067 Tested-by: Ben Crocker <bcrocker@redhat.com> 9068 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 9069 9070commit 1519475a4382066707276d7463c7c6e52c382caa 9071Author: Adam Jackson <ajax@redhat.com> 9072Date: Wed Mar 14 15:29:15 2018 -0400 9073 9074 appveyor: use meson 9075 9076 i686 cygwin build time with autotools: 16:22 9077 i686 cygwin build time with meson: 6:56 9078 9079 Signed-off-by: Adam Jackson <ajax@redhat.com> 9080 9081commit 2415ef5b228f5a235377decde95cb146c6642ea9 9082Author: Adam Jackson <ajax@redhat.com> 9083Date: Tue Mar 20 14:00:33 2018 -0400 9084 9085 meson: Add the x(org)proto headers to the include path 9086 9087 Without this meson won't properly use headers installed into a 9088 non-default location. 9089 9090 Signed-off-by: Adam Jackson <ajax@redhat.com> 9091 9092commit 21e8a4a1b993a9b1abd4a7510cd3d8a152be998b 9093Author: Emil Velikov <emil.velikov@collabora.com> 9094Date: Wed Mar 7 18:45:33 2018 +0000 9095 9096 modesetting: remove fallback DRM_CAP_* defines 9097 9098 All the macros are available in the libdrm that we depend on. 9099 9100 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 9101 Reviewed-by: Adam Jackson <ajax@redhat.com> 9102 9103commit d2568c058745ea66353fefa8f7296a6f37795905 9104Author: Emil Velikov <emil.velikov@collabora.com> 9105Date: Wed Mar 7 18:45:32 2018 +0000 9106 9107 modesetting: remove always true DRM_IOCTL_CRTC_QUEUE_SEQUENCE guard 9108 9109 We already require libdrm 2.4.89 which provides the definition plus 9110 guarding kernel UABI like that is generally a bad idea. 9111 9112 See previous commit for details why :-) 9113 9114 Cc: Keith Packard <keithp@keithp.com> 9115 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 9116 Reviewed-by: Adam Jackson <ajax@redhat.com> 9117 9118commit f368428ab84c14f32f52e50b880d9353fbb87181 9119Author: Emil Velikov <emil.velikov@collabora.com> 9120Date: Wed Mar 7 18:45:31 2018 +0000 9121 9122 modesetting: remove always true defined(DRM_CAP_PRIME) guards 9123 9124 The macro was available in libdrm for ages. Furthermore having a guard 9125 like this is a very bad idea. 9126 9127 Building on an old server will result in a missing run-time functionality. 9128 Since it's UABI one can use a local fallback, old kernels will return 9129 -EINVAL and the fallback path will kick in. 9130 9131 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 9132 Reviewed-by: Adam Jackson <ajax@redhat.com> 9133 9134commit 22b489d27365d2513bf1cab6da92254e3bdbcf41 9135Author: Emil Velikov <emil.velikov@collabora.com> 9136Date: Wed Mar 7 18:45:30 2018 +0000 9137 9138 Remove always true GLAMOR_HAS_DRM_* guards 9139 9140 With earlier commit the required version was bumped to 2.4.89, thus the 9141 guards always evaluate to true. 9142 9143 Fixes: e4e3447603b ("Add RandR leases with modesetting driver support 9144 [v6]") 9145 Cc: Keith Packard <keithp@keithp.com> 9146 Cc: Daniel Stone <daniels@collabora.com> 9147 Cc: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9148 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 9149 Reviewed-by: Adam Jackson <ajax@redhat.com> 9150 9151commit e0748b10a8621066b404d713a76826907da777b9 9152Author: Emil Velikov <emil.velikov@collabora.com> 9153Date: Wed Mar 7 18:45:29 2018 +0000 9154 9155 configure: remove libdrm version check 9156 9157 We already require said version. 9158 9159 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 9160 Reviewed-by: Adam Jackson <ajax@redhat.com> 9161 9162commit 8718d46157b9317554aba3e01b9919a1c0c28026 9163Author: Adam Jackson <ajax@redhat.com> 9164Date: Wed Mar 14 16:43:09 2018 -0400 9165 9166 ci: Ignore builds for branches aimed at other CI services 9167 9168 Make appveyor ignore builds for /travis.*/ and vice versa. If you're 9169 only testing a Windows change, building for OSX and Linux too is a bit 9170 rude. 9171 9172 Signed-off-by: Adam Jackson <ajax@redhat.com> 9173 9174commit ae5c0dd199a5fbfbdf7a2d6b8c1b28c410289106 9175Author: Emil Velikov <emil.velikov@collabora.com> 9176Date: Tue Mar 13 18:38:49 2018 +0000 9177 9178 dri3: cap the version returned to the client 9179 9180 As per the protocol, the server should not return version greater than 9181 the one supported by the client. 9182 9183 Add a spec quote and tweak the numbers accordingly. 9184 9185 Fixes: 563138298868 ("dri3: Add DRI3 extension") 9186 Cc: Daniel Stone <daniels@collabora.com> 9187 Cc: Keith Packard <keithp@keithp.com> 9188 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 9189 Reviewed-by: Adam Jackson <ajax@redhat.com> 9190 9191commit 6a5d51e0823b43280e3646b7a0c919a3b76146ea 9192Author: Emil Velikov <emil.velikov@collabora.com> 9193Date: Mon Mar 19 16:04:43 2018 +0000 9194 9195 present: cap the version returned to the client 9196 9197 As per the protocol, the server should not return version greater than 9198 the one supported by the client. 9199 9200 Add a spec quote and tweak the numbers accordingly. 9201 9202 Fixes: 5c5c1b77982 ("present: Add Present extension") 9203 Cc: Thierry Reding <treding@nvidia.com> 9204 Cc: Daniel Stone <daniels@collabora.com> 9205 Cc: Keith Packard <keithp@keithp.com> 9206 Reviewed-by: Adam Jackson <ajax@redhat.com> 9207 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 9208 9209commit abb9b58d1af9a0286162e52ef9db390d0c950fc1 9210Author: Thierry Reding <treding@nvidia.com> 9211Date: Fri Mar 16 14:24:21 2018 +0100 9212 9213 present: Advertise protocol version 1.2 9214 9215 Everything is implemented to support protocol version 1.2. Make it 9216 official. 9217 9218 Reviewed-by: Daniel Stone <daniels@collabora.com> 9219 Signed-off-by: Thierry Reding <treding@nvidia.com> 9220 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 9221 9222commit 7fc89251ef5e7363dfbf6d831ed448bbcd8519b8 9223Author: Alan Coopersmith <alan.coopersmith@oracle.com> 9224Date: Sun Mar 4 13:00:05 2018 -0800 9225 9226 Revert "modesetting: Remove #ifdefs XF86_PDEV_SERVER_FD" 9227 9228 This reverts commit 8c455db0ebb6e5313ca81428bb6dd75ef12aaa15. 9229 9230 Since xf86platformBus.h is only included when XSERVER_PLATFORM_BUS is 9231 defined, and configure.ac only defines that on systems with udev, this 9232 commit breaks the build on non-udev systems like Solaris. 9233 9234 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 9235 9236commit edf08bd654dcdf31da9a592513b218bbd9f64cc6 9237Author: Adam Jackson <ajax@redhat.com> 9238Date: Wed Mar 14 12:10:27 2018 -0400 9239 9240 appveyor: Switch to xorgproto 9241 9242 Also build with -j2, the builder VM has two cores. 9243 9244 Signed-off-by: Adam Jackson <ajax@redhat.com> 9245 9246commit 8f628033bed408c4de215ca117fa2fa2d9f9d5cd 9247Author: Adam Jackson <ajax@redhat.com> 9248Date: Wed Mar 14 12:54:16 2018 -0400 9249 9250 travis: Update OSX build 9251 9252 Install and use xorgproto, and build with optimization as clang has 9253 warnings that only work above -O0. Also expand the build matrix for both 9254 Xcode 8.3 and 9.2. Both images are OSX 10.12 at the moment so this is 9255 really just a way to test with clang 8 vs 9. 9256 9257 Signed-off-by: Adam Jackson <ajax@redhat.com> 9258 9259commit 1ca092f990e67c7dccf5790a72c41e84711c28a8 9260Author: Adam Jackson <ajax@redhat.com> 9261Date: Wed Mar 14 13:13:54 2018 -0400 9262 9263 travis: Switch linux build from stretch to rawhide 9264 9265 Stretch's Mesa is too far behind, and building Mesa to build xserver is 9266 not the sort of problem we're trying to solve here. But our Travis 9267 scripts run inside a docker image, and that image can be whatever, so 9268 since I have to keep rawhide building anyway let's switch to that for 9269 now. The dockerfile can be found at: 9270 9271 https://github.com/nwnk/xserver-travis 9272 9273 The $DISTRO part of the change is foreshadowing for the day we readd a 9274 (probably sid-based) debian image; patches for other environments are 9275 welcome. 9276 9277 Signed-off-by: Adam Jackson <ajax@redhat.com> 9278 9279commit 17d30369f4445e1693a1b75dc6620bbc21ec8e61 9280Author: Michel Dänzer <michel.daenzer@amd.com> 9281Date: Wed Mar 14 16:48:23 2018 +0100 9282 9283 randr: Initialize RROuptutRec::nonDesktop 9284 9285 Flagged by valgrind: 9286 9287 ==13695== Conditional jump or move depends on uninitialised value(s) 9288 ==13695== at 0x22461C: RRNoticePropertyChange (rrproperty.c:150) 9289 ==13695== by 0x22461C: RRChangeOutputProperty (rrproperty.c:263) 9290 ==13695== by 0x222FC4: RROutputSetNonDesktop (rroutput.c:333) 9291 ==13695== by 0x22319C: RROutputCreate (rroutput.c:122) 9292 ==13695== by 0x1E1CE9: xf86RandR12CreateObjects12 (xf86RandR12.c:1734) 9293 ==13695== by 0x1E1CE9: xf86RandR12Init12 (xf86RandR12.c:2375) 9294 ==13695== by 0x1E1CE9: xf86RandR12Init (xf86RandR12.c:895) 9295 ==13695== by 0x1D469B: xf86CrtcScreenInit (xf86Crtc.c:778) 9296 ==13695== by 0xC095A54: RADEONScreenInit_KMS (radeon_kms.c:2436) 9297 ==13695== by 0x161444: AddGPUScreen (dispatch.c:3966) 9298 ==13695== by 0x1A3E46: InitOutput (xf86Init.c:763) 9299 ==13695== by 0x1654A7: dix_main (main.c:193) 9300 ==13695== by 0x7041A86: (below main) (libc-start.c:310) 9301 ==13695== Uninitialised value was created by a heap allocation 9302 ==13695== at 0x4C2CB8F: malloc (vg_replace_malloc.c:299) 9303 ==13695== by 0x223083: RROutputCreate (rroutput.c:83) 9304 ==13695== by 0x1E1CE9: xf86RandR12CreateObjects12 (xf86RandR12.c:1734) 9305 ==13695== by 0x1E1CE9: xf86RandR12Init12 (xf86RandR12.c:2375) 9306 ==13695== by 0x1E1CE9: xf86RandR12Init (xf86RandR12.c:895) 9307 ==13695== by 0x1D469B: xf86CrtcScreenInit (xf86Crtc.c:778) 9308 ==13695== by 0xC095A54: RADEONScreenInit_KMS (radeon_kms.c:2436) 9309 ==13695== by 0x161444: AddGPUScreen (dispatch.c:3966) 9310 ==13695== by 0x1A3E46: InitOutput (xf86Init.c:763) 9311 ==13695== by 0x1654A7: dix_main (main.c:193) 9312 ==13695== by 0x7041A86: (below main) (libc-start.c:310) 9313 9314 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9315 Reviewed-by: Keith Packard <keithp@keithp.com> 9316 Signed-off-by: Keith Packard <keithp@keithp.com> 9317 9318commit 52048ce29f872192478fece3887b7bc1c8aace7d 9319Author: Keith Packard <keithp@keithp.com> 9320Date: Mon Mar 12 12:14:44 2018 -0700 9321 9322 Xext/saver: Swap ScreenSaverSuspend 'suspend' field. Handle old XCB clients. 9323 9324 This field was defined as a Bool in the protocol headers and BOOL in 9325 xcb. Bool is not a valid type for protocol fields. It is defined as 9326 'int' by Xdefs.h, which we expect to be 32-bits on all machines. 9327 9328 The protocol headers and xcb have patches posted to switch to CARD32, 9329 which is at least well defined. 9330 9331 This change adds the necessary byte swapping to handle other-endian 9332 clients with this 32-bit field, and then changes the request 9333 processing to compare all 32-bits against zero so that it works with 9334 both new and old clients. 9335 9336 On MSB machines, Xlib will continue to work properly, but old XCB will 9337 not interoperate with the X server (either before or after this patch). 9338 9339 Signed-off-by: Keith Packard <keithp@keithp.com> 9340 Reviewed-by: Mihai Moldovan <ionic@ionic.de> 9341 Reviewed-by: Adam Jackson <ajax@redhat.com> 9342 9343commit beda6afb3c33adbf77637141f05631081894e93a 9344Author: Mario Kleiner <mario.kleiner.de@gmail.com> 9345Date: Fri Mar 9 04:58:20 2018 +0100 9346 9347 glamor/xwayland: Add depth 30 format mapping for DRI 3.2 as well. 9348 9349 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 9350 Reviewed-by: Daniel Stone <daniels@collabora.com> 9351 9352commit 07c369e05ac116901dacc20f4742622bdca2784f 9353Author: Jon Turney <jon.turney@dronecode.org.uk> 9354Date: Thu Mar 8 12:34:26 2018 +0000 9355 9356 meson: Require libdrm for dri1/2/3 when configured 'auto' as well as 'true' 9357 9358 If dri1/2/3 are configured for auto-detection, libdrm is required, as well 9359 as the corresponding proto. (Practically we will always have the 9360 corresponding protos now, as they are part of xorgproto). 9361 9362 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 9363 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 9364 9365commit f69cd2024ef25e7fd2bd2d7012e33ce16eeb79ff 9366Author: Adam Jackson <ajax@redhat.com> 9367Date: Thu Mar 8 11:13:16 2018 -0500 9368 9369 os: Define {ReadFdFrom,WriteFdTo}Client unconditionally 9370 9371 Otherwise this is broken on cygwin: 9372 9373 rrlease.c: In function ‘ProcRRCreateLease’: 9374 rrlease.c:305:9: error: implicit declaration of function ‘WriteFdToClient’ [-Werror=implicit-function-declaration] 9375 if (WriteFdToClient(client, fd, TRUE) < 0) { 9376 9377 Signed-off-by: Adam Jackson <ajax@redhat.com> 9378 9379commit 90e0cdd42dfda2accfadffa5c550712696902e14 9380Author: Rodrigo Vivi <rodrigo.vivi@intel.com> 9381Date: Wed Mar 7 07:46:20 2018 -0800 9382 9383 dri2: Sync i965_pci_ids.h from Mesa. 9384 9385 Copied from Mesa with no modifications. 9386 9387 Gives us Geminilake and Kaby Lake platform names updates and 9388 sync on Coffee Lake PCI IDs. 9389 9390 Cc: Timo Aaltonen <timo.aaltonen@canonical.com> 9391 Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> 9392 Reviewed-by: Adam Jackson <ajax@redhat.com> 9393 9394commit 43576b901151a1f32209f476249a4de6980b654f 9395Author: Michel Dänzer <michel.daenzer@amd.com> 9396Date: Tue Mar 6 10:53:19 2018 +0100 9397 9398 glamor: Restore glamor_fd_from_pixmap and glamor_pixmap_from_fd 9399 9400 They're part of the 1.20 RC1 ABI, and actually used by external drivers. 9401 Also, requiring drivers which don't support the new functionality in 9402 DRI3 1.2 to switch to the new interfaces seems unreasonable. 9403 9404 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 9405 Reviewed-by: Daniel Stone <daniels@collabora.com> 9406 9407commit 5d3fd3856b718cfbc87d17f1de698ced94923c78 9408Author: Samir Benmendil <me@rmz.io> 9409Date: Sat Mar 3 01:41:12 2018 +0000 9410 9411 xkb: XkbSetMap on the lastSlave needs to change the master 9412 9413 The layout needs to be set on the master too if the layout is changed on 9414 the device that was last used by the master. 9415 9416 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=85256 9417 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=91571 9418 9419 Signed-off-by: Samir Benmendil <me@rmz.io> 9420 Acked-by: Daniel Stone <daniels@collabora.com> 9421 9422commit ce8da936464cc78ec31de6e970e1353ffecf2547 9423Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9424Date: Wed Feb 28 01:19:46 2018 +0000 9425 9426 dri3: Enable DRI3 version 1.2 9427 9428 Enable DRI3 v1.2 now that all functions have been implemented and 9429 that there is at least one backend implementing the driver hooks 9430 (modesetting/glamor). 9431 9432 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9433 Reviewed-by: Daniel Stone <daniels@collabora.com> 9434 Acked-by: Keith Packard <keithp@keithp.com> 9435 Reviewed-by: Adam Jackson <ajax@redhat.com> 9436 9437commit 8d0d89715984e321315631dd6667e05813d26e03 9438Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9439Date: Wed Feb 28 01:19:45 2018 +0000 9440 9441 glamor: Use gbm_bo_create_with_modifiers for internal pixmap allocation 9442 9443 Using modifier might allow the driver to use a more optimal format 9444 (e.g. tiled/compressed). Let's try to use those if possible. 9445 9446 v2: Don't filter out multi-plane modifiers 9447 9448 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9449 Reviewed-by: Daniel Stone <daniels@collabora.com> 9450 Acked-by: Keith Packard <keithp@keithp.com> 9451 Reviewed-by: Adam Jackson <ajax@redhat.com> 9452 9453commit cef12efc15ca1444d6d8cd839116b318a4668692 9454Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9455Date: Wed Feb 28 01:19:44 2018 +0000 9456 9457 glamor: Implement GetSupportedModifiers 9458 9459 Implement function added in DRI3 v1.1. 9460 9461 A newest version of libepoxy (>= 1.4.4) is required as earlier 9462 versions use a problematic version of Khronos 9463 EXT_image_dma_buf_import_modifiers spec. 9464 9465 v4: Only send scanout-supported modifiers if flipping is possible 9466 v5: Fix memory corruption in XWayland (uninitialized pointer) 9467 9468 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9469 Reviewed-by: Daniel Stone <daniels@collabora.com> 9470 Acked-by: Keith Packard <keithp@keithp.com> 9471 Reviewed-by: Adam Jackson <ajax@redhat.com> 9472 9473commit c8c276c9569b3ca1e695682a5443f1b615c606bd 9474Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9475Date: Wed Feb 28 01:19:43 2018 +0000 9476 9477 glamor: Implement PixmapFromBuffers and BuffersFromPixmap 9478 9479 It relies on GBM >= 17.1.0 where we can import BO with multiple 9480 planes and a format modifier (GBM_BO_IMPORT_FD_MODIFIER). 9481 9482 v2: Properly free fds in Xwayland 9483 9484 [Also add glamor_egl_ext.h to Makefile.am for distcheck's sake - ajax] 9485 9486 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9487 Reviewed-by: Daniel Stone <daniels@collabora.com> 9488 Acked-by: Keith Packard <keithp@keithp.com> 9489 Reviewed-by: Adam Jackson <ajax@redhat.com> 9490 9491commit 9d147305b4048dcec7ea4eda3eeea83f843f7788 9492Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9493Date: Wed Feb 28 01:19:42 2018 +0000 9494 9495 modesetting: Check if buffer format is supported when flipping 9496 9497 Add support for 'check_flip2' so that the present core can know 9498 why it is impossible to flip in that scenario. The core can then 9499 let know the client that the buffer format/modifier is suboptimal. 9500 9501 v2: No longer need to implement 'check_flip' 9502 9503 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9504 Reviewed-by: Daniel Stone <daniels@collabora.com> 9505 Acked-by: Keith Packard <keithp@keithp.com> 9506 Reviewed-by: Adam Jackson <ajax@redhat.com> 9507 9508commit e375f29662ad7589cc6d8d179846da9b8a897122 9509Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9510Date: Wed Feb 28 01:19:41 2018 +0000 9511 9512 modesetting: Create scanout buffers using supported modifiers 9513 9514 Use most optimal buffer format (e.g. tiled/compressed) available 9515 for scanout. 9516 9517 v2: Don't use multi-plane modifier to create scanout buffer 9518 9519 v3: Add flag to retrieve modifiers set from enabled CRTCs only 9520 9521 v4: Fix uses when GBM/EGL driver doesn't support modifiers 9522 9523 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9524 Reviewed-by: Daniel Stone <daniels@collabora.com> 9525 Acked-by: Keith Packard <keithp@keithp.com> 9526 Reviewed-by: Adam Jackson <ajax@redhat.com> 9527 9528commit ca1c390ec7bf617c0147d69cf874dadce28e903b 9529Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9530Date: Wed Feb 28 01:19:40 2018 +0000 9531 9532 modesetting: Get supported formats/modifiers for scanout 9533 9534 Retrieve IN_FORMATS property from the plane. It gives the 9535 allowed formats and modifiers for BO allocation. 9536 9537 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9538 Reviewed-by: Daniel Stone <daniels@collabora.com> 9539 Acked-by: Keith Packard <keithp@keithp.com> 9540 Reviewed-by: Adam Jackson <ajax@redhat.com> 9541 9542commit 9817c14f6a2ea5db44459659131c13f403716df1 9543Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9544Date: Wed Feb 28 01:19:39 2018 +0000 9545 9546 modesetting: Use atomic modesetting to configure output/CRTCs 9547 9548 To make sure we also use the same primary plane and to avoid 9549 mixing uses of two APIs, it is better to always use the atomic 9550 modesetting API when possible. 9551 9552 v2: Don't use mode_output->connector_id 9553 9554 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9555 Reviewed-by: Daniel Stone <daniels@collabora.com> 9556 Acked-by: Keith Packard <keithp@keithp.com> 9557 Reviewed-by: Adam Jackson <ajax@redhat.com> 9558 9559commit 2f807c2324b46fbd9a557e7a4f177a392d5851b3 9560Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9561Date: Wed Feb 28 01:19:38 2018 +0000 9562 9563 modesetting: Add support for multi-plane pixmaps when page-flipping 9564 9565 This allows the uses of CCS compressed or tiled pixmaps as BOs when 9566 page-flipping. 9567 9568 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9569 Reviewed-by: Daniel Stone <daniels@collabora.com> 9570 Acked-by: Keith Packard <keithp@keithp.com> 9571 Reviewed-by: Adam Jackson <ajax@redhat.com> 9572 9573commit 4023d537341b9c36b9bb248c206e40694e4b08d2 9574Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9575Date: Wed Feb 28 01:19:37 2018 +0000 9576 9577 modesetting: Use atomic modesetting API for pageflip if available 9578 9579 In order to flip between compressed and uncompressed buffers - 9580 something drmModePageFlip explicitly bans us from doing - we need 9581 to port use the atomic modesetting API. It's only 'fake' atomic 9582 though given we still commit for each CRTC separately and 9583 CRTC and connector properties are not set with the atomic API. 9584 9585 The helper functions to retrieve DRM properties have been borrowed 9586 from Weston. 9587 9588 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9589 Reviewed-by: Daniel Stone <daniels@collabora.com> 9590 Acked-by: Keith Packard <keithp@keithp.com> 9591 Reviewed-by: Adam Jackson <ajax@redhat.com> 9592 9593commit e2ef3b44fa697f2ed453767e98da72ea56d4046f 9594Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9595Date: Wed Feb 28 01:19:36 2018 +0000 9596 9597 present: Send PresentCompleteModeSuboptimalCopy appropriately 9598 9599 Add 'check_flip2' hook for driver to let know the core 9600 about why flipping is not possible ('reason'). 9601 If it is because of unsupported buffer format/modifier, 9602 a PresentCompleteNotify event is sent to the client with 9603 the PresentCompleteModeSuboptimalCopy mode. 9604 9605 v2: Check for PresentOptionSuboptimal and check driver version 9606 before using 'check_flip2'. 9607 9608 v3: Only require one of 'check_flip' or 'check_flip2' to be 9609 implemented by the driver. 9610 Refactor reasons list to enum 9611 9612 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9613 Reviewed-by: Daniel Stone <daniels@collabora.com> 9614 Acked-by: Keith Packard <keithp@keithp.com> 9615 Reviewed-by: Adam Jackson <ajax@redhat.com> 9616 9617commit 6e7c40f62db66f5b59b2cf59f1e6a26e4fd21bda 9618Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9619Date: Wed Feb 28 01:19:35 2018 +0000 9620 9621 dri3: Add multi-planar/modifier buffer requests 9622 9623 Initial implementation for DRI3 v1.1. Only the DRI3 implementation 9624 is there, backends need to implement the proper hooks. 9625 9626 Version is still set to 1.0 so clients shouldn't use the new 9627 requests yet. 9628 9629 v2: Use depth/bpp instead of DRM formats in requests 9630 9631 v3: Remove DMA fence requests from v1.1 9632 Add screen/drawable modifier sets 9633 9634 v4: Free array returned by 'get_drawable_modifiers()' 9635 9636 v5: Fix FD leak 9637 9638 Signed-off-by: Daniel Stone <daniels@collabora.com> 9639 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 9640 Acked-by: Keith Packard <keithp@keithp.com> 9641 Reviewed-by: Adam Jackson <ajax@redhat.com> 9642 9643commit 0ff4074224ea370e021d79526c63cef91356730d 9644Author: Keith Packard <keithp@keithp.com> 9645Date: Fri Mar 2 17:05:04 2018 -0800 9646 9647 Require libdrm 2.4.89 or newer 9648 9649 Both autotools and meson build systems had complicated logic around 9650 what version of libdrm to require for various options. Remove that and 9651 just check for a new enough version to support all of the options 9652 which need libdrm. 9653 9654 Signed-off-by: Keith Packard <keithp@keithp.com> 9655 Reviewed-by: Adam Jackson <ajax@redhat.com> 9656 9657commit 43ffd572592d26bb78decfdf55e643bdfb011d3f 9658Author: Laurent Carlier <lordheavym@gmail.com> 9659Date: Thu Mar 1 12:58:16 2018 +0100 9660 9661 meson: Make SHM extension optional 9662 9663 v2: check for header 'sys/shm.h' 9664 9665 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 9666 Reviewed-by: Adam Jackson <ajax@redhat.com> 9667 9668commit de4565727a0bb199365d33e3c98d9ddd3e0de6f2 9669Author: Laurent Carlier <lordheavym@gmail.com> 9670Date: Thu Mar 1 12:57:27 2018 +0100 9671 9672 meson: get rid of useless HAVE_SYSV_IPC 9673 9674 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 9675 Reviewed-by: Adam Jackson <ajax@redhat.com> 9676 9677commit e02afb94783a25c21b07e44a7b615acdd86fa3f7 9678Author: Laurent Carlier <lordheavym@gmail.com> 9679Date: Thu Mar 1 12:56:28 2018 +0100 9680 9681 meson: Make ACPI support optional 9682 9683 v2: Define HAVE_ACPI in dix-config.h 9684 9685 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 9686 Reviewed-by: Adam Jackson <ajax@redhat.com> 9687 9688commit 238219e7312e2b5d972a8620b02379009686a005 9689Author: Laurent Carlier <lordheavym@gmail.com> 9690Date: Thu Mar 1 12:55:11 2018 +0100 9691 9692 meson: Make APM support optional 9693 9694 v2: Define HAVE_APM in dix-config.h 9695 9696 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 9697 Reviewed-by: Adam Jackson <ajax@redhat.com> 9698 9699commit 2af0a50a4bb9be9f58681d417ceb9a7029caaf3b 9700Author: Adam Jackson <ajax@redhat.com> 9701Date: Wed Feb 28 11:23:41 2018 -0500 9702 9703 randr: Fix a crash on initialization with GPU screens 9704 9705 RRSetChanged (immediately above) was immune to screens with no master, 9706 but RRTellChanged was not: 9707 9708 Thread 1 "X" received signal SIGSEGV, Segmentation fault. 9709 RRTellChanged (pScreen=<optimized out>) at ../../randr/randr.c:576 9710 576 mastersp = rrGetScrPriv(master); 9711 (gdb) bt 9712 #0 RRTellChanged (pScreen=<optimized out>) at ../../randr/randr.c:576 9713 #1 0x000055555566f1e9 in RRNoticePropertyChange (value=0x555555bfbf28, property=70, output=0x555555bfef10) at ../../randr/rrproperty.c:153 9714 #2 RRChangeOutputProperty (output=output@entry=0x555555bfef10, property=<optimized out>, type=type@entry=19, format=format@entry=32, mode=<optimized out>, mode@entry=0, len=len@entry=1, value=0x7fffffffe77c, sendevent=1, pending=0) 9715 at ../../randr/rrproperty.c:263 9716 #3 0x000055555566dba5 in RROutputSetNonDesktop (output=output@entry=0x555555bfef10, nonDesktop=nonDesktop@entry=0) at ../../randr/rroutput.c:333 9717 ... 9718 9719 Reported-by: Michel Dänzer <michel@daenzer.net> 9720 Signed-off-by: Adam Jackson <ajax@redhat.com> 9721 9722commit cf485c659d93c92b26c4390402575f061eb175f6 9723Author: Adam Jackson <ajax@redhat.com> 9724Date: Wed Feb 28 12:18:28 2018 -0500 9725 9726 xserver 1.20 RC 1 9727 9728 Signed-off-by: Adam Jackson <ajax@redhat.com> 9729 9730commit afccb266edc8b2ca014aec5b2a67bb1627f29fe9 9731Author: Adam Jackson <ajax@redhat.com> 9732Date: Wed Feb 28 12:26:26 2018 -0500 9733 9734 glx: Add vndserver.h to the SDK 9735 9736 Also unbreak 'make distcheck', sigh autotools. 9737 9738 Signed-off-by: Adam Jackson <ajax@redhat.com> 9739 9740commit 85ff1cdbd3286838e512fe7c70af149a8743b6fe 9741Author: Emil Velikov <emil.velikov@collabora.com> 9742Date: Tue Feb 27 17:05:50 2018 +0000 9743 9744 protocol.txt: add GLX req. 35 - SetClientInfo2ARB 9745 9746 Noticed while skimming for the typo'd version ;-) 9747 9748 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 9749 Reviewed-by: Adam Jackson <ajax@redhat.com> 9750 9751commit e4e3447603b5fd3a38a92c3f972396d1f81168ad 9752Author: Keith Packard <keithp@keithp.com> 9753Date: Mon Feb 12 13:51:56 2018 -0800 9754 9755 Add RandR leases with modesetting driver support [v6] 9756 9757 This adds support for RandR CRTC/Output leases through the modesetting 9758 driver, creating a lease using new kernel infrastructure and returning 9759 that to a client through an fd which will have access to only those 9760 resources. 9761 9762 v2: Restore CRTC mode when leases terminate 9763 9764 When a lease terminates for a crtc we have saved data for, go 9765 ahead and restore the saved mode. 9766 9767 v3: Report RR_Rotate_0 rotations for leased crtcs. 9768 9769 Ignore leased CRTCs when selecting screen size. 9770 9771 Stop leasing encoders, the kernel doesn't do that anymore. 9772 9773 Turn off crtc->enabled while leased so that modesetting 9774 ignores them. 9775 9776 Check lease status before calling any driver mode functions 9777 9778 When starting a lease, mark leased CRTCs as disabled and hide 9779 their cursors. Also, check to see if there are other 9780 non-leased CRTCs which are driving leased Outputs and mark 9781 them as disabled as well. Sometimes an application will lease 9782 an idle crtc instead of the one already associated with the 9783 leased output. 9784 9785 When terminating a lease, reset any CRTCs which are driving 9786 outputs that are no longer leased so that they start working 9787 again. 9788 9789 This required splitting the DIX level lease termination code 9790 into two pieces, one to remove the lease from the system 9791 (RRLeaseTerminated) and a new function that frees the lease 9792 data structure (RRLeaseFree). 9793 9794 v4: Report RR_Rotate_0 rotation for leased crtcs. 9795 9796 v5: Terminate all leases on server reset. 9797 9798 Leases hang around after the associated client exits so that 9799 the client doesn't need to occupy an X server client slot and 9800 consume a file descriptor once it has gotten the output 9801 resources necessary. 9802 9803 Any leases still hanging around when the X server resets or 9804 shuts down need to be cleaned up by calling the kernel to 9805 terminate the lease and freeing any DIX structures. 9806 9807 Note that we cannot simply use the existing 9808 drmmode_terminate_lease function on each lease as that wants 9809 to also reset the video mode, and during server shut down that 9810 9811 modesetting: Validate leases on VT enter 9812 9813 The kernel doesn't allow any master ioctls to run when another 9814 VT is active, including simple things like listing the active 9815 leases. To deal with that, we check the list of leases 9816 whenever the X server VT is activated. 9817 9818 xfree86: hide disabled cursors when resetting after lease termination 9819 9820 The lessee may well have played with cursors and left one 9821 active on our screen. Just tell the kernel to turn it off. 9822 9823 v6: Add meson build infrastructure 9824 9825 [Also bumped libdrm requirement - ajax] 9826 9827 Signed-off-by: Keith Packard <keithp@keithp.com> 9828 Reviewed-by: Adam Jackson <ajax@redhat.com> 9829 9830commit 023d4aba8d45e9e3630b944ecfb650c081799b96 9831Author: Keith Packard <keithp@keithp.com> 9832Date: Mon Feb 12 13:51:55 2018 -0800 9833 9834 xf86-video-modesetting: Create CONNECTOR_ID properties for outputs [v2] 9835 9836 This lets a DRM client map between X outputs and kernel connectors. 9837 9838 v2: 9839 Change CONNECTOR_ID to enum -- Adam Jackson <ajax@nwnk.net> 9840 9841 Signed-off-by: Keith Packard <keithp@keithp.com> 9842 Reviewed-by: Adam Jackson <ajax@nwnk.net> 9843 9844commit 39573605050f6253d14a4786bb79ce2b48537039 9845Author: Keith Packard <keithp@keithp.com> 9846Date: Mon Feb 12 13:51:54 2018 -0800 9847 9848 randr: Support "non-desktop" property 9849 9850 Tracks changes to the non-desktop property so that when non-zero, 9851 outputs will always appear to be disconnected. 9852 9853 Signed-off-by: Keith Packard <keithp@keithp.com> 9854 Reviewed-by: Adam Jackson <ajax@nwnk.net> 9855 9856commit b91c787c4cd2d20685db69426c539938c556128a 9857Author: Keith Packard <keithp@keithp.com> 9858Date: Mon Feb 12 13:51:53 2018 -0800 9859 9860 xf86-video-modesetting: Record non-desktop kernel property at PreInit time 9861 9862 Save any value of the kernel non-desktop property in the xf86Output 9863 structure to avoid non-desktop outputs in the default configuration. 9864 9865 [Also bump randrproto requirement to a version that defines 9866 RR_PROPERTY_NON_DESKTOP - ajax] 9867 9868 Signed-off-by: Keith Packard <keithp@keithp.com> 9869 Reviewed-by: Adam Jackson <ajax@nwnk.net> 9870 9871commit fabbaa8d6260da14fe5b7ca0f3f8748af16253a2 9872Author: Keith Packard <keithp@keithp.com> 9873Date: Mon Feb 12 13:51:52 2018 -0800 9874 9875 xfree86/modes: Check for non-desktop monitors during PreInit [v2] 9876 9877 At startup, we want to ignore non-desktop monitors unless we don't 9878 find any desktop monitors. Because there are no DIX RandR resources 9879 allocated, let the driver store this information in a new field in the 9880 xf86Output structure and then use that value to help decide whether to 9881 include an output as part of the default configuration. 9882 9883 v2: 9884 Suggested-by: Michel Dänzer <michel@daenzer.net> 9885 9886 Bump XF86_CRTC_VERSION from 7 to 8. This will let out-of-tree 9887 drivers know whether this field is available. 9888 9889 Signed-off-by: Keith Packard <keithp@keithp.com> 9890 Reviewed-by: Adam Jackson <ajax@nwnk.net> 9891 9892commit 80d4f4b6c20288aa3e0c9908d5d845fc0088a662 9893Author: Mario Kleiner <mario.kleiner.de@gmail.com> 9894Date: Tue Feb 27 03:05:32 2018 +0100 9895 9896 modesetting: Enable screen color depth 30 support. 9897 9898 glamor now supports depth 30, so allow use of it. 9899 9900 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 9901 Reviewed-by: Eric Anholt <eric@anholt.net> 9902 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> 9903 9904commit 7bc86c7d2346eadb3aafe1550ec55e63f2a3b3e5 9905Author: Mario Kleiner <mario.kleiner.de@gmail.com> 9906Date: Tue Feb 27 03:05:31 2018 +0100 9907 9908 modesetting: Adapt xf86HandleColorMaps() for > 24 color depth. (v2) 9909 9910 This retains old behavior for depths <= 24, but allows gamma 9911 table and colormap updates to work properly at depth 30. 9912 9913 This needs the xf86Randr12CrtcComputeGamma() fix for depth 30 9914 from a previous commit to work. Otherwise the server will work, 9915 but gamma table updates will silently fail, iow. the server 9916 would always run with a default identity gamma lut. 9917 9918 v2: Simplify as proposed by Michel. 9919 9920 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 9921 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> (v1) 9922 9923commit b5f9fcd50a999a00128c0cc3f6e7d1f66182c9d5 9924Author: Mario Kleiner <mario.kleiner.de@gmail.com> 9925Date: Tue Feb 27 03:05:30 2018 +0100 9926 9927 xfree86/modes: Adapt xf86Randr12CrtcComputeGamma() for depth 30. (v2) 9928 9929 At screen depths > 24 bit, the color palettes passed into 9930 xf86Randr12CrtcComputeGamma() can have a larger number of slots 9931 than the crtc's hardware lut. E.g., at depth 30, 1024 palette 9932 slots vs. 256 hw lut slots. This palette size > crtc gamma size 9933 case is not handled yet and leads to silent failure, so gamma 9934 table updates do not happen. 9935 9936 Add a new subsampling path for this case. 9937 9938 This makes lut updates work again, as tested with the xgamma 9939 utility (uses XF86VidMode extension) and some RandR based 9940 gamma ramp animation. 9941 9942 v2: Better resampling when subsampling the palette, as 9943 proposed by Ville. Now reaches the max index of the 9944 palette and deals with non-power-of-two sizes. Thanks. 9945 9946 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 9947 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> (v1) 9948 Cc: <ville.syrjala@linux.intel.com> 9949 9950commit 125201408ddd51ca8ea961a7338a6ad299e5816c 9951Author: Mario Kleiner <mario.kleiner.de@gmail.com> 9952Date: Tue Feb 27 03:05:29 2018 +0100 9953 9954 glamor: Bump version to 1.0.1 to signal depth 30 support. 9955 9956 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 9957 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> 9958 9959commit 7a687da24c9d3c9e40ffa426631bf355322d470e 9960Author: Mario Kleiner <mario.kleiner.de@gmail.com> 9961Date: Tue Feb 27 03:05:28 2018 +0100 9962 9963 glamor: Enable composite acceleration for rgb10 formats. (v2) 9964 9965 Specifically for xrgb2101010 format. 9966 9967 Tested on KDE Plasma-5 with XRender based composite 9968 acceleration backend. Much smoother and faster. 9969 9970 (v2) Dropped argb2101010, because of depth 32 confusion with 9971 argb8888, as pointed out by Eric. Thanks! 9972 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 9973 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> 9974 9975commit c326c717c8dbf4a3f0c64a031fbe6030e73ff355 9976Author: Mario Kleiner <mario.kleiner.de@gmail.com> 9977Date: Tue Feb 27 03:05:27 2018 +0100 9978 9979 glamor: Fix loose ends in color depth 30 support. 9980 9981 This makes it work properly with OpenGL based desktop 9982 compositing, as tested with EGL and GLX based compositing 9983 under OpenGL-2/3, and also artifact free with XRender 9984 based 2D compositing. 9985 9986 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 9987 Reviewed-by: Eric Anholt <eric@anholt.net> 9988 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> 9989 9990commit 9ab5d91c6c228e8f5b34e4b30c5104aed6c41958 9991Author: Mario Kleiner <mario.kleiner.de@gmail.com> 9992Date: Tue Feb 27 03:05:26 2018 +0100 9993 9994 glamor: Make Xv extension initialize at depth 30. 9995 9996 Support x-screens of depth 30, so init doesn't fail. 9997 9998 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 9999 Reviewed-by: Eric Anholt <eric@anholt.net> 10000 Reviewed-by: Antoine Martin <antoine@nagafix.co.uk> 10001 10002commit efe9e3e9ff96a06f50d94dc93ad2c19085fa3d09 10003Author: Mario Kleiner <mario.kleiner.de@gmail.com> 10004Date: Tue Feb 27 06:51:59 2018 +0100 10005 10006 modesetting: Fix fallback for lack of new vblank kernel API. 10007 10008 Turns out that the kernel DRM ioctl handling returns EINVAL 10009 instead of ENOTTY if one tries to call the new drmCrtcGetSequence() 10010 or drmCrtcQueueSequence() ioctl's introduced in Linux 4.15 on an 10011 older kernel where they are missing. This causes the fallback code 10012 not to fall back to the old drmWaitVblank() ioctl and thereby 10013 failure of vblank stuff. 10014 10015 E.g., on Linux 4.13, glxgears -info runs unthrottled at 10000 fps 10016 instead of 60 fps. Also breakage of OML_sync_control extension. 10017 10018 Check for errno != EINVAL before setting has_queue_sequence = TRUE. 10019 10020 Additionally in case of supported drmCrtcQueueSequence(), set 10021 has_queue_sequence = TRUE on success, or we might get at 10022 least a temporary failure in ms_queue_vblank(). 10023 10024 One slight ambiguity is that we can also get EINVAL if 10025 drm_crtc_vblank_get() fails in the kernel, so if that 10026 happened at first invocation of the new api, we'd fall 10027 back to drmWaitVblank() and then fail there, instead of 10028 failing in the new api, but the end result would be the 10029 same. 10030 10031 Fixes: 44d5f2eb8a2f ("xf86-video-modesetting: Support new vblank kernel API [v2]") 10032 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 10033 Cc: Keith Packard <keithp@keithp.com> 10034 Reviewed-by: Keith Packard <keithp@keithp.com> 10035 10036commit 8171d4c2d67b2990a278dc018ac32534c3afe606 10037Author: Adam Jackson <ajax@redhat.com> 10038Date: Fri Feb 23 16:08:10 2018 -0500 10039 10040 render: Store and use all 16bpc of precision for solid pixels (v2.1) 10041 10042 This plumbs the full width color for solid pictures through to fb, exa, 10043 and glamor. External drivers and acceleration code may wish to make a 10044 similar change for sufficiently new servers. 10045 10046 v2: Don't break ABI (Michel Dänzer) 10047 v2.1: Use the (correct) full color in fb too (Michel Dänzer) 10048 10049 Signed-off-by: Adam Jackson <ajax@redhat.com> 10050 Reviewed-by: Keith Packard <keithp@keithp.com> 10051 10052commit 92272378064a878ce7a62852b877d674bcad89cd 10053Author: Jeffrey Smith <whydoubt@gmail.com> 10054Date: Thu Feb 15 06:46:55 2018 -0600 10055 10056 modesetting: Do not close uninitialized dri2 screen 10057 10058 If a dri2 screen is not successfully initialized, attempting to close it 10059 results in a null dereference. 10060 10061 Maintain a flag indicating whether the dri2 screen was successfully 10062 initialized, and check it before attempting to close the dri2 screen. 10063 10064 https://bugzilla.redhat.com/show_bug.cgi?id=1485811 10065 https://bugzilla.redhat.com/show_bug.cgi?id=1493805 10066 https://bugzilla.redhat.com/show_bug.cgi?id=1534459 10067 https://bugzilla.redhat.com/show_bug.cgi?id=1541745 10068 https://bugs.freedesktop.org/show_bug.cgi?id=101282 10069 10070 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10071 10072commit aa29298416a6323fad1681a019dbcd54867b66c4 10073Author: Adam Jackson <ajax@redhat.com> 10074Date: Mon Feb 26 15:26:35 2018 -0500 10075 10076 glamor: unifdef XORG_VERSION_CURRENT 10077 10078 This is always true now that glamor is in-tree. 10079 10080 Signed-off-by: Adam Jackson <ajax@redhat.com> 10081 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 10082 10083commit 059879b3edb104014959426f953e1faf103d99ff 10084Author: Adam Jackson <ajax@redhat.com> 10085Date: Mon Feb 26 13:58:26 2018 -0500 10086 10087 xquartz: Fix linking with glxvnd 10088 10089 Signed-off-by: Adam Jackson <ajax@redhat.com> 10090 10091commit 22a3ffe68c9c498e0b6b74ebcf1404becda8e2b1 10092Author: Michel Dänzer <michel.daenzer@amd.com> 10093Date: Mon Feb 26 17:30:57 2018 +0100 10094 10095 glx: Don't pass NULL to glxGetClient 10096 10097 These guards were dropped by the commit below, but it turns out they're 10098 needed. Fixes crash on VT switch. 10099 10100 Fixes: d8ec33fe0542 ("glx: Use vnd layer for dispatch (v4)") 10101 Bugzilla: https://bugs.freedesktop.org/105233 10102 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 10103 Reviewed-by: Adam Jackson <ajax@redhat.com> 10104 10105commit 528c94c6501db936882cd06338e2250a0b884b36 10106Author: Adam Jackson <ajax@redhat.com> 10107Date: Tue Feb 20 11:03:46 2018 -0500 10108 10109 glx: Fix indirect no-config contexts 10110 10111 We would throw BadValue here for the GLX_SCREEN attribute. The upper 10112 dispatch layer already checks this, we can ignore it here. 10113 10114 Signed-off-by: Adam Jackson <ajax@redhat.com> 10115 10116commit a63ab81f6be7ad46b7e2d98a50aef4da5e664847 10117Author: Adam Jackson <ajax@redhat.com> 10118Date: Tue Feb 20 11:03:45 2018 -0500 10119 10120 glx: Enable GLX_ARB_create_context_no_error (v2) 10121 10122 This is mostly for the client library's convenience, if this extension 10123 is listed then it can know the attribute won't be rejected. Note that we 10124 don't honor this attribute for indirect contexts. That's fine, we don't 10125 want to introduce undefined behavior into a potentially privileged 10126 process. 10127 10128 v2: Remember to ignore the attribute (Eric Anholt) 10129 10130 Signed-off-by: Adam Jackson <ajax@redhat.com> 10131 10132commit 701f3a206d6f71c4754f457e95e2b954cf0b80c7 10133Author: Adam Jackson <ajax@redhat.com> 10134Date: Fri Feb 23 16:08:11 2018 -0500 10135 10136 render: Simplify miCompositeRects 10137 10138 Make a solid-fill picture for this instead of a 1x1 pixmap. In principle 10139 the backend can accelerate this directly, and we also get to preserve 10140 all the bits of the fill color. 10141 10142 Signed-off-by: Adam Jackson <ajax@redhat.com> 10143 Reviewed-by: Keith Packard <keithp@keithp.com> 10144 10145commit 9339fa87d03204b29d47eb8c9406f701517b3514 10146Author: Adam Jackson <ajax@redhat.com> 10147Date: Thu Feb 22 16:53:22 2018 -0500 10148 10149 render: Cosmetic cleanup to default format creation 10150 10151 Signed-off-by: Adam Jackson <ajax@redhat.com> 10152 Reviewed-by: Keith Packard <keithp@keithp.com> 10153 10154commit ddf752eafe951bb9723aabf2579c4725a6ffc9de 10155Author: Adam Jackson <ajax@redhat.com> 10156Date: Thu Feb 22 16:53:21 2018 -0500 10157 10158 render: Remove some 24bpp handling 10159 10160 This can't happen anymore. 10161 10162 Signed-off-by: Adam Jackson <ajax@redhat.com> 10163 Reviewed-by: Keith Packard <keithp@keithp.com> 10164 10165commit efd84bff238f8e12bf652525990d36baada8785b 10166Author: Peter Harris <pharris@opentext.com> 10167Date: Thu Feb 22 18:07:38 2018 -0500 10168 10169 composite: Fix use-after-free in compReparentWindow 10170 10171 If an implicitly redirected window is unredirected by the reparent 10172 operation, cw will be a stale pointer. 10173 10174 Signed-off-by: Peter Harris <pharris@opentext.com> 10175 Reviewed-by: Keith Packard <keithp@keithp.com> 10176 10177commit ac13d740bf14ba8799d4dfbee1f6f99a297dbc7e 10178Author: Emil Velikov <emil.velikov@collabora.com> 10179Date: Wed Feb 21 17:12:15 2018 +0000 10180 10181 mi: Mention extension loading in verbose logs 10182 10183 Listing the extensions is useful, despite being annoying for normal 10184 usecases. Print it only when extra (lvl 3) vebose is requested. 10185 10186 v2: Move the logging to InitExtensions(), as requested by Adam. 10187 10188 Reviewed-by: Adam Jackson <ajax@redhat.com> 10189 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 10190 10191commit 9237c5e2878c088a6cd26ea380763465a9e67012 10192Author: Emil Velikov <emil.velikov@collabora.com> 10193Date: Mon Feb 19 15:18:11 2018 +0000 10194 10195 docs: correct LoadExtensionList() documentation 10196 10197 Add the missing arguments to the function signature. 10198 10199 Fixes: e46820fb897 ("miinitext: introduce LoadExtensionList() to replace 10200 over LoadExtension()") 10201 10202 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 10203 Reviewed-by: Adam Jackson <ajax@redhat.com> 10204 10205commit 089a1b0f9fa2e80457dc3d5f05c84565c9113a21 10206Author: Emil Velikov <emil.velikov@collabora.com> 10207Date: Mon Feb 19 15:18:10 2018 +0000 10208 10209 glx: set the version we implement in GlxServerExports 10210 10211 Set the major/minor version it's currently implemented, not the one 10212 provided by the GLXSERVER_VENDOR_ABI_{MAJOR,MINOR}_VERSION macros. 10213 10214 Those are identical for now, but can change in the future. 10215 10216 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 10217 Reviewed-by: Adam Jackson <ajax@redhat.com> 10218 10219commit edc00e020310d8f53ebc16295c373fe84ec6ec87 10220Author: Emil Velikov <emil.velikov@collabora.com> 10221Date: Mon Feb 19 15:18:09 2018 +0000 10222 10223 glx: use C99 initializers for GlxServerExports 10224 10225 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 10226 Reviewed-by: Adam Jackson <ajax@redhat.com> 10227 10228commit 7fc757986947ad89d76fc0fd3d69f5fdeefc9055 10229Author: Emil Velikov <emil.velikov@collabora.com> 10230Date: Mon Feb 19 15:18:08 2018 +0000 10231 10232 glx: NULL check the correct argument in dispatch_GLXVendorPriv 10233 10234 malloc can return NULL, unlike GetVendorDispatchFunc. The latter 10235 provides DispatchBadRequest. 10236 10237 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 10238 Reviewed-by: Adam Jackson <ajax@redhat.com> 10239 10240commit 1310241e8d606687811c7394811fe337c3e4de11 10241Author: Emil Velikov <emil.velikov@collabora.com> 10242Date: Mon Feb 19 15:18:07 2018 +0000 10243 10244 glx: keep glvnd_vendor a private [static] variable 10245 10246 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 10247 Reviewed-by: Adam Jackson <ajax@redhat.com> 10248 10249commit a054532668cbbb152d0d7acfcce1e03e884bb491 10250Author: Jason Ekstrand <jason@jlekstrand.net> 10251Date: Tue Feb 20 09:41:39 2018 -0800 10252 10253 xwayland: Fix backwards need_rotate logic (v2) 10254 10255 When xdg_output support was added to Xwayland, need_rotate parameter was 10256 added to output_get_new_size where true gave you the old pre-xdg_output 10257 behavior and false gave the new behavior. Unfortunately, the two places 10258 where this is called, need_rotate was set backwards. This caused input 10259 get clampped to the wrong dimensions. Also, the logic for deciding 10260 whether or not to flip was wrong because, if need_rotate was false, it 10261 would always flip which is not what you want. 10262 10263 v2 (Daniel Stone): 10264 - Fix output_get_new_size so that it doesn't flip the dimensions when 10265 need_rotate is false. 10266 10267 Signed-off-by: Jason Ekstrand <jason@jlekstrand.net> 10268 Reviewed-by: Daniel Stone <daniels@collabora.com> 10269 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 10270 10271commit 343ee7d07565a75b90e6489c5c29f772dca79c43 10272Author: Alexander Volkov <a.volkov@rusbitech.ru> 10273Date: Tue Feb 20 18:07:18 2018 +0300 10274 10275 Xephyr: Avoid calling xcb_shm_detach() twice 10276 10277 This call was forgotten to be removed in 10278 90996f5909aab4bc9aa4011a6a6d0555a7aa3adf in which 10279 hostx_destroy_shm_segment() was introduced, which 10280 itself does it. 10281 10282 Signed-off-by: Alexander Volkov <a.volkov@rusbitech.ru> 10283 Reviewed-by: Adam Jackson <ajax@redhat.com> 10284 10285commit 5ba0ba93e127a8c460b013e3c6851dd186377c27 10286Author: Laurent Carlier <lordheavym@gmail.com> 10287Date: Sun Feb 18 10:58:35 2018 +0100 10288 10289 meson: Make DGA extension optional 10290 10291 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 10292 10293commit e790bc621b8f55489e35b1de260c649d8d105d63 10294Author: Laurent Carlier <lordheavym@gmail.com> 10295Date: Sun Feb 18 10:58:34 2018 +0100 10296 10297 dga: #if XFreeXDGA → #ifdef XFreeXDGA 10298 10299 ../include/events.h:32:14: error: #if with no expression 10300 10301 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 10302 10303commit ce4e20f80ff64a5a8b9f344e2507128f379e0715 10304Author: Laurent Carlier <lordheavym@gmail.com> 10305Date: Sun Feb 18 10:58:33 2018 +0100 10306 10307 meson: Make Xv and XvMC extensions optional 10308 10309 Just mimic autoconf file, XvMC can't be enabled without Xv 10310 10311 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 10312 10313commit 82d9e99bf99660537a11d9475369c43814cad664 10314Author: Laurent Carlier <lordheavym@gmail.com> 10315Date: Sun Feb 18 10:58:32 2018 +0100 10316 10317 meson: Xephyr Requires xcb-shm version 1.9.3 or newer 10318 10319 Sync with commit 8510f542e5474b719f7d7edba7ef0cc05af0271e 10320 10321 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 10322 10323commit 272fa1817091117d078a22252a03d3161dfc7b0f 10324Author: Jeffrey Smith <whydoubt@gmail.com> 10325Date: Thu Feb 15 23:47:46 2018 -0600 10326 10327 os: avoid uninitialized offsets in backtrace 10328 10329 When xorg_backtrace calls unw_get_proc_name and an error occurs, offset 10330 might not be set for the current frame. 10331 10332 Initialize offset for each frame so that the offset from another frame 10333 cannot be used inadvertently. 10334 10335 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10336 Reviewed-by: Adam Jackson <ajax@redhat.com> 10337 10338commit 6f9d29040cd9f4723a2e6c1e5d2ec8104efc0710 10339Author: Jon Turney <jon.turney@dronecode.org.uk> 10340Date: Fri Feb 16 16:44:37 2018 +0000 10341 10342 hw/xwin: Update for glxvnd 10343 10344 - Link with libglxvnd in meson.build 10345 - Call xorgGlxCreateVendor() like all other DDX 10346 10347 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 10348 10349commit 500cc4a029efce9b962ab8fbb10da9f87d83372e 10350Author: Adam Jackson <ajax@redhat.com> 10351Date: Wed Feb 14 17:21:43 2018 -0500 10352 10353 dmx: warning fix 10354 10355 /home/ajax/git/xserver/at/../hw/dmx/input/dmxinputinit.c: In function ‘dmxInputInit’: 10356 /home/ajax/git/xserver/at/../hw/dmx/input/dmxinputinit.c:1019:9: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement] 10357 int found; 10358 10359 Signed-off-by: Adam Jackson <ajax@redhat.com> 10360 10361commit 9bcddaf9619d01cdbf965cedb9f86a08ed6dc524 10362Author: Adam Jackson <ajax@redhat.com> 10363Date: Wed Feb 14 17:20:16 2018 -0500 10364 10365 dri2: warning fix 10366 10367 /home/ajax/git/xserver/at/../glx/glxdri2.c: In function ‘create_driver_context’: 10368 /home/ajax/git/xserver/at/../glx/glxdri2.c:444:5: warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement] 10369 const __DRIconfig *driConfig = config ? config->driConfig : NULL; 10370 10371 I don't know why autotools has decided not to be C99 but at least it's 10372 trivial to fix. 10373 10374 Signed-off-by: Adam Jackson <ajax@redhat.com> 10375 10376commit 133bd4443be6b76a36afadfae0905b832e998165 10377Author: Adam Jackson <ajax@redhat.com> 10378Date: Wed Jan 10 13:05:48 2018 -0500 10379 10380 glx: Large commands are context state, not client state 10381 10382 There's no reason a multithreaded client shouldn't be allowed to 10383 interleave other requests (for other contexts) with a RenderLarge. Move 10384 the check into __glXForceCurrent, and store the state in the context not 10385 the client. 10386 10387 Signed-off-by: Adam Jackson <ajax@redhat.com> 10388 10389commit 5d87e4f71355272c8da38c83f493b826623301cc 10390Author: Adam Jackson <ajax@redhat.com> 10391Date: Wed Jan 10 13:05:47 2018 -0500 10392 10393 miinitext: Remove separate extension toggle list 10394 10395 This was only separate because GLX was loadable. The frontend is now 10396 linked statically, so we can use the static extension list directly. 10397 10398 Signed-off-by: Adam Jackson <ajax@redhat.com> 10399 10400commit 67c303fff303f94b62f03a76de97116c6ebcfda9 10401Author: Adam Jackson <ajax@redhat.com> 10402Date: Wed Jan 10 13:05:46 2018 -0500 10403 10404 miinitext: Load GLX on the mi path 10405 10406 Add a stub for Xnest so it continues to link, but otherwise we support 10407 GLX on every server so there's no need to make every DDX add it. 10408 10409 Signed-off-by: Adam Jackson <ajax@redhat.com> 10410 10411commit d8ec33fe0542141aed1d9016d2ecaf52da944b4b 10412Author: Adam Jackson <ajax@redhat.com> 10413Date: Wed Jan 10 13:05:45 2018 -0500 10414 10415 glx: Use vnd layer for dispatch (v4) 10416 10417 The big change here is MakeCurrent and context tag tracking. We now 10418 delegate context tags entirely to the vnd layer, and simply store a 10419 pointer to the context state as the tag data. If a context is deleted 10420 while it's current, we allocate a fake ID for the context and move the 10421 context state there, so the tag data still points to a real context. As 10422 a result we can stop trying so hard to detach the client from contexts 10423 at disconnect time and just let resource destruction handle it. 10424 10425 Since vnd handles all the MakeCurrent protocol now, our request handlers 10426 for it can just be return BadImplementation. We also remove a bunch of 10427 LEGAL_NEW_RESOURCE, because now by the time we're called vnd has already 10428 allocated its tracking resource on that XID. 10429 10430 v2: Update to match v2 of the vnd import, and remove more redundant work 10431 like request length checks. 10432 10433 v3: Add/remove the XID map from the vendor private thunk, not the 10434 backend. (Kyle Brenneman) 10435 10436 v4: Fix deletion of ghost contexts (Kyle Brenneman) 10437 10438 Signed-off-by: Adam Jackson <ajax@redhat.com> 10439 10440commit 8753218beae641e5c5ac2c2ba598cfb99a893cf4 10441Author: Kyle Brenneman <kbrenneman@nvidia.com> 10442Date: Wed Jan 10 13:05:44 2018 -0500 10443 10444 glx: Import glxvnd server module (v2) 10445 10446 This is based on an out-of-tree module written by Kyle: 10447 10448 https://github.com/kbrenneman/libglvnd/tree/server-libglx 10449 10450 I (ajax) did a bunch of cosmetic fixes, ported it off xfree86 API, 10451 added request length checks, and fixed a minor bug or two. 10452 10453 v2: Use separate functions to set/get a context tag's private data, and 10454 call the backend's MakeCurrent when a client disconnects to unbind the 10455 context. (Kyle Brenneman) 10456 10457 Signed-off-by: Adam Jackson <ajax@redhat.com> 10458 10459commit d1fdddeb76328ab84d9a181a5c72b90013f0f0a6 10460Author: Laurent Carlier <lordheavym@gmail.com> 10461Date: Tue Feb 13 22:33:51 2018 +0100 10462 10463 meson: Make Security extension optional 10464 10465 Just mimic autoconf file, Security extension needs X-ACE extension 10466 10467 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 10468 Reviewed-by: Adam Jackson <ajax@redhat.com> 10469 10470commit 43ef94068e072228446484bf58c0624b9b9649ac 10471Author: Laurent Carlier <lordheavym@gmail.com> 10472Date: Tue Feb 13 22:33:50 2018 +0100 10473 10474 meson: Make more extensions optional 10475 10476 Just mimic autoconf file for xf86bigfont, screensaver, 10477 xres, xace and xinerama extensions 10478 10479 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 10480 Reviewed-by: Adam Jackson <ajax@redhat.com> 10481 10482commit 3203dd175bd71ef4d154578165d930eec29afddb 10483Author: Laurent Carlier <lordheavym@gmail.com> 10484Date: Tue Feb 13 22:33:49 2018 +0100 10485 10486 meson: Make DPMS extension optional 10487 10488 Just mimic autoconf file, DPMS is disabled with Xquartz 10489 10490 Signed-off-by: Laurent Carlier <lordheavym@gmail.com> 10491 Reviewed-by: Adam Jackson <ajax@redhat.com> 10492 10493commit 2c3e137b786dab4a8002a4934fbcf4ef0742f473 10494Author: Adam Jackson <ajax@redhat.com> 10495Date: Wed Feb 14 14:10:44 2018 -0500 10496 10497 dmx: #if PANORAMIX → #ifdef PANORAMIX 10498 10499 Signed-off-by: Adam Jackson <ajax@redhat.com> 10500 10501commit ceaf4277447407a7430b9bc3205462fbdbae0e39 10502Author: Adam Jackson <ajax@redhat.com> 10503Date: Mon Feb 12 15:35:05 2018 -0500 10504 10505 xfixes: Restore monitoring of animated cursors 10506 10507 This was broken by: 10508 10509 commit aa6651f83c698e8cc40de61aad36c73ca7a6adcf 10510 Author: Adam Jackson <ajax@redhat.com> 10511 Date: Thu Jun 8 16:23:12 2017 -0400 10512 10513 xfixes: Remove the CursorCurrent array 10514 10515 As of that change we look up the current cursor dynamically instead of 10516 trying to track every time it's set through ->DisplayCursor. That would 10517 work, except the 'bits' of an animated cursor is a transparent 1x1 10518 pixel. So now, look up whether there's an animated cursor, and use its 10519 current frame if so. 10520 10521 Signed-off-by: Adam Jackson <ajax@redhat.com> 10522 10523commit 0b7a5700ad3d65e408a2fae00bdee867a8d2e1a8 10524Author: Adam Jackson <ajax@redhat.com> 10525Date: Mon Feb 12 15:35:04 2018 -0500 10526 10527 animcur: Change which CursorPtr we save in external state 10528 10529 Formerly spriteInfo->anim.pCursor would point to the animated cursor (or 10530 NULL if not animated). That value would also be available in 10531 spriteInfo->sprite->current, so instead lets use anim.pCursor to point 10532 to the current animation element. 10533 10534 Signed-off-by: Adam Jackson <ajax@redhat.com> 10535 10536commit 8510f542e5474b719f7d7edba7ef0cc05af0271e 10537Author: Alexander Volkov <a.volkov@rusbitech.ru> 10538Date: Wed Feb 7 14:08:50 2018 +0300 10539 10540 Xephyr: Require xcb-shm version 1.9.3 or newer 10541 10542 It's needed for FD-passing. 10543 10544 Signed-off-by: Alexander Volkov <a.volkov@rusbitech.ru> 10545 Reviewed-by: Adam Jackson <ajax@redhat.com> 10546 10547commit 2ff6eaad85fd702e3cc597c44ccf2210570004ea 10548Author: Roman Gilg <subdiff@gmail.com> 10549Date: Fri Feb 9 23:40:42 2018 +0100 10550 10551 Correct xorg_list_is_empty return value description 10552 10553 The helper xorg_list_is_empty returns True when the list is empty and not when 10554 it contains one or more elements. 10555 10556 Signed-off-by: Roman Gilg <subdiff@gmail.com> 10557 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 10558 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 10559 10560commit bebcc8477c8070ade9dd4be7299c718baeab3d7a 10561Author: Mario Kleiner <mario.kleiner.de@gmail.com> 10562Date: Mon Feb 5 11:20:41 2018 +0100 10563 10564 glx: Only assign 8 bpc fbconfigs for composite visuals. 10565 10566 Commit 91c42093b248 ("glx: Duplicate relevant fbconfigs for 10567 compositing visuals") adds many new depth 32 fbconfigs as 10568 composite visuals. On a X-Screen running at depth 24, this 10569 also adds bgra 10-10-10-2 fbconigs, as they also have 10570 config.rgbBits == 32, but these are not displayable on a 10571 depth 24 screen, leading to visually corrupted desktops 10572 under some compositors, e.g., fdo bug 104597 "Compton 10573 weird colors" when running compton with 10574 "compton --backend glx". 10575 10576 Be more conservative for now and only select fbconfigs with 10577 8 bpc red, green, blue components for composite visuals. 10578 10579 Fixes: 91c42093b248 ("glx: Duplicate relevant fbconfigs for 10580 compositing visuals") 10581 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104597 10582 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 10583 Reviewed-by: Thomas Hellstrom <thellstrom@vmware.com> 10584 Reviewed-by: Adam Jackson <ajax@redhat.com> 10585 10586commit 98edb9a35e2cdd61355656b82975d7f2b6a9f0e6 10587Author: Lyude Paul <lyude@redhat.com> 10588Date: Tue Feb 6 12:41:47 2018 -0500 10589 10590 xwayland: Don't process cursor warping without an xwl_seat 10591 10592 Unfortunately, on my machine Xwayland immediately crashes when I try to 10593 start it. gdb backtrace: 10594 10595 #0 0x00007ffff74f0e79 in wl_proxy_marshal () from target:/lib64/libwayland-client.so.0 10596 #1 0x0000000000413172 in zwp_confined_pointer_v1_destroy (zwp_confined_pointer_v1=0x700000000) 10597 at hw/xwayland/Xwayland@exe/pointer-constraints-unstable-v1-client-protocol.h:612 10598 #2 0x0000000000418bc0 in xwl_seat_destroy_confined_pointer (xwl_seat=0x8ba2a0) 10599 at /home/lyudess/Projects/xserver/hw/xwayland/xwayland-input.c:2839 10600 #3 0x0000000000418c09 in xwl_seat_unconfine_pointer (xwl_seat=0x8ba2a0) 10601 at /home/lyudess/Projects/xserver/hw/xwayland/xwayland-input.c:2849 10602 #4 0x0000000000410d97 in xwl_cursor_confined_to (device=0xa5a000, screen=0x8b9d80, window=0x9bdb70) 10603 at /home/lyudess/Projects/xserver/hw/xwayland/xwayland.c:328 10604 #5 0x00000000004a8571 in ConfineCursorToWindow (pDev=0xa5a000, pWin=0x9bdb70, generateEvents=1, 10605 confineToScreen=0) at /home/lyudess/Projects/xserver/dix/events.c:900 10606 #6 0x00000000004a94b7 in ScreenRestructured (pScreen=0x8b9d80) 10607 at /home/lyudess/Projects/xserver/dix/events.c:1387 10608 #7 0x0000000000502386 in RRScreenSizeNotify (pScreen=0x8b9d80) 10609 at /home/lyudess/Projects/xserver/randr/rrscreen.c:160 10610 #8 0x000000000041a83c in update_screen_size (xwl_output=0x8e7670, width=3840, height=2160) 10611 at /home/lyudess/Projects/xserver/hw/xwayland/xwayland-output.c:203 10612 #9 0x000000000041a9f0 in apply_output_change (xwl_output=0x8e7670) 10613 at /home/lyudess/Projects/xserver/hw/xwayland/xwayland-output.c:252 10614 #10 0x000000000041aaeb in xdg_output_handle_done (data=0x8e7670, xdg_output=0x8e7580) 10615 at /home/lyudess/Projects/xserver/hw/xwayland/xwayland-output.c:307 10616 #11 0x00007ffff50e9d1e in ffi_call_unix64 () at ../src/x86/unix64.S:76 10617 #12 0x00007ffff50e968f in ffi_call (cif=<optimized out>, fn=<optimized out>, rvalue=<optimized out>, 10618 avalue=<optimized out>) at ../src/x86/ffi64.c:525 10619 #13 0x00007ffff74f3d8b in wl_closure_invoke () from target:/lib64/libwayland-client.so.0 10620 #14 0x00007ffff74f0928 in dispatch_event.isra () from target:/lib64/libwayland-client.so.0 10621 #15 0x00007ffff74f1be4 in wl_display_dispatch_queue_pending () from target:/lib64/libwayland-client.so.0 10622 #16 0x00007ffff74f200b in wl_display_roundtrip_queue () from target:/lib64/libwayland-client.so.0 10623 #17 0x0000000000418cad in InitInput (argc=12, argv=0x7fffffffd9c8) 10624 at /home/lyudess/Projects/xserver/hw/xwayland/xwayland-input.c:2867 10625 #18 0x00000000004a20e3 in dix_main (argc=12, argv=0x7fffffffd9c8, envp=0x7fffffffda30) 10626 at /home/lyudess/Projects/xserver/dix/main.c:250 10627 #19 0x0000000000420cb2 in main (argc=12, argv=0x7fffffffd9c8, envp=0x7fffffffda30) 10628 at /home/lyudess/Projects/xserver/dix/stubmain.c:34 10629 10630 This appears to be the result of xwl_cursor_confined_to() and 10631 xwl_screen_get_default_seat(). While not against protocol, mutter ends 10632 up sending xdg_output before wl_seat. xwl_screen_get_default_seat() 10633 makes the naïve assumption that we always have a valid seat, we end up 10634 returning a pointer to the empty list itself instead of an actual seat 10635 and causing ourselves to segfault. 10636 10637 So, actually return NULL in xwl_screen_get_default_seat() if the seat 10638 list is empty, and skip any pointer confinement processing in 10639 xwl_cursor_confined_to() when we don't have a seat setup yet. 10640 10641 Signed-off-by: Lyude Paul <lyude@redhat.com> 10642 Reviewed-by: Adam Jackson <ajax@redhat.com> 10643 10644commit fd21b282dc88936043a23baa4ec053a2811319a7 10645Author: Jeff Smith <whydoubt@gmail.com> 10646Date: Sun Feb 4 23:17:54 2018 -0600 10647 10648 xfree86: Only call PreInit handler if it exists for device 10649 10650 DoConfigure() attempts to call the PreInit handler on a device without 10651 checking that the handler exists. 10652 10653 Check that the PreInit handler exists for a device before attempting to 10654 call it. 10655 10656 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10657 Reviewed-by: Adam Jackson <ajax@redhat.com> 10658 10659commit e81031f3fda0f8b4237224b13c016759eaa52449 10660Author: Jeff Smith <whydoubt@gmail.com> 10661Date: Sun Feb 4 23:17:53 2018 -0600 10662 10663 xfree86: Allocate sufficienct space for dev2screen array 10664 10665 When the dev2screen is sized to xf86NumDrivers in DoConfigure(), 10666 subsequent code may attempt to write past the end of the array. 10667 10668 Size the dev2screen array to nDevToConfig instead. 10669 10670 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10671 Reviewed-by: Adam Jackson <ajax@redhat.com> 10672 10673commit 1a24a0ae7b1a7400735530a21ac8c0247723223d 10674Author: Jeff Smith <whydoubt@gmail.com> 10675Date: Sun Feb 4 23:17:52 2018 -0600 10676 10677 xfree86: Do not use uninitialized pointer during probe 10678 10679 Commits b5dffbb and d75ffcd introduce code in xf86platformProbe() that 10680 references a member of xf86configptr. However, when using the 10681 "-configure" option, xf86configptr may not be initialized when 10682 xf86platformProbe() is called. 10683 10684 Avoid referencing a member of xf86configptr if uninitialized. 10685 10686 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100405 10687 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10688 Reviewed-by: Adam Jackson <ajax@redhat.com> 10689 10690commit 1e23f03dd5fb6c981ef6d64c084f72fc6820ed71 10691Author: Jan Beich <jbeich@freebsd.org> 10692Date: Fri Jul 7 16:31:48 2017 +0200 10693 10694 shm, xwayland: prefer atomic close-on-exec without O_TMPFILE 10695 10696 Signed-off-by: Jan Beich <jbeich@FreeBSD.org> 10697 Reviewed-by: Adam Jackson <ajax@redhat.com> 10698 10699commit 3e3b8a40fee77d1af6ca0c2946ff276d555ddea8 10700Author: Michal Srb <msrb@suse.com> 10701Date: Mon Mar 27 15:03:38 2017 +0300 10702 10703 modesetting: Check for -1 before converting to unsigned int. 10704 10705 dri2.c:516:21: warning: comparison of unsigned expression < 0 is always false [-Wtautological-compare] 10706 if (front->name < 0) 10707 10708 Prevents a failure from being ignored. 10709 10710commit a9f68688f313752b1c6dfcbf6bf54f8c41fd811d 10711Author: Michal Srb <msrb@suse.com> 10712Date: Mon Mar 27 15:03:37 2017 +0300 10713 10714 xfree86: Silence always true condition warning. 10715 10716 xf86pciBus.c:1464:21: warning: comparison of constant 256 with expression of type 'uint8_t' (aka 'unsigned char') is always true [-Wtautological-constant-out-of-range-compare] 10717 if (pVideo->bus < 256) 10718 10719 The code used to be in xf86FormatPciBusNumber and compared parameter which was int, but since b967bf2a it was inlined now it works with uint8_t. 10720 10721commit b75835a4358065378142691bf4a20140981a410f 10722Author: Raimonds Cicans <ray@apollo.lv> 10723Date: Mon Nov 6 15:26:33 2017 +0200 10724 10725 Xdmx: fix input driver initialization code paths 10726 10727 Problem: Xdmx exit with error "(Fatal Error) dmx: Unknown input argument: " 10728 if local input driver is specified (example: -input local,usb-kbd,usb-mou). 10729 10730 Cause: code block for remote input drivers is placed in code path common 10731 for all drivers. 10732 10733 Fix: affected code block moved to remote input driver code path part. 10734 10735 Signed-off-by: Raimonds Cicans <ray@apollo.lv> 10736 Reviewed-by: Adam Jackson <ajax@redhat.com> 10737 10738commit cbf09619f2355d14c103eff69aaf191cca614c06 10739Author: Adam Jackson <ajax@redhat.com> 10740Date: Mon Nov 6 12:56:05 2017 -0500 10741 10742 meson: Get more xkb configuration from xkbcomp.pc 10743 10744 Signed-off-by: Adam Jackson <ajax@redhat.com> 10745 Reviewed-by: Daniel Stone <daniels@collabora.com> 10746 10747commit 93c16b0524711cd3cd8e32f151c53b5d6522e967 10748Author: Alexander Volkov <a.volkov@rusbitech.ru> 10749Date: Fri Feb 2 15:17:15 2018 +0300 10750 10751 Xephyr: Call forgotten XShmDetach if can't mmap SHM segment 10752 10753 Reviewed-by: Adam Jackson <ajax@redhat.com> 10754 10755commit 6b26a7bda9efa93440734ede0382a3e9a6761365 10756Author: Alex Goins <agoins@nvidia.com> 10757Date: Mon Jan 8 18:44:25 2018 -0800 10758 10759 randr: Fix rotation check in ProcRRSetScreenSize() 10760 10761 ProcRRSetScreenSize() does bounds checking to ensure that none of the CRTCs have 10762 a viewport that extends beyond the new screen size. In doing so, it accounts for 10763 if the CRTC is rotated 90 or 270 degrees, swapping width and height. 10764 10765 However, it does so by testing if crtc->rotation is equal to RR_Rotate_90 or 10766 RR_Rotate_270. crtc->rotation is a bit mask, and it includes reflection as well 10767 as rotation. If a CRTC is reflected as well as rotated, it will incorrectly fail 10768 this test, resulting in incorrect dimensions being used to verify the validity 10769 of the new screen size. In some cases, this can cause valid uses of 10770 ProcRRSetScreenSize() to fail with BadMatch. 10771 10772 This patch fixes the issue by testing that the bits RR_Rotate_90 or 10773 RR_Rotate_270 are set, rather than testing for equality. 10774 10775 Signed-off-by: Alex Goins <agoins@nvidia.com> 10776 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 10777 10778commit 90996f5909aab4bc9aa4011a6a6d0555a7aa3adf 10779Author: Alexander Volkov <a.volkov@rusbitech.ru> 10780Date: Wed Jan 31 15:36:36 2018 +0300 10781 10782 Xephyr: Prefer using MIT-SHM FD-passing when possible 10783 10784 This makes the shared memory visible only for the Xephyr 10785 and the X server to which it is connected. 10786 10787 Signed-off-by: Alexander Volkov <a.volkov@rusbitech.ru> 10788 Reviewed-by: Keith Packard <keithp@keithp.com> 10789 10790commit 8a220bd83c3e23de7e07d3976bfc1248c38558d4 10791Author: Alexander Volkov <a.volkov@rusbitech.ru> 10792Date: Wed Jan 31 15:36:35 2018 +0300 10793 10794 Xephyr: Extract functions to create/delete shared memory segments 10795 10796 Signed-off-by: Alexander Volkov <a.volkov@rusbitech.ru> 10797 Reviewed-by: Keith Packard <keithp@keithp.com> 10798 10799commit 510e7d0d869dcc008b1e11e21961f2c1c3cb9cd3 10800Author: Jeff Smith <whydoubt@gmail.com> 10801Date: Tue Dec 26 22:10:52 2017 -0600 10802 10803 os: Always return microseconds from GetTimeInMicros() 10804 10805 When a monotonic clock is not available, GetTimeInMicros() returns the 10806 time in nanoseconds. Instead, return the time in microseconds, as the 10807 name indicates. 10808 10809 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10810 Reviewed-by: Adam Jackson <ajax@redhat.com> 10811 10812commit 94e7ca6a5667eeb13c078496a07d3e6cf1ec95cc 10813Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 10814Date: Wed Jan 31 17:33:48 2018 +0100 10815 10816 rrmonitor: allocate using the correct type 10817 10818 Monitor outputs are of type RROutput, not RRCrtc. 10819 10820 (Which are both XID, so this makes no difference in practice, other than 10821 being technically correct.) 10822 10823 Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 10824 Reviewed-by: Adam Jackson <ajax@redhat.com> 10825 10826commit ff410c56bfd6dcdf5d252ef0ba3f4c6fde91774b 10827Author: Jeff Smith <whydoubt@gmail.com> 10828Date: Fri Jan 26 06:25:22 2018 -0600 10829 10830 glamor: tidy up some gradient color formulas 10831 10832 - Avoid some needlessly complex equations for calculating the color at a 10833 point in a gradient field. 10834 - Avoid calculating certain values multiple times. 10835 - Use similar variable names across the two versions of the get_color 10836 function where practical. 10837 10838 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10839 Reviewed-by: Adam Jackson <ajax@redhat.com> 10840 10841commit 3bd17dff1ac5a4685a5da92a9982f621effee658 10842Author: Jeff Smith <whydoubt@gmail.com> 10843Date: Fri Jan 26 06:25:21 2018 -0600 10844 10845 glamor: remove unused variables in linear gradient shader 10846 10847 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10848 Reviewed-by: Adam Jackson <ajax@redhat.com> 10849 10850commit 5815c7b5951fd46d69e5c40144b64e516c7afdbf 10851Author: Jeff Smith <whydoubt@gmail.com> 10852Date: Fri Jan 26 06:25:20 2018 -0600 10853 10854 glamor: fix no-reflect case for gradients 10855 10856 When compositing a no-reflect gradient, 'before' the gradient is empty, 10857 but 'after' the gradient is padded with the final color. Both sides are 10858 supposed to be empty. 10859 10860 This is fixed by moving the virtual stops to match the first and last 10861 client-supplied stops for no-reflect gradients, then causing everything 10862 'before' the initial virtual stop and 'after' the final virtual stop to 10863 emit rgba(0,0,0,0). This does not impact gradients using the other 10864 reflect modes. 10865 10866 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10867 Reviewed-by: Adam Jackson <ajax@redhat.com> 10868 10869commit 3e377e238f7257fd01e56a4a25dfd77e033673e4 10870Author: Jeff Smith <whydoubt@gmail.com> 10871Date: Fri Jan 26 06:25:19 2018 -0600 10872 10873 glamor: fix repeat-reflect case in linear gradient shader 10874 10875 If _pt_distance is negative, it causes the final distance to be negative 10876 in the repeat-reflect case. Moving the scaling by _pt_distance earlier 10877 avoids this problem, and simplifies some equations as a bonus. 10878 10879 Bugzilla: https://bugs.freedesktop.org/98508 10880 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10881 Reviewed-by: Adam Jackson <ajax@redhat.com> 10882 10883commit 9b7b8720ebc8028b14796a66d0a21b002682a83c 10884Author: Jeff Smith <whydoubt@gmail.com> 10885Date: Sat Jan 27 00:14:02 2018 -0600 10886 10887 edid: use value returned from pruning duplicate modes 10888 10889 xf86PruneDuplicateModes is passed a linked list of modes, and after 10890 pruning the duplicate nodes, the new head of the list is returned. If 10891 the first element is removed, the head of the list will change and the 10892 returned value needs to be assigned. 10893 10894 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103722 10895 Thanks: John Lumby <johnlumby@hotmail.com> 10896 Signed-off-by: Jeff Smith <whydoubt@gmail.com> 10897 Reviewed-by: Adam Jackson <ajax@redhat.com> 10898 10899commit dd00e5466a0e4ea313d1860824da4123692827ed 10900Author: Adam Jackson <ajax@redhat.com> 10901Date: Wed Jan 24 15:40:22 2018 -0500 10902 10903 xfree86: Remove broken RANDR disabling logic (v4) 10904 10905 The only way to get at xf86Info.disableRandR from configuration is 10906 Option "RANDR" "foo" in ServerFlags, which probably nobody is using 10907 seeing as it's not documented. The other way it could be set is if a 10908 screen supports RANDR 1.2, in which case we set it to avoid trying to 10909 use the RANDR 1.1 compat code. If the second screen is not 1.2-aware 10910 then this would mean we don't do RANDR setup on the second screen at 10911 all, which would almost certainly crash the first time you try to do 10912 RANDR operations on the second screen. 10913 10914 Fix that all by deletion, and just check whether the screen already has 10915 RANDR initialized before installing the stub support. If you want to 10916 disable RANDR, use the Extensions section of xorg.conf instead. 10917 10918 v2: Also remove a now entirely pointless log message, telling you to 10919 ignore a line we will no longer print. 10920 10921 v3: Explain the fallback path in InitOutput. (Keith Packard) 10922 10923 v4: Check whether the RANDR private key is initialized before trying to 10924 use it to look up the screen private. 10925 10926 Signed-off-by: Adam Jackson <ajax@redhat.com> 10927 Reviewed-by: Keith Packard <keithp@keithp.com> 10928 10929commit 9a416a478c1a29d1aa12c394acf21fb1b4a3aca9 10930Author: Hawking Zhang <Hawking.Zhang@amd.com> 10931Date: Thu Jan 25 18:03:57 2018 +0100 10932 10933 glamor: Add 30bit RGB color format support 10934 10935 Signed-off-by: Hawking Zhang <Hawking.Zhang@amd.com> 10936 10937 [ Michel Dänzer: Adapt to glamor changes since 1.19 ] 10938 10939 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 10940 Acked-by: Alex Deucher <alexander.deucher@amd.com> 10941 10942commit fc8b7d05e74a6351df56ad8b17216aeb0dcc016b 10943Author: Olivier Fourdan <ofourdan@redhat.com> 10944Date: Thu Jan 25 13:43:09 2018 +0100 10945 10946 xwayland: place a manual redirect on windows 10947 10948 Place a manual redirect on windows on xwl_realize_window() and remove 10949 it on xwl_unrealize_window() to avoid the X11 window manager removing 10950 its redirect before Xwayland has unrealized the window (e.g. if the X11 10951 window manager has terminated unexpectedly) 10952 10953 Suggested by Daniel Stone <daniel@fooishbar.org> 10954 10955 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 10956 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 10957 10958commit 3362422e8413dd9f231cfac50ce0a0862525b1bf 10959Author: Olivier Fourdan <ofourdan@redhat.com> 10960Date: Wed Jan 24 17:45:37 2018 +0100 10961 10962 xwayland: remove dirty window unconditionally on unrealize 10963 10964 This is a rare occurrence of a crash in Xwayland for which I don't have 10965 the reproducing steps, just a core file. 10966 10967 The backtrace looks as follow: 10968 10969 #0 raise () from /usr/lib64/libc.so.6 10970 #1 abort () from /usr/lib64/libc.so.6 10971 #2 OsAbort () at utils.c:1361 10972 #3 AbortServer () at log.c:877 10973 #4 FatalError () at log.c:1015 10974 #5 OsSigHandler () at osinit.c:154 10975 #6 <signal handler called> 10976 #7 xwl_glamor_pixmap_get_wl_buffer () at xwayland-glamor.c:162 10977 #8 xwl_screen_post_damage () at xwayland.c:514 10978 #9 block_handler () at xwayland.c:665 10979 #10 BlockHandler () at dixutils.c:388 10980 #11 WaitForSomething () at WaitFor.c:219 10981 #12 Dispatch () at dispatch.c:422 10982 #13 dix_main () at main.c:287 10983 10984 The crash is caused by dereferencing “xwl_pixmap->buffer” in 10985 xwl_glamor_pixmap_get_wl_buffer() because “xwl_pixmap” is NULL. 10986 10987 Reason for this is because the corresponding pixmap is from the root 10988 window and xwayland is rootless by default. 10989 10990 This can happen if the window was mapped, redirected, damaged and 10991 unredirected immediately, before the damage is processed by Xwayland. 10992 10993 Make sure to remove the dirty window from the damage list on unrealize 10994 to prevent this from happening. 10995 10996 Credit goes to Adam Jackson <ajax@nwnk.net> and Daniel Stone 10997 <daniel@fooishbar.org> for finding the root cause the issue. 10998 10999 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 11000 Reviewed-by: Daniel Stone <daniels@collabora.com> 11001 11002commit 29a5423abd4bc6cb7bb800924a27b5d90951e408 11003Author: Jason Gerecke <killertofu@gmail.com> 11004Date: Tue Nov 7 11:09:44 2017 -0800 11005 11006 xwayland: Support for BTN_STYLUS3 kernel events 11007 11008 BTN_STYLUS3 has been introduced by the Linux 4.15 kernel to report the 11009 status of the third button present on Wacom's new "Pro Pen 3D" stylus. 11010 Treat this button like xf86-input-wacom and send a button 8 event 11011 ("navigate back") when received from Wayland. 11012 11013 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 11014 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 11015 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 11016 11017commit 71348e99a8e6a95542e483b93839168ca8f51f81 11018Author: Michal Srb <msrb@suse.com> 11019Date: Mon Nov 27 09:59:01 2017 +0100 11020 11021 os/inputthread: Force unlock when stopping thread. 11022 11023 The inputthread is kept locked all the time while X server's VT is not active. 11024 If the X server is terminated while not active, it will be stuck forever in 11025 InputThreadFini waiting for the thread to join, but it wouldn't because it is 11026 locked. 11027 11028 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=103782 11029 Signed-off-by: Michal Srb <msrb@suse.com> 11030 Reviewed-by: Adam Jackson <ajax@redhat.com> 11031 11032commit d5c23b29487d5ff440abf5ed0beb22c00f21e387 11033Author: Peter Harris <pharris@opentext.com> 11034Date: Thu Dec 14 12:21:19 2017 -0500 11035 11036 composite: Propagate damagedDescendants when reparented 11037 11038 If a window that is fully covered by an automatic-redirected descendant 11039 (even implicitly, eg. via BackingStores) is reparented, the automatic 11040 updates could be broken if the new parent is not marked as having 11041 damaged descendants. 11042 11043 Fix this issue by propagating the damagedDescendants flag whenever a 11044 window is reparented. 11045 11046 Reviewed-by: Adam Jackson <ajax@redhat.com> 11047 Signed-off-by: Peter Harris <pharris@opentext.com> 11048 11049commit ac138f9b31b0fba00742edbc3326afe66e28099a 11050Author: Rinat Ibragimov <ibragimovrinat@mail.ru> 11051Date: Tue Oct 24 01:43:43 2017 +0300 11052 11053 modesetting: setup colormap 11054 11055 Signed-off-by: Rinat Ibragimov <ibragimovrinat@mail.ru> 11056 Reviewed-by: Michel Dänzer <michel.daenzer at amd.com> 11057 11058commit 5cb330cd5d887b20351604ac080e8b71a4e28be7 11059Author: Adam Jackson <ajax@redhat.com> 11060Date: Wed Jan 24 15:24:00 2018 -0500 11061 11062 Revert "xfree86: Remove broken RANDR disabling logic (v3)" 11063 11064 Tsk. This broke vesa for me, the rrGetScrPriv in InitOutput will crash 11065 if randr's screen private key hasn't been initialized yet. That seems 11066 dumb, but let's not leave it broken. 11067 11068 This reverts commit c08d7c1cdde6a844338ed4c3645b00bf25843a31. 11069 11070commit c08d7c1cdde6a844338ed4c3645b00bf25843a31 11071Author: Adam Jackson <ajax@redhat.com> 11072Date: Tue Jan 23 12:10:52 2018 -0500 11073 11074 xfree86: Remove broken RANDR disabling logic (v3) 11075 11076 The only way to get at xf86Info.disableRandR from configuration is 11077 Option "RANDR" "foo" in ServerFlags, which probably nobody is using 11078 seeing as it's not documented. The other way it could be set is if a 11079 screen supports RANDR 1.2, in which case we set it to avoid trying to 11080 use the RANDR 1.1 compat code. If the second screen is not 1.2-aware 11081 then this would mean we don't do RANDR setup on the second screen at 11082 all, which would almost certainly crash the first time you try to do 11083 RANDR operations on the second screen. 11084 11085 Fix that all by deletion, and just check whether the screen already has 11086 RANDR initialized before installing the stub support. If you want to 11087 disable RANDR, use the Extensions section of xorg.conf instead. 11088 11089 v2: Also remove a now entirely pointless log message, telling you to 11090 ignore a line we will no longer print. 11091 11092 v3: Explain the fallback path in InitOutput. (Keith Packard) 11093 11094 Signed-off-by: Adam Jackson <ajax@redhat.com> 11095 Reviewed-by: Keith Packard <keithp@keithp.com> 11096 11097commit da8de2a7f6ab52ef52039b0dc9978260232a34a6 11098Author: Olivier Fourdan <ofourdan@redhat.com> 11099Date: Thu Sep 7 17:43:16 2017 +0200 11100 11101 xwayland: Add optional xdg-output support 11102 11103 The xdg-output protocol aims at describing outputs in way which is 11104 more in line with the concept of an output on desktop oriented systems. 11105 11106 For now it just features the position and logical size which describe 11107 the output position and size in the global compositor space. 11108 11109 This is however much useful for Xwayland to advertise the output size 11110 and position to X11 clients which need this to configure their surfaces 11111 in the global compositor space as the compositor may apply a different 11112 scale from what is advertised by the output scaling property (to achieve 11113 fractional scaling, for example). 11114 11115 This was added in wayland-protocols 1.10. 11116 11117 Reviewed-by: Adam Jackson <ajax@redhat.com> 11118 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 11119 11120commit 75408f53d4e203e462b2f13ea4b06264f0e59ad2 11121Author: Michel Dänzer <michel.daenzer@amd.com> 11122Date: Wed Jan 24 12:28:38 2018 +0100 11123 11124 xf86XvMCScreenInit: Clear pScreenPriv->dixinfo when freeing pAdapt 11125 11126 Fixes double-free later in xf86XvMCCloseScreen, which would generally 11127 cause fireworks. 11128 11129 Reviewed-by: Adam Jackson <ajax@redhat.com> 11130 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 11131 11132commit 76732f498f1e73fb081841a04faf068660f3d5c7 11133Author: Michel Dänzer <michel.daenzer@amd.com> 11134Date: Wed Jan 24 11:40:50 2018 +0100 11135 11136 Revert "present: Only send PresentCompleteNotify events to the presenting client" 11137 11138 This reverts commit 5c00e693631475679c1c2504e03177652ec7de28. 11139 11140 It broke the Gallium nine state tracker, which makes PresentPixmap 11141 requests on one display connection and processes the corresponding 11142 events on another one. 11143 11144 The issue that motivated this change is prevented on the client side by 11145 https://cgit.freedesktop.org/mesa/mesa/commit/?id=7b0e8264dd21ae05521d08d41fecd84139401fef 11146 11147 Bugzilla: https://bugs.freedesktop.org/104678 11148 11149commit f704620e5db8305c54c6b6981657d801dbf6849b 11150Author: Adam Jackson <ajax@redhat.com> 11151Date: Wed Jan 10 13:05:43 2018 -0500 11152 11153 glx: Move provider setup interface to glx_extinit.h 11154 11155 Keeps us from needing to add glx/ to the include path everywhere, since 11156 we can't add it to the dix include path because the header file names 11157 conflict with glxproxy. 11158 11159 Signed-off-by: Adam Jackson <ajax@redhat.com> 11160 Reviewed-by: Eric Anholt <eric@anholt.net> 11161 11162commit c2b2f06aa03fe738301ebaa82f220d9cfed5ab1a 11163Author: Adam Jackson <ajax@redhat.com> 11164Date: Wed Jan 10 13:05:42 2018 -0500 11165 11166 miinitext: General cleanup (v2) 11167 11168 This really just wants to be the list of disable booleans and 11169 initialization functions, and nothing else. Stop including the protocol 11170 headers from extinit.h, remove a stray mention of xgl, and move an 11171 XInput declaration to a better place. 11172 11173 v2: A bunch of drivers assume they'll get the DPMS tokens implicitly, 11174 so add it to globals.h. 11175 11176 Signed-off-by: Adam Jackson <ajax@redhat.com> 11177 Reviewed-by: Eric Anholt <eric@anholt.net> 11178 11179commit 4d5aab66c052795c7f0381a3dfc7293c9a41e441 11180Author: Keith Packard <keithp@keithp.com> 11181Date: Thu Dec 21 18:54:39 2017 -0800 11182 11183 xfree86: Disable cursor whenever turning off CRTC during modeset 11184 11185 This makes sure the CRTC's cursor is hidden before we hand the CRTC 11186 over to some other application. 11187 11188 Signed-off-by: Keith Packard <keithp@keithp.com> 11189 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 11190 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 11191 11192commit 29f79bedf2c80241ba4b482db6ead08a5709a982 11193Author: Keith Packard <keithp@keithp.com> 11194Date: Thu Dec 21 18:54:37 2017 -0800 11195 11196 randr: Declare incoming property values const 11197 11198 RRChangeOutputProperty and RRConfigureOutputProperty should not modify 11199 their parameters, and callers may want to pass pointers to fixed data, 11200 so declare the value pointers as const in both cases. 11201 11202 Reviewed-by: Adam Jackson <ajax@redhat.com> 11203 Signed-off-by: Keith Packard <keithp@keithp.com> 11204 11205commit a12485ed846b852ca14d17d1e58c8b0f2399e577 11206Author: Keith Packard <keithp@keithp.com> 11207Date: Thu Dec 21 18:54:34 2017 -0800 11208 11209 xf86-video-modesetting: Update property values at detect and uevent time 11210 11211 We were updating the link-status property when a uevent came in, but 11212 we also want to update the non-desktop property, and potentially 11213 others as well. We also want to check at detect time in case we don't 11214 get a hotplug event. 11215 11216 This patch updates every property provided by the kernel, sending 11217 changes to DIX so it can track things as well. 11218 11219 Reviewed-by: Adam Jackson <ajax@redhat.com> 11220 Signed-off-by: Keith Packard <keithp@keithp.com> 11221 11222commit 44d5f2eb8a2f92571698adec39ac569b71da5a1b 11223Author: Keith Packard <keithp@keithp.com> 11224Date: Thu Dec 21 18:54:33 2017 -0800 11225 11226 xf86-video-modesetting: Support new vblank kernel API [v2] 11227 11228 drmCrtcGetSequence returns the current vblank sequence and time. 11229 11230 drmCrtcQueueSequence queues an event for delivery at a specified 11231 vblank sequence. 11232 11233 Use these (when available) in preference to drmWaitVBlank. 11234 11235 v2: Remove FIRST_PIXEL_OUT_FLAG. This has been removed from the kernel 11236 API. 11237 11238 Reviewed-by: Adam Jackson <ajax@redhat.com> 11239 Signed-off-by: Keith Packard <keithp@keithp.com> 11240 11241commit 627dfc2f836a3a1fc6afbd3f2335f88f5ca55ffd 11242Author: Adam Jackson <ajax@redhat.com> 11243Date: Mon Jan 22 15:43:05 2018 -0500 11244 11245 automake: Fix 'make distcheck' 11246 11247 Add some missing meson.builds to the distball, and update the meson 11248 build test to work when the distcheck srcdir != builddir. The test build 11249 directory will be $(srcdir)/_distcheck_build so srcdir will need to be 11250 writable; this shouldn't be too much to ask I hope. 11251 11252 Signed-off-by: Adam Jackson <ajax@redhat.com> 11253 11254commit 2d29daf4c69385513a322002570ec2c41cfc838f 11255Author: Jon Turney <jon.turney@dronecode.org.uk> 11256Date: Thu Jan 18 16:18:12 2018 +0000 11257 11258 meson.build: Fix hw/xwin build when dependencies are installed in a non-default location 11259 11260 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 11261 Reviewed-by: Eric Anholt <eric@anholt.net> 11262 11263commit 514d2c243e01b114de304ea72226bbf8772b1b11 11264Author: Jon Turney <jon.turney@dronecode.org.uk> 11265Date: Wed Jan 17 13:43:47 2018 +0000 11266 11267 meson: Use and prefer tirpc for Secure RPC authentication 11268 11269 Reviewed-by: Adam Jackson <ajax@redhat.com> 11270 11271commit f72587ecc7e1dedfb20a999a0e600b83c06a1b29 11272Author: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 11273Date: Wed Dec 20 13:18:45 2017 +0200 11274 11275 xwayland: reduce over-damage 11276 11277 If an X11 app draws a little here, some there, and a tiny bit in the 11278 opposite corner, using RegionExtents for the damage to be sent to the 11279 Wayland compositor will cause massive over-damaging. 11280 11281 However, we cannot blindly send an arbitrary number of damage 11282 rectangles, because there is a risk of overflowing the Wayland 11283 connection. If that happens, it triggers an abort in libwayland-client. 11284 11285 Try to be more accurate with the damage by sending up to 256 rectangles 11286 per window, and fall back to extents otherwise. The number is completely 11287 arbitrary. 11288 11289 Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 11290 Reviewed-by: Daniel Stone <daniels@collabora.com> 11291 11292commit a5e9bcad7ad0887f804905b482894b85751519fb 11293Author: Martin Wilck <mwilck@suse.com> 11294Date: Tue Jan 9 20:33:09 2018 +0100 11295 11296 xfree86: add default modes for 16:9 and 16:10 11297 11298 Improve the user experience for users with wide screens by adding standard 11299 16:9 and 16:10 modes to extramodes, as suggested previously 11300 (https://lists.x.org/archives/xorg-devel/2016-February/048866.html). 11301 Tested successfully on my laptop. Feedback welcome. 11302 11303 See also https://bugs.freedesktop.org/show_bug.cgi?id=37858. 11304 11305 Signed-off-by: Martin Wilck <mwilck@suse.com> 11306 Reviewed-by: Adam Jackson <ajax@redhat.com> 11307 11308commit 528cf6fcee1839cb829652a49c1b1bf45d6a37fb 11309Author: David Weinehall <david.weinehall@linux.intel.com> 11310Date: Thu Jan 18 15:12:46 2018 +0200 11311 11312 sync: Fix diffgreater comparison 11313 11314 xsync: Fix diffgreater comparison 11315 11316 While transitioning from CARD64 to int64, 11317 the GreaterThan call was mistakenly transformed into ">=". 11318 Part of this was fixed already in 11319 commit 8060196a3e80a3c0ad2c0abbe459416821cd366c 11320 11321 This patch fixes the remaining issue. 11322 11323 Signed-off-by: David Weinehall <david.weinehall@linux.intel.com> 11324 Reviewed-by: Keith Packard <keithp@keithp.com> 11325 11326commit 32b42627217917337553121191e9bc3b423b8d99 11327Author: Keith Packard <keithp@keithp.com> 11328Date: Tue Jan 16 11:44:37 2018 -0800 11329 11330 modesetting: Use seq instead of msc in ms_queue_vblank failure path 11331 11332 When the call to queue a vblank event fails, we need to clean up by 11333 removing the user-space queue entry. That is indexed by the local 11334 sequence number, not by the kernel vblank count. The call in this 11335 case was just passing the wrong value. 11336 11337 Signed-off-by: Keith Packard <keithp@keithp.com> 11338 Reviewed-by: Eric Anholt <eric@anholt.net> 11339 11340commit 15d91df47424127b9e8d0d73692e2196c79dd3fe 11341Author: Adam Jackson <ajax@redhat.com> 11342Date: Wed Jan 10 12:07:41 2018 -0500 11343 11344 x86emu: Teach the debug code about varargs 11345 11346 With -Wformat-nonliteral and a debug build you'd get yelled at here: 11347 11348 ../hw/xfree86/x86emu/x86emu/debug.h:188:9: warning: format not a string literal, argument types not checked [-Wformat-nonliteral] 11349 11350 To fix this, rewrite the printf code to actually use varargs and the 11351 appropriate format attribute. All callers of DECODE_PRINTF() pass a 11352 string with no % specifiers, so we pass that as the argument to 11353 printf("%s"). For DECODE_PRINTF2() we just pass the args through. 11354 11355 Signed-off-by: Adam Jackson <ajax@redhat.com> 11356 Reviewed-by: Eric Anholt <eric@anholt.net> 11357 11358commit 1274015186a8457c38c3b5dcc9965c62f1d2a7a6 11359Author: Adam Jackson <ajax@redhat.com> 11360Date: Mon Jan 8 15:50:29 2018 -0500 11361 11362 build: Remove <*dbm.h> checks 11363 11364 Formerly used by the rgb database code, which hasn't been a thing in 11365 over a decade. 11366 11367 Reviewed-by: Eric Anholt <eric@anholt.net> 11368 Signed-off-by: Adam Jackson <ajax@redhat.com> 11369 11370commit c3fbe2bbff19c67179f52b8dcd27b576a958fde7 11371Author: Adam Jackson <ajax@redhat.com> 11372Date: Mon Jan 8 15:50:28 2018 -0500 11373 11374 meson: Enable SUN-DES-1 auth code 11375 11376 Signed-off-by: Adam Jackson <ajax@redhat.com> 11377 Reviewed-by: Eric Anholt <eric@anholt.net> 11378 11379commit 4d82a150b2ee29c1025408cdb9ece255452a81bd 11380Author: Adam Jackson <ajax@redhat.com> 11381Date: Tue Jan 9 11:48:10 2018 -0500 11382 11383 animcur: Handle allocation failure for the animation timer 11384 11385 Signed-off-by: Adam Jackson <ajax@redhat.com> 11386 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 11387 Tested-by: Aaron Plattner <aplattner@nvidia.com> 11388 11389commit de60245e05c0d2528d4ff42557a044387e53315c 11390Author: Adam Jackson <ajax@redhat.com> 11391Date: Tue Jan 9 10:54:05 2018 -0500 11392 11393 animcur: Fix transitions between animated cursors 11394 11395 We weren't cancelling the old timer when changing cursors, making things 11396 go all crashy. Logically we could always cancel the timer first, but 11397 then we'd have to call TimerSet to re-arm ourselves, and GetTimeInMillis 11398 is potentially expensive. 11399 11400 Reported-by: https://devtalk.nvidia.com/default/topic/1028172/linux/titan-v-ubuntu-16-04lts-and-387-34-driver-crashes-badly/post/5230967/#5230967 11401 Signed-off-by: Adam Jackson <ajax@redhat.com> 11402 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 11403 Tested-by: Aaron Plattner <aplattner@nvidia.com> 11404 11405commit a09fbe6c82efc674fc3790c7ffad7df70866a7df 11406Author: Adam Jackson <ajax@redhat.com> 11407Date: Wed Dec 13 11:44:38 2017 -0500 11408 11409 glx: Stop printing messages about what GLX extensions we enable 11410 11411 glxinfo already exists, use it. 11412 11413 Signed-off-by: Adam Jackson <ajax@redhat.com> 11414 Reviewed-by: Eric Anholt <eric@anholt.net> 11415 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 11416 11417commit 4ed8d5c9463cbef0693d98eb9faf7f9c2222c383 11418Author: Adam Jackson <ajax@redhat.com> 11419Date: Tue Nov 14 15:33:54 2017 -0500 11420 11421 ddc: Port some paranoia from drm_edid.c 11422 11423 Avoid adding a "standard" mode timing if we've already got a matching 11424 detailed timing. To help with that, parse CEA blocks for detailed 11425 timings before doing standard or established timings. 11426 11427 Signed-off-by: Adam Jackson <ajax@redhat.com> 11428 11429commit 652913cd9474946bcb29271602bacfd98f46ad0b 11430Author: Adam Jackson <ajax@redhat.com> 11431Date: Wed Dec 13 14:53:56 2017 -0500 11432 11433 os: Fix a type error in the IPv6 XDMCP code 11434 11435 Building with strict-aliasing rightly chirps here: 11436 11437 ../os/xdmcp.c: In function ‘XdmcpRegisterConnection’: 11438 ../os/xdmcp.c:489:31: warning: dereferencing type-punned pointer will break strict-aliasing rules [-Wstrict-aliasing] 11439 &((struct sockaddr_in6 *) &address)->sin6_addr.s6_addr[12]; 11440 ^~~~~~~~~~~~ 11441 11442 We have "const char *address", so &address here is a char ** (i.e., it 11443 points to the slot on the stack containing the pointer to the character 11444 array passed in as an argument). Casting that to a struct sockaddr_in6 * 11445 is wrong, because it means that area of the stack will be reinterpreted 11446 as a struct sockaddr_in6. 11447 11448 Instead, cast address, not &address. 11449 11450 Signed-off-by: Adam Jackson <ajax@redhat.com> 11451 11452commit da4ffb2f6a0b5a039ae1362ae71e9b47441f90d2 11453Author: Jon Turney <jon.turney@dronecode.org.uk> 11454Date: Fri Dec 22 18:44:01 2017 +0000 11455 11456 meson: Correct the option for disabled int10 from 'disabled' to 'false' 11457 11458 Fix meson_option.txt to align with the check of the int10 option against 11459 'disabled', not 'false' in hw/xfree/meson.build, to see if it shouldn't be 11460 built at all. 11461 11462 This keeps everything consistent that 'false' always turns things off. 11463 11464 Not noticed before as options weren't validated against choices until meson 11465 0.43 11466 11467 Reviewed-by: Adam Jackson <ajax@redhat.com> 11468 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 11469 11470commit c601c8faf54ff9e3bcbc653421828d71042deef7 11471Author: Helmut Grohne <helmut@subdivi.de> 11472Date: Thu Dec 21 11:48:15 2017 +0100 11473 11474 build: guess availability of monotonic clock for cross compilation 11475 11476 When cross compiling, the value of MONOTONIC_CLOCK would be "cross 11477 compiling", because AC_RUN_IFELSE doesn't work. However when enabling 11478 wayland, a monotonic clock is required and configure aborts. 11479 11480 We change detection of CLOCK_MONOTONIC to degrade it gracefully from a 11481 run check to a declaration check in case of cross compilation based on 11482 the assumption that most systems will have a monotonic clock and those 11483 that don't won't be able to run Xwayland anyway. The trade-off 11484 essentially is either "always fail cross compilation" or "produce an 11485 unusable Xwayland for unusual platform" and this commit switches to the 11486 latter. 11487 11488 Signed-off-by: Helmut Grohne <helmut@subdivi.de> 11489 Bug-Debian: https://bugs.debian.org/882531 11490 Reviewed-by: Adam Jackson <ajax@redhat.com> 11491 11492commit a309323328d9d6e0bf5d9ea1d75920e53b9beef3 11493Author: Peter Hutterer <peter.hutterer@who-t.net> 11494Date: Fri Jan 5 11:58:42 2018 +1000 11495 11496 config: fix NULL value detection for ID_INPUT being unset 11497 11498 Erroneous condition caused us to keep going with all devices that didn't have 11499 ID_INPUT set. 11500 11501 Fixes: 5aad81445c8c3d6 11502 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104382 11503 Reviewed-by: Adam Jackson <ajax@redhat.com> 11504 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 11505 11506commit dbfbe58b94ec07a45fd6170f96bffec55051724e 11507Author: Corentin Rossignon <corentin.rossignon@gmail.com> 11508Date: Thu Jan 4 08:34:01 2018 +0100 11509 11510 glamor: Specify GLSL version for xv shader 11511 11512 SAMPLE is a GLSL keyword in newer OpenGL version. 11513 This fix issue with gnome-shell and playing video using xv 11514 11515 Signed-off-by: Corentin Rossignon <corentin.rossignon@gmail.com> 11516 Bugzilla: https://bugs.freedesktop.org/104405 11517 Reviewed-by: Adam Jackson <ajax@redhat.com> 11518 11519commit ab54bc295cd05281e55bd4d9c37216c0a929fc83 11520Author: Adam Jackson <ajax@redhat.com> 11521Date: Thu Oct 26 15:33:14 2017 -0400 11522 11523 animcur: Stop tracking the last display time in the SpriteInfoRec 11524 11525 Reviewed-by: Robert Morell <rmorell@nvidia.com> 11526 Tested-by: Robert Morell <rmorell@nvidia.com> 11527 Signed-off-by: Adam Jackson <ajax@redhat.com> 11528 11529commit 094a63d56fbfb9e23210cc9ac538fb198af37cee 11530Author: Adam Jackson <ajax@redhat.com> 11531Date: Thu Oct 26 15:24:39 2017 -0400 11532 11533 animcur: Run the timer from the device, not the screen 11534 11535 This is very slightly more efficient since the callback now doesn't need 11536 to walk every input device, instead we know exactly which device's 11537 cursor is being updated. AnimCurTimerNotify() gets outdented nicely as a 11538 result. A more important side effect is that we can stop using the 11539 TimerAbsolute mode and just pass in the relative delay. 11540 11541 In AnimCurSetCursorPosition, we no longer need to rearm the timer with 11542 the new screen; it is enough to update the device's state. In 11543 AnimCurDisplayCursor we need to notice when we're switching from 11544 animated cursor to regular and cancel the existing timer. 11545 11546 Reviewed-by: Robert Morell <rmorell@nvidia.com> 11547 Tested-by: Robert Morell <rmorell@nvidia.com> 11548 Signed-off-by: Adam Jackson <ajax@redhat.com> 11549 11550commit cc3241a712684f8c7147f5688e9ee3ecb5a93b87 11551Author: Adam Jackson <ajax@redhat.com> 11552Date: Thu Oct 26 13:53:06 2017 -0400 11553 11554 animcur: Return the next interval directly from the timer callback 11555 11556 If the return value is non-zero here, DoTimer() will automatically rearm 11557 the timer for the new (relative) delay. 'soonest' is in absolute time, 11558 so subtract off 'now' and return that. 11559 11560 Reviewed-by: Robert Morell <rmorell@nvidia.com> 11561 Tested-by: Robert Morell <rmorell@nvidia.com> 11562 Signed-off-by: Adam Jackson <ajax@redhat.com> 11563 11564commit 3abbdb7318018584a27220737bd92081ce8ee67c 11565Author: Adam Jackson <ajax@redhat.com> 11566Date: Thu Oct 26 13:40:57 2017 -0400 11567 11568 animcur: Use fixed-size screen private 11569 11570 Reviewed-by: Robert Morell <rmorell@nvidia.com> 11571 Tested-by: Robert Morell <rmorell@nvidia.com> 11572 Signed-off-by: Adam Jackson <ajax@redhat.com> 11573 11574commit f615cb62d47cb24ea31718e8226df53ce6651c91 11575Author: Lukáš Krejčí <lskrejci@gmail.com> 11576Date: Sat Dec 30 23:46:45 2017 +0100 11577 11578 Xorg.wrap: Ensure correct ordering of post-install hook 11579 11580 The install rule of Xorg.wrap is currently a dependency of the 11581 install-data target instead of the install-exec target. The build also 11582 uses install-exec-hook to change the ownership and set the SUID bit on 11583 the Xorg.wrap binary. The problem is that install-exec-hook is only 11584 ordered respective to the install-exec target, the rules of install-data 11585 may or may not have been executed. 11586 11587 If install-exec-hook runs before the Xorg.wrap binary is in place, 11588 a message similar to the following will be present in the build log: 11589 11590 chown: cannot access '/pkgdir/usr/lib/xorg-server/Xorg.wrap': No such file or directory 11591 make[6]: [Makefile:1151: install-exec-hook] Error 1 (ignored) 11592 11593 All that needs to be done is to change the name of the program variable 11594 to contain 'exec' for the install rule to depend on the install-exec 11595 target. 11596 11597 Excerpt from the Automake manual, chapter 12.2 The Two Parts of Install: 11598 "Any variable using a user-defined directory prefix with ‘exec’ in the 11599 name (e.g., myexecbin_PROGRAMS) is installed by install-exec. All other 11600 user-defined prefixes are installed by install-data." 11601 11602 https://bugs.freedesktop.org/show_bug.cgi?id=104419 11603 11604 Signed-off-by: Lukáš Krejčí <lskrejci@gmail.com> 11605 Acked-by: Hans de Goede <hdegoede@redhat.com> 11606 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 11607 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 11608 11609commit 170c95978530f6373bdf4488116902b273f3abf4 11610Author: Olivier Fourdan <ofourdan@redhat.com> 11611Date: Fri Dec 15 16:43:47 2017 +0100 11612 11613 xwayland: avoid race condition on new keymap 11614 11615 When the Wayland compositor notifies of a new keymap, for the first X11 11616 client using the keyboard, the last slave keyboard used might still not 11617 be set (i.e. “lastSlave” is still NULL). 11618 11619 As a result, the new keymap is not applied, and the first X11 window 11620 will have the wrong keymap set initially. 11621 11622 Apply the new keymap to the master keyboard as long as there's one. 11623 11624 Bugzilla: https://bugzilla.gnome.org/show_bug.cgi?id=791383 11625 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 11626 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 11627 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 11628 11629commit c72049530503ebde493cfcd22156105557ea18d3 11630Author: Konstantin Kharlamov <Hi-Angel@yandex.ru> 11631Date: Sun Dec 17 23:23:02 2017 +0300 11632 11633 modesetting: simplify bailing on calloc fail 11634 11635 The "done" label restores crtc-> {x,y,rotation,mode}, frees output_id. 11636 Doing the calloc() before writing to those values frees us from 11637 necessity to restore them if calloc fails, and allows to merge 11638 "if (mode)" block. 11639 11640 Signed-off-by: Konstantin Kharlamov <Hi-Angel@yandex.ru> 11641 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 11642 11643commit c2954b16c8730c7ed8441fd8dba25900f3aed265 11644Author: Tapani Pälli <tapani.palli@intel.com> 11645Date: Tue Nov 28 09:23:29 2017 +0200 11646 11647 glx: do not pick sRGB config for 32-bit RGBA visual 11648 11649 This fixes blending issues seen with kwin and gnome-shell when 11650 32bit visual has sRGB capability set. 11651 11652 Reviewed-by: Adam Jackson <ajax@redhat.com> 11653 Signed-off-by: Tapani Pälli <tapani.palli@intel.com> 11654 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103699 11655 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103646 11656 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103655 11657 11658commit cf7517675d988c2d1ff967d6d162a17acbdad466 11659Author: Keith Packard <keithp@keithp.com> 11660Date: Wed Aug 2 21:34:52 2017 -0700 11661 11662 xfree86: Hold input_lock across SPRITE functions in VGA arbiter 11663 11664 Avoid scrambling the sprite functions wrapper. 11665 11666 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=101995 11667 Signed-off-by: Keith Packard <keithp@keithp.com> 11668 Reviewed-by: Adam Jackson <ajax@redhat.com> 11669 11670commit 3265d0c81f7a501258fa91e49fcc137714b4af5e 11671Author: Jon Turney <jon.turney@dronecode.org.uk> 11672Date: Tue Oct 10 13:43:40 2017 +0100 11673 11674 meson: Add dependency on generated code fragments in hw/xwin/glx/ 11675 11676 Somehow I'd managed to write this with this dependency missing, so this only 11677 works correctly when the generated files already exist and the correct 11678 automatic dependencies generated, but fails on a clean build. 11679 11680 Including generated files with a .c extension into the sources for a target 11681 causes meson to want to compile them (and it seems to be hard to say "make 11682 the directory containing this generated file available to include"). 11683 11684 So, change the extension of included generated C fragments to .ic 11685 11686 Update the autotools build to align. 11687 11688 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 11689 11690commit edda951fa5145a50915611ee0e9e459074117700 11691Author: Peter Harris <pharris@opentext.com> 11692Date: Thu Oct 12 16:48:37 2017 -0400 11693 11694 os: Add epoll-like port implementation for Solaris 11695 11696 x11perf -noop with 200 xlogos connected is slightly faster with ports: 11697 11698 before after Operation 11699 ---------- ----------------- -------------------- 11700 18400000.0 19200000.0 (1.04) X protocol NoOperation 11701 11702 Acked-by: Keith Packard <keithp@keithp.com> 11703 Signed-off-by: Peter Harris <pharris@opentext.com> 11704 11705commit 83c04ee6eae1edc80528203fb515425108171cd8 11706Author: Peter Harris <pharris@opentext.com> 11707Date: Thu Oct 12 16:48:36 2017 -0400 11708 11709 os: Add epoll-like pollset implementation for AIX 11710 11711 AIX's poll only allows FD_SETSIZE entries in the fd list, which is 11712 insufficient for expanded MaxClients. 11713 11714 As a bonus, x11perf -noop with ~250 xlogos connected is slightly faster 11715 with pollset: 11716 11717 before after Operation 11718 --------- ---------------- -------------------- 11719 5750000.0 5990000.0 (1.04) X protocol NoOperation 11720 11721 Signed-off-by: Peter Harris <pharris@opentext.com> 11722 Acked-by: Keith Packard <keithp@keithp.com> 11723 11724commit bed28300999a07514d741abe5c748adf234e18a6 11725Author: Adam Jackson <ajax@redhat.com> 11726Date: Mon Nov 20 15:43:07 2017 -0500 11727 11728 kdrive: remove KdSignalWrapper etc. 11729 11730 This no longer does anything useful. 11731 11732 Signed-off-by: Adam Jackson <ajax@redhat.com> 11733 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 11734 Reviewed-by: Dave Airlie <airlied@redhat.com> 11735 11736commit 4353d83f60766824a65f183716616eee8e17bb24 11737Author: Adam Jackson <ajax@redhat.com> 11738Date: Mon Nov 20 15:43:06 2017 -0500 11739 11740 xfree86: remove xf86CaughtSignal etc. 11741 11742 This no longer does anything useful. 11743 11744 Signed-off-by: Adam Jackson <ajax@redhat.com> 11745 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 11746 Reviewed-by: Dave Airlie <airlied@redhat.com> 11747 11748commit 9c72887939f319e185d2726d9d9a4191b9d12efd 11749Author: Adam Jackson <ajax@redhat.com> 11750Date: Mon Nov 20 15:43:05 2017 -0500 11751 11752 os: Make OsSignalHandler ask for core dumps for signo != SIGQUIT 11753 11754 SIGQUIT is a normal termination request, but any other signal we handle 11755 here wants a core. This has the effect of making FatalError's call to 11756 AbortServer trigger the 11757 11758 if (CoreDump) 11759 OsAbort(); 11760 11761 path. This will allow us to remove some DDX code that has the same net 11762 effect. 11763 11764 Signed-off-by: Adam Jackson <ajax@redhat.com> 11765 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 11766 Reviewed-by: Dave Airlie <airlied@redhat.com> 11767 11768commit 0a255dceb79ee28a88667d5bd23cf989dbf9bed8 11769Author: Adam Jackson <ajax@redhat.com> 11770Date: Mon Nov 20 15:43:04 2017 -0500 11771 11772 xfree86: Remove xf86InterceptSignals 11773 11774 The only consumer of this is the Linux vm86 backend for int10 (which you 11775 should not use), and there all it serves to do is make signals generated 11776 by the vm86 task non-fatal. In practice this error appears never to 11777 happen, and marching ahead with root privileges after arbitrary code has 11778 raised a signal seems like a poor plan. 11779 11780 Remove the usage in the vm86 code, making this error fatal. 11781 11782 Signed-off-by: Adam Jackson <ajax@redhat.com> 11783 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 11784 Reviewed-by: Dave Airlie <airlied@redhat.com> 11785 11786commit 722c8035dcf3ae0b18841066fe4ee030277274bc 11787Author: Adam Jackson <ajax@redhat.com> 11788Date: Mon Nov 20 15:43:03 2017 -0500 11789 11790 xfree86: Remove xf86InterceptSigIll 11791 11792 This was added in ~2004 for the sis driver, to detect whether it could 11793 use SSE for memcpy. Charmingly, the code to check whether that feature 11794 exists in the server is: 11795 11796 #if XORG_VERSION_CURRENT >= XORG_VERSION_NUMERIC(6,8,99,13,0) 11797 #define SISCHECKOSSSE /* Automatic check OS for SSE; requires SigIll facility */ 11798 #endif 11799 11800 Which means it has never worked in any modular server release. 11801 11802 A less gross way to do this is to check for SSE support with getauxval() 11803 or /proc/cpuinfo or similar. Since no driver is using the existing 11804 intercept mechanism, drop it. 11805 11806 Signed-off-by: Adam Jackson <ajax@redhat.com> 11807 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 11808 Reviewed-by: Dave Airlie <airlied@redhat.com> 11809 11810commit fe46cbea0f19959d469ca4d1f09be379dc7b1e45 11811Author: Olivier Fourdan <ofourdan@redhat.com> 11812Date: Tue Nov 21 14:45:13 2017 +0100 11813 11814 xwayland: Give up “cleanly“ on Wayland socket errors 11815 11816 Xwayland is a pretty standard Wayland client, we want to be able to 11817 capture core dumps on crashes. 11818 11819 Yet using "-core" causes any FatalError() to generate a core dump, 11820 meaning that we would get a core file for all Wayland server crashes, 11821 which would generate a lot of false positives. 11822 11823 Instead of using FatalError() on Wayland socket errors, give up cleanly 11824 to avoid dumping core files when "-core" is used. 11825 11826 See also: https://bugzilla.gnome.org/show_bug.cgi?id=790502 11827 and: https://bugzilla.gnome.org/show_bug.cgi?id=789086 11828 11829 Reviewed-by: Adam Jackson <ajax@redhat.com> 11830 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 11831 11832commit 6883ae43eb72fe4e2651c1dca209563323fad2db 11833Author: Tomasz Śniatowski <kailoran@gmail.com> 11834Date: Wed Dec 6 12:16:17 2017 +0100 11835 11836 os: Fix strtok/free crash in ComputeLocalClient 11837 11838 Don't reuse cmd for strtok output to ensure the proper pointer is 11839 freed afterwards. 11840 11841 The code incorrectly assumed the pointer returned by strtok(cmd, ":") 11842 would always point to cmd. However, strtok(str, sep) != str if str 11843 begins with sep. This caused an invalid-free crash when running 11844 a program under X with a name beginning with a colon. 11845 11846 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=104123 11847 Signed-off-by: Tomasz Śniatowski <kailoran@gmail.com> 11848 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 11849 11850commit aa6651f83c698e8cc40de61aad36c73ca7a6adcf 11851Author: Adam Jackson <ajax@redhat.com> 11852Date: Thu Jun 8 16:23:12 2017 -0400 11853 11854 xfixes: Remove the CursorCurrent array 11855 11856 We're not wrapping all the ways a cursor can be destroyed, so this array 11857 ends up with stale data. Rather than try harder to wrap more code paths, 11858 just look up the cursor when we need it. 11859 11860 Signed-off-by: Adam Jackson <ajax@redhat.com> 11861 11862commit 3db7707df3cfd8ba62c10d08c7c731ec18ea8ddf 11863Author: Adam Jackson <ajax@redhat.com> 11864Date: Wed Dec 6 15:12:20 2017 -0500 11865 11866 test: Fix build dependency for bigreq test 11867 11868 libxcb-xinput isn't a thing in whichever Ubuntu it is that Travis is 11869 using. The test is already optional, make it more so. 11870 11871 Signed-off-by: Adam Jackson <ajax@redhat.com> 11872 11873commit 9f7a9be13d6449c00c86d3035374f4f543654b3f 11874Author: Olivier Fourdan <ofourdan@redhat.com> 11875Date: Tue Dec 5 09:59:06 2017 +0100 11876 11877 dix: avoid deferencing NULL PtrCtrl 11878 11879 PtrCtrl really makes sense for relative pointing device only, absolute 11880 devices such as touch devices do not have any PtrCtrl set. 11881 11882 In some cases, if the client issues a XGetPointerControl() immediatlely 11883 after a ChangeMasterDeviceClasses() copied the touch device to the VCP, 11884 a NULL pointer dereference will occur leading to a crash of Xwayland. 11885 11886 Check whether the PtrCtrl is not NULL in ProcGetPointerControl() and 11887 return the default control values otherwise, to avoid the NULL pointer 11888 dereference. 11889 11890 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1519533 11891 Reviewed-by: Adam Jackson <ajax@redhat.com> 11892 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 11893 11894commit 60f4646ae10f0b57790fce46682baa531512b53e 11895Author: Carlos Garnacho <carlosg@gnome.org> 11896Date: Mon Dec 4 16:55:13 2017 +0100 11897 11898 xwayland: Keep separate variables for pointer and tablet foci 11899 11900 The tablet/stylus interfaces reused xwl_seat->focus_window, which 11901 would leave a somewhat inconsistent state of that variable for 11902 wl_pointer purposes (basically, everything) if the pointer happened 11903 to lay on the same surface than the stylus while proximity_out 11904 happens. 11905 11906 We just want the stylus xwl_window to correctly determine we have 11907 stylus focus, and to correctly translate surface-local coordinates 11908 to root coordinates, this can be done using a different variable. 11909 11910 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 11911 Acked-by: Jason Gerecke <jason.gerecke@wacom.com> 11912 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 11913 11914commit 97ac59b1ed3624f7c04e54dd3e3dadfa46a8f170 11915Author: Olivier Fourdan <ofourdan@redhat.com> 11916Date: Wed Sep 27 18:01:01 2017 +0200 11917 11918 xwayland: Fix non-argb cursor conversion 11919 11920 From the bug: "What happens if bits->width is less than 8? :)" 11921 11922 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=103012 11923 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 11924 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 11925 Reviewed-by: Daniel Stone <daniels@collabora.com> 11926 11927commit a4c1e290824d0e40191ad26019b705f61c11e113 11928Author: Adam Jackson <ajax@redhat.com> 11929Date: Tue Nov 14 15:15:03 2017 -0500 11930 11931 glx: Implement GLX_EXT_no_config_context (v2) 11932 11933 Only enabled for the DRI backends at the moment. In principle WGL/CGL 11934 could support this - it's sort of implied by GL 3.0 support - but in 11935 practice their implementations back GLX drawables with native drawables 11936 (and not anonymous FBOs), so they would need either a corresponding 11937 window system binding extension or significant implementation work. 11938 11939 v2: Require that the two screen numbers match, per v4 of spec. 11940 11941 Khronos: https://github.com/KhronosGroup/OpenGL-Registry/pull/102 11942 Signed-off-by: Adam Jackson <ajax@redhat.com> 11943 Reviewed-by: Eric Anholt <eric@anholt.net> 11944 11945commit f0fffa926a5771e0e604fe9a48178b0514ca5d41 11946Author: Adam Jackson <ajax@redhat.com> 11947Date: Tue Nov 14 15:15:02 2017 -0500 11948 11949 glx: Prepare __glXGetDrawable for no-config contexts 11950 11951 Any proper (GLX 1.3) drawable will already have a bound config, but if 11952 we're doing the GLX 1.2 thing of making a Window current, we need to 11953 infer the config from the window's Visual. 11954 11955 Signed-off-by: Adam Jackson <ajax@redhat.com> 11956 Reviewed-by: Eric Anholt <eric@anholt.net> 11957 11958commit 5d667df6ea1634191a26f9a7c26bc883701d62b0 11959Author: Adam Jackson <ajax@redhat.com> 11960Date: Tue Nov 14 15:15:01 2017 -0500 11961 11962 glx: Fix glXQueryContext for GLX_FBCONFIG_ID and GLX_RENDER_TYPE (v2) 11963 11964 Just never filled in, oops. Seems to have gone unnoticed because 11965 normally glXQueryContext simply returns the values filled in by the 11966 client library when the context was created. The only path by which you 11967 normally get to a GLXQueryContext request is glXImportContext, and then 11968 only if the context is already indirect. 11969 11970 However, that's a statement about Mesa's libGL (and anything else that 11971 inherited that bit of the SGI SI more or less intact). Nothing prevents 11972 a mischeivous client from issuing that request of a direct context, and 11973 if they did we'd be in trouble because we never bothered to preserve the 11974 associated fbconfig in the context state, so we'd crash looking up 11975 GLX_VISUAL_ID_EXT. So let's fix that too. 11976 11977 v2: Fixed missing preservation of the config in DRI2 (Eric Anholt) 11978 11979 Signed-off-by: Adam Jackson <ajax@redhat.com> 11980 Reviewed-by: Eric Anholt <eric@anholt.net> 11981 11982commit 78b2ce1410dda758537d52398223ca9b757883bc 11983Author: Daniel Martin <consume.noise@gmail.com> 11984Date: Mon Nov 20 10:47:43 2017 +0100 11985 11986 modesetting: Remove #if 0 CMAP_LOAD_EVEN_IF_OFFSCREEN 11987 11988 CMAP_LOAD_EVEN_IF_OFFSCREEN has been encapsulated since the import of 11989 xf86-video-modesetting into the tree. 11990 11991 Reviewed-by: Adam Jackson <ajax@redhat.com> 11992 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 11993 11994commit d563443381142fbc0b102422d7e7389bf3621331 11995Author: Daniel Martin <consume.noise@gmail.com> 11996Date: Mon Nov 20 10:47:42 2017 +0100 11997 11998 modesetting: Remove unused define DRV_ERROR 11999 12000 Reviewed-by: Adam Jackson <ajax@redhat.com> 12001 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12002 12003commit 6804875662363764683a86c1614e4cf3cc70a20a 12004Author: Daniel Martin <consume.noise@gmail.com> 12005Date: Mon Nov 20 10:47:41 2017 +0100 12006 12007 modesetting: Reset output_id if drmModeGetConnector failed 12008 12009 If drmModeGetConnector() fails in drmmode_output_detect(), we have to 12010 reset the output_id to -1 too. 12011 12012 Yet another spot leading to a potential NULL dereference when handling 12013 the mode_output member as output_id was != -1. Though, this case should 12014 be very hard to hit. 12015 12016 Reviewed-by: Adam Jackson <ajax@redhat.com> 12017 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12018 12019commit e20b1016de22e893d5468960988ddd5677c11d98 12020Author: Daniel Martin <consume.noise@gmail.com> 12021Date: Mon Nov 20 10:47:40 2017 +0100 12022 12023 modesetting: Fix log msg if pixmap creation failed 12024 12025 Add a missing new-line character and make the message more verbose than 12026 "Failed". 12027 12028 Reviewed-by: Adam Jackson <ajax@redhat.com> 12029 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12030 12031commit 02981fe1a844e101c81037df79fc147f217f3bf9 12032Author: Daniel Martin <consume.noise@gmail.com> 12033Date: Mon Nov 20 10:47:39 2017 +0100 12034 12035 xfree86: Fix set but not used warnings in lnx_platform 12036 12037 Those warnings are generated, when building without systemd support: 12038 12039 ../hw/xfree86/os-support/linux/lnx_platform.c: In function ‘get_drm_info’: 12040 ../hw/xfree86/os-support/linux/lnx_platform.c:29:16: warning: variable ‘minor’ set but not used [-Wunused-but-set-variable] 12041 int major, minor, fd; 12042 ^~~~~ 12043 ../hw/xfree86/os-support/linux/lnx_platform.c:29:9: warning: variable ‘major’ set but not used [-Wunused-but-set-variable] 12044 int major, minor, fd; 12045 ^~~~~ 12046 12047 In this case the functions are macros, which don't use theese arguments. 12048 12049 v2: Add comments, why the warnings appear. Suggested by Emil Velikov 12050 12051 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 12052 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12053 12054commit 918afeecbc63d70413e222efdb2ac4cfb16eae9e 12055Author: Daniel Martin <consume.noise@gmail.com> 12056Date: Mon Nov 20 10:47:38 2017 +0100 12057 12058 os/xdmcp: Honour -once when session is dead 12059 12060 Terminate a dead session when -once was passed. Don't restart it. 12061 12062 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12063 Reviewed-by: Walter Harms <wharms@bfs.de> 12064 12065commit 559954aaa8d811a22cf918cc16a7d618e12201a0 12066Author: Michel Dänzer <michel.daenzer@amd.com> 12067Date: Mon Oct 2 11:33:43 2017 +0200 12068 12069 present: Only send PresentCompleteNotify events to the presenting client 12070 12071 We were sending the events to all clients listening for them on the 12072 window. But clients can get confused by events from another client, and 12073 I can't imagine any case where receiving events from other clients would 12074 be required. 12075 12076 v2: 12077 * Also restrict events sent to additional windows to the presenting 12078 client 12079 * Don't shorten line lengths 12080 12081 Reviewed-by: Keith Packard <keithp@keithp.com> 12082 12083commit fc7fb5bbe1c8f787e53500b9a2ca4af815f310d1 12084Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12085Date: Thu Nov 9 10:21:21 2017 +0100 12086 12087 randr: free crtc->outputs on destroy 12088 12089 Reviewed-by: Adam Jackson <ajax@redhat.com> 12090 Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12091 12092commit 16381d186e7c791031392ed8afcfd33009854e9e 12093Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12094Date: Thu Nov 9 10:21:20 2017 +0100 12095 12096 randr: always realloc crtcs and outputs 12097 12098 When the last crtc (resp. output) is destroyed, the rrScrPriv crtcs 12099 (resp. outputs) fields do not get cleared, which can lead to a situation 12100 where the private's numCrtcs (resp. numOutputs) field is zero, but the 12101 associated memory is still allocated. Just checking if numCrtcs (resp. 12102 numOutputs) is zero is thus not a good criteria to determine whetehr to 12103 use a realloc or a malloc. 12104 12105 Since crtcs (resp. outputs) are NULL-initialized anyway, relying on 12106 numCrtcs (resp. numOutputs) is actually unnecessary, because 12107 reallocation of a NULL ptr is equivalent to a malloc anyway. 12108 12109 Therefore, just use realloc() unconditionally, and ensure that the 12110 fields are properly initialized. 12111 12112 Reviewed-by: Adam Jackson <ajax@redhat.com> 12113 Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12114 12115commit fb5ee77b91a93e27801006be8ee34d27984e7fa6 12116Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12117Date: Thu Nov 9 10:21:19 2017 +0100 12118 12119 randr: rrGetScreenResources: initialize memory 12120 12121 Similarly to bb766ef11227bd8c71ac65845d1930edd0eda40d, ensure that the 12122 extra padding is set to 0. 12123 12124 Reviewed-by: Adam Jackson <ajax@redhat.com> 12125 Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12126 12127commit 307c124d6bcfe26057767b2c0990dc9ac66b9c93 12128Author: Adam Jackson <ajax@redhat.com> 12129Date: Tue Nov 14 15:59:35 2017 -0500 12130 12131 glx: Only flush indirect contexts in MakeCurrent (v2) 12132 12133 If the context is direct none of the GL commands were issued by this 12134 process, the server couldn't flush them even if it wanted to. 12135 12136 v2: Fix embarassingly obvious boolean inversion (Michel Dänzer) 12137 12138 Signed-off-by: Adam Jackson <ajax@redhat.com> 12139 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 12140 12141commit 66e8eaa456a0f200c0649c56a2e34914daa88065 12142Author: Adam Jackson <ajax@redhat.com> 12143Date: Thu Nov 2 16:10:50 2017 -0400 12144 12145 glamor: Drop the non-VAO rendering path 12146 12147 GLES spells this extension as GL_OES_vertex_array_object, but it is 12148 functionally equivalent to the GL_ARB version. Mesa has supported both 12149 since 9.0, let's go ahead and require it. 12150 12151 Signed-off-by: Adam Jackson <ajax@redhat.com> 12152 Reviewed-by: Eric Anholt <eric@anholt.net> 12153 12154commit dbda03d5d15bec8fdff64bc8545875c39da81777 12155Author: Adam Jackson <ajax@redhat.com> 12156Date: Mon Nov 6 16:41:14 2017 -0500 12157 12158 glx: Drop references to <GL/internal/glcore.h> 12159 12160 This header is a truly ancient remnant of the SGI SI code, which used 12161 the same struct for visual/fbconfig state on both sides of the wire. 12162 We're not using this struct ourselves so let's stop depending on the 12163 header. 12164 12165 Signed-off-by: Adam Jackson <ajax@redhat.com> 12166 12167commit eafcd009f1fe3e554219a428237ef1c71d1ba506 12168Author: Adam Jackson <ajax@redhat.com> 12169Date: Tue Oct 24 14:33:19 2017 -0400 12170 12171 xres: Return the atom naming the type, not its internal type number 12172 12173 Signed-off-by: Adam Jackson <ajax@redhat.com> 12174 12175commit 68556d74b49e99d3490166c446079f7d5de26ca4 12176Author: Hector Martin <marcan@marcan.st> 12177Date: Wed Nov 15 03:12:31 2017 +0900 12178 12179 edid: fix off-by-one error in CEA mode numbering 12180 12181 The CEA extension short video descriptors contain the VIC, which starts 12182 at 1, not 0. 12183 12184 Reviewed-by: Adam Jackson <ajax@redhat.com> 12185 Signed-off-by: Hector Martin <marcan@marcan.st> 12186 12187commit 9bd5a198dc5383d0d2a1e28f7aa4270132eca5db 12188Author: Adam Jackson <ajax@redhat.com> 12189Date: Tue Jul 11 15:37:07 2017 -0400 12190 12191 glamor: Remove unused glamor_create_screen_resources 12192 12193 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 12194 Signed-off-by: Adam Jackson <ajax@redhat.com> 12195 12196commit 081675f7ffda0cce20b395874d6e5755fb6600a0 12197Author: Adam Jackson <ajax@redhat.com> 12198Date: Fri Jul 14 14:52:01 2017 -0400 12199 12200 glamor: Un-inline glamor_[gs]et_screen_private 12201 12202 text data bss dec hex filename 12203 2134764 45210 128704 2308678 233a46 build/hw/kdrive/ephyr/Xephyr.before 12204 2129972 45210 128704 2303886 23278e build/hw/kdrive/ephyr/Xephyr.after 12205 12206 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 12207 Signed-off-by: Adam Jackson <ajax@redhat.com> 12208 12209commit 6dff5e5d49a21bc749e44f0e8cb390690ece750d 12210Author: Adam Jackson <ajax@redhat.com> 12211Date: Mon Jul 10 10:56:58 2017 -0400 12212 12213 xwayland: Stop printing the EGL version 12214 12215 It doesn't matter, none of this matters. 12216 12217 Acked-by: Olivier Fourdan <ofourdan@redhat.com> 12218 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 12219 Signed-off-by: Adam Jackson <ajax@redhat.com> 12220 12221commit c57f63829f0077b5a3e5968f8405e087a4490485 12222Author: Adam Jackson <ajax@redhat.com> 12223Date: Wed Sep 13 15:44:15 2017 -0400 12224 12225 dmx: Use noGlxExtension like other DDXes 12226 12227 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 12228 Signed-off-by: Adam Jackson <ajax@redhat.com> 12229 12230commit fd0eafb18426da14601d5c0d0a50092c49a7aff8 12231Author: Adam Jackson <ajax@redhat.com> 12232Date: Mon Nov 6 16:07:41 2017 -0500 12233 12234 glx: Fix typos that break GLX_ARB_context_flush_control 12235 12236 The trailing \n are just wrong here, __glXEnableExtension wants a string 12237 without them. 12238 12239 Signed-off-by: Adam Jackson <ajax@redhat.com> 12240 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 12241 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 12242 12243commit 2e7f790b5770589ae2f97ee363ecd33fdb362c24 12244Author: Adam Jackson <ajax@redhat.com> 12245Date: Mon Nov 6 15:25:34 2017 -0500 12246 12247 dix: Remove ffs.c 12248 12249 Your libc has ffs, I promise. 12250 12251 Signed-off-by: Adam Jackson <ajax@redhat.com> 12252 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 12253 12254commit 4d15a2645f00995618d8687edf1fb0f4b2316ace 12255Author: Adam Jackson <ajax@redhat.com> 12256Date: Mon Nov 6 15:25:33 2017 -0500 12257 12258 os: Remove mffs() 12259 12260 This was always wide enough to work on an fd_mask ("mask" ffs 12261 presumably). We don't operate on fd_masks anymore, so this can go. 12262 12263 Signed-off-by: Adam Jackson <ajax@redhat.com> 12264 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 12265 12266commit 4b0a3cbab131eb453e2b3fc0337121969258a7be 12267Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12268Date: Sat Nov 4 23:06:31 2017 +0100 12269 12270 glx: free fbconfigs on destroy 12271 12272 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 12273 Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12274 12275commit bb766ef11227bd8c71ac65845d1930edd0eda40d 12276Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12277Date: Sat Nov 4 23:06:29 2017 +0100 12278 12279 randr: ProcRRGetOutputInfo: initialize memory 12280 12281 Running Xephyr under valgrind reveals that we're sending some 12282 uninitialized memory over the wire (particularly, the leftover padding 12283 that comes from rounding extraLen to the next 32-bit multiple). 12284 12285 Solve by calloc()ing the memory instead of malloc()ing (the alternative 12286 would be to memset just the padding, but I'm not sure it's more 12287 convenient.) 12288 12289 Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12290 Reviewed-by: Adam Jackson <ajax@redhat.com> 12291 12292commit 6828645916505a5925db5c2c2e816fee4e1050e5 12293Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12294Date: Sat Nov 4 23:06:28 2017 +0100 12295 12296 Xephyr: free driverPrivates on Fini 12297 12298 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 12299 12300commit b2167015043a458e9cf93b827b43eb5b7c552ce9 12301Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12302Date: Sat Nov 4 23:06:27 2017 +0100 12303 12304 xkb: initialize tsyms 12305 12306 This fixes some “Conditional jump depends on uninitialized value(s)” 12307 errors spotted by valgrind. 12308 12309 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 12310 Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12311 12312commit 2dafa1bdafe583587431a5f6ebee6ec371c61a8d 12313Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 12314Date: Thu Nov 2 21:27:59 2017 +0100 12315 12316 dix/window: fix typos 12317 12318 Reviewed-by: Adam Jackson <ajax@redhat.com> 12319 12320commit f84e59a4f474d22860bac8aec2947798a86db69b 12321Author: Thomas Hellstrom <thellstrom@vmware.com> 12322Date: Tue Sep 26 17:28:12 2017 -0700 12323 12324 glx: Duplicate relevant fbconfigs for compositing visuals 12325 12326 Previously, before GLX_OML_swap_method was fixed, both the X server and 12327 client ignored the swapMethod fbconfig value, which meant that, if the dri 12328 driver thought it exposed more than one swapMethod, it actually just 12329 exported a duplicated set of fbconfigs. When fixing GLX_OML_swap_method 12330 and restricting the choice for built-in visuals to a single swap method 12331 that meant we didn't have that many fbconfigs to choose from when pairing 12332 the compositing visual with an fbconfig, resulting in the fbconfig paired 12333 with the compositing visual becoming too restrictive for some applications, 12334 (at least for kwin). This problem would also happen if the dri driver 12335 only exposed a single swap method to begin with. 12336 12337 So, to make sure the compositing visual gets a good enough fbconfig, 12338 duplicate fbconfigs that are suitable for compositing visuals and make 12339 sure these duplicated fbconfigs can be used only by compositing visuals. 12340 For duplicated fbconfigs not paired with a compositing visual, construct 12341 new compositing visuals, making compositing clients able to choose visuals 12342 / fbconfig more adapted to their needs. 12343 12344 This is in some sense equivalent to adding a new "TRUECOLOR_COMPOSITING" 12345 GLX visualtype. 12346 12347 Fixes: 4486d199bd3b ("glx: Fix visual fbconfig matching with respect to 12348 swap method") 12349 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102806 12350 Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> 12351 Tested-By: Nick Sarnie <commendsarnex@gmail.com> 12352 Tested-by: Fredrik Höglund <fredrik@kde.org> 12353 Reviewed-by: Adam Jackson <ajax@redhat.com> 12354 12355commit 30f4d440ebc3517fdcc1d3c6a422a8fbf3af1f23 12356Author: Eric Anholt <eric@anholt.net> 12357Date: Tue Oct 31 12:22:31 2017 -0700 12358 12359 xkb: Print the xkbcomp path being executed when we fail to compile. 12360 12361 I don't know how many times I've had a broken server due to a bad 12362 directory to xkbcomp, and only finding the whole path has shown me 12363 where I went wrong. 12364 12365 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 12366 12367commit 14af8bee242fe40af0e91c61465d6720aaa60e97 12368Author: Eric Anholt <eric@anholt.net> 12369Date: Mon Oct 9 17:14:32 2017 -0700 12370 12371 test: Add a test for the overflow bug in bigreqs. 12372 12373 The failing struct comes from the python test written by Michal Srb 12374 <msrb@suse.com>. 12375 12376 v2: Use a drawable (root window) and gc, so that PolyLines hopefully 12377 actually tries processing things. However, the request seems to 12378 process successfully so the poll() just stalls out. However, this 12379 does let us distinguish between detecting the bigrequests error 12380 and not, at least. 12381 v3: Clean up the description of what we expect the poll() call to do. 12382 v4: Use XI2 instead of PolyLine to trigger a predictable error. We know the 12383 server replies with BadValue for a zero num_masks argument. So if we send 12384 a bigreq with a num_masks 0 and a length 0, we can just check whether we 12385 get killed (good) or a BadValue (bad). It doesn't test for specific memory 12386 overflows or crashes, but based on the assumption that we shouldn't look 12387 at *any* BigReq of size 0, this seems to be sufficient. 12388 12389 Signed-off-by: Eric Anholt <eric@anholt.net> 12390 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 12391 12392commit 5aad81445c8c3d6b7b30d503cfe26027fa482870 12393Author: Peter Hutterer <peter.hutterer@who-t.net> 12394Date: Tue Sep 26 15:21:59 2017 +1000 12395 12396 config/udev: consider ID_INPUT_FOO=0 as 'unset' 12397 12398 Historically we didn't need to care about this case but more devices are 12399 having invalid types set and they cannot be unset with a hwdb entry (which 12400 doesn't handle the empty string). Allow for "0" to mean "unset" because 12401 anything else would be crazy anyway. 12402 12403 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 12404 12405commit d5379b350fb63e42e604361c21ad9832b4c791b9 12406Author: Daniel Martin <consume.noise@gmail.com> 12407Date: Fri Oct 27 16:11:56 2017 +0200 12408 12409 Use ARRAY_SIZE all over the tree 12410 12411 Roundhouse kick replacing the various (sizeof(foo)/sizeof(foo[0])) with 12412 the ARRAY_SIZE macro from dix.h when possible. A semantic patch for 12413 coccinelle has been used first. Additionally, a few macros have been 12414 inlined as they had only one or two users. 12415 12416 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12417 Reviewed-by: Adam Jackson <ajax@redhat.com> 12418 12419commit 15a32ee5d1fffa171bb05af9a0e5b472e4af1488 12420Author: Daniel Martin <consume.noise@gmail.com> 12421Date: Fri Oct 27 16:11:55 2017 +0200 12422 12423 test: signal-logging: Fix looping signed number tests 12424 12425 unsigned_tests[] was used to compute the amount of signed numbers to 12426 test. 12427 12428 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12429 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 12430 12431commit 0bcc65f2bf479c6a74ac70bb5b5181d6834dded6 12432Author: Daniel Martin <consume.noise@gmail.com> 12433Date: Fri Oct 27 16:11:54 2017 +0200 12434 12435 test: input: Fix used uninitialized warning in dix_event_to_core 12436 12437 input.c: In function ‘dix_event_to_core’: 12438 ../include/inputstr.h:61:55: warning: ‘*((void *)&ev+80)’ is used uninitialized in this function [-Wuninitialized] 12439 #define SetBit(ptr, bit) (((BYTE *) (ptr))[(bit)>>3] |= (1 << ((bit) & 7))) 12440 ^~ 12441 12442 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12443 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 12444 12445commit 04a305121fbc08ecc2ef345ee7155d6087a43fd1 12446Author: Daniel Martin <consume.noise@gmail.com> 12447Date: Fri Oct 27 16:11:53 2017 +0200 12448 12449 modesetting: Fix potential buffer overflow 12450 12451 If one misconfigures a ZaphodHeads value (more than 20 characters 12452 without a delimiter), we get an overflow of our buffer. Use 12453 xstrtokenize() instead of writing/fixing our own tokenizer. 12454 12455 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12456 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 12457 12458commit 2230e6c8af92b041821eee0ea6210eda82c74106 12459Author: Adam Jackson <ajax@redhat.com> 12460Date: Tue Oct 24 14:28:34 2017 -0400 12461 12462 glamor: Unconditionalize GLAMOR_TEXTURED_LARGE_PIXMAP 12463 12464 Signed-off-by: Adam Jackson <ajax@redhat.com> 12465 Reviewed-by: Eric Anholt <eric@anholt.net> 12466 12467commit 1edac5c1ceb8b21af45b20397ff64177ad22466c 12468Author: Adam Jackson <ajax@redhat.com> 12469Date: Tue Oct 24 14:28:33 2017 -0400 12470 12471 glamor: Unconditionalize GLAMOR_PIXMAP_DYNAMIC_UPLOAD 12472 12473 Signed-off-by: Adam Jackson <ajax@redhat.com> 12474 Reviewed-by: Eric Anholt <eric@anholt.net> 12475 12476commit 13409b91b1936330dbf19bacefa4b48d0b6843b8 12477Author: Adam Jackson <ajax@redhat.com> 12478Date: Tue Oct 24 14:28:32 2017 -0400 12479 12480 glamor: Unconditionalize GLAMOR_GRADIENT_SHADER 12481 12482 Effectively always true anyway. 12483 12484 Signed-off-by: Adam Jackson <ajax@redhat.com> 12485 Reviewed-by: Eric Anholt <eric@anholt.net> 12486 12487commit 266d9868ca1cf77b7d315d607b515f081a9f45c3 12488Author: Alex Goins <agoins@nvidia.com> 12489Date: Thu Oct 19 20:02:30 2017 -0700 12490 12491 xf86-video-modesetting: Fix ms_queue_vblank(flags = MS_QUEUE_RELATIVE) 12492 12493 Change 677c32bc refactored all usages of drmWaitVBlank() into a helper function, 12494 ms_queue_vblank(). 12495 12496 ms_queue_vblank() takes in an MS_QUEUE_RELATIVE flag to indicate that the 12497 sequence number is relative rather than absolute, but still treats the actual 12498 sequence number as absolute, passing it through ms_crtc_msc_to_kernel_msc() 12499 unconditionally before calling drmWaitVBlank(). 12500 12501 ms_crtc_msc_to_kernel_msc() works by subtracting a vblank offset from the 12502 provided sequence number, which only makes sense for absolute sequence numbers. 12503 In the case of PRIME Sync, drmmode_SharedPixmapPrsentOnVBlank() passes in 1, 12504 which results in a large negative vblank offset. After subtracting, we're left 12505 with a relative sequence number of 100,000+, i.e. wait for 100,000+ vblanks... 12506 12507 In the relative case we want to pass in the sequence number unmodified. Simply 12508 add a check to do this. 12509 12510 Signed-off-by: Alex Goins <agoins@nvidia.com> 12511 Reviewed-by: Keith Packard <keithp@keithp.com> 12512 12513commit 68d95e759f8b6ebca6bd52e69e6bc34cc174f8ca 12514Author: Alex Goins <agoins@nvidia.com> 12515Date: Tue Oct 24 18:39:13 2017 -0700 12516 12517 ramdac: Check ScreenPriv != NULL in xf86ScreenSetCursor() 12518 12519 Similar to change cba5a10f, xf86ScreenSetCursor() would dereference ScreenPriv 12520 without NULL checking it. If Option "SWCursor" is specified, ScreenPriv == NULL. 12521 12522 Without this fix, it is observed that setting Option "SWCursor" "on" on the 12523 modesetting driver in a PRIME configuration will segfault the server. 12524 12525 It is important to return success rather than failure in the instance that 12526 ScreenPriv == NULL and pCurs == NullCursor, because otherwise xf86SetCursor() 12527 can fall into infinite recursion: xf86SetCursor(pCurs) calls 12528 xf86ScreenSetCursor(pCurs), and if FALSE, calls xf86SetCursor(NullCursor). If 12529 xf86ScreenSetCursor(NullCursor) returns FALSE, it calls 12530 xf86SetCursor(NullCursor) again and this repeats forever. 12531 12532 Signed-off-by: Alex Goins <agoins@nvidia.com> 12533 Reviewed-by: Dave Airlie <airlied@redhat.com> 12534 12535commit 04163fe8c66ae6683a1384fd1bd91ea85f9892ee 12536Author: Adam Jackson <ajax@redhat.com> 12537Date: Tue Oct 24 13:03:09 2017 -0400 12538 12539 meson: Use [ true, false, auto ] for tristate values 12540 12541 For symmetry with the boolean options. I really do not want to care 12542 whether an option is a tristate if I'm trying to set it explicitly. 12543 12544 Signed-off-by: Adam Jackson <ajax@redhat.com> 12545 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 12546 12547commit c2c6e9e68a8815420233c996acdd29ba572b1f0e 12548Author: Adam Jackson <ajax@redhat.com> 12549Date: Tue Oct 24 14:38:13 2017 -0400 12550 12551 dix: Don't track the XKB client versions in the ClientRec 12552 12553 XKB stores some stuff in the ClientRec that, style-wise, should probably 12554 be in a client private. vMinor tracks the client's idea of the XKB 12555 minor version, but is never read, we can just nuke it. vMajor is only 12556 used for a bug-compat workaround for X11R6.0-vintage clients. We're 12557 only using though (1<<4) for xkbClientFlags in the protocol, so we can 12558 pack that field down to a u8 and store the bug-compat flag there. 12559 12560 Signed-off-by: Adam Jackson <ajax@redhat.com> 12561 Reviewed-by: Julien Cristau <jcristau@debian.org> 12562 12563commit bc5fb8c0928498c32dc33680d40d50e6db5879b6 12564Author: Adam Jackson <ajax@redhat.com> 12565Date: Tue Oct 24 14:38:12 2017 -0400 12566 12567 dix: Don't vary the ClientRec ABI at build time 12568 12569 Just no. 12570 12571 Signed-off-by: Adam Jackson <ajax@redhat.com> 12572 Reviewed-by: Julien Cristau <jcristau@debian.org> 12573 12574commit 0c3ad9d1e06294b4f66717798afa4f5dfa505469 12575Author: Adam Jackson <ajax@redhat.com> 12576Date: Mon Oct 23 15:34:26 2017 -0400 12577 12578 travis: Bump linux Dockerfile to v7 for new meson 12579 12580 Signed-off-by: Adam Jackson <ajax@redhat.com> 12581 12582commit 4d53e30651c0fe5f7be38ae8529fa49846d39549 12583Author: Lyude Paul <lyude@redhat.com> 12584Date: Mon Oct 23 16:21:19 2017 -0400 12585 12586 meson: Don't use '' in link_with, ever 12587 12588 String arguments as elements in the array passed to the link_with 12589 argument in meson's executable() functions are not valid and will end up 12590 causing the build file generation to file. This actually ended up 12591 exposing a bug in meson that caused it not to report where in the 12592 meson.build file it was failing: 12593 12594 https://github.com/mesonbuild/meson/pull/2527 12595 12596 The proper way to have a variable that can contain either an empty link 12597 target or an actual link target is: 12598 12599 some_target = [] 12600 if some_cond 12601 some_target = static_library(...) 12602 endif 12603 12604 This way if some_cond is False, some_target gets set to [], gets passed 12605 to executable() in the link_with array, and then gets removed by array 12606 flattening. 12607 12608 This also unbreaks Xwayland builds with -Dglx=false, the thing that 12609 originally made me notice this. 12610 12611 Signed-off-by: Lyude Paul <lyude@redhat.com> 12612 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 12613 12614commit 5893e72a20ffa5cfcd6d0c7c5c934b92546d76c3 12615Author: Jon Turney <jon.turney@dronecode.org.uk> 12616Date: Tue Oct 24 12:07:08 2017 +0100 12617 12618 travis: Fix OSX build 12619 12620 Turn off homebrew autoupdating before installing ccache, to avoid: 12621 12622 /usr/local/Homebrew/Library/Homebrew/brew.rb:12:in `<main>': Homebrew must 12623 be run under Ruby 2.3! You're running 2.0.0. (RuntimeError) 12624 12625 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 12626 Reviewed-by: Adam Jackson <ajax@redhat.com> 12627 12628commit f44935cdb7321af242ce9f242975f096807b97f7 12629Author: Daniel Martin <consume.noise@gmail.com> 12630Date: Mon Oct 23 10:31:21 2017 +0200 12631 12632 modesetting: Use helper to fetch drmModeProperty(Blob)s 12633 12634 Replace the various loops to lookup drmModeProperty(Blob)s by 12635 introducing helper functions. 12636 12637 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12638 12639commit 6abdb54a11dac4e8854ff94ecdcb90a14321ab31 12640Author: Daniel Martin <consume.noise@gmail.com> 12641Date: Mon Oct 23 10:31:20 2017 +0200 12642 12643 modesetting: Fix leak of tile_blob in drmmode_output_destroy 12644 12645 And drmModeFreePropertyBlob() can handle NULL pointers, no need to check 12646 edid_blob. 12647 12648 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12649 12650commit 8d7f7e24261e68459e6f0a865e243473f65fe7ad 12651Author: Daniel Martin <consume.noise@gmail.com> 12652Date: Fri Oct 20 10:05:35 2017 +0200 12653 12654 modesetting: Check crtc before searching link-status property 12655 12656 No need to lookup the link-status property if we don't have a crtc. 12657 12658 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12659 Reviewed-by: Adam Jackson <ajax@redhat.com> 12660 12661commit 8c455db0ebb6e5313ca81428bb6dd75ef12aaa15 12662Author: Daniel Martin <consume.noise@gmail.com> 12663Date: Fri Oct 20 10:05:34 2017 +0200 12664 12665 modesetting: Remove #ifdefs XF86_PDEV_SERVER_FD 12666 12667 XF86_PDEV_SERVER_FD is defined since: 12668 12669 commit 5fb641a29bfb4a33da964e1e9af523f3472015c6 12670 Author: Hans de Goede <hdegoede@redhat.com> 12671 Date: Mon Jan 13 12:03:46 2014 +0100 12672 12673 hotplug: Extend OdevAttributes for server-managed fd support 12674 12675 ifdef'ing for it is a leftover from the external xf86-video-modesetting. 12676 12677 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12678 Reviewed-by: Adam Jackson <ajax@redhat.com> 12679 12680commit 66d8cbf8ce9285a8771118e46daa44faa73ad847 12681Author: Daniel Martin <consume.noise@gmail.com> 12682Date: Fri Oct 20 10:05:33 2017 +0200 12683 12684 modesetting: Fix warning of unused variable if not GLAMOR_HAS_GBM 12685 12686 ../hw/xfree86/drivers/modesetting/driver.c: In function ‘redisplay_dirty’: 12687 ../hw/xfree86/drivers/modesetting/driver.c:586:20: warning: unused variable ‘ms’ [-Wunused-variable] 12688 modesettingPtr ms = modesettingPTR(xf86ScreenToScrn(screen)); 12689 12690 Move the variable ms into #ifdef GLAMOR_HAS_GBM, where it is used. 12691 12692 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 12693 Reviewed-by: Adam Jackson <ajax@redhat.com> 12694 12695commit 6d7e1d1de06336c9b49a253810afda8ac4e9f7b2 12696Author: Lyude Paul <lyude@redhat.com> 12697Date: Fri Oct 13 15:44:32 2017 -0400 12698 12699 meson: Don't forget to define DEBUG! 12700 12701 Changes since v2: 12702 - Don't enable by default for debugoptimized builds 12703 12704 Signed-off-by: Lyude Paul <lyude@redhat.com> 12705 Reviewed-by: Adam Jackson <ajax@redhat.com> 12706 12707commit 0debe011901b87f686e2a76ce5edc150b04bf9d1 12708Author: Lyude Paul <lyude@redhat.com> 12709Date: Fri Oct 13 15:44:31 2017 -0400 12710 12711 meson: Silence -Wformat-nonliteral for x86emu 12712 12713 Signed-off-by: Lyude Paul <lyude@redhat.com> 12714 Reviewed-by: Adam Jackson <ajax@redhat.com> 12715 12716commit cbca18c5516084ee540255df52e116209f1c1cbe 12717Author: Lyude Paul <lyude@redhat.com> 12718Date: Fri Oct 13 15:44:30 2017 -0400 12719 12720 x86emu: Fix type conversion warnings on x86_64 with DEBUG 12721 12722 Warnings come from the fact that PRIx32 is not used for printing 32 bit 12723 values instead of "%lx", and "%lx" evaluates to a 64 bit long on 64 bit 12724 systems while PRIx32 always evaluates to the right type for the 12725 respective arch. 12726 12727 Signed-off-by: Lyude Paul <lyude@redhat.com> 12728 Reviewed-by: Adam Jackson <ajax@redhat.com> 12729 12730commit 01470ce0a9628abc8af4fe7b960f0d1eced8cd46 12731Author: Lyude Paul <lyude@redhat.com> 12732Date: Fri Oct 13 15:44:29 2017 -0400 12733 12734 fbdevhw: Fix inconsistent #if DEBUG usage 12735 12736 fbdevhw is the only file in X's source that actually uses #if DEBUG to 12737 check for debugging instead of #ifdef DEBUG. This is contrary to 12738 everything else that checks the DEBUG macro in the source, so let's make 12739 it consistent and in turn, make our meson files a little simpler. 12740 12741 Signed-off-by: Lyude Paul <lyude@redhat.com> 12742 Reviewed-by: Adam Jackson <ajax@redhat.com> 12743 12744commit c66d65a645332bbf055a1c49f092636139eb9285 12745Author: Eric Anholt <eric@anholt.net> 12746Date: Wed Jul 26 16:54:40 2017 -0700 12747 12748 glamor: Use GL_MESA_tile_raster_order for overlapping blits. 12749 12750 Improves Raspberry Pi 3 x11perf -copywinwin500 from ~480/sec to 12751 ~700/sec. 12752 12753 Signed-off-by: Eric Anholt <eric@anholt.net> 12754 Reviewed-by: Keith Packard <keithp@keithp.com> 12755 12756commit 885636b7d42b3c7b151fc386d358184db004ce45 12757Author: Nikolay Martynov <mar.kolya@gmail.com> 12758Date: Wed Sep 13 23:23:13 2017 -0400 12759 12760 XShmGetImage: fix censoring 12761 12762 It looks like offsets calculated during image censoring are wrong. 12763 This results in black (empty) images returns. 12764 12765 This fix is very similar to 6c6f09aac7f1d1367a042087b7681c7fdf1d1e0f 12766 that was applied to XGetImage 12767 12768 Visually this fixes chromium/firefox window sharing in multiscreen 12769 configurations - without this patch most of the windows on 'secodnary' 12770 screens are black. 12771 12772 This also should fix https://bugs.freedesktop.org/show_bug.cgi?id=101730. 12773 12774 Signed-off-by: Nikolay Martynov <mar.kolya@gmail.com> 12775 Reviewed-by: Adam Jackson <ajax@redhat.com> 12776 12777commit a10b4fcdc016d7ac56e5bb35f7f844af41016efe 12778Author: Jon Turney <jon.turney@dronecode.org.uk> 12779Date: Tue Oct 10 14:28:33 2017 +0100 12780 12781 meson: Fix linkage of loadable modules for PE/COFF 12782 12783 For the loadable modules it makes sense to build for PE/COFF targets, link 12784 those loadable modules with the import library for the Xorg executable, so 12785 that symbols provided by the executable can be satisfied at link time (as 12786 required by PE/COFF). 12787 12788 Since this uses the syntax of using the returned build target object from an 12789 executable() with an implib: kwarg to link_with:, introduced in meson 0.42 12790 and a syntax error with older meson, also update the minimum meson version 12791 which we require in project() to that. 12792 12793 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 12794 12795commit d43b1ca852562882fc6930afc36a982801e3be87 12796Author: Jon Turney <jon.turney@dronecode.org.uk> 12797Date: Tue Oct 10 14:28:32 2017 +0100 12798 12799 meson: Fix underlinkage of shadow loadable module 12800 12801 Future work: probably some other modules are underlinked? 12802 12803 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 12804 12805commit 10cba7d54668698741510374ee63dec7a3cc3647 12806Author: Lyude Paul <lyude@redhat.com> 12807Date: Wed Oct 11 18:03:45 2017 -0400 12808 12809 meson: Add xkb_bin_dir option 12810 12811 Now that we can actually configure all of the directories xkb uses for 12812 finding things, we can (finally, but only with meson) finally make it so 12813 that with the correct meson configuration the Xserver will "just work" 12814 without any additional changes to the installation prefix after 12815 building. 12816 12817 For the people like me who have since scripted this part out of their 12818 build process and forgotten about it, building and installing the X 12819 server into a non-standard prefix has always required the following (or 12820 something else that makes sure that X has a valid xkbcomp configuration) 12821 commands be run right after doing the installation: 12822 12823 # start in root of prefix you installed X to 12824 mkdir -pv share/X11/xkb/rules 12825 ln -s /usr/share/X11/xkb/rules/evdev share/X11/xkb/rules/ 12826 rm -f bin/xkbcomp 12827 ln -s /usr/bin/xkbcomp bin/ 12828 12829 The one last piece of getting rid of this post-install junk is making 12830 sure that we can control the directory that X uses for finding the 12831 xkbcomp binary from meson so we can point it at the system provided 12832 xkbcomp (/usr/bin/xkbcomp or similar). So, this patch adds a 12833 configuration option for controlling this called xkb_bin_dir. 12834 12835 Signed-off-by: Lyude Paul <lyude@redhat.com> 12836 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 12837 Reviewed-by: Daniel Stone <daniels@collabora.com> 12838 12839commit b747da5e25be944337a9cd1415506fc06b70aa81 12840Author: Nathan Kidd <nkidd@opentext.com> 12841Date: Fri Jan 9 10:15:46 2015 -0500 12842 12843 Unvalidated extra length in ProcEstablishConnection (CVE-2017-12176) 12844 12845 Reviewed-by: Julien Cristau <jcristau@debian.org> 12846 Signed-off-by: Nathan Kidd <nkidd@opentext.com> 12847 Signed-off-by: Julien Cristau <jcristau@debian.org> 12848 12849commit 4ca68b878e851e2136c234f40a25008297d8d831 12850Author: Nathan Kidd <nkidd@opentext.com> 12851Date: Fri Jan 9 10:09:14 2015 -0500 12852 12853 dbe: Unvalidated variable-length request in ProcDbeGetVisualInfo (CVE-2017-12177) 12854 12855 v2: Protect against integer overflow (Alan Coopersmith) 12856 12857 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 12858 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 12859 Reviewed-by: Julien Cristau <jcristau@debian.org> 12860 Signed-off-by: Nathan Kidd <nkidd@opentext.com> 12861 Signed-off-by: Julien Cristau <jcristau@debian.org> 12862 12863commit 859b08d523307eebde7724fd1a0789c44813e821 12864Author: Nathan Kidd <nkidd@opentext.com> 12865Date: Wed Dec 24 16:22:18 2014 -0500 12866 12867 Xi: fix wrong extra length check in ProcXIChangeHierarchy (CVE-2017-12178) 12868 12869 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 12870 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 12871 Reviewed-by: Julien Cristau <jcristau@debian.org> 12872 Signed-off-by: Nathan Kidd <nkidd@opentext.com> 12873 Signed-off-by: Julien Cristau <jcristau@debian.org> 12874 12875commit d088e3c1286b548a58e62afdc70bb40981cdb9e8 12876Author: Nathan Kidd <nkidd@opentext.com> 12877Date: Fri Jan 9 10:04:41 2015 -0500 12878 12879 Xi: integer overflow and unvalidated length in (S)ProcXIBarrierReleasePointer 12880 12881 [jcristau: originally this patch fixed the same issue as commit 12882 211e05ac85 "Xi: Test exact size of XIBarrierReleasePointer", with the 12883 addition of these checks] 12884 12885 This addresses CVE-2017-12179 12886 12887 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 12888 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 12889 Reviewed-by: Julien Cristau <jcristau@debian.org> 12890 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 12891 Signed-off-by: Nathan Kidd <nkidd@opentext.com> 12892 Signed-off-by: Julien Cristau <jcristau@debian.org> 12893 12894commit 1b1d4c04695dced2463404174b50b3581dbd857b 12895Author: Nathan Kidd <nkidd@opentext.com> 12896Date: Sun Dec 21 01:10:03 2014 -0500 12897 12898 hw/xfree86: unvalidated lengths 12899 12900 This addresses: 12901 CVE-2017-12180 in XFree86-VidModeExtension 12902 CVE-2017-12181 in XFree86-DGA 12903 CVE-2017-12182 in XFree86-DRI 12904 12905 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 12906 Reviewed-by: Julien Cristau <jcristau@debian.org> 12907 Signed-off-by: Nathan Kidd <nkidd@opentext.com> 12908 Signed-off-by: Julien Cristau <jcristau@debian.org> 12909 12910commit 55caa8b08c84af2b50fbc936cf334a5a93dd7db5 12911Author: Nathan Kidd <nkidd@opentext.com> 12912Date: Fri Jan 9 11:43:05 2015 -0500 12913 12914 xfixes: unvalidated lengths (CVE-2017-12183) 12915 12916 v2: Use before swap (Jeremy Huddleston Sequoia) 12917 12918 v3: Fix wrong XFixesCopyRegion checks (Alan Coopersmith) 12919 12920 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 12921 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 12922 Reviewed-by: Julien Cristau <jcristau@debian.org> 12923 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 12924 Signed-off-by: Nathan Kidd <nkidd@opentext.com> 12925 Signed-off-by: Julien Cristau <jcristau@debian.org> 12926 12927commit cad5a1050b7184d828aef9c1dd151c3ab649d37e 12928Author: Nathan Kidd <nkidd@opentext.com> 12929Date: Fri Jan 9 09:57:23 2015 -0500 12930 12931 Unvalidated lengths 12932 12933 v2: Add overflow check and remove unnecessary check (Julien Cristau) 12934 12935 This addresses: 12936 CVE-2017-12184 in XINERAMA 12937 CVE-2017-12185 in MIT-SCREEN-SAVER 12938 CVE-2017-12186 in X-Resource 12939 CVE-2017-12187 in RENDER 12940 12941 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 12942 Reviewed-by: Julien Cristau <jcristau@debian.org> 12943 Signed-off-by: Nathan Kidd <nkidd@opentext.com> 12944 Signed-off-by: Julien Cristau <jcristau@debian.org> 12945 12946commit 9c23685009aa96f4b861dcc5d2e01dbee00c4dd9 12947Author: Michal Srb <msrb@suse.com> 12948Date: Fri Jul 7 17:04:03 2017 +0200 12949 12950 os: Make sure big requests have sufficient length. 12951 12952 A client can send a big request where the 32B "length" field has value 12953 0. When the big request header is removed and the length corrected, 12954 the value will underflow to 0xFFFFFFFF. Functions processing the 12955 request later will think that the client sent much more data and may 12956 touch memory beyond the receive buffer. 12957 12958 Signed-off-by: Eric Anholt <eric@anholt.net> 12959 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 12960 12961commit c2f2b25ab55c67f9f3ad07c02fa746eae7c61196 12962Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 12963Date: Wed Sep 27 01:19:58 2017 -0400 12964 12965 present: Check the whole exec queue on event 12966 12967 Later events are sometimes added in front of the queue (e.g. 12968 if page flipping fails) so we need to check the whole queue 12969 on event. 12970 12971 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 12972 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 12973 12974commit 677c32bcda98a96585bb1f66b57e0755a157b772 12975Author: Keith Packard <keithp@keithp.com> 12976Date: Fri Sep 29 08:48:33 2017 -0700 12977 12978 xf86-video-modesetting: Add ms_queue_vblank helper [v3] 12979 12980 This provides an API wrapper around the kernel interface for queueing 12981 a vblank event, simplifying all of the callers. 12982 12983 v2: Fix missing '|' in computing vbl.request.type 12984 12985 v3: Remove spurious bit of next patch (thanks, Michel Dänzer) 12986 12987 Signed-off-by: Keith Packard <keithp@keithp.com> 12988 Reviewed-by: Adam Jackson <ajax@redhat.com> 12989 12990commit 94f11ca5cf011ef123bd222cabeaef6f424d76ac 12991Author: Keith Packard <keithp@keithp.com> 12992Date: Thu Jul 27 10:08:32 2017 -0700 12993 12994 xkb: Handle xkb formated string output safely (CVE-2017-13723) 12995 12996 Generating strings for XKB data used a single shared static buffer, 12997 which offered several opportunities for errors. Use a ring of 12998 resizable buffers instead, to avoid problems when strings end up 12999 longer than anticipated. 13000 13001 Reviewed-by: Michal Srb <msrb@suse.com> 13002 Signed-off-by: Keith Packard <keithp@keithp.com> 13003 Signed-off-by: Julien Cristau <jcristau@debian.org> 13004 13005commit eaf1f72ed8994b708d94ec2de7b1a99f5c4a39b8 13006Author: Michal Srb <msrb@suse.com> 13007Date: Thu Jul 27 11:54:26 2017 +0200 13008 13009 xkb: Escape non-printable characters correctly. 13010 13011 XkbStringText escapes non-printable characters using octal numbers. Such escape 13012 sequence would be at most 5 characters long ("\0123"), so it reserves 5 bytes 13013 in the buffer. Due to char->unsigned int conversion, it would print much longer 13014 string for negative numbers. 13015 13016 Reviewed-by: Keith Packard <keithp@keithp.com> 13017 Signed-off-by: Julien Cristau <jcristau@debian.org> 13018 13019commit b95f25af141d33a65f6f821ea9c003f66a01e1f1 13020Author: Michal Srb <msrb@suse.com> 13021Date: Fri Jul 28 16:27:10 2017 +0200 13022 13023 Xext/shm: Validate shmseg resource id (CVE-2017-13721) 13024 13025 Otherwise it can belong to a non-existing client and abort X server with 13026 FatalError "client not in use", or overwrite existing segment of another 13027 existing client. 13028 13029 Signed-off-by: Julien Cristau <jcristau@debian.org> 13030 13031commit db465bae533f85e7f900deb96efecc831c9d550b 13032Author: Adam Jackson <ajax@redhat.com> 13033Date: Wed Sep 27 14:31:39 2017 -0400 13034 13035 meson: Port default warning flags from xorg-macros 13036 13037 Well, almost all of them. No -Wdeclaration-after-statement because 13038 that's legal in C99, and in the limited ways we use it, more readable. 13039 13040 Signed-off-by: Adam Jackson <ajax@redhat.com> 13041 Reviewed-by: Daniel Stone <daniels@collabora.com> 13042 13043commit 40c90ead044aa3fab79280d7a5ca5ad8e390029e 13044Author: Adam Jackson <ajax@redhat.com> 13045Date: Wed Sep 27 14:31:38 2017 -0400 13046 13047 meson: Check for HAVE_TYPEOF 13048 13049 Without this, exa's gc swap macros trigger gcc's -Wdiscarded-qualifiers. 13050 13051 Signed-off-by: Adam Jackson <ajax@redhat.com> 13052 Reviewed-by: Daniel Stone <daniels@collabora.com> 13053 13054commit 712b02ec72b98291b78e17dcb1073a13c4a1086e 13055Author: Adam Jackson <ajax@redhat.com> 13056Date: Wed Sep 27 14:31:37 2017 -0400 13057 13058 meson: Default to gnu99 13059 13060 We don't really require all of C99, but enough that it's not worth 13061 bothering with the distinction, especially if your toolchain is new 13062 enough that meson is a thing for you. We could do strict C99 if we 13063 really insisted on spelling it __typeof__, but who wants that? Nobody, 13064 that's who. 13065 13066 Signed-off-by: Adam Jackson <ajax@redhat.com> 13067 Reviewed-by: Eric Anholt <eric@anholt.net> 13068 13069commit 041f25afa81dfc7e7ef350d5b955d2d5912846c9 13070Author: Adam Jackson <ajax@redhat.com> 13071Date: Wed Sep 27 14:31:36 2017 -0400 13072 13073 test: const correctness fix 13074 13075 ../test/sync/sync.c: In function ‘main’: 13076 ../test/sync/sync.c:288:40: warning: initialization discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers] 13077 xcb_query_extension_reply_t *ext = xcb_get_extension_data(c, &xcb_sync_id); 13078 ^~~~~~~~~~~~~~~~~~~~~~ 13079 13080 Signed-off-by: Adam Jackson <ajax@redhat.com> 13081 Reviewed-by: Eric Anholt <eric@anholt.net> 13082 13083commit 1d1ff1142a5a3904175f45d2f52bbdffb0249e57 13084Author: Adam Jackson <ajax@redhat.com> 13085Date: Wed Sep 27 14:31:35 2017 -0400 13086 13087 dmx: More const correctness 13088 13089 Fixes several dozen cases like: 13090 13091 ../hw/dmx/examples/ev.c: In function ‘main’: 13092 ../hw/dmx/examples/ev.c:147:29: warning: assignment discards ‘const’ qualifier from pointer target type [-Wdiscarded-qualifiers] 13093 tmp = "X"; 13094 ^ 13095 13096 Signed-off-by: Adam Jackson <ajax@redhat.com> 13097 Reviewed-by: Eric Anholt <eric@anholt.net> 13098 13099commit eb25facb37ef74eab83060b75f8205ce1538817f 13100Author: Adam Jackson <ajax@redhat.com> 13101Date: Wed Sep 27 14:31:34 2017 -0400 13102 13103 dmx: Fix a silly redeclaration bug 13104 13105 ../hw/dmx/dmx.c:66:12: warning: redundant redeclaration of ‘PanoramiXNumScreens’ [-Wredundant-decls] 13106 extern int PanoramiXNumScreens; 13107 ^~~~~~~~~~~~~~~~~~~ 13108 In file included from ../hw/dmx/dmx.c:65:0: 13109 ../Xext/panoramiXsrv.h:11:22: note: previous declaration of ‘PanoramiXNumScreens’ was here 13110 extern _X_EXPORT int PanoramiXNumScreens; 13111 ^~~~~~~~~~~~~~~~~~~ 13112 13113 Signed-off-by: Adam Jackson <ajax@redhat.com> 13114 Reviewed-by: Eric Anholt <eric@anholt.net> 13115 13116commit 8060196a3e80a3c0ad2c0abbe459416821cd366c 13117Author: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 13118Date: Wed Sep 27 01:24:28 2017 -0400 13119 13120 sync: Fix delta value check 13121 13122 While transitionning from CARD64 to int64, the GreaterThan call 13123 as mistakenly been transformed into ">=". That was at least 13124 causing problems with Mutter. 13125 13126 Signed-off-by: Louis-Francis Ratté-Boulianne <lfrb@collabora.com> 13127 Reviewed-by: Adam Jackson <ajax@redhat.com> 13128 13129commit 37f4e7651a2fd51efa613a08a1e705553be33e76 13130Author: Keith Packard <keithp@keithp.com> 13131Date: Mon Sep 25 16:18:22 2017 -0700 13132 13133 modesetting: Skip no-longer-present connectors when resetting BAD links 13134 13135 Outputs may have NULL mode_output (connector) pointers if the 13136 connector disappears while the server is running. Skip these when 13137 resetting outputs with BAD link status. 13138 13139 Signed-off-by: Keith Packard <keithp@keithp.com> 13140 Reviewed-by: Adam Jackson <ajax@redhat.com> 13141 13142commit 147b4602f9c47e8977d13b1cbb6566f86dba8647 13143Author: Adam Jackson <ajax@redhat.com> 13144Date: Mon Sep 25 15:06:44 2017 -0400 13145 13146 xfree86: Hush some warnings when Xv is disabled 13147 13148 Spotted by Appveyor: 13149 13150 xf86Crtc.c:3281:1: warning: ‘xf86_crtc_box_area’ defined but not used [-Wunused-function] 13151 xf86_crtc_box_area(BoxPtr box) 13152 ^~~~~~~~~~~~~~~~~~ 13153 xf86Crtc.c:3268:1: warning: ‘x86_crtc_box’ defined but not used [-Wunused-function] 13154 x86_crtc_box(xf86CrtcPtr crtc, BoxPtr crtc_box) 13155 ^~~~~~~~~~~~ 13156 xf86Crtc.c:3256:1: warning: ‘x86_crtc_box_intersect’ defined but not used [-Wunused-function] 13157 x86_crtc_box_intersect(BoxPtr dest, BoxPtr a, BoxPtr b) 13158 ^~~~~~~~~~~~~~~~~~~~~~ 13159 13160 Signed-off-by: Adam Jackson <ajax@redhat.com> 13161 Reviewed-by: Eric Anholt <eric@anholt.net> 13162 13163commit 0888b22fea71118b1a9a238134b3b8d1dc659734 13164Author: Adam Jackson <ajax@redhat.com> 13165Date: Mon Sep 25 15:01:32 2017 -0400 13166 13167 test: Fix a thinko in simple-xinit 13168 13169 Spotted by clang courtesy of the shiny new OSX Travis target: 13170 13171 simple-xinit.c:90:65: warning: sizeof on pointer operation will return size of 'char *' instead of 'char [10]' [-Wsizeof-array-decay] 13172 ret = read(displayfd, display_string, sizeof(display_string - 1)); 13173 13174 Signed-off-by: Adam Jackson <ajax@redhat.com> 13175 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 13176 13177commit 84e3b96b531363e47f6789aacfcae4aa60135e2e 13178Author: Nick Sarnie <commendsarnex@gmail.com> 13179Date: Sat Sep 23 17:35:48 2017 -0400 13180 13181 suid: Include sysmacros.h to fix build after glibc-2.25 13182 13183 [Added HAVE_SYS_SYSMACROS_H guard - ajax] 13184 13185 Signed-off-by: Nick Sarnie <commendsarnex@gmail.com> 13186 Reviewed-by: Adam Jackson <ajax@redhat.com> 13187 13188commit 0b00440678fa22c1c1ca4a292b67db3ab8676969 13189Author: Jon Turney <jon.turney@dronecode.org.uk> 13190Date: Fri Sep 22 15:25:05 2017 +0100 13191 13192 travis: Also build on OSX 13193 13194 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 13195 Signed-off-by: Eric Anholt <eric@anholt.net> 13196 Reviewed-by: Eric Anholt <eric@anholt.net> 13197 13198commit c30eca688781deb06b25a6a08297cb855aeb6ed2 13199Author: Jon Turney <jon.turney@dronecode.org.uk> 13200Date: Fri Sep 22 13:54:01 2017 +0100 13201 13202 meson: Make it possible to build for 32-bit targets 13203 13204 Setting glx_align64 to '' gives a null string in the arguments list passed 13205 to the compiler. This is taken as an input filename, leading to: 13206 13207 "cc: error: : No such file or directory" 13208 13209 Instead, assign an empty list to glx_align64, which gets flattened to 13210 nothing in the arguments list. 13211 13212 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 13213 Signed-off-by: Eric Anholt <eric@anholt.net> 13214 Reviewed-by: Eric Anholt <eric@anholt.net> 13215 13216commit 2c2e1b6f57753a82c7ca144a57950f9e2d76199e 13217Author: Adam Jackson <ajax@redhat.com> 13218Date: Wed Aug 30 14:58:24 2017 -0400 13219 13220 glx: Be a bit more paranoid in glx client cleanup 13221 13222 This would probably crash (via double-free) if you had multiple GPUs and 13223 an indirect context. 13224 13225 Signed-off-by: Adam Jackson <ajax@redhat.com> 13226 13227commit d770f9293296d2d3c4e49e94130ff68c0890f625 13228Author: Daniel Stone <daniels@collabora.com> 13229Date: Wed Sep 20 07:22:13 2017 -0700 13230 13231 meson: Use dependency version_compare() 13232 13233 We can check the version on an existing dependency, rather than spinning 13234 up pkg-config again just to check the right version. 13235 13236 Signed-off-by: Daniel Stone <daniels@collabora.com> 13237 13238commit 86a89dd6020f59964c0e924acbfae9d5a51e3148 13239Author: Daniel Stone <daniels@collabora.com> 13240Date: Wed Sep 20 07:22:12 2017 -0700 13241 13242 Build: Use dri3proto/libdrm CFLAGS 13243 13244 Make sure we get the CFLAGS required for building DRI3 into the 13245 command line. 13246 13247 Signed-off-by: Daniel Stone <daniels@collabora.com> 13248 13249commit 294670682120c65001b36369d6395003704f4ac1 13250Author: Eric Anholt <eric@anholt.net> 13251Date: Mon Sep 18 17:34:33 2017 -0700 13252 13253 sync: Clean up a bit of header formatting. 13254 13255 Signed-off-by: Eric Anholt <eric@anholt.net> 13256 Reviewed-by: Keith Packard <keithp@keithp.com> 13257 13258commit e0f872207aa203adb85e825c311ed50fe3a3af60 13259Author: Eric Anholt <eric@anholt.net> 13260Date: Mon Sep 18 17:34:32 2017 -0700 13261 13262 sync: Convert from "CARD64" to int64_t. 13263 13264 The extension was using the name CARD64 to represent 64-bit values, 13265 with a #define from CARD64 to XSyncValue, a struct with a pair of 13266 32-bit values representing a signed 64-bit value. This interfered 13267 with protocol headers using CARD64 to try to actually store a 13268 uint64_t. Now that stdint.h exists, let's just use that here, 13269 instead. 13270 13271 v2: Fix alarm delta changes. 13272 v3: Do the potentially overflowing math as uint and convert to int 13273 afterward, out of C spec paranoia. 13274 13275 Signed-off-by: Eric Anholt <eric@anholt.net> 13276 Reviewed-by: Keith Packard <keithp@keithp.com> 13277 13278commit 5cbfa276541e6a621cf9c4b44b75323e90a5bd4c 13279Author: Eric Anholt <eric@anholt.net> 13280Date: Mon Sep 18 17:34:31 2017 -0700 13281 13282 test: Add basic SYNC tests. 13283 13284 I couldn't find any, and I was modifying the implementation, so I had 13285 to write some. I would like the test to end with a "make sure there 13286 weren't any stray unchecked errors", but I didn't figure out how to do 13287 that. 13288 13289 v2: Extend sync tests to cover alarm delta and waitvalue changes. 13290 13291 Signed-off-by: Eric Anholt <eric@anholt.net> 13292 Reviewed-by: Keith Packard <keithp@keithp.com> 13293 13294commit 3336291fc68444ee65b48ba675ec947e505fed57 13295Author: Eric Anholt <eric@anholt.net> 13296Date: Mon Sep 18 17:34:30 2017 -0700 13297 13298 test: Return error from simple-xinit if the client crashes. 13299 13300 I want to be able to call client tests with simple-xinit, so assertion 13301 failures should be an error. 13302 13303 v2: Clean up identical returns. 13304 13305 Signed-off-by: Eric Anholt <eric@anholt.net> 13306 13307commit a8eeb332ccf4d13b3fdcc382397bd3ea45e76212 13308Author: Eric Anholt <eric@anholt.net> 13309Date: Mon Sep 18 17:34:29 2017 -0700 13310 13311 meson: Add Xvfb and Xephyr-glamor testing. 13312 13313 The Xvfb tests are passing and Xephyr-glamor is failing for me, but it 13314 fails identically on autotools. It's disabled on Travis for now 13315 because the >10 minutes of silence during testing times out the entire 13316 build. 13317 13318 v2: Fix the disable on travis. 13319 13320 Signed-off-by: Eric Anholt <eric@anholt.net> 13321 Reviewed-by: Adam Jackson <ajax@redhat.com> 13322 13323commit a09743c9300f805d6527368ddcf44f5dccd4b366 13324Author: Eric Anholt <eric@anholt.net> 13325Date: Mon Sep 18 17:34:28 2017 -0700 13326 13327 meson: Move Xvfb build under an option. 13328 13329 Autotools also had it as an option. 13330 13331 Signed-off-by: Eric Anholt <eric@anholt.net> 13332 Reviewed-by: Adam Jackson <ajax@redhat.com> 13333 13334commit de3b61869140768335daed0e855a5ae4bdb75020 13335Author: Jon Turney <jon.turney@dronecode.org.uk> 13336Date: Fri Sep 15 14:36:52 2017 +0100 13337 13338 Add an .appveyor.yml for AppVeyor CI 13339 13340 This currently does an autotools build using Cygwin. 13341 13342 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 13343 Reviewed-by: Adam Jackson <ajax@redhat.com> 13344 13345commit 7d0728d6c42f9200b90c58e0357776018de18496 13346Author: Jon Turney <jon.turney@dronecode.org.uk> 13347Date: Fri Sep 15 14:36:51 2017 +0100 13348 13349 Revert "dmx: fix linking" 13350 13351 Since commit 3ef16dfb9830bd6b41ae428f4f213ae0c35c1056, "dmx: fix 13352 linking", linking dmx is broken for me: 13353 13354 CCLD Xdmx.exe 13355 ../../render/.libs/librender.a(glyph.o): In function `HashGlyph': 13356 /wip/xserver/build.x86_64/../render/glyph.c:168: undefined reference to `x_sha1_init' 13357 /wip/xserver/build.x86_64/../render/glyph.c:174: undefined reference to `x_sha1_update' 13358 /wip/xserver/build.x86_64/../render/glyph.c:177: undefined reference to `x_sha1_update' 13359 /wip/xserver/build.x86_64/../render/glyph.c:180: undefined reference to `x_sha1_final' 13360 ../../render/.libs/librender.a(mipict.o): In function `miClipPictureReg': 13361 /wip/xserver/build.x86_64/../render/mipict.c:233: undefined reference to `pixman_region_n_rects' 13362 /wip/xserver/build.x86_64/../render/mipict.c:234: undefined reference to `pixman_region_n_rects' 13363 /wip/xserver/build.x86_64/../render/mipict.c:235: undefined reference to `pixman_region_rectangles' 13364 /wip/xserver/build.x86_64/../render/mipict.c:236: undefined reference to `pixman_region_rectangles' 13365 /wip/xserver/build.x86_64/../render/mipict.c:248: undefined reference to `pixman_region_init' 13366 /wip/xserver/build.x86_64/../render/mipict.c:251: undefined reference to `pixman_region_not_empty' 13367 /wip/xserver/build.x86_64/../render/mipict.c:261: undefined reference to `pixman_region_not_empty' 13368 ../../render/.libs/librender.a(mipict.o): In function `miComputeCompositeRegion': 13369 /wip/xserver/build.x86_64/../render/mipict.c:340: undefined reference to `pixman_region_init' 13370 13371 The change this was fixing appears to be effectively reverted by 13372 542d9f6807ac06b70f564ccab10af69fa21a1221, so just revert commit 13373 3ef16dfb9830bd6b41ae428f4f213ae0c35c1056. 13374 13375 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 13376 Reviewed-by: Adam Jackson <ajax@redhat.com> 13377 13378commit fbd80b2c8ebe9fd41229dc5438524d107c071ff1 13379Author: Dawid Kurek <dawid.kurek@displaylink.com> 13380Date: Thu Jul 6 14:51:11 2017 +0200 13381 13382 modesetting: Blacklist EVDI devices from PRIME sync 13383 13384 UDL (usb) devices are blacklisted because of they weird behaviour when 13385 it comes to vblank events. As EVDI uses very similar model of handling 13386 vblanks it should be treated similarly. 13387 13388 When doing a page flip, EVDI does not wait for real vblank, but 13389 simulates it by adding constant delay. It also does not support 13390 DRM_IOCTL_WAIT_VBLANK. 13391 13392 In contrast to UDL, EVDI uses platform devices, thus instead of 'usb' in 13393 path they all have 'platform'. 13394 13395 It is possible to blacklist by 'platform', so without explicitly saying 13396 'evdi', but it might be misleading when it comes to real reason for it. 13397 13398 Signed-off-by: Dawid Kurek <dawid.kurek@displaylink.com> 13399 13400commit eac1a2e37b833c179b587107230805ea1fb3dfda 13401Author: Emil Velikov <emil.velikov@collabora.com> 13402Date: Thu Aug 3 19:43:28 2017 +0100 13403 13404 dri2: sort DRI2InfoPtr::version checking in ascending order 13405 13406 Makes it easer to follow if 8 is between 7 and 9 ;-) 13407 13408 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 13409 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 13410 13411commit 82df2ce38c560915f8c6574052bd56215b649072 13412Author: Roman Gilg <subdiff@gmail.com> 13413Date: Tue Aug 22 15:38:26 2017 +0200 13414 13415 xwayland: Avoid repeatedly looping through window ancestor chain 13416 13417 Calling xwl_window_from_window means looping through the window ancestor 13418 chain whenever it is called on a child window or on an automatically 13419 redirected window. 13420 13421 Since these properties and the potential ancestor's xwl_window are constant 13422 between window realization and unrealization, we can omit the looping by 13423 always putting the respective xwl_window in the Window's private field on 13424 its realization. If the Window doesn't feature an xwl_window on its own, 13425 it's the xwl_window of its first ancestor with one. 13426 13427 Signed-off-by: Roman Gilg <subdiff@gmail.com> 13428 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 13429 13430commit 1089d5d518a315963a8cda6c7d47a0ce09de0979 13431Author: Olivier Fourdan <ofourdan@redhat.com> 13432Date: Thu Mar 2 11:03:15 2017 +0100 13433 13434 xwayland: add envvar XWAYLAND_NO_GLAMOR 13435 13436 Not all compositors allow for customizing the Xwayland command line, 13437 gnome-shell/mutter for example have the command line and path to 13438 Xwayland binary hardcoded, which makes it harder for users to disable 13439 glamor acceleration in Xwayland (glamor being used by default). 13440 13441 Add an environment variable XWAYLAND_NO_GLAMOR to disable glamor support 13442 in Xwayland. 13443 13444 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 13445 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 13446 13447commit 5abaa50b292798370a665ace5eec56fd830be226 13448Author: Eric Anholt <eric@anholt.net> 13449Date: Fri Sep 1 12:22:59 2017 -0700 13450 13451 meson: Move the BUILD_DATE/TIME setup to configure time. 13452 13453 By having it as a custom_target with build_always, every "ninja -C 13454 build" would rebuild Xorg for the new date/time, even if the rest of 13455 Xorg didn't change. 13456 13457 We could build the rest of Xorg into a static lib, and regenerate 13458 date/time when the static lib changes and link that into a final Xorg, 13459 but BUILD_DATE/TIME is such a dubious feature (compared to including a 13460 git sha, which is easy with meson) it doesn't seem worth the build 13461 time cost. 13462 13463 Signed-off-by: Eric Anholt <eric@anholt.net> 13464 Reviewed-by: Adam Jackson <ajax@redhat.com> 13465 13466commit 05e7e8b587dd9d37e8beadc72ab993f028c47fa1 13467Author: Eric Anholt <eric@anholt.net> 13468Date: Fri Sep 1 12:22:58 2017 -0700 13469 13470 meson: Include BUILD_DATE in the meson xf86Build.h. 13471 13472 Due to a typo, I only had BUILD_TIME present. 13473 13474 Signed-off-by: Eric Anholt <eric@anholt.net> 13475 Reviewed-by: Adam Jackson <ajax@redhat.com> 13476 13477commit 2b080a14c87fc9e5f77fc3361297ac332aa04f02 13478Author: Eric Anholt <eric@anholt.net> 13479Date: Fri Sep 1 12:22:57 2017 -0700 13480 13481 meson: Respect SOURCE_DATE_EPOCH for reproducible builds. 13482 13483 This just copies over Chris Lamb's code from autotools. 13484 13485 Signed-off-by: Eric Anholt <eric@anholt.net> 13486 Reviewed-by: Adam Jackson <ajax@redhat.com> 13487 13488commit 937ed782ae5e4e0da739f0630d1681b3754f0281 13489Author: Chris Lamb <lamby@debian.org> 13490Date: Fri Sep 1 12:22:56 2017 -0700 13491 13492 configure.ac: Make BUILD_{DATE, TIME} respect SOURCE_DATE_EPOCH if set 13493 13494 Whilst working on the Reproducible Builds effort [0], we noticed that 13495 xorg-server could not be built reproducibly. One reason is because it 13496 embeds a "current" build and date time. 13497 13498 This should be compatible with both GNU and BSD date(1). 13499 13500 [0] https://reproducible-builds.org/ 13501 13502 v2: Fix change in Y-M-D format that broke the build. 13503 13504 Signed-off-by: Eric Anholt <eric@anholt.net> 13505 Reviewed-by: Adam Jackson <ajax@redhat.com> 13506 13507commit a49379b6045453c7b787cc638db6afd0d14dce9c 13508Author: Adam Jackson <ajax@redhat.com> 13509Date: Tue Sep 12 16:53:24 2017 -0400 13510 13511 fb: Check whether the window is enabled directly 13512 13513 ... instead of its root window. Xwayland's rootless mode empties the 13514 root window border clip since its root window has no storage, but 13515 redirected windows (the only kind it can show) will have a non-empty 13516 border clip anyway, cf. the #ifdef COMPOSITE in miComputeClips. With 13517 this change, non-glamor Xwayland's GetImage actually works. 13518 13519 Other acceleration layers may need to change to account for this, but 13520 this appears to be safe for the existing open source drivers. Only the 13521 xfree86 DDX has any problem with losing its framebuffer on VT switch, 13522 and even then only for UMS drivers (which excludes glamor, uxa, and sna 13523 from consideration). This leaves exa, which already contains code to 13524 evict pixmaps to host memory on VT switch. Since the xfree86 core will 13525 still empty the root clip on VT switch, while the root window itself may 13526 not contain a valid image we won't try to touch it, but GetImage from a 13527 redirected window will now work even when switched away. 13528 13529 Acked-by: Keith Packard <keithp@keithp.com> 13530 Signed-off-by: Adam Jackson <ajax@redhat.com> 13531 13532commit 9869dcb349b49f6d4cc2fab5d927cd8b1d1f463c 13533Author: Olivier Fourdan <ofourdan@redhat.com> 13534Date: Wed Jul 26 16:00:38 2017 +0200 13535 13536 glamor: Avoid overflow between box32 and box16 box 13537 13538 glamor_compute_transform_clipped_regions() uses a temporary box32 13539 internally which is copied back to a box16 to init the regions16, 13540 thus causing a potential overflow. 13541 13542 If an overflow occurs, the given region is invalid and the pixmap 13543 init region will fail. 13544 13545 Simply check that the coordinates won't overflow when copying back to 13546 the box16, avoiding a crash later down the line in glamor. 13547 13548 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=101894 13549 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 13550 Tested-by: Fabrice Bellet <fabrice@bellet.info> 13551 Reviewed-by: Adam Jackson <ajax@redhat.com> 13552 13553commit bd353e9b84e013fc34ed730319d5b63d20977903 13554Author: Olivier Fourdan <ofourdan@redhat.com> 13555Date: Wed Jul 26 16:00:37 2017 +0200 13556 13557 glamor: handle NULL source picture 13558 13559 COMPOSITE_REGION() can pass NULL as a source picture, make sure we 13560 handle that nicely in both glamor_composite_clipped_region() and 13561 glamor_composite_choose_shader(). 13562 13563 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=101894 13564 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 13565 Reviewed-by: Adam Jackson <ajax@redhat.com> 13566 13567commit 4486d199bd3bcb5b2b8ad9bc54eb11604d9bd653 13568Author: Thomas Hellstrom <thellstrom@vmware.com> 13569Date: Wed Sep 6 16:27:54 2017 +0200 13570 13571 glx: Fix visual fbconfig matching with respect to swap method 13572 13573 For the built in visuals, we'd typically select the "best" fbconfig 13574 without considering the swap method. If the client then requests a 13575 specific swap method, say GLX_SWAP_COPY_OML, it may well happen that the 13576 first fbconfig matching requirements would have been paired with the 32-bit 13577 compositing visual, and the client would render a potentially transparent 13578 window. 13579 13580 Fix this so that we try to match fbconfigs with the same swap method to all 13581 built-in visuals. That would guarantee that selecting a specific swap- 13582 method would not influence the chance of getting a compositing visual. 13583 13584 Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> 13585 Reviewed-by: Adam Jackson <ajax@redhat.com> 13586 13587commit 0fc26310d5b09213c65f50bde444a1758172b016 13588Author: Thomas Hellstrom <thellstrom@vmware.com> 13589Date: Wed Sep 6 16:27:53 2017 +0200 13590 13591 glx: Work around a GLX_OML swap method in older dri drivers 13592 13593 The swapMethod config member would typically contain an arbitrary value 13594 on older dri drivers. Fix this so that if we detect an illegal value, 13595 return GLX_SWAP_UNDEFINED_OML. 13596 13597 Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> 13598 Reviewed-by: Adam Jackson <ajax@redhat.com> 13599 13600commit da29a15874229d78ec61946eacdbbed9871e3664 13601Author: Emil Velikov <emil.l.velikov@gmail.com> 13602Date: Thu Sep 29 18:36:43 2016 +0100 13603 13604 glx: remove unused systemTimeExtension 13605 13606 Not even a single DRI2/DRISW driver in mesa ever used this. Appears to be a 13607 dri1 artefact copy/pasted in the dri2/drisw codebase. 13608 13609 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13610 Signed-off-by: Eric Anholt <eric@anholt.net> 13611 Reviewed-by: Eric Anholt <eric@anholt.net> 13612 13613commit ea82ececbf85a7ac3d0931687f44c57534fde17c 13614Author: Peter Hutterer <peter.hutterer@who-t.net> 13615Date: Wed Sep 6 11:53:02 2017 +1000 13616 13617 test: fix compiler warning 13618 13619 signal-logging.c:182:12: warning: suggest parentheses around assignment used as truth value [-Wparentheses] 13620 13621 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 13622 13623commit e12e02b2e6ee4582ba902ec1d1e5309abc78e282 13624Author: Emil Velikov <emil.l.velikov@gmail.com> 13625Date: Sun Apr 17 17:24:53 2016 +0100 13626 13627 glamor: use the lowercase xnf.*alloc API 13628 13629 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13630 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 13631 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 13632 13633commit 17e488570067cdf499c7ad17ab198508c6f1531d 13634Author: Emil Velikov <emil.l.velikov@gmail.com> 13635Date: Sun Apr 17 17:24:52 2016 +0100 13636 13637 xwayland: use the lowercase xnf.*alloc API 13638 13639 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13640 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 13641 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 13642 13643commit cdd0352ba05d4d8482aaca41797e05d40e58da36 13644Author: Olivier Fourdan <ofourdan@redhat.com> 13645Date: Thu Aug 31 10:23:00 2017 +0200 13646 13647 xwayland: Fix a segfault with pointer locking 13648 13649 Xwayland would crash in some circumstances while trying to issue a 13650 pointer locking when the cursor is hidden when there is no seat focus 13651 window set. 13652 13653 The crash signature looks like: 13654 13655 #0 zwp_pointer_constraints_v1_lock_pointer () 13656 #1 xwl_pointer_warp_emulator_lock () at xwayland-input.c:2584 13657 #2 xwl_seat_maybe_lock_on_hidden_cursor () at xwayland-input.c:2756 13658 #3 xwl_seat_maybe_lock_on_hidden_cursor () at xwayland-input.c:2765 13659 #4 xwl_seat_cursor_visibility_changed () at xwayland-input.c:2768 13660 #5 xwl_set_cursor () at xwayland-cursor.c:245 13661 #6 miPointerUpdateSprite () at mipointer.c:468 13662 #7 miPointerDisplayCursor () at mipointer.c:206 13663 #8 CursorDisplayCursor () at cursor.c:150 13664 #9 AnimCurDisplayCursor () at animcur.c:220 13665 #10 ChangeToCursor () at events.c:936 13666 #11 ActivatePointerGrab () at events.c:1542 13667 #12 GrabDevice () at events.c:5120 13668 #13 ProcGrabPointer () at events.c:4908 13669 #14 Dispatch () at dispatch.c:478 13670 #15 dix_main () at main.c:276 13671 13672 xwl_pointer_warp_emulator_lock() tries to use the surface from the 13673 xwl_seat->focus_window leading to a NULL pointer dereference when that 13674 value is NULL. 13675 13676 Check that xwl_seat->focus_window is not NULL earlier in the stack in 13677 xwl_seat_maybe_lock_on_hidden_cursor() and return early if not the case 13678 to avoid the crash. 13679 13680 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=102474 13681 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 13682 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 13683 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 13684 13685commit 3fbc3c3eefd3803d0c4e86e42595d8a891aeb5f2 13686Author: Olivier Fourdan <ofourdan@redhat.com> 13687Date: Fri Sep 1 09:35:33 2017 +0200 13688 13689 xwayland: No grab handler without protocol support 13690 13691 If the compositor has no support for the Xwayland keyboard grab 13692 protocol, there is no need to set-up our keyboard grab handler. 13693 13694 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 13695 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 13696 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 13697 13698commit aabf65d2a0206bd1a9c6e9a9f3153ded873dfd43 13699Author: Adam Jackson <ajax@redhat.com> 13700Date: Wed Aug 30 15:11:45 2017 -0400 13701 13702 os: Fix warning in LockServer 13703 13704 The meson build gives me: 13705 13706 ../os/utils.c: In function ‘LockServer’: 13707 ../os/utils.c:310:40: warning: ‘snprintf’ output may be truncated before the last format character [-Wformat-truncation=] 13708 snprintf(pid_str, sizeof(pid_str), "%10ld\n", (long) getpid()); 13709 ^~~~~~~~~ 13710 ../os/utils.c:310:5: note: ‘snprintf’ output between 12 and 13 bytes into a destination of size 12 13711 snprintf(pid_str, sizeof(pid_str), "%10ld\n", (long) getpid()); 13712 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 13713 13714 Which seems to be due to the %d part meaning that a negative number's - 13715 sign would be one wider than we're expecting. Fine, just coerce it to 13716 unsigned. 13717 13718 Signed-off-by: Adam Jackson <ajax@redhat.com> 13719 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 13720 13721commit 3bc32a7f593cbd2747237621135ef5196bc1e15a 13722Author: Adam Jackson <ajax@redhat.com> 13723Date: Tue Aug 22 12:19:56 2017 -0400 13724 13725 glx: Fix error generation for non-reply vendor private requests 13726 13727 Discarding the return value here is just wrong. 13728 13729 Signed-off-by: Adam Jackson <ajax@redhat.com> 13730 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 13731 13732commit 591ac95f73b1c1ea681c45e4da9afa2b1d505abe 13733Author: Emil Velikov <emil.l.velikov@gmail.com> 13734Date: Tue Aug 22 11:48:31 2017 +0100 13735 13736 xwin: automake: remove unused {SRCS, DEFS}_{NATIVEGDI, PRIMARYFB} 13737 13738 Left over from the following commits: 13739 13740 8465ee788fd xwin: Remove native GDI engine (v2) 13741 c79f824bf66 xwin: Remove primary DirectDraw engine 13742 13743 v2: drop leading - in the makefile 13744 13745 Cc: Adam Jackson <ajax@redhat.com> 13746 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13747 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> (v1) 13748 13749commit 69fe6156ec6cd37657d4e5af1c6f21ebf72b8fd7 13750Author: Emil Velikov <emil.l.velikov@gmail.com> 13751Date: Tue Aug 22 11:48:30 2017 +0100 13752 13753 xwin: remove always true/set XWIN_RANDR conditional/define 13754 13755 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13756 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 13757 13758commit 8f1a200d0be261a60bb9e8e72d8704cef2d91cd4 13759Author: Emil Velikov <emil.l.velikov@gmail.com> 13760Date: Tue Aug 22 11:48:29 2017 +0100 13761 13762 xwin: remove always true/set XWIN_MULTIWINDOW conditional/define 13763 13764 v2: drop trailing endif (Jon) 13765 13766 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13767 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 13768 13769commit 9d00f6190a8e5cc4ebda9b315d7c0a0782b58501 13770Author: Emil Velikov <emil.l.velikov@gmail.com> 13771Date: Tue Aug 22 11:48:28 2017 +0100 13772 13773 xwin: remove always true/set XWIN_CLIPBOARD conditional/define 13774 13775 v2: drop trailing endif (Jon) 13776 13777 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13778 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 13779 13780commit 1ef65692258136d6e36d0d9d6c5efad9877c9ceb 13781Author: Emil Velikov <emil.l.velikov@gmail.com> 13782Date: Tue Aug 22 11:48:27 2017 +0100 13783 13784 os: make MitGenerateCookie() independent of XCSECURITY 13785 13786 Analogous to previous commit. 13787 13788 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13789 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 13790 13791commit 292ee7151631cabbd928847a5a2ca47e14a06b75 13792Author: Emil Velikov <emil.l.velikov@gmail.com> 13793Date: Tue Aug 22 11:48:26 2017 +0100 13794 13795 os: make GenerateRandomData() independent of XCSECURITY 13796 13797 The function itself does not depend on the macro. Move it outside 13798 of the ifdef guard and remove the identical copy in XWIN. 13799 13800 This is step 1 towards removing the duplication in winauth.c and moving 13801 the OS specifics to os/ 13802 13803 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13804 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 13805 13806commit 8aee1f40eaa3527c90698d6826ce8314b8b5a92a 13807Author: Emil Velikov <emil.l.velikov@gmail.com> 13808Date: Tue Aug 22 11:48:25 2017 +0100 13809 13810 xwin/glx: remove unused __GLXWinScreen::glx_enable_bits 13811 13812 All the relevant code already uses the ::base::glx_enable_bits one. 13813 13814 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 13815 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 13816 13817commit 43878fa8772d8642f601d65e78d2f702ecc0de41 13818Author: Adam Jackson <ajax@redhat.com> 13819Date: Wed Aug 16 14:49:18 2017 -0400 13820 13821 glx: Remove some unused stuff from glxserver.h 13822 13823 Signed-off-by: Adam Jackson <ajax@redhat.com> 13824 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 13825 Reviewed-by: Keith Packard <keithp@keithp.com> 13826 13827commit 3d81abba9ca2f9d57b4c293e3b7fcb32ef853f44 13828Author: Adam Jackson <ajax@redhat.com> 13829Date: Wed Aug 16 14:49:17 2017 -0400 13830 13831 glx: Inline some reply swapping code 13832 13833 Signed-off-by: Adam Jackson <ajax@redhat.com> 13834 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 13835 13836commit 51bab63b7353319c51ec2f1adaed95d86e4f6119 13837Author: Adam Jackson <ajax@redhat.com> 13838Date: Wed Aug 16 14:49:16 2017 -0400 13839 13840 glx: Remove True/False defines 13841 13842 Those are xlib spellings, we say TRUE/FALSE pretty consistently 13843 elsewhere in the server. 13844 13845 Signed-off-by: Adam Jackson <ajax@redhat.com> 13846 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 13847 Reviewed-by: Keith Packard <keithp@keithp.com> 13848 13849commit 8e3b26ceaa86eaf98a78f6b84f46a4a86aed8ef8 13850Author: Michel Dänzer <michel.daenzer@amd.com> 13851Date: Fri Apr 14 19:08:41 2017 +0900 13852 13853 Make PixmapDirtyUpdateRec::src a DrawablePtr 13854 13855 This allows making the master screen's pixmap_dirty_list entries 13856 explicitly reflect that we're now tracking the root window instead of 13857 the screen pixmap, in order to allow Present page flipping on master 13858 outputs while there are active slave outputs. 13859 13860 Define HAS_DIRTYTRACKING_DRAWABLE_SRC for drivers to check, but leave 13861 HAS_DIRTYTRACKING_ROTATION defined as well to make things slightly 13862 easier for drivers. 13863 13864 Reviewed-by: Adam Jackson <ajax@redhat.com> 13865 13866commit c52f77e4ca2cda878da341a6228e6411eec7b1a0 13867Author: Peter Harris <pharris@opentext.com> 13868Date: Mon Aug 14 15:54:36 2017 -0400 13869 13870 meson: Fix epoll detection 13871 13872 The epoll code depends on epoll_create1, not epoll_create. 13873 13874 Signed-off-by: Peter Harris <pharris@opentext.com> 13875 Signed-off-by: Eric Anholt <eric@anholt.net> 13876 Reviewed-by: Eric Anholt <eric@anholt.net> 13877 13878commit 27500ee82e97ef8a6b3199c2d7b623523c1ee2c1 13879Author: Eric Anholt <eric@anholt.net> 13880Date: Mon Jul 10 12:08:29 2017 -0700 13881 13882 glamor: Scissor Render composite operations to the bounds of the drawing. 13883 13884 Unlike the previous two fixes, this one introduces new GL calls and 13885 statechanges of the scissor. However, given that our Render drawing 13886 already does CPU side transformation and inefficient box upload, this 13887 shouldn't be a limiting factor for Render acceleration. 13888 13889 Surprisingly, it improves x11perf -comppixwin10 -repeat 1 -reps 10000 13890 on i965 by 3.21191% +/- 1.79977% (n=50). 13891 13892 v2: Make the jump to the exit land after scissor disable. 13893 13894 Signed-off-by: Eric Anholt <eric@anholt.net> 13895 Reviewed-by: Keith Packard <keithp@keithp.com> 13896 13897commit e6ab3b1109e72a1512c6b7b92dd84525ad8c8052 13898Author: Eric Anholt <eric@anholt.net> 13899Date: Thu Jul 6 16:15:17 2017 -0700 13900 13901 glamor: Scissor CopyArea to the bounds of the drawing. 13902 13903 Like the previous fix to rectangles, this reduces the area drawn on 13904 tiled renderers by letting the CPU-side tile setup know what tiles 13905 might be drawn at all. 13906 13907 Surprisingly, it improves x11perf -copypixwin1 -repeat 1 -reps 10000 13908 on i965 by 2.93185% +/- 1.5561% (n=90). 13909 13910 v2: Drop extra glamor_bounds_union_box() from previous debugging 13911 (caught by Mark Marshall). 13912 13913 Signed-off-by: Eric Anholt <eric@anholt.net> 13914 Reviewed-by: Keith Packard <keithp@keithp.com> (v1) 13915 13916commit 60cc7e367a2a5e6014f193105dafd47a4d598fd9 13917Author: Eric Anholt <eric@anholt.net> 13918Date: Thu Jul 6 15:43:14 2017 -0700 13919 13920 glamor: Scissor rectangle drawing to the bounds of the rects. 13921 13922 Scissors provide a critical hint to tiled renderers as to what tiles 13923 need to be load/stored because they could be modified by the 13924 rendering. 13925 13926 The bounds calculation here is limited to when we have a small number 13927 of rects (large enough to cover rounded window corners, but probably 13928 not xeyes) to avoid overhead on desktop GL. 13929 13930 No performance difference on i965 with x11perf -rect1 -repeat 1 -reps 13931 10000 (n=50) 13932 13933 v2: Clamp rectangle bounds addition. 13934 13935 Signed-off-by: Eric Anholt <eric@anholt.net> 13936 Reviewed-by: Keith Packard <keithp@keithp.com> 13937 13938commit f2110157713cf22c8b5c46a0d5416fdb033c27d3 13939Author: Eric Anholt <eric@anholt.net> 13940Date: Tue May 9 15:48:27 2017 -0700 13941 13942 test: Remove unused ddxstubs.c 13943 13944 Signed-off-by: Eric Anholt <eric@anholt.net> 13945 Reviewed-by: Keith Packard <keithp@keithp.com> 13946 13947commit f80119120c487581ac050ce741808f7c8f438f35 13948Author: Adam Jackson <ajax@redhat.com> 13949Date: Thu Jul 27 16:02:28 2017 -0400 13950 13951 composite: Make compIsAlternateVisual safe even if Composite is off 13952 13953 As of ea483af9 we're calling this unconditionally from the GLX code so 13954 the synthetic visual is in a lower select group. If Composite has been 13955 disabled then GetCompScreen() will return NULL, and this would crash. 13956 13957 Rather than force the caller to check first, just always return FALSE if 13958 Composite is disabled (which is correct, since none of the visuals will 13959 be synthetic in that case). 13960 13961 Signed-off-by: Adam Jackson <ajax@redhat.com> 13962 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 13963 13964commit b06a85e7fd6ddac58dd97e5d7ce3b5ea9553c8f6 13965Author: Emil Velikov <emil.velikov@collabora.com> 13966Date: Mon Jul 31 14:13:48 2017 +0100 13967 13968 glx: remove unused GlxSetVisualConfigs stub 13969 13970 The function was an empty since 2008 at least. Drop it since no 13971 drivers use it any more. 13972 13973 Reviewed-by: Adam Jackson <ajax@redhat.com> 13974 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 13975 13976commit e8f190b445ae10b12896ee396cd8d57f6b6b8cc4 13977Author: Eric Anholt <eric@anholt.net> 13978Date: Fri Aug 4 13:19:01 2017 -0700 13979 13980 travis: Request a new docker image with xkb-data and xkbcomp installed. 13981 13982 Xvfb was erroring out with: 13983 13984 XKB: Failed to compile keymap 13985 Keyboard initialization failed. This could be a missing or incorrect 13986 setup of xkeyboard-config. 13987 (EE) Fatal server error: 13988 (EE) Failed to activate virtual core keyboard: 2(EE) 13989 13990 With this change, we can now run my xsync regression test on Travis. 13991 13992 Reviewed-by: Adam Jackson <ajax@redhat.com> 13993 Signed-off-by: Eric Anholt <eric@anholt.net> 13994 13995commit b62e486c4e3d7411f3e70b610af7f8b9dd5e8a73 13996Author: Eric Anholt <eric@anholt.net> 13997Date: Fri Aug 4 13:19:00 2017 -0700 13998 13999 travis: Set prefix to /usr, so we find xkbcomp. 14000 14001 Reviewed-by: Adam Jackson <ajax@redhat.com> 14002 Signed-off-by: Eric Anholt <eric@anholt.net> 14003 14004commit 2d50e32e0ba1045f8c73a14335743ae98121d102 14005Author: Eric Anholt <eric@anholt.net> 14006Date: Tue Aug 1 13:11:25 2017 -0700 14007 14008 meson: Fix xwayland build since xwayland-keyboard-grab. 14009 14010 The version detect was erroring out with 1.9 protos installed, and we 14011 weren't building the new code. 14012 14013 Signed-off-by: Eric Anholt <eric@anholt.net> 14014 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14015 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14016 14017commit 0a448d133f4f1c913b1c2cb05accff31c74a3dbf 14018Author: Olivier Fourdan <ofourdan@redhat.com> 14019Date: Wed Jul 12 11:51:08 2017 +0200 14020 14021 xwayland: Add grab protocol support 14022 14023 The keyboard grabbing protocol for Xwayland is included in 14024 wayland-protocol 1.9. 14025 14026 Update the wayland-protocol required version in both configure and meson 14027 builds and add support for this new protocol in Xwayland. 14028 14029 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 14030 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14031 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14032 14033commit abe49b009064c38823fac17c373fd5f1a390b3ab 14034Author: Jan Beich <jbeich@freebsd.org> 14035Date: Fri Jul 7 19:59:05 2017 +0200 14036 14037 dix: unbreak --with-dtrace (default) on FreeBSD 11.1+ 14038 14039 gmake[2]: Entering directory '/path/to/xserver/dix' 14040 /usr/sbin/dtrace -G -C -o dtrace-dix.o -s ../dix/Xserver.d .libs/atom.o ... 14041 dtrace: failed to compile script ../dix/Xserver.d: line 26: useless declaration 14042 gmake[2]: *** [Makefile:1007: dtrace-dix.o] Error 1 14043 14044 Signed-off-by: Jan Beich <jbeich@FreeBSD.org> 14045 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 14046 14047commit c967e4cd8ba50034b6e02d87df10d04edf15ccf6 14048Author: Emil Velikov <emil.velikov@collabora.com> 14049Date: Mon Jul 31 14:06:48 2017 +0100 14050 14051 glx: don't export __glXDRISWRastProvider 14052 14053 The symbol is used only internally and is not part of the API/ABI. 14054 14055 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 14056 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 14057 14058commit c09e906d61a0855a8202e58ba72ed6bb3b08f37f 14059Author: Emil Velikov <emil.velikov@collabora.com> 14060Date: Mon Jul 31 14:06:47 2017 +0100 14061 14062 glx: fix typo becuase -> because 14063 14064 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 14065 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 14066 14067commit db0dd06ddf47d5190f2349ed8b0e71caa9e8e1eb 14068Author: Adam Jackson <ajax@redhat.com> 14069Date: Thu Jul 27 16:02:33 2017 -0400 14070 14071 glx: Allow arbitrary context attributes for direct contexts 14072 14073 For direct contexts, most context attributes don't require any 14074 particular awareness on the part of the server. Examples include 14075 GLX_ARB_create_context_no_error and GLX_ARB_context_flush_control, where 14076 all of the behavior change lives in the renderer; since that's on the 14077 client side for a direct context, there's no reason for the X server to 14078 validate the attribute. 14079 14080 The context attributes will still be validated on the client side, and 14081 we still validate attributes for indirect contexts since the server 14082 implementation might need to handle them. For example, the indirect 14083 code might internally use ARB_context_flush_control for all contexts, in 14084 which case it would need to manually emit glFlush when the client 14085 switches between two indirect contexts that didn't request the no-flush 14086 attribute. 14087 14088 Signed-off-by: Adam Jackson <ajax@redhat.com> 14089 Reviewed-by: Eric Anholt <eric@anholt.net> 14090 14091commit 3050d277616ef472a24d8ccd24afcbf8e31ed80e 14092Author: Adam Jackson <ajax@redhat.com> 14093Date: Tue Jun 27 11:18:48 2017 -0400 14094 14095 xfree86: Fix X -configure driver sort yet again 14096 14097 There were two bugs here: The comparison function was not stable when 14098 one or more of the drivers being compared is a fallback, and the last 14099 driver in the list would never be moved. 14100 14101 Signed-off-by: Adam Jackson <ajax@redhat.com> 14102 14103commit 6f9939525c31f1f2d8d6c137a46404278384b4bf 14104Author: Adam Jackson <ajax@redhat.com> 14105Date: Mon Jun 26 12:43:54 2017 -0400 14106 14107 modesetting: Fix PCI initialization on non-zero domains 14108 14109 libdrm's busid matching for the legacy three-integer bus string format 14110 simply ignores the domain number, rather than what we were doing here of 14111 packing the domain into the bus number. Whatever, just use the existing 14112 code to build a busid string, since that gets the domain right. 14113 14114 Signed-off-by: Adam Jackson <ajax@redhat.com> 14115 14116commit d16133e3e293fe16b291114d344fbd5a61d5ed65 14117Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 14118Date: Thu Jul 20 19:48:26 2017 -0700 14119 14120 glx: Only include compositeext.h if COMPOSITE is supported 14121 14122 Regressed-in: ea483af99a6351323afe00a0b630cd63310efdb1 14123 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 14124 14125commit dd0b721c977e88d7c3830cc2a672c2793bca4fa0 14126Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 14127Date: Tue Jul 18 21:35:49 2017 -0700 14128 14129 XQuartz: Hack around an issue that can occur on macOS due to background apps incorrectly stealing focus 14130 14131 Works around <rdar://problem/7150340>. 14132 14133 Tested-by: Martin Otte <martinjotte@gmail.com> 14134 Tested-by: Tom Lane <tgl@sss.pgh.pa.us> 14135 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 14136 14137commit 211e05ac85a294ef361b9f80d689047fa52b9076 14138Author: Michal Srb <msrb@suse.com> 14139Date: Fri Jul 7 17:21:46 2017 +0200 14140 14141 Xi: Test exact size of XIBarrierReleasePointer 14142 14143 Otherwise a client can send any value of num_barriers and cause reading or swapping of values on heap behind the receive buffer. 14144 14145 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14146 14147commit abb031e731f5c159add1b3351de9c4bb121bf00a 14148Author: Rodrigo Vivi <rodrigo.vivi@intel.com> 14149Date: Thu Jun 29 13:29:58 2017 -0700 14150 14151 dri2: Sync i965_pci_ids.h from Mesa. 14152 14153 Copied from Mesa with no modifications. 14154 14155 Gives us Coffee Lake and Cannon Lake PCI IDs. 14156 14157 Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com> 14158 Acked-by: Kenneth Graunke <kenneth@whitecape.org> 14159 14160commit 05a4396f08c83d9261b8ad16908d9e2730bd4bbf 14161Author: Emil Velikov <emil.velikov@collabora.com> 14162Date: Thu Jun 29 02:17:06 2017 +0100 14163 14164 glamor: update "required EGL extensions" comment 14165 14166 The extensions listed have not been needed in a while. Replace with the 14167 only remaining requirement. 14168 14169 Reviewed-by: Adam Jackson <ajax@redhat.com> 14170 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 14171 14172commit 04511a0476b5c860e7d157b01080dff94d935f74 14173Author: Adam Jackson <ajax@redhat.com> 14174Date: Thu Jun 29 10:32:00 2017 -0400 14175 14176 wayland: Sync drm.xml with Mesa 14177 14178 ... where it is named src/egl/wayland/wayland-drm/wayland-drm.xml and 14179 has its requests sorted by protocol version number, avoiding a warning 14180 from wayland-scanner. 14181 14182 Signed-off-by: Adam Jackson <ajax@redhat.com> 14183 Reviewed-by: Daniel Stone <daniels@collabora.com> 14184 14185commit ce393de0efb8626d15f3b97c97916971a6aefebd 14186Author: Dave Airlie <airlied@redhat.com> 14187Date: Fri Jun 23 09:29:13 2017 +1000 14188 14189 modesetting: handle NULL cursor in drmmode_set_cursor. 14190 14191 We had a bug reported with a touchscreen where we could end up 14192 in here with a NULL cursor, so let's not crash the X server. 14193 14194 Signed-off-by: Dave Airlie <airlied@redhat.com> 14195 Reviewed-and-Tested-by: Daniel Martin <consume.noise@gmail.com> 14196 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14197 14198commit 43527d31810fe94a5eb4b2543bbc8eb26b1654b7 14199Author: Peter Hutterer <peter.hutterer@who-t.net> 14200Date: Tue Jul 4 06:51:58 2017 +1000 14201 14202 xfree86: link to libinput(4) from the xorg.conf man page 14203 14204 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14205 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 14206 14207commit edaad0eff48219787d27f7723cd3d4d9b293de96 14208Author: Adam Jackson <ajax@redhat.com> 14209Date: Thu Jun 29 13:57:37 2017 -0400 14210 14211 automake: Add hw/xfree86/xkb/meson.build to EXTRA_DIST 14212 14213 Signed-off-by: Adam Jackson <ajax@redhat.com> 14214 14215commit 06419a5b17180126d2507c8699229e07f5b2c434 14216Author: Emil Velikov <emil.velikov@collabora.com> 14217Date: Thu Jun 29 01:55:25 2017 +0100 14218 14219 glamor: remove no longer needed KHR_gl_texture_2D_image requirement 14220 14221 The code that needed it was introduced with commit 7cfd9cc2327 ("Add 14222 DRI3 support to glamor") back in 2013. And was nuked a couple of years 14223 ago with commit 51984dddfcc ("glamor: Delay making pixmaps shareable 14224 until we need to.") 14225 14226 Signed-off-by: Emil Velikov <emil.velikov@collabora.com> 14227 Reviewed-by: Eric Anholt <eric@anholt.net> 14228 14229commit 47a7b63f7455eaeacb634fd800e924f2087dacbd 14230Author: Jon TURNEY <jon.turney@dronecode.org.uk> 14231Date: Mon Jun 26 14:54:06 2017 +0100 14232 14233 meson: Use --export-all-symbols when building PE/COFF objects 14234 14235 Reviewed-by: Adam Jackson <ajax@redhat.com> 14236 14237commit b34abb3d2da46339d59a0feefd2240790a6a6a0e 14238Author: Jon TURNEY <jon.turney@dronecode.org.uk> 14239Date: Mon Jun 26 14:54:05 2017 +0100 14240 14241 meson: Shuffle around subdirs so we build Xorg loadable modules after Xorg 14242 14243 Reviewed-by: Adam Jackson <ajax@redhat.com> 14244 14245commit fbdd73fac68383c93f6f5c6a7615860503039999 14246Author: Jon TURNEY <jon.turney@dronecode.org.uk> 14247Date: Mon Jun 26 14:54:04 2017 +0100 14248 14249 Move statically linked xorgxkb files from dixmods to a separate directory 14250 14251 [ajax: Fixed test/Makefile.am as well] 14252 14253 Reviewed-by: Adam Jackson <ajax@redhat.com> 14254 14255commit 5c8fb7d128e312e6c01cd40c7ada072c4b5f3776 14256Author: Adam Jackson <ajax@redhat.com> 14257Date: Wed Jun 21 15:58:43 2017 -0400 14258 14259 xfree86: Move DRICreatePCIBusID to xf86Pci.h 14260 14261 xf86str.h is parsed into sdksyms unconditionally but the symbol is only 14262 defined when building with PCI support. Move the decl to a header that 14263 sdksyms only parses when building PCI support. 14264 14265 Signed-off-by: Adam Jackson <ajax@redhat.com> 14266 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 14267 14268commit 525db17e04d5cabaeefd20e38cebe7d606047a76 14269Author: Adam Jackson <ajax@redhat.com> 14270Date: Fri Jun 16 15:44:52 2017 -0400 14271 14272 glx/drisw: Use scratch GCs for swrastPutImage 14273 14274 The scratch GC defaults to the same state as our persistent GCs. Except 14275 for the "draw" GC, which would generate graphics exposures for... well, 14276 no reason really, PutImage doesn't generate graphics exposures. 14277 14278 Signed-off-by: Adam Jackson <ajax@redhat.com> 14279 Reviewed-by: Keith Packard <keithp@keithp.com> 14280 14281commit ecc5e362985d1f649c29ac5f93be19851eaf7d87 14282Author: Adam Jackson <ajax@redhat.com> 14283Date: Fri Jun 16 15:44:51 2017 -0400 14284 14285 glx/drisw: Remove unused glx_enable_bits 14286 14287 Sloppy of me! 14288 14289 Signed-off-by: Adam Jackson <ajax@redhat.com> 14290 Acked-by: Keith Packard <keithp@keithp.com> 14291 14292commit a6c23ef31a4e5aea19ddabdf4a0760f4c48f1ebe 14293Author: Adam Jackson <ajax@redhat.com> 14294Date: Fri Jun 16 15:44:50 2017 -0400 14295 14296 glx/dri2: Don't chirp when falling back to software 14297 14298 This isn't an error if the screen isn't accelerated in the first place. 14299 14300 Signed-off-by: Adam Jackson <ajax@redhat.com> 14301 Acked-by: Keith Packard <keithp@keithp.com> 14302 14303commit 0b1831d043028f7dd6accca19a81e2abd9a145b5 14304Author: Adam Jackson <ajax@redhat.com> 14305Date: Fri Jun 16 15:44:49 2017 -0400 14306 14307 glx: Remove some indirection around EXT_texture_from_pixmap 14308 14309 Signed-off-by: Adam Jackson <ajax@redhat.com> 14310 Reviewed-by: Keith Packard <keithp@keithp.com> 14311 14312commit c33541e59e338cd53f89d94a6898555d350e84aa 14313Author: Adam Jackson <ajax@redhat.com> 14314Date: Fri Jun 16 15:44:48 2017 -0400 14315 14316 glx: Remove __glXReply 14317 14318 Static data bad, hulk smash. 14319 14320 Signed-off-by: Adam Jackson <ajax@redhat.com> 14321 Reviewed-by: Keith Packard <keithp@keithp.com> 14322 14323commit d6db66811643d3762716f6b144a7358572216a4f 14324Author: Adam Jackson <ajax@redhat.com> 14325Date: Fri Jun 16 15:44:47 2017 -0400 14326 14327 dmx: Remove some not-very-interesting debug prints 14328 14329 gcc/glibc think the snprintf in dmxExecOS() might truncate. Yes, it 14330 might, and we also don't care. Just delete all this. 14331 14332 Signed-off-by: Adam Jackson <ajax@redhat.com> 14333 Acked-by: Keith Packard <keithp@keithp.com> 14334 14335commit 17ad6e5d5616039021455bc821d6ee2497f7ebde 14336Author: Adam Jackson <ajax@redhat.com> 14337Date: Fri Jun 16 15:44:46 2017 -0400 14338 14339 dmx: Silence an unused-result warning 14340 14341 Modern glibc is very insistent that you care about whether write() 14342 succeeds: 14343 14344 ../hw/dmx/input/usb-keyboard.c: In function ‘kbdUSBCtrl’: 14345 ../hw/dmx/input/usb-keyboard.c:292:9: warning: ignoring return value of ‘write’, declared with attribute warn_unused_result [-Wunused-result] 14346 write(priv->fd, &event, sizeof(event)); 14347 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 14348 14349 Signed-off-by: Adam Jackson <ajax@redhat.com> 14350 Reviewed-by: Keith Packard <keithp@keithp.com> 14351 14352commit fbc4da6fef2a0d2111284a19c26d83686ebcbd1f 14353Author: Adam Jackson <ajax@redhat.com> 14354Date: Fri Jun 16 15:44:44 2017 -0400 14355 14356 xfree86: Move DRICreatePCIBusID to the PCI code 14357 14358 This symbol is used by some DRI2+ drivers and there's nothing 14359 DRI1-specific about it. 14360 14361 Signed-off-by: Adam Jackson <ajax@redhat.com> 14362 Acked-by: Keith Packard <keithp@keithp.com> 14363 14364commit b723da8390d2fe1bf048247302416b37972a0cc3 14365Author: Adam Jackson <ajax@redhat.com> 14366Date: Fri Jun 16 15:44:43 2017 -0400 14367 14368 xfree86: Remove unused xf86EnableVTSwitch 14369 14370 Signed-off-by: Adam Jackson <ajax@redhat.com> 14371 Acked-by: Keith Packard <keithp@keithp.com> 14372 14373commit f44e0af4daaad5b18d79f2076bc98d6e79f638af 14374Author: Adam Jackson <ajax@redhat.com> 14375Date: Fri Jun 16 15:44:42 2017 -0400 14376 14377 dix: Remove extension aliases 14378 14379 This appears to be essentially unused. The only known client-side 14380 library for the SELinux extension is xcb, which does not look for the 14381 name "Flask". The "SGI-GLX" alias for GLX appears to be a bit of 14382 superstition at this point, NVIDIA's driver does not expose it and Mesa 14383 does not check for it. 14384 14385 Signed-off-by: Adam Jackson <ajax@redhat.com> 14386 Acked-by: Keith Packard <keithp@keithp.com> 14387 14388commit 88e807d4fa90165868929190f601beef18252f57 14389Author: Adam Jackson <ajax@redhat.com> 14390Date: Fri Jun 16 15:44:41 2017 -0400 14391 14392 include: Remove some unused macros 14393 14394 Signed-off-by: Adam Jackson <ajax@redhat.com> 14395 Reviewed-by: Keith Packard <keithp@keithp.com> 14396 14397commit d2e57f71ad16f1e677dd542a8bc0db87dfa9791d 14398Author: Michel Dänzer <michel.daenzer@amd.com> 14399Date: Mon Jun 19 19:05:32 2017 +0900 14400 14401 xfree86: Document BusID PCI domain format in ddxDesign.xml 14402 14403 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 14404 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 14405 14406commit 3b0fd9421add9f0ca8958ced0f39b2da21143c22 14407Author: Michel Dänzer <michel.daenzer@amd.com> 14408Date: Mon Jun 19 19:05:31 2017 +0900 14409 14410 xfree86: Document BusID PCI domain format in xorg.conf manpage 14411 14412 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 14413 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 14414 14415commit 8e00dc59b51da7e0d3978c26b1884c4767aa2dc2 14416Author: Michel Dänzer <michel.daenzer@amd.com> 14417Date: Mon Jun 19 19:05:30 2017 +0900 14418 14419 xfree86: Fix printing of PCI domain/bus in xf86MatchPciInstances 14420 14421 It was attempting to use the <bus>@<domain> format accepted by the BusID 14422 stanza, but the two values were swapped. 14423 14424 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 14425 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 14426 14427commit e905b19a53f96013c4417bec993a1dea5a3b0a5f 14428Author: Michel Dänzer <michel.daenzer@amd.com> 14429Date: Mon Jun 19 19:05:29 2017 +0900 14430 14431 xfree86: Print BusID stanza compatible bus IDs for found devices 14432 14433 The PCI domain has to be specified like this: 14434 14435 "PCI:<bus>@<domain>:<device>:<function>" 14436 14437 Example before: 14438 14439 (--) PCI:*(0:0:1:0) 1002:130f:1043:85cb [...] 14440 (--) PCI: (0:1:0:0) 1002:6939:1458:229d [...] 14441 14442 after: 14443 14444 (--) PCI:*(0@0:1:0) 1002:130f:1043:85cb [...] 14445 (--) PCI: (1@0:0:0) 1002:6939:1458:229d [...] 14446 14447 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 14448 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 14449 14450commit fb212f2c1bcfb015112aa665dd1f25d9b80f86fc 14451Author: Jon Turney <jon.turney@dronecode.org.uk> 14452Date: Sat Jun 17 14:32:30 2017 +0100 14453 14454 meson: Make VBE and VGAHW modules optional 14455 14456 Don't build them on platforms where they aren't meaningful. 14457 14458 Note that autoconf defines WITH_VGAHW when building the VGAHW module, but 14459 that doesn't seem to be used anywhere, so we just drop that. 14460 14461 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 14462 Reviewed-by: Eric Anholt <eric@anholt.net> 14463 14464commit 6d2b7d33678e7a881ebdb5ceb4a32509ad3ee930 14465Author: Jon Turney <jon.turney@dronecode.org.uk> 14466Date: Thu May 4 14:08:48 2017 +0100 14467 14468 meson: Restore stub ossupport 14469 14470 Don't build BSD ossupport when there is no specific support, build stubs 14471 14472 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 14473 Reviewed-by: Eric Anholt <eric@anholt.net> 14474 14475commit 4212c884c423e5ce2cd3b4d67c0d656475fddc79 14476Author: Michel Dänzer <michel.daenzer@amd.com> 14477Date: Fri Jun 16 11:30:03 2017 +0900 14478 14479 xfree86/modes: Use RRTransformEqual in xf86RandR12CrtcSet 14480 14481 The memcmp didn't catch when e.g. only the filter changed. Tested by 14482 alternately running 14483 14484 xrandr --output DVI-I-0 --scale-from 3840x2160 --filter bilinear 14485 xrandr --output DVI-I-0 --scale-from 3840x2160 --filter nearest 14486 14487 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 14488 14489commit 091af80be48c37f16c679d35fc12ad33e6b0cd74 14490Author: Aaron Plattner <aplattner@nvidia.com> 14491Date: Thu Jun 15 14:28:27 2017 -0700 14492 14493 randr: Use RRTransformEqual in RRCrtcPendingTransform 14494 14495 Currently, RRCrtcPendingTransform returns false unless the 14496 transformation matrix itself is changing. This makes RRCrtcSet skip 14497 doing anything if the only thing that is changing is the transform 14498 filter. 14499 14500 There's already a function for comparing RRTransformPtrs, so use that 14501 instead. 14502 14503 Tested by running 14504 14505 xrandr --output DP-1 --mode 1920x1080 --rate 144 --scale 0.5x0.5 --filter nearest 14506 14507 follwed by 14508 14509 xrandr --output DP-1 --mode 1920x1080 --rate 144 --scale 0.5x0.5 --filter bilinear 14510 14511 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 14512 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 14513 14514commit ba336b24052122b136486961c82deac76bbde455 14515Author: Michal Srb <msrb@suse.com> 14516Date: Wed May 24 15:54:42 2017 +0300 14517 14518 Xi: Do not try to swap GenericEvent. 14519 14520 The SProcXSendExtensionEvent must not attempt to swap GenericEvent because 14521 it is assuming that the event has fixed size and gives the swapping function 14522 xEvent-sized buffer. 14523 14524 A GenericEvent would be later rejected by ProcXSendExtensionEvent anyway. 14525 14526 Signed-off-by: Michal Srb <msrb@suse.com> 14527 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14528 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14529 14530commit 8caed4df36b1f802b4992edcfd282cbeeec35d9d 14531Author: Michal Srb <msrb@suse.com> 14532Date: Wed May 24 15:54:41 2017 +0300 14533 14534 Xi: Verify all events in ProcXSendExtensionEvent. 14535 14536 The requirement is that events have type in range 14537 EXTENSION_EVENT_BASE..lastEvent, but it was tested 14538 only for first event of all. 14539 14540 Signed-off-by: Michal Srb <msrb@suse.com> 14541 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14542 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14543 14544commit 215f894965df5fb0bb45b107d84524e700d2073c 14545Author: Michal Srb <msrb@suse.com> 14546Date: Wed May 24 15:54:40 2017 +0300 14547 14548 dix: Disallow GenericEvent in SendEvent request. 14549 14550 The SendEvent request holds xEvent which is exactly 32 bytes long, no more, 14551 no less. Both ProcSendEvent and SProcSendEvent verify that the received data 14552 exactly match the request size. However nothing stops the client from passing 14553 in event with xEvent::type = GenericEvent and any value of 14554 xGenericEvent::length. 14555 14556 In the case of ProcSendEvent, the event will be eventually passed to 14557 WriteEventsToClient which will see that it is Generic event and copy the 14558 arbitrary length from the receive buffer (and possibly past it) and send it to 14559 the other client. This allows clients to copy unitialized heap memory out of X 14560 server or to crash it. 14561 14562 In case of SProcSendEvent, it will attempt to swap the incoming event by 14563 calling a swapping function from the EventSwapVector array. The swapped event 14564 is written to target buffer, which in this case is local xEvent variable. The 14565 xEvent variable is 32 bytes long, but the swapping functions for GenericEvents 14566 expect that the target buffer has size matching the size of the source 14567 GenericEvent. This allows clients to cause stack buffer overflows. 14568 14569 Signed-off-by: Michal Srb <msrb@suse.com> 14570 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14571 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14572 14573commit 05442de962d3dc624f79fc1a00eca3ffc5489ced 14574Author: Michal Srb <msrb@suse.com> 14575Date: Wed May 24 15:54:39 2017 +0300 14576 14577 Xi: Zero target buffer in SProcXSendExtensionEvent. 14578 14579 Make sure that the xEvent eventT is initialized with zeros, the same way as 14580 in SProcSendEvent. 14581 14582 Some event swapping functions do not overwrite all 32 bytes of xEvent 14583 structure, for example XSecurityAuthorizationRevoked. Two cooperating 14584 clients, one swapped and the other not, can send 14585 XSecurityAuthorizationRevoked event to each other to retrieve old stack data 14586 from X server. This can be potentialy misused to go around ASLR or 14587 stack-protector. 14588 14589 Signed-off-by: Michal Srb <msrb@suse.com> 14590 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14591 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14592 14593commit d82c3cee02a99cf7861d1effaa5c7d38683a7783 14594Author: Eric Anholt <eric@anholt.net> 14595Date: Thu Jun 1 13:59:54 2017 -0700 14596 14597 travis: Enable ccache. 14598 14599 We bind-mount the cache directory into the container. Cuts build time 14600 from about 4 minutes to 2. 14601 14602 Signed-off-by: Eric Anholt <eric@anholt.net> 14603 14604commit 01dfb65ee2e34148718d3f95dfd4629024edba84 14605Author: Eric Anholt <eric@anholt.net> 14606Date: Thu Jun 1 12:29:12 2017 -0700 14607 14608 travis: Add a build script for Travis CI. 14609 14610 This currently does a meson build using a docker image I've prepared. 14611 The Dockerfile source is at: 14612 14613 https://github.com/anholt/xserver-travis 14614 14615 Docker proved to be necessary to cut the build time per Travis push. 14616 If some day we end up using meson in more of the X stack, we may be 14617 able to move more dependencies out of the docker image and into the CI 14618 build (putting the I in CI). Until then, we'll have to do docker 14619 image rebuilds when dependencies are added/updated. 14620 14621 To enable Travis CI on your github repository, see the first two steps 14622 of the docs at: 14623 14624 https://docs.travis-ci.com/user/getting-started/ 14625 14626 Once you do that, pushing branches to your github repo will trigger 14627 builds, which will send you email if they fail. Current build status 14628 can be veiewed your account on travis-ci.org: 14629 14630 https://travis-ci.org/anholt/xserver 14631 14632commit fda2b064a0ebd223e8a39fd19f5c5ad1da0776e8 14633Author: Eric Anholt <eric@anholt.net> 14634Date: Thu Jun 1 12:29:11 2017 -0700 14635 14636 meson: Fix test for whether we've built glamor-EGL. 14637 14638 This matches the test we use for going into the glamor_egl subdir in 14639 ../../meson.build. 14640 14641commit 75bf302ed20f4b9794f71f84c1da2b9fcd689356 14642Author: Eric Anholt <eric@anholt.net> 14643Date: Fri Jun 2 09:12:41 2017 -0700 14644 14645 dmx: Remove includes of glxtokens.h 14646 14647 Our top-level glx.h include already provides all of the tokens we use, 14648 and fixes redefinition warnings in the meson build. 14649 14650 Reviewed-by: Adam Jackson <ajax@redhat.com> 14651 Signed-off-by: Eric Anholt <eric@anholt.net> 14652 14653commit f3689f637f5ac0fb6c231a470e65b39aa5e9ba20 14654Author: Keith Packard <keithp@keithp.com> 14655Date: Wed May 17 09:57:29 2017 -0700 14656 14657 os: Set oc->fd to -1 when connection is closed 14658 14659 This ensures that we don't use the now-closed file descriptor in the 14660 future. 14661 14662 Reviewed-by: Adam Jackson <ajax@redhat.com> 14663 Signed-off-by: Keith Packard <keithp@keithp.com> 14664 14665commit d05c754e1bde895589fb514d8f518afeccecbc05 14666Author: Keith Packard <keithp@keithp.com> 14667Date: Wed May 17 09:57:28 2017 -0700 14668 14669 os: Check oc->trans_conn before using oc->fd in YieldControlNoInput 14670 14671 oc->trans_conn is set to NULL when the connection is closed. At this 14672 point, oc->fd is no longer valid and shouldn't be used. Move 14673 dereference of oc->fd up into YieldControlNoInput where the state of 14674 oc->trans_conn can be checked in a single place. 14675 14676 Reviewed-by: Adam Jackson <ajax@redhat.com> 14677 Signed-off-by: Keith Packard <keithp@keithp.com> 14678 14679commit 448a5586e9235bee9648d89e4103ed48e6237c15 14680Author: Keith Packard <keithp@keithp.com> 14681Date: Wed May 17 09:57:27 2017 -0700 14682 14683 os: Don't call ospoll_listen/ospoll_mute after connection is closed 14684 14685 In set_poll_client, check oc->trans_conn to make sure the connection 14686 is still running before changing the ospoll configuration of the file 14687 descriptor in case some other bit of the server is now using this file 14688 descriptor. 14689 14690 Reviewed-by: Adam Jackson <ajax@redhat.com> 14691 Signed-off-by: Keith Packard <keithp@keithp.com> 14692 14693commit 523d35e3e1c703a655386f6348a4bfb4291c3969 14694Author: Keith Packard <keithp@keithp.com> 14695Date: Wed May 17 09:57:26 2017 -0700 14696 14697 os: Use CloseDownFileDescriptor from AbortClient, including ospoll_remove 14698 14699 AbortClient performs most of the same operations as 14700 CloseDownFileDescriptor except that it doesn't call ospoll_remove, 14701 leaving that unaware that the file descriptor has been closed. 14702 14703 If the file descriptor is re-used before the server comes back around 14704 to clean up, and that new file descriptor is passed to SetNotifyFd, 14705 then that function will mistakenly re-interpret the stale ClientPtr 14706 returned by ospoll_data as a struct notify * instead and mangle data 14707 badly. 14708 14709 To fix this, the patch does: 14710 14711 1) Change CloseDownFileDescriptor so that it can be called multiple 14712 times on the same OsCommPtr. The calls related to the file 14713 descriptor are moved inside the check for trans_conn and 14714 oc->trans_conn is set to NULL after cleaning up. 14715 14716 2) Move the XdmcpCloseDisplay call into CloseDownFileDescriptor. I 14717 don't think the actually matters as we just need to know at some 14718 point that the session client has exited. Moving it avoids the 14719 possibility of having this accidentally trigger from another client 14720 with the same fd which closes down at around the same time. 14721 14722 3) Change AbortClient to call CloseDownFileDescriptor. This makes sure 14723 that all of the fd-related clean up happens in the same way 14724 everywhere, in particular ensures that ospoll is notified about the 14725 closed file descriptor at the time it is closed and not some time later. 14726 14727 Debian-bug: https://bugs.debian.org/862824 14728 Reviewed-by: Adam Jackson <ajax@redhat.com> 14729 Signed-off-by: Keith Packard <keithp@keithp.com> 14730 14731commit 5d941ccb0b30399d505b48bff894c95cc3023bbe 14732Author: Keith Packard <keithp@keithp.com> 14733Date: Wed May 17 09:57:25 2017 -0700 14734 14735 os: Eliminate ConnectionTranslation 14736 14737 This infrastructure is no longer read, only written; the mapping 14738 from fd to client is now handled by ospoll. 14739 14740 Reviewed-by: Adam Jackson <ajax@redhat.com> 14741 Signed-off-by: Keith Packard <keithp@keithp.com> 14742 14743commit c86fc56b10b603b41ae37057eedfa9c86b609752 14744Author: Keith Packard <keithp@keithp.com> 14745Date: Mon Jun 12 17:46:51 2017 -0700 14746 14747 glamor: Clarify variable names in glamor_copy_cpu_fbo 14748 14749 This function creates a temporary pixmap to hold data being moved from 14750 the source to the destination. However, it labeled all of the 14751 variables associated with this as src_, which makes it confusing to 14752 read the code. Rename them tmp_ instead. Also fix the comment 14753 describing the function to note that it copies from CPU to GPU, not 14754 GPU to GPU. 14755 14756 Signed-off-by: Keith Packard <keithp@keithp.com> 14757 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 14758 14759commit ffda82ed04d28feae2e001dbd0c32d6c795d90b1 14760Author: Michel Dänzer <michel.daenzer@amd.com> 14761Date: Tue Jun 6 18:42:06 2017 +0900 14762 14763 glamor: Fix temporary pixmap coordinate offsets 14764 14765 The previous values happened to work in basic cases, but not in general 14766 if the destination is a subwindow or has a border. 14767 14768 Fixes crash with xli, which moves a large subwindow inside a smaller 14769 parent window for scrolling. 14770 14771 No regressions with xterm, x11perf -copyplane or the xscreensaver 14772 phosphor hack. 14773 14774 Bug: https://bugs.debian.org/857983 14775 Reviewed-by: Keith Packard <keithp@keithp.com> 14776 14777commit d4995a3936ae283b9080fdaa0905daa669ebacfc 14778Author: Adam Jackson <ajax@redhat.com> 14779Date: Mon Jun 12 14:43:23 2017 -0400 14780 14781 modesetting: Validate the atom for enum properties 14782 14783 The client could have said anything here, and if what they said doesn't 14784 actually name an atom NameForAtom() will return NULL, and strcmp() will 14785 be unhappy about that. 14786 14787 Signed-off-by: Adam Jackson <ajax@redhat.com> 14788 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14789 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14790 14791commit 7c7a540f1e1d6b5466e1c9aa28476a2d7273d5ed 14792Author: Jason Gerecke <killertofu@gmail.com> 14793Date: Fri Jun 9 16:02:07 2017 -0700 14794 14795 xwayland: Implement tablet_tool_wheel for scrolling 14796 14797 The 'tablet_tool_wheel' function for tablet scrolling was added back in 14798 8a1defcc634 but left unimplemented. This commit fills in the necessary 14799 details, using the "clicks" count as the number of discrete scroll up/down 14800 events to send. 14801 14802 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 14803 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14804 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14805 14806commit fbc9814975fe82be25becf1a55d4f8d34298a956 14807Author: Jason Gerecke <killertofu@gmail.com> 14808Date: Fri Jun 9 16:02:06 2017 -0700 14809 14810 xwayland: Correct off-by-one error in tablet button numbering 14811 14812 The 'tablet_tool_frame' function treats the button masks as though they 14813 are zero-indexed, but 'tablet_tool_button_state' treats them as one- 14814 indexed. The result is that an e.g. middle click event recieved from 14815 Wayland will be sent from the X server as a right-click instead. 14816 14817 Fixes: 773b04748d0 ("xwayland: handle button events after motion events") 14818 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 14819 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14820 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14821 14822commit d5e2f271ad93e50412ff3605fb25cb9622f437e0 14823Author: Carlos Garnacho <carlosg@gnome.org> 14824Date: Sun May 28 15:56:22 2017 +0200 14825 14826 xwayland: Remove two unused proc pointers. 14827 14828 Xwayland doesn't override these, so we don't need defining those 14829 in the xwl_screen struct. 14830 14831 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 14832 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14833 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14834 14835commit 95febc42cadf392a888104ad6d5cf4f34fdde7d5 14836Author: Carlos Garnacho <carlosg@gnome.org> 14837Date: Sun May 28 15:56:21 2017 +0200 14838 14839 Xi: Use WarpPointerProc hook on XI pointer warping implementation 14840 14841 Just like we do with XWarpPointer's. 14842 14843 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 14844 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14845 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14846 14847commit ca17f3e9fd3b59fdc5ffd0e5d78e4db6ddc87aa1 14848Author: Carlos Garnacho <carlosg@gnome.org> 14849Date: Sun May 28 15:56:20 2017 +0200 14850 14851 xwayland: Lock the pointer if it is confined and has no cursor 14852 14853 In the typical pattern in games of "hide cursor, grab with a confineTo, 14854 warp constantly the pointer to the middle of the window" the last warping 14855 step is actually rather optional. Some games may choose to just set up a 14856 grab with confineTo argument, and trust that they'll get correct relative 14857 X/Y axis values despite the hidden cursor hitting the confinement window 14858 edge. 14859 14860 To cater for these cases, lock the pointer whenever there is a pointer 14861 confinement and the cursor is hidden. This ensures the pointer position 14862 is in sync with the compositor's when it's next shown again, and more 14863 importantly resorts to the relative pointer for event delivery. 14864 14865 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 14866 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14867 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14868 14869commit 513e3bd3870fdb8a8e0e2e52c0fa93872300bc8b 14870Author: Carlos Garnacho <carlosg@gnome.org> 14871Date: Sun May 28 15:56:19 2017 +0200 14872 14873 xwayland: Update root window size when desktop size changes 14874 14875 This fixes grabs on InputOnly windows whose parent is the root window 14876 failing with GrabNotViewable. This is due to window->borderSize/windowSize 14877 being computed as clipped by its parent, resulting in a null region. 14878 14879 Setting up the right size on the root window makes the InputOnly size 14880 correct too, so the GrabNotViewable paths aren't hit anymore. 14881 14882 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 14883 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 14884 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14885 14886commit fafdb0cc9697eb53635ed1e78bec1d4cd87ab3a2 14887Author: Carlos Garnacho <carlosg@gnome.org> 14888Date: Sun May 28 15:56:18 2017 +0200 14889 14890 xwayland: "Accept" confineTo on InputOnly windows 14891 14892 Of sorts, actually make it confine to the pointer focus, as the 14893 InputOnly window is entirely invisible to xwayland accounting, 14894 we don't have a xwl_window for it. 14895 14896 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 14897 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14898 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14899 14900commit c217fcb4c4640ffd2fefee63c6fcd7ea5e64b942 14901Author: Carlos Garnacho <carlosg@gnome.org> 14902Date: Sun May 28 15:56:17 2017 +0200 14903 14904 xwayland: Allow pointer warp on root/None window 14905 14906 Of sorts, as we can't honor pointer warping across the whole root window 14907 coordinates, peek the pointer focus in these cases. 14908 14909 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 14910 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 14911 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14912 14913commit f065721e6fa6e9057cca4eadaf145a4dc22a08d2 14914Author: Carlos Garnacho <carlosg@gnome.org> 14915Date: Sun May 28 15:56:16 2017 +0200 14916 14917 dix: assume warping on the RootWindow always happens on visible coords 14918 14919 If the root window borderClip region is null, the PointInWindowIsVisible() 14920 check fails if pointer warping is attempted on the root window, making 14921 the warping operation bail out early. 14922 14923 Assume coordinates always lay inside the root window for this case, 14924 the actual position will be clamped later within screen coordinates anyway. 14925 14926 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 14927 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 14928 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 14929 14930commit 7c88977d338a01aca866e52c9e736f8857fb9ae4 14931Author: Michel Dänzer <michel.daenzer@amd.com> 14932Date: Fri May 26 12:30:13 2017 +0900 14933 14934 glamor: Store the actual EGL/GLX context pointer in lastGLContext 14935 14936 Fixes subtle breakage which could sometimes trigger after a server reset 14937 with multiple screens using glamor: 14938 14939 Screen A enters glamor_close_screen last and calls various cleanup 14940 functions, which at some point call glamor_make_current to make sure 14941 screen A's GL context is current. This sets lastGLContext to screen A's 14942 &glamor_priv->ctx. Finally, glamor_close_screen calls 14943 glamor_release_screen_priv, which calls free(glamor_priv). 14944 14945 Later, screen B enters glamor_init, which allocates a new glamor_priv. 14946 With bad luck, this can return the same pointer which was previously 14947 used for screen A's glamor_priv. So when screen B's glamor_init calls 14948 glamor_make_current, lastGLContext == &glamor_priv->ctx, so MakeCurrent 14949 isn't called for screen B's GL context, and the following OpenGL API 14950 calls triggered by glamor_init mess up screen A's GL context. 14951 14952 The observed end result of this was a crash in glamor_get_vbo_space 14953 because glamor_priv->vbo didn't match the GL context, though there might 14954 be other possible outcomes. 14955 14956 Assigning the actual GL context pointer to lastGLContext prevents this 14957 by preventing the false negative test in glamor_make_current. 14958 14959 Reviewed-by: Keith Packard <keithp@keithp.com> 14960 Reviewed-by: Eric Anholt <eric@anholt.net> 14961 14962commit d164c10850609c96fd46c8441efd40940b06dfe0 14963Author: Michel Dänzer <michel.daenzer@amd.com> 14964Date: Thu May 25 16:19:35 2017 +0900 14965 14966 xfree86: Add Option "PreferCloneMode" 14967 14968 When the default behaviour was changed from clone mode to horizontal 14969 extended layout, a boolean ScrnInfoRec member preferClone was introduced 14970 to choose the old default behaviour. Option "PreferCloneMode" allows 14971 setting this preferClone member. 14972 14973 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 14974 14975commit 87db150d8a572d8983a234663aaaab9cb8b6c98d 14976Author: Eric Anholt <eric@anholt.net> 14977Date: Thu Jun 1 10:07:55 2017 -0700 14978 14979 glamor: Remove the "delayed fallback" code. 14980 14981 The usage of this died with the old core rendering code. 14982 14983 Signed-off-by: Eric Anholt <eric@anholt.net> 14984 Reviewed-by: Keith Packard <keithp@keithp.com> 14985 14986commit 72ddad7a97fa062bf9967d221c9a5520ebe33352 14987Author: Eric Anholt <eric@anholt.net> 14988Date: Wed May 17 11:36:07 2017 -0700 14989 14990 glamor: Drop glamor_set_screen_pixmap(). 14991 14992 All that was left here was updating the FBO's size. However, the FBO 14993 size was always set correctly already through 14994 glamor_set_pixmap_texture() from whoever had attached a new BO to the 14995 pixmap. 14996 14997 Signed-off-by: Eric Anholt <eric@anholt.net> 14998 Reviewed-by: Keith Packard <keithp@keithp.com> 14999 15000commit 49b12cb7360608b14caef5facb12f146d0bd15e6 15001Author: Eric Anholt <eric@anholt.net> 15002Date: Wed May 17 11:33:16 2017 -0700 15003 15004 glamor: Stop tracking the screen_fbo. 15005 15006 This means we no longer get "s" for on-screen drawing in glamor_debug, 15007 and there's only "m" (CPU memory) or "f" (Any GPU memory, aka FBOs). 15008 That seems fine to me. 15009 15010 Signed-off-by: Eric Anholt <eric@anholt.net> 15011 Reviewed-by: Keith Packard <keithp@keithp.com> 15012 15013commit fd0d2523a33620b36cd478925ddac5cf70f9c409 15014Author: Eric Anholt <eric@anholt.net> 15015Date: Tue May 9 15:33:21 2017 -0700 15016 15017 glamor_egl: Stop saving the EGL major/minor version. 15018 15019 We don't use them for anything. 15020 15021 Signed-off-by: Eric Anholt <eric@anholt.net> 15022 Reviewed-by: Keith Packard <keithp@keithp.com> 15023 15024commit 1f38a31ed3969471cbed69c61edb971f6cff5287 15025Author: Jon Turney <jon.turney@dronecode.org.uk> 15026Date: Sun May 7 20:53:04 2017 +0100 15027 15028 Add meson.build for XWin server (v2) 15029 15030 This needs a meson with PRs #1784, #1792 and #1794 15031 15032 Future work: remove conditionals which are always on, and simplify redundant 15033 CYGDEBUG conditionals 15034 15035 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15036 Reviewed-by: Eric Anholt <eric@anholt.net> 15037 15038commit 36b9dac212a0f8a287cdbd35db152a5eb5cbc744 15039Author: Jon Turney <jon.turney@dronecode.org.uk> 15040Date: Tue May 9 12:59:19 2017 +0100 15041 15042 hw/xwin: Remove pretense of Xv support 15043 15044 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15045 Reviewed-by: Eric Anholt <eric@anholt.net> 15046 15047commit 793af4d3f945b1d59eb2f84e625b581ea90b0066 15048Author: Jon Turney <jon.turney@dronecode.org.uk> 15049Date: Sun May 7 20:31:55 2017 +0100 15050 15051 hw/xwin: Don't unconditionally include rootless.h 15052 15053 Don't unconditionally include rootless.h, and so we don't need to add 15054 rootless to the include path unless building MWEXTWM. 15055 15056 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15057 Reviewed-by: Eric Anholt <eric@anholt.net> 15058 15059commit d8ccfb132602be88e640cc87ea58496e0445aab7 15060Author: Eric Anholt <eric@anholt.net> 15061Date: Thu Jun 1 16:46:02 2017 -0700 15062 15063 meson: Fix enabling of xshmfence. 15064 15065 I misspelled the enable flag, so DRI3 would throw BadImplementation 15066 when you tried to start any GL app. Same as in 15067 c7be7a688a78a34f61b90c0d95914e14b90b0cdc, we also convert it to #ifdef 15068 for consistency. 15069 15070 Signed-off-by: Eric Anholt <eric@anholt.net> 15071 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 15072 15073commit 4f29366f1e5678505fb882143c9b4a892d5b8273 15074Author: Lyude <lyude@redhat.com> 15075Date: Tue May 30 16:39:49 2017 -0400 15076 15077 xwayland: Don't load extension list more than once 15078 15079 When running an Xwayland server from the command line, we end up 15080 resetting the server every time all of the clients connected to the 15081 server leave. This would be fine, except that xwayland makes the mistake 15082 of unconditionally calling LoadExtensionList(). This causes us to setup 15083 the glxExtension twice in a row which means that when we lose our last 15084 client on the second server generation, we end up trying to call the glx 15085 destructors twice in a row resulting in a segfault: 15086 15087 (EE) 15088 (EE) Backtrace: 15089 (EE) 0: Xwayland (OsSigHandler+0x3b) [0x4982f9] 15090 (EE) 1: /lib64/libpthread.so.0 (__restore_rt+0x0) [0x70845bf] 15091 (EE) 2: /usr/lib64/dri/swrast_dri.so (__driDriverGetExtensions_virtio_gpu+0x32897d) [0x1196e5bd] 15092 (EE) 3: /usr/lib64/dri/swrast_dri.so (__driDriverGetExtensions_virtio_gpu+0x328a45) [0x1196e745] 15093 (EE) 4: /usr/lib64/dri/swrast_dri.so (__driDriverGetExtensions_virtio_gpu+0x32665f) [0x11969f7f] 15094 (EE) 5: Xwayland (__glXDRIscreenDestroy+0x30) [0x54686e] 15095 (EE) 6: Xwayland (glxCloseScreen+0x3f) [0x5473db] 15096 (EE) 7: Xwayland (glxCloseScreen+0x53) [0x5473ef] 15097 (EE) 8: Xwayland (dix_main+0x7b6) [0x44c8c9] 15098 (EE) 9: Xwayland (main+0x28) [0x61c503] 15099 (EE) 10: /lib64/libc.so.6 (__libc_start_main+0xf1) [0x72b1401] 15100 (EE) 11: Xwayland (_start+0x2a) [0x4208fa] 15101 (EE) 12: ? (?+0x2a) [0x2a] 15102 (EE) 15103 (EE) Segmentation fault at address 0x18 15104 (EE) 15105 Fatal server error: 15106 (EE) Caught signal 11 (Segmentation fault). Server aborting 15107 (EE) 15108 15109 Easy reproduction recipe: 15110 - Start an Xwayland session with the default settings 15111 - Open a window 15112 - Close that window 15113 - Open another window 15114 - Close that window 15115 - Total annihilation occurs 15116 15117 Signed-off-by: Lyude <lyude@redhat.com> 15118 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 15119 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15120 15121commit 2fbf62b2fb3dcb29551251d09aa695715bb754f4 15122Author: Jason Gerecke <killertofu@gmail.com> 15123Date: Fri May 26 14:27:19 2017 -0700 15124 15125 xfree86: Fix interpretation of xf86WaitForInput timeout 15126 15127 Commit aa6717ce2 switched xf86WaitForInput from using select(2) to using 15128 poll(2). Before this change, the timeout was interpreted as being in 15129 microseconds; afterwards it is fed directly to xorg_poll which interprets 15130 it as being in milliseconds. This results in the function potentially 15131 blocking 1000x longer than intended. This commit scales down the timeout 15132 argument before passing it to xorg_poll, being careful to ensure the result 15133 is not rounded down due to integer division. 15134 15135 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 15136 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 15137 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15138 15139commit 96af794dc648eadcd596893412d7530e92cb5421 15140Author: Peter Hutterer <peter.hutterer@who-t.net> 15141Date: Fri May 5 09:04:35 2017 +1000 15142 15143 xfree86: up the path name size to 512 in xf86MatchDriverFromFiles 15144 15145 ./hw/xfree86/common/xf86pciBus.c: In function ‘xf86MatchDriverFromFiles’: 15146 ../hw/xfree86/common/xf86pciBus.c:1330:52: warning: ‘snprintf’ output may be 15147 truncated before the last format character [-Wformat-truncation=] 15148 snprintf(path_name, sizeof(path_name), "%s/%s", ^~~~~~~ 15149 ../hw/xfree86/common/xf86pciBus.c:1330:13: note: ‘snprintf’ output between 2 15150 15151 dirent->d_name is 256, so sprintf("%s/%s") into a 256 buffer gives us: 15152 15153 and 257 bytes into a destination of size 256 15154 15155 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15156 15157commit bf3a591fdc5c98f7cab232bf5ec39902947ae2e5 15158Author: Peter Hutterer <peter.hutterer@who-t.net> 15159Date: Mon May 8 10:07:51 2017 +1000 15160 15161 fb: remove some superfluous empty lines 15162 15163 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15164 15165commit 933d80ba93cf9b5d32ce8673d60810c128debdab 15166Author: Adam Jackson <ajax@redhat.com> 15167Date: Thu May 18 11:19:53 2017 -0400 15168 15169 misprite: Clean up some preprocessor abuse 15170 15171 Signed-off-by: Adam Jackson <ajax@redhat.com> 15172 Reviewed-by: Eric Anholt <eric@anholt.net> 15173 Reviewed-by: Keith Packard <keithp@keithp.com> 15174 15175commit 57c223002c61c8ceb84fa7a02a3c1579517996f3 15176Author: Adam Jackson <ajax@redhat.com> 15177Date: Thu May 18 11:19:52 2017 -0400 15178 15179 misprite: Remove some dead cache variables 15180 15181 Unused since Keith ported misprite to damage in '04. 15182 15183 Signed-off-by: Adam Jackson <ajax@redhat.com> 15184 Reviewed-by: Eric Anholt <eric@anholt.net> 15185 Reviewed-by: Keith Packard <keithp@keithp.com> 15186 15187commit d0675e96f1fc8501324c88afa648a9d394f35587 15188Author: Eric Anholt <eric@anholt.net> 15189Date: Tue May 16 11:37:23 2017 -0700 15190 15191 glamor_egl: Drop glamor_egl_create_textured_screen_ext(). 15192 15193 The function hasn't been doing anything useful since keithp's resource 15194 freeing fixes in 2014. 15195 15196 Reviewed-by: Adam Jackson <ajax@redhat.com> 15197 Signed-off-by: Eric Anholt <eric@anholt.net> 15198 15199commit c36f56808ffbefc4a3dcf61dfad17f62aeb8e16c 15200Author: Eric Anholt <eric@anholt.net> 15201Date: Wed May 17 17:06:44 2017 -0700 15202 15203 glamor_egl: Automatically choose a GLES2 context if desktop GL fails. 15204 15205 GLES2 support has been requested multiple times, and we've had this 15206 code laying around trying to implement it. The GLES2 implementation 15207 is not quite there yet (some pixel transfer failures), but it 15208 shouldn't take much fixing at this point. 15209 15210 Reviewed-by: Adam Jackson <ajax@redhat.com> 15211 Signed-off-by: Eric Anholt <eric@anholt.net> 15212 15213commit f6bf5f1c7475ff1b19b0e8c3e71c1224d5b6bc82 15214Author: Eric Anholt <eric@anholt.net> 15215Date: Wed May 17 17:06:43 2017 -0700 15216 15217 glamor_egl: Remove check for KHR_surfaceless_context_* 15218 15219 Those extensions don't exist. There's only surfaceless_context. 15220 15221 Signed-off-by: Eric Anholt <eric@anholt.net> 15222 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 15223 15224commit 6e793ae08510639d011f638bcfc5246b9cccd1ae 15225Author: Eric Anholt <eric@anholt.net> 15226Date: Tue May 16 11:37:21 2017 -0700 15227 15228 glamor_egl: Drop warning about indirect GLX and GLES2. 15229 15230 Indirect GLX uses its own context and doesn't care what glamor is 15231 using. 15232 15233 Reviewed-by: Adam Jackson <ajax@redhat.com> 15234 Signed-off-by: Eric Anholt <eric@anholt.net> 15235 15236commit 59de50d6ef3331ee03eb82d4acba45aa75c70d42 15237Author: Eric Anholt <eric@anholt.net> 15238Date: Tue May 16 11:37:20 2017 -0700 15239 15240 glamor_egl: Avoid flink names in glamor_egl_create_textured_pixmap(). 15241 15242 Using flink is banned on render nodes, and they needlessly expose our 15243 screen pixmap contents to any authenticated client. This also 15244 incidentally drops the dependency on EGL_MESA_drm_image. 15245 15246 Reviewed-by: Adam Jackson <ajax@redhat.com> 15247 Signed-off-by: Eric Anholt <eric@anholt.net> 15248 15249commit 5d6f17d913bd4de0b4594585c0a8ba4f717a2a78 15250Author: Eric Anholt <eric@anholt.net> 15251Date: Tue May 16 11:37:19 2017 -0700 15252 15253 glamor_egl: Drop dead "cpp" field 15254 15255 It's been unused since the initial import. 15256 15257 Reviewed-by: Adam Jackson <ajax@redhat.com> 15258 Signed-off-by: Eric Anholt <eric@anholt.net> 15259 15260commit 37618e810b70c24bb8a97dce844963ed1a99f897 15261Author: Eric Anholt <eric@anholt.net> 15262Date: Tue May 16 11:37:18 2017 -0700 15263 15264 glamor_egl: Drop dead gl_context_depth. 15265 15266 This was replaced in 4afe15d8bfd575c010ed1868697a7922a37ab378, but not 15267 deleted. 15268 15269 Reviewed-by: Adam Jackson <ajax@redhat.com> 15270 Signed-off-by: Eric Anholt <eric@anholt.net> 15271 15272commit a6a751e5d8962b01abc2a92ec35300f68cb38d94 15273Author: Eric Anholt <eric@anholt.net> 15274Date: Tue May 16 11:37:17 2017 -0700 15275 15276 glamor_egl: Drop unnecessary check for KHR_gl_renderbuffer_image. 15277 15278 I couldn't find it being used anywhere in the history of the code. 15279 15280 Reviewed-by: Adam Jackson <ajax@redhat.com> 15281 Signed-off-by: Eric Anholt <eric@anholt.net> 15282 15283commit 40a153ee452390f713f7e5638af320a70defe106 15284Author: Eric Anholt <eric@anholt.net> 15285Date: Tue May 16 11:37:16 2017 -0700 15286 15287 glamor_egl: Always require the gbm-based image import path. 15288 15289 This has been associated with dri3 for now, but we need to use it 15290 elsewhere in order to avoid flink. The extensions have been 15291 implemented for long enough that I couldn't find when it was that we 15292 turned them on. Oddly, we already required renderbuffer import 15293 support, which is basically as hard to implement as texture import. 15294 15295 Reviewed-by: Adam Jackson <ajax@redhat.com> 15296 Signed-off-by: Eric Anholt <eric@anholt.net> 15297 Acked-by: Daniel Stone <daniels@collabora.com> 15298 15299commit b1b71b7a8b2d2c5d8f4afa67b75d10d4e7cbaa20 15300Author: Eric Anholt <eric@anholt.net> 15301Date: Tue May 16 11:37:15 2017 -0700 15302 15303 glamor_egl: Drop the has_gem flag. 15304 15305 We're using GBM, so we know we've got GEM. 15306 15307 Reviewed-by: Adam Jackson <ajax@redhat.com> 15308 Signed-off-by: Eric Anholt <eric@anholt.net> 15309 15310commit 32957d9faeb8b0e968a23d91ed7391c0465dd1b9 15311Author: Eric Anholt <eric@anholt.net> 15312Date: Tue May 16 11:37:14 2017 -0700 15313 15314 modesetting: Drop code for GLAMOR && !GLAMOR_HAS_GBM. 15315 15316 The glamor_egl module that the GLAMOR paths are using is only built if 15317 GLAMOR_HAS_GBM is true, and there's no plan for implementing the 15318 module without GBM. Simplify modesetting's code as a result. 15319 15320 Reviewed-by: Adam Jackson <ajax@redhat.com> 15321 Signed-off-by: Eric Anholt <eric@anholt.net> 15322 15323commit 5102ea38eb7331aeb1258cd148fd374cf0cbf8f2 15324Author: Eric Anholt <eric@anholt.net> 15325Date: Tue May 16 11:37:13 2017 -0700 15326 15327 glamor_egl: Unifdef GLAMOR_HAS_GBM. 15328 15329 We only build this code with GBM, and supporting non-GBM well would be 15330 invasive. 15331 15332 Reviewed-by: Adam Jackson <ajax@redhat.com> 15333 Signed-off-by: Eric Anholt <eric@anholt.net> 15334 15335commit 5064153c79d8b18767c0ade7635680a2f396f296 15336Author: Eric Anholt <eric@anholt.net> 15337Date: Tue May 16 11:37:12 2017 -0700 15338 15339 glamor_egl: Print a useful identifying string on initialization. 15340 15341 The EGL version is not used anywhere in the glamor code, so it's not 15342 interesting. And when saying that we've started using GL 15343 acceleration, it's nice to know what GL we're actually using. 15344 15345 Reviewed-by: Adam Jackson <ajax@redhat.com> 15346 Signed-off-by: Eric Anholt <eric@anholt.net> 15347 15348commit fd96ff7fb325f54cb06c2bf5b8c37c75457b5c42 15349Author: Jon Turney <jon.turney@dronecode.org.uk> 15350Date: Tue Apr 18 10:44:55 2017 +0100 15351 15352 hw/xwin: Fix -Wunused-function warning for MessageName() in non-debug builds 15353 15354 ../hw/xwin/winmultiwindowwm.c:188:1: warning: ‘MessageName’ defined but not used [-Wunused-function] 15355 15356 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15357 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 15358 15359commit e49f5947eda578d8b029f42fb4ec9c90f413b4b3 15360Author: Jon Turney <jon.turney@dronecode.org.uk> 15361Date: Tue Apr 18 10:45:07 2017 +0100 15362 15363 hw/xwin: Fix -Wmaybe-uninitialized warnings in engine CloseScreen wrappers 15364 15365 ../hw/xwin/winshadgdi.c: In function ‘winCloseScreenShadowGDI’: 15366 ../hw/xwin/winshadgdi.c:632:12: warning: ‘fReturn’ may be used uninitialized in this function [-Wmaybe-uninitialized] 15367 ../hw/xwin/winshadgdi.c:579:10: note: ‘fReturn’ was declared here 15368 ../hw/xwin/winshadddnl.c: In function ‘winCloseScreenShadowDDNL’: 15369 ../hw/xwin/winshadddnl.c:711:12: warning: ‘fReturn’ may be used uninitialized in this function [-Wmaybe-uninitialized] 15370 ../hw/xwin/winshadddnl.c:661:10: note: ‘fReturn’ was declared here 15371 15372 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15373 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 15374 15375commit 319daa7a9fa8bcfdefb08d7770912775af22ad99 15376Author: Jon Turney <jon.turney@dronecode.org.uk> 15377Date: Wed Apr 19 13:30:27 2017 +0100 15378 15379 hw/xwin: Fix -Wmaybe-uninitialized warning in winWindowProc 15380 15381 This is possibly an actual bug in failing to check we successfully retrieved 15382 the monitor size before using it to set the X screen size. 15383 15384 ../hw/xwin/winwndproc.c: In function ‘winWindowProc’: 15385 ../hw/xwin/winwndproc.c:283:55: warning: ‘dwHeight’ may be used uninitialized in this function [-Wmaybe-uninitialized] 15386 ../hw/xwin/winwndproc.c:240:32: note: ‘dwHeight’ was declared here 15387 ../hw/xwin/winwndproc.c:281:54: warning: ‘dwWidth’ may be used uninitialized in this function [-Wmaybe-uninitialized] 15388 ../hw/xwin/winwndproc.c:240:23: note: ‘dwWidth’ was declared here 15389 15390 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15391 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 15392 15393commit 7a90c9b24e6776c8f1fa43df4268f79a465c0d4e 15394Author: Jon Turney <jon.turney@dronecode.org.uk> 15395Date: Tue Apr 18 10:32:31 2017 +0100 15396 15397 xfree86: Hide unused variables when !XSERVER_LIBPCIACCESS 15398 15399 ../hw/xfree86/common/xf86Config.c: In function ‘xf86HandleConfigFile’: 15400 ../hw/xfree86/common/xf86Config.c:2278:10: warning: unused variable ‘singlecard’ [-Wunused-variable] 15401 ../hw/xfree86/common/xf86Config.c:2277:17: warning: unused variable ‘scanptr’ [-Wunused-variable] 15402 15403 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15404 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 15405 15406commit 7a19a3e093a62a779c3cab8c28930adcb1604e0d 15407Author: Peter Hutterer <peter.hutterer@who-t.net> 15408Date: Mon May 8 10:24:34 2017 +1000 15409 15410 xfree86: build a shared lib identical to Xorg 15411 15412 Because we can use a shared lib in dlsym() for the symbol loader test, 15413 but we can't use the executable. 15414 15415 v2: Drop a stray ' ' and rebase on 820a4cbe9fe, by anholt 15416 15417 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15418 Reviewed-by: Eric Anholt <eric@anholt.net> 15419 15420commit 49c742557457ff9200a10f32daadfcea50d4911c 15421Author: Adam Jackson <ajax@redhat.com> 15422Date: Thu May 11 17:17:55 2017 -0400 15423 15424 meson: Fix CLOCK_MONOTONIC test 15425 15426 C99 isn't enough on its own to get POSIX feature sets defined. 15427 15428 Signed-off-by: Adam Jackson <ajax@redhat.com> 15429 Reviewed-by: Julien Cristau <jcristau@debian.org> 15430 Tested-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 15431 15432commit 152375f4e4f1ee4833129802730b36af8d0f7e1a 15433Author: Adam Jackson <ajax@redhat.com> 15434Date: Thu May 11 17:17:54 2017 -0400 15435 15436 os, xfree86: Stop being so weird about <limits.h> 15437 15438 Whatever problem this is trying to fix, we don't care. Just include the 15439 thing and stop worrying about whether _POSIX_SOURCE is defined. 15440 15441 Signed-off-by: Adam Jackson <ajax@redhat.com> 15442 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 15443 Reviewed-by: Julien Cristau <jcristau@debian.org> 15444 Tested-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 15445 15446commit d9e23ea4228575344e3b4c0443cecc5eb75356e4 15447Author: Keith Packard <keithp@keithp.com> 15448Date: Wed May 10 21:50:45 2017 -0700 15449 15450 dix: Remove clients from input and output ready queues after closing 15451 15452 Delay removing the client from these two queues until all potential 15453 I/O has completed in case we mark the client as ready for reading or 15454 with pending output during the close operation. 15455 15456 Bugzilla: https://bugs.freedesktop.org/100957 15457 Signed-off-by: Keith Packard <keithp@keithp.com> 15458 Tested-by: Nick Sarnie <commendsarnex@gmail.com> 15459 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 15460 15461commit c394f6b27329d212d2a8b9b3c6da1ef7847c46f3 15462Author: Julien Cristau <jcristau@debian.org> 15463Date: Thu May 11 08:42:16 2017 +0200 15464 15465 Replace all checks for 'linux' macro with '__linux__' 15466 15467 gcc -std=c99 does not define the former, and it's a horrible namespace 15468 confusion anyway. 15469 15470 Signed-off-by: Julien Cristau <jcristau@debian.org> 15471 Reviewed-by: Adam Jackson <ajax@redhat.com> 15472 Tested-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 15473 15474commit 6634ffc4d26846dcf892f27682f9021f6d9956a9 15475Author: Julien Cristau <jcristau@debian.org> 15476Date: Thu May 11 08:42:15 2017 +0200 15477 15478 Drop workaround for pre-glibc linux 15479 15480 It seems unlikely anyone still needs to build against libc4/libc5. 15481 15482 Reviewed-by: Adam Jackson <ajax@redhat.com> 15483 Signed-off-by: Julien Cristau <jcristau@debian.org> 15484 15485commit 318257029578a30e29c191befa94a20ad07f97d4 15486Author: Frank Binns <frank.binns@imgtec.com> 15487Date: Thu May 11 14:00:19 2017 +0100 15488 15489 hw/xfree86: fix comment typo 15490 15491 Signed-off-by: Frank Binns <frank.binns@imgtec.com> 15492 15493commit 0848390d51f93ecfb9342cffe1cda46a65cfc07b 15494Author: Adam Jackson <ajax@redhat.com> 15495Date: Wed May 10 11:14:57 2017 -0400 15496 15497 xephyr: Fix regeneration 15498 15499 I had said: 15500 15501 commit c42311a9d7d2e5a67bdb7f4fa32032b4feba59b1 15502 Author: Adam Jackson <ajax@redhat.com> 15503 Date: Fri Mar 24 15:58:54 2017 -0400 15504 15505 kdrive: Remove KdOsFuncs 15506 15507 Only the Init slot was used, and Xephyr can just as easily do 15508 that initialization directly. 15509 15510 And I'd've been right, but I forgot to make that initialization only 15511 happen on startup (i.e. when serverGeneration == 1). 15512 15513 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 15514 Signed-off-by: Adam Jackson <ajax@redhat.com> 15515 15516commit 0ff2fb128bf35dffaa302f15e9ae004b65ee0827 15517Author: Peter Harris <pharris@opentext.com> 15518Date: Tue May 9 19:39:47 2017 -0400 15519 15520 meson: Detect strlcat/strlcpy/reallocarray in libbsd 15521 15522 If we're linking with libbsd anyway, we might as well use the functions 15523 it provides instead of compiling our replacements. 15524 15525 Signed-off-by: Peter Harris <pharris@opentext.com> 15526 Reviewed-by: Eric Anholt <eric@anholt.net> 15527 15528commit c4c002d1ca80bd69776387dafb9c5bb082c72e48 15529Author: Peter Harris <pharris@opentext.com> 15530Date: Tue May 9 19:39:46 2017 -0400 15531 15532 meson: Only detect each function once 15533 15534 Use conf_data outside of include/ to avoid re-running detection of the 15535 same functions. 15536 15537 Reviewed-by: Eric Anholt <eric@anholt.net> 15538 Signed-off-by: Peter Harris <pharris@opentext.com> 15539 15540commit 03d6275e6094a5ede5a70f05bbbdde653a9fd9e0 15541Author: Peter Harris <pharris@opentext.com> 15542Date: Tue May 9 19:39:45 2017 -0400 15543 15544 meson: Detect more functions 15545 15546 Set HAVE_REALLOCARRAY, HAVE_SIGACTION, HAVE_STRCASESTR, HAVE_STRLCAT, 15547 HAVE_STRLCPY, HAVE_TIMINGSAFE_MEMCMP, and BUSFAULT. 15548 15549 Reviewed-by: Eric Anholt <eric@anholt.net> 15550 Signed-off-by: Peter Harris <pharris@opentext.com> 15551 15552commit ba1599610b1889545be8a9314dce8a740a2764d3 15553Author: Peter Harris <pharris@opentext.com> 15554Date: Tue May 9 19:39:44 2017 -0400 15555 15556 meson: Detect arc4random_buf 15557 15558 Reviewed-by: Eric Anholt <eric@anholt.net> 15559 Signed-off-by: Peter Harris <pharris@opentext.com> 15560 15561commit e67ca940473e249c7b9712ce6d515310ed1e4cf9 15562Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15563Date: Mon May 8 11:40:22 2017 +0100 15564 15565 meson: An empty array is not a dependency object 15566 15567 Using the meson idiom of initializing a variaible to an empty array to 15568 represent a null value does not work well for a variable expected to hold a 15569 dependency object, as trying to apply found() to it will fail (currently). 15570 15571 I think this can be demonstrated by configuing with Dglamor=no, then gbm_dep 15572 is set to [] in meson.build, and then tested with found() in 15573 include/meson.build. 15574 15575 Use dependency('') to create an empty dependency. 15576 15577 Future work: In the autotools build GBM is required if build_xorg & 15578 build_glamor. I suspect that it was also required for build_xwayland & 15579 build_glamor, but that was never checked. That test seems to have gone 15580 missing in meson.build, the required:false when checking for gbm should 15581 possibly be dropped? 15582 15583 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15584 Reviewed-by: Eric Anholt <eric@anholt.net> 15585 15586commit 16054647abfc9fe5f71d0098e17415e45fa28199 15587Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15588Date: Mon May 8 11:40:21 2017 +0100 15589 15590 meson: Refine build_dbus 15591 15592 As in autotools build, only turn on build_dbus (build dbus-core.c) if it's 15593 needed by build_hal or build_systemd_logind, not just because dbus is 15594 available. 15595 15596 Building dbus-core.c without NEED_DBUS defined fails, as the contents of 15597 dbus-core.h are turned off. 15598 15599 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15600 Acked-by: Eric Anholt <eric@anholt.net> 15601 15602commit 820a4cbe9fef36bc97a7107ccc74a785180e227a 15603Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15604Date: Tue May 9 12:23:48 2017 +0100 15605 15606 meson: Make driprotos and libdrm optional 15607 15608 Add options for DRI{1,2,3} 15609 shmfence is required for DRI3 15610 libdrm is required for any DRI{1,2,3} 15611 Consolidate calls to dependency('libdrm') 15612 Set WITH_LIBDRM when building with libdrm 15613 15614 v2: 15615 Initialize libxserver_dri3 to [] 15616 Manually flatten libxserver, since meson doesn't (currently) 15617 Use version_compare rather than circumloctions with dependency() 15618 15619 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15620 Reviewed-by: Eric Anholt <eric@anholt.net> 15621 15622commit fdb11e2274f1d1069cdc9ff7b3655288c1b42cf4 15623Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15624Date: Mon May 8 11:40:18 2017 +0100 15625 15626 meson: Don't require xf86dgaproto 15627 15628 This dependency is never used, and build_dga is hardwired to false later on, 15629 anyhow. 15630 15631 Reviewed-by: Eric Anholt <eric@anholt.net> 15632 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15633 15634commit 5bff2a68560837fc402905fef6244ce794c47843 15635Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15636Date: Mon May 8 11:40:17 2017 +0100 15637 15638 meson: Nettle is required if it's the only SHA1 choice 15639 15640 Reviewed-by: Eric Anholt <eric@anholt.net> 15641 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15642 15643commit 95e5fa04762fa0f04ac1349708d9e33c30205a96 15644Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15645Date: Mon May 8 11:40:16 2017 +0100 15646 15647 meson: Remove stray whitespace 15648 15649 Reviewed-by: Eric Anholt <eric@anholt.net> 15650 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15651 15652commit e9dbecf7c259f7e8b610fa93f97ea55f5dafa7af 15653Author: Adam Jackson <ajax@redhat.com> 15654Date: Wed May 10 10:56:16 2017 -0400 15655 15656 timingsafe_memcmp: Fix meson build 15657 15658 Include dix-config.h first to pick up _GNU_SOURCE so we get the 15659 definition for sigset_t. 15660 15661 Reported-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 15662 Signed-off-by: Adam Jackson <ajax@redhat.com> 15663 15664commit 5d3b6cc221aff5b9322b31aa6f5ae922271bc7d8 15665Author: Karol Kosik <kkosik@nvidia.com> 15666Date: Wed Jul 22 16:42:07 2015 -0700 15667 15668 xfree86: Improved autoconfig drivers matching 15669 15670 Implementation of new drivers matching algorithm. New approach 15671 doesn't add duplicate drivers and ease drivers matching phase. 15672 15673 v2: Re-commit the patch reverted in 15674 2388f5e583d4ab2ee12f2b087d381b64aed3f7d5, with Aaron Plattner's 15675 fix squashed in (by anholt). 15676 15677 Signed-off-by: Karol Kosik <kkosik@nvidia.com> 15678 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 15679 Signed-off-by: Eric Anholt <eric@anholt.net> 15680 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> (v1) 15681 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> (v1) 15682 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 15683 Tested-by: Eric Anholt <eric@anholt.net> 15684 15685commit 2388f5e583d4ab2ee12f2b087d381b64aed3f7d5 15686Author: Eric Anholt <eric@anholt.net> 15687Date: Tue May 9 13:37:59 2017 -0700 15688 15689 Revert "xfree86: Improved autoconfig drivers matching" 15690 15691 This reverts commit 112d0d7d01b98fb0d67910281dd1feeec125247b. 15692 15693 It broke Xorg for Adam, Peter, and myself, by failing hard when a 15694 module load failed. 15695 15696 Signed-off-by: Eric Anholt <eric@anholt.net> 15697 15698commit 1dd14e7a499428cf7215b49a9319199545bcd6cb 15699Author: Eric Anholt <eric@anholt.net> 15700Date: Mon May 8 10:36:30 2017 -0700 15701 15702 meson: Use link_whole() not link_with(), for the Xorg dixmods. 15703 15704 I clearly hadn't run ninja test, since fb no longer had any of the fb 15705 symbols in it. 15706 15707 Signed-off-by: Eric Anholt <eric@anholt.net> 15708 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 15709 15710commit a06bb73053d9df56d9070ce325a43af3a3c7a6a2 15711Author: Lyude <lyude@redhat.com> 15712Date: Thu May 4 18:04:31 2017 -0400 15713 15714 xwayland: Unconditionally initialize lists in init_tablet_manager_seat() 15715 15716 In the event that xwayland gets launched on a wayland compositor that 15717 doesn't yet have support for wp_tablet_manager, we end up skipping the 15718 initialization of the lists. This is wrong, because regardless of 15719 whether or not a tablet is present we still attempt to traverse these 15720 lists later in xwl_set_cursor(), expecting that if the lists are empty 15721 from no tablet manager that we simply won't execute any loop iterations. 15722 15723 (EE) 15724 (EE) Backtrace: 15725 (EE) 0: Xwayland (OsSigHandler+0x3b) [0x4982f9] 15726 (EE) 1: /lib64/libpthread.so.0 (__restore_rt+0x0) [0x7f73722545bf] 15727 (EE) 2: Xwayland (xwl_set_cursor+0x9f) [0x429974] 15728 (EE) 3: Xwayland (miPointerUpdateSprite+0x261) [0x4fe1ca] 15729 (EE) 4: Xwayland (mieqProcessInputEvents+0x239) [0x4f8d33] 15730 (EE) 5: Xwayland (ProcessInputEvents+0x9) [0x4282f0] 15731 (EE) 6: Xwayland (Dispatch+0x42) [0x43e2d4] 15732 (EE) 7: Xwayland (dix_main+0x5c9) [0x44c6dc] 15733 (EE) 8: Xwayland (main+0x28) [0x61c523] 15734 (EE) 9: /lib64/libc.so.6 (__libc_start_main+0xf1) [0x7f7371e9d401] 15735 (EE) 10: Xwayland (_start+0x2a) [0x4208fa] 15736 (EE) 11: ? (?+0x2a) [0x2a] 15737 (EE) 15738 (EE) Segmentation fault at address 0x28 15739 (EE) 15740 Fatal server error: 15741 (EE) Caught signal 11 (Segmentation fault). Server aborting 15742 (EE) 15743 15744 Reproduced when trying to run upstream xwayland under fedora 25's weston 15745 package. 15746 15747 Signed-off-by: Lyude <lyude@redhat.com> 15748 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 15749 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15750 15751commit 5f83b951ac7e0e6bbd766a9bb860283e32808c90 15752Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15753Date: Thu May 4 13:46:26 2017 +0100 15754 15755 meson: Fix typo in xephyr option description 15756 15757 Reviewed-by: Adam Jackson <ajax@redhat.com> 15758 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15759 15760commit ce6741f6ac63493d6b903f8f58b83bf571baf4b9 15761Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15762Date: Thu May 4 13:46:25 2017 +0100 15763 15764 meson: Fix typoed filename 15765 15766 Reviewed-by: Adam Jackson <ajax@redhat.com> 15767 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15768 15769commit 492534813302ef95538f485401fbcf6eccbd0026 15770Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15771Date: Thu May 4 13:46:24 2017 +0100 15772 15773 meson: Make xf86vidmode optional 15774 15775 Reviewed-by: Adam Jackson <ajax@redhat.com> 15776 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15777 15778commit 3b3ce4a55b0c1f83eff37fb8544156e8961f3fa3 15779Author: Jon TURNEY <jon.turney@dronecode.org.uk> 15780Date: Thu May 4 13:46:23 2017 +0100 15781 15782 meson: Test to build xserver_poll.c was inverted 15783 15784 Test to build xserver_poll.c was inverted compared to autoconf. Build 15785 xserver_poll.c if poll is missing. 15786 15787 Reviewed-by: Adam Jackson <ajax@redhat.com> 15788 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 15789 15790commit 4b7dd6d1aa53bdeee2e3fd0c0a73f6ce40495682 15791Author: Adam Jackson <ajax@redhat.com> 15792Date: Wed May 3 15:40:08 2017 -0400 15793 15794 Makefile.am: make 'make distcheck' verify the meson build 15795 15796 Signed-off-by: Adam Jackson <ajax@redhat.com> 15797 Reviewed-by: Eric Anholt <eric@anholt.net> 15798 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15799 15800commit 111a5615a60565f7b68a7fa69dbac3d289d80c9e 15801Author: Peter Hutterer <peter.hutterer@who-t.net> 15802Date: Wed May 3 15:40:07 2017 -0400 15803 15804 Makefile.am: add the meson files to the tarball 15805 15806 While we're shipping both build systems, we'll likely have 'make dist' 15807 generated tarballs - those tarballs should be buildable with meson to 15808 have more exposure. List them all explicitly since the list isn't likely 15809 to change much. 15810 15811 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15812 Reviewed-by: Eric Engestrom <eric@engestrom.ch> 15813 Reviewed-by: Eric Anholt <eric@anholt.net> 15814 15815commit 11c133aafe63c776bde502057243d67fe2c630c8 15816Author: Eric Anholt <eric@anholt.net> 15817Date: Wed May 3 15:33:46 2017 -0700 15818 15819 meson: Fix kdrive build. 15820 15821 Signed-off-by: Eric Anholt <eric@anholt.net> 15822 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 15823 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15824 15825commit b7376fb933ac1b0c1771535c14cf77029eccb7f5 15826Author: Adam Jackson <ajax@redhat.com> 15827Date: Fri Mar 24 15:58:55 2017 -0400 15828 15829 kdrive: Remove dead slots from KdCardFuncs 15830 15831 Signed-off-by: Adam Jackson <ajax@redhat.com> 15832 Reviewed-by: Eric Anholt <eric@anholt.net> 15833 15834commit c42311a9d7d2e5a67bdb7f4fa32032b4feba59b1 15835Author: Adam Jackson <ajax@redhat.com> 15836Date: Fri Mar 24 15:58:54 2017 -0400 15837 15838 kdrive: Remove KdOsFuncs 15839 15840 Only the Init slot was used, and Xephyr can just as easily do that 15841 initialization directly. 15842 15843 Reviewed-by: Eric Anholt <eric@anholt.net> 15844 Signed-off-by: Adam Jackson <ajax@redhat.com> 15845 15846commit c0375dced38674ed98562529530d89ff02c48100 15847Author: Adam Jackson <ajax@redhat.com> 15848Date: Fri Mar 24 15:58:53 2017 -0400 15849 15850 kdrive: static and dead code cleanup 15851 15852 Reviewed-by: Eric Anholt <eric@anholt.net> 15853 Signed-off-by: Adam Jackson <ajax@redhat.com> 15854 15855commit d732c36597fab2e9bc4f2aa72cf1110997697557 15856Author: Adam Jackson <ajax@redhat.com> 15857Date: Thu Apr 27 14:45:25 2017 -0400 15858 15859 xfree86: Silence a new glibc warning 15860 15861 glibc would like to stop declaring major()/minor() macros in 15862 <sys/types.h> because that header gets included absolutely everywhere 15863 and unix device major/minor is perhaps usually not what's expected. Fair 15864 enough. If one includes <sys/sysmacros.h> as well then glibc knows we 15865 meant it and doesn't warn, so do that if it exists. 15866 15867 Signed-off-by: Adam Jackson <ajax@redhat.com> 15868 15869commit c3147a20065b212fac78eb29c9bb9e150f9b22f5 15870Author: Adam Jackson <ajax@redhat.com> 15871Date: Thu Apr 27 14:45:24 2017 -0400 15872 15873 dmx: Fix decl mismatch for XRT_WINDOW 15874 15875 Building xserver with meson -Db_lto=true found this! Check it: 15876 15877 ../hw/dmx/dmx.c:66:22: warning: type of ‘XRT_WINDOW’ does not match 15878 original declaration [-Wlto-type-mismatch] 15879 extern unsigned long XRT_WINDOW; 15880 ^ 15881 ../include/resource.h:58:18: note: type ‘RESTYPE’ should match type ‘long 15882 unsigned int’ 15883 typedef uint32_t RESTYPE; 15884 ^ 15885 ../Xext/panoramiX.c:85:9: note: ‘XRT_WINDOW’ was previously declared here 15886 RESTYPE XRT_WINDOW; 15887 ^ 15888 ../Xext/panoramiX.c:85:9: note: code may be misoptimized unless 15889 -fno-strict-aliasing is used 15890 15891 I have never managed successfully to build xserver with LTO with 15892 autotools, so, score one for meson. 15893 15894 Get the decl from panoramiXsrv.h, which includes panoramiX.h first anyway. 15895 15896 Signed-off-by: Adam Jackson <ajax@redhat.com> 15897 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 15898 15899commit ea275cd3d5fb561c6aed7876f486e1124082426c 15900Author: Eric Anholt <eric@anholt.net> 15901Date: Fri Apr 28 15:09:38 2017 -0700 15902 15903 meson: Fix glx build with DRI2 disabled. 15904 15905 A common meson pattern is to use empty arrays for optional sources and 15906 dependencies, since it generally iterates over trees of arrays looking 15907 for values. A value of '' will actually go looking for ''. 15908 15909 Daniel Stone <daniels@collabora.com> 15910 Signed-off-by: Eric Anholt <eric@anholt.net> 15911 15912commit 62f727c7ea359b7bdf58d9b827731a082b47553d 15913Author: Eric Anholt <eric@anholt.net> 15914Date: Fri Apr 28 15:09:37 2017 -0700 15915 15916 meson: use link_with for Xorg's shared modules as well as Xorg. 15917 15918 I converted Xorg when meson 0.40 came out, but fumbled my grep and 15919 forgot that I had instances of the .extract_all_objects() workaround 15920 in dixmods, too. 15921 15922 Signed-off-by: Eric Anholt <eric@anholt.net> 15923 Reviewed-by: Daniel Stone <daniels@collabora.com> 15924 15925commit e2f68296ffb8e40035c0ebd949b67d1e2e424e11 15926Author: Keith Packard <keithp@keithp.com> 15927Date: Sat Apr 29 00:26:10 2017 -0700 15928 15929 os: Mark client as ready to read when closing due to write failure [100863] 15930 15931 This makes sure the server will go look at the client again, notice 15932 that the FD is no longer valid and close the client down. 15933 15934 Bugzilla: https://bugs.freedesktop.org/100863 15935 Signed-off-by: Keith Packard <keithp@keithp.com> 15936 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 15937 15938commit a82971b07035ee9a4e3ed01326e7c1eab34b5a19 15939Author: Keith Packard <keithp@keithp.com> 15940Date: Sat Apr 29 00:21:47 2017 -0700 15941 15942 os: un-duplicate code to close client on write failure 15943 15944 There are three copies of the same short sequence of operations to 15945 close down a client when a write error occurs. Create a new function, 15946 AbortClient, which performs these operations and then call it from the 15947 three places. 15948 15949 Signed-off-by: Keith Packard <keithp@keithp.com> 15950 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 15951 15952commit 8475e6360ce31551d50fd63a26f7a44d1e8928f2 15953Author: Peter Hutterer <peter.hutterer@who-t.net> 15954Date: Tue Feb 7 15:04:46 2017 +1000 15955 15956 xwayland: add tablet pad support 15957 15958 Hooked up a bit differently to the other tools. Those tools can be static for 15959 all and be re-used. The wacom driver initializes the pad with the correct 15960 number of buttons though and we can't do this until we have the pad done event. 15961 15962 If the tablet is removed and we plug a different one in, we should initialize 15963 that correctly, so unlike the other tools the pad is properly removed and 15964 re-initialized on plug. 15965 15966 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15967 Acked-by: Ping Cheng <ping.cheng@wacom.com> 15968 15969commit f471b5b8eb451b442554517c7cb6f0aa90d218c4 15970Author: Carlos Garnacho <carlosg@gnome.org> 15971Date: Fri Nov 4 19:58:04 2016 +0100 15972 15973 xwayland: update cursor on tablet tools in proximity 15974 15975 Each xwl_tablet_tool gets a xwl_cursor, as on wayland each of those 15976 will get an independent cursor that can be set through 15977 zwp_tablet_tool.set_cursor. 15978 15979 However, all tools (and the pointer) share conceptually the same VCP 15980 on Xwayland, so have cursor changes trigger a xwl_cursor update on 15981 every tool (and the pointer, again). Maybe Xwayland could keep track 15982 of the most recent device and only update that cursor to get better 15983 visual results, but this is simpler, and it's going to be odd 15984 anyway... 15985 15986 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 15987 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 15988 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 15989 Acked-by: Ping Cheng <ping.cheng@wacom.com> 15990 15991commit 6d1ad39fe6c18220dd39b0653fd1e4145140e2dc 15992Author: Carlos Garnacho <carlosg@gnome.org> 15993Date: Fri Nov 4 19:36:10 2016 +0100 15994 15995 xwayland: Refactor cursor management into xwl_cursor 15996 15997 This struct takes away the cursor info in xwl_seat, and has 15998 an update function so we can share the frame handling code 15999 across several xwl_cursors. 16000 16001 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 16002 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16003 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16004 Acked-by: Ping Cheng <ping.cheng@wacom.com> 16005 16006commit 773b04748d0c839bc8b12e33f74bb8d11c447f5b 16007Author: Peter Hutterer <peter.hutterer@who-t.net> 16008Date: Tue Feb 7 12:23:46 2017 +1000 16009 16010 xwayland: handle button events after motion events 16011 16012 Make sure the button events are sent after the motion events into the new 16013 position. 16014 16015 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16016 Acked-by: Ping Cheng <ping.cheng@wacom.com> 16017 16018commit 8a1defcc634daddbb3570519d69ec5c9e39a8b56 16019Author: Jason Gerecke <killertofu@gmail.com> 16020Date: Fri Oct 14 14:31:46 2016 -0700 16021 16022 xwayland: Handle tablet_tool events 16023 16024 Translates Wayland tablet events into corresponding X11 tablet events. As 16025 with the prior commit, these events are modeled after those created by the 16026 xf86-input-wacom driver to maximize compatibility with existing applications. 16027 16028 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 16029 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 16030 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16031 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16032 Acked-by: Ping Cheng <ping.cheng@wacom.com> 16033 16034commit 5812d1c28f4fb7b7de8b96a81415a21425561fd4 16035Author: Jason Gerecke <killertofu@gmail.com> 16036Date: Fri Jan 15 17:01:38 2016 -0800 16037 16038 xwayland: Handle wp_tablet events 16039 16040 Creates and maintains the canonical trio of X devices (stylus, eraser, 16041 and cursor) to be shared by all connected tablets. A per-tablet trio 16042 could be created instead, but there are very few benefits to such a 16043 configuration since all tablets still ultimately share control of a 16044 single master pointer. 16045 16046 The three X devices are modeled after those created by xf86-input-wacom 16047 but use a generic maximum X and Y that should be large enough to 16048 accurately represent values from even the largest currently-available 16049 tablets. 16050 16051 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 16052 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 16053 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16054 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16055 Acked-by: Ping Cheng <ping.cheng@wacom.com> 16056 16057commit 47c4415912b5b16b115135be365beb370858df76 16058Author: Jason Gerecke <killertofu@gmail.com> 16059Date: Fri Oct 14 14:50:18 2016 -0700 16060 16061 xwayland: Listen for wp_tablet_seat events 16062 16063 The wp_tablet_seat interface provides us with notifications as tablets, 16064 tools, and pads are connected to the system. Add listener functions and 16065 store references to the obtained devices. 16066 16067 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 16068 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 16069 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16070 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16071 Acked-by: Ping Cheng <ping.cheng@wacom.com> 16072 16073commit 7d48b758a601ce0252ebd21297a7c42263adfaaf 16074Author: Jason Gerecke <killertofu@gmail.com> 16075Date: Fri Jan 15 17:29:37 2016 -0800 16076 16077 xwayland: Bind to wp_tablet_manager if available and get its seats 16078 16079 If we're notified about the existence of the wp_tablet_manager interface, 16080 we bind to it so that we can make use of any tablets that are (or later 16081 become) available. For each seat that exists or comes into existance at 16082 a later point, obtain the associated tablet_seat. 16083 16084 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 16085 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 16086 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16087 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16088 Acked-by: Ping Cheng <ping.cheng@wacom.com> 16089 16090commit 89c841915ac4fba6d2a5ad0051c778f1a76ffbf3 16091Author: Jason Gerecke <killertofu@gmail.com> 16092Date: Thu Oct 13 10:39:46 2016 -0700 16093 16094 xwayland: Depend on wayland-protocols to build tablet protocol headers 16095 16096 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 16097 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 16098 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16099 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16100 Acked-by: Ping Cheng <ping.cheng@wacom.com> 16101 16102commit 59554eaa6c4d3d3c170a4be4fc385a86d05c5558 16103Author: Adam Jackson <ajax@redhat.com> 16104Date: Thu Apr 27 13:23:56 2017 -0400 16105 16106 meson: Build libvbe.so 16107 16108 Reviewed-by: Eric Anholt <eric@anholt.net> 16109 Signed-off-by: Adam Jackson <ajax@redhat.com> 16110 16111commit 2f1a1be46c3366a47657fe7550bc59fb99acf91c 16112Author: Adam Jackson <ajax@redhat.com> 16113Date: Thu Apr 27 13:19:11 2017 -0400 16114 16115 meson: Build libshadowfb.so 16116 16117 Reviewed-by: Eric Anholt <eric@anholt.net> 16118 Signed-off-by: Adam Jackson <ajax@redhat.com> 16119 16120commit bdce17959cd17fa6d46a1df361b53b5855ff62e8 16121Author: Adam Jackson <ajax@redhat.com> 16122Date: Thu Apr 27 13:15:55 2017 -0400 16123 16124 meson: Build libvgahw.so 16125 16126 Reviewed-by: Eric Anholt <eric@anholt.net> 16127 Signed-off-by: Adam Jackson <ajax@redhat.com> 16128 16129commit 5c7d92362b25670856091663e70d3e0dc2350efb 16130Author: Adam Jackson <ajax@redhat.com> 16131Date: Thu Apr 27 13:10:33 2017 -0400 16132 16133 meson: Build libfbdevhw.so 16134 16135 Reviewed-by: Eric Anholt <eric@anholt.net> 16136 Signed-off-by: Adam Jackson <ajax@redhat.com> 16137 16138commit 6413a87bcd3bfb91ac1536b41c37af0c6e47a9b6 16139Author: Adam Jackson <ajax@redhat.com> 16140Date: Thu Apr 27 12:58:54 2017 -0400 16141 16142 meson: Build libwfb.so 16143 16144 Reviewed-by: Eric Anholt <eric@anholt.net> 16145 Signed-off-by: Adam Jackson <ajax@redhat.com> 16146 16147commit 1597297067fd14d479a2279ad1ab20082317271a 16148Author: Adam Jackson <ajax@redhat.com> 16149Date: Thu Apr 27 12:50:13 2017 -0400 16150 16151 meson: Build libglx.so 16152 16153 Reviewed-by: Eric Anholt <eric@anholt.net> 16154 Signed-off-by: Adam Jackson <ajax@redhat.com> 16155 16156commit 47f047462dcb492203ff569a2b1528bff8cd3ff3 16157Author: Adam Jackson <ajax@redhat.com> 16158Date: Thu Apr 27 11:49:11 2017 -0400 16159 16160 meson: Fix linking the dmx utilities 16161 16162 Reviewed-by: Eric Anholt <eric@anholt.net> 16163 Signed-off-by: Adam Jackson <ajax@redhat.com> 16164 16165commit 5d785693a8843d18be806ea9f8caecc2dcbf1007 16166Author: Adam Jackson <ajax@redhat.com> 16167Date: Thu Apr 27 11:47:11 2017 -0400 16168 16169 meson: Factor out the AC_REPLACE_FUNCS stuff to its own library 16170 16171 This is all just stuff we wish we had in libc, and some of this gets 16172 used in eg. the dmx utilities build, so split it to its own library to 16173 avoid pulling in xserver stuff. 16174 16175 Reviewed-by: Eric Anholt <eric@anholt.net> 16176 Signed-off-by: Adam Jackson <ajax@redhat.com> 16177 16178commit 8659aac2ccc6649477f601908e1a4062fad8635e 16179Author: Adam Jackson <ajax@redhat.com> 16180Date: Thu Apr 27 12:52:29 2017 -0400 16181 16182 meson: Fix a typo 16183 16184 "bulid" isn't valid anything, and build_xwayland was already defined 16185 above. 16186 16187 Reviewed-by: Eric Anholt <eric@anholt.net> 16188 Signed-off-by: Adam Jackson <ajax@redhat.com> 16189 16190commit 112d0d7d01b98fb0d67910281dd1feeec125247b 16191Author: Karol Kosik <kkosik@nvidia.com> 16192Date: Wed Jul 22 16:42:07 2015 -0700 16193 16194 xfree86: Improved autoconfig drivers matching 16195 16196 Implementation of new drivers matching algorithm. New approach 16197 doesn't add duplicate drivers and ease drivers matching phase. 16198 16199 Signed-off-by: Karol Kosik <kkosik@nvidia.com> 16200 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 16201 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 16202 16203commit 1549e30372756cfdb85379fd8d93c769297ab6e6 16204Author: Eric Anholt <eric@anholt.net> 16205Date: Sat Mar 11 10:24:04 2017 -0800 16206 16207 Add a Meson build system alongside autotools. 16208 16209 This is a work in progress that builds Xvfb, Xephyr, Xwayland, Xnest, 16210 and Xdmx so far. The outline of Xquartz/Xwin support is in tree, but 16211 hasn't been built yet. The unit tests are also not done. 16212 16213 The intent is to build this as a complete replacement for the 16214 autotools system, then eventually replace autotools. meson is faster 16215 to generate the build, faster to run the bulid, shorter to write the 16216 build files in, and less error-prone than autotools. 16217 16218 v2: Fix indentation nits, move version declaration to project(), use 16219 existing meson_options for version-config.h's vendor name/web. 16220 16221 Signed-off-by: Eric Anholt <eric@anholt.net> 16222 Acked-by: Keith Packard <keithp@keithp.com> 16223 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16224 16225commit c7be7a688a78a34f61b90c0d95914e14b90b0cdc 16226Author: Eric Anholt <eric@anholt.net> 16227Date: Tue Apr 25 16:03:00 2017 -0700 16228 16229 Use #ifdef instead of #if for features to make Meson easier. 16230 16231 We mostly use #ifdef throughout the tree, and this lets the generated 16232 config.h files just be #define TOKEN instead of #define TOKEN 1. 16233 16234 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16235 Reviewed-by: Keith Packard <keithp@keithp.com> 16236 Reviewed-by: Adam Jackson <ajax@redhat.com> 16237 Signed-off-by: Eric Anholt <eric@anholt.net> 16238 16239commit ace6bfd5901e19045371f9bd18b997b37f55b589 16240Author: Eric Anholt <eric@anholt.net> 16241Date: Tue Apr 25 16:02:59 2017 -0700 16242 16243 dix: Remove a redundant #define 16244 16245 Found when the meson conversion set the symbol to defined, instead of 16246 defined to 1. 16247 16248 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16249 Reviewed-by: Keith Packard <keithp@keithp.com> 16250 Reviewed-by: Adam Jackson <ajax@redhat.com> 16251 Signed-off-by: Eric Anholt <eric@anholt.net> 16252 16253commit 27a6b9f7c84c914d0f5909ec1069d72f5035bc04 16254Author: Michel Dänzer <michel.daenzer@amd.com> 16255Date: Wed Apr 26 18:31:08 2017 +0900 16256 16257 os: Handle SIGABRT 16258 16259 Without this, assertion failures can make life hard for users and those 16260 trying to help them. 16261 16262 v2: 16263 * Change commit log wording slightly to "can make life hard", since 16264 apparently e.g. logind can alleviate that somewhat. 16265 * Set default handler for SIGABRT in 16266 hw/xfree86/common/xf86Init.c:InstallSignalHandlers() and 16267 hw/xquartz/quartz.c:QuartzInitOutput() (Eric Anholt) 16268 16269 Reviewed-by: Eric Anholt <eric@anholt.net> 16270 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 16271 16272commit 23f2f1932a1d3f36468eaf735ae34934d246567b 16273Author: Eric Anholt <eric@anholt.net> 16274Date: Tue Apr 25 11:33:50 2017 -0700 16275 16276 modesetting: Add the "DPI" connector type. 16277 16278 The number for it was merged to drm_mode.h in kernel 4.7, and the 16279 output_names[] array just requires that we slot in new strings in 16280 order. 16281 16282 Signed-off-by: Eric Anholt <eric@anholt.net> 16283 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 16284 16285commit fec9607c8e0a84dc86466c638d00b502f21ec622 16286Author: Eric Anholt <eric@anholt.net> 16287Date: Fri Apr 21 12:03:45 2017 -0700 16288 16289 Remove support for unaligned swaps. 16290 16291 The previous misc.h code went out of its way to allow swapping of 16292 unaligned pointers to values. However, the members of an X 16293 request/response are always naturally aligned within the struct, and 16294 the buffers containing a request/response will also be aligned to at 16295 least 8 bytes, so we can just drop it. 16296 16297 text data bss dec hex filename 16298 before: 2215167 51552 132016 2398735 249a0f hw/xfree86/Xorg 16299 after: 2214919 51552 132016 2398487 249917 hw/xfree86/Xorg 16300 16301 Signed-off-by: Eric Anholt <eric@anholt.net> 16302 Reviewed-by: Keith Packard <keithp@keithp.com> 16303 16304commit 4552238960fc05ff885bcabbc24d1489370fbd89 16305Author: Eric Anholt <eric@anholt.net> 16306Date: Tue Mar 28 13:09:30 2017 -0700 16307 16308 Add unit tests for the byte swapping macros. 16309 16310 Peter noted a weirdness in my new bswap code, which could use some 16311 tests to justify it. 16312 16313 Signed-off-by: Eric Anholt <eric@anholt.net> 16314 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16315 16316commit 563b6ee873b898c0f3e3671cf6adaf91def5d92a 16317Author: Eric Anholt <eric@anholt.net> 16318Date: Mon Mar 27 14:59:06 2017 -0700 16319 16320 Rewrite the byte swapping macros. 16321 16322 The clever pointer tricks were actually not working, and we were doing 16323 the byte-by-byte moves in general. By just doing the memcpy and 16324 obvious byte swap code, we end up generating actual byte swap 16325 instructions, thanks to optimizing compilers. 16326 16327 text data bss dec hex filename 16328 before: 2240807 51552 132016 2424375 24fe37 hw/xfree86/Xorg 16329 after: 2215167 51552 132016 2398735 249a0f hw/xfree86/Xorg 16330 16331 Signed-off-by: Eric Anholt <eric@anholt.net> 16332 Reviewed-by: Keith Packard <keithp@keithp.com> 16333 16334commit 5ef4e785131bb30e774a8175099c0432537533fa 16335Author: Eric Anholt <eric@anholt.net> 16336Date: Fri Mar 31 17:38:27 2017 -0700 16337 16338 Remove default defines of some directories. 16339 16340 The build defines these, so having the defaults is just a way for the 16341 build system's configuration to get out of sync with the code. 16342 16343 v2: Drop #ifndefs around the other two defines. 16344 16345 Signed-off-by: Eric Anholt <eric@anholt.net> 16346 Reviewed-by: Keith Packard <keithp@keithp.com> 16347 16348commit da27ca84b4324b68037d6ec19a73b9e0a5ef10bc 16349Author: Eric Anholt <eric@anholt.net> 16350Date: Fri Mar 31 17:33:52 2017 -0700 16351 16352 xorg: Change __XCONFIGFILE__ to XCONFIGFILE (and DIR) to fix scan.c. 16353 16354 parser/scan.c was checking for #ifdef XCONFIGFILE and XCONFIGDIR and 16355 defaulting to "xorg.conf", and "xorg.conf.d", so if you had changed 16356 __XCONFIGFILE__ to anything else, it would have got out of sync. 16357 Settle on the name without gratuitous underscores. 16358 16359 Signed-off-by: Eric Anholt <eric@anholt.net> 16360 Reviewed-by: Keith Packard <keithp@keithp.com> 16361 16362commit f81343dd2fff49c70f53e3efdd40f8f89a38e460 16363Author: Alan Coopersmith <alan.coopersmith@oracle.com> 16364Date: Sat Apr 22 18:10:42 2017 -0700 16365 16366 Mark test_data static in Xinput protocol tests to avoid clashes 16367 16368 Fixes warning from Solaris linker: 16369 ld: warning: symbol 'test_data' has differing sizes: 16370 (file tests-protocol-xigetselectedevents.o value=0x418; 16371 file tests-protocol-xigetclientpointer.o value=0x18); 16372 tests-protocol-xigetselectedevents.o definition taken 16373 16374 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 16375 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16376 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16377 16378commit 7bfb87a2137853295ecc9e544a15626cfd773a02 16379Author: Olivier Fourdan <ofourdan@redhat.com> 16380Date: Fri Apr 21 09:05:51 2017 +0200 16381 16382 glamor: an FBO is not needed for Xv pixmaps 16383 16384 It appears that on some hardware/diver combo such as nv30/nouveau, using 16385 GL_ALPHA as format for 8-bit depth will cause an incomplete attachment 16386 error (GL_FRAMEBUFFER_INCOMPLETE_ATTACHMENT) when trying to bind the 16387 texture. 16388 16389 As a result, the FBO is NULL and glamor segfaults when trying to access 16390 the FBO width/height in pixmap_priv_get_scale() in glamor_xv_render(). 16391 16392 This happens with glamor-xv which uses 8-bit pixmaps, meaning that on 16393 such hardware/driver, trying to play a video using Xv will lead to a 16394 crash of the Xserver. This affects Xwayland, Xephyr, modesetting driver 16395 with glamor accel. 16396 16397 But the use of an FBO is not actually needed for glamox-xv, so by 16398 disabling FBO at pixmap creation, we can avoid the issue entirely. 16399 16400 Fix suggested by Eric Anholt <eric@anholt.net> 16401 16402 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 16403 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=100710 16404 Fixes: https://bugzilla.redhat.com/1412814 16405 Reviewed-by: Eric Anholt <eric@anholt.net> 16406 16407commit be80a3cb48a7860b9ed985b123f4d8a3b4ae3c89 16408Author: Eric Anholt <eric@anholt.net> 16409Date: Mon Mar 27 14:21:43 2017 -0700 16410 16411 glx: Use the same endian swapping as the rest of the server. 16412 16413 This dumps a ton of configure-time checks for system endian macros. 16414 Given that we're marking the mixed-endian fixup code as cold, getting 16415 at the system macros is a waste of code. 16416 16417 Signed-off-by: Eric Anholt <eric@anholt.net> 16418 Reviewed-by: Keith Packard <keithp@keithp.com> 16419 16420commit dae97e1bb4d4f86db118c22dfeea0eef0d3f8bdd 16421Author: Eric Anholt <eric@anholt.net> 16422Date: Thu Mar 23 17:06:17 2017 -0700 16423 16424 configure: Stop checking for XdmcpWrap 16425 16426 As far back as the initial import, it seems to have been exposed, and 16427 there's no explanation why the test happened in the initial xserver 16428 import. 16429 16430 Signed-off-by: Eric Anholt <eric@anholt.net> 16431 Reviewed-by: Keith Packard <keithp@keithp.com> 16432 16433commit cc882af336c8963a2cba0c93e29869ef59b19818 16434Author: Roman Gilg <subdiff@gmail.com> 16435Date: Tue Apr 11 15:52:58 2017 +0200 16436 16437 xwayland: Small comment edit 16438 16439 Be more precise in describing the return value. 16440 16441 Signed-off-by: Roman Gilg <subdiff@gmail.com> 16442 Reviewed-by: Daniel Stone <daniels@collabora.com> 16443 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16444 16445commit 41dafcc2a2942fc4c94ce3cbafc4a1b413c460c3 16446Author: Michel Dänzer <michel.daenzer@amd.com> 16447Date: Wed Apr 12 17:58:05 2017 +0900 16448 16449 xfree86/modes: Make colormap/gamma glue code work with RandR disabled 16450 16451 E.g. because Xinerama is enabled. 16452 16453 Fixes crash on startup and wrong colours in that case. 16454 16455 Bugzilla: https://bugs.freedesktop.org/100293 16456 Bugzilla: https://bugs.freedesktop.org/100294 16457 Fixes: 62f44052573b ("xfree86/modes: Move gamma initialization to 16458 xf86RandR12Init12 v2") 16459 Tested-by: Mariusz Bialonczyk <manio@skyboo.net> 16460 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 16461 16462commit bcee1b76aa0db8525b491485e90b8740763d7de6 16463Author: Martin Peres <martin.peres@linux.intel.com> 16464Date: Mon Apr 10 16:48:21 2017 +0300 16465 16466 modesetting: re-set the crtc's mode when link-status goes BAD 16467 16468 Despite all the careful planning of the kernel, a link may become 16469 insufficient to handle the currently-set mode. At this point, the 16470 kernel should mark this particular configuration as being broken 16471 and potentially prune the mode before setting the offending connector's 16472 link-status to BAD and send the userspace a hotplug event. This may 16473 happen right after a modeset or later on. 16474 16475 Upon receiving a hot-plug event, we iterate through the connectors to 16476 re-apply the currently-set mode on all the connectors that have a 16477 link-status property set to BAD. The kernel may be able to get the 16478 link to work by dropping to using a lower link bpp (with the same 16479 display bpp). However, the modeset may fail if the kernel has pruned 16480 the mode, so to make users aware of this problem a warning is outputed 16481 in the logs to warn about having a potentially-black display. 16482 16483 This patch does not modify the current behaviour of always propagating 16484 the events to the randr clients. This allows desktop environments to 16485 re-probe the connectors and select a new resolution based on the new 16486 (currated) mode list if a mode disapeared. This behaviour is expected in 16487 order to pass the Display Port compliance tests. 16488 16489 Signed-off-by: Martin Peres <martin.peres@linux.intel.com> 16490 Reviewed-by: Eric Anholt <eric@anholt.net> 16491 16492commit ea91996a9c2081d0692f5c2f79a9fda9effba358 16493Author: Daniel Stone <daniels@collabora.com> 16494Date: Tue Apr 11 07:55:00 2017 +0100 16495 16496 README: Add link to fd.o CoC 16497 16498 The Contributor Covenant applies to all fd.o-hosted resources, including 16499 lists and Bugzilla. Add a note to the README, reminding people to act 16500 like human beings. 16501 16502 Reviewed-by: Adam Jackson <ajax@redhat.com> 16503 Signed-off-by: Daniel Stone <daniels@collabora.com> 16504 16505commit 8bdf340e478524262c85fb81687d7f02b14ad71c 16506Author: Daniel Stone <daniels@collabora.com> 16507Date: Tue Apr 11 07:54:59 2017 +0100 16508 16509 README: HTTPS everywhere 16510 16511 All fd.o HTTP services are now HTTPS-only (with redirects, but still). 16512 16513 Reviewed-by: Adam Jackson <ajax@redhat.com> 16514 Signed-off-by: Daniel Stone <daniels@collabora.com> 16515 16516commit 14d2fe74f4e51c5b37eab4b7475c804a0340b530 16517Author: Adam Jackson <ajax@redhat.com> 16518Date: Fri Apr 7 10:24:54 2017 -0400 16519 16520 xephyr: Check for host XVideo support before trying to use it 16521 16522 Otherwise xcb will treat our attempt to send xv requests as a connection 16523 error (quite reasonably: we're asking it to emit a request for which 16524 there is no defined major opcode), and we'll die quietly the first time 16525 we hit KdBlockhandler. 16526 16527 Signed-off-by: Adam Jackson <ajax@redhat.com> 16528 Reviewed-by: Eric Anholt <eric@anholt.net> 16529 16530commit 7dfb9608d080b46a7b21b57dc8aa2f52670addce 16531Author: Adam Jackson <ajax@redhat.com> 16532Date: Wed Apr 5 13:21:10 2017 -0400 16533 16534 glx: Remove enter/leave server API 16535 16536 Dead code since: 16537 16538 commit 8aacf47e1778d8b72811b025a82452b933d3c1f2 16539 Author: Adam Jackson <ajax@redhat.com> 16540 Date: Fri Oct 4 12:58:19 2013 -0400 16541 16542 glx: Remove DRI1 AIGLX (v2) 16543 16544 Signed-off-by: Adam Jackson <ajax@redhat.com> 16545 16546commit 0c8e6ed85810e96d84173a52d628863802a78d82 16547Author: Daniel Stone <daniels@collabora.com> 16548Date: Fri Apr 7 14:27:58 2017 +0100 16549 16550 modesetting: Set correct DRM event context version 16551 16552 DRM_EVENT_CONTEXT_VERSION is the latest context version supported by 16553 whatever version of libdrm is present. modesetting was blindly asserting 16554 it supported whatever version that may be, even if it actually didn't. 16555 16556 With libdrm 2.4.78, setting a higher context version than 2 will attempt 16557 to call the page_flip_handler2 vfunc if it was non-NULL, which being a 16558 random chunk of stack memory, it might well have been. 16559 16560 Set the version as 2, which should be bumped only with the appropriate 16561 version checks. 16562 16563 Reviewed-by: Adam Jackson <ajax@redhat.com> 16564 Signed-off-by: Daniel Stone <daniels@collabora.com> 16565 16566commit e4d0757fc2620fd3182448c5e4b5a55a509594d3 16567Author: Adam Jackson <ajax@redhat.com> 16568Date: Wed Mar 29 11:35:32 2017 -0400 16569 16570 xfree86: Remove driver entity hooks and private 16571 16572 No driver is using these, as far as I know. 16573 16574 v2: Tripwire the entity hook arguments to xf86Config*Entity, fix 16575 documentation (Eric Anholt) 16576 16577 Signed-off-by: Adam Jackson <ajax@redhat.com> 16578 Reviewed-by: Eric Anholt <eric@anholt.net> 16579 16580commit 33604187674ec78b2c0bf7f67af250acc80cf23a 16581Author: Adam Jackson <ajax@redhat.com> 16582Date: Fri Mar 24 12:30:58 2017 -0400 16583 16584 dpms: Consolidate a bunch of stuff into Xext/dpms.c 16585 16586 Most of this is a legacy of the old "extmod" design where you could load 16587 _some_ extensions dynamically but only if the server had been built with 16588 support for them in the first place. 16589 16590 Note that since we now only initialize the DPMS extension if at least 16591 one screen supports it, we no longer need DPMSCapableFlag: if it would 16592 be false, we would never read its value. 16593 16594 Signed-off-by: Adam Jackson <ajax@redhat.com> 16595 Reviewed-by: Eric Anholt <eric@anholt.net> 16596 16597commit 7f1ef9289d974fc2bacc968ae0b5d7714382cb9e 16598Author: Adam Jackson <ajax@redhat.com> 16599Date: Fri Mar 24 12:30:57 2017 -0400 16600 16601 dix: Lift DPMS to a screen hook 16602 16603 Following on from the previous change, this adds a DPMS hook to the 16604 ScreenRec and uses that to infer DPMS support. As a result we can drop 16605 the dpms stub code from Xext. 16606 16607 Signed-off-by: Adam Jackson <ajax@redhat.com> 16608 Reviewed-by: Eric Anholt <eric@anholt.net> 16609 16610commit 8ed0b00fceb34cdb54a0ea113c3cdff3b4c9e7e1 16611Author: Adam Jackson <ajax@redhat.com> 16612Date: Fri Mar 24 12:30:56 2017 -0400 16613 16614 xfree86: Clean up DPMS support 16615 16616 Rather than setting up a per-screen private, just conditionally 16617 initialize ScrnInfoRec::DPMSSet based on the config options, and inspect 16618 that to determine whether DPMS is supported. 16619 16620 We also move the "turn the screen back on at CloseScreen" logic into the 16621 DPMS extension's (new) reset hook. This would be a behavior change for 16622 the non-xfree86 servers, if any of them had non-stub DPMS support. 16623 16624 Signed-off-by: Adam Jackson <ajax@redhat.com> 16625 Reviewed-by: Eric Anholt <eric@anholt.net> 16626 16627commit 21eda7464d0e13ac6558edaf6531c3d3251e05df 16628Author: Tobias Stoeckmann <tobias@stoeckmann.org> 16629Date: Sun Mar 12 14:21:38 2017 +0100 16630 16631 dmx: Fix null pointer dereference 16632 16633 A null pointer dereference can occur in dmxSync, because TimerForce 16634 does not handle a null pointer. 16635 16636 dmxSyncTimer is set to NULL a few lines above on a certain condition, 16637 which happened on my machine. The explicit NULL check allowed me to 16638 start Xdmx again without a segmentation fault. 16639 16640 Reviewed-by: Adam Jackson <ajax@redhat.com> 16641 16642commit 5cdfcdef819efa6406807254cb3c7aba6ba5475e 16643Author: Matt Turner <mattst88@gmail.com> 16644Date: Thu Mar 23 19:37:46 2017 -0700 16645 16646 xkb: Correct type of XkbDfltAccessXOptions. 16647 16648 Declared as unsigned short in xkb/xkbAccessX.c. 16649 16650 Signed-off-by: Matt Turner <mattst88@gmail.com> 16651 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16652 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16653 16654commit f4b78286ea8d4c94a913a02a3579cbe55e8f1f6b 16655Author: Adam Jackson <ajax@redhat.com> 16656Date: Mon Mar 20 13:08:19 2017 -0400 16657 16658 glx: Fix MakeCurrent with no drawables 16659 16660 GLX_ARB_create_context, which we aspire to support, allows making GL 3.0 16661 or newer contexts current with null current drawables. Strictly this 16662 might not be legal for pre-3.0 contexts, but there's no harm in allowing 16663 it anyway. 16664 16665 Signed-off-by: Adam Jackson <ajax@redhat.com> 16666 16667commit e2e50c509718ea8c7f4684dc18f5ce66e526792b 16668Author: Eric Anholt <eric@anholt.net> 16669Date: Wed Mar 22 15:11:35 2017 -0700 16670 16671 xserver: Unifdef HAVE_CONFIG_H 16672 16673 The X Server never generates a global config.h, and instead all these 16674 paths are including dix-config.h or xorg-config.h. 16675 16676 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16677 Reviewed-by: Daniel Stone <daniels@collabora.com> 16678 Signed-off-by: Eric Anholt <eric@anholt.net> 16679 16680commit 266af991c8e41e62f054ac9870468a9ba2aaa9f2 16681Author: Eric Anholt <eric@anholt.net> 16682Date: Wed Mar 22 15:11:34 2017 -0700 16683 16684 glamor: Fix some formatting that confused the unifdef command. 16685 16686 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16687 Reviewed-by: Daniel Stone <daniels@collabora.com> 16688 Signed-off-by: Eric Anholt <eric@anholt.net> 16689 16690commit 2da405759fbabd4b918838a0de728c1911b4a2fa 16691Author: Eric Anholt <eric@anholt.net> 16692Date: Wed Mar 22 15:11:33 2017 -0700 16693 16694 xserver: Check the right HAVE_*_CONFIG_H. 16695 16696 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16697 Reviewed-by: Daniel Stone <daniels@collabora.com> 16698 Signed-off-by: Eric Anholt <eric@anholt.net> 16699 16700commit 190c2adf4a237a4c0dc13d7566ca91662271d3e4 16701Author: Eric Anholt <eric@anholt.net> 16702Date: Wed Mar 22 15:11:32 2017 -0700 16703 16704 kdrive: Drop kdrive-config.h. 16705 16706 It had nothing left in it that was used but wasn't in dix-config.h. 16707 16708 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16709 Reviewed-by: Daniel Stone <daniels@collabora.com> 16710 Signed-off-by: Eric Anholt <eric@anholt.net> 16711 16712commit 5c7ed785e3bdb9f0fbf8fbfdc93b5fdd2b2c7dbf 16713Author: Eric Anholt <eric@anholt.net> 16714Date: Wed Mar 22 15:11:31 2017 -0700 16715 16716 kdrive: Remove vestiges of TSLIB support. 16717 16718 The actual code was gone in 27819950e4158326e0f83a30f2e8968b932625ef, 16719 but some checks remained. 16720 16721 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16722 Reviewed-by: Daniel Stone <daniels@collabora.com> 16723 Signed-off-by: Eric Anholt <eric@anholt.net> 16724 16725commit 1b1a8c481773e15142202a7b761d614b33f12577 16726Author: Eric Anholt <eric@anholt.net> 16727Date: Wed Mar 22 15:11:30 2017 -0700 16728 16729 kdrive: Unifdef KDRIVE_EVDEV. 16730 16731 ajax deleted the evdev driver in the removal of fbdev and the linux 16732 backend. 16733 16734 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 16735 Reviewed-by: Daniel Stone <daniels@collabora.com> 16736 Signed-off-by: Eric Anholt <eric@anholt.net> 16737 16738commit ad6689b7db0a4b0d22716acedae98462a68ea1ae 16739Author: Mariusz Bialonczyk <manio@skyboo.net> 16740Date: Fri Mar 17 11:20:22 2017 +0100 16741 16742 configure.ac: fix checking for libdrm version after 9232835bd 16743 16744 No matter what libdrm version was installed, it always set 16745 the GLAMOR_HAS_DRM_NAME_FROM_FD_2 conditional to 1. 16746 This obviously leads to compilation problems. 16747 16748 Fixes: 9232835bd16b ("glamor: use drmGetDeviceNameFromFD2 when available") 16749 Signed-off-by: Mariusz Bialonczyk <manio@skyboo.net> 16750 Reviewed-and-Tested-by: Qiang Yu <Qiang.Yu@amd.com> 16751 16752commit eb2cf11724df5cbe0be3cce1ad4c61ba61e8ec6b 16753Author: Michel Dänzer <michel.daenzer@amd.com> 16754Date: Thu Mar 23 12:42:37 2017 +0900 16755 16756 Xephyr: Handle source-only pictures in ephyrPrepare/DoneComposite 16757 16758 There is no pixmap associated with source-only pictures. 16759 16760 Fixes Xephyr -fakexa crashing on startup. 16761 16762 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 16763 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 16764 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 16765 16766commit fdc79fe72bc0b97776df2c3a664076c60e08a87c 16767Author: Michel Dänzer <michel.daenzer@amd.com> 16768Date: Thu Mar 9 17:34:55 2017 +0900 16769 16770 edid: Prune duplicates after adding modes from DDC 16771 16772 Multiple calls to xf86EdidMonitorSet (which can be triggered e.g. by 16773 running xrandr) would potentially keep adding the same modes, causing 16774 the Monitor->Modes list to keep growing larger and using up more memory. 16775 16776 Fix this by calling xf86PruneDuplicateModes after adding the modes 16777 returned by xf86DDCGetModes. This makes sure there's only one instance 16778 of each unique mode in the list. 16779 16780 v2: 16781 * Replace semicolon with {} for empty for loop (Emil Velikov) 16782 * Slightly tweak commit log to avoid minor inaccuracy about what 16783 xf86PruneDuplicateModes does 16784 16785 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99521 16786 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 16787 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 16788 16789commit 4e76c2711740b775242170d8e5ee56c7e6829cce 16790Author: Adam Jackson <ajax@redhat.com> 16791Date: Mon Mar 20 13:42:12 2017 -0400 16792 16793 dix: Remove PRIVATE_DAMAGE 16794 16795 None of this is actually wired up to anything, so we can also remove the 16796 devPrivates from the DamageRec. The DamageExtRec is what would need 16797 devPrivates for selinux labeling, in principle. 16798 16799 Signed-off-by: Adam Jackson <ajax@redhat.com> 16800 Reviewed-by: Eric Anholt <eric@anholt.net> 16801 16802commit 40c12a76c2ae57adefd3b1d412387ebbfe2fb784 16803Author: Tobias Stoeckmann <tobias@stoeckmann.org> 16804Date: Sun Mar 19 17:55:07 2017 +0100 16805 16806 record: Fix OOB access in ProcRecordUnregisterClients 16807 16808 If a client sends a RecordUnregisterClients request with an nClients 16809 field larger than INT_MAX / 4, an integer overflow leads to an 16810 out of boundary access in RecordSanityCheckClientSpecifiers. 16811 16812 An example line with libXtst would be: 16813 XRecordUnregisterClients(dpy, rc, clients, 0x40000001); 16814 16815 Reviewed-by: Adam Jackson <ajax@redhat.com> 16816 16817commit 1ad230682338a9d2fc6eca6966a5bebb007df32c 16818Author: Adam Jackson <ajax@redhat.com> 16819Date: Mon Sep 26 13:28:42 2016 -0400 16820 16821 glamor: Avoid software fallback for planemasked ZPixmap GetImage 16822 16823 Same trick as in fb: just do a normal GetImage and deal with the 16824 planemask on the CPU if you have to. Since the software fallback hit for 16825 glamor is pretty brutal, this is a much more impressive win for glamor 16826 than it was for fb: 16827 16828 11100.0 87700.0 (7.901) (copy 0xaaaaaaaa) ShmGetImage 10x10 square 16829 9840.0 47800.0 (4.858) (copy 0xaaaaaaaa) ShmGetImage 100x100 square 16830 1550.0 4240.0 (2.735) (copy 0xaaaaaaaa) ShmGetImage 500x500 square 16831 9450.0 78900.0 (8.349) (0xaaaaaaaa) GetImage 10x10 square 16832 6910.0 30900.0 (4.472) (0xaaaaaaaa) GetImage 100x100 square 16833 431.0 2020.0 (4.687) (0xaaaaaaaa) GetImage 500x500 square 16834 16835 Measured with Xephyr -glamor on Skylake GT3e. 16836 16837 Reviewed-by: Eric Anholt <eric@anholt.net> 16838 Signed-off-by: Adam Jackson <ajax@redhat.com> 16839 16840commit 4aa35c46dab72bc945981f6fd29e494133bc2b0a 16841Author: Adam Jackson <ajax@redhat.com> 16842Date: Wed Sep 28 11:18:34 2016 -0400 16843 16844 fb: Handle ZPixmap planemask in GetImage the other way around 16845 16846 Formerly we'd zero the image data and then pull out a plane at a time. 16847 It's faster to apply the planemask after the fact, since that turns the 16848 GetImage into a memcpy: 16849 16850 100000.0 101000.0 (1.010) (copy 0xaaaaaaaa) ShmGetImage 10x10 square 16851 42400.0 59400.0 (1.401) (copy 0xaaaaaaaa) ShmGetImage 100x100 square 16852 3040.0 5280.0 (1.737) (copy 0xaaaaaaaa) ShmGetImage 500x500 square 16853 96100.0 95200.0 (0.991) (0xaaaaaaaa) GetImage 10x10 square 16854 29600.0 36800.0 (1.243) (0xaaaaaaaa) GetImage 100x100 square 16855 1850.0 2620.0 (1.416) (0xaaaaaaaa) GetImage 500x500 square 16856 16857 Measured with Xvfb at depth 24 on Skylake i7-6560U. 16858 16859 Reviewed-by: Eric Anholt <eric@anholt.net> 16860 Signed-off-by: Adam Jackson <ajax@redhat.com> 16861 16862commit 368f60d461421fe5e2bbd90652d6ac858dbff8fe 16863Author: Kenneth Graunke <kenneth@whitecape.org> 16864Date: Fri Mar 17 13:45:04 2017 -0700 16865 16866 dri2: Sync i965_pci_ids.h from Mesa. 16867 16868 Copied from Mesa with no modifications. Gives us Geminilake PCI IDs. 16869 16870 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 16871 Acked-by: Eric Anholt <eric@anholt.net> 16872 16873commit 8805a48ed35afb2ca66315656c1575ae5a01c639 16874Author: Olivier Fourdan <ofourdan@redhat.com> 16875Date: Fri Mar 17 15:58:26 2017 +0100 16876 16877 glamor: avoid a crash if texture allocation failed 16878 16879 Texture creation in _glamor_create_tex() can fail if a GL_OUT_OF_MEMORY 16880 is raised, in which case the texture returned is zero. 16881 16882 But the texture value is not checked in glamor_create_fbo() and glamor 16883 will abort in glamor_pixmap_ensure_fb() because the fbo->tex is 0: 16884 16885 Truncated backtrace: 16886 Thread no. 1 (10 frames) 16887 #4 glamor_pixmap_ensure_fb at glamor_fbo.c:57 16888 #5 glamor_create_fbo_from_tex at glamor_fbo.c:112 16889 #6 glamor_create_fbo at glamor_fbo.c:159 16890 #7 glamor_create_fbo_array at glamor_fbo.c:210 16891 #8 glamor_create_pixmap at glamor.c:226 16892 #9 compNewPixmap at compalloc.c:536 16893 #10 compAllocPixmap at compalloc.c:605 16894 #11 compCheckRedirect at compwindow.c:167 16895 #12 compRealizeWindow at compwindow.c:267 16896 #13 RealizeTree at window.c:2617 16897 16898 Check the value returned by _glamor_create_tex() in glamor_create_fbo() 16899 and return NULL in the texture is zero. 16900 16901 All callers of glamor_create_fbo() actually check the returned value and 16902 will use a fallback code path if it's NULL. 16903 16904 Please cherry-pick this to active stable branches. 16905 16906 Bugzilla: https://bugzilla.redhat.com/1433305 16907 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 16908 Reviewed-by: Eric Anholt <eric@anholt.net> 16909 16910commit 0803918e64262482035f042e5e1f2a571d3dea1b 16911Author: Adam Jackson <ajax@redhat.com> 16912Date: Sat Sep 24 19:40:52 2016 -0400 16913 16914 fb: Remove 24bpp support (v3) 16915 16916 v2: 16917 - Require power-of-two bpp in ScreenInit 16918 - Eliminate fbCreatePixmapBpp 16919 16920 v3 16921 - Squash in the exa and glamor changes so we can remove pRotatedPixmap 16922 in the same stroke. 16923 16924 Reviewed-by: Eric Anholt <eric@anholt.net> 16925 Signed-off-by: Adam Jackson <ajax@redhat.com> 16926 16927commit e33be78e2ab63abc84aa0baddff90bcefa9c183a 16928Author: Adam Jackson <ajax@redhat.com> 16929Date: Fri Feb 28 13:19:00 2014 -0500 16930 16931 xfree86: Remove 24bpp pixmap format support (v2) 16932 16933 There's really no reason to pretend to support this, apps hate it, all 16934 we're doing is giving people a way to injure themselves. It doesn't work 16935 anyway with any Radeon, any NVIDIA chip, or any Intel chip since i810. 16936 Rip out all the logic for handling 24bpp pixmaps and framebuffers, and 16937 silently ignore the old options that would ask for it. 16938 16939 The cirrus alpine driver has been updated to default to 16bpp, and both 16940 it and the i810 driver can now use the 32->24 conversion code in shadow 16941 if they want. All other drivers support 32bpp. Configurations that 16942 explicitly request 24bpp in order to fit in VRAM will be broken now 16943 though. 16944 16945 v2: Fix command line options to silently ignore 24bpp rather than fail 16946 16947 Reviewed-by: Eric Anholt <eric@anholt.net> 16948 Signed-off-by: Adam Jackson <ajax@redhat.com> 16949 16950commit 83c4297d2c4fd501a9d36bc0cb7d357a8d22394c 16951Author: Adam Jackson <ajax@redhat.com> 16952Date: Fri Mar 17 12:40:03 2017 -0400 16953 16954 ephyr: Don't clobber bitsPerPixel when using glamor 16955 16956 This ends up passing 0 as the bpp argument to fb screen setup, which is 16957 not really the best plan. 16958 16959 Reviewed-by: Eric Anholt <eric@anholt.net> 16960 Signed-off-by: Adam Jackson <ajax@redhat.com> 16961 16962commit fe0b297420fc1de8a7fab28457d0864b3182e967 16963Author: Eric Anholt <eric@anholt.net> 16964Date: Wed Mar 15 17:51:46 2017 -0700 16965 16966 glamor: Fix dashed line rendering. 16967 16968 We were binding the screen pixmap as the dash and sampling its alpha, 16969 which is usually just 1.0 (no dashing at all). 16970 16971 Please cherry-pick this to active stable branches. 16972 16973 Signed-off-by: Eric Anholt <eric@anholt.net> 16974 Reviewed-by: Keith Packard <keithp@keithp.com> 16975 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 16976 16977commit 5b5ba7dc3d544868035b7088e9aac62eb2511e10 16978Author: Adam Jackson <ajax@redhat.com> 16979Date: Thu Mar 16 14:19:37 2017 -0400 16980 16981 Revert "sdksyms: Tighten up the symbols we add to the magic table" 16982 16983 Enough people are reporting build issues pointing at this commit that 16984 it's not worth the benefit. 16985 16986 This reverts commit 3dad57b121fdf001e75fffa7e3007a9a0a154f67. 16987 16988commit e337de2d488a124e5fee0fdcb882567b68f1767d 16989Author: Adam Jackson <ajax@redhat.com> 16990Date: Wed Nov 2 12:49:25 2016 -0400 16991 16992 xinerama: Implement graphics exposures for window->pixmap copies (v4) 16993 16994 This code is using GetImage to accumulate a logical view of the window 16995 image (since the windows will be clipped to their containing screen), 16996 and then PutImage to load that back into the pixmap. What it wasn't 16997 doing was constructing a region for the obscured areas of the window and 16998 emitting graphics exposures for same. 16999 17000 v2: Fix coordinate translation when the source is the root window 17001 v3: Create sourceBox with the right coordinates initially instead of 17002 translating (Keith Packard) 17003 v4: Clamp the region to 15 bits to avoid overflow (Keith Packard) 17004 17005 Signed-off-by: Adam Jackson <ajax@redhat.com> 17006 17007commit f1f865e909090406841a9b9416ea6259a75c2086 17008Author: Adam Jackson <ajax@redhat.com> 17009Date: Mon Jan 23 11:31:36 2017 -0500 17010 17011 parser: Fix crash when xf86nameCompare(s1 = x, s2 = NULL) 17012 17013 Signed-off-by: Adam Jackson <ajax@redhat.com> 17014 17015commit 455051a0f1d2bc84f605c325f647bd64d414c47d 17016Author: Olivier Fourdan <ofourdan@redhat.com> 17017Date: Tue Mar 14 15:06:34 2017 +0100 17018 17019 glamor: Check glamor_set_destination_drawable() return value 17020 17021 Check the value returned by glamor_set_destination_drawable() and use 17022 the fallback code path where possible. 17023 17024 Bugzilla: https://bugzilla.redhat.com/1417575 17025 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17026 17027commit 04b4bad7c048fd077fe839f10634c99ef1e488af 17028Author: Olivier Fourdan <ofourdan@redhat.com> 17029Date: Tue Mar 14 14:58:26 2017 +0100 17030 17031 glamor: glamor_set_destination_drawable() can fail 17032 17033 The fbo_array of a given glamor pixmap can be NULL in some cases, as 17034 glamor_create_fbo_array() can fail to allocate the FBO array. 17035 17036 If this is the case, glamor_pixmap_fbo_at() will return NULL even though 17037 the box index is valid, and glamor_set_destination_drawable() simply 17038 assumes glamor_pixmap_fbo_at() will return an FBO prior to pass the 17039 value to glamor_set_destination_pixmap_fbo(), which will segfault. 17040 17041 We need a way for glamor_set_destination_drawable() to fail safely and 17042 let the caller know about the failure. 17043 17044 Add a boolean return value to glamor_set_destination_drawable() for that 17045 purpose. 17046 17047 Bugzilla: https://bugzilla.redhat.com/1417575 17048 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17049 17050commit b0ce1d088a863492f5de11e4dbde10af4261d892 17051Author: Olivier Fourdan <ofourdan@redhat.com> 17052Date: Tue Mar 14 15:22:32 2017 +0100 17053 17054 Xephyr: Check screen resources creation success 17055 17056 If the screen pixmap or the corresponding texture creation with glamor 17057 fails, exit cleanly with an error message instead of segfaulting. 17058 17059 Fixes: https://bugzilla.redhat.com/1431633 17060 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 17061 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17062 17063commit f40ff18c96e02ff18a367bf53feeb4bd8ee952a0 17064Author: Olivier Fourdan <ofourdan@redhat.com> 17065Date: Tue Mar 14 15:22:31 2017 +0100 17066 17067 glamor: Check for NULL pixmap in glamor_get_pixmap_texture() 17068 17069 glamor_create_pixmap() would return a NullPixmap if the given size is 17070 larger than the maximum size of a pixmap. 17071 17072 But glamor_get_pixmap_texture() won't check if the given pixmap is 17073 non-null, leading to a segfault if glamor_create_pixmap() failed. 17074 17075 This can be reproduced by passing Xephyr a very large screen width, 17076 e.g.: 17077 17078 $ Xephyr -glamor -screen 32768x1024 :10 17079 17080 (EE) 17081 (EE) Backtrace: 17082 (EE) 0: Xephyr (OsSigHandler+0x29) 17083 (EE) 1: /lib64/libpthread.so.0 (__restore_rt+0x0) 17084 (EE) 2: Xephyr (glamor_get_pixmap_texture+0x30) 17085 (EE) 3: Xephyr (ephyr_glamor_create_screen_resources+0xc6) 17086 (EE) 4: Xephyr (ephyrCreateResources+0x98) 17087 (EE) 5: Xephyr (dix_main+0x275) 17088 (EE) 6: /lib64/libc.so.6 (__libc_start_main+0xf1) 17089 (EE) 7: Xephyr (_start+0x2a) 17090 (EE) 8: ? (?+0x2a) [0x2a] 17091 (EE) 17092 (EE) Segmentation fault at address 0x0 17093 (EE) 17094 Fatal server error: 17095 (EE) Caught signal 11 (Segmentation fault). Server aborting 17096 (EE) 17097 Aborted (core dumped) 17098 17099 Bugzilla: https://bugzilla.redhat.com/1431633 17100 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 17101 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17102 17103commit 646bc74cded8b77d57312ae858a2ed01b0d17c21 17104Author: Adam Jackson <ajax@redhat.com> 17105Date: Fri Mar 3 14:51:51 2017 -0500 17106 17107 test: Fix distcheck failures 17108 17109 Signed-off-by: Adam Jackson <ajax@redhat.com> 17110 17111commit bca221609004aecbcc8928382ad29ce319d2eb6f 17112Author: Jon TURNEY <jon.turney@dronecode.org.uk> 17113Date: Mon Mar 6 17:44:19 2017 +0000 17114 17115 xfree86: work around a sdksyms problem with gcc5 on Cygwin 17116 17117 The linemarkers in the preprocessor output from gcc5 on Cygwin have 17118 canonicalized paths to included files (e.g. xserver/build/../include/misc.h 17119 is canonicalized to xserver/build/include/misc.h). (see gcc svn rev 210264, 17120 which causes the transformation performed by -fcanonical-system-headers to 17121 be applied to all include pathnames) 17122 17123 These canonicalized paths won't match $topdir, so sdksyms doesn't look at 17124 the contents of those headers for sdk exported symbols. 17125 17126 Workaround this by canonicalizing all the paths we consider, using readlink. 17127 17128 v2: 17129 Keep a cache of readlink results so it isn't quite so dreadfully slow. 17130 17131 Reviewed-by: Adam Jackson <ajax@redhat.com> 17132 17133commit ac15d4cecca377c5c31ab852c39bbd554ca48fe2 17134Author: Tobias Stoeckmann <tobias@stoeckmann.org> 17135Date: Mon Mar 13 19:13:14 2017 +0100 17136 17137 render: Fix out of boundary heap access 17138 17139 ProcRenderCreateRadialGradient and ProcRenderCreateConicalGradient must 17140 be protected against an integer overflow during length check. This is 17141 already included in ProcRenderCreateLinearGradient since the fix for 17142 CVE-2008-2362. 17143 17144 This can only be successfully exploited on a 32 bit system for an 17145 out of boundary read later on. Validated by using ASAN. 17146 17147 Reviewed-by: Adam Jackson <ajax@redhat.com> 17148 17149commit 0c1574d9882a91b2c1a046bf4ac5a9b138a37965 17150Author: Michel Dänzer <michel.daenzer@amd.com> 17151Date: Thu Mar 9 18:13:31 2017 +0900 17152 17153 glamor: Fix typo: "vec2_pos" -> "vec2 pos" 17154 17155 Fixes crash when trying to use dashed lines: 17156 17157 Failed to compile VS: 0:8(2): error: `vec2_pos' undeclared 17158 17159 Trivial. 17160 17161 Fixes: d8161aeb5089 ("glamor: Fix missing declaration in dash vertex 17162 shader") 17163 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99943 17164 17165commit 937527f9798d573ec82c2c508821899c229c018f 17166Author: Olivier Fourdan <ofourdan@redhat.com> 17167Date: Wed Mar 8 10:32:22 2017 +0100 17168 17169 xwayland: Monitor client states to destroy callbacks 17170 17171 Client resources can survive the client itself, in which case we 17172 may end up in our sync callback trying to access client's data after 17173 it's been freed/reclaimed. 17174 17175 Add a ClientStateCallback handler to monitor the client state changes 17176 and clear the sync callback set up by the glamor drm code if any. 17177 17178 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=100040 17179 Tested-by: Mark B <mark.blakeney@bullet-systems.net> 17180 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17181 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 17182 17183commit 669eb0a65430678fe97de9405e4e42dcb4417ed2 17184Author: Jon Turney <jon.turney@dronecode.org.uk> 17185Date: Wed Mar 8 23:46:27 2017 +0000 17186 17187 test: Use correct signature for XISetEventMask wrapper 17188 17189 This looks like it's always been wrong, but didn't matter before the single 17190 test binary changes as the wrapper was only in place for tests which didn't 17191 care about it's effects, so didn't forward to the real implementation. (see 17192 the changes to protocol-xiselectevents.c in 773fc07e) 17193 17194 This fixes the protocol_xigetselectedevents_test for me. I'm a bit 17195 surprised that passes for anyone at the moment. 17196 17197 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 17198 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 17199 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 17200 17201commit 9232835bd16b6948442f7a4588fb9376782cb814 17202Author: Qiang Yu <Qiang.Yu@amd.com> 17203Date: Tue Mar 7 11:53:23 2017 +0800 17204 17205 glamor: use drmGetDeviceNameFromFD2 when available 17206 17207 This is for glamor can support fd from DRM render node which is useful 17208 for a render only DDX. 17209 17210 Reviewed-by: Adam Jackson <ajax@redhat.com> 17211 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 17212 17213commit d4b7e0eaa4b2e97ce1dec653a2ae7d9621fe1431 17214Author: Olivier Fourdan <ofourdan@redhat.com> 17215Date: Tue Mar 7 10:28:35 2017 +0100 17216 17217 xwayland: clear cursor frame callback 17218 17219 After an X cursor is unrealized, the seat's corresponding x_cursor is 17220 cleared, but if a frame callback was pending at the time, it will 17221 remain and thus prevent any further cursor update, leaving the window 17222 with no cursor. 17223 17224 Make sure to destroy the frame callback, if any, when that occurs, so 17225 that next time a cursor needs to be set, it won't be ignored for a frame 17226 callback that will never be triggered. 17227 17228 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1389327 17229 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17230 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 17231 Reviewed-by: Rui Matos <tiagomatos@gmail.com> 17232 17233commit 64ca14b85e45b13628396f21d1903e311f92a9e1 17234Author: Olivier Fourdan <ofourdan@redhat.com> 17235Date: Thu Mar 2 10:19:26 2017 +0100 17236 17237 xwayland: make sure client is not gone in sync callback 17238 17239 in XWayland, dri3_send_open_reply() is called from a sync callback, so 17240 there is a possibility that the client might be gone when we get to the 17241 callback eventually, which leads to a crash in _XSERVTransSendFd() from 17242 WriteFdToClient() . 17243 17244 Check if clientGone has been set in the sync callback handler to avoid 17245 this. 17246 17247 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99149 17248 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100040 17249 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1416553 17250 Reviewed-by: Adam Jackson <ajax@redhat.com> 17251 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17252 Tested-by: Mark B <mark.blakeney@bullet-systems.net> 17253 17254commit 5c44169caed811e59a65ba346de1cadb46d266ec 17255Author: Adam Jackson <ajax@redhat.com> 17256Date: Thu Mar 2 17:20:30 2017 -0500 17257 17258 os: Squash missing declaration warning for timingsafe_memcmp 17259 17260 timingsafe_memcmp.c:21:1: warning: no previous prototype for ‘timingsafe_memcmp’ [-Wmissing-prototypes] 17261 timingsafe_memcmp(const void *b1, const void *b2, size_t len) 17262 17263 Signed-off-by: Adam Jackson <ajax@redhat.com> 17264 17265commit 0d7f05ed99b71a4641415c9f26e245c3bb24a9a0 17266Author: Adam Jackson <ajax@redhat.com> 17267Date: Wed Mar 1 16:13:59 2017 -0500 17268 17269 miarc: "Cache" arc span data for dashed arcs 17270 17271 This avoids recomputing the span data for every dash. x11perf thinks 17272 this is a pretty modest speedup: 17273 17274 832919.4 840471.1 ( 1.009) 100-pixel dashed ellipse 17275 672353.1 680652.2 ( 1.012) 100-pixel double-dashed ellipse 17276 13748.9 24287.9 ( 1.767) 100-pixel wide dashed ellipse 17277 9236.3 21298.2 ( 2.306) 100-pixel wide double-dashed ellipse 17278 17279 But part of the reason it's so modest there is that the arcs are 17280 relatively small (100 pixel diameter at line width 10, so ~6000 pixels) 17281 and the dashes relatively large (30 on 20 off so ~6 dashes per 17282 quadrant). 17283 17284 With larger arcs and finer dashes this is much more impressive. A fairly 17285 trivial testcase of a single 15000x13000 arc with the default {2, 2} 17286 dash pattern drops from ~3500 milliseconds to 10 milliseconds. 17287 17288 Reviewed-by: Keith Packard <keithp@keithp.com> 17289 Signed-off-by: Adam Jackson <ajax@redhat.com> 17290 17291commit 849c825855b82caf44d25edb8155bc9f17869256 17292Author: Adam Jackson <ajax@redhat.com> 17293Date: Wed Mar 1 16:13:58 2017 -0500 17294 17295 miarc: Make the caller free the arc span data 17296 17297 drawArc does some fairly expensive computation, but it's only sensitive 17298 to arc width/height. Thread the span data up through the call chain so 17299 it's at least possible for the caller to cache things. 17300 17301 Reviewed-by: Keith Packard <keithp@keithp.com> 17302 Signed-off-by: Adam Jackson <ajax@redhat.com> 17303 17304commit 9426c5500b72e1fe004fef4c3b259023c4ec49f7 17305Author: Adam Jackson <ajax@redhat.com> 17306Date: Wed Mar 1 16:13:57 2017 -0500 17307 17308 miarc: Style cleanup for miWideArc 17309 17310 Outdent, normalize comment formatting, and use 'goto out' idiom for 17311 error paths. No functional change. 17312 17313 Reviewed-by: Keith Packard <keithp@keithp.com> 17314 Signed-off-by: Adam Jackson <ajax@redhat.com> 17315 17316commit fe5c340046c8cf755b92763a49b2dc475e856a76 17317Author: Olivier Fourdan <ofourdan@redhat.com> 17318Date: Wed Mar 1 08:51:31 2017 +0100 17319 17320 xwayland: do not set checkRepeat on master kbd 17321 17322 keyboard_check_repeat() fetches the XWayland seat from the 17323 dev->public.devicePrivate do do its thing. 17324 17325 If a key event is sent programmatically through Xtest, our device is the 17326 virtual core keyboard and that has a dev->public.devicePrivate of NULL, 17327 leading to a segfault in keyboard_check_repeat(). 17328 17329 This is the case with "antimicro" which sends key events based on the 17330 joystick buttons. 17331 17332 Don't set the checkRepeat handler on the VCK since it cannot possibly work 17333 anyway and it has no effect on the actual checkRepeat intended functionality. 17334 17335 Bugzilla: https://bugzilla.redhat.com/1416244 17336 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17337 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 17338 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 17339 17340commit 27819950e4158326e0f83a30f2e8968b932625ef 17341Author: Adam Jackson <ajax@redhat.com> 17342Date: Thu Feb 16 12:46:23 2017 -0500 17343 17344 kdrive: Remove now-unused linux backend 17345 17346 With Xfbdev gone this has no consumers. 17347 17348 Signed-off-by: Adam Jackson <ajax@redhat.com> 17349 Acked-by: Alex Deucher <alexander.deucher@amd.com> 17350 17351commit feed7e3f982a7ac14f6fe85ed2e1ec4a83700841 17352Author: Adam Jackson <ajax@redhat.com> 17353Date: Thu Feb 16 12:46:22 2017 -0500 17354 17355 xfbdev: Remove 17356 17357 With the shadow framebuffer overallocation bug fixed (ref below), Xorg + 17358 fbdev has tens to hundreds of kilobytes more baseline memory usage than 17359 Xfbdev. That's not nothing, but it's little enough that we should focus 17360 our efforts on the server that actually gets development attention. 17361 17362 https://cgit.freedesktop.org/xorg/driver/xf86-video-fbdev/commit/?id=2c5eba8 17363 17364 Signed-off-by: Adam Jackson <ajax@redhat.com> 17365 Acked-by: Alex Deucher <alexander.deucher@amd.com> 17366 17367commit 35fbcb3f9987758bc26a87d5d7c033f4367cbd39 17368Author: Adam Jackson <ajax@redhat.com> 17369Date: Thu Feb 16 12:46:21 2017 -0500 17370 17371 xfake: Remove 17372 17373 We already have Xvfb for a dummy DDX. 17374 17375 Signed-off-by: Adam Jackson <ajax@redhat.com> 17376 Acked-by: Alex Deucher <alexander.deucher@amd.com> 17377 17378commit 3dad57b121fdf001e75fffa7e3007a9a0a154f67 17379Author: Adam Jackson <ajax@redhat.com> 17380Date: Thu Feb 16 14:00:03 2017 -0500 17381 17382 sdksyms: Tighten up the symbols we add to the magic table 17383 17384 The code as written would match anything declared extern. _X_EXPORT is 17385 what we really mean here. That's a macro, so check for what it expands 17386 to and skip if not found. 17387 17388 Signed-off-by: Adam Jackson <ajax@redhat.com> 17389 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 17390 17391commit 386fbbe410a1168b724136e54cf3bd37fb64ad4e 17392Author: Matthieu Herrb <matthieu@herrb.eu> 17393Date: Wed Mar 1 20:28:34 2017 +0100 17394 17395 Brown bag commit to fix 957e8d (arc4random_buf() support) 17396 17397 - typo in #ifdef check 17398 - also need to add AC_CHECK_FUNCS([arc4random_buf]) 17399 17400 Reported-by Eric Engestrom. Thanks 17401 17402 Reviewed-by: Adam Jackson <ajax@redhat.com> 17403 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 17404 17405commit dc7ceda90fdedb2212f105095fe8d449bfdb0eb4 17406Author: Adam Jackson <ajax@redhat.com> 17407Date: Thu Feb 16 14:56:45 2017 -0500 17408 17409 dispatch: Mark swapped dispatch as _X_COLD 17410 17411 This touches everything that ends up in the Xorg binary; the big missing 17412 part is GLX since that's all generated code. Cuts about 14k from the 17413 binary on amd64. 17414 17415 Signed-off-by: Adam Jackson <ajax@redhat.com> 17416 Reviewed-by: Eric Anholt <eric@anholt.net> 17417 17418commit 8c9909a99292b2fb4a86de694bb0029f61e35662 17419Author: Olivier Fourdan <ofourdan@redhat.com> 17420Date: Tue Feb 28 14:27:52 2017 +0100 17421 17422 xwayland: Make sure we have a focus window 17423 17424 During the InitInput() phase, the wayland events get dequeued so we 17425 can possibly end up calling dispatch_pointer_motion_event(). 17426 17427 If this occurs before xwl_seat->focus_window is set, it leads to a NULL 17428 pointer derefence and a segfault. 17429 17430 Check for xwl_seat->focus_window in both pointer_handle_frame() and 17431 relative_pointer_handle_relative_motion() prior to calling 17432 dispatch_pointer_motion_event() like it's done in 17433 pointer_handle_motion(). 17434 17435 Bugzilla: https://bugzilla.redhat.com/1410804 17436 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17437 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 17438 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 17439 17440commit c9cbdada755f1d00d164bef7833f1aef139ffeb0 17441Author: Adam Jackson <ajax@redhat.com> 17442Date: Mon Feb 20 10:24:14 2017 -0500 17443 17444 vfb: Bump default depth to 24 17445 17446 Signed-off-by: Adam Jackson <ajax@redhat.com> 17447 Acked-by: Keith Packard <keithp@keithp.com> 17448 Reviewed-by: Eric Anholt <eric@anholt.net> 17449 17450commit 2855f759b1e7bf7f5e57cac36c1f0d0e5ac1a683 17451Author: Matthieu Herrb <matthieu@herrb.eu> 17452Date: Tue Feb 28 19:18:56 2017 +0100 17453 17454 auth: remove AuthToIDFunc and associated functions. Not used anymore. 17455 17456 And the current code for MitToId has a use-after-free() issue. 17457 17458 [Also remove the actual implementations - ajax] 17459 17460 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 17461 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 17462 17463commit 957e8db38f27932d353e86e9aa69cf16778b18f1 17464Author: Matthieu Herrb <matthieu@herrb.eu> 17465Date: Tue Feb 28 19:18:43 2017 +0100 17466 17467 Use arc4random_buf(3) if available to generate cookies. 17468 17469 Reviewed-by: Adam Jackson <ajax@redhat.com> 17470 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 17471 17472commit d7ac755f0b618eb1259d93c8a16ec6e39a18627c 17473Author: Matthieu Herrb <matthieu@herrb.eu> 17474Date: Tue Feb 28 19:18:25 2017 +0100 17475 17476 Use timingsafe_memcmp() to compare MIT-MAGIC-COOKIES CVE-2017-2624 17477 17478 Provide the function definition for systems that don't have it. 17479 17480 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 17481 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 17482 17483commit 1b12249fd6d4d3b566043d556503f3f625b9b4ba 17484Author: Peter Hutterer <peter.hutterer@who-t.net> 17485Date: Thu Feb 23 10:21:49 2017 +1000 17486 17487 os: log a bug whenever WriteToClient is called from the input thread 17488 17489 The input thread should generate events, not send them. Make it easier to 17490 find the instances where it's doing so. 17491 17492 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 17493 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 17494 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 17495 17496commit 7d7788e0f2fbbf5d05cb918df11a5e1b7e047e48 17497Author: Olivier Fourdan <ofourdan@redhat.com> 17498Date: Fri Feb 17 09:16:24 2017 +0100 17499 17500 Revert "xwayland: bump wayland-protocols version to 1.7" 17501 17502 This reverts commit 371ff0c969a38a0013688391bbd7375bc7b6f933. 17503 17504 Reviewed-by: Adam Jackson <ajax@redhat.com> 17505 17506commit 9f4d308cdad957a354912f17febe5bcad7f44812 17507Author: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 17508Date: Wed Nov 23 13:30:53 2016 +0200 17509 17510 xwayland: use _XWAYLAND_ALLOW_COMMITS property 17511 17512 The X11 window manager (XWM) of a Wayland compositor can use the 17513 _XWAYLAND_ALLOW_COMMITS property to control when Xwayland sends 17514 wl_surface.commit requests. If the property is not set, the behaviour 17515 remains what it was. 17516 17517 XWM uses the property to inhibit commits until the window is ready to be 17518 shown. This gives XWM time to set up the window decorations and internal 17519 state before Xwayland does the first commit. XWM can use this to ensure 17520 the first commit carries fully drawn decorations and the window 17521 management state is correct when the window becomes visible. 17522 17523 Setting the property to zero inhibits further commits, and setting it to 17524 non-zero allows commits. Deleting the property allows commits. 17525 17526 When the property is changed from zero to non-zero, there will be a 17527 commit on next block_handler() call provided that some damage has been 17528 recorded. 17529 17530 Without this patch (i.e. with the old behaviour) Xwayland can and will 17531 commit the surface very soon as the application window has been realized 17532 and drawn into. This races with XWM and may cause visible glitches. 17533 17534 v3: 17535 - introduced a simple setter for xwl_window::allow_commits 17536 - split xwl_window_property_allow_commits() out of 17537 xwl_property_callback() 17538 - check MakeAtom(_XWAYLAND_ALLOW_COMMITS) 17539 17540 v2: 17541 - use PropertyStateCallback instead of XACE, based on the patch 17542 "xwayland: Track per-window support for netwm frame sync" by 17543 Adam Jackson 17544 - check property type is XA_CARDINAL 17545 - drop a useless memcpy() 17546 17547 Weston Bug: https://phabricator.freedesktop.org/T7622 17548 Reviewed-by: Adam Jackson <ajax@redhat.com> 17549 Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 17550 17551commit a6308cea602f688ac653e3466cd57767e02093a9 17552Author: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 17553Date: Thu Nov 24 11:54:44 2016 +0200 17554 17555 xwayland: fix 'buffer' may be used uninitialized warning 17556 17557 Fix the following warning due to --disable-glamor: 17558 17559 CC Xwayland-xwayland.o 17560 In file included from /home/pq/local/include/wayland-client.h:40:0, 17561 from xwayland.h:35, 17562 from xwayland.c:26: 17563 xwayland.c: In function ‘block_handler’: 17564 /home/pq/local/include/wayland-client-protocol.h:3446:2: warning: ‘buffer’ may be used uninitialized in this function [-Wmaybe-uninitialized] 17565 wl_proxy_marshal((struct wl_proxy *) wl_surface, 17566 ^ 17567 xwayland.c:466:23: note: ‘buffer’ was declared here 17568 struct wl_buffer *buffer; 17569 ^ 17570 17571 Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 17572 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 17573 17574commit f7b8560f23ac5582e2f97dc9f6de32a42e61e520 17575Author: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 17576Date: Thu Nov 24 11:45:25 2016 +0200 17577 17578 xwayland: refactor into xwl_window_post_damage() 17579 17580 Refactor xwl_screen_post_damage() and split the window specific code 17581 into a new function xwl_window_post_damage(). 17582 17583 This is a pure refactoring, there are no behavioral changes. An assert 17584 is added to xwl_window_post_damage() to ensure frame callbacks are not 17585 leaked if a future patch changes the call. 17586 17587 Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 17588 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 17589 17590commit 8e3f9ce6c06e7605832c55dfd180396f66ec8b66 17591Author: Adam Jackson <ajax@redhat.com> 17592Date: Sat Jul 12 12:45:23 2014 -0400 17593 17594 dix: Add a callback chain for window property state change 17595 17596 This will be used by in-server features that need to react to property 17597 changes. The first one will be _XWAYLAND_ALLOW_COMMITS. 17598 17599 Signed-off-by: Adam Jackson <ajax@redhat.com> 17600 [Pekka: add commit message body] 17601 Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 17602 17603commit 50bcea8be337ea983e464f2b5b8b2dc6d1024532 17604Author: Adam Jackson <ajax@redhat.com> 17605Date: Sat Jul 12 12:39:21 2014 -0400 17606 17607 dix: Pass the whole property into deliverPropertyNotifyEvent 17608 17609 Instead of just the atom. No functional change. 17610 17611 Signed-off-by: Adam Jackson <ajax@redhat.com> 17612 Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 17613 17614commit fde5cd78488f3370c92397868ca78929296f7ad1 17615Author: Mihail Konev <k.mvc@ya.ru> 17616Date: Tue Jan 31 20:16:35 2017 +0500 17617 17618 tests: Fix guards for ld -wrap 17619 17620 Fixes XQuartz build. 17621 17622 Regressed-in: 23f6dbc96e47be6cbeed78cc9ef303987c3e29a9 17623 Reviewed-by: Adam Jackson <ajax@redhat.com> 17624 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 17625 17626commit d8161aeb50891ae10c5656487ce8f982deed5f9f 17627Author: Dr.-Ing. Dieter Jurzitza <dieter.jurzitza@t-online.de> 17628Date: Thu Feb 23 12:57:26 2017 -0500 17629 17630 glamor: Fix missing declaration in dash vertex shader 17631 17632 Fixes a GLSL compilation error: 17633 17634 Failed to compile VS: 0:13(43): error: `pos' undeclared 17635 0:13(14): error: operands to arithmetic operators must be numeric 17636 0:13(13): error: operands to arithmetic operators must be numeric 17637 17638 Tested-by: Stefan Dirsch <sndirsch@suse.com> 17639 Reviewed-by: Adam Jackson <ajax@redhat.com> 17640 17641commit acdb5bf2de57c0080d2a6e730c788a0a428e13dc 17642Author: Chris Wilson <chris@chris-wilson.co.uk> 17643Date: Fri Feb 17 08:18:52 2017 +0000 17644 17645 os: Fix iteration over busfaults 17646 17647 Fixes a regression from 17648 17649 commit 41da295eb50fa08eaacd0ecde99f43a716fcb41a 17650 Author: Keith Packard <keithp@keithp.com> 17651 Date: Sun Nov 3 13:12:40 2013 -0800 17652 17653 Trap SIGBUS to handle truncated shared memory segments 17654 17655 that causes the SIGBUS handler to fail to chain up correctly and 17656 corrupts nearby memory instead. 17657 17658 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 17659 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 17660 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 17661 17662commit dcb63deb1b6337781943a7bea5bb2b04c6d76b55 17663Author: Chris Wilson <chris@chris-wilson.co.uk> 17664Date: Wed Feb 22 14:56:28 2017 +0000 17665 17666 inputthread: Initialise inputThreadInfo->changed before use 17667 17668 ==8734== Thread 2 InputThread: 17669 ==8734== Conditional jump or move depends on uninitialised value(s) 17670 ==8734== at 0x2FDB05: InputThreadDoWork (inputthread.c:333) 17671 ==8734== by 0x6924423: start_thread (pthread_create.c:333) 17672 ==8734== by 0x6C229BE: clone (clone.S:105) 17673 17674 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 17675 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 17676 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 17677 17678commit 23d85c54421ebdd12460566918fcb7f309d2ee30 17679Author: Olivier Fourdan <ofourdan@redhat.com> 17680Date: Fri Feb 17 14:13:51 2017 +0100 17681 17682 os: remove unused define MAX_TIMES_PER 17683 17684 Remove leftover from commit e10ba9e, MAX_TIMES_PER is not used anymore. 17685 17686 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17687 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 17688 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 17689 17690commit 8b170c4447868c8d4656cf2143f009db20885f3a 17691Author: Adam Jackson <ajax@redhat.com> 17692Date: Thu Feb 16 13:00:24 2017 -0500 17693 17694 mi: Add include guards to migc.h 17695 17696 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17697 Signed-off-by: Adam Jackson <ajax@redhat.com> 17698 17699commit e161853f380a0b176c81284024066cb2776136c3 17700Author: Adam Jackson <ajax@redhat.com> 17701Date: Thu Feb 16 13:00:24 2017 -0500 17702 17703 fb: Remove stray references to fbHasVisualTypes 17704 17705 Removed in 84e8de12. 17706 17707 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17708 Signed-off-by: Adam Jackson <ajax@redhat.com> 17709 17710commit 4c7d229582a97e07014091909745ac7dc4d5e93e 17711Author: Adam Jackson <ajax@redhat.com> 17712Date: Thu Feb 16 13:00:24 2017 -0500 17713 17714 fb: Remove unused fbGetFreeCompClip macro 17715 17716 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17717 Signed-off-by: Adam Jackson <ajax@redhat.com> 17718 17719commit 706e6d9cd074da606016ed4ecff51e9c2a822087 17720Author: Adam Jackson <ajax@redhat.com> 17721Date: Thu Feb 16 13:00:24 2017 -0500 17722 17723 wfb: Hide init function decls behind FB_ACCESS_WRAPPER 17724 17725 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17726 Signed-off-by: Adam Jackson <ajax@redhat.com> 17727 17728commit 790c1c0b10968776440a962e1fc6d1d3157b1559 17729Author: Adam Jackson <ajax@redhat.com> 17730Date: Thu Feb 16 13:00:24 2017 -0500 17731 17732 wfb: Add missing rename for fbGlyphs 17733 17734 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17735 Signed-off-by: Adam Jackson <ajax@redhat.com> 17736 17737commit a6566f9e4dbf9ea9568a14e22cb5d004e10dbd4d 17738Author: Michel Dänzer <michel.daenzer@amd.com> 17739Date: Thu Feb 16 16:13:56 2017 +0900 17740 17741 prime: Clear PixmapDirtyUpdateRec::damage when it's destroyed 17742 17743 The root window, and by extension any damage records referencing it, 17744 may be destroyed before shared pixmaps referencing it, which resulted in 17745 use-after-free / double-free in PixmapStopDirtyTracking. 17746 17747 Fixes: b5b292896f64 ("prime: Sync shared pixmap from root window instead of screen pixmap") 17748 Reviewed-by: Adam Jackson <ajax@redhat.com> 17749 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 17750 17751commit 371ff0c969a38a0013688391bbd7375bc7b6f933 17752Author: Olivier Fourdan <ofourdan@redhat.com> 17753Date: Thu Feb 16 10:43:06 2017 +0100 17754 17755 xwayland: bump wayland-protocols version to 1.7 17756 17757 Xwayland support for pointer locking in confinement requires 17758 wayland-protocols version 1.7 or later. 17759 17760 Update the required version in configure.ac to match the minimal 17761 required version of wayland-protocols. 17762 17763 Reviewed-by: Adam Jackson <ajax@redhat.com> 17764 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17765 17766commit 5d29b493666f728eccd8dd91992ea020bd4be9f9 17767Author: Adam Jackson <ajax@redhat.com> 17768Date: Tue Feb 14 15:30:56 2017 -0500 17769 17770 xfree86: Reorganize the reserved ScrnInfoRec slots 17771 17772 First, move them to the end of the struct, for marginally better cache 17773 locality for the struct members that actually have meaning; move the 17774 existing slots at the end of the struct up near some others with similar 17775 meanings. Second, only keep four slots each of integer, data pointer, 17776 and function pointer; we've rarely used this escape hatch so this is 17777 still plenty. 17778 17779 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17780 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 17781 Acked-by: Keith Packard <keithp@keithp.com> 17782 Signed-off-by: Adam Jackson <ajax@redhat.com> 17783 17784commit 47db92473f7da2b7d6775ff681ec747009f8cf3f 17785Author: Adam Jackson <ajax@redhat.com> 17786Date: Tue Feb 14 15:30:55 2017 -0500 17787 17788 xfree86: Remove unused chipID/Rev from ScrnInfoRec 17789 17790 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17791 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 17792 Acked-by: Keith Packard <keithp@keithp.com> 17793 Signed-off-by: Adam Jackson <ajax@redhat.com> 17794 17795commit 4dcf68e41857cdf78387a6232160cba5cb467fb0 17796Author: Adam Jackson <ajax@redhat.com> 17797Date: Tue Feb 14 15:30:54 2017 -0500 17798 17799 xfree86: Remove memClk from ScrnInfoRec 17800 17801 Never set by the core, not used in any modern driver. 17802 17803 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17804 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 17805 Acked-by: Keith Packard <keithp@keithp.com> 17806 Signed-off-by: Adam Jackson <ajax@redhat.com> 17807 17808commit 0b6f2806029493afa246b4472ee88fc173976c5a 17809Author: Adam Jackson <ajax@redhat.com> 17810Date: Tue Feb 14 15:30:53 2017 -0500 17811 17812 xfree86: Remove Option "BiosBase" (v2) 17813 17814 Just no. 17815 17816 The ddxDesign chunk removes the whole para about xf86FixPciResource, 17817 since it turns out that function doesn't exist at all anymore. 17818 17819 The only drivers that reference this at all are i128 and mga, and even 17820 then only in the non-pciaccess path. 17821 17822 v2: 17823 - Update commentary about i128/mga 17824 - Don't remove the BiosBase keyword from the config parser since that 17825 would turn a no-op into a fatal error (Aaron Plattner) 17826 17827 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17828 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 17829 Acked-by: Keith Packard <keithp@keithp.com> 17830 Signed-off-by: Adam Jackson <ajax@redhat.com> 17831 17832commit 76ef102be3b47d23ebfaaaa04fb5cfc0511d43f1 17833Author: Adam Jackson <ajax@redhat.com> 17834Date: Tue Feb 14 15:30:52 2017 -0500 17835 17836 xfree86: Drop virtualFrom from ScrnInfoRec 17837 17838 Seriously not worth the effort of tracking this, especially now that 17839 competent drivers don't have a limit. The sis driver does inspect this 17840 member, but hilariously does so only so it can print the same information 17841 as the core does. 17842 17843 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17844 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 17845 Acked-by: Keith Packard <keithp@keithp.com> 17846 Signed-off-by: Adam Jackson <ajax@redhat.com> 17847 17848commit 5a945f54e738f6e9c5a489febdd8b87cd8a41210 17849Author: Adam Jackson <ajax@redhat.com> 17850Date: Tue Feb 14 15:30:51 2017 -0500 17851 17852 xfree86: Remove max[HV]Value from ScrnInfoRec 17853 17854 Only mach64 and rendition actually use this feature. Everyone else just 17855 checks it in their ValidMode hook, they can too. 17856 17857 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17858 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 17859 Acked-by: Keith Packard <keithp@keithp.com> 17860 Signed-off-by: Adam Jackson <ajax@redhat.com> 17861 17862commit 3e1a23281dd64ee7457f36208cac77c01846a28f 17863Author: Adam Jackson <ajax@redhat.com> 17864Date: Tue Feb 14 15:30:50 2017 -0500 17865 17866 xfree86: Fix a comment about ScrnInfoRec 17867 17868 We don't actually need (or intend) to keep this struct the same across 17869 revisions. 17870 17871 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17872 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 17873 Acked-by: Keith Packard <keithp@keithp.com> 17874 Signed-off-by: Adam Jackson <ajax@redhat.com> 17875 17876commit 4c00609c370dab130b69207cb2684c2b92bd3084 17877Author: Alan Coopersmith <alan.coopersmith@oracle.com> 17878Date: Sun Feb 12 09:35:21 2017 -0800 17879 17880 DetermineClientCmd: try using /proc/pid/cmdline on Solaris too 17881 17882 Solaris 11.3.5 introduced support for /proc/pid/cmdline, so try it 17883 first, and if we can't open it, then fallback to /proc/pid/psinfo 17884 as we did before. 17885 17886 Reviewed-by: Adam Jackson <ajax@redhat.com> 17887 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 17888 17889commit 3f9507ed2f7246b2c8cf2bbc430cc99c5f35c92a 17890Author: Adam Jackson <ajax@redhat.com> 17891Date: Thu Feb 9 13:53:41 2017 -0500 17892 17893 xfree86: Bump video ABI to 24.0 17894 17895 c02f6a68 changed the layout of xf86CrtcFuncs. 17896 17897 Signed-off-by: Adam Jackson <ajax@redhat.com> 17898 17899commit 058809c43ec578a407cf40d4c3e54a42503e3562 17900Author: Olivier Fourdan <ofourdan@redhat.com> 17901Date: Wed Feb 8 09:23:20 2017 +0100 17902 17903 xwayland: Apply output rotation for screen size 17904 17905 Previously, we would swap the width/height of the Xwayland output based 17906 on the output rotation, so that the overall screen size would match the 17907 actual rotation of each output. 17908 17909 Problem is the RandR's ConstrainCursorHarder() handler will also apply 17910 the output rotation, meaning that when the output is rotated, the 17911 pointer will be constrained within the wrong dimension. 17912 17913 Moreover, XRandR assumes the original output width/height are unchanged 17914 when the output is rotated, so by changing the Xwayland output width and 17915 height based on rotation, Xwayland causes XRandr to report the wrong 17916 output sizes (an output of size 1024x768 rotated left or right should 17917 remain 1024x768, not 768x1024). 17918 17919 So to avoid this issue and keep things consistent between Wayland and 17920 Xwayland outputs, leave the actual width/height unchanged but apply the 17921 rotation when computing the screen size. This fixes both the output size 17922 being wrong in "xrandr -q" and the pointer being constrained in the 17923 wrong dimension with rotated with weston. 17924 17925 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99663 17926 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17927 Reviewed-by: Adam Jackson <ajax@redhat.com> 17928 17929commit afeace27d3818274b75d59375771dc964d2f56bb 17930Author: Olivier Fourdan <ofourdan@redhat.com> 17931Date: Tue Feb 7 15:31:22 2017 +0100 17932 17933 xwayland: CRTC should support all rotations 17934 17935 If the Wayland compositor sets a rotation on the output, Xwayland 17936 translates the transformation as an xrandr rotation for the given 17937 output. 17938 17939 However, if the rotation is not supported by the CRTC, this is not 17940 a valid setup and xrandr queries will fail. 17941 17942 Pretend we support all rotations and reflections so that the 17943 configuration remains a valid xrandr setup. 17944 17945 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99663 17946 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 17947 Reviewed-by: Adam Jackson <ajax@redhat.com> 17948 17949commit 38696ea56854e055c31bd2730adfc7c39aa115b0 17950Author: Michel Dänzer <michel.daenzer@amd.com> 17951Date: Tue Feb 7 17:38:23 2017 +0900 17952 17953 damage: Validate source pictures bound to windows before unwrapping 17954 17955 The lower layers also do this, but no damage may be reported there, 17956 since we unwrap before calling down. 17957 17958 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99220 17959 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 17960 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17961 17962commit 1c78bec9ca3cd1975a38bf5ebdba7dea65b309ab 17963Author: Svitozar Cherepii <razotivs@gmail.com> 17964Date: Fri Feb 3 01:49:04 2017 +0200 17965 17966 xwayland: Add hack for FWXGA resolution #99574 17967 17968 For some applications (like fullscreen games) it matters for XRandr 17969 resolution to be correctly set and equal to root window resolution. 17970 17971 In XServer there is already hack for this, adapted it for XWayland. 17972 17973 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=99574 17974 17975 Signed-off-by: Svitozar Cherepii <razotivs@gmail.com> 17976 Tested-by: Svitozar Cherepii <razotivs@gmail.com> 17977 Acked-by: Olivier Fourdan <ofourdan@redhat.com> 17978 17979commit 542d9f6807ac06b70f564ccab10af69fa21a1221 17980Author: Michel Dänzer <michel.daenzer@amd.com> 17981Date: Wed Feb 1 18:35:57 2017 +0900 17982 17983 present: Allow flipping with PRIME slave outputs 17984 17985 Works fine now. 17986 17987 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 17988 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 17989 17990commit b5b292896f647c85f03f53b20b2f03c0e94de428 17991Author: Michel Dänzer <michel.daenzer@amd.com> 17992Date: Wed Feb 1 18:35:56 2017 +0900 17993 17994 prime: Sync shared pixmap from root window instead of screen pixmap 17995 17996 The screen pixmap doesn't receive updates while there's a Present flip 17997 window. 17998 17999 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 18000 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 18001 18002commit eb04b20160d706d4e0e67122d0adb1e723c0da92 18003Author: Michael Thayer <michael.thayer@oracle.com> 18004Date: Fri Sep 30 15:18:45 2016 +0200 18005 18006 modesetting: allow switching from software to hardware cursors (v5). 18007 18008 Currently if modesetting ever fails to set a hardware cursor it will switch 18009 to using a software cursor and never go back. Change this to only 18010 permanently switch to a software cursor if -ENXIO is returned (which means 18011 hardware cursors not supported), and to otherwise still try a hardware 18012 cursor first every time a new one is set. This is needed because hardware 18013 may be able to handle some cursors in hardware and others not, or virtual 18014 hardware may be able to handle hardware cursors at some times and not 18015 others. 18016 18017 Changes since v1, v2 and v3: 18018 * take into account the switch to load_cursor_argb_check 18019 * keep the permanent software cursor fall-back if -ENXIO is returned 18020 * move parts of v3 into separate patches 18021 18022 Reviewed-by: Adam Jackson <ajax@redhat.com> 18023 Signed-off-by: Michael Thayer <michael.thayer@oracle.com> 18024 18025commit ecd0a62323f26b333c49bddd7237dd5118482a35 18026Author: Michael Thayer <michael.thayer@oracle.com> 18027Date: Fri Sep 30 08:02:09 2016 +0200 18028 18029 modesetting: Immediately handle failure to set HW cursor, v5 18030 18031 Based on v4 by Alexandre Courbot <acourbot@nvidia.com> 18032 18033 There is currently no reliable way to report failure to set a HW 18034 cursor. Still such failures can happen if e.g. the MODE_CURSOR DRM 18035 ioctl fails (which currently happens at least with modesetting on Tegra 18036 for format incompatibility reasons). 18037 18038 As failures are currently handled by setting the HW cursor size to 18039 (0,0), the fallback to SW cursor will not happen until the next time the 18040 cursor changes and xf86CursorSetCursor() is called again. In the 18041 meantime, the cursor will be invisible to the user. 18042 18043 This patch addresses that by adding _xf86CrtcFuncs::set_cursor_check and 18044 _xf86CursorInfoRec::ShowCursorCheck hook variants that return booleans. 18045 This allows to propagate errors up to xf86CursorSetCursor(), which can 18046 then fall back to using the SW cursor immediately. 18047 18048 v5: 18049 - Removed parts of patch already committed as part of 14c21ea1. 18050 - Adjusted code slightly to match surrounding code. 18051 - Effectively reverted af916477 which is made unnecessary by this patch. 18052 18053 Reviewed-by: Adam Jackson <ajax@redhat.com> 18054 Signed-off-by: Michael Thayer <michael.thayer@oracle.com> 18055 18056commit c02f6a687c3d6bd0727322b055ee788f8fefa005 18057Author: Michael Thayer <michael.thayer@oracle.com> 18058Date: Thu Sep 29 20:23:12 2016 +0200 18059 18060 xfree86: Immediately handle failure to set HW cursor, v5 18061 18062 Based on v4 by Alexandre Courbot <acourbot@nvidia.com> 18063 18064 There is currently no reliable way to report failure to set a HW 18065 cursor. Still such failures can happen if e.g. the MODE_CURSOR DRM 18066 ioctl fails (which currently happens at least with modesetting on Tegra 18067 for format incompatibility reasons). 18068 18069 As failures are currently handled by setting the HW cursor size to 18070 (0,0), the fallback to SW cursor will not happen until the next time the 18071 cursor changes and xf86CursorSetCursor() is called again. In the 18072 meantime, the cursor will be invisible to the user. 18073 18074 This patch addresses that by adding _xf86CrtcFuncs::set_cursor_check and 18075 _xf86CursorInfoRec::ShowCursorCheck hook variants that return booleans. 18076 This allows to propagate errors up to xf86CursorSetCursor(), which can 18077 then fall back to using the SW cursor immediately. 18078 18079 v5: Updated the patch to apply to current git HEAD, split up into two 18080 patches (server and modesetting driver) and adjusted the code slightly 18081 to match surrounding code. I also removed the new exported function 18082 ShowCursorCheck(), as instead just changing ShowCursor() to return Bool 18083 should not affect its current callers. 18084 18085 Reviewed-by: Adam Jackson <ajax@redhat.com> 18086 Signed-off-by: Michael Thayer <michael.thayer@oracle.com> 18087 18088commit 3ef16dfb9830bd6b41ae428f4f213ae0c35c1056 18089Author: Mihail Konev <k.mvc@ya.ru> 18090Date: Sat Feb 4 00:03:24 2017 +0500 18091 18092 dmx: fix linking 18093 18094 Reviewed-by: Adam Jackson <ajax@redhat.com> 18095 18096commit e50da50118408a195d4d2e1b39817fe7c4447c56 18097Author: Adam Jackson <ajax@redhat.com> 18098Date: Wed Jan 25 15:08:25 2017 -0500 18099 18100 dri1: Remove some dead event code 18101 18102 Signed-off-by: Adam Jackson <ajax@redhat.com> 18103 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 18104 18105commit 0702f2e840370be53d129fc5e3c243dac47ee720 18106Author: Peter Hutterer <peter.hutterer@who-t.net> 18107Date: Tue Feb 7 11:44:51 2017 +1000 18108 18109 xwayland: replace hardcoded function name with __func__ in error msg 18110 18111 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 18112 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 18113 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 18114 18115commit 3eb964e25243056dd998f52d3b00171b71c89189 18116Author: Chris Wilson <chris@chris-wilson.co.uk> 18117Date: Fri Jan 20 09:49:19 2017 +0000 18118 18119 xfree86: Take input_lock() for xf86ScreenCheckHWCursor 18120 18121commit cfddd919cce4178baba07959e5e862d02e166522 18122Author: Chris Wilson <chris@chris-wilson.co.uk> 18123Date: Mon Jan 16 22:36:34 2017 +0000 18124 18125 xfree86: Take input lock for xf86TransparentCursor 18126 18127commit 7198a6d4e74f684cb383b3e0f70dd2bae405e6e7 18128Author: Chris Wilson <chris@chris-wilson.co.uk> 18129Date: Mon Jan 16 22:17:36 2017 +0000 18130 18131 xfree86: Take the input lock for xf86RecolorCursor 18132 18133 xf86RecolorCursor() may be called directly from XRecolorCursor as well 18134 as from xf86ScreenSetCursor(). In the latter case, the input lock is 18135 already held, but not for the former and so we need to add a wrapper 18136 function that acquires the input lock before performing 18137 xf86RecolorCursor() 18138 18139 References: https://bugs.freedesktop.org/show_bug.cgi?id=99358 18140 18141commit bb9128fdc86decd6f6e3b0e145011a8c08b1d2b5 18142Author: Qiang Yu <Qiang.Yu@amd.com> 18143Date: Thu Jan 26 18:13:53 2017 +0800 18144 18145 present: disable page flip only when a slave crtc is active 18146 18147 This prevents the tearing of moving window in a composite WM 18148 desktop when output slave is attached but none of its crtc is 18149 really active. 18150 18151 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 18152 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 18153 18154commit 45e0eb4b156f2155687cce268b07f10540fc507b 18155Author: Michel Dänzer <michel.daenzer@amd.com> 18156Date: Thu Jan 26 16:24:03 2017 +0900 18157 18158 loader: Handle mod->VersionInfo == NULL 18159 18160 This can happen when a module fails to load: 18161 18162 Program received signal SIGSEGV, Segmentation fault. 18163 UnloadModule (_mod=0x5555559d9280) at ../../../../hw/xfree86/loader/loadmod.c:848 18164 848 name = mod->VersionInfo->modname; 18165 (gdb) bt 18166 #0 UnloadModule (_mod=0x5555559d9280) at ../../../../hw/xfree86/loader/loadmod.c:848 18167 #1 0x00005555555ddd1b in LoadModule (module=module@entry=0x5555559c7ce0 "fbdev", options=0x0, modreq=modreq@entry=0x0, errmaj=errmaj@entry=0x7fffffffe8ec) at ../../../../hw/xfree86/loader/loadmod.c:824 18168 #2 0x00005555555edfe9 in xf86LoadModules (list=list@entry=0x5555559dcf50, optlist=optlist@entry=0x0) at ../../../../hw/xfree86/common/xf86Init.c:1506 18169 #3 0x00005555555ee7bc in InitOutput (pScreenInfo=pScreenInfo@entry=0x5555559abf80 <screenInfo>, argc=argc@entry=4, argv=argv@entry=0x7fffffffeb18) at ../../../../hw/xfree86/common/xf86Init.c:484 18170 #4 0x00005555555a885c in dix_main (argc=4, argv=0x7fffffffeb18, envp=<optimized out>) at ../../dix/main.c:197 18171 #5 0x00007ffff5d582b1 in __libc_start_main (main=0x555555593130 <main>, argc=4, argv=0x7fffffffeb18, init=<optimized out>, fini=<optimized out>, rtld_fini=<optimized out>, stack_end=0x7fffffffeb08) at ../csu/libc-start.c:291 18172 #6 0x000055555559316a in _start () 18173 18174 Fixes: 8e83eacb9e2d ("loader: Remove unused path and name from ModuleDescPtr") 18175 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 18176 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 18177 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 18178 18179commit 7617a0a180a2cd3427a8ffa9534152df6a8fecbf 18180Author: Qiang Yu <Qiang.Yu@amd.com> 18181Date: Wed Jan 11 13:59:22 2017 +0800 18182 18183 dri2: refine dri2_probe_driver_name (v2) 18184 18185 V2: 18186 1. update comment 18187 2. check bustype if PCI 18188 3. configure add libdrm version check for drmGetDevice 18189 18190 Get PCI information from info->fd with drmGetDevice instead of 18191 assuming the info->fd is the first entity of scrn which is not 18192 true for multi entities scrn. 18193 18194 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 18195 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 18196 18197commit b1a6986395f85b0194646a34b49367086dc9239d 18198Author: Martin Peres <martin.peres@linux.intel.com> 18199Date: Wed Jan 25 17:29:15 2017 +0200 18200 18201 tests: shuffle around the linking order to please GNU ld 18202 18203 Classic GNU ld resolves symbol dependencies only forward, while GOLD 18204 seems to work regardless of the specified library order. 18205 18206 Suggested-by: Eero Tamminen <eero.t.tamminen@intel.com> 18207 Reviewed-by: Mihail Konev <k.mvc@ya.ru> 18208 Signed-off-by: Martin Peres <martin.peres@linux.intel.com> 18209 18210commit cc0f173ea2936d1405e382329c1bd58c7af67ea7 18211Author: Adam Jackson <ajax@redhat.com> 18212Date: Thu Apr 14 15:56:46 2016 -0400 18213 18214 loader: Learn about the joy of snprintf 18215 18216 Reviewed-by: Julien Cristau <jcristau@debian.org> 18217 Signed-off-by: Adam Jackson <ajax@redhat.com> 18218 18219commit 8e83eacb9e2d2c6c2b9f8cdb9e82c976a0237f24 18220Author: Adam Jackson <ajax@redhat.com> 18221Date: Thu Apr 14 15:30:35 2016 -0400 18222 18223 loader: Remove unused path and name from ModuleDescPtr 18224 18225 Just a waste of memory. Path was never referenced at all, and name was 18226 only used when unloading the module; we can just as well get the 18227 module's internal idea of its name from VersionInfo. 18228 18229 Reviewed-by: Julien Cristau <jcristau@debian.org> 18230 Signed-off-by: Adam Jackson <ajax@redhat.com> 18231 18232commit 8920dca0091675f1202c1198336cd4d8e0259100 18233Author: Adam Jackson <ajax@redhat.com> 18234Date: Thu Apr 14 11:45:58 2016 -0400 18235 18236 loader: Remove unused arguments from LoadModule 18237 18238 Nobody was ever calling this with a non-null argument for subdir list or 18239 pattern list. Having done this, InitSubdirs is only ever called with a 18240 NULL argument, so it's really just a complicated way of duplicating the 18241 default list; we can remove that and just walk the list directly. 18242 18243 The minor error code was only ever used to distinguish among two cases 18244 of LDR_BADUSAGE. Whatever. 18245 18246 Reviewed-by: Julien Cristau <jcristau@debian.org> 18247 Signed-off-by: Adam Jackson <ajax@redhat.com> 18248 18249commit ba726ba6a73efe1bd19708b058f481f28068a85b 18250Author: Adam Jackson <ajax@redhat.com> 18251Date: Thu Apr 14 11:30:02 2016 -0400 18252 18253 loader: Turn LoaderListDirs into LoaderListDir 18254 18255 Callers only ever use this for a single directory anyway. 18256 18257 While we're at it, also move xf86DriverListFromCompile near its only 18258 user in the X -configure code (and inline it out of existence), and 18259 remove LoaderFreeDirList as it's unused (since X -configure is just 18260 going to exit anyway, none of that code cares about cleanup). 18261 18262 Reviewed-by: Julien Cristau <jcristau@debian.org> 18263 Signed-off-by: Adam Jackson <ajax@redhat.com> 18264 18265commit c54a9ca152898ec2ffe50f6d5b70d483b85c1c34 18266Author: Adam Jackson <ajax@redhat.com> 18267Date: Thu Apr 14 11:17:22 2016 -0400 18268 18269 loader: Move loader list details to internal header 18270 18271 There's no reason a driver should ever care about this. 18272 18273 Reviewed-by: Eric Anholt <eric@anholt.net> 18274 Reviewed-by: Julien Cristau <jcristau@debian.org> 18275 Signed-off-by: Adam Jackson <ajax@redhat.com> 18276 18277commit d55284e8638ede15be851aa8a19a98dbbff0ce20 18278Author: Adam Jackson <ajax@redhat.com> 18279Date: Wed Apr 13 16:06:50 2016 -0400 18280 18281 xfree86: Remove a stray reference to font modules 18282 18283 Reviewed-by: Eric Anholt <eric@anholt.net> 18284 Reviewed-by: Julien Cristau <jcristau@debian.org> 18285 Signed-off-by: Adam Jackson <ajax@redhat.com> 18286 18287commit d7879c46724a36bfd9a3a08e49903d001ffbf93e 18288Author: Adam Jackson <ajax@redhat.com> 18289Date: Wed Apr 13 15:58:33 2016 -0400 18290 18291 xfree86: Fix up some bad indentation 18292 18293 indent(1) gets confused by function-like macros with no trailing 18294 semicolon, which is fair enough really. 18295 18296 Reviewed-by: Eric Anholt <eric@anholt.net> 18297 Reviewed-by: Julien Cristau <jcristau@debian.org> 18298 Signed-off-by: Adam Jackson <ajax@redhat.com> 18299 18300commit 5c577da5f3a65c68d2ee12e4afca8f20c3e8ccf4 18301Author: Adam Jackson <ajax@redhat.com> 18302Date: Thu Jul 14 15:42:18 2016 -0400 18303 18304 xfree86: Remove DriverRec1 compat struct 18305 18306 The idea here is that the driver might have once been old enough to not 18307 have the driverFunc slot in DriverRec, with the module ABI not having 18308 changed when it was added. That was ages ago, and drivers always declare 18309 themselves with DriverRec not DriverRec1, so uninitialized slots will 18310 simply be zero. 18311 18312 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 18313 Reviewed-by: Eric Anholt <eric@anholt.net> 18314 Reviewed-by: Julien Cristau <jcristau@debian.org> 18315 Signed-off-by: Adam Jackson <ajax@redhat.com> 18316 18317commit 2e3ad7e2506d9eb6667a5f229b5213d215451a5a 18318Author: Adam Jackson <ajax@redhat.com> 18319Date: Wed Apr 13 15:50:39 2016 -0400 18320 18321 loader: Remove silly "unspecified" version handling 18322 18323 Everybody using this functionality specifies a major version, which 18324 makes sense. If you don't care about a minor version, that's equivalent 18325 to saying you require minor >= 0, so just say so; likewise patch level. 18326 18327 Likewise ABI class is always specified. 18328 18329 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 18330 Reviewed-by: Eric Anholt <eric@anholt.net> 18331 Reviewed-by: Julien Cristau <jcristau@debian.org> 18332 Signed-off-by: Adam Jackson <ajax@redhat.com> 18333 18334commit ef533a912d18db31456b29a18c8bced649309565 18335Author: Adam Jackson <ajax@redhat.com> 18336Date: Tue Mar 1 08:29:06 2016 -0500 18337 18338 loader: Remove unused loader error codes and dead enum 18339 18340 The enum has been unused since at least the removal of elfloader. 18341 18342 Reviewed-by: Eric Anholt <eric@anholt.net> 18343 Reviewed-by: Julien Cristau <jcristau@debian.org> 18344 Signed-off-by: Adam Jackson <ajax@redhat.com> 18345 18346commit 7e3cccf8e4426a4b25a7a94d52775d334693572f 18347Author: Adam Jackson <ajax@redhat.com> 18348Date: Wed Apr 13 14:59:27 2016 -0400 18349 18350 loader: Include fewer headers from xf86Module.h 18351 18352 This looks like more, but only if you don't compare it to the number 18353 pulled in by misc.h. 18354 18355 Reviewed-by: Eric Anholt <eric@anholt.net> 18356 Reviewed-by: Julien Cristau <jcristau@debian.org> 18357 Signed-off-by: Adam Jackson <ajax@redhat.com> 18358 18359commit 49fa76801348f6d044128f7ec743693d0759d683 18360Author: Adam Jackson <ajax@redhat.com> 18361Date: Fri Jun 5 11:01:30 2015 -0400 18362 18363 loader: Don't add internal/ to the search path 18364 18365 Reviewed-by: Eric Anholt <eric@anholt.net> 18366 Reviewed-by: Julien Cristau <jcristau@debian.org> 18367 Signed-off-by: Adam Jackson <ajax@redhat.com> 18368 18369commit 97bd6e453676516891250389ec0fd695c110087c 18370Author: Adam Jackson <ajax@redhat.com> 18371Date: Fri Jun 5 10:58:20 2015 -0400 18372 18373 loader: Remove *GetOS 18374 18375 This API is dumb. uname(3) exists, feel free to use it, but ideally 18376 write to the interface not to the OS. There are a couple of drivers 18377 using this API, they could all reasonably just not. 18378 18379 This also removes the OS name from the loader subdirectory path search. 18380 Having /usr/lib/xorg shared across OSes is a non-goal here. 18381 18382 Reviewed-by: Eric Anholt <eric@anholt.net> 18383 Reviewed-by: Julien Cristau <jcristau@debian.org> 18384 Signed-off-by: Adam Jackson <ajax@redhat.com> 18385 18386commit a6fcb15472bb7663ae917f5913bf07b6d3c7f186 18387Author: Adam Jackson <ajax@redhat.com> 18388Date: Mon Jun 16 11:30:18 2014 -0400 18389 18390 loader: Port from xfree86 to dix API 18391 18392 Reviewed-by: Eric Anholt <eric@anholt.net> 18393 Reviewed-by: Julien Cristau <jcristau@debian.org> 18394 Signed-off-by: Adam Jackson <ajax@redhat.com> 18395 18396commit 778cfc59762cdf528cf8672bfb5696844e91ebc3 18397Author: Emil Velikov <emil.l.velikov@gmail.com> 18398Date: Sun Apr 17 21:07:30 2016 +0100 18399 18400 xfree86: flatten pathlist management in the loader 18401 18402 Now that users can set the path only via LoaderSetPath(), we can simplify 18403 things. 18404 18405 Reviewed-by: Eric Anholt <eric@anholt.net> 18406 Reviewed-by: Julien Cristau <jcristau@debian.org> 18407 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 18408 18409commit 7b71055fc65242a9c4b651e72bb07dbb3f00c4e6 18410Author: Emil Velikov <emil.l.velikov@gmail.com> 18411Date: Sun Apr 17 21:07:29 2016 +0100 18412 18413 xfree86: remove dummy/dead function prototype for LoadDriver 18414 18415 Reviewed-by: Julien Cristau <jcristau@debian.org> 18416 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 18417 18418commit 57eec704c3ddea0fecea2d8fbd1c3547f0b8fccf 18419Author: Emil Velikov <emil.l.velikov@gmail.com> 18420Date: Sun Apr 17 21:07:27 2016 +0100 18421 18422 xfree86: remove references to LoadSubModule's path from the doc 18423 18424 Afaics the argument hasn't been part of the API since the documentation 18425 has been converted to xml with commit fc6ebe1e1d3 "Convert LinuxDoc 18426 documents to DocBook/XML" 18427 18428 Reviewed-by: Julien Cristau <jcristau@debian.org> 18429 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 18430 18431commit 2196bb50383bd96d364f799018c0693c9309ed7f 18432Author: Emil Velikov <emil.l.velikov@gmail.com> 18433Date: Sun Apr 17 21:07:28 2016 +0100 18434 18435 xfree86: remove unused path from the LoadModule API 18436 18437 Similar to its little brother - LoadSubModule. Currently all call sites 18438 provide NULL anyway ;-) 18439 18440 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 18441 Reviewed-by: Julien Cristau <jcristau@debian.org> 18442 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 18443 18444commit ea483af99a6351323afe00a0b630cd63310efdb1 18445Author: Adam Jackson <ajax@redhat.com> 18446Date: Thu Jan 19 17:20:49 2017 -0500 18447 18448 glx: Move Composite's synthetic visuals to a different select group 18449 18450 Should you find yourself using a 16bpp display while also using a 18451 compositor, you poor soul, you may find that your GLX applications 18452 behave strangely; in particular, glxgears will be transparent. This is 18453 because it clears to (0,0,0,0) which is transparent if you honor alpha, 18454 and it will choose the synthetic visual because it has the most 18455 available r/g/b bits. 18456 18457 To avoid this, bump synthetic visuals to a higher (less-preferred) 18458 select group. Unless the client explicitly asks for non-zero alpha bits, 18459 this will prefer any rgb565 visual ahead of the argb8888 visual. 18460 18461 Reviewed-by: Eric Anholt <eric@anholt.net> 18462 Signed-off-by: Adam Jackson <ajax@redhat.com> 18463 18464commit 43efaa6e4fd903229dc9c508da4177ad4bbdd4d8 18465Author: Adam Jackson <ajax@redhat.com> 18466Date: Thu Jan 19 17:20:49 2017 -0500 18467 18468 glx: Send GLX_VISUAL_SELECT_GROUP_SGIX attribute for visuals 18469 18470 We already send this for fbconfigs. Mesa happens to implement 18471 glXChooseVisual relative to the fbconfig data, but that might not be 18472 true of NVIDIA's libGL. 18473 18474 Reviewed-by: Eric Anholt <eric@anholt.net> 18475 Signed-off-by: Adam Jackson <ajax@redhat.com> 18476 18477commit ef2345aaa28461a76f77c65240ce5facc180f98e 18478Author: Adam Jackson <ajax@redhat.com> 18479Date: Thu Jan 19 17:20:49 2017 -0500 18480 18481 composite: Export compIsAlternateVisual 18482 18483 Reviewed-by: Eric Anholt <eric@anholt.net> 18484 Signed-off-by: Adam Jackson <ajax@redhat.com> 18485 18486commit bccbaf7c113b1300071601879002682ebbe8b1c1 18487Author: Adam Jackson <ajax@redhat.com> 18488Date: Thu Jan 19 17:20:49 2017 -0500 18489 18490 composite: Remove a misleading comment 18491 18492 This comment is above compIsAlternateVisual, which used to be the only 18493 thing determining whether implicit redirect was needed. It's not anymore 18494 due to the redirection exception list. That job is now performed by 18495 compImplicitRedirect, whose code is self-explanitory. 18496 18497 Reviewed-by: Eric Anholt <eric@anholt.net> 18498 Signed-off-by: Adam Jackson <ajax@redhat.com> 18499 18500commit 86463981361064dd0352ec215abf1696ce7fc5ea 18501Author: Olivier Fourdan <ofourdan@redhat.com> 18502Date: Tue Jan 24 18:08:30 2017 +0100 18503 18504 glamor: Two pass won't work on memory pixmaps 18505 18506 When selecting "CA_TWO_PASS" in glamor_composite_clipped_region() when 18507 the hardware does not support "GL_ARB_blend_func_extended", we call 18508 glamor_composite_choose_shader() twice in a row, which in turn calls 18509 glamor_pixmap_ensure_fbo(). 18510 18511 On memory pixmaps, the first call will set the FBO and the second one 18512 will fail an assertion in glamor_upload_picture_to_texture() because 18513 the FBO is already set. 18514 18515 Bail out earlier when the mask pixmap is in memory and the hardware 18516 capabilities would require to use two pass, so that the assertion is not 18517 failed and the rendering is correct. 18518 18519 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99346 18520 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 18521 Reviewed-by: Eric Anholt <eric@anholt.net> 18522 18523commit e7b8b7b131d8283c96ed0aff4593ab41441b5d3b 18524Author: Adam Jackson <ajax@redhat.com> 18525Date: Thu Jan 19 17:30:40 2017 -0500 18526 18527 kdrive: Remove non-evdev input drivers 18528 18529 Use evdev. 18530 18531 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 18532 Signed-off-by: Adam Jackson <ajax@redhat.com> 18533 18534commit d740e1830a7ef32433fdf338c1e04dc6c9a927cb 18535Author: Adam Jackson <ajax@redhat.com> 18536Date: Thu Jan 19 17:30:39 2017 -0500 18537 18538 dmx: Remove non-evdev hardware input drivers 18539 18540 Use evdev. 18541 18542 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 18543 Signed-off-by: Adam Jackson <ajax@redhat.com> 18544 18545commit 6f531b0cdf8c9d7f711dd46071a26c50db3f1a16 18546Author: Mihail Konev <k.mvc@ya.ru> 18547Date: Sat Jan 14 12:51:45 2017 +0500 18548 18549 tests: Fix typo 18550 18551 Remove redundrant/incorrect automake assignment. 18552 18553 Buglink: https://lists.x.org/archives/xorg-devel/2017-January/052320.html 18554 Reported-by: Michel Dänzer <michel@daenzer.net> 18555 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18556 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 18557 18558commit 3f8c2f94483bf0b96e129c97ef4950170a3f05b4 18559Author: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 18560Date: Fri Jan 13 18:44:01 2017 +0100 18561 18562 tests: Fix without-dtrace build 18563 18564 Signed-off-by: Giuseppe Bilotta <giuseppe.bilotta@gmail.com> 18565 Reviewed-by: Mihail Konev <k.mvc@ya.ru> 18566 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18567 18568commit b44bef513bd83249036cb5a237dc20eb6088c17c 18569Author: Adam Jackson <ajax@redhat.com> 18570Date: Mon Jan 16 14:59:40 2017 -0500 18571 18572 Revert "tests: fix --disable-xorg build" 18573 18574 Pushed the wrong thing. 18575 18576 This reverts commit 730fd8c05f56da21894691bbd2e7ff37f67b45f4. 18577 18578commit 730fd8c05f56da21894691bbd2e7ff37f67b45f4 18579Author: Mihail Konev <k.mvc@ya.ru> 18580Date: Sat Jan 14 15:19:30 2017 +0500 18581 18582 tests: fix --disable-xorg build 18583 18584commit 704a867f8fb7652a8b7d5569bbe44e188457db4e 18585Author: Mihail Konev <k.mvc@ya.ru> 18586Date: Thu Jan 12 13:21:12 2017 +0500 18587 18588 tests: Remove obsolete libxservertest.la 18589 18590 Reviewed-by: Adam Jackson <ajax@redhat.com> 18591 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18592 18593commit 23f6dbc96e47be6cbeed78cc9ef303987c3e29a9 18594Author: Mihail Konev <k.mvc@ya.ru> 18595Date: Thu Jan 12 13:21:11 2017 +0500 18596 18597 tests: Move test/{xi1,xi2}/tests to test/ 18598 18599 Reviewed-by: Adam Jackson <ajax@redhat.com> 18600 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18601 18602commit ead5064581665ff40c177dd1b447949f1420e209 18603Author: Mihail Konev <k.mvc@ya.ru> 18604Date: Thu Jan 12 13:21:10 2017 +0500 18605 18606 tests: Convert test/ to single binary 18607 18608 Reviewed-by: Adam Jackson <ajax@redhat.com> 18609 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18610 18611commit 773fc07e4c2f816966809492a924a9bd1a15d059 18612Author: Mihail Konev <k.mvc@ya.ru> 18613Date: Thu Jan 12 13:21:09 2017 +0500 18614 18615 tests: Convert test/xi2/ to single binary 18616 18617 Reviewed-by: Adam Jackson <ajax@redhat.com> 18618 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18619 18620commit 371576f64baa99c4ab1e736fbae7975a32577275 18621Author: Mihail Konev <k.mvc@ya.ru> 18622Date: Thu Jan 12 13:21:08 2017 +0500 18623 18624 tests: Convert test/xi1/ to single binary 18625 18626 Reviewed-by: Adam Jackson <ajax@redhat.com> 18627 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18628 18629commit ff66bca3e8797db709e03572d296358dc4b95653 18630Author: Mihail Konev <k.mvc@ya.ru> 18631Date: Thu Jan 12 13:21:07 2017 +0500 18632 18633 tests: Refactor wraps into protocol-common.c 18634 18635 Part of refactoring the tests into a single binary, 18636 to make partial rebuild slightly faster and less verbose. 18637 18638 Prepares for joining test/xi2/protocol-* into a single binary. 18639 18640 Reviewed-by: Adam Jackson <ajax@redhat.com> 18641 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18642 18643commit 45546219e117c1b6129fd9f9be2ce5b592486feb 18644Author: Mihail Konev <k.mvc@ya.ru> 18645Date: Tue Jan 3 23:40:18 2017 +0500 18646 18647 tests: Avoid libtool message 18648 18649 The "copying selected object files" message appears as some source 18650 files have the same name, and some objects are included twice. 18651 18652 Reviewed-by: Adam Jackson <ajax@redhat.com> 18653 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18654 18655commit 785053d033e73d2deb0ded4b97eabfd881991978 18656Author: Keith Packard <keithp@keithp.com> 18657Date: Mon Jan 9 18:10:21 2017 -0800 18658 18659 AttendClient of grab-pervious client must queue to saved_ready_clients [v2] 18660 18661 A client which is attended while a grab is blocking execution of its 18662 requests needs to be placed in the saved_ready_clients list so that it 18663 will get scheduled once the grab terminates. Otherwise, if the client 18664 never sends another request, there is no way for it to be placed in 18665 the ready_clients list. 18666 18667 v2: Wrap comment above mark_client_saved_ready. 18668 Remove test for OS_COMM_IGNORED which will always be true. 18669 18670 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=99333 18671 Reviewed-by: Adam Jackson <ajax@redhat.com> 18672 Signed-off-by: Keith Packard <keithp@keithp.com> 18673 18674commit 436da935bde43076f3052a758a463ed080e8a747 18675Author: Qiang Yu <Qiang.Yu@amd.com> 18676Date: Tue Jan 10 18:51:57 2017 +0800 18677 18678 modesetting: cleanup pci device open 18679 18680 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 18681 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 18682 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 18683 18684commit 555e0a42d138ac8d83af62638752a1bebad602d6 18685Author: Qiang Yu <Qiang.Yu@amd.com> 18686Date: Tue Jan 10 18:51:55 2017 +0800 18687 18688 randr: fix xserver crash when xrandr setprovideroutputsource 18689 18690 xrandr --setprovideroutputsource <screen> <gpu screen> 18691 Xorg: ../../../xserver/dix/dispatch.c:4018: AttachOutputGPU: 18692 Assertion `new->isGPU' failed. 18693 18694 GPUScreen is not allowed to be sink output. 18695 18696 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 18697 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 18698 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 18699 18700commit 1012510620de7dadd0ab18b19a8e11facd884601 18701Author: Qiang Yu <Qiang.Yu@amd.com> 18702Date: Tue Jan 10 18:51:54 2017 +0800 18703 18704 xfree86: fix wrong usage of xf86optionListMerge 18705 18706 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 18707 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 18708 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 18709 18710commit e473b2bc016adacfe3fa47fdf6a8ce9f8cddff62 18711Author: Michel Dänzer <michel.daenzer@amd.com> 18712Date: Fri Nov 25 17:34:05 2016 +0900 18713 18714 present: Only call present_flip_notify if vblank->queued == FALSE 18715 18716 We are no longer using the present_flip_queue list only for presents 18717 which have already been submitted to the driver for page flipping, but 18718 also for those which we are queueing up to be flipped later, marked 18719 with vblank->queued == TRUE. We were incorrectly calling 18720 present_flip_notify for such entries, failing the assertion in 18721 present_flip_notify (or presumably resulting in other undesirable 18722 behaviour with assertions disabled). 18723 18724 Reproduction recipe: Run the JavaFX test case referenced by 18725 https://bugs.freedesktop.org/show_bug.cgi?id=98831#c6 and alt-tab out 18726 of it while it's fullscreen. May take a few attempts to hit the 18727 assertion failure. 18728 18729 Fixes: bab0f450a719 ("present: Fix presentation of flips out of order") 18730 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98854 18731 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 18732 18733commit 9d32b71c93cf6187e9320c99ae857e34a51b7102 18734Author: Mihail Konev <k.mvc@ya.ru> 18735Date: Wed Jan 4 07:08:51 2017 +0500 18736 18737 xkb: Match key releases with an overlaid press 18738 18739 Testcase: 18740 18741 In ~/.xbindkeysrc: 18742 "xterm &" 18743 XF86LaunchA 18744 18745 In ~/ov.xkb: 18746 xkb_keymap { 18747 xkb_keycodes { include "evdev" }; 18748 xkb_types { include "complete" }; 18749 xkb_compat { include "complete" 18750 interpret Overlay1_Enable+AnyOfOrNone(all) { 18751 action= SetControls(controls=Overlay1); 18752 }; 18753 }; 18754 xkb_symbols { include "pc+inet(evdev)+us" 18755 key <INS> { [ Overlay1_Enable ] }; 18756 key <AE01> { overlay1 = <AE02> }; // Insert+1 => 2 18757 key <TLDE> { overlay1 = <I128> }; // Insert+~ => XF86LaunchA 18758 }; 18759 xkb_geometry { include "pc(pc104)" }; 18760 }; 18761 18762 Apply this layout: 'xkbcomp ~/ov.xkb $DISPLAY'. 18763 Run "xbindkeys -n -v" 18764 In the exact order: 18765 - press Insert 18766 - press Tilde 18767 - release Insert 18768 - wait 18769 - release Tilde 18770 Keyboard input in the new terminal window(s) would be locked 18771 until another Insert+Tilde . 18772 18773 Reported-by: Mariusz Mazur <mariusz.g.mazur@gmail.com> 18774 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18775 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 18776 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 18777 18778commit 29a4f3db60fdeaef7bca3aa2746bb43a1850fddd 18779Author: Adam Jackson <ajax@redhat.com> 18780Date: Mon Jan 2 15:38:08 2017 -0500 18781 18782 glamor: Squash unused variable warning 18783 18784 Accidentally introduced in 05e19644. 18785 18786 Signed-off-by: Adam Jackson <ajax@redhat.com> 18787 18788commit 9874f73e88678c9eacbcba05e52336fc63a32712 18789Author: Kai-Heng Feng <kai.heng.feng@canonical.com> 18790Date: Thu Dec 15 13:56:38 2016 +0800 18791 18792 edid: Add quirk for ADA 1024x600 7" display. 18793 18794 Detailed mode reports 108 mm x 68 mm which is for smaller display. 18795 Maximum image size reports 15 cm x 10 cm which aligns with its physical 18796 size, use this size instead. 18797 18798 Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com> 18799 Acked-by: Alex Deucher <alexander.deucher@amd.com> 18800 18801commit 1b42f9505ff3a39b441464f553442079b750fe88 18802Author: Peter Hutterer <peter.hutterer@who-t.net> 18803Date: Thu Dec 8 14:32:06 2016 +1000 18804 18805 os: return 0 from check_timers if we touched any of them 18806 18807 Fixes a regression introduced in 0b2f30834b1a9f. If a driver posts input 18808 events during a timer function (wacom and synaptics do this during tap 18809 timeouts), ProcessInputEvents() is not called for these events. There are no 18810 new events on any fds, so the events just sit in the queue waiting for 18811 something else to happen. 18812 18813 Fix this by simply returning 0 from check_timers if we ran at least one of 18814 them or reset them all. This way the callers ospoll_wait will exit and 18815 continue with normal processing. 18816 18817 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 18818 Reviewed-by: Keith Packard <keithp@keithp.com> 18819 18820commit c96c860b6420adf0d004707a323af30491a1d7d3 18821Author: Michal Srb <msrb@suse.com> 18822Date: Mon Dec 12 17:45:22 2016 +0200 18823 18824 xinerama: Swap the response in RRXineramaWriteMonitor 18825 18826 Reviewed-by: Adam Jackson <ajax@redhat.com> 18827 18828commit 05e19644250698aa126a60bc671e85425df784d1 18829Author: Hans De Goede <hdegoede@redhat.com> 18830Date: Tue Dec 20 13:00:43 2016 +0100 18831 18832 glamor: Trust eglGetPlatformDisplayEXT if it exists 18833 18834 If the libEGL we are using has eglGetPlatformDisplayEXT, yet it still 18835 returns NULL, then this very likely means that it does not support the 18836 type (e.g. EGL_PLATFORM_GBM_MESA) passed in, and then returning NULL is 18837 the right thing to do. 18838 18839 This avoids falling back to an eglGetDisplay() implementation which does 18840 not understands the passed in gbm handle, treats it as a pointer to 18841 something else completely, followed by a crash sooner or later. 18842 18843 Specifically this fixes using the nvidia binary driver, with nvidia's 18844 libEGL + the modesetting driver on a secondary GPU crashing inside 18845 glamor_egl_init() sometimes. 18846 18847 Cc: Eric Anholt <eric@anholt.net> 18848 Reviewed-by: Adam Jackson <ajax@redhat.com> 18849 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 18850 18851commit 5b74e260e009d8bdf26433724495802b85cce7c3 18852Author: Mihail Konev <k.mvc@ya.ru> 18853Date: Thu Dec 29 12:37:18 2016 +0500 18854 18855 os,dix: Depend custom libs on libs, not objects 18856 18857 The custom os/os.O library reuses *.o files of os/libos.la. 18858 18859 The current rule assumes automake puts all the objects into per-target 18860 am__*_la_OBJECTS variable. At least with AC_REPLACE_FUNCS, this no 18861 longer holds (as wanted objects are put into LTLIBOBJS instead). 18862 18863 Depend on automake's result, the *.la library instead, to express demand 18864 of any its dependencies being built. 18865 18866 Should be fixing randomly occuring "undefined reference to `strlcpy'" 18867 errors when linking Xvfb and other DDX-es that could use os.O. 18868 18869 Reviewed-by: Adam Jackson <ajax@redhat.com> 18870 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 18871 18872commit b5dffbbac193aa640ffcfa0a431c21b862854e53 18873Author: Hans De Goede <hdegoede@redhat.com> 18874Date: Mon Dec 12 17:03:17 2016 +0100 18875 18876 xfree86: Add ModulePath support for OutputClass config Sections 18877 18878 Allow OutputClass config snippets to modify the module-path. 18879 18880 Note that any specified ModulePaths will be pre-pended to the normal 18881 ModulePath. The idea behind this is that any output hardware specific 18882 modules should have preference over the normal modules. 18883 18884 One use-case for this is the nvidia binary driver, this allows a 18885 config snippet like this: 18886 18887 Section "OutputClass" 18888 MatchDriver "nvidia" 18889 Modulepath "/usr/lib64/nvidia/modules" 18890 EndSection 18891 18892 To get the nvidia glx specific glx module loaded, but only when the 18893 nvidia kernel driver is loaded. 18894 18895 Together with the glvnd work done recently, this allows the nouveau 18896 + mesa and nvidia-binary userspace stacks to co-exist on the same 18897 system without any ldconfig / xorg.conf tweaking and the xserver will 18898 automatically do the right thing depending on which kernel driver 18899 (nouveau or nvidia) is loaded. 18900 18901 Reviewed-by: Adam Jackson <ajax@redhat.com> 18902 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 18903 18904commit d75ffcdbf8c1e3c8e0d46debcd533a9f2560f0a8 18905Author: Hans De Goede <hdegoede@redhat.com> 18906Date: Mon Dec 12 17:03:16 2016 +0100 18907 18908 xfree86: Allow overriding primary GPU detection from an OutputClass section 18909 18910 Allow using: 18911 18912 Option "PrimaryGPU" "yes" 18913 18914 In an OutputClass section to override the default primary GPU device 18915 selection which selects the GPU used as output by the firmware. 18916 18917 If multiple output devices match an OutputClass section with 18918 the PrimaryGPU option set, the first one enumerated becomes the 18919 primary GPU. 18920 18921 Reviewed-by: Adam Jackson <ajax@redhat.com> 18922 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 18923 18924commit ab1a65b7755d081b41188104b21f4d21eaa3187b 18925Author: Hans De Goede <hdegoede@redhat.com> 18926Date: Mon Dec 12 17:03:15 2016 +0100 18927 18928 xfree86: xf86platformProbe: split finding pci-info and setting primary GPU 18929 18930 This is a preparation patch for allowing an OutputClass section to 18931 override the default primary GPU device selection. 18932 18933 Reviewed-by: Adam Jackson <ajax@redhat.com> 18934 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 18935 18936commit 9cd3cc75269d9196898487b5712ee47b8291e077 18937Author: Hans De Goede <hdegoede@redhat.com> 18938Date: Mon Dec 12 17:03:14 2016 +0100 18939 18940 xfree86: Add options support for OutputClass Options 18941 18942 Add support for setting options in OutputClass Sections and having these 18943 applied to any matching output devices. 18944 18945 Reviewed-by: Adam Jackson <ajax@redhat.com> 18946 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 18947 18948commit 08b84d72878e43401e99059c3c926dfa42a360c3 18949Author: Hans De Goede <hdegoede@redhat.com> 18950Date: Mon Dec 12 17:03:13 2016 +0100 18951 18952 xfree86: Make OutputClassMatches take a xf86_platform_device 18953 18954 Make OutputClassMatches directly take a xf86_platform_device as argument, 18955 rather then an index into xf86_platform_devices. This makes things 18956 easier for callers which already have a xf86_platform_device pointer. 18957 18958 Reviewed-by: Adam Jackson <ajax@redhat.com> 18959 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 18960 18961commit c57c1e53ea3d76ebba5b2a23b7260817d3e6b921 18962Author: Hans De Goede <hdegoede@redhat.com> 18963Date: Mon Dec 12 17:03:12 2016 +0100 18964 18965 xfree86: Free devlist returned by xf86MatchDevice 18966 18967 xf86MatchDevice returns a dynamically allocated list of GDevPtr-s, 18968 free this when we're done with it. 18969 18970 Reviewed-by: Adam Jackson <ajax@redhat.com> 18971 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 18972 18973commit 32e632e85894eddc3ace83f16f1e973b1be478fe 18974Author: Adam Jackson <ajax@redhat.com> 18975Date: Mon Dec 12 13:08:17 2016 -0500 18976 18977 Revert "damage: Make damageRegionProcessPending take a damage not a drawable" 18978 18979 The commit message makes the assertion that the code below damage is not 18980 allowed to change whether there's a damage monitor for the drawable. 18981 That turns out not to be the case! exa's mixed code, at least, will 18982 create and destroy a damage in PrepareAccess. The destroy path can then 18983 be catastrophic, as damageRegionProcessPending will attempt to 18984 RegionEmpty memory from the middle of a freed block. 18985 18986 I'd wanted that invariant for performance, but faster isn't worth 18987 broken, so revert it. I think what exa's doing is reasonable, so the 18988 better way to improve performance for the unmonitored case is to either 18989 revisit dynamically wrapping into the GC, or inline damage into dix. 18990 18991 This reverts commit 4e124203f2260daaf54155f4a05fe469733e0b97. 18992 18993 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1389886 18994 Signed-off-by: Adam Jackson <ajax@redhat.com> 18995 18996commit 6dd65a6e9bd41237afe7615bea1cc9510e82781a 18997Author: Adam Jackson <ajax@redhat.com> 18998Date: Fri Dec 9 14:52:38 2016 -0500 18999 19000 shadow: Remove unused shadowAlloc 19001 19002 Signed-off-by: Adam Jackson <ajax@redhat.com> 19003 Reviewed-by: Keith Packard <keithp@keithp.com> 19004 19005commit 8d985c1e696bf46edaea0021f03f6f613ce37860 19006Author: Adam Jackson <ajax@redhat.com> 19007Date: Fri Dec 9 14:52:37 2016 -0500 19008 19009 shadow: Macro cleanup 19010 19011 shadowDamage is just obfuscation. The other two macros won't work 19012 outside shadow.c since the private key is in fact static there (meaning 19013 the extern decl is a lie). 19014 19015 Signed-off-by: Adam Jackson <ajax@redhat.com> 19016 Reviewed-by: Keith Packard <keithp@keithp.com> 19017 19018commit a530dc2a62349d922de77a75786a3795102aec11 19019Author: Adam Jackson <ajax@redhat.com> 19020Date: Fri Dec 9 14:52:36 2016 -0500 19021 19022 shadow: Remove indirection stubs 19023 19024 These are no longer used in the drivers. 19025 19026 Signed-off-by: Adam Jackson <ajax@redhat.com> 19027 Reviewed-by: Keith Packard <keithp@keithp.com> 19028 19029commit 1ae09800863992cfb1d5d21c600e2bc29e3b6ab2 19030Author: Adam Jackson <ajax@redhat.com> 19031Date: Fri Dec 9 14:52:35 2016 -0500 19032 19033 shadow: Lift 32->24 conversion from modesetting to dix 19034 19035 Signed-off-by: Adam Jackson <ajax@redhat.com> 19036 Reviewed-by: Keith Packard <keithp@keithp.com> 19037 19038commit 2b486f052d8a7d3bada9eb2615aa19d79e999cbe 19039Author: Adam Jackson <ajax@redhat.com> 19040Date: Fri Dec 9 14:52:34 2016 -0500 19041 19042 shadowfb: Rename this particular shadow.c to shadowfb.c 19043 19044 So as not to conflict with the one in miext/shadow. 19045 19046 Signed-off-by: Adam Jackson <ajax@redhat.com> 19047 Reviewed-by: Keith Packard <keithp@keithp.com> 19048 19049commit 98caba9f5669291e84f8610cc3071a9ac2df205f 19050Author: Adam Jackson <ajax@redhat.com> 19051Date: Fri Dec 9 14:52:29 2016 -0500 19052 19053 dix: Clear graphicsExposures for scratch GCs in one place 19054 19055 ... instead of in all the CreateScratchGC callers. 19056 19057 Signed-off-by: Adam Jackson <ajax@redhat.com> 19058 Reviewed-by: Keith Packard <keithp@keithp.com> 19059 19060commit e900a00f8dc5a1db63bafddd4ad1cde744975db1 19061Author: Adam Jackson <ajax@redhat.com> 19062Date: Fri Dec 9 14:52:28 2016 -0500 19063 19064 dix: Rename (and retype) PixmapPerDepth[1] to defaultStipple 19065 19066 Signed-off-by: Adam Jackson <ajax@redhat.com> 19067 Reviewed-by: Keith Packard <keithp@keithp.com> 19068 19069commit 8b335d9068fe4e1f1423a4d86c22b69ffcb819a5 19070Author: Ben Crocker <bcrocker@redhat.com> 19071Date: Tue Nov 15 16:34:38 2016 -0500 19072 19073 Fix a segfault that occurs if xorg.conf.d is absent: 19074 19075 In InitOutput, if xf86HandleConfigFile returns CONFIG_NOFILE 19076 (which it does if no config file or directory is present), the 19077 autoconfig flag is set, causing xf86AutoConfig to be called 19078 later on. 19079 19080 xf86AutoConfig calls xf86OutputClassDriverList via the 19081 call tree: 19082 19083 xf86AutoConfig => 19084 listPossibleVideoDrivers => 19085 xf86PlatformMatchDriver => 19086 xf86OutputClassDriverList 19087 19088 and xf86OutputClassDriverList attempts to traverse a linked list 19089 that is a member of the XF86ConfigRec struct pointed to by the 19090 global xf86configptr, which is NULL at this point because the 19091 XF86ConfigRec struct is only allocated (by xf86readConfigFile) 19092 AFTER the config file and directory have been successfully 19093 opened; the CONFIG_NOFILE return from xf86HandleConfigFile 19094 occurs BEFORE the call to xf86readConfigFile which allocates 19095 the XF86ConfigRec struct. 19096 19097 Rx: In read.c (for symmetry with xf86freeConfig, which already 19098 appears in this file), add a new function xf86allocateConfig 19099 which tests the value of xf86configptr and, if it's NULL, 19100 allocates the XF86ConfigRec struct and deposits the pointer 19101 in xf86configptr. In xf86Parser.h, add a prototype for the 19102 new xf86allocateConfig function. 19103 19104 Back in read.c, #include "xf86Config.h". In xf86readConfigFile, 19105 change the open-code call to calloc to a call to the new 19106 xf86allocateConfig function. 19107 19108 In xf86AutoConfig.c, add a call to the new xf86allocateConfig function 19109 to the beginning of xf86AutoConfig to make sure the XF86ConfigRec struct 19110 is allocated. 19111 19112 Reviewed-by: Adam Jackson <ajax@redhat.com> 19113 Signed-off-by: Ben Crocker <bcrocker@redhat.com> 19114 19115commit b365c5d16894a259dbf29db4ca2640d8ed768063 19116Author: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 19117Date: Wed Dec 7 15:43:06 2016 +0200 19118 19119 test: fix distributing scripts 19120 19121 Fix the following error on 'make distcheck': 19122 19123 make[6]: *** No rule to make target 'scripts/xvfb-piglit.sh', needed by 'scripts/xvfb-piglit.sh.log'. Stop. 19124 make[6]: Leaving directory '/home/pq/git/xserver/xorg-server-1.19.99.1/_build/sub/test' 19125 Makefile:1367: recipe for target 'check-TESTS' failed 19126 19127 The setup to trigger this is: 19128 $ ./configure --prefix=/home/pq/local --disable-docs 19129 --disable-devel-docs --enable-xwayland --disable-xorg --disable-xvfb 19130 --disable-xnest --disable-xquartz --disable-xwin --enable-debug 19131 19132 SCRIPT_TESTS is populated conditionally, but we should distribute the 19133 scripts in any case. 19134 19135 Reviewed-by: Adam Jackson <ajax@redhat.com> 19136 Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 19137 19138commit f31875510d818ba517f082e124adb294db906e51 19139Author: Adam Jackson <ajax@redhat.com> 19140Date: Tue Dec 6 14:11:11 2016 -0500 19141 19142 composite: Fix repaint of borders (v2) 19143 19144 When going from border width zero to a non-zero border width, the 19145 Composite extension is informed via the ConfigNotify callback. The 19146 call-chain looks like this: compConfigNotify -> compReallocPixmap -> 19147 compSetPixmap -> TraverseTree -> compSetPixmapVisitWindow. However, at 19148 this time, pWindow->borderWidth was not yet updated. Thus, HasBorder() 19149 is false and the window border will not be repainted. 19150 19151 To fix this, thread the new bw through to the window visitor, and 19152 inspect that rather than HasBorder(). For the other callers of 19153 compSetPixmap the border does not change size, so we can pass 19154 pWin->borderWidth instead. 19155 19156 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98499 19157 Signed-off-by: Adam Jackson <ajax@redhat.com> 19158 Reviewed-by: Keith Packard <keithp@keithp.com> 19159 19160commit 75c4f6e412e07c5d416fa9ad8d7defd972d2baa9 19161Author: Hans de Goede <hdegoede@redhat.com> 19162Date: Fri Sep 30 12:29:09 2016 +0200 19163 19164 xfree86: Try harder to find atleast 1 non GPU Screen 19165 19166 If we did not find any non GPU Screens, try again ignoring the notion 19167 of any video devices being the primary device. This fixes Xorg exiting 19168 with a "no screens found" error when using virtio-vga in a 19169 virtual-machine and when using a device driven by simpledrm. 19170 19171 This is a somewhat ugly solution, but it is the best I can come up with 19172 without major surgery to the bus and probe code. 19173 19174 Reviewed-by: Adam Jackson <ajax@redhat.com> 19175 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19176 19177commit 7121b03d324fccf687d49b63c53da7c8d93038c9 19178Author: Hans de Goede <hdegoede@redhat.com> 19179Date: Fri Sep 30 11:59:04 2016 +0200 19180 19181 xfree86: Make adding unclaimed devices as GPU devices a separate step 19182 19183 This is primarily a preparation patch for fixing the xserver exiting with 19184 a "no screens found" error even though there are supported video cards, 19185 due to the server not recognizing any card as the primary card. 19186 19187 This also fixes the (mostly theoretical) case of a platformBus capable 19188 driver adding a device as GPUscreen before a driver which only supports 19189 the old PCI probe method gets a chance to claim it as a normal screen. 19190 19191 Reviewed-by: Adam Jackson <ajax@redhat.com> 19192 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19193 19194commit b72d161cad15bec9dd082d30521c5383be526c56 19195Author: Hans de Goede <hdegoede@redhat.com> 19196Date: Tue Sep 27 14:30:10 2016 +0200 19197 19198 xfree86: Remove redundant ServerIsNotSeat0 check from xf86CallDriverProbe 19199 19200 If foundScreen is TRUE, then all the code below the removed if 19201 will not execute until we reach the return foundScreen; at the 19202 end, so this entire if block is redundant. 19203 19204 Reviewed-by: Adam Jackson <ajax@redhat.com> 19205 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19206 19207commit a46afee84d45fbff4e4dad9376afc95bbcc31d7c 19208Author: Hans de Goede <hdegoede@redhat.com> 19209Date: Tue Nov 22 14:53:20 2016 +0100 19210 19211 randr: rrCheckPixmapBounding: do not shrink the screen_pixmap 19212 19213 The purpose of rrCheckPixmapBounding is to make sure that the 19214 screen_pixmap is *large* enough for the slave-output which crtc is 19215 being configured. 19216 19217 However until now rrCheckPixmapBounding would also shrink the 19218 screen_pixmap in certain scenarios leading to various problems. 19219 19220 For example: Take a laptop with its internalscreen on a slave-output and 19221 currently disabled and an external monitor at 1920x1080+0+0. 19222 Now lets say that we want to drive the external monitor at its native 19223 resolution of 2560x1440 and have the internal screen mirror the top left 19224 part of the external monitor, so we run: 19225 19226 $ xrandr --output eDP --mode 1920x1080 --pos 0x0 --output HDMI \ 19227 --mode 2560x1440 --pos 0x0 19228 19229 Here xrandr utility first calls RRSetScreenSize to 2560x1440, then it 19230 calls RRSetCrtc 1920x1080+0+0 on the eDP, since this is a slave output, 19231 rrCheckPixmapBounding gets called and resizes the screen_pixmap to 19232 1920x1080, undoing the RRSetScreenSize. Then RRSetCrtc 2560x1440+0+0 19233 gets called on the HDMI, depending on crtc->transforms this will 19234 either result in a BadValue error from ProcRRSetCrtcConfig; or 19235 it will succeed, but the monitor ends up running at 2560x1440 19236 while showing a 1920x1080 screen_pixmap + black borders on the right 19237 and bottom. Neither of which is what we want. 19238 19239 This commit removes the troublesome shrinking behavior, fixing this. 19240 19241 Note: 19242 19243 1) One could argue that this will leave us with a too large screen_pixmap 19244 in some cases, but rrCheckPixmapBounding only gets called for slave 19245 outputs, so xrandr clients already must manually shrink the screen_pixmap 19246 after disabling crtcs in normal setups. 19247 19248 2) An alternative approach would be to also call rrCheckPixmapBounding 19249 on RRSetCrtc on normal (non-slave) outputs, but that would result in 19250 2 unnecessary resizes of the screen_pixmap in the above example, which 19251 seems undesirable. 19252 19253 Cc: Nikhil Mahale <nmahale@nvidia.com> 19254 Cc: Dave Airlie <airlied@redhat.com> 19255 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19256 Reviewed-by: Dave Airlie <airlied@redhat.com> 19257 19258commit 3b624aa9a9df86dc7d48149e0f18ca223b4355f1 19259Author: Hans de Goede <hdegoede@redhat.com> 19260Date: Tue Nov 22 14:48:14 2016 +0100 19261 19262 randr: rrCheckPixmapBounding: Do not substract crtc non 0 x,y from screen size 19263 19264 The purpose of rrCheckPixmapBounding is to make sure that the 19265 screen_pixmap is large enough for the slave-output which crtc is 19266 being configured. 19267 19268 This should include crtc->x and crtc->y, otherwise the crtc might 19269 still end up scanning out an area outside of the screen-pixmap. 19270 19271 For example: Take a laptop with an external monitor on a slave-output at 19272 1920x1080+0+0 and its internal-screen at 3840x2160+1920+0 and in 19273 gnome-settings-daemon move the external monitor to be on the ri ght of 19274 the internal screen rather then on the left. First g-s-d will do a 19275 RRSetScreenSize to 5760*2160 (which is a nop), then it calls RRSetCrtc 19276 to move the slave output to 1920x1080+3840+0, since this is a slave 19277 output, rrCheckPixmapBounding gets called, since the 2 crtcs now overlap 19278 the code before this commit would shrinks the screen_pixmap to 3180*2160. 19279 Then g-s-d calls RRSetCrtc to move the internal screen to 3180*2160+0+0. 19280 19281 And we end up with the slave-output configured to scan-out an area 19282 which completely falls outside of the screen-pixmap (and end up with 19283 a black display on the external monitor). 19284 19285 This commit fixes this by not substracting the x1 and y1 coordinates 19286 of the union-ed region when determining the new screen_pixmap size. 19287 19288 Cc: Nikhil Mahale <nmahale@nvidia.com> 19289 Cc: Dave Airlie <airlied@redhat.com> 19290 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19291 Reviewed-by: Dave Airlie <airlied@redhat.com> 19292 19293commit e1d30075c923f96a375895d74ea12a3c92a640c6 19294Author: Adam Jackson <ajax@redhat.com> 19295Date: Tue Nov 22 10:51:24 2016 -0500 19296 19297 configure: Enable glamor when building just Xwayland 19298 19299 Signed-off-by: Adam Jackson <ajax@redhat.com> 19300 Acked-by: Daniel Stone <daniels@collabora.com> 19301 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 19302 Tested-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 19303 Reviewed-by: Eric Anholt <eric@anholt.net> 19304 19305commit 59ad0e6a416d8e23f9d962af67a16ee28ec7867b 19306Author: Olivier Fourdan <ofourdan@redhat.com> 19307Date: Wed Nov 30 09:42:43 2016 +0100 19308 19309 xwayland: Fix use after free of cursors 19310 19311 Sometimes, Xwayland will try to use a cursor that has just been freed, 19312 leading to a crash when trying to access that cursor data either in 19313 miPointerUpdateSprite() or AnimCurTimerNotify(). 19314 19315 CheckMotion() updates the pointer's cursor based on which xwindow 19316 XYToWindow() returns, and Xwayland implements its own xwl_xy_to_window() 19317 to fake a crossing to the root window when the pointer has left the 19318 Wayland surface but is still within the xwindow. 19319 19320 But after an xwindow is unrealized, the last xwindow used to match the 19321 xwindows is cleared so two consecutive calls to xwl_xy_to_window() may 19322 not return the same xwindow. 19323 19324 To avoid this issue, update the last_xwindow based on enter and leave 19325 notifications instead of xwl_xy_to_window(), and check if the xwindow 19326 found by the regular miXYToWindow() is a child of the known last 19327 xwindow, so that multiple consecutive calls to xwl_xy_to_window() 19328 return the same xwindow, being either the one found by miXYToWindow() 19329 or the root window. 19330 19331 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1385258 19332 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 19333 Tested-by: Vít Ondruch <vondruch@redhat.com> 19334 Tested-by: Satish Balay <balay@fastmail.fm> 19335 Reviewed-by: Jonas Ådahl <jadahl@gmail.com> 19336 19337commit f43207c1c4a8487600cf3ea116c10437417c861b 19338Author: Olivier Fourdan <ofourdan@redhat.com> 19339Date: Thu Nov 3 09:59:30 2016 +0100 19340 19341 glamor: restore vfunc handlers on init failure 19342 19343 In glamor_init(), if the minimum requirements are not met, glamor may 19344 fail after setting up its own CloseScreen() and DestroyPixmap() 19345 routines, leading to a crash when either of the two routines is called 19346 if glamor failed to complete its initialization, e.g: 19347 19348 (EE) Backtrace: 19349 (EE) 0: Xwayland (OsSigHandler+0x29) 19350 (EE) 1: /lib64/libpthread.so.0 (__restore_rt+0x0) 19351 (EE) 2: Xwayland (glamor_sync_close+0x2a) 19352 (EE) 3: Xwayland (glamor_close_screen+0x52) 19353 (EE) 4: Xwayland (CursorCloseScreen+0x88) 19354 (EE) 5: Xwayland (AnimCurCloseScreen+0xa4) 19355 (EE) 6: Xwayland (present_close_screen+0x42) 19356 (EE) 7: Xwayland (dix_main+0x4f9) 19357 (EE) 8: /lib64/libc.so.6 (__libc_start_main+0xf1) 19358 (EE) 9: Xwayland (_start+0x2a) 19359 19360 Restore the previous CloseScreen() and DestroyPixmap() vfunc handlers in 19361 case of failure when checking for the minimum requirements, so that if 19362 any of the requirement is not met we don't leave the CloseScreen() and 19363 DestroyPixmap() from glamor handlers in place. 19364 19365 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1390018 19366 19367 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 19368 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19369 Reviewed-by: Eric Anholt <eric@anholt.net> 19370 19371commit d6a6e1d6abb110ff00ad31b94cd29d92ca7c71a5 19372Author: Peter Hutterer <peter.hutterer@who-t.net> 19373Date: Fri Nov 11 15:28:50 2016 +1000 19374 19375 Xi: when creating a new master device, update barries for all clients 19376 19377 The previous code only worked when the barrier was created by the same client 19378 as the one calling XIChangeDeviceHierarchy. 19379 19380 http://bugzilla.redhat.com/show_bug.cgi?id=1384432 19381 19382 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 19383 Reviewed-by: Daniel Stone <daniels@collabora.com> 19384 19385commit 5611585b87ce48428a66f98ece319a083f55d205 19386Author: Rui Matos <tiagomatos@gmail.com> 19387Date: Thu Nov 24 19:56:18 2016 +0100 19388 19389 xwayland: Don't send KeyRelease events on wl_keyboard::leave 19390 19391 Commits 816015648ffe660ddaa0f7d4d192e555b723c372 and 19392 fee0827a9a695600765f3d04376fc9babe497401 made it so that 19393 wl_keyboard::enter doesn't result in X clients getting KeyPress events 19394 while still updating our internal xkb state to be in sync with the 19395 host compositor. 19396 19397 wl_keyboard::leave needs to be handled in the same way as its 19398 semantics from an X client POV should be the same as an X grab getting 19399 triggered, i.e. X clients shouldn't get KeyRelease events for keys 19400 that are still down at that point. 19401 19402 This patch uses LeaveNotify for these events on wl_keyboard::leave and 19403 changes the current use of KeymapNotify to EnterNotify instead just to 19404 keep some symmetry between both cases. 19405 19406 On ProcessDeviceEvent() we still need to deactivate X grabs if needed 19407 for KeyReleases. 19408 19409 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 19410 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 19411 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 19412 19413commit 2de37eb71b928f6f50c109536c2f7b77293ad8e3 19414Author: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 19415Date: Wed Nov 23 09:54:27 2016 +0200 19416 19417 xwayland: fix order of calloc() args 19418 19419 The definition by the manual is: 19420 calloc(size_t nmemb, size_t size) 19421 19422 Swap the arguments of calloc() calls to be the right way around. 19423 19424 Presumably this makes no functional difference, but better follow the 19425 spec. 19426 19427 Signed-off-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 19428 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 19429 19430commit cf8860786c3e301486cd2853bc82977ba75e6b17 19431Author: Rhys Kidd <rhyskidd@gmail.com> 19432Date: Fri Oct 28 22:44:30 2016 -0400 19433 19434 test: Fix stray Makefile reference to removed os test 19435 19436 Fixes the following warning: 19437 19438 test/Makefile.am:69: warning: variable 'os_LDADD' is defined but no program or 19439 test/Makefile.am:69: library has 'os' as canonical name (possible typo) 19440 19441 Introduced upon the removal of test/os in: 19442 19443 commit 6a5a4e60373c1386b311b2a8bb666c32d68a9d99 19444 Author: Keith Packard <keithp@keithp.com> 19445 Date: Tue Dec 8 14:39:46 2015 -0800 19446 19447 Remove SIGIO support for input [v5] 19448 19449 This removes all of the SIGIO handling support used for input 19450 throughout the X server, preparing the way for using threads for input 19451 handling instead. 19452 19453 Places calling OsBlockSIGIO and OsReleaseSIGIO are marked with calls 19454 to stub functions input_lock/input_unlock so that we don't lose this 19455 information. 19456 19457 xfree86 SIGIO support is reworked to use internal versions of 19458 OsBlockSIGIO and OsReleaseSIGIO. 19459 19460 v2: Don't change locking order (Peter Hutterer) 19461 v3: Comment weird && FALSE in xf86Helper.c 19462 Leave errno save/restore in xf86ReadInput 19463 Squash with stub adding patch (Peter Hutterer) 19464 v4: Leave UseSIGIO config parameter so that 19465 existing config files don't break (Peter Hutterer) 19466 v5: Split a couple of independent patch bits out 19467 of kinput.c (Peter Hutterer) 19468 19469 Signed-off-by: Keith Packard <keithp@keithp.com> 19470 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 19471 19472 Signed-off-by: Rhys Kidd <rhyskidd@gmail.com> 19473 Reviewed-by: Keith Packard <keithp@keithp.com> 19474 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 19475 19476commit f0f8d5bb535814d5803df541b0ef4cb0354fabbe 19477Author: Keith Packard <keithp@keithp.com> 19478Date: Fri Nov 18 09:48:28 2016 -0800 19479 19480 Bump to post-1.19 development version 19481 19482 Signed-off-by: Keith Packard <keithp@keithp.com> 19483 19484commit 45af7fcb522464922831005d63cc362dc09f19a9 19485Author: Olivier Fourdan <ofourdan@redhat.com> 19486Date: Thu Nov 10 09:36:40 2016 +0100 19487 19488 xwayland: Remove MIPOINTER() definition 19489 19490 Not needed anymore now that mipointer exposes an API for that, 19491 miPointerInvalidateSprite() 19492 19493 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 19494 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 19495 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 19496 19497commit 7e5c9bdf73e4185ac4cf7f044da0793a5cc3e62c 19498Author: Keith Packard <keithp@keithp.com> 19499Date: Tue Nov 15 08:12:05 2016 -0800 19500 19501 Bump version to 1.19.0 19502 19503 Signed-off-by: Keith Packard <keithp@keithp.com> 19504 19505commit ea8efb3a997e15d27d362762e6c85d59487f9864 19506Author: Keith Packard <keithp@keithp.com> 19507Date: Wed Nov 2 13:25:33 2016 -0700 19508 19509 dix: Make sure client is not in output_pending chain after closed (RH 1382444) 19510 19511 I think it is possible that output could get queued to a client during 19512 CloseDownClient. After it is removed from the pending queue, active 19513 grabs are released, the client is awoken if sleeping and any work 19514 queue entries related to the client are processed. 19515 19516 To fix this, move the call removing it from the output_pending chain 19517 until after clientGone has been set and then check clientGone in 19518 output_pending_mark. 19519 19520 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1382444 19521 Signed-off-by: Keith Packard <keithp@keithp.com> 19522 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19523 19524commit 7513da40a656317ad3aa101651d29373de99c798 19525Author: Timo Aaltonen <tjaalton@ubuntu.com> 19526Date: Wed Nov 2 17:18:11 2016 +0200 19527 19528 dri2: Sync i965_pci_ids.h from mesa 19529 19530 Import changes from these mesa commits: 19531 85ea8deb26da420 i965: Removing PCI IDs that are no longer listed as Kabylake. 19532 bdff2e554735ed9 i956: Add more Kabylake PCI IDs. 19533 f1fa8b4a1ca73fa i965/bxt: Add 2x6 variant 19534 d1ab544bb883d04 i965/chv: Display proper branding 19535 20e8ee36627f874 i965/skl: Update Skylake renderer strings 19536 644c8a515192d28 i965/skl: Add two missing device IDs 19537 19538 Reviewed-by: Adam Jackson <ajax@redhat.com> 19539 Signed-off-by: Timo Aaltonen <tjaalton@ubuntu.com> 19540 19541commit 4cfee398726adf89db4b632ade7d6cab8b78282e 19542Author: Ian Ray <ian.ray@ge.com> 19543Date: Wed May 25 10:41:53 2016 +0300 19544 19545 xwayland-shm: block signals during fallocate 19546 19547 posix_fallocate() does an explicit rollback if it gets EINTR, and 19548 this is a problem on slow systems because when the allocation size 19549 is sufficiently large posix_fallocate() will always be interrupted 19550 by the smart scheduler's SIGALRM. 19551 19552 Changes since v1 - big comment in the code to explain what is going on 19553 19554 Reviewed-by: Adam Jackson <ajax@redhat.com> 19555 Signed-off-by: Ian Ray <ian.ray@ge.com> 19556 Acked-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 19557 Acked-by: Daniel Stone <daniels@collabora.com> 19558 19559commit 512e75a23b7f6339593c3ae6942f0c420a975fd3 19560Author: Keith Packard <keithp@keithp.com> 19561Date: Fri Oct 28 09:32:42 2016 -0700 19562 19563 Bump to 1.18.99.902 (1.19 RC2) 19564 19565 Signed-off-by: Keith Packard <keithp@keithp.com> 19566 19567commit c85f81825e196e96337347e0ce3a538fb2e38f16 19568Author: Keith Packard <keithp@keithp.com> 19569Date: Fri Oct 28 09:28:32 2016 -0700 19570 19571 dix: Bump MAXHASHSIZE for the resource db [v2] 19572 19573 [This was originally a workaround for a client-side resource leak: 19574 19575 http://lists.freedesktop.org/archives/xorg-devel/2012-November/034555.html 19576 19577 Obviously that's a broken app, but the performance problem it 19578 illustrates - that walking the linked list ends up burning all your CPU 19579 time - is real enough. - ajax] 19580 19581 v2: Replace with a shorter code sequence which computes the same 19582 results for all but numBits == 7 19583 19584 Reviewed-by: Adam Jackson <ajax@redhat.com> 19585 Signed-off-by: Keith Packard <keithp@keithp.com> 19586 19587commit 356db2340f5b473a7191c7969586ca5b0396c48f 19588Merge: 9ed5b2635 03d99ef72 19589Author: Keith Packard <keithp@keithp.com> 19590Date: Fri Oct 28 09:05:54 2016 -0700 19591 19592 Merge remote-tracking branch 'jturney/master' 19593 19594commit 9ed5b263542e5245317927828f0515db6c0a54c8 19595Author: Keith Packard <keithp@keithp.com> 19596Date: Fri Oct 28 08:04:43 2016 -0700 19597 19598 os: Recompute whether any clients are ready after ProcessWorkQueue() (bug 98030) 19599 19600 If a work proc wakes up a sleeping client and it is ready to execute, 19601 we need to re-compute the local 'are_ready' value before deciding 19602 what timeout value to use in WaitForSomething. 19603 19604 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=98030 19605 Signed-off-by: Keith Packard <keithp@keithp.com> 19606 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19607 19608commit 2c91f3235a4e1655e18cb014922529161051ffbb 19609Author: Keith Packard <keithp@keithp.com> 19610Date: Wed Oct 5 09:41:44 2016 -0700 19611 19612 ephyr: Leave window unmapped for -glamor-skip-present [v2] 19613 19614 If we're never painting anything in the window, we probably don't need 19615 to map it. 19616 19617 v2: Drop ephyr_glamor_gles2 from hostx.c 19618 19619 Signed-off-by: Keith Packard <keithp@keithp.com> 19620 Reviewed-by: Eric Anholt <eric@anholt.net> 19621 19622commit cba5a10fd93310702cad9dbe1e6d48da99f5552f 19623Author: Alex Goins <agoins@nvidia.com> 19624Date: Wed Oct 26 11:03:49 2016 -0700 19625 19626 ramdac: Check sPriv != NULL in xf86CheckHWCursor() 19627 19628 xf86CheckHWCursor() would dereference sPriv without NULL checking it. If Option 19629 "SWCursor" is specified, sPriv == NULL. In this case we should assume that HW 19630 cursors are not supported. 19631 19632 Signed-off-by: Alex Goins <agoins@nvidia.com> 19633 Reviewed-by: Andy Ritger <aritger@nvidia.com> 19634 Signed-off-by: Keith Packard <keithp@keithp.com> 19635 19636commit 03d99ef729178dd99268f185fb45320a29382091 19637Author: Jon Turney <jon.turney@dronecode.org.uk> 19638Date: Thu Oct 6 22:13:07 2016 +0100 19639 19640 glx/dri2: Don't build DRI loader if DRI2 isn't enabled 19641 19642 This partially reverts 501d8e2b. 19643 19644 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 19645 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 19646 19647commit 5cb328338684d8e5b03913c47475bfcd7acffec4 19648Author: Peter Hutterer <peter.hutterer@who-t.net> 19649Date: Wed Oct 26 12:21:16 2016 +0200 19650 19651 inputthread: On Linux leave the main thread's name as-is 19652 19653 On Linux, setting the main thread's name changes the program name 19654 (/proc/self/comm). Setting it to MainThread breaks scripts that rely on 19655 the command name, e.g. ps -C Xorg. 19656 19657 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 19658 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19659 19660commit 007f8ee61a35ceda36b43e772a9a1074b8e27a06 19661Author: Olivier Fourdan <ofourdan@redhat.com> 19662Date: Fri Oct 21 10:11:45 2016 +0200 19663 19664 xwayland: Activate and enable touch devices 19665 19666 On some random condition, a touch event may trigger a crash in Xwayland 19667 in GetTouchEvents(). 19668 19669 The (simplified) backtrace goes as follow: 19670 19671 (gdb) bt 19672 #0 GetTouchEvents() at getevents.c:1892 19673 #1 QueueTouchEvents() at getevents.c:1866 19674 #2 xwl_touch_send_event() at xwayland-input.c:652 19675 #5 wl_closure_invoke() from libwayland-client.so.0 19676 #6 dispatch_event() from libwayland-client.so.0 19677 #7 wl_display_dispatch_queue_pending() from libwayland-client.so.0 19678 #8 xwl_read_events() at xwayland.c:483 19679 #9 ospoll_wait() at ospoll.c:412 19680 #10 WaitForSomething() at WaitFor.c:222 19681 #11 Dispatch() at dispatch.c:412 19682 #12 dix_main() at main.c:287 19683 #13 __libc_start_main() at libc-start.c:289 19684 #14 _start () 19685 19686 The crash occurs when trying to access the sprite associated with the 19687 touch device, which appears to be NULL. Reason being the device itself 19688 is more a keyboard device than a touch device. 19689 19690 Moreover, it appears the device is neither enabled nor activated 19691 (inited=0, enabled=0) which doesn't seem right, but matches the code in 19692 init_touch() from xwayland-input.c which would enable the device if it 19693 was previously existing and otherwise would create the device but not 19694 activate it. 19695 19696 Make sure we do activate and enable touch devices just like we do for 19697 other input devices such as keyboard and pointer. 19698 19699 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 19700 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19701 19702commit f68ba7b81ffe765380664fccc92f3e689c6c48c2 19703Author: Rui Matos <tiagomatos@gmail.com> 19704Date: Tue Oct 25 19:24:49 2016 +0200 19705 19706 xwayland: Transform pointer enter event coordinates 19707 19708 Pointer enter event coordinates are surface relative and we need them to 19709 be screen relative for pScreen->SetCursorPosition(). 19710 19711 https://bugzilla.gnome.org/show_bug.cgi?id=758283 19712 19713 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 19714 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 19715 Reviewed-by: Jonas Ådahl <jadahl@gmail.com> 19716 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19717 19718commit f5c6d751d08c6de77c2ca49ba2a48f8023758cef 19719Author: Nikhil Mahale <nmahale@nvidia.com> 19720Date: Sat Oct 22 17:40:51 2016 +0530 19721 19722 modesetting: unifdef MODESETTING_OUTPUT_SLAVE_SUPPORT 19723 19724 Commit c7e8d4a6ee9542f56cd241cf7a960fb8223a6b22 had already unifdef 19725 MODESETTING_OUTPUT_SLAVE_SUPPORT but commit 19726 9257b1252da9092ddc676fec9aabe2b33dfad272 didn't notice that. 19727 19728 Signed-off-by: Nikhil Mahale <nmahale@nvidia.com> 19729 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19730 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19731 19732commit 4aaeeda4774397dd6d80aa240ca623ae795ec5dc 19733Author: Hans de Goede <hdegoede@redhat.com> 19734Date: Tue Oct 18 16:10:19 2016 +0200 19735 19736 xfree86: Xorg.wrap: Do not require root rights for cards with 0 outputs 19737 19738 Prior to this commit the Xorg.wrap code to detect if root rights are 19739 necessary checked for DRM_IOCTL_MODE_GETRESOURCES succeeding *and* 19740 reporting more then 0 output connectors. 19741 19742 DRM_IOCTL_MODE_GETRESOURCES succeeding alone is enough to differentiate 19743 between old drm only cards (which need ums and thus root) and kms capable 19744 cards. 19745 19746 Some hybrid gfx laptops have 0 output connectors on one of their 2 GPUs, 19747 resulting in Xorg needlessly running as root. This commits removes the 19748 res.count_connectors > 0 check, fixing this. 19749 19750 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19751 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 19752 19753commit 7d91063aca4e4d326c294e246bc2dc36cb05318e 19754Author: Michel Dänzer <michel.daenzer@amd.com> 19755Date: Mon Oct 17 18:48:44 2016 +0900 19756 19757 DRI2: Sync radeonsi_pci_ids.h from Mesa 19758 19759 Fixes DRI2 client driver name mapping for newer AMD GPUs with the 19760 modesetting driver, allowing the DRI2 extension to initialize. 19761 19762 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 19763 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 19764 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19765 19766commit f6ff2e974c5de3071c899eba828789f1d4d8645a 19767Author: Mihail Konev <k.mvc@ya.ru> 19768Date: Sat Oct 15 04:13:43 2016 +0000 19769 19770 modesetting: fix glamor ifdef 19771 19772 Add a missing ifdef needed for --disable-glamor. 19773 19774 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 19775 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 19776 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19777 19778commit 8fee6a917b6468e1b116d922f86484498874fb5c 19779Author: Mihail Konev <k.mvc@ya.ru> 19780Date: Fri Oct 14 02:40:59 2016 +0000 19781 19782 xwin: make glx optional again 19783 19784 Commit 501d8e2b removed --enable-aiglx, but made xwin always be 19785 --enable-glx. 19786 19787 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 19788 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 19789 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 19790 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 19791 19792commit d13cb974426f7f1110b0bdb08c4ebb46ff8975f7 19793Author: Peter Hutterer <peter.hutterer@who-t.net> 19794Date: Fri Oct 21 06:27:57 2016 +1000 19795 19796 ddx: add new call to purge input devices that weren't added 19797 19798 Special case for the systemd-logind case in xfree86: when we're vt-switched 19799 away and a device is plugged in, we get a paused fd from logind. Since we 19800 can't probe the device or do anything with it, we store that device in the 19801 xfree86 and handle it later when we vt-switch back. The device is not added to 19802 inputInfo.devices until that time. 19803 19804 When the device is removed while still vt-switched away, the the config system 19805 never notifies the DDX. It only runs through inputInfo.devices and our device 19806 was never added to that. 19807 19808 When a device is plugged in, removed, and plugged in again while vt-switched 19809 away, we have two entries in the xfree86-specific list that refer to the same 19810 device node, both pending for addition later. On VT switch back, the first one 19811 (the already removed one) will be added successfully, the second one (the 19812 still plugged-in one) fails. Since the fd is correct, the device works until 19813 it is removed again. The removed devices' config_info (i.e. the syspath) 19814 doesn't match the actual device we addded tough (the input number increases 19815 with each plug), it doesn't get removed, the fd remains open and we lose track 19816 of the fd count. Plugging the device in again leads to a dead device. 19817 19818 Fix this by adding a call to notify the DDX to purge any remainders of devices 19819 with the given config_info, that's the only identifiable bit we have at this 19820 point. 19821 19822 https://bugs.freedesktop.org/show_bug.cgi?id=97928 19823 19824 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 19825 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19826 19827commit 8fcf2fa78f09257933b17e7dc9a03a2034e3076f 19828Author: Peter Hutterer <peter.hutterer@who-t.net> 19829Date: Thu Oct 20 15:45:46 2016 +1000 19830 19831 xfree86: swap the list of paused devices to an xorg_list 19832 19833 No functional changes but it makes it easier to remove elements from the 19834 middle of the list (future patch). 19835 19836 We don't have an init call into this file, so the list is manually 19837 initialized. 19838 19839 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 19840 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19841 19842commit 28d8855cd4e3be8831fb1c17bf1d205bd8465d1e 19843Author: Peter Hutterer <peter.hutterer@who-t.net> 19844Date: Thu Oct 20 15:25:51 2016 +1000 19845 19846 xfree86: use the right option traversal list to search for an option 19847 19848 They're identically laid-out structs but let's use the right type to search 19849 for our desired value. 19850 19851 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 19852 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19853 19854commit 7fc96fb02dade4a86f2fc038f3cf5f2d9c0cda00 19855Author: Emil Velikov <emil.velikov@collabora.com> 19856Date: Fri Oct 21 19:06:00 2016 +0100 19857 19858 glamor: don't look for non-existing EGL_KHR_platform_base 19859 19860 The extension does not exist in the registry, thus needs to know they're 19861 using EGL 1.5 in order to determine the eglGetPlatformDisplay function 19862 pointer is valid. 19863 19864 Thus brings us into some lovely circular dependency. 19865 19866 Since mesa won't be able (in the foreseeable future) to export the KHR 19867 flavour of extension (another way one could assume that EGL 1.5 is 19868 available) just drop all the heuristics and use the 19869 EGL_EXT_platform_base extension. 19870 19871 In practise (checked with the Mali driver) any EGL 1.5 driver will 19872 advertise support for EGL_EXT_platform_base. 19873 19874 Reviewed-by: Adam Jackson <ajax@redhat.com> 19875 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 19876 19877commit 5dcb0666b82f5ab00f3d22e86f05ac14b0d5341e 19878Author: Mihail Konev <k.mvc@ya.ru> 19879Date: Sat Oct 15 17:59:05 2016 +0000 19880 19881 os/inputthread: Ensure pollfd refreshing 19882 19883 When putting a device node into a poll-request list, do not overwrite a 19884 "please-remove" element with the same fd, so that a closed device file 19885 is ospoll_remove'd prior to being ospoll_add'ed. 19886 19887 Before, the opposite order was possible, resulting in ospoll_add 19888 considering the newly opened file being already polled, should it have a 19889 fd for which the "please-remove" has not been procesed yet. In this 19890 case, no further events would be seen from the device. 19891 19892 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 19893 Regressed-in: 52d6a1e832a5e62289dd4f32824ae16a78dfd7e8 19894 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97880 19895 Patchwork: https://patchwork.freedesktop.org/patch/113763/ 19896 Hit-and-Reduced-by: Hans de Goede <hdegoede@redhat.com> 19897 Reviewed-and-Reduced-by: Peter Hutterer <peter.hutterer@who-t.net> 19898 Reviewed-by: Keith Packard <keithp@keithp.com> 19899 Signed-off-by: Keith Packard <keithp@keithp.com> 19900 19901commit 9cf0bd4d4507dca6234024605b14724713f2109e 19902Author: Michel Dänzer <michel.daenzer@amd.com> 19903Date: Wed Oct 5 18:28:45 2016 +0900 19904 19905 xf86Cursor: Take the input lock in xf86Set/MoveCursor 19906 19907 Prevents the HW cursor from intermittently jumping around when the 19908 cursor image is changed while the cursor is being moved. This is hardly 19909 noticeable in normal operation but can be quite confusing when stepping 19910 through these codepaths in a debugger. 19911 19912 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19913 19914commit 011ce3297d924e78ef0254b0451561946bd8be8d 19915Author: Michel Dänzer <michel.daenzer@amd.com> 19916Date: Wed Oct 5 18:01:29 2016 +0900 19917 19918 xf86Cursor: Use PRIME master xf86CursorScreenRec::HotX/Y for slaves 19919 19920 xf86CursorScreenRec::HotX/Y contain 0 for PRIME slave screens. 19921 19922 Fixes incorrect HW cursor position on PRIME slave screens. 19923 19924 Also hoist the hotspot translation out from xf86ScreenSet/MoveCursor to 19925 xf86Set/MoveCursor, since the hotspot position is a property of the 19926 cursor, not the screen. 19927 19928 v2: 19929 * Squash patches 1 & 2 of the v1 series, since it's basically the same 19930 problem 19931 * Use the master screen's xf86CursorScreenRec::HotX/Y instead of 19932 CursorRec::bits->x/yhot, since CursorRec::bits can be NULL (Hans de 19933 Goede) 19934 19935 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 19936 19937commit e3f53aa3a8dfc323c75837cd41e35b8e895d547e 19938Author: Adam Jackson <ajax@redhat.com> 19939Date: Thu Oct 6 14:42:34 2016 -0400 19940 19941 glx/dri2: Don't include drm headers 19942 19943 They're not needed, and they won't be present on win32. 19944 19945 Signed-off-by: Adam Jackson <ajax@redhat.com> 19946 Reviewed-by: Emil Velikov <emil.l.veikov@gmail.com> 19947 19948commit 97a8353ec1192d8d3bd2ebb99e5687cb91427e09 19949Author: Peter Harris <pharris@opentext.com> 19950Date: Thu Sep 29 11:17:48 2016 -0400 19951 19952 Fix id in error when resource does not exist 19953 19954 Always set client->errorValue before returning an error. 19955 19956 Test program: 19957 19958 int main(int argc, char *argv[]) { 19959 int screen = 0; 19960 xcb_connection_t *c = xcb_connect(NULL, &screen); 19961 if (!c) { 19962 printf("Cannot connect\n"); 19963 return 1; 19964 } 19965 19966 xcb_void_cookie_t tok = xcb_free_gc_checked(c, BAD_VALUE); 19967 xcb_g_context_error_t *err = (xcb_g_context_error_t *)xcb_request_check(c, tok); 19968 if (!err) { 19969 printf("Unexpected request success\n"); 19970 return 1; 19971 } 19972 if (err->bad_value != BAD_VALUE) { 19973 printf("Error: Got 0x%X, expected 0x%X\n", err->bad_value, BAD_VALUE); 19974 return 1; 19975 } 19976 printf("Success! Got expected bad value of 0x%X\n", BAD_VALUE); 19977 return 0; 19978 } 19979 19980 Reviewed-by: Adam Jackson <ajax@redhat.com> 19981 Signed-off-by: Peter Harris <pharris@opentext.com> 19982 19983commit f4a41155479e68bf55740c1dfffafc78e4c02087 19984Author: Adam Jackson <ajax@redhat.com> 19985Date: Wed Oct 5 12:34:34 2016 -0400 19986 19987 glamor: Use eglGetPlatformDisplay{,EXT} if we can 19988 19989 eglGetDisplay forces the implementation to guess which kind of display 19990 it's been handed. glvnd does something different from Mesa, and in 19991 general it's impossible for the library to get this right. Add a new 19992 inline that gets the logic right, and works around a quirk in epoxy. 19993 19994 Signed-off-by: Adam Jackson <ajax@redhat.com> 19995 Reviewed-by: Eric Anholt <eric@anholt.net> 19996 19997commit 95d3980c7c991b2cc8dcadac173635641ae15865 19998Author: Michel Dänzer <michel.daenzer@amd.com> 19999Date: Wed Oct 5 17:29:36 2016 +0900 20000 20001 test: Use $XSERVER_BUILDDIR for Xvfb executable path 20002 20003 Fixes make check with out-of-tree builds. 20004 20005 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 20006 Reviewed-by: Keith Packard <keithp@keithp.com> 20007 Reviewed-by: Eric Anholt <eric@anholt.net> 20008 20009commit cc69d4f110bd8a87b9e6c238ebebf8d85cbd93ec 20010Author: Emil Velikov <emil.l.velikov@gmail.com> 20011Date: Thu Sep 29 18:35:04 2016 +0100 20012 20013 xfree86/dri: remove libdrm runtime checks 20014 20015 As of last commit all the places in our configure.ac require version 20016 2.3.1 (released back in 2007) or later. With the latter introducing the 20017 1.3.0 version, as returned by drmGetLibVersion. 20018 20019 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 20020 Reviewed-by: Adam Jackson <ajax@redhat.com> 20021 20022commit 962962e978e5cededc20e577a36f77eb24d14bda 20023Author: Emil Velikov <emil.l.velikov@gmail.com> 20024Date: Thu Sep 29 18:35:03 2016 +0100 20025 20026 configure.ac: bump the required libdrm version to 2.3.1 20027 20028 It fixes a bug where the version reported by drmGetLibVersion() was 20029 'wrong'. See libdrm commit 79038751ffe("libdrm: add support for server 20030 side functionality in libdrm") for details. 20031 20032 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 20033 Reviewed-by: Adam Jackson <ajax@redhat.com> 20034 20035commit 74a8b320fa5068cd86d8b8b8e73fa92caae9a8a6 20036Author: Emil Velikov <emil.l.velikov@gmail.com> 20037Date: Thu Sep 29 18:35:02 2016 +0100 20038 20039 configure.ac: use $LIBDRM over libdrm when using pkg-config 20040 20041 The former contains the minimum required version which is required. 20042 Strictly speaking Xephyr/Xwayland may require version greater than the 20043 current 2.3.0, although I've personally haven't checked the specifics. 20044 20045 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 20046 Reviewed-by: Adam Jackson <ajax@redhat.com> 20047 20048commit 45e1220486139c483a69366323f8f590beffa19d 20049Author: Emil Velikov <emil.l.velikov@gmail.com> 20050Date: Thu Sep 29 18:35:01 2016 +0100 20051 20052 configure.ac: default to DRI=yes on solaris platforms 20053 20054 Afaict there's little-to-no reason/way one would want xserver without 20055 DRI support on Solaris platforms. 20056 20057 This will allow us to simplify/fix all the libdrm detection in the next 20058 commit. 20059 20060 Cc: Alan Coopersmith <alan.coopersmith@oracle.com> 20061 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 20062 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 20063 Reviewed-by: Adam Jackson <ajax@redhat.com> 20064 20065commit 501d8e2beb337e072c93c9310fcd927a099b9c3b 20066Author: Emil Velikov <emil.l.velikov@gmail.com> 20067Date: Thu Sep 29 18:41:19 2016 +0100 20068 20069 configure.ac: remove --enable-aiglx option 20070 20071 Presently the option guards both direct and accelerated indirect GLX. As 20072 such when one toggles it off they end up without any acceleration. 20073 20074 Remove the option all together until we have the time to split/rework 20075 things. 20076 20077 Cc: Jon Turney <jon.turney@dronecode.org.uk> 20078 Reviewed-by: Adam Jackson <ajax@redhat.com> 20079 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 20080 20081commit 7ec350ddd42479595f0ea88f86085af941913617 20082Author: Emil Velikov <emil.l.velikov@gmail.com> 20083Date: Thu Sep 29 18:41:18 2016 +0100 20084 20085 xfree86: remove aiglx cmd/xorg.conf option 20086 20087 The option is misleading and using it leads to disabling both direct and 20088 accelerated indirect GLX. In such cases the xserver GLX attempts to 20089 match DRISW (IGLX) configs with the DRI2/3 ones (direct GLX) leading to 20090 all sorts of fun experience. 20091 20092 Remove the option until we get a clear split and control over direct vs 20093 indirect GLX. 20094 20095 Reviewed-by: Adam Jackson <ajax@redhat.com> 20096 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 20097 20098commit 04ef8558a731bf070abf1b40c7e6b54aad8f5f31 20099Author: Emil Velikov <emil.l.velikov@gmail.com> 20100Date: Thu Sep 29 18:41:17 2016 +0100 20101 20102 glx: drisw is not accelerated IGLX, reflect that in log messages 20103 20104 The messages from glxdricommon.c (used by drisw) still have the A, but 20105 at least we're don't have it locally. 20106 20107 Reviewed-by: Adam Jackson <ajax@redhat.com> 20108 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 20109 20110commit 1c2fcb95484777ca9aa80b3f814ad64e81f825f1 20111Author: Olivier Fourdan <ofourdan@redhat.com> 20112Date: Wed Oct 5 08:36:21 2016 +0200 20113 20114 glamor: Fix pixmap offset for bitplane in glamor_copy_fbo_cpu 20115 20116 Commit cba28d5 - "glamor: Handle bitplane in glamor_copy_fbo_cpu" 20117 introduced a regression as the computed pixmap offset would not match 20118 the actual coordinates and write data elsewhere in memory causing a 20119 segfault in fbBltOne(). 20120 20121 Translate the pixmap coordinates so that the data is read and written at 20122 the correct location. 20123 20124 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97974 20125 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 20126 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 20127 20128commit a6e85e6330adcdcbcd939c0963daaecc96d41a2a 20129Author: Jonas Ådahl <jadahl@gmail.com> 20130Date: Tue Sep 13 15:17:08 2016 +0800 20131 20132 xwayland: Add pointer warp emulator 20133 20134 Emulate pointer warps by locking the pointer and sending relative 20135 motion events instead of absolute. X will keep track of the "fake" 20136 pointer cursor position given the relative motion events, and the 20137 client warping the cursor will warp the faked cursor position. 20138 20139 Various requirements need to be met for the pointer warp emulator to 20140 enable: 20141 20142 The cursor must be invisible: since it would not be acceptable that a 20143 fake cursor position would be different from the visual representation 20144 of the cursor, emulation can only be done when there is no visual 20145 representation done by the Wayland compositor. Thus, for the emulator 20146 to enable, the cursor must be hidden, and would the cursor be displayed 20147 while the emulator is active, the emulator would be destroyed. 20148 20149 The window that is warped within must be likely to have pointer focus. 20150 For example, warping outside of the window region will be ignored. 20151 20152 The pointer warp emulator will disable itself once the fake cursor 20153 position leaves the window region, or the cursor is made visible. 20154 20155 This makes various games depending on pointer warping (such as 3D 20156 first-person shooters and stategy games using click-to-drag-map like 20157 things) work. 20158 20159 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20160 20161commit 467ab142fff926e1475440dd5f649a49f45808fa 20162Author: Jonas Ådahl <jadahl@gmail.com> 20163Date: Tue Sep 13 15:17:07 2016 +0800 20164 20165 xwayland: Translate a pointer grab with confineTo to pointer confinement 20166 20167 Translate grabbing a pointer device with confineTo set to a window into 20168 confining the Wayland pointer using the pointer constraints protocol. 20169 This makes clients that depend on the pointer not going outside of the 20170 window region, such as certain games and virtual machines viewers, to 20171 function more properly. 20172 20173 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20174 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 20175 20176commit ca7b593fbe54bc9a0b44037e62e4b4401cbd375e 20177Author: Jonas Ådahl <jadahl@gmail.com> 20178Date: Thu Sep 29 10:42:13 2016 +0800 20179 20180 xwayland: Bind pointer constraints global 20181 20182 Will be used by the pointer warp emulator. 20183 20184 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20185 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 20186 20187commit c14a8c6cc0fcd56c380d1220c2a8f04b74edee93 20188Author: Jonas Ådahl <jadahl@gmail.com> 20189Date: Tue Sep 13 15:17:05 2016 +0800 20190 20191 xwayland: Put getting a xwl_window from a Window in a helper 20192 20193 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20194 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 20195 20196commit b4644ce8d3420447a8e5a2339238968da0a59de7 20197Author: Jonas Ådahl <jadahl@gmail.com> 20198Date: Tue Sep 13 15:17:04 2016 +0800 20199 20200 xwayland: Set unaccelerated pointer motion delta if available 20201 20202 If there was an relative pointer motion within the same frame as an 20203 absolute pointer motion, provide both the absolute coordinate and the 20204 unaccelerated delta when setting the valuator mask. 20205 20206 If a frame contained only a relative motion, queue an absolute motion 20207 with an unchanged position, but still pass the unaccelerated motion 20208 event. 20209 20210 If the wl_seat advertised by the compositor is not new enough, assume 20211 each relative and absolute pointer motion arrives within their own 20212 separate frames. 20213 20214 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20215 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 20216 20217commit aa9634d03bc2434dfd25476529eccb14e46fcfdc 20218Author: Jonas Ådahl <jadahl@gmail.com> 20219Date: Tue Sep 13 15:17:03 2016 +0800 20220 20221 xwayland: Dispatch pointer motion events on wl_pointer.frame if possible 20222 20223 Wait until wl_pointer.frame with dispatching the pointer motion event, 20224 if wl_pointer.frame is supported by the compositor. This will later be 20225 used to combine unaccelerated motion deltas with the absolute motion 20226 delta. 20227 20228 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20229 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 20230 20231commit 42e8902395cb27af5c28306abd577a92c467a62d 20232Author: Krzysztof Sobiecki <sobkas@gmail.com> 20233Date: Tue Sep 13 22:23:49 2016 +0800 20234 20235 xwayland: Add a new input device used for pointer warping/locking 20236 20237 Generating relative and absolute movement events from the same input 20238 device is problematic, because an absolute pointer device doesn't 20239 expect to see any relative motion events. To be able to generate 20240 relative pointer motion events including unaccelerated deltas, create a 20241 secondary pointer device 'xwayland-relative-pointer', and use that for 20242 emitting relative motion events. 20243 20244 Signed-off-by: Krzysztof Sobiecki <sobkas@gmail.com> 20245 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20246 20247commit 011ada724afdba8955f1d4844b306e61390eead8 20248Author: Jonas Ådahl <jadahl@gmail.com> 20249Date: Tue Sep 13 15:17:01 2016 +0800 20250 20251 xwayland: Move pointer button initialization into helper 20252 20253 We'll later use this for initializing buttons for the relative pointer 20254 since they need to be the same. 20255 20256 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20257 20258commit a77d0715c6272cc1778a54dccd8cb68dc28cd761 20259Author: Jonas Ådahl <jadahl@gmail.com> 20260Date: Tue Sep 13 15:17:00 2016 +0800 20261 20262 xwayland: Split up device class init/release into functions 20263 20264 Put device class initialization in init_[device_class](xwl_seat) and 20265 releasing in release_[device class](xwl_seat). The purpose is to make 20266 it easier to add more type of initialization here later, without making 20267 the function too large. 20268 20269 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20270 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 20271 20272commit 9037ba736a0424feee2fb6ac20cf7687613dc452 20273Author: Jonas Ådahl <jadahl@gmail.com> 20274Date: Thu Sep 29 10:40:01 2016 +0800 20275 20276 xwayland: Bind the relative pointer manager 20277 20278 Will be used for getting unaccelerated motion events and later for 20279 relative motions used by a pointer warp emulator. 20280 20281 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20282 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 20283 20284commit 0fae3be0686cae746e03d6e4592f97278cc2275d 20285Author: Jonas Ådahl <jadahl@gmail.com> 20286Date: Tue Sep 13 15:16:57 2016 +0800 20287 20288 dix: Add valuator_mask_set_absolute_unaccelerated 20289 20290 Add a valuator mask setter for setting absolute coordinate combined 20291 with unaccelerated motion deltas. This will later be used by Xwayland 20292 to combine a wl_pointer.motion() event with the unaccelerated delta of 20293 a wp_relative_pointer.relative_motion() event. 20294 20295 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 20296 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 20297 20298commit bbd4854f81ebba1119202c5f6ff4679c62afec1f 20299Author: Carlos Garnacho <carlosg@gnome.org> 20300Date: Tue Sep 27 19:03:26 2016 +0200 20301 20302 xwayland: Apply touch abs axes transformation before posting events 20303 20304 The way we map the touch absolute device to screen coordinates can't 20305 work across wl_output mode and geometry events. Instead, set up 20306 a fixed coordinate space, and transform touch events according to 20307 the screen coordinate space as they happen. 20308 20309 Reviewed-by: Adam Jackson <ajax@redhat.com> 20310 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 20311 20312commit ee526285882995289846648f3122c4295e3e8284 20313Author: Carlos Garnacho <carlosg@gnome.org> 20314Date: Wed Sep 28 12:35:36 2016 +0200 20315 20316 xwayland: Apply "last pointer window" check only to the pointer device 20317 20318 The checks in xwayland's XYToWindow handler pretty much assumes that the 20319 sprite is managed by the wl_pointer, which is not entirely right, given 20320 1) The Virtual Core Pointer may be controlled from other interfaces, and 20321 2) there may be other SpriteRecs than the VCP's. 20322 20323 This makes XYToWindow calls return a sprite trace with just the root 20324 window if any of those two assumptions are broken, eg. on touch events. 20325 20326 So turn the check upside down, first assume that the default XYToWindow 20327 proc behavior is right, and later cut down the spriteTrace if the 20328 current device happens to be the pointer and is out of focus. We work 20329 our way to the device's lastSlave here so as not to break assumption #1 20330 above. 20331 20332 Reviewed-by: Adam Jackson <ajax@redhat.com> 20333 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 20334 Acked-by: Olivier Fourdan <ofourdan@redhat.com> 20335 20336commit 2b6a068d21997ca812e665ed058e72eb4626c129 20337Author: Keith Packard <keithp@keithp.com> 20338Date: Mon Oct 3 15:37:03 2016 -0700 20339 20340 fb: XYPixmap format PutImage includes all planes in depth 20341 20342 Unlike GetImage, for which the provided planemask restricts the data 20343 delivered, for PutImage in XYPixmap format, all of the planes in the 20344 drawable depth are sent and those outside the plane mask are simply 20345 ignored. 20346 20347 Reviewed-by: Adam Jackson <ajax@redhat.com> 20348 Signed-off-by: Keith Packard <keithp@keithp.com> 20349 20350commit c9b8ce73927a43ed82b267d1c1985445f24827cc 20351Author: Mark Yao <mark.yao@rock-chips.com> 20352Date: Fri Sep 30 16:42:35 2016 +0800 20353 20354 glamor: spans: fixup wrong count on glDrawArrays 20355 20356 In commit 9e9fcf5 (glamor: Add a helper function for the common 20357 GL_QUADS fallback pattern.), the glDrawArrays count change was 20358 accidentally changed to nbox. 20359 20360 Fixes xlogo with MESA_GL_VERSION_OVERRIDE=2.1 and 20361 MESA_GLSL_VERSION_OVERRIDE=120 20362 20363 Signed-off-by: Mark Yao <mark.yao@rock-chips.com> 20364 Reviewed-by: Eric Anholt <eric@anholt.net> 20365 20366commit 2aca2dadda4cc9c378049457885d33a4eede9768 20367Author: Eric Anholt <eric@anholt.net> 20368Date: Sat Sep 24 14:42:28 2016 -0700 20369 20370 glamor: Fix link failure on GLES2. 20371 20372 Current Mesa requires that the precision qualifier on uniforms matches 20373 between stages, even if (as in this case) the uniform isn't used in 20374 one of the stages. 20375 20376 Signed-off-by: Eric Anholt <eric@anholt.net> 20377 Reviewed-by: Keith Packard <keithp@keithp.com> 20378 20379commit 20fcdfcf39497fe44d2a3200338523effb8e2bc6 20380Author: Eric Anholt <eric@anholt.net> 20381Date: Sat Sep 24 14:30:27 2016 -0700 20382 20383 glamor: Remove #if 0-ed picture dumping code. 20384 20385 I don't think anybody has run this code since it was pulled into the 20386 server. 20387 20388 Signed-off-by: Eric Anholt <eric@anholt.net> 20389 Reviewed-by: Keith Packard <keithp@keithp.com> 20390 20391commit 4b5326aeba539249fcded91bf7806a708eeca651 20392Author: Eric Anholt <eric@anholt.net> 20393Date: Sat Sep 24 14:04:14 2016 -0700 20394 20395 glamor: Remove many unused glamor util functions. 20396 20397 Signed-off-by: Eric Anholt <eric@anholt.net> 20398 Reviewed-by: Keith Packard <keithp@keithp.com> 20399 20400commit 117d614d1ba324bdb02c50a430c2e0d22a384f03 20401Author: Eric Anholt <eric@anholt.net> 20402Date: Sat Sep 24 22:26:17 2016 +0300 20403 20404 glamor: Require GL_OES_texture_border_clamp for GLES2. 20405 20406 The extension came out in 2000, and all Mesa-supported hardware that 20407 can do glamor supports it. We were already relying on the ARB version 20408 being present on desktop. 20409 20410 Signed-off-by: Eric Anholt <eric@anholt.net> 20411 Reviewed-by: Keith Packard <keithp@keithp.com> 20412 20413commit 7a5ddf8da5ed817aa2dd6f7af1b9197a5ceec60b 20414Author: Adam Jackson <ajax@redhat.com> 20415Date: Wed Sep 28 15:27:20 2016 -0400 20416 20417 test: Re-enable a couple of GetImage tests 20418 20419 Fixed in 6c6f09aac. 20420 20421 Signed-off-by: Adam Jackson <ajax@redhat.com> 20422 Reviewed-by: Eric Anholt <eric@anholt.net> 20423 20424commit c7a9161da5eb80ef0444601621b52314255b0ce5 20425Author: Francois Tigeot <ftigeot@wolfpond.org> 20426Date: Fri Sep 23 11:36:49 2016 +0200 20427 20428 Enable XTRANS_SEND_FDS on FreeBSD, DragonFly and OpenBSD 20429 20430 This code is based on local patches which had been sitting in 20431 FreeBSD and OpenBSD ports. 20432 20433 Reviewed-by: Matthieu Herrb <matthieu@herrb.eu> 20434 Signed-off-by: François Tigeot <ftigeot@wolfpond.org> 20435 20436commit 6c6f09aac7f1d1367a042087b7681c7fdf1d1e0f 20437Author: Keith Packard <keithp@keithp.com> 20438Date: Sun Sep 25 14:19:08 2016 +0300 20439 20440 xace: Don't censor window borders 20441 20442 GetImage is allowed to return window border contents, so don't remove 20443 that from the returned image. 20444 20445 Reviewed-by: Adam Jackson <ajax@redhat.com> 20446 Signed-off-by: Keith Packard <keithp@keithp.com> 20447 20448commit 380c2ca25ec4dd330f938e382ff4af60bc253785 20449Author: Hans de Goede <hdegoede@redhat.com> 20450Date: Fri Sep 23 14:11:52 2016 +0300 20451 20452 XF86VidMode: Fix free() on walked pointer 20453 20454 Based on: https://patchwork.freedesktop.org/patch/85636/ 20455 20456 Rewritten to just not walk the pointer. 20457 20458 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20459 Reviewed-by: Emi Velikov <emil.l.velikov@gmail.com> 20460 20461commit 220d327ee00ec1278d50d4ba8cd20612e8b77c8f 20462Author: Hans de Goede <hdegoede@redhat.com> 20463Date: Fri Sep 23 14:06:52 2016 +0300 20464 20465 Xext: Fix a memory leak 20466 20467 Based on: https://patchwork.freedesktop.org/patch/85636/ 20468 20469 Rewritten to also free the resources allocated by 20470 panoramix_setup_ids(). 20471 20472 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20473 Reviewed-by: Emi Velikov <emil.l.velikov@gmail.com> 20474 20475commit 3abf791ab82cb3dc7eacd0317fe0262a71340753 20476Author: Michael Thayer <michael.thayer@oracle.com> 20477Date: Fri Sep 16 17:51:25 2016 +0200 20478 20479 modesetting: only fall back to drmModeSetCursor() on -EINVAL 20480 20481 This change effectively reverts commit 074cf58. We were falling back from 20482 drmModeSetCursor2() to drmModeSetCursor() whenever the first failed. This 20483 fall-back only makes sense on pre-mid-2013 kernels which implemented the 20484 cursor_set hook but not cursor_set2, and in this case the call to 20485 drmModeSetCursor2() will always return -EINVAL. Specifically, a return 20486 value of -ENXIO usually means that neither are supported. 20487 20488 Signed-off-by: Michael Thayer <michael.thayer@oracle.com> 20489 [hdegoede@redhat.com: initialize ret to -EINVAL] 20490 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 20491 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20492 20493commit 363f4273dd4aec3e26cc57ecb6c20f27e6c813d8 20494Author: Daniel Martin <consume.noise@gmail.com> 20495Date: Fri Dec 11 12:05:22 2015 +0100 20496 20497 modesetting: Consume all available udev events at once 20498 20499 We get multiple udev events for actions like docking a laptop into its 20500 station or plugging a monitor to the station. By consuming as much 20501 events as we can, we reduce the number of output re-evalutions. 20502 20503 I.e. having a Lenovo X250 in a ThinkPad Ultra Dock and plugging a 20504 monitor to the station generates 5 udev events. Or having 2 monitors 20505 attached to the station and docking the laptop generates 7 events. 20506 20507 It depends on the timing how many events can consumed at once. 20508 20509 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 20510 [hdegoede@redhat.com: Keep goto out so that we always call RRGetInfo()] 20511 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 20512 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20513 20514commit ea91db4b83312e791dcae935b8720f49a3861d72 20515Author: Qiang Yu <Qiang.Yu@amd.com> 20516Date: Thu Sep 8 21:24:58 2016 +0800 20517 20518 config: fix GPUDevice fail when AutoAddGPU off + BusID 20519 20520 This fix is for the following xorg.conf can work: 20521 20522 Section "ServerFlags" 20523 Option "AutoAddGPU" "off" 20524 EndSection 20525 20526 Section "Device" 20527 Identifier "Amd" 20528 Driver "ati" 20529 BusID "PCI:1:0:0" 20530 EndSection 20531 20532 Section "Device" 20533 Identifier "Intel" 20534 Driver "modesetting" 20535 BusID "pci:0:2:0" 20536 EndSection 20537 20538 Section "Screen" 20539 Identifier "Screen0" 20540 Device "Intel" 20541 GPUDevice "Amd" 20542 EndSection 20543 20544 Without AutoAddGPU off, modesetting DDX will also be loaded 20545 for GPUDevice. 20546 20547 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 20548 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 20549 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20550 20551commit ca8d88e50310a0d440a127c22a0a383cc149f408 20552Author: Laszlo Ersek <lersek@redhat.com> 20553Date: Wed Sep 7 15:08:06 2016 +0200 20554 20555 xfree86: recognize primary BUS_PCI device in xf86IsPrimaryPlatform() 20556 20557 The new platform bus code and the old PCI bus code overlap. Platform bus 20558 can handle any type of device, including PCI devices, whereas the PCI code 20559 can only handle PCI devices. Some drivers only support the old style 20560 PCI-probe methods, but the primary device detection code is server based, 20561 not driver based; so we might end up with a primary device which only has 20562 a PCI bus-capable driver, but was detected as primary by the platform 20563 code, or the other way around. 20564 20565 (The above paragraph was shamelessly stolen from Hans de Goede, and 20566 customized.) 20567 20568 The latter case applies to QEMU's virtio-gpu-pci device: it is detected as 20569 a BUS_PCI primary device, but we actually probe it first (with the 20570 modesetting driver) through xf86platformProbeDev(). The 20571 xf86IsPrimaryPlatform() function doesn't recognize the device as primary 20572 (it bails out as soon as it sees BUS_PCI); instead, we add the device as a 20573 secondary graphics card under "autoAddGPU". In turn, the success of this 20574 automatic probing-as-GPU prevents xf86CallDriverProbe() from proceeding to 20575 the PCI probing. 20576 20577 The result is that the server exits with no primary devices detected. 20578 20579 Commit cf66471353ac ("xfree86: use udev to provide device enumeration for 20580 kms devices (v10)") added "cross-bus" matching to xf86IsPrimaryPci(). Port 20581 that now to xf86IsPrimaryPlatform(), so that we can probe virtio-gpu-pci 20582 as a primary card in platform bus code. 20583 20584 Cc: Adam Jackson <ajax@redhat.com> 20585 Cc: Dave Airlie <airlied@redhat.com> 20586 Cc: Hans de Goede <hdegoede@redhat.com> 20587 Cc: Keith Packard <keithp@keithp.com> 20588 Cc: Marcin Juszkiewicz <mjuszkiewicz@redhat.com> 20589 Signed-off-by: Laszlo Ersek <lersek@redhat.com> 20590 Tested-By: Marcin Juszkiewicz <mjuszkiewicz@redhat.com> 20591 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 20592 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20593 20594commit 44968da376c880e096225f149f78673aa4c6e5e0 20595Author: Kyle Guinn <elyk03@gmail.com> 20596Date: Fri Sep 23 15:03:34 2016 +0300 20597 20598 xfree86: Fix null pointer dereference 20599 20600 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=93675 20601 Signed-off-by: Kyle Guinn <elyk03@gmail.com> 20602 [hdegoede@redhat.com: Simplify by adding 2 if conds together with &&] 20603 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20604 20605commit 127e0569ca8dbf774df8f83d4c59bbbbf2e25a20 20606Author: David CARLIER <devnexen@gmail.com> 20607Date: Fri Sep 23 14:17:12 2016 +0300 20608 20609 xfree86: small memory leaks fixes 20610 20611 A couple of memory leaks fixes and avoiding bit shifting on an 20612 unitialized value. 20613 20614 [hdegoede@redhat.com: Split out some non free fixes in separate patches] 20615 [hdegoede@redhat.com: Don't touch ancient (and weird) os/rpcauth.c code] 20616 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20617 20618commit d51cce7992508f366758fa13dd93be24df0947ed 20619Author: Adam Jackson <ajax@redhat.com> 20620Date: Tue Jul 5 13:07:09 2016 -0400 20621 20622 xephyr: Don't crash if the server advertises zero xv adaptors 20623 20624 Useless as an XVideo implementation with zero adaptors might be, it's 20625 apparently a thing in the wild. Catch this case and bail out of xv init 20626 if it happens. 20627 20628 Signed-off-by: Adam Jackson <ajax@redhat.com> 20629 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 20630 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20631 20632commit 445271ec00f25405fa5ed0e149dd4a9d9f6098d4 20633Author: Hans de Goede <hdegoede@redhat.com> 20634Date: Fri Jul 1 17:36:02 2016 +0200 20635 20636 glx: Always enable EXT_texture_from_pixmap for DRI swrast glx 20637 20638 Prior to commit f95645c6f700 ("glx: Don't enable EXT_texture_from_pixmap 20639 unconditionally") DRI glx would always advertise EXT_texture_from_pixmap. 20640 20641 That commit moved the setting of the extension in the extension bits from 20642 __glXInitExtensionEnableBits to its callers so that 20643 __glXInitExtensionEnableBits can be used more generally, but at the same 20644 time made the setting of EXT_texture_from_pixmap conditionally on 20645 __DRI_TEX_BUFFER being present. 20646 20647 This has result in an unintended behavior change which breaks e.g. 20648 compositors running on llvmpipe. This commit makes the DRI swrast glx code 20649 advertise EXT_texture_from_pixmap unconditionally again fixing this. 20650 20651 Fixes: f95645c6f700 ("glx: Don't enable EXT_texture_from_pixmap unconditionally") 20652 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20653 --- 20654 Changes in v2: 20655 -Only add unconditional advertising of GLX_EXT_texture_from_pixmap 20656 to glxdriswrast.c, do not also add it to glxdri2.c 20657 20658commit 02ff0a5d7e32ce8460d6d0669f532d65ad212fcd 20659Author: Hans de Goede <hdegoede@redhat.com> 20660Date: Sat Sep 17 11:33:13 2016 +0200 20661 20662 xf86RandR12: Fix XF86VidModeSetGamma triggering a BadImplementation error 20663 20664 Commit b4e46c0444bb ("xfree86: Hook up colormaps and RandR 1.2 gamma code") 20665 dropped the providing of a pScrn->ChangeGamma callback from the xf86RandR12 20666 code. Leaving pScrn->ChangeGamma NULL in most cases. 20667 20668 This triggers the BadImplementation error in xf86ChangeGamma() : 20669 20670 if (pScrn->ChangeGamma) 20671 return (*pScrn->ChangeGamma) (pScrn, gamma); 20672 20673 return BadImplementation; 20674 20675 Which causes X-apps using XF86VidModeSetGamma to crash with a 20676 X protocol error. 20677 20678 This commit fixes this by re-introducing the xf86RandR12ChangeGamma 20679 helper removed by the commit and adjusting it to work with the new 20680 combined palette / gamma code. 20681 20682 Fixes: b4e46c0444bb ("xfree86: Hook up colormaps and RandR 1.2 gamma code") 20683 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20684 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 20685 20686commit a58dd678bf952560e5422845e186d80a189953fe 20687Author: Hans de Goede <hdegoede@redhat.com> 20688Date: Sat Sep 17 11:42:09 2016 +0200 20689 20690 xf86RandR12: Move calculating of shift inside init_one_component 20691 20692 This is a preparation patch to allow easier usage of init_one_component 20693 outside of xf86RandR12CrtcInitGamma. 20694 20695 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20696 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 20697 20698commit 299cbb92670fdb83c24c3ef6327eb9d66b8be92a 20699Author: Hans de Goede <hdegoede@redhat.com> 20700Date: Thu Sep 15 14:35:52 2016 +0200 20701 20702 modesetting: Fix reverse prime update lagging on secondary GPU outputs 20703 20704 When using secondary GPU outputs the primary GPU's blockhandler 20705 will copy changes from its framebuffer to a pixmap shared with the 20706 secondary GPU. 20707 20708 In reverse prime setups the secondary GPU's blockhandler will do another 20709 copy from the shared pixmap to its own framebuffer. 20710 20711 Before this commit, if the primary GPU's blockhandler would run after 20712 the secondary GPU's blockhandler and no events were pending, then the 20713 secondary GPU's blockhandler would not run until some events came in 20714 (WaitForSomething() would block in the poll call), resulting in the 20715 secondary GPU output sometimes showing stale contents (e.g. a just closed 20716 window) for easily up to 10 seconds. 20717 20718 This commit fixes this by setting the timeout passed into the 20719 blockhandler to 0 if any shared pixmaps were updated by the primary GPU, 20720 forcing an immediate re-run of all blockhandlers. 20721 20722 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20723 Reviewed-by: Dave Airlie <airlied@redhat.com> 20724 20725commit e8695100b17b758359fc4897dbe995231ed224fc 20726Author: Hans de Goede <hdegoede@redhat.com> 20727Date: Thu Sep 15 14:28:35 2016 +0200 20728 20729 modesetting: Fix reverse prime partial update issues on secondary GPU outputs 20730 20731 When using reverse prime we do 2 copies, 1 from the primary GPU's 20732 framebuffer to a shared pixmap and 1 from the shared pixmap to the 20733 secondary GPU's framebuffer. 20734 20735 This means that on the primary GPU side the copy MUST be finished, 20736 before we start the second copy (before the secondary GPU's driver 20737 starts processing the damage on the shared pixmap). 20738 20739 This fixes secondary outputs sometimes showning (some) old fb contents, 20740 because of the 2 copies racing with each other, for an example of 20741 what this looks like see: 20742 20743 https://fedorapeople.org/~jwrdegoede/IMG_20160915_130555.jpg 20744 20745 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 20746 Reviewed-by: Dave Airlie <airlied@redhat.com> 20747 Reviewed-by: Eric Anholt <eric@anholt.net> 20748 20749commit 862a3dab287b5186a958d0131d70779468348e3e 20750Author: Olivier Fourdan <ofourdan@redhat.com> 20751Date: Thu Sep 22 09:38:50 2016 +0200 20752 20753 xwayland: Clear up x_cursor on UnrealizeCursor() 20754 20755 In Xwayland's xwl_unrealize_cursor(), the x_cursor is cleared up only 20756 when a device value is provided to the UnrealizeCursor() routine, but 20757 if the device is NULL as called from FreeCursor(), the corresponding 20758 x_cursor for the xwl_seat is left untouched. 20759 20760 This might cause a segfault when trying to access the unrealized 20761 cursor's devPrivates in xwl_seat_set_cursor(). 20762 20763 A possible occurrence of this is the client changing the cursor, the 20764 Xserver calling FreeCursor() which does UnrealizeCursor() and then 20765 the Wayland server sending a pointer enter event, which invokes 20766 xwl_seat_set_cursor() while the seat's x_cursor has just been 20767 unrealized. 20768 20769 To avoid this, walk through all the xwl_seats and clear up all x_cursor 20770 matching the cursor being unrealized. 20771 20772 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 20773 Reviewed-by: Jonas Ådahl <jadahl@gmail.com> 20774 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 20775 20776commit b79eaf1184f6514ede9dcd9baaa24a40ef724a15 20777Author: Olivier Fourdan <ofourdan@redhat.com> 20778Date: Thu Sep 15 15:59:07 2016 +0200 20779 20780 xwayland: handle EAGAIN on Wayland fd 20781 20782 wl_display_flush() can fail with EAGAIN and Xwayland would make this a 20783 fatal error. 20784 20785 When this happens, it means that Xwayland has flooded the Wayland file 20786 descriptor, either because the Wayland compositor cannot cope or more 20787 likely because of a deadlock situation where the Wayland compositor is 20788 blocking, waiting for an X reply while Xwayland tries to write data to 20789 the Wayland file descriptor. 20790 20791 The general consensus to avoid the deadlock is for the Wayland 20792 compositor to never issue blocking X11 roundtrips, but in practice 20793 blocking rountrips can occur in various places, including Xlib calls 20794 themselves so this is not always achievable without major surgery in the 20795 Wayland compositor/Window manager. 20796 20797 What this patch does is to avoid dispatching to the Wayland file 20798 descriptor until it becomes available for writing again, while at the 20799 same time continue processing X11 requests to release the deadlock. 20800 20801 This is not perfect, as there is still the possibility of another X 20802 client hammering the connection and we'll still fail writing to the 20803 Wayland connection eventually, but this improves things enough to avoid 20804 a 100% repeatable crash with vlc and gtkperf. 20805 20806 Also, it is worth considering that window managers and Wayland 20807 compositors such as mutter already have a higher priority than other 20808 regular X clients thanks to XSyncSetPriority(), mitigating the risk. 20809 20810 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1278159 20811 Bugzilla: https://bugzilla.gnome.org/show_bug.cgi?id=763400 20812 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 20813 Reviewed-by: Daniel Stone <daniels@collabora.com> 20814 20815commit 36e1a058c5826398ceea9dba6c166ae40c75646e 20816Author: Rui Matos <tiagomatos@gmail.com> 20817Date: Fri Feb 5 14:44:29 2016 +0100 20818 20819 xwayland: Close the shm fd as early as possible 20820 20821 Keeping the shm fd open beyond pixmap creation means we can easily 20822 reach the open file descriptor limit if an X client asks us to create 20823 that many pixmaps. Instead, let's get the wl_buffer immediatly so that 20824 we can destroy the shm pool and close the fd before being asked to 20825 create more. 20826 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 20827 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 20828 20829commit fd4a6dc06fcf8671d616af52025e19200c79e404 20830Author: Adam Jackson <ajax@redhat.com> 20831Date: Tue Sep 27 17:15:12 2016 -0400 20832 20833 glamor: Fall back to software for CopyPlane if we need to 20834 20835 glUniform4ui is available starting in GL{,ES} 3.0. Technically it's 20836 also in EXT_gpu_shader4, but that's not worth supporting. There was also 20837 a MESA_shading_language_130 spec proposed at one point; if that ever 20838 gets finished, we can update epoxy to know about it and fix up the 20839 feature check. 20840 20841 Signed-off-by: Adam Jackson <ajax@redhat.com> 20842 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 20843 20844commit c49e820f796c27cbd2907709576a3fb672acd453 20845Author: Eric Anholt <eric@anholt.net> 20846Date: Sun Sep 25 13:30:22 2016 -0700 20847 20848 test: Run xts against Xephyr -glamor when present 20849 20850 v2: Drop x8r8g8b8 skip, now that it's fixed. 20851 20852 Reviewed-by: Adam Jackson <ajax@redhat.com> 20853 Signed-off-by: Eric Anholt <eric@anholt.net> 20854 20855commit 5392ea5525082ceeb3aad50d9168baee2abc860e 20856Author: Eric Anholt <eric@anholt.net> 20857Date: Sun Sep 25 13:30:21 2016 -0700 20858 20859 test: Switch our testing X server to being spawned with simple-xinit 20860 20861 Once I introduced a second X server being tested, I found that startx 20862 hit races in choosing a display. 20863 20864 Reviewed-by: Adam Jackson <ajax@redhat.com> 20865 Signed-off-by: Eric Anholt <eric@anholt.net> 20866 20867commit 283840c295c69aa3729f3f0aa2bf3ef83b1c220f 20868Author: Eric Anholt <eric@anholt.net> 20869Date: Sun Sep 25 13:30:20 2016 -0700 20870 20871 test: Update piglit HTML even when tests all pass 20872 20873 I was confused by the behavior I'd written before. keithp and 20874 mattst88 responded with shock that I would have made it so surprising, 20875 as well. 20876 20877 v2: Point to index.html instead of problems.html, which won't exist if 20878 we had no problems. 20879 20880 Reviewed-by: Adam Jackson <ajax@redhat.com> 20881 Signed-off-by: Eric Anholt <eric@anholt.net> 20882 20883commit 022fea767cd29d256fb99159df89cc25d339e787 20884Author: Eric Anholt <eric@anholt.net> 20885Date: Sun Sep 25 13:30:19 2016 -0700 20886 20887 test: Fix parsing of piglit results 20888 20889 The "dmesg-fail" line was matching our "fail" regex, so if you didn't 20890 have those we would ignore fails. 20891 20892 Reviewed-by: Adam Jackson <ajax@redhat.com> 20893 Signed-off-by: Eric Anholt <eric@anholt.net> 20894 20895commit 7fe5b9c02cf90b81532cb6dee3ec75c90c72c27c 20896Author: Eric Anholt <eric@anholt.net> 20897Date: Sun Sep 25 13:30:18 2016 -0700 20898 20899 test: Make the piglit-running script callable with an arbitrary server 20900 20901 v2: Check that SERVER_COMMAND is set. 20902 20903 Reviewed-by: Adam Jackson <ajax@redhat.com> 20904 Signed-off-by: Eric Anholt <eric@anholt.net> 20905 20906commit 7ac130a0664b43b6ba5324548c8f7ab8230f7199 20907Author: Eric Anholt <eric@anholt.net> 20908Date: Sun Sep 25 13:30:17 2016 -0700 20909 20910 test: Add a little xinit-like program for starting servers for testing 20911 20912 The normal xinit is racy because it doesn't use -displayfd. This 20913 implements the bare minimum for testing purposes, using -displayfd to 20914 sequence starting the client, and avoids adding yet another dependency 20915 to the server. 20916 20917 v2: Fix asprintf error checks. 20918 v3: Add error checking for fork(), clarify calloc() arg. 20919 20920 Reviewed-by: Adam Jackson <ajax@redhat.com> 20921 Signed-off-by: Eric Anholt <eric@anholt.net> 20922 20923commit 8d5e0c07b965d545c925c34be95d41c29135a63a 20924Author: Eric Anholt <eric@anholt.net> 20925Date: Sun Sep 25 13:30:16 2016 -0700 20926 20927 test: Handle srcdir != builddir in Xvfb testing 20928 20929 Reviewed-by: Adam Jackson <ajax@redhat.com> 20930 Signed-off-by: Eric Anholt <eric@anholt.net> 20931 20932commit add49792608171feef8090f7520cfa02d2daae9c 20933Author: Eric Anholt <eric@anholt.net> 20934Date: Sun Sep 25 13:30:15 2016 -0700 20935 20936 ephyr: Add a mode for skipping redisplay in glamor 20937 20938 This speeds up headless testing of Xephyr -glamor with softpipe from 20939 "a test per minute or so" to "a test every few seconds". 20940 20941 Reviewed-by: Adam Jackson <ajax@redhat.com> 20942 Signed-off-by: Eric Anholt <eric@anholt.net> 20943 20944commit 453f813bb4afd39eb5b6b3c4a822894e04b6b11b 20945Author: Eric Anholt <eric@anholt.net> 20946Date: Sat Sep 24 14:37:28 2016 +0300 20947 20948 glamor: Properly handle mask formats without alpha. 20949 20950 Even if the pixmap's storage has alpha, it may have been uploaded with 20951 garbage in the alpha channel, so we need to force the shader to set 20952 alpha to 1. This was broken way back in 20953 355334fcd99e4dce62e2be1e27290c9a74ea944f. 20954 20955 Fixes rendercheck -t composite -f x8r8g8b8. 20956 20957 Signed-off-by: Eric Anholt <eric@anholt.net> 20958 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 20959 20960commit 905c877f094eb6cfe905d64e34fdd01a72dd8103 20961Author: Eric Anholt <eric@anholt.net> 20962Date: Tue Sep 20 12:24:38 2016 +0100 20963 20964 glamor: Require that pixmap depths match for Render copies. 20965 20966 The copy optimization in d37329cba42fa8e72fe4be8a7be18e512268b5bd 20967 replicated a bug from last time we did a copy optimization: CopyArea 20968 is only defined for matching depths. This is only a problem at 15 vs 20969 16 depth right now (24 vs 32 would also have matching Render formats, 20970 but they should work) but be strict in case we store other depths 20971 differently in the future. 20972 20973 Fixes rendercheck -t blend -o src -f x4r4g4b4,x3r4g4b4 20974 20975 v2: Drop excessive src->depth == dst->depth check that snuck in. 20976 v3: Switch back to src->depth == dst->depth 20977 v4: Touch up commit message (s/bpp/depth). 20978 20979 Signed-off-by: Eric Anholt <eric@anholt.net> 20980 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 20981 20982commit b87a07567421af05622197eec91a71e4fca4f79e 20983Author: Eric Anholt <eric@anholt.net> 20984Date: Tue Sep 20 12:14:36 2016 +0100 20985 20986 glamor: Fix some awful formatting of some fallback debug code. 20987 20988 This was clearly x-indent.sh damage. 20989 20990 Signed-off-by: Eric Anholt <eric@anholt.net> 20991 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 20992 20993commit ba199cb90157cefab01183f2e2c909895df73321 20994Author: Mark Kettenis <kettenis@openbsd.org> 20995Date: Tue Nov 17 22:31:21 2015 +0100 20996 20997 glamor: Make glamor_sync_init work with --disable-xshmfence 20998 20999 Signed-off-by: Mark Kettenis <kettenis@openbsd.org> 21000 Signed-off-by: Eric Anholt <eric@anholt.net> 21001 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 21002 21003commit 8bb4b11298c285d2cd1eb28e65729933ec386829 21004Author: Matthieu Herrb <matthieu@herrb.eu> 21005Date: Fri Sep 23 16:56:06 2016 +0300 21006 21007 glamor: Make glamor_name_from_pixmap work without DRI3 21008 21009 This function is used by the modesetting driver to implement DRI2 and 21010 shouldn't fail on systems that don't support DRI3. 21011 21012 v2: Drop stale commit message wording, fix compiler warning (by anholt) 21013 21014 Signed-off-by: Eric Anholt <eric@anholt.net> 21015 Reviewed-by: Eric Anholt <eric@anholt.net> 21016 21017commit 128d40b2dd0a26a075a67854455d4208afaa01ca 21018Author: Eric Engestrom <eric@engestrom.ch> 21019Date: Sat Apr 2 19:53:18 2016 +0100 21020 21021 glamor: fix spelling mistakes 21022 21023 v2: Fix "orignal" too (review feedback by ajax, change by anholt)_ 21024 21025 Signed-off-by: Eric Engestrom <eric@engestrom.ch> 21026 Reviewed-by: Eric Anholt <eric@anholt.net> 21027 21028commit ade315386cee9a65a3885c65e96256880ac6f00a 21029Author: Keith Packard <keithp@keithp.com> 21030Date: Fri Sep 23 09:52:21 2016 +0300 21031 21032 Require xproto 7.0.31 21033 21034 xproto 7.0.30 had a typo which caused compiles to fail on anything 21035 other than clang 21036 21037 Signed-off-by: Keith Packard <keithp@keithp.com> 21038 21039commit f72ff1f7ac784f3911fe2f8d746f62f7709f0edc 21040Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21041Date: Mon Sep 19 09:34:50 2016 -0700 21042 21043 XQuartz: Silence an expected TSan warning 21044 21045 This code is safe. If the data race fails, the result is that we take the 21046 lock and recheck. 21047 21048 ================== 21049 WARNING: ThreadSanitizer: data race (pid=31401) 21050 Read of size 1 at 0x00010f5d2500 by thread T11: 21051 #0 wait_for_mieq_init darwinEvents.c:102 (X11.bin+0x00010003155a) 21052 #1 -[X11Application(Private) sendX11NSEvent:] X11Application.m:1330 (X11.bin+0x00010001d652) 21053 #2 __28-[X11Application sendEvent:]_block_invoke X11Application.m:476 (X11.bin+0x00010001887f) 21054 #3 __tsan::invoke_and_release_block(void*) <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x00000005d97b) 21055 #4 _dispatch_client_callout <null>:33 (libdispatch.dylib+0x0000000020ef) 21056 21057 Previous write of size 1 at 0x00010f5d2500 by thread T8: 21058 [failed to restore the stack] 21059 21060 Location is global 'mieqInitialized' at 0x00010f5d2500 (X11.bin+0x000100599500) 21061 21062 Thread T11 (tid=4367138, running) created by thread T-1 21063 [failed to restore the stack] 21064 21065 Thread T8 (tid=4367130, running) created by main thread at: 21066 #0 pthread_create <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x000000024490) 21067 #1 create_thread quartzStartup.c:78 (X11.bin+0x000100039d2d) 21068 #2 QuartzInitServer quartzStartup.c:95 (X11.bin+0x000100039b96) 21069 #3 X11ApplicationMain X11Application.m:1238 (X11.bin+0x00010001cd54) 21070 #4 X11ControllerMain X11Controller.m:984 (X11.bin+0x00010002a5b2) 21071 #5 server_main quartzStartup.c:136 (X11.bin+0x000100039fbb) 21072 #6 do_start_x11_server bundle-main.c:436 (X11.bin+0x000100002e25) 21073 #7 _Xstart_x11_server mach_startupServer.c:189 (X11.bin+0x000100004e09) 21074 #8 mach_startup_server mach_startupServer.c:399 (X11.bin+0x0001000056a4) 21075 #9 mach_msg_server mach_msg.c:563 (libsystem_kernel.dylib+0x000000012186) 21076 #10 start <null>:29 (libdyld.dylib+0x000000005254) 21077 21078 SUMMARY: ThreadSanitizer: data race darwinEvents.c:102 in wait_for_mieq_init 21079 ================== 21080 21081 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21082 21083commit 7d6ebf3f4e4d517bb846d15a5deb131da19a267c 21084Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21085Date: Mon Sep 19 00:21:44 2016 -0700 21086 21087 XQuartz: Adopt input_lock() and input_unlock() 21088 21089 This allows us to remove darwinEvents_lock() and darwinEvents_unlock() 21090 and remove the serverRunning hack from dix 21091 21092 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21093 21094commit 8bc4727f475321a8adcbbd859f3f8dd9eeb64aa4 21095Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21096Date: Sun Sep 18 23:17:19 2016 -0700 21097 21098 XQuartz: pbproxy shouldn't need to wait for server initialization. 21099 21100 Just block on the socket like every other client does. 21101 21102 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21103 21104commit 81493d30981d7deb38f57bca2a35487add5ea509 21105Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21106Date: Sun Sep 18 23:13:20 2016 -0700 21107 21108 XQuartz: Remove X11ApplicationFatalError 21109 21110 AppKit handles crashes on app launch with their own dialog now, so we shouldn't need to do this ourselves. 21111 21112 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21113 21114commit 9153ec8464336c280f2a6e79ffa443104bbeb98c 21115Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21116Date: Sun Sep 18 22:45:16 2016 -0700 21117 21118 XQuartz: Don't respond to SIGALRM on the AppKit thread 21119 21120 ================== 21121 WARNING: ThreadSanitizer: data race (pid=69627) 21122 Write of size 8 at 0x00010dae73f8 by main thread (mutexes: write M262): 21123 #0 SmartScheduleTimer utils.c:1245 (X11.bin+0x0001004b21f9) 21124 #1 __tsan::CallUserSignalHandler(__tsan::ThreadState*, bool, bool, bool, int, my_siginfo_t*, void*) <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x0000000279f6) 21125 #2 __CFRunLoopRun <null>:77 (CoreFoundation+0x000000087e17) 21126 #3 X11ControllerMain X11Controller.m:984 (X11.bin+0x00010002a062) 21127 #4 server_main quartzStartup.c:127 (X11.bin+0x000100039b6b) 21128 #5 do_start_x11_server bundle-main.c:436 (X11.bin+0x0001000022c5) 21129 #6 _Xstart_x11_server mach_startupServer.c:189 (X11.bin+0x0001000042a9) 21130 #7 mach_startup_server mach_startupServer.c:399 (X11.bin+0x000100004b44) 21131 #8 mach_msg_server mach_msg.c:563 (libsystem_kernel.dylib+0x000000012186) 21132 #9 start <null>:29 (libdyld.dylib+0x000000005254) 21133 21134 Previous read of size 8 at 0x00010dae73f8 by thread T7: 21135 [failed to restore the stack] 21136 21137 Location is global 'SmartScheduleTime' at 0x00010dae73f8 (X11.bin+0x0001005b03f8) 21138 21139 Mutex M262 (0x7d300000bd10) created at: 21140 #0 pthread_mutex_init <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x0000000253c3) 21141 #1 __CFRunLoopCreate <null>:77 (CoreFoundation+0x000000054e63) 21142 #2 X11ControllerMain X11Controller.m:984 (X11.bin+0x00010002a062) 21143 #3 server_main quartzStartup.c:127 (X11.bin+0x000100039b6b) 21144 #4 do_start_x11_server bundle-main.c:436 (X11.bin+0x0001000022c5) 21145 #5 _Xstart_x11_server mach_startupServer.c:189 (X11.bin+0x0001000042a9) 21146 #6 mach_startup_server mach_startupServer.c:399 (X11.bin+0x000100004b44) 21147 #7 mach_msg_server mach_msg.c:563 (libsystem_kernel.dylib+0x000000012186) 21148 #8 start <null>:29 (libdyld.dylib+0x000000005254) 21149 21150 Thread T7 (tid=4051693, running) created by main thread at: 21151 #0 pthread_create <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x000000024490) 21152 #1 create_thread quartzStartup.c:78 (X11.bin+0x0001000398dd) 21153 #2 QuartzInitServer quartzStartup.c:95 (X11.bin+0x000100039813) 21154 #3 X11ApplicationMain X11Application.m:1286 (X11.bin+0x00010001c804) 21155 #4 X11ControllerMain X11Controller.m:984 (X11.bin+0x00010002a062) 21156 #5 server_main quartzStartup.c:127 (X11.bin+0x000100039b6b) 21157 #6 do_start_x11_server bundle-main.c:436 (X11.bin+0x0001000022c5) 21158 #7 _Xstart_x11_server mach_startupServer.c:189 (X11.bin+0x0001000042a9) 21159 #8 mach_startup_server mach_startupServer.c:399 (X11.bin+0x000100004b44) 21160 #9 mach_msg_server mach_msg.c:563 (libsystem_kernel.dylib+0x000000012186) 21161 #10 start <null>:29 (libdyld.dylib+0x000000005254) 21162 21163 SUMMARY: ThreadSanitizer: data race utils.c:1245 in SmartScheduleTimer 21164 ================== 21165 ================== 21166 WARNING: ThreadSanitizer: signal handler spoils errno (pid=69627) 21167 #0 SmartScheduleTimer utils.c:1244 (X11.bin+0x0001004b21a0) 21168 #1 __CFRunLoopRun <null>:77 (CoreFoundation+0x000000087e17) 21169 #2 X11ControllerMain X11Controller.m:984 (X11.bin+0x00010002a062) 21170 #3 server_main quartzStartup.c:127 (X11.bin+0x000100039b6b) 21171 #4 do_start_x11_server bundle-main.c:436 (X11.bin+0x0001000022c5) 21172 #5 _Xstart_x11_server mach_startupServer.c:189 (X11.bin+0x0001000042a9) 21173 #6 mach_startup_server mach_startupServer.c:399 (X11.bin+0x000100004b44) 21174 #7 mach_msg_server mach_msg.c:563 (libsystem_kernel.dylib+0x000000012186) 21175 #8 start <null>:29 (libdyld.dylib+0x000000005254) 21176 21177 SUMMARY: ThreadSanitizer: signal handler spoils errno utils.c:1244 in SmartScheduleTimer 21178 ================== 21179 21180 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21181 21182commit 2740dc1937e8ab32b4930f4086c302e6602a7bad 21183Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21184Date: Mon Sep 19 01:25:13 2016 -0700 21185 21186 dix: Silence TSan warnings when checking for pending input 21187 21188 V2: Moves InputCheckPending() into dix.h 21189 21190 Bumps required version of xproto to 7.0.30 21191 21192 ================== 21193 WARNING: ThreadSanitizer: data race (pid=4943) 21194 Read of size 4 at 0x00010c4e3854 by thread T8: 21195 #0 WaitForSomething WaitFor.c:237 (X11.bin+0x00010049216c) 21196 #1 Dispatch dispatch.c:413 (X11.bin+0x000100352ed9) 21197 #2 dix_main main.c:287 (X11.bin+0x00010036e894) 21198 #3 server_thread quartzStartup.c:66 (X11.bin+0x000100039e63) 21199 21200 Previous write of size 4 at 0x00010c4e3854 by thread T12 (mutexes: write M856, write M1976): 21201 #0 mieqEnqueue mieq.c:263 (X11.bin+0x000100448d14) 21202 #1 DarwinSendDDXEvent darwinEvents.c:641 (X11.bin+0x000100033613) 21203 #2 DarwinProcessFDAdditionQueue_thread darwinEvents.c:338 (X11.bin+0x000100032039) 21204 21205 Location is global 'miEventQueue' at 0x00010c4e3850 (X11.bin+0x0001005ab854) 21206 21207 Mutex M856 (0x00010c4c8c80) created at: 21208 #0 pthread_mutex_lock <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x0000000321fe) 21209 #1 DarwinListenOnOpenFD darwinEvents.c:300 (X11.bin+0x000100031607) 21210 #2 socket_handoff bundle-main.c:288 (X11.bin+0x000100002b40) 21211 #3 __do_request_fd_handoff_socket_block_invoke bundle-main.c:379 (X11.bin+0x0001000029ba) 21212 #4 __tsan::invoke_and_release_block(void*) <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x00000005d97b) 21213 #5 _dispatch_client_callout <null>:33 (libdispatch.dylib+0x0000000020ef) 21214 21215 Mutex M1976 (0x00010c4e3d68) created at: 21216 #0 pthread_mutex_init <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x0000000253c3) 21217 #1 input_lock inputthread.c:103 (X11.bin+0x00010049fd10) 21218 #2 TimerSet WaitFor.c:343 (X11.bin+0x0001004926c2) 21219 #3 RootlessQueueRedisplay rootlessScreen.c:594 (X11.bin+0x000100065d7f) 21220 #4 RootlessInstallColormap rootlessScreen.c:514 (X11.bin+0x000100069f1a) 21221 #5 miSpriteInstallColormap misprite.c:562 (X11.bin+0x000100467095) 21222 #6 miCreateDefColormap micmap.c:270 (X11.bin+0x000100440399) 21223 #7 DarwinScreenInit darwin.c:285 (X11.bin+0x0001000303bb) 21224 #8 AddScreen dispatch.c:3908 (X11.bin+0x00010036c417) 21225 #9 InitOutput darwin.c:671 (X11.bin+0x00010002fdeb) 21226 #10 dix_main main.c:197 (X11.bin+0x00010036e228) 21227 #11 server_thread quartzStartup.c:66 (X11.bin+0x000100039e63) 21228 21229 Thread T8 (tid=4198779, running) created by main thread at: 21230 #0 pthread_create <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x000000024490) 21231 #1 create_thread quartzStartup.c:78 (X11.bin+0x000100039dad) 21232 #2 QuartzInitServer quartzStartup.c:95 (X11.bin+0x000100039c16) 21233 #3 X11ApplicationMain X11Application.m:1238 (X11.bin+0x00010001cde4) 21234 #4 X11ControllerMain X11Controller.m:984 (X11.bin+0x00010002a642) 21235 #5 server_main quartzStartup.c:136 (X11.bin+0x00010003a03b) 21236 #6 do_start_x11_server bundle-main.c:436 (X11.bin+0x000100002eb5) 21237 #7 _Xstart_x11_server mach_startupServer.c:189 (X11.bin+0x000100004e99) 21238 #8 mach_startup_server mach_startupServer.c:399 (X11.bin+0x000100005734) 21239 #9 mach_msg_server mach_msg.c:563 (libsystem_kernel.dylib+0x000000012186) 21240 #10 start <null>:29 (libdyld.dylib+0x000000005254) 21241 21242 Thread T12 (tid=4198797, running) created by thread T8 at: 21243 #0 pthread_create <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x000000024490) 21244 #1 create_thread darwinEvents.c:121 (X11.bin+0x000100031ecf) 21245 #2 DarwinEQInit darwinEvents.c:365 (X11.bin+0x000100031860) 21246 #3 InitInput darwin.c:571 (X11.bin+0x00010002ea09) 21247 #4 dix_main main.c:261 (X11.bin+0x00010036e7ce) 21248 #5 server_thread quartzStartup.c:66 (X11.bin+0x000100039e63) 21249 21250 SUMMARY: ThreadSanitizer: data race WaitFor.c:237 in WaitForSomething 21251 ================== 21252 ================== 21253 WARNING: ThreadSanitizer: data race (pid=22841) 21254 Write of size 4 at 0x000105bbd864 by main thread (mutexes: write M1945): 21255 #0 mieqEnqueue mieq.c:263 (X11.bin+0x000100448cf4) 21256 #1 DarwinSendDDXEvent darwinEvents.c:642 (X11.bin+0x000100033693) 21257 #2 -[X11Controller set_window_menu:] X11Controller.m:275 (X11.bin+0x0001000222fd) 21258 #3 -[X11Application set_window_menu:] X11Application.m:486 (X11.bin+0x000100018b44) 21259 #4 -[X11Application handleMachMessage:] X11Application.m:177 (X11.bin+0x000100016678) 21260 #5 __NSFireMachPort <null>:69 (Foundation+0x00000009b62b) 21261 #6 X11ControllerMain X11Controller.m:984 (X11.bin+0x00010002a5f2) 21262 #7 server_main quartzStartup.c:136 (X11.bin+0x000100039ffb) 21263 #8 do_start_x11_server bundle-main.c:436 (X11.bin+0x000100002e65) 21264 #9 _Xstart_x11_server mach_startupServer.c:189 (X11.bin+0x000100004e49) 21265 #10 mach_startup_server mach_startupServer.c:399 (X11.bin+0x0001000056e4) 21266 #11 mach_msg_server mach_msg.c:563 (libsystem_kernel.dylib+0x000000012186) 21267 #12 start <null>:29 (libdyld.dylib+0x000000005254) 21268 21269 Previous read of size 4 at 0x000105bbd864 by thread T7: 21270 #0 Dispatch dispatch.c:434 (X11.bin+0x000100352fc8) 21271 #1 dix_main main.c:287 (X11.bin+0x00010036e874) 21272 #2 server_thread quartzStartup.c:66 (X11.bin+0x000100039e23) 21273 21274 Location is global 'miEventQueue' at 0x000105bbd860 (X11.bin+0x0001005ab864) 21275 21276 Mutex M1945 (0x000105bbdd78) created at: 21277 #0 pthread_mutex_init <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x0000000253c3) 21278 #1 input_lock inputthread.c:103 (X11.bin+0x00010049fd10) 21279 #2 TimerSet WaitFor.c:348 (X11.bin+0x0001004926c2) 21280 #3 RootlessQueueRedisplay rootlessScreen.c:594 (X11.bin+0x000100065d3f) 21281 #4 RootlessInstallColormap rootlessScreen.c:514 (X11.bin+0x000100069eda) 21282 #5 miSpriteInstallColormap misprite.c:562 (X11.bin+0x000100467075) 21283 #6 miCreateDefColormap micmap.c:270 (X11.bin+0x000100440379) 21284 #7 DarwinScreenInit darwin.c:285 (X11.bin+0x00010003036b) 21285 #8 AddScreen dispatch.c:3914 (X11.bin+0x00010036c3f7) 21286 #9 InitOutput darwin.c:671 (X11.bin+0x00010002fd9b) 21287 #10 dix_main main.c:197 (X11.bin+0x00010036e208) 21288 #11 server_thread quartzStartup.c:66 (X11.bin+0x000100039e23) 21289 21290 Thread T7 (tid=4257217, running) created by main thread at: 21291 #0 pthread_create <null>:144 (libclang_rt.tsan_osx_dynamic.dylib+0x000000024490) 21292 #1 create_thread quartzStartup.c:78 (X11.bin+0x000100039d6d) 21293 #2 QuartzInitServer quartzStartup.c:95 (X11.bin+0x000100039bd6) 21294 #3 X11ApplicationMain X11Application.m:1238 (X11.bin+0x00010001cd94) 21295 #4 X11ControllerMain X11Controller.m:984 (X11.bin+0x00010002a5f2) 21296 #5 server_main quartzStartup.c:136 (X11.bin+0x000100039ffb) 21297 #6 do_start_x11_server bundle-main.c:436 (X11.bin+0x000100002e65) 21298 #7 _Xstart_x11_server mach_startupServer.c:189 (X11.bin+0x000100004e49) 21299 #8 mach_startup_server mach_startupServer.c:399 (X11.bin+0x0001000056e4) 21300 #9 mach_msg_server mach_msg.c:563 (libsystem_kernel.dylib+0x000000012186) 21301 #10 start <null>:29 (libdyld.dylib+0x000000005254) 21302 21303 SUMMARY: ThreadSanitizer: data race mieq.c:263 in mieqEnqueue 21304 ================== 21305 21306 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21307 Reviewed-by: Keith Packard <keithp@keithp.com> 21308 21309commit 7ff8a74945c0cb525a2d197e28426fd098a407cf 21310Author: Keith Packard <keithp@keithp.com> 21311Date: Thu Sep 22 02:53:46 2016 +0300 21312 21313 os: Clear saved poll events in listen so that edge triggering works 21314 21315 When a client is marked as write blocked, clear any old 'write ready' 21316 bit in the osfds structure so that a new indication of write ready 21317 (which is marked as edge trigggered) will trigger the callback. 21318 21319 Signed-off-by: Keith Packard <keithp@keithp.com> 21320 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21321 Tested-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21322 Tested-by: Matthieu Herrb <matthieu@herrb.eu> 21323 21324commit 1d365f1ffe0362732b64e565db6ae8af718758f2 21325Author: Keith Packard <keithp@keithp.com> 21326Date: Thu Sep 22 02:53:45 2016 +0300 21327 21328 os: Ready clients with pending output aren't flushed, so set NewOutputPending 21329 21330 When a client with pending output is ready (has request data pending), 21331 FlushAllOutput will skip it to get all of the requests processed 21332 before sending any queued output. That means FlushAllOutput is going 21333 to return with some output pending to a client which isn't known to be 21334 write blocked. And that means NewOutputPending needs to be set so that 21335 FlushAllOutput will get called again to actually go flush this client. 21336 21337 It might be interesting to try just flushing the client to send any 21338 queued data along the way. This patch just restores the server 21339 behavior to what it was before the ospoll changes. 21340 21341 Signed-off-by: Keith Packard <keithp@keithp.com> 21342 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21343 Tested-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21344 Tested-by: Matthieu Herrb <matthieu@herrb.eu> 21345 21346commit d0c5d205a919fc1d2eb599356090b58b1bf0176d 21347Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21348Date: Mon Sep 19 01:13:02 2016 -0700 21349 21350 dix: Make InitCoreDevices() failures more verbose. 21351 21352 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21353 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 21354 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 21355 21356commit 5794bdd52821463acf691c4230741f6b4289669b 21357Author: Mihail Konev <k.mvc@ya.ru> 21358Date: Wed Sep 14 11:47:15 2016 +0100 21359 21360 xkb: fix turbo-repeat of RedirectKey-ed keysyms 21361 21362 RedirectKey() action had been broken by commit 2e6190. 21363 A dropped check caused over-intense autorepeat of keysyms enriched 21364 with the action. 21365 21366 Previous to this commit, the check wrapped the entire switch() block, 21367 which was dropped with the move to a separate function. 21368 21369 Restore the checking. 21370 21371 Signed-off-by: Mihail Konev <k.mvc@ya.ru> 21372 Reviewed-by: Daniel Stone <daniels@collabora.com> 21373 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 21374 21375commit fc1c358b955258ce675e49490de7c3b8ddee2b76 21376Author: Keith Packard <keithp@keithp.com> 21377Date: Mon Sep 19 09:02:57 2016 -0700 21378 21379 Bump to 1.18.99.901 (1.19 RC1) 21380 21381 Signed-off-by: Keith Packard <keithp@keithp.com> 21382 21383commit 5998da7855750dd947288349a0b14e65db1e3c8c 21384Author: Jonas Ådahl <jadahl@gmail.com> 21385Date: Tue Sep 13 15:16:56 2016 +0800 21386 21387 dix: Incroduce CursorConfinedTo vfunc in Screen 21388 21389 This function will be called when a pointer is grabbed non-root window 21390 set as the 'confineTo'. This will enable the ddx to handle the 21391 confinement their own way. 21392 21393 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 21394 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 21395 Signed-off-by: Keith Packard <keithp@keithp.com> 21396 21397commit daf48a3aba7d5c42d7156f0d0e2b1d8aae423303 21398Author: Jonas Ådahl <jadahl@gmail.com> 21399Date: Tue Sep 13 15:16:55 2016 +0800 21400 21401 dix: Introduce CursorWarpedTo vfunc in Screen 21402 21403 This new vfunc will be called, if set, after a client has issued a 21404 WarpPointer request. This is necessary for implementing pointer warp 21405 emulation in Xwayland. 21406 21407 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 21408 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 21409 Signed-off-by: Keith Packard <keithp@keithp.com> 21410 21411commit 3ef51c5c74297937002668b0c507fa88550db67e 21412Author: Keith Packard <keithp@keithp.com> 21413Date: Fri Sep 16 10:03:49 2016 -0700 21414 21415 Bump version to 1.18.99.2 21416 21417 Prepare for 1.19 RC1 21418 21419 Signed-off-by: Keith Packard <keithp@keithp.com> 21420 21421commit 86038dfd42d6f6da9646ef2f87b2ab8ccba98e02 21422Author: Keith Packard <keithp@keithp.com> 21423Date: Fri Sep 16 10:56:15 2016 -0700 21424 21425 os: Add 'ospoll.h' to Makefile.am 21426 21427 Distribute this source file. 21428 21429 Signed-off-by: Keith Packard <keithp@keithp.com> 21430 21431commit 56972d007dcde93693cb9e39931978b744942e5b 21432Author: Keith Packard <keithp@keithp.com> 21433Date: Fri Sep 16 10:55:27 2016 -0700 21434 21435 include: Add 'xserver_poll.h' to Makefile.am 21436 21437 Make sure this file is included in the tarball. 21438 21439 Signed-off-by: Keith Packard <keithp@keithp.com> 21440 21441commit 61b159f5a79130a9661064d3f0143e5d6107985a 21442Author: Keith Packard <keithp@keithp.com> 21443Date: Fri Sep 16 10:54:37 2016 -0700 21444 21445 hw/xwin: Add 'dri' to DIST_SUBDIRS 21446 21447 This creates the needed Makefile.in files during 'make dist' or 'make 21448 distcheck' 21449 21450 Signed-off-by: Keith Packard <keithp@keithp.com> 21451 21452commit 93f474e49b24abc673cb9cf5ea16fac296dce312 21453Author: Keith Packard <keithp@keithp.com> 21454Date: Fri Sep 16 10:43:09 2016 -0700 21455 21456 Remove 'dixfontstubs.h' from include EXTRA_DIST 21457 21458 This file was removed when moving to the new libXfont API, but 21459 accidentally left in EXTRA_DIST. 21460 21461 Signed-off-by: Keith Packard <keithp@keithp.com> 21462 21463commit 5b9f3ea2501a886fb74e5248e82a95e76443f1e8 21464Author: Michel Dänzer <michel.daenzer@amd.com> 21465Date: Thu Jan 14 18:09:44 2016 +0900 21466 21467 xfree86/modes: Set RandR primary output from CreateScreenResources 21468 21469 Fixes XRRGetOutputPrimary and xrandr not reporting a primary output after 21470 startup. This was especially confusing when an output was explicitly 21471 marked as primary using Option "Primary" in Section "Monitor". 21472 21473 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 21474 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 21475 Signed-off-by: Keith Packard <keithp@keithp.com> 21476 21477commit bd8ecd9be1568b59258b6d2f78530ede345b7c7a 21478Author: Michel Dänzer <michel.daenzer@amd.com> 21479Date: Fri Sep 16 09:51:31 2016 -0700 21480 21481 Add SyncSharedPixmap ScreenRec hook 21482 21483 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 21484 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 21485 Signed-off-by: Keith Packard <keithp@keithp.com> 21486 21487commit 7f6fa4e449f5a7215bbf60ce8c874e163a7453a9 21488Author: Qiang Yu <Qiang.Yu@amd.com> 21489Date: Mon Sep 5 18:05:42 2016 +0800 21490 21491 modesetting: fix compile error when --disable-glamor 21492 21493 Move ms_flush_drm_events out of GLAMOR ifdef. 21494 21495 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97586 21496 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 21497 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 21498 Signed-off-by: Keith Packard <keithp@keithp.com> 21499 21500commit d81f9ce12aa4ac54b9c2b8c74c2f827c1f3e739a 21501Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21502Date: Sun Sep 11 03:07:44 2016 -0700 21503 21504 dix/dispatch: Fix SmartScheduleClient interval adjustment to use best rather than pClient 21505 21506 pClient does not contain a live value after the transition to lists 21507 21508 https://bugs.freedesktop.org/show_bug.cgi?id=97765 21509 21510 Application Specific Information: 21511 X.Org X Server 1.18.99.1 Build Date: 20160910 21512 ================================================================= 21513 ==16921==ERROR: AddressSanitizer: global-buffer-overflow on address 0x000108ce3834 at pc 0x000108880766 bp 0x7000045f76c0 sp 0x7000045f76b8 21514 READ of size 4 at 0x000108ce3834 thread T6 21515 #0 0x108880765 in SmartScheduleClient dispatch.c:365 21516 #1 0x10887ecc5 in Dispatch dispatch.c:422 21517 #2 0x1088c05f1 in dix_main main.c:301 21518 #3 0x1082aabba in server_thread quartzStartup.c:66 21519 #4 0x7fffc5f16aaa in _pthread_body (libsystem_pthread.dylib+0x3aaa) 21520 #5 0x7fffc5f169f6 in _pthread_start (libsystem_pthread.dylib+0x39f6) 21521 #6 0x7fffc5f161fc in thread_start (libsystem_pthread.dylib+0x31fc) 21522 21523 Regressed-in: 8f1edf4bd3a1f050ce9eeb5eac45dd1a8f7a6d5e 21524 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21525 Reviewed-by: Keith Packard <keithp@keithp.com> 21526 Signed-off-by: Keith Packard <keithp@keithp.com> 21527 21528commit 265c4e66533243b60d92cffc7769ecadb0292c8d 21529Merge: 3caf671ed f5f4d32ac 21530Author: Keith Packard <keithp@keithp.com> 21531Date: Thu Sep 15 21:32:46 2016 -0700 21532 21533 Merge remote-tracking branch 'jturney/master' 21534 21535commit 3caf671ed2a8a79856409a109355fc0f393d7aed 21536Author: Keith Packard <keithp@keithp.com> 21537Date: Wed Sep 14 22:40:32 2016 -0700 21538 21539 config/dbus: Initialize dbus fd to -1 so teardown doesn't use fd 0 21540 21541 The dbus teardown code is called when the server fatal errors even if 21542 that is before dbus has ever been initialized. By statically 21543 initializing the value of bus_info.fd, we avoid calling RemoveNotifyFd 21544 on stdin. 21545 21546 Signed-off-by: Keith Packard <keithp@keithp.com> 21547 Reviewed-by: Julien Cristau <jcristau@debian.org> 21548 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 21549 21550commit f5f4d32ac7c250cfbfb94883ce7d7e46151e89f0 21551Author: Jon Turney <jon.turney@dronecode.org.uk> 21552Date: Thu Jul 24 12:29:41 2014 +0100 21553 21554 Add Windows-DRI extension 21555 21556 If windowsdriproto headers are available, build a Windows-DRI extension, 21557 which supports requests to enable local clients to directly render GL to a 21558 Windows drawable: 21559 21560 - a query to check if WGL is being used on a screen 21561 - a query to map a fbconfigID to a native pixelformatindex 21562 - a query to map a drawable to a native handle 21563 21564 Windows-DRI can only be useful if we are using WGL, so make an note if WGL 21565 is active on a screen. 21566 21567 Make validGlxDrawable() public 21568 21569 Adjust glxWinSetPixelFormat() so it doesn't require a context, just a 21570 screen and config. 21571 21572 That enables factoring out the deferred drawable creation code as 21573 glxWinDeferredCreateDrawable() 21574 21575 Enhance glxWinDeferredCreateDrawable(), so that pixmaps are placed into a 21576 file mapping, so they exist in memory which can be shared with the direct 21577 rendering process. 21578 21579 Currently, this file mapping is accessed by a name generated from the XID. 21580 This will not be unique across multiple server instances. It would perhaps 21581 be better, although more complicated, to use an anonymous file mapping, and 21582 then duplicate the handle for the direct rendering process. 21583 21584 Use glxWinDeferredCreateDrawable() to ensure the native handle exists for 21585 the Windows-DRI query to map a drawable to native handle. 21586 21587 v2: 21588 Various printf format warning fixes 21589 21590 v3: 21591 Fix format warnings on x86 21592 Move some uninteresting windows-dri output to debug log level 21593 21594 v4: 21595 check for windowsdriproto when --enable-windowsdri 21596 use windowsdriproto_CFLAGS 21597 21598 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 21599 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 21600 21601commit 52d6a1e832a5e62289dd4f32824ae16a78dfd7e8 21602Author: Keith Packard <keithp@keithp.com> 21603Date: Thu Sep 8 11:02:30 2016 -0600 21604 21605 Hold input lock for deviceProc 21606 21607 This ensures that the deviceProc is never called while the input 21608 thread is processing data from the device. 21609 21610 Signed-off-by: Keith Packard <keithp@keithp.com> 21611 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 21612 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 21613 21614commit dfc91f0f630f9e0500fc28b18e3f2851024d719f 21615Author: Keith Packard <keithp@keithp.com> 21616Date: Thu Sep 8 11:02:31 2016 -0600 21617 21618 Hold input lock while calling input device SetProperty callback 21619 21620 This keeps the input driver SetProperty function from being called 21621 while input events are being processed. 21622 21623 Signed-off-by: Keith Packard <keithp@keithp.com> 21624 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 21625 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 21626 21627commit c4799f186b31e579721f5874c897f3f46db6ad0a 21628Author: Alan Coopersmith <alan.coopersmith@oracle.com> 21629Date: Sat Sep 10 21:14:19 2016 -0700 21630 21631 os: Use pthread_setname_np to set thread names if available 21632 21633 Autoconf logic borrowed from glib 21634 21635 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 21636 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21637 Tested-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21638 21639commit 75c1d04650f63464263c159d2e95364482be724f 21640Author: Alan Coopersmith <alan.coopersmith@oracle.com> 21641Date: Sat Sep 10 21:08:01 2016 -0700 21642 21643 os: OsSigHandler should not show rtld errors for unrelated signals 21644 21645 If RTLD_DI_SETSIGNAL is set to let us turn runtime linker/loader errors 21646 into catchable signals, then we should only show the errors when catching 21647 that signal, instead of tossing out red herrings to distract people with 21648 unrelated crashes long after their last failed symbol lookup (especially 21649 when using drivers built to support multiple API's by checking which 21650 symbols are available before calling them). 21651 21652 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 21653 Reviewed-by: Keith Packard <keithp@keithp.com> 21654 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 21655 21656commit 065eb6612492bacf4d7caaad90e35dafc2cbf7ea 21657Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21658Date: Sun Sep 11 19:56:52 2016 -0700 21659 21660 os/inputthread: Fix setting of cloexec on file descriptors 21661 21662 O_CLOEXEC is not a file bit. It is not setable with F_SETFL. One must 21663 use it when calling open(2). To set it cloexec on an existing fd, 21664 F_SETFD and FD_CLOEXEC must be used. 21665 21666 This also fixes a build failure regression on configurations that don't 21667 have O_CLOEXEC defined. 21668 21669 cf: http://pubs.opengroup.org/onlinepubs/9699919799/functions/fcntl.html 21670 Regressed-in: 30ac7567980a1eb79d084a63e0e74e1d9a3af673 21671 21672 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21673 Reviewed-by: Julien Cristau <jcristau@debian.org> 21674 21675commit a5769de0f5399053e9864b753fa9755220d65ae0 21676Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21677Date: Sun Sep 11 19:47:47 2016 -0700 21678 21679 Xext/shm: Fix usage of F_GETFD to match standard 21680 21681 flags = fcntl(fd, F_GETFD) is compliant. 21682 21683 fcntl(fd, F_GETFD, &flags) is non-compliant (Linux extension?) 21684 21685 cf: http://pubs.opengroup.org/onlinepubs/9699919799/functions/fcntl.html 21686 21687 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21688 Reviewed-by: Julien Cristau <jcristau@debian.org> 21689 21690commit 7def2fea30060d47780dc1eedc91fada5ae1934f 21691Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21692Date: Sat Sep 10 23:37:46 2016 -0700 21693 21694 Xquartz: Update for removal of AddEnabledDevice and RemoveEnabledDevice 21695 21696 Regressed-in: be5a513fee6cbf29ef7570e57eb0436d70fbd88c 21697 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21698 21699commit 4f4ecd0f41cf1c710d3ef1626b747847e6184f4c 21700Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21701Date: Sat Sep 10 22:16:11 2016 -0700 21702 21703 XQuartz: Cleanup CPPFLAGS that are no longer necessary on darwin 21704 21705 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21706 21707commit 33d595255d4206df0d136014de33100817cbe344 21708Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21709Date: Sun May 29 14:01:38 2016 -0700 21710 21711 XQuartz: Fix the issue where the h key could be come "stuck" after hiding XQuartz with cmd-h 21712 21713 The issue was that we set a flag to ignore the k key's up event when sent 21714 the cmd-h down event, but because the cmd-h keycode hides XQuartz, we 21715 became !_x_active by the time the event is delivered which caused us to 21716 go down a differnet codepath rather than getting a chance to ignore it. 21717 We then incorrectly ignored the next h up key. 21718 21719 https://bugs.freedesktop.org/show_bug.cgi?id=92648 21720 21721 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 21722 21723commit d8e05c04758cbcd7b5c11362cb28ce017d50098b 21724Author: Hans de Goede <hdegoede@redhat.com> 21725Date: Wed Aug 24 15:00:13 2016 +0200 21726 21727 modesetting: Fall back to primary crtc for vblank for drawables on slave outputs 21728 21729 This fixes glxgears running at 1 fps when fully covering a slave-output 21730 and the modesetting driver is used for the master gpu. 21731 21732 Reported-by: Peter Wu <peter@lekensteyn.nl> 21733 Reviewed-by: Adam Jackson <ajax@redhat.com> 21734 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21735 21736commit 7ade8ba10e1e767bb510343c86573bc5d4804b92 21737Author: Hans de Goede <hdegoede@redhat.com> 21738Date: Wed Aug 24 14:55:27 2016 +0200 21739 21740 modesetting: ms_covering_crtc: Allow calling on non modesetting Screens 21741 21742 99% of the code in ms_covering_crtc is video-driver agnostic. Add a 21743 screen_is_ms parameter when when FALSE skips the one ms specific check, 21744 this will allow calling ms_covering_crtc on slave GPUs. 21745 21746 Reviewed-by: Adam Jackson <ajax@redhat.com> 21747 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21748 21749commit 238248d67e6a422f31e8864c0b15d693a658cdac 21750Author: Hans de Goede <hdegoede@redhat.com> 21751Date: Mon Aug 15 12:02:54 2016 +0200 21752 21753 modesetting: Implement DRI2InfoRec version 9 callbacks 21754 21755 Implement the CreateBuffer2 / DestroyBuffer2 / CopyRegion2 DRI2InfoRec 21756 version 9 callbacks, this is necessary for being an offload source 21757 provider with DRI2. 21758 21759 Reviewed-by: Adam Jackson <ajax@redhat.com> 21760 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21761 21762commit 03a7c50202f61030830ff639fccf52091e02156c 21763Author: Hans de Goede <hdegoede@redhat.com> 21764Date: Mon Aug 15 10:44:57 2016 +0200 21765 21766 modesetting: ms_dri2_create_buffer: check screen of existing front buffers 21767 21768 If a frontbuffer drawable already has a pixmap, make sure it was created 21769 on the right screen. 21770 21771 Reviewed-by: Adam Jackson <ajax@redhat.com> 21772 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21773 21774commit 35c4e96ed1d372dd161480be8cddcd2d4549e449 21775Author: Hans de Goede <hdegoede@redhat.com> 21776Date: Mon Sep 12 12:47:59 2016 +0200 21777 21778 randr: Fix crtc_bounds when using rotation combined with reflection 21779 21780 Before this commit crtc_bounds() did not take reflection into account, 21781 when using reflection with 0 / 180 degree rotation this was not an 21782 issue because of the default in the switch-case doing the right thing. 21783 21784 But when using 90 / 270 degree rotation we would also end up in the 21785 default which is wrong in this case. This would lead to the cursor 21786 being constrained to a height x height area of the monitor. 21787 21788 This commit masks out the reflection bits for the switch-case, 21789 making crtc_bounds return the correct bounds and fixing the 21790 problematic cursor constraining. 21791 21792 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21793 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 21794 21795commit 7b634067c13045671685a9f00bfbac626ed68f94 21796Author: Dave Airlie <airlied@redhat.com> 21797Date: Wed Jul 15 10:15:51 2015 +1000 21798 21799 xf86Cursor: Add hw cursor support for prime 21800 21801 Currently with PRIME if we detect a secondary GPU, 21802 we switch to using SW cursors, this isn't optimal, 21803 esp for the intel/nvidia combinations, we have 21804 no choice for the USB offload devices. 21805 21806 This patch checks on each slave screen if hw 21807 cursors are enabled, and also calls set cursor 21808 and move cursor on all screens. 21809 21810 Cc: Aaron Plattner <aplattner@nvidia.com> 21811 Signed-off-by: Dave Airlie <airlied@redhat.com> 21812 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21813 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 21814 Reviewed-by: Keith Packard <keithp@keithp.com> 21815 21816commit df88008f92f85ef96d9fe48ac509d027570424eb 21817Author: Hans de Goede <hdegoede@redhat.com> 21818Date: Tue Sep 6 11:48:31 2016 +0200 21819 21820 xf86Cursor: Deal with rotation on GPU screens using a hw-cursor 21821 21822 When a slave-output is rotated the transformation is done on the blit 21823 from master to slave GPU, so crtc->transform_in_use is not set, but we 21824 still need to adjust the mouse position for things to work. 21825 21826 This commit modifies xf86_crtc_transform_cursor_position to not rely 21827 on crtc->f_framebuffer_to_crtc, so that it can be used with GPU screens 21828 too and always calls it for crtcs with any form of rotation. 21829 21830 Note not using crtc->f_framebuffer_to_crtc means that crtc->transform 21831 will not be taken into account, that is ok, because when we've a transform 21832 active hw-cursors are not used and xf86_crtc_transform_cursor_position 21833 will never get called. 21834 21835 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21836 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 21837 21838commit f82fd47016628e8bcdcba3aab506a919fe8c49d8 21839Author: Hans de Goede <hdegoede@redhat.com> 21840Date: Tue Sep 6 11:50:50 2016 +0200 21841 21842 xf86Cursor: Fix xf86_crtc_rotate_coord using width/height wrongly 21843 21844 xf86_crtc_rotate_coord should be the exact inverse operation of 21845 xf86_crtc_rotate_coord_back, but when calculating x / y for 90 / 270 21846 degrees rotation it was using height to calculate x / width to calculate y, 21847 instead of the otherway around. 21848 21849 This was likely not noticed before since xf86_crtc_rotate_coord 21850 until now was only used with cursor_info->MaxWidth and 21851 cursor_info->MaxHeight, which are usally the same. 21852 21853 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21854 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 21855 21856commit b0b04cb266a62675dd7cde97111ebe7c1552db9a 21857Author: Hans de Goede <hdegoede@redhat.com> 21858Date: Tue Sep 6 13:14:35 2016 +0200 21859 21860 xf86Cursor: Fix xf86CurrentCursor to work on slave GPU Screens 21861 21862 The CurrentCursor is always attached to the master GPU. 21863 21864 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21865 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 21866 21867commit 71fecc84e9ceb11ff61c912bdaa3fc959ec36bef 21868Author: Dave Airlie <airlied@redhat.com> 21869Date: Tue Sep 6 13:02:32 2016 +0200 21870 21871 xf86Cursor: Add xf86CheckHWCursor() helper function 21872 21873 This is a preparation patch for adding prime hw-cursor support. 21874 21875 Signed-off-by: Dave Airlie <airlied@redhat.com> 21876 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21877 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 21878 21879commit 2eefb53f58854ef9d34859583207ec37d3c3047a 21880Author: Hans de Goede <hdegoede@redhat.com> 21881Date: Tue Sep 6 13:15:36 2016 +0200 21882 21883 randr: Add RRHasScanoutPixmap helper function 21884 21885 This is a preparation patch for adding prime hw-cursor support. 21886 21887 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21888 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 21889 21890commit a52530a655438f03919d47f6edd11287efff47bb 21891Author: Dave Airlie <airlied@redhat.com> 21892Date: Tue Sep 6 12:50:14 2016 +0200 21893 21894 dix: Add dixPrivatesCreated helper function 21895 21896 This is a preparation patch for adding prime hw-cursor support. 21897 21898 Signed-off-by: Dave Airlie <airlied@redhat.com> 21899 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21900 Reviewed-by: Keith Packard <keithp@keithp.com> 21901 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 21902 21903commit 3fe4107643ba029dd48e3d12ec9bc97d07112300 21904Author: Hans de Goede <hdegoede@redhat.com> 21905Date: Thu Sep 8 09:30:31 2016 +0200 21906 21907 glamor: Fix crash when master gpu is using glamor and another gpu is hotplugged 21908 21909 When a GPU gets hotplugged while X is already running, glamor_egl_init() 21910 gets called and changes the current egl context, without updating 21911 lastGLContext, potentially causing the next glamor call on another GPU to 21912 run in the wrong context. 21913 21914 This causes glamor to e.g. crash in the next glamor_create_pixmap() call 21915 (called through the master's screen->CreatePixmap), note this is not the 21916 only troublesome entry point I've seen other backtraces when using a 21917 compositing window manager. 21918 21919 Set lastGLContext to NULL to force the next glamor_make_current() call 21920 to set the right context. 21921 21922 Note that we cannot use glamor_make_current() here to replace the 21923 eglMakeCurrent() call and update lastGLContext for us because 21924 glamor_make_current takes a glamor_priv struct as argument and that 21925 has not been created yet when glamor_egl_init() gets called. 21926 21927 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21928 Reviewed-by: Keith Packard <keithp@keithp.com> 21929 21930commit 1075af8a6c26009c04db30a6d6d1f10070568ab1 21931Author: Hans de Goede <hdegoede@redhat.com> 21932Date: Mon Aug 15 11:10:15 2016 +0200 21933 21934 modesetting: Remove some dead code 21935 21936 The "if (pixmap) ..." block this commit removes is inside an 21937 "if (pixmap == NULL) ..." block, so it will never execute. 21938 21939 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21940 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 21941 21942commit dfa295b29c20b174f80ab823eef41e5211a6a921 21943Author: Hans de Goede <hdegoede@redhat.com> 21944Date: Wed Aug 24 14:13:19 2016 +0200 21945 21946 modesetting: ms_covering_crtc: Remove unused arguments, make static 21947 21948 Remove unused arguments from ms_covering_crtc, make it static as it is 21949 only used in vblank.c. 21950 21951 While at it also change its first argument from a ScrnInfoPtr to a 21952 ScreenPtr, this makes the next patch in this patch-set cleaner. 21953 21954 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21955 Reviewed-by: Keith Packard <keithp@keithp.com> 21956 21957commit cb7b145a25452de8b549e8c8e9ec3bcc752e55dc 21958Author: Hans de Goede <hdegoede@redhat.com> 21959Date: Wed Aug 17 12:03:41 2016 +0200 21960 21961 modesetting: Fix msSharePixmapBacking returning a non-linear bo 21962 21963 glamor_fd_from_pixmap() may return a tiled bo, which is not suitable 21964 for sharing with another GPU as tiling usually is GPU specific. 21965 21966 Switch to glamor_shareable_fd_from_pixmap(), which always returns a 21967 linear bo. This fixes mis-rendering when running the mode setting 21968 driver on the master gpu in a dual-gpu setup and running an opengl 21969 app with DRI_PRIME=1. 21970 21971 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21972 Reviewed-by: Keith Packard <keithp@keithp.com> 21973 21974commit a74d553cb97d545148bd2f81b7bd021cca94e076 21975Author: Hans de Goede <hdegoede@redhat.com> 21976Date: Tue Aug 23 12:18:56 2016 +0200 21977 21978 glamor: Add glamor_shareable_fd_from_pixmap() 21979 21980 Add glamor_shareable_fd_from_pixmap function to get dma-buf fds suitable 21981 for sharing across GPUs (not using GPU specific tiling). 21982 21983 This is necessary for the modesetting driver to correctly implement 21984 the DRI2 SharePixmapBacking callback. 21985 21986 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 21987 Reviewed-by: Keith Packard <keithp@keithp.com> 21988 21989commit 527c6baa294d17c5eca1d87ac941844872e90dac 21990Author: Ran Benita <ran234@gmail.com> 21991Date: Wed Aug 29 12:33:34 2012 +0300 21992 21993 xkb: fix check for appending '|' character when applying rules 21994 21995 There are two ways to separate multiple files in XKB include statements: 21996 '+' will cause the later file to override the first in case of conflict, 21997 while '|' will cause it augment it (this is done by xkbcomp). '!' is 21998 unrelated here. 21999 Currently, if someone tries to use '|' in a rule instead of '+', it 22000 won't have any effect. Since '|' is practically never used, this wasn't 22001 noticed. 22002 22003 Signed-off-by: Ran Benita <ran234@gmail.com> 22004 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 22005 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 22006 22007commit 392da389d7b0e9dd970741dcd5321a4e0fd3aef5 22008Author: Adam Jackson <ajax@redhat.com> 22009Date: Wed Mar 30 14:04:04 2016 -0400 22010 22011 glx: Fix computation of GLX_X_RENDERABLE fbconfig attribute 22012 22013 >From the GLX spec: 22014 22015 "GLX_X_RENDERABLE is a boolean indicating whether X can be used to 22016 render into a drawable created with the GLXFBConfig. This attribute 22017 is True if the GLXFBConfig supports GLX windows and/or pixmaps." 22018 22019 Every backend was setting this to true unconditionally, and then the 22020 core ignored that value and sent true unconditionally on its own. This 22021 is broken for ARB_fbconfig_float and EXT_fbconfig_packed_float, which 22022 only apply to pbuffers, which are not renderable from non-GLX APIs. 22023 22024 Instead compute GLX_X_RENDERABLE from the supported drawable types. The 22025 dri backends were getting _that_ wrong too, so fix that as well. 22026 22027 This is not a functional change, as there are no mesa drivers that claim 22028 to support __DRI_ATTRIB_{UNSIGNED_,}FLOAT_BIT yet. 22029 22030 Signed-off-by: Adam Jackson <ajax@redhat.com> 22031 Reviewed-by: Eric Anholt <eric@anholt.net> 22032 22033commit 589f42e9830e66a7e26475fc9a8b91034b5aad86 22034Author: Rui Matos <tiagomatos@gmail.com> 22035Date: Sun Jun 26 19:48:23 2016 +0200 22036 22037 xwayland: Process queued events before making wayland mods effective 22038 22039 Since xwayland's initial commit we have had a check to not process 22040 wayland modifier events while one of our surfaces has keyboard focus 22041 since the normal xkb event processing keeps our internal modifier 22042 state up to date and if we use the modifiers we get from the 22043 compositor we mess up that state. 22044 22045 This was slightly changed in commit 22046 10e9116b3f709bec6d6a50446c1341441a0564e4 to allow the xkb group to be 22047 set from the wayland event while we have focus in case the compositor 22048 triggers a group switch. 22049 22050 There's a better solution to the original problem though. Processing 22051 queued events before overriding the xkb state with the compositor's 22052 allows those events to be sent properly modified to X clients while 22053 any further events will be modified with the wayland modifiers as 22054 intended. 22055 22056 This allows us to fully take in the wayland modifiers, including 22057 depressed ones, which fixes an issue where we wouldn't be aware of 22058 already pressed modifiers on enter. 22059 22060 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 22061 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 22062 Reviewed-by: Daniel Stone <daniels@collabora.com> 22063 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 22064 22065commit deae9c7e846e244e5d62b2dcfb6663fde0e12cbb 22066Author: Eric Anholt <eric@anholt.net> 22067Date: Wed Aug 31 17:00:00 2016 -0700 22068 22069 test: Run some XTS5 integration tests against Xvfb if possible. 22070 22071 By default the tests will be skipped. However, if you set XTEST_DIR 22072 to the repo of a built X Test Suite and PIGLIT_DIR to a piglit repo 22073 (no build necessary), make check will run piglit's xts-render tests 22074 against Xvfb. 22075 22076 We could run more of XTS5, but I haven't spent the time identifying 22077 what additional subset would be worth running, since much of it is 22078 only really testing the client libraries. We want to make sure that 22079 we keep the runtime down, and this subset of the test suite took 92 22080 seconds according to piglit. 22081 22082 Reviewed-by: Adam Jackson <ajax@redhat.com> 22083 22084commit f06aef31c0b2ab62d915a469a399d82a1a73ac28 22085Author: Qiang Yu <Qiang.Yu@amd.com> 22086Date: Mon Aug 22 18:14:18 2016 +0800 22087 22088 modesetting: add DRI2 page flip support 22089 22090 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 22091 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 22092 22093commit 4a839da6274b76424ee7cbc2deba2693e97e8186 22094Author: Qiang Yu <Qiang.Yu@amd.com> 22095Date: Mon Aug 22 18:14:17 2016 +0800 22096 22097 modesetting: move common page flip handle to pageflip.c 22098 22099 The common page flip handle framework can be shared with DRI2 22100 page flip. 22101 22102 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 22103 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 22104 22105commit 4f1eb7864b61d0c992c5eff3339085e887792048 22106Author: Qiang Yu <Qiang.Yu@amd.com> 22107Date: Mon Aug 22 18:14:16 2016 +0800 22108 22109 modesetting: move ms_do_pageflip to pageflip.c 22110 22111 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 22112 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 22113 22114commit a586bf9ccf60b2cc66953555e82eeb6a8e0b4ca9 22115Author: Qiang Yu <Qiang.Yu@amd.com> 22116Date: Mon Aug 22 18:14:15 2016 +0800 22117 22118 modesetting: make ms_do_pageflip generic for share with DRI2 22119 22120 Signed-off-by: Qiang Yu <Qiang.Yu@amd.com> 22121 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 22122 22123commit 7d33ab0f8c7958b205076f71e4b47c24aace77fd 22124Author: Adam Jackson <ajax@redhat.com> 22125Date: Tue Jun 28 15:54:44 2016 -0400 22126 22127 dri2: Don't make reference to noClientException 22128 22129 noClientException is now never filled in with a meaningful value, it's 22130 always -1. The sole caller of this function disregards the error value 22131 in any case. 22132 22133 Reviewed-by: Eric Anholt <eric@anholt.net> 22134 Signed-off-by: Adam Jackson <ajax@redhat.com> 22135 22136commit dff435568b40286df2bf7a66e870f9ae5f4eb9bd 22137Author: Adam Jackson <ajax@redhat.com> 22138Date: Tue Jun 28 15:54:43 2016 -0400 22139 22140 dri3: Don't do return client->noClientException 22141 22142 Hasn't been necessary since: 22143 22144 commit 92ed75ac59e2d3af149cddb962efd05fc8487750 22145 Author: Jamey Sharp <jamey@minilop.net> 22146 Date: Mon May 10 20:22:05 2010 -0700 22147 22148 Eliminate boilerplate around client->noClientException. 22149 22150 Reviewed-by: Eric Anholt <eric@anholt.net> 22151 Signed-off-by: Adam Jackson <ajax@redhat.com> 22152 22153commit 65493c0b73f6ee65b2a8fcaa4a15b0c453ff72b1 22154Author: Adam Jackson <ajax@redhat.com> 22155Date: Tue Jun 28 15:54:42 2016 -0400 22156 22157 sync: Don't do return client->noClientException 22158 22159 Hasn't been necessary since: 22160 22161 commit 92ed75ac59e2d3af149cddb962efd05fc8487750 22162 Author: Jamey Sharp <jamey@minilop.net> 22163 Date: Mon May 10 20:22:05 2010 -0700 22164 22165 Eliminate boilerplate around client->noClientException. 22166 22167 Signed-off-by: Adam Jackson <ajax@redhat.com> 22168 Reviewed-by: Eric Anholt <eric@anholt.net> 22169 22170commit bc3eed379ee5b4176db848708799fb2d5db16531 22171Author: Hans De Goede <hdegoede@redhat.com> 22172Date: Wed Aug 31 15:37:20 2016 +0200 22173 22174 modesetting: Hide cursor when initializing crtc 22175 22176 When Xorg gets started directly from a wayland-gdm the crtc still has the 22177 wayland hw cursor set. Combine this with Xorg immediately falling back to 22178 a sw cursor because a slave-output has a monitor attached at startup; and 22179 we end up with the wayland hardware cursor overlay fixed in its last 22180 position + the Xorg sw cursor resulting in 2 cursors. 22181 22182 This commit fixes this by hiding any left-over cursors when initializing 22183 the crtc. 22184 22185 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 22186 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 22187 22188commit 6c984ac9a737304ab5e26de90c76a79d970b6e33 22189Author: Hans De Goede <hdegoede@redhat.com> 22190Date: Wed Aug 31 16:11:11 2016 +0200 22191 22192 modesetting: Do not use function local static variables 22193 22194 The modesetting driver may be driving 2 screens (slave and master 22195 gpu), which may have different behavior wrt hardware cursor support. 22196 22197 So stop using static variables and instead store the hw-cursor support 22198 related data in a per screen struct. While at it actually make it per 22199 crtc data as in theory different crtc's could have different hw-cursor 22200 support. 22201 22202 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 22203 Reviewed-by: Keith Packard <keithp@keithp.com> 22204 22205commit d8c288ec371a853bcd023217f34893a0efae65ea 22206Author: Michael Thayer <michael.thayer@oracle.com> 22207Date: Wed Aug 24 08:48:13 2016 +1000 22208 22209 xi2: fix FocusIn grabs 22210 22211 Fix a couple of copy-and-paste errors preventing FocusIn grabs from working. 22212 Perhaps the extension version should be bumped though to distinguish between 22213 working and non-working extension versions. 22214 22215 Signed-off-by: Michael Thayer <michael.thayer@oracle.com> 22216 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 22217 22218commit 25e4f9ee68b99c2810efdb6cd8c56affa45e1fea 22219Author: Peter Hutterer <peter.hutterer@who-t.net> 22220Date: Tue Aug 23 13:43:42 2016 +1000 22221 22222 xfree86: print the module name together with the load failure message 22223 22224 We're happily printing the error to stdout but not which module caused it... 22225 That's in the Xorg.log but that's at least one click away. 22226 22227 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 22228 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 22229 22230commit be334f42a198a25e817e6dab43dd0e30aa1cd4f8 22231Author: Michel Dänzer <michel.daenzer@amd.com> 22232Date: Thu Aug 18 09:40:34 2016 +0900 22233 22234 glamor: Declare "pos" in the composite glyph GLSL 1.20 vertex shader 22235 22236 Fixes shader compile failure: 22237 22238 Failed to compile VS: 0:13(43): error: `pos' undeclared 22239 0:13(14): error: operands to arithmetic operators must be numeric 22240 0:13(13): error: operands to arithmetic operators must be numeric 22241 22242 Program source: 22243 #define ATLAS_DIM_INV 0.000976562500000000 22244 attribute vec2 primitive; 22245 attribute vec2 source; 22246 varying vec2 glyph_pos; 22247 uniform vec2 fill_offset; 22248 uniform vec2 fill_size_inv; 22249 varying vec2 fill_pos; 22250 uniform vec4 v_matrix; 22251 void main() { 22252 gl_Position.xy = primitive.xy * v_matrix.xz + v_matrix.yw; 22253 gl_Position.zw = vec2(0.0,1.0); 22254 glyph_pos = source.xy * ATLAS_DIM_INV; 22255 fill_pos = (fill_offset + primitive.xy + pos) * fill_size_inv; 22256 } 22257 (EE) Fatal server error: 22258 (EE) GLSL compile failure 22259 22260 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97300 22261 Reviewed-by: Keith Packard <keithp@keithp.com> 22262 22263commit cba28d572ac799391beacd89d57e69d0d7ed70e7 22264Author: Michel Dänzer <michel.daenzer@amd.com> 22265Date: Tue Jul 12 12:25:13 2016 +0900 22266 22267 glamor: Handle bitplane in glamor_copy_fbo_cpu 22268 22269 This can significantly speed up at least some CopyPlane cases, e.g. 22270 indirectly for stippled fills. 22271 22272 v2: 22273 * Make temporary pixmap the same size as the destination pixmap 22274 (instead of the destination drawable size), and fix coordinate 22275 parameters passed to fbCopyXtoX and glamor_upload_boxes. Fixes 22276 incorrect rendering with x11perf -copyplane* and crashes with the 22277 xscreensaver phosphor hack. 22278 v3: 22279 * Make the change a bit more compact and hopefully more readable by 22280 re-using the existing src_* locals in the bitplane case as well. 22281 22282 Reported-by: Keith Raghubar <keith.raghubar@amd.com> 22283 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 22284 Acked-by: Eric Anholt <eric@anholt.net> 22285 22286commit 6e5bec261c3f7af069b57618d6c82b070dc4579d 22287Author: Olivier Fourdan <ofourdan@redhat.com> 22288Date: Thu Jun 23 15:31:30 2016 +0200 22289 22290 wayland: Emulate crossing for native window 22291 22292 Emitting a LeaveNotify event every time the pointer leaves an X11 window 22293 may confuse focus follow mouse mode in window managers such as 22294 mutter/gnome-shell. 22295 22296 Keep the previously found X window and compare against the new one, and 22297 if they match then it means the pointer has left an Xwayland window for 22298 a native Wayland surface, only in this case fake the crossing to the 22299 root window. 22300 22301 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 22302 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 22303 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 22304 22305commit 48c5c23a1b250c7f9d7a1747c76e4669ebf752cf 22306Author: Hans De Goede <hdegoede@redhat.com> 22307Date: Mon Aug 8 14:53:59 2016 +0200 22308 22309 Fix Xorg -configure not working anymore 22310 22311 Xorg -configure relies on the bus implementation, e.g. 22312 xf86pciBus.c to call xf86AddBusDeviceToConfigure(). The new 22313 xf86platformBus code does not have support for this. 22314 22315 Almost all drivers support both the xf86platformBus and xf86pciBus 22316 nowadays, and the generic xf86Bus xf86CallDriverProbe() function 22317 prefers the new xf86platformBus probe method when available. 22318 22319 Since the platformBus paths do not call xf86AddBusDeviceToConfigure() 22320 this results in Xorg -configure failing with the following error: 22321 "No devices to configure. Configuration failed.". 22322 22323 Adding support for the xf86Configure code to xf86platformBus.c 22324 is non trivial and since we advise users to normally run without 22325 any Xorg.conf at all not worth the trouble. 22326 22327 However some users still want to use Xorg -configure to generate a 22328 template config file, this commit implements a minimal fix to make 22329 things work again for PCI devices by skipping the platform 22330 probe method when xf86DoConfigure is set. 22331 22332 This has been tested on a system with integrated intel graphics, 22333 with both the intel and modesetting drivers and restores Xorg -configure 22334 functionality on both cases. 22335 22336 Reviewed-by: Adam Jackson <ajax@redhat.com> 22337 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 22338 22339commit 6acd0d098add5a14f23d326ac6ee8cebab14b86a 22340Author: Adam Jackson <ajax@redhat.com> 22341Date: Wed Aug 17 10:32:28 2016 -0400 22342 22343 test: Remember to swap the window argument to XIQueryPointer 22344 22345 Before 5c69cb60 this wouldn't matter, because ProcXIQueryPointer 22346 manually emitted its own error before (bogusly) returning Success to the 22347 main loop. Since these tests only look at the return value of the 22348 dispatch function we'd think things succeeded even when we'd generated 22349 an error. 22350 22351 With that fixed, the test code's failure to swap the window id would 22352 make dixLookupWindow (rightly) throw BadWindow. 22353 22354 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 22355 Signed-off-by: Adam Jackson <ajax@redhat.com> 22356 22357commit 5c69cb604a144bdced7d066eaca574aa44e8cb36 22358Author: Adam Jackson <ajax@redhat.com> 22359Date: Mon Jun 20 12:57:12 2016 -0400 22360 22361 xinput: Let top-level dispatch generate the error 22362 22363 ... instead of calling SendErrorToClient ourselves. 22364 22365 Signed-off-by: Adam Jackson <ajax@redhat.com> 22366 Reviewed-by: Daniel Stone <daniels@collabora.com> 22367 22368commit 2f981c06a5b9b9eab5e8a9b1d0a18281c0d1033f 22369Author: Adam Jackson <ajax@redhat.com> 22370Date: Mon Jun 20 12:57:11 2016 -0400 22371 22372 xv: Remove some dumb calls SendErrorToClient 22373 22374 We already generate errors from the top level when non-Success is 22375 returned from a dispatch function, so really we were emitting errors 22376 twice. 22377 22378 Signed-off-by: Adam Jackson <ajax@redhat.com> 22379 Reviewed-by: Daniel Stone <daniels@collabora.com> 22380 22381commit 0bfa6bf9def43fd14bc79a302b912454e82999b1 22382Author: Adam Jackson <ajax@redhat.com> 22383Date: Mon Jun 20 12:57:10 2016 -0400 22384 22385 xwayland: Fix relinking when dix changes 22386 22387 Without this a change in eg Xext/ wouldn't relink Xwayland, making you 22388 wonder why your changes didn't have any effect. 22389 22390 Signed-off-by: Adam Jackson <ajax@redhat.com> 22391 Reviewed-by: Daniel Stone <daniels@collabora.com> 22392 22393commit bf2ca502b9cd07044cba56499d5c092deb9d73f5 22394Author: Keith Packard <keithp@keithp.com> 22395Date: Mon Aug 15 22:25:56 2016 -0700 22396 22397 config: Hold input lock across udev socket handling 22398 22399 This avoids having the server running inside libudev from two threads 22400 at once, which it appears to not like. 22401 22402 Signed-off-by: Keith Packard <keithp@keithp.com> 22403 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 22404 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 22405 22406commit de5291c04b05772e6da599a475baa1b19dcae07a 22407Author: Olivier Fourdan <ofourdan@redhat.com> 22408Date: Mon Aug 8 17:25:35 2016 +0200 22409 22410 present: Free the fake_present OsTimerPtr 22411 22412 Plug a leak in present_fake_queue_vblank() where the OsTimer would not 22413 be freed. 22414 22415 492,608 (482,816 direct, 9,792 indirect) bytes in 15,088 blocks 22416 are definitely lost in loss record 3,954 of 3,954 22417 at 0x4C2ABDE: malloc (in vgpreload_memcheck-amd64-linux.so) 22418 by 0x586B19: TimerSet (WaitFor.c:433) 22419 by 0x4F1AA9: present_fake_queue_vblank (present_fake.c:108) 22420 by 0x4F15E0: present_pixmap (present.c:954) 22421 by 0x4F23B4: proc_present_pixmap (present_request.c:138) 22422 by 0x552BCE: Dispatch (dispatch.c:430) 22423 by 0x556C22: dix_main (main.c:300) 22424 by 0x6F0D290: (below main) (in /usr/lib/libc-2.24.so) 22425 22426 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97065 22427 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 22428 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 22429 22430commit d21b53673b9de052004e761fd1299d2e788b44b8 22431Author: Michel Dänzer <michel.daenzer@amd.com> 22432Date: Wed Aug 3 10:19:20 2016 +0900 22433 22434 present: Call set_abort_flip / restore_screen_pixmap in clear_window_flip 22435 22436 We were asserting that these were called before from other places, but 22437 that isn't always the case, e.g. during server shutdown. 22438 22439 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96951 22440 Reported-and-Tested-by: Tod Jackson <tod.jackson@gmail.com> 22441 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 22442 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 22443 22444commit 9f04fa3babaaabe0ed5bf4aa2563755bc6c45723 22445Author: Michel Dänzer <michel.daenzer@amd.com> 22446Date: Wed Aug 3 10:19:19 2016 +0900 22447 22448 present: Make present_restore_screen_pixmap handle screen->root == NULL 22449 22450 Easier than dealing with it in all paths that can end up here during 22451 server shutdown. 22452 22453 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 22454 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 22455 22456commit 4cbf1fb1f978ecd975770cebbb330dc10f712b77 22457Author: Olivier Fourdan <ofourdan@redhat.com> 22458Date: Mon Aug 8 17:57:57 2016 +0200 22459 22460 xwayland: Avoid double free of RRCrtc and RROutput 22461 22462 At shutdown, the Xserver will free all its resources which includes the 22463 RRCrtc and RROutput created. 22464 22465 Xwayland would do the same in its xwl_output_destroy() called from 22466 xwl_close_screen(), leading to a double free of existing RRCrtc 22467 RROutput: 22468 22469 Invalid read of size 4 22470 at 0x4CDA10: RRCrtcDestroy (rrcrtc.c:689) 22471 by 0x426E75: xwl_output_destroy (xwayland-output.c:301) 22472 by 0x424144: xwl_close_screen (xwayland.c:117) 22473 by 0x460E17: CursorCloseScreen (cursor.c:187) 22474 by 0x4EB5A3: AnimCurCloseScreen (animcur.c:106) 22475 by 0x4EF431: present_close_screen (present_screen.c:64) 22476 by 0x556D40: dix_main (main.c:354) 22477 by 0x6F0D290: (below main) (in /usr/lib/libc-2.24.so) 22478 Address 0xbb1fc30 is 0 bytes inside a block of size 728 free'd 22479 at 0x4C2BDB0: free (in 22480 /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 22481 by 0x4CCE5F: RRCrtcDestroyResource (rrcrtc.c:719) 22482 by 0x577541: doFreeResource (resource.c:895) 22483 by 0x5787B5: FreeClientResources (resource.c:1161) 22484 by 0x578862: FreeAllResources (resource.c:1176) 22485 by 0x556C54: dix_main (main.c:323) 22486 by 0x6F0D290: (below main) (in /usr/lib/libc-2.24.so) 22487 Block was alloc'd at 22488 at 0x4C2CA6A: calloc (in 22489 /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 22490 by 0x4CC6DB: RRCrtcCreate (rrcrtc.c:76) 22491 by 0x426D1C: xwl_output_create (xwayland-output.c:264) 22492 by 0x4232EC: registry_global (xwayland.c:431) 22493 by 0x76CB1C7: ffi_call_unix64 (in /usr/lib/libffi.so.6.0.4) 22494 by 0x76CAC29: ffi_call (in /usr/lib/libffi.so.6.0.4) 22495 by 0x556CEFD: wl_closure_invoke (connection.c:935) 22496 by 0x5569CBF: dispatch_event.isra.4 (wayland-client.c:1310) 22497 by 0x556AF13: dispatch_queue (wayland-client.c:1456) 22498 by 0x556AF13: wl_display_dispatch_queue_pending 22499 (wayland-client.c:1698) 22500 by 0x556B33A: wl_display_roundtrip_queue (wayland-client.c:1121) 22501 by 0x42371C: xwl_screen_init (xwayland.c:631) 22502 by 0x552F60: AddScreen (dispatch.c:3864) 22503 22504 And: 22505 22506 Invalid read of size 4 22507 at 0x522890: RROutputDestroy (rroutput.c:348) 22508 by 0x42684E: xwl_output_destroy (xwayland-output.c:302) 22509 by 0x423CF4: xwl_close_screen (xwayland.c:118) 22510 by 0x4B6377: CursorCloseScreen (cursor.c:187) 22511 by 0x539503: AnimCurCloseScreen (animcur.c:106) 22512 by 0x53D081: present_close_screen (present_screen.c:64) 22513 by 0x43DBF0: dix_main (main.c:354) 22514 by 0x7068730: (below main) (libc-start.c:289) 22515 Address 0xc403190 is 0 bytes inside a block of size 154 free'd 22516 at 0x4C2CD5A: free (vg_replace_malloc.c:530) 22517 by 0x521DF3: RROutputDestroyResource (rroutput.c:389) 22518 by 0x45DA61: doFreeResource (resource.c:895) 22519 by 0x45ECFD: FreeClientResources (resource.c:1161) 22520 by 0x45EDC2: FreeAllResources (resource.c:1176) 22521 by 0x43DB04: dix_main (main.c:323) 22522 by 0x7068730: (below main) (libc-start.c:289) 22523 Block was alloc'd at 22524 at 0x4C2BBAD: malloc (vg_replace_malloc.c:299) 22525 by 0x52206B: RROutputCreate (rroutput.c:84) 22526 by 0x426763: xwl_output_create (xwayland-output.c:270) 22527 by 0x422EDC: registry_global (xwayland.c:432) 22528 by 0x740FC57: ffi_call_unix64 (unix64.S:76) 22529 by 0x740F6B9: ffi_call (ffi64.c:525) 22530 by 0x5495A9D: wl_closure_invoke (connection.c:949) 22531 by 0x549283F: dispatch_event.isra.4 (wayland-client.c:1274) 22532 by 0x5493A13: dispatch_queue (wayland-client.c:1420) 22533 by 0x5493A13: wl_display_dispatch_queue_pending 22534 (wayland-client.c:1662) 22535 by 0x5493D2E: wl_display_roundtrip_queue (wayland-client.c:1085) 22536 by 0x4232EC: xwl_screen_init (xwayland.c:632) 22537 by 0x439F50: AddScreen (dispatch.c:3864) 22538 22539 Split xwl_output_destroy() into xwl_output_destroy() which frees the 22540 wl_output and the xwl_output structure, and xwl_output_remove() which 22541 does the RRCrtcDestroy() and RROutputDestroy() and call the latter only 22542 when an output is effectively removed. 22543 22544 An additional benefit, on top of avoiding a double free, is to avoid 22545 updating the screen size at shutdown. 22546 22547 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 22548 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 22549 22550commit b380f3ac51f40ffefcde7d3db5c4c149f274246d 22551Author: Michel Dänzer <michel.daenzer@amd.com> 22552Date: Tue Aug 2 17:53:01 2016 +0900 22553 22554 dix: Pass ClientPtr to FlushCallback 22555 22556 This change has two effects: 22557 22558 1. Only calls FlushCallbacks when we're actually flushing data to a 22559 client. The unnecessary FlushCallback calls could cause significant 22560 performance degradation with compositing, which is significantly 22561 reduced even without any driver changes. 22562 22563 2. By passing the ClientPtr to FlushCallbacks, drivers can completely 22564 eliminate unnecessary flushing of GPU commands by keeping track of 22565 whether we're flushing any XDamageNotify events to the client for 22566 which the corresponding rendering commands haven't been flushed to 22567 the GPU yet. 22568 22569 Reviewed-by: Adam Jackson <ajax@redha.com> 22570 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 22571 22572commit 92b3cd32066aa5befa67a408cc079cd2ce4c077e 22573Author: Aaron Plattner <aplattner@nvidia.com> 22574Date: Tue Aug 2 10:34:07 2016 -0700 22575 22576 xace: Fix XaceCensorImage to actually censor the right part of the image 22577 22578 The caller passes arguments into XaceCensorImage that are in 22579 window-relative coordinates. However, the pBuf that it uses to construct 22580 a temporary pixmap has its origin at (x, y) relative to the window in 22581 question. The code to convert the censor region into boxes adjusts for 22582 the Y coordinate, but leaves the X coordinate alone. The result is that 22583 if x is not zero, it censors the wrong part of the image. 22584 22585 Fix this by just translating censorRegion into pixmap-relative 22586 coordinates and using the resulting boxes as-is. 22587 22588 Reported-by: Fabien Lelaquais <Fabien.Lelaquais@roguewave.com> 22589 Link: https://lists.x.org/archives/xorg/2016-August/058165.html 22590 Reviewed-by: Adam Jackson <ajax@redhat.com> 22591 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 22592 22593commit 4d586118c113f3c0a6e95ed2d3fc7f9d03a4e362 22594Author: Olivier Fourdan <ofourdan@redhat.com> 22595Date: Tue Aug 2 11:24:41 2016 +0200 22596 22597 xwayland: Plug memleak in frame callbacks 22598 22599 The frame callback set up via wl_surface_frame() needs to be freed with 22600 wl_callback_destroy() or we'll leak memory. 22601 22602 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97065 22603 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 22604 Reviewed-by: Daniel Stone <daniels@collabora.com> 22605 22606commit 611b3b1d40f58035582480a3d7cfa77cb288cbc6 22607Author: Matthieu Herrb <matthieu@herrb.eu> 22608Date: Sun Jul 31 12:13:03 2016 +0200 22609 22610 Autoconfig: Fix the loop adding dectected drivers to the list of screens 22611 22612 This loop was written in a buggy style, causing a NULL driver ptr to be 22613 passed to copyScreen(). copyScreen() only uses that to generate an 22614 identifier string, so this is mostly harmless on systems that accept 22615 NULL for asprintf() "%s" format. (the generated identifiers are off 22616 by one wrt the driver names and the last one contains NULL. 22617 22618 For systems that don't accept NULL for '%s' this would cause a 22619 segmentation fault when this code is used (no xorg.conf, but partial 22620 config in xorg.conf.d for instance). 22621 22622 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 22623 Reviewed-by: Keith Packard <keithp@keithp.com> 22624 22625commit 88820f1c7b66cbc98d3f19efca24c9f52410d9f9 22626Author: Emily Deng <Emily.Deng@amd.com> 22627Date: Mon Jul 25 16:12:53 2016 +0800 22628 22629 xfree86: vgaarb: Initialize local variable rsrc_decodes 22630 22631 In function xf86VGAarbiterScrnInit when the "pEnt->bus.type" is 22632 BUS_PLATFORM, the "pScrn->vgaDev" won't be set, so the "pScrn->vgaDev" is 22633 equal to zero. 22634 22635 The variable "rsrc_decodes" in function "xf86VGAarbiterAllowDRI" is not 22636 initialized. So it will occur error when "pScrn->vgaDev == 0", and 22637 "vga_count > 1". For this case, as "pScrn->vgaDev == 0", the function 22638 "pci_device_vgaarb_get_info" will only set the value of "vga_count", 22639 but won't set the value of "rsrc_decodes", so it will has two different 22640 return values for function "xf86VGAarbiterAllowDRI" in different 22641 platforms. One platform will return TRUE, as the "rsrc_decodes" 's 22642 default value is 0, but another platform will return FALSE, as the 22643 "rsrc_decodes" 's default value is "32767", this will cause disable 22644 direct rendering. 22645 22646 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96937 22647 22648 Signed-off-by: Emily Deng <Emily.Deng@amd.com> 22649 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 22650 22651commit fa2bc06407665ee074f5914d0f83d4114b20fc8e 22652Author: Keith Packard <keithp@keithp.com> 22653Date: Fri Aug 12 17:40:55 2016 -0700 22654 22655 dix: Avoid writing uninitialized bytes in deliverPropertyNotifyEvent 22656 22657 Make sure the whole event is initialized, instead of leaving the pad 22658 bytes unset. 22659 22660 Signed-off-by: Keith Packard <keithp@keithp.com> 22661 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 22662 22663commit 67fc5d68f967d19236221b2c0c2e0fe45acf560f 22664Author: Keith Packard <keithp@keithp.com> 22665Date: Sat Aug 13 09:02:15 2016 -0700 22666 22667 os: Abandon loop after poll call when array of fds has changed 22668 22669 If a file descriptor is added or removed from an ospoll callback, then 22670 the arrays containing file descriptor information will have all of 22671 their indices changed, so the loop state is no longer consistent. Just 22672 bail out and let the caller come back around to try again. 22673 22674 Signed-off-by: Keith Packard <keithp@keithp.com> 22675 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 22676 22677commit 7d6fffb6013cfc718d884cc549d7a8fef37240cf 22678Author: Keith Packard <keithp@keithp.com> 22679Date: Fri Aug 12 15:23:55 2016 -0700 22680 22681 os: Delay freeing of epoll_wait return pointers until epoll is idle. 22682 22683 Instead of freeing the struct ospollfd elements when the fd is 22684 removed by the user, delay that until epoll is idle so that we are 22685 sure no epoll_event structures could contain the stale pointer. This 22686 handles cases where an fd is removed from the ospoll callback 22687 interface, and also in case the OS keeps stale pointers around after 22688 the call to epoll_ctl with EPOLL_CTL_DEL. 22689 22690 Signed-off-by: Keith Packard <keithp@keithp.com> 22691 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 22692 22693commit 69b782aa75bc06f11b8f9b532d5213f252c4c6c4 22694Author: Keith Packard <keithp@keithp.com> 22695Date: Fri Jul 29 17:45:45 2016 -0700 22696 22697 xfree86: Set pScrn->pScreen before driver ScreenInit is called 22698 22699 Any code called from the driver ScreenInit may want to refer to 22700 pScrn->pScreen. As the function passed to AddScreen is the first place 22701 the DDX sees a new screen, the generic code needs to make sure that 22702 value is set before passing control to the video driver's 22703 initialization code. 22704 22705 This was found by running a driver which didn't bother to set this 22706 value when the initial colormap was installed; xf86RandR12LoadPalette 22707 tried to use pScrn->pScreen and crashed. 22708 22709 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97124 22710 Signed-off-by: Keith Packard <keithp@keithp.com> 22711 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 22712 22713commit 387d6b5df3e11de5cb88db8a559128fcb62e4e5a 22714Merge: d2558f063 8bb9d460f 22715Author: Keith Packard <keithp@keithp.com> 22716Date: Sat Aug 13 09:12:19 2016 -0700 22717 22718 Merge remote-tracking branch 'daenzer/for-master' 22719 22720commit d2558f063a034a827247b66b4c5adbffe79dc0e7 22721Author: Keith Packard <keithp@keithp.com> 22722Date: Thu Aug 11 17:28:38 2016 -0700 22723 22724 Build glamor when Xorg or Xephyr are built. 22725 22726 Requires gbm when building Xorg so that xf86-video-modesetting will 22727 work. 22728 22729 Signed-off-by: Keith Packard <keithp@keithp.com> 22730 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 22731 Reviewed-by: Eric Anholt <eric@anholt.net> 22732 22733commit 2b9f8ae98602d2abda7c363900f39a5bd3768a4c 22734Author: Keith Packard <keithp@keithp.com> 22735Date: Thu Aug 11 21:32:59 2016 -0700 22736 22737 mi: Remove spurious call to OsReleaseSignals from mieqGrowQueue 22738 22739 This call wasn't converted to 'input_unlock()' when the SIGIO code was 22740 removed from the server, and so when the queue growing was reworked to 22741 be done from the input thread, it got left sitting here. As the caller 22742 now manages the lock, we don't need to switch this to input_unlock at 22743 this point. 22744 22745 Signed-off-by: Keith Packard <keithp@keithp.com> 22746 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 22747 22748commit f399919e13c994452f7219163b2a4b1fa015242e 22749Author: Peter Hutterer <peter.hutterer@who-t.net> 22750Date: Fri Aug 12 13:59:56 2016 +1000 22751 22752 xfree86: lock input during PreInit 22753 22754 This is a problem for the libinput driver that uses the same context across 22755 multiple devices. The driver may be halfway through setting up an input device 22756 (and the only way to do so is to add it to libinput) when the input thread 22757 comes in an reads events. This then causes mayhem when data is dereferenced 22758 that hasn't been set up yet. 22759 22760 In my case the cause was the call to libinput_path_remove_device() inside 22761 preinit racing with evdev_dispatch_device() handling of ENODEV. The sequence 22762 was: 22763 - thread 2 gets an event and calls evdev_dispatch_device() 22764 - thread 1 calls libinput_path_remove_device() which sets the device->source 22765 to NULL 22766 - thread 2 reads from the fd, gets ENODEV and now removes the device->source, 22767 dereferencing the null-pointer 22768 22769 This is the one I could reproduce the most, but there are other potential 22770 pitfalls that affect any driver that uses the same fd for multiple devices. 22771 Avoid all this and wrap PreInit into the lock. 22772 22773 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 22774 Reviewed-by: Keith Packard <keithp@keithp.com> 22775 22776commit dd4e21cb3a6e692e834ec34bec42944202b3e085 22777Author: Peter Hutterer <peter.hutterer@who-t.net> 22778Date: Fri Aug 12 12:06:18 2016 +1000 22779 22780 xfree86: fix unbalanced input_lock/unlock in xf86NewInputDevice() 22781 22782 If a device couldn't be enabled we left the lock hanging. 22783 22784 This patch also removes the leftover OsReleaseSignals() call, now unnecessary. 22785 Note that input_unlock() is later than previously OsReleaseSignals(). 22786 RemoveDevice() manipulates the input device and its file descriptors, it's 22787 safer to put the input_unlock() call after RemoveDevice() to avoid events 22788 coming in while the device is being removed. 22789 22790 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 22791 Reviewed-by: Keith Packard <keithp@keithp.com> 22792 22793commit bf31d6f43e5ce04891a96b226a975379e2e2ba71 22794Author: Keith Packard <keithp@keithp.com> 22795Date: Thu Aug 11 12:34:54 2016 -0700 22796 22797 os: Allow re-registering fd with InputThreadRegisterDev 22798 22799 Calling InputThreadRegisterDev twice with the same fd should replace 22800 the existing function and args instead of creating a new entry with 22801 the same fd. 22802 22803 Signed-off-by: Keith Packard <keithp@keithp.com> 22804 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 22805 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 22806 22807commit 2df2815d6ae0c2ffb9e0a084d61ee3585f5116df 22808Author: Peter Hutterer <peter.hutterer@who-t.net> 22809Date: Tue Aug 9 17:35:58 2016 +1000 22810 22811 os: append, rather than prepend, any new input thread fds 22812 22813 xf86AddEnabledDevice() prepends the new fd to the list, 22814 xf86RemoveEnabledDevice() then searches for a matching fd and removes that 22815 entry. If this is done for the same fd (and since we lose all information but 22816 the actual fd) we usually unregister virtual devices in reverse order, causing 22817 a dereference of already released memory. 22818 22819 Case in point: 22820 - the wacom driver calls xf86AddEnabledDevice() once for the physical device, 22821 then multiple times for the virtual subdevices 22822 - when the physical device is unplugged, the driver calls 22823 xf86RemoveEnabledDevice() for the physical device 22824 - all we have is the fd, so we end up removing the last virtual device from 22825 the fd set 22826 - xf86DeleteInput() frees the physical device's pInfo 22827 - the fd goes crazy with ENODEV, but a read_input() now passes the already 22828 freed pInfo for the physical device 22829 - boom 22830 22831 Fix this by appending to the fd list to provide bug-for-bug compatibility with 22832 the old SIGIO code. This needs to be fixed in the driver, but meanwhile not 22833 crashing the server provides for better user experience. 22834 22835 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 22836 Reviewed-by: Keith Packard <keithp@keithp.com> 22837 22838commit a446ff84de2dd29439521f6e87d75bde3bbf002c 22839Author: Michel Dänzer <michel.daenzer@amd.com> 22840Date: Fri Jul 29 17:38:21 2016 +0900 22841 22842 xfree86/modes: Handle no palette case better in xf86RandR12CrtcSetGamma 22843 22844 Just use the RandR gamma ramp directly. 22845 22846 Fixes random on-monitor colours with drivers which don't call 22847 xf86HandleColormaps, e.g. modesetting. 22848 22849 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97154 22850 22851 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 22852 22853commit d3d4ff0ed4d4679b7983b1a3ea267ff945f7643c 22854Author: Kenneth Graunke <kenneth@whitecape.org> 22855Date: Thu Jul 28 07:28:51 2016 -0700 22856 22857 modesetting: Delete dead drmmode_bo_for_pixmap function. 22858 22859 Embarassingly, it looks like I introduced this dead function in 22860 commit 13c7d53df8dac45ea2a685826cd45a39bcb51657 a year ago. 22861 Nothing ever used it, not even then. 22862 22863 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 22864 Reviewed-by: Eric Anholt <eric@anholt.net> 22865 22866commit 8bb9d460fbc82ee5d6f87e1d11cfa52f4f3072ee 22867Author: Michel Dänzer <michel.daenzer@amd.com> 22868Date: Fri Jul 29 17:38:21 2016 +0900 22869 22870 xfree86/modes: Handle no palette case better in xf86RandR12CrtcSetGamma 22871 22872 Just use the RandR gamma ramp directly. 22873 22874 Fixes random on-monitor colours with drivers which don't call 22875 xf86HandleColormaps, e.g. modesetting. 22876 22877 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97154 22878 22879 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 22880 22881commit c833c0866f2f8f829185667efe3d6dfa5979a9e8 22882Author: Michel Dänzer <michel.daenzer@amd.com> 22883Date: Fri Jul 29 18:41:06 2016 +0900 22884 22885 present: Handle event mask updates as specified v2 22886 22887 From the Present extension specification: 22888 22889 An event context is associated with a specific window; using 22890 an existing event context with a different window generates 22891 a Match error. 22892 22893 If eventContext specifies an existing event context, then if 22894 eventMask is empty, PresentSelectInput deletes the specified 22895 context, otherwise the specified event context is changed to 22896 select a different set of events. 22897 22898 If eventContext is an unused XID, then if eventMask is empty 22899 no operation is performed. Otherwise, a new event context is 22900 created selecting the specified events. 22901 22902 Without this change, there's no way for a client to explicitly change 22903 or destroy an existing event mask entry. Trying to do so as specified 22904 above would just result in a protocol error. 22905 22906 v2: (Keith Packard) 22907 * Use dixLookupResourceByType instead of walking window_priv->events 22908 * Return BadMatch if the existing event context is associated with a 22909 different window or client 22910 * Call LEGAL_NEW_RESOURCE again when creating a new event context 22911 * Drop invalid "leak fix" 22912 22913 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 22914 Signed-off-by: Keith Packard <keithp@keithp.com> 22915 Reviewed-by: Keith Packard <keithp@keithp.com> 22916 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 22917 22918commit 0924ac014d7caadab0b15ba69cd0a09cfe8a01da 22919Author: Michel Dänzer <michel.daenzer@amd.com> 22920Date: Tue Jul 26 17:28:33 2016 +0900 22921 22922 present: Only call restore_screen_pixmap once from set_abort_flip 22923 22924 present_restore_screen_pixmap's work doesn't need to be done several 22925 times for the same pending flip. 22926 22927 Fixes a crash if the X server quits while a flip is pending, in which 22928 case present_set_abort_flip may be called several times, including when 22929 screen->root is already cleared to NULL. 22930 22931 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 22932 22933commit b4e46c0444bb09f4af59d9d13acc939a0fbbc6d6 22934Author: Michel Dänzer <michel.daenzer@amd.com> 22935Date: Sat Nov 28 16:50:47 2015 +0900 22936 22937 xfree86: Hook up colormaps and RandR 1.2 gamma code v6 22938 22939 Instead of breaking the former when the driver supports the latter, 22940 hook them up so that the hardware LUTs reflect the combination of the 22941 current colourmap and gamma states. I.e. combine the colourmap, the 22942 global gamma value/ramp and the RandR 1.2 per-CRTC gamma ramps into one 22943 combined LUT per CRTC. 22944 22945 Fixes e.g. gamma sliders not working in games. 22946 22947 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=27222 22948 22949 v2: 22950 * Initialize palette_size and palette struct members, fixes crash on 22951 server startup. 22952 v3: 22953 * Free randrp->palette in xf86RandR12CloseScreen, fixes memory leak. 22954 v4: 22955 * Call CMapUnwrapScreen if xf86RandR12InitGamma fails (Emil Velikov). 22956 * Still allow xf86HandleColormaps to be called with a NULL loadPalette 22957 parameter in the xf86_crtc_supports_gamma case. 22958 v5: 22959 * Clean up inner loops in xf86RandR12CrtcComputeGamma (Keith Packard) 22960 * Move palette update out of per-CRTC loop in xf86RandR12LoadPalette 22961 (Keith Packard) 22962 v6: 22963 * Handle reallocarray failure in xf86RandR12LoadPalette (Keith Packard) 22964 22965 Reviewed-by: Keith Packard <keithp@keithp.com> 22966 22967commit 17213b74fd7fc4c4e2fe7a3781e7422dd482a0ab 22968Author: Michel Dänzer <michel.daenzer@amd.com> 22969Date: Tue Jun 21 16:44:20 2016 +0900 22970 22971 xfree86/modes: Remove xf86RandR12CrtcGetGamma 22972 22973 This would normally return the same values the core RandR code passed to 22974 xf86RandR12CrtcSetGamma before, which is rather pointless. The only 22975 possible exception would be if a driver tried initializing 22976 crtc->gamma_red/green/blue to reflect the hardware LUT state on startup, 22977 but that can't work correctly if whatever set the LUT before the server 22978 started was running at a different depth. 22979 22980 Even the pointless round-trip case will no longer work with the 22981 following change. 22982 22983 Reviewed-by: Keith Packard <keithp@keithp.com> 22984 22985commit 62f44052573b475a7b4c24f8e2da9bd6a8f1bd70 22986Author: Michel Dänzer <michel.daenzer@amd.com> 22987Date: Tue Jun 21 15:41:45 2016 +0900 22988 22989 xfree86/modes: Move gamma initialization to xf86RandR12Init12 v2 22990 22991 RRCrtcGammaSetSize cannot be used yet in xf86InitialConfiguration, 22992 because randr_crtc isn't allocated yet at that point, but a following 22993 change will require RRCrtcGammaSetSize to be called from 22994 xf86RandR12CrtcInitGamma. 22995 22996 v2: 22997 * Bail from xf86RandR12CrtcInitGamma if !crtc->funcs->gamma_set (Keith 22998 Packard) 22999 23000 Reviewed-by: Keith Packard <keithp@keithp.com> 23001 23002commit 0b2f30834b1a9f4a03542e25c5f54ae800df57e2 23003Author: Keith Packard <keithp@keithp.com> 23004Date: Sun May 29 17:47:29 2016 -0700 23005 23006 os: Clean up WaitFor.c 23007 23008 Do all timer stuff before blocking, avoiding a bunch of duplicate code 23009 and merge common code in WaitForSomething. 23010 23011 The WaitForSomething changes need a bit of explanation to show that 23012 the new code is effectively equivalent to the old. Eliding error 23013 checking and trivial bits we've got: 23014 23015 Before: 23016 23017 if (ready clients) 23018 timeout = 0 23019 else 23020 compute timeout 23021 i = poll 23022 if (i <= 0) { 23023 if (ready clients) 23024 return TRUE; 23025 if (input) 23026 return FALSE; 23027 if (any ready timers) { 23028 run timers 23029 return FALSE; 23030 } 23031 } else { 23032 if (input) 23033 return FALSE; 23034 if (any ready timers) { 23035 run timers 23036 return FALSE; 23037 } 23038 if (ready clients) 23039 return TRUE; 23040 } 23041 23042 After: 23043 23044 if (ready clients) 23045 timeout = 0; 23046 else 23047 compute timeout 23048 run_timers 23049 poll 23050 23051 if (input) 23052 return FALSE; 23053 23054 if (ready clients) 23055 return TRUE; 23056 23057 The old code would return TRUE if there were ready clients and input 23058 pending. Dispatch would then schedule that ready client, but before 23059 processing any requests, it would notice that there was input pending 23060 and go process it. The new code just checks for input first, which is 23061 effectively the same. 23062 23063 If the poll timed out and there weren't clients ready, then timers 23064 would get run. 23065 23066 If the poll didn't time out, then timers would get run, even if there 23067 were clients now ready. Now, if the timeout interval was zero, that 23068 means that the timers must have been ready *before* poll was 23069 invoked. In this case, we should simply run the timers before calling 23070 poll -- no sense calling poll just to discard any data that it 23071 generates. 23072 23073 If the timeout interval was non-zero, and poll didn't timeout, then 23074 either there aren't any timers to run, or we got a surprise and hit a 23075 timer exactly as a client became ready to run. This is the one case 23076 where the new code is different from the old; the new code delays the 23077 timer call until the next time WaitForSomething is called. 23078 23079 Signed-off-by: Keith Packard <keithp@keithp.com> 23080 Reviewed-by: Adam Jackson <ajax@redhat.com> 23081 23082commit 2ab8b1dcd37900e5e2f11007d7c0cd0545209482 23083Author: Keith Packard <keithp@keithp.com> 23084Date: Sun May 29 19:48:25 2016 -0700 23085 23086 os: Use xorg_list for struct _OsTimerRec 23087 23088 No sense having an open-coded linked list here, plus the doubly linked 23089 list is more efficient 23090 23091 Signed-off-by: Keith Packard <keithp@keithp.com> 23092 Reviewed-by: Adam Jackson <ajax@redhat.com> 23093 23094commit 50779c494d4682103b96db440021e56e44311b6b 23095Author: Keith Packard <keithp@keithp.com> 23096Date: Sun May 29 12:45:53 2016 -0700 23097 23098 os: Remove CheckConnections 23099 23100 poll provides per-fd notification of failure, so we don't need 23101 CheckConnections anymore. 23102 23103 Signed-off-by: Keith Packard <keithp@keithp.com> 23104 Reviewed-by: Adam Jackson <ajax@redhat.com> 23105 23106commit f0275b1e5a4787da1f4d5507165315a000c22b33 23107Author: Keith Packard <keithp@keithp.com> 23108Date: Thu May 26 00:19:55 2016 -0700 23109 23110 os: Leave stdin and stdout open 23111 23112 There's no reason to close these now that we don't care what file 23113 descriptors we use. 23114 23115 Signed-off-by: Keith Packard <keithp@keithp.com> 23116 Reviewed-by: Adam Jackson <ajax@redhat.com> 23117 23118commit 8217c29d2d8b29bd66bc54fee3fe5cb3385a05c7 23119Author: Keith Packard <keithp@keithp.com> 23120Date: Wed May 25 23:43:49 2016 -0700 23121 23122 Allow 1024 and 2048 for LimitClients 23123 23124 There's no reason not to offer ridiculous numbers of clients; only a 23125 few static data structures are arrays of this length. 23126 23127 Signed-off-by: Keith Packard <keithp@keithp.com> 23128 Reviewed-by: Adam Jackson <ajax@redhat.com> 23129 23130commit e0edb963fe09582f23a4b55da4f8840173e7a1ee 23131Author: Keith Packard <keithp@keithp.com> 23132Date: Tue May 24 21:04:14 2016 -0700 23133 23134 os: eliminate fd value limits for clients 23135 23136 With no code depending on the range of file descriptors, checking 23137 for that can be eliminated. 23138 23139 Signed-off-by: Keith Packard <keithp@keithp.com> 23140 Reviewed-by: Adam Jackson <ajax@redhat.com> 23141 23142commit 30bc0732f959bbc63f318c06d48de080d495da32 23143Author: Keith Packard <keithp@keithp.com> 23144Date: Tue May 24 21:12:33 2016 -0700 23145 23146 os: Use ospoll for input thread [v2] 23147 23148 Replace use of select(2) to avoid fd limits. Note that 23149 InputThreadFillPipe used select as well, but none of the files passed 23150 were non-blocking, so there was no need for that code at all. 23151 23152 v2: Keep ospoll API usage single threaded to avoid re-entrancy issues 23153 23154 Signed-off-by: Keith Packard <keithp@keithp.com> 23155 Reviewed-by: Adam Jackson <ajax@redhat.com> 23156 23157commit f993091e7db81b0420e23c485378cba112278839 23158Author: Keith Packard <keithp@keithp.com> 23159Date: Thu May 26 10:40:44 2016 -0700 23160 23161 os: Switch server to poll(2) [v3] 23162 23163 Eliminates all of the fd_set mangling in the server main thread 23164 23165 v2: Listen for POLLOUT while writes are blocked. 23166 23167 v3: Only mark client not ready on EAGAIN return from read 23168 23169 Signed-off-by: Keith Packard <keithp@keithp.com> 23170 Reviewed-by: Adam Jackson <ajax@redhat.com> 23171 23172commit 8f1edf4bd3a1f050ce9eeb5eac45dd1a8f7a6d5e 23173Author: Keith Packard <keithp@keithp.com> 23174Date: Thu May 19 13:59:54 2016 -0700 23175 23176 dix: Use list for ready clients 23177 23178 This converts the dispatch loop into using a list of ready clients 23179 instead of an array. This changes the WaitForSomething API so that it 23180 notifies DIX when a client becomes ready to read, instead of returning 23181 the set of ready clients. 23182 23183 Signed-off-by: Keith Packard <keithp@keithp.com> 23184 Reviewed-by: Adam Jackson <ajax@redhat.com> 23185 23186commit d6eff3c31e8289881a3aa9b858e5710d0f741db0 23187Author: Keith Packard <keithp@keithp.com> 23188Date: Thu May 26 10:20:45 2016 -0700 23189 23190 os: Add ospoll interface [v2] 23191 23192 This provides a wrapper around poll or epoll providing a 23193 callback-based interface for monitoring activity on a large set of 23194 file descriptors. 23195 23196 v2: use xserver_poll API instead of poll. Don't use WSAPoll as 23197 that is broken. 23198 23199 Signed-off-by: Keith Packard <keithp@keithp.com> 23200 Reviewed-by: Adam Jackson <ajax@redhat.com> 23201 23202commit d403aca70a07e1401cb93738f1af5961582a2e47 23203Author: Keith Packard <keithp@keithp.com> 23204Date: Mon Jul 18 15:58:42 2016 -0700 23205 23206 Switch poll() users to xserver_poll() 23207 23208 This uses the wrapper in case we need to emulate poll with select 23209 as we do on Windows. 23210 23211 Reviewed-by: Adam Jackson <ajax@redhat.com> 23212 Signed-off-by: Keith Packard <keithp@keithp.com> 23213 23214commit 711c36558f50943c8342f25ad210281134887a3d 23215Author: Brian M. Clapper <bmc@clapper.org> 23216Date: Mon Jul 11 18:30:05 2016 -0400 23217 23218 os: Add poll emulation for mingw [v2] 23219 23220 v2: rename as 'xserver_poll' to avoid potential library name 23221 collisions. Provide 'xserver_poll.h' which uses the system 23222 poll where available and falls back to this emulation otherwise. 23223 Autodetects when this is required, building the emulation only 23224 then 23225 23226 Source: https://github.com/bmc/poll 23227 Signed-off-by: Adam Jackson <ajax@redhat.com> 23228 23229commit 2a79be9e4dd1b6ba00b69ff40bc5257ec84d34da 23230Author: Jon Turney <jon.turney@dronecode.org.uk> 23231Date: Tue Jul 19 13:35:37 2016 +0100 23232 23233 hw/xwin: Update BlockHandler function signature 23234 23235 Update for removal of fdset from Block/Wakeup handler API in 9d15912a 23236 23237 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 23238 Reviewed-by: Keith Packard <keithp@keithp.com> 23239 23240commit 60a91031d13e4d29c383087120e318f6b528b6e5 23241Author: Jon Turney <jon.turney@dronecode.org.uk> 23242Date: Tue Jul 19 13:35:36 2016 +0100 23243 23244 hw/xwin: Update for removal of AddEnabledDevice 23245 23246 Update for removal of AddEnabledDevice in be5a513f. Use SetNotifyFd instead. 23247 23248 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 23249 Reviewed-by: Keith Packard <keithp@keithp.com> 23250 23251commit 7f2d690725292e7b5a44c813c9456da426256873 23252Author: Peter Hutterer <peter.hutterer@who-t.net> 23253Date: Mon Jun 27 10:17:14 2016 +1000 23254 23255 xfree86: if ATTR_KEYBOARD is set, match for keyboards 23256 23257 ATTR_KEY maps to ID_INPUT_KEY which is set for any device with keys. 23258 ID_INPUT_KEYBOARD and thus ATTR_KEYBOARD is set for devices that are actual 23259 keyboards (and have a set of expected keys). 23260 23261 Hand-written match rules may only apply ID_INPUT_KEYBOARD, so make sure we 23262 match on that too. 23263 23264 Arguably we should've been matching on ATTR_KEYBOARD only all along but 23265 changing that likely introduces regressions. 23266 23267 Reported-by: Marty Plummer <netz.kernel@gmail.com> 23268 Reviewed-by: Adam Jackson <ajax@redhat.com> 23269 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 23270 23271commit 4b311d23e84356bd0e9e736aeed7448dd6382118 23272Author: Adam Jackson <ajax@redhat.com> 23273Date: Mon Jul 18 12:46:51 2016 -0400 23274 23275 modesetting: resubmit dirty rects on EINVAL (v2) 23276 23277 This error code can mean we're submitting more rects at once than the 23278 driver can handle. If that happens, resubmit one at a time. 23279 23280 v2: Make the rect submit loop more error-proof (Walter Harms) 23281 23282 Signed-off-by: Adam Jackson <ajax@redhat.com> 23283 Reviewed-by: Michael Thayer <michael.thayer@oracle.com> 23284 23285commit 8d3a368d8980e37e7e8c57065dc901ce809887c6 23286Author: Keith Packard <keithp@keithp.com> 23287Date: Wed Jun 1 22:47:29 2016 -0700 23288 23289 os: InputThreadFillPipe doesn't need select or poll 23290 23291 The file descriptors passed to InputThreadFillPipe are always 23292 blocking, so there's no need to use Select (or poll). 23293 23294 Signed-off-by: Keith Packard <keithp@keithp.com> 23295 Reviewed-by: Adam Jackson <ajax@redhat.com> 23296 23297commit ef7ddbe242ed4c461f816663fb88646e41f1c21b 23298Author: Keith Packard <keithp@keithp.com> 23299Date: Wed Jun 1 22:35:09 2016 -0700 23300 23301 os: Move ETEST macro from io.c to osdep.h 23302 23303 This lets other code share this functionality 23304 23305 Signed-off-by: Keith Packard <keithp@keithp.com> 23306 Reviewed-by: Adam Jackson <ajax@redhat.com> 23307 23308commit 0d294462a5af08ada654c588fad921ed7a22749b 23309Author: Keith Packard <keithp@keithp.com> 23310Date: Sun May 29 15:06:36 2016 -0700 23311 23312 os: Add X_NOTIFY_ERROR value 23313 23314 This provides a way to report errors on file descriptors that is 23315 better defined than "any bits which are not READ or WRITE". 23316 23317 Signed-off-by: Keith Packard <keithp@keithp.com> 23318 Reviewed-by: Adam Jackson <ajax@redhat.com> 23319 23320commit e6636b438322a9a2f2270ad9d60bf3dfc72be0b3 23321Author: Keith Packard <keithp@keithp.com> 23322Date: Thu May 26 10:30:56 2016 -0700 23323 23324 os: Compute timeout in milliseconds instead of struct timeval 23325 23326 The timeout resolution offered in the AdjustWaitForDelay call is 23327 only milliseconds, so passing around the timeout as a pointer to a 23328 struct timeval is not helpful. Doing everything in milliseconds up to 23329 the point of the select call simplifies the code without affecting 23330 functionality at all. 23331 23332 Signed-off-by: Keith Packard <keithp@keithp.com> 23333 Reviewed-by: Adam Jackson <ajax@redhat.com> 23334 23335commit a414db021575accff64abad6f1047245e81c7476 23336Author: Keith Packard <keithp@keithp.com> 23337Date: Thu May 19 15:08:05 2016 -0700 23338 23339 dix: Intermediate GrabServer state 'GrabKickout' not needed 23340 23341 The intermediate grabState, "GrabKickout", was used to trigger 23342 dispatch into going back to WaitForSomething after doing a GrabServer 23343 so that the set of ready clients would be recomputed to match what the 23344 server should be processing. As we only process one client per 23345 WaitForSomething call, we will always hit WaitForSomething after 23346 finishing the current client, and so don't need any special case here. 23347 23348 Signed-off-by: Keith Packard <keithp@keithp.com> 23349 Reviewed-by: Adam Jackson <ajax@redhat.com> 23350 23351commit 7762a602c1dfdd8cfcf2b8c2281cf4d683d05216 23352Author: Keith Packard <keithp@keithp.com> 23353Date: Thu May 19 15:05:55 2016 -0700 23354 23355 dix/os: Merge priority computation into SmartScheduleClient 23356 23357 Instead of having scheduling done in two places (one in 23358 WaitForSomething, and the other in SmartScheduleClient), just stick 23359 all of the scheduling in SmartScheduleClient. 23360 23361 Signed-off-by: Keith Packard <keithp@keithp.com> 23362 Reviewed-by: Adam Jackson <ajax@redhat.com> 23363 23364commit 4af00242ef1e39499b932d12423fdf449296090a 23365Author: Keith Packard <keithp@keithp.com> 23366Date: Mon May 30 01:49:46 2016 -0700 23367 23368 Bump ABI versions to reflect block/wakeup handler API changes 23369 23370 Signed-off-by: Keith Packard <keithp@keithp.com> 23371 Reviewed-by: Adam Jackson <ajax@redhat.com> 23372 23373commit be5a513fee6cbf29ef7570e57eb0436d70fbd88c 23374Author: Keith Packard <keithp@keithp.com> 23375Date: Mon Dec 7 15:12:14 2015 -0800 23376 23377 Remove AddEnabledDevice and AddGeneralSocket APIs 23378 23379 All uses of these interfaces should instead be using the NotifyFd API 23380 instead. 23381 23382 Signed-off-by: Keith Packard <keithp@keithp.com> 23383 Reviewed-by: Adam Jackson <ajax@redhat.com> 23384 23385commit 9d15912aa475b733bbb20efc367a67dacad63bf1 23386Author: Keith Packard <keithp@keithp.com> 23387Date: Tue Sep 1 18:51:14 2015 -0700 23388 23389 Remove fd_set from Block/Wakeup handler API 23390 23391 This removes the last uses of fd_set from the server interfaces 23392 outside of the OS layer itself. 23393 23394 Signed-off-by: Keith Packard <keithp@keithp.com> 23395 Reviewed-by: Adam Jackson <ajax@redhat.com> 23396 23397commit fb0802113b4c57819cba15d64baf79bf4148607e 23398Author: Keith Packard <keithp@keithp.com> 23399Date: Tue Sep 1 11:20:04 2015 -0700 23400 23401 Remove readmask from screen block/wakeup handler 23402 23403 With no users of the interface needing the readmask anymore, we can 23404 remove it from the argument passed to these functions. 23405 23406 Signed-off-by: Keith Packard <keithp@keithp.com> 23407 Reviewed-by: Adam Jackson <ajax@redhat.com> 23408 23409commit 410bc047480a9f98df678dc850bc6b99c3cfb5bf 23410Author: Keith Packard <keithp@keithp.com> 23411Date: Mon Dec 7 15:03:10 2015 -0800 23412 23413 dmx: Eliminate use of AddEnabledDevice [v2] 23414 23415 Use SetNotifyFd instead, with the hope that someday someone will come 23416 fix this to be more efficient -- right now, the wakeup handler is 23417 doing the event reading, instead of the notify callback. 23418 23419 v2: no need to patch dmxsigio.c as it has been removed. 23420 23421 Reviewed-by: Adam Jackson <ajax@redhat.com> 23422 Signed-off-by: Keith Packard <keithp@keithp.com> 23423 23424commit 6299ef3d749d6f978d3d38d42f711ac56bf382eb 23425Author: Keith Packard <keithp@keithp.com> 23426Date: Thu Aug 27 11:49:35 2015 -0700 23427 23428 modesetting: Use passed-in fd for drm event monitoring NotifyFd callback 23429 23430 This is a cleanup, proposed by Adam Jackson, but wasn't merged with 23431 the original NotifyFD changes. 23432 23433 Signed-off-by: Keith Packard <keithp@keithp.com> 23434 Reviewed-by: Adam Jackson <ajax@redhat.com> 23435 23436commit 6bf7b49f6711c7ed1837283dc04f93f4c1b77ecc 23437Author: Keith Packard <keithp@keithp.com> 23438Date: Thu Aug 27 11:47:26 2015 -0700 23439 23440 hw/kdrive: Use passed-in fd for kdrive/linux APM monitoring [v2] 23441 23442 This is a cleanup, proposed by Adam Jackson, but wasn't merged with 23443 the original NotifyFD changes. 23444 23445 Signed-off-by: Keith Packard <keithp@keithp.com> 23446 Reviewed-by: Adam Jackson <ajax@redhat.com> 23447 23448commit 55c2e1a3aa587c58a74066724e11e30b3df267b8 23449Author: Keith Packard <keithp@keithp.com> 23450Date: Mon Dec 7 15:11:33 2015 -0800 23451 23452 xnest: Use SetNotifyFd to receive events 23453 23454 Signed-off-by: Keith Packard <keithp@keithp.com> 23455 Reviewed-by: Adam Jackson <ajax@redhat.com> 23456 23457commit 24e65bf0db57bf4ac70386c0a0e8275b73edd2fb 23458Author: Keith Packard <keithp@keithp.com> 23459Date: Thu Dec 3 01:02:02 2015 -0600 23460 23461 hw/xfree86: Use NotifyFd for other input fd wakeups 23462 23463 Remove code in xf86Wakeup for dealing with other input and switch to 23464 using the new NotifyFd interface. 23465 23466 Signed-off-by: Keith Packard <keithp@keithp.com> 23467 Reviewed-by: Adam Jackson <ajax@redhat.com> 23468 23469commit c3fea428aed919826130ef8ebdb2cceb445a845b 23470Author: Keith Packard <keithp@keithp.com> 23471Date: Tue May 24 20:51:31 2016 -0700 23472 23473 os: Use NotifyFd for ErrorConnMax 23474 23475 Instead of open-coding a single FD wait, use NotifyFd to wait for the 23476 FD to become readable before returning the error message. 23477 23478 Signed-off-by: Keith Packard <keithp@keithp.com> 23479 Reviewed-by: Adam Jackson <ajax@redhat.com> 23480 23481commit 559aac2d71250e3137aaa582e2a59a918ddf21b7 23482Author: Keith Packard <keithp@keithp.com> 23483Date: Tue May 24 21:59:38 2016 -0700 23484 23485 dmx: Switch from select(2) to poll(2) for input 23486 23487 Signed-off-by: Keith Packard <keithp@keithp.com> 23488 Reviewed-by: Adam Jackson <ajax@redhat.com> 23489 23490commit aa6717ce213e79735c72afc5ec9cc1f9c0297e09 23491Author: Keith Packard <keithp@keithp.com> 23492Date: Tue May 24 21:36:18 2016 -0700 23493 23494 xfree86: Switch from select(2) to poll(2) 23495 23496 xf86WaitForInput and the xf86 SIGIO handling code. 23497 23498 Signed-off-by: Keith Packard <keithp@keithp.com> 23499 Reviewed-by: Adam Jackson <ajax@redhat.com> 23500 23501commit 81135991a583b3b30a90e82ddc1d5c86d57bf00b 23502Author: Keith Packard <keithp@keithp.com> 23503Date: Tue May 24 21:19:13 2016 -0700 23504 23505 kdrive: switch from select(2) to poll(2) 23506 23507 This avoids fd limits 23508 23509 Signed-off-by: Keith Packard <keithp@keithp.com> 23510 Reviewed-by: Adam Jackson <ajax@redhat.com> 23511 23512commit 05a793f5b3c40747d5a92a076def7f4fb673c7e7 23513Author: Keith Packard <keithp@keithp.com> 23514Date: Tue Sep 1 18:50:55 2015 -0700 23515 23516 dix: Switch to the libXfont2 API (v2) 23517 23518 This new libXfont API eliminates exposing internal X server symbols to 23519 the font library, replacing those with a struct full of the entire API 23520 needed to use that library. 23521 23522 v2: Use libXfont2 instead of libXfont_2 23523 23524 Signed-off-by: Keith Packard <keithp@keithp.com> 23525 Reviewed-by: Adam Jackson <ajax@redhat.com> 23526 23527commit 950ffb8d6fd1480f305e38c571bda44f247f1de2 23528Author: Eric Anholt <eric@anholt.net> 23529Date: Fri Nov 6 12:27:42 2015 -0800 23530 23531 glamor: Remove the FBO cache. 23532 23533 It is a modest performance improvement (2.7% on Intel), with the 23534 significant downside that it keeps extra pixmap contents laying around 23535 for 1000 BlockHandlers without the ability for the system to purge 23536 them when under memory pressure, and tiled renderers don't know that 23537 we could avoid reading their current contents when beginning to render 23538 again. We could use the FB invalidate functions, but they aren't 23539 always available, aren't hooked up well in Mesa, and would eat into 23540 the performance gains of having the cache. 23541 23542 [ajax: rebased to master] 23543 23544 Reviewed-by: Adam Jackson <ajax@redhat.com> 23545 23546commit e8e36755abb17872d669b88d33ca9adc511029a0 23547Author: Michel Dänzer <michel.daenzer@amd.com> 23548Date: Fri Jul 1 12:34:20 2016 +0900 23549 23550 glamor: Translate solid text background region after clipping 23551 23552 Fixes incorrect clipping for redirected windows which don't happen to be 23553 located at the top left corner of the screen. 23554 23555 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96742 23556 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 23557 Reviewed-by: Adam Jackson <ajax@redhat.com> 23558 23559commit 32a9504c69183485b0b796fa3966cd1e39992365 23560Author: Adam Jackson <ajax@redhat.com> 23561Date: Tue Jul 12 10:04:18 2016 -0400 23562 23563 xfree86: Fix fallback driver sort order for Xorg -configure (v2) 23564 23565 The intent here was that fallback drivers would be at the end of the 23566 list in order, but if a fallback driver happened to be at the end of the 23567 list already that's not what would happen. Rather than open-code 23568 something smarter, just use qsort. 23569 23570 Note that qsort puts things in ascending order, so somewhat backwardsly 23571 fallbacks are greater than native drivers, and vesa is greater than 23572 modesetting. 23573 23574 v2: Use strcmp to compare non-fallback drivers so we get a predictable 23575 result if your libc's qsort isn't stable (Keith Packard) 23576 23577 Reviewed-by: Keith Packard <keithp@keithp.com> 23578 Signed-off-by: Adam Jackson <ajax@redhat.com> 23579 23580commit 4926845a57fa8b53e18ea7d3434bf5539e9b7782 23581Author: Andrew Eikum <aeikum@codeweavers.com> 23582Date: Wed Jul 6 14:13:09 2016 -0500 23583 23584 shm: Also censor images returned by ShmGetImage 23585 23586 We currently censor images from dix's GetImage, but not from 23587 ShmGetImage. This is a method to bypass XACE, creating a potential 23588 leak. We should censor in both methods. 23589 23590 Reviewed-by: Adam Jackson <ajax@redhat.com> 23591 Signed-off-by: Andrew Eikum <aeikum@codeweavers.com> 23592 23593commit 9fcb554e9bfdf3eed2c2250d89150e3e7b907f01 23594Author: Adam Jackson <ajax@redhat.com> 23595Date: Fri Jul 8 15:17:05 2016 -0400 23596 23597 xwayland: Only force monotonic clock once 23598 23599 Otherwise on regeneration we get: 23600 23601 (EE) BUG: triggered 'if (clockid)' 23602 (EE) BUG: utils.c:440 in ForceClockId() 23603 (EE) 23604 (EE) Backtrace: 23605 (EE) 0: ./hw/xwayland/Xwayland (ForceClockId+0x5c) [0x47713c] 23606 (EE) 1: ./hw/xwayland/Xwayland (OsInit+0x25) [0x4763d5] 23607 (EE) 2: ./hw/xwayland/Xwayland (dix_main+0x11c) [0x43e60c] 23608 (EE) 3: /lib64/libc.so.6 (__libc_start_main+0xf1) [0x7f627b2f9731] 23609 (EE) 4: ./hw/xwayland/Xwayland (_start+0x29) [0x4238e9] 23610 (EE) 5: ? (?+0x29) [0x29] 23611 23612 Signed-off-by: Adam Jackson <ajax@redhat.com> 23613 Reviewed-by: Jonas Ådahl <jadahl@gmail.com> 23614 23615commit cf6730c503f8090a5d1b80918fe253fc2c5bc090 23616Author: Rui Matos <tiagomatos@gmail.com> 23617Date: Wed Jul 13 19:19:09 2016 +0200 23618 23619 xwayland: Update RR state on wl_output.done instead of wl_output.mode 23620 23621 Otherwise if the geometry changes but the mode doesn't we end up with 23622 the previous geometry from RR's point of view. 23623 23624 Fixes https://bugzilla.gnome.org/show_bug.cgi?id=768710 23625 23626 Reviewed-by: Jonas Ådahl <jadahl@gmail.com> 23627 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 23628 23629commit 401a8d6e1379133863e3271374dc21850d0d3cab 23630Author: Michel Dänzer <michel.daenzer@amd.com> 23631Date: Tue Jun 28 17:22:47 2016 +0900 23632 23633 dix: Work around non-premultiplied ARGB cursor data 23634 23635 Some games incorrectly use non-premultiplied ARGB cursor data, presumably 23636 because that's what Windows uses. On some hardware (and with SWcursor), 23637 this breaks areas of the cursor which are supposed to be transparent 23638 (and presumably also translucent areas, but that's less noticeable). 23639 23640 This change checks for pixels with alpha == 0 and any non-alpha component 23641 != 0. If any such pixel is found, the data is assumed to be 23642 non-premultiplied and fixed up by multiplying the RGB components with the 23643 alpha component. 23644 23645 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92309 23646 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 23647 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 23648 23649commit 033888e7766d226a179357d970223428c19c4b53 23650Author: Hans De Goede <hdegoede@redhat.com> 23651Date: Thu Jul 7 10:55:04 2016 +0200 23652 23653 linux: Do not try to open /dev/vc/0, fix error msg when /dev/tty0 open fails 23654 23655 /dev/vc/0 is a devfs thing which is long dead, so stop trying to open 23656 /dev/vc/0, besides being a (small) code cleanup this will also fix the 23657 "parse_vt_settings: Cannot open /dev/tty0 (%s)\n" error message to 23658 display the actual error, rather then the -ENOENT from also trying 23659 /dev/vc/0. 23660 23661 BugLink: https://patchwork.freedesktop.org/patch/8768/ 23662 Reported-by: Chad Versace <chad.versace@linux.intel.com> 23663 Suggested-by: Julien Cristau <jcristau@debian.org> 23664 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 23665 Reviewed-by: Julien Cristau <jcristau@debian.org> 23666 Reviewed-by: Chad Versace <chad.versace@intel.com> 23667 23668commit ce24f3aa5815e03c2d501d1504f6c0162b219801 23669Author: Alex Goins <agoins@nvidia.com> 23670Date: Thu Jul 7 14:50:03 2016 -0700 23671 23672 modesetting: NULL assignment for drmmode_set_target_scanout_target_cpu 23673 23674 Commit 80e64dae: "modesetting: Implement PRIME syncing as a sink" originally was 23675 supposed to have this line, but it was dropped as part of the merge process. 23676 23677 Foregoing the NULL assignment causes a ton of problems with dereferencing 23678 uninitialized memory. 23679 23680 Signed-off-by: Alex Goins <agoins@nvidia.com> 23681 Reviewed-by: Dave Airlie <airlied@redhat.com> 23682 23683commit df8e86931eb19c196ed2afc85d89525ef8cb711b 23684Author: agoins <agoins@nvidia.com> 23685Date: Thu Jun 16 20:06:48 2016 -0700 23686 23687 randr: Add ability to turn PRIME sync off 23688 23689 Adds an output parameter to disable PRIME synchronization. 23690 23691 Output parameter is created when the user calls 'xrandr 23692 --setprovideroutputsource <sink> <source>' to prevent polluting output 23693 parameters of non-PRIME configurations. 23694 23695 Defaults to on, so if the user wants PRIME synchronization they don't need 23696 to do anything. 23697 23698 If the user wishes to disable PRIME synchronization when they first set up 23699 PRIME, they can run 'xrandr --output <output> --set "PRIME Synchronization" 23700 0' after running 'xrandr --setprovideroutputsource <sink> <source>', but 23701 before 'xrandr --auto'. 23702 23703 If the user wishes to enable or disable PRIME synchronization after PRIME has 23704 already been set up, they can run 'xrandr --output <output> --set "PRIME 23705 Synchronization" <0 or 1>' at any time, xrandr will trigger a modeset, which 23706 will tear down and setup PRIME in the configuration they requested on CRTCs 23707 associated with that output. 23708 23709 randrstr.h: 23710 Add central definition of the output property name. 23711 23712 rrcrtc.c: 23713 Add function rrGetPixmapSharingSyncProp() to query the status of the 23714 output property. 23715 23716 Add function rrSetPixmapSharingSyncProp() to set the output property. 23717 23718 Add 'sync' parameter to rrSetupPixmapSharing(), which when false will 23719 use single buffering even if the required ABI functions are supported. 23720 Changes rrSetupPixmapSharing() to only report an error if falling back 23721 to single buffering when the user requested synchronization. 23722 23723 Change RRCrtcSet() to use rrPixmapSharingSyncProp() to query the status 23724 of the output property and feed it into rrSetupPixmapSharing() using 23725 the 'sync' parameter. 23726 23727 rrprovider.c: 23728 Add RR(Init/Fini)PrimeSyncProps(), functions to create and destroy the 23729 PRIME synchronization output property. 23730 23731 Add a call to RRInitPrimeSyncProps() in 23732 ProcRRSetProviderOutputSource(), such that the output property is 23733 created when the user requests PRIME. 23734 23735 Add a call to RRFiniPrimeSyncProps() in RRProviderDestroy(). 23736 23737 v1: Initial commit 23738 v2: Unchanged 23739 v3: Add /* TODO */ for handling different sources with different outputs 23740 Make rrSetupPixmapSharing() set the output property to 0 if it has to fall 23741 back, to avoid user confusion. 23742 Make rr(Get)PixmapSharingSyncProp() check the current value if there isn't a 23743 pending value 23744 v4: Unchanged 23745 v5: Unchanged 23746 v6: Rebase onto ToT 23747 v7: Unchanged 23748 23749 Signed-off-by: Alex Goins <agoins@nvidia.com> 23750 23751commit 60ad701a6a8cb9f1eacb72acfe2cb8d3b7a865dc 23752Author: Hans de Goede <hdegoede@redhat.com> 23753Date: Tue Jun 14 11:58:01 2016 +0200 23754 23755 modesetting: Load on GPU-s with 0 outputs 23756 23757 In newer laptops with switchable graphics, the GPU may have 0 outputs, 23758 in this case the modesetting driver should still load if the GPU is 23759 SourceOffload capable, so that it can be used as an offload source provider. 23760 23761 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 23762 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 23763 23764commit 94a1c77259ce39ba59ad87615df39b570ffab435 23765Author: Hans de Goede <hdegoede@redhat.com> 23766Date: Thu Jun 16 15:21:03 2016 +0200 23767 23768 modesetting: Fix swapping of provider sink / source capabilities 23769 23770 When a card has import capability it can be an offload _sink_, not 23771 a source and vice versa for export capability. 23772 23773 This commit fixes the modesetting driver to properly set these 23774 capabilities, this went unnoticed sofar because most gpus have both 23775 import and export capability. 23776 23777 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 23778 Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> 23779 23780commit fcbafdfd79fd03205bc324e23b6d4e4d4f7180e1 23781Author: Adam Jackson <ajax@redhat.com> 23782Date: Wed Jul 22 12:14:08 2015 -0400 23783 23784 modesetting: Implement a double-buffered shadow mode 23785 23786 Server GPUs often have a VNC feature attached to allow remote console. 23787 The controller implementing this feature is usually not very powerful, 23788 and we can easily swamp it with work. This is made somewhat worse by 23789 damage over-reporting the size of the dirty region, and a whole lot 23790 worse by applications (or shells) that update the screen with identical 23791 pixel content as was already there. 23792 23793 Fix this by double-buffering the shadow fb, using memcmp to identify 23794 dirty tiles on each update pass. Since both shadows are in host memory 23795 the memcmp is cheap, and worth it given the win in network bandwidth. 23796 The tile size is somewhat arbitrarily chosen to be one cacheline wide at 23797 32bpp on Intel Core. 23798 23799 By default we enable this behaviour for (a subset of) known server GPUs; 23800 the heuristic could use work. 23801 23802 Signed-off-by: Adam Jackson <ajax@redhat.com> 23803 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 23804 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 23805 23806commit 75e660e379d921a53eb7c3bc9c2e412fa58aec02 23807Author: Adam Jackson <ajax@redhat.com> 23808Date: Wed Jul 22 12:14:07 2015 -0400 23809 23810 modesetting: Drop some non-functional triple-buffering variables 23811 23812 Signed-off-by: Adam Jackson <ajax@redhat.com> 23813 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 23814 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 23815 23816commit 21217d02168d1883b2d1f64399aec494f96a8b9d 23817Author: Dave Airlie <airlied@redhat.com> 23818Date: Wed Jul 22 12:14:06 2015 -0400 23819 23820 modesetting: Implement 32->24 bpp conversion in shadow update 23821 23822 24bpp front buffers tend to be the least well tested path for client 23823 rendering. On the qemu cirrus emulation, and on some Matrox G200 server 23824 chips, the hardware can't do 32bpp at all. It's better to just allocate 23825 a 32bpp shadow and downconvert in the upload hook than expose a funky 23826 pixmap format to clients. 23827 23828 [ajax: Ported from RHEL and separate modesetting driver, lifted kbpp 23829 into the drmmode struct, cleaned up commit message, fixed 16bpp] 23830 23831 Reviewed-by: Adam Jackson <ajax@redhat.com> 23832 Signed-off-by: Dave Airlied <airlied@redhat.com> 23833 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 23834 [hdegoede@redhat.com: rebase, also use kbpp for rotate shadow fb] 23835 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 23836 23837commit af916477c65a083ec496ac3f088d766b410e8b6e 23838Author: Takashi Iwai <tiwai@suse.de> 23839Date: Mon Feb 16 17:00:55 2015 +0100 23840 23841 modesetting: Fix hw cursor check at the first call 23842 23843 With the previous patch, the modesetting driver can now return whether 23844 the driver supports hw cursor. However, it alone doesn't suffice, 23845 unfortunately. drmmode_load_cursor_argb_check() is called in the 23846 following chain: 23847 23848 xf86CursorSetCursor() 23849 -> xf86SetCursor() 23850 -> xf86DriverLoadCursorARGB() 23851 -> xf86_load_cursor_argb() 23852 -> xf86_crtc_load_cursor_argb() 23853 -> drmmode_load_cursor_argb_check() 23854 23855 *but* at first with drmmode_crtc->cursor_up = FALSE. Then the 23856 function doesn't actually set the cursor but returns TRUE 23857 unconditionally. The actual call of drmmode_set_cursor() is done at 23858 first via the show_cursor callback, and there is no check of sw cursor 23859 fallback any longer at this place. Since it's called only once per 23860 cursor setup, so the xserver still thinks as if the hw cursor is 23861 supported. 23862 23863 This patch is an ad hoc fix to correct the behavior somehow: it does 23864 call drmmode_set_cursor() at the very first time even if cursor_up is 23865 FALSE, then quickly hides again. In that way, whether the hw cursor 23866 is supported is evaluated in the right place at the right time. 23867 23868 Of course, it might be more elegant if we have a more proper mechanism 23869 to fall back to sw cursor at any call path. But it'd need more 23870 rework, so I leave this workaround as is for now. 23871 23872 Signed-off-by: Takashi Iwai <tiwai@suse.de> 23873 Reviewed-by: Adam Jackson <ajax@redhat.com> 23874 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 23875 23876commit 14c21ea1c9496638b1feb8e6145c440fb4f1d14b 23877Author: Takashi Iwai <tiwai@suse.de> 23878Date: Mon Feb 16 17:00:54 2015 +0100 23879 23880 modesetting: Use load_cursor_argb_check for sw cursor fallback 23881 23882 The modesetting driver still has an everlasting bug of invisible 23883 cursor on cirrus and other KMS drivers where no hardware cursor is 23884 supported. This patch is a part of an attempt to address it. 23885 23886 This patch particularly converts the current load_cursor_argb callback 23887 of modesetting driver to load_cursor_argb_check so that it can return 23888 whether the driver handles the hw cursor or falls back to the sw 23889 cursor. 23890 23891 Signed-off-by: Takashi Iwai <tiwai@suse.de> 23892 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 23893 [hdegoede@redhat.com: Add extra comment suggested by Kenneth] 23894 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 23895 23896commit 074cf58769b38550a3f35d5ecb73333330cb2aed 23897Author: Takashi Iwai <tiwai@suse.de> 23898Date: Mon Feb 16 17:00:53 2015 +0100 23899 23900 modesetting: Fix the error check from DRM_IOCTL_MODE_CURSOR2 23901 23902 The error value isn't always -EINVAL, e.g. the kernel drm core returns 23903 -ENXIO when the corresponding ops doesn't exist. Without this fix, 23904 DRM_IOCTL_MODE_CURSOR2 would be dealt as success even if it 23905 shouldn't. 23906 23907 Signed-off-by: Takashi Iwai <tiwai@suse.de> 23908 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 23909 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 23910 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 23911 23912commit 3762edded8e6e973f71419937e16bea262a266b8 23913Author: Adam Jackson <ajax@redhat.com> 23914Date: Wed Jun 29 14:07:32 2016 -0400 23915 23916 configure: Tell AC_REPLACE_FUNCS where to find replacements 23917 23918 Fixes weird link errors of the form: 23919 23920 CCLD Xvfb 23921 ../../Xext/.libs/libXext.a(xvmc.o): In function `xf86XvMCRegisterDRInfo': 23922 /home/ajax/git/xserver/Xext/xvmc.c:828: undefined reference to `strlcpy' 23923 /home/ajax/git/xserver/Xext/xvmc.c:829: undefined reference to `strlcpy' 23924 ../../os/os.O: In function `siHostnameAddrMatch': 23925 /home/ajax/git/xserver/os/access.c:1821: undefined reference to `strlcpy' 23926 ../../os/os.O: In function `AuthAudit': 23927 /home/ajax/git/xserver/os/connection.c:555: undefined reference to `strlcpy' 23928 /home/ajax/git/xserver/os/connection.c:574: undefined reference to `strlcpy' 23929 ../../os/os.O:/home/ajax/git/xserver/os/log.c:972: more undefined references to `strlcpy' follow 23930 collect2: error: ld returned 1 exit status 23931 Makefile:688: recipe for target 'Xvfb' failed 23932 make[3]: *** [Xvfb] Error 1 23933 Makefile:749: recipe for target 'all-recursive' failed 23934 make[2]: *** [all-recursive] Error 1 23935 Makefile:608: recipe for target 'all-recursive' failed 23936 make[1]: *** [all-recursive] Error 1 23937 Makefile:776: recipe for target 'all-recursive' failed 23938 make: *** [all-recursive] Error 1 23939 23940 Signed-off-by: Adam Jackson <ajax@redhat.com> 23941 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 23942 23943commit d135100d6b17e54262a12aeaebe0fe2fe48da3bb 23944Author: Peter Hutterer <peter.hutterer@who-t.net> 23945Date: Tue Jun 28 11:42:41 2016 +1000 23946 23947 xkb: add a cause to the xkb indicator update after a keymap change 23948 23949 Regression introduce by ac164e58870d which calls 23950 XkbUpdateAllDeviceIndicators() with two NULL arguments. A few layers down into 23951 the stack and we triggered a NULL-pointer dereference. In theory a NULL cause 23952 is acceptable since we don't actually change modifier state here. Instead of 23953 updating all places to check for NULL just set the cause to the client 23954 request and go to the pub. 23955 23956 https://bugs.freedesktop.org/show_bug.cgi?id=96384 23957 23958 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 23959 Reviewed-by: Julien Cristau <jcristau@debian.org> 23960 Reviewed-by: Daniel Stone <daniels@collabora.com> 23961 23962commit b83dede9cb930cf55249ad8e935f3c4d4328e2d9 23963Author: Alex Goins <agoins@nvidia.com> 23964Date: Thu Jun 16 20:06:56 2016 -0700 23965 23966 modesetting: Implement PRIME syncing as a source 23967 23968 Implements (Start/Stop)FlippingPixmapTracking, PresentSharedPixmap, and 23969 RequestSharedPixmapNotifyDamage, the source functions for PRIME 23970 synchronization and double buffering. Allows modesetting driver to be used 23971 as a source with PRIME synchronization. 23972 23973 v1: N/A 23974 v2: N/A 23975 v3: N/A 23976 v4: Initial commit 23977 v5: Move disabling of reverse PRIME on sink to sink commit 23978 v6: Rebase onto ToT 23979 v7: Unchanged 23980 23981 Reviewed-by: Dave Airlie <airlied@redhat.com> 23982 Signed-off-by: Alex Goins <agoins@nvidia.com> 23983 23984commit 44cb9578c0e5e10568826bc3ecbed97d358bba3c 23985Author: Alex Goins <agoins@nvidia.com> 23986Date: Thu Jun 16 20:06:55 2016 -0700 23987 23988 modesetting: Disable Reverse PRIME for i915 23989 23990 Reverse PRIME seems to be designed with discrete graphics as a sink in 23991 mind, designed to do an extra copy from sysmem to vidmem to prevent a 23992 discrete chip from needing to scan out from sysmem. 23993 23994 The criteria it used to detect this case is if we are a GPU screen and 23995 Glamor accelerated. It's possible for i915 to fulfill these conditions, 23996 despite the fact that the additional copy doesn't make sense for i915. 23997 23998 Normally, you could just set AccelMethod = none as an option for the device 23999 and call it a day. However, when running with modesetting as both the sink 24000 and the source, Glamor must be enabled. 24001 24002 Ideally, you would be able to set AccelMethod individually for devices 24003 using the same driver, but there seems to be a bug in X option parsing that 24004 makes all devices on a driver inherit the options from the first detected 24005 device. Thus, glamor needs to be enabled for all or for none until that bug 24006 (if it's even a bug) is fixed. 24007 24008 Nonetheless, it probably doesn't make sense to do the extra copy on i915 24009 even if Glamor is enabled for the device, so this is more user friendly by 24010 not requiring users to disable acceleration for i915. 24011 24012 v1: N/A 24013 v2: N/A 24014 v3: N/A 24015 v4: Initial commit 24016 v5: Unchanged 24017 v6: Rebase onto ToT 24018 v7: NULL check and free drmVersionPtr 24019 24020 Reviewed-by: Dave Airlie <airlied@redhat.com> 24021 Signed-off-by: Alex Goins <agoins@nvidia.com> 24022 24023commit f6fef2a171366156c4c6807de7fe086f04f41b7b 24024Author: Alex Goins <agoins@nvidia.com> 24025Date: Thu Jun 16 20:06:54 2016 -0700 24026 24027 modesetting: Blacklist USB transport devices from PRIME sync 24028 24029 UDL (USB 2.0 DisplayLink DRM driver) and other drivers for USB transport devices 24030 have strange semantics when it comes to vblank events, due to their inability to 24031 get the actual vblank info. 24032 24033 When doing a page flip, UDL instantly raises a vblank event without waiting for 24034 vblank. It also has no support for DRM_IOCTL_WAIT_VBLANK, and has some strange 24035 behavior with how it handles damage when page flipping. 24036 24037 It's possible to get something semi-working by hacking around these issues, 24038 but even then there isn't much value-add vs single buffered PRIME, and it 24039 reduces maintainability and adds additional risks to the modesetting driver 24040 when running with more well-behaved DRM drivers. 24041 24042 Work needs to be done on UDL in order to properly support synchronized 24043 PRIME. For now, just blacklist it, causing RandR to fall back to 24044 unsynchronized PRIME. 24045 24046 This patch originally blacklisted UDL by name, but it was pointed out that there 24047 are other USB transport device drivers with similar limitations, so it was 24048 expanded to blacklist all USB transport devices. 24049 24050 v1: N/A 24051 v2: N/A 24052 v3: Initial commit 24053 v4: Move check to driver.c for consistency/visibility 24054 v5: Refactor to accomodate earlier changes 24055 v6: Rebase onto ToT 24056 v7: Expand to blacklist all USB transport devices, not just UDL 24057 24058 Signed-off-by: Alex Goins <agoins@nvidia.com> 24059 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 24060 Reviewed-by: Dave Airlie <airlied@redhat.com> 24061 24062commit 500853086dd5fbfe6d2b3e30923fdc4d8c262cf0 24063Author: Alex Goins <agoins@nvidia.com> 24064Date: Thu Jun 16 20:06:53 2016 -0700 24065 24066 modesetting: Suspend and resume flipping with DPMS 24067 24068 DPMS would prevent page flip / vblank events from being raised, freezing 24069 the screen until PRIME flipping was reinitialized. To handle DPMS cleanly, 24070 suspend PRIME page flipping when DPMS mode is not on, and resume it when 24071 DPMS mode is on. 24072 24073 v1: Initial commit 24074 v2: Moved flipping_active check from previous commit to here 24075 v3: Unchanged 24076 v4: Unchanged 24077 v5: Move flipping_active check to sink support commit 24078 v6: Rebase onto ToT 24079 v7: Unchanged 24080 24081 Reviewed-by: Dave Airlie <airlied@redhat.com> 24082 Signed-off-by: Alex Goins <agoins@nvidia.com> 24083 24084commit 80e64dae8af1eb3bb225b00fd800c6924883cf46 24085Author: Alex Goins <agoins@nvidia.com> 24086Date: Thu Jun 16 20:06:52 2016 -0700 24087 24088 modesetting: Implement PRIME syncing as a sink 24089 24090 Implements (Enable/Disable)SharedPixmapFlipping and 24091 SharedPixmapNotifyDamage, the sink functions for PRIME synchronization and 24092 double buffering. Allows modesetting driver to be used as a sink with PRIME 24093 synchronization. 24094 24095 Changes dispatch_slave_dirty to flush damage from both scanout pixmaps. 24096 24097 Changes drmmode_set_scanout_pixmap*() functions to 24098 drmmode_set_target_scanout_pixmap*() that take an additional parameter 24099 PixmapPtr *target. Then, treat *target as it did prime_pixmap. This allows 24100 me to use it to explicitly set both prime_pixmap and prime_pixmap_back 24101 individually. drmmode_set_scanout_pixmap() without the extra parameter 24102 remains to cover the single-buffered case, but only works if we aren't 24103 already double buffered. 24104 24105 driver.c: 24106 Add plumbing for rr(Enable/Disable)SharedPixmapFlipping and 24107 SharedPixmapNotifyDamage. 24108 24109 Change dispatch_dirty_crtc to dispatch_dirty_pixmap, which functions the 24110 same but flushes damage associated with a ppriv instead of the crtc, and 24111 chanage dispatch_slave_dirty to use it on both scanout pixmaps if 24112 applicable. 24113 24114 drmmode_display.h: 24115 Add flip_seq field to msPixmapPrivRec to keep track of the event handler 24116 associated with a given pixmap, if any. 24117 24118 Add wait_for_damage field to msPixmapPrivRec to keep track if we have 24119 requested a damage notification from the source. 24120 24121 Add enable_flipping field to drmmode_crtc_private_rec to keep track if 24122 flipping is enabled or disabled. 24123 24124 Add prime_pixmap_back to drmmode_crtc_private_rec to keep track of back 24125 buffer internally. 24126 24127 Add declarations for drmmode_SetupPageFlipFence(), 24128 drmmode_EnableSharedPixmapFlipping(), 24129 drmmode_DisableSharedPixmapFlipping, drmmode_SharedPixmapFlip(), and 24130 drmmode_SharedPixmapPresentOnVBlank(). 24131 24132 Move slave damage from crtc to ppriv. 24133 24134 drmmode_display.c: 24135 Change drmmode_set_scanout_pixmap*() functions to 24136 drmmode_set_target_scanout_pixmap*() that take an additional parameter 24137 PixmapPtr *target for explicitly setting different scanout pixmaps. 24138 24139 Add definitions for functions drmmode_SharedPixmapFlip(), 24140 drmmode_SharedPixmapPresentOnVBlank(), 24141 drmmode_SharedPixmapPresent(), 24142 drmmode_SharedPixmapVBlankEventHandler(), 24143 drmmode_SharedPixmapVBlankEventAbort(), 24144 drmmode_EnableSharedPixmapFlipping(), and 24145 drmmode_DisableSharedPixmapFlipping, 24146 drmmode_InitSharedPixmapFlipping(), and 24147 drmmode_FiniSharedPixmapFlipping, along with struct 24148 vblank_event_args. 24149 24150 The control flow is as follows: 24151 pScrPriv->rrEnableSharedPixmapFlipping() makes its way to 24152 drmmode_EnableSharedPixmapFlipping(), which sets enable_flipping to 24153 TRUE and sets both scanout pixmaps prime_pixmap and 24154 prime_pixmap_back. 24155 24156 When setting a mode, if prime_pixmap is defined, modesetting 24157 driver will call drmmode_InitSharedPixmapFlipping(), which if 24158 flipping is enabled will call drmmode_SharedPixmapPresent() on 24159 scanout_pixmap_back. 24160 24161 drmmode_SharedPixmapPresent() requests that for the source to 24162 present on the given buffer using master->PresentSharedPixmap(). If 24163 it succeeds, it will then attempt to flip to that buffer using 24164 drmmode_SharedPixmapFlip(). Flipping shouldn't fail, but if it 24165 does, it will raise a warning and try drmmode_SharedPixmapPresent() 24166 again on the next vblank using 24167 drmmode_SharedPixmapPresentOnVBlank(). 24168 24169 master->PresentSharedPixmap() could fail, in most cases because 24170 there is no outstanding damage on the mscreenpix tracked by the 24171 shared pixmap. In this case, drmmode_SharedPixmapPresent() will 24172 attempt to use master->RequestSharedPixmapNotifyDamage() to request 24173 for the source driver to call slave->SharedPixmapNotifyDamage() in 24174 response to damage on mscreenpix. This will ultimately call 24175 into drmmode_SharedPixmapPresentOnVBlank() to retry 24176 drmmode_SharedPixmapPresent() on the next vblank after 24177 accumulating damage. 24178 24179 drmmode_SharedPixmapFlip() sets up page flip event handler by 24180 packing struct vblank_event_args with the necessary parameters, and 24181 registering drmmode_SharedPixmapVBlankEventHandler() and 24182 drmmode_SharedPixmapVBlankEventAbort() with the modesetting DRM 24183 event handler queue. Then, it uses the drmModePageFlip() to flip on 24184 the next vblank and raise an event. 24185 24186 drmmode_SharedPixmapPresentOnVBlank() operates similarly to 24187 drmmode_SharedPixmapFlip(), but uses drmWaitVBlank() instead of 24188 drmModePageFlip() to raise the event without flipping. 24189 24190 On the next vblank, DRM will raise an event that will ultimately be 24191 handled by drmmode_SharedPixmapVBlankEventHandler(). If we flipped, 24192 it will update prime_pixmap and prime_pixmap_back to reflect that 24193 frontTarget is now being displayed, and use 24194 drmmode_SharedPixmapPresent(backTarget) to start the process again 24195 on the now-hidden shared pixmap. If we didn't flip, it will just 24196 use drmmode_SharedPixmapPresent(frontTarget) to start the process 24197 again on the still-hidden shared pixmap. 24198 24199 Note that presentation generally happens asynchronously, so with 24200 these changes alone tearing is reduced, but we can't always 24201 guarantee that the present will finish before the flip. These 24202 changes are meant to be paired with changes to the sink DRM driver 24203 that makes flips wait on fences attached to dmabuf backed buffers. 24204 The source driver is responsible for attaching the fences and 24205 signaling them when presentation is finished. 24206 24207 Note that because presentation is requested in response to a 24208 vblank, PRIME sources will now conform to the sink's refresh rate. 24209 24210 At teardown, pScrPriv->rrDisableSharedPixmapFlipping() will be 24211 called, making its way to drmmode_FiniSharedPixmapFlipping(). 24212 There, the event handlers for prime_pixmap and prime_pixmap_back 24213 are aborted, freeing the left over parameter structure. Then, 24214 prime_pixmap and prime_pixmap back are unset as scanout pixmaps. 24215 24216 Register and tear down slave damage per-scanout pixmap instead of 24217 per-crtc. 24218 24219 v1: Initial commit 24220 v2: Renamed PresentTrackedFlippingPixmap to PresentSharedPixmap 24221 Renamed flipSeq to flip_seq 24222 Warn if flip failed 24223 Use SharedPixmapNotifyDamage to retry on next vblank after damage 24224 v3: Refactor to accomodate moving (rr)StartFlippingPixmapTracking and 24225 (rr)(Enable/Disable)SharedPixmapFlipping to rrScrPrivRec from ScreenRec 24226 Do damage tracking on both scanout pixmaps 24227 v4: Tweaks to commit message 24228 v5: Revise for internal storage of prime pixmap ptrs 24229 Move disabling for reverse PRIME from source commit to here 24230 Use drmmode_set_target_scanout_pixmap*() to set scanout pixmaps 24231 internally to EnableSharedPixmapFlipping(). 24232 Don't support flipping if ms->drmmode.pageflip == FALSE. 24233 Move flipping_active check to this commit 24234 v6: Rebase onto ToT 24235 v7: Unchanged 24236 24237 Reviewed-by: Dave Airlie <airlied@redhat.com> 24238 Signed-off-by: Alex Goins <agoins@nvidia.com> 24239 24240commit 378c85a8848679eaa3c2881f3ba8b686e59df25e 24241Author: Alex Goins <agoins@nvidia.com> 24242Date: Thu Jun 16 20:06:51 2016 -0700 24243 24244 modesetting: Always load ms->drmmode.pageflip 24245 24246 ms->drmmode.pageflip was only loaded from options if ms->drmmode.glamor was 24247 defined, otherwise it would always assume FALSE. 24248 24249 PRIME Synchronization requires ms->drmmode.pageflip even if we aren't using 24250 glamor, so load it unconditionally. 24251 24252 v1: N/A 24253 v2: N/A 24254 v3: N/A 24255 v4: N/A 24256 v5: Initial commit 24257 v6: Rebase onto ToT 24258 v7: Unchanged 24259 24260 Reviewed-by: Dave Airlie <airlied@redhat.com> 24261 Signed-off-by: Alex Goins <agoins@nvidia.com> 24262 24263commit b773a9c8126222e5fed2904d012fbf917a9f22fd 24264Author: Alex Goins <agoins@nvidia.com> 24265Date: Thu Jun 16 20:06:50 2016 -0700 24266 24267 modesetting: Always tear down scanout pixmap 24268 24269 drmmode_set_scanout_pixmap_(cpu/gpu) would only do teardown if ppix == 24270 NULL. This meant that if there were consecutive calls to 24271 SetScanoutPixmap(ppix != NULL) without calls to SetScanoutPixmap(ppix == 24272 NULL) in between, earlier calls would be leaked. RRReplaceScanoutPixmap() 24273 does this today. 24274 24275 Instead, when setting a scanout pixmap, always do teardown of the existing 24276 scanout pixmap before setting up the new one. Then, if there is no new one 24277 to set up, stop there. 24278 24279 This maintains the previous behavior in all cases except those with 24280 multiple consecutive calls to SetScanoutPixmap(ppix != NULL). 24281 24282 v1: N/A 24283 v2: N/A 24284 v3: N/A 24285 v4: N/A 24286 v5: Initial commit 24287 v6: Rebase onto ToT 24288 v7: Unchanged 24289 24290 Reviewed-by: Dave Airlie <airlied@redhat.com> 24291 Signed-off-by: Alex Goins <agoins@nvidia.com> 24292 24293commit f4c37eeee7953df1fe0e3196eda452acf0078e61 24294Author: Alex Goins <agoins@nvidia.com> 24295Date: Thu Jun 16 20:06:49 2016 -0700 24296 24297 modesetting: Internal storage of scanout pixmaps 24298 24299 modesetting relied on randr_crtc->scanout_pixmap being consistent with 24300 calls to SetScanoutPixmap, which is very fragile and makes a lot of 24301 assumptions about the caller's behavior. 24302 24303 For example, RRReplaceScanoutPixmap(), when dropping off with !size_fits, 24304 will set randr_crtc->scanout_pixmap = NULL and then call SetScanoutPixmap. 24305 Without this patch, drmmode_set_scanout_pixmap_(cpu/gpu) will think that 24306 there is no scanout pixmap to tear down, because it's already been set to 24307 NULL. 24308 24309 By keeping track of the scanout pixmap in its internal state, modesetting 24310 can avoid these types of bugs and reduce constraints on calling 24311 conventions. 24312 24313 v1: N/A 24314 v2: N/A 24315 v3: N/A 24316 v4: N/A 24317 v5: Initial commit 24318 v6: Rebase onto ToT 24319 v7: Unchanged 24320 24321 Reviewed-by: Dave Airlie <airlied@redhat.com> 24322 Signed-off-by: Alex Goins <agoins@nvidia.com> 24323 24324commit 1bdbc7e764ed7bf7c1ae46287dec368aa7c7e80d 24325Author: Alex Goins <agoins@nvidia.com> 24326Date: Thu Jun 16 20:06:47 2016 -0700 24327 24328 randr/xf86: Add PRIME Synchronization / Double Buffer 24329 24330 Changes PRIME to use double buffering and synchronization if all required 24331 driver functions are available. 24332 24333 rrcrtc.c: 24334 Changes rrSetupPixmapSharing() to use double buffering and 24335 synchronization in the case that all required driver functions are 24336 available. Otherwise, falls back to unsynchronized single buffer. 24337 24338 Changes RRCrtcDetachScanoutPixmap() to properly clean up in the case of 24339 double buffering. 24340 24341 Moves StopPixmapTracking() from rrDestroySharedPixmap() to 24342 RRCrtcDetachScanoutPixmap(). 24343 24344 Changes RRReplaceScanoutPixmap() to fail if we are using double buffering, 24345 as it would need a second ppix parameter to function with double buffering, 24346 and AFAICT no driver I've implemented double buffered source support in uses 24347 RRReplaceScanoutPixmap(). 24348 24349 randrstr.h: 24350 Adds scanout_pixmap_back to struct _rrCrtc to facilitate PRIME 24351 double buffering. 24352 24353 xf86Crtc.h: 24354 Adds current_scanout_back to _xf86Crtc to facilitate detection 24355 of changes to it in xf86RandR12CrtcSet(). 24356 24357 xf86RandR12.c: 24358 Changes xf86RandR12CrtcSet() to detect changes in 24359 scanout_pixmap_back. 24360 24361 Adds scanout_pixmap_back to struct _rrCrtc to facilitate PRIME double 24362 buffering. 24363 24364 v1: Initial commit 24365 v2: Rename PresentTrackedFlippingPixmap to PresentSharedPixmap 24366 v3: Refactor to accomodate moving (rr)StartFlippingPixmapTracking and 24367 (rr)(Enable/Disable)SharedPixmapFlipping to rrScrPrivRec from ScreenRec 24368 Add fallback if flipping funcs fail 24369 v4: Detach scanout pixmap when destroying scanout_pixmap_back, to avoid 24370 dangling pointers in some drivers 24371 v5: Disable RRReplaceScanoutPixmap for double-buffered PRIME, it would need an 24372 ABI change with support for 2 pixmaps if it were to be supported, but AFAICT 24373 no driver that actually supports double-buffered PRIME uses it. 24374 Refactor to use rrEnableSharedPixmapFlipping() as a substitute for 24375 rrCrtcSetScanoutPixmap() in the flipping case. 24376 Remove extraneous pSlaveScrPriv from DetachScanoutPixmap() 24377 Remove extraneous protopix and pScrPriv from rrSetupPixmapSharing() 24378 v6: Rebase onto ToT 24379 v7: Unchanged 24380 24381 Reviewed-by: Dave Airlie <airlied@redhat.com> 24382 Signed-off-by: Alex Goins <agoins@nvidia.com> 24383 24384commit b601f96a5915a2c486b389483b291797e6fdf617 24385Author: Alex Goins <agoins@nvidia.com> 24386Date: Thu Jun 16 20:06:46 2016 -0700 24387 24388 xf86: Add PRIME flipping functions to Screen 24389 24390 Adds typedefs for (*RRStartFlippingPixmapTrackingProcPtr), 24391 (*RREnableSharedPixmapFlippingProcPtr), 24392 and (*RRDisableSharedPixmapFlippingProcPtr) in randrstr.h. 24393 24394 Adds typedefs for (*PresentSharedPixmapProcPtr), 24395 (*SharedPixmapNotifyDamageProcPtr), 24396 (*RequestSharedPixmapNotifyDamageProcPtr), and 24397 (*StopFlippingPixmapTrackingProcPtr) in scrnintstr.h. 24398 24399 Adds RR(Enable/Disable)SharedPixmapFlipping, and 24400 RRStartFlippingPixmapTracking to rrScrnPrivRec. 24401 24402 Adds StopFlippingPixmapTracking, PresentSharedPixmap, 24403 SharedPixmapNotifyDamage, and RequestSharedPixmapNotifyDamage to ScreenRec. 24404 24405 rrScrnPrivRec used for functions that use RandR-private data types, and 24406 ScreenRec used for the rest. 24407 24408 RREnableSharedPixmapFlipping will allow the sink driver to setup for 24409 flipping between two shared pixmaps. 24410 24411 RRDisableSharedPixmapFlipping will allow the sink driver to do teardown 24412 associated with flipping between two shared pixmaps. 24413 24414 (RRStart/Stop)FlippingPixmapTracking are merely the double-buffered 24415 equivalents of (Start/Stop)PixmapTracking, allowing the source driver to do 24416 whatever setup and teardown necessary for presenting on the two shared 24417 pixmaps. 24418 24419 PresentSharedPixmap is a function exposed by the source driver for the X 24420 server or sink driver to call to request a present on a given shared 24421 pixmap. This way, presents can be driven by the sink's vblank instead of a 24422 timer or similar mechanism. 24423 24424 SharedPixmapNotifyDamage and RequestSharedPixmapNotifyDamage are OPTIONAL 24425 (even for double-buffered PRIME) functions exposed by the sink driver and 24426 the source driver, respectively. By calling 24427 master->RequestSharedPixmapNotifyDamage(ppix), the sink driver can request 24428 for the source driver to call slave->SharedPixmapNotifyDamage(ppix) in 24429 response to damage on the master screen pixmap tracked by ppix. 24430 24431 v1: Initial commit 24432 v2: Rename PresentTrackedFlippingPixmap to PresentSharedPixmap 24433 Add SharedPixmapNotifyDamage / RequestSharedPixmapNotifyDamage 24434 v3: Add RRCrtcPtr as a parameter to StartFlippingPixmapTracking 24435 Move functions that use RandR-private data types to rrScrnPrivRec. 24436 v4: Unchanged 24437 v5: Add front and back parameters to RREnableSharedPixmapFlippingProcPtr 24438 v6: Rebase onto ToT 24439 v7: Unchanged 24440 24441 Reviewed-by: Dave Airlie <airlied@redhat.com> 24442 Signed-off-by: Alex Goins <agoins@nvidia.com> 24443 24444commit ef1578e736887aadf209172f48daa9eaec25b3db 24445Author: Jon Turney <jon.turney@dronecode.org.uk> 24446Date: Sun Feb 28 00:53:24 2016 +0000 24447 24448 hw/xwin: Fix a typo in "Remove Shadow DirectDraw engine" 24449 24450 Commit 7a22912e "Remove Shadow DirectDraw engine" contained a typo, changing 24451 the fullscreen && DirectDraw check in WM_DISPLAYCHANGE to fullscreen || 24452 DirectDraw 24453 24454 This causes disruptive depth changes to be improperly handled 24455 24456 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24457 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24458 24459commit 0a0c1bd93259c208184d6ad974ba55658191af0c 24460Author: Jon Turney <jon.turney@dronecode.org.uk> 24461Date: Fri Feb 26 18:03:04 2016 +0000 24462 24463 hw/xwin: Fix a crash trying to reload window icons when not in multiwindow mode 24464 24465 ReloadEnumWindowsProc() accesses window privates, which are only valid in 24466 multiwindow mode, but is called in all modes. 24467 24468 Fix this potential crash by not doing this unless in multiwindow mode. 24469 24470 Reproduction steps: 24471 1/ XWin -mwextwm 24472 2/ Run a client which creates an X window e.g. xterm 24473 3/ Right click on notification area icon, and choose 'Reload .XWinrc' from the menu 24474 24475 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24476 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24477 24478commit 91ae2571458c50d9f782190d9f80815b770aefac 24479Author: Colin Harrison <colin.harrison@virgin.net> 24480Date: Fri Feb 26 16:46:15 2016 +0000 24481 24482 hw/xwin: Fix a crash which occurs if focus returns to XWin after xkbcomp has failed 24483 24484 If WM_FOCUS is received while the "core devices failed" fatal error (due to 24485 xkbcomp failing) is displayed, winRestoreModeKeyState() attempts to 24486 dereference a NULL InputInfo.keyboard->key pointer. 24487 24488 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 24489 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 24490 24491commit 4b123e0f61eb55e59d30bf442499cf3b3f621c3b 24492Author: Jon Turney <jon.turney@dronecode.org.uk> 24493Date: Tue Aug 4 16:05:05 2015 +0100 24494 24495 hw/xwin: Make window maximizable if a maximium size larger than virtual desktop size is specified 24496 24497 Firefox 38 has a WM_NORMAL_HINTS with a maximum size of 32767x32767. 24498 24499 Don't remove the maximize control from the window frame if the maximum size 24500 is bigger than the virtual desktop size, as maximizing the window will not 24501 exceed the maximium size. 24502 24503 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24504 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24505 24506commit 504bf495f9b3adea1ba650469223511f81709714 24507Author: Jon Turney <jon.turney@dronecode.org.uk> 24508Date: Wed Jul 8 19:12:22 2015 +0100 24509 24510 hw/xwin: Detect invalid options in combination with -nodecoration 24511 24512 Detect invalid options in combination with -nodecoration 24513 24514 These are particularly problematic as -nodecoration implies a default of 24515 -nomultimonitors, for some reason, which will gives rendering issues with 24516 -multiwindow. 24517 24518 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24519 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24520 24521commit e1b983b55e4cefcf976c8f92d608af8216a56927 24522Author: Jon Turney <jon.turney@dronecode.org.uk> 24523Date: Tue Nov 11 11:46:25 2014 +0000 24524 24525 hw/xwin: Default to -noresize when -fullscreen is used 24526 24527 Currently, just using -fullscreen fails in winValidateArgs(), as the default 24528 -resize=randr is incompatible with -fullscreen. 24529 24530 Set the default resize mode to -noresize if -fullscreen is used. 24531 24532 Also, rename enum value notAllowed -> resizeNotAllowed for clarity. 24533 24534 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24535 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24536 24537commit 42f7cd5d92f2046e1b5c264b3d76c3afda624a55 24538Author: Jon Turney <jon.turney@dronecode.org.uk> 24539Date: Sat Oct 11 14:49:16 2014 +0100 24540 24541 hw/xwin: Tell LogInit() to backup previous logfile as .old 24542 24543 Future work: Do we really need to call LogInit() in so many different 24544 places? 24545 24546 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24547 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24548 24549commit 197419838273eddbd0bd34890771799f4bebbc07 24550Author: Jon Turney <jon.turney@dronecode.org.uk> 24551Date: Fri Sep 26 13:18:33 2014 +0100 24552 24553 hw/xwin: Downgrade some uninformative, always-emitted log output to debug 24554 24555 Downgrade from error to debug some uninformative, always-emitted log output 24556 about thread synchronization during initialization 24557 24558 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24559 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24560 24561commit 17c8bf348eea4f12ce7cb4ca7db0d0576e28c982 24562Author: Jon Turney <jon.turney@dronecode.org.uk> 24563Date: Tue Apr 28 16:04:40 2015 +0100 24564 24565 hw/xwin: Check for just the hostname in window title 24566 24567 When -hostintitle is enabled, only use the hostname, not a FQDN from 24568 WM_CLIENT_MACHINE, when checking if the window title already contains it 24569 24570 Also restructure GetWindowName() to fix a potential memory leak. 24571 24572 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24573 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24574 24575commit c05c4360eea245b3ef5f3a355b95dcd63244ef70 24576Author: Jon Turney <jon.turney@dronecode.org.uk> 24577Date: Mon Sep 22 14:40:41 2014 +0100 24578 24579 hw/xwin: Use _NET_WM_NAME for window titles in multiwindow mode 24580 24581 Use _NET_WM_NAME in preference to WM_NAME for window title 24582 24583 Update window title when _NET_WM_NAME property changes 24584 24585 We should always have been doing this, but some qt5 examples only set 24586 _NET_WM_NAME, so now it's become more important... 24587 24588 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24589 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24590 24591commit 866d8299abc4315d4836292aeac4abe0b686fc10 24592Author: Jon Turney <jon.turney@dronecode.org.uk> 24593Date: Thu Mar 13 17:53:44 2014 +0000 24594 24595 hw/xwin: Remove decorations from _NET_WM_WINDOW_TYPE_SPLASH type windows 24596 24597 In multiwindow mode, remove decorations from _NET_WM_WINDOW_TYPE_SPLASH type 24598 windows. 24599 24600 Some programs use _NET_WM_WINDOW_TYPE_SPLASH_SCREEN in error, so also accept 24601 that as equivalent. 24602 24603 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24604 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24605 24606commit 356b9129067dd0e1dc62e893d47a1e9b033885bd 24607Author: Jon Turney <jon.turney@dronecode.org.uk> 24608Date: Thu Apr 7 17:40:56 2016 +0100 24609 24610 hw/xwin: Use Bool type in winShowWindowOnTaskbar() prototype 24611 24612 Use the Bool type from X11/Xdefs.h for winShowWindowOnTaskbar(). 24613 24614 This is the boolean type we should be using inside the X server, rather than 24615 BOOL, which evaluates to either the Win32 API type, or the Xlib API type, 24616 depending on the context... 24617 24618 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 24619 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 24620 24621commit 7397a2191f333a8632109873cc60d13ca4d550b0 24622Author: Olivier Fourdan <ofourdan@redhat.com> 24623Date: Tue Jun 21 13:54:35 2016 +0200 24624 24625 xwayland-input: Fake crossing to rootwin 24626 24627 This partially reverts commit c1565f3. 24628 24629 When the pointer moves from an X11 window to a Wayland native window, 24630 no LeaveNotify event is emitted which can lead to various unexpected 24631 behaviors like tooltips remaining visible after the pointer has left the 24632 window. 24633 24634 Yet the pointer_handle_leave() is called and so is the DIX CheckMotion() 24635 but since the pointer enters a Wayland native window with no other 24636 Xwayland window matching, DoEnterLeaveEvents() does not get invoked and 24637 therefore no LeaveNotify event is sent to the X11 client at the time the 24638 pointer leaves the window for a Wayland native surface. 24639 24640 Restore the XYToWindow() handler in xwayland-input that was previously 24641 removed with commit c1565f3 and use that handler to pretend that the 24642 pointer entered the root window in this case so that the LeaveNotify 24643 event is emitted. 24644 24645 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96437 24646 24647 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 24648 Acked-by: Daniel Stone <daniels@collabora.com> 24649 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 24650 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 24651 24652commit 111a045dcf0912dc5b178c18acda479b3972876b 24653Author: Adam Jackson <ajax@redhat.com> 24654Date: Tue Jun 21 11:45:17 2016 -0400 24655 24656 Revert "XDMCP: For IPv6 add link local addresses to the end of the list" 24657 24658 This reverts commit fdd448cd39b4ca84a28d73fc628911070437b703. 24659 24660commit fdd448cd39b4ca84a28d73fc628911070437b703 24661Author: Reinhard Max <max@suse.de> 24662Date: Thu Apr 28 15:21:15 2016 +0200 24663 24664 XDMCP: For IPv6 add link local addresses to the end of the list 24665 24666 For link local addresses the XDMCP server would need to either know the 24667 interface thru a scope identifier or try all available interfaces. If 24668 they don't this address will fail in which case the XDMCP server could 24669 still try the other addresses passed - however some only try the first 24670 address and then give up. 24671 24672 Even if this seems to be the wrong place to fix this it seems to be 24673 easier than fixing all display servers. 24674 24675 [ajax: Cleaned up commit message] 24676 24677 Reviewed-by: Adam Jackson <ajax@redhat.com> 24678 24679commit ce82ae1964b2115a780786676c614546297a79c9 24680Author: Adam Jackson <ajax@redhat.com> 24681Date: Tue Feb 2 14:10:49 2016 -0500 24682 24683 res: Account for GLXPixmap references too 24684 24685 GLX_EXT_tetxure_from_pixmap operates on a GLXPixmap, which takes a 24686 reference on the backing pixmap; that GLXPixmap might be long-lived, so 24687 we should account for it in ResQueryClientPixmapBytes. 24688 24689 Acked-by: Michel Dänzer <michel.daenzer@amd.com> 24690 Signed-off-by: Adam Jackson <ajax@redhat.com> 24691 24692commit 5d6ad0d3a7611c4dff2d79b6af9eb5c0885656ef 24693Author: Adam Jackson <ajax@redhat.com> 24694Date: Tue Feb 2 13:49:37 2016 -0500 24695 24696 res: Fix accounting of redirected window pixmaps for Composite 24697 24698 The previous change removed the special case that matched resources of 24699 CompositeClientWindowType and walked back from that to the window 24700 pixmap. That was intentional, since that logic was broken anyway. CCWTs 24701 don't map 1:1 to references on the backing pixmap; a window redirected 24702 by multiple clients (say, by the server since it's on the synthetic 24703 visual, and then manually by a compositor) would have a window pixmap 24704 refcount of 1, but would have those bytes accounted twice. 24705 24706 The right thing is to have Composite wrap window accounting, and add the 24707 pixmap bytes once and only once for the redirection reference. 24708 24709 Note that the view from the client can still be non-intuitive in the 24710 face of Composite. xcompmgr, for example, holds _two_ references to 24711 each window pixmap (one each from CompositeNameWindowPixmap and 24712 RenderCreatePicture), so a synthetic-visual window will have its bytes 24713 split 2/3 to xcompmgr and 1/3 to the server-client. Nothing to be done 24714 about that, and at least this way we're not over-accounting. 24715 24716 Acked-by: Michel Dänzer <michel.daenzer@amd.com> 24717 Signed-off-by: Adam Jackson <ajax@redhat.com> 24718 24719commit 4f8a72034c4a2654bfc16d929e09a69d1cc138ef 24720Author: Adam Jackson <ajax@redhat.com> 24721Date: Tue Feb 2 13:46:42 2016 -0500 24722 24723 res: Simplify QueryClientPixmapBytes 24724 24725 I suspect this code predates the common resource hooks for computing 24726 sizes. It's ugly in any case since the Resource extension shouldn't 24727 need to know which extensions can take a reference on pixmaps. Instead, 24728 let's just walk every resource for the client and sum up all the pixmap 24729 bytes that way. 24730 24731 This might be slightly slower since we're calling the size func once for 24732 every resource. On the other hand, it might be slightly faster since we 24733 only walk the resource table once instead of 3-5 times. Probably a 24734 wash, and not really a performance path in any case. 24735 24736 Acked-by: Michel Dänzer <michel.daenzer@amd.com> 24737 Signed-off-by: Adam Jackson <ajax@redhat.com> 24738 24739commit 263c5333a54107efac702a54d7e6329ae25ff4e0 24740Author: Michel Dänzer <michel.daenzer@amd.com> 24741Date: Fri Dec 25 18:32:46 2015 +0900 24742 24743 xfree86/modes: Simplify in_range logic in xf86_crtc_set_cursor_position 24744 24745 Consolidate to a single if/else statement and eliminate the redundant 24746 local variable in_range and assignments to x/y. 24747 24748 Reviewed-by: Adam Jackson <ajax@redhat.com> 24749 24750commit a991b1ec30344c16d318d2c0640f9e5d380193e0 24751Author: Michel Dänzer <michel.daenzer@amd.com> 24752Date: Mon Feb 1 16:07:22 2016 +0900 24753 24754 xfree86/modes: Disambiguate driverIsPerformingTransform 24755 24756 The driver can now specify exactly which aspects of the transform it 24757 wants to handle via XF86DriverTransform* flags. 24758 24759 Since the driver can now choose whether it wants to receive transformed 24760 or untransformed cursor coordinates, xf86CrtcTransformCursorPos no 24761 longer needs to be available to drivers, so make it static. 24762 24763 Reviewed-by: Adam Jackson <ajax@redhat.com> 24764 24765commit aad96f85005f8eab27df62049d619092865a9b16 24766Author: Michel Dänzer <michel.daenzer@amd.com> 24767Date: Fri Dec 25 18:28:47 2015 +0900 24768 24769 xfree86/modes: Fix HW cursor clipping for driverIsPerformingTransform (v2) 24770 24771 Even if the driver is handling the transform, we still need to transform 24772 the cursor position for clipping, otherwise we may hide the HW cursor 24773 when the cursor is actually inside the area covered by the CRTC. 24774 24775 v2: Use crtc_x/y local variables for clarity 24776 24777 Reviewed-by: Adam Jackson <ajax@redhat.com> 24778 24779commit 828887b6f4a997b6468da565a88b6fe9afcda191 24780Author: Keith Packard <keithp@keithp.com> 24781Date: Tue Jun 14 13:45:27 2016 -0700 24782 24783 ephyr: Process only the last expose or configure available from the server 24784 24785 Delay expose or configure processing until the event queue is empty so 24786 that we don't end up processing a long series of events one at a 24787 time. Expose events already have a check waiting for the last in a 24788 series, this further improves that by discarding multiple 24789 series of events. 24790 24791 Signed-off-by: Keith Packard <keithp@keithp.com> 24792 Reviewed-by: Adam Jackson <ajax@redhat.com> 24793 24794commit c17a41794507d6d04c850e9e1bc04fc60c31de18 24795Author: Keith Packard <keithp@keithp.com> 24796Date: Sat Jun 4 19:55:07 2016 -0700 24797 24798 ephyr: Process queued X events before blocking [v2] 24799 24800 If we end up reading all pending X events in the course of other server 24801 execution, then our notify FD callback won't get invoked and we won't 24802 process them. Fix this by noting that there are queued events in the 24803 block handler, setting the poll timeout to zero and queuing a work 24804 proc to clear the event queue. 24805 24806 v2: use a work proc to clear the event queue rather than doing it in 24807 the block handler directly. 24808 24809 Signed-off-by: Keith Packard <keithp@keithp.com> 24810 Reviewed-by: Adam Jackson <ajax@redhat.com> 24811 24812commit f3248eba6e2d0c099025f8e9a4874b431246eac7 24813Author: Keith Packard <keithp@keithp.com> 24814Date: Tue Jun 14 13:19:49 2016 -0700 24815 24816 ephyr: Handle window resize when using glamor 24817 24818 Under glamor, we need to re-create the screen pixmap at the new size 24819 so that we can ask glamor for the associated texture. Fortunately, we 24820 can simply use ephyr_glamor_create_screen_resources to create the new 24821 pixmap. 24822 24823 Because this is being done after the server has started, we need to 24824 walk the window heirarchy and reset any windows pointing at the old 24825 pixmap. I could easily be convinced that this TraverseTree should be 24826 moved to miSetScreenPixmap. 24827 24828 Signed-off-by: Keith Packard <keithp@keithp.com> 24829 Reviewed-by: Adam Jackson <ajax@redhat.com> 24830 24831commit 235d21670dcff224807ff719c7fa86212058ec46 24832Author: Keith Packard <keithp@keithp.com> 24833Date: Tue Jun 14 13:16:03 2016 -0700 24834 24835 ephyr: Don't configure window while responding to configure events 24836 24837 This leads to and endless sequence of window resizes. 24838 24839 Signed-off-by: Keith Packard <keithp@keithp.com> 24840 Reviewed-by: Adam Jackson <ajax@redhat.com> 24841 24842commit fb1edccf3c90c626f120b3c399657f24d3f7901e 24843Author: Keith Packard <keithp@keithp.com> 24844Date: Thu May 26 12:11:46 2016 -0700 24845 24846 dix: Call screen block/wakeup handlers closest to blocking [v3] 24847 24848 The screen block and wakeup handlers are the only ones which provide a 24849 well known ordering between the wrapping layers; placing these as 24850 close as possible to the server blocking provides a way for the driver 24851 to control the flow of execution correctly. 24852 24853 Switch the shadow code to run in the screen block handler so that it 24854 now occurrs just before the server goes to sleep. 24855 24856 Switch glamor to call down to the driver after it has executed its own 24857 block handler piece, in case the driver needs to perform additional 24858 flushing work after glamor has called glFlush. 24859 24860 These changes ensure that the following modules update the screen in 24861 the correct order: 24862 24863 animated cursors (uses RegisterBlockAndWakeupHandlers dynamically) 24864 composite (dynamic wrapping) 24865 misprite (dynamic wrapping) 24866 shadow (static wrapping) 24867 glamor (static wrapping) 24868 driver (static wrapping) 24869 24870 It looks like there's still a bit of confusion between composite and 24871 misprite; if composite updates after misprite, then it's possible 24872 you'd exit the block handler chain with the cursor left hidden. To fix 24873 that, misprite should be wrapping during ScreenInit time and not 24874 unwrapping. And composite might as well join in that fun, just to make 24875 things consistent. 24876 24877 [v2] Unwrap BlockHandler in shadowCloseScreen (ajax) 24878 [v3] ephyr: Use screen block handler for flushing changes 24879 24880 ephyr needs to make sure it calls glXSwapBuffers after glamor finishes 24881 its rendering. As the screen block handler is now called last, we have 24882 to use that instead of a registered block/wakeup handler to make sure 24883 the GL rendering is done before we copy it to the front buffer. 24884 24885 Signed-off-by: Keith Packard <keithp@keithp.com> 24886 Reviewed-by: Adam Jackson <ajax@redhat.com> 24887 24888commit a134d1e7eada0ac90fb622f45833c87b72c9aa06 24889Author: Guilherme Quentel Melo <gqmelo@gmail.com> 24890Date: Tue May 10 00:29:58 2016 +0000 24891 24892 glx: avoid memory leak when using indirect rendering 24893 24894 When multiple processes are using GL with indirect rendering a race 24895 condition can make drawables refcount never drop to zero. 24896 24897 This situation could happen when there are many X clients using indirect 24898 GLX: 24899 24900 1 - client1: calls glXMakeCurrent 24901 24902 2 - client2: calls glXMakeCurrent 24903 This is the first context switch for this client. So old_context_tag=0 24904 24905 3 - client1: calls glXRender 24906 For the client, its context is already current. 24907 For the server side lastGLContext points to client2's context. 24908 24909 Reviewed-by: Adam Jackson <ajax@redhat.com> 24910 Signed-off-by: Guilherme Quentel Melo <gqmelo@gmail.com> 24911 24912commit 266cf39a8f108e2f365a9772f4ee4d9fca88eb36 24913Merge: fa7b70a9b 848089e0d 24914Author: Adam Jackson <ajax@redhat.com> 24915Date: Mon Jun 20 11:21:40 2016 -0400 24916 24917 Merge remote-tracking branch 'hans/for-master' 24918 24919commit fa7b70a9b895789ec57562bbba0bc052fc91d134 24920Author: Keith Packard <keithp@keithp.com> 24921Date: Tue Jun 14 17:36:16 2016 -0700 24922 24923 kdrive: Only enable threaded input if we have input devices 24924 24925 When there aren't any devices, the input thread is going to be pretty 24926 lonely, so don't bother to even start it. 24927 24928 Signed-off-by: Keith Packard <keithp@keithp.com> 24929 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 24930 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 24931 24932commit 6f2a5b8cdf7b1e913a1e0581e65195dd10f04ca3 24933Author: Jason Gerecke <killertofu@gmail.com> 24934Date: Fri Jun 17 08:44:41 2016 -0700 24935 24936 xwayland: Expose all NBUTTONS buttons on the pointer 24937 24938 The call to 'InitButtonClassDeviceStruct' which initializes the pointer 24939 buttons only results in the first three buttons being created due to a 24940 hardcoded '3'. In order to expose all the buttons defined in the 24941 btn_labels array, we subtitute 'NBUTTONS' in its place. 24942 24943 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 24944 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 24945 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 24946 24947commit 72df6e2a3a5020696902b70fe940934ef0a681c4 24948Author: Jason Gerecke <killertofu@gmail.com> 24949Date: Fri Jun 17 08:44:40 2016 -0700 24950 24951 xwayland: Use correct labels when initializing pointer valuators 24952 24953 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 24954 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 24955 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 24956 24957commit dab5b3922c50f73712a8ea0cb9742c4bc7d7ab1d 24958Author: Jason Gerecke <killertofu@gmail.com> 24959Date: Fri Jun 17 08:44:39 2016 -0700 24960 24961 xwayland: Fix whitespace errors 24962 24963 Substitute a few errant tab characters with eight spaces to conform to the 24964 prevailing style. 24965 24966 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 24967 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 24968 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 24969 24970commit 848089e0dde38f043c85332785520946103e77c1 24971Author: Lyude Paul <cpaul@redhat.com> 24972Date: Mon Jun 13 15:31:02 2016 -0400 24973 24974 modesetting: Clear drmmode->fb_id before unflipping 24975 24976 [fix copied from 40191d82370e in xf86-video-ati] 24977 24978 Without this, we end up setting rotated CRTCs back to their previous 24979 framebuffer right after we perform a rotation. Reproducer: 24980 24981 - Have two monitors connected at the same resolution 24982 - Rotate one monitor from normal straight to inverted 24983 - Watch as the monitor you didn't rotate either freezes or shows intense 24984 flickering 24985 24986 Signed-off-by: Lyude <cpaul@redhat.com> 24987 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 24988 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 24989 24990commit 4313122dea0df9affc280ee698e929489061ccc6 24991Author: Hans de Goede <hdegoede@redhat.com> 24992Date: Wed Jun 1 15:14:32 2016 +0200 24993 24994 modesetting: Only add main fb if necessary 24995 24996 If we're doing reverse-prime; or doing rotation the main fb is not used, 24997 and there is no reason to add it in this case. 24998 24999 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 25000 Reviewed-by: Dave Airlie <airlied@redhat.com> 25001 25002commit 877453212166fdc912e0d687cdecee11aba563b5 25003Author: Hans de Goede <hdegoede@redhat.com> 25004Date: Wed Jun 1 15:11:05 2016 +0200 25005 25006 modesetting: Remove unnecessary fb addition from drmmode_xf86crtc_resize 25007 25008 drmmode_set_mode_major() is the only user of drmmode->fb_id and will 25009 create it if necessary. 25010 25011 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 25012 Reviewed-by: Dave Airlie <airlied@redhat.com> 25013 25014commit 210d83ad492f0e91889472eaae549106d9b4ebf5 25015Author: Hans de Goede <hdegoede@redhat.com> 25016Date: Wed Jun 1 15:07:22 2016 +0200 25017 25018 modesetting: Set ppix->fb_id to 0 after removing the fb 25019 25020 This ensures the fb gets re-added when a shared pixmap is re-used for 25021 a second drmmode_set_scanout_pixmap_cpu call. 25022 25023 Note currently the xserver never re-uses a shared pixmap in this way, 25024 so this is mostly a sanity fix. 25025 25026 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 25027 Reviewed-by: Dave Airlie <airlied@redhat.com> 25028 25029commit b8ef71fb07a8ba9587aeaca942b4de20b59266ca 25030Author: Hans de Goede <hdegoede@redhat.com> 25031Date: Wed Jun 1 14:59:38 2016 +0200 25032 25033 modesetting: Properly cleanup fb for reverse-prime-offload 25034 25035 drmmode_set_scanout_pixmap_gpu(pix) adds drmmod->fb_id through a call 25036 to drmmode_xf86crtc_resize(), but on a subsequent 25037 drmmode_set_scanout_pixmap_gpu(NULL) it would not remove the fb. 25038 25039 This keeps the crtc marked as busy, which causes the dgpu to not 25040 being able to runtime suspend, after an output attached to the dgpu 25041 has been used once. Which causes burning through an additional 10W 25042 of power and the laptop to run quite hot. 25043 25044 This commit adds the missing remove fb call, allowing the dgpu to runtime 25045 suspend after an external monitor has been plugged into the laptop. 25046 25047 Note this also makes drmmode_set_scanout_pixmap_gpu(NULL) match the 25048 behavior of drmmode_set_scanout_pixmap_cpu(NULL) which was already 25049 removing the fb. 25050 25051 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 25052 Reviewed-by: Dave Airlie <airlied@redhat.com> 25053 25054commit 15595d29c2c3992f260a3314f26fa691f1e5cec8 25055Author: Nikhil Mahale <nmahale@nvidia.com> 25056Date: Thu May 26 21:50:16 2016 +0530 25057 25058 randr: Adjust master's last set time with slaves 25059 25060 In prime configurations master's last set time may not be latest 25061 and greatest, adjust it with slaves last set time, pick up greatest 25062 one. Otherwise xserver may end with events which has 25063 lastSetTime < lastConfigTime even if that's not 25064 the case and confuse xrandr client. 25065 25066 [hdegoede: rebase on "xrandrprovider: Do not use separate lists for unbound / 25067 source / offload slaves"] 25068 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 25069 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 25070 25071commit 5c7af02b103790ac1fb6a71822788892c70290b6 25072Author: Hans de Goede <hdegoede@redhat.com> 25073Date: Fri May 13 15:58:10 2016 +0200 25074 25075 xrandrprovider: Do not use separate lists for unbound / source / offload slaves 25076 25077 A single provider can be both a offload and source slave at the same time, 25078 the use of seperate lists breaks in this case e.g. : 25079 25080 xrandr --listproviders 25081 Providers: number : 2 25082 Provider 0: id: 0x7b cap: 0xf, Source Output, Sink Output, Source Offload, Sink Offload crtcs: 3 outputs: 2 associated providers: 0 name:modesetting 25083 Provider 1: id: 0x46 cap: 0xf, Source Output, Sink Output, Source Offload, Sink Offload crtcs: 2 outputs: 5 associated providers: 0 name:modesetting 25084 25085 xrandr --setprovideroutputsource 1 0x7b 25086 xrandr --listproviders 25087 Providers: number : 2 25088 Provider 0: id: 0x7b cap: 0xf, Source Output, Sink Output, Source Offload, Sink Offload crtcs: 3 outputs: 2 associated providers: 1 name:modesetting 25089 Provider 1: id: 0x46 cap: 0xf, Source Output, Sink Output, Source Offload, Sink Offload crtcs: 2 outputs: 5 associated providers: 1 name:modesetting 25090 25091 xrandr --setprovideroffloadsink 1 0x7b 25092 xrandr --listproviders 25093 Providers: number : 3 25094 Provider 0: id: 0x7b cap: 0xf, Source Output, Sink Output, Source Offload, Sink Offload crtcs: 3 outputs: 2 associated providers: 2 name:modesetting 25095 Provider 1: id: 0x46 cap: 0xf, Source Output, Sink Output, Source Offload, Sink Offload crtcs: 2 outputs: 5 associated providers: 2 name:modesetting 25096 Provider 2: id: 0x46 cap: 0xf, Source Output, Sink Output, Source Offload, Sink Offload crtcs: 2 outputs: 5 associated providers: 2 name:modesetting 25097 25098 Not good. The problem is that the provider with id 0x46 now is on both 25099 the output_slave_list and the offload_slave_list of the master screen. 25100 25101 This commit fixes this by unifying all 3 lists into a single slaves list. 25102 25103 Note that this does change the struct _Screen definition, so this is an ABI 25104 break. I do not expect any of the drivers to actually use the removed / changed 25105 fields so a recompile should suffice. 25106 25107 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 25108 Reviewed-by: Dave Airlie <airlied@redhat.com> 25109 25110commit bab0f450a719a11799491043b82c2f293fed27fe 25111Author: Chris Wilson <chris@chris-wilson.co.uk> 25112Date: Sat Feb 14 09:58:44 2015 +0000 25113 25114 present: Fix presentation of flips out of order 25115 25116 The flip queue currently only holds events submitted to the driver for 25117 flipping, awaiting the completion notifier. It is short. We therefore 25118 can speed up interrupt processing by keeping the small number of events 25119 ready to be flipped on the end of the flip queue. By appending the 25120 events to the flip_queue in the order that they become ready, we also 25121 resolve one issue causing Present to display frames out of order. 25122 25123 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 25124 Reviewed-and-tested-by: Mario Kleiner <mario.kleiner.de@gmail.com> 25125 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 25126 25127commit fbe660b7652b945bdd8e3eb0780179d83cc7de96 25128Author: Chris Wilson <chris@chris-wilson.co.uk> 25129Date: Sat Feb 14 09:58:43 2015 +0000 25130 25131 present: Improve scaling of vblank handler 25132 25133 With large numbers of queued vblank, the list iteration on every 25134 interupt dominates processing time. If we reorder the list to be in 25135 ascending event order, then not only is also likely to be in order for 25136 notification queries (i.e. the notification will be near the start of 25137 the list), we can also stop iterating when past the target event_id. 25138 25139 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 25140 Reviewed-and-tested-by: Mario Kleiner <mario.kleiner.de@gmail.com> 25141 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 25142 25143commit c6511d0142040654140bdedd6f03d43af0abba21 25144Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 25145Date: Wed Oct 2 15:47:54 2013 +0200 25146 25147 exa: only draw valid trapezoids 25148 25149 Fixes freedesktop.org bug https://bugs.freedesktop.org/show_bug.cgi?id=67484 25150 25151 If t->bottom is close to MIN_INT, removing top can wraparound, so do the check properly. 25152 A similar fix should also be applied to pixman. 25153 25154 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 25155 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 25156 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 25157 25158commit 2d6230e77348e550c90e00c007b113ce38c81d12 25159Author: Keith Packard <keithp@keithp.com> 25160Date: Tue Jun 14 17:09:07 2016 -0700 25161 25162 glamor: Use DestroyPixmap instead of FreePicture for glyph atlas pixmap 25163 25164 Nice of FreePicture to take a void * instead of a PicturPtr so that 25165 this error wasn't caught by the compiler. 25166 25167 Noticed when resetting the X server left a dangling pixmap around. 25168 25169 Signed-off-by: Keith Packard <keithp@keithp.com> 25170 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 25171 25172commit edd24aa50b17849b904a3ac5429e333bac9bac9c 25173Author: Olivier Fourdan <ofourdan@redhat.com> 25174Date: Mon Jun 6 09:22:28 2016 +0200 25175 25176 wayland: clear resource for pixmap on unrealize 25177 25178 On cursor unrealize, the associated pixmap is destroyed, make sure we 25179 clear the pointer from the private resource and check for the value 25180 being non-null when setting or destroying the cursor. 25181 25182 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 25183 Reviewed-by: Adam Jackson <ajax@redhat.com> 25184 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96246 25185 25186commit 941aeb3b92e644923bd112eef8023f033a140ee6 25187Author: Olivier Fourdan <ofourdan@redhat.com> 25188Date: Fri May 13 08:58:58 2016 +0200 25189 25190 randr: Do not update ConnectionInfo if NULL 25191 25192 RRScreenSizeNotify() will update the connection information block, but 25193 if this occurs during initialization before ConnectionInfo is even 25194 initialized, this will lead to a crash. 25195 25196 Simply check for ConnectionInfo prior to update it to avoid the crash. 25197 25198 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=95337 25199 Reviewed-by: Adam Jackson <ajax@redhat.com> 25200 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 25201 25202commit 6a6bbc3b756706bdade01434f1ee8d8cbe7dd854 25203Author: Olivier Fourdan <ofourdan@redhat.com> 25204Date: Thu May 12 14:51:51 2016 +0200 25205 25206 xwayland: Restore wl_display_roundtrip() in InitInput 25207 25208 This partially revert commit 984be78 25209 25210 The rountrip in Xwayland's InitInput() is unlikely the culprit for the 25211 crash reported in bug 95337, even though it's triggered from 25212 InitInput(). 25213 25214 Startup goes like this: 25215 25216 xwl_screen_init() 25217 xwl_output_create() 25218 wl_display_roundtrip() 25219 InitInput() 25220 wl_display_roundtrip() 25221 ConnectionInfo initialized 25222 25223 What happens in bug 95337 is that some output data is already available 25224 when we reach InitInput()'s wl_display_roundtrip() and therefore we end 25225 up trying to update the ConnectionInfo's data from RR routines before 25226 ConnectionInfo is actually initialized. 25227 25228 Removing the wl_display_roundtrip() from InitInput() will not fix the 25229 issue (although it would make it less lileky to happen), because 25230 xwl_screen_init() also does a wl_display_roundtrip() after creating the 25231 output, so the race that led to bug 95337 remains. 25232 25233 However, re-setting the xwl_screen->expecting_event to 0 again in 25234 InitInput() still doesn't seem right. so this part is not restored 25235 (thus a partial revert). 25236 25237 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 25238 Reviewed-by: Adam Jackson <ajax@redhat.com> 25239 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=95337 25240 25241commit 6cb34816afa95d9214199c363f9b4bb5ecbae77b 25242Author: Adam Jackson <ajax@redhat.com> 25243Date: Fri Apr 29 14:22:52 2016 -0400 25244 25245 xace: Remove the audit hooks and tune dispatch 25246 25247 There are no in-tree consumers of the audit hooks, and they are in any 25248 case redundant with the dtrace dispatch hooks. Neither is there any 25249 in-tree user of the core request dispatch hook. The extension hook is 25250 only used for non-default security cases, but in the absence of LTO we 25251 always have to take the function call into XaceHookDispatch to find out 25252 that there's no callback registered. 25253 25254 Cc: Eamon Walsh <ewalsh@tycho.nsa.gov> 25255 Signed-off-by: Adam Jackson <ajax@redhat.com> 25256 Reviewed-by: Keith Packard <keithp@keithp.com> 25257 25258commit da9fec4eddd554b4b709ba58b4436aef5a76cd51 25259Author: Eric Biggers <ebiggers3@gmail.com> 25260Date: Thu Jun 9 19:45:15 2016 -0500 25261 25262 Allow ConstantDeceleration < 1.0 25263 25264 As documented in xorg.conf(5), a value of ConstantDeceleration between 0 25265 and 1 will speed up the pointer. However, values less than 1 actually 25266 had no effect. Fix this. 25267 25268 Note that this bug only affected "ConstantDeceleration" as configured 25269 through xorg.conf, not "Device Accel Constant Deceleration" as configured 25270 through xinput. The property handler AccelSetDecelProperty() also did 25271 not need to be changed, as it did not limit the values of the property. 25272 25273 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=92766 25274 25275 Signed-off-by: Eric Biggers <ebiggers3@gmail.com> 25276 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25277 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 25278 25279commit 7cf80b9714864c3c4eb9898ff1b7d657ecd16df7 25280Author: Adam Jackson <ajax@redhat.com> 25281Date: Wed Jun 1 14:27:23 2016 -0400 25282 25283 xfree86: Don't swallow ±iglx command line flag 25284 25285 We want to notice that it's set, but still pass it through to dix. 25286 Return 0 to indicate this. 25287 25288 Signed-off-by: Adam Jackson <ajax@redhat.com> 25289 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25290 25291commit 4653793de3860f2d83a39919af348527bc719794 25292Author: Olivier Fourdan <ofourdan@redhat.com> 25293Date: Wed Jun 8 17:10:11 2016 +0200 25294 25295 wayland: Remove unused field in xwl_screen 25296 25297 Can't find any reference of pointer_limbo_window in the code, let's 25298 remove it. 25299 25300 Reviewed-by: Adam Jackson <ajax@redhat.com> 25301 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 25302 25303commit c69bd15e00aea2610834d05f63777f9bf20eca34 25304Author: Peter Hutterer <peter.hutterer@who-t.net> 25305Date: Wed Jun 1 14:38:54 2016 +1000 25306 25307 Allow compile-time selection of a fallback input driver 25308 25309 A new --with-fallback-input-driver=foo option allows selecting a 25310 fallback driver for the server if the driver configured for the device 25311 is not found. Note that this only applies when the device has a driver 25312 assigned and that module fails to load, devices without a driver are 25313 ignored as usual. 25314 25315 This avoids the situation where a configuration assigns e.g. the 25316 synaptics driver but that driver is not available on the system, 25317 resulting in a dead device. A fallback driver can at least provides some 25318 functionality. 25319 25320 This becomes more important as we move towards making other driver true 25321 leaf nodes that can be installed/uninstalled as requested. Specifically, 25322 wacom and synaptics, a config that assigns either driver should be 25323 viable even when the driver itself is not (yet) installed on the system. 25324 25325 It is up to the distributions to make sure that the fallback driver is 25326 always installed. The fallback driver can be disabled with 25327 --without-fallback-input-driver and is disabled by default on non-Linux 25328 systems because we don't have generic drivers on those platforms. 25329 Default driver on Linux is libinput, evdev is the only other serious 25330 candidate here. 25331 25332 Sample log output: 25333 [ 3274.421] (II) config/udev: Adding input device SynPS/2 Synaptics TouchPad (/dev/input/event4) 25334 [ 3274.421] (**) SynPS/2 Synaptics TouchPad: Applying InputClass "touchpad weird driver" 25335 [ 3274.421] (II) LoadModule: "banana" 25336 [ 3274.422] (WW) Warning, couldn't open module banana 25337 [ 3274.422] (II) UnloadModule: "banana" 25338 [ 3274.422] (II) Unloading banana 25339 [ 3274.422] (EE) Failed to load module "banana" (module does not exist, 0) 25340 [ 3274.422] (EE) No input driver matching `banana' 25341 [ 3274.422] (II) Falling back to input driver `libinput' 25342 .. server proceeds to assign libinput, init the device, world peace and rainbows 25343 everywhere, truly what a sight. Shame about the banana though. 25344 25345 Reviewed-by: Adam Jackson <ajax@redhat.com> 25346 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 25347 25348commit a6b6e8ba026acedef6336b17adf06aebddd5f22f 25349Author: Dave Airlie <airlied@gmail.com> 25350Date: Fri May 6 10:46:14 2016 +1000 25351 25352 prime: clean up slave bo properly. (v3) 25353 25354 This is an ABI break, in that we now pass NULL to a function that hasn't 25355 accepted it before. 25356 25357 Alex Goins had a different patch for this but it wasn't symmetrical, it 25358 freed something in a very different place than it allocated it, this 25359 attempts to retain symmetry in the releasing of the backing bo. 25360 25361 v2: use a new toplevel API, though it still passes NULL to something 25362 that wasn't expecting it. 25363 v3: pass -1 instead of 0. 25364 25365 Signed-off-by: Dave Airlie <airlied@redhat.com> 25366 Tested-by: Hans de Goede <hdegoede@redhat.com> 25367 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 25368 Reviewed-by: Alex Goins <agoins at nvidia.com> 25369 25370commit aa5390aa83be55f887e3b4f78681b4f3d3386eb5 25371Author: Adam Jackson <ajax@redhat.com> 25372Date: Mon May 9 14:18:32 2016 -0400 25373 25374 xfree86: Remove redundant parse of AIGLX server flag 25375 25376 Not visible in the patch, but the same stanza is repeated below inside 25377 the #ifdef GLXEXT. There's no reason to bother with checking it if we 25378 built without GLXEXT so remove the unconditional one. 25379 25380 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 25381 Signed-off-by: Adam Jackson <ajax@redhat.com> 25382 25383commit 166d4c817897f5aa197f639bf06b5b3e04994496 25384Author: Michel Dänzer <michel.daenzer@amd.com> 25385Date: Fri Jun 3 10:25:20 2016 +0900 25386 25387 xwayland: Call eglBindAPI after eglInitialize 25388 25389 Current Mesa Git master checks that the EGL display actually supports 25390 the API passed to eglBindAPI, which can only succeed after 25391 eglInitialize. 25392 25393 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 25394 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 25395 25396commit d798b8318adf298bc23166e74f31e49805f0d881 25397Author: Michel Dänzer <michel.daenzer@amd.com> 25398Date: Fri Jun 3 10:25:19 2016 +0900 25399 25400 glamor: Call eglBindAPI after eglInitialize 25401 25402 Current Mesa Git master checks that the EGL display actually supports 25403 the API passed to eglBindAPI, which can only succeed after 25404 eglInitialize. 25405 25406 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96344 25407 Reviewed-by: Adam Jackson <ajax@redhat.com> 25408 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 25409 25410commit 5bcdd1cc753d8e60c249eeacba695653af1a39d8 25411Author: Adam Jackson <ajax@redhat.com> 25412Date: Tue Jun 7 15:52:15 2016 -0400 25413 25414 dix: Update some comments to reflect the new non-SIGIO input model 25415 25416 Signed-off-by: Adam Jackson <ajax@redhat.com> 25417 Reviewed-by: Keith Packard <keithp@keithp.com> 25418 25419commit 48a9b29b0a09b865492d60e8a7cd2a94864fbb3f 25420Author: Adam Jackson <ajax@redhat.com> 25421Date: Tue Jun 7 15:52:14 2016 -0400 25422 25423 xfree86: Undocument UseSIGIO in xorg.conf 25424 25425 The doc text is wrong at this point, input processing isn't going to 25426 vary based on this, so we shouldn't say it does. The only thing this 25427 _does_ get used for is DRI1 SwapBuffers (on everything but savage), and 25428 if you disable it you're not going to get DRI1 at all, so we really 25429 shouldn't even mention it. 25430 25431 Still, leave the option wired up to the parser so we don't break any 25432 DRI1-driver-using setup relying on it being disabled, and so we don't 25433 complain about unused options elsewhere. 25434 25435 Signed-off-by: Adam Jackson <ajax@redhat.com> 25436 Acked-by: Keith Packard <keithp@keithp.com> 25437 25438commit 2725dd024de33e96555cbf6ff28184a64d515c3c 25439Author: Adam Jackson <ajax@redhat.com> 25440Date: Tue Jun 7 15:52:13 2016 -0400 25441 25442 dri1: Hide the SIGIO details from drivers 25443 25444 Not being used, and not likely to be useful. 25445 25446 Signed-off-by: Adam Jackson <ajax@redhat.com> 25447 Acked-by: Keith Packard <keithp@keithp.com> 25448 25449commit 95ce2bccdcdc88355262674a636c342c946914da 25450Author: Adam Jackson <ajax@redhat.com> 25451Date: Tue Jun 7 15:52:12 2016 -0400 25452 25453 dmx: Remove SIGIO input support here too 25454 25455 This code was broken anyway. Note that DEVICE_OFF would make dmx think 25456 _no_ devices were using SIGIO anymore, which means 'xinput disable' on 25457 your mouse would probably do weird things to your keyboard too. Rather 25458 than try to repair that and keep SIGIO working on this one niche DDX, 25459 just rip it out and use the thread model like everyone else. 25460 25461 Signed-off-by: Adam Jackson <ajax@redhat.com> 25462 Acked-by: Keith Packard <keithp@keithp.com> 25463 25464commit 6178b1c91cfc9e860914acc6f0be2f2d2e07a124 25465Author: Adam Jackson <ajax@redhat.com> 25466Date: Tue Jun 7 15:52:11 2016 -0400 25467 25468 dix: Use OsSignal() not signal() 25469 25470 As the man page for the latter states: 25471 25472 The effects of signal() in a multithreaded process are unspecified. 25473 25474 We already have an interface to call sigaction() instead, use it. 25475 25476 Signed-off-by: Adam Jackson <ajax@redhat.com> 25477 Reviewed-by: Keith Packard <keithp@keithp.com> 25478 25479commit 8174daa6bd3f0c792425a5ebef63a6a9ce7d00a4 25480Author: Keith Packard <keithp@keithp.com> 25481Date: Sun Jun 5 12:32:19 2016 -0700 25482 25483 os: Do timers under input lock, not blocked signals 25484 25485 Timer processing can happen on either the main thread or the input 25486 thread. As a result, it must be done under the input lock. 25487 25488 Signals are unrelated to timers now that SIGIO isn't used for input 25489 processing, so stop blocking signals while processing timers. 25490 25491 Signed-off-by: Keith Packard <keithp@keithp.com> 25492 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25493 25494commit 88e981e7088198fabea6c322c58f371d91578b6a 25495Author: Olivier Fourdan <ofourdan@redhat.com> 25496Date: Wed Mar 9 10:33:50 2016 +0100 25497 25498 xwayland: sync event queue to check compositor reply 25499 25500 Read and dispatch pending Wayland events to make sure we do not miss a 25501 possible reply from the compositor prior to discard a key repeat. 25502 25503 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 25504 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25505 25506commit 26ad25a0ed1d99f3cacb711793c830cb5249580e 25507Author: Olivier Fourdan <ofourdan@redhat.com> 25508Date: Wed Mar 9 11:17:27 2016 +0100 25509 25510 xwayland: refactor Wayland event handling 25511 25512 To be able to reuse some code. 25513 25514 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 25515 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25516 25517commit 239705a6feefaddf90be9541a054ebd8c8c10d81 25518Author: Olivier Fourdan <ofourdan@redhat.com> 25519Date: Wed Mar 9 10:31:58 2016 +0100 25520 25521 xwayland: add a server sync before repeating keys 25522 25523 Key repeat is handled by the X server, but input events need to be 25524 processed and forwarded by the Wayland compositor first. 25525 25526 Make sure the Wayland compositor is actually processing events, to 25527 avoid repeating keys in Xwayland while the Wayland compositor cannot 25528 deal with input events for whatever reason, thus not dispatching key 25529 release events, leading to repeated keys while the user has already 25530 released the key. 25531 25532 Bugzilla: https://bugzilla.gnome.org/show_bug.cgi?id=762618 25533 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 25534 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25535 25536commit fda5675f9d257d583ea0683adf863c5e1e51e30c 25537Author: Olivier Fourdan <ofourdan@redhat.com> 25538Date: Wed Mar 9 10:31:13 2016 +0100 25539 25540 xkb: add hook to allow/deny AccessX key repeat 25541 25542 The xserver generates the key repeat by itself. 25543 25544 But when used with another server processing inputs first (e.g. a 25545 Wayland compositor), the other server may be busy dealing with some 25546 other things and not queue up key release events in time. 25547 25548 Add a vfunc in XkbSrvInfo to possibly add a check before re-emitting a 25549 keypress event in the AccessX timer handler, so that the key repeat has 25550 a chance to be denied if the server processing the input is not ready. 25551 25552 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 25553 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25554 25555commit 3735ab965a4642273d070840bda4e827991e5219 25556Merge: 7c77c42fe 43dbc556f 25557Author: Keith Packard <keithp@keithp.com> 25558Date: Thu Jun 2 07:47:16 2016 -0700 25559 25560 Merge remote-tracking branch 'daenzer/for-master' 25561 25562commit 7c77c42fe87316e5c68611f03b0bda96656c4a4c 25563Author: Keith Packard <keithp@keithp.com> 25564Date: Fri May 27 17:36:57 2016 -0700 25565 25566 dix: Don't update current time in the middle of input event processing 25567 25568 In patch 137ac094e7ab8c871f3b36e40ad826ac797f0e26, Adam moved an 25569 expensive call to UpdateCurrentTime out of the main dispatch 25570 loop. That's a good change as the original fix from Chase was a bit 25571 expensive. However, it breaks grab processing and so a couple of the 25572 calls to UpdateCurrenTime need to be removed. 25573 25574 Input event processing can generate a stream of events; a button press 25575 that activates a grab will send a press followed by a sequence of 25576 enter/leave events. All of these should have the same time stamp on 25577 the wire as they occur at the 'same' time. 25578 25579 More importantly, the grab time recorded in the device is pulled from 25580 currentTime after all of the events are delivered, so if currentTime 25581 doesn't match the time in the device event, then future grab 25582 modifications will fail as the time marked in the device will be 25583 'later' than the grab time known to the client (which is defined as 25584 the timestamp from the activating input event). 25585 25586 A bit of history here -- it used to be that currentTime was driven 25587 *entirely* by input events; those timestamps didn't even have to be 25588 related to the system time in any way. Then we started doing ICCCM 25589 stuff and people got confused when PropertyNotify events would have 25590 the same timestamp even when delivered minutes apart because no input 25591 events were delivered. 25592 25593 We added code in the server to go update the time, but only if no 25594 input events were pending (so that the clock "wouldn't" go 25595 backwards). The only places where this is necessary is in request 25596 processing which may generate an event with a timestamp, and there 25597 only at the very top of the request processing code so that the whole 25598 request would be processed at the 'same time', just like events. 25599 25600 cc: Chase Douglas <chase.douglas@canonical.com> 25601 cc: Peter Hutterer <peter.hutterer@who-t.net> 25602 cc: Adam Jackson <ajax@redhat.com> 25603 Signed-off-by: Keith Packard <keithp@keithp.com> 25604 Tested-by: Michel Dänzer <michel.daenzer@amd.com> 25605 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 25606 25607commit ce6546337487c052b5dd3c04d3d8d4b09d691c3d 25608Author: Keith Packard <keithp@keithp.com> 25609Date: Tue May 31 09:38:17 2016 -0700 25610 25611 os: Initialize NotifyFds earlier in startup 25612 25613 If the server calls AbortServer during the first-time initialization 25614 (which can happen if you start the server on an already using 25615 DISPLAY), then the dbus code will shut down and call the notify fd 25616 interface. If the notify fd list hasn't been initialized, the server 25617 will crash. 25618 25619 Signed-off-by: Keith Packard <keithp@keithp.com> 25620 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25621 25622commit f0756793e4c30278164d7a5cc483ce6a311c58dc 25623Author: Keith Packard <keithp@keithp.com> 25624Date: Tue May 31 09:14:17 2016 -0700 25625 25626 os: Lock input while messing with input device list 25627 25628 The list of input devices may be changed by hotplugging while the 25629 server is active, and those changes may come from either the main 25630 thread or the input thread. That means the list of input devices needs 25631 to be protected by a mutex. 25632 25633 This prevents input drivers from receiving I/O ready callbacks after 25634 removing a device. 25635 25636 Signed-off-by: Keith Packard <keithp@keithp.com> 25637 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25638 25639commit a779fda224bee0c4d27636503367e55ae93b33c2 25640Author: Jonas Ådahl <jadahl@gmail.com> 25641Date: Tue May 24 18:46:42 2016 +0800 25642 25643 xwayland: Use the CLOCK_MONOTONIC clock 25644 25645 By default the X server will try CLOCK_MONOTONIC_COARSE before 25646 CLOCK_MONOTONIC, while A Wayland compositor may only support getting 25647 their timestamps from the CLOCK_MONOTONIC clock. This causes various 25648 issues since it may happen that a timestamp from CLOCK_MONOTONIC 25649 retrieved before a sending an X request will still be "later" than the 25650 timestamp the X server than gets after receiving the request, due to the 25651 fact that CLOCK_MONOTONIC_COARSE has a lower resolution. 25652 25653 To avoid these issues, make Xwayland always use CLOCK_MONOTONIC, so 25654 that it becomes possible for Wayland compositor only supporting 25655 CLOCK_MONOTONIC and X server to use the same clock. 25656 25657 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 25658 Acked-by: Daniel Stone <daniels@collabora.com> 25659 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 25660 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 25661 25662commit 43dbc556f3a4d743b9121d6cfc21961be4a9da56 25663Author: Michel Dänzer <michel.daenzer@amd.com> 25664Date: Thu Dec 24 16:43:44 2015 +0900 25665 25666 xfree86/modes: Remove xf86_reload_cursors v2 25667 25668 No longer needed now that xf86CursorResetCursor is getting called for 25669 each CRTC configuration change. 25670 25671 v2: Keep xf86_reload_cursors as a deprecated empty inline function 25672 until all drivers stop calling it. (Adam Jackson) 25673 25674 Reviewed-by: Adam Jackson <ajax@redhat.com> 25675 25676commit e156c0ccb530897d3a428255bd5585f7ea7b9b41 25677Author: Michel Dänzer <michel.daenzer@amd.com> 25678Date: Wed Mar 30 18:23:04 2016 +0900 25679 25680 os: Use strtok instead of xstrtokenize in ComputeLocalClient 25681 25682 Fixes leaking the memory pointed to by the members of the array returned 25683 by xstrtokenize. 25684 25685 Reviewed-by: Adam Jackson <ajax@redhat.com> 25686 25687commit 5ff75da317539e87cca429185d710d0eeb9d9222 25688Author: Michel Dänzer <michel.daenzer@amd.com> 25689Date: Thu May 26 19:01:59 2016 +0900 25690 25691 glamor: Cannot use copies when accessing outside of composite source 25692 25693 Commit b64108fa ("glamor: Check for composite operations which are 25694 equivalent to copies") failed to copy conditions from exaComposite which 25695 ensure that the composite operation doesn't access outside of the source 25696 picture. 25697 25698 This fixes rendercheck regressions from the commit above. 25699 25700 Reviewed-by: Keith Packard <keithp@keithp.com> 25701 25702commit 84e0d5d63c393fc36e91433cd7897e776e82528c 25703Author: Michel Dänzer <michel.daenzer@amd.com> 25704Date: Fri May 27 16:58:19 2016 +0900 25705 25706 xfree86/modes: Assign xf86_config->cursor in xf86_load_cursor_image v2 25707 25708 Fixes a crash on startup in the radeon driver's drmmode_show_cursor() 25709 due to xf86_config->cursor == NULL, because no CRTC was enabled yet, so 25710 xf86_crtc_load_cursor_image was never called. 25711 25712 (Also use scrn->pScreen instead of xf86ScrnToScreen(scrn)) 25713 25714 v2: Set xf86_config->cursor at the beginning of xf86_load_cursor_image 25715 instead of at the end. 25716 25717 Reviewed-by: Keith Packard <keithp@keithp.com> 25718 25719commit 28b2c880a9dbfd91e82397a0f22482d258a8b670 25720Author: Keith Packard <keithp@keithp.com> 25721Date: Sun May 29 19:11:50 2016 -0700 25722 25723 xfree86: Provide xf86BlockSIGIO and xf86ReleaseSIGIO as wrappers for input mutex 25724 25725 Threaded input doesn't use SIGIO anymore, but existing drivers using 25726 xf86BlockSIGIO and xf86ReleaseSIGIO probably want to lock the input 25727 mutex during those operations. Provide inline functions to do this 25728 which are marked as 'deprecated' so that drivers will get warnings 25729 until they are changed. 25730 25731 Signed-off-by: Keith Packard <keithp@keithp.com> 25732 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25733 25734commit 2bed654be7001407ae65f7128440340fd3c1f083 25735Author: Keith Packard <keithp@keithp.com> 25736Date: Sun May 29 19:05:13 2016 -0700 25737 25738 dix: Remove bogus input_mutex and input_mutex_count definitions from globals.c 25739 25740 These are local to inputthread.c; the ones in globals.c aren't 25741 declared in any header file. 25742 25743 Signed-off-by: Keith Packard <keithp@keithp.com> 25744 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25745 25746commit 06bdc3bc1aeca2635d0f711ac7eef34796e28339 25747Author: Keith Packard <keithp@keithp.com> 25748Date: Sun May 29 01:14:59 2016 -0700 25749 25750 os: fix input_mutex_count off-by-one in input_force_unlock 25751 25752 input_force_unlock was mis-using input_mutex_lock and leaving it set 25753 to -1. As this is executed from OsInit at each server generation, on 25754 the second time through, the mutex would be left locked (!) due to the 25755 trylock call. This caused input to fail after the first server reset. 25756 25757 Signed-off-by: Keith Packard <keithp@keithp.com> 25758 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25759 25760commit de3620065945f199a203afb831cac99bb8bff75b 25761Author: Keith Packard <keithp@keithp.com> 25762Date: Fri May 27 17:51:01 2016 -0700 25763 25764 xfree86: Remove event reading code from xf86Wakeup 25765 25766 Oops. This didn't get removed when xfree86 was converted over to use 25767 the input thread. 25768 25769 Signed-off-by: Keith Packard <keithp@keithp.com> 25770 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25771 25772commit c9703007ea77a8aa5aba4e7bc7367814526056fe 25773Author: Keith Packard <keithp@keithp.com> 25774Date: Sat May 28 10:34:07 2016 -0700 25775 25776 Add ax_pthread.m4 to m4/ 25777 25778 Threaded input requires AX_PTHREAD, which is provided by this file. 25779 25780 Signed-off-by: Keith Packard <keithp@keithp.com> 25781 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 25782 25783commit c07b796b47ae7f5a616d5e8e02ab0f84dc438a93 25784Author: Keith Packard <keithp@keithp.com> 25785Date: Sun May 29 18:38:18 2016 -0700 25786 25787 xfree86: Bump ABI versions 25788 25789 threaded input can affect drivers that use OsBlockSIGIO when dealing 25790 with cursors. 25791 25792 Signed-off-by: Keith Packard <keithp@keithp.com> 25793 Requested-by: Peter Hutterer <peter.hutterer@who-t.net> 25794 25795commit 24e4c6db25aef7a057f3f384f1e2ce67e66fa40e 25796Author: Keith Packard <keithp@keithp.com> 25797Date: Sun May 29 18:43:00 2016 -0700 25798 25799 test: Actually verify that two equivalent touch points are the same 25800 25801 I typo'd when fixing this at Peter's request and left off the '2' from 25802 the second variable name. 25803 25804 Signed-off-by: Keith Packard <keithp@keithp.com> 25805 25806commit 1338bfa81c6eddc66f07c15225c3feff062182ce 25807Author: Keith Packard <keithp@keithp.com> 25808Date: Fri May 27 01:56:39 2016 -0700 25809 25810 test: Make touch test reflect new ability to realloc touch array [v2] 25811 25812 Threaded input allows the input code to call malloc while processing 25813 events. In this case, that's in the middle of processing touch events 25814 and needing to resize the touch buffer. 25815 25816 This test was expecting the old behaviour where touch points would get 25817 dropped if the buffer was full. The fix is to check for the new 25818 behaviour instead. 25819 25820 [v2] 25821 25822 * make sure two finding two equivalent touches return the same touch 25823 object 25824 25825 * check to make sure the queue resizes by the expected amount 25826 25827 Changes provided by Peter Hutterer <peter.hutterer@who-t.net> 25828 25829 Signed-off-by: Keith Packard <keithp@keithp.com> 25830 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25831 25832commit 8b9b4387e8473810f6174519ee76818fcaae725d 25833Author: Keith Packard <keithp@keithp.com> 25834Date: Fri May 27 10:55:02 2016 -0700 25835 25836 glamor: Adjust for drawable x/y in composite's copy optimization 25837 25838 Patch b64108fa305e956e4edaae9d53071ff0abee268e added a short cut that 25839 identifies composite operations that can be performed with a simple 25840 copy instead. 25841 25842 glamor_copy works in absolute coordinates, so the dx and dy values 25843 passed in need to be converted from drawable-relative to absolute by 25844 adding the drawable x/y values. 25845 25846 Signed-off-by: Keith Packard <keithp@keithp.com> 25847 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 25848 25849commit 0d16a0c3b95d4ac446a07d8b95e634de5cb8055f 25850Author: Keith Packard <keithp@keithp.com> 25851Date: Thu May 26 10:22:17 2016 -0700 25852 25853 os: Increase default client buffer to 16kB 25854 25855 This matches a change made in xcb and improves performance for a small 25856 increase in memory usage. 25857 25858 Signed-off-by: Keith Packard <keithp@keithp.com> 25859 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 25860 25861commit 714736124f2bfba9ede076eff3bf017f986ce6c4 25862Author: Keith Packard <keithp@keithp.com> 25863Date: Fri May 27 02:02:48 2016 -0700 25864 25865 modesetting: Use new xf86CurrentCursor API 25866 25867 Use this instead of the (now deprecated) cursor pointer in the 25868 xf86CrtcConfigRec. 25869 25870 Signed-off-by: Keith Packard <keithp@keithp.com> 25871 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 25872 25873commit e69061e605c66084056cf2064dcc5d4c6c8aa0cf 25874Author: Keith Packard <keithp@keithp.com> 25875Date: Tue Dec 8 14:26:50 2015 -0800 25876 25877 kdrive: Use threaded input 25878 25879 Signed-off-by: Keith Packard <keithp@keithp.com> 25880 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 25881 25882commit a977c9c4d07a04c0c7b3713a21eefb92a8d3005b 25883Author: Keith Packard <keithp@keithp.com> 25884Date: Tue Dec 8 14:53:08 2015 -0800 25885 25886 xfree86: Use threaded input mechanism [v2] 25887 25888 Switch the XFree86 DDX over to threaded input 25889 25890 v2: Rewrite comment in xf86Helper about silken mouse 25891 25892 Signed-off-by: Keith Packard <keithp@keithp.com> 25893 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25894 25895commit f84703b50cc908a127f4ad923ebbf56f8f244c0d 25896Author: Keith Packard <keithp@keithp.com> 25897Date: Tue Dec 8 14:20:21 2015 -0800 25898 25899 dix: Reallocate touchpoint buffer at input event time [v2] 25900 25901 Now that input is threaded, malloc can be used at event time to resize 25902 the touchpoint buffer as needed.x 25903 25904 v2: Remove "Need to grow the queue means dropping events." 25905 from comment as it no longer applies. (Peter Hutterer) 25906 25907 Signed-off-by: Keith Packard <keithp@keithp.com> 25908 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25909 25910commit e2df803fcabd7c2d2fc7991c794856378dcb5489 25911Author: Keith Packard <keithp@keithp.com> 25912Date: Tue Dec 8 19:53:36 2015 -0800 25913 25914 mi: Grow event queue while reading events 25915 25916 Now that events are read at normal process time, we can use malloc to 25917 grow the event queue instead of discarding events. 25918 25919 Signed-off-by: Keith Packard <keithp@keithp.com> 25920 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25921 25922commit 30ac7567980a1eb79d084a63e0e74e1d9a3af673 25923Author: Keith Packard <keithp@keithp.com> 25924Date: Tue Dec 8 11:37:51 2015 -0800 25925 25926 Create a threaded mechanism for input [v7] 25927 25928 The current SIGIO signal handler method, used at generation of input events, 25929 has a bunch of oddities. This patch introduces an alternative way using a 25930 thread, which is used to select() all input device file descriptors. 25931 25932 A mutex was used to control the access to input structures by the main and input 25933 threads. Two pipes to emit alert events (such hotplug ones) and guarantee the 25934 proper communication between them was also used. 25935 25936 Co-authored-by: Fernando Carrijo <fcarrijo@freedesktop.org> 25937 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 25938 25939 v2: Fix non-Xorg link. Enable where supported by default. 25940 25941 This also splits out the actual enabling of input threads to 25942 DDX-specific patches which follow 25943 25944 v3: Make the input lock recursive 25945 25946 v4: Use regular RECURSIVE_MUTEXes instead of rolling our own 25947 Respect the --disable-input-thread configuration option by 25948 providing stubs that expose the same API/ABI. 25949 25950 Respond to style comments from Peter Hutterer. 25951 25952 v5: use __func__ in inputthread debug and error mesages. 25953 25954 Respond to style comments from Peter Hutterer. 25955 25956 v6: use AX_PTHREAD instead of inlining pthread tests. 25957 25958 Suggested by Emil Velikov <emil.l.velikov@gmail.com> 25959 25960 v7: Use pthread_sigmask instead of sigprocmask when using threads 25961 25962 Suggested by Adam Jackson <ajax@redhat.com> 25963 25964 Signed-off-by: Adam Jackson <ajax@redhat.com> 25965 Signed-off-by: Keith Packard <keithp@keithp.com> 25966 Reviewed-by: Adam Jackson <ajax@redhat.com> 25967 25968commit 728c9570a05f03bd90343ff6f5b1a8fd3988864c 25969Author: Keith Packard <keithp@keithp.com> 25970Date: Wed Dec 9 14:27:05 2015 -0800 25971 25972 xkb: Hold input lock across injected key event processing 25973 25974 This makes the code more consistent with other versions of 25975 out-of-queue event processing 25976 25977 Signed-off-by: Keith Packard <keithp@keithp.com> 25978 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25979 25980commit 05d549d60440126105d5a70272e72ea2319724bd 25981Author: Keith Packard <keithp@keithp.com> 25982Date: Thu Dec 17 22:21:46 2015 -0800 25983 25984 xfree86: Remove unnecessary errno save/restore in xf86ReadInput 25985 25986 When this code was called from SIGIO, saving and restoring errno could 25987 possibly have made sense in some strange environment. Now that this 25988 will not be called from a signal handler, there is no reason to do that. 25989 25990 Signed-off-by: Keith Packard <keithp@keithp.com> 25991 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 25992 25993commit 0bbb5aabf794720fa99ae5ea3f55138ded159705 25994Author: Keith Packard <keithp@keithp.com> 25995Date: Mon May 16 21:33:41 2016 -0500 25996 25997 kdrive: Don't lock input across read in KdNotifyFd 25998 25999 We won't need these locks with the new threaded input code as it holds 26000 the input lock across all of the input device I/O operations. 26001 26002 Signed-off-by: Keith Packard <keithp@keithp.com> 26003 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 26004 26005commit 6a5a4e60373c1386b311b2a8bb666c32d68a9d99 26006Author: Keith Packard <keithp@keithp.com> 26007Date: Tue Dec 8 14:39:46 2015 -0800 26008 26009 Remove SIGIO support for input [v5] 26010 26011 This removes all of the SIGIO handling support used for input 26012 throughout the X server, preparing the way for using threads for input 26013 handling instead. 26014 26015 Places calling OsBlockSIGIO and OsReleaseSIGIO are marked with calls 26016 to stub functions input_lock/input_unlock so that we don't lose this 26017 information. 26018 26019 xfree86 SIGIO support is reworked to use internal versions of 26020 OsBlockSIGIO and OsReleaseSIGIO. 26021 26022 v2: Don't change locking order (Peter Hutterer) 26023 v3: Comment weird && FALSE in xf86Helper.c 26024 Leave errno save/restore in xf86ReadInput 26025 Squash with stub adding patch (Peter Hutterer) 26026 v4: Leave UseSIGIO config parameter so that 26027 existing config files don't break (Peter Hutterer) 26028 v5: Split a couple of independent patch bits out 26029 of kinput.c (Peter Hutterer) 26030 26031 Signed-off-by: Keith Packard <keithp@keithp.com> 26032 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 26033 26034commit 8cf832c288dec13cbf3c25478a8ccef52d61f3db 26035Author: Keith Packard <keithp@keithp.com> 26036Date: Mon May 16 21:28:48 2016 -0500 26037 26038 kdrive: Remove unneeded AddEnabledDevice/RemoveEnabledDevice calls 26039 26040 kdrive uses the NotifyFd interface, which handles all of the necessary 26041 fd configuration in the OS layer. Having it also use the old 26042 EnableDevice interfaces is incorrect. 26043 26044 Signed-off-by: Keith Packard <keithp@keithp.com> 26045 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 26046 26047commit 4fca18dc03669bada75175df289fbee41275208e 26048Author: Adam Jackson <ajax@redhat.com> 26049Date: Mon May 9 13:28:08 2016 -0400 26050 26051 xfree86: Add IndirectGLX server flag (v2) 26052 26053 Not all display managers make it easy (or possible) to modify the 26054 command line flags passed to the server, so add a way to get to it from 26055 xorg.conf. 26056 26057 v2: Fix the FlagOptions list to not have IGLX after the terminator (Alan 26058 Coopersmith) 26059 26060 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 26061 Signed-off-by: Adam Jackson <ajax@redhat.com> 26062 26063commit 181a4bd0cc436f89582408196038ff37032f9bac 26064Author: Keith Packard <keithp@keithp.com> 26065Date: Fri May 13 16:19:38 2016 -0700 26066 26067 glamor: Preserve GL_RED bits in R channel when destination is GL_RED [v2] 26068 26069 A1 and A8 pixmaps are usually stored in the Red channel to conform 26070 with more recent GL versions. When using these pixmaps as mask values, 26071 that works great. When using these pixmaps as source values, then the 26072 value we want depends on what the destination looks like. 26073 26074 For RGBA or RGB destinations, then we want to use the Red channel 26075 for A values and leave RGB all set to zero. 26076 26077 For A destinations, then we want to leave the R values in the Red 26078 channel so that they end up in the Red channel of the output. 26079 26080 This patch adds a helper function, glamor_bind_texture, which performs 26081 the glBindTexture call along with setting the swizzle parameter 26082 correctly for the Red channel. The swizzle parameter for the Alpha 26083 channel doesn't depend on the destination as it's safe to leave it 26084 always swizzled from the Red channel. 26085 26086 This fixes incorrect rendering in firefox for this page: 26087 26088 https://gfycat.com/HoarseCheapAmericankestrel 26089 26090 while not breaking rendering for this page: 26091 26092 https://feedly.com 26093 26094 v2: Add change accidentally left in patch for missing 26095 glDisable(GL_COLOR_LOGIC_OP). 26096 Found by Emil Velikov <emil.l.velikov@gmail.com> 26097 26098 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=63397 26099 Signed-off-by: Keith Packard <keithp@keithp.com> 26100 Tested-by: Michel Dänzer <michel.daenzer@amd.com> 26101 26102commit b07bc700b3cf2f5c8912fc5b9e0dad2baf395525 26103Author: Keith Packard <keithp@keithp.com> 26104Date: Sat May 14 08:22:17 2016 -0700 26105 26106 glamor: glamor_make_current sooner in glamor_composite_with_shader 26107 26108 glamor_make_current is supposed to be called before any GL APIs. 26109 26110 Signed-off-by: Keith Packard <keithp@keithp.com> 26111 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26112 26113commit 743b6f231e999d8b2909228412266dc13cc433c5 26114Author: Keith Packard <keithp@keithp.com> 26115Date: Fri May 13 04:25:43 2016 -0700 26116 26117 glamor: Disable logic ops when doing compositing [v4] 26118 26119 If the logic op gets left enabled, it overrides the blending 26120 operation, causing incorrect contents on the display. 26121 26122 v2: Disable only on non-ES2, but disable even for PictOpSrc 26123 26124 v3: Found another place this is needed in 26125 glamor_composite_set_shader_blend 26126 26127 v4: Remove change dependent on new glamor_set_composite_texture 26128 API. This belongs in a different patch. 26129 Found by Emil Velikov <emil.l.velikov@gmail.com> 26130 26131 Signed-off-by: Keith Packard <keithp@keithp.com> 26132 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26133 26134commit f5670b4a7eb3b43fa20017f89192394930eae6f4 26135Author: Keith Packard <keithp@keithp.com> 26136Date: Wed Mar 9 11:13:14 2016 -0800 26137 26138 xfree86: Set xf86CrtcConfigRec cursor pointer to NULL in HideCursor 26139 26140 This makes the cursor pointer held by xf86Cursors.c get reset to NULL 26141 whenever the cursor isn't displayed, and means that the reference 26142 count held in xf86Cursor.c is sufficient to cover the reference in 26143 xf86Cursors.c. 26144 26145 As HideCursor may be called in the cursor loading path after 26146 UseHWCursor or UseHWCursorARGB when HARDWARE_CURSOR_UPDATE_UNHIDDEN 26147 isn't set in the Flags field, the setting of the cursor pointer had to 26148 be moved to the LoadCursor paths. 26149 26150 LoadCursorARGBCheck gets the cursor pointer, but LoadCursorImageCheck 26151 does not. For LoadCursorImageCheck, I added a new function, 26152 xf86CurrentCursor, which returns the current cursor. With this new 26153 function, we can eliminate the cursor pointer from the 26154 xf86CrtcConfigRec, once drivers are converted over to use it. 26155 26156 Signed-off-by: Keith Packard <keithp@keithp.com> 26157 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 26158 26159commit b64108fa305e956e4edaae9d53071ff0abee268e 26160Author: Michel Dänzer <michel.daenzer@amd.com> 26161Date: Wed May 25 19:09:23 2016 +0900 26162 26163 glamor: Check for composite operations which are equivalent to copies 26164 26165 Increases x11perf -compwinwin500 numbers by a factor of 10 for me with 26166 radeonsi. 26167 26168 Conditions copied from exaComposite(). 26169 26170 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 26171 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26172 26173commit 4711ebc174206b5a50e6ae8a7f974cd835e4ebd3 26174Author: Michel Dänzer <michel.daenzer@amd.com> 26175Date: Tue May 24 18:12:42 2016 +0900 26176 26177 glamor: Fix sampling outside of RGBx source/mask pictures 26178 26179 RENDER requires that sampling outside of any source/mask picture results 26180 in alpha == 0.0. 26181 26182 The OpenGL border colour cannot set alpha = 0.0 if the texture format 26183 doesn't have an alpha channel, so we have to use the RepeatFix handling 26184 in that case. 26185 26186 Also, only force alpha = 1.0 when sampling inside of RGBx source/mask 26187 pictures. 26188 26189 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94514 26190 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 26191 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26192 26193commit 4d649d51770cace4d7c1e51d9a199ac7a056c30a 26194Author: Keith Packard <keithp@keithp.com> 26195Date: Wed May 18 16:03:48 2016 -0500 26196 26197 mi: Remove miPointerRec from API 26198 26199 This moves the definition of miPointerRec from mipointrst.h to 26200 mipointer.c so that it is no longer visible in the API, allowing it to 26201 be changed while the API/ABI is frozen. 26202 26203 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 26204 Signed-off-by: Keith Packard <keithp@keithp.com> 26205 26206commit 3f9015b6dc9e7e9c97f8717dea6af9f4d8523f2e 26207Author: Keith Packard <keithp@keithp.com> 26208Date: Wed May 18 16:03:47 2016 -0500 26209 26210 xwayland: Move sprite invalidation logic into mipointer 26211 26212 This creates a function that invalidates the current sprite and forces 26213 a sprite image reload the next time the sprite is checked, moving that 26214 logic out of the xwayland sources and allowing the miPointerRec 26215 structure to be removed from the server API. 26216 26217 Signed-off-by: Keith Packard <keithp@keithp.com> 26218 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 26219 26220commit 4e124203f2260daaf54155f4a05fe469733e0b97 26221Author: Adam Jackson <ajax@redhat.com> 26222Date: Tue May 3 11:24:58 2016 -0400 26223 26224 damage: Make damageRegionProcessPending take a damage not a drawable 26225 26226 In the case where there's no damage monitor on the drawable, we look 26227 that fact up twice: once before rendering to decide whether to compute 26228 damage, and again after to decide whether to append it. This is wasted 26229 effort, as the layer below us is effectively not allowed to change 26230 whether there's a damage monitor for the drawable, but there's no way 26231 the compiler can know that. 26232 26233 Instead, look it up once up front, and change the check macros and 26234 damageRegionProcessPending to take a damage not a drawable. 26235 26236 v2: Explicitly pass pDamage to the macros as well (Michel Dänzer) 26237 26238 Signed-off-by: Adam Jackson <ajax@redhat.com> 26239 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26240 26241commit 1bb932bf1417ff319a594d7e40f6813a26936784 26242Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26243Date: Sat May 7 00:43:33 2016 -0700 26244 26245 XQuartz: Fix default CFBundleVersion 26246 26247 m4/shell variable name collision broke the case when the configure 26248 option was not used 26249 26250 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26251 26252commit 2fbf5c2f91d33efbda573c4be036248b1d8ed7f1 26253Author: Andrew Comminos <andrew@comminos.com> 26254Date: Tue Dec 29 23:41:27 2015 -0800 26255 26256 Input: Send XI2 FocusOut NotifyPointer events to the pointer window. 26257 26258 This changes XInput 2's propagation of NotifyPointer focus out events to 26259 include the pointer window as well, similar to core events. This fixes 26260 a potential permanent focus in GDK when the focus moves to PointerRoot. 26261 26262 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=93539 26263 26264 Signed-off-by: Andrew Comminos <andrew@comminos.com> 26265 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 26266 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 26267 26268commit 984be789d5935bc7f695dc61234ef0251ed33679 26269Author: Olivier Fourdan <ofourdan@redhat.com> 26270Date: Wed May 11 09:15:19 2016 +0200 26271 26272 xwayland: don't check events as early as InitInput 26273 26274 If data is received during XWayland startup, it will be read early in 26275 InitInput() before the connection data is initialized, causing a crash. 26276 26277 Remove the wayland rountrips from InitInput() as this is done again in 26278 xwl_screen_init() where it seems more appropriate. 26279 26280 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 26281 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=95337 26282 Reviewed-by: Adam Jackson <ajax@redhat.com> 26283 26284commit 3f569435e10df515c6f86289dc986bfc59c6bc18 26285Author: Adam Jackson <ajax@redhat.com> 26286Date: Thu Mar 31 16:56:44 2016 -0400 26287 26288 glx: Stop tracking hasUnflushedCommands 26289 26290 This is only meaningful for indirect contexts, and all it does is 26291 (maybe) prevent a flush when switching away from an indirect context. 26292 Indirect contexts aren't worth optimizing for, and Mesa tracks whether 26293 a flush is needed anyway. 26294 26295 Careful readers will note that ReadPixels would reset the flag even 26296 though it doesn't imply a flush! 26297 26298 Reviewed-by: Eric Anholt <eric@anholt.net> 26299 Signed-off-by: Adam Jackson <ajax@redhat.com> 26300 26301commit 0ebb58f6b6f66afbc3a48bd10693fe190aed8694 26302Author: Adam Jackson <ajax@redhat.com> 26303Date: Thu Mar 31 16:20:40 2016 -0400 26304 26305 glx: Remove some redundant zero-init of GLX context state 26306 26307 The context is already calloc'd. 26308 26309 Reviewed-by: Eric Anholt <eric@anholt.net> 26310 Signed-off-by: Adam Jackson <ajax@redhat.com> 26311 26312commit d5ba09598210063926aeff55ac58d2b7020d9a03 26313Author: Adam Jackson <ajax@redhat.com> 26314Date: Thu Mar 31 15:51:39 2016 -0400 26315 26316 glx/dri2: Implement dri2FlushFrontBuffer explicitly 26317 26318 No functional change. 26319 26320 Reviewed-by: Eric Anholt <eric@anholt.net> 26321 Signed-off-by: Adam Jackson <ajax@redhat.com> 26322 26323commit 0ba4e251a1012bd5a2eab83e2c51006ec3805394 26324Author: Adam Jackson <ajax@redhat.com> 26325Date: Thu Mar 31 15:50:34 2016 -0400 26326 26327 glx: Code motion 26328 26329 No functional change, just rearranging some code to make later commits 26330 more obvious. 26331 26332 Reviewed-by: Eric Anholt <eric@anholt.net> 26333 Signed-off-by: Adam Jackson <ajax@redhat.com> 26334 26335commit f523ebb5492a2e12b76a8b5b7f1b0e5efd4a8040 26336Author: Adam Jackson <ajax@redhat.com> 26337Date: Mon Apr 18 12:56:17 2016 -0400 26338 26339 dix: Remove pointless client-state callbacks 26340 26341 Private storage is pre-zeroed by the private system itself. 26342 26343 Signed-off-by: Adam Jackson <ajax@redhat.com> 26344 Reviewed-by: Eric Anholt <eric@anholt.net> 26345 26346commit 140c2f1a54dbc2503311e9a011e841aa1f277cbd 26347Author: Michel Dänzer <michel.daenzer@amd.com> 26348Date: Fri Apr 1 18:24:00 2016 +0900 26349 26350 EXA: Honour op parameter to exaGlyphs even if maskFormat == NULL 26351 26352 Reported-by: Uli Schlachter <psychon@znc.in> 26353 26354 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94775 26355 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 26356 Reviewed-by: Uli Schlachter <psychon@znc.in> 26357 26358commit 7f0494671f95cfa33f6f2c7c74b20a21f239f8c5 26359Author: Dave Airlie <airlied@redhat.com> 26360Date: Fri May 6 05:41:58 2016 +1000 26361 26362 modesetting: fix build with glamor disabled. 26363 26364 Fix build without --enable-glamor. 26365 26366 Caught by the arm tinderbox. 26367 26368 Signed-off-by: Dave Airlie <airlied@redhat.com> 26369 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 26370 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 26371 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 26372 26373commit 66fdeb880aac0966be9aa41219047ade0148c35f 26374Author: Emil Velikov <emil.l.velikov@gmail.com> 26375Date: Mon May 2 23:47:22 2016 +0100 26376 26377 xfree86: drop unneeded strdup for modulepath/logfile 26378 26379 The destination variable is never freed, thus we even plug some memory 26380 leaks. 26381 26382 v2: Rebase against updated xf86CheckPrivs() helper. 26383 26384 Reviewed-by: Adam Jackson <ajax@redhat.com> 26385 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 26386 26387commit 032b1d79b7d04d47814a5b3a9fdd162249fea74c 26388Author: Emil Velikov <emil.l.velikov@gmail.com> 26389Date: Mon May 2 23:47:21 2016 +0100 26390 26391 xfree86: use the xf86CheckPrivs() helper for modulepath/logfile 26392 26393 v2: Rebase against updated xf86CheckPrivs() helper. 26394 26395 Reviewed-by: Adam Jackson <ajax@redhat.com> 26396 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 26397 26398commit 0fdd47560468e7b6147c666cc6b9eebcecb68572 26399Author: Emil Velikov <emil.l.velikov@gmail.com> 26400Date: Mon May 2 23:47:20 2016 +0100 26401 26402 xfree86: factor out the check priviliges and print a big warning 26403 26404 Current message was quite off "file specified must be a relative path" 26405 and alike. Just factor it out and use "path/file" as needed. 26406 26407 v2: Rework error message, drop "Using default", print actual arg value. 26408 26409 Reviewed-by: Adam Jackson <ajax@redhat.com> 26410 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 26411 26412commit 2378adde6770385c06f7efcd32f270d00d49ce3f 26413Author: Dave Airlie <airlied@gmail.com> 26414Date: Fri Apr 29 14:01:33 2016 +1000 26415 26416 modesetting: set capabilities up after glamor and enable offload caps. 26417 26418 This moves the capabilites setting to after glamor is initialised, and 26419 enables the offload caps in cases where they work. This enables DRI2 26420 PRIME support with modesetting. 26421 26422 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26423 Signed-off-by: Dave Airlie <airlied@redhat.com> 26424 26425commit 258588224d6b35de12f97d9092efa5cb3e426550 26426Author: Dave Airlie <airlied@gmail.com> 26427Date: Fri Apr 29 14:01:32 2016 +1000 26428 26429 xf86Crtc: don't set the root window property on slave GPUs. 26430 26431 Slave GPUs don't have a root window to set this on, so don't. 26432 26433 This fixes some crashes I saw just playing around. 26434 26435 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26436 Signed-off-by: Dave Airlie <airlied@redhat.com> 26437 26438commit a41a171bcbae9aeafac2865faa904f15d9b59925 26439Author: Dave Airlie <airlied@gmail.com> 26440Date: Fri Apr 29 14:01:31 2016 +1000 26441 26442 modesetting: set driverPrivate to NULL after closing fd. 26443 26444 Otherwise ms_ent_priv will return NULL and things will fall apart. 26445 26446 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26447 Signed-off-by: Dave Airlie <airlied@redhat.com> 26448 26449commit fc1f61b75c9712e24ff665f95d62771f39eba1c7 26450Author: Dave Airlie <airlied@gmail.com> 26451Date: Fri Apr 29 14:01:30 2016 +1000 26452 26453 randr/provider: only allow slave gpu to be offload sources. 26454 26455 The other way around makes no sense. 26456 26457 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26458 Signed-off-by: Dave Airlie <airlied@redhat.com> 26459 26460commit 2d0f151c6195ba94d0320b55227a108e6301a74c 26461Author: Alex Goins <agoins@nvidia.com> 26462Date: Fri Apr 29 14:01:29 2016 +1000 26463 26464 modesetting: Consistent whitespace in driver.c 26465 26466 For some reason a couple of the dirty functions in driver.c used 8 26467 spaces per tab instead of 4 like the rest of the file. Fix this to make 26468 it more consistent and give me more room to work in ms_dirty_update in 26469 subsequent commits. 26470 26471 v1: N/A 26472 v2: N/A 26473 v3: N/A 26474 v4: Initial commit 26475 26476 Signed-off-by: Alex Goins <agoins@nvidia.com> 26477 Reviewed-by: Dave Airlie <airlied@redhat.com> 26478 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26479 26480commit 137ac094e7ab8c871f3b36e40ad826ac797f0e26 26481Author: Adam Jackson <ajax@redhat.com> 26482Date: Fri Apr 1 22:44:26 2016 -0400 26483 26484 dix: Push UpdateCurrentTimeIf down out of the main loop 26485 26486 This was added in: 26487 26488 commit 312910b4e34215aaa50fc0c6092684d5878dc32f 26489 Author: Chase Douglas <chase.douglas@canonical.com> 26490 Date: Wed Apr 18 11:15:40 2012 -0700 26491 26492 Update currentTime in dispatch loop 26493 26494 Unfortunately this is equivalent to calling GetTimeInMillis() once per 26495 request. In the absolute best case (as on Linux) you're only hitting the 26496 vDSO; on other platforms that's a syscall. Either way it puts a pretty 26497 hard ceiling on request throughput. 26498 26499 Instead, push the call down to the requests that need it; basically, 26500 grab processing and event generation. 26501 26502 Cc: Chase Douglas <chase.douglas@canonical.com> 26503 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 26504 Signed-off-by: Adam Jackson <ajax@redhat.com> 26505 26506commit ac164e58870d70640381e68b776eb95578c7fbd3 26507Author: Peter Hutterer <peter.hutterer@who-t.net> 26508Date: Wed May 4 20:36:18 2016 +1000 26509 26510 xkb: after changing the keymap, force an indicator update 26511 26512 When NumLock is on and a new keymap is applied, the next modifier state 26513 change will turn off that LED (but leave the state enabled). The cause 26514 for this is a bit convoluted: 26515 26516 * the SLI explicitState is copied from the current state in 26517 ProcXkbGetKbdByName. Thus, if NumLock is on, that state is 0x2. 26518 * on the next modifier key press (e.g. Shift), XkbApplyState() calls into 26519 XkbUpdateIndicators() -> XkbUpdateLedAutoState() to update SLIs (if any) 26520 for the currently changed modifier. But it does so with a mask only for 26521 the changed modifier (i.e. for Shift). 26522 * XkbUpdateLedAutoState() calculates the state based on this mask and 26523 ends up with 0 because we don't have a Shift LED and we masked out the 26524 others. 26525 * XkbUpdateLedAutoState() compares that state with the previous state 26526 (which is still 0x2) and then proceeds to turn the LED off 26527 26528 This doesn't happen in the normal case because either the mask 26529 encompasses all modifiers or the state matches of the masked-out 26530 modifiers matches the old state. 26531 26532 Avoid this issue by forcing an SLI update after changing the keymap. 26533 This updates the sli->effectiveState and thus restores everything to 26534 happy working order. 26535 26536 https://bugzilla.redhat.com/show_bug.cgi?id=1047151 26537 26538 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 26539 Reviewed-by: Daniel Stone <daniels@collabora.com> 26540 26541commit 059d5ef30490233f410ca87084c7697b87e5b05e 26542Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26543Date: Tue May 3 23:43:06 2016 -0700 26544 26545 XQuartz: Update copyright years 26546 26547 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26548 26549commit d6ba4f2c52da150a9a92bdb00efe7902d17033bd 26550Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26551Date: Tue May 3 23:42:34 2016 -0700 26552 26553 XQuartz: Add --with-bundle-version and --with-bundle-version-string configure options 26554 26555 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26556 26557commit c1614928c10a8f8400f99acfd1b7f96d503af7ec 26558Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26559Date: Tue May 3 23:21:38 2016 -0700 26560 26561 XQuartz: Add --with-sparkle-feed-url configure option 26562 26563 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26564 26565commit 299b01eabf827a7435b5d6004d50637ac710bbc7 26566Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26567Date: Tue May 3 23:14:24 2016 -0700 26568 26569 XQuartz: Update release feed URL to use new https URL 26570 26571 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26572 26573commit 16d6733c63727d910eb516d7f6950f4675281f2d 26574Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26575Date: Tue May 3 23:24:44 2016 -0700 26576 26577 XQuartz: Fix the help text for --with-bundle-id-prefix 26578 26579 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26580 26581commit 214a66b661dcb56ebb9776e34049753f65c7510a 26582Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26583Date: Tue May 3 23:16:46 2016 -0700 26584 26585 XQuartz: Remove --with-launchd-id-prefix 26586 26587 It's been deprecated for years. 26588 26589 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 26590 26591commit 2285fe78c04714561a0d1a164a41a38c48263f89 26592Author: Peter Hutterer <peter.hutterer@who-t.net> 26593Date: Fri Apr 22 15:04:37 2016 +1000 26594 26595 xfree86: add support for MatchIsTabletPad 26596 26597 The tablet pads have been separate kernel devices for a while now and 26598 libwacom has labelled them with the udev ID_INPUT_TABLET_PAD for over a year 26599 now. Add a new MatchIsTabletPad directive to apply configuration options 26600 specifically to the Pad part of a tablet. 26601 26602 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 26603 Reviewed-by: Adam Jackson <ajax@redhat.com> 26604 26605commit fa02b05645080c285da5972262a8d37403e39d7e 26606Author: Dave Airlie <airlied@redhat.com> 26607Date: Tue May 3 06:54:57 2016 +1000 26608 26609 modesetting: port clean start code from amdgpu. (v2) 26610 26611 Both radeon and amdgpu don't set the mode until the first blockhandler, 26612 this means everything should be rendered on the screen correctly by 26613 then. 26614 26615 This ports this code, it also removes the tail call of EnterVT from 26616 ScreenInit, it really isn't necessary and causes us to set a dirty mode 26617 with -modesetting always anyways. 26618 26619 v2: reorder set desired modes vs block handler as done for amdgpu. 26620 26621 Reviewed-by: Eric Anholt <eric@anholt.net> 26622 Signed-off-by: Dave Airlie <airlied@redhat.com> 26623 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26624 26625commit caabc4e85540dcd4225f2780b5616f7d870fbb06 26626Author: Dave Airlie <airlied@redhat.com> 26627Date: Tue May 3 06:54:56 2016 +1000 26628 26629 modesetting: add support for background none. 26630 26631 This adds support using glamor for background None. 26632 26633 loosely based off the amdgpu code. relies on the glamor_finish code. 26634 26635 Acked-by: Eric Anholt <eric@anholt.net> 26636 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26637 Signed-off-by: Dave Airlie <airlied@redhat.com> 26638 26639commit c33250945b45adc447154239f0cf48fb9b2d7335 26640Author: Adam Jackson <ajax@redhat.com> 26641Date: Fri Apr 8 11:26:36 2016 -0400 26642 26643 kdrive: Nuke a bunch of dead code 26644 26645 gcc6 says: 26646 26647 keyboard.c:46:21: warning: ‘linux_to_x’ defined but not used 26648 26649 Only referenced by a bunch of long if-0'd code, so chuck it all out. 26650 26651 Reviewed-by: Julien Cristau <jcristau@debian.org> 26652 Signed-off-by: Adam Jackson <ajax@redhat.com> 26653 26654commit a5dd7b890f4f3a5245639591c73303c5a087b38a 26655Author: Adam Jackson <ajax@redhat.com> 26656Date: Fri Apr 8 11:24:50 2016 -0400 26657 26658 dix: Squash some new gcc6 warnings 26659 26660 -Wlogical-op now tells us: 26661 26662 devices.c:1685:23: warning: logical ‘and’ of equal expressions 26663 26664 Reviewed-by: Julien Cristau <jcristau@debian.org> 26665 Signed-off-by: Adam Jackson <ajax@redhat.com> 26666 26667commit 23dfa017298ceceac818f83779858e490c7757b6 26668Author: Andreas Schwab <schwab@suse.de> 26669Date: Thu Apr 28 14:47:33 2016 +0200 26670 26671 x86emu: Change include order to avoid conflict with system header 26672 26673 R_SP is also defined in <sys/ucontext.h> on m68k. Also remove duplicate 26674 definitions. 26675 26676 Reviewed-by: Adam Jackson <ajax@redhat.com> 26677 Signed-off-by: Andreas Schwab <schwab@suse.de> 26678 26679commit aa4e757130010dd3202f10ec6cb0c306c1dbcfbc 26680Author: Dave Airlie <airlied@redhat.com> 26681Date: Fri Mar 11 09:22:00 2016 +1000 26682 26683 glamor: add glamor_finish API 26684 26685 Some drivers are calling glFinish, they really should be doing this. 26686 26687 This also is needed for some reverse prime scenarios. 26688 26689 Signed-off-by: Dave Airlie <airlied@redhat.com> 26690 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26691 Reviewed-by: Eric Anholt <eric@anholt.net> 26692 26693commit f48b0534f110397246809d279225afedb28aa233 26694Author: Marek Chalupa <mchqwerty@gmail.com> 26695Date: Mon Apr 25 11:33:00 2016 +0200 26696 26697 xwayland-shm: fortify fallocate against EINTR 26698 26699 If posix_fallocate or ftruncate is interrupted by signal while working, 26700 we return -1 as fd and the allocation process returns BadAlloc error. 26701 That causes xwayland clients to abort with 'BadAlloc (insufficient 26702 resources for operation)' even when there's a lot of resources 26703 available. 26704 26705 Fix it by trying again when we get EINTR. 26706 26707 Signed-off-by: Marek Chalupa <mchqwerty@gmail.com> 26708 Reviewed-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 26709 26710commit 4cc32880737c2d3e568fdb4867b2dba10fb3998a 26711Author: Alexandre Courbot <acourbot@nvidia.com> 26712Date: Wed Mar 23 13:47:37 2016 +0900 26713 26714 configure.ac: Keep environment CFLAGS when testing 26715 26716 DRI2 detection could fail if configure is invoked with a sysroot passed 26717 as CFLAGS. Ideally configure should invoke gcc with the sysroot argument 26718 passed to the configure script, but for some reason this is not done by 26719 AC_COMPILE_IFELSE. 26720 26721 Fix this by ensuring CFLAGS are preserved when checking for stuff. 26722 26723 Signed-off-by: Alexandre Courbot <acourbot@nvidia.com> 26724 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 26725 26726commit 16e4bce9e5257c50c80c66efee0d07c2483619e1 26727Author: Simon Thum <simon.thum@gmx.de> 26728Date: Fri Apr 8 13:24:39 2016 +0200 26729 26730 dix/ptraccel: Remove float literals 26731 26732 This was fine back when valuators were integer. Device 26733 properties are float (not double), so some instances remain. 26734 26735 Signed-off-by: Simon Thum <simon.thum@gmx.de> 26736 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 26737 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 26738 26739commit c8e5fc30575a309c25970fc68b9184c07bb74df4 26740Author: Simon Thum <simon.thum@gmx.de> 26741Date: Tue Apr 5 14:29:47 2016 +0200 26742 26743 dix/ptraccel: Fix memory leak in InitPredictableAccelerationScheme 26744 26745 This was quite unlikely except in situations where a proper startup 26746 would have been impossible anyway, but since automated checks don't 26747 grade likelyhood just fix it. 26748 26749 Detected by Jeremy Huddleston's clang checks. 26750 26751 Signed-off-by: Simon Thum <simon.thum@gmx.de> 26752 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 26753 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 26754 26755commit f641ae412287ecb7a3437987e2ba1646a8443aa4 26756Author: Peter Hutterer <peter.hutterer@who-t.net> 26757Date: Wed Mar 9 10:45:48 2016 +1000 26758 26759 Xi: don't deliver emulated motion events for non-emulating touches 26760 26761 The touchpoint knows whether it should be emulating or not and we have a check 26762 for that later. Check for this before we generate the event and try to deliver 26763 it, lest we trigger a bug warning. 26764 26765 https://bugzilla.redhat.com/show_bug.cgi?id=1282252 26766 26767 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 26768 26769commit 577bebe2067293bb154068e99a2ef085b066cb67 26770Author: Emil Velikov <emil.l.velikov@gmail.com> 26771Date: Sun Apr 17 20:30:46 2016 +0100 26772 26773 xfree86/parser: simplify #ifdef ladder 26774 26775 Rather than 'hacking' around symbol names and providing macros such as 26776 'Local' just fold things and make the code more readable. 26777 26778 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 26779 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26780 26781commit 537276a5b86b7341169ea4eb36d479a503ba5d84 26782Author: Emil Velikov <emil.l.velikov@gmail.com> 26783Date: Sun Apr 17 20:30:45 2016 +0100 26784 26785 xfree86/parser: reuse StringToToken() in xf86getToken() 26786 26787 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26788 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 26789 26790commit 944ea03d5be2ffe22a3f1c4c287760261c31235f 26791Author: Emil Velikov <emil.l.velikov@gmail.com> 26792Date: Sun Apr 17 20:30:44 2016 +0100 26793 26794 xfree86/parser: move StringToToken() definition further up 26795 26796 ... so that we can use it without the forward declaration. Plus we're 26797 doing to reuse it in the next commit ;-) 26798 26799 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26800 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 26801 26802commit b93be14b7d339e4e46d941729dad853452fae8c0 26803Author: Emil Velikov <emil.l.velikov@gmail.com> 26804Date: Sun Apr 17 20:30:43 2016 +0100 26805 26806 xfree86/parser: annotate xf86ConfigSymTabRec as constant data 26807 26808 Add the const notation to all the static storage as well as the 26809 functions that use it - xf86getToken(), xf86getSubTokenWithTab(), 26810 StringToToken() and xf86getStringToken(). 26811 26812 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26813 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 26814 26815commit 3981dcdd489b60fbf356534a509ca93dcbedf769 26816Author: Emil Velikov <emil.l.velikov@gmail.com> 26817Date: Sun Apr 17 20:34:43 2016 +0100 26818 26819 dri3: remove unused file dri3int.h 26820 26821 Copied during the prototyping stage and never used. 26822 26823 Cc: Keith Packard <keithp@keithp.com> 26824 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 26825 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 26826 26827commit a1b13cda6169a98d694451fec75e63352e9d90bd 26828Author: Adam Jackson <ajax@redhat.com> 26829Date: Wed Jan 6 09:09:21 2016 -0500 26830 26831 xfree86: Remove xf86RegisterRootWindowProperty 26832 26833 All consumers have been ported to the root window callback, so this can 26834 all be nuked. 26835 26836 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26837 Signed-off-by: Adam Jackson <ajax@redhat.com> 26838 26839commit e89c7f1c2a0ea3480b21446e413073c1427285ae 26840Author: Adam Jackson <ajax@redhat.com> 26841Date: Wed Jan 6 09:04:15 2016 -0500 26842 26843 xfree86: Create EDID atom from the root window callback (v2) 26844 26845 v2: Fix swapped callback args 26846 26847 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26848 Signed-off-by: Adam Jackson <ajax@redhat.com> 26849 26850commit 8e3010d7d8e8c49c8859b576de1808ae7b2859be 26851Author: Adam Jackson <ajax@redhat.com> 26852Date: Wed Jan 6 08:54:47 2016 -0500 26853 26854 xfree86: Remove a never-hit diagnostic message 26855 26856 Practically speaking, the EDID major version is never not 1. 26857 26858 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26859 Signed-off-by: Adam Jackson <ajax@redhat.com> 26860 26861commit 7961377567f15dfad9d96c5c0a0992b38013d973 26862Author: Adam Jackson <ajax@redhat.com> 26863Date: Wed Jan 6 08:47:37 2016 -0500 26864 26865 xfree86: Make xf86SetDDCproperties work more than once (v2) 26866 26867 We can call this more than once via xf86OutputSetEDID since hotplug is 26868 actually a thing in RANDR 1.2, but xf86RegisterRootWindowProperty merely 26869 adds the data to a list to be applied to the root at CreateWindow time, 26870 so calls past the first (for a given screen) would have no effect until 26871 server regen. 26872 26873 Once we've initialised pScrn->pScreen is filled in, so we can just set 26874 the property directly. 26875 26876 v2: Removed pointless version check, deobfuscate math (Walter Harms) 26877 26878 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26879 Signed-off-by: Adam Jackson <ajax@redhat.com> 26880 26881commit 8be83fff04a009109a956837ca983a96fd279711 26882Author: Adam Jackson <ajax@redhat.com> 26883Date: Wed Jan 6 08:35:43 2016 -0500 26884 26885 xfree86: Remove some leftovers from DisplayID support 26886 26887 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26888 Signed-off-by: Adam Jackson <ajax@redhat.com> 26889 26890commit 0cd2a24b61ef1583fc6b3fec7d01c7481cc97d52 26891Author: Adam Jackson <ajax@redhat.com> 26892Date: Wed Jan 6 08:30:07 2016 -0500 26893 26894 xfree86: Unexport xf86Initialising, remove xf86ServerIsInitialising 26895 26896 Neither of these are used from outside the server. 26897 26898 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26899 Signed-off-by: Adam Jackson <ajax@redhat.com> 26900 26901commit e70ee11a39b957141fbc565d79d128a46fac5f34 26902Author: Adam Jackson <ajax@redhat.com> 26903Date: Tue Jan 5 16:27:41 2016 -0500 26904 26905 xfree86: Create VT atoms from the root window callback (v2) 26906 26907 v2: Fix swapped callback args 26908 26909 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26910 Signed-off-by: Adam Jackson <ajax@redhat.com> 26911 26912commit da9ee1eddd65d00788cef8a3becfad948c0a2168 26913Author: Adam Jackson <ajax@redhat.com> 26914Date: Tue Jan 5 16:11:42 2016 -0500 26915 26916 xfree86: Create seat atom from the root window callback (v2) 26917 26918 v2: Fix swapped callback args 26919 26920 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26921 Signed-off-by: Adam Jackson <ajax@redhat.com> 26922 26923commit 2c3a3afb5104714b637c1c4aea195df73e0fa918 26924Author: Adam Jackson <ajax@redhat.com> 26925Date: Tue Jan 5 15:56:42 2016 -0500 26926 26927 dix: Add RootWindowFinalizeCallback 26928 26929 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 26930 Signed-off-by: Adam Jackson <ajax@redhat.com> 26931 26932commit 8437955515ad59b0bfcd6598248e7f0ffc706370 26933Author: Olivier Fourdan <ofourdan@redhat.com> 26934Date: Thu Mar 17 09:53:58 2016 +0100 26935 26936 glamor: fix wrong offset on composite rectangles 26937 26938 When using PictOpSrc, the destination is wrongly shifted back to (0, 0). 26939 26940 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94568 26941 Reviewed-by: Adam Jackson <ajax@redhat.com> 26942 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 26943 26944commit e8e5d839968e22cf42a6e1982a07b02c6f4a4562 26945Author: Olivier Fourdan <ofourdan@redhat.com> 26946Date: Fri Apr 1 13:38:11 2016 +0200 26947 26948 xwayland: Fix compiler warning in GLAMOR Xv 26949 26950 XvWindowMask is defined as 0x00020000 and cannot fit in the XvAdaptor 26951 type which is defined as an unsigned char, thus causing a compiler 26952 warning: 26953 26954 xwayland-glamor-xv.c: In function ‘xwl_glamor_xv_add_adaptors’: 26955 xwayland-glamor-xv.c:339:16: warning: large integer implicitly 26956 truncated to unsigned type [-Woverflow] 26957 26958 This XvWindowMask value is actually not used for XvAdaptor itself but by 26959 the server in its xf86xv implementation, so we don't even need that mask 26960 in our xwayland-glamor-xv implementation. 26961 26962 Reviewed-by: Adam Jackson <ajax@redhat.com> 26963 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 26964 26965commit f9b5bbaa3a7fce1a4efb2084b8d9d82b98d4ee83 26966Author: Peter Hutterer <peter.hutterer@who-t.net> 26967Date: Tue Mar 8 15:42:42 2016 +1000 26968 26969 xkb: fix SlowKeys release/reject beeps 26970 26971 Wrong use of the mask here caused a beep whenever a key was rejected but 26972 also when it was released after being accepted. Fix the mask to check 26973 for the correct enabled controls. 26974 26975 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 26976 Reviewed-by: Daniel Stone <daniels@collabora.com> 26977 26978commit a6288f0954cf97499e78849a87847062ee962c17 26979Author: Jon Turney <jon.turney@dronecode.org.uk> 26980Date: Tue Nov 3 19:07:43 2015 +0000 26981 26982 hw/xwin: xcbify internal window manager 26983 26984 Convert the code for the multiwindow mode internal window manager to xcb 26985 26986 xcb conversion avoids xlib/xserver namespace collision and _XSERVER64 type 26987 sizing issues 26988 26989 v2: Various fixes 26990 v3: Don't include X11/extensions/windowswmstr.h, which uses the Display type 26991 and thus depends on Xlib.h, just for _WINDOWSWM_NATIVE_HWND 26992 v4: Fix indentation, add some error handling. 26993 Fix a bug with ConfigureNotify handling 26994 v5: Fix a bug which prevented WM_NORMAL_HINTS from being checked 26995 26996 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 26997 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 26998 26999commit 8114b8127f01fc81390fc13e3d09bcc50e41a66f 27000Author: Jon Turney <jon.turney@dronecode.org.uk> 27001Date: Wed Jul 1 16:30:36 2015 +0100 27002 27003 hw/xwin: In multiwindow mode, do window minimization entirely in the WM 27004 27005 Remove winMinimizeWindow(), implement as UpdateState() in the WM instead, 27006 which uses getHwnd() to map a Window XID to a HWND (like everything else in 27007 the WM), rather than peering into the servers internal data structures. 27008 27009 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27010 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27011 27012commit 6a64b9d7af70dc7ff2cac8b35a1f7b0797823733 27013Author: Jon Turney <jon.turney@dronecode.org.uk> 27014Date: Mon Nov 2 17:55:19 2015 +0000 27015 27016 hw/xwin: xcbify code for converting X11 icon to Win32 icon 27017 27018 Convert the code for converting an X11 icon to Win32 icon from Xlib to xcb. 27019 27020 v2: some warning fixes in winXIconToHICON() 27021 v3: declaration-after-statement warning fixes 27022 v4: printf format fixes 27023 v5: convert in place rather than in a library 27024 27025 This also avoids the xlib/xserver namespace collision issues, so 27026 winmultiwindowicons.h can be included everywhere it should be, which fixes 27027 compilation with -Werror=implicit-function-declaration 27028 27029 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27030 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27031 27032commit 0a69c1e2fa0ea63b02fff98e68d9f56a369e882b 27033Author: Jon Turney <jon.turney@dronecode.org.uk> 27034Date: Wed Mar 30 18:31:38 2016 +0100 27035 27036 xwin/glx: Build fix for warnings about missing WGL extensioons 27037 27038 Reviewed-by: Adam Jackson <ajax@redhat.com> 27039 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27040 27041commit b08526eecf1e165ed9ec2e6b571a5a616a9b696e 27042Author: Adam Jackson <ajax@redhat.com> 27043Date: Wed Mar 16 11:38:13 2016 -0400 27044 27045 glx: Implement GLX_EXT_libglvnd (v2) 27046 27047 For the dri2 backend, we depend on xfree86 already, so we can walk the 27048 options for the screen looking for a vendor string from xorg.conf. For 27049 the swrast backend we don't have that luxury, so just say mesa. This 27050 extension isn't really meaningful on Windows or OSX yet (since libglvnd 27051 isn't really functional there yet), so on those platforms we don't say 27052 anything and return BadValue for the token from QueryServerString. 27053 27054 v2: Use xnf* allocators when parsing options (Eric and Emil) 27055 27056 Reviewed-by: Eric Anholt <eric@anholt.net> 27057 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27058 Signed-off-by: Adam Jackson <ajax@redhat.com> 27059 27060commit 2e8781ead3067b195baec2e76a28091575679383 27061Author: Adam Jackson <ajax@redhat.com> 27062Date: Wed Mar 23 15:41:24 2016 -0400 27063 27064 glx: Compute the GLX extension string from __glXScreenInit 27065 27066 Now that the enable bits are in the screen base class we can compute 27067 this in one place, rather than making every backend do it. 27068 27069 Reviewed-by: Eric Anholt <eric@anholt.net> 27070 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27071 Signed-off-by: Adam Jackson <ajax@redhat.com> 27072 27073commit e21de4bf3c5ff8cbb9c5ea023d04162e5e56b3df 27074Author: Adam Jackson <ajax@redhat.com> 27075Date: Wed Mar 23 15:36:52 2016 -0400 27076 27077 glx: Move glx_enable_bits up to the GLX screen base class 27078 27079 Reviewed-by: Eric Anholt <eric@anholt.net> 27080 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27081 Signed-off-by: Adam Jackson <ajax@redhat.com> 27082 27083commit 23cce73221c0b96e7778da34616f8c3f4d6aa819 27084Author: Adam Jackson <ajax@redhat.com> 27085Date: Wed Mar 23 15:13:51 2016 -0400 27086 27087 xquartz/glx: Remove unused fields from the glx screen subclass 27088 27089 dmt:~/git/xserver% git grep -E '\<(index|num_vis)\>' hw/xquartz/GL 27090 hw/xquartz/GL/indirect.c: int index; 27091 hw/xquartz/GL/indirect.c: int num_vis; 27092 27093 Reviewed-by: Eric Anholt <eric@anholt.net> 27094 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27095 Signed-off-by: Adam Jackson <ajax@redhat.com> 27096 27097commit 36bcbf76dcc7e88cac093f8fb656c525bfeaf65d 27098Author: Adam Jackson <ajax@redhat.com> 27099Date: Wed Mar 23 15:26:23 2016 -0400 27100 27101 glx: Enable GLX 1.4 unconditionally 27102 27103 Reviewed-by: Eric Anholt <eric@anholt.net> 27104 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27105 Signed-off-by: Adam Jackson <ajax@redhat.com> 27106 27107commit 2a72789ee8e88f612dff48ebe2ebe9fecda7a95d 27108Author: Adam Jackson <ajax@redhat.com> 27109Date: Wed Mar 16 16:28:13 2016 -0400 27110 27111 xwin/glx: Drop GLWIN_NO_WGL_EXTENSIONS hack 27112 27113 This doesn't seem very useful, and we're about to implement 1.4 across 27114 the board, so some WGL extensions will become required. 27115 27116 Reviewed-by: Eric Anholt <eric@anholt.net> 27117 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27118 Signed-off-by: Adam Jackson <ajax@redhat.com> 27119 27120commit 77bdaa1313aa55191b49ec73c1e377928ca294fe 27121Author: Adam Jackson <ajax@redhat.com> 27122Date: Tue Mar 22 14:40:37 2016 -0400 27123 27124 glx: Use __glXInitExtensionEnableBits in all backends (v2) 27125 27126 On xquartz this enables SGI_make_current_read, which is a mostly 27127 harmless lie as CGL doesn't implement it, as well as SGIX_pbuffer, which 27128 is fine because no pbuffer-enabled configs are created. 27129 27130 On xwin this enables SGIX_pbuffer and ARB_multisample in all cases. 27131 Again this is harmless if the backend doesn't support the features, 27132 since no fbconfigs will be created to expose them. 27133 27134 It also adds SGIX_visual_select_group to both xquartz and xwin. 27135 Amusingly, both were filling in the appropriate field in the fbconfig 27136 already. 27137 27138 v2: Warn about missing WGL extensions (Emil) 27139 27140 Reviewed-by: Eric Anholt <eric@anholt.net> 27141 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27142 Signed-off-by: Adam Jackson <ajax@redhat.com> 27143 27144commit 15af78fc56569dc3b6a7f2c5a6a49edb602111b7 27145Author: Adam Jackson <ajax@redhat.com> 27146Date: Wed Mar 23 15:19:15 2016 -0400 27147 27148 glx: Enable GLX_SGI_make_current_read in the core 27149 27150 GLX 1.3 implies equivalent functionality, so this is safe to enable 27151 unconditionally, and bindContext always takes both drawable and readable 27152 arguments in any case. Mesa stopped exporting the __DRI_READ_DRAWABLE 27153 extension in 8.0 (when the DRI1 drivers were removed) so this will 27154 restore the extension string. 27155 27156 Reviewed-by: Eric Anholt <eric@anholt.net> 27157 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27158 Signed-off-by: Adam Jackson <ajax@redhat.com> 27159 27160commit 9b2fc6d98691966f1c9186edad956f78c31f3698 27161Author: Adam Jackson <ajax@redhat.com> 27162Date: Wed Mar 23 14:57:25 2016 -0400 27163 27164 xwin/glx: Enable GLX_SGI_make_current_read unconditionally (v2) 27165 27166 This seems to be fairly universal these days, and if it doesn't exist 27167 the only thing you break is separate drawable and readable, which is a 27168 rare feature to use. So pretend it's always there and just throw an 27169 error on MakeCurrent if it isn't, and don't consider it when computing 27170 the GLX version number. 27171 27172 v2: Fix type-o for glxWinScreen (Jon Turney) 27173 27174 Reviewed-by: Eric Anholt <eric@anholt.net> 27175 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27176 Signed-off-by: Adam Jackson <ajax@redhat.com> 27177 27178commit b2ef7df476af619903ef7f6b6962b371ae14306c 27179Author: Adam Jackson <ajax@redhat.com> 27180Date: Sat Mar 19 13:44:10 2016 -0400 27181 27182 xquartz/glx: Error out for MakeContextCurrent(draw != read) 27183 27184 CGL doesn't have a way to express this directly, unlike EGL WGL and GLX. 27185 It might be implementable, but it's never actually worked, and it's a 27186 fairly niche feature so we're better off throwing an error if someone 27187 attempts it. 27188 27189 Reviewed-by: Eric Anholt <eric@anholt.net> 27190 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27191 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 27192 Signed-off-by: Adam Jackson <ajax@redhat.com> 27193 27194commit f95645c6f70019316f8ad77b7beb84530fc0505f 27195Author: Adam Jackson <ajax@redhat.com> 27196Date: Tue Mar 22 14:37:19 2016 -0400 27197 27198 glx: Don't enable EXT_texture_from_pixmap unconditionally 27199 27200 Not all backend servers implement this. Those that don't happen to not 27201 use __glXInitExtensionEnableBits, but we'd like that to change, so fix 27202 it up before we switch them over. 27203 27204 Reviewed-by: Eric Anholt <eric@anholt.net> 27205 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27206 Signed-off-by: Adam Jackson <ajax@redhat.com> 27207 27208commit 410aec82556def5395f51299bcefbeb7d0bda604 27209Author: Adam Jackson <ajax@redhat.com> 27210Date: Tue Mar 22 14:29:06 2016 -0400 27211 27212 glx: Remove server-side mention of GLX_MESA_swap_control 27213 27214 This extension is direct-only and has no GLX protocol. We don't even 27215 track an enable bit for it, trying to turn it on is pointless. 27216 27217 Reviewed-by: Eric Anholt <eric@anholt.net> 27218 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27219 Signed-off-by: Adam Jackson <ajax@redhat.com> 27220 27221commit 3a21da59e59cf11a9113d71e3431c4bd394ff1e8 27222Author: Adam Jackson <ajax@redhat.com> 27223Date: Wed Mar 16 19:05:17 2016 -0400 27224 27225 glx: Remove default server glx extension string 27226 27227 This existed only to be strdup'd and then immediately freed. 27228 27229 Reviewed-by: Eric Anholt <eric@anholt.net> 27230 Reviewed-by: Emil Velikov <emil.velikov@collabora.com> 27231 Signed-off-by: Adam Jackson <ajax@redhat.com> 27232 27233commit 44e1c97ca6fe992bbb6ef9ecb0b82a113adfa57e 27234Author: Olivier Fourdan <ofourdan@redhat.com> 27235Date: Mon Mar 21 09:53:17 2016 +0100 27236 27237 xwayland: Pretend we support viewport in vidmode 27238 27239 Some games (namely openttd) will raise an XError and fail with a 27240 BadValue if their request to XF86VidModeSetViewPort fails. 27241 27242 Support only the default zoom and viewport, fail for everything else. 27243 27244 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 27245 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 27246 27247commit 6e3a6e30a6ac66942a0756a5d079993181f02e34 27248Author: Olivier Fourdan <ofourdan@redhat.com> 27249Date: Thu Mar 17 14:39:45 2016 +0100 27250 27251 xwayland: do not include frequency in mode name 27252 27253 Some applications (e.g. using lwjgl) try to parse the output of the 27254 xrandr command and get confused with the mode name returned by Xwayland, 27255 because it contains "@[frequency]" (e.g. "1024x640@60.0Hz"). 27256 27257 Remove the @[frequency] part of the mode name to match what is found in 27258 usual mode names on regular X servers to please those applications. 27259 27260 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94589 27261 27262 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 27263 Reviewed-by: Daniel Stone <daniels@collabora.com> 27264 27265commit 418fe365b45a143680d3b4143dc60f7cdc5a3507 27266Author: Michel Dänzer <michel.daenzer@amd.com> 27267Date: Thu Mar 24 17:34:23 2016 +0900 27268 27269 xfree86/modes: Make sure the HW cursor is hidden when it should be 27270 27271 When the HW cursor is hidden (e.g. because xf86CursorResetCursor 27272 triggers a switch from HW cursor to SW cursor), the driver isn't 27273 notified of this for disabled CRTCs. If the HW cursor was shown when the 27274 CRTC was disabled, it may still be displayed when the CRTC is enabled 27275 again. 27276 27277 Prevent this by explicitly hiding the HW cursor again after setting a 27278 mode if it's currently supposed to be hidden. 27279 27280 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94560 27281 Reviewed-by: Adam Jackson <ajax@redhat.com> 27282 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 27283 27284commit adefbaee499b9679c6cac21f52ec6545af2b36b5 27285Author: Adam Jackson <ajax@redhat.com> 27286Date: Mon Mar 28 18:11:09 2016 +0900 27287 27288 os: Treat ssh as a non-local client (v4) 27289 27290 By the time we get to ComputeLocalClient, we've already done 27291 NextAvailableClient → ReserveClientIds → DetermineClientCmd (assuming 27292 we're built with #define CLIENTIDS), so we can look up the name of the 27293 client process and refuse to treat ssh's X forwarding as if it were 27294 local. 27295 27296 v2: (Michel Dänzer) 27297 * Only match "ssh" itself, not other executable names starting with 27298 that prefix. 27299 * Ignore executable path for the match. 27300 v3: (Michel Dänzer) 27301 * Use GetClientCmdName (Mark Kettenis) 27302 * Perform check on Windows as well, but only ignore path on Cygwin 27303 (Martin Peres, Emil Velikov, Jon Turney) 27304 v4: (Michel Dänzer) 27305 * Cut of any colon and whatever comes after it. (Adam Jackson) 27306 * Add bugzilla reference. 27307 27308 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93261 27309 27310 Signed-off-by: Adam Jackson <ajax@redhat.com> 27311 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 27312 27313commit 1c90797565385426ad63bd2108085c8466695c0b 27314Author: Sonny Jiang <sonny.jiang@amd.com> 27315Date: Mon Mar 28 16:36:50 2016 +0900 27316 27317 DRI2: add Polaris PCI IDs 27318 27319 Signed-off-by: Sonny Jiang <sonny.jiang@amd.com> 27320 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> (Polaris10) 27321 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> (Polaris11) 27322 27323 (Ported from Mesa commit f00c840578a70e479ffb99f6b64c73dc420179fa) 27324 27325 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 27326 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 27327 27328commit 3b385105b2d19a1c55e9779ae88d775185eea231 27329Author: Michel Dänzer <michel.daenzer@amd.com> 27330Date: Thu Mar 24 17:42:47 2016 +0900 27331 27332 present: Only requeue for next MSC after flip failure 27333 27334 This code was added to deal with the driver present hook failing, in 27335 which case we need to wait for the next MSC before executing the 27336 presentation. 27337 27338 However, it could also take effect in cases where the driver incorrectly 27339 thinks the current MSC matches the target one (e.g. due to the kernel 27340 interface only supporting 32-bit MSC values), in which case it could 27341 result in the presentation getting requeued over and over. 27342 27343 To prevent such issues, check specifically for the target MSC 27344 immediately following the current MSC. 27345 27346 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94596 27347 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 27348 Reviewed-by: Keith Packard <keithp@keithp.com> 27349 27350commit 8ac0e05cc6c6e87b223ba7cb31d8856771c5d41a 27351Author: Adam Jackson <ajax@redhat.com> 27352Date: Wed Mar 23 14:42:08 2016 -0400 27353 27354 vfb: Re-add LD_EXPORT_SYMBOLS_FLAG to LDFLAGS 27355 27356 Accidentally removed, breaks Xvfb on cygwin. 27357 27358 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 27359 Signed-off-by: Adam Jackson <ajax@redhat.com> 27360 27361commit a288cf58a0dc0f965a6f964c76bb86bb1989d797 27362Author: Dave Airlie <airlied@redhat.com> 27363Date: Wed Mar 16 10:45:54 2016 +1000 27364 27365 glamor: swizzle RED to 0 for alpha textures 27366 27367 I'm pretty sure Eric suspected this could cause a problem, and we 27368 couldn't find a test. Well loading feedly in firefox seems to trigger 27369 badness that this solves. 27370 27371 bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94554 27372 Reviewed-by: Adam Jackson <ajax@redhat.com> 27373 Signed-off-by: Dave Airlie <airlied@redhat.com> 27374 Tested-by: Timo Aaltonen <tjaalton@ubuntu.com> 27375 27376commit 4583fa9a6cd3095d7497c075d68e7430ea3c5cb0 27377Author: Ángel González <ingenit@zoho.com> 27378Date: Mon Oct 15 22:04:22 2012 +0200 27379 27380 security: Fix reversed comment 27381 27382 Commit 6045506be0cebca4ebbe943ae77f020aafa703d4 changed back 27383 the behavior to only allow the trusted extensions to the untrusted clients, 27384 but left the 8b5d21cc1d1f4e9d20e5d5eca44cb1e60a419763 27385 comment intended for Security*Un*trustedExtensions saying that 27386 "untrusted clients shouldn't have access to these". 27387 27388 Reviewed-by: Adam Jackson <ajax@redhat.com> 27389 Signed-off-by: Ángel González <ingenit@zoho.com> 27390 27391commit b78897d0a04a833720698901fbac9535570875e4 27392Author: Evgeny M. Zubok <evgeny.zubok@tochka.ru> 27393Date: Thu Apr 8 03:58:21 2010 -0700 27394 27395 xfree86: Change VBE version early-out to 1.2. (#22672) 27396 27397 Reporter has an S3 Trio with DDC and VESA 1.2. 27398 27399 Signed-off-by: Corbin Simpson <MostAwesomeDude@gmail.com> 27400 Reviewed-by: Matt Turner <mattst88@gmail.com> 27401 27402commit 184fbf7541012090b8716c7eaf00895efd16d0ac 27403Author: Adam Jackson <ajax@redhat.com> 27404Date: Mon Mar 14 11:37:32 2016 -0400 27405 27406 xfree86: Finish removing font modules 27407 27408 Signed-off-by: Adam Jackson <ajax@redhat.com> 27409 27410commit 69d1528bc35073331e281448dfb04eb4240ff472 27411Author: Adam Jackson <ajax@redhat.com> 27412Date: Thu Apr 26 16:42:20 2012 -0400 27413 27414 xfree86: Font modules aren't a real thing 27415 27416 There are no longer any loadable font modules (not that they ever did 27417 much in the first place), so stop pretending they're a defined ABI 27418 surface. 27419 27420 Signed-off-by: Adam Jackson <ajax@redhat.com> 27421 Reviewed-by: Julien Cristau <jcristau@debian.org> 27422 27423commit 4f8151c7a4bcf2d55848e70c83b2f9ce458cf316 27424Author: Adam Jackson <ajax@redhat.com> 27425Date: Mon Mar 14 11:20:39 2016 -0400 27426 27427 Fix the typo from the previous patch, d'oh 27428 27429 Signed-off-by: Adam Jackson <ajax@redhat.com> 27430 27431commit 75eecf28ae3709181a51571132b0accd9cae316e 27432Author: Chris Wilson <chris@chris-wilson.co.uk> 27433Date: Sun Mar 13 13:54:01 2016 +0000 27434 27435 Xext/vidmode: Reduce verbosity of GetModeLine debug messages 27436 27437 In commit f175cf45aebcdda53f3ae49c0eaf27da1f194e92 27438 Author: Olivier Fourdan <ofourdan@redhat.com> 27439 Date: Wed Feb 10 09:34:34 2016 +0100 27440 27441 vidmode: move to a separate library of its own 27442 27443 the verbosity of some old debug messages (which print the reply to every 27444 GetModeLine client request and others) was increased leading to lots of 27445 log spam. Downgrade the logging back to DebugF. 27446 27447 [ajax: Fix a typo so it compiles.] 27448 27449 Reviewed-by: Adam Jackson <ajax@redhat.com> 27450 References: https://bugs.freedesktop.org/show_bug.cgi?id=94515 27451 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 27452 Cc: Olivier Fourdan <ofourdan@redhat.com> 27453 27454commit 744c292ae49dd1f9d33b860d2b4f0ae27023809b 27455Author: Adam Jackson <ajax@redhat.com> 27456Date: Fri Mar 11 13:50:32 2016 -0500 27457 27458 vidmode: Remove stray vidmodeproc.h from EXTRA_DIST 27459 27460 Was removed from the tree in: 27461 27462 commit f175cf45aebcdda53f3ae49c0eaf27da1f194e92 27463 Author: Olivier Fourdan <ofourdan@redhat.com> 27464 Date: Wed Feb 10 09:34:34 2016 +0100 27465 27466 vidmode: move to a separate library of its own 27467 27468 but not removed from the Makefile, which broke 'make dist'. 27469 27470 Signed-off-by: Adam Jackson <ajax@redhat.com> 27471 27472commit 47bc7fcaa567b0d9e2627b4a21b112e96f81725b 27473Author: Dave Airlie <airlied@gmail.com> 27474Date: Tue Feb 9 16:54:22 2016 +1000 27475 27476 present: fail flipping if we have any slave outputs 27477 27478 Due to the way present currently works, we don't ever check with the 27479 secondary adapters if we can flip at all. 27480 27481 We shouldn't flip if the secondary adapters are attached to the pixmap 27482 currently, however using the current check_flip callback isn't possible 27483 as it passes the Window to the driver (something we shouldn't be doing), 27484 so the slave driver can never get it's own screen ptr back. 27485 27486 For now to fix the problem just block flips if we have any slaves 27487 configured. We can fix the ABI up later, but this fix can be backported 27488 to stable. 27489 27490 Signed-off-by: Dave Airlie <airlied@redhat.com> 27491 Reviewed-by: Keith Packard <keithp@keithp.com> 27492 27493commit 316948734c2bba72d42c4cf4530932b5862c109f 27494Author: Michel Dänzer <michel.daenzer@amd.com> 27495Date: Fri Mar 11 12:21:06 2016 +0900 27496 27497 glamor: Make context current in glamor_pixmap_fbo_cache_put 27498 27499 Without this, we may be manipulating the context of another screen. 27500 27501 In a system with two GPUs using glamor, this fixes lots of 27502 27503 (EE) glamor256: GL error: GL_INVALID_OPERATION in glBindTexture(non-gen name) 27504 27505 spew since 0b4c0c75 ('glamor: Replace "finish access" shader with texture 27506 swizzling'). 27507 27508 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 27509 Reviewed-by: Dave Airlie <airlied@redhat.com> 27510 27511commit ef3005da3d5dc92b3ee5a0ee78164e739a3216dc 27512Author: Adam Jackson <ajax@redhat.com> 27513Date: Thu Jan 21 12:47:57 2016 -0500 27514 27515 glx: Implement GLX_EXT_fbconfig_packed_float 27516 27517 The tokens for this are already defined by GLX_ARB_fbconfig_float, which 27518 we already support, so just add the extension to the list and let the 27519 driver provide those configs if it wants. 27520 27521 Reviewed-by: Eric Anholt <eric@anholt.net> 27522 Signed-off-by: Adam Jackson <ajax@redhat.com> 27523 27524commit d15cb654b6ba365dac9a62064e277adebde2fdab 27525Author: Adam Jackson <ajax@redhat.com> 27526Date: Mon May 18 13:00:02 2015 -0400 27527 27528 glx: Implement GLX_EXT_stereo_tree 27529 27530 This is correct as it is, but only because we know no DRI drivers 27531 implement stereo. 27532 27533 v2: Use new ATTRIB macro 27534 27535 Reviewed-by: James Jones <jajones@nvidia.com> 27536 Reviewed-by: Eric Anholt <eric@anholt.net> 27537 Signed-off-by: Adam Jackson <ajax@redhat.com> 27538 27539commit a18238877bbf9aab95843d849a6f434275e9cd6c 27540Author: Adam Jackson <ajax@redhat.com> 27541Date: Thu Mar 3 16:50:02 2016 -0500 27542 27543 glx: Add GLX_SCREEN to the GetDrawableAttributes response 27544 27545 libglvnd would like to use this to map from drawable to screen, so it 27546 can know which driver to dispatch to. Refer to the spec proposal here: 27547 27548 https://lists.freedesktop.org/archives/mesa-dev/2016-March/109543.html 27549 27550 Reviewed-by: Eric Anholt <eric@anholt.net> 27551 Signed-off-by: Adam Jackson <ajax@redhat.com> 27552 27553commit 47c1d6b7abcfb1c6b478367bbc2e869c91485bc0 27554Author: Adam Jackson <ajax@redhat.com> 27555Date: Thu Mar 10 12:29:34 2016 -0500 27556 27557 glx: Macroize building the attribute list in DoGetDrawableAttributes 27558 27559 No functional change, just a little easier to read and harder to get 27560 wrong. 27561 27562 Reviewed-by: Eric Anholt <eric@anholt.net> 27563 Signed-off-by: Adam Jackson <ajax@redhat.com> 27564 27565commit c01094c5312fbd84146dd83122e5256a8e57d092 27566Author: Eric Anholt <eric@anholt.net> 27567Date: Mon Feb 1 13:58:15 2016 -0800 27568 27569 ephyr: Fix redisplay with glamor on GLES. 27570 27571 glamor_transfer.c is still totally broken, though. 27572 27573 Reviewed-by: Adam Jackson <ajax@redhat.com> 27574 Signed-off-by: Eric Anholt <eric@anholt.net> 27575 27576commit 0b4c0c75d06f3dbe92be1a26a637e9f05529cb3d 27577Author: Eric Anholt <eric@anholt.net> 27578Date: Mon Feb 1 13:58:14 2016 -0800 27579 27580 glamor: Replace "finish access" shader with texture swizzling. 27581 27582 For pictures without alpha, and for most other formats for GLES2, we 27583 would make a temporary FBO, make another temporary texture, upload our 27584 GLAMOR_MEMORY pixmap to the texture, then run the "finish access" shader 27585 across it to swizzle its values around into the temporary FBO (which we 27586 would use for a single Render operation and then throw away). 27587 27588 We can simplify everything by using GL_ARB_texture_swizzle (or its 27589 GLES3 counterpart). It's just not worth the complexity to try to 27590 improve the performance of this already low-performance path (SHM 27591 pixmaps + Render) on GLES2. 27592 27593 Reviewed-by: Adam Jackson <ajax@redhat.com> 27594 Signed-off-by: Eric Anholt <eric@anholt.net> 27595 27596commit b0cc04992ced5d96bb5c52fc1e5c868797cc0a17 27597Author: Eric Anholt <eric@anholt.net> 27598Date: Mon Feb 1 13:58:13 2016 -0800 27599 27600 glamor: Drop dead large-pixmap handling code in temp picture uploads. 27601 27602 The glamor_pixmap_ensure_fbo() in glamor_pixmap_upload_prepare() will 27603 always fail on a large pixmap, so we can just be explicit about 27604 bailing out here and then dump the rest of this garbage. 27605 27606 Reviewed-by: Adam Jackson <ajax@redhat.com> 27607 Signed-off-by: Eric Anholt <eric@anholt.net> 27608 27609commit 094b1bea8bd08eec029e4d61ba40d81441c37905 27610Author: Eric Anholt <eric@anholt.net> 27611Date: Mon Feb 1 13:58:12 2016 -0800 27612 27613 glamor: Drop unused PBO code in temporary picture uploading. 27614 27615 Reviewed-by: Adam Jackson <ajax@redhat.com> 27616 Signed-off-by: Eric Anholt <eric@anholt.net> 27617 27618commit 6112fecc3a4fd7dfb0ef77a98cfd1f7c91ccea0c 27619Author: Eric Anholt <eric@anholt.net> 27620Date: Mon Feb 1 13:58:11 2016 -0800 27621 27622 glamor: Generalize the a1-to-a8 conversion path. 27623 27624 Pixman is quite qualified to allocate our temporary memory, and all we 27625 need to do is decide what formats to convert from and to. 27626 27627 Reviewed-by: Adam Jackson <ajax@redhat.com> 27628 Signed-off-by: Eric Anholt <eric@anholt.net> 27629 27630commit 8f1411c3847a742f84ff07c4a0bb610801f5bfee 27631Author: Eric Anholt <eric@anholt.net> 27632Date: Mon Feb 1 13:58:10 2016 -0800 27633 27634 glamor: Drop the REVERT_UPLOADING_1_5_5_5 path. 27635 27636 There was only a pretty special case that could have even worked -- 27637 you've got a GLES2 renderer, you've got a SHM pixmap, it's 1555 (not 27638 the usual 565 for 16-bit), and you're little endian (BE was broken, 27639 since GL's 5_5_5_1 picks the 1 bit from the lowest bit of the short, 27640 and on BE we weren't doing the conversion path that swaps around the 27641 channels). This is just not worth the complexity. 27642 27643 Reviewed-by: Adam Jackson <ajax@redhat.com> 27644 Signed-off-by: Eric Anholt <eric@anholt.net> 27645 27646commit 2cc7a0815e5e2c2b1b1267cae9c348a8e95b1082 27647Author: Eric Anholt <eric@anholt.net> 27648Date: Mon Feb 1 13:58:09 2016 -0800 27649 27650 glamor: Drop the GLES2 REVERT_UPLOADING_2_10_10_10 paths. 27651 27652 These just smash your 2_10_10_10 data into 8888, despite what the 27653 comments said. That's not valid rendering, so just ditch this path 27654 and fall back to software. One might also note in the code being 27655 removed here that the REVERT_UPLOADING_10_10_10_2 path wasn't even 27656 connected. 27657 27658 Reviewed-by: Adam Jackson <ajax@redhat.com> 27659 Signed-off-by: Eric Anholt <eric@anholt.net> 27660 27661commit f667d5177024d3fdfb1b51694bdaeba6cee67962 27662Author: Eric Anholt <eric@anholt.net> 27663Date: Mon Feb 1 13:58:08 2016 -0800 27664 27665 glamor: Merge the two GL-type-from-pictformat paths. 27666 27667 It clarifies what the difference is between the two paths, and would 27668 potentially encourage us to handle GLES extensions that expose 27669 additional types. 27670 27671 Reviewed-by: Adam Jackson <ajax@redhat.com> 27672 Signed-off-by: Eric Anholt <eric@anholt.net> 27673 27674commit c7574c63c618d3a017105c380542eb04341b04a2 27675Author: Eric Anholt <eric@anholt.net> 27676Date: Mon Feb 1 13:58:07 2016 -0800 27677 27678 glamor: Propagate that is_upload is always true. 27679 27680 Reviewed-by: Adam Jackson <ajax@redhat.com> 27681 Signed-off-by: Eric Anholt <eric@anholt.net> 27682 27683commit 1bed5ef2b80c77c1bb9b62971367bea864fd8f66 27684Author: Eric Anholt <eric@anholt.net> 27685Date: Mon Feb 1 13:58:06 2016 -0800 27686 27687 glamor: Drop dead fbo handling from GLAMOR_MEMORY pict uploads. 27688 27689 The previous commit asserts that we don't have one. 27690 27691 Reviewed-by: Adam Jackson <ajax@redhat.com> 27692 Signed-off-by: Eric Anholt <eric@anholt.net> 27693 27694commit ee7ca670b1695d64bc12cb37302913acc066a569 27695Author: Eric Anholt <eric@anholt.net> 27696Date: Mon Feb 1 13:58:05 2016 -0800 27697 27698 glamor: Make sure that GLAMOR_MEMORY pixmaps don't retain an FBO. 27699 27700 glamor_composite_choose_shader() may upload our scratch pixmaps to get 27701 a Render operation completed. We don't want to hang onto GL memory 27702 for our scratch pixmaps, since we'll just have to reallocate them at a 27703 new w/h next time around, and the contents will be updated as well. 27704 27705 Reviewed-by: Adam Jackson <ajax@redhat.com> 27706 Signed-off-by: Eric Anholt <eric@anholt.net> 27707 27708commit a96c6d4658e3f386002f96eede660af3b01e5209 27709Author: Eric Anholt <eric@anholt.net> 27710Date: Mon Feb 1 13:58:04 2016 -0800 27711 27712 glamor: Simplify temporary picture uploading call stack. 27713 27714 glamor_upload_sub_pixmap_to_texture() only had the one caller, so we 27715 can merge it in, fix its silly return value, and propagate a bunch of 27716 constants. 27717 27718 Reviewed-by: Adam Jackson <ajax@redhat.com> 27719 Signed-off-by: Eric Anholt <eric@anholt.net> 27720 27721commit 25ce263fd88684be9370025f93ba3a2bfc72ff1a 27722Author: Olivier Fourdan <ofourdan@redhat.com> 27723Date: Wed Mar 9 16:45:18 2016 +0100 27724 27725 glamor: do not build Xv support when --disable-xv 27726 27727 Reviewed-by: Adam Jackson <ajax@redhat.com> 27728 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 27729 27730commit da7724d3d277c6c8a814881785b716896802629a 27731Author: Olivier Fourdan <ofourdan@redhat.com> 27732Date: Wed Mar 9 16:21:18 2016 +0100 27733 27734 xwayland: add glamor Xv adaptor 27735 27736 This adds an Xv adaptor using glamor. 27737 27738 Reviewed-by: Adam Jackson <ajax@redhat.com> 27739 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 27740 27741commit d11fdff50c91575e977a63617806a61bca98cd35 27742Author: Jon Turney <jon.turney@dronecode.org.uk> 27743Date: Fri Jan 3 13:21:40 2014 +0000 27744 27745 hw/xwin: Tidy-up of winmsg.h 27746 27747 - winVMsg() has no uses, so remove 27748 - winMsgVerb() has only one use, with default verbosity, so remove 27749 - winMsg() is identical to LogMessage() 27750 - Put winDrvMsg() and winDrvMsgVerb() under XWIN_XF86CONFIG 27751 - Include what you use Xfuncproto.h for _X_ATTRIBUTE_PRINTF 27752 27753 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27754 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27755 27756commit 519b98765f0c7d083a744ae7beb641753e4eb751 27757Author: Jon Turney <jon.turney@dronecode.org.uk> 27758Date: Thu Mar 3 22:32:28 2016 +0000 27759 27760 hw/xwin: Remove GC privates, unused since native GDI engine removal 27761 27762 Unused since native GDI engine removal in commit 8465ee78 27763 27764 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27765 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27766 27767commit 9d28ff2a9be86662f56463aa1fd46d12988e30fa 27768Author: Jon Turney <jon.turney@dronecode.org.uk> 27769Date: Fri Jul 31 20:23:59 2015 +0100 27770 27771 hw/xwin: Use NULL rather than NoopDDA for unimplemented engine functions 27772 27773 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27774 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27775 27776commit a309085a56de4d30dfbc44d9ff5302c7d9fdbf73 27777Author: Jon Turney <jon.turney@dronecode.org.uk> 27778Date: Fri Jul 31 20:20:00 2015 +0100 27779 27780 hw/xwin: Remove unused FinishCreateWindowsWindow engine function 27781 27782 This only ever had an (unused) implementation in the DDNL engine, which was 27783 removed in commit 57bbf6e2. 27784 27785 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27786 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27787 27788commit fa6f9d06a3d1a90134d9349f5ce4ec5b4eeff5e3 27789Author: Jon Turney <jon.turney@dronecode.org.uk> 27790Date: Fri Jul 31 20:12:37 2015 +0100 27791 27792 hw/xwin: Remove unused HotKeyAltTab engine function 27793 27794 This was only ever used by the primaryfb engine, removed in commit c79f824b 27795 27796 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27797 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27798 27799commit 7bd25aa8437ec410e7a7de1b0636ba33298fc8d6 27800Author: Jon Turney <jon.turney@dronecode.org.uk> 27801Date: Fri Jul 31 19:44:34 2015 +0100 27802 27803 hw/xwin: Return FALSE to indicate failure in winSetEngine() 27804 27805 Return FALSE to indicate failure in winSetEngine(), if it couldn't find a 27806 drawing engine to use 27807 27808 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27809 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27810 27811commit 52e05b9282f0f220d7c762793ce0b8f606a45deb 27812Author: Jon Turney <jon.turney@dronecode.org.uk> 27813Date: Tue Feb 23 23:09:43 2016 +0000 27814 27815 hw/xwin: Remove WM_WM_MAP message, which is now unused 27816 27817 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27818 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27819 27820commit c42217aa3d372acaa5ca7c64895edbfbd20c8475 27821Author: Jon Turney <jon.turney@dronecode.org.uk> 27822Date: Mon Nov 2 20:24:33 2015 +0000 27823 27824 hw/xwin: Remove WM_(UN|)MANAGE messages, which are now never sent 27825 27826 Remove fAnotherWMRunning which tracks this message (although since it was 27827 never initialized, I doubt this worked reliably), and the only use of that, 27828 which was to prevent winMWExtWMRestackWindows() from being used when the 27829 internalwm is running 27830 27831 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27832 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27833 27834commit b6bdf368420355332e41c604c523584bd39933f9 27835Author: Jon Turney <jon.turney@dronecode.org.uk> 27836Date: Mon Nov 2 20:51:51 2015 +0000 27837 27838 hw/xwin: Remove allowOtherWM, which is now always FALSE 27839 27840 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27841 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27842 27843commit 8407d3096287868b5c5e50ca5a98d470918c85a9 27844Author: Jon Turney <jon.turney@dronecode.org.uk> 27845Date: Mon Nov 2 20:34:09 2015 +0000 27846 27847 hw/xwin: Remove winIsInternalWMRunning(), which now always returns FALSE 27848 27849 Also remove then unused variables and IsRaiseonClick() 27850 27851 v2: 27852 Also remove unused pScreenInfo variable in winMWEXtWMRestackFrame() 27853 27854 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27855 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27856 27857commit 2779a28a86a13831b334e8678cd4e838b3b38472 27858Author: Jon Turney <jon.turney@dronecode.org.uk> 27859Date: Mon Nov 2 20:07:04 2015 +0000 27860 27861 hw/xwin: Remove fInternalWM flag 27862 27863 Remove the fInternalWM flag as it is now always FALSE after removing the 27864 -internalwm option 27865 27866 v2: 27867 Also remove then unused pRLWinPriv local from pRLWinPriv() 27868 27869 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27870 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27871 27872commit 98238ece5756801a8a67b9235e42cb9ab2318633 27873Author: Jon Turney <jon.turney@dronecode.org.uk> 27874Date: Mon Nov 2 19:37:40 2015 +0000 27875 27876 hw/xwin: Ignore the obsolete, undocumented -internalwm option 27877 27878 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27879 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27880 27881commit cdeaebad9889d81d2698b8a10fec5e55d8dec7a1 27882Author: Jon Turney <jon.turney@dronecode.org.uk> 27883Date: Tue Mar 1 21:37:05 2016 +0000 27884 27885 hw/xwin: Remove the long-broken -silent-dup-error option 27886 27887 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 27888 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 27889 27890commit 24042b4e367803dd64f3fcdc1bef7b2bf36c4145 27891Author: Michel Dänzer <michel.daenzer@amd.com> 27892Date: Mon Dec 21 17:54:08 2015 +0900 27893 27894 modesetting: Allow CRTC transforms to actually take effect 27895 27896 Setting crtc->transformPresent to FALSE was preventing the transform 27897 from actually taking effect and putting RandR into a confused state. 27898 27899 Now that the RandR 1.2 cursor code handles transforms correctly, we can 27900 allow them to properly take effect. 27901 27902 Reviewed-by: Keith Packard <keithp@keithp.com> 27903 27904commit b04767c84deafc44993723add4b1c5163fc11711 27905Author: Michel Dänzer <michel.daenzer@amd.com> 27906Date: Wed Oct 21 18:33:46 2015 +0900 27907 27908 xfree86: Re-set current cursor after RandR 1.2 CRTC configuration change 27909 27910 Add xf86CursorResetCursor, which allows switching between HW and SW 27911 cursor depending on the current state. 27912 27913 Call it from xf86DisableUnusedFunctions, which is called after any CRTC 27914 configuration change such as setting a mode or disabling a CRTC. This 27915 makes sure that SW cursor is used e.g. while a transform is in use on 27916 any CRTC or while there are active PRIME output slaves, and enables HW 27917 cursor again once none of those conditions are true anymore. 27918 27919 Reviewed-by: Keith Packard <keithp@keithp.com> 27920 27921commit a4ffa8721debb34bd36fd4624890d9c26886c618 27922Author: Michel Dänzer <michel.daenzer@amd.com> 27923Date: Thu Dec 24 12:56:03 2015 +0900 27924 27925 xfree86/modes: Check for CRTC transforms in xf86_use_hw_cursor(_argb) (v2) 27926 27927 We currently don't handle transforms for the HW cursor image, so return 27928 FALSE to signal a software cursor must be used if a transform is in use 27929 on any CRTC. 27930 27931 v2: Check crtc->transformPresent instead of crtc->transform_in_use. The 27932 latter is TRUE for rotation as well, which we handle correctly. 27933 27934 Reviewed-by: Keith Packard <keithp@keithp.com> 27935 27936commit c3e4e9fc5d84bfc17b3ed63f67488ea25ba150ce 27937Author: Michel Dänzer <michel.daenzer@amd.com> 27938Date: Thu Dec 24 16:20:49 2015 +0900 27939 27940 xfree86/modes: Refactor xf86_use_hw_cursor_argb to use xf86_use_hw_cursor (v2) 27941 27942 This reduces code duplication. 27943 27944 v2: No functional change this time. 27945 27946 Reviewed-by: Keith Packard <keithp@keithp.com> 27947 27948commit a3e681eafa5355b8bb3b099d47983f14f0d5e197 27949Author: Michel Dänzer <michel.daenzer@amd.com> 27950Date: Tue Feb 23 17:19:03 2016 +0900 27951 27952 glamor: Source pictures are always depth 32 27953 27954 We were using the destination pixmap depth to determine the source 27955 picture format. 27956 27957 Fixes incorrect text rendering with some MATE desktop GTK3 themes. 27958 27959 Reviewed-by: Adam Jackson <ajax@redhat.com> 27960 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94246 27961 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 27962 27963commit b05ae79ee3bebef9790c97eedc033d1ffb3ec39a 27964Author: Michel Dänzer <michel.daenzer@amd.com> 27965Date: Tue Feb 23 17:19:02 2016 +0900 27966 27967 glamor: Factor out glamor_set_color_depth from glamor_set_color 27968 27969 The former takes explicit screen and depth parameters instead of 27970 deriving them from a pixmap. 27971 27972 Reviewed-by: Adam Jackson <ajax@redhat.com> 27973 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 27974 27975commit 947e94a341fa153258e9e86060b83af95934672b 27976Author: Hans de Goede <hdegoede@redhat.com> 27977Date: Wed Feb 10 16:40:43 2016 +0100 27978 27979 glamor: Fix XvPutImage when src_y != 0 27980 27981 We already take src_y into account when uploading the src data by 27982 starting at the top line of the src data when uploading. 27983 27984 Adjust src_y accordingly when rendering. 27985 27986 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 27987 27988commit 40a164b8f4e720b0d6ebf228ee175eb397ffeec2 27989Author: Rui Matos <tiagomatos@gmail.com> 27990Date: Fri Mar 4 16:24:48 2016 +0100 27991 27992 build: Enable vidmode independently from Xorg 27993 27994 This allows building Xwayland without Xorg and still include the 27995 vidmode extension. 27996 27997 v2: Use PKG_CHECK_EXISTS instead of PKG_CHECK_MODULES 27998 27999 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 28000 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 28001 28002commit 2be527b1d4ce2b0412c4484539a8c9607645ec6d 28003Author: Julien Cristau <jcristau@debian.org> 28004Date: Mon Mar 7 23:20:34 2016 +0100 28005 28006 xfixes: avoid double free if AddResource fails 28007 28008 pChc is already freed through CursorFreeHideCount → 28009 deleteCursorHideCount. 28010 28011 Reviewed-by: Rémi Cardona <remi@gentoo.org> 28012 Signed-off-by: Julien Cristau <jcristau@debian.org> 28013 28014commit 4217db89ecd480fda2ee74fecba06c6713c2a0f0 28015Author: Julien Cristau <jcristau@debian.org> 28016Date: Mon Mar 7 23:20:33 2016 +0100 28017 28018 render: free already allocated formats in PictureInit failure case 28019 28020 Probably pointless, if this fails you're not likely to get far... 28021 28022 Reviewed-by: Rémi Cardona <remi@gentoo.org> 28023 Signed-off-by: Julien Cristau <jcristau@debian.org> 28024 28025commit 054f80717812d4781741cd05393623fe6f6c627f 28026Author: Julien Cristau <jcristau@debian.org> 28027Date: Mon Mar 7 23:20:32 2016 +0100 28028 28029 record: don't call RecordDeleteContext when AddResource fails 28030 28031 Reviewed-by: Rémi Cardona <remi@gentoo.org> 28032 Signed-off-by: Julien Cristau <jcristau@debian.org> 28033 28034commit d0c1a5bc61a3d151f2234aa3820862f16c0f00c7 28035Author: Julien Cristau <jcristau@debian.org> 28036Date: Mon Mar 7 23:20:31 2016 +0100 28037 28038 xwin: no need to free auth data if AddResource fails 28039 28040 This is taken care of by SecurityDeleteAuthorization 28041 28042 Reviewed-by: Rémi Cardona <remi@gentoo.org> 28043 Signed-off-by: Julien Cristau <jcristau@debian.org> 28044 28045commit acf263df81ad6813e0233033610fb44521cab1b4 28046Author: Julien Cristau <jcristau@debian.org> 28047Date: Mon Mar 7 23:20:30 2016 +0100 28048 28049 modesetting: avoid double free if AddResource fails 28050 28051 ms_dri2_frame_event_client_gone or ms_dri2_frame_event_drawable_gone 28052 already free the resource. 28053 28054 Reviewed-by: Rémi Cardona <remi@gentoo.org> 28055 Signed-off-by: Julien Cristau <jcristau@debian.org> 28056 28057commit 164753f158e78f615f903467bfd234d7c58244ef 28058Author: Julien Cristau <jcristau@debian.org> 28059Date: Mon Mar 7 23:20:29 2016 +0100 28060 28061 dmx/glxProxy: don't free the glx pixmap twice if AddResource fails 28062 28063 Reviewed-by: Rémi Cardona <remi@gentoo.org> 28064 Signed-off-by: Julien Cristau <jcristau@debian.org> 28065 28066commit 59b9c3d5e4bf05aeaaac2ee4ea12c301a67aae2c 28067Author: Julien Cristau <jcristau@debian.org> 28068Date: Mon Mar 7 23:20:28 2016 +0100 28069 28070 glx: don't call pGlxDraw->destroy() if AddResource fails 28071 28072 AddResource will have called DrawableGone, which takes care of the 28073 destruction. 28074 28075 Reviewed-by: Rémi Cardona <remi@gentoo.org> 28076 Signed-off-by: Julien Cristau <jcristau@debian.org> 28077 28078commit ac97fb2b804809c39b12fe0032d96fb076657258 28079Author: Julien Cristau <jcristau@debian.org> 28080Date: Mon Mar 7 23:20:27 2016 +0100 28081 28082 dri3: return an error if AddResource fails 28083 28084 Reviewed-by: Rémi Cardona <remi@gentoo.org> 28085 Signed-off-by: Julien Cristau <jcristau@debian.org> 28086 28087commit 119d5c0e2f800737c949ef760c5fe25d963200bf 28088Author: Julien Cristau <jcristau@debian.org> 28089Date: Mon Mar 7 23:20:26 2016 +0100 28090 28091 xvmc: Fix unchecked AddResource 28092 28093 Reviewed-by: Rémi Cardona <remi@gentoo.org> 28094 Signed-off-by: Julien Cristau <jcristau@debian.org> 28095 28096commit a2c3c34b44b866440a152511e682c98879ee13b7 28097Author: Jonas Ådahl <jadahl@gmail.com> 28098Date: Tue Mar 8 20:05:33 2016 +0800 28099 28100 xwayland: Correctly detect whether posix_fallocate exists 28101 28102 We had HAVE_POSIX_FALLOCATE checks, but no such macros were ever 28103 defined anywhere. This commit makes it so that this macro is defined if 28104 the posix_fallocate is detected during configure. 28105 28106 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 28107 Reviewed-by: Julien Cristau <jcristau@debian.org> 28108 28109commit 939ce0bae68b682b57675f65c901653c1a094ebb 28110Author: Julien Cristau <jcristau@debian.org> 28111Date: Tue Mar 1 21:39:01 2016 +0100 28112 28113 xv: fix double free in AddResource failure case 28114 28115 XvdiDestroyVideoNotifyList already frees the list if AddResource fails, 28116 so don't do it twice. And set tpn->client to NULL explicitly to avoid 28117 confusing uninitialized memory with a valid value. 28118 28119 Reviewed-by: Adam Jackson <ajax@redhat.com> 28120 Signed-off-by: Julien Cristau <jcristau@debian.org> 28121 28122commit 05e1bcf56e1c511a1ef539acfe11e37727e1179e 28123Author: Adam Jackson <ajax@redhat.com> 28124Date: Tue Mar 1 14:09:30 2016 -0500 28125 28126 dri1: Fix unchecked AddResource 28127 28128 Signed-off-by: Adam Jackson <ajax@redhat.com> 28129 Reviewed-by: Julien Cristau <jcristau@debian.org> 28130 28131commit 093f9505c12565cc19bdf6e33b263f31d104c3ef 28132Author: Adam Jackson <ajax@redhat.com> 28133Date: Tue Mar 1 14:09:29 2016 -0500 28134 28135 xv: Fix unchecked AddResource 28136 28137 Reviewed-by: Julien Cristau <jcristau@debian.org> 28138 Signed-off-by: Adam Jackson <ajax@redhat.com> 28139 28140commit ac4d8c7cee13947b688ebb26035f06f7744db201 28141Author: Olivier Fourdan <ofourdan@redhat.com> 28142Date: Tue Mar 1 17:03:44 2016 +0100 28143 28144 vidmode: build without xf86vidmodeproto 28145 28146 git commit f175cf45: 28147 28148 vidmode: move to a separate library of its own 28149 28150 introduced a regression where the xserver would not build when 28151 xf86vidmodeproto is not installed even if the configure option 28152 "--disable-xf86vidmode" is specified. 28153 28154 Fix build failure when xf86vidmodeproto is not installed. 28155 28156 Reviewed-by: Adam Jackson <ajax@redhat.com> 28157 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28158 28159commit 9c88cb9b059111e0531852f3fa8fa571c0306f57 28160Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28161Date: Thu Feb 25 16:37:57 2016 -0300 28162 28163 kdrive/ephyr: map host X server's keymap into Xephyr, if supported 28164 28165 Currently Xephyr doesn't inherit host X server's keymap, which 28166 may lead to keymap mismatches when using a non-US keyboard in a 28167 window inside Xephyr. This patch makes Xephyr change its keymap 28168 to match host X server's one (unless XKB support is disabled), 28169 using xcb-xkb to retrieve the needed XKB controls. 28170 This implementation is analogous to Xnest one at commit 83fef4235. 28171 28172 Supersedes: https://patchwork.freedesktop.org/patch/67504 28173 28174 Reviewed-by: Adam Jackson <ajax@redhat.com> 28175 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28176 28177commit daa6d2d58f65b9301b1b1f3c6df07719ecb5c03d 28178Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28179Date: Fri Feb 12 14:18:02 2016 -0200 28180 28181 config/udev: distinguish between real keyboards and other key devices 28182 28183 This patch introduces a new flag ATTR_KEY for hotplugged input devices, 28184 so we can better distinguish between real keyboards (i.e. devices with 28185 udev property ID_INPUT_KEYBOARD="1") and other key input devices like 28186 lid switches, power buttons, etc. 28187 28188 All supported hotplug backends (udev, hal, and wscons) will set both 28189 flags ATTR_KEY and ATTR_KEYBOARD for real keyboards, but udev backend 28190 will set ATTR_KEY, but not ATTR_KEYBOARD, for non-keyboard key input 28191 devices (hal and wscons will set both flags in any case). With this 28192 distinction, kdrive input hotplugging mechanism will be allowed to only 28193 grab real keyboards, as other key input devices are currently not 28194 supported. 28195 28196 In order to don't break current behaviour, this patch will replace all 28197 ATTR_KEYBOARD occurrences with ATTR_KEY in hw/xfree86/common/xf86Xinput.c. 28198 28199 [ajax: Just add ATTR_KEY, don't re-number the other attributes] 28200 28201 Reviewed-by: Adam Jackson <ajax@redhat.com> 28202 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28203 28204commit 851ff9ec04b73412c7dbad7b4911a1feac21f354 28205Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28206Date: Fri Feb 12 14:18:01 2016 -0200 28207 28208 ephyr: enable option -sw-cursor by default in multi-seat mode 28209 28210 Option -seat passed to Xephyr requires -sw-cursor to be passed as well, 28211 otherwise the mouse cursor will remain invisible for the given seat. 28212 This patch takes care of enabling -sw-cursor if -seat is passed. 28213 28214 Reviewed-by: Adam Jackson <ajax@redhat.com> 28215 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28216 28217commit 40e32e9fc9f3a1bd8287ee03dd399d8161cb98dd 28218Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28219Date: Fri Feb 12 14:18:00 2016 -0200 28220 28221 kdrive: add options to set default XKB properties 28222 28223 This patch introduces convenient command-line options -xkb-rules, 28224 -xkb-model, -xkb-layout, -xkb-variant, and -xkb-options, to set default 28225 values for these properties. 28226 28227 These options can be handful for cases in which compile-time default 28228 values don't match user locale, since kdrive doesn't support InputClass 28229 matching rules yet and not all Linux distros provide default rules to 28230 store these values in udev properties (which by the way is a discouraged 28231 practice). 28232 28233 Reviewed-by: Adam Jackson <ajax@redhat.com> 28234 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28235 28236commit 0cf3d72be6bd99cd2c66b7885339322c7e5bf73d 28237Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28238Date: Fri Feb 12 14:17:59 2016 -0200 28239 28240 kdrive: introduce input hot-plugging support for udev and hal backends (#33140) 28241 28242 This patch introduces input hot-plugging support for kdrive-based 28243 applications in multi-seat context. This feature is enabled by passing 28244 -seat option with desired seat name. All keyboard/mouse devices assigned 28245 to that seat will be automatically grabbed by kdrive. 28246 28247 It supports udev and hal backends for input hot-plugging support. 28248 Another patches may be required for wscons backend. 28249 28250 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=33140 28251 28252 Reviewed-by: Adam Jackson <ajax@redhat.com> 28253 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28254 28255commit 2116f03be04240e961649ca750a7aa5438b8446c 28256Author: Olivier Fourdan <ofourdan@redhat.com> 28257Date: Mon Feb 8 17:48:26 2016 +0100 28258 28259 xwayland: fix a crash on output removal 28260 28261 On output removal, the CRTC that was added in xwl_output_create() 28262 is not removed in xwl_output_destroy() and would cause a segmentation 28263 fault later on in ProcRRGetMonitors(): 28264 28265 (EE) Segmentation fault at address 0x100000001 28266 (EE) 28267 (EE) 10: ? (?+0x29) [0x29] 28268 (EE) 9: /usr/bin/Xwayland (_start+0x29) [0x423299] 28269 (EE) 8: /lib64/libc.so.6 (__libc_start_main+0xf0) [0x7fdd80e7f580] 28270 (EE) 7: /usr/bin/Xwayland (dix_main+0x3b3) [0x544ef3] 28271 (EE) 6: /usr/bin/Xwayland (Dispatch+0x31e) [0x54109e] 28272 (EE) 5: /usr/bin/Xwayland (ProcRRGetMonitors+0x9b) [0x4ca18b] 28273 (EE) 4: /usr/bin/Xwayland (RRMonitorMakeList+0x269) [0x4c9ba9] 28274 (EE) 3: /usr/bin/Xwayland (RRMonitorSetFromServer+0x118) [0x4c9198] 28275 (EE) 2: /usr/bin/Xwayland (MakeAtom+0x30) [0x530710] 28276 (EE) 1: /lib64/libc.so.6 (__restore_rt+0x0) [0x7fdd80e93b1f] 28277 (EE) 0: /usr/bin/Xwayland (OsSigHandler+0x29) [0x5792d9] 28278 28279 Remove the output CRTC in xwl_output_destroy() to avoid the crash. 28280 28281 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28282 Reviewed-by: Daniel Stone <daniels@collabora.com> 28283 28284commit 1bee4e254ca0305cb23e574b4c8b250d276ee998 28285Author: Michel Dänzer <michel.daenzer@amd.com> 28286Date: Thu Feb 18 17:33:19 2016 +0900 28287 28288 present: Call present_restore_screen_pixmap from present_set_abort_flip 28289 28290 After present_set_abort_flip, the screen pixmap will be used for all 28291 screen drawing, so we need to restore the current flip pixmap contents 28292 to the screen pixmap here as well. 28293 28294 Improves flashing / stutter e.g. when something like a popup menu appears 28295 on top of a flipping fullscreen window or when switching out of 28296 fullscreen. 28297 28298 Note that this means present_set_abort_flip now relies on screen->root 28299 being non-NULL, but that's already the case in other present code. 28300 28301 Reviewed-by: Keith Packard <keithp@keithp.com> 28302 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 28303 28304commit 4611e902c291b8a789f374cff3300f74645bc2b2 28305Author: Michel Dänzer <michel.daenzer@amd.com> 28306Date: Thu Feb 18 17:20:45 2016 +0900 28307 28308 present: Factor code for restoring screen pixmap out of present_unflip (v2) 28309 28310 The following fix will use the refactored function. 28311 28312 The logic in the refactored function is slightly simplified, exploiting 28313 the fact that this function is only ever called with a valid flip 28314 pixmap. 28315 28316 v2: Assert that flip_pixmap is non-NULL instead of testing and bailing 28317 on NULL, preserve test for flip_window being non-NULL before calling 28318 present_set_tree_pixmap for it (Keith Packard) 28319 28320 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> (v1) 28321 Reviewed-by: Keith Packard <keithp@keithp.com> 28322 28323commit 72328e5eb98a3f27e1f0a0e17beae6db447bd87c 28324Author: Michel Dänzer <michel.daenzer@amd.com> 28325Date: Thu Feb 18 18:23:47 2016 +0900 28326 28327 present: Only update screen pixmap from flip pixmap once per unflip 28328 28329 present_unflip may be called several times from present_check_flip_window 28330 during the same unflip. We can only copy to the screen pixmap the first 28331 time, otherwise we may scribble over other windows. The flip pixmap 28332 contents don't get updated after the first time anyway. 28333 28334 Fixes at least the following problems, which were introduced by commit 28335 806470b9 ("present: Copy unflip contents back to the Screen Pixmap"): 28336 28337 On xfwm4 without compositing, run glxgears and put its window into 28338 fullscreen mode to start flipping. While in fullscreen, open the xfwm4 28339 window menu by pressing Alt-Space. The window menu was invisible most 28340 of the time because it was getting scribbled over by a repeated unflip 28341 copy. 28342 28343 When switching a flipping window out of fullscreen, a repeated unflip 28344 copy could leave artifacts of the flip pixmap on the desktop. 28345 28346 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94325 28347 28348 Reviewed-by: Keith Packard <keithp@keithp.com> 28349 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 28350 28351commit 43eb5b6047c9b35c337e553ec054f08bdc835abb 28352Author: Michel Dänzer <michel.daenzer@amd.com> 28353Date: Tue Dec 8 12:52:17 2015 +0900 28354 28355 dri3: Refuse to work for remote clients (v2) 28356 28357 Prevents clients forwarded via SSH from hanging while waiting for the 28358 reply from the DRI3Open request. 28359 28360 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93261 28361 28362 v2: Return BadMatch instead of BadRequest (Keith Packard) 28363 28364 Reviewed-by: Keith Packard <keithp@keithp.com> 28365 28366commit 6070a749d953951bacbfb149c5c36451293aad35 28367Author: Olivier Fourdan <ofourdan@redhat.com> 28368Date: Wed Feb 10 09:35:39 2016 +0100 28369 28370 xwayland: add partial xvidmode extension support 28371 28372 Older games (mostly those based on SDL 1.x) rely on the XVidMode 28373 extension and would refuse to run without. 28374 28375 Add a simple, limited and read-only xvidmode support that reports the 28376 current mode used so that games that rely on xvidmode extension can run 28377 on XWayland. 28378 28379 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=87806 28380 Reviewed-by: Adam Jackson <ajax@redhat.com> 28381 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28382 28383commit b430f53bb753f9b064ab62d014820c1c3c76a841 28384Author: Olivier Fourdan <ofourdan@redhat.com> 28385Date: Fri Feb 5 09:48:25 2016 +0100 28386 28387 vidmode: remove redundant DIX function 28388 28389 The API signature of the DIX xf86VidModeGetGammaRampSize() is now 28390 identical to the xf86cmap's xf86GetGammaRampSize() and all it does is 28391 actually call xf86GetGammaRampSize() so we can save one vfunc. 28392 28393 Remove uneeded xf86VidModeGetGammaRampSize() function. 28394 28395 Reviewed-by: Adam Jackson <ajax@redhat.com> 28396 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28397 28398commit 48fccde2bfb60efdbf45a96fa53bcd9a6570bf89 28399Author: Olivier Fourdan <ofourdan@redhat.com> 28400Date: Fri Feb 5 09:48:24 2016 +0100 28401 28402 vidmode: remove redundant check 28403 28404 The DIX already checks for VidModePrivateKey to get the vfunc, so 28405 checking for this again in the DDX is redundant. 28406 28407 Remove the redundant function xf86VidModeAvailable() from the DDX. 28408 28409 Reviewed-by: Adam Jackson <ajax@redhat.com> 28410 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28411 28412commit f175cf45aebcdda53f3ae49c0eaf27da1f194e92 28413Author: Olivier Fourdan <ofourdan@redhat.com> 28414Date: Wed Feb 10 09:34:34 2016 +0100 28415 28416 vidmode: move to a separate library of its own 28417 28418 XVidMode extension might be useful to non hardware servers as well (e.g. 28419 Xwayand) so that applications that rely on it (e.g. lot of older games) 28420 can at least have read access to XVidMode. 28421 28422 But the implementation is very XFree86 centric, so the idea is to add 28423 a bunch of vfunc that other non-XFree86 servers can hook up into to 28424 provide a similar functionality. 28425 28426 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=87806 28427 Reviewed-by: Adam Jackson <ajax@redhat.com> 28428 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28429 28430commit 17097e083b2392c8989474f6e0da8cc234329e9c 28431Author: Olivier Fourdan <ofourdan@redhat.com> 28432Date: Fri Feb 5 09:48:22 2016 +0100 28433 28434 vidmode: rename DDX functions 28435 28436 To avoid confusion as to what belongs on the DDX and what not. 28437 28438 Reviewed-by: Adam Jackson <ajax@redhat.com> 28439 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28440 28441commit ddfb8c009ac651209eb0087aaf86b54e1446e8b2 28442Author: Olivier Fourdan <ofourdan@redhat.com> 28443Date: Fri Feb 5 09:48:21 2016 +0100 28444 28445 vidmode: move display mode definitions 28446 28447 To be able to reuse the VidMode extension in a non-hardware server, the 28448 display mode definitions need to be accessible from DIX. 28449 28450 Reviewed-by: Adam Jackson <ajax@redhat.com> 28451 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28452 28453commit e29a64de662112b8ebcd3f20c89df0e8c51890ef 28454Author: Olivier Fourdan <ofourdan@redhat.com> 28455Date: Fri Feb 5 09:48:20 2016 +0100 28456 28457 vidmode: remove mode access from public API 28458 28459 The mode access functions (namely VidModeCreateMode(), 28460 VidModeCopyMode(), VidModeGetModeValue() and VidModeSetModeValue()) are 28461 used only in xf86VidMode code and do not need to be available anywhere 28462 else. 28463 28464 Remove these functions from the public VidMode API and move them as 28465 static where they are used. 28466 28467 Reviewed-by: Adam Jackson <ajax@redhat.com> 28468 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28469 28470commit b7962ade5265a21ac7c60da6cc07ece15ef7e648 28471Author: Olivier Fourdan <ofourdan@redhat.com> 28472Date: Fri Feb 5 09:48:19 2016 +0100 28473 28474 vidmode: use appropriate DisplayModePtr type 28475 28476 The API uses an untyped pointer (void *) where a DisplayModePtr is 28477 expected. 28478 28479 Clean up the API to use the appropriate type, as DisplayModePtr is 28480 really all that will be passed there. 28481 28482 Reviewed-by: Adam Jackson <ajax@redhat.com> 28483 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28484 28485commit 12f714fd95dc9d912c0bf2524005a73ec6e8ee4f 28486Author: Olivier Fourdan <ofourdan@redhat.com> 28487Date: Fri Feb 5 09:48:18 2016 +0100 28488 28489 vidmode: remove VidModeGetMonitor() 28490 28491 VidModeGetMonitor() is used solely in ProcXF86VidModeGetMonitor() to 28492 get a untyped monitor pointer that is passed back straight again to 28493 VidModeGetMonitorValue(). 28494 28495 This is actually useless as VidModeGetMonitorValue() could as well get 28496 the monitor from the ScreenPtr just like VidModeGetMonitor() does. 28497 28498 Reviewed-by: Adam Jackson <ajax@redhat.com> 28499 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28500 28501commit f6f7e21133c13c34f306a191137d566e83b40929 28502Author: Olivier Fourdan <ofourdan@redhat.com> 28503Date: Fri Feb 5 09:48:17 2016 +0100 28504 28505 vidmode: use ScreenPtr instead of screen index 28506 28507 New code passes ScreenPtr instead of the screen index. 28508 28509 Change the VidMode functions to take a ScreenPtr. 28510 28511 Reviewed-by: Adam Jackson <ajax@redhat.com> 28512 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28513 28514commit 6e898ef080df93e885ead9d6fee8854b34e0216f 28515Author: Olivier Fourdan <ofourdan@redhat.com> 28516Date: Fri Feb 5 09:48:16 2016 +0100 28517 28518 vidmode: get rid of the CloseScreen wrapper 28519 28520 As we rely on dixRegisterPrivateKey() to allocate the memory for us that 28521 will be free automatically, we do not need the CloseScreen wrapper 28522 anymore. 28523 28524 Reviewed-by: Adam Jackson <ajax@redhat.com> 28525 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28526 28527commit 341f3bccafde71754a9ed2303df9908e509c6d31 28528Author: Olivier Fourdan <ofourdan@redhat.com> 28529Date: Fri Feb 5 09:48:15 2016 +0100 28530 28531 vidmode: use appropriate API 28532 28533 dixRegisterPrivateKey() can allocate memory that will be freed when the 28534 screen is teared down. 28535 28536 No need to calloc() and free the memory ourself using a broken ref 28537 counting method. 28538 28539 Reviewed-by: Adam Jackson <ajax@redhat.com> 28540 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 28541 28542commit d8882954570aba656d5a7be7d357feaba21cb099 28543Author: Chris Wilson <chris@chris-wilson.co.uk> 28544Date: Fri Feb 12 11:59:53 2016 +0000 28545 28546 dri2: Allow many blocked clients per-drawable 28547 28548 This patch was motivated by the need to fix the use-after-free in 28549 dri2ClientWake, but in doing so removes an arbitrary restriction that 28550 limits DRI2 to only blocking the first client on each drawable. In order 28551 to fix the use-after-free, we need to avoid touching our privates in the 28552 ClientSleep callback and so we want to only use that external list as 28553 our means of controlling sleeps and wakeups. We thus have a list of 28554 sleeping clients at our disposal and can manage multiple events and 28555 sources. 28556 28557 Reviewed-by: Adam Jackson <ajax@redhat.com> 28558 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 28559 28560commit bc3634010c096dffd1935c0c6cf8ba37534ae3d8 28561Author: Chris Wilson <chris@chris-wilson.co.uk> 28562Date: Fri Feb 12 11:59:52 2016 +0000 28563 28564 dix: Add ClientSignalAll() 28565 28566 This is a variant of ClientSignal() that signals all clients with an 28567 optional matching sleeping client, function and closure. 28568 28569 Reviewed-by: Adam Jackson <ajax@redhat.com> 28570 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 28571 28572commit 9fecc4cd571641f8920f25473ce5abb3fdca8b40 28573Author: Jon Turney <jon.turney@dronecode.org.uk> 28574Date: Tue Jan 12 16:18:35 2016 +0000 28575 28576 xwin: Remove unhelpful debug about WM message queue size 28577 28578 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28579 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28580 28581commit f7d1e5acdf5ed4ab4ed5c18727aa6f3d379ef560 28582Author: Jon Turney <jon.turney@dronecode.org.uk> 28583Date: Tue Aug 11 15:45:57 2015 +0100 28584 28585 xwin: Add SKIPTASKBAR hint to _NET_WM_WINDOW_TYPE_DOCK type windows 28586 28587 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28588 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28589 28590commit e7f87f8f76e5ac9479a71e3daf2cfdefd4b5f684 28591Author: Jon Turney <jon.turney@dronecode.org.uk> 28592Date: Thu Jul 2 19:21:11 2015 +0100 28593 28594 xwin: In multiwindow mode, look up the HWND for the parent window 28595 28596 Rather than only looking at the foreground window to see if it matches 28597 the WM_TRANSIENT_FOR window XID, lookup that XID and fetch the HWND from 28598 the window privates. 28599 28600 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28601 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28602 28603commit 856a28f63739bffe32f5a8156fd2680e5c5259a1 28604Author: Jon Turney <jon.turney@dronecode.org.uk> 28605Date: Thu Jul 2 13:32:40 2015 +0100 28606 28607 xwin: Factor out MessageName() debug helper 28608 28609 Factor out the MessageName() debug helper for message id -> text, and 28610 use it on message queue and dequeue. 28611 28612 Reorder in numerical order to match winwindow.h 28613 28614 Add missing WM_WM_ICON_EVENT 28615 28616 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28617 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28618 28619commit 8c2006ddc5abbd4ac374dabf1cfdd3df8fc88779 28620Author: Jon Turney <jon.turney@dronecode.org.uk> 28621Date: Mon Feb 22 17:41:14 2016 +0000 28622 28623 xwin: Keyboard layout updates 28624 28625 layout zh_TW doesn't exist (anymore), try something else for that. 28626 28627 layout it variant mac doesn't seem to exist anymore, try to handle 28628 Macintosh keyboards (running under Parallels on Mac) and other oddities 28629 in a more generic way, by falling back to matching only on the language 28630 identifer part of the input locale identifer. 28631 28632 v2: 28633 Fix typo of 0xa0000 for 0xa000 28634 28635 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28636 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28637 28638commit 8c97a0078e7fe22e6159fa53490dfca2f4d267a9 28639Author: Colin Harrison <colin.harrison@virgin.net> 28640Date: Wed Nov 19 22:33:39 2014 +0000 28641 28642 xwin: Add a tentative entry for the Korean keyboard to the list of known keyboard layouts 28643 28644 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 28645 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 28646 28647commit a4d8a64c4ba467964476c4a1486da698bd6aed9e 28648Author: Jon Turney <jon.turney@dronecode.org.uk> 28649Date: Tue Feb 10 14:36:37 2015 +0000 28650 28651 xwin: Update to XRANDR 1.2 internal interface to ensure an output is reported by XRANDR 28652 28653 If using the X server internal XRANDR 1.0 interface, it seems we must 28654 register a display size with RRRegisterSize() in order to make 28655 RRGetInfo() call RRScanOldConfig() to generate an output and crtc for 28656 us. 28657 28658 Without this, the following GDM bug means that an XDMCP session to GDM 28659 cannot be started. 28660 28661 https://bugzilla.gnome.org/show_bug.cgi?id=736054 28662 28663 Instead, use the more recent XRANDR 1.2 internal interface to explicitly 28664 create the output and crtc, and maintain a single mode which represents 28665 the current display size. 28666 28667 Also don't emit a RRScreenSizeNotify when a RRScreenSizeSize is done 28668 which has no effect, this seems to throw the GDM greeter into a loop... 28669 28670 v2: Maintain reference count for the mode we maintain more correctly, to 28671 avoid double free causing a crash on shutdown 28672 28673 Connect crtc to output, so a subsequent RRSetCrtcConfig request doesn't 28674 change anything, so we don't fail due to our lack of rrSetConfig or 28675 rrCrtcSet hooks. 28676 28677 See https://cygwin.com/ml/cygwin-xfree/2015-02/msg00032.html 28678 28679 v3: 28680 Raise limit on X display size from 4Kx4K to 32Kx32K 28681 28682 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28683 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28684 28685commit 008efebda801b9b80e2ab3f2c95aeef8c82102ee 28686Author: Jon Turney <jon.turney@dronecode.org.uk> 28687Date: Sat Oct 18 17:31:57 2014 +0100 28688 28689 xwin: Use WM_CLIPBOARDUPDATE clipboard API 28690 28691 Windows Vista and later have a saner clipboard API where the clipboard 28692 viewer linked list is no longer maintained by applications. Use it 28693 where available. 28694 28695 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28696 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28697 28698commit de7f1fd6f8f10f07b366ae5428a8c65a224bda98 28699Author: Jon Turney <jon.turney@dronecode.org.uk> 28700Date: Wed Jul 9 14:26:54 2014 +0100 28701 28702 xwin: Check that window position is visible on non-rectangular virtual desktops 28703 28704 Improve the check that window position is visible to work correctly for 28705 non-rectangular virtual desktops 28706 28707 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28708 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28709 28710commit a9e73131b6453e0fa2da5360e84af7a2eae3b205 28711Author: Jon Turney <jon.turney@dronecode.org.uk> 28712Date: Thu May 8 11:40:39 2014 +0100 28713 28714 xwin: Correctly interpret WM_HINTS, WM_NORMAL_HINTS properties on x86_64 28715 28716 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28717 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28718 28719commit d7cef6fbe23381b31c163468f349feee2f1b1eba 28720Author: Jon Turney <jon.turney@dronecode.org.uk> 28721Date: Thu Mar 13 18:04:17 2014 +0000 28722 28723 xwin: Improve handling of no-decoration motif hint 28724 28725 When motif decoration hint asks for no decoration, don't add sysmenu, 28726 mimimize or maximimize controls. 28727 28728 (This fixes a problem with e.g. fbpanel having a minimize control, but 28729 gtk's panel_configure_event() doesn't like the state we put the window 28730 into when we minimize it, causing it to spin) 28731 28732 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28733 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28734 28735commit f75404be3ad94c8da493688397712c65ea66cb90 28736Author: Jon Turney <jon.turney@dronecode.org.uk> 28737Date: Fri Mar 28 16:57:48 2014 +0000 28738 28739 xwin: XGetWMNormalHints() returns non-zero on success 28740 28741 XGetWMNormalHints() doesn't actually return a Status value. On success 28742 it returns a non-zero value, not Success. 28743 28744 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28745 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28746 28747commit 9dc32746f2cd161b512cce8f39c95287bdf1a3a3 28748Author: Jon Turney <jon.turney@dronecode.org.uk> 28749Date: Tue Nov 10 10:27:07 2015 +0000 28750 28751 xwin: Fix format warnings when ./configured --enable-debug --enable-windowswm 28752 28753 Fix format warnings (mainly pointer format fixes) which show up when 28754 ./configured --enable-debug --enable-windowswm 28755 28756 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28757 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 28758 28759commit b4ac7b142fa3c536e9b283cfd34b94d82c03aac6 28760Author: Michel Dänzer <michel.daenzer@amd.com> 28761Date: Wed Feb 24 16:52:59 2016 +0900 28762 28763 present: Only requeue if target MSC is not reached after an unflip 28764 28765 While present_pixmap decrements target_msc by 1 for present_queue_vblank, 28766 it leaves the original vblank->target_msc intact. So incrementing the 28767 latter for requeueing resulted in the requeued presentation being 28768 executed too late. 28769 28770 Also, no need to requeue if the target MSC is already reached. 28771 28772 This further reduces stutter when a popup menu appears on top of a 28773 flipping fullscreen window. 28774 28775 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 28776 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 28777 28778commit e7a35b9e16aa12970908f5d55371bb1b862f8f24 28779Author: Michel Dänzer <michel.daenzer@amd.com> 28780Date: Wed Feb 24 16:52:58 2016 +0900 28781 28782 present: Requeue if flip driver hook fails and target MSC not reached 28783 28784 For flipping, we wait for the MSC before the target MSC and then call 28785 the driver flip hook. If the latter fails, we have to wait for the 28786 target MSC before falling back to a copy, or else it's executed too 28787 early. 28788 28789 Fixes glxgears running at unbounded framerate (not synchronized to the 28790 refresh rate) in fullscreen if the driver flip hook fails. 28791 28792 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 28793 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 28794 28795commit 1a9f8c4623c4e6b6955cb6d5f44d29c244dfd32a 28796Author: Michel Dänzer <michel.daenzer@amd.com> 28797Date: Wed Feb 24 16:52:57 2016 +0900 28798 28799 present: Move msc_is_(equal_or_)after to the top of present.c 28800 28801 To make them usable from any other function in the file. No functional 28802 change. 28803 28804 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 28805 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 28806 28807commit 0461bca0cb2f7918c77ed45d2cbc756cf65021be 28808Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28809Date: Mon Feb 22 16:04:12 2016 -0300 28810 28811 kdrive/evdev: update keyboard LEDs (#22302) 28812 28813 Implement missing parts in kdrive evdev driver for 28814 correct update of evdev keyboard LEDs. 28815 28816 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=22302 28817 28818 [ajax: Fixed deref-before-null-check bug] 28819 28820 Reviewed-by: Adam Jackson <ajax@redhat.com> 28821 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 28822 28823commit 30b7d7995ef70b7473e0fb170eb8ae23b8d1f4a7 28824Author: Jon TURNEY <jon.turney@dronecode.org.uk> 28825Date: Mon Feb 22 16:20:00 2016 +0000 28826 28827 Fix build on Cygwin by ensuring WIN32 check triggers only on MinGW 28828 28829 The type of fd_mask was changed in Cygwin 2.4.0 headers from 'long' to 28830 'unsigned long'. This exposes an existing problem with winauth.c, which 28831 includes Xwindows.h (which includes windows.h, which defines WIN32), 28832 before including osdep.h, which causes the now conflicting definition of 28833 fd_mask in osdep.h to be exposed: 28834 28835 In file included from ../os/osdep.h:198:18: error: conflicting types for 28836 ‘fd_mask’ typedef long int fd_mask; /usr/include/sys/select.h:46:23: 28837 note: previous declaration of ‘fd_mask’ was here typedef unsigned long 28838 fd_mask; 28839 28840 Adjust the include guards in osdep.h to make sure we only use WIN32 28841 guarded code when not compiling for Cygwin (i.e. WIN32 && !__CYGWIN__) 28842 28843 This isn't a very elegant, but unfortunately appears to be the best 28844 solution, since it doesn't seem to be possible to write the test in a 28845 positive form. 28846 28847 Future work: Should also audit of all the other uses of WIN32 in 28848 xserver, and make sure they are correct. 28849 28850 Reviewed-by: Adam Jackson <ajax@redhat.com> 28851 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 28852 28853commit 544b4149268561d3d794aa540172831fa7550a20 28854Author: Jonas Ådahl <jadahl@gmail.com> 28855Date: Fri Feb 19 15:08:12 2016 +0800 28856 28857 xwayland: Prefix shm tmp file names with xwayland 28858 28859 Prefix the temporary file names used for allocating pixmaps with 28860 "xwayland-" instead of "weston-". This makes it less confusing while 28861 looking at the file names of the currently open fds of the Xwayland 28862 process. 28863 28864 Reviewed-by: Adam Jackson <ajax@redhat.com> 28865 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 28866 28867commit 5627708e5f447d0e360bbc9eb8d1c4e58c5046d0 28868Author: Marc-Andre Lureau <marcandre.lureau@gmail.com> 28869Date: Fri Feb 12 22:52:07 2016 +0100 28870 28871 dri2: add virtio-gpu pci ids 28872 28873 Add virtio-gpu legacy + 1.0 pci ids, allowing them to use 28874 modesetting + glamor with dri2. 28875 28876 Signed-off-by: Marc-André Lureau <marcandre.lureau@gmail.com> 28877 Reviewed-by: Dave Airlie <airlied@redhat.com> 28878 28879commit e957a2e5dd288f515f3e93724823542c20333f6a 28880Author: Daniel Stone <daniels@collabora.com> 28881Date: Fri Feb 12 16:36:59 2016 +0000 28882 28883 dix: Add hybrid full-size/empty-clip mode to SetRootClip 28884 28885 216bdbc735 removed the SetRootClip call in the XWayland output-hotplug 28886 handler when running rootless (e.g. as a part of Weston/Mutter), since 28887 the root window has no storage, so generating exposures will result in 28888 writes to invalid memory. 28889 28890 Unfortunately, preventing the segfault also breaks sprite confinement. 28891 SetRootClip updates winSize and borderSize for the root window, which 28892 when combined with RRScreenSizeChanged calling ScreenRestructured, 28893 generates a new sprite-confinment area to update it to the whole screen. 28894 28895 Removing this call results in the window geometry being reported 28896 correctly, but winSize/borderSize never changing from their values at 28897 startup, i.e. out of sync with the root window geometry / screen 28898 information in the connection info / XRandR. 28899 28900 This patch introduces a hybrid mode, where we update winSize and 28901 borderSize for the root window, enabling sprite confinement to work 28902 correctly, but keep the clip emptied so exposures are never generated. 28903 28904 Signed-off-by: Daniel Stone <daniels@collabora.com> 28905 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 28906 Reviewed-by: Adam Jackson <ajax@redhat.com> 28907 28908commit dbe8d03c42f01332b3dc41fe9290aed142f1436f 28909Author: Keith Packard <keithp@keithp.com> 28910Date: Tue Feb 16 20:03:57 2016 -0800 28911 28912 randr: Send ConfigNotify when manual monitor list changes 28913 28914 This lets clients know that the layout of the monitors on the screen 28915 has changed so they can adapt appropriately. 28916 28917 Signed-off-by: Keith Packard <keithp@keithp.com> 28918 Reviewed-by: Dave Airlie <airlied@redhat.com> 28919 28920commit eddf848c44349c7ebc7da9957bffb3630f3faaa9 28921Author: Adam Jackson <ajax@redhat.com> 28922Date: Wed Feb 10 11:13:04 2016 -0500 28923 28924 dri2: Use the work queue to manage client sleeps 28925 28926 In commit e43abdce964f5ed9689cf908af8c305b39a5dd36 28927 Author: Chris Wilson <chris@chris-wilson.co.uk> 28928 Date: Wed Feb 3 09:54:46 2016 +0000 28929 28930 dri2: Unblock Clients on Drawable release 28931 28932 we try to wake up any blocked clients at drawable destruction. But by 28933 the time we get there, CloseDownConnection has already torn down state 28934 that AttendClient wants to modify. 28935 28936 Using ClientSleep instead of IgnoreClient puts a wakeup function on a 28937 workqueue, and the queue will be cleared for us in CloseDownClient 28938 before (non-neverretain) resource teardown. 28939 28940 Tested-by: Chris Wilson <chris@chris-wilson.co.uk> 28941 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 28942 Signed-off-by: Adam Jackson <ajax@redhat.com> 28943 28944commit b3e9c534e2b0dc2c9acd2fe9b942e1fc5227339b 28945Author: Adam Jackson <ajax@redhat.com> 28946Date: Mon Feb 8 17:36:03 2016 -0500 28947 28948 os: unifdef STREAMSCONN 28949 28950 Removed from xtrans in 2012, and never wired up in the modular build 28951 anyway. 28952 28953 Signed-off-by: Adam Jackson <ajax@redhat.com> 28954 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 28955 28956commit e43abdce964f5ed9689cf908af8c305b39a5dd36 28957Author: Chris Wilson <chris@chris-wilson.co.uk> 28958Date: Wed Feb 3 09:54:46 2016 +0000 28959 28960 dri2: Unblock Clients on Drawable release 28961 28962 If the Window is destroyed by another client, such as the window 28963 manager, the original client may be blocked by DRI2 awaiting a vblank 28964 event. When this happens, DRI2DrawableGone forgets to unblock that 28965 client and so the wait never completes. 28966 28967 Note Present/xshmfence is also suspectible to this race. 28968 28969 Testcase: dri2-race/manager 28970 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 28971 Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> 28972 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> 28973 28974commit a722d617a092f08f69086630f5cfb598d4a21cc7 28975Author: Adam Jackson <ajax@redhat.com> 28976Date: Wed Dec 9 09:10:13 2015 -0500 28977 28978 autogen: Set a default subject prefix for patches 28979 28980 Per discussion at XDC2015, we want this so we can easily distinguish 28981 which module a patch is for. There's no way to set this in the 28982 server-side config, so setting a default at autogen time is about the 28983 best we can do. 28984 28985 Reviewed-by: Eric Anholt <eric@anholt.net> 28986 Signed-off-by: Adam Jackson <ajax@redhat.com> 28987 28988commit 7bb64d8c1de9659f11da7917772919b071e9db82 28989Author: Jan Burgmeier <jan.burgmeier@unicon-software.com> 28990Date: Thu Feb 4 14:06:43 2016 +0100 28991 28992 Fix XineramaQueryScreens for reverse prime 28993 28994 Make sure we account for slave CRTCs when building the monitor list, 28995 since that's what rrxinerama uses to fake Xinerama geometry. 28996 28997 [ajax: Slightly more informative commit message.] 28998 28999 Bugzilla: https://bugs.freedesktop.org/92313 29000 Reviewed-by: Dave Airlie <airlied@redhat.com> 29001 29002commit 87d5534f701242d7c23aa20545a6292a0779c89c 29003Author: Rui Matos <tiagomatos@gmail.com> 29004Date: Wed Feb 3 16:14:09 2016 +0100 29005 29006 xwayland: Clear pending cursor frame callbacks on pointer enter 29007 29008 The last cursor frame we commited before the pointer left one of our 29009 surfaces might not have been shown. In that case we'll have a cursor 29010 surface frame callback pending which we need to clear so that we can 29011 continue submitting new cursor frames. 29012 29013 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 29014 Reviewed-by: Bryce Harrington <bryce@osg.samsung.com> 29015 Acked-by: Pekka Paalanen <pekka.paalanen@collabora.co.uk> 29016 Reviewed-by: Daniel Stone <daniels@collabora.com> 29017 Reviewed-by: Jonas Ådahl <jadahl@gmail.com> 29018 29019commit b7d392931a309d0fe754869efb456ccd0385f3aa 29020Author: Chris Wilson <chris@chris-wilson.co.uk> 29021Date: Wed Feb 3 09:54:43 2016 +0000 29022 29023 dri2: Only invalidate the immediate Window upon SetWindowPixmap 29024 29025 All callers of SetWindowPixmap will themselves be traversing the Window 29026 heirarchy updating the backing Pixmap of each child and so we can forgo 29027 doing the identical traversal inside the DRI2SetWindowPixmap handler. 29028 29029 Reported-by: Loïc Yhuel <loic.yhuel@gmail.com> 29030 Link: http://lists.x.org/archives/xorg-devel/2015-February/045638.html 29031 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 29032 Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> 29033 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> 29034 29035commit da69f2f15a9917c3ed04e305061683c41338126e 29036Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29037Date: Fri Dec 11 11:43:14 2015 -0200 29038 29039 ephyr: don't load ephyr input driver if -seat option is passed 29040 29041 When used for single-GPU multi-seat purposes, there's no need to enable 29042 ephyr virtual input devices, since Xephyr is supposed to handle its own 29043 hardware devices. 29044 29045 Reviewed-by: Adam Jackson <ajax@redhat.com> 29046 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29047 29048commit 7213e99cbc38a60f0076bc2115b144798ea4c3ba 29049Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29050Date: Fri Dec 11 11:43:13 2015 -0200 29051 29052 ephyr: ignore Xorg multiseat command line options 29053 29054 Multi-seat-capable display managers commonly pass command-line options 29055 like "-novtswitch", "-sharevts", or "-layout seatXXXX" to Xorg server, 29056 but Xephyr currently refuses to start if these options are passed to it, 29057 which may break Xephyr-based single-GPU multiseat setups. 29058 29059 [ajax: shortened summary] 29060 29061 Reviewed-by: Adam Jackson <ajax@redhat.com> 29062 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29063 29064commit edd443f69ea385a957b8eae0b066ad8e77fb947b 29065Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29066Date: Fri Dec 11 11:43:11 2015 -0200 29067 29068 kdrive: don't let evdev driver overwrite existing device names 29069 29070 KDrive evdev driver deliberately name grabbed devices as "Evdev mouse" 29071 or "Evdev keyboard". This patch will make it skip this step if grabbed 29072 devices are already named (i.e. from udev). 29073 29074 Reviewed-by: Adam Jackson <ajax@redhat.com> 29075 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29076 29077commit 0b80da0d18713df4712c05973388b6226bb0227f 29078Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29079Date: Fri Dec 11 11:43:07 2015 -0200 29080 29081 kdrive: set "evdev" driver for input devices automatically, if available. 29082 29083 If kdrive input driver "evdev" is available, no other driver was 29084 explicitly set for a given input device, and its kernel device node is 29085 /dev/input/event*, this patch will make kdrive set "evdev" driver 29086 automatically for such device. 29087 29088 Reviewed-by: Adam Jackson <ajax@redhat.com> 29089 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29090 29091commit 6d6fd688ecf95f2e84f2af276d681ff42f9d5610 29092Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29093Date: Fri Dec 11 11:43:06 2015 -0200 29094 29095 kdrive: fix up NewInputDeviceRequest() implementation 29096 29097 This patch simplifies NewInputDeviceRequest() implementation in 29098 kinput.c, making use of improved KdParseKbdOptions() / 29099 KdParsePointerOptions() and merging several "if (ki)"/"if (pi)" clauses. 29100 29101 Reviewed-by: Adam Jackson <ajax@redhat.com> 29102 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29103 29104commit 2c3e8768443caa66d78775ea79bb16a5faae3a3c 29105Author: Eric Anholt <eric@anholt.net> 29106Date: Wed Jan 27 11:26:16 2016 -0800 29107 29108 glamor: Flip around conditionals in RepeatNone fixups. 29109 29110 Signed-off-by: Eric Anholt <eric@anholt.net> 29111 Reviewed-by: Dave Airlie <airlied@redhat.com> 29112 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29113 29114commit e82c8c81df80de487aa795d69e874a0811c537ea 29115Author: Eric Anholt <eric@anholt.net> 29116Date: Wed Jan 27 11:21:05 2016 -0800 29117 29118 glamor: Cut down a bunch of conditional handling for RepeatFix. 29119 29120 For hardware that doesn't do actual jumps for conditionals (i915, 29121 current vc4 driver), this reduces the number of texture fetches 29122 performed (assuming the driver isn't really smart about noticing that 29123 the same sampler is used on each side of an if just with different 29124 coordinates). 29125 29126 No performance difference on i965 with x11perf -magpixwin100 (n=40). 29127 Improves -magpixwin100 by 12.9174% +/- 0.405272% (n=5) on vc4. 29128 29129 Signed-off-by: Eric Anholt <eric@anholt.net> 29130 Reviewed-by: Dave Airlie <airlied@redhat.com> 29131 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29132 29133commit 2c3273861cdf874b165ce5a1953102187f71b48e 29134Author: Eric Anholt <eric@anholt.net> 29135Date: Wed Jan 27 11:15:27 2016 -0800 29136 29137 glamor: Clarify how the repeat values being passed around work. 29138 29139 Signed-off-by: Eric Anholt <eric@anholt.net> 29140 Reviewed-by: Dave Airlie <airlied@redhat.com> 29141 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29142 29143commit 990a8ee01324332ee9b4a4bb124ce8f73be24349 29144Author: Eric Anholt <eric@anholt.net> 29145Date: Wed Jan 27 11:10:14 2016 -0800 29146 29147 glamor: Clean up formatting of RepeatFix shader code. 29148 29149 All sorts of weird indentation, and some cuddled conditional 29150 statements deep in the if tree. 29151 29152 Signed-off-by: Eric Anholt <eric@anholt.net> 29153 Reviewed-by: Dave Airlie <airlied@redhat.com> 29154 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29155 29156commit 20cb5b2d65ce63ea7934b77f1520387550c778a8 29157Author: Eric Anholt <eric@anholt.net> 29158Date: Wed Jan 27 10:24:17 2016 -0800 29159 29160 glamor: Clarify some logic in RepeatFix handling. 29161 29162 wh ratios are != 1.0 only when large, so with that we can simplify 29163 down how we end up with RepeatFix being used. 29164 29165 Signed-off-by: Eric Anholt <eric@anholt.net> 29166 Reviewed-by: Dave Airlie <airlied@redhat.com> 29167 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29168 29169commit 07f0d90e4a8b05ef968b1ef47acda7c9f4580340 29170Author: Eric Anholt <eric@anholt.net> 29171Date: Tue Jan 26 17:29:48 2016 -0800 29172 29173 glamor: Simplify the pixmap box looping. 29174 29175 We had a double loop across h and w, and passed the current x and y 29176 out to callers who then used w to multiply/add to an index. Instead, 29177 just single loop across w * h. 29178 29179 Signed-off-by: Eric Anholt <eric@anholt.net> 29180 Reviewed-by: Dave Airlie <airlied@redhat.com> 29181 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29182 29183commit 0dbce65b08f4812dcaa4b77cd37aebac334c47a2 29184Author: Eric Anholt <eric@anholt.net> 29185Date: Tue Jan 26 15:56:27 2016 -0800 29186 29187 glamor: Reuse the glamor_program_alpha_* enums for Render. 29188 29189 This is a step toward using glamor_program.c for Render acceleration. 29190 29191 Signed-off-by: Eric Anholt <eric@anholt.net> 29192 Reviewed-by: Dave Airlie <airlied@redhat.com> 29193 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29194 29195commit 9b676786de32f06aedf9d4c9535c10fda247335a 29196Author: Eric Anholt <eric@anholt.net> 29197Date: Tue Jan 26 15:47:01 2016 -0800 29198 29199 glamor: Drop extra SHADER_IN type for no mask present. 29200 29201 We can just hand in a constant mask and the driver will optimize away 29202 the multiplication for us. 29203 29204 Signed-off-by: Eric Anholt <eric@anholt.net> 29205 Reviewed-by: Dave Airlie <airlied@redhat.com> 29206 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29207 29208commit 03f34f85563c81e1655626e10f75fd7e21393c92 29209Author: Eric Anholt <eric@anholt.net> 29210Date: Tue Jan 26 15:19:50 2016 -0800 29211 29212 glamor: Convert XV to using glamor_program.c. 29213 29214 One less custom path! By following the common glamor_program.c use 29215 pattern, we get the ability to handle large pixmaps as the 29216 destination. It's also one less place where glamor_utils.h coordinate 29217 transformation happens. 29218 29219 Signed-off-by: Eric Anholt <eric@anholt.net> 29220 Reviewed-by: Dave Airlie <airlied@redhat.com> 29221 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29222 29223commit f368a0ba3aa58e5260d839d11d2f3aef75feaeaf 29224Author: Eric Anholt <eric@anholt.net> 29225Date: Tue Jan 26 14:52:08 2016 -0800 29226 29227 glamor: Simplify XV vertex setup. 29228 29229 We were clipping the drawn rectangle to each clip box, then expanding 29230 the box to a big triangle to avoid tearing, then drawing each triangle 29231 to the destination through a scissor. If we're using a scissor for 29232 clipping, though, then we don't need to clip the drawn primitive on 29233 the CPU in the first place. 29234 29235 Signed-off-by: Eric Anholt <eric@anholt.net> 29236 Reviewed-by: Dave Airlie <airlied@redhat.com> 29237 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29238 29239commit 294e45b60d99cf7d11c657288bbe2670b56775f3 29240Author: Eric Anholt <eric@anholt.net> 29241Date: Tue Jan 26 14:18:48 2016 -0800 29242 29243 glamor: Set up XV sampler uniforms once at program build time. 29244 29245 No sense doing it on every draw. 29246 29247 Signed-off-by: Eric Anholt <eric@anholt.net> 29248 Reviewed-by: Dave Airlie <airlied@redhat.com> 29249 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29250 29251commit 5d7bef2eedfd965359dd4eebb6ab806cdad5b83f 29252Author: Eric Anholt <eric@anholt.net> 29253Date: Tue Jan 26 13:39:18 2016 -0800 29254 29255 glamor: Drop dead glamor_pict_format_is_compatible(). 29256 29257 This hasn't been used since 2f80c7791bb0b11f261cb1e3e0d89163dcdd0342 29258 (GLAMOR_SEPARATE_TEXTURE removal). 29259 29260 Signed-off-by: Eric Anholt <eric@anholt.net> 29261 Reviewed-by: Dave Airlie <airlied@redhat.com> 29262 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29263 29264commit 4494a450405cf539743cbcfe6907bf5bdd2d80cb 29265Author: Eric Anholt <eric@anholt.net> 29266Date: Tue Jan 26 15:08:17 2016 -0800 29267 29268 glamor: Drop comment about dead yInverted flag. 29269 29270 Wait long enough, and you don't need to think about it at all. 29271 29272 Signed-off-by: Eric Anholt <eric@anholt.net> 29273 Reviewed-by: Dave Airlie <airlied@redhat.com> 29274 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29275 29276commit f7c24e6ac345aab91df5fc959f239a33f37113b1 29277Author: Eric Anholt <eric@anholt.net> 29278Date: Tue Jan 26 13:34:00 2016 -0800 29279 29280 glamor: Rename the *y_inverted helpers to not say "inverted". 29281 29282 Signed-off-by: Eric Anholt <eric@anholt.net> 29283 Reviewed-by: Dave Airlie <airlied@redhat.com> 29284 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29285 29286commit 1fcb6f4cbf3d6514716435a0e79c0e6d53c31a3a 29287Author: Eric Anholt <eric@anholt.net> 29288Date: Tue Jan 26 13:31:59 2016 -0800 29289 29290 glamor: Drop dead *_from_x_coord_y() functions. 29291 29292 They've been dead since the yInverted removal 29293 (e310387f443b6333edf02c8980daa303505382b4). 29294 29295 Signed-off-by: Eric Anholt <eric@anholt.net> 29296 Reviewed-by: Dave Airlie <airlied@redhat.com> 29297 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29298 29299commit 9ef11f13af7f552dadb4a90c248e525a257e0a2c 29300Author: Eric Anholt <eric@anholt.net> 29301Date: Thu Jan 21 16:01:14 2016 -0800 29302 29303 glamor: Clarify when Render fallbacks happen due to an unsupported op. 29304 29305 Signed-off-by: Eric Anholt <eric@anholt.net> 29306 Reviewed-by: Dave Airlie <airlied@redhat.com> 29307 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29308 29309commit b8229cc5f5298a37a4735dd002b0e0ebfc8bc75a 29310Author: Eric Anholt <eric@anholt.net> 29311Date: Wed Jan 27 11:35:03 2016 -0800 29312 29313 glamor: Label programs before linking them. 29314 29315 i965 does most of its compiling at link time, so our debug output for 29316 its shaders didn't have the name on. 29317 29318 Signed-off-by: Eric Anholt <eric@anholt.net> 29319 Reviewed-by: Dave Airlie <airlied@redhat.com> 29320 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29321 29322commit 68f236ebd4b268a9e525d623986999d230feb453 29323Author: Eric Anholt <eric@anholt.net> 29324Date: Wed Jan 27 16:11:17 2016 -0800 29325 29326 ephyr: Make sure we have GLX_ARB_create_context before calling it. 29327 29328 This should fix aborts()s from epoxy on old software stacks. 29329 29330 Signed-off-by: Eric Anholt <eric@anholt.net> 29331 Reviewed-by: Dave Airlie <airlied@redhat.com> 29332 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 29333 29334commit 623ff251dd025929f5bb6174ca86580c5e707261 29335Author: Adam Jackson <ajax@redhat.com> 29336Date: Tue Dec 8 17:41:38 2015 -0500 29337 29338 xephyr: Remove DRI1 29339 29340 This only worked if the backend server supported DRI1, which is 29341 stunningly unlikely these days. 29342 29343 Signed-off-by: Adam Jackson <ajax@redhat.com> 29344 Reviewed-by: Eric Anholt <eric@anholt.net> 29345 29346commit 953b71270cf12c59c8a836c9be403d07fb01fa22 29347Author: Adam Jackson <ajax@redhat.com> 29348Date: Tue Dec 8 17:41:37 2015 -0500 29349 29350 xfree86: Build parser for DRI config file subsection unconditionally 29351 29352 This applies regardless of which DRI you're asking for. Worse, leaving 29353 it out means breaking the config file syntax in a pointless way, since 29354 non-DRI servers can safely just parse it and ignore it. 29355 29356 Reviewed-by: Eric Anholt <eric@anholt.net> 29357 Signed-off-by: Adam Jackson <ajax@redhat.com> 29358 29359commit 1a48a5863e4bceee6b99c1feda1b2c584ff3657c 29360Author: Adam Jackson <ajax@redhat.com> 29361Date: Tue Dec 8 17:41:36 2015 -0500 29362 29363 xfree86: Remove ancient DRI build instructions 29364 29365 Reviewed-by: Eric Anholt <eric@anholt.net> 29366 Signed-off-by: Adam Jackson <ajax@redhat.com> 29367 29368commit d8ecbe563991cc689e95a8cb9d510e920eaceea0 29369Author: Dave Airlie <airlied@redhat.com> 29370Date: Wed Jan 27 16:46:06 2016 -0800 29371 29372 ephyr: catch X errors if we try to create a core context and fail. 29373 29374 This stops Xephyr failing on GLXBadFBConfig. 29375 29376 Signed-off-by: Dave Airlie <airlied@redhat.com> 29377 Reviewed-by: Dave Airlie <airlied@redhat.com> 29378 29379commit 50ca286d79f6304b972ea74487308e7794a170fb 29380Author: Timo Aaltonen <tjaalton@ubuntu.com> 29381Date: Wed Jan 27 14:18:50 2016 +0200 29382 29383 dri2: Sync i915_pci_ids.h and i965_pci_ids.h from mesa 29384 29385 Adds Skylake, Kabylake and Broxton allowing them to use 29386 modesetting + glamor with dri2. 29387 29388 Signed-off-by: Timo Aaltonen <timo.aaltonen@canonical.com> 29389 Reviewed-by: Andreas Boll <andreas.boll.dev@gmail.com> 29390 29391commit bf23db42a4e5943129501223a47b48884cdeb62f 29392Author: Adam Jackson <ajax@redhat.com> 29393Date: Wed Jan 27 11:50:13 2016 -0500 29394 29395 modesetting: Require sufficiently new libdrm 29396 29397 Bugzilla: https://bugs.freedesktop.org/93883 29398 Signed-off-by: Adam Jackson <ajax@redhat.com> 29399 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 29400 Reviewed-by: Julien Cristau <jcristau@debian.org> 29401 29402commit 6978c8ee666a9224213173e7680e2d71b1094bab 29403Author: Dave Airlie <airlied@redhat.com> 29404Date: Tue Jan 19 08:06:25 2016 +1000 29405 29406 xwayland: add support for use core profile for glamor. (v2) 29407 29408 This adds support to Xwayland to try and use OpenGL core 29409 profile for glamor first. 29410 29411 v1.1: use version defines. 29412 v2: let glamor work out core profile itself. 29413 29414 Signed-off-by: Dave Airlie <airlied@redhat.com> 29415 Reviewed-by: Eric Anholt <eric@anholt.net> 29416 29417commit 79c3925532bc0d098c9a4da6d5117bdada56e0af 29418Author: Dave Airlie <airlied@redhat.com> 29419Date: Tue Jan 19 07:59:59 2016 +1000 29420 29421 glamor: add core profile support to EGL glamor. (v2) 29422 29423 v1.1: use version defines. 29424 v2: let glamor work it out itself 29425 29426 Signed-off-by: Dave Airlie <airlied@redhat.com> 29427 Reviewed-by: Eric Anholt <eric@anholt.net> 29428 29429commit 98c3504dcfcec227b9c7798a0bd287941cec0691 29430Author: Keith Packard <keithp@keithp.com> 29431Date: Wed Sep 10 19:05:08 2014 -0700 29432 29433 ephyr: Create 3.1 core profile context if possible (v3) 29434 29435 On desktop GL, ask for a 3.1 core profile context if that's available, 29436 otherwise create a generic context. 29437 29438 v2: tell glamor the profile is a core one. 29439 v2.1: add/use GL version defines 29440 v3: let glamor work out core itself 29441 29442 Signed-off-by: Keith Packard <keithp@keithp.com> 29443 Signed-off-by: Dave Airlie <airlied@redhat.com> 29444 Reviewed-by: Eric Anholt <eric@anholt.net> 29445 29446commit 564d9f0f8c17bb3c13aa3ca36da7825454dc5de3 29447Author: Dave Airlie <airlied@redhat.com> 29448Date: Mon Jan 11 14:00:35 2016 +1000 29449 29450 glamor: add core profile support. (v2) 29451 29452 Glamor works out from the profile if it is 29453 core. 29454 29455 This flag is used to disable quads for rendering. 29456 29457 v1.1: split long line + make whitespace conform (Michel) 29458 v1.2: add GL 3.1 version defines 29459 v2: move to having glamor work out the profile. 29460 29461 Signed-off-by: Dave Airlie <airlied@redhat.com> 29462 Reviewed-by: Eric Anholt <eric@anholt.net> 29463 29464commit e6754dcb59ee21abb42421a28f4e467295584f67 29465Author: Keith Packard <keithp@keithp.com> 29466Date: Wed Sep 10 19:02:55 2014 -0700 29467 29468 glamor: Use GL_RED instead of GL_ALPHA if we have texture_swizzle (v3) 29469 29470 GL_RED is supported by core profiles while GL_ALPHA is not; use GL_RED 29471 for one channel objects (depth 1 to 8), and then swizzle them into the 29472 alpha channel when used as a mask. 29473 29474 [airlied: updated to master, add swizzle to composited glyphs and xv paths] 29475 29476 v2: consolidate setting swizzle into the texture creation code, it 29477 should work fine there. Handle swizzle when setting color as well. 29478 v3: Fix drawing to a8 with Render (changes by anholt, reviewed by airlied). 29479 29480 Signed-off-by: Keith Packard <keithp@keithp.com> 29481 Signed-off-by: Dave Airlie <airlied@redhat.com> 29482 Reviewed-by: Eric Anholt <eric@anholt.net> 29483 29484commit 5042b0652b9fe5fed57a233880c3429ba390d86d 29485Author: Eric Anholt <eric@anholt.net> 29486Date: Wed Jan 20 12:33:25 2016 -0800 29487 29488 glamor: Drop duplicated GLAMOR_DEFAULT_PRECISIONs in render accel. 29489 29490 We only need it once at the top of the shader, so just put it 29491 there. 29492 29493 Signed-off-by: Eric Anholt <eric@anholt.net> 29494 Reviewed-by: Dave Airlie <airlied@redhat.com> 29495 29496commit 1fd82c764d5b24107e60f2173e30e5d24a2f2667 29497Author: Dave Airlie <airlied@redhat.com> 29498Date: Tue Jan 19 02:01:09 2016 +0000 29499 29500 glamor: don't do copy if we have 0 boxes to copy. 29501 29502 This happens if you run twm + mplayer + xclock and drag 29503 the clock over the mplayer. If we don't catch it, we cause 29504 an illegal draw elements command to be passed to GL. 29505 29506 Signed-off-by: Dave Airlie <airlied@redhat.com> 29507 Reviewed-by: Eric Anholt <eric@anholt.net> 29508 29509commit e7308b6c77561df44c04f81509f8ada678705d94 29510Author: Dave Airlie <airlied@redhat.com> 29511Date: Tue Jan 12 18:13:46 2016 +1000 29512 29513 glamor: Add support for CA rendering in a single pass. 29514 29515 It's been on the list to add dual source blending support to avoid the 29516 two pass componentAlpha code. Radeon has done this for a while in 29517 EXA, so let's add support to bring glamor up to using it. 29518 29519 This adds dual blend to both render and composite glyphs paths. 29520 29521 Initial results show close to doubling of speed of x11perf -rgb10text. 29522 29523 v2: Fix breakage of all of CA acceleration for systems without 29524 GL_ARB_blend_func_extended. Add CA support for all the ops we 29525 support in non-CA mode when blend_func_extended is present. Clean 29526 up some comments and formatting. (changes by anholt) 29527 29528 Signed-off-by: Dave Airlie <airlied@redhat.com> 29529 Signed-off-by: Eric Anholt <eric@anholt.net> 29530 29531commit cab14a9a08ff06bc4cbef79c7be8f1d07c07ebf9 29532Author: Eric Anholt <eric@anholt.net> 29533Date: Thu Jan 21 12:01:02 2016 -0800 29534 29535 glamor: Drop the composite_with_copy path entirely. 29536 29537 I originally inherited this from the EXA code, without determining 29538 whether it was really needed. Regular composite should end up doing 29539 the same thing, since it's all just shaders anyway. To the extent 29540 that it doesn't, we should fix composite. 29541 29542 Signed-off-by: Eric Anholt <eric@anholt.net> 29543 Reviewed-by: Dave Airlie <airlied@redhat.com> 29544 29545commit 510c8605641803f1f5b5d2de6d3bb422b148e0e7 29546Author: Eric Anholt <eric@anholt.net> 29547Date: Thu Jan 21 11:30:15 2016 -0800 29548 29549 glamor: Fix copy-like Render operations between 15 and 16 depth. 29550 29551 Reading and writing to 16-depth pixmaps using PICT_x1r5g5b5 ends up 29552 failing, unless you're doing a straight copy at the same bpp where the 29553 misinterpretation matches on both sides. 29554 29555 Fixes rendercheck/blend/over and renderhceck/blend/src in piglit. 29556 29557 Please cherry-pick this to active stable branches. 29558 29559 Signed-off-by: Eric Anholt <eric@anholt.net> 29560 Reviewed-by: Dave Airlie <airlied@redhat.com> 29561 29562commit bc415fb1e0031ad23bda6e9c3f4664532876a0e5 29563Author: Adam Jackson <ajax@redhat.com> 29564Date: Wed Jan 20 15:43:10 2016 -0500 29565 29566 glx: Fix GLX_EXT_create_context_es2_profile support 29567 29568 As of v4 of this extension, any GLES version number may be requested (to 29569 enable GLES3 and later). To comply with this, simply remove the API 29570 version checks and leave it to the DRI driver to validate. This happens 29571 to also enable using GLES1 in direct contexts, so if that's the dire 29572 situation you find yourself in, your client driver at least stands a 29573 chance of working. 29574 29575 v4 also specifies that both extension strings should be advertised for 29576 compatibility with clients written against v1 of the extension spec, so 29577 add the es_profile bit to the extension list and enable it whenever we 29578 would enable es2_profile. 29579 29580 Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu> 29581 Signed-off-by: Adam Jackson <ajax@redhat.com> 29582 29583commit 49aa5e3ea4eecea0562c05a4e52962985a56e510 29584Author: Keith Packard <keithp@keithp.com> 29585Date: Wed Sep 10 16:17:52 2014 -0700 29586 29587 glamor: Use vertex array objects 29588 29589 Core contexts require the use of vertex array objects, so switch both glamor 29590 and ephyr/glamor over. 29591 29592 Signed-off-by: Keith Packard <keithp@keithp.com> 29593 Reviewed-by: Eric Anholt <eric@anholt.net> 29594 Signed-off-by: Dave Airlie <airlied@redhat.com> 29595 29596commit d99204fb5e09ce7be36485d4226f7ad6d6eb24cc 29597Author: Dave Airlie <airlied@redhat.com> 29598Date: Tue Jan 19 10:34:14 2016 +1000 29599 29600 glamor/xv: add vbo support (v2.1) 29601 29602 This converts the Xv code to using VBOs instead of 29603 client ptrs. This is necessary to move towards using 29604 the core profile later. 29605 29606 v2: put all boxes into single vbo, use draw arrays 29607 to offset things. (Eric) 29608 v2.1: brown paper bag with releasing vbo. 29609 29610 Reviewed-by: Eric Anholt <eric@anholt.net> 29611 Signed-off-by: Dave Airlie <airlied@redhat.com> 29612 29613commit 5582ad1b9b29934498cf3fef305d3a988130cd52 29614Author: Dave Airlie <airlied@redhat.com> 29615Date: Mon Jan 11 14:00:04 2016 +1000 29616 29617 glamor: use vbos in gradient/picture code. 29618 29619 This converts two client arrays users to using vbos, 29620 this is necessary to move to using core profile later. 29621 29622 Reviewed-by: Eric Anholt <eric@anholt.net> 29623 Signed-off-by: Dave Airlie <airlied@redhat.com> 29624 29625commit 25eca80265654cfbf8768024e027426fedeb0918 29626Author: Michel Dänzer <michel.daenzer@amd.com> 29627Date: Fri Jan 15 18:22:21 2016 +0900 29628 29629 present: Handle wraparound when comparing MSC values 29630 29631 When a window moves from one CRTC to another, present_window_to_crtc_msc 29632 updates window_priv->msc_offset according to the delta between the 29633 current MSC values of the old and new CRTC: 29634 29635 window_priv->msc_offset += new_msc - old_msc; 29636 29637 window_priv->msc_offset is initially 0, so if new_msc < old_msc, 29638 window_priv->msc_offset wraps around and becomes a large number. If the 29639 window_msc parameter passed in is small (in particular if it's 0, such as 29640 is the case when the client just wants to know the current window MSC 29641 value), the returned CRTC MSC value may still be a large number. In that 29642 case, the existing MSC comparisons in pixmap_present weren't working as 29643 intended, resulting in scheduling a wait far into the future when the 29644 target MSC had actually already passed. This would result in the client 29645 (e.g. the Chromium browser) hanging when moving its window between CRTCs. 29646 29647 In order to fix this, introduce msc_is_(equal_or_)after helper functions 29648 which take the wraparound into account for comparing two MSC values. 29649 29650 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 29651 Reviewed-by: Keith Packard <keithp@keithp.com> 29652 Reviewed-by: Martin Peres <martin.peres@linux.intel.com> 29653 Signed-off-by: Keith Packard <keithp@keithp.com> 29654 29655commit 1db6de7b6a6ee240eb50a13fe1fa1e135d7cb93b 29656Author: Michel Dänzer <michel.daenzer@amd.com> 29657Date: Tue Jan 12 15:42:47 2016 +0900 29658 29659 glamor: Disable debugging messages other than GL API errors 29660 29661 According to Nicolai Hähnle, the relevant specification says "All 29662 messages are initially enabled unless their assigned severity is 29663 DEBUG_SEVERITY_LOW", so we need to explicitly disable the messages we 29664 don't want to get. Failing that, we were accidentally logging e.g. 29665 shader stats intended for shader-db. 29666 29667 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93659 29668 Tested-by: Laurent Carlier <lordheavym@gmail.com> 29669 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 29670 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 29671 Reviewed-by: Eric Anholt <eric@anholt.net> 29672 29673commit 8116fd8a760b5935645def1b2c3b155c05927850 29674Author: Dave Airlie <airlied@redhat.com> 29675Date: Tue Jan 12 14:19:24 2016 +1000 29676 29677 glamor: store old fonts in double width textures. 29678 29679 There is a problem with some fonts that the height necessary 29680 to store the font is greater than the max texture size, which 29681 causes a fallback to occur. We can avoid this by storing two 29682 macro columns side-by-side in the texture and adjusting 29683 the calculations to suit. 29684 29685 This fixes 29686 xfd -fn -*-*-*-*-*-*-*-*-*-*-*-*-*-* 29687 falling back here, when it picks 29688 -arabic-newspaper-medium-r-normal--32-246-100-100-p-137-iso10646-1 29689 29690 Reviewed-by: Keith Packard <keithp@keithp.com> 29691 Signed-off-by: Dave Airlie <airlied@redhat.com> 29692 29693commit 64081d0eacf3e53a029b8e8b63096cc153e98549 29694Author: Dave Airlie <airlied@redhat.com> 29695Date: Mon Jan 11 17:02:57 2016 +1000 29696 29697 glamor: fallback if font is too large for FBO size. 29698 29699 running xfontsel on haswell here, with a max texture size 29700 of 8kx8k, one font wants 9711 height. This fallsback to 29701 sw in this case. 29702 29703 A proper solution probably involves using an array texture. 29704 29705 Reviewed-by: Keith Packard <keithp@keithp.com> 29706 Signed-off-by: Dave Airlie <airlied@redhat.com> 29707 29708commit 6dcb73375e0ce389315d55587623eb84e9d13543 29709Author: Adam Jackson <ajax@redhat.com> 29710Date: Wed Jan 6 10:03:23 2016 -0500 29711 29712 os: Failure to remove a non-existent log file is not an error 29713 29714 Signed-off-by: Adam Jackson <ajax@redhat.com> 29715 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 29716 29717commit 862cbf4c870c9ed913206c6ef4988bdb470e1c39 29718Author: Thomas Klausner <wiz@NetBSD.org> 29719Date: Wed Nov 11 13:32:05 2015 +0100 29720 29721 Fix build when XSERVER_PLATFORM_BUS is not defined. 29722 29723 Reviewed-by: Adam Jackson <ajax@redhat.com> 29724 Signed-off-by: Thomas Klausner <wiz@NetBSD.org> 29725 29726commit 63f83d1b7f496d05b409352749cdb6674d71cf80 29727Author: Thomas Klausner <wiz@NetBSD.org> 29728Date: Tue Jan 5 12:51:41 2016 -0500 29729 29730 Fix uninitialized variable warnings reported by clang 29731 29732 v2: Move initializing pos into the first clause of the for statement. We 29733 have to keep this macro equivalent to a plain for statement from the 29734 user's perspective, otherwise callers need to {} things to keep control 29735 flow correct. [ajax] 29736 29737 Signed-off-by: Thomas Klausner <wiz@NetBSD.org> 29738 Acked-by: Michel Dänzer <michel.daenzer@amd.com> 29739 29740commit ba71b69f94f00a6f6910597185610668e79c10be 29741Author: Alan Coopersmith <alan.coopersmith@oracle.com> 29742Date: Fri Jan 1 17:34:41 2016 -0800 29743 29744 Avoid segfault in CloseWellKnownConnections when using -displayfd 29745 29746 When -displayfd is looping through the possible display ids to use, 29747 if it can't open all the listening sockets for one (say when :0 is 29748 already in use), it calls CloseWellKnownConnections to close all 29749 the ListenTransConns entries before the point that ListenTransFds 29750 was allocated & initialized, so CloseWellKnownConnections would 29751 segfault trying to read entries from a NULL ListenTransFds pointer. 29752 29753 Introduced by commit 7b02f0b8 29754 29755 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 29756 Reviewed-by: Keith Packard <keithp@keithp.com> 29757 29758commit edcb6426f20c3be5dd5f50b76a686754aef2f64e 29759Author: Alan Coopersmith <alan.coopersmith@oracle.com> 29760Date: Fri Jan 1 18:11:14 2016 -0800 29761 29762 Use unique logfile names when starting server with -displayfd 29763 29764 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=93212 29765 29766 Previously all X servers started with -displayfd would overwrite 29767 Xorg.0.log - now a temporary name of Xorg.pid-<pid>.log is used 29768 until after -displayfd finds an open display - then it is renamed 29769 to the traditional Xorg.<display>.log name. 29770 29771 Reviewed-by: Adam Jackson <ajax@redhat.com> 29772 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 29773 29774commit fe8562f5316d8c74ca074ad145295c65ddff5fc2 29775Author: Alan Coopersmith <alan.coopersmith@oracle.com> 29776Date: Fri Jan 1 18:10:08 2016 -0800 29777 29778 modesetting should not reference gbm when it's not defined 29779 29780 Fixes build errors of: 29781 present.c: In function 'ms_do_pageflip': 29782 present.c:410:17: error: 'drmmode_bo' has no member named 'gbm' 29783 new_front_bo.gbm = glamor_gbm_bo_from_pixmap(screen, new_front); 29784 ^ 29785 present.c:412:22: error: 'drmmode_bo' has no member named 'gbm' 29786 if (!new_front_bo.gbm) { 29787 ^ 29788 present.c: In function 'ms_present_check_flip': 29789 present.c:536:36: error: 'drmmode_bo' has no member named 'gbm' 29790 if (drmmode_crtc->rotate_bo.gbm) 29791 ^ 29792 Introduced by commit 13c7d53d 29793 29794 Reviewed-by: Adam Jackson <ajax@redhat.com> 29795 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 29796 29797commit 792176417f7e62a99c4f07801f74f804d6080de2 29798Author: Adam Jackson <ajax@redhat.com> 29799Date: Tue Jan 5 12:03:54 2016 -0500 29800 29801 Revert "Fix uninitialized variable warnings reported by clang" 29802 29803 Crashes at startup: 29804 29805 dmt:~/git/xserver% ./hw/kdrive/ephyr/Xephyr :1 29806 (EE) Backtrace: 29807 (EE) 0: ./hw/kdrive/ephyr/Xephyr (OsSigHandler+0x29) [0x47c8f9] 29808 (EE) 1: /lib64/libc.so.6 (__restore_rt+0x0) [0x7f38d4de6b1f] 29809 (EE) 2: ./hw/kdrive/ephyr/Xephyr (InitNotifyFds+0x17) [0x478697] 29810 (EE) 3: ./hw/kdrive/ephyr/Xephyr (OsInit+0x1e) [0x47c99e] 29811 (EE) 4: ./hw/kdrive/ephyr/Xephyr (dix_main+0x11c) [0x446efc] 29812 (EE) 5: /lib64/libc.so.6 (__libc_start_main+0xf0) [0x7f38d4dd2580] 29813 (EE) 6: ./hw/kdrive/ephyr/Xephyr (_start+0x29) [0x427099] 29814 (EE) 7: ? (?+0x29) [0x29] 29815 (EE) 29816 (EE) Segmentation fault at address 0x0 29817 29818 This reverts commit a221d4737c167589da44595c795d54f2c36b439a. 29819 29820commit a221d4737c167589da44595c795d54f2c36b439a 29821Author: Thomas Klausner <wiz@NetBSD.org> 29822Date: Wed Dec 16 08:36:24 2015 +0100 29823 29824 Fix uninitialized variable warnings reported by clang 29825 29826 Signed-off-by: Thomas Klausner <wiz@NetBSD.org> 29827 Acked-by: Michel Dänzer <michel.daenzer@amd.com> 29828 29829commit f3593918a0507cc080875788b2776ebe4f57c694 29830Author: Peter Hutterer <peter.hutterer@who-t.net> 29831Date: Wed Dec 16 09:21:39 2015 +1000 29832 29833 xfree86: move check for driver->PreInit up 29834 29835 No real change, but if the driver is broken and doesn't provide a PreInit 29836 function, then we don't need to worry about logind. 29837 29838 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 29839 Reviewed-by: Keith Packard <keithp@keithp.com> 29840 29841commit fc4fbe8224fdc8d7774120b6e978c652a70a50ee 29842Author: Peter Hutterer <peter.hutterer@who-t.net> 29843Date: Fri Dec 11 09:31:30 2015 +1000 29844 29845 xfree86: add NoMatchFoo directives for InputClass sections 29846 29847 InputClass sections use various MatchFoo directives to decide which device to 29848 apply to. This usually works fine for specific snippets but has drawbacks for 29849 snippets that apply more generally to a multitude of devices. 29850 29851 This patch adds a NoMatchFoo directive to negate a match, thus allowing 29852 snippets that only apply if a given condition is not set. Specifically, this 29853 allows for more flexible fallback driver matching, it is now possible to use a 29854 snippet that says "assign driver foo, but only if driver bar wasn't already 29855 assigned to it". For example: 29856 29857 Section "InputClass" 29858 Identifier "libinput for tablets" 29859 MatchIsTablet "true" 29860 NoMatchDriver "wacom" 29861 Driver "libinput" 29862 EndSection 29863 29864 The above only assigns libinput to tablet devices if wacom isn't already 29865 assigned to this device, making it possible to select a specific driver by 29866 installing/uninstalling it. 29867 29868 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 29869 Reviewed-by: Benjamin Tissoires <benjamin.tissoires@gmail.com> 29870 29871commit eb671b804e902011ba901d0833a215799177bad0 29872Author: Peter Hutterer <peter.hutterer@who-t.net> 29873Date: Fri Dec 11 07:26:02 2015 +1000 29874 29875 xfree86: whitespace fix 29876 29877 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 29878 29879commit 3d68d1f26709ecb5ce22a9baa0d3d8162574ed6a 29880Author: Siim Põder <siim@p6drad-teel.net> 29881Date: Mon Jun 8 22:14:12 2015 +0000 29882 29883 vfb: add randr support (v2) 29884 29885 The motivation for getting this is chrome remote desktop that runs under 29886 Xvfb and wants to use RANDR to adjust screen size according to the 29887 remote desktop client screen size. Apparently there are other use cases 29888 as well, the bug mentions gnome-settings-daemon testing. 29889 29890 [ajax: massaged commit message] 29891 29892 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=26391 29893 Reviewed-by: Adam Jackson <ajax@redhat.com> 29894 Signed-off-by: Lambros Lambrou <lambroslambrou@google.com> 29895 Signed-off-by: Mike Frysinger <vapier@gentoo.org> 29896 Signed-off-by: Michal Srb <msrb@suse.com> 29897 Signed-off-by: Siim Põder <siim@p6drad-teel.net> 29898 29899commit e98b36bc69d805ab52aba95b070a614427d46e05 29900Author: Adam Jackson <ajax@redhat.com> 29901Date: Tue Dec 8 16:22:16 2015 -0500 29902 29903 randr: Silence unused variable warning 29904 29905 Signed-off-by: Adam Jackson <ajax@redhat.com> 29906 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 29907 29908commit 8c49c883dc196ea57e3f64425c2a5b0a8ce3ba75 29909Author: Adam Jackson <ajax@redhat.com> 29910Date: Tue Dec 8 15:08:34 2015 -0500 29911 29912 sync: Don't allow creating a system counter before sync is initialized 29913 29914 It probably doesn't work very well since there's other extension setup 29915 we're not doing on this path, and in any event it's not a thing that 29916 happens currently. 29917 29918 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 29919 Signed-off-by: Adam Jackson <ajax@redhat.com> 29920 29921commit b5f04a79df8b5aab8b558461261d14721f0b3e41 29922Author: Adam Jackson <ajax@redhat.com> 29923Date: Tue Dec 8 15:56:17 2015 -0500 29924 29925 glxproxy: Silence shadowed-variable warnings 29926 29927 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 29928 Signed-off-by: Adam Jackson <ajax@redhat.com> 29929 29930commit 18729a211a5fdd4f733e44eded754a0e6210b687 29931Author: Adam Jackson <ajax@redhat.com> 29932Date: Tue Dec 8 15:53:35 2015 -0500 29933 29934 glxproxy: Silence set-but-unused-variable warnings 29935 29936 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 29937 Signed-off-by: Adam Jackson <ajax@redhat.com> 29938 29939commit bc996fa4e3c06f65eafe0a88ef491dbf01f31422 29940Author: Adam Jackson <ajax@redhat.com> 29941Date: Tue Dec 8 16:04:44 2015 -0500 29942 29943 dmx: Run 'doxygen -u' to upgrade the doxygen config file 29944 29945 Also change the dot font setting back to the default of Helvetica as 29946 doxygen no longer ships FreeSans. 29947 29948 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 29949 Signed-off-by: Adam Jackson <ajax@redhat.com> 29950 29951commit a55e0bc56fd8f9da8f066dc344af19535dd092ca 29952Author: Adam Jackson <ajax@redhat.com> 29953Date: Tue Dec 8 16:00:19 2015 -0500 29954 29955 dmx: Silence unused variable warning in dmxcompat 29956 29957 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 29958 Signed-off-by: Adam Jackson <ajax@redhat.com> 29959 29960commit 2730ccb803c55af74dbdd1bfd982fc23e643554d 29961Author: Adam Jackson <ajax@redhat.com> 29962Date: Tue Dec 8 15:37:12 2015 -0500 29963 29964 dmx: Silence lex/yacc-related config parser warnings 29965 29966 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 29967 Signed-off-by: Adam Jackson <ajax@redhat.com> 29968 29969commit 718223d27452862eedcf1bee6278eae6040d45ea 29970Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29971Date: Thu Dec 3 08:05:59 2015 -0200 29972 29973 systemd-logind.c: don't parse VT settings for non-seat0 X servers 29974 29975 Since non-seat0 X servers no longer touch VTs, I believe these settings 29976 are unnecessary. 29977 29978 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 29979 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 29980 29981commit 7c0ba32ddd5f1d5c65279365fa307dec3433caf3 29982Author: Bob Ham <bob.ham@collabora.com> 29983Date: Fri Dec 4 12:30:47 2015 +0000 29984 29985 xserver: Fix configure.ac check for libsystemd/-daemon 29986 29987 The configure script looks for the libsystemd-daemon pkg-config 29988 module. If the configure script finds it, the script will add 29989 libsystemd-daemon to a list of modules which are used to consolidate 29990 CFLAGS and LIBS. 29991 29992 The check for libsystemd-daemon was altered to fallback to libsystemd 29993 if libsystemd-daemon was not found (libsystemd-daemon was brought into 29994 libsystemd). Unfortunately, the configure script still adds 29995 "libsystemd-daemon" to the list of modules to consolidate, instead of 29996 "libsystemd". With this patch, we set a variable depending on which 29997 pkg-config module is found and add that to the module list instead. 29998 29999 Changes since v1: 30000 - Rearranged logic so that we do a versioned check for libsystemd 30001 first, then look for libsystemd-daemon. 30002 - Cleaned up the check a bit, only performing the module checks if we 30003 don't have --with-systemd-daemon=no, in a similar style to 30004 --with-dtrace. 30005 - Changed the variable name to LIBSYSTEMD_DAEMON as per feedback. 30006 30007 Reviewed-by: Adam Jackson <ajax@redhat.com> 30008 Signed-off-by: Bob Ham <bob.ham@collabora.com> 30009 30010commit 530d3e5ca0a02039b04ec6a677bbb4e05b78e5f4 30011Author: Michel Dänzer <michel.daenzer@amd.com> 30012Date: Thu Dec 3 17:04:09 2015 +0900 30013 30014 prime: Damage full destination rectangle when we start dirty tracking 30015 30016 This makes sure that the destination pixmap contents will be fully 30017 initialized. Without this, a PRIME output starts out with garbage. 30018 30019 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 30020 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 30021 30022commit 2a52c06e235bd79f91851121f53f7c1808fde321 30023Author: Adam Jackson <ajax@redhat.com> 30024Date: Mon Dec 7 17:03:02 2015 -0500 30025 30026 x86emu: Squash a warning 30027 30028 Apologies, should have caught this one when applying the previous x86emu 30029 patch. 30030 30031 Signed-off-by: Adam Jackson <ajax@redhat.com> 30032 30033commit 59b618227ebd024e57720aaaea17596953f5b80e 30034Author: Julian Pidancet <julian.pidancet@gmail.com> 30035Date: Sun Jul 1 18:49:25 2012 +0100 30036 30037 x86emu: Correctly handle 0x66 prefix for some instructions 30038 30039 (Sorry for double posting) 30040 30041 I repost this patch because I havn't got any replies from maintainers 30042 since I posted the initial patch back in March. 30043 30044 Some instructions are not emulated correctly by x86emu when they 30045 are prefixed by the 0x66 opcode. 30046 I've identified problems in the emulation of these intructions: ret, 30047 enter, leave, iret and some forms of call. 30048 30049 Most of the time, the problem is that these instructions should push or 30050 pop 32-bit values to/from the stack, instead of 16bit, when they are 30051 prefixed by the 0x66 special opcode. 30052 30053 The SeaBIOS project aims to produce a complete legacy BIOS 30054 implementation as well as a VGA option ROM, entirely written in C and 30055 using the GCC compiler. 30056 30057 In 16bit code produced by the GCC compiler, the 0x66 prefix is used 30058 almost everywhere. This patch is necessary to allow the SeaBIOS VGA 30059 option ROM to function with Xorg when using the vesa driver. 30060 30061 SeaBIOS currently use postprocessing on the ROM assembly output to 30062 replace the affected instruction with alternative unaffected instructions. 30063 This is obviously not very elegant, and this fix in x86emu would be 30064 more appropriate. 30065 30066 v2: - Decrement BP instead of EBP in accordance with the Intel Manual 30067 - Assign EIP instead of IP when poping the return address from the 30068 stack in 32-bit operand size mode in ret_far_IMM, ret_far, and iret 30069 - When poping EFLAGS from the stack in iret in 32-bit operand size 30070 mode, apply some mask to preserve Read-only flags. 30071 30072 v3: - Rebase 30073 30074 Reviewed-by: Adam Jackson <ajax@redhat.com> 30075 Signed-off-by: Julian Pidancet <julian.pidancet@gmail.com> 30076 30077commit 72f0724cdc65dc9abbbf70b9feb6cce7c2b9f8a0 30078Author: Chris Wilson <chris@chris-wilson.co.uk> 30079Date: Sun Feb 8 09:47:42 2015 +0000 30080 30081 present: Do not replace Pixmaps on redirected Window on unflip 30082 30083 When unflipping, we may find that our flip window has been redirected. 30084 If we replace the redirected Window with the Screen Pixmap we then have 30085 mutliple fullscreen Windows believing that their own the Screen Pixmap - 30086 multiple fullscreen Windows that are being flipped by Clients, and so 30087 continue to flip causing popping between e.g. the compositor and the 30088 game. 30089 30090 [ajax: Fix up present_execute() hunk to account for changes introduced 30091 in fe07ec19e212a68076560d243a2a935c54589068] 30092 30093 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 30094 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30095 30096commit 180b09912c0d2c4a43b5a08678bcad4b818924c7 30097Author: Chris Wilson <chris@chris-wilson.co.uk> 30098Date: Sun Feb 8 09:47:41 2015 +0000 30099 30100 present: When cancelling a pending synchronous flip, requeue it 30101 30102 The vblank event request for a synchronous flip is scheduled for the 30103 vblank before the target flip msc (so that the flip itself appears at 30104 the right frame). If we cancel that flip and so wish to schedule a 30105 copy instead, that copy needs to be postponed by a frame in order for it 30106 be performed at the requested time. 30107 30108 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 30109 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30110 30111commit b2d55338f6b8f43ebcb49994abad123a797248cf 30112Author: Chris Wilson <chris@chris-wilson.co.uk> 30113Date: Sun Feb 8 09:47:40 2015 +0000 30114 30115 present: Requery pending flips with the right sync_flip mode 30116 30117 When verifying whether a pending flip is still valid, we need to pass 30118 down the orignal sync_flip mode (e.g. if the driver only supports sync 30119 flips, verifying a async flip will falsely fail). 30120 30121 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 30122 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30123 30124commit 548a3d5fd69bb059bbaf26ededdc94c212712cd7 30125Author: Dave Airlie <airlied@gmail.com> 30126Date: Mon Nov 16 09:05:27 2015 +1000 30127 30128 modesetting: create entities for pci and old probe. (v2) 30129 30130 This moves the code from the platform case into 30131 a common function, and calls that from the 30132 other two. 30133 30134 v2: Emil convinced me we don't need to lookup pEnt 30135 here, so let's not bother. 30136 30137 Reported-by: Mark Kettenis <mark.kettenis@xs4all.nl> 30138 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 30139 Signed-off-by: Dave Airlie <airlied@redhat.com> 30140 30141commit 771016f0705909c908917b4ccaeafc950ba93c05 30142Author: Dave Airlie <airlied@gmail.com> 30143Date: Mon Nov 16 09:05:26 2015 +1000 30144 30145 modesetting: drop platform_dev pointer. 30146 30147 This isn't used anywhere, so no point storing it until we need it. 30148 30149 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 30150 Signed-off-by: Dave Airlie <airlied@redhat.com> 30151 30152commit 19b0249a5e07b9fc008e5d8709d7e489874415de 30153Author: Arkadiusz Miśkiewicz <arekm@maven.pl> 30154Date: Mon Nov 16 11:06:57 2015 +0100 30155 30156 Xorg.wrap: activate libdrm based detection for KMS drivers 30157 30158 Xorg.wrap includes code guarded with WITH_LIBDRM for detecting KMS drivers. 30159 Unfortunately it is never activated since code missed to include file 30160 which defines WITH_LIBDRM. 30161 30162 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92894 30163 Signed-off-by: Arkadiusz Miśkiewicz <arekm@maven.pl> 30164 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30165 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 30166 30167commit 5bccde749db93296b7784e4cdc5e54c4443656c1 30168Author: Michel Dänzer <michel.daenzer@amd.com> 30169Date: Wed Dec 2 18:21:12 2015 +0900 30170 30171 randr: Stop dirty tracking for shared pixmap being destroyed 30172 30173 Otherwise, we leave a dangling reference to the destroyed pixmap in the 30174 master screen's pixmap_dirty_list. 30175 30176 Fixes regression from commit cf5d6414 ("randr: Factor out shared pixmap 30177 destruction"). 30178 30179 Reviewed-by: Adam Jackson <ajax@redhat.com> 30180 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 30181 30182commit 2e3d9623ae3f562f81e513cb183ca1b1b68f279c 30183Author: Adam Jackson <ajax@redhat.com> 30184Date: Wed Dec 2 10:42:36 2015 -0500 30185 30186 Revert "hw/xfree86: Use NotifyFd for device and other input fd wakeups" 30187 30188 Reported to break libinput: 30189 30190 http://lists.freedesktop.org/archives/xorg-devel/2015-December/048091.html 30191 30192 This reverts commit 1df07dc36ca145c59f51176d9ab2651112506d75. 30193 30194commit 7b02f0b87ec2fa0cc5a65307a1fd55c671cec884 30195Author: Keith Packard <keithp@keithp.com> 30196Date: Wed Nov 11 22:02:17 2015 -0800 30197 30198 os: Use NotifyFd interface for listen descriptors 30199 30200 Replace the custom path for dealing with new incoming connections with 30201 the general-purpose NotifyFd API. 30202 30203 Reviewed-by: Adam Jackson <ajax@redhat.com> 30204 Signed-off-by: Keith Packard <keithp@keithp.com> 30205 30206commit 1df07dc36ca145c59f51176d9ab2651112506d75 30207Author: Keith Packard <keithp@keithp.com> 30208Date: Wed Nov 11 22:02:16 2015 -0800 30209 30210 hw/xfree86: Use NotifyFd for device and other input fd wakeups 30211 30212 Remove code in xf86Wakeup for dealing with device and other input and 30213 switch to using the new NotifyFd interface. 30214 30215 Reviewed-by: Adam Jackson <ajax@redhat.com> 30216 Signed-off-by: Keith Packard <keithp@keithp.com> 30217 30218commit e51ea53b26bd9ec05b9209825960af28d0b6bbe1 30219Author: Keith Packard <keithp@keithp.com> 30220Date: Wed Nov 11 22:02:15 2015 -0800 30221 30222 render: Use OsTimer for animated cursor timing 30223 30224 This replaces the block/wakeup handlers with an OsTimer. This also 30225 avoids problems with performing rendering during the wakeup handler. 30226 30227 Reviewed-by: Adam Jackson <ajax@redhat.com> 30228 Signed-off-by: Keith Packard <keithp@keithp.com> 30229 30230commit 49c0f2413d32fdfe36e45861fcb32aaeab633094 30231Author: Keith Packard <keithp@keithp.com> 30232Date: Wed Nov 11 22:02:14 2015 -0800 30233 30234 os/xdmcp: Replace xdmcp block/wakeup handlers with timer and NotifyFd 30235 30236 This removes the block and wakeup handlers and replaces them with a 30237 combination of a NotifyFd callback and timers. 30238 30239 Reviewed-by: Adam Jackson <ajax@redhat.com> 30240 Signed-off-by: Keith Packard <keithp@keithp.com> 30241 30242commit 6a121f55381ecbb9cdaaef36ba6135890f3e006f 30243Author: Keith Packard <keithp@keithp.com> 30244Date: Wed Nov 11 22:02:13 2015 -0800 30245 30246 Xext/xselinux: Use NotifyFd interface 30247 30248 Replace block/wakeup handlers with SetNotifyFd. Much nicer now. 30249 30250 Reviewed-by: Adam Jackson <ajax@redhat.com> 30251 Signed-off-by: Keith Packard <keithp@keithp.com> 30252 30253commit f933a1b38ed1c65cc39fce1ee37aa18e3022c3f0 30254Author: Keith Packard <keithp@keithp.com> 30255Date: Mon Nov 23 17:36:53 2015 -0800 30256 30257 hw/xwayland: Use NotifyFd handler to monitor wayland socket 30258 30259 Replace the block/wakeup handler with a NotifyFd callback instead. 30260 30261 Reviewed-by: Adam Jackson <ajax@redhat.com> 30262 Signed-off-by: Keith Packard <keithp@keithp.com> 30263 30264commit 8543d4d8bc0526d1c910913b76259e5dee108e74 30265Author: Keith Packard <keithp@keithp.com> 30266Date: Wed Nov 11 22:02:11 2015 -0800 30267 30268 modesetting: Use NotifyFd for drm event monitoring 30269 30270 Replace the block/wakeup handlers with a NotifyFd callback. 30271 30272 Reviewed-by: Adam Jackson <ajax@redhat.com> 30273 Signed-off-by: Keith Packard <keithp@keithp.com> 30274 30275commit 58354fcf472cefc35b9c19cf84bf079cadfa2e1d 30276Author: Keith Packard <keithp@keithp.com> 30277Date: Wed Nov 11 22:02:10 2015 -0800 30278 30279 kdrive/ephyr: Use NotifyFd for XCB connection input [v2] 30280 30281 Eliminates polling every 20ms for device input. 30282 30283 v2: rename ephyrPoll to ephyrXcbNotify and fix the API so it can be 30284 used directly for SetNotifyFd. Thanks to Daniel Martin 30285 <consume.noise@gmail.com> 30286 30287 Reviewed-by: Adam Jackson <ajax@redhat.com> 30288 Signed-off-by: Keith Packard <keithp@keithp.com> 30289 Cc: Daniel Martin <consume.noise@gmail.com> 30290 30291commit 483c2a1adf044ba1da844687b908c1c802060d2d 30292Author: Keith Packard <keithp@keithp.com> 30293Date: Wed Nov 11 22:02:09 2015 -0800 30294 30295 hw/kdrive: Use NotifyFd for kdrive input devices 30296 30297 This switches the kdrive code to use FD notification for input 30298 devices, rather than the block and wakeup handlers. 30299 30300 Reviewed-by: Adam Jackson <ajax@redhat.com> 30301 Signed-off-by: Keith Packard <keithp@keithp.com> 30302 30303commit 21c1680e83865a52d88cf8c80fb236d212931e5c 30304Author: Keith Packard <keithp@keithp.com> 30305Date: Wed Nov 11 22:02:08 2015 -0800 30306 30307 hw/kdrive: Use NotifyFd interface for kdrive/linux APM monitoring 30308 30309 Replace the block/wakeup handlers with a NotifyFd callback 30310 30311 Reviewed-by: Adam Jackson <ajax@redhat.com> 30312 Signed-off-by: Keith Packard <keithp@keithp.com> 30313 30314commit 8c8f3567fe4b4e372e22420810443c43e008da2c 30315Author: Keith Packard <keithp@keithp.com> 30316Date: Wed Nov 11 22:02:07 2015 -0800 30317 30318 config: Use NotifyFd interface for udev 30319 30320 This uses the NotifyFd interface to monitor the udev file descriptor 30321 rather than adding another block/wakeup handler 30322 30323 Reviewed-by: Adam Jackson <ajax@redhat.com> 30324 Signed-off-by: Keith Packard <keithp@keithp.com> 30325 30326commit bf920b2390dc27947f87dd0b228518290a5ed85d 30327Author: Keith Packard <keithp@keithp.com> 30328Date: Wed Nov 11 22:02:06 2015 -0800 30329 30330 config: Use NotifyFd for dbus interface 30331 30332 This uses the NotifyFd interface to monitor the dbus socket rather 30333 than a block/wakeup handler. 30334 30335 Reviewed-by: Adam Jackson <ajax@redhat.com> 30336 Signed-off-by: Keith Packard <keithp@keithp.com> 30337 30338commit 91ea0965dd4dfeba0a914c47ad4a64768e983b1b 30339Author: Keith Packard <keithp@keithp.com> 30340Date: Wed Nov 11 22:02:04 2015 -0800 30341 30342 dix: Move InitFonts up above screen initialization 30343 30344 Font initialization was split into two stages, the first was to set up 30345 font privates with a call to ResetFontPrivateIndex, then much later 30346 the call to InitFonts to set up all of the FPEs. Doing the full font 30347 initialization before initializing the video drivers means that we can 30348 move the call to ResetFontPrivateIndex inside InitFonts. 30349 30350 Reviewed-by: Adam Jackson <ajax@redhat.com> 30351 Signed-off-by: Keith Packard <keithp@keithp.com> 30352 30353commit 4020aacd1fc5b9c63369f011aeb9120af9c55218 30354Author: Keith Packard <keithp@keithp.com> 30355Date: Wed Nov 11 22:02:03 2015 -0800 30356 30357 os: Implement support for NotifyFd X_NOTIFY_WRITE 30358 30359 This adds the ability to be notified when a file descriptor is 30360 available for writing. 30361 30362 Reviewed-by: Adam Jackson <ajax@redhat.com> 30363 Signed-off-by: Keith Packard <keithp@keithp.com> 30364 30365commit 0c41b7af4ab0c8d22b88f201293f59524d1e7317 30366Author: Keith Packard <keithp@keithp.com> 30367Date: Wed Nov 11 22:02:02 2015 -0800 30368 30369 os: Add NotifyFd interfaces 30370 30371 This provides a callback-based interface to monitor file 30372 descriptors beyond the usual client and device interfaces. 30373 30374 Modules within the server using file descriptors for reading and/or 30375 writing can call 30376 30377 Bool SetNotifyFd(int fd, NotifyFdProcPtr notify_fd, int mask, void *data); 30378 30379 mask can be any combination of X_NOTIFY_READ and X_NOTIFY_WRITE. 30380 30381 When 'fd' becomes readable or writable, the notify_fd function will be 30382 called with the 'fd', the ready conditions and 'data' values as arguments, 30383 30384 When the module no longer needs to monitor the fd, it will call 30385 30386 void RemoveNotifyFd(int fd); 30387 30388 RemoveNotifyFd may be called from the notify function. 30389 30390 Reviewed-by: Adam Jackson <ajax@redhat.com> 30391 Signed-off-by: Keith Packard <keithp@keithp.com> 30392 30393commit e10ba9e4b52269b2ac75c4802dce4ca47d169657 30394Author: Keith Packard <keithp@keithp.com> 30395Date: Wed Nov 11 22:02:01 2015 -0800 30396 30397 Remove non-smart scheduler. Don't require setitimer. 30398 30399 This allows the server to call GetTimeInMillis() after each request is 30400 processed to avoid needing setitimer. -dumbSched now turns off the 30401 setitimer. 30402 30403 Reviewed-by: Adam Jackson <ajax@redhat.com> 30404 Signed-off-by: Keith Packard <keithp@keithp.com> 30405 30406commit c7f4aef8f45e500c900d59f68c653477148907ea 30407Author: agoins <agoins@nvidia.com> 30408Date: Wed Nov 25 18:39:31 2015 -0800 30409 30410 randr: Cleanup rrSetupPixmapSharing() 30411 30412 protopix is completely redundant with mscreenpix. Get rid of it. 30413 30414 We don't need rrScrPriv, so remove it. 30415 30416 [ajax: also squash an unused variable in RRCrtcDetachScanoutPixmap, 30417 though it'll come back when the rest of this series lands] 30418 30419 Reviewed-by: Adam Jackson <ajax@redhat.com> 30420 Signed-off-by: Alex Goins <agoins@nvidia.com> 30421 30422commit 8d3f0e964e399dcfa8eb5e85d405217fdc5dbcd4 30423Author: agoins <agoins@nvidia.com> 30424Date: Wed Nov 25 18:39:27 2015 -0800 30425 30426 xf86: Bump ABI version to 21 30427 30428 Reviewed-by: Adam Jackson <ajax@redhat.com> 30429 Signed-off-by: Alex Goins <agoins@nvidia.com> 30430 30431commit 7006b4e7ff759c899d5391b7d12db889cbc0b535 30432Author: agoins <agoins@nvidia.com> 30433Date: Wed Nov 25 18:39:26 2015 -0800 30434 30435 randr: Factor out shared pixmap creation 30436 30437 The old version of rrCreateSharedPixmap(), in addition to actually creating 30438 a shared pixmap with scanout, also set up pixmap tracking on the source 30439 driver. 30440 30441 I will be needing to create multiple shared pixmaps for PRIME double 30442 buffering, so factor the part that does shared pixmap creation into its own 30443 function, the new rrCreateSharedPixmap(). Rename the old 30444 rrCreateSharedPixmap() to rrSetupPixmapSharing(), a function that 30445 replicates the old functionality of rrCreateSharedPixmap() using the new 30446 rrCreateSharedPixmap(). 30447 30448 Reviewed-by: Adam Jackson <ajax@redhat.com> 30449 Signed-off-by: Alex Goins <agoins@nvidia.com> 30450 30451commit cf5d6414e0c21140f763d618bde1e91ad2b1cb49 30452Author: agoins <agoins@nvidia.com> 30453Date: Wed Nov 25 18:39:25 2015 -0800 30454 30455 randr: Factor out shared pixmap destruction 30456 30457 Shared pixmap destruction is done by unrefing the master pixmap twice: once 30458 for the original reference, and once for the reference implicitly added by 30459 PixmapShareToSlave. Then, unrefing the slave pixmap once. 30460 30461 When I add PRIME double buffering and synchronization, I will need to do 30462 this in multiple places. To avoid duplication of code and comments 30463 explaining it everywhere, factor it out into its own function and use that 30464 in place of where it was before. 30465 30466 Reviewed-by: Adam Jackson <ajax@redhat.com> 30467 Signed-off-by: Alex Goins <agoins@nvidia.com> 30468 30469commit ab9837cc6a11f46b9df780f131b69de3822c3dd9 30470Author: Olivier Fourdan <ofourdan@redhat.com> 30471Date: Tue Dec 1 17:16:03 2015 +0100 30472 30473 xwayland: Update screen size on output removal 30474 30475 When unplugging an output, it's still listed in xrandr and the size 30476 of the root window still includes the removed output. 30477 30478 The RR output should be destroyed when its Wayland counterpart is 30479 destroyed and the screen dimensions must be updated in both the done 30480 and the destroy handlers. 30481 30482 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92914 30483 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 30484 Reviewed-by: Marek Chalupa <mchqwerty@gmail.com> 30485 30486commit 07941a50a547f2ca094e242588298695f48903ed 30487Author: Jonas Ådahl <jadahl@gmail.com> 30488Date: Wed Oct 7 12:01:53 2015 +0800 30489 30490 xwayland: Always update the wl_pointer cursor on pointer focus 30491 30492 In Wayland, a client (in this case XWayland) should set the cursor 30493 surface when it receives pointer focus. Not doing this will leave the 30494 curser at whatever it was previously. 30495 30496 When running on XWayland, the X server will not be the entity that 30497 controls what actual pointer cursor is displayed, and it wont be notified 30498 about the pointer cursor changes done by the Wayland compositor. This 30499 causes X11 clients running via XWayland to end up with incorrect pointer 30500 cursors because the X server believes that, if the cursor was previously 30501 set to the cursor C, if we receive Wayland pointer focus over window W 30502 which also has the pointer cursor C, we do not need to update it. This 30503 will cause us to end up with the wrong cursor if cursor C was not the 30504 same one that was already set by the Wayland compositor. 30505 30506 This patch works around this by, when receiving pointer focus, getting 30507 the private mipointer struct changing the "current sprite" pointer to 30508 an invalid cursor in order to trigger the update path next time a cursor 30509 is displayed by dix. 30510 30511 Reviewed-by: Adam Jackson <ajax@redhat.com> 30512 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 30513 30514commit 51a4399b94f9adfac5f7718d4cbf73f793dcca56 30515Author: Olivier Fourdan <ofourdan@redhat.com> 30516Date: Mon Nov 23 08:51:48 2015 +0100 30517 30518 xwayland: Do not set root clip when rootless 30519 30520 Otherwise the server may try to draw onto the root window when closing 30521 down, but when running rootless the root window has no storage thus 30522 causing a memory corruption. 30523 30524 Thanks to Adam Jackson <ajax@redhat.com> for helping tracking this down! 30525 30526 Reviewed-by: Adam Jackson <ajax@redhat.com> 30527 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93045 30528 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 30529 Tested-by: Marek Chalupa <mchqwerty@gmail.com> 30530 30531commit 5b2ca3413203210d112a08a4e20d14382abae437 30532Author: Marek Chalupa <mchqwerty@gmail.com> 30533Date: Fri Nov 27 14:27:46 2015 +0100 30534 30535 xwayland: check if creating xwl_output succeeded 30536 30537 check return values of RR.*Create calls 30538 30539 v2. do not bail out if we don't have any output 30540 30541 Signed-off-by: Marek Chalupa <mchqwerty@gmail.com> 30542 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 30543 30544commit 646ebea456b4c5251ae997eab8d9b971f97de836 30545Author: Marek Chalupa <mchqwerty@gmail.com> 30546Date: Fri Nov 27 14:59:27 2015 +0100 30547 30548 xwayland: fix memory leaks on error paths in xwl_realize_window 30549 30550 don't leak memory when realizing window fails 30551 30552 v2. take care of all memory allocation and return values, 30553 not just one leak 30554 30555 Signed-off-by: Marek Chalupa <mchqwerty@gmail.com> 30556 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 30557 30558commit e6b106715f24112d1dc7a84c6e37df4b5debb2d0 30559Author: Richard PALO <richard@NetBSD.org> 30560Date: Tue Nov 17 07:02:07 2015 +0100 30561 30562 Replace 'sun' with '__sun' 30563 30564 Globally replace #ifdef and #if defined usage of 'sun' with '__sun' 30565 such that strict ISO compiler modes such as -ansi or -std=c99 can be used. 30566 30567 Signed-off-by: Richard PALO <richard@NetBSD.org> 30568 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 30569 30570commit 44d0fd435a4eaf45e252b4f00409152a6d599dfc 30571Author: Egbert Eich <eich@suse.de> 30572Date: Tue Nov 24 17:37:36 2015 +0100 30573 30574 kdrive/UnregisterFd: Fix off by one 30575 30576 The number of FDs has been decremented already, therefore the 30577 number contained the index of the top one that is to me moved down. 30578 30579 This problem was introduced by: 30580 commit 1110b71e360195aab040d835b54540ab558638c5 30581 Author: Chris Clayton <chris2553@googlemail.com> 30582 30583 kdrive: fix build error on gcc 4.8 for out-of-bounds array access 30584 30585 The reason for the warning was likely a confused compiler. 30586 Hoping to reduce the confusion by moving the decrement behind the end 30587 if the copy loop. 30588 30589 Signed-off-by: Egbert Eich <eich@suse.de> 30590 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 30591 30592commit eb36924ead40564325aa56d54a973dc8fb4eae83 30593Author: Adam Jackson <ajax@redhat.com> 30594Date: Tue Nov 24 08:31:21 2015 -0500 30595 30596 dix: Remove redundant ChangeWindowProperty 30597 30598 Use dixChangeWindowProperty(serverClient, ...) instead. 30599 30600 Signed-off-by: Adam Jackson <ajax@redhat.com> 30601 Reviewed-by: Keith Packard <keithp@keithp.com> 30602 30603commit 4affa75a90d2455c81087b930126ad7adfd019f0 30604Author: Adam Jackson <ajax@redhat.com> 30605Date: Thu Nov 19 12:21:08 2015 -0500 30606 30607 xnest: Fix needless build dependency on xcb-util-keysyms 30608 30609 This was added in: 30610 30611 commit 43014795087a0a8774dd9687f5967329b15f06a2 30612 Author: Olivier Fourdan <ofourdan@redhat.com> 30613 Date: Mon Jan 5 16:44:22 2015 +0100 30614 30615 Synchronize capslock in Xnest and Xephyr 30616 30617 Which is fine if you're building both, but if you don't happen to have 30618 xcb-util-keysyms' headers installed Xnest will configure as enabled but 30619 fail to build. 30620 30621 Fortunately <X11/X.h> has a corresponding #define, so use that instead. 30622 30623 Signed-off-by: Adam Jackson <ajax@redhat.com> 30624 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 30625 30626commit fee0827a9a695600765f3d04376fc9babe497401 30627Author: Daniel Stone <daniels@collabora.com> 30628Date: Fri Nov 20 15:37:31 2015 +0000 30629 30630 XWayland: Use FocusIn events for keyboard enter 30631 30632 wl_keyboard::enter is the equivalent of FocusIn + KeymapNotify: it 30633 notifies us that the surface/window has now received the focus, and 30634 provides us a set of keys which are currently down. 30635 30636 We should use these keys to update the current state, but not to send 30637 any events to clients. 30638 30639 Signed-off-by: Daniel Stone <daniels@collabora.com> 30640 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 30641 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 30642 30643commit 816015648ffe660ddaa0f7d4d192e555b723c372 30644Author: Daniel Stone <daniels@collabora.com> 30645Date: Fri Nov 20 15:37:30 2015 +0000 30646 30647 Input: Add focus-in event source 30648 30649 Add a new event source type for keypress events synthesised from focus 30650 notifications (e.g. KeymapNotify from the parent server, when running 30651 nested). This is used to keep the keys-down array in sync with the host 30652 server's, without sending actual keypress events to clients. 30653 30654 Signed-off-by: Daniel Stone <daniels@collabora.com> 30655 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 30656 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 30657 30658commit c3788394e9190130a8eed44c5c93eeb93c2a9893 30659Author: Daniel Stone <daniels@collabora.com> 30660Date: Fri Nov 20 15:37:29 2015 +0000 30661 30662 Input: Add DeviceEventSource enum 30663 30664 Add a flag to DeviceEvents, giving the source of the event. Currently 30665 this only supports a 'normal' flag, but will be used later to add a 30666 'focus-in' flag, noting events synthesised from key/button arrays on 30667 focus-in notifications. 30668 30669 Signed-off-by: Daniel Stone <daniels@collabora.com> 30670 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 30671 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 30672 30673commit 2e61901e46d28ce2f436219ad1a495aa0dcd0fba 30674Author: Daniel Stone <daniels@collabora.com> 30675Date: Fri Nov 20 15:37:28 2015 +0000 30676 30677 XKB: Split filter execution into a separate function 30678 30679 Move the giant state machine which maps from a key action to actually 30680 running the filters into a separate function, to be used when adding 30681 KeyFocusIn. 30682 30683 Signed-off-by: Daniel Stone <daniels@collabora.com> 30684 Tested-by: Giulio Camuffo <giuliocamuffo@gmail.com> 30685 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 30686 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 30687 30688commit 71ba82690158f46d50a455e69a83ee0d685bb274 30689Author: Peter Hutterer <peter.hutterer@who-t.net> 30690Date: Mon Nov 23 07:59:24 2015 +1000 30691 30692 xfree86: fix minor memory leak 30693 30694 xf86*StrOption returns a strdup 30695 30696 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 30697 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 30698 30699commit 51984dddfcc7133ed3c1f20d03514aa98c9a7831 30700Author: Eric Anholt <eric@anholt.net> 30701Date: Thu Jun 18 11:14:41 2015 -0700 30702 30703 glamor: Delay making pixmaps shareable until we need to. 30704 30705 If a pixmap isn't getting exported as a dmabuf, then we don't need to 30706 make an EGLImage/GBM bo for it. This should reduce normal pixmap 30707 allocation overhead, and also lets the driver choose non-scanout 30708 formats which may be much higher performance. 30709 30710 On Raspberry Pi, where scanout isn't usable as a texture source, this 30711 improves x11perf -copypixwin100 from about 4300/sec to 5780/sec under 30712 xcompmgr -a, because we no longer need to upload our x11perf window to 30713 a tiled temporary in order to render it to the screen. 30714 30715 v2: Just use pixmap->usage_hint instead of a new field. Drop the 30716 changes that started storing gbm_bos in the pixmap priv due to 30717 lifetime issues. 30718 v3: Fix a missing gbm_bo_destroy() on the pixmap-from-fd success path. 30719 30720 Signed-off-by: Eric Anholt <eric@anholt.net> 30721 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30722 30723commit 7cd495a88807698b4ebaf9c1fb3db6edf31dd7e6 30724Author: Eric Anholt <eric@anholt.net> 30725Date: Thu Oct 15 13:25:12 2015 -0700 30726 30727 glamor: Make glamor_get_name_from_bo static. 30728 30729 Signed-off-by: Eric Anholt <eric@anholt.net> 30730 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30731 30732commit 6be33fd044949330e0b2b4185882c9664d2f90b4 30733Author: Eric Anholt <eric@anholt.net> 30734Date: Thu Jun 18 11:26:46 2015 -0700 30735 30736 glamor: Simplify DRI3 pixmap-from-fd, using GBM. 30737 30738 This GBM import path was introduced in 10.2, which we already require. 30739 30740 Signed-off-by: Eric Anholt <eric@anholt.net> 30741 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30742 30743commit 1b8f16d8e659fb483453e1123a9fa876adb758ff 30744Author: Eric Anholt <eric@anholt.net> 30745Date: Thu Jun 18 11:21:10 2015 -0700 30746 30747 glamor: Use real types for glamor_egl's public gbm functions. 30748 30749 I think void * was just used to avoid needing to #include gbm.h, but 30750 we can just forward-declare the structs and be fine. 30751 30752 Signed-off-by: Eric Anholt <eric@anholt.net> 30753 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30754 30755commit f80758f32a7b922baf8fbf3ac6d8c9aae5fea1c4 30756Author: Eric Anholt <eric@anholt.net> 30757Date: Thu Jun 18 11:15:40 2015 -0700 30758 30759 glamor: Use the GBM function for getting an FD from a GBM BO. 30760 30761 We were rolling ioctl calls ourselves, when there's a nice interface 30762 for it. 30763 30764 Signed-off-by: Eric Anholt <eric@anholt.net> 30765 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30766 30767commit ff2850424c99652506d0d6bc43506b4c16bf2ad5 30768Author: Eric Anholt <eric@anholt.net> 30769Date: Sun Oct 18 19:26:14 2015 -0700 30770 30771 glamor: Hook up EGL DestroyPixmap through the normal wrap chain. 30772 30773 One less layering violation (EGL should call glamor, if anything, not 30774 the other way around). 30775 30776 v2: Move glamor.c's DestroyPixmap wrapping up above the 30777 glamor_egl_screen_init() call, since glamor.c's DestroyPixmap 30778 needs to be the bottom of the stack (it calls fb directly and 30779 doesn't wrap). Caught by Michel. 30780 30781 Signed-off-by: Eric Anholt <eric@anholt.net> 30782 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30783 30784commit e91fd30049ba9ebfb6ee8aded74eebe006af3f57 30785Author: Eric Anholt <eric@anholt.net> 30786Date: Sun Oct 18 21:34:45 2015 -0700 30787 30788 glamor: Unexport glamor_destroy_textured_pixmap(). 30789 30790 This is just a bit of the DestroyPixmap chain. 30791 30792 Signed-off-by: Eric Anholt <eric@anholt.net> 30793 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30794 30795commit 3dd202933fd94615aeeaec7e4cfd05a68954a3f3 30796Author: Eric Anholt <eric@anholt.net> 30797Date: Sun Oct 18 21:28:19 2015 -0700 30798 30799 glamor: Remove glamor_egl_destroy_textured_pixmap(). 30800 30801 The DestroyPixmap chain and CloseScreen chain all do pixmap teardown 30802 already, and calling it manually would be redundant. 30803 30804 Signed-off-by: Eric Anholt <eric@anholt.net> 30805 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30806 30807commit 9d2b76652f0bca5680b9e3ae2aacd508d5525684 30808Author: Eric Anholt <eric@anholt.net> 30809Date: Fri Jun 19 15:56:13 2015 -0700 30810 30811 modesetting: No need to free the EGLImage just before freeing the pixmap. 30812 30813 DestroyPixmap handles that just fine. This also lets us drop our use 30814 of the manual image destruction function (Note that the radeon driver 30815 still uses it in a similar fashion, though). 30816 30817 Signed-off-by: Eric Anholt <eric@anholt.net> 30818 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30819 30820commit 98a1993536add730b7ec29a9e37f62b1cd70ad31 30821Author: Eric Anholt <eric@anholt.net> 30822Date: Sun Oct 18 19:16:20 2015 -0700 30823 30824 glamor: No need to glFlush before destroying a pixmap. 30825 30826 I assume this was a workaround for an old, broken, closed driver. The 30827 driver doesn't get to throw away rendering just because the rendering 30828 context's shared-across-processes render target is getting freed from 30829 the local address space. If the rendering isn't to a shared render 30830 target, then we *do* want to throw away the rendering to it. 30831 30832 Signed-off-by: Eric Anholt <eric@anholt.net> 30833 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30834 30835commit 9a5972801f7789833062e5711e77483b643eef92 30836Author: Eric Anholt <eric@anholt.net> 30837Date: Thu Nov 5 16:35:56 2015 -0800 30838 30839 glamor: Fix segfault in fallback picture uploading. 30840 30841 If the source/mask pixmap is a pixmap that doesn't have an FBO 30842 attached, and it doesn't match the Render operation's size, then we'll 30843 composite it to a CPU temporary (not sure why). We would take the 30844 PictFormatShort from the source Picture, make a pixmap of that depth, 30845 and try to look up the PictFormat description from the depth and the 30846 PictFormatShort. However, the screen's PictFormats are only attached 30847 to the screen's visuals' depths. So, with an x2r10g10b10 short format 30848 (depth 30), we wouldn't find the screen's PictFormat for it 30849 (associated with depth 32). 30850 30851 Instead of trying to look up from the screen, just use the pFormat 30852 that came from our source picture. The only time we need to look up a 30853 PictFormat when we're doing non-shader gradients, which we put in 30854 a8r8g8b8. 30855 30856 Signed-off-by: Eric Anholt <eric@anholt.net> 30857 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30858 30859commit e7aa4d3c7420d45cca2b7e1e69e22cebc64d5b74 30860Author: Eric Anholt <eric@anholt.net> 30861Date: Thu Nov 5 16:03:14 2015 -0800 30862 30863 glamor: Fix assert failures when fallback picture upload alloc fails. 30864 30865 If the glTexImage (or glTexSubImage) out-of-memories, error out 30866 cleanly so that we can fall back to software. 30867 30868 Signed-off-by: Eric Anholt <eric@anholt.net> 30869 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30870 30871commit ff8ef975df9cd99ec6f0b8b8047445091bf35ef0 30872Author: Eric Anholt <eric@anholt.net> 30873Date: Thu Nov 5 15:13:55 2015 -0800 30874 30875 glamor: Fix rendering when core font texture allocation fails. 30876 30877 Signed-off-by: Eric Anholt <eric@anholt.net> 30878 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30879 30880commit a6b05d10da2fe476f46e6dc4ad8a603964735905 30881Author: Eric Anholt <eric@anholt.net> 30882Date: Thu Nov 5 15:05:11 2015 -0800 30883 30884 glamor: Fix crashes when the glyph atlas allocation fails. 30885 30886 We already have a fallback path, so we just need to jump to it when we 30887 hit the failure. 30888 30889 Signed-off-by: Eric Anholt <eric@anholt.net> 30890 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30891 30892commit de959ec939b262cb1cb4c0b6146826e3092843f9 30893Author: Eric Anholt <eric@anholt.net> 30894Date: Thu Nov 5 14:47:42 2015 -0800 30895 30896 glamor: Handle GL_OUT_OF_MEMORY when allocating texture images. 30897 30898 The spec allows general undefined behavior when GL_OOM is thrown. But 30899 if the driver happens to throw the error at this point, it probably 30900 means the pixmap was just too big, so we should delete that texture 30901 and have this pixmap fall back to software. 30902 30903 Signed-off-by: Eric Anholt <eric@anholt.net> 30904 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30905 30906commit 74be466d40080545117628c376cb59b696db33bc 30907Author: Eric Anholt <eric@anholt.net> 30908Date: Mon Nov 9 15:47:05 2015 -0800 30909 30910 glamor: Avoid GL errors from mapping with size == 0. 30911 30912 GL 4.5 / GLES 3.0 require throwing GL errors at map time, and Mesa 30913 before that might throw errors accidentally if a malloc(0) call was 30914 made to return the mapping. 30915 30916 Signed-off-by: Eric Anholt <eric@anholt.net> 30917 Reviewed-by: Rob Clark <robdclark@gmail.com> 30918 30919commit a6cddb8c04ddc3c48aae3f3611ad9f336fecb09d 30920Author: Michael Stapelberg <stapelberg@google.com> 30921Date: Tue Nov 3 03:51:48 2015 -0800 30922 30923 Also dump passive grabs on XF86LogGrabInfo 30924 30925 Signed-off-by: Michael Stapelberg <stapelberg@google.com> 30926 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 30927 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 30928 30929commit f2ceb683c2da1ec08c8e07d7ec146f14864f2386 30930Author: Adam Jackson <ajax@redhat.com> 30931Date: Mon Nov 9 16:22:12 2015 -0500 30932 30933 Post-release version bump for 1.19 30934 30935 Signed-off-by: Adam Jackson <ajax@redhat.com> 30936 30937commit 43fb888bd01cf9d3d277e77a52a3d0c93ccff8bd 30938Author: Adam Jackson <ajax@redhat.com> 30939Date: Mon Nov 9 16:00:26 2015 -0500 30940 30941 xserver 1.18.0 30942 30943 Signed-off-by: Adam Jackson <ajax@redhat.com> 30944 30945commit 2203735887ab548b3ee004400d1b89149aae412c 30946Author: Axel Davy <axel.davy@ens.fr> 30947Date: Wed Nov 4 18:42:42 2015 +0100 30948 30949 present: Fix Async swap logic 30950 30951 According to the spec, PresentOptionAsync should only 30952 trigger a different behaviour when the target msc has been reached. 30953 30954 In this case if the driver is able to do async swaps, we use 30955 them to avoid a screen copy. 30956 30957 When the target msc hasn't been reached yet, we want to use sync swaps. 30958 30959 v2: Fix indentation and simplify checks for Async flips 30960 30961 Signed-off-by: Axel Davy <axel.davy@ens.fr> 30962 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30963 30964commit 3f35909acba117dc8934920d788c7ce612bce444 30965Author: Jammy Zhou <Jammy.Zhou@amd.com> 30966Date: Wed Oct 28 18:39:10 2015 +0800 30967 30968 present: Execute right away if target_msc equals current_msc 30969 30970 It is according to the protocol: 30971 30972 "If 'options' contains PresentOptionAsync, and the 'target-msc' 30973 is less than or equal to the current msc for 'window', then 30974 the operation will be performed as soon as possible, not 30975 necessarily waiting for the next vertical blank interval." 30976 30977 Signed-off-by: Jammy Zhou <Jammy.Zhou@amd.com> 30978 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 30979 Reviewed-by: Axel Davy <axel.davy@ens.fr> 30980 30981commit 7d1e4783853f9830344d101ceab087feb19995be 30982Author: Daniel Martin <consume.noise@gmail.com> 30983Date: Thu Oct 29 14:58:11 2015 +0100 30984 30985 modesetting: Remove XF86_CRTC_VERSION checks 30986 30987 The ifdef checks for XF86_CRTC_VERSION >= 3/5 are remnants from the 30988 out-of-tree driver. Within the tree, we can rely on: 30989 xf86Crtc.h:#define XF86_CRTC_VERSION 6 30990 30991 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 30992 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 30993 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 30994 30995commit 45c83a266d18eb515313aa3f1a4d7ff6af53be5d 30996Author: Daniel Martin <consume.noise@gmail.com> 30997Date: Thu Oct 29 14:58:10 2015 +0100 30998 30999 modesetting: Free output_ids in drmmode_set_mode_major() 31000 31001 We calloc() output_ids. Let's free() it, too. 31002 31003 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 31004 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 31005 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 31006 31007commit 2674d424020bd71d4f99b8d8de8b0b21aa490d54 31008Author: Daniel Martin <consume.noise@gmail.com> 31009Date: Thu Oct 29 14:58:09 2015 +0100 31010 31011 modesetting: Handle failures in setting a CRTC to a DRM mode properly 31012 31013 This fixes a bug where running the card out of PPLL's when hotplugging 31014 another monitor would result in all of the displays going blank and 31015 failing to work properly until X was restarted or the user switched to 31016 another VT. 31017 31018 [Michel Dänzer: Pass errno instead of -ret to strerror()] 31019 [Daniel Martin: Add \n to log message] 31020 31021 Picked from xf86-video-ati 31022 7186a87 Handle failures in setting a CRTC to a DRM mode properly 31023 31024 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 31025 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 31026 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 31027 31028commit 250666586e2b6f3ed1371340452dc2be2d094d40 31029Author: Adam Jackson <ajax@redhat.com> 31030Date: Thu Oct 29 10:08:17 2015 -0400 31031 31032 vidmode: Drop the unused event code 31033 31034 As the code says, this is "far from complete". So far, in fact, that 31035 it's been basically untouched for twenty years (XFree86 3.1!). As far 31036 as I can tell it was never enabled in any XFree86 build, and certainly 31037 has never been enabled since Xorg 7.0. 31038 31039 Also, K&R. 31040 31041 Signed-off-by: Adam Jackson <ajax@redhat.com> 31042 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 31043 31044commit 478efe285a440c33b053bdf0bfbfdd482f429f01 31045Author: Alan Coopersmith <alan.coopersmith@oracle.com> 31046Date: Tue Oct 27 19:44:28 2015 -0700 31047 31048 Xserver.man: document more transports for -nolisten & -listen options 31049 31050 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31051 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31052 31053commit 8fc295bde9a736f3c8c047031a6698d140d5266f 31054Author: Alan Coopersmith <alan.coopersmith@oracle.com> 31055Date: Tue Oct 27 19:44:27 2015 -0700 31056 31057 Xorg.man: update to reflect -nolisten tcp becoming default 31058 31059 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31060 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31061 31062commit 75157b7dbf4ed4db0492328a44e4e67dda83f769 31063Author: Alan Coopersmith <alan.coopersmith@oracle.com> 31064Date: Tue Oct 27 19:44:26 2015 -0700 31065 31066 Xorg.man: move XLOCAL details to X(7) man page instead 31067 31068 These settings affect clients, not server, so belong there, next to 31069 the information about how to set $DISPLAY. 31070 31071 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31072 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31073 31074commit 50c167164700e8ead9b7ccf9f9eafc7541baac75 31075Author: Martin Peres <martin.peres@linux.intel.com> 31076Date: Mon Jul 20 10:37:30 2015 +0300 31077 31078 os: make sure the clientsWritable fd_set is initialized before use 31079 31080 In WaitForSomething(), the fd_set clientsWritable may be used 31081 unitialized when the boolean AnyClientsWriteBlocked is set in the 31082 WakeupHandler(). This leads to a crash in FlushAllOutput() after 31083 x11proto's commit 2c94cdb453bc641246cc8b9a876da9799bee1ce7. 31084 31085 The problem did not manifest before because both the XFD_SIZE and the 31086 maximum number of clients were set to 256. As the connectionTranslation 31087 table was initalized for the 256 clients to 0, the test on the index not 31088 being 0 was aborting before dereferencing the client #0. 31089 31090 As of commit 2c94cdb453bc641246cc8b9a876da9799bee1ce7 in x11proto, the 31091 XFD_SIZE got bumped to 512. This lead the OutputPending fd_set to have 31092 any fd above 256 to be uninitialized which in turns lead to reading an 31093 index after the end of the ConnectionTranslation table. This index would 31094 then be used to find the client corresponding to the fd marked as 31095 pending writes and would also result to an out-of-bound access which 31096 would usually be the fatal one. 31097 31098 Fix this by zeroing the clientsWritable fd_set at the beginning of 31099 WaitForSomething(). In this case, the bottom part of the loop, which 31100 would indirectly call FlushAllOutput, will not do any work but the next 31101 call to select will result in the execution of the right codepath. This 31102 is exactly what we want because we need to know the writable clients 31103 before handling them. In the end, it also makes sure that the fds above 31104 MaxClient are initialized, preventing the crash in FlushAllOutput(). 31105 31106 Thanks to everyone involved in tracking this one down! 31107 31108 Reported-by: Karol Herbst <freedesktop@karolherbst.de> 31109 Reported-by: Tobias Klausmann <tobias.klausmann@mni.thm.de> 31110 Signed-off-by: Martin Peres <martin.peres@linux.intel.com> 31111 Tested-by: Tobias Klausmann <tobias.klausmann@mni.thm.de> 31112 Tested-by: Martin Peres <martin.peres@linux.intel.com> 31113 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91316 31114 Cc: Ilia Mirkin <imirkin@alum.mit.edu> 31115 Cc: Olivier Fourdan <ofourdan@redhat.com 31116 Cc: Adam Jackson <ajax@redhat.com> 31117 Cc: Alan Coopersmith <alan.coopersmith@oracle.com 31118 Cc: Chris Wilson <chris@chris-wilson.co.uk> 31119 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31120 31121commit bb78c464f09f515db557182a458b12b63c3b52d7 31122Author: Adam Jackson <ajax@redhat.com> 31123Date: Tue Oct 27 13:28:37 2015 -0400 31124 31125 build: Remove stale miext/cw include paths 31126 31127 Signed-off-by: Adam Jackson <ajax@redhat.com> 31128 Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com> 31129 31130commit 524844c8c18e226aad30feb371b19ef491d83449 31131Author: Julien Cristau <jcristau@debian.org> 31132Date: Tue Oct 27 13:23:13 2015 +0100 31133 31134 Xext: fix build with --disable-xace 31135 31136 Regression from 990cf5b2828f73dc7a07f1e38f608af39acfd81d 31137 31138 Signed-off-by: Julien Cristau <jcristau@debian.org> 31139 Cc: Andrew Eikum <aeikum@codeweavers.com> 31140 Cc: Peter Hutterer <peter.hutterer@who-t.net> 31141 Signed-off-by: Andrew Eikum <aeikum@codeweavers.com> 31142 31143commit ac2f27f1a9fa8cd88c5dbe7ec0f96238eecf2c3e 31144Author: Michel Dänzer <michel.daenzer@amd.com> 31145Date: Tue Oct 27 11:51:49 2015 +0900 31146 31147 DRI2: Sync radeonsi_pci_ids.h from Mesa 31148 31149 Fixes DRI2 client driver name mapping for newer AMD GPUs with the 31150 modesetting driver, allowing the DRI2 extension to initialize. 31151 31152 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 31153 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 31154 31155commit d8b7a900cf912cadb5915b3924dd6ce5a74505e7 31156Author: Adam Jackson <ajax@redhat.com> 31157Date: Mon Oct 26 13:47:04 2015 -0400 31158 31159 xserver 1.17.99.902 (1.18 RC2) 31160 31161 Signed-off-by: Adam Jackson <ajax@redhat.com> 31162 31163commit 47b00fa4bf3b67736957296492310f7fdd6c0a25 31164Author: Adam Jackson <ajax@redhat.com> 31165Date: Wed Oct 21 12:15:34 2015 -0400 31166 31167 xfree86: Use same inb/outb asm code for i386 amd64 and ia64 31168 31169 This matches the GCCUSESGAS path from the old monolith build (where that 31170 macro was actually set), and fixes the build on modern OSX. 31171 31172 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31173 Signed-off-by: Adam Jackson <ajax@redhat.com> 31174 31175commit c99fb550e06207e83ec89463fe32bd6bceca45f8 31176Author: Dave Airlie <airlied@gmail.com> 31177Date: Sun Sep 13 07:45:15 2015 +1000 31178 31179 xf86: don't add gpus from udev if autoAddGPU is set 31180 31181 At startup the server wasn't adding devices, but nothing 31182 was blocking hotplug devices by the look of it. 31183 31184 bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91388 31185 Signed-off-by: Dave Airlie <airlied@redhat.com> 31186 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 31187 31188commit 912f1fe2bb1b640d55fc44fcf636b6ca40d7f40b 31189Author: Chris Wilson <chris@chris-wilson.co.uk> 31190Date: Sun Apr 5 10:32:03 2015 +0100 31191 31192 Xv: Only stop the adaptors when the Pixmap is finally destroyed 31193 31194 Pixmaps are reference counted and DestroyPixmap is called for the 31195 removal of every reference. However, we only want to stop the adaptors 31196 writing into the Pixmap just before the Pixmap is finally destroyed, 31197 similar to how Windows are handled. 31198 31199 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 31200 Cc: Ville Syrjälä <ville.syrjala@linux.intel.com> 31201 Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com> 31202 31203commit 413cb2ff1d754b5f19d77ff19cddf40793989c03 31204Author: Chris Wilson <chris@chris-wilson.co.uk> 31205Date: Mon Feb 16 09:49:18 2015 +0000 31206 31207 present: Fix missed notify MSC computation 31208 31209 Only treat divisor==0 as async to immediately report the actual vblank. 31210 If the user species a non-zero divisor, we should compute the missed 31211 vblank properly or else we report too early. 31212 31213 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 31214 Reviewed-by: Mario Kleiner <mario.kleiner.de@gmail.com> 31215 31216commit 702c0a247f5523b43652befaf6be548ddbbe9dee 31217Author: Alberto Milone <alberto.milone@canonical.com> 31218Date: Thu Dec 12 10:00:09 2013 +0100 31219 31220 randr: make RROutputChanged change the main protocol screen not the gpu screen 31221 31222 We only set changes on the main protocol screen as, for example 31223 in RRSetChanged() and RRTellChanged(), therefore we should follow 31224 the same logic when reporting that an output changed in 31225 RROutputChanged(). 31226 31227 This means that RRTellChanged() will then update the relevant 31228 timestamps also when events come from gpu screens. 31229 31230 [ajax: Fix mixed code and decls] 31231 31232 Reviewed-by: Dave Airlie <airlied@redhat.com> 31233 Signed-off-by: Alberto Milone <alberto.milone@canonical.com> 31234 31235commit 2092f12a243b9f7682f542b593b77c96d455ec89 31236Author: Hans de Goede <hdegoede@redhat.com> 31237Date: Wed Oct 21 11:36:06 2015 +0200 31238 31239 linux: Do not call FatalError from xf86CloseConsole 31240 31241 FatalError ends up calling xf86CloseConsole itself, so calling FatalError 31242 from within xf86CloseConsole is not a good idea. 31243 31244 Make switch_to log errors using xf86Msg(X_WARNING, ...) and return success 31245 (or failure). 31246 31247 This makes switch_to match the other error checking done in xf86CloseConsole 31248 which all logs warnings and continues. 31249 31250 Add checking of the return value in xf86OpenConsole and call 31251 FatalError there when switch_to fails, to preserve the error-handling 31252 behavior of xf86OpenConsole. 31253 31254 BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1269210 31255 Reviewed-by: Adam Jackson <ajax@redhat.com> 31256 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 31257 31258commit ee06f674bbcd796324d6daf69bfb5d8856e94008 31259Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31260Date: Thu Oct 15 22:28:49 2015 -0700 31261 31262 Xi: Silence some tautological warnings 31263 31264 xichangehierarchy.c:424:23: warning: comparison of constant 536870911 with expression of type 'uint16_t' 31265 (aka 'unsigned short') is always false [-Wtautological-constant-out-of-range-compare,Semantic Issue] 31266 if (stuff->length > (INT_MAX >> 2)) 31267 ~~~~~~~~~~~~~ ^ ~~~~~~~~~~~~~~ 31268 xichangehierarchy.c:438:26: warning: comparison of constant 536870911 with expression of type 'uint16_t' 31269 (aka 'unsigned short') is always false [-Wtautological-constant-out-of-range-compare,Semantic Issue] 31270 if ((any->length > (INT_MAX >> 2)) || (len < (any->length << 2))) 31271 ~~~~~~~~~~~ ^ ~~~~~~~~~~~~~~ 31272 31273 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31274 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 31275 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 31276 31277commit 113c0bb4fd764da8c08d8f30abe350c7650c9fc2 31278Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 31279Date: Mon Oct 19 23:15:36 2015 +0200 31280 31281 hurd: fix xorg-wrapper build 31282 31283 hurd does not have any PATH_MAX limitation. misc.h provides a default value 31284 which is fine here. 31285 31286 Reviewed-by: Adam Jackson <ajax@redhat.com> 31287 Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 31288 31289commit f9a04d19aef77bf787b8d322305a6971d24a6ba1 31290Author: Keith Packard <keithp@keithp.com> 31291Date: Mon Sep 21 07:16:12 2015 +0100 31292 31293 fonts: Continue when font calls return Suspended more than once 31294 31295 Patch 3ab6cd31cbdf8095b2948034fce5fb645422d8da fixed Xinerama 31296 interactions with font servers by not putting clients to sleep 31297 multiple times. However, it introduced additional changes dealing with 31298 libXfont routine returning Suspended more than once for the same 31299 request. This additional change was to abandon processing of the 31300 current request and free the closure data by jumping to 31301 'xinerama_sleep' in each of the functions. 31302 31303 Font library functions shouldn't return Suspended more than once, 31304 except for ListFontsWithInfo, which produces multiple replies, and 31305 thus ends up returning Suspended many times during processing. 31306 31307 With the jump to xinerama_sleep occurring after the first reply was 31308 processed, the closure for the request was freed and future calls into 31309 the ListFontsWithInfo callback resulted in dereferencing freed 31310 memory. 31311 31312 This patch removes the added branches, reverting the code to its 31313 previous behaviour, which permitted multiple Suspended returns and 31314 simply waited for the client to be signaled again so that the callback 31315 could continue processing the request. 31316 31317 Signed-off-by: Keith Packard <keithp@keithp.com> 31318 Reviewed-by: Adam Jackson <ajax@redhat.com> 31319 31320commit 5b582a4a0350c253d729efb31b710851ae9a958e 31321Merge: 1d4aa6724 27ad21254 31322Author: Adam Jackson <ajax@redhat.com> 31323Date: Mon Oct 19 12:23:22 2015 -0400 31324 31325 Merge remote-tracking branch 'jeremyhu/master' 31326 31327commit 1d4aa672424d8b1629fda11400b88607b5066965 31328Author: Julien Cristau <jcristau@debian.org> 31329Date: Mon Oct 19 15:42:30 2015 +0200 31330 31331 xorg-wrapper: when starting the server as root, reset its environment 31332 31333 When the server is privileged, we shouldn't be passing the user's 31334 environment directly. 31335 31336 Clearing the environment is recommended by the libdbus maintainers, see 31337 https://bugs.freedesktop.org/show_bug.cgi?id=52202 31338 31339 v2: rename envp to empty_envp (Jeremy) 31340 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 31341 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83849 31342 Signed-off-by: Julien Cristau <jcristau@debian.org> 31343 31344commit 08c4912406b965bbac0a3a52413c374a073c051b 31345Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 31346Date: Mon Oct 19 14:47:43 2015 +0200 31347 31348 xorg-wrapper: fix build without libdrm 31349 31350 Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 31351 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 31352 31353commit a4cd8ee5f8e8de776c7f764656770311cdbde0d2 31354Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 31355Date: Mon Oct 19 14:47:42 2015 +0200 31356 31357 hurd: disable detecting drm 31358 31359 Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 31360 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 31361 31362commit 27ad21254f257bac6c647315d749ee69f20c24e0 31363Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31364Date: Sun Oct 18 23:01:53 2015 -0700 31365 31366 XQuartz: Cleanup formatting of DarwinEQInit that was butchered by automation a few years ago 31367 31368 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31369 31370commit 3db7e332d374bf8cee581c31b7d50d7ac0509187 31371Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31372Date: Sun Oct 18 23:03:23 2015 -0700 31373 31374 XQuartz: Make sure that darwin_all_modifier_mask_additions is 0-terminated 31375 31376 Found by ASan 31377 31378 X.Org X Server 1.17.99.901 Build Date: 20151018 31379 ================================================================ 31380 ==40471==ERROR: AddressSanitizer: global-buffer-overflow on address 0x000101fed7a4 at pc 0x000101584030 bp 0x70000029f920 sp 0x70000029f918 31381 READ of size 4 at 0x000101fed7a4 thread T7 31382 #0 0x10158402f in DarwinEQInit darwinEvents.c:377 31383 #1 0x10157f3bc in InitInput darwin.c:566 31384 #2 0x101be87ad in dix_main main.c:268 31385 #3 0x10159131b in server_thread quartzStartup.c:66 31386 #4 0x7fff8a535c12 in _pthread_body (/usr/lib/system/libsystem_pthread.dylib+0x3c12) 31387 #5 0x7fff8a535b8f in _pthread_start (/usr/lib/system/libsystem_pthread.dylib+0x3b8f) 31388 #6 0x7fff8a533374 in thread_start (/usr/lib/system/libsystem_pthread.dylib+0x1374) 31389 31390 0x000101fed7a4 is located 0 bytes to the right of global variable 'darwin_all_modifier_mask_additions' defined in 'darwinEvents.c:181:12' 31391 (0x101fed7a0) of size 4 31392 SUMMARY: AddressSanitizer: global-buffer-overflow darwinEvents.c:377 DarwinEQInit 31393 Shadow bytes around the buggy address: 31394 0x1000203fdaa0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 31395 0x1000203fdab0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 31396 0x1000203fdac0: f9 f9 f9 f9 00 00 00 00 00 00 f9 f9 f9 f9 f9 f9 31397 0x1000203fdad0: 00 00 00 00 00 00 00 00 f9 f9 f9 f9 00 00 00 00 31398 0x1000203fdae0: 00 00 f9 f9 f9 f9 f9 f9 00 00 00 00 00 f9 f9 f9 31399 =>0x1000203fdaf0: f9 f9 f9 f9[04]f9 f9 f9 f9 f9 f9 f9 00 00 00 00 31400 0x1000203fdb00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 31401 0x1000203fdb10: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 31402 0x1000203fdb20: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 31403 0x1000203fdb30: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 31404 0x1000203fdb40: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 31405 Shadow byte legend (one shadow byte represents 8 application bytes): 31406 Addressable: 00 31407 Partially addressable: 01 02 03 04 05 06 07 31408 Heap left redzone: fa 31409 Heap right redzone: fb 31410 Freed heap region: fd 31411 Stack left redzone: f1 31412 Stack mid redzone: f2 31413 Stack right redzone: f3 31414 Stack partial redzone: f4 31415 Stack after return: f5 31416 Stack use after scope: f8 31417 Global redzone: f9 31418 Global init order: f6 31419 Poisoned by user: f7 31420 Container overflow: fc 31421 Array cookie: ac 31422 Intra object redzone: bb 31423 ASan internal: fe 31424 Left alloca redzone: ca 31425 Right alloca redzone: cb 31426 Thread T7 created by T0 here: 31427 #0 0x10242ee99 in wrap_pthread_create 31428 (/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/lib/clang/7.0.0/lib/darwin/libclang_rt.asan_osx_dynamic.dylib+0x37e99) 31429 #1 0x101591089 in create_thread quartzStartup.c:78 31430 #2 0x101590ed9 in QuartzInitServer quartzStartup.c:95 31431 #3 0x1015697eb in X11ApplicationMain X11Application.m:1277 31432 #4 0x101575dc0 in X11ControllerMain X11Controller.m:984 31433 #5 0x10159171a in server_main quartzStartup.c:127 31434 #6 0x101540fc0 in do_start_x11_server bundle-main.c:436 31435 #7 0x101544869 in _Xstart_x11_server mach_startupServer.c:189 31436 #8 0x101545c96 in mach_startup_server mach_startupServer.c:398 31437 #9 0x7fff8d1b70f3 in mach_msg_server (/usr/lib/system/libsystem_kernel.dylib+0x110f3) 31438 #10 0x1015416e7 in main bundle-main.c:774 31439 #11 0x7fff8bd975ac in start (/usr/lib/system/libdyld.dylib+0x35ac) 31440 #12 0x0 (<unknown module>) 31441 31442 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31443 31444commit ec6294116cc41ff1c3be081b626952fb7e614244 31445Author: Jeremy Huddleston <jeremyhu@apple.com> 31446Date: Sun Oct 18 23:12:52 2015 -0700 31447 31448 osinit: Silence -Wunused-variable warnings 31449 31450 osinit.c:161:24: warning: unused variable 'devnull' [-Wunused-variable,Unused Entity Issue] 31451 static const char *devnull = "/dev/null"; 31452 ^ 31453 osinit.c:162:10: warning: unused variable 'fname' [-Wunused-variable,Unused Entity Issue] 31454 char fname[PATH_MAX]; 31455 ^ 31456 31457 Reviewed-by: Adam Jackson <ajax@redhat.com> 31458 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31459 31460commit 0a5d54f721de7e3ab3b6f4b080190a92bbe3429b 31461Author: Jeremy Huddleston <jeremyhu@apple.com> 31462Date: Sun Oct 18 23:12:51 2015 -0700 31463 31464 dix: Silence -Wunused-variable warning by moving window.c off of legacy region defines 31465 31466 window.c:223:15: warning: unused variable 'pScreen' [-Wunused-variable,Unused Entity Issue] 31467 ScreenPtr pScreen = pWin->drawable.pScreen; 31468 ^ 31469 31470 Reviewed-by: Adam Jackson <ajax@redhat.com> 31471 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31472 31473commit e09875701b980b8c4578fb310a922c9934c34eef 31474Author: Jeremy Huddleston <jeremyhu@apple.com> 31475Date: Sun Oct 18 23:12:50 2015 -0700 31476 31477 glx: Fix header length error checking in __glXDisp_RenderLarge 31478 31479 glxcmds.c:2206:46: warning: comparison of unsigned expression < 0 is always false [-Wtautological-compare,Semantic Issue] 31480 if ((cmdlen = safe_pad(hdr->length)) < 0) 31481 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~ 31482 31483 Reviewed-by: Adam Jackson <ajax@redhat.com> 31484 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31485 31486commit 5dc415048e4091b18cd7d123ebeae8f95ed5a4f5 31487Author: Jeremy Huddleston <jeremyhu@apple.com> 31488Date: Sun Oct 18 23:12:49 2015 -0700 31489 31490 randr: Silence -Wshift-negative-value warnings 31491 31492 rrtransform.c:199:23: warning: shifting a negative signed value is undefined [-Wshift-negative-value,Semantic Issue] 31493 rot_cos = F(-1); 31494 ^~~~~ 31495 rrtransform.c:114:14: note: expanded from macro 'F' 31496 ^~~~~~~~~~~~~~ 31497 ../render/picture.h:200:24: note: expanded from macro 'IntToxFixed' 31498 ^~~~~~~~~~~~~~~~~~~~~~ 31499 /opt/X11/include/pixman-1/pixman.h:130:56: note: expanded from macro 'pixman_int_to_fixed' 31500 ~~~ ^ 31501 31502 Reviewed-by: Adam Jackson <ajax@redhat.com> 31503 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31504 31505commit 9a2a05a9a7ba02921fa29844c4cad41243c41326 31506Author: Jeremy Huddleston <jeremyhu@apple.com> 31507Date: Wed Oct 14 15:13:42 2015 -0700 31508 31509 xdmcp: Declare XdmcpFatal _X_NORETURN 31510 31511 xdmcp.c:1404:1: warning: function 'XdmcpFatal' could be declared with attribute 'noreturn' 31512 [-Wmissing-noreturn,Semantic Issue] 31513 31514 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31515 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31516 31517commit a7a00e4bd8153e0386c7e58d6b30a7a96f15fdb0 31518Author: Jeremy Huddleston <jeremyhu@apple.com> 31519Date: Wed Oct 14 15:13:39 2015 -0700 31520 31521 security: Silence some benign -Wformat warnings 31522 31523 XID may be either 'unsigned long' or 'unsigned int' depending on: 31524 31525 typedef unsigned long CARD64; 31526 typedef unsigned int CARD32; 31527 typedef unsigned long long CARD64; 31528 typedef unsigned long CARD32; 31529 31530 typedef unsigned long XID; 31531 typedef CARD32 XID; 31532 31533 so when building with -Wformat, we get some warnings that are benign. This silences them. 31534 31535 security.c:215:52: warning: format specifies type 'int' but the argument has type 'XID' (aka 'unsigned long') 31536 [-Wformat,Format String Issue] 31537 SecurityAudit("revoked authorization ID %d\n", pAuth->id); 31538 ~~ ^~~~~~~~~ 31539 %lu 31540 CC dpmsstubs.lo 31541 security.c:553:25: warning: format specifies type 'int' but the argument has type 'XID' (aka 'unsigned long') 31542 [-Wformat,Format String Issue] 31543 client->index, pAuth->id, pAuth->trustLevel, pAuth->timeout, 31544 ^~~~~~~~~ 31545 security.c:553:55: warning: format specifies type 'int' but the argument has type 'CARD32' (aka 'unsigned long') 31546 [-Wformat,Format String Issue] 31547 client->index, pAuth->id, pAuth->trustLevel, pAuth->timeout, 31548 ^~~~~~~~~~~~~~ 31549 security.c:554:10: warning: format specifies type 'int' but the argument has type 'XID' (aka 'unsigned long') 31550 [-Wformat,Format String Issue] 31551 pAuth->group, eventMask); 31552 ^~~~~~~~~~~~ 31553 security.c:554:24: warning: format specifies type 'int' but the argument has type 'Mask' (aka 'unsigned long') 31554 [-Wformat,Format String Issue] 31555 pAuth->group, eventMask); 31556 ^~~~~~~~~ 31557 security.c:781:19: warning: format specifies type 'unsigned int' but the argument has type 'Mask' (aka 'unsigned 31558 long') 31559 [-Wformat,Format String Issue] 31560 requested, rec->id, cid, 31561 ^~~~~~~~~ 31562 security.c:781:30: warning: format specifies type 'unsigned int' but the argument has type 'XID' (aka 'unsigned long') 31563 [-Wformat,Format String Issue] 31564 requested, rec->id, cid, 31565 ^~~~~~~ 31566 security.c:863:23: warning: format specifies type 'unsigned int' but the argument has type 'XID' (aka 'unsigned long') 31567 [-Wformat,Format String Issue] 31568 rec->pWin->drawable.id, wClient(rec->pWin)->index, 31569 ^~~~~~~~~~~~~~~~~~~~~~ 31570 security.c:893:31: warning: format specifies type 'unsigned int' but the argument has type 'XID' (aka 'unsigned long') 31571 [-Wformat,Format String Issue] 31572 rec->pWin->drawable.id, 31573 ^~~~~~~~~~~~~~~~~~~~~~ 31574 security.c:915:39: warning: format specifies type 'unsigned int' but the argument has type 'XID' (aka 'unsigned long') 31575 [-Wformat,Format String Issue] 31576 rec->client->index, rec->pWin->drawable.id, 31577 ^~~~~~~~~~~~~~~~~~~~~~ 31578 31579 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31580 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31581 31582commit aa73d587fece225753d8e6b8773a9c8b85823bd9 31583Author: Jeremy Huddleston <jeremyhu@apple.com> 31584Date: Wed Oct 14 15:13:38 2015 -0700 31585 31586 xres: Silence -Wunused-function warnings when building !COMPOSITE or !RENDER 31587 31588 xres.c:422:1: warning: unused function 'ResFindCompositeClientWindowPixmaps' [-Wunused-function,Unused Entity Issue] 31589 ResFindCompositeClientWindowPixmaps (void *value, XID id, void *cdata) 31590 ^ 31591 31592 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31593 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31594 31595commit 85eb90ea45e89033b97bf71a13c5c70fec8f6871 31596Author: Jeremy Huddleston <jeremyhu@apple.com> 31597Date: Wed Oct 14 15:13:36 2015 -0700 31598 31599 xdmauth: Correct miscall of abs() to instrad call labs() 31600 31601 xdmauth.c:230:13: warning: absolute value function 'abs' given an argument of type 'long' but has parameter of 31602 type 31603 'int' 31604 which may cause truncation of value [-Wabsolute-value,Semantic Issue] 31605 if (abs(now - client->time) > TwentyFiveMinutes) { 31606 ^ 31607 xdmauth.c:230:13: note: use function 'labs' instead [Semantic Issue] 31608 if (abs(now - client->time) > TwentyFiveMinutes) { 31609 ^~~ 31610 labs 31611 xdmauth.c:302:9: warning: absolute value function 'abs' given an argument of type 'long' but has parameter of type 31612 'int' which 31613 may cause truncation of value [-Wabsolute-value,Semantic Issue] 31614 if (abs(client->time - now) > TwentyMinutes) { 31615 ^ 31616 xdmauth.c:302:9: note: use function 'labs' instead [Semantic Issue] 31617 if (abs(client->time - now) > TwentyMinutes) { 31618 ^~~ 31619 labs 31620 31621 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31622 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31623 31624commit 9f0fcd14b52f8481cbb3b3b9c6e06f64ff003cc8 31625Author: Jeremy Huddleston <jeremyhu@apple.com> 31626Date: Wed Oct 14 15:13:35 2015 -0700 31627 31628 randr: Correct a miscall of abs() to instead call fabs() 31629 31630 rrtransform.c:124:22: warning: using integer absolute value function 'abs' when 31631 argument is of floating point type [-Wabsolute-value,Semantic Issue] 31632 if ((v = abs(f_transform->m[j][i])) > max) 31633 ^ 31634 rrtransform.c:124:22: note: use function 'fabs' instead [Semantic Issue] 31635 if ((v = abs(f_transform->m[j][i])) > max) 31636 ^~~ 31637 fabs 31638 31639 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31640 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31641 31642commit 610dd8a58a75bb6a5b7d7abbae476d1cc4be519e 31643Author: Jeremy Huddleston <jeremyhu@apple.com> 31644Date: Wed Oct 14 15:13:34 2015 -0700 31645 31646 mi: Correct a miscall of abs() to instead call fabs() 31647 31648 miarc.c:1714:9: warning: using integer absolute value function 31649 'abs' when 31650 argument is of floating point type [-Wabsolute-value,Semantic Issue] 31651 if (abs(parc->angle2) >= 360.0) 31652 ^ 31653 miarc.c:1714:9: note: use function 'fabs' instead [Semantic Issue] 31654 if (abs(parc->angle2) >= 360.0) 31655 ^~~ 31656 fabs 31657 31658 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31659 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31660 31661commit 4513f924a7065edbd267cf96837af94ce5a58a6f 31662Author: Ken Thomases <ken@codeweavers.com> 31663Date: Mon Oct 12 16:45:41 2015 -0700 31664 31665 XQuartz: Fix how we calculate the height of the OSX menu bar 31666 31667 +[NSScreen mainScreen] does not mean the primary display. It used to mean the 31668 one with the key window. When "Displays have separate spaces" is enabled, it 31669 means the active screen, the one whose menu bar is mostly opaque. As such, it 31670 may not be the screen whose lower-left corner is located at (0, 0). That's 31671 why its max-Y is not necessarily comparable to its height. That only works 31672 for the primary display. 31673 31674 This code could use [[NSScreen screens] firstObject]. This is always the 31675 primary display, the one whose lower-left corner is at (0, 0). 31676 31677 Once that's done, the above change should be reverted. The height of the 31678 visible frame would be the full height of the screen minus the menu bar _and 31679 the Dock_ if the Dock is along the bottom of the screen. 31680 31681 Actually, there's a theoretically-simpler approach: use 31682 -[NSMenu menuBarHeight]. That replaces a long-deprecated method 31683 +[NSMenuView menuBarHeight]. However, there was a bug in Tiger that led to 31684 the former not working while the latter still worked. I haven't actually 31685 checked recently. 31686 31687 CrossOver's still-kicking X server code uses this code, which tries all of 31688 the above: 31689 31690 NSScreen* primaryScreen = [[NSScreen screens] objectAtIndex:0]; 31691 aquaMenuBarHeight = [[NSApp mainMenu] menuBarHeight]; 31692 if (!aquaMenuBarHeight) aquaMenuBarHeight = [NSMenuView menuBarHeight]; 31693 if (!aquaMenuBarHeight) aquaMenuBarHeight = 31694 NSHeight([primaryScreen frame]) - NSMaxY([primaryScreen visibleFrame]); 31695 31696 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31697 Signed-off-by: Ken Thomases <ken@codeweavers.com> 31698 31699commit 6e6827aac33eddeacfb37ef3d7b0397b1446bf66 31700Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31701Date: Mon Oct 12 02:06:10 2015 -0700 31702 31703 XQuartz: Remove InfoPlist.strings 31704 31705 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31706 31707commit df80e2649aa47957986554928e6bf41b86c15240 31708Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31709Date: Mon Oct 12 02:04:11 2015 -0700 31710 31711 XQuartz: Relax App Transport Security for communicating with the update server 31712 31713 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 31714 31715commit 880d4e78b1823bcc0e66e68270af1a475f097bb5 31716Author: Jonas Ådahl <jadahl@gmail.com> 31717Date: Wed Oct 7 12:02:39 2015 +0800 31718 31719 xwayland: Set physical screen size to something 31720 31721 When we have a single output, just set it to the physical size of that 31722 output. Otherwise try to approximate it calculating a mean m.m. per 31723 dot. Last fallback is to default to 96 DPI. 31724 31725 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 31726 Reviewed-by: Daniel Stone <daniels@collabora.com> 31727 31728commit 216bdbc7351e2078d94857a3323f6cd8f0d2f191 31729Author: Jonas Ådahl <jadahl@gmail.com> 31730Date: Wed Oct 7 12:02:38 2015 +0800 31731 31732 xwayland: Update actual screen and root window size on output hot plug 31733 31734 When a new output is hot-plugged we need to not only update our internal 31735 screen dimensions, but also the dix screen dimensions, screenInfo 31736 dimensions and the root window dimensions. 31737 31738 https://bugs.freedesktop.org/show_bug.cgi?id=92273 31739 31740 Signed-off-by: Jonas Ådahl <jadahl@gmail.com> 31741 Reviewed-by: Daniel Stone <daniels@collabora.com> 31742 31743commit 211d4c2d353b5e379716484055a3f58235ea65f4 31744Author: Chris Wilson <chris@chris-wilson.co.uk> 31745Date: Wed Dec 14 15:55:22 2011 +0000 31746 31747 render: Propagate allocation failure from createSourcePicture() 31748 31749 All the callers were already checking for failure, except that 31750 createSourcePicture() itself was failing to check whether it 31751 successfully allocated the Picture. 31752 31753 [ajax: Rebase, fix line wrap of preceding line] 31754 31755 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 31756 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 31757 31758commit 8a26a4aff570744347a8bed19f693bf631be464e 31759Author: Alexandr Shadchin <alexandr.shadchin@gmail.com> 31760Date: Sun Nov 27 02:49:02 2011 +0600 31761 31762 bsd: Remove LINKKIT 31763 31764 Reviewed-by: Adam Jackson <ajax@redhat.com> 31765 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 31766 31767commit ab3e34bf9ddff90261fbbe0e2604887459cd70ad 31768Author: Alexandr Shadchin <alexandr.shadchin@gmail.com> 31769Date: Sun Nov 27 02:49:01 2011 +0600 31770 31771 xfree86: Remove 386BSD 31772 31773 Reviewed-by: Adam Jackson <ajax@redhat.com> 31774 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 31775 31776commit c877048c8509a03f6aa69e533ab86c1039cfb6df 31777Author: Jon TURNEY <jon.turney@dronecode.org.uk> 31778Date: Thu Jun 30 15:30:40 2011 +0100 31779 31780 glx: Synchronize Xserver glx/rensize.c with mesa src/glx/compsize.c 31781 31782 Noticed while investigating https://bugs.freedesktop.org/show_bug.cgi?id=30102 31783 31784 Reviewed-by: Adam Jackson <ajax@redhat.com> 31785 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 31786 31787commit d08ac3660651c8614a1780750b76c48eeb0870b7 31788Author: Jamey Sharp <jamey@minilop.net> 31789Date: Thu May 12 15:21:52 2011 -0700 31790 31791 kdrive: Delete unused TOUCHSCREEN define. 31792 31793 There's nothing in configure to enable this, and KdTsPhyScreen isn't 31794 defined anywhere. 31795 31796 [ajax: Rebase, also clean up Xfbdev] 31797 31798 Reviewed-by: Adam Jackson <ajax@redhat.com> 31799 Signed-off-by: Jamey Sharp <jamey@minilop.net> 31800 31801commit e31fe8115ee080b58b2e96a5106f38e64944ce5e 31802Author: Olivier Fourdan <ofourdan@redhat.com> 31803Date: Wed Sep 23 15:35:58 2015 +0200 31804 31805 xwayland: Adjust screen size with rotation 31806 31807 bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92076 31808 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 31809 Tested-by: Artem Chudinov <arzeth0@gmail.com> 31810 Reviewed-by: Dave Airlie <airlied@redhat.com> 31811 31812commit f6ce23fbfc8804204fa103f98b94478387b94040 31813Author: Fredrik Höglund <fredrik@kde.org> 31814Date: Thu Sep 10 23:33:20 2015 +0200 31815 31816 present: Don't stash the MSC value when present_get_ust_msc fails 31817 31818 Otherwise we stash an uninitalized value, and later use it to compute 31819 the msc_offset for the window. Also initialize ust and crtc_msc so we 31820 never use uninitalized values when present_get_ust_msc fails. 31821 31822 This fixes clients getting stuck waiting indefinitely for an idle 31823 event when a CRTC is turned off. 31824 31825 Signed-off-by: Fredrik Höglund <fredrik@kde.org> 31826 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 31827 31828commit c1f5f9022f31ca4d662cc125fa80cb25386da6d1 31829Author: Alan Coopersmith <alan.coopersmith@oracle.com> 31830Date: Sun Sep 27 18:41:35 2015 -0700 31831 31832 glamor: Get rid of an extraneous ; at the end of a C source line 31833 31834 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31835 Signed-off-by: Eric Anholt <eric@anholt.net> 31836 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 31837 31838commit 10e9116b3f709bec6d6a50446c1341441a0564e4 31839Author: Rui Matos <tiagomatos@gmail.com> 31840Date: Tue Jul 15 15:57:20 2014 +0200 31841 31842 xwayland-input: Always set the xkb group index on modifiers events 31843 31844 While we have keyboard focus, the server's xkb code is already locking 31845 and latching modifiers appropriately while processing keyboard 31846 events. 31847 31848 Since there is no guaranteed order between wl_keyboard key and 31849 modifiers events, if we got the modifiers event with a locked or 31850 latched modifier and then process the key press event for that 31851 modifier we would wrongly unlock/unlatch. To prevent this, we ignore 31852 locked and latched modifiers while any of our surfaces has keyboard 31853 focus. 31854 31855 But we always need to set the xkb group index since this might be 31856 triggered programatically by the wayland compositor at any time. 31857 31858 Reviewed-by: Daniel Stone <daniels@collabora.com> 31859 31860commit f937032ee6526af3c9206a2d8c6eacdeeb490d3b 31861Author: Andreas Wettstein <wettstein509@solnet.ch> 31862Date: Sat Feb 15 17:37:24 2014 +0100 31863 31864 xkb: Keyboard mouse button emulation should not suppress other keyboard events 31865 31866 With this change, when a key mapped to an action to emulate mouse button 31867 presses and releases is held down, other keys pressed during that time are 31868 still processed normally. This is a prerequisite for proper support of 31869 ISOLock. 31870 31871 Signed-off-by: Andreas Wettstein <wettstein509@solnet.ch> 31872 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 31873 31874commit 5447ac45bc090e8f3269979af4db55f619c5f278 31875Author: Andreas Wettstein <wettstein509@solnet.ch> 31876Date: Sat Feb 15 17:36:51 2014 +0100 31877 31878 xkb: Suppress autorepeat for Set and Lock of Mods, Groups, and Controls 31879 31880 The autorepeat for these actions was not correctly implemented, as the key 31881 repeat would be mistakenly interpreted as key releases. Rather than fixing 31882 this, this change simply disables autorepeat for Set/Lock actions, for two 31883 reasons: 31884 31885 - Autorepeating Set/Lock keys make complicate the interactions of actions. 31886 31887 - Autorepeating Set/Lock keys have no apparent benefit, but hurt in the real 31888 world for layouts such as de(neo): Neo has a Level5 shift on the LSGT key, 31889 and a Level5 lock on Level5 of the same key. This is unusable if LSGT 31890 autorepeats. However, disabling autorepeat for key LSGT completely is not 31891 ideal for users that have a "usual" layout besides Neo, where LSGT carries 31892 symbols. 31893 31894 Signed-off-by: Andreas Wettstein <wettstein509@solnet.ch> 31895 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 31896 31897commit fbac451724f7e1b8ce6b6e0505cc33ccd41c5158 31898Author: Arcady Goldmints-Orlov <arcadyg@nvidia.com> 31899Date: Thu Oct 4 14:22:13 2012 -0700 31900 31901 Fix alphamap interactions with wfb 31902 31903 Set a destructor function on pixman images and call fbFinishAccess() 31904 from there, rather than directly from free_pixman_pict(). This ensures 31905 that fbFinishAccess() gets called even if pixman still has a reference 31906 to the image after free_pixman_pict(), as is the case for alphamaps. 31907 31908 [ajax: Squash an unused variable warning in the non-wfb build] 31909 31910 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 31911 Signed-off-by: Arcady Goldmints-Orlov <arcadyg@nvidia.com> 31912 Reviewed-by: Søren Sandmann <ssp@redhat.com> 31913 31914commit 19a87257f92e7f57a167fc03fba014ee22d8e1ab 31915Author: Emil Velikov <emil.l.velikov@gmail.com> 31916Date: Fri Sep 25 14:04:34 2015 +0100 31917 31918 glamor: move 'dri3_capable' extension checks after eglMakeCurrent 31919 31920 Otherwise we'll fail and/or crash as no context is bound. 31921 31922 Fixes: 64e6124f27e (glamor: move GL_OES_EGL_image check next to EGL_EXT_image_dma_buf_import) 31923 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92105 31924 Cc: Michel Dänzer <michel@daenzer.net> 31925 Cc: Nick Sarnie <commendsarnex@gmail.com> 31926 Reported-by: Nick Sarnie <commendsarnex@gmail.com> 31927 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 31928 Tested-by: Nick Sarnie <commendsarnex@gmail.com> 31929 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 31930 31931commit dca5770af9e20bb1148374ebfd60931a81b148a2 31932Merge: a31bbc450 21f384b7b 31933Author: Adam Jackson <ajax@redhat.com> 31934Date: Fri Sep 25 10:23:28 2015 -0400 31935 31936 Merge remote-tracking branch 'mchalupa/output-bugs' 31937 31938commit a31bbc450a08622aadafc264b3efe57e465eaecb 31939Author: Alan Coopersmith <alan.coopersmith@oracle.com> 31940Date: Tue Sep 22 19:25:43 2015 -0700 31941 31942 protocol.txt: Add MIT-SHM 1.2 requests 31943 31944 Reviewed-by: Adam Jackson <ajax@redhat.com> 31945 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31946 31947commit 424ba5de56e2ddf173bb0d26daa821f12897f18f 31948Author: Alan Coopersmith <alan.coopersmith@oracle.com> 31949Date: Tue Sep 22 19:25:42 2015 -0700 31950 31951 protocol.txt: Add DRI3 1.0 requests 31952 31953 Reviewed-by: Adam Jackson <ajax@redhat.com> 31954 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31955 31956commit 7e6b7c799455fdeac082cf2a064fe1ba432e92ca 31957Author: Alan Coopersmith <alan.coopersmith@oracle.com> 31958Date: Tue Sep 22 19:25:41 2015 -0700 31959 31960 protocol.txt: Add Present 1.0 requests 31961 31962 Reviewed-by: Adam Jackson <ajax@redhat.com> 31963 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31964 31965commit fd653194be731e4e32f95f38446e0dcbf56d04c9 31966Author: Alan Coopersmith <alan.coopersmith@oracle.com> 31967Date: Tue Sep 22 19:25:40 2015 -0700 31968 31969 protocol.txt: Add RANDR 1.5 requests 31970 31971 Reviewed-by: Adam Jackson <ajax@redhat.com> 31972 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31973 31974commit 1ea8fb96bc6a95d98dd67fc9d4c4e4b858383d34 31975Author: Alan Coopersmith <alan.coopersmith@oracle.com> 31976Date: Tue Sep 22 19:25:39 2015 -0700 31977 31978 protocol.txt: Remove extensions that died before Xorg was born 31979 31980 Adobe-DPS-Extension, DEC-XTRAP, X3D-PEX, XIE 31981 31982 Reviewed-by: Adam Jackson <ajax@redhat.com> 31983 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 31984 31985commit 21f384b7b8b571151805674c9d384e2ad7f8b7ea 31986Author: Marek Chalupa <mchqwerty@gmail.com> 31987Date: Thu May 21 15:43:43 2015 +0200 31988 31989 xwayland: do not add output into output_list multiple times 31990 31991 output.done event can be sent even on some property change, not only 31992 when announcing the output. Therefore we must check if we already have it 31993 otherwise we may corrupt the list by adding it multiple times. 31994 31995 This fixes bug when xwayland looped indefinitely in output.done handler 31996 and that can be reproduced following these steps (under X without 31997 multi-monitor setup): 31998 1) run weston --output-count=2 31999 2) run xterm, move it so that half is on one output 32000 and half on the other 32001 3) close second output, try run weston-terminal 32002 32003 weston sends updated outputs which trigger this bug. 32004 32005 v2. factor out common code into function 32006 move expecting_events into right branch 32007 32008 Signed-off-by: Marek Chalupa <mchqwerty@gmail.com> 32009 32010commit 95014ad2a7815282efdecc895bf14742b4b3ccb3 32011Author: Dima Ryazanov <dima@gmail.com> 32012Date: Tue May 12 10:21:18 2015 -0700 32013 32014 xwayland: Remove the output from the list after destroying it 32015 32016 [Marek Chalupa]: rebased to master 32017 32018 Signed-off-by: Dima Ryazanov <dima@gmail.com> 32019 32020commit 550984c95ef0bee1886dbb54e35edc0852772e2f 32021Author: Dima Ryazanov <dima@gmail.com> 32022Date: Fri May 15 22:38:28 2015 -0700 32023 32024 xwayland: Destroy xwl_output when wl_output gets removed 32025 32026 This makes Xwayland correctly handle a monitor getting unplugged. 32027 32028 [Marek]: use xorg_list_for_each_entry_safe 32029 32030 Signed-off-by: Dima Ryazanov <dima@gmail.com> 32031 Signed-off-by: Marek Chalupa <mchqwerty@gmail.com> 32032 32033commit 6ca496b7c3ccfd677c8c1bee88cc509a5e3c9e04 32034Author: Alan Coopersmith <alan.coopersmith@oracle.com> 32035Date: Mon Sep 21 22:36:41 2015 -0700 32036 32037 dri2: better checks for integer overflow in GetBuffers* 32038 32039 Check for integer overflow before using stuff->count in a multiplication, 32040 to avoid compiler optimizing out due to undefined behaviour, but only 32041 after we've checked to make sure stuff->count is in the range of the 32042 request we're parsing. 32043 32044 Reported-by: jes@posteo.de 32045 Reviewed-by: Adam Jackson <ajax@redhat.com> 32046 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32047 32048commit 6c2c6fb5a7b44f50811722eb8621afb272aff2e8 32049Author: Keith Packard <keithp@keithp.com> 32050Date: Mon Sep 21 07:16:14 2015 +0100 32051 32052 Xext, os: Remove OS-internal usages within XACE and XSELinux 32053 32054 These extensions were accessing internal OS functions and 32055 structures. Expose the necessary functionality to them and remove 32056 their use of osdep.h 32057 32058 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32059 Signed-off-by: Keith Packard <keithp@keithp.com> 32060 32061commit 9af111fe045d82192bacc722940e1d92a8ba4cbd 32062Author: Keith Packard <keithp@keithp.com> 32063Date: Mon Sep 21 07:16:13 2015 +0100 32064 32065 Xext/xselinux: Warning fixes 32066 32067 There's a 'const char *' adventure here that I'm mostly ignoring; some 32068 client information gets const poisoned. Worked around by adding a 32069 couple of casts. Ick. 32070 32071 Added an _X_ATTRIBUTE_PRINTF to SELinuxLog. 32072 32073 Ignore a couple of unused return values. 32074 32075 Reviewed-by: Adam Jackson <ajax@redhat.com> 32076 Signed-off-by: Keith Packard <keithp@keithp.com> 32077 32078commit 9ee05c8987245350976ebc0df4953c60da6bcc70 32079Author: Jon TURNEY <jon.turney@dronecode.org.uk> 32080Date: Thu Aug 13 13:11:31 2015 +0100 32081 32082 debug output format fix in RRMonitorCrtcName() 32083 32084 xorg/xserver/randr/rrmonitor.c:35:5: error: format ‘%x’ expects argument of type ‘unsigned int’, but argument 3 has type ‘RRCrtc’ [-Werror=format=] 32085 32086 RRCrtc is XID is CARD32, which inside the server is unsigned long or int 32087 depending on architecture, so a cast is required. 32088 32089 Reviewed-by: Adam Jackson <ajax@redhat.com> 32090 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 32091 32092commit fe253296035277f7a3adfad6eb239fee41714d22 32093Author: Jon TURNEY <jon.turney@dronecode.org.uk> 32094Date: Thu Aug 13 13:11:37 2015 +0100 32095 32096 debug output format fix in xf86Helper.c 32097 32098 xserver/hw/xfree86/common/xf86Helper.c:1834:12: error: format ‘%ld’ expects argument of type ‘long int’, but argument 3 has type ‘Atom’ [-Werror=format=] 32099 xserver/hw/xfree86/common/xf86Helper.c:1834:12: error: format ‘%ld’ expects argument of type ‘long int’, but argument 4 has type ‘Atom’ [-Werror=format=] 32100 32101 Atom is unfortunately unsigned long or unsigned int depending on the 32102 architecture, so a cast is required. 32103 32104 Reviewed-by: Adam Jackson <ajax@redhat.com> 32105 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 32106 32107commit dc2998bf060957b1fdd6bd2ea4f76c5154233a1b 32108Author: Dima Ryazanov <dima@gmail.com> 32109Date: Tue May 12 10:21:19 2015 -0700 32110 32111 xwayland: Remove a useless out-of-memory check 32112 32113 snprintf does not allocate memory, so we can never get an out-of-memory 32114 error. 32115 32116 (Also, the error handler would free xwl_output after it was already 32117 registered as an event listener.) 32118 32119 Signed-off-by: Dima Ryazanov <dima@gmail.com> 32120 Reviewed-by: Marek Chalupa <mchqwerty@gmail.com> 32121 32122commit b469fc72d2dd0cf60760fa0828ed73771c2a0512 32123Author: Adam Jackson <ajax@redhat.com> 32124Date: Wed Feb 25 16:03:13 2015 -0500 32125 32126 composite: Factor out backing store transition 32127 32128 No functional change. 32129 32130 Reviewed-by: Keith Packard <keithp@keithp.com> 32131 Signed-off-by: Adam Jackson <ajax@redhat.com> 32132 32133commit 6da3f5d04f6a1cda0c858280f9561f9fbc323275 32134Author: Dave Airlie <airlied@redhat.com> 32135Date: Wed Sep 23 11:19:00 2015 +1000 32136 32137 glx: fix regression with copy sub buffer disappearing 32138 32139 So copy sub buffer isn't a core extensions it's a driver extension 32140 which means we are using totally the wrong interface to query for it 32141 here, which means bad things happen when you roll out this code, 32142 for instance MESA_copy_sub_buffer stops working. 32143 32144 This is just the hack I'm sticking in Fedora to avoid the regression 32145 for now, but hopefully will inspire us. 32146 32147 Reviewed-by: Adam Jackson <ajax@redhat.com> 32148 Signed-off-by: Dave Airlie <airlied@redhat.com> 32149 32150commit 5e9457c41c05b65da7e39bd8944ec761479f19dd 32151Author: Eric Anholt <eric@anholt.net> 32152Date: Mon May 5 11:10:07 2014 -0700 32153 32154 fbdevhw: Fix a const qualifier warning 32155 32156 We're passing string literals to these functions. 32157 32158 Reviewed-by: Adam Jackson <ajax@redhat.com> 32159 Signed-off-by: Eric Anholt <eric@anholt.net> 32160 32161commit 88355cedfc9b3940b9fe1599fbf6e0272c57f8a9 32162Author: Eric Anholt <eric@anholt.net> 32163Date: Mon May 5 11:10:08 2014 -0700 32164 32165 kdrive: Remove dead debug arrays. 32166 32167 These have never been used in the history of the tree, and were 32168 producing string literal const loss warnings. 32169 32170 Reviewed-by: Adam Jackson <ajax@redhat.com> 32171 Signed-off-by: Eric Anholt <eric@anholt.net> 32172 32173commit e6f8a0340b72113474ccfbf1ea447558b705995a 32174Author: Eric Anholt <eric@anholt.net> 32175Date: Mon May 5 11:10:06 2014 -0700 32176 32177 x86emu: Fix some set-but-not-used warnings. 32178 32179 Reviewed-by: Adam Jackson <ajax@redhat.com> 32180 Signed-off-by: Eric Anholt <eric@anholt.net> 32181 32182commit fca98ce5e7d6424265c35a1219c2240d3342f627 32183Author: Egbert Eich <eich@freedesktop.org> 32184Date: Fri Feb 7 16:21:48 2014 +0100 32185 32186 libxf86config: libxf86config_internal -> libxf86config 32187 32188 Now since the installable libxf86config is gone, rename 32189 libxf86config_internal to libxf86config. 32190 32191 Signed-off-by: Egbert Eich <eich@freedesktop.org> 32192 Reviewed-by: Keith Packard <keithp@keithp.com> 32193 32194commit dabf7c3590b427c185089128ecfd967f188f2658 32195Author: Egbert Eich <eich@freedesktop.org> 32196Date: Fri Feb 7 16:21:47 2014 +0100 32197 32198 libxf86config: Remove 32199 32200 The library used by the Xserver to read and parse the configuration file 32201 could be built so that it culd be installed as a separate lib and used 32202 by external programs. 32203 32204 Apparently there has not been any interest in this for quite a while as 32205 this library has been broken for a long time now in the sense that it 32206 was calling functions provided by the Xserver which were not implemented 32207 for the external library. 32208 32209 Since this library is useless as it is anyway when built let's drop 32210 support for it. 32211 32212 Signed-off-by: Egbert Eich <eich@freedesktop.org> 32213 Reviewed-by: Keith Packard <keithp@keithp.com> 32214 32215commit c7f84c880bbbb508016ad80390a11961b1c4214d 32216Author: Chris Wilson <chris@chris-wilson.co.uk> 32217Date: Mon Jan 27 09:44:48 2014 +0000 32218 32219 render: Fix leak of filter params 32220 32221 ==11097== 2,048 (+1,640) bytes in 32 (+26) blocks are definitely lost in loss record 1,570 of 1,719 32222 ==11097== at 0x4C2A2DB: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 32223 ==11097== by 0x225EF3: SetPicturePictFilter (filter.c:339) 32224 ==11097== by 0x22DF4F: ProcRenderSetPictureFilter (render.c:1773) 32225 ==11097== by 0x15D25D: Dispatch (dispatch.c:432) 32226 ==11097== by 0x14C7B9: main (main.c:298) 32227 32228 [ajax: Fixed whitespace] 32229 32230 Reviewed-by: Adam Jackson <ajax@redhat.com> 32231 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 32232 32233commit 98e170971c6b06d64a5bc7418bf6ae6ae985a3ae 32234Author: Ville Syrjälä <ville.syrjala@linux.intel.com> 32235Date: Mon Oct 14 17:07:29 2013 +0300 32236 32237 xfree86: Fix the 1792x1344-75 EST III mode 32238 32239 The correct refresh rate for this mode is 75, not 85. 32240 32241 Reviewed-by: Adam Jackson <ajax@redhat.com> 32242 Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> 32243 32244commit 91c7acfc3447a0073dd1db260cf05115d94ce7e8 32245Author: Michal Srb <msrb@suse.com> 32246Date: Mon Oct 7 17:57:14 2013 +0300 32247 32248 randr: Allow RRSelectInput for ProviderChange and ResourceChange events 32249 32250 [ajax: To be clear, we already have code to emit these events, and it 32251 looks like it works, but to get them you'd have had to also ask for one 32252 of the other notify types. This makes it possible to listen for e.g. 32253 ProviderChange alone.] 32254 32255 Reviewed-by: Adam Jackson <ajax@redhat.com> 32256 Signed-off-by: Michal Srb <msrb@suse.com> 32257 32258commit 5437949a51fa6d9c9c83336d46ae57398f390ee3 32259Author: Egbert Eich <eich@freedesktop.org> 32260Date: Sat Oct 5 13:57:13 2013 +0200 32261 32262 fbdevhw: Disable FBIOBLANK ioctl if not supported 32263 32264 Some ioctls may not be supported by the kernel however their failure 32265 is non-fatal to the driver. Unfortunately we only know once we try 32266 to execute the ioctl however the sematics of the fbdev driver API 32267 doesn't allow upper layers to disable the call. 32268 Instead of changing the fbdevHW driver API just disable the call to 32269 this ioctl on the module level when detecting such a case. 32270 32271 Reviewed-by: Adam Jackson <ajax@redhat.com> 32272 Signed-off-by: Egbert Eich <eich@freedesktop.org> 32273 32274commit 49fe4ee7b7085b616028d6afe67767a60a8adf82 32275Author: Matt Turner <mattst88@gmail.com> 32276Date: Thu Feb 7 17:29:00 2013 -0800 32277 32278 compiler.h: Remove dead STANDALONE_MMIO 32279 32280 The only drivers I can find that used this are the r128 and radeon DRI 32281 drivers. r128 is dead and the radeon driver wasn't including Xorg's 32282 compiler.h and still worked. 32283 32284 Reviewed-by: Adam Jackson <ajax@redhat.com> 32285 Signed-off-by: Matt Turner <mattst88@gmail.com> 32286 32287commit d2a6962474978437545120ebc1fcfb909d901d55 32288Author: Andy Ritger <aritger@nvidia.com> 32289Date: Wed Jun 20 11:52:34 2012 -0700 32290 32291 mi: Log an error if mieqProcessInputEvents() recurses. 32292 32293 v2: 32294 32295 Uses BUG_WARN_MSG to also provide a stack trace. (Peter Hutterer) 32296 32297 Signed-off-by: Keith Packard <keithp@keithp.com> 32298 Signed-off-by: Andy Ritger <aritger@nvidia.com> 32299 32300commit cc57ab4fc8b9dc15531f0dd9f5b352fa99a80cd0 32301Author: Andy Ritger <aritger@nvidia.com> 32302Date: Wed Jun 20 11:52:33 2012 -0700 32303 32304 Remove static qualifier on mieqProcessInputEvents()'s 'event' 32305 32306 This isn't necessary. 32307 32308 Signed-off-by: Andy Ritger <aritger@nvidia.com> 32309 Signed-off-by: Keith Packard <keithp@keithp.com> 32310 32311commit 44ea291b78b39ed5fd03f9693ac9525b866e183b 32312Author: Tomas Carnecky <tomas.carnecky@gmail.com> 32313Date: Mon May 21 19:47:01 2012 +0000 32314 32315 Remove NO_LIBCWRAPPER define 32316 32317 I'll just refer to 1faba79 (Death to libcwrapper., 2007-12-03). 32318 32319 Reviewed-by: Adam Jackson <ajax@redhat.com> 32320 Signed-off-by: Tomas Carnecky <tomas.carnecky@gmail.com> 32321 32322commit 1e95a7df5f2e919ce59c5188d2ad1b30c3ac8595 32323Author: Peter Hutterer <peter.hutterer@who-t.net> 32324Date: Thu Mar 15 13:46:56 2012 +1000 32325 32326 dix: put warning in for scroll increments of 0 32327 32328 If the increment is 0 but this is a scroll axis, it's definitely a bug. 32329 Nonetheless, it has happened, so put a warning in and a return statement 32330 that we avoid the infinite loop and hopefully be able to reproduce later. 32331 32332 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 32333 Reviewed-by: Keith Packard <keithp@keithp.com> 32334 32335commit 63efa2e7e91349dfecc849d30c5de84b3c911fe9 32336Author: Bryce Harrington <bryce@canonical.com> 32337Date: Wed Nov 30 12:54:29 2011 -0800 32338 32339 test: Fix typo and drop documentation references to glib. 32340 32341 The X server doesn't use glib for the tests any more. 32342 32343 Signed-off-by: Bryce Harrington <bryce@canonical.com> 32344 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 32345 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 32346 32347commit 4c7cefe24001279a558dd875949d40ede49ef702 32348Author: Dave Airlie <airlied@redhat.com> 32349Date: Thu Oct 20 11:44:28 2011 +0100 32350 32351 exa: initialise mask_off_x and mask_off_y 32352 32353 These get used at the end of the function in a calculation, 32354 even though the result isn't used its not pretty. 32355 32356 Pointed out by coverity. 32357 32358 Signed-off-by: Dave Airlie <airlied@redhat.com> 32359 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32360 32361commit 668cce3f5a175109442f0ef9b50354db95e66b44 32362Author: Gaetan Nadon <memsize@videotron.ca> 32363Date: Fri Sep 23 14:11:17 2011 -0400 32364 32365 dmx: fix typo in the title 32366 32367 [ajax: fix typo in the commit message] 32368 32369 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 32370 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32371 32372commit 7ed0c3912e1cbab960df308b9d4ea618be757fd2 32373Author: Andrea Canciani <ranma42@gmail.com> 32374Date: Wed Jul 27 10:52:31 2011 +0200 32375 32376 render: Allow single-stop gradients 32377 32378 The Render specification allows single-stop gradients and pixman 0.22 32379 (X server requires >= 0.27.2 now) can rasterize them correctly. 32380 32381 [ajax: update commit message] 32382 32383 Signed-off-by: Andrea Canciani <ranma42@gmail.com> 32384 Reviewed-by: Soren Sandmann <ssp@redhat.com> 32385 32386commit 3d9f578e6f0d13ecada7dc677fe61a4dd3d05e18 32387Author: Cyril Brulebois <kibi@debian.org> 32388Date: Mon Jun 6 02:25:11 2011 +0200 32389 32390 doc: Convert an \fP pair into appropriate tags. 32391 32392 Reviewed-by: Adam Jackson <ajax@redhat.com> 32393 Signed-off-by: Cyril Brulebois <kibi@debian.org> 32394 32395commit d97ffef1ab95546634a74618ed2661cfd6d980b1 32396Author: Cyril Brulebois <kibi@debian.org> 32397Date: Mon Jun 6 02:25:10 2011 +0200 32398 32399 doc: Try to avoid repetitions. 32400 32401 Reviewed-by: Adam Jackson <ajax@redhat.com> 32402 Signed-off-by: Cyril Brulebois <kibi@debian.org> 32403 32404commit 2cce6ffc350d3840337a4d045d5cf14bb7bf2b32 32405Author: Cyril Brulebois <kibi@debian.org> 32406Date: Mon Jun 6 02:25:08 2011 +0200 32407 32408 doc: Fix punctuation. 32409 32410 [ajax: s/ponct/punct/] 32411 32412 Signed-off-by: Cyril Brulebois <kibi@debian.org> 32413 Reviewed-by: Jamey Sharp <jamey@minilop.net> 32414 32415commit e7e3320098cad585d46f74d8a0c6946a080b0263 32416Author: Cyril Brulebois <kibi@debian.org> 32417Date: Mon Jun 6 02:25:07 2011 +0200 32418 32419 doc: Fix extraneous > 32420 32421 It looks like it serves no special purpose. 32422 32423 Reviewed-by: Jamey Sharp <jamey@minilop.net> 32424 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32425 Signed-off-by: Cyril Brulebois <kibi@debian.org> 32426 32427commit 9b4f10d7490f6610a65d401958296c95ce891c8b 32428Author: Cyril Brulebois <kibi@debian.org> 32429Date: Mon Jun 6 02:25:06 2011 +0200 32430 32431 doc: Fix grammar and typos. 32432 32433 Reviewed-by: Jamey Sharp <jamey@minilop.net> 32434 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32435 Signed-off-by: Cyril Brulebois <kibi@debian.org> 32436 32437commit 5b92a0a84a067b653cc457d8ec0fe25f36a9d207 32438Author: Cyril Brulebois <kibi@debian.org> 32439Date: Mon Jun 6 02:25:05 2011 +0200 32440 32441 doc: Fix alignment. 32442 32443 Reviewed-by: Jamey Sharp <jamey@minilop.net> 32444 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32445 Signed-off-by: Cyril Brulebois <kibi@debian.org> 32446 32447commit 0b8fca667b45f00249a3dd9ee79d2a77a4177b24 32448Author: Cyril Brulebois <kibi@debian.org> 32449Date: Mon Jun 6 02:24:59 2011 +0200 32450 32451 doc: Fix typo. 32452 32453 Signed-off-by: Cyril Brulebois <kibi@debian.org> 32454 Reviewed-by: Jamey Sharp <jamey@minilop.net> 32455 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32456 32457commit 1226f7ddd78da5dc8b6f6442600ac2deb3986e8d 32458Author: Cyril Brulebois <kibi@debian.org> 32459Date: Mon Jun 6 02:25:01 2011 +0200 32460 32461 doc: Get rid of RCS tags. 32462 32463 Reviewed-by: Adam Jackson <ajax@redhat.com> 32464 Signed-off-by: Cyril Brulebois <kibi@debian.org> 32465 32466commit 4803f7139131cee7bbf50b32d667d6e60302df09 32467Author: Cyril Brulebois <kibi@debian.org> 32468Date: Sun Jun 5 23:32:26 2011 +0200 32469 32470 man: Fix case for MIT-unspecified. 32471 32472 include/site.h says that COMPILEDDISPLAYCLASS is MIT-unspecified, rather 32473 than MIT-Unspecified. Fix the manpage accordingly. 32474 32475 Signed-off-by: Cyril Brulebois <kibi@debian.org> 32476 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32477 32478commit 593270038d8983b99aa0aaa24509da3cb944be5c 32479Author: Tiago Vignatti <tiago.vignatti@nokia.com> 32480Date: Tue Sep 7 15:50:26 2010 +0300 32481 32482 dix: remove dead extensions to protocol.txt 32483 32484 lbx and mbe. 32485 32486 Reviewed-by: Adam Jackson <ajax@redhat.com> 32487 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 32488 32489commit 3cd56dd2cfbfcce6ba68c22450d51139208993ab 32490Author: Robert Hooker <sarvatt@ubuntu.com> 32491Date: Fri Dec 3 23:42:27 2010 -0500 32492 32493 Add new GLX events errors and requests to protocol.txt 32494 32495 Based off of glproto 1.4.11. 32496 32497 Reviewed-by: Adam Jackson <ajax@redhat.com> 32498 Signed-off-by: Robert Hooker <robert.hooker@canonical.com> 32499 32500commit 272ba9f021c6d08ebd6a2d96da3737f5630bcbd5 32501Author: Robert Hooker <sarvatt@ubuntu.com> 32502Date: Fri Dec 3 23:39:51 2010 -0500 32503 32504 Add DRI2 2.2/2.3 events and requests to protocol.txt 32505 32506 Reviewed-by: Adam Jackson <ajax@redhat.com> 32507 Signed-off-by: Robert Hooker <robert.hooker@canonical.com> 32508 32509commit 64e6124f27e9900809d0009b95c7c0b5b8788e6b 32510Author: Emil Velikov <emil.l.velikov@gmail.com> 32511Date: Tue Sep 22 15:35:40 2015 +0100 32512 32513 glamor: move GL_OES_EGL_image check next to EGL_EXT_image_dma_buf_import 32514 32515 We're using the former only as the latter is present. Thus in some cases 32516 we might incorrectly error out if it's missing. 32517 32518 Namely - glamor_glx, glamor_egl without gbm, EGL_KHR_gl_texture_2D_image 32519 or EGL_EXT_image_dma_buf_import. 32520 32521 Fixes 58d54ee82df(glamor: explicitly check for GL_OES_EGL_image) 32522 Cc: Eric Anholt <eric@anholt.net> 32523 Reviewed-by: Adam Jackson <ajax@redhat.com> 32524 Suggested-by: Eric Anholt <eric@anholt.net> 32525 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 32526 32527commit aa5606be1c0ea72b38de48ca333c7da5436368ef 32528Author: Alan Coopersmith <alan.coopersmith@oracle.com> 32529Date: Fri May 15 10:25:46 2015 -0700 32530 32531 Make configure test for LD_NO_UNDEFINED_FLAG on Solaris work w/autoconf 2.69 32532 32533 After upgrading from autoconf 2.68 to 2.69, this test started failing with 32534 "conftest.c", line 149: undefined symbol: NULL 32535 so use a raw 0 pointer to avoid header dependencies in the autoconf 32536 generated test case. 32537 32538 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32539 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 32540 32541commit fa0bb01899d586e02cf51892c17edd22e462a46c 32542Author: Adam Jackson <ajax@redhat.com> 32543Date: Mon Sep 21 15:23:23 2015 -0400 32544 32545 rootless: Fix bogus handling of broken root clip 32546 32547 gcc quite correctly complains about this: 32548 32549 In file included from ../../include/scrnintstr.h:51:0, 32550 from rootlessValTree.c:98: 32551 In function 'RegionUninit.isra.1', 32552 inlined from 'RegionEmpty' at ../../include/regionstr.h:194:5, 32553 inlined from 'RootlessMiValidateTree' at rootlessValTree.c:490:9: 32554 ../../include/regionstr.h:166:9: warning: attempt to free a non-heap object 'RegionBrokenData' [-Wfree-nonheap-object] 32555 free((_pReg)->data); 32556 32557 So that'd crash if you ever got there. RegionNull will do almost the 32558 same thing only without the free(), so let's do that instead; it might 32559 still not be an entirely sane way to recover, but it at least won't 32560 crash. 32561 32562 Reviewed-by: Keith Packard <keithp@keithp.com> 32563 Signed-off-by: Adam Jackson <ajax@redhat.com> 32564 32565commit b25f7d0c165c4d32c73fbdf9080012503d319ba1 32566Merge: db1089eaf 1ba4fde10 32567Author: Adam Jackson <ajax@redhat.com> 32568Date: Mon Sep 21 17:34:48 2015 -0400 32569 32570 Merge commit '1ba4fde1015fc6da19dfbfdf0f77f1071416e215' 32571 32572 This is ofourdan/xwayland, minus: 32573 32574 xwayland: Always update the wl_pointer cursor on pointer focus 32575 32576 which was rejected for ABI and design issues. 32577 32578commit db1089eafc1c5371fa0030202de588d2e2b4f8e5 32579Author: Keith Packard <keithp@keithp.com> 32580Date: Mon Sep 21 07:16:17 2015 +0100 32581 32582 os/xdmcp: Just send XDMCP keepalive packets once every three minutes 32583 32584 There was a complicated scheme to increase the time between keepalives 32585 from 3 minutes up to as much as 24 hours in an attempt to reduce 32586 network traffic from idle X terminals. X terminals receiving X 32587 traffic, or receiving user input would use the 3 minute value; X 32588 terminals without any network traffic would use a longer value. 32589 32590 However, this was actually broken -- any activity in the X server, 32591 either client requests or user input, would end up resetting the 32592 keepalive timeout, so a user mashing on the keyboard would never 32593 discover that the XDMCP master had disappeared and have the session 32594 terminated, which was precisely the design goal of the XDMCP keepalive 32595 mechanism. 32596 32597 Instead of attempting to fix this, accept the cost of a pair of XDMCP 32598 packets once every three minutes and just perform keepalives 32599 regularly. 32600 32601 This will also make reworking the block and wakeup handler APIs to 32602 eliminate select masks easier. 32603 32604 Reviewed-by: Adam Jackson <ajax@redhat.com> 32605 Signed-off-by: Keith Packard <keithp@keithp.com> 32606 32607commit a3a40291330bad10401fe2bcdbc097ce742b026a 32608Author: Keith Packard <keithp@keithp.com> 32609Date: Mon Sep 21 07:16:16 2015 +0100 32610 32611 os/xdmcp: Remove dead 'restart' code 32612 32613 The X server used to wait for the user to hit a key or move the mouse 32614 before restarting the session after a keepalive failure. This, 32615 presumably, was to avoid having the X server continuously spew XDMCP 32616 protocol on the network while the XDM server was dead. 32617 32618 Switching into this state was removed from the server some time before 32619 XFree86 4.3.99.16, so the remaining bits of code have been dead for 32620 over a decade, and no-one ever noticed. 32621 32622 Reviewed-by: Adam Jackson <ajax@redhat.com> 32623 Signed-off-by: Keith Packard <keithp@keithp.com> 32624 32625commit 3cd7d33380953f2a315dd79ab74d7340c83fd80a 32626Author: Egbert Eich <eich@freedesktop.org> 32627Date: Tue Mar 31 09:06:46 2015 +0200 32628 32629 randr: Remove senseless checks for xf86RandR12Key 32630 32631 When xf86RandR12Key is not set we will not get to the places where 32632 these tests are done as the functions in question are not called. 32633 In most cases we would have crashed before these checks anyway. 32634 32635 Signed-off-by: Egbert Eich <eich@freedesktop.org> 32636 Reviewed-by: Keith Packard <keithp@keithp.com> 32637 32638commit 245040f0d0b937efe14d947468a641a95398776a 32639Author: Chris Wilson <chris@chris-wilson.co.uk> 32640Date: Mon Oct 6 10:02:22 2014 +0100 32641 32642 randr/prime: Don't stop on the first pipe when disabling ReplaceScanoutPixmap 32643 32644 As we define sizeFits based on whether a CRTC is active, and skip trying 32645 to redirect the scanout on a disable pipe, we then attempt to undo it 32646 later and fail because crtc->scanout_pixmap != DRI2_Pixmap and 32647 !sizeFits. Paper over this failure by skipping unredirected CRTC when 32648 disabling. 32649 32650 v2: Unwind upon failure 32651 32652 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=84653 32653 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 32654 Cc: Dave Airlie <airlied@redhat.com> 32655 Reported-by: Christoph Haag <haagch@frickel.club> 32656 Tested-by: Christoph Haag <haagch@frickel.club> 32657 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 32658 32659commit 9d65a0de95a20632f2a6a04ba596dcdd76d9f5e2 32660Author: Alan Coopersmith <alan.coopersmith@oracle.com> 32661Date: Fri Sep 18 08:44:07 2015 -0700 32662 32663 Remove DECnet from Xserver.man 32664 32665 DECnet support died in modularization (X11R7.0) 32666 32667 Reviewed-by: Adam Jackson <ajax@redhat.com> 32668 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32669 32670commit 82eb490b0a341729f25d0c42a937fe29d99da521 32671Author: Aaron Plattner <aplattner@nvidia.com> 32672Date: Thu Sep 17 16:04:33 2015 -0700 32673 32674 privates: Clear screen-specific keys during CloseScreen 32675 32676 The modesetting driver corrupts memory when used after a server regeneration 32677 because not enough memory is allocated for its pixmap privates. This happens 32678 because its call to dixRegisterScreenSpecificPrivateKey() does nothing because 32679 key->initialized is still TRUE from the first server generation. However, the 32680 key is not in the screen's linked list of screen-specific privates because 32681 that's freed and reallocated during the server generation loop in dix_main(). 32682 32683 Fix this by clearing key->initialized before CloseScreen and add a call to 32684 dixFreeScreenSpecificPrivates() for GPU screens. 32685 32686 v2: Just set key->initialized to FALSE and move dixFreeScreenSpecificPrivates() 32687 calls to after CloseScreen. 32688 32689 v3: Move dixFreeScreenSpecificPrivates() calls back to just before CloseScreen. 32690 32691 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 32692 Reviewed-by: Keith Packard <keithp@keithp.com> 32693 32694commit 18a93da9b1f429bc54a8b236fcdd38dab58375cd 32695Author: Jasper St. Pierre <jstpierre@mecheye.net> 32696Date: Wed Feb 18 12:41:25 2015 -0800 32697 32698 test/xi1: Update .gitignore 32699 32700 Reviewed-by: Adam Jackson <ajax@redhat.com> 32701 32702commit cdd1d58159140759052ec043bc9828d94ba97bc7 32703Author: Jon TURNEY <jon.turney@dronecode.org.uk> 32704Date: Mon Jul 6 17:15:44 2015 +0100 32705 32706 mingw: Fix NO_LOCAL_CLIENT_CRED build 32707 32708 Commit 4b4b9086 "os: support new implicit local user access mode [CVE-2015-3164 32709 2/3]" carefully places the relevant code it adds under !NO_LOCAL_CLIENT_CRED, 32710 but unfortunately doesn't notice that NO_LOCAL_CLIENT_CRED is defined as a 32711 side-effect in the middle of GetLocalClientCreds(), so many of these checks 32712 precede its definition. 32713 32714 Move the check if NO_LOCAL_CLIENT_CRED should be defined to configure.ac, so it 32715 always occurs before it's first use. 32716 32717 v2: 32718 Move check to configure.ac 32719 32720 v3: 32721 Use AC_CACHE_CHECK and name cache varaible appropriately 32722 32723 [ajax: Massaged commit message] 32724 32725 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 32726 Reviewed-by: Ray Strode <rstrode@redhat.com> 32727 32728commit f8a2c087c89ed5b638a2e62987f255d3f04f914e 32729Author: Javier Jardón <jjardon@gnome.org> 32730Date: Sun Mar 22 03:52:03 2015 +0000 32731 32732 configure.ac: Add support for new versions of systemd 32733 32734 ... that only provide a unique libsystemd.pc file 32735 32736 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 32737 Signed-off-by: Javier Jardón <jjardon@gnome.org> 32738 32739commit 67dd34fa734085ed67ba7790820fe4743d48b60f 32740Author: Adam Jackson <ajax@redhat.com> 32741Date: Wed Jun 10 13:20:59 2015 -0400 32742 32743 render: Simplify SProcRenderScale 32744 32745 Since ProcRenderScale throws BadImplementation anyway it's pointless to 32746 waste time carefully swapping the request. 32747 32748 Signed-off-by: Adam Jackson <ajax@redhat.com> 32749 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32750 32751commit a579e6ba7763749d51d16e3a54941e06939df07a 32752Author: Ian Scott <ian.scott@arteris.com> 32753Date: Thu May 21 16:13:12 2015 -0700 32754 32755 Xephyr: Paint with subimage for non-Glamor & non-XSHM case 32756 32757 This improves the case for when we paint an area without SHM. 32758 xcb_image_subimage() is used to create a subimage for the damaged area, which 32759 is converted to native format if necessary. 32760 32761 Signed-off-by: Ian Scott <ian.scott@arteris.com> 32762 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 32763 32764commit a6c0564f7f00aa66fd19180f75b6b09904444f06 32765Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 32766Date: Mon Mar 23 10:58:03 2015 -0300 32767 32768 ephyr: move host_has_extension() implementation to hostx.c 32769 32770 This is a trivial patch that moves host_has_extension() implementation 32771 from ephyr.c to hostx.c so that it can be called by hostx.c internal 32772 functions. Also rename function to hostx_has_extension() for consistency. 32773 32774 Reviewed-by: Adam Jackson <ajax@redhat.com> 32775 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 32776 32777commit 9414ba3683231afe9bee0f941c53998dea218615 32778Author: Jürg Billeter <j@bitron.ch> 32779Date: Sun Feb 8 11:34:18 2015 +0100 32780 32781 ephyr: fix -no-host-grab argument processing 32782 32783 Signed-off-by: Jürg Billeter <j@bitron.ch> 32784 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 32785 32786commit 1f915e8b524dd02011158aa038935970684c7630 32787Author: Daniel Drake <drake@endlessm.com> 32788Date: Wed May 20 13:16:12 2015 -0600 32789 32790 Keep SIGALRM restart flag after Popen 32791 32792 Commit 94ab7455 added SA_RESTART to the SIGALRM handler. However, the 32793 Popen code tears down and recreates the SIGALRM handler via OsSignal(), 32794 and this flag is dropped at this time. 32795 32796 Clean the code to use just a single codepath for creating this signal 32797 handler, always applying SA_RESTART. 32798 32799 [ajax: Fixed commit id] 32800 32801 Reviewed-by: Adam Jackson <ajax@redhat.com> 32802 Signed-off-by: Daniel Drake <drake@endlessm.com> 32803 32804commit 94ab7455abc213fc96760e29ab2e943ec682fb22 32805Author: Daniel Drake <drake@endlessm.com> 32806Date: Tue May 12 16:39:22 2015 -0600 32807 32808 Allow system call restarts upon signal interruption 32809 32810 The X server frequently deals with SIGIO and SIGALRM interruptions. 32811 If process execution is inside certain blocking system calls 32812 when these signals arrive, e.g. with the kernel blocked on 32813 a contended semaphore, the system calls will be interrupted. 32814 32815 Some system calls are automatically restartable (the kernel re-executes 32816 them with the same parameters once the signal handler returns) but 32817 only if the signal handler allows it. 32818 32819 Set SA_RESTART on the signal handlers to enable this convenient 32820 behaviour. 32821 32822 Reviewed-by: Adam Jackson <ajax@redhat.com> 32823 Signed-off-by: Daniel Drake <drake@endlessm.com> 32824 32825commit 1ba4fde1015fc6da19dfbfdf0f77f1071416e215 32826Author: Boyan Ding <stu_dby@126.com> 32827Date: Mon Aug 4 21:16:34 2014 +0800 32828 32829 xwayland: Activate and enable device on first capability reporting 32830 32831 Commit 2172714c changed behavior of capability handling, but it only 32832 solved part of the problem. If Xwayland is launched without a capability 32833 (e.g. no pointer device is connected when Xwayland was spinned up), and 32834 later that capability comes, the device added will not be automatically 32835 initialized. This patch initializes the device when the capability is 32836 reported for the first time, thus avoiding the problem. 32837 32838 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=81819 32839 Signed-off-by: Boyan Ding <stu_dby@126.com> 32840 Reviewed-by: Daniel Stone <daniels@collabora.com> 32841 32842commit f1995de1c43733a9e586605674c4071357a58e78 32843Author: Peter Hutterer <peter.hutterer@who-t.net> 32844Date: Fri Aug 28 14:28:11 2015 +1000 32845 32846 xwayland: call through to miPointerWarpCursor on warp_cursor 32847 32848 This is (eventually) called during 32849 InitializeSprite() → *pScreen->SetCursorPosition → miPointerSetCursorPosition 32850 when a device is set to floating. We don't do anything special outselves, but 32851 we need to pass on to the next layer to make sure the device is initialized 32852 properly. Otherwise, pScreen stays NULL and eventually crashes the server when 32853 we try to clean up behind us. 32854 32855 Test case: grab a device → floats it, ungrab again → crash 32856 32857 Reported-by: Jason Gerecke <killertofu@gmail.com> 32858 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 32859 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 32860 32861commit 373599ab008f0ecf8f3fb62455a5474f5ec8f499 32862Author: Peter Hutterer <peter.hutterer@who-t.net> 32863Date: Fri Aug 28 14:28:10 2015 +1000 32864 32865 xwayland: fix crash on enter/leave for a grabbed slave device 32866 32867 When grabbed, the slave device is floating, i.e. the master device is NULL. 32868 CheckMotion() isn't happy with NULL. Make sure we pass the right device in, 32869 either the master device when the device is attached, or the device itself 32870 when it is floating. 32871 32872 Reported-by: Jason Gerecke <killertofu@gmail.com> 32873 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 32874 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 32875 32876commit f1ba8858d5baa63cde6361492e6d349ce1b38d21 32877Author: Peter Hutterer <peter.hutterer@who-t.net> 32878Date: Fri Aug 28 14:28:09 2015 +1000 32879 32880 xwayland: use the device pointer we already have 32881 32882 No functional changes. 32883 32884 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 32885 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 32886 32887commit 44f250a7e8eb86d3f41da5d186d839e7a14648a3 32888Author: Carlos Garnacho <carlosg@gnome.org> 32889Date: Wed May 27 18:42:00 2015 +0200 32890 32891 xwayland: Remove related touchpoints when unrealizing windows 32892 32893 These sequences are forgotten to all purposes. 32894 32895 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 32896 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 32897 32898commit b7cd48f71d269101e28ae6855b895057378715f9 32899Author: Carlos Garnacho <carlosg@gnome.org> 32900Date: Wed May 27 18:41:59 2015 +0200 32901 32902 xwayland: Implement the wl_touch interface 32903 32904 A DeviceIntPtr with touch valuators is also created in order to deliver 32905 the translated touch events. The lifetime of xwl_touch structs is tied 32906 to the wayland ones, finishing in either wl_touch.up() or wl_touch.cancel() 32907 32908 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 32909 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 32910 32911commit d96eccc057094be31c1ee4e3b3246675934cb19d 32912Author: Carlos Garnacho <carlosg@gnome.org> 32913Date: Wed May 27 18:41:58 2015 +0200 32914 32915 xwayland: Add xwl_touch struct 32916 32917 This struct holds information about each individual, ongoing touchpoint. 32918 A list of these is held by the xwl_seat. 32919 32920 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 32921 Reviewed-by: Olivier Fourdan <ofourdan@redhat.com> 32922 32923commit c1565f3ca663aa4ba4316d7248737ab817116fbd 32924Author: Jasper St. Pierre <jstpierre@mecheye.net> 32925Date: Sat Feb 21 11:53:59 2015 -0800 32926 32927 xwayland-input: Remove our XYToWindow handler 32928 32929 This was built as a hack for simple Wayland compositors like Weston 32930 which were lazy and didn't want to configure windows server-side when 32931 moved. 32932 32933 Since comboboxes and menus are separate toplevel O-R windows, this hack 32934 breaks input as it needs to be traced normally, not simply sent to the 32935 focused window. 32936 32937 X11 toolkits really do need their windows to be configured correctly 32938 for their O-R windows comboboxes or menus other things, so let's fix 32939 the lazy compositors and remove this. 32940 32941 I have tested this patch with both Weston and Mutter and neither of 32942 them require any changes, and it fixes comboboxes and menus. 32943 32944 If somebody then wants to revert 73698d4, that's fine by me, so we 32945 reduce the amount of API that DDXen have. 32946 32947 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 32948 Reviewed-by: Daniel Stone <daniels@collabora.com> 32949 32950commit 58d54ee82dfae5486bc09d04d2760c922d54d631 32951Author: Emil Velikov <emil.l.velikov@gmail.com> 32952Date: Wed Sep 16 22:10:22 2015 +0100 32953 32954 glamor: explicitly check for GL_OES_EGL_image 32955 32956 Otherwise we'll fail miserably later on as we try to use 32957 glEGLImageTargetTexture2DOES. 32958 32959 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 32960 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 32961 32962commit 139e36dd5cbab80a9296129f3d25379dc01442b3 32963Author: Ilia Mirkin <imirkin@alum.mit.edu> 32964Date: Fri Sep 4 18:56:19 2015 -0400 32965 32966 modesetting: fix up output naming convention 32967 32968 A user on a nouveau-driven card ran into a problem where DVI-D-1 and 32969 DVI-I-1 were aliasing. The simplest fix is to provide the full connector 32970 names. While we're at it, rename the output names to match what is in 32971 the kernel, and start counting the connectors from 1 rather than 0. The 32972 only deviation is HDMI vs HDMI-A, which kept its original name. 32973 32974 This will break backwards compatibility with existing xorg.conf's that 32975 reference output names, but the alternative is to create a separate 32976 counting system, further disconnecting from the kernel names. 32977 32978 Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu> 32979 Reviewed-by: Dave Airlie <airlied@redhat.com> 32980 32981commit a907ead7864ac5720854fa3333b66fd5187d5eaf 32982Author: Marek Chalupa <mchqwerty@gmail.com> 32983Date: Thu Aug 20 07:43:51 2015 +0200 32984 32985 xwayland: implement repeat_info event 32986 32987 xwayland windows ignored any key repeating settings 32988 advertised by a compositor 32989 32990 v2. don't hardcode version 4 of seat 32991 use AutoRepeatModeOn/Off 32992 32993 v3. use min(version, 4) when binding seat 32994 32995 Reviewed-by: Adam Jackson <ajax@redhat.com> 32996 Signed-off-by: Marek Chalupa <mchqwerty@gmail.com> 32997 32998commit e8e3368298c98a1292bb1c727e2f3d56539e2eac 32999Author: Felix Janda <felix.janda@posteo.de> 33000Date: Mon Aug 17 20:25:53 2015 +0200 33001 33002 On linux use <termios.h> instead of <termio.h> 33003 33004 <termio.h> is obsolete. Using <termios.h> instead fixes building with 33005 musl libc. 33006 33007 Reviewed-by: Adam Jackson <ajax@redhat.com> 33008 Signed-off-by: Felix Janda <felix.janda@posteo.de> 33009 33010commit dfa8b37cfb2589f2474f6fcc755dde77bfdd6fed 33011Author: Adam Jackson <ajax@redhat.com> 33012Date: Wed Jul 29 11:30:15 2015 -0400 33013 33014 glxproxy: Fix an obvious thinko in QueryExtensionsString 33015 33016 gcc 5.1 throws a rather amusing warning here: 33017 33018 glxcmdsswap.c: In function ‘__glXSwapQueryExtensionsString’: 33019 glxcmdsswap.c:439:1: warning: function might be candidate for attribute 33020 ‘noreturn’ [-Wsuggest-attribute=noreturn] 33021 __glXSwapQueryExtensionsString(__GLXclientState * cl, GLbyte * pc) 33022 33023 This is entirely accurate, albeit curiously phrased: swapping some bits 33024 hanging off the end of a null pointer will make Xdmx crash, so from the 33025 perspective of code generation the function will indeed not return. 33026 33027 Reviewed-by: Keith Packard <keithp@keithp.com> 33028 Signed-off-by: Adam Jackson <ajax@redhat.com> 33029 33030commit 75ee874e432d959a70fbc2c9208d20dbfa5cbc5f 33031Author: Jon TURNEY <jon.turney@dronecode.org.uk> 33032Date: Fri May 1 17:06:47 2015 +0100 33033 33034 Build fix for DMX 33035 33036 ./../dix/.libs/libdix.a(pixmap.o): In function `PixmapStartDirtyTracking': 33037 xorg/xserver/dix/pixmap.c:194: undefined reference to `RRTransformCompute' 33038 33039 I'm not quite sure how 0db457b fixes this 33040 33041 Move RANDR_LIB to after DIX_LIB, so the reference to RRTransformCompute in 33042 DIX_LIB is satisfied. 33043 33044 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 33045 Tested-by: Olivier Fourdan <ofourdan@redhat.com> 33046 Signed-off-by: Keith Packard <keithp@keithp.com> 33047 33048commit 634e357be2ec400f63bd5c42f706b709b6ddabc3 33049Author: Ingo Schwarze <schwarze@usta.de> 33050Date: Sun Aug 30 15:26:40 2015 +0200 33051 33052 remove bogus \/ escapes 33053 33054 some X manuals use then escape sequence \/ when they want to render 33055 a slash. That's bad because \/ is not a slash but an italic 33056 correction, never producing any output, having no effect at all in 33057 terminal output, and only changing spacing in a minor way in typeset 33058 output. 33059 33060 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 33061 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 33062 Signed-off-by: Keith Packard <keithp@keithp.com> 33063 33064commit c74c074d8e4981eb4509e120e14d15387bdc94ef 33065Author: Olivier Fourdan <ofourdan@redhat.com> 33066Date: Mon Aug 31 08:18:59 2015 +0200 33067 33068 dix: export ResourceClientBits() 33069 33070 ResourceClientBits() might be used by the modules as well, need to 33071 mark the symbol as visible to that the linker can resolve it at 33072 run time. 33073 33074 /usr/bin/X: symbol lookup error: 33075 /usr/lib64/xorg/modules/extensions/libglx.so: undefined symbol: 33076 ResourceClientBits 33077 33078 bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91763 33079 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 33080 Reviewed-by: Keith Packard <keithp@keithp.com> 33081 Signed-off-by: Keith Packard <keithp@keithp.com> 33082 33083commit dbfed3a291619c2530f314394bd73d71e2567e78 33084Author: Keith Packard <keithp@keithp.com> 33085Date: Tue Sep 1 19:03:33 2015 -0700 33086 33087 Bump version to 1.17.99.901 (1.18 RC1) 33088 33089 Signed-off-by: Keith Packard <keithp@keithp.com> 33090 33091commit bcb60a49c5e74aa11d0256874659afddea91e53d 33092Author: Jon TURNEY <jon.turney@dronecode.org.uk> 33093Date: Tue Jul 7 23:11:07 2015 +0100 33094 33095 debug output format fix in TouchEventHistoryPush() 33096 33097 xserver/build.x86_64/../dix/touch.c:468:16: error: format ‘%d’ expects argument of type ‘int’, but argument 3 has type ‘size_t’ [-Werror=format=] 33098 33099 %zu is C99, but is already used in a few places. Perhaps doc/c-extensions needs 33100 to mention it. 33101 33102 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 33103 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 33104 33105commit 1f96a0d273e56d01cd240e6a4ef8765e0ffe058c 33106Author: Jon TURNEY <jon.turney@dronecode.org.uk> 33107Date: Fri May 1 17:40:06 2015 +0100 33108 33109 debug output format fix in ephyrProcessMouseMotion() 33110 33111 xorg/xserver/hw/kdrive/ephyr/ephyr.c:979:9: error: format ‘%d’ expects argument of type ‘int’, but argument 4 has type ‘ScreenPtr’ [-Werror=format=] 33112 33113 This looks like a genuine bug, and ephyrCursorScreen->myNum was meant here 33114 rather than ephyrCursorScreen 33115 33116 v2: 33117 Insert a ":" as well 33118 33119 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 33120 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 33121 33122commit 8cc88fbe9e4699cd2570641eb2752977b5505d16 33123Author: Jon TURNEY <jon.turney@dronecode.org.uk> 33124Date: Fri May 1 17:33:35 2015 +0100 33125 33126 debug output format fix in DRISwapContext() 33127 33128 xorg/xserver/hw/xfree86/dri/dri.c:1695:19: error: format ‘%x’ expects argument of type ‘unsigned int’, but argument 4 has type ‘DRIContextPrivPtr’ [-Werror=format=] ^ 33129 xorg/xserver/hw/xfree86/dri/dri.c:1695:19: error: format ‘%x’ expects argument of type ‘unsigned int’, but argument 5 has type ‘DRIContextPrivPtr’ [-Werror=format=] 33130 33131 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 33132 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 33133 33134commit 27cf5840272810ae20fdae550db71ec48cf1b02d 33135Author: Jon TURNEY <jon.turney@dronecode.org.uk> 33136Date: Tue Jul 7 23:29:51 2015 +0100 33137 33138 debug output format fix in TI.c 33139 33140 xserver/hw/xfree86/ramdac/TI.c:118:12: error: format ‘%d’ expects argument of type ‘int’, but argument 4 has type ‘long unsigned int’ [-Werror=format=] 33141 xserver/hw/xfree86/ramdac/TI.c:118:12: error: format ‘%d’ expects argument of type ‘int’, but argument 5 has type ‘long unsigned int’ [-Werror=format=] 33142 xserver/hw/xfree86/ramdac/TI.c:118:12: error: format ‘%d’ expects argument of type ‘int’, but argument 6 has type ‘long unsigned int’ [-Werror=format=] 33143 33144 Use %lu for an unsigned long 33145 33146 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 33147 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 33148 33149commit 6cc0f3d95d62e33a5d86f865159cc3526d3a9dc5 33150Author: Jon TURNEY <jon.turney@dronecode.org.uk> 33151Date: Tue Jul 7 23:20:29 2015 +0100 33152 33153 debug output format fix in xf86Events.c 33154 33155 xserver/hw/xfree86/common/xf86Events.c:183:5: error: format ‘%x’ expects argument of type ‘unsigned int’, but argument 3 has type ‘void *’ [-Werror=format=] 33156 33157 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 33158 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 33159 33160commit d206c240c0b85c4da44f073d6e9a692afb6b96d2 33161Author: Olivier Fourdan <ofourdan@redhat.com> 33162Date: Tue Jul 7 18:19:50 2015 +0200 33163 33164 configurable maximum number of clients 33165 33166 Make the maximum number of clients user configurable, either from the command 33167 line or from xorg.conf 33168 33169 This patch works by using the MAXCLIENTS (raised to 512) as the maximum 33170 allowed number of clients, but allowing the actual limit to be set by the 33171 user to a lower value (keeping the default of 256). 33172 33173 There is a limit size of 29 bits to be used to store both the client ID and 33174 the X resources ID, so by reducing the number of clients allowed to connect to 33175 the X server, the user can increase the number of X resources per client or 33176 vice-versa. 33177 33178 Parts of this patch are based on a similar patch from Adam Jackson 33179 <ajax@redhat.com> 33180 33181 This now requires at least xproto 7.0.28 33182 33183 Signed-off-by: Adam Jackson <ajax@redhat.com> 33184 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 33185 Reviewed-by: Adam Jackson <ajax@redhat.com> 33186 Signed-off-by: Keith Packard <keithp@keithp.com> 33187 33188commit 7ecdfbf0af3547295b245efa754123db65cabb43 33189Merge: bca4f4b56 a0b4f30b1 33190Author: Keith Packard <keithp@keithp.com> 33191Date: Mon Aug 17 18:53:25 2015 -0700 33192 33193 Merge remote-tracking branch 'airlied/modesetting-zaphod' 33194 33195commit bca4f4b56c7619d17f9df46f7aac392ea01a9429 33196Author: Keith Packard <keithp@keithp.com> 33197Date: Mon Aug 17 18:37:10 2015 -0700 33198 33199 mi: Always initialize edge1 and edge2 in miLineArc 33200 33201 This eliminates a warning generated when miLineArcD is inlined and the 33202 compiler can't figure out that edge1 and edge2 are always initialized 33203 before being used. 33204 33205 Signed-off-by: Keith Packard <keithp@keithp.com> 33206 Reviewed-by: Dave Airlie <airlied@redhat.com> 33207 33208commit 5daf1295cb4f71510da595721d10f7ea3b74d1b9 33209Author: Keith Packard <keithp@keithp.com> 33210Date: Mon Aug 17 18:35:20 2015 -0700 33211 33212 dixfont.h: Include fontutil.h before GetGlyphs is #defined 33213 33214 This eliminates the duplicate declaration of dixGetGlyphs when 33215 fontutil.h gets included after dixfont.h has defined GetGlyphs as 33216 dixGetGlyphs. 33217 33218 Signed-off-by: Keith Packard <keithp@keithp.com> 33219 Reviewed-by: Dave Airlie <airlied@redhat.com> 33220 33221commit d4710004367ab8eab794a12385ca7ed3acc9a0e0 33222Author: Dave Airlie <airlied@redhat.com> 33223Date: Thu Aug 13 09:25:37 2015 +1000 33224 33225 rootless: rename w->_w to avoid shadow warnings 33226 33227 fixes: 33228 In file included from rootlessWindow.c:51:0: 33229 rootlessWindow.c: In function 'RootlessResizeWindow': 33230 rootlessCommon.h:198:19: warning: declaration of 'w' shadows a parameter [-Wshadow] 33231 WindowPtr w = pWin; \ 33232 ^ 33233 rootlessWindow.c:1292:9: note: in expansion of macro 'HUGE_ROOT' 33234 HUGE_ROOT(pWin); 33235 ^ 33236 rootlessWindow.c:1262:35: note: shadowed declaration is here 33237 unsigned int w, unsigned int h, WindowPtr pSib) 33238 ^ 33239 In file included from rootlessWindow.c:51:0: 33240 rootlessCommon.h:207:19: warning: declaration of 'w' shadows a parameter [-Wshadow] 33241 WindowPtr w = pWin; \ 33242 ^ 33243 rootlessWindow.c:1296:9: note: in expansion of macro 'NORMAL_ROOT' 33244 NORMAL_ROOT(pWin); 33245 ^ 33246 rootlessWindow.c:1262:35: note: shadowed declaration is here 33247 unsigned int w, unsigned int h, WindowPtr pSib) 33248 ^ 33249 33250 Signed-off-by: Dave Airlie <airlied@redhat.com> 33251 Reviewed-by: Keith Packard <keithp@keithp.com> 33252 Signed-off-by: Keith Packard <keithp@keithp.com> 33253 33254commit 7f506b8099d23c2f045ad26cc5cff00109b69d7b 33255Author: Dave Airlie <airlied@redhat.com> 33256Date: Thu Aug 13 09:25:36 2015 +1000 33257 33258 rootless: fix warnings due to lack of const keeping. 33259 33260 Signed-off-by: Dave Airlie <airlied@redhat.com> 33261 Reviewed-by: Keith Packard <keithp@keithp.com> 33262 Signed-off-by: Keith Packard <keithp@keithp.com> 33263 33264commit b923443816320d0636d6fd40c3c1125b93332907 33265Author: Dave Airlie <airlied@redhat.com> 33266Date: Thu Aug 13 09:25:35 2015 +1000 33267 33268 mioverlay.c: remove shadowed pScreen. 33269 33270 This is already defined at the function entry. 33271 33272 fixes warning: 33273 CC mivaltree.lo 33274 mioverlay.c: In function 'miOverlayWindowExposures': 33275 mioverlay.c:993:23: warning: declaration of 'pScreen' shadows a previous local [-Wshadow] 33276 ScreenPtr pScreen = pWin->drawable.pScreen; 33277 ^ 33278 mioverlay.c:986:15: note: shadowed declaration is here 33279 ScreenPtr pScreen = pWin->drawable.pScreen; 33280 33281 Signed-off-by: Dave Airlie <airlied@redhat.com>> 33282 Reviewed-by: Keith Packard <keithp@keithp.com> 33283 Signed-off-by: Keith Packard <keithp@keithp.com> 33284 33285commit a0b4f30b1f584d2957c9c99e430da5740d2d4ab2 33286Author: Mario Kleiner <mario.kleiner.de@gmail.com> 33287Date: Wed Aug 5 00:08:40 2015 +0200 33288 33289 modesetting: Add more missing options to man page. 33290 33291 Descriptions for Options PageFlip and SWCursor. 33292 33293 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 33294 Reviewed-by: Dave Airlie <airlied@redhat.com> 33295 Signed-off-by: Dave Airlie <airlied@redhat.com> 33296 33297commit 19e1dc8f6ea6d7ff5ba4a5caa0e2f40a47879408 33298Author: Dave Airlie <airlied@redhat.com> 33299Date: Wed Jul 22 03:56:13 2015 +0100 33300 33301 modesetting: add zaphod support (v3) 33302 33303 This adds zaphod and ZaphodHeads support 33304 to the the in-server modesetting driver. 33305 33306 this is based on a request from Mario, 33307 and on the current radeon driver, along 33308 with some patches from Mario to bring things 33309 up to the state of the art in Zaphod. 33310 33311 v2: fixup vblank fd registring. 33312 v3: squash Mario's fixes. 33313 modesetting: Allow/Fix use of multiple ZaphodHead outputs per x-screen. 33314 modesetting: Take shift in crtc positions for ZaphodHeads configs into account. 33315 modesetting: Add ZaphodHeads description to man page. 33316 small cleanups (airlied). 33317 33318 Reviewed-and-tested-by: Mario Kleiner <mario.kleiner.de@gmail.com> 33319 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 33320 Signed-off-by: Dave Airlie <airlied@redhat.com> 33321 33322commit 533fb627398e20f863234d780f4463e37007515b 33323Author: Adam Jackson <ajax@redhat.com> 33324Date: Tue Aug 4 13:17:02 2015 -0400 33325 33326 xwayland: Don't (double) destroy input resources in CloseScreen 33327 33328 By the time we get here we've already done CloseDownDevices, so on the 33329 second regeneration you get: 33330 33331 Invalid read of size 4 33332 at 0x43402A: RemoveDevice (devices.c:1125) 33333 by 0x427902: xwl_seat_destroy (xwayland-input.c:568) 33334 by 0x42649C: xwl_close_screen (xwayland.c:116) 33335 by 0x4B7F67: CursorCloseScreen (cursor.c:187) 33336 by 0x536003: AnimCurCloseScreen (animcur.c:106) 33337 by 0x539831: present_close_screen (present_screen.c:64) 33338 by 0x43E486: dix_main (main.c:351) 33339 by 0x30D70206FF: (below main) (libc-start.c:289) 33340 Address 0x980e1a0 is 64 bytes inside a block of size 904 33341 33342 at 0x4A07D6A: free (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) 33343 by 0x434158: RemoveDevice (devices.c:1157) 33344 by 0x42F77B: CloseDeviceList (devices.c:1017) 33345 by 0x430246: CloseDownDevices (devices.c:1047) 33346 by 0x43E3EB: dix_main (main.c:333) 33347 by 0x30D70206FF: (below main) (libc-start.c:289) 33348 33349 Signed-off-by: Adam Jackson <ajax@redhat.com> 33350 Reviewed-by: Daniel Stone <daniels@collabora.com> 33351 Signed-off-by: Keith Packard <keithp@keithp.com> 33352 33353commit 11f4cc47a815c5d6284b1338e748e2605b2dd779 33354Author: Adam Jackson <ajax@redhat.com> 33355Date: Tue Aug 4 13:17:01 2015 -0400 33356 33357 xwayland: Don't (double) destroy RANDR resources in CloseScreen 33358 33359 By the time we get here we've already been through FreeAllResources, 33360 which has already torn down the RANDR objects, so on the second 33361 regeneration you get: 33362 33363 Invalid read of size 4 33364 at 0x51C6F0: RRCrtcDestroy (rrcrtc.c:659) 33365 by 0x4285F5: xwl_output_destroy (xwayland-output.c:191) 33366 by 0x426464: xwl_close_screen (xwayland.c:112) 33367 by 0x4B7F77: CursorCloseScreen (cursor.c:187) 33368 by 0x536013: AnimCurCloseScreen (animcur.c:106) 33369 by 0x539841: present_close_screen (present_screen.c:64) 33370 by 0x43E496: dix_main (main.c:351) 33371 by 0x30D70206FF: (below main) (libc-start.c:289) 33372 Address 0x4cc6640 is 0 bytes inside a block of size 728 free'd 33373 at 0x4A07D6A: free (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so) 33374 by 0x51BCCF: RRCrtcDestroyResource (rrcrtc.c:689) 33375 by 0x45CD91: doFreeResource (resource.c:872) 33376 by 0x45DE56: FreeClientResources (resource.c:1138) 33377 by 0x45DF06: FreeAllResources (resource.c:1153) 33378 by 0x43E3BD: dix_main (main.c:321) 33379 by 0x30D70206FF: (below main) (libc-start.c:289) 33380 33381 Signed-off-by: Adam Jackson <ajax@redhat.com> 33382 Reviewed-by: Daniel Stone <daniels@collabora.com> 33383 Signed-off-by: Keith Packard <keithp@keithp.com> 33384 33385commit ea03e314f98e5d8ed7bf7a508006a3d84014bde5 33386Author: Eric Anholt <eric@anholt.net> 33387Date: Mon Jul 27 13:54:00 2015 -0700 33388 33389 glamor: Don't try to free the pixmap priv if we fail to allocate FBO. 33390 33391 Fixes a regression since a2a2f6e34bd49e7ae31779274d52e800595660bc. I 33392 missed this in testing on x86, because we never fail to allocate an 33393 FBO. We do hit this path on VC4, though. 33394 33395 Signed-off-by: Eric Anholt <eric@anholt.net> 33396 Reviewed-by: Dave Airlie <airlied@redhat.com> 33397 33398commit 6916d32ab3ec1aa922094abce7fac22884ee4e83 33399Author: Alan Coopersmith <alan.coopersmith@oracle.com> 33400Date: Sun Jul 19 08:24:39 2015 -0700 33401 33402 Xserver.man: -retro is used when starting the server, not the stipple 33403 33404 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 33405 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33406 Signed-off-by: Eric Anholt <eric@anholt.net> 33407 33408commit 0a458a908ec071a4da5d22c760581e0c5ec885ce 33409Author: Michel Dänzer <michel.daenzer@amd.com> 33410Date: Wed Jul 22 15:37:23 2015 +0900 33411 33412 glamor: Make our EGL context current before calling into GL in glamor_init 33413 33414 Without this, the context of another screen may be current, or no context 33415 at all if glamor_egl_init failed for another screen. 33416 33417 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 33418 Signed-off-by: Eric Anholt <eric@anholt.net> 33419 Reviewed-by: Dave Airlie <airlied@redhat.com> 33420 Reviewed-by: Eric Anholt <eric@anholt.net> 33421 33422commit a8a0f6464a33c12c1de495d74fd478c0d952643e 33423Author: Aaron Plattner <aplattner@nvidia.com> 33424Date: Thu Jul 16 15:12:52 2015 -0700 33425 33426 xfree86: Bump video driver ABI version to 20 33427 33428 Commit 90db5edf119187f8b1b9207c8c384d6cd7ef9edc modified the signature of 33429 StartPixmapTrackingProcPtr, so drivers implementing that need to use the updated 33430 definition. 33431 33432 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 33433 Signed-off-by: Eric Anholt <eric@anholt.net> 33434 33435commit 2fcfa532532fbe4a7f668556808e6245ff4e36bc 33436Merge: cb695b0f3 7b0f94062 33437Author: Eric Anholt <eric@anholt.net> 33438Date: Fri Jul 17 10:15:01 2015 -0700 33439 33440 Merge remote-tracking branch 'ajax/xserver-next' into master 33441 33442commit cb695b0f3b8def4d1ac02458fe630b65f6a5a3e6 33443Author: Armin K <krejzi@email.com> 33444Date: Thu Jul 16 14:54:48 2015 +0200 33445 33446 xfree86/os-support/linux: Fix make distcheck 33447 33448 Header was added in 1dba5a0b194653b0effb01a918bd7338b0c6bcb9 33449 but not in Makefile.am, resulting in missing header in the 33450 distribution tarball. 33451 33452 Signed-off-by: Eric Anholt <eric@anholt.net> 33453 33454commit 29efa905ec1e22d66b517378c6e5ad47a09d6977 33455Author: Robert Ancell <robert.ancell@canonical.com> 33456Date: Thu Jul 16 12:26:05 2015 +1200 33457 33458 modesetting: Use correct types for return values of glamor BO exports. 33459 33460 glamor_name_from_pixmap and glamor_fd_from_pixmap return CARD16 and 33461 CARD32 values via pointers. The current code uses uint16_t and 33462 uint32_t which will probably be the same but it's safer to use the 33463 datatypes as specified by the function. 33464 33465 Signed-off-by: Robert Ancell <robert.ancell@canonical.com> 33466 Signed-off-by: Eric Anholt <eric@anholt.net> 33467 33468commit db5337afb248edf81087cf8d74006fc496d70589 33469Author: Dave Airlie <airlied@redhat.com> 33470Date: Wed Jul 15 17:56:11 2015 +1000 33471 33472 glamor: make current in prepare paths 33473 33474 Lots of the accel paths only make current once they start 33475 doing someting, so a lot of them call the bail paths without 33476 make current, which means on PRIME systems for example 33477 we end up in the wrong context. 33478 33479 Add a prepare pixmap in the prepare fallback path. 33480 33481 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90667 33482 Signed-off-by: Dave Airlie <airlied@redhat.com> 33483 Signed-off-by: Eric Anholt <eric@anholt.net> 33484 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 33485 Reviewed-by: Eric Anholt <eric@anholt.net> 33486 33487commit 1a18513a4eb3fa22459dd9f7d8f0e275aff679ec 33488Author: Michel Dänzer <michel.daenzer@amd.com> 33489Date: Fri Jul 10 15:28:52 2015 +0900 33490 33491 glamor: Use glamor_prepare_access_box() for PutImage/GetImage fallback 33492 33493 Fixes slow text display in xdvi. 33494 33495 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=91260 33496 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 33497 Signed-off-by: Eric Anholt <eric@anholt.net> 33498 Reviewed-by: Dave Airlie <airlied@redhat.com> 33499 Reviewed-by: Eric Anholt <eric@anholt.net> 33500 33501commit fde13565c1b2462ee38f2a446ad3c9157261afa2 33502Author: Eric Anholt <eric@anholt.net> 33503Date: Thu Jul 2 13:57:05 2015 -0700 33504 33505 glamor: Drop unused box translation/bounds code. 33506 33507 These are dead since the glamor_copy.c replacement. 33508 33509 Signed-off-by: Eric Anholt <eric@anholt.net> 33510 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33511 Reviewed-by: Dave Airlie <airlied@redhat.com> 33512 33513commit ab5aa270c79d70f095bc7abadeef227b4062027c 33514Author: Eric Anholt <eric@anholt.net> 33515Date: Thu Jul 2 13:52:05 2015 -0700 33516 33517 glamor: Move cache_format to glamor_fbo.c, where it's used. 33518 33519 Signed-off-by: Eric Anholt <eric@anholt.net> 33520 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33521 Reviewed-by: Dave Airlie <airlied@redhat.com> 33522 33523commit 0ca783e8ee5dfb086fbfa64d26173bcca78b4010 33524Author: Eric Anholt <eric@anholt.net> 33525Date: Thu Jul 2 13:50:42 2015 -0700 33526 33527 glamor: Drop another dead function. 33528 33529 This hasn't been used since the format swap/revert stuff for pictures 33530 was added back in 2012. 33531 33532 Signed-off-by: Eric Anholt <eric@anholt.net> 33533 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33534 Reviewed-by: Dave Airlie <airlied@redhat.com> 33535 33536commit c1111710628cc2a9dfaee0d74ea9fceee990095b 33537Author: Eric Anholt <eric@anholt.net> 33538Date: Thu Jul 2 13:33:01 2015 -0700 33539 33540 glamor: Take transforms into account when preparing for a fallback. 33541 33542 This function takes the start x/y and the destination's width/height, 33543 so it only works if there's no transform. We could potentially 33544 transform this box and take its bounds with some rounding, but this at 33545 least gets us to read out enough data. 33546 33547 Note that this does the same overshoot on destination pictures with a 33548 transform attached, but that seems unlikely to be used anyway. 33549 33550 v2: Add XXX comment for the commit message note (Suggested by Michel). 33551 33552 Signed-off-by: Eric Anholt <eric@anholt.net> 33553 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> (v1) 33554 Reviewed-by: Dave Airlie <airlied@redhat.com> (v1) 33555 33556commit c16e086dace81151f399485e5c686617a43d5b94 33557Author: Eric Anholt <eric@anholt.net> 33558Date: Thu Jul 2 13:25:40 2015 -0700 33559 33560 glamor: Drop dead drm_stride field. 33561 33562 Signed-off-by: Eric Anholt <eric@anholt.net> 33563 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33564 Reviewed-by: Dave Airlie <airlied@redhat.com> 33565 33566commit 8097c887023b72744cb2cb943c2deca7ca0c404a 33567Author: Eric Anholt <eric@anholt.net> 33568Date: Thu Jul 2 13:11:20 2015 -0700 33569 33570 glamor: Drop tracking of the last picture attached to pixmaps. 33571 33572 Signed-off-by: Eric Anholt <eric@anholt.net> 33573 Acked-by: Michel Dänzer <michel.daenzer@amd.com> 33574 Reviewed-by: Dave Airlie <airlied@redhat.com> 33575 33576commit 0dbdb83b0d0254ac67fb33da8bf9ebc0ef4e8081 33577Author: Eric Anholt <eric@anholt.net> 33578Date: Thu Jul 2 12:55:27 2015 -0700 33579 33580 glamor: Use the actual picture's format when uploading memory pixmaps. 33581 33582 The pixmap->picture is just the *last* picture attached to the pixmap, 33583 so you'd potentially be looking at the wrong one when trying to 33584 temporarily upload to avoid a composite fallback. 33585 33586 There's some trickiness in glamor_render.c when we're dealing with the 33587 upload of a GLAMOR_MEMORY pixmap as both the source and mask using 33588 different formats, where we smash the source's format to a new value 33589 so that the mask can use the same uploaded bits. Dropping most of 33590 that should be safe, since it will be uploaded as the source first, so 33591 the smashed format will still be used. 33592 33593 Signed-off-by: Eric Anholt <eric@anholt.net> 33594 Acked-by: Michel Dänzer <michel.daenzer@amd.com> 33595 Reviewed-by: Dave Airlie <airlied@redhat.com> 33596 33597commit 1fd11c46aa086e4b80cb1bafa87c6b19a8014fd1 33598Author: Eric Anholt <eric@anholt.net> 33599Date: Thu Jul 2 12:49:36 2015 -0700 33600 33601 glamor: Move glamor_render.c pict handling to glamor_picture.c 33602 33603 These functions aren't used by anything else, and are specific to the 33604 temporary-upload-as-a-weird-format path of glamor_render.c, called 33605 through glamor_upload_picture_to_texture(). 33606 33607 Signed-off-by: Eric Anholt <eric@anholt.net> 33608 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33609 Reviewed-by: Dave Airlie <airlied@redhat.com> 33610 33611commit 53aed06fb45aed4145754cd622508563ad262c4a 33612Author: Eric Anholt <eric@anholt.net> 33613Date: Thu Jul 2 12:48:01 2015 -0700 33614 33615 glamor: Drop dead glamor_upload_bits_to_pixmap_texture() proto. 33616 33617 Signed-off-by: Eric Anholt <eric@anholt.net> 33618 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33619 Reviewed-by: Dave Airlie <airlied@redhat.com> 33620 33621commit b51e079e1877411fde72372c3953d71e42ac1d6e 33622Author: Eric Anholt <eric@anholt.net> 33623Date: Thu Jul 2 12:40:05 2015 -0700 33624 33625 glamor: Use the new upload/download interface for XV uploads. 33626 33627 We don't need any of its weird handling of picture formats, since our 33628 XV pixmaps don't have any pictures attached. 33629 33630 Signed-off-by: Eric Anholt <eric@anholt.net> 33631 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33632 Reviewed-by: Dave Airlie <airlied@redhat.com> 33633 33634commit 1bd966d16d87ed908547144c68850f2cdee05469 33635Author: Eric Anholt <eric@anholt.net> 33636Date: Thu Jul 2 12:21:41 2015 -0700 33637 33638 glamor: Ignore picture formats when choosing color for core rendering. 33639 33640 Attaching a picture to a pixmap doesn't change its GL format, so it 33641 doesn't change how core rendering should be assigning colors to it. 33642 33643 (Imagine XCreatePixmap(), optional XCreatePicture(pixmap) with various 33644 formats, XFillRectangle, XGetImage(). If the XGetImage results 33645 changed, this would be wrong). 33646 33647 Fixes all failures in "rendercheck -t fill" and, as a result, the 33648 remaining failures in "rendercheck -t blend -o src -f 33649 a8r8g8b8,x2r10g10b10" 33650 33651 Signed-off-by: Eric Anholt <eric@anholt.net> 33652 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33653 Reviewed-by: Dave Airlie <airlied@redhat.com> 33654 33655commit 5a81de8284f851751a8943aa6a905377be315c7b 33656Author: Eric Anholt <eric@anholt.net> 33657Date: Wed Jul 8 12:18:28 2015 -0700 33658 33659 glamor: Don't try to do rendering with unsupported formats. 33660 33661 I'm amazed we've made it as far as we have without these checks: if 33662 you made an unusual format picture that wasn't the normal a8r8g8b8 or 33663 x8r8g8b8 or a8, we'd go ahead and try to render with it, ignoring that 33664 the sampler would fetch totally wrong bits. 33665 33666 Fixes 260 tests in rendercheck -t blend -o src -f a8r8g8b8,x2r10g10b10 33667 33668 Signed-off-by: Eric Anholt <eric@anholt.net> 33669 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33670 Reviewed-by: Dave Airlie <airlied@redhat.com> 33671 33672commit 0d7cbd6f5a2b5c9c690979da9c1483e8f6133903 33673Author: Hans de Goede <hdegoede@redhat.com> 33674Date: Fri Jul 3 20:42:48 2015 +0200 33675 33676 glamor_egl: Properly free resources on init-error and exit 33677 33678 glamor_egl_init() was not undoing any of the init steps on init error, 33679 add an glamor_egl_cleanup() function and use this both on error and on exit 33680 to cleanup the various resources. 33681 33682 Even on a clean exit eglTerminate() was not being called, causing the fd 33683 dup()-ed by eglInitialize() to stay open, call eglTerminate() from the new 33684 glamor_egl_cleanup() to fix this. 33685 33686 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 33687 Signed-off-by: Eric Anholt <eric@anholt.net> 33688 Reviewed-by: Eric Anholt <eric@anholt.net> 33689 33690commit 80b6652c9fb9cdff91e92fa16c07e2cc63b06f44 33691Author: Eric Anholt <eric@anholt.net> 33692Date: Tue Jul 7 17:21:55 2015 -0700 33693 33694 glamor: Drop a bunch of glamor_priv == NULL checks. 33695 33696 Now that it's always non-null when the pixmap is non-null, we don't 33697 need so much of this. glamor_get_pixmap_private() itself still 33698 accepts a NULL pixmap and returns NULL, because of glamor_render.c 33699 33700 Signed-off-by: Eric Anholt <eric@anholt.net> 33701 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33702 Reviewed-by: Keith Packard <keithp@keithp.com> 33703 33704commit a2a2f6e34bd49e7ae31779274d52e800595660bc 33705Author: Eric Anholt <eric@anholt.net> 33706Date: Tue Jun 30 16:20:18 2015 -0700 33707 33708 glamor: Ask the server to always allocate our private. 33709 33710 This avoids a lot of screwing around to attach our privates later. It 33711 means that non-glamor pixmaps now gain 120 bytes of glamor privates on 33712 64-bit (which has quite a bit of fixable bloat), and glamor pixmaps 33713 take one less pointer of storage (not counting malloc overhead). 33714 33715 Note that privates start out zero-filled, which matches the callocs we 33716 were doing when making our own privates, and in the case of an fb 33717 pixmap that has a priv where it didn't before, the type ends up being 33718 GLAMOR_MEMORY as we would want. 33719 33720 v2: Clarify that the GLAMOR_MEMORY enum must be 0 (as it was 33721 previosuly), so that the new pixmap private behavior is as 33722 expected. Suggested by keithp. 33723 33724 Signed-off-by: Eric Anholt <eric@anholt.net> 33725 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> (v1) 33726 Reviewed-by: Keith Packard <keithp@keithp.com> 33727 33728commit d278c30e68d79e86a3e2207b629d1e461856f907 33729Author: Eric Anholt <eric@anholt.net> 33730Date: Tue Jun 30 20:35:59 2015 -0700 33731 33732 glamor: Drop dead glamor_is_large_picture(). 33733 33734 It died as of keithp's new glyphs code. 33735 33736 Signed-off-by: Eric Anholt <eric@anholt.net> 33737 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33738 Reviewed-by: Keith Packard <keithp@keithp.com> 33739 33740commit cd608a516d2c59ebfb95fadfecf68227888c0e4e 33741Author: Eric Anholt <eric@anholt.net> 33742Date: Tue Jun 30 20:25:06 2015 -0700 33743 33744 glamor: Reuse the glamor_is_memory helper. 33745 33746 Signed-off-by: Eric Anholt <eric@anholt.net> 33747 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 33748 Reviewed-by: Keith Packard <keithp@keithp.com> 33749 33750commit 7b0f940625a0f6eee2ef91dfbfa4a206246940f5 33751Author: Adam Jackson <ajax@redhat.com> 33752Date: Fri Dec 19 13:46:37 2014 -0500 33753 33754 atom: make FreeAtom static 33755 33756 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 33757 Signed-off-by: Adam Jackson <ajax@redhat.com> 33758 33759commit 2d2cf9e322a5ec3ead175edc136cdaeea00ec1b1 33760Author: Adam Jackson <ajax@redhat.com> 33761Date: Fri Dec 19 12:57:41 2014 -0500 33762 33763 atom: Bump initial table size 33764 33765 We're always creating ~230 atoms at startup, might as well tune it so we 33766 don't hit the realloc path before Dispatch. 33767 33768 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 33769 Signed-off-by: Adam Jackson <ajax@redhat.com> 33770 33771commit 995ff11422eb49160abfe276f471e412b16cab9d 33772Author: Adam Jackson <ajax@redhat.com> 33773Date: Tue May 19 11:31:25 2015 -0400 33774 33775 glx: Implement GLX_ARB_context_flush_control 33776 33777 This extension allows clients to opt out of the implicit glFlush on 33778 context release, which is quite nice for performance for clients using 33779 multiple contexts. The server doesn't really need to be aware of the 33780 client's decision, at least for direct contexts, but it does need to not 33781 reject the context attribute out of hand. 33782 33783 This patch won't do anything unless built against a Mesa that defines 33784 the __DRI2_FLUSH_CONTROL extension (and a new enough glxext.h, but 33785 that's been there since 10.3 at least). 33786 33787 Reviewed-by: James Jones <jajones@nvidia.com> 33788 Signed-off-by: Adam Jackson <ajax@redhat.com> 33789 33790commit 2d7194334a9f84e417ec90e220b2fe476f704612 33791Author: Adam Jackson <ajax@redhat.com> 33792Date: Fri May 15 11:27:31 2015 -0400 33793 33794 glx/swrast: Do more GLX extension setup 33795 33796 This gets you nice things like core contexts when using Xvfb. 33797 33798 Also, no, MESA_copy_sub_buffer is not enabled automatically. 33799 33800 Reviewed-by: James Jones <jajones@nvidia.com> 33801 Reviewed-by: Jon Turney <jon.turney@dronecode.org.uk> 33802 Signed-off-by: Adam Jackson <ajax@redhat.com> 33803 33804commit cbd3cfbad3f07b20e90ea9498110f255813eb441 33805Author: Adam Jackson <ajax@redhat.com> 33806Date: Tue Jun 2 13:58:30 2015 -0400 33807 33808 dix: Restore PaintWindow screen hook 33809 33810 Removes the last cpp conditional on ROOTLESS from dix code. 33811 33812 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 33813 Signed-off-by: Adam Jackson <ajax@redhat.com> 33814 33815commit 74d64ceea02bffad1caf3d1a727edbd38d968059 33816Author: Adam Jackson <ajax@redhat.com> 33817Date: Tue Jun 2 13:58:30 2015 -0400 33818 33819 mi: Remove questionable optimization from the rootless path 33820 33821 This is effectively a revert of 7b506fdc840aebed6b5acb91437a2cb620b5bddc 33822 except the coding style reindent broke that. The code makes no sense in 33823 any case. drawable can never be null since it's the first member of 33824 WindowRec, and we're never called with a null window. Neither can it be 33825 an UNDRAWABLE_WINDOW since those are InputOnly windows; the rootless 33826 code does not set the root window to either UNDRAWABLE or InputOnly, so. 33827 33828 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 33829 Signed-off-by: Adam Jackson <ajax@redhat.com> 33830 33831commit 315661a425018a546f7bcc18ad3e5f4578473ca6 33832Author: Adam Jackson <ajax@redhat.com> 33833Date: Tue Jun 2 13:58:30 2015 -0400 33834 33835 fb: Make rootless-agnostic 33836 33837 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 33838 Signed-off-by: Adam Jackson <ajax@redhat.com> 33839 33840commit 1594fb912996e9776d9e36383bd1d9f5dd2877eb 33841Author: Adam Jackson <ajax@redhat.com> 33842Date: Tue Jun 2 13:58:30 2015 -0400 33843 33844 rootless: Build unconditionally 33845 33846 It's not a lot of code, and if you're doing things to the core that 33847 affect rootless you should be alerted early. 33848 33849 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 33850 Signed-off-by: Adam Jackson <ajax@redhat.com> 33851 33852commit bdc917a41691250701ccc9a0b087448b50892fa9 33853Author: Adam Jackson <ajax@redhat.com> 33854Date: Tue Jun 2 13:58:30 2015 -0400 33855 33856 rootless: Genericize the config header a bit 33857 33858 ROOTLESS_RESIZE_GRAVITY is an optimization, so let's default it to off 33859 unless the backing window system is known to support it. 33860 33861 ROOTLESS_PROTECT_ALPHA looks like it has different values for xquartz 33862 and win32, but hilariously rootlessGC.c merely checks if it is defined, 33863 not what the value is. It's effectively the more conservative path 33864 anyway, so let's just turn it on. 33865 33866 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 33867 Signed-off-by: Adam Jackson <ajax@redhat.com> 33868 33869commit 6f3332b9f411fd7ac9cab0085faa56f6c2ff3323 33870Author: Adam Jackson <ajax@redhat.com> 33871Date: Tue Jun 2 13:58:30 2015 -0400 33872 33873 dix: unifdef pWin->rootlessUnhittable 33874 33875 No reason to vary the dix ABI over this. 33876 33877 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 33878 Signed-off-by: Adam Jackson <ajax@redhat.com> 33879 33880commit 36fac0dd1a9a9cf41b631cf7e1ea015bcf920db7 33881Author: Adam Jackson <ajax@redhat.com> 33882Date: Fri May 23 13:23:20 2014 -0400 33883 33884 xfixes: Unexport xfixes.h 33885 33886 There's nothing a driver could want to use this for. 33887 33888 Acked-by: Chris Wilson <chris@chris-wilson.co.uk> 33889 Signed-off-by: Adam Jackson <ajax@redhat.com> 33890 33891commit 4da66d9e0399141feffd5f002e4195a438fe5020 33892Author: Adam Jackson <ajax@redhat.com> 33893Date: Tue Jul 8 14:27:02 2014 -0400 33894 33895 vidmode: Hide implementation details 33896 33897 Also remove vidmodeproc.h from the SDK since no drivers are using it. 33898 33899 Acked-by: Chris Wilson <chris@chris-wilson.co.uk> 33900 Signed-off-by: Adam Jackson <ajax@redhat.com> 33901 33902commit 49d7bae7f4c6a93f92df828440e3478513f930f4 33903Author: Adam Jackson <ajax@redhat.com> 33904Date: Tue Jul 8 14:18:07 2014 -0400 33905 33906 xge: Hide some implementation details 33907 33908 Acked-by: Chris Wilson <chris@chris-wilson.co.uk> 33909 Signed-off-by: Adam Jackson <ajax@redhat.com> 33910 33911commit 2377690709de11087edc9d1116c9feaeb8f118ed 33912Author: Adam Jackson <ajax@redhat.com> 33913Date: Tue Jul 8 14:03:13 2014 -0400 33914 33915 dga: Hide a bunch of implementation details 33916 33917 Acked-by: Chris Wilson <chris@chris-wilson.co.uk> 33918 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 33919 Signed-off-by: Adam Jackson <ajax@redhat.com> 33920 33921commit b5fbe9c6327fac88c09afce598238a7534521b69 33922Author: Adam Jackson <ajax@redhat.com> 33923Date: Tue Jul 8 13:37:34 2014 -0400 33924 33925 xfree86: Hide some pre-randr mode validation details 33926 33927 Acked-by: Chris Wilson <chris@chris-wilson.co.uk> 33928 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 33929 Signed-off-by: Adam Jackson <ajax@redhat.com> 33930 33931commit 70b402be2626d4f4afb02b6f5baa543262548bd7 33932Author: Adam Jackson <ajax@redhat.com> 33933Date: Fri May 30 16:19:08 2014 -0400 33934 33935 render: Hide/unexport some implementation details 33936 33937 Acked-by: Chris Wilson <chris@chris-wilson.co.uk> 33938 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 33939 Signed-off-by: Adam Jackson <ajax@redhat.com> 33940 33941commit a1c80bb7d5871e1c5ee21fc650834858054103bf 33942Author: Adam Jackson <ajax@redhat.com> 33943Date: Fri May 30 15:49:38 2014 -0400 33944 33945 randr: Unexport some implementation details 33946 33947 Acked-by: Chris Wilson <chris@chris-wilson.co.uk> 33948 Signed-off-by: Adam Jackson <ajax@redhat.com> 33949 33950commit b51f7f8582ab6c3cc9fa56c8d9721d0f240915e7 33951Author: Adam Jackson <ajax@redhat.com> 33952Date: Fri May 23 13:37:41 2014 -0400 33953 33954 dix: Unexport various implementation details 33955 33956 Acked-by: Chris Wilson <chris@chris-wilson.co.uk> 33957 Signed-off-by: Adam Jackson <ajax@redhat.com> 33958 33959commit c4a0d6c9139d2c0107b80420cc2342614bbe95ef 33960Author: Adam Jackson <ajax@redhat.com> 33961Date: Wed Oct 22 10:47:03 2014 -0400 33962 33963 dri3: static cleanup 33964 33965 Also remove unused window private key. 33966 33967 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 33968 Signed-off-by: Adam Jackson <ajax@redhat.com> 33969 33970commit a69e9027e167a83281b96e13f3a77100876b72ae 33971Author: Adam Jackson <ajax@redhat.com> 33972Date: Wed Oct 22 10:44:36 2014 -0400 33973 33974 dri3: Remove unbuilt dri3_event.c 33975 33976 The config notifies live in present, not dri3. 33977 33978 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 33979 Signed-off-by: Adam Jackson <ajax@redhat.com> 33980 33981commit 91f0d71c18254746e1d519f0d19d11f756a9588c 33982Author: Adam Jackson <ajax@redhat.com> 33983Date: Wed Oct 22 10:38:11 2014 -0400 33984 33985 parser: static cleanup 33986 33987 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 33988 Signed-off-by: Adam Jackson <ajax@redhat.com> 33989 33990commit c732bb2799ad7ff6b0cdd9c65487466e6a9b76e7 33991Author: Adam Jackson <ajax@redhat.com> 33992Date: Wed Oct 22 10:31:29 2014 -0400 33993 33994 present: static cleanup 33995 33996 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 33997 Signed-off-by: Adam Jackson <ajax@redhat.com> 33998 33999commit 0db457b2b5b6e314e912efc62f158f8f0b8d48af 34000Author: Adam Jackson <ajax@redhat.com> 34001Date: Wed Jul 8 15:56:55 2015 -0400 34002 34003 dmx: Fix the build 34004 34005 Broken since: 34006 34007 commit 4fd81823fafcd103e8d890f4c0f7c2f90e822336 34008 Author: Dave Airlie <airlied@gmail.com> 34009 Date: Tue Jun 30 14:54:42 2015 +1000 34010 34011 prime: add rotation support for offloaded outputs (v2) 34012 34013 Signed-off-by: Adam Jackson <ajax@redhat.com> 34014 34015commit 90db5edf119187f8b1b9207c8c384d6cd7ef9edc 34016Author: Dave Airlie <airlied@gmail.com> 34017Date: Tue Jun 30 14:54:42 2015 +1000 34018 34019 prime: add rotation support for offloaded outputs (v2) 34020 34021 One of the lacking features with output offloading was 34022 that screen rotation didn't work at all. 34023 34024 This patch makes 0/90/180/270 rotation work with USB output 34025 and GPU outputs. 34026 34027 When it allocates the shared pixmap it allocates it rotated, 34028 and any updates to the shared pixmap are done using a composite 34029 path that does the rotation. The slave GPU then doesn't need 34030 to know about the rotation and just displays the pixmap. 34031 34032 v2: 34033 rewrite the sync dirty helper to use the dst pixmap, and 34034 avoid any strange hobbits and rotations. 34035 34036 This breaks ABI in two places. 34037 34038 Signed-off-by: Dave Airlie <airlied@redhat.com> 34039 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34040 Signed-off-by: Keith Packard <keithp@keithp.com> 34041 34042commit 991712f1e8deeb6289ee0abd9910e279d6396246 34043Merge: 0cd228073 3c859112d 34044Author: Keith Packard <keithp@keithp.com> 34045Date: Wed Jul 8 10:43:31 2015 -0700 34046 34047 Merge remote-tracking branch 'ofourdan/for-keith' 34048 34049commit 0cd228073ad3b7eb72cef8b61128324895ee6cd4 34050Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34051Date: Mon Jun 29 15:08:51 2015 +0100 34052 34053 hw/xwin: printf format fixes for Pixel type 34054 34055 Pixel is CARD32, so inside the server has type unsigned int (x86_64) or unsigned 34056 long (x86) 34057 34058 Cast to unsigned int and use a %u format 34059 34060 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34061 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34062 34063commit e3cfeb949a9c9363beacdb00acdd9723ed54fac2 34064Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34065Date: Fri Feb 6 20:22:11 2015 +0000 34066 34067 hw/xwin: printf format fixes for WPARAM and LPARAM types 34068 34069 Some Win32 API types are different fundamental types in the 32-bit and 64-bit 34070 34071 This problem is then further compounded by the fact that whilst both 32-bit 34072 Cygwin and 32-bit MinGW use the ILP32 data model, 64-bit MinGW uses the LLP64 34073 data model, but 64-bit Cygwin uses the LP64 data model. 34074 34075 This makes it impossible to write printf format specifiers which are correct for 34076 all those targets, so we use some macros to provide the correct specifier for 34077 the target. 34078 34079 LPARAM and WPARAM are integer types which can contain a pointer 34080 34081 LPARAM is long in ILP32 and long long in LLP64 34082 WPARAM is unsigned int in ILP32 and unsigned long long in LLP64 34083 34084 Generally, these are just used to passs integer parameters, so for simplicity, 34085 cast to int and use an int-compatible format 34086 34087 In the specific case of WM_CHANGECBCHAIN, they are used to pass HWND, so cast to 34088 that type and print using an appropriate format. 34089 34090 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34091 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34092 34093commit 4f8661fac985306c56330cae69ffc19e5dd0af61 34094Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34095Date: Fri Feb 6 19:19:41 2015 +0000 34096 34097 hw/xwin: printf format fixes for LONG type 34098 34099 Some Win32 API types are different fundamental types in the 32-bit and 64-bit 34100 versions. 34101 34102 This problem is then further compounded by the fact that whilst both 32-bit 34103 Cygwin and 32-bit MinGW use the ILP32 data model, 64-bit MinGW uses the LLP64 34104 data model, but 64-bit Cygwin uses the LP64 data model. 34105 34106 This makes it impossible to write printf format specifiers which are correct for 34107 all those targets 34108 34109 In the Win32 API, DWORD is an signed, 32-bit type. It is defined in terms of a 34110 long, except in the LP64 data model, where it is an int. 34111 34112 It should always be safe to cast it to int and use %d. 34113 34114 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34115 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34116 34117commit aa83c61f510121da20b56e8f7de700193f7d16b5 34118Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34119Date: Fri Feb 6 19:46:45 2015 +0000 34120 34121 hw/xwin: printf format fixes for DWORD type 34122 34123 Some Win32 API types are different fundamental types in the 32-bit and 64-bit 34124 versions. 34125 34126 This problem is then further compounded by the fact that whilst both 32-bit 34127 Cygwin and 32-bit MinGW use the ILP32 data model, 64-bit MinGW uses the LLP64 34128 data model, but 64-bit Cygwin uses the LP64 data model. 34129 34130 This makes it impossible to write printf format specifiers which are correct for 34131 all those targets 34132 34133 In the Win32 API, DWORD is an unsigned, 32-bit type. It is defined in terms of 34134 an unsigned long, except in the LP64 data model, where it is an unsigned int. 34135 34136 It should always be safe to cast it to unsigned int and use %u or %x. 34137 34138 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34139 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34140 34141commit 487f2595c9dd9a5c3c600168a108963e87602561 34142Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34143Date: Fri Feb 6 20:48:26 2015 +0000 34144 34145 hw/xwin: printf format fixes in xevents.c 34146 34147 Window and Atom types derive from XID, which is always unsigned long in client 34148 code, so use %ld format specifier 34149 34150 XTextProperty.nitems is of type unsigned long, so use %lu format specifier 34151 34152 ulReturnBytesLeft is of type unsigned long, so use %lu format specifier 34153 34154 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34155 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34156 34157commit 98798fcf0d1dc66c539abb2ecec0acc676d0b574 34158Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34159Date: Fri Feb 6 20:42:54 2015 +0000 34160 34161 hw/xwin: printf format fix in winProcessXEventsTimeout() 34162 34163 remainingTime is computed as a long int, so use %ld format specifier 34164 34165 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34166 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34167 34168commit 5b6f511c2fcad4edaccedbeb9842680cf3aeffb1 34169Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34170Date: Fri Feb 6 20:40:46 2015 +0000 34171 34172 hw/xwin: printf format fix in winCreateDefColormap() 34173 34174 Use %lu for unsigned long 34175 34176 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34177 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34178 34179commit eb679671560b21206a47476dfbc23518d8d46588 34180Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34181Date: Fri Feb 6 20:38:26 2015 +0000 34182 34183 hw/xwin: printf format fixes in winAllocatePrivates() 34184 34185 serverGeneration is of type unsigned long, so use %lu format specifier 34186 34187 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34188 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34189 34190commit 0d6c499b9976d305e13685eb79b4bb7cd4250f58 34191Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34192Date: Fri Feb 6 20:27:11 2015 +0000 34193 34194 hw/xwin: printf format fixes in winConfigKeyboard() 34195 34196 struct winInfoRec.keyboard members are of type long, not type int 34197 34198 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34199 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34200 34201commit ba468e003e32c406f4885c8cb3deeb1c8175cd45 34202Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34203Date: Mon Jun 29 14:25:49 2015 +0100 34204 34205 hw/xwin: printf format fixes for XID type 34206 34207 XID inside the server has type unsigned int (x86_64) or unsigned long (x86) 34208 34209 Follow the example of the rest of the server and cast to unsigned int and use 34210 a %u or %x format. 34211 34212 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34213 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34214 34215commit dcb797b31f0affdc4b08b000dd92ec38fa261953 34216Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34217Date: Thu Oct 2 14:28:46 2014 +0100 34218 34219 hw/xwin: printf format fixes for HWND type 34220 34221 HWND derives from HANDLE, a pointer type, so we should use the %p format 34222 34223 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34224 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34225 34226commit f1d14266672c8130e992dd9d0c08fd1cef0f08cb 34227Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34228Date: Wed Oct 1 22:12:26 2014 +0100 34229 34230 hw/xwin: Ensure format warnings in winclipboard/ 34231 34232 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34233 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34234 34235commit 38f340b13e693a0d456a9ab22a6474c8f55375fa 34236Author: Jon TURNEY <jon.turney@dronecode.org.uk> 34237Date: Mon Jun 29 15:30:36 2015 +0100 34238 34239 hw/xwin: Remove GetTickCount() from various pieces of debugging output 34240 34241 The use of %d format for the DWORD return value of GetTickCount() isn't 34242 portable, but it doesn't seem to be worth fixing it when this information isn't 34243 very useful (and is redundant to the timestamping of log messages we now have) 34244 34245 Instead just remove these uses of GetTickCount() 34246 34247 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 34248 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 34249 34250commit 732e3b9c08532f40656010eac9d128601cc88c3f 34251Author: Michel Dänzer <michel.daenzer@amd.com> 34252Date: Mon Jul 6 18:45:51 2015 -0700 34253 34254 Instead of one glTexSubImage2D call for each glyph. 34255 34256 This significantly reduces the amount of time it takes for xterm to start 34257 up on a fresh X server with the radeonsi driver. 34258 34259 v2: Use GLYPHWIDTHBYTESPADDED instead of hardcoding 4 bytes glyph 34260 alignment (Keith Packard) 34261 34262 Reviewed-by: Keith Packard <keithp@keithp.com> 34263 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 34264 Signed-off-by: Keith Packard <keithp@keithp.com> 34265 34266commit ac94cdb9df94dd26688de6d3d9143ac48f7b3c7b 34267Author: Julien Cristau <jcristau@debian.org> 34268Date: Thu Jul 2 20:16:21 2015 +0200 34269 34270 mi: fix typo in warning about overflowing queue 34271 34272 Debian bug#726041 34273 34274 Reported-by: Vincent Hobeïka <vincent.hobeika@gmail.com> 34275 Signed-off-by: Julien Cristau <jcristau@debian.org> 34276 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 34277 Signed-off-by: Keith Packard <keithp@keithp.com> 34278 34279commit e3624aa5fde658c44724a7ea21f5df9e3c586923 34280Author: Dave Airlie <airlied@gmail.com> 34281Date: Wed Jul 1 20:41:15 2015 +1000 34282 34283 xserver: fix build with glamor disabled. 34284 34285 This fixes modesetting when glamor is disabled. 34286 34287 Signed-off-by: Dave Airlie <airlied@redhat.com> 34288 Signed-off-by: Eric Anholt <eric@anholt.net> 34289 Reviewed-by: Eric Anholt <eric@anholt.net> 34290 Reviewed-by: Adam Jackson <ajax@redhat.com> 34291 34292commit 3703c88ac18063a35fd8c0fc52fee52d8c669638 34293Author: Eric Anholt <eric@anholt.net> 34294Date: Wed Jul 1 16:23:45 2015 -0700 34295 34296 glamor: Use GL_ARB_debug_output to log GL errors. 34297 34298 This should help people debugging when glamor does something stupid on 34299 their driver. 34300 34301 Signed-off-by: Eric Anholt <eric@anholt.net> 34302 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 34303 34304commit f4dfd282d311e5f23631d33e0251bad1854b1ab9 34305Author: Eric Anholt <eric@anholt.net> 34306Date: Wed Jul 1 14:46:55 2015 -0700 34307 34308 glamor: Add an assert to catch the previous bug. 34309 34310 Signed-off-by: Eric Anholt <eric@anholt.net> 34311 Acked-by: Michel Dänzer <michel.daenzer@amd.com> 34312 34313commit c268d2fc33318f9ea0019053fa5de18f193e704e 34314Author: Keith Packard <keithp@keithp.com> 34315Date: Thu Jul 2 17:06:27 2015 -0600 34316 34317 glamor: Fix bad rendering of glyphs after an a1 glyph upload. (v3) 34318 34319 Fixes regressions since Eric's "don't make an FBO for the glyph atlas" 34320 change. The a1 upload was a fallback, as expected. However, fallback 34321 reads use glReadPixels() because there's no glGetTexSubImage2D() to 34322 match glTexSubImage2D(). We were just binding the 0 FBO value, so the 34323 glReadPixels() would throw a GL error instead of getting any data. 34324 After the fallback was done we'd write back the undefined data to the 34325 atlas, blowing away the entire rest of the atlas because we didn't 34326 specify any bounds on our prepare. 34327 34328 To fix the fallbacks to actually work, we'd need a prepare path that 34329 allocates some memory memory do a full glGetTexImage() into, then 34330 memcpy out of that. Instead, just dodge the general fallback by 34331 implementing the specific upload we need to do here, which should also 34332 be *much* faster at uploading a1 glyphs since it's not 34333 readpixels/texsubimaging back and forth. 34334 34335 v3: Use CopyPlane to a temp pixmap for the upload 34336 v4: Rewrite anholt's commit message to be from keithp's perspective 34337 (changes by anholt) 34338 34339 Signed-off-by: Keith Packard <keithp@keithp.com> 34340 Signed-off-by: Eric Anholt <eric@anholt.net> 34341 Reviewed-by: Eric Anholt <eric@anholt.net> 34342 34343commit 5c440817f7c128c9a9d78ef9672f22641dbeeaa1 34344Author: Eric Anholt <eric@anholt.net> 34345Date: Tue Jun 30 15:01:53 2015 -0700 34346 34347 glamor: Use ARRAY_SIZE in a couple more places for consistency. 34348 34349 Signed-off-by: Eric Anholt <eric@anholt.net> 34350 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34351 34352commit 7b6edb52435f474c43130f6ac4c6b2d2b28814f5 34353Author: Eric Anholt <eric@anholt.net> 34354Date: Tue Jun 30 14:59:55 2015 -0700 34355 34356 glamor: Mark a bunch of single-file data static. 34357 34358 This gives the compiler a chance to optimize when the data is never 34359 changed -- for example, with pict_format_combine_tab, the compiler 34360 ends up inlining the 24 bytes of data into just 10 more bytes of code. 34361 34362 Signed-off-by: Eric Anholt <eric@anholt.net> 34363 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34364 34365commit 3fe6731b17ae953d49db4c1bf95e0022244513e1 34366Author: Eric Anholt <eric@anholt.net> 34367Date: Tue Jun 30 14:50:43 2015 -0700 34368 34369 glamor: Make a bunch of single-file glamor functions static. 34370 34371 Signed-off-by: Eric Anholt <eric@anholt.net> 34372 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34373 34374commit a8e84e8c3b2047775fe2da9371329290eee23655 34375Author: Eric Anholt <eric@anholt.net> 34376Date: Tue Jun 30 14:45:43 2015 -0700 34377 34378 glamor: Drop dead glamor_es2_pixmap_read_prepare(). 34379 34380 It's been unused since I killed glamor_download_pixmap_to_cpu(). 34381 34382 Signed-off-by: Eric Anholt <eric@anholt.net> 34383 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34384 34385commit cad56dc62dbf2ba097cc0fa61217897ba40d8d66 34386Author: Eric Anholt <eric@anholt.net> 34387Date: Tue Jun 30 14:41:27 2015 -0700 34388 34389 glamor: Restore the hook to glamor_composite_rectangles(). 34390 34391 It was apparently accidentally dropped in keithp's removal of _nf 34392 functions in 90d326fcc687e6d6d4b308f6272ededcf8145a17. 34393 34394 Signed-off-by: Eric Anholt <eric@anholt.net> 34395 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34396 34397commit 7adb38a0d18de2e8b0762d9dd82b851883ad75fe 34398Author: Eric Anholt <eric@anholt.net> 34399Date: Tue Jun 30 14:25:16 2015 -0700 34400 34401 glamor: Fix up some weird formatting in _glamor_create_fbo_array(). 34402 34403 Signed-off-by: Eric Anholt <eric@anholt.net> 34404 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34405 34406commit 835067a4095e5e48703aa6561fea3ee95923e832 34407Author: Eric Anholt <eric@anholt.net> 34408Date: Tue Jun 30 14:14:50 2015 -0700 34409 34410 glamor: Drop a redundant check. 34411 34412 Above, we've already checked for ->fbo && ->fbo->fb and returned. 34413 34414 Signed-off-by: Eric Anholt <eric@anholt.net> 34415 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34416 34417commit 47b868492c7ff0909ff71e3ee1aaa79daaade513 34418Author: Eric Anholt <eric@anholt.net> 34419Date: Thu Feb 5 12:34:41 2015 -0800 34420 34421 glamor: Simplify some temp pixmap extents calculations. 34422 34423 Signed-off-by: Eric Anholt <eric@anholt.net> 34424 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 34425 34426commit bfb6a290afabe492d93ec8024fac6b28f672d7d0 34427Author: Eric Anholt <eric@anholt.net> 34428Date: Thu Feb 5 11:46:32 2015 -0800 34429 34430 glamor: Drop a dead flag to glamor_create_fbo_array(). 34431 34432 v2: Don't forget to set priv->block_w/block_h like the wrapper used 34433 to. 34434 34435 Signed-off-by: Eric Anholt <eric@anholt.net> 34436 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> (v1) 34437 34438commit fe3fedf280596c2c2e69dddd30bb9ff17dbe611e 34439Author: Eric Anholt <eric@anholt.net> 34440Date: Thu Feb 5 11:33:54 2015 -0800 34441 34442 glamor: Drop dead GLAMOR_FBO_DOWNLOADED flag. 34443 34444 Signed-off-by: Eric Anholt <eric@anholt.net> 34445 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 34446 34447commit 447274094147f31025700a2ff6f94e3b0c96fa17 34448Author: Eric Anholt <eric@anholt.net> 34449Date: Thu Feb 5 11:20:50 2015 -0800 34450 34451 glamor: Move doxygen for private structs into the structs. 34452 34453 This should hopefully keep the comments more up to date with the 34454 structure comments. While I'm here, I've reworded a few of them to be 34455 more accurate, and dropped a bunch of stale comments. 34456 34457 Signed-off-by: Eric Anholt <eric@anholt.net> 34458 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 34459 34460commit 33d6384b29db1b28a9d1a035dc7cc82da8f33c99 34461Author: Eric Anholt <eric@anholt.net> 34462Date: Thu Feb 5 11:18:24 2015 -0800 34463 34464 glamor: Drop the dead "PBO" field in an FBO cache struct. 34465 34466 PBOs are only used at the whole pixmap level. 34467 34468 Signed-off-by: Eric Anholt <eric@anholt.net> 34469 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 34470 34471commit 1734aa285663d83070e6313bb976f976aa2bea4c 34472Author: Eric Anholt <eric@anholt.net> 34473Date: Wed Feb 4 18:07:06 2015 -0800 34474 34475 glamor: Remove dead gl_tex flag in the fbo struct. 34476 34477 This used to be used in the old copy_area path. 34478 34479 Signed-off-by: Eric Anholt <eric@anholt.net> 34480 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 34481 34482commit 82d23fc7290a2986efbc2982eeaa0de0ad5ad0d1 34483Author: Eric Anholt <eric@anholt.net> 34484Date: Wed Feb 4 18:02:34 2015 -0800 34485 34486 glamor: Propagate the fact that pbo_valid is never set. 34487 34488 The code to set it was deleted in keithp's big rewrite. 34489 34490 Signed-off-by: Eric Anholt <eric@anholt.net> 34491 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 34492 34493commit 0e3f1252dacdc3194a99a2d090b5c13f070f8799 34494Author: Eric Anholt <eric@anholt.net> 34495Date: Fri Jun 19 18:40:47 2015 -0700 34496 34497 glamor: Avoid using GL_QUADS on VC4. 34498 34499 Improves text rendering from about 284k glyphs per second to 320k 34500 glyphs per second. There's no GL extension for probing this, because 34501 of the philosophy of "Don't expose whether things are really in 34502 hardware or not." 34503 34504 Signed-off-by: Eric Anholt <eric@anholt.net> 34505 Reviewed-by: Dave Airlie <airlied@redhat.com> 34506 34507commit f47e3f539568d19e22d10d7dd4ec09c7d570e716 34508Author: Eric Anholt <eric@anholt.net> 34509Date: Fri Jun 19 18:36:28 2015 -0700 34510 34511 glamor: Use proper Bools for some boolean values. 34512 34513 Signed-off-by: Eric Anholt <eric@anholt.net> 34514 Reviewed-by: Dave Airlie <airlied@redhat.com> 34515 34516commit 4001a7465e6fb27fc82fa20ea5a65283d33a5e2e 34517Author: Eric Anholt <eric@anholt.net> 34518Date: Fri Jun 19 16:56:35 2015 -0700 34519 34520 glamor: Provide a fallback path for using an index buffer to do quads. 34521 34522 Improves x11perf -aa10text performance by 1377.59% +/- 23.8198% (n=93) 34523 on Intel with GLES2. 34524 34525 Signed-off-by: Eric Anholt <eric@anholt.net> 34526 Reviewed-by: Dave Airlie <airlied@redhat.com> 34527 34528commit e8fc929d4a140666420a64d592371f5d3eb10cca 34529Author: Eric Anholt <eric@anholt.net> 34530Date: Fri Jun 19 17:26:46 2015 -0700 34531 34532 glamor: Use GL_EXT_map_buffer_range if present. 34533 34534 We were only looking for the desktop GL version of the extension, so 34535 GLES2 missed out. 34536 34537 Signed-off-by: Eric Anholt <eric@anholt.net> 34538 Reviewed-by: Dave Airlie <airlied@redhat.com> 34539 34540commit 4fc4cde0ce7d1a35bab3b3bbd787d688cd415d78 34541Author: Eric Anholt <eric@anholt.net> 34542Date: Fri Jun 19 16:34:55 2015 -0700 34543 34544 glamor: Use the normal GL_QUADS drawing helper in the render code. 34545 34546 We use this for all of our other performance-sensitive rendering, too. 34547 34548 Signed-off-by: Eric Anholt <eric@anholt.net> 34549 Reviewed-by: Dave Airlie <airlied@redhat.com> 34550 34551commit c5e6fffbdd88ad714454b9de1ac1c3eb9f6bf196 34552Author: Eric Anholt <eric@anholt.net> 34553Date: Fri Jun 19 17:05:28 2015 -0700 34554 34555 glamor: Drop CloseScreen-time GL resource cleanup code. 34556 34557 These will all be freed when the context is freed. 34558 34559 Signed-off-by: Eric Anholt <eric@anholt.net> 34560 Reviewed-by: Dave Airlie <airlied@redhat.com> 34561 34562commit 3d6dcad00d67bff407385ad37c7efde7b73ba719 34563Author: Eric Anholt <eric@anholt.net> 34564Date: Fri Jun 19 19:04:16 2015 -0700 34565 34566 glamor: Fix text rendering on GLES2. 34567 34568 The GL_QUADS helper takes a number of quads, not a number of vertices. 34569 34570 Signed-off-by: Eric Anholt <eric@anholt.net> 34571 Reviewed-by: Dave Airlie <airlied@redhat.com> 34572 34573commit b4061cf5f76241157b2dc81dec053012075311c0 34574Author: Peter Harris <pharris@opentext.com> 34575Date: Tue May 12 14:19:15 2015 -0400 34576 34577 Fix border tile origin when background is ParentRelative 34578 34579 According to 34580 http://www.x.org/releases/X11R7.7/doc/xproto/x11protocol.html#requests:CreateWindow 34581 "The border tile origin is always the same as the background tile 34582 origin." 34583 34584 ChangeWindowAttributes goes to some effort to make sure it repaints 34585 the border tile whenever the background origin may have changed, but 34586 miPaintWindow was ignoring the background origin. 34587 34588 Found by xts XChangeWindowAttributes-3 34589 34590 Signed-off-by: Peter Harris <pharris@opentext.com> 34591 Reviewed-by: Keith Packard <keithp@keithp.com> 34592 Signed-off-by: Keith Packard <keithp@keithp.com> 34593 34594commit 8a8d51358cea782147bb9b5b2e0902b1e47fcf4a 34595Merge: b22534d8e 13c7d53df 34596Author: Keith Packard <keithp@keithp.com> 34597Date: Mon Jun 29 21:04:53 2015 -0700 34598 34599 Merge remote-tracking branch 'airlied/for-keithp' 34600 34601commit b22534d8e653575196180970292eaade33c090a8 34602Merge: 0a625adee 780a69aff 34603Author: Keith Packard <keithp@keithp.com> 34604Date: Mon Jun 29 21:04:11 2015 -0700 34605 34606 Merge remote-tracking branch 'jwrdecoede/for-keith' 34607 34608commit 0a625adeec465d6c7dcdb8622c53157b4e932bb0 34609Merge: 84128c10b 790311cec 34610Author: Keith Packard <keithp@keithp.com> 34611Date: Mon Jun 29 21:01:34 2015 -0700 34612 34613 Merge remote-tracking branch 'anholt/glamor-next' 34614 34615commit 84128c10bb6d34f78f7a4148b3c640748efe58e1 34616Merge: 9003a3e5c 8c86dd516 34617Author: Keith Packard <keithp@keithp.com> 34618Date: Mon Jun 29 20:59:31 2015 -0700 34619 34620 Merge remote-tracking branch 'whot/for-keith' 34621 34622commit 13c7d53df8dac45ea2a685826cd45a39bcb51657 34623Author: Kenneth Graunke <kenneth@whitecape.org> 34624Date: Mon Jan 26 21:29:23 2015 -0800 34625 34626 modesetting: Implement page flipping support for Present. 34627 34628 Based on code by Keith Packard, Eric Anholt, and Jason Ekstrand. 34629 34630 v2: 34631 - Fix double free and flip_count underrun (caught by Mario Kleiner). 34632 - Don't leak flip_vblank_event on the error_out path (Mario). 34633 - Use the updated ms_flush_drm_events API (Mario, Ken). 34634 34635 v3: Hack around DPMS shenanigans. If all monitors are DPMS off, then 34636 there is no active framebuffer; attempting to pageflip will hit the 34637 error_undo paths, causing us to drmModeRmFB with no framebuffer, 34638 which confuses the kernel into doing full modesets and generally 34639 breaks things. To avoid this, make ms_present_check_flip check that 34640 some CRTCs are enabled and DPMS on. This is an ugly hack that would 34641 get better with atomic modesetting, or some core Present work. 34642 34643 v4: 34644 - Don't do pageflipping if CRTCs are rotated (caught by Jason Ekstrand). 34645 - Make pageflipping optional (Option "PageFlip" in xorg.conf.d), but 34646 enabled by default. 34647 34648 v5: Initialize num_crtcs_on to 0 (caught by Michel Dänzer). 34649 34650 [airlied: took over] 34651 v6: merge async flip support from Mario Kleiner 34652 free sequence after failed vblank queue 34653 handle unflip while DPMS'ed off (Michel) 34654 move flip tracking into its own structure, and 34655 fix up reference counting issues, and add comments. 34656 34657 Signed-off-by: Dave Airlie <airlied@redhat.com> 34658 Acked-by: Eric Anholt <eric@anholt.net> 34659 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 34660 34661commit 4342369e843ca2c6ca0a8c42475893cffb077a54 34662Author: Dave Airlie <airlied@redhat.com> 34663Date: Fri Jun 26 09:20:13 2015 +1000 34664 34665 modesetting: don't try and load dri2 34666 34667 Since we are shipped with the server and the server has it built-in, 34668 don't bother trying to load it. 34669 34670 Don't remove or invert the if statement on purpose as a later 34671 patch adds stuff in here. 34672 34673 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 34674 Reviewed-by: Eric Anholt <eric@anholt.net> 34675 Signed-off-by: Dave Airlie <airlied@redhat.com> 34676 34677commit edec6394a47be590f3133f6a90186356afd00829 34678Author: Dave Airlie <airlied@redhat.com> 34679Date: Fri Jun 26 09:19:37 2015 +1000 34680 34681 modesetting: drop unused struct in vblank.c 34682 34683 this isn't used anywhere here. 34684 34685 Reviewed-by: Eric Anholt <eric@anholt.net> 34686 Signed-off-by: Dave Airlie <airlied@redhat.com> 34687 34688commit bf262b43007050180d336defd48251de80894801 34689Author: Kenneth Graunke <kenneth@whitecape.org> 34690Date: Mon Jan 26 21:25:58 2015 -0800 34691 34692 modesetting: Implement an ms_drm_abort_seq() function. 34693 34694 This is a specialization of ms_drm_abort that matches based on the drm 34695 event queue's sequence number. 34696 34697 Based on code by Keith Packard. 34698 34699 Reviewed-by: Eric Anholt <eric@anholt.net> 34700 Reviewed-by: Dave Airlie <airlied@redhat.com> 34701 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 34702 34703commit b8ccc5d36b0436bac57f07969ff3eb8d479a8511 34704Author: Jason Ekstrand <jason.ekstrand@intel.com> 34705Date: Tue Jan 13 17:00:28 2015 -0800 34706 34707 glamor: Add a new glamor_gbm_bo_from_pixmap() function. 34708 34709 v2: Rebase on commit e0788a03144 (drop "base"). 34710 34711 Reviewed-by: Eric Anholt <eric@anholt.net> 34712 Reviewed-by: Dave Airlie <airlied@redhat.com> 34713 34714commit f2171d0a20d3b3858a282c51c1561432de1b29ef 34715Author: Kenneth Graunke <kenneth@whitecape.org> 34716Date: Tue Jan 6 21:02:44 2015 -0800 34717 34718 modesetting: Make ms_crtc_on non-static. 34719 34720 I want to use this in present.c. 34721 34722 Reviewed-by: Dave Airlie <airlied@redhat.com> 34723 Reviewed-by: Eric Anholt <eric@anholt.net> 34724 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 34725 34726commit f6853baebaec6a68078fa7f90486e9547d59dcbd 34727Author: Kenneth Graunke <kenneth@whitecape.org> 34728Date: Sun Feb 8 17:21:58 2015 -0800 34729 34730 modesetting: Improve the ms_flush_drm_events() API. 34731 34732 Previously, ms_flush_drm_events() returned a boolean value, and it was 34733 very easy to interpret the meaning incorrectly. Now, we return an 34734 integer value. 34735 34736 The possible outcomes of this call are: 34737 - poll() raised an error (formerly TRUE, now -1 - poll's return value) 34738 - poll() said there are no events (formerly TRUE, now 0). 34739 - drmHandleEvent() raised an error (formerly FALSE, now the negative 34740 value returned by drmHandleEvent). 34741 - An event was successfully handled (formerly TRUE, now 1). 34742 34743 The nice part is that this allows you to distinguish errors (< 0), 34744 nothing to do (= 0), and success (1). We no longer conflate errors 34745 with success. 34746 34747 v2: Change ms_present_queue_vblank to < 0 instead of <= 0, fixing an 34748 unintentional behavior change. libdrm may return EBUSY if it's 34749 received EINTR for more than a second straight; just keep retrying 34750 in that case. Suggested by Jasper St. Pierre. 34751 34752 Reviewed-by: Dave Airlie <airlied@redhat.com> 34753 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 34754 34755commit 7328fb3f2b468048faf4ed4c29db720b5bf00b05 34756Author: Dave Airlie <airlied@gmail.com> 34757Date: Wed Jun 10 13:41:02 2015 +1000 34758 34759 modesetting: reverse prime support (v1.1) 34760 34761 This adds support for reverse prime to the modesetting driver. 34762 34763 Reverse prime is where we have two GPUs in the display chain, 34764 but the second GPU can't scanout from the shared pixmap, so needs 34765 an extra copy to the on screen pixmap. 34766 34767 This allows modesetting to support this scenario while still 34768 supporting the USB offload one. 34769 34770 v1.1: 34771 fix comment + ret = bits (Eric) 34772 34773 Reviewed-by: Eric Anholt <eric@anholt.net> 34774 Signed-off-by: Dave Airlie <airlied@redhat.com> 34775 34776commit 1f10ee221edb87561900f5d2b1336db1c4a9f05c 34777Author: Dave Airlie <airlied@gmail.com> 34778Date: Wed Jun 10 13:38:39 2015 +1000 34779 34780 glamor: egl: add function to back a pixmap with a dma-buf. (v1.1) 34781 34782 Rather than create the pixmap, this uses the file descriptor 34783 to change an existing pixmaps backing store. 34784 34785 This is required for reverse prime slaves, where we create 34786 the slave pixmap, then set the backing store. 34787 34788 v1.1: use local pScreen (Eric) 34789 34790 Reviewed-by: Eric Anholt <eric@annholt.net> 34791 Signed-off-by: Dave Airlie <airlied@redhat.com> 34792 34793commit a79fbfd707e1d745b670def08287241c87e11c28 34794Author: Dave Airlie <airlied@gmail.com> 34795Date: Tue Jun 9 17:15:44 2015 +1000 34796 34797 modesetting: add output master support 34798 34799 This allows a glamor enabled master device to have 34800 slave USB devices attached. 34801 34802 Tested with modesetting on SNB + USB. 34803 34804 It relies on the previous patch to export linear 34805 buffers from glamor. 34806 34807 Acked-by: Eric Anholt <eric@anholt.net> 34808 Signed-off-by: Dave Airlie <airlied@redhat.com> 34809 34810commit ea0e4d752b778c1f2132f8d29542f7dabc296415 34811Author: Dave Airlie <airlied@gmail.com> 34812Date: Wed Jun 10 14:46:22 2015 +1000 34813 34814 glamor: add support for allocating linear buffers (v2) 34815 34816 We need this for doing USB offload scenarios using glamor 34817 and modesetting driver. 34818 34819 unfortunately only gbm in mesa 10.6 has support for the 34820 linear API. 34821 34822 v1.1: fix bad define 34823 v2: update the configure.ac test as per amdgpu. (Michel) 34824 set linear bos to external to avoid cache. (Eric) 34825 34826 Reviewed-by: Eric Anholt <eric@anholt.net> 34827 Signed-off-by: Dave Airlie <airlied@redhat.com> 34828 34829commit 5c555605386c76be2f02193d434a4617218b67be 34830Author: Dave Airlie <airlied@redhat.com> 34831Date: Thu Jun 11 14:29:48 2015 +1000 34832 34833 xf86: restrict when we auto add devices as gpu devices. 34834 34835 Michel pointed out I broke Zaphod with the initial auto add 34836 gpu devices change, 34837 34838 Fix this, by only auto adding GPU devices if we are screen 0 34839 and there are no other screens in the layout. Anyone who 34840 wants to assign GPU devices can specify it in the xorg.conf 34841 for this use case. 34842 34843 Signed-off-by: Dave Airlie <airlied@redhat.com> 34844 Tested-by: Michel Dänzer <michel.daenzer@amd.com> 34845 34846commit 5064ee276fe0f6cebc33ea08e57f1de91f092d88 34847Author: Dave Airlie <airlied@redhat.com> 34848Date: Fri Jun 12 15:13:47 2015 +1000 34849 34850 xf86Rotate: remove unused macros. 34851 34852 These macros aren't used anywhere. 34853 34854 v1.1: drop comment (Aaron) 34855 34856 Signed-off-by: Dave Airlie <airlied@redhat.com> 34857 34858commit 8218dadd30eb0c87efbac6db5d8e3862a236b6a4 34859Author: Dave Airlie <airlied@redhat.com> 34860Date: Tue Jun 9 15:30:22 2015 +1000 34861 34862 cursor: drop ARGB_CURSOR 34863 34864 I doubt anyone builds with this turned off or has done for a long 34865 time. 34866 34867 It helps my eyes bleed slightly less when reading the code, I've left 34868 the define in place as some drivers use it. 34869 34870 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 34871 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 34872 Signed-off-by: Dave Airlie <airlied@redhat.com> 34873 34874commit 3c859112d38860764bcdb205ab461d00db502f45 34875Author: Chris Wilson <chris@chris-wilson.co.uk> 34876Date: Tue Jun 23 15:05:31 2015 +0200 34877 34878 xwayland: keep temp files out of the client mask 34879 34880 Xwayland opens anonymous files for its sharing buffers, move these 34881 file descriptors out of the range of the client select mask to avoid 34882 reaching the maximum number of clients prematurely. 34883 34884 https://bugs.freedesktop.org/show_bug.cgi?id=91072 34885 34886 Tested-by: Olivier Fourdan <fourdan@gmail.com> 34887 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 34888 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 34889 34890commit 8c86dd51606486f4794c503b9c05c6875bdc693a 34891Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 34892Date: Tue Jun 23 08:41:57 2015 -0300 34893 34894 configure.ac: fix code misplacing for KDRIVE_{KBD, MOUSE, EVDEV} auto-detection 34895 34896 KDrive input drivers "kbd", "mouse", and "evdev" should be automatically built 34897 for linux systems. However, due to a code misplacing in configure.ac, the 34898 auto-detection test is being performed AFTER the corresponding AC_DEFINEs' calls, 34899 so that the drivers are NOT being built, regardless the test result. 34900 34901 This bug should be affecting previous xorg-server releases. Please consider 34902 backporting this patch to all maintained ones! 34903 34904 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 34905 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 34906 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 34907 34908commit 780a69aff0ff57ef813179a2f09556eaf488cdfc 34909Author: David Herrmann <dh.herrmann@gmail.com> 34910Date: Mon Jun 22 21:13:05 2015 +0200 34911 34912 systemd-logind: do not rely on directed signals 34913 34914 Right now, Xorg does not install DBus matches for "PauseDevice" / 34915 "ResumeDevice". Therefore, it should usually not receive those DBus 34916 signals from logind. It is just a coincidence that systemd-logind sends 34917 those signals in a directed manner right now. Therefore, dbus-daemon 34918 bypasses the broadcast matches. 34919 34920 However, this is not ABI and Xorg should not rely on this. systemd-logind 34921 is free to send those signals as broadcasts, in which case Xorg will 34922 freeze the VT. Fix this by always installing those matches. 34923 34924 Cc: Hans de Goede <hdegoede@redhat.com> 34925 Cc: Keith Packard <keithp@keithp.com> 34926 Reported-by: Jan Alexander Steffens <jan.steffens@gmail.com> 34927 Signed-off-by: David Herrmann <dh.herrmann@gmail.com> 34928 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 34929 Tested-by: Hans de Goede <hdegoede@redhat.com> 34930 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 34931 34932commit d3352d098724e96be8fb8538a3640ce271d8f8b9 34933Author: Hans de Goede <hdegoede@redhat.com> 34934Date: Mon Jun 22 13:51:21 2015 +0200 34935 34936 Xorg.wrap.man: Fix spelling mistake in Xorg.wrap.1 34937 34938 This commit fixes a small mistake in Xorg.wrap.1 . 34939 34940 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 34941 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 34942 34943commit 790311cec30ac3d35e580b9f1266236f558033d4 34944Author: Eric Anholt <eric@anholt.net> 34945Date: Sun May 31 16:08:10 2015 -0700 34946 34947 glamor: Don't try to disable attrib divisors without the extension. 34948 34949 Fixes epoxy assertion failures on vc4. 34950 34951 Signed-off-by: Eric Anholt <eric@anholt.net> 34952 Reviewed-by: Keith Packard <keithp@keithp.com> 34953 34954commit 9c679d06055cc62aa9209318705e87dc33fba4c8 34955Author: Eric Anholt <eric@anholt.net> 34956Date: Sun May 31 16:07:01 2015 -0700 34957 34958 glamor: Skip actual FBO setup in our glyph atlas. 34959 34960 VC4 (and many GLES2 renderers) can't render to GL_ALPHA, so our pixmap 34961 would end up as GLAMOR_MEMORY and our dereference of the FBO would 34962 setfault. Instead, tell the pixmap creation that we don't need an FBO 34963 at all. Our glyph upload path was already glTexImage for non-a1, and 34964 a more general software fallback for a1 (since the glyph is also in 34965 system memory). 34966 34967 Signed-off-by: Eric Anholt <eric@anholt.net> 34968 Reviewed-by: Keith Packard <keithp@keithp.com> 34969 34970commit 077bb1bdea0fa9af846c02896df680293cf9e25c 34971Author: Eric Anholt <eric@anholt.net> 34972Date: Sun May 31 15:46:55 2015 -0700 34973 34974 glamor: Clean up some declarations in glyph rendering. 34975 34976 Signed-off-by: Eric Anholt <eric@anholt.net> 34977 Reviewed-by: Keith Packard <keithp@keithp.com> 34978 34979commit bf7a3bcb00d0dce22ff3fe42fe6236beeab4403b 34980Author: Eric Anholt <eric@anholt.net> 34981Date: Sun May 31 15:28:39 2015 -0700 34982 34983 glamor: Actually allow glyphs of dimension 65 to 128 in the cache. 34984 34985 The cache was trying to allow glyph_max_dim in, but since we were 34986 putting over 64x64 into HW memory, it would end up in the 34987 single-glyph-per-render bail_one path. 34988 34989 Signed-off-by: Eric Anholt <eric@anholt.net> 34990 Reviewed-by: Keith Packard <keithp@keithp.com> 34991 34992commit f4c3cac010cff0980905eeedd8e8d7baf0aff91b 34993Author: Alexander Volkov <a.volkov@rusbitech.ru> 34994Date: Wed Jun 17 12:41:35 2015 +0300 34995 34996 Xi: call UpdateDeviceState() even when the client is missing 34997 34998 The client window can be closed in the middle of a touch sequence, 34999 e.g. Qt 4 closes popup windows on MousePress and Qt 5.5 will do it 35000 on TouchBegin. In this case the state of mouse buttons will not be 35001 updated on TouchEnd because ProcessTouchEvent() calls UpdateDeviceState() 35002 only when the event has been sent to the client. It results in a 35003 stuck left mouse button. 35004 35005 This patch leads to calling UpdateDeviceState() in case the client 35006 can't be found. 35007 35008 Signed-off-by: Alexander Volkov <a.volkov@rusbitech.ru> 35009 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 35010 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35011 35012commit 9003a3e5c55903ce4e371b2d5cb5030b5f97ae0e 35013Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35014Date: Tue Jun 9 23:41:12 2015 -0700 35015 35016 XQuartz: Silence -Wformat-security for NSRunAlertPanel 35017 35018 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35019 35020commit 3a6fa115759c787ec34483437021ad1e55c52423 35021Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35022Date: Tue Jun 9 23:34:50 2015 -0700 35023 35024 XQuartz: Silence -Wunused-function 35025 35026 quartzKeyboard.c:741:1: warning: unused function 'macroman2ucs' [-Wunused-function,Unused Entity Issue] 35027 macroman2ucs(unsigned char c) 35028 ^ 35029 1 warning generated. 35030 35031 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35032 35033commit 9fe7f5ccada37e2d2a2fa92064b92a0334a3fcdd 35034Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35035Date: Tue Jun 9 23:30:50 2015 -0700 35036 35037 XQuartz: Silence -Wunused-variable 35038 35039 X11Controller.m:939:9: warning: unused variable 'remain' [-Wunused-variable,Unused Entity Issue] 35040 int remain; 35041 ^ 35042 35043 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35044 35045commit 0b9c32489131a5723bd78decf5d2557b94207cf4 35046Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35047Date: Tue Jun 9 23:29:40 2015 -0700 35048 35049 XQuartz: Silence -Wpointer-bool-conversion 35050 35051 X11Controller.m:417:17: error: address of function 'asl_log_descriptor' will always evaluate to 'true' 35052 [-Werror,-Wpointer-bool-conversion,Value Conversion Issue] 35053 if (asl_log_descriptor) { 35054 ~~ ^~~~~~~~~~~~~~~~~~ 35055 X11Controller.m:417:17: note: prefix with the address-of operator to silence this warning [Semantic Issue] 35056 if (asl_log_descriptor) { 35057 ^ 35058 & 35059 35060 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35061 35062commit 77611066397747411f348e4a77871da5cff3b71e 35063Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35064Date: Sun Jun 1 04:29:19 2014 -0700 35065 35066 XQuartz: GLX: Use __glXEnableExtension to build extensions list 35067 35068 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 35069 35070commit 990cf5b2828f73dc7a07f1e38f608af39acfd81d 35071Author: Andrew Eikum <aeikum@codeweavers.com> 35072Date: Thu Jun 4 13:24:53 2015 -0500 35073 35074 dix: Send KeyPress and KeyRelease events to the XACE_KEY_AVAIL hook 35075 35076 While it's documented in the XACE spec, the XACE_KEY_AVAIL hook is 35077 currently never actually invoked by the xserver. 35078 35079 This hook was added in 13c6713c82 (25 Aug 2006), but as the keyboard 35080 processing was moved into XKB, the hook was forgotten and silently 35081 dropped. The code calling this hook was removed by 7af53799c (4 Jan 35082 2009), but it was probably already unused before that. 35083 35084 This patch re-adds support for this hook. The "count" hook parameter is 35085 unused. 35086 35087 Signed-off-by: Andrew Eikum <aeikum@codeweavers.com> 35088 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 35089 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35090 35091commit 5dc2a9aae46352f1f35d3ba122da04727789a15d 35092Author: Andrew Eikum <aeikum@codeweavers.com> 35093Date: Thu Jun 4 13:23:19 2015 -0500 35094 35095 xace: Add XaceHookIsSet helper function 35096 35097 Signed-off-by: Andrew Eikum <aeikum@codeweavers.com> 35098 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 35099 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35100 35101commit c88a325899381d1133e63e6c33c29db079e20a87 35102Author: Hans de Goede <hdegoede@redhat.com> 35103Date: Wed Jan 21 10:13:20 2015 +0100 35104 35105 systemd-logind: Only use systemd-logind integration together with keeptty 35106 35107 systemd-logind integration does not work when starting X on a new tty, as 35108 that detaches X from the current session and after hat systemd-logind revokes 35109 all rights any already open fds and refuses to open new fds for X. 35110 35111 This means that currently e.g. "startx -- vt7" breaks, and breaks badly, 35112 requiring ssh access to the system to kill X. 35113 35114 The fix for this is easy, we must not use systemd-logind integration when 35115 not using KeepTty, or iow we may only use systemd-logind integration together 35116 with KeepTty. 35117 35118 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 35119 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 35120 Tested-by: Aaron Plattner <aplattner@nvidia.com> 35121 35122commit 81bcada14e339fe2a2fb3f3a040566d94dc20bad 35123Author: Hans de Goede <hdegoede@redhat.com> 35124Date: Wed May 13 13:17:09 2015 +0200 35125 35126 linux: Add a may_fail paramter to linux_parse_vt_settings 35127 35128 linux_parse_vt_settings() was split out of xf86OpenConsole so that it can 35129 be called earlier during systemd-logind init, but it is possible to run 35130 the xserver in such a way that xf86OpenConsole() is never used. 35131 35132 The FatalError calls in linux_parse_vt_settings() may stop the Xorg xserver 35133 from working when e.g. no /dev/tty0 is present in such a setup. 35134 35135 This commit adds a may_fail parameter to linux_parse_vt_settings() which 35136 can be used to make linux_parse_vt_settings() fail silenty with an error 35137 return in this case, rather then calling FatalError(). 35138 35139 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 35140 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 35141 35142commit 1dba5a0b194653b0effb01a918bd7338b0c6bcb9 35143Author: Hans de Goede <hdegoede@redhat.com> 35144Date: Wed Jan 21 09:31:39 2015 +0100 35145 35146 linux: Add linux_parse_vt_settings and linux_get_keeptty helpers 35147 35148 systemd-logind integration does not work when starting X on a new tty, as 35149 that detaches X from the current session and after hat systemd-logind revokes 35150 all rights on any already open fds and refuses to open new fds for X. 35151 35152 This means that currently e.g. "startx -- vt7" breaks, and breaks badly, 35153 requiring ssh access to the system to kill X. 35154 35155 The fix for this is easy, we must not use systemd-logind integration when 35156 not using KeepTty, or iow we may only use systemd-logind integration together 35157 with KeepTty. 35158 35159 But the final KeepTty value is not known until the code to chose which vtno to 35160 run on has been called, which currently happens after intializing 35161 systemd-logind. 35162 35163 This commit is step 1 in fixing the "startx -- vt7" breakage, it factors out 35164 the linux xf86OpenConsole bits which set xf86Info.vtno and keepTty so that 35165 these can be called earlier. Calling this earlier is safe as this code has 35166 no side effects other than setting xf86Info.vtno and keepTty. 35167 35168 Note this basically only moves a large chunk of xf86OpenConsole() into 35169 linux_parse_vt_settings() without changing a single line of it, this is 35170 hard to see in the diff because the identation level has changed. 35171 35172 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 35173 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 35174 35175commit fa12f2c150b2f50de9dac4a2b09265f13af353af 35176Author: Dave Airlie <airlied@redhat.com> 35177Date: Thu May 28 16:21:37 2015 +1000 35178 35179 glamor: don't do render ops with matching source/dest (v3) 35180 35181 XRender defines this, GL really doesn't like it. 35182 35183 kwin 4.x and qt 4.x seem to make this happen for the 35184 gradient in the titlebar, and on radeonsi/r600 hw 35185 this draws all kinds of wrong. 35186 35187 v2: bump this up a level, and check it earlier. 35188 (I assume the XXXX was for this case.) 35189 v3: add same code to largepixmap paths (Keith) 35190 35191 Reviewed-by: Eric Anholt <eric@anholt.net> 35192 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 35193 Signed-off-by: Dave Airlie <airlied@redhat.com> 35194 35195commit cbb7eb73b5399e31a7afb800363504d539df0ecf 35196Author: Rui Matos <tiagomatos@gmail.com> 35197Date: Wed May 27 12:08:45 2015 +0200 35198 35199 xwayland: Throttle our cursor surface updates with a frame callback 35200 35201 In some extreme cases with animated cursors at a high frame rate we 35202 could end up filling the wl_display outgoing buffer and end up with 35203 wl_display_flush() failing. 35204 35205 In any case, using the frame callback to throttle ourselves is the 35206 right thing to do. 35207 35208 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 35209 Reviewed-by: Daniel Stone <daniels@collabora.com> 35210 Signed-off-by: Keith Packard <keithp@keithp.com> 35211 35212commit 806470b9f623089dc81b985f250f0c3a4e8edbe8 35213Author: Chris Wilson <chris@chris-wilson.co.uk> 35214Date: Fri Feb 6 08:25:42 2015 +0000 35215 35216 present: Copy unflip contents back to the Screen Pixmap 35217 35218 As we unflip after the flip Window no longer passes the pixel ownership 35219 test for the full Screen Pixmap, we can no longer utilize that Window to 35220 copy the contents back to the backing pixmap. To first flip means that 35221 the Window was originally backed by the Screen Pixmap and wholly covered 35222 the Pixmap, thus we need to copy the last frame contents to the Screen 35223 Pixmap when the flip chain is complete. 35224 35225 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 35226 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 35227 35228commit baa50f60acd9e9f4293107435645ab072b6110e1 35229Author: Vicente Olivert Riera <Vincent.Riera@imgtec.com> 35230Date: Mon Jan 12 17:10:02 2015 +0000 35231 35232 backtrace.c: Fix word cast to a pointer 35233 35234 backtrace.c uses a word size provided by libunwind. In some 35235 architectures like MIPS, libunwind makes that word size 64-bit for all 35236 variants of the architecture. 35237 35238 In the lines #90 and #98, backtrace.c tries to do a cast to a pointer, 35239 which fails in all MIPS variants with 32-bit pointers, like MIPS32 or 35240 MIPS64 n32, because it's trying to do a cast from a 64-bit wide variable 35241 to a 32-bit pointer: 35242 35243 Making all in os 35244 make[2]: Entering directory 35245 `/home/test/test/1/output/build/xserver_xorg-server-1.15.1/os' 35246 CC WaitFor.lo 35247 CC access.lo 35248 CC auth.lo 35249 CC backtrace.lo 35250 backtrace.c: In function 'xorg_backtrace': 35251 backtrace.c:90:20: error: cast to pointer from integer of different size 35252 [-Werror=int-to-pointer-cast] 35253 if (dladdr((void *)(pip.start_ip + off), &dlinfo) && 35254 dlinfo.dli_fname && 35255 ^ 35256 backtrace.c:98:13: error: cast to pointer from integer of different size 35257 [-Werror=int-to-pointer-cast] 35258 (void *)(pip.start_ip + off)); 35259 ^ 35260 cc1: some warnings being treated as errors 35261 make[2]: *** [backtrace.lo] Error 1 35262 make[2]: *** Waiting for unfinished jobs.... 35263 35264 Making the cast to a pointer-sized integer, and then to a pointer fixes 35265 the problem. 35266 35267 Related: 35268 https://bugs.freedesktop.org/show_bug.cgi?id=79939 35269 35270 Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com> 35271 Reviewed-by: Keith Packard <keithp@keithp.com> 35272 Signed-off-by: Keith Packard <keithp@keithp.com> 35273 35274commit 76636ac12f2d1dbdf7be08222f80e7505d53c451 35275Author: Ray Strode <rstrode@redhat.com> 35276Date: Tue May 5 16:43:44 2015 -0400 35277 35278 xwayland: default to local user if no xauth file given. [CVE-2015-3164 3/3] 35279 35280 Right now if "-auth" isn't passed on the command line, we let 35281 any user on the system connect to the Xwayland server. 35282 35283 That's clearly suboptimal, given Xwayland is generally designed 35284 to be used by one user at a time. 35285 35286 This commit changes the behavior, so only the user who started the 35287 X server can connect clients to it. 35288 35289 Signed-off-by: Ray Strode <rstrode@redhat.com> 35290 Reviewed-by: Daniel Stone <daniels@collabora.com> 35291 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 35292 Signed-off-by: Keith Packard <keithp@keithp.com> 35293 35294commit 4b4b9086d02b80549981d205fb1f495edc373538 35295Author: Ray Strode <rstrode@redhat.com> 35296Date: Tue May 5 16:43:43 2015 -0400 35297 35298 os: support new implicit local user access mode [CVE-2015-3164 2/3] 35299 35300 If the X server is started without a '-auth' argument, then 35301 it gets started wide open to all local users on the system. 35302 35303 This isn't a great default access model, but changing it in 35304 Xorg at this point would break backward compatibility. 35305 35306 Xwayland, on the other hand is new, and much more targeted 35307 in scope. It could, in theory, be changed to allow the much 35308 more secure default of a "user who started X server can connect 35309 clients to that server." 35310 35311 This commit paves the way for that change, by adding a mechanism 35312 for DDXs to opt-in to that behavior. They merely need to call 35313 35314 LocalAccessScopeUser() 35315 35316 in their init functions. 35317 35318 A subsequent commit will add that call for Xwayland. 35319 35320 Signed-off-by: Ray Strode <rstrode@redhat.com> 35321 Reviewed-by: Daniel Stone <daniels@collabora.com> 35322 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 35323 Signed-off-by: Keith Packard <keithp@keithp.com> 35324 35325commit c4534a38b68aa07fb82318040dc8154fb48a9588 35326Author: Ray Strode <rstrode@redhat.com> 35327Date: Tue May 5 16:43:42 2015 -0400 35328 35329 xwayland: Enable access control on open sockets [CVE-2015-3164 1/3] 35330 35331 Xwayland currently allows wide-open access to the X sockets 35332 it listens on, ignoring Xauth access control. 35333 35334 This commit makes sure to enable access control on the sockets, 35335 so one user can't snoop on another user's X-over-wayland 35336 applications. 35337 35338 Signed-off-by: Ray Strode <rstrode@redhat.com> 35339 Reviewed-by: Daniel Stone <daniels@collabora.com> 35340 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 35341 Signed-off-by: Keith Packard <keithp@keithp.com> 35342 35343commit ad02d0df75318660c3f7cd6063eac409327fe560 35344Author: Peter Hutterer <peter.hutterer@who-t.net> 35345Date: Wed May 6 15:25:28 2015 +1000 35346 35347 test: add tests for new valuator mask features 35348 35349 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35350 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 35351 35352commit da10d0cb4283ad5a9c4290555751f61795d11b49 35353Author: Peter Hutterer <peter.hutterer@who-t.net> 35354Date: Tue May 5 14:48:41 2015 +1000 35355 35356 dix: hook up the unaccelerated valuator masks 35357 35358 If present, access the unaccelerated valuator mask values for DGA and XI2 raw 35359 events. 35360 35361 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35362 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 35363 35364commit 4c2f2cb4c8ca1ab894a65828fdd39aea9b014f69 35365Author: Peter Hutterer <peter.hutterer@who-t.net> 35366Date: Tue May 5 14:18:54 2015 +1000 35367 35368 dix: Add unaccelerated valuators to the ValuatorMask 35369 35370 Allows a mask to carry both accelerated and unaccelerated motion at the same 35371 time. 35372 35373 This is required for xf86-input-libinput where the pointer acceleration 35374 happens in libinput already, but parts of the server, specifically raw events 35375 and DGA rely on device-specific unaccelerated data. 35376 35377 To ease integration add this as a second set to the ValuatorMask rather than 35378 extending all APIs to carry a second, possibly NULL set of valuators. 35379 35380 Note that a valuator mask should only be used in either accel/unaccel or 35381 standard mode at any time. Switching requires either a valuator_mask_zero() 35382 call or unsetting all valuators one-by-one. Trying to mix the two will produce 35383 a warning. 35384 35385 The server has a shortcut for changing a mask with the 35386 valuator_mask_drop_unaccelerated() call. This saves us from having to loop 35387 through all valuators on every event, we can just drop the bits we know we 35388 don't want. 35389 35390 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35391 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 35392 35393commit 3f0d3201f38ef9d1651fcaf94e45c640786edcc0 35394Author: Peter Hutterer <peter.hutterer@who-t.net> 35395Date: Mon Apr 13 13:52:28 2015 -0700 35396 35397 dix: fix indentation 35398 35399 from 9ff89a2e469ab0ac5af57d0fc115127feb1c0d99 35400 35401 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35402 35403commit f82dc6bd91a545a844fb688ba3552e4f5df8a0d2 35404Author: Peter Hutterer <peter.hutterer@who-t.net> 35405Date: Tue Mar 24 09:43:17 2015 +1000 35406 35407 xfree86: drop if 0 hunk 35408 35409 Disabled in 2005. bye bye. 35410 35411 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35412 35413commit fc59c8fe8d941b0ec1e98c59bc57b1f97dba149d 35414Author: Peter Hutterer <peter.hutterer@who-t.net> 35415Date: Tue May 5 15:54:20 2015 +1000 35416 35417 config: remove 10-evdev.conf, let the evdev driver install that file 35418 35419 Now that we have two likely drivers that fight for control over the input 35420 devices (evdev and libinput) let's move the respective driver assignment to 35421 each package. So you get what you install. 35422 35423 https://bugs.freedesktop.org/show_bug.cgi?id=89023 35424 35425 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35426 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 35427 35428commit bf6344e1913a5d24c2d68eaca999ea3d71e1b707 35429Author: Keith Packard <keithp@keithp.com> 35430Date: Thu May 14 14:58:29 2015 -0700 35431 35432 Revert "glx/dri2: Disable AIGLX if indirect GLX is disabled" 35433 35434 This reverts commit d61ae18074e53c2cdfb13cc37693b526160d6ca7. 35435 35436 If the DRI2 provider is disabled, then we don't advertise the correct 35437 GLX extensions to clients, and things like GLES fail. 35438 35439 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90442 35440 Reviewed-by: Eric Anholt <eric@anholt.net> 35441 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 35442 35443commit b0d2e010316d710eb4052963de3a1e2dc7ba356e 35444Author: Keith Packard <keithp@keithp.com> 35445Date: Fri Oct 10 09:25:51 2014 +0200 35446 35447 glamor: Replace CompositeGlyphs code [v2] 35448 35449 New composite glyphs code uses the updated glamor program 35450 infrastructure to create efficient shaders for drawing render text. 35451 35452 Glyphs are cached in two atlases (one 8-bit, one 32-bit) in a simple 35453 linear fashion. When the atlas fills, it is discarded and a new one 35454 constructed. 35455 35456 v2: Eric Anholt changed the non-GLSL 130 path to use quads instead of 35457 two triangles for a significant performance improvement on hardware 35458 with quads. Someone can fix the GLES quads emulation if they want to 35459 make it faster there. 35460 35461 v3: Eric found more dead code to delete 35462 35463 Signed-off-by: Keith Packard <keithp@keithp.com> 35464 Reviewed-by: Eric Anholt <eric@anholt.net> 35465 35466commit 1b745e0c1ff45e014aa21c3d8edf93227bec99bf 35467Author: Keith Packard <keithp@keithp.com> 35468Date: Mon Oct 13 11:40:06 2014 -0700 35469 35470 glamor: Adapt glamor_program API to handle render acceleration 35471 35472 This extends the existing API to support options needed for render 35473 accleration, including an additional fragment, 'combine', (which 35474 provides a place to perform the source IN mask operation before the 35475 final OP dest state) and an additional 'defines' parameter which 35476 provides a way to add target-dependent values without using a uniform. 35477 35478 Signed-off-by: Keith Packard <keithp@keithp.com> 35479 Reviewed-by: Eric Anholt <eric@anholt.net> 35480 35481commit ff3195aadde95c8e89f77f389a7dfb418dd2426c 35482Author: Keith Packard <keithp@keithp.com> 35483Date: Mon Oct 20 21:31:56 2014 -0700 35484 35485 glamor: Compute GLSL version from GL_SHADING_LANGUAGE_VERSION (v3) 35486 35487 Use code from Piglit project to compute GLSL version for either GL or 35488 GLES. The Piglit code was originally written by Chad Versace. 35489 35490 v2: bail if the parse fails (requested by Eric Anholt) 35491 v3: Use version 1.20 for GLES until we fix our programs (Eric Anholt) 35492 35493 Signed-off-by: Keith Packard <keithp@keithp.com> 35494 Reviewed-by: Eric Anholt <eric@anholt.net> 35495 35496commit dcb3d74ba8861e7b0a592e92b5b2247b84e843f3 35497Author: Keith Packard <keithp@keithp.com> 35498Date: Mon Oct 13 12:35:40 2014 -0700 35499 35500 glamor: Remove destination drawable argument from glamor_set_texture 35501 35502 This argument wasn't used at all. 35503 35504 Signed-off-by: Keith Packard <keithp@keithp.com> 35505 Reviewed-by: Eric Anholt <eric@anholt.net> 35506 35507commit 2bf34fe8d9b7628d164392c2d11ace78f7cf17b9 35508Author: Keith Packard <keithp@keithp.com> 35509Date: Mon Oct 13 12:32:27 2014 -0700 35510 35511 glamor: Pass depth to glamor_pm_is_solid and glamor_set_planemask 35512 35513 Instead of passing the destination drawable, just pass the depth, as 35514 the underlying functions need only that to check whether the planemask 35515 is going to work. 35516 35517 This API change will allow higher level functions to not need the 35518 destination pixmap. 35519 35520 Signed-off-by: Keith Packard <keithp@keithp.com> 35521 Reviewed-by: Eric Anholt <eric@anholt.net> 35522 35523commit 910ddf85219f114744e8996a4ac044c4eafc62ac 35524Author: Egbert Eich <eich@freedesktop.org> 35525Date: Tue May 12 09:52:48 2015 -0700 35526 35527 Xephyr: Fix broken image when endianess of client machine and host-Xserver differ 35528 35529 The image is created in the native byte order of the machine Xephyr is 35530 rendered on however drawn in the image byte order of the Xephyr server. 35531 Correct byte order in the xcb_image_t structure and convert to native 35532 before updating the window. 35533 If depths of Xephyr and host server differ this is already taken care of 35534 by the depth conversion routine. 35535 It is a terrible wase to always convert and transmit the entire image 35536 no matter of the size of the damaged area. One should probably use 35537 sub-images here. For now we leave this as an exercise. 35538 35539 Signed-off-by: Egbert Eich <eich@freedesktop.org> 35540 Reviewed-by: Keith Packard <keithp@keithp.com> 35541 Signed-off-by: Keith Packard <keithp@keithp.com> 35542 35543commit c65eda5e6676d942e80eaf2650a670174c8bd84a 35544Author: Egbert Eich <eich@freedesktop.org> 35545Date: Tue Mar 31 09:14:28 2015 +0200 35546 35547 Xephyr: Fix screen image draw for the non-Glamor & non-XHSM case 35548 35549 xcb_image_put() prints the entire image, therefore don't use an offset. 35550 35551 Signed-off-by: Egbert Eich <eich@freedesktop.org> 35552 Reviewed-by: Keith Packard <keithp@keithp.com> 35553 Signed-off-by: Keith Packard <keithp@keithp.com> 35554 35555commit 66212ca0d2f194fd16db65e863f0a2d613e180ea 35556Author: Egbert Eich <eich@freedesktop.org> 35557Date: Tue Mar 31 09:14:27 2015 +0200 35558 35559 Xephyr: Fix compile when debugging is enabled 35560 35561 Signed-off-by: Egbert Eich <eich@freedesktop.org> 35562 Reviewed-by: Keith Packard <keithp@keithp.com> 35563 Signed-off-by: Keith Packard <keithp@keithp.com> 35564 35565commit b536d56aef21739b6da44693bbf19d0e7541392d 35566Author: Egbert Eich <eich@freedesktop.org> 35567Date: Tue Mar 31 09:14:26 2015 +0200 35568 35569 Xephyr: Print default server display number if none is specified 35570 35571 Signed-off-by: Egbert Eich <eich@freedesktop.org> 35572 Reviewed-by: Keith Packard <keithp@keithp.com> 35573 Signed-off-by: Keith Packard <keithp@keithp.com> 35574 35575commit 5af73f490870da9265eeb9b3ce59a2be026be0c8 35576Author: Egbert Eich <eich@suse.de> 35577Date: Tue Mar 31 09:14:25 2015 +0200 35578 35579 Xephyr: Don't crash when no command line argument is specified 35580 35581 The DDX specific command line parsing function only gets called 35582 if command line arguments are present. Therefore this function 35583 is not suitable to initialize mandatory global variables. 35584 Replace main() instead. 35585 35586 Signed-off-by: Egbert Eich <eich@freedesktop.org> 35587 Reviewed-by: Keith Packard <keithp@keithp.com> 35588 Signed-off-by: Keith Packard <keithp@keithp.com> 35589 35590commit 00f79416b19f0cde68291aced44ab07b9b76f7b8 35591Author: Jonathan Gray <jsg@jsg.id.au> 35592Date: Wed Apr 15 21:29:58 2015 +1000 35593 35594 glamor: fix build when DRI3 is not defined 35595 35596 Signed-off-by: Jonathan Gray <jsg@jsg.id.au> 35597 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 35598 Signed-off-by: Keith Packard <keithp@keithp.com> 35599 35600commit 7c609c911a3a33b7e4ddad46b8fc42878a073ee7 35601Author: Jonathan Gray <jsg@jsg.id.au> 35602Date: Wed Apr 15 21:29:07 2015 +1000 35603 35604 glamor: remove const from the return type of glamor_get_drawable_location() 35605 35606 Fixes a build error with gcc 4.2.1 on OpenBSD due to 35607 -Werror=return-type from xorg-macros. 35608 35609 error: type qualifiers ignored on function return type 35610 35611 Signed-off-by: Jonathan Gray <jsg@jsg.id.au> 35612 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 35613 Signed-off-by: Keith Packard <keithp@keithp.com> 35614 35615commit 145ae03814cb3b700b6fe1fd19f8fb15da84d1c8 35616Author: Michel Dänzer <michel.daenzer@amd.com> 35617Date: Tue Mar 17 10:21:13 2015 +0900 35618 35619 modesetting: Include dix-config.h from dumb_bo.c 35620 35621 Fixes mmap failures with 32-bit builds. 35622 35623 Signed-off-by: Michel Dänzer <michel@daenzer.net> 35624 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 35625 Signed-off-by: Keith Packard <keithp@keithp.com> 35626 35627commit 4962c8c08842d9d3ca66d254b1ce4cacc4fb3756 35628Author: Michel Dänzer <michel.daenzer@amd.com> 35629Date: Tue Mar 17 10:21:12 2015 +0900 35630 35631 Add AC_SYS_LARGEFILE defines to dix-config.h 35632 35633 Without this, AC_SYS_LARGEFILE doesn't actually have any effect. 35634 35635 Signed-off-by: Michel Dänzer <michel@daenzer.net> 35636 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 35637 Signed-off-by: Keith Packard <keithp@keithp.com> 35638 35639commit 0409b6e6d63e9cfb5dc71bb27de4b1ed0152dd9b 35640Merge: c39c3a975 23702dd26 35641Author: Keith Packard <keithp@keithp.com> 35642Date: Mon May 11 16:50:43 2015 -0700 35643 35644 Merge remote-tracking branch 'evelikov/master' 35645 35646commit c39c3a97508dc384c0757a0990c07b5d7b2fe97a 35647Merge: 6b65e9618 747057852 35648Author: Keith Packard <keithp@keithp.com> 35649Date: Mon May 11 16:34:48 2015 -0700 35650 35651 Merge remote-tracking branch 'ajax/xserver-next' 35652 35653commit 6b65e961894b9ed53066d22cfd218b12c3f361ca 35654Author: Robert Ancell <robert.ancell@canonical.com> 35655Date: Wed May 6 15:07:21 2015 +1200 35656 35657 xwayland: Fix error strings 35658 35659 Fix missing newlines from error string and fix grammar. 35660 35661 Signed-off-by: Robert Ancell <robert.ancell@canonical.com> 35662 Reviewed-by: Daniel Stone <daniels@collabora.com> 35663 Signed-off-by: Keith Packard <keithp@keithp.com> 35664 35665commit d7091a21d90cf463ae39ec5e8741123218ec5686 35666Merge: c3ce9d8fd 8fb8bbb30 35667Author: Keith Packard <keithp@keithp.com> 35668Date: Mon May 11 15:49:34 2015 -0700 35669 35670 Merge remote-tracking branch 'airlied/for-keithp' 35671 35672commit c3ce9d8fd404ca1d4697a104ce4899525dd43c51 35673Author: Adel Gadllah <adel.gadllah@gmail.com> 35674Date: Fri May 1 17:43:41 2015 +0200 35675 35676 modesetting: Fix software cursor fallback 35677 35678 The code in drmmode_set_cursor does not properly handle the case where 35679 drmModeSetCursor2 returns any other error than EINVAL and silently fails to set 35680 a cursor. 35681 35682 So only return when the drmModeSetCursor2 succeeds (i.e returns 0) and disable 35683 the cursor2 usage on EINVAL. 35684 35685 References: https://bugzilla.redhat.com/show_bug.cgi?id=1205725 35686 Signed-off-by: Adel Gadllah <adel.gadllah@gmail.com> 35687 Reviewed-by: Michel Dänzer <michel@daenzer.net> 35688 Reviewed-by: Rob Clark <robdclark@gmail.com> 35689 Signed-off-by: Keith Packard <keithp@keithp.com> 35690 35691commit 81a51a6cace6fdb54999ccdf1533dc28a2222bb9 35692Author: Dima Ryazanov <dima@gmail.com> 35693Date: Wed Apr 29 22:39:18 2015 -0700 35694 35695 xwayland: Implement smooth scrolling 35696 35697 We don't even need to simulate button clicks; it's done automatically. 35698 This also fixes scrolling in Qt5 apps. 35699 35700 Signed-off-by: Dima Ryazanov <dima@gmail.com> 35701 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 35702 Signed-off-by: Keith Packard <keithp@keithp.com> 35703 35704commit 9ff89a2e469ab0ac5af57d0fc115127feb1c0d99 35705Author: Jason Gerecke <killertofu@gmail.com> 35706Date: Wed Apr 29 15:08:58 2015 +1000 35707 35708 dix: Do not allow device transform to be set on valuatorless devices 35709 35710 If a device does not have any valuators, it makes no sense to set the 35711 device transformation. Return a BadMatch error to let the caller know 35712 that they're trying something stupid. 35713 35714 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 35715 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 35716 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 35717 Signed-off-by: Keith Packard <keithp@keithp.com> 35718 35719commit fa62ca68ef8be7f63cd0b899dfae3366ca9350cf 35720Author: Rui Matos <tiagomatos@gmail.com> 35721Date: Sat Apr 25 21:19:25 2015 +0200 35722 35723 dix/events: Set currentTime to the given time stamp in NoticeTime 35724 35725 The refactoring in commit efc1035ca958f2c9d266338a308518a0834b1773 35726 removed the actual update of currentTime. 35727 35728 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 35729 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 35730 Reviewed-by: Keith Packard <keithp@keithp.com> 35731 Signed-off-by: Keith Packard <keithp@keithp.com> 35732 35733commit 26e50e8b2cbd01d050b5ecc02d47488b53ecc08b 35734Merge: 28ff661e7 c7b49bdbb 35735Author: Keith Packard <keithp@keithp.com> 35736Date: Mon May 11 15:36:53 2015 -0700 35737 35738 Merge remote-tracking branch 'jturney/mingw-build-fixes' 35739 35740commit 7470578520e90b6402b2509cd0c51fd4fd84849f 35741Author: Brent Collins <bcollins@trustedcs.com> 35742Date: Thu May 7 15:10:19 2015 -0400 35743 35744 shm: Fix xselinux resource initialization for xinerama pixmaps 35745 35746 This is necessary to avoid a NULL pointer deference when the pixmap is 35747 used later. 35748 35749 [ajax: massaged commit message, fixed it to compile] 35750 35751 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89748 35752 Reviewed-by: Adam Jackson <ajax@redhat.com> 35753 Signed-off-by: Brent Collins <bcollins@trustedcs.com> 35754 35755commit bcec9f867d19e954a46c2654a79782bff6c65fce 35756Author: Chris Wilson <chris@chris-wilson.co.uk> 35757Date: Sat Feb 14 10:36:44 2015 +0000 35758 35759 shm: Fix use-after-free in ShmDestroyPixmap 35760 35761 We pass the pPixmap->drawable.id to the ShmDetachSegment function after 35762 the pPixmap is freed. Fortunately, we don't use the value inside 35763 ShmDetachSegment and can simply pass zero instead. 35764 35765 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 35766 Reviewed-by: Michel Dänzer <michel@daenzer.net> 35767 35768commit d61ae18074e53c2cdfb13cc37693b526160d6ca7 35769Author: Chris Wilson <chris@chris-wilson.co.uk> 35770Date: Wed Mar 4 12:16:29 2015 +0000 35771 35772 glx/dri2: Disable AIGLX if indirect GLX is disabled 35773 35774 There is no point in setting up the acceleration for indirect GLX if 35775 indirect GLX is itself disabled. 35776 35777 Reviewed-by: Adam Jackson <ajax@redhat.com> 35778 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 35779 35780commit 1af15aaf278edcf6f6de94774350e34a80883c24 35781Author: Olivier Fourdan <ofourdan@redhat.com> 35782Date: Fri Apr 10 10:07:38 2015 +0200 35783 35784 dix: Fix image byte order on big endian hardware 35785 35786 Make sure X_BIG_ENDIAN/X_LITTLE_ENDIAN are defined before actually using 35787 them. 35788 35789 Otherwise, image byte order could be wrong on big endian hardware even 35790 though endianess detection is correct. 35791 35792 Reported-by: Tim Waugh <twaugh@redhat.com> 35793 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 35794 Reviewed-by: Adam Jackson <ajax@redhat.com> 35795 35796commit 28159eff6badf6181b255f26d1f444abe81c05b7 35797Author: Jason Gerecke <killertofu@gmail.com> 35798Date: Thu Apr 30 18:06:14 2015 -0700 35799 35800 xfree86: Return NULL from xf86CompatOutput if no compat_output is defined 35801 35802 If no compat_output is defined, we inadvertently (attempt to) return 35803 whatever data is at index -1. Instead, return NULL since that's what 35804 callers are expecting. 35805 35806 Reviewed-by: Adam Jackson <ajax@redhat.com> 35807 Signed-off-by: Jason Gerecke <jason.gerecke@wacom.com> 35808 35809commit cad831f39861435cb94f4c90a73f7f0b62d1a4a1 35810Author: Robert Ancell <robert.ancell@canonical.com> 35811Date: Wed May 6 15:07:21 2015 +1200 35812 35813 xwayland: Fix error strings 35814 35815 Fix missing newlines from error string and fix grammar. 35816 35817 Signed-off-by: Robert Ancell <robert.ancell@canonical.com> 35818 Reviewed-by: Daniel Stone <daniels@collabora.com> 35819 35820commit 8fb8bbb3062f1a06621ab7030a9e89d5e8367b35 35821Author: Dave Airlie <airlied@redhat.com> 35822Date: Mon Mar 23 11:33:23 2015 +1000 35823 35824 modesetting: add tile property support (v2.1) 35825 35826 This adds tiling support to the server modesetting driver, 35827 it retrieves the tile info from the kernel and translates 35828 it into the server format and exposes the property. 35829 35830 v2.1: fix resetting tile property (Chris) 35831 35832 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 35833 Signed-off-by: Dave Airlie <airlied@redhat.com> 35834 35835commit 9257b1252da9092ddc676fec9aabe2b33dfad272 35836Author: Dave Airlie <airlied@redhat.com> 35837Date: Wed Jan 28 17:35:21 2015 +1000 35838 35839 modesetting: add dynamic connector hotplug support (MST) (v3) 35840 35841 This is ported from the same code in the ati and intel drivers, 35842 35843 It uses the same option name as nvidia and the other DDXes to 35844 disable tearing down outputs as it is hard to avoid racing with clients. 35845 35846 v2: address two issues with DeleteUnusedDP12 enabled, reported 35847 by Daniel Martin, 35848 a) check we have a mode_output before destroying it 35849 b) only delete *unused* displays (thanks Aaron for clarifying) 35850 so we check if the output has a crtc and if it does we don't 35851 delete it. 35852 35853 v3: drop the option to delete unused displays, just encode 35854 behaviour into the randr spec. 35855 35856 Signed-off-by: Dave Airlie <airlied@redhat.com> 35857 35858commit 33422d160bff3117b413a62d82b168e84f1aa8f6 35859Author: Dave Airlie <airlied@redhat.com> 35860Date: Wed Jan 28 17:25:00 2015 +1000 35861 35862 modesetting: stop caching mode resources 35863 35864 There is no need to cache the mode resources and with dynamic 35865 connectors for mst support we don't want to. So first clean that 35866 up before adding dynamic connector support. 35867 35868 Reviewed-by: Keith Packard <keithp@keithp.com> 35869 Signed-off-by: Dave Airlie <airlied@redhat.com> 35870 35871commit a9ac02f6949357619684dd98ff6cf05489e62e55 35872Author: Dave Airlie <airlied@redhat.com> 35873Date: Wed Apr 1 12:13:51 2015 +1000 35874 35875 xf86Crtc/monitors: create initial monitors for tiled outputs 35876 35877 This creates an automatic monitor for a tiled monitor at startup. 35878 35879 Reviewed-by: Keith Packard <keithp@keithp.com> 35880 Signed-off-by: Dave Airlie <airlied@redhat.com> 35881 35882commit afd18bce6a81106a12fd750d5fa09d05c09d37a8 35883Author: Dave Airlie <airlied@redhat.com> 35884Date: Wed Apr 1 14:32:00 2015 +1000 35885 35886 xf86Crtc: setup tiled monitors correctly in right of 35887 35888 This puts the tiles of the monitor in the right place at 35889 X server startup. 35890 35891 Reviewed-by: Keith Packard <keithp@keithp.com> 35892 Signed-off-by: Dave Airlie <airlied@redhat.com> 35893 35894commit e472dd89420f671685c11b06d376ff146d54c3b8 35895Author: Adam Jackson <ajax@redhat.com> 35896Date: Tue Jul 28 11:07:13 2009 -0400 35897 35898 xf86Crtc: right-of placement by default. 35899 35900 Change the X server default to do right-of placement 35901 at startup. This gives an option to allow drivers to 35902 override this placement, which has been used for server 35903 drivers where both heads are not in the same physical 35904 place. 35905 35906 Been in Fedora for a few years, but for tiled monitors 35907 we really want something along these lines. 35908 35909 This is an ABI break. 35910 35911 Reviewed-by: Keith Packard <keithp@keithp.com> 35912 Signed-off-by: Dave Airlie <airlied@redhat.com> 35913 35914commit 69e4b8e602ecc7b69c75988a447ec5b509b22402 35915Author: Dave Airlie <airlied@redhat.com> 35916Date: Tue Mar 31 16:56:42 2015 +1000 35917 35918 xfree86: attempt to autoconfig gpu slave devices (v3) 35919 35920 This allows us to skip the screen section, the first 35921 Device section will get assigned to the screen, 35922 any remaining ones will get assigned to the GPUDevice 35923 sections for the screen. 35924 35925 v2: fix the skipping unsuitable screen logic (Aaron) 35926 v3: fix segfault if not conf file (me, 5s after sending v2) 35927 35928 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 35929 Signed-off-by: Dave Airlie <airlied@redhat.com> 35930 35931commit 3b6930c5d02d8fc0d22fe7955e1ef2af61727705 35932Author: Dave Airlie <airlied@redhat.com> 35933Date: Tue Mar 31 16:42:36 2015 +1000 35934 35935 xserver: add xorg.conf support for gpu devices. (v2.1) 35936 35937 This allows gpu devices to be specified in xorg.conf Screen sections. 35938 35939 Section "Device" 35940 Driver "intel" 35941 Identifier "intel0" 35942 Option "AccelMethod" "uxa" 35943 EndSection 35944 35945 Section "Device" 35946 Driver "modesetting" 35947 Identifier "usb0" 35948 EndSection 35949 35950 Section "Screen" 35951 Identifier "screen" 35952 Device "intel0" 35953 GPUDevice "usb0" 35954 EndSection 35955 35956 This should allow for easier tweaking of driver options which 35957 currently mess up the GPU device discovery process. 35958 35959 v2: add error handling for more than 4 devices, (Emil) 35960 fixup CONF_ defines to consistency 35961 add MAX_GPUDEVICES define 35962 (yes there is two defines, this is consistent 35963 with everywhere else). 35964 remove braces around slp (Mark Kettenis) 35965 man page fixups (Aaron) 35966 v2.1: fixup whitespace (Aaron) 35967 35968 Reviewed-by: Keith Packard <keithp@keithp.com> 35969 Signed-off-by: Dave Airlie <airlied@redhat.com> 35970 35971commit c7b49bdbb9321fe9a7dc35f47b91cac85516988f 35972Author: Colin Harrison <colin.harrison@virgin.net> 35973Date: Fri Mar 20 14:25:17 2015 +0000 35974 35975 os/utils.c: Fix prototype for Win32TempDir() 35976 35977 xorg/xserver/os/utils.c: In function ‘Win32TempDir’: 35978 xorg/xserver/os/utils.c:1643:1: warning: old-style function definition [-Wold-style-definition] 35979 35980 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 35981 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 35982 35983commit a9b4b7b79682dd367ce26c29aa5dc85807201851 35984Author: Colin Harrison <colin.harrison@virgin.net> 35985Date: Fri Mar 20 14:25:53 2015 +0000 35986 35987 os/utils.c: Don't try to build os_move_fd() for WIN32 35988 35989 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 35990 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 35991 35992commit 2b114d6a516ee584ff89b96b12acf91799b6d677 35993Author: Jon TURNEY <jon.turney@dronecode.org.uk> 35994Date: Fri Mar 20 14:29:58 2015 +0000 35995 35996 hw/xnest: Fix build for MinGW 35997 35998 Include the wrapped windows.h via X11/Xwindows.h before xcb_keysyms.h to avoid 35999 type clashes caused by the unwrapped windows.h that includes. 36000 36001 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36002 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 36003 36004commit 5bf3e5c8326fb81655827dbd4c527f1a2ba09511 36005Author: Jon TURNEY <jon.turney@dronecode.org.uk> 36006Date: Wed Mar 18 13:29:22 2015 +0000 36007 36008 hw/xwin/winclipboard: Link xwinclip with -lpthread 36009 36010 Link xwinclip with -lpthread to fix build for MinGW 36011 36012 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36013 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 36014 36015commit 8363ef2764fe9c6877d1cb21d3ce7b6cf869f00d 36016Author: Colin Harrison <colin.harrison@virgin.net> 36017Date: Fri Mar 20 14:06:49 2015 +0000 36018 36019 os/xdmcp.c: Include Xtrans.h when building for WIN32 36020 36021 Xtrans.h must be included on WIN32 to prototype _XSERVTransWSAStartup() 36022 36023 xserver/os/xdmcp.c: In function ‘get_addr_by_name’: 36024 xserver/os/xdmcp.c:1483:5: error: implicit declaration of function ‘_XSERVTransWSAStartup’ [-Werror=implicit-function-declaration] 36025 36026 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 36027 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36028 36029commit 28ff661e73f4bdf0a9b7c84e70fa0a08fd30e482 36030Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36031Date: Fri Apr 3 19:06:21 2015 -0700 36032 36033 Remove empty stub of $host_cpu case statement in configure.ac 36034 36035 Left behind when commit 5c12399b6c3a8d moved the xorg_bus_* settings 36036 elsewhere. 36037 36038 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36039 Reviewed-by: Thomas Klausner <wiz@NetBSD.org> 36040 36041commit d4e85afac6578d5babc8bf97e00e85b00746fca9 36042Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36043Date: Fri Apr 3 19:01:40 2015 -0700 36044 36045 Accept x86_64 as well as i*86 for $host_cpu in Solaris on x86 36046 36047 Needed when using a compiler that defaults to 64-bit output when 36048 configure is checking for $host_cpu. 36049 36050 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36051 Reviewed-by: Thomas Klausner <wiz@NetBSD.org> 36052 36053commit 0018784cdde19444a8f970bc414796fc2a523a51 36054Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36055Date: Sat Mar 21 17:59:13 2015 -0700 36056 36057 Convert hw/dmx to new *allocarray functions 36058 36059 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36060 Reviewed-by: Matt Turner <mattst88@gmail.com> 36061 36062commit dc5acaa28ab9ed091f087e56046400d63f1f192a 36063Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36064Date: Sat Mar 21 17:30:05 2015 -0700 36065 36066 Convert hw/kdrive to new *allocarray functions 36067 36068 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36069 Reviewed-by: Matt Turner <mattst88@gmail.com> 36070 36071commit ae2dc01cf1a371db69d5deb987f4185e7c3ccedd 36072Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36073Date: Sat Mar 21 17:23:33 2015 -0700 36074 36075 Convert hw/xnest & hw/vfb to new *allocarray functions 36076 36077 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36078 Reviewed-by: Matt Turner <mattst88@gmail.com> 36079 36080commit 4cb1034906eeec8c8442d70918bea0f4eb1f6e44 36081Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36082Date: Sat Mar 21 17:12:06 2015 -0700 36083 36084 Convert hw/xfree86 to new *allocarray functions 36085 36086 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36087 Reviewed-by: Matt Turner <mattst88@gmail.com> 36088 36089commit f59236c2865d22c6f0b2d1ba303213afd10cd02e 36090Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36091Date: Sat Mar 21 15:05:30 2015 -0700 36092 36093 Convert glamor & glx to new *allocarray functions 36094 36095 v2: fixup whitespace 36096 36097 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36098 Reviewed-by: Matt Turner <mattst88@gmail.com> 36099 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 36100 36101commit 7ac280287491fe06127d9fefc504217e21c780e6 36102Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36103Date: Sat Mar 21 14:58:50 2015 -0700 36104 36105 Convert mi & miext to new *allocarray functions 36106 36107 v2: remove now useless parentheses 36108 36109 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36110 Reviewed-by: Matt Turner <mattst88@gmail.com> 36111 36112commit 70f4a0e6bd18055cc9cb6685253bf5e07b125657 36113Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36114Date: Sat Mar 21 14:40:39 2015 -0700 36115 36116 Convert exa & fb to new *allocarray functions 36117 36118 v2: fixup whitespace 36119 36120 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36121 Reviewed-by: Matt Turner <mattst88@gmail.com> 36122 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 36123 36124commit 4fe6b03b97ab8dbb32e4908e46be350d7f7d336f 36125Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36126Date: Sat Mar 21 14:36:07 2015 -0700 36127 36128 Convert XKB to new *allocarray functions 36129 36130 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36131 Reviewed-by: Matt Turner <mattst88@gmail.com> 36132 Acked-by: Daniel Stone <daniels@collabora.com> 36133 36134commit 1c56ac63c040280498c4a9d67b48c35b60070821 36135Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36136Date: Sat Mar 21 13:42:12 2015 -0700 36137 36138 Convert top level extensions to new *allocarray functions 36139 36140 v2: remove now useless parentheses 36141 36142 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36143 Reviewed-by: Matt Turner <mattst88@gmail.com> 36144 36145commit b9e665c8b2f048feb3064ce412e0b3e9eb6797b0 36146Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36147Date: Sat Mar 21 10:52:29 2015 -0700 36148 36149 Convert dix/* to new *allocarray functions 36150 36151 v2: remove now useless parentheses 36152 36153 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36154 Reviewed-by: Matt Turner <mattst88@gmail.com> 36155 36156commit df4e41fdb47cdeab5155224557ed8ab4ec4dc659 36157Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36158Date: Sat Mar 21 10:08:25 2015 -0700 36159 36160 Convert os/* to new *allocarray functions 36161 36162 v2: remove now useless parentheses 36163 36164 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36165 Reviewed-by: Matt Turner <mattst88@gmail.com> 36166 36167commit ae75d50395fdd7a6bc382ba73e923c460764c702 36168Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36169Date: Sat Mar 21 09:28:07 2015 -0700 36170 36171 Add no-fail equivalents of allocarray & reallocarray 36172 36173 v2: Remove extra 's' from comment 36174 36175 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36176 Reviewed-by: Matt Turner <mattst88@gmail.com> 36177 36178commit c213b29d14aaf2f4523d638abc762eaaa873cf83 36179Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36180Date: Sat Mar 21 08:58:04 2015 -0700 36181 36182 Add xallocarray() helper macro 36183 36184 Uses reallocarray to perform integer overflow detection when allocating 36185 an array, using NULL as the previous pointer to force a new allocation. 36186 36187 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36188 Reviewed-by: Matt Turner <mattst88@gmail.com> 36189 36190commit 0887c9463fc7d36d7874e7f978df4109542f7307 36191Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36192Date: Sat Mar 21 08:42:19 2015 -0700 36193 36194 Import reallocarray() from OpenBSD 36195 36196 Wrapper for realloc() that checks for overflow when multiplying 36197 arguments together, so we don't have to add overflow checks to 36198 every single call. For documentation on usage, see: 36199 http://www.openbsd.org/cgi-bin/man.cgi/OpenBSD-current/man3/calloc.3 36200 36201 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36202 Acked-by: Matt Turner <mattst88@gmail.com> 36203 36204commit b96dc999968320ad6322b50c79d847efcfcd02b2 36205Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36206Date: Sat Mar 21 15:21:14 2015 -0700 36207 36208 Add XNFcallocarray() to allow xnfcalloc() to check for overflow 36209 36210 The xnfcalloc() macro took two arguments but simply multiplied them 36211 together without checking for overflow and defeating any overflow 36212 checking that calloc() might have done. Let's not do that. 36213 36214 The original XNFcalloc() function is left for now to preserve driver 36215 ABI, but is marked as deprecated so it can be removed in a future round 36216 of ABI break/cleanup. 36217 36218 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36219 Reviewed-by: Matt Turner <mattst88@gmail.com> 36220 36221commit f3ba909753cd216fc9eca4618a76cc283ccbf51e 36222Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36223Date: Sat Mar 21 11:07:24 2015 -0700 36224 36225 Let calloc handle multiplication 36226 36227 It's going to multiply anyway, so if we have non-constant values, might 36228 as well let it do the multiplication instead of adding another multiply, 36229 and good versions of calloc will check for & avoid overflow in the process. 36230 36231 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36232 Reviewed-by: Matt Turner <mattst88@gmail.com> 36233 36234commit a28202a148508837911c5932a0d14af4b145bece 36235Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36236Date: Sat Mar 21 08:19:30 2015 -0700 36237 36238 Remove remaining doc references to Xalloc, Xrealloc, and Xfree 36239 36240 Functions were removed in commit cad9b053d52f62432 36241 36242 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36243 Reviewed-by: Matt Turner <mattst88@gmail.com> 36244 36245commit 2a6d253b3f2ec4775412b4ae96a96b0b668ec827 36246Author: Alan Coopersmith <alan.coopersmith@oracle.com> 36247Date: Sat Feb 28 10:04:09 2015 -0800 36248 36249 Remove apSolaris.shar (old "aperture" kernel driver for Solaris) 36250 36251 Modern Solaris releases provide this functionality in the OS via the 36252 xsvc driver. Since the move to libpciaccess, nothing in Xorg uses 36253 this aperture driver any more. 36254 36255 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 36256 36257commit b1029716e41e252f149b82124a149da180607c96 36258Author: Ray Strode <rstrode@redhat.com> 36259Date: Thu Apr 16 11:28:16 2015 -0400 36260 36261 systemd-logind: don't second guess D-Bus default timeout 36262 36263 At the moment, the X server uses a non-default timeout for D-Bus 36264 messages to systemd-logind. The only timeouts normally used with 36265 D-Bus are: 36266 36267 1) Infinite 36268 2) Default 36269 36270 Anything else is just as arbitrary as Default, and so rarely makes 36271 sense to use instead of Default. 36272 36273 Put another way, there's little reason to be fault tolerant against 36274 a local root running daemon (logind), that in some configurations, the 36275 X server already depends on for proper functionality. 36276 36277 This commit changes systemd-logind to just use the default timeouts. 36278 36279 Downstream-bug: https://bugzilla.redhat.com/show_bug.cgi?id=1209347 36280 Signed-off-by: Ray Strode <rstrode@redhat.com> 36281 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 36282 Signed-off-by: Keith Packard <keithp@keithp.com> 36283 36284commit 792e9251670ce94210df5c6d354059bbb97f4478 36285Author: Ray Strode <rstrode@redhat.com> 36286Date: Thu Apr 16 11:28:15 2015 -0400 36287 36288 systemd-logind: filter out non-signal messages from message filter 36289 36290 It's possible to receive a message reply in the message filter if a 36291 previous message call timed out locally before the reply arrived. 36292 36293 The message_filter function only handles signals, at the moment, and 36294 does not properly handle message replies. 36295 36296 This commit changes the message_filter function to filter out all 36297 non-signal messages, including spurious message replies. 36298 36299 Downstream-bug: https://bugzilla.redhat.com/show_bug.cgi?id=1209347 36300 Signed-off-by: Ray Strode <rstrode@redhat.com> 36301 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 36302 Signed-off-by: Keith Packard <keithp@keithp.com> 36303 36304commit 41932dfbc841a1adc6512d41085ea3f8ebecb42c 36305Author: Keith Packard <keithp@keithp.com> 36306Date: Wed Apr 8 07:45:28 2015 -0700 36307 36308 mi: Partial pie-slice filled arcs may need more space for spans 36309 36310 The mi filled arc code estimates that a filled arc will produce no 36311 more spans than the arc is tall. This is true for most arcs except 36312 for pie-slice arcs strictly between 180 and 360 degrees where the missing 36313 portion of the arc faces up or down such that we get two spans on some 36314 scanlines. 36315 36316 For those, we need to reserve room for another height/2 spans. This 36317 patch just does it for all partial pie-sliced arcs to make the test 36318 easier to understand; it's just over-allocating a bit of memory, so 36319 that's safe. 36320 36321 Signed-off-by: Keith Packard <keithp@keithp.com> 36322 Reviewed-by: Adam Jackson <ajax@redhat.com> 36323 36324commit 21e7d2bb5cd810688bb905b6ed092497ca8a2c59 36325Author: Hans de Goede <hdegoede@redhat.com> 36326Date: Wed Apr 8 14:07:52 2015 +0200 36327 36328 Re-enable non serverfd input devices immediately on vtenter 36329 36330 Non serverfd input devices will never get a systemd-logind dbus resume signal, 36331 causing them to never get re-enabled. 36332 36333 This commit changes xf86VTEnter() to enable them immediately, fixing this. 36334 36335 BugLink: https://bugs.freedesktop.org/show_bug.cgi?id=89756 36336 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 36337 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 36338 Signed-off-by: Keith Packard <keithp@keithp.com> 36339 36340commit f1da6bf5d94911e78d2e27e6accf0c6e3aefb331 36341Author: Keith Packard <keithp@keithp.com> 36342Date: Tue Mar 31 17:07:43 2015 -0700 36343 36344 Require randrproto version 1.5.0 or newer 36345 36346 Signed-off-by: Keith Packard <keithp@keithp.com> 36347 Reviewed-by: Dave Airlie <airlied@redhat.com> 36348 36349commit e36236eade412dd3894f75f78a7b3d7c1037e6c3 36350Author: Aaron Plattner <aplattner@nvidia.com> 36351Date: Tue Dec 30 09:13:16 2014 -0800 36352 36353 xfree86: Add GPU screens even if there are no active GDevs 36354 36355 xf86platformProbeDev creates GPU screens for any platform devices that were not 36356 matched by a GDev in the loop above, but only if there was at least one device. 36357 This means that it's impossible to configure a device as a GPU screen if there 36358 is only one platform device that matches that driver. 36359 36360 Instead, create a GPU screen (if possible) for any platform device that was not 36361 claimed by the GDev loop. 36362 36363 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 36364 Reviewed-by: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36365 Acked-by: Alex Deucher <alexander.deucher@amd.com> 36366 36367commit 4ecda362594d771f401de467c2d58c0f552227a8 36368Author: Aaron Plattner <aplattner@nvidia.com> 36369Date: Tue Dec 30 09:13:15 2014 -0800 36370 36371 xfree86: Fix xf86_check_platform_slot's handling of PCI 36372 36373 If a PCI entity is found, xf86_check_platform_slot performs a device ID check 36374 against the xf86_platform_device passed in. However, it just returns 36375 immediately without checking the rest of the entities first. This leads to this 36376 situation happening: 36377 36378 1. The nvidia driver creates an entity 0 with bus.type == BUS_PCI 36379 2. The intel driver creates entity 1 for its platform device, opening 36380 /dev/dri/card0 36381 3. xf86platformProbeDev calls probeSingleDevice on the Intel platform device, 36382 which calls doPlatformProbe, which calls xf86_check_platform_slot. 36383 4. xf86_check_platform_slot compares the Intel platform device against the 36384 NVIDIA PCI entity. Since they don't have the same device ID, it returns 36385 TRUE. 36386 5. doPlatformProbe calls xf86ClaimPlatformSlot, which creates a duplicate entity 36387 for the Intel one. 36388 36389 Fix this by only returning FALSE if the PCI ID matches, and continuing the loop 36390 otherwise. In the scenario above, this allows it to continue on to find the 36391 Intel platform device that matches the second entity. 36392 36393 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 36394 Reviewed-by: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36395 Acked-by: Alex Deucher <alexander.deucher@amd.com> 36396 36397commit e608f3521eaaab972a3eea62aa04a65958351c1c 36398Merge: d3b9c47c8 5de138307 36399Author: Keith Packard <keithp@keithp.com> 36400Date: Tue Mar 31 09:06:08 2015 -0700 36401 36402 Merge remote-tracking branch 'airlied/for-keithp' 36403 36404commit d3b9c47c849d039493c5e0ca5c3af8affe738746 36405Author: Jon Turney <jon.turney@dronecode.org.uk> 36406Date: Sun Mar 29 23:18:49 2015 +0100 36407 36408 Fix XQuartz build 36409 36410 Fix XQuartz build since commit e036cbfc "Make PseudoramiXExtensionInit() 36411 prototype more generally available" 36412 36413 Add #include "nonsdk_extinit.h" to xprScreen.c 36414 36415 Add #include "nonsdk_extinit.h" to miinitext.c under INXQUARTZ to provide 36416 declarations used under INXQUARTZ 36417 36418 Signed-off-by: Jon Turney <jon.turney@dronecode.org.uk> 36419 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 36420 Tested-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 36421 Signed-off-by: Keith Packard <keithp@keithp.com> 36422 36423commit 5de13830709a7f2d4d112d71e062f710ef466ab6 36424Author: Keith Packard <keithp@keithp.com> 36425Date: Tue Dec 16 09:56:50 2014 -0800 36426 36427 randr: Use Monitor list for Xinerama 36428 36429 This replaces the CRTC-based Xinerama implementation with one which 36430 uses Monitors instead, allowing clients to manipulate the Xinerama 36431 configuration through the RandR Monitor list. 36432 36433 Reviewed-by: Dave Airlie <airlied@redhat.com> 36434 Signed-off-by: Keith Packard <keithp@keithp.com> 36435 36436commit 7e1f86d42b54fb7f6492875e47a718eaeca3069b 36437Author: Keith Packard <keithp@keithp.com> 36438Date: Tue Dec 16 01:59:03 2014 -0800 36439 36440 randr: Add Monitor support (v1.1) 36441 36442 Store the user-defined monitors in the RandR screen private. 36443 36444 Generate a list of monitors from both the user-defined ones and from 36445 any outputs not mentioned in one of the user-defined monitors. This list 36446 covers both the outputs in the main screen as well as any slaves. 36447 36448 v1.1: airlied: fix up primary skipping bug, 36449 fix wrong height initialiser 36450 add get_active flag from updated protocol. 36451 36452 Reviewed-by: Dave Airlie <airlied@redhat.com> 36453 Signed-off-by: Keith Packard <keithp@keithp.com> 36454 36455commit 9c2b4f8e0e2b5d4b5e1102d6eea7bdb4211baa68 36456Author: Dave Airlie <airlied@redhat.com> 36457Date: Tue Mar 31 11:18:44 2015 +1000 36458 36459 xf86Crtc: add tile prop setting 36460 36461 Add support for drivers to set the tiling 36462 property. This is used by clients to 36463 work out the monitor tiles for DisplayID 36464 monitors. 36465 36466 Reviewed-by: Keith Packard <keithp@keithp.com> 36467 Signed-off-by: Dave Airlie <airlied@redhat.com> 36468 36469commit 7088816fee0ca7d609c7bca41ef8c3fc938556f5 36470Merge: 95e83ff87 e977b404d 36471Author: Keith Packard <keithp@keithp.com> 36472Date: Fri Mar 27 06:45:56 2015 -0700 36473 36474 Merge remote-tracking branch 'anholt/glamor-next' 36475 36476commit e977b404d7d1c6cd2be2168f4fdce0ae31cd1f9f 36477Author: Keith Packard <keithp@keithp.com> 36478Date: Sat Jan 17 07:44:27 2015 +1300 36479 36480 glamor: * 1/size is faster than / size in VS 36481 36482 Pass the inverse of the texture size to glamor vertex shaders so that 36483 we multiply by that instead of dividing by the size as multiplication 36484 is generally faster than division. 36485 36486 Signed-off-by: Keith Packard <keithp@keithp.com> 36487 Signed-off-by: Eric Anholt <eric@anholt.net> 36488 Reviewed-by: Eric Anholt <eric@anholt.net> 36489 Reviewed-by: Matt Turner <mattst88@gmail.com> 36490 36491commit 82634d2b69950e7e9a066c8c96bf2624ac26193a 36492Author: Jon TURNEY <jon.turney@dronecode.org.uk> 36493Date: Mon Feb 9 13:45:44 2015 +0000 36494 36495 ephyr: Avoid a segfault with 'DISPLAY= Xephyr -glamor' 36496 36497 ephyr_glamor_connect() returns NULL if we failed, but applying 36498 xcb_connection_has_error() to NULL is not permitted. 36499 36500 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36501 Signed-off-by: Eric Anholt <eric@anholt.net> 36502 Reviewed-by: Daniel Martin <consume.noise@gmail.com> 36503 36504commit 391bcf77db96ff227e504c213aa5d5db09b6a050 36505Author: Adam Jackson <ajax@redhat.com> 36506Date: Mon Feb 23 11:58:59 2015 -0500 36507 36508 glamor: Fix up indentation 36509 36510 Signed-off-by: Adam Jackson <ajax@redhat.com> 36511 Signed-off-by: Eric Anholt <eric@anholt.net> 36512 Reviewed-by: Eric Anholt <eric@anholt.net> 36513 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 36514 36515commit 0669babf2b5b50cbc185b0f714671b2c2b368778 36516Author: Matt Turner <mattst88@gmail.com> 36517Date: Wed Mar 4 13:42:48 2015 -0800 36518 36519 glamor: Perform texture2D() separately from swizzle. 36520 36521 The texture2D() happens in each branch, so we may as well do it as early 36522 as possible and hide some of its latency in the branching instructions. 36523 Moving it outside the (uniform) control flow reduces the number of 36524 instructions in the fs_source shader from 64 to 46 and in the 36525 set_alpha_source shader from 69 to 47 on Haswell. 36526 36527 Signed-off-by: Eric Anholt <eric@anholt.net> 36528 Reviewed-by: Eric Anholt <eric@anholt.net> 36529 36530commit 9e9fcf578063b4155aab4adab83f8d956bde5d1a 36531Author: Eric Anholt <eric@anholt.net> 36532Date: Wed Feb 4 14:38:15 2015 -0800 36533 36534 glamor: Add a helper function for the common GL_QUADS fallback pattern. 36535 36536 We should do better than this with an index buffer, but for now at 36537 least make it so that we don't have to copy the same code to new 36538 places. 36539 36540 Signed-off-by: Eric Anholt <eric@anholt.net> 36541 Reviewed-by: Keith Packard <keithp@keithp.com> 36542 36543commit 909a406aa239b8d231d6f63ce05a3e4a2bc3cb07 36544Author: Eric Anholt <eric@anholt.net> 36545Date: Sun Dec 28 19:39:45 2014 -1000 36546 36547 glamor: Don't optimize out scissor updates in CopyArea. 36548 36549 This possibly is a minor hit for immediate mode renderers (no 36550 difference on copypixin100 on my hsw, n=12), but it gives important 36551 information about drawing bounds to a deferred renderer (3.1x 36552 improvement in copypixwin100 on vc4). 36553 36554 Signed-off-by: Eric Anholt <eric@anholt.net> 36555 Reviewed-by: Keith Packard <keithp@keithp.com> 36556 36557commit 6ba6cc57e17aefb7db0201a1f3180ab55076eb48 36558Author: Eric Anholt <eric@anholt.net> 36559Date: Sat Dec 27 09:00:58 2014 -1000 36560 36561 glamor: Just set the logic op to what we want at the start of all rendering. 36562 36563 By dropping the unconditional logic op disable at the end of 36564 rendering, this fixes GL errors being thrown in GLES2 contexts (which 36565 don't have logic ops). On desktop, this also means a little less 36566 overhead per draw call from taking one less trip through the 36567 glEnable/glDisable switch statement of doom in Mesa. 36568 36569 The exchange here is that we end up taking a trip through it in the 36570 XV, Render, and gradient-generation paths. If the glEnable() is 36571 actually costly, we should probably cache our logic op state in our 36572 screen, since there's no way the GL could make that switch statement 36573 as cheap as the caller caching it would be. 36574 36575 v2: Don't forget to set the logic op in Xephyr's drawing. 36576 36577 Signed-off-by: Eric Anholt <eric@anholt.net> 36578 Reviewed-by: Keith Packard <keithp@keithp.com> 36579 36580commit 8102927282d5134493e5009a876a6b01a68d1f97 36581Author: Jon TURNEY <jon.turney@dronecode.org.uk> 36582Date: Thu Feb 5 12:52:02 2015 +0000 36583 36584 glamor: Fix build when configured --enable-glamor --disable-xshmfence 36585 36586 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36587 Signed-off-by: Eric Anholt <eric@anholt.net> 36588 Reviewed-by: Eric Anholt <eric@anholt.net> 36589 36590commit 4218a1e066cf39bb980ebbc9f69536c85232da5c 36591Author: Olivier Fourdan <ofourdan@redhat.com> 36592Date: Thu Feb 5 11:59:22 2015 +0100 36593 36594 glamor: check max native ALU instructions 36595 36596 When using glamor (either in Xephyr or Xwayland) on hardware with too 36597 low instructions limit, glamor fallbacks to sw due to large shaders. 36598 36599 This makes glamor unbearably slow on such hardware. 36600 36601 Check reported value for GL_MAX_PROGRAM_NATIVE_ALU_INSTRUCTIONS_ARB 36602 and fail in glamor_init() if the limit is lower than 128. 36603 36604 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=88316 36605 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 36606 Signed-off-by: Eric Anholt <eric@anholt.net> 36607 Reviewed-by: Eric Anholt <eric@anholt.net> 36608 36609commit e0788a03144186522fd0ef0e9c954e2744b40275 36610Author: Keith Packard <keithp@keithp.com> 36611Date: Wed Oct 29 23:47:23 2014 -0700 36612 36613 glamor: Eliminate GLAMOR_TEXTURE_LARGE pixmap type 36614 36615 Initialize full pixmap private for all pixmaps, including block 36616 dimensions and counts so that no checks are needed when walking the 36617 fbos. 36618 36619 Signed-off-by: Keith Packard <keithp@keithp.com> 36620 Signed-off-by: Eric Anholt <eric@anholt.net> 36621 Reviewed-by: Eric Anholt <eric@anholt.net> 36622 36623commit cc731ce0ca4d6c5c8d9c1130864e652a814644b4 36624Author: Keith Packard <keithp@keithp.com> 36625Date: Wed Oct 29 23:20:11 2014 -0700 36626 36627 glamor: Create inline tests for small/large pixmaps 36628 36629 This will let us eliminate the pixmap types shortly 36630 36631 Signed-off-by: Keith Packard <keithp@keithp.com> 36632 Signed-off-by: Eric Anholt <eric@anholt.net> 36633 Reviewed-by: Eric Anholt <eric@anholt.net> 36634 36635commit 020fcc582837b80b20843a09c1e28fa96096fe6b 36636Author: Keith Packard <keithp@keithp.com> 36637Date: Wed Oct 29 23:08:57 2014 -0700 36638 36639 glamor: Eliminate separate 'large' pixmap private structure 36640 36641 Just embed the large elements in the regular pixmap private and 36642 collapse the union to a single struct. 36643 36644 Signed-off-by: Keith Packard <keithp@keithp.com> 36645 Signed-off-by: Eric Anholt <eric@anholt.net> 36646 Reviewed-by: Eric Anholt <eric@anholt.net> 36647 36648commit 9ef5cbf8ab8a51941d469bfcaca6cbfb7838d1ef 36649Author: Keith Packard <keithp@keithp.com> 36650Date: Wed Oct 29 22:54:27 2014 -0700 36651 36652 glamor: Remove unused glamor_pixmap_private_atlas_t 36653 36654 Signed-off-by: Keith Packard <keithp@keithp.com> 36655 Signed-off-by: Eric Anholt <eric@anholt.net> 36656 Reviewed-by: Eric Anholt <eric@anholt.net> 36657 36658commit af687396f1875de0a45683aad4fd5a9a2e00f10d 36659Author: Keith Packard <keithp@keithp.com> 36660Date: Wed Oct 29 22:50:53 2014 -0700 36661 36662 glamor: Remove screen private and pixmap ptrs from pixmap private and fbo 36663 36664 There's no reason to waste memory storing redundant copies of the same 36665 pointer all over the system; just pass in pointers as necessary to 36666 each function. 36667 36668 Signed-off-by: Keith Packard <keithp@keithp.com> 36669 Signed-off-by: Eric Anholt <eric@anholt.net> 36670 Reviewed-by: Eric Anholt <eric@anholt.net> 36671 36672commit 1eb954c3830d46c27bf2a61f825b59f12092728c 36673Author: Keith Packard <keithp@keithp.com> 36674Date: Wed Oct 29 21:31:32 2014 -0700 36675 36676 glamor: Remove remaining support for FBOs not matching pixmap size 36677 36678 The core rendering code already requires that FBOs be allocated at 36679 exactly the pixmap size so that tiling and stippling work 36680 correctly. Remove the allocation support for that, along with the 36681 render code. 36682 36683 Signed-off-by: Keith Packard <keithp@keithp.com> 36684 Signed-off-by: Eric Anholt <eric@anholt.net> 36685 Reviewed-by: Eric Anholt <eric@anholt.net> 36686 36687commit adb847faeb51c567933fab97b97e79ee3d184dc8 36688Author: Keith Packard <keithp@keithp.com> 36689Date: Wed Oct 29 20:49:49 2014 -0700 36690 36691 glamor: Eliminate GLAMOR_TEXTURE_PACK pixmap type 36692 36693 This is not used anywhere 36694 36695 Signed-off-by: Keith Packard <keithp@keithp.com> 36696 Signed-off-by: Eric Anholt <eric@anholt.net> 36697 Reviewed-by: Eric Anholt <eric@anholt.net> 36698 36699commit 2f80c7791bb0b11f261cb1e3e0d89163dcdd0342 36700Author: Keith Packard <keithp@keithp.com> 36701Date: Wed Oct 29 20:48:40 2014 -0700 36702 36703 glamor: Eliminate GLAMOR_SEPARATE_TEXTURE pixmap type 36704 36705 This was only used to signal when we couldn't ask the DDX to draw to a 36706 pixmap; now that we have no DDX-based fallbacks, we don't need to have 36707 this type. 36708 36709 Signed-off-by: Keith Packard <keithp@keithp.com> 36710 Signed-off-by: Eric Anholt <eric@anholt.net> 36711 Reviewed-by: Eric Anholt <eric@anholt.net> 36712 36713commit c6ab13566798c7adff23a609575a7ac2d1ce2df6 36714Author: Keith Packard <keithp@keithp.com> 36715Date: Wed Oct 29 20:44:31 2014 -0700 36716 36717 glamor: Remove ddx fallback check functions 36718 36719 With no DDX-based fallback support, we can remove these functions as 36720 they are no longer called. 36721 36722 Signed-off-by: Keith Packard <keithp@keithp.com> 36723 Signed-off-by: Eric Anholt <eric@anholt.net> 36724 Reviewed-by: Eric Anholt <eric@anholt.net> 36725 36726commit 90d326fcc687e6d6d4b308f6272ededcf8145a17 36727Author: Keith Packard <keithp@keithp.com> 36728Date: Wed Oct 29 20:40:21 2014 -0700 36729 36730 glamor: Remove _nf rendering functions 36731 36732 These were used by the non-standard glamor implementation in the intel 36733 driver. 36734 36735 Signed-off-by: Keith Packard <keithp@keithp.com> 36736 Signed-off-by: Eric Anholt <eric@anholt.net> 36737 Reviewed-by: Eric Anholt <eric@anholt.net> 36738 36739commit 697f8581e04a93862a6049cc982d01d25f9a0410 36740Author: Keith Packard <keithp@keithp.com> 36741Date: Wed Oct 29 20:30:12 2014 -0700 36742 36743 glamor: Eliminate GLAMOR_USE_SCREEN and GLAMOR_USE_PICTURE_SCREEN 36744 36745 Remove these defines as we start to remove support for non-standard 36746 glamor layering as used by the intel driver. 36747 36748 v2: Rebase on the blockhandler change and the Xephyr init failure 36749 change (by anholt), fix stray NO_DRI3 addition to xwayland. 36750 36751 Signed-off-by: Keith Packard <keithp@keithp.com> 36752 Signed-off-by: Eric Anholt <eric@anholt.net> 36753 Reviewed-by: Eric Anholt <eric@anholt.net> 36754 36755commit 28ff815c4b32c88bd9007fd79f2acf3682a4b4db 36756Author: Keith Packard <keithp@keithp.com> 36757Date: Wed Oct 29 20:20:29 2014 -0700 36758 36759 glamor: Eliminate GLAMOR_CREATE_PIXMAP_MAP and GLAMOR_MEMORY_MAP 36760 36761 GLAMOR_MEMORY_MAP was only used with GLAMOR_CREATE_PIXMAP_MAP, and 36762 GLAMOR_CREATE_PIXMAP_MAP doesn't appear to be used anywhere, so just 36763 remove both of them. 36764 36765 v2: Fix a stray whitespace bug that was introduced (change by anholt). 36766 36767 Signed-off-by: Keith Packard <keithp@keithp.com> 36768 Signed-off-by: Eric Anholt <eric@anholt.net> 36769 Reviewed-by: Eric Anholt <eric@anholt.net> 36770 36771commit 0e1372e1bd8f79fa295738bc1e983cb3648ba4be 36772Author: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36773Date: Mon Jan 19 12:37:55 2015 +0100 36774 36775 glamor: GL_TEXTURE_MAX_LEVEL is not available on GLES2 36776 36777 Remove the calls to GL_TEXTURE_MAX_LEVEL. Setting the filtering is 36778 a sufficient hint to the driver about texture mipmap allocation. 36779 36780 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36781 Signed-off-by: Eric Anholt <eric@anholt.net> 36782 Reviewed-by: Eric Anholt <eric@anholt.net> 36783 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 36784 36785commit c1f35c3d862f77cbd2fb871be5613fb0ca7e508b 36786Author: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36787Date: Mon Jan 12 15:29:34 2015 +0100 36788 36789 glamor: Use GL_FRAMEBUFFER instead of GL_READ_FRAMEBUFFER 36790 36791 The latter might not be available on GLES2. 36792 36793 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36794 Signed-off-by: Eric Anholt <eric@anholt.net> 36795 Reviewed-by: Eric Anholt <eric@anholt.net> 36796 36797commit b66501b4fd9c392e0f971ea356b27eb71c9c9e79 36798Author: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36799Date: Mon Jan 19 12:36:52 2015 +0100 36800 36801 glamor: do not check for gl errors in glamor_build_program 36802 36803 According to Eric Anholt the check for glGetError is not needed here. 36804 Because a opengl error might be set before this function is called 36805 keeping the check could result in glamor_build_program returning 36806 failure when building the shader succeeded. 36807 36808 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36809 Signed-off-by: Eric Anholt <eric@anholt.net> 36810 Reviewed-by: Eric Anholt <eric@anholt.net> 36811 36812commit 7c6f483670770e4e534cafd4e70d0b1490f4cca6 36813Author: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36814Date: Mon Jan 19 12:32:54 2015 +0100 36815 36816 glamor: only use (un)pack_subimage when available 36817 36818 Check for GL_EXT_unpack_subimage and GL_NV_pack_subimage to 36819 check if GL_(UN)PACK_ROW_LENGTH is available. Set the offsets 36820 manually to prevent calls to GL_(UN)PACK_SKIP_*. 36821 36822 v2: Check support for GL_NV_pack_subimage as suggested by Matt Turner. 36823 36824 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> 36825 Signed-off-by: Eric Anholt <eric@anholt.net> 36826 Reviewed-by: Eric Anholt <eric@anholt.net> 36827 36828commit 4f534c26c64cc1edd33b04d30e4df818b7ae55d9 36829Author: Olivier Fourdan <ofourdan@redhat.com> 36830Date: Wed Jan 28 16:08:26 2015 +0100 36831 36832 xwayland: Add dependency on glamor libs 36833 36834 So that Xwayland gets re-linked each time glamor is modified. 36835 36836 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 36837 Signed-off-by: Eric Anholt <eric@anholt.net> 36838 Reviewed-by: Eric Anholt <eric@anholt.net> 36839 36840commit 251a067993658fd02f158fef36476f898c238e3c 36841Author: Olivier Fourdan <ofourdan@redhat.com> 36842Date: Mon Feb 2 10:41:06 2015 +0100 36843 36844 ephyr: Fail if glamor is requested but not usable 36845 36846 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 36847 Signed-off-by: Eric Anholt <eric@anholt.net> 36848 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 36849 Reviewed-by: Eric Anholt <eric@anholt.net> 36850 36851commit 23702dd2689e2e1e65be5767ac0303a985bb04a0 36852Author: Emil Velikov <emil.l.velikov@gmail.com> 36853Date: Thu Mar 19 12:26:29 2015 +0000 36854 36855 randr: coding style fixes 36856 36857 In most of xserver code-base we define new functions at column 0, with 36858 their return type provided on the previous line. Two functions did not 36859 follow this rule so update them, and get them wrapped up to 80 as an 36860 added bonus. 36861 36862 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 36863 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 36864 36865commit 739e8fac0e9d8d4c1653e53a8f3ce2d38b3de320 36866Author: Emil Velikov <emil.l.velikov@gmail.com> 36867Date: Thu Mar 19 12:22:18 2015 +0000 36868 36869 randr: wrap long line 36870 36871 Also make use of total_name_len variable for consistency. 36872 36873 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 36874 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 36875 36876commit 363cd0e0b499ea8c32b2aa5cf7ea0f0a56b4c3ef 36877Author: Emil Velikov <emil.l.velikov@gmail.com> 36878Date: Thu Mar 19 12:19:23 2015 +0000 36879 36880 randr: use local variables where possible 36881 36882 This will allow us to make the code more readable, and the lines will 36883 fit within 80 columns. 36884 36885 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 36886 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 36887 36888commit 93ef0e580e9cdbe739046a0873971d402525ef00 36889Author: Emil Velikov <emil.l.velikov@gmail.com> 36890Date: Thu Mar 19 12:40:06 2015 +0000 36891 36892 randr: use randr: prefix in ErrorF() 36893 36894 To provide some information about the origin of the message. 36895 36896 Cc: Dave Airlie <airlied@redhat.com> 36897 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 36898 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 36899 36900commit a08ee773983c44ebb893f10a1dcfa443f2734277 36901Author: Emil Velikov <emil.l.velikov@gmail.com> 36902Date: Thu Mar 19 12:36:50 2015 +0000 36903 36904 randr: remove chatty error messages 36905 36906 All of these seem like left over from developments stage. Remove them as 36907 they can cause excessive flood in the logs. 36908 36909 Cc: Dave Airlie <airlied@redhat.com> 36910 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 36911 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 36912 36913commit a34d29c2edd786f0baa0e7c334f4174eeecb71d2 36914Author: Emil Velikov <emil.l.velikov@gmail.com> 36915Date: Sun Mar 8 17:56:58 2015 +0000 36916 36917 configure.ac: remove remaining TLS references 36918 36919 No longer used with the removal of the GL dispatch (glapi) from libglx a 36920 few releases ago. 36921 36922 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 36923 Reviewed-by: Adam Jackson <ajax@redhat.com> 36924 Reviewed-by: Eric Anholt <eric@anholt.net> 36925 36926commit 95e83ff87ab0149ab01c9299524dfbe37d9b21a2 36927Author: Jon TURNEY <jon.turney@dronecode.org.uk> 36928Date: Thu Jul 12 02:00:43 2012 +0100 36929 36930 Don't allow both RandR XINERAMA and pseudoramiX XINERAMA extensions to register 36931 36932 Prevent RRXinerama from activating if PseudoramiX is, so we don't get XINERAMA 36933 listed twice in the list of extensions. I think this is otherwise benign, as 36934 the PseudoramiX XINERAMA gets registered first and thus handles all requests. 36935 36936 Perhaps AddExtension() ought to warn us if the extension name is already 36937 registered? 36938 36939 This appears to be a long-standing bug seen in XQuartz, and now in XWin as well. 36940 36941 Future work: Perhaps since RRXinerama isn't actually doing anything useful but 36942 faking it when we have one screen, it seems that the PseudoramiX code could be 36943 also used in that case. 36944 36945 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36946 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 36947 36948commit 234fe391802b114e96d60a6630f717f8e4fb726c 36949Author: Jon TURNEY <jon.turney@dronecode.org.uk> 36950Date: Fri Nov 7 13:38:20 2014 +0000 36951 36952 hw/xwin: Report OS version in log 36953 36954 Report OS version in log 36955 Report if WoW64 is in use 36956 Manifest for compatbility with Windows versions, so we don't get lied to by GetVersionEx() 36957 Also, make the description in the manifest a bit more generic 36958 36959 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36960 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 36961 36962commit 7a22912edbda826778463f8fefa9c8e689858bbe 36963Author: Jon TURNEY <jon.turney@dronecode.org.uk> 36964Date: Thu Nov 6 15:22:58 2014 +0000 36965 36966 hw/xwin: Remove Shadow DirectDraw engine 36967 36968 Maybe a long time ago this made some kind of sense, but now there's no good 36969 reason to ever use this, rather than the Shadow DirectDraw NonLocking engine. 36970 36971 Also remove screen private data members used by other removed engines. 36972 36973 Also remove no longer needed OS version check in winDetectSupportedEngines() 36974 36975 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36976 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 36977 36978commit 55a84be085e9f479f7907bed5fb68fc66ce81ae8 36979Author: Jon TURNEY <jon.turney@dronecode.org.uk> 36980Date: Sat Mar 3 20:13:19 2012 +0000 36981 36982 hw/xwin: Only set native positions if XINERAMA is enabled 36983 36984 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36985 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 36986 36987commit 23e07d71b607c99c3a0a0ea362c70dcae73485c6 36988Author: Jon TURNEY <jon.turney@dronecode.org.uk> 36989Date: Thu Nov 6 14:28:17 2014 +0000 36990 36991 hw/xwin: Register native screens with pseudoramiX 36992 36993 Update man page to document pseudo-xinerama 36994 36995 v2: Make the use of PseudoramiXExtensionInit() match the prototype 36996 v3: Update for nonsdk_extinit.h 36997 36998 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 36999 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 37000 37001commit e036cbfccbe33775524a469082306913843c63df 37002Author: Jon TURNEY <jon.turney@dronecode.org.uk> 37003Date: Thu Nov 6 13:54:11 2014 +0000 37004 37005 Make PseudoramiXExtensionInit() prototype more generally available 37006 37007 Make PseudoramiXExtensionInit() prototype available to hw/xwin 37008 37009 Rather than avoiding a reference to it being pulled in to Xorg by sdksyms by 37010 hiding this prototype behind the INXQUARTZ define, which is only defined when 37011 building Xquartz, introduce nonsdk_extinit.h and move it there. 37012 37013 (The only remaining use of INXQUARTZ is in mi/miiniext.c, in order 37014 to do PseudoramiXExtensionInit() at the point apparently needed by Xquartz) 37015 37016 Also remove duplicate declaration of noPseudoramiXExtension from pseudoramiX.h 37017 37018 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 37019 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 37020 37021commit 3aad9b7556391b0f55e46ba13ced5e58d8fd47f7 37022Author: Jon TURNEY <jon.turney@dronecode.org.uk> 37023Date: Mon Nov 7 20:54:10 2011 +0000 37024 37025 hw/xwin: Turn on -hostintitle by default 37026 37027 Turn on -hostintitle by default 37028 Provide -nohostintitle to disable if needed 37029 37030 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 37031 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 37032 37033commit b2aaf69e62f4109ffb8aaf39e9bd2571abb29dfb 37034Author: Jon TURNEY <jon.turney@dronecode.org.uk> 37035Date: Thu Jun 28 14:22:07 2012 +0100 37036 37037 hw/xwin: Report Cygwin version information in log 37038 37039 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 37040 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 37041 37042commit d02f9611c8d3bafca4be17c65efca3c5527f71e2 37043Author: Jon TURNEY <jon.turney@dronecode.org.uk> 37044Date: Mon Mar 16 16:29:29 2015 +0000 37045 37046 hw/xwin/glx: Improve code generator to deal with latest Khronos OpenGL registry XML 37047 37048 Improve the parsing of the <proto> XML element 37049 37050 Include all text from the param element, in the order it appears in the xml 37051 document, as part of the formal parameter declaration 37052 37053 This is needed to correctly handle the XML description added in svn r27498 of 37054 glPathGlyphIndexRangeNV()'s baseAndCount parameter of type GLuint[2] 37055 37056 This fixes the way the parameter declaration is generated so it is in the 37057 correct form 'GLuint baseAndCount_[2]' and not 'GLuint baseAndCount[2]_' 37058 37059 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 37060 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 37061 37062commit 5071cb7e0ac5f17c48ed7f3f3b6f242dec802f66 37063Author: Jon TURNEY <jon.turney@dronecode.org.uk> 37064Date: Mon Mar 16 16:14:10 2015 +0000 37065 37066 hw/xwin/glx: Refactor parsing of the <proto> XML element 37067 37068 Factor out duplicated code used in parsing of the <proto> XML element in the 37069 code generator 37070 37071 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 37072 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 37073 37074commit d3080d421bf0d91daea2e39bfc391c43d7fdad75 37075Author: Jon TURNEY <jon.turney@dronecode.org.uk> 37076Date: Thu Feb 5 20:29:48 2015 +0000 37077 37078 os: Teach vpnprintf() how to handle "%*.*s" 37079 37080 XdmcpFatal uses the format specifier %*.*s, which vpnprintf() doesn't 37081 understand, which causes a backtrace and prevents the reason for the XDMCP 37082 failure being logged. 37083 37084 See also: 37085 https://bugs.freedesktop.org/show_bug.cgi?id=66862 37086 https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=758574 37087 37088 "%*.*s" is also currently used in a few other places, so teach vpnprintf() how 37089 to handle it 37090 37091 $ fgrep -r "%*.*s" * 37092 hw/dmx/config/scanner.l: fprintf(stderr, "parse error on line %d at token \"%*.*s\"\n", 37093 hw/dmx/dmxlog.c: ErrorF("(%s) dmx[i%d/%*.*s]: ", type, 37094 hw/dmx/input/dmxinputinit.c: dmxLogCont(dmxInfo, "\t[i%d/%*.*s", 37095 os/access.c: ErrorF("Xserver: siAddrMatch(): type = %s, value = %*.*s -- %s\n", 37096 os/access.c: ("Xserver: siCheckAddr(): type = %s, value = %*.*s, len = %d -- %s\n", 37097 os/xdmcp.c: FatalError("XDMCP fatal error: %s %*.*s\n", type, 37098 37099 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 37100 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 37101 37102commit 491cf02e191e70c5ce24c19da880bb79bebfc03c 37103Author: Jon TURNEY <jon.turney@dronecode.org.uk> 37104Date: Tue Feb 10 14:37:26 2015 +0000 37105 37106 os: XDMCP options like -query etc. should imply -listen tcp 37107 37108 In X server 1.17, the default configuration is now -nolisten tcp. In this 37109 configuration, XDMCP options don't work usefully, as the X server is not 37110 listening on the port for the display that it tells the display manager to 37111 connect to. 37112 37113 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 37114 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37115 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 37116 37117commit f42520c5f151bda25e52ae8b0408a421c3fba7e5 37118Author: Jon TURNEY <jon.turney@dronecode.org.uk> 37119Date: Wed Feb 4 17:04:45 2015 +0000 37120 37121 ephyr: Avoid a segfault with 'DISPLAY= Xephy -glamor' 37122 37123 ephyr_glamor_connect() returns NULL if we failed, but applying 37124 xcb_connection_has_error() to NULL is not permitted. 37125 37126 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 37127 Reviewed-by: Daniel Martin <consume.noise@gmail.com> 37128 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 37129 37130commit 0a78b599b34cc8b5fe6fe82f90e90234e8ab7a56 37131Author: Jürg Billeter <j@bitron.ch> 37132Date: Sat Feb 7 18:13:21 2015 +0100 37133 37134 int10: Fix error check for pci_device_map_legacy 37135 37136 pci_device_map_legacy returns 0 on success. 37137 37138 Signed-off-by: Jürg Billeter <j@bitron.ch> 37139 Reviewed-by: Adam Jackson <ajax@redhat.com> 37140 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 37141 37142commit 21b896939c5bb242f3aacc37baf12379e43254b6 37143Author: Egbert Eich <eich@freedesktop.org> 37144Date: Tue Mar 3 16:27:05 2015 +0100 37145 37146 symbols: Fix sdksyms.sh to cope with gcc5 37147 37148 Gcc5 adds additional lines stating line numbers before and 37149 after __attribute__() which need to be skipped. 37150 37151 Signed-off-by: Egbert Eich <eich@freedesktop.org> 37152 Tested-by: Daniel Stone <daniels@collabora.com> 37153 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 37154 37155commit 7ea64fb4374504bd3d524fc08c90efdab9f253ea 37156Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37157Date: Mon Mar 9 09:55:57 2015 -0700 37158 37159 Clear ListenTransConns entries in CloseWellKnownConnections 37160 37161 Since _XSERVTransClose frees the connection pointer passed to it, 37162 remove that pointer from the array, so we don't try to double free it 37163 if we come back into CloseWellKnownConnections again. 37164 37165 Should fix https://bugzilla.yoctoproject.org/show_bug.cgi?id=6665 in which 37166 the shutdown section of the main() loop called CloseWellKnownConnections() 37167 and then moved on to ddxGiveUp(), which failed to release the VT and thus 37168 called AbortServer(), which called CloseWellKnownConnections() again. 37169 37170 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37171 Reviewed-by: Adam Jackson <ajax@redhat.com> 37172 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 37173 37174commit 6d3cf35a6f0856ac44a7be560e2265461f9bb32b 37175Author: Emil Velikov <emil.l.velikov@gmail.com> 37176Date: Mon Mar 9 12:00:52 2015 +0000 37177 37178 autogen.sh: use quoted string variables 37179 37180 Place quotes around the $srcdir, $ORIGDIR and $0 variables to prevent 37181 fall-outs, when they contain space. 37182 37183 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 37184 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 37185 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 37186 37187commit 5c4202ea85aaea2a4dc7eb29776357a2ba13e191 37188Author: Michal Srb <msrb@suse.com> 37189Date: Thu Feb 19 14:57:27 2015 +0200 37190 37191 Expose GetMaster to modules. 37192 37193 Add _X_EXPORT to GetMaster function. It is required by tigervnc's VNC module. 37194 37195 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 37196 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 37197 37198commit f485a1af64bb00c696ea9f79961786bd791eaec1 37199Author: Peter Hutterer <peter.hutterer@who-t.net> 37200Date: Tue Dec 16 14:43:29 2014 +1000 37201 37202 Drop valuator mask argument from GetKeyboardEvents 37203 37204 Nothing was using it and if anyone had they would've gotten a warning and 37205 noticed that it doesn't actually work. Drop this, it has been unused for years. 37206 37207 Input ABI 22 37208 37209 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 37210 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 37211 37212commit 9d9bd38fe1454590c303dc936ddac913808bf881 37213Author: Dave Airlie <airlied@redhat.com> 37214Date: Tue Feb 17 14:40:27 2015 +1000 37215 37216 os/access: fix regression in server interpreted auth 37217 37218 This was reported on irc on Fedora when rawhide went to 1.17.1. 37219 37220 regression occured in: 2566835b4374edb3e5a8353d4f7c9e7ec4851c57 37221 os: Eliminate uninitialized value warnings from access.c 37222 37223 siAddrMatch doesn't need addr to be a useful value, it checks 37224 some things like localuser without having an address at all. 37225 37226 Signed-off-by: Dave Airlie <airlied@redhat.com> 37227 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 37228 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 37229 37230commit 3a06faf3fcdb7451125a46181f9152e8e59e9770 37231Author: Keith Packard <keithp@keithp.com> 37232Date: Tue Feb 10 22:16:49 2015 -0800 37233 37234 Mark development version for 1.18 37235 37236 Signed-off-by: Keith Packard <keithp@keithp.com> 37237 37238commit 4a758f59a8a65c313244e06852745c8e180ff247 37239Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37240Date: Fri Dec 12 22:38:26 2014 -0800 37241 37242 dmx: include header for DMXExtensionInit() in dmx.c 37243 37244 Gets rid of gcc 4.8 warning: 37245 37246 dmx.c:1193:1: warning: no previous prototype for ‘DMXExtensionInit’ [-Wmissing-prototypes] 37247 DMXExtensionInit(void) 37248 ^ 37249 37250 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37251 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37252 Reviewed-by: Keith Packard <keithp@keithp.com> 37253 37254commit a0e44ddfb0c641e7896a55abf26621ae3d17e337 37255Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37256Date: Fri Dec 12 22:33:54 2014 -0800 37257 37258 dmx: constify GCOps & GCFuncs pointers 37259 37260 Gets rid of 16 instances of gcc 4.8 warnings: 37261 37262 In file included from dmxgc.c:41:0: 37263 dmx.h:327:23: warning: assignment discards ‘const’ qualifier from 37264 pointer target type [enabled by default] 37265 (_saved)->_entry = (_actual)->_entry; \ 37266 ^ 37267 dmxgc.h:80:5: note: in expansion of macro ‘DMX_WRAP’ 37268 DMX_WRAP(funcs, &dmxGCFuncs, _pGCPriv, (_pGC)); \ 37269 ^ 37270 dmxgc.c:192:5: note: in expansion of macro ‘DMX_GC_FUNC_EPILOGUE’ 37271 DMX_GC_FUNC_EPILOGUE(pGC); 37272 ^ 37273 37274 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37275 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37276 Reviewed-by: Keith Packard <keithp@keithp.com> 37277 37278commit 9682c47e228962e4008855c4da7467748742685e 37279Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37280Date: Fri Dec 12 22:20:24 2014 -0800 37281 37282 dmx: Mark glxIsExtensionSupported as a const char * 37283 37284 Gets rid of 9 instances of gcc 4.8 warning: 37285 37286 glxcmds.c: In function ‘CreateContext’: 37287 glxcmds.c:378:13: warning: passing argument 1 of ‘glxIsExtensionSupported’ 37288 discards ‘const’ qualifier from pointer target type [enabled by default] 37289 else if (glxIsExtensionSupported("GLX_SGIX_fbconfig")) { 37290 ^ 37291 In file included from glxserver.h:49:0, 37292 from glxcmds.c:41: 37293 glxscreens.h:53:12: note: expected ‘char *’ but argument is of type 37294 ‘const char *’ 37295 extern int glxIsExtensionSupported(char *ext); 37296 37297 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37298 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37299 Reviewed-by: Keith Packard <keithp@keithp.com> 37300 37301commit 45ec646d77086eeb5f94766b055a93b5c75bdef0 37302Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37303Date: Fri Dec 12 22:16:49 2014 -0800 37304 37305 dmx: remove redundant redeclarations of variables from other headers 37306 37307 Gets rid of these gcc 4.8 warnings: 37308 37309 dmxcb.c:50:12: warning: redundant redeclaration of ‘PanoramiXPixWidth’ 37310 [-Wredundant-decls] 37311 extern int PanoramiXPixWidth; 37312 ^ 37313 In file included from dmxcb.c:49:0: 37314 ../../Xext/panoramiXsrv.h:12:22: note: previous declaration of 37315 ‘PanoramiXPixWidth’ was here 37316 extern _X_EXPORT int PanoramiXPixWidth; 37317 ^ 37318 dmxcb.c:51:12: warning: redundant redeclaration of ‘PanoramiXPixHeight’ 37319 [-Wredundant-decls] 37320 extern int PanoramiXPixHeight; 37321 ^ 37322 In file included from dmxcb.c:49:0: 37323 ../../Xext/panoramiXsrv.h:13:22: note: previous declaration of 37324 ‘PanoramiXPixHeight’ was here 37325 extern _X_EXPORT int PanoramiXPixHeight; 37326 ^ 37327 dmxcb.c:52:12: warning: redundant redeclaration of ‘PanoramiXNumScreens’ 37328 [-Wredundant-decls] 37329 extern int PanoramiXNumScreens; 37330 ^ 37331 In file included from dmxcb.c:49:0: 37332 ../../Xext/panoramiXsrv.h:11:22: note: previous declaration of 37333 ‘PanoramiXNumScreens’ was here 37334 extern _X_EXPORT int PanoramiXNumScreens; 37335 ^ 37336 37337 dmxpict.c:60:12: warning: redundant redeclaration of ‘RenderErrBase’ 37338 [-Wredundant-decls] 37339 extern int RenderErrBase; 37340 ^ 37341 In file included from ../../render/glyphstr.h:29:0, 37342 from ../../render/picturestr.h:28, 37343 from dmx.h:65, 37344 from dmxpict.c:42: 37345 ../../render/picture.h:176:22: note: previous declaration of ‘RenderErrBase’ 37346 was here 37347 extern _X_EXPORT int RenderErrBase; 37348 ^ 37349 37350 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37351 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37352 Reviewed-by: Keith Packard <keithp@keithp.com> 37353 37354commit 1033b85196e0f25877b4b9081993aff16bf32138 37355Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37356Date: Fri Dec 12 22:13:28 2014 -0800 37357 37358 dmx: remove redundant declaration of dmxFontPrivateIndex from dmxfont.h 37359 37360 Gets rid of 8 instances of gcc 4.8 warning: 37361 37362 In file included from glxcmds.c:38:0: 37363 ../../../hw/dmx/dmxfont.h:57:12: warning: redundant redeclaration of 37364 ‘dmxFontPrivateIndex’ [-Wredundant-decls] 37365 extern int dmxFontPrivateIndex; 37366 ^ 37367 In file included from glxcmds.c:35:0: 37368 ../../../hw/dmx/dmx.h:388:12: note: previous declaration of 37369 ‘dmxFontPrivateIndex’ was here 37370 extern int dmxFontPrivateIndex; /**< Private index for Fonts */ 37371 ^ 37372 37373 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37374 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37375 Reviewed-by: Keith Packard <keithp@keithp.com> 37376 37377commit ac64a653fb4935a799baca7545ba6ca401e20fa1 37378Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37379Date: Fri Dec 12 22:09:14 2014 -0800 37380 37381 dmx: glxproxy prototype cleanup 37382 37383 Gets rid of 107 -Wmissing-prototypes & -Wredundant-decls warnings 37384 from gcc 4.8 37385 37386 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37387 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37388 Reviewed-by: Keith Packard <keithp@keithp.com> 37389 37390commit bd563fd48f909094d6660854e97811ad7e676ddf 37391Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37392Date: Fri Dec 12 21:11:17 2014 -0800 37393 37394 dmx: more unshadowing of variables 37395 37396 Gets rid of gcc 4.8 warnings: 37397 37398 dmxinputinit.c: In function ‘dmxInputScanForExtensions’: 37399 dmxinputinit.c:877:14: warning: declaration of ‘display’ shadows a global declaration [-Wshadow] 37400 Display *display; 37401 ^ 37402 In file included from ../../../include/windowstr.h:60:0, 37403 from dmxinputinit.c:72: 37404 ../../../include/opaque.h:52:30: warning: shadowed declaration is here [-Wshadow] 37405 extern _X_EXPORT const char *display; 37406 ^ 37407 37408 glxcmds.c: In function ‘__glXCreatePbuffer’: 37409 glxcmds.c:3397:21: warning: declaration of ‘pc’ shadows a parameter [-Wshadow] 37410 CARD32 *pc = (CARD32 *) (be_req + 1); 37411 ^ 37412 glxcmds.c:3314:52: warning: shadowed declaration is here [-Wshadow] 37413 __glXCreatePbuffer(__GLXclientState * cl, GLbyte * pc) 37414 ^ 37415 37416 glxscreens.c: In function ‘CalcServerVersionAndExtensions’: 37417 glxscreens.c:139:35: warning: declaration of ‘req’ shadows a previous local [-Wshadow] 37418 xGLXQueryServerStringReq *req; 37419 ^ 37420 glxscreens.c:68:26: warning: shadowed declaration is here [-Wshadow] 37421 xGLXQueryVersionReq *req; 37422 ^ 37423 glxscreens.c:140:36: warning: declaration of ‘reply’ shadows a previous local [-Wshadow] 37424 xGLXQueryServerStringReply reply; 37425 ^ 37426 glxscreens.c:69:27: warning: shadowed declaration is here [-Wshadow] 37427 xGLXQueryVersionReply reply; 37428 ^ 37429 37430 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37431 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37432 Reviewed-by: Keith Packard <keithp@keithp.com> 37433 37434commit 57e08fae82f733304200a473f55b86e689404c13 37435Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37436Date: Fri Dec 12 21:07:12 2014 -0800 37437 37438 dmx: attempt to untangle nested loops using same index variable 37439 37440 This doesn't just make gcc sad, it makes my brain sad. 37441 37442 Change from: 37443 for (i = 0; i < dmxNumScreens; i++) { 37444 int i; 37445 for (i = 0; i < nconfigs; i++) { 37446 for (j = 0; j < dmxScreen->beNumVisuals; j++) { 37447 37448 to the easier to follow: 37449 for (i = 0; i < dmxNumScreens; i++) { 37450 for (j = 0; j < nconfigs; j++) { 37451 for (k = 0; k < dmxScreen->beNumVisuals; k++) { 37452 37453 Gets rid of gcc 4.8 warning: 37454 37455 dmxinit.c: In function ‘InitOutput’: 37456 dmxinit.c:765:17: warning: declaration of ‘i’ shadows a previous local [-Wshadow] 37457 int i; 37458 ^ 37459 dmxinit.c:608:9: warning: shadowed declaration is here [-Wshadow] 37460 int i; 37461 ^ 37462 37463 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37464 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37465 Reviewed-by: Keith Packard <keithp@keithp.com> 37466 37467commit 0fbebad72428abbc9fc3fa9f406f7a7e1b9d95b2 37468Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37469Date: Fri Dec 12 20:53:33 2014 -0800 37470 37471 dmx: remove unused variables 37472 37473 Gets rid of gcc 4.8 warnings: 37474 37475 dmxscrinit.c: In function ‘dmxBEScreenInit’: 37476 dmxscrinit.c:83:15: warning: unused variable ‘gcvals’ [-Wunused-variable] 37477 XGCValues gcvals; 37478 ^ 37479 37480 dmxwindow.c: In function ‘dmxResizeWindow’: 37481 dmxwindow.c:860:19: warning: variable ‘pSibPriv’ set but not used [-Wunused-but-set-variable] 37482 dmxWinPrivPtr pSibPriv; 37483 ^ 37484 37485 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37486 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37487 Reviewed-by: Keith Packard <keithp@keithp.com> 37488 37489commit f358f0d50cadaf10f7d97e21da00419a4a37c4e1 37490Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37491Date: Fri Dec 12 19:09:36 2014 -0800 37492 37493 dmx: fix printf format argument warnings 37494 37495 Gets rid of these gcc 4.8 warnings: 37496 37497 dmxinit.c: In function ‘dmxErrorHandler’: 37498 dmxinit.c:167:16: warning: format ‘%x’ expects argument of type ‘unsigned int’, 37499 but argument 3 has type ‘XID64’ [-Wformat=] 37500 ev->resourceid); 37501 ^ 37502 dmxinit.c:171:16: warning: format ‘%x’ expects argument of type ‘unsigned int’, 37503 but argument 3 has type ‘XID64’ [-Wformat=] 37504 ev->resourceid); 37505 ^ 37506 dmxinit.c:175:16: warning: format ‘%x’ expects argument of type ‘unsigned int’, 37507 but argument 3 has type ‘XID64’ [-Wformat=] 37508 ev->resourceid); 37509 ^ 37510 dmxinit.c:181:12: warning: format ‘%d’ expects argument of type ‘int’, 37511 but argument 3 has type ‘long unsigned int’ [-Wformat=] 37512 ev->serial); 37513 ^ 37514 dmxinit.c:183:12: warning: format ‘%d’ expects argument of type ‘int’, 37515 but argument 3 has type ‘long unsigned int’ [-Wformat=] 37516 dpy->request); 37517 ^ 37518 dmxinit.c: In function ‘InitOutput’: 37519 dmxinit.c:637:9: warning: format ‘%d’ expects argument of type ‘int’, 37520 but argument 3 has type ‘long unsigned int’ [-Wformat=] 37521 dmxLog(dmxInfo, "Generation: %d\n", dmxGeneration); 37522 ^ 37523 37524 dmxprop.c: In function ‘dmxPropertyCheckOtherWindows’: 37525 dmxprop.c:223:24: warning: format ‘%lu’ expects argument of type ‘long 37526 unsigned int’, but argument 4 has type ‘Window’ [-Wformat=] 37527 dmxScreen->name, win, tp.value); 37528 ^ 37529 dmxprop.c: In function ‘dmxPropertyWindow’: 37530 dmxprop.c:364:16: warning: format ‘%lu’ expects argument of type ‘long 37531 unsigned int’, but argument 5 has type ‘Window’ [-Wformat=] 37532 other->index, other->name, other->scrnWin); 37533 ^ 37534 dmxprop.c:364:16: warning: format ‘%lu’ expects argument of type ‘long 37535 unsigned int’, but argument 8 has type ‘Window’ [-Wformat=] 37536 37537 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37538 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37539 Reviewed-by: Keith Packard <keithp@keithp.com> 37540 37541commit 11af200b004b4810c9267e3e16449c3175c64d4a 37542Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37543Date: Fri Dec 12 18:58:01 2014 -0800 37544 37545 dmx: move format strings inline so gcc can check 37546 37547 Gets rid of gcc 4.8 warnings: 37548 37549 dmxprint.c: In function ‘dmxConfigPrintPair’: 37550 dmxprint.c:284:25: warning: format not a string literal, 37551 argument types not checked [-Wformat-nonliteral] 37552 p->ysign < 0 ? '-' : '+', p->y); 37553 ^ 37554 dmxprint.c:289:9: warning: format not a string literal, 37555 argument types not checked [-Wformat-nonliteral] 37556 dmxConfigOutput(addSpace, 0, p->comment, format, p->x, p->y); 37557 ^ 37558 37559 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37560 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37561 Reviewed-by: Keith Packard <keithp@keithp.com> 37562 37563commit 5e01eac10e915568324aff81b51d3af558757042 37564Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37565Date: Fri Dec 12 18:52:05 2014 -0800 37566 37567 dmx: Add _X_ATTRIBUTE_PRINTF to dmxConfigLog() & dmxConfigOutput() 37568 37569 Gets rid of gcc 4.8 warnings: 37570 37571 dmxparse.c: In function ‘dmxConfigLog’: 37572 dmxparse.c:61:5: warning: function might be possible candidate for 37573 ‘gnu_printf’ format attribute [-Wsuggest-attribute=format] 37574 vprintf(format, args); /* RATS: All calls to dmxConfigLog from 37575 ^ 37576 37577 dmxprint.c: In function ‘dmxConfigOutput’: 37578 dmxprint.c:149:9: warning: function might be possible candidate for 37579 ‘gnu_printf’ format attribute [-Wsuggest-attribute=format] 37580 pos += vfprintf(str, format, args); /* assumes no newlines! */ 37581 37582 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37583 Reviewed-by: Rémi Cardona <remi@gentoo.org> 37584 Reviewed-by: Keith Packard <keithp@keithp.com> 37585 37586commit 9e002dfcd71439fd81563db1f70e8c1964d52fab 37587Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37588Date: Tue Dec 9 18:27:11 2014 -0800 37589 37590 Get rid of const warnings in XSERVER_INPUT_EVENT dtrace probe calls 37591 37592 Use typedefs to work around dtrace dropping const qualifiers from probe 37593 arguments when generating Xserver-dtrace.h. Add new probes.h header to 37594 avoid having to replicate these typedefs in every file with dtrace probes. 37595 37596 Gets rid of these warnings from gcc 4.8: 37597 getevents.c:1096:9: 37598 warning: passing argument 6 of '__dtrace_Xserver___input__event' discards 37599 'const' qualifier from pointer target type [enabled by default] 37600 getevents.c:1096:9: 37601 warning: passing argument 7 of '__dtrace_Xserver___input__event' disards 37602 'const' qualifier from pointer target type [enabled by default] 37603 getevents.c:1651:9: 37604 warning: passing argument 6 of '__dtrace_Xserver___input__event' disards 37605 'const' qualifier from pointer target type [enabled by default] 37606 getevents.c:1651:9: 37607 warning: passing argument 7 of '__dtrace_Xserver___input__event' disards 37608 'const' qualifier from pointer target type [enabled by default] 37609 getevents.c:1791:9: 37610 warning: passing argument 6 of '__dtrace_Xserver___input__event' disards 37611 'const' qualifier from pointer target type [enabled by default] 37612 getevents.c:1791:9: 37613 warning: passing argument 7 of '__dtrace_Xserver___input__event' disards 37614 'const' qualifier from pointer target type [enabled by default] 37615 getevents.c:1921:9: 37616 warning: passing argument 6 of '__dtrace_Xserver___input__event' disards 37617 'const' qualifier from pointer target type [enabled by default] 37618 getevents.c:1921:9: 37619 warning: passing argument 7 of '__dtrace_Xserver___input__event' disards 37620 'const' qualifier from pointer target type [enabled by default] 37621 37622 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37623 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 37624 37625commit 20079c36cf7d377938ca5478447d8b9045cb7d43 37626Author: Olivier Fourdan <ofourdan@redhat.com> 37627Date: Fri Jan 16 08:44:45 2015 +0100 37628 37629 xkb: Check strings length against request size 37630 37631 Ensure that the given strings length in an XkbSetGeometry request remain 37632 within the limits of the size of the request. 37633 37634 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 37635 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 37636 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 37637 37638commit 81c90dc8f0aae3b65730409b1b615b5fa7280ebd 37639Author: Olivier Fourdan <ofourdan@redhat.com> 37640Date: Fri Jan 16 20:08:59 2015 +0100 37641 37642 xkb: Don't swap XkbSetGeometry data in the input buffer 37643 37644 The XkbSetGeometry request embeds data which needs to be swapped when the 37645 server and the client have different endianess. 37646 37647 _XkbSetGeometry() invokes functions that swap these data directly in the 37648 input buffer. 37649 37650 However, ProcXkbSetGeometry() may call _XkbSetGeometry() more than once 37651 (if there is more than one keyboard), thus causing on swapped clients the 37652 same data to be swapped twice in memory, further causing a server crash 37653 because the strings lengths on the second time are way off bounds. 37654 37655 To allow _XkbSetGeometry() to run reliably more than once with swapped 37656 clients, do not swap the data in the buffer, use variables instead. 37657 37658 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 37659 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 37660 37661commit 28f6427aec1f5a1982e1c01eff45af0d401bf659 37662Author: Keith Packard <keithp@keithp.com> 37663Date: Mon Feb 2 07:41:06 2015 +0100 37664 37665 Update to version 1.17.0 37666 37667 Signed-off-by: Keith Packard <keithp@keithp.com> 37668 37669commit 697b696e5e24d0679f133183a3bb0852025377c2 37670Author: Dave Airlie <airlied@redhat.com> 37671Date: Fri Jan 30 09:59:49 2015 +1000 37672 37673 config/udev: Respect seat assignments when assigned devices 37674 37675 Jonathan Dieter posted a few patches to do this inside the Xorg 37676 server but it makes no sense to do it there, just have the code 37677 we use to probe the device list at startup check seat assignments 37678 using the same code we check at hotplug time. 37679 37680 Bugilla: https://bugzilla.redhat.com/show_bug.cgi?id=1183654 37681 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 37682 Acked-by: Hans de Goede <hdegoede@redhat.com> 37683 Tested-by: Jonathan Dieter <jdieter@lesbg.com> 37684 Signed-off-by: Dave Airlie <airlied@redhat.com> 37685 Signed-off-by: Keith Packard <keithp@keithp.com> 37686 37687commit df1b401f57ad4b4925bad66684445b476562f26f 37688Author: Dave Airlie <airlied@redhat.com> 37689Date: Wed Jan 7 09:19:27 2015 +1000 37690 37691 randr: attempt to fix primary on slave output (v2) 37692 37693 If the user wants to set one of the slave devices as 37694 the primary output, we shouldn't fail to do so, 37695 we were returning BadMatch which was tripping up 37696 gnome-settings-daemon and bad things ensues. 37697 37698 Fix all the places we use primaryOutput to work 37699 out primaryCrtc and take it into a/c when slave 37700 gpus are in use. 37701 37702 v2: review from Aaron, fix indent, unhide has_primary from 37703 macro. I left the int vs Bool alone to be consistent with 37704 code below, a future patch could fix both. 37705 37706 Signed-off-by: Dave Airlie <airlied@redhat.com> 37707 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 37708 Signed-off-by: Keith Packard <keithp@keithp.com> 37709 37710commit 62fcd364ac8c71a2db1db84b17b17cade6832492 37711Author: Adel Gadllah <adel.gadllah@gmail.com> 37712Date: Sat Jan 3 21:12:25 2015 +0100 37713 37714 dri2: Set vdpau driver name if ddx does not provide any driver name 37715 37716 Currently when the ddx does not set any driver name we set DRI2 driver but 37717 not the VDPAU driver name. The result is that VDPAU drivers will not get found 37718 by libvdpau when the modesetting driver is being used. 37719 37720 Just assume that the VDPAU driver matches the DRI2 driver name, this is true 37721 for nouveau, r300, r600 and radeonsi i.e all VDPAU drivers currently supported 37722 by mesa. 37723 37724 Signed-off-by: Adel Gadllah <adel.gadllah@gmail.com> 37725 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 37726 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37727 Signed-off-by: Keith Packard <keithp@keithp.com> 37728 37729commit fe4c774c572e3f55a7417f0ca336ae1479a966ad 37730Author: Nikhil Mahale <nmahale@nvidia.com> 37731Date: Sat Jan 24 17:06:59 2015 -0800 37732 37733 os: Fix timer race conditions 37734 37735 Fixing following kind of race-conditions - 37736 37737 WaitForSomething() 37738 | 37739 ----> // timers -> timer-1 -> timer-2 -> null 37740 while (timers && (int) (timers->expires - now) <= 0) 37741 // prototype - DoTimer(OsTimerPtr timer, CARD32 now, OsTimerPtr *prev) 37742 DoTimer(timers, now, &timers) 37743 | 37744 | 37745 ----> OsBlockSignals(); .... OS Signal comes just before blocking it, 37746 .... timer-1 handler gets called. 37747 // timer-1 gets served and scheduled again; 37748 // timers -> timer-2 -> timer-1 -> null 37749 .... 37750 *prev = timer->next; 37751 timer->next = NULL; // timers -> null 37752 // timers list gets corrupted here and timer-2 gets removed from list. 37753 37754 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=86288 37755 Signed-off-by: Nikhil Mahale <nmahale@nvidia.com> 37756 Reviewed-by: Julien Cristau <jcristau@debian.org> 37757 37758 v2: Apply warning fixes from Keith Packard <keithp@keithp.com> 37759 37760 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 37761 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 37762 Signed-off-by: Keith Packard <keithp@keithp.com> 37763 37764commit 58f28b0427f0a0c0c445f314bd42721ca8e1e844 37765Author: Keith Packard <keithp@keithp.com> 37766Date: Fri Jan 23 10:59:39 2015 -0800 37767 37768 Update to version 1.16.99.902 37769 37770 Signed-off-by: Keith Packard <keithp@keithp.com> 37771 37772commit fef2f6357b40b238ae01c4c80b0d29b17b839686 37773Author: Jason Ekstrand <jason@jlekstrand.net> 37774Date: Tue Jan 13 15:08:38 2015 -0800 37775 37776 modesetting: Return the crtc for a drawable even if it's rotated 37777 37778 All of our checks for what crtc we are on take rotation into account so we 37779 select the correct crtc. The only problem is that we weren't returning it 37780 we were rotated. This caused X to think DRI3 apps were not on any crtc and 37781 limit them to 1 FPS. 37782 37783 Signed-off-by: Jason Ekstrand <jason.ekstrand@intel.com> 37784 Reviewed-by: Keith Packard <keithp@keithp.com> 37785 Signed-off-by: Keith Packard <keithp@keithp.com> 37786 37787commit 3dcd591fa9b71a3dce58d612ca5970209d8386eb 37788Author: Jason Ekstrand <jason@jlekstrand.net> 37789Date: Tue Jan 13 15:08:37 2015 -0800 37790 37791 modesetting: Add support for using RandR shadow buffers 37792 37793 This replaces the stubs for shadow buffer creation/allocation with actual 37794 functions and adds a shadow_destroy function. With this, we actually get 37795 shadow buffers and RandR now works properly. Most of this is copied from 37796 the xf86-video-intel driver and modified for modesetting. 37797 37798 v2 Jason Ekstrand <jason.ekstrand@intel.com>: 37799 - Fix build with --disable-glamor 37800 - Set the pixel data pointer in the pixmap header for dumb shadow bo's 37801 - Call drmmode_create_bo with the right bpp 37802 37803 v2 Jason Ekstrand <jason.ekstrand@intel.com>: 37804 - Make shadow buffers per-crtc and leave shadow_enable alone 37805 37806 Signed-off-by: Jason Ekstrand <jason.ekstrand@intel.com> 37807 Reviewed-by: Keith Packard <keithp@keithp.com> 37808 Signed-off-by: Keith Packard <keithp@keithp.com> 37809 37810commit 7c656bfcae1d68aeffd5e202b3c1569885f5d13d 37811Author: Jason Ekstrand <jason@jlekstrand.net> 37812Date: Tue Jan 13 15:08:36 2015 -0800 37813 37814 modesetting: Add drmmode_bo_has_bo and drmmode_bo_map helper function 37815 37816 Signed-off-by: Jason Ekstrand <jason.ekstrand@intel.com> 37817 Reviewed-by: Keith Packard <keithp@keithp.com> 37818 Signed-off-by: Keith Packard <keithp@keithp.com> 37819 37820commit b4703a5a6e529b78810db8d8782317f0b4e2f265 37821Author: Jason Ekstrand <jason@jlekstrand.net> 37822Date: Tue Jan 13 15:08:35 2015 -0800 37823 37824 modesetting: Refactor drmmode_glamor_new_screen_pixmap 37825 37826 The original drmmode_glamor_new_screen_pixmap function was specific to the 37827 primary screen pixmap. This commit pulls the guts out into a new, more 37828 general, drmmode_set_pixmap_bo function for setting a buffer on a pixmap. 37829 The new function also properly tears down the glamor bits if the buffer 37830 being set is NULL. The drmmode_glamor_new_screen_pixmap function is now 37831 just a 3-line wrapper around drmmode_set_pixmap_bo. 37832 37833 v2 Jason Ekstrand <jason.ekstrand@intel.com>: 37834 - Re-arranged code in drmmode_set_pixmap_bo and 37835 drmmode_glamor_handle_new_screen_pixmap so that glamor_set_screen_pixmap 37836 only gets called for the screen pixmap 37837 - Guard the call to glamor_set_screen_pixmapa with a drmmode->glamor check 37838 37839 Signed-off-by: Jason Ekstrand <jason.ekstrand@intel.com> 37840 Reviewed-by: Keith Packard <keithp@keithp.com> 37841 Signed-off-by: Keith Packard <keithp@keithp.com> 37842 37843commit bb23fbf5bb278113c9c481875423b4d128180972 37844Author: Adam Jackson <ajax@redhat.com> 37845Date: Mon Jan 5 16:48:11 2015 -0500 37846 37847 dix: make RegionInit legal C++ 37848 37849 The CVE fix in: 37850 37851 commit 97015a07b9e15d8ec5608b95d95ec0eb51202acb 37852 Author: Alan Coopersmith <alan.coopersmith@oracle.com> 37853 Date: Wed Jan 22 22:37:15 2014 -0800 37854 37855 dix: integer overflow in RegionSizeof() [CVE-2014-8092 3/4] 37856 37857 offended the C++ demons: 37858 37859 ../../include/regionstr.h:147:45: error: invalid conversion from 'void*' to 37860 'pixman_region16_data_t* {aka pixman_region16_data*}' [-fpermissive] 37861 37862 Normally this isn't a problem, because around here we have the sense and 37863 common decency to not use C++, but this does make tigervnc fail to build, 37864 which is a little rude of us. 37865 37866 Signed-off-by: Adam Jackson <ajax@redhat.com> 37867 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37868 Signed-off-by: Keith Packard <keithp@keithp.com> 37869 37870commit 082931014811e587a9734cbf4d88fd948979b641 37871Author: Chris Wilson <chris@chris-wilson.co.uk> 37872Date: Sat Jan 17 10:09:54 2015 +0000 37873 37874 dri2: SourceOffloads may be for DRI3 only 37875 37876 As a DDX may declare offload support without supporting DRI2 37877 (because it is using an alternative acceleration mechanism like DRI3), 37878 when iterating the list of offload_source Screens to find a matching 37879 DRI2 provider we need to check before assuming it is DRI2 capable. 37880 37881 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88514 37882 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 37883 Reviewed-by: Dave Airlie <airlied@redhat.com> 37884 Signed-off-by: Keith Packard <keithp@keithp.com> 37885 37886commit f27d743c1899f307ec8063febbb3198c8945d372 37887Author: Carlos Olmedo Escobar <carlos.olmedo.e@gmail.com> 37888Date: Wed Jan 21 01:44:54 2015 +0100 37889 37890 Avoid possible null pointer dereference. 37891 37892 Signed-off-by: Carlos Olmedo Escobar <carlos.olmedo.e@gmail.com> 37893 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 37894 Signed-off-by: Keith Packard <keithp@keithp.com> 37895 37896commit 437d2ec5f250f8ab4b44cbae56f938719802e1cc 37897Author: Carlos Sánchez de La Lama <csanchezdll@gmail.com> 37898Date: Wed Jan 21 10:22:05 2015 +0100 37899 37900 randr: swap num-preferred field on RRGetOutputInfo reply 37901 37902 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=88614 37903 Signed-off-by: Carlos Sánchez de La Lama <csanchezdll@gmail.com> 37904 Reviewed-by: Dave Airlie <airlied@redhat.com> 37905 Signed-off-by: Keith Packard <keithp@keithp.com> 37906 37907commit 3d12941b408de7a3bdc579e34e119f8aa81ea926 37908Author: Keith Packard <keithp@keithp.com> 37909Date: Thu Jan 22 22:28:34 2015 -0800 37910 37911 drivers/modesetting: Save current BlockHandler on return in msBlockHandler 37912 37913 If the BlockHandler chain is modified while it is active, we need to 37914 re-fetch the current value and store it in our private for use the 37915 next time through. 37916 37917 Signed-off-by: Dave Airlie <airlied@redhat.com> 37918 Signed-off-by: Keith Packard <keithp@keithp.com> 37919 37920commit b3e496c6d21058147de9a11b78860e73c74db3cb 37921Author: Dave Airlie <airlied@redhat.com> 37922Date: Fri Jan 23 16:28:28 2015 +1000 37923 37924 glamor: use screen blockhandler rather than dix one (v3) 37925 37926 This adds glamor into the block handler call chain 37927 in the correct place. 37928 37929 This should fix interactions between glamor and drivers 37930 requiring damage from glamor. 37931 37932 v2: okay don't consolidate, just leave things wierd for now 37933 remove blcokhandler in screen close. 37934 37935 v3: block handler wrapping the right way. 37936 37937 Signed-off-by: Dave Airlie <airlied@redhat.com> 37938 Reviewed-by: Keith Packard <keithp@keithp.com> 37939 Signed-off-by: Keith Packard <keithp@keithp.com> 37940 37941commit 5af2f5b7d2f955586d0cb40eb30812f1893db22e 37942Author: Markus Wick <markus@selfnet.de> 37943Date: Thu Jan 15 22:03:18 2015 +0100 37944 37945 xwayland: Set glamor filter to nearest 37946 37947 glEGLImageTargetTexture2DOES only set the first level. 37948 Mesa handles this new texture as incomplete and renders a black screen. 37949 We also want to prevent linear filtering. 37950 37951 https://bugs.freedesktop.org/show_bug.cgi?id=81800 37952 37953 Signed-off-by: Markus Wick <markus@selfnet.de> 37954 Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer@amd.com> 37955 Reviewed-by: Eric Anholt <eric@anholt.net> 37956 Signed-off-by: Keith Packard <keithp@keithp.com> 37957 37958commit 5f2e8ac51ccbf7c02f25c8cb7617df0238418cd1 37959Merge: 4e12d7b6f 430147950 37960Author: Keith Packard <keithp@keithp.com> 37961Date: Sat Jan 10 14:51:57 2015 +1300 37962 37963 Merge remote-tracking branch 'whot/for-keith' 37964 37965commit 4e12d7b6f4489fa06475465993a3e1e1d896390b 37966Author: Jasper St. Pierre <jstpierre@mecheye.net> 37967Date: Sun Jan 4 23:27:32 2015 -0800 37968 37969 modesetting: Update the cursor without hiding it 37970 37971 In the new KMS APIs, the legacy drmModeSetCursor ioctl actually waits 37972 for a vblank after changing the cursor image before returning, meaning 37973 that the X server, in attempting to hide the cursor before updating 37974 its image, actually makes that hide *visible* for a full vblank. 37975 37976 It's unknown why the X server does this by default, but turn it off. 37977 37978 If we're with a legacy driver that doesn't support the modern 37979 drmModeSetCursor by waiting for a vblank before returning, we're going 37980 to get a tiny bit of tearing on the cursor plane. But between tearing 37981 with a new cursor image and tearing with a blank cursor image, I'd 37982 rather the former. 37983 37984 The only proper solution to this is an atomic ioctl that page flips 37985 all planes, including the cursor plane, at vblank time and at the same 37986 time. 37987 37988 Signed-off-by: Keith Packard <keithp@keithp.com> 37989 Reviewed-by: Keith Packard <keithp@keithp.com> 37990 37991commit 43014795087a0a8774dd9687f5967329b15f06a2 37992Author: Olivier Fourdan <ofourdan@redhat.com> 37993Date: Mon Jan 5 16:44:22 2015 +0100 37994 37995 Synchronize capslock in Xnest and Xephyr 37996 37997 In Xnest or Xephyr, pressing CapsLock when focus is on another 37998 window does not update the state in the nested X server. 37999 38000 This is because when synchronizing the lock modifier, sending a 38001 keypress or a key release only is not sufficient to toggle the state, 38002 unlike regular modifiers, one has to emulate a full press/release 38003 to lock or unlock the modifier. 38004 38005 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 38006 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 38007 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 38008 38009commit 24b943132f90bc72ce8b5dc954fe9ee8484edfc2 38010Author: Olivier Fourdan <fourdan@xfce.org> 38011Date: Fri Jan 2 18:50:17 2015 +0100 38012 38013 Fix subwindow in Xi emulated events 38014 38015 Bug: 70790 38016 38017 Signed-off-by: Olivier Fourdan <fourdan@xfce.org> 38018 38019commit b058dec281568d6a9c5b5e230c20eed096cbdc6d 38020Author: Peter Hutterer <peter.hutterer@who-t.net> 38021Date: Mon Jan 5 11:19:46 2015 +1000 38022 38023 mi: fix accidental x/y coordinate swap 38024 38025 Reported-by: Adam Greenblatt <adam.greenblatt@gmail.com> 38026 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 38027 Reviewed-by: Eric Anholt <eric@anholt.net> 38028 Signed-off-by: Keith Packard <keithp@keithp.com> 38029 38030commit 23a11fd85e12e94d29ee6d33715ac49684867b16 38031Author: Keith Packard <keithp@keithp.com> 38032Date: Sun Jan 4 19:13:35 2015 -0800 38033 38034 doc: Create a script to filter xmlto output 38035 38036 This reduces the build log spam while still preserving the xmlto 38037 status to catch build failures correctly. 38038 38039 Signed-off-by: Keith Packard <keithp@keithp.com> 38040 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 38041 38042commit 1c01633877caa4239f901f02fbe113926318d030 38043Merge: 357385551 e774663fa 38044Author: Keith Packard <keithp@keithp.com> 38045Date: Sun Jan 4 17:02:25 2015 -0800 38046 38047 Merge remote-tracking branch 'ajax/xserver-next' 38048 38049commit 3573855514557a518de40a93b3c578f28c7d9c2b 38050Author: Olivier Fourdan <ofourdan@redhat.com> 38051Date: Wed Dec 3 13:49:37 2014 +0100 38052 38053 Remove explicit dependency on $(WAYLAND_LIBS) 38054 38055 Xwayland Makefile explicitely set its dependencies on 38056 WAYLAND_LIBS. If the ibrairies are installed in a non-standard 38057 path, WAYLAND_LIBS contains '-L/path/to/the/lib' which will fail 38058 at build time with: 38059 38060 "No rule to make target '-L/path/to/the/lib', needed by 'Xwayland'. 38061 Stop" 38062 38063 Remove that explicit dependency to avoid the problem (LDADD ought 38064 to be enough to get the right libraries linked). 38065 38066 Signed-off-by: Olivier Fourdan <ofourdan@redhat.com> 38067 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 38068 Signed-off-by: Keith Packard <keithp@keithp.com> 38069 38070commit de89c6b8c6e81bad131c7f432e355cb42d233e87 38071Author: Peter Hutterer <peter.hutterer@who-t.net> 38072Date: Tue Dec 16 13:59:45 2014 +1000 38073 38074 xfree86: rename Xorg.bin to Xorg 38075 38076 If the suid wrapper is enabled, /usr/bin/Xorg is just a shell script that 38077 execs either /usr/libexec/Xorg.bin directly or the Xorg.wrap binary which then 38078 execve's /usr/libexec/Xorg.bin. 38079 38080 Either way, we end up with Xorg.bin, which is problematic for two reasons: 38081 * ps shows the command as Xorg.bin 38082 * _COMM and _EXE in systemd's journal will both show Xorg.bin as well 38083 38084 There's not much we can do about the path, but having the actual command stay 38085 as Xorg means better compatibility to existing scripts. And, the reason for 38086 this path: the command 38087 journalctl _COMM=Xorg 38088 works universally, regardless of whether the wrapper is used or not. 38089 38090 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 38091 Reviewed-by: Keith Packard <keithp@keithp.com> 38092 Acked-by: Hans de Goede <hdegoede@redhat.com> 38093 38094commit ee21be1324de1d6ef14e529fed7b75992e971beb 38095Author: Peter Hutterer <peter.hutterer@who-t.net> 38096Date: Thu Dec 11 10:32:45 2014 +1000 38097 38098 dix: offset touch root coordinates by ScreenRec origins (#86655) 38099 38100 For two ScreenRecs abs pointer positioning was working fine, but touch events 38101 stuck to the lower/right edge on any screen but the one with a 0/0 origin. 38102 Cause is a missing offset by the screen coordinates, causing the root 38103 coordinates in the event to desktop-wide, not screen-wide. 38104 38105 Offset properly, just like we do for pointer events. 38106 38107 X.Org Bug 86655 <http://bugs.freedesktop.org/show_bug.cgi?id=86655> 38108 38109 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 38110 38111commit dc777c346d5d452a53b13b917c45f6a1bad2f20b 38112Author: Keith Packard <keithp@keithp.com> 38113Date: Sat Jan 3 08:46:45 2015 -0800 38114 38115 dix: Allow zero-height PutImage requests 38116 38117 The length checking code validates PutImage height and byte width by 38118 making sure that byte-width >= INT32_MAX / height. If height is zero, 38119 this generates a divide by zero exception. Allow zero height requests 38120 explicitly, bypassing the INT32_MAX check. 38121 38122 Signed-off-by: Keith Packard <keithp@keithp.com> 38123 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38124 38125commit 924996c41c419dda0f02a96aafdf52f7670ff4ea 38126Author: Michele Baldessari <michele@redhat.com> 38127Date: Wed Dec 3 11:53:10 2014 -0500 38128 38129 ephyr: Implement per-screen colormaps 38130 38131 Xephyr's pseudocolor emulation added in: 38132 38133 commit 81a3b6fe27567b4f91033ece69996aa6bf8d01a3 38134 Author: Matthew Allum <breakfast@10.am> 38135 Date: Mon Nov 8 22:39:47 2004 +0000 38136 38137 Add support to Xephyr for lower depths than hosts 38138 38139 only tracks one global colormap for the whole (Xephyr) display. Move 38140 this to per-screen state so each screen's colormap can be correct. 38141 38142 [ajax: rebased to 1.17, cleaned up commit message] 38143 Reviewed-by: Adam Jackson <ajax@redhat.com> 38144 Signed-off-by: Michele Baldessari <michele@redhat.com> 38145 Reviewed-by: Eric Anholt <eric@anholt.net> 38146 Signed-off-by: Keith Packard <keithp@keithp.com> 38147 38148commit 7b076fdfc0e7a98430b6cb1c43079c137d7d3a6c 38149Author: Carl Worth <cworth@cworth.org> 38150Date: Thu Dec 11 19:09:48 2014 -0800 38151 38152 os/xsha1.c: Add license and copyright attribution. 38153 38154 I'm interested in copying this code to the mesa project, but before 38155 doing that it seems prudent to have the license and copyright 38156 attributions in place before copying that. To get this list of names I 38157 went through: 38158 38159 git log -- os/xsha1.c 38160 and: 38161 git log -- render/glyph.c 38162 38163 Signed-off-by: Keith Packard <keithp@keithp.com> 38164 38165commit 66726064204b869a24e5a8dc6e24daa5acb92bab 38166Author: Kenneth Graunke <kenneth@whitecape.org> 38167Date: Tue Dec 30 11:51:43 2014 -0800 38168 38169 modesetting: Fix build with --disable-glamor. 38170 38171 present.c: In function 'ms_present_flush': 38172 present.c:204:9: error: implicit declaration of function 38173 'glamor_block_handler' 38174 38175 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=87858 38176 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38177 Reviewed-by: Eric Anholt <eric@anholt.net> 38178 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 38179 Reviewed-by: Rob Clark <robdclark@gmail.com> 38180 Signed-off-by: Keith Packard <keithp@keithp.com> 38181 38182commit d723928870eea9fc2554abe137d6e57f7ad29e98 38183Author: Keith Packard <keithp@keithp.com> 38184Date: Sat Dec 27 09:11:33 2014 -0800 38185 38186 glamor: Fix nlines in glamor_xv_put_image when src_y is odd 38187 38188 The number of lines of video to update in the texture needs to be 38189 computed from the height of the updated source, not the full height of 38190 the source. 38191 38192 Signed-off-by: Keith Packard <keithp@keithp.com> 38193 Reviewed-by: Eric Anholt <eric@anholt.net> 38194 38195commit 09230a2d435aa7f893ad5022220e3eb39ffb6db5 38196Author: Kenneth Graunke <kenneth@whitecape.org> 38197Date: Thu Dec 11 13:37:14 2014 -0800 38198 38199 modesetting: Add vblank synchronization support when using Present. 38200 38201 modesetting hooked up vblank support for DRI2, but was missing support 38202 for vblanks in Present. 38203 38204 This is mostly copy and pasted from Keith's code in the intel driver. 38205 38206 v2: Use ms_crtc_msc_to_kernel_msc in ms_present_queue_vblank to hook 38207 up the vblank_offset workaround for bogus MSC values (which the 38208 DRI2 code already did). 38209 38210 Also simplify the ms_present_get_crtc function. vblank.c already 38211 implements the functionality; we just need to convert types. 38212 38213 v3: Fix ms_flush_drm_events return code. I'd copied code where 0 meant 38214 success into a function that returned a boolean, so the return code 38215 was always backwards. 38216 38217 Also add DebugPresent calls in ms_present_vblank_{handler,abort}. 38218 38219 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38220 Reviewed-by: Keith Packard <keithp@keithp.com> 38221 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38222 Signed-off-by: Keith Packard <keithp@keithp.com> 38223 38224commit 3b5be33fc55b3090d996014fa56603edb9c1d411 38225Author: Kenneth Graunke <kenneth@whitecape.org> 38226Date: Thu Dec 11 13:48:04 2014 -0800 38227 38228 modesetting: Include glamor.h from driver.h. 38229 38230 We basically want it throughout the driver. 38231 38232 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38233 Reviewed-by: Keith Packard <keithp@keithp.com> 38234 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38235 Signed-off-by: Keith Packard <keithp@keithp.com> 38236 38237commit 7d1fc04d27857c999bd732d54a0b36f5079449b8 38238Author: Kenneth Graunke <kenneth@whitecape.org> 38239Date: Thu Dec 18 17:58:18 2014 -0800 38240 38241 modesetting: Check DPMS mode in ms_covering_crtc(). 38242 38243 crtc->enabled is insufficient; we should also make sure DPMS is on. 38244 38245 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38246 Reviewed-by: Keith Packard <keithp@keithp.com> 38247 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38248 Signed-off-by: Keith Packard <keithp@keithp.com> 38249 38250commit 8affaade2c127ea08989c86e7d71cc9da3db1824 38251Author: Kenneth Graunke <kenneth@whitecape.org> 38252Date: Thu Dec 18 17:55:29 2014 -0800 38253 38254 modesetting: Track the CRTC's DPMS mode. 38255 38256 We don't want to try to vblank synchronize to monitors which are off. 38257 38258 In order to handle that properly, we need to know the CRTC's DPMS mode. 38259 38260 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38261 Reviewed-by: Keith Packard <keithp@keithp.com> 38262 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38263 Signed-off-by: Keith Packard <keithp@keithp.com> 38264 38265commit b51f804b1ca826699296aff9de27d19257d78dae 38266Author: Kenneth Graunke <kenneth@whitecape.org> 38267Date: Thu Dec 18 17:22:00 2014 -0800 38268 38269 present: If present_queue_vblank() fails, do present_execute(). 38270 38271 Previously, if present_queue_vblank() failed, we simply dropped the 38272 present request on the floor, and returned an error. This was rather 38273 mean to clients - after presenting, they wait for a PresentComplete 38274 event to come back. But since the present never happens, they end up 38275 waiting forever, and lock up in poll(). 38276 38277 This patch falls back to present_execute if present_queue_vblank fails. 38278 We still print a debugging message to warn when queueing fails, which 38279 allows us to continue debugging problems, but makes Present robust 38280 enough to not lock up people's compositing manager when vblank bugs 38281 happen. 38282 38283 v2: Don't do present_queue_vblank() /and/ present_execute() (a bug that 38284 snuck in during last minute tidying). 38285 38286 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38287 Reviewed-by: Keith Packard <keithp@keithp.com> 38288 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38289 Signed-off-by: Keith Packard <keithp@keithp.com> 38290 38291commit f9e22cefcb9684bf495ada3ec4b0231ab2dc6cff 38292Author: Dima Ryazanov <dima@gmail.com> 38293Date: Mon Dec 22 11:35:29 2014 -0800 38294 38295 Fix "Back", "Forward", and other special mouse buttons in XWayland. 38296 38297 Currently, the indexes are off by 4 because of the scroll buttons. 38298 38299 Signed-off-by: Dima Ryazanov <dima@gmail.com> 38300 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 38301 Signed-off-by: Keith Packard <keithp@keithp.com> 38302 38303commit 70a6f65f9e2b26ef7539dcacfcfea927bc1f13fd 38304Author: Michel Dänzer <michel.daenzer@amd.com> 38305Date: Thu Dec 25 11:42:03 2014 +0900 38306 38307 glamor: Make sure Xvideo source image data is properly aligned 38308 38309 _glamor_upload_bits_to_pixmap_texture currently ignores the stride 38310 parameter, but __glamor_upload_pixmap_to_texture uses 4-byte alignment 38311 via glPixelStorei(GL_UNPACK_ALIGNMENT, 4). 38312 38313 Also fix up the stride argument passed in though, in case it starts 38314 being used properly in the future. 38315 38316 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=87455 38317 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 38318 Reviewed-by: Keith Packard <keithp@keithp.com> 38319 Signed-off-by: Keith Packard <keithp@keithp.com> 38320 38321commit 11b85ab2fc944c0645df43180a6c7e2d6a06483a 38322Author: Michel Dänzer <michel.daenzer@amd.com> 38323Date: Wed Dec 24 11:46:46 2014 +0900 38324 38325 glamor: Make glamor_purge_fbo static 38326 38327 Only called from glamor_fbo.c now. 38328 38329 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 38330 Reviewed-by: Keith Packard <keithp@keithp.com> 38331 Signed-off-by: Keith Packard <keithp@keithp.com> 38332 38333commit 8323d2e90150bb86a12d00a16269c69a24efc896 38334Author: Michel Dänzer <michel.daenzer@amd.com> 38335Date: Wed Dec 24 11:46:18 2014 +0900 38336 38337 glamor: Call glamor_pixmap_destroy_fbo from glamor_set_pixmap_private 38338 38339 Calling glamor_purge_fbo directly was incorrect for large pixmaps. 38340 38341 Fixes use-after free with large pixmaps: 38342 38343 ==2029== Invalid write of size 8 ~ 38344 ==2029== at 0x85F93AD: __xorg_list_del (list.h:184) 38345 ==2029== by 0x85F93AD: xorg_list_del (list.h:204) 38346 ==2029== by 0x85F93AD: glamor_fbo_expire (glamor_fbo.c:280) 38347 ==2029== by 0x85F95CA: glamor_pixmap_fbo_cache_put (glamor_fbo.c:159) 38348 ==2029== by 0x85D7AB5: glamor_destroy_textured_pixmap (glamor.c:228) 38349 ==2029== by 0xC1BDDC4: radeon_glamor_destroy_pixmap (radeon_glamor.c:272) 38350 ==2029== by 0x519D00: damageDestroyPixmap (damage.c:1473) 38351 ==2029== by 0x4DD307: XvDestroyPixmap (xvmain.c:370) 38352 ==2029== by 0x4DB975: ShmDestroyPixmap (shm.c:258) 38353 ==2029== by 0x5098F6: FreePicture (picture.c:1425) 38354 ==2029== by 0x85E678E: glamor_composite_clipped_region (glamor_render.c:1558) 38355 ==2029== by 0x85F763A: glamor_composite_largepixmap_region (glamor_largepixmap.c:1347) 38356 ==2029== by 0x85E7964: _glamor_composite (glamor_render.c:1679) 38357 ==2029== by 0x85E7A38: glamor_composite (glamor_render.c:1758) 38358 ==2029== Address 0x1141d3c0 is 0 bytes inside a block of size 64 free'd 38359 ==2029== at 0x4C29E90: free (vg_replace_malloc.c:473) 38360 ==2029== by 0x85D7167: glamor_set_pixmap_private (glamor.c:570) 38361 ==2029== by 0xC1BDDC4: radeon_glamor_destroy_pixmap (radeon_glamor.c:272) 38362 ==2029== by 0x519D00: damageDestroyPixmap (damage.c:1473) 38363 ==2029== by 0x4DD307: XvDestroyPixmap (xvmain.c:370) 38364 ==2029== by 0x4DB975: ShmDestroyPixmap (shm.c:258) 38365 ==2029== by 0x45B246: doFreeResource (resource.c:875) 38366 ==2029== by 0x45BD5E: FreeResource (resource.c:905) 38367 ==2029== by 0x43444B: ProcFreePixmap (dispatch.c:1422) 38368 ==2029== by 0x43856E: Dispatch (dispatch.c:432) 38369 ==2029== by 0x43C96F: dix_main (main.c:298) 38370 ==2029== by 0x6CFAB44: (below main) (libc-start.c:287) 38371 38372 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 38373 Reviewed-by: Keith Packard <keithp@keithp.com> 38374 Signed-off-by: Keith Packard <keithp@keithp.com> 38375 38376commit 0d37c7e4b5d63308a60ca59c9de981b380af8689 38377Author: Keith Packard <keithp@keithp.com> 38378Date: Fri Dec 19 18:40:19 2014 -0800 38379 38380 modesetting: Detect whether damage tracking is needed 38381 38382 Call drmModeDirtyFB and check the return value to detect whether the 38383 driver support for damage tracking is present, only initialize it in 38384 that case. 38385 38386 Signed-off-by: Keith Packard <keithp@keithp.com> 38387 Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com> 38388 38389commit 7804305673ca63720e8736c0c88afa711ebe208a 38390Author: Keith Packard <keithp@keithp.com> 38391Date: Fri Dec 19 19:34:34 2014 -0800 38392 38393 modesetting: Fix damage tracking auto-disable code 38394 38395 dispatch_dirty_region was only returning -EINVAL error codes, 38396 otherwise it would return 0. The kernel returns -ENOSYS when the 38397 driver doesn't support damage tracking, so dispatch_dirty would never 38398 see the error and never disable damage tracking. 38399 38400 Pass all errors back from dispatch_dirty_region and let dispatch_dirty 38401 deal with them. 38402 38403 Signed-off-by: Keith Packard <keithp@keithp.com> 38404 Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com> 38405 38406commit 826e7c2b36f192fbbe7ddff37eb559f4d6301146 38407Author: Alan Coopersmith <alan.coopersmith@oracle.com> 38408Date: Wed Dec 17 20:50:02 2014 -0800 38409 38410 Solaris: Move shared declarations to xf86_OSlib.h 38411 38412 Gets rid of gcc 4.8 warnings: 38413 xf86AutoConfig.c:211:9: warning: nested extern declaration of 38414 'xf86SolarisFbDev' [-Wnested-externs] 38415 sun_VTsw.c:44:1: warning: no previous prototype for 'xf86VTRelease' 38416 [-Wmissing-prototypes] 38417 sun_VTsw.c:59:1: warning: no previous prototype for 'xf86VTAcquire' 38418 [-Wmissing-prototypes] 38419 38420 and ensures caller & definition stay in sync. 38421 38422 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38423 Reviewed-by: Keith Packard <keithp@keithp.com> 38424 Signed-off-by: Keith Packard <keithp@keithp.com> 38425 38426commit bebcac0cf790b220e849e8905139d0744a606928 38427Author: Alan Coopersmith <alan.coopersmith@oracle.com> 38428Date: Wed Dec 17 20:50:01 2014 -0800 38429 38430 Move RTLD_DI_SETSIGNAL code into a separate block to quiet warning 38431 38432 Gets rid of gcc 4.8 warning: 38433 osinit.c:211:9: warning: ISO C90 forbids mixed declarations and code 38434 [-Wdeclaration-after-statement] 38435 38436 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38437 Reviewed-by: Keith Packard <keithp@keithp.com> 38438 Signed-off-by: Keith Packard <keithp@keithp.com> 38439 38440commit 2c7111235ca5c3bd23cd824589dfbc0ebd477585 38441Author: Keith Packard <keithp@keithp.com> 38442Date: Sat Dec 13 22:22:30 2014 -0800 38443 38444 modesetting: Enable Xv when using glamor 38445 38446 This just calls the existing function to create the relevant Xv 38447 adaptor and hook it up. 38448 38449 Signed-off-by: Keith Packard <keithp@keithp.com> 38450 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 38451 38452commit 5a541bd5e7adc9fdcc1883558be9218f82a284e0 38453Author: Keith Packard <keithp@keithp.com> 38454Date: Tue Dec 16 23:14:34 2014 -0800 38455 38456 modesetting: [v2] Don't re-enable the cursor when loading the image 38457 38458 Hidden cursors also have their image updated; re-enabling the cursor 38459 each time the image is set will cause it to re-appear. 38460 38461 * Unifies the code that was in drmmode_load_cursor_argb and 38462 drm_mode_show_cursor and moves it to a new drmmode_set_cursor 38463 38464 * Add a new boolean, 'cursor_up', to the per-crtc 38465 private data to track whether the cursor should be displayed. 38466 38467 * Call drmmode_set_cursor from drm_mode_show_cursor and, if 38468 the cursor should be displayed, from drm_mode_load_cursor_argb. 38469 38470 v2: Call drmModeSetCursor2 when loading a new cursor image if the 38471 cursor should be displayed. 38472 38473 Signed-off-by: Keith Packard <keithp@keithp.com> 38474 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 38475 38476commit 0f5fdaf600bfeada966aea942cb3e347d4efee30 38477Author: Michel Dänzer <michel.daenzer@amd.com> 38478Date: Thu Dec 11 16:50:40 2014 +0900 38479 38480 glamor: Make sure glamor_egl_close_screen wraps glamor_close_screen 38481 38482 The other way around fails to destroy the screen pixmap EGL image: 38483 38484 ==1782== 80 (32 direct, 48 indirect) bytes in 1 blocks are definitely lost in loss record 981 of 2,171 38485 ==1782== at 0x4C28C20: malloc (vg_replace_malloc.c:296) 38486 ==1782== by 0xF9D4BD2: dri2_create_image_from_dri (egl_dri2.c:1264) 38487 ==1782== by 0xF9D4BD2: dri2_create_image_dma_buf (egl_dri2.c:1764) 38488 ==1782== by 0xF9D4BD2: dri2_create_image_khr (egl_dri2.c:1798) 38489 ==1782== by 0xF9C7937: eglCreateImageKHR (eglapi.c:1494) 38490 ==1782== by 0x85D5655: _glamor_egl_create_image (glamor_egl.c:134) 38491 ==1782== by 0x85D5655: glamor_egl_create_textured_pixmap (glamor_egl.c:302) 38492 ==1782== by 0x85D579B: glamor_egl_create_textured_screen (glamor_egl.c:225) 38493 ==1782== by 0xC1BE05D: radeon_glamor_create_screen_resources (radeon_glamor.c:67) 38494 ==1782== by 0xC1B6153: RADEONCreateScreenResources_KMS (radeon_kms.c:258) 38495 ==1782== by 0x4B2105: xf86CrtcCreateScreenResources (xf86Crtc.c:709) 38496 ==1782== by 0x43C823: dix_main (main.c:223) 38497 ==1782== by 0x6CFAB44: (below main) (libc-start.c:287) 38498 38499 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 38500 Reviewed-by: Keith Packard <keithp@keithp.com> 38501 Signed-off-by: Keith Packard <keithp@keithp.com> 38502 38503commit c8d4c0a25dba49d1486761438b96374672b29645 38504Author: Michel Dänzer <michel.daenzer@amd.com> 38505Date: Thu Dec 11 16:50:39 2014 +0900 38506 38507 glamor: Make glamor_destroy_textured_pixmap idempotent 38508 38509 For robustness against drivers which may call both 38510 glamor_(egl_)destroy_textured_pixmap and glamor_destroy_pixmap. 38511 38512 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 38513 Reviewed-by: Keith Packard <keithp@keithp.com> 38514 Signed-off-by: Keith Packard <keithp@keithp.com> 38515 38516commit 0355e280a39dee68981b2cbebfa2573cfde2f5bd 38517Author: Michel Dänzer <michel.daenzer@amd.com> 38518Date: Thu Dec 11 16:50:38 2014 +0900 38519 38520 glamor: Make glamor_set_pixmap_private not crash if the pixmap has no fbo 38521 38522 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 38523 Reviewed-by: Keith Packard <keithp@keithp.com> 38524 Signed-off-by: Keith Packard <keithp@keithp.com> 38525 38526commit 45b333525e4626523f2b3b7bb360404b6366750d 38527Author: Michel Dänzer <michel.daenzer@amd.com> 38528Date: Thu Dec 11 16:50:37 2014 +0900 38529 38530 glamor: Fix use-after-free in glamor_destroy_textured_pixmap 38531 38532 ==25551== Invalid read of size 8 38533 ==25551== at 0x85D5F2C: glamor_egl_destroy_pixmap_image (glamor_egl.c:527) 38534 ==25551== by 0x85D7750: glamor_destroy_pixmap (glamor.c:235) 38535 ==25551== by 0xC1BDD9B: radeon_glamor_destroy_pixmap (radeon_glamor.c:278) 38536 ==25551== by 0x5098F6: FreePicture (picture.c:1425) 38537 ==25551== by 0x85DD7A9: glamor_unrealize_glyph_caches (glamor_glyphs.c:257) 38538 ==25551== by 0x85D7B50: glamor_close_screen (glamor.c:586) 38539 ==25551== by 0x4B1A82: xf86CrtcCloseScreen (xf86Crtc.c:734) 38540 ==25551== by 0x4CFFC7: CursorCloseScreen (cursor.c:187) 38541 ==25551== by 0x513A44: AnimCurCloseScreen (animcur.c:106) 38542 ==25551== by 0x51529B: present_close_screen (present_screen.c:64) 38543 ==25551== by 0x43CA83: dix_main (main.c:351) 38544 ==25551== by 0x6CFAB44: (below main) (libc-start.c:287) 38545 ==25551== Address 0x83dafa0 is 96 bytes inside a block of size 152 free'd 38546 ==25551== at 0x4C29E90: free (vg_replace_malloc.c:473) 38547 ==25551== by 0x85D76B4: glamor_destroy_textured_pixmap (glamor.c:225) 38548 ==25551== by 0x85D7750: glamor_destroy_pixmap (glamor.c:235) 38549 ==25551== by 0xC1BDD9B: radeon_glamor_destroy_pixmap (radeon_glamor.c:278) 38550 ==25551== by 0x5098F6: FreePicture (picture.c:1425) 38551 ==25551== by 0x85DD7A9: glamor_unrealize_glyph_caches (glamor_glyphs.c:257) 38552 ==25551== by 0x85D7B50: glamor_close_screen (glamor.c:586) 38553 ==25551== by 0x4B1A82: xf86CrtcCloseScreen (xf86Crtc.c:734) 38554 ==25551== by 0x4CFFC7: CursorCloseScreen (cursor.c:187) 38555 ==25551== by 0x513A44: AnimCurCloseScreen (animcur.c:106) 38556 ==25551== by 0x51529B: present_close_screen (present_screen.c:64) 38557 ==25551== by 0x43CA83: dix_main (main.c:351) 38558 38559 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 38560 Reviewed-by: Keith Packard <keithp@keithp.com> 38561 Signed-off-by: Keith Packard <keithp@keithp.com> 38562 38563commit 913416b718e73e7416c2128a82d2e5e5cd962d28 38564Author: Alan Coopersmith <alan.coopersmith@oracle.com> 38565Date: Tue Dec 9 10:15:08 2014 -0800 38566 38567 Solaris: delete undocumented, unuseful -protect0 flag 38568 38569 Solaris already makes the page at address 0 inaccessible by default to 38570 catch NULL pointer bugs, we don't need a double secret undocumented flag 38571 to try to make our own hacky attempt at it. 38572 38573 As a bonus, deleting this code removes gcc warning of: 38574 38575 sun_init.c: In function 'xf86OpenConsole': 38576 sun_init.c:103:17: warning: declaration of 'fd' shadows a previous local [-Wshadow] 38577 int fd = -1; 38578 ^ 38579 sun_init.c:89:9: warning: shadowed declaration is here [-Wshadow] 38580 int fd; 38581 ^ 38582 38583 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38584 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 38585 Signed-off-by: Keith Packard <keithp@keithp.com> 38586 38587commit 7b784df51bd6be38238d7101b1f5e320eb4aa5b8 38588Author: Kenneth Graunke <kenneth@whitecape.org> 38589Date: Tue Dec 9 12:43:57 2014 -0800 38590 38591 modesetting: Use GBM for buffer allocations if Glamor supports it. 38592 38593 For performance, Glamor wants to render to tiled buffers, not linear 38594 ones. Using GBM allows us to pick the 3D driver's preferred tiling 38595 modes. 38596 38597 v2: Declare drmmode->gbm as void * if !GLAMOR_HAS_GBM. 38598 v3: Just use a forward declaration of struct gbm_device. 38599 38600 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38601 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38602 Reviewed-by: Keith Packard <keithp@keithp.com> 38603 38604commit cfef64b0cabe7677c7584a72d7432c20343d9361 38605Author: Dave Airlie <airlied@redhat.com> 38606Date: Tue Dec 9 12:28:38 2014 -0800 38607 38608 glamor: Add an accessor for the GBM device. 38609 38610 (Originally written by Dave Airlie; split into a separate patch by 38611 Kenneth Graunke.) 38612 38613 Signed-off-by: Dave Airlie <airlied@redhat.com> 38614 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38615 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38616 Reviewed-by: Keith Packard <keithp@keithp.com> 38617 38618commit 980535757d38db5c812c8afa32726d8cc36abfa4 38619Author: Kenneth Graunke <kenneth@whitecape.org> 38620Date: Tue Dec 9 15:20:44 2014 -0800 38621 38622 modesetting: Create a drmmode_bo wrapper; use it for front_bo. 38623 38624 This code is going to be extended to support GBM BOs soon. This small 38625 abstraction removes a lot of direct dumb_bo access, so we can add that 38626 support in one place, rather than putting conditionals at every 38627 pitch/handle/etc access. 38628 38629 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38630 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38631 Reviewed-by: Keith Packard <keithp@keithp.com> 38632 38633commit c6388964b079b3828c0b64951ff03008dda8f44a 38634Author: Kenneth Graunke <kenneth@whitecape.org> 38635Date: Tue Dec 9 11:26:39 2014 -0800 38636 38637 modesetting: Drop dumb_bo::map_count field and dead unmap code. 38638 38639 The drm kernel API for dumb BOs apparently doesn't include an unmap 38640 ioctl, so we can't do much here. It looks like this code was copied 38641 from libkms, which was also unfinished. 38642 38643 We may as well delete the dead variable that simply gets incremented 38644 and never read. 38645 38646 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38647 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38648 Reviewed-by: Keith Packard <keithp@keithp.com> 38649 38650commit 87cc0c0d317c5eac4fcf9bf4a7ad36454c05f6b8 38651Author: Kenneth Graunke <kenneth@whitecape.org> 38652Date: Tue Dec 9 11:17:30 2014 -0800 38653 38654 modesetting: Move dumb_bo into its own source files. 38655 38656 Eventually, drmmode_display will be able to use GBM for handling 38657 buffers, and won't need dumb_bo. Keeping the display related logic 38658 and buffer object abstraction in separate files seems a bit tidier. 38659 38660 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38661 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38662 Reviewed-by: Keith Packard <keithp@keithp.com> 38663 38664commit af4f94b08eb8e5961799a8a74a06449fd3fe8ad9 38665Author: Kenneth Graunke <kenneth@whitecape.org> 38666Date: Tue Dec 9 15:18:39 2014 -0800 38667 38668 modesetting: Create helper for glamor_egl_create_textured_screen call. 38669 38670 This will need to change when we add GBM support; by pulling it into a 38671 helper function, we should only have to edit one place. 38672 38673 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38674 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38675 Reviewed-by: Keith Packard <keithp@keithp.com> 38676 38677commit b4324c6a2382dc96295bdb061316f699e4ffabb9 38678Author: Kenneth Graunke <kenneth@whitecape.org> 38679Date: Tue Dec 9 14:38:39 2014 -0800 38680 38681 modesetting: Move ModifyPixmapHeader calls out of if/else branches. 38682 38683 Both branches called ModifyPixmapHeader with essentially the same 38684 parameters. By using new_pixels in the shadowfb case, we can make 38685 them completely the same, and move them out a level, for simplicity. 38686 38687 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38688 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38689 Reviewed-by: Keith Packard <keithp@keithp.com> 38690 38691commit 35e9924484ec7bafb749c88ed7f78e6a71f8dfb3 38692Author: Kenneth Graunke <kenneth@whitecape.org> 38693Date: Tue Dec 9 13:01:09 2014 -0800 38694 38695 modesetting: Stop using glamor_egl_create_textured_screen_ext(). 38696 38697 The _ext variant takes an additional pointer argument, which it now 38698 ignores, thanks to Keith's recent patches. 38699 38700 Signed-off-by: Kenneth Graunke <kenneth@whitecape.org> 38701 Tested-by: Jason Ekstrand <jason.ekstrand@intel.com> 38702 Reviewed-by: Keith Packard <keithp@keithp.com> 38703 38704commit 91651e7c15892aa846fc406fbb13b37f094dd3f0 38705Author: Michel Dänzer <michel.daenzer@amd.com> 38706Date: Wed Dec 10 16:21:44 2014 +0900 38707 38708 glamor: Reinstate glamor_(egl_)destroy_textured_pixmap 38709 38710 They are part of the ABI. 38711 38712 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 38713 Reviewed-by: Keith Packard <keithp@keithp.com> 38714 Signed-off-by: Keith Packard <keithp@keithp.com> 38715 38716commit e774663fa5209ff469d920821934bb1f5964a72f 38717Author: Michele Baldessari <michele@redhat.com> 38718Date: Wed Dec 3 11:30:29 2014 -0500 38719 38720 ephyr: Implement per-screen colormaps 38721 38722 Xephyr's pseudocolor emulation added in: 38723 38724 commit 81a3b6fe27567b4f91033ece69996aa6bf8d01a3 38725 Author: Matthew Allum <breakfast@10.am> 38726 Date: Mon Nov 8 22:39:47 2004 +0000 38727 38728 Add support to Xephyr for lower depths than hosts 38729 38730 only tracks one global colormap for the whole (Xephyr) display. Move 38731 this to per-screen state so each screen's colormap can be correct. 38732 38733 [ajax: rebased to 1.17, cleaned up commit message] 38734 Reviewed-by: Adam Jackson <ajax@redhat.com> 38735 Reviewed-by: Eric Anholt <eric@anholt.net> 38736 Signed-off-by: Michele Baldessari <michele@redhat.com> 38737 38738commit cadd70c809232c9a6601fb8baab665a7ab10045d 38739Author: Adam Jackson <ajax@redhat.com> 38740Date: Tue Dec 2 14:52:35 2014 -0500 38741 38742 glx: Add hack for GLX-1.2-style naked windows to GetDrawableAttributes 38743 38744 Some people like to call this on bare Window XIDs and expect reasonable 38745 results. I sure wish they wouldn't, but since they do, if we're given 38746 a window without any glx decoration just fill in as much as we can. This 38747 means you won't actually get an answer for GLX_FBCONFIG_ID and friends, 38748 but there's not much to be done about that, and it matches what NVIDIA's 38749 driver seems to do. 38750 38751 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=54080 38752 Reviewed-by: Eric Anholt <eric@anholt.net> 38753 Signed-off-by: Adam Jackson <ajax@redhat.com> 38754 38755commit f452b4a47b3bebf883e84df804e8d54830ccfe83 38756Author: Adam Jackson <ajax@redhat.com> 38757Date: Tue Dec 2 14:47:45 2014 -0500 38758 38759 glx: Dynamically compute attribute slot in GetDrawableAttributes 38760 38761 No functional change. 38762 38763 Reviewed-by: Eric Anholt <eric@anholt.net> 38764 Signed-off-by: Adam Jackson <ajax@redhat.com> 38765 38766commit c1455f76c6b1aa4ecaacb2221a687244285aa44b 38767Author: Neil Roberts <neil@linux.intel.com> 38768Date: Mon Dec 1 16:06:17 2014 -0500 38769 38770 glx: Add implementation of __GLXContext->loseCurrent for direct ctxts 38771 38772 This adds a dummy implementation for the loseCurrent function in 38773 __GLXContext for direct contexts which just returns GL_TRUE. Without 38774 this then the X server can crash if receives a MakeCurrent message for 38775 a direct context because it will attempt to call loseCurrent when 38776 cleaning up the client in the callback for ClientStateGone. 38777 38778 [ajax: added assumed s-o-b line] 38779 38780 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=86531 38781 Reviewed-by: Adam Jackson <ajax@redhat.com> 38782 Signed-off-by: Neil Roberts <neil@linux.intel.com> 38783 Signed-off-by: Keith Packard <keithp@keithp.com> 38784 38785commit 6704bb0ed7a10dabe8ef3bb3adf8b8a7f29a78f0 38786Merge: 8aa23f27c 1559a9439 38787Author: Keith Packard <keithp@keithp.com> 38788Date: Tue Dec 9 11:47:11 2014 -0800 38789 38790 Merge remote-tracking branch 'alanc/master' 38791 38792commit 1559a94395258fd73e369f1a2c98a44bfe21a486 38793Author: Keith Packard <keithp@keithp.com> 38794Date: Tue Dec 9 09:31:00 2014 -0800 38795 38796 dix: GetHosts bounds check using wrong pointer value [CVE-2014-8092 pt. 6] 38797 38798 GetHosts saves the pointer to allocated memory in *data, and then 38799 wants to bounds-check writes to that region, but was mistakenly using 38800 a bare 'data' instead of '*data'. Also, data is declared as void **, 38801 so we need a cast to turn it into a byte pointer so we can actually do 38802 pointer comparisons. 38803 38804 Signed-off-by: Keith Packard <keithp@keithp.com> 38805 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38806 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38807 38808commit 9802a0162f738de03585ca3f3b8a8266494f7d45 38809Author: Keith Packard <keithp@keithp.com> 38810Date: Tue Dec 9 09:30:59 2014 -0800 38811 38812 Missing parens in REQUEST_FIXED_SIZE macro [CVE-2014-8092 pt. 5] 38813 38814 The 'n' parameter must be surrounded by parens in both places to 38815 prevent precedence from mis-computing things. 38816 38817 Signed-off-by: Keith Packard <keithp@keithp.com> 38818 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38819 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38820 38821commit 61b17c0f10307e25e51e30e6fb1d3e3127f82d86 38822Author: Keith Packard <keithp@keithp.com> 38823Date: Tue Dec 9 09:30:58 2014 -0800 38824 38825 glx: Can't mix declarations and code in X.org sources [CVE-2014-8098 pt. 9] 38826 38827 We're using compiler compatibility settings which generate warnings 38828 when a variable is declared after the first statement. 38829 38830 Signed-off-by: Keith Packard <keithp@keithp.com> 38831 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38832 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38833 38834commit b20912c3d45cbbde3c443e6c3d9e189092fe65e1 38835Author: Keith Packard <keithp@keithp.com> 38836Date: Tue Dec 9 09:30:57 2014 -0800 38837 38838 dbe: Call to DDX SwapBuffers requires address of int, not unsigned int [CVE-2014-8097 pt. 2] 38839 38840 When the local types used to walk the DBE request were changed, this 38841 changed the type of the parameter passed to the DDX SwapBuffers API, 38842 but there wasn't a matching change in the API definition. 38843 38844 At this point, with the API frozen, I just stuck a new variable in 38845 with the correct type. Because we've already bounds-checked nStuff to 38846 be smaller than UINT32_MAX / sizeof(DbeSwapInfoRec), we know it will 38847 fit in a signed int without overflow. 38848 38849 Signed-off-by: Keith Packard <keithp@keithp.com 38850 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38851 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38852 38853commit 8aa23f27c7e29c62d23867440b0bb00fc288b5ba 38854Author: Keith Packard <keithp@keithp.com> 38855Date: Mon Dec 8 12:59:31 2014 -0800 38856 38857 glamor: Free existing EGL image when assigning new one 38858 38859 When reallocating the framebuffer on screen resize, the old EGL image 38860 was getting leaked. Check for an existing EGL image and free it in 38861 this case. 38862 38863 Signed-off-by: Keith Packard <keithp@keithp.com> 38864 Revewied-by: Zhigang Gong <zhigang.gong@linux.intel.com> 38865 38866commit c22433d55ea16e4879b092ee28d284fc895c9956 38867Author: Keith Packard <keithp@keithp.com> 38868Date: Fri Dec 5 11:02:11 2014 -0800 38869 38870 glamor: Remove redundant reference to screen pixmap EGL image 38871 38872 There's no reason to store this in the egl screen private as the 38873 screen pixmap will always hold a reference to it anyways. 38874 38875 Signed-off-by: Keith Packard <keithp@keithp.com> 38876 Revewied-by: Zhigang Gong <zhigang.gong@linux.intel.com> 38877 38878commit 5064ffab631dcbc4265079fb32a02d3e2f1c4cd8 38879Author: Keith Packard <keithp@keithp.com> 38880Date: Fri Dec 5 10:58:28 2014 -0800 38881 38882 glamor: Always destroy EGL image associated with destroyed pixmap 38883 38884 There were three paths that called eglDestroyImageKHR: 38885 38886 * The front buffer 38887 * The intel driver's flip buffer 38888 * pixmaps under DRI3 38889 38890 This patch unifies the second two by having glamor_destroy_pixmap 38891 always destroy any associaged EGL image. This allows us to stop 38892 storing the back_pixmap pointer in glamor as that was only used to 38893 make sure that buffer was freed at server reset time. 38894 38895 v2: check for valid pixmap_priv before using it in 38896 glamor_egl_destroy_pixmap_image 38897 38898 Signed-off-by: Keith Packard <keithp@keithp.com> 38899 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 38900 38901commit 7e7630bbb775573eea2a2335adb9d190c3e1e971 38902Author: Robert Morell <rmorell@nvidia.com> 38903Date: Wed Nov 12 18:51:43 2014 -0800 38904 38905 glx: Fix mask truncation in __glXGetAnswerBuffer [CVE-2014-8093 6/6] 38906 38907 On a system where sizeof(unsigned) != sizeof(intptr_t), the unary 38908 bitwise not operation will result in a mask that clears all high bits 38909 from temp_buf in the expression: 38910 temp_buf = (temp_buf + mask) & ~mask; 38911 38912 Signed-off-by: Robert Morell <rmorell@nvidia.com> 38913 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38914 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38915 38916commit e883c170c15493ab3637c0a01890f5a7ca4e16a5 38917Author: Adam Jackson <ajax@redhat.com> 38918Date: Mon Nov 10 12:13:48 2014 -0500 38919 38920 glx: Pass remaining request length into ->varsize (v2) [CVE-2014-8098 8/8] 38921 38922 v2: Handle more multiplies in indirect_reqsize.c (Julien Cristau) 38923 38924 Reviewed-by: Julien Cristau <jcristau@debian.org> 38925 Reviewed-by: Michal Srb <msrb@suse.com> 38926 Reviewed-by: Andy Ritger <aritger@nvidia.com> 38927 Signed-off-by: Adam Jackson <ajax@redhat.com> 38928 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38929 38930commit 984583a497c813df5827ae22483133e704fee79c 38931Author: Adam Jackson <ajax@redhat.com> 38932Date: Mon Nov 10 12:13:47 2014 -0500 38933 38934 glx: Length checking for non-generated single requests (v2) [CVE-2014-8098 7/8] 38935 38936 v2: 38937 Fix single versus vendor-private length checking for ARB_imaging subset 38938 extensions. (Julien Cristau) 38939 38940 v3: 38941 Fix single versus vendor-private length checking for ARB_imaging subset 38942 extensions. (Julien Cristau) 38943 38944 Reviewed-by: Michal Srb <msrb@suse.com> 38945 Reviewed-by: Andy Ritger <aritger@nvidia.com> 38946 Signed-off-by: Adam Jackson <ajax@redhat.com> 38947 Signed-off-by: Julien Cristau <jcristau@debian.org> 38948 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38949 38950commit 44ba149f28ece93c2fbfc9cc980588de5322dd4b 38951Author: Adam Jackson <ajax@redhat.com> 38952Date: Mon Nov 10 12:13:46 2014 -0500 38953 38954 glx: Length-checking for non-generated vendor private requests [CVE-2014-8098 6/8] 38955 38956 Reviewed-by: Keith Packard <keithp@keithp.com> 38957 Reviewed-by: Michal Srb <msrb@suse.com> 38958 Reviewed-by: Andy Ritger <aritger@nvidia.com> 38959 Signed-off-by: Adam Jackson <ajax@redhat.com> 38960 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38961 38962commit afe177020d1fb776c6163f21eddc82cb185b95ca 38963Author: Adam Jackson <ajax@redhat.com> 38964Date: Mon Nov 10 12:13:45 2014 -0500 38965 38966 glx: Request length checks for SetClientInfoARB [CVE-2014-8098 5/8] 38967 38968 Reviewed-by: Keith Packard <keithp@keithp.com> 38969 Reviewed-by: Julien Cristau <jcristau@debian.org> 38970 Reviewed-by: Michal Srb <msrb@suse.com> 38971 Reviewed-by: Andy Ritger <aritger@nvidia.com> 38972 Signed-off-by: Adam Jackson <ajax@redhat.com> 38973 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38974 38975commit c91e4abc3b892f42802efa20fef7ada442c2d3f5 38976Author: Adam Jackson <ajax@redhat.com> 38977Date: Mon Nov 10 12:13:44 2014 -0500 38978 38979 glx: Top-level length checking for swapped VendorPrivate requests [CVE-2014-8098 4/8] 38980 38981 Reviewed-by: Keith Packard <keithp@keithp.com> 38982 Reviewed-by: Julien Cristau <jcristau@debian.org> 38983 Reviewed-by: Michal Srb <msrb@suse.com> 38984 Reviewed-by: Andy Ritger <aritger@nvidia.com> 38985 Signed-off-by: Adam Jackson <ajax@redhat.com> 38986 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 38987 38988commit a33a939e6abb255b14d8dbc85fcbd2c55b958bae 38989Author: Adam Jackson <ajax@redhat.com> 38990Date: Mon Nov 10 12:13:43 2014 -0500 38991 38992 glx: Length checking for RenderLarge requests (v2) [CVE-2014-8098 3/8] 38993 38994 This is a half-measure until we start passing request length into the 38995 varsize function, but it's better than the nothing we had before. 38996 38997 v2: Verify that there's at least a large render header's worth of 38998 dataBytes (Julien Cristau) 38999 39000 Reviewed-by: Michal Srb <msrb@suse.com> 39001 Reviewed-by: Andy Ritger <aritger@nvidia.com> 39002 Signed-off-by: Adam Jackson <ajax@redhat.com> 39003 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39004 39005commit 698888e6671d54c7ae41e9d456f7f5483a3459d2 39006Author: Adam Jackson <ajax@redhat.com> 39007Date: Mon Nov 10 12:13:42 2014 -0500 39008 39009 glx: Integer overflow protection for non-generated render requests (v3) [CVE-2014-8093 5/6] 39010 39011 v2: 39012 Fix constants in __glXMap2fReqSize (Michal Srb) 39013 Validate w/h/d for proxy targets too (Keith Packard) 39014 39015 v3: 39016 Fix Map[12]Size to correctly reject order == 0 (Julien Cristau) 39017 39018 Reviewed-by: Keith Packard <keithp@keithp.com> 39019 Reviewed-by: Michal Srb <msrb@suse.com> 39020 Reviewed-by: Andy Ritger <aritger@nvidia.com> 39021 Signed-off-by: Adam Jackson <ajax@redhat.com> 39022 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39023 39024commit be09e0c988ffdb0371293af49fb4ea8f49ed324a 39025Author: Julien Cristau <jcristau@debian.org> 39026Date: Mon Nov 10 12:13:41 2014 -0500 39027 39028 glx: Length checking for GLXRender requests (v2) [CVE-2014-8098 2/8] 39029 39030 v2: 39031 Remove can't-happen comparison for cmdlen < 0 (Michal Srb) 39032 39033 Reviewed-by: Adam Jackson <ajax@redhat.com> 39034 Reviewed-by: Michal Srb <msrb@suse.com> 39035 Reviewed-by: Andy Ritger <aritger@nvidia.com> 39036 Signed-off-by: Julien Cristau <jcristau@debian.org> 39037 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39038 39039commit 2a5cbc17fc72185bf0fa06fef26d1f782de72595 39040Author: Adam Jackson <ajax@redhat.com> 39041Date: Mon Nov 10 12:13:40 2014 -0500 39042 39043 glx: Add safe_{add,mul,pad} (v3) [CVE-2014-8093 4/6] 39044 39045 These are paranoid about integer overflow, and will return -1 if their 39046 operation would overflow a (signed) integer or if either argument is 39047 negative. 39048 39049 Note that RenderLarge requests are sized with a uint32_t so in principle 39050 this could be sketchy there, but dix limits bigreqs to 128M so you 39051 shouldn't ever notice, and honestly if you're sending more than 2G of 39052 rendering commands you're already doing something very wrong. 39053 39054 v2: Use INT_MAX for consistency with the rest of the server (jcristau) 39055 v3: Reject negative arguments (anholt) 39056 39057 Reviewed-by: Keith Packard <keithp@keithp.com> 39058 Reviewed-by: Julien Cristau <jcristau@debian.org> 39059 Reviewed-by: Michal Srb <msrb@suse.com> 39060 Reviewed-by: Andy Ritger <aritger@nvidia.com> 39061 Signed-off-by: Adam Jackson <ajax@redhat.com> 39062 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39063 39064commit 13d36923e0ddb077f4854e354c3d5c80590b5d9d 39065Author: Adam Jackson <ajax@redhat.com> 39066Date: Mon Nov 10 12:13:39 2014 -0500 39067 39068 glx: Fix image size computation for EXT_texture_integer [CVE-2014-8098 1/8] 39069 39070 Without this we'd reject the request with BadLength. Note that some old 39071 versions of Mesa had a bug in the same place, and would _send_ zero 39072 bytes of image data; these will now be rejected, correctly. 39073 39074 Reviewed-by: Keith Packard <keithp@keithp.com> 39075 Reviewed-by: Julien Cristau <jcristau@debian.org> 39076 Reviewed-by: Michal Srb <msrb@suse.com> 39077 Reviewed-by: Andy Ritger <aritger@nvidia.com> 39078 Signed-off-by: Adam Jackson <ajax@redhat.com> 39079 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39080 39081commit 717a1b37767b41e14859e5022ae9e679152821a9 39082Author: Adam Jackson <ajax@redhat.com> 39083Date: Mon Nov 10 12:13:38 2014 -0500 39084 39085 glx: Additional paranoia in __glXGetAnswerBuffer / __GLX_GET_ANSWER_BUFFER (v2) [CVE-2014-8093 3/6] 39086 39087 If the computed reply size is negative, something went wrong, treat it 39088 as an error. 39089 39090 v2: Be more careful about size_t being unsigned (Matthieu Herrb) 39091 v3: SIZE_MAX not SIZE_T_MAX (Alan Coopersmith) 39092 39093 Reviewed-by: Julien Cristau <jcristau@debian.org> 39094 Reviewed-by: Michal Srb <msrb@suse.com> 39095 Reviewed-by: Andy Ritger <aritger@nvidia.com> 39096 Signed-off-by: Adam Jackson <ajax@redhat.com> 39097 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39098 39099commit ab2ba9338aa5e85b4487bc7fbe69985c76483e01 39100Author: Adam Jackson <ajax@redhat.com> 39101Date: Mon Nov 10 12:13:37 2014 -0500 39102 39103 glx: Be more strict about rejecting invalid image sizes [CVE-2014-8093 2/6] 39104 39105 Before this we'd just clamp the image size to 0, which was just 39106 hideously stupid; if the parameters were such that they'd overflow an 39107 integer, you'd allocate a small buffer, then pass huge values into (say) 39108 ReadPixels, and now you're scribbling over arbitrary server memory. 39109 39110 Reviewed-by: Keith Packard <keithp@keithp.com> 39111 Reviewed-by: Julien Cristau <jcristau@debian.org> 39112 Reviewed-by: Michal Srb <msrb@suse.com> 39113 Reviewed-by: Andy Ritger <aritger@nvidia.com> 39114 Signed-off-by: Adam Jackson <ajax@redhat.com> 39115 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39116 39117commit 23fe7718bb171e71db2d1a30505c2ca2988799d9 39118Author: Adam Jackson <ajax@redhat.com> 39119Date: Mon Nov 10 12:13:36 2014 -0500 39120 39121 glx: Be more paranoid about variable-length requests [CVE-2014-8093 1/6] 39122 39123 If the size computation routine returns -1 we should just reject the 39124 request outright. Clamping it to zero could give an attacker the 39125 opportunity to also mangle cmdlen in such a way that the subsequent 39126 length check passes, and the request would get executed, thus passing 39127 data we wanted to reject to the renderer. 39128 39129 Reviewed-by: Keith Packard <keithp@keithp.com> 39130 Reviewed-by: Julien Cristau <jcristau@debian.org> 39131 Reviewed-by: Michal Srb <msrb@suse.com> 39132 Reviewed-by: Andy Ritger <aritger@nvidia.com> 39133 Signed-off-by: Adam Jackson <ajax@redhat.com> 39134 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39135 39136commit f4afd53f2aeaddf509bf9f71d1716dd273fd6e14 39137Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39138Date: Sun Feb 9 22:42:47 2014 -0800 39139 39140 Add REQUEST_FIXED_SIZE testcases to test/misc.c 39141 39142 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39143 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39144 39145commit 2df83bb122debc3c20cfc3d3b0edc85cd0270f79 39146Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39147Date: Sun Feb 9 21:28:05 2014 -0800 39148 39149 Add request length checking test cases for some Xinput 2.x requests 39150 39151 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39152 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39153 39154commit d153a85f7478a7a67ccb02fbca6390b0ab1732ee 39155Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39156Date: Sun Feb 9 21:27:27 2014 -0800 39157 39158 Add request length checking test cases for some Xinput 1.x requests 39159 39160 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39161 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39162 39163commit a0ece23a8bd300c8be10812d368dc8058c97c63e 39164Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39165Date: Sun Jan 26 20:02:20 2014 -0800 39166 39167 xfixes: unvalidated length in SProcXFixesSelectSelectionInput [CVE-2014-8102] 39168 39169 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39170 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39171 39172commit 5d3a788aeb2fbd3ca2812747dc18c94a8b981c63 39173Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39174Date: Sun Jan 26 19:51:29 2014 -0800 39175 39176 render: unvalidated lengths in Render extn. swapped procs [CVE-2014-8100 2/2] 39177 39178 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39179 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39180 39181commit b5f9ef03df6a650571b29d3d1c1d2b67c6e84336 39182Author: Julien Cristau <jcristau@debian.org> 39183Date: Tue Oct 28 10:30:04 2014 +0100 39184 39185 render: check request size before reading it [CVE-2014-8100 1/2] 39186 39187 Otherwise we may be reading outside of the client request. 39188 39189 Signed-off-by: Julien Cristau <jcristau@debian.org> 39190 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39191 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39192 39193commit 3df2fcf12499ebdb26b9b67419ea485a42041f33 39194Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39195Date: Sun Jan 26 19:38:09 2014 -0800 39196 39197 randr: unvalidated lengths in RandR extension swapped procs [CVE-2014-8101] 39198 39199 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39200 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39201 39202commit d155b7a8e38e74aee96bf52c20c8b6a330d7d462 39203Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39204Date: Sun Jan 26 19:33:34 2014 -0800 39205 39206 present: unvalidated lengths in Present extension procs [CVE-2014-8103 2/2] 39207 39208 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39209 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39210 Reviewed-by: Julien Cristau <jcristau@debian.org> 39211 39212commit 0a6085aaf3581cca558d960ea176ddf3a41a2213 39213Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39214Date: Sun Jan 26 19:28:05 2014 -0800 39215 39216 dri3: unvalidated lengths in DRI3 extension swapped procs [CVE-2014-8103 1/2] 39217 39218 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39219 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39220 39221commit 32a95fb7c7dbe22c9441c62762dfa4a8ec54d6c3 39222Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39223Date: Sun Jan 26 19:23:17 2014 -0800 39224 39225 Xv: unvalidated lengths in XVideo extension swapped procs [CVE-2014-8099] 39226 39227 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39228 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39229 39230commit 7553082b9b883b5f130044f3d53bce2f0b660e52 39231Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39232Date: Sun Jan 26 17:18:54 2014 -0800 39233 39234 xcmisc: unvalidated length in SProcXCMiscGetXIDList() [CVE-2014-8096] 39235 39236 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39237 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39238 39239commit 73c63afb93c0af1bfd1969bf6e71c9edca586c77 39240Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39241Date: Sun Jan 26 10:54:41 2014 -0800 39242 39243 Xi: unvalidated lengths in Xinput extension [CVE-2014-8095] 39244 39245 Multiple functions in the Xinput extension handling of requests from 39246 clients failed to check that the length of the request sent by the 39247 client was large enough to perform all the required operations and 39248 thus could read or write to memory outside the bounds of the request 39249 buffer. 39250 39251 This commit includes the creation of a new REQUEST_AT_LEAST_EXTRA_SIZE 39252 macro in include/dix.h for the common case of needing to ensure a 39253 request is large enough to include both the request itself and a 39254 minimum amount of extra data following the request header. 39255 39256 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39257 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39258 39259commit 2ef42519c41e793579c9cea699c866fee3d9321f 39260Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39261Date: Wed Jan 22 23:12:04 2014 -0800 39262 39263 dbe: unvalidated lengths in DbeSwapBuffers calls [CVE-2014-8097] 39264 39265 ProcDbeSwapBuffers() has a 32bit (n) length value that it uses to read 39266 from a buffer. The length is never validated, which can lead to out of 39267 bound reads, and possibly returning the data read from out of bounds to 39268 the misbehaving client via an X Error packet. 39269 39270 SProcDbeSwapBuffers() swaps data (for correct endianness) before 39271 handing it off to the real proc. While doing the swapping, the 39272 length field is not validated, which can cause memory corruption. 39273 39274 v2: reorder checks to avoid compilers optimizing out checks for overflow 39275 that happen after we'd already have done the overflowing multiplications. 39276 39277 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 39278 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39279 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39280 39281commit 6692670fde081bbfe9313f17d84037ae9116702a 39282Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39283Date: Wed Jan 22 23:40:18 2014 -0800 39284 39285 dri2: integer overflow in ProcDRI2GetBuffers() [CVE-2014-8094] 39286 39287 ProcDRI2GetBuffers() tries to validate a length field (count). 39288 There is an integer overflow in the validation. This can cause 39289 out of bound reads and memory corruption later on. 39290 39291 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 39292 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39293 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39294 Reviewed-by: Julien Cristau <jcristau@debian.org> 39295 39296commit e0e11644622a589129a01e11e5d105dc74a098de 39297Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39298Date: Wed Jan 22 23:44:46 2014 -0800 39299 39300 dix: integer overflow in REQUEST_FIXED_SIZE() [CVE-2014-8092 4/4] 39301 39302 Force use of 64-bit integers when evaluating data provided by clients 39303 in 32-bit fields which can overflow when added or multiplied during 39304 checks. 39305 39306 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 39307 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39308 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39309 39310commit 97015a07b9e15d8ec5608b95d95ec0eb51202acb 39311Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39312Date: Wed Jan 22 22:37:15 2014 -0800 39313 39314 dix: integer overflow in RegionSizeof() [CVE-2014-8092 3/4] 39315 39316 RegionSizeof contains several integer overflows if a large length 39317 value is passed in. Once we fix it to return 0 on overflow, we 39318 also have to fix the callers to handle this error condition 39319 39320 v2: Fixed limit calculation in RegionSizeof as pointed out by jcristau. 39321 39322 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 39323 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39324 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39325 Reviewed-by: Julien Cristau <jcristau@debian.org> 39326 39327commit bc8e20430b6f6378daf6ce4329029248a88af08b 39328Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39329Date: Mon Jan 6 23:30:14 2014 -0800 39330 39331 dix: integer overflow in GetHosts() [CVE-2014-8092 2/4] 39332 39333 GetHosts() iterates over all the hosts it has in memory, and copies 39334 them to a buffer. The buffer length is calculated by iterating over 39335 all the hosts and adding up all of their combined length. There is a 39336 potential integer overflow, if there are lots and lots of hosts (with 39337 a combined length of > ~4 gig). This should be possible by repeatedly 39338 calling ProcChangeHosts() on 64bit machines with enough memory. 39339 39340 This patch caps the list at 1mb, because multi-megabyte hostname 39341 lists for X access control are insane. 39342 39343 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 39344 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39345 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39346 39347commit eeae42d60bf3d5663ea088581f6c28a82cd17829 39348Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39349Date: Wed Jan 22 21:11:16 2014 -0800 39350 39351 dix: integer overflow in ProcPutImage() [CVE-2014-8092 1/4] 39352 39353 ProcPutImage() calculates a length field from a width, left pad and depth 39354 specified by the client (if the specified format is XYPixmap). 39355 39356 The calculations for the total amount of memory the server needs for the 39357 pixmap can overflow a 32-bit number, causing out-of-bounds memory writes 39358 on 32-bit systems (since the length is stored in a long int variable). 39359 39360 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 39361 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39362 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39363 39364commit 90cc925c5991fcb203f72d00b04419cd754a9b2c 39365Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39366Date: Fri Jan 17 18:54:03 2014 -0800 39367 39368 unchecked malloc may allow unauthed client to crash Xserver [CVE-2014-8091] 39369 39370 authdes_ezdecode() calls malloc() using a length provided by the 39371 connection handshake sent by a newly connected client in order 39372 to authenticate to the server, so should be treated as untrusted. 39373 39374 It didn't check if malloc() failed before writing to the newly 39375 allocated buffer, so could lead to a server crash if the server 39376 fails to allocate memory (up to UINT16_MAX bytes, since the len 39377 field is a CARD16 in the X protocol). 39378 39379 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 39380 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39381 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39382 39383commit 3e7218a6c23354d66f508b18164cac98a346b3ee 39384Merge: 6f4c398a0 bc71081f0 39385Author: Keith Packard <keithp@keithp.com> 39386Date: Mon Dec 8 18:07:55 2014 -0800 39387 39388 Merge remote-tracking branch 'jturney/indirect-glx-fixes' 39389 39390commit 6f4c398a0e632b0c92e3cb8ee03ca7f5b5cc018e 39391Merge: aae646069 5920433c3 39392Author: Keith Packard <keithp@keithp.com> 39393Date: Mon Dec 8 18:04:08 2014 -0800 39394 39395 Merge remote-tracking branch 'jturney/master' 39396 39397commit aae6460694ac3667abb8c34fdf3a7dae524827a4 39398Author: Mario Kleiner <mario.kleiner.de@gmail.com> 39399Date: Sat Dec 6 05:40:08 2014 +0100 39400 39401 present: Fix use of vsynced pageflips and honor PresentOptionAsync. (v4) 39402 39403 Pageflips for Pixmap presents were not synchronized to vblank on 39404 drivers with support for PresentCapabilityAsync, due to some 39405 missing init for vblank->sync_flips. The PresentOptionAsync 39406 flag was completely ignored for pageflipped presents. 39407 39408 Vsynced flips only worked by accident on the intel-ddx, as that 39409 driver doesn't have PresentCapabilityAsync support. 39410 39411 On nouveau-ddx, which supports PresentCapabilityAsync, this 39412 always caused non-vsynced pageflips with pretty ugly tearing. 39413 39414 This patch fixes the problem, as tested on top of XOrg 1.16.2 39415 on nouveau and intel. 39416 39417 v4: Add additional PresentCapabilityAsync caps check, as 39418 suggested by Eric Anholt. 39419 39420 Please also apply to XOrg 1.17 and XOrg 1.16.2 stable. 39421 39422 Applying on top of XOrg 1.16.2 requires cherry-picking 39423 commit 2051514652481a83bd7cf22e57cb0fcd40333f33 39424 which trivially fixes lack of support for protocol option 39425 PresentOptionCopy - get two bug fixes for the price of one! 39426 39427 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 39428 Reviewed-by: Eric Anholt <eric@anholt.net> 39429 Signed-off-by: Keith Packard <keithp@keithp.com> 39430 39431commit 32d3100bd78efe41d468f7d66861296aee468b6f 39432Author: Mario Kleiner <mario.kleiner.de@gmail.com> 39433Date: Sat Dec 6 05:40:07 2014 +0100 39434 39435 present: Avoid crashes in DebugPresent(), a bit more info. 39436 39437 DebugPresent() crashed the server when a dri3 drawable 39438 was closed while a pageflipped present was still pending, 39439 due to vblank->window-> Null-Ptr deref, so debug builds 39440 caused new problems to debug. 39441 39442 E.g., 39443 39444 glXSwapBuffers(...); 39445 glXDestroyWindow(...); 39446 -> Pageflip for non-existent window completes -> boom. 39447 39448 Also often happens when switching desktop compositor on/off 39449 due to Present unflips, or when logging out of session. 39450 39451 Also add info if a Present is queued for copyswap or pageflip, 39452 if the present is vsynced, and the serial no of the Present 39453 request, to aid debugging of pageflip and vsync issues. The 39454 serial number is useful as Mesa's dri3/present backend encodes 39455 its sendSBC in the serial number, so one can easily correlate 39456 server debug output with Mesa and with the SBC values returned 39457 to actual OpenGL client applications via OML_sync_control and 39458 INTEL_swap_events extension, makes debugging quite a bit more 39459 easy. 39460 39461 Please also cherry-pick this for a 1.16.x stable update. 39462 39463 Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com> 39464 Reviewed-by: Eric Anholt <eric@anholt.net> 39465 Signed-off-by: Keith Packard <keithp@keithp.com> 39466 39467commit 32ca85c9e075e40b8a4712a79a51702facba0332 39468Author: Daniel Martin <consume.noise@gmail.com> 39469Date: Fri Nov 28 11:20:50 2014 +0100 39470 39471 modesetting: Remove unused params from drmmode_output_init() 39472 39473 drmmode_output_init() doesn't touch (the int*) num_dvi and num_hdmi. 39474 Remove both parameters. 39475 39476 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 39477 Reviewed-by: Eric Anholt <eric@anholt.net> 39478 Signed-off-by: Keith Packard <keithp@keithp.com> 39479 39480commit c0ea476b9bb91f03061b891ffb897c438fa03879 39481Author: Daniel Martin <consume.noise@gmail.com> 39482Date: Fri Nov 28 11:20:49 2014 +0100 39483 39484 modesetting: Fix ifdefs s/HAVE_UDEV/CONFIG_UDEV_KMS/ 39485 39486 We don't define HAVE_UDEV, that's a remnant from xf86-video-modesetting. 39487 But, we have CONFIG_UDEV_KMS. 39488 39489 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 39490 Reviewed-by: Eric Anholt <eric@anholt.net> 39491 Signed-off-by: Keith Packard <keithp@keithp.com> 39492 39493commit 8a16620dcd7a22fd9b38876ffc711629c4fbda0a 39494Author: Daniel Martin <consume.noise@gmail.com> 39495Date: Fri Nov 28 11:20:48 2014 +0100 39496 39497 modesetting: Create new EGL screen in drmmode_xf86crtc_resize 39498 39499 If we don't glamor_egl_create_textured_screen_ext() in 39500 drmmode_xf86crtc_resize() we end up with a black screen and no client 39501 window(s) visible. 39502 39503 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 39504 Reviewed-by: Eric Anholt <eric@anholt.net> 39505 Signed-off-by: Keith Packard <keithp@keithp.com> 39506 39507commit 0d63fa5850fe9d8b045b8f1a83e72b75f9065a7a 39508Author: Daniel Martin <consume.noise@gmail.com> 39509Date: Fri Nov 28 11:20:47 2014 +0100 39510 39511 modesetting: Move Bool glamor into drmmode struct 39512 39513 Move the boolean glamor from struct modesetting into struct drmmode for 39514 later re-use in drmmode_display. 39515 39516 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 39517 Reviewed-by: Eric Anholt <eric@anholt.net> 39518 Reviewed-by: Keith Packard <keithp@keithp.com> 39519 Signed-off-by: Keith Packard <keithp@keithp.com> 39520 39521commit 62a4eeaa25099872682d6c2f9f13a0e73fc5ce1e 39522Author: Daniel Martin <consume.noise@gmail.com> 39523Date: Fri Nov 28 11:20:46 2014 +0100 39524 39525 config/udev: Prefix and shift "removing GPU" message 39526 39527 The message "removing GPU device ..." appeared even if the removal was 39528 skipped (when path == NULL). Move it below the path check and make it a 39529 LogMessage with config/udev prefix. 39530 39531 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 39532 Reviewed-by: Eric Anholt <eric@anholt.net> 39533 Reviewed-by: Keith Packard <keithp@keithp.com> 39534 Signed-off-by: Keith Packard <keithp@keithp.com> 39535 39536commit bc71081f0e3d8ce3aecf2cb168431dbc9fe6a87b 39537Author: Jon TURNEY <jon.turney@dronecode.org.uk> 39538Date: Fri Apr 18 12:17:06 2014 +0100 39539 39540 glx: Fix crash when a client exits without deleting GL contexts 39541 39542 With the previous patches applied, we now have crash due to use-after-free when 39543 a client exits without deleting all it's GL contexts 39544 39545 On client exit, CloseDownClient first calls glxClientCallback() with 39546 ClientStateGone, which calls __glXFreeContext() directly. 39547 39548 Subsequently CloseDownClient() frees all the clients resources, which leads to 39549 ContextGone() being called for a context resource where the context has already 39550 been freed. 39551 39552 Fix this by modifiying glxClientCallback() to free the context resource. 39553 39554 Also make __glXFreeContext() static, as calling it directly leads to this 39555 problem, instead the context resource should be released. 39556 39557 With the previous patches applied, this can be demonstrated with e.g. glxinfo, 39558 which doesn't delete it's context before exit. 39559 39560 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 39561 Reviewed-by: Adam Jackson <ajax@redhat.com> 39562 39563commit 5c606c0a89e74fa223a99864be11cc3be60a159b 39564Author: Jon TURNEY <jon.turney@dronecode.org.uk> 39565Date: Fri Apr 18 12:17:05 2014 +0100 39566 39567 glx: Flush context which is being made non-current due to drawable going away 39568 39569 Some sequences of glean tests fail with GLXBadCurrentWindow when using indirect 39570 rendering, e.g. glean -t 'fpexceptions getString'. 39571 39572 Flush a context which is being made non-current due to the drawable on which is 39573 it is current going away. Waiting until another context is made current is too 39574 late, as the drawable no longer exists. 39575 39576 v2: Rewrite for direct GL dispatch 39577 39578 v3: Inline FlushContext(), doesn't need to be a separate function 39579 39580 e.g. LIBGL_ALWAYS_INDIRECT=1 ./glean -r results -o --quick -t "fpexceptions 39581 getString" fails with a BadContextTag error. 39582 39583 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 39584 Reviewed-by: Adam Jackson <ajax@redhat.com> 39585 39586commit 437b27494f127854d75e59b4e2aac264e9f913e9 39587Author: Jon TURNEY <jon.turney@dronecode.org.uk> 39588Date: Fri Apr 18 12:17:04 2014 +0100 39589 39590 Revert "glx: Simplify glXDestroyContext" 39591 39592 This reverts commit 7f5adf73a0f9a951a6df201532b4031d38054369. 39593 39594 This seems to miss the whole point of the idExists flag, as it makes the 39595 lifetime of that being true the same as the lifetime of the Context resource. 39596 39597 The previously current context tag is always given in a MakeContextCurrent 39598 request, even if that context tag is no longer valid (for example, the context 39599 has been deleted), so this leads to BadContextTag errors. 39600 39601 See fd.o bug #30089 for the makecurrenttest.c testcase, and some discussion of 39602 previous manifestations of this bug. 39603 39604 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 39605 Reviewed-by: Adam Jackson <ajax@redhat.com> 39606 39607commit 4b0d0df34f10a88c10cb23dd50087b59f5c4fece 39608Author: Peter Harris <pharris@opentext.com> 39609Date: Mon Nov 17 14:31:24 2014 -0500 39610 39611 Fix overflow of ConnectionOutput->size and ->count 39612 39613 When (long) is larger than (int), and when realloc succeeds with sizes 39614 larger than INT_MAX, ConnectionOutput->size and ConnectionOutput->count 39615 overflow and become negative. 39616 39617 When ConnectionOutput->count is negative, InsertIOV does not actually 39618 insert an IOV, and FlushClient goes into an infinite loop of writev(fd, 39619 iov, 0) [an empty list]. 39620 39621 Avoid this situation by killing the client when it has more than INT_MAX 39622 unread bytes of data. 39623 39624 Signed-off-by: Peter Harris <pharris@opentext.com> 39625 Reviewed-by: Keith Packard <keithp@keithp.com> 39626 Signed-off-by: Keith Packard <keithp@keithp.com> 39627 39628commit 802932d112a3f6a09420be9e4a13fa78ac43840b 39629Author: Adam Jackson <ajax@redhat.com> 39630Date: Mon Nov 17 15:28:58 2014 -0500 39631 39632 render: fix ChangePicture when Xinerama is active (v2) (#49170) 39633 39634 ChangePicture takes wire XIDs, but didn't do any Xinerama translation, 39635 which meant setting a clip pixmap or a separate alpha picture would 39636 result in those elements pointing at the instance of the pixmap on 39637 screen 0. Which is, you know, bad. 39638 39639 v2: This one actually builds. 39640 39641 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=49170 39642 Signed-off-by: Adam Jackson <ajax@redhat.com> 39643 Reviewed-by: Keith Packard <keithp@keithp.com> 39644 Signed-off-by: Keith Packard <keithp@keithp.com> 39645 39646commit c2994001680a7dcf9c167886d49b7326c97cd5d1 39647Author: Aaron Plattner <aplattner@nvidia.com> 39648Date: Fri Nov 21 08:39:02 2014 -0800 39649 39650 os: "Server terminated successfully" is not an error 39651 39652 ErrorFSigSafe calls LogVMessageVerbSigSafe with the message type set to X_ERROR. 39653 That generates this in the log: 39654 39655 (EE) Server terminated successfully (0). Closing log file. 39656 39657 People periodically report this as an error, sometimes quoting this "error" 39658 rather than an earlier error that actually caused a problem. 39659 39660 v2: Use X_INFO instead of X_NOTICE 39661 39662 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 39663 Reviewed-by: Keith Packard <keithp@keithp.com> 39664 Reviewed-by: Eric Anholt <eric@anholt.net> 39665 Signed-off-by: Keith Packard <keithp@keithp.com> 39666 39667commit b09d59342804db7dbb8056dca43dd39f54e290aa 39668Author: Alan Coopersmith <alan.coopersmith@oracle.com> 39669Date: Wed Nov 26 13:01:29 2014 -0800 39670 39671 Add -iglx & +iglx to Xserver.man 39672 39673 Covers the current state after commits 99f0365b1fbdfd9238b9f, 39674 d0da0e9c3bb8fe0cd4879, & e3aa13b8d63ea2fba6eb4 were all applied. 39675 39676 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39677 Reviewed-by: James Jones <jajones@nvidia.com> 39678 Reviewed-by: Robert Morell <rmorell@nvidia.com> 39679 Signed-off-by: Keith Packard <keithp@keithp.com> 39680 39681commit c52a2b1ebad56820af932dfbc871701a8b04fd9c 39682Author: Adam Jackson <ajax@redhat.com> 39683Date: Tue Nov 11 12:30:07 2014 -0500 39684 39685 mi: Fix regression in arc drawing 39686 39687 xts' XDrawArcs/15 regressed (turning into a server-side infinite loop) 39688 after: 39689 39690 commit 7679afd4da8b86aed27e5916ba723116a3c8bb4a 39691 Author: Adam Jackson <ajax@redhat.com> 39692 Date: Fri Sep 26 12:01:37 2014 -0400 39693 39694 mi: Fold mifpolycon.c into miarc.c 39695 39696 The reason is miarc.c provided its own definitions (sigh) of min/max, 39697 that both accept int arguments and return an int. Since miFillSppPoly 39698 uses a double (sigh) and some min-involving math for its loop index 39699 variable, things do not go well. 39700 39701 Since the integer versions of min/max are redundant, nuke 'em. 39702 39703 Signed-off-by: Adam Jackson <ajax@redhat.com> 39704 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 39705 Tested-by: Keith Packard <keithp@keithp.com> 39706 Signed-off-by: Keith Packard <keithp@keithp.com> 39707 39708commit ec0ac8970b508adcbc3d104b14a127118e4979d0 39709Merge: 65cc0982a 0f88d4e79 39710Author: Keith Packard <keithp@keithp.com> 39711Date: Mon Nov 24 15:39:51 2014 -0800 39712 39713 Merge remote-tracking branch 'whot/for-keith' 39714 39715commit 65cc0982afb1d47d17e655efaef40d4a82e7574a 39716Author: Keith Packard <keithp@keithp.com> 39717Date: Mon Oct 13 10:31:52 2014 -0700 39718 39719 glamor: Don't insert fbos from external objects into fbo cache 39720 39721 Mark fbos created from external buffers so that when the associated 39722 pixmap is destroyed, they aren't put into the fbo cache for later 39723 re-use and are instead freed immediately. 39724 39725 Signed-off-by: Keith Packard <keithp@keithp.com> 39726 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 39727 39728commit 0f88d4e7937116dbc438f98ddf1f3d94cfc5bead 39729Author: Peter Hutterer <peter.hutterer@who-t.net> 39730Date: Wed Nov 12 10:29:36 2014 +1000 39731 39732 mi: fix documentation for miPointerSetPosition 39733 39734 Changed when we added barriers, documentation didn't get updated. 39735 39736 Reported-by: Dave Airlie <airlied@redhat.com> 39737 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39738 Reviewed-by: Dave Airlie <airlied@redhat.com> 39739 39740commit 6a117af7e7ef6eac796ff0680a915c866fce4019 39741Author: John Hunter <zhaojunwang@pku.edu.cn> 39742Date: Mon Nov 3 15:45:38 2014 +0800 39743 39744 fix an annotation mistake 39745 39746 Signed-off-by: John Hunter <zhaojunwang@pku.edu.cn> 39747 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 39748 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39749 39750commit 732fd7e571b81d63aa44d6e4cf55883479e88f2f 39751Author: Peter Hutterer <peter.hutterer@who-t.net> 39752Date: Fri Oct 31 09:45:12 2014 +1000 39753 39754 Drop trailing whitespaces 39755 39756 sed -i "s/[ ]\+$//g" **/*.(c|h) 39757 39758 happy reviewing... 39759 git diff -w is an empty diff. 39760 39761 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39762 39763commit 96143d6dc35c98cb6b4e409dd89687e4ced3c1f0 39764Author: Peter Hutterer <peter.hutterer@who-t.net> 39765Date: Fri Oct 31 09:32:35 2014 +1000 39766 39767 dix: silence compiler warning comparing CARD32 to -1 39768 39769 window.c:3246:36: warning: comparison of constant -1 with expression of type 39770 'CARD32' (aka 'unsigned int') is always true 39771 [-Wtautological-constant-out-of-range-compare] 39772 if (optional->backingBitPlanes != ~0L) 39773 ~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~~~ 39774 39775 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39776 Reviewed-by: Michel Dänzer <michel@daenzer.net> 39777 39778commit 04ca16997994c7561ed8948b89fc1c922233da13 39779Author: Peter Hutterer <peter.hutterer@who-t.net> 39780Date: Fri Oct 31 09:29:33 2014 +1000 39781 39782 dix: silence compiler warning 39783 39784 inpututils.c:243:26: warning: comparison of constant 256 with expression of 39785 type 'KeyCode' (aka 'unsigned char') is always false 39786 [-Wtautological-constant-out-of-range-compare] 39787 if (modkeymap[i] >= MAP_LENGTH) 39788 ~~~~~~~~~~~~ ^ ~~~~~~~~~~ 39789 39790 MAP_LENGTH depends on MAX_BUTTONS which is somewhat arbitrarily chosen. We 39791 don't expect this to ever change, but just in case leave the condition there 39792 so the code is correct if we drop the number down. 39793 39794 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39795 39796commit 79b9d33545034f2952ebc7e444c201693b45f44b 39797Author: Peter Hutterer <peter.hutterer@who-t.net> 39798Date: Fri Oct 31 09:26:28 2014 +1000 39799 39800 xwayland: declare fatal log handler as noreturn 39801 39802 xwayland.c:661:1: warning: function 'xwl_log_handler' could be declared with 39803 attribute 'noreturn' [-Wmissing-noreturn] 39804 39805 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39806 39807commit adee4f239b5074d318dfc48beb0ec4e91973b449 39808Author: Peter Hutterer <peter.hutterer@who-t.net> 39809Date: Fri Oct 31 09:24:50 2014 +1000 39810 39811 xfree86: drop double-typedef of DBusConnection 39812 39813 Just include dbus.h, it's been typedef'd there since 2003. 39814 39815 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39816 39817commit 25315ee555dfeb96d9ed83064b1022ed852b612e 39818Author: Peter Hutterer <peter.hutterer@who-t.net> 39819Date: Fri Oct 31 09:16:28 2014 +1000 39820 39821 Xext: fix clang compiler warning 39822 39823 shm.c:1150:24: warning: equality comparison with extraneous parentheses 39824 [-Wparentheses-equality] 39825 if ((shmdesc->addr == ((char *) -1))) { 39826 39827 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39828 39829commit 995ecfb51d4ab8197e4591d5c0957e08a0bd6a59 39830Author: Peter Hutterer <peter.hutterer@who-t.net> 39831Date: Thu Oct 30 09:00:21 2014 +1000 39832 39833 include: change RegionSize() to take a size_t 39834 39835 /usr/include/xorg/regionstr.h:130:36: warning: implicit conversion changes 39836 signedness: 'int' to 'unsigned long' [-Wsign-conversion] 39837 return (sizeof(RegDataRec) + ((n) * sizeof(BoxRec))); 39838 ^ ~ 39839 39840 Really only just pushes the problem to the caller, but maybe that motivates 39841 someone to fix it. 39842 39843 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39844 39845commit 3ccec1b5ba6a26a30f82ba3aa9c5218dfa336869 39846Author: Peter Hutterer <peter.hutterer@who-t.net> 39847Date: Thu Oct 30 08:56:55 2014 +1000 39848 39849 include: fix documentation for list.h 39850 39851 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 39852 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 39853 39854commit 5920433c3a30f5f1c0ba1ab39a0c2ff388df6b23 39855Author: Colin Harrison <colin.harrison@virgin.net> 39856Date: Sat Sep 27 11:50:11 2014 +0100 39857 39858 hw/xwin: Don't allocate one wchar_t too much for unicode text placed on the Windows clipboard 39859 39860 The count of wchar_t returned by MultiByteToWideChar() includes the terminating 39861 null character, so don't add one to it. 39862 39863 Also, reduce the scope of various length variables 39864 39865 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 39866 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 39867 39868commit d172cd630dae9c991b84b4c367a4caf8199266ac 39869Author: Jon TURNEY <jon.turney@dronecode.org.uk> 39870Date: Mon Apr 28 12:48:15 2014 +0100 39871 39872 hw/xwin: Fix hang on shutdown when we own the clipboard. 39873 39874 If we are the clipboard owner when we are shutdown, we recieve a 39875 WM_RENDERALLFORMATS, to render the clipboard, so it's contents will remain 39876 available to other applications. Unfortunately, this is far too late to do 39877 anything useful with, as the server is waiting for the clipboard thread to exit, 39878 and so can't process requests to convert clipboard contents. 39879 39880 Change so we just do nothing on WM_RENDERALLFORMATS. (I'm not convinced that 39881 WM_RENDERALLFORMATS has ever worked usefully, in any case). 39882 39883 (To make this work, I guess we would need to rearrange the way shutdown works 39884 completely: first synchronously stop the clipboard, then stop the X server) 39885 39886 We also then receive a WM_DRAWCLIPBOARD, perhaps telling us that the available 39887 clipboard formats have changed (as ones which haven't been rendered are now 39888 removed), but the clipboard owner is now the system, not us, which we have to 39889 arrange to ignore. 39890 39891 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 39892 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 39893 39894commit 94d433c8cb64c9167050d02473176f888decf1d8 39895Author: Jon TURNEY <jon.turney@dronecode.org.uk> 39896Date: Fri Feb 21 23:20:00 2014 +0000 39897 39898 hw/xwin: Fix clipboard thread restart 39899 39900 It seems that the clipboard thread restart mechanism has been broken for a 39901 while, which can be demonstrated using XDMCP with KDM (e.g. to a Kubutunu 12.04 39902 host) 39903 39904 KDM kills all attached clients, including the clipboard integration client, 39905 which restarts, but then exits on WM_QUIT. 39906 39907 Using PostQuitMessage() in WM_DESTROY is unhelpful, as we may not actually be 39908 quitting the thread, if we just destroyed the window because the clipboard 39909 thread is about to retry, because he WM_QUIT message sticks around, and is 39910 noticed the next time we look at the window message queue and confuses us into 39911 thinking we need to quit. 39912 39913 Sending a WM_DESTROY is apparently never correct anyhow, see [1] 39914 39915 So: 39916 39917 1/ Use DestroyWindow() to destroy the clipboard messaging window when cleaning 39918 up for retry or exit in winClipboardProc (the clipboard thread main proc) 39919 39920 2/ Send a special WM_WM_QUIT message in winClipboardWindowDestroy() from the X 39921 server thread when the X server is resetting. 39922 39923 3/ When processing that WM_WM_QUIT message in the clipboard thread, cause the 39924 clipboard window to PostQuitMessage(), which causes the clipboard thread to 39925 exit. 39926 39927 [1] http://blogs.msdn.com/b/oldnewthing/archive/2011/09/26/10216420.aspx 39928 39929 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 39930 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 39931 39932commit b4a08e642b977b4bbc892ff1d96ecc0cf6e2ca54 39933Author: Jon TURNEY <jon.turney@dronecode.org.uk> 39934Date: Thu Nov 21 15:18:48 2013 +0000 39935 39936 hw/xwin: Improve reliability of clipboard X->Windows pastes 39937 39938 Sometimes, particularly with large clipboard pastes to Windows, we could end up 39939 waiting for the timeout to expire, rather than pasting the data. 39940 39941 Various changes to improve reliability: 39942 39943 1. Use XFlush() not XSync() in winProcessXEventsTimeout(). 39944 39945 It makes no sense to ensure we have received replies to outstanding requests if 39946 we are going to wait for them using select() 39947 39948 2. Add XFlush() to winClipboardProc() 39949 39950 Make sure we have sent any requests before we wait using select() 39951 39952 3. Don't use FD_ISSET() to check which fd is ready 39953 39954 This looks like a Cygwin select() bug in that it sometimes returns 0 with an 39955 empty fd set before the timeout expires, but a fd appears to be ready. 39956 39957 Add select() return value to debug output when we are warning that this has 39958 happened. 39959 39960 4. Drain event queues before entering select() 39961 39962 Unconditionally drain event queues before entering select(). This seems to be 39963 the recommended way of writing select() and X event processing loops. 39964 39965 winClipboardFlushXEvents() checks using XPending(), and 39966 winClipboardFlushWindowsMessageQueue() checks using PeekMessage() so this is 39967 safe against blocking, but means that may not need to enter select() at all 39968 sometimes. 39969 39970 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 39971 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 39972 39973commit c03f9e23c2b75af410b2fc76ca3f1aa9e979dcc4 39974Author: Jon TURNEY <jon.turney@dronecode.org.uk> 39975Date: Tue Sep 24 15:09:22 2013 +0100 39976 39977 hw/xwin: Add controls for enabling/disabling monitoring of PRIMARY selection 39978 39979 xwinclip: Add -noprimary option 39980 Xwin: Add -primary and -noprimary options and tray-menu control 39981 39982 v2: 39983 Use Bool type for fPrimarySelection 39984 Add -noprimary to usage message 39985 Fix indentation in hw/xwin/winwndproc.c 39986 39987 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 39988 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 39989 39990commit 851b50419907276fdc781b6be2d42b38849bbd77 39991Author: Jon TURNEY <jon.turney@dronecode.org.uk> 39992Date: Tue Sep 24 16:02:37 2013 +0100 39993 39994 hw/xwin: Retrieve TARGETS to avoid unnecessary failing conversion attempts 39995 39996 See http://cygwin.com/ml/cygwin-xfree/2013-07/msg00016.html 39997 39998 It looks like the change in a9aca218f557c723e637287272819a7c17174e1e had some 39999 unforseen consequences. 40000 40001 If the X11 selection contents are not convertable to COMPOUND_TEXT, UTF8_STRING 40002 or STRING format (for example, if it is an image), after those conversion 40003 attempts have failed, we sit in winProcessXEventsTimeout() until the timeout 40004 expires. 40005 40006 It also seems that maybe gnuplot doesn't respond correctly to this sequence of 40007 conversion requests and doesn't reply to some of them, which also causes us to 40008 sit in winProcessXEventsTimeout() until the timeout expires. 40009 40010 The Windows application which has requested the clipboard contents via 40011 GetClipboardContents() is blocked until we return from WM_RENDERFORMAT, so 40012 sitting waiting for this timeout to expire should be avoided. 40013 40014 So instead, explicitly request conversion to the TARGETS target, choose 40015 the most preferred format, and request conversion to that. 40016 40017 Also: if there is no owned selection, there is nothing to paste, so don't bother 40018 trying to convert it. 40019 40020 v2: Fix compilation with -Werror=declaration-after-statement 40021 40022 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40023 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40024 40025commit 4db1241037e3fe8f0a46888377b8fef40bae9065 40026Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40027Date: Tue Sep 24 13:30:32 2013 +0100 40028 40029 hw/xwin: In SelectionNotify, don't pointlessly retrieve just the size of the property 40030 40031 Don't pointlessly retrieve just the size of the property, if we are then going 40032 to assume we can retrieve the whole property in one request anyhow... 40033 40034 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40035 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40036 40037commit c5ad92077e7eb32590fafa92d697cc4173f7e57b 40038Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40039Date: Tue Sep 24 12:23:42 2013 +0100 40040 40041 hw/xwin: In SelectionNotify, delete the property containing returned data after we have retrieved it 40042 40043 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40044 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40045 40046commit ff018d88b9f0fe23ba96c81b53d58532baf4df42 40047Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40048Date: Thu Oct 2 14:10:50 2014 +0100 40049 40050 hw/xwin: Remove some redundant clipboard externs, now defined in winglobals.h 40051 40052 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40053 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40054 40055commit 63bb5c5ef16edf652179770294dcca4fc07dc992 40056Author: Thierry Reding <treding@nvidia.com> 40057Date: Wed Oct 29 16:59:27 2014 +0100 40058 40059 xv: Add missing gcstruct.h include 40060 40061 Commit ea3f3b0786d5 (xv: Move xf86 XV color key helper to core.) added 40062 code that uses internals of struct _GC. This structure is defined in the 40063 include/gcstruct.h header which wasn't included by the source file, only 40064 gc.h was. That caused the following build failure: 40065 40066 CC xvmain.lo 40067 Xext/xvmain.c: In function 'XvFillColorKey': 40068 Xext/xvmain.c:1114:13: error: dereferencing pointer to incomplete type 40069 (*gc->ops->PolyFillRect) (pDraw, gc, nbox, rects); 40070 ^ 40071 40072 Fix this by including the correct header file. 40073 40074 Signed-off-by: Thierry Reding <treding@nvidia.com> 40075 Tested-by: Keith Packard <keithp@keithp.com> 40076 Reviewed-by: Keith Packard <keithp@keithp.com> 40077 Signed-off-by: Keith Packard <keithp@keithp.com> 40078 40079commit 882f2d10d99a04a96afc0ce0c8937e16bec3afb5 40080Author: Axel Davy <axel.davy@ens.fr> 40081Date: Wed Oct 29 13:31:42 2014 +0100 40082 40083 Fix present_notify to return right away when querying current or past msc. 40084 40085 When the target msc is past or is the current one, we want to get immediate 40086 feedback. This patch fixes this behaviour. 40087 40088 Signed-off-by: Axel Davy <axel.davy@ens.fr> 40089 Reviewed-by: Keith Packard <keithp@keithp.com> 40090 Signed-off-by: Keith Packard <keithp@keithp.com> 40091 40092commit 72a0754f4605493fd60471557422f88a97eb356a 40093Author: Aaron Plattner <aplattner@nvidia.com> 40094Date: Wed Oct 29 19:19:18 2014 -0700 40095 40096 xfree86: Bump ABI versions (video: 19, extension: 9) 40097 40098 Among other things, commit b851ca968b7cce6d1a6438c05d3d5c8832249704 added a 40099 NameWindowPixmap function pointer to ScreenRec, shifting some of the fields 40100 around. 40101 40102 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 40103 Reviewed-by: Keith Packard <keithp@keithp.com> 40104 Signed-off-by: Keith Packard <keithp@keithp.com> 40105 40106commit b227d974569b4c315a72b85fe839c5f455396678 40107Author: Peter Hutterer <peter.hutterer@who-t.net> 40108Date: Thu Oct 30 08:54:11 2014 +1000 40109 40110 include: fix compiler warning about casting int to uint16_t 40111 40112 /usr/include/xorg/misc.h:141:30: warning: implicit conversion loses integer 40113 precision: 'int' to 'uint16_t' (aka 'unsigned short') [-Wconversion] 40114 return ((x & 0xff) << 8) | ((x >> 8) & 0xff); 40115 ~~~~~~ ~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~ 40116 40117 Function sig is a uint16_t, so just force the cast. 40118 40119 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 40120 40121commit e9db7682028bb0464c211c1f7bb6983fcfb6f37b 40122Author: Keith Packard <keithp@keithp.com> 40123Date: Tue Oct 28 21:32:24 2014 -0700 40124 40125 Update to version 1.16.99.901 40126 40127 1.17 RC1 40128 40129 Signed-off-by: Keith Packard <keithp@keithp.com> 40130 40131commit 5ff74d68b0cce85e041251efdc76595eac3c89e4 40132Author: Keith Packard <keithp@keithp.com> 40133Date: Tue Oct 28 21:31:40 2014 -0700 40134 40135 xfree86: Include 'drivers' directory with distribution 40136 40137 Even if we don't build it locally, make sure it gets included with the release. 40138 40139 Signed-off-by: Keith Packard <keithp@keithp.com> 40140 40141commit f36f16f0551aa091c1ff154875aebe3d29cf51e7 40142Author: Keith Packard <keithp@keithp.com> 40143Date: Tue Oct 28 21:30:23 2014 -0700 40144 40145 dri2: Distribute new pci_ids headers 40146 40147 Create hw/xfree86/dri2/pci_ids/Makefile.am which includes all of the new 40148 pci id files in the tarballs. Build that from configure.ac, and run it 40149 from dri2/Makefile.am 40150 40151 Signed-off-by: Keith Packard <keithp@keithp.com> 40152 40153commit 65dd1ba7b3e00243a76ec691e566ef7c8dd94855 40154Author: Keith Packard <keithp@keithp.com> 40155Date: Wed Oct 22 14:48:10 2014 -0700 40156 40157 dix: Untwist transformAbsolute logic, eliminate uninitialized value warnings 40158 40159 tranformAbsolute has a pretty simple job, that of running the X/Y 40160 values from a device through the transformation matrix. The tricky bit 40161 comes when the current device state doesn't include one of the 40162 values. In that case, the last delivered value is back-converted to 40163 device space and used instead. 40164 40165 The logic was twisted though, confusing GCC's uninitialized value 40166 detection logic and emitting warnings. 40167 40168 This has been fixed by changing the code to: 40169 40170 1) Detect whether the ValuatorMask includes X/Y values 40171 2) If either are missing, back-convert the current values into ox/oy 40172 3) When X/Y are present, set ox/oy to the current value 40173 4) Transform 40174 5) Store X/Y values if changed or if they were set before. 40175 40176 Signed-off-by: Keith Packard <keithp@keithp.com> 40177 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 40178 40179commit 0fbbdb37c87b5824729f65c7fbac05223024fd27 40180Author: Adam Jackson <ajax@redhat.com> 40181Date: Wed Oct 22 13:19:18 2014 -0400 40182 40183 composite: Wrap GetSpans 40184 40185 GetSpans should flush composition from children to parent, just like 40186 GetImage and SourceValidate. Fortunately no one is likely to have 40187 noticed, since to hit this you're already deep into failure town. 40188 40189 Signed-off-by: Adam Jackson <ajax@redhat.com> 40190 Reviewed-by: Keith Packard <keithp@keithp.com> 40191 Signed-off-by: Keith Packard <keithp@keithp.com> 40192 40193commit da70c7d556bbf21ad495c26e982e2e0f2d7de6c1 40194Author: Keith Packard <keithp@keithp.com> 40195Date: Wed Oct 22 14:27:26 2014 -0700 40196 40197 xkb: Initialize 'bad' Atom in _XkbSetNamesCheck 40198 40199 When _XkbCheckAtoms returns NULL for an error, it always sets the 40200 error return code, but GCC can't figure that out, so just initialize 40201 the local variable, 'bad', in _XkbSetNamesCheck to eliminate the warning. 40202 40203 Signed-off-by: Keith Packard <keithp@keithp.com> 40204 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 40205 40206commit 2566835b4374edb3e5a8353d4f7c9e7ec4851c57 40207Author: Keith Packard <keithp@keithp.com> 40208Date: Wed Oct 22 14:24:55 2014 -0700 40209 40210 os: Eliminate uninitialized value warnings from access.c 40211 40212 The ConvertAddr function doesn't reliably set the 'addr' return value, 40213 and so callers are getting flagged for using potentially uninitialized 40214 values. Initialize the value in the callers to NULL and then go ahead 40215 and check for NULL values before using them. 40216 40217 Signed-off-by: Keith Packard <keithp@keithp.com> 40218 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 40219 40220commit 1b94fd77792310c80b0a2bcf4bf6d4e4c4c23bca 40221Author: Alex Orange <crazycasta@gmail.com> 40222Date: Fri Oct 3 15:41:38 2014 -0600 40223 40224 fb: Fix Bresenham algorithms for commonly used small segments. 40225 40226 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=54168 40227 40228 Fix errors introducted in 863d528a9f76d0e8f122aebf19f8564a4c67a938. Said 40229 patch does indeed remove the problematic writes to bad memory, however 40230 it also introduces errors in the algoritm. This patch has the effect of 40231 reverting said patch and adding an if in the proper location to catch 40232 the out of bounds memory write without causing problems to the overall 40233 algorithm. 40234 40235 Signed-off-by: Alex Orange <crazycasta@gmail.com> 40236 Reviewed-by: Peter Harris <pharris@opentext.com> 40237 Tested-by: Peter Harris <pharris@opentext.com> 40238 Signed-off-by: Keith Packard <keithp@keithp.com> 40239 40240commit ea5b2b0a2e2143ad1414fcbdc081b5d584588346 40241Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40242Date: Tue Oct 21 15:03:55 2014 +0100 40243 40244 os: -displayfd should check ports up to 65535 40245 40246 -displayfd should check ports up to 65535 40247 40248 Noticed during https://cygwin.com/ml/cygwin-xfree/2014-07/msg00024.html 40249 40250 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40251 Reviewed-by: Adam Jackson <ajax@redhat.com> 40252 Signed-off-by: Keith Packard <keithp@keithp.com> 40253 40254commit 3a123fbdda56b24dc01a2308137144dc960e0d61 40255Merge: 9b29fa957 7ea4a21fe 40256Author: Keith Packard <keithp@keithp.com> 40257Date: Mon Oct 27 15:39:26 2014 -0700 40258 40259 Merge remote-tracking branch 'jturney/master' 40260 40261 Conflicts: 40262 hw/xwin/winpixmap.c 40263 40264 Adam removed a pile of windows code, including this file. 40265 40266commit 9b29fa957a397664463c7c78fbcc2f34d1993271 40267Author: Chris Wilson <chris@chris-wilson.co.uk> 40268Date: Thu Oct 16 14:09:08 2014 +0100 40269 40270 Xext/shm: Detach SHM segment after Pixmap is released 40271 40272 The GPU may still have a reference to the SHM segment which would only 40273 be finally released when the Pixmap is destroy. So we can only detach 40274 the SHM segment (and thereby making the memory unaccessible) after the 40275 backend has had a chance to flush any remaining references. 40276 40277 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=85058 40278 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 40279 Reported-and-tested-by: gedgon@gmail.com 40280 Reviewed-by: Adam Jackson <ajax@redhat.com> 40281 Signed-off-by: Keith Packard <keithp@keithp.com> 40282 40283commit 5adc20179e9818d51e1cd79bfc8a8271786c3949 40284Author: Keith Packard <keithp@keithp.com> 40285Date: Mon Oct 27 15:28:14 2014 -0700 40286 40287 modesetting: Skip kernel work-around on error in crtc to kernel msc 40288 40289 ms_crtc_msc_to_kernel_msc attempts to work around kernel 40290 inconsistencies in reporting msc values by comparing the expected 40291 value with the reported value. If the kernel fails to 40292 actually provide its current values, then just skip the work around 40293 steps as there's really nothing better we can do. 40294 40295 Signed-off-by: Keith Packard <keithp@keithp.com> 40296 40297commit 59b12c454d9c4b5a26c6ad87c53abc671b15ce37 40298Merge: 5574a0a07 cac4b064f 40299Author: Keith Packard <keithp@keithp.com> 40300Date: Mon Oct 27 14:48:55 2014 -0700 40301 40302 Merge remote-tracking branch 'anholt/modesetting-dri2-no-pageflip' 40303 40304commit 5574a0a07e83ff989c074c44d3ea9db0a819f472 40305Author: Keith Packard <keithp@keithp.com> 40306Date: Mon Oct 27 13:38:21 2014 -0700 40307 40308 composite: Skip SetWindowPixmap when pixmap isn't changing 40309 40310 Check the current window pixmap to see if it matches the new pixmap 40311 and avoid calling SetWindowPixmap in that case. 40312 40313 x11perf -ucreate highlights an instance where compCreateWindow reassigns 40314 the same Pixmap to the Window. Currently this triggers an expensive 40315 invalidation of the entire window hierachy, making sure that DRI2 40316 clients are kept informed of the handle changes. However, as the backing 40317 Pixmap for the Window is actually unchanged, there is no need to 40318 do anything in this case. 40319 40320 Signed-off-by: Keith Packard <keithp@keithp.com> 40321 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 40322 Tested-by: Chris Wilson <chris@chris-wilson.co.uk> 40323 40324commit df6a3a07f60c40bab2b59b9b3ba0c20c8fb7e95d 40325Author: Yogish Kulkarni <yogishk@nvidia.com> 40326Date: Wed Oct 22 12:55:28 2014 +0530 40327 40328 Skip SetCursor when vtSema is FALSE. 40329 40330 Consider below sequence - 40331 1) Cursor is removed : isUp will be FALSE if HW cursor is set. 40332 2) VT switched away from X : vtSema becomes FALSE. 40333 3) xf86CursorSetCursor is called with non-null CursorPtr : 40334 Saves the passed in CursorPtr, fallbacks to SW cursor and invokes 40335 spriteFuncs->SetCursor which saves the area under cursor and restores 40336 the cursor. This sets isUp to TRUE and as vtSema is FALSE saved data 40337 is garbage. 40338 4) VT switched to X : vtSema becomes TRUE. xf86Cursor enable fb access 40339 is called which will remove the SW cursor, i.e copies saved data in #3 40340 to screen. 40341 40342 This results to momentary garbage data on screen. Hence when !vtSema 40343 skip spriteFuncs->SetCursor. 40344 40345 X.Org Bug 85313 <https://bugs.freedesktop.org/show_bug.cgi?id=85313> 40346 40347 Signed-off-by: Yogish Kulkarni <yogishk@nvidia.com> 40348 Reviewed-by: Adam Jackson <ajax@redhat.com> 40349 Signed-off-by: Keith Packard <keithp@keithp.com> 40350 40351commit cac4b064f9f66435430f61568c6a516c54bf3c40 40352Author: Eric Anholt <eric@anholt.net> 40353Date: Mon Dec 30 17:23:38 2013 -0800 40354 40355 modesetting: Add support for DRI2 with glamor. 40356 40357 This is derived from the intel driver DRI2 code, with swapchain and 40358 pageflipping dropped, functions renamed, and vblank event management 40359 shared code moved to a vblank.c for reuse by Present. 40360 40361 This allows AIGLX to load, which means that you get appropriate 40362 visuals exposed in GL, along with many extensions under 40363 direct-rendering that require presence in GLX (which aren't supported 40364 in glxdriswrast.c). 40365 40366 v2: Drop unused header includes in pageflip.c, wrap in #ifdef GLAMOR. 40367 Drop triple-buffering, which was totally broken in practice (I'll 40368 try to fix this later). Fix up some style nits. Document the 40369 general flow of pageflipping and why, rename the DRI2 frame event 40370 type enums to reflect what they're for, and handle them in a 40371 single switch statement so you can understand the state machine 40372 more easily. 40373 v3: Drop pageflipping entirely -- it's unstable on my Intel laptop 40374 (not that the normal 2D driver is stable with pageflipping for 40375 me), and I won't get it fixed before the merge window. It now 40376 passes all of the OML_sync_control tests from Jamey and Theo 40377 (except for occasional warns in timing -fullscreen -divisor 2). 40378 v4: Fix doxygen at the top of vblank.c 40379 40380 Signed-off-by: Eric Anholt <eric@anholt.net> 40381 Reviewed-by: Adam Jackson <ajax@redhat.com> 40382 40383commit 7064b00d478646cb428e427df2d3f91d7022c582 40384Author: Eric Anholt <eric@anholt.net> 40385Date: Mon Dec 30 00:25:37 2013 -0800 40386 40387 modesetting: Export two functions I want to reuse from DRI2/Present. 40388 40389 This renames dumb_get_bo_from_handle(), since it wasn't using a handle 40390 (GEM terminology) but a dmabuf fd. 40391 40392 Signed-off-by: Eric Anholt <eric@anholt.net> 40393 Reviewed-by: Adam Jackson <ajax@redhat.com> 40394 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 40395 40396commit 34a98c8103a6f725283d0d1ee4084bf387b93019 40397Author: Eric Anholt <eric@anholt.net> 40398Date: Wed Oct 8 04:12:49 2014 -0700 40399 40400 dri2: Automatically fill in the driver name if the DDX doesn't provide it. 40401 40402 This will be used by the modesetting driver to support DRI2 across all 40403 hardware that can support glamor, and could potentially be used by 40404 other drivers that have to support DRI2 on sets of hardware with 40405 multiple Mesa drivers. 40406 40407 This logic is the same as what's present in the Mesa driver loader, 40408 except for the lack of nouveau_vieux support (which requires a 40409 predicate on the device). 40410 40411 v2: Fix duplicated assignment of info->driverName. 40412 40413 Signed-off-by: Eric Anholt <eric@anholt.net> 40414 Reviewed-by: Adam Jackson <ajax@redhat.com> 40415 40416commit 470d97e1d176981dbd8087c621e54caab0d6c73c 40417Author: Eric Anholt <eric@anholt.net> 40418Date: Tue Dec 31 13:52:49 2013 -0800 40419 40420 dri2: Import a copy of Mesa's PCI ID -> driver name mappings. 40421 40422 This comes from Mesa commit acdcef6788beaa2a1532e13ff84c3e246b8025ed 40423 40424 Previously, each driver had to tell DRI2 what GL driver object should 40425 be loaded. Originally for a 2D driver that was a matter of giving the 40426 constant string for the vendor name, same as the driver's name. For a 40427 driver that's trying to handle multiple generations of hardware with 40428 different Mesa driver filenames, the driver had to bake in a mapping 40429 from PCI ID to the appropriate driver name in Mesa, which seems like a 40430 pretty awful layering violation (and one that was fixed with DRI3) 40431 40432 As of January, Mesa now handles the mapping from a DRI fd to the 40433 driver name on its own, but the AIGLX loader still relies on DRI2 for 40434 choosing the filename. Instead of propagating the PCI ID list from 40435 each 2D driver to the modesetting driver, import a central copy of the 40436 PCI ID list so that drivers can stop handling this themselves. (Some 40437 day, when AIGLX transitions to EGL, we can drop the DRI2 filename 40438 setup entirely). 40439 40440 Signed-off-by: Eric Anholt <eric@anholt.net> 40441 Reviewed-by: Adam Jackson <ajax@redhat.com> 40442 40443commit 839f05329ddb6f9b5b988edf3a2c3a1f856727f7 40444Author: Adam Jackson <ajax@redhat.com> 40445Date: Fri Feb 28 16:35:10 2014 -0500 40446 40447 dix: Retype and repack GC 40448 40449 LP64: 160 bytes before, 128 after. 40450 40451 Signed-off-by: Adam Jackson <ajax@redhat.com> 40452 Reviewed-by: Keith Packard <keithp@keithp.com> 40453 40454commit 73e2383b7350723256894c5076d0c731aec5cd1f 40455Author: Adam Jackson <ajax@redhat.com> 40456Date: Fri Jun 20 13:31:20 2014 -0400 40457 40458 dix: Always store GC client clip as a region (v2) 40459 40460 Again, this changes FixesCreateRegionFromGC to throw BadMatch when fed a 40461 GC with no client clip. 40462 40463 v2: Fix Xnest and some variable names (Keith) 40464 40465 Reviewed-by: Keith Packard <keithp@keithp.com> 40466 Signed-off-by: Adam Jackson <ajax@redhat.com> 40467 40468commit e7b9295551d55b42b416ba70cce11dae79ece5e7 40469Author: Adam Jackson <ajax@redhat.com> 40470Date: Fri Sep 26 13:41:59 2014 -0400 40471 40472 fb: Hide fbPush{Pattern,Fill} 40473 40474 Reviewed-by: Keith Packard <keithp@keithp.com> 40475 Signed-off-by: Adam Jackson <ajax@redhat.com> 40476 40477commit e31d16a82d83738740f67743b759c006d9845e79 40478Author: Adam Jackson <ajax@redhat.com> 40479Date: Fri Sep 26 13:32:37 2014 -0400 40480 40481 fb: Hide fbDots 40482 40483 Reviewed-by: Keith Packard <keithp@keithp.com> 40484 Signed-off-by: Adam Jackson <ajax@redhat.com> 40485 40486commit 5db0f067ed4ff24c48cdb7998e4a52ad80037cb7 40487Author: Adam Jackson <ajax@redhat.com> 40488Date: Fri Sep 26 13:30:06 2014 -0400 40489 40490 fb: Hide Bresenham line details 40491 40492 Reviewed-by: Keith Packard <keithp@keithp.com> 40493 Signed-off-by: Adam Jackson <ajax@redhat.com> 40494 40495commit b184a863c111834e28564d963c0733a30833f128 40496Author: Adam Jackson <ajax@redhat.com> 40497Date: Tue Jul 8 12:23:48 2014 -0400 40498 40499 fb: Hide some zero-width line details 40500 40501 Reviewed-by: Keith Packard <keithp@keithp.com> 40502 Signed-off-by: Adam Jackson <ajax@redhat.com> 40503 40504commit 266cd552bd54fd3e7a83ae7178a303285857631e 40505Author: Adam Jackson <ajax@redhat.com> 40506Date: Tue Jul 8 12:13:13 2014 -0400 40507 40508 fb: Hide glyph implementation details 40509 40510 Reviewed-by: Keith Packard <keithp@keithp.com> 40511 Signed-off-by: Adam Jackson <ajax@redhat.com> 40512 40513commit e572bcc7f4236b7e0f23ab762f225b3bce37db59 40514Author: Adam Jackson <ajax@redhat.com> 40515Date: Mon Jun 23 15:24:35 2014 -0400 40516 40517 fb: Remove even/odd tile slow-pathing 40518 40519 Again, clearly meant to be a fast path, but this turns out not to be the 40520 case. 40521 40522 Reviewed-by: Keith Packard <keithp@keithp.com> 40523 Signed-off-by: Adam Jackson <ajax@redhat.com> 40524 40525commit 836bb27726441e048bb300664343a136bc596a5b 40526Author: Adam Jackson <ajax@redhat.com> 40527Date: Tue Jun 17 12:25:42 2014 -0400 40528 40529 fb: Remove unused fbReduceRasterOp 40530 40531 Appears to have been cargo-culted in from cfb and then never used. 40532 40533 Reviewed-by: Keith Packard <keithp@keithp.com> 40534 Signed-off-by: Adam Jackson <ajax@redhat.com> 40535 40536commit 167ccd33e0d2c4bd68aebcdba5a512d24c5eb49a 40537Author: Adam Jackson <ajax@redhat.com> 40538Date: Tue Jun 17 12:21:19 2014 -0400 40539 40540 fb: Move fbStipple*Bits near their only consumer 40541 40542 And remove fbStippleTable since gcc can't figure that out itself. 40543 40544 Reviewed-by: Keith Packard <keithp@keithp.com> 40545 Signed-off-by: Adam Jackson <ajax@redhat.com> 40546 40547commit 7430fdb689678b98ac63f5a8dad13719bac777e0 40548Author: Adam Jackson <ajax@redhat.com> 40549Date: Tue Dec 17 14:46:38 2013 -0500 40550 40551 fb: Remove even/odd stipple slow-pathing 40552 40553 This is clearly meant to be a fast path, but it appears to be a net 40554 loss at this point. 40555 40556 If you really wanted to ricer-tune here, note that the inner loop of 40557 fbBltOne is a bit too complicated for gcc (at least 4.8.2), it doesn't 40558 specialize the loop for the handful of legal values of bitsPerDst, which 40559 means computing at runtime what could have been const-propped. 40560 40561 Reviewed-by: Keith Packard <keithp@keithp.com> 40562 Signed-off-by: Adam Jackson <ajax@redhat.com> 40563 40564commit a1983736853814d34450c0a0de1c4f6ee308423d 40565Author: Adam Jackson <ajax@redhat.com> 40566Date: Tue Jun 17 11:24:06 2014 -0400 40567 40568 fb: FB_SHIFT is 5 (and FB_UNIT is 32) 40569 40570 The other paths don't build or work, PCI and other buses are almost 40571 always 32 bit data paths, and X doesn't really support pixels bigger 40572 than that anyway. 40573 40574 Reviewed-by: Keith Packard <keithp@keithp.com> 40575 Signed-off-by: Adam Jackson <ajax@redhat.com> 40576 40577commit 3d35bd6b79957a3441abd87a377f0beb058997ed 40578Author: Adam Jackson <ajax@redhat.com> 40579Date: Mon Mar 24 12:01:54 2014 -0400 40580 40581 fb: Eliminate fbLaneTable, staticize fb{8,16,32}Lane 40582 40583 gcc doesn't appear to be smart enough to fold away the indirection here, 40584 even if you make fbLaneTable const. 40585 40586 Reviewed-by: Keith Packard <keithp@keithp.com> 40587 Signed-off-by: Adam Jackson <ajax@redhat.com> 40588 40589commit 86ce6262975a7d53d9331723beb49cbe7bd02b48 40590Author: Adam Jackson <ajax@redhat.com> 40591Date: Fri Sep 26 12:39:38 2014 -0400 40592 40593 mi: Fold micursor.c into mipointer.c 40594 40595 Reviewed-by: Keith Packard <keithp@keithp.com> 40596 Signed-off-by: Adam Jackson <ajax@redhat.com> 40597 40598commit 3b63900e90f44cf85a4d3391d23407b3bb6db52b 40599Author: Adam Jackson <ajax@redhat.com> 40600Date: Fri Sep 26 12:34:15 2014 -0400 40601 40602 mi: Fold mipolyutil.c into mipoly.c 40603 40604 Reviewed-by: Keith Packard <keithp@keithp.com> 40605 Signed-off-by: Adam Jackson <ajax@redhat.com> 40606 40607commit 21b041ef48c49ff6fc06f6e4482ffe54cb763493 40608Author: Adam Jackson <ajax@redhat.com> 40609Date: Fri Sep 26 12:27:22 2014 -0400 40610 40611 mi: Fold mipoly{con,gen}.c into mipoly.c 40612 40613 Reviewed-by: Keith Packard <keithp@keithp.com> 40614 Signed-off-by: Adam Jackson <ajax@redhat.com> 40615 40616commit 7679afd4da8b86aed27e5916ba723116a3c8bb4a 40617Author: Adam Jackson <ajax@redhat.com> 40618Date: Fri Sep 26 12:01:37 2014 -0400 40619 40620 mi: Fold mifpolycon.c into miarc.c 40621 40622 Also put mifpoly.h on a diet, and stop including it from places that 40623 don't need it. 40624 40625 Reviewed-by: Keith Packard <keithp@keithp.com> 40626 Signed-off-by: Adam Jackson <ajax@redhat.com> 40627 40628commit f307ef10f4c33da4b5ae59800931741b0a431d75 40629Author: Adam Jackson <ajax@redhat.com> 40630Date: Mon Jul 21 17:22:07 2014 -0400 40631 40632 mi: Fold mispans.c into miwideline.c 40633 40634 Reviewed-by: Keith Packard <keithp@keithp.com> 40635 Signed-off-by: Adam Jackson <ajax@redhat.com> 40636 40637commit 707965407a3c907058b89610e73e02989fd0b552 40638Author: Adam Jackson <ajax@redhat.com> 40639Date: Fri May 23 12:10:29 2014 -0400 40640 40641 mi: Unexport arc fill implementation details 40642 40643 Reviewed-by: Keith Packard <keithp@keithp.com> 40644 Signed-off-by: Adam Jackson <ajax@redhat.com> 40645 40646commit ce8fff5cd681523088c14b97ce4fb6e4fb658321 40647Author: Adam Jackson <ajax@redhat.com> 40648Date: Fri May 23 12:06:03 2014 -0400 40649 40650 mi: Unexport subpixel polygon implementation details 40651 40652 Reviewed-by: Keith Packard <keithp@keithp.com> 40653 Signed-off-by: Adam Jackson <ajax@redhat.com> 40654 40655commit a085ba82bde375b5c884a40b410b563815c58a9e 40656Author: Adam Jackson <ajax@redhat.com> 40657Date: Fri May 23 11:55:05 2014 -0400 40658 40659 mi: Unexport span group implementation details 40660 40661 Reviewed-by: Keith Packard <keithp@keithp.com> 40662 Signed-off-by: Adam Jackson <ajax@redhat.com> 40663 40664commit c15c886ee70f9144bc4603b918310f229312aadc 40665Author: Adam Jackson <ajax@redhat.com> 40666Date: Fri May 23 11:50:07 2014 -0400 40667 40668 mi: Unexport polygon edge table implementation details 40669 40670 Reviewed-by: Keith Packard <keithp@keithp.com> 40671 Signed-off-by: Adam Jackson <ajax@redhat.com> 40672 40673commit 9bdc9b0113915de3d536b2b3f025915379c2126f 40674Author: Adam Jackson <ajax@redhat.com> 40675Date: Fri May 23 11:21:28 2014 -0400 40676 40677 mi: Unexport wide line details 40678 40679 XAA wanted these, once upon a time, but that's gone now. 40680 40681 Reviewed-by: Keith Packard <keithp@keithp.com> 40682 Signed-off-by: Adam Jackson <ajax@redhat.com> 40683 40684commit d181e52ceb9ae44e1faa8d5af8805f43328da6c2 40685Author: Keith Packard <keithp@keithp.com> 40686Date: Fri Oct 24 11:56:23 2014 -0700 40687 40688 glamor: Free converted bits in _glamor_upload_bits_to_pixmap_texture fast path 40689 40690 When uploading bits to a texture which need reformatting to match a 40691 supported GL format, a temporary buffer is allocated to hold the 40692 reformatted bits. This gets freed in the general path, but is not 40693 freed in the fast path because that includes an early return before 40694 the call to free. 40695 40696 This patch removes the early return and places the general case under 40697 an 'else' block, so that both paths reach the call to free. 40698 40699 Signed-off-by: Keith Packard <keithp@keithp.com> 40700 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 40701 40702commit 55b27ed70cf9dfa9b9dffe46e9a3191bfda38f7c 40703Author: Andreas Hartmetz <ahartmetz@gmail.com> 40704Date: Sat Oct 4 18:13:04 2014 +0200 40705 40706 glamor: Don't free memory we are going to use. 40707 40708 glamor_color_convert_to_bits() returns its second argument on 40709 success, NULL on error, and need_free_bits already makes sure that 40710 "bits" aliasing converted_bits is freed in the success case. 40711 Looks like the memory leak that was supposed to be fixed in 40712 6e50bfa706cd3ab884c933bf1f17c221a6208aa4 only occurred in the error 40713 case. 40714 40715 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 40716 Signed-off-by: Keith Packard <keithp@keithp.com> 40717 40718commit 16a32c53f6e9ad1f3284d4596edfa33e9efb740e 40719Author: Alan Coopersmith <alan.coopersmith@oracle.com> 40720Date: Thu Oct 9 05:42:09 2014 -0700 40721 40722 If fork fails in System(), don't fallthrough to exec() 40723 40724 In the unlikely event of a failure in creating processes, signal 40725 masks will fall from the panels above you. Secure your mask before 40726 telling your child what to do, since it won't exist, and you will 40727 instead cause the server itself to be replaced by a shell running 40728 the target program. 40729 40730 Found by Coverity #53397: Missing break in switch 40731 Execution falls through to the next case statement or default; 40732 this might indicate a common typo. 40733 In System: Missing break statement between cases in switch statement (CWE-484) 40734 40735 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 40736 Reviewed-by: Matthieu Herrb <matthieu@herrb.eu> 40737 Signed-off-by: Keith Packard <keithp@keithp.com> 40738 40739commit 7e5bc49d1ed2c78c321da79bdbc99b90c5f95b38 40740Author: Alan Coopersmith <alan.coopersmith@oracle.com> 40741Date: Thu Oct 9 05:42:08 2014 -0700 40742 40743 Allocate enough room for both reset & flags attributes 40744 40745 ctx_attribs had room for 3 pairs of attributes, but if both flags & reset 40746 attributes were being returned it was storing 4 pairs in the array. 40747 40748 Found by Coverity #53442: Out-of-bounds write 40749 This could cause an immediate crash or incorrect computations. 40750 In create_driver_context: Out-of-bounds write to a buffer (CWE-119) 40751 40752 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 40753 Reviewed-by: Adam Jackson <ajax@redhat.com> 40754 Signed-off-by: Keith Packard <keithp@keithp.com> 40755 40756commit cffd4e4a4ee615d8583eae78b35017e0d1bfa4f0 40757Author: Adam Jackson <ajax@redhat.com> 40758Date: Thu Oct 9 09:58:06 2014 +0200 40759 40760 mi: Mark the span blit routines as _X_COLD 40761 40762 On gcc, __attribute__((cold)) means: 40763 40764 - consider calls to the function to be unlikely for branch prediction 40765 - optimize the function for size 40766 - emit the function in a dedicated cold text section 40767 40768 It's not worth deleting these routines even though there are no longer 40769 in-tree consumers, but we can at least keep them out of i$ at runtime. 40770 40771 Signed-off-by: Adam Jackson <ajax@redhat.com> 40772 Reviewed-by: Keith Packard <keithp@keithp.com> 40773 Signed-off-by: Keith Packard <keithp@keithp.com> 40774 40775commit de55aafa8f6a9e0dec364bec920d6f91ef2b39f0 40776Merge: 462bf87c4 c79f824bf 40777Author: Keith Packard <keithp@keithp.com> 40778Date: Thu Oct 23 17:28:24 2014 -0700 40779 40780 Merge remote-tracking branch 'ajax/xwin' 40781 40782commit 462bf87c4d1c2211dd49a5ce62d01ff84ff33970 40783Author: Adam Jackson <ajax@redhat.com> 40784Date: Wed Oct 8 17:20:33 2014 +0200 40785 40786 render: Always store client clip as a region 40787 40788 This does have one semantic change. FixesCreateRegionFromPicture used to 40789 throw BadImplementation if you tried to create a region from a picture 40790 with no client clip. I changed that to BadMatch here since that more 40791 honestly describes what's going on. 40792 40793 Signed-off-by: Adam Jackson <ajax@redhat.com> 40794 Reviewed-by: Keith Packard <keithp@keithp.com> 40795 Signed-off-by: Keith Packard <keithp@keithp.com> 40796 40797commit 3f4edd2e3ff84c38df563b09c2e8c32404db38f7 40798Author: Adam Jackson <ajax@redhat.com> 40799Date: Wed Oct 8 17:18:32 2014 +0200 40800 40801 xinerama: Fix access mode in GetImage's drawable lookup 40802 40803 Signed-off-by: Adam Jackson <ajax@redhat.com> 40804 Reviewed-by: Julien Cristau <jcristau@debian.org> 40805 Signed-off-by: Keith Packard <keithp@keithp.com> 40806 40807commit 942e18e17e334d7a25f8c1dcc6f5c3711eeb6a7d 40808Author: William ML Leslie <william.leslie.ttg@gmail.com> 40809Date: Wed May 21 10:28:52 2014 +1000 40810 40811 Xephyr: option to disable grabbing the host 40812 40813 This patch makes it possible to use C-S key combinations 40814 within Xephyr without losing access to the host window manager's 40815 commands. 40816 40817 Signed-off-by: Keith Packard <keithp@keithp.com> 40818 Reviewed-by: Eric Anholt <eric@anholt.net> 40819 40820commit 7ea4a21feed5ac5e71cc6d8141c97d6362402c5d 40821Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40822Date: Fri Sep 19 13:37:50 2014 +0100 40823 40824 configure.ac: Force --disable-libdrm on Cygwin 40825 40826 This is now needed to avoid trying to build the modesetting driver by default 40827 when building the Xorg DDX on Cygwin. 40828 40829 http://tinderbox.x.org/builds/2014-09-18-0011/logs/xserver/#build 40830 40831 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40832 Reviewed-by: Adam Jackson <ajax@redhat.com> 40833 40834commit 883927d2e8733f4070b2a8bcc5ec8cc2d7a661b3 40835Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40836Date: Tue Sep 30 13:55:59 2014 +0100 40837 40838 configure.ac: Avoid "Your OS is unknown" warning when configuring for Cygwin 40839 40840 Don't emit "Your OS is unknown" warning when configuring with --enable-xorg to 40841 build the XOrg DDX for Cygwin. 40842 40843 The list of supported OSes is getting a bit unwieldy, so just remove it. 40844 40845 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40846 Reviewed-by: Adam Jackson <ajax@redhat.com> 40847 40848commit 549b3175afc46cee6e1b61841f854a5272eaf1f7 40849Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40850Date: Thu Oct 2 15:37:50 2014 +0100 40851 40852 hw/xwin: Fix unused variable warning in winCreateMsgWindow() 40853 40854 winmsgwindow.c:99:11: warning: variable ‘winClass’ set but not used [-Wunused-but-set-variable] 40855 40856 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40857 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40858 40859commit 5adfb566fe05bcf5f40b4adc1457862936d6d038 40860Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40861Date: Fri Oct 3 14:22:30 2014 +0100 40862 40863 hw/xwin: Fix redundant declaration warning in winprefslex.l 40864 40865 winprefslex.l:40:12: warning: redundant redeclaration of ‘yyparse’ [-Wredundant-decls] 40866 winprefsyacc.h:130:5: note: previous declaration of ‘yyparse’ was here 40867 40868 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40869 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40870 40871commit 8f062f73805765f5769dfd7b738c8dd10062fba6 40872Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40873Date: Fri Oct 3 14:12:57 2014 +0100 40874 40875 hw/xwin: Fix warning in yyerror() 40876 40877 winprefsyacc.y:257:3: warning: nested extern declaration of ‘yylineno’ [-Wnested-externs] 40878 40879 Promote yylineno declaration to file scope 40880 40881 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40882 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40883 40884commit f0f0c92a08f0ccf5a7ac447d117eb0341fe136d6 40885Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40886Date: Fri Oct 3 14:00:06 2014 +0100 40887 40888 hw/xwin: Fix const warning in winPrefsLoadPreferences() 40889 40890 winprefs.c:643:14: warning: assignment discards ‘const’ qualifier from pointer target type [enabled by default] 40891 40892 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40893 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40894 40895commit 593c6b19349538ec3b89bc48eb3f0bb4c0cebb23 40896Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40897Date: Thu Oct 2 15:35:54 2014 +0100 40898 40899 hw/xwin: Fix const warnings in winprefsyacc.y 40900 40901 winprefsyacc.y:174:5: warning: passing argument 1 of ‘AddMenuLine’ discards ‘const’ qualifier from pointer target type [enabled by default] 40902 winprefsyacc.y:67:13: note: expected ‘char *’ but argument is of type ‘const char *’ 40903 winprefsyacc.y:174:5: warning: passing argument 3 of ‘AddMenuLine’ discards ‘const’ qualifier from pointer target type [enabled by default] 40904 winprefsyacc.y:67:13: note: expected ‘char *’ but argument is of type ‘const char *’ 40905 winprefsyacc.y:175:5: warning: passing argument 3 of ‘AddMenuLine’ discards ‘const’ qualifier from pointer target type [enabled by default] 40906 winprefsyacc.y:67:13: note: expected ‘char *’ but argument is of type ‘const char *’ 40907 winprefsyacc.y:178:5: warning: passing argument 3 of ‘AddMenuLine’ discards ‘const’ qualifier from pointer target type [enabled by default] 40908 winprefsyacc.y:67:13: note: expected ‘char *’ but argument is of type ‘const char *’ 40909 winprefsyacc.c:1737:9: warning: passing argument 1 of ‘yyerror’ discards ‘const’ qualifier from pointer target type [enabled by default] 40910 winprefsyacc.y:82:12: note: expected ‘char *’ but argument is of type ‘const char *’ 40911 winprefsyacc.c:1854:3: warning: passing argument 1 of ‘yyerror’ discards ‘const’ qualifier from pointer target type [enabled by default] 40912 40913 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40914 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40915 40916commit 394ad259596801860d32d531408a07b63e469427 40917Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40918Date: Thu Oct 2 14:31:35 2014 +0100 40919 40920 hw/xwin: Fix const warning in winCheckDisplayNumber() 40921 40922 InitOutput.c:1032:19: warning: assignment discards ‘const’ qualifier from pointer target type [enabled by default] 40923 40924 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40925 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40926 40927commit 701492e5d9487545afeb5ebfcebccc0408ca7bc4 40928Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40929Date: Thu Oct 2 14:30:56 2014 +0100 40930 40931 hw/xwin: Fix warning in winXIconToHICON() 40932 40933 winmultiwindowicons.c:403:29: warning: passing argument 8 of ‘XGetWindowProperty’ from incompatible pointer type [enabled by default] 40934 /usr/include/X11/Xlib.h:2688:12: note: expected ‘Atom *’ but argument is of type ‘long unsigned int *’ 40935 40936 Looks like this has been wrong since I added it in 527cf131 :-( 40937 40938 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40939 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40940 40941commit a14f1d94d5bb87055f7e4812e9ef75771ea65bc5 40942Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40943Date: Thu Oct 2 14:09:40 2014 +0100 40944 40945 hw/xwin: Fix format warnings with debug printing of pointers on 64-bit 40946 40947 Fix various pieces of debug output, mainly under --enable-debug, which use a 40948 "%08x" printf format for a pointer type. Use "%p" format for 64-bit portability. 40949 40950 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40951 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40952 40953commit 03d462ddd2cacce92dc9e6bad00710d899509387 40954Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40955Date: Thu Oct 2 12:04:20 2014 +0100 40956 40957 hw/xwin: Fix warnings in glx/glshim.c 40958 40959 glext.h currently requires GL_GLEXT_PROTOTYPES in order to prototype 40960 glCompressedTexImmage* functions 40961 40962 generated_gl_shim.c:2859:6: warning: no previous prototype for 'glCompressedTexImage3DARB' [-Wmissing-prototypes] 40963 generated_gl_shim.c:2866:6: warning: no previous prototype for 'glCompressedTexImage2DARB' [-Wmissing-prototypes] 40964 generated_gl_shim.c:2873:6: warning: no previous prototype for 'glCompressedTexImage1DARB' [-Wmissing-prototypes] 40965 generated_gl_shim.c:2880:6: warning: no previous prototype for 'glCompressedTexSubImage3DARB' [-Wmissing-prototypes] 40966 generated_gl_shim.c:2887:6: warning: no previous prototype for 'glCompressedTexSubImage2DARB' [-Wmissing-prototypes] 40967 generated_gl_shim.c:2894:6: warning: no previous prototype for 'glCompressedTexSubImage1DARB' [-Wmissing-prototypes] 40968 generated_gl_shim.c:2901:6: warning: no previous prototype for 'glGetCompressedTexImageARB' [-Wmissing-prototypes] 40969 40970 Also, explicitly prototype glXGetProcAddressARB(), as glx/glxdri*.c does, as 40971 it's not practical to include glx.h here... 40972 40973 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40974 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40975 40976commit 329e8125aa3b4c83121f8290a5436c2cb4c1cf96 40977Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40978Date: Wed Oct 1 12:14:43 2014 +0100 40979 40980 hw/xwin: Fix compilation with -Werror=declaration-after-statement 40981 40982 xevents.c: In function 'winClipboardInitMonitoredSelections': 40983 xevents.c:129:5: error: 'for' loop initial declarations are only allowed in C99 mode 40984 for (int i = 0; i < CLIP_NUM_SELECTIONS; ++i) 40985 40986 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 40987 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 40988 40989commit 97c3298caab7c5a5396fcbde02d862e5380d7d2e 40990Author: Jon TURNEY <jon.turney@dronecode.org.uk> 40991Date: Wed Oct 1 12:02:04 2014 +0100 40992 40993 hw/xwin: Fix compilation with -Werror=return-type 40994 40995 winshadddnl.c: In function ‘winRedrawScreenShadowDDNL’: 40996 winshadddnl.c:991:9: error: ‘return’ with no value, in function returning non-void [-Werror=return-type] 40997 40998 Just wrong in 1c34e774 40999 41000 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 41001 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 41002 41003commit 61a292adf45405641de1c522a04c148e0a152acd 41004Author: Keith Packard <keithp@keithp.com> 41005Date: Thu Oct 9 15:17:17 2014 +0200 41006 41007 glx: check return from __glXGetAnswerBuffer 41008 41009 This function can return NULL; make sure every caller tests for that. 41010 41011 Reviewed-by: Adam Jackson <ajax@redhat.com> 41012 Signed-off-by: Keith Packard <keithp@keithp.com> 41013 41014commit d634ecdf82f244ff8ce75d351fc175792d254e5c 41015Merge: 6622f0cb1 7ebf480f5 41016Author: Keith Packard <keithp@keithp.com> 41017Date: Thu Oct 9 15:08:31 2014 +0200 41018 41019 Merge remote-tracking branch 'ajax/dead-code' 41020 41021commit 6622f0cb1761501f3287ffa8d3cbca3fd30f9f0c 41022Merge: f12e7f498 5ecd7866f 41023Author: Keith Packard <keithp@keithp.com> 41024Date: Thu Oct 9 15:05:26 2014 +0200 41025 41026 Merge remote-tracking branch 'ajax/mi-cleanup' 41027 41028commit f12e7f4980416b74988052a04d0f2487b2c4a7e9 41029Merge: da887726e aaf5e2d64 41030Author: Keith Packard <keithp@keithp.com> 41031Date: Thu Oct 9 14:21:19 2014 +0200 41032 41033 Merge remote-tracking branch 'anholt/modesetting-glamor' 41034 41035commit da887726eeaece0396962dd3912578d3372881b1 41036Author: Alan Coopersmith <alan.coopersmith@oracle.com> 41037Date: Sun Sep 21 09:49:06 2014 -0700 41038 41039 Use unique display name for each xi2 test program [v3] 41040 41041 make -j 8 check was sporadically failing in different xi2 tests. 41042 After adding the asserts in the previous commit to catch xkb failure 41043 it became easier to catch the failures and see that multiple tests 41044 were running at once trying to write to /tmp/server-(null).xkm and 41045 then delete it, and interfering with each other. 41046 41047 Putting a unique string into the display variable let them each write 41048 to their own file and not interfere with others. 41049 41050 v2: Fix Linux bits: 41051 41052 Add #include <errno.h> to get a declaration of 41053 program_invocation_name on Linux. 41054 41055 Use only the last portion of the pathname so that the resulting 41056 display name doesn't contain any slashes. 41057 41058 v3: use program_invocation_short_name on Linux 41059 41060 This is the same as program_invocation_name, except is has 41061 stripped off any path prefix. 41062 41063 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 41064 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 41065 Reviewed-by: Adam Jackson <ajax@redhat.com> 41066 Signed-off-by: Keith Packard <keithp@keithp.com> 41067 41068commit 5ecd7866f7587d620f000e802a262c7cd4b9df33 41069Author: Adam Jackson <ajax@redhat.com> 41070Date: Thu Sep 11 10:11:39 2014 -0400 41071 41072 misc: Fold together some redundant conditionals 41073 41074 Reviewed-by: Julien Cristau <jcristau@debian.org> 41075 Signed-off-by: Adam Jackson <ajax@redhat.com> 41076 41077commit 5b07f1db6de9f08f757206454da03ee0ae872bfc 41078Author: Adam Jackson <ajax@redhat.com> 41079Date: Thu Sep 18 16:05:40 2014 -0400 41080 41081 fb: fb{Map,Unmap}Window -> fb{Realize,Unrealize}Window 41082 41083 Make the function names match the screen slot name. 41084 41085 Reviewed-by: Julien Cristau <jcristau@debian.org> 41086 Signed-off-by: Adam Jackson <ajax@redhat.com> 41087 41088commit 9ed83a694b6492c9db226b115ed282cb127e798e 41089Author: Adam Jackson <ajax@redhat.com> 41090Date: Thu Sep 18 15:36:57 2014 -0400 41091 41092 mi: miSlideAndSizeWindow -> miResizeWindow 41093 41094 Make the function name match the screen slot name. 41095 41096 Reviewed-by: Julien Cristau <jcristau@debian.org> 41097 Signed-off-by: Adam Jackson <ajax@redhat.com> 41098 41099commit 939ca767c7603923a91fd37901175a42e802d40e 41100Author: Adam Jackson <ajax@redhat.com> 41101Date: Thu Sep 18 14:50:28 2014 -0400 41102 41103 dix: Remove an obfuscatory macro 41104 41105 Reviewed-by: Julien Cristau <jcristau@debian.org> 41106 Signed-off-by: Adam Jackson <ajax@redhat.com> 41107 41108commit d138d9ccc62a9472925574089f346bf9aa3dab1f 41109Author: Adam Jackson <ajax@redhat.com> 41110Date: Mon Sep 15 13:05:24 2014 -0400 41111 41112 dix: Remove some pointless casting of NULL 41113 41114 Reviewed-by: Julien Cristau <jcristau@debian.org> 41115 Signed-off-by: Adam Jackson <ajax@redhat.com> 41116 41117commit 277330075bfbd4343c4efdd136e0be716e1a491f 41118Author: Adam Jackson <ajax@redhat.com> 41119Date: Mon Sep 15 12:30:44 2014 -0400 41120 41121 dix: Remove an empty if 41122 41123 Reviewed-by: Julien Cristau <jcristau@debian.org> 41124 Signed-off-by: Adam Jackson <ajax@redhat.com> 41125 41126commit 81d76a835b2f647e3051b1d93606e59db7998d76 41127Author: Adam Jackson <ajax@redhat.com> 41128Date: Mon Sep 15 12:10:55 2014 -0400 41129 41130 dix: Lower backStorage to a bit instead of a pointer 41131 41132 Reviewed-by: Julien Cristau <jcristau@debian.org> 41133 Signed-off-by: Adam Jackson <ajax@redhat.com> 41134 41135commit 322ba42c23a3a107f7a62fb1c449792b616e5eba 41136Author: Adam Jackson <ajax@redhat.com> 41137Date: Mon Sep 15 12:05:26 2014 -0400 41138 41139 dix: Remove DIXsaveUnder bit from the Window 41140 41141 Reviewed-by: Julien Cristau <jcristau@debian.org> 41142 Signed-off-by: Adam Jackson <ajax@redhat.com> 41143 41144commit 1e56b2dfc6377234ffdcdf206528d476b04d13af 41145Author: Adam Jackson <ajax@redhat.com> 41146Date: Fri Sep 12 12:51:13 2014 -0400 41147 41148 mi: Move pScreen->SendGraphicsExpose up to dix 41149 41150 No DDX is overriding this and it's fairly absurd to expose it as a 41151 screen operation anyway. 41152 41153 Reviewed-by: Julien Cristau <jcristau@debian.org> 41154 Signed-off-by: Adam Jackson <ajax@redhat.com> 41155 41156commit 5d3bd8a3dc6456ea1ccf7b5f71b972379d7565ec 41157Author: Adam Jackson <ajax@redhat.com> 41158Date: Fri Sep 12 11:56:28 2014 -0400 41159 41160 mi: Drop plane argument from miHandleExposures 41161 41162 This existed to be passed to the bs recovery routine; since we back all 41163 planes, we don't care. 41164 41165 Reviewed-by: Julien Cristau <jcristau@debian.org> 41166 Signed-off-by: Adam Jackson <ajax@redhat.com> 41167 41168commit 7eddc80bc4a6c1902746a9fbe79ab553f9fd5821 41169Author: Adam Jackson <ajax@redhat.com> 41170Date: Tue Aug 5 14:07:17 2014 -0400 41171 41172 mi: Simplify composite border clip redirection interface 41173 41174 There's not really a good reason for mi to not just call the composite 41175 code directly. 41176 41177 Reviewed-by: Keith Packard <keithp@keithp.com> 41178 Signed-off-by: Adam Jackson <ajax@redhat.com> 41179 41180commit 7b35dc9b1533d4b99a3f84f460ac66354a6fe347 41181Author: Adam Jackson <ajax@redhat.com> 41182Date: Tue Sep 23 10:35:37 2014 -0400 41183 41184 mi: Deobfuscate miOverlayWindowExposures 41185 41186 Reviewed-by: Julien Cristau <jcristau@debian.org> 41187 Signed-off-by: Adam Jackson <ajax@redhat.com> 41188 41189commit fb34f273481b90a90a988f0f5c5a2d642a5661b8 41190Author: Adam Jackson <ajax@redhat.com> 41191Date: Tue Sep 23 10:30:56 2014 -0400 41192 41193 mi: Deobfuscate miWindowExposures 41194 41195 Reviewed-by: Julien Cristau <jcristau@debian.org> 41196 Signed-off-by: Adam Jackson <ajax@redhat.com> 41197 41198commit 0d30d44a8cdacfbc99d8193f76c133b803464622 41199Author: Adam Jackson <ajax@redhat.com> 41200Date: Thu Sep 11 12:44:54 2014 -0400 41201 41202 dix: Drop the third argument from WindowExposuresProcPtr 41203 41204 A careful read shows that it was always NULL. It hasn't always been; as 41205 the DDX spec indicates, it was the "occluded region that has backing 41206 store", but since that backing store code is long gone, we can nuke it. 41207 41208 mi{,Overlay}WindowExposures get slightly simpler here, and will get even 41209 simpler in just a moment. 41210 41211 Reviewed-by: Julien Cristau <jcristau@debian.org> 41212 Signed-off-by: Adam Jackson <ajax@redhat.com> 41213 41214commit aaf5e2d643af590366b2d02ad1c2db05d1fc4131 41215Author: Eric Anholt <eric@anholt.net> 41216Date: Sat Dec 28 11:32:10 2013 -0800 41217 41218 modesetting: Add support for rendering using glamor. 41219 41220 By default modesetting now tries to enable X acceleration using 41221 glamor, but falls back to normal shadowfb if GL fails to initialize. 41222 41223 Signed-off-by: Eric Anholt <eric@anholt.net> 41224 Reviewed-by: Keith Packard <keithp@keithp.com> 41225 41226commit 6d41bdb23ca92803e3a51dc317d2b80cdb92f43a 41227Author: Eric Anholt <eric@anholt.net> 41228Date: Tue Aug 26 16:59:37 2014 -0700 41229 41230 modesetting: Deduplicate some scrn setup. 41231 41232 Signed-off-by: Eric Anholt <eric@anholt.net> 41233 Reviewed-by: Keith Packard <keithp@keithp.com> 41234 41235commit 20a3b482191711bc15be0410ccc3651924271349 41236Author: Eric Anholt <eric@anholt.net> 41237Date: Wed Oct 8 00:41:13 2014 -0700 41238 41239 modesetting: Fix some weird formatting after x-indent-all.sh. 41240 41241 Signed-off-by: Eric Anholt <eric@anholt.net> 41242 Reviewed-by: Keith Packard <keithp@keithp.com> 41243 41244commit 3119acdab9eecff90b8fd9a265e2a456e7bc8779 41245Author: Eric Anholt <eric@anholt.net> 41246Date: Wed Oct 8 00:39:15 2014 -0700 41247 41248 modesetting: Run x-indent-all.sh. 41249 41250 As I was editing code, the top-level .dir-locals.el was making my new 41251 stuff conflict with the existing style. Make it consistently use the 41252 xorg style, instead. 41253 41254 Signed-off-by: Eric Anholt <eric@anholt.net> 41255 Reviewed-by: Keith Packard <keithp@keithp.com> 41256 41257commit c79f824bf6617816aaf10393beec8dddfa591f7b 41258Author: Adam Jackson <ajax@redhat.com> 41259Date: Thu Sep 25 15:49:26 2014 -0400 41260 41261 xwin: Remove primary DirectDraw engine 41262 41263 Again, as the documentation says, "unsupported, obsolete". 41264 41265 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 41266 Signed-off-by: Adam Jackson <ajax@redhat.com> 41267 41268commit 8465ee788fd541fa37681aa0a44103c7f944d437 41269Author: Adam Jackson <ajax@redhat.com> 41270Date: Thu Sep 25 15:37:33 2014 -0400 41271 41272 xwin: Remove native GDI engine (v2) 41273 41274 As the man page says, "unsupported, experimental, and barely 41275 functional". The last even minor updates to any of this were back in 41276 2004, presumably it's not getting better any time soon. 41277 41278 This is also the only GC ops implementation in the tree that actually 41279 falls all the way down to the spans routines for everything, so that's 41280 pretty nice to be rid of. 41281 41282 v2: Fix stray break statement (Jon) 41283 41284 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 41285 Signed-off-by: Adam Jackson <ajax@redhat.com> 41286 41287commit e2a3e91a56cb598f2c8257614d92ccd46b7fc094 41288Author: Adam Jackson <ajax@redhat.com> 41289Date: Fri Jul 25 11:01:22 2014 -0400 41290 41291 rootless: remove redundant RootlessShapedWindowIn 41292 41293 miShapedWindowIn is identical. Which you would expect, since neither 41294 one takes an actual WindowPtr argument, so they couldn't possibly make 41295 reference to rootlessness. 41296 41297 Reviewed-by: Julien Cristau <jcristau@debian.org> 41298 Signed-off-by: Adam Jackson <ajax@redhat.com> 41299 41300commit 7e8ac6dcd24d56e384ad09c7c23259cb3f4f8dc6 41301Author: Adam Jackson <ajax@redhat.com> 41302Date: Thu Sep 11 12:30:15 2014 -0400 41303 41304 xquartz: Remove useless DRIWindowExposures 41305 41306 Reviewed-by: Keith Packard <keithp@keithp.com> 41307 Signed-off-by: Adam Jackson <ajax@redhat.com> 41308 41309commit a8c6fc46c088bedb36c6441b78c98ca4cb9761b4 41310Author: Adam Jackson <ajax@redhat.com> 41311Date: Thu Sep 11 10:24:27 2014 -0400 41312 41313 kdrive: Remove vestigial reference to fbInitValidateTree 41314 41315 Reviewed-by: Keith Packard <keithp@keithp.com> 41316 Signed-off-by: Adam Jackson <ajax@redhat.com> 41317 41318commit 7ebf480f5eb75e06271ec924769e51bdcbbdc69c 41319Author: Adam Jackson <ajax@redhat.com> 41320Date: Fri Feb 28 12:34:49 2014 -0500 41321 41322 xfree86: Remove some can't-happen printf from xf86CreateRootWindow 41323 41324 Reviewed-by: Keith Packard <keithp@keithp.com> 41325 Signed-off-by: Adam Jackson <ajax@redhat.com> 41326 41327commit 3689be96ac125c9bfca9d37de4878f05f4e2f584 41328Author: Adam Jackson <ajax@redhat.com> 41329Date: Tue Jun 10 12:54:41 2014 -0400 41330 41331 vfb: Don't reimplement micmap so much 41332 41333 We know we're atop fb which is atop micmap, the only thing we need to 41334 hook is InstallColormap to handle the xwd colormap change. 41335 41336 Reviewed-by: Keith Packard <keithp@keithp.com> 41337 Signed-off-by: Adam Jackson <ajax@redhat.com> 41338 41339commit 692676debbc2227054d67078c6442f1ab1811192 41340Author: Adam Jackson <ajax@redhat.com> 41341Date: Mon Jun 16 12:35:28 2014 -0400 41342 41343 loader: Remove some baklava code 41344 41345 Reviewed-by: Keith Packard <keithp@keithp.com> 41346 Signed-off-by: Adam Jackson <ajax@redhat.com> 41347 41348commit c695a1430b80bac0b95e4d7f559e1bd223ecb1ef 41349Author: Adam Jackson <ajax@redhat.com> 41350Date: Tue Jul 8 14:38:34 2014 -0400 41351 41352 input: Remove unused SetKeySymsMap 41353 41354 Reviewed-by: Daniel Stone <daniels@collabora.com> 41355 Signed-off-by: Adam Jackson <ajax@redhat.com> 41356 41357commit 645063cd9384a26353775e92a79ed8710c6ca66c 41358Author: Adam Jackson <ajax@redhat.com> 41359Date: Tue Mar 4 12:51:49 2014 -0500 41360 41361 exa, kdrive: Remove redundant BitsPerPixel macros 41362 41363 We already get this from servermd.h 41364 41365 Reviewed-by: Keith Packard <keithp@keithp.com> 41366 Signed-off-by: Adam Jackson <ajax@redhat.com> 41367 41368commit cccba52d152a7d233c0b762a54844ecef831483e 41369Author: Alan Coopersmith <alan.coopersmith@oracle.com> 41370Date: Sun Sep 21 09:49:05 2014 -0700 41371 41372 Abort xi2 tests if ActivateDevice() fails 41373 41374 I was getting segfaults in xi2 tests from trying to copy XKB keyboard 41375 state to NULL pointers with a stack of: 41376 key=key@entry=0) at xkbActions.c:1189 41377 sendevent=sendevent@entry=0 '\000') at devices.c:420 41378 at protocol-xiquerydevice.c:338 41379 41380 which turned out to be due to xkbcomp failure, which was logged in the 41381 test logs as: 41382 XKB: Failed to compile keymap 41383 Keyboard initialization failed. This could be a missing or incorrect setup of xkeyboard-config. 41384 41385 but which was overlooked because the ActivateDevice() return code wasn't 41386 checked and the tests went forward assuming the structures were all 41387 correctly initialized. This catches the failure closer to the point of 41388 failure, to save debugging time. 41389 41390 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 41391 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 41392 Signed-off-by: Keith Packard <keithp@keithp.com> 41393 41394commit 95a5b92e37f73f497d547fd91c543c16d2cc73de 41395Author: Keith Packard <keithp@keithp.com> 41396Date: Thu Oct 2 10:26:53 2014 -0700 41397 41398 xfree86: Remove remaining return FALSE from configServerFlags 41399 41400 Remove the error return path from the FLAG_PIXMAP path and leave the 41401 default value in place. There's no point skipping the rest of this 41402 function. 41403 41404 Signed-off-by: Keith Packard <keithp@keithp.com> 41405 41406commit e9651e3dbc32650fcaf90888b4cbef81da77cfcd 41407Merge: 9bc01dfc7 4433bc6d5 41408Author: Keith Packard <keithp@keithp.com> 41409Date: Thu Oct 2 10:24:33 2014 -0700 41410 41411 Merge remote-tracking branch 'ajax/dead-code' 41412 41413commit 9bc01dfc7070a40f5948588895b3a11dd1636d0e 41414Author: Axel Davy <axel.davy@ens.fr> 41415Date: Sat Sep 27 23:17:13 2014 +0200 41416 41417 Fix present_pixmap when using present_notify_msc 41418 41419 Calling present_notify_msc could cancel a pending pixmap presentation. 41420 41421 Signed-off-by: Axel Davy <axel.davy@ens.fr> 41422 Reviewed-by: Keith Packard <keithp@keithp.com> 41423 Signed-off-by: Keith Packard <keithp@keithp.com> 41424 41425commit b3e9791fd375eb71b6150a77b8009b046b213004 41426Author: Daniel Martin <consume.noise@gmail.com> 41427Date: Sat Sep 20 09:07:41 2014 +0200 41428 41429 xfree86: Delete file modinit.h 41430 41431 All references to modinit.h have been remove with: 41432 41433 a1d41e3 Move extension initialisation prototypes into extinit.h 41434 41435 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 41436 Reviewed-by: Eric Anholt <eric@anholt.net> 41437 Signed-off-by: Keith Packard <keithp@keithp.com> 41438 41439commit 2aeb9d8f8b22ceb1de883c231a0a75d4c75d3f4a 41440Author: Daniel Martin <consume.noise@gmail.com> 41441Date: Sat Sep 20 09:07:40 2014 +0200 41442 41443 security: Remove XpExtension from SecurityTrustedExtensions 41444 41445 There's no XPrint extension (anymore). 41446 41447 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 41448 Reviewed-by: Keith Packard <keithp@keithp.com> 41449 Signed-off-by: Keith Packard <keithp@keithp.com> 41450 41451commit 12c19bbff4b7b2d0b350a7b93b38ebbf69748534 41452Author: Daniel Martin <consume.noise@gmail.com> 41453Date: Sat Sep 20 09:07:39 2014 +0200 41454 41455 dix: Remove XpExtension leftovers from protocol.txt 41456 41457 There's no XPrint extension (anymore). 41458 41459 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 41460 Reviewed-by: Keith Packard <keithp@keithp.com> 41461 Signed-off-by: Keith Packard <keithp@keithp.com> 41462 41463commit d3d845ca9e92f0a2ccde93f4242d7769cfe14164 41464Author: Michel Dänzer <michel.daenzer@amd.com> 41465Date: Thu Sep 25 15:27:22 2014 +0900 41466 41467 glamor: Use GL_STREAM_READ also for read/write access to a PBO 41468 41469 Otherwise the CPU may end up reading from non-cacheable memory, which is 41470 very slow. 41471 41472 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=84178 41473 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 41474 Reviewed-by: Markus Wick <markus@selfnet.de> 41475 Reviewed-by: Eric Anholt <eric@anholt.net> 41476 Signed-off-by: Keith Packard <keithp@keithp.com> 41477 41478commit 78c27d12e1f4857bfcfc92afbb69d5dd989232b9 41479Merge: 6e50bfa70 1d6334dd0 41480Author: Keith Packard <keithp@keithp.com> 41481Date: Mon Sep 29 13:04:49 2014 -0700 41482 41483 Merge remote-tracking branch 'jturney/xwin-clipboard-refactoring' 41484 41485commit 6e50bfa706cd3ab884c933bf1f17c221a6208aa4 41486Author: Michel Dänzer <michel.daenzer@amd.com> 41487Date: Wed Sep 24 16:48:10 2014 +0900 41488 41489 glamor: Fix leak of converted_bits in _glamor_upload_bits_to_pixmap_texture 41490 41491 ==9530== 808,575,600 bytes in 5,904 blocks are definitely lost in loss record 4,602 of 4,602 41492 ==9530== at 0x4C28C20: malloc (vg_replace_malloc.c:296) 41493 ==9530== by 0xAD29C98: _glamor_upload_bits_to_pixmap_texture (glamor_pixmap.c:771) 41494 ==9530== by 0xAD2AE95: glamor_upload_sub_pixmap_to_texture (glamor_pixmap.c:1031) 41495 ==9530== by 0xAD2BD55: glamor_upload_pixmap_to_texture (glamor_pixmap.c:1057) 41496 ==9530== by 0xAD1C2E6: glamor_composite_choose_shader (glamor_render.c:1025) 41497 ==9530== by 0xAD1C629: glamor_composite_with_shader (glamor_render.c:1174) 41498 ==9530== by 0xAD1DA77: glamor_composite_clipped_region (glamor_render.c:1542) 41499 ==9530== by 0xAD1E849: _glamor_composite (glamor_render.c:1689) 41500 ==9530== by 0xAD1ED90: glamor_composite (glamor_render.c:1758) 41501 ==9530== by 0x519FD6: damageComposite (damage.c:502) 41502 ==9530== by 0xAD27AA3: glamor_trapezoids (glamor_trapezoid.c:147) 41503 ==9530== by 0xAD27B51: glamor_trapezoids (glamor_trapezoid.c:101) 41504 41505 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=84176 41506 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 41507 Reviewed-by: Eric Anholt <eric@anholt.net> 41508 Signed-off-by: Keith Packard <keithp@keithp.com> 41509 41510commit 4433bc6d53e651b3dc4112099a153c2574f30797 41511Author: Adam Jackson <ajax@redhat.com> 41512Date: Thu May 1 12:48:15 2014 -0400 41513 41514 x86emu: Undefine _NO_INLINE 41515 41516 Never defined by the server. 41517 41518 Reviewed-by: Eric Anholt <eric@anholt.net> 41519 Signed-off-by: Adam Jackson <ajax@redhat.com> 41520 41521commit 1c3cb68d0cd15d29fd3aaa9eedbfa772ec8e3a47 41522Author: Adam Jackson <ajax@redhat.com> 41523Date: Tue Jul 29 14:15:24 2014 -0400 41524 41525 xfree86: Remove pointless xf86PrintMarkers 41526 41527 Reviewed-by: Eric Anholt <eric@anholt.net> 41528 Signed-off-by: Adam Jackson <ajax@redhat.com> 41529 41530commit 30fa6da6f140d713041942a81f5d6da6736e4f7f 41531Author: Adam Jackson <ajax@redhat.com> 41532Date: Tue Jul 29 14:07:39 2014 -0400 41533 41534 xfree86: configServerFlags never fails, make it return void 41535 41536 Reviewed-by: Eric Anholt <eric@anholt.net> 41537 Signed-off-by: Adam Jackson <ajax@redhat.com> 41538 41539commit 2bf9db1930528255a37f9787aeb82e2851f814c2 41540Author: Adam Jackson <ajax@redhat.com> 41541Date: Tue Jul 29 14:02:01 2014 -0400 41542 41543 xfree86: Remove unused xf86Info.useDefaultFontPathFrom 41544 41545 Reviewed-by: Eric Anholt <eric@anholt.net> 41546 Signed-off-by: Adam Jackson <ajax@redhat.com> 41547 41548commit fffea07b4cb0992fed2cca33be08440ea94090d9 41549Author: Adam Jackson <ajax@redhat.com> 41550Date: Tue Jul 29 13:48:10 2014 -0400 41551 41552 xfree86: Remove xf86Info.log (v2) 41553 41554 SUBCLASS ALL THE THINGS 41555 41556 v2: Remove the enum too (anholt) 41557 41558 Reviewed-by: Eric Anholt <eric@anholt.net> 41559 Signed-off-by: Adam Jackson <ajax@redhat.com> 41560 41561commit 2f5cfbee5409b088156d4cd84a3376499e291631 41562Author: Adam Jackson <ajax@redhat.com> 41563Date: Tue Jul 8 14:50:22 2014 -0400 41564 41565 xfree86: Remove DisplayID support 41566 41567 Not actually wired up so it's fairly useless. 41568 41569 Reviewed-by: Eric Anholt <eric@anholt.net> 41570 Signed-off-by: Adam Jackson <ajax@redhat.com> 41571 41572commit b56fcce761534d3e28b7ea19e2a13213c3157a1f 41573Author: Adam Jackson <ajax@redhat.com> 41574Date: Tue Jun 17 10:36:44 2014 -0400 41575 41576 xfree86: Remove useless back-pointer to pScrn from colormap code 41577 41578 Reviewed-by: Eric Anholt <eric@anholt.net> 41579 Signed-off-by: Adam Jackson <ajax@redhat.com> 41580 41581commit ef22655311efa0d5a5395ba2f19aa20d71a05a4f 41582Author: Adam Jackson <ajax@redhat.com> 41583Date: Fri May 2 13:36:13 2014 -0400 41584 41585 xfree86: Remove spurious xf86ConfigError 41586 41587 READABLE. 41588 41589 Reviewed-by: Eric Anholt <eric@anholt.net> 41590 Signed-off-by: Adam Jackson <ajax@redhat.com> 41591 41592commit 35f0257360fa6e11b9ecc08e467163718e012bb1 41593Author: Adam Jackson <ajax@redhat.com> 41594Date: Tue Mar 4 11:12:16 2014 -0500 41595 41596 xfree86: Remove Option "TextClockFreq" 41597 41598 No modern driver pays attention to this. Presumably there existed 41599 hardware once where you couldn't just read the right values out of the 41600 CRTC. 41601 41602 Reviewed-by: Eric Anholt <eric@anholt.net> 41603 Signed-off-by: Adam Jackson <ajax@redhat.com> 41604 41605commit 6755aa2c1244fee11846c1e0b274bd34900e3299 41606Author: Adam Jackson <ajax@redhat.com> 41607Date: Fri Jul 25 12:31:54 2014 -0400 41608 41609 xfree86: Remove xv clip notify driver hook 41610 41611 Nothing's using it. 41612 41613 Reviewed-by: Eric Anholt <eric@anholt.net> 41614 Signed-off-by: Adam Jackson <ajax@redhat.com> 41615 41616commit e89ccacea9a1e4bc3693bb78e717343addd517e5 41617Author: Adam Jackson <ajax@redhat.com> 41618Date: Thu Jan 23 10:27:58 2014 -0500 41619 41620 mi: Remove Windows 8-bpp colormap compatibility 41621 41622 Nice, but not something our Windows servers build, and not something 41623 that belongs in mi anyway. 41624 41625 Reviewed-by: Eric Anholt <eric@anholt.net> 41626 Signed-off-by: Adam Jackson <ajax@redhat.com> 41627 41628commit 7cd192edc195c089c6df32e9f22e57da7467b6fd 41629Author: Adam Jackson <ajax@redhat.com> 41630Date: Mon Jun 16 11:55:30 2014 -0400 41631 41632 loader: Deobfuscate RTLD_* macro stuff 41633 41634 POSIX requires that these be named correctly, no need to be clever. 41635 41636 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 41637 Signed-off-by: Adam Jackson <ajax@redhat.com> 41638 41639commit cc59be38b7eff52a1d003b390f2994c73ee0b3e9 41640Author: Keith Packard <keithp@keithp.com> 41641Date: Fri Sep 12 11:33:48 2014 -0700 41642 41643 os: Don't listen to 'tcp' by default. Add '-listen' option. [v2] 41644 41645 This disables the tcp listen socket by default. Then, it 41646 uses a new xtrans interface, TRANS(Listen), to provide a command line 41647 option to re-enable those if desired. 41648 41649 v2: Leave unix socket enabled by default. Add configure options. 41650 41651 Signed-off-by: Keith Packard <keithp@keithp.com> 41652 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 41653 41654commit 8ada3fb32cd7dd8948eb55620de18ba03df6131d 41655Author: Keith Packard <keithp@keithp.com> 41656Date: Sat Sep 20 10:04:11 2014 -0700 41657 41658 Require xtrans 1.3.5 or newer 41659 41660 This version of xtrans offers the TRANS(Listen) function. 41661 41662 Signed-off-by: Keith Packard <keithp@keithp.com> 41663 41664commit 18935b41e13b34571ffda84bde5257b8e84e6d1f 41665Author: Keith Packard <keithp@keithp.com> 41666Date: Sat Sep 20 04:04:43 2014 -0700 41667 41668 XSERVER_DTRACE needs request names from registry too 41669 41670 The dtrace code in the server wants to log the name of each executed 41671 request, which it gets from the registry. Use that as an additional 41672 indication of when that portion of the registry should be included in 41673 the server build. 41674 41675 See: 41676 41677 http://tinderbox.x.org/builds/2014-09-19-0003/logs/xserver/#build 41678 41679 Signed-off-by: Keith Packard <keithp@keithp.com> 41680 Reviewed-by: Dave Airlie <airlied@gmail.com> 41681 41682commit 28337cb14e4347e1dd7936c5393a22e042866687 41683Author: Jon TURNEY <jon.turney@dronecode.org.uk> 41684Date: Fri Sep 19 14:22:13 2014 +1000 41685 41686 xserver: Move 'pragma GCC diagnostic' outside functions 41687 41688 $ gcc --version 41689 gcc (Gentoo 4.4.3-r2 p1.2) 4.4.3 41690 41691 /jhbuild/checkout/xorg/xserver/os/log.c: In function ‘LogInit’: 41692 /jhbuild/checkout/xorg/xserver/os/log.c:199: error: #pragma GCC diagnostic not allowed inside functions 41693 /jhbuild/checkout/xorg/xserver/os/log.c:201: warning: format not a string literal, argument types not checked 41694 /jhbuild/checkout/xorg/xserver/os/log.c:212: error: #pragma GCC diagnostic not allowed inside functions 41695 /jhbuild/checkout/xorg/xserver/os/log.c:214: warning: format not a string literal, argument types not checked 41696 41697 etc. 41698 41699 Reviewed-by: Dave Airlie <airlied@redhat.com> 41700 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 41701 Signed-off-by: Dave Airlie <airlied@redhat.com> 41702 Signed-off-by: Keith Packard <keithp@keithp.com> 41703 41704commit 7e6bd546846964fd9b8c2a06dea4782a552b62d7 41705Author: Keith Packard <keithp@keithp.com> 41706Date: Wed Aug 6 12:58:38 2014 -0700 41707 41708 glamor: Remove shader-based trapezoid implementation. Fixes Bug 76213. 41709 41710 I can't find any performance benefit to using the GL path and the code 41711 renders this trapezoid incorrectly: 41712 41713 top: FIXED 29.50 41714 bottom: FIXED 30.00 41715 left top: POINT 0.00, 29.50 41716 left bottom: POINT 0.00, 30.50 41717 right top: POINT -127.50, 29.50 41718 right bottom: POINT 52.50, 30.00 41719 41720 This should render a solid line from 0,30 to 52,30 but draws nothing. 41721 41722 The code also uses an area computation for trapezoid coverage which 41723 does not conform to the Render specification which requires a specific 41724 point sampling technique. 41725 41726 Signed-off-by: Keith Packard <keithp@keithp.com> 41727 Reviewed-by: Adam Jackson <ajax@redhat.com> 41728 41729commit b2452311bd1d67b4d78612570d4a25c685c78a0c 41730Author: Keith Packard <keithp@keithp.com> 41731Date: Tue Aug 19 12:52:05 2014 -0700 41732 41733 glamor: Check large pixmap users in glamor_largepixmap.c 41734 41735 This enables the assertion that all users of the large pixmap member 41736 are restricted to pixmaps which are actually large. 41737 41738 Signed-off-by: Keith Packard <keithp@keithp.com> 41739 Reviewed-by: Adam Jackson <ajax@redhat.com> 41740 41741commit 5fc3e99f537b10dd9c2adfd37cf2c4ba5ed4bd44 41742Author: Keith Packard <keithp@keithp.com> 41743Date: Tue Aug 19 12:44:41 2014 -0700 41744 41745 glamor: Handle compositing from large to small pixmaps 41746 41747 glamor_composite_largepixmap_region is given the job of dealing with 41748 compositing between a mixture of large and small pixmaps. However, it 41749 was assuming that the destination pixmap was large and fetching 41750 members of the large structure even for small pixmaps. 41751 41752 This manifested with assertion failures when compositing from a large 41753 pixmap to a small pixmap. 41754 41755 Fixed by using the pixmap size for the destination block size for 41756 small pixmaps. 41757 41758 Signed-off-by: Keith Packard <keithp@keithp.com> 41759 Reviewed-by: Adam Jackson <ajax@redhat.com> 41760 41761commit 6e78d7f5e6edf56180e2ecfd25300bb2523876ab 41762Author: Keith Packard <keithp@keithp.com> 41763Date: Tue Aug 19 12:41:45 2014 -0700 41764 41765 glamor: Don't abuse large_pixmap members for regular pixmaps 41766 41767 glamor_compute_clipped_regions_ext wants to treat small and large 41768 pixmaps uniformly and did that by writing into the large pixmap 41769 union member in small pixmaps to construct something that looks like a 41770 one texture large pixmap. 41771 41772 Instead of doing that, simply allocate the necessary elements locally 41773 on the stack and use them from there. 41774 41775 Signed-off-by: Keith Packard <keithp@keithp.com> 41776 Reviewed-by: Adam Jackson <ajax@redhat.com> 41777 41778commit bab319763c9734da3cff7b91a582f54989dc4cbe 41779Author: Keith Packard <keithp@keithp.com> 41780Date: Tue Aug 19 12:39:07 2014 -0700 41781 41782 glamor: Add macro __glamor_large in glamor_largepixmap.c 41783 41784 For now, this simply fetches the large member of the pixmap private. 41785 It will be changed to assert that the pixmap is large once bugs 41786 related to that have been fixed. 41787 41788 Signed-off-by: Keith Packard <keithp@keithp.com> 41789 Reviewed-by: Adam Jackson <ajax@redhat.com> 41790 41791commit 16e429f1f96f984b1ff575875a7a4d7d29b482e2 41792Author: Keith Packard <keithp@keithp.com> 41793Date: Tue Aug 19 12:34:41 2014 -0700 41794 41795 glamor: Change SET_PIXMAP_FBO_CURRENT from macro to static inline 41796 41797 This is the last function-like macro in glamor_priv.h; change to 41798 static inline like all of the other functions there. 41799 41800 Signed-off-by: Keith Packard <keithp@keithp.com> 41801 Reviewed-by: Adam Jackson <ajax@redhat.com> 41802 41803commit ea7357de90792cbceca6d8d9fdc144e7df8a0fba 41804Author: Keith Packard <keithp@keithp.com> 41805Date: Wed Sep 10 14:01:59 2014 -0700 41806 41807 present: Clear pending flip pixmaps at CloseScreen 41808 41809 If a flip is active at server reset time, the associated window will 41810 get destroyed which will queue an unflip operation. If that isn't 41811 synchronous, then it won't have finished by the time CloseScreen is 41812 called. 41813 41814 Calling present_flip_idle will signal the fence and remove the 41815 reference to the fence and pixmap, freeing these in the X server and 41816 allowing a DRM client to clean up as well. 41817 41818 This also rewords other comments in present_flip_destroy, removing 41819 scary words about needing synchronous operation (everything in this 41820 function is synchronous now) and describing what effect we actually 41821 need from present_set_abort_flip. 41822 41823 Signed-off-by: Keith Packard <keithp@keithp.com> 41824 Reviewed-by: Adam Jackson <ajax@redhat.com> 41825 41826commit 2051514652481a83bd7cf22e57cb0fcd40333f33 41827Author: Keith Packard <keithp@keithp.com> 41828Date: Wed Sep 10 14:02:13 2014 -0700 41829 41830 present: Support PresentOptionCopy 41831 41832 We added this option to the present protocol before 1.0 but somehow 41833 never implemented it in the server. It's pretty simple; just don't 41834 ever do flips if the application specifies Copy. 41835 41836 Signed-off-by: Keith Packard <keithp@keithp.com> 41837 Reviewed-by: Adam Jackson <ajax@redhat.com> 41838 41839commit 69d8572ae4cd1bce17223ea8aff87916a974c861 41840Author: Keith Packard <keithp@keithp.com> 41841Date: Wed Sep 10 15:21:32 2014 -0700 41842 41843 Build required portions of registry.c automatically [v2] 41844 41845 Instead of making the inclusion of the registry code a global 41846 conditional, split the registry into two pieces; the bits required by 41847 the X-Resource extension (the resource names) and the bits required by 41848 the XCSECURITY extension (the protocol names). Build each set of code 41849 if the related extension is being built. 41850 41851 v2: Check for both XCSECURITY and XSELINUX. 41852 41853 Signed-off-by: Keith Packard <keithp@keithp.com> 41854 Reviewed-by: Adam Jackson <ajax@redhat.com> 41855 41856commit a11fc2493e85e4a532f4954805a7c6d1c601b08f 41857Author: Keith Packard <keithp@keithp.com> 41858Date: Wed Sep 10 15:04:00 2014 -0700 41859 41860 dix: Close protocol.txt after we're done loading extension names 41861 41862 Don't leave this file open during the whole server execution process; 41863 close it once all of the extensions are initialized. 41864 41865 Signed-off-by: Keith Packard <keithp@keithp.com> 41866 Reviewed-by: Adam Jackson <ajax@redhat.com> 41867 41868commit 9e07f3a3d28b01402beecb3280c94fc525bd8075 41869Author: Keith Packard <keithp@keithp.com> 41870Date: Wed Sep 10 15:41:08 2014 -0700 41871 41872 test: Only build hashtable tests if building X-Resource extension 41873 41874 The hash table functions are only included in the server when the 41875 X-Resource extension is built, so don't try to build and test them 41876 unless the X-Resource extension is being built. 41877 41878 Signed-off-by: Keith Packard <keithp@keithp.com> 41879 Reviewed-by: Adam Jackson <ajax@redhat.com> 41880 41881commit 670ee0757febfb965d718fce2d5688717def667d 41882Author: Adam Jackson <ajax@redhat.com> 41883Date: Tue Sep 16 11:10:58 2014 -0400 41884 41885 ephyr: Properly implement hardware cursors (v3) 41886 41887 When dix hands us a new cursor we proxy it through to the host server; 41888 since we keep the host XID on the cursor bits private we can switch 41889 among them with just ChangeWindowAttributes. 41890 41891 v2: 41892 Use xcb-renderutil for argb format lookup (Uli, Keith) 41893 Fall back to core cursors for host RENDER < 0.5 (Keith) 41894 Drop useless ephyrEnableCursor 41895 Consistently create/destroy the cursor image GC on both paths 41896 Treat null cursor from dix as invisible 41897 41898 v3: 41899 Initialize the invisible cursor's image (Keith) 41900 41901 Signed-off-by: Adam Jackson <ajax@redhat.com> 41902 Reviewed-by: Keith Packard <keithp@keithp.com> 41903 Signed-off-by: Keith Packard <keithp@keithp.com> 41904 41905commit f839caadd4e567bf89f893aa7e5ba1327b712b4b 41906Author: Adam Jackson <ajax@redhat.com> 41907Date: Wed Sep 17 11:07:22 2014 -0400 41908 41909 loader: Remove compatibility with ancient XFree86 version numbering 41910 41911 Pretty sure I'm guilty of adding this. I think I was thinking of trying 41912 to be compatible with some really old binary-only driver that I had 41913 vague aspirations of reverse-engineering, but since I haven't gotten 41914 around to it in the intervening decade... 41915 41916 Signed-off-by: Adam Jackson <ajax@redhat.com> 41917 Reviewed-by: Eric Anholt <eric@anholt.net> 41918 Signed-off-by: Keith Packard <keithp@keithp.com> 41919 41920commit af40913797e6595fb5466c2ff3110e9526a37b9f 41921Author: Julien Cristau <jcristau@debian.org> 41922Date: Wed Sep 17 07:41:27 2014 +0200 41923 41924 xwayland: always include drm.xml in tarballs 41925 41926 Move drm.xml out of the automake conditional so make dist includes it 41927 even if glamor-egl is disabled. 41928 41929 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=83960 41930 41931 Signed-off-by: Julien Cristau <jcristau@debian.org> 41932 Reviewed-by: Eric Anholt <eric@anholt.net> 41933 Signed-off-by: Keith Packard <keithp@keithp.com> 41934 41935commit eaee6572beefca240c42791f9a3a6e547bedd410 41936Merge: e3aa13b8d b84d25fd5 41937Author: Keith Packard <keithp@keithp.com> 41938Date: Wed Sep 17 15:49:24 2014 -0700 41939 41940 Merge remote-tracking branch 'anholt/modesetting-import' 41941 41942commit e3aa13b8d63ea2fba6eb47ddd0a7dde5a6d2b661 41943Author: Jon TURNEY <jon.turney@dronecode.org.uk> 41944Date: Wed Sep 17 15:38:39 2014 -0700 41945 41946 Update help text since -iglx is now the default 41947 41948 This should have been part of d0da0e9c3bb8fe0cd4879ecb24d21715bfaa209b 41949 41950 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 41951 Signed-off-by: Eric Anholt <eric@anholt.net> 41952 Reviewed-by: Eric Anholt <eric@anholt.net> 41953 Signed-off-by: Keith Packard <keithp@keithp.com> 41954 41955commit b84d25fd53e97105b99495af84abaacc4c901967 41956Author: Eric Anholt <eric@anholt.net> 41957Date: Tue Aug 26 11:22:02 2014 -0700 41958 41959 modesetting: Drop remaining old-xserver compatibility. 41960 41961 Signed-off-by: Eric Anholt <eric@anholt.net> 41962 Reviewed-by: Keith Packard <keithp@keithp.com> 41963 41964commit 7b0965a826612988dd721ddbd67cfbec392b67b5 41965Author: Eric Anholt <eric@anholt.net> 41966Date: Tue Aug 26 11:37:46 2014 -0700 41967 41968 modesetting: Drop use of compat-api.h 41969 41970 Signed-off-by: Eric Anholt <eric@anholt.net> 41971 Reviewed-by: Keith Packard <keithp@keithp.com> 41972 41973commit 2feda3b6b58f46cef91ea41312aac9021a703777 41974Author: Eric Anholt <eric@anholt.net> 41975Date: Tue Aug 26 11:06:23 2014 -0700 41976 41977 modesetting: Connect the driver to the build. 41978 41979 v2: Fix libdrm version check, and use XORG_VERSION_* instead of a 41980 static 1.0.0 version for the driver module. 41981 41982 Signed-off-by: Eric Anholt <eric@anholt.net> 41983 Reviewed-by: Keith Packard <keithp@keithp.com> 41984 41985commit 30dcc10f6c9c4bacbc0e3652917c298ca9fb3898 41986Author: Adam Jackson <ajax@redhat.com> 41987Date: Mon May 5 11:18:33 2014 -0400 41988 41989 xfree86: Clean up some more alpha awareness 41990 41991 Since the sparse stuff is gone none of these variables get used for 41992 anything, they're just dead side-effect-less execution. 41993 41994 Reviewed-by: Eric Anholt <eric@anholt.net> 41995 Signed-off-by: Adam Jackson <ajax@redhat.com> 41996 41997commit 2e21c081d34d52ee49ac00a294d2e74caf529818 41998Author: Adam Jackson <ajax@redhat.com> 41999Date: Fri May 2 13:00:21 2014 -0400 42000 42001 xfree86: Remove the memory of MTRR awareness 42002 42003 pciaccess does this for us, and none of our internal hooks really 42004 remain. This does remove a cleanup pass from the BSD code, but the case 42005 it's covering (a previous server leaving MTRRs around) can't happen 42006 anymore. 42007 42008 Reviewed-by: Eric Anholt <eric@anholt.net> 42009 Signed-off-by: Adam Jackson <ajax@redhat.com> 42010 42011commit c42712ab4d8eec56704ad5422d8f54befeb1e886 42012Author: Adam Jackson <ajax@redhat.com> 42013Date: Fri May 2 12:58:23 2014 -0400 42014 42015 xfree86: Remove xf86CheckMTRR 42016 42017 Die die die die die. 42018 42019 Reviewed-by: Eric Anholt <eric@anholt.net> 42020 Signed-off-by: Adam Jackson <ajax@redhat.com> 42021 42022commit 8686463de7e1ef9814824d46526d8177a7a492a0 42023Author: Adam Jackson <ajax@redhat.com> 42024Date: Fri May 2 12:55:37 2014 -0400 42025 42026 xfree86: Remove xf86ReadBIOS 42027 42028 Not being used externally or internally anymore. 42029 42030 Reviewed-by: Eric Anholt <eric@anholt.net> 42031 Signed-off-by: Adam Jackson <ajax@redhat.com> 42032 42033commit 21b216ad6ce2e9c89359b95e4196e42d91bf9420 42034Author: Adam Jackson <ajax@redhat.com> 42035Date: Fri May 2 12:46:45 2014 -0400 42036 42037 int10: Port off xf86ReadBIOS 42038 42039 If the linux vm86 backend changes look somewhat horrifying to you, 42040 that's because you have taste. 42041 42042 Reviewed-by: Eric Anholt <eric@anholt.net> 42043 Signed-off-by: Adam Jackson <ajax@redhat.com> 42044 42045commit 96c0da812a563d8229dc1396ac801329b3bcba91 42046Author: Adam Jackson <ajax@redhat.com> 42047Date: Fri May 2 12:39:01 2014 -0400 42048 42049 xfree86: Remove xf86LinearVidMem 42050 42051 The only driver even pretending to check the result is mach64 anyway. 42052 42053 Reviewed-by: Eric Anholt <eric@anholt.net> 42054 Signed-off-by: Adam Jackson <ajax@redhat.com> 42055 42056commit ec0e29ed5bc71516de7feb226b548ce45cb9ffdf 42057Author: Adam Jackson <ajax@redhat.com> 42058Date: Fri May 2 12:32:02 2014 -0400 42059 42060 xfree86: Remove {set,undo}WC from VidMemInfo 42061 42062 Now that mapMem is gone this can never actually get called. 42063 42064 Reviewed-by: Eric Anholt <eric@anholt.net> 42065 Signed-off-by: Adam Jackson <ajax@redhat.com> 42066 42067commit 9db2af6f757ef9680c70eb731698b7c9f9aca203 42068Author: Adam Jackson <ajax@redhat.com> 42069Date: Fri May 2 12:14:10 2014 -0400 42070 42071 xfree86: Remove xf86{Map,Unmap}VidMem 42072 42073 This API sucks. Fortunately it's mostly unused at this point. geode, 42074 sis, and xgi need minor patches to use the corresponding pciaccess code, 42075 neomagic will (more explicitly) lose its non-PCI support, and newport 42076 will need to be ported to /dev/mem or the platform bus or something. 42077 42078 This should also make it pretty clear that alpha's sparse memory support 42079 was basically not a thing anymore, very few tears shed. 42080 42081 Reviewed-by: Eric Anholt <eric@anholt.net> 42082 Signed-off-by: Adam Jackson <ajax@redhat.com> 42083 42084commit 6d6aa8ab6b6d62d65f1354b4cd9af0bf4f2e81d7 42085Author: Eric Anholt <eric@anholt.net> 42086Date: Tue Aug 26 11:43:54 2014 -0700 42087 42088 modesetting: Update for the platform device API change. 42089 42090 Signed-off-by: Eric Anholt <eric@anholt.net> 42091 Reviewed-by: Keith Packard <keithp@keithp.com> 42092 42093commit 9c080f44960d2338c7a59eee132a02d45186bfb0 42094Author: Eric Anholt <eric@anholt.net> 42095Date: Tue Aug 26 11:41:46 2014 -0700 42096 42097 modesetting: Drop use of "pointer" for "void *" 42098 42099 The xserver doesn't have that typedef. 42100 42101 Signed-off-by: Eric Anholt <eric@anholt.net> 42102 Reviewed-by: Keith Packard <keithp@keithp.com> 42103 42104commit a10ea01475c6d0795b1a00cfd8be9912f4b54235 42105Author: Eric Anholt <eric@anholt.net> 42106Date: Tue Aug 26 11:24:31 2014 -0700 42107 42108 modesetting: xserver already checks XEXTPROTO >= 7.2.99.901. 42109 42110 Signed-off-by: Eric Anholt <eric@anholt.net> 42111 Reviewed-by: Keith Packard <keithp@keithp.com> 42112 42113commit c7e8d4a6ee9542f56cd241cf7a960fb8223a6b22 42114Author: Eric Anholt <eric@anholt.net> 42115Date: Tue Aug 26 11:21:24 2014 -0700 42116 42117 modesetting: unifdef MODESETTING_OUTPUT_SLAVE_SUPPORT 42118 42119 The server will always have it. 42120 42121 v2: Clean up some weird formatting from the unifdeffing. 42122 42123 Signed-off-by: Eric Anholt <eric@anholt.net> 42124 Reviewed-by: Keith Packard <keithp@keithp.com> 42125 42126commit 0c07bfb38e61a8a6ad511dc8827e2dd20d68665d 42127Author: Eric Anholt <eric@anholt.net> 42128Date: Mon Aug 25 15:42:45 2014 -0700 42129 42130 modesetting: Merge modesetting's COPYING into the xserver's. 42131 42132 Signed-off-by: Eric Anholt <eric@anholt.net> 42133 Reviewed-by: Keith Packard <keithp@keithp.com> 42134 42135commit 1d6334dd0ca99923b6b55c5c047ef6b5c325f66c 42136Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42137Date: Wed Jun 19 13:09:35 2013 +0100 42138 42139 hw/xwin: Undefine _XSERVER64 in hw/xwin/winclipboard 42140 42141 Including any server header might define the macro _XSERVER64 on 64 bit 42142 machines. That macro must _NOT_ be defined for Xlib client code, otherwise bad 42143 things happen. So let's undef that macro if necessary. 42144 42145 Remove server directories from include path to ensure no server includes are 42146 included 42147 42148 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42149 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42150 42151commit 7ca93301940647219dbed2d14fcd3f37e49fb477 42152Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42153Date: Tue Jun 18 19:53:52 2013 +0100 42154 42155 hw/xwin: Only add to XSetIOErrorHandler() handler chain once 42156 42157 Only use XSetIOErrorHandler() to add to the global XSetIOErrorHandler() chain 42158 once. If we do it every restart, then we make a loop in the handler chain, and 42159 we end up with a thread spinning in that loop when the server shuts down... 42160 42161 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42162 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42163 42164commit 5819a02d461349dbfc86317c742406d2217f7f0c 42165Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42166Date: Tue Jun 18 19:35:20 2013 +0100 42167 42168 hw/xwin: Add a basic manpage for xwinclip 42169 42170 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42171 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42172 42173commit 492c924c8531b013d50637e29213da287b9a7fa2 42174Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42175Date: Tue Jun 18 19:28:53 2013 +0100 42176 42177 hw/xwin: In clipboard function prototypes, use Window type, not int type 42178 42179 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42180 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42181 42182commit 4a38c3bda2b0a868d7611f9423d8a0f304b435aa 42183Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42184Date: Tue Jun 18 19:18:08 2013 +0100 42185 42186 hw/xwin: Remove XOpenDisplay() retry code 42187 42188 Remove XOpenDisplay() retry code. This isn't a sensible thing for the 42189 application to be doing, and XWin server needs to retry much more than just 42190 XOpenDisplay(). 42191 42192 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42193 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42194 42195commit 4fc62aef8079266f8878cb355e6eb894000d8b8b 42196Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42197Date: Wed Jun 19 14:00:28 2013 +0100 42198 42199 hw/xwin: Improve comment about why we can't XCloseDisplay() 42200 42201 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42202 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42203 42204commit 89eb7bdcad092821b61a48832e1db82975a2252b 42205Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42206Date: Wed Jun 19 13:59:40 2013 +0100 42207 42208 hw/xwin: Remove setjmp()/longjmp() error constants 42209 42210 Check specially that setjmp() returned a value which we don't pass to longjmp() 42211 seems a bit over-complex. 42212 42213 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42214 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42215 42216commit d0353bbe57aab910a6ef7da8778023a9d39c37df 42217Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42218Date: Thu Jun 20 00:04:16 2013 +0100 42219 42220 hw/xwin: Move WIN_MSG_QUEUE_FNAME to where it's used 42221 42222 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42223 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42224 42225commit fa0da5a020c2ea9366568375b7f35d3f2ddad680 42226Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42227Date: Thu Jun 20 13:04:35 2013 +0100 42228 42229 hw/xwin: Use header for prototypes in textconv.c 42230 42231 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42232 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42233 42234commit cbe133752d1e817246b532a274fda3f7d74ebca1 42235Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42236Date: Thu Jun 20 12:50:18 2013 +0100 42237 42238 hw/xwin: Remove unused X includes from internal.h and add them where needed 42239 42240 Also removing server headers we might clash with and no longer need. Make 42241 a few adjustments to allow for this change: 42242 42243 - provide a prototype of ErrorF() 42244 - use the MAX() macro provided by sys/param.h, not the max() macro provided by misc.h 42245 - use the X 'Bool' type rather than the unwrapped Windows 'BOOL' type 42246 42247 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42248 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42249 42250commit 52f3cf3c612b2f33e39fd6f985b840a0d7192a50 42251Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42252Date: Wed Jun 19 18:09:07 2013 +0100 42253 42254 hw/xwin: Remove standard includes from internal.h and add them where needed 42255 42256 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42257 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42258 42259commit 5992550c4cf8818d1c47c1e474c3d8c3383983fa 42260Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42261Date: Mon Jun 17 19:01:19 2013 +0100 42262 42263 hw/xwin: Add xwinclip test client 42264 42265 Add xwinclip test client, which includes stubs for winDebug(), ErrorF() 42266 42267 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42268 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42269 42270commit 5ae674ae6d6c0b02fd86c5e7b23b8c38b19f180a 42271Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42272Date: Mon Jun 17 20:03:04 2013 +0100 42273 42274 hw/xwin: Changes in error logging to make libwinclipboard more useful as a library 42275 42276 Consistently use ErrorF() rather than winErrorFVerb() 42277 42278 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42279 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42280 42281commit e1cf2b47bea9ff3e64c0ef1e4d0744d9df4bc1b7 42282Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42283Date: Mon Jun 17 19:58:20 2013 +0100 42284 42285 hw/xwin: Look up atoms in winClipboardProc() 42286 42287 Look up all atoms of interest in clipboard code in winClipboardProc() and pass 42288 them down. 42289 42290 This avoids the need to check serverGeneration to notice when we need to 42291 invalidate cached atom values. 42292 42293 Also consistently use cached atom values everywhere, rather than sometimes just 42294 doing XInternAtom() again. 42295 42296 Remove WIN_LOCAL_PROPERTY as unused now, as we only refer to CYGX_CUT_BUFFER 42297 once and do that directly. 42298 42299 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42300 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42301 42302commit be61a1fc7e25a380f8d506e13b22017492a71913 42303Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42304Date: Mon Jun 17 19:07:05 2013 +0100 42305 42306 hw/xwin: Add fUseUnicode as parameter to winClipboardProc() 42307 42308 Add fUseUnicode as parameter to winClipboardProc() 42309 Access g_fUseUnicode global when calling it 42310 42311 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42312 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42313 42314commit 126c1cfaa5b5978026d44a1f8b044d367056965b 42315Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42316Date: Mon Jun 17 18:27:56 2013 +0100 42317 42318 hw/xwin: Separate libwinclipboard interface and internal header files 42319 42320 Rename the libwinclipboard internal header from winclipboard.h to internal.h 42321 42322 Put libwinclipboard's public interface into winclipboard.h 42323 42324 This lets winclipboardinit.c partake of that public interface, and all X server 42325 headers without clashes 42326 42327 winInitClipboard() prototype belongs in a server header 42328 42329 v2: Remove duplicate declaration of winClipboardWindowDestroy() 42330 42331 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42332 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42333 42334commit 30c535219ef2c251d90a1a1d3d07c8a7ec936da9 42335Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42336Date: Tue Jun 18 00:32:53 2013 +0100 42337 42338 hw/xwin: Return a shutdown flag from winClipboardProc() if we should stop trying 42339 42340 Return a shutdown flag from winClipboardProc(), and use it in 42341 winClipboardThreadProc() to determine if we should stop. 42342 42343 Currently this is set if the clipboard messaging window received a WM_QUIT. 42344 42345 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42346 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42347 42348commit 88d44597046086953032fec7e035fc036596e255 42349Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42350Date: Mon Jun 17 00:38:47 2013 +0100 42351 42352 hw/xwin: Move clipboard integration code down to a subdirectory 42353 42354 Move clipboard integration code down to a subdirectory and build as a 42355 convenience library 42356 42357 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42358 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42359 42360commit e965001a732a6b573c374f6a3503e172df01e8ec 42361Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42362Date: Mon Jun 17 13:18:16 2013 +0100 42363 42364 hw/xwin: Make g_hwndClipboard static 42365 42366 Move winFixClipboardChain() into winclipboardthread.c 42367 Add winCLipboardWindowDestroy() function to access it for WM_DESTROY 42368 42369 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42370 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42371 42372commit 03a59e7f787a160c6cc07c7a37df64a793242ef2 42373Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42374Date: Mon Jun 17 02:01:39 2013 +0100 42375 42376 hw/xwin: Add remaining clipboard globals to winglobals.h 42377 42378 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42379 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42380 42381commit 8f9fba5bc1499804a6c4a3287d99fe7dab8d6b8f 42382Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42383Date: Mon Jun 17 02:01:27 2013 +0100 42384 42385 hw/xwin: Hoist use of winSetAuthorization() and winGetDisplayName() up one level 42386 42387 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42388 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42389 42390commit a3c1e405cb78df9ef18a1158dcaed3c27b3cbffe 42391Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42392Date: Sun Jun 16 22:35:22 2013 +0100 42393 42394 hw/xwin: Eliminate g_pClipboardDisplay and g_iClipboardWindow globals 42395 42396 Eliminate the g_pClipboardDisplay and g_iClipboardWindow globals used to make 42397 those values available to the clipboard wndproc, by passing them in via the 42398 WM_CREATE message instead. 42399 42400 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42401 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42402 42403commit 229a0a83a44c94ce76eb937d58dc1773a38baa3e 42404Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42405Date: Sun Jun 16 22:13:26 2013 +0100 42406 42407 hw/xwin: Move winClipboardCreateMessagingWindow() to winclipboardthread.c 42408 42409 Move winClipboardCreateMessagingWindow() from winclipboardinit.c to 42410 winclipboardthread.c, the only place that uses it, and make it static. 42411 42412 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42413 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42414 42415commit a70c2384a2689cc0346868bb27366c008c01758d 42416Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42417Date: Mon Feb 13 19:58:37 2012 +0000 42418 42419 hw/xwin: Remove SetSelectionOwner wrapper, use XFixesSetSelectionOwnerNotify event instead 42420 42421 Use the XFixesSetSelectionNotify event instead of a SetSelectionOwner wrapper, 42422 the completely equivalent client-side mechanism. 42423 42424 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42425 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42426 42427commit ab55746c1b179172a82b7a89ec6429b268a1f166 42428Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42429Date: Mon Jun 17 01:55:46 2013 +0100 42430 42431 hw/xwin: Remove g_fClipboardLaunched, it's value is identical to g_fClipboardStarted 42432 42433 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42434 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42435 42436commit 42d13810822f6b48771d77b09c0ef7cc2356ede5 42437Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42438Date: Mon Jun 17 01:30:09 2013 +0100 42439 42440 hw/xwin: Hoist setting of g_fClipboardStarted flag up one level 42441 42442 Hoist the setting of g_fClipboardStarted flag up one level. 42443 42444 Also move up the clearing of the g_fClipboardLaunched at the end of clipboard 42445 function. 42446 42447 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42448 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42449 42450commit 290dbff0753ddf60f04da92d5a408c567a2ee3e0 42451Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42452Date: Mon Jun 17 01:55:03 2013 +0100 42453 42454 hw/xwin: winProcEstablishConnection doesn't need to check if clipboard started 42455 42456 winProcEstablishConnection doesn't need to check if clipboard has already been 42457 started. 42458 42459 It should be clear that we start the thread only once when the wrapper tells us 42460 to, as the wrapper unhooks itself thereafter. 42461 42462 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42463 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42464 42465commit 91e55691ef07735bb866c0f4d3c1a6e2ca167992 42466Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42467Date: Sun Jun 16 23:57:17 2013 +0100 42468 42469 hw/xwin: Hoist clipboard thread restart up one level 42470 42471 Hoist clipboard thread restart up one level. 42472 42473 Note that currently g_fClipboardLaunched is set the first time in the 42474 winProcEstablishConnection wrapper, and subsequent times when the clipboard 42475 thread restarts itself. 42476 42477 Try to clarify this and just set g_fClipboardLaunched before starting the 42478 thread. 42479 42480 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42481 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42482 42483commit 0bb6eae4e34634abb2679004ace94504a4d65964 42484Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42485Date: Mon Jun 17 00:07:01 2013 +0100 42486 42487 hw/xwin: Push winClipboardShutdown() into winclipboardinit.c 42488 42489 Push winClipboardShutdown() into winclipboardinit.c 42490 This lets us make g_ptClipboardProc static 42491 42492 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42493 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42494 42495commit 6343b531d786dd4a9bb52050c9ef267a04374b57 42496Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42497Date: Wed Jan 1 17:54:48 2014 +0000 42498 42499 hw/xwin: Remove left-over pthread_exit() in clipboard code 42500 42501 Commit c1bf3baa44fbd8af33a2b3ce045324485b85a7a7 removed all but one of the 42502 pthread_exit() calls which used to call winClipboardThreadExit() 42503 42504 Fix the final remaining one to exit via done label on IOError instead. 42505 42506 Also fix a comment and report pre-flush failure to log, but do not exit 42507 42508 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42509 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42510 42511commit d3427717f2c6a473dc3d20631dff653e4e37228e 42512Author: Adam Jackson <ajax@redhat.com> 42513Date: Wed Jul 2 12:09:52 2014 -0400 42514 42515 xwayland: Snap damage reports to the bounding box 42516 42517 Instead of sending every little rect. Lets x11perf run to completion, 42518 makes 'while true; do gtkperf -a; done' take longer to crash. 42519 42520 This is effectively a resend of the same logic against the old 42521 xfree86+xwayland branch: 42522 42523 http://lists.x.org/archives/xorg-devel/2013-October/038453.html 42524 42525 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 42526 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 42527 Signed-off-by: Adam Jackson <ajax@redhat.com> 42528 42529commit 5ecc0315a2b9bea44ace886f5706a7a3f5cbc4f6 42530Author: Jasper St. Pierre <jstpierre@mecheye.net> 42531Date: Mon Jun 30 13:53:50 2014 -0400 42532 42533 xwayland: Implement throttling for surfaces based on the frame event 42534 42535 This implements simple throttling that keeps us to one attach per 42536 frame. There isn't really an active performance benefit, since the 42537 buffers will be redrawn only once per frame anyway, but it does cut down 42538 on the chatty network traffic. Since the Wayland sockets might fill 42539 up as well, the cut down on the volume of data we send out also provides 42540 us with a big stability benefit. 42541 42542 Namely, mutter is a lot more stable running gtkperf, a fairly intensive 42543 X11 application, after this change. 42544 42545 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 42546 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 42547 42548commit 6c442fc4f7d84508665dfdee8a75ec91f50395dc 42549Author: Jasper St. Pierre <jstpierre@mecheye.net> 42550Date: Mon Jun 30 15:28:16 2014 -0400 42551 42552 xwayland-input: Fix a crasher for a race with the Wayland compositor 42553 42554 If something quickly maps and unmaps a window, then we'll immediately 42555 create and destroy the Wayland surface that cooresponds to that 42556 window. If our mouse pointer is over the window when the surface is 42557 created, we'll receive a enter on the window. 42558 42559 Since resource creation and destruction is not synchronous, that 42560 means that the compositor will queue up an event for a resource that's 42561 eventually destroyed. On the client-side, when we receive this message, 42562 we note that the resource isn't allocated, and get a NULL surface in our 42563 enter handler. We immediately try to dereference this, and then crash. 42564 42565 This was caused by running gtkperf while moving the window a lot. 42566 42567 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 42568 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 42569 42570commit 1d32004ed5a3a20b2d5879867e213122dced69d0 42571Author: Michel Dänzer <michel.daenzer@amd.com> 42572Date: Wed Sep 10 16:20:48 2014 +0900 42573 42574 glamor: Make glyph mask cache per-screen 42575 42576 It references a pixmap, which is a per-screen resource. 42577 42578 Fixes broken text rendering in xfwm4-tweak-settings in Zaphod mode. 42579 42580 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 42581 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 42582 Signed-off-by: Keith Packard <keithp@keithp.com> 42583 42584commit 710d314ad16cf1eaca3b97b1736ff0ca43e598c4 42585Author: Michel Dänzer <michel.daenzer@amd.com> 42586Date: Wed Sep 10 16:20:47 2014 +0900 42587 42588 glamor: Track glyph caching information per screen 42589 42590 This is necessary because the glyph caches are per screen. 42591 42592 Fixes broken menu text in gnome-terminal in Zaphod mode. 42593 42594 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 42595 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 42596 Signed-off-by: Keith Packard <keithp@keithp.com> 42597 42598commit a4e8e6dff9a0c26944a325e4017faa68d93ae5b8 42599Author: Michel Dänzer <michel.daenzer@amd.com> 42600Date: Wed Sep 10 16:20:46 2014 +0900 42601 42602 glamor: Move 'glyph caches realized' check to glamor_realize_glyph_caches 42603 42604 And rename the boolean to reflect what it's about. 42605 42606 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 42607 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 42608 Signed-off-by: Keith Packard <keithp@keithp.com> 42609 42610commit 0acff6e4374340dc391014f396f190daf73d579c 42611Author: Michel Dänzer <michel.daenzer@amd.com> 42612Date: Wed Sep 10 16:20:45 2014 +0900 42613 42614 glamor: Call glamor_glyphs_init from glamor_create_screen_resources 42615 42616 The comment above glamor_glyphs_init was already saying so. 42617 42618 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 42619 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 42620 Signed-off-by: Keith Packard <keithp@keithp.com> 42621 42622commit 54a1d4db825be90fb7be138e3267a15922d62f03 42623Merge: 8f8dcfee2 f92df22a0 42624Author: Keith Packard <keithp@keithp.com> 42625Date: Thu Sep 11 18:27:41 2014 -0700 42626 42627 Merge remote-tracking branch 'jturney/master' 42628 42629commit 8f8dcfee2085ba82107a8bf3872a0bb241493409 42630Author: Alan Coopersmith <alan.coopersmith@oracle.com> 42631Date: Thu Sep 4 23:18:03 2014 -0700 42632 42633 Make list tests actually test lists 42634 42635 Coverity scan detected that asserts were setting values, not checking them: 42636 42637 CID 53252: Side effect in assertion (ASSERT_SIDE_EFFECT) 42638 assignment_where_comparison_intended: Assignment item->b = i * 2 42639 has a side effect. This code will work differently in a non-debug build. 42640 Did you intend to use a comparison ("==") instead? 42641 42642 CID 53259: Side effect in assertion (ASSERT_SIDE_EFFECT) 42643 assignment_where_comparison_intended: Assignment item->a = i 42644 has a side effect. This code will work differently in a non-debug build. 42645 Did you intend to use a comparison ("==") instead? 42646 42647 CID 53260: Side effect in assertion (ASSERT_SIDE_EFFECT) 42648 assignment_where_comparison_intended: Assignment item->a = i 42649 has a side effect. This code will work differently in a non-debug build. 42650 Did you intend to use a comparison ("==") instead? 42651 42652 CID 53261: Side effect in assertion (ASSERT_SIDE_EFFECT) 42653 assignment_where_comparison_intended: Assignment item->b = i * 2 42654 has a side effect. This code will work differently in a non-debug build. 42655 Did you intend to use a comparison ("==") instead? 42656 42657 Fixing those to be == caused test_nt_list_insert to start failing as 42658 part assumed append order, part assumed insert order, so it had to be 42659 fixed to use consistent ordering. 42660 42661 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 42662 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 42663 Signed-off-by: Keith Packard <keithp@keithp.com> 42664 42665commit d0da0e9c3bb8fe0cd4879ecb24d21715bfaa209b 42666Author: Eric Anholt <eric@anholt.net> 42667Date: Thu Sep 4 10:59:40 2014 -0700 42668 42669 glx: Disable indirect GLX contexts by default. 42670 42671 Almost every situation of someone running indirect GLX is a mistake 42672 that results in X Server crashes. Indirect GLX is the cause of 42673 regular security vulnerabilities, and rarely provides any capability 42674 to the user. Just disable it unless someone wants to enable it for 42675 their special use case (using +iglx on the command line). 42676 42677 Signed-off-by: Eric Anholt <eric@anholt.net> 42678 Acked-by: Keith Packard <keithp@keithp.com> 42679 Signed-off-by: Keith Packard <keithp@keithp.com> 42680 42681commit bf338efc678258d2d366dff2ed873752f98f0bfc 42682Author: Keith Packard <keithp@keithp.com> 42683Date: Thu Sep 4 08:36:07 2014 -0700 42684 42685 glx/present: Only send GLX_BufferSwapComplete for PresentCompleteKindPixmap 42686 42687 Present didn't provide the 'kind' argument to the 42688 present_complete_notify hook that GLX uses to construct 42689 GLX_BufferSwapComplete events, so GLX was reporting events for 42690 PresentCompleteKindMSC notifications, which resulted in duplicate 42691 GLX_BufferSwapComplete events and crashes in clutter. 42692 42693 See the gnome bug: https://bugzilla.gnome.org/show_bug.cgi?id=733282 42694 42695 Signed-off-by: Keith Packard <keithp@keithp.com> 42696 Reviewed-by: Eric Anholt <eric@anholt.net> 42697 42698commit 7a0c79c8c478bf22ee71c1ea35886a6f555ef2bb 42699Author: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> 42700Date: Tue Sep 2 22:03:14 2014 +0200 42701 42702 os/log: adjust gcc version conditions for #pragma 42703 42704 In commit e67f2d7e0f9189beb2907fa06cff5ecc7f35f922 ("gcc 4.2.1 doesn't 42705 support #pragma GCC diagnostic ignored"), some compile time 42706 conditionals were added around the #pragma usage. Those conditionals 42707 ensure that the #pragma are not used on gcc <= 4.2. 42708 42709 However, the usage of #pragma diagnostic inside functions was only 42710 added in gcc 4.6, and a build failure is therefore experienced with 42711 gcc 4.5: 42712 42713 log.c: In function 'LogInit': 42714 log.c:199:9: error: #pragma GCC diagnostic not allowed inside functions 42715 log.c:201:9: warning: format not a string literal, argument types not checked 42716 log.c:212:9: error: #pragma GCC diagnostic not allowed inside functions 42717 log.c:214:17: warning: format not a string literal, argument types not checked 42718 42719 $ ./host/usr/bin/powerpc-linux-gnu-gcc -v 42720 [...] 42721 gcc version 4.5.2 (Sourcery G++ Lite 2011.03-38) 42722 42723 This patch therefore adjusts the compile time conditionals to make 42724 sure the #pragma is not used on gcc <= 4.5, and only used on gcc >= 42725 4.6. 42726 42727 Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> 42728 Acked-by: Keith Packard <keithp@keithp.com> 42729 Signed-off-by: Keith Packard <keithp@keithp.com> 42730 42731commit da4bad620a6f1a58978f5279fda74da3c1c1d443 42732Author: Dave Airlie <airlied@gmail.com> 42733Date: Tue Jul 30 13:50:48 2013 +1000 42734 42735 pixmap: fix reverse optimus support with multiple heads 42736 42737 For doing reverese optimus to multiple outputs on a secondary GPU 42738 the GPU can store the blits into a large screen pixmap, unfortunately 42739 this means we need a destination offset into the dirty code, so 42740 add a new API that just adds this interface. 42741 42742 Signed-off-by: Dave Airlie <airlied@redhat.com> 42743 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 42744 Signed-off-by: Keith Packard <keithp@keithp.com> 42745 42746commit e1cc0d3df1fdb6c4393ac455ca038e1e0680bb22 42747Author: Maks Naumov <maksqwe1@ukr.net> 42748Date: Thu Aug 21 12:58:16 2014 -0700 42749 42750 glx: Fix 'y ' value in swrastGetDrawableInfo() 42751 42752 Signed-off-by: Maks Naumov <maksqwe1@ukr.net> 42753 Reviewed-by: Eric Anholt <eric@anholt.net> 42754 Signed-off-by: Keith Packard <keithp@keithp.com> 42755 42756commit f92df22a037cb8b672dab2e7aef18a22275f5660 42757Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42758Date: Fri Jul 18 15:08:17 2014 +0100 42759 42760 hw/xwin/glx: Remove an incorrect assertion in glxWinDrawableSwapBuffers 42761 42762 The piglit test glx_make_current triggers this assertion, by making the context 42763 current on a different drawable before issuing a glXSwapBuffers() 42764 42765 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42766 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42767 42768commit b21321e515fc778f87077b7ce47922c43a4ba103 42769Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42770Date: Mon Jul 14 19:41:25 2014 +0100 42771 42772 hw/xwin/glx: Downgrade "forcing window to exist" message to debug 42773 42774 It happens whenever a GLX client uses GL on a window before it's been mapped, so 42775 don't log it like an error. 42776 42777 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42778 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42779 42780commit 31d089633683ea2a2716651c3c48e1456cc595bf 42781Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42782Date: Wed Jul 16 13:44:33 2014 +0100 42783 42784 hw/xwin/glx: Fix a couple of typo bugs in indirect.c 42785 42786 fbConfigToPixelFormatIndex()'s drawableTypeOverride parameter is a drawable 42787 type bitmask, not a drawable type enum value 42788 42789 WGL_SWAP_COPY_ARB is a value of the WGL_SWAP_METHOD_ARB attribute for 42790 wglChoosePixelFormatARB(), not an attribute itself 42791 42792 also remove duplicate error reporting for wglChoosePixelFormat() and fix a 42793 comment 42794 42795 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42796 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42797 42798commit b1ea714cdce8444c6ad7778dae04e35c6bcc1750 42799Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42800Date: Sun Apr 27 22:10:31 2014 +0100 42801 42802 hw/xwin/glx: Update .gitignore 42803 42804 Update to align with rewrite of wrapper generation script in commit 42805 583a1146233f16d861706926706e5feec3baffba 42806 42807 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42808 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42809 42810commit d1d3bd359670a8f55c9350fd905af32ba6e67d33 42811Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42812Date: Sun Mar 30 20:51:36 2014 +0100 42813 42814 hw/xwin/glx: Don't override the server supported GL extensions string. 42815 42816 Don't override the server supported GL extensions string. 42817 42818 The string reported to the client is the intersection of client, server and GL 42819 implementation extensions. 42820 42821 Overriding the server supported GL extensions string like this causes extensions 42822 which are supported by the client and implementation, but not by the server, to 42823 be erroneously reported, so don't change it. 42824 42825 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42826 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42827 42828commit c99cd058a89bbfc9ceab3c91b93451f70351ddce 42829Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42830Date: Tue Mar 4 22:37:13 2014 +0000 42831 42832 hw/xwin: Add an idempotency guard to glwindows.h 42833 42834 Add an idempotency guard to glwindows.h 42835 42836 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42837 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42838 42839commit 1c34e774eff6c5a22501833444a10440cf3f915a 42840Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42841Date: Mon Nov 11 15:09:18 2013 +0000 42842 42843 hw/xwin: Fix a potential crash in winRedrawScreenShadowDDNL() 42844 42845 Seen during shutdown when using '-fullscreen' and '-depth 8' 42846 42847 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42848 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42849 42850commit 16d9da08861fe504de4be8c31708592e30687156 42851Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42852Date: Thu Feb 18 23:36:27 2010 -0600 42853 42854 hw/xwin: Improve choice of display name used by internal clients 42855 42856 Choose the display name used to connect to internal clients and exported into 42857 environment of processes started from the traymenu so that it uses a transport 42858 we know is working 42859 42860 This should mean the server can start correctly with -multiwindow and/or 42861 -clipboard and any two of -nolisten inet6, -nolisten inet and -nolisten unix 42862 (the server will correctly refuse to start if all 3 are used, as it must be 42863 listening on at least one socket) 42864 42865 v2: 42866 Place prototype for winGetDisplayName() in windisplay.h, and include it where 42867 needed. 42868 42869 v3: 42870 Include xwin-config.h, so that _XSERVER64 is defined, just in case anything 42871 relies on that. 42872 42873 v4: 42874 Replace grovelling around in the server's list of listeners with new Xtrans 42875 TransIsListening() interface, added in Xtrans 1.3.3 42876 42877 See also [1] 42878 42879 [1] https://sourceware.org/bugzilla/show_bug.cgi?id=10725 42880 42881 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42882 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 42883 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42884 42885commit 95357539c1178a97d8a167461806a280d2803ec4 42886Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42887Date: Wed Jan 1 20:34:36 2014 +0000 42888 42889 hw/xwin: Makefile.am cleanup 42890 42891 - Rename XWIN_GLX_LINK_FLAGS -> XWIN_GLX_SYS_LIBS for consistency 42892 - Rename MULTIWINDOW_LIBS -> MULTIWINDOW_SYS_LIBS for consistency 42893 - Don't link with XWin with $(MAIN_LIB), it provides it's own main() 42894 - Put one library per line for more intelligible diffs when one is added 42895 42896 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42897 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42898 42899commit aa40d0c07173209a2c13b332f1168e0df499a19e 42900Author: Jon TURNEY <jon.turney@dronecode.org.uk> 42901Date: Fri Jun 13 16:16:51 2014 +0100 42902 42903 hw/xwin/glx: When generating shims, limit the considered features to GL version <=1.2 42904 42905 This fixes a problem when using Khronos registry data since the change of 42906 2013-08-16 removed glBlend(Color|Equation) from GL1.2 and added them to 42907 GL_ARB_imaging. 42908 42909 If shim generation considers all features, no shims are generated for 42910 glBlend(Color|Equation) as they are first emitted for GL 1.4 (which we ignore as 42911 shims are only generated for GL version <=1.2), then emission for GL_ARB_imaging 42912 is skipped as they have already been emitted. 42913 42914 Also improve feature name matching so it is exact, not on an initial substring, 42915 so 'GL_ARB_texture_compression_bptc' and 'GL_ARB_texture_compression_rgtc' 42916 aren't matched by 'GL_ARB_texture_compression'. 42917 42918 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 42919 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 42920 42921commit 35dc7c75150733dbcef8a18b6796f49a7c48ebee 42922Merge: 3a51418b2 5f3485b06 42923Author: Eric Anholt <eric@anholt.net> 42924Date: Mon Aug 25 15:52:58 2014 -0700 42925 42926 Merge branch 'modesetting-import' into master 42927 42928 This reintroduces a "hardware" driver to the xfree86 directory. 42929 Unlike the drivers that xorg used to include in the source tree, that 42930 needed independent release schedules to get hardware support out the 42931 door, the modesetting driver shouldn't change much as new hardware 42932 gets released. A lot of what this driver needs to do is just keep up 42933 with server ABI changes. 42934 42935 This import was done by taking xf86-video-modesetting-0.9.0, and 42936 running this script with 'git-filter-branch -f --tree-filter 42937 ~/bin/modesetting-filter': 42938 42939 mkdir -p hw/xfree86/drivers/modesetting 42940 42941 rm -f README autogen.sh configure.ac Makefile.am .gitignore 42942 rm -f man/Makefile.am 42943 42944 mv man/modesetting.man hw/xfree86/drivers/modesetting/ 42945 mv COPYING hw/xfree86/drivers/modesetting/ 42946 mv src/* hw/xfree86/drivers/modesetting/ 42947 42948commit 3a51418b2db353519a1779cf3cebbcc9afba2520 42949Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 42950Date: Mon Aug 18 08:45:43 2014 -0300 42951 42952 ephyr: set screen size & origin from host X server output's CRTC geometry 42953 42954 If a given output is passed via new -output option, Xephyr will query 42955 host X server for its info. If the following conditions are met: 42956 42957 a. RandR extension is enabled in host X server; 42958 b. supported RandR version in host X server is 1.2 or newer; 42959 c. the given output name is valid; 42960 d. the given output is connected; 42961 42962 then Xephyr will get output's CRTC geometry and use it to set its own 42963 screen size and origin. It's just like starting Xephyr in fullscreen mode, 42964 but restricted to the given output's CRTC geometry (fake "Zaphod mode"). 42965 42966 This is the main feature needed for Xephyr-based single-card multiseat 42967 setups where we don't have separate screens to start Xephyr in fullscreen 42968 mode safely. 42969 42970 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 42971 Reviewed-by: Keith Packard <keithp@keithp.com> 42972 Signed-off-by: Keith Packard <keithp@keithp.com> 42973 42974commit 84b02469ef97e6f85d074d220a517d752180045f 42975Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 42976Date: Mon Aug 18 08:45:42 2014 -0300 42977 42978 ephyr: enable screen window placement following kdrive -screen option extended syntax 42979 42980 With this patch, one can launch Xephyr with option "-screen WxH+X+Y" 42981 to place its window origin at (X,Y). This patch relies on a previous 42982 one that extends kdrive -screen option syntax to parse +X+Y substring 42983 as expected. 42984 42985 If +X+Y is not passed in -screen argument string, let the WM place 42986 the window for us, as before. 42987 42988 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 42989 Reviewed-by: Keith Packard <keithp@keithp.com> 42990 Signed-off-by: Keith Packard <keithp@keithp.com> 42991 42992commit 376f4de8ae927748417046390c24afbda24b0583 42993Author: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 42994Date: Mon Aug 18 08:45:41 2014 -0300 42995 42996 kdrive: add support to +X+Y syntax in -screen option parsing 42997 42998 This patch enhances current -screen option parsing for kdrive-based 42999 applications. It can parse strings like 43000 <WIDTH>x<HEIGHT>+<XOFFSET>+<YOFFSET>, storing X and Y offsets 43001 in KdScreenInfo instances. 43002 43003 For negative values, this patch supports +-X+-Y (not -X-Y) syntax. 43004 43005 It will allow e.g. proper Xephyr window placement for multiseat 43006 purposes. 43007 43008 Signed-off-by: Laércio de Sousa <laerciosousa@sme-mogidascruzes.sp.gov.br> 43009 Reviewed-by: Keith Packard <keithp@keithp.com> 43010 Signed-off-by: Keith Packard <keithp@keithp.com> 43011 43012commit fe5018e0564118a7a8198fa286186fdb9ed818c7 43013Author: Takashi Iwai <tiwai@suse.de> 43014Date: Tue Aug 19 15:57:22 2014 -0500 43015 43016 fb: Fix invalid bpp for 24bit depth window 43017 43018 We have a hack in fb layer for a 24bpp screen to use 32bpp images, and 43019 fbCreateWindow() replaces its drawable.bitsPerPixel field 43020 appropriately. But, the problem is that it always replaces when 32bpp 43021 is passed. If the depth is 32, this results in bpp < depth, which is 43022 actually invalid. 43023 43024 Meanwhile, fbCreatePixmap() has a more check and it creates with 24bpp 43025 only when the passed depth <= 24 for avoiding such a problem. 43026 43027 This oneliner patch just adds the similar check in fbCreateWindow(). 43028 This (hopefully) fixes the long-standing broken graphics mess of 43029 cirrus KMS with 24bpp. 43030 43031 Signed-off-by: Takashi Iwai <tiwai@suse.de> 43032 Reviewed-by: Keith Packard <keithp@keithp.com> 43033 43034commit 5d133276de9c50146e80ffc69edd429c2afe98e6 43035Author: Thierry Reding <treding@nvidia.com> 43036Date: Thu Feb 13 13:36:12 2014 +0100 43037 43038 xfree86: Allow non-PCI devices as primary 43039 43040 On platforms that don't support PCI or have no GPU attached to the PCI 43041 bus, there can still be a primary device on a non-PCI bus. 43042 43043 Signed-off-by: Thierry Reding <treding@nvidia.com> 43044 Reviewed-by: Rob Clark <robdclark@gmail.com> 43045 Tested-by: Rob Clark <robdclark@gmail.com> 43046 Signed-off-by: Keith Packard <keithp@keithp.com> 43047 43048commit 2f0183222b4279266e6ef60b923738ff55f0afba 43049Author: Thierry Reding <treding@nvidia.com> 43050Date: Thu Feb 13 13:31:31 2014 +0100 43051 43052 xfree86: Fallback to first platform device as primary 43053 43054 When neither of the various bus implementations was able to find a 43055 primary bus and device, fallback to using the platform bus as primary 43056 bus and the first platform device as primary device. 43057 43058 Signed-off-by: Thierry Reding <treding@nvidia.com> 43059 Reviewed-by: Rob Clark <robdclark@gmail.com> 43060 Tested-by: Rob Clark <robdclark@gmail.com> 43061 Signed-off-by: Keith Packard <keithp@keithp.com> 43062 43063commit 33aeec8a11bc75a06d32956bbdd9efb216790a02 43064Author: Keith Packard <keithp@keithp.com> 43065Date: Thu Jan 30 16:24:35 2014 -0800 43066 43067 dri3: Don't enable the DRI3 extension unless some screen supports it 43068 43069 There's no reason to advertise this extension unless one of the 43070 hardware drivers actually supports it. Not listing it means it's 43071 slightly easier for users to tell what's going on. 43072 43073 On the other hand, not listing it means we may have applications that 43074 only check for the extension and not for appropriate per-screen 43075 support. I don't think that's a real risk as DRI3 is only useful for 43076 systems with deep knowledge of the hardware. 43077 43078 Signed-off-by: Keith Packard <keithp@keithp.com> 43079 Reviewed-by: Eric Anholt <eric@anholt.net> 43080 43081commit 0d7306ffe2e069f0091c4f174f138127100abb43 43082Author: Michel Dänzer <michel.daenzer@amd.com> 43083Date: Wed Aug 13 17:10:21 2014 +0900 43084 43085 glamor: Don't call glamor_copy_fbo_cpu() for GLAMOR_DRM_ONLY pixmaps 43086 43087 It results in a crash. 43088 43089 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 43090 Reviewed-by: Eric Anholt <eric@anholt.net> 43091 Signed-off-by: Keith Packard <keithp@keithp.com> 43092 43093commit c2548d71f7a754251d89696ae49aab2fe07446df 43094Author: Joerg Sonnenberger <joerg@NetBSD.org> 43095Date: Mon Aug 11 10:36:01 2014 +0200 43096 43097 Fix format string. 43098 43099 Reviewed-by: Keith Packard <keithp@keithp.com> 43100 Signed-off-by: Thomas Klausner <wiz@NetBSD.org> 43101 Signed-off-by: Keith Packard <keithp@keithp.com> 43102 43103commit d4111239e698f0126123830429d74da7d24910ed 43104Author: Keith Packard <keithp@keithp.com> 43105Date: Wed Aug 13 11:30:11 2014 -0700 43106 43107 glamor: Free Xv put image data immediately after use 43108 43109 The Xv StopVideo callback is not invoked on textured video ports, so 43110 the temporary pixmaps allocated for the video planes are never freed. 43111 43112 Freeing the storage immediately after use is a simple solution to this 43113 problem which doesn't appear to have any visible performance impact. 43114 43115 Signed-off-by: Keith Packard <keithp@keithp.com> 43116 Reviewed-by: Eric Anholt <eric@anholt.net> 43117 43118commit c47ee880e742e15fdbd5f5ac3eba64db6228c09f 43119Author: Eric Anholt <eric@anholt.net> 43120Date: Mon Aug 11 15:23:02 2014 -0700 43121 43122 glamor: Don't try to set up core fonts textures when we won't use them. 43123 43124 This happens to avoid GL errors on hardware without 43125 EXT_texture_integer (which implies < GLSL 130, and thus glamor_text.c 43126 programs not compiling anyway). 43127 43128 Signed-off-by: Eric Anholt <eric@anholt.net> 43129 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 43130 Signed-off-by: Keith Packard <keithp@keithp.com> 43131 43132commit d09c9ddb866dea31e2adf82be329d5d3489718f5 43133Merge: 606b701c5 ab32ee358 43134Author: Keith Packard <keithp@keithp.com> 43135Date: Tue Aug 12 16:34:24 2014 -0700 43136 43137 Merge remote-tracking branch 'jeremyhu/master' 43138 43139commit 606b701c563c10f01dec147e061fa00cda784a9d 43140Author: Michel Dänzer <michel.daenzer@amd.com> 43141Date: Wed Aug 6 10:30:25 2014 +0900 43142 43143 glamor: Add glamor_copy_fbo_cpu() for CopyArea to non-GPU destination 43144 43145 This provides a speedup e.g. when the destination is an SHM pixmap. 43146 43147 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=76285 43148 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 43149 Reviewed-by: Eric Anholt <eric@anholt.net> 43150 Signed-off-by: Keith Packard <keithp@keithp.com> 43151 43152commit 1e30fc1b99bda040038e4fd56d1b27c686b44c75 43153Author: Peter Hutterer <peter.hutterer@who-t.net> 43154Date: Mon Aug 4 10:47:03 2014 +1000 43155 43156 xkb: ignore floating slave devices when updating from master (#81885) 43157 43158 Introduced in 45fb3a934dc0db51584aba37c2f9d73deff9191d. When a device is 43159 enabled, the master's locked state is pushed to the slave. If the device is 43160 floating, no master exists and we triggered a NULL-pointer dereference 43161 in XkbPushLockedStateToSlaves. 43162 43163 X.Org Bug 81885 <http://bugs.freedesktop.org/show_bug.cgi?id=81885> 43164 43165 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 43166 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 43167 Signed-off-by: Keith Packard <keithp@keithp.com> 43168 43169commit 4599a4492ca6a1c87cd9c537c54fbda11f13c15c 43170Author: Daphne Pfister <daphnediane@mac.com> 43171Date: Wed Jul 23 21:51:57 2014 -0400 43172 43173 os: Regenerate os/oscolor.c from app/rgb #52289 43174 43175 Regenerate os/oscolor.c from rgb.txt. This adds the following 43176 colors: aqua, lime, fuchsia, crimson, indigo, olive, rebecca 43177 purple, silver and teal. It also adds versions of gray, grey, 43178 green, maroon and purple prefixed with web and x11 for the 43179 colors that are different between X11 and HTML/CSS web colors. 43180 43181 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=52289 43182 Related: https://bugs.freedesktop.org/show_bug.cgi?id=80371 43183 43184 Signed-off-by: nobody 43185 Tested-by: Keith Packard <keithp@keithp.com> 43186 Signed-off-by: Keith Packard <keithp@keithp.com> 43187 43188commit ee9f766abf6fbe0cc36234bfe727aba924e9b5fd 43189Author: Keith Packard <keithp@keithp.com> 43190Date: Mon Jul 28 12:09:51 2014 -0700 43191 43192 xfree86: [v2] Remove duplicate InputInfoPtr typedef from xf86Xinput.h 43193 43194 GCC 4.2 doesn't accept 2 typedef declarations of the same type, so 43195 remove the extra one from xf86Xinput.h and have xf86Xinput.h #include 43196 xf86.h to make sure everyone using just that file gets the typedef. 43197 43198 Signed-off-by: Keith Packard <keithp@keithp.com> 43199 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 43200 Reviewed-by: Matthieu Herrb <matthieu@herrb.eu> 43201 43202commit ab32ee35890be51137b56525826ffe219b63eb88 43203Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 43204Date: Mon Aug 11 13:15:24 2014 -0700 43205 43206 XQuartz: Better support turning off "Displays have separate Spaces" on OS X Mavericks 43207 43208 http://xquartz.macosforge.org/trac/ticket/1876 43209 43210 Follow-up to: 1c10b37380d228b35db8a8616a6312ac54f5e59b 43211 43212 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 43213 43214commit dfbc6a1a78ff3192baae6f7f2af98b6e51422f75 43215Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 43216Date: Wed Jul 23 11:53:33 2014 -0700 43217 43218 ListenOnOpenFD: Remove Resets since this is intended to be for hotplugging connections 43219 43220 pharris says that the resets should not be done in the hotplugging case. 43221 43222 This may fix a crash reported against XQuartz: 43223 http://xquartz.macosforge.org/trac/ticket/869 43224 43225 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 43226 Reviewed-by: Peter Harris <pharris@opentext.com> 43227 43228commit 3714f5401690b288045090c4bcd9cb01c6e4860e 43229Author: Eric Anholt <eric@anholt.net> 43230Date: Sun Apr 6 07:57:44 2014 +0100 43231 43232 kdrive: Remove a dead struct. 43233 43234 It's never been used in the history of the tree. 43235 43236 Signed-off-by: Eric Anholt <eric@anholt.net> 43237 Reviewed-by: Keith Packard <keithp@keithp.com> 43238 43239commit ef258fa4fd0c215b1327fb3d94b37e75af9fae6b 43240Author: Eric Anholt <eric@anholt.net> 43241Date: Wed Jul 30 21:11:51 2014 -0700 43242 43243 xv: Drop the ClientPtr from the interface to the DDX. 43244 43245 Nobody was using it. 43246 43247 v2: Merge the hunk that was accidentally in the previous commit into 43248 this one. 43249 43250 Signed-off-by: Eric Anholt <eric@anholt.net> 43251 Reviewed-by: Keith Packard <keithp@keithp.com> 43252 43253commit 850b268e2b9e49445a22e041a356d8ff26ae2147 43254Author: Eric Anholt <eric@anholt.net> 43255Date: Sat Apr 5 10:55:45 2014 +0100 43256 43257 xv: Remove the no-op AllocatePort/FreePort interfaces. 43258 43259 v2: Fix accidentally squashed-in change for dropping client from the 43260 arguments, which should have been in the next commit. 43261 43262 Signed-off-by: Eric Anholt <eric@anholt.net> 43263 Reviewed-by: Keith Packard <keithp@keithp.com> (v2) 43264 43265commit 41d4a626c6ceb15d5b008dca00c6f5032ddd9556 43266Author: Eric Anholt <eric@anholt.net> 43267Date: Sat Apr 5 10:35:52 2014 +0100 43268 43269 kdrive: Don't bother explicitly clearing new window privates to NULL. 43270 43271 Privates are initially cleared to zero by dixInitPrivates(). 43272 43273 Signed-off-by: Eric Anholt <eric@anholt.net> 43274 Reviewed-by: Keith Packard <keithp@keithp.com> 43275 43276commit e311318b35ef4f82996c6e277137569005893b44 43277Author: Eric Anholt <eric@anholt.net> 43278Date: Sat Apr 5 09:11:03 2014 +0100 43279 43280 kdrive: Remove dead KXVPaintRegion(). 43281 43282 It's been unused since mach64 was deleted, and now there's a helper in 43283 core XV. 43284 43285 Signed-off-by: Eric Anholt <eric@anholt.net> 43286 Reviewed-by: Keith Packard <keithp@keithp.com> 43287 43288commit 7eac142fb6e8abb011ce59feda3c2d7b04f2b225 43289Author: Eric Anholt <eric@anholt.net> 43290Date: Sat Apr 5 09:09:17 2014 +0100 43291 43292 xv: Fix malloc-failure cases in the fill color key helper. 43293 43294 Signed-off-by: Eric Anholt <eric@anholt.net> 43295 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 43296 43297commit ea3f3b0786d58b2ba6a9bbf0a32e734670f54b73 43298Author: Eric Anholt <eric@anholt.net> 43299Date: Sat Apr 5 09:07:18 2014 +0100 43300 43301 xv: Move xf86 XV color key helper to core. 43302 43303 Color key overlay implementations want to reuse this code, and XF86's 43304 had bugs (to be fixed in the next commit). 43305 43306 Signed-off-by: Eric Anholt <eric@anholt.net> 43307 Reviewed-by: Keith Packard <keithp@keithp.com> 43308 43309commit b01cfe5f23766b9c13ed6bd889263d5d7a8a351d 43310Author: Eric Anholt <eric@anholt.net> 43311Date: Fri Apr 4 22:58:37 2014 +0100 43312 43313 xv: Move CloseScreen setup from a DIX hook to normal wrapping. 43314 43315 Signed-off-by: Eric Anholt <eric@anholt.net> 43316 Reviewed-by: Keith Packard <keithp@keithp.com> 43317 43318commit 6a71ee79aabe6b611cae443a49786824a0322313 43319Author: Eric Anholt <eric@anholt.net> 43320Date: Fri Apr 4 18:02:14 2014 +0100 43321 43322 xv: Drop unused XvdiVideoStopped(). 43323 43324 Signed-off-by: Eric Anholt <eric@anholt.net> 43325 Reviewed-by: Keith Packard <keithp@keithp.com> 43326 43327commit 4a095902a37cb98d7954d0873dcf1376cbab54b8 43328Author: Eric Anholt <eric@anholt.net> 43329Date: Fri Apr 4 18:01:07 2014 +0100 43330 43331 xv: Drop unused XvdiPreemptVideo(). 43332 43333 Signed-off-by: Eric Anholt <eric@anholt.net> 43334 Reviewed-by: Keith Packard <keithp@keithp.com> 43335 43336commit e7dde86f2343832c52ef74f5c4fe44a72054c79b 43337Author: Eric Anholt <eric@anholt.net> 43338Date: Fri Apr 4 17:54:58 2014 +0100 43339 43340 xv: Move the DDX XV screen private allocation into the DDXes. 43341 43342 XV was going against convention by having the core infrastructure 43343 allocate the private on behalf of the DDX. I was interested in this 43344 because I was trying to make multiple pieces of DDX be able to 43345 allocate adaptors, and that wasn't going to work if DDX-specific code 43346 was hung off of a single global screen private. 43347 43348 Signed-off-by: Eric Anholt <eric@anholt.net> 43349 Reviewed-by: Keith Packard <keithp@keithp.com> 43350 43351commit a146c6d4212ed199002e40419b4aa22da8f49e3a 43352Author: Eric Anholt <eric@anholt.net> 43353Date: Fri Apr 4 13:12:34 2014 +0100 43354 43355 xv: Drop the ddQueryAdaptors() interface. 43356 43357 The core was passing pointers to pxvs's nAdaptors and pAdaptors, and 43358 the two hardware implementations were copying pxvs's nAdaptors and 43359 pAdaptors into those pointers. 43360 43361 Signed-off-by: Eric Anholt <eric@anholt.net> 43362 Reviewed-by: Keith Packard <keithp@keithp.com> 43363 43364commit cb42805c6675bbb9f0b37609fe2ffecc07f46290 43365Author: Eric Anholt <eric@anholt.net> 43366Date: Fri Dec 27 21:54:26 2013 -0800 43367 43368 xv: Move common code for adaptor cleanup to xvmain.c 43369 43370 Since any DDX XV screen cleanup would need this same code for freeing 43371 the tree of pointers for xv adaptors, move it to the dix. 43372 43373 v2: Unconditionalize the pPorts freeing, to match the block above it. 43374 43375 Signed-off-by: Eric Anholt <eric@anholt.net> 43376 Reviewed-by: Keith Packard <keithp@keithp.com> (v1) 43377 43378commit 8cb0da29404d3fed13c1cbedb5c6164e4733ea82 43379Author: Eric Anholt <eric@anholt.net> 43380Date: Fri Dec 27 20:59:12 2013 -0800 43381 43382 xv: Remove dead VIDEO_NO_CLIPPING from the xorg and kdrive DDXes. 43383 43384 As far as I can see, nothing has ever used this flag except possibly 43385 the i.mx6 xorg ddx debug during bringup. 43386 43387 Signed-off-by: Eric Anholt <eric@anholt.net> 43388 Reviewed-by: Keith Packard <keithp@keithp.com> 43389 43390commit a6ec7d5278405d69e60cca2964006fb7ce32c556 43391Author: Eric Anholt <eric@anholt.net> 43392Date: Fri Dec 27 20:53:00 2013 -0800 43393 43394 xv: Remove dead VIDEO_INVERT_CLIPLIST from the xorg and kdrive DDXes. 43395 43396 As far as I can see (looking at trees on my disk, plus googling for 43397 the term), nothing has ever used this flag 43398 43399 Signed-off-by: Eric Anholt <eric@anholt.net> 43400 Reviewed-by: Keith Packard <keithp@keithp.com> 43401 43402commit e6c8c7e46c79b2837a7d0b12079a47734eff1eb7 43403Author: Egbert Eich <eich@freedesktop.org> 43404Date: Mon Aug 4 19:16:30 2014 +0200 43405 43406 BellProc: Send bell event on core protocol bell when requested 43407 43408 XKB allows to override the BellProc() ringing the 'keyboard bell': 43409 instead an event is sent to an X client which can perform an 43410 appropriate action. 43411 In most cases this effectively prevents the core protocol bell 43412 from ringing: if no BellProc() is set for the device, no attempt 43413 is made to ring a bell. 43414 This patch ensures that an XKB bell event is sent also when 43415 the core protocol bell is rung end thus an appropriate action 43416 can be taken by a client. 43417 43418 Signed-off-by: Egbert Eich <eich@freedesktop.org> 43419 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 43420 Signed-off-by: Keith Packard <keithp@keithp.com> 43421 43422commit e8373e4b6a793b48487b1a82d798d97b5c9a10a3 43423Author: Keith Packard <keithp@keithp.com> 43424Date: Thu Jul 31 19:57:53 2014 -0700 43425 43426 Add X_BYTE_ORDER to xorg-server.h 43427 43428 Drivers don't get to use dix-config.h, they use xorg-server.h 43429 instead. Add X_BYTE_ORDER to that file so drivers can see the value. 43430 43431 Signed-off-by: Keith Packard <keithp@keithp.com> 43432 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 43433 43434commit e31564e1a21e73f4d20d6471da4fc7a9b63e4062 43435Author: Michel Dänzer <michel@daenzer.net> 43436Date: Thu Jul 31 08:35:13 2014 -0700 43437 43438 Check for dix-config.h or xorg-server.h before using X_BYTE_ORDER 43439 43440 Now that servermd.h depends on X_BYTE_ORDER being defined in 43441 dix-config.h or xorg-server.h, check to make sure one of those has 43442 been included before using the value. 43443 43444 Reviewed-by: Keith Packard <keithp@keithp.com> 43445 Signed-off-by: Keith Packard <keithp@keithp.com> 43446 43447commit 61afe950e6a1a640ad9c5368549914ea32b90d48 43448Author: Keith Packard <keithp@keithp.com> 43449Date: Mon Jul 21 19:27:20 2014 -0700 43450 43451 xfree86/modes: rotation damage is automatically destroyed on close 43452 43453 Don't try to destroy rotation_damage in the xf86RotateCloseScreen; it 43454 will have been destroyed when the screen pixmap was destroyed. 43455 43456 Signed-off-by: Keith Packard <keithp@keithp.com> 43457 Reviewed-by: Eric Anholt <eric@anholt.net> 43458 43459commit 40dc81154ad38514793f2181447d597b57d39e80 43460Author: Keith Packard <keithp@keithp.com> 43461Date: Sun Jul 20 18:57:42 2014 -0700 43462 43463 present: Avoid crash at server shutdown 43464 43465 When a present flip operation is still in process during server reset, 43466 the call to present_set_abort_flip may not happen until the screen is 43467 being closed, at which point there is no root window to set pixmaps 43468 for. Check to make sure there's a window before resetting window pixmaps. 43469 43470 Signed-off-by: Keith Packard <keithp@keithp.com> 43471 Reviewed-by: Eric Anholt <eric@anholt.net> 43472 43473commit 627bce80894647c681b146a709aad4c390b4374e 43474Author: Keith Packard <keithp@keithp.com> 43475Date: Sun Jul 20 18:53:52 2014 -0700 43476 43477 present: Make window MSC offset 0 initially 43478 43479 The MSC offset used by a window is adjusted as the window moves 43480 between screens, and between shown/unshown. The value shouldn't 43481 matter, but it's helpful for debugging to have window MSC values be 43482 the same as the CRTC MSC at first. 43483 43484 This patch introduces a unique CRTC value so that Present can detect 43485 the first time a window is a PresentPixmap destination and set the MSC 43486 offset to zero, rather than using the fake MSC value as the previous 43487 window MSC. 43488 43489 Signed-off-by: Keith Packard <keithp@keithp.com> 43490 Reviewed-by: Eric Anholt <eric@anholt.net> 43491 43492commit bd4198b01f2baf2284e61ea7ebf4600f554800e9 43493Author: Keith Packard <keithp@keithp.com> 43494Date: Fri Jul 18 11:19:01 2014 -0700 43495 43496 xfree86: Avoid compiler warning for unused vars without systemd 43497 43498 When systemd isn't being used, systemd_logind_release_fd is defined 43499 as an empty macro, leaving the arguments unused. Fix the compiler 43500 warnings by simply removing the local variables and referencing the 43501 structure within the macro call. 43502 43503 Signed-off-by: Keith Packard <keithp@keithp.com> 43504 Reviewed-by: Eric Anholt <eric@anholt.net> 43505 43506commit b063a185ab9d2c54669fb6e036fdbae3707c9e4b 43507Author: David Ung <davidu@nvidia.com> 43508Date: Tue Jul 29 15:01:39 2014 -0700 43509 43510 randr: Fix logic in RRPointerToNearestCrtc 43511 43512 RRPointerToNearestCrtc is suppose to snap to the nearest Crtc, 43513 but best_x and best_y is always positive, hence when calling 43514 SetCursorPosition it will make the cursor even further away. 43515 Correct delta x/y to allow negative values and also use 43516 "width/height -1" in the calculation. Also choose the closest 43517 Crtc by setting the "best" value. 43518 43519 Signed-off-by: David Ung <davidu@nvidia.com> 43520 Reviewed-by: Keith Packard <keithp@keithp.com> 43521 Signed-off-by: Keith Packard <keithp@keithp.com> 43522 43523commit 90803042bcbb9dc6261b5d112acf5613829d9f70 43524Author: Keith Packard <keithp@keithp.com> 43525Date: Wed Jul 30 10:13:53 2014 -0700 43526 43527 Fix files including xorg-server.h by mistake 43528 43529 A few files in the server are including xorg-server.h, which is only 43530 for use by Xorg server drivers. This fixes those errors and then adds 43531 a check to make sure it doesn't happen again. 43532 43533 Signed-off-by: Keith Packard <keithp@keithp.com> 43534 Reviewed-by: Adam Jackson <ajax@redhat.com> 43535 43536commit a08e7a098aa37351294a25fe62b7ff6e95622e64 43537Author: Adam Jackson <ajax@redhat.com> 43538Date: Tue Jun 24 14:21:57 2014 -0400 43539 43540 dri1: Remove unused DRI{,Post}ValidateTree 43541 43542 i810, mga, savage, and tdfx do reference these slots, but only to set 43543 them to NULL, so while this does have API impact it's not actually used. 43544 43545 Reviewed-by: Keith Packard <keithp@keithp.com> 43546 Signed-off-by: Adam Jackson <ajax@redhat.com> 43547 43548commit 5486c834fa3eebd79a4811143e513d6a92b7e504 43549Author: Adam Jackson <ajax@redhat.com> 43550Date: Tue Jun 24 14:16:55 2014 -0400 43551 43552 xquartz: Remove unused DRI{,Post}ValidateTree 43553 43554 Cargo-culted from DRI1, not actually used for anything. 43555 43556 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 43557 Reviewed-by: Keith Packard <keithp@keithp.com> 43558 Signed-off-by: Adam Jackson <ajax@redhat.com> 43559 43560commit a5e7701058c676231d2bc3f9a0c7c0fa9da7ad45 43561Author: Adam Jackson <ajax@redhat.com> 43562Date: Wed May 28 14:48:14 2014 -0400 43563 43564 xkb: Remove some fascinating paranoia from event emission 43565 43566 XkbInterestPtrs are created by clients that already exist, meaning, 43567 clients that have already had ProcVector installed as something other 43568 than InitialProcVector. 43569 43570 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 43571 Signed-off-by: Adam Jackson <ajax@redhat.com> 43572 43573commit ac3af4bd21394291ef6dd9cfcdd8ce6fceefec40 43574Author: Adam Jackson <ajax@redhat.com> 43575Date: Tue Jun 17 12:58:41 2014 -0400 43576 43577 xfree86: Remove unused HardEdges 43578 43579 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 43580 Signed-off-by: Adam Jackson <ajax@redhat.com> 43581 43582commit 4100687ea667970733c3aa69f704be0e3747f9fd 43583Author: Adam Jackson <ajax@redhat.com> 43584Date: Mon Jun 23 15:42:23 2014 -0400 43585 43586 xfree86: Remove useless #include "fb.h" 43587 43588 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 43589 Signed-off-by: Adam Jackson <ajax@redhat.com> 43590 43591commit 2d451c5cdc7f81ea0c94c3542fab45027f829736 43592Author: Adam Jackson <ajax@redhat.com> 43593Date: Thu May 22 14:04:16 2014 -0400 43594 43595 xfree86: Remove pointless xf86RandRCreateScreenResources 43596 43597 Given the #if 0 this was wrapping for no effect. 43598 43599 Reviewed-by: Keith Packard <keithp@keithp.com> 43600 Signed-off-by: Adam Jackson <ajax@redhat.com> 43601 43602commit a8cca7bd5ad7ea36a9585f4927ed53cfa6db1daf 43603Author: Adam Jackson <ajax@redhat.com> 43604Date: Fri May 2 11:43:14 2014 -0400 43605 43606 xfree86: Remove protectMem from VidMemInfo 43607 43608 Never filled in. 43609 43610 Reviewed-by: Keith Packard <keithp@keithp.com> 43611 Signed-off-by: Adam Jackson <ajax@redhat.com> 43612 43613commit e4cf1e58f5745dd39a9dd5eb29acd1cb28710094 43614Author: Adam Jackson <ajax@redhat.com> 43615Date: Fri May 2 11:11:14 2014 -0400 43616 43617 xfree86: Remove unused xf86{Map,Unmap}LegacyIO 43618 43619 I ported these to pciaccess in: 43620 43621 commit 858fbbb40d7c69540cd1fb5315cebf811c6e7b3f 43622 Author: Adam Jackson <ajax@redhat.com> 43623 Date: Fri Sep 16 13:33:04 2011 -0400 43624 43625 pci: Port xf86MapLegacyIO to pciaccess 43626 43627 As of yet there are still no drivers using them, and there's not a lot 43628 of value in having the wrappers when they just trivially call pciaccess 43629 anyway. Nuke 'em. 43630 43631 Reviewed-by: Keith Packard <keithp@keithp.com> 43632 Signed-off-by: Adam Jackson <ajax@redhat.com> 43633 43634commit 96206cf47752af3fc8ca14dc1e03c59ed482167f 43635Author: Adam Jackson <ajax@redhat.com> 43636Date: Tue Jul 8 13:19:08 2014 -0400 43637 43638 xfree86: Remove deprecated PCI types 43639 43640 Reviewed-by: Keith Packard <keithp@keithp.com> 43641 Signed-off-by: Adam Jackson <ajax@redhat.com> 43642 43643commit 46fd5a28ef7191b5ad0244f849c756dd2c5f43b6 43644Author: Adam Jackson <ajax@redhat.com> 43645Date: Fri May 2 11:09:53 2014 -0400 43646 43647 xfree86: Remove some unused pre-pciaccess types 43648 43649 Reviewed-by: Keith Packard <keithp@keithp.com> 43650 Signed-off-by: Adam Jackson <ajax@redhat.com> 43651 43652commit 3bb9f9862b5bea3720ba7922714af729cab38776 43653Author: Adam Jackson <ajax@redhat.com> 43654Date: Fri Feb 28 13:31:08 2014 -0500 43655 43656 xfree86: Remove xf86ConfigActivePciEntity 43657 43658 The giant OBSOLETE DO NOT USE comment has been there since 2000, 43659 probably it's safe to nuke by now. 43660 43661 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 43662 Signed-off-by: Adam Jackson <ajax@redhat.com> 43663 43664commit 51531a67179e342f6ad798de21accf014748a04f 43665Author: Adam Jackson <ajax@redhat.com> 43666Date: Tue Jun 10 13:10:28 2014 -0400 43667 43668 vgahw: Nuke unused vgaCmap.c 43669 43670 Never been built since m12n, can't be needed. 43671 43672 Reviewed-by: Keith Packard <keithp@keithp.com> 43673 Signed-off-by: Adam Jackson <ajax@redhat.com> 43674 43675commit 2db71b232c4756773562182cadb7ccdf5194a62d 43676Author: Adam Jackson <ajax@redhat.com> 43677Date: Tue Jun 10 13:00:11 2014 -0400 43678 43679 saver: Don't open-code IsMapInstalled 43680 43681 Reviewed-by: Keith Packard <keithp@keithp.com> 43682 Signed-off-by: Adam Jackson <ajax@redhat.com> 43683 43684commit dd0e8491f75afbd72bea17172a042aa45ff7e30b 43685Author: Adam Jackson <ajax@redhat.com> 43686Date: Fri Jun 20 12:36:10 2014 -0400 43687 43688 render: Remove unused CopyPicture 43689 43690 Arguably this would be useful API, but it's never called, and a careful 43691 reading of the CPClipMask path reveals that callers would be fairly 43692 disappointed. 43693 43694 Reviewed-by: Keith Packard <keithp@keithp.com> 43695 Signed-off-by: Adam Jackson <ajax@redhat.com> 43696 43697commit 578026fcd58d12dd0be4c0976ee5ecf95a9d61f0 43698Author: Adam Jackson <ajax@redhat.com> 43699Date: Tue May 20 12:43:59 2014 -0400 43700 43701 os: Remove LocalClientCred 43702 43703 The comment lies, shm hasn't used this code since: 43704 43705 commit fdef7be5c8d5989e0aa453d0a5b86d0a6952e960 43706 Author: Alan Coopersmith <alan.coopersmith@sun.com> 43707 Date: Tue Oct 9 18:44:04 2007 -0700 43708 43709 Sun bug 6589829: include zoneid of shm segment in access [...] 43710 43711 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 43712 Signed-off-by: Adam Jackson <ajax@redhat.com> 43713 43714commit cad9b053d52f62432dfd70e42e0240de77027cae 43715Author: Adam Jackson <ajax@redhat.com> 43716Date: Tue Jul 8 13:24:25 2014 -0400 43717 43718 os: Remove deprecated malloc/free wrappers 43719 43720 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 43721 Signed-off-by: Adam Jackson <ajax@redhat.com> 43722 43723commit d5b279977236c4f80a8ba81db1f81a314484facf 43724Author: Adam Jackson <ajax@redhat.com> 43725Date: Thu May 22 13:58:45 2014 -0400 43726 43727 miext/shadow: Remove shadowInit 43728 43729 This code is nonsensical. You end up creating a screen-sized pixmap 43730 that's totally detached from everything else, which you then listen for 43731 damage on, which means you'll never hear any damage, which means your 43732 shadow update hooks will never get called. Any driver using this would 43733 be sorely disappointed. 43734 43735 Reviewed-by: Keith Packard <keithp@keithp.com> 43736 Signed-off-by: Adam Jackson <ajax@redhat.com> 43737 43738commit cf4793d99ecee4dfd6094e02b1ccb89eb744d313 43739Author: Adam Jackson <ajax@redhat.com> 43740Date: Fri Feb 28 15:36:31 2014 -0500 43741 43742 miext/shadow: Remove ancient backwards-compatibility hack 43743 43744 Here's a trip down memory lane. Back when we merged kdrive we adopted 43745 kdrive's version of shadow, which used damage directly instead of 43746 hand-rolling it. However a couple of Xorg drivers referred to the 43747 accumulated damage region in the shadow private directly, so I added a 43748 hack to copy the damage region around. 43749 43750 That was 9148d8700b7c5afc2644e5820c57c509378f93ce, back in early 2006. 43751 Eight years is unusually patient for me. The neomagic and trident drivers 43752 were still relying on this, but they've been modified to ask the damage 43753 code for the region instead. 43754 43755 Reviewed-by: Keith Packard <keithp@keithp.com> 43756 Signed-off-by: Adam Jackson <ajax@redhat.com> 43757 43758commit 35eabf2e5272e17e6765027d4baea43b34c66933 43759Author: Adam Jackson <ajax@redhat.com> 43760Date: Tue Jul 8 14:11:22 2014 -0400 43761 43762 mi: Remove apparently unused miSegregateChildren 43763 43764 This came in between XFree86 4.3 and 4.4, I'm not entirely sure what it 43765 was meant to do. 43766 43767 Reviewed-by: Keith Packard <keithp@keithp.com> 43768 Signed-off-by: Adam Jackson <ajax@redhat.com> 43769 43770commit 150acef3af9bb43cb60cbcc558ef3ad4a710e624 43771Author: Adam Jackson <ajax@redhat.com> 43772Date: Mon May 19 13:46:12 2014 -0400 43773 43774 kdrive: Remove some dead bits of the man page 43775 43776 Reviewed-by: Keith Packard <keithp@keithp.com> 43777 Signed-off-by: Adam Jackson <ajax@redhat.com> 43778 43779commit 71ef49062ee33ac49033fbcd67bab88120882668 43780Author: Adam Jackson <ajax@redhat.com> 43781Date: Fri Feb 28 14:05:25 2014 -0500 43782 43783 dix: Make some LBX-era code more obvious 43784 43785 isItTimeToYield in the conditional effectively didn't do anything here. 43786 Take it out, and remove the comment since LBX proxies aren't a thing for 43787 us anymore. 43788 43789 Reviewed-by: Keith Packard <keithp@keithp.com> 43790 Signed-off-by: Adam Jackson <ajax@redhat.com> 43791 43792commit a317e0a974c1597d55fe05a38b4a495c4e4e8d71 43793Author: Adam Jackson <ajax@redhat.com> 43794Date: Fri May 2 11:50:17 2014 -0400 43795 43796 bsd: Remove some reference arm code behind #if 0 43797 43798 git history is reference enough, thanks. 43799 43800 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 43801 Signed-off-by: Adam Jackson <ajax@redhat.com> 43802 43803commit 8a60d1c3f4a62675cad005107f56413f946ecd77 43804Author: Keith Packard <keithp@keithp.com> 43805Date: Mon Jul 28 11:10:15 2014 -0700 43806 43807 glamor: Eliminate diagonal tearing in xv 43808 43809 This uses a single large triangle and a scissor to draw the video 43810 instead of two triangles. 43811 43812 Signed-off-by: Keith Packard <keithp@keithp.com> 43813 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 43814 43815commit 29eaa61cb210e0c67004bc567af0c15d48aad626 43816Author: Adam Jackson <ajax@redhat.com> 43817Date: Tue Jul 22 11:05:32 2014 -0400 43818 43819 mi: Remove semi-arbitrary arch awareness in packed coordinate macros 43820 43821 The majority of arches end up on the right-shift path here. I can't 43822 think of any arch where that'd be slower than a divide, and semantically 43823 it makes more sense to think of this as a shift operation anyway. 43824 43825 Signed-off-by: Adam Jackson <ajax@redhat.com> 43826 Reviewed-by: Keith Packard <keithp@keithp.com> 43827 Signed-off-by: Keith Packard <keithp@keithp.com> 43828 43829commit 6ddd164508f81e9582b66061b385bd34ac1ccb7e 43830Author: Adam Jackson <ajax@redhat.com> 43831Date: Tue Jul 22 10:59:13 2014 -0400 43832 43833 xfree86: Unify the ppc/sparc mmio-swap-or-not conditionals 43834 43835 Map SPARC_MMIO_IS_BE and PPC_MMIO_IS_BE to MMIO_IS_BE and use the same 43836 macros for both since they're identical. 43837 43838 Reviewed-by: Julien Cristau <jcristau@debian.org> 43839 Signed-off-by: Adam Jackson <ajax@redhat.com> 43840 Signed-off-by: Keith Packard <keithp@keithp.com> 43841 43842commit 8ffd1c066ab1e993d8964aef12a7c585329c677c 43843Author: Adam Jackson <ajax@redhat.com> 43844Date: Tue Jul 22 10:59:12 2014 -0400 43845 43846 xfree86: Clean up some silly __sparc macro usage 43847 43848 The top of this file already defines __sparc__ if __sparc is defined. 43849 43850 Reviewed-by: Julien Cristau <jcristau@debian.org> 43851 Signed-off-by: Adam Jackson <ajax@redhat.com> 43852 Signed-off-by: Keith Packard <keithp@keithp.com> 43853 43854commit 1c1711b57f2c278c0df03bf9e7e9a1776a546968 43855Author: Adam Jackson <ajax@redhat.com> 43856Date: Tue Jul 22 10:59:11 2014 -0400 43857 43858 xfree86: Pull generic barrier() definition up to top level 43859 43860 And remove the redundant redecl from the nds32 section. 43861 43862 Reviewed-by: Julien Cristau <jcristau@debian.org> 43863 Signed-off-by: Adam Jackson <ajax@redhat.com> 43864 Signed-off-by: Keith Packard <keithp@keithp.com> 43865 43866commit c73929bbfc14ba80397c53ca98d3b503c807ae78 43867Author: Adam Jackson <ajax@redhat.com> 43868Date: Tue Jul 22 10:59:10 2014 -0400 43869 43870 xfree86: Remove MMIO_ONB* and friends 43871 43872 Non-barrier-emitting MMIO writes. They appear to be utterly unused, 43873 burn it all down. 43874 43875 Reviewed-by: Julien Cristau <jcristau@debian.org> 43876 Signed-off-by: Adam Jackson <ajax@redhat.com> 43877 Signed-off-by: Keith Packard <keithp@keithp.com> 43878 43879commit 9b33e31d4160dc484ee569740f883f4d34e10653 43880Author: Adam Jackson <ajax@redhat.com> 43881Date: Tue Jul 22 10:59:09 2014 -0400 43882 43883 xfree86: Clean up powerpc barrier decls 43884 43885 I think the externs are there for the non-gcc case? And maybe there was 43886 some assembly code to implement that once? Whatever, at this point on 43887 ppc the compiler is either gcc or willing to pretend. The macros below 43888 the decls take care of the actual eieio so the externs can just go. 43889 43890 Also remove a comment that maybe made sense once upon a time. 43891 43892 Reviewed-by: Julien Cristau <jcristau@debian.org> 43893 Signed-off-by: Adam Jackson <ajax@redhat.com> 43894 Signed-off-by: Keith Packard <keithp@keithp.com> 43895 43896commit b5141a1fab4d45c1af62db8dc712deb9776668a9 43897Author: Adam Jackson <ajax@redhat.com> 43898Date: Tue Jul 22 10:59:08 2014 -0400 43899 43900 xfree86: Simplify a bunch of OS and arch conditionals 43901 43902 All of this is inside #ifdef __GNUC__, between that and configure.ac we 43903 can assume there's a unixy thing under us. Given that there's no real 43904 reason to limit the arch paths to particular OSes, so let's not. 43905 43906 The final #elif here, combined with the ones before it, effectively said 43907 "if not (alpha amd64 sparc* mips* ppc* arm* nds32 m68k sh hppa s390 m32r)", 43908 and as the comment above it hints, it's meant to cover i386 (and happens to 43909 also cover itanic). Flip the conditional around to be sensible. 43910 43911 Reviewed-by: Julien Cristau <jcristau@debian.org> 43912 Signed-off-by: Adam Jackson <ajax@redhat.com> 43913 Signed-off-by: Keith Packard <keithp@keithp.com> 43914 43915commit 6d3ba80f197b04b6e0814190cf5c678c786c14d6 43916Author: Adam Jackson <ajax@redhat.com> 43917Date: Tue Jul 22 10:59:07 2014 -0400 43918 43919 xfree86: Remove pre-2.6 Linux ppc support 43920 43921 2.6.0 was December 2003, you've had plenty of time to get your head in 43922 the game. 43923 43924 Reviewed-by: Julien Cristau <jcristau@debian.org> 43925 Signed-off-by: Adam Jackson <ajax@redhat.com> 43926 Signed-off-by: Keith Packard <keithp@keithp.com> 43927 43928commit e242e82eba17414603c8806fffc9ecd235eebabb 43929Author: Adam Jackson <ajax@redhat.com> 43930Date: Tue Jul 22 10:59:06 2014 -0400 43931 43932 xfree86: Remove a useless !__SUNPRO_C guard 43933 43934 You can't tell from context here, but this is all inside #ifdef 43935 __GNUC__, so this conditional can't do squat. 43936 43937 Reviewed-by: Julien Cristau <jcristau@debian.org> 43938 Signed-off-by: Adam Jackson <ajax@redhat.com> 43939 Signed-off-by: Keith Packard <keithp@keithp.com> 43940 43941commit 93745a6c1afc830759219f8632ec565d83c7bc12 43942Author: Adam Jackson <ajax@redhat.com> 43943Date: Tue Jul 22 10:59:05 2014 -0400 43944 43945 xfree86: Undef GCCUSESGAS 43946 43947 Can't be needed, we've never defined it in modular xserver. 43948 43949 Reviewed-by: Julien Cristau <jcristau@debian.org> 43950 Signed-off-by: Adam Jackson <ajax@redhat.com> 43951 Signed-off-by: Keith Packard <keithp@keithp.com> 43952 43953commit b1d9bc8d1d828b0dedda46da98628b47915e6a04 43954Author: Adam Jackson <ajax@redhat.com> 43955Date: Tue Jul 22 10:59:04 2014 -0400 43956 43957 xfree86: Remove an unlikely bit of #pragma 43958 43959 __USLC__ appears to mean the SCO OpenServer compiler, which configure.ac 43960 doesn't think is an OS the xfree86 ddx supports. The conditionals 43961 surrounding these pragmas effectively mean "if not gcc and not Sun C", 43962 and probably arbitrary pragmas aren't supported by arbitrary compilers. 43963 43964 Reviewed-by: Julien Cristau <jcristau@debian.org> 43965 Signed-off-by: Adam Jackson <ajax@redhat.com> 43966 Signed-off-by: Keith Packard <keithp@keithp.com> 43967 43968commit f7f9ccef1dc251a1b95ca698ac39bd0a47f0c8fa 43969Author: Adam Jackson <ajax@redhat.com> 43970Date: Tue Jul 22 10:59:03 2014 -0400 43971 43972 xfree86: Undef __HIGHC__ 43973 43974 MetaWare High C++ compiler? xfree86 cvs history shows this being added 43975 in a commit whose text is, classically, "updates". metaware.com 43976 redirects to a 404 on synopsys.com, which to me indicates it's not super 43977 important to them, and their order form won't even tell you how much the 43978 thing costs. At any rate if this is worth worrying about it's worth 43979 letting autoconf worry about for us. 43980 43981 Reviewed-by: Julien Cristau <jcristau@debian.org> 43982 Signed-off-by: Adam Jackson <ajax@redhat.com> 43983 Signed-off-by: Keith Packard <keithp@keithp.com> 43984 43985commit 80446086b9cfcc5e23a400d7fa38ec773fae68fc 43986Author: Adam Jackson <ajax@redhat.com> 43987Date: Tue Jul 22 10:59:02 2014 -0400 43988 43989 xfree86: Undef FAKEIT 43990 43991 I guess this is meant to stub out all I/O port calls? Whatever, it's 43992 not been defined by the buildsystem at least as far back as monolith 43993 6.8.2. 43994 43995 Reviewed-by: Julien Cristau <jcristau@debian.org> 43996 Signed-off-by: Adam Jackson <ajax@redhat.com> 43997 Signed-off-by: Keith Packard <keithp@keithp.com> 43998 43999commit 8002b1a8cb18ecc6526e8ebfd769767f8991ebd6 44000Author: Adam Jackson <ajax@redhat.com> 44001Date: Tue Jul 22 10:59:01 2014 -0400 44002 44003 xfree86: Remove a few random ppc decls 44004 44005 Whatever these are, they're not something grep can find, they must not 44006 be used. 44007 44008 Reviewed-by: Julien Cristau <jcristau@debian.org> 44009 Signed-off-by: Adam Jackson <ajax@redhat.com> 44010 Signed-off-by: Keith Packard <keithp@keithp.com> 44011 44012commit 5f5af5d6692b625f6231c8bca3ac801f2ac75717 44013Author: Adam Jackson <ajax@redhat.com> 44014Date: Tue Jul 22 10:59:00 2014 -0400 44015 44016 xfree86: Remove remaining unused unaligned accessors 44017 44018 Reviewed-by: Julien Cristau <jcristau@debian.org> 44019 Signed-off-by: Adam Jackson <ajax@redhat.com> 44020 Signed-off-by: Keith Packard <keithp@keithp.com> 44021 44022commit d28b788e11182d134bce5a414359841b37b62ac0 44023Author: Adam Jackson <ajax@redhat.com> 44024Date: Tue Jul 22 10:58:59 2014 -0400 44025 44026 xfree86: Move generic unaligned helpers into int10 code 44027 44028 This is the only place they're actually used (well, aside from some XAA 44029 code in the s3 driver, but one s3 and 2 XAA). 44030 44031 Reviewed-by: Julien Cristau <jcristau@debian.org> 44032 Signed-off-by: Adam Jackson <ajax@redhat.com> 44033 Signed-off-by: Keith Packard <keithp@keithp.com> 44034 44035commit 956a8d5c92a66ddd8d00e63a95b17cfd376fe424 44036Author: Adam Jackson <ajax@redhat.com> 44037Date: Tue Jul 22 10:58:58 2014 -0400 44038 44039 xfree86: Remove unused unaligned int64 helpers 44040 44041 Reviewed-by: Julien Cristau <jcristau@debian.org> 44042 Signed-off-by: Adam Jackson <ajax@redhat.com> 44043 Signed-off-by: Keith Packard <keithp@keithp.com> 44044 44045commit da2b2671863655776d60102ddc7af61470d14022 44046Author: Adam Jackson <ajax@redhat.com> 44047Date: Tue Jul 22 10:58:57 2014 -0400 44048 44049 xfree86: Unspecialize gcc variants of unaligned memory access 44050 44051 Yes yes, very clever, memmove works fine on gcc too, let's just do the 44052 portable thing since none of this is performance code. 44053 44054 Reviewed-by: Julien Cristau <jcristau@debian.org> 44055 Signed-off-by: Adam Jackson <ajax@redhat.com> 44056 Signed-off-by: Keith Packard <keithp@keithp.com> 44057 44058commit e03c902e03dea7a6a98ad5fb00580b9404b67c4d 44059Author: Adam Jackson <ajax@redhat.com> 44060Date: Tue Jul 22 10:58:56 2014 -0400 44061 44062 xfree86: Undefine NO_INLINE 44063 44064 Nothing in the server defines this, nor do any drivers. 44065 44066 Reviewed-by: Julien Cristau <jcristau@debian.org> 44067 Signed-off-by: Adam Jackson <ajax@redhat.com> 44068 Signed-off-by: Keith Packard <keithp@keithp.com> 44069 44070commit f63b8e44ab278dd65be93146790bf150717fc1c8 44071Author: Adam Jackson <ajax@redhat.com> 44072Date: Tue Jul 22 10:58:55 2014 -0400 44073 44074 xfree86: Remove MMIO_MOVE32 44075 44076 Only used by mach64's XAA code, which isn't built if XAA isn't 44077 available, and it isn't. 44078 44079 Reviewed-by: Julien Cristau <jcristau@debian.org> 44080 Signed-off-by: Adam Jackson <ajax@redhat.com> 44081 Signed-off-by: Keith Packard <keithp@keithp.com> 44082 44083commit 1100935650cc0032709d095d6d5a5f136bec99b9 44084Author: Adam Jackson <ajax@redhat.com> 44085Date: Tue Jul 22 10:58:54 2014 -0400 44086 44087 xfree86: Remove nds32_flush_icache 44088 44089 I guess this might have been needed for elfloader, except we didn't 44090 support nds32 back then, so I assume this was cargo-culted from 44091 ppc_flush_icache, which is also dead now. 44092 44093 Reviewed-by: Julien Cristau <jcristau@debian.org> 44094 Signed-off-by: Adam Jackson <ajax@redhat.com> 44095 Signed-off-by: Keith Packard <keithp@keithp.com> 44096 44097commit eb76228080da6fb28c34c903a9854aebe157d2c7 44098Author: Adam Jackson <ajax@redhat.com> 44099Date: Tue Jul 22 10:58:53 2014 -0400 44100 44101 xfree86: Remove #include "compiler.h" from places that don't need it 44102 44103 Reviewed-by: Julien Cristau <jcristau@debian.org> 44104 Signed-off-by: Adam Jackson <ajax@redhat.com> 44105 Signed-off-by: Keith Packard <keithp@keithp.com> 44106 44107commit 4afedf545b673282f2e214c0e2c759c9be9b9a2a 44108Author: Ross Burton <ross.burton@intel.com> 44109Date: Wed Jul 9 11:33:23 2014 +0100 44110 44111 configure.ac: add option for xshmfence 44112 44113 xshmfence is usable outside of DRI3, and is currently autodetected which isn't 44114 good for distributions where deterministic builds are desired. 44115 44116 Signed-off-by: Ross Burton <ross.burton@intel.com> 44117 Reviewed-by: Matt Turner <mattst88@gmail.com> 44118 Signed-off-by: Keith Packard <keithp@keithp.com> 44119 44120commit 2172714c67d8701aa54c202e89f246f1dddac80a 44121Author: Carlos Garnacho <carlosg@gnome.org> 44122Date: Tue Jul 22 17:55:25 2014 +0200 44123 44124 xwayland: Only disable/enable devices on capabilities change 44125 44126 Anytime a capability is first reported, the device is created, but after 44127 that, it is only disabled/enabled. 44128 44129 This is a closer behavior to what Xorg does on VT switch, at the expense 44130 of maybe leaving a dangling "physical" device if a capability goes for good. 44131 Otherwise, any DeviceIntPtr (re)created after server initialization will be 44132 left floating, and bad things happen when the wayland enter event handler 44133 tries to update cursor position based on a floating device. 44134 44135 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 44136 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 44137 Signed-off-by: Keith Packard <keithp@keithp.com> 44138 44139commit 556cdf8fe870bc23e0393c0eed15c86f49a0b9f8 44140Merge: 83701c4ee 1faa76670 44141Author: Keith Packard <keithp@keithp.com> 44142Date: Thu Jul 24 16:51:00 2014 -0700 44143 44144 Merge remote-tracking branch 'jeremyhu/master' 44145 44146commit 83701c4ee8f79f69f4286ef262cd3ae3922629b9 44147Author: Adam Jackson <ajax@redhat.com> 44148Date: Mon Jul 21 10:21:07 2014 -0400 44149 44150 xfree86: Remove i2c multimedia modules 44151 44152 These came in with the GATOS merge I think. The only driver using them 44153 was radeon, and then only in UMS mode. The radeon driver dropped UMS 44154 support from the main branch about two years ago, the UMS branch hasn't 44155 been touched in about fifteen months, and does not build against 1.16 in 44156 any case, so this is all dead code. 44157 44158 Signed-off-by: Adam Jackson <ajax@redhat.com> 44159 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 44160 Reviewed-by: Daniel Stone <daniels@collabora.com> 44161 Signed-off-by: Keith Packard <keithp@keithp.com> 44162 44163commit 4302484ef2e67b2e05756c7ee0834dce938f1b44 44164Author: Adam Jackson <ajax@redhat.com> 44165Date: Mon Jul 21 10:06:53 2014 -0400 44166 44167 dix: Remove arch awareness from servermd.h (v2) 44168 44169 At this point we have no architectures where image byte order is 44170 different from bitmap bit order, or where either of those two are not 44171 also the native word endianness. Hooray, one more place where we don't 44172 have to worry about enabling new CPU architectures. 44173 44174 v2: Rebase to master to handle the addition of ppc64le, arc, and xtensa, 44175 and use autoconf's endianness detection instead of gcc predefines. 44176 44177 Signed-off-by: Adam Jackson <ajax@redhat.com> 44178 Reviewed-by: Daniel Stone <daniels@collabora.com> 44179 Signed-off-by: Keith Packard <keithp@keithp.com> 44180 44181commit 860cd209679a2e56460977882b3b3826ddcb8710 44182Author: Adam Jackson <ajax@redhat.com> 44183Date: Mon Jul 21 10:06:52 2014 -0400 44184 44185 dix: Remove a weird case of little-endian s390 44186 44187 I really don't think this was ever correct, but I'm also not sure what 44188 non-Linux Unix this was meant to enable. The only one I know of was 44189 OS/390 / z/OS / OpenEdition, but I think that was big-endian too. 44190 44191 At any rate this is all about to go away, so this is just removing an 44192 edge case. 44193 44194 Signed-off-by: Adam Jackson <ajax@redhat.com> 44195 Reviewed-by: Daniel Stone <daniels@collabora.com> 44196 Signed-off-by: Keith Packard <keithp@keithp.com> 44197 44198commit f593d76de76a645e29af5dd457ad1fd5a30736cc 44199Author: Adam Jackson <ajax@redhat.com> 44200Date: Mon Jul 21 10:06:51 2014 -0400 44201 44202 dix: Remove wacky sparc special casing (v2) 44203 44204 This appears to be defining sparc if ever __sparc or __sparc__ were 44205 defined, which is almost reasonable, but these days we want to be using 44206 the __arch__ style. Why any of this would ever be triggered on m68k is 44207 truly a mystery for the ages. 44208 44209 v2: Fix commit message, as noted by nix 44210 44211 Signed-off-by: Adam Jackson <ajax@redhat.com> 44212 Reviewed-by: Daniel Stone <daniels@collabora.com> 44213 Signed-off-by: Keith Packard <keithp@keithp.com> 44214 44215commit 17c3347f14822b9f7da4253c71f6ed51be2b38d1 44216Author: Adam Jackson <ajax@redhat.com> 44217Date: Mon Jul 21 10:06:50 2014 -0400 44218 44219 dix: Default GLYPHPADBYTES to 4 44220 44221 This effectively no longer varied across architectures anyway. 44222 44223 Signed-off-by: Adam Jackson <ajax@redhat.com> 44224 Reviewed-by: Daniel Stone <daniels@collabora.com> 44225 Signed-off-by: Keith Packard <keithp@keithp.com> 44226 44227commit 094b76735cb99b9f26e21f32170614454c644653 44228Author: Adam Jackson <ajax@redhat.com> 44229Date: Mon Jul 21 10:06:49 2014 -0400 44230 44231 dix: Remove some cfb leftovers 44232 44233 These macros meant something in cfb, but not in fb. 44234 44235 Signed-off-by: Adam Jackson <ajax@redhat.com> 44236 Reviewed-by: Daniel Stone <daniels@collabora.com> 44237 Signed-off-by: Keith Packard <keithp@keithp.com> 44238 44239commit fee9a1536651ace9324b956df809ef22a63938f8 44240Author: Adam Jackson <ajax@redhat.com> 44241Date: Mon Jul 21 10:06:48 2014 -0400 44242 44243 dix: Remove some dead macros 44244 44245 Signed-off-by: Adam Jackson <ajax@redhat.com> 44246 Reviewed-by: Daniel Stone <daniels@collabora.com> 44247 Signed-off-by: Keith Packard <keithp@keithp.com> 44248 44249commit 634aa310c5130ef4d38f339bb26b01b5a7bad4e9 44250Author: Adam Jackson <ajax@redhat.com> 44251Date: Mon Jul 21 10:06:47 2014 -0400 44252 44253 dix: (Don't) change BITMAP_SCANLINE_UNIT on Linux s390{, x} 44254 44255 Every other architecture sets this to 32, and I can't think of any 44256 benefit s390 would derive from changing it. It is, at any rate, 44257 something the client already copes with, and the only internal code 44258 impact seems to be some complicated math in miGetPlane, which you never 44259 hit if you're using fb. 44260 44261 Signed-off-by: Adam Jackson <ajax@redhat.com> 44262 Reviewed-by: Daniel Stone <daniels@collabora.com> 44263 Signed-off-by: Keith Packard <keithp@keithp.com> 44264 44265commit f6469f1910c72add358b7263d69bd116b4450ece 44266Author: Adam Jackson <ajax@redhat.com> 44267Date: Mon Jul 21 10:06:46 2014 -0400 44268 44269 dix: Remove some XFree86 3.x leftovers 44270 44271 Signed-off-by: Adam Jackson <ajax@redhat.com> 44272 Reviewed-by: Daniel Stone <daniels@collabora.com> 44273 Signed-off-by: Keith Packard <keithp@keithp.com> 44274 44275commit 5aaacb501ff26bad391664cc7367bfe5164f72c7 44276Author: Adam Jackson <ajax@redhat.com> 44277Date: Mon Jul 21 10:06:45 2014 -0400 44278 44279 dix: Remove an ancient IBM configuration 44280 44281 Whatever unix this was meant to be is either no longer in circulation, 44282 or is AIX, which we don't claim to support anyway. 44283 44284 Signed-off-by: Adam Jackson <ajax@redhat.com> 44285 Reviewed-by: Daniel Stone <daniels@collabora.com> 44286 Signed-off-by: Keith Packard <keithp@keithp.com> 44287 44288commit 06268a0c55b7ce2737bdaa39f11cce07efa8bfa3 44289Author: Aaron Plattner <aplattner@nvidia.com> 44290Date: Sat Jul 19 09:59:07 2014 -0700 44291 44292 xfree86: Remove stale /usr/share/X11/xorg.conf.d hard-code from the xorg.conf man page 44293 44294 I forgot that the old behavior of searching in /usr/share/X11/xorg.conf.d was 44295 documented in the man page. 44296 44297 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 44298 Fixes: acc0b5edd1dc ("xfree86: Only support one sysconfigdir") 44299 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 44300 Signed-off-by: Keith Packard <keithp@keithp.com> 44301 44302commit 1faa76670572e3478965fd2cd9ab60ab2d865e3a 44303Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 44304Date: Sat Jul 19 17:08:09 2014 -0700 44305 44306 mieq: Fix a crash regression in mieqProcessDeviceEvent 44307 44308 (lldb) bt 44309 * thread #6: tid = 0x92d4eb, 0x00000001001dee94 X11.bin`mieqProcessDeviceEvent(dev=0x0000000000000000, event=0x0000000100298bb0, 44310 screen=0x0000000000000000) + 36 at mieq.c:519, stop reason = EXC_BAD_ACCESS (code=1, address=0x44) 44311 * frame #0: 0x00000001001dee94 X11.bin`mieqProcessDeviceEvent(dev=0x0000000000000000, event=0x0000000100298bb0, screen=0x0000000000000000) + 36 at 44312 mieq.c:519 44313 frame #1: 0x00000001001df3eb X11.bin`mieqProcessInputEvents + 555 at mieq.c:631 44314 frame #2: 0x0000000100017674 X11.bin`ProcessInputEvents + 20 at darwinEvents.c:422 44315 frame #3: 0x0000000100175eaa X11.bin`Dispatch + 154 at dispatch.c:357 44316 frame #4: 0x0000000100181b4a X11.bin`dix_main(argc=4, argv=0x00007fff5fbff750, envp=0x00007fff5fbff650) + 1594 at main.c:296 44317 frame #5: 0x000000010001ba80 X11.bin`server_thread(arg=0x0000000101208220) + 64 at quartzStartup.c:66 44318 frame #6: 0x00007fff89bb9899 libsystem_pthread.dylib`_pthread_body + 138 44319 frame #7: 0x00007fff89bb972a libsystem_pthread.dylib`_pthread_start + 137 44320 frame #8: 0x00007fff89bbdfc9 libsystem_pthread.dylib`thread_start + 13 44321 44322 Regression from: 9fb08310b51b46736f3ca8dbc04efdf502420403 44323 44324 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 44325 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 44326 44327commit cff12936275db2f71f6d24f9ea0985a0d14af454 44328Author: Keith Packard <keithp@keithp.com> 44329Date: Fri Jul 18 21:46:23 2014 -0700 44330 44331 glamor: sync_fence_set_triggered should use glFlush, not glFinish 44332 44333 I intended to use glFlush all along, but somehow managed to type 44334 glFinish instead. glFlush is sufficient (for a single-queue GPU) to 44335 ensure serialization between queued rendering in the X server and 44336 future rendering from the client. 44337 44338 Signed-off-by: Keith Packard <keithp@keithp.com> 44339 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 44340 44341commit cfa302d6224d10860e60491333950544c4fb9b04 44342Author: Keith Packard <keithp@keithp.com> 44343Date: Fri Jul 18 11:16:27 2014 -0700 44344 44345 glamor: Add support for SHM sync fences 44346 44347 This hooks up SHM sync fences to complete the requirements for DRI3 44348 running on Glamor. 44349 44350 Signed-off-by: Keith Packard <keithp@keithp.com> 44351 Reviewed-by: Eric Anholt <eric@anholt.net> 44352 44353commit ac3acab1311c47cc73e2681e575f0407284c3c66 44354Author: Colin Walters <walters@verbum.org> 44355Date: Fri Jul 18 09:13:54 2014 -0400 44356 44357 xf86platformBus: Add assertion to avoid (fatal) compiler warning 44358 44359 Compilation of -video-intel started failing in gnome-continuous, 44360 it's because xserver has -Werror=return-type on, and gcc can't 44361 prove this function always returns a value: 44362 44363 /usr/include/xorg/xf86platformBus.h:119:1: error: control reaches end of non-void function [-Werror=return-type] 44364 44365 Let's add assertions to the accessor functions to fix this. 44366 44367 Reviewed-by: Keith Packard <keithp@keithp.com> 44368 Signed-off-by: Keith Packard <keithp@keithp.com> 44369 44370commit 5c2e9fa3d6c4bdf626fededce866056ba8eca502 44371Merge: e678b4971 6d4954884 44372Author: Keith Packard <keithp@keithp.com> 44373Date: Thu Jul 17 20:06:21 2014 -0700 44374 44375 Merge remote-tracking branch 'anholt/glamor-next' 44376 44377commit 6d4954884908ea9894fcfe9836db1ba7bb45be61 44378Merge: 9ddcb20f4 55f5bfb57 44379Author: Eric Anholt <eric@anholt.net> 44380Date: Thu Jul 17 18:00:44 2014 -0700 44381 44382 Merge remote-tracking branch 'origin/master' into glamor-next 44383 44384 I've done this merge manually to resolve the minor conflict in glamor.c. 44385 44386 Signed-off-by: Eric Anholt <eric@anholt.net> 44387 44388commit 9ddcb20f47b5e199989c8990512b0bca1354af86 44389Author: Eric Anholt <eric@anholt.net> 44390Date: Mon Jun 16 07:24:09 2014 +0100 44391 44392 glamor: Drop the "are we doing a series of blits or draws" logic. 44393 44394 It's unused since keithp's copy acceleration code completely replaced 44395 glamor_copyarea.c and removed the blit path. 44396 44397 Signed-off-by: Eric Anholt <eric@anholt.net> 44398 Reviewed-by: Keith Packard <keithp@keithp.com> 44399 44400commit b03a581d8cbe3f29140935063b865285e2a00333 44401Author: Eric Anholt <eric@anholt.net> 44402Date: Mon Jun 16 07:15:58 2014 +0100 44403 44404 glamor: Remove a dead prototype. 44405 44406 The corresponding code was deleted in 44407 2ff41008494e6c5909c058f1f80b4f66617dada1 (2012) 44408 44409 Signed-off-by: Eric Anholt <eric@anholt.net> 44410 Reviewed-by: Keith Packard <keithp@keithp.com> 44411 44412commit b5f94df319469ad44b0e88374a0d389414803f7a 44413Author: Eric Anholt <eric@anholt.net> 44414Date: Mon Jun 16 07:06:30 2014 +0100 44415 44416 glamor: Drop dead glamor_restore_pixmap_to_texture(). 44417 44418 Unused since the glamor_prepare.c replacement of glamor_finish_access(). 44419 44420 Signed-off-by: Eric Anholt <eric@anholt.net> 44421 Reviewed-by: Keith Packard <keithp@keithp.com> 44422 44423commit 98155bd9d9b2a15a4dbcf80b2b57a7636efb14da 44424Author: Eric Anholt <eric@anholt.net> 44425Date: Mon Jun 16 07:01:28 2014 +0100 44426 44427 glamor: Drop dead glamor_download_pixmap_to_cpu() 44428 44429 Signed-off-by: Eric Anholt <eric@anholt.net> 44430 Reviewed-by: Keith Packard <keithp@keithp.com> 44431 44432commit b6181007de357da58125e022992f165b10eda65d 44433Author: Eric Anholt <eric@anholt.net> 44434Date: Mon Jun 16 07:04:01 2014 +0100 44435 44436 glamor: Drop dead get/pub sub pixmap functions. 44437 44438 These were replaced by the new glamor_prepare.c code. 44439 44440 Signed-off-by: Eric Anholt <eric@anholt.net> 44441 Reviewed-by: Keith Packard <keithp@keithp.com> 44442 44443commit e310387f443b6333edf02c8980daa303505382b4 44444Author: Eric Anholt <eric@anholt.net> 44445Date: Sun Jan 12 10:19:10 2014 -0800 44446 44447 glamor: Remove always-true yInverted flag. 44448 44449 All users of glamor had the same value set, and it complicated things 44450 for no reason. 44451 44452 Signed-off-by: Eric Anholt <eric@anholt.net> 44453 Reviewed-by: Keith Packard <keithp@keithp.com> 44454 44455commit d71ecbb458c15fda608654b78d99683ebeab3c60 44456Author: Eric Anholt <eric@anholt.net> 44457Date: Sun Apr 6 07:22:26 2014 -0700 44458 44459 glamor: Drop constant arguments to glamor_solid(). 44460 44461 After keithp's change to drop the old glamor_fill() code, nothing ever 44462 changed these values. 44463 44464 Signed-off-by: Eric Anholt <eric@anholt.net> 44465 Reviewed-by: Keith Packard <keithp@keithp.com> 44466 44467commit db9bff5c389f96991df8d6c62df1174e62e937d3 44468Author: Eric Anholt <eric@anholt.net> 44469Date: Wed Apr 2 12:09:47 2014 -0700 44470 44471 glamor: Drop unnecessary glTexParameteri() in upload of texture data. 44472 44473 We're not drawing, and we're not initially setting up the texture for 44474 later drawing. 44475 44476 Signed-off-by: Eric Anholt <eric@anholt.net> 44477 Reviewed-by: Keith Packard <keithp@keithp.com> 44478 44479commit 1140a89d7762540965ff865d154cb0528049066f 44480Author: Eric Anholt <eric@anholt.net> 44481Date: Wed Apr 2 12:12:06 2014 -0700 44482 44483 glamor: Use MIN/MAX macros to clean up glamor_transfer.c 44484 44485 Signed-off-by: Eric Anholt <eric@anholt.net> 44486 Reviewed-by: Keith Packard <keithp@keithp.com> 44487 44488commit 2ab0fba9df695f19f0b43321d01395736f865c69 44489Author: Eric Anholt <eric@anholt.net> 44490Date: Wed Apr 2 13:41:09 2014 -0700 44491 44492 glamor: Drop unnecessary glTexParameteri() in SetSpans(). 44493 44494 If this path needed the filters set, so would all the other 44495 glDrawArrays() callers. But they don't. 44496 44497 Signed-off-by: Eric Anholt <eric@anholt.net> 44498 Reviewed-by: Keith Packard <keithp@keithp.com> 44499 44500commit e678b4971b666e65a99c141e0a5f33969aa701b7 44501Author: Keith Packard <keithp@keithp.com> 44502Date: Tue Jul 15 17:31:59 2014 -0700 44503 44504 xfree86: Provide backward-compatibilty API for OdevAttributes 44505 44506 This allows drivers to compile using the old OdevAttributes API 44507 against a new server. It generates compiler errors if the caller uses 44508 the wrong or undefined attribute types, or if the caller provides an 44509 incorrect default value for an integer attribute. 44510 44511 Signed-off-by: Keith Packard <keithp@keithp.com> 44512 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 44513 44514commit 60c05ce1ab5d9dc7c034b6b3723f43a42ea637d8 44515Author: Keith Packard <keithp@keithp.com> 44516Date: Tue Jul 15 17:31:58 2014 -0700 44517 44518 config: Replace OdevAttributes linked list with struct 44519 44520 OdevAttributes are a fixed set of values with known types; instead of 44521 storing them in a linked list and requiring accessor/settor functions, 44522 replace the list header, struct OdevAttributes, with a struct that 44523 directly contains the values. This provides for compile-time 44524 typechecking of the values, eliminates a significant amount of code 44525 and generally simplifies using this datatype. 44526 44527 Signed-off-by: Keith Packard <keithp@keithp.com> 44528 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 44529 44530commit 55f5bfb578e934319d1308cbb56c900c5ac7cfa7 44531Author: Keith Packard <keithp@keithp.com> 44532Date: Wed Jul 16 16:03:23 2014 -0700 44533 44534 glamor: Fix temp picture coordinates in glamor_composite_clipped_region 44535 44536 To understand this patch, let's start at the protocol interface where 44537 the relationship between the coordinate spaces is documented: 44538 44539 static Bool 44540 _glamor_composite(CARD8 op, 44541 PicturePtr source, 44542 PicturePtr mask, 44543 PicturePtr dest, 44544 INT16 x_source, 44545 INT16 y_source, 44546 INT16 x_mask, 44547 INT16 y_mask, 44548 INT16 x_dest, INT16 y_dest, 44549 CARD16 width, CARD16 height, Bool fallback) 44550 44551 The coordinates are passed to this function directly off the wire and 44552 are all relative to their respective drawables. For Windows, this means 44553 that they are relative to the upper left corner of the window, in 44554 whatever pixmap that window is getting drawn to. 44555 44556 _glamor_composite calls miComputeCompositeRegion to construct a clipped 44557 region to actually render to. In reality, miComputeCompositeRegion clips 44558 only to the destination these days; source clip region based clipping 44559 would have to respect the transform, which isn't really possible. The 44560 returned region is relative to the screen in which dest lives; offset by 44561 dest->drawable.x and dest->drawable.y. 44562 44563 What is important to realize here is that, because of clipping, the 44564 composite region may not have the same position within the destination 44565 drawable as x_dest, y_dest. The protocol coordinates now exist solely to 44566 'pin' the three objects together. 44567 44568 extents->x1,y1 Screen origin of clipped operation 44569 width,height Extents of the clipped operation 44570 x_dest,y_dest Unclipped destination-relative operation coordinate 44571 x_source,y_source Unclipped source-relative operation coordinate 44572 x_mask,y_mask Unclipped mask-relative operation coordinate 44573 44574 One thing we want to know is what the offset is from the original 44575 operation origin to the clipped origin 44576 44577 Destination drawable relative coordinates of the clipped operation: 44578 44579 x_dest_clipped = extents->x1 - dest->drawable.x 44580 y_dest_clipped = extents->y1 - dest->drawable.y 44581 44582 Offset from the original operation origin: 44583 44584 x_off_clipped = x_dest_clipped - x_dest 44585 y_off_clipped = y_dest_clipped - y_dest 44586 44587 Source drawable relative coordinates of the clipped operation: 44588 44589 x_source_clipped = x_source + x_off_clipped; 44590 y_source_clipped = y_source + y_off_clipped; 44591 44592 Mask drawable relative coordinates of the clipped operation: 44593 44594 x_mask_clipped = x_source + x_off_clipped; 44595 y_mask_clipped = y_source + y_off_clipped; 44596 44597 This is where the original code fails -- it doesn't subtract the 44598 destination drawable location when computing the distance that the 44599 operation has been moved by clipping. Here's what it does when 44600 constructing a temporary source picture: 44601 44602 temp_src = 44603 glamor_convert_gradient_picture(screen, source, 44604 extent->x1 + x_source - x_dest, 44605 extent->y1 + y_source - y_dest, 44606 width, height); 44607 ... 44608 x_temp_src = -extent->x1 + x_dest; 44609 y_temp_src = -extent->y1 + y_dest; 44610 44611 glamor_convert_gradient_picture needs source drawable relative 44612 coordinates, but that is not what it's getting; it's getting 44613 screen-relative coordinates for the destination, adjusted by the 44614 distance between the provided source and destination operation 44615 coordinates. We want x_source_clipped and y_source_clipped: 44616 44617 x_source_clipped = x_source + x_off_clipped 44618 = x_source + x_dest_clipped - x_dest 44619 = x_source + extents->x1 - dest->drawable.x - x_dest 44620 44621 x_temp_src/y_temp_src are supposed to be the coordinates of the original 44622 operation translated to the temporary picture: 44623 44624 x_temp_src = x_source - x_source_clipped; 44625 y_temp_src = y_source - y_source_clipped; 44626 44627 Note that x_source_clipped/y_source_clipped will never be less than 44628 x_source/y_source because all we're doing is clipping. This means that 44629 x_temp_src/y_temp_src will always be non-positive; the original source 44630 coordinate can never be strictly *inside* the temporary image or we 44631 could have made the temporary image smaller. 44632 44633 x_temp_src = x_source - x_source_clipped 44634 = x_source - (x_source + x_off_clipped) 44635 = -x_off_clipped 44636 = x_dest - x_dest_clipped 44637 = x_dest - (extents->x1 - dest->drawable.x) 44638 44639 Again, this is off by the destination origin within the screen 44640 coordinate space. 44641 44642 The code should look like: 44643 44644 temp_src = 44645 glamor_convert_gradient_picture(screen, source, 44646 extent->x1 + x_source - x_dest - dest->pDrawable->x, 44647 extent->y1 + y_source - y_dest - dest->pDrawable->y, 44648 width, height); 44649 44650 x_temp_src = -extent->x1 + x_dest + dest->pDrawable->x; 44651 y_temp_src = -extent->y1 + y_dest + dest->pDrawable->y; 44652 44653 Signed-off-by: Keith Packard <keithp@keithp.com> 44654 Reviewed-by: Markus Wick <markus@selfnet.de> 44655 44656commit 16fbad3c7a22d2fb33549bcd422de985a60982b8 44657Author: Keith Packard <keithp@keithp.com> 44658Date: Thu Jul 17 16:17:53 2014 -0700 44659 44660 Post 1.16 version bump to 1.16.99.1 44661 44662 And we're off towards 1.17; this version bump serves to keep 44663 development versions distinct from stable versions. 44664 44665 Signed-off-by: Keith Packard <keithp@keithp.com> 44666 44667commit bfa5c73a36230f77fb211f185152212541c9d56d 44668Author: Keith Packard <keithp@keithp.com> 44669Date: Wed Jun 4 22:00:45 2014 -0700 44670 44671 ephyr: Free damage structure at server reset time 44672 44673 The usual mechanism for freeing a damage structure when the pixmap is 44674 destroyed does not work for the screen pixmap as it isn't freed in the 44675 normal way. 44676 44677 The existing driver cleanup function, scrfini, is called after the 44678 wrapped CloseScreen functions, including damageCloseScreen, are called 44679 and thus ephyr can't free the damage structure at that point. 44680 44681 Deal with this by providing an early CloseScreen hook in KdCloseScreen 44682 which ephyr can use to free the damage structure before damage itself 44683 shuts down. 44684 44685 Signed-off-by: Keith Packard <keithp@keithp.com> 44686 Reviewed-by: Jamey Sharp <jamey@minilop.net> 44687 44688commit 3319e7041ff89bb01b16a1dbbac85e28b1976ae3 44689Author: Keith Packard <keithp@keithp.com> 44690Date: Fri Apr 18 14:24:29 2014 -0700 44691 44692 hw/xfree86: Let xf86Rotate leave the BlockHandler unwrapped when possible 44693 44694 When no shadow frame buffer is needed, the rotate block handler 44695 doesn't need to be called any more. Remove it from the chain. 44696 44697 Signed-off-by: Keith Packard <keithp@keithp.com> 44698 Reviewed-by: Eric Anholt <eric@anholt.net> 44699 44700commit 79a2733005202af43821d8fd8e4c9fb77bf8f69e 44701Author: Keith Packard <keithp@keithp.com> 44702Date: Fri Apr 18 14:11:17 2014 -0700 44703 44704 hw/xfree86: Fix VGA arbiter screen proc wrapping 44705 44706 Change the screen proc epilog code to re-fetch the current screen 44707 function in case a nested proc changes how things work. This isn't a 44708 problem with the current code as all of the wrapping layers that are 44709 set up at server init time (like the VGA arbiter) leave themselves in 44710 the screen proc chain forever. But, this makes the code conform with 44711 the expected norms. 44712 44713 Signed-off-by: Keith Packard <keithp@keithp.com> 44714 Reviewed-by: Eric Anholt <eric@anholt.net> 44715 44716commit a1189fe322724ab1b524aaad5b700287777252bd 44717Author: Keith Packard <keithp@keithp.com> 44718Date: Fri Apr 18 13:57:55 2014 -0700 44719 44720 mi: Fix block handler wrapping in miSprite 44721 44722 miSpriteBlockHandler was leaving the BlockHandler wrapped until just 44723 before calling any nested block handler. If any code executed before 44724 that added or removed block handlers, the wrapping chain would have 44725 been broken. 44726 44727 Signed-off-by: Keith Packard <keithp@keithp.com> 44728 Reviewed-by: Eric Anholt <eric@anholt.net> 44729 44730commit 08fc33042c858568e7244eb9ad25a8d0270754f0 44731Author: Keith Packard <keithp@keithp.com> 44732Date: Fri Apr 18 13:55:50 2014 -0700 44733 44734 hw/xfree86: Fix block handler wrapping in xf86Rotate 44735 44736 xf86Rotate, it was delaying unwrapping the BlockHandler until after 44737 calling xf86RotateRedisplay. If there was a software cursor on the 44738 screen, the redisplay operation would cause cursor to be removed from 44739 the frame buffer and the misprite block handler to be inserted into 44740 the block handler chain with the misprite screen private saved block 44741 handler now set to xf86RotateBlockHandler. 44742 44743 When xf86RotateRedisplay returned, xf86RotateBlockHandler would then 44744 set screen->BlockHandler to its saved value, call down and then reset 44745 screen->BlockHandler to xf86RotateBlockHandler. miSpriteBlockHandler 44746 would never be called after that, which meant that the software cursor 44747 will now disappear from the screen whenever rendering overlapped and 44748 would only reappear when the cursor was moved. 44749 44750 To correct this, all that is needed is to move the restoration of 44751 screen->BlockHandler to the top of xf86RotateBlockHandler, before the 44752 call to xf86RotateRedisplay. 44753 44754 Signed-off-by: Keith Packard <keithp@keithp.com> 44755 Reviewed-by: Eric Anholt <eric@anholt.net> 44756 44757commit c75fee79ace394f6f51aa6fdda1c0436eb8a2026 44758Author: Keith Packard <keithp@keithp.com> 44759Date: Fri Apr 18 13:54:11 2014 -0700 44760 44761 Document how to correctly wrap screen procedures 44762 44763 This adds a large comment to include/scrnintstr.h which should serve 44764 to document the correct way to wrap any screen procedure, with a 44765 particular focus on how to dynamically add/remove wrapping layers 44766 while the server is running. 44767 44768 Signed-off-by: Keith Packard <keithp@keithp.com> 44769 Reviewed-by: Eric Anholt <eric@anholt.net> 44770 44771commit 5eb77697ea35e7dc8cb8af2c3b5d8ffdba0fb632 44772Author: Peter Harris <pharris@opentext.com> 44773Date: Mon Mar 10 18:31:33 2014 -0400 44774 44775 Avoid starting a comment with */* 44776 44777 Even though -Wcomment doesn't mind it (in gcc or clang), the appearance 44778 of */* confuses the syntax highlighter of some editors (eg. vim), and 44779 causes warnings in MSVC. 44780 44781 Signed-off-by: Peter Harris <pharris@opentext.com> 44782 Reviewed-by: Keith Packard <keithp@keithp.com> 44783 Signed-off-by: Keith Packard <keithp@keithp.com> 44784 44785commit 8b36e1ec8dd9f53e9f4e10422c2100844e9e549c 44786Author: Keith Packard <keithp@keithp.com> 44787Date: Thu Jul 17 00:03:33 2014 -0700 44788 44789 Update to version 1.16.0 44790 44791 Signed-off-by: Keith Packard <keithp@keithp.com> 44792 44793commit 4dbb641bb2d4037f107b58b31e80963dc8b72c0e 44794Author: Hans de Goede <hdegoede@redhat.com> 44795Date: Mon Jul 14 14:01:47 2014 +0200 44796 44797 config_odev_add_attribute*: Check for right attribute type 44798 44799 Don't allow setting string attributes to integers and vice versa. 44800 44801 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 44802 Reviewed-by: Keith Packard <keithp@keithp.com> 44803 Signed-off-by: Keith Packard <keithp@keithp.com> 44804 44805commit 25eca7ce35973577c8d85704c270f7fb53e6732e 44806Author: Hans de Goede <hdegoede@redhat.com> 44807Date: Mon Jul 14 14:01:46 2014 +0200 44808 44809 Fix ODEV_ATTRIB_DRIVER overlapping with ODEV_ATTRIB_FD 44810 44811 Looks like the value of ODEV_ATTRIB_DRIVER was not updated when the patch 44812 adding it got rebased on top of a newer server version. 44813 44814 This fixes the xserver crashing when systemd-logind integration is used. 44815 44816 https://bugzilla.redhat.com/show_bug.cgi?id=1118540 44817 44818 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 44819 Reviewed-by: Keith Packard <keithp@keithp.com> 44820 Signed-off-by: Keith Packard <keithp@keithp.com> 44821 44822commit 9a19bf06b5b409fa0d5b5932e29cd4c5545052c5 44823Author: Peter Hutterer <peter.hutterer@who-t.net> 44824Date: Mon Jul 14 15:05:58 2014 +1000 44825 44826 Revert "dix: fix up coordinate scaling when external monitors are present" 44827 44828 This reverts commit d90b5f83010248be65b2039b0b2d0b9e6a4e93cf. 44829 44830 Reverting for two reasons: 44831 * the scaling does not work on devices that don't advertise resolution, and 44832 the default resolution used (100 units/mm) is higher than most devices, 44833 resulting in a significant slowdown of the touchpads. 44834 * the scaling is still affected by resolution changing. The patch worked 44835 before acceleration but since it maps into resolution-dependent dx/dy 44836 coordinates the acceleration may distort the movement after the fact. So the 44837 same input data generates different movements depending on the resolution. 44838 This can't easily be fixed for all affected devices as synaptics has its own 44839 velocity calculation method whereas wacom doesn't. So anything in the server 44840 won't work for both at the same time. 44841 44842 Revert this for now, until a more integrated solution can be implemented. 44843 44844commit acc0b5edd1dc560b5c39dc44872b46581ec23903 44845Author: Aaron Plattner <aplattner@nvidia.com> 44846Date: Fri Jul 11 15:56:35 2014 -0700 44847 44848 xfree86: Only support one sysconfigdir 44849 44850 When the X server is compiled with --prefix set to something other than /usr, 44851 then it ends up with a nonstandard sysconfigdir in its .pc file. This causes 44852 various other components to install their xorg.conf.d snippets there. 44853 44854 However, the X server first looks for /usr/share/X11/xorg.conf.d before looking 44855 in sysconfigdir. That means that if the system administrator installed anything 44856 that created that path, the user's custom sysconfigdir is not searched. 44857 44858 Rather than doing that, just look in the configured sysconfdir and nowhere else. 44859 44860 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 44861 Reviewed-by: Julien Cristau <jcristau@debian.org> 44862 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 44863 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 44864 44865commit 578b9283bcc129d0a35fabe2637c7622085ef1e8 44866Author: Peter Hutterer <peter.hutterer@who-t.net> 44867Date: Thu Jul 10 10:39:50 2014 +1000 44868 44869 xfree86: don't force the screensaver off on DPMS unblank, merely suggest it 44870 44871 Commit 41d4beb2616ceb3f1a1b8694733e85bae70de59a added symmetry to the 44872 screensaver/DPMS invocations so that one (en|dis)ables the other. Having 44873 dependencies between DPMS and the screensaver is subject to further arguments, 44874 but in this particular case using SCREENSAVER_FORCER is detrimental. 44875 SCREENSAVER_FORCER(ScreenSaverReset) resets the idle time for all 44876 devices on DPMS unblank. 44877 44878 It prevents at least one use-case that GNOME tries to implement: 44879 GNOME displays a notification before suspending. If the display is 44880 currently blanked, GNOME lights it up to display the message. With the 44881 original patch in place DPMS unblank also resets the device idle times, thus 44882 restarting the timeout ad infinitum. 44883 44884 Switch this to a more suggestive SCREENSAVER_OFF(ScreenSaverReset). This keeps 44885 the symmetry in blanking mode (DPMS and screensaver turn each other on/off as 44886 expected) but does not reset the idle time on the devices. 44887 44888 https://bugzilla.gnome.org/show_bug.cgi?id=731241 44889 44890 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 44891 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 44892 Reviewed-By: Egbert Eich <eich@freedesktop.org> 44893 44894commit daa1a9d22db8e83d1933d8403acf72626199ee2a 44895Author: Peter Hutterer <peter.hutterer@who-t.net> 44896Date: Mon Jul 7 08:27:44 2014 +1000 44897 44898 os: prevent negative array index access (#80890) 44899 44900 If an empty string is provided to LogMessageVerbSigSafe, the length of the 44901 printed string is 0. 44902 44903 Read-only access only and the only effect it had was adding a linebreak or not. 44904 44905 X.Org Bug 80890 <http://bugs.freedesktop.org/show_bug.cgi?id=80890> 44906 44907 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 44908 44909commit 9de3cc8daa4c6e877d30a0e8ccfe0cc159f1dbe3 44910Merge: 9308eafb7 a5499870e 44911Author: Keith Packard <keithp@keithp.com> 44912Date: Mon Jul 7 16:53:07 2014 -0700 44913 44914 Merge remote-tracking branch 'jturney/master' 44915 44916commit 9308eafb7d303739b81634ed2ee0da88554fd429 44917Author: Keith Packard <keithp@keithp.com> 44918Date: Mon Jul 7 16:28:26 2014 -0700 44919 44920 Update to version 1.15.99.904 44921 44922 One more RC to get the non-PCI patches tested before release 44923 44924 Signed-off-by: Keith Packard <keithp@keithp.com> 44925 44926commit b6cc489838dca0bcec7e9dbb4663b871e8cb7bc8 44927Author: Thierry Reding <treding@nvidia.com> 44928Date: Thu Feb 13 21:09:49 2014 +0100 44929 44930 xfree86: Make error message more readable 44931 44932 While at it also replace a tab by four spaces for consistency. 44933 44934 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 44935 Tested-By: Aaron Plattner <aplattner@nvidia.com> 44936 Signed-off-by: Thierry Reding <treding@nvidia.com> 44937 Reviewed-by: Rob Clark <robdclark@gmail.com> 44938 Tested-by: Rob Clark <robdclark@gmail.com> 44939 Signed-off-by: Keith Packard <keithp@keithp.com> 44940 44941commit eeefecd9df88920d4dca4100a84a135f7f53dd82 44942Author: Thierry Reding <treding@nvidia.com> 44943Date: Thu Feb 13 13:54:23 2014 +0100 44944 44945 xfree86: Support driver loading via OutputClass 44946 44947 Use the OutputClass configuration to determine what drivers to autoload 44948 for a given device. 44949 44950 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 44951 Tested-By: Aaron Plattner <aplattner@nvidia.com> 44952 Signed-off-by: Thierry Reding <treding@nvidia.com> 44953 Tested-by: Rob Clark <robdclark@gmail.com> 44954 Signed-off-by: Keith Packard <keithp@keithp.com> 44955 44956commit a270bb18baa6e3b45fa4a105a8d2be51ac5270bc 44957Author: Thierry Reding <treding@nvidia.com> 44958Date: Thu Feb 13 13:42:05 2014 +0100 44959 44960 xfree86: Introduce OutputClass configuration 44961 44962 The OutputClass section provides a way to match output devices to a set 44963 of given attributes and configure them. For now, only matching by kernel 44964 driver name is supported. This can be used to determine what DDX module 44965 to load for non-PCI output devices. DDX modules can ship an xorg.conf.d 44966 snippet (e.g. in /usr/share/X11/xorg.conf.d) that looks like this: 44967 44968 Section "OutputClass" 44969 Identifer "NVIDIA Tegra open-source driver" 44970 MatchDriver "tegra" 44971 Driver "opentegra" 44972 EndSection 44973 44974 This will cause any device that's driven by the kernel driver named 44975 "tegra" to use the "opentegra" DDX module. 44976 44977 See the OUTPUTCLASS section in xorg.conf(5) for more details. 44978 44979 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 44980 Tested-By: Aaron Plattner <aplattner@nvidia.com> 44981 Signed-off-by: Thierry Reding <treding@nvidia.com> 44982 Tested-by: Rob Clark <robdclark@gmail.com> 44983 Signed-off-by: Keith Packard <keithp@keithp.com> 44984 44985commit 856bb80cea695106a8f6767d00918b38795b51c6 44986Author: Thierry Reding <treding@nvidia.com> 44987Date: Wed Feb 12 16:43:29 2014 +0100 44988 44989 xfree86: Store kernel driver name in platform device attribute 44990 44991 When opening a DRM device, query the version and store the driver name 44992 as a new attribute for future reference. 44993 44994 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 44995 Tested-By: Aaron Plattner <aplattner@nvidia.com> 44996 Signed-off-by: Thierry Reding <treding@nvidia.com> 44997 Reviewed-by: Rob Clark <robdclark@gmail.com> 44998 Tested-by: Rob Clark <robdclark@gmail.com> 44999 Signed-off-by: Keith Packard <keithp@keithp.com> 45000 45001commit 5a4e15c3f6fb8d674879e54458328e9f595d9451 45002Author: Thierry Reding <treding@nvidia.com> 45003Date: Fri Feb 14 15:45:33 2014 +0100 45004 45005 xfree86: Make driver matching consistent 45006 45007 Most of the driver enumeration functions take an array and a maximum 45008 number of entries that they are allowed to fill in. Upon success, they 45009 return the number of entries filled in. This allows them to be easily 45010 used to consecutively. 45011 45012 One exception is the xf86MatchDriverFromFiles() function, which doesn't 45013 return a value, so callers have to manually search the array for the 45014 first empty entry. 45015 45016 This commit modifies the xf86MatchDriverFromFiles() to behave the same 45017 way as others, which makes it easier to deal with. 45018 45019 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 45020 Tested-By: Aaron Plattner <aplattner@nvidia.com> 45021 Tested-by: Rob Clark <robdclark@gmail.com> (on arm / platform device) 45022 Signed-off-by: Thierry Reding <treding@nvidia.com> 45023 Signed-off-by: Keith Packard <keithp@keithp.com> 45024 45025commit a61ca6f006d70343c88fe45206fae0669d1e8971 45026Author: Tomasz Borowik <timon37@gmail.com> 45027Date: Thu Jul 3 13:04:44 2014 -0700 45028 45029 glamor: Fix stack corruption in glamor_init 45030 45031 glGet on GL_MAX_VIEWPORT_DIMS returns two values 45032 45033 Reviewed-by: Markus Wick <markus@selfnet.de> 45034 Signed-off-by: Keith Packard <keithp@keithp.com> 45035 45036commit 10d2805dbc6b96a159b8c5acedcd53f34df362bf 45037Author: Keith Packard <keithp@keithp.com> 45038Date: Thu Jun 26 14:12:24 2014 -0700 45039 45040 fb: Don't free NULL pixmap in fbCloseScreen. Bug #80313 45041 45042 We fixed fbCloseScreen to use the FreePixmap function so that the 45043 private counts would be updated correctly during CloseScreen. Xvfb 45044 calls FreePixmap and sets devPrivate to NULL before fbCloseScreen is 45045 called; not checking devPrivate before calling would result in a NULL 45046 pointer dereference. 45047 45048 Signed-off-by: Keith Packard <keithp@keithp.com> 45049 Reviewed-by: Julien Cristau <jcristau@debian.org> 45050 45051commit a5499870e2f88822f52b1b54cad0db69856597c0 45052Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 45053Date: Sun Apr 6 04:47:15 2014 -0500 45054 45055 hw/xwin: link dynamically and export symbols 45056 45057 With my patch to fix shared libXfont to work correctly on Cygwin/Win32, 45058 there is no need for -static anymore. But, XWin.exe must export its 45059 symbols in order for them to override libXfont's stubs. 45060 45061 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 45062 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 45063 45064commit 5f3485b06e180674af640e7e9d7d56f67b61735b 45065Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 45066Date: Thu Jun 26 10:37:19 2014 +0200 45067 45068 bump to 0.9.0 45069 45070commit a3b44ad8db1fa2f3b81c1ff9498f31c5323edd37 45071Author: Thierry Reding <treding@nvidia.com> 45072Date: Wed Jun 25 16:03:29 2014 +0200 45073 45074 exa: Fix a warning when enabling DEBUG_TRACE_FALL 45075 45076 The format string wants a picture and a character, but the argument list 45077 contains only a character, causing GCC to complain. Add the missing 45078 argument. 45079 45080 Signed-off-by: Thierry Reding <treding@nvidia.com> 45081 Reviewed-by: Keith Packard <keithp@keithp.com> 45082 Signed-off-by: Keith Packard <keithp@keithp.com> 45083 45084commit 2f113d68f6c1572576bc57ecca12e44cc9e438eb 45085Author: Kristian Høgsberg <krh@bitplanet.net> 45086Date: Mon Jun 16 11:34:55 2014 -0700 45087 45088 xwayland: Add glamor and DRI3 support 45089 45090 Reviewed-by: Axel Davy <axel.davy@ens.fr> 45091 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 45092 Signed-off-by: Keith Packard <keithp@keithp.com> 45093 45094commit fd16555c2fc606fc43236050deba558c20e184e8 45095Author: Michel Dänzer <michel.daenzer@amd.com> 45096Date: Tue Jun 24 16:09:19 2014 +0900 45097 45098 Revert "glamor: Fix coordinates handling for composite source/mask pictures" 45099 45100 This reverts commit 4e9aabb6fc15d8052934f20c6a07801c197ec36a. 45101 45102 It broke kwin decorations with XRender compositing. 45103 45104 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 45105 Reviewed-by: Eric Anholt <eric@anholt.net> 45106 Signed-off-by: Keith Packard <keithp@keithp.com> 45107 45108commit 18744907d0766b1b57be12df5adafd0f93221006 45109Author: Chris Wilson <chris@chris-wilson.co.uk> 45110Date: Tue Jun 10 07:45:05 2014 +0100 45111 45112 dri2: Invalidate DRI2Buffers upon SetWindowPixmap updates 45113 45114 When transitioning to a redirected or unredirected Window, the Composite 45115 layer modifies the Window's Pixmap. However, the DRI2Buffer for the 45116 Drawable is still pointing to the backing bo of the old Pixmap with the 45117 result that rendering goes astray. 45118 45119 This now also effects DRI2 Drawables that are touched by PresentPixmap. 45120 45121 v2: Fixup the function name after rebasing 45122 45123 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 45124 Cc: Reinis Danne <reinis.danne@gmail.com> 45125 Reviewed-by: Dave Airlie <airlied@redhat.com> 45126 Cc: Keith Packard <keithp@keithp.com> 45127 Signed-off-by: Keith Packard <keithp@keithp.com> 45128 45129commit 4d92fab39c4225e89f2d157a1f559cb0618a6eaa 45130Author: Chris Wilson <chris@chris-wilson.co.uk> 45131Date: Wed Jun 18 11:14:43 2014 +0100 45132 45133 dri2: Use the PrimeScreen when creating/reusing buffers 45134 45135 This fixes a segfault when we attempt to call ds->ReuseBufferNotify() 45136 passing a Prime DRI2BufferPtr to the master backend. 45137 45138 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=80001 45139 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 45140 Reviewed-by: Dave Airlie <airlied@redhat.com> 45141 Signed-off-by: Keith Packard <keithp@keithp.com> 45142 45143commit 0d9ad781807022d78239327ed508758a3f88b090 45144Author: Eric Anholt <eric@anholt.net> 45145Date: Mon Jun 16 08:06:53 2014 +0100 45146 45147 glamor: Fix GLES2 non-VBO temporary memory allocation. 45148 45149 We'd get a request for like 16 bytes, claim to have allocated 45150 GLAMOR_VBO_SIZE, and then not reallocate when something a request 45151 bigger than 16 came along. The intent was to always allocate at least 45152 GLAMOR_VBO_SIZE. 45153 45154 Fixes segfaults with Xephyr -glamor_gles2 and running gnome-terminal. 45155 45156 Signed-off-by: Eric Anholt <eric@anholt.net> 45157 Reviewed-by: Keith Packard <keithp@keithp.com> 45158 Signed-off-by: Keith Packard <keithp@keithp.com> 45159 45160commit 736bed2aaf4bb672b6c7b33146fe84700baf22bd 45161Merge: ce581ac3f d90b5f830 45162Author: Keith Packard <keithp@keithp.com> 45163Date: Mon Jun 23 14:18:03 2014 -0700 45164 45165 Merge remote-tracking branch 'whot/for-keith' 45166 45167commit ce581ac3fa80c4c52d87bd54fa92bc566b7d6adc 45168Author: Axel Davy <axel.davy@ens.fr> 45169Date: Wed Jun 18 22:22:09 2014 -0400 45170 45171 present: fix bad logic in cancelling scheduled operations. 45172 45173 If we present several pixmaps in advance for different msc, the later one 45174 shouldn't cancel the previous ones. 45175 45176 This reverts a change made by commit 45177 e6f5d9d7b7efdacea0f22f1808efca849bcede4c 45178 45179 Without this fix, vblank_mode=0 glxgears doesn't update 45180 with the present fallback. 45181 45182 Signed-off-by: Axel Davy <axel.davy@ens.fr> 45183 Reviewed-by: Keith Packard <keithp@keithp.com> 45184 Signed-off-by: Keith Packard <keithp@keithp.com> 45185 45186commit adb7bc3386559dfee34b359dadcbb6796bc416e7 45187Author: Dinar Valeev <dvaleev@suse.com> 45188Date: Mon Feb 24 11:36:54 2014 +0100 45189 45190 arch: Fix image and bitmap byte order for ppc64le 45191 45192 So far PPC was big endian for sure. For ppc64le this is no longer 45193 true. 45194 45195 Signed-off-by: Egbert Eich <eich@freedesktop.org> 45196 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 45197 Signed-off-by: Keith Packard <keithp@keithp.com> 45198 45199commit ea47341792cf06eda5bbdf0aca48aacc024a5be0 45200Merge: 334faabe6 a11bbd875 45201Author: Keith Packard <keithp@keithp.com> 45202Date: Mon Jun 23 13:50:54 2014 -0700 45203 45204 Merge remote-tracking branch 'anholt/glamor-fixes' 45205 45206commit 334faabe682a422075ba214501c7554dd5ee5563 45207Author: Frank Binns <frank.binns@imgtec.com> 45208Date: Tue Jun 10 13:43:31 2014 +0100 45209 45210 present: restore screen pixmap when aborting a flip 45211 45212 If a 2D application is started on top of a fullscreen 3D application, which 45213 is flipping, then we need to stop flipping and restore the root window, and 45214 possibly the flip window, to using the screen pixmap. Normally this would 45215 be done as part of an unflip. However, in the case that there is a pending 45216 flip there is no mechanism to abort so the unflip is deferred until the 45217 pending flip completes. This provides a window of opportunity for the 2D 45218 application to draw to the wrong pixmap. 45219 45220 Restore the screen pixmap at the point a pending flip is marked as aborted, 45221 thus avoiding this issue. 45222 45223 Reviewed-by: Keith Packard <keithp@keithp.com> 45224 Signed-off-by: Frank Binns <frank.binns@imgtec.com> 45225 Signed-off-by: Keith Packard <keithp@keithp.com> 45226 45227commit d90b5f83010248be65b2039b0b2d0b9e6a4e93cf 45228Author: Peter Hutterer <peter.hutterer@who-t.net> 45229Date: Fri May 30 09:56:37 2014 +1000 45230 45231 dix: fix up coordinate scaling when external monitors are present 45232 45233 The goal of all this is to get an x/y motion reflecting the motion 45234 on the device, i.e. a circle on the device is a circle on the screen. 45235 45236 This is currently done by scaling the y coordinate depending on the screen 45237 ratio vs device ratio. Depending on that ratio the movement on the y axis may 45238 be accelerated (ratio < 1) or slowed (ratio > 1). This leads to the weird 45239 effect that changing the screen ratio by plugging a new monitor changes the 45240 speed of the touchpad. 45241 45242 Use a different algorithm: calculate the physical movement on the device, map 45243 that to the same-ish distance on the screen, then convert that back into a 45244 device-specific vector. This way we get the same mapping regardless of the 45245 current screen dimensions. 45246 45247 Since the pointer accel code doesn't take device resolution into account, make 45248 sure we apply our crazy mapping before we accelerate. This way we accelerate 45249 resolution-independent. 45250 45251 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 45252 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 45253 45254commit 382ff4a306b97b0ddcdac03ce8611b026ca5323b 45255Author: Keith Packard <keithp@keithp.com> 45256Date: Tue Jun 17 15:50:48 2014 -0700 45257 45258 present: Remove executing vblank from window list. Bug# 79709. 45259 45260 Once the vblank is actually getting executed, it's lifetime is no 45261 longer tied to the window, and so it shouldn't be controlled by window 45262 destruction. In particular, if the vblank is queued for flip, it will 45263 get stored in the flip_pending field, and will be correctly destroyed 45264 when the flip completes. 45265 45266 Signed-off-by: Keith Packard <keithp@keithp.com> 45267 45268commit 199d9a6a942af9b443616f6ef6d26052ef3f48b9 45269Author: Eric Anholt <eric@anholt.net> 45270Date: Fri Dec 27 11:12:27 2013 -0800 45271 45272 xephyr: Allow initializing glamor with gles2 (on GLX). 45273 45274 This should be useful for glamor development, so you can test both 45275 paths (which are significantly different, and apparently 45276 glamor_gradient.c was broken on GLES2 as of the import). 45277 45278 Signed-off-by: Eric Anholt <eric@anholt.net> 45279 Reviewed-by: Keith Packard <keithp@keithp.com> 45280 Reviewed-by: Adam Jackson <ajax@redhat.com> 45281 45282commit 98b6158bc1e32aaca375829452266e013a520e14 45283Author: Eric Anholt <eric@anholt.net> 45284Date: Sun Apr 6 07:44:20 2014 +0100 45285 45286 glamor: Share code for put_image handling. 45287 45288 The difference between the two is that XF86 has the clip helper that 45289 lets you upload less data when rendering video that's clipped. I 45290 don't think that's really worth the trouble, especially in a world of 45291 compositors, so I've dropped it to get to shared code. 45292 45293 It turns out the clipping code was broken on xf86-video-intel anyway. 45294 To reproduce, run without a compositor, and use another window to clip 45295 the top half of your XV output on the glamor XV adaptor: the rendering 45296 got confused about which half of the window was being drawn to. 45297 45298 Signed-off-by: Eric Anholt <eric@anholt.net> 45299 Reviewed-by: Adam Jackson <ajax@redhat.com> 45300 45301commit 34884e16bf7c97434e7883d025c6814e083b0def 45302Author: Eric Anholt <eric@anholt.net> 45303Date: Sat Apr 5 12:30:28 2014 +0100 45304 45305 ephyr: Add support for XV using glamor. 45306 45307 Signed-off-by: Eric Anholt <eric@anholt.net> 45308 Reviewed-by: Adam Jackson <ajax@redhat.com> 45309 45310commit 23d303bf905e76a70bda942037bdfbdcd06e55d2 45311Author: Eric Anholt <eric@anholt.net> 45312Date: Sat Apr 5 12:57:16 2014 +0100 45313 45314 kdrive: Mark XV names const to avoid warnings. 45315 45316 No code modifies it at runtime, and it's common to store string 45317 literals to it. 45318 45319 Signed-off-by: Eric Anholt <eric@anholt.net> 45320 Reviewed-by: Adam Jackson <ajax@redhat.com> 45321 45322commit 65efc14b6ae1ee73bf6db379d7826b6bc9fd6d33 45323Author: Eric Anholt <eric@anholt.net> 45324Date: Sat Apr 5 11:50:51 2014 +0100 45325 45326 glamor: Split the XV code into XF86-dependent parts and generic. 45327 45328 I want to expose this from Xephyr as well, both to be able to test XV 45329 changes rapidly, and beause the XV passthrough to the host's overlay 45330 really doesn't work out well when we glXSwapBuffers() over the 45331 colorkey. 45332 45333 Signed-off-by: Eric Anholt <eric@anholt.net> 45334 Reviewed-by: Adam Jackson <ajax@redhat.com> 45335 45336commit a5662193f1c3dd63f615d96d1a300f70086ccbc6 45337Author: Eric Anholt <eric@anholt.net> 45338Date: Sat Apr 5 12:26:33 2014 +0100 45339 45340 kdrive: Simplify the adaptor setup interface. 45341 45342 Now that we don't have to worry about the generic adaptors code, 45343 there's no need to have a list of pointers to different sets of 45344 adaptors. 45345 45346 Signed-off-by: Eric Anholt <eric@anholt.net> 45347 Reviewed-by: Adam Jackson <ajax@redhat.com> 45348 45349commit 55aad7399d4470bc46c064aafe07d12a6c293982 45350Author: Eric Anholt <eric@anholt.net> 45351Date: Fri Apr 4 12:09:01 2014 +0100 45352 45353 kdrive: Remove dead generic XV adaptors code. 45354 45355 I couldn't find any callers in the history of the tree. 45356 45357 Signed-off-by: Eric Anholt <eric@anholt.net> 45358 Reviewed-by: Adam Jackson <ajax@redhat.com> 45359 45360commit 0edc0a78fbfac4578b0f809aef17332c1eb461e3 45361Author: Eric Anholt <eric@anholt.net> 45362Date: Sun Apr 6 08:12:35 2014 +0100 45363 45364 kdrive: Do a little more cleanup from the XV struct deduplication. 45365 45366 Signed-off-by: Eric Anholt <eric@anholt.net> 45367 Reviewed-by: Adam Jackson <ajax@redhat.com> 45368 45369commit 1d90e8811a7d4db328c0c944bec0aa3ed6afb70d 45370Author: Eric Anholt <eric@anholt.net> 45371Date: Fri Dec 27 21:34:44 2013 -0800 45372 45373 xorg: Remove duplicated definitions of some XV-related structs. 45374 45375 These were field-for-field identical, so we can just typedef them to 45376 be the same, and memcpy their contents. 45377 45378 v2: Fix missed strdup(). 45379 45380 Signed-off-by: Eric Anholt <eric@anholt.net> 45381 Reviewed-by: Adam Jackson <ajax@redhat.com> 45382 45383commit a6aaa51752f301de24abce264976ba3c3a50863c 45384Author: Keith Packard <keithp@keithp.com> 45385Date: Fri Mar 21 14:55:47 2014 -0700 45386 45387 glamor: Remove stubbed-out glamor_stipple function 45388 45389 This function isn't used anymore. 45390 45391 Signed-off-by: Keith Packard <keithp@keithp.com> 45392 Reviewed-by: Eric Anholt <eric@anholt.net> 45393 45394commit ef2bf0e645ed8242a0b637ed6a9d5b8c03b6b481 45395Author: Keith Packard <keithp@keithp.com> 45396Date: Fri Mar 21 14:30:33 2014 -0700 45397 45398 glamor: Remove 'tiling' shader code 45399 45400 The core rendering paths all use the glamor_program fill functions now 45401 45402 Signed-off-by: Keith Packard <keithp@keithp.com> 45403 Reviewed-by: Eric Anholt <eric@anholt.net> 45404 45405commit 18c09e60bf16b28060ade5d24110f2aa6bc19b57 45406Author: Keith Packard <keithp@keithp.com> 45407Date: Sun Mar 16 20:49:28 2014 -0700 45408 45409 glamor: Replace glamor_solid_boxes and glamor_solid with GC using code 45410 45411 This provides glamor_solid_boxes and glamor_solid using regular GC 45412 operations instead of calling directly to underlying rendering 45413 functions. This will allow the old rendering code to be removed. 45414 45415 Signed-off-by: Keith Packard <keithp@keithp.com> 45416 Reviewed-by: Eric Anholt <eric@anholt.net> 45417 45418commit bd3b2c48f69a5169aefb261c041462271c69a07a 45419Author: Keith Packard <keithp@keithp.com> 45420Date: Thu Apr 3 14:22:52 2014 -0700 45421 45422 glamor: Add accelerated stipple support 45423 45424 This copies the stipple to a 8bpp pixmap and uses that to paint the 45425 texture from. 45426 45427 v2: Create deep stipple pixmap without GLAMOR_CREATE_FBO_NO_FBO 45428 45429 v3: Fix stipple origin sign (matches tiles now). Track changes 45430 to original stipple with damage. This isn't required by the 45431 X spec, but java appears to depend on it, so we'll just do it. 45432 When Glamor switches to 8bpp bitmaps, we'll be able to render 45433 directly from them and not need this anymore. 45434 45435 v4: Review comments from Eric: 45436 45437 * Remove stray whitespace change 45438 * Avoid "large" pixmap for stipple by using GLAMOR_CREATE_NO_LARGE 45439 * Wrap to 80 columns 45440 45441 v5: Don't crash when stipple damage tracker is destroyed 45442 45443 The stipple damage tracker is automatically destroyed when the 45444 associated stipple pixmap is destroyed. When this happens, just 45445 clear the pointer from the GC rather than calling 45446 glamor_invalidate_stipple; that function would call 45447 DamageUnregister on the now invalid stipple damage pointer and 45448 crash. 45449 45450 Signed-off-by: Keith Packard <keithp@keithp.com> 45451 Reviewed-by: Eric Anholt <eric@anholt.net> 45452 45453commit d18f5801c9a632dd4d9f8b7912491b6623e943d5 45454Author: Keith Packard <keithp@keithp.com> 45455Date: Wed Apr 2 14:07:20 2014 -0700 45456 45457 glamor: Add glamor_program based 0-width dashed lines 45458 45459 This makes sure the pixelization for dashed lines matches non-dashed 45460 lines, while also speeding them up. 45461 45462 v2: Switch to glamor_make_current 45463 45464 v3: Create dash pattern pixmap without GLAMOR_CREATE_FBO_NO_FBO 45465 45466 v4: Adopt suggestions from Eric's review: 45467 45468 - Drops power-of-two alignment of our line vertex data, simplifying 45469 the code. 45470 45471 - Stops reading from the VBO. While on keithp's and my machines the 45472 VBO is mapped cached, on many implementations it will be mapped WC, 45473 making those reads extremely expensive. 45474 45475 - Style fixes (line wrapping, spaces around operators). 45476 45477 v5: Adopt suggestions from Markus' review: 45478 45479 - Use max when computing zero-width dashed line length. 45480 45481 Don't open code max here. 45482 45483 - Embed CoordModePrevious into VBO writing for dashed lines 45484 45485 Instead of pre-computing the coord mode previous results, just 45486 embed this in the loop which fills the vertex buffer. Saves 45487 re-writing the request buffer, and shortens the code a bit 45488 45489 v6: Export glamor_destroy_gc for UXA 45490 45491 UXA needs to call glamor_destroy_gc from its GCFuncs, so export 45492 it. 45493 45494 Signed-off-by: Keith Packard <keithp@keithp.com> 45495 Reviewed-by: Eric Anholt <eric@anholt.net> 45496 45497commit dc9fa9080a1cb994b4e54a341d2245f442dac576 45498Author: Keith Packard <keithp@keithp.com> 45499Date: Sun Mar 23 20:59:02 2014 -0700 45500 45501 glamor: Use glamor_program and GL_LINES for 0-width lines 45502 45503 GL lines are nearly X compliant; you just need to fill in the last 45504 pixel when the client hasn't requested CapNotLast. 45505 45506 v2: switch to glamor_make_current 45507 45508 v3: use miPolylines instead of custom glamor fallback path. Wrap 45509 code to 80 columns. 45510 45511 Signed-off-by: Keith Packard <keithp@keithp.com> 45512 Reviewed-by: Eric Anholt <eric@anholt.net> 45513 45514commit 51075ebd37dca8d17c42425fb756ad3090e157c4 45515Author: Keith Packard <keithp@keithp.com> 45516Date: Fri Mar 21 18:03:07 2014 -0700 45517 45518 glamor: Use glamor_program for glamor_push_pixels 45519 45520 This uses the same shaders as glamor_poly_glyph_blt. 45521 45522 v2: Wrap some long lines (changes by anholt). 45523 45524 Signed-off-by: Keith Packard <keithp@keithp.com> 45525 Reviewed-by: Eric Anholt <eric@anholt.net> 45526 45527commit 45ebc4e3fac7f1a85167d05e2833949b89f02d64 45528Author: Keith Packard <keithp@keithp.com> 45529Date: Sat Mar 15 13:31:18 2014 -0700 45530 45531 glamor: Add glamor_program based copy acceleration 45532 45533 Paints with textures, using a temporary buffer for overlapping copies 45534 45535 Performs CPU to GPU transfers for pixmaps in memory. Accelerates copy 45536 plane when both objects are in the GPU. Includes copy_window 45537 acceleration too. 45538 45539 v2: Use NV_texture_barrier for non-overlapping copies within the same 45540 drawable 45541 45542 v3: Switch to glamor_make_current 45543 45544 v4: Do overlap check on the bounding box of the region rather than 45545 on individual boxes 45546 45547 v5: Use Eric Anholt's re-written comments which provide a more accurate 45548 description of the code 45549 45550 v6: Use floating point uniform for copy plane bit multiplier. This 45551 avoids an int to float conversion in the copy plane fragment shader. 45552 45553 Use round() instead of adding 0.5 in copy plane. round() and +0.5 45554 end up generating equivalent code, and performance measurements 45555 confirm that they are the same speed. Round() is a bit clearer 45556 though, so we'll use it. 45557 45558 Signed-off-by: Keith Packard <keithp@keithp.com> 45559 Reviewed-by: Eric Anholt <eric@anholt.net> 45560 Reviewed-by: Markus Wick <markus@selfnet.de> 45561 45562commit 0e08a79599c773f77c0667d557376a5ccee3f89c 45563Author: Keith Packard <keithp@keithp.com> 45564Date: Wed Apr 2 14:05:35 2014 -0700 45565 45566 glamor: Directly reference the private key records 45567 45568 There's no reason to use a pointer here, it just wastes time. 45569 45570 Signed-off-by: Keith Packard <keithp@keithp.com> 45571 Reviewed-by: Eric Anholt <eric@anholt.net> 45572 45573commit 15e4d14dfae054c026b7e965ac33985e5cf6a168 45574Author: Keith Packard <keithp@keithp.com> 45575Date: Sat Mar 15 13:27:14 2014 -0700 45576 45577 glamor: Replace fallback preparation code 45578 45579 These offer a simpler and more efficient means for temporarily 45580 transitioning to CPU-accessible memory for fallback implementations. 45581 45582 v2: Do not attempt fallbacks with GLAMOR_DRM_ONLY pixmaps 45583 45584 glamor cannot transfer pixels for GLAMOR_DRM_ONLY pixmaps using 45585 glReadPixels and glTexSubImage2D, and so there's no way to perform 45586 fallback operations with these pixmaps. 45587 45588 v3: Clear ->pbo field when deleting the PBO. Otherwise, we'd reuse 45589 the old name next time we fall back on the pixmap, which would 45590 potentially conflict with some other pixmap that genned a new 45591 name, or just do a lazy allocation of the name (compat GL context, 45592 like we currently use) or error out (core GL context, like we hope 45593 to use some day). Also, style fixes. Changes by anholt, acked by 45594 keithp. 45595 45596 Signed-off-by: Keith Packard <keithp@keithp.com> 45597 Reviewed-by: Eric Anholt <eric@anholt.net> 45598 45599commit 3ac481c9dac591ffc33812ab047fc793d8a43d4f 45600Author: Keith Packard <keithp@keithp.com> 45601Date: Wed May 7 10:31:16 2014 -0700 45602 45603 mi: Draw multiple lines in one FillSpans call in miZeroLine 45604 45605 miZeroLine allocates enough space to draw a line spanning the entire 45606 width/height of the target drawable. When drawing multiple shorter 45607 lines, this leaves most of the space in that buffer unfilled. Let 45608 multiple lines be drawn into the buffer if there is plenty of space. 45609 45610 Speeds up glamor fallback zero-width lines: 45611 45612 Before 45613 6000000 trep @ 0.0020 msec (508000.0/sec): 1-pixel line 45614 6000000 trep @ 0.0020 msec (492000.0/sec): 10-pixel line 45615 6000000 trep @ 0.0023 msec (427000.0/sec): 100-pixel line 45616 4000000 trep @ 0.0035 msec (282000.0/sec): 500-pixel line 45617 45618 After: 45619 600000000 trep @ 0.0000 msec (43400000.0/sec): 1-pixel line 45620 140000000 trep @ 0.0001 msec (13000000.0/sec): 10-pixel line 45621 16000000 trep @ 0.0008 msec (1300000.0/sec): 100-pixel line 45622 4000000 trep @ 0.0038 msec (261000.0/sec): 500-pixel line 45623 45624 (500 pixel lines do not change in performance because the buffer can 45625 only one one of them.) 45626 45627 Signed-off-by: Keith Packard <keithp@keithp.com> 45628 Reviewed-by: Eric Anholt <eric@anholt.net> 45629 45630commit ea678a73c5688f73071d5581b6406808b7a0230f 45631Author: Keith Packard <keithp@keithp.com> 45632Date: Fri Apr 25 22:43:51 2014 -0700 45633 45634 mi: Fill spans for multiple arcs in miPolyFillArc 45635 45636 This allocates span data for multiple arcs and draws the 45637 whole set in one call, rather than doing them one at a time. For 45638 modern hardware, this is a significant performance improvement. 45639 45640 v2: Limit the number of spans per buffer to 4M to avoid 45641 integer overflow in computing the malloc size. 45642 45643 Signed-off-by: Keith Packard <keithp@keithp.com> 45644 Reviewed-by: Eric Anholt <eric@anholt.net> 45645 45646commit a7fce36affb8211990e5b4956adea4d75f0e73c9 45647Author: Keith Packard <keithp@keithp.com> 45648Date: Wed May 7 09:58:26 2014 -0700 45649 45650 mi: Make miPolyArc draw fast zero-width when possible 45651 45652 Instead of forcing drivers to figure out when to call miZeroPolyArc, 45653 have miPolyArc call that when possible. 45654 45655 This involved renaming the existing miPolyArc call to miWideArc and 45656 creating a new miPolyArc wrapper function as miZeroPolyArc falls back 45657 to miWideArc when the arc is too large to be drawn with the zero-width 45658 code (ellipses larger than 800x800). 45659 45660 Signed-off-by: Keith Packard <keithp@keithp.com> 45661 Reviewed-by: Eric Anholt <eric@anholt.net> 45662 45663commit bf1429b2034a577c994ff16b60d809d05ca99241 45664Author: Keith Packard <keithp@keithp.com> 45665Date: Wed May 7 09:56:39 2014 -0700 45666 45667 mi: Create miPolylines as a general-purpose line drawing function 45668 45669 Instead of requiring all drivers to figure out which mi function to 45670 call for each of the four cases, create a single wrapper in mi that 45671 handles them correctly. Now drivers can simply use miPolylines in all 45672 cases. 45673 45674 Signed-off-by: Keith Packard <keithp@keithp.com> 45675 Reviewed-by: Eric Anholt <eric@anholt.net> 45676 45677commit 14d82a2bc3179160803c62c20746630d14e1b7d7 45678Author: Keith Packard <keithp@keithp.com> 45679Date: Thu Mar 13 23:29:54 2014 -0700 45680 45681 ephyr: Deal with non-root visual for window 45682 45683 glx will sometimes select a non-root visual, deal with that by 45684 creating a suitable colormap and using that instead of attempting to 45685 use the default colormap. 45686 45687 Signed-off-by: Keith Packard <keithp@keithp.com> 45688 Reviewed-by: Eric Anholt <eric@anholt.net> 45689 45690commit a11bbd875f3f90a3d02d727778cb1d3524cf59fd 45691Author: Eric Anholt <eric@anholt.net> 45692Date: Fri May 30 10:39:30 2014 -0700 45693 45694 glamor: Don't leak a prepare_access_gc() in putimage fallbacks. 45695 45696 It turns out putimage doesn't use the GC tile or stipple anyway, so 45697 there's no need to do this. 45698 45699 Signed-off-by: Eric Anholt <eric@anholt.net> 45700 Reviewed-by: Keith Packard <keithp@keithp.com> 45701 45702commit 8da1e4e2bf28c3610cdbe1770a57be89578d37f5 45703Author: Markus Wick <markus@selfnet.de> 45704Date: Wed May 14 23:08:18 2014 +0200 45705 45706 glamor: Choose max fbo size by texture + viewport size 45707 45708 The max size of renderbuffers and texture often match by accident, but 45709 as we always use textures, we should check for the right flag. Also 45710 check for viewport size as this may be lower and we want to render to 45711 almost every pixmap. 45712 45713 Signed-off-by: Markus Wick <markus@selfnet.de> 45714 Signed-off-by: Eric Anholt <eric@anholt.net> 45715 Reviewed-by: Eric Anholt <eric@anholt.net> 45716 45717commit 3778fab34bc00334edec4f845d0c6d43440f265a 45718Author: Markus Wick <markus@selfnet.de> 45719Date: Wed May 14 23:08:17 2014 +0200 45720 45721 glamor: Fix no-mipmap allocations 45722 45723 With GL_TEXTURE_MIN_FILTER, we configure not to use mipmaps, but 45724 there's no real way until GL_ARB_texture_storage to dictate whether 45725 memory should be allocated for mipmap levels or not. 45726 45727 GL_TEXTURE_MAX_LEVEL is a stronger hint to the driver than the 45728 filtering that we really don't want mipmap allocations. Stops VARM 45729 wasting warnings from the nvidia driver. 45730 45731 Signed-off-by: Markus Wick <markus@selfnet.de> 45732 Signed-off-by: Eric Anholt <eric@anholt.net> 45733 Reviewed-by: Eric Anholt <eric@anholt.net> 45734 Reviewed-by: Keith Packard <keithp@keithp.com> 45735 45736commit 1095c262502332bc570a9b288b01da51c1fa10ec 45737Author: Alan Coopersmith <alan.coopersmith@oracle.com> 45738Date: Fri Jun 6 17:54:22 2014 -0700 45739 45740 config: show default path in help for --with-xkb-bin-directory 45741 45742 Now shows: 45743 --with-xkb-bin-directory=DIR 45744 Directory containing xkbcomp program (default: 45745 ${bindir}) 45746 45747 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 45748 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 45749 Signed-off-by: Keith Packard <keithp@keithp.com> 45750 45751commit e27a839bf0488d5b1cc2e2a887f2ea0e3d790790 45752Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 45753Date: Thu Jun 5 20:38:44 2014 -0700 45754 45755 mi: Build fix: mieqProcessDeviceEvent returns void 45756 45757 mieq.c:520:9: error: void function 'mieqProcessDeviceEvent' should not return a value [-Wreturn-type,Semantic Issue] 45758 return 0; 45759 ^ ~ 45760 1 error generated. 45761 45762 Regression-from: 9fb08310b51b46736f3ca8dbc04efdf502420403 45763 Found-by: Tinderbox 45764 45765 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 45766 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 45767 45768commit 08820f03768e904f1746ef93ca01660330dc5406 45769Author: Keith Packard <keithp@keithp.com> 45770Date: Wed Jun 4 22:28:57 2014 -0700 45771 45772 Update to version 1.15.99.903 45773 45774 Signed-off-by: Keith Packard <keithp@keithp.com> 45775 45776commit 9fb08310b51b46736f3ca8dbc04efdf502420403 45777Author: Peter Hutterer <peter.hutterer@who-t.net> 45778Date: Wed May 21 10:07:31 2014 +1000 45779 45780 mi: don't process events from disabled devices (#77884) 45781 45782 Once a device is disabled, it doesn't have a sprite pointer anymore. If an 45783 event is still in the queue and processed after DisableDevice finished, a 45784 dereference causes a crash. Example backtrace (crash forced by injecting an 45785 event at the right time): 45786 45787 (EE) 0: /opt/xorg/bin/Xorg (OsSigHandler+0x3c) [0x48d334] 45788 (EE) 1: /lib64/libpthread.so.0 (__restore_rt+0x0) [0x37fcc0f74f] 45789 (EE) 2: /opt/xorg/bin/Xorg (mieqMoveToNewScreen+0x38) [0x609240] 45790 (EE) 3: /opt/xorg/bin/Xorg (mieqProcessDeviceEvent+0xd4) [0x609389] 45791 (EE) 4: /opt/xorg/bin/Xorg (mieqProcessInputEvents+0x206) [0x609720] 45792 (EE) 5: /opt/xorg/bin/Xorg (ProcessInputEvents+0xd) [0x4aeb58] 45793 (EE) 6: /opt/xorg/bin/Xorg (xf86VTSwitch+0x1a6) [0x4af457] 45794 (EE) 7: /opt/xorg/bin/Xorg (xf86Wakeup+0x2bf) [0x4af0a7] 45795 (EE) 8: /opt/xorg/bin/Xorg (WakeupHandler+0x83) [0x4445cb] 45796 (EE) 9: /opt/xorg/bin/Xorg (WaitForSomething+0x3fe) [0x491bf6] 45797 (EE) 10: /opt/xorg/bin/Xorg (Dispatch+0x97) [0x435748] 45798 (EE) 11: /opt/xorg/bin/Xorg (dix_main+0x61d) [0x4438a9] 45799 (EE) 12: /opt/xorg/bin/Xorg (main+0x28) [0x49ba28] 45800 (EE) 13: /lib64/libc.so.6 (__libc_start_main+0xf5) [0x37fc821d65] 45801 (EE) 14: /opt/xorg/bin/Xorg (_start+0x29) [0x425e69] 45802 (EE) 15: ? (?+0x29) [0x29] 45803 45804 xf86VTSwitch() calls ProcessInputEvents() before disabling a device, and 45805 DisableDevice() calls mieqProcessInputEvents() again when flushing touches and 45806 button events. Between that and disabling the device (which causes new events 45807 to be refused) there is a window where events may be triggered and enqueued. 45808 On the next call to PIE that event is processed on a now defunct device, 45809 causing the crash. 45810 45811 The simplest fix to this is to discard events from disabled devices. We flush 45812 the queue often enough before disabling that when we get here, we really don't 45813 care about the events from this device. 45814 45815 X.Org Bug 77884 <http://bugs.freedesktop.org/show_bug.cgi?id=77884> 45816 45817 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 45818 Reported-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 45819 Tested-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 45820 Reviewed-by: Keith Packard <keithp@keithp.com> 45821 Signed-off-by: Keith Packard <keithp@keithp.com> 45822 45823commit e6f5d9d7b7efdacea0f22f1808efca849bcede4c 45824Author: Keith Packard <keithp@keithp.com> 45825Date: Mon Jan 27 11:23:58 2014 -0800 45826 45827 present: Queue flips for later execution 45828 45829 When a flip (or unflip) is pending and a flip request comes in, leave 45830 it queued until the pending flip completes and then execute it. 45831 45832 This fixes a bug where an application submitting back-to-back 45833 present_pixmap requests for sequential frames would alternate between 45834 flipping and copying as the pending flip would cause the new 45835 present_pixmap request to not use a flip. 45836 45837 Signed-off-by: Keith Packard <keithp@keithp.com> 45838 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 45839 Tested-by: Frank Binns <frank.binns@imgtec.com> 45840 45841commit 2f5cf9ff9a0f713b7e038636484c77f113a5f10a 45842Author: Frank Binns <frank.binns@imgtec.com> 45843Date: Thu Apr 24 11:37:48 2014 +0100 45844 45845 xfixes: disable cursor on X server reset 45846 45847 The initial state of the cursor is set to disabled but this was 45848 never be re-disabled during X server reset. This meant any 45849 application run after an X server reset would have the cursor 45850 displayed even if it hadn't requested this to be the case. 45851 45852 Signed-off-by: Frank Binns <frank.binns@imgtec.com> 45853 Reviewed-by: Keith Packard <keithp@keithp.com> 45854 Signed-off-by: Keith Packard <keithp@keithp.com> 45855 45856commit b92d86a982504af2605042c19ef796d135074ee1 45857Author: Brendan King <brendan.king@imgtec.com> 45858Date: Thu Apr 24 11:37:47 2014 +0100 45859 45860 dix: fix pixmap leak on server reset 45861 45862 The server is leaking a pixmap (created by CreateDefaultStipple()) on 45863 reset. The leak is caused by some X Server graphics contexts not being 45864 freed on reset by the machine independent cursor code in the server, 45865 which in turn is caused by the cursor cleanup code 45866 (miSpriteDeviceCursorCleanup()) not being called. 45867 45868 Ensures the DeviceCursorCleanup() function is called when the associated 45869 input device is closed on server reset. 45870 45871 Signed-off-by: Frank Binns <frank.binns@imgtec.com> 45872 Reviewed-by: Keith Packard <keithp@keithp.com> 45873 Signed-off-by: Keith Packard <keithp@keithp.com> 45874 45875commit a99c061aafd4aaa5728c776dca7d0a1789ac739a 45876Author: Brendan King <brendan.king@imgtec.com> 45877Date: Thu Apr 24 11:37:46 2014 +0100 45878 45879 dix: reference the cursor just once in InitializeSprite() 45880 45881 The new current cursor was being referenced twice, resulting in a 45882 memory leak when the current server generation ended. 45883 45884 Signed-off-by: Frank Binns <frank.binns@imgtec.com> 45885 Reviewed-by: Keith Packard <keithp@keithp.com> 45886 Signed-off-by: Keith Packard <keithp@keithp.com> 45887 45888commit 444a1f7a8802999e27ecf5f6eb598df2206f7277 45889Author: Brendan King <brendan.king@imgtec.com> 45890Date: Thu Apr 24 11:37:45 2014 +0100 45891 45892 fb: fix screen pixmap leak on server reset 45893 45894 Call FreePixmap() instead of free() to destroy the screen pixmap in 45895 fbCloseScreen(). 45896 45897 Signed-off-by: Frank Binns <frank.binns@imgtec.com> 45898 Reviewed-by: Keith Packard <keithp@keithp.com> 45899 Signed-off-by: Keith Packard <keithp@keithp.com> 45900 45901commit 7ca458493aa2f0aa091c989ea0768611e0730bf5 45902Author: Chris Wilson <chris@chris-wilson.co.uk> 45903Date: Wed May 28 08:14:00 2014 +0100 45904 45905 xfree86: Report Present as a built-in module 45906 45907 This is so that drivers can do a runtime check that Present is available, 45908 similar to existing runtime checks performed by the drivers for DRI. 45909 45910 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 45911 Signed-off-by: Keith Packard <keithp@keithp.com> 45912 45913commit 746be5a03ebbda4ab411ca3efb2ed95f99e9ea46 45914Author: Chris Wilson <chris@chris-wilson.co.uk> 45915Date: Wed May 28 08:13:59 2014 +0100 45916 45917 xfree86: Report DRI3 as a built-in module 45918 45919 This is so that drivers can do a runtime check that DRI3 is available, 45920 similar to existing runtime checks performed by the drivers for DRI and 45921 DRI2. 45922 45923 v2: Only add DRI3 to the list if the module was actually built into the 45924 server (Mark Kettenis). 45925 45926 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 45927 Cc: Mark Kettenis <mark.kettenis@xs4all.nl> 45928 Signed-off-by: Keith Packard <keithp@keithp.com> 45929 45930commit b5526141f78df15bc5df2497bc4b4a17f1a6368d 45931Merge: dc8d06884 a205ec3ab 45932Author: Keith Packard <keithp@keithp.com> 45933Date: Mon Jun 2 13:08:23 2014 -0700 45934 45935 Merge remote-tracking branch 'jturney/master' 45936 45937commit dc8d0688471695ec5a8db5fef93fbcfc064891d5 45938Author: Max Filippov <jcmvbkbc@gmail.com> 45939Date: Mon Jun 2 01:17:31 2014 +0400 45940 45941 xtensa: add support for xtensa architecture 45942 45943 Signed-off-by: Max Filippov <jcmvbkbc@gmail.com> 45944 Reviewed-by: Adam Jackson <ajax@redhat.com> 45945 Signed-off-by: Keith Packard <keithp@keithp.com> 45946 45947commit a6ffdc69c4ac45bc39924159c7fab5f02f720f1f 45948Merge: 63c48de63 1c10b3738 45949Author: Keith Packard <keithp@keithp.com> 45950Date: Mon Jun 2 11:30:24 2014 -0700 45951 45952 Merge remote-tracking branch 'jeremyhu/master' 45953 45954commit 63c48de63becdee680e90cb2363984c2f8f77fd1 45955Author: Søren Sandmann <soren.sandmann@gmail.com> 45956Date: Fri May 30 16:50:35 2014 -0400 45957 45958 xfree86: Add "modesetting" to list of fallback drivers 45959 45960 To make X -configure work properly, the output of fixup_video_driver_list() 45961 should be in order of preference. Otherwise, the config file may use 45962 the incorrect driver for some devices. 45963 45964 In particular, the drivers that work for all (or many) devices need to be 45965 last in the list. Since the modesetting driver works for many devices, 45966 it needs to be considered a fallback driver. 45967 45968 Signed-off-by: Søren Sandmann <ssp@redhat.com> 45969 Acked-by: Daniel Stone <daniels@collabora.com> 45970 Signed-off-by: Keith Packard <keithp@keithp.com> 45971 45972commit ab47ec9636b21a7f542d2e8612834915e7c8cc64 45973Merge: 0df871cf3 cfaf2abba 45974Author: Keith Packard <keithp@keithp.com> 45975Date: Mon Jun 2 11:22:48 2014 -0700 45976 45977 Merge remote-tracking branch 'whot/for-keith' 45978 45979commit 0df871cf34ee5f1a85586206027de9b02fb364ec 45980Author: Robert Ancell <robert.ancell@canonical.com> 45981Date: Thu May 22 10:43:52 2014 +1200 45982 45983 Fix overflow checking extension versions 45984 45985 The easiest way to check for the version of an extension is to send the maximum 45986 possible version numbers in the QueryVersion request. The X server overflows on 45987 these as it assumes you will send a reasonable version number. 45988 45989 Reviewed-by: Keith Packard <keithp@keithp.com> 45990 Signed-off-by: Keith Packard <keithp@keithp.com> 45991 45992commit 4cbd3fabaa41e89579530bceb01b9239fa2afdb7 45993Author: Adam Jackson <ajax@redhat.com> 45994Date: Wed May 21 09:22:49 2014 -0400 45995 45996 configure: Require sufficiently new wayland-client for xwayland ddx 45997 45998 The explicit release requests were added in 1.3.0, don't try to build 45999 against older. 46000 46001 Signed-off-by: Adam Jackson <ajax@redhat.com> 46002 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 46003 Signed-off-by: Keith Packard <keithp@keithp.com> 46004 46005commit a205ec3ab77c6765d44f77c526a832345fef77b0 46006Author: Jon TURNEY <jon.turney@dronecode.org.uk> 46007Date: Sat Apr 5 15:15:46 2014 +0100 46008 46009 configure: Restore DEFAULT_LOGDIR to xwin-config.h, it is used 46010 46011 Commit f37a46913489397d8628ffe578c8d4ed50b6ca72 removed this from 46012 xwin-config.h.in for some reason, but it is used. 46013 46014 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 46015 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 46016 46017commit aaaa41345674a1c76a70836394cb5e57f82e9a26 46018Author: Jon TURNEY <jon.turney@dronecode.org.uk> 46019Date: Sun Apr 27 20:55:51 2014 +0100 46020 46021 hw/xwin: Update for __glXLastContext -> lastGLContext 46022 46023 Update for __glXLastContext -> lastGLContext. 46024 46025 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 46026 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 46027 46028commit 1c10b37380d228b35db8a8616a6312ac54f5e59b 46029Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 46030Date: Sat May 31 17:45:02 2014 -0700 46031 46032 XQuartz: Update logic to account for title bar on every display in Mavericks 46033 46034 <rdar://problem/14205847> 46035 http://xquartz.macosforge.org/trac/ticket/832 46036 46037 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 46038 46039commit ad0ff649c68b18c4b95b078c2d1d1e7de71a7c6f 46040Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 46041Date: Sat May 31 17:22:51 2014 -0700 46042 46043 XQuartz: Dead code removal (ENABLE_DEBUG_LOG) 46044 46045 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 46046 46047commit e281288d5afedb154f52c31f70a350a013ecfb06 46048Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 46049Date: Sat May 31 13:04:43 2014 -0700 46050 46051 XQuartz: Bump to 2.7.7 46052 46053 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 46054 46055commit cfaf2abbac3f01e57d00845d8908bf01559263f9 46056Author: Peter Hutterer <peter.hutterer@who-t.net> 46057Date: Sat May 24 20:05:53 2014 +1000 46058 46059 man: drop specific mention of DontZap in -retro (#71113) 46060 46061 DontZap off is the default anyway, don't mention it specifically to avoid 46062 confusion 46063 46064 X.Org Bug 71113 <http://bugs.freedesktop.org/show_bug.cgi?id=71113> 46065 46066 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46067 46068commit e48a132b6d187f355abd7021be47edde972e7091 46069Author: Peter Hutterer <peter.hutterer@who-t.net> 46070Date: Sat May 24 20:02:56 2014 +1000 46071 46072 xfree86: fix wrong DontZap documentation (#71113) 46073 46074 X.Org Bug 71113 <http://bugs.freedesktop.org/show_bug.cgi?id=71113> 46075 46076 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46077 46078commit 548fc937b22d4dfe7f96e0bd77522261603a2c2f 46079Author: Robert Ancell <robert.ancell@canonical.com> 46080Date: Thu May 22 10:43:52 2014 +1200 46081 46082 Fix overflow checking extension versions 46083 46084 The easiest way to check for the version of an extension is to send the maximum 46085 possible version numbers in the QueryVersion request. The X server overflows on 46086 these as it assumes you will send a reasonable version number. 46087 46088 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46089 Reviewed-by: Keith Packard <keithp@keithp.com> 46090 46091commit db2e708f31a162c6c66643d3559dd5f3e21ee06b 46092Author: Adam Jackson <ajax@redhat.com> 46093Date: Thu Sep 27 17:44:42 2012 -0400 46094 46095 glx: Require at least one True/DirectColor visual 46096 46097 Mesa no longer supports rendering to anything less. GLX 1.2 requires 46098 that the server advertise at least one GLX visual. GLX 1.3 and 1.4 are 46099 more subtle: they require at least one fbconfig capable of window 46100 rendering, and _also_ require that window-capable fbconfigs have a 46101 non-zero value for the GLX_VISUAL_ID. In either case we should refuse 46102 to init GLX if there's not at least one GL-capable visual on at least 46103 one screen. 46104 46105 Signed-off-by: Adam Jackson <ajax@redhat.com> 46106 Signed-off-by: Keith Packard <keithp@keithp.com> 46107 46108commit d7a2df0a7499864cb005b098b79c1bdf884f6600 46109Author: Steven McDonald <steven@steven-mcdonald.id.au> 46110Date: Sun May 18 13:42:08 2014 +0200 46111 46112 Xi: block SIGIOs while copying device classes around 46113 46114 I've been seeing sporadic (anywhere from once every few days to 3-4 46115 times a day) crashes and freezes in X. The problematic behaviour isn't 46116 always the same, but I chose a particular incident to debug, and found 46117 that X was segfaulting in updateMotionHistory, on line 575 of 46118 dix/getevents.c. 46119 46120 After some further investigation, I found that the bug was being 46121 triggered when a SIGIO was received in DeepCopyPointerClasses, between 46122 the AllocValuatorClass call (line 540) and updating the to->valuator 46123 pointer (line 545). AllocValuatorClass calls realloc() on to->valuator, 46124 so between these lines, it's not guaranteed to point to allocated 46125 memory. 46126 46127 It seems the SIGIO handler is calling updateMotionHistory, which is 46128 reading the memory pointed to by to->valuator and getting a wrong value 46129 for last_motion, which updates buff to point to wildly the wrong place 46130 and thus generates a segfault when a memcpy() is done into buff. 46131 46132 I am attaching a patch which I've been running on that machine for the 46133 past three days, and haven't yet observed any more crashing or freezing 46134 behaviour. The patch simply calls OsBlockSIGIO while 46135 DeepCopyDeviceClasses is in progress, as the state of the X server's 46136 device data structures is not guaranteed to be in a consistent state 46137 during that time. 46138 46139 Debian bug#744303 <https://bugs.debian.org/744303> 46140 46141 Signed-off-by: Julien Cristau <jcristau@debian.org> 46142 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 46143 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46144 46145commit d3a3ee7a0e5cbd3fcaba77d2d5664e1524eda11c 46146Author: Peter Hutterer <peter.hutterer@who-t.net> 46147Date: Thu May 8 09:39:36 2014 +1000 46148 46149 os: automatically switch to sigsafe logging when needed 46150 46151 If we're smart enough to warn, we should be smart enough to just pass it 46152 through to the right function. Worst case we lose some formatting specifiers 46153 which pnprintf will complain about anyway. And in most cases it won't matter. 46154 46155 This requires renaming pnprintf to vpnprintf and changing the size_t to int to 46156 be compatible with Xvscnprintf. pnprintf is internal only, the others are 46157 exported API so we can't change them as easily. 46158 46159 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46160 Reviewed-by: Keith Packard <keithp@keithp.com> 46161 46162commit 6d684f916b4c33327d32ef1f48803bf90357287f 46163Author: Adam Jackson <ajax@redhat.com> 46164Date: Wed May 21 09:23:19 2014 -0400 46165 46166 shadowfb: Fix initialization 46167 46168 This has to run at initial CreateWindow time, at CreateScreenResources 46169 the root window doesn't actually exist yet. 46170 46171 Tested-by: Michael Thayer <michael.thayer@oracle.com> 46172 Signed-off-by: Adam Jackson <ajax@redhat.com> 46173 Reviewed-by: Michael Thayer <michael.thayer@oracle.com> 46174 Signed-off-by: Keith Packard <keithp@keithp.com> 46175 46176commit d9e99edd4dcf109492abdaa73ccf8f29d5bfd159 46177Author: Adam Jackson <ajax@redhat.com> 46178Date: Wed Apr 30 13:49:06 2014 -0400 46179 46180 glxproxy: Fix memory leak on error path in CreateGLXPixmap (#50281) (v2) 46181 46182 v2: Fix another path spotted by keithp 46183 46184 Signed-off-by: Adam Jackson <ajax@redhat.com> 46185 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 46186 Reviewed-by: Keith Packard <keithp@keithp.com> 46187 Signed-off-by: Keith Packard <keithp@keithp.com> 46188 46189commit dff1f5e1a9608eda2992fc9e3304edee7b5bee9b 46190Author: Adam Jackson <ajax@redhat.com> 46191Date: Wed Apr 30 13:49:05 2014 -0400 46192 46193 dmx: Fix memory leak in dmxBELoadFont (#50281) 46194 46195 Signed-off-by: Adam Jackson <ajax@redhat.com> 46196 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 46197 Signed-off-by: Keith Packard <keithp@keithp.com> 46198 46199commit 3dba171f5e82101aa8ea2e47d2253c3410907f6a 46200Author: Michal Srb <msrb@suse.com> 46201Date: Tue Apr 15 18:54:35 2014 +0300 46202 46203 dri2: Fix detection of wrong prime_id in GetScreenPrime. 46204 46205 Checking the iterating variable ("slave") against null can not detect if the 46206 xorg_list_for_each_entry finished without break being invoked - slave variable 46207 will be always non-null. This caused segfault whenever someone tried to use 46208 DRI_PRIME with incorrect id while having at least one render offloading slave 46209 configured. 46210 46211 Restructurize the GetScreenPrime to work as expected. 46212 46213 Reviewed-by: Dave Airlie <airlied@redhat.com> 46214 Signed-off-by: Keith Packard <keithp@keithp.com> 46215 46216commit c08d2b8ccb0158dbf1f529f80cc3236e66236cce 46217Author: Alexey Brodkin <Alexey.Brodkin@synopsys.com> 46218Date: Thu May 1 15:39:29 2014 +0300 46219 46220 ARC: Add support for ARC architecture 46221 46222 Xorg server could be built for and run on Synopsys DesignWare ARC cores. 46223 These changes are required for successful building and execution of the server. 46224 46225 Both little-endian and big-endian flavors of ARC cores are supported. 46226 46227 Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com> 46228 Acked-by: Adam Jackson <ajax@redhat.com> 46229 Reviewed-by: Egbert Eich <eich@freedesktop.org> 46230 Signed-off-by: Keith Packard <keithp@keithp.com> 46231 46232commit 042bae0b49f6192ec3cb8bb51423c90c44638134 46233Author: Adam Jackson <ajax@redhat.com> 46234Date: Wed May 21 10:16:39 2014 -0400 46235 46236 Use own thunk function instead of shadowUpdatePackedWeak 46237 46238 I plan to remove the Weak functions from a future server. 46239 46240 Signed-off-by: Adam Jackson <ajax@redhat.com> 46241 46242commit d5587cf3acb7ad6f81e1c095f1b88f8fb25ee953 46243Author: Peter Hutterer <peter.hutterer@who-t.net> 46244Date: Wed May 14 10:32:30 2014 +1000 46245 46246 config: revert quirk for the Evoluent Vertical Mouse 46247 46248 This quick makes the mouse behaviour more obvious (LMR buttons generate LMR 46249 events) but goes against the manufacturer's intended defaults: 46250 46251 Top button - left click. 46252 Wheel button - .middle click (e.g., for pan and rotate in CAD programs). 46253 Middle button - right click. 46254 Bottom button - back. 46255 Thumb button - forward. 46256 46257 from http://www.evoluent.com/vm3.html 46258 46259 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46260 46261commit 01e18af17f8dc91451fbd0902049045afd1cea7e 46262Author: Dave Airlie <airlied@redhat.com> 46263Date: Mon May 12 14:55:04 2014 -0700 46264 46265 rrcrtc: brackets are hard, lets go shopping. 46266 46267 Slaving two outputs on a secondary GPU to a primary GPU testing 46268 picked this up, in that we'd try to resize to the totally the 46269 wrong thing, then as usual segfault in the rotation code. 46270 46271 Signed-off-by: Dave Airlie <airlied@redhat.com> 46272 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 46273 46274commit 6dd86f3cba7cca67f3a9102db2e90722d32a34cb 46275Author: Eric Anholt <eric@anholt.net> 46276Date: Mon May 5 17:01:14 2014 -0700 46277 46278 glamor: Don't forget to check whether we can fall back in polysegment. 46279 46280 Part of the _nf contract is that glamor will only return FALSE if 46281 glamor has checked that UXA can actually map the pixmaps (UXA only 46282 allocates the BO itself in the screen pixmap and DRI2 cases, and can't 46283 map it otherwise). Fixes server segfaults zooming in and out of 46284 libreoffice spreadsheets. 46285 46286 Signed-off-by: Eric Anholt <eric@anholt.net> 46287 Reviewed-by: Keith Packard <keithp@keithp.com> 46288 Signed-off-by: Keith Packard <keithp@keithp.com> 46289 46290commit fb24ac0a2c317099bdf7d1d8e4b7c443e3ee644f 46291Author: Laércio de Sousa <lbsousajr@gmail.com> 46292Date: Fri May 9 09:09:16 2014 -0300 46293 46294 xfree86: fix warnings after MatchSeat patch 46295 46296 This patch fixes some compile warnings that arise after 46297 commit 7070ebeebaca1b51f8a2801989120784a1c374ae 46298 (xfree86: add new key MatchSeat to xorg.conf sections "Device", "Screen", and "ServerLayout") 46299 available at git repository 46300 git://people.freedesktop.org/~whot/xserver for-keith 46301 46302 Signed-off-by: Keith Packard <keithp@keithp.com> 46303 46304commit bbfed454b2a0fd5e2723524d145cb7c6e5aa9193 46305Author: Eric Anholt <eric@anholt.net> 46306Date: Thu May 8 08:45:22 2014 -0700 46307 46308 glamor: Stop disabling asserts by default. 46309 46310 Disabling asserts is something the user gets to manage. 46311 46312 Signed-off-by: Eric Anholt <eric@anholt.net> 46313 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46314 Signed-off-by: Keith Packard <keithp@keithp.com> 46315 46316commit 8e2fefe3ef247f8b5d74e32e8d37c619b06fc60c 46317Author: Peter Hutterer <peter.hutterer@who-t.net> 46318Date: Tue Apr 29 16:52:01 2014 +1000 46319 46320 Xi: don't copy a DeviceEvent into an InternalEvent 46321 46322 ==26141== Invalid read of size 8 46323 ==26141== at 0x58FAEA: DeliverEmulatedMotionEvent (exevents.c:1484) 46324 46325 An InternalEvent is bigger than a DeviceEvent, thus copying one to the other 46326 reads past the allocated boundary. Shouldn't have any real effect since we 46327 shouldn't access anything past the DeviceEvent boundary if the event type is 46328 correct. 46329 46330 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46331 Reviewed-by: Keith Packard <keithp@keithp.com> 46332 46333commit d7ac9aff061f2961e6b76557dda97b57988ce362 46334Author: Peter Hutterer <peter.hutterer@who-t.net> 46335Date: Wed Apr 30 10:55:26 2014 +1000 46336 46337 mi: start calculating the sprite trace from the first child of root 46338 46339 Introduced in 46340 73698d41e41ce76bef2d9a90b46ac0c24ae148dd "Make XYToWindow a screen function" 46341 46342 Moving the code into miwindow.c changed the start of the loop from 46343 RootWindow()->firstChild to DeepestSpriteWindow(). This function is only 46344 supposed to be called from miXYToWindow which resets spriteTraceGood to 1, 46345 thus DeepestSpriteWindow() is always the root window anyway. 46346 What got dropped was the firstChild as the first window to handle, so we may 46347 end up with the root window twice in the sprite trace. 46348 46349 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46350 Reviewed-by: Keith Packard <keithp@keithp.com> 46351 46352commit d60724b752613ce2ad11c6ae102bf4513f9768ff 46353Merge: a5b975714 15460eaed 46354Author: Keith Packard <keithp@keithp.com> 46355Date: Thu May 8 12:58:41 2014 -0700 46356 46357 Merge commit 'who/for-keith' 46358 46359 Skipping two unreviewed patches; will mark them reviewed and cherry pick 46360 46361commit a5b9757142a2ab471ca26651dce9cc5f5e351f3d 46362Author: Keith Packard <keithp@keithp.com> 46363Date: Fri Apr 25 15:07:03 2014 -0700 46364 46365 glamor: Publish change_window_attributes and copy_window 46366 46367 Because uxa doesn't just use glamor directly, it keeps these two 46368 functions from being wrapped so that they get called 46369 automatically. Publishing these will allow uxa to call them directly. 46370 46371 Signed-off-by: Keith Packard <keithp@keithp.com> 46372 Signed-off-by: Eric Anholt <eric@anholt.net> 46373 Reviewed-by: Eric Anholt <eric@anholt.net> 46374 46375commit 4711182033ec579caff8c930d420f90ecdbe54cf 46376Author: Keith Packard <keithp@keithp.com> 46377Date: Fri Apr 25 20:25:56 2014 -0700 46378 46379 glamor: Work around libXfont when it fails to use defaultChar 46380 46381 GetGlyphs is supposed to always return the full list of characters 46382 when there is a default character available. However, if an 46383 application opens a 16-bit two dimensional font and then draws with 46384 8-bit requests, the bitmapGetGlyphs function in libXfont versions up 46385 through 1.4.7 will return zero glyphs if there is no 0th row. 46386 46387 While this is a bug in libXfont and should be fixed there, it's easy 46388 to protect glamor from it by simply falling through to the case that 46389 handles GetGlyphs failures for fonts without a default character. 46390 46391 Signed-off-by: Keith Packard <keithp@keithp.com> 46392 Signed-off-by: Eric Anholt <eric@anholt.net> 46393 Reviewed-by: Eric Anholt <eric@anholt.net> 46394 46395commit 91767a32874790f8a8c7340be0e81a0b719ca4f6 46396Author: Keith Packard <keithp@keithp.com> 46397Date: Fri Apr 25 23:38:35 2014 -0700 46398 46399 glamor: Fix uxa-entry point for ImageText16 46400 46401 Was interpreting the incoming chars as 8-bits instead of 16-bits, 46402 resulting in the wrong characters being drawn. 46403 46404 Signed-off-by: Keith Packard <keithp@keithp.com> 46405 Signed-off-by: Eric Anholt <eric@anholt.net> 46406 Reviewed-by: Eric Anholt <eric@anholt.net> 46407 46408commit 15460eaed28b5353a94a17091d7795cfaeda9146 46409Author: Peter Hutterer <peter.hutterer@who-t.net> 46410Date: Fri May 2 14:14:58 2014 +1000 46411 46412 systemd-logind: let the logind code decided whether to close an fd 46413 46414 We can only request one fd per device from systemd-logind. If a fd is re-used 46415 by the same device, releasing the fd from one device doesn't mean we can close 46416 it. The systemd code knows when it's really released, so let it close the fd. 46417 46418 Test case: xorg.conf section for an input device with hotplugging enabled. 46419 evdev detects the duplicate and closes the hotplugged device, which closes the 46420 fd. The other instance of evdev thinks the fd is still valid so now you're 46421 playing a double lottery. First, which client(s) will get the evdev fd? 46422 Second, which requests will be picked up by evdev and which ones will be 46423 picked up by the client? You'll never know, but the fun is in finding out. 46424 46425 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46426 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 46427 46428commit 219ce9e6c34bfc946005fb8e81ab4fb74ae1d25a 46429Author: Dave Airlie <airlied@redhat.com> 46430Date: Fri May 2 13:16:05 2014 +1000 46431 46432 modesetting: fix build regression against older servers. 46433 46434 Signed-off-by: Dave Airlie <airlied@redhat.com> 46435 46436commit 795af9ff017113788b4bb6d9dbeb5370c5b8165c 46437Author: Dave Airlie <airlied@redhat.com> 46438Date: Fri May 2 13:14:30 2014 +1000 46439 46440 modesetting: fix use after free. 46441 46442 Signed-off-by: Dave Airlie <airlied@redhat.com> 46443 46444commit 42a4873cc7fe0c7a729e48f806e877c4dac5c07f 46445Author: Keith Packard <keithp@keithp.com> 46446Date: Wed Apr 30 09:32:29 2014 -0700 46447 46448 hw/xfree86: Video Driver ABI version 18.0 46449 46450 With the change in the cursor interface in 46451 4c3932620c29c91dfbbc8eb09c84efcaa7ec873e, we need to bump the video 46452 driver ABI number to ensure that drivers are rebuilt to match the new 46453 interface. 46454 46455 Signed-off-by: Keith Packard <keithp@keithp.com> 46456 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 46457 46458commit a6f5ffd5879e7fb052d343592951cc476b699bb4 46459Author: Laércio de Sousa <lbsousajr@gmail.com> 46460Date: Thu Apr 3 11:19:15 2014 -0300 46461 46462 xfree86: add short description about MatchSeat key in xorg.conf man page 46463 46464 Signed-off-by: Laércio de Sousa <lbsousajr@gmail.com> 46465 Reviewed-by: Dave Airlie <airlied@redhat.com> 46466 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46467 46468commit 7070ebeebaca1b51f8a2801989120784a1c374ae 46469Author: Oleg Samarin <osamarin68@gmail.com> 46470Date: Thu Apr 3 11:19:14 2014 -0300 46471 46472 xfree86: add new key MatchSeat to xorg.conf sections "Device", "Screen", and "ServerLayout" 46473 46474 This patch introduces a new key MatchSeat in xorg.conf (also applies to 46475 any .conf file in xorg.conf.d). It will allow targeting a given 46476 "Device", "Screen", and/or "ServerLayout" section to a particular 46477 seat only (specified by option "-seat" in X server command line), 46478 so that other seats won't be affected. 46479 46480 Without this patch, one needs to write a separate xorg.conf.custom 46481 file and pass it to X server via "-config" option, if one wants that 46482 these settings only apply for the right seat. However, in some cases, 46483 this solution is undesirable or even impossible (e.g. when using GDM, 46484 which doesn't allow X server command line customization). 46485 46486 Example file (/etc/X11/xorg.conf.d/seat1.conf), which would be ignored 46487 by X server unless it was started with "-seat seat1" option: 46488 46489 Section "Device" 46490 Identifier "card0" 46491 Driver "nvidia" 46492 Option "NoLogo" "True" 46493 MatchSeat "seat1" 46494 EndSection 46495 46496 Signed-off-by: Oleg Samarin <osamarin68@gmail.com> 46497 Signed-off-by: Laércio de Sousa <lbsousajr@gmail.com> 46498 Reviewed-by: Dave Airlie <airlied@redhat.com> 46499 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46500 46501commit 29b1484bb9555e45067669cbfe68a3c40596f4ff 46502Author: Laércio de Sousa <lbsousajr@gmail.com> 46503Date: Thu Apr 3 11:19:13 2014 -0300 46504 46505 xfree86: allow fallback to PCI bus probe for graphics devices on non-seat0 X servers (#66851) 46506 46507 Currently non-seat0 X servers only probe platform bus for graphics devices, 46508 which is OK for most KMS-compliant drivers. However, for non-KMS drivers 46509 (like NVIDIA proprietary ones), graphics devices can't be reached 46510 by platform bus probe, resulting in a "No devices detected" error. 46511 46512 This patch allows a fallback to PCI bus probe for non-seat0 X servers 46513 in case no platform bus graphics device is found. 46514 46515 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=66851 46516 46517 Signed-off-by: Laércio de Sousa <lbsousajr@gmail.com> 46518 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 46519 Reviewed-by: Dave Airlie <airlied@redhat.com> 46520 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46521 46522commit 2535b76c0d32bc1dd0ddaca06a419a68a4757df1 46523Author: Robert Ancell <robert.ancell@canonical.com> 46524Date: Wed Apr 23 08:19:35 2014 +1200 46525 46526 os: Add -displayfd into -help text 46527 46528 Add -displayfd into -help text. It was mentioned in the man page but seem to have been missed from the -help text. 46529 46530 Signed-off-by: Keith Packard <keithp@keithp.com> 46531 46532commit a26578dbb93dbaab1910b52f9fa1c87f8c9e6d1e 46533Author: Ray Strode <rstrode@redhat.com> 46534Date: Thu Apr 24 13:38:29 2014 -0400 46535 46536 selinux: don't checkout for AVCs on select() EINTR 46537 46538 wakeup handlers are called even when select() returns EINTR, 46539 and when they're called the passed fd set is undefined. 46540 46541 This commit fixes the selinux wakeup handler to avoid checking 46542 for AVCs over the netlink socket spuriously. 46543 46544 Reviewed-by: Adam Jackson <ajax@redhat.com> 46545 Signed-off-by: Ray Strode <rstrode@redhat.com> 46546 Signed-off-by: Keith Packard <keithp@keithp.com> 46547 46548commit bcb17b99873f020f2639b860ce99d260def0cd97 46549Author: Ray Strode <rstrode@redhat.com> 46550Date: Thu Apr 24 13:38:28 2014 -0400 46551 46552 dbus: don't dispatch dbus events on select() EINTR 46553 46554 wakeup handlers are called even when select() returns EINTR, 46555 and when they're called the passed fd set is undefined. 46556 46557 This commit fixes the dbus wakeup handler to avoid dispatching 46558 into dbus spuriously. 46559 46560 Reviewed-by: Adam Jackson <ajax@redhat.com> 46561 Signed-off-by: Ray Strode <rstrode@redhat.com> 46562 Signed-off-by: Keith Packard <keithp@keithp.com> 46563 46564commit 6bd7ad914470322ddd02b50078e159234ebec643 46565Merge: b92a9289f e06fbe4dc 46566Author: Keith Packard <keithp@keithp.com> 46567Date: Tue Apr 29 09:37:37 2014 -0700 46568 46569 Merge remote-tracking branch 'whot/for-keith' 46570 46571commit e06fbe4dc81110e01d36f958849c4ab5f2a9df0e 46572Author: Peter Hutterer <peter.hutterer@who-t.net> 46573Date: Mon Apr 28 07:57:00 2014 +1000 46574 46575 dix: remove a now-superfluous comment 46576 46577 Obsolete since 93945b0a74aa8156a88f52b8ba77f1210042f396 46578 46579 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46580 46581commit 6a848122def0f2a439e21c50f22814db11760132 46582Author: Michal Srb <msrb@suse.com> 46583Date: Wed Apr 2 17:14:05 2014 +0300 46584 46585 input: Remove invalid bug checks. 46586 46587 Commit 2f1aedcaed8fd99b823d451bf1fb02330c078f67 added several bug checks. Some 46588 of them are not correct. 46589 46590 Checks in Init(Ptr|String|Bell|Led|Integer)FeedbackClassDeviceStruct verify 46591 that no feedback struct was set yet, but that is not required. If any feedback 46592 structs are already present, the function will chain them behind the new one. 46593 46594 Signed-off-by: Michal Srb <msrb@suse.com> 46595 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 46596 46597commit f073b5e58262d5fcc0102d773091d11d175227ce 46598Author: Hans de Goede <hdegoede@redhat.com> 46599Date: Tue Mar 18 15:48:22 2014 +0100 46600 46601 Add support for server managed fds 46602 46603 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 46604 46605commit b92a9289fef4831e9af69e5f6f64b5454f550559 46606Author: Thierry Reding <thierry.reding@gmail.com> 46607Date: Wed Feb 12 17:11:40 2014 +0100 46608 46609 xfree86: int10: Fix build on ARM 46610 46611 Commit 7353ec7cb6fc235b03e59e35425201429c83ee72 "xfree86: Switch int10 46612 code to stdint types" uses designated initializers to setup the fields 46613 of the X86EMU_pioFuncs. 46614 46615 This breaks compilation on ARM, since out{b,w,l}() are redefined using 46616 the preprocessor and therefore cause the compiler to complain about 46617 non-existent fields being assigned to. 46618 46619 It seems like the compiler.h header that contains these redefinitions 46620 isn't actually needed in xf86x86emu.c, so the easiest "fix" is to not 46621 include it. 46622 46623 Signed-off-by: Thierry Reding <treding@nvidia.com> 46624 Reviewed-by: Rob Clark <robdclark@gmail.com> 46625 46626commit 774fea38c48671ed44e54278f0280bae04822491 46627Author: Rob Clark <robdclark@gmail.com> 46628Date: Thu Apr 3 15:35:53 2014 -0400 46629 46630 int10: fix build error 46631 46632 Fixes: 46633 46634 stub.c:66:1: error: conflicting types for 'xf86int10Addr' 46635 In file included from stub.c:14:0: 46636 xf86int10.h:72:53: note: previous declaration of 'xf86int10Addr' was here 46637 46638 Signed-off-by: Rob Clark <robdclark@gmail.com> 46639 46640commit b241934238feedcb86c113fe3beb69a7470de02d 46641Author: Rob Clark <robdclark@gmail.com> 46642Date: Thu Apr 3 14:54:02 2014 -0400 46643 46644 default to stub int10 implementation on arm 46645 46646 There should be no reason to need a real int10 implementation on arm, 46647 and switching to stub is an easy way to fix: 46648 46649 xf86x86emu.c: In function 'xf86Int10ExecSetup': 46650 xf86x86emu.c:56:9: error: unknown field 'xf_outb' specified in initializer 46651 xf86x86emu.c:57:9: error: unknown field 'xf_outw' specified in initializer 46652 xf86x86emu.c:58:9: error: unknown field 'xf_outl' specified in initializer 46653 46654 which is caused by the following in compiler.h: 46655 46656 #define outb xf_outb 46657 #define outw xf_outw 46658 #define outl xf_outl 46659 46660 Signed-off-by: Rob Clark <robdclark@gmail.com> 46661 Acked-by: Daniel Stone <daniels@collabora.com> 46662 46663commit 4c3932620c29c91dfbbc8eb09c84efcaa7ec873e 46664Author: Keith Packard <keithp@keithp.com> 46665Date: Fri Apr 25 08:22:15 2014 -0700 46666 46667 hw/xfree86: Restore API compatibility for cursor loading functions 46668 46669 Create load_cursor_image_check, load_cursor_argb_check, 46670 LoadCursorImageCheck and LoadCursorARGBCheck that can return failure 46671 and use them in preference to the old unchecked variants. 46672 46673 Signed-off-by: Keith Packard <keithp@keithp.com> 46674 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 46675 Reviewed-by: Michael Thayer <michael.thayer@oracle.com> 46676 46677commit 99f0365b1fbdfd9238b9f5cc28491e4e6c7324f1 46678Author: Eric Anholt <eric@anholt.net> 46679Date: Wed Apr 23 09:52:17 2014 -0700 46680 46681 Add a command line argument for disabling indirect GLX. 46682 46683 The attack surface for indirect GLX is huge, and it's of no use to 46684 most people (if you get an indirect GL context, you're better served 46685 by a immediate X error than actually trying to use an indirect GL 46686 context and finding out that it doesn't support doing anything you 46687 want, slowly). This flag gives you a chance to disable indirect GLX 46688 in environments where you just don't need it. 46689 46690 I put in both the '+' and '-' arguments right now, so that it's easy 46691 to patch the value to change the default policy. 46692 46693 Signed-off-by: Eric Anholt <eric@anholt.net> 46694 Acked-by: Julien Cristau <jcristau@debian.org> 46695 Reviewed-by: Keith Packard <keithp@keithp.com> 46696 Signed-off-by: Keith Packard <keithp@keithp.com> 46697 46698commit a4d96afdbddb7a636df8e336059d3a5624f2e6ae 46699Author: Zhigang Gong <zhigang.gong@intel.com> 46700Date: Wed Apr 23 18:54:43 2014 +0900 46701 46702 glamor: Fallback to system memory when fail to allocate one big fbo. 46703 46704 Even when create a pixmap which smaller than the max_fbo_size, 46705 it may fail due to some low level driver limitation. If that is 46706 the case, we don't need to crash the xserver. We just need to 46707 fallback to system memory. 46708 46709 See the related bug at: 46710 https://bugs.freedesktop.org/show_bug.cgi?id=71190 46711 46712 Signed-off-by: Zhigang Gong <zhigang.gong@intel.com> 46713 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46714 Tested-by: Kai Wasserbach <kai@dev.carbon-project.org> 46715 Tested-by: Erich Seifert <eseifert@error-reports.org> 46716 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 46717 Signed-off-by: Eric Anholt <eric@anholt.net> 46718 46719commit 21e0e373858bd7f3458172ebd465397e33b90162 46720Author: Michel Dänzer <michel.daenzer@amd.com> 46721Date: Wed Apr 23 18:54:42 2014 +0900 46722 46723 glamor: Fix memory leak in _glamor_copy_n_to_n() 46724 46725 It would leak the memory allocated for the region rects in some cases. 46726 Found with valgrind. 46727 46728 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 46729 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 46730 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 46731 Signed-off-by: Eric Anholt <eric@anholt.net> 46732 46733commit 4e9aabb6fc15d8052934f20c6a07801c197ec36a 46734Author: Anthony Waters <awaters1@gmail.com> 46735Date: Wed Apr 23 18:54:41 2014 +0900 46736 46737 glamor: Fix coordinates handling for composite source/mask pictures 46738 46739 There were actually two issues with the original code I believe, the 46740 first is that the call to glamor_convert_gradient_picture wasn't 46741 properly referencing the coordinates of the source/mask pictures. The 46742 second, was that the updated references (x_temp/y_temp) were also 46743 improperly set, they should always be 0 because the temp pictures are 46744 new ones that start at (0, 0). The reason it worked in certain cases 46745 and it didn't in others (notably the tray icons) was due to the 46746 numbers working out based on the call to glamor_composite. In the 46747 cases that it did work extent->x1 would equal x_dest and extent->y1 46748 would equal y_dest, making it so what was actually passed into 46749 glamor_convert_gradient_picture and the settings for x_temp/y_temp 46750 were correct. However, for the case when extent->x1 wouldn't equal 46751 x_dest and extent->y1 wouldn't equal y_dest (for example with the tray 46752 icons) then the wrong parameters get passed into 46753 glamor_convert_gradient_picture and x_temp/y_temp are set improperly. 46754 46755 Fixes issues with tray icons not appearing properly in certain cases. 46756 46757 Bug: 46758 https://bugs.freedesktop.org/show_bug.cgi?id=64738 46759 46760 Signed-Off-by: Anthony Waters <awaters1@gmail.com> 46761 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 46762 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 46763 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 46764 Signed-off-by: Eric Anholt <eric@anholt.net> 46765 46766commit 5062b4fadd977d044e54b53b7f3e02e466eac9a9 46767Author: Keith Packard <keithp@keithp.com> 46768Date: Sun Mar 16 20:44:48 2014 -0700 46769 46770 glamor: Add glamor_transfer based glamor_get_image and glamor_put_image 46771 46772 These use the upload_boxes and download_boxes helpers to provide 46773 reasonably efficient image transfer. 46774 46775 Fixes segfaults in Xephyr with x11perf -reps 1. 46776 46777 Performance improvements: 46778 46779 Improves -putimage10 by 548.218% +/- 88.601% (n=10). 46780 Improves -putimage500 by 3.71014% +/- 1.5049% (n=10). 46781 Improves -getimage10 by 8.37004% +/- 4.58274% (n=10). 46782 No statistically significant difference on -getimage500 (n=10). 46783 46784 v2: Fix rebase failures, don't forget to check/prepare the gc in 46785 putimage fallbacks (changes by anholt). 46786 46787 Signed-off-by: Keith Packard <keithp@keithp.com> 46788 Signed-off-by: Eric Anholt <eric@anholt.net> 46789 Reviewed-by: Eric Anholt <eric@anholt.net> 46790 46791commit 747160016ba2d0cd42ad6b174cbf927d67c01875 46792Author: Keith Packard <keithp@keithp.com> 46793Date: Tue Apr 22 16:33:45 2014 -0700 46794 46795 glamor: Wire alpha to 1 for pictures without alpha bits 46796 46797 When sourcing a picture that has no alpha values, make sure any 46798 texture fetches wire the alpha value to one. This ensures that bits 46799 beyond the depth of the pixmap, or bits other than the RGB values 46800 aren't used. 46801 46802 Signed-off-by: Keith Packard <keithp@keithp.com> 46803 Signed-off-by: Eric Anholt <eric@anholt.net> 46804 Reviewed-by: Eric Anholt <eric@anholt.net> 46805 46806commit a69907288d59792fd783f2f1756cde03d4a06f97 46807Author: Michel Dänzer <michel.daenzer@amd.com> 46808Date: Mon Apr 21 17:47:15 2014 +0900 46809 46810 glx: If DRI2GetBuffers changes the GL context, call it again 46811 46812 By changing the context, it may also invalidate the DRI2 buffer 46813 information, so we need to get that again. 46814 46815 Fixes crashes due to use-after-free with LIBGL_ALWAYS_INDIRECT=1 46816 glxgears and piglit. 46817 46818 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 46819 Signed-off-by: Eric Anholt <eric@anholt.net> 46820 Reviewed-by: Eric Anholt <eric@anholt.net> 46821 Reviewed-by: Adam Jackson <ajax@redhat.com> 46822 46823commit f6abfece3e59fa8c2e14a61430133816837855f8 46824Author: Eric Anholt <eric@anholt.net> 46825Date: Thu Apr 17 18:50:13 2014 -0700 46826 46827 glamor: Move a make_current before the first GL call entrypoint. 46828 46829 Fixes a usage of the wrong context with swrast GLX's GetImage entrypoint. 46830 46831 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46832 Reviewed-by: Adam Jackson <ajax@redhat.com> 46833 46834commit 707726b155a2a086d12e0a6abef7f1e59def5370 46835Author: Eric Anholt <eric@anholt.net> 46836Date: Fri Apr 4 07:03:46 2014 +0100 46837 46838 glamor: Do the same MakeCurrent(None) for GLX as we do for EGL. 46839 46840 Signed-off-by: Eric Anholt <eric@anholt.net> 46841 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46842 Reviewed-by: Adam Jackson <ajax@redhat.com> 46843 46844commit 482b06a95a0a2c9a58bc196eb3dd928d796853df 46845Author: Eric Anholt <eric@anholt.net> 46846Date: Fri Apr 4 07:03:31 2014 +0100 46847 46848 glamor: Explain the weird EGL_NO_CONTEXT code. 46849 46850 Signed-off-by: Eric Anholt <eric@anholt.net> 46851 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46852 Reviewed-by: Adam Jackson <ajax@redhat.com> 46853 46854commit fab0a4a4c9dad2275bb398d09632f4274b5e16f2 46855Author: Eric Anholt <eric@anholt.net> 46856Date: Fri Mar 14 17:31:18 2014 -0700 46857 46858 glamor: Replace glamor_get/put_context() with just glamor_make_current(). 46859 46860 Now that we have the DIX global state for the current context, we 46861 don't need to track nesting to try to reduce MakeCurrent overhead. 46862 46863 v2: Fix a mistaken replacement of a put_context with make_current in 46864 glamor_fill_spans_gl() (caught by keithp). 46865 46866 Signed-off-by: Eric Anholt <eric@anholt.net> 46867 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> (v1) 46868 Reviewed-by: Adam Jackson <ajax@redhat.com> (v1) 46869 46870commit 11ff12d4e74c0b514cb62cc47df9770c91e5002a 46871Author: Eric Anholt <eric@anholt.net> 46872Date: Fri Mar 14 17:29:31 2014 -0700 46873 46874 glamor: Stop unsetting the EGL context in put_context(). 46875 46876 This matches the Xephyr behavior. Now that we know when to reset the 46877 context in the presence of GLX, we don't need to try to keep our stuff 46878 from being smashed by GLX. 46879 46880 Signed-off-by: Eric Anholt <eric@anholt.net> 46881 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46882 Reviewed-by: Adam Jackson <ajax@redhat.com> 46883 46884commit b5e394b3f5d80749af0148611df2eb009e7bf823 46885Author: Eric Anholt <eric@anholt.net> 46886Date: Fri Mar 14 17:20:12 2014 -0700 46887 46888 glamor: Use lastGLContext to coordinate the context with GLX. 46889 46890 This gets us some more context changes that are needed to make sure 46891 the two sides render to the right drawables and manipulate the right 46892 objects. 46893 46894 Signed-off-by: Eric Anholt <eric@anholt.net> 46895 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46896 Reviewed-by: Adam Jackson <ajax@redhat.com> 46897 46898commit ab6e958a2e611b03e475c16d10beb9961d8dffc8 46899Author: Eric Anholt <eric@anholt.net> 46900Date: Wed Apr 9 10:23:01 2014 -0700 46901 46902 glx: Make sure that DRI2/swrast calls haven't changed the GL context. 46903 46904 These functions are called from the GL driver, in some series of GL 46905 calls by GLX. If some server component (like glamor CreatePixmap for 46906 GetBuffers()) changes the GL context on us, we need to set it back or 46907 the later GL calls will land in the glamor context instead of the GLX 46908 context. 46909 46910 Signed-off-by: Eric Anholt <eric@anholt.net> 46911 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46912 Reviewed-by: Adam Jackson <ajax@redhat.com> 46913 46914commit f3f2fb6baac3d2b248eb4b0da13fe95e9dc3de7d 46915Author: Eric Anholt <eric@anholt.net> 46916Date: Wed Apr 9 11:25:32 2014 -0700 46917 46918 glx: Refactor DRI2CopyRegion calls. 46919 46920 I needed to add some code to each one, so it's a good time to make a 46921 helper func. 46922 46923 Signed-off-by: Eric Anholt <eric@anholt.net> 46924 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46925 Reviewed-by: Adam Jackson <ajax@redhat.com> 46926 46927commit 008f1ab31e8ac6be8f3fb11f19cdf8674be318dd 46928Author: Eric Anholt <eric@anholt.net> 46929Date: Thu Apr 17 18:28:54 2014 -0700 46930 46931 glx: Unconditionally clear lastGLContext on loseCurrent(). 46932 46933 This hook calls unbindContext in the DRI driver interface, which 46934 unsets the dispatch table, regardless of whether the context argument 46935 was the current one or not. 46936 46937 Signed-off-by: Eric Anholt <eric@anholt.net> 46938 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46939 Reviewed-by: Adam Jackson <ajax@redhat.com> 46940 46941commit 789509ef5324ce9141222fa92dea51b5cc777209 46942Author: Eric Anholt <eric@anholt.net> 46943Date: Thu Apr 17 18:23:51 2014 -0700 46944 46945 glx: Move GLX MakeCurrent lastGLContext updates next to makeCurrent(). 46946 46947 We want to make sure that lastGLContext is set correctly during 46948 makeCurrent, because we may have recursive GL context changes in the 46949 DRI2 interfaces due to glamor. 46950 46951 Signed-off-by: Eric Anholt <eric@anholt.net> 46952 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46953 Reviewed-by: Adam Jackson <ajax@redhat.com> 46954 46955commit abf12027063dbe22bec1d055ad8db365cf07395a 46956Author: Eric Anholt <eric@anholt.net> 46957Date: Thu Apr 17 18:18:49 2014 -0700 46958 46959 glx: Move the GLX variable caching what GL context is current to dix. 46960 46961 GLX is trying to track whether the context it wants is current, to 46962 avoid the glFlush() (and the rest of the overhead) that occurs on all 46963 MakeCurrent calls. However, its cache can be incorrect now that 46964 glamor exists. This is a step toward getting glamor to coordinate 46965 with GLX. 46966 46967 Signed-off-by: Eric Anholt <eric@anholt.net> 46968 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46969 Reviewed-by: Adam Jackson <ajax@redhat.com> 46970 46971commit f12221cbd8ff33070fa2ca086bccf7ed32115f0e 46972Author: Eric Anholt <eric@anholt.net> 46973Date: Wed Apr 16 16:16:21 2014 -0700 46974 46975 glamor: Fix a missing set of the GL context. 46976 46977 Signed-off-by: Eric Anholt <eric@anholt.net> 46978 Reviewed-by: Keith Packard <keithp@keithp.com> 46979 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 46980 Reviewed-by: Adam Jackson <ajax@redhat.com> 46981 46982commit e924034269532979f1e88947ee9c33d30461edae 46983Author: Eric Anholt <eric@anholt.net> 46984Date: Mon Apr 21 10:56:38 2014 -0700 46985 46986 glamor: Fix accelerated rendering of GTK's ARGB vs xBGR composites. 46987 46988 There is some complicated code to support tweaking the format as we 46989 upload from a SHM pixmap (aka the GTK icon cache), but if we weren't 46990 sourcing from a SHM pixmap we just forgot to check that the formats 46991 matched at all. 46992 46993 We could potentially be a little more discerning here (xRGB source and 46994 ARGB mask would be fine, for example), but this will all change with 46995 texture views anyway, so just get the rendering working for 1.16 46996 release. 46997 46998 Fixes the new rendercheck gtk_argb_xbgr test. 46999 47000 v2: Squash in keithp's fix for checking that we have a non-NULL 47001 pixmap, and reword the comment even more. 47002 47003 Signed-off-by: Eric Anholt <eric@anholt.net> 47004 Reviewed-by: Keith Packard <keithp@keithp.com> 47005 47006commit 4bbee6761d49478fa40ec5daa2797f13ae96ce5c 47007Author: Jamey Sharp <jamey@minilop.net> 47008Date: Fri Apr 18 09:31:09 2014 -0700 47009 47010 Make glamor build with --enable-debug. 47011 47012 Bad anholt, no biscuit. Broken in commit 47013 4c9a20072552c52b3763bd73e7a7e9b9cb8b4993. 47014 47015 Signed-off-by: Jamey Sharp <jamey@minilop.net> 47016 Cc: Eric Anholt <eric@anholt.net> 47017 Signed-off-by: Eric Anholt <eric@anholt.net> 47018 Reviewed-by: Eric Anholt <eric@anholt.net> 47019 47020commit ba2432a020a9f9bd0892f643117795336ba0fc16 47021Author: Adam Jackson <ajax@redhat.com> 47022Date: Thu Apr 10 11:34:28 2014 -0400 47023 47024 xfixes: Forbid manipulating clip for source-only pictures (#28968) 47025 47026 Just throw BadPicture instead of crashing. It's not currently a 47027 meaningful thing to do anyway, RenderSetPictureRectangles would error if 47028 you tried (which this patch changes to BadPicture as well for 47029 consistency). The problem with trying to do it is if the clip is 47030 specified as a pixmap then we try to convert it to a region, and 47031 ->BitmapToRegion requires a ScreenPtr, and source-only pictures don't 47032 have one. 47033 47034 I can imagine a use for client clip on source-only pictures, so if we 47035 really wanted to allow this, probably the way forward is to always store 47036 the clip as a region internally, and when setting the clip _from_ a 47037 pixmap, look up BitmapToRegion relative to the pixmap not the picture. 47038 But since clearly nobody can be relying on it working... 47039 47040 Signed-off-by: Adam Jackson <ajax@redhat.com> 47041 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 47042 Reviewed-by: Keith Packard <keithp@keithp.com> 47043 Signed-off-by: Keith Packard <keithp@keithp.com> 47044 47045commit 70e564104b69bc53d29633f392f2c1ab94caddc9 47046Author: Dominik Behr <dbehr@chromium.org> 47047Date: Tue Apr 1 20:36:13 2014 -0700 47048 47049 xf86RandR12: use correct gamma size when allocating gamma table 47050 47051 When setting crtc->gamma_size to randr_crtc->gammaSize we should 47052 use randr_crtc->gammaSize to allocate new gamma table in crtc. 47053 Currently, if randr_crtc->gammaSize > crtc->gammaSize the subsequent 47054 memcpy will overwrite memory beyond the end of gamma table. 47055 47056 Signed-off-by: Dominik Behr <dbehr@chromium.org> 47057 Reviewed-by: Stéphane Marchesin <marcheu@chromium.org> 47058 Signed-off-by: Keith Packard <keithp@keithp.com> 47059 47060commit 35d275c7519570ceaf82cd5e7a663a8a5be4d441 47061Author: Eric Anholt <eric@anholt.net> 47062Date: Fri Apr 4 08:38:33 2014 +0100 47063 47064 os: Initialize the set of signals to be suppressed during our handler. 47065 47066 Fixes a valgrind complaint: 47067 47068 ==8805== Syscall param rt_sigaction(act->sa_mask) points to uninitialised byte(s) 47069 ==8805== at 0x5EB8315: __libc_sigaction (sigaction.c:66) 47070 ==8805== by 0x5B13DA: busfault_init (busfault.c:145) 47071 ==8805== by 0x5A60A2: OsInit (osinit.c:191) 47072 ==8805== by 0x46EBA2: dix_main (main.c:163) 47073 47074 Signed-off-by: Eric Anholt <eric@anholt.net> 47075 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 47076 Reviewed-by: Adam Jackson <ajax@redhat.com> 47077 Signed-off-by: Keith Packard <keithp@keithp.com> 47078 47079commit 19e5a13970ad9d440c0789395e94931429ef4fbb 47080Author: YunQiang Su <wzssyqa@gmail.com> 47081Date: Sat Oct 12 13:22:35 2013 +0800 47082 47083 Fix hw/xfree86/common/compiler.h for mips64 47084 47085 Mark mips64 as 64bit 47086 Use long as PORT_SIZE 47087 47088 Signed-off-by: YunQiang Su <wzssyqa@gmail.com> 47089 Reviewed-by: Matt Turner <mattst88@gmail.com> 47090 Signed-off-by: Keith Packard <keithp@keithp.com> 47091 47092commit f466fb2432292d74123d701764c432d3aa75ef1d 47093Author: Adam Jackson <ajax@redhat.com> 47094Date: Thu Apr 10 11:45:37 2014 -0400 47095 47096 xres: Fix size estimation for <8bpp pixmaps (#69057) 47097 47098 Just use floats, it's not like this is a performance path. 47099 47100 Signed-off-by: Adam Jackson <ajax@redhat.com> 47101 Reviewed-by: Julien Cristau <jcristau@debian.org> 47102 Signed-off-by: Keith Packard <keithp@keithp.com> 47103 47104commit 5df66822c9f6bac1ed3952971958068600ccd095 47105Author: Keith Packard <keithp@keithp.com> 47106Date: Mon Apr 21 21:38:28 2014 -0700 47107 47108 shadowfb: Fix compilation errors 47109 47110 Ok, that's embarassing -- I didn't even make sure Adam's patch 47111 compiled. These are minimal fixes to make it build. 47112 47113 Signed-off-by: Keith Packard <keithp@keithp.com> 47114 47115commit 2dcb1672620ef5f16111517e9e7a98cf97deb678 47116Author: Adam Jackson <ajax@redhat.com> 47117Date: Thu Apr 10 13:42:42 2014 -0400 47118 47119 shadowfb: Port to miext/damage (#31303, #65547) 47120 47121 Somewhat shocking how much simpler this is, isn't it? We no longer need 47122 to wrap the screen or GC or Picture, because damage does it for us, 47123 which is doubly great since the old shadowfb code didn't wrap _enough_ 47124 things (border updates and Render glyphs, at least). The only real 47125 difference now between this and shadow is a) shadow will let you track 47126 arbitrary pixmaps, and b) shadow's update hook runs off the BlockHandler 47127 whereas shadowfb is immediate. 47128 47129 Tested on nouveau. 47130 47131 Signed-off-by: Adam Jackson <ajax@redhat.com> 47132 Reviewed-by: Keith Packard <keithp@keithp.com> 47133 Signed-off-by: Keith Packard <keithp@keithp.com> 47134 47135commit 8edeac202ecbace859b7889dd811424471dcca73 47136Author: Peter Harris <pharris@opentext.com> 47137Date: Wed Apr 16 20:24:34 2014 -0400 47138 47139 composite: Fix memory leak in compCopyWindow 47140 47141 Signed-off-by: Peter Harris <pharris@opentext.com> 47142 Reviewed-by: Keith Packard <keithp@keithp.com> 47143 Signed-off-by: Keith Packard <keithp@keithp.com> 47144 47145commit 935ea0d7b5e18ad8f22ee074a08cbdc7390d0f5d 47146Author: Alan Coopersmith <alan.coopersmith@oracle.com> 47147Date: Mon Apr 21 15:18:45 2014 -0700 47148 47149 dmxConfigCanvasUpdate: Delay dereference until after checking for NULL 47150 47151 Flagged by cppcheck 1.64: 47152 [hw/dmx/config/xdmxconfig.c:306] -> [hw/dmx/config/xdmxconfig.c:323]: 47153 (warning) Possible null pointer dereference: fs - otherwise 47154 it is redundant to check it against null. 47155 47156 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 47157 Reviewed-by: Keith Packard <keithp@keithp.com> 47158 Signed-off-by: Keith Packard <keithp@keithp.com> 47159 47160commit 983e30361f49a67252d0b5d82630e70724d69dbf 47161Author: Peter Harris <pharris@opentext.com> 47162Date: Fri Apr 11 17:44:59 2014 -0400 47163 47164 fb: Fix origin of source picture in fbGlyphs 47165 47166 If a source picture doesn't repeat and a mask format is specified, the 47167 incorrect calulation of the origin of the glyphs caused the glyphs to 47168 not be drawn at all. 47169 47170 Noticed when running gtk-demo from RHEL 6.5 and selecting "Rotated 47171 Text". 47172 47173 Signed-off-by: Peter Harris <pharris@opentext.com> 47174 Reviewed-by: Keith Packard <keithp@keithp.com> 47175 Signed-off-by: Keith Packard <keithp@keithp.com> 47176 47177 /* Test for this bug 47178 47179 cc -std=c99 -o glyph glyph.c `pkg-config --cflags --libs xcb-render` 47180 47181 */ 47182 47183 // 16 x 16 pictfmt_a8 "glyph" 47184 static const char glyph[] = { 47185 0xff, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0xff, 47186 0, 0xff, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0xff, 0, 47187 0, 0, 0xff, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0xff, 0, 0, 47188 0, 0, 0, 0xff, 0, 0, 0, 0, 0, 0, 0, 0, 0xff, 0, 0, 0, 47189 0, 0, 0, 0, 0xff, 0, 0, 0, 0, 0, 0, 0xff, 0, 0, 0, 0, 47190 0, 0, 0, 0, 0, 0xff, 0, 0, 0, 0, 0xff, 0, 0, 0, 0, 0, 47191 0, 0, 0, 0, 0, 0, 0xff, 0, 0, 0xff, 0, 0, 0, 0, 0, 0, 47192 0, 0, 0, 0, 0, 0, 0, 0xff, 0xff, 0, 0, 0, 0, 0, 0, 0, 47193 0, 0, 0, 0, 0, 0, 0, 0xff, 0xff, 0, 0, 0, 0, 0, 0, 0, 47194 0, 0, 0, 0, 0, 0, 0xff, 0, 0, 0xff, 0, 0, 0, 0, 0, 0, 47195 0, 0, 0, 0, 0, 0xff, 0, 0, 0, 0, 0xff, 0, 0, 0, 0, 0, 47196 0, 0, 0, 0, 0xff, 0, 0, 0, 0, 0, 0, 0xff, 0, 0, 0, 0, 47197 0, 0, 0, 0xff, 0, 0, 0, 0, 0, 0, 0, 0, 0xff, 0, 0, 0, 47198 0, 0, 0xff, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0xff, 0, 0, 47199 0, 0xff, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0xff, 0, 47200 0xff, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0xff, 47201 }; 47202 47203 static struct { 47204 uint8_t len; 47205 uint8_t pad[3]; 47206 uint16_t deltax, deltay; 47207 uint8_t glyph; 47208 uint8_t pad2[3]; 47209 } elt = { len:1, glyph:1, deltax:WIN_SIZE/2 - GLYPH_SIZE/2, deltay:WIN_SIZE/2 - GLYPH_SIZE/2 }; 47210 47211 int main(int argc, char *argv[]) 47212 { 47213 int screen; 47214 xcb_connection_t *c = xcb_connect(NULL, &screen); 47215 if (!c || xcb_connection_has_error(c)) { 47216 fprintf(stderr, "Cannot open default display \"%s\"\n", getenv("DISPLAY")); 47217 return EXIT_FAILURE; 47218 } 47219 47220 // Find root window and depth 47221 const xcb_setup_t *setup = xcb_get_setup(c); 47222 if (screen >= setup->roots_len) 47223 screen = 0; 47224 xcb_screen_iterator_t si = xcb_setup_roots_iterator(setup); 47225 for (int i=0; i < screen; i++) 47226 xcb_screen_next(&si); 47227 xcb_window_t root = si.data->root; 47228 uint8_t depth = si.data->root_depth; 47229 xcb_visualid_t visual = si.data->root_visual; 47230 47231 // Find picture formats 47232 xcb_render_query_pict_formats_reply_t *qpf; 47233 qpf = xcb_render_query_pict_formats_reply(c, xcb_render_query_pict_formats(c), NULL); 47234 if (!qpf) { 47235 fprintf(stderr, "Cannot query RENDER picture formats\n"); 47236 return EXIT_FAILURE; 47237 } 47238 xcb_render_pictformat_t fmt_a8 = 0; 47239 xcb_render_pictforminfo_iterator_t pfi = 47240 xcb_render_query_pict_formats_formats_iterator(qpf); 47241 for (int i = 0; i < xcb_render_query_pict_formats_formats_length(qpf); i++) { 47242 47243 if (pfi.data->depth == 8 && 47244 pfi.data->type == XCB_RENDER_PICT_TYPE_DIRECT && 47245 pfi.data->direct.alpha_mask == 0xFF) { 47246 fmt_a8 = pfi.data->id; 47247 break; 47248 } 47249 xcb_render_pictforminfo_next(&pfi); 47250 } 47251 if (!fmt_a8) { 47252 fprintf(stderr, "Cannot find a8 RENDER picture format\n"); 47253 return EXIT_FAILURE; 47254 } 47255 47256 xcb_render_pictformat_t fmt_visual = 0; 47257 xcb_render_pictscreen_iterator_t psi = 47258 xcb_render_query_pict_formats_screens_iterator(qpf); 47259 for (int i = 0; i < xcb_render_query_pict_formats_screens_length(qpf); i++) { 47260 xcb_render_pictdepth_iterator_t pdi = 47261 xcb_render_pictscreen_depths_iterator(psi.data); 47262 for (int j = 0; i < xcb_render_pictscreen_depths_length(psi.data); i++) { 47263 xcb_render_pictvisual_iterator_t pvi = 47264 xcb_render_pictdepth_visuals_iterator(pdi.data); 47265 for (int k = 0; k < xcb_render_pictdepth_visuals_length(pdi.data); i++) { 47266 if (pvi.data->visual == visual) { 47267 fmt_visual = pvi.data->format; 47268 goto found_visual; 47269 } 47270 xcb_render_pictvisual_next(&pvi); 47271 } 47272 xcb_render_pictdepth_next(&pdi); 47273 } 47274 xcb_render_pictscreen_next(&psi); 47275 } 47276 found_visual: 47277 if (!fmt_visual) { 47278 fprintf(stderr, "Cannot find visual RENDER picture format\n"); 47279 return EXIT_FAILURE; 47280 } 47281 47282 xcb_render_glyphset_t glyphset = xcb_generate_id(c); 47283 xcb_render_create_glyph_set(c, glyphset, fmt_a8); 47284 uint32_t glyph_ids[] = {1}; 47285 xcb_render_add_glyphs(c, glyphset, 1, glyph_ids, 47286 &(xcb_render_glyphinfo_t){width:GLYPH_SIZE, height:GLYPH_SIZE}, sizeof(glyph), glyph); 47287 47288 // Create window, pixmap, and gc 47289 xcb_window_t window = xcb_generate_id(c); 47290 uint32_t list[] = { si.data->black_pixel, XCB_EVENT_MASK_EXPOSURE }; 47291 xcb_create_window(c, XCB_COPY_FROM_PARENT, window, root, 0, 0, WIN_SIZE, WIN_SIZE, 47292 0, XCB_WINDOW_CLASS_INPUT_OUTPUT, XCB_COPY_FROM_PARENT, 47293 XCB_CW_BACK_PIXEL | XCB_CW_EVENT_MASK, list); 47294 xcb_map_window(c, window); 47295 xcb_render_picture_t winpic = xcb_generate_id(c); 47296 xcb_render_create_picture(c, winpic, window, fmt_visual, 0, NULL); 47297 47298 xcb_pixmap_t pixmap = xcb_generate_id(c); 47299 xcb_create_pixmap(c, depth, pixmap, window, GLYPH_SIZE, GLYPH_SIZE); 47300 xcb_render_picture_t pixpic = xcb_generate_id(c); 47301 xcb_render_create_picture(c, pixpic, pixmap, fmt_visual, 0, NULL); 47302 xcb_render_fill_rectangles(c, XCB_RENDER_PICT_OP_SRC, pixpic, 47303 (xcb_render_color_t){green:0xFFFF, alpha:0xFFFF}, 1, 47304 &(xcb_rectangle_t){width:GLYPH_SIZE, height:GLYPH_SIZE} ); 47305 47306 xcb_flush(c); 47307 for (xcb_generic_event_t *ev = xcb_wait_for_event(c); ev; ev = xcb_wait_for_event(c)) { 47308 int type = ev->response_type; 47309 free(ev); 47310 if (type == XCB_EXPOSE) { 47311 xcb_clear_area(c, 0, window, 0, 0, 0, 0); 47312 xcb_render_composite_glyphs_8(c, XCB_RENDER_PICT_OP_SRC, pixpic, winpic, fmt_a8, 47313 glyphset, 0, 0, sizeof(elt), (uint8_t *)&elt); 47314 xcb_flush(c); 47315 } 47316 } 47317 47318 return EXIT_SUCCESS; 47319 } 47320 47321commit 80a0b29c3b61ba9024bd0278ea8cc51686afe3f4 47322Author: Michel Dänzer <michel.daenzer@amd.com> 47323Date: Fri Apr 11 11:01:51 2014 +0900 47324 47325 glamor: Add remaining header files to SOURCES 47326 47327 It wasn't possible to build glamor from tarballs. 47328 47329 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=64297#c9 47330 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 47331 Reviewed-by: Keith Packard <keithp@keithp.com> 47332 Signed-off-by: Keith Packard <keithp@keithp.com> 47333 47334commit ea15f8b4c90ebef21d4031bf7f412ef9241d5438 47335Author: Jasper St. Pierre <jstpierre@mecheye.net> 47336Date: Mon Apr 21 16:30:49 2014 -0400 47337 47338 dixfonts: Turn a missing directory ErrorF into a DebugF 47339 47340 On systems without these directories, we don't need to be complaining 47341 loudly. 47342 47343 Reviewed-by: Kristian Hoegsberg <krh@bitplanet.net> 47344 47345 Signed-off-by: Keith Packard <keithp@keithp.com> 47346 47347commit ba5c2b6c876a15fe18bc1ff1ab4e773c46bc9009 47348Merge: 9c86d5f4f 66b602474 47349Author: Keith Packard <keithp@keithp.com> 47350Date: Mon Apr 21 14:59:51 2014 -0700 47351 47352 Merge remote-tracking branch 'krh/xwayland-for-keithp' 47353 47354commit 9c86d5f4f6f5d6d135a9f2f8eae904c6441e84e5 47355Author: Kristian Høgsberg <krh@bitplanet.net> 47356Date: Tue Mar 25 14:57:12 2014 -0700 47357 47358 .gitignore: Add new autotools file 'test-driver' 47359 47360 Automake 1.12 introduces a new parallel test framework that uses a shell 47361 script helper and generates *.log and *.trs files. Add to .gitignore. 47362 47363 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 47364 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 47365 Signed-off-by: Keith Packard <keithp@keithp.com> 47366 47367commit 28fa5dd5cad3a5fb51018250dfa2b605a8aea85c 47368Author: Julien Cristau <jcristau@debian.org> 47369Date: Mon Apr 14 15:27:19 2014 +0200 47370 47371 configure: require xproto 7.0.26 for pointer typedef removal 47372 47373 Signed-off-by: Julien Cristau <jcristau@debian.org> 47374 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 47375 Signed-off-by: Keith Packard <keithp@keithp.com> 47376 47377commit d2274f2f46db300ac75ca0fdcba7f5655865b8b6 47378Author: Julien Cristau <jcristau@debian.org> 47379Date: Mon Apr 14 15:27:18 2014 +0200 47380 47381 configure: require fontsproto 2.1.3 for pointer typedef removal 47382 47383 Signed-off-by: Julien Cristau <jcristau@debian.org> 47384 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 47385 Signed-off-by: Keith Packard <keithp@keithp.com> 47386 47387commit 6d892ad112fc887e184f50c5dc2ba593668e9e11 47388Author: Robert Morell <rmorell@nvidia.com> 47389Date: Fri Apr 18 18:29:44 2014 -0700 47390 47391 randr: Implement RandR 1.4 request swapping 47392 47393 The protocol handlers all have support for swapping variable data and 47394 replies, but the top-level dispatch plumbing was missing. 47395 47396 Signed-off-by: Robert Morell <rmorell@nvidia.com> 47397 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 47398 Signed-off-by: Keith Packard <keithp@keithp.com> 47399 47400commit ed4ee7c34a7fbb58a449647fb8d968618c499a61 47401Author: Robert Morell <rmorell@nvidia.com> 47402Date: Fri Apr 18 18:29:43 2014 -0700 47403 47404 randr: Fix size checks for SetProvider* reqs 47405 47406 Both xRRSetProviderOutputSourceReq and xRRSetProviderOffloadSinkReq are 47407 fixed-size requests, so the length on the wire should match exactly. 47408 47409 Signed-off-by: Robert Morell <rmorell@nvidia.com> 47410 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 47411 Signed-off-by: Keith Packard <keithp@keithp.com> 47412 47413commit 668321e7e5acb41c0c7b299f57165e86dc74d7ed 47414Author: Robert Morell <rmorell@nvidia.com> 47415Date: Fri Apr 18 18:29:42 2014 -0700 47416 47417 randr: Fix crash for NULL swap dispatch procs 47418 47419 The previous code was checking the wrong table for function pointers. 47420 47421 Signed-off-by: Robert Morell <rmorell@nvidia.com> 47422 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 47423 Signed-off-by: Keith Packard <keithp@keithp.com> 47424 47425commit 66b602474047c499b8a888267a489790fc9f9d85 47426Author: Kristian Høgsberg <krh@bitplanet.net> 47427Date: Mon Apr 21 11:14:51 2014 -0700 47428 47429 xwayland: Remove left-over ErrorF logging 47430 47431 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 47432 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 47433 47434commit 138bf5ac9703b410a6066c303feea067680edf5a 47435Author: Jamey Sharp <jamey@minilop.net> 47436Date: Fri Apr 18 12:22:19 2014 -0700 47437 47438 DRI2SwapBuffers: Fix uninitialized target SBC. 47439 47440 Fixes Piglit test "swapbuffersmsc-return swap_interval 0". 47441 47442 Ensure that *swap_target gets initialized on any 'return Success' path, 47443 even if the swap request can't be completed by the driver and the server 47444 falls back to a simple blit. That path can also be triggered by setting 47445 swap_interval to 0, which disables sync to vertical retrace. 47446 47447 We originally found this bug because for some reason SDL2 automatically 47448 sets swap_interval to 0, when we were trying to test OML_sync_control in 47449 an SDL2 test application. We then discovered that the above-mentioned 47450 Piglit test has been failing for the same reason since it was 47451 introduced. 47452 47453 Signed-off-by: Jamey Sharp <jamey@minilop.net> 47454 Signed-off-by: Theo Hill <Theo0x48@gmail.com> 47455 Reviewed-by: Keith Packard <keithp@keithp.com> 47456 Reviewed-by: Eric Anholt <eric@anholt.net> 47457 Signed-off-by: Keith Packard <keithp@keithp.com> 47458 47459commit 4927af4475bc7d020190d9f468c90366525c1109 47460Author: Jamey Sharp <jamey@minilop.net> 47461Date: Fri Apr 18 12:22:18 2014 -0700 47462 47463 DRI2SwapBuffers: Don't reuse swap_target variable. 47464 47465 swap_target is an out-parameter that needs to be set to the value that 47466 SBC will take on after this SwapBuffers request completes. 47467 47468 However, it was also being used as a temporary variable to hold the MSC 47469 at which the SwapBuffers request got scheduled to occur. This confusion 47470 makes it harder to reason about whether swap_target is being set 47471 correctly for its out-parameter usage. (Hint: It isn't.) 47472 47473 For the latter use, it makes more sense to use the existing target_msc 47474 variable, which already has the right value unless target_msc, divisor, 47475 and remainder are all 0, in which case we can set it using swap_interval 47476 as usual. 47477 47478 Signed-off-by: Jamey Sharp <jamey@minilop.net> 47479 Signed-off-by: Theo Hill <Theo0x48@gmail.com> 47480 Reviewed-by: Keith Packard <keithp@keithp.com> 47481 Reviewed-by: Eric Anholt <eric@anholt.net> 47482 Signed-off-by: Keith Packard <keithp@keithp.com> 47483 47484commit d2c6e801b548674b8636e6a64b42c9038d46263c 47485Author: Kristian Høgsberg <krh@bitplanet.net> 47486Date: Tue Apr 8 09:23:42 2014 -0700 47487 47488 configure.ac: Remove check for WAYLAND_SCANNER_RULES 47489 47490 This makes configure fail if the wayland autoconf macros aren't found. 47491 We don't need the scanner for shm-only xwayland so just drop this line for 47492 now. 47493 47494 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 47495 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 47496 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 47497 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 47498 47499commit f618455c802abb77de8d64eb442422cfe5e26ea8 47500Author: Kristian Høgsberg <krh@bitplanet.net> 47501Date: Tue Apr 8 09:19:13 2014 -0700 47502 47503 xwayland: Build without xshmfence 47504 47505 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 47506 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 47507 47508commit 552d40b26a082df6f1816bb461793b8fa8b553c6 47509Author: Kristian Høgsberg <krh@bitplanet.net> 47510Date: Tue Apr 8 09:17:51 2014 -0700 47511 47512 xwayland: Build without GLX extension 47513 47514 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 47515 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 47516 47517commit c7011249d2abe6cc7af82ee4b79d8f6873444707 47518Author: Keith Packard <keithp@keithp.com> 47519Date: Fri Apr 18 15:12:14 2014 -0700 47520 47521 xkb: Verify reads of compiled keymap header and TOC 47522 47523 Check the return values from fread to make sure the elements are 47524 actually getting read from the file. 47525 47526 Signed-off-by: Keith Packard <keithp@keithp.com> 47527 Reviewed-by: Jamey Sharp <jamey@minilop.net> 47528 47529commit 0af8788579c2f52cc1172952c9004483bf863932 47530Author: Keith Packard <keithp@keithp.com> 47531Date: Fri Apr 18 15:09:50 2014 -0700 47532 47533 os: Ignore log file write failures 47534 47535 There's no place to log the message if writing to the log file fails, 47536 and we surely don't want to crash in that case, so just ignore errors 47537 and keep going. 47538 47539 Signed-off-by: Keith Packard <keithp@keithp.com> 47540 Reviewed-by: Jamey Sharp <jamey@minilop.net> 47541 47542commit 7abd28685066369ded807f59493c1159cfb286bf 47543Author: Keith Packard <keithp@keithp.com> 47544Date: Fri Apr 18 15:05:00 2014 -0700 47545 47546 os: Make sure that writing our pid to the lock file actually worked 47547 47548 There's no sense verifying that we can create the lock file and then 47549 ignoring the return value from write. 47550 47551 Signed-off-by: Keith Packard <keithp@keithp.com> 47552 Reviewed-by: Jamey Sharp <jamey@minilop.net> 47553 47554commit d72f691c0c9cace857975a6608a4cb431c8b6846 47555Author: Keith Packard <keithp@keithp.com> 47556Date: Fri Apr 18 15:00:30 2014 -0700 47557 47558 os: FatalError if -displayfd writes fail 47559 47560 When the server is started with the -displayfd option, check to make 47561 sure that the writes succeed and give up running if they don't. 47562 47563 Signed-off-by: Keith Packard <keithp@keithp.com> 47564 Reviewed-by: Jamey Sharp <jamey@minilop.net> 47565 47566commit 4957e986841225e9984daca76f1a0ee08df125bb 47567Author: Keith Packard <keithp@keithp.com> 47568Date: Fri Apr 18 15:00:35 2014 -0700 47569 47570 os: Clear the -displayfd option after closing the file 47571 47572 Failing to clear this means that we'll attempt to write the display 47573 number to a random file descriptor on subsequent X server generations. 47574 47575 Signed-off-by: Keith Packard <keithp@keithp.com> 47576 Reviewed-by: Jamey Sharp <jamey@minilop.net> 47577 47578commit 0c0feddbcda238efa82a47f456ef3008ffa53195 47579Author: Keith Packard <keithp@keithp.com> 47580Date: Fri Apr 18 14:54:02 2014 -0700 47581 47582 kdrive: Ignore failure to chown console tty to current user 47583 47584 I'm not sure what we'd do in this case anyways, other than fatal 47585 error. 47586 47587 Signed-off-by: Keith Packard <keithp@keithp.com> 47588 Reviewed-by: Jamey Sharp <jamey@minilop.net> 47589 47590commit 696e08f8e820449bb4b0ce7a60bf5b5bc5097935 47591Author: Keith Packard <keithp@keithp.com> 47592Date: Fri Apr 18 14:50:14 2014 -0700 47593 47594 kdrive: Explicitly ignore errors from the -switchCmd script 47595 47596 Make it clear that we intentionally ignore the -switchCmd return 47597 value. This keeps GCC from emitting a warning when the server is 47598 compiled with -D_FORTIFY_SOURCE=2. 47599 47600 Signed-off-by: Keith Packard <keithp@keithp.com> 47601 Reviewed-by: Jamey Sharp <jamey@minilop.net> 47602 47603commit 570b1c79942b237022be3594ae3a25e7b833cb73 47604Author: Keith Packard <keithp@keithp.com> 47605Date: Fri Apr 18 14:47:00 2014 -0700 47606 47607 test: [v2] Validate server log reading more carefully in signal-logging test 47608 47609 Check return value from fgets and strchr instead of assuming they 47610 worked. 47611 47612 [v2] 47613 47614 Don't do any necessary work inside the assert call. 47615 Also make sure the return value was long enough. 47616 47617 Signed-off-by: Keith Packard <keithp@keithp.com> 47618 Reviewed-by: Jamey Sharp <jamey@minilop.net> 47619 47620commit 3119aae80782b7ff5613c719fbcc8a01feab1134 47621Author: Gaetan Nadon <memsize@videotron.ca> 47622Date: Fri Apr 4 14:59:21 2014 -0400 47623 47624 test: create a link to the generated hw/xfree86/sdksyms.c at build time 47625 47626 Automake 1.14 gives us warning about source code specified in _SOURCES 47627 that comes from directories other than the current one. It suggests to enable 47628 the subdir-objects feature which only supports code in sub directories. 47629 47630 The test directory needs source from hw/xfree86 which is neither under test 47631 nor under a sub directory of test. In 1.14 we get a warning, in 2.0 it will 47632 break as it will overwrite the object code in xfree86. 47633 47634 The solution in this case is to create a link to hw/xfree86/sdksyms.c at build 47635 time. It's just like any other built source file. 47636 47637 There are no links created in git. 47638 47639 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 47640 Reviewed-by: Keith Packard <keithp@keithp.com> 47641 Signed-off-by: Keith Packard <keithp@keithp.com> 47642 47643commit 58ef01648b18f85ad79620da110ece2339f638c3 47644Author: Adam Jackson <ajax@redhat.com> 47645Date: Thu Apr 17 10:49:46 2014 -0400 47646 47647 dri1: Don't bother asking the loader about drmGetLibVersion 47648 47649 This was added for DRM_IOCTL_SET_VERSION support, which has been around 47650 for over ten years now. Since we require ≥2.3.0 in configure.ac this 47651 would really only protect you if you managed to build against a modern 47652 libdrm but run against one that's more than 7½ years old, which, doctor 47653 it hurts when I do this. 47654 47655 Archaeology: http://cgit.freedesktop.org/~ajax/dri/commit/xc/programs/Xserver/GL/dri/dri.c?id=77d62efca033dced96ab7998b7c62a4e2df907d5 47656 Signed-off-by: Adam Jackson <ajax@redhat.com> 47657 Reviewed-by: Eric Anholt <eric@anholt.net> 47658 Signed-off-by: Keith Packard <keithp@keithp.com> 47659 47660commit 461ed561b9f75cee43c0dfd39228716f977956e7 47661Author: Hans de Goede <hdegoede@redhat.com> 47662Date: Fri Apr 18 11:30:16 2014 +0200 47663 47664 Xorg.wrap manpages: use __appmansuffix__ instead of hardcoding 1 47665 47666 Cc: Gaetan Nadon <memsize@videotron.ca> 47667 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 47668 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 47669 47670commit 584961c86427b400a5ba65b0f145d00748d2da1e 47671Author: Hans de Goede <hdegoede@redhat.com> 47672Date: Fri Apr 18 11:26:00 2014 +0200 47673 47674 man/Makefile.am: Fix Xorg.wrap.man Xwrapper.config.man missing from make dist 47675 47676 Fix suggested by: Gaetan Nadon <memsize@videotron.ca> 47677 47678 Cc: Gaetan Nadon <memsize@videotron.ca> 47679 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 47680 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 47681 47682commit ec01d51a9973a9cf5f32f14f00058f1fdc9ed25e 47683Author: Guillem Jover <guillem@hadrons.org> 47684Date: Mon Apr 14 18:13:25 2014 +0200 47685 47686 Xorg.wrap: Make the console check portable 47687 47688 Handle the unported case by issuing a build-time and run-time warning. 47689 47690 And add support for FreeBSD kernel based systems, by using the 47691 VT_GETINDEX ioctl to check if the file descriptor is on a virtual 47692 console. 47693 47694 Signed-off-by: Guillem Jover <guillem@hadrons.org> 47695 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 47696 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 47697 47698commit 3a469917b585914ba2421e305f3b6a837b232e93 47699Author: Guillem Jover <guillem@hadrons.org> 47700Date: Mon Apr 14 18:13:24 2014 +0200 47701 47702 Xorg.wrap: Clarify error messages 47703 47704 Not printing the program name produces very confusing messages that 47705 might be difficult to attribute while trying to diagnose problems, 47706 let's be explicit about who we are. 47707 47708 Also add a missing "/" between SUID_WRAPPER_DIR and "Xorg.bin". 47709 47710 Signed-off-by: Guillem Jover <guillem@hadrons.org> 47711 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 47712 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 47713 47714commit 50b6e1b0d786d95ee8eab030b0d1fd7420e2fbeb 47715Author: Guillem Jover <guillem@hadrons.org> 47716Date: Mon Apr 14 18:13:23 2014 +0200 47717 47718 Xorg.wrap: Use <drm.h> instead of hardcoding libdrm include path 47719 47720 The libdrm.pc file gives us the correct include path, do not try to 47721 hardcode it on the source, as it might vary on the installed system, 47722 for example on Debian-based systems it's under /user/include/libdrm/. 47723 47724 Signed-off-by: Guillem Jover <guillem@hadrons.org> 47725 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 47726 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 47727 47728commit 7bffceb9f939c0bf5529b00500cbb14ae01bb61a 47729Author: Guillem Jover <guillem@hadrons.org> 47730Date: Mon Apr 14 18:13:22 2014 +0200 47731 47732 Xorg: Add Xorg.wrap to hw/xfree86/.gitignore 47733 47734 Signed-off-by: Guillem Jover <guillem@hadrons.org> 47735 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 47736 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 47737 47738commit 208157a76c6f340e0161e18ac77149d7110af2f5 47739Author: Julien Cristau <jcristau@debian.org> 47740Date: Sun Apr 13 19:14:34 2014 +0200 47741 47742 configure: fix help text for dmx 47743 47744 As of 93fa64e17d7bd600ebf18ecab85f5b2d17fe30ce it's disabled by default, 47745 so stop saying it's automatically enabled in configure --help. 47746 47747 Signed-off-by: Julien Cristau <jcristau@debian.org> 47748 Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> 47749 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 47750 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 47751 47752commit c917fa411502d402b9c69de30f6a88e102bb8610 47753Author: Keith Packard <keithp@keithp.com> 47754Date: Tue Apr 8 01:01:28 2014 -0700 47755 47756 glamor: Always allocate precisely the requested pixmap size 47757 47758 Using a pixmap as a tile or stipple means that we must have the 47759 underlying FBO match the pixmap geometry exactly. We may want to add 47760 some complexity here to migrate pixmaps into exact sized objects as 47761 necessary, but for now, make the server work correctly by skipping 47762 this optimization. 47763 47764 Signed-off-by: Keith Packard <keithp@keithp.com> 47765 Reviewed-by: Eric Anholt <eric@anholt.net> 47766 47767commit 3028ae6c9aa37168e249e0d847b29f8e3efb05b2 47768Merge: bd3de10f8 ec6007e6f 47769Author: Keith Packard <keithp@keithp.com> 47770Date: Wed Apr 9 21:59:17 2014 -0700 47771 47772 Merge remote-tracking branch 'jeremyhu/master' 47773 47774commit bd3de10f86a511886b0748d55fa19860314a0150 47775Author: Keith Packard <keithp@keithp.com> 47776Date: Tue Apr 8 14:25:18 2014 -0700 47777 47778 Bump to version 1.15.99.902. Closing the 1.16 merge window. 47779 47780 Signed-off-by: Keith Packard <keithp@keithp.com> 47781 47782commit 6ec04a75deb1695820e337abbefe1929d5928f15 47783Author: Keith Packard <keithp@keithp.com> 47784Date: Fri Apr 4 16:28:43 2014 -0700 47785 47786 dri3: Fix dri3_open API change by adding new dri3_open_client 47787 47788 Xwayland will eventually need the current client in dri3_open. Simply 47789 changing that API is not an option though as other drivers that 47790 implement DRI3 will not have a matching function signature and will 47791 crash when called. 47792 47793 Add a new dri3_open_client function pointer and bump 47794 DRI3_SCREEN_INFO_VERSION so that drivers can be aware of the new 47795 function which will be used in preference to the old function when 47796 available. 47797 47798 Signed-off-by: Keith Packard <keithp@keithp.com> 47799 Reviewed-by: Eric Anhole <eric@anholt.net> 47800 47801commit ec6007e6f7772a90713c9c51c64359229961ce27 47802Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 47803Date: Sun Apr 6 05:32:00 2014 -0700 47804 47805 XQuartz: Ensure we wait for the server thread to terminate 47806 47807 AKA: XQuartz 2.7.5 doesn't delete its /tmp/.X$d-lock 47808 47809 http://xquartz.macosforge.org/trac/ticket/823 47810 47811 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 47812 47813commit 901fbfbbbd71c0d82080957f8ba09eebbc786f2b 47814Author: Michael Thayer <michael.thayer@oracle.com> 47815Date: Mon Mar 31 11:21:42 2014 +0200 47816 47817 Add a return value to load_cursor_argb() to allow it to report failure 47818 47819 load_cursor_argb() may need to be able to fail and have the server fall back 47820 to a software cursor in at least the following circumstances. 47821 1) The hardware can only support some ARGB cursors and this does not just 47822 depend on cursor size. 47823 2) Virtual hardware may not wish to pass through a cursor to the host at a 47824 particular time but may wish to accept the same cursor at another time. 47825 This patch adds a return value to the API and makes the server do the 47826 software fall-back on failure. 47827 47828 Signed-off-by: Michael Thayer <michael.thayer@oracle.com> 47829 Reviewed-by: Dave Airlie <airlied@redhat.com> 47830 Signed-off-by: Keith Packard <keithp@keithp.com> 47831 47832commit 62ab4102260fd3342a0e5ae3a4f77b430af64f4e 47833Author: Michael Thayer <michael.thayer@oracle.com> 47834Date: Mon Mar 31 11:19:16 2014 +0200 47835 47836 Set a flag property on the root window to say if the X server VT is active 47837 47838 An X11 client may need to know whether the X server virtual terminal is 47839 currently the active one. This change adds a root window property which 47840 provides that information. Intended interface user: the VirtualBox Guest 47841 Additions. 47842 47843 Signed-off-by: Michael Thayer <michael.thayer@oracle.com> 47844 Reviewed-by: Dave Airlie <airlied@redhat.com> 47845 Signed-off-by: Keith Packard <keithp@keithp.com> 47846 47847commit b851ca968b7cce6d1a6438c05d3d5c8832249704 47848Author: Daniel Stone <daniels@collabora.com> 47849Date: Thu Apr 3 18:22:50 2014 +0100 47850 47851 Add pScreen->NameWindowPixmap hook 47852 47853 This hook allows drivers to be notified when a pixmap gains a new ID. 47854 47855 (ABI break.) 47856 47857 Signed-off-by: Daniel Stone <daniels@collabora.com> 47858 Reviewed-by: Adam Jackson <ajax@redhat.com> 47859 Signed-off-by: Keith Packard <keithp@keithp.com> 47860 47861commit b4d0bec22c15930abf13a7fb9d684208ccd56b1d 47862Merge: 3c34dd360 6e539d881 47863Author: Keith Packard <keithp@keithp.com> 47864Date: Thu Apr 3 15:35:01 2014 -0700 47865 47866 Merge remote-tracking branch 'krh/xwayland-for-keithp' 47867 47868commit 6e539d8817f738289dc2dea13d0720116287ab9d 47869Author: Kristian Høgsberg <krh@bitplanet.net> 47870Date: Tue Mar 11 16:11:39 2014 -0700 47871 47872 Xwayland DDX 47873 47874 Started out as an Xorg module to be used from Xorg drivers to let 47875 Xorg run under a wayland server. The idea was to be able to reuse the 47876 2D acceleration from the Xorg driver. Now with glamor being credible, 47877 a better plan is to just make Xwayland its own DDX, similar to Xwin 47878 and Xquartz. This is a much better fit, as much of the code in the 47879 original approach had to hack around Xorg doing Xorg things like take 47880 over the VT, probe input devices and read config files. Another big win 47881 is that Xwayland dosn't need to be setuid root. 47882 47883 The Xwayland support for DRI3, Glamor and render nodes was done by 47884 Axel Davy <axel.davy@ens.fr>, who also did a lot of work on the rebase 47885 to the Xwayland DDX. 47886 47887 Contributions from: 47888 47889 Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 47890 Corentin Chary <corentin.chary@gmail.com> 47891 Daniel Stone <daniel@fooishbar.org> 47892 Kristian Høgsberg <krh@bitplanet.net> 47893 Robert Bragg <robert@linux.intel.com> 47894 Scott Moreau <oreaus@gmail.com> 47895 Tiago Vignatti <tiago.vignatti@intel.com> 47896 Giovanni Campagna <gcampagn@redhat.com> 47897 Jonas Ådahl <jadahl@gmail.com> 47898 Ray Strode <rstrode@redhat.com> 47899 Trevor McCort <tjmccort@gmail.com> 47900 Rui Matos <tiagomatos@gmail.com> 47901 Axel Davy <axel.davy@ens.fr> 47902 Jasper St. Pierre <jstpierre@mecheye.net> 47903 47904 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 47905 Reviewed-by: Axel Davy <axel.davy@ens.fr> 47906 47907commit 3c34dd3603989c0365654ca1b6809395c7f3b169 47908Author: Keith Packard <keithp@keithp.com> 47909Date: Tue Apr 1 21:15:48 2014 -0700 47910 47911 glamor: Add glamor_program based poly_text and image_text 47912 47913 Accelerates text painting with GPU-based geometry computation and stippling 47914 47915 v2: Simplify get_glyphs, expand single character variable names to 47916 more descriptive ones. (Markus Wick) 47917 v3: Rebase against the glamor_prepare_* un-renaming (changes by anholt). 47918 47919 Improves x11perf -f8text by 417.908% +/- 11.0144% (n=10) 47920 47921 Signed-off-by: Keith Packard <keithp@keithp.com> 47922 Signed-off-by: Eric Anholt <eric@anholt.net> 47923 Reviewed-by: Eric Anholt <eric@anholt.net> 47924 47925commit 5f700c3ac31db52f02f6ef11ea7823a8ce4f6f6f 47926Author: Keith Packard <keithp@keithp.com> 47927Date: Tue Apr 1 21:15:47 2014 -0700 47928 47929 glamor: Use glamor_program for glamor_glyphblt 47930 47931 This constructs suitable shaders using the glamor_program 47932 infrastructure for poly glyph blt, and then gets rid of the no-op 47933 wrapper of miImageGlyphBlt. 47934 47935 Improves x11perf -f8text by 11.6221% +/- 1.04585% (n=10) 47936 47937 Signed-off-by: Keith Packard <keithp@keithp.com> 47938 Signed-off-by: Eric Anholt <eric@anholt.net> 47939 Reviewed-by: Eric Anholt <eric@anholt.net> 47940 47941commit 3411e8c538476443cafbb70082ef7457932e4a38 47942Author: Keith Packard <keithp@keithp.com> 47943Date: Tue Apr 1 21:15:46 2014 -0700 47944 47945 glamor: Add glamor_program based poly_fill_rect 47946 47947 This accelerates poly_fill_rect using GPU-based geometry computation 47948 47949 Improves x11perf -rect100 by 41.5127% +/- 7.63888% (n=10) 47950 Improves x11perf -rect10 by 3745.72% +/- 94.7503% (n=6) 47951 47952 v2: Rebase on skipping the prepare rewrite for now, and fix the GLSL 47953 1.20 and GLES2 cases (changes by anholt). 47954 47955 Signed-off-by: Keith Packard <keithp@keithp.com> 47956 Signed-off-by: Eric Anholt <eric@anholt.net> 47957 Reviewed-by: Eric Anholt <eric@anholt.net> 47958 47959commit 0a6d31161811c422799d6f065ea2842d42707530 47960Author: Keith Packard <keithp@keithp.com> 47961Date: Tue Apr 1 21:15:45 2014 -0700 47962 47963 glamor: Add glamor_program based fill/set/get spans 47964 47965 This accelerates spans operations using GPU-based geometry computation 47966 47967 -wellipse500 goes from about 4k/sec before the patch, to ~8k/sec in 47968 the GLES2 fallback loop, to ~100k/sec in desktop mode. 47969 47970 v2: Rebase on skipping the prepare rewrite for now, and fix the GLSL 47971 1.20 and GLES2 cases (changes by anholt). 47972 47973 Signed-off-by: Keith Packard <keithp@keithp.com> 47974 Signed-off-by: Eric Anholt <eric@anholt.net> 47975 Reviewed-by: Eric Anholt <eric@anholt.net> 47976 47977commit d78c257f523908760c1b872cc4bef3d42eb31f50 47978Author: Keith Packard <keithp@keithp.com> 47979Date: Tue Apr 1 21:15:43 2014 -0700 47980 47981 glamor: Add simple upload/download functions in glamor_transfer 47982 47983 These use glTexSubimage2D for upload and glReadPixels for 47984 download. There are a variety of interfaces to the basic function as 47985 needed by the callers. 47986 47987 Signed-off-by: Keith Packard <keithp@keithp.com> 47988 Signed-off-by: Eric Anholt <eric@anholt.net> 47989 47990commit 693e6bea897e4c9473a9f22dd8b90df3437e04a1 47991Author: Keith Packard <keithp@keithp.com> 47992Date: Tue Apr 1 21:15:42 2014 -0700 47993 47994 glamor: Use plain GLSL 1.20 features for fill code. 47995 47996 This prevents performance regressions from losing acceleration support 47997 on older hardware as we transition to using glamor_program.c for 47998 acceleration. 47999 48000 Signed-off-by: Keith Packard <keithp@keithp.com> 48001 Signed-off-by: Eric Anholt <eric@anholt.net> 48002 Reviewed-by: Eric Anholt <eric@anholt.net> 48003 48004commit 5b76b3978d792c92b23a691204b473209782476e 48005Author: Keith Packard <keithp@keithp.com> 48006Date: Tue Apr 1 21:15:42 2014 -0700 48007 48008 glamor: Compute supported GLSL version and save in screen private 48009 48010 This currently computes the GLSL version in a fairly naïve fashion, 48011 and leaves that in the screen private for other users. This will let 48012 us update the version computation in one place later on. 48013 48014 v2: Drop an accidental rebase-squashed hunk (change by anholt). 48015 48016 Signed-off-by: Keith Packard <keithp@keithp.com> 48017 Signed-off-by: Eric Anholt <eric@anholt.net> 48018 Reviewed-by: Eric Anholt <eric@anholt.net> 48019 48020commit 3c2c59eed3c68c0e5a93c38cf01eedad015e3157 48021Author: Keith Packard <keithp@keithp.com> 48022Date: Tue Apr 1 21:15:41 2014 -0700 48023 48024 fb: Publish fbGlyphs and fbUnrealizeGlyph 48025 48026 This lets other code jump directly into the fb code for fallbacks 48027 48028 Signed-off-by: Keith Packard <keithp@keithp.com> 48029 Signed-off-by: Eric Anholt <eric@anholt.net> 48030 Reviewed-by: Eric Anholt <eric@anholt.net> 48031 48032commit 9dffcda804b2609f0b319d35dd47fd401d4fafce 48033Author: Keith Packard <keithp@keithp.com> 48034Date: Tue Apr 1 20:54:22 2014 -0700 48035 48036 glamor: Bail from composite when pixmap cannot be uploaded 48037 48038 I think the sense of the return value was just flipped here; if you 48039 return TRUE, then the calling code assumes that the pixmap *has* been 48040 uploaded and that an FBO is available. When it tries to use it, it 48041 crashes though. Returning false makes the caller bail back to software. 48042 48043 Signed-off-by: Keith Packard <keithp@keithp.com> 48044 Signed-off-by: Eric Anholt <eric@anholt.net> 48045 Reviewed-by: Eric Anholt <eric@anholt.net> 48046 48047commit 15fec3bc3a2d29c903f35b5a30ac09e196eb36b1 48048Author: Keith Packard <keithp@keithp.com> 48049Date: Tue Apr 1 20:54:21 2014 -0700 48050 48051 glamor: Remove warning message when pixmap cannot be stored in a texture 48052 48053 This happens when you have 4bpp pixmaps; it's not an error, so stop 48054 flooding the log file when it happens. 48055 48056 Signed-off-by: Keith Packard <keithp@keithp.com> 48057 Signed-off-by: Eric Anholt <eric@anholt.net> 48058 Reviewed-by: Eric Anholt <eric@anholt.net> 48059 48060commit 5f177f3582323b6c1791e9e2c107548b607369de 48061Author: Keith Packard <keithp@keithp.com> 48062Date: Tue Apr 1 20:54:20 2014 -0700 48063 48064 glamor: Public polyLines function is glamor_poly_lines_nf 48065 48066 There was a spurious declaratoin in glamor.h for glamor_poly_line_nf 48067 48068 Signed-off-by: Keith Packard <keithp@keithp.com> 48069 Signed-off-by: Eric Anholt <eric@anholt.net> 48070 48071commit 6aac97198f60de9d4e415fb4fa2fc11cac6c6790 48072Author: Keith Packard <keithp@keithp.com> 48073Date: Tue Apr 1 20:54:19 2014 -0700 48074 48075 glamor: Initialize XV shaders from glamor_xv_init instead of glamor_init 48076 48077 The glamor_init calls to glamor_init_xv_shader were never getting run 48078 because GLAMOR_XV was never defined. Instead of trying to make that 48079 work, fix glamor_xv_init to make the call instead. 48080 48081 Further, just get rid of the glamor_fini_xv_shader function entirely 48082 as the shader program will be destroyed when the context is destroyed 48083 at server reset time. 48084 48085 Signed-off-by: Keith Packard <keithp@keithp.com> 48086 Signed-off-by: Eric Anholt <eric@anholt.net> 48087 Reviewed-by: Eric Anholt <eric@anholt.net> 48088 48089commit 1707faf36d49a63efc462dfef133a555fe093832 48090Author: Keith Packard <keithp@keithp.com> 48091Date: Tue Apr 1 20:54:18 2014 -0700 48092 48093 glamor: SetWindowPixmap is not related to RENDER 48094 48095 Move the configuration of screen->SetWindowPixmap out from under it. 48096 48097 Signed-off-by: Keith Packard <keithp@keithp.com> 48098 Signed-off-by: Eric Anholt <eric@anholt.net> 48099 Reviewed-by: Eric Anholt <eric@anholt.net> 48100 48101commit 8d88b90953471ead3b480b6ae9d7280b8bd6b216 48102Author: Keith Packard <keithp@keithp.com> 48103Date: Tue Apr 1 20:54:15 2014 -0700 48104 48105 glamor: glamor_poly_point_nf cannot fail for non-DDX pixmaps 48106 48107 All of the glamor _nf functions must check to see if the DDX can 48108 access the pixmap directly before returning failure back to the 48109 driver; this restructures the point code to split out the _nf checking 48110 from the _gl code. 48111 48112 Signed-off-by: Keith Packard <keithp@keithp.com> 48113 Signed-off-by: Eric Anholt <eric@anholt.net> 48114 Reviewed-by: Eric Anholt <eric@anholt.net> 48115 48116commit d5dd1d6371021c2aad94e33c5837ccd4bf9bf0f4 48117Author: Eric Anholt <eric@anholt.net> 48118Date: Fri Mar 21 13:13:27 2014 -0700 48119 48120 Revert "glx: Make sure we get an FBConfig the root window's visual." 48121 48122 This reverts commit b5a61239e2fef167c229154d7919ff862503e3f3. 48123 48124 Not only did I screw up and introduce a warning, it turns out 48125 glXChooseFBConfig() explicitly ignores this attribute. Thanks, GLX. 48126 48127 Signed-off-by: Eric Anholt <eric@anholt.net> 48128 Reviewed-by: Julien Cristau <jcristau@debian.org> 48129 48130commit 84f977467b514af88019ab2791bf7a74530b54df 48131Merge: 9d20d18fb 98924719d 48132Author: Keith Packard <keithp@keithp.com> 48133Date: Thu Apr 3 10:44:28 2014 -0700 48134 48135 Merge remote-tracking branch 'whot/for-keith' 48136 48137commit 98924719d524bf87cdf301063cd744d1271c33ff 48138Author: Peter Hutterer <peter.hutterer@who-t.net> 48139Date: Wed Apr 2 13:55:10 2014 +1000 48140 48141 Revert "xkb: ProcesssPointerEvent must work on the VCP if it gets the VCP" 48142 48143 This was the wrong fix to the problem, and it triggered a change in XKB 48144 behavior: previously a button event would unlock a latched modifier, now it 48145 doesn't anymore. 48146 https://bugs.freedesktop.org/show_bug.cgi?id=73155 48147 48148 Note that the new behavior is is strictly spec compliant but we've had the 48149 other behavior for a long time so we shouldn't break it. 48150 48151 The bug this patch originally fixed was a null-pointer dereference when 48152 releasing button events on server shutdown. This was addressed by the commit 48153 below, so the need for this patch has gone away anyway. 48154 48155 commit 3e4be4033aed78b2bb3a18d51f0963989efd1af3 48156 Author: Peter Hutterer <peter.hutterer@who-t.net> 48157 Date: Fri Jan 25 11:47:32 2013 +1000 48158 48159 dix: when shutting down slave devices, shut down xtest devices last 48160 48161 This reverts commit 2decff6393a44b56d80d53570718f95354fde454. 48162 48163 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48164 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 48165 48166commit 9d65c515d83d2158b5949e249777ca2a02b31901 48167Author: Hans de Goede <hdegoede@redhat.com> 48168Date: Tue Apr 1 11:24:17 2014 +0200 48169 48170 xf86LogInit: log to XDG_DATA_HOME when not running as root 48171 48172 When no logfile was specified (xf86LogFileFrom == X_DEFAULT) and we're not 48173 running as root log to $XDG_DATA_HOME/xorg/Xorg.#.log as Xorg won't be able to 48174 log to the default /var/log/... when it is not running as root. 48175 48176 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 48177 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48178 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48179 48180commit e738276e96590b2230dd9d9f5a3f9485f592e1e6 48181Author: Kristian Høgsberg <krh@bitplanet.net> 48182Date: Mon Mar 31 18:04:07 2014 -0700 48183 48184 glamor: Expose glamor_destroy_pixmap() 48185 48186 When we create a glamor pixmap by calling glamor_create_pixmap() 48187 directly, we need to call glamor_destroy_pixmap() to destroy it. 48188 48189 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 48190 Reviewed-by: Keith Packard <keithp@keithp.com> 48191 48192commit d14803437582668f5ffea14071ea962b6b8e450d 48193Author: Kristian Høgsberg <krh@bitplanet.net> 48194Date: Mon Mar 31 18:00:54 2014 -0700 48195 48196 glamor: Add new GLAMOR_CREATE_PIXMAP_NO_TEXTURE pixmap create flag 48197 48198 This flag lets a DDX allocate a glamor pixmap without allocating the 48199 texture that backs it. The DDX can then allocate the texture itself 48200 and then set it later. 48201 48202 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 48203 Reviewed-by: Keith Packard <keithp@keithp.com> 48204 48205commit fe204185bc801bb17785dc8d1cde09d48b5e257f 48206Author: Kristian Høgsberg <krh@bitplanet.net> 48207Date: Mon Mar 31 17:59:18 2014 -0700 48208 48209 glamor: Move glamor_egl_screen_init() prototype to glamor.h 48210 48211 A DDX that implements the glamor EGL functions need to pull in this 48212 prototype but shouldn't need to pull in glamor_priv.h 48213 48214 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 48215 Reviewed-by: Keith Packard <keithp@keithp.com> 48216 48217commit 9ba2084321b1ac8d271feefee4f5b968d34226d6 48218Author: Kristian Høgsberg <krh@bitplanet.net> 48219Date: Thu Mar 27 23:47:59 2014 -0700 48220 48221 dri3: Allow asynchronous implementation for dri3_open 48222 48223 By passing the client pointer to the dri3_open implementation, we allow 48224 the clients to implement the open callback asynchronously. If the 48225 client ignore count is positive after returning from dri3_open, we 48226 assume that authentication is in progress and doesn't send the reply. 48227 The code to send the reply is moved into a helper function, which the 48228 implementation can call upon receiving its authenticaion reply. 48229 48230 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 48231 Reviewed-by: Keith Packard <keithp@keithp.com> 48232 48233commit 4ba7b594464887ae68623e150348d90812eec36c 48234Author: Kristian Høgsberg <krh@bitplanet.net> 48235Date: Thu Mar 27 22:15:28 2014 -0700 48236 48237 composite: Add exception mechanism for implicit redirection policy 48238 48239 Normally composite will decide to add implicit redirection when a 48240 window with an alternate visual is a parent of a window with a regular 48241 visual or vice versa. This uses extra pixmap memory and incurs an extra 48242 copy. This exception mechanism provides a way for a DDX to override this 48243 if the DDX knows that its acceleration architecture will render correctly. 48244 48245 The relevant case is that of an RGB window reparented into a ARGB parent 48246 frame window. If the DDX knows that the acceleration architecture in use 48247 will pad the alpha channel to opaque when rendering to the RGB window, 48248 the implicit redirection can be avoided. 48249 48250 This patch adds a new composite function: 48251 48252 CompositeRegisterImplicitRedirectionException() 48253 48254 which lets a DDX register a pair of parent and child window visuals, that 48255 will not be implicitly redirected even if the normal policy would have 48256 made that choice. 48257 48258 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 48259 Reviewed-by: Keith Packard <keithp@keithp.com> 48260 48261commit 73698d41e41ce76bef2d9a90b46ac0c24ae148dd 48262Author: Keith Packard <keithp@keithp.com> 48263Date: Mon Mar 31 23:55:25 2014 -0700 48264 48265 Make XYToWindow a screen function 48266 48267 This allows DDXen to override the window picking to account for 48268 native windows not seen by the X server. The bulk of the picking logic 48269 is exposed as a new helper function, miSpriteTrace(). This function 48270 completes the sprite trace filled out by the caller, and can be set up 48271 to start the search from a given toplevel window. 48272 48273 v2: Leave existing XYToWindow API in place for API compatibility 48274 48275 Signed-off-by: Keith Packard <keithp@keithp.com> 48276 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 48277 48278commit f37a46913489397d8628ffe578c8d4ed50b6ca72 48279Author: Hans de Goede <hdegoede@redhat.com> 48280Date: Wed Mar 26 12:24:49 2014 +0100 48281 48282 configure: Change DEFAULT_LOGPREFIX to really be a filename prefix 48283 48284 Rather then a full path prefix, this is a preparation patch for adding 48285 support for logging to another location when not running as root. 48286 48287 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 48288 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48289 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48290 48291commit 9d20d18fb9dcc74bfa5392a2da40fd41b3e640d3 48292Merge: 8468e2443 bc348bd2c 48293Author: Keith Packard <keithp@keithp.com> 48294Date: Mon Mar 31 23:24:56 2014 -0700 48295 48296 Merge remote-tracking branch 'jturney/master' 48297 48298commit 8468e24430988058b150dc1daef3aad6ba9fa698 48299Author: Adam Jackson <ajax@redhat.com> 48300Date: Thu Mar 27 11:48:39 2014 -0400 48301 48302 mi: Hush extension initialization (#75870) 48303 48304 Printing these as ErrorF is fairly obnoxious, since it means the 48305 non-hardware servers now spew stuff to the console for entirely routine 48306 events. And actually, printing these at all is fairly obnoxious, since 48307 a) we're printing a line for every extension, whether it's enabled or 48308 not, and b) we're not actually initializing the extension at this point. 48309 48310 Signed-off-by: Adam Jackson <ajax@redhat.com> 48311 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 48312 Signed-off-by: Keith Packard <keithp@keithp.com> 48313 48314commit a2880699e8f1f576e1a48ebf25e8982463323f84 48315Author: Keith Packard <keithp@keithp.com> 48316Date: Tue Mar 25 08:21:16 2014 -0700 48317 48318 fb: fix fast-path blt detection 48319 48320 The width parameter is used to disable the blit fast-path (memcpy) when 48321 source and destination rows overlap in memory. This check was added in [0]. 48322 48323 Unfortunately, the calculation to determine if source and destination 48324 lines overlapped was incorrect: 48325 (1) it converts width from pixels to bytes, but width is actually in 48326 bits, not pixels. 48327 (2) it adds this byte offset to dst/srcLine, which implicitly converts 48328 the offset from bytes to sizeof(FbBits). 48329 48330 Fix both of these by converting addresses to byte pointers and width 48331 to bytes and doing comparisons on the resulting byte address. 48332 48333 For example: 48334 A 32-bpp 1366 pixel-wide row will have 48335 width = 1366 * 32 = 43712 bits 48336 bpp = 32 48337 (bpp >> 3) = 4 48338 width * (bpp >> 3) = 174848 FbBits 48339 (FbBits *)width => 699392 bytes 48340 48341 So, "careful" was true if the destination line was within 699392 bytes, 48342 instead of just within its 1366 * 4 = 5464 byte row. 48343 48344 This bug causes us to take the slow path for large non-overlapping rows 48345 that are "close" in memory. As a data point, XGetImage(1366x768) on my 48346 ARM chromebook was taking ~140 ms, but with this fixed, it now takes 48347 about 60 ms. 48348 XGetImage() -> exaGetImage() -> fbGetImage -> fbBlt() 48349 48350 [0] commit e32cc0b4c85c78cd8743a6e1680dcc79054b57ce 48351 Author: Adam Jackson <ajax@redhat.com> 48352 Date: Thu Apr 21 16:37:11 2011 -0400 48353 48354 fb: Fix memcpy abuse 48355 48356 The memcpy fast path implicitly assumes that the copy walks 48357 left-to-right. That's not something memcpy guarantees, and newer glibc 48358 on some processors will indeed break that assumption. Since we walk a 48359 line at a time, check the source and destination against the width of 48360 the blit to determine whether we can be sloppy enough to allow memcpy. 48361 (Having done this, we can remove the check for !reverse as well.) 48362 48363 v3: Convert to byte units 48364 48365 This first checks to make sure the blt is byte aligned, converts all 48366 of the data to byte units and then compares for byte address range 48367 overlap between source and dest. 48368 48369 Signed-off-by: Keith Packard <keithp@keithp.com> 48370 Reviewed-by: Daniel Kurtz <djkurtz@chromium.org> 48371 48372commit bc348bd2c42f3f18786085ccef2f010eff5bf3d2 48373Author: Jon TURNEY <jon.turney@dronecode.org.uk> 48374Date: Mon Mar 11 14:34:32 2013 +0000 48375 48376 Handle -displayfd and an explicit display number sensibly 48377 48378 Handle -displayfd and an explicit display number sensibly, e.g. use the 48379 explicitly specified display number, and write it to the displayfd 48380 48381 v2: displayfd might be 0, so use -1 as invalid value 48382 v3: Rebase for addition of NoListenAll flag 48383 48384 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 48385 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 48386 48387commit a3730271d52542757ec0f41acc29e7730501f666 48388Author: Jon TURNEY <jon.turney@dronecode.org.uk> 48389Date: Mon Mar 24 17:32:10 2014 +0000 48390 48391 Fix build when configured --enable-debug 48392 48393 Include os.h for ErrorF() to fix implicit-function-declaration warnings when 48394 configured with --enable-debug. 48395 48396 hw/xfree86/parser/DRI.c: In function 'xf86parseDRISection': 48397 hw/xfree86/parser/DRI.c:87:5: error: implicit declaration of function 'ErrorF' [-Werror=implicit-function-declaration] 48398 hw/xfree86/parser/Extensions.c: In function 'xf86parseExtensionsSection': 48399 hw/xfree86/parser/Extensions.c:77:5: error: implicit declaration of function 'ErrorF' [-Werror=implicit-function-declaration] 48400 48401 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 48402 Reviewed-by: Julien Cristau <jcristau@debian.org> 48403 48404commit 5870bd398d6073938d821c060eccf24adc07a2f1 48405Author: Jon TURNEY <jon.turney@dronecode.org.uk> 48406Date: Mon Mar 24 17:32:09 2014 +0000 48407 48408 Fix ephyr build with --disable-glamor 48409 48410 See http://tinderbox.x.org/builds/2014-03-23-0010/logs/xserver/#build 48411 48412 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 48413 Tested-by: Julien Cristau <jcristau@debian.org> 48414 Reviewed-by: Julien Cristau <jcristau@debian.org> 48415 48416commit 432fc37e2ca833f1ec3cb614cbdc3a7ae13f22b7 48417Author: Jon TURNEY <jon.turney@dronecode.org.uk> 48418Date: Mon Feb 24 20:45:51 2014 +0000 48419 48420 Build fbcmap_mi.c once, rather than once for each DDX 48421 48422 Build fbcmap_mi.c once, rather than once for each DDX, and make it part of libfb 48423 or libwfb convenience library. 48424 48425 Since 84e8de1271bb11b5b4b9747ae4647f47333a8ab7 we don't have fbcmap.c 48426 48427 This is a sort of revert of 17d85387d1e6851d35474b65929e268ca64ef65b 48428 48429 v2: Remove libkdrivestubs.la from configure.ac 48430 48431 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 48432 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 48433 48434commit f23dd02b969089b66685f70887f19041049f7a24 48435Author: Jon TURNEY <jon.turney@dronecode.org.uk> 48436Date: Fri Feb 21 18:11:33 2014 +0000 48437 48438 Build Xi/stubs.c once as a convenience library, rather than once for each DDX which wants to use it 48439 48440 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 48441 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 48442 48443commit 45b223e1c2bdc447263a9de7af62c93f4058c33c 48444Author: Jon TURNEY <jon.turney@dronecode.org.uk> 48445Date: Wed Sep 7 12:50:20 2011 +0100 48446 48447 Build dpmsstubs.c once as a convenience library, rather than once for each DDX which wants to use it 48448 48449 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 48450 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 48451 48452commit 7b2a517ba99f6756e98c4aef47d9b9399b997157 48453Author: Julien Cristau <jcristau@debian.org> 48454Date: Wed Mar 26 23:24:20 2014 +0100 48455 48456 Xephyr: restore initial window resize lost in xcb conversion 48457 48458 The XResizeWindow call wasn't replaced by the xcb equivalent, so we 48459 were no longer setting the initial window size, only wm size hints. 48460 48461 Regression from commit a2b73da "Xephyr: start converting hostx.c over to 48462 xcb" 48463 48464 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=74849 48465 48466 Signed-off-by: Julien Cristau <jcristau@debian.org> 48467 Reported-by: Laércio de Sousa <lbsousajr@gmail.com> 48468 Tested-by: Jon TURNEY <jon.turney@dronecode.org.uk> 48469 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 48470 Signed-off-by: Keith Packard <keithp@keithp.com> 48471 48472commit 1b5d7e78460ed686ba2da398dab341d4867fd4a0 48473Author: Keith Packard <keithp@keithp.com> 48474Date: Wed Mar 12 14:35:27 2014 -0700 48475 48476 glamor: Add glamor_program PolyPoint implementation 48477 48478 This accelerates poly point when possible by off-loading all geometry 48479 computation to the GPU. 48480 48481 Improves x11perf -dot performance by 28109.5% +/- 1022.01% (n=3) 48482 48483 Signed-off-by: Keith Packard <keithp@keithp.com> 48484 Signed-off-by: Eric Anholt <eric@anholt.net> 48485 Reviewed-by: Eric Anholt <eric@anholt.net> 48486 48487commit 72a4beff6d3aed767df9a30bca0c1d860715928f 48488Author: Keith Packard <keithp@keithp.com> 48489Date: Sat Mar 22 15:20:49 2014 -0700 48490 48491 glamor: Move glamor_poly_segment to separate glamor_segment.c file 48492 48493 There's no reason to mix PolyPoint and PolySegment in the same file. 48494 48495 Signed-off-by: Keith Packard <keithp@keithp.com> 48496 Signed-off-by: Eric Anholt <eric@anholt.net> 48497 Reviewed-off-by: Eric Anholt <eric@anholt.net> 48498 48499commit 029b64c30a8bdf87edfdb1c988f4e8dcee1bb4ef 48500Author: Keith Packard <keithp@keithp.com> 48501Date: Wed Mar 12 14:27:59 2014 -0700 48502 48503 glamor: Add infrastructure for generating shaders on the fly 48504 48505 This just adds a bunch of support code to construct shaders from 48506 'facets', which bundle attributes needed for each layer of the 48507 rendering system. At this point, that includes only the primitive and 48508 the fill stuff. 48509 48510 v2: Correct comment in glamor transform about 1/2 pixel correction needed 48511 for GL_POINT. (Eric Anholt) 48512 v3: Rebase on Markus's cleanups (change by anholt) 48513 48514 Signed-off-by: Keith Packard <keithp@keithp.com> 48515 Signed-off-by: Eric Anholt <eric@anholt.net> 48516 Reviewed-by: Eric Anholt <eric@anholt.net> 48517 48518commit 0ca7223742e9ec0594203b3a99b11441730cca1a 48519Author: Keith Packard <keithp@keithp.com> 48520Date: Thu Mar 13 23:12:27 2014 -0700 48521 48522 glamor: Add helper functions to walk pixmap tiling 48523 48524 This adds a few helper functions to make pixmap fbo access symmetrical 48525 between the single fbo and tiled cases. 48526 48527 Signed-off-by: Keith Packard <keithp@keithp.com> 48528 Reviewed-by: Eric Anholt <eric@anholt.net> 48529 Signed-off-by: Eric Anholt <eric@anholt.net> 48530 48531commit 209d004469391420130262059af43a813b2c07d7 48532Author: Keith Packard <keithp@keithp.com> 48533Date: Thu Mar 13 23:11:40 2014 -0700 48534 48535 glamor: Add bounding box to one-fbo pixmaps 48536 48537 This lets code treat the one-fbo pixmaps more symmetrically with the 48538 tiled pixmaps. 48539 48540 Signed-off-by: Keith Packard <keithp@keithp.com> 48541 Reviewed-by: Eric Anholt <eric@anholt.net> 48542 Signed-off-by: Eric Anholt <eric@anholt.net> 48543 48544commit 82f91433e277e3711678c1b16be85e89b6c7d530 48545Author: Keith Packard <keithp@keithp.com> 48546Date: Tue Mar 18 21:23:24 2014 -0700 48547 48548 glamor: Get testing code using small FBOs working again 48549 48550 Glamor has a mode where pixmaps will be constructed from numerous 48551 small FBOs. This allows testing of the tiled pixmap code without 48552 needing to create huge pixmaps. 48553 48554 However, the render glyph code assumed that it could create a pixmap 48555 large enough for the glyph atlas. Instead of attempting to fix that 48556 (which would be disruptive and not helpful), I've added a new pixmap 48557 creation usage, GLAMOR_CREATE_NO_LARGE which forces allocation of a 48558 single large FBO. 48559 48560 Now that we have pixmaps with varying FBO sizes, I then went around 48561 and fixed the few places using the global FBO max size and replaced 48562 that with the per-pixmap FBO tiling sizes, which were already present 48563 in each large pixmap. 48564 48565 Xephyr has been changed to pass GLAMOR_CREATE_NO_LARGE when it creates 48566 the screen pixmap as it doesn't want to deal with tiling either. 48567 48568 Signed-off-by: Keith Packard <keithp@keithp.com> 48569 Signed-off-by: Eric Anholt <eric@anholt.net> 48570 Reviewed-by: Eric Anholt <eric@anholt.net> 48571 48572commit 11e2f0de71fa341f8b0b1da0a1b9ccbfa6550a50 48573Author: Keith Packard <keithp@keithp.com> 48574Date: Tue Mar 18 21:35:41 2014 -0700 48575 48576 mi: miPutImage with XYPixmap failed at depth 32 on 64-bit machines 48577 48578 The X server still has 'unsigned long' in a few places to hold 32 bit 48579 values. One of those is in miPutImage where it's holding the temporary 48580 planemask for XYPixmap format images. 48581 48582 It computed the highest plane in the source image with 1 << (depth - 48583 1). On 64-bit machines, taking that value and storing it in an 48584 unsigned long promotes it to a signed 64-bit value 48585 (0xffffffff80000000). 48586 48587 Then, it loops over that value, shifting one bit right each time, 48588 waiting for it to go to zero.. That takes 64 iterations, and ends up 48589 with some mystic planemask values *and* walking off the end of the 48590 source image data and out into space. 48591 48592 A simple cast is all that is required to compute the correct initial 48593 plane mask (0x0000000080000000), at which point the loop operates 48594 correctly once again. 48595 48596 Checking the fbPutImage code, I note that this same bug was fixed 48597 in 2006 by Aaron Plattner in commit 48598 f39fd4242902eaa862321d39337f429dd14ebacf 48599 48600 Signed-off-by: Keith Packard <keithp@keithp.com> 48601 Signed-off-by: Eric Anholt <eric@anholt.net> 48602 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 48603 48604commit ae87b536155207e6e28b68963593a7ab09792e08 48605Author: Keith Packard <keithp@keithp.com> 48606Date: Sat Mar 15 23:53:53 2014 -0700 48607 48608 dix: Allow NULL stipple in ChangeGC 48609 48610 miOpqStipDrawable resets the stipple after painting. When that stipple 48611 was NULL, ChangeGC needs to handle that and not crash. 48612 48613 Signed-off-by: Keith Packard <keithp@keithp.com> 48614 Signed-off-by: Eric Anholt <eric@anholt.net> 48615 Reviewed-off-by: Eric Anholt <eric@anholt.net> 48616 48617commit 2e040f41de18f4d46e649561b6efb043195fdfbe 48618Author: Markus Wick <markus@selfnet.de> 48619Date: Tue Mar 18 09:42:55 2014 +0100 48620 48621 glamor: Drop feature dependent optimization on startup. 48622 48623 We don't care that much about startup time to write different code paths... 48624 48625 Signed-off-by: Eric Anholt <eric@anholt.net> 48626 Reviewed-by: Eric Anholt <eric@anholt.net> 48627 48628commit 2e75d974bcfdd879d7417cc02892b78eaea5e888 48629Author: Markus Wick <markus@selfnet.de> 48630Date: Tue Mar 18 09:42:54 2014 +0100 48631 48632 glamor: Remove unneeded unbindings. 48633 48634 They are already cleared in glamor_put_vbo_space. 48635 48636 Signed-off-by: Eric Anholt <eric@anholt.net> 48637 Reviewed-by: Eric Anholt <eric@anholt.net> 48638 48639commit 7c4e1472966d9d3d6432621c3114f491a5378604 48640Author: Markus Wick <markus@selfnet.de> 48641Date: Tue Mar 18 09:42:53 2014 +0100 48642 48643 glamor: Always keep GL_ELEMENT_ARRAY_BUFFER bound to the same IB. 48644 48645 We never used glDrawElements() with a different index buffer. 48646 48647 Signed-off-by: Eric Anholt <eric@anholt.net> 48648 Reviewed-by: Eric Anholt <eric@anholt.net> 48649 48650commit 82168b1e6ede48898be8a64b1c93b7d82ef65702 48651Author: Markus Wick <markus@selfnet.de> 48652Date: Tue Mar 18 09:42:52 2014 +0100 48653 48654 glamor: Select VBO path by ARB_mbr extension. 48655 48656 The mbr path was hard coded enabled for desktop gl and disabled for 48657 gles. But there are both desktop without mbr and GLES with mbr. 48658 48659 v2: Don't forget to update the fini path, too (change by anholt) 48660 48661 Signed-off-by: Eric Anholt <eric@anholt.net> 48662 Reviewed-by: Eric Anholt <eric@anholt.net> 48663 48664commit 53df6e8c3bd7025883b9614a549fbf268c110b74 48665Author: Markus Wick <markus@selfnet.de> 48666Date: Tue Mar 18 09:42:51 2014 +0100 48667 48668 glamor: Update GL requirements to 2.1. 48669 48670 We will never ever run on OpenGL 1.2 as we use shaders everywhere. 48671 2.0 may be enough, but we also often use PBOs and our big shaders 48672 won't fit into the first GLSL limits. 48673 48674 Signed-off-by: Eric Anholt <eric@anholt.net> 48675 Reviewed-by: Eric Anholt <eric@anholt.net> 48676 48677commit 15d36444acc7a2890ade8e320dc16d627a9d2035 48678Author: Markus Wick <markus@selfnet.de> 48679Date: Tue Mar 18 09:42:50 2014 +0100 48680 48681 glamor: Use epoxy_gl_version() instead of rolling our own. 48682 48683 Signed-off-by: Eric Anholt <eric@anholt.net> 48684 Reviewed-by: Eric Anholt <eric@anholt.net> 48685 48686commit 9d87f66e862a521553d48aa73e5ee97431d36ad3 48687Author: Markus Wick <markus@selfnet.de> 48688Date: Tue Mar 18 09:42:49 2014 +0100 48689 48690 glamor: Use epoxy_has_gl_extension() instead of rolling our own. 48691 48692 Signed-off-by: Eric Anholt <eric@anholt.net> 48693 Reviewed-by: Eric Anholt <eric@anholt.net> 48694 48695commit 708fe0625f31ad39ba54f27f04d17ed2aa621cad 48696Author: Markus Wick <markus@selfnet.de> 48697Date: Tue Mar 18 09:42:48 2014 +0100 48698 48699 glamor: Use glsl "fract/mod" instead of "while" in gradient shaders. 48700 48701 This fixes gtkperf. It seemed to hang forever. 48702 48703 Signed-off-by: Eric Anholt <eric@anholt.net> 48704 Reviewed-by: Eric Anholt <eric@anholt.net> 48705 48706commit 84ff8960722fa9baf6ad70756b9c335bf2a5e515 48707Author: Eric Anholt <eric@anholt.net> 48708Date: Tue Mar 18 09:42:47 2014 +0100 48709 48710 kdrive: Remove duplicated definitions of some XV-related structs. 48711 48712 v2: Fix crash because of removed strdup. (by Markus Wick) 48713 48714 Signed-off-by: Eric Anholt <eric@anholt.net> 48715 Reviewed-by: Eric Anholt <eric@anholt.net> (v2) 48716 48717commit 370c8c75ca6aebb5d93d2ab87c9c8ffac21fc57d 48718Author: Gaetan Nadon <memsize@videotron.ca> 48719Date: Wed Mar 19 11:28:57 2014 -0400 48720 48721 xfree86: glamor_egl subdir must be distributed - breaks distcheck 48722 48723 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 48724 Signed-off-by: Eric Anholt <eric@anholt.net> 48725 Reviewed-by: Eric Anholt <eric@anholt.net> 48726 48727commit 870bbf85e63a59f252f3ea9d63e28f083359d990 48728Author: Gaetan Nadon <memsize@videotron.ca> 48729Date: Wed Mar 19 14:17:15 2014 -0400 48730 48731 ephyr: typo where "()" should be "$()" in the Makefile - breaks make dist 48732 48733 make[3]: Entering directory `/home/nadon/xorg/src/xserver/hw/kdrive/ephyr' 48734 make[3]: *** No rule to make target `()', needed by `distdir'. Stop. 48735 make[3]: Leaving directory `/home/nadon/xorg/src/xserver/hw/kdrive/ephyr' 48736 make[2]: *** [distdir] Error 1 48737 48738 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 48739 Signed-off-by: Eric Anholt <eric@anholt.net> 48740 Reviewed-by: Eric Anholt <eric@anholt.net> 48741 48742commit dc48bd653c7e1013e2d69e3f59ae3cbc0c893473 48743Author: Hans de Goede <hdegoede@redhat.com> 48744Date: Mon Mar 24 16:32:01 2014 +0100 48745 48746 systemd-logind: Monitor systemd-logind going away 48747 48748 When we're using server managed-fds through systemd-logind, systemd-logind 48749 *must* keep running while we are using it, as it does things like drmSetMaster 48750 and drmDropMaster for us on vt-switch. 48751 48752 On a systemd-logind restart, we cannot simply re-connect since we will then 48753 get a different fd for the /dev/dri/card# node, and we've tied a lot of 48754 state to the old fd. I've discussed this with the systemd people, and in the 48755 future there may be a restart mechanism were systemd-logind passed fds from 48756 the old logind to the new logind. But for now there answer is simply: 48757 "Don't restart systemd-logind", and there never really is a good reason to 48758 restart it. 48759 48760 So to ensure unpleasentness if people do decide to restart systemd-logind 48761 anyways (or when it crashes), monitor logind going away and make this a fatal 48762 error. This avoids getting a hard-hung machine on the next vt-switch and will 48763 hopefully quickly educate users to not restart systemd-logind while they have 48764 an X session using it active. 48765 48766 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 48767 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48768 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48769 48770commit c0d3a13e7ee42be26d6d1923c9f3ae0355497869 48771Author: Hans de Goede <hdegoede@redhat.com> 48772Date: Mon Mar 24 16:31:59 2014 +0100 48773 48774 Buildsys: Create SUID_WRAPPER_DIR before using it 48775 48776 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 48777 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48778 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48779 48780commit 6289f71494453d34614b7ff1238840681c2472af 48781Author: Gaetan Nadon <memsize@videotron.ca> 48782Date: Tue Mar 25 14:00:13 2014 -0400 48783 48784 suid: adding Xorg.sh.in to EXTRA_DIST is redundant 48785 48786 All files specified in AC_CONFIG_FILES get distributed automatically. 48787 48788 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 48789 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48790 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48791 48792commit b82ef51509bf24e8eafeb615a3e2f4fcd559554e 48793Author: Gaetan Nadon <memsize@videotron.ca> 48794Date: Tue Mar 25 13:49:02 2014 -0400 48795 48796 suid: add generated Xorg.sh to hw/xfree86/.gitignore 48797 48798 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 48799 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48800 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48801 48802commit ca8f26f63694c27fa488a3b58ea6f00b7dd50f81 48803Author: Gaetan Nadon <memsize@videotron.ca> 48804Date: Tue Mar 25 13:48:52 2014 -0400 48805 48806 suid: replace deprecated AC_HELP_STRING with AS_HELP_STRING 48807 48808 Fixes automake warning. 48809 48810 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 48811 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48812 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48813 48814commit bda6fdc71c25b0c3f3747f445103a1995a5713f4 48815Merge: e46820fb8 78167a98a 48816Author: Keith Packard <keithp@keithp.com> 48817Date: Tue Mar 25 16:06:03 2014 -0700 48818 48819 Merge remote-tracking branch 'whot/for-keith' 48820 48821commit e46820fb897600800b5f3297a59039556774e9c5 48822Author: Emil Velikov <emil.l.velikov@gmail.com> 48823Date: Thu Sep 26 20:28:37 2013 +0100 48824 48825 miinitext: introduce LoadExtensionList() to replace over LoadExtension() 48826 48827 Looping around LoadExtension() meant that ExtensionModuleList was reallocated 48828 on every extension. Using LoadExtensionList() we pass an array thus the 48829 function can do the reallocation in one go, and then loop and setup the 48830 ExtensionModuleList. 48831 48832 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 48833 Reviewed-by: Keith Packard <keithp@keithp.com> 48834 48835 v2: Update ephyr [Keith Packard] 48836 v3: Eliminate const warnings in LoadExtensionList [Keith Packard] 48837 48838 Signed-off-by: Keith Packard <keithp@keithp.com> 48839 48840commit 78167a98a8631ee3fad145ddc051ceb8487b9683 48841Author: Adam Jackson <ajax@redhat.com> 48842Date: Fri Nov 8 09:44:24 2013 -0500 48843 48844 xkb: Restore XkbCopyDeviceKeymap 48845 48846 Removed in d35a02a767017f13db4bd4742eef49293d5a30ea, tigervnc 1.2.80 and 48847 xf86-video-nested need it for now. 48848 48849 Signed-off-by: Adam Jackson <ajax@redhat.com> 48850 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48851 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48852 48853commit 215f3d2e0f2a4ef6fc2f2c08fb991e1f00b747c8 48854Author: Kristian Høgsberg <krh@bitplanet.net> 48855Date: Wed Mar 19 20:46:41 2014 -0700 48856 48857 os: Add AddClientOnOpenFD() to create a new client for an file descriptor 48858 48859 When the Xwayland server is socket-activated, we need to connect and 48860 initialize the window manager before the activating client gets to 48861 proceed with connecting. We do this by passing a socket file 48862 descriptor for the window manager connection to the Xwayland server, 48863 which then uses this new function to set it up as an X client. 48864 48865 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 48866 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48867 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 48868 48869commit d415b9d69fdaab4ce3fc05d3d26b2d8413403aa4 48870Author: Emil Velikov <emil.l.velikov@gmail.com> 48871Date: Thu Sep 26 20:16:31 2013 +0100 48872 48873 miinitext: avoid allocating a sentinel ExtensionModule 48874 48875 With all the logic now in place there is no need to allocate a separate 48876 ExtensionModule to be used as a sentinel. 48877 48878 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 48879 Reviewed-by: Keith Packard <keithp@keithp.com> 48880 Reviewed-by: Adam Jackson <ajax@redhat.com> 48881 48882commit b421ac8e026f464fdba2082aaedc00c19fb8b6c7 48883Author: Emil Velikov <emil.l.velikov@gmail.com> 48884Date: Thu Sep 26 20:11:40 2013 +0100 48885 48886 miinitext: constify staticExtensions 48887 48888 The array is readonly and should not be tampered with. 48889 48890 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 48891 Reviewed-by: Keith Packard <keithp@keithp.com> 48892 48893commit 5ff34b5a0b3b333bf0f0334717cc8630cc785510 48894Author: Emil Velikov <emil.l.velikov@gmail.com> 48895Date: Thu Sep 26 20:08:56 2013 +0100 48896 48897 miinitext: drop sentinel detection during LoadExtension 48898 48899 All the sentinels from the extension lists were removed, thus the only 48900 case were this would trigger is when the code is broken badly. 48901 48902 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 48903 Reviewed-by: Keith Packard <keithp@keithp.com> 48904 48905commit 6fea450d41a03d32c10c7da162dfaadd8ead4ece 48906Author: Emil Velikov <emil.l.velikov@gmail.com> 48907Date: Thu Sep 26 20:08:16 2013 +0100 48908 48909 miinitext: move AddStaticExtensions() to LoadExtension() 48910 48911 Separate the function from NewExtensionModule() as the former does 48912 only memory reallocation. No functional change. 48913 48914 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 48915 Reviewed-by: Keith Packard <keithp@keithp.com> 48916 48917commit bf087659f0fb747c471e26c5b287c35877818040 48918Author: Alan Coopersmith <alan.coopersmith@oracle.com> 48919Date: Sat Mar 22 14:30:04 2014 -0700 48920 48921 Add necessary headers for major()/minor() on Solaris to xf86Xinput.c 48922 48923 Without these, after commit fdb4ec86c29d85c, it fails to build on Solaris, 48924 with errors of: 48925 xf86Xinput.c: In function 'xf86stat': 48926 xf86Xinput.c:816:5: error: implicit declaration of function 'major' [-Werror=implicit-function-declaration] 48927 xf86Xinput.c:817:5: error: implicit declaration of function 'minor' [-Werror=implicit-function-declaration] 48928 48929 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 48930 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 48931 Signed-off-by: Keith Packard <keithp@keithp.com> 48932 48933commit d9fdae5f4a648a10653e18dbc602646f3e22e522 48934Merge: 99bee6c05 0e5d2996a 48935Author: Keith Packard <keithp@keithp.com> 48936Date: Sat Mar 22 14:09:00 2014 -0700 48937 48938 Merge remote-tracking branch 'jturney/master' 48939 48940commit 99bee6c0549b36ef15f625bab3997cbd70c7b879 48941Author: Gaetan Nadon <memsize@videotron.ca> 48942Date: Fri Jan 17 09:54:01 2014 -0500 48943 48944 Default font path: remove the check for ${sysconfdir}/X11/fontpath.d 48945 48946 The location ${sysconfdir}/X11/fontpath.d is unknown at configuration time 48947 (only at make time) as evidenced by the configuration output: 48948 48949 checking for ${prefix}/etc/X11/fontpath.d... no 48950 48951 Unlike font-util for the X fonts, there is no mechanism to query where 48952 fontpath.d is. Fedora have chosen /etc/X11 and others have followed, but this 48953 is not a standard. It might also be installed at another location, it may or may 48954 not be under the xserver installation prefix. We just don't know. Debian does 48955 not use this at all. 48956 48957 Distros are using --with-default-path when they support fontpath.d, so they 48958 never relied on the server default as it never worked. 48959 48960 The patch essentially is a noop, which leaves the door open for anyone who 48961 could potentially provide a reliable implementation of choosing fontpath.d 48962 when applicable. 48963 48964 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 48965 Signed-off-by: Keith Packard <keithp@keithp.com> 48966 48967commit 1782316744866a181ed92486bb4dca6ac46d40bd 48968Merge: 4fb31e482 2b77b208d 48969Author: Keith Packard <keithp@keithp.com> 48970Date: Sat Mar 22 13:46:15 2014 -0700 48971 48972 Merge remote-tracking branch 'jwrdecoede/for-keith' 48973 48974commit 0e5d2996ac872aca5995897fc518bdf9116bd246 48975Author: Jon TURNEY <jon.turney@dronecode.org.uk> 48976Date: Wed Sep 11 13:36:18 2013 +0100 48977 48978 hw/xwin: Fix WM_ENDSESSION crash on x86_64 48979 48980 We need to include xwin-config.h into winmsgwindow.c, so that _XSERVER64 is 48981 defined, so that the layout of ScreenRec type is correct, so that it's privates 48982 can be accessed correctly, so that the WM_GIVEUP message can be sent to the 48983 screen window. 48984 48985 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 48986 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 48987 48988commit 896b53ffa72d91d7d604967028291525562b60dd 48989Author: Jon TURNEY <jon.turney@dronecode.org.uk> 48990Date: Sun Jun 16 13:55:51 2013 +0100 48991 48992 hw/xwin: Improve NET_WM_ICON validation 48993 48994 Check that we don't overrun the end of the property data while converting icons 48995 48996 See http://cygwin.com/ml/cygwin-xfree/2013-06/msg00040.html for testcase. 48997 48998 Also, some warning fixes in winXIconToHICON() 48999 49000 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49001 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 49002 49003commit ab61d070024a4776f011e71d762d4c6c0cf58b12 49004Author: Jon TURNEY <jon.turney@dronecode.org.uk> 49005Date: Sun Mar 31 18:29:07 2013 +0100 49006 49007 hw/xwin: Remove obsolete control handling for About dialog 49008 49009 Remove the unused, cygwin-specific handling for ChangeLog, UG and CG buttons in 49010 the About... Dialog. The buttons themselves were removed in commmit 49011 34269a90ea2087f883f5dc8805894fc4998e4b81. 49012 49013 Also remove those window control IDs which are now obsolete. 49014 49015 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49016 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 49017 49018commit 3b4d472b72601922bac264283eb6b611d8d524fc 49019Author: Oliver Schmidt <oschmidt-mailinglists@gmx.de> 49020Date: Wed Sep 7 14:00:56 2011 +0100 49021 49022 hw/xwin: Minimize redraw events after resizing/moving windows in multiwindow mode 49023 49024 In multiwindow mode the modal moving/resizing of windows causes a lot of redraw 49025 events to be sent to the X clients after the user releases the mouse button. 49026 During the moving/resizing client windows are not redrawn as long as the mouse 49027 button is pressed, but all redraw/resizing events are queued and executed step 49028 after step after the moving/resizing ends. 49029 49030 Some clients collect and combine multiple redraw or resizing events, other 49031 clients (e.g. xterm) simply execute each redraw or sizing event. 49032 49033 The enclosed patch minimizes the events for clients to only one event after the 49034 user releases the mouse button to end the moving/resizing. This improves the 49035 user experience and reduces strange screen flickerings, especially on slow 49036 platforms. 49037 49038 The enclosed patch modifies winmultiwindowwndproc.c such that the windows events 49039 WM_ENTERSIZEMOVE and WM_EXITSIZEMOVE that are sent by Windows when the modal 49040 window resizing/moving begins or ends are considered. Only after WM_EXITSIZEMOVE 49041 is the redraw/resizing executed. 49042 49043 Signed-off-by: Oliver Schmidt <oschmidt-mailinglists@gmx.de> 49044 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49045 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 49046 49047commit 03e1cc6f250a3f5cf17b34639adbbc9850c681cd 49048Author: Jon TURNEY <jon.turney@dronecode.org.uk> 49049Date: Mon Nov 7 20:54:10 2011 +0000 49050 49051 hw/xwin: Add '@<WM_CLIENT_MACHINE>' to window name when it's useful to do so 49052 49053 Enhance GetWindowName() so it appends the result of XGetWMClientMachine() when 49054 it is available and useful to do so 49055 49056 Add -hostintitle option to control this behaviour. Add documentation for this 49057 option to man page and -help text. 49058 49059 Also, fix warning in UpdateName() 49060 49061 v2: Provide a HOST_NAME_MAX definition for MinGW 49062 v3: Use '@host' rather than ' (on host)'. Don't add host if it's already in the 49063 title. 49064 49065 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49066 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 49067 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 49068 49069commit 6804acfe4fabc8ff8491bbc7edb6260440d3d4d3 49070Author: Colin Harrison <colin.harrison@virgin.net> 49071Date: Fri Mar 14 15:44:54 2014 +0000 49072 49073 hw/xwin: Remove prototype for non-existent winMWExtWMUpdateIcon() 49074 49075 winMWExtWMUpdateIcon() was removed in commit 527cf13135cfd279733060e0028bbfbe02be5167 49076 49077 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 49078 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49079 49080commit 7fb5d765ba1e28eb775b847cef3d6079eba3bac1 49081Author: Colin Harrison <colin.harrison@virgin.net> 49082Date: Fri Mar 14 15:40:43 2014 +0000 49083 49084 hw/xwin: Use AllocDevicePair() 49085 49086 Use AllocDevicePair() rather than allocating Windows keyboard and pointer 49087 devices individually. 49088 49089 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 49090 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49091 49092commit 061e5eba00a9a0dc27ff76946dc5f77004fc6a56 49093Author: Colin Harrison <colin.harrison@virgin.net> 49094Date: Fri Mar 14 15:36:59 2014 +0000 49095 49096 hw/xwin: Consistently use 'L' for long int constants 49097 49098 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 49099 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49100 49101commit 46df614784cbbffcf6368859285e949010f1ceb8 49102Author: Colin Harrison <colin.harrison@virgin.net> 49103Date: Fri Mar 14 15:36:24 2014 +0000 49104 49105 hw/xwin: Remove an unneeded include 49106 49107 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 49108 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49109 49110commit d48749492dc492fd7430ffdfd29842153618f778 49111Author: Colin Harrison <colin.harrison@virgin.net> 49112Date: Fri Mar 14 15:34:04 2014 +0000 49113 49114 hw/xwin: Use boolean AND rather than bitwise AND in WIN_POLLING_MOUSE_TIMER_ID 49115 49116 For clarity, use boolean AND rather than bitwise AND in 49117 WIN_POLLING_MOUSE_TIMER_ID processing. 49118 49119 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 49120 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49121 49122commit 80ac4a85d56130d09bbc72ed071759a361ded689 49123Author: Colin Harrison <colin.harrison@virgin.net> 49124Date: Fri Mar 14 15:23:16 2014 +0000 49125 49126 hw/xwin: Fix declaration after statement warning in ddxGiveUp() 49127 49128 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 49129 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49130 49131commit cf59f4888e202c6a68176bd3de1f837ca8480370 49132Author: Colin Harrison <colin.harrison@virgin.net> 49133Date: Fri Mar 14 15:09:24 2014 +0000 49134 49135 hw/xwin: Fix typo in comment 49136 49137 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 49138 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49139 49140commit 561bca469d3ade1ff14faddaf70bf12dcdc6aa74 49141Author: Jon TURNEY <jon.turney@dronecode.org.uk> 49142Date: Thu Mar 13 22:38:43 2014 +0000 49143 49144 hw/xwin: Fix const discarded warning in winGenerateAuthorization() 49145 49146 Fix const discarded warning in winGenerateAuthorization() in !XCSECURITY case 49147 49148 In function ‘winGenerateAuthorization’: 49149 hw/xwin/winauth.c:123:38: warning: passing argument 2 of ‘GenerateAuthorization’ discards ‘const’ qualifier from pointer target type [enabled by default] 49150 hw/xwin/winauth.c:99:1: note: expected ‘char *’ but argument is of type ‘const char *’ 49151 49152 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 49153 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 49154 49155commit 44fe1b8ea284df6bbaef67e246016d104665b2fe 49156Author: Kristian Høgsberg <krh@bitplanet.net> 49157Date: Wed Mar 19 14:03:13 2014 -0700 49158 49159 os: Add a mechanism to prevent creating any listen sockets 49160 49161 A socket-activated server will receive its listening sockets from the 49162 parent process and should not create its own sockets. This patch 49163 introduces a NoListen flag that can be set by a DDX to prevent 49164 the server from creating the sockets. When NoListen is enabled, we 49165 also disable the server lock checking, since the parent process is 49166 responsible for checking the lock before picking the display name and 49167 creating the sockets. 49168 49169 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 49170 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49171 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 49172 49173commit 62d584d63190c6c8cef5fd49d87b1cb29358a434 49174Author: Kristian Høgsberg <krh@bitplanet.net> 49175Date: Tue Mar 18 22:05:58 2014 -0700 49176 49177 os: Always compile ListenOnOpenFD() and export it 49178 49179 This function was written to allow the X server to inherit the listen 49180 socket from launchd on OS X. The code is not specific to OS X though 49181 and will be useful for on-demand launched Xwayland servers. 49182 49183 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 49184 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49185 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 49186 49187commit 797781c7bfd7a0df354fb85aa73df4560828b90b 49188Author: Kristian Høgsberg <krh@bitplanet.net> 49189Date: Wed Mar 19 10:27:39 2014 -0700 49190 49191 test: Don't add TEST_LDADD to list test 49192 49193 The list test case is always enabled, even if Xorg is disabled. 49194 TEST_LDADD pulls in Xorg files which breaks linking when Xorg is disabled. 49195 The list test doesn't need any libraries, so just remove list_LDADD. 49196 49197 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 49198 Cc: Peter Hutterer <peter.hutterer@who-t.net> 49199 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49200 49201commit 4fb31e4824d46edc80bb49b4065152899faa5ac6 49202Merge: d18d3f6d1 0e531fbb9 49203Author: Keith Packard <keithp@keithp.com> 49204Date: Wed Mar 19 06:43:14 2014 -0700 49205 49206 Merge remote-tracking branch 'whot/for-keith' 49207 49208commit 0e531fbb97868b9a869044fc5a4f6cb58de6751e 49209Author: Kristian Høgsberg <krh@bitplanet.net> 49210Date: Wed Mar 12 16:31:25 2014 +1000 49211 49212 xkb: add XkbLoadKeymapFromString 49213 49214 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49215 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 49216 49217commit d18d3f6d18aa7827748f7ba14efd405b525883d0 49218Merge: 81a4952d3 6649d0059 49219Author: Keith Packard <keithp@keithp.com> 49220Date: Mon Mar 17 17:29:56 2014 -0700 49221 49222 Merge remote-tracking branch 'anholt/glamor-megaseries' 49223 49224commit 6649d0059e1bc136b84a9e457ae81c07521fadc0 49225Author: Michel Dänzer <michel.daenzer@amd.com> 49226Date: Mon Mar 17 13:16:48 2014 +0900 49227 49228 glamor: Move up glamor_priv->flags assignment in glamor_init() 49229 49230 It wasn't assigned yet when it was tested for GLAMOR_NO_DRI3. 49231 49232 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 49233 Signed-off-by: Eric Anholt <eric@anholt.net> 49234 Reviewed-by: Eric Anholt <eric@anholt.net> 49235 49236commit b5a61239e2fef167c229154d7919ff862503e3f3 49237Author: Eric Anholt <eric@anholt.net> 49238Date: Fri Mar 14 13:10:04 2014 -0700 49239 49240 glx: Make sure we get an FBConfig the root window's visual. 49241 49242 Signed-off-by: Eric Anholt <eric@anholt.net> 49243 Reviewed-by: Keith Packard <keithp@keithp.com> 49244 49245commit 9999a660135fafe5f1fdf9452e5ca74ffb069dc2 49246Author: Markus Wick <markus@selfnet.de> 49247Date: Thu Mar 13 11:17:39 2014 +0100 49248 49249 glamor: don't reset the GLSL program 49250 49251 We don't use fixed function rendering, so there is no need to reset 49252 the program at all. This lets the driver avoid checking for state 49253 changes between draw calls when we rebind the same program. 49254 49255 Improves xephyr x11perf -f8text performance by 6.03062% +/- 1.64928% 49256 (n=20) 49257 49258 Signed-off-by: Eric Anholt <eric@anholt.net> 49259 Reviewed-by: Eric Anholt <eric@anholt.net> 49260 49261commit aa2635b804e7d28322f62d802190bdf0bb799634 49262Author: Markus Wick <markus@selfnet.de> 49263Date: Thu Mar 13 00:24:04 2014 +0100 49264 49265 glamor: remove disabled code 49266 49267 This block was disabled since 2011, so there is likely no need to keep it any more. 49268 49269 Signed-off-by: Eric Anholt <eric@anholt.net> 49270 Reviewed-by: Eric Anholt <eric@anholt.net> 49271 49272commit 6227f07b69b8ab3cd39cf9d257daf2a7202b4ef5 49273Author: Alan Coopersmith <alan.coopersmith@oracle.com> 49274Date: Tue Mar 11 17:28:10 2014 -0700 49275 49276 Remove duplicate assignment of repeat_type_uniform_location 49277 49278 Flagged by cppcheck 1.64: 49279 [glamor/glamor_gradient.c:987] -> [glamor/glamor_gradient.c:991]: 49280 (performance) Variable 'repeat_type_uniform_location' is 49281 reassigned a value before the old one has been used. 49282 49283 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 49284 Reviewed-by: Eric Anholt <eric@anholt.net> 49285 Signed-off-by: Eric Anholt <eric@anholt.net> 49286 49287commit 1a4b24993974200e26345ea6c501d4aa093169db 49288Author: Eric Anholt <eric@anholt.net> 49289Date: Thu Feb 20 15:50:03 2014 -0800 49290 49291 glamor: Apply debug labels to our shaders. 49292 49293 This will help tools like fips, apitrace, or INTEL_DEBUG=shader_time 49294 provide useful information about the shaders in use. 49295 49296 Signed-off-by: Eric Anholt <eric@anholt.net> 49297 Reviewed-by: Markus Wick <markus@selfnet.de> 49298 49299commit d07d2c3c5f95fd644ef563004dfa7f3c53904168 49300Author: Eric Anholt <eric@anholt.net> 49301Date: Sat Jan 11 21:26:01 2014 -0800 49302 49303 glamor: Don't forget to unmap our PBOs before using them to upload. 49304 49305 From the GL_ARB_vertex_buffer_object spec: 49306 49307 After the client has specified the contents of a mapped data store, 49308 and before the data in that store are dereferenced by any GL commands, 49309 the mapping must be relinquished by calling 49310 49311 boolean UnmapBufferARB(enum target); 49312 49313 Our mappings were only getting reaped at PBO destroy time, after the 49314 upload. If the GL implementation wasn't coherent, it would have used 49315 stale data to do the texture upload. 49316 49317 Signed-off-by: Eric Anholt <eric@anholt.net> 49318 Reviewed-by: Markus Wick <markus@selfnet.de> 49319 49320commit 64146e4560fdf3c3a40d023ac8e22259915b05ad 49321Author: Eric Anholt <eric@anholt.net> 49322Date: Sat Jan 11 21:11:18 2014 -0800 49323 49324 glamor: Add an assert about an invariant the upload code relies on. 49325 49326 Signed-off-by: Eric Anholt <eric@anholt.net> 49327 Reviewed-by: Markus Wick <markus@selfnet.de> 49328 49329commit 7eb2bafe22dcc90c2fb94d2d9cae370b683dba7c 49330Author: Eric Anholt <eric@anholt.net> 49331Date: Fri Jan 10 15:54:25 2014 +0800 49332 49333 glamor: Fix ignoring the ALU during SetSpans(). 49334 49335 Signed-off-by: Eric Anholt <eric@anholt.net> 49336 Reviewed-by: Markus Wick <markus@selfnet.de> 49337 49338commit 923c8db7ed1adfe4689f0a36496262faca44b79d 49339Author: Eric Anholt <eric@anholt.net> 49340Date: Thu Jan 9 11:36:30 2014 +0800 49341 49342 glamor: Improve the performance of line fallbacks. 49343 49344 If the lines aren't solid-filled vert/horiz solid-filled rectangles, 49345 we fall back. libreoffice has some diagonal lines, and the 49346 performance of the fallback path was atrocious. Just fall back to 49347 mi's spans instead, so that we don't do an upload/download. 49348 49349 Improves x11perf -seg100 by 863.652% +/- 9.8968% (n=5) 49350 49351 Signed-off-by: Eric Anholt <eric@anholt.net> 49352 49353commit b6953045566dae15437d4220e68986db7a6b4ca7 49354Author: Eric Anholt <eric@anholt.net> 49355Date: Thu Jan 9 10:53:04 2014 +0800 49356 49357 glamor: Improve the performance of PolyGlyphBlt. 49358 49359 Using the same idea as the previous PushPixels code, just make points 49360 for each point in the glyph. This is an advantage over the pushpixels 49361 fallback because we can batch the BO mappings and draw calls across 49362 glyphs. 49363 49364 Improves performance of x11perf -f8text by 773.389% +/- 3.50754% (n=10). 49365 49366 Signed-off-by: Eric Anholt <eric@anholt.net> 49367 Reviewed-by: Markus Wick <markus@selfnet.de> 49368 49369commit b885a639144cdadcb0dae8249a168db158770604 49370Author: Eric Anholt <eric@anholt.net> 49371Date: Sun Jan 5 21:54:48 2014 +0800 49372 49373 glamor: Improve the performance of PushPixels by, well, pushing pixels. 49374 49375 Otherwise, mi will fall back to GetSpans()ing the bitmap, walking the 49376 bitmap, computing spans to be filled, and calling FillSpans(). 49377 49378 Improves x11perf -f8text by 759.373% +/- 3.33096% (n=166) 49379 49380 Signed-off-by: Eric Anholt <eric@anholt.net> 49381 Reviewed-by: Markus Wick <markus@selfnet.de> 49382 49383commit ab68982dcc53d29d5e3c5ea092bd91dab09e54c9 49384Author: Eric Anholt <eric@anholt.net> 49385Date: Sat Jan 11 00:06:21 2014 +0800 49386 49387 glamor: Add missing prepares on the GC during fb fallbacks. 49388 49389 We had regressions in CopyPlane reported by xts5, because we were 49390 (successfully!) dereferencing the null pixmap->devPrivate.ptr for a 49391 tile or stipple without having done a prepare. 49392 49393 Signed-off-by: Eric Anholt <eric@anholt.net> 49394 Reviewed-by: Markus Wick <markus@selfnet.de> 49395 49396commit 93f1824a0b19346f0e1759bedfa9cf10772067af 49397Author: Eric Anholt <eric@anholt.net> 49398Date: Sat Jan 11 00:00:00 2014 +0800 49399 49400 glamor: Rely on nested mappings to handle src==dst and !prepare bugs. 49401 49402 Now that the core deals with that for us, we can avoid all this extra 49403 carefulness. 49404 49405 Signed-off-by: Eric Anholt <eric@anholt.net> 49406 Reviewed-by: Markus Wick <markus@selfnet.de> 49407 49408commit 4c9a20072552c52b3763bd73e7a7e9b9cb8b4993 49409Author: Eric Anholt <eric@anholt.net> 49410Date: Fri Jan 10 23:43:09 2014 +0800 49411 49412 glamor: Allow nested mapping of pixmaps. 49413 49414 The common pattern is to do nested if statements making calls to 49415 prepare_access() and then pop those mappings back off in each set of 49416 braces. Some cases checked for src == dst to avoid leaking mappings, 49417 but others didn't. Others didn't even do the nested mappings, so a 49418 failure in the outer map would result in trying to umap the inner and 49419 failing. 49420 49421 By allowing nested mappings, we can fix both problems by not requiring 49422 the care from the caller, plus we can allow a simpler nesting of all 49423 the prepares in one if statement. 49424 49425 v2: Add a comment about nested unmap behavior, and just reuse the 49426 glamor_access_t enum. 49427 49428 Signed-off-by: Eric Anholt <eric@anholt.net> 49429 Reviewed-by: Markus Wick <markus@selfnet.de> 49430 49431commit d86eacedab443f172baccf544d17e09090f71f3c 49432Author: Eric Anholt <eric@anholt.net> 49433Date: Fri Jan 10 23:39:02 2014 +0800 49434 49435 glamor: Drop unused GLAMOR_ACCESS_WO. 49436 49437 Nothing was using it, and it was going to complicate the 49438 glamor_prepare_access bugfixing I'm going to do next. 49439 49440 Signed-off-by: Eric Anholt <eric@anholt.net> 49441 Reviewed-by: Markus Wick <markus@selfnet.de> 49442 49443commit c36b903f240664f7222cfefc480e60fca936de6a 49444Author: Eric Anholt <eric@anholt.net> 49445Date: Thu Mar 13 10:09:08 2014 -0700 49446 49447 glamor: Drop stale comment. 49448 49449 The old Xephyr codebase was using the GL window system framebuffer for 49450 the screen pixmap, but that meant you couldn't texture from it to do 49451 operations sourcing from the screen, so in the version that landed I 49452 instead had the screen just be a plain texture. 49453 49454 Signed-off-by: Eric Anholt <eric@anholt.net> 49455 Reviewed-by: Keith Packard <keithp@keithp.com> 49456 49457commit f7cd1189d068471373f51d932c558f8151182d58 49458Author: Eric Anholt <eric@anholt.net> 49459Date: Fri Jan 10 22:42:02 2014 +0800 49460 49461 glamor: Replace some goofy enum-likes with a real enum. 49462 49463 This unpacks the bitfield into an int size, but my experience has been 49464 that packing bitfields doesn't matter for performance. 49465 49466 v2: Convert more comparisons against numbers or implicit bool 49467 comparisons to comparisons against the enum names, and fix up some 49468 comments. 49469 49470 Signed-off-by: Eric Anholt <eric@anholt.net> 49471 Reviewed-by: Markus Wick <markus@selfnet.de> 49472 49473commit 575e3e1bf08c418430da86228d2774e8fda4b8c0 49474Author: Eric Anholt <eric@anholt.net> 49475Date: Mon Jan 6 07:22:47 2014 +0800 49476 49477 glamor: Fix up doxygen for glamor_fill.c. 49478 49479 Signed-off-by: Eric Anholt <eric@anholt.net> 49480 Reviewed-by: Markus Wick <markus@selfnet.de> 49481 49482commit 53996e252e51c8053537f485616a078cb7b0f738 49483Author: Eric Anholt <eric@anholt.net> 49484Date: Mon Jan 6 07:05:02 2014 +0800 49485 49486 glamor: Rename more solid fill variables to clean up the code. 49487 49488 Now the error path of allocation is more obvious: We leave things in 49489 the a-few-boxes-at-a-time stack memory state. 49490 49491 Signed-off-by: Eric Anholt <eric@anholt.net> 49492 Reviewed-by: Markus Wick <markus@selfnet.de> 49493 49494commit ec3ab2f67baa00326a4964ed2047e8a137915578 49495Author: Eric Anholt <eric@anholt.net> 49496Date: Mon Jan 6 07:01:17 2014 +0800 49497 49498 glamor: Rename a variable to be more descriptive. 49499 49500 The "valid_" prefix was used above to describe our allocation that 49501 gets reused multiple times, which is totally unrelated. 49502 49503 Signed-off-by: Eric Anholt <eric@anholt.net> 49504 Reviewed-by: Markus Wick <markus@selfnet.de> 49505 49506commit 01e30d2043f5df104947908f14a377dc77896a98 49507Author: Eric Anholt <eric@anholt.net> 49508Date: Mon Jan 6 06:55:15 2014 +0800 49509 49510 glamor: Fix some integer overflow errors. 49511 49512 Imagine a nbox that was (UINT_MAX + small number) / (4 * 2 * 49513 sizeof(float)). We'd malloc a few bytes after the integer overflow, 49514 but glamor_set_normalize_vcoords would write over gigabytes of heap. 49515 49516 Signed-off-by: Eric Anholt <eric@anholt.net> 49517 Reviewed-by: Markus Wick <markus@selfnet.de> 49518 49519commit 20bcda977755559294c3a951f80af2ec861595fc 49520Author: Eric Anholt <eric@anholt.net> 49521Date: Mon Jan 6 06:53:46 2014 +0800 49522 49523 glamor: Drop bogus _X_UNLIKELY. 49524 49525 nbox > 4 is actually quite common for spans handling. 49526 49527 Signed-off-by: Eric Anholt <eric@anholt.net> 49528 Reviewed-by: Markus Wick <markus@selfnet.de> 49529 49530commit ad3dd80720676dcdc11ab5db63ccbed7c9509d7e 49531Author: Eric Anholt <eric@anholt.net> 49532Date: Mon Jan 6 06:52:32 2014 +0800 49533 49534 glamor: Fix stack overflow in glamor_solid vertex handling. 49535 49536 ARRAY_SIZE(vertices) is 32 (floating point values), so we need to 49537 divide by the number of floats in a box like we do in the overflow 49538 case below. 49539 49540 Signed-off-by: Eric Anholt <eric@anholt.net> 49541 Reviewed-by: Markus Wick <markus@selfnet.de> 49542 49543commit d036d22a95b8516a817dd4fa4dd35be363709636 49544Author: Eric Anholt <eric@anholt.net> 49545Date: Sat Jan 4 13:28:52 2014 -0800 49546 49547 glamor: Drop duplicated lines for getting pixmap state in GetImage. 49548 49549 No change in generated code size -- apparently the compiler figured it 49550 out. 49551 49552 Signed-off-by: Eric Anholt <eric@anholt.net> 49553 Reviewed-by: Markus Wick <markus@selfnet.de> 49554 49555commit e23dd41195dfaf1f3e303197845a071d01523618 49556Author: Eric Anholt <eric@anholt.net> 49557Date: Mon Dec 30 18:41:57 2013 -0800 49558 49559 glamor: Return the stride/size for glamor_dri3_name_from_pixmap(), too. 49560 49561 Just like for a caller of glamor_dri3_fd_from_pixmap(), otherwise the 49562 consumer of that named buffer has no idea what GL chose for the 49563 stride. 49564 49565 Signed-off-by: Eric Anholt <eric@anholt.net> 49566 Reviewed-by: Markus Wick <markus@selfnet.de> 49567 49568commit 4e21b7ee49bc8f33c6211411d80aa70b16998adc 49569Author: Eric Anholt <eric@anholt.net> 49570Date: Mon Dec 30 18:33:09 2013 -0800 49571 49572 glamor: Drop the body of the function for enabling DRI3 extensions. 49573 49574 The flag is already being set at glamor_egl_screen_init() time, so no 49575 need for the driver to separately call this. That said, leave the 49576 function around to keep the ABI compatibility. 49577 49578 Signed-off-by: Eric Anholt <eric@anholt.net> 49579 Reviewed-by: Markus Wick <markus@selfnet.de> 49580 49581commit 92e2125c5219c7b270785f063fe2560583b2baff 49582Author: Eric Anholt <eric@anholt.net> 49583Date: Mon Dec 30 18:29:50 2013 -0800 49584 49585 sync: Add a header include necessary to use misyncstr.h 49586 49587 Signed-off-by: Eric Anholt <eric@anholt.net> 49588 Reviewed-by: Keith Packard <keithp@keithp.com> 49589 49590commit 12b2adaaeb091ad48825f439fa0359e5641c86d1 49591Author: Eric Anholt <eric@anholt.net> 49592Date: Sat Dec 28 13:46:17 2013 -0800 49593 49594 glamor: Do glyph private init at screeninit time, and other stuff at CSR. 49595 49596 This hasn't actually been a problem, since the server hasn't allocated 49597 any glyphs before our glyph private initialization during 49598 CreateScreenResources. But it's generally not X Server style to do 49599 things this way. 49600 49601 Now that glamor itself drives both parts of glyphs setup, DDX drivers 49602 no longer need to tell glamor to initialize glyphs. We do retain the 49603 old public symbol so they can keep running with no changes. 49604 49605 Signed-off-by: Eric Anholt <eric@anholt.net> 49606 Reviewed-by: Markus Wick <markus@selfnet.de> 49607 49608commit df1ef90fa0f98c779c3f312130183dbf32f1e447 49609Author: Eric Anholt <eric@anholt.net> 49610Date: Fri Dec 27 18:45:44 2013 -0800 49611 49612 glamor: Move the EGL image to the normal pixmap private. 49613 49614 There's no reason to hide EGL from the rest of glamor, now that we 49615 have epoxy. 49616 49617 Signed-off-by: Eric Anholt <eric@anholt.net> 49618 Reviewed-by: Markus Wick <markus@selfnet.de> 49619 49620commit 732faea542c5ecab353536f93bab5dc6c6068d0a 49621Author: Eric Anholt <eric@anholt.net> 49622Date: Fri Dec 27 18:27:56 2013 -0800 49623 49624 glamor: Use epoxy_has_egl_extension() instead of rolling our own. 49625 49626 Signed-off-by: Eric Anholt <eric@anholt.net> 49627 Reviewed-by: Markus Wick <markus@selfnet.de> 49628 49629commit a895f6a16588024e03218d8d0e77dd8bd8494329 49630Author: Eric Anholt <eric@anholt.net> 49631Date: Sun Jan 5 02:45:05 2014 -0800 49632 49633 glamor: Drop the set of the context to NULL at the end of glamor ops. 49634 49635 The theory here was that it (which I copied from EGL) existed to fix 49636 up context switching with indirect GLX. But indirect GLX won't even 49637 try to set the context again unless its lastContext field is cleared, 49638 so we need to solve this a different way. 49639 49640 Signed-off-by: Eric Anholt <eric@anholt.net> 49641 Reviewed-by: Keith Packard <keithp@keithp.com> 49642 49643commit f31911ff8f3f3707101b7771d5dc994ed1ca5c70 49644Author: Eric Anholt <eric@anholt.net> 49645Date: Thu Feb 27 17:55:50 2014 -0800 49646 49647 xephyr: Don't forget to glViewport() before drawing the screen. 49648 49649 Fixes misrendering with cairogears. I had noticed the failure while 49650 trying to figure out what was going on with traps. Cairogears was 49651 apparently putting its results on the screen through putimage, which 49652 is a texture upload, so the last GL drawing was done to the size of 49653 the cairogears window, not the size of the xephyr screen. 49654 49655 Signed-off-by: Eric Anholt <eric@anholt.net> 49656 Reviewed-by: Keith Packard <keithp@keithp.com> 49657 49658commit 8b6c79e19cfabd9f343884ddd91a738720c94583 49659Author: Kristian Høgsberg <krh@bitplanet.net> 49660Date: Wed Mar 12 17:20:17 2014 +1000 49661 49662 xkb: add KeymapOrDefault 49663 49664 Helper function to return a default map if the keymap compilation failed. 49665 49666 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49667 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 49668 49669commit cb9a1d01468732196fd25c6422991e784639ffce 49670Author: Kristian Høgsberg <krh@bitplanet.net> 49671Date: Wed Mar 12 17:18:22 2014 +1000 49672 49673 xkb: factor out xkb loading to LoadXkm 49674 49675 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49676 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 49677 49678commit 4391cf27f40ec53d41ac7e437cde49848eb32bde 49679Author: Peter Hutterer <peter.hutterer@who-t.net> 49680Date: Wed Mar 12 16:08:15 2014 +1000 49681 49682 xkb: add a callback to xkbcomp 49683 49684 This provides a callback to write to xkbcomp's buffer once everything is 49685 prepared. 49686 49687 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49688 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 49689 49690commit 6ebd838d80074075827dedb75ed4169ea1946be3 49691Author: Peter Hutterer <peter.hutterer@who-t.net> 49692Date: Wed Mar 12 17:17:25 2014 +1000 49693 49694 xkb: constify XkbDDXOpenConfigFile 49695 49696 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49697 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 49698 49699commit 2b77b208daf9402472ba7fb709156a14eb487299 49700Author: Hans de Goede <hdegoede@redhat.com> 49701Date: Wed Mar 12 16:00:23 2014 +0100 49702 49703 xf86Xinput: release server managed fd before removing the device from the list 49704 49705 So that the fd in use test in systemd_logind_release_fd works properly. 49706 49707 Note we cannot change the test inside systemd_logind_release_fd as it must 49708 work for devices which were never added to the xf86InputDevs too. 49709 49710 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 49711 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 49712 49713commit 0e972b6037d3709c13d46adef9d14b702f477abc 49714Author: Hans de Goede <hdegoede@redhat.com> 49715Date: Wed Mar 12 14:57:24 2014 +0100 49716 49717 systemd-logind: Correctly deal with InputDevs sharing a device-node 49718 49719 InputDevices may share a single device-node, this happens ie with Wacom 49720 tablets. 49721 49722 This patch makes take_fd and release_fd properly deal with this, together 49723 with the earlier patch for updating the fd in all matching xf86InputDevs 49724 on pause / resume this completes support for such shared device-nodes. 49725 49726 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 49727 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 49728 49729commit 8d3f63dbe9bfd816beb6475fd0e00df4dbba269f 49730Author: Hans de Goede <hdegoede@redhat.com> 49731Date: Wed Mar 12 12:58:22 2014 +0100 49732 49733 systemd-logind: Add a systemd_logind_set_input_fd_for_all_devs helper 49734 49735 And use it where appropriate. 49736 49737 Setting the fd for all matching InputDevices is necessary when we've 49738 multiple InputDevices sharing a single device-node, such as happens with 49739 Wacom tablets. 49740 49741 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 49742 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 49743 49744commit 4e3d9690e1868d286dcb766b429f9c99313f2401 49745Author: Hans de Goede <hdegoede@redhat.com> 49746Date: Wed Mar 12 13:19:32 2014 +0100 49747 49748 systemd_logind_find_info_ptr_by_devnum: Move to higher inside the file 49749 49750 This is a preparation patch for adding support for server managed fds 49751 for InputDevices where multiple input devices share the same device node (and 49752 thus also their major and minor). 49753 49754 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 49755 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 49756 49757commit ddc3888bbaaddc47706a9cb96459738683d72cb3 49758Author: Hans de Goede <hdegoede@redhat.com> 49759Date: Wed Mar 12 12:45:40 2014 +0100 49760 49761 systemd_logind_find_info_ptr_by_devnum: Add a start argument 49762 49763 Modify systemd_logind_find_info_ptr_by_devnum to take a start argument, so 49764 that it can be used to find all occurences of a devnum in an InputInfo list, 49765 rather then just the first. 49766 49767 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 49768 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 49769 49770commit 40e3c79a591909ab64822cc86fd07513317bf19b 49771Author: Hans de Goede <hdegoede@redhat.com> 49772Date: Tue Mar 11 11:38:09 2014 +0100 49773 49774 Remove config_odev_add_*attribute checks in various places 49775 49776 Note that there are more callers but those were already not doing any 49777 error checking. 49778 49779 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 49780 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 49781 49782commit 92ff79f1a804d63d2f2bb59dfbf3a2869627609a 49783Author: Hans de Goede <hdegoede@redhat.com> 49784Date: Tue Mar 11 11:28:15 2014 +0100 49785 49786 config_odev*: Use XNF alloc functions 49787 49788 config_odev* functions are called in code-paths were we already use 49789 XNF* functions in other places, so which are not oom safe already. 49790 49791 Besides that oom is something which should simply never happen, so aborting 49792 when it does is as good a response as any other. 49793 49794 While switching to XNF functions also fixup an unchecked strdup case. 49795 49796 Note the function prototypes are kept unchanged, as they are part of the 49797 server ABI. 49798 49799 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 49800 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 49801 49802commit e7b84ca46944895971a8f048c7e34869b7de01c0 49803Author: Hans de Goede <hdegoede@redhat.com> 49804Date: Wed Mar 5 16:41:14 2014 +0100 49805 49806 Xorg: Add a suid root wrapper 49807 49808 With the recent systemd-logind changes it is possible to install the Xorg 49809 binary without suid root rights and still have everything working as it 49810 should *if* the user only has cards which are supported by kms. 49811 49812 This commit adds a little suid root wrapper, which is a bit weird, first we 49813 strip the suid-root bit of the Xorg binary, and then we add a wrapper ? 49814 49815 The function of this wrapper is to see if a system still needs root-rights, 49816 if it does not (it supports kms and the kms drivers are properly loaded), 49817 then it will immediately drop all elevated rights before executing the real 49818 Xorg binary. If it finds (some) cards which don't support kms, or no cards 49819 at all, then it will execute the Xorg server with elevated rights so that 49820 ie the nvidia binary driver and the vesa driver can keep working normally. 49821 49822 To make it possible for security concious users who don't need the root 49823 rights to completely remove the wrapper, Xorg is started in a 3 step process 49824 when the wrapper is enabled during build time: 49825 49826 1) A simple shell script which checks if the wrapper is there, if it is 49827 it executes the wrapper, if not it directly executes the real Xorg binary 49828 49829 2) The wrapper gets executed, does its checks, normally drops all elevated 49830 rights and then executes the real Xorg binary 49831 49832 3) The real Xorg binary does its thing 49833 49834 This allows distributions to put the wrapper binary in a separate package, and 49835 will allow users to remove this package. IE the plan with Fedora is to make 49836 "legacy" drivers depend on the wrapper pkg, and since our default install 49837 contains some legacy drivers it will be part of the default install, but 49838 users can later yum remove it (which will also automatically remove the 49839 legacy driver packages as those won't work without it anyways). 49840 49841 The wrapper is loosely modelled after the existing Debian Xwrapper, it 49842 uses the same config-file + config-file format, and also allows restricting 49843 Xserver execution (through the wrapper) to console users only. 49844 49845 There also is a new needs_root_rights config file directive, which can 49846 be used to override the auto-detection the wrapper does. 49847 49848 Hopefully this will allow Debian to replace their own wrapper with this 49849 upstream one. 49850 49851 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 49852 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 49853 49854commit d35a02a767017f13db4bd4742eef49293d5a30ea 49855Author: Rui Matos <tiagomatos@gmail.com> 49856Date: Fri Mar 7 14:32:27 2014 -0800 49857 49858 xkb: Repurpose XkbCopyDeviceKeymap to apply a given keymap to a device 49859 49860 This will also make it useful for cases when we have a new keymap to 49861 apply to a device but don't have a source device. 49862 49863 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 49864 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 49865 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49866 49867commit 361f405d3c866bc4d22ef84975fe771691156336 49868Author: Rui Matos <tiagomatos@gmail.com> 49869Date: Fri Mar 7 14:32:28 2014 -0800 49870 49871 xkb: Factor out a function to copy a keymap's controls onto another 49872 49873 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 49874 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49875 49876commit 81a4952d3dcab9ca3a1ee399c773f5ac352036c4 49877Merge: 96a28e9c9 795066477 49878Author: Keith Packard <keithp@keithp.com> 49879Date: Tue Mar 11 22:04:36 2014 -0700 49880 49881 Merge remote-tracking branch 'whot/for-keith' 49882 49883commit 795066477ee81b5b82e490eac8bed6b656d01f17 49884Author: Peter Hutterer <peter.hutterer@who-t.net> 49885Date: Wed Feb 26 07:54:56 2014 +1000 49886 49887 config: search for PnPID on all parents (#75513) 49888 49889 The PnPID for a device may not be on the immediate parent, so search up the 49890 device tree until we find one. 49891 49892 X.Org Bug 75513 <http://bugs.freedesktop.org/show_bug.cgi?id=75513> 49893 49894 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49895 Tested-by: Benjamin Tissoires <benjamin.tissoires@gmail.com> 49896 49897commit 45fb3a934dc0db51584aba37c2f9d73deff9191d 49898Author: Peter Hutterer <peter.hutterer@who-t.net> 49899Date: Wed Feb 26 16:20:08 2014 +1000 49900 49901 xkb: push locked modifier state down to attached slave devices 49902 49903 Whenever the master changes, push the locked modifier state to the attached 49904 slave devices, then update the indicators. This way, when NumLock or CapsLock 49905 are hit on any device, the LED will light up on all devices. Likewise, a new 49906 keyboard attached to a master device will light up with the correct 49907 indicators. 49908 49909 The indicators are handled per-keyboard, depending on the layout, i.e. if one 49910 keyboard has grp_led:num set, the NumLock LED won't light up on that keyboard. 49911 49912 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49913 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 49914 49915commit 656841798c99bcd79da47c03ec666a48b855541f 49916Author: Peter Hutterer <peter.hutterer@who-t.net> 49917Date: Wed Feb 26 16:16:10 2014 +1000 49918 49919 xkb: factor out state update into a function 49920 49921 No functional changes 49922 49923 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49924 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 49925 49926commit dda2468e579762dbd1fed2c75b5587d98f841e9c 49927Author: Peter Hutterer <peter.hutterer@who-t.net> 49928Date: Wed Feb 26 16:03:19 2014 +1000 49929 49930 xkb: factor out the StateNotify flag check 49931 49932 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49933 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 49934 49935commit fdb4ec86c29d85c1f68418a26e64bcc05b9c14ae 49936Author: Peter Hutterer <peter.hutterer@who-t.net> 49937Date: Tue Mar 11 14:21:27 2014 +1000 49938 49939 xfree86: handle xorg.conf devices with logind 49940 49941 Only devices from the config backend have their attributes set, devices from 49942 the xorg.conf only have Option "Device". That option is also set by the 49943 config backend, so use it. 49944 49945 And since the config backend sets our major/minor but xorg.conf devices don't 49946 have that set, make sure we try to stat it first where needed. 49947 49948 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49949 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 49950 49951commit 4754af9ddbd55c5637b56bd9f37d0d3414c839a8 49952Author: Peter Hutterer <peter.hutterer@who-t.net> 49953Date: Tue Mar 11 14:20:51 2014 +1000 49954 49955 config: only free odev->attrib_name for string types 49956 49957 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 49958 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 49959 49960commit 96a28e9c914d7ae9b269f73a27b99cbd3c465ac8 49961Author: Eric Anholt <eric@anholt.net> 49962Date: Thu Mar 6 18:50:07 2014 -0800 49963 49964 glx: Clear new FBConfig attributes to 0 by default. 49965 49966 The visualSelectGroup wasn't getting set (since our DRI drivers don't 49967 use it), and and since it's the top priority in the sort order, you 49968 got random sorting of your visuals unless malloc really returned you 49969 new memory. This manifested as Xephyr -glamor rendering to a 49970 multisampled window on my system, which as you might guess was 49971 slightly lower performance than expected. 49972 49973 Signed-off-by: Eric Anholt <eric@anholt.net> 49974 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 49975 49976commit 2843a2f9b58a4aae2949111722c5c1bf3878674b 49977Author: Eric Anholt <eric@anholt.net> 49978Date: Tue Feb 25 16:56:54 2014 -0800 49979 49980 glamor: Use buffer_storage 49981 49982 v2: 49983 - Make the default buffer size a #define. (by Markus Wick) 49984 - Fix the return offset for mapping with buffer_storage. (oops!) 49985 v3: 49986 - Avoid GL error at first rendering from unmapping no buffer. 49987 - Rebase on the glBindBuffer(GL_ARRAY_BUFFER, 0) change. 49988 v4: Rebase on Markus's vbo init changes. 49989 v5: Fix missing put_context() in the buffer_storage fallback path. 49990 49991 Signed-off-by: Eric Anholt <eric@anholt.net> 49992 Reviewed-by: Markus Wick <markus at selfnet.de> 49993 49994commit 9b86f7a1ab7904fbbe04283b6461a31470c77972 49995Author: Eric Anholt <eric@anholt.net> 49996Date: Sun Jan 12 07:54:51 2014 -0800 49997 49998 glamor: Fix requested composite VBO size. 49999 50000 The argument to setup_composte_vbo is the number of verts. 50001 50002 v2: Drop the now-unused vert_stride value. 50003 50004 Signed-off-by: Eric Anholt <eric@anholt.net> 50005 Reviewed-by: Markus Wick <markus at selfnet.de> 50006 50007commit d310d566b18e55bce27301e9ac78a7894d5194dd 50008Author: Eric Anholt <eric@anholt.net> 50009Date: Mon Jan 6 09:29:57 2014 +0800 50010 50011 glamor: Extract the streamed vertex data code used by Render. 50012 50013 We should be uploading any vertex data using this kind of upload 50014 style, since it saves a bunch of extra copies of our vertex data. 50015 50016 v2: 50017 - Add a simple comment about what the function does. 50018 - Use get_vbo_space()'s return in trapezoids, instead of dereffing 50019 glamor_priv->vb (by Markus Wick). 50020 - Fix the double-unmapping by moving put_vbo_space() outside of 50021 flush_composite_rects(). 50022 - Remove the rest of the composite_vbo_offset usage, and just always 50023 use get_vbo_space()'s return value. 50024 v3: 50025 - Fix failure to put_vbo_space in traps when no prims were 50026 generated. 50027 - Unbind the VBO from put_vbo_space(). Keeps callers from 50028 forgetting to do so. 50029 v4: 50030 - Split out some changes into the previous 3 commits while trying to 50031 track down a regression. 50032 - Fix regression due to rebase fail where glamor_priv->vbo_offset 50033 wasn't incremented. 50034 v5: 50035 - Fix GLES2 VBO sizing. 50036 - Add a comment about resize behavior. 50037 - Move glamor_vbo.c init code to glamor_vbo.c from 50038 glamor_render.c. (Derived from Markus's changes, but the GLES2 fix 50039 dropped almost all of the code in the functions). 50040 v6: 50041 - Drop the initial BufferData on GLES2 (it happens at put() time). 50042 - Don't forget to set vbo_offset to the size on GLES2. 50043 - Use char * instead of void * in the cast to return the vbo_offset. 50044 - Resize the default FBO to 512kb, to be similar to previous 50045 behavior. +1.66124% +/- 0.284223% (n=679) on aa10text. 50046 50047 Signed-off-by: Eric Anholt <eric@anholt.net> 50048 Reviewed-by: Markus Wick <markus at selfnet.de> 50049 50050commit ffc8423077d8b5af36ab59f59e4e18109586aec9 50051Author: Eric Anholt <eric@anholt.net> 50052Date: Thu Mar 6 17:05:31 2014 -0800 50053 50054 glamor: Track the next vertex offset as we go for AA traps. 50055 50056 Signed-off-by: Eric Anholt <eric@anholt.net> 50057 Reviewed-by: Markus Wick <markus at selfnet.de> 50058 50059commit ca307fd18d4dfded04979903737ec5dd03487625 50060Author: Eric Anholt <eric@anholt.net> 50061Date: Thu Mar 6 17:03:14 2014 -0800 50062 50063 glamor: Track the next vertex offset as we go for non-AA traps. 50064 50065 v2: Use sizeof(float) instead of 4. 50066 50067 Signed-off-by: Eric Anholt <eric@anholt.net> 50068 Reviewed-by: Markus Wick <markus at selfnet.de> 50069 50070commit 03a33048a74d986e27668a4d01861ead0f8be76b 50071Author: Eric Anholt <eric@anholt.net> 50072Date: Thu Mar 6 16:57:46 2014 -0800 50073 50074 glamor: Track the next vertex offset as we go in render accel. 50075 50076 I want to extract the VBO mapping code, and as part of that I need to 50077 get the global vbo_offset munging to stop. 50078 50079 Signed-off-by: Eric Anholt <eric@anholt.net> 50080 Reviewed-by: Markus Wick <markus at selfnet.de> 50081 50082commit 438d8aceec2d4af48917a7dfecacf11cd4f4b5ce 50083Author: Eric Anholt <eric@anholt.net> 50084Date: Thu Mar 6 13:29:08 2014 -0800 50085 50086 glamor: Move glamor_emit_composite_vert() to traps, where it's used. 50087 50088 It's only used in the nonantialiased, triangle-based trapezoids path. 50089 50090 Signed-off-by: Eric Anholt <eric@anholt.net> 50091 Reviewed-by: Markus Wick <markus at selfnet.de> 50092 50093commit bce5ec4f411134114c6ee2715f293a2250ee8800 50094Author: Eric Anholt <eric@anholt.net> 50095Date: Tue Dec 24 12:14:04 2013 -0800 50096 50097 glamor: Don't forget to set GL_INVALIDATE_RANGE_BIT on GL_ARB_mbr. 50098 50099 We don't need any current contents of the buffer, and this allows an 50100 implementation to make a temporary BO for a streamed upload if it 50101 wants to. 50102 50103 Signed-off-by: Eric Anholt <eric@anholt.net> 50104 Reviewed-by: Markus Wick <markus at selfnet.de> 50105 50106commit 25be6b384298a7769fedb6d9b5bdc370398c9c95 50107Author: Eric Anholt <eric@anholt.net> 50108Date: Sun Jan 12 07:18:50 2014 +0800 50109 50110 glamor: Fix a mismatched glamor_get/put_context(). 50111 50112 We don't call GL in this function any more, so we can just drop the 50113 get. 50114 50115 Signed-off-by: Eric Anholt <eric@anholt.net> 50116 Reviewed-by: Markus Wick <markus at selfnet.de> 50117 50118commit bd09292debbb810db06b187d4e61c9416b5e565e 50119Author: Eric Anholt <eric@anholt.net> 50120Date: Fri Dec 27 14:03:26 2013 -0800 50121 50122 glamor: Add a note about the state of GL_ARB_map_buffer_range. 50123 50124 GLES2 Xephyr is failing due to lack of glMapBuffer() with the read 50125 bits set, and I decided to see if we can just switch everything to 50126 glMapBufferRange(). I'm undecided, and it largely depends on whether 50127 we find people are interested in using glamor for the windows X server. 50128 50129 Signed-off-by: Eric Anholt <eric@anholt.net> 50130 Reviewed-by: Markus Wick <markus at selfnet.de> 50131 50132commit ca507d215f54e878055de8da13877bc0225bece1 50133Author: Eric Anholt <eric@anholt.net> 50134Date: Tue Dec 24 16:07:12 2013 -0800 50135 50136 glamor: Fix a spelling mistake in GLAMOR_PIXMAP_FBO_NOT_EXACT_SIZE. 50137 50138 Signed-off-by: Eric Anholt <eric@anholt.net> 50139 Reviewed-by: Markus Wick <markus at selfnet.de> 50140 50141commit 4e86f4236b32b61796d820714dc12457dbc92e28 50142Author: Eric Anholt <eric@anholt.net> 50143Date: Mon Dec 23 18:16:01 2013 -0800 50144 50145 glamor: remove dead global variable. 50146 50147 Signed-off-by: Eric Anholt <eric@anholt.net> 50148 Reviewed-by: Markus Wick <markus at selfnet.de> 50149 50150commit 7f15e5aed79803fa6057fbfda7dfbacacc2e019d 50151Author: Hans de Goede <hdegoede@redhat.com> 50152Date: Mon Mar 3 16:58:54 2014 +0100 50153 50154 Require video drivers to report that they support server managed fds 50155 50156 This makes how we handle video drivers identical to what we do for input 50157 drivers, and this should make live easier for old non kms drivers. 50158 50159 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50160 Reviewed-by: Dave Airlie <airlied@redhat.com> 50161 50162commit 76af81bdf8ade158d7cef718ff6cf31cbf641a9a 50163Author: Hans de Goede <hdegoede@redhat.com> 50164Date: Mon Mar 3 16:07:16 2014 +0100 50165 50166 xf86platformBus: Make doPlatformProbe accept a NULL gdev argument 50167 50168 And use it from xf86platformAddDevice too, instead of directly calling 50169 drvp->platformProbe. 50170 50171 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50172 Reviewed-by: Dave Airlie <airlied@redhat.com> 50173 50174commit 749d25f6f0faed00a4cd906171ea6f96d76591e7 50175Author: Hans de Goede <hdegoede@redhat.com> 50176Date: Fri Mar 7 04:59:47 2014 -0500 50177 50178 systemd-logind: Fix vt-enter not working when using a legacy video driver 50179 50180 If there is only a single non kms video device (tested with the vesa driver), 50181 then we will never get a resume signal for a drm node, so also call vtenter 50182 when we get a resume for an input device. 50183 50184 Notes: 50185 1) vtenter checks if it is ok to do the vtenter, so if there are kms video 50186 devices the calls for input device resumes are a nop 50187 2) This assumes that there will always be at least one server event fd 50188 supporting input device. Since all non legacy input-drivers will be patched 50189 to supported server fds this seems a safe assumption. 50190 50191 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50192 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50193 50194commit c6dea929a4951c3c048554a1ef54cf8143d9577a 50195Author: Hans de Goede <hdegoede@redhat.com> 50196Date: Mon Mar 3 17:00:57 2014 +0100 50197 50198 config_odev_add*_attribute: Don't add the attribute to the list twice 50199 50200 Don't do the xorg_list_append a 2nd time when updating existing attributes, 50201 while at it add a helper function for this to avoid code duplication. 50202 50203 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50204 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50205 50206commit 2f2967173ba0b0a4f7ab013edeaaf11e7c181beb 50207Author: Hans de Goede <hdegoede@redhat.com> 50208Date: Mon Mar 3 09:51:36 2014 +0100 50209 50210 hashtabletest: Fix warning: format ‘%ld’ expects argument of type ... 50211 50212 This fixes the following compiler warning: 50213 50214 hashtabletest.c: In function ‘print_xid’: 50215 hashtabletest.c:15:5: warning: format ‘%ld’ expects argument of type ‘long int’, but argument 2 has type ‘XID’ [-Wformat=] 50216 printf("%ld", *x); 50217 ^ 50218 50219 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50220 Reviewed-by: Eric Anholt <eric@anholt.net> 50221 50222commit b972851ce47b5e3b681a212a43dd71aab94e7a1b 50223Author: Dave Airlie <airlied@gmail.com> 50224Date: Mon Mar 10 08:58:37 2014 +1000 50225 50226 xserver: fix build since system-logind.h 50227 50228 on tinderbox and irc 50229 50230 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 50231 Reviewed-by: Matthieu Herrb <matthieu@herrb.eu> 50232 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 50233 Signed-off-by: Dave Airlie <airlied@redhat.com> 50234 Signed-off-by: Keith Packard <keithp@keithp.com> 50235 50236commit 1c61d38528a573caadee2468ee59ea558c822e09 50237Merge: 5350ae1d3 bf3543739 50238Author: Keith Packard <keithp@keithp.com> 50239Date: Fri Mar 7 22:07:19 2014 -0800 50240 50241 Merge remote-tracking branch 'jwrdecoede/for-keith' 50242 50243commit 5350ae1d38f3c69a26421e0866ede0d2ccc3aea4 50244Author: Jon TURNEY <jon.turney@dronecode.org.uk> 50245Date: Fri Mar 7 15:59:10 2014 +0000 50246 50247 Fix xephyr build in --disable-glamor case 50248 50249 It broke after commit 9fe052d90cca90fdf750d3a45b151be2ac7f0ebd 50250 "xephyr: Build support for rendering with glamor using a -glamor 50251 option." 50252 50253 See http://tinderbox.x.org/builds/2014-03-07-0004/logs/xserver/#build 50254 50255 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 50256 Signed-off-by: Eric Anholt <eric@anholt.net> 50257 Reviewed-by: Eric Anholt <eric@anholt.net> 50258 50259commit 78e508c9379b3976ac8e76b7aaa90ba86ad4e443 50260Author: Andrew Eikum <aeikum@codeweavers.com> 50261Date: Mon Dec 30 09:15:41 2013 -0600 50262 50263 dix: Clear any existing selections before initializing privates 50264 50265 If there is a selection left over from a previous execution of the 50266 main loop, and that selection has privates allocated for it, the X 50267 server will crash. This is because dixResetPrivates() resets the 50268 privates refcounts to zero without accounting for the reference held 50269 by the selection object. When the selection is then deleted in 50270 InitSelections() after the call to dixResetPrivates(), the refcount 50271 for its privates type goes negative and bad things happen. 50272 50273 To fix this, we should delete any existing selections before calling 50274 dixResetPrivates(). This will properly release the selection's 50275 privates and avoid the crash. 50276 50277 A more thorough description of the problem and a test case to 50278 reproduce the crash is available at a previous mail: 50279 "Negative Selection devPrivates refcount?" 50280 By Andrew Eikum to xorg-devel on 10 Dec 2013 50281 http://lists.freedesktop.org/archives/xorg-devel/2013-December/039492.html 50282 50283 Signed-off-by: Andrew Eikum <aeikum@codeweavers.com> 50284 Reviewed-by: Adam Jackson <ajax@redhat.com> 50285 Signed-off-by: Keith Packard <keithp@keithp.com> 50286 50287commit da08316605b26830b4d8f8fb2d9e69471cdc80ab 50288Author: Eric Anholt <eric@anholt.net> 50289Date: Tue Dec 31 00:07:42 2013 -0800 50290 50291 glamor: Add support for DRI3. 50292 50293 The render-nodes case is untested. 50294 50295 v2: Add a flag for wayland to suppress the native DRI3 support. 50296 Wayland isn't running as a master itself, so it can't do the auth 50297 on its own and has to ask the compositor to do it for us. Dropped 50298 XXX about randr provider -- the conclusion from discussion with 50299 keithp was that if the driver's dri3_open for a provider on a 50300 different screen, that's a core dri3 bug. 50301 v3: Don't put quite so much under GLAMOR_NO_DRI3, and add a comment 50302 explaining what this is about. 50303 50304 Signed-off-by: Eric Anholt <eric@anholt.net> 50305 Reviewed-by: Adam Jackson <ajax@redhat.com> 50306 50307commit fb4a1e6ef6f80a7670e92cab2bc490d4afd80a9b 50308Author: Eric Anholt <eric@anholt.net> 50309Date: Mon Dec 30 18:49:46 2013 -0800 50310 50311 glamor: Rename the DRI-related pixmap functions. 50312 50313 There was confusion over whether they should have egl in the name, and 50314 they had DRI3 in the name even though they're useful to have without 50315 DRI3. 50316 50317 v2: Just rename glamor_name_from_pixmap for now -- I'd accidentally 50318 conflict-resolved in adding new parameters from a later commit. 50319 50320 Signed-off-by: Eric Anholt <eric@anholt.net> 50321 Reviewed-by: Adam Jackson <ajax@redhat.com> 50322 50323commit 2d20f75b60276508424997d9fa8c5b97d8f1d92b 50324Author: Eric Anholt <eric@anholt.net> 50325Date: Fri Dec 27 22:19:07 2013 -0800 50326 50327 xorg: Connect up the glamor XV code, xorg DDX-only for now. 50328 50329 Porting this code to be non-xorg-dependent is going to take 50330 significant hacking, so just dump it in the glamoregl module for the 50331 moment, so I can hack on it while regression testing. 50332 50333 v2: Fix compiler warnings by adding #include dix-config.h at the top, 50334 don't try to auto-init (I'll try to fix the xv ABI later). 50335 v3: Fix last minute breakage of having reintroduced xf86ScrnToScreen 50336 (one of the compat macros). Just use the drawable's pScreen instead. 50337 50338 Signed-off-by: Eric Anholt <eric@anholt.net> 50339 Reviewed-by: Adam Jackson <ajax@redhat.com> 50340 50341commit 8d1cca30638f8a12c09efee27e9dedd90322b40e 50342Author: Eric Anholt <eric@anholt.net> 50343Date: Fri Dec 27 16:44:15 2013 -0800 50344 50345 xorg: Build a glamor_egl module. 50346 50347 This is not exposing the API we want long term, but it should get 50348 existing DDX drivers up and running while we massage the API into 50349 shape. 50350 50351 v2: Use LIBADD instead of LDFLAGS to fix deps on libglamor.la, and use 50352 version 0.5.1 (the point it was forked from the external repo). 50353 50354 Signed-off-by: Eric Anholt <eric@anholt.net> 50355 Reviewed-by: Adam Jackson <ajax@redhat.com> 50356 50357commit fa2e78788327c41bfd45b3de3c71bf9c26dcc85e 50358Author: Eric Anholt <eric@anholt.net> 50359Date: Fri Dec 20 10:41:43 2013 -0800 50360 50361 xephyr: Pass incoming XCB events to the Xlib event filter. 50362 50363 This is the same thing that Qt ended up doing to get DRI2's event 50364 mangling to happen despite using an XCB event loop. 50365 50366 Signed-off-by: Eric Anholt <eric@anholt.net> 50367 Reviewed-by: Keith Packard <keithp@keithp.com> 50368 Reviewed-by: Adam Jackson <ajax@redhat.com> 50369 50370commit 9fe052d90cca90fdf750d3a45b151be2ac7f0ebd 50371Author: Eric Anholt <eric@anholt.net> 50372Date: Wed Dec 18 13:24:14 2013 -0800 50373 50374 xephyr: Build support for rendering with glamor using a -glamor option. 50375 50376 v2: Avoid making the Ximage for the screen that we'll never use, and 50377 drive the screen pixmap creation for glamor ourselves. 50378 50379 Signed-off-by: Eric Anholt <eric@anholt.net> 50380 Reviewed-by: Keith Packard <keithp@keithp.com> (v1) 50381 Reviewed-by: Adam Jackson <ajax@redhat.com> 50382 50383commit b634e909895f6001e7d9543e1350b20c82c8c01c 50384Author: Jon TURNEY <jon.turney@dronecode.org.uk> 50385Date: Thu Jun 27 23:56:23 2013 +0100 50386 50387 hw/xwin: More closely follow ICCCM for setting input focus 50388 50389 In multiwindow mode, more closely follow ICCCM section 4.1.7 when setting X 50390 input focus to a window when the native Windows window acquires input focus: 50391 50392 - If InputHint is FALSE, don't use XSetInputFocus() 50393 - If the window supports the WM_TAKE_FOCUS protocol, send a WM_TAKE_FOCUS message 50394 50395 This helps JDK 1.7 clients acquire the focus correctly. 50396 50397 Also, factor out checking client support for a given WM_PROTOCOLS protocol as a 50398 separate function. 50399 50400 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 50401 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 50402 50403commit 0fc84a2bb6970f6b05a19cd8b32a7f3f7fd148b3 50404Author: Colin Harrison <colin.harrison@virgin.net> 50405Date: Fri Feb 28 15:02:14 2014 +0000 50406 50407 hw/xwin: Remove unnecessary casts from malloc/realloc/calloc calls 50408 50409 Remove unnecessary casts from malloc/realloc/calloc calls. This is the style 50410 used for the majority of X server code. 50411 50412 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 50413 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 50414 50415commit 94f709cfce62e716f8d3adea388625850de71e78 50416Author: Colin Harrison <colin.harrison@virgin.net> 50417Date: Fri Feb 28 14:23:00 2014 +0000 50418 50419 hw/xwin: Add missing FORCEEXIT token to XWin configuration file lexer 50420 50421 Somehow this was left out of commmit f3fad371cce0f3836514ad5b29e59fa1ca0627a7 50422 50423 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 50424 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 50425 50426commit 6432d44020443bbda90bd46ffcb572b51be803a1 50427Author: Colin Harrison <colin.harrison@virgin.net> 50428Date: Fri Feb 28 14:21:46 2014 +0000 50429 50430 hw/xwin: Silence bell when volume is zero 50431 50432 Allow the bell to be turned off with X server option '-f 0', or by 'xset b off'. 50433 50434 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 50435 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 50436 50437commit 7e37c4f727609d2d992ca46ffce56311c8d8225c 50438Author: Colin Harrison <colin.harrison@virgin.net> 50439Date: Fri Feb 28 14:21:06 2014 +0000 50440 50441 hw/xwin: Fix typo in comment 50442 50443 'i' before 'e' except after 'c' 50444 50445 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 50446 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 50447 50448commit d75195b62677f5b0f17bbe089b3aea5b295d5f2c 50449Author: Colin Harrison <colin.harrison@virgin.net> 50450Date: Fri Feb 28 14:20:48 2014 +0000 50451 50452 hw/xwin: Align parameter names in prototypes with definition 50453 50454 A follow up to commits 2d9123fd, 451c5d91 and efe96a17, which changed the 50455 parameter name in the definition from index to i, to fix shadowing index() but 50456 didn't adjust the prototype declaration. 50457 50458 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 50459 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 50460 50461commit e53568e2c5004a434a16e3971fb2cd0823e6487b 50462Author: Jon TURNEY <jon.turney@dronecode.org.uk> 50463Date: Wed Jan 1 16:43:38 2014 +0000 50464 50465 hw/xwin: Just generate the WGL wrappers we need 50466 50467 Just generate the WGL wrappers we need, rather than for everything in wgl.xml 50468 50469 This avoids generating a lot of unused wrappers, and also avoids compilation 50470 requiring a wglext.h at least as new as wgl.xml 50471 50472 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 50473 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 50474 50475commit dec5e9899bfee2a83f8a64f975790ecd2390256d 50476Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 50477Date: Wed Apr 10 18:04:17 2013 +0100 50478 50479 hw/xwin: Fix implicit-function-declaration warning in XwinExtensionInit() when compiled with XWIN_GLX_WINDOWS defined 50480 50481 InitOutput.c: In function ‘XwinExtensionInit’: 50482 InitOutput.c:170:9: error: implicit declaration of function ‘glxWinPushNativeProvider’ [-Werror=implicit-function-declaration] 50483 glxWinPushNativeProvider(); 50484 ^ 50485 InitOutput.c:170:9: warning: nested extern declaration of ‘glxWinPushNativeProvider’ [-Wnested-externs] 50486 50487 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 50488 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 50489 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 50490 50491commit bf3543739db18c6cd52908f7c693cb64f43e3f23 50492Author: Hans de Goede <hdegoede@redhat.com> 50493Date: Sat Feb 1 13:21:15 2014 +0100 50494 50495 xf86OpenSerial: Add support server managed fds 50496 50497 Try to get a server managed fd from the Options before trying to open the 50498 device node ourselves. 50499 50500 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50501 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50502 50503commit d9a4059d400bf288aadd17c04885cb9b0fb93989 50504Author: Hans de Goede <hdegoede@redhat.com> 50505Date: Tue Jan 14 15:47:48 2014 +0100 50506 50507 config-udev: Don't try to add drm devices twice 50508 50509 Before this commit we were trying to add most drm devices twice, once 50510 from xf86platformProbe() and once from config_udev_init(). 50511 50512 This results in somewhat confusing messages in Xorg.log, ie: 50513 50514 (II) xfree86: Adding drm device (/dev/dri/card0) 50515 Later followed by: 50516 (II) config/udev: Adding drm device (/dev/dri/card0) 50517 50518 By filtering out duplicate drm devices we avoid these confusing messages. 50519 50520 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50521 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 50522 50523commit 687afa3f6475060e6c460aa9496ba1a6a7982102 50524Author: Hans de Goede <hdegoede@redhat.com> 50525Date: Fri Jan 31 15:21:38 2014 +0100 50526 50527 systemd-logind: Add delayed input device probing 50528 50529 With systemd-logind we cannot probe input devices while switched away, so 50530 if we're switched away, put the pInfo on a list, and probe everything on 50531 that list on VT-Enter. 50532 50533 This is using an array grown by re-alloc, rather than a xorg_list since 50534 creating a new data-type to store a pInfo + list-entry just for this seems 50535 overkill. 50536 50537 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50538 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50539 50540commit cac39219898f5e5a59ff8d8d6524f5fe0d111469 50541Author: Hans de Goede <hdegoede@redhat.com> 50542Date: Tue Jan 14 15:37:33 2014 +0100 50543 50544 systemd-logind: Hookup systemd-logind integration 50545 50546 This commits makes the changes necessary outside of the systemd-logind core 50547 to make the server use systemd-logind managed fds for input devices and drm 50548 nodes. 50549 50550 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50551 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50552 50553commit 82863656ec449644cd34a86388ba40f36cea11e9 50554Author: Hans de Goede <hdegoede@redhat.com> 50555Date: Tue Jan 7 12:15:12 2014 +0100 50556 50557 systemd-logind: Add systemd-logind "core" 50558 50559 This commits add the bulk of the systemd-logind integration code, but does 50560 not hook it up yet other then calling its init and fini functions, which 50561 don't do that much. 50562 50563 Note the configure bits check for udev since systemd-logind use will only be 50564 supported in combination with udev. Besides that it only checks for dbus 50565 since all communication with systemd-logind is happening over dbus, so 50566 no further libs are needed. 50567 50568 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50569 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50570 50571commit 5fb641a29bfb4a33da964e1e9af523f3472015c6 50572Author: Hans de Goede <hdegoede@redhat.com> 50573Date: Mon Jan 13 12:03:46 2014 +0100 50574 50575 hotplug: Extend OdevAttributes for server-managed fd support 50576 50577 With systemd-logind support, the xserver, rather than the drivers will be 50578 responsible for opening/closing the fd for drm nodes. 50579 50580 This commit adds a fd member to OdevAttributes to store the fd to pass it 50581 along to the driver. 50582 50583 systemd-logind tracks devices by their chardev major + minor numbers, so 50584 also add OdevAttributes to store the major and minor. 50585 50586 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50587 Reviewed-by: Dave Airlie <airlied@redhat.com> 50588 50589commit a8d802cb019da0549da57576ba0cc2cef1c8a98c 50590Author: Hans de Goede <hdegoede@redhat.com> 50591Date: Sun Feb 2 13:54:33 2014 +0100 50592 50593 OdevAttribute: Remove unowned flag from OdevAttributes head 50594 50595 The OdevAttributes struct should just be a head of the attributes list, and 50596 not contain various unrelated flags. Instead add a flags field to 50597 struct xf86_platform_device and use that. 50598 50599 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50600 Reviewed-by: Dave Airlie <airlied@redhat.com> 50601 50602commit bc9d17fb566d57eabe4a3372773654fcf6ae5b44 50603Author: Hans de Goede <hdegoede@redhat.com> 50604Date: Sun Feb 2 15:28:31 2014 +0100 50605 50606 OdevAttribute: Add support for integer attributes 50607 50608 Add a couple of new functions for dealing with storing integer values into 50609 OdevAttributes. 50610 50611 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50612 Reviewed-by: Dave Airlie <airlied@redhat.com> 50613 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50614 50615commit 350559dcdcd8b0de57801302e385e25debcb91f1 50616Author: Hans de Goede <hdegoede@redhat.com> 50617Date: Mon Feb 3 15:47:35 2014 +0100 50618 50619 OdevAttribute: config_odev_add_attribute: replace existing values 50620 50621 If an attribute already exists replace its value rather than adding a 50622 duplicate attribute with the new value to the list. 50623 50624 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50625 Reviewed-by: Dave Airlie <airlied@redhat.com> 50626 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50627 50628commit 10c64e80569bbd35e97916a067f553cab27454b5 50629Author: Hans de Goede <hdegoede@redhat.com> 50630Date: Sun Feb 2 11:25:34 2014 +0100 50631 50632 OdevAttribute: Add config_odev_get_attribute helper 50633 50634 Add a config_odev_get_attribute helper, and replace the diy looping over all 50635 the attributes done in various places with calls to this helper. 50636 50637 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50638 Reviewed-by: Dave Airlie <airlied@redhat.com> 50639 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50640 50641commit 3346166a6596115421702c1f7e2732215674ab8e 50642Author: Hans de Goede <hdegoede@redhat.com> 50643Date: Mon Jan 13 11:57:38 2014 +0100 50644 50645 xf86Xinput: Modify API for server-managed fd support 50646 50647 With systemd-logind support, the xserver, rather than the drivers will be 50648 responsible for opening/closing the fd for input devices. 50649 50650 This commit adds a new capabilities field to the InputDriverRec and a 50651 XI86_DRV_CAP_SERVER_FD flag for drivers to indicate that they support server 50652 managed fds. 50653 50654 This commit adds a new XI86_SERVER_FD flag to indicate to drivers when the 50655 server is managing the fd and they should not open/close it. Note that even 50656 if drivers declare they support server managed fds there is no guarantee they 50657 will actually get them. 50658 50659 Since this changes the input driver ABI, this commit bumps it. 50660 50661 systemd-logind tracks devices by their chardev major + minor numbers, since 50662 we are breaking ABI anyways also add major and minor fields for easy storage / 50663 retrieval of these. 50664 50665 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50666 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50667 50668commit 6b79f28f5cb259f2168ecc3034fe2161e969bf7d 50669Author: Hans de Goede <hdegoede@redhat.com> 50670Date: Wed Jan 22 11:32:56 2014 +0100 50671 50672 linux: xf86OpenConsole: Don't detach from controlling tty when it is our vt 50673 50674 Detaching from our controlling tty makes little sense when it is the same 50675 as the vt we're asked to run on. So automatically assume -keeptty in this case. 50676 50677 This is useful to do because when not running as root the server can only make 50678 various VT related ioctls when it does not detach from the tty. 50679 50680 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50681 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50682 50683commit 701b7ab34984b0f9f0c357b0ff8b53c5132c6324 50684Author: Hans de Goede <hdegoede@redhat.com> 50685Date: Thu Jan 16 10:37:28 2014 +0100 50686 50687 linux: xf86OpenConsole remove root-rights check for keeptty option 50688 50689 There is no reason why keeptty cannot be used without root-rights. 50690 50691 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 50692 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50693 50694commit 98f59a33dfc2d47a57c68420f226dc0c17992164 50695Author: Dave Airlie <airlied@redhat.com> 50696Date: Fri Feb 28 12:04:30 2014 +1000 50697 50698 modesetting: fix cursor rendering with hotspots. 50699 50700 older kernels report EINVAL not ENOSYS, doh. 50701 50702 Signed-off-by: Dave Airlie <airlied@redhat.com> 50703 50704commit 2029c028fcffbe1aa175ab75636ae954a211c82a 50705Author: Adam Jackson <ajax@redhat.com> 50706Date: Tue Feb 25 09:06:55 2014 -0500 50707 50708 modesetting: Don't (brokenly) double-track software cursor 50709 50710 Signed-off-by: Adam Jackson <ajax@redhat.com> 50711 50712commit caf1dec2a76fbbd21259fe4cc809e24a55ff79b4 50713Author: David Sodman <dsodman@chromium.org> 50714Date: Mon Feb 10 09:07:33 2014 -0800 50715 50716 V2: Add check for link from output to crtc before optimizing out a CrtcSet call 50717 50718 The function RRCrtcSet call checks to see if the config being set is 50719 already configured, but, doesn't check that the selected outputs are 50720 connected to the crtc before skipping. This means that the following 50721 sequence will omit the final CrtcSet call to the driver: 50722 50723 CRTC c1 connect to output o 50724 CRTC c2 connect to output o 50725 CRTC c1 connect to output o 50726 50727 This change adds the check to ensure that each of the calls are made to 50728 the driver. 50729 50730 Signed-off-by: David Sodman <dsodman@chromium.org> 50731 Reviewed-by: Stéphane Marchesin <marcheu@chromium.org> 50732 Signed-off-by: Keith Packard <keithp@keithp.com> 50733 50734commit 249565a07d1d243e27440e2a5ecf4c95490903c6 50735Author: Alan Coopersmith <alan.coopersmith@oracle.com> 50736Date: Sun Feb 9 23:13:51 2014 -0800 50737 50738 Fix test/os to work on Solaris 50739 50740 Due to bad decisions made decades ago at AT&T, on SVR4 OS'es the signal() 50741 function resets the signal handler before calling the signal handler 50742 (equivalent to sigaction flag SA_RESETHAND). This is why the X server 50743 has a OsSignal() helper function in os/utils.c that uses the portable 50744 POSIX sigaction function to provide BSD/Linux semantics in a signal() 50745 style API, so we switch to use that in this test case, allowing it to 50746 pass on Solaris. 50747 50748 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 50749 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50750 Signed-off-by: Keith Packard <keithp@keithp.com> 50751 50752commit 0031f6b073c7a4cca553c1d3e185a63436a75cbe 50753Author: Alan Coopersmith <alan.coopersmith@oracle.com> 50754Date: Sun Feb 9 23:13:50 2014 -0800 50755 50756 Fix test/signal-logging to work on Solaris 50757 50758 For some reason, Solaris libc sprintf() doesn't add "0x" to the %p output 50759 as glibc does, causing the test to fail for not matching the exact output. 50760 Since the 0x is desirable, we add it ourselves to the test string. 50761 50762 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 50763 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50764 Signed-off-by: Keith Packard <keithp@keithp.com> 50765 50766commit e67f2d7e0f9189beb2907fa06cff5ecc7f35f922 50767Author: Matthieu Herrb <matthieu.herrb@laas.fr> 50768Date: Sun Feb 9 11:20:59 2014 +0100 50769 50770 gcc 4.2.1 doesn't support #pragma GCC diagnostic ignored 50771 50772 Signed-off-by: Matthieu Herrb <matthieu@herrb.eu> 50773 Signed-off-by: Keith Packard <keithp@keithp.com> 50774 50775commit 1940508a4af33d44a7a8ef24bbdcd1e31e228dab 50776Merge: c85ea92af ddeca9274 50777Author: Keith Packard <keithp@keithp.com> 50778Date: Mon Feb 24 16:27:52 2014 -0800 50779 50780 Merge remote-tracking branch 'herrb/master' 50781 50782commit c85ea92af145c98a4a3fc6cf7b9eac796a48cc2e 50783Author: Keith Packard <keithp@keithp.com> 50784Date: Mon Feb 24 09:51:39 2014 -0800 50785 50786 Bump to 1.15.99.901 50787 50788 Signed-off-by: Keith Packard <keithp@keithp.com> 50789 50790commit 0840a303c47b9b10ba8e24e62956da0f1675e963 50791Merge: 0f10cfd4b 3d71df48e 50792Author: Keith Packard <keithp@keithp.com> 50793Date: Mon Feb 24 09:40:23 2014 -0800 50794 50795 Merge remote-tracking branch 'anholt/glamor-pull-request' 50796 50797commit fac2bcf1da09fba8bc8b22631e31f3c5b2737efd 50798Author: Dave Airlie <airlied@redhat.com> 50799Date: Mon Jan 20 11:06:42 2014 +1000 50800 50801 modesetting: try and use hotspot cursor support 50802 50803 Signed-off-by: Dave Airlie <airlied@redhat.com> 50804 50805commit 4aab7aa7c1a15af7211996fa30e17a645f120ff9 50806Author: Dave Airlie <airlied@redhat.com> 50807Date: Fri Feb 21 12:48:42 2014 +1000 50808 50809 modesetting: move closing fd to after we check outputs 50810 50811 On something like cirrus, start X, then attempt to start a second 50812 X while the first is running, if fbdev is installed it'll fail 50813 hard. 50814 50815 Signed-off-by: Dave Airlie <airlied@redhat.com> 50816 50817commit 0f10cfd4b903d4db293ec47c8a9a0d8b33965803 50818Author: Peter Hutterer <peter.hutterer@who-t.net> 50819Date: Thu Feb 20 13:18:05 2014 +1000 50820 50821 Xi: check for invalid modifiers for XI2 passive grabs 50822 50823 The other values are checked correctly, but if a modifier was outside the 50824 allowed range, it would go unnoticed and cause a out-of-bounds read error for 50825 any mask equal or larger than 256. The DetailRec where we store the grab masks 50826 is only sized to 8 * sizeof(Mask). 50827 50828 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 50829 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 50830 50831commit ec6a44612565e0716bd5b2e2b80a8d381691feb6 50832Author: Gaetan Nadon <memsize@videotron.ca> 50833Date: Wed Feb 19 13:48:06 2014 -0500 50834 50835 test: remove source file from hashtabletest LDADD 50836 50837 LDADD is for libraries and not for source code. 50838 50839 Introduced in commit: ccb3e78124fb05defd0c9b438746b79d84dfc3ae 50840 50841 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 50842 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 50843 50844commit 31b0be69e5eea3d1c82d6610bd37bbdb4dca779c 50845Author: Thierry Reding <treding@nvidia.com> 50846Date: Wed Feb 19 17:16:48 2014 +0100 50847 50848 test/input: Fix alignment assertion for doubles 50849 50850 The code previously tried to compute the offset of a field in the 50851 valuator by subtracting the address of the valuator from the _value_ of 50852 the field (rather than the field's address). The correct way to do it 50853 would have been (note the &'s): 50854 50855 assert(((void *) &v->axisVal - (void *) v) % sizeof(double) == 0); 50856 assert(((void *) &v->axes - (void *) v) % sizeof(double) == 0); 50857 50858 That's essentially what the offsetof() macro does. Using offsetof() has 50859 the added benefit of not using void pointer arithmetic and therefore 50860 silencing a warning on some compilers. 50861 50862 Signed-off-by: Thierry Reding <treding@nvidia.com> 50863 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 50864 50865commit 9368bdec1d37127e97411b684f0b1fce5ee97907 50866Author: Laércio de Sousa <lbsousajr@gmail.com> 50867Date: Mon Feb 17 08:43:49 2014 +1000 50868 50869 Get rid of config/non-seat0.conf.multi-seat 50870 50871 This file is deprecated by commits c73c36b53 and 46cf2a6093. 50872 50873 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 50874 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 50875 50876commit 3d71df48e70884deccb293cf1fcffbdba8efd94b 50877Author: Eric Anholt <eric@anholt.net> 50878Date: Fri Dec 27 14:21:05 2013 -0800 50879 50880 glamor: Move the EGL DRI3 code to GLAMOR_HAS_GBM. 50881 50882 There's nothing dependent on the presence of DRI3 code in the server 50883 for this, but it does rely on GBM. 50884 50885 Signed-off-by: Eric Anholt <eric@anholt.net> 50886 Reviewed-by: Keith Packard <keithp@keithp.com> 50887 Reviewed-by: Adam Jackson <ajax@redhat.com> 50888 50889commit 80cdbb20f9dcdfdf3e77ecf5cdd7dfb1a87d1a83 50890Author: Eric Anholt <eric@anholt.net> 50891Date: Fri Dec 27 13:21:21 2013 -0800 50892 50893 glamor: Fix linking of the gradient shaders on GLES2. 50894 50895 GLES2 sensibly doesn't allow you to attach multiple shaders for the 50896 same stage to a single program. This means we have to attach the 50897 whole thing in one glShaderSource call. 50898 50899 Signed-off-by: Eric Anholt <eric@anholt.net> 50900 Reviewed-by: Keith Packard <keithp@keithp.com> 50901 Reviewed-by: Adam Jackson <ajax@redhat.com> 50902 50903commit 76bd0f994959f30a37d41eaf06f43f23c3b23faf 50904Author: Eric Anholt <eric@anholt.net> 50905Date: Fri Dec 27 13:09:25 2013 -0800 50906 50907 glamor: Don't bother keeping references to shader stages for gradients. 50908 50909 They never get reattached to any other program, so saving them to 50910 unreference later is a waste of code. 50911 50912 Signed-off-by: Eric Anholt <eric@anholt.net> 50913 Reviewed-by: Keith Packard <keithp@keithp.com> 50914 Reviewed-by: Adam Jackson <ajax@redhat.com> 50915 50916commit 6e62cdf66d2f0baa4a26eff6f4917f73baf3b008 50917Author: Eric Anholt <eric@anholt.net> 50918Date: Fri Dec 27 11:29:27 2013 -0800 50919 50920 glamor: Fix typo in setting v_position's attrib location. 50921 50922 Assuming it was the first attribute assigned by the GL, it would have 50923 ended up with location 0 anyway. 50924 50925 Signed-off-by: Eric Anholt <eric@anholt.net> 50926 Reviewed-by: Keith Packard <keithp@keithp.com> 50927 Reviewed-by: Adam Jackson <ajax@redhat.com> 50928 50929commit f8d384fa8f0fa2a3ac04c4ad21fb04400490e8e6 50930Author: Eric Anholt <eric@anholt.net> 50931Date: Fri Dec 27 02:13:27 2013 -0800 50932 50933 glamor: Move shader precision stuff from build time to shader compile time. 50934 50935 This is the last desktop-versus-ES2 build ifdef in core glamor. 50936 50937 Signed-off-by: Eric Anholt <eric@anholt.net> 50938 Reviewed-by: Keith Packard <keithp@keithp.com> 50939 Reviewed-by: Adam Jackson <ajax@redhat.com> 50940 50941commit 1adac62a7d9da099c779e65527ece611cf4d2f2c 50942Author: Eric Anholt <eric@anholt.net> 50943Date: Fri Dec 27 02:10:53 2013 -0800 50944 50945 glamor: Unifdef the picture-format-to-format-and-type functions. 50946 50947 Signed-off-by: Eric Anholt <eric@anholt.net> 50948 Reviewed-by: Keith Packard <keithp@keithp.com> 50949 Reviewed-by: Adam Jackson <ajax@redhat.com> 50950 50951commit 3747c260816a36352bb91ce06befcbffbbd747c7 50952Author: Eric Anholt <eric@anholt.net> 50953Date: Fri Dec 27 16:12:15 2013 -0800 50954 50955 glamor: Move glamor_get_tex_format_type_from_pictformat to a .c file. 50956 50957 A pair of 150 lines of inlined switch statements in a header file is 50958 crazy. 50959 50960 Signed-off-by: Eric Anholt <eric@anholt.net> 50961 Reviewed-by: Keith Packard <keithp@keithp.com> 50962 Reviewed-by: Adam Jackson <ajax@redhat.com> 50963 50964commit 0e4f3414189b1820443b35248e4c9e03f0c2e34e 50965Author: Eric Anholt <eric@anholt.net> 50966Date: Fri Dec 27 02:04:11 2013 -0800 50967 50968 glamor: Unifdef the cache format indices. 50969 50970 We only ask for GL_RGB on desktop GL as far as I can see, but now if 50971 GLES2 did happen to ask for GL_RGB it would return a cache index 50972 instead of -1. 50973 50974 Signed-off-by: Eric Anholt <eric@anholt.net> 50975 Reviewed-by: Keith Packard <keithp@keithp.com> 50976 Reviewed-by: Adam Jackson <ajax@redhat.com> 50977 50978commit d63283860a7c04a12838dead0dfd6d04fb73a093 50979Author: Eric Anholt <eric@anholt.net> 50980Date: Fri Dec 27 01:23:15 2013 -0800 50981 50982 glamor: Pass pixmaps around to unifdef glamor_iformat_for_depth(). 50983 50984 v2: Just pass in the PicturePtr to glamor_pict_format_is_compatible() 50985 (suggestion by keithp) 50986 50987 Signed-off-by: Eric Anholt <eric@anholt.net> 50988 Reviewed-by: Keith Packard <keithp@keithp.com> 50989 Reviewed-by: Adam Jackson <ajax@redhat.com> 50990 50991commit f3f4fc7a65589a200a086ea7b1527f91941bc19b 50992Author: Eric Anholt <eric@anholt.net> 50993Date: Fri Dec 27 01:17:10 2013 -0800 50994 50995 glamor: Add a screen argument to drop an ifdef from glamor_set_alu(). 50996 50997 Signed-off-by: Eric Anholt <eric@anholt.net> 50998 Reviewed-by: Keith Packard <keithp@keithp.com> 50999 Reviewed-by: Adam Jackson <ajax@redhat.com> 51000 51001commit 9553020b7184ed7a7ef3f02d60556d519ea3e769 51002Author: Eric Anholt <eric@anholt.net> 51003Date: Wed Dec 25 11:39:13 2013 -0800 51004 51005 glamor: Drop a bunch of GLES2 ifdefs. 51006 51007 Now that we're using epoxy, we can write code using both desktop and 51008 ES symbols and decide what to use at runtime. 51009 51010 v2: Fix a spelling mistake (latter), since the lines were moved 51011 anyway (noticed by Rémi Cardona). Fix condition invert in 51012 glamor_set_composite_texture (caught by Michel Dänzer). 51013 51014 Signed-off-by: Eric Anholt <eric@anholt.net> 51015 Reviewed-by: Keith Packard <keithp@keithp.com> (v1) 51016 Reviewed-by: Adam Jackson <ajax@redhat.com> (v1) 51017 51018commit c3c8a5f36034b6a2ce48d1d42e3de5af36406c38 51019Author: Eric Anholt <eric@anholt.net> 51020Date: Wed Dec 25 11:36:35 2013 -0800 51021 51022 glamor: yInverted is a boolean value, so use the Bool type. 51023 51024 Signed-off-by: Eric Anholt <eric@anholt.net> 51025 Reviewed-by: Keith Packard <keithp@keithp.com> 51026 Reviewed-by: Adam Jackson <ajax@redhat.com> 51027 51028commit 12cbfeed81d3ae73e10ee8a2b6619fb4f403b8f6 51029Author: Eric Anholt <eric@anholt.net> 51030Date: Mon Dec 23 18:00:46 2013 -0800 51031 51032 glamor: Drop fixed function transformation matrix setup. 51033 51034 gl_ModelViewProjection and friends aren't used in our shaders, so this 51035 setup didn't do anything. 51036 51037 Signed-off-by: Eric Anholt <eric@anholt.net> 51038 Reviewed-by: Keith Packard <keithp@keithp.com> 51039 Reviewed-by: Adam Jackson <ajax@redhat.com> 51040 51041commit 62965d278c347ff323f2262d767978794e32f841 51042Author: Eric Anholt <eric@anholt.net> 51043Date: Mon Dec 23 17:50:46 2013 -0800 51044 51045 glamor: Drop useless glEnable/glDisable(GL_TEXTURE_2D) calls. 51046 51047 Those calls are only for enabling texture handling in the fixed 51048 function pipeline, while everything we do is with shaders. 51049 51050 Signed-off-by: Eric Anholt <eric@anholt.net> 51051 Reviewed-by: Keith Packard <keithp@keithp.com> 51052 Reviewed-by: Adam Jackson <ajax@redhat.com> 51053 51054commit 4afe15d8bfd575c010ed1868697a7922a37ab378 51055Author: Eric Anholt <eric@anholt.net> 51056Date: Wed Dec 18 15:27:52 2013 -0800 51057 51058 glamor: Put in a pluggable context switcher for GLX versus EGL. 51059 51060 The GLX side just gets the context from the current state. That's 51061 also something I want to do for EGL, so that the making a context is 51062 separate from initializing glamor, but I think I need the modesetting 51063 driver in the server before I think about hacking on that more. 51064 51065 The previous code was rather incestuous, along with pulling in xf86 51066 dependencies to our dix code. The new code just initializes itself 51067 from the current state. 51068 51069 Signed-off-by: Eric Anholt <eric@anholt.net> 51070 Reviewed-by: Keith Packard <keithp@keithp.com> 51071 51072commit 781c692cf970642865f0e537f4905c43192f9935 51073Author: Eric Anholt <eric@anholt.net> 51074Date: Wed Dec 18 13:10:24 2013 -0800 51075 51076 glamor: Rename glamor_get/put_dispatch to glamor_get/put_context. 51077 51078 It used to be the thing that returned your dispatch table and happeend 51079 to set up the context, but now it just sets up the context. 51080 51081 Signed-off-by: Eric Anholt <eric@anholt.net> 51082 Reviewed-by: Keith Packard <keithp@keithp.com> 51083 51084commit 0373b3f4f7c7aec633468b37f9236a2734dbcc74 51085Author: Eric Anholt <eric@anholt.net> 51086Date: Wed Dec 18 12:18:20 2013 -0800 51087 51088 glamor: Convert to using libepoxy. 51089 51090 Libepoxy hides all the GL versus GLES2 dispatch handling for us, with 51091 higher performance. 51092 51093 v2: Squash in the later patch to drop the later of two repeated 51094 glamor_get_dispatch()es instead (caught by keithp) 51095 51096 Signed-off-by: Eric Anholt <eric@anholt.net> 51097 Reviewed-by: Keith Packard <keithp@keithp.com> 51098 51099commit f34dc7fa96457ea6a0703493d74e63cca357712e 51100Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 51101Date: Fri Feb 14 00:31:38 2014 -0800 51102 51103 XQuartz: pointer -> void * 51104 51105 Fix build regression from 60014a4a98ff924ae7f6840781f768c1cc93bbab 51106 51107 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 51108 51109commit 72237e0908527e9261b91c7db99f32d895947d8d 51110Author: Keith Packard <keithp@keithp.com> 51111Date: Thu Feb 13 15:25:56 2014 -0800 51112 51113 DIST_SUBDIRS needs to include glamor, even if it isn't built 51114 51115 Otherwise, glamor won't get cleaned for 'make distclean' 51116 51117 Signed-off-by: Keith Packard <keithp@keithp.com> 51118 51119commit 783991b1beeb71d91068601789d179d10eb8b544 51120Author: Gaetan Nadon <memsize@videotron.ca> 51121Date: Thu Feb 13 17:36:34 2014 -0500 51122 51123 config: fails to create tarball as xorg-server.conf file removed 51124 51125 Just need to update EXTRA_DIST 51126 51127 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51128 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 51129 Signed-off-by: Keith Packard <keithp@keithp.com> 51130 51131commit 02fbae85e7aa17b213e1411fe558256998d367c4 51132Author: Alex Deucher <alexander.deucher@amd.com> 51133Date: Wed Feb 12 13:06:51 2014 -0500 51134 51135 modesetting: query cursor size from the kernel 51136 51137 Use new drm caps. This allows hw cursors to work 51138 correctly on gpus with non-64x64 cursors. 51139 51140 Signed-off-by: Alex Deucher <alexander.deucher@amd.com> 51141 51142commit 30c3852bda7f60b0ffb1bb2f6ed8ba8800001b32 51143Author: Alan Coopersmith <alan.coopersmith@oracle.com> 51144Date: Sun Feb 9 16:56:40 2014 -0800 51145 51146 Delete stray ; in struct _DeviceChangedEvent 51147 51148 Caused Solaris Studio cc to complain in every file which included it: 51149 "../include/eventstr.h", line 179: warning: syntax error: 51150 empty member declaration 51151 51152 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 51153 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51154 Signed-off-by: Keith Packard <keithp@keithp.com> 51155 51156commit 5300212ce8e9364ba26497605f3edc089af20130 51157Merge: ae796d43c 0b193b3ac 51158Author: Keith Packard <keithp@keithp.com> 51159Date: Sun Feb 9 16:05:42 2014 -0800 51160 51161 Merge remote-tracking branch 'whot/for-keith' 51162 51163commit 0b193b3ac94e078d9291d1b1dfd4814e5f2d9e34 51164Author: Peter Hutterer <peter.hutterer@who-t.net> 51165Date: Thu Feb 6 07:59:07 2014 +1000 51166 51167 xfree86: use xnfstrdup in the Xorg -configure code 51168 51169 Just for consistency, I'm pretty sure the code is generally not happy for 51170 malloc failures anyway. 51171 51172 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51173 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51174 Reviewed-by: Keith Packard <keithp@keithp.com> 51175 51176commit 7b1b7fb3be47672454837a3f7be2d1440433ec27 51177Author: Peter Hutterer <peter.hutterer@who-t.net> 51178Date: Wed Feb 5 15:12:22 2014 +1000 51179 51180 mi: fix printf warning about size_t format specifier 51181 51182 mieq.c:290:26: warning: format '%u' expects argument of type 'unsigned int', 51183 but argument 2 has type 'size_t' [-Wformat=] 51184 51185 pnprintf supports size_t since 5ea21560dd071ea4ab87430000d087fd5fe1f092 51186 51187 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51188 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51189 Reviewed-by: Keith Packard <keithp@keithp.com> 51190 51191commit ddeca927498c9df3b5e62c7bf05e31e2a3aeaa52 51192Author: Mark Kettenis <kettenis@openbsd.org> 51193Date: Sun Dec 15 14:31:10 2013 +0100 51194 51195 sync: Avoid ridiculously long timeouts 51196 51197 On OpenBSD, passing a timeout longer than 100000000 seconds to select(2) will 51198 make it fail with EINVAL. As this is original 4.4BSD behaviour it is not 51199 inconceivable that other systems suffer from the same problem. And Linux, 51200 though not suffering from any 4.4BSD heritage, briefly did something similar: 51201 51202 <https://lkml.org/lkml/2012/8/31/263> 51203 51204 So avoid calling AdjustWaitForDelay() instead of setting the timeout to 51205 (effectively) ULONG_MAX milliseconds. 51206 51207 Signed-off-by: Mark Kettenis <kettenis@openbsd.org> 51208 Reviewed-by: Matthieu Herrb <matthieu@herrb.eu> 51209 51210commit ae796d43c934ba378c9a618adc81c6729a14b2f8 51211Author: Keith Packard <keithp@keithp.com> 51212Date: Thu Feb 6 19:17:50 2014 -0800 51213 51214 ephyr: Repaint entire screen when colormap is updated 51215 51216 Any time the colormap is changed, the entire screen needs to be 51217 repainted to match. 51218 51219 Signed-off-by: Keith Packard <keithp@keithp.com> 51220 Reviewed-by: Adam Jackson <ajax@redhat.com> 51221 Reviewed-by: Eric Anholt <eric@anholt.net> 51222 51223commit 9eecc9ac73aa06dca1420e0a89fb0cbd432a9bd7 51224Author: David Heidelberger <david.heidelberger@ixit.cz> 51225Date: Fri Feb 7 02:06:04 2014 +0100 51226 51227 configure.ac: add missing "test" 51228 51229 Signed-off-by: Keith Packard <keithp@keithp.com> 51230 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51231 51232commit 08c7df9b0870fadf7b4655825459ff2e5a5c47da 51233Author: Alan Coopersmith <alan.coopersmith@oracle.com> 51234Date: Mon Feb 3 17:38:24 2014 -0800 51235 51236 __glGetProcAddress: explictly cast function pointers to void * 51237 51238 Fixes Solaris Studio compiler warning & error: 51239 51240 "glxext.c", line 557: warning: assignment type mismatch: 51241 pointer to void "=" pointer to function(void) returning void 51242 "glxext.c", line 559: error: operands have incompatible types: 51243 pointer to void ":" pointer to function(void) returning void 51244 51245 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 51246 Signed-off-by: Keith Packard <keithp@keithp.com> 51247 51248commit 76eedb039fc8d515a76c1df944fd5a85ac674019 51249Author: Eric Anholt <eric@anholt.net> 51250Date: Mon Jan 27 11:36:08 2014 -0800 51251 51252 xfree86: Fix a compiler warning on 64-bit. 51253 51254 asm/mtrr.h makes this an unsigned long on 32, but a u64 on 64. Cast 51255 it to a long to win. 51256 51257 Signed-off-by: Eric Anholt <eric@anholt.net> 51258 Reviewed-by: Adam Jackson <ajax@redhat.com> 51259 Signed-off-by: Keith Packard <keithp@keithp.com> 51260 51261commit 9f8f6657cdd3a5b166771695addb6fe76d93c378 51262Author: Eric Anholt <eric@anholt.net> 51263Date: Mon Jan 27 11:36:07 2014 -0800 51264 51265 glx: Delete dead NV program string functions. 51266 51267 These have been throwing a compiler warning about missing prototypes, 51268 since the generated code to define the prototypes stopped being 51269 generated (possibly because the code was dead). 51270 51271 Signed-off-by: Eric Anholt <eric@anholt.net> 51272 Reviewed-by: Adam Jackson <ajax@redhat.com> 51273 Signed-off-by: Keith Packard <keithp@keithp.com> 51274 51275commit 0c774d53c5c1e8845e7da9b01814d7b98f621f51 51276Author: Eric Anholt <eric@anholt.net> 51277Date: Mon Jan 27 11:36:06 2014 -0800 51278 51279 glx: Reduce compiler warnings by not requesting GL extension prototypes. 51280 51281 They're not officially in the ABI, so you shouldn't use them anyway. 51282 51283 Signed-off-by: Eric Anholt <eric@anholt.net> 51284 Reviewed-by: Adam Jackson <ajax@redhat.com> 51285 Signed-off-by: Keith Packard <keithp@keithp.com> 51286 51287commit 87c4551c9cc9e382e6cfb59df1759d27c776abc8 51288Author: Eric Anholt <eric@anholt.net> 51289Date: Mon Jan 27 11:36:05 2014 -0800 51290 51291 glx: Stop relying on libGL ABI bugs for glGetCompressedTexImage(). 51292 51293 In theory, the linux libGL ABI exposes just GL 1.2 plus GLX 1.3. But, 51294 thanks to libglapi, we're letting glGetCompressedTexImageARB() be 51295 exposed too. The GLX code was inappropriately relying on it by using 51296 GL_GLEXT_PROTOTYPES. 51297 51298 Signed-off-by: Eric Anholt <eric@anholt.net> 51299 Reviewed-by: Adam Jackson <ajax@redhat.com> 51300 Signed-off-by: Keith Packard <keithp@keithp.com> 51301 51302commit bf4f02337c97ffc3f7fbba9dba8ab72fa6c4a5fd 51303Author: Jon TURNEY <jon.turney@dronecode.org.uk> 51304Date: Mon Jan 6 14:13:24 2014 +0000 51305 51306 glx: Remove left-over glthread.c 51307 51308 Commit be668096 "glx: convert to direct GL dispatch (v2)" removes glthread.c 51309 from Makefile.am along with the rest of the dispatch table code, but doesn't 51310 remove glthread.c itself. 51311 51312 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 51313 Reviewed-by: Eric Anholt <eric@anholt.net> 51314 Reviewed-by: Adam Jackson <ajax@redhat.com> 51315 Signed-off-by: Keith Packard <keithp@keithp.com> 51316 51317commit 7305153643622269e14f3564f7a8a66ecaf49f78 51318Author: Adam Jackson <ajax@redhat.com> 51319Date: Wed Feb 5 11:10:47 2014 -0500 51320 51321 configure: Don't add GLX_SYS_LIBS to XORG_SYS_LIBS 51322 51323 libglx.so is linked against libGL.so, but Xorg itself should not be. 51324 51325 Signed-off-by: Adam Jackson <ajax@redhat.com> 51326 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51327 Signed-off-by: Keith Packard <keithp@keithp.com> 51328 51329commit c64130c13bd6914a998a8e8b808a6cedd18f9384 51330Author: Peter Hutterer <peter.hutterer@who-t.net> 51331Date: Wed Feb 5 15:10:27 2014 +1000 51332 51333 dix: fix a -Wshadow warning 51334 51335 dispatch.c: In function 'SetVendorString': 51336 dispatch.c:481:29: warning: declaration of 'string' shadows a global declaration [-Wshadow] 51337 SetVendorString(const char *string) 51338 ^ 51339 dispatch.c:135:21: warning: shadowed declaration is here [-Wshadow] 51340 typedef const char *string; 51341 51342 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51343 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51344 Reviewed-by: Keith Packard <keithp@keithp.com> 51345 51346commit 72967d6c153b0d1109df23967e1a05c3c397a1e0 51347Author: Peter Hutterer <peter.hutterer@who-t.net> 51348Date: Wed Feb 5 14:32:23 2014 +1000 51349 51350 xfree86: unconstify driver in the XF86ConfInputClassRec 51351 51352 No const value is ever assigned to it, let's not pretend it's const. 51353 51354 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51355 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51356 Reviewed-by: Keith Packard <keithp@keithp.com> 51357 51358commit f14d6647c0883d4e7e4de7645d7a18e4d301845e 51359Author: Peter Hutterer <peter.hutterer@who-t.net> 51360Date: Wed Feb 5 14:30:50 2014 +1000 51361 51362 xfree86: unconstify another string 51363 51364 Only Xorg -configure uses a hardcoded value here, so let's not change the rest 51365 of the server for that. 51366 51367 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51368 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51369 Reviewed-by: Keith Packard <keithp@keithp.com> 51370 51371commit 46ae9d67e4118fa79bef0f9119d20559dfd6b6c0 51372Author: Peter Hutterer <peter.hutterer@who-t.net> 51373Date: Wed Feb 5 14:24:54 2014 +1000 51374 51375 xfree86: un-constify inp_driver/inp_identifier 51376 51377 The only place this isn't allocated is during Xorg -configure where we just 51378 statically assing "mouse"/"kbd" and the identifiers for it. Everywhere else 51379 it's strdup'd and then free'd already. 51380 51381 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51382 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51383 Reviewed-by: Keith Packard <keithp@keithp.com> 51384 51385commit 93bf9544712a39f10557533993d8826b2b67fc9a 51386Author: Peter Hutterer <peter.hutterer@who-t.net> 51387Date: Wed Feb 5 14:16:01 2014 +1000 51388 51389 xfree86: device name and driver are not const char 51390 51391 Allocated in one place, freed in another. 51392 51393 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51394 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51395 Reviewed-by: Keith Packard <keithp@keithp.com> 51396 51397commit 25d10464f440b8b34594b7c988a99a830ea39a29 51398Author: Peter Hutterer <peter.hutterer@who-t.net> 51399Date: Fri Jan 24 18:16:54 2014 +1000 51400 51401 dix: fix button state check before changing a button mapping 51402 51403 dev->button->down is a bitmask, not a normal array. Use the helper function to 51404 check, we technically allow the mapping to change after the physical button 51405 has been pressed (but not yet processed yet), so only check BUTTON_PROCESSED. 51406 51407 From XSetPointerMapping(3): 51408 "If any of the buttons to be altered are logically in the down state, 51409 XSetPointerMapping returns MappingBusy, and the mapping is not changed." 51410 51411 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 51412 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51413 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 51414 51415commit 73926622b91fde01148727f26d6aad5e6827c1d2 51416Author: Peter Hutterer <peter.hutterer@who-t.net> 51417Date: Fri Jan 24 18:33:54 2014 +1000 51418 51419 include: make the various button array lengths more obvious 51420 51421 No functional changes, just making a better case for why MAP_LENGTH is 256. 51422 "But can't we remove MAP_LENGTH then?" I hear you say? "Why, yes. Go for it!" 51423 51424 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51425 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 51426 51427commit 87ca80a7196949597113225405f3e4ee03bbee13 51428Author: Peter Hutterer <peter.hutterer@who-t.net> 51429Date: Fri Jan 24 18:32:54 2014 +1000 51430 51431 dix: prevent a driver from initializing or submitting buttons > MAX_BUTTONS 51432 51433 The server internally relies on arrays with a MAX_BUTTONS maximum size (which 51434 is the max the core protocol can transport). Make sure a driver adheres to 51435 that. 51436 51437 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51438 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 51439 51440commit 14fb6cf92c009d726f0a6b6e8b89cc48f2dd50eb 51441Author: Peter Hutterer <peter.hutterer@who-t.net> 51442Date: Thu Jan 30 09:16:23 2014 +1000 51443 51444 Revert "xfree86/common: handle string constants in xf86Xinput configuration" 51445 51446 This reverts commit 22592855e90d23013ba7f9e945d567725cb44bf3. 51447 51448 What warning was this supposed to fix? 51449 51450 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51451 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51452 51453commit a553444b5841522836f7437ebb96e40be270fd8b 51454Author: Peter Hutterer <peter.hutterer@who-t.net> 51455Date: Thu Jan 30 09:12:41 2014 +1000 51456 51457 Revert "xfree86/parser: make strings in xf86MatchGroup const" 51458 51459 This reverts commit f71de60355cc76810657f40c7b5461af86b34bf7. 51460 51461 What warnings? 51462 51463 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51464 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51465 51466commit 45f1d527f39a296104f2fa79a774446e7e1560e0 51467Author: Peter Hutterer <peter.hutterer@who-t.net> 51468Date: Thu Jan 30 09:01:09 2014 +1000 51469 51470 input: un-constify dev->name 51471 51472 Fallout from fecc7eb1cf66db64728ee2d68cd9443df7e70879, and reverts most of the 51473 rest of that patch. 51474 51475 The device name is allocated and may even change during PreInit. The const 51476 warnings came from the test codes, the correct fix here is to fix the test 51477 code. 51478 51479 touch.c: In function ‘touch_init’: 51480 touch.c:254:14: warning: assignment discards ‘const’ qualifier from pointer target type [enabled by default] 51481 dev.name = "test device"; 51482 51483 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51484 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51485 51486commit 675f215af291135ee3ece5414e4a5a2e89bf4ed3 51487Author: Peter Hutterer <peter.hutterer@who-t.net> 51488Date: Thu Jan 30 08:52:32 2014 +1000 51489 51490 Revert "os: xstrtokenize takes and returns const char * now" 51491 51492 This reverts commit d0339a5c66846c9f14e3b584e34688520a0916ab. 51493 51494 seriously, what the fuck? Are we making xstrdup() return a const char now too? 51495 51496 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51497 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51498 51499commit ce3df579e3f24a169189b288230959527e059080 51500Author: Peter Hutterer <peter.hutterer@who-t.net> 51501Date: Thu Jan 30 08:50:06 2014 +1000 51502 51503 input: un-constify InputAttributes 51504 51505 Introduced in fecc7eb1cf66db64728ee2d68cd9443df7e70879 and reverts most of 51506 that but it's helpfully mixed with other stuff. 51507 51508 InputAttributes are not const, they're strdup'd everywhere but the test code 51509 and freed properly. Revert the const char changes and fix the test up instead. 51510 51511 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51512 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51513 51514commit 2fc38d1e299587d25ca8225051e0ea9403164b15 51515Author: Peter Hutterer <peter.hutterer@who-t.net> 51516Date: Wed Jan 29 11:16:51 2014 +1000 51517 51518 xkb: add a call to init an XkbRMLVOSet from const chars 51519 51520 Just forcing everything to const char* is not helpful, compiler warnings are 51521 supposed to warn about broken code. Forcing everything to const when it 51522 clearly isn't less than ideal. 51523 51524 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51525 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51526 51527commit 6307d60dd592f4b438f880d02bde9fd8d50ae264 51528Author: Peter Hutterer <peter.hutterer@who-t.net> 51529Date: Wed Jan 29 10:47:49 2014 +1000 51530 51531 Xi: remove superfluous cast. 51532 51533 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51534 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51535 51536commit 83e38eb73fd8c852513aac2da2975b4c01070ec2 51537Author: Arun Raghavan <arun@accosted.net> 51538Date: Mon Jan 6 20:29:50 2014 +0530 51539 51540 edid: Add quirk for Sony Vaio Pro 13 51541 51542 The detailed timings are for a 15.6" display when max image size 51543 correctly reports 13.3". 51544 51545 Signed-off-by: Arun Raghavan <arun@accosted.net> 51546 Reviewed-by: Adam Jackson <ajax@redhat.com> 51547 Signed-off-by: Keith Packard <keithp@keithp.com> 51548 51549commit bf83843b92ce21d11f6ff1a407ff3d014e017c9b 51550Author: Hans de Goede <hdegoede@redhat.com> 51551Date: Thu Dec 19 14:10:18 2013 +0100 51552 51553 xf86Events: add Enable/DisableInputDeviceForVTSwitch functions 51554 51555 Factor this code out into functions so that it can be re-used for the 51556 systemd-logind device pause/resume paths. 51557 51558 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 51559 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51560 51561commit 48b489769e78fa20911630173ab708feecb0fb0e 51562Author: Hans de Goede <hdegoede@redhat.com> 51563Date: Thu Dec 19 11:28:40 2013 +0100 51564 51565 xf86Events: refactor xf86VTLeave error handling 51566 51567 Use kernel goto style error handling for xf86VTSwitchAway() failure. This 51568 makes it much easier to read the straight path. 51569 51570 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 51571 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51572 51573commit 78f0667d6df9cc43a397d9f1490e540936a435d6 51574Author: Hans de Goede <hdegoede@redhat.com> 51575Date: Thu Dec 19 11:26:36 2013 +0100 51576 51577 xf86Events: split xf86VTSwitch into xf86VTLeave and xf86VTEnter functions 51578 51579 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 51580 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51581 51582commit 33cec8af55d829cd77b297ae356ed7a00ce8523c 51583Author: Hans de Goede <hdegoede@redhat.com> 51584Date: Tue Dec 17 09:29:07 2013 +0100 51585 51586 dbus-core: Attempt to connect to dbus ASAP 51587 51588 For systemd-logind integration we need the dbus connection to be available 51589 before enumerating input and gfx devices. 51590 51591 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 51592 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 51593 51594commit 480590b90c3966536451d2a2fecc42a66082ed77 51595Author: Hans de Goede <hdegoede@redhat.com> 51596Date: Wed Dec 4 11:10:06 2013 +0100 51597 51598 dbus-core: Make dbus-core no longer mutually exclusive with udev 51599 51600 With systemd-logind the dbus-core will be used for more then just config, so 51601 it should be possible to build it even when using a non dbus dependent config 51602 backend. 51603 51604 This patch also removes the config_ prefix from the dbus-core symbols. 51605 51606 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 51607 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 51608 51609commit c29454ae9d2e8e647732077fdfd97b351095f122 51610Author: Peter Hutterer <peter.hutterer@who-t.net> 51611Date: Thu Dec 12 12:48:57 2013 +1000 51612 51613 config: drop the dbus API 51614 51615 This API has been disabled by default since 1.4, the first release it came in. 51616 There a no known users of it and even its direct replacement (HAL) has 51617 been superseeded by udev on supported platforms since 1.8. 51618 51619 This code is untested, probably hasn't been compiled in years and should not 51620 be shipped. 51621 51622 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51623 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 51624 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51625 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 51626 51627commit 46cf2a60934076bf568062eb83121ce90b6ff596 51628Author: Laércio de Sousa <lbsousajr@gmail.com> 51629Date: Thu Dec 12 14:22:48 2013 -0200 51630 51631 xfree86: Keep a non-seat0 X server from touching VTs (#71258) 51632 51633 Updated patch following Hans de Goede's advice. 51634 51635 If -seat option is passed with a value different from seat0, 51636 X server won't call xf86OpenConsole(). 51637 51638 This is needed to avoid any race condition between seat0 and 51639 non-seat0 X servers. If a non-seat0 X server opens a given VT 51640 before a seat0 one which expects to open the same VT, one can 51641 get an inactive systemd-logind graphical session for seat0. 51642 51643 This patch was first tested in a multiseat setup with multiple 51644 video cards and works quite well. 51645 51646 I suppose it can also make things like DontVTSwitch and -sharevts 51647 meaningless for non-seat0 seats, so it may fix bug #69477, too. 51648 51649 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=71258 51650 https://bugs.freedesktop.org/show_bug.cgi?id=69477 (maybe) 51651 51652 See also: http://lists.x.org/archives/xorg-devel/2013-October/038391.html 51653 https://bugzilla.redhat.com/show_bug.cgi?id=1018196 51654 51655 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 51656 Reviewed-by: Hans de Goede <hdegoede@redhat.com> 51657 51658commit b3d3ffd19937827bcbdb833a628f9b1814a6e189 51659Author: Łukasz Stelmach <l.stelmach@samsung.com> 51660Date: Mon Nov 25 11:54:07 2013 +0100 51661 51662 configure.ac: enable systemd socket activation in libxtrans 51663 51664 Signed-off-by: Łukasz Stelmach <l.stelmach@samsung.com> 51665 Signed-off-by: Hans de Goede <hdegoede@redhat.com> 51666 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 51667 51668commit 435098a0dce6bca8870ec9725bf0af0969cd84fa 51669Author: Alan Coopersmith <alan.coopersmith@oracle.com> 51670Date: Tue Jan 28 20:27:52 2014 -0800 51671 51672 Add RANDR 1.4 requests & events to dix/protocol.txt 51673 51674 Checked against randrproto.txt & randr.h 51675 51676 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 51677 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51678 Signed-off-by: Keith Packard <keithp@keithp.com> 51679 51680commit c1ac89c793614797e08d3d8e7fc9ba55be899130 51681Author: Alan Coopersmith <alan.coopersmith@oracle.com> 51682Date: Tue Jan 28 20:27:51 2014 -0800 51683 51684 xf86DeleteScreen: move check for NULL pScrn before first dereference 51685 51686 Flagged by cppcheck 1.62: 51687 [hw/xfree86/common/xf86Helper.c:220] -> [hw/xfree86/common/xf86Helper.c:231]: 51688 (warning) Possible null pointer dereference: pScrn - otherwise it is 51689 redundant to check it against null. 51690 51691 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 51692 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51693 Signed-off-by: Keith Packard <keithp@keithp.com> 51694 51695commit e6733ae91b7be52930f22a87de15fa05819ef948 51696Author: Alan Coopersmith <alan.coopersmith@oracle.com> 51697Date: Tue Jan 28 20:27:50 2014 -0800 51698 51699 On realloc failure, free font_path_string instead of leaking it 51700 51701 Flagged by cppcheck 1.62: 51702 [dix/dixfonts.c:1792]: (error) Common realloc mistake: 51703 'font_path_string' nulled but not freed upon failure 51704 51705 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 51706 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51707 Signed-off-by: Keith Packard <keithp@keithp.com> 51708 51709commit 910b5b245425f0a866a703303b78768b0de5cb2b 51710Author: Alan Coopersmith <alan.coopersmith@oracle.com> 51711Date: Tue Jan 28 20:27:49 2014 -0800 51712 51713 Link libvgahw with $(PCIACCESS_LIBS) as well 51714 51715 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 51716 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51717 Signed-off-by: Keith Packard <keithp@keithp.com> 51718 51719commit da1660deeb9032ecca61f4bcdc9fc2eec2ada445 51720Merge: 0fbb3d711 9fc19168e 51721Author: Keith Packard <keithp@keithp.com> 51722Date: Mon Jan 27 19:02:15 2014 -0800 51723 51724 Merge remote-tracking branch 'whot/for-keith' 51725 51726commit 0fbb3d711efec5222a57b45a70d28fc98380f3a1 51727Merge: 7ddef4f70 b98e49379 51728Author: Keith Packard <keithp@keithp.com> 51729Date: Mon Jan 27 14:11:09 2014 -0800 51730 51731 Merge remote-tracking branch 'anholt/glamor-reformat' 51732 51733 Conflicts: 51734 Makefile.am 51735 51736 Conflict caused by adding PSEUDORAMIX and GLAMOR directory defines in 51737 separate branches 51738 51739commit 7ddef4f7033c10b6e92866182d4475a4d49c5083 51740Author: Keith Packard <keithp@keithp.com> 51741Date: Wed Jan 22 14:30:28 2014 -0800 51742 51743 Add _XITYPEDEF_POINTER to dix-config.h 51744 51745 Just like the pointer type from Xdefs.h, the Pointer type from 51746 XIproto.h collides with local declarations of variables using the same 51747 name. XIproto.h can use _XITYPEDEF_POINTER to avoid declaring the 51748 unnecessary pointer type. 51749 51750 Signed-off-by: Keith Packard <keithp@keithp.com> 51751 Reviewed-by: Eric Anholt <eric@anholt.net> 51752 51753commit 0b5a87f37d5dcd2ebff977a3b9a50a75ace93c83 51754Author: Keith Packard <keithp@keithp.com> 51755Date: Wed Jan 22 19:17:24 2014 -0800 51756 51757 pseudoramiX: Only compile on XQUARTZ and XWIN 51758 51759 PseudoramiXExtensionInit() is not defined in extinit.h if it won't be 51760 used and we get a compiler warning when compiling the pseudoramiX code. 51761 51762 Signed-off-by: Keith Packard <keithp@keithp.com> 51763 Reviewed-by: Eric Anholt <eric@anholt.net> 51764 51765commit 61cb6c9aa95aa369573c4e02c023bbc33c7678c9 51766Author: Keith Packard <keithp@keithp.com> 51767Date: Wed Jan 22 19:16:36 2014 -0800 51768 51769 pseudoramiX: Add _X_ATTRIBUTE_PRINTF attributes to debug functions. 51770 51771 Signed-off-by: Keith Packard <keithp@keithp.com> 51772 Reviewed-by: Eric Anholt <eric@anholt.net> 51773 51774commit 9e45a1a030a5e70318441f33a132269a19a43df5 51775Author: Keith Packard <keithp@keithp.com> 51776Date: Fri Nov 15 18:00:53 2013 +0900 51777 51778 Warning fixes in glx 51779 51780 Signed-off-by: Keith Packard <keithp@keithp.com> 51781 Reviewed-by: Eric Anholt <eric@anholt.net> 51782 51783commit b98e49379c8d7cecce991207048489f51b10028c 51784Author: Eric Anholt <eric@anholt.net> 51785Date: Wed Dec 18 15:00:19 2013 -0800 51786 51787 glamor: Remove more out-of-tree compat code. 51788 51789 Signed-off-by: Eric Anholt <eric@anholt.net> 51790 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51791 Reviewed-by: Keith Packard <keithp@keithp.com> 51792 51793commit 40a8186f0f78b7c36ef7e3d902c53bdc9bb7c650 51794Author: Eric Anholt <eric@anholt.net> 51795Date: Fri Dec 27 12:22:08 2013 -0800 51796 51797 glamor: Remove an extra copy of RegionNil(). 51798 51799 Signed-off-by: Eric Anholt <eric@anholt.net> 51800 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51801 Reviewed-by: Keith Packard <keithp@keithp.com> 51802 51803commit d26f5335151a9e962afa4bbf29ca5b94660a33ca 51804Author: Eric Anholt <eric@anholt.net> 51805Date: Wed Dec 18 11:35:53 2013 -0800 51806 51807 glamor: Hook the module back up to the build. 51808 51809 For now we're just building an uninstalled library. The extra EGL 51810 stubs are required so that we can get the DIX building and usable 51811 without pulling in the xf86 DDX code in glamor_egl.c. 51812 51813 Signed-off-by: Eric Anholt <eric@anholt.net> 51814 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51815 Reviewed-by: Keith Packard <keithp@keithp.com> 51816 51817commit 54e78ec31e030d488765341a0c143c5a060c3768 51818Author: Eric Anholt <eric@anholt.net> 51819Date: Wed Jan 22 15:03:26 2014 -0800 51820 51821 glamor: Convert use of the old "pointer" typedef to "void *". 51822 51823 Reviewed-by: Keith Packard <keithp@keithp.com> 51824 51825commit 3c3a4eeaa1f24b7534b332739158a2a36987ea6b 51826Author: Eric Anholt <eric@anholt.net> 51827Date: Wed Dec 18 11:45:33 2013 -0800 51828 51829 glamor: Silence warnings for non-debug builds. 51830 51831 Signed-off-by: Eric Anholt <eric@anholt.net> 51832 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51833 Reviewed-by: Keith Packard <keithp@keithp.com> 51834 51835commit 9af66851e2770bcd8408a0e5ddf8bb8ea816feaa 51836Author: Eric Anholt <eric@anholt.net> 51837Date: Fri Jan 24 10:58:04 2014 -0800 51838 51839 glamor: Disable definitions of GL extension prototypes to avoid warnings. 51840 51841 We're not using the extension prototypes, since you have to dlsym them 51842 anyway. Disabling their definitions prevents them from being defined 51843 twice (once by gl.h, once by glext.h). 51844 51845 Signed-off-by: Eric Anholt <eric@anholt.net> 51846 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51847 51848commit a7b8ce8b4207a93c0455a0b796cfc47917c04a9c 51849Author: Eric Anholt <eric@anholt.net> 51850Date: Wed Dec 18 11:41:33 2013 -0800 51851 51852 glamor: Drop xfree86 dependencies from this dix module. 51853 51854 Signed-off-by: Eric Anholt <eric@anholt.net> 51855 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51856 Reviewed-by: Keith Packard <keithp@keithp.com> 51857 51858commit 7759e4d090f56cdcad1c8a2acb2c1be20021de5d 51859Author: Eric Anholt <eric@anholt.net> 51860Date: Wed Dec 18 11:43:48 2013 -0800 51861 51862 glamor: Disable the XV code for now. 51863 51864 We're going to want to make this DIX code instead of XF86 if at all 51865 possible, but for now just disable it so we can work on the rest of 51866 the build. 51867 51868 Signed-off-by: Eric Anholt <eric@anholt.net> 51869 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51870 Reviewed-by: Keith Packard <keithp@keithp.com> 51871 51872commit b3acb47e98023da898ffc4b6a5bac38a78bd7727 51873Author: Adam Jackson <ajax@redhat.com> 51874Date: Wed Oct 30 11:30:23 2013 -0400 51875 51876 glamor: Use dix-config.h not project config.h 51877 51878 v2: Also edit the one in glamor_egl.c (by anholt) 51879 v3: Also edit the one in glamor_eglmodule.c (by anholt) 51880 51881 Signed-off-by: Adam Jackson <ajax@redhat.com> 51882 Signed-off-by: Eric Anholt <eric@anholt.net> 51883 Reviewed-by: Eric Anholt <eric@anholt.net> 51884 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51885 Reviewed-by: Keith Packard <keithp@keithp.com> 51886 51887commit 6cc0b7b01599b94af07e9fc8b16134f751ede077 51888Author: Adam Jackson <ajax@redhat.com> 51889Date: Wed Oct 30 11:05:40 2013 -0400 51890 51891 glamor/egl: Remove glapi awareness 51892 51893 We only needed this because glx was directly bashing glapi tables. 51894 Since that's not the case anymore, we should just MakeCurrent like a 51895 real GL app. 51896 51897 v2: Hand-resolve against rebase onto newer server (by anholt) 51898 51899 Signed-off-by: Adam Jackson <ajax@redhat.com> 51900 Signed-off-by: Eric Anholt <eric@anholt.net> 51901 Reviewed-by: Eric Anholt <eric@anholt.net> 51902 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51903 Reviewed-by: Keith Packard <keithp@keithp.com> 51904 51905commit 0c5a7c208601110a0b36d24cdacb4d844af03f75 51906Author: Eric Anholt <eric@anholt.net> 51907Date: Tue Dec 17 16:32:17 2013 -0800 51908 51909 glamor: Remove compat code for building out of tree. 51910 51911 Signed-off-by: Eric Anholt <eric@anholt.net> 51912 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51913 Reviewed-by: Keith Packard <keithp@keithp.com> 51914 51915commit 82efb90efba685678638f94804d773b59ca1643f 51916Author: Adam Jackson <ajax@redhat.com> 51917Date: Wed Oct 30 09:56:46 2013 -0400 51918 51919 glamor: Remove copy of sna's compiler.h 51920 51921 Xfuncproto.h has equivalents for these already. 51922 51923 v2: Adjust a couple more likelies after the rebase (anholt) 51924 51925 Signed-off-by: Adam Jackson <ajax@redhat.com> 51926 Signed-off-by: Eric Anholt <eric@anholt.net> 51927 Reviewed-by: Eric Anholt <eric@anholt.net> 51928 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51929 Reviewed-by: Keith Packard <keithp@keithp.com> 51930 51931commit 714926b090344a17a2f7aee9f4bc6aea4e45f48c 51932Author: Eric Anholt <eric@anholt.net> 51933Date: Thu Jan 9 18:23:39 2014 +0800 51934 51935 glamor: Fix up some indentation damage on header prototypes. 51936 51937 Signed-off-by: Eric Anholt <eric@anholt.net> 51938 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51939 Reviewed-by: Keith Packard <keithp@keithp.com> 51940 51941commit 7f6e865359c2d055db4eb7d82b4779b3d7c5d264 51942Author: Eric Anholt <eric@anholt.net> 51943Date: Fri Dec 27 11:46:30 2013 -0800 51944 51945 glamor: Fix some indent damage of putting a ' ' after the '*' for pointers. 51946 51947 Signed-off-by: Eric Anholt <eric@anholt.net> 51948 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51949 Reviewed-by: Keith Packard <keithp@keithp.com> 51950 51951commit 5f57d436c391c51f3f90958b033f6ee3eb7a1136 51952Author: Eric Anholt <eric@anholt.net> 51953Date: Tue Dec 24 15:59:14 2013 -0800 51954 51955 glamor: Fix some mangling of shader strings by indent. 51956 51957 Signed-off-by: Eric Anholt <eric@anholt.net> 51958 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51959 Reviewed-by: Keith Packard <keithp@keithp.com> 51960 51961commit d84d71029ae9e462559d64eff7259e2cc7732fac 51962Author: Eric Anholt <eric@anholt.net> 51963Date: Wed Dec 18 11:59:07 2013 -0800 51964 51965 glamor: Apply x-indent.sh. 51966 51967 Signed-off-by: Eric Anholt <eric@anholt.net> 51968 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51969 Reviewed-by: Keith Packard <keithp@keithp.com> 51970 51971commit e8e9a54c47deedbc13b8969ed29e16463b0314a0 51972Author: Eric Anholt <eric@anholt.net> 51973Date: Wed Dec 18 11:56:15 2013 -0800 51974 51975 glamor: Touch up some code so indent doesn't get confused. 51976 51977 Signed-off-by: Eric Anholt <eric@anholt.net> 51978 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 51979 Reviewed-by: Keith Packard <keithp@keithp.com> 51980 51981commit 9fc19168e7ca6308275bf8769d1ccb982f88465b 51982Author: Alan Coopersmith <alan.coopersmith@oracle.com> 51983Date: Fri Jan 24 23:42:49 2014 -0800 51984 51985 Check for calloc() failure in add_master() 51986 51987 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 51988 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 51989 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 51990 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 51991 51992commit 76b3be75b62657e346731444736f7e4d200beb5b 51993Author: Peter Hutterer <peter.hutterer@who-t.net> 51994Date: Fri Jan 24 16:51:02 2014 +1000 51995 51996 Xi: fix modifier offset in XIPassiveGrab swapping function 51997 51998 The request is followed by mask_len 4-byte units, then followed by the actual 51999 modifiers. 52000 52001 Also fix up the swapping test, which had the same issue. 52002 52003 Reported-by: Alan Coopersmith <alan.coopersmith@oracle.com> 52004 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 52005 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 52006 52007commit c1ce807d9f18f215332d7eeb844e8c640f71c53c 52008Author: Keith Packard <keithp@keithp.com> 52009Date: Wed Jan 22 11:01:59 2014 -0800 52010 52011 dix: Praise clients which haven't run for a while, rather than idle clients 52012 52013 A client which is ready, but hasn't run for a while, should receive 52014 the same benefit as one which has simply been idle for a while. Use 52015 the smart_stop_tick to see how long it has been since a client has 52016 run instead of smart_check_tick, which got reset each time a client 52017 was ready, even if it didn't get to run. 52018 52019 Reported-by: Chris Wilson <chris@chris-wilson.co.uk> 52020 Signed-off-by: Keith Packard <keithp@keithp.com> 52021 Tested-by: Chris Wilson <chris@chris-wilson.co.uk> 52022 52023commit 76b275d7acbfd6b28cb9e74fa4862faa6d08217d 52024Author: Keith Packard <keithp@keithp.com> 52025Date: Sun Jan 12 10:53:31 2014 -0800 52026 52027 test/signal-formatting: Ignore compiler warnings 52028 52029 The signal formatting tests intentionally include code which generates 52030 warnings with the current X server warning flags. Turn the compiler 52031 warnings off 52032 52033 Signed-off-by: Keith Packard <keithp@keithp.com> 52034 Reviewed-by: Eric Anholt <eric@anholt.net> 52035 52036commit 12ea81792868c10a89a06ce9a9e8f707be283569 52037Author: Keith Packard <keithp@keithp.com> 52038Date: Sun Jan 12 10:44:22 2014 -0800 52039 52040 test/hashtabletest: Clean up -Wshadow errors 52041 52042 Declare 'XID id' local to each scope it is used in, rather than having 52043 the first use be a function-wide declaration. 52044 52045 Signed-off-by: Keith Packard <keithp@keithp.com> 52046 Reviewed-by: Eric Anholt <eric@anholt.net> 52047 52048commit 7104f0f338fa96824a463331d6099a76ee3c21ac 52049Author: Keith Packard <keithp@keithp.com> 52050Date: Sun Jan 12 10:42:37 2014 -0800 52051 52052 test/xi2: Clean up -Wshadow warnings 52053 52054 protocol-common declares a bunch of pretty generic names; fix shadows 52055 of these names. 52056 52057 Signed-off-by: Keith Packard <keithp@keithp.com> 52058 Reviewed-by: Eric Anholt <eric@anholt.net> 52059 52060commit 92a9495800cf59b917c30d5fbaf4b7eca10d19cf 52061Author: Keith Packard <keithp@keithp.com> 52062Date: Sun Jan 12 10:42:01 2014 -0800 52063 52064 kdrive/ephyr: Remove extra 'i' variable in ProcXF86DRIGetDrawableInfo 52065 52066 Just re-use the top-level one 52067 52068 Signed-off-by: Keith Packard <keithp@keithp.com> 52069 Reviewed-by: Eric Anholt <eric@anholt.net> 52070 52071commit 988877065486d0f25cc85be5be05f4523cc883c7 52072Author: Keith Packard <keithp@keithp.com> 52073Date: Sun Dec 15 01:57:09 2013 -0800 52074 52075 randr: Eliminate -Wshadow warnings 52076 52077 Don't use rrScrPriv for nested screen private fetching. 52078 52079 Eliminate a duplicate fetch of the randr screen private in rrCheckPixmapBounding. 52080 52081 Signed-off-by: Keith Packard <keithp@keithp.com> 52082 Reviewed-by: Eric Anholt <eric@anholt.net> 52083 52084commit f46c487625e1ffbc78df5153128c68bd6bc41104 52085Author: Keith Packard <keithp@keithp.com> 52086Date: Sun Dec 15 01:56:36 2013 -0800 52087 52088 hw/xnest: Eliminate shadowed names 52089 52090 Just rename stuff; nothing fancy here. 52091 52092 Signed-off-by: Keith Packard <keithp@keithp.com> 52093 Reviewed-by: Eric Anholt <eric@anholt.net> 52094 52095commit 1f3676a81ef2c52699bdfcf1fb9da7e4e79b00b7 52096Author: Keith Packard <keithp@keithp.com> 52097Date: Sun Dec 15 01:55:43 2013 -0800 52098 52099 hw/vfb: Rename screenInfo parameter to screen_info 52100 52101 Avoid shadowing the global screenInfo value. 52102 52103 Signed-off-by: Keith Packard <keithp@keithp.com> 52104 Reviewed-by: Eric Anholt <eric@anholt.net> 52105 52106commit 89c0fa3d6c8ae67084a30395085db6d872d9d6de 52107Author: Keith Packard <keithp@keithp.com> 52108Date: Sun Dec 15 01:55:06 2013 -0800 52109 52110 kdrive: Fix -Wshadow errors 52111 52112 Just rename stuff to avoid conflicts. 52113 52114 Signed-off-by: Keith Packard <keithp@keithp.com> 52115 Reviewed-by: Eric Anholt <eric@anholt.net> 52116 52117commit 9c108b3ccdb4086002b83d9ad66f5619e9ec95bb 52118Author: Keith Packard <keithp@keithp.com> 52119Date: Sun Dec 15 01:53:50 2013 -0800 52120 52121 exa: Fix -Wshadow warnings 52122 52123 In exa_accel, there was a duplicate fetch of a pixmap private field. 52124 52125 exa_render just had a regular shadowed value. 52126 52127 Signed-off-by: Keith Packard <keithp@keithp.com> 52128 Reviewed-by: Eric Anholt <eric@anholt.net> 52129 52130commit 04e320a4e4e1f46bf95e36078d93b4d18a0ef855 52131Author: Keith Packard <keithp@keithp.com> 52132Date: Sun Dec 15 01:52:58 2013 -0800 52133 52134 composite: Remove duplicate window pixmap fetch 52135 52136 There's nothing between the previous fetch and this one that could 52137 cause the window pixmap to have changed. 52138 52139 Signed-off-by: Keith Packard <keithp@keithp.com> 52140 Reviewed-by: Eric Anholt <eric@anholt.net> 52141 52142commit 07b03e721e4dbec1a640c9b3bb8a2d5902587e63 52143Author: Keith Packard <keithp@keithp.com> 52144Date: Sun Dec 15 01:41:19 2013 -0800 52145 52146 xfree86: Fix -Wshadow warnings 52147 52148 Just rename variables to eliminate -Wshadow warnings. 52149 52150 Signed-off-by: Keith Packard <keithp@keithp.com> 52151 Reviewed-by: Eric Anholt <eric@anholt.net> 52152 52153commit 3c94b89f149c2d63424faed0d44a6911393c3569 52154Author: Keith Packard <keithp@keithp.com> 52155Date: Sun Dec 15 01:40:04 2013 -0800 52156 52157 Xi: Use 'void *' instead of 'Pointer' (which is going away) 52158 52159 There's no reason for XI to declare 'typedef char *Pointer' in a 52160 shared header file; assume it will eventually go away and stop using 52161 it here. 52162 52163 Signed-off-by: Keith Packard <keithp@keithp.com> 52164 Reviewed-by: Eric Anholt <eric@anholt.net> 52165 52166commit 08d0481e299c28b64a0db9bb0782ba2b551028fd 52167Author: Keith Packard <keithp@keithp.com> 52168Date: Sun Dec 15 01:12:22 2013 -0800 52169 52170 os: Fix -Wshadow errors 52171 52172 Rename variables to avoid shadowing globals 52173 52174 Signed-off-by: Keith Packard <keithp@keithp.com> 52175 Reviewed-by: Eric Anholt <eric@anholt.net> 52176 52177commit 1487145b41390e689f8b03c1e8a2df70e00cf519 52178Author: Keith Packard <keithp@keithp.com> 52179Date: Fri Dec 13 11:25:45 2013 -0800 52180 52181 Add 'FONT_PATH_NAME_ELEMENT_CONST' to dix-config.h and xorg-server.h 52182 52183 This signals to the fontsproto code that the X server has been fixed 52184 to allow the name member in a FontPathElement struct to be declared 52185 const to eliminate piles of warnings when assigning string constants 52186 to them. 52187 52188 Signed-off-by: Keith Packard <keithp@keithp.com> 52189 Reviewed-by: Eric Anholt <eric@anholt.net> 52190 52191commit 1ad8d12e7fcd7d55a37099d7426e3290297ba6d3 52192Author: Keith Packard <keithp@keithp.com> 52193Date: Fri Dec 13 11:24:00 2013 -0800 52194 52195 Ignore a couple of format-nonliteral warnings 52196 52197 These are generated in code which uses sprintf as a convenient way to 52198 construct strings from various pieces. 52199 52200 Signed-off-by: Keith Packard <keithp@keithp.com> 52201 Reviewed-by: Eric Anholt <eric@anholt.net> 52202 52203commit 9ef53e2267ea3dd60651961f66c65e8b00b7d823 52204Author: Keith Packard <keithp@keithp.com> 52205Date: Wed Dec 11 11:55:37 2013 -0800 52206 52207 include: GetClientResolutions is declared in font.h 52208 52209 No need to have a duplicate definition here. 52210 52211 Signed-off-by: Keith Packard <keithp@keithp.com> 52212 Reviewed-by: Eric Anholt <eric@anholt.net> 52213 52214commit 25ebb9dbc9df659dec2bf6c27654a5bad2d11f94 52215Merge: 409e8e29f 71baa466b 52216Author: Keith Packard <keithp@keithp.com> 52217Date: Wed Jan 22 11:33:53 2014 -0800 52218 52219 Merge remote-tracking branch 'whot/for-keith' 52220 52221commit 409e8e29fbe16122ba5a4249256fc56e2e68ea93 52222Merge: 457bc8354 d6c8d7509 52223Author: Keith Packard <keithp@keithp.com> 52224Date: Wed Jan 22 11:32:35 2014 -0800 52225 52226 Merge remote-tracking branch 'dlespiau/20131216-4k' 52227 52228commit 457bc83549e58bb87de96bed02988db3275a7611 52229Author: Alan Coopersmith <alan.coopersmith@oracle.com> 52230Date: Sat Jan 4 00:09:45 2014 -0800 52231 52232 If EAGAIN == EWOULDBLOCK, only need to check errno for one of them 52233 52234 Solaris <sys/errno.h> has: 52235 #define EWOULDBLOCK EAGAIN 52236 so checking (errno == EAGAIN || errno == EWOULDBLOCK) is overkill. 52237 52238 This leads cppcheck 1.62 to complain: 52239 [xserver/os/io.c:365] -> [xserver/os/io.c:365]: (style) Same expression on both sides of '||'. 52240 [xserver/os/io.c:941] -> [xserver/os/io.c:941]: (style) Same expression on both sides of '||'. 52241 52242 This quiets it, and reduces the number of calls Solaris Studio cc 52243 generates to the __errno() function to get the thread-specific errno value. 52244 52245 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 52246 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 52247 Signed-off-by: Keith Packard <keithp@keithp.com> 52248 52249commit 044a6ef2ff5f2c83a80b11e747f62f48830b526b 52250Merge: 0b932cf47 77df653ae 52251Author: Keith Packard <keithp@keithp.com> 52252Date: Wed Jan 22 11:28:54 2014 -0800 52253 52254 Merge remote-tracking branch 'jeremyhu/master' 52255 52256commit 0b932cf47a4df8ea0f7488e285d218fe7c10dd77 52257Merge: 1d76b02fa 295d41fa2 52258Author: Keith Packard <keithp@keithp.com> 52259Date: Wed Jan 22 11:27:36 2014 -0800 52260 52261 Merge remote-tracking branch 'anholt/xserver-unifdef' 52262 52263commit 1d76b02fac79c0360ae201e4d1a8ba0e9a00e810 52264Merge: 771f390ef 4dd62d780 52265Author: Keith Packard <keithp@keithp.com> 52266Date: Wed Jan 22 11:22:35 2014 -0800 52267 52268 Merge remote-tracking branch 'anholt/glamor-external-rebase' 52269 52270commit 771f390efdf48cb7c44fe20857f06f8ffff3b2ce 52271Author: Gaetan Nadon <memsize@videotron.ca> 52272Date: Sat Dec 21 19:58:32 2013 -0500 52273 52274 autoconf: fix warning by replacing AC_WARN with AC_MSG_WARN 52275 52276 A warning about the macro generating the warnings to the user... 52277 52278 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 52279 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 52280 Signed-off-by: Keith Packard <keithp@keithp.com> 52281 52282commit c3819da87ff2e8c6292066f965c098704edc442c 52283Author: Gaetan Nadon <memsize@videotron.ca> 52284Date: Sat Dec 21 19:58:31 2013 -0500 52285 52286 xorg-tls: fix warning, replace AC_TRY_COMPILE with AC_COMPILE_IFELSE 52287 52288 The code produced in the configure script is identical. 52289 52290 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 52291 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 52292 Signed-off-by: Keith Packard <keithp@keithp.com> 52293 52294commit 02d866a088a2fc60c6aaebe23c48b372f6b6aa9b 52295Author: Gaetan Nadon <memsize@videotron.ca> 52296Date: Sat Dec 21 19:58:30 2013 -0500 52297 52298 autoconf: fix warning, replace AC_TRY_COMPILE with AC_COMPILE_IFELSE 52299 52300 The code produced in the configure script is identical. 52301 52302 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 52303 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 52304 Signed-off-by: Keith Packard <keithp@keithp.com> 52305 52306commit e38c98fe74394a138ebadb29b1cfe1d4ea7f7071 52307Author: Gaetan Nadon <memsize@videotron.ca> 52308Date: Sat Dec 21 19:58:29 2013 -0500 52309 52310 autoconf: fix warning, replace AC_TRY_LINK with AC_LINK_IFELSE 52311 52312 The code produced in the configure script is identical. 52313 52314 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 52315 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 52316 Signed-off-by: Keith Packard <keithp@keithp.com> 52317 52318commit 4993590c90b1289171a2ca0e74d9c4310f5318c7 52319Author: Gaetan Nadon <memsize@videotron.ca> 52320Date: Sat Dec 21 19:58:28 2013 -0500 52321 52322 autoconf: fix warning by replacing deprecated AC_HELP_STRING 52323 52324 We can also make do without the workaround introduced in 2005. 52325 The 2.60 autoconf minimum version covers that now. 52326 52327 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 52328 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 52329 Signed-off-by: Keith Packard <keithp@keithp.com> 52330 52331commit 1d4dbc2f2b4d2acca2691c3c7464a45aac59d73f 52332Author: Gaetan Nadon <memsize@videotron.ca> 52333Date: Wed Dec 4 13:28:04 2013 -0500 52334 52335 libtool: bump minimum required version to 2.2 52336 52337 LT_PREREQ([2.2]) 52338 LT_INIT 52339 52340 AC_CONFIG_MACRO_DIR([m4]) is recommended to locate local macros, 52341 which is fully used starting automake 1.14. 52342 52343 ACLOCAL_AMFLAGS in Makefile.am is deprecated since automake 1.14. 52344 The comment is to prevent developers from removing the statement 52345 as suggested in 1.14 which would break earlier versions. 52346 52347 Automake 1.14 is already in use by many. 52348 This patch works on versions prior and post 1.14 52349 52350 References: 52351 Building the X Window System from Source: 52352 http://www.x.org/wiki/Building_the_X_Window_System/#index2h3 52353 52354 Discussion on xorg minimum autotools requirements: 52355 http://lists.x.org/archives/xorg-devel/2013-October/038325.html 52356 52357 Tested with autoconf 2.60, automake 1.10.3 and libtool 2.2 52358 52359 The same upgrade will be applied to libraries and drivers. 52360 52361 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 52362 Signed-off-by: Keith Packard <keithp@keithp.com> 52363 52364commit 71baa466b1f6b02fe503f9a3089b7b9d61aa0f80 52365Author: Peter Hutterer <peter.hutterer@who-t.net> 52366Date: Mon Jan 13 17:00:23 2014 +1000 52367 52368 os: restrict display names to digits 52369 52370 We call atoi() on the server's display to get the socket but otherwise use the 52371 unmodified display for log file name, xkb paths, etc. This results in 52372 Xorg :banana being the equivalent of Xorg :0, except for the log files being 52373 in /var/log/Xorg.banana.log. I'm not sure there's a good use-case for this 52374 behaviour. 52375 52376 Check the display for something that looks reasonable, i.e. digits only, but 52377 do allow for :0.0 (i.e. digits, followed by a period, followed by one or two 52378 digits). 52379 52380 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 52381 Reviewed-by: Keith Packard <keithp@keithp.com> 52382 52383commit 77df653ae3d8448be21221711851acde12c6bc1a 52384Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52385Date: Wed Jan 1 11:10:41 2014 -0800 52386 52387 XQuartz: Avoid passing uninitialized pointers to X11ApplicationSetWindowMenu in AppleWMSetWindowMenu 52388 52389 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52390 52391commit 3bc608a361a01043b226fb9aaebf88f6fd852925 52392Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52393Date: Wed Jan 1 11:04:07 2014 -0800 52394 52395 XQuartz: Check for allocated memory before using it in AppleWMSetWindowMenu 52396 52397 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52398 52399commit 64327226ddfba8f0653615cd678d2d4336fb993d 52400Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52401Date: Wed Jan 1 11:00:40 2014 -0800 52402 52403 XQuartz: Silence a clang static analysis warning about a memory leak 52404 52405 It seems the alanyzer can't comprehend dixSetPrivate(). 52406 52407 quartz.c:119:12: warning: Potential leak of memory pointed to by 'displayInfo' 52408 return quartzProcs->AddScreen(index, pScreen); 52409 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 52410 52411 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52412 52413commit b2f6b3497c33a4897afae80a2cf69c596b9f81e8 52414Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52415Date: Wed Jan 1 10:55:10 2014 -0800 52416 52417 XQuartz: Silence a clang static analysis warning about a possible memory leak on exit 52418 52419 stub.c:356:9: warning: Potential leak of memory pointed to by 'newargv' 52420 asl_log(aslc, NULL, ASL_LEVEL_ERR, 52421 ^~~~~~~ 52422 stub.c:356:9: warning: Potential leak of memory pointed to by 'newenvp' 52423 asl_log(aslc, NULL, ASL_LEVEL_ERR, 52424 ^~~~~~~ 52425 2 warnings generated. 52426 52427 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52428 52429commit a03f096a85537d9e881cedaa6cb71aca43a97086 52430Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52431Date: Wed Jan 1 10:47:52 2014 -0800 52432 52433 XQuartz: Validate length in appledri before swapping 52434 52435 Avoids potential memory corruption from bad requests 52436 52437 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52438 52439commit b3572c0d1ab7888ac26d6b2b8be6d1d19ed9af3f 52440Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52441Date: Wed Jan 1 10:39:56 2014 -0800 52442 52443 XQuartz: Validate screen in AppleDRIQueryDirectRenderingCapable requests 52444 52445 Return an error to the caller rather than crashing the server on 52446 invalid screens. 52447 52448 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52449 52450commit 959e8f23af7850fcaf40d6c67f5228241a36a9ab 52451Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52452Date: Sun Dec 29 12:45:23 2013 -0800 52453 52454 XQuartz: Simplify hook_run to quiet static analyzer 52455 52456 x-hook.c:96:9: warning: Called function pointer is an uninitalized pointer value 52457 (*fun[i])(arg, data[i]); 52458 ^~~~~~~~~~~~~~~~~~~~~~~ 52459 1 warning generated. 52460 52461 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52462 52463commit f79af1941776fd6f1ec26c50603fcc35ca7d514b 52464Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52465Date: Sun Dec 29 12:41:18 2013 -0800 52466 52467 XQuartz: Mark applicationWillTerminate: noreturn 52468 52469 X11Controller.m:938:1: warning: method 'applicationWillTerminate:' could be declared with attribute 'noreturn' 52470 [-Wmissing-noreturn,Semantic Issue] 52471 { 52472 ^ 52473 52474 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52475 52476commit ea80279e292e59a9fe9651489f03e9f2f39810d9 52477Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52478Date: Sun Dec 29 12:36:51 2013 -0800 52479 52480 XQuartz: Fix get_proc_address signature 52481 52482 indirect.c:675:28: warning: incompatible pointer types passing 'glx_gpa_proc (*)(const char *)' to parameter of type 52483 'glx_gpa_proc' (aka 'glx_func_ptr (*)(const char *)') [-Wincompatible-pointer-types,Semantic Issue] 52484 __glXsetGetProcAddress(&get_proc_address); 52485 ^~~~~~~~~~~~~~~~~ 52486 ../../../glx/glxserver.h:122:42: note: passing argument to parameter 'get_proc_address' here [Semantic Issue] 52487 void __glXsetGetProcAddress(glx_gpa_proc get_proc_address); 52488 ^ 52489 52490 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52491 52492commit 2e3ebec9520719a8e5c3c92390e83bcb5216f978 52493Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52494Date: Sun Dec 29 12:31:23 2013 -0800 52495 52496 XQuartz: Fix darwinfb.h header guard 52497 52498 ./darwinfb.h:28:9: warning: '_DARWIN_FB_H' is used as a header guard here, followed by #define of a different macro 52499 [-Wheader-guard,Lexical or Preprocessor Issue] 52500 ^~~~~~~~~~~~ 52501 ./darwinfb.h:29:9: note: '_DARWIN_DB_H' is defined here; did you mean '_DARWIN_FB_H'? [Lexical or Preprocessor Issue] 52502 ^~~~~~~~~~~~ 52503 _DARWIN_FB_H 52504 52505 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52506 52507commit 9da6c0918f40359f28fe8889d5b7cae7efcc8377 52508Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52509Date: Sun Dec 29 12:22:55 2013 -0800 52510 52511 XQuartz: Silence some static analyzer warnings by annotating referencing counts 52512 52513 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 52514 52515commit 2d2d49dab5c5718989de97d7227aac793479745e 52516Author: Keith Packard <keithp@keithp.com> 52517Date: Sun Dec 29 15:18:28 2013 -0800 52518 52519 Clean up a few function prototypes to not place formals in /**/ 52520 52521 This just removes the comment markers from around the formals in 52522 several function prototypes near where pointer -> void * changes were 52523 made. There are plenty more of these to fix. 52524 52525 Signed-off-by: Keith Packard <keithp@keithp.com> 52526 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 52527 52528commit 60014a4a98ff924ae7f6840781f768c1cc93bbab 52529Author: Keith Packard <keithp@keithp.com> 52530Date: Sun Dec 15 01:05:51 2013 -0800 52531 52532 Replace 'pointer' type with 'void *' 52533 52534 This lets us stop using the 'pointer' typedef in Xdefs.h as 'pointer' 52535 is used throughout the X server for other things, and having duplicate 52536 names generates compiler warnings. 52537 52538 Signed-off-by: Keith Packard <keithp@keithp.com> 52539 Reviewed-by: Eric Anholt <eric@anholt.net> 52540 52541commit 93fa64e17d7bd600ebf18ecab85f5b2d17fe30ce 52542Author: Keith Packard <keithp@keithp.com> 52543Date: Wed Dec 11 12:18:49 2013 -0800 52544 52545 Don't build dmx by default 52546 52547 It's clearly un-loved, having piles and piles of warnings. If someone 52548 wants to fix it up to compile without warnings, we can re-enable it. 52549 52550 Signed-off-by: Keith Packard <keithp@keithp.com> 52551 Reviewed-by: Adam Jackson <ajax@redhat.com> 52552 52553commit 319dff750f99a28a1544bcb81d52172c641ad1ca 52554Author: Keith Packard <keithp@keithp.com> 52555Date: Wed Dec 11 12:18:16 2013 -0800 52556 52557 hw/xfree86: nobus.c shouldn't define a static function 52558 52559 Having this function be static generates a compiler warning. 52560 52561 Signed-off-by: Keith Packard <keithp@keithp.com> 52562 Reviewed-by: Adam Jackson <ajax@redhat.com> 52563 52564commit 7915791bac3ff609e94f355fbb637d7c570a5b32 52565Author: Keith Packard <keithp@keithp.com> 52566Date: Wed Dec 11 12:17:45 2013 -0800 52567 52568 kdrive/ephyr; Don't redeclare monitorResolution 52569 52570 It's already declared in globals.h 52571 52572 Signed-off-by: Keith Packard <keithp@keithp.com> 52573 Reviewed-by: Adam Jackson <ajax@redhat.com> 52574 52575commit aa47a4409660634d80c3423ad49b70ca929a8756 52576Author: Keith Packard <keithp@keithp.com> 52577Date: Wed Dec 11 12:16:51 2013 -0800 52578 52579 kdrive/ephyr: ddxUseMsg is supposed to return 52580 52581 I'm not sure why ephyr thinks that ddxUseMsg shouldn't return, but 52582 it's not declared to exit. 52583 52584 Signed-off-by: Keith Packard <keithp@keithp.com> 52585 Reviewed-by: Adam Jackson <ajax@redhat.com> 52586 52587commit 86d68825c2b75ea9e10954fec00b44b33b38d8b2 52588Author: Keith Packard <keithp@keithp.com> 52589Date: Wed Dec 11 11:55:01 2013 -0800 52590 52591 vfb: Remove unused pXWDCmap variable 52592 52593 Signed-off-by: Keith Packard <keithp@keithp.com> 52594 Reviewed-by: Adam Jackson <ajax@redhat.com> 52595 52596commit 7353ec7cb6fc235b03e59e35425201429c83ee72 52597Author: Keith Packard <keithp@keithp.com> 52598Date: Wed Dec 11 11:32:19 2013 -0800 52599 52600 xfree86: Switch int10 code to stdint types 52601 52602 CARD32 is not type compatible with uint32_t and ends up generating a 52603 pile of warnings. Fix this by replacing all of the CARD* types with 52604 stdint types. 52605 52606 Signed-off-by: Keith Packard <keithp@keithp.com> 52607 Reviewed-by: Adam Jackson <ajax@redhat.com> 52608 52609commit 93b15b1a47eb414338e8b229f25f11bdca099471 52610Author: Keith Packard <keithp@keithp.com> 52611Date: Wed Dec 11 11:31:46 2013 -0800 52612 52613 xfree86: Don't complain when the SDK dependency file doesn't exist yet 52614 52615 It won't exist until the build is complete, so don't complain about it 52616 52617 Signed-off-by: Keith Packard <keithp@keithp.com> 52618 Reviewed-by: Adam Jackson <ajax@redhat.com> 52619 52620commit eda935627116dd106b99dbcc95d90a723a2de330 52621Author: Keith Packard <keithp@keithp.com> 52622Date: Sat Nov 16 23:35:21 2013 -0800 52623 52624 xkb: Make XkbWriteCountedString take a const char * input parameter 52625 52626 Signed-off-by: Keith Packard <keithp@keithp.com> 52627 Reviewed-by: Adam Jackson <ajax@redhat.com> 52628 52629commit d0339a5c66846c9f14e3b584e34688520a0916ab 52630Author: Keith Packard <keithp@keithp.com> 52631Date: Sat Nov 16 23:34:59 2013 -0800 52632 52633 os: xstrtokenize takes and returns const char * now 52634 52635 Signed-off-by: Keith Packard <keithp@keithp.com> 52636 Reviewed-by: Adam Jackson <ajax@redhat.com> 52637 52638commit c608560dbbac18837cb00ef0d774a66ea7706534 52639Author: Keith Packard <keithp@keithp.com> 52640Date: Sat Nov 16 23:33:59 2013 -0800 52641 52642 xfree86/vbe: Make VBEValidateModes take const char ** 52643 52644 mode names are now const 52645 52646 Signed-off-by: Keith Packard <keithp@keithp.com> 52647 Reviewed-by: Adam Jackson <ajax@redhat.com> 52648 52649commit 3835e8b0a1cc7b1e971bcc3f9453422fec014b10 52650Author: Keith Packard <keithp@keithp.com> 52651Date: Sat Nov 16 23:33:18 2013 -0800 52652 52653 xfree86/shadowfb: GCFuncs and GCOps are now const 52654 52655 Change GC private to match and fix resulting warnings 52656 52657 Signed-off-by: Keith Packard <keithp@keithp.com> 52658 Reviewed-by: Adam Jackson <ajax@redhat.com> 52659 52660commit f71de60355cc76810657f40c7b5461af86b34bf7 52661Author: Keith Packard <keithp@keithp.com> 52662Date: Sat Nov 16 23:32:40 2013 -0800 52663 52664 xfree86/parser: make strings in xf86MatchGroup const 52665 52666 and fix resulting warnings 52667 52668 Signed-off-by: Keith Packard <keithp@keithp.com> 52669 Reviewed-by: Adam Jackson <ajax@redhat.com> 52670 52671commit 2a93e75ff81b095bc9cdb80906e23ec705d52df5 52672Author: Keith Packard <keithp@keithp.com> 52673Date: Sat Nov 16 23:31:55 2013 -0800 52674 52675 xfree86/int10: mark printk as _X_ATTRIBUTE_PRINTF 52676 52677 Signed-off-by: Keith Packard <keithp@keithp.com> 52678 Reviewed-by: Adam Jackson <ajax@redhat.com> 52679 52680commit 7fe436a7b786ff02ed55cf9ed65edfd2fe07ac22 52681Author: Keith Packard <keithp@keithp.com> 52682Date: Sat Nov 16 23:31:06 2013 -0800 52683 52684 xfree86/fbdevhw: Fix warnings 52685 52686 Unused fPtr variable. Deal with string constants. 52687 52688 Signed-off-by: Keith Packard <keithp@keithp.com> 52689 Reviewed-by: Adam Jackson <ajax@redhat.com> 52690 52691commit 6990de00eb23478d9343660ad3981fb1d33f8be4 52692Author: Keith Packard <keithp@keithp.com> 52693Date: Sat Nov 16 23:30:29 2013 -0800 52694 52695 xfree86/exa: xf86GetOptValString returns const char * now 52696 52697 fix exaDDXDriverInit to match. 52698 52699 Signed-off-by: Keith Packard <keithp@keithp.com> 52700 Reviewed-by: Adam Jackson <ajax@redhat.com> 52701 52702commit 017307f0b41476c3c73a9414a7ce9eb42373c3f4 52703Author: Keith Packard <keithp@keithp.com> 52704Date: Sat Nov 16 23:29:30 2013 -0800 52705 52706 xfree86/dri: Mark DRIDrvMsg and dri_drm_debug_print _X_ATTRIBUTE_PRINTF 52707 52708 And fix resulting warnings. 52709 52710 v2: (Adam Jackson) Cast handles through uintptr_t to avoid size change warnings 52711 52712 Signed-off-by: Keith Packard <keithp@keithp.com> 52713 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 52714 Reviewed-by: Adam Jackson <ajax@redhat.com> 52715 52716commit 22592855e90d23013ba7f9e945d567725cb44bf3 52717Author: Keith Packard <keithp@keithp.com> 52718Date: Sat Nov 16 23:28:48 2013 -0800 52719 52720 xfree86/common: handle string constants in xf86Xinput configuration 52721 52722 Signed-off-by: Keith Packard <keithp@keithp.com> 52723 Reviewed-by: Adam Jackson <ajax@redhat.com> 52724 52725commit 3a163d2af48a7cf8fd5c2db6ac68166a5fdbeb8d 52726Author: Keith Packard <keithp@keithp.com> 52727Date: Sat Nov 16 23:28:16 2013 -0800 52728 52729 xfree86/common: Const GC funcs and ops in xf86VAarbiter 52730 52731 Signed-off-by: Keith Packard <keithp@keithp.com> 52732 Reviewed-by: Adam Jackson <ajax@redhat.com> 52733 52734commit 8a9aa44a45377d8953d11a0b035c86eb139283ba 52735Author: Keith Packard <keithp@keithp.com> 52736Date: Sat Nov 16 23:26:19 2013 -0800 52737 52738 xfree86/config: Kludge around const strings 52739 52740 defaultFontPath is now a const char * so that it can be initialized 52741 from a string constant. This patch kludges around that by inserting 52742 suitable casts to eliminate warnings. Fixing this 'correctly' would 52743 involve inserting some new variables and conditionals to use them. 52744 52745 Signed-off-by: Keith Packard <keithp@keithp.com> 52746 Reviewed-by: Adam Jackson <ajax@redhat.com> 52747 52748commit 86647e7279163b13492d39ecb9c44414b752a61e 52749Author: Keith Packard <keithp@keithp.com> 52750Date: Sat Nov 16 23:25:50 2013 -0800 52751 52752 config/udev: handle const strings 52753 52754 Signed-off-by: Keith Packard <keithp@keithp.com> 52755 Reviewed-by: Adam Jackson <ajax@redhat.com> 52756 52757commit b7633c6ff2c273621df6c0fac9c269f8ab217a46 52758Author: Keith Packard <keithp@keithp.com> 52759Date: Sat Nov 16 23:24:03 2013 -0800 52760 52761 composite: Remove unused pScreen variables 52762 52763 Signed-off-by: Keith Packard <keithp@keithp.com> 52764 Reviewed-by: Adam Jackson <ajax@redhat.com> 52765 52766commit 5bc5684d4cb4905cd82fe1c036cc8a48c3c4868f 52767Author: Keith Packard <keithp@keithp.com> 52768Date: Fri Nov 15 06:02:05 2013 -0800 52769 52770 test: Warning cleanup 52771 52772 const char in test/xfree86.c. Cast values to (intmax_t) for %ju format 52773 in test/signal-logging.c. 52774 52775 Signed-off-by: Keith Packard <keithp@keithp.com> 52776 Reviewed-by: Adam Jackson <ajax@redhat.com> 52777 52778commit fecc7eb1cf66db64728ee2d68cd9443df7e70879 52779Author: Keith Packard <keithp@keithp.com> 52780Date: Fri Nov 15 06:00:44 2013 -0800 52781 52782 xi: More warning cleanup for input 52783 52784 Lots more const char stuff. 52785 52786 Remove duplicate defs of CoreKeyboardProc and CorePointerProc from 52787 test/xi2/protocol-common.c 52788 52789 Signed-off-by: Keith Packard <keithp@keithp.com> 52790 Reviewed-by: Adam Jackson <ajax@redhat.com> 52791 52792commit af04cf6968b16cc9efd17905471047e7de62058a 52793Author: Keith Packard <keithp@keithp.com> 52794Date: Fri Nov 15 05:55:38 2013 -0800 52795 52796 Xext: Clean up warnings in hashtable code 52797 52798 Make keys const void * 52799 52800 Signed-off-by: Keith Packard <keithp@keithp.com> 52801 Reviewed-by: Adam Jackson <ajax@redhat.com> 52802 52803commit 00438c9f943b219ba33055969ae7f9ba07214b2a 52804Author: Keith Packard <keithp@keithp.com> 52805Date: Fri Nov 15 05:41:54 2013 -0800 52806 52807 exa: Clean up warnings 52808 52809 Declare GC ops/funcs as const. 52810 Use 'typeof' in the 'swap' macro to capture the right type. 52811 52812 Signed-off-by: Keith Packard <keithp@keithp.com> 52813 Reviewed-by: Adam Jackson <ajax@redhat.com> 52814 52815commit 6e51645b4796fc3a02206fefd416b84beb2fcdf7 52816Author: Keith Packard <keithp@keithp.com> 52817Date: Fri Nov 15 05:28:47 2013 -0800 52818 52819 xkb: Clean up warnings 52820 52821 Add const to lots of strings. 52822 52823 Signed-off-by: Keith Packard <keithp@keithp.com> 52824 Reviewed-by: Adam Jackson <ajax@redhat.com> 52825 52826commit abce3206cbc82f632abae5344e0ce46622e00f24 52827Author: Keith Packard <keithp@keithp.com> 52828Date: Fri Nov 15 05:19:24 2013 -0800 52829 52830 os: Clean up warnings 52831 52832 Just const char stuff. 52833 52834 Signed-off-by: Keith Packard <keithp@keithp.com> 52835 Reviewed-by: Adam Jackson <ajax@redhat.com> 52836 52837commit a1cb69dc28fdbfbdfaf954e0bec221f759462399 52838Author: Keith Packard <keithp@keithp.com> 52839Date: Fri Nov 15 22:05:44 2013 +0900 52840 52841 Xext: Clean up warnings 52842 52843 GC funcs and ops are const. 52844 Remove unused variables. 52845 52846 Signed-off-by: Keith Packard <keithp@keithp.com> 52847 Reviewed-by: Adam Jackson <ajax@redhat.com> 52848 52849commit 2f6e8eb70d527541178433933d6230466421bd15 52850Author: Keith Packard <keithp@keithp.com> 52851Date: Fri Nov 15 22:02:58 2013 +0900 52852 52853 damage: Clean up warnings 52854 52855 GC funcs and ops are const now, so all wrappers need to declare them 52856 as such. 52857 52858 Signed-off-by: Keith Packard <keithp@keithp.com> 52859 Reviewed-by: Adam Jackson <ajax@redhat.com> 52860 52861commit c706fb0db86d6946482700d65ad6803c4daaa6f9 52862Author: Keith Packard <keithp@keithp.com> 52863Date: Fri Nov 15 22:00:29 2013 +0900 52864 52865 Clean up warnings in mi. 52866 52867 A coupel of unused variables, and some debug code with mis-matching 52868 printf format and variable types. 52869 52870 Signed-off-by: Keith Packard <keithp@keithp.com> 52871 Reviewed-by: Adam Jackson <ajax@redhat.com> 52872 52873commit d89b42bda46d36fc0879611cc3b3566957ce36d0 52874Author: Keith Packard <keithp@keithp.com> 52875Date: Fri Nov 15 21:57:05 2013 +0900 52876 52877 Clean up warnings in DIX 52878 52879 As usual, mostly const char changes. However, filter_device_events had 52880 a potentially uninitialized value, 'raw', which I added a bunch of 52881 checks for. I suspect most of those are 'can't happen', but it's hard 52882 to see that inside the function. 52883 52884 Signed-off-by: Keith Packard <keithp@keithp.com> 52885 Reviewed-by: Adam Jackson <ajax@redhat.com> 52886 52887commit d6da9f23cca562fc0c6ae398665ab7fa770a4fa8 52888Author: Keith Packard <keithp@keithp.com> 52889Date: Fri Nov 15 18:07:54 2013 +0900 52890 52891 hw/xfree86: More const declarations for strings 52892 52893 Signed-off-by: Keith Packard <keithp@keithp.com> 52894 Reviewed-by: Adam Jackson <ajax@redhat.com> 52895 52896commit 644725ac5e28a00d8c24372f55ae7e7e2fe0cb2e 52897Author: Keith Packard <keithp@keithp.com> 52898Date: Fri Nov 15 18:06:37 2013 +0900 52899 52900 Just remove dpms functsion from xf86.h 52901 52902 Signed-off-by: Keith Packard <keithp@keithp.com> 52903 Reviewed-by: Adam Jackson <ajax@redhat.com> 52904 52905commit c85e26d59901fbf8c7a437cf934b51fa64160073 52906Author: Keith Packard <keithp@keithp.com> 52907Date: Fri Nov 15 17:55:15 2013 +0900 52908 52909 Bunch of DMX warning fixes 52910 52911 Signed-off-by: Keith Packard <keithp@keithp.com> 52912 Reviewed-by: Adam Jackson <ajax@redhat.com> 52913 52914commit 493d992501b586d95823e045d1dc994bd6c00d27 52915Author: Keith Packard <keithp@keithp.com> 52916Date: Fri Nov 15 17:43:46 2013 +0900 52917 52918 More warning fixes in hw/xfree86 52919 52920 Signed-off-by: Keith Packard <keithp@keithp.com> 52921 Reviewed-by: Adam Jackson <ajax@redhat.com> 52922 52923commit c78be3a4b714deb7ad75cacd54042ca1b51d6261 52924Author: Keith Packard <keithp@keithp.com> 52925Date: Fri Nov 15 17:36:26 2013 +0900 52926 52927 xfree86 warning reduction 52928 52929 This gets the easy warnings, mostly constant string problems. 52930 52931 Signed-off-by: Keith Packard <keithp@keithp.com> 52932 Reviewed-by: Adam Jackson <ajax@redhat.com> 52933 52934commit 1ce15ed5bae60a184c6bd9440e3c017488a5f415 52935Author: Keith Packard <keithp@keithp.com> 52936Date: Sat Nov 16 23:44:51 2013 -0800 52937 52938 Xext: _X_EXPORT DPMSSet and DPMSSupported 52939 52940 These are needed by drivers, and it's better to export them from here 52941 rather than redefining them in hw/xfree86 and exporting them from there. 52942 52943 Signed-off-by: Keith Packard <keithp@keithp.com> 52944 Reviewed-by: Adam Jackson <ajax@redhat.com> 52945 52946commit 1f407763be28745f18d224077d6b07b9431ee16b 52947Author: Keith Packard <keithp@keithp.com> 52948Date: Sat Nov 16 23:43:49 2013 -0800 52949 52950 include: Make xkbrules structures all const char * 52951 52952 This lets them be initialized with string constants 52953 52954 Signed-off-by: Keith Packard <keithp@keithp.com> 52955 Reviewed-by: Adam Jackson <ajax@redhat.com> 52956 52957commit e1e01d2e33c632e395d7e396f73fba8ae606b15a 52958Author: Keith Packard <keithp@keithp.com> 52959Date: Fri Nov 15 16:56:15 2013 +0900 52960 52961 xfree86/common: Warning fixes. Mostly const string handling. 52962 52963 Also removes DPMS functiosn from Xext/dpmsproc.h 52964 52965 Signed-off-by: Keith Packard <keithp@keithp.com> 52966 Reviewed-by: Adam Jackson <ajax@redhat.com> 52967 52968commit 27b44949a3d2e34ac10e801bd8a8fc2c28791e7e 52969Author: Keith Packard <keithp@keithp.com> 52970Date: Fri Nov 15 16:24:15 2013 +0900 52971 52972 hw/xfree86: Lots of constant string support 52973 52974 Make lots of string pointers 'const char' so that we can use constant 52975 strings with them without eliciting warnings. 52976 52977 Signed-off-by: Keith Packard <keithp@keithp.com> 52978 Reviewed-by: Adam Jackson <ajax@redhat.com> 52979 52980commit 6f77e2645ea36e324ccc664aae1d36464418bdea 52981Author: Keith Packard <keithp@keithp.com> 52982Date: Fri Nov 15 13:40:30 2013 +0900 52983 52984 hw/xfree86: Make strings in DriverRec and ScrnInfoRec const 52985 52986 This avoids compiler warnings when initializing with string constants. 52987 52988 Signed-off-by: Keith Packard <keithp@keithp.com> 52989 Reviewed-by: Eric Anholt <eric@anholt.net> 52990 Reviewed-by: Adam Jackson <ajax@redhat.com> 52991 52992commit e81902872176fa9848211fcd7a5eafa4f861a1b7 52993Author: Peter Hutterer <peter.hutterer@who-t.net> 52994Date: Thu Jan 9 15:29:23 2014 +1000 52995 52996 ephyr: don't allow a shift+ctrl keygrab if mod1 was enabled 52997 52998 Xephyr wants ctrl+shift to grab the window, but that conflicts with 52999 ctrl+alt+shift key combos. Remember the modifier state on key presses and 53000 releases, if mod1 is pressed, we need ctrl, shift and mod1 released 53001 before we allow a shift-ctrl grab activation. 53002 53003 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 53004 53005commit b2d5ee2e3684951b611fd2068d57cc65fd8305a3 53006Author: Carlos Garnacho <carlosg@gnome.org> 53007Date: Thu Jan 2 21:33:30 2014 +0100 53008 53009 Xi: Ensure DeviceChanged is emitted after grabs are deactivated 53010 53011 When a grab on a slave device is deactivated, the master device must 53012 be checked, just in case there were events from other devices while 53013 the slave device was stolen away by the passive grab. This may 53014 introduce misbehaviors on mismatching valuators and device features 53015 later on UpdateDeviceState(). 53016 53017 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 53018 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 53019 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 53020 53021commit 863d2ad5c02cccde9a4d1a392a7cae78d001c8a9 53022Author: Alan Coopersmith <alan.coopersmith@oracle.com> 53023Date: Mon Jan 6 17:10:40 2014 -0800 53024 53025 CheckPassiveGrabsOnWindow() needs to handle NULL return value from AllocGrab() 53026 53027 CheckPassiveGrabsOnWindow() calls AllocGrab() which can fail and return NULL. 53028 This return value is not checked, and can cause NULL pointer dereferences. 53029 53030 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 53031 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 53032 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 53033 53034commit 5493a67ec256d22a8a41597a345d8e1c54d6e335 53035Author: Alan Coopersmith <alan.coopersmith@oracle.com> 53036Date: Mon Jan 6 17:10:39 2014 -0800 53037 53038 GrabDevice() needs to handle NULL return value from AllocGrab() 53039 53040 GrabDevice() calls AllocGrab() which can fail and return NULL. 53041 This return value is not checked, and can cause NULL pointer dereferences. 53042 53043 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 53044 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 53045 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 53046 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 53047 53048commit 3a113815a0cc86d64789494e905da9778174f738 53049Author: Alan Coopersmith <alan.coopersmith@oracle.com> 53050Date: Mon Jan 6 17:10:38 2014 -0800 53051 53052 If AllocGrab() fails to set up grab, don't copy to a NULL grab 53053 53054 If either the initial calloc or the xi2mask_new fails, grab is NULL, 53055 but if a src grab is passed in, it was always being written to by 53056 CopyGrab (and if that failed, dereferenced again in teardown). 53057 53058 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 53059 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 53060 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 53061 53062commit a92c6406e7f6c0b74fb1cb9361ad904facc0f722 53063Author: Keith Packard <keithp@keithp.com> 53064Date: Thu Jan 9 11:10:11 2014 +0800 53065 53066 Bump release to 1.15.99.900 53067 53068 The merge window for 1.16 is open 53069 53070 Signed-off-by: Keith Packard <keithp@keithp.com> 53071 53072commit d6c8d7509727060b8e2358b9ed1c0e17b2ec3401 53073Author: Damien Lespiau <damien.lespiau@intel.com> 53074Date: Wed Jul 31 19:16:45 2013 +0100 53075 53076 xfree86: Use the TMDS maximum frequency to prune modes 53077 53078 Instead of only relying on the Range section, we can do better on 53079 HDMI to find out what is the max dot clock the monitor supports. The 53080 HDMI CEA vendor block adds a TMDS max freq we can use. 53081 53082 This makes X not prune 4k resolutions on HDMI. 53083 53084 v2: Replace X_INFO by X_PROBED in the message that prints the max 53085 TMDS frequency (Chris Wilson) 53086 53087 Reviewed-by: Chris Wilson <chris at chris-wilson.co.uk> 53088 Signed-off-by: Damien Lespiau <damien.lespiau@intel.com> 53089 53090commit 95c2287465138ac251bf792f354cee3626e76b44 53091Author: Damien Lespiau <damien.lespiau@intel.com> 53092Date: Wed Aug 7 15:16:21 2013 +0100 53093 53094 xfree86: Refactor xf86MonitorIsHDMI() using xf86MonitorFindHDMIBlock() 53095 53096 Reviewed-by: Chris Wilson <chris at chris-wilson.co.uk> 53097 Signed-off-by: Damien Lespiau <damien.lespiau@intel.com> 53098 53099commit a279fb3ff3f3cfc41530aff1d9ff5620279348f6 53100Author: Damien Lespiau <damien.lespiau@intel.com> 53101Date: Wed Jul 31 19:11:45 2013 +0100 53102 53103 xfree86: Add a xf86MonitorFindHDMIBlock() 53104 53105 The HDMI CEA vendor specific block has some interesting information, 53106 such as the maximum TMDS dot clock. 53107 53108 v2: Don't parse CEA blocks with invalid offsets, remove spurious 53109 brackets (Chris Wilson) 53110 53111 v3: Fix the looping through the CEA data blocks, it had a typo using the 53112 wrong variable coming from the code it was ported from. 53113 Replace x << 16 + y << 8 + z by x << 16 | y << 8 | z 53114 (Chris Wilson) 53115 53116 v4: Remove the stray ';' at the end of "if (*end == 0)". 53117 (Dominik Behr on IRC) 53118 53119 Reviewed-by: Chris Wilson <chris at chris-wilson.co.uk> 53120 Signed-off-by: Damien Lespiau <damien.lespiau@intel.com> 53121 53122commit 295d41fa2aa97b74c1b9ffd7ef4ccf52f3e97dde 53123Author: Eric Anholt <eric@anholt.net> 53124Date: Mon Dec 16 10:52:19 2013 -0800 53125 53126 glx: unifdef swrast dri_interface.h values from Mesa 7.1. 53127 53128 We can't remove all the ifdefs (__DRI_TEX_BUFFER_VERSION) because 53129 configure.ac is only checking for that version of Mesa in the absence 53130 of dri2. 53131 53132 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 53133 53134commit 4ab1a2797ba366179f8a2e621334dd45df2a5a65 53135Author: Eric Anholt <eric@anholt.net> 53136Date: Tue Nov 5 11:41:06 2013 -0800 53137 53138 glx: unifdef for DRI2 dri_interface.h things in mesa 9.2. 53139 53140 Thanks to configure.ac's check, we know that we have a new enough 53141 dri_interface.h that we don't need to conditionalize all this code. 53142 53143 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 53144 53145commit 2ea973e12f5d954211e1d10085a4c74581b43aca 53146Author: Keith Packard <keithp@keithp.com> 53147Date: Fri Dec 27 09:50:55 2013 -0800 53148 53149 Bump version to 1.15.0 53150 53151 Signed-off-by: Keith Packard <keithp@keithp.com> 53152 53153commit 2eb9915eea358f941702d3dad7434197991885c5 53154Author: Keith Packard <keithp@keithp.com> 53155Date: Thu Dec 12 15:48:08 2013 -0800 53156 53157 present: Set complete notify mode to Skip as needed 53158 53159 Skipped present pixmap calls were not setting the mode to 53160 PresentCompleteModeSkip for skipped operations. 53161 53162 Signed-off-by: Keith Packard <keithp@keithp.com> 53163 53164commit e487babd525ef1bd016ec7b283fa08cf9e6c6f4f 53165Author: Keith Packard <keithp@keithp.com> 53166Date: Thu Dec 12 14:52:35 2013 -0800 53167 53168 present: Don't abandon presents which are already queued for flip 53169 53170 Presents which are not marked 'queued' and are in the window present 53171 list are waiting for the flip event; discarding those won't work very 53172 well (it'll end up trashing displayed content for the next frame), so 53173 skip over those when looking for duplicate frame presents 53174 53175 Signed-off-by: Keith Packard <keithp@keithp.com> 53176 53177commit ca3a5b2a8f9f627bbaa9883e16512a21c2a0b81d 53178Author: Keith Packard <keithp@keithp.com> 53179Date: Thu Dec 12 14:17:40 2013 -0800 53180 53181 present: Handle PresentOptionAsync for copy mode 53182 53183 Check for Async flag and execute immediately if set, otherwise wait 53184 for the next appropriate vblank before copying. 53185 53186 Signed-off-by: Keith Packard <keithp@keithp.com> 53187 53188commit a68df147421da21528b5be2d34678383922fa352 53189Author: Keith Packard <keithp@keithp.com> 53190Date: Thu Dec 19 14:31:07 2013 -0800 53191 53192 Bump version to 1.14.99.905 (1.15 RC5) 53193 53194 Another week, another RC. This should be the last before 1.15 final 53195 53196 Signed-off-by: Keith Packard <keithp@keithp.com> 53197 53198commit 4b1ead9d3400acc3402c2480d7cc0527750c32f0 53199Merge: 4d6264614 929795d50 53200Author: Keith Packard <keithp@keithp.com> 53201Date: Thu Dec 19 14:14:59 2013 -0800 53202 53203 Merge remote-tracking branch 'whot/for-keith' 53204 53205commit 4dd62d7807b47efbc9065ae8f17f73e1ec6e9d26 53206Author: Gaetan Nadon <memsize@videotron.ca> 53207Date: Tue Dec 10 11:13:32 2013 -0500 53208 53209 libglamoregl: remove -I$(top_srcdir)/src 53210 53211 Automake always provide -I. It is at the beginning of the list of compiler 53212 options. 53213 53214 Not needed either to find glamor_egl.c source. 53215 53216 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 53217 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53218 53219commit 08c23ff8aefe80f2940ecb90adda27156084f57c 53220Author: Gaetan Nadon <memsize@videotron.ca> 53221Date: Tue Dec 10 11:13:31 2013 -0500 53222 53223 Make: remove cruft copied over from the X server makefiles 53224 53225 In toplevel makefile: 53226 nostdinc: only xserver, no other X modules 53227 aclocaldir: no local macros to install 53228 xkb_path: xserver only 53229 "Gross hack": xserver only 53230 53231 In src/makefile: 53232 SOLARIS_ASM_CFLAGS; server only 53233 XORG_INCS: undefined variable 53234 DIX_CFLAGS: undefined variable 53235 53236 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 53237 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53238 53239commit b8055bd1f6c9dea0fe8f7a786d2a1522f5f32129 53240Author: Dave Airlie <airlied@redhat.com> 53241Date: Fri Dec 13 11:33:46 2013 +1000 53242 53243 glamor: fix leak in xv code. 53244 53245 This loop needed to go one higher, not sure if this fixes the leak 53246 MrCooper was seeing on irc, but it fixes a leak. 53247 53248 Signed-off-by: Dave Airlie <airlied@redhat.com> 53249 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 53250 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53251 53252commit 752a79562eb13f59fa54b4181d65367c8488c0a5 53253Author: Gaetan Nadon <memsize@videotron.ca> 53254Date: Sun Dec 8 16:43:08 2013 -0500 53255 53256 Fix glamor_egl->egl_create_image_khr makes pointer from integer 53257 53258 This is a warning, but a real problem can occur on some system. 53259 53260 Reported-by: Fabio Pedretti <fabio.ped@libero.it> 53261 Reviewed-by: Axel Davy <davyaxel@free.fr> 53262 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 53263 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53264 53265commit 7cfd9cc23270c0246d584e167fe14bf5af4571df 53266Author: Axel Davy <axel.davy@ens.fr> 53267Date: Thu Dec 5 08:49:15 2013 +0100 53268 53269 Add DRI3 support to glamor 53270 53271 This implements some DRI3 helpers to help the DDXs using 53272 glamor to support DRI3. 53273 53274 Signed-off-by: Axel Davy <axel.davy@ens.fr> 53275 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53276 53277commit 06ba3fdfd6162a64c149ea6fda85f9f56d7f3c31 53278Author: Zhigang Gong <zhigang.gong@intel.com> 53279Date: Tue Nov 19 15:16:57 2013 +0800 53280 53281 Fixed some compilation warning/error or error checking. 53282 53283 There is one compilation error ,cast int to pointer, when built without 53284 libgbm, reported by Gaetan Nadon. 53285 And some error checking after memory allocation, reported by Seth Arnold. 53286 There are still some similar issues in the largepixmap implementation. 53287 They are relatively more complicate due to the heavy usage of RegionXXX 53288 APIs which may allocate implicitly. Will fix them in the future. 53289 53290 Signed-off-by: Zhigang Gong <zhigang.gong@intel.com> 53291 Tested-by: Gaetan Nadon <memsize@videotron.ca> 53292 53293commit a5321ea431376feca2dcd55cf04925dc492270fc 53294Author: Axel Davy <axel.davy@ens.fr> 53295Date: Mon Nov 18 22:52:22 2013 +0100 53296 53297 Allow to create textured pixmaps from gbm_bo without using gem names 53298 53299 This implements glamor_egl_create_textured_pixmap_from_gbm_bo, 53300 which is similar to glamor_egl_create_textured_pixmap, except 53301 it takes a gbm_bo as argument. 53302 53303 Signed-off-by: Axel Davy <axel.davy@ens.fr> 53304 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53305 53306commit 403004fe799f293bf299fe78da821635b19d9a1a 53307Author: Fabio Pedretti <fabio.ped@libero.it> 53308Date: Mon Nov 4 12:08:42 2013 +0100 53309 53310 glamor: remove unused variable 53311 53312 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53313 53314commit 842cd7eb4353492bc9c29439f975c3bd33445cbd 53315Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 53316Date: Wed Nov 6 10:25:27 2013 +0100 53317 53318 fixup picture in SetWindowPixmap 53319 53320 When creating a window with recordmydesktop running, the following may happen: 53321 53322 create picture 0x1cd457e0, with drawable 0x1327d1f0 53323 (SetWindowPixmap is called) 53324 destroy picture 0x1cd457e0, with drawable 0x1cd65820 53325 53326 Obtaining format for pixmap 0x1327d1f0 and picture 0x1cd457e0 53327 ==7989== Invalid read of size 4 53328 ==7989== at 0x8CAA0CA: glamor_get_tex_format_type_from_pixmap (glamor_utils.h:1252) 53329 ==7989== by 0x8CAD1B7: glamor_download_sub_pixmap_to_cpu (glamor_pixmap.c:1074) 53330 ==7989== by 0x8CA8BB7: _glamor_get_image (glamor_getimage.c:66) 53331 ==7989== by 0x8CA8D2F: glamor_get_image (glamor_getimage.c:92) 53332 ==7989== by 0x29AEF2: miSpriteGetImage (misprite.c:413) 53333 ==7989== by 0x1E7674: compGetImage (compinit.c:148) 53334 ==7989== by 0x1F5E5B: ProcShmGetImage (shm.c:684) 53335 ==7989== by 0x1F686F: ProcShmDispatch (shm.c:1121) 53336 ==7989== by 0x15D00D: Dispatch (dispatch.c:432) 53337 ==7989== by 0x14C569: main (main.c:298) 53338 ==7989== Address 0x1cd457f0 is 16 bytes inside a block of size 120 free'd 53339 ==7989== at 0x4C2B60C: free (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 53340 ==7989== by 0x228897: FreePicture (picture.c:1477) 53341 ==7989== by 0x228B23: PictureDestroyWindow (picture.c:73) 53342 ==7989== by 0x234C19: damageDestroyWindow (damage.c:1646) 53343 ==7989== by 0x1E92C0: compDestroyWindow (compwindow.c:590) 53344 ==7989== by 0x20FF85: DbeDestroyWindow (dbe.c:1389) 53345 ==7989== by 0x185D46: FreeWindowResources (window.c:907) 53346 ==7989== by 0x1889A7: DeleteWindow (window.c:975) 53347 ==7989== by 0x17EBF1: doFreeResource (resource.c:873) 53348 ==7989== by 0x17FC1B: FreeClientResources (resource.c:1139) 53349 ==7989== by 0x15C4DE: CloseDownClient (dispatch.c:3402) 53350 ==7989== by 0x2AB843: CheckConnections (connection.c:1008) 53351 ==7989== 53352 (II) fail to get matched format for dfdfdfdf 53353 53354 The fix is to update the picture pointer when the window pixmap is changed, 53355 so it moves the picture around with the window rather than the pixmap. 53356 53357 This makes FreePicture work correctly. 53358 53359 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 53360 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=71088 53361 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53362 53363commit 2925db6616944c8d1e911caee0637a00e5586576 53364Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53365Date: Sat Nov 2 00:08:11 2013 +0800 53366 53367 Fixed an incorrect printf format. 53368 53369 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=69573 53370 53371 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53372 53373commit 644e05562e401f192c93964be06c33fe92099089 53374Author: Brian Paul <brianp@vmware.com> 53375Date: Sat Oct 19 16:10:54 2013 -0600 53376 53377 Remove useless return statement 53378 53379 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53380 53381commit 8c51eb8239247c98057eb9daf06c905a1dc0588c 53382Author: Brian Paul <brianp@vmware.com> 53383Date: Sat Oct 19 16:10:53 2013 -0600 53384 53385 Remove redundant dispatch->glEnable(GL_TEXTURE_2D) 53386 53387 The same call was already made a few lines earlier. 53388 53389 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53390 53391commit 3bf1eb577e2d2b4d55d38b5a0043b0efb9abb385 53392Author: Brian Paul <brianp@vmware.com> 53393Date: Sat Oct 19 16:10:52 2013 -0600 53394 53395 Fix _glamor_set_spans() bug (re-used 'n' variable) 53396 53397 n was used as a function parameter. But inside the for (i=1..n) loop, 53398 n got reassigned as REGION_NUM_RECTS() and then decremented to zero by 53399 the while loop. This caused the for loop to only iterate once instead 53400 of 'n' times. 53401 53402 This patch renames the n parameter to numPoints. 53403 53404 Found by code inspection. Untested. 53405 53406 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53407 53408commit 2f62bd46cc002af57f8fe3246b9aea9e0c0bbcc9 53409Author: Grigori Goronzy <greg@chown.ath.cx> 53410Date: Wed Oct 2 00:37:57 2013 +0200 53411 53412 glamor_render: fix PictFilters 53413 53414 Add Fast/Good/Best and appropriately map to Nearest and 53415 Bilinear. Additionally, add a fallback path for unsupported filters. 53416 53417 Notably, this fixes window shadow rendering with Compiz, which uses 53418 PictFilterConvolution for some odd reason. 53419 53420 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 53421 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53422 53423commit 5695708ecd2a26fcb9c05985c6758e719071995a 53424Author: Grigori Goronzy <greg@chown.ath.cx> 53425Date: Wed Oct 2 00:37:56 2013 +0200 53426 53427 Use GL_STATIC_DRAW for element index buffer 53428 53429 The buffer never changes anyway. 53430 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 53431 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53432 53433commit 8afa008ec4b393666bb3c506a711b4d50cc4e756 53434Author: Grigori Goronzy <greg@chown.ath.cx> 53435Date: Wed Oct 2 00:37:55 2013 +0200 53436 53437 Use glDrawRangeElements instead of glDrawElements 53438 53439 This lets us explicitly specify the range of vertices that are used, 53440 which the OpenGL driver can use for optimization. Particularly, 53441 it results in lower CPU overhead with Mesa-based drivers. 53442 53443 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 53444 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53445 53446commit 229601e080aefad21927c1449cd520733317858b 53447Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53448Date: Fri Sep 27 05:25:54 2013 +0800 53449 53450 Shoud return null subpixmap if we fail to get a valid map address. 53451 53452 The patch is prepared by Raul Fernandes. 53453 53454 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=86693 53455 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53456 53457commit e3d1d4e3caaab8076eba89b58d037d24e203e506 53458Author: Dave Airlie <airlied@redhat.com> 53459Date: Mon Sep 23 06:42:24 2013 +0100 53460 53461 glamor: add initial Xv support 53462 53463 This does YV12 and I420 for now, not sure if we can do packed without 53464 a GL extension. 53465 53466 Signed-off-by: Dave Airlie <airlied@redhat.com> 53467 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 53468 53469commit 39e95cd0f50cf471d8220d5428788d5be6a59d3f 53470Author: Michel Dänzer <michel.daenzer@amd.com> 53471Date: Fri Sep 20 10:41:10 2013 +0200 53472 53473 Reset traps_count and ptrap when necessary for the next trapezoid cliprect 53474 53475 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=64912 53476 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 53477 Reviewed-by: He Junyan <junyan.he@inbox.com> 53478 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53479 53480commit 61fca4342a65be2dbc7f890f2e67da56a50db978 53481Author: Michel Dänzer <michel.daenzer@amd.com> 53482Date: Tue Sep 17 13:25:02 2013 +0200 53483 53484 Fix RegionContainsRect test for PutImage 53485 53486 The return value of RegionContainsRect() is not a boolean but an enum 53487 indicating that the region contains the rectangle completely, partially 53488 or not at all. We can only take the PutImage fastpath when the region 53489 contatins the rectangle completely. 53490 53491 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=65964 53492 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 53493 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53494 53495commit 2ea618f2cf906fec9807b321f9e8ef3a4706c708 53496Author: Christian König <christian.koenig at amd.com> 53497Date: Mon Aug 26 14:57:47 2013 +0800 53498 53499 Use GBM_LIBS and GBM_CFLAGS 53500 53501 Signed-off-by: Christian König <christian.koenig at amd.com> 53502 Reviewed-by: Michel Dänzer <michel.daenzer at amd.com> 53503 53504commit 7e818f7d39cfef2701fe9cf95c7854ee8c9f3be6 53505Author: Armin K <krejzi@email.com> 53506Date: Wed Mar 13 18:49:32 2013 +0100 53507 53508 First attempt to make libglamor.so shared versioned library 53509 53510 As recommended by Michel in this thread reply: 53511 http://lists.freedesktop.org/archives/glamor/2013-March/000305.html 53512 53513 v2: Correct shared library location in glamor.pc.in 53514 53515 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=62259 53516 53517 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 53518 53519commit b0318b6de7a2f8c5d9e75b55211dcc0b12f811fc 53520Author: Armin K <krejzi@email.com> 53521Date: Wed Mar 13 18:28:58 2013 +0100 53522 53523 Properly dist necesary headers 53524 53525 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 53526 53527commit fc179bb863778c03288bebaa258358ccbdb6aa0c 53528Author: Armin K <krejzi@email.com> 53529Date: Wed Mar 13 18:28:57 2013 +0100 53530 53531 Silence Automake 1.13 warnings 53532 53533 warning: 'INCLUDES' is the old name for 'AM_CPPFLAGS' (or '*_CPPFLAGS') 53534 53535 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 53536 53537commit 97416e3f144606728a963aa2a337b1283e156ad2 53538Author: Michel Dänzer <michel.daenzer@amd.com> 53539Date: Wed Mar 13 17:15:33 2013 +0100 53540 53541 glamoregl: Use xf86ScreenToScrn() 53542 53543 Fixes crashes when glamor is used for a GPU screen with xserver 1.13 or 53544 newer. 53545 53546 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=57200#c17 53547 53548 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 53549 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 53550 53551commit 4c5bfd529f6f182a3f0ac1b9bb5c3170a297f460 53552Author: Dave Airlie <airlied@redhat.com> 53553Date: Sat Dec 29 06:42:30 2012 +0000 53554 53555 glamor_utils: fix unlikely define use 53556 53557 using a define across a split line expression is failure, compiling 53558 with warnings shows this up. 53559 53560 Signed-off-by: Dave Airlie <airlied@redhat.com> 53561 53562commit 6b954880c25af353dd4679cfdad6664b107d4df6 53563Author: Dave Airlie <airlied@redhat.com> 53564Date: Sat Dec 29 06:42:10 2012 +0000 53565 53566 glamor: add compiler.h 53567 53568 This is also required for distchecking. 53569 53570 Signed-off-by: Dave Airlie <airlied@redhat.com> 53571 53572commit efbdc9e90f36494b07508aeaba40c9eb01eab398 53573Author: Dave Airlie <airlied@redhat.com> 53574Date: Sat Dec 29 06:28:17 2012 +0000 53575 53576 glamor: fix make distcheck part 1 53577 53578 This just adds the headers, then it falls over on the sdk_HEADERS 53579 as it overrides proper install paths by the looks of it. 53580 53581 Signed-off-by: Dave Airlie <airlied@redhat.com> 53582 53583commit 80f5e21daeed79654d6e1976f2766eb528f01d47 53584Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53585Date: Tue Nov 13 10:08:02 2012 +0800 53586 53587 glamor_compositerects: Need to initialize region before fallback. 53588 53589 As we need to call DamageRegionAppend even for fallback path, 53590 we must initialize the region before do that. Pointed by 53591 Igor Vagulin. 53592 53593 https://bugs.freedesktop.org/show_bug.cgi?id=56940 53594 53595 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53596 53597commit 14e02f5132d7b4f50c8d9f8e5c6a0f285a3f3c14 53598Author: Michel Dänzer <michel.daenzer@amd.com> 53599Date: Wed Oct 31 16:56:00 2012 +0100 53600 53601 Don't use glBlitFramebufferEXT for overlapping copies. 53602 53603 According to the GL_EXT_framebuffer_blit spec, the result of doing so is 53604 undefined. But we need well-defined results. :) 53605 53606 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 53607 53608commit e846f48f489e33e270db682b1b6c8e6f03fdf313 53609Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53610Date: Wed Aug 8 20:43:38 2012 +0800 53611 53612 Increase vbo size to 64K verts. 53613 53614 This commit will benefit vertex stressing cases such as 53615 aa10text/rgb10text, and can get about 15% performance gain. 53616 53617 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53618 Acked-by: Junyan <junyan.he@linux.intel.com> 53619 53620commit b8f0a2188295f6cc2c997b6639b6928323eae617 53621Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53622Date: Wed Aug 8 20:11:43 2012 +0800 53623 53624 Silence compilation warnings. 53625 53626 After increase to gcc4.7, it reports more warnings, now 53627 fix them. 53628 53629 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53630 Tested-by: Junyan He<junyan.he@linux.intel.com> 53631 53632commit 50614451adafc816ae5ffbe9c2a97a58f493b927 53633Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53634Date: Wed Aug 8 20:08:12 2012 +0800 53635 53636 glamor_largepixmap: Fixed a bug in repeat clipping. 53637 53638 If the repeat direction only has one block, then we need to set the 53639 dx/dy to cover all the extent. This commit also silence some compilation 53640 warnings. 53641 53642 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53643 53644commit 7eb434918bd7d58162c53fac9316aceda136cb6f 53645Author: Michel Dänzer <michel.daenzer@amd.com> 53646Date: Tue Aug 7 18:16:58 2012 +0200 53647 53648 Prefer KHR_surfaceless_context EGL extension over KHR_surfaceless_opengl/gles2. 53649 53650 Current Mesa Git only advertises the former instead of the latter. 53651 53652 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 53653 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53654 53655commit 59653fa08a67a0694b3408407f5f92af99470739 53656Author: Michel Dänzer <michel.daenzer@amd.com> 53657Date: Tue Aug 7 18:13:32 2012 +0200 53658 53659 Print space between name of missing EGL extension and 'required'. 53660 53661 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 53662 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53663 53664commit c3096c5a568c9c7b5f8a68ca882d866b8fb8e3ad 53665Author: Junyan He <junyan.he@linux.intel.com> 53666Date: Tue Aug 7 05:01:59 2012 +0800 53667 53668 Fallback to pixman when trapezoid mask is big. 53669 53670 The trapezoid generating speed of the shader is relatively 53671 slower when the trapezoid area is big. We fallback when 53672 the trapezoid's width and height is bigger enough. 53673 The big traps number will also slow down the render because 53674 of the VBO size. We fallback if ntrap > 256 53675 53676 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 53677 Reviewed-By: Zhigang Gong <zhigang.gong@linux.intel.com> 53678 53679commit f62f4d53ef91ba05cc9456b1468cd1b24252454f 53680Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53681Date: Thu Aug 2 18:07:07 2012 +0800 53682 53683 glamor_glyphs: When dst arg point to a NULL buffer, dont't flush. 53684 53685 This is a corner case, when we render glyphs via mask cache, and 53686 when we need to upload new glyphs cache, we need to flush both the 53687 mask and dest buffer. But we the dest arg may point to a NULL buffer 53688 at that time, we need to check it firstly. If the dest buffer is NULL. 53689 Then we don't need to flush both the dest and mask buffer. 53690 53691 This commit fix a potential crash. 53692 53693 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53694 53695commit e7af7cb76dade31b9e40cede7ac21d32bc95831c 53696Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53697Date: Wed Aug 1 18:27:29 2012 +0800 53698 53699 glamor_trapezoid: workaround a glsl like problem. 53700 53701 It seems that the following statement cann't run as expected on SNB. 53702 bool trap_left_vertical = (abs(trap_left_vertical_f - 1.0) <= 0.0001); 53703 53704 Have to rewrite it to another style to let the vertical edge trapezoid 53705 to be rendered correctly. 53706 53707 Reviewed-by: Junyan He <junyan.he@linux.intel.com> 53708 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53709 53710commit 5512c14e346a988f6eb8cdf9385a868ef705450c 53711Author: Junyan He <junyan.he@linux.intel.com> 53712Date: Fri Jul 20 05:52:57 2012 +0800 53713 53714 Fix the problem of VBO leak. 53715 53716 In some cases we allocate the VBO but have no vertex to 53717 emit, which cause the VBO fail to be released. Fix it. 53718 53719 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 53720 53721commit 9f78e22fa62d8f1a08d6a0688e2eafba35206a3b 53722Author: Junyan He <junyan.he@linux.intel.com> 53723Date: Fri Jul 20 05:52:50 2012 +0800 53724 53725 Just use the shader to generate trapezoid if PolyMode == Imprecise 53726 53727 The precise mode of trapezoid rendering need to sample the trapezoid on 53728 the centre points of an (2*n+1)x(2*n-1) subpixel grid. It is computationally 53729 expensive in shader, and we use inside area ratio to replace it. 53730 The result has some difference, and we just use it if the polymode == Imprecise. 53731 53732 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 53733 53734commit fe024a78220a6d61c6b1df89166da953b19b633a 53735Author: Junyan He <junyan.he@linux.intel.com> 53736Date: Fri Jul 20 05:52:43 2012 +0800 53737 53738 Change the trapezoid render to use VBO. 53739 53740 Because some uniform variables need to be set for every 53741 trapezoid rendering, we can not use vbo to render multi 53742 trapezoids one time, which have performance big loss. 53743 We now add attributes which contain the same value to bypass 53744 the uniform variable problem. The uniform value for one 53745 trapezoid will be set to the same value to all the vertex 53746 of that trapezoid as an attribute, then in FS, it is still 53747 a constant. 53748 53749 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 53750 53751commit 9dff3378e5c39b346985924f9ab6291069cc7b6e 53752Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53753Date: Mon Jul 16 17:27:22 2012 +0800 53754 53755 Added the missed header file for xorg 1.13 compat. 53756 53757 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53758 53759commit bc1b412b3b0bb716702ec89ae512f2b5ec62c17a 53760Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53761Date: Mon Jul 16 11:25:09 2012 +0800 53762 53763 Synch with xorg 1.13 change. 53764 53765 As xorg 1.13 change the scrn interaces and remove those 53766 global arrays. Some API change cause we can't build. Now 53767 fix it. 53768 53769 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53770 53771commit 4c27ca4700e4ba4ae19d77377a7776eb32f74647 53772Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53773Date: Fri Jul 13 09:20:02 2012 +0800 53774 53775 gles2: Fixed the compilation problem and some bugs. 53776 53777 Previous patch doesn't set the offset to zero for GLESv2 53778 path. Now fix it. 53779 53780 This patch also fix a minor problem in pixmap uploading 53781 preparation. If the revert is not REVERT_NORMAL, then we 53782 don't need to prepare a fbo for it. As current mesa i965 53783 gles2 driver doesn't support to set a A8 texture as a fbo 53784 target, we must fix this problem. As some A1/A8 picture 53785 need to be uploaded, this is the only place a A8 texture 53786 may be attached to a fbo. 53787 53788 This patch also enable the shader gradient for GLESv2. 53789 The reason we disable it before is that some glsl linker 53790 doesn't support link different objects which have cross 53791 reference. Now we don't have that problem. 53792 53793 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53794 53795commit 006fe0e66df4b214ca5c50241b3cca22d31161df 53796Author: Michel Dänzer <michel.daenzer@amd.com> 53797Date: Thu Jul 12 18:57:06 2012 +0800 53798 53799 Stream vertex data to VBOs. 53800 53801 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 53802 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53803 53804commit 551ca11c77d0524ceb37eb26caf0c25e4d7d806f 53805Author: Michel D=C3=A4nzer <michel.daenzer@amd.com> 53806Date: Wed Jul 11 15:01:15 2012 +0800 53807 53808 Fix translation of clip region for composite fallback. 53809 53810 Fixes incorrectly clipped rendering. E.g. the cursor in Evolution 53811 composer windows became invisible. 53812 53813 Signed-off-by: Michel Daenzer <michel.daenzer@amd.com> 53814 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53815 53816commit 88c317fb1e6a9056acab4a76c0ee0bf283e001ce 53817Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53818Date: Wed Jul 4 17:59:25 2012 +0800 53819 53820 glamor_glyphs: Don't merge extents for different lists. 53821 53822 If we merge all lists's extent together, than we may have 53823 some fail overlap checking. Here is a simple: 53824 A E 53825 B F 53826 C 53827 D 53828 53829 The first list has vertical "ABCD". And the second list 53830 has two char "EF". When detecting E, it can successfully 53831 find it doesn't overlap with previous glyphs. But after 53832 that, the original code will merge the previous extent with 53833 E's extent, then the extent will cover "F", so when detecting 53834 F, it will be treated as overlapped. 53835 53836 We can simply solve this issue by not merge extent from different 53837 list. We can union different list's extent to a global region. 53838 And then do the intersect checkint between that region and 53839 current glyph extent, then we can avoid that fail checking. 53840 53841 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53842 53843commit 32a7438bf79573100862ad91c5722d12623a8bf9 53844Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53845Date: Tue Jul 3 18:24:07 2012 +0800 53846 53847 glamor_copyarea: Use blitcopy if current state is not render. 53848 53849 Practically, for pure 2D blit, the blit copy is much faster 53850 than textured copy. For the x11perf copywinwin100, it's about 53851 3x faster. But if we have heavy rendering/compositing, then use 53852 textured copy will get much better (>30%)performance for most 53853 of the cases. 53854 53855 So we simply add a data element to track current state. For 53856 rendering state we use textured copy, otherwise, we use blit 53857 copy. 53858 53859 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53860 53861commit 0706423bcfd7a589bab3b41fe9f13d0b636ecdef 53862Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53863Date: Tue Jul 3 18:11:12 2012 +0800 53864 53865 glamor_glyphs: Use cache picture to store mask picture if possible. 53866 53867 By default, mask picture is newly created, and each time we need to 53868 clear the whole mask picture, and then composite glyphs to the mask 53869 picture and then composite the mask picture to destination. 53870 53871 Testing results shows that the filling of the mask picture takes a 53872 big portion of the rendering time. As we don't really need to clear 53873 the whole region, we just need to clear the real overlapped region. 53874 53875 This commit is to solve this issue. We split a large glyphs list to 53876 serval lists and each list is non-overlapped or overlapped. 53877 53878 we can reduce the length of overlapped glyphs to do the glyphs_via_mask 53879 to 2 or 3 glyphs one time for most cases. Thus it give us a case to allocate a 53880 small portion of the corresponding cache directly as the mask picture. 53881 Then we can rendering the glyphs to this mask picture, and latter we 53882 can accumulate the second steps, composite the mask to the dest with 53883 the other non-overlapped glyphs's rendering process. 53884 It also make us implement a batch mask cache blocks clearing 53885 algorithm to avoid too frequently small region clearing. 53886 53887 If there is no any overlapping, this method will not get performance gain. 53888 If there is some overlapping, then this algorithm can get about 15% performance 53889 gain. 53890 53891 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53892 53893commit 4d1a2173f2e5a200d1535a4a459fffd75cd5f779 53894Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53895Date: Wed Jun 27 12:53:40 2012 +0800 53896 53897 glamor_compositerects: Implement optimized version. 53898 53899 Don't call miCompositeRects. Use glamor_composite_clipped_region 53900 to render those boxes at once. 53901 Also add a new function glamor_solid_boxes to fill boxes at once. 53902 53903 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53904 53905commit dd7924339803b51332e3fe94d36eb48549e9d24d 53906Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53907Date: Tue Jun 26 20:00:21 2012 +0800 53908 53909 optimize: Use likely and unlikely. 53910 53911 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53912 53913commit d5f03ba0109bc50e159c82d3d67cffa8fc174134 53914Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53915Date: Tue Jun 26 19:56:27 2012 +0800 53916 53917 create_pixmap: use texture for large glyphs. 53918 53919 As we only cache glyphs smaller than 64x64, we need to use 53920 texutre for the large glyphs. 53921 53922 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53923 53924commit 90dd6ddbab864aa6c3add55c46a85d7db2c125f6 53925Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53926Date: Tue Jun 26 17:19:16 2012 +0800 53927 53928 glamor_copyarea: Fixed a bug introduced by 996194... 53929 53930 Default return value should be FALSE. 53931 53932 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53933 53934commit b8dd2a597de455ae985302d1b93d865ef9f37c7b 53935Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53936Date: Tue Jun 26 16:24:32 2012 +0800 53937 53938 glamor_glyphs: Slightly performance tuning. 53939 53940 As glamor_glyphs never fallback, we don't need to keep the 53941 underlying glyphs routines, just override the ps->glys 53942 53943 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53944 53945commit 3873d412f018b975feaa000cb7ef337feaeee37d 53946Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53947Date: Tue Jun 26 15:39:24 2012 +0800 53948 53949 glamor_render: Don't allocate buffer for vbo each time. 53950 53951 We can reuse the last one if the last one is big enough 53952 to contain current vertext data. In the meantime, Use 53953 MapBufferRange instead of MapBuffer. 53954 53955 Testing shows, this patch brings some benefit for 53956 aa10text/rgb10text. Not too much, but indeed faster. 53957 53958 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53959 53960commit 682f5d2989306856c0c193b0b4abd8c104a9ba92 53961Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53962Date: Tue Jun 26 13:09:05 2012 +0800 53963 53964 glamor_largepixmap: Walkaround for large texture's upload. 53965 53966 I met a problem with large texture (larger than 7000x7000)'s 53967 uploading on SNB platform. The map_gtt get back a mapped VA 53968 without error, but write to that virtual address triggers 53969 BUS error. This work around is to avoid that direct uploading. 53970 53971 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53972 53973commit 37d4022f01483eefe06b5f12e13ca2ed4003c8f0 53974Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53975Date: Mon Jun 25 23:24:37 2012 +0800 53976 53977 glamor_render: Optimize the two pass ca rendering. 53978 53979 For the componentAlpha with PictOpOver, we use two pass 53980 rendering to implement it. Previous implementation call 53981 two times the glamor_composite_... independently which is 53982 very inefficient. Now we change the control flow, and do 53983 the two pass internally and avoid duplicate works. 53984 53985 For the x11perf -rgb10text, this optimization can get about 53986 30% improvement. 53987 53988 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53989 53990commit 21916cf84f0cd9ada8701650d39b5cf67646eaf7 53991Author: Zhigang Gong <zhigang.gong@linux.intel.com> 53992Date: Mon Jun 25 16:37:22 2012 +0800 53993 53994 glamor_composite_glyph: Optimize glyphs with non-solid pattern. 53995 53996 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 53997 53998commit c1bd50d58d2c7b39e2b5c529bd86fde1ab14d8e6 53999Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54000Date: Thu Jun 21 19:30:51 2012 +0800 54001 54002 glamor_glyphs: Detect fake or real glyphs overlap. 54003 54004 To split a glyph's extent region to three sub-boxes 54005 as below. 54006 54007 left box 2 x h 54008 center box (w-4) x h 54009 right box 2 x h 54010 54011 Take a simple glyph A as an example: 54012 * 54013 __* *__ 54014 ***** 54015 * * 54016 ~~ ~~ 54017 54018 The left box and right boxes are both 2 x 2. The center box 54019 is 2 x 4. 54020 54021 The left box has two bitmaps 0001'b and 0010'b to indicate 54022 the real inked area. 54023 The right box also has two bitmaps 0010'b and 0001'b. 54024 54025 And then we can check the inked area in left and right boxes with 54026 previous glyph. If the direction is from left to right, then we 54027 need to check the previous right bitmap with current left bitmap. 54028 54029 And if we found the center box has overlapped or we overlap with 54030 not only the previous glyph, we will treat it as real overlapped 54031 and will render the glyphs via mask. 54032 54033 If we only intersect with previous glyph on the left/right edge. 54034 Then we further compute the real overlapped bits. We set a loose 54035 check criteria here, if it has less than two pixel overlapping, we 54036 treat it as non-overlapping. 54037 54038 With this patch, The aa10text boost fom 1660000 to 320000. 54039 Almost double the performance! And the cairo test result is the 54040 same as without this patch. 54041 54042 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54043 54044commit ea4c22716ca1544e924c7462db6a2797afebff59 54045Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54046Date: Thu Jun 21 19:29:06 2012 +0800 54047 54048 glamor_render: Don't fallback when rendering glyphs with OpOver. 54049 54050 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54051 54052commit 7acbe895618837305cf4050731ccd1cd3bc11589 54053Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54054Date: Thu Jun 21 19:27:07 2012 +0800 54055 54056 glamor_create_pixmap: Allocate glyphs pixmap in memory. 54057 54058 As we have glyphs atlas cache, we don't need to hold each 54059 glyphs on GPU. And for the subsequent optimization, we need 54060 to store the original glyphs pixmap on system memory. 54061 54062 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54063 54064commit 1e4fc85a71cc6498a7e11872026062fceb5bbdf1 54065Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54066Date: Thu Jun 21 19:26:28 2012 +0800 54067 54068 glamor_fbo: fix a memory leak for large pixmap. 54069 54070 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54071 54072commit 2122e60bf9027b63ddc59c0aa2a441af3687cb3a 54073Author: Junyan He <junyan.he@linux.intel.com> 54074Date: Fri Jun 15 09:00:15 2012 +0800 54075 54076 Fix a bug for trapezoid clip 54077 54078 We find in some cases the trapezoid will be render as a triangle and 54079 the left edge and right edge will cross with each other just bellow 54080 the top or over the bottom. The distance between the cross poind and 54081 the top or bottom is less than pixman_fixed_1_minus_e, so after the 54082 fixed converted to int, the cross point has the same value with the 54083 top or botton and the triangle should not be affected. But in our 54084 clip logic, the cross point will be clipped out. So add a logic 54085 to fix this problem. 54086 54087 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54088 54089commit 6ed418d17b5143f32b3b415103f3157a1b05e3db 54090Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54091Date: Fri Jun 15 13:31:30 2012 +0800 54092 54093 gles2_largepixmap: force clip for a non-large pixmap. 54094 54095 One case we need force clip when download/upload a drm_texture 54096 pixmap. Actually, this is only meaningful for testing purpose. 54097 As we may set the max_fbo_size to a very small value, but the 54098 drm texture may exceed this value but the drm texture pixmap 54099 is not largepixmap. This is not a problem with OpenGL. But for 54100 GLES2, we may need to call glamor_es2_pixmap_read_prepare to 54101 create a temporary fbo to do the color conversion. Then we have 54102 to force clip the drm pixmap here to avoid large pixmap handling 54103 at glamor_es2_pixmap_read_prepare. 54104 54105 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54106 54107commit c41d5c79e7d8772be38b77122817fb872f2d721d 54108Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54109Date: Fri Jun 15 09:03:42 2012 +0800 54110 54111 glamor_emit_composite_vert: Optimize to don't do two times vert coping. 54112 54113 We change some macros to put the vert to the vertex buffer 54114 directly when we cacluating it. This way, we can get about 54115 4% performance gain. 54116 54117 This commit also fixed one RepeatPad bug, when we RepeatPad 54118 a not eaxct size fbo. We need to calculate the edge. The edge 54119 should be 1.0 - half point, not 1.0. 54120 54121 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54122 54123commit 8656ddbbe7a38e5cd15e6d006bba19778e34e9e9 54124Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54125Date: Fri Jun 15 08:46:24 2012 +0800 54126 54127 glamor_glyphs: Before get upload to cache flush is needed. 54128 54129 When we can't get a cache hit and have to evict one cache 54130 entry to upload new picture, we need to flush the previous 54131 buffer. Otherwise, we may get corrupt glyphs rendering result. 54132 54133 This is the reason why user-font-proxy may fail sometimes. 54134 54135 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54136 54137commit 016995334e6acfad41574a9270b716db98ec697f 54138Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54139Date: Fri Jun 15 08:43:37 2012 +0800 54140 54141 copyarea: Cleanup the error handling logic. 54142 54143 Should use ok rather than mixed ok or ret. 54144 54145 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54146 54147commit b4a499b7db068117801770b7ab80416014822101 54148Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54149Date: Tue Jun 12 18:21:10 2012 +0800 54150 54151 trapezoid: Fallback to sw-rasterize for largepixmap. 54152 54153 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54154 54155commit 8f31aed48ca9412f8f5df9c3d47d938c7c90bccb 54156Author: Junyan He <junyan.he@linux.intel.com> 54157Date: Tue Jun 5 03:21:03 2012 +0800 54158 54159 Use the direct render path for A1 54160 54161 Because when mask depth is 1, there is no Anti-Alias at all, 54162 in this case, the directly render can work well and it is faseter. 54163 54164 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54165 54166commit fa74a213ad175cb0920905a8011c62efdd348d45 54167Author: Junyan He <junyan.he@linux.intel.com> 54168Date: Mon Jun 4 07:52:05 2012 +0800 54169 54170 Add the trapezoid direct render logic 54171 54172 We firstly get the render area by clipping the trapezoid 54173 with the clip rect, then split the clipped area into small 54174 triangles and use the composite logic to generate the result 54175 directly. This manner is fast but have the problem that 54176 some implementation of GL do not implement the Anti-Alias 54177 of triangles fill, so the edge sometimes has sawtooth. It is 54178 not acceptable when use trapezoid to approximate circles and 54179 wide lines. 54180 54181 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54182 54183commit 5f1560c84aa386bb12dc6ec3daeb93e0f63c09c2 54184Author: Junyan He <junyan.he@linux.intel.com> 54185Date: Mon Jun 4 07:22:27 2012 +0800 54186 54187 Modilfy the composite logic to two phases 54188 54189 We seperate the composite to two phases, firstly to 54190 select the shader according to source type and logic 54191 op, setting the right parameters. Then we emit the 54192 vertex array to generate the dest result. 54193 The reason why we do this is that the shader may be 54194 used to composite no only rect, trapezoid and triangle 54195 render function can also use it to render triangles and 54196 polygens. The old function glamor_composite_with_shader 54197 do the whole two phases work and can not match the 54198 new request. 54199 54200 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54201 54202commit 0b0391765f3ca0192b00f4970c9453934d529397 54203Author: Junyan He <junyan.he@linux.intel.com> 54204Date: Mon Jun 4 03:18:40 2012 +0800 54205 54206 Add macro of vertex setting for triangle stripe 54207 54208 Add macro of vertex setting for triangle stripe draw, 54209 and make the code clear. 54210 54211 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54212 54213commit bd180be619c2ef469ca40cf95023340c59960540 54214Author: RobinHe <robinhe@robinhe-desktop.lan> 54215Date: Sat Jun 2 22:00:09 2012 +0800 54216 54217 Use shader to generate the temp trapezoid mask 54218 54219 The old manner of trapezoid render uses pixman to 54220 generate a mask pixmap and upload it to the GPU. 54221 This effect the performance. We now use shader to 54222 generate the temp trapezoid mask to avoid the 54223 uploading of this pixmap. 54224 We implement a anti-alias manner in the shader 54225 according to pixman, which will caculate the area 54226 inside the trapezoid dividing total area for every 54227 pixel and assign it to the alpha value of that pixel. 54228 The pixman use a int-to-fix manner to approximate but 54229 the shader use float, so the result may have some 54230 difference. 54231 Because the array in the shader has optimization problem, 54232 we need to emit the vertex of every trapezoid every 54233 time, which will effect the performance a lot. Need to 54234 improve it. 54235 54236 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54237 54238commit 6dd81c5939197eb51bc31ffbc3e6f359f2aad191 54239Author: RobinHe <robinhe@robinhe-desktop.lan> 54240Date: Sat Jun 2 21:52:25 2012 +0800 54241 54242 Create the file glamor_triangles.c 54243 54244 Create the file glamor_trapezoid.c, extract the logic 54245 relating to trapezoid from glamor_render.c to this file. 54246 54247 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54248 54249commit bf38ee407ba1038a25efa58b8475bcf084a21042 54250Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54251Date: Mon Jun 11 14:02:59 2012 +0800 54252 54253 Enable large pixmap by default. 54254 54255 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54256 54257commit 8ca16754f7e95e47423183bbd5f72966b937f6c8 54258Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54259Date: Mon Jun 11 13:05:48 2012 +0800 54260 54261 largepixmap: Fix the selfcopy issue. 54262 54263 If the source and destination are the same pixmap/fbo, and we 54264 need to split the copy to small pieces. Then we do need to 54265 consider the sequence of the small pieces when the copy area 54266 has overlaps. This commit take the reverse/upsidedown into 54267 the clipping function, thus it can generate correct sequence 54268 and avoid corruption self copying. 54269 54270 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54271 54272commit 5325c800f706f46085735d608f57d513da63cddf 54273Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54274Date: Mon Jun 11 11:42:46 2012 +0800 54275 54276 largepixmap: Support self composite for large pixmap. 54277 54278 The simplest way to support large pixmap's self compositing 54279 is to just clone a pixmap private data structure, and change 54280 the fbo and box to point to the correct postions. Don't need 54281 to copy a new box. 54282 54283 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54284 54285commit 1d2d858b8daacdd349084433a8af60d559fb6a9d 54286Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54287Date: Mon Jun 11 09:36:56 2012 +0800 54288 54289 largepixmap: Add transform/repeat/reflect/pad support. 54290 54291 This commit implement almost all the needed functions for 54292 the large pixmap support. It's almost complete. 54293 54294 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54295 54296commit 48916a23a92366128f5a1eeb08949d8d32383c32 54297Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54298Date: Mon Jun 11 02:27:00 2012 +0800 54299 54300 glamor_getimage: should call miGetimage if failed to get sub-image. 54301 54302 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54303 54304commit 56d6e7a85fbe8e50c38efda0f59f09aac52b769a 54305Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54306Date: Mon Jun 11 02:24:41 2012 +0800 54307 54308 glamor_putimage: Correct the wrong stride value. 54309 54310 We should not use the destination pixmap's devkind as the input 54311 image data's stride. 54312 54313 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54314 54315commit eb6f981ba4d3a4aff1b3651bdad3cd7a52233b41 54316Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54317Date: Mon Jun 11 02:15:53 2012 +0800 54318 54319 largepixmap: Enable glamor_composite. 54320 54321 Now we start to enable glamor_composite on large pixmap. 54322 We need to do a three layer clipping to split the dest/source/mask 54323 to small pieces. This commit only support non-transformation and 54324 repeat normal case. 54325 54326 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54327 54328commit e96ea02010874a3a46f212f42134083bd29fefe3 54329Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54330Date: Mon Jun 11 01:52:14 2012 +0800 54331 54332 largepixmap: Implement infrastructure for large pixmap. 54333 54334 Added infrastructure for largepixmap, this commit implemented: 54335 1. Create/Destroy large pixmap. 54336 2. Upload/Download large pixmap. 54337 3. Implement basic repeat normal support. 54338 3. tile/fill/copyarea large pixmap get supported. 54339 54340 The most complicated part glamor_composite still not implemented. 54341 54342 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54343 54344commit ace35e408cd7a79c5215bbd0f14b624d8d949e34 54345Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54346Date: Mon Jun 11 01:02:30 2012 +0800 54347 54348 glamor_largepixmap: first commit for large pixmap. 54349 54350 This is the first commit to add support for large pixmap. 54351 The large here means a pixmap is larger than the texutre's 54352 size limitation thus can't fit into one single texutre. 54353 54354 The previous implementation will simply fallback to use a 54355 in memory pixmap to contain the large pixmap which is 54356 very slow in practice. 54357 54358 The basic idea here is to use an array of texture to hold 54359 the large pixmap. And when we need to get a specific area 54360 of the pixmap, we just need to compute/clip the correct 54361 region and find the corresponding fbo. 54362 54363 We need to implement some auxiliary routines to clip every 54364 rendering operations into small pieces which can fit into 54365 one texture. 54366 54367 The complex part is the transformation/repeat/repeatReflect 54368 and repeat pad and their comination. We will support all of 54369 them step by step. 54370 54371 This commit just add some necessary data structure to represent 54372 the large pixmap, and doesn't change any rendering process. 54373 This commit doesn't add real large pixmap support. 54374 54375 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54376 54377commit 4c174f4c9ce1514ef226e9de97e5c87a46a75524 54378Author: Junyan He <junyan.he@linux.intel.com> 54379Date: Tue May 15 10:08:03 2012 +0800 54380 54381 Fix the problem of x_source and y_source causing radial error 54382 54383 The x_source and y_source cause some problem in 54384 gradient. The old way to handle it by recaulate P1 P2 54385 to minus the x_source and y_source, but this causes 54386 problem in radial shader. Now we modify the manner to 54387 set the texture coordinates: (x_source, y_source) --> 54388 (x_source + width, y_source + height) to handle all the 54389 cases. 54390 54391 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54392 54393 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54394 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54395 54396commit 553910d08b93ef80cc25fcbfd7876726778bc655 54397Author: Junyan He <junyan.he@linux.intel.com> 54398Date: Tue May 15 10:07:55 2012 +0800 54399 54400 Fix the problem of vertical and horizontal case error in linear gradient. 54401 54402 1. The vertical and horizontal judgement in linear 54403 gradient have problem when p1 point and p2 point 54404 distance is very small but the gradient pict have a 54405 transform matrix which will convert the X Y coordinates 54406 to small values. So the judgement is not suitable. 54407 Because this judgement's purpose is to assure the 54408 divisor not to be zero, so we simply it to enter 54409 horizontal judgement when p1 and p2's Y is same. 54410 Vertical case is deleted. 2. Delete the unused p1 p2 54411 uniform variable. 54412 54413 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54414 54415 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54416 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54417 54418commit 41aa93c393c8760aae357725db63fa4a8f798557 54419Author: Junyan He <junyan.he@linux.intel.com> 54420Date: Tue May 15 10:07:46 2012 +0800 54421 54422 Fix the problem of set the same stop several times. 54423 54424 Some gradient set the stops at the same position, for 54425 example: firstly 0.5 to red color and then set 0.5 to 54426 blue. This kind of setting will cause the shader work 54427 not correctly because the percentage caculating need to 54428 use the stop[i] - stop[i-1] as dividend. The previous 54429 patch we just kill some stop if the distance between 54430 them is 0. But this cause the problem that the color 54431 for next stop is wrong. We now modify to handle it in 54432 the shader to avoid the 0 as dividend. 54433 54434 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54435 54436 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54437 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54438 54439commit 09de37ec1c0543c8073f934274c84b3b7d5f31ae 54440Author: Junyan He <junyan.he@linux.intel.com> 54441Date: Tue May 15 10:07:35 2012 +0800 54442 54443 Fix a bugy macro definition. 54444 54445 The macro like "#define LINEAR_SMALL_STOPS 6 + 2" causes 54446 the problem. When use it to define like "GLfloat 54447 stop_colors_st[LINEAR_SMALL_STOPS*4];" The array is 54448 small than what we supposed it to be. Cause memory 54449 corruption problem and cause the bug of render wrong 54450 result. Fix it. 54451 54452 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54453 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54454 54455commit d900f553c2bb8d6e01529524fb7125918291c406 54456Author: Junyan He <junyan.he@linux.intel.com> 54457Date: Tue May 15 10:07:24 2012 +0800 54458 54459 Extract the gradient related code out. 54460 54461 1. Extract the logic of gradient from the glamor_render.c 54462 to the file glamor_gradient.c. 54463 2. Modify the logic of gradient pixmap gl draw. Use the 54464 logic like composite before, but the gradient always just 54465 have one rect to render, so no need to set the VB and EB, 54466 replace it with just call glDrawArrays. 3.Kill all the 54467 warning in glamor_render.c 54468 54469 Reviewed-by: Zhigang Gong<zhigang.gong@linux.intel.com> 54470 54471 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54472 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54473 54474commit 81692804645394b43832c8ec586e8e67f3e4f606 54475Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54476Date: Tue May 15 15:25:41 2012 +0800 54477 54478 glamor_set_destination_pixmap_priv_nc: set drawable's width x height. 54479 54480 Previous implementation set the whole fbo's width and height as the 54481 viewpoint. This may increase the numerical error as we may only has 54482 a partial region as the valid pixmap. So add a new marco 54483 pixmap_priv_get_dest_scale to get proper scale factor for the 54484 destination pixmap. For the source/mask pixmap, we still need to 54485 consider the whole fbo's size. 54486 54487 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54488 54489commit 7f55e48499ea7bed73cb1adeac00c480263583f8 54490Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54491Date: Tue May 15 10:52:37 2012 +0800 54492 54493 Remove the texture cache code. 54494 54495 Caching texture objects is not necessary based on previous testing. 54496 To keep the code simple, we remove it. 54497 54498 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54499 54500commit c5b3c2cedc8b2e486b1e3727f288c42869310387 54501Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54502Date: Tue May 15 10:42:41 2012 +0800 54503 54504 Added strict warning flags to CFLAGS. 54505 54506 We miss the strict warning flags for a long time, now add it back. 54507 This commit also fixed most of the warnings after enable the strict 54508 flags. 54509 54510 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54511 54512commit 6839996b0b0fe5f4e6ef28d1dfe527092d60d28a 54513Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54514Date: Sat Apr 28 18:23:09 2012 +0800 54515 54516 We should not call gradient finalization code if we disable it. 54517 54518 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54519 54520commit 1035fc72b9017eb2466760133ca7bbc9155c8c46 54521Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54522Date: Sat Apr 28 14:55:34 2012 +0800 54523 54524 Fixed all unused variables warnings. 54525 54526 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54527 54528commit 33e11cd6149294060269ed693de67b135868e094 54529Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54530Date: Sat Apr 28 14:54:38 2012 +0800 54531 54532 Fixed an uninitialized problem at gradient shader functions. 54533 54534 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54535 54536commit c0f75c657f36642faea4ff8c51f7e4f6971c3d19 54537Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54538Date: Sat Apr 28 14:51:27 2012 +0800 54539 54540 Fixed one typo bug when fixup a mask picture. 54541 54542 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54543 54544commit 5c1f15fac26b86b6cb73776db1a644d6af570da7 54545Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54546Date: Sat Apr 28 14:13:47 2012 +0800 54547 54548 Added some copyright and author information. 54549 54550 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54551 54552commit 0d846d95699fadcddcc77b8d6e432e969467dab2 54553Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54554Date: Sat Apr 28 13:43:39 2012 +0800 54555 54556 Added --enable-debug configuration option. 54557 54558 For release version, we disable asserts. 54559 54560 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54561 54562commit 503f8ec1a670e73f41314a5e94cdde8782d7cbab 54563Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54564Date: Sat Apr 28 13:42:48 2012 +0800 54565 54566 Remove unecessary header file. 54567 54568 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54569 54570commit 9dfd10dc75efdf4c26b7ff46b55e4a2d2453803b 54571Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54572Date: Fri Apr 27 15:34:15 2012 +0800 54573 54574 glamor_render: Fix the repeat none for GLES2. 54575 54576 As GLES2 doesn't support clamp to the border, we have to 54577 handle it seprately from the normal case. 54578 54579 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54580 54581commit 9fcd123aed80430b220fc4141eaa3723d7cb611e 54582Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54583Date: Thu Apr 26 20:35:09 2012 +0800 54584 54585 glamor_blockhandler: Don't do glFinish every time. 54586 54587 To do glfinish every time bring some performance overhead. 54588 54589 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54590 54591commit 1f83411c9a69a44944fd8afe0c167cece60c1ecb 54592Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54593Date: Thu Apr 26 19:59:58 2012 +0800 54594 54595 glamor_copyarea: Return earlier if have zero nbox. 54596 54597 Almost all callers will check whether the regions is empty 54598 before call to this internal API, but it seems the 54599 glamor_composite_with_copy may call into here with a zero 54600 nbox. A little weird, as the miComputeCompositeRegion return 54601 a Non-NULL, but the region is empty. 54602 54603 Also remove a unecessary glflush. 54604 54605 So let's check it here. 54606 54607 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54608 54609commit 20cbaa61cdca41e61526a57f13475cb31e17e5dd 54610Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54611Date: Thu Apr 26 19:57:21 2012 +0800 54612 54613 glamor_render: Have to use eaxct size pixmap for transformation. 54614 54615 Use partial texture as the pixmap for the transformation 54616 source/mask may introduce extra errors. have to use 54617 eaxct size. 54618 54619 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54620 54621commit 6e50ee9c108e6d7ce4ebfcd08cfc97896e8e194e 54622Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54623Date: Thu Apr 26 18:31:07 2012 +0800 54624 54625 glamor_fbo: Added a threshold value for the fbo cache pool. 54626 54627 Currently set it to 256MB. If cache pool watermark increases 54628 to this value, then don't push any fbo to this pool, will purge 54629 the fbo directly. 54630 54631 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54632 54633commit 540846204cce9239233c5608a29bfe0607d77e44 54634Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54635Date: Thu Apr 26 18:28:17 2012 +0800 54636 54637 Fixed a1 bug. 54638 54639 It seems that mesa has bugs when uploading bitmap to texture. 54640 We switch to convert bitmap to a8 format and then upload the 54641 a8 texture. 54642 54643 Also added a helper routine to dump 1bpp pixmap. 54644 54645 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54646 54647commit 9f53cc1c3369fc61630b238f1b347a92fabf5a5a 54648Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54649Date: Wed Apr 25 22:35:12 2012 +0800 54650 54651 glamor_render.c: Fixed repeatPad and repeatRelect. 54652 54653 We should use difference calculation for these two repeat mode 54654 when we are a sub region within one texture. 54655 54656 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54657 54658commit 67cf3838e4acd788b0ce413dcbe9896e2ca20e56 54659Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54660Date: Wed Apr 25 19:50:57 2012 +0800 54661 54662 gradient: Don't need fixup flag when creating pixmap. 54663 54664 Gradient can use a larger texture/fbo directly, don't need 54665 an eaxct size texture. 54666 54667 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54668 54669commit 8a85071edbd90780b286fa4b19205540fb276815 54670Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54671Date: Wed Apr 25 18:58:16 2012 +0800 54672 54673 glamor_copyarea: Don't access a DRM only pixmap. 54674 54675 As EGL image/gbm only support ARGB8888 image, we don't support 54676 other format. We may change the way to use gbm directly latter. 54677 But now, we have to face this limitation, and thus if a client 54678 create a 16bpp drawable, and call get texture from pixmap then 54679 a copy to here may occur and thus we have to force retur a TRUE 54680 without do nothing. 54681 54682 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54683 54684commit 0b6867dddbed186c46048c610299ac0dde6a9ef0 54685Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54686Date: Wed Apr 25 13:54:40 2012 +0800 54687 54688 Disable A8 texture format for GLES2. 54689 54690 As PVR's GLES2 implementation doesn't support A8 texture as 54691 rendering target, we disable it for now. 54692 54693 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54694 54695commit 6b664dda69afa6c47ca083739093fa15fc674380 54696Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54697Date: Wed Apr 25 13:51:58 2012 +0800 54698 54699 gradient: Disable gradient for gles2. 54700 54701 As PVR glsl compiler seems doesn't support external fragment 54702 function, and fails at compile gradient shader. Disable it 54703 for now. We may need to modify gradient shader to don't use 54704 external function. 54705 54706 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54707 54708commit 686a322c76fa20ed45d5bbfc9742024300e83e7d 54709Author: Junyan He <junyan.he@linux.intel.com> 54710Date: Wed Apr 18 08:04:26 2012 +0800 54711 54712 Fix the bug caused by gradient picture set the stops at the same percentage. 54713 54714 Fix the bug caused by gradient picture set the stops at 54715 the same percentage. The (stops[i] - stops[i-1]) will 54716 be used as divisor in the shader, which will cause 54717 problem. We just keep the later one if stops[i] == 54718 stops[i-1]. 54719 54720 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 54721 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54722 54723commit 3d96929596fd3a6da41aab5cb9d7fb1cf28b2a03 54724Author: Junyan He <junyan.he@linux.intel.com> 54725Date: Wed Apr 25 14:25:39 2012 +0800 54726 54727 Fix the problem of memory leak in gradient pixmap generating. 54728 54729 Fix the problem of memory leak in gradient pixmap 54730 generating. The problem caused by we do not call 54731 glDeleteShader when destroy a shader program. This patch 54732 will split the gradient pixmap generating to three 54733 category. If nstops < 6, we will use the no array version 54734 of the shader, which has the best performance. Else if 54735 nstops < 16, we use array version of the shader, which is 54736 compiled and linked at screen init stage. Else if nstops > 54737 16, we dynamically create a new shader program, and this 54738 program will be cached until bigger nstops. 54739 54740 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54741 54742commit 05da99106b81465488c9879cfd709fd4f0c7b9e5 54743Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54744Date: Wed Apr 25 11:59:47 2012 +0800 54745 54746 glamor_putimage: Optimize for direct uploading and fallback path. 54747 54748 This commit optimize two cases: 54749 1. When the clip contains the whole area, we can directly upload 54750 the texel data to the pixmap, and don't need to do one extra 54751 clipped copy. 54752 54753 2. At fallback path, we don't read back the whole pixmap, just 54754 need a sub region. 54755 54756 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54757 54758commit ea70ebe0ac9fe5b3d0ad553f9aeb4d2829bf9a62 54759Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54760Date: Wed Apr 25 11:56:43 2012 +0800 54761 54762 Fixed one potential texture size mismatch problem. 54763 54764 There are two cases which we may use a wrong texture size. 54765 1. A pixmap is modified by the client side after it created 54766 it. Then the pixmap's width may mismatch the original fbo/tex's 54767 size. Thus we need to check this condition when preparing 54768 upload the pixmap. 54769 54770 2. We provide two API to download/upload sub region of a 54771 textured pixmap. The caller may pass in a larger width then 54772 the original pixmap's size, this may happen at putimage 54773 and setspans. We need to validate the width and height 54774 when do the downloading/uploading. 54775 54776 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54777 54778commit 08e8c00fe6a21741ff9f38652c2b9fd2310f1ce5 54779Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54780Date: Wed Apr 25 09:27:06 2012 +0800 54781 54782 glamor_getimage: Don't fallback to miGetImage. 54783 54784 As miGetImage is very inefficient, we don't fallback to it. 54785 If the format is not ZPixmap, we download the required sub- 54786 region, and then call fbGetImage to do the conversion. 54787 This way is much faster than previous. 54788 54789 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54790 54791commit 9bcddff93b79fd8978426d9832a5edd60ac410c0 54792Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54793Date: Fri Apr 13 18:15:49 2012 +0800 54794 54795 pending_op: Remove the pending operations handling. 54796 54797 We have disabled this feature for a long time, and previous 54798 testing shows that this(pending fill) will not bring observed 54799 performance gain. Now remove it. 54800 54801 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54802 54803commit 1761768f49a356f50645da53305e6b4bdef5c5f4 54804Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54805Date: Sun Apr 15 23:50:09 2012 +0800 54806 54807 glamor_upload_pixmap: Use glTexImage2D for a fully update. 54808 54809 Currently, intel's mesa dri driver will not check pbo for 54810 a TexSubImage2D. So we use glTexImage2D if we are a fully 54811 updating. 54812 54813 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54814 54815commit 2806f1eaced851e4c88055c53d706f7beef8b555 54816Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54817Date: Sun Apr 15 23:49:38 2012 +0800 54818 54819 glamor_setspans: Reuse glamor_upload_sub_pixmap. 54820 54821 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54822 54823commit e15bc1207480d7d198862861d40af58903b4d0f0 54824Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54825Date: Sun Apr 15 23:18:47 2012 +0800 54826 54827 code clean up. 54828 54829 Remove unused variables. 54830 54831 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54832 54833commit 65c5605c9693c8d30e597ac029be936495f23927 54834Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54835Date: Sun Apr 15 23:16:51 2012 +0800 54836 54837 glamor_getspans: Reuse glamor_download_sub_pixmap. 54838 54839 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54840 54841commit 68a5cc6f371e1f07ce176e154331dafaf5d9ef0e 54842Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54843Date: Sun Apr 15 19:46:09 2012 +0800 54844 54845 glamor_render: Don't download whole picture when fallback. 54846 54847 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54848 54849commit e38eb675321dce1bbc39cbd177a6398de567dd79 54850Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54851Date: Sun Apr 15 19:43:44 2012 +0800 54852 54853 glamor_put_sub_pixmap: Change to use glamor_upload_sub_pixmap. 54854 54855 As the pixmap may be attached to a picture, we need to use 54856 glamor_upload_sub_pixmap to process it. glamor_copy_n_to_n 54857 will not consider the picture case. 54858 54859 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54860 54861commit ff3d2c796363ea603ab92995091a967a3f8636d7 54862Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54863Date: Sun Apr 15 17:09:06 2012 +0800 54864 54865 Fixed a stride problem for textured_drm pixmap. 54866 54867 As a textured_drm pixmap has a drm bo attached to it, and 54868 it's the DDX layer to set it stride value. In some case, 54869 the stride value is not equal to PixmapBytePad(w, depth) 54870 which is used within glamor. 54871 54872 Then if it is the case, we have two choice, one is to set 54873 the GL_PACK_ROW_LENGTH/GL_UNPACK_ROW_LENGTH when we need 54874 to download or upload the pixmap. The other option is to 54875 change the pixmap's devKind to match the one glamor is using 54876 when downloading the pixmap, and restore it to the drm stride 54877 after uploading the pixmap. 54878 54879 We choose the 2nd option, as GLES doesn't support the first 54880 method. 54881 54882 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54883 54884commit 70b71718e737872dfdbf4f6c7285d4260a099d17 54885Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54886Date: Sun Apr 15 14:36:09 2012 +0800 54887 54888 glamor_putimage: Reuse copy area to do the clipped copy. 54889 54890 If no clip set, we load the bits to the pixmap directly. 54891 Otherwise, load the bits to a temporary pixmap and call 54892 glamor_copy_area to do the clipped copy. 54893 54894 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54895 54896commit e1be714312df8d596f6be268cb8a4e390e634c36 54897Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54898Date: Sun Apr 15 10:30:02 2012 +0800 54899 54900 Fixed a unbalanced glamor_put_dispatch. 54901 54902 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54903 54904commit bd53e24dc32206dc978d3dc2408e1832f5a803e7 54905Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54906Date: Sun Apr 15 22:47:48 2012 +0800 54907 54908 glamor_pixmap_priv: Always return a valid private pixmap. 54909 54910 If a pixmap doesn't have a private, then set its type to 54911 GLAMOR_MEMORY, and thus it will get a valid private. 54912 54913 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54914 54915commit 420af44a3aaefd4848aa24a28c330cab36049078 54916Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54917Date: Sat Apr 14 22:40:48 2012 +0800 54918 54919 Don't need to set GL_PACK_ROW_LENGTH/GL_UNPACK_ROW_LENGTH. 54920 54921 We already adjust the stride of the pixmap, and keep the alignment 54922 as 4 should be ok to let the GL/GLES match the stride. 54923 54924 Previous version has a unbalanced PACK ROW length seting, and is 54925 buggy, now fixed it. 54926 54927 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54928 54929commit 18d69fb0142088f7df230ec876c7ce2b55a41ad9 54930Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54931Date: Fri Apr 13 13:47:05 2012 +0800 54932 54933 glamor_gl: Use GL_ALPHA for depth 8 pixmap. 54934 54935 Use GL_RGBA to represent a8 pixmap is not efficient. 54936 54937 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54938 54939commit 428f2a3f58a91300835aaa083ac0b291ce394dd0 54940Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54941Date: Fri Apr 13 13:45:42 2012 +0800 54942 54943 glamor_pixmap_ensure_fbo: Should allocate tex if we don't have one. 54944 54945 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54946 54947commit cf0e206a0ff6538eb286c06098023b9e29b00c74 54948Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54949Date: Wed Apr 11 18:59:27 2012 +0800 54950 54951 glamor_polylines: Don't fallback for non-solid fill. 54952 54953 As glamor_fill/fbFill will handle non-solid fill correctly. 54954 We don't fallback it here. 54955 54956 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54957 54958commit b5bd9a2d902d44834fc43199167d7dee71c9b709 54959Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54960Date: Wed Apr 11 18:57:33 2012 +0800 54961 54962 glamor_upload/download: fix 1bpp bug. 54963 54964 For A1 to A8's conversion, the stride is different for the 54965 source and destination. Previous implementation use the same 54966 stride, and may allocate less memory than required. Thus may 54967 crash the server when uploading a A1 pixmap. Now fix it. 54968 54969 Tested-by: Peng Li <peng.li@intel.com> 54970 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54971 54972commit b0e91f0f5a3a4f74800b89ed4003d300ae138151 54973Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54974Date: Tue Apr 10 20:50:14 2012 +0800 54975 54976 glamor_pixmap_upload_texture: Support to upload a sub region of data. 54977 54978 Just as the downloading side, we can upload an sub region data to 54979 a pixmap's specified region. The data could be in memory or in a 54980 pbo buffer. 54981 54982 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54983 54984commit 3061f348ca2f05d88ca2391e1ad81ce8216d69f2 54985Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54986Date: Tue Apr 10 10:40:17 2012 +0800 54987 54988 glamor_getimage: Use glamor_download_sub_pixmap_to_cpu to get image. 54989 54990 Reduce the duplicate logic. 54991 54992 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 54993 54994commit 3a91f169122ea0556f628f11fb473b0feed7e44b 54995Author: Zhigang Gong <zhigang.gong@linux.intel.com> 54996Date: Tue Apr 10 10:19:30 2012 +0800 54997 54998 glamor_polyfillrect: Fixed a potential bug if fallback at glamor_fill. 54999 55000 We should advance the prect after we successfully excuted the 55001 glamor_fill. And if failed, we need to add the failed 1 box 55002 back to nbox. 55003 55004 Although, this bug will never happen currently, as glamor_fill 55005 will never fallback. 55006 55007 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55008 55009commit 1f657f72cacaa1170e7e55f4b5149c69492db6a5 55010Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55011Date: Mon Apr 9 20:20:45 2012 +0800 55012 55013 glamor_polyfillrect: Optimize fallback path. 55014 55015 Download/upload required region only. 55016 55017 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55018 55019commit cea0fe3e1f49fc8d86f0cc653c8089f1ea2c9f1f 55020Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55021Date: Mon Apr 9 20:16:07 2012 +0800 55022 55023 fallback_optimize: Prepare for downloading/uploading subregion. 55024 55025 Introduced two function glamor_get_sub_pixmap/glamor_put_sub_pixmap, 55026 can easily used to get and put sub region of a big textured pixmap. 55027 And it can use pbo if possible. 55028 55029 To support download a big textured pixmap's sub region to another 55030 pixmap's pbo, we introduce a new type of pixmap GLAMOR_MEMORY_MAP. 55031 This type of pixmap has a valid devPrivate.ptr pointer, and that 55032 pointer points to a pbo mapped address. 55033 55034 Now, we are ready to refine those 55035 glamor_prepare_access/glamor_finish_access pairs. 55036 55037 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55038 55039commit d9dfc3d795b7e567d53cfeed61126164be36e233 55040Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55041Date: Mon Apr 9 16:39:42 2012 +0800 55042 55043 glamor_download_sub_pixmap_to_cpu: New function to download subregion. 55044 55045 Prepare to optimize the fallback path. We choose the important 55046 rendering pathes to optimzie it by using shader. For other pathes, 55047 we have to fallback. We may continue to optimize more pathes in 55048 the future, but now we have to face those fallbacks. 55049 55050 The original fallback is very slow and will download/upload the whole 55051 pixmap. From this commit, I will refine it to just download/upload 55052 needed part. 55053 55054 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55055 55056commit d96226ac6f34aa61fc00ad15ef58c1ed1253160e 55057Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55058Date: Mon Apr 9 15:57:05 2012 +0800 55059 55060 glamor_es2_pixmap_read_prepare: Just prepare the required region. 55061 55062 Don't need to prepare the whole source pixmap. 55063 55064 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55065 55066commit 3dbdd40c6ce4203619f2a3492029a444bb9217af 55067Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55068Date: Mon Apr 9 10:00:57 2012 +0800 55069 55070 glamor_color_convert: Let the caller to provide destination buffer. 55071 55072 As we don't need to allocate new buffer when downloading pixmap 55073 to CPU, we change the prototype of the color converting function 55074 and let the caller to provide the buffer to hold the result. 55075 55076 All the color conversion function supports store the result 55077 just at the same place of the source. 55078 55079 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55080 55081commit 4dc6d4e84b4904540d7701cfc88a9c945464f833 55082Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55083Date: Fri Apr 6 21:20:15 2012 +0800 55084 55085 glyphblt/polyops: Use miFunctions by default. 55086 55087 Calling to miFunctions give some opportunities to jump to 55088 accelerated path, so we switch to call miFunctions rather 55089 than fallback to fbFunctions directly. 55090 55091commit 49e3b44aa813c98c05fcb10c19882e10d751580a 55092Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55093Date: Fri Apr 6 21:17:46 2012 +0800 55094 55095 glamor_set_alu: Added GXclear support at glamor_solid. 55096 55097 We don't need to issue the glamor_fallback at the glamor_set_alu 55098 routine, as the caller may support GXclear or other most frequent 55099 Ops. Leave it to the caller to determine fallback or not. 55100 55101 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55102 55103commit 3b8b2c77fc4449c7b63fd2597f73562b33dc1722 55104Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55105Date: Fri Apr 6 21:15:12 2012 +0800 55106 55107 getimage: Enable getimage by default. 55108 55109 Fixed one bug when calculate the coords, should consider the 55110 drawable's x and y. Now enable it by default. Most of the time, 55111 it should be more efficient than miGetImage. 55112 55113 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55114 55115commit c6ce44d88134115b42edc76e1ee961b57bae86ff 55116Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55117Date: Fri Apr 6 17:01:57 2012 +0800 55118 55119 render: Enable more componentAlpha support. 55120 55121 Actually only PictOpAtop,PictOpAtopReverse and PictOpXor 55122 can't be implemented by using single source blending. 55123 All the other can be easily support. Slightly change 55124 the code to support them. Consider those three Ops 55125 are not frequenly used in real application. We simply 55126 fallback them currently. 55127 55128 PictOpAtop: s*mask*dst.a + (1 - s.a*mask)*dst 55129 PictOpAtopReverse: s*mask*(1 - dst.a) + dst *s.a*mask 55130 PictOpXor: s*mask*(1 - dst.a) + dst * (1 - s.a*mask) 55131 55132 The two oprands in the above three ops are all reated to dst and 55133 the blend factors are not constant (0 or 1), it's hardly to 55134 convert it to single source blend. 55135 55136 Now, the rendercheck is runing more smoothly. 55137 55138 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55139 55140commit 3e9c35bdcbdb96a67c9f2a1ea76c382aaacca7e9 55141Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55142Date: Fri Apr 6 11:02:36 2012 +0800 55143 55144 glamor_set_alu: Fallback for non GXcopy ops with GLES2. 55145 55146 As GLES2 doesn't support LogiOps, we have to fallback 55147 here. GLES2 programing guide's statement is as below: 55148 55149 "In addition, LogicOp is removed as it is very 55150 infrequently used by applications and the OpenGL ES 55151 working group did not get requests from independent 55152 software vendors (ISVs) to support this feature in 55153 OpenGL ES 2.0." 55154 55155 So, I think, fallback here may not a big deal ;). 55156 55157 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55158 55159commit 1a238e89f39fd5aeaf6975399971123cd3e15f24 55160Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55161Date: Thu Apr 5 22:39:41 2012 +0800 55162 55163 glamor_putimage: Reuse the function in pixmap.c to do the uploading. 55164 55165 We reuse glamor_upload_bits_to_pixmap_texture to do the 55166 data uploading to texture in putimage. Besides to avoid 55167 duplicate code, this also fixed the potential problem 55168 when the data format need extra reversion which is not 55169 supported by the finish shader, as 55170 glamor_upload_bits_to_pixmap_texture will handle all 55171 conditions. 55172 55173 Tested-by: Junyan He <junyan.he@linux.intel.com> 55174 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55175 55176commit 0650c7d4be6d4c21510c953543599aea32780f24 55177Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55178Date: Thu Apr 5 22:31:44 2012 +0800 55179 55180 gles2: Added 1555/2101010 formats support. 55181 55182 Added color conversion code to support 1555/2101010 55183 formats,now gles2 can pass the render check with all 55184 formats. 55185 55186 We use 5551 to represent 1555, and do the revertion 55187 if downloading/uploading is needed. 55188 55189 For 2101010, as gles2 doesn't support reading the 55190 identical formats. We have to use 8888 to represent, 55191 thus we may introduce some accurate problem. But anyway, 55192 we can pass the error checking in render check, so that 55193 may not be a big problem. 55194 55195 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55196 55197commit 3add3750658107bd18592a8672caa8bed0c8931c 55198Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55199Date: Thu Apr 5 21:53:44 2012 +0800 55200 55201 gles2: Fixed color conversion for the formats except 1555 and 2101010. 55202 55203 This patch fixed two major problems when we do the color convesion with 55204 GLES2. 55205 55206 1. lack of necessary formats in FBO pool. 55207 GLES2 has three different possible texture formats, GL_RGBA, 55208 GL_BGRA and GL_ALPHA. Previous implementation only has one bucket 55209 for all the three formats which may reuse a incorrect texture format 55210 when do the cache lookup. After this fix, we can enable fbo safely 55211 when running with GLES2. 55212 55213 2. Refine the format matching method in 55214 glamor_get_tex_format_type_from_pictformat. 55215 If both revertion and swap_rb are needed, for example use GL_RGBA 55216 to represent PICT_b8g8r8a8. Then the downloading and uploading should 55217 be handled differently. 55218 55219 The picture's format is PICT_b8g8r8a8, 55220 Then the expecting color layout is as below (little endian): 55221 0 1 2 3 : address 55222 a r g b 55223 55224 Now the in GLES2 the supported color format is GL_RGBA, type is 55225 GL_UNSIGNED_TYPE, then we need to shuffle the fragment 55226 color as : 55227 frag_color = sample(texture).argb; 55228 before we use glReadPixel to get it back. 55229 55230 For the uploading process, the shuffle is a revert shuffle. 55231 We still use GL_RGBA, GL_UNSIGNED_BYTE to upload the color 55232 to a texture, then let's see 55233 0 1 2 3 : address 55234 a r g b : correct colors 55235 R G B A : GL_RGBA with GL_UNSIGNED_BYTE 55236 55237 Now we need to shuffle again, the mapping rule is 55238 r = G, g = B, b = A, a = R. Then the uploading shuffle is as 55239 below: 55240 frag_color = sample(texture).gbar; 55241 55242 After this commit, gles2 version can pass render check with all 55243 the formats except those 1555/2101010. 55244 55245 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55246 55247commit 55fdc7b196904a4e537f429d06d36081a0c9a60d 55248Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55249Date: Thu Apr 5 20:27:35 2012 +0800 55250 55251 glamor_utils: Added debug function to dump depth 15/16 pixmap. 55252 55253 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55254 55255commit 57e29ebdc13d0ef145e4c70c1406af3229a92ad8 55256Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55257Date: Fri Apr 6 11:14:21 2012 +0800 55258 55259 glamor_render: Disable gradient shader conversion due to bug. 55260 55261 I found when enable the gradient shader, the firefox's tab's 55262 background has incorrect rendering result. 55263 55264 Need furthr investigation, for now, just disable it. 55265 55266 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55267 55268commit 7036cfdd0d29f4e13d432893a4f386cd3d632de7 55269Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55270Date: Fri Mar 16 16:42:46 2012 +0800 55271 55272 glamor_fbo: Added one macro to disable fbo cache. 55273 55274 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55275 55276commit 94186db527b33b2623eb26a7f1ae0c4c0fca66a4 55277Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55278Date: Mon Mar 26 19:03:20 2012 +0800 55279 55280 glamor_fill: Should restore alu to GXcopy. 55281 55282 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55283 55284commit 1f4486c10bb2201830af251eede8cc4dbb857953 55285Author: Junyan He <junyan.he@linux.intel.com> 55286Date: Fri Mar 23 04:06:06 2012 +0800 55287 55288 Add the feature for radial gradient using shader. 55289 55290 Add the feature for radial gradient using shader. The 55291 transform matrix and the 4 type of repeat mode are 55292 supported. Less than 2/255 difference for every color 55293 component comparing to pixman's result. Extract the 55294 common logic of linear and radial's to another shader. 55295 55296 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 55297 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55298 55299 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55300 55301commit 1026327cdc90bc0c801a493b7b76b3efc5f33fe2 55302Author: Junyan He <junyan.he@linux.intel.com> 55303Date: Fri Mar 23 04:05:56 2012 +0800 55304 55305 Add the feature of generating linear gradient picture by using shader. 55306 55307 Add the feature of generating linear gradient picture 55308 by using shader. This logic will replace the original 55309 linear gradient picture generating manner in glamor 55310 which firstly use pixman and then upload it to GPU. 55311 Compare it to the result generated by pixman, the 55312 difference of each color component of each pixel is 55313 normally 0, sometimes 1/255, and 2/255 at most. The 55314 pixman use fixed-point but shader use float-point, so may have 55315 difference. The feature of transform matrix and 4 types 55316 of repeat modes have been supported. The array usage in 55317 shader seems slow, so use 8 uniform variables to avoid 55318 using array when stops number is not very big. This 55319 make code look verbose but the performance improved a 55320 lot. 55321 55322 We still have slightly performance regression compare to 55323 original pixman version. There are one further optimization 55324 opportunity which is to merge the gradient pixmap generation 55325 and the latter compositing into one shader, then we don't need 55326 to generate the extra texture, we can use the gradient value 55327 directly at the compositing shader. Hope that can beat pixman 55328 version. Will do that latter. 55329 55330 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 55331 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55332 55333 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55334 55335commit ccf5d7f52bae664f90d2c33c9fcff099a820575f 55336Author: Junyan He <junyan.he@linux.intel.com> 55337Date: Fri Mar 23 04:05:48 2012 +0800 55338 55339 Prepare for modification of gradient using shader. 55340 55341 Prepare for modification of gradient using shader. The 55342 gradient pixmaps now is generated by pixman and we will 55343 replace them with shader. Add structure fields and 55344 dispatch functions which will be needed. Some auxiliary 55345 macro for vertex convert. 55346 55347 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 55348 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55349 55350 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55351 55352commit a57bf66d492a0eec7b1f2a08b2f424835694e3fb 55353Author: Junyan He <junyan.he@linux.intel.com> 55354Date: Mon Mar 26 13:51:13 2012 +0800 55355 55356 glamor_utils: Add some assistant functions to compare pixmaps/pictures. 55357 55358 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 55359 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55360 55361 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55362 55363commit cd75e85ff36827b8438d965750faab0615d01e86 55364Author: Junyan He <junyan.he@linux.intel.com> 55365Date: Mon Mar 5 08:24:20 2012 +0800 55366 55367 Fixup For list.h change in xorg 55368 55369 Because the file list.h in xorg/include has changed the 55370 functions and struct names, adding xorg_ prefix before 55371 the original name. So Modify glamor_screen_private 55372 struct and the code which use list's functions in 55373 glamor_fbo.c. We hack at glamor_priv.h avoid the 55374 compile error when using old version xserver header 55375 file. 55376 55377 Signed-off-by: Junyan He <junyan.he@linux.intel.com> 55378 55379commit 213285f2b8c6578e01783a77d954ccc3ec313663 55380Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55381Date: Fri Feb 17 16:56:05 2012 +0800 55382 55383 For DRI swap buffers. 55384 55385 This commit added two APIs to support the DRI swap buffer. 55386 one is glamor_egl_exchange_buffers() which can swap two 55387 pixmaps' underlying KHRimages/fbos/texs. The DDX layer should 55388 exchange the DRM bos to make them consistent to each other. 55389 55390 Another API is glamor_egl_create_textured_screen_ext(), which 55391 extent one more parameters to track the DDX layer's back pixmap 55392 pointer. This is for the triple buffer support. When using triple 55393 buffer, the DDX layer will keep a back pixmap rather then the 55394 front pixmap and the pixmap used by the DRI2 client. And during 55395 the closing screen stage, we have to dereference all the back 55396 pixmap's glamor resources. Thus we have to extent this API to 55397 register it when create new screen. 55398 55399 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55400 55401commit 8012b030c3144b02af036107179c5b4c94567292 55402Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55403Date: Fri Feb 17 16:26:59 2012 +0800 55404 55405 glamor_copyarea: Don't use GL_CLAMP_TO_BORDER when GLES2 enabled. 55406 55407 We may need to modify all the shader to handle GL_CLAMP_TO_BORDER 55408 when using GLES2. XXX, for now, we just ignore them. 55409 55410 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55411 55412commit 5ccf721d386a62f411d77bbc9a142e0c395162b3 55413Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55414Date: Tue Feb 14 17:39:11 2012 +0800 55415 55416 glamor_fbo: Fix a bug when create No gl FBO pixmap. 55417 55418 Need to get format before goto create a new fbo. 55419 55420 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55421 55422commit ce634e84d4bb559f01203653c5ffd6397f4b0366 55423Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55424Date: Sun Feb 12 09:18:51 2012 +0800 55425 55426 glamor_render: Only recalculate texture for repeat case. 55427 55428 Slightly optimize the fragment shader, as if we are not 55429 repeat case and not exceed the valid texture range, then 55430 we don't need to recalculate the coords. 55431 55432 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55433 55434commit 53387728ddc0c871821b68d728bb5f96a53ba227 55435Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55436Date: Fri Feb 10 17:40:37 2012 +0800 55437 55438 glamor_tile/composite: Modify fs to re-calculate texture coords. 55439 55440 Then we don't need to fixup the larger pixmap to the exact 55441 size, just need to let the shader to re-calculate the correct 55442 texture coords. 55443 55444 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55445 55446commit 556adfa6b90f4c1ef12635cf78fa0bba046cf123 55447Author: Chris Wilson <chris@chris-wilson.co.uk> 55448Date: Fri Feb 10 12:54:44 2012 +0000 55449 55450 Fixup glx support 55451 55452 Renaming glamor_priv->dispatch and wrapping the access to 55453 the dispatch table with a function that also ensured the 55454 context was bound. 55455 55456 dispatch = glamor_get_dispatch(glamor_priv); 55457 ... 55458 glamor_put_dispatch(glamor_priv); 55459 55460 So that we catch all places where we attempt to call into GL withouta 55461 context. As an optimisation we can then do glamor_get_context(); 55462 glamor_put_context() around the rendering entry points to reduce the 55463 frequency of having to restore the old context. (Along with allowing 55464 the context to be recursively acquired and making the old context part of 55465 the glamor_egl state.) 55466 55467 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55468 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55469 55470commit 430bc16ca03b3ea00255a4045c8e9fd11aea95ad 55471Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55472Date: Fri Feb 10 16:04:07 2012 +0800 55473 55474 GLX: Enable glx support. 55475 55476 If we are using MESA as our GL library, then both xserver's 55477 GLX and glamor are link to the same library. As xserver's 55478 GLX has its own _glapi_get/set_context/dispatch etc, and it 55479 is a simplified version derived from mesa thus is not 55480 sufficient for mesa/egl's dri loader which is used by glamor. 55481 55482 Then if glx module is loaded before glamoregl module, the 55483 initialization of mesa/egl/opengl will not be correct, and 55484 will fail at a very early stage, most likely fail to map 55485 the element buffer. 55486 55487 Two methodis to fix this problem, first is to modify the xserver's 55488 glx's glapi.c to fit mesa's requirement. The second is to put 55489 a glamor.conf as below, to the system's xorg.conf path. 55490 55491 Section "Module" 55492 Load "glamoregl" 55493 EndSection 55494 55495 Then glamor will be loaded firstly, and the mesa's libglapi.so 55496 will be used. As current xserver's dispatch table is the same 55497 as mesa's, then the glx's dri loader can work without problem. 55498 55499 We took the second method as it don't need any change to xorg.:) 55500 Although this is not a graceful implementation as it depends 55501 on the xserver's dispatch table and the mesa's dispatch table 55502 is the same and the context set and get is using the same method. 55503 Anyway it works. 55504 55505 As by default, xserver will enable GLX_USE_TLS. But mesa will not 55506 enable it, you may need to enable that when build mesa. 55507 55508 Three pre-requirements to make this glamor version work: 55509 55510 0. Make sure xserver has commit 66e603, if not please pull the latest 55511 master branch. 55512 1. Rebuild mesa by enable GLX_USE_TLS. 55513 2. Put the glamor.conf to your system's xorg.conf path and make sure 55514 it loaded prior to glx module. 55515 55516 Preliminary testing shows indirect glxgears works fine. 55517 55518 If user want to use GLES2 for glamor by using MESA, GLX will not 55519 work correctly. 55520 55521 If you are not using normal MESA, for example PVR's private GLES 55522 implementation, then it should be ok to use GLES2 glamor and the 55523 GLX should work as expected. In this commit, I use gbm to check 55524 whether we are using MESA or non-mesa. Maybe not the best way. 55525 55526 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55527 55528commit 0a8fb8563f3cecf2019d3a35d5a61ec2b3a069cd 55529Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55530Date: Fri Feb 10 15:34:43 2012 +0800 55531 55532 glamor_pixmap: Should bind unpack buffer to 0 after the uploading. 55533 55534 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55535 55536commit e03ad27dfb885e58298e1fd2accac0b6ad409fce 55537Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55538Date: Sat Feb 11 07:37:17 2012 +0800 55539 55540 glamor_picture: Fix the wrong order of header file. 55541 55542 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55543 55544commit bf24c2c06870dc38b21795f4f6fc51b053134181 55545Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55546Date: Fri Feb 3 11:46:01 2012 +0800 55547 55548 glamor_dump_pixmap: Add helper routine to dump pixmap. 55549 55550 For debug purpose only to dump the pixmap's content. 55551 As this function will call glamor_prepare_access/glamor_finish_access 55552 internally. Please use it carefully. 55553 55554 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55555 55556commit 91891b3711b1fe1402bb7742c20e976fab63c6fe 55557Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55558Date: Fri Feb 3 11:44:51 2012 +0800 55559 55560 glamor_fill/tile: Fixed a tileX/tileY calculation bug. 55561 55562 The previous's calculation is incorrect, now fix it and then 55563 we don't need to fallback at glamor_tile. 55564 55565 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55566 Tested-by: Peng Li <peng.li@intel.com> 55567 55568commit 39d9e6c693a4c3ad12c6569f1fd56e0a87b164d2 55569Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55570Date: Fri Feb 3 11:21:37 2012 +0800 55571 55572 prepare_access: Don't use fbo after it was downloaded. 55573 55574 We add a new gl_fbo status GLAMOR_FBO_DOWNLOADED to indicate 55575 the fbo was already downloaded to CPU. Then latter the access 55576 to this pixmap will be treated as pure CPU access. In glamor, 55577 if we fallback to DDX/fbXXX, then we fallback everything 55578 currently. We don't support to jump into glamor acceleration 55579 layer between a prepare_access/finish_access. Actually, fbCopyPlane 55580 is such a function which may call to acceleration function within 55581 it. Then we must mark the downloaded pixmap to another state 55582 rather than a normal fbo textured pixmap, and then stick to use 55583 it as a in-memory pixmap. 55584 55585 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55586 Tested-by: Peng Li <peng.li@intel.com> 55587 55588commit 1817b6c0cf20aa6fc1e1aa5b68e47473f341b85a 55589Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55590Date: Thu Feb 2 11:21:05 2012 +0800 55591 55592 glamor_eglmodule: Change module name according to normalize naming rule. 55593 55594 As Xorg module loader will normalize module name which will 55595 remove '_' when we put "glamor_egl" to the configure file, 55596 then it will fail to find us. 55597 55598 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55599 55600commit 1ab40028740efad1ba2ee16d425b0fbdd822c6ab 55601Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55602Date: Thu Feb 2 09:34:42 2012 +0800 55603 55604 Don't call dixSetPrivate directly. 55605 55606 We may change the way to set/get those private data latter. 55607 consolidate to glamor_set_pixmap/screen_private is better 55608 than call those dixSetPrivate directly. 55609 55610 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55611 55612commit bf7d79dc0ac2e97ed97e22c1d2c95e77bf959327 55613Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55614Date: Tue Jan 31 18:16:58 2012 +0800 55615 55616 Refine CloseScreen and FreeScreen processes. 55617 55618 This commit move the calling to glamor_close_screen from 55619 glamor_egl_free_screen to glamor_egl_close_screen, as this 55620 is the right place to do this. 55621 55622 We should detach screen fbo and destroy the corresponding 55623 KHR image at glamor_egl_close_screen stage. As latter 55624 DDX driver will call DestroyPixmap to destroy screen pixmap, 55625 if the fbo and image are still there but glamor screen private 55626 data pointer has been freed, then it causes segfault. 55627 55628 This commit also introduces a new flag GLAMOR_USE_EGL_SCREEN. 55629 if DDX driver is using EGL layer then should set this bit 55630 when call to glamor_init and then both glamor_close_screen 55631 and glamor_egl_close_screen will be registered correctly, 55632 DDX layer will not need to call these two functions manually. 55633 This way is also the preferred method within Xorg domain. 55634 55635 As interfaces changed, bump the version to 0.3.0. 55636 55637 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55638 Tested-by: Peng Li <peng.li@intel.com> 55639 55640commit 97efbd25fed0f86338853d81b974b7f4deb9f1b4 55641Author: Chris Wilson <chris@chris-wilson.co.uk> 55642Date: Fri Jan 27 13:22:13 2012 +0000 55643 55644 Use CLAMP_TO_BORDER in copy_n_to_n so we can sample outside of the source 55645 55646 In order to reduce a composite operation to a source, we need to provide 55647 Render semantics for the pixel values of samples outside of the source 55648 pixmap, i.e. they need to be rgba(0, 0, 0, 0). This is provided by using 55649 the CLAMP_TO_BORDER repeat mode, but only if the texture has an alpha 55650 channel. 55651 55652 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 55653 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55654 55655commit 864153bb9eeadb539ed5e9488b829c9f3f8b9482 55656Author: Chris Wilson <chris@chris-wilson.co.uk> 55657Date: Fri Jan 27 13:22:12 2012 +0000 55658 55659 Do not reduce a composite to a copy if we need to sample outside of the source 55660 55661 In order to maintain Render semantics, samples outside of the source 55662 should return CLEAR. The copy routines instead are based on the core 55663 protocol and expects the source rectangle to be wholly contained within 55664 the drawable and so does no fixup. 55665 55666 Fixes the rendering of GTK icons. 55667 55668 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 55669 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55670 55671commit 566cca59e14b223b8dcd045ccc2bbcf08d05b297 55672Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55673Date: Sun Jan 29 14:14:36 2012 +0800 55674 55675 glamor-gles2: Fixup the pixmap before read back if it is not readable. 55676 55677 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55678 55679commit 36ac9b7191a4bbbe4fb25ced9ee27d1e91308a15 55680Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55681Date: Sat Jan 21 14:41:44 2012 +0800 55682 55683 glamor-fbo: Tweek the cache bucket calculation. 55684 55685 And also reduce the expire count to 100 which should be 55686 good enough on x11perf and cairo-trace testing. 55687 55688 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55689 55690commit a1de528c56ff1cfd05a04c4ad127456ec00707fd 55691Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55692Date: Sat Jan 21 13:22:14 2012 +0800 55693 55694 glamor_create_fbo: Concentrate the fbo size/depth checking. 55695 55696 Concentrate checking the size/depth when creating fbo. Simply 55697 the pixmap creation and the uploading fbo/texture preparing. 55698 55699 Also slightly change the uploading fbo's preparation. If don't 55700 need fbo, then a fbo only has valid texture should be enough 55701 to return. 55702 55703 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55704 55705commit 1bfe5957117ba5916236caa021124734228e5aa9 55706Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55707Date: Fri Jan 20 20:38:29 2012 +0800 55708 55709 glamor-pixmap-upload: Create a uploading fbo with a texture only. 55710 55711 Just an initial implementation and disabled by default. 55712 When uploading a pixmap to a texture, we don't really want 55713 to attach the texture to any fbo. So add one fbo type 55714 which doesn't has a gl FBO attached to it. 55715 This commit can increase the cairo-trace's performance by 55716 10-20%. Now the firefox-planet-gnome is 8.3s. SNA is still 55717 the best, only take 3.5s. 55718 55719 Thanks for Chris to point out the A1 pixmap uploading bug. 55720 55721 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55722 55723commit 6fb92e67b3f3dee38d8d440c8ab9f8cc5dcef9bc 55724Author: Li Peng <peng.li@intel.com> 55725Date: Fri Jan 20 16:23:17 2012 +0800 55726 55727 glamor: check driver support GEM or not 55728 55729 glamor calls DRM_IOCTL_GEM_FLINK to get a name for a buffer object. 55730 It only works for driver that support GEM, such as intel i915 driver. 55731 But for pvr driver who doesn't has GEM, we can't do it this way. 55732 55733 According to Chris's comments, we check the has_gem as the following 55734 method: 55735 55736 Here we just try to flink handle 0. If that fails with ENODEV or 55737 ENOTTY instead of ENOENT (or EINVAL on older kernels), set has_gem=0. 55738 55739 Signed-off-by: Li Peng <peng.li@intel.com> 55740 Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55741 55742commit 68789b23e791d81c6987c755a56851961cbb262f 55743Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55744Date: Fri Jan 20 15:56:25 2012 +0800 55745 55746 glamor_gles2: Consolidate gles2 pixmap format readable check to one function. 55747 55748 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55749 55750commit 3373d2c696028a9f2f637430c48b9de1a3776800 55751Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55752Date: Fri Jan 20 12:10:06 2012 +0800 55753 55754 glamor_egl: Add support for the platform doesn't have gbm. 55755 55756 Maybe we should use eglGetDisplayDRM to get display, but current 55757 PVR's driver is using eglGetDisplay. 55758 55759 Signed-off-by: Peng Li <peng.li@intel.com> 55760 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55761 55762commit 92671e3ac8cc955379085c143ee8cf8b37e760ec 55763Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55764Date: Fri Jan 20 11:58:34 2012 +0800 55765 55766 glamor_egl: Don't call eglDestroyImageKHR directly. 55767 55768 Some implementation doesn't have it. 55769 55770 Signed-off-by: Peng Li <peng.li@intel.com> 55771 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55772 55773commit 125f317d90b69fa4d6f98da9badc1946f0a099ec 55774Author: Peng Li <peng.li@intel.com> 55775Date: Fri Jan 20 11:52:18 2012 +0800 55776 55777 glamor_gl_dispatch: fix the dispatch initialization on GLES2. 55778 55779 Some gles2 implementation doesn's support get_proc_address. 55780 And we also need to avoid get those missing functions pointers 55781 when we are GLES2. 55782 55783 Signed-off-by: Peng Li <peng.li@intel.com> 55784 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55785 55786commit 64fef665c9297ddd110b8472943f96b55db120ba 55787Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55788Date: Fri Jan 20 10:56:17 2012 +0800 55789 55790 glamor_render: Add non-Map/Unmap vertex array for GLES. 55791 55792 As some GLES implementations' glMapOES /glUnmapOES is 55793 not so efficient, we implement the in memory vertex array 55794 for them. 55795 55796 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55797 55798commit c244969b331e08679be9a9618d75fe5ee9cc9d86 55799Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55800Date: Fri Jan 20 10:55:16 2012 +0800 55801 55802 glamor_init: Should set gl_flavor before sub-module intialization. 55803 55804 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55805 55806commit 62e536535168827be76dafb1f5b5e0807c1d5ec9 55807Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55808Date: Fri Jan 20 10:04:21 2012 +0800 55809 55810 glamor_composite: Fix one bug when we have too more vertices. 55811 55812 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55813 55814commit 9c6fd931a63fb8a5300014265e4f1cacc746857a 55815Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55816Date: Fri Jan 20 09:59:44 2012 +0800 55817 55818 glamor-fbo-pool: Enable to reuse different size fbo/texture. 55819 55820 Fixup three special cases, one is in tile and the other is in 55821 composite. Both cases are due to repeat texture issue. Maybe 55822 we can refine the shader to recalculate texture coords to 55823 support partial texture's repeating. 55824 55825 The third is when upload a memory pixmap to texture, as now 55826 the texture may not have the exact size as the pixmap, we 55827 should not use the full rect coords. 55828 55829 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55830 55831commit c7e79d6acff5abd7a327d5f4d6698ae5d7583834 55832Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55833Date: Thu Jan 19 20:47:55 2012 +0800 55834 55835 glamor-fbo-pool: Implement fbo cache mechanism. 55836 55837 We classify the cache according to the texture's format/width/height. 55838 As openGL doesn't allow us to change a texture's format/width/height 55839 after the internal texture object is already allocated, we can't 55840 just calculate the size and then according ths size to put the 55841 fbo to an bucket which is just like SNA does. We can only put 55842 the fbo to the corresponding format/width/height bucket. 55843 55844 This commit only support the exact size match. The following patch 55845 will remove this restriction, just need to handle the repeat/tile 55846 case when the size is not exactly match. 55847 55848 Should use fls instead of ffs when decide the width/height bucket, 55849 thanks for Chris to point this out. 55850 55851 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55852 55853commit 2ff41008494e6c5909c058f1f80b4f66617dada1 55854Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55855Date: Wed Jan 18 19:21:36 2012 +0800 55856 55857 glamor_fbo: Introduce glamor fbo to manage all the fb/tex. 55858 55859 This is the first patch to implement a fbo/tex pool mechanism which 55860 is like the sna's BO cache list. We firstly need to decopule the 55861 fbo/tex from each pixmap. The new glamor_pixmap_fbo data 55862 structure is for that purpose. It's somehow independent to each 55863 pixmap and can be reused latter by other pixmaps once it's detached 55864 from the current pixmap. 55865 55866 And this commit also slightly change the way to create a 55867 memory pixmap. We will not create a pixmap private data structure 55868 by default, instead we will crete that structure when a memory 55869 pixmap is attaching a fbo to it. 55870 55871 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55872 55873commit ca2ddd33a114fe83584b5fa9f73b7534abdb96fa 55874Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55875Date: Wed Jan 18 17:12:32 2012 +0800 55876 55877 glamor_set_pixmap_texture/screen_pixmap: Remove useless parameters. 55878 55879 As after we got a texture, no matter the texture is created 55880 on the glamor_create_pixmap or on the egl layer, we all already 55881 know the texture's width and height there. We don't need 55882 to pass them in. 55883 55884 This commit also simply the glamor_egl_create_textured_screen to 55885 reuse the egl_create_textured_pixmap. And also remove the useless 55886 root image from the egl private structure. As now the root image 55887 is bound to the screen image, we don't take care it separately 55888 here. It will be freed at the screen closing. 55889 55890 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55891 55892commit 15166bba973206dcb98121eb3932660529cfc997 55893Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55894Date: Wed Jan 18 17:10:26 2012 +0800 55895 55896 Add debug message for all the uploading path. 55897 55898 The previous message missed the texture restoring path. 55899 55900 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55901 55902commit 994a9ff7f58161bf5651f83d810eb77b7718ab00 55903Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55904Date: Wed Jan 18 17:07:25 2012 +0800 55905 55906 glamor_create_picture: Fix the format matching method. 55907 55908 We should not simply set a TEXTURE_DRM pixmap to a separated 55909 texture pixmap. If the format is compatible with current fbo 55910 then it is just fine to keep it as TEXTURE_DRM type and we 55911 can safely fallback to DDX layer on it. 55912 55913 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55914 55915commit 28fcd7cd01edfdf68c370e6c6ad0238d45477b3f 55916Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55917Date: Fri Jan 13 11:18:28 2012 +0800 55918 55919 Rearrange data structure and remove unused fileds. 55920 55921 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55922 55923commit bdd72da0c525faf2aac38a7a8afa1cd88cd8dc1b 55924Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55925Date: Tue Jan 10 17:04:48 2012 +0800 55926 55927 Release previous textre/fb when bind a new texture to a pixmap. 55928 55929 As we want to support DRI2 drawable which may create a new textured_drm 55930 to a pre-existing texture_only pixmap, we have to add this logical. 55931 55932 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55933 Tested-by: He Junyan<junyan.he@linux.intel.com> 55934 55935commit d7352d57b9ceb809d47c922ed8c820a44f8a0ee5 55936Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55937Date: Tue Jan 10 15:24:36 2012 +0800 55938 55939 Add glFinish after glFlush. 55940 55941 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55942 55943commit 8b943ce2030b02bb8d279ce2c284cb5d2910ced6 55944Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55945Date: Mon Jan 9 12:50:39 2012 +0800 55946 55947 Set glamor's initial version to 0.2.0. 55948 55949 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55950 55951commit 7329414bf2ca9873c04150a1d9386cf37f70b663 55952Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55953Date: Mon Jan 9 07:16:06 2012 +0800 55954 55955 Silence a compilation warning. 55956 55957 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55958 55959commit 069a6d1746ef4c180f48a75e655e5e06e33327b5 55960Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55961Date: Mon Jan 9 07:11:26 2012 +0800 55962 55963 glamor_composite: Allocate VBO on demand. 55964 55965 Use a fixed VBO is not efficient. Some times we may only has less than 55966 100 verts, and some times we may have larger than 4K verts. We change 55967 it to allocate VBO buffer dynamically, and this can bring about 10% 55968 performance gain for both aa10text/rgb10text and some cairo benchmarks. 55969 55970 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55971 55972commit 42a0261cb3065200887f81816b1bc3850593da4c 55973Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55974Date: Mon Jan 9 05:03:08 2012 +0800 55975 55976 glamor_getimage: Add the optimization path of getImage. 55977 55978 This optimization will only call glReadPixels once. It should get 55979 some performance gain. But it seems it even get worse performance 55980 at SNB, disable it by default. 55981 55982 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55983 55984commit 96085017c8da96e6d882c87b69186f5aba20131d 55985Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55986Date: Mon Jan 9 05:01:48 2012 +0800 55987 55988 Consolidate the choose of internal texture format to one function. 55989 55990 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 55991 55992commit a74596be0e7e02b43fb0db844b71b68f384ab599 55993Author: Zhigang Gong <zhigang.gong@linux.intel.com> 55994Date: Sun Jan 8 08:50:41 2012 +0800 55995 55996 Set filter to GL_NEAREST by default. 55997 55998 This is the fastest filter and let's use it by default. 55999 56000 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56001 56002commit 4cd07871a417b15b0382c07fecec497e93697a5d 56003Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56004Date: Sun Jan 8 08:45:08 2012 +0800 56005 56006 glamor-composite: Use glDrawElements to reduce the count of vertices. 56007 56008 To split a rectangle (0,1,2,3) to two separated triangles need to feed 56009 6 vertices, (0,1,2) and (0,2,3). use glDrawElements can reuse the shared 56010 vertices. 56011 56012 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56013 56014commit 9dafd6fce547a6824c1de98d4309f8f2197ebbd0 56015Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56016Date: Sun Jan 8 01:14:12 2012 +0800 56017 56018 glamor-composite: Optimize the computation of composite region. 56019 56020 Computing the composite region at the composite_with_shader is very 56021 inefficient. As when we call to here from the glamor_glyph's temproary 56022 picture, we don't need to compute this region at all. So we move this 56023 computing out from this function and do that at the glamor_composite 56024 function. This can get about 5% performance gain for aa10text/rgb10text. 56025 56026 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56027 56028commit 2511a00cdd576512f5d7f45707a4a725df2f4618 56029Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56030Date: Thu Jan 5 14:26:01 2012 +0800 56031 56032 Fixed a configure bug. 56033 56034 Should check the enable-glamor-gles2 before use the variable. 56035 And should include the config.h as the GLAMOR_GLES2 macro is 56036 defined there. 56037 56038 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56039 56040commit a65e1c736aaa1bb2440a83c42c0410d9d9b58007 56041Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56042Date: Wed Jan 4 14:40:24 2012 +0800 56043 56044 Reduce the double check of pixmap's private pointer. 56045 56046 As we now add the checking to the Macro, we don't need to check 56047 the pointer outside the Macro. 56048 56049 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56050 56051commit e1789893e5fe07fdbd8f21b7fa8a15f6db066592 56052Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56053Date: Wed Jan 4 14:37:37 2012 +0800 56054 56055 get_spans: Check whether have a valid fbo before check format. 56056 56057 If a pixmap is a pure in-memory pixmap, we do not need to 56058 check its format. Format checking has more overhead than 56059 checking FBO, so we change to check fbo firtly. 56060 56061 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56062 56063commit 057f52a04d47e6ca5f6dead9c1bc71765d4c7ae8 56064Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56065Date: Wed Jan 4 14:34:39 2012 +0800 56066 56067 Track all picture's drawable pict format. 56068 56069 Even if a picture's pixmap is a pure in memory pixmap, we still need 56070 to track its format. The reason is we may need to upload this drawable 56071 to texture thus we need to know its real picture format. 56072 56073 As to the MACRO to check whether a pixmap is a picture, we should 56074 check whether the priv is non-NULL before we touch its field. 56075 56076 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56077 56078commit 8a4758a358e56b1db481607ff3f7e9375238e5d8 56079Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56080Date: Wed Jan 4 13:11:52 2012 +0800 56081 56082 Need to check pixmap priv before touch its field. 56083 56084 As now the pixmap may be allocated by DDX and doesn't have a 56085 valid pixmap private field. We must check pixmap private 56086 pointer before touch its field value. If a pixmap doesn't 56087 have a non-NULL private pointer, it doesn't have a valid 56088 FBO. 56089 56090 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56091 56092commit 9264335347da641b2626a6b9f56aa05ab313239e 56093Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56094Date: Sat Dec 31 19:29:17 2011 +0800 56095 56096 Added more drawing functions. 56097 56098 As we want to take over all the possible GC ops from the DDX 56099 layer, we need to add all the missed functions. 56100 This commit also fixed one bug at polylines. 56101 We simply drop the bugy optimized code now, as it did not 56102 consider of clip info. 56103 56104 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56105 56106commit d42eb04c29d015fb1e4ed4e9ded4c0dd8bc7dc3c 56107Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56108Date: Sat Dec 31 19:27:33 2011 +0800 56109 56110 Remove useless output messages. 56111 56112 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56113 56114commit fbccc4bbbc84a5459010d3238a00a0dd49111eec 56115Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56116Date: Sat Dec 31 19:20:10 2011 +0800 56117 56118 Fixed a rendering bug at fillspans. 56119 56120 We should not change the points coords when loop for the clip 56121 rects. Change to use another variable to store the clipped 56122 coords and keep the original coords. This bug cause some 56123 XTS failures. Now fix it. 56124 56125 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56126 56127commit 70b6341538cc11f074b02f9dd97d387418245d77 56128Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56129Date: Sat Dec 31 19:14:06 2011 +0800 56130 56131 Fixed a bug at putImage. 56132 56133 fbPutImage wants the input drawable is the target drawable rather 56134 than the backing pixmap. This bug cause some XTS failures, now 56135 fix it. 56136 56137 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56138 56139commit 27c4c0457dac307500859ca7a14a1e58465d5e0b 56140Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56141Date: Tue Dec 27 16:44:17 2011 +0800 56142 56143 Remove the assertion which is not safe sometimes. 56144 56145 The original version assumes that each drawable pixmap should 56146 have a valid private pixmap pointer. But this is not true after 56147 we create this libglamor. As the DDX layer may create a pure 56148 software drawable pixmap which doesn't have a private pixmap 56149 pointer. 56150 56151 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56152 56153commit c65dc686228262894e0afc9fbe9fe6635d420e2f 56154Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56155Date: Tue Dec 27 16:42:32 2011 +0800 56156 56157 Export glamor_validate_gc to DDX. 56158 56159 This is also a function which may direct access pixmaps which 56160 may be a glamor only pixmap and DDX doesn't know how to access 56161 it. We have to export this API to DDX driver and let the DDX 56162 driver use it to do the validation. 56163 56164 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56165 56166commit f2809745fb641239be5b0281eac569c5b371e55e 56167Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56168Date: Fri Dec 23 16:21:44 2011 +0800 56169 56170 Flush gl operations when destroy an textured image. 56171 56172 Before destroy an image which was attached to a texture. 56173 we must call glFlush to make sure the operation on that 56174 texture has been done. 56175 56176 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56177 56178commit b36d264ec98440707ed3a94fab430e23d4a1582a 56179Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56180Date: Fri Dec 23 16:19:39 2011 +0800 56181 56182 Add a new API glamor_create_gc. 56183 56184 As at ValidateGC stage, it may need to touch the pixmap directly, for 56185 example the tile pixmap. We must export this interface to DDX driver 56186 and let the DDX driver to route the processing to us. As this pixmap 56187 may be a texture only pixmap, and DDX don't know how to handle it. 56188 56189 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56190 56191commit 3ba546b6998bf1c5d68ff0083017f804ffc48a93 56192Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56193Date: Fri Dec 23 16:18:22 2011 +0800 56194 56195 Code cleanup. 56196 56197 Remove useless code, and after a fbCreatePixmap, we 56198 do not need to modify its header there. 56199 56200 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56201 56202commit 2583620c6cd6a6fedeb7082b526f2751eb6f75b4 56203Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56204Date: Fri Dec 23 14:23:56 2011 +0800 56205 56206 Remove useless egl functions. 56207 56208 We only need to create image fron external name rather 56209 than use drm_image_mesa to create drm image, so remove 56210 them. 56211 56212 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56213 56214commit 7285f2e836c56c31a42b8c18a7ebd0e8a7d78fa1 56215Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56216Date: Fri Dec 16 20:15:33 2011 +0800 56217 56218 Make sure the devPrivate.ptr of a textured pixmap is NULL. 56219 56220 Just in case when wrongly fallback to DDX layer and cause 56221 random memory corruption. Pointed out by Chris. 56222 56223 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56224 56225commit 5769d8603ddef6a8023f12bba1d3d3cde83e82d4 56226Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56227Date: Fri Dec 16 15:04:33 2011 +0800 56228 56229 Export glamor_create_texture to DDX driver. 56230 56231 If DDX failed to create textured pixmap from its BO's handle, 56232 it can turn to call this API to create a brand new glamor 56233 rather than fallback to pure in memory pixmap. 56234 56235 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56236 56237commit 3f5d53b33978bdb848dd605a19d3fb4ea5f259e0 56238Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56239Date: Fri Dec 16 15:02:54 2011 +0800 56240 56241 Correct the logic of glamor_egl_create_textured_pixmap. 56242 56243 Discussed with Chris and found the previous logic is not 56244 good. Now change it in this commit, this API will just 56245 try to create a textured pixmap from the handle provided 56246 by DDX driver, if failed simply return FALSE without touch 56247 the pixmap. And the DDX driver can choose how to do next 56248 step. 56249 56250 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56251 56252commit 8c7fcefb965bfdaaf0576347da20155d0415caa5 56253Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56254Date: Tue Dec 13 22:48:34 2011 +0800 56255 56256 glamor-for-ddx: Exports all rendering/drawing functions. 56257 56258 This commit exports all the rest rendering/drawing functions 56259 to the DDX drivers. And introduce some new pixmap type. For 56260 a pixmap which has a separated texture, we never fallback 56261 it to the DDX layer. 56262 56263 This commit also adds the following new functions: 56264 glamor_composite_rects, glamor_get_image_nf which are needed 56265 by UXA framework. Just a simple wrapper function of miXXX. 56266 Will consider to optimize them next few weeks. 56267 56268 This commit also Fixed a glyphs rendering bug pointed by Chris. 56269 56270 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56271 56272commit f7539d9bff0a65b9430dcae8745bf5d4409a9f04 56273Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56274Date: Mon Dec 12 09:58:03 2011 +0800 56275 56276 Enable texture dynamic uploading. 56277 56278commit 2d0ea392ec574efb96ac7a04ee72ba580417ba1e 56279Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56280Date: Mon Dec 12 09:49:06 2011 +0800 56281 56282 Call screen's method to create pixmap. 56283 56284 As we may need to fallback to DDX's rendering path 56285 during the glyphs, we have to call screen's create pixmap 56286 method to create pixmap. 56287 56288 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56289 56290commit 47e86eea563601c4e1e356f7267b935fcdd542f7 56291Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56292Date: Mon Dec 12 09:48:04 2011 +0800 56293 56294 Fallback if the pixmap is drm only. 56295 56296 Glamor can't do anything with a drm only pixmap. 56297 56298 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56299 56300commit e3d16c9ebdffe8e5793ee253cf9f72b884b42805 56301Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56302Date: Mon Dec 12 07:09:03 2011 +0800 56303 56304 Classfy glamor pixmap to different types. 56305 56306 During the integration with intel driver, we introduce two 56307 new type of pixmap, one is TEXTURE_DRM, the other is DRM_ONLY. 56308 TEXTURE_DRM means we create a texture bind to the DRM buffer 56309 successfully. And then the texture and the external BO is 56310 consistent. DRM_ONLY means that we failed to create a texture 56311 from the external DRM BO. We need to handle those different 56312 types carefully, so we have to track them in the data structure. 56313 56314 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56315 56316commit 36d424feaeec765d131c015df77d24db1a36fc38 56317Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56318Date: Mon Dec 12 06:56:36 2011 +0800 56319 56320 Call glamor_create_pixmap directly in glamor rendering path. 56321 56322 When glamor is rendering pixmaps, and needs to create some 56323 temporary pixmap, it's better to use glamor version create 56324 pixmap directly. As if goes to external DDX's create pixmap, 56325 it may create a external DRM buffer which is not necessary. 56326 All the case within glamor scope is to create a texture only 56327 pixmap or a in memory pixmap. 56328 56329 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56330 56331commit 10ad332dedad2df13212dcd198d947a2f5c76862 56332Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56333Date: Sun Dec 11 08:06:13 2011 +0800 56334 56335 Remove garbage file. 56336 56337 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56338 56339commit 92fd83872f5509a2792338b4665dc2b3e1ed68cd 56340Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56341Date: Sun Dec 11 02:26:15 2011 +0800 56342 56343 Remove the access mode from private pixmap structure. 56344 56345 Change the finish_access to pass in the access mode, and remove 56346 the access mode from the pixmap structure. This element should 56347 not be a pixmap's property. 56348 56349 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56350 56351commit d135e879a6c8698bf21ae44a02315f370b961b14 56352Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56353Date: Fri Dec 9 16:01:45 2011 +0800 56354 56355 Can't create KHR image if the depth is uncompatible. 56356 56357 Currently, KHR image only support one color format ARGB32. 56358 For all other format, we have to fail to create corresponding 56359 image and texture here. 56360 56361 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56362 56363commit b5630663cf9438383166f59cdfc7889571f2cd62 56364Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56365Date: Fri Dec 9 15:58:19 2011 +0800 56366 56367 exports more rendering functions to DDX driver. 56368 56369 Exports all necessary rendering functions to DDx drivers, including 56370 CopyArea, Glyphs, Composite, Triangles, .... 56371 56372 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56373 56374commit b5480e64ac4a72bdc31e587ff9e51d67a1a0f11d 56375Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56376Date: Fri Dec 9 14:57:24 2011 +0800 56377 56378 Override create/destroy picture to track pict format. 56379 56380 When create a piccture, we need to attach a pixmap to it. 56381 A pixmap only has a depth, which is not sufficant for glamor. 56382 As in openGL texture only has a few internal formats which 56383 is not sufficant to represent all the possible picture 56384 format. So we always transform the picture format to GL_RGBA. 56385 And when we need to read back the picture, we must know the 56386 original picture format. So we have to override create 56387 and destroy picture to track a pixmap's real picture format 56388 if it is attached to a picture. 56389 56390 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56391 56392commit 7b2310834d83198fc88b5a5fb062cb1c9ee19ead 56393Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56394Date: Tue Nov 22 18:28:10 2011 +0800 56395 56396 Added a missed header file. 56397 56398 Pointed by Chris, we must add xorg-server.h at the top 56399 of each file before we include any other xorg header files. 56400 Otherwise, it may cause incorrect XID length. 56401 56402 This commit also fixes one compilation warning in X86_64 56403 platform. 56404 56405 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56406 56407commit 90eaac96aeb53b3e45b82dc7dd4e98778c8c9343 56408Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56409Date: Tue Nov 15 20:17:18 2011 +0800 56410 56411 If caller is not using glamor screen, we can't register 56412 the block handler. 56413 56414 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56415 56416commit 9ff39e53eb20673d42f42c94adb55bf9a9f1c285 56417Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56418Date: Tue Nov 15 20:16:16 2011 +0800 56419 56420 Revert "Disable glamor_tile temporary." 56421 56422 The previous corruption output is caused by GLAMOR/UXA flushing 56423 issue. Now get fixed. 56424 56425 This reverts commit 0d2d3e4f051e48c3499e1c972e012803a8623526. 56426 56427commit 13ab4d59afb824f2a4c03280fffe79c6ba9aa26c 56428Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56429Date: Fri Nov 11 14:55:12 2011 +0800 56430 56431 Disable glamor_tile temporary. 56432 56433 Enable glamor tile cause corrupted output. Need more investigation. 56434 56435 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56436 56437commit 9b6a484df0724d72dd27ad620672e7a762a460df 56438Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56439Date: Fri Nov 11 10:26:00 2011 +0800 56440 56441 Add new version glamor_poly_fill_rect without internal fallback. 56442 56443 If need fallback, this new version just returns FALSE without 56444 doing anything. It's the caller's responsibility to implement 56445 fallback method. 56446 56447 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56448 56449commit ba1b3b53240b5d499791570ba9f598a4e421c719 56450Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56451Date: Wed Nov 9 16:09:03 2011 +0800 56452 56453 Add new version glamor_fillspans without internal fallback. 56454 56455 For the purpose of incrementally intergration of existing intel 56456 driver, for the GC operations we may don't want to use glamor's 56457 internal fallback which is in general much slower than the 56458 implementation in intel driver. If the parameter "fallback" is 56459 false when call the glamor_fillspans, then if glamor found it 56460 can't accelerate it then it will just return a FALSE rather than 56461 fallback to a slow path. 56462 56463 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56464 56465commit b861aad8e2fcf6fe1fae4f26abb650bb4eb499c6 56466Author: Zhigang Gong <zhigang.gong@linux.intel.com> 56467Date: Wed Nov 2 13:44:50 2011 +0800 56468 56469 Initial version. 56470 56471 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 56472 56473commit 2dbbe2565052cc024ce0e98918ed34c1239b780a 56474Merge: 4d6264614 7982eca62 56475Author: Eric Anholt <eric@anholt.net> 56476Date: Wed Dec 18 11:21:07 2013 -0800 56477 56478 Merge branch 'glamor-gongzg-merge' 56479 56480commit 7982eca622bbc4b6a4845801a77da8a16138004a 56481Author: Eric Anholt <eric@anholt.net> 56482Date: Wed Dec 18 11:16:15 2013 -0800 56483 56484 Revert changes outside of glamor from the glamor branch. 56485 56486 We want to merge the external glamor code to the xserver, with the 56487 internal history retained. However, we don't want a bunch of 56488 non-building changes along the way, so remove all the build system and 56489 support code outside of glamor for now. 56490 56491commit 4d62646142718024b0981eb4f1fd0131e829161f 56492Merge: f4bfb14f5 81ba89d67 56493Author: Keith Packard <keithp@keithp.com> 56494Date: Mon Dec 16 09:27:57 2013 -0800 56495 56496 Merge remote-tracking branch 'jeremyhu/master' 56497 56498commit f4bfb14f53a939574da1f5995f0dad949898b86a 56499Author: Adam Jackson <ajax@redhat.com> 56500Date: Thu Dec 12 10:57:40 2013 -0500 56501 56502 configure: Fix a typo near dri3/shmfence detection 56503 56504 " is not the mate of ]. 56505 56506 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 56507 Signed-off-by: Adam Jackson <ajax@redhat.com> 56508 56509commit 8248b4af31ef71ac03158871b77f20eb456dbe38 56510Author: Adam Jackson <ajax@redhat.com> 56511Date: Fri Dec 13 15:07:48 2013 -0500 56512 56513 glx: Add null pointer protection to __glGetProcAddress 56514 56515 This can't happen when GLX is the backing window system, but can 56516 elsewhere. We may as well protect against it at a high level. 56517 56518 Reviewed-by: Eric Anholt <eric@anholt.net> 56519 Signed-off-by: Adam Jackson <ajax@redhat.com> 56520 56521commit 128449dd6498a2f74c3770f89a9dae0f70e2b351 56522Author: Adam Jackson <ajax@redhat.com> 56523Date: Fri Dec 13 11:39:16 2013 -0500 56524 56525 present: Don't use the major/minor version from the protocol headers 56526 56527 We want to advertise the version we implement, not the version the 56528 protocol headers happen to describe. 56529 56530 Reviewed-by: Jasper St. Pierre <<jstpierre@mecheye.net> 56531 Signed-off-by: Adam Jackson <ajax@redhat.com> 56532 56533commit ec6087bf070131bc3a26848fb9fb840f09b0e19d 56534Author: Adam Jackson <ajax@redhat.com> 56535Date: Fri Dec 13 11:36:38 2013 -0500 56536 56537 dri3: Don't use the major/minor version from the protocol headers 56538 56539 We want to advertise the version we implement, not the version the 56540 protocol headers happen to describe. 56541 56542 Reviewed-by: Jasper St. Pierre <<jstpierre@mecheye.net> 56543 Signed-off-by: Adam Jackson <ajax@redhat.com> 56544 56545commit e6fafd3de70d315130fb6e5732cfd02a5901b788 56546Author: Adam Jackson <ajax@redhat.com> 56547Date: Fri Dec 13 11:28:59 2013 -0500 56548 56549 dri3: Guard against buggy clients 56550 56551 There's nothing to stop a client from sending these requests to screens 56552 without DRI3 support, and if they do, we'll crash. Let's not do that. 56553 56554 Reviewed-by: Jasper St. Pierre <<jstpierre@mecheye.net> 56555 Signed-off-by: Adam Jackson <ajax@redhat.com> 56556 56557commit 81ba89d6703a22178a153aa39478ba2d4bde262b 56558Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 56559Date: Fri Dec 13 01:39:44 2013 -0800 56560 56561 configure.ac: Add PRESENT_LIB to XQUARTZ_LIBS 56562 56563 Undefined symbols for architecture x86_64: 56564 "_present_extension_init", referenced from: 56565 _staticExtensions in libXquartz.a(miinitext.o) 56566 "_present_register_complete_notify", referenced from: 56567 ___glXregisterPresentCompleteNotify in libglx.a(glxcmds.o) 56568 56569 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 56570 56571commit d7c9235ee261b0f780320985233e00dec5e2689c 56572Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 56573Date: Sat Dec 7 01:36:33 2013 -0800 56574 56575 XQuartz: Use asl_log_descriptor to log stdout/stderr of child processes 56576 56577 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 56578 56579commit ad8111d7c971ce448905c733d65ba0cfc72bdca4 56580Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 56581Date: Sat Dec 7 01:14:37 2013 -0800 56582 56583 darwin: Don't leave stdin/stdout closed 56584 56585 <rdar://problem/15609419> 56586 56587 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 56588 56589commit 929795d50d788358d6269ce423f72c6cc40e334b 56590Author: Peter Hutterer <peter.hutterer@who-t.net> 56591Date: Tue Dec 3 10:14:51 2013 +1000 56592 56593 dix: fix check for grab type 56594 56595 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 56596 56597commit 23394c7fea0f5c33333198c87ecfecc9f6c6a791 56598Author: Peter Hutterer <peter.hutterer@who-t.net> 56599Date: Tue Dec 3 08:36:45 2013 +1000 56600 56601 Xi: ungrab device when releasing a passive grab without ButtonReleaseMask (#71878) 56602 56603 If an touch triggers an async button grab and that grab does not have the 56604 ButtonReleaseMask set, the TouchEnd is never delivered, deliveries is 0 and 56605 the grab is never deactivated. 56606 56607 If the grab is pointer async and keyboard sync, the keyboard events are stuck 56608 in EnqueueEvent until some other pointer event terminates the grab. 56609 56610 Change this to check for the number of listeners. If we're about to deliver a 56611 TouchEnd to a passive pointer grab, the number of listeners is already 1 - 56612 pointer grabs always accept so other listeners were removed. 56613 56614 X.Org Bug 71878 <http://bugs.freedesktop.org/show_bug.cgi?id=71878> 56615 56616 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 56617 56618commit c1d30b5bd7f90e68bc38404fd0cc32578d6d3018 56619Author: Ben Gamari <bgamari.foss@gmail.com> 56620Date: Thu Nov 21 21:24:20 2013 -0500 56621 56622 Xi: Don't ActivateEarlyAccept POINTER_REGULAR listeners 56623 56624 Bug #71878 describes a bug resulting in the server ceasing to respond to 56625 keyboard input after a touch event. The problem might be the following: 56626 56627 DeliverTouchBeginEvent tries to deliver an event to a listener of type 56628 LISTENER_POINTER_REGULAR, taking the following if branch, 56629 56630 if (listener->type == LISTENER_POINTER_REGULAR || 56631 listener->type == LISTENER_POINTER_GRAB) { 56632 rc = DeliverTouchEmulatedEvent(dev, ti, ev, listener, client, win, 56633 grab, xi2mask); 56634 if (rc == Success) { 56635 listener->state = LISTENER_IS_OWNER; 56636 /* async grabs cannot replay, so automatically accept this touch */ 56637 if (dev->deviceGrab.grab && 56638 dev->deviceGrab.fromPassiveGrab && 56639 dev->deviceGrab.grab->pointerMode == GrabModeAsync) 56640 ActivateEarlyAccept(dev, ti); 56641 } 56642 goto out; 56643 } 56644 56645 DeliverTouchEmulatedEvent succeeds. The deviceGrab meets all 56646 three of the conditions of the inner if, enters 56647 ActivateEarlyAccept which then fails due to, 56648 56649 BUG_RETURN(ti->listeners[0].type != LISTENER_GRAB && 56650 ti->listeners[0].type != LISTENER_POINTER_GRAB); 56651 56652 That is, despite listener->type == LISTENER_POINTER_REGULAR. With my 56653 non-existent knowledge of XINPUT, it seems like the solution here 56654 might be to only ActivateEarlyAccept when listener->type == 56655 LISTENER_POINTER_GRAB. 56656 56657 Signed-off-by: Ben Gamari <bgamari.foss@gmail.com> 56658 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 56659 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 56660 56661commit fe07ec19e212a68076560d243a2a935c54589068 56662Author: Keith Packard <keithp@keithp.com> 56663Date: Tue Dec 10 11:27:47 2013 -0800 56664 56665 present: recursively set window pixmaps on flip 56666 56667 Newly created windows inherit the pixmap of their parent, similarly, 56668 reparenting a tree inherits the pixmap of the destination tree. 56669 56670 Making present preserve the invariant that unredirected windows always 56671 have the same pixmap as their parent ensures that the above cases work 56672 correctly. 56673 56674 v2: name the recursive function to 'set_tree_pixmap' instead of 'set_window_pixmap' 56675 56676 Signed-off-by: Keith Packard <keithp@keithp.com> 56677 Reviewed-by: Adam Jackson <ajax@redhat.com> 56678 56679commit b3533d0b212b6747a8f9a01931253d6bdb648ee2 56680Author: Keith Packard <keithp@keithp.com> 56681Date: Tue Dec 10 10:29:11 2013 -0800 56682 56683 Bump version to 1.14.99.904 (1.15 RC4) 56684 56685 Signed-off-by: Keith Packard <keithp@keithp.com> 56686 56687commit 9a4d7572fe3be16bb5aded5f48d4217fdf725b1d 56688Author: Jon TURNEY <jon.turney@dronecode.org.uk> 56689Date: Thu Sep 27 18:59:45 2012 +0100 56690 56691 hw/xwin/glx: Rewrite WGL wrappers after Xserver conversion to direct GL dispatch 56692 56693 v1: Rewrite by Marc Haesen of the WGL wrapper function generation script to use 56694 Khronos group XML. 56695 56696 v2: Remove -dispatchheader option, since dispatch.h doesn't exist anymore, use 56697 the private glapi interface to construct the GL dispatch table for the native 56698 WGL thunks. 56699 56700 v3: 56701 Rewrite to generate shims for the OpenGL 1.2.1 (GL 1.2 + GL_ARB_imaging 56702 +GL_ARB_multitexture + GL_ARB_texture_compression(?)) functions the server links 56703 directly with rather than libGL. 56704 56705 These shims dispatch to either the mesa GL DLL, or a thunking DLL containing 56706 cdecl-to-stcall wrapper functions for the native GL DLL. 56707 56708 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 56709 Reviewed-by: Adam Jackson <ajax@redhat.com> 56710 56711commit cb48877a3c4cfb1ae2248000fcd3443e3092bdaa 56712Author: Jon TURNEY <jon.turney@dronecode.org.uk> 56713Date: Fri Dec 6 14:00:21 2013 +0000 56714 56715 glx: Consistently use ARB-suffixed names for ARB_multitexture functions 56716 56717 At the moment we have a mix of ARB and non-ARB suffixed forms for ARB_multitexture functions 56718 e.g. glMultiTexCoord1fvARB and glMultiTexCoord1dv 56719 56720 Consistently use the ARB-suffixed form, assuming that is present in all libGL 56721 which provide the OpenGL 1.2.1 ABI we expect to be able to directly link with. 56722 56723 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 56724 Reviewed-by: Adam Jackson <ajax@redhat.com> 56725 56726commit 1a021f57a1e75ab63e8f87704f0394162402c4d4 56727Author: Jon TURNEY <jon.turney@dronecode.org.uk> 56728Date: Sun Dec 1 22:25:46 2013 +0000 56729 56730 ephyr: Fix compilation when ./configure'd with --enable-debug 56731 56732 /jhbuild/checkout/xorg/xserver/hw/kdrive/ephyr/ephyr.c: In function ‘ephyrProcessMouseMotion’: 56733 /jhbuild/checkout/xorg/xserver/hw/kdrive/ephyr/ephyr.c:946:188: error: ‘ephyrCurScreen’ undeclared (first use in this function) 56734 /jhbuild/checkout/xorg/xserver/hw/kdrive/ephyr/ephyr.c: In function ‘ephyrProcessButtonPress’: 56735 /jhbuild/checkout/xorg/xserver/hw/kdrive/ephyr/ephyr.c:980:186: error: ‘ephyrCurScreen’ undeclared (first use in this function) 56736 /jhbuild/checkout/xorg/xserver/hw/kdrive/ephyr/ephyr.c: In function ‘ephyrProcessButtonRelease’: 56737 /jhbuild/checkout/xorg/xserver/hw/kdrive/ephyr/ephyr.c:1007:186: error: ‘ephyrCurScreen’ undeclared (first use in this function) 56738 56739 Fix ephyr compilation when ./configure'd with --enable-debug after commit 56740 46cf6bf5692ef751ec9d17ae2292565d4b13f14b, some instances of ephyrCurScreen were 56741 not converted to screen->pScreen->myNum. 56742 56743 v2: Don't use a trivial local variable which will be unused when ./configure'd 56744 with --disable-debug 56745 56746 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 56747 Reviewed-by: Adam Jackson <ajax@redhat.com> 56748 56749commit 875dbcef5bfcb6a94ef8456be6151e133408793f 56750Author: Jon TURNEY <jon.turney@dronecode.org.uk> 56751Date: Fri Nov 29 14:29:20 2013 +0000 56752 56753 configure.ac: Link XWin with present extension if we are building with it enabled 56754 56755 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 56756 Reviewed-by: Adam Jackson <ajax@redhat.com> 56757 56758commit e61e19959d9138d5b81b1f25b7aa3e257918170d 56759Author: Adam Jackson <ajax@redhat.com> 56760Date: Tue Dec 3 13:45:43 2013 -0500 56761 56762 xquartz/glx: Convert to non-glapi dispatch 56763 56764 CGL doesn't have anything like glXGetProcAddress, and the old code just 56765 called down to dlsym in any case. It's a little mind-warping since 56766 dlopening a framework actually loads multiple dylibs, but that's just 56767 how OSX rolls. 56768 56769 Signed-off-by: Adam Jackson <ajax@redhat.com> 56770 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 56771 56772commit a668aa0e41bc33ff2db7f9c53b6dc321a96926a9 56773Author: Adam Jackson <ajax@redhat.com> 56774Date: Tue Dec 3 15:03:13 2013 -0500 56775 56776 drisw: Wire up GetProcAddress 56777 56778 Signed-off-by: Adam Jackson <ajax@redhat.com> 56779 Reviewed-by: Keith Packard <keithp@keithp.com> 56780 56781commit 4fcdfeb7bc55d00d50d7f89788d9b929946ae3d9 56782Author: Adam Jackson <ajax@redhat.com> 56783Date: Tue Dec 3 15:02:57 2013 -0500 56784 56785 dri2: wire up GetProcAddress 56786 56787 Signed-off-by: Adam Jackson <ajax@redhat.com> 56788 Reviewed-by: Keith Packard <keithp@keithp.com> 56789 56790commit 47f00b3920be9d8df8b148263d744934f3a02cd4 56791Author: Adam Jackson <ajax@redhat.com> 56792Date: Tue Dec 3 15:01:44 2013 -0500 56793 56794 glx: Untangle the prototypes around the GetProcAddress thunk 56795 56796 Signed-off-by: Adam Jackson <ajax@redhat.com> 56797 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 56798 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 56799 56800commit c1fd143f2847ef425ec9891d9e1cba44a1bb007d 56801Author: Adam Jackson <ajax@redhat.com> 56802Date: Tue Dec 3 13:11:52 2013 -0500 56803 56804 glx: Remove function stubs 56805 56806 Now that we're calling non-1.2 ABI things by function pointer this is no 56807 longer needed. 56808 56809 Signed-off-by: Adam Jackson <ajax@redhat.com> 56810 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 56811 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 56812 56813commit 6b93e1f5ff39c6ac5b0d97d4245aeac9113eed8b 56814Author: Adam Jackson <ajax@redhat.com> 56815Date: Tue Dec 3 13:11:09 2013 -0500 56816 56817 glx: Convert non-generated function pointer thunking 56818 56819 Same concept as the generated code conversion. 56820 56821 Signed-off-by: Adam Jackson <ajax@redhat.com> 56822 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 56823 56824commit c4567a376083eb7b142a7f003ddf8372d376ea86 56825Author: Adam Jackson <ajax@redhat.com> 56826Date: Tue Dec 3 12:40:24 2013 -0500 56827 56828 glx: Convert generated code function pointer thunking 56829 56830 We're meant not to call these by name due to ABI. Rather than try to 56831 generate a bunch of little stub functions that do the lookup, just 56832 inline it all directly into the calling function. 56833 56834 This does not cache results. That's fine, this is not a performance 56835 path, and if we're atop WGL then we effectively have to do this every 56836 time anyway because wglGetProcAddress results are context-dependent. 56837 56838 Signed-off-by: Adam Jackson <ajax@redhat.com> 56839 Jon TURNEY <jon.turney@dronecode.org.uk> 56840 56841commit f69f4f417c6807405e45289b39a80f1b9a55f9e6 56842Author: Keith Packard <keithp@keithp.com> 56843Date: Tue Dec 10 07:09:17 2013 -0800 56844 56845 Depend on latest glproto (1.4.17) 56846 56847 Signed-off-by: Keith Packard <keithp@keithp.com> 56848 56849commit 576f3d36dd324bd0d16c9968f1ded5c5eead3569 56850Author: Adam Jackson <ajax@redhat.com> 56851Date: Mon Dec 9 13:16:01 2013 -0500 56852 56853 damageext: Die if we can't create the Xinerama resource type 56854 56855 Reviewed-by: Keith Packard <keithp@keithp.com> 56856 Signed-off-by: Adam Jackson <ajax@redhat.com> 56857 56858commit f10f36d91db1b21c2ce5a531b4fa5f96fcbdc2f0 56859Author: Adam Jackson <ajax@redhat.com> 56860Date: Mon Sep 16 15:17:26 2013 -0400 56861 56862 damageext: Xineramify (v7) 56863 56864 v7: Don't bother making resources for the backing listeners. [keithp] 56865 56866 This is now slightly unlike how other resources are xineramified. We 56867 create N+1 internal damage listeners, one that's a real resource and 56868 reflects the protocol view, and then one per backend screen where the 56869 report function piles onto the protocol view. The internal listeners 56870 are not stored in the resource database directly, they just hang off the 56871 xinerama resource. We don't wrap Subtract at the dispatch level, but we 56872 do extend it for the Xinerama case to clip to the root window geometry. 56873 56874 As a result of the N+1 design here, the damage reports we generate are 56875 not quite minimal. However they are indistinguishable from sequential 56876 rendering events happening before the client hears damage, and we don't 56877 need to add a post-dispatch callback just for this one extension. 56878 56879 Add is probably (still) somewhat broken since it will only hit screen 0, 56880 but Add really only exists for DRI1's sake, and DRI1 disables itself 56881 with Xinerama enabled anyway. In the absence of a use case, I'm leaving 56882 it unwrapped under Xinerama; if someone wants to define how it ought to 56883 work, be my guest. 56884 56885 Reviewed-by: Keith Packard <keithp@keithp.com> 56886 Signed-off-by: Adam Jackson <ajax@redhat.com> 56887 56888commit 808303fe5232ba0320ae5a4310b1ed1322e85f1d 56889Author: Adam Jackson <ajax@redhat.com> 56890Date: Mon Nov 4 17:49:34 2013 -0500 56891 56892 fixes: Fix PanoramiXSetWindowShapeRegion for root windows (v2) 56893 56894 Root windows in Xinerama are in the coordinate space of their root window 56895 pixmap, not in protocol space. 56896 56897 v2: Only translate for root windows, sice the window shape is 56898 drawable-relative. 56899 56900 Reviewed-by: Keith Packard <keithp@keithp.com> 56901 Signed-off-by: Adam Jackson <ajax@redhat.com> 56902 56903commit 5c10c7ea2129b70015e745523918d143cc29318d 56904Author: Adam Jackson <ajax@redhat.com> 56905Date: Mon Nov 4 17:49:33 2013 -0500 56906 56907 fixes: Fix PanoramiXSetPictureClipRegion for root windows (v2) 56908 56909 Root windows in Xinerama are in the coordinate space of their root 56910 window pixmap, not in protocol space. This fixes 'xcompmgr -n' when 56911 Xinerama is active. 56912 56913 v2: Only translate for root windows, since the clip origin is 56914 drawable-relative. 56915 56916 Reviewed-by: Keith Packard <keithp@keithp.com> 56917 Signed-off-by: Adam Jackson <ajax@redhat.com> 56918 56919commit 1dd839a425adc6e5a1dc377003ed86a374d81f0b 56920Author: Adam Jackson <ajax@redhat.com> 56921Date: Thu Nov 7 13:59:30 2013 -0500 56922 56923 composite: Fix COW creation for Xinerama (v2) 56924 56925 Say you have two 800x600 screens left/right of each other. A window 56926 that's 200x200 at +700+0 in protocol coordinate space will appear to be 56927 at -100+0 in the coordinate space of the right hand screen. Put another 56928 way: windows are in the coordinate space of their root window pixmap. 56929 56930 We weren't doing this translation for the COW, so when rendering came in 56931 to it you'd see the top-left chunk of the COW on all screens. Cool 56932 effect and all, but wrong. 56933 56934 v2: Only translate when Xinerama is active [keithp] 56935 56936 Reviewed-by: Keith Packard <keithp@keithp.com> 56937 Signed-off-by: Adam Jackson <ajax@redhat.com> 56938 56939commit 8dbe456abae1a32fb6c5c74fa456bc7301cccabe 56940Author: Adam Jackson <ajax@redhat.com> 56941Date: Tue Oct 29 10:44:07 2013 -0400 56942 56943 xinerama: Export the screen region 56944 56945 damageext wants this so it can intersect subtract requests against the 56946 root window geometry. 56947 56948 Reviewed-by: Keith Packard <keithp@keithp.com> 56949 Signed-off-by: Adam Jackson <ajax@redhat.com> 56950 56951commit c6d4c2a24140feee4ceae48c4c74fedcae9b68b3 56952Author: Adam Jackson <ajax@redhat.com> 56953Date: Fri Nov 15 15:13:31 2013 -0500 56954 56955 xfree86: Prefer fbdev to vesa 56956 56957 On UEFI machines you'd prefer fbdev to grab efifb instead of vesa trying 56958 to initialize and failing in a way we can't unwind from. On BIOS 56959 machines this is harmless: either there is an fbdev driver and it'll 56960 probably be more capable, or there's not and vesa will kick in anyway. 56961 56962 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 56963 Reviewed-by: Eric Anholt <eric@anholt.net> 56964 Signed-off-by: Adam Jackson <ajax@redhat.com> 56965 56966commit 7b5d4f147fdef9edfeaa9c6565375111079efd11 56967Author: Adam Jackson <ajax@redhat.com> 56968Date: Tue Nov 5 10:12:28 2013 -0500 56969 56970 composite: Don't double-redirect if someone asks for backing store twice 56971 56972 v2: Belt-and-suspenders the unredirection path [vsyrjala] 56973 56974 Reviewed-by: Keith Packard <keithp@keithp.com> 56975 Signed-off-by: Adam Jackson <ajax@redhat.com> 56976 56977commit a2b2c271e0ca87d3188ba2741b6db9bbbdc599f5 56978Author: Adam Jackson <ajax@redhat.com> 56979Date: Tue Nov 5 10:08:17 2013 -0500 56980 56981 composite: Automatically enable backing store support on the screen 56982 56983 ... unless you explicitly disabled it with -bs on the command line, or 56984 with the corresponding thing in xorg.conf. 56985 56986 v2: Drop a bogus hunk from compChangeWindowAttributes [vsyrjala] 56987 v3: s/TRUE/WhenMapped/ [jcristau] 56988 56989 Reviewed-by: Keith Packard <keithp@keithp.com> 56990 Signed-off-by: Adam Jackson <ajax@redhat.com> 56991 56992commit e0cac005608a2e5618c7be59701318d684e0bb93 56993Author: Adam Jackson <ajax@redhat.com> 56994Date: Tue Nov 5 09:58:17 2013 -0500 56995 56996 bs: Set the screen's bs support level to WhenMapped 56997 56998 Since we're using RedirectAutomatic to do this, we don't actually 56999 preserve contents when unmapped. 57000 57001 v2: Don't say WhenMapped if Composite didn't initialize [vsyrjala] 57002 57003 Reviewed-by: Julien Cristau <jcristau@debian.org> 57004 Signed-off-by: Adam Jackson <ajax@redhat.com> 57005 57006commit b61ccd5d9d368f3fbbae27ce14ac02a3db1884c4 57007Author: Adam Jackson <ajax@redhat.com> 57008Date: Tue Nov 5 10:20:04 2013 -0500 57009 57010 smartsched: Tweak the default scheduler intervals 57011 57012 A default timeslice of 20ms means a pathological client can ruin up to 57013 two frames per scheduler tick. And a fifth of a second is just insane. 57014 57015 Pick two different numbers out of the hat. A 5ms slice means you can 57016 probably keep up with two or three abusive clients, and letting it burst 57017 to 15ms should give you about all the timeslice you need for a 57018 fullscreen game (that's doing server-side rendering for some reason). 57019 57020 If you're running on a system with a 10ms granularity on SIGALRM, then 57021 this effectively changes the intervals to 10ms and 30ms. Which is still 57022 better, just not as better. 57023 57024 I suspect this is about as good as we can do without actually going 57025 preemptive, which is an entire other nightmare. 57026 57027 Reviewed-by: Julien Cristau <jcristau@debian.org> 57028 Signed-off-by: Adam Jackson <ajax@redhat.com> 57029 57030commit 66310ea2893811614b3c3aade8992624a95a6ee7 57031Author: Adam Jackson <ajax@redhat.com> 57032Date: Fri Nov 8 09:22:37 2013 -0500 57033 57034 dri3: Disable when Xinerama is active 57035 57036 Pretty sure this can't work. 57037 57038 Reviewed-by: Julien Cristau <jcristau@debian.org> 57039 Signed-off-by: Adam Jackson <ajax@redhat.com> 57040 57041commit 793fd5eefb0e417614d77fe1522c6548587fbd4e 57042Author: Adam Jackson <ajax@redhat.com> 57043Date: Fri Nov 8 09:20:29 2013 -0500 57044 57045 dri2: Disable when Xinerama is active 57046 57047 Would only work on ScreenRec 0, which means it's broken. 57048 57049 Reviewed-by: Julien Cristau <jcristau@debian.org> 57050 Signed-off-by: Adam Jackson <ajax@redhat.com> 57051 57052commit 2bf92108294805c5c6fef249c2f7de6ecf4975d8 57053Author: Adam Jackson <ajax@redhat.com> 57054Date: Fri Nov 8 09:06:59 2013 -0500 57055 57056 present: Disable when Xinerama is active 57057 57058 Among much else Present depends on RANDR types, and RANDR isn't properly 57059 Xinerama-aware yet anyway. 57060 57061 Reviewed-by: Julien Cristau <jcristau@debian.org> 57062 Signed-off-by: Adam Jackson <ajax@redhat.com> 57063 57064commit ac772cb187ddf7e04b8f4b3a071b90f18f4488d0 57065Author: Eric Anholt <eric@anholt.net> 57066Date: Thu Nov 14 17:40:47 2013 -0800 57067 57068 glx: Fix incorrect use of dri_interface.h version defines in driver probing. 57069 57070 If we extend __DRI_CORE or __DRI_SWRAST in dri_interface.h to allow a 57071 new version, it shouldn't make old server code retroactively require 57072 the new version from swrast drivers. 57073 57074 Notably, new Mesa defines __DRI_SWRAST version 4, but we still want to 57075 be able to probe version 1 drivers, since we don't use any features 57076 beyond version 1 of the struct. 57077 57078 Signed-off-by: Keith Packard <keithp@keithp.com> 57079 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 57080 57081commit 6e926b18ca1b182253bac435a1d53caaff7ffff6 57082Author: Eric Anholt <eric@anholt.net> 57083Date: Thu Nov 14 17:40:46 2013 -0800 57084 57085 glx: Fix incorrect use of dri_interface.h version defines in extensions. 57086 57087 Those defines are so you can compile-time check "do I have a 57088 dri_interface.h that defines this new field of the struct?" You don't 57089 want the server to claim it implements the new struct just because you 57090 installed a new copy of Mesa. 57091 57092 Signed-off-by: Keith Packard <keithp@keithp.com> 57093 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 57094 57095commit 57a8ce927332e855dd29ff30210a211a907adb25 57096Author: Jon TURNEY <jon.turney@dronecode.org.uk> 57097Date: Fri Nov 29 12:26:43 2013 +0000 57098 57099 configure.ac: Fixup for "Require libpciaccess for int10" 57100 57101 On 16/11/2013 01:00, Connor Behan wrote: 57102 > A --disable-pciaccess build will fail with an int10 module other than 57103 > stub. 57104 > 57105 > Signed-off-by: Connor Behan <connor.behan-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> 57106 > --- 57107 > configure.ac | 3 +++ 57108 > 1 file changed, 3 insertions(+) 57109 > 57110 > diff --git a/configure.ac b/configure.ac 57111 > index 5e621e0..a843770 100644 57112 > --- a/configure.ac 57113 > +++ b/configure.ac 57114 > @@ -1822,6 +1822,9 @@ if test "x$XORG" = xyes; then 57115 > if test "x$CONFIG_UDEV_KMS" = xyes; then 57116 > AC_MSG_ERROR([Platform device enumeration requires libpciaccess]) 57117 > fi 57118 > + if test "x$INT10" != xstub; then 57119 > + AC_MSG_ERROR([Cannot build int10 without libpciaccess]) 57120 > + fi 57121 > fi 57122 > AC_MSG_RESULT([$PCI]) 57123 > 57124 57125 This causes my build to fail where --disable-int10-module --disable-pciaccess 57126 is the default (as INT10 still has the default value 'x86emu') 57127 57128 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 57129 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57130 Tested-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57131 Signed-off-by: Keith Packard <keithp@keithp.com> 57132 57133commit 653d33941b0808ef910aaa5f3aeab05d9c1a100b 57134Author: Keith Packard <keithp@keithp.com> 57135Date: Tue Nov 26 12:18:12 2013 -0800 57136 57137 present: Report damage when flipping 57138 57139 Limit damage to the 'update' region. 57140 57141 Signed-off-by: Keith Packard <keithp@keithp.com> 57142 Reviewed-by: Adam Jackson <ajax@redhat.com> 57143 57144commit 5cf12c9569ac3d83fe1b7a8376c15f8f0b01655e 57145Author: Keith Packard <keithp@keithp.com> 57146Date: Tue Nov 26 12:10:48 2013 -0800 57147 57148 present: Also set the root window pixmap when flipping 57149 57150 This makes sure that things like software cursors continue to work 57151 while the screen is flipped. 57152 57153 Signed-off-by: Keith Packard <keithp@keithp.com> 57154 Reviewed-by: Adam Jackson <ajax@redhat.com> 57155 57156commit 4aa77378de69efdc10bced6ba650b0ebff50c112 57157Author: Keith Packard <keithp@keithp.com> 57158Date: Tue Nov 26 12:06:57 2013 -0800 57159 57160 present: Clear target_crtc if driver lacks Present support 57161 57162 If the driver doesn't have the necessary hooks for Present, then the 57163 target_crtc needs to be set to NULL to make sure the extension uses 57164 the present_fake code. 57165 57166 Signed-off-by: Keith Packard <keithp@keithp.com> 57167 Tested-by: Fredrik Höglund <fredrik@kde.org> 57168 Reviewed-by: Adam Jackson <ajax@redhat.com> 57169 57170commit 3dd5bfe540b295bb37a2c2fd0ba4a31fb217612b 57171Author: Keith Packard <keithp@keithp.com> 57172Date: Thu Nov 21 22:48:31 2013 -0800 57173 57174 present: Send GLX_BufferSwapComplete events from present extension 57175 57176 This allows GL to support the GLX_INTEL_swap_event extension. 57177 57178 v2: Return GLX_BLIT_COMPLETE_INTEL for unknown swap types 57179 57180 Signed-off-by: Keith Packard <keithp@keithp.com> 57181 Reviewed-by: Adam Jackson <ajax@redhat.com> 57182 57183commit cde86e68fcb716f34c90f5a16eb868870f5c85e4 57184Author: Keith Packard <keithp@keithp.com> 57185Date: Mon Nov 25 23:06:08 2013 -0800 57186 57187 present: Set window pixmap to flipped pixmap 57188 57189 This makes other drawing to the window appear on the screen. 57190 57191 Note that no child windows can be affected because only full-screen 57192 windows are eligible for flipping, and so we only need to set pixmap 57193 for the window itself. 57194 57195 Signed-off-by: Keith Packard <keithp@keithp.com> 57196 Reviewed-by: Adam Jackson <ajax@redhat.com> 57197 57198commit 04e138846e128670d409798aa2e797c3c5503a47 57199Author: Keith Packard <keithp@keithp.com> 57200Date: Mon Nov 25 23:10:17 2013 -0800 57201 57202 present: Leave vblank on window list until flip complete 57203 57204 If the window is destroyed, then we've got cleanup work to do, even if 57205 the vblank has already been executed -- we need to clear the window 57206 pointer so that we don't try to deliver events to it. 57207 57208 Leaving it on the window list meant that when walking that list, we 57209 need to know whether the vblank is waiting to be executed or waiting 57210 for the flip to complete, so a new 'queued' flag was added to the 57211 vblank to distinguish between the two states. 57212 57213 Signed-off-by: Keith Packard <keithp@keithp.com> 57214 Reviewed-by: Adam Jackson <ajax@redhat.com> 57215 57216commit b121d62accb8c346b4e1b1bce99586e13712f04a 57217Author: Keith Packard <keithp@keithp.com> 57218Date: Tue Nov 26 03:07:55 2013 -0800 57219 57220 present: Add a debug output line when skipping a pending present 57221 57222 When an application provides two pixmaps for the same MSC, the 57223 previous one is skipped. This just dumps out some information at that point 57224 57225 Signed-off-by: Keith Packard <keithp@keithp.com> 57226 Reviewed-by: Adam Jackson <ajax@redhat.com> 57227 57228commit ae24f5bf6b31c035af50e14727966055ed8767bd 57229Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 57230Date: Tue Dec 3 17:01:28 2013 +0100 57231 57232 bump to 0.8.1 57233 57234commit 59e324abd7cbda5d301b7b67125cd6d96684f000 57235Author: Thierry Reding <thierry.reding@gmail.com> 57236Date: Tue Nov 26 17:08:03 2013 +0100 57237 57238 Properly identify DSI outputs 57239 57240 Newer Linux kernels support DSI outputs. To be able to identify them 57241 properly, add DSI to the list of output names. 57242 57243 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 57244 Signed-off-by: Thierry Reding <treding@nvidia.com> 57245 57246commit 190289f6393191bb3c58067ce1df994809475077 57247Author: Thierry Reding <thierry.reding@gmail.com> 57248Date: Tue Nov 26 17:08:02 2013 +0100 57249 57250 Staticise and constify output names 57251 57252 This array isn't used anywhere outside this file, so it can be made 57253 static. While at it, make the array const as well. 57254 57255 Signed-off-by: Thierry Reding <treding@nvidia.com> 57256 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 57257 57258commit eafba23b34be31c141ddafb8380520ac9a0622ac 57259Author: Keith Packard <keithp@keithp.com> 57260Date: Thu Nov 21 22:45:18 2013 -0800 57261 57262 miext/sync: Handle libxshmfence API change 57263 57264 libxshmfence had an unfortunate 'int32_t' type for the mapped fence. 57265 That changed to exposing a 'struct shmfence' instead, which is nice 57266 and opaque and offers fine type checking across the API. 57267 57268 This patch requires the newer version of the library and uses 57269 the new interface type. 57270 57271 Signed-off-by: Keith Packard <keithp@keithp.com> 57272 Reviewed-by: Julien Cristau <jcristau@debian.org> 57273 57274commit b6d7ed4d787a652e8150532f384bfdf51760f3c2 57275Author: Keith Packard <keithp@keithp.com> 57276Date: Thu Nov 21 22:12:34 2013 -0800 57277 57278 miext: Move SyncShm FDs out of the way of clients 57279 57280 Applications may end up allocating a bunch of shmfence objects, each 57281 of which uses a file descriptor, which must be kept open lest some 57282 other client ask for a copy of it later on. 57283 57284 Lacking an API that can turn a memory mapping back into a file 57285 descriptor, about the best we can do is push the file descriptors out 57286 of the way of other X clients so that we don't run out of the ability 57287 to accept new connections. 57288 57289 This uses fcntl F_GETFD to push the FD up above MAXCLIENTS. 57290 57291 Signed-off-by: Keith Packard <keithp@keithp.com> 57292 Reviewed-by: Julien Cristau <jcristau@debian.org> 57293 57294commit cc63204926c6da83d9221c5f8c0dc8f5e2f2481d 57295Author: Keith Packard <keithp@keithp.com> 57296Date: Wed Nov 13 14:16:33 2013 +0900 57297 57298 Xext: Use SHMDIR and O_TMPFILE when creating mapping files 57299 57300 ShmCreateSegment asks for a file descriptor for a memory mapped file 57301 created by the X server. This patch uses O_TMPFILE where available, 57302 and also uses the SHMDIR directory to store the files, both for the 57303 O_TMPFILE and mkstemp cases. 57304 57305 Signed-off-by: Keith Packard <keithp@keithp.com> 57306 Reviewed-by: Julien Cristau <jcristau@debian.org> 57307 57308commit 5a969f0928b84da5cfe0777dfb542caaacc915ad 57309Author: Keith Packard <keithp@keithp.com> 57310Date: Wed Nov 13 12:17:10 2013 +0900 57311 57312 Select directory for MIT-SHM temp files at configure time 57313 57314 By default, this looks through a list of directories to find one which 57315 exists, but can be overridden with --with-shared-memory-dir=PATH 57316 57317 This patch doesn't actually do anything with this directory, just 57318 makes it available in the configuration 57319 57320 Signed-off-by: Keith Packard <keithp@keithp.com> 57321 Reviewed-by: Julien Cristau <jcristau@debian.org> 57322 57323commit d695cb7fb6bb8fd159a332d902673edbbd2237d7 57324Author: Alan Coopersmith <alan.coopersmith@oracle.com> 57325Date: Mon Nov 25 22:19:34 2013 -0800 57326 57327 Xdmx: Initialize DMX extension even if not built with GLX support 57328 57329 dmxAddExtensions takes an argument to determine if it should just 57330 initialize the DMX extension, or both DMX & GLX, but if GLX wasn't 57331 compiled in, the entire call was #ifdef'ed out, leaving the DMX 57332 extension unavailable. 57333 57334 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 57335 Reviewed-by: Julien Cristau <jcristau@debian.org> 57336 Signed-off-by: Keith Packard <keithp@keithp.com> 57337 57338commit bd70def07754357d26b86c287037eb3884631daf 57339Author: Connor Behan <connor.behan@gmail.com> 57340Date: Fri Nov 15 17:00:40 2013 -0800 57341 57342 configure.ac: Require libpciaccess for int10 57343 57344 A --disable-pciaccess build will fail with an int10 module other than 57345 stub. 57346 57347 Signed-off-by: Connor Behan <connor.behan@gmail.com> 57348 Reviewed-by: Adam Jackson <ajax@redhat.com> 57349 Signed-off-by: Keith Packard <keithp@keithp.com> 57350 57351commit a575c1dc64c10bc88bd0e963ebb8ab137f59a078 57352Author: Connor Behan <connor.behan@gmail.com> 57353Date: Fri Nov 15 17:00:20 2013 -0800 57354 57355 configure.ac: Require libpciaccess for platform bus support 57356 57357 There is currently no reason to build with --enable-config-udev-kms and 57358 --disable-pciaccess but anyone who tries this should know that the build 57359 will fail. 57360 57361 Signed-off-by: Connor Behan <connor.behan@gmail.com> 57362 Reviewed-by: Adam Jackson <ajax@redhat.com> 57363 Signed-off-by: Keith Packard <keithp@keithp.com> 57364 57365commit 361dfe74eaa2f1eddb54abdecca1d8b1b3f58704 57366Author: Connor Behan <connor.behan@gmail.com> 57367Date: Fri Nov 15 17:00:00 2013 -0800 57368 57369 configure.ac: Add whitespace near PCI configuration 57370 57371 Signed-off-by: Connor Behan <connor.behan@gmail.com> 57372 Reviewed-by: Adam Jackson <ajax@redhat.com> 57373 Signed-off-by: Keith Packard <keithp@keithp.com> 57374 57375commit 80481267662c8687e73081237913fa561e7a6561 57376Author: Keith Packard <keithp@keithp.com> 57377Date: Sat Nov 23 22:22:37 2013 -0800 57378 57379 Bump release to 1.14.99.903 (1.15 RC3) 57380 57381 Signed-off-by: Keith Packard <keithp@keithp.com> 57382 57383commit f1604002a32b7f098c2a16b4a8649c694af570c8 57384Author: Keith Packard <keithp@keithp.com> 57385Date: Mon Nov 18 22:36:17 2013 -0800 57386 57387 miext: Ensure xshmfence is only called when driver supports it 57388 57389 This provides a place for drivers to insert their own FD-based 57390 SyncFence implementations, and prevents applications from using DRI3 57391 SyncFence creation functions unless the driver has some support for 57392 them. 57393 57394 Signed-off-by: Keith Packard <keithp@keithp.com> 57395 Tested-by: Fredrik Höglund <fredrik@kde.org> 57396 57397commit 037566c57caff93fd7717f385d4b532b81f19c77 57398Author: Keith Packard <keithp@keithp.com> 57399Date: Mon Nov 18 22:33:27 2013 -0800 57400 57401 Xext: Recover from miSyncInitFenceFromFD failure without crashing 57402 57403 miSyncDestroyFence must not be called unless miSyncInitFence has been 57404 invoked, so if miSyncInitFenceFromFD fails, we must free the fence 57405 manually. 57406 57407 Signed-off-by: Keith Packard <keithp@keithp.com> 57408 Tested-by: Fredrik Höglund <fredrik@kde.org> 57409 57410commit e7000534a456fdf9cd4eaada3193846c8525f740 57411Author: Chris Wilson <chris@chris-wilson.co.uk> 57412Date: Sat Oct 5 08:47:03 2013 +0100 57413 57414 glx/glxdri2: Unwrap EnterVT/LeaveVT upon CloseScreen 57415 57416 In a similar spirit to 57417 57418 commit d75e8146c414bfd512ba5dbd4a83acb334bbe19b 57419 Author: Keith Packard <keithp@keithp.com> 57420 Date: Mon Jul 12 16:01:34 2010 -0700 57421 57422 Unwrap/rewrap EnterVT/LeaveVT completely, Fixes 28998 57423 57424 we need to unwrap our pScrn->EnterVT/LeaveVT hooks around server 57425 regeneration or else we cause an infinite recursion on the next VT 57426 switch afterwards. 57427 57428 Bugzilla: https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/1235516 57429 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 57430 Reviewed-by: Keith Packard <keithp@keithp.com> 57431 Signed-off-by: Keith Packard <keithp@keithp.com> 57432 57433commit 6d5883bd7e5b765f8f0548501b825d9e56840799 57434Author: Keith Packard <keithp@keithp.com> 57435Date: Sat Nov 23 16:19:46 2013 -0800 57436 57437 xnest: Ignore GetImage() error in xnestGetImage() (v3) 57438 57439 (I found an amended version of this patch and applied the difference 57440 here - keithp) 57441 57442 v3: Don't call Xsync before restoring error handler as any errors 57443 generated by XGetImage() should be processed when this call 57444 returns as suggested by Jamey Sharp <jamey@minilop.net> 57445 57446 Signed-off-by: Egbert Eich <eich@freedesktop.org> 57447 Reviewed-by: Jamey Sharp <jamey@minilop.net> 57448 57449commit 6403cbb143c67872ca9c58e3116ae7942def0ae1 57450Author: Keith Packard <keithp@keithp.com> 57451Date: Tue Nov 19 22:13:54 2013 -0800 57452 57453 present: When unflipping, copy to flip window rather than screen 57454 57455 unflip happens after the clip lists have been updated, so instead of 57456 smashing the whole screen and drawing over other windows, just draw to 57457 the original flip window; it'll have the right clip list and so the 57458 copy will work just fine. 57459 57460 Signed-off-by: Keith Packard <keithp@keithp.com> 57461 Reviewed-by: Adam Jackson <ajax@redhat.com> 57462 57463commit 8bdd2ccc776ded3f527596b5009ef25129aa3287 57464Author: Keith Packard <keithp@keithp.com> 57465Date: Mon Nov 11 18:03:42 2013 -0800 57466 57467 present: Block for wait_fence in present_execute 57468 57469 Pend presentation until wait_fence is also triggered by having the 57470 SyncFence trigger invoke present_execute once triggered. 57471 57472 Signed-off-by: Keith Packard <keithp@keithp.com> 57473 Reviewed-by: Adam Jackson <ajax@redhat.com> 57474 57475commit e5a188cb919edee2e3a03054276bce0db02f7b62 57476Author: Keith Packard <keithp@keithp.com> 57477Date: Sat Nov 9 12:36:47 2013 -0800 57478 57479 present: Signal destroyed flip window with vblank->window == NULL 57480 57481 This eliminates dereferencing freed window pointers when there is a 57482 flip for that window in progress. The flip will complete, and then 57483 immediately get undone (as we can't stop an in-progress flip). 57484 57485 Remove the vblank->window_destroyed field as we can signal this with 57486 vblank->window == NULL instead. 57487 57488 Change check to vblank->window == NULL in: 57489 57490 present_flip_notify 57491 57492 Add check for vblank->window == NULL in: 57493 present_vblank_notify 57494 present_execute 57495 57496 present_flip_notify was also using vblank->window->drawable.pScreen, 57497 so stop doing that and use vblank->screen instead. 57498 57499 Signed-off-by: Keith Packard <keithp@keithp.com> 57500 Reviewed-by: Adam Jackson <ajax@redhat.com> 57501 57502commit a5bcc4f7b9499caf8993f0a6ef96088553399ca3 57503Author: Keith Packard <keithp@keithp.com> 57504Date: Sat Nov 9 12:33:02 2013 -0800 57505 57506 present: Ignore event_id 0 from driver vblank notify 57507 57508 We use event_id 0 to mean 'no such event'; if a driver sends us that 57509 event_id, make sure we don't accidentally match it. 57510 57511 Signed-off-by: Keith Packard <keithp@keithp.com> 57512 Reviewed-by: Adam Jackson <ajax@redhat.com> 57513 57514commit 4f3c37a1f17ffcfbbff71d217e1caad3d0148c90 57515Author: Keith Packard <keithp@keithp.com> 57516Date: Thu Nov 7 14:17:12 2013 -0800 57517 57518 present: Round fake MSC computations intead of truncating 57519 57520 If the timer fired too early, we'd sometimes mis-compute the MSC for 57521 fake vblanks. Rounding the computation to the nearest MSC fixes this nicely. 57522 57523 Signed-off-by: Keith Packard <keithp@keithp.com> 57524 Reviewed-by: Adam Jackson <ajax@redhat.com> 57525 57526commit da9997f89f14ab619f244d5b2e80a423b028c789 57527Author: Peter Hutterer <peter.hutterer@who-t.net> 57528Date: Tue Nov 19 08:07:09 2013 +1000 57529 57530 configure: allow for --enable-libunwind and --disable-libunwind 57531 57532 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 57533 Reviewed-by: Keith Packard <keithp@keithp.com> 57534 Signed-off-by: Keith Packard <keithp@keithp.com> 57535 57536commit a307ac27382dd0c377cd1a436cb25cb5aaa54416 57537Author: Keith Packard <keithp@keithp.com> 57538Date: Fri Nov 15 13:39:52 2013 +0900 57539 57540 Handle new DamageUnregister API which has only one argument 57541 57542 API change in 1.15 57543 57544 Signed-off-by: Keith Packard <keithp@keithp.com> 57545 57546commit d6ffce4b0e2376e1f8f04ee70b323a6aea3e517e 57547Author: Keith Packard <keithp@keithp.com> 57548Date: Fri Nov 15 13:26:36 2013 +0900 57549 57550 Clean up compiler warnings. 57551 57552 Add const to any immutable string pointers. 57553 Rename 'range' to 'prop_range' to avoid redefined warning. 57554 Eliminate some unused return values. 57555 57556 Signed-off-by: Keith Packard <keithp@keithp.com> 57557 57558commit 29240e5cbf6e7f875b128fc7bfc4d56b2350835a 57559Merge: 0492deb8f 8ff7e32c3 57560Author: Keith Packard <keithp@keithp.com> 57561Date: Thu Nov 14 17:02:04 2013 +0900 57562 57563 Merge remote-tracking branch 'whot/for-keith' 57564 57565commit 0492deb8f8238b7782e5a706ec6219d88aa1091d 57566Author: Adam Jackson <ajax@redhat.com> 57567Date: Tue Oct 29 12:09:27 2013 -0400 57568 57569 mieq: Bump default queue size to 512 57570 57571 Based on some bugzilla scraping I did around November 2012. Of xserver 57572 bugs in Red Hat bugzilla with an EQ size message in the log, the 57573 distribution looked like: 57574 57575 String | Matches 57576 ------------------------------------- 57577 Increasing EQ size to 512 | 460 57578 Increasing EQ size to 1024 | 52 57579 Increasing EQ size to 2048 | 6 57580 Increasing EQ size to 4096 | 0 57581 57582 Most of the "512" ones appear to be mostly harmless, some relatively 57583 expensive path in either rendering or resource destruction simply taking 57584 too long due to external pressures like paging or CPU contention. So 57585 let's raise the initial queue size, both to reduce the number of 57586 spurious abrt reports and to drop fewer events in all but the most 57587 pathological cases. 57588 57589 Signed-off-by: Adam Jackson <ajax@redhat.com> 57590 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 57591 Signed-off-by: Keith Packard <keithp@keithp.com> 57592 57593commit d1440783a7367ff0d0c47d256bbca3b3cf8a5034 57594Author: Dave Airlie <airlied@redhat.com> 57595Date: Tue Oct 29 12:09:26 2013 -0400 57596 57597 xfree86: return NULL for compat output if no outputs. 57598 57599 With outputless GPUs showing up we crash here if there are not outputs 57600 try and recover with a bit of grace. 57601 57602 Reviewed-by: Adam Jackson <ajax@redhat.com> 57603 Signed-off-by: Dave Airlie <airlied@redhat.com> 57604 Signed-off-by: Keith Packard <keithp@keithp.com> 57605 57606commit d7ee27e5e415778240919082c83a65226c6f17e6 57607Author: Dan Horák <dan@danny.cz> 57608Date: Tue Oct 29 12:09:25 2013 -0400 57609 57610 test: build the touch test only when building Xorg 57611 57612 Reviewed-by: Adam Jackson <ajax@redhat.com> 57613 Signed-off-by: Dan Horák <dan@danny.cz> 57614 Signed-off-by: Keith Packard <keithp@keithp.com> 57615 57616commit 8ff7e32c3ef7b0c13c4ab9664f651e9782d35a85 57617Author: Peter Hutterer <peter.hutterer@who-t.net> 57618Date: Wed Nov 13 17:14:11 2013 +1000 57619 57620 include: export key_is_down and friends 57621 57622 VNC needs key_is_down to check if a key is processed as down before it 57623 simulates various key releases. Make it available, because I seriously can't 57624 be bothered thinking about how to rewrite VNC to not need that. 57625 57626 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 57627 Acked-by: Keith Packard <keithp@keithp.com> 57628 57629commit 0ba7fc8472f1227fd1cd51bd58048f164b8ea1ab 57630Author: Alan Coopersmith <alan.coopersmith@oracle.com> 57631Date: Tue Nov 12 14:46:04 2013 -0800 57632 57633 Stop including inline assembly .il file for Solaris Studio builds 57634 57635 Since all the inb/outb/etc. use in the X server itself (except for 57636 xf86SlowBcopy) has been replaced by calls to libpciaccess, we no 57637 longer need to pass inline assembly files to replace the gcc inline 57638 assembly from hw/xfree86/common/compiler.h when building Xorg itself. 57639 57640 The .il files are still generated and installed in the SDK for the 57641 benefit of drivers who may use them. 57642 57643 Binary diff of before and after showed that xf86SlowBcopy was the 57644 only function changed across the Xorg binary and all modules built 57645 in the Xserver build, it just calls the outb() function now instead 57646 of having the outb instructions inlined, making it a slightly slower 57647 bcopy. 57648 57649 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 57650 Reviewed-by: Keith Packard <keithp@keithp.com> 57651 Signed-off-by: Keith Packard <keithp@keithp.com> 57652 57653commit 550baf38f6096658f0bcf0ad647c4fedf93132f2 57654Author: Peter Hutterer <peter.hutterer@who-t.net> 57655Date: Fri Oct 4 10:55:52 2013 +1000 57656 57657 kdrive: fix cursor jumps on CursorOffScreen behavior 57658 57659 This patch fixes cursor jumps when there is a grab on the Xephyr window and 57660 the pointer moves outside the window. 57661 57662 So on two side-by-side 640x480 screens, a coordinate of 0/481 57663 triggers KdCursorOffscreen. 57664 57665 If the delta between two screens is 0, they share the same offset for 57666 that dimension. When searching for the new screen, the loop always rules out 57667 the current screen. So we get to the second screen, trigger the conditions 57668 where dy <= 0 and decide that this new screen is the correct one. The result 57669 is that whenever KdCursorOffScreen is called, the pointer jumps to the other 57670 screen. 57671 57672 Change to check for dy < 0 etc. so that the cursor stays on the same screen if 57673 there is no other screen at the target location. 57674 57675 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 57676 Reviewed-by: Keith Packard <keithp@keithp.com> 57677 57678commit a94d945065177d73f3ee8dc0b9147264ba281136 57679Author: Peter Hutterer <peter.hutterer@who-t.net> 57680Date: Tue Oct 29 14:24:41 2013 +1000 57681 57682 kdrive: modify ephyr events to use POINTER_DESKTOP and scale them to that 57683 57684 A multi-head Xephyr instance has the pointer stuck on one screen 57685 because of bad coordinate calculation. The coordinates passed to 57686 GetPointerEvents are per-screen, so the cursor gets stuck on the left-most 57687 screen by default. 57688 57689 Adjust and mark the events as POINTER_DESKTOP, so the DIX 57690 can adjust them accordingly. 57691 57692 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 57693 Reviewed-by: Keith Packard <keithp@keithp.com> 57694 57695commit d66832a3b8a8675f1e5f3656dcb1bbe95598f0ea 57696Author: Keith Packard <keithp@keithp.com> 57697Date: Thu Nov 14 13:21:54 2013 +1000 57698 57699 kdrive: handle WxH as valid geometry spec 57700 57701 If a screen size was specified as WxH, the loop returned early and kdOrigin 57702 was never advanced. Thus, screen->origin was always 0 (or whatever was given 57703 at the -origin commandline flag). 57704 57705 If a screen size was given with a bit depth (WxHxD), kdOrigin would always 57706 advance by the current screen, offsetting the next screen. 57707 57708 Signed-off-by: Keith Packard <keithp@keithp.com> 57709 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 57710 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 57711 57712commit c8c5105c1d5c093675d7f571f158147f22f7572b 57713Author: Peter Hutterer <peter.hutterer@who-t.net> 57714Date: Tue Oct 29 14:33:56 2013 +1000 57715 57716 ephyr: xcb_connect returns an error, not NULL 57717 57718 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 57719 Reviewed-by: Keith Packard <keithp@keithp.com> 57720 57721commit 2aa67ccef6ad8f51deb94b66defc668ddb7d226c 57722Author: Keith Packard <keithp@keithp.com> 57723Date: Thu Nov 14 10:26:20 2013 +0900 57724 57725 Update to version 1.14.99.902 57726 57727 1.15 RC2 57728 57729 Signed-off-by: Keith Packard <keithp@keithp.com> 57730 57731commit 9309d2a009a6a3b6090fc377fc7a3eded1a34dc0 57732Author: Keith Packard <keithp@keithp.com> 57733Date: Wed Nov 13 07:01:45 2013 +0900 57734 57735 xext: Fix shmint.h to not use headers outside of sdk_HEADERS 57736 57737 shmint.h is part of sdk_HEADERS, and so can't use anything not 57738 included in sdk_HEADERS. 57739 57740 busfault.h includes dix-config.h which is not. Leave the use of 57741 struct busfault in shmint.h and move the include of busfault.h to 57742 shm.c. 57743 57744 protocol-versions.h is not part of sdk_HEADERS, so instead of using 57745 that, just use XTRANS_SEND_FDS to choose whether to expose the fd 57746 passing requests directly. 57747 57748 Reported-by: Julien Cristau <jcristau@debian.org> 57749 Signed-off-by: Keith Packard <keithp@keithp.com> 57750 Tested-by: Knut Petersen <Knut_Petersen@t-online.de> 57751 Reviewed-by: Julien Cristau <jcristau@debian.org> 57752 57753 v2: also avoid using protocol-versions.h 57754 57755commit 4a251f5883b042cd902c192060a0be2b11148f2b 57756Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57757Date: Fri Nov 8 14:55:33 2013 -0800 57758 57759 xfree86: Fix build without libpciaccess 57760 57761 Regression fix from commit 04ab07ca19236d6c9a947e065fb69b0dd0d16639 57762 57763 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57764 Reviewed-by: Connor Behan <connor.behan@gmail.com> 57765 57766commit 5b02d5b7aaabf1ba8dcbdfe4525b7995b4e79f92 57767Author: Alan Coopersmith <alan.coopersmith@oracle.com> 57768Date: Thu Nov 7 21:08:56 2013 -0800 57769 57770 Enable XTRANS_SEND_FDS on Solaris too. 57771 57772 Requires passing through the __EXTENSIONS__ and _XOPEN_SOURCE defines 57773 in order to expose the msg_control members in struct msghdr. 57774 57775 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 57776 Reviewed-by: Adam Jackson <ajax@redhat.com> 57777 57778commit c4c154d18ef42dc550f2675a9ee88e07b1ca6bed 57779Author: Alan Coopersmith <alan.coopersmith@oracle.com> 57780Date: Thu Nov 7 21:05:33 2013 -0800 57781 57782 Avoid conflicts with Solaris <sys/regset.h> defines that clash with our names 57783 57784 When building on Solaris with _XOPEN_SOURCE set to a recent XPG release, 57785 <stdlib.h> and other core headers start including <sys/regset.h>, which 57786 has a bunch of unfortunately named macros such as "CS", "ES", etc. for 57787 x86 & x64 registers which clash with existing variable & struct member 57788 names in Xorg - so #undef these so they don't interfere with our use. 57789 57790 (Yes, have filed a bug against the system headers for exposing these, 57791 but this solves the problem for building on existing releases.) 57792 57793 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 57794 Reviewed-by: Adam Jackson <ajax@redhat.com> 57795 57796commit d7f9be0f8541368bf1a095ad1fbe7f38be6d3d15 57797Author: Keith Packard <keithp@keithp.com> 57798Date: Mon Nov 11 15:48:41 2013 -0800 57799 57800 Proper spelling of MAP_ANONYMOUS is MAP_ANON. 57801 57802 The former doesn't exist on BSD and the latter is available everywhere 57803 AFAIK (checked Solaris and Linux). 57804 57805 You also might want to wrap that line ;). 57806 57807 Reported-by: Mark Kettenis <mark.kettenis@xs4all.nl> 57808 Signed-off-by: Keith Packard <keithp@keithp.com> 57809 57810commit fc84166e65c35ad75f566b135dcfc305ad5a2fde 57811Author: Keith Packard <keithp@keithp.com> 57812Date: Thu Nov 7 15:18:28 2013 -0800 57813 57814 Get rid of the rest of the FD passing code when XTRANS_SEND_FDS isn't set 57815 57816 req_fds and SetReqFds in include/dixstruct.h 57817 57818 ReadFdFromClient, WriteFdToClient and the FD flushing in os/io.c 57819 57820 Signed-off-by: Keith Packard <keithp@keithp.com> 57821 57822commit a239e6faf3fce848ac0d10c48f8e817db68a493c 57823Merge: 43e5a43fa f70a8bf37 57824Author: Keith Packard <keithp@keithp.com> 57825Date: Mon Nov 11 15:26:12 2013 -0800 57826 57827 Merge remote-tracking branch 'jeremyhu/master' 57828 57829commit 43e5a43fa8994e50cf01dd954118f4ef2c4e7933 57830Author: Adam Jackson <ajax@redhat.com> 57831Date: Thu Nov 7 12:01:41 2013 -0500 57832 57833 present: Don't try to initialize when building without present support 57834 57835 There's a --disable-present, so it'd be nice if it worked. 57836 57837 Signed-off-by: Adam Jackson <ajax@redhat.com> 57838 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 57839 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 57840 Signed-off-by: Keith Packard <keithp@keithp.com> 57841 57842commit 0822a23e048b12c98c654e8b6af711c5f2c97141 57843Author: Keith Packard <keithp@keithp.com> 57844Date: Thu Nov 7 03:19:42 2013 -0800 57845 57846 present: Change debug output a bit to help diagnose missing vblank signals 57847 57848 Signed-off-by: Keith Packard <keithp@keithp.com> 57849 Reviewed-by: Adam Jackson <ajax@redhat.com> 57850 57851commit 20bb49ae9ba11d3dccfba191483cd682d9c9d96c 57852Author: Keith Packard <keithp@keithp.com> 57853Date: Wed Nov 6 23:56:26 2013 -0800 57854 57855 Present: Check for window/fence destroyed when idling pixmaps 57856 57857 A client destroying objects in the middle of an unflip can end up 57858 having the screen flip window or fence set to NULL in the unflip 57859 notify path. Check for these and don't try to use those objects. 57860 57861 Signed-off-by: Keith Packard <keithp@keithp.com> 57862 Reviewed-by: Adam Jackson <ajax@redhat.com> 57863 57864commit 41da295eb50fa08eaacd0ecde99f43a716fcb41a 57865Author: Keith Packard <keithp@keithp.com> 57866Date: Sun Nov 3 13:12:40 2013 -0800 57867 57868 Trap SIGBUS to handle truncated shared memory segments 57869 57870 If a client passes a section of memory via file descriptor and then 57871 subsequently truncates that file, the underlying pages will be freed 57872 and the addresses invalidated. Subsequent accesses to the page will 57873 fail with a SIGBUS error. 57874 57875 Trap that SIGBUS, figure out which segment was causing the error and 57876 then allocate new pages to fill in for that region. Mark the offending 57877 shared segment as invalid and free the resource ID so that the client 57878 will be able to tell when subsequently attempting to use the segment. 57879 57880 Signed-off-by: Keith Packard <keithp@keithp.com> 57881 57882 v2: Use MAP_FIXED to simplify the recovery logic (Mark Kettenis) 57883 v3: Also catch errors in ShmCreateSegment 57884 57885 Conflicts: 57886 include/dix-config.h.in 57887 include/xorg-config.h.in 57888 57889commit 719e880d7698d92f9b854b217ef9680aaa446f2e 57890Author: Keith Packard <keithp@keithp.com> 57891Date: Wed Nov 6 23:22:51 2013 -0800 57892 57893 Require libXtrans version 1.3.2 57894 57895 This has the FD passing support included 57896 57897 Signed-off-by: Keith Packard <keithp@keithp.com> 57898 Reviewed-by: Adam Jackson <ajax@redhat.com> 57899 57900commit bee2ec54049377e0033d49abff20d7bd069c62aa 57901Author: Keith Packard <keithp@keithp.com> 57902Date: Sun Nov 3 10:08:15 2013 -0800 57903 57904 Xext: Enable MIT-SHM FD-passing request definitions only when possible 57905 57906 Check to see if xtrans FD passing is available and use that to 57907 advertise the appropriate version of the SHM extension 57908 57909 Signed-off-by: Keith Packard <keithp@keithp.com> 57910 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 57911 57912commit ea8acfe3e2f74a46c3f1e91809b4b99af18502b7 57913Author: Keith Packard <keithp@keithp.com> 57914Date: Wed Nov 6 23:06:28 2013 -0800 57915 57916 Require xextproto version 7.2.99.901 57917 57918 This includes the MIT-SHM FD passing requests 57919 57920 Signed-off-by: Keith Packard <keithp@keithp.com> 57921 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 57922 57923commit 914672fefacc15386041e7c2bc3beed4faf45e9d 57924Author: Keith Packard <keithp@keithp.com> 57925Date: Thu Nov 7 14:20:26 2013 -0800 57926 57927 Remove dix-config.h config variables from xorg-config.h 57928 57929 xorg-config.h includes dix-config, so there's no need to duplicate. 57930 57931 Signed-off-by: Keith Packard <keithp@keithp.com> 57932 57933commit 7aad79c5a582ece301d950bd65f0bcb4b9956e86 57934Author: Keith Packard <keithp@keithp.com> 57935Date: Wed Nov 6 23:05:46 2013 -0800 57936 57937 Make XTrans FD passing support optional. Define only on Linux 57938 57939 Until other operating systems have a libXtrans port for FD passing, 57940 disable this on non-Linux systems. 57941 57942 Note that this define affects how libXtrans gets built into the X 57943 server, which is why it need only define the symbol 57944 57945 Signed-off-by: Keith Packard <keithp@keithp.com> 57946 57947commit f70a8bf3714d89bccaad36841ef9149e91ad3bba 57948Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57949Date: Sun Nov 10 11:40:27 2013 -0800 57950 57951 XQuartz: Don't sleep forever on exit if usleep() returns an error 57952 57953 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57954 57955commit ebcc1c214c466582d7b92826b4860256fd9c582a 57956Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57957Date: Sun Nov 10 11:26:49 2013 -0800 57958 57959 XQuartz: Update Info.plist copyright years 57960 57961 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57962 57963commit 81c123ea2dd833864f7ba217791e59acca0f7c97 57964Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57965Date: Mon Nov 4 12:13:45 2013 -0800 57966 57967 configure.ac: Move GLX section after DRI 57968 57969 The GLX section of configura.ac checks the state of DRI2, so it 57970 needs to be after DRI2=auto is resolved. 57971 57972 Also reset libgl requirement to 7.1.0 in non-dri2 case. 57973 57974 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 57975 Reviewed-by: Adam Jackson <ajax@redhat.com> 57976 57977commit ab4b1fb38a61feb73d8336cc7a3399eb9d3d25be 57978Author: Andreas Schwab <schwab@suse.de> 57979Date: Thu Aug 15 12:03:42 2013 +0200 57980 57981 ARM64: Add support for aarch64 57982 57983 Signed-off-by: Andreas Schwab <schwab@suse.de> 57984 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 57985 Reviewed-by: Adam Jackson <ajax@redhat.com> 57986 Signed-off-by: Keith Packard <keithp@keithp.com> 57987 57988commit 8fcf82cbf015609e00b3feb9bc567133c7aee5cf 57989Author: Keith Packard <keithp@keithp.com> 57990Date: Tue Nov 5 18:39:11 2013 -0800 57991 57992 Use $GL_LIBS instead of -lGL for linking 57993 57994 -lGL presumes that the GL library is in the system path, while 57995 $GL_LIBS is auto-detected. 57996 57997 Signed-off-by: Keith Packard <keithp@keithp.com> 57998 57999commit f348935e7d2c84a438aba08eee2f2c4e0dc77d20 58000Author: Keith Packard <keithp@keithp.com> 58001Date: Tue Nov 5 18:37:12 2013 -0800 58002 58003 Link with xshmfence, reference miSyncShmScreenInit in sdksyms 58004 58005 This gets the server to link with xshmfence again, and also ensures 58006 that the miSyncShm code is linked into the server with the reference 58007 from sdksyms. 58008 58009 Signed-off-by: Keith Packard <keithp@keithp.com> 58010 58011commit 5f1e832694e57986c0185048a941b3af51b2f85f 58012Author: Julien Cristau <jcristau@debian.org> 58013Date: Tue Nov 5 07:08:21 2013 -0800 58014 58015 os: Actually use the computed clockid in GetTimeInMicros 58016 58017 The selection of which clock to use for this function was not actually 58018 getting used when fetching the final clock value. 58019 58020 Reported-by: Julien Cristau <jcristau@debian.org> 58021 Signed-off-by: Keith Packard <keithp@keithp.com> 58022 58023commit 903a058370645ea075ea98d380fd565efb6160c9 58024Author: Keith Packard <keithp@keithp.com> 58025Date: Mon Nov 4 19:01:26 2013 -0800 58026 58027 hw/xfree86: Link libdri3 only when DRI3 is defined 58028 58029 Don't attempt to link non-existant libraries... 58030 58031 Signed-off-by: Keith Packard <keithp@keithp.com> 58032 58033commit aaf0e29619196a283fee7ead2020a91032d84f48 58034Author: Keith Packard <keithp@keithp.com> 58035Date: Sun Nov 3 09:56:02 2013 -0800 58036 58037 Disable DRI3 and sync fence FD functions if xshmfence isn't available 58038 58039 Make sure the server can build when the xshmfence library isn't present 58040 58041 Signed-off-by: Keith Packard <keithp@keithp.com> 58042 58043commit 33c85beed521c9db140cadd8c5aa9992398ee1fe 58044Author: Keith Packard <keithp@keithp.com> 58045Date: Fri Nov 1 01:01:58 2013 -0700 58046 58047 Set version to 1.14.99.901 (1.15 RC1) 58048 58049 Signed-off-by: Keith Packard <keithp@keithp.com> 58050 58051commit 9db1d66f6d2c6857de220d8f1c08965962fe1416 58052Author: Keith Packard <keithp@keithp.com> 58053Date: Fri Nov 1 01:43:59 2013 -0700 58054 58055 Add dri3 and present directorys to distribution 58056 58057 Signed-off-by: Keith Packard <keithp@keithp.com> 58058 58059commit 977e2644b1d9e13b44debcb9a372bbc832ee32a3 58060Merge: 66c5ee0ff f36f5a65f 58061Author: Keith Packard <keithp@keithp.com> 58062Date: Fri Nov 1 00:34:18 2013 -0700 58063 58064 Merge remote-tracking branch 'whot/for-keith' 58065 58066commit 66c5ee0ff4690e630a69656331fd9d545c99c26c 58067Author: Egbert Eich <eich@freedesktop.org> 58068Date: Wed Aug 14 18:18:56 2013 +0200 58069 58070 config/APM: Add option to disable building of APM support on Linux 58071 58072 APM support in the Xserver was used to restore the console mode 58073 prior to a power management event. This was to ensure the mode 58074 upon suspend/resume was one that the system firmware or kernel 58075 could deal with. 58076 APM support is now largely obsolete, KMS drivers don't require a 58077 mode restoration anyhow. Therefore it should be possible to disable 58078 this feature. 58079 58080 (small modification by keithp - move test for XF86PM flag after check 58081 for APM, then move XF86PM flag to xorg-config.h.in) 58082 58083 Signed-off-by: Egbert Eich <eich@freedesktop.org> 58084 Tested-by: Keith Packard <keithp@keithp.com> 58085 Signed-off-by: Keith Packard <keithp@keithp.com> 58086 58087commit f36f5a65f639b6524191d888d5bf89e73027156c 58088Author: Peter Hutterer <peter.hutterer@who-t.net> 58089Date: Wed Oct 30 15:40:58 2013 +1000 58090 58091 sync: fix corner-case in triggering idle alarms 58092 58093 ProcessInputEvent() resets the device idle times. If idle time was higher than 58094 the lower bracket, this should trigger an event in the idle time wakeup 58095 handler. 58096 58097 If processing is slow, the idle time may advance past the lower bracket 58098 between the reset and the time the BlockHandler is called. In that case, we'd 58099 never schedule a wakeup to handle the event, causing us to randomly miss 58100 events. 58101 58102 Ran tests with a neg transition trigger on 5ms with 200 repeats of the test 58103 and it succeeded. Anything below that gets a bit tricky to make sure the 58104 server sees the same idle time as the client usleeps for. 58105 58106 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58107 Reviewed-by: Keith Packard <keithp@keithp.com> 58108 58109commit f994d99539192a10a8c83047e11991f52e0ca677 58110Author: Egbert Eich <eich@freedesktop.org> 58111Date: Wed Aug 14 18:18:39 2013 +0200 58112 58113 config/ACPI: Add option to disable building of ACPI support on Linux 58114 58115 ACPI support in the Xserver was used to restore the console mode 58116 prior to a power management event. This was to ensure the mode 58117 upon suspend/resume was one that the system firmware or kernel 58118 could deal with. 58119 The feature depended on acpid to be running. Most of this functionality 58120 is now take over by systemd, KMS drivers don't require a mode restoration 58121 anyhow. Therefore it should be possible to disable this feature under 58122 some circumstances. 58123 58124 Tested-by: Keith Packard <keithp@keithp.com> 58125 Signed-off-by: Keith Packard <keithp@keithp.com> 58126 58127commit 2aa5092b88174cb9988076ae96298217c97ad75f 58128Author: Egbert Eich <eich@freedesktop.org> 58129Date: Mon Aug 12 18:21:33 2013 +0200 58130 58131 DDX/Randr: Avoid server crash when xrandr SetConfig is called while switched away 58132 58133 A call to Xrandr SetScreenConfig (for randr 1.1) causes the Xserver to 58134 crash when xf86SetViewport() which does not check if the hardware is 58135 accessible. 58136 Wrap accesses to xf86SetViewport() with if (vtSema) { ... } to avoid that. 58137 58138 Signed-off-by: Egbert Eich <eich@freedesktop.org> 58139 Reviewed-by: Keith Packard <keithp@keithp.com> 58140 Signed-off-by: Keith Packard <keithp@keithp.com> 58141 58142commit 508e05777aba823c93a2417cc26d18c9f061ca26 58143Author: Egbert Eich <eich@freedesktop.org> 58144Date: Mon Aug 12 18:20:36 2013 +0200 58145 58146 DDX/Events: Distinguish between Input- and GeneralHandlers in xf86VTSwitch() 58147 58148 When enabling/disabling input handlers in xf86VTSwitch() we treat Input- 58149 and GeneralHandlers equally. The result is that after a VT switch the 58150 masks for EnabledDevices and AllSockets are equal and the distiction 58151 between both types is lost. 58152 58153 Signed-off-by: Egbert Eich <eich@freedesktop.org> 58154 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 58155 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 58156 Signed-off-by: Keith Packard <keithp@keithp.com> 58157 58158commit 0cb33ce34081d7d123ec565582d7d19bdb3964ad 58159Author: Egbert Eich <eich@freedesktop.org> 58160Date: Sat Oct 5 08:02:28 2013 +0200 58161 58162 DDX/modes: Add a sanity check when using screen sizes from EDID 58163 58164 EDID sometimes lies about screen sizes. Since the screen size is used 58165 by clients to determine the DPI a wrong ration will lead to terrible 58166 looking fonts. 58167 Add a sanity check for the h/v ratio cutting off at 2.4. This would 58168 still accept the cinemascope aspect ratio as valid. 58169 Also add message suggesting to add a quirk table entry. 58170 58171 Signed-off-by: Egbert Eich <eich@freedesktop.org> 58172 Reviewed-by: Keith Packard <keithp@keithp.com> 58173 Signed-off-by: Keith Packard <keithp@keithp.com> 58174 58175commit 41d4beb2616ceb3f1a1b8694733e85bae70de59a 58176Author: Egbert Eich <eich@freedesktop.org> 58177Date: Wed Aug 14 11:50:26 2013 +0200 58178 58179 DDX/DPMS: Call dixSaveScreens() also when screen is turned on 58180 58181 DMPS calls dixSaveScreens() when turned off but not when turned 58182 on. In most cases this is irrelevant as DPMS is done when a 58183 key is hit in which case dixSaveScreens() will be called to 58184 unblank anyhow. This isn't the case if we use xset (or the 58185 DPMS extension directly) to unblank. 58186 Check screenIsSaved to make sure the state needs to be changed 58187 before calling dixSaveScreens(). 58188 58189 Signed-off-by: Egbert Eich <eich@freedesktop.org> 58190 Reviewed-by: Keith Packard <keithp@keithp.com> 58191 Signed-off-by: Keith Packard <keithp@keithp.com> 58192 58193commit 623c4147650d0404cfbea0f9b7df66dc7d928e00 58194Author: Søren Sandmann Pedersen <ssp@redhat.com> 58195Date: Mon Oct 21 17:11:56 2013 -0400 58196 58197 ephyr: Ensure stride of private framebuffer is multiple of 4 58198 58199 The fb layer of X can't deal with strides that are not a multiple of 58200 4, so when Xephyr allocates its own framebuffer it should make sure to 58201 align it. 58202 58203 This fixes crashes and rendering corruption when Xephyr runs in a 58204 depth that is different from the host X server and its screen size is 58205 not a multiple of 4 / depth. (This is particularly easy to trigger if 58206 you use the -resizeable option). 58207 58208 Reviewed-by: Eric Anholt <eric@anholt.net> 58209 Signed-off-by: Soren Sandmann <ssp@redhat.com> 58210 Reviewed-by: Adam Jackson <ajax@redhat.com> 58211 58212commit 97cf53cc2ad7ecfdd495133bad31d0ec7d939326 58213Author: Søren Sandmann Pedersen <ssp@redhat.com> 58214Date: Mon Oct 21 16:58:54 2013 -0400 58215 58216 ephyr: hostx_screen_init(): Fix bits_per_pixel and bytes_per_line 58217 58218 When the depth of the Xephyr server matches that of the host X server, 58219 Xephyr simply uses the buffer associated with the XImage as its 58220 framebuffer. In this case, it is correct to get the bits_per_pixel and 58221 bytes_per_line values returned from hostx_screen_init() from the XImage. 58222 58223 However, when the depth doesn't match the host, Xephyr uses a private 58224 framebuffer that is periodically copied to the XImage. In this case, 58225 the returned values of bits_per_pixel and bytes_per_line should be 58226 those of the private framebuffer, not those of the XImage. 58227 58228 Reviewed-by: Eric Anholt <eric@anholt.net> 58229 Signed-off-by: Soren Sandmann <ssp@redhat.com> 58230 Reviewed-by: Adam Jackson <ajax@redhat.com> 58231 58232commit 55246b67b755d4c1039d54971fe3f77ea60d604e 58233Author: Søren Sandmann Pedersen <ssp@redhat.com> 58234Date: Tue Oct 8 17:45:40 2013 -0400 58235 58236 xf86AddBusDeviceToConfigure(): Store device in DevToConfig[i].pVideo 58237 58238 After fc3ab84d the pVideo field in DevToConfig[i] is no longer 58239 initialized, so it's always NULL. This causes the duplicate finding 58240 algorithm in the beginning of the function to not work anymore as it 58241 is based on this field. 58242 58243 The symptom of this bug is that X -configure reports 58244 58245 Number of created screens does not match number of detected devices. 58246 Configuration failed. 58247 Server terminated with error (2). Closing log file. 58248 58249 rather than producing a working config file. 58250 58251 This patch fixes that bug by initializing the field before calling 58252 xf86PciConfigureNewDev(). 58253 58254 Cc: tvignatti@gmail.com 58255 Signed-off-by: Soren Sandmann <ssp@redhat.com> 58256 Reviewed-by: Adam Jackson <ajax@redhat.com> 58257 58258commit 95bf478b78e466002c382bcde7d4d62591e9215d 58259Merge: abf5d5ac1 bb745f2b4 58260Author: Keith Packard <keithp@keithp.com> 58261Date: Thu Oct 31 18:18:19 2013 -0700 58262 58263 Merge remote-tracking branch 'jeremyhu/master' 58264 58265commit abf5d5ac12437ebe156b4dd500c2acd69eea3654 58266Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 58267Date: Thu Oct 31 08:57:56 2013 -0700 58268 58269 miext/damage: Partial revert of "Only wrap into the GC ops chain if there's a listener (v3)" 58270 58271 Fixes regression from: 4dc2a76740d921c824a4d8193f39dd373475f02a 58272 58273 http://lists.x.org/archives/xorg-devel/2013-May/036241.html 58274 https://trac.macports.org/ticket/38993 58275 58276 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 58277 Reviewed-by: Adam Jackson <ajax@redhat.com> 58278 Signed-off-by: Keith Packard <keithp@keithp.com> 58279 58280commit e4636072949a76477fe7c9d54662a0b3536b1372 58281Merge: 17ed7ac1f 5c5c1b779 58282Author: Keith Packard <keithp@keithp.com> 58283Date: Thu Oct 31 17:05:48 2013 -0700 58284 58285 Merge remote-tracking branch 'keithp/dri3' 58286 58287commit 17ed7ac1fe7426d66f688087c09da5ff1dbb41d3 58288Author: Adam Jackson <ajax@redhat.com> 58289Date: Thu Oct 31 18:39:22 2013 -0400 58290 58291 glx: Lie about GLX_Y_INVERTED_EXT 58292 58293 Well, that was lame. The problem with reporting y inversion honestly is 58294 that libGL asks the driver _its_ opinion of Y inversion, which it just 58295 fabricates from whole cloth. So then when libGL goes to compare the 58296 driver's idea of fbconfigs with that of the server - a fairly dumb idea 58297 to begin with - nothing matches, and direct rendering fails, and 58298 sadness. 58299 58300 So until the DRI drivers are fixed we should just continue to lie about 58301 Y inversion. GLX_DONT_CARE is what libGL would make up for that 58302 attribute if we hadn't sent it, so just send that instead. 58303 58304 Signed-off-by: Adam Jackson <ajax@redhat.com> 58305 Tested-by: Keith Packard <keithp@keithp.com> 58306 Signed-off-by: Keith Packard <keithp@keithp.com> 58307 58308commit 5c5c1b77982a9af7279a90bc3c2be48adaa9c778 58309Author: Keith Packard <keithp@keithp.com> 58310Date: Thu Jul 11 16:11:57 2013 -0700 58311 58312 present: Add Present extension 58313 58314 Provides both a software implementation using timers and driver hooks 58315 to base everything on vblank intervals. 58316 58317 Signed-off-by: Keith Packard <keithp@keithp.com> 58318 Reviewed-by: Adam Jackson <ajax@redhat.com> 58319 58320commit 563138298868f62501875d3016f03469dcffaad0 58321Author: Keith Packard <keithp@keithp.com> 58322Date: Tue Apr 9 19:59:39 2013 -0700 58323 58324 dri3: Add DRI3 extension 58325 58326 Adds DRM compatible fences using futexes. 58327 Uses FD passing to get pixmaps from DRM applications. 58328 58329 Signed-off-by: Keith Packard <keithp@keithp.com> 58330 Reviewed-by: Adam Jackson <ajax@redhat.com> 58331 58332commit fdec793cdc2ef9a6ea66b311cb1068a7bd4a3be3 58333Author: Keith Packard <keithp@keithp.com> 58334Date: Thu Jan 17 13:46:55 2013 -0800 58335 58336 Add support for MIT-SHM AttachFd request 58337 58338 This passes a file descriptor from the client to the server, which is 58339 then mmap'd 58340 58341 Signed-off-by: Keith Packard <keithp@keithp.com> 58342 Reviewed-by: Adam Jackson <ajax@redhat.com> 58343 58344commit 9fd35daa3160fd36f00ed354bfcbefefa1353cce 58345Author: Keith Packard <keithp@keithp.com> 58346Date: Thu Jan 17 13:43:02 2013 -0800 58347 58348 Add interfaces to get FDs from clients over the socket 58349 58350 This adds two interfaces: 58351 58352 void SetReqFds(ClientPtr client, int req_fds) 58353 58354 Marks the number of file descriptors expected for this 58355 request. Call this before any request processing so that 58356 any un-retrieved file descriptors will be closed 58357 automatically. 58358 58359 int ReadFdFromClient(ClientPtr client) 58360 58361 Reads the next queued file descriptor from the connection. If 58362 this request is not expecting any more file descriptors, or 58363 if there are no more file descriptors available from the 58364 connection, then this will return -1. 58365 58366 Signed-off-by: Keith Packard <keithp@keithp.com> 58367 Reviewed-by: Adam Jackson <ajax@redhat.com> 58368 58369commit 264fc3abe5f18341d0cf9ddb6766e10e4154e447 58370Author: Keith Packard <keithp@keithp.com> 58371Date: Thu Oct 31 13:01:46 2013 -0700 58372 58373 misync: Don't smash custom screen sync functions 58374 58375 There was a check to avoid smashing custom functions, but the sense 58376 was backwards causing it to always smash them, and also not set them otherwise. 58377 58378 Signed-off-by: Keith Packard <keithp@keithp.com> 58379 Reviewed-by: Adam Jackson <ajax@redhat.com> 58380 58381commit 2d96948ab5c952b68875ac63844cf7d778d4bf63 58382Author: Keith Packard <keithp@keithp.com> 58383Date: Thu Jul 11 16:10:34 2013 -0700 58384 58385 os: Add GetTimeInMicros 58386 58387 64-bit higher resolution current time value. 58388 58389 Signed-off-by: Keith Packard <keithp@keithp.com> 58390 Reviewed-by: Adam Jackson <ajax@redhat.com> 58391 58392commit 0c33f47281c36726848daf513fb0483cdea57bff 58393Author: Keith Packard <keithp@keithp.com> 58394Date: Thu Jul 11 16:09:34 2013 -0700 58395 58396 Add swapll to byte swap 64-bit datatypes 58397 58398 Signed-off-by: Keith Packard <keithp@keithp.com> 58399 Reviewed-by: Adam Jackson <ajax@redhat.com> 58400 58401commit 26f013ba45b08a02bb028a461af68288a86fadb1 58402Author: Keith Packard <keithp@keithp.com> 58403Date: Thu Jul 11 16:08:41 2013 -0700 58404 58405 Add a RegionDuplicate function 58406 58407 This allocates a new region structure and copies a source region into 58408 it in a single API rather than forcing the caller to do both steps themselves. 58409 58410 Signed-off-by: Keith Packard <keithp@keithp.com> 58411 Reviewed-by: Adam Jackson <ajax@redhat.com> 58412 58413commit d25c217964eb1fe54c3a54bca4cac7f47b4b9fdf 58414Author: Keith Packard <keithp@keithp.com> 58415Date: Mon Jan 14 14:24:36 2013 -0800 58416 58417 Clean up a couple of warnings in os/ 58418 58419 Signed-off-by: Keith Packard <keithp@keithp.com> 58420 Reviewed-by: Adam Jackson <ajax@redhat.com> 58421 58422commit 7710f2b927ec8e3c631f72d66d494b523377b48e 58423Author: Keith Packard <keithp@keithp.com> 58424Date: Fri Jan 18 21:43:40 2013 -0800 58425 58426 Xext: Move MIT-SHM 'ShmDesc' to shmint.h 58427 58428 This data structure is required to use shared memory objects in any 58429 extension. That includes the Xv extension, which (before this patch) 58430 duplicated the definition of this structure in its own code. 58431 58432 Signed-off-by: Keith Packard <keithp@keithp.com> 58433 Reviewed-by: Adam Jackson <ajax@redhat.com> 58434 58435commit e8961b718d82f1c081ec110d8d962f64e8406b82 58436Author: Peter Hutterer <peter.hutterer@who-t.net> 58437Date: Tue Oct 22 14:24:52 2013 +1000 58438 58439 os: use a constant for backtrace array size 58440 58441 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58442 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 58443 58444commit f12a9ed870017f35cf6d2a82b1405e843aae42ac 58445Author: Peter Hutterer <peter.hutterer@who-t.net> 58446Date: Tue Oct 29 15:24:11 2013 +1000 58447 58448 configure: remove a comment 58449 58450 94ed0ba1b5043ad9fc33b42756af447d5ab15bbd moved backtracing into the DIX, so 58451 this comment is outdated. since no-one noticed and it's easier to just grep 58452 than update file references, remove the comment. 58453 58454 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58455 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 58456 58457commit bb745f2b45768d0936f3960951e1a5cee8b03194 58458Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 58459Date: Thu Oct 31 08:50:57 2013 -0700 58460 58461 XQuartz: Bump bundle version to 2.7.6 58462 58463 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 58464 58465commit 2eedf42c229e2251035d716a96b654222155cd51 58466Author: Adam Jackson <ajax@redhat.com> 58467Date: Thu Oct 31 09:23:52 2013 -0400 58468 58469 dri: Fix GLX_Y_INVERTED_EXT fbconfig attribute 58470 58471 We're Y-inverted from GL's coordinates, so this is correct. gnome-shell 58472 doesn't seem to check this - somewhat reasonable since the only server 58473 that answered the other way around was Xglx - but kwin does, and 58474 upside-down hilarity ensues. 58475 58476 Tested-by: maelcum on #xorg-devel 58477 Signed-off-by: Adam Jackson <ajax@redhat.com> 58478 Signed-off-by: Keith Packard <keithp@keithp.com> 58479 58480commit 902ff0b3497d202b86bf9a411e17db7b694d6eaa 58481Author: Adam Jackson <ajax@redhat.com> 58482Date: Tue Oct 29 12:58:14 2013 -0400 58483 58484 xfree86: Bump video/input/extension ABIs for 1.15 58485 58486 Signed-off-by: Adam Jackson <ajax@redhat.com> 58487 Reviewed-by: Julien Cristau <jcristau@debian.org> 58488 Signed-off-by: Keith Packard <keithp@keithp.com> 58489 58490commit 899451ae595f5f51a500295c624bf7ad14aca602 58491Author: Keith Packard <keithp@keithp.com> 58492Date: Wed Oct 30 08:54:24 2013 -0700 58493 58494 kdrive/ephyr: Don't discard one-time driver structure at server reset 58495 58496 KdScreenInfo is constructed at server startup time, and not 58497 re-generated at server reset time. Freeing the 'driver' element at 58498 reset time means this information is lost, and the server crashes 58499 pretty quickly afterwards. 58500 58501 Signed-off-by: Keith Packard <keithp@keithp.com> 58502 Reviewed-by: Adam Jackson <ajax@redhat.com> 58503 58504commit 85ae44f07fa21b4a48d95d736dc8135bd4f751d0 58505Author: Michal Srb <msrb@suse.com> 58506Date: Wed Oct 30 13:33:51 2013 +0200 58507 58508 randr: deliver Output and Crtc events of attached output providers. 58509 58510 Consider all attached output providers when looking for changed outputs and 58511 crtcs. 58512 58513 Reviewed-by: Dave Airlie <airlied@redhat.com> 58514 Signed-off-by: Michal Srb <msrb@suse.com> 58515 Signed-off-by: Keith Packard <keithp@keithp.com> 58516 58517commit a9ca93dcf9a711b3d60fbad31cbd44e0c82b6f42 58518Author: Michal Srb <msrb@suse.com> 58519Date: Wed Oct 30 13:33:50 2013 +0200 58520 58521 randr: send RRResourceChangeNotify event 58522 58523 Send RRResourceChangeNotify event when provider, output or crtc was created or 58524 destroyed. I.e. when the list of resources returned by RRGetScreenResources and 58525 RRGetProviders changes. 58526 58527 Reviewed-by: Dave Airlie <airlied@redhat.com> 58528 Signed-off-by: Michal Srb <msrb@suse.com> 58529 Signed-off-by: Keith Packard <keithp@keithp.com> 58530 58531commit 4b39ea8a918fe6117bf34dc1cba67a4e56d0d1fb 58532Author: Michal Srb <msrb@suse.com> 58533Date: Wed Oct 30 13:33:49 2013 +0200 58534 58535 randr: send RRProviderChangeNotify event 58536 58537 Send RRProviderChangeNotify event when a provider becomes output source or 58538 offload sink. 58539 58540 Reviewed-by: Dave Airlie <airlied@redhat.com> 58541 Signed-off-by: Michal Srb <msrb@suse.com> 58542 Signed-off-by: Keith Packard <keithp@keithp.com> 58543 58544commit 1470c0a132806fb2ce817515ad4ddfe66eb7c98f 58545Merge: 04ab07ca1 9f4afe7e8 58546Author: Keith Packard <keithp@keithp.com> 58547Date: Tue Oct 29 21:00:13 2013 -0700 58548 58549 Merge remote-tracking branch 'whot/for-keith' 58550 58551commit 9f4afe7e835cf9c513387ae01eaf8b7a6ba4aac9 58552Author: Gaetan Nadon <memsize@videotron.ca> 58553Date: Mon Oct 28 18:32:36 2013 -0400 58554 58555 config: replace deprecated use of AC_OUTPUT with AC_CONFIG_FILES 58556 58557 Fix Automake warning: AC_OUTPUT should be used without arguments. 58558 www.gnu.org/software/autoconf/manual/autoconf.html#Configuration-Files 58559 58560 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 58561 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 58562 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58563 58564commit e16dea72d6f45dd6d7788362bb7dbe76263719b1 58565Author: Peter Hutterer <peter.hutterer@who-t.net> 58566Date: Wed Oct 23 15:16:25 2013 +1000 58567 58568 dmx: queue button events with a flags of 0 58569 58570 Setting POINTER_SCREEN with a unset valuator mask causes a jump to 0/0. Set 58571 the flags to 0 so we don't generate any motion on a button event. 58572 58573 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58574 Reviewed-by: Keith Packard <keithp@keithp.com> 58575 58576commit 04ab07ca19236d6c9a947e065fb69b0dd0d16639 58577Author: Connor Behan <connor.behan@gmail.com> 58578Date: Thu Oct 17 18:26:28 2013 -0700 58579 58580 xfree86: Find primary entity when bus types are nominally different 58581 58582 As of server 1.13, systems with DRM and Udev will have BUS_PLATFORM as 58583 their primary bus type. However, drivers not implementing a 58584 platformProbe function will still create entities of type BUS_PCI. We 58585 need to account for this when checking for the primary entity. 58586 58587 Signed-off-by: Connor Behan <connor.behan@gmail.com> 58588 Acked-by: Tormod Volden <debian.tormod@gmail.com> 58589 Reviewed-by: Adam Jackson <ajax@redhat.com> 58590 Signed-off-by: Keith Packard <keithp@keithp.com> 58591 58592commit b32a4c91ccd479638c2bd2c0143b6ea170c717d1 58593Merge: be6680967 c671e935e 58594Author: Keith Packard <keithp@keithp.com> 58595Date: Tue Oct 29 09:37:30 2013 -0700 58596 58597 Merge remote-tracking branch 'idr/glx-float-fbconfig' 58598 58599commit be6680967a479eedbcab2fe1718c5f981e1029c7 58600Author: Adam Jackson <ajax@redhat.com> 58601Date: Wed Jul 10 10:00:46 2013 -0400 58602 58603 glx: convert to direct GL dispatch (v2) 58604 58605 We now expect to be linked against something that provides the GL API, 58606 instead of manually grubbing about in the DRI driver's dispatch table. 58607 Since the GLX we expose calls GL functions that are meant to be looked 58608 up dynamically, also add a way to thunk through to GetProcAddress. 58609 58610 This includes a refresh of the generated sources, which requires a 58611 correspondingly new Mesa. 58612 58613 The GetProcAddress stubs are at the moment merely enough to make this 58614 link against Mesa 9.2, but should really be provided for everything not 58615 in the OpenGL 1.2 ABI. 58616 58617 v2: Explicitly hide the GetProcAddress stubs so we can't conflict with 58618 libGL symbols; fix leading tab/space issues [anholt] 58619 58620 Reviewed-by: Keith Packard <keithp@keithp.com> 58621 Signed-off-by: Adam Jackson <ajax@redhat.com> 58622 58623commit 8aacf47e1778d8b72811b025a82452b933d3c1f2 58624Author: Adam Jackson <ajax@redhat.com> 58625Date: Fri Oct 4 12:58:19 2013 -0400 58626 58627 glx: Remove DRI1 AIGLX (v2) 58628 58629 Mesa doesn't ship DRI1 drivers as of 8.0, which is about 18 months and 58630 three releases ago. The main reason to have wanted DRI1 AIGLX was to 58631 get a GLX compositor working, but DRI1's (lack of) memory management API 58632 meant that the cost of a GLX compositor was breaking direct GLX apps, 58633 which isn't a great tradeoff. 58634 58635 Of the DRI1 drivers Mesa has dropped, I believe only mga stands to lose 58636 some functionality here, since it and only it has support for 58637 NV_texture_rectangle. Since that's required for every extant GLX 58638 compositor I know of, I conclude that anybody with a savage, say, would 58639 probably not notice AIGLX going away, since they wouldn't be running a 58640 GLX compositor in the first place. 58641 58642 In the future we'd like to use GL in the server in a more natural way, 58643 as just another EGL client, including in the GLX implementation itself. 58644 Since there's no EGL implemented for DRI1 drivers, this would already 58645 doom AIGLX on DRI1 (short of entirely forking the GLX implementation, 58646 which I'm not enthusiastic about). 58647 58648 v2: Remove DRI1 from AIGLX conditionals in configure.ac [anholt] 58649 58650 Reviewed-by: Eric Anholt <eric@anholt.net> 58651 Signed-off-by: Adam Jackson <ajax@redhat.com> 58652 58653commit 7f5adf73a0f9a951a6df201532b4031d38054369 58654Author: Adam Jackson <ajax@redhat.com> 58655Date: Wed Oct 2 17:38:05 2013 -0400 58656 58657 glx: Simplify glXDestroyContext 58658 58659 We can just free the resource unconditionally here. ContextGone (which 58660 FreeResourceByType will call) already does: 58661 58662 cx->idExists = GL_FALSE; 58663 if (!cx->currentClient) { 58664 __glXFreeContext(cx); 58665 } 58666 58667 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 58668 Signed-off-by: Adam Jackson <ajax@redhat.com> 58669 58670commit 53653f4889dcaea0cf67f8a4180af71eeb4b2f6b 58671Author: Adam Jackson <ajax@redhat.com> 58672Date: Mon Aug 5 15:46:11 2013 -0400 58673 58674 glx: Fix a corner case in DrawableGone 58675 58676 Losing the drawable does not change our notion of current client. Since 58677 the GL under us doesn't understand having a current context without 58678 current drawables (sigh), we do still need to loseCurrent so that we 58679 re-bind the context on the next request. 58680 58681 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=65030 58682 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 58683 Signed-off-by: Adam Jackson <ajax@redhat.com> 58684 58685commit 276d8057aa7f6e36c9c874257a6383f416c5cedd 58686Author: Adam Jackson <ajax@redhat.com> 58687Date: Sat Aug 3 09:47:55 2013 -0400 58688 58689 glx: Fix memory leak in context garbage collection (v2) 58690 58691 I broke this, back in: 58692 58693 commit a48dadc98a28c969741979b70b7a639f24f4cbbd 58694 Author: Adam Jackson <ajax@redhat.com> 58695 Date: Mon Mar 21 11:59:29 2011 -0400 58696 58697 glx: Reimplement context tags 58698 58699 In that, I changed the glx client state to not explicitly track the list 58700 of current contexts for the client (since that was what we were deriving 58701 tags from). The bug was that I removed the code for same from 58702 glxClientCallback without noticing that it had the side effect of 58703 effectively de-currenting those contexts, so that ContextGone could free 58704 them. So, if you had a client exit with a context still current, the 58705 context's memory would leak. Not a huge deal for direct clients, but 58706 viciously bad for indirect, since the swrast context state at the bottom 58707 of Mesa is like 15M. 58708 58709 Fix this by promoting Bool isCurrent to ClientPtr currentClient, so that 58710 we have a back-pointer to chase when walking the list of contexts when 58711 ClientStateGone happens. 58712 58713 v2: Explicitly call __glXFreeContext on the ClientStateGone path. Our 58714 current context might be one we got from EXT_import_context and whose 58715 creating client has since died. Without the explicit call, the creating 58716 client's FreeClientResources would not free the context because it's 58717 still current, and the using client's FreeClientResources would not free 58718 the context because it's not an XID it created. This matches the logic 58719 from a48dadc. 58720 58721 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 58722 Signed-off-by: Adam Jackson <ajax@redhat.com> 58723 58724commit 75b362763cca8b6139582ad85e3b58f0b2e30a1c 58725Author: Adam Jackson <ajax@redhat.com> 58726Date: Tue Oct 8 15:02:44 2013 -0400 58727 58728 glx: Only take a Pixmap reference if creating the GLXPixmap succeeded 58729 58730 Reviewed-by: Julien Cristau <jcristau@debian.org> 58731 Signed-off-by: Adam Jackson <ajax@redhat.com> 58732 58733commit 4292a39c8014f9f4920bed3001802e4a3e523c5e 58734Author: Peter Hutterer <peter.hutterer@who-t.net> 58735Date: Mon Oct 28 12:10:10 2013 +1000 58736 58737 Xi: fix logic error when calculating emulated motion events 58738 58739 gcc -Wlogical-op 58740 exevents.c: In function 'DeliverEmulatedMotionEvent': 58741 exevents.c:1480:13: warning: logical 'or' of collectively exhaustive 58742 tests is always true [-Wlogical-op] 58743 58744 The relevant snippet of exevents.c: 58745 58746 1479 if (ti->listeners[0].type != LISTENER_POINTER_REGULAR || 58747 1480 ti->listeners[0].type != LISTENER_POINTER_GRAB) 58748 1481 return; 58749 58750 This condition was always true, causing dropped motion events. 58751 58752 Reported-by: Alan Coopersmith <alan.coopersmith@oracle.com> 58753 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58754 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 58755 58756commit c671e935e7c44d4711ba9588ee96cb9796b879e6 58757Author: Daniel Czarnowski <daniel.czarnowski@intel.com> 58758Date: Mon Oct 21 19:48:38 2013 +0200 58759 58760 glx: Enable GLX_ARB_fbconfig_float for DRI2 drivers 58761 58762 v2 (idr): Maintain sort order of extension list. Fixed rebase error 58763 (comment header for initializeExtensions was in front of 58764 __glXDRIinvalidateBuffers). Re-write commit message. 58765 58766 Signed-off-by: Daniel Czarnowski <daniel.czarnowski@intel.com> 58767 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 58768 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 58769 58770commit ccc8bb1153ec669f7540a174a81a8ec0e0d6dd67 58771Author: Daniel Czarnowski <daniel.czarnowski@intel.com> 58772Date: Wed Oct 23 12:53:59 2013 +0200 58773 58774 glx: Handle float config types in glxConvertConfigs 58775 58776 Replaces old use of floatMode attribute with new, extended range of 58777 values in __DRI_ATTRIB_RENDER_TYPE. Also adds new conditions, where the 58778 float modes support requires it. Enables support for not only float 58779 configs, but packed float configs as well. 58780 58781 v2 (idr): Whitespace and formatting fixes. Refactor render type 58782 vs. pbuffer checking to a separate function that includes a quote from 58783 the spec. Re-write commit message. Fix compiler warnings: 58784 58785 glxdricommon.c: In function 'glxConvertConfigs': 58786 glxdricommon.c:212:35: warning: pointer targets in passing argument 3 of 'core->getConfigAttrib' differ in signedness [-Wpointer-sign] 58787 glxdricommon.c:212:35: note: expected 'unsigned int *' but argument is of type 'int *' 58788 glxdricommon.c:230:35: warning: pointer targets in passing argument 3 of 'core->getConfigAttrib' differ in signedness [-Wpointer-sign] 58789 glxdricommon.c:230:35: note: expected 'unsigned int *' but argument is of type 'int *' 58790 58791 Signed-off-by: Daniel Czarnowski <daniel.czarnowski@intel.com> 58792 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 58793 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 58794 58795commit 4e5eb15b4cc170499cf87e6633c1737f0b1c514b 58796Author: Daniel Czarnowski <daniel.czarnowski@intel.com> 58797Date: Mon Oct 21 19:57:53 2013 +0200 58798 58799 glx: Correctly set render type enum 58800 58801 The type specified is bits, but the context wants an enum. 58802 58803 v2 (idr): Fix various whitespace and formatting problems. Rename 58804 convFBconfRenderTypeBits2CtxRenderType to 58805 renderTypeBitsToRenderTypeEnum. Re-write commit message. 58806 58807 Signed-off-by: Daniel Czarnowski <daniel.czarnowski@intel.com> 58808 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 58809 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 58810 58811commit eabb523aa45f3d8940cadac197ef1225fbfd5c56 58812Author: Daniel Czarnowski <daniel.czarnowski@intel.com> 58813Date: Mon Oct 21 19:45:35 2013 +0200 58814 58815 glx: Allow float renderType in glXCreateContextAttribsARB 58816 58817 This enables GLX_RGBA_FLOAT_TYPE_ARB and 58818 GLX_RGBA_UNSIGNED_FLOAT_TYPE_EXT as valid renderType parameters to 58819 glXCreateContextAttribsARB. 58820 58821 v2 (idr) : Use C-style comments and fix word wrapping. Re-write commit 58822 message. 58823 58824 Signed-off-by: Daniel Czarnowski <daniel.czarnowski@intel.com> 58825 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 58826 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 58827 58828commit 7ecfab47eb221dbb996ea6c033348b8eceaeb893 58829Author: Eric Anholt <eric@anholt.net> 58830Date: Tue Oct 22 14:22:04 2013 -0700 58831 58832 glx: Add support for the new DRI loader entrypoint. 58833 58834 This is going to be exposed (and not the old entrypoint) for some DRI 58835 drivers once the megadrivers series lands, and the plan is to 58836 eventually transition all drivers to that. Hopefully this is 58837 unobtrusive enough to merge to stable X servers so that they can be 58838 compatible with new Mesa versions. 58839 58840 v2: typo fix in the comment 58841 58842 Signed-off-by: Eric Anholt <eric@anholt.net> 58843 Reviewed-by: Adam Jackson <ajax@redhat.com> 58844 Signed-off-by: Keith Packard <keithp@keithp.com> 58845 58846commit 83bb366e71ef41165998eed2934208f84714a37e 58847Author: Strake <strake888@gmail.com> 58848Date: Thu Oct 17 23:20:20 2013 -0500 58849 58850 unbreak KdEnableInput 58851 58852 pi = ki = 0 here, so not follow them. 58853 58854 Signed-off-by: Strake <strake888@gmail.com> 58855 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 58856 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58857 58858commit 8db4121a3c7104548fd4a138ac1d1bdea5fd22d9 58859Author: Keith Packard <keithp@keithp.com> 58860Date: Fri Oct 18 17:31:01 2013 -0700 58861 58862 Update to version 1.14.99.3 58863 58864 Signed-off-by: Keith Packard <keithp@keithp.com> 58865 58866commit e3220d7bae4548bf64e5879c981ad070ac1f78f7 58867Merge: 73b2660d7 2523a445a 58868Author: Keith Packard <keithp@keithp.com> 58869Date: Fri Oct 18 17:22:52 2013 -0700 58870 58871 Merge commit '2523a445a09a75a8baf642608d099a5e12d5437f' 58872 58873 This is merging the commits along Peter's for-keith branch, without 58874 also merging in the spurious merge at the top of that branch. 58875 58876commit f8662a133690773692dfa392ecfa3b8b56af8fa9 58877Merge: 7cf1b595c 2523a445a 58878Author: Peter Hutterer <peter.hutterer@who-t.net> 58879Date: Fri Oct 18 16:50:12 2013 +1000 58880 58881 Merge branch 'xsync-transition-bug-59644' into for-keith 58882 58883commit 2523a445a09a75a8baf642608d099a5e12d5437f 58884Author: Peter Hutterer <peter.hutterer@who-t.net> 58885Date: Thu Oct 17 12:02:27 2013 +1000 58886 58887 sync: split updating and triggering a counter up 58888 58889 Now that the brackets are always the nearest change points (regardless of 58890 transition) we need to update the counters whenever we check for any updates. 58891 58892 Otherwise we end up with a situation where counter->value is out of date and 58893 an alarm doesn't trigger because we're still using the value from last time 58894 something actually triggered. 58895 58896 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58897 Reviewed-by: Adam Jackson <ajax@redhat.com> 58898 Reviewed-by: Keith Packard <keithp@keithp.com> 58899 58900commit e57ec99b03b2ad840c384a97ab2766ce9da0f5cc 58901Author: Peter Hutterer <peter.hutterer@who-t.net> 58902Date: Wed Oct 16 16:31:15 2013 +1000 58903 58904 sync: always set the brackets (#59644) 58905 58906 The current code sets bracket_greater to the first trigger after the current 58907 value, and bracket_less to the last trigger before the current value. 58908 58909 For example, the idle timer with three negative and three positive transitions 58910 would set this: 58911 58912 nt1 nt2 nt3 58913 |--------|------|--|------- idle --|---|--|-----> t 58914 pt1 pt2 pt3 58915 bracket_less == nt2 58916 bracket_greater == pt2 58917 58918 This is an optimization so we can skip code paths in the block/wakeup handlers 58919 if the current value doesn't meet any of the trigger requirements. Those 58920 handlers largely do a 58921 if (bracket_less is less than current value && 58922 bracket_greater is greater than current value) 58923 return, nothing to do 58924 58925 However, unless the bracket values are updated at the correct time, the 58926 following may happen: 58927 58928 nt 58929 |--------------|---------- idle ------|--------> t 58930 pt 58931 58932 In this case, neither bracket is set, we're past the pos transition and not 58933 yet at the neg transition. idle may now go past nt, but the brackets are not 58934 updated. If idle is then reset to 0, no alarm is triggered for nt. Likewise, 58935 idle may now go past pt and no alarm is triggered. 58936 58937 Changing an alarm or triggering an alarm will re-calculate the brackets, so 58938 this bug is somewhat random. If any other client triggers an alarm when the 58939 brackets are wrongly NULL, the recalculation will set them this bug may not 58940 appear. 58941 58942 This patch changes the behavior, so that the brackets are always the nearest 58943 positive or negative transitions to the current counter value. In the example 58944 above, nt will trigger a wakeup and a re-calculation of the brackets, so that 58945 going past it in the negative direction will then cause the proper alarm 58946 triggers. 58947 58948 Or, in Keith's words: 58949 58950 Timer currently past a positive trigger 58951 No bracket values, because no trigger in range 58952 58953 Timer moves backwards before the positive trigger 58954 Brackets not reset, even though there is now a trigger in range 58955 58956 Timer moves forward past the positive trigger 58957 Trigger doesn't fire because brackets not set 58958 58959 Setting the LT bracket in this case will cause everything to get 58960 re-evaluated when the sync value moves backwards before the trigger 58961 value. 58962 58963 X.Org Bug 59644 <http://bugs.freedesktop.org/show_bug.cgi?id=59644> 58964 58965 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58966 Reviewed-by: Adam Jackson <ajax@redhat.com> 58967 Reviewed-by: Keith Packard <keithp@keithp.com> 58968 58969commit 06b87aa528d7a739ba20101a1f83b1a428691a01 58970Author: Peter Hutterer <peter.hutterer@who-t.net> 58971Date: Wed Oct 16 10:08:46 2013 +1000 58972 58973 sync: if the idle time was reset, force alarms to trigger (#70476) 58974 58975 The time between the idle reset and the IdleTimeWakeupHandler to be called is 58976 indeterminate. Clients with an PositiveTransition or NegativeTransition alarm 58977 on a low threshold may miss an alarm. 58978 58979 Work around this by keeping a reset flag for each device. When the 58980 WakeupHandler triggers and the reset flag is set, we force a re-calculation of 58981 everything and pretend the current idle time is zero. Immediately after is the 58982 next calculation with the real idle time. 58983 58984 Relatively reproducible test case: Set up a XSyncNegativeTransition alarm for 58985 a threshold of 1 ms. May trigger, may not. 58986 58987 X.Org Bug 70476 <http://bugs.freedesktop.org/show_bug.cgi?id=70476> 58988 58989 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 58990 Reviewed-by: Adam Jackson <ajax@redhat.com> 58991 Reviewed-by: Keith Packard <keithp@keithp.com> 58992 58993commit efc1035ca958f2c9d266338a308518a0834b1773 58994Author: Peter Hutterer <peter.hutterer@who-t.net> 58995Date: Wed Oct 16 09:36:01 2013 +1000 58996 58997 dix: provide accessor methods for the last device event time 58998 58999 And now that we have the accessors, localize it. No functional changes, just 59000 preparing for a future change. 59001 59002 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59003 Reviewed-by: Adam Jackson <ajax@redhat.com> 59004 Reviewed-by: Keith Packard <keithp@keithp.com> 59005 59006commit b7c9bd9cf276e92a73be57ff2ed32b47a80f13fb 59007Author: Peter Hutterer <peter.hutterer@who-t.net> 59008Date: Wed Oct 16 09:21:47 2013 +1000 59009 59010 sync: supply the counter from IdleTimeBlockHandler 59011 59012 The main idletime counter has an initialized deviceid, might as well be 59013 supplying it properly. Without this, we'd only ever check the XIAllDevices 59014 counter, so the wait would never be adjusted for the device-specific triggers. 59015 59016 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59017 Reviewed-by: Adam Jackson <ajax@redhat.com> 59018 Reviewed-by: Keith Packard <keithp@keithp.com> 59019 59020commit 2efe49c1029f959fe80879bcf50df42e8b80451d 59021Author: Peter Hutterer <peter.hutterer@who-t.net> 59022Date: Wed Oct 16 13:01:01 2013 +1000 59023 59024 sync: always call BracketValues when recalculating upper/lower brackets 59025 59026 Both ServertimeBracketValues and IdleTimeBracketValues copy the value into 59027 there SysCounter privates. Call it for a NULL set as well, so we don't end up 59028 with stale pointers and we can remove the block/wakeup handlers. 59029 59030 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59031 Reviewed-by: Adam Jackson <ajax@redhat.com> 59032 Reviewed-by: Keith Packard <keithp@keithp.com> 59033 59034commit 5c7cfd4c6978834551848e1be62af240102e39b5 59035Author: Peter Hutterer <peter.hutterer@who-t.net> 59036Date: Tue Oct 15 10:11:20 2013 +1000 59037 59038 sync: compress two if statements 59039 59040 No functional changes, just merges a > and == condition into a >= condition. 59041 59042 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59043 Reviewed-by: Adam Jackson <ajax@redhat.com> 59044 Reviewed-by: Keith Packard <keithp@keithp.com> 59045 59046commit 73b2660d7273d175d279d22f8ca0c3932a14ff1c 59047Author: Alan Coopersmith <alan.coopersmith@oracle.com> 59048Date: Mon Sep 16 21:47:16 2013 -0700 59049 59050 Avoid use-after-free in dix/dixfonts.c: doImageText() [CVE-2013-4396] 59051 59052 Save a pointer to the passed in closure structure before copying it 59053 and overwriting the *c pointer to point to our copy instead of the 59054 original. If we hit an error, once we free(c), reset c to point to 59055 the original structure before jumping to the cleanup code that 59056 references *c. 59057 59058 Since one of the errors being checked for is whether the server was 59059 able to malloc(c->nChars * itemSize), the client can potentially pass 59060 a number of characters chosen to cause the malloc to fail and the 59061 error path to be taken, resulting in the read from freed memory. 59062 59063 Since the memory is accessed almost immediately afterwards, and the 59064 X server is mostly single threaded, the odds of the free memory having 59065 invalid contents are low with most malloc implementations when not using 59066 memory debugging features, but some allocators will definitely overwrite 59067 the memory there, leading to a likely crash. 59068 59069 Reported-by: Pedro Ribeiro <pedrib@gmail.com> 59070 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 59071 Reviewed-by: Julien Cristau <jcristau@debian.org> 59072 59073commit 8afe20d4e34adcfd29bdf43a01d55335ea2c5dba 59074Author: Alan Coopersmith <alan.coopersmith@oracle.com> 59075Date: Mon Sep 16 21:45:29 2013 -0700 59076 59077 Update GLX dependencies now that DRI & DRI2 are builtins, not modules 59078 59079 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 59080 59081commit 2704bdb24a2c7bac65b90e05f1a68438b34ecf58 59082Author: Alan Coopersmith <alan.coopersmith@oracle.com> 59083Date: Tue Sep 3 22:53:28 2013 -0700 59084 59085 DMX glxproxy: Don't allocate & copy data just to free it unused 59086 59087 Two functions in the DMX glxproxy code loop over all the backend 59088 screens, starting at the highest numbered and counting down to 59089 the lowest. 59090 59091 Previously, for each screen, the code would allocate a buffer 59092 large enough to read the reply from the backend, copy that reply 59093 into the buffer, and then if it wasn't the final screen, free it. 59094 Only the buffer from the final screen is used, to pass on to the 59095 client in the reply. 59096 59097 This modifies it to just immediately discard the responses from 59098 the screens as we loop through it, only doing the allocate & copy 59099 work for the one buffer we pass back to the client. 59100 59101 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 59102 Reviewed-by: Alex Deucher <aleander.deucher@amd.com> 59103 59104commit 6c06c268adbab63ebe7490489aa030724cbdc54b 59105Author: Alan Coopersmith <alan.coopersmith@oracle.com> 59106Date: Sun Aug 18 18:02:49 2013 -0700 59107 59108 Skip damage calls if DamageCreate fails in exa functions 59109 59110 Fixes parfait errors such as: 59111 Null pointer dereference (CWE 476): Write to null pointer pDamage 59112 at line 1833 of miext/damage/damage.c in function 'DamageRegister'. 59113 Function DamageCreate may return constant 'NULL' at line 1775, 59114 called at line 232 of exa/exa_migration_mixed.c 59115 in function 'exaPrepareAccessReg_mixed'. 59116 Constant 'NULL' passed into function DamageRegister, 59117 argument pDamage, from call at line 237. 59118 Null pointer introduced at line 1775 of miext/damage/damage.c 59119 in function 'DamageCreate'. 59120 Null pointer dereference (CWE 476): Write to null pointer pDamage 59121 at line 1833 of miext/damage/damage.c in function 'DamageRegister'. 59122 Function DamageCreate may return constant 'NULL' at line 1775, 59123 called at line 104 of exa/exa_mixed.c 59124 in function 'exaCreatePixmap_mixed'. 59125 Constant 'NULL' passed into function DamageRegister, 59126 argument pDamage, from call at line 109. 59127 Null pointer introduced at line 1775 of miext/damage/damage.c 59128 in function 'DamageCreate'. 59129 59130 Checks are similar to handling results of other calls to DamageCreate. 59131 59132 [ This bug was found by the Parfait 1.3.0 bug checking tool. 59133 http://labs.oracle.com/pls/apex/f?p=labs:49:::::P49_PROJECT_ID:13 ] 59134 59135 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 59136 59137commit 7cf1b595c8c8f9776a39559d2878cf90af3f2859 59138Author: Peter Hutterer <peter.hutterer@who-t.net> 59139Date: Wed Sep 4 15:34:29 2013 +1000 59140 59141 dix: only deliver for the current grab type 59142 59143 Use the grabtype to determine which type of event to send - all other events 59144 are pointless and may result in erroneous events being delivered. 59145 59146 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59147 59148commit 78944d62ffc7fec6f75b6f514ab7a139ba9bc77b 59149Author: Peter Hutterer <peter.hutterer@who-t.net> 59150Date: Wed Sep 4 15:34:28 2013 +1000 59151 59152 dix: don't attempt to deliver an event for a different grabtype 59153 59154 For an active grab, grab->eventMask can be either the core or the XI1 mask. 59155 With the overlap of event filters, calling DeliverOneGrabbedEvent(XI1) for a 59156 ProximityOut event will trigger if the client has selected for enter events - 59157 the filter is the same for both. 59158 59159 Thus, we end up delivering a proximity event to a client not expecting one. 59160 59161 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59162 59163commit 6159811a1aaf848016dbfa7bde68df097027870c 59164Author: Peter Hutterer <peter.hutterer@who-t.net> 59165Date: Wed Sep 4 15:34:27 2013 +1000 59166 59167 include: change grabtypes to start at 1 59168 59169 Avoid erroneous detection of an unset grabtype as CORE 59170 59171 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59172 59173commit be6ea80b79b3d6eb97106ea142e01acbdf059995 59174Author: Peter Hutterer <peter.hutterer@who-t.net> 59175Date: Wed Sep 4 15:34:26 2013 +1000 59176 59177 dix: only allow button and key events to freeze a sync'd pointer 59178 59179 If a client calls XAllowEvents(SyncPointer) it expects events as normal until 59180 the next button press or release event - that freezes the device. An e.g. 59181 proximity event must thus not freeze the pointer. 59182 59183 As per the spec, only button and key events may do so, so narrow it to these 59184 cases. 59185 59186 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59187 59188commit 93a27b2dd0d7ed51120f42456a91a7c59902ebc2 59189Author: Michele Baldessari <michele@acksyn.org> 59190Date: Sun Sep 15 19:30:38 2013 +0100 59191 59192 Xephyr: restore cursor visibility 59193 59194 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=69388 59195 59196 Commit c100211034ab69ce453a1644fb61c6808d7e3eda (dix: only show the cursor 59197 if a window defines one (#58398)) broke the default cursor behaviour in 59198 Xephyr (unless run with -retro). Restore the default cursor visibility 59199 so that '-retro' or '-host-cursor' are not needed to have a visible 59200 cursor. 59201 59202 Signed-off-by: Michele Baldessari <michele@acksyn.org> 59203 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59204 59205 as of ba387cf21f7d95987211f75d8024601e7d64e322 "ephyr: Use host (HW) cursors 59206 by default." this only applies if -sw-cursor is given on the cmdline. 59207 59208 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59209 59210commit 70efc799cbeec26f19831b3b14c0d4646198a077 59211Author: Michele Baldessari <michele@acksyn.org> 59212Date: Sun Sep 15 19:56:34 2013 +0100 59213 59214 Fix Xephyr compilation when DEBUG is enabled 59215 59216 When DEBUG is enabled Xephyr compilation fails: 59217 ephyrdriext.c:343:133: error: 'is_ok' undeclared (first use in this 59218 function) 59219 EPHYR_LOG("leave. is_ok:%d\n", is_ok); 59220 59221 Just reemove bogus is_ok variable. 59222 59223 Signed-off-by: Michele Baldessari <michele@acksyn.org> 59224 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59225 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59226 59227commit 5ac4bfca64bd75474e550b26bc1195a5ca245752 59228Author: Adam Jackson <ajax@redhat.com> 59229Date: Tue Sep 10 14:18:18 2013 -0400 59230 59231 input: calloc minimization for xi2mask_new 59232 59233 There's no reason to do this as (nmasks + 2) callocs, and it's a 59234 surprisingly hot path. Turns out you hit this ~once per passive grab, 59235 and you do a few bajillion passive grab changes every time you enter or 59236 leave the overview in gnome-shell. According to a callgrind of Xorg 59237 with gnome-shell-perf-tool run against it: 59238 59239 Ir before: 721437275 59240 Ir after: 454227086 59241 59242 Signed-off-by: Adam Jackson <ajax@redhat.com> 59243 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 59244 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59245 59246commit 44d502c6f1d03a9e548f0941cde0228b6fde8394 59247Author: Jon TURNEY <jon.turney@dronecode.org.uk> 59248Date: Sun Oct 6 16:45:52 2013 +0100 59249 59250 hw/xwin: Fix for "glx: Remove screen number from __GLXconfig" 59251 59252 Fix compilation after commit c3c976f54c3c282d6fa6c8360688e036bc43d210 "glx: 59253 Remove screen number from __GLXconfig" 59254 59255 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 59256 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 59257 Reviewed-by: Adam Jackson <ajax@redhat.com> 59258 59259commit 6a9bd103cb993e873cb82664b677ceed73c0bd85 59260Merge: ea3b7db4b 8aae28e3c 59261Author: Keith Packard <keithp@keithp.com> 59262Date: Mon Oct 7 17:29:38 2013 -0700 59263 59264 Merge remote-tracking branch 'jeremyhu/master' 59265 59266commit ea3b7db4b747f883ef7b8b524ee13bf8a63add14 59267Author: Emil Velikov <emil.l.velikov@gmail.com> 59268Date: Mon Oct 7 18:29:25 2013 +0100 59269 59270 configure.ac: Include missing proto declarations in SDK_REQUIRED_MODULES 59271 59272 Commits a1d41e311c21e, 7d859bd87834d & 3ed2c6e11298c made extinit.h require 59273 the XF86 Big Font, XRes & ScrnSaver proto headers, but failed to add them 59274 to the SDK_REQUIRED_MODULES so pkg-config would find them for driver builds. 59275 59276 Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com> 59277 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 59278 Signed-off-by: Keith Packard <keithp@keithp.com> 59279 59280commit 8aae28e3cbdd49bfa8d940a4a517c94a53b94f86 59281Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 59282Date: Sun Oct 6 09:20:45 2013 -0700 59283 59284 XQuartz: Dont't set screen on __GLXConfig 59285 59286 Fixes regression introduced by c3c976f54c3c282d6fa6c8360688e036bc43d210 59287 59288 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 59289 59290commit c5f8eb968efa970201712105a1ac880b517e7b6d 59291Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 59292Date: Sun Oct 6 09:18:15 2013 -0700 59293 59294 rootless: Use miCopyRegion instead of fbCopyRegion 59295 59296 Fixes regression introduced by e657635dbe6b92875b0e88370557c2cbab673a49 59297 59298 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 59299 59300commit ccbe17b1c6da1ad9d085fc8133cdd15dc7004a4a 59301Author: Pino Toscano <toscano.pino@tiscali.it> 59302Date: Wed Aug 28 17:15:03 2013 +0200 59303 59304 os: move <arpa/inet.h> for any !win32 system 59305 59306 It is needed in IPv6 configurations (for inet_pton) also when 59307 SIOCGIFCONF is not defined. 59308 59309 Signed-off-by: Pino Toscano <toscano.pino@tiscali.it> 59310 Acked-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 59311 Signed-off-by: Julien Cristau <jcristau@debian.org> 59312 59313commit c079b8e675dd2762c9e07fe5d8c07612299dceb7 59314Author: Pino Toscano <toscano.pino@tiscali.it> 59315Date: Wed Aug 28 17:04:48 2013 +0200 59316 59317 xfree86/hurd: include <hurd.h> 59318 59319 Needed for using get_privileged_port. 59320 59321 Signed-off-by: Pino Toscano <toscano.pino@tiscali.it> 59322 Acked-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 59323 Signed-off-by: Julien Cristau <jcristau@debian.org> 59324 59325commit f0659a7f68a5a334879af01a6a63742eab1bf5de 59326Author: Keith Packard <keithp@keithp.com> 59327Date: Fri Oct 4 16:44:46 2013 -0700 59328 59329 Bump version to 1.14.99.2 59330 59331 Signed-off-by: Keith Packard <keithp@keithp.com> 59332 59333commit f2d149b6d10edea7233d93c7e19f71decf11dc62 59334Author: Keith Packard <keithp@keithp.com> 59335Date: Fri Oct 4 16:43:41 2013 -0700 59336 59337 kdrive/ephyr: Remove duplicate ephyrExtensions and ephyrExtensionInit decls 59338 59339 These were duplicated when GLX support was re-added on two different branches. 59340 59341 Signed-off-by: Keith Packard <keithp@keithp.com> 59342 59343commit e0a678f0596f3547b98d74d5f0438b484b059458 59344Author: Gaetan Nadon <memsize@videotron.ca> 59345Date: Sat Sep 28 15:26:38 2013 -0400 59346 59347 xfree86: add a comment as to why the logdir is created 59348 59349 Without the logdir, the xserver will write the content of the log file on the 59350 terminal stating that it cannot be written and will stop. 59351 59352 Refer to https://bugs.freedesktop.org/show_bug.cgi?id=3889 59353 59354 Reviewed-By: Matt Dew <marcoz@osource.org> 59355 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 59356 Signed-off-by: Keith Packard <keithp@keithp.com> 59357 59358commit 5bdbf2dba3f7da4a593b286020eab0c7857e940c 59359Author: Gaetan Nadon <memsize@videotron.ca> 59360Date: Sat Sep 28 15:26:37 2013 -0400 59361 59362 xfree86: Use $(MKDIR_P) for better code portability 59363 59364 Still true that we should not use the lower case $(mkdir_p) version. 59365 However, remove the 2005 comment as the MKDIR_P is widely used now. 59366 59367 Reviewed-By: Matt Dew <marcoz@osource.org> 59368 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 59369 Signed-off-by: Keith Packard <keithp@keithp.com> 59370 59371commit b8cfb0dc8426754b477b561bc158b8a24c143ddf 59372Author: Gaetan Nadon <memsize@videotron.ca> 59373Date: Sat Sep 28 15:26:36 2013 -0400 59374 59375 Uninstall X link and CYGWIN libXorg.exe.a in local install targets 59376 59377 It is our duty to uninstall any files and/or directories that we installed 59378 through install-data-local and install-exec-hook. 59379 59380 Currently the X symbolic link to Xorg remains on disk after running 59381 make uninstall. 59382 59383 Note the exception for logdir which is usually shared by other modules. 59384 59385 Reviewed-By: Matt Dew <marcoz@osource.org> 59386 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 59387 Signed-off-by: Keith Packard <keithp@keithp.com> 59388 59389commit 35a528e49269c3df75486cc606e51a8c1c083c06 59390Author: Gaetan Nadon <memsize@videotron.ca> 59391Date: Sat Sep 28 15:26:35 2013 -0400 59392 59393 Xorg binary: use install-exec-hook rather than install-exec-local 59394 59395 The former was explicitly designed to execute additional code after the binary 59396 has been installed. The latter can be executed in any order, hence it's 59397 current dependency on install-binPROGRAMS as a workaround. 59398 59399 The CYGWIN libXorg.exe.a target is an installation target rather than 59400 a post-installation one, so it should not be done as a hook. It does not depend 59401 on the Xorg executable being installed. 59402 59403 Automake: 59404 "These hooks are run after all other install rules of the appropriate type, 59405 exec or data, have completed. So, for instance, it is possible to perform 59406 post-installation modifications using an install hook". 59407 59408 "With the -local targets, there is no particular guarantee of execution order; 59409 typically, they are run early, but with parallel make, there is no way 59410 to be sure of that". 59411 59412 Reviewed-by: Eric Anholt <eric@anholt.net> 59413 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 59414 Signed-off-by: Keith Packard <keithp@keithp.com> 59415 59416commit a1d87576a30008253eb0cc92b44c9718f6665062 59417Author: Gaetan Nadon <memsize@videotron.ca> 59418Date: Sat Sep 28 15:26:34 2013 -0400 59419 59420 Use $(LN_S) provided by AC_PROG_LN_S macro to create links 59421 59422 For better code portability. 59423 59424 Reviewed-By: Matt Dew <marcoz@osource.org> 59425 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 59426 Signed-off-by: Keith Packard <keithp@keithp.com> 59427 59428commit 28c8e1910797970dbf1d40794f1804a8fcc812bc 59429Author: Gaetan Nadon <memsize@videotron.ca> 59430Date: Sat Sep 28 15:26:33 2013 -0400 59431 59432 The Xorg binary is missing the extension $(EXEEXT) in the makefile 59433 59434 This is not a problem on UNIX platforms, but on CYGWIN it creates a broken 59435 link to Xorg rather than a link to Xorg.exe. 59436 59437 From the CYGWIN log on tinderbox, we can see that the executable Xorg.exe is 59438 installed correctly. We can see the command used to create the link: 59439 59440 (cd /jhbuild/install/[...]/install/bin && rm -f X && ln -s Xorg X) 59441 59442 Note that the "relink" makefile target correctly appends $(EXEEXT) to Xorg. 59443 59444 Reviewed-By: Matt Dew <marcoz@osource.org> 59445 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 59446 Signed-off-by: Keith Packard <keithp@keithp.com> 59447 59448commit f3b529bf25db2700addbafc7aa4fc5b126ba9929 59449Merge: 4399bd383 abc2bfca1 59450Author: Keith Packard <keithp@keithp.com> 59451Date: Fri Oct 4 14:04:48 2013 -0700 59452 59453 Merge remote-tracking branch 'anholt/ephyr-fixes' 59454 59455commit 4399bd38326b81f7b92365346504d7c6354772a9 59456Author: Gaetan Nadon <memsize@videotron.ca> 59457Date: Mon Sep 30 13:30:46 2013 -0400 59458 59459 miext/shadow: missing c2p_core.h breaks "make distcheck" target. 59460 59461 Reviewed-by: Julien Cristau <jcristau@debian.org> 59462 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 59463 Signed-off-by: Keith Packard <keithp@keithp.com> 59464 59465commit 9ccb4a63982942fdce19609d2cadbd8e51708462 59466Merge: 5d2ec6933 da5e20127 59467Author: Keith Packard <keithp@keithp.com> 59468Date: Fri Oct 4 13:54:55 2013 -0700 59469 59470 Merge remote-tracking branch 'whot/for-keith' 59471 59472commit 5d2ec6933f6bb5116666ca1bfb07f0f20bf2a0b5 59473Merge: 7d3d4ae55 6ee4d9f94 59474Author: Keith Packard <keithp@keithp.com> 59475Date: Fri Oct 4 13:50:04 2013 -0700 59476 59477 Merge remote-tracking branch 'ajax/xserver-next' 59478 59479commit da5e20127a5b6e11a47eb6e61f3ea87b056cdf5f 59480Author: Gaetan Nadon <memsize@videotron.ca> 59481Date: Thu Sep 19 19:13:35 2013 -0400 59482 59483 test: add new os executable to .gitignore 59484 59485 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 59486 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59487 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59488 59489commit c73c36b537f996574628e69681833ea37dec2b6e 59490Author: Laércio de Sousa <lbsousajr@gmail.com> 59491Date: Wed Sep 18 13:42:17 2013 -0300 59492 59493 xserver: enable InputClass option "GrabDevice" by default for non-seat0 seats (#69478) 59494 59495 This patch contributes to fill the remaining gaps which make 59496 systemd-multi-seat-x wrapper still necessary in some multiseat setups. 59497 59498 This also replaces previous evdev patch that does the same thing 59499 for that particular driver. 59500 59501 When option "-seat" is passed with an argument different from "seat0", 59502 option "GrabDevice" for input devices is enabled by default 59503 (no need of enabling it in xorg.conf's "InputClass" section). 59504 59505 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=69478 59506 59507 Signed-off-by: Laércio de Sousa <lbsousajr@gmail.com> 59508 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59509 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59510 59511commit 160c2db32d0b9ee5316a44179a6af4f4722e3d0a 59512Author: Peter Hutterer <peter.hutterer@who-t.net> 59513Date: Wed Sep 18 16:04:44 2013 +1000 59514 59515 dmx: provide enough space for axis mappings 59516 59517 relmap/absmap is used as a evdev-axis-to-x-axis mapping. ABS_X maps to 59518 axis 0, ABS_Y to 1, etc. skipping over non-existing axes so that the third bit 59519 set in the ABS_* range is axis 2, and so on. This requires us to actually have 59520 enough space to have all the ABS_*/REL_* range. 59521 59522 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 59523 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 59524 59525commit 7d3d4ae55dd6ee338439e2424ac423b1df80501b 59526Author: Keith Packard <keithp@keithp.com> 59527Date: Fri Sep 20 14:43:41 2013 -0500 59528 59529 damage: Must translate initial window damage by window offset 59530 59531 Damage is reported relative to the drawable origin, but the window 59532 borderClip is absolute. Translate the region by the window position 59533 before reporting damage to adjust. 59534 59535 Reported-by: Adam Jackson <ajax@redhat.com> 59536 Signed-off-by: Keith Packard <keithp@keithp.com> 59537 Reviewed-by: Adam Jackson <ajax@redhat.com> 59538 59539commit 8010d3a48bd0b224dcb0883e39c2351ad364d846 59540Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 59541Date: Sun Sep 15 08:10:33 2013 -0700 59542 59543 XQuartz: pbproxy: Fix build with -DDEBUG 59544 59545 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 59546 59547commit 39c548da0cf0522ad246630605b96751ab6d39c0 59548Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 59549Date: Sun Sep 15 08:07:44 2013 -0700 59550 59551 XQuartz: Fix build with moved pseudoramiX 59552 59553 Regression from: e716baedc4d8e52a60f43ef21aba771b340d8c8b 59554 59555 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 59556 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 59557 59558commit 6ee4d9f94ae2cc27c39381d7554134d9b4807a65 59559Author: Adam Jackson <ajax@redhat.com> 59560Date: Tue Jul 9 11:19:11 2013 -0400 59561 59562 glx: Fill in some missing attributes from DoGetFBConfigs 59563 59564 Reviewed-by: Eric Anholt <eric@anholt.net> 59565 Signed-off-by: Adam Jackson <ajax@redhat.com> 59566 59567commit b257fabff00ae925bb7e70364dd1d738634087c9 59568Author: Adam Jackson <ajax@redhat.com> 59569Date: Thu Jul 4 10:31:40 2013 -0400 59570 59571 glx: Remove a dead comment 59572 59573 Reviewed-by: Eric Anholt <eric@anholt.net> 59574 Signed-off-by: Adam Jackson <ajax@redhat.com> 59575 59576commit abd086502123086ce3d1fba25c159bb022d2750f 59577Author: Adam Jackson <ajax@redhat.com> 59578Date: Mon Jul 8 17:01:13 2013 -0400 59579 59580 glx: Catch another failure case in drawable creation 59581 59582 Reviewed-by: Eric Anholt <eric@anholt.net> 59583 Signed-off-by: Adam Jackson <ajax@redhat.com> 59584 59585commit c3c976f54c3c282d6fa6c8360688e036bc43d210 59586Author: Adam Jackson <ajax@redhat.com> 59587Date: Mon Jul 8 12:28:56 2013 -0400 59588 59589 glx: Remove screen number from __GLXconfig 59590 59591 Not used. There's no real reason to match against this instead of 59592 matching against fbconfig or visual ID anyway. 59593 59594 Reviewed-by: Eric Anholt <eric@anholt.net> 59595 Signed-off-by: Adam Jackson <ajax@redhat.com> 59596 59597commit 1d1484e9bd55d1fa8316b1574754f28d5d31a076 59598Author: Adam Jackson <ajax@redhat.com> 59599Date: Mon Jul 8 12:09:13 2013 -0400 59600 59601 glx: Remove pixmapMode from __GLXconfig 59602 59603 This has never been filled in with anything meaningful afaict, and you 59604 can't get to it from the client in any event. 59605 59606 Reviewed-by: Eric Anholt <eric@anholt.net> 59607 Signed-off-by: Adam Jackson <ajax@redhat.com> 59608 59609commit 34e6e60105796ad015469c10921c5ff09acf7eb0 59610Author: Adam Jackson <ajax@redhat.com> 59611Date: Tue Jul 2 12:33:36 2013 -0400 59612 59613 glx: Remove support for NV_vertex_program and NV_fragment_program 59614 59615 Mesa doesn't implement these anymore, never really did outside of swrast 59616 anyway. 59617 59618 Reviewed-by: Eric Anholt <eric@anholt.net> 59619 Signed-off-by: Adam Jackson <ajax@redhat.com> 59620 59621commit acf14c1de766c4ea206cd7ce7f25b780a589d33c 59622Author: Adam Jackson <ajax@redhat.com> 59623Date: Tue Jul 2 10:35:19 2013 -0400 59624 59625 glx: realloc style fix in RenderLarge 59626 59627 Reviewed-by: Eric Anholt <eric@anholt.net> 59628 Signed-off-by: Adam Jackson <ajax@redhat.com> 59629 59630commit 9ebf739a6864c9ec38bf72f63ef2e3b9cd1951db 59631Author: Adam Jackson <ajax@redhat.com> 59632Date: Tue Jul 2 10:30:40 2013 -0400 59633 59634 glx: Eliminate a small malloc from QueryContext 59635 59636 No reason to have that be a failure path. 59637 59638 Reviewed-by: Eric Anholt <eric@anholt.net> 59639 Signed-off-by: Adam Jackson <ajax@redhat.com> 59640 59641commit b99f7975407f111b99d772cd28224d7dc1b34fe4 59642Author: Adam Jackson <ajax@redhat.com> 59643Date: Fri Nov 9 18:05:27 2012 -0500 59644 59645 glx: Handle failure to create the pixmap backing the pbuffer 59646 59647 We happen not to sanitize the width/height we pass to CreatePixmap here, 59648 oops. It's not exploitable, but it's certainly a crash, so let's just 59649 throw BadAlloc instead. 59650 59651 Reviewed-by: Eric Anholt <eric@anholt.net> 59652 Signed-off-by: Adam Jackson <ajax@redhat.com> 59653 59654commit 22fbfdcb31f335c22053a0e29f0e73c03cd70a76 59655Author: Adam Jackson <ajax@redhat.com> 59656Date: Fri Nov 9 17:52:12 2012 -0500 59657 59658 glx: Implement GLX_PRESERVED_CONTENTS drawable attribute 59659 59660 We back pixmaps with pbuffers so they're never actually clobbered. Say 59661 so when asked. 59662 59663 Reviewed-by: Eric Anholt <eric@anholt.net> 59664 Signed-off-by: Adam Jackson <ajax@redhat.com> 59665 59666commit 2e20b8382cdf37ae5a2802f85b8f06a370f9645d 59667Author: Adam Jackson <ajax@redhat.com> 59668Date: Fri Nov 9 17:12:49 2012 -0500 59669 59670 glx: Implement GLX_FBCONFIG_ID in GetDrawableAttributes 59671 59672 Required by GLX 1.4, section 3.3.6, "Querying Attributes". 59673 59674 Reviewed-by: Eric Anholt <eric@anholt.net> 59675 Signed-off-by: Adam Jackson <ajax@redhat.com> 59676 59677commit 0d76191baeee005982754ac18ca995742e32d8c1 59678Author: Adam Jackson <ajax@redhat.com> 59679Date: Fri Nov 9 17:10:57 2012 -0500 59680 59681 glx: Implement GLX_{WIDTH,HEIGHT} in GetDrawableAttributes 59682 59683 Required by GLX 1.4, section 3.3.6, "Querying Attributes". 59684 59685 Reviewed-by: Eric Anholt <eric@anholt.net> 59686 Signed-off-by: Adam Jackson <ajax@redhat.com> 59687 59688commit d11f13e383dda1e9cabe32793d261d46edc44418 59689Author: Adam Jackson <ajax@redhat.com> 59690Date: Fri Nov 9 17:07:20 2012 -0500 59691 59692 glx: Compute number of attributes in GetDrawableAttributes on the fly 59693 59694 This doesn't have any effect yet, but is needed to properly build the 59695 reply for pbuffers. 59696 59697 Reviewed-by: Eric Anholt <eric@anholt.net> 59698 Signed-off-by: Adam Jackson <ajax@redhat.com> 59699 59700commit 468b57324f838144d79e3a9d85a0bde853578c4b 59701Author: Adam Jackson <ajax@redhat.com> 59702Date: Fri Nov 9 16:48:16 2012 -0500 59703 59704 glx: Style fixes 59705 59706 Reviewed-by: Eric Anholt <eric@anholt.net> 59707 Signed-off-by: Adam Jackson <ajax@redhat.com> 59708 59709commit 2b181ad0952851f47f23dcb1a36471685313c13e 59710Author: Adam Jackson <ajax@redhat.com> 59711Date: Fri Nov 9 16:33:54 2012 -0500 59712 59713 glx: Pull GLX vendor string out of __GLXscreen 59714 59715 Given how we're currently implementing GLX this can't meaningfully vary 59716 per-screen. 59717 59718 Reviewed-by: Eric Anholt <eric@anholt.net> 59719 Signed-off-by: Adam Jackson <ajax@redhat.com> 59720 59721commit b9c489cf13e1ad35b7d9bea9de8d85e929374d17 59722Author: Adam Jackson <ajax@redhat.com> 59723Date: Wed Oct 31 10:05:24 2012 -0400 59724 59725 glx: Remove unused bits from the context struct 59726 59727 Reviewed-by: Eric Anholt <eric@anholt.net> 59728 Signed-off-by: Adam Jackson <ajax@redhat.com> 59729 59730commit f579cc85c29f52b575c62e6ae1e16eb87c506bd3 59731Author: Adam Jackson <ajax@redhat.com> 59732Date: Thu Jul 4 10:08:40 2013 -0400 59733 59734 glxproxy: Don't track GLCLientm{aj,in}orVersion 59735 59736 Basically just a port of 62f06b0d to glxproxy. 59737 59738 Reviewed-by: Eric Anholt <eric@anholt.net> 59739 Signed-off-by: Adam Jackson <ajax@redhat.com> 59740 59741commit e549ffb5f19cf044bc3e2d368a6f081f8b84bfaa 59742Author: Adam Jackson <ajax@redhat.com> 59743Date: Tue Oct 30 18:20:43 2012 -0400 59744 59745 dmx/glx: Remove unused __glXNop 59746 59747 Reviewed-by: Eric Anholt <eric@anholt.net> 59748 Signed-off-by: Adam Jackson <ajax@redhat.com> 59749 59750commit d962b8ed272d4a8921c0dadf16cd926484e30d10 59751Author: Adam Jackson <ajax@redhat.com> 59752Date: Thu Aug 22 16:42:23 2013 -0400 59753 59754 damageext: Style fix 59755 59756 Reviewed-by: Eric Anholt <eric@anholt.net> 59757 Signed-off-by: Adam Jackson <ajax@redhat.com> 59758 59759commit c90add0f1ce0c85aea7a60635de765848d778266 59760Author: Adam Jackson <ajax@redhat.com> 59761Date: Wed Aug 21 11:57:24 2013 -0400 59762 59763 damage: Remove "post-rendering" hooks 59764 59765 This is a revert of 974db58f5b730c3770ee461665a02dd4334d1dea. There are 59766 no consumers of this API, we must not need it. 59767 59768 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 59769 Reviewed-by: Eric Anholt <eric@anholt.net> 59770 Signed-off-by: Adam Jackson <ajax@redhat.com> 59771 59772commit 0f3f8e4295d4c05f33b93e0ffd2f6f8cd5759bf7 59773Author: Adam Jackson <ajax@redhat.com> 59774Date: Wed Aug 21 11:13:11 2013 -0400 59775 59776 damage: Don't rewrite Text ops to GlyphBlt ops 59777 59778 There's no particularly good reason to, and it breaks Xnest. 59779 59780 Bugzilla: http://bugs.freedesktop.org/2454 59781 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 59782 Reviewed-by: Eric Anholt <eric@anholt.net> 59783 Signed-off-by: Adam Jackson <ajax@redhat.com> 59784 59785commit d08966227e7d567df8d26eebc80f35f886e59a4a 59786Author: Adam Jackson <ajax@redhat.com> 59787Date: Mon Aug 26 14:03:51 2013 -0400 59788 59789 damage: Simplify DamageUnregister 59790 59791 You can only register one drawable on a given damage, so there's no 59792 reason to require the caller to specify the drawable, the damage is 59793 enough. The implementation would do something fairly horrible if you 59794 _did_ pass mismatched drawable and damage, so let's avoid the problem 59795 entirely. 59796 59797 v2: Simplify xf86RotateDestroy even more [anholt] 59798 59799 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 59800 Reviewed-by: Eric Anholt <eric@anholt.net> 59801 Signed-off-by: Adam Jackson <ajax@redhat.com> 59802 59803commit 28708a045de7d9043d20fb06b61c44a46eb5526b 59804Author: Adam Jackson <ajax@redhat.com> 59805Date: Mon Aug 26 13:52:14 2013 -0400 59806 59807 damage: Implicitly unregister on destroy 59808 59809 There's no reason not to, and it simplifies quite a few callers. 59810 59811 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 59812 Reviewed-by: Eric Anholt <eric@anholt.net> 59813 Signed-off-by: Adam Jackson <ajax@redhat.com> 59814 59815commit e657635dbe6b92875b0e88370557c2cbab673a49 59816Author: Adam Jackson <ajax@redhat.com> 59817Date: Thu Oct 18 14:24:23 2012 -0400 59818 59819 fb: Remove unused compatibility wrappers 59820 59821 Originally added in early 2009, not being used elsewhere anymore. 59822 59823 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59824 Signed-off-by: Adam Jackson <ajax@redhat.com> 59825 59826commit e4b15125bb06c97f1f34f7074fcdb597e718399c 59827Author: Adam Jackson <ajax@redhat.com> 59828Date: Thu Oct 18 14:24:22 2012 -0400 59829 59830 mipointer: Remove deprecated miPointerCurrentScreen 59831 59832 The only remaining use was in some debugging code in DMX. 59833 59834 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59835 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 59836 Signed-off-by: Adam Jackson <ajax@redhat.com> 59837 59838commit b5d66b9513a0f1e13f5c5bc9e0958f4eac9bf6c3 59839Author: Adam Jackson <ajax@redhat.com> 59840Date: Thu Oct 18 14:24:21 2012 -0400 59841 59842 xfree86: Remove deprecated unimplemented xf86MapReadSideEffects 59843 59844 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59845 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 59846 Signed-off-by: Adam Jackson <ajax@redhat.com> 59847 59848commit abbd85742a390e18497b96902a0c0d976739e3bd 59849Author: Adam Jackson <ajax@redhat.com> 59850Date: Thu Oct 18 14:24:20 2012 -0400 59851 59852 dix: FIXES is not optional 59853 59854 It's already not optional at configure time, this just makes it so at 59855 build time too. 59856 59857 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59858 Signed-off-by: Adam Jackson <ajax@redhat.com> 59859 59860commit dff81687f5eac3eac9b49f58d4654cc30add2547 59861Author: Adam Jackson <ajax@redhat.com> 59862Date: Thu Oct 18 14:24:19 2012 -0400 59863 59864 vbe: Don't try to load the ddc submodule 59865 59866 DDC is built into the server now. 59867 59868 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59869 Signed-off-by: Adam Jackson <ajax@redhat.com> 59870 59871commit bb82191bcfa8042f8d98ec8b1c6616dc31d2c5c8 59872Author: Adam Jackson <ajax@redhat.com> 59873Date: Thu Oct 18 14:24:18 2012 -0400 59874 59875 misprite: Delete some dead private storage 59876 59877 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59878 Signed-off-by: Adam Jackson <ajax@redhat.com> 59879 59880commit 3639ab4ce25dc82a3f8592b7322198f1d2aca5f9 59881Author: Adam Jackson <ajax@redhat.com> 59882Date: Tue Oct 16 14:56:25 2012 -0400 59883 59884 dri2: Don't bother with xf86LoaderCheckSymbol("DRI2Connect") 59885 59886 The DRI2 code is now built-in to the server, even for Xorg. The only 59887 thing this could protect against is trying to run a libglx built with 59888 DRI2 support against an Xorg built without it, which is firmly in 59889 "doctor it hurts when I do this" territory. 59890 59891 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59892 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 59893 Signed-off-by: Adam Jackson <ajax@redhat.com> 59894 59895commit 13b507409f9e01adebba50e7eb801b52e52692ba 59896Author: Adam Jackson <ajax@redhat.com> 59897Date: Tue Oct 16 14:56:23 2012 -0400 59898 59899 list: Avoid using X types 59900 59901 In particular, Bool. This is not an ABI break: 59902 59903 /usr/include/X11/Xdefs.h:typedef int Bool; 59904 59905 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59906 Signed-off-by: Adam Jackson <ajax@redhat.com> 59907 59908commit dbe10ef0e35659ee0e5db73f29670a3894a3e3f5 59909Author: Adam Jackson <ajax@redhat.com> 59910Date: Tue Oct 16 14:56:22 2012 -0400 59911 59912 dbe: Fold the window private private into the window private 59913 59914 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59915 Signed-off-by: Adam Jackson <ajax@redhat.com> 59916 59917commit 1489637d567b0034a916f88079b5f360e7fce99c 59918Author: Adam Jackson <ajax@redhat.com> 59919Date: Tue Oct 16 14:56:21 2012 -0400 59920 59921 dbe: Fold midbe reset into dix 59922 59923 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59924 Signed-off-by: Adam Jackson <ajax@redhat.com> 59925 59926commit 44d770750069486a8452fb2384f9b85745268e61 59927Author: Adam Jackson <ajax@redhat.com> 59928Date: Tue Oct 16 14:56:20 2012 -0400 59929 59930 dbe: Drop the unused idiom support 59931 59932 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59933 Signed-off-by: Adam Jackson <ajax@redhat.com> 59934 59935commit 8cce0cf4d378f6fa52bc2cd277737b8986ab683c 59936Author: Adam Jackson <ajax@redhat.com> 59937Date: Tue Oct 16 14:56:19 2012 -0400 59938 59939 damage: Remove the 'damage window' resource type. 59940 59941 Note that the existing code never actually creates a resource of this 59942 type, so it can't possibly do anything. This is clearly meant to track 59943 window destruction and clean up any associated Damages, but that's 59944 already handled by miext/damage's DestroyWindow wrapper. 59945 59946 Previous discussion: 59947 http://lists.freedesktop.org/archives/xorg-devel/2011-March/020847.html 59948 59949 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59950 Signed-off-by: Adam Jackson <ajax@redhat.com> 59951 59952commit e674815e0a217ac01ec2d6a8517f8466b27fb39f 59953Author: Adam Jackson <ajax@redhat.com> 59954Date: Tue Oct 16 14:56:18 2012 -0400 59955 59956 dmx: Remove ShadowFB support 59957 59958 This has been listed as deprecated ever since DMX was merged. 59959 59960 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59961 Signed-off-by: Adam Jackson <ajax@redhat.com> 59962 59963commit 19d3a7cb5db3188210830efa746987ba50d839e9 59964Author: Adam Jackson <ajax@redhat.com> 59965Date: Tue Oct 16 14:56:17 2012 -0400 59966 59967 dix: Remove a dead parameter from DoGetImage 59968 59969 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59970 Signed-off-by: Adam Jackson <ajax@redhat.com> 59971 59972commit 4dca026880da90cff9734cf5787c5b57c02e7310 59973Author: Adam Jackson <ajax@redhat.com> 59974Date: Tue Oct 16 14:56:16 2012 -0400 59975 59976 mipointer: Remove EnqueueEvent from miPointerScreenFuncRec 59977 59978 No DDX overrode this, and we never actually called through that slot 59979 anyway. 59980 59981 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59982 Signed-off-by: Adam Jackson <ajax@redhat.com> 59983 59984commit ad076dc6e82019defa422500710b3b42c6309e39 59985Author: Adam Jackson <ajax@redhat.com> 59986Date: Tue Oct 16 14:56:15 2012 -0400 59987 59988 mipointer: Flatten calls to mieqSwitchScreen 59989 59990 No DDX was overriding this. 59991 59992 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 59993 Signed-off-by: Adam Jackson <ajax@redhat.com> 59994 59995commit e2b14a1d3bb27e86ff1d7aaf59af850d06263ef5 59996Author: Adam Jackson <ajax@redhat.com> 59997Date: Tue Oct 16 14:56:14 2012 -0400 59998 59999 os: Factor out some common code in input buffer handling 60000 60001 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 60002 Signed-off-by: Adam Jackson <ajax@redhat.com> 60003 60004commit 65fca558848acd907483d3c85c7cd4db5560f1b4 60005Author: Radek Doulik <rodo@novell.com> 60006Date: Tue Aug 13 08:45:47 2013 +0200 60007 60008 xnest: Ignore GetImage() error in xnestGetImage() 60009 60010 When an Xnest instance is not viewable it will crash when a client in 60011 that instance calls GetImage. This is because the Xnest server will 60012 itself receives a BadMatch error. 60013 This patch ignores the error. The application which has requested the 60014 image will receive garbage - this however is fully legal according 60015 to the specs as obscured areas will always contain garbage if there 60016 isn't some sort of backing store as discussed in 60017 https://bugs.freedesktop.org/show_bug.cgi?id=9488 60018 The applied patch is a version from Dadek Doulik. 60019 60020 v2: Call XSync() before changing error handlers as suggested by 60021 Daniel Stone <daniel@fooishbar.org>. 60022 60023 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 60024 Signed-off-by: Egbert Eich <eich@freedesktop.org> 60025 60026commit b902c8abb6d03e68c93e13881a350523b5ac900c 60027Author: Michal Srb <msrb@novell.com> 60028Date: Tue Aug 6 16:56:53 2013 +0200 60029 60030 Xnest: Implement xnestModifyPixmapHeader 60031 60032 Xnest variant of ModifyPixmapHeader that creates new Pixmap in parent X 60033 server if it's size is modified from 0x0 to anything bigger. 60034 60035 xnestCreatePixmap doesn't create pixmap in parent X server if it has 60036 dimensions 0x0. If it is later resized and accessed, Xnest will be 60037 aborted with BadDrawable error from parent X server because it will 60038 use XID 0. This happens with ScratchPixmap, for example as used from 60039 XaceCensorImage. Applications using XACE crash Xnest. 60040 60041 Reviewed-by: Adam Jackson <ajax@redhat.com> 60042 60043commit 132507eba93e4dfd466d5c30c9b818cdc6f8536e 60044Author: Peter Harris <pharris@opentext.com> 60045Date: Mon Jul 15 19:44:45 2013 -0400 60046 60047 security: Don't count RetainPermanent clients twice 60048 60049 If a RetainPermanent client is subsequently killed by a KillClient 60050 request, the reference count is decremented twice. This can cause the 60051 server to prematurely kill other clients using the same Authorization. 60052 60053 Reviewed-by: Adam Jackson <ajax@redhat.com> 60054 Signed-off-by: Peter Harris <pharris@opentext.com> 60055 60056commit 20c2a3bcb11b5baf564e2c73a477ba23f5ae2b10 60057Author: Peter Harris <pharris@opentext.com> 60058Date: Mon Jul 15 19:44:29 2013 -0400 60059 60060 mi: Avoid stack smash when drawing dashed lines 60061 60062 X.org Bug 54013 <https://bugs.freedesktop.org/show_bug.cgi?id=54013> 60063 60064 Reviewed-by: Adam Jackson <ajax@redhat.com> 60065 Signed-off-by: Peter Harris <pharris@opentext.com> 60066 60067commit 9bf46610a9d20962854016032de4567974e87957 60068Author: Chris Wilson <chris@chris-wilson.co.uk> 60069Date: Fri Jun 21 22:58:31 2013 +0100 60070 60071 os: Immediately queue initial WriteToClient 60072 60073 If we immediately put the WriteToClient() buffer into the socket's write 60074 queue, not only do we benefit from sending the response back to client 60075 earlier, but we also avoid the overhead of copying the data into our own 60076 staging buffer and causing extra work in the next select(). The write is 60077 effectively free as typically we may only send one reply per client per 60078 select() call, so the cost of the FlushClient() is the same. 60079 60080 shmget10: 26400 -> 110000 60081 getimage10: 25000 -> 108000 60082 60083 shmget500: 3160 -> 13500 60084 getimage500: 1000 -> 1010 60085 60086 The knock-on effect is that on a mostly idle composited desktop, the CPU 60087 overhead is dominated by the memmove in WriteToClient, which is in turn 60088 eliminated by this patch. 60089 60090 Reviewed-by: Adam Jackson <ajax@redhat.com> 60091 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 60092 60093commit ceffb34774d44ada83cc1994d20c8d65b46bd555 60094Author: Laurent Carlier <lordheavym@gmail.com> 60095Date: Sat Jun 8 00:43:56 2013 +0200 60096 60097 ephyr: restore GLX support 60098 60099 It was removed since version 1.13 60100 60101 Signed-off-by: Sebastien Bacher <seb128@ubuntu.com> 60102 Acked-by: Daniel Stone <daniel@fooishbar.org> 60103 60104commit f37972f4a67b49ba9649310a6778bd1bf31e0130 60105Author: Keith Packard <keithp@keithp.com> 60106Date: Wed Apr 24 14:58:27 2013 -0700 60107 60108 miext/sync: Fix multi-screen support in SyncFence initialization 60109 60110 miSyncSetup was checking to see if the screen private key had been 60111 registered, and if so, skipping the setup of the provided screen. 60112 60113 Instead, it should register the private index only once, but then 60114 initialize the screen unless it has already been initialized. This 60115 latter step allows drivers to initialize the sync private structures 60116 before the Sync extension itself is initialized. 60117 60118 Reviewed-by: Adam Jackson <ajax@redhat.com> 60119 Signed-off-by: Keith Packard <keithp@keithp.com> 60120 60121commit 1c8beedfe4343377fa3504f8ab71430cc137b22c 60122Author: Adam Jackson <ajax@redhat.com> 60123Date: Wed Aug 21 14:12:52 2013 -0400 60124 60125 randr: Fix a copypasta bug in CRTC confinement 60126 60127 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 60128 Signed-off-by: Adam Jackson <ajax@redhat.com> 60129 60130commit 47ff382d1fce25a8b097d45b79489e891f1f1228 60131Merge: 6f89ae3e6 3752a8158 60132Author: Keith Packard <keithp@keithp.com> 60133Date: Mon Sep 9 15:13:42 2013 -0700 60134 60135 Merge remote-tracking branch 'jturney/master' 60136 60137commit 6f89ae3e64c4dfeea508813e546c10ba1da3ea8e 60138Author: Thomas Klausner <wiz@NetBSD.org> 60139Date: Wed Sep 4 20:06:07 2013 +0200 60140 60141 Fix typo in configure warning. 60142 60143 Signed-off-by: Thomas Klausner <wiz@NetBSD.org> 60144 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 60145 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 60146 60147commit 47218a6e09549781fd61dbf5e0d3d5c81da64323 60148Author: Thomas Klausner <wiz@NetBSD.org> 60149Date: Wed Sep 4 20:05:51 2013 +0200 60150 60151 Fix bug in cursor handling. 60152 60153 CreateCursor (Xlib call XCreatePixmapCursor) with a non-bitmap 60154 source pixmap and a None mask is supposed to error out with BadMatch, 60155 but didn't. 60156 60157 From der Mouse <mouse@Rodents-Montreal.ORG>, changed following 60158 comments by Alan Coopersmith <alan.coopersmith@oracle.com>. 60159 60160 Signed-off-by: Thomas Klausner <wiz@NetBSD.org> 60161 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 60162 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 60163 60164commit 1110b71e360195aab040d835b54540ab558638c5 60165Author: Chris Clayton <chris2553@googlemail.com> 60166Date: Wed Sep 4 15:42:04 2013 +1000 60167 60168 kdrive: fix build error on gcc 4.8 for out-of-bounds array access 60169 60170 I'm getting a error building xorg-server-1.14.1.902 with thelatest snapshot 60171 of gcc-4.8: 60172 60173 input.c:225:43: error: array subscript is above array bounds 60174 [-Werror=array-bounds] 60175 60176 This is because kdNumInputFds can become equal to KD_MAX_INPUT_FDS in 60177 KdRegisterFd(). This means that in KdUnregisterFd(), kdInputFds[j + 1] can 60178 be beyond the end of the array. 60179 60180 Signed-off-by: Chris Clayton <chris2553@googlemail.com> 60181 Reviewed-by: Keith Packard <keithp@keithp.com> 60182 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 60183 60184commit abc2bfca16adcd1e5657d4ce54f4e1335d5ed53f 60185Author: Sebastien Bacher <seb128@ubuntu.com> 60186Date: Fri Aug 23 21:21:44 2013 +0200 60187 60188 kdrive: initialize GLX for xephyr 60189 60190 Like commit ac1a60e7b6f06fd075cc5bf55d6bc67206a01d29, re-add 60191 initialization of GLX after it was accidentally dropped from non-Xorg 60192 servers in 5f5bbbe543f65c48ecbb5cce80116a86ca3fbe86. 60193 60194 Signed-off-by: Sebastien Bacher <seb128@ubuntu.com> 60195 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=62346 60196 Reviewed-by: Eric Anholt <eric@anholt.net> 60197 Reviewed-by: Adam Jackson <ajax@redhat.com> 60198 Reviewed-by: Julien Cristau <jcristau@debian.org> 60199 Signed-off-by: Eric Anholt <eric@anholt.net> 60200 60201commit aa5534ec6931ffcc81cca76b98785d57de8e0e52 60202Author: Eric Anholt <eric@anholt.net> 60203Date: Fri Aug 23 11:23:00 2013 -0700 60204 60205 ephyr: Do grab/ungrab for ctrl+shift, not just shift+ctrl. 60206 60207 Given that the window title says "ctrl+shift", having pressing those 60208 keys in that order not ungrab you is fairly mean. 60209 60210 Signed-off-by: Eric Anholt <eric@anholt.net> 60211 Reviewed-by: Adam Jackson <ajax@redhat.com> 60212 Reviewed-by: Julien Cristau <jcristau@debian.org> 60213 60214commit 5cc15e1c427bacd73db7381df3474a23f7616624 60215Author: Eric Anholt <eric@anholt.net> 60216Date: Fri Aug 23 11:19:17 2013 -0700 60217 60218 ephyr: Flush the X connection when updating the window title. 60219 60220 Otherwise when you're doing the ctrl-shift mouse grab thing, you 60221 don't know what state you're in until the next rendering occurs. 60222 60223 Signed-off-by: Eric Anholt <eric@anholt.net> 60224 Reviewed-by: Adam Jackson <ajax@redhat.com> 60225 Reviewed-by: Julien Cristau <jcristau@debian.org> 60226 60227commit ba387cf21f7d95987211f75d8024601e7d64e322 60228Author: Eric Anholt <eric@anholt.net> 60229Date: Fri Aug 23 11:17:14 2013 -0700 60230 60231 ephyr: Use host (HW) cursors by default. 60232 60233 Unless you're working on the sw cursor rendering code, you surely want 60234 to have real hardware cursors. 60235 60236 Signed-off-by: Eric Anholt <eric@anholt.net> 60237 Reviewed-by: Adam Jackson <ajax@redhat.com> 60238 Reviewed-by: Julien Cristau <jcristau@debian.org> 60239 60240commit 0f5a2f13dc284dd5f0f28e2c4323b2e08d253a7d 60241Author: Eric Anholt <eric@anholt.net> 60242Date: Fri Aug 23 11:04:22 2013 -0700 60243 60244 ephyr: Garbage collect some DOA host window clipping code. 60245 60246 Introduced in 79782726617d4b5a4f9b376f21936d035fc870e1 but never used. 60247 60248 Signed-off-by: Eric Anholt <eric@anholt.net> 60249 Reviewed-by: Julien Cristau <jcristau@debian.org> 60250 60251commit 090de17d361dc9828203a2562565d6499998db46 60252Author: Eric Anholt <eric@anholt.net> 60253Date: Fri Aug 23 10:56:22 2013 -0700 60254 60255 ephyr: Garbage collect some dead XV clipping code. 60256 60257 Signed-off-by: Eric Anholt <eric@anholt.net> 60258 Reviewed-by: Julien Cristau <jcristau@debian.org> 60259 60260commit 347d0fa6961c2b3dc4efd4e7cca8ef98cd3ea06e 60261Author: Eric Anholt <eric@anholt.net> 60262Date: Fri Aug 23 10:44:23 2013 -0700 60263 60264 ephyr: Remove the helper libs for each of the optional components. 60265 60266 Note that EXTRA_DIST was always unnecessary, because automake notices 60267 conditional compile of source files and includes them. 60268 60269 Copyright header is added because git noted that this was a 61% 60270 rewrite. 60271 60272 Signed-off-by: Eric Anholt <eric@anholt.net> 60273 Reviewed-by: Julien Cristau <jcristau@debian.org> 60274 60275commit 74dff87790dbb3dcb23634f1e8cd54018e5ddc9d 60276Author: Eric Anholt <eric@anholt.net> 60277Date: Fri Aug 23 10:40:52 2013 -0700 60278 60279 ephyr: Remove the remaining bits of host/server XV split. 60280 60281 Signed-off-by: Eric Anholt <eric@anholt.net> 60282 Reviewed-by: Julien Cristau <jcristau@debian.org> 60283 60284commit 2f994dfe9d1add9ea8cb9c88ddb0d15843854788 60285Author: Eric Anholt <eric@anholt.net> 60286Date: Fri Aug 23 10:14:36 2013 -0700 60287 60288 ephyr: Remove the host/server split of the actual XV video operations. 60289 60290 I suspect there's more cleanup possible in ephyrHostXVPutImage() by 60291 sticking a bunch of the args in the port priv earlier. 60292 60293 Signed-off-by: Eric Anholt <eric@anholt.net> 60294 Reviewed-by: Julien Cristau <jcristau@debian.org> 60295 60296commit 0b7ce74315213ccd31802847e5a83e67090ded14 60297Author: Eric Anholt <eric@anholt.net> 60298Date: Fri Aug 23 10:01:23 2013 -0700 60299 60300 ephyr: Remove another host/server split for XV image formats. 60301 60302 Signed-off-by: Eric Anholt <eric@anholt.net> 60303 Reviewed-by: Julien Cristau <jcristau@debian.org> 60304 60305commit 1dcc28bcb402d7286d18b41b469062a5c176d7eb 60306Author: Eric Anholt <eric@anholt.net> 60307Date: Thu Aug 22 17:00:41 2013 -0700 60308 60309 ephyr: Remove the host/server split for video encodings. 60310 60311 Signed-off-by: Eric Anholt <eric@anholt.net> 60312 Reviewed-by: Julien Cristau <jcristau@debian.org> 60313 60314commit 28a37757e11068b4594f4128a0d64c6bc9a53ee0 60315Author: Eric Anholt <eric@anholt.net> 60316Date: Thu Aug 22 17:26:21 2013 -0700 60317 60318 ephyr: Remove the host/server split for XV formats. 60319 60320 Signed-off-by: Eric Anholt <eric@anholt.net> 60321 Reviewed-by: Julien Cristau <jcristau@debian.org> 60322 60323commit 41134d393a37dd5b971b018a3888fbfea0b9c736 60324Author: Eric Anholt <eric@anholt.net> 60325Date: Thu Aug 22 17:21:37 2013 -0700 60326 60327 ephyr: Remove the host/server split for XV attributes. 60328 60329 v2: Fix leaks of xcb replies (caught by Julien) 60330 60331 Signed-off-by: Eric Anholt <eric@anholt.net> 60332 Reviewed-by: Julien Cristau <jcristau@debian.org> 60333 60334commit 1342a93c49beb6e4924534304119d57fa587ad0f 60335Author: Eric Anholt <eric@anholt.net> 60336Date: Thu Aug 22 17:26:08 2013 -0700 60337 60338 ephyr: Remove the host/server split for XV adaptors. 60339 60340 Signed-off-by: Eric Anholt <eric@anholt.net> 60341 Reviewed-by: Julien Cristau <jcristau@debian.org> 60342 60343commit fad79d2e3e53de4f25b3d1dd7accb558464a4a46 60344Author: Eric Anholt <eric@anholt.net> 60345Date: Sun Aug 18 20:08:32 2013 +0200 60346 60347 ephyr: Refactor XV adaptor feature detection. 60348 60349 This obviously wanted a helper function beforehand, but even more so 60350 now that we have XCB. 60351 60352 Signed-off-by: Eric Anholt <eric@anholt.net> 60353 Reviewed-by: Julien Cristau <jcristau@debian.org> 60354 60355commit 7a9c311add62026972419ef0b04a55680072bfeb 60356Author: Eric Anholt <eric@anholt.net> 60357Date: Sun Aug 18 20:02:53 2013 +0200 60358 60359 ephyr: Remove some pointless indirection in the XV code. 60360 60361 Now that we have XCB on the server side, we don't need to split these 60362 functions out. 60363 60364 Signed-off-by: Eric Anholt <eric@anholt.net> 60365 Reviewed-by: Julien Cristau <jcristau@debian.org> 60366 60367commit 46cf6bf5692ef751ec9d17ae2292565d4b13f14b 60368Author: Eric Anholt <eric@anholt.net> 60369Date: Sun Aug 18 18:38:29 2013 +0200 60370 60371 ephyr: Move event processing into ephyr.c. 60372 60373 No more extra event structure to translate between hostx.c and 60374 ephyr.c! 60375 60376 Signed-off-by: Eric Anholt <eric@anholt.net> 60377 Reviewed-by: Julien Cristau <jcristau@debian.org> 60378 60379commit 847c856effde482c13cdefcdca8176cee3bf67c4 60380Author: Eric Anholt <eric@anholt.net> 60381Date: Sun Aug 18 17:28:08 2013 +0200 60382 60383 ephyr: Move the host screen info into the kdrive screen private. 60384 60385 We can include xcb bits from the same place as server headers, so 60386 there's no need to hide them any more. 60387 60388 Signed-off-by: Eric Anholt <eric@anholt.net> 60389 Reviewed-by: Julien Cristau <jcristau@debian.org> 60390 60391commit 8dadc78e233f341f1ecb8d404de03d98ab210620 60392Author: Eric Anholt <eric@anholt.net> 60393Date: Sun Aug 18 16:45:06 2013 +0200 60394 60395 ephyr: Rename and use the proper type for what was host_screen->info. 60396 60397 Now that we can include server headers in talking to host X, we don't 60398 need to hide any more. 60399 60400 Signed-off-by: Eric Anholt <eric@anholt.net> 60401 Reviewed-by: Julien Cristau <jcristau@debian.org> 60402 60403commit 18d836f6dde46bcac8b794d34f76032688c28851 60404Author: Eric Anholt <eric@anholt.net> 60405Date: Sun Aug 18 14:12:45 2013 +0200 60406 60407 ephyr: Expose a single function for detecting extensions. 60408 60409 v2: Fix trying to include xcb-dri in the non-dri-build case (Noted by 60410 Julien) 60411 60412 Signed-off-by: Eric Anholt <eric@anholt.net> 60413 Reviewed-by: Julien Cristau <jcristau@debian.org> 60414 60415commit 3aacb22d6473480c31fb08f0b2527747dbf5be57 60416Author: Eric Anholt <eric@anholt.net> 60417Date: Sun Aug 18 14:01:53 2013 +0200 60418 60419 ephyr: Drop dead ephyrHostAtomToLocal code. 60420 60421 Signed-off-by: Eric Anholt <eric@anholt.net> 60422 Reviewed-by: Julien Cristau <jcristau@debian.org> 60423 60424commit 4aaad0bfbcef018eb36839b1d034c3f60022694a 60425Author: Julien Cristau <jcristau@debian.org> 60426Date: Wed Oct 20 20:10:43 2010 +0200 60427 60428 Xephyr: we're not using Xlib anymore, no need to undef _XSERVER64 60429 60430 Drop obsolete comments about interaction between Xlib code and that 60431 macro, and stop undefining it. 60432 60433 Signed-off-by: Julien Cristau <jcristau@debian.org> 60434 Reviewed-by: Eric Anholt <eric@anholt.net> 60435 60436commit 0fc2f9f04a44599f598380d4df992bec60c0e3a1 60437Author: Julien Cristau <jcristau@debian.org> 60438Date: Sat Sep 25 16:05:07 2010 +0200 60439 60440 Xephyr: handle errors in event loop 60441 60442 Signed-off-by: Julien Cristau <jcristau@debian.org> 60443 Reviewed-by: Eric Anholt <eric@anholt.net> 60444 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60445 60446commit 3309a73c826f02c4b2e14180a3254546314ec323 60447Author: Julien Cristau <jcristau@debian.org> 60448Date: Fri Jun 25 14:20:57 2010 +0100 60449 60450 Xephyr: drop remaining Xlib dependency 60451 60452 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60453 Reviewed-by: Eric Anholt <eric@anholt.net> 60454 Signed-off-by: Julien Cristau <jcristau@debian.org> 60455 60456commit 7da2174cf93f59b7e0372a6dc9d37a510076e187 60457Author: Julien Cristau <jcristau@debian.org> 60458Date: Mon Jun 21 14:39:03 2010 +0100 60459 60460 Xephyr: move glx code to xcb 60461 60462 v2: Rebase on master, notably adding XCB for 60463 X_GLXvop_MakeCurrentReadSGI (anholt). 60464 v3: Fix use-after-free (caught by Julien). 60465 60466 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> (v1) 60467 Signed-off-by: Julien Cristau <jcristau@debian.org> 60468 Signed-off-by: Eric Anholt <eric@anholt.net> 60469 60470commit 19c2699b954e2978301c4449dfccc7ca40bf6ae6 60471Author: Julien Cristau <jcristau@debian.org> 60472Date: Wed Jun 23 18:36:00 2010 +0100 60473 60474 Xephyr: remove unused DRI1 code 60475 60476 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60477 Signed-off-by: Julien Cristau <jcristau@debian.org> 60478 60479commit 68b7db19e8d3c1ac55a811327d3058f4c8fb45ec 60480Author: Julien Cristau <jcristau@debian.org> 60481Date: Sun Jun 20 00:07:08 2010 +0100 60482 60483 Xephyr: move ephyrdri over to xcb 60484 60485 Require new xcb-xf86dri to get fixed GetDrawableInfo. 60486 60487 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60488 Signed-off-by: Julien Cristau <jcristau@debian.org> 60489 60490commit aea119cd05807ed77336f96a40f8e4607c6ce169 60491Author: Julien Cristau <jcristau@debian.org> 60492Date: Sun Jun 20 11:13:26 2010 +0100 60493 60494 Xephyr: use xcb-xv instead of libXv 60495 60496 v2: Massive, conflict-filled rebase. I think I resolved everything 60497 right, but this host xv code makes my eyes bleed. Touch-tested XV 60498 after the rebase (anholt). 60499 60500 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> (v1) 60501 Signed-off-by: Julien Cristau <jcristau@debian.org> 60502 Signed-off-by: Eric Anholt <eric@anholt.net> 60503 60504commit 3ea75a59d5fec3c4099388724942e216f2344328 60505Author: Julien Cristau <jcristau@debian.org> 60506Date: Mon Jun 21 17:20:32 2010 +0100 60507 60508 Xephyr: use xcb for event handling 60509 60510 v2: Rebase on indentation changes, squash in a simpler variant of the 60511 later event compression patch, fix server hang or segfault on 60512 window close by reimplementing the x_io_error_handler in the new 60513 XCB event loop (anholt). 60514 60515 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> (v1) 60516 Signed-off-by: Julien Cristau <jcristau@debian.org> 60517 Signed-off-by: Eric Anholt <eric@anholt.net> 60518 60519commit ff26845252889f7cc3258a0d3aa116189acc6aae 60520Author: Julien Cristau <jcristau@debian.org> 60521Date: Thu Jun 24 23:45:16 2010 +0100 60522 60523 Xephyr: some more hostx.c xcb-ification 60524 60525 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60526 Reviewed-by: Eric Anholt <eric@anholt.net> 60527 Signed-off-by: Julien Cristau <jcristau@debian.org> 60528 60529commit 9b55f76a265aab0d9943a40208a38800e5782ba5 60530Author: Julien Cristau <jcristau@debian.org> 60531Date: Thu Jun 24 23:35:15 2010 +0100 60532 60533 Xephyr: move HostX.visual to xcb 60534 60535 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60536 Reviewed-by: Eric Anholt <eric@anholt.net> 60537 Signed-off-by: Julien Cristau <jcristau@debian.org> 60538 60539commit 126312fb4a6e06024a306341e675163108426390 60540Author: Julien Cristau <jcristau@debian.org> 60541Date: Thu Jun 24 23:26:48 2010 +0100 60542 60543 Xephyr: replace XKeycodeToKeysym with xcb-keysyms 60544 60545 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60546 Reviewed-by: Eric Anholt <eric@anholt.net> 60547 Signed-off-by: Julien Cristau <jcristau@debian.org> 60548 60549commit 8543d3fdfd47d2a0423f3ef8e7429c23c5d6b053 60550Author: Julien Cristau <jcristau@debian.org> 60551Date: Thu Jun 24 23:17:32 2010 +0100 60552 60553 Xephyr: delete unused hostx_get_extension_info function 60554 60555 v2: Also remove the prototype (anholt) 60556 60557 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> (v1) 60558 Signed-off-by: Julien Cristau <jcristau@debian.org> 60559 Signed-off-by: Eric Anholt <eric@anholt.net> 60560 60561commit d72c11b9df4a98fd80cb6635d34964483e33bc41 60562Author: Julien Cristau <jcristau@debian.org> 60563Date: Mon Jun 21 14:25:11 2010 +0100 60564 60565 Xephyr: delete unused proxy code 60566 60567 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60568 Reviewed-by: Eric Anholt <eric@anholt.net> 60569 Signed-off-by: Julien Cristau <jcristau@debian.org> 60570 60571commit a71ea3f96769da157dde82a591e43264820772ac 60572Author: Julien Cristau <jcristau@debian.org> 60573Date: Sun Jun 20 23:34:54 2010 +0100 60574 60575 Xephyr: xcb-ify visual list construction 60576 60577 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60578 Reviewed-by: Eric Anholt <eric@anholt.net> 60579 Signed-off-by: Julien Cristau <jcristau@debian.org> 60580 60581commit ac2b0348c00d692ec1bd8d6c8ab99e9198e755a7 60582Author: Julien Cristau <jcristau@debian.org> 60583Date: Sun Jun 20 23:34:39 2010 +0100 60584 60585 Xephyr: xcb-ify pointer/keyboard grab 60586 60587 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60588 Reviewed-by: Eric Anholt <eric@anholt.net> 60589 Signed-off-by: Julien Cristau <jcristau@debian.org> 60590 60591commit 1f93e6b350d1b9a058009052a666a2e8796708cb 60592Author: Julien Cristau <jcristau@debian.org> 60593Date: Sun Jun 20 23:34:03 2010 +0100 60594 60595 Xephyr: no need for XDisplayKeycodes 60596 60597 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60598 Reviewed-by: Eric Anholt <eric@anholt.net> 60599 Signed-off-by: Julien Cristau <jcristau@debian.org> 60600 60601commit 73398e5bf61c30e69a40a19a1057c2263fbb91ef 60602Author: Julien Cristau <jcristau@debian.org> 60603Date: Thu Jun 24 23:00:38 2010 +0100 60604 60605 Xephyr: use xcb-shape instead of XShape* 60606 60607 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 60608 Reviewed-by: Eric Anholt <eric@anholt.net> 60609 Signed-off-by: Julien Cristau <jcristau@debian.org> 60610 60611commit a2b73da78de4e627965213d24a6c33f243a60eb6 60612Author: Julien Cristau <jcristau@debian.org> 60613Date: Sun Jun 20 00:05:40 2010 +0100 60614 60615 Xephyr: start converting hostx.c over to xcb 60616 60617 v2: Dropped the hostx_load_keymap changes, now that that function is 60618 gutted (anholt). 60619 60620 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> (v1) 60621 Signed-off-by: Julien Cristau <jcristau@debian.org> 60622 Signed-off-by: Eric Anholt <eric@anholt.net> 60623 60624commit 762606b4cdfb57c5202ae128ce4dbb8afe172d4e 60625Author: Julien Cristau <jcristau@debian.org> 60626Date: Fri Jun 25 17:59:13 2010 +0100 60627 60628 Xephyr: stop loading the host's keymap 60629 60630 This isn't used anywhere. 60631 60632 v2: Rebase to the top of the patch series (anholt) 60633 60634 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> (v1) 60635 Signed-off-by: Julien Cristau <jcristau@debian.org> 60636 Signed-off-by: Eric Anholt <eric@anholt.net> 60637 60638commit 3752a815800626ae93f64a856f5889aa6636344e 60639Author: Marc Haesen <marha@users.sourceforge.net> 60640Date: Sat Aug 10 17:38:38 2013 +0100 60641 60642 hw/xwin: Remove unnecessary and incorrect HWND casts 60643 60644 Signed-off-by: Marc Haesen <marha@users.sourceforge.net> 60645 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60646 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60647 60648commit 34932cef6cc800fbdf2f375506057ed8cd1e07f6 60649Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60650Date: Mon Apr 1 13:43:00 2013 +0100 60651 60652 hw/xwin: Fix compilation of winauth.c with -Werror=implicit-function-declaration 60653 60654 /jhbuild/checkout/xorg/xserver/hw/xwin/winauth.c: In function ‘MitGenerateCookie’: 60655 /jhbuild/checkout/xorg/xserver/hw/xwin/winauth.c:87:5: error: implicit declaration of function ‘MitAddCookie’ [-Werror=implicit-function-declaration] 60656 60657 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60658 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60659 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60660 60661commit 2a08554a0bf80b862c15676c31ecd463b616518c 60662Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60663Date: Mon Jun 17 23:54:01 2013 +0100 60664 60665 hw/xwin: Remove unused extern g_fUseUnicode from winclipboardwndproc.c 60666 60667 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk 60668 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60669 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60670 60671commit 69401f38a697c941ea58351028f4cd4b810f9188 60672Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60673Date: Mon Jun 17 18:37:26 2013 +0100 60674 60675 hw/xwin: Remove unneeded forward declaration of winProcessXEventsTimeout() 60676 60677 Remove unneeded forward declaration of winProcessXEventsTimeout(), the actual 60678 definition immediately follows. 60679 60680 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60681 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60682 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60683 60684commit 9c45a1ad44f12ac25e1c3b7a18a3bceb85a2cf38 60685Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60686Date: Mon Jun 17 16:58:47 2013 +0100 60687 60688 hw/xwin: Remove prototype for non-existent winDeinitClipboard() 60689 60690 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60691 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60692 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60693 60694commit acc46fc217352b94f77a6b5b14808ea89875a2a1 60695Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60696Date: Mon Jun 17 00:40:13 2013 +0100 60697 60698 hw/xwin: Remove unused externs from winclipboardwrappers.c 60699 60700 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60701 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60702 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60703 60704commit fcc7767f8cd6c318861b1060ef6a3936750a0dd2 60705Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60706Date: Mon Jun 10 17:31:12 2013 +0100 60707 60708 hw/xwin: Remove unused winProcQueryTreeOrig variable 60709 60710 Unused since 47c7b6d3e626497747ae2780f259a15b8e6c846f "Remove no-longer needed 60711 tricks used to prevent the clipboard client from being killed" 60712 60713 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60714 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60715 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60716 60717commit 8fa882a980abea6f05b4b59447a187aea2c793a1 60718Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60719Date: Sat Aug 10 12:41:07 2013 +0100 60720 60721 hw/xwin: Fix ARGB cursor conversion on x86_64 60722 60723 Fix erroneous use of unsigned long * for lpBits in winXCursorToHCURSOR() which 60724 leads to ARGB cursors being vertically streched on x86_64 by interleaving blank 60725 rows of pixels. 60726 60727 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60728 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60729 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60730 60731commit 4bfb2dce5eea4923eaf86eca33b96087b28235f5 60732Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60733Date: Sun Apr 7 23:59:44 2013 +0100 60734 60735 hw/xwin: Fix an issue in winSetSpansNativeGDI() identifed by -Warray-bounds 60736 60737 The BITMAPINFO local only has room for a single RBGQUAD in bmiColors, but we 60738 access two (black and white for a mono-color DIB). Fix by changing to a dynamic 60739 allocation big enough for a BITMAPINFO and a RGBQUAD. 60740 60741 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60742 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60743 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60744 60745commit 784c006adb22c6693ad30dd0b1b05369cfaa7b5e 60746Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60747Date: Sat Aug 10 11:55:23 2013 +0100 60748 60749 hw/xwin: Fix -Warray-bounds warning in winXCursorToHCURSOR() 60750 60751 Rewrite winXCursorToHCURSOR() so access to BITMAPINFO bmiColors member doesn't 60752 trigger an -Warray-bounds warning. 60753 60754 Note that the underlying storage is allocated as a BITMAPV4HEADER, so has 60755 sufficent room for the extra RGBQUADs bmiColors after the BITMAPINFO bmiHeader. 60756 60757 wincursor.c: In function 'winSetCursor': 60758 wincursor.c:293:24: error: array subscript is above array bounds [-Werror=array-bounds] 60759 wincursor.c:294:24: error: array subscript is above array bounds [-Werror=array-bounds] 60760 wincursor.c:295:24: error: array subscript is above array bounds [-Werror=array-bounds] 60761 wincursor.c:296:24: error: array subscript is above array bounds [-Werror=array-bounds] 60762 wincursor.c:297:24: error: array subscript is above array bounds [-Werror=array-bounds] 60763 wincursor.c:298:24: error: array subscript is above array bounds [-Werror=array-bounds] 60764 wincursor.c:299:24: error: array subscript is above array bounds [-Werror=array-bounds] 60765 wincursor.c:300:24: error: array subscript is above array bounds [-Werror=array-bounds] 60766 60767 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60768 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60769 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60770 60771commit 41d03c9ccce665b919b180bbb819b78c750aa325 60772Author: Marc Haesen <marha@users.sourceforge.net> 60773Date: Fri Aug 2 18:19:22 2013 +0100 60774 60775 hw/xwin: Correct size of _WINDOWSWM_NATIVE_HWND property on x86_64 60776 60777 Use the correct size of a HWND on x86_64 in XChangeProperty() and 60778 XGetWindowProperty() calls for the _WINDOWSWM_NATIVE_HWND property. 60779 60780 Signed-off-by: Marc Haesen <marha@users.sourceforge.net> 60781 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60782 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60783 60784commit 390cfec10ef7d52391f890c7593ac82a7d77d856 60785Author: Colin Harrison <colin.harrison@virgin.net> 60786Date: Fri Aug 9 16:48:12 2013 +0100 60787 60788 hw/xwin: Correct winprefs.c function signatures for x64. 60789 60790 Correct SetupRootMenu(), SetupSysMenu(), HandleCustomWM_INITMENU() and 60791 HandleCustomWM_COMMAND() function signatures which use unsigned long parameters, 60792 where just a specific HWND or HMENU handle type should have been used. 60793 60794 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 60795 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60796 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60797 60798commit e716baedc4d8e52a60f43ef21aba771b340d8c8b 60799Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60800Date: Sat Mar 31 14:42:45 2012 +0100 60801 60802 Move pseudoramiX code where it can be shared between Xwin and Xquartz 60803 60804 Move pseudoramiX code to a separate top-level directory. Link Xwin and Xquartz 60805 with libPseudoramiX 60806 60807 I'm not sure moving this to a top-level directory is appropriate, but I'm not 60808 sure where else it fits. 60809 60810 Future work: pseudoramiX can probably be consolidated with the rrxinerama code 60811 (which I think provides fake xinerama data when real XINERAMA is disabled and we 60812 only have one screen) 60813 60814 v2: fix distcheck 60815 60816 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60817 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60818 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60819 60820commit ce0217b98261dcf8b20cfb743831965dda1950ec 60821Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60822Date: Sat Aug 10 12:35:55 2013 +0100 60823 60824 hw/xwin: Fix winglobals.h for MinGW64 build 60825 60826 Include pthread.h in winglobals.h to fix build with latest MinGW64 headers 60827 60828 In file included from 60829 winmultiwindowicons.c:47:0: 60830 winglobals.h:92:1: error: unknown type name ‘pthread_mutex_t’ 60831 60832 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60833 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60834 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60835 60836commit b27756dbe6a1be4633ec47a8fe92a4c217f68f93 60837Author: Jon TURNEY <jon.turney@dronecode.org.uk> 60838Date: Sun Aug 11 12:15:11 2013 +0100 60839 60840 hw/xwin: Fix inclusion of shlobj.h in MinGW64 build 60841 60842 Fix build with latest MinGW64 headers by wrapping Status type in shlobj.h as well 60843 60844 In file included from InitOutput.c:51:0: 60845 /usr/i686-w64-mingw32/sys-root/mingw/include/shlobj.h:1231:44: error: expected identifier or ‘(’ before ‘int’ 60846 /usr/i686-w64-mingw32/sys-root/mingw/include/shlobj.h:1248:44: error: expected identifier or ‘(’ before ‘int’ 60847 60848 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 60849 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 60850 Reviewed-by: Marc Haesen <marha@users.sourceforge.net> 60851 60852commit 94d4e29aedc69431fa9b299ca1b67947173d7a24 60853Author: Peter Hutterer <peter.hutterer@who-t.net> 60854Date: Mon Aug 26 15:51:46 2013 +1000 60855 60856 Xi: allow for XIAllowEvent requests larger than XI < 2.2 size (#68554) 60857 60858 XIAllowEvents changed length in XI 2.2 (for the touchid). A bug in libXi 60859 causes libXi to always use the new request length if the server supports 60860 2.2, regardless of the client's XIQueryVersion request. 60861 60862 The server takes the client's XIQueryVersion request into account though, 60863 resulting in a BadLength error if a 2.[0,1] client calls XIAllowEvents on a 60864 XI 2.2+ server. 60865 60866 Can't fix this in libXi, so work around this in the server. 60867 60868 X.Org Bug 68554 <http://bugs.freedesktop.org/show_bug.cgi?id=68554> 60869 60870 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 60871 Reviewed-by: Adam Jackson <ajax@redhat.com> 60872 60873commit 82939e02392cbb880313fe92957091ff89ce2f2b 60874Author: Peter Hutterer <peter.hutterer@who-t.net> 60875Date: Fri Aug 16 15:30:59 2013 +1000 60876 60877 dix: check for grab type before checking XI2 mask 60878 60879 if the grab type isn't XI2, grab->xi2mask is random. That random data may 60880 have the enter/leave mask set, leading to events sent to the client that the 60881 client can't handler. 60882 60883 Source of these errors: 60884 _xgeWireToEvent: Unknown extension 131, this should never happen. 60885 60886 Simplest reproducer: 60887 Start Xephyr, press button inside window, move out. As the pointer leaves 60888 the Xephyr window, the errors appear. 60889 60890 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 60891 Reviewed-by: Adam Jackson <ajax@redhat.com> 60892 60893commit 16894df1ffb6fe21bf723dd954578e8d934799ca 60894Author: Peter Hutterer <peter.hutterer@who-t.net> 60895Date: Fri Aug 16 10:55:09 2013 +1000 60896 60897 Xi: replace loop with memset 60898 60899 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 60900 Reviewed-by: Adam Jackson <ajax@redhat.com> 60901 60902commit 8f2292b683f591052da2fe94fb3831a3c9a5aa61 60903Author: Peter Hutterer <peter.hutterer@who-t.net> 60904Date: Fri Aug 16 11:05:49 2013 +1000 60905 60906 include: wrap EMASKSIZE in parentheses 60907 60908 Otherwise things like EMASKSIZE * foo will yield interesting results. 60909 60910 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 60911 Reviewed-by: Adam Jackson <ajax@redhat.com> 60912 60913commit d803f296c616ccf55711cf0af9e7d3eae2edda16 60914Author: Peter Hutterer <peter.hutterer@who-t.net> 60915Date: Fri Aug 23 09:04:01 2013 +1000 60916 60917 test: fix the gcc diagnostics pragma 60918 60919 pop without push restores the commandline options. The proper way is to 60920 push, then ignore, then pop. 60921 60922 And while we're at it, change the pop argument to a comment - pop ignores 60923 the argument, but be proper about it. 60924 60925 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 60926 60927commit 05ea6307fa4f0bee3d3bf536396059f3906791a0 60928Author: Egbert Eich <eich@freedesktop.org> 60929Date: Fri Aug 16 19:52:16 2013 +0200 60930 60931 DIX/Xi: Pass correct client to CheckDeviceGrabAndHintWindow() 60932 60933 If we have a client which has registered for a DeviceButton grab 60934 be sure to pass this to CheckDeviceGrabAndHintWindow(). Since the 60935 order of clients is arbitrary there is no guarantee that the last 60936 client in the list is the one that belongs to this class. 60937 60938 Signed-off-by: Egbert Eich <eich@freedesktop.org> 60939 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 60940 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 60941 60942commit e01a3f65d3e6d92f92ef2992b338cc9625bde082 60943Author: Eric Anholt <eric@anholt.net> 60944Date: Sat Aug 17 14:50:37 2013 +0200 60945 60946 ephyr: Fix warning about XID vs unsigned long * by changing function args 60947 60948 There's no reason to pass the data back out to the caller, since the 60949 caller was dropping it on the floor. The original data is a CARD32, 60950 so no need to mess with weird unsigned longs. 60951 60952 Signed-off-by: Eric Anholt <eric@anholt.net> 60953 Reviewed-by: Julien Cristau <jcristau@debian.org> 60954 60955commit f1cefe020bd21c07dfcc9f204e79939499abbd35 60956Author: Eric Anholt <eric@anholt.net> 60957Date: Sat Aug 17 14:38:08 2013 +0200 60958 60959 ephyr: Drop is_ok logging. 60960 60961 The only cases that lead to !is_ok are already EPHYR_LOG_ERROR, and it 60962 fixes set-but-unused compiler warnings. 60963 60964 Signed-off-by: Eric Anholt <eric@anholt.net> 60965 Reviewed-by: Julien Cristau <jcristau@debian.org> 60966 60967commit 77c7a2014e3942b55c0b78308edf02eab33f2e60 60968Author: Eric Anholt <eric@anholt.net> 60969Date: Sat Aug 17 14:36:09 2013 +0200 60970 60971 kdrive: Fix const cast warnings in arguments processing. 60972 60973 Signed-off-by: Eric Anholt <eric@anholt.net> 60974 Reviewed-by: Julien Cristau <jcristau@debian.org> 60975 60976commit 7107937e473a3557c37d719cb100b2e9b09075dd 60977Author: Eric Anholt <eric@anholt.net> 60978Date: Sat Aug 17 13:11:17 2013 +0200 60979 60980 kdrive: Fix const cast warnings in driver name handling. 60981 60982 Signed-off-by: Eric Anholt <eric@anholt.net> 60983 Reviewed-by: Julien Cristau <jcristau@debian.org> 60984 60985commit 56c405d46a86853771a0b808da4c8a9ef626960c 60986Author: Eric Anholt <eric@anholt.net> 60987Date: Sat Aug 17 13:07:23 2013 +0200 60988 60989 kdrive: Fix a few easy cast warnings. 60990 60991 Signed-off-by: Eric Anholt <eric@anholt.net> 60992 Reviewed-by: Julien Cristau <jcristau@debian.org> 60993 60994commit a9c48a07eea50468fa47ab178cdd91edb0b5727f 60995Author: Eric Anholt <eric@anholt.net> 60996Date: Sat Aug 17 13:06:23 2013 +0200 60997 60998 kdrive: Drop dead code. 60999 61000 This is unused as of the introduction of a helper in 61001 e7150db5350bc2113ff4126019b489847a4dc217 61002 61003 Signed-off-by: Eric Anholt <eric@anholt.net> 61004 Reviewed-by: Julien Cristau <jcristau@debian.org> 61005 61006commit 873ae962bdb7559c952b5955ba262285701982a8 61007Author: Eric Anholt <eric@anholt.net> 61008Date: Sat Aug 17 13:01:57 2013 +0200 61009 61010 ephyr: Fix dead code warning. 61011 61012 Signed-off-by: Eric Anholt <eric@anholt.net> 61013 Reviewed-by: Julien Cristau <jcristau@debian.org> 61014 61015commit 34af2eebe19e2f8afe35a5614d4d44fa4fd89d6b 61016Author: Eric Anholt <eric@anholt.net> 61017Date: Sat Aug 17 13:01:32 2013 +0200 61018 61019 ephyr: Fix const-cast warnings for setting window title. 61020 61021 Signed-off-by: Eric Anholt <eric@anholt.net> 61022 Reviewed-by: Julien Cristau <jcristau@debian.org> 61023 61024commit 7edf5066db48434739ac64e0576fbb4fdf0f37f9 61025Author: Eric Anholt <eric@anholt.net> 61026Date: Sat Aug 17 12:54:33 2013 +0200 61027 61028 ephyr: Fix const cast warnings in ephyrhostvideo. 61029 61030 Signed-off-by: Eric Anholt <eric@anholt.net> 61031 Reviewed-by: Julien Cristau <jcristau@debian.org> 61032 61033commit fe7463b8ce0de301c2f82b108c93963424f77219 61034Author: Julien Cristau <jcristau@debian.org> 61035Date: Sat Jul 27 12:09:07 2013 +0200 61036 61037 dix: add missing include for DeleteWindowFromAnySelections 61038 61039 Fixes build error with XACE disabled: 61040 window.c:886:5: error: implicit declaration of function 'DeleteWindowFromAnySelections' [-Werror=implicit-function-declaration] 61041 DeleteWindowFromAnySelections(pWin); 61042 ^ 61043 61044 Debian bug#701372 61045 61046 Reported-by: Matthias Klose <doko@debian.org> 61047 Signed-off-by: Julien Cristau <jcristau@debian.org> 61048 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61049 Signed-off-by: Keith Packard <keithp@keithp.com> 61050 61051commit 6a49f8836a649322d38a364d9ae084be40fb2305 61052Author: Keith Packard <keithp@keithp.com> 61053Date: Sat Aug 17 12:17:36 2013 +0200 61054 61055 Add .dir-locals.el 61056 61057 Enforce X.org style with an emacs configuration file 61058 61059 Signed-off-by: Keith Packard <keithp@keithp.com> 61060 61061commit f67d022523c59a27f3bf8791aa9ca6624318b1fd 61062Merge: 5a36cdd91 9680f6a12 61063Author: Keith Packard <keithp@keithp.com> 61064Date: Tue Aug 6 07:12:12 2013 +0200 61065 61066 Merge commit '9680f6a12d642c2e4bfa09bf459d678214059d74' 61067 61068commit 5a36cdd91530d27627e39159a89b53f9fbb75280 61069Author: Alan Coopersmith <alan.coopersmith@oracle.com> 61070Date: Mon Aug 5 20:46:45 2013 -0700 61071 61072 Allow disabling XFree86-DGA, DRI, VidModeExtension extensions 61073 61074 Code to recognize these in extension enable/disable options was wrapped 61075 in #ifdef XorgLoader, but that's not defined when building miinitext.c 61076 since the great module merge of 1.13. Change to an #ifdef that is defined. 61077 61078 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61079 Reviewed-by: Robert Hooker <robert.hooker@canonical.com> 61080 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 61081 Signed-off-by: Keith Packard <keithp@keithp.com> 61082 61083commit 9680f6a12d642c2e4bfa09bf459d678214059d74 61084Author: Peter Hutterer <peter.hutterer@who-t.net> 61085Date: Tue Aug 6 13:14:56 2013 +1000 61086 61087 xfree86: de-duplicate some AM_CPPFLAGS 61088 61089 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61090 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61091 61092commit bbef8e46f2fed8f823ee4b97e96c3fcee28f3b75 61093Author: Peter Hutterer <peter.hutterer@who-t.net> 61094Date: Tue Aug 6 13:08:13 2013 +1000 61095 61096 Replace INCLUDES with AM_CPPFLAGS 61097 61098 newer automake gets quite noisy about this. 61099 hw/xfree86/ddc/Makefile.am:7: warning: 61100 'INCLUDES' is the old name for 'AM_CPPFLAGS' (or '*_CPPFLAGS') 61101 and many more of these. 61102 61103 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61104 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61105 61106commit c218ba8423a73a7a643cb17789db8a1dd0901ca6 61107Author: Julien Cristau <jcristau@debian.org> 61108Date: Fri Aug 2 20:07:36 2013 +0200 61109 61110 xfree86: improve check for posix saved ids 61111 61112 Replace hardcoded SVR4 || linux || CSRG_BASED with an autoconf check and 61113 the _POSIX_SAVED_IDS macro. 61114 61115 Suggested-by: Mark Kettenis <mark.kettenis@xs4all.nl> 61116 Signed-off-by: Julien Cristau <jcristau@debian.org> 61117 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61118 Reviewed-by: Mark Kettenis <kettenis@openbsd.org>. 61119 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61120 61121commit 930c6ff15d437dfb0f897e8cb4253abba70819cc 61122Author: Julien Cristau <jcristau@debian.org> 61123Date: Fri Aug 2 23:46:00 2013 +0200 61124 61125 test: include dix-config.h in hashtabletest.c 61126 61127 Missing _XSERVER64 define caused inconsistent sizeof(XID) between the 61128 test and hashtable code, leading to test failures on 64bit big endian 61129 archs like s390x or ppc64. 61130 61131 Signed-off-by: Julien Cristau <jcristau@debian.org> 61132 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 61133 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61134 61135commit b6e5c4669e0db391966deb397e8c975ec7f0124d 61136Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61137Date: Tue Jul 30 15:31:24 2013 +0200 61138 61139 test/xi2: fix protocol-xiqueryversion test 61140 61141 The old code was broken and allowed setting client version >= XIVersion, 61142 this was fixed in the previous patch, but updating the value for XIVersion 61143 broke the tests, so fix the tests too. 61144 61145 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61146 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 61147 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61148 61149commit 500e844a24962c9e70abb3d614f1973013b2de73 61150Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61151Date: Tue Jul 30 14:45:21 2013 +0200 61152 61153 Xi: Clamp XIClient maximal version to XIVersion 61154 61155 Do not allow setting client version to an arbitrary value >= XIVersion. 61156 Fixes a test error with test/xi2/protocol-xiqueryversion.c, introduced by 61157 commit 4360514d1c "Xi: Allow clients to ask for 2.3 and then 2.2 without failing" 61158 61159 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61160 Reviewed-by: Keith Packard <keithp@keithp.com> 61161 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61162 61163commit e1ab8f239bdc6af17ef6cd0e9c6187c4e2e2f5bc 61164Author: Dave Airlie <airlied@redhat.com> 61165Date: Wed Jul 31 11:35:44 2013 +1000 61166 61167 xfree86/man: document AutoAddGPU 61168 61169 This at least mentions AutoAddGPU and hints at when you might 61170 want to disable it. 61171 61172 Signed-off-by: Dave Airlie <airlied@redhat.com> 61173 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 61174 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61175 61176commit 836daf4c7a276108a9f950e524828744aeda0a3c 61177Author: Dave Airlie <airlied@redhat.com> 61178Date: Wed Jul 31 11:24:48 2013 +1000 61179 61180 modesetting: change output names for secondary GPUs 61181 61182 if we are a secondary GPU modify the output name to avoid clashes. 61183 61184 Signed-off-by: Dave Airlie <airlied@redhat.com> 61185 61186commit ff38bbe81ace85bf675bbaa0a9ca5f3b32ede449 61187Merge: bdd1e22cb 4360514d1 61188Author: Keith Packard <keithp@keithp.com> 61189Date: Thu Jul 25 22:50:27 2013 -0700 61190 61191 Merge remote-tracking branch 'whot/for-keith' 61192 61193commit 4360514d1cc8e3132f93f56172d291074e8c770f 61194Author: Keith Packard <keithp@keithp.com> 61195Date: Wed Jul 10 22:42:55 2013 -0700 61196 61197 Xi: Allow clients to ask for 2.3 and then 2.2 without failing 61198 61199 This allows different sub-systems within the same application to 61200 request different Xi versions without either getting old behaviour 61201 everywhere or simply failing with a BadValue. 61202 61203 Signed-off-by: Keith Packard <keithp@keithp.com> 61204 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61205 61206commit bdd1e22cbde9ea2324e4e1991c9e152f22f88151 61207Author: Aaron Plattner <aplattner@nvidia.com> 61208Date: Tue Apr 30 14:14:23 2013 -0700 61209 61210 xfree86: detach scanout pixmaps when detaching output GPUs 61211 61212 Commit 8f4640bdb9d3988148e09a08d2c7e3bab1d538d6 fixed a bit of a 61213 chicken-and-egg problem by detaching GPU screens when their providers 61214 are destroyed, which happens before CloseScreen is called. However, 61215 this created a new problem: the GPU screen tears down its RandR crtc 61216 objects during CloseScreen and if one of them is active, it tries to 61217 detach the scanout pixmap then. This crashes because 61218 RRCrtcDetachScanoutPixmap tries to get the master screen's screen 61219 pixmap, but crtc->pScreen->current_master is already NULL at that 61220 point. 61221 61222 It doesn't make sense for an unbound GPU screen to still be scanning 61223 out its former master screen's pixmap, so detach them first when the 61224 provider is destroyed. 61225 61226 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 61227 Reviewed-by: Dave Airlie <airlied@redhat.com> 61228 Signed-off-by: Keith Packard <keithp@keithp.com> 61229 61230commit 3d8756631070c440cefa31b35fea3d407f187810 61231Author: Peter Hutterer <peter.hutterer@who-t.net> 61232Date: Mon Jul 22 13:43:07 2013 +1000 61233 61234 dix: set the valuator mask to ensure XI 1.x events have data 61235 61236 XI 1.x only allows for first + num valuators, so if a device sends data for 61237 valuators 0 and 2+ only (i.e. valuator 1 is missing) we still need to get 61238 the data for that from somewhere. 61239 XI 1.x uses the hack of an unset valuator mask to get the right coordinates, 61240 i.e. we set the value but don't set the mask for it so XI2 events have the 61241 right mask. 61242 61243 For an absolute device in relative mode, this broke in b28a1af55cf, the 61244 value was now always 0. This wasn't visible on the cursor, only in an XI 1.x 61245 client. The GIMP e.g. sees jumps to x/0 every few events. 61246 61247 Drop the condition introduced in b28a1af55cf, data in valuators is always 61248 absolute, regardless of the mode. 61249 61250 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61251 61252commit 4fb686d6a6777950f0e0d55b848cd2af4cbad372 61253Author: Peter Hutterer <peter.hutterer@who-t.net> 61254Date: Wed Jul 24 11:50:00 2013 +1000 61255 61256 dix: check the xi2mask, not the grab type for touch listeners 61257 61258 grab->type is only non-zero for passive grabs. We're checking an active grab 61259 here, so we need to check if the touch mask is set on the grab. 61260 61261 Test case: grab the device, then start two simultaneous touches. The 61262 grabbing client won't see the second touchpoints because grab->type is 0 61263 and the second touch is not an emulating pointer. 61264 61265 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61266 61267commit 43ac0491e36cdbb716b5c9d39c97d0aba3bebd75 61268Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61269Date: Mon Apr 8 01:28:36 2013 +0100 61270 61271 hw/xwin: Fix numerous 64-bit format/type cast issues with debug printing of pointers 61272 61273 Numerous pieces of debug output cast a pointer to an int and then use a "%08x" 61274 printf format. 61275 61276 Use "%p" format for 64-bit portability. 61277 61278 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61279 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61280 61281commit e95bb97073ca77193e4c51bd1504403fbe245533 61282Author: Colin Harrison <colin.harrison@virgin.net> 61283Date: Tue Jun 25 21:34:43 2013 +0100 61284 61285 hw/xwin: Fix possible crash in winMultiWindowGetClassHint 61286 61287 Fix a possible crash in winMultiWindowGetClassHint() when an application doesn't 61288 null terminate the WM_CLASS property class name (which is an ICCCM conformance 61289 bug in the application) 61290 61291 (Reported for running the contiki cooja simulator in multiwindow mode, although 61292 it seems that many Java clients may have this problem, see [1]) 61293 61294 Based on a patch by Marc Haesen. 61295 61296 v2: Avoid using strnlen() which is missing on MinGW 61297 v3: Align with Xming patch 61298 61299 [1] http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6961123 61300 61301 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 61302 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 61303 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61304 61305commit 4bc375aa2f08bd092f60d3e358d345fc9a0d3637 61306Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61307Date: Tue Jun 18 19:21:02 2013 +0100 61308 61309 hw/xwin: Remove obsolete WIN_XEVENTS_SHUTDOWN 61310 61311 Remove obsolete WIN_XEVENTS_SHUTDOWN. This event is never generated. 61312 61313 (I think the idea was to listen for WM_DELETE_WINDOW, but that's not a very 61314 useful thing to do for a hidden window.) 61315 61316 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61317 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61318 61319commit 2663203cd6994166e107e796c4724a4ed3b59370 61320Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61321Date: Tue Jun 18 18:03:05 2013 +0100 61322 61323 configure.ac: Check for python at configure time when building XWin with AIGLX 61324 61325 Check for python at configure time when building XWin with AIGLX, it's used to 61326 generate the wrapper code for native GL functions. 61327 61328 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61329 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61330 61331commit 9fe360bf2be5c959fb21835955ef550098ccbbf0 61332Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61333Date: Sun Jun 16 00:07:57 2013 +0100 61334 61335 hw/xwin: Avoid a null dereference if CreateDIBSection() fails in NetWMToWinIconAlpha() 61336 61337 Avoid a null dereference of DIB_pixels if CreateDIBSection() fails in 61338 NetWMToWinIconAlpha() 61339 61340 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61341 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61342 61343commit 71b5f56302bbd8be62f63f0dd62cbcd33aab3ac5 61344Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61345Date: Thu Feb 21 17:12:17 2013 +0000 61346 61347 hw/xwin: Handle WM_MOUSEHWHEEL 61348 61349 Handle WM_MOUSEHWHEEL tilt wheel messages, similarly to WM_MOUSEWHEEL scroll 61350 wheel messages, to generate X button 6 and 7 presses and releases. 61351 61352 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61353 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61354 61355commit cf9c777ee094d660e0c95559373fd23ee910362e 61356Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61357Date: Thu Feb 21 17:10:45 2013 +0000 61358 61359 hw/xwin: Map extra mouse buttons 1 and 2 to X buttons 8 and 9 61360 61361 Map extra mouse buttons 1 and 2 to X buttons 8 and 9, as conventional, leaving X 61362 buttons 6 and 7 for tilt wheel. 61363 61364 Also add button labels for buttons 6, 7, 8 and 9 and change btn_labels in from a 61365 dynamic allocation to a fixed one of the required size for all the labels we 61366 use. 61367 61368 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61369 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61370 61371commit c9d7b9516f4a04cb9012c6c1e9466491a468aa07 61372Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61373Date: Tue Feb 19 14:38:40 2013 +0000 61374 61375 hw/xwin: Remove unneeded WM_XBUTTON message defines 61376 61377 Remove unneeded WM_XBUTTON message defines, they have been provided by w32api 61378 for a long time now. 61379 61380 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61381 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61382 61383commit 5dccfc63f9c7056e7cd30884e1d3ccea86a2f419 61384Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61385Date: Mon Dec 17 22:38:00 2012 +0000 61386 61387 hw/xwin: Change winTranslateKey() to return it's result as it's return value 61388 61389 Change winTranslateKey() to return it's result as it's return value, and change 61390 it's uses as well. 61391 61392 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61393 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61394 61395commit 9b4cec76f1d7792d4bf64fa069ea6b64bc42da0d 61396Author: Oliver Schmidt <oschmidt-mailinglists@gmx.de> 61397Date: Mon Nov 5 15:05:32 2012 +0000 61398 61399 hw/xwin: Consider left and right modifier keys independently on gaining focus 61400 61401 Handle left and right ctrl and shift keys independently 61402 61403 Assume that all modifiers are cleared when all keys are released on focus lost, 61404 as internalKeyState doesn't record which modifier key was pressed. 61405 61406 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61407 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61408 61409commit 235149d0b4af9097e47b3af0ba56d4eb0179518c 61410Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61411Date: Mon Apr 5 14:22:39 2010 +0100 61412 61413 hw/xwin: Remove an extra '\n' from some log messages 61414 61415 Remove an extra '\n' from internal client IOError log messages 61416 61417 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61418 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61419 61420commit c7aa9f7578e721b0c9e565e7a085b6aeb30bf2ac 61421Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61422Date: Sat Feb 4 17:04:11 2012 +0000 61423 61424 hw/xwin: _NET_WM_STATE is ATOM[] not ATOM 61425 61426 _NET_WM_STATE is ATOM[] not ATOM, a list of window state hints, so check all of 61427 the atoms, not just the first one 61428 61429 See EWMH specifcation, section "Application Window Properties" 61430 61431 v2: Actually use [] on the returned atom data 61432 61433 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61434 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61435 61436commit 682ccac90b18f293520f77db5b163fcb40328e2b 61437Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61438Date: Wed Aug 31 21:35:14 2011 +0100 61439 61440 hw/xwin: Improve WM_ENDSESSION handling using separate messaging window thread 61441 61442 Currently, WM_ENDSESSION just calls GiveUp() to set the DE_TERMINATE flag. But 61443 for the X server to exit cleanly, we also need the X server dispatch loop to be 61444 unblocked so it can notice that DE_TERMINATE has been set and exit, removing 61445 it's lock file and any unix domain socket. 61446 61447 It appears that the system will terminate the process when the last UI thread in 61448 that process returns from processing WM_ENDSESSION for the last top-level 61449 window. 61450 61451 Since WM_ENDSESSION appears to sent by the system via SendMessage() 61452 (synchronously) and the wndproc is called to process it in the message thread 61453 for that window (the X server thread), we can't easily terminate the X server 61454 dispatch loop from inside the WM_ENDSESSION message processing. 61455 61456 So, create a messaging window, a hidden, top-level window, with a separate 61457 thread to catch this message, and process it by calling GiveUp() and then 61458 blocking on a mutex until the X server dispatch loop exits. 61459 61460 Also, notice when this is a shutdown cancel WM_ENDSESSION message and take no 61461 action. 61462 61463 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61464 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61465 61466commit 15febb05d77cc7e7185c942f35459838f75cfdc0 61467Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61468Date: Mon Sep 3 15:19:43 2012 +0100 61469 61470 Allow DDX to provide a main() 61471 61472 XQuartz already conditionally renames main() as dix_main() so it can provide 61473 it's own main(). This isn't the ideal way of doing this, as it prevents libdix 61474 built this way from being useful with any other DDX. 61475 61476 So instead, always name that function dix_main(), and also provide a stub main() 61477 which just calls dix_main(), which can be overriden in the DDX. 61478 61479 Add a main() to XWin (XQuartz already has one, of course). 61480 61481 It's no longer neccessary to link XWin and XQuartz with libmain. 61482 61483 v2: Remove unneeded stub main hw/xwin/InitOutput.c 61484 61485 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61486 Reviewed-by: Julien Cristau <jcristau@debian.org> 61487 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61488 61489commit d5ebe20f9ba9569351c4a41449866679fd60ba45 61490Merge: 21ea7ebb6 33f7e6078 61491Author: Keith Packard <keithp@keithp.com> 61492Date: Mon Jul 22 20:34:41 2013 -0700 61493 61494 Merge remote-tracking branch 'alanc/master' 61495 61496commit 21ea7ebb6a1ad56f3060747af0df235ca6ff27fa 61497Author: Peter Hutterer <peter.hutterer@who-t.net> 61498Date: Fri Jul 19 08:56:38 2013 +1000 61499 61500 dix: scale y back instead of x up when pre-scaling coordinates 61501 61502 The peculiar way we handle coordinates results in relative coordinates on 61503 absolute devices being added to the last value, then that value is mapped to 61504 the screen (taking the device dimensions into account). From that mapped 61505 value we get the final coordinates, both screen and device coordinates. 61506 61507 To avoid uneven scaling on relative coordinates, they are pre-scaled by 61508 screen ratio:resolution:device ratio factor before being mapped. This 61509 ensures that a circle drawn on the device is a circle on the screen. 61510 61511 Previously, we used the ratio to scale x up. Synaptics already does its own 61512 scaling based on the resolution and that is done by scaling y down by the 61513 ratio. So we can remove the code from the driver and get approximately the 61514 same behaviour here. 61515 61516 Minor ABI bump, so we can remove this from synaptics. 61517 61518 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61519 Tested-by: Emmanuel Benisty <benisty.e@gmail.com> 61520 61521commit 33f7e60785f12770ce10558c2ca7ce1323eefc59 61522Author: Thomas Klausner <wiz@NetBSD.org> 61523Date: Fri Jul 12 08:21:19 2013 +0200 61524 61525 Fix typo in header guard. 61526 61527 Signed-off-by: Thomas Klausner <wiz@NetBSD.org> 61528 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61529 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61530 61531commit ddc52b9cbae017f04c7a232af4e8b16c9f96823d 61532Author: Alan Coopersmith <alan.coopersmith@oracle.com> 61533Date: Fri Jul 5 22:43:17 2013 -0700 61534 61535 Xephyr: Use _XEatDataWords (from Xlib 1.6) instead of _XEatData 61536 61537 Simplifies code and reduces risk of overflow from converting length 61538 field in X replies from words to bytes. 61539 61540 One call to _XEatData is left in ephyrHostGLXGetStringFromServer where 61541 it's already been checked for overflow, and other values have been 61542 subtracted from it to reduce the size of data remaining to be eaten. 61543 61544 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61545 61546commit a3d43edf71847f4b486f971405d2b457f81b73d1 61547Author: Alan Coopersmith <alan.coopersmith@oracle.com> 61548Date: Fri Jul 5 22:35:32 2013 -0700 61549 61550 glxproxy: Use _XEatDataWords (from Xlib 1.6) instead of _XEatData 61551 61552 Reduces risk of overflow from converting length field in X replies 61553 from words to bytes. (Also seems to be what several calls were 61554 already incorrectly passing to _XEatData.) 61555 61556 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61557 61558commit 87b0cabc145a9b5f6faffdfb544ce1c647b8ab72 61559Author: Alan Coopersmith <alan.coopersmith@oracle.com> 61560Date: Fri Jul 5 22:32:10 2013 -0700 61561 61562 glxproxy: Use _XReadPad instead of _XEatData to clean up the slop 61563 61564 Xlib already provides a function to eat padding bytes after the 61565 data read, so use it instead of calculating it ourselves. 61566 61567 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61568 61569commit 1cb182cbdc11fc1c97507c57875f1d2453f27328 61570Author: Alan Coopersmith <alan.coopersmith@oracle.com> 61571Date: Mon Apr 15 08:41:14 2013 -0700 61572 61573 Xephyr: integer overflow in XF86DRIGetClientDriverName() 61574 61575 clientDriverNameLength is a CARD32 and needs to be bounds checked before 61576 adding one to it to come up with the total size to allocate, to avoid 61577 integer overflow leading to underallocation and writing data from the 61578 network past the end of the allocated buffer. 61579 61580 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 61581 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61582 Reviewed-by: Julien Cristau <jcristau@debian.org> 61583 61584commit 20644e53b385c54d73242c86a7d3f981d18a3843 61585Author: Alan Coopersmith <alan.coopersmith@oracle.com> 61586Date: Mon Apr 15 08:39:03 2013 -0700 61587 61588 Xephyr: integer overflow in XF86DRIOpenConnection() 61589 61590 busIdStringLength is a CARD32 and needs to be bounds checked before adding 61591 one to it to come up with the total size to allocate, to avoid integer 61592 overflow leading to underallocation and writing data from the network past 61593 the end of the allocated buffer. 61594 61595 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 61596 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61597 Reviewed-by: Julien Cristau <jcristau@debian.org> 61598 61599commit c37ceda76bf9ec6f5166122e864663e10f106546 61600Author: Alan Coopersmith <alan.coopersmith@oracle.com> 61601Date: Sun Apr 14 10:50:50 2013 -0700 61602 61603 Xephyr: integer overflow in ephyrHostGLXGetStringFromServer() 61604 61605 reply.length & reply.size are CARD32s and need to be bounds checked before 61606 multiplying or adding to come up with the total size to allocate, to avoid 61607 integer overflow leading to underallocation and writing data from the 61608 network past the end of the allocated buffer. 61609 61610 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 61611 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61612 61613commit 5fbd8c45b46ab93522e417240aa770466c30b735 61614Author: Alan Coopersmith <alan.coopersmith@oracle.com> 61615Date: Sun Apr 14 09:45:26 2013 -0700 61616 61617 Xdmx: integer overflow in GetGLXFBConfigs() 61618 61619 numFBConfigs & numAttribs are CARD32s and need to be bounds checked before 61620 multiplying by structure sizes to come up with the total size to allocate, 61621 to avoid integer overflow leading to underallocation and writing data from 61622 the network past the end of the allocated buffer. 61623 61624 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 61625 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61626 61627commit c89bcd12f4fc7233830a8dbe7863c312f657da3c 61628Author: Alan Coopersmith <alan.coopersmith@oracle.com> 61629Date: Sun Apr 14 09:33:57 2013 -0700 61630 61631 Xdmx: integer overflow in GetGLXVisualConfigs() 61632 61633 numVisuals & numProps are both CARD32 and need to be bounds checked before 61634 multiplying by structure sizes to come up with the total size to allocate, 61635 to avoid integer overflow leading to underallocation and writing data from 61636 the network past the end of the allocated buffer. 61637 61638 Reported-by: Ilja Van Sprundel <ivansprundel@ioactive.com> 61639 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61640 61641commit 74469895e39fa38337f59edd64c4031ab9bb51d8 61642Author: Peter Hutterer <peter.hutterer@who-t.net> 61643Date: Mon Jul 15 13:06:28 2013 +1000 61644 61645 dix: allow a ConstantDeceleration between 0 and 1 (#66134) 61646 61647 A constant deceleration of x simply means (delta * 1/x). We limited that to 61648 values >= 1.0f for obvious reasons, but can also allow values from 0-1. 61649 That means that ConstantDeceleration is actually a ConstantAcceleration, but 61650 hey, if someone needs it... 61651 61652 X.Org Bug 66134 <http://bugs.freedesktop.org/show_bug.cgi?id=66134> 61653 61654 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61655 Reviewed-by: Keith Packard <keithp@keithp.com> 61656 61657commit 0e3be0b25fcfeff386bad132526352c2e45f1932 61658Author: Peter Hutterer <peter.hutterer@who-t.net> 61659Date: Fri Jul 12 14:10:10 2013 +1000 61660 61661 dix: UpdateTouchesForGrab must only free the listener grab if it is non-NULL 61662 61663 If a client calls XIGrabDevice in response to a ButtonPress event (regular 61664 event selection), the device will have a grab, but listener->grab is NULL. 61665 61666 Check for that, to avoid logspam. 61667 61668 [ 26293.863] (EE) BUG: triggered 'if (!pGrab)' 61669 [ 26293.863] (EE) BUG: grabs.c:256 in FreeGrab() 61670 [ 26293.863] (EE) 61671 [ 26293.863] (EE) Backtrace: 61672 [ 26293.864] (EE) 0: /usr/bin/Xorg (FreeGrab+0x54) [0x45d3fc] 61673 [ 26293.864] (EE) 1: /usr/bin/Xorg (UpdateTouchesForGrab+0x135) [0x447d4e] 61674 [ 26293.864] (EE) 2: /usr/bin/Xorg (ActivatePointerGrab+0x1ba) [0x447f3d] 61675 [ 26293.864] (EE) 3: /usr/bin/Xorg (GrabDevice+0x3e6) [0x4503bc] 61676 [ 26293.864] (EE) 4: /usr/bin/Xorg (ProcXIGrabDevice+0x1f9) [0x5981b1] 61677 [ 26293.865] (EE) 5: /usr/bin/Xorg (ProcIDispatch+0x78) [0x58aa17] 61678 [ 26293.865] (EE) 6: /usr/bin/Xorg (Dispatch+0x30d) [0x43347e] 61679 [ 26293.865] (EE) 7: /usr/bin/Xorg (main+0x61d) [0x498175] 61680 [ 26293.865] (EE) 8: /lib64/libc.so.6 (__libc_start_main+0xf5) [0x3df5621b75] 61681 [ 26293.865] (EE) 9: /usr/bin/Xorg (_start+0x29) [0x423a19] 61682 [ 26293.866] (EE) 10: ? (?+0x29) [0x29] 61683 [ 26293.866] (EE) 61684 61685 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61686 Reviewed-by: Keith Packard <keithp@keithp.com> 61687 61688commit 24cbb47ca80cd7a47749ece0ff3a98657ddf3ae3 61689Author: Peter Hutterer <peter.hutterer@who-t.net> 61690Date: Fri Jul 12 12:17:47 2013 +1000 61691 61692 config: provide example configuration for multi-seat setups 61693 61694 Seats other than seat0 need custom configuration. Provide that with a 61695 default configuration file so we can share it across distros. 61696 61697 This file intentionally does not end in .conf so it won't get picked up by 61698 the server after a normal installation. gdm, or whatever starts up the 61699 servers will have to explicitly specifiy this config file. 61700 61701 This file replaces the one currently written by systemd's multi-seat-x 61702 binary: 61703 http://cgit.freedesktop.org/systemd/systemd/tree/src/login/multi-seat-x.c 61704 61705 CC: Lennart Poettering <lennart@poettering.net> 61706 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61707 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 61708 61709commit 791121e00692e12cb4c38c8d018bd72eb5081eb7 61710Author: François Tigeot <ftigeot@wolfpond.org> 61711Date: Mon Jul 15 21:54:26 2013 +0200 61712 61713 Fix mouse header include on DragonFly and FreeBSD 61714 61715 * __FreeBSD_kernel_version doesn't exist anymore 61716 61717 * The removed check was for FreeBSD versions from before September 2000 61718 which are no longer supported anyway 61719 61720 * Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=66045 61721 61722 Signed-off-by: François Tigeot <ftigeot@wolfpond.org> 61723 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61724 61725commit 59a6d3f1eb03010ab831b32b03706a1d6143c732 61726Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61727Date: Thu Nov 8 13:41:13 2012 +0000 61728 61729 os: Remove any old logfile before trying to write to it 61730 61731 If we are not backing up logfiles, remove the old logfile before trying to write 61732 a new logfile, as otherwise the operation may fail if the previous logfile was 61733 created by a different user. 61734 61735 This change is useful when: 61736 - The DDX doesn't use the logfile backup mechanism (i.e. not Xorg) 61737 - The DDX is run by a non-root user, and then by a different non-root user 61738 - The logfile directory doesn't have the restricted-deletion flag set 61739 61740 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61741 Acked-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 61742 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 61743 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61744 61745commit 8eeaa74bc241acb41f1d3ed64971e0b01e794776 61746Author: Peter Hutterer <peter.hutterer@who-t.net> 61747Date: Tue Jul 9 13:27:19 2013 +1000 61748 61749 dix: when ungrabbing an active grab, accept pointer grabs (#66720) 61750 61751 Ungrabbing a device during an active touch grab rejects the grab. Ungrabbing 61752 a device during an active pointer grab accepts the grab. 61753 61754 Rejection is not really an option for a pointer-emulated grab, if a client 61755 has a button mask on the window it would get a ButtonPress emulated after 61756 UngrabDevice. That is against the core grab behaviour. 61757 61758 X.Org Bug 66720 <http://bugs.freedesktop.org/show_bug.cgi?id=66720> 61759 61760 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61761 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 61762 61763commit a2d6932ad48f2cbfcc4c5fb8272e47106d08ac78 61764Author: Julien Cristau <jcristau@debian.org> 61765Date: Mon Jul 1 19:39:30 2013 +0200 61766 61767 xkb: don't call atoi(NULL) when parsing argv 61768 61769 If the -ardelay or -arinterval options have no argument, there's no 61770 point trying to read it. 61771 61772 See 61773 http://www.forallsecure.com/bug-reports/feb3db57fc206d8df22ca53a6907f74973876272/ 61774 61775 Reported-by: Alexandre Rebert <alexandre@cmu.edu> 61776 Signed-off-by: Julien Cristau <jcristau@debian.org> 61777 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 61778 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61779 61780commit 77e51d5bbb97eb5c9d9dbff9a7c44d7e53620e68 61781Author: Eric Anholt <eric@anholt.net> 61782Date: Mon Jun 17 15:51:19 2013 -0700 61783 61784 Revert "DRI2: re-allocate DRI2 drawable if pixmap serial changes" 61785 61786 This reverts commit 3209b094a3b1466b579e8020e12a4f3fa78a5f3f. After a 61787 long debug session by Paul Berry, it appears that this was the commit 61788 that has been producing sporadic failures in piglit front buffer 61789 rendering tests for the last several years. 61790 61791 GetBuffers may return fresh buffers with invalid contents at a couple 61792 reasonable times: 61793 61794 - When first asked for a non-fake-front buffer. 61795 - When the drawable size is changed, an Invalidate has been sent, and 61796 obviously the app needs to redraw the whole buffer. 61797 - After a glXSwapBuffers(), GL allows the backbuffer to be undefined, 61798 and an Invalidate was sent to tell the GL that it should grab these 61799 appropriate new buffers to avoid stalling. 61800 61801 But with the patch being reverted, GetBuffers would also return fresh 61802 invalid buffers when the drawable serial number changed, which is 61803 approximately "whenever, for any reason". The app is not expecting 61804 invalid buffer contents "whenever", nor is it valid. Because the GL 61805 usually only GetBuffers after an Invalidate is sent, and the new 61806 buffer allocation only happened during a GetBuffers, most apps saw no 61807 problems. But apps that do (fake-)frontbuffer rendering do frequently 61808 ask the server for the front buffer (since we drop the fake front 61809 allocation when we're not doing front buffer rendering), and if the 61810 drawable serial got bumped midway through a draw, the server would 61811 pointlessly ditch the front *and* backbuffer full of important 61812 drawing, resulting in bad rendering. 61813 61814 The patch was originally to fix bugzilla: 61815 https://bugs.freedesktop.org/show_bug.cgi?id=28365 61816 Specifically: 61817 61818 To reproduce, start with a large-ish display (i.e. 1680x1050 on my 61819 laptop), use the patched glxgears from bug 28252 to add the 61820 -override option. Then run glxgears -override -geometry 640x480 61821 to create a 640x480 window in the top left corner, which will work 61822 fine. Next, run xrandr -s 640x480 and watch the fireworks. 61823 61824 I've tested with an override-redirect glxgears, both with vblank sync 61825 enabled and disabled, both with gnome-shell and no window manager at 61826 all, before and after this patch. The only problem observed was that 61827 before and after the revert, sometimes when alt-tabbing to kill my 61828 gears after completing the test gnome-shell would get confused about 61829 override-redirectness of the glxgears window (according to a log 61830 message) and apparently not bother doing any further compositing. 61831 61832 Signed-off-by: Eric Anholt <eric@anholt.net> 61833 Reviewed-by: Keith Packard <keithp@keithp.com> 61834 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 61835 Tested-by: Chris Wilson <chris@chris-wilson.co.uk> 61836 Signed-off-by: Keith Packard <keithp@keithp.com> 61837 61838commit 227015d6544115258b74c31b700f9bd163540e24 61839Merge: d9005d58c 80054fcaf 61840Author: Keith Packard <keithp@keithp.com> 61841Date: Tue Jun 18 10:53:08 2013 -0700 61842 61843 Merge remote-tracking branch 'jturney/master' 61844 61845commit 951d2f83fc65a136717fb8090510d053e3f3a6c0 61846Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61847Date: Wed Jun 12 14:11:53 2013 +0200 61848 61849 bump to 0.8.0 61850 61851 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61852 61853commit 525ac7fb9a0d639253fc5d6556db3edb7239f932 61854Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61855Date: Wed Jun 12 14:05:19 2013 +0200 61856 61857 modesetting: probe only succeeds if connectors are detected 61858 61859 This will prevent modesetting being used for outputless intel or nvidia cards. 61860 61861 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61862 61863commit d4791dd97b85ddac1295761bf6ab4f10c08346d7 61864Author: Dave Airlie <airlied@redhat.com> 61865Date: Tue Jun 11 10:29:25 2013 +1000 61866 61867 modesetting: fix adjust frame crash 61868 61869 When SDL called this it was totally broken, actually hook 61870 up to the underlying drmmode function. 61871 61872 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=64808 61873 61874 Thanks to Peter Wu <lekensteyn@gmail.com> for harassing me. 61875 61876 Signed-off-by: Dave Airlie <airlied@redhat.com> 61877 61878commit 80054fcafb095aff76f6df00d348c7fd85e8ddbc 61879Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61880Date: Sat Mar 30 19:44:14 2013 +0000 61881 61882 hw/xwin: Update manifest to target all architectures, not just x86 61883 61884 Update manifest to target all architectures, not just x86 61885 61886 Also: Write explicit dependencies for Xwin.rc. XWin.exe.manifest and X.ico are 61887 included the resource compiler output, so add a dependency on them to cause it 61888 to be recompiled if they change. 61889 61890 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61891 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61892 61893commit 689e7ad9b0fccd582cd63796cc90a999f141f589 61894Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61895Date: Sun Mar 31 18:36:37 2013 +0100 61896 61897 hw/xwin: Win32 Windows properties are of pointer type HANDLE, not an integer type 61898 61899 Win32 Windows properties are of pointer type HANDLE, not an integer type, but we 61900 use the Windows property WINDOW_WID_PROP to store the X window XID. 61901 61902 Add appropriate casts to show it doesn't matter that an XID is smaller than a 61903 pointer on 64-bit build. 61904 61905 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61906 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61907 61908commit e1a0945e8d4a3858838a14e64022524efd73c905 61909Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61910Date: Sun Mar 31 18:35:35 2013 +0100 61911 61912 hw/xwin: winMultiWindowGetTransientFor() returns an XID not a WindowPtr 61913 61914 winMultiWindowGetTransientFor() accesses the WM_TRANSIENT_FOR property, which is 61915 an X window XID (which we compare with WINDOW_WID_PROP) and not a WindowPtr. 61916 61917 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61918 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61919 61920commit be201ac257687355bf0916d7c525d598ed9b1282 61921Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61922Date: Sat Mar 30 19:47:16 2013 +0000 61923 61924 hw/xwin: Change winIconIsOverride() to take a HICON parameter 61925 61926 Change winIconIsOverride() to take a HICON parameter, so some unneccessary 61927 casts, which weren't portable to 64-bit Cygwin, can be removed. 61928 61929 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61930 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61931 61932commit fc40f9a92cf5981bccc45bbc0b959cc06c5accc2 61933Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61934Date: Sat Mar 30 19:41:06 2013 +0000 61935 61936 hw/xwin: Fix uses of CreateDialogParam() for 64-bit builds 61937 61938 Fix uses of CreateDialogParam() to be more strictly correct, for 64-bit builds 61939 61940 dwInitParam parameter is LPARAM, not int 61941 Return type of DLGPROC is INT_PTR, not WINBOOL 61942 61943 Also: Reporting the value of g_hDlgDepthChange is completely uninteresting, we 61944 are just interested in GetLastError() if it is NULL 61945 61946 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61947 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61948 61949commit b542976846eb5d4c82910272af32ec8298da4177 61950Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61951Date: Sat Mar 30 19:40:31 2013 +0000 61952 61953 hw/xwin: deprecated interface cygwin_conv_to_win32_path() doesn't exist in 64-bit cygwin 61954 61955 The deprecated interface cygwin_conv_to_win32_path() doesn't exist in 64-bit 61956 cygwin, but both the ID_ABOUT_CHANGELOG control and the referenced file who's 61957 path we are converting haven't existed for a while, so just remove this unused, 61958 cygwin-specific code. 61959 61960 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61961 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61962 61963commit 5a47c6420cfe298143310938f08cc36a503e8d3d 61964Author: Jon TURNEY <jon.turney@dronecode.org.uk> 61965Date: Mon Apr 8 01:27:26 2013 +0100 61966 61967 hw/xwin: Fix a 64-bit portability issue with casting HINSTANCE result of ShellExecute() to an integer 61968 61969 Fix a 64-bit portability issue with casting HINSTANCE result of ShellExecute() 61970 to an integer: Since HINSTANCE is a pointer type, use INT_PTR rather than int. 61971 61972 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 61973 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 61974 61975commit d9005d58c0413dcc3a99150c13d05163881cf782 61976Author: Peter Hutterer <peter.hutterer@who-t.net> 61977Date: Tue Jun 4 14:39:33 2013 +1000 61978 61979 config: add a quirks file for devices needing special configuration 61980 61981 This is the fedora quirks file configuring the special handling some devices 61982 need. Rather than keeping this distro-specific add it to the tree so we can 61983 share these quirks. 61984 61985 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61986 Reviewed-by: Jamey Sharp <jamey@minilop.net> 61987 61988commit 04ce6f6f6c880281e42c9ebf21d6f980ddae79b4 61989Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61990Date: Thu May 30 11:23:07 2013 +0200 61991 61992 dix: add missing dixFreeRegistry define 61993 61994 Fixes a build failure on debian's udeb builds. 61995 61996 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 61997 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 61998 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 61999 62000commit c21344add2fc589df83b29be5831c36a372201bd 62001Author: Peter Hutterer <peter.hutterer@who-t.net> 62002Date: Mon May 27 13:46:49 2013 +1000 62003 62004 dix: remove logspam in RefCursor() 62005 62006 This shouldn't have been in the patch 62007 62008 Reported-by: Colin Harrison <colin.harrison@virgin.net> 62009 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62010 Reviewed-by: Keith Packard <keithp@keithp.com> 62011 Signed-off-by: Keith Packard <keithp@keithp.com> 62012 62013commit cb3018d8a15add48efb3335e1f79b66378166008 62014Merge: c76a1b343 6589f3b55 62015Author: Keith Packard <keithp@keithp.com> 62016Date: Thu May 23 19:58:46 2013 -0600 62017 62018 Merge remote-tracking branch 'whot/unreviewed' 62019 62020commit c76a1b343d6a56aa9529e87f0eda8d61355d562b 62021Merge: 891123ce9 9a5ad6533 62022Author: Keith Packard <keithp@keithp.com> 62023Date: Thu May 23 19:58:36 2013 -0600 62024 62025 Merge remote-tracking branch 'whot/touch-grab-race-condition-56578-v3' 62026 62027commit 891123ce9dac5e52685b9a921b33d8279ba52956 62028Merge: 7e97166c4 49521edaf 62029Author: Keith Packard <keithp@keithp.com> 62030Date: Thu May 23 10:30:15 2013 -0600 62031 62032 Merge remote-tracking branch 'whot/for-keith' 62033 62034commit 6589f3b55e335eef6c658c8ba1fe15a062f7e31c 62035Author: Peter Hutterer <peter.hutterer@who-t.net> 62036Date: Tue May 14 08:31:19 2013 +1000 62037 62038 dix: devices must have valuators before touch is initialized 62039 62040 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62041 62042commit 756ab88d93542f0589c9bf46f40ccc57df64f0fd 62043Author: Peter Hutterer <peter.hutterer@who-t.net> 62044Date: Fri Apr 26 15:10:08 2013 +1000 62045 62046 dix: fix device scaling to use a [min,max[ range. 62047 62048 defmin/defmax are screen coords and thus use a min-inclusive, max-exclusive 62049 range. device axes ranges are inclusive, so bump the max up by one to get the 62050 scaling right. 62051 62052 This fixes off-by-one coordinate errors if the coordinate matrix is used to 62053 bind the device to a fraction of the screen. It introduces an off-by-one 62054 scaling error in the device coordinate range, but since most devices have a 62055 higher resolution than the screen (e.g. a Wacom I4 has 5080 dpi) the effect 62056 of this should be limited. 62057 62058 This error manifests when we have numScreens > 1, as the scaling from 62059 desktop size back to screen size drops one device unit. 62060 62061 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62062 62063commit 5cc2c96f824dbb28b9f8da61efc41596f8bd0561 62064Author: Peter Hutterer <peter.hutterer@who-t.net> 62065Date: Sat Mar 9 20:43:51 2013 +1000 62066 62067 dix: pre-scale x by the screen:device:resolution ratio 62068 62069 commit 61a99aff9d33728a0b67920254d2d4d79f80cf39 62070 dix: pre-scale relative events from abs devices to desktop ratio (#31636) 62071 62072 added pre-scaling of relative coordinates coming from absolute devices to 62073 undo uneven scaling based on the screen dimensions. 62074 62075 Devices have their own device width/height ratio as well (in a specific 62076 resolution) and this must be applied for relative devices as well to avoid 62077 scaling of the relative events into the device's ratio. 62078 62079 e.g. a Wacom Intuos4 6x9 is in 16:10 format with equal horiz/vert 62080 resolution (dpi). A movement by 1000/1000 coordinates is a perfect diagonal 62081 on the tablet and must be reflected as such on the screen. 62082 62083 However, we map the relative device-coordinate events to absolute screen 62084 coordinates based on the axis ranges. This results in an effective scaling 62085 of 1000/(1000 * 1.6) and thus an uneven x/y axis movement - the y 62086 axis is always faster. 62087 62088 So we need to pre-scale not only by the desktop dimenstions but also by the 62089 device width/height ratio _and_ the resolution ratio. 62090 62091 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62092 62093commit 9a5ad65330693b3273972b63d10f2907d9ab954a 62094Author: Peter Hutterer <peter.hutterer@who-t.net> 62095Date: Wed May 15 19:01:11 2013 +1000 62096 62097 Abstract cursor refcounting 62098 62099 Too many callers relied on the refcnt being handled correctly. Use a simple 62100 wrapper to handle that case. 62101 62102 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62103 62104commit 35c2e263db01b2b61354298e5e85aa3cae8ac317 62105Author: Peter Hutterer <peter.hutterer@who-t.net> 62106Date: Tue May 14 14:51:31 2013 +1000 62107 62108 dix: call UpdateDeviceState() for emulated TouchEndEvents 62109 62110 ProcessTouchEvents() calls UDS for all touch events, but if the event type 62111 was switched to TouchUpdate(pending end) UDS is a noop. 62112 62113 Daniel Drake found this can cause stuck buttons if a touch grab is 62114 activated, rejected and the touch event is passed to a regular listener. 62115 This sequence causes the TouchEnd to be changed to TouchUpdate(pending end). 62116 62117 The actual TouchEnd event is later generated by the server once it is passed 62118 to the next listener. UDS is never called for this event, thus the button 62119 remains logically down. 62120 62121 A previous patch suggested for UDS to handle TouchUpdate events [1], however 62122 this would release the button when the first TouchEvent is processed, not 62123 when the last grab has been released (as is the case for sync pointer 62124 grabs). A client may thus have the grab on the device, receive a ButtonPress 62125 but see the button logically up in an XQueryPointer request. 62126 62127 This patch adds a call to UDS to TouchEmitTouchEnd(). The device state must 62128 be updated once a TouchEnd event was sent to the last grabbing listener and 62129 the number of grabs on the touchpoint is 0. 62130 62131 [1] http://patchwork.freedesktop.org/patch/13464/ 62132 62133 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62134 62135commit 8b9dc2628115dcb3f3601ad19b1ae157df21b9ee 62136Author: Peter Hutterer <peter.hutterer@who-t.net> 62137Date: Tue May 14 07:46:25 2013 +1000 62138 62139 dix: devices must have valuators before touch is initialized 62140 62141 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62142 62143commit fd5ea0237db6d725a48f76b706135df9d3246b82 62144Author: Peter Hutterer <peter.hutterer@who-t.net> 62145Date: Mon May 13 15:22:12 2013 +1000 62146 62147 Xi: fix warning - remove unused 'rc' 62148 62149 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62150 62151commit 481702101b86fff003430e952dc65fb41eb56400 62152Author: Peter Hutterer <peter.hutterer@who-t.net> 62153Date: Fri May 3 15:07:58 2013 +1000 62154 62155 dix: fix cursor refcounting 62156 62157 The cursor is referenced during CopyGrab(), thus doesn't need to be handled 62158 manually anymore. It does need to be refcounted for temp grabs though. 62159 62160 The oldGrab handling in ProcGrabPointer is a leftover from the cursor in the 62161 grab being refcounted, but the grab itself being a static struct in the 62162 DeviceIntRec. Now that all grabs are copied, this lead to a double-free of 62163 the cursor (Reproduced in Thunderbird, dragging an email twice (or more 62164 often) causes a crash). 62165 62166 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62167 62168commit 49521edaf8930b335afac2185209e26e7000820b 62169Author: Peter Hutterer <peter.hutterer@who-t.net> 62170Date: Wed May 15 08:51:51 2013 +1000 62171 62172 ephyr: add -resizeable to the man page 62173 62174 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62175 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 62176 62177commit 7a4bbfd9f2ce7aad23148ba01bb3f1071da606a3 62178Author: Peter Hutterer <peter.hutterer@who-t.net> 62179Date: Wed May 15 08:46:38 2013 +1000 62180 62181 ephyr: fix typo in man page 62182 62183 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62184 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 62185 62186commit 7e97166c4fc38076b6224e4dbdd632b43444149f 62187Author: Geert Uytterhoeven <geert@linux-m68k.org> 62188Date: Tue May 14 13:04:53 2013 +0200 62189 62190 Shadow: Switch the Amiga/Atari bitplane code to the canonical X.Org license 62191 62192 Also add my name to the list of authors in COPYING 62193 62194 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 62195 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 62196 Signed-off-by: Keith Packard <keithp@keithp.com> 62197 62198commit 3093f78d17e48a506aab170a9089cd10e21af299 62199Author: Peter Hutterer <peter.hutterer@who-t.net> 62200Date: Fri May 3 15:02:05 2013 +1000 62201 62202 dix: free the old grab when activating a new grab 62203 62204 A client may call XIGrabDevice twice, overwriting the existing grab. Thus, 62205 make sure we free the old copy after we copied it. Free it last, to make 62206 sure our refcounts don't run to 0 and inadvertantly free something on the 62207 way. 62208 62209 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62210 62211commit 2566bdd8bc996cccde77b846819808c6239a89d2 62212Author: Peter Hutterer <peter.hutterer@who-t.net> 62213Date: Wed Apr 24 14:46:06 2013 +1000 62214 62215 Xi: check for HAS_ACCEPTED only for grab listeners 62216 62217 If we have one listener left but it's not a grab, it cannot be in 62218 LISTENER_HAS_ACCEPTED state. 62219 62220 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62221 62222commit 5b00fc52270e9cfdfe7ac1838a21defe50fc3d31 62223Author: Peter Hutterer <peter.hutterer@who-t.net> 62224Date: Wed Apr 24 14:40:31 2013 +1000 62225 62226 Move TouchListenerGone call to CloseDownClient 62227 62228 TouchListenerGone cleans up if a client disappears. Having this in 62229 FreeGrab() triggers cyclic removal of grabs, emitting wrong events. In 62230 particular, it would clean up a passive grab record while that grab is 62231 active. 62232 62233 Move it to CloseDownClient() instead, cleaning up before we go. 62234 62235 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62236 62237commit 34c9b39d9937c2e19c2dffc9748605f90d40f965 62238Author: Peter Hutterer <peter.hutterer@who-t.net> 62239Date: Wed Apr 24 12:53:52 2013 +1000 62240 62241 dix: remove all listeners when freeing a touch 62242 62243 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62244 62245commit 395124bd2782823de37e5c5b2f15dba49cff05f6 62246Author: Peter Hutterer <peter.hutterer@who-t.net> 62247Date: Wed Apr 17 20:15:35 2013 +1000 62248 62249 dix: always copy grabs, don't reference them 62250 62251 Introduced in xorg-server-1.13.99.901-2-g9ad0fdb. Storing the grab pointer 62252 in the listener turns out to be a bad idea. If the grab is not an active 62253 grab or an implicit grab, the pointer stored is the one to the grab attached 62254 on the window. This grab may be removed if the client calls UngrabButton or 62255 similar while the touch is still active, leaving a dangling pointer. 62256 62257 To avoid this, copy the grab wherever we need to reference it later. This 62258 is also what we do for pointer/keyboard grabs, where we copy the grab as 62259 soon as it becomes active. 62260 62261 Reported-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 62262 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62263 62264commit 925e35122ebad877395bcf13676e9dbeb254bdfa 62265Author: Peter Hutterer <peter.hutterer@who-t.net> 62266Date: Tue Apr 23 15:52:18 2013 +1000 62267 62268 dix: AllocGrab can copy if an argument is passed in 62269 62270 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62271 62272commit 4980bcef9973ba1f90f53028f061669ee5d2661b 62273Author: Peter Hutterer <peter.hutterer@who-t.net> 62274Date: Tue Apr 23 15:46:04 2013 +1000 62275 62276 dix: freeing a null grab is a bug, complain if doing so 62277 62278 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62279 62280commit ccfa0f2d5de557546815a5e4f59552e2af46b578 62281Author: Peter Hutterer <peter.hutterer@who-t.net> 62282Date: Tue Apr 23 15:39:32 2013 +1000 62283 62284 dix: use a temporary variable for listeners[0] 62285 62286 no functional changes 62287 62288 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62289 62290commit 5363433a5cc64e2f83859aa1c32a89e5e1ddf9e4 62291Author: Peter Hutterer <peter.hutterer@who-t.net> 62292Date: Thu Apr 18 10:32:11 2013 +1000 62293 62294 dix: drop DeviceIntRec's activeGrab struct 62295 62296 Obsolete since 4bc2761ad5ec2d0668aec639780ffb136605fbc8. This struct 62297 existed so copying a passive grab could be simply done by 62298 activeGrab = *grab 62299 62300 and thus have a copy of the GrabPtr we'd get from various sources but still 62301 be able to check device->grab for NULL. 62302 62303 Since 4bc2761 activeGrab is a pointer itself and points to the same memory 62304 as grabinfo->grab, leaving us with the potential of dangling pointers if 62305 either calls FreeGrab() and doesn't reset the other one. 62306 62307 There is no reader of activeGrab anyway, so simply removing it is 62308 sufficient. 62309 62310 Note: field is merely renamed to keep the ABI. Should be removed in the 62311 future. 62312 62313 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62314 62315commit 7dbf61817d3bd4b1fc71710677e56c5d8cfcdb4e 62316Author: Peter Hutterer <peter.hutterer@who-t.net> 62317Date: Wed Apr 17 20:14:56 2013 +1000 62318 62319 dix: use a tmp variable for the to-be-removed touch listener 62320 62321 No functional changes. 62322 62323 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62324 62325commit a71a283934406d870bcd8dc376eb1c9ce1c8bbb4 62326Author: Peter Hutterer <peter.hutterer@who-t.net> 62327Date: Wed Apr 17 20:13:34 2013 +1000 62328 62329 dix: invert a loop condition 62330 62331 Change the single if condition in the loop body to a 62332 if (!foo) continue; 62333 and re-indent the rest. 62334 62335 No functional changes. 62336 62337 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62338 62339commit 5174b1f98204beee79eba74c4cda5f2be0a60a8f 62340Author: Peter Hutterer <peter.hutterer@who-t.net> 62341Date: Thu Feb 28 11:02:40 2013 +1000 62342 62343 dix: XAllowEvents() on a touch event means accepting it 62344 62345 A sync grab is the owner once it gets events. If it doesn't replay the 62346 event it will get all events from this touch, equivalent to accepting it. 62347 62348 If the touch has ended before XAllowEvents() is called, we also now need to 62349 send the TouchEnd event and clean-up since we won't see anything more from 62350 this touch. 62351 62352 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62353 62354commit e7f79c48b0faea910dc881034c00eb807fcd6210 62355Author: Peter Hutterer <peter.hutterer@who-t.net> 62356Date: Fri Mar 1 12:52:35 2013 +1000 62357 62358 dix: move EmitTouchEnd to touch.c 62359 62360 No functional changes, this just enables it to be re-used easier. 62361 62362 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62363 62364commit 0eb9390f6048049136347d5a5834d88bfc67cc09 62365Author: Peter Hutterer <peter.hutterer@who-t.net> 62366Date: Fri Mar 1 10:41:01 2013 +1000 62367 62368 Xi: Don't emit a TouchEnd event to a frozen device 62369 62370 EmitTouchEnd calls DeliverTouchEvents directly instead of through 62371 public.processInputProc. If a device is frozen, the TouchEnd is 62372 processed while the device is waiting for a XAllowEvents and thus ends the 62373 touch point (and the grab) before the client decided what to do with it. In 62374 the case of ReplayPointer, this loses the event. 62375 62376 This is a hack, but making EmitTouchEnd use processInputProc breaks 62377 approximately everything, especially the touch point is cleaned up during 62378 ProcessTouchEvents. Working around that is a bigger hack than this. 62379 62380 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62381 62382commit a7d989d335f903ffd8b168cd2beeb82c78d30c21 62383Author: Peter Hutterer <peter.hutterer@who-t.net> 62384Date: Thu Feb 28 13:07:26 2013 +1000 62385 62386 Xi: use public.processInputProc to replay the touch history 62387 62388 If a device is frozen in results to a grab, we need to enqueue the events. 62389 62390 This makes things complicated, and hard to follow since touch events are now 62391 replayed in the history, pushed into EnqueueEvent, then replayed later 62392 during PlayReleasedEvents in response to an XAllowEvents. 62393 62394 While the device is frozen, no touch events are processed, so if there is a 62395 touch client with ownership mask _below_ the grab this will delay the 62396 delivery and potentially screw gesture recognition. However, this is the 62397 behaviour we have already anyway if the top-most client is a sync pgrab or 62398 there is a sync grab active on the device when the TouchBegin was generated. 62399 62400 (also note, such a client would only reliably work in case of ReplayPointer 62401 anyway) 62402 62403 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62404 62405commit 214d11d3fcdac51fc7afbf3770516ec14f9e13c1 62406Author: Peter Hutterer <peter.hutterer@who-t.net> 62407Date: Wed Feb 27 15:05:54 2013 +1000 62408 62409 Xi: when punting to a new owner, always create TouchEnd events 62410 62411 If a touch is pending_finish and we just punted it to the next owner, that 62412 client must receive a TouchEnd event. 62413 62414 If we just punted to the last owner and that owner not a touch grab, we need 62415 to end the touch since this is the last event to be sent, and the client 62416 cannot accept/reject this. 62417 62418 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62419 62420commit 026627fe19aad007544eccf209f0dea05e67a7a7 62421Author: Peter Hutterer <peter.hutterer@who-t.net> 62422Date: Fri Mar 1 09:15:37 2013 +1000 62423 62424 Xi: save state for early acceptance 62425 62426 Delivering an event changes the state to LISTENER_IS_OWNER and we thus lose 62427 the information of early acceptance. 62428 62429 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62430 62431commit d905348134c80f19793eefb761731b00559ddf3a 62432Author: Peter Hutterer <peter.hutterer@who-t.net> 62433Date: Thu Feb 28 15:28:46 2013 +1000 62434 62435 Xi: if a passive async grab is activated from an emulated touch, accept 62436 62437 Async grabs cannot replay events, they cannot reject, so we can do an early 62438 accept here. 62439 62440 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62441 62442commit 8b0d21044956f3810199d5e2f38ce33069e97be7 62443Author: Peter Hutterer <peter.hutterer@who-t.net> 62444Date: Thu Feb 28 13:04:36 2013 +1000 62445 62446 Xi: fix lookup in ActivateEarlyAccept 62447 62448 ActivateEarlyAccept() can only be called from a grabbing client, so we can 62449 ignore the rest. And it's easy enough to get the client from that since 62450 9ad0fdb135a1c336771aee1f6eab75a6ad874aff. 62451 62452 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62453 62454commit d08bae297f9d7651edb1923d6b0d6b14b3d674fc 62455Author: Peter Hutterer <peter.hutterer@who-t.net> 62456Date: Thu Feb 28 13:08:27 2013 +1000 62457 62458 Xi: update the core listener state if we delivered the touch event 62459 62460 If a TouchBegin is sent to a core client, that client is now the owner. 62461 62462 By the time the TouchEnd is being processed, the client cannot replay 62463 anymore, so we can assume that this is the final touch end and we can clean 62464 up the touch record. 62465 62466 Note: DeliverTouchEmulatedEvent is called for all listeners and immediately 62467 bails out if the client is not the owner and thus shouldn't yet get the 62468 event. Thus, check the return code. 62469 62470 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62471 62472commit 81554d274f04951c55ea7f2e38d0455e2025e08d 62473Author: Peter Hutterer <peter.hutterer@who-t.net> 62474Date: Mon Feb 25 11:21:07 2013 +1000 62475 62476 Xi: if we delivered a TouchEnd to a passive grab, end it 62477 62478 ef64b5ee97099618cf2e2cbbd3e471095695ae24 (which introduced the 62479 TOUCH_CLIENT_ID check) has a wrong assumption that generated touch events 62480 (TOUCH_CLIENT_ID) should not terminate passive grabs. 62481 This is untrue, a TouchEnd may be generated in response to a TouchReject 62482 higher up. If we _deliver_ an event to a client, terminate the passive grab. 62483 62484 This requires us to count the actually delivered events too (first hunk). 62485 62486 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62487 62488commit 363b6387da6e669099a2da3861c73a29808295a6 62489Author: Peter Hutterer <peter.hutterer@who-t.net> 62490Date: Wed Feb 13 11:26:11 2013 +1000 62491 62492 dix: don't prepend an activated passive grab to the listeners 62493 62494 If the device is currently grabbed as the result of a passive grab 62495 activating, do not prepend that grab to the listeners (unlike active grabs). 62496 Otherwise, a client with a passive pointer grab will prevent touch grabs 62497 from activating higher up in the window stack. 62498 62499 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62500 62501commit 9cc45c18ad1511adf3fb163dd4cefbef106edb23 62502Author: Peter Hutterer <peter.hutterer@who-t.net> 62503Date: Wed Feb 13 10:49:23 2013 +1000 62504 62505 Xi: not having an ownership mask does not mean automatic acceptance 62506 62507 If we only have a single touch-grabbing client, setting the client as owner 62508 would clean up the touch once the TouchEnd was processed. If the client then 62509 calls XIAllowTouches() it will receive a BadValue for the touch ID (since 62510 the internal record is already cleaned up). 62511 62512 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62513 62514commit 2f1aedcaed8fd99b823d451bf1fb02330c078f67 62515Author: Peter Hutterer <peter.hutterer@who-t.net> 62516Date: Thu May 9 14:30:50 2013 +1000 62517 62518 input: print warnings if drivers don't initialize properly 62519 62520 If drivers supply incorrect values don't just quietly return False, spew to 62521 the log so we can detect what's going on. All these cases are driver bugs 62522 and should be fixed immediately. 62523 62524 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62525 Reviewed-by: Dave Airlie <airlied@redhat.com> 62526 62527commit 8a88b0ab52ba375ae84463a90503db88af10e368 62528Author: Peter Hutterer <peter.hutterer@who-t.net> 62529Date: Thu May 9 14:30:49 2013 +1000 62530 62531 dix: don't overwrite proximity/focus classes 62532 62533 InitPointerClassDeviceStruct/InitKeyboardDeviceStruct allocate a 62534 proximity/focus class, respectively. If a driver calls 62535 InitFocusClassDeviceStruct or InitProximityClassDeviceStruct beforehand, 62536 the previously allocated class is overwritten, leaking the memory. 62537 62538 Neither takes a parameter other than the device, so we can simply skip 62539 initialising it if we already have one. 62540 62541 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62542 Reviewed-by: Dave Airlie <airlied@redhat.com> 62543 62544commit 34b0d07ebf4a7874fe7fd336bef5bbdd8debda1c 62545Author: Peter Hutterer <peter.hutterer@who-t.net> 62546Date: Fri May 10 09:44:29 2013 +1000 62547 62548 dix: reset the OsBuffers after killing all clients 62549 62550 ==21860== 24 bytes in 1 blocks are still reachable in loss record 85 of 397 62551 ==21860== at 0x4C2B3F8: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 62552 ==21860== by 0x61ED93: AllocateOutputBuffer (io.c:1037) 62553 ==21860== by 0x61E15A: WriteToClient (io.c:764) 62554 ==21860== by 0x457B30: ProcQueryExtension (extension.c:275) 62555 ==21860== by 0x43596B: Dispatch (dispatch.c:432) 62556 ==21860== by 0x425DAB: main (main.c:295) 62557 62558 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62559 Reviewed-by: Dave Airlie <airlied@redhat.com> 62560 62561commit 67c66606c760c263d7a4c2d1bba43ed6225a4e7c 62562Author: Robert Morell <rmorell@nvidia.com> 62563Date: Thu May 9 13:09:02 2013 -0700 62564 62565 os: Reset input buffer's 'ignoreBytes' field 62566 62567 If a client sends a request larger than maxBigRequestSize, the server is 62568 supposed to ignore it. 62569 62570 Before commit cf88363d, the server would simply disconnect the client. After 62571 that commit, it attempts to gracefully ignore the request by remembering how 62572 long the client specified the request to be, and ignoring that many bytes. 62573 However, if a client sends a BigReq header with a large size and disconnects 62574 before actually sending the rest of the specified request, the server will 62575 reuse the ConnectionInput buffer without resetting the ignoreBytes field. This 62576 makes the server ignore new X clients' requests. 62577 62578 This fixes that behavior by resetting the ignoreBytes field when putting the 62579 ConnectionInput buffer back on the FreeInputs list. 62580 62581 Signed-off-by: Robert Morell <rmorell@nvidia.com> 62582 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 62583 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62584 62585commit ddc11397a56c745b5d1fb377e9d9b1fcc73802c8 62586Author: Peter Hutterer <peter.hutterer@who-t.net> 62587Date: Mon Feb 27 18:36:39 2012 +1000 62588 62589 dix: delete all callbacks before reset 62590 62591 DeleteCallbackManager() introduced for better symmetry in the caller, they 62592 do the same thing. 62593 62594 ==20085== 24 bytes in 1 blocks are still reachable in loss record 11 of 103 62595 ==20085== at 0x4C2A4CD: malloc (vg_replace_malloc.c:236) 62596 ==20085== by 0x43A097: CreateCallbackList (dixutils.c:837) 62597 ==20085== by 0x43A1D3: AddCallback (dixutils.c:869) 62598 ==20085== by 0x4B1736: GEExtensionInit (geext.c:209) 62599 ==20085== by 0x41C8A8: InitExtensions (miinitext.c:389) 62600 ==20085== by 0x5AC918: main (main.c:208) 62601 62602 ==2042== 8 bytes in 1 blocks are still reachable in loss record 2 of 97 62603 ==2042== at 0x4C2A4CD: malloc (vg_replace_malloc.c:236) 62604 ==2042== by 0x4C2A657: realloc (vg_replace_malloc.c:525) 62605 ==2042== by 0x4802F5: XNFrealloc (utils.c:1095) 62606 ==2042== by 0x43A17A: CreateCallbackList (dixutils.c:855) 62607 ==2042== by 0x43A1EF: AddCallback (dixutils.c:870) 62608 ==2042== by 0x4B1752: GEExtensionInit (geext.c:209) 62609 ==2042== by 0x41C8A8: InitExtensions (miinitext.c:389) 62610 ==2042== by 0x5AC9E4: main (main.c:208) 62611 ==2042== 62612 62613 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62614 Reviewed-by: Keith Packard <keithp@keithp.com> 62615 62616commit 6993f8b459e25b1a7a0a03e209688a28ce6c2c56 62617Author: Peter Hutterer <peter.hutterer@who-t.net> 62618Date: Wed Apr 24 14:59:11 2013 +1000 62619 62620 Xi: free barrier code at reset time 62621 62622 ==29423== 16 bytes in 1 blocks are definitely lost in loss record 73 of 328 62623 ==29423== at 0x4A06B6F: calloc (vg_replace_malloc.c:593) 62624 ==29423== by 0x5987C0: XIBarrierInit (xibarriers.c:908) 62625 ==29423== by 0x58F370: XInputExtensionInit (extinit.c:1300) 62626 ==29423== by 0x4F33C3: InitExtensions (miinitext.c:337) 62627 ==29423== by 0x4997DB: main (main.c:208) 62628 62629 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62630 Reviewed-by: Keith Packard <keithp@keithp.com> 62631 62632commit 042c6d861f7bb7038ddcdd6b59766fd9094d0e52 62633Author: Peter Hutterer <peter.hutterer@who-t.net> 62634Date: Tue Apr 30 15:44:37 2013 +1000 62635 62636 os: Use ErrorFSigSafe from FatalError and it's friends 62637 62638 Backtrace logging etc. is already sigsafe, but the actual FatalError message 62639 in response is not yet, leading to amusing logs like this: 62640 62641 (EE) Segmentation fault at address 0x0 62642 (EE) BUG: triggered 'if (inSignalContext)' 62643 (EE) BUG: log.c:499 in LogVMessageVerb() 62644 (EE) Warning: attempting to log data in a signal unsafe manner while in 62645 signal context. 62646 Please update to check inSignalContext and/or use LogMessageVerbSigSafe() or 62647 ErrorFSigSafe(). 62648 The offending log format message is: 62649 62650 Fatal server error: 62651 62652 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62653 Reviewed-by: Keith Packard <keithp@keithp.com> 62654 62655commit d9848fb4b182ca21bacf28ed7410d1a502cb000e 62656Author: Peter Hutterer <peter.hutterer@who-t.net> 62657Date: Thu Feb 14 16:24:53 2013 +1000 62658 62659 os: complain about unsupported pnprintf directives 62660 62661 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62662 Reviewed-by: Keith Packard <keithp@keithp.com> 62663 62664commit d903d17d7f006fa333265b8476063b189c20d082 62665Author: Peter Hutterer <peter.hutterer@who-t.net> 62666Date: Thu Feb 14 16:19:34 2013 +1000 62667 62668 os: support %c in pnprintf 62669 62670 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62671 Reviewed-by: Keith Packard <keithp@keithp.com> 62672 62673commit 58ef34ee6d0f68aa28f6f1a26e56f49ec85ed9bf 62674Author: Peter Hutterer <peter.hutterer@who-t.net> 62675Date: Thu Feb 14 16:13:22 2013 +1000 62676 62677 os: support %% in pnprintf 62678 62679 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62680 Reviewed-by: Keith Packard <keithp@keithp.com> 62681 62682commit 5ea21560dd071ea4ab87430000d087fd5fe1f092 62683Author: Peter Hutterer <peter.hutterer@who-t.net> 62684Date: Thu Feb 14 15:34:32 2013 +1000 62685 62686 os: support pnprintf length modifiers for integers 62687 62688 Mainly for %ld, smaller than int is propagated anyway, and %lld isn't really 62689 used. 62690 62691 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62692 Reviewed-by: Keith Packard <keithp@keithp.com> 62693 62694commit d3d4af5f9e505d444b6c82c6ea238206433fd24b 62695Author: Peter Hutterer <peter.hutterer@who-t.net> 62696Date: Mon Feb 27 18:59:45 2012 +1000 62697 62698 dix: reset the registry before quitting 62699 62700 Heaps of these: 62701 ==2042== 15,360 bytes in 120 blocks are still reachable in loss record 94 of 62702 97 62703 ==2042== at 0x4C2A4CD: malloc (vg_replace_malloc.c:236) 62704 ==2042== by 0x4C2A657: realloc (vg_replace_malloc.c:525) 62705 ==2042== by 0x45FB91: double_size (registry.c:65) 62706 ==2042== by 0x45FC97: RegisterRequestName (registry.c:85) 62707 ==2042== by 0x460095: RegisterExtensionNames (registry.c:179) 62708 ==2042== by 0x460729: dixResetRegistry (registry.c:334) 62709 ==2042== by 0x5AC992: main (main.c:201) 62710 62711 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62712 Reviewed-by: Keith Packard <keithp@keithp.com> 62713 62714commit 6f44d672aa34d343f63f0ea81ad58154a66b57ec 62715Author: Peter Hutterer <peter.hutterer@who-t.net> 62716Date: Mon Feb 27 17:17:41 2012 +1000 62717 62718 xkb: free XkbRulesUsed and XkbRulesDflt on extension cleanup 62719 62720 ==2547== 1 bytes in 1 blocks are still reachable in loss record 1 of 111 62721 ==2547== at 0x4C2A4CD: malloc (vg_replace_malloc.c:236) 62722 ==2547== by 0x64D1551: strdup (strdup.c:43) 62723 ==2547== by 0x4802FB: Xstrdup (utils.c:1113) 62724 ==2547== by 0x585B6C: XkbSetRulesUsed (xkbInit.c:219) 62725 ==2547== by 0x58700F: InitKeyboardDeviceStruct (xkbInit.c:595) 62726 ==2547== by 0x419FA3: vfbKeybdProc (InitInput.c:74) 62727 ==2547== by 0x425A3D: ActivateDevice (devices.c:540) 62728 ==2547== by 0x425F65: InitAndStartDevices (devices.c:713) 62729 ==2547== by 0x5ACA57: main (main.c:259) 62730 62731 and a few more of the above. 62732 62733 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62734 Reviewed-by: Keith Packard <keithp@keithp.com> 62735 62736commit 623981ddaeb8836f3b0939e527c0e943f9c4e974 62737Author: Peter Hutterer <peter.hutterer@who-t.net> 62738Date: Fri Feb 1 09:06:20 2013 +1000 62739 62740 If neither HAL nor udev backends are enabled, warn the user 62741 62742 If both are missing, input device hotplugging will not work out of the box. 62743 While we still have a DBus-API or the user may want to set AAD off all the 62744 time, the most likely source of this is misconfiguration (i.e. lack of the 62745 udev/hal devel packages). 62746 62747 Message printed last to make it more obvious to the user. 62748 62749 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62750 Reviewed-by: Keith Packard <keithp@keithp.com> 62751 62752commit dce40e2266200421647044ec7e856656d2ef952d 62753Author: Peter Hutterer <peter.hutterer@who-t.net> 62754Date: Tue Feb 26 12:34:21 2013 +1000 62755 62756 Xi: always return BadMatch for XTest devices ChangeDeviceControl requests 62757 62758 The only controls that still do something are DEVICE_RESOLUTION and 62759 DEVICE_ENABLE. 62760 62761 XTest devices have no resolution to change, and they cannot be disabled. So 62762 skip the lot, and prevent a crash in the DDX when it's trying to 62763 de-reference pInfo->control_proc on device with no pInfo struct. 62764 62765 Likewise, don't allow setting device mode or the valuators. 62766 XTest pointers are always relative, they don't have a mode. 62767 62768 Test cases: 62769 xts5/XI/ChangeDeviceControl (1/10) 62770 xts5/XI/SetDeviceValuators (1/6) 62771 and a few others 62772 62773 Reported-by: Knut Petersen <Knut_Petersen@t-online.de> 62774 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62775 Reviewed-by: Keith Packard <keithp@keithp.com> 62776 62777commit 23d1bc69f305edd5a6e2cfec3dfc84befda0881c 62778Author: Peter Hutterer <peter.hutterer@who-t.net> 62779Date: Thu Mar 14 11:07:57 2013 +1000 62780 62781 dix: send the current axis value in DeviceChangedEvents (#62321) 62782 62783 X.Org Bug 62321 <http://bugs.freedesktop.org/show_bug.cgi?id=62321> 62784 62785 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62786 Reviewed-by: Keith Packard <keithp@keithp.com> 62787 62788commit 27b11f5c19d67e2f6784d86cb7df86bbab3b58bb 62789Author: Peter Hutterer <peter.hutterer@who-t.net> 62790Date: Fri Feb 15 11:19:10 2013 +1000 62791 62792 Xi: fix comment - XI2 grabs aren't keysym grabs 62793 62794 Comment dates back to a pre-release version of XI2 that supported keysym 62795 grabs. That never made it into a release, it was ditched before. 62796 62797 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 62798 Reviewed-by: Keith Packard <keithp@keithp.com> 62799 62800commit 2746c681639f9512e6e45fb8d0151b996b6aff7f 62801Merge: 1e6cf8ec2 9878e097a 62802Author: Keith Packard <keithp@keithp.com> 62803Date: Mon May 6 10:52:40 2013 -0700 62804 62805 Merge remote-tracking branch 'alanc/master' 62806 62807commit 1e6cf8ec20d07b73a11116564aba71b4e4291dcd 62808Merge: dbfeaf706 5860408a1 62809Author: Keith Packard <keithp@keithp.com> 62810Date: Mon May 6 10:20:21 2013 -0700 62811 62812 Merge remote-tracking branch 'jturney/unused-but-set-variable-warning-fix' 62813 62814commit dbfeaf70623a83e1a3f3255c94d52e0e04702837 62815Author: Aaron Plattner <aplattner@nvidia.com> 62816Date: Tue Apr 30 14:30:18 2013 -0700 62817 62818 xfree86: don't enable anything in xf86InitialConfiguration for GPU screens 62819 62820 There's no point in turning on outputs connected to GPU screens during initial 62821 configuration. Not only does this cause them to just display black, it also 62822 confuses clients when these screens are attached to a master screen and RandR 62823 reports that the outputs are already on. 62824 62825 Also, don't print the warning about no outputs being found on GPU screens, 62826 since that's expected. 62827 62828 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 62829 Reviewed-by: Dave Airlie <airlied@gmail.com> 62830 62831commit 5860408a19d956f8cebb7ae39cfdbfcebd5e6360 62832Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62833Date: Wed Apr 3 18:43:26 2013 +0100 62834 62835 hw/xwin: Fix unused-but-set-variable warning in ProcWindowsWMFrameGetRect() 62836 62837 /jhbuild/checkout/xorg/xserver/hw/xwin/winwindowswm.c: In function ‘ProcWindowsWMFrameGetRect’: 62838 /jhbuild/checkout/xorg/xserver/hw/xwin/winwindowswm.c:322:12: error: variable ‘ir’ set but not used [-Werror=unused-but-set-variable] 62839 62840 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62841 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62842 62843commit bcfd523c150b34e9304d174ba812160c611f270d 62844Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62845Date: Wed Apr 3 18:34:38 2013 +0100 62846 62847 hw/xwin: Fix unused-but-set-variable warning in winMWExtWMUpdateWindowDecoration() 62848 62849 /jhbuild/checkout/xorg/xserver/hw/xwin/winwin32rootlesswindow.c: In function ‘winMWExtWMUpdateWindowDecoration’: 62850 /jhbuild/checkout/xorg/xserver/hw/xwin/winwin32rootlesswindow.c:189:11: error: variable ‘dwStyle’ set but not used [-Werror=unused-but-set-variable] 62851 62852 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62853 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62854 62855commit fbf819c24dc080f166cff29bf46b0feb604c6b8c 62856Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62857Date: Wed Apr 3 16:40:35 2013 +0100 62858 62859 hw/xwin: Fix unused-but-set-variable warning in winHotKeyAltTabPrimaryDD() 62860 62861 /jhbuild/checkout/xorg/xserver/hw/xwin/winpfbdd.c: In function ‘winHotKeyAltTabPrimaryDD’: 62862 /jhbuild/checkout/xorg/xserver/hw/xwin/winpfbdd.c:518:20: error: variable ‘rcSrc’ set but not used [-Werror=unused-but-set-variable] 62863 62864 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62865 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62866 62867commit 7dae1e59ce6077f475c04a41fa00096a74114064 62868Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62869Date: Wed Apr 3 16:33:39 2013 +0100 62870 62871 hw/xwin: Fix unused-but-set-variable warning in winMinimizeWindow() when built !XWIN_MULTIWINDOWEXTWM 62872 62873 /jhbuild/checkout/xorg/xserver/hw/xwin/winmultiwindowwindow.c: In function ‘winMinimizeWindow’: 62874 /jhbuild/checkout/xorg/xserver/hw/xwin/winmultiwindowwindow.c:813:20: error: variable ‘pScreenInfo’ set but not used [-Werror=unused-but-set-variable] 62875 62876 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62877 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62878 62879commit b284666f09b77f9c2f048fee22015a353152a57f 62880Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62881Date: Wed Apr 3 16:27:45 2013 +0100 62882 62883 hw/xwin: Fix unused-but-set-variable warning in winClipboardProc() 62884 62885 /jhbuild/checkout/xorg/xserver/hw/xwin/winclipboardthread.c: In function ‘winClipboardProc’: 62886 /jhbuild/checkout/xorg/xserver/hw/xwin/winclipboardthread.c:88:25: error: variable ‘atomClipboardManager’ set but not used [-Werror=unused-but-set-variable] 62887 62888 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62889 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62890 62891commit d414a09bb7f8d4440b36cbc6f0b4ca0eaefe5cbf 62892Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62893Date: Wed Apr 3 16:25:54 2013 +0100 62894 62895 hw/xwin: Fix unused-but-set-variable warning in winWindowProc() 62896 62897 /jhbuild/checkout/xorg/xserver/hw/xwin/winwndproc.c: In function ‘winWindowProc’: 62898 /jhbuild/checkout/xorg/xserver/hw/xwin/winwndproc.c:65:22: error: variable ‘s_hInstance’ set but not used [-Werror=unused-but-set-variable] 62899 62900 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62901 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62902 62903commit 7726102671690eb28138d032c8398ac903b97606 62904Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62905Date: Wed Apr 3 16:24:08 2013 +0100 62906 62907 hw/xwin: Fix unused-but-set-variable warning in winCopyWindowNativeGDI() 62908 62909 /jhbuild/checkout/xorg/xserver/hw/xwin/winwindow.c: In function ‘winCopyWindowNativeGDI’: 62910 /jhbuild/checkout/xorg/xserver/hw/xwin/winwindow.c:131:15: error: variable ‘pwinRoot’ set but not used [-Werror=unused-but-set-variable] 62911 62912 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62913 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62914 62915commit b9a2566a6e2f56a0335cd4a089b2066c06a67455 62916Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62917Date: Wed Apr 3 16:50:38 2013 +0100 62918 62919 hw/xwin: Fix unused-but-set-variable warning in winFinishScreenInitFB() 62920 62921 /jhbuild/checkout/xorg/xserver/hw/xwin/winscrinit.c:264:11: error: variable ‘pbits’ set but not used [-Werror=unused-but-set-variable] 62922 62923 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62924 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62925 62926commit c97fbd39adbc0a8cb8537f40b12c091aee29221a 62927Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62928Date: Wed Apr 3 16:18:27 2013 +0100 62929 62930 hw/xwin: Fix unused-but-set-variable warning in winIsFakeCtrl_L() 62931 62932 /jhbuild/checkout/xorg/xserver/hw/xwin/winkeybd.c:331:17: error: variable ‘lastMessage’ set but not used [-Werror=unused-but-set-variable] 62933 62934 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62935 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62936 62937commit cdb74fe17dcb3b5a9c61951f897deaf07f86775f 62938Author: Jon TURNEY <jon.turney@dronecode.org.uk> 62939Date: Wed Apr 3 16:16:06 2013 +0100 62940 62941 hw/xwin: Fix unused-but-set-variable warnings in windialog.c 62942 62943 /jhbuild/checkout/xorg/xserver/hw/xwin/windialogs.c: In function ‘winChangeDepthDlgProc’: 62944 /jhbuild/checkout/xorg/xserver/hw/xwin/windialogs.c:432:22: error: variable ‘s_pScreen’ set but not used [-Werror=unused-but-set-variable] 62945 /jhbuild/checkout/xorg/xserver/hw/xwin/windialogs.c: In function ‘winAboutDlgProc’: 62946 /jhbuild/checkout/xorg/xserver/hw/xwin/windialogs.c:564:22: error: variable ‘s_pScreen’ set but not used [-Werror=unused-but-set-variable] 62947 62948 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 62949 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 62950 62951commit f2fd8ec3725a61abbc831f0a9ec28fa2b7020c47 62952Author: Dave Airlie <airlied@redhat.com> 62953Date: Wed Jan 9 12:52:13 2013 +1000 62954 62955 gpu: call CreateScreenResources for GPU screens 62956 62957 I didn't think we needed this before, but after doing some more 62958 work with reverse optimus it seems like it should be called. 62959 62960 Reviewed-by: Keith Packard <keithp@keithp.com> 62961 Signed-off-by: Dave Airlie <airlied@redhat.com> 62962 62963commit 8fcb9d91b69abc72ddef31b9f2e8585580c6cad2 62964Author: Dave Airlie <airlied@redhat.com> 62965Date: Wed Jan 9 12:52:08 2013 +1000 62966 62967 dix: allow pixmap dirty helper to be used for non-shared pixmaps 62968 62969 this allows the pixmap dirty helper to be used for reverse optimus, 62970 where the GPU wants to copy from the shared pixmap to its VRAM copy. 62971 62972 [airlied: slave_dst is wrong name now but pointless ABI churn at this point] 62973 Reviewed-by: Keith Packard <keithp@keithp.com> 62974 Signed-off-by: Dave Airlie <airlied@redhat.com> 62975 62976commit 16077b81c502e04d77f81f683e0c213b9fe75393 62977Author: Dave Airlie <airlied@redhat.com> 62978Date: Wed Jan 9 12:52:03 2013 +1000 62979 62980 xf86crtc: don't use scrn->display for gpu screens 62981 62982 scrn->display is a property of the main screen really, and we don't 62983 want to have the GPU screens use it for anything when picking modes 62984 or a front buffer size. 62985 62986 This fixes a bug where when you plugged a display link device, it 62987 would try and allocate a screen the same size as the current running 62988 one (3360x1050 in this case), which was too big for the device. Avoid 62989 doing this and just pick sizes based on whats plugged into this device. 62990 62991 Reviewed-by: Keith Packard <keithp@keithp.com> 62992 Signed-off-by: Dave Airlie <airlied@redhat.com> 62993 62994commit f0d0d75bfe62553dde353f89e46ff13dd863fbe8 62995Author: Dave Airlie <airlied@redhat.com> 62996Date: Wed Jan 9 12:51:55 2013 +1000 62997 62998 dix/gpu: remove asserts for output/offload from same slave 62999 63000 We should have no problem allowing output/offload from the same slave, 63001 I asserted here, but in order to implement reverse optimus this makes 63002 perfect sense. (reverse optimus is intel outputting to nvidia). 63003 63004 Reviewed-by: Keith Packard <keithp@keithp.com> 63005 Signed-off-by: Dave Airlie <airlied@redhat.com> 63006 63007commit 9d26e8eaf5a2d7c3e65670ac20254c60f665c463 63008Author: Dave Airlie <airlied@redhat.com> 63009Date: Wed Jan 9 14:26:35 2013 +1000 63010 63011 randr: report changes when we disconnect a GPU slave 63012 63013 When we disconnect an output/offload slave set the changed bits, 63014 so a later TellChanged can do something. 63015 63016 Then when we remove a GPU slave device, sent change notification 63017 to the protocol screen. 63018 63019 This allows hot unplugged USB devices to disappear in clients. 63020 63021 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63022 Signed-off-by: Dave Airlie <airlied@redhat.com> 63023 63024commit b724324252d13ff95f62eebd12d125b194d2ccc2 63025Author: Dave Airlie <airlied@redhat.com> 63026Date: Wed Jan 9 14:25:43 2013 +1000 63027 63028 randr: only respected changed on the protocol screen 63029 63030 We don't want to know about changes on the non-protocol screen, 63031 we will fix up setchanged to make sure non-protocol screens update 63032 the protocol screens when they have a change. 63033 63034 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63035 Signed-off-by: Dave Airlie <airlied@redhat.com> 63036 63037commit b3f70f38edebac87afe9351538365292f1aaaff3 63038Author: Dave Airlie <airlied@redhat.com> 63039Date: Wed Jan 9 14:29:47 2013 +1000 63040 63041 randr: make SetChanged modify the main protocol screen not the gpu screen 63042 63043 When SetChanged is called we now modify the main protocol screen, 63044 not the the gpu screen. Since changed stuff should work at the protocol level. 63045 63046 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63047 Signed-off-by: Dave Airlie <airlied@redhat.com> 63048 63049commit f9c8248b8326ad01f33f31531c6b2479baf80f02 63050Author: Dave Airlie <airlied@redhat.com> 63051Date: Wed Jan 9 14:23:57 2013 +1000 63052 63053 randr: don't directly set changed bits in randr screen 63054 63055 Introduce a wrapper interface so we can fix things up for multi-gpu 63056 situations later. 63057 63058 This just introduces the API for now. 63059 63060 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63061 Signed-off-by: Dave Airlie <airlied@redhat.com> 63062 63063commit 451ba4bd41b82acd4aec6236ba121e00cfeb311b 63064Author: Chris Wilson <chris@chris-wilson.co.uk> 63065Date: Thu Jan 10 03:26:33 2013 +0000 63066 63067 hw/xfree86: Only report SetDesiredModes() failed if at least one modeset fails 63068 63069 commit 6703a7c7cf1a349c137e247a0c8eb462ff7b07be 63070 Author: Keith Packard <keithp@keithp.com> 63071 Date: Tue Jan 8 20:24:32 2013 -0800 63072 63073 hw/xfree86: Require only one working CRTC to start the server. 63074 63075 changed the logic to try to set the mode on all connected outputs rather 63076 than abort upon the first failure. The return error code was then 63077 tweaked such that it reported success if it set a mode on any crtc. 63078 However, this confuses the headless case where we never enable any crtcs 63079 and also, importantly, never fail to set a crtc. 63080 63081 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=59190 63082 63083 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 63084 Also-written-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63085 Reviewed-by: Keith Packard <keithp@keithp.com> 63086 Signed-off-by: Keith Packard <keithp@keithp.com> 63087 63088commit 9878e097a7de2f86eff0dcfd9fe5d83b162197ec 63089Author: Alan Coopersmith <alan.coopersmith@oracle.com> 63090Date: Wed Apr 24 15:24:31 2013 -0700 63091 63092 Only call xf86platformVTProbe() when it's defined 63093 63094 Fixes build on non-udev systems, since XSERVER_PLATFORM_BUS is only 63095 defined in configure.ac if $CONFIG_UDEV_KMS is true. 63096 63097 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 63098 Reviewed-by: Dave Airlie <airlied@redhat.com> 63099 63100commit e779402d531625ba4ed5d628d06c6b1e06b2ec7b 63101Author: Keith Packard <keithp@keithp.com> 63102Date: Thu Apr 25 09:02:31 2013 -0700 63103 63104 shadow: Define c2p_unsupported 63105 63106 Just make this call BUG_WARN(1) to indicate that something unexpected 63107 happened 63108 63109 Signed-off-by: Keith Packard <keithp@keithp.com> 63110 63111commit 2b361fbda536f0125e4b87a6d455acc58f4e8690 63112Author: Alan Coopersmith <alan.coopersmith@oracle.com> 63113Date: Sun Feb 10 10:24:59 2013 -0800 63114 63115 sparcPromPathname2Node: free name when returning error, instead of leaking it 63116 63117 Reported with other leaks found by cppcheck in bugzilla #50281 63118 https://bugs.freedesktop.org/show_bug.cgi?id=50281 63119 63120 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 63121 Reviewed-by: Keith Packard <keithp@keithp.com> 63122 63123commit 174ccd84931ece5a92a09c4a1d6a47e0958ebf7e 63124Author: Alan Coopersmith <alan.coopersmith@oracle.com> 63125Date: Sun Feb 10 10:24:59 2013 -0800 63126 63127 xf86SbusCmapLoadPalette: Delay malloc until needed, avoiding leak on error 63128 63129 Reported with other leaks found by cppcheck in bugzilla #50281 63130 https://bugs.freedesktop.org/show_bug.cgi?id=50281 63131 63132 V2: check for malloc failure 63133 63134 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 63135 63136commit 6bca0184d167388cd417d113031317990489987d 63137Author: Alan Coopersmith <alan.coopersmith@oracle.com> 63138Date: Sun Feb 10 10:18:02 2013 -0800 63139 63140 dmxVDLRead: if we opened a file, close it instead of leaking it 63141 63142 Reported with other leaks found by cppcheck in bugzilla #50281 63143 https://bugs.freedesktop.org/show_bug.cgi?id=50281 63144 63145 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 63146 Reviewed-by: Keith Packard <keithp@keithp.com> 63147 63148commit 8c5ff2e93f73b24adff741b25fc1e31b2f5dd0ac 63149Author: Alan Coopersmith <alan.coopersmith@oracle.com> 63150Date: Sun Feb 10 10:24:59 2013 -0800 63151 63152 __glXDRIscreenProbe: free screen when DRI2Connect fails, instead of leaking it 63153 63154 Reported with other leaks found by cppcheck in bugzilla #50281 63155 https://bugs.freedesktop.org/show_bug.cgi?id=50281 63156 63157 V2: goto existing error handler, instead of replicating more of it here 63158 63159 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 63160 Reviewed-by: Keith Packard <keithp@keithp.com> 63161 63162commit 9567fa9a7269765bc59ac71b931e16df261d794d 63163Author: Alan Coopersmith <alan.coopersmith@oracle.com> 63164Date: Sun Feb 10 10:24:59 2013 -0800 63165 63166 __glXDRIscreenCreateContext: free context on failure, instead of leaking it 63167 63168 Reported with other leaks found by cppcheck in bugzilla #50281 63169 https://bugs.freedesktop.org/show_bug.cgi?id=50281 63170 63171 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 63172 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 63173 63174commit 7ab98bafc9a3426fd40f8ae693430491333ba4fc 63175Merge: 53da26afb 5ab260317 63176Author: Keith Packard <keithp@keithp.com> 63177Date: Wed Apr 24 14:14:45 2013 -0700 63178 63179 Merge remote-tracking branch 'geertu/master' 63180 63181commit 53da26afb76bcce70b91c5df23d25e7e85c010cf 63182Merge: 5ece86e92 d8d3c78b6 63183Author: Keith Packard <keithp@keithp.com> 63184Date: Wed Apr 24 10:27:19 2013 -0700 63185 63186 Merge remote-tracking branch 'whot/for-keith' 63187 63188commit 5ece86e921a1df457a9baab6a2357ccda4ad6957 63189Merge: ef0a726bc 22cab8a28 63190Author: Keith Packard <keithp@keithp.com> 63191Date: Wed Apr 24 10:23:51 2013 -0700 63192 63193 Merge remote-tracking branch 'airlied/for-keithp-gpu-vt-owner' 63194 63195commit ef0a726bc8d5c415244cb4f86d664938b0127a5d 63196Author: Jeremy White <jwhite@codeweavers.com> 63197Date: Thu Mar 21 10:58:42 2013 -0500 63198 63199 Eliminate the use of xf86Rename.h 63200 63201 Signed-off-by: Jeremy White <jwhite@codeweavers.com> 63202 Reviewed-by: Keith Packard <keithp@keithp.com> 63203 Signed-off-by: Keith Packard <keithp@keithp.com> 63204 63205commit 2b539239e465faa9e8c6d96f1c6f00e593d46c18 63206Author: Keith Packard <keithp@keithp.com> 63207Date: Wed Apr 24 10:11:03 2013 -0700 63208 63209 Bump to 1.14.99.1 63210 63211 Snapshot with fix for CVE-2013-1940. 63212 63213 Signed-off-by: Keith Packard <keithp@keithp.com> 63214 63215commit d8d3c78b6ebdf63836951d2dbd2d3d621ff26da2 63216Author: Andreas Müller <schnitzeltony@googlemail.com> 63217Date: Tue Apr 16 14:30:43 2013 +0200 63218 63219 dixstruct.h: fix segfaults - char is unsigned for ARM and PowerPC architectures 63220 63221 see ARM related bug reports [1-3] 63222 63223 [1] https://github.com/archlinuxarm/PKGBUILDs/issues/446I 63224 [2] http://www.raspberrypi.org/phpBB3/viewtopic.php?t=38568&p=321673 63225 [3] http://lists.linuxtogo.org/pipermail/openembedded-core/2013-April/037805.html 63226 63227 Signed-off-by: Andreas Müller <schnitzeltony@googlemail.com> 63228 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 63229 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63230 63231commit 7d722796c678532e8c5897c673c43184da353f44 63232Author: Peter Hutterer <peter.hutterer@who-t.net> 63233Date: Wed Apr 17 19:47:42 2013 +1000 63234 63235 dix: plug memory leak in freeing TouchClass 63236 63237 ==15562== 1,800 bytes in 1 blocks are definitely lost in loss record 298 of 330 63238 ==15562== at 0x4A06B6F: calloc (vg_replace_malloc.c:593) 63239 ==15562== by 0x4312C7: InitTouchClassDeviceStruct (devices.c:1644) 63240 63241 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63242 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 63243 63244commit 5ab260317ad3b2aafff31a97df21620db52eacd1 63245Author: Geert Uytterhoeven <geert@linux-m68k.org> 63246Date: Sun Mar 10 21:04:19 2013 +0100 63247 63248 Xfbdev: Wire up Amiga afb4 and afb8 support 63249 63250 Add support for Amiga-style bitplanes, with 4 or 8 bits per pixel. 63251 63252 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63253 Acked-by: Keith Packard <keithp@keithp.com> 63254 63255commit 672bc5bb38918304cf68114a1112cd48651a5e83 63256Author: Geert Uytterhoeven <geert@linux-m68k.org> 63257Date: Sun Mar 10 17:17:26 2013 +0100 63258 63259 Xfbdev: Wire up Atari iplan2p4 and iplan2p8 support 63260 63261 Add support for Atari-style interleaved bitplanes, with 2 bytes interleave 63262 and 4 or 8 bits per pixel. 63263 63264 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63265 Reviewed-by: Keith Packard <keithp@keithp.com> 63266 63267commit 95a3c7536c1a4afe97aaf955980034cc69af9c2f 63268Author: Geert Uytterhoeven <geert@linux-m68k.org> 63269Date: Sun Mar 10 17:17:26 2013 +0100 63270 63271 Xfbdev: Force shadowfb for frame buffers with non-packed pixels 63272 63273 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63274 Reviewed-by: Keith Packard <keithp@keithp.com> 63275 63276commit 0e808110df216649e05503baecd06cd5a3e50421 63277Author: Geert Uytterhoeven <geert@linux-m68k.org> 63278Date: Sun Mar 10 17:17:26 2013 +0100 63279 63280 Xfbdev: Reject unsupported frame buffer types 63281 63282 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63283 Reviewed-by: Keith Packard <keithp@keithp.com> 63284 63285commit cfd10576812c36f5844805eb95ed1f2d093d1691 63286Author: Geert Uytterhoeven <geert@linux-m68k.org> 63287Date: Sun Mar 10 21:03:44 2013 +0100 63288 63289 Shadow: Add support for Amiga afb8 63290 63291 Add support for Amiga-style bitplanes, with 8 bits per pixel. 63292 63293 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63294 Acked-by: Keith Packard <keithp@keithp.com> 63295 63296commit a1b8e7f1e6118b611ba9d332b8763ee2b44f550c 63297Author: Geert Uytterhoeven <geert@linux-m68k.org> 63298Date: Wed Mar 27 09:37:02 2013 +0100 63299 63300 Shadow: Add support for Amiga afb4 63301 63302 Add support for Amiga-style bitplanes, with 4 bits per pixel. 63303 63304 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63305 Acked-by: Keith Packard <keithp@keithp.com> 63306 63307commit d7181e567d1629c387b834da7eecdf618d14718e 63308Author: Geert Uytterhoeven <geert@linux-m68k.org> 63309Date: Sun Mar 10 17:16:17 2013 +0100 63310 63311 Shadow: Add support for Atari iplan2p8 63312 63313 Add support for Atari-style interleaved bitplanes, with 2 bytes interleave 63314 and 8 bits per pixel. 63315 63316 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63317 Acked-by: Keith Packard <keithp@keithp.com> 63318 63319commit 3f7506b0fca72b2462b85bc4f613de809ae95859 63320Author: Geert Uytterhoeven <geert@linux-m68k.org> 63321Date: Sun Mar 10 21:01:02 2013 +0100 63322 63323 Shadow: Add support for Atari iplan2p4 63324 63325 Add support for Atari-style interleaved bitplanes, with 2 bytes interleave 63326 and 4 bits per pixel. 63327 63328 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63329 Acked-by: Keith Packard <keithp@keithp.com> 63330 63331commit 87af9ab7d52578b0ea315fc4f8dfd906d19755fb 63332Author: Geert Uytterhoeven <geert@linux-m68k.org> 63333Date: Sun Mar 10 17:15:05 2013 +0100 63334 63335 Shadow: Add c2p core 63336 63337 Add Chunky-to-Planar core functionality, to be used by the Atari and Amiga 63338 (interleaved) bitplanes code. 63339 63340 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63341 Acked-by: Keith Packard <keithp@keithp.com> 63342 63343commit 4ee2566c3e420a717be36f79126a14e15edac30f 63344Author: Geert Uytterhoeven <geert@linux-m68k.org> 63345Date: Tue Mar 12 18:10:40 2013 +0100 63346 63347 Xfbdev: Treat 1 bpp pseudocolor as monochrome 63348 63349 miCreateDefColormap() only preallocates black and white pixels if 63350 depth > 1. 63351 Hence override the visual, so fbdevCreateColormap() takes care of it. 63352 63353 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63354 Reviewed-by: Keith Packard <keithp@keithp.com> 63355 63356commit 6dfb94a891ed445f47832ba1364fcbf1de018f4a 63357Author: Geert Uytterhoeven <geert@linux-m68k.org> 63358Date: Tue Mar 12 14:16:10 2013 +0100 63359 63360 Xfbdev: Add support for monochrome visuals 63361 63362 Monochrome supports StaticGray, with hardcoded black and white pixels. 63363 63364 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63365 Reviewed-by: Keith Packard <keithp@keithp.com> 63366 63367commit 1049b32166760bdc00106625e213d31a8fc60bad 63368Author: Geert Uytterhoeven <geert@linux-m68k.org> 63369Date: Tue Mar 12 14:19:25 2013 +0100 63370 63371 Xfbdev: Handle unset fix.line_length 63372 63373 Older frame buffer devices may not fill in fix.line_length, in which 63374 case it must be calculated by the application. 63375 63376 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63377 63378commit 5c509c360d97bea78ef461596ab4b7bde69020ac 63379Author: Geert Uytterhoeven <geert@linux-m68k.org> 63380Date: Fri Feb 22 13:21:28 2013 +0100 63381 63382 Xfbdev: Make char *fbdevDevicePath const 63383 63384 This fixes: 63385 63386 hw/kdrive/fbdev/fbdev.c: In function 'fbdevInitialize': 63387 hw/kdrive/fbdev/fbdev.c:41:25: warning: assignment discards 'const' qualifier from pointer target type [enabled by default] 63388 63389 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63390 Reviewed-by: Keith Packard <keithp@keithp.com> 63391 63392commit b08afbc53c1d583c3913b92e67db44823077b112 63393Author: Geert Uytterhoeven <geert@linux-m68k.org> 63394Date: Tue Mar 12 15:20:00 2013 +0100 63395 63396 KDrive: Bail out if screen initialization failed 63397 63398 Else we may get a segmentation fault later. 63399 63400 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63401 Reviewed-by: Keith Packard <keithp@keithp.com> 63402 63403commit e7045c9dd208a1afe36526ab21ef15f8b01c8bd2 63404Author: Geert Uytterhoeven <geert@linux-m68k.org> 63405Date: Sat Feb 23 20:09:22 2013 +0100 63406 63407 test/input: Fix double-aligned test in dix_valuator_alloc() on m68k 63408 63409 On m68k, doubles are not 64-bit aligned, just like on i386 and sh. 63410 63411 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63412 63413commit 8bb0616cc2799c2c34448da6337ea8cb9438cda7 63414Author: Geert Uytterhoeven <geert@linux-m68k.org> 63415Date: Fri Feb 22 21:10:07 2013 +0100 63416 63417 miext/shadow/shpacked.c: Remove unused PickBit() define 63418 63419 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63420 Reviewed-by: Keith Packard <keithp@keithp.com> 63421 63422commit 6ca03b9161d33b1d2b55a3a1a913cf88deb2343f 63423Author: Dave Airlie <airlied@gmail.com> 63424Date: Wed Apr 10 16:09:01 2013 +1000 63425 63426 xf86: fix flush input to work with Linux evdev devices. 63427 63428 So when we VT switch back and attempt to flush the input devices, 63429 we don't succeed because evdev won't return part of an event, 63430 since we were only asking for 4 bytes, we'd only get -EINVAL back. 63431 63432 This could later cause events to be flushed that we shouldn't have 63433 gotten. 63434 63435 This is a fix for CVE-2013-1940. 63436 63437 Signed-off-by: Dave Airlie <airlied@redhat.com> 63438 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63439 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63440 63441commit ddc149beaf708e84fb5573c4322d4f859c517917 63442Author: Peter Hutterer <peter.hutterer@who-t.net> 63443Date: Mon Feb 27 18:50:46 2012 +1000 63444 63445 Stop the shouting 63446 63447 Meanwhile, here in the future lowercase letters have been invented. 63448 63449 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63450 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 63451 63452commit 27356a45b4f6bad4d0bbf356e6d976d0c8dd257b 63453Author: Peter Hutterer <peter.hutterer@who-t.net> 63454Date: Mon Feb 27 17:22:16 2012 +1000 63455 63456 dix: only allocate unused classes for master devices 63457 63458 Slave devices don't need these and the matching code in CloseDevice() has a 63459 IsMaster() condition on freeing these, causing a leak. 63460 63461 ==16111== 384 bytes in 4 blocks are definitely lost in loss record 72 of 105 63462 ==16111== at 0x4C28BB4: calloc (vg_replace_malloc.c:467) 63463 ==16111== by 0x42AEE2: AllocDevicePair (devices.c:2707) 63464 ==16111== by 0x4BAA27: AllocXTestDevice (xtest.c:617) 63465 ==16111== by 0x4BA89A: InitXTestDevices (xtest.c:570) 63466 ==16111== by 0x425F5E: InitCoreDevices (devices.c:690) 63467 ==16111== by 0x5ACB2D: main (main.c:257) 63468 63469 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63470 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 63471 63472commit 98b94c36d6b1d286bbd4cb414e54b4b95a1484b0 63473Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63474Date: Mon Apr 15 16:53:48 2013 +0200 63475 63476 dix: copy event in TouchConvertToPointerEvent correctly 63477 63478 Fixes reading random memory read beyond the end of original event. 63479 63480 sizeof device_event: 424 63481 sizeof internal_event: 2800 63482 63483 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63484 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63485 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63486 63487commit 4bff442ec5aa4b93a3f5c11782d4b7b9d1ae13ac 63488Author: Knut Petersen <Knut_Petersen@t-online.de> 63489Date: Tue Feb 26 07:52:59 2013 +0100 63490 63491 Never try to execute BellProcPtr NULL. 63492 63493 This prevents xts XI/XDeviceBell-2 test 63494 from segfaulting the server. 63495 63496 Signed-off-by: Knut Petersen <Knut_Petersen@t-online.de> 63497 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63498 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63499 63500commit 9003399708936481083424b4ff8f18a16b88b7b3 63501Author: Rui Matos <tiagomatos@gmail.com> 63502Date: Sat Apr 13 04:22:54 2013 +0200 63503 63504 Xi: fix swapped grab mode args for keyboard devices in XIGrabDevice 63505 63506 The protocol says that the grab_mode argument applies to the device 63507 being grabbed and paired_device_mode to the paired master 63508 device. GrabDevice() however takes in a pointer mode and a keyboard 63509 mode and so we have to swap the values according the type of device 63510 being grabbed. 63511 63512 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 63513 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63514 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63515 63516commit d16284687ddeffde32a561b8c67b2e2e3d66aeaa 63517Author: Bryce Harrington <bryce@canonical.com> 63518Date: Wed Apr 10 23:05:40 2013 -0700 63519 63520 xfree86: Revert workaround for drm race condition. 63521 63522 Revert 70739e817b2d64bc020ea491f23a3574bdb6155e and mostly revert 63523 c31eac647a9ecf0fb20dc98266cadf0ba923ba14. 63524 63525 Further investigation shows the encountered race condition is between 63526 lightdm and plymouth-splash, as implemented in the Ubuntu distribution 63527 within the limitations of upstart's job coordination logic, and can (and 63528 should) be fixed within those limiations. Not in xserver itself. 63529 63530 This leaves some of the diagnostic improvements from the recent patch 63531 series, in case others run into a similar situation. 63532 63533 Signed-off-by: Bryce Harrington <bryce@canonical.com> 63534 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63535 63536commit 131f883f85b9b734e5e5652f16ba1d3b5f4de12f 63537Author: Peter Hutterer <peter.hutterer@who-t.net> 63538Date: Fri Apr 12 09:14:53 2013 +1000 63539 63540 xfree86: change a log message 63541 63542 This path is technically executed through config/udev, but having two 63543 messages in the form "config/udev: Adding drm device" makes it appear as if 63544 the udev filters are wrong and it's trying to add the same device twice. In 63545 fact, it's only one device, only added once, but a duplicate log message. 63546 63547 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63548 Reviewed-by: Dave Airlie <airlied@redhat.com> 63549 63550commit 22cab8a28a433d03a4e6ba97f9a160271d73cb52 63551Author: Dave Airlie <airlied@gmail.com> 63552Date: Wed Apr 10 16:32:15 2013 +1000 63553 63554 xf86: don't hotplug output devices while VT switched. 63555 63556 We don't want to hotplug output devices while we are VT switched, 63557 as we get races between multiple X servers on the device open, and 63558 drm device master status. This just queues device opens until we return 63559 from VT switch. 63560 63561 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 63562 Signed-off-by: Dave Airlie <airlied@redhat.com> 63563 63564commit 5b359cf6135ca173d8f65cb92926332f07f91efe 63565Author: Dave Airlie <airlied@gmail.com> 63566Date: Wed Apr 10 16:32:11 2013 +1000 63567 63568 xf86: use new xf86VTOwner interface in a few places 63569 63570 This replaces some previous uses of direct xf86Screens[0] accesses. 63571 63572 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 63573 Signed-off-by: Dave Airlie <airlied@gmail.com> 63574 63575commit d61ea1f64db45201c1a2b39c39293c5768d98092 63576Author: Dave Airlie <airlied@gmail.com> 63577Date: Wed Apr 10 16:32:02 2013 +1000 63578 63579 xfree86: add VT owner interface 63580 63581 This is just a simple interface to avoid accessing x86Screens[0] 63582 directly. 63583 63584 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 63585 Signed-off-by: Dave Airlie <airlied@gmail.com> 63586 63587commit 7347f39f94d8cebbf73ce1a2f94d1abdaf7ff383 63588Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63589Date: Tue Apr 9 11:19:07 2013 +0200 63590 63591 Xi: Do not handle ET_TouchOwnership in ProcessTouchEvent 63592 63593 The event struct is different, causing memory corruption on 1.13 and 1.14, 63594 63595 as can be witnessed in https://bugs.freedesktop.org/show_bug.cgi?id=56578 63596 63597 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63598 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63599 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63600 63601commit dbba50a1280cbda9ecff6f37884b4c5756c30bab 63602Author: Keith Packard <keithp@keithp.com> 63603Date: Tue Apr 9 16:23:19 2013 -0700 63604 63605 Xi: Use correct destination when swapping barrier events 63606 63607 Write the swapped values to the destination rather than the source. 63608 63609 Signed-off-by: Keith Packard <keithp@keithp.com> 63610 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 63611 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63612 63613commit b86b3d10bb2fee1a922b8831e8bb415c339f3d99 63614Author: Peter Hutterer <peter.hutterer@who-t.net> 63615Date: Mon Mar 4 07:58:41 2013 +1000 63616 63617 dix: don't set non-exisiting flags on touch events 63618 63619 Unlike pointer/keyboard events, the flags field for ET_Touch* is a set of 63620 server-internal defines that we need to convert to XI protocol defines. 63621 Currently only two of those defines actually translate to the protocol, so 63622 make sure we don't send internal garbage down the wire. 63623 63624 No effect to current clients since they shouldn't look at undefined bits 63625 anyway. 63626 63627 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63628 63629commit ecf62755086fd65898998d5a509aee5f29a9583d 63630Author: Jeremy White <jwhite@codeweavers.com> 63631Date: Thu Mar 21 10:58:57 2013 -0500 63632 63633 Define prototypes for hw/xfree86/modes/xf86Modes.c only in xf86Modes.h. 63634 63635 This removes a large number of redundant declaration warnings. 63636 63637 Signed-off-by: Jeremy White <jwhite@codeweavers.com> 63638 Reviewed-by: Robert Morell <rmorell@nvidia.com> 63639 Signed-off-by: Keith Packard <keithp@keithp.com> 63640 63641commit e13f29984251dbbe6076264ed710c2922312eed1 63642Author: Bryce Harrington <bryce@canonical.com> 63643Date: Tue Mar 19 12:12:46 2013 -0700 63644 63645 xfree86: Be verbose if waiting on opening the drm device 63646 63647 Signed-off-by: Bryce Harrington <bryce@canonical.com> 63648 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 63649 Signed-off-by: Keith Packard <keithp@keithp.com> 63650 63651commit 70739e817b2d64bc020ea491f23a3574bdb6155e 63652Author: Bryce Harrington <bryce@canonical.com> 63653Date: Tue Mar 19 12:12:45 2013 -0700 63654 63655 xfree86: Fix race condition failure opening drm. 63656 63657 If other processes have had drm open previously, xserver may attempt to 63658 open the device too early and fail, with xserver error exit "Cannot 63659 run in framebuffer mode" or Xorg.0.log messages about "setversion 1.4 63660 failed". 63661 63662 In this situation, we're receiving back -EACCES from libdrm. To address 63663 this we need to re-set ourselves as the drm master, and keep trying to 63664 set the interface until it works (or until we give up). 63665 63666 See https://bugs.launchpad.net/ubuntu/+source/libdrm/+bug/982889 63667 63668 Signed-off-by: Bryce Harrington <bryce@canonical.com> 63669 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 63670 Signed-off-by: Keith Packard <keithp@keithp.com> 63671 63672commit c31eac647a9ecf0fb20dc98266cadf0ba923ba14 63673Author: Bryce Harrington <bryce@canonical.com> 63674Date: Tue Mar 19 12:12:44 2013 -0700 63675 63676 xfree86: Keep trying to set interface on drm for 2 seconds. 63677 63678 And if we've had to delay booting due to not being able to set the 63679 interface, fess up. 63680 63681 Signed-off-by: Bryce Harrington <bryce@canonical.com> 63682 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 63683 Signed-off-by: Keith Packard <keithp@keithp.com> 63684 63685commit d1cc210de8c13f2db9f6f284ecc652305c28801e 63686Author: Bryce Harrington <bryce@canonical.com> 63687Date: Tue Mar 19 12:12:43 2013 -0700 63688 63689 xfree86: Provide more details on failure 63690 63691 Signed-off-by: Bryce Harrington <bryce@canonical.com> 63692 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 63693 Signed-off-by: Keith Packard <keithp@keithp.com> 63694 63695commit f059d0dabc553a5f748d86de9115da00be5997d5 63696Author: Bryce Harrington <bryce@canonical.com> 63697Date: Tue Mar 19 12:12:42 2013 -0700 63698 63699 xfree86: Track error code and add label for error handling. 63700 63701 Signed-off-by: Bryce Harrington <bryce@canonical.com> 63702 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 63703 Signed-off-by: Keith Packard <keithp@keithp.com> 63704 63705commit 4d7052bd7bbf49b573dc4d34ad14e7f058a0d884 63706Author: Bryce Harrington <bryce@canonical.com> 63707Date: Tue Mar 19 12:12:41 2013 -0700 63708 63709 xfree86: (Cleanup) Close fd if drm interface 1.4 could not be set. 63710 63711 Signed-off-by: Bryce Harrington <bryce@canonical.com> 63712 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 63713 Signed-off-by: Keith Packard <keithp@keithp.com> 63714 63715commit c64fa9a2961f82719cc7734a4222f31297a2e593 63716Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63717Date: Wed Apr 3 12:08:44 2013 +0200 63718 63719 support 32 bpp pixmaps when 24 bpp fb is used. 63720 63721 Fixes background corruption in ubuntu. 63722 63723 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63724 63725commit 8928f8fa0bb154ce437af703ff702016f0dcf127 63726Author: Geert Uytterhoeven <geert@linux-m68k.org> 63727Date: Mon Mar 18 21:38:10 2013 +0100 63728 63729 kdrive/fbdev: revive randr new screen size logic 63730 63731 hw/kdrive/fbdev/fbdev.c: In function 'fbdevRandRSetConfig': 63732 hw/kdrive/fbdev/fbdev.c:470:19: warning: variable 'newheight' set but not used [-Wunused-but-set-variable] 63733 hw/kdrive/fbdev/fbdev.c:470:9: warning: variable 'newwidth' set but not used [-Wunused-but-set-variable] 63734 63735 Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org> 63736 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63737 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63738 63739commit 808c87bff710263f5a7b3c12a7e7fec54672fecd 63740Author: Peter Hutterer <peter.hutterer@who-t.net> 63741Date: Mon Mar 25 14:51:58 2013 +1000 63742 63743 Revert "kdrive: fix "set but not used" warnings" 63744 63745 Partial revert of commit 4149ee8ec0193acbf3812c7ee2627b93b9a89997, better 63746 fix coming up. 63747 63748 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63749 63750commit 3a9a6c972e4bc88dc3dde9399df3b0cda759cd70 63751Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63752Date: Wed Mar 27 10:33:46 2013 +0100 63753 63754 bump version to 0.7.0 63755 63756commit ac4c2abe985ce6816a3389c41eb23462c6ac2486 63757Author: vdb@picaros.org <vdb@picaros.org> 63758Date: Thu Mar 7 16:07:31 2013 +0100 63759 63760 xserver: add monitor Option "ZoomModes" [v2] 63761 63762 Section "Monitor" 63763 Identifier "a21inch" 63764 Option "PreferredMode" "1600x1200" 63765 Option "ZoomModes" "1600x1200 1280x1024 1280x1024 640x480" 63766 EndSection 63767 63768 The option's effect is to search for and mark once each named mode in 63769 the output modes list. So the specification order is free and the zoom 63770 modes sequence follows the order of the output modes list. All marked 63771 modes are available via the Ctrl+Alt+Keypad-{Plus,Minus} key 63772 combination. 63773 63774 See also http://bugs.freedesktop.org/show_bug.cgi?id=17954. 63775 63776 This option has its use for combined monitor and television setups. 63777 It allows for easy switching between 60 Hz and 50 Hz modes even when a 63778 monitor refuses to display the input signal. 63779 63780 (Includes a few minor changes suggested by Aaron for v2) 63781 63782 Signed-off-by: Servaas Vandenberghe <vdb@picaros.org> 63783 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 63784 Signed-off-by: Keith Packard <keithp@keithp.com> 63785 63786commit 862bc2836835b93f4b2cfd2de6ca344f65484a00 63787Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63788Date: Tue Mar 26 15:19:52 2013 +0100 63789 63790 modesetting: return null for get_modes if output could not be retrieved 63791 63792 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63793 63794commit 6b79a8791d35bbd1a13ebdec1c582f66c559c039 63795Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63796Date: Tue Mar 26 15:19:33 2013 +0100 63797 63798 modesetting: clean up leaks 63799 63800 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 63801 63802commit 2967391c6d35f03121afa8003e0fb94b62495129 63803Author: Peter Hutterer <peter.hutterer@who-t.net> 63804Date: Thu Feb 28 10:43:05 2013 +1000 63805 63806 Xi: add a comment to make a condition a bit clearer 63807 63808 The commit message to 676447190190d8546165e21be242cf16dd69f5ae explains it, 63809 but that doesn't stop the WTF moment when reading the code. 63810 63811 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63812 Reviewed-by: Keith Packard <keithp@keithp.com> 63813 63814commit 2fdde2c40d83695438b6f5615f98bd7ae801b43d 63815Author: Peter Hutterer <peter.hutterer@who-t.net> 63816Date: Mon Feb 25 15:21:32 2013 +1000 63817 63818 dix: update coords for touch events in PlayReleasedEvents 63819 63820 Note: this is only hit for #ifdef PANORAMIX 63821 63822 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63823 Reviewed-by: Keith Packard <keithp@keithp.com> 63824 63825commit 697071ab2b7f8910e01ed74618138538291cd1e8 63826Author: Peter Hutterer <peter.hutterer@who-t.net> 63827Date: Thu Feb 28 10:30:15 2013 +1000 63828 63829 Xi: compress two if statements with the same body 63830 63831 We do the same thing here, compress them into one body. 63832 63833 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63834 Reviewed-by: Keith Packard <keithp@keithp.com> 63835 63836commit 11bead1fa205a1353e6a33c6024c7e8ace80be7c 63837Author: Peter Hutterer <peter.hutterer@who-t.net> 63838Date: Mon Feb 25 11:13:52 2013 +1000 63839 63840 dix: fix a comment 63841 63842 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63843 Reviewed-by: Keith Packard <keithp@keithp.com> 63844 63845commit fc504a44d12d537d4e07f659f1863f200a0272ad 63846Author: Peter Hutterer <peter.hutterer@who-t.net> 63847Date: Fri Mar 1 08:26:06 2013 +1000 63848 63849 Xi: use a temp variable for the new listener 63850 63851 Instead of accessing ti->listener[0] all the time. 63852 63853 No functional changes. 63854 63855 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63856 Reviewed-by: Keith Packard <keithp@keithp.com> 63857 63858commit 9978b57b8d94f061d72a67b99a02b0ba16a11429 63859Author: Peter Hutterer <peter.hutterer@who-t.net> 63860Date: Fri Mar 1 08:43:58 2013 +1000 63861 63862 Xi: return !Success from DeliverTouchEmulatedEvent if we didn't deliver 63863 63864 All callers currently ignore the new value, so this patch has no effect. 63865 Inverse call graph: 63866 63867 DeliverTouchEmulatedEvent 63868 DeliverEmulatedMotionEvent Ignores value 63869 DeliverTouchBeginEvent 63870 DeliverTouchEvent 63871 DeliverTouchEvents Ignores value 63872 DeliverTouchEndEvent 63873 DeliverTouchEvent 63874 DeliverTouchEvents Ignores value 63875 63876 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63877 Reviewed-by: Keith Packard <keithp@keithp.com> 63878 63879commit 1754973206c09a901747e71d4abdf666293da59e 63880Author: Aaron Plattner <aplattner@nvidia.com> 63881Date: Wed Feb 20 17:27:16 2013 -0800 63882 63883 modesetting: match PCI class 3, any subclass 63884 63885 If a device is not primary, the PCI device match fails because the 63886 xf86-video-modesetting driver looks specifically for a PCI class match of 63887 0x30000 with a mask of 0xffffff. This fails to match, for example, a 63888 non-primary Intel VGA device, because it is reported as having a class of 63889 0x38000. 63890 63891 Fix that by ignoring the low 16 bits of the class in the pci_id_match table. 63892 63893 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 63894 Reviewed on IRC by Adam Jackson <ajax@redhat.com> 63895 63896commit cc3d1a5a6120e721a46c67446ba68f5596055633 63897Author: Piotr Dziwinski <piotrdz@gmail.com> 63898Date: Sat Feb 23 13:14:45 2013 +0100 63899 63900 glx: fix uninitialized var in __glXDRIscreenProbe 63901 63902 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=59825 63903 63904 Signed-off-by: Piotr Dziwinski <piotrdz@gmail.com> 63905 Reviewed-by: Keith Packard <keithp@keithp.com> 63906 Signed-off-by: Keith Packard <keithp@keithp.com> 63907 63908commit c2ede8f92f0d6f15b1b2a04889cb54cd1728780b 63909Merge: 190b03215 7050aae69 63910Author: Keith Packard <keithp@keithp.com> 63911Date: Mon Mar 18 11:26:24 2013 -0700 63912 63913 Merge remote-tracking branch 'whot/for-keith' 63914 63915commit 190b0321510b99d4738915f540cea5c0c51e07e2 63916Author: Zack Rusin <zackr@vmware.com> 63917Date: Thu Feb 14 15:06:57 2013 -0800 63918 63919 GLX/DRI2: Do not expose INTEL_swap_event without swap control 63920 63921 Swap events depent on the implementation of ScheduleSwap. By 63922 unconditionally enabling GLX_INTEL_swap_event we're breaking 63923 the system with drivers that don't support it because the apps 63924 are forever stuck waiting for an event that will never be 63925 delivered. So lets enable the extension only if the hooks it 63926 depends on are actually there. 63927 63928 Signed-off-by: Zack Rusin <zackr@vmware.com> 63929 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 63930 Signed-off-by: Keith Packard <keithp@keithp.com> 63931 63932commit 116f020102fd6c2a603069a639b113dfa31b48b7 63933Merge: cf89aa537 0f537da72 63934Author: Keith Packard <keithp@keithp.com> 63935Date: Mon Mar 18 11:18:58 2013 -0700 63936 63937 Merge remote-tracking branch 'whot/next' 63938 63939commit cf89aa53748b964f9d9eceaa12a7d6f1a076d1ee 63940Author: Tomasz Lis <tomasz.lis@intel.com> 63941Date: Mon Mar 11 10:21:28 2013 +0100 63942 63943 Full support of sRGB capable fbconfigs. 63944 63945 Changes to correctly initialize the sRGB capability attribute and 63946 transfer it between XServer and the client. Modifications include 63947 extension string, transferring visual config attribs and fbconfig 63948 attribs. Also, attribute is initialized in the modules which do not 63949 really use it (xquartz and xwin). 63950 This version advertises both ARB and EXT strings, and initializes 63951 the capability to default value of FALSE. It has corrected required 63952 GLX version and does not influence swrast. The sRGB capable attribute 63953 is attached only to those configs which do have this capability. 63954 Both ARB and EXT versions share the same GLX extension enabling bit. 63955 63956 Signed-off-by: Tomasz Lis <tomasz.lis@intel.com> 63957 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 63958 Signed-off-by: Keith Packard <keithp@keithp.com> 63959 63960commit 679ccecd8bc6d797e99f3b707f8041b727ca7a49 63961Author: Keith Packard <keithp@keithp.com> 63962Date: Mon Mar 18 10:01:20 2013 -0700 63963 63964 Bump release to 1.14.99.0 63965 63966 Get us off the release number so that we don't conflict with the 63967 stable branch. 63968 63969 Signed-off-by: Keith Packard <keithp@keithp.com> 63970 63971commit beb1715e96b649a0903fa61f30dd1eb69411da2f 63972Author: Colin Walters <walters@verbum.org> 63973Date: Wed Jan 4 22:37:06 2012 +0000 63974 63975 autogen.sh: Implement GNOME Build API 63976 63977 http://people.gnome.org/~walters/docs/build-api.txt 63978 63979 Signed-off-by: Adam Jackson <ajax@redhat.com> 63980 63981commit 7050aae69c2a55dfdbb5c6af7882307e90ba4275 63982Author: Robert Morell <rmorell@nvidia.com> 63983Date: Tue Mar 12 09:40:16 2013 -0700 63984 63985 list.h: Make xorg_list_init inline 63986 63987 Otherwise this file is emitted in every unit that includes it. 63988 63989 Signed-off-by: Robert Morell <rmorell@nvidia.com> 63990 Reviewed-by: Jamey Sharp <jamey@minilop.net> 63991 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 63992 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 63993 63994commit 31595b528645a7e4903eb81da0fc332d78407f25 63995Author: Robert Morell <rmorell@nvidia.com> 63996Date: Tue Mar 12 09:37:43 2013 -0700 63997 63998 configure.ac: Require inputproto 2.3 63999 64000 This picks up support for Xi pointer barriers in the protocol. 64001 64002 Signed-off-by: Robert Morell <rmorell@nvidia.com> 64003 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64004 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64005 64006commit 3ac2e61705432951f9e1b96b42b4214e7f748f94 64007Author: Peter Hutterer <peter.hutterer@who-t.net> 64008Date: Sat Mar 9 17:12:53 2013 +1000 64009 64010 xfixes: ifdef PanoramiXFixes* (#62015) 64011 64012 Fixes build failure with --disable-xinerama introduced by 64013 482e0cb cursor: Move pointer barrier code over to XI 64014 64015 Reason is new include order: sdksyms.sh includes xfixes.h, which previously 64016 did not include xfixesint.h. 64017 64018 As of 482e0cb xfixes.h includes xibarriers.h which includes xfixesint.h 64019 64020 X.Org Bug 62015 <http://bugs.freedesktop.org/show_bug.cgi?id=62015> 64021 64022 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64023 Tested-by: Daniel Martin <consume.noise@gmail.com> 64024 64025commit 5047810a4c20fab444b8c6eb146c55dcdb0d4219 64026Author: Aaron Plattner <aplattner@nvidia.com> 64027Date: Wed Mar 6 09:40:28 2013 -0800 64028 64029 fb: Rename wfbDestroyGlyphCache 64030 64031 Renaming this function was missed in commit 64032 9cbcb5bd6a5360a128d15b77a02d8d3351f74366, so both libfb.so and libwfb.so define 64033 functions named fbDestroyGlyphCache. 64034 64035 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 64036 Reviewed-by: Søren Sandmann <ssp@redhat.com> 64037 Signed-off-by: Keith Packard <keithp@keithp.com> 64038 64039commit 103b77c59e3638a45179bf6d7908f5c738d2d872 64040Author: Keith Packard <keithp@keithp.com> 64041Date: Tue Mar 5 22:31:17 2013 -0800 64042 64043 Version bumped to 1.14 64044 64045 Signed-off-by: Keith Packard <keithp@keithp.com> 64046 64047commit 0f537da72d414ed84e3cd14e3bb7e08565136bd7 64048Author: Andreas Wettstein <wettstein509@solnet.ch> 64049Date: Sun Mar 3 20:25:44 2013 +0100 64050 64051 xkb: Fixes to LatchMods/LatchGroup 64052 64053 The main problem this patch addresses is that if a latch is put on 64054 multi-level key with a Latch/Lock/Set, it is possible that after all 64055 keys are released, still base modifiers are set, which typically will 64056 make the keyboard unusable. To see how it happens (without the patch), 64057 assume that key AltGr sets Mod5 when pressed by itself, and latches Mod3 64058 when pressed together with Shift. Now press Shift, then AltGr and 64059 release both keys in reverse order. Mod3 is now latched, and the 64060 LatchMods filter remains active as the second filter. Now press AltGr; 64061 Mod5 base modifier gets set, and the SetMods filter will become active 64062 as the first filter. Release AltGr: First, the SetMods filter will set 64063 clearMods to Mod5, then the LatchMods filter will overwrite clearMods 64064 with Mod3. Result: the Mod5 base modifier will remain set. This 64065 example becomes practically relevant for the revised German standard 64066 layout (DIN 2137-1:2012-06). 64067 64068 Other changes implement the latch behaviour more accurately according to 64069 the specification. For example, releasing a modifier latching key can 64070 at the same time clear a locked modifier, promote another modifier that 64071 is latched to locked, and latch a third modifier. Overall, what the 64072 code does should be straightforward to compare what the XKB protocol 64073 specification demands, see the table in section 6.3. 64074 64075 Finally, releasing a key no longer cancels a latch that has not become 64076 pending yet. In my opinion, the specification is not clear; it speaks 64077 of "operating" a key, which the patch effectivly interprets as "press" 64078 rather than "press or release". From my experience, using the latter 64079 interpretation makes latches on higher levels practically unusable. In 64080 the example given above, one would have to release AltGr always before 64081 Shift to get the Mod3-Latch. The practical relevance of latches on 64082 higher levels is once more given by the revised German standard layout. 64083 64084 Signed-off-by: Andreas Wettstein <wettstein509@solnet.ch> 64085 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64086 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64087 64088commit eadda231091aa6feb68207ee22f6bc3a390d4556 64089Author: Peter Hutterer <peter.hutterer@who-t.net> 64090Date: Fri Feb 15 11:19:09 2013 +1000 64091 64092 xephyr: fix "set but not used warnings" 64093 64094 ephyrvideo.c: In function 'ephyrPutVideo': 64095 ephyrvideo.c:1009:42: warning: variable 'drw_h' set but not used 64096 ephyrvideo.c:1009:31: warning: variable 'drw_w' set but not used 64097 ephyrvideo.c:1009:20: warning: variable 'drw_y' set but not used 64098 ephyrvideo.c:1009:9: warning: variable 'drw_x' set but not used 64099 ephyrvideo.c: In function 'ephyrGetVideo': 64100 ephyrvideo.c:1058:42: warning: variable 'drw_h' set but not used 64101 ephyrvideo.c:1058:31: warning: variable 'drw_w' set but not used 64102 ephyrvideo.c:1058:20: warning: variable 'drw_y' set but not used 64103 ephyrvideo.c:1058:9: warning: variable 'drw_x' set but not used 64104 ephyrvideo.c: In function 'ephyrPutStill': 64105 ephyrvideo.c:1107:42: warning: variable 'drw_h' set but not used 64106 ephyrvideo.c:1107:31: warning: variable 'drw_w' set but not used 64107 ephyrvideo.c:1107:20: warning: variable 'drw_y' set but not used 64108 ephyrvideo.c:1107:9: warning: variable 'drw_x' set but not used 64109 ephyrvideo.c: In function 'ephyrGetStill': 64110 ephyrvideo.c:1156:42: warning: variable 'drw_h' set but not used 64111 ephyrvideo.c:1156:31: warning: variable 'drw_w' set but not used 64112 ephyrvideo.c:1156:20: warning: variable 'drw_y' set but not used 64113 ephyrvideo.c:1156:9: warning: variable 'drw_x' set but not used 64114 64115 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64116 Reviewed-by: Robert Morell <rmorell@nvidia.com> 64117 64118commit 4149ee8ec0193acbf3812c7ee2627b93b9a89997 64119Author: Peter Hutterer <peter.hutterer@who-t.net> 64120Date: Fri Feb 15 11:19:08 2013 +1000 64121 64122 kdrive: fix "set but not used" warnings 64123 64124 kinput.c: In function 'KdEnqueueKeyboardEvent': 64125 kinput.c:1845:16: warning: variable 'ctrl' set but not used 64126 kinput.c:1844:17: warning: variable 'keyc' set but not used 64127 64128 kinput.c: In function 'KdEnqueuePointerEvent': 64129 kinput.c:1887:12: warning: variable 'ms' set but not used 64130 64131 kxv.c: In function 'KdXVDisable': 64132 kxv.c:1181:19: warning: variable 'ScreenPriv' set but not used 64133 64134 mouse.c: In function 'ps2SkipInit': 64135 mouse.c:444:9: warning: variable 'skipping' set but not used 64136 mouse.c: In function 'ps2Init': 64137 mouse.c:473:10: warning: variable 'waiting' set but not used 64138 mouse.c:472:9: warning: variable 'skipping' set but not used 64139 64140 fbdev.c: In function 'fbdevRandRSetConfig': 64141 fbdev.c:468:19: warning: variable 'newheight' set but not used 64142 fbdev.c:468:9: warning: variable 'newwidth' set but not used 64143 64144 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64145 Reviewed-by: Robert Morell <rmorell@nvidia.com> 64146 64147commit 44fc062f85df7288c17d2d64b73aa4957b91fd6d 64148Author: Peter Hutterer <peter.hutterer@who-t.net> 64149Date: Mon Feb 18 14:57:58 2013 +1000 64150 64151 os: document pnprintf as sigsafe snprintf 64152 64153 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64154 64155commit c100211034ab69ce453a1644fb61c6808d7e3eda 64156Author: Peter Hutterer <peter.hutterer@who-t.net> 64157Date: Tue Dec 18 14:12:40 2012 +1000 64158 64159 dix: only show the cursor if a window defines one (#58398) 64160 64161 e02f864fdf "Suppress cursor display until the first XDefineCursor() request" 64162 disabled cursor display a priori unless -retro is given. 64163 64164 On a plain server, caling XFixesHideCursor() and XFixesShowCursor() would 64165 show the default root cursor, despite no client actually defining a cursor. 64166 64167 Change the logic, disable CursorVisible by default and only enable it from 64168 the window's CWCursor logic. If no window ever defines a cursor, said cursor 64169 stays invisible. 64170 64171 X.Org Bug 58398 <http://bugs.freedesktop.org/show_bug.cgi?id=58398> 64172 64173 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64174 Tested-by: Bastien Nocera <hadess@hadess.net> 64175 Reviewed-by: Daniel Martin <consume.noise@gmail.com> 64176 64177commit 6238bd68bd71323f8b4f1808f34dabe2ae447fe3 64178Author: Aaron Plattner <aplattner@nvidia.com> 64179Date: Tue Mar 5 13:04:46 2013 -0800 64180 64181 DPMS: include GPU screens in DPMS code 64182 64183 Otherwise, displays driven by GPU screens remain on all the time. 64184 64185 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 64186 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 64187 Signed-off-by: Keith Packard <keithp@keithp.com> 64188 64189commit dd4ab8b572956f9457616869d6e383c8ed964c73 64190Author: Peter Hutterer <peter.hutterer@who-t.net> 64191Date: Tue Mar 5 10:14:29 2013 +1000 64192 64193 Xi: force dtime to 0 on the first BarrierHit 64194 64195 dtime to the previous event is 0 on the first BarrierHit event. 64196 64197 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64198 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 64199 Signed-off-by: Keith Packard <keithp@keithp.com> 64200 64201commit 604169af8b67afc74a292cdb9070a3a1f2d7c536 64202Merge: 8f4640bdb 6ea59dc2d 64203Author: Keith Packard <keithp@keithp.com> 64204Date: Mon Mar 4 21:09:59 2013 -0800 64205 64206 Merge commit '6ea59dc2d8887102bfd8d7c838d2e7ab17645aec' 64207 64208commit 6ea59dc2d8887102bfd8d7c838d2e7ab17645aec 64209Author: Jasper St. Pierre <jstpierre@mecheye.net> 64210Date: Sun Mar 3 04:50:55 2013 -0500 64211 64212 xibarriers: Remove accidental use of the comma operator 64213 64214 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64215 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64216 64217commit bd58ebe4cf3b0ce60f87fb26a3715f774dabd349 64218Author: Daniel Martin <consume.noise@gmail.com> 64219Date: Thu Dec 20 13:50:17 2012 +0100 64220 64221 ephyr: Fix crash on 24bpp host framebuffer 64222 64223 Use bytes_per_line and bits_per_pixel from the created XImage to fix 64224 https://bugzilla.redhat.com/show_bug.cgi?id=518960 64225 64226 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 64227 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64228 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64229 64230commit 8f4640bdb9d3988148e09a08d2c7e3bab1d538d6 64231Author: Dave Airlie <airlied@redhat.com> 64232Date: Wed Jan 9 12:58:28 2013 +1000 64233 64234 randr: cleanup provider properly 64235 64236 So in the cold plug server shutdown case, we reap the resources 64237 before we call CloseScreen handlers, so the config->randr_provider 64238 is a dangling pointer when the xf86CrtcCloseScreen handler is called, 64239 64240 however in the hot screen unplug case, we can't rely on automatically 64241 reaped resources, so we need to clean up the provider in the xf86CrtcCloseScreen 64242 case. 64243 64244 This patch provides a cleanup callback from the randr provider removal 64245 into the DDX so it can cleanup properly, this then gets called by the automatic 64246 code for cold plug, or if hot unplug it gets called explicitly. 64247 64248 Fixes a number of random server crashes on shutdown 64249 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=58174 64250 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=891140 64251 64252 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64253 Signed-off-by: Dave Airlie <airlied@redhat.com> 64254 64255commit 3ec35c45ca17f5ed6fd02c50fc49ae7b8d128dcb 64256Author: Dave Airlie <airlied@redhat.com> 64257Date: Wed Jan 9 12:53:14 2013 +1000 64258 64259 xf86: actually set the compat output in the failure case 64260 64261 The previous fix for the previous fix, didn't fully work, 64262 64263 If we don't set compat_output we end up doing derferences 64264 of arrays with -1, leading to valgrind warnings. 64265 64266 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 64267 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64268 Signed-off-by: Dave Airlie <airlied@redhat.com> 64269 64270commit da8ee26023fc2868fe970471195a5f3c86fb574b 64271Author: Dave Airlie <airlied@redhat.com> 64272Date: Wed Jan 9 12:51:45 2013 +1000 64273 64274 xfree86/hotplug: cleanup properly if the screen fails to initialise 64275 64276 Due to another bug, the modesetting/udl driver would fail to init properly 64277 on hotplug, when it did the code didn't clean up properly, and on removing 64278 the device the server could crash. 64279 64280 Found in F18 testing. 64281 64282 Signed-off-by: Dave Airlie <airlied@redhat.com> 64283 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 64284 64285commit e21e183059df5975e7086850d1931edb2c1bbd06 64286Author: Marcin Slusarz <marcin.slusarz@gmail.com> 64287Date: Thu Feb 21 09:02:44 2013 +1000 64288 64289 os: use libunwind to generate backtraces 64290 64291 Libunwind generates backtraces much more reliably than glibc's "backtrace". 64292 64293 Before: 64294 0: /opt/xserver/bin/X (0x400000+0x18ce36) [0x58ce36] 64295 1: /opt/xserver/bin/X (xorg_backtrace+0x9) [0x58d119] 64296 2: /opt/xserver/bin/X (0x400000+0x190d69) [0x590d69] 64297 3: /lib64/libpthread.so.0 (0x7fb904268000+0x10a90) [0x7fb904278a90] 64298 4: /lib64/libc.so.6 (ioctl+0x7) [0x7fb902fbf987] 64299 5: /usr/lib64/libdrm.so.2 (drmIoctl+0x28) [0x7fb90405ffa8] 64300 6: /usr/lib64/libdrm.so.2 (drmCommandWrite+0x1b) [0x7fb90406235b] 64301 7: /usr/lib64/libdrm_nouveau.so.2 (nouveau_bo_wait+0x89) [0x7fb902009719] 64302 8: /opt/xserver/lib/xorg/modules/drivers/nouveau_drv.so (0x7fb90220e000+0x76f3) [0x7fb9022156f3] 64303 9: /opt/xserver/lib/xorg/modules/libexa.so (0x7fb9019c7000+0xbae0) [0x7fb9019d2ae0] 64304 10: /opt/xserver/bin/X (0x400000+0x17d2b3) [0x57d2b3] 64305 11: /opt/xserver/bin/X (0x400000+0xc9930) [0x4c9930] 64306 12: /opt/xserver/bin/X (0x400000+0x3a81a) [0x43a81a] 64307 13: /opt/xserver/bin/X (0x400000+0x3d6a1) [0x43d6a1] 64308 14: /opt/xserver/bin/X (0x400000+0x2c2ca) [0x42c2ca] 64309 15: /lib64/libc.so.6 (__libc_start_main+0xf5) [0x7fb902f019b5] 64310 16: /opt/xserver/bin/X (0x400000+0x2c60d) [0x42c60d] 64311 17: ?? [0x0] 64312 64313 After: 64314 0: /opt/xserver/bin/X (OsSigHandler+0x39) [0x590d69] 64315 1: /lib64/libpthread.so.0 (__restore_rt+0x0) [0x7fb904278a8f] 64316 2: /lib64/libc.so.6 (ioctl+0x7) [0x7fb902fbf987] 64317 3: /usr/lib64/libdrm.so.2 (drmIoctl+0x28) [0x7fb90405ffa8] 64318 4: /usr/lib64/libdrm.so.2 (drmCommandWrite+0x1b) [0x7fb90406235b] 64319 5: /usr/lib64/libdrm_nouveau.so.2 (nouveau_bo_wait+0x89) [0x7fb902009719] 64320 6: /opt/xserver/lib/xorg/modules/drivers/nouveau_drv.so (nouveau_exa_download_from_screen+0x1a3) [0x7fb9022156f3] 64321 7: /opt/xserver/lib/xorg/modules/libexa.so (exaGetImage+0x1f0) [0x7fb9019d2ae0] 64322 8: /opt/xserver/bin/X (miSpriteGetImage+0x173) [0x57d2b3] 64323 9: /opt/xserver/bin/X (compGetImage+0xb0) [0x4c9930] 64324 10: /opt/xserver/bin/X (ProcGetImage+0x55a) [0x43a81a] 64325 11: /opt/xserver/bin/X (Dispatch+0x341) [0x43d6a1] 64326 12: /opt/xserver/bin/X (main+0x3ba) [0x42c2ca] 64327 13: /lib64/libc.so.6 (__libc_start_main+0xf5) [0x7fb902f019b5] 64328 14: /opt/xserver/bin/X (_start+0x29) [0x42c60d] 64329 15: ? (?+0x29) [0x29] 64330 64331 Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com> 64332 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64333 Tested-by: Knut Petersen <knut.petersen@t-online.de> 64334 64335commit 75815dbb373d5a74d57cbec9d469f9f88a8ee3ef 64336Author: Chris Wilson <chris@chris-wilson.co.uk> 64337Date: Mon Dec 3 13:25:19 2012 +0000 64338 64339 Add missing GTF modes 64340 64341 A fixed-mode output device like a panel will often only inform of its 64342 preferred mode through its EDID. However, the driver will adjust user 64343 specified modes for display through use of a panel-fitter allowing 64344 greater flexibility in upscaling. This is often used by games to set a 64345 low resolution for performance and use the panel fitter to fill the 64346 screen. 64347 64348 v2: Use the presence of the 'scaling mode' connector property as an 64349 indication that a panel fitter is attached to that pipe. 64350 64351 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=55564 64352 64353commit e27b2e61632b220ddc36d0d0387581a9f4774f21 64354Author: Peter Harris <pharris@opentext.com> 64355Date: Tue Feb 19 16:08:52 2013 -0500 64356 64357 xkb: Set nIndicators in XkbGetIndicatorMap 64358 64359 Xlib doesn't use this value (it computes it from the reply length 64360 instead) which is why nobody has noticed yet. But the spec 64361 http://www.x.org/releases/X11R7.7/doc/kbproto/xkbproto.html 64362 says that it should be set. 64363 64364 Signed-off-by: Peter Harris <pharris@opentext.com> 64365 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64366 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64367 64368commit 3aac7a59dc6ef2d8bbf46ba5d37acdf6013e9450 64369Author: Daniel Martin <daniel.martin@secunet.com> 64370Date: Tue Dec 11 17:23:55 2012 +0100 64371 64372 ephyr: Add -resizeable option 64373 64374 With this option passed, ephyr windows can be resized like normal 64375 windows on the fly, without the need of an explicit parent window. 64376 64377 Signed-off-by: Daniel Martin <daniel.martin@secunet.com> 64378 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64379 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64380 64381commit f9198e278becec158b570204cf9fc1de822ac76b 64382Author: Peter Hutterer <peter.hutterer@who-t.net> 64383Date: Mon Feb 27 16:43:10 2012 +1000 64384 64385 dix: FreeAllAtoms() on reset 64386 64387 ==5712== 6 bytes in 1 blocks are still reachable in loss record 17 of 585 64388 ==5712== at 0x4A074CD: malloc (vg_replace_malloc.c:236) 64389 ==5712== by 0x3D1DE885B1: strndup (strndup.c:46) 64390 ==5712== by 0x41CB71: MakeAtom (atom.c:121) 64391 ==5712== by 0x55AE3E: XIGetKnownProperty (xiproperty.c:401) 64392 ==5712== by 0x4251C9: AddInputDevice (devices.c:312) 64393 ==5712== by 0x42AC0C: AllocDevicePair (devices.c:2657) 64394 ==5712== by 0x425E6E: InitCoreDevices (devices.c:677) 64395 ==5712== by 0x5ACA05: main (main.c:257) 64396 64397 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64398 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 64399 64400commit 88517ced1fa2e621be2d05a319b522b3111da516 64401Merge: 82425c66e 73974dd7e 64402Author: Peter Hutterer <peter.hutterer@who-t.net> 64403Date: Fri Feb 15 11:58:52 2013 +1000 64404 64405 Merge branch 'master' of git+ssh://people.freedesktop.org/~alanc/xserver into next 64406 64407commit 82425c66e7a8c436b5be7ad44880dd4be6c1f06a 64408Author: Peter Hutterer <peter.hutterer@who-t.net> 64409Date: Thu Feb 14 10:21:47 2013 +1000 64410 64411 xfree86: remove redundant declaration of inputInfo 64412 64413 xf86Cursor.c:19:18: warning: redundant redeclaration of 'inputInfo' 64414 [-Wredundant-decls] 64415 In file included from xf86Cursor.c:18:0: 64416 ../../../include/inputstr.h:614:57: note: previous declaration of 64417 'inputInfo' was here 64418 64419 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64420 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64421 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 64422 64423commit 6133c417595a9c8fc158b0d68fd4a7a2c58fdc47 64424Author: Peter Hutterer <peter.hutterer@who-t.net> 64425Date: Thu Feb 14 10:15:33 2013 +1000 64426 64427 xkb: remove unused variable 'names' 64428 64429 xkb.c: In function '_XkbSetNamesCheck': 64430 xkb.c:3987:18: warning: variable 'names' set but not used 64431 [-Wunused-but-set-variable] 64432 64433 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64434 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64435 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 64436 64437commit 63d00c6b146d7f5fb80f20f19066088059042a5e 64438Author: Peter Hutterer <peter.hutterer@who-t.net> 64439Date: Thu Feb 14 10:13:53 2013 +1000 64440 64441 Xext: rename two shadowing variables 64442 64443 panoramiX.c: In function 'PanoramiXCreateConnectionBlock': 64444 panoramiX.c:599:10: warning: declaration of 'disableBackingStore' shadows a 64445 global declaration [-Wshadow] 64446 In file included from ../include/windowstr.h:60:0, 64447 from panoramiX.c:47: 64448 ../include/opaque.h:56:52: warning: shadowed declaration is here [-Wshadow] 64449 64450 panoramiX.c: In function 'PanoramiXConsolidate': 64451 panoramiX.c:834:19: warning: declaration of 'pScreen' shadows a previous 64452 local [-Wshadow] 64453 panoramiX.c:813:15: warning: shadowed declaration is here [-Wshadow] 64454 64455 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64456 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64457 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 64458 64459commit 8bbea9f48f43e54c55c8b60ea36bda3134a86af2 64460Author: Peter Hutterer <peter.hutterer@who-t.net> 64461Date: Thu Feb 14 10:09:53 2013 +1000 64462 64463 Xext: renaming shadowing variable 64464 64465 xvdisp.c: In function 'ProcXvStopVideo': 64466 xvdisp.c:712:11: warning: declaration of 'rc' shadows a previous local 64467 [-Wshadow] 64468 xvdisp.c:705:17: warning: shadowed declaration is here [-Wshadow] 64469 64470 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64471 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64472 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 64473 64474commit 7b79a2e4a11b5c5f0ebaa495828725e235d2b08e 64475Author: Peter Hutterer <peter.hutterer@who-t.net> 64476Date: Thu Feb 14 10:06:55 2013 +1000 64477 64478 fb: drop two unneeded shadowing variables 64479 64480 fbpict.c: In function 'fbGlyphs': 64481 fbpict.c:188:6: warning: declaration of 'x' shadows a previous local 64482 [-Wshadow] 64483 fbpict.c:111:9: warning: shadowed declaration is here [-Wshadow] 64484 fbpict.c:188:9: warning: declaration of 'y' shadows a previous local 64485 [-Wshadow] 64486 fbpict.c:111:12: warning: shadowed declaration is here [-Wshadow] 64487 64488 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64489 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64490 Reviewed-by: Soren Sandmann <ssp@redhat.com> 64491 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 64492 64493commit 858d8b19b3a5bb59e5c5f9a9e68adce2495b5e31 64494Author: Peter Hutterer <peter.hutterer@who-t.net> 64495Date: Thu Feb 14 10:04:06 2013 +1000 64496 64497 xfree86: drop unused prevSIGIO 64498 64499 Unused as of 5d309af2ed93e91c7d72f548a11052051efbb40f 64500 64501 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64502 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64503 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 64504 64505commit 592d35aef0a8089a344543cf5a425e0537c0431b 64506Author: Peter Hutterer <peter.hutterer@who-t.net> 64507Date: Thu Feb 14 10:02:02 2013 +1000 64508 64509 randr: fix "set but unused" warnings 64510 64511 rrcrtc.c: In function 'RRCrtcDetachScanoutPixmap': 64512 rrcrtc.c:366:9: warning: variable 'ret' set but not used 64513 [-Wunused-but-set-variable] 64514 rrcrtc.c: In function 'rrCheckPixmapBounding': 64515 rrcrtc.c:505:13: warning: variable 'ret' set but not used 64516 [-Wunused-but-set-variable] 64517 rrcrtc.c:445:9: warning: unused variable 'i' [-Wunused-variable] 64518 64519 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64520 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64521 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 64522 64523commit 90642948cc78834d95f7a3bddaac7ff77b68ed7e 64524Merge: 9a35d4240 eda7dbff5 64525Author: Keith Packard <keithp@keithp.com> 64526Date: Thu Feb 14 11:05:48 2013 -0800 64527 64528 Merge remote-tracking branch 'jeremyhu/master' 64529 64530commit 9a35d4240e2aa91ac104f0f9f86f83ff9a2d3d04 64531Author: Peter Hutterer <peter.hutterer@who-t.net> 64532Date: Thu Feb 14 16:31:13 2013 +1000 64533 64534 os: fix pnprintf OOB buffer read for unterminated length modifiers 64535 64536 Format strings with length modifiers but missing format specifier like "%0" 64537 will read one byte past the array size. 64538 64539 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64540 Reviewed-by: Keith Packard <keithp@keithp.com> 64541 Signed-off-by: Keith Packard <keithp@keithp.com> 64542 64543commit eda7dbff5a9f35fefe1e3dedacb82daadbf5945e 64544Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 64545Date: Sat Feb 9 20:53:02 2013 -0800 64546 64547 XORG_TLS: Pick the first option that works (ie: prefer __thread) 64548 64549 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 64550 64551commit 5da82f872daf25939a8fa04f1561b01f0315a623 64552Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 64553Date: Sat Feb 9 20:40:10 2013 -0800 64554 64555 os: Ensure <dix-config.h> is included in strndup.c 64556 64557 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 64558 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64559 64560commit 67eaf4b46f678bc904f47dfcab6655e170d843e0 64561Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 64562Date: Sat Feb 9 20:34:33 2013 -0800 64563 64564 XQuartz: Ensure <dix-config.h> is included in capabilities.c 64565 64566 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 64567 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64568 64569commit 955d434f4d755d00a24ae4068b9957765989d672 64570Author: Keith Packard <keithp@keithp.com> 64571Date: Wed Feb 13 21:39:37 2013 -0800 64572 64573 Update to version 1.3.99.902 (1.14 RC2) 64574 64575 Signed-off-by: Keith Packard <keithp@keithp.com> 64576 64577commit 7115f6c709898a5124b67e19c61dc01334471358 64578Author: Aaron Plattner <aplattner@nvidia.com> 64579Date: Wed Jan 23 16:58:47 2013 -0800 64580 64581 randr: unref the provider shared pixmap the appropriate number of times 64582 64583 When an RandR shared pixmap is created in rrCreateSharedPixmap, it has a refcnt 64584 of 1. Then, PixmapShareToSlave bumps the refcnt to 2. However, there's no 64585 corresponding PixmapUnshareFromSlave where the refcnt can be decreased again, 64586 and there's no convenient common place where the refcnt can be decremented when 64587 the slave pixmap is destroyed. 64588 64589 Fix this by just unreffing the pixmap twice in RRCrtcDetachScanoutPixmap. 64590 64591 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 64592 Reviewed-by: Dave Airlie <airlied@redhat.com> 64593 Signed-off-by: Keith Packard <keithp@keithp.com> 64594 64595commit da92690107d90061205340d4cdc98b73b59db9b2 64596Author: Aaron Plattner <aplattner@nvidia.com> 64597Date: Mon Feb 11 14:31:56 2013 -0800 64598 64599 xf86: use nt_list_for_each_entry_safe to walk InputHandlers in xf86Wakeup 64600 64601 This is necessary when the input handler deletes itself from the 64602 list. Bug found by Maarten Lankhorst, this patch uses the list macros 64603 instead of open-coding the fix. 64604 64605 Signed-off-by: Keith Packard <keithp@keithp.com> 64606 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64607 64608commit d0a14877872b6a33172fb8c8f335792a28b9499b 64609Author: Bryce Harrington <bryce@canonical.com> 64610Date: Fri Feb 8 15:56:02 2013 -0800 64611 64612 xfree86: Man page shouldn't say Device is mandatory anymore 64613 64614 man xorg.conf states that the 'Device' identifier is required in the 64615 'Screen' section, yet current xserver defaults properly and boots up 64616 fine without it. 64617 64618 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=20742 64619 Signed-off-by: Bryce Harrington <bryce@canonical.com> 64620 Signed-off-by: Keith Packard <keithp@keithp.com> 64621 64622commit 5e91054aa070ea09b61325b7ec479d767b02730e 64623Author: Bryce Harrington <bryce@canonical.com> 64624Date: Fri Feb 8 15:56:01 2013 -0800 64625 64626 xfree86: Use fbdev/vesa driver on Oaktrail, Medfield, CDV rather than -intel 64627 64628 Instead of defaulting to -intel for Oaktrail, Medfield, and CDV chips, 64629 default to -fbdev. For Poulsbo (only), attempt to use -psb if it's 64630 installed, and fallback to fbdev otherwise. All other Intel chips 64631 should use -intel. 64632 64633 This fixed an issue where -intel would load on these chips and cause a 64634 boot failure. Newer -intel drivers avoid the boot hang, but it's still 64635 the wrong driver to load, so why take chances. 64636 64637 The patch was originally created by Stefan Dirsch for OpenSUSE. We have 64638 included it in our stable release (Ubuntu "quantal" 12.10) since 64639 December. 64640 64641 ref: https://bugzilla.novell.com/show_bug.cgi?id=772279 64642 ref: https://bugs.launchpad.net/ubuntu/+bug/1069031 64643 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=60514 64644 Signed-off-by: Bryce Harrington <bryce@canonical.com> 64645 Signed-off-by: Keith Packard <keithp@keithp.com> 64646 64647commit c1602d1c17967bdd4db9db19b3a9c0dfca6a58aa 64648Author: Dave Airlie <airlied@gmail.com> 64649Date: Tue Feb 5 07:46:06 2013 -0800 64650 64651 randr: bump advertised RandR version to 1.4 64652 64653 Signed-off-by: Dave Airlie <airlied@redhat.com> 64654 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 64655 Tested-by: Aaron Plattner <aplattner@nvidia.com> 64656 Signed-off-by: Keith Packard <keithp@keithp.com> 64657 64658commit 9f79e93b6b3416055d08a0e8f9f16d5fd0649e36 64659Author: Peter Hutterer <peter.hutterer@who-t.net> 64660Date: Wed Jan 30 03:20:07 2013 +0000 64661 64662 Short-cut the input device cleanup process during AbortServer() 64663 64664 If we're about to abort, we're already in the signal handler and cannot call 64665 down to the default device cleanup routines (which reset, free, alloc, and 64666 do a bunch of other things). 64667 64668 Add a new DEVICE_ABORT mode to signal a driver's DeviceProc that it must 64669 reset the hardware if needed but do nothing else. An actual HW reset is only 64670 required for some drivers dealing with the HW directly. 64671 64672 This is largely backwards-compatible, hence the input ABI minor bump only. 64673 64674 Drivers we care about either return BadValue on a mode that's not 64675 DEVICE_{INIT|ON|OFF|CLOSE} or print an error and return BadValue. Exception 64676 here is vmmouse, which currently ignores it and would not reset anything. 64677 This should be fixed if the reset is required. 64678 64679 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64680 Reviewed-by: Keith Packard <keithp@keithp.com> 64681 Signed-off-by: Keith Packard <keithp@keithp.com> 64682 64683commit b58221f9da8c549d979215271359c6cd88b5568a 64684Author: Peter Hutterer <peter.hutterer@who-t.net> 64685Date: Fri Feb 8 14:52:02 2013 +1000 64686 64687 dix: support the transformation matrix for relative devices. 64688 64689 The transformation matrix we previously stored was a scaled matrix based on 64690 the axis ranges of the device. For relative movements, the scaling is not 64691 required (or desired). 64692 64693 Store two separate matrices, one as requested by the client, one as the 64694 product of [scale . matrix . inv_scale]. Depending on the type of movement, 64695 apply the respective matrix. 64696 64697 For relative movements, also drop the translation component since it doesn't 64698 really make sense to use that bit. 64699 64700 Input ABI 19 64701 64702 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64703 Reviewed-by: Keith Packard <keithp@keithp.com> 64704 Signed-off-by: Keith Packard <keithp@keithp.com> 64705 64706commit b173eb2ae3349c557db1ff9e424fa540b8289bb2 64707Author: Keith Packard <keithp@keithp.com> 64708Date: Fri Feb 8 08:57:43 2013 -0800 64709 64710 os: Round fraction in pnprintf %f format 64711 64712 Truncating the fraction part leads to a test failure where -1203.30 is 64713 printed as -1203.29. Round this to the nearest value instead by adding 64714 0.5 before converting to an integer 64715 64716 Signed-off-by: Keith Packard <keithp@keithp.com> 64717 64718commit 509b3c3dc82e7abce1900d5e1cddd90f23be5a87 64719Author: Carlos Garnacho <carlosg@gnome.org> 64720Date: Wed Feb 6 14:07:22 2013 +0100 64721 64722 dix: Set focus field on XI2 crossing events 64723 64724 Set on DeviceEnterLeaveEvent() the xXIEnterEvent->focus field 64725 similarly to how the CoreEnterLeaveEvent() function above does 64726 for core events. 64727 64728 This fixes bug https://bugzilla.gnome.org/show_bug.cgi?id=677329 64729 reported to GTK+, where focus handling on window managers with 64730 sloppy focus or no window manager present was broken due to this 64731 field being always set to FALSE. 64732 64733 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 64734 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64735 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64736 64737commit 3e4be4033aed78b2bb3a18d51f0963989efd1af3 64738Author: Peter Hutterer <peter.hutterer@who-t.net> 64739Date: Fri Jan 25 11:47:32 2013 +1000 64740 64741 dix: when shutting down slave devices, shut down xtest devices last 64742 64743 XTest devices are the first ones in the list, being initialised together 64744 with the master devices. If we disable the devices in-order and a device has 64745 a button down when being disabled, the XTest device is checked for a 64746 required button release (xkbAccessX.c's ProcessPointerEvent). This fails if 64747 the device is already NULL. 64748 64749 Instead of putting the check there, disable the devices in the reverse order 64750 they are initialised. Disable physical slaves first, then xtest devices, 64751 then the master devices. 64752 64753 Testcase: shut down server with a button still held down on a physical 64754 device 64755 64756 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64757 Reviewed-by: Keith Packard <keithp@keithp.com> 64758 64759commit 0d5bb882600ee7734af034fbea935a79d21d1e70 64760Merge: b33fcb149 61a99aff9 64761Author: Peter Hutterer <peter.hutterer@who-t.net> 64762Date: Fri Feb 8 14:10:52 2013 +1000 64763 64764 Merge branch 'ptraccel-fixes' into for-keith 64765 64766commit 61a99aff9d33728a0b67920254d2d4d79f80cf39 64767Author: Peter Hutterer <peter.hutterer@who-t.net> 64768Date: Fri Jan 11 14:22:07 2013 +1000 64769 64770 dix: pre-scale relative events from abs devices to desktop ratio (#31636) 64771 64772 Absolute devices may send relative events depending on the mode (synaptics 64773 by default, wacom per option). The relative events are added to the previous 64774 position, converted into device coordinates and then scaled into desktop 64775 coordinates for pointer movement. 64776 64777 Because the device range must be mapped into the desktop coordinate range, 64778 this results in uneven scaling depending dimensions, e.g. on a setup with 64779 width == 2 * height, a relative movement of 10/10 in device coordinates 64780 results in a cursor movement of 20/10 (+ acceleration) 64781 64782 Other commonly user-visible results: 64783 * the touchpad changing acceleration once an external monitor as added. 64784 * drawing a circle on a wacom tablet in relative mode gives an ellipsis in 64785 the same ratio as the desktop dimensions. 64786 64787 Solution: pre-scale the incoming relative x/y coordinates by width/height 64788 ratio of the total desktop size. Then add them to the previous 64789 coordinates and scale back with the previous mapping, which will undo the 64790 pre-scaling and give us the right movement. 64791 64792 X.Org Bug 31636 <http://bugs.freedesktop.org/show_bug.cgi?id=31636> 64793 64794 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64795 Reviewed-by: Keith Packard <keithp@keithp.com> 64796 64797commit a6ba2b79ae8ad0fdee3f208d5e030b012df48785 64798Author: Peter Hutterer <peter.hutterer@who-t.net> 64799Date: Fri Jan 11 08:53:24 2013 +1000 64800 64801 dix: unify prefix for ptraccel debugging in DebugAccelF macro 64802 64803 If we're already using our own custom macro, might as well use it properly. 64804 64805 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64806 Reviewed-by: Keith Packard <keithp@keithp.com> 64807 64808commit 0d7d79406011169ad95e23c6e937f6d15a5ec3cc 64809Author: Peter Hutterer <peter.hutterer@who-t.net> 64810Date: Fri Jan 11 08:52:08 2013 +1000 64811 64812 dix: use BUG_RETURN_VAL for an error message 64813 64814 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64815 Reviewed-by: Keith Packard <keithp@keithp.com> 64816 64817commit a0c38ea6cbad61edcfefff0e5dd6330edb706f13 64818Author: Peter Hutterer <peter.hutterer@who-t.net> 64819Date: Fri Jan 11 08:49:15 2013 +1000 64820 64821 dix: add some more info to a ptraccel debug msg 64822 64823 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64824 Reviewed-by: Keith Packard <keithp@keithp.com> 64825 64826commit 95125a7c0cbbbae40216a0497acdd863ddc645ed 64827Author: Peter Hutterer <peter.hutterer@who-t.net> 64828Date: Thu Jan 10 13:19:27 2013 +1000 64829 64830 dix: fix ptraccel debugging printfs 64831 64832 This is mostly sigsafe code, so use sigsave printf. And update some fields 64833 to double that used to be int. 64834 64835 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64836 Reviewed-by: Keith Packard <keithp@keithp.com> 64837 64838commit b33fcb149710a28fd8767b2307a97bf367de695e 64839Author: Andreas Wettstein <wettstein509@solnet.ch> 64840Date: Tue Jan 29 21:49:20 2013 +0100 64841 64842 xkb: Fix repeat behaviour of redirect and message actions 64843 64844 The redirect and the message action filter functions implicitly assumed that 64845 when they receive an event for the same keycode they were activated for, that 64846 this is the a release of the key that activated the filter. This is not true 64847 if the key autorepeats. Due to the incorrect assumption, the effective key 64848 repeat rate was effectively halved. 64849 64850 Signed-off-by: Andreas Wettstein <wettstein509@solnet.ch> 64851 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64852 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64853 64854commit 8571c648a79444bcee9a0fe6e395129116372f49 64855Author: Peter Hutterer <peter.hutterer@who-t.net> 64856Date: Sat Jan 26 15:53:08 2013 +1000 64857 64858 Xext: if a root window is given in XTestFakeInput, move to that 64859 64860 For absolute events, if the client specifies a screen number offset the 64861 coordinates by that. And add a new flag so we know when _not_ to add the 64862 screen offset in GPE. 64863 64864 Without this offset and the flag, GPE would simply add the offset of the 64865 current screen if POINTER_SCREEN is set. 64866 64867 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64868 Reviewed-by: Keith Packard <keithp@keithp.com> 64869 64870commit 9fd6cb89539fde44a41ae5183c89ef9c8831c8dd 64871Author: Peter Hutterer <peter.hutterer@who-t.net> 64872Date: Tue Jan 29 15:13:44 2013 +1000 64873 64874 Xext: pass the current screen to miProcessDeviceEvent() from xtest calls 64875 64876 Not passing in a screen means we skip the screen crossing updates, so a 64877 xtest event that changes between ScreenRecs won't do so until the next 64878 physical event comes in or never, whichever comes earlier. 64879 64880 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64881 Reviewed-by: Keith Packard <keithp@keithp.com> 64882 64883commit a191dbfe850ed9c6440346f59cb0078e0e844edc 64884Author: Sybren van Elderen <sowmestno@msn.com> 64885Date: Tue Jan 29 15:43:57 2013 +1000 64886 64887 dix: when scaling from desktop coord, take the total desktop size (#51904) 64888 64889 Scaled is already in desktop coordinates, take the total width into account, 64890 not just the current screen's width. 64891 64892 Fixes Xdmx pointer position calculation. 64893 64894 X.Org Bug 51904 <http://bugs.freedesktop.org/show_bug.cgi?id=51904> 64895 64896 Signed-off-by: Sybren van Elderen <sowmestno@msn.com> 64897 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64898 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64899 64900commit 1cb19803f0f8dfd1e0fb9d189afe2262e24a0be5 64901Author: Peter Hutterer <peter.hutterer@who-t.net> 64902Date: Tue Jan 29 12:51:15 2013 +1000 64903 64904 include: fix typo in list description 64905 64906 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64907 64908commit 3d35dfcf5bad1b0a028fbecd65cb6cf6ebf12503 64909Author: Ted Felix <ted@tedfelix.com> 64910Date: Tue Jan 29 16:36:48 2013 +1000 64911 64912 xfree86: bail on misformed acpi strings (#73227) 64913 64914 If acpid sends a string in a format that we can't parse, bail out instead of 64915 potentially dereferencing a NULL-pointer. 64916 64917 X.Org Bug 73227 <http://bugs.freedesktop.org/show_bug.cgi?id=73227> 64918 64919 Signed-off-by: Ted Felix <ted@tedfelix.com> 64920 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64921 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64922 64923commit fdc451588816c4bc798d54e56316530e9066be80 64924Author: Peter Hutterer <peter.hutterer@who-t.net> 64925Date: Tue Jan 29 11:01:29 2013 +1000 64926 64927 Xi: limit valuator copy to valuator array size (#59939) 64928 64929 mask[(MAX_VALUATORS + 7)/8] is larger than data[MAX_VALUATORS], so static 64930 code checkers think we may be running OOB on the data array. Mask is 64931 initialized to 0, so this should not happen, but change it anyway to shut up 64932 code analyzer noise. 64933 64934 X.Org Bug 59939 <http://bugs.freedesktop.org/show_bug.cgi?id=59939> 64935 64936 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64937 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64938 64939commit 48bc30c5413a1be0039fa77affcbbb4fe677479f 64940Author: Alan Coopersmith <alan.coopersmith@oracle.com> 64941Date: Tue Jan 29 10:24:32 2013 +1000 64942 64943 Xext: avoid null-pointer dereference in XTestFakeInput (#59937) 64944 64945 dv is still NULL at this point, so return firstValuator instead (which is 64946 the same value dv->firstValuator would be once initialized) 64947 64948 X.Org Bug 59937 <http://bugs.freedesktop.org/show_bug.cgi?id=59937> 64949 64950 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64951 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64952 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64953 64954commit 1058fcf57fdcb94d92e7b5f4483b347853d5f8e6 64955Author: Peter Hutterer <peter.hutterer@who-t.net> 64956Date: Sat Jan 26 14:13:33 2013 +1000 64957 64958 dmx: don't include dmx-config.h from xdmxconfig (#37502) 64959 64960 dmx-config.h is a server header which includes dix-config.h. That again 64961 defines a bunch of server-specifics, including setting the size of XID to 64962 32 bit. 64963 64964 libX11 uses unsigned long (8 bits on x86_64). XGCValues thus ends up being 64965 16 bytes smaller in xdmxconfig than in the library, causing garbage to be 64966 sent to the server. 64967 64968 X.Org Bug 37502 <http://bugs.freedesktop.org/show_bug.cgi?id=37502> 64969 64970 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 64971 Reviewed-by: Keith Packard <keithp@keithp.com> 64972 64973commit ac34281b8a487640c61f258ae45a6f17a4fb3cc0 64974Author: Dave Airlie <airlied@redhat.com> 64975Date: Thu Feb 7 12:24:20 2013 +1000 64976 64977 modesetting: provide dummy hooks for shadow 64978 64979 Since in some wierd cases the server can call these without checking they 64980 exist. 64981 64982 Signed-off-by: Dave Airlie <airlied@redhat.com> 64983 64984commit 73974dd7ea9ca4d4cdd5464cb813088a6ee9770b 64985Author: Alan Coopersmith <alan.coopersmith@oracle.com> 64986Date: Sun Jan 27 15:42:02 2013 -0800 64987 64988 Avoid memory leak in ddc resort() if find_header() fails 64989 64990 Call find_header first, returning on failure before calling malloc. 64991 64992 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 64993 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 64994 64995commit b1129a1f1771c9d1653cc15aae94a614f081638a 64996Author: Alan Coopersmith <alan.coopersmith@oracle.com> 64997Date: Sun Jan 27 14:00:54 2013 -0800 64998 64999 xf86XvMCScreenInit: Avoid leak if dixRegisterPrivateKey fails 65000 65001 Found by parfait 1.1 memory analyser: 65002 Memory leak of pointer 'pAdapt' allocated with malloc((88 * num_adaptors)) 65003 at line 162 of hw/xfree86/common/xf86xvmc.c in function 'xf86XvMCScreenInit'. 65004 'pAdapt' allocated at line 158 with malloc((88 * num_adaptors)). 65005 65006 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 65007 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65008 65009commit 563db909bcf965b6103c1807bf9f00ede957077d 65010Author: Alan Coopersmith <alan.coopersmith@oracle.com> 65011Date: Sun Jan 27 13:55:50 2013 -0800 65012 65013 Avoid memory leak on realloc failure in localRegisterFreeBoxCallback 65014 65015 Also avoids leaving invalid pointers in structures if realloc had to 65016 move them elsewhere to make them larger. 65017 65018 Found by parfait 1.1 code analyzer: 65019 Memory leak of pointer 'newCallbacks' allocated with realloc(((char*)offman->FreeBoxesUpdateCallback), (8 * (offman->NumCallbacks + 1))) 65020 at line 328 of hw/xfree86/common/xf86fbman.c in function 'localRegisterFreeBoxCallback'. 65021 'newCallbacks' allocated at line 320 with realloc(((char*)offman->FreeBoxesUpdateCallback), (8 * (offman->NumCallbacks + 1))). 65022 newCallbacks leaks when newCallbacks != NULL at line 327. 65023 Memory leak of pointer 'newPrivates' allocated with realloc(((char*)offman->devPrivates), (8 * (offman->NumCallbacks + 1))) 65024 at line 328 of hw/xfree86/common/xf86fbman.c in function 'localRegisterFreeBoxCallback'. 65025 'newPrivates' allocated at line 324 with realloc(((char*)offman->devPrivates), (8 * (offman->NumCallbacks + 1))). 65026 newPrivates leaks when newCallbacks == NULL at line 327. 65027 65028 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 65029 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65030 65031commit 08f75d3a9661c6c32800e1b4f150626200b889d9 65032Author: Alan Coopersmith <alan.coopersmith@oracle.com> 65033Date: Sun Jan 27 13:50:30 2013 -0800 65034 65035 Avoid NULL pointer dereference in xf86TokenToOptinfo if token not found 65036 65037 Reported by parfait 1.1 code analyzer: 65038 65039 Error: Null pointer dereference (CWE 476) 65040 Read from null pointer 'p' 65041 at line 746 of hw/xfree86/common/xf86Option.c in function 'xf86TokenToOptName'. 65042 Function 'xf86TokenToOptinfo' may return constant 'NULL' at line 721, called at line 745. 65043 Null pointer introduced at line 721 in function 'xf86TokenToOptinfo'. 65044 65045 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 65046 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65047 65048commit c1c01e350834a23161b33bd34b2fa9c01d02a65b 65049Author: Alan Coopersmith <alan.coopersmith@oracle.com> 65050Date: Sun Jan 27 13:10:08 2013 -0800 65051 65052 Make xf86ValidateModes actually copy clock range list to screen pointer 65053 65054 Our in-house parfait 1.1 code analysis tool complained that every exit 65055 path from xf86ValidateModes() in hw/xfree86/common/xf86Mode.c leaks the 65056 storeClockRanges allocation made at line 1501 with XNFalloc. 65057 65058 Investigating, it seems that this code to copy the clock range list to 65059 the clockRanges list in the screen pointer is just plain insane, and 65060 according to git, has been since we first imported it from XFree86. 65061 65062 We start at line 1495 by walking the linked list from scrp->clockRanges 65063 until we find the end. But that was just a diversion, since we've found 65064 the end and immediately forgotten it, and thus at 1499 we know that 65065 storeClockRanges is NULL, but that's not a problem since we're going to 65066 immediately overwrite that value as the first thing in the loop. 65067 65068 So we move on through this loop at 1499, which takes us through the 65069 linked list from the clockRanges variable, and for every entry in 65070 that list allocates a new structure and copies cp to it. If we've 65071 not filled in the screen's clockRanges pointer yet, we set it to 65072 the first storeClockRanges we copied from cp. Otherwise, as best 65073 I can tell, we just drop it into memory and let it leak away, as 65074 parfait warned. 65075 65076 And then we hit the loop action, which if we haven't hit the end of 65077 the cp list, advances cp to the next item in the list, and then just 65078 for the fun of it, also sets storeClockRanges to the ->next pointer it 65079 has just copied from cp as well, even though it's going to overwrite 65080 it as the very first instruction in the loop body. 65081 65082 v2: rewritten using nt_list_* macros from Xorg's list.h header 65083 65084 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 65085 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65086 65087commit 89badba082c81d20fe35cb064c16e131ff288ca3 65088Author: Alan Coopersmith <alan.coopersmith@oracle.com> 65089Date: Sun Jan 27 12:08:47 2013 -0800 65090 65091 Free keymap on error in Xephyr's hostx_load_keymap 65092 65093 Found by parfait 1.1 code analyser: 65094 Memory leak of pointer 'keymap' allocated with XGetKeyboardMapping(HostX.dpy, min_keycode, ((max_keycode - min_keycode) + 1), &host_width) 65095 at line 861 of hw/kdrive/ephyr/hostx.c in function 'hostx_load_keymap'. 65096 'keymap' allocated at line 845 with XGetKeyboardMapping(HostX.dpy, min_keycode, ((max_keycode - min_keycode) + 1), &host_width). 65097 65098 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 65099 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65100 65101commit 48b94651205b175760904e448f94111d1ab85e13 65102Author: Alan Coopersmith <alan.coopersmith@oracle.com> 65103Date: Sun Jan 27 11:20:45 2013 -0800 65104 65105 Stop leaking overlayWin in PanoramiXCompositeGetOverlayWindow error paths 65106 65107 Found by parfait 1.1 code analyzer: 65108 65109 Error: Memory leak (CWE 401) 65110 Memory leak of pointer 'overlayWin' allocated with malloc(72) 65111 at line 806 of composite/compext.c in function 'PanoramiXCompositeGetOverlayWindow'. 65112 pointer allocated at line 794 with malloc(72). 65113 <unknown> leaks when rc != 0 at line 804. 65114 at line 816 of composite/compext.c in function 'PanoramiXCompositeGetOverlayWindow'. 65115 pointer allocated at line 794 with malloc(72). 65116 <unknown> leaks when pOc == NULL at line 815. 65117 at line 825 of composite/compext.c in function 'PanoramiXCompositeGetOverlayWindow'. 65118 pointer allocated at line 794 with malloc(72). 65119 <unknown> leaks when cs->pOverlayWin == NULL at line 822 65120 and compCreateOverlayWindow(pScreen) == 0 at line 823. 65121 at line 834 of composite/compext.c in function 'PanoramiXCompositeGetOverlayWindow'. 65122 pointer allocated at line 794 with malloc(72). 65123 <unknown> leaks when rc != 0 at line 832. 65124 65125 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 65126 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65127 65128commit f4a9332ad149ed15353a9c482563bdd042d0b403 65129Author: Alan Coopersmith <alan.coopersmith@oracle.com> 65130Date: Sun Jan 27 10:06:42 2013 -0800 65131 65132 Handle failure to create counter in init_system_idle_counter 65133 65134 Check for NULL pointer (which can be returned for multiple reasons) 65135 before trying to dereference it to add privates. To avoid memory leak 65136 in error path, delay malloc of privates until we're ready to add them. 65137 65138 In case we do return NULL up through SyncInitDeviceIdleTime, handle the 65139 possibility of getting NULL passed back down to SyncRemoveDeviceIdleTime. 65140 65141 As reported by parfait 1.1: 65142 Error: Null pointer dereference (CWE 476) 65143 Read from null pointer 'idle_time_counter' 65144 at line 2764 of xserver/Xext/sync.c in function 'init_system_idle_counter'. 65145 Function 'SyncCreateSystemCounter' may return constant 'NULL' at line 952, called at line 2756. 65146 Null pointer introduced at line 952 in function 'SyncCreateSystemCounter'. 65147 65148 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 65149 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65150 65151commit 7fe5e6dfa5c1e71d8b7540b28c1d508687a2fbee 65152Author: Jasper St. Pierre <jstpierre@mecheye.net> 65153Date: Wed Jan 23 13:11:55 2013 -0500 65154 65155 protocol-versions: Bump minor version of XI 65156 65157 This was accidentally excluded when we added barriers. 65158 65159 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65160 Signed-off-by: Keith Packard <keithp@keithp.com> 65161 65162commit 70b127c9f1c53bdb42f078265e67f76b464deae2 65163Author: Aaron Plattner <aplattner@nvidia.com> 65164Date: Thu Jan 10 17:01:17 2013 -0800 65165 65166 config/udev: fix "removing GPU device" format string mistake 65167 65168 udev.c: In function 'device_removed': 65169 udev.c:270:9: warning: format '%d' expects argument of type 'int', but argument 3 has type 'const char *' [-Wformat] 65170 65171 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 65172 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65173 Signed-off-by: Keith Packard <keithp@keithp.com> 65174 65175commit 605dfc6804a05ff2bda5692fec26c37344fd95cb 65176Author: Dave Airlie <airlied@gmail.com> 65177Date: Tue Jan 22 07:39:53 2013 +1000 65178 65179 xserver: fix build regression since 91ab237358c6e33da854914d3de493a9cbea7637 65180 65181 inputstr, double defines TouchListener typedef, maybe some gcc handles it, 65182 but not all. 65183 65184 fixes tinderbox 65185 65186 Reported-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65187 Signed-off-by: Dave Airlie <airlied@redhat.com> 65188 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65189 Signed-off-by: Keith Packard <keithp@keithp.com> 65190 65191commit 069d8ed3eb659c48dd2b0f8b7b8c11f092fdb362 65192Merge: 591c06277 d6dcde7a0 65193Author: Keith Packard <keithp@keithp.com> 65194Date: Sun Jan 20 15:58:38 2013 -0800 65195 65196 Merge remote-tracking branch 'jturney/xserver-next' 65197 65198commit 591c06277bb120ab9615633f2d28addbd3a2aa5f 65199Merge: 6703a7c7c fa6ab7d9b 65200Author: Keith Packard <keithp@keithp.com> 65201Date: Sun Jan 20 15:52:26 2013 -0800 65202 65203 Merge remote-tracking branch 'whot/for-keith' 65204 65205commit cde7cbe9674e8a771f9a4e646c1772a46a8230fb 65206Author: Peter Hutterer <peter.hutterer@who-t.net> 65207Date: Thu Jan 10 13:20:12 2013 +1000 65208 65209 os: add support for %f to pnprintf 65210 65211 This is the lazy man's %f support. Print the decimal part of the number, 65212 then append a decimal point, then print the first two digits of the 65213 fractional part. So %f in sigsafe printing is really %.2f. 65214 65215 No boundary checks in place here. 65216 65217 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65218 Reviewed-by: Keith Packard <keithp@keithp.com> 65219 65220commit 20def57632583aef095ca18792c7fce16d2d9004 65221Author: Peter Hutterer <peter.hutterer@who-t.net> 65222Date: Thu Jan 10 13:24:05 2013 +1000 65223 65224 os: silently ignore length modifiers in pnprintf 65225 65226 Until we have support for them, ignore any length modifiers so we don't need 65227 to update all callers. 65228 65229 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65230 Reviewed-by: Keith Packard <keithp@keithp.com> 65231 65232commit f53b2012f39085d866f267dda1442a48ace3c5a5 65233Author: Peter Hutterer <peter.hutterer@who-t.net> 65234Date: Thu Jan 17 16:19:51 2013 +1000 65235 65236 test/signal-logging: simplify tests using sprintf 65237 65238 Ever looked at your own code and thought 'WTF was I thinking?'. yeah, that. 65239 65240 Instead of passing in the expected string just use sprintf to print the 65241 number for us and compare. In the end we're just trying to emulate printf 65242 behaviour anyway. 65243 65244 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65245 Reviewed-by: Keith Packard <keithp@keithp.com> 65246 65247commit d6dcde7a03bb38c17ffc4ec5f0ca1c161e54569f 65248Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65249Date: Sat Feb 11 12:22:17 2012 +0000 65250 65251 hw/xwin: Stop assuming WS_EX_APPWINDOW style in WM_SHOWWINDOW 65252 65253 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65254 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65255 65256commit 56e94403f8f9182e05428d895a983371c7737d2a 65257Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65258Date: Wed Jan 9 20:15:01 2013 +0000 65259 65260 hw/xwin: Use ITaskBarList interface to ensure show-on-taskbar state is updated correctly 65261 65262 Use ITaskBarList interface to ensure that the taskbar notices if the window has 65263 changed it's style in a way which affects if the taskbar shows it or not. 65264 65265 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65266 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65267 65268commit c94d1cb0a49106f44714f4511720a197cc549164 65269Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65270Date: Thu Jan 10 14:35:56 2013 +0000 65271 65272 hw/xwin: Ensure full styling is applied when the window is mapped 65273 65274 Move styling update code from WM_WM_HINTS_EVENT to a function UpdateStyle(), 65275 which is also invoked from WM_WM_MAP3, so everything which needs to be done 65276 to style the window happens when it is mapped 65277 65278 (Otherwise, the appearance of the window is sensitive to the timing of the 65279 notification of the windows appearance hint properties being set relative to 65280 window creation. e.g. see [1]) 65281 65282 [1] http://sourceware.org/ml/cygwin-xfree/2012-06/msg00004.html 65283 65284 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65285 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65286 65287commit ef61f8cacc84080c9156675f9ce26a27e8a90ac1 65288Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65289Date: Sat Mar 31 18:45:28 2012 +0100 65290 65291 hw/xwin: Make sure that WM_WM_HINTS_EVENT does nothing for override-redirect windows 65292 65293 Future work: It looks like this code could be rationalized quite a lot: It might 65294 make sense to pull the checking for override-redirect up out of UpdateIcon() and 65295 UpdateName() and consolidate WM_WM_MAP2 and WM_WM_MAP3 65296 65297 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65298 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65299 65300commit 3628559e594fcbdfcc14b1e8fa60aa841f184e19 65301Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65302Date: Sun Feb 5 11:25:39 2012 +0000 65303 65304 hw/xwin: Add a new WM_WM_HINTS_EVENT event to update window style 65305 65306 Add a new WM_WM_HINTS_EVENT event to update window style if any of the 65307 properties which affect window style change 65308 65309 Check PropertyNotify events for any of the window properties which we consider 65310 to decide on the window style, and update the window style by sending a 65311 WM_WM_HINTS_EVENT message to the WM. 65312 65313 This allows the styling of the window to change during it's lifetime. 65314 65315 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65316 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65317 65318commit 066ecbd11d516ea68d7ebc7470232d01c5717546 65319Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65320Date: Thu Jan 10 14:37:45 2013 +0000 65321 65322 hw/xwin: Move reshape code from winUpdateWindowPosition() to the map event handler 65323 65324 Move reshape code, which was only used when handling a map event, from 65325 winUpdateWindowPosition(), to put it explicitly in the map event handler. 65326 65327 Remove 'reshape' parameter from winUpdatePosition(). 65328 65329 (Note that there's no handling of the ShapeNotify event to notice when the 65330 window shape changes, instead we hook the screen SetShape procedure and reshape 65331 the native window then) 65332 65333 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65334 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65335 65336commit 852d1fb042f4160fe023a015f1c9a34126bf911a 65337Author: Ryan Pavlik <rpavlik@iastate.edu> 65338Date: Sat Dec 1 16:58:40 2012 +0000 65339 65340 hw/xwin: Add missing include xwin-config.h to winglobals.h 65341 65342 winglobals.h checks if RELOCATE_PROJECTROOT is defined to see if a declaration 65343 of g_fLogFileChanged is needed, so must include xwin-config.h 65344 65345 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 65346 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65347 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65348 65349commit ab686ce029208abf970a4bcd1435bf8411a44de9 65350Author: Ryan Pavlik <rpavlik@iastate.edu> 65351Date: Wed Oct 26 17:03:25 2011 -0500 65352 65353 include: Add RELOCATE_PROJECTROOT to xwin-config.h header 65354 65355 RELOCATE_PROJECTROOT is AC_DEFINED in configure.ac, but currently has no effect 65356 as it doesn't appear in any AC_CONFIG_HEADER header. 65357 65358 When packaged for Windows, we do not have a unix-style filesystem tree, where 65359 file needed by the X server can be found in fixed, absolute paths under the 65360 prefix (PROJECTROOT). 65361 65362 Instead, the filesystem tree containing files needed by the X server and clients 65363 will be installed with the directory containing the X server executable as the 65364 root directory of that tree. 65365 65366 (Typically, this will be in the Program Files directory, which does not have a 65367 fixed name, as it can be moved, localized, or added to to indicate x86 or x64 65368 binaries) 65369 65370 So, RELOCATE_PROJECTROOT is used to make a native Windows build of the X server 65371 look for various files (fonts, xkb data) in locations relative to the X server 65372 rather than at absolute paths, by translating those paths at run-time. 65373 65374 Additionally the XKEYSYMDB, XERRORDB, XLOCALEDIR env vars checked by libX11 are 65375 set appropriately for clients started by the X server. 65376 65377 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 65378 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65379 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65380 65381commit fa6ab7d9b2d7fd8184f1e068360607845f5c33ab 65382Merge: adde4e644 0e1ab433f 65383Author: Peter Hutterer <peter.hutterer@who-t.net> 65384Date: Fri Jan 11 14:58:17 2013 +1000 65385 65386 Merge branch 'pointer-emulation-fixes-56558-v2' into for-keith 65387 65388commit adde4e64480315dc5b47a727ee37d86f5cd8584f 65389Author: Peter Hutterer <peter.hutterer@who-t.net> 65390Date: Thu Jan 10 10:33:05 2013 +1000 65391 65392 dix: typo fix in comment 65393 65394 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65395 65396commit 05ed095dd8d6cf939b4ebd9a59d70ce32705df7c 65397Author: Benjamin Tissoires <benjamin.tissoires@gmail.com> 65398Date: Wed Jan 9 19:32:19 2013 +0100 65399 65400 dix: fix error logging occuring in signal context of GetTouchEvents 65401 65402 GetTouchEvents is usually called in a signal context. 65403 Calling ErrorF for the error messages leads to X complaining about log: 65404 65405 (EE) BUG: triggered 'if (inSignalContext)' 65406 (EE) BUG: log.c:484 in LogVMessageVerb() 65407 (EE) Warning: attempting to log data in a signal unsafe manner while in signal context. 65408 Please update to check inSignalContext and/or use LogMessageVerbSigSafe() or ErrorFSigSafe(). 65409 The offending log format message is: 65410 %s: Attempted to start touch without x/y (driver bug) 65411 65412 Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com> 65413 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65414 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65415 65416commit f4a58469a298c226668fd8dce375bf22331c902d 65417Author: Peter Hutterer <peter.hutterer@who-t.net> 65418Date: Wed Jan 9 13:58:56 2013 +1000 65419 65420 xfree86: don't access the old input handler after freeing it 65421 65422 Introduced in 323869f3298cbbfe864af9404a8aed1bf7995d79 65423 65424 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65425 Reviewed-by: Dave Airlie <airlied@redhat.com> 65426 65427commit 205cfbd6d9824fb9a67c21b19bc8f1e66c9df4d2 65428Author: Dave Airlie <airlied@gmail.com> 65429Date: Sat Jan 5 18:35:42 2013 +1000 65430 65431 xf86: bump input ABI version to 19 65432 65433 The changes to miPointerSetPosition interface from int->double breaks 65434 the SIS driver build, so time to bump this. 65435 65436 Signed-off-by: Dave Airlie <airlied@redhat.com> 65437 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65438 65439commit ad3bc571348a7007a2960bf87ae739397c5511ee 65440Author: Peter Hutterer <peter.hutterer@who-t.net> 65441Date: Tue Jan 8 11:19:09 2013 +1000 65442 65443 xfree86: update the device state for all DGA events (#59100) 65444 65445 DGA only handles master devices but it does intercept slave device events as 65446 well (since the event handlers are per event type, not per device). 65447 65448 The DGA code must thus call into UpdateDeviceState to reset the button/key 65449 state on the slave device before it discards the remainder of the event. 65450 65451 Test case: 65452 - Passive GrabModeSync on VCP 65453 - Press button 65454 - Enable DGA after ButtonPress 65455 - AllowEvents(SyncPointer) 65456 - Release button 65457 65458 The button release is handled by DGAProcessPointerEvent but the device state 65459 is never updated, so the slave ends up with the button permanently down. 65460 And since the master's button state is the union of the slave states, the 65461 master has the button permanently down. 65462 65463 X.Org Bug 59100 <http://bugs.freedesktop.org/show_bug.cgi?id=59100> 65464 65465 Reported-by: Steven Elliott <selliott4@austin.rr.com> 65466 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65467 Reviewed-by: Dave Airlie <airlied@redhat.com> 65468 65469commit c5f2818edbec2f87383baa6c6be5c389b73ca6f9 65470Author: Peter Hutterer <peter.hutterer@who-t.net> 65471Date: Tue Jan 8 10:13:53 2013 +1000 65472 65473 xfree86: set event->detail for DGA pointer events 65474 65475 Reported-by: Steven Elliott <selliott4@austin.rr.com> 65476 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65477 Reviewed-by: Dave Airlie <airlied@redhat.com> 65478 65479commit 519d183d78e0b0eaf47a473e94f5d8611baf8463 65480Author: Peter Hutterer <peter.hutterer@who-t.net> 65481Date: Mon Jan 7 10:44:33 2013 +1000 65482 65483 Fix two typos "requires an string value" 65484 65485 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65486 65487commit 4e13dd90144dde47550aceea4db4b4329e531279 65488Author: Peter Hutterer <peter.hutterer@who-t.net> 65489Date: Wed Aug 22 10:34:07 2012 +1000 65490 65491 dix: don't filter RawEvents if the grab window is not the root window (#53897) 65492 65493 If a XI2.1+ client has a grab on a non-root window, it must still receive 65494 raw events on the root window. 65495 65496 Test case: register for XI_ButtonPress on window and XI_RawMotion on root. 65497 No raw events are received once the press activates an implicit grab on the 65498 window. 65499 65500 X.Org Bug 53897 <http://bugs.freedesktop.org/show_bug.cgi?id=53897> 65501 65502 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65503 Reviewed-by: Keith Packard <keithp@keithp.com> 65504 65505commit a2037d7080ae64ea55f7d76971716346aa3ec6d3 65506Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65507Date: Wed Nov 28 22:25:35 2012 +0000 65508 65509 hw/xwin: Fix MinGW build of winSetAppModelID.c 65510 65511 Add missing #include <pthread.h> 65512 65513 In file included from /jhbuild/checkout/xorg/xserver/hw/xwin/winSetAppUserModelID.c:31:0: 65514 /jhbuild/checkout/xorg/xserver/hw/xwin/winwindow.h:140:11: error: expected declaration specifiers or ‘...’ before ‘pthread_t’ 65515 /jhbuild/checkout/xorg/xserver/hw/xwin/winwindow.h:141:11: error: expected declaration specifiers or ‘...’ before ‘pthread_t’ 65516 /jhbuild/checkout/xorg/xserver/hw/xwin/winwindow.h:142:11: error: expected declaration specifiers or ‘...’ before ‘pthread_mutex_t’ 65517 65518 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65519 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65520 65521commit f57100bb36eae3b4d75f3c315973405f705b8de6 65522Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65523Date: Tue Feb 23 13:38:48 2010 +0000 65524 65525 hw/xwin: Process one Windows message per wakeup, rather than all of them. 65526 65527 De-queuing Windows messages and X events happens in the same thread of 65528 execution. Draining the windows message queue can lead to the X event queue 65529 overflowing if lots of those windows messages cause X events (e.g. if a keyboard 65530 macro program has just dumped thousands of keypresses into the Windows message 65531 queue). See the mailing list thread [1] for more details. 65532 65533 Processing one Windows message per wakeup, rather than all of them gives the X 65534 server a chance to do stuff as well after each message. 65535 65536 [1] http://cygwin.com/ml/cygwin-xfree/2010-01/msg00056.html 65537 65538 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65539 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65540 65541commit e30e1ea98720acc583f34c830a1c1b7e3e88f694 65542Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65543Date: Mon Dec 17 22:38:25 2012 +0000 65544 65545 hw/xwin: Fix some comments in winkeybd.c 65546 65547 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65548 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65549 65550commit 6f4a48f8a55bc54b6d3e9d80734be05750c024de 65551Author: Jon TURNEY <jon.turney@dronecode.org.uk> 65552Date: Thu Apr 1 15:08:26 2010 +0100 65553 65554 hw/xwin: Bring the X screen window to the front on a single left-click on the tray icon 65555 65556 Bring the X screen window to the front on a single left click on the tray icon, 65557 like the comment says we do 65558 65559 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 65560 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 65561 65562commit 6703a7c7cf1a349c137e247a0c8eb462ff7b07be 65563Author: Keith Packard <keithp@keithp.com> 65564Date: Tue Jan 8 20:24:32 2013 -0800 65565 65566 hw/xfree86: Require only one working CRTC to start the server. 65567 65568 Instead of requiring every mode set to complete successfully, start up 65569 as long as at least one CRTC is working. This avoids failures when one 65570 or more CRTCs can't start due to mode setting conflicts. 65571 65572 Signed-off-by: Keith Packard <keithp@keithp.com> 65573 Reviewed-by: Dave Airlie <airlied@redhat.com> 65574 65575commit 34a571c5b9781025897ba5e5c5e2a34ae5262266 65576Author: Dave Airlie <airlied@redhat.com> 65577Date: Wed Jan 9 13:04:00 2013 +1000 65578 65579 modesetting: bump to 0.6.0 65580 65581commit 8650ff14a52047173fa32f12f22ec6f4e38ff433 65582Author: Dave Airlie <airlied@redhat.com> 65583Date: Wed Jan 9 12:48:30 2013 +1000 65584 65585 modesetting: fix crashes caused by udev race conditions 65586 65587 So the kernel removes the device, and the driver processes the first 65588 udev event, and gets no output back from the kernel, so it check 65589 and don't fall over. 65590 65591 This fixes a couple of crashes seen when hotplugging USB devices. 65592 65593 Signed-off-by: Dave Airlie <airlied@redhat.com> 65594 65595commit 0e1ab433f4048b3367bb2f01d16cd00502538e4d 65596Author: Peter Hutterer <peter.hutterer@who-t.net> 65597Date: Fri Jan 4 12:26:58 2013 +1000 65598 65599 dix: remove already-moved hunk 65600 65601 Should've been removed in bc1f90a615018c05994fae3e678dd2341256cd82a, but got 65602 left here due to a botched rebase. 65603 65604 Fixes stray button events sent to clients after deactivating an async 65605 pointer grab on a pointer-emulating-touch. 65606 65607 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65608 Reviewed-by: Keith Packard <keithp@keithp.com> 65609 65610commit 32a6d8a6b59c42f8d65002d7ca1cafb1957b656f 65611Author: Peter Hutterer <peter.hutterer@who-t.net> 65612Date: Thu Dec 20 16:25:43 2012 +1000 65613 65614 dix: check for the right device's xi2 mask 65615 65616 events.c: In function 'DeactivatePointerGrab': 65617 events.c:1524:51: warning: 'dev' may be used uninitialized in this function 65618 [-Wuninitialized 65619 65620 dev is unset when we get here, the device to check is "mouse". 65621 Introduced in ece8157a59751b3ed492fb2e1eb8d5f20221e195. 65622 65623 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65624 Reviewed-by: Keith Packard <keithp@keithp.com> 65625 65626commit f59499b5d05fde83813709e9848152951592120d 65627Author: Peter Hutterer <peter.hutterer@who-t.net> 65628Date: Tue Oct 30 12:44:08 2012 +1000 65629 65630 dix: add resource type to touch listeners 65631 65632 Instead of guessing what resource type the listener is and what property to 65633 retrieve, store the resource type in the listener directly. 65634 65635 Breaks XIT test cases: 65636 TouchGrabTestMultipleTaps.PassiveGrabPointerEmulationMultipleTouchesFastSuccession 65637 65638 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=56557 65639 65640 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65641 Reviewed-by: Keith Packard <keithp@keithp.com> 65642 Reviewed-by: Chase Douglas <chase.douglas@ubuntu.com> 65643 65644commit 9ad0fdb135a1c336771aee1f6eab75a6ad874aff 65645Author: Keith Packard <keithp@keithp.com> 65646Date: Tue Nov 27 11:21:17 2012 -0800 65647 65648 input: Record grab pointer in TouchListener 65649 65650 This places a pointer to the grab related to a TouchListener directly 65651 in the TouchListener structure rather than hoping to find the grab 65652 later on using the resource ID. 65653 65654 Passive grabs have resource ID in the resource DB so they can be 65655 removed when a client exits, and those resource IDs get copied when 65656 activated, but implicit grabs are constructed on-the-fly and have no 65657 resource DB entry. 65658 65659 Signed-off-by: Keith Packard <keithp@keithp.com> 65660 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65661 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65662 65663commit 91ab237358c6e33da854914d3de493a9cbea7637 65664Author: Keith Packard <keithp@keithp.com> 65665Date: Tue Nov 27 11:21:16 2012 -0800 65666 65667 input: Pull TouchListener declaration to top-level 65668 65669 No reason to have a struct declared inside another struct 65670 65671 Signed-off-by: Keith Packard <keithp@keithp.com> 65672 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65673 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65674 65675commit 3578cc3c2e1b5cb8eb191e2d12ad88e1bc9e6e1e 65676Author: Andreas Wettstein <wettstein509@solnet.ch> 65677Date: Wed Dec 19 18:13:21 2012 +0100 65678 65679 xkb: Do not use base group as an array index. 65680 65681 The base group is not brought into range and, therefore, using it as an array 65682 index crashed the X server. Also, at this place, we should ignore locked 65683 groups, but not latched groups. Therefore, use sum of base and latched groups, 65684 brought into range. 65685 65686 Reproducible with: 65687 key <FK07> { 65688 type= "ONE_LEVEL", 65689 symbols[Group1]= [ NoSymbol ], 65690 actions[Group1]= [ LatchGroup(group=-1, clearLocks) ] 65691 }; 65692 65693 And hitting F7 will exceed the group level and access arbitrary memory. 65694 65695 Signed-off-by: Andreas Wettstein <wettstein509@solnet.ch> 65696 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65697 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65698 65699commit df746a73410b892a4d41a2934cf9cd2e8ad7ba51 65700Author: Carlos Garnacho <carlosg@gnome.org> 65701Date: Wed Dec 19 18:42:39 2012 +0100 65702 65703 render: Unwrap early on the animated cursor BlockHandler 65704 65705 The loop above the previous call may end up triggering other 65706 handlers attaching to the same function slot, so unwrapping 65707 the handler after that could leave the just attached handler 65708 in a dangling but not unset state. 65709 65710 This issue was most visible on the XO, where destroying a 65711 window with an animated cursor set and running would trigger 65712 this inconsistent state, never calling the miSpriteBlockHandler 65713 again after the animated cursor is unset. 65714 65715 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 65716 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65717 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65718 65719commit 0fbd779a82919d5dbf8776be9b57a76c0eae6b14 65720Author: Carlos Garnacho <carlosg@gnome.org> 65721Date: Wed Dec 19 18:42:38 2012 +0100 65722 65723 mi: Ensure pointer emulating touch events update the sprite 65724 65725 Different miPointerSpriteFuncRec implementations do a varying 65726 business at ultimately calling miPointerUpdateSprite(), this 65727 particularly fails when using the plain mi sprite on touch events, 65728 where the sprite is just moved/updated on cursor changes. 65729 65730 So, ensure miPointerUpdateSprite() is called generically for 65731 pointer emulating touch events as with regular motion events. 65732 65733 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 65734 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65735 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65736 65737commit 8b328d4ee3873bc0a7a34f2cb9d301827244b98c 65738Author: Aaron Plattner <aplattner@nvidia.com> 65739Date: Fri Dec 21 07:37:33 2012 -0800 65740 65741 dix: Make small bitfields that store enums unsigned 65742 65743 Commit 31bf81772e146af79b0c456aae2159eba8b0280f changed the clientState field 65744 from a signed int to a signed int 2-bit bitfield. The ClientState enum that is 65745 expected to be assigned to this field has four values: ClientStateInitial (0), 65746 ClientStateRunning (1), ClientStateRetained (2), and ClientStateGone (3). 65747 However, because this bitfield is signed, ClientStateRetained becomes -2 when 65748 assigned, and ClientStateGone becomes -1. This causes warnings: 65749 65750 test.c:54:10: error: case label value exceeds maximum value for type [-Werror] 65751 test.c:55:10: error: case label value exceeds maximum value for type [-Werror] 65752 65753 The code here is a switch statement: 65754 65755 53 switch (client->clientState) { 65756 54 case ClientStateGone: 65757 55 case ClientStateRetained: 65758 56 [...] 65759 57 break; 65760 58 65761 59 default: 65762 60 [...] 65763 61 break; 65764 62 } 65765 65766 It also causes bizarre problems like this: 65767 65768 client->clientState = ClientStateGone; 65769 assert(client->clientState == ClientStateGone); // this assert fails 65770 65771 Also change the signedness of nearby bitfields to match. 65772 65773 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 65774 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 65775 Reviewed-by: Colin Harrison <colin.harrison at virgin.net> 65776 Signed-off-by: Keith Packard <keithp@keithp.com> 65777 65778commit 8f4820be7a2e0f6e286ddc85c4b75bccdbe8a730 65779Author: Peter Hutterer <peter.hutterer@who-t.net> 65780Date: Thu Dec 20 12:44:16 2012 +1000 65781 65782 test/xi2: fix compiler warning 65783 65784 protocol-xiwarppointer.c: In function ‘ScreenSetCursorPosition’: 65785 protocol-xiwarppointer.c:71:53: warning: declaration of ‘screen’ shadows a 65786 global declaration [-Wshadow] 65787 65788 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65789 65790commit bd91b05b631f13afd1f7a9d6cbc4f0c5408b523a 65791Author: Keith Packard <keithp@keithp.com> 65792Date: Wed Dec 19 12:47:35 2012 -0800 65793 65794 Update to version 1.13.99.901 (1.14 RC1) 65795 65796 Signed-off-by: Keith Packard <keithp@keithp.com> 65797 65798commit 2a0b544f5c09f79e4ab04ea8103a5ecf59ee2e7b 65799Merge: 0eb1559eb ba4bb3bc1 65800Author: Keith Packard <keithp@keithp.com> 65801Date: Wed Dec 19 12:26:00 2012 -0800 65802 65803 Merge remote-tracking branch 'jeremyhu/master' 65804 65805commit 0eb1559eb29d11e63c2b33e317590a88d86fe313 65806Merge: 386e4d76b ea1d76d1b 65807Author: Keith Packard <keithp@keithp.com> 65808Date: Wed Dec 19 12:21:31 2012 -0800 65809 65810 Merge remote-tracking branch 'yselkowitz/master' 65811 65812 I checked this patch with diff -w to check that it only affected 65813 whitespace. 65814 65815commit 386e4d76baaffe226d2d561ff936509454eb0ac2 65816Merge: 014a5c8a9 9ff2e8315 65817Author: Keith Packard <keithp@keithp.com> 65818Date: Wed Dec 19 12:17:59 2012 -0800 65819 65820 Merge remote-tracking branch 'alanc/master' 65821 65822commit 014a5c8a9d86f2f992183bff9106354fac2c3b0e 65823Merge: f793b5fd3 2eefa5d6e 65824Author: Keith Packard <keithp@keithp.com> 65825Date: Wed Dec 19 12:09:31 2012 -0800 65826 65827 Merge remote-tracking branch 'whot/barriers' 65828 65829 Conflicts: 65830 Xi/xichangehierarchy.c 65831 65832 Small conflict with the patch from 65833 65834 Xi: don't use devices after removing them 65835 65836 Was easily resolved by hand. 65837 65838 Signed-off-by: Keith Packard <keithp@keithp.com> 65839 65840commit 9ff2e831517875f96477862f979abff394e8d551 65841Author: Alan Coopersmith <alan.coopersmith@oracle.com> 65842Date: Tue Dec 18 00:41:08 2012 -0800 65843 65844 EnableDisableExtensionError: Use ARRAY_SIZE rather than sentinel 65845 65846 d785368e0e converted the other miinitext functions to use ARRAY_SIZE, 65847 and removed the sentinel, but missed EnableDisableExtensionError so 65848 passing an invalid extension name could cause the server to walk off 65849 the end off the list looking for a sentinel that wasn't there. 65850 65851 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 65852 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 65853 65854commit ba4bb3bc1b87eb57cc34d2ad1f302c9d2a15c847 65855Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 65856Date: Tue Dec 18 01:29:12 2012 -0800 65857 65858 XQuartz: Don't add the 15bit visual any more 65859 65860 Mountain Lion only supports 32bit backing stores, so don't use 15bit visuals until libXplugin adapts 65861 65862 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 65863 65864commit c298f9c42ec05486ac88fc91a84739d7c3e3ea9e 65865Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 65866Date: Tue Dec 18 01:28:02 2012 -0800 65867 65868 XQuartz: Revert some unfortunate auto-indenting mishaps with our super-nested for-loops 65869 65870 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 65871 65872commit 07a91fa6c6d535f3f05d4c3bd9c4d2b8c382c475 65873Author: Rob Clark <robdclark@gmail.com> 65874Date: Wed Dec 5 23:21:27 2012 +0000 65875 65876 hw/dmx: fix build without GLX 65877 65878 Fixes this compile break that showed up on arm recently: 65879 65880 dmxinit.c:746:26: error: 'glxSupported' undeclared (first use in this function) 65881 dmxinit.c:746:26: note: each undeclared identifier is reported only once for each function it appears in 65882 65883 Signed-off-by: Rob Clark <robdclark@gmail.com> 65884 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 65885 65886commit 5692a1e8f5befd3698134b1a5516a4dadda00115 65887Author: Alan Coopersmith <alan.coopersmith@oracle.com> 65888Date: Thu Dec 6 16:15:50 2012 -0800 65889 65890 Support compilers with alternate spellings of typeof 65891 65892 The AC_C_TYPEOF adds a #undef typeof to its autogenerated config.h.in 65893 template, but b8ab93dfbc7f292 didn't copy that to dix-config.h.in 65894 when HAVE_TYPEOF was, so the macro could claim typeof support but not 65895 make it work, when used with compilers like Solaris Studio 12.1 which 65896 only recognize it as __typeof__. 65897 65898 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 65899 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 65900 65901commit f793b5fd3eb16a2ada130367c2ffebeede69a322 65902Author: Peter Hutterer <peter.hutterer@who-t.net> 65903Date: Tue Dec 11 13:36:02 2012 +1000 65904 65905 dix: don't copy the wrong event mask when activating a passive grab 65906 65907 GrabMask is a union of core, XI1 and XI2 masks. If a XI2 grab is activated, 65908 the value is a random pointer value, using it as mask has unpredictable 65909 effects. 65910 65911 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65912 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 65913 Reviewed-by: Keith Packard <keithp@keithp.com> 65914 65915commit dd3242c87a0a58cba055eb99c0c3fcf03153e4b8 65916Author: Peter Hutterer <peter.hutterer@who-t.net> 65917Date: Fri Dec 14 11:34:15 2012 +1000 65918 65919 dix: don't allow overriding a grab with a different type of grab (#58255) 65920 65921 If a client has a core grab, don't allow re-grabbing with type XI2, etc. 65922 This was the intent of the original commit 65923 xorg-server-1.5.99.1-782-g09f9a86, but ineffective. 65924 65925 X.Org Bug 58255 <http://bugs.freedesktop.org/show_bug.cgi?id=58255> 65926 65927 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65928 Reviewed-by: Keith Packard <keithp@keithp.com> 65929 65930commit 2eefa5d6e870c57ac6a5930883d8cfe3a3882a43 65931Author: Peter Hutterer <peter.hutterer@who-t.net> 65932Date: Thu Dec 13 14:42:32 2012 +1000 65933 65934 Xi: if a MD is removed, send a barrier leave event (if applicable) 65935 65936 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65937 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 65938 65939commit e2423b627ef7e4d08ea0685af5e771ddd2b511ce 65940Author: Peter Hutterer <peter.hutterer@who-t.net> 65941Date: Thu Dec 13 14:04:13 2012 +1000 65942 65943 Xi: don't store the window pointer in barriers, store the window ID 65944 65945 When a client shuts down and resources are being freed, the window may have 65946 been freed already, so accessing it to get the window ID is bad. Plus, we 65947 never care about the window anyway other than for stuffing it into the 65948 event. 65949 65950 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65951 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 65952 65953commit f71c2f895c9e2f9d0d42feaac2a3e1d2deb71f67 65954Author: Peter Hutterer <peter.hutterer@who-t.net> 65955Date: Wed Dec 12 15:41:32 2012 +1000 65956 65957 Xi: fix per-device barrier handling 65958 65959 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65960 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 65961 65962commit 58bff17e43a80eb21b3ff6d4bb1596230e61f707 65963Author: Peter Hutterer <peter.hutterer@who-t.net> 65964Date: Thu Dec 13 11:49:54 2012 +1000 65965 65966 Xi: don't use devices after removing them 65967 65968 RemoveDevice() frees the DeviceIntPtr, we shouldn't use the pointer after 65969 that 65970 65971 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65972 Reviewed-by: Keith Packard <keithp@keithp.com> 65973 65974commit 3420a7778c7d5eaa638327f31dd460554c257bb1 65975Author: Peter Hutterer <peter.hutterer@who-t.net> 65976Date: Wed Dec 12 11:02:19 2012 +1000 65977 65978 xfree86: print message to the log when zapping the server 65979 65980 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 65981 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 65982 Signed-off-by: Keith Packard <keithp@keithp.com> 65983 65984commit 6d508b81857edaed03c7ee06410434ea56d9b701 65985Merge: d982d8774 bb6f3514c 65986Author: Keith Packard <keithp@keithp.com> 65987Date: Mon Dec 17 11:45:45 2012 -0800 65988 65989 Merge remote-tracking branch 'whot/for-keith' 65990 65991 I've looked at these patches, but I can't say I've actually 65992 reviewed them... 65993 65994commit d982d877436377597b8ed04bca1438d2edaf53ee 65995Merge: da3eaf6bd 6b4aa8a35 65996Author: Keith Packard <keithp@keithp.com> 65997Date: Mon Dec 17 10:56:12 2012 -0800 65998 65999 Merge remote-tracking branch 'airlied/for-keithp' 66000 66001commit da3eaf6bdbd7ce3bebf2c490cbe2448b4c402dba 66002Author: Dave Airlie <airlied@redhat.com> 66003Date: Mon Dec 17 15:40:17 2012 +1000 66004 66005 glx/dri2: initialise api to avoid indirect rendering failing randomly 66006 66007 Running glxinfo under indirect rendering would randomly fail against the 66008 intel driver, as it would create a context with no attribs, and then the 66009 api value would be passed to the driver uninitialised. 66010 66011 Signed-off-by: Dave Airlie <airlied@redhat.com> 66012 Reviewed-by: Keith Packard <keithp@keithp.com> 66013 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 66014 Signed-off-by: Keith Packard <keithp@keithp.com> 66015 66016commit cc10ac8f0e07854647e1fd7cc70b7e9f8d919fd1 66017Author: Peter Hutterer <peter.hutterer@who-t.net> 66018Date: Mon Dec 10 11:00:46 2012 +1000 66019 66020 Xi: fix swapping for barrier events 66021 66022 Protocol events don't contain pointers, so it's easier to copy everything 66023 over, then swap in-place. 66024 66025 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66026 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66027 66028commit 151d44149a09dd125e25b3d94f22a609f0221548 66029Author: Peter Hutterer <peter.hutterer@who-t.net> 66030Date: Mon Dec 10 10:53:49 2012 +1000 66031 66032 Xi: swap sequence number and evtype in barrier events 66033 66034 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66035 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66036 66037commit d4065493b211e90b46f9d4178f23f347afd8043f 66038Author: Peter Hutterer <peter.hutterer@who-t.net> 66039Date: Mon Dec 10 11:02:32 2012 +1000 66040 66041 tests/xi2: at protocol conversion test for barrier events 66042 66043 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66044 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66045 66046commit 938187f2fec006daf4cc677df26d5b0b6999b54b 66047Author: Peter Hutterer <peter.hutterer@who-t.net> 66048Date: Fri Dec 7 15:55:59 2012 +1000 66049 66050 Require inputproto 2.2.99.1 66051 66052 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66053 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66054 66055commit 78376a9217058aa66e52a0399fc2be469d70ce32 66056Author: Peter Hutterer <peter.hutterer@who-t.net> 66057Date: Fri Dec 7 09:26:13 2012 +1000 66058 66059 mi: rename mipointer's internal event list 66060 66061 Avoid name shadowing warnings, change the event list to a more specific 66062 name. 66063 66064 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66065 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66066 66067commit 88a2cccc37cac4e171f9bfc020017ddfa66ae417 66068Author: Peter Hutterer <peter.hutterer@who-t.net> 66069Date: Fri Dec 7 09:11:13 2012 +1000 66070 66071 Xi: if the device is currently grabbed, flag the barrier event 66072 66073 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66074 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66075 66076commit 353aa515922e1095047161ec47a2722772218f20 66077Author: Peter Hutterer <peter.hutterer@who-t.net> 66078Date: Thu Dec 6 15:09:27 2012 +1000 66079 66080 Xi: deliver barrier events as grabbed events where necessary 66081 66082 If the grab_window is the barrier window and the client owns the grab, 66083 deliver as normal grabbed event (respecting owner_events). Otherwise, 66084 deliver as usual. 66085 66086 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66087 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66088 66089commit 3b161401700a2d916da0f81f99b39e75fdbe78df 66090Author: Peter Hutterer <peter.hutterer@who-t.net> 66091Date: Thu Dec 6 14:13:26 2012 +1000 66092 66093 dix: ignore barrier events in FixUpEventFromWindow 66094 66095 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66096 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66097 66098commit a1eeb6fbecf0bd9665a9158b27c273e83b3ab351 66099Author: Peter Hutterer <peter.hutterer@who-t.net> 66100Date: Thu Dec 6 13:25:29 2012 +1000 66101 66102 dix: handle barrier events properly when converting to core/XI 1.x 66103 66104 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66105 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66106 66107commit 1b83775f6712a7e48c49a3411a87103e8e906e67 66108Author: Peter Hutterer <peter.hutterer@who-t.net> 66109Date: Thu Dec 6 12:39:04 2012 +1000 66110 66111 dix: skip delivery if it's not the right pointer barrier client 66112 66113 Only deliver to the client that created the barrier, not to other clients. 66114 66115 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66116 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66117 66118commit c50db6faba4ee1c27b735c6e9c98a4276ba3c7ff 66119Author: Peter Hutterer <peter.hutterer@who-t.net> 66120Date: Thu Dec 6 12:58:55 2012 +1000 66121 66122 Xi: fill in barrier root x/y after clamping to RandR outputs 66123 66124 x/y for barrier events should contain the actual pointer position. 66125 66126 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66127 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66128 66129commit 21a15f9a04ec0a6c8f654eef943561e98db2475d 66130Author: Peter Hutterer <peter.hutterer@who-t.net> 66131Date: Mon Dec 3 10:41:16 2012 +1000 66132 66133 Pass the event list through to the pointer barrier code to return it 66134 66135 Instead of having the pointer barrier code enqueue events separately from 66136 GetPointerEvents, pass the event list through and let it add to it. 66137 66138 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66139 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66140 66141commit 707b4dc61f18960611409ef5ad8947be189f7296 66142Author: Jasper St. Pierre <jstpierre@mecheye.net> 66143Date: Tue Nov 27 15:31:52 2012 -0500 66144 66145 barriers: Support line and ray barriers 66146 66147 This allows clients to add barriers that extend to the edge of the 66148 screen. Clients are encouraged to use these instead of precise coordinates 66149 in these cases to help prevent pointer leaks. 66150 66151 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66152 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66153 66154commit 0a5a0d7c242add4d79721c1b90073b603a475534 66155Author: Jasper St. Pierre <jstpierre@mecheye.net> 66156Date: Tue Nov 27 14:29:52 2012 -0500 66157 66158 barriers: Replace complex intersection test with simpler math 66159 66160 Since barriers are axis-aligned, we can do the intersection test with 66161 simple interpolation rather than line-segment intersection. This also 66162 helps us out in the future when we want the barriers to extend to be 66163 rays and lines rather than just segments. 66164 66165 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66166 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66167 66168commit 6401317bdcf58b2717214ac65c8f47b0c384a7db 66169Author: Jasper St. Pierre <jstpierre@mecheye.net> 66170Date: Sun Dec 9 19:45:48 2012 -0500 66171 66172 barriers: Send an XI_BarrierLeave event when a barrier is destroyed 66173 66174 This ensures that we always complete an event sequence. 66175 66176 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66177 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66178 66179commit 207e8dee00dd26ff4363edb72058b48b8a76b824 66180Author: Jasper St. Pierre <jstpierre@mecheye.net> 66181Date: Sun Dec 9 19:44:13 2012 -0500 66182 66183 barriers: Clean up code 66184 66185 Rename a variable. This is to make the diff in the next commit cleaner. 66186 66187 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66188 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66189 66190commit 7fabecd8e3efe76caeb740d905a02ae4256e3246 66191Author: Jasper St. Pierre <jstpierre@mecheye.net> 66192Date: Wed Dec 5 19:58:05 2012 -0500 66193 66194 barriers: Send a BarrierLeave event when we leave the hitbox 66195 66196 Additionally, add flags when the pointer is released. 66197 66198 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66199 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66200 66201commit dac9e13a6c874f3b99ba4cf9085ed0e63beb0871 66202Author: Jasper St. Pierre <jstpierre@mecheye.net> 66203Date: Mon Nov 26 20:37:56 2012 -0500 66204 66205 barriers: Increment event ID on hit box leave 66206 66207 We eventually want to send a new notify event on hitbox leave, 66208 which signifies the dawn of a new barrier event ID, so it's 66209 convenient if we can put the code here. 66210 66211 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66212 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66213 66214commit e3a734d081ebda4460e6c1eb7dcdd0f7c10ba9b4 66215Author: Jasper St. Pierre <jstpierre@mecheye.net> 66216Date: Sat Nov 24 21:42:56 2012 -0500 66217 66218 barriers: Add a couple pixels of elbow room for the hit detection 66219 66220 Pointers (and the hands that drive them) aren't very precise, and the 66221 slightest amount of nudging to either side might be enough to reset 66222 the event ID, making clients think they have an entirely new hit. Allow 66223 for a slightly bigger "hit box" before these barriers get reset. 66224 66225 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66226 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66227 66228commit 3b2fbcfa6c75ab072926ab9659adf785b324ce28 66229Author: Jasper St. Pierre <jstpierre@mecheye.net> 66230Date: Thu Nov 8 01:17:40 2012 -0500 66231 66232 barriers: Add support for edge cases when releasing barriers 66233 66234 If we release a barrier, we want to ensure that we block all 66235 other barriers afterwards, rather than capping the limit to 66236 the two nearest barriers. 66237 66238 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66239 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66240 66241commit a1ee26e624e6a7030ecb37a608190cb8e3d1f65b 66242Author: Peter Hutterer <peter.hutterer@who-t.net> 66243Date: Thu Dec 6 12:06:28 2012 +1000 66244 66245 barriers: Don't allow releasing the pointer on other client's barriers 66246 66247 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66248 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66249 66250commit 2884b249294e3e8fc749bbb81e2a897314fa0818 66251Author: Peter Hutterer <peter.hutterer@who-t.net> 66252Date: Thu Dec 6 11:59:53 2012 +1000 66253 66254 barriers: Don't allow destroying other client's barriers 66255 66256 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66257 Reviewed-by: Jasper St. Pierre <jstpierre@mecheye.net> 66258 66259commit e130a46ab45839f91b29ee75bf19a597dfe3da28 66260Author: Jasper St. Pierre <jstpierre@mecheye.net> 66261Date: Thu Nov 1 16:06:40 2012 -0400 66262 66263 Add support for XI2.3: Pointer barrier events and releases. 66264 66265 This adds support for clients that would like to get a notification 66266 every time a barrier is hit, and allows clients to temporarily release 66267 a barrier so that pointers can go through them, without having to 66268 destroy and recreate barriers. 66269 66270 Based on work by Chris Halse Rogers <chris.halse.rogers@canonical.com> 66271 66272 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66273 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66274 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66275 66276commit 85a37ddcc2ff49adaec4075532b60e7d015f901c 66277Author: Jasper St. Pierre <jstpierre@mecheye.net> 66278Date: Sun Dec 9 22:48:18 2012 -0500 66279 66280 barriers: Reindent the constrainment hook 66281 66282 This is to make future diffs much cleaner. Best viewed with -w. 66283 66284 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66285 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66286 66287commit 97da74c80e8b8eb03f7c1e34180cf5cd343bd608 66288Author: Jasper St. Pierre <jstpierre@mecheye.net> 66289Date: Sun Dec 9 22:19:14 2012 -0500 66290 66291 barriers: Switch to finding the nearest barrier client 66292 66293 When we add events, we eventually want to add more state to the 66294 PointerBarrierClient, so return one of these instead of the dummy 66295 public structure that's not very interesting. 66296 66297 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66298 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66299 66300commit 7e16dd3628334a5991b6713d778a46c1ce3b8b78 66301Author: Jasper St. Pierre <jstpierre@mecheye.net> 66302Date: Sat Nov 24 19:39:47 2012 -0500 66303 66304 barriers: Switch to an explicit hook for barrier constrainment 66305 66306 Rather than riding on the ConstrainCursorHarder hook, which has 66307 several issues, move to an explicit hook, which will help us with 66308 some RANDR interaction issues. 66309 66310 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66311 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66312 66313commit 2868a93945f043e1efd897d56543fe673e341faf 66314Author: Jasper St. Pierre <jstpierre@mecheye.net> 66315Date: Sun Dec 9 18:41:52 2012 -0500 66316 66317 barriers: Don't loop over the server to destroy a barrier 66318 66319 This is completely pointless as far as I can tell. 66320 66321 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66322 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66323 66324commit 482e0cb35225dc3b0664f9429db19c61b0c98ef6 66325Author: Jasper St. Pierre <jstpierre@mecheye.net> 66326Date: Thu Nov 1 14:14:13 2012 -0400 66327 66328 cursor: Move pointer barrier code over to XI 66329 66330 In order to send events to specific windows associated with the barrier, 66331 we need to move the code that handles barriers to somewhere where it's 66332 easier to construct and send events. Rather than duplicating XSync with 66333 its XSyncSelectAlarm, re-use the existing XI infrastructure. 66334 66335 For now, just move a bunch of code over, rename some things, and initialize 66336 the new structures, but still consider it a separate codebase. Pointer barrier 66337 requests are still handled by XFixes, so this is a weird intermediate state. 66338 It's unknown whether we'll add explicit requests to pointer barriers inside 66339 XI. 66340 66341 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66342 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66343 66344commit bb6f3514ca17d993c1af380e8d4480d61e5bbcae 66345Merge: f961c3a3b 08da994a0 66346Author: Peter Hutterer <peter.hutterer@who-t.net> 66347Date: Wed Dec 12 17:25:28 2012 +1000 66348 66349 Merge branch 'stack-smash-on-touchpoint' into for-keith 66350 66351commit f961c3a3b9dfbe1201da317c24797ba7f979731e 66352Merge: 36740d02b 39f19b3f3 66353Author: Peter Hutterer <peter.hutterer@who-t.net> 66354Date: Wed Dec 12 17:25:23 2012 +1000 66355 66356 Merge branch 'touch-selection-conflict-fixes' into for-keith 66357 66358commit 39f19b3f3b8c9b714e70e339dfb0083ff629ab2a 66359Author: Peter Hutterer <peter.hutterer@who-t.net> 66360Date: Tue Nov 20 11:48:31 2012 +1000 66361 66362 Xi: fix touch event selction conflicts (#57301) 66363 66364 There are limits on which client may select for touch events on a given 66365 window, with restrictions being that no two clients can select on the same 66366 device, but narrower selections are allowed, i.e. if one client has 66367 XIAllDevices, a second client may still select for device X. 66368 66369 The current code had a dependency on which client selected first and which 66370 device, resulting in inconsistencies when selecting for events. Fix that, 66371 responding with the right errors regardless of who selected what first. 66372 66373 X.Org Bug 57301 <http://bugs.freedesktop.org/show_bug.cgi?id=57301> 66374 66375 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66376 66377commit a7c97d737ef0d14ec97869082decd049637cfe7a 66378Author: Peter Hutterer <peter.hutterer@who-t.net> 66379Date: Tue Nov 20 11:38:39 2012 +1000 66380 66381 dix: split xi2_mask_isset into a per-device function 66382 66383 For touch selection conflicts, we need to check not only if the mask is set 66384 for the device, but if it is set for only that specific device (regardless 66385 of XIAll*Devices) 66386 66387 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66388 66389commit 08da994a08bb74afae81176c56fb525d0439274b 66390Author: Peter Hutterer <peter.hutterer@who-t.net> 66391Date: Mon Nov 26 12:33:29 2012 +1000 66392 66393 dix: add FIXME, TouchRemovePointerGrab does nothing 66394 66395 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66396 66397commit 00def5144557cfe8bf535f926212a8e084dc7cf6 66398Author: Peter Hutterer <peter.hutterer@who-t.net> 66399Date: Mon Nov 26 14:55:13 2012 +1000 66400 66401 Xi: if a TouchEnd appears on a actively grabbing client, always accept 66402 66403 Once the TouchEnd appears on the device, the touch is done. If the client 66404 still has a pointer grab, accept it to avoid clients with TouchOwnership 66405 selections to wait indefinitely for the actual touch event. 66406 66407 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66408 66409commit ece8157a59751b3ed492fb2e1eb8d5f20221e195 66410Author: Peter Hutterer <peter.hutterer@who-t.net> 66411Date: Mon Nov 26 15:14:19 2012 +1000 66412 66413 dix: when deactivating pointer-only grabs, don't emulate TouchEnd events 66414 66415 A client with a pointer grab on a touch device must reject the touch when 66416 detactivating the grab while the touch is active. However, such a rejecting 66417 must not trigger a ButtonRelease event to be emulated and sent to the 66418 client. 66419 Set the grabbing listener's state to HAS_END, so we simply skip delivery to 66420 that client. 66421 66422 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66423 66424commit bc1f90a615018c05994fae3e678dd2341256cd82 66425Author: Peter Hutterer <peter.hutterer@who-t.net> 66426Date: Mon Nov 26 12:23:54 2012 +1000 66427 66428 dix: only reject active grabs on ungrab and do it before actually ungrabbing 66429 66430 An active grab ungrabbing is the same as rejecting the grab, since the 66431 client is no longer interested in those events. So reject any touch grab, 66432 but do so before actually deactivating since we're interested in the 66433 TouchEnd for the current grabbing client. 66434 66435 A passive grab otoh is _not_ like rejecting a grab, since it deactivates 66436 automatically when the touch ends. 66437 66438 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66439 66440commit 146f48c2934fc85ec095496da5c8f0102bc7f5b5 66441Author: Peter Hutterer <peter.hutterer@who-t.net> 66442Date: Thu Nov 22 13:49:34 2012 +1000 66443 66444 dix: don't call ProcessInputEvents() when accepting/rejecting touches 66445 66446 TouchListenerAcceptReject may be called during normal event processing, but 66447 ProcessInputEvents is not reentrant and calling it here smashes the event 66448 queue. 66449 66450 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66451 66452commit ead21f9426122536adfb4787ac181008ae83cd4b 66453Author: Peter Hutterer <peter.hutterer@who-t.net> 66454Date: Mon Nov 19 16:16:10 2012 +1000 66455 66456 Xi: fix typo "mechansims" → "mechanisms" 66457 66458 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66459 66460commit 6b4aa8a359822cc70666de3b18744f05dd9b5f45 66461Author: Dave Airlie <airlied@redhat.com> 66462Date: Wed Dec 12 12:56:46 2012 +1000 66463 66464 Revert "xf86: Fix non-PCI configuration-less setups" 66465 66466 This reverts commit 76d9c62eb2be2010a19bf36285012d086cdd180b. 66467 66468 This breaks multi-GPU setups here, so lets drop it for now. 66469 66470 Signed-off-by: Dave Airlie <airlied@redhat.com> 66471 66472commit 785af88ab0120036e0ce3d0139f3c560ff71e10b 66473Author: Dave Airlie <airlied@redhat.com> 66474Date: Wed Sep 26 16:16:40 2012 +1000 66475 66476 dri1: fix dri1 startup since 459c6da0f907ba33d733c7e62a116184ba2f14e5 66477 66478 This commit regresses dri1 since it moves the drmSetServerInfo from being 66479 called at module load time to extension init time. However DRIScreenInit 66480 relies on this being called before it gets control. 66481 66482 This patches moves the call into DRIScreenInit and seems to work here. 66483 66484 Signed-off-by: Dave Airlie <airlied@redhat.com> 66485 66486commit 36740d02b9ca117f1404e077367fbbbe271a17d6 66487Author: Daniel Martin <consume.noise@gmail.com> 66488Date: Fri Dec 7 19:38:55 2012 +0100 66489 66490 xfixes: Fix minor number in QueryVersion 66491 66492 Due to a typo the major version number was passed as minor version to 66493 version_compare(). 66494 66495 Regression-from: ffd4874798ba54f86acac75779a15b4babeaa5f3 66496 66497 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 66498 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66499 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66500 66501commit b34cffd27b9ef46e1dd477219cbb90e36d5609e3 66502Author: Peter Hutterer <peter.hutterer@who-t.net> 66503Date: Mon Dec 3 11:39:08 2012 +1000 66504 66505 include: fix comment 66506 66507 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66508 66509commit ea1d76d1b68c7cbcedcaa1f967c5027e0cc1c7a2 66510Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66511Date: Mon Nov 5 23:50:58 2012 -0600 66512 66513 Fix formatting of address operators 66514 66515 The formatter confused address operators preceded by casts with 66516 bitwise-and expressions, placing spaces on either side of both. 66517 That syntax isn't used by ordinary address operators, however, 66518 so fix them for consistency. 66519 66520 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66521 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66522 66523commit 5daa442fe15d9a87112a2def673c99a5f2506dcf 66524Author: Peter Hutterer <peter.hutterer@who-t.net> 66525Date: Tue Dec 4 15:28:14 2012 +1000 66526 66527 xkb: only post a XTest release if the XTest device has the button down 66528 66529 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66530 66531commit c4fee9d2ecd6fdd4c05b04c9f02b23f10f2938b8 66532Author: Peter Hutterer <peter.hutterer@who-t.net> 66533Date: Tue Dec 4 15:03:37 2012 +1000 66534 66535 xkb: always post XTest button up when the physical button released (#28808) 66536 66537 Regression introduced by commit 2decff6393a44b56d80d53570718f95354fde454 66538 xkb: ProcesssPointerEvent must work on the VCP if it gets the VCP 66539 66540 XTest buttons must be released when a physical button is released. This was 66541 fixed in 14327858391ebe929b806efb53ad79e789361883, but 66542 2decff6393a44b56d80d53570718f95354fde454 changed a condition that this code 66543 didn't get triggered anymore. 66544 66545 "dev" for pointer events is now always the VCP which doesn't have a xkbi 66546 struct. So move this condition out and always trigger the XTest released for 66547 button events. 66548 66549 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66550 Tested-by: Frank Roscher <Frank-Roscher@gmx.net> 66551 66552commit b51a1bd2766e7dc975ca8f1cacc3f8bd0e1a68a3 66553Merge: e9ea96d6a ce6b65292 66554Author: Keith Packard <keithp@keithp.com> 66555Date: Fri Nov 30 09:31:46 2012 -0800 66556 66557 Merge remote-tracking branch 'whot/for-keith' 66558 66559commit e9ea96d6a88f197928c82b008953f1810de6f792 66560Author: Dave Airlie <airlied@gmail.com> 66561Date: Fri Nov 30 20:54:38 2012 +1000 66562 66563 randr: call RRProviderInit in the proper place. 66564 66565 No idea where this got lost across development cycles, but its 66566 definitely missing. 66567 66568 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=57448 66569 66570 Signed-off-by: Dave Airlie <airlied@redhat.com> 66571 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 66572 Signed-off-by: Keith Packard <keithp@keithp.com> 66573 66574commit ce6b6529290f7d42717aad3fd8a83aaaaf8962de 66575Merge: 6e18599d6 3b9f1c701 66576Author: Peter Hutterer <peter.hutterer@who-t.net> 66577Date: Thu Nov 29 14:49:22 2012 +1000 66578 66579 Merge branch 'high-resolution-touch-devices' into for-keith 66580 66581commit 6e18599d691036eca6ff082c17727d9ffb926984 66582Author: Sybren van Elderen <sowmestno@msn.com> 66583Date: Tue Nov 27 19:27:10 2012 +0100 66584 66585 hw/dmx: add update_desktop_dimensions() call 66586 66587 The Xdmx server did not update the desktop dimensions when computing screen 66588 origins. 66589 66590 Signed-off-by: Sybren van Elderen <sowmestno@msn.com> 66591 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66592 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66593 66594commit 59d70b30e93f8be93db9e873423353f7e4419811 66595Author: Peter Hutterer <peter.hutterer@who-t.net> 66596Date: Fri Nov 23 10:05:02 2012 +1000 66597 66598 dix: use pixman for fp1616 conversions 66599 66600 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66601 Reviewed-by: Søren Sandmann <ssp@redhat.com>> 66602 66603commit 2c4388a00ec308bc2d48ba751ff510cd5c1b2384 66604Author: Peter Hutterer <peter.hutterer@who-t.net> 66605Date: Tue Jun 26 10:30:35 2012 +1000 66606 66607 xkb: fill in keycode and event type for slow keys enablement 66608 66609 eventType is set for the type that triggered a XkbControlsNotify event. 66610 Technically, SlowKeys is triggered by a timer which doesn't have a matching 66611 core event type. So we used to use 0 here. 66612 66613 Practically, the timer is triggered by a key press + hold and cancelled when 66614 the key is released before the timeout expires. So we might as well set 66615 KeyPress (keycode) in the ControlsNotify to give clients a chance to differ 66616 between timer-triggered SlowKeys and client-triggered ones. 66617 66618 This is a chance in behaviour, though I suspect with little impact. 66619 66620 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66621 Acked-by: Daniel Stone <daniel@fooishbar.org> 66622 66623commit 2dc6d9228456cbc274a1fcea478258e5ae4dc629 66624Author: Peter Hutterer <peter.hutterer@who-t.net> 66625Date: Thu Nov 15 14:00:57 2012 +1000 66626 66627 When resetting device idle time, reset XIAll(Master)Devices too (#56649) 66628 66629 When the screen saver is forcibly deactivated, the idle time counter is 66630 reset for all devices but not for the fake XIAllDevices and 66631 XIAllMasterDevices. XScreenSaverQueryInfo uses XIAlldevices to fill the 66632 "idle" field, thus returning the wrong value. 66633 66634 Regression introduced in 66635 commit 6aef209ebc2e54f5465da505a780f7b4cc273ee0 66636 Author: Peter Hutterer <peter.hutterer@who-t.net> 66637 Date: Mon Mar 12 13:51:02 2012 +1000 66638 66639 Change lastDeviceIdleTime to be per-device 66640 66641 X.Org Bug 56649 <http://bugs.freedesktop.org/show_bug.cgi?id=56649> 66642 66643 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66644 Tested-by: Giacomo Perale <ghepeu@virgilio.it> 66645 Reviewed-by: Keith Packard <keithp@keithp.com> 66646 66647commit a51b2c3913fc8556f6bd1c76805d045fc424c4bb 66648Author: Jasper St. Pierre <jstpierre@mecheye.net> 66649Date: Tue Nov 20 14:50:41 2012 -0500 66650 66651 cursor: Revise edge cases for the pointer moving towards barriers 66652 66653 Since barriers block the invisible line between pixels, that means 66654 that we need to explicitly check the boundaries, or else we'll have 66655 a potential off-by-one error. This fixes issues when trying to move 66656 down or right across a barrier and having the pointer visibly bounce. 66657 66658 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 66659 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 66660 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 66661 66662commit 0b198248ecfbcb7e7c61688145be4005366da23d 66663Author: Thierry Reding <thierry.reding@avionic-design.de> 66664Date: Thu Nov 8 14:28:10 2012 +0100 66665 66666 Remove call to miInitializeBackingStore() 66667 66668 Recent versions of the X server no longer provide this function, which 66669 has been obsolete for over 2 years now. 66670 66671 Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de> 66672 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 66673 Signed-off-by: Dave Airlie <airlied@redhat.com> 66674 66675commit 739d6c067a9167226ce248aa82997bfc2cca7259 66676Author: Thierry Reding <thierry.reding@avionic-design.de> 66677Date: Thu Nov 8 14:28:09 2012 +0100 66678 66679 build: Put m4 macros in m4 subdirectory 66680 66681 This silences a warning from libtoolize when running the autogen.sh 66682 script. 66683 66684 Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de> 66685 Signed-off-by: Dave Airlie <airlied@redhat.com> 66686 66687commit 1712a45422a63f11b2146541279616fcfda09ec6 66688Merge: e54f71a2c 338bec300 66689Author: Keith Packard <keithp@keithp.com> 66690Date: Wed Nov 28 07:54:13 2012 -0800 66691 66692 Merge remote-tracking branch 'jturney/mingw-patches' 66693 66694commit 338bec30054bb5d41cb6fdb5d37439f3c8124cad 66695Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66696Date: Fri Nov 9 12:19:31 2012 -0600 66697 66698 hw/xwin: Fix for MinGW-w64 DirectDraw headers 66699 66700 mingw-w64 headers handle NONAMELESSUNION earlier than mingw.org's, so it must be 66701 defined before including any headers. It also provides a ddraw.h, so use it. 66702 66703 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66704 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66705 Reviewed-by: Keith Packard <keithp@keithp.com> 66706 66707commit fb170498abc746a850864b59db2ddcba7ee29215 66708Author: Jon TURNEY <jon.turney@dronecode.org.uk> 66709Date: Tue Nov 13 13:22:44 2012 +0000 66710 66711 dix/dispatch.c, os/utils.c: Disable smart scheduler on WIN32 66712 66713 setitimer() and SIGALRM aren't available on WIN32, so smart scheduler 66714 code cannot be built. Provide only stubs for smart scheduler timer 66715 code, and disable smart scheduler by default. 66716 66717 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 66718 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66719 Tested-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66720 Reviewed-by: Keith Packard <keithp@keithp.com> 66721 66722commit b20d2998cd9a1320ef0e45ab6b741c7e25f646e0 66723Author: Ryan Pavlik <rpavlik@iastate.edu> 66724Date: Thu Nov 3 19:25:57 2011 -0500 66725 66726 os/osinit.c: no getpgrp() and setpgrp() on WIN32 66727 66728 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 66729 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66730 Tested-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66731 Reviewed-by: Keith Packard <keithp@keithp.com> 66732 66733commit 2c712d094a49de0b96e2e9f5bcc1043b54ad0f61 66734Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66735Date: Sun Oct 14 23:05:30 2012 -0500 66736 66737 hw/xwin: use raise() instead of kill() 66738 66739 MinGW doesn't have kill(), so use raise() instead 66740 66741 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66742 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66743 Reviewed-by: Keith Packard <keithp@keithp.com> 66744 66745commit 98d5acc1213b2ec71c2a7bd3a3b77a1b5536e963 66746Author: Jon TURNEY <jon.turney@dronecode.org.uk> 66747Date: Fri Aug 17 13:43:29 2012 +0100 66748 66749 os/utils.c: Fix compilation of OsBlockSIGIO when SIGIO isn't defined 66750 66751 Fix compilation of OsBlockSIGIO with -Werror=return-type when SIGIO isn't 66752 defined. 66753 66754 /jhbuild/checkout/xorg/xserver/os/utils.c: In function 'OsBlockSIGIO': 66755 /jhbuild/checkout/xorg/xserver/os/utils.c:1248:1: error: control reaches end of non-void function [-Wreturn-type] 66756 66757 v2: Shuffle around to avoid writing unreachable code 66758 66759 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66760 Tested-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66761 Reviewed-by: Keith Packard <keithp@keithp.com> 66762 66763commit 8e2bac0a69173b51e17a39d70e1d154b47fd0139 66764Author: Ryan Pavlik <rpavlik@iastate.edu> 66765Date: Fri Apr 20 13:46:21 2012 +0100 66766 66767 os/osinit.c, os/utils.c: Exclude sigaction code when building for MinGW 66768 66769 MinGW doesn't have sigaction, so this patch is needed for building. 66770 66771 No attempt is made to actually install the fatal error signal handler, as MinGW 66772 will simply terminate the process rather than deliver a fatal signal. 66773 66774 Also avoid using strsignal 66775 66776 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 66777 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66778 Tested-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66779 Reviewed-by: Keith Packard <keithp@keithp.com> 66780 66781commit e54f71a2c7a1677a5c5032d1e7dc099d5204144b 66782Author: Chris Wilson <chris@chris-wilson.co.uk> 66783Date: Sat Nov 10 10:26:08 2012 +0000 66784 66785 xf86: select a fake output for headless servers 66786 66787 Following commit 37d956e3ac9513b74078882dff489f9b0a7a5a28 66788 Author: Keith Packard <keithp@keithp.com> 66789 Date: Mon Sep 10 11:14:20 2012 +1000 66790 66791 xf86: fix compat output selection for no output GPUs 66792 66793 headless servers can no longer startup as we no longer select a compat 66794 output for the fake framebuffer. 66795 66796 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=56343 66797 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 66798 Reviewed-by: Dave Airlie <airlied@redhat.com> 66799 Signed-off-by: Keith Packard <keithp@keithp.com> 66800 66801commit 76d9c62eb2be2010a19bf36285012d086cdd180b 66802Author: Thierry Reding <thierry.reding@avionic-design.de> 66803Date: Wed Oct 17 12:06:47 2012 +0200 66804 66805 xf86: Fix non-PCI configuration-less setups 66806 66807 For non-PCI video devices, such as those found on many ARM embedded 66808 systems, the X server currently requires the BusID option to specify the 66809 full path to the DRM device's sysfs node in order to properly match it 66810 against the probed platform devices. 66811 66812 In order to allow X to start up properly if either the BusID option was 66813 omitted or no configuration is present at all, the first video device is 66814 used by default. 66815 66816 Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de> 66817 Reviewed-by: Dave Airlie <airlied@redhat.com> 66818 Signed-off-by: Keith Packard <keithp@keithp.com> 66819 66820commit d50a94568e06a06df682a0028579d37a7b209d15 66821Merge: 87b60efb8 cf05db8dd 66822Author: Keith Packard <keithp@keithp.com> 66823Date: Tue Nov 27 15:14:26 2012 -0800 66824 66825 Merge remote-tracking branch 'daniels/master' 66826 66827commit 87b60efb859d86fd9e96306112bc5453a549fa69 66828Author: Colin Harrison <colin.harrison@virgin.net> 66829Date: Sun Sep 2 22:39:32 2012 +0100 66830 66831 hw/xwin: Don't spam MotionNotify events when the mouse hasn't moved 66832 66833 Don't spam MotionNotify events when the mouse hasn't moved, when polling the 66834 mouse position outside any X window 66835 66836 (Test with 'xev -root' after mouse polling has started.) 66837 66838 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 66839 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66840 66841commit 886f4c389266bf759414f15a5e25e50b5935dc9d 66842Author: Jon TURNEY <jon.turney@dronecode.org.uk> 66843Date: Mon Jun 25 15:58:21 2012 +0100 66844 66845 hw/xwin: Don't log all fbConfigs and GL/WGL extensions 66846 66847 Don't log all fbConfigs and GL/WGL extensions, unless verbose logging 66848 is requested 66849 Log the number of pixelFormats which gave rise to the fbConfigs 66850 66851 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66852 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66853 66854commit 9b10f20f472e6b0d6114fb61c88906e349e47f2b 66855Author: Jon TURNEY <jon.turney@dronecode.org.uk> 66856Date: Fri Feb 10 13:36:46 2012 +0000 66857 66858 hw/xwin: Use pre-computed extent of damage for creating GDI region 66859 66860 When -clipupdates option is specified, use the pre-computed extent of damage, 66861 rather than computing a GDI region which combines all the individual boxes in 66862 the damage. 66863 66864 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66865 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66866 66867commit 3ab95a2a5c848e8ce4c39be243fd5e2ac6bb1c1a 66868Author: Jon TURNEY <jon.turney@dronecode.org.uk> 66869Date: Fri Feb 24 15:32:26 2012 +0000 66870 66871 hw/xwin: Ensure we own the clipboard before checking the format of it's contents 66872 66873 Ensure we own the clipboard before checking the format of it's contents, this 66874 prevents the contents from changing underneath us. 66875 66876 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66877 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66878 66879commit a4f357c620ae9e2270316197a61fb8e1d5183a51 66880Author: Colin Harrison <colin.harrison@virgin.net> 66881Date: Mon Apr 9 15:48:08 2012 +0100 66882 66883 hw/xwin: Disable minimize button on window with skip-taskbar state 66884 66885 If a window has had its taskbar button removed, disable its minimize 66886 button to prevent it becoming lost 66887 66888 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 66889 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66890 66891commit a2983452ee1e2051ab54cded7a3a7735be020db6 66892Author: Jon TURNEY <jon.turney@dronecode.org.uk> 66893Date: Wed Jul 8 17:48:03 2009 +0100 66894 66895 hw/xwin: Process _NET_WM_STATE_SKIP_TASKBAR hint in multiwindow mode. 66896 66897 Set WS_EX_TOOLWINDOW style to hide window from Alt-Tab switcher 66898 66899 Use ITaskBarList interface to ensure that the taskbar notices if the window has 66900 changed it's style in a way which affects if the taskbar shows it or not 66901 66902 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66903 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66904 66905commit de759cc38ccc8a7811923eb9620f01301aedc10e 66906Author: Jon TURNEY <jon.turney@dronecode.org.uk> 66907Date: Sun Feb 5 17:57:04 2012 +0000 66908 66909 hw/xwin: Add a simple interface to the ITaskbarList COM interface 66910 66911 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66912 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66913 66914commit 19d875187a113ca7772a770767ea2ceb2ade5a43 66915Author: Jon TURNEY <jon.turney@dronecode.org.uk> 66916Date: Wed Jan 4 17:40:14 2012 +0000 66917 66918 hw/xwin: Enable RANDR resize by default 66919 66920 Also fix erroneous use of '--resize' not '-resize', and document '-noresize' in man page 66921 66922 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66923 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66924 66925commit 444580673221bf3f89dfa3cbc518b5ed4ecabc23 66926Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66927Date: Sat Dec 31 22:11:18 2011 -0600 66928 66929 hw/xwin: Enable AIGLX by default 66930 66931 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 66932 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66933 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66934 66935commit 11bb32e561b3f1c657a99b3902f9beea786babbb 66936Author: Oliver Schmidt <oschmidt-mailinglists@gmx.de> 66937Date: Sun Jan 8 20:30:02 2012 +0000 66938 66939 hw/xwin: Restore non-latching modifier key state when an X window gains focus 66940 66941 In multiwindow mode, the state of the modifier keys was lost when a window is 66942 created (or raised) and focus moved to that window. 66943 66944 For example: In window A Ctrl + some key opens a window B, then in window B Ctrl 66945 + some other key triggers the next action. However after the opening of window B 66946 the Ctrl key has to be released and pressed again. If the user keeps the Ctrl 66947 key held down when the window B is opened, the next key press X will be 66948 interpreted as X and not as Ctrl+X. 66949 66950 Extended the function winRestoreModeKeyStates in winkeybd.c to consider not only 66951 the latching modifier keys but also the modifiers Ctrl, Shift, Alt/AltGr by 66952 using the Windows function GetAsyncKeyState. 66953 66954 A combined Ctrl+AltGr modifier state cannot be restored correctly, as Windows 66955 always fakes a Ctrl-L when AltGr is pressed. 66956 66957 Signed-off-by: Oliver Schmidt <oschmidt-mailinglists@gmx.de> 66958 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66959 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66960 66961commit 47291d0b7d7bfee74156f83badae15a0818c68ad 66962Author: Jon TURNEY <jon.turney@dronecode.org.uk> 66963Date: Mon Nov 5 14:54:51 2012 +0000 66964 66965 hw/xwin: Give our logical xor operator a more logical name 66966 66967 Also, rather than a comment about why we need a logical operator, let's have a 66968 comment about what we are doing to the keyboard state... 66969 66970 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66971 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66972 66973commit 8aa27ae82109e4fab0ff3ed86ad1d152438a2585 66974Author: Tobias Häußler <tobias.haeussler@gmx.de> 66975Date: Mon Jul 18 14:48:23 2011 +0100 66976 66977 hw/xwin: Add correct taskbar grouping of X windows on Windows 7 66978 66979 I created a small patch for XWin that adds correct grouping of taskbar icons 66980 when 'Always combine, hide labels' is set in the taskbar properties. It uses the 66981 new taskbar APIs introduced in Windows 7 to set an application id for each 66982 window. The id is based on the X11 class hints. 66983 66984 v2: Add file to _SOURCES to fix distcheck 66985 66986 v3 : Fix compilation with mingw-w64 w32api headers 66987 Include propkey.h, propsys.h rather than defining necessary stuff ourselves 66988 66989 v4: Fix up names taskbar->propertystore, AppID->AppUserModelID, etc. 66990 Link directly with ole32 for PropVariantClear(), prototyping it if neccessary. 66991 66992 v5: Put winSetAppUserModelID()-related code in a separate file. 66993 Drop a superfluous assign to hr of ignored HRESULT of SetValue() 66994 66995 Signed-off-by: Tobias Häußler <tobias.haeussler@gmx.de> 66996 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 66997 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 66998 66999commit f3aef7f9561d2723da0d1438a8b276b77a1e672e 67000Author: Jon TURNEY <jon.turney@dronecode.org.uk> 67001Date: Sun Sep 2 16:04:25 2012 +0100 67002 67003 hw/xwin: Fix function name in log message 67004 67005 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67006 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 67007 67008commit 127a7068b75258d396776cc101aba0bef10a5cef 67009Author: Jon TURNEY <jon.turney@dronecode.org.uk> 67010Date: Thu Apr 14 23:10:23 2011 +0100 67011 67012 hw/xwin: Allow XScreenSaverSuspend() to effect Windows screen saver or monitor power off 67013 67014 Reset the idle timer in the block handler if screenSaverSuspended is set. 67015 67016 This isn't quite a complete solution. We should also set the block timeout to 67017 something less than the idle timer timeout to ensure we will reset the idle 67018 timer before it times out if we are blocking. 67019 67020 The idle timer timeout is presumably the first one to expire of the screen saver 67021 activation or monitor low power or power down timeout, depending on 67022 configuration. 67023 67024 Unfortunately this is probably not straightforward to do. Whilst 67025 SystemParametersInfo(SPI_GETSCREENSAVETIMEOUT) is portable, apparently 67026 SPI_GETLOWPOWERTIMEOUT and SPI_GETPOWEROFFTIMEOUT are not supported by Windows 67027 versions 6.0 or later, and the interface for discovering equivalent value is 67028 complex. 67029 67030 This doesn't matter in the case where a media player or similar application is 67031 the one making the XScreenSaverSuspend() requests, as it will be continuously 67032 drawing, causing the X server to become unblocked often. 67033 67034 In the case where slide show presentation application or similar is the one 67035 making the XScreenSaverSuspend() request, this might be a problem. 67036 67037 Since "Powerpoint is Evil" [1], I think we'll leave it like this till someone 67038 complains :-) 67039 67040 [1] http://www.edwardtufte.com/tufte/books_pp 67041 67042 v2: conditionalize on SCREENSAVER 67043 67044 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67045 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 67046 67047commit 190f165317c4a58adf38e3e32acfe7cf04850e91 67048Author: Jon TURNEY <jon.turney@dronecode.org.uk> 67049Date: Mon Nov 5 13:02:46 2012 +0000 67050 67051 hw/xwin: Avoid logging an extra blank line if BUILDERSTRING is empty 67052 67053 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67054 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 67055 67056commit ef9dbea24f5ce7e1e400463ab2a333076d98a28e 67057Author: Jon TURNEY <jon.turney@dronecode.org.uk> 67058Date: Wed Aug 31 20:57:05 2011 +0100 67059 67060 hw/xwin: Report which drawing engines are being enabled when checking for support 67061 67062 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67063 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 67064 67065commit 5824166256c8e96c91e225f4ccb2b094b8fb045b 67066Author: Jon TURNEY <jon.turney@dronecode.org.uk> 67067Date: Fri Mar 12 14:38:51 2010 +0000 67068 67069 hw/xwin: Show any fatal error message 67070 67071 Report the fatal error message in the dialog we pop up, rather than just 67072 referring the user to the logfile. 67073 67074 v2: Do this a better way since the "Pass the FatalError message to 67075 OsVendorFatalError" patch has landed, and OsVendorFatalError() now gets passed 67076 the fatal error message 67077 67078 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67079 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 67080 67081commit cf05db8ddec948d907915e1c9569d743e8771def 67082Author: Sjoerd Simons <sjoerd.simons@collabora.co.uk> 67083Date: Tue Oct 30 17:02:58 2012 +0100 67084 67085 Ephyr: Find the right host screen when embedded 67086 67087 When Xephyr is embedded into another application (e.g. in xoo), the 67088 input events may come from the parent Window instead of our own. So make 67089 sure we find the host screen in that case as well instead of crashing. 67090 67091 Signed-off-by: Sjoerd Simons <sjoerd.simons@collabora.co.uk> 67092 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 67093 67094commit ac1a60e7b6f06fd075cc5bf55d6bc67206a01d29 67095Author: Raphael Kubo da Costa <raphael.kubo.da.costa@intel.com> 67096Date: Fri Nov 16 19:51:58 2012 +0200 67097 67098 vfb: Initialize the GLX extension again. 67099 67100 This should fix a regression in the 1.13.0 release: commit 67101 5f5bbbe543f65c48ecbb5cce80116a86ca3fbe86 removed a code path used by Xvfb 67102 and made it use the default one when initializing extensions. However, this 67103 meant the GLX extension was not initialized anymore since it is not part of 67104 the `staticExtensions' array. 67105 67106 Since it is not possible to just add it to that array after commit 67107 aad428b8e21c77397c623b78706eb64b1fea77c9, adopt an approach similar to 67108 xwin's and xquartz's and initialize the extension from vfb's `InitOutput'. 67109 67110 Signed-off-by: Raphael Kubo da Costa <raphael.kubo.da.costa@intel.com> 67111 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 67112 67113commit 353b14ea77324b3f7b3ea29d6f03618fd9916d5f 67114Author: Daniel Stone <daniel@fooishbar.org> 67115Date: Mon Nov 19 15:00:22 2012 +1100 67116 67117 DMX: Add DMX and GLX extensions 67118 67119 Unfortunately this also got lost in the extmod fallout, leaving the DMX 67120 server not exposing the DMX or GLX extensions. 67121 67122 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 67123 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67124 67125commit 3556d43010f0a37bf30aa2e0ceee9a8bbcad311d 67126Author: Daniel Stone <daniel@fooishbar.org> 67127Date: Mon Nov 19 15:29:34 2012 +1100 67128 67129 Constify extensions in LoadExtension users 67130 67131 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 67132 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67133 67134commit db83830eef38a86a471378121213162da180537d 67135Author: Daniel Stone <daniel@fooishbar.org> 67136Date: Mon Nov 19 15:28:07 2012 +1100 67137 67138 Constify argument to LoadExtension 67139 67140 Since we never modify it. 67141 67142 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 67143 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67144 67145commit 3b9f1c701787965246638c1a6fd99fb2b6078114 67146Author: Yuly Novikov <ynovikov@chromium.org> 67147Date: Mon Nov 19 21:04:57 2012 -0500 67148 67149 dix: Save touchpoint last coordinates before transform. #49347 67150 67151 DDXTouchPointInfoRec.valuators used to store axis values after transform. 67152 This resulted in Coordinate Transformation Matrix 67153 being applied multiple times to the last coordinates, 67154 in the case when only pressure changes in the last touch event. 67155 67156 Changed DDXTouchPointInfoRec.valuators to store values before transform. 67157 67158 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=49347 67159 67160 Signed-off-by: Yuly Novikov <ynovikov@chromium.org> 67161 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67162 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67163 67164commit d0fd592fc7be1706084923f43fa057e50122562c 67165Author: Thomas Jaeger <ThJaeger@gmail.com> 67166Date: Sat Oct 13 23:18:50 2012 -0400 67167 67168 Simplify GetTouchEvents 67169 67170 With only one callee left, we are free to assume that 67171 !(flags & TOUCH_CLIENT_ID) 67172 67173 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 67174 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67175 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67176 67177commit fe59774c55e5d423633405e0869c22f4ce382548 67178Author: Thomas Jaeger <ThJaeger@gmail.com> 67179Date: Sat Oct 13 23:08:27 2012 -0400 67180 67181 Don't use GetTouchEvents in EmitTouchEnd 67182 67183 As before GetTouchEvents causes unwanted side effects. Add a new 67184 function GetDixTouchEnd, which generates a touch event from the touch 67185 point. We fill in the event's screen coordinates from the MD's current 67186 sprite position. 67187 67188 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 67189 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67190 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67191 67192commit cc79107a5b60d2926e16ddbee04149e8d5acc969 67193Author: Thomas Jaeger <ThJaeger@gmail.com> 67194Date: Sat Oct 13 22:51:24 2012 -0400 67195 67196 Don't use GetTouchEvents when replaying events 67197 67198 GetTouchEvents has plenty of side effects such as moving the pointer or 67199 updating the master device, which we don't want to happen when 67200 replaying. The only reason for calling it was to generate a DCCE event, 67201 but GetTouchEvents doesn't even do that right (we might need a DCCE 67202 event even when replaying a master event, or clients could interpret 67203 valuator data incorrectly). 67204 67205 This discussion is moot at the moment anyway, since DeliverTouchEvents 67206 doesn't appear to deliver DCCE events. 67207 67208 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 67209 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67210 67211 [Added call to processInputProc instead of direct call to DeliverTouchEvents] 67212 67213 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67214 67215commit 90b177e5cbb9047453af426d2abfe9c0514fe866 67216Author: Thomas Jaeger <ThJaeger@gmail.com> 67217Date: Sat Oct 13 22:43:26 2012 -0400 67218 67219 Update the MD's position when a touch event is received 67220 67221 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 67222 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67223 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67224 67225commit 6a6c3afe71ac82a93d9fd0034dd5bbdcf0eae1ea 67226Merge: 14c9e245b 6f145084d 67227Author: Keith Packard <keithp@keithp.com> 67228Date: Sun Nov 18 21:50:20 2012 -0800 67229 67230 Merge remote-tracking branch 'whot/for-keith' 67231 67232commit 6f145084d5b9ca4a023dfc538a69bbf30edeac4a 67233Author: Adam Jackson <ajax@redhat.com> 67234Date: Fri Nov 16 13:46:32 2012 -0500 67235 67236 linux: Prefer ioctl(KDSKBMUTE, 1) over ioctl(KDSKBMODE, K_OFF) 67237 67238 K_OFF is a slightly broken interface, since if some other process 67239 (cough, systemd) sets the console state to K_UNICODE then it undoes 67240 K_OFF, and now Alt-F2 will switch terminals instead of summoning the 67241 Gnome "run command" dialog. 67242 67243 KDSKBMUTE separates the "don't enqueue events" logic from the keymap, so 67244 doesn't have this problem. Try it first, then continue falling back to 67245 older methods. 67246 67247 Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=859485 67248 Tested-by: Josh Boyer <jwboyer@redhat.com> 67249 Signed-off-by: Adam Jackson <ajax@redhat.com> 67250 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67251 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67252 67253commit 0ff1c77b17062a3a3ff520709ef719316bc1d8bb 67254Author: Peter Hutterer <peter.hutterer@who-t.net> 67255Date: Fri Nov 16 14:21:47 2012 +1000 67256 67257 xfixes: let PointerBarrierClient->device_ids point to allocated memory 67258 67259 "Fixed" up Jasper's original patch [1] to use pointers instead of 67260 zero-length arrays but didn't actually set the pointer to anywhere. 67261 67262 [1] 9a536820f6174befb22431bab375ef1af0a1ec29 67263 67264 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67265 Reviewed-By: Jasper St. Pierre <jstpierre@mecheye.net> 67266 67267commit aa9a9ad08b8a6e7e95de7c2bf45d93dd50f9ca87 67268Author: Daniel Martin <consume.noise@gmail.com> 67269Date: Wed Nov 7 11:31:01 2012 +0100 67270 67271 Xi: Fix modifier swapping in XIPassiveGrabDevice 67272 67273 XIPassiveGrabDevice uses a list of uint32_t as modifier sets. 67274 67275 The ModifierInfo struct represents the current modifier states and is 67276 therefor used in XIQueryPointer and various events. 67277 67278 Signed-off-by: Daniel Martin <consume.noise@gmail.com> 67279 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67280 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67281 67282commit aa9da5eae109154edf740a5c1eee4bf98e8fbd4f 67283Author: Thomas Jaeger <ThJaeger@gmail.com> 67284Date: Sat Oct 13 22:39:27 2012 -0400 67285 67286 remove init_event 67287 67288 The function is identical to init_device_event from inpututils.c with 67289 the first two arguments swapped. 67290 67291 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 67292 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67293 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67294 67295commit 710065da37b68a9ec7a0c9dc995247f39e3d01e9 67296Author: Daniel Stone <daniel@fooishbar.org> 67297Date: Mon Nov 5 15:58:25 2012 +1100 67298 67299 XKB: Remove component listing support 67300 67301 No-one uses this - not xkbcomp, not GNOME, not KDE. The preferred way 67302 to deal with component listing (which gives you RMLVO rather than 67303 KcCGST) is to use the XML files on the client side. 67304 67305 Indeed, a couple of hours after making this commit, it emerged that all 67306 *.dir files built with xkbcomp 1.1.1 (released two years ago) and later 67307 have been catastrophically broken and nearly empty. So I think that's 67308 reasonable proof that no-one uses them. 67309 67310 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 67311 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67312 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67313 67314commit fd214aabf77c44ba0ca3fd438f0c3c8eb22c7211 67315Author: Peter Hutterer <peter.hutterer@who-t.net> 67316Date: Tue Nov 6 10:27:25 2012 +1000 67317 67318 input: drop FP1616 macro 67319 67320 The double_to_f1616() functions do the same thing, and they're tested. 67321 67322 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67323 Reviewed-by: Keith Packard <keithp@keithp.com> 67324 67325commit 14c9e245bae2447f0e57a3ea4daa8d2085fa8b8f 67326Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 67327Date: Sat Nov 17 11:52:10 2012 -0500 67328 67329 XQuartz: Opt-in to NSSupportsAutomaticGraphicsSwitching 67330 67331 http://xquartz.macosforge.org/trac/ticket/668 67332 67333 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 67334 67335commit 011f8458805e443ac9130865d2840a929a00cabf 67336Merge: 54ba26cb1 b4e44b285 67337Author: Keith Packard <keithp@keithp.com> 67338Date: Mon Nov 5 17:16:07 2012 -0800 67339 67340 Merge remote-tracking branch 'whot/for-keith' 67341 67342commit b4e44b285ed0eee1d06514215a4b01d54f40094b 67343Author: Carlos Garnacho <carlosg@gnome.org> 67344Date: Wed Oct 31 19:29:45 2012 +0100 67345 67346 Xi: Set modifier mask on touch events 67347 67348 Button mask should be out-of-band with the emulated 67349 pointer events as touch devices don't truly have 67350 "buttons". Even though, it's handy to have the modifier 67351 mask from the paired keyboard on touch events. 67352 67353 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 67354 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67355 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67356 67357commit 863f32c930d71073ee5f78452b78bd459d024867 67358Author: Carlos Garnacho <carlosg@gnome.org> 67359Date: Wed Oct 31 19:32:57 2012 +0100 67360 67361 Xi: Update the device after delivering the emulated pointer event(#56558) 67362 67363 Ensure emulated pointer events contain the state that applies before the 67364 event was processed, so the device state must be updated after delivering 67365 such emulated events. 67366 67367 Co-authored-by: Peter Hutterer <peter.hutterer@who-t.net> 67368 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 67369 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67370 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67371 67372commit 9a536820f6174befb22431bab375ef1af0a1ec29 67373Author: Jasper St. Pierre <jstpierre@mecheye.net> 67374Date: Sat Nov 3 20:23:46 2012 -0400 67375 67376 cursor: Fix up implementation for per-device barriers 67377 67378 Support multiple mast devices being specified as a parameter to the 67379 barrier. This should implement all parts of the XFixes specification, 67380 minus the existing non-specification detail where specifying no devices 67381 is the same as specifying XIAllDevices. 67382 67383 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 67384 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67385 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67386 67387commit 944213bbc7099f1e0238f39973653d459efdc2c9 67388Author: Jasper St. Pierre <jstpierre@mecheye.net> 67389Date: Sat Nov 3 20:23:45 2012 -0400 67390 67391 cursor: Clean up barrier finding code a bit 67392 67393 The current code manually unrolls the loop, by finding a barrier, 67394 clamping it, and then repeaing the code. Since we want to modify 67395 both passes of the loop to support device-specific barriers, make 67396 it into an actual loop. 67397 67398 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 67399 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67400 67401commit 1536bc2d213713af45541a15b5e49c552d8d0592 67402Author: Jasper St. Pierre <jstpierre@mecheye.net> 67403Date: Sat Nov 3 20:23:44 2012 -0400 67404 67405 cursor: Clean up pointer barrier creation code a tiny bit 67406 67407 This will make it much simpler when we add more error paths to the code 67408 that constructs pointer barrier clients. 67409 67410 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 67411 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67412 67413commit 04c885de715a7c989e48fc8cf2e61db2b401de2d 67414Author: Jasper St. Pierre <jstpierre@mecheye.net> 67415Date: Sat Nov 3 20:23:43 2012 -0400 67416 67417 cursor: CreatePointerBarrier has a variable request length 67418 67419 Support this, and swap the devices in the SProc as well. Don't 67420 actually do anything with the devices just yet -- specifying any 67421 should still emit a BadImplementation. 67422 67423 Based on a patch by Peter Hutterer <peter.hutterer@who-t.net> 67424 67425 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 67426 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67427 67428commit d74b63f95fc622afe83c53580f1b55346d55fa7a 67429Author: Jasper St. Pierre <jstpierre@mecheye.net> 67430Date: Sat Nov 3 20:23:42 2012 -0400 67431 67432 cursor: Fix a minor unused variable warning 67433 67434 Signed-off-by: Jasper St. Pierre <jstpierre@mecheye.net> 67435 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67436 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67437 67438commit aad65415bff12c6860c19beac42e4165e598a40f 67439Author: Peter Hutterer <peter.hutterer@who-t.net> 67440Date: Thu Oct 11 13:02:27 2012 +1000 67441 67442 dix: don't allow disabling XTest devices 67443 67444 Disabling a XTest device followed by an XTest API call crashes the server. 67445 This could be fixed elsewhere but disabled devices must not send events 67446 anyway. The use-case for disabled XTest devices is somewhat limited, so 67447 simply disallow disabling the devices. 67448 67449 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67450 Reviewed-by: Keith Packard <keithp@keithp.com> 67451 67452commit 5b7384a3154a95a805b040e1910e276b52aada96 67453Author: Alexey Ten (Lynn) <alexeyten@gmail.com> 67454Date: Tue Oct 23 12:32:31 2012 +0400 67455 67456 Apply partial matches for option (#25873) 67457 67458 Rules which match star (*) and option, like one below, should be applied 67459 67460 layout[2] option = symbols 67461 * misc:typo = +typo(base) 67462 67463 This is port of patch from #19563 (https://bugs.freedesktop.org/19563) 67464 because here we have own copy of maprules.c 67465 67466 X.Org Bug 25873 <http://bugs.freedesktop.org/show_bug.cgi?id=25873> 67467 67468 Signed-off-by: Alexey Ten (Lynn) <alexeyten@gmail.com> 67469 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67470 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67471 67472commit 54ba26cb1f9c59559cc3c449abeb31b2ce23bdba 67473Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67474Date: Mon Oct 29 22:37:37 2012 -0500 67475 67476 os: Add libnettle as a choice of SHA1 implementation 67477 67478 libnettle is smaller than libgcrypt, currently being released more 67479 frequently, and has replaced the latter in gnutls-3.x (which is used 67480 by TigerVNC, so they can avoid pulling in two crypto libraries 67481 simultaneously). 67482 67483 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67484 Reviewed-by: Julien Cristau <jcristau@debian.org> 67485 67486commit 2ff56033de2b493a11d2bdf411b7057b1b3a22d7 67487Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67488Date: Tue Oct 16 02:11:59 2012 -0500 67489 67490 Xi: fix fprint format warning 67491 67492 exevents.c: In function 'ProcessTouchEvent': 67493 exevents.c:1601:20: warning: too many arguments for format 67494 67495 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67496 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67497 67498commit d631dbe9b6a1faa4cd18ab7a4a0276db02b84b77 67499Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67500Date: Sat Oct 27 21:50:01 2012 -0500 67501 67502 Xext: fix shadow warnings 67503 67504 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67505 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67506 67507commit 20cf7918ed954a932a4cc404b3d0c4ca87a6ae16 67508Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67509Date: Tue Oct 16 21:08:52 2012 -0500 67510 67511 Xext: fix redundant redeclaration warnings 67512 67513 panoramiX.c:595:13: warning: redundant redeclaration of 'CreateConnectionBlock' 67514 ../include/dix.h:167:23: note: previous declaration of 'CreateConnectionBlock' was here 67515 xres.c:193:13: warning: redundant redeclaration of 'ResExtensionInit' 67516 ../include/extinit.h:109:13: note: previous declaration of 'ResExtensionInit' 67517 xtest.c:60:12: warning: redundant redeclaration of 'DeviceValuator' 67518 ../Xi/exglobals.h:61:12: note: previous declaration of 'DeviceValuator' was here 67519 67520 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67521 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67522 67523commit 89447e1b50d2fb0e046102664045d1f9d4efc542 67524Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67525Date: Mon Oct 15 01:33:05 2012 -0500 67526 67527 Xext: fix unused variable warnings in xf86bigfont.c 67528 67529 These were rendered unused by commit 2c7c520cfe0df30f4bc3adba59d9c62582823bf8. 67530 67531 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67532 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67533 67534commit 344eea237fc07dedfd733d14f95ed0ad26bb5f81 67535Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67536Date: Tue Oct 16 02:04:34 2012 -0500 67537 67538 xkb: fix shadow warnings 67539 67540 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67541 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67542 67543commit 7f1d74e8a906210eafc637df81ded62c3adff748 67544Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67545Date: Tue Oct 16 02:06:54 2012 -0500 67546 67547 render: fix shadow warnings 67548 67549 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67550 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67551 67552commit 8e86123998d26d26f0f60dcbb9836c38e5a4c3a5 67553Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67554Date: Tue Oct 16 01:19:12 2012 -0500 67555 67556 randr: export more provider property symbols 67557 67558 These were added as part of commit 66d92afeaeed9f4a19267d95a1f81b9bf27162a5 67559 but never declared or exported. Fixes warnings: 67560 67561 rrproviderproperty.c:255:1: warning: no previous prototype for 'RRPostProviderPendingProperties' 67562 rrproviderproperty.c:327:1: warning: no previous prototype for 'RRConfigureProviderProperty' 67563 67564 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67565 Reviewed-by: Dave Airlie <airlied@redhat.com> 67566 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67567 67568commit f02e27e4fcc34413b2051e5a01edc92172fa8612 67569Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67570Date: Tue Oct 16 02:16:17 2012 -0500 67571 67572 mi: fix shadow warnings 67573 67574 mibitblt.c: In function 'miGetImage': 67575 mibitblt.c:617:20: warning: declaration of 'pt' shadows a previous local 67576 mibitblt.c:609:17: warning: shadowed declaration is here 67577 mispans.c: In function 'miFillUniqueSpanGroup': 67578 mispans.c:456:33: warning: declaration of 'i' shadows a previous local 67579 mispans.c:382:9: warning: shadowed declaration is here 67580 mispans.c:488:17: warning: declaration of 'i' shadows a previous local 67581 mispans.c:382:9: warning: shadowed declaration is here 67582 67583 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67584 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67585 67586commit 1fe30c00679bd36a6355b48b94b87564d528ff28 67587Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67588Date: Tue Oct 16 02:21:04 2012 -0500 67589 67590 fb: fix shadow warnings 67591 67592 fbblt.c: In function 'fbBlt': 67593 fbblt.c:76:16: warning: declaration of 'src' shadows a previous local 67594 fbblt.c:52:13: warning: shadowed declaration is here 67595 fbblt.c:77:16: warning: declaration of 'dst' shadows a previous local 67596 fbblt.c:52:19: warning: shadowed declaration is here 67597 fbbltone.c: In function 'fbBltPlane': 67598 fbbltone.c:742:13: warning: declaration of 'w' shadows a previous local 67599 fbbltone.c:725:9: warning: shadowed declaration is here 67600 67601 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67602 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67603 67604commit e8d45f301845f70b76407577b92363934ca4f19e 67605Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67606Date: Tue Oct 16 20:54:56 2012 -0500 67607 67608 dix: fix shadow warnings 67609 67610 dispatch.c: In function 'ProcCopyArea': 67611 dispatch.c:1608:5: warning: declaration of 'rc' shadows a previous local 67612 dispatch.c:1604:9: warning: shadowed declaration is here 67613 dispatch.c: In function 'ProcCopyPlane': 67614 dispatch.c:1647:5: warning: declaration of 'rc' shadows a previous local 67615 dispatch.c:1643:9: warning: shadowed declaration is here 67616 events.c: In function 'GetClientsForDelivery': 67617 events.c:2030:68: warning: declaration of 'clients' shadows a global declaration 67618 ../include/dix.h:124:28: warning: shadowed declaration is here 67619 events.c: In function 'DeliverEventToWindowMask': 67620 events.c:2113:19: warning: declaration of 'clients' shadows a global declaration 67621 ../include/dix.h:124:28: warning: shadowed declaration is here 67622 events.c: In function 'EventSuppressForWindow': 67623 events.c:4420:12: warning: declaration of 'free' shadows a global declaration 67624 67625 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67626 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67627 67628commit 1aa783754e21a263b0973516850656b13fd18f0d 67629Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67630Date: Tue Oct 16 14:06:30 2012 -0500 67631 67632 dix: fix redundant redeclaration warnings in dixfont 67633 67634 These functions are already declared in <X11/fonts/fontproto.h>. 67635 Redeclaring them just for _X_EXPORT causes tons of warnings throughout 67636 xserver, but they need to be declared somewhere to be picked up by 67637 sdksyms.sh. Doing so in a private header limits the warnings to 67638 sdksyms.c; fixing those as well would require changes to fontsproto. 67639 67640 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67641 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67642 67643commit 27c5966de35d4726dd9795b4828d4236851f6a88 67644Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67645Date: Mon Oct 15 01:25:17 2012 -0500 67646 67647 xfree86: os-support: fix old-style function definition warnings 67648 67649 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67650 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 67651 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67652 67653commit 64b961bb21369aaea694d883f361a36bc23b19b9 67654Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67655Date: Wed Oct 24 23:57:11 2012 -0500 67656 67657 macros: clarify documentation 67658 67659 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67660 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67661 67662commit 04f980ead5e9af066143a921e0a0e98f26f28ee5 67663Merge: a194630f7 a4941ce49 67664Author: Keith Packard <keithp@keithp.com> 67665Date: Mon Nov 5 08:45:21 2012 -0800 67666 67667 Merge remote-tracking branch 'jturney/mingw-w64-w32api' 67668 67669commit a4941ce4961f254ac31e2cef6dd0863866246bab 67670Author: Jon TURNEY <jon.turney@dronecode.org.uk> 67671Date: Tue Aug 21 15:31:16 2012 +0100 67672 67673 hw/xwin: Wrap 'Status' when including ddraw.h 67674 67675 Status is #defined as an alias for a type in xkbsrv.h, which conflicts with it's 67676 use as a parameter name in rpcdce.h 67677 67678 This fixes compilation with MinGW-w64 w32api headers 67679 67680 (The MinGW-w64 w32api headers actually provide a ddraw.h, so this fix tries to 67681 do things in a way which will still work if our private copy of ddraw.h is 67682 eventually removed) 67683 67684 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67685 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 67686 67687commit 57bbf6e2aead6cce76a9f5d43829b58868a26059 67688Author: Jon TURNEY <jon.turney@dronecode.org.uk> 67689Date: Mon Oct 29 14:35:45 2012 +0000 67690 67691 hw/xwin: Remove pointless winFinishCreateWindowsWindowDDNL() 67692 67693 Remove pointless winFinishCreateWindowsWindowDDNL() and the useless DirectDraw 67694 surface pointers in the window privates 67695 67696 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67697 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 67698 67699commit 5bc4c53d23d02d3d7297a8f00a92b6e4d5e0d6fe 67700Author: Jon TURNEY <jon.turney@dronecode.org.uk> 67701Date: Tue Sep 4 17:28:16 2012 +0100 67702 67703 hw/xwin: Link with libdxguid rather than defining the DirectX GUIDs ourselves 67704 67705 Decruftify by linking with libdxguid rather than defining the DirectX GUIDs 67706 ourselves. This fixes compilation with MinGW-w64 w32api headers, which do not 67707 provide GUID_SECT (which only ever did anything for gcc prior to 2.95, anyhow) 67708 67709 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67710 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 67711 67712commit a194630f7f7b287cb4ea4a459df0745f4c0d4c1a 67713Merge: 74a313042 d511a3016 67714Author: Keith Packard <keithp@keithp.com> 67715Date: Thu Nov 1 13:38:54 2012 -0700 67716 67717 Merge remote-tracking branch 'whot/for-keith' 67718 67719commit d511a3016a79c50cb38e7504d4831a9ae128e422 67720Author: Benjamin Tissoires <benjamin.tissoires@gmail.com> 67721Date: Fri Oct 26 16:27:34 2012 +0200 67722 67723 Add missing labels for multitouch valuators 67724 67725 ABS_MT_DISTANCE exists since kernel v2.6.38, 67726 ABS_MT_TOOL_X|Y appeared in v3.6. 67727 67728 Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com> 67729 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67730 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67731 67732commit ced56f322ead10d1bc93fcd1f8e0ec3ae51292a3 67733Author: Carlos Garnacho <carlosg@gnome.org> 67734Date: Thu Oct 25 15:03:50 2012 +0200 67735 67736 Sync TouchListener memory allocation with population in TouchSetupListeners() 67737 67738 The allocated TouchListener array may fall short by 1 if hitting the worst case 67739 situation where there's an active grab, passive grabs on each window in the 67740 sprite trace and event selection for touch in one of the windows. This may lead 67741 to memory corruptions as the array is overflown. 67742 67743 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 67744 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67745 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67746 67747commit 74a3130426d7e1b456e45daa517769cd3e4b58c0 67748Merge: 1ca096d5e f32ad6dd3 67749Author: Keith Packard <keithp@keithp.com> 67750Date: Mon Oct 29 09:09:16 2012 -0700 67751 67752 Merge remote-tracking branch 'jturney/master' 67753 67754commit 1ca096d5e07221025c4c4110528772b7d94f15ee 67755Author: Michel Dänzer <michel.daenzer@amd.com> 67756Date: Mon Oct 29 12:57:54 2012 +0100 67757 67758 EXA: Track source/mask pixmaps more explicitly for Composite fallback regions. 67759 67760 In particular, make sure pExaScr->src/maskPix are cleared when the 67761 corresponding pictures aren't associated with drawables, i.e. solid or gradient 67762 pictures. Without this, we would in some cases associate the source/mask region 67763 with unrelated pixmaps from previous Composite fallbacks, resulting in random 67764 corruption. 67765 67766 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=47266 67767 67768 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 67769 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 67770 Signed-off-by: Keith Packard <keithp@keithp.com> 67771 67772commit f32ad6dd314a587c301b7f1a6e4ee74e5d8cc06c 67773Author: Ryan Pavlik <rpavlik@iastate.edu> 67774Date: Fri Oct 21 12:32:49 2011 -0500 67775 67776 os/utils.c: Provide only stubs for Lock/UnlockServer on WIN32 67777 67778 MinGW and MSVC lack the POSIX functions to compile the lock file code. 67779 67780 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 67781 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67782 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67783 67784commit 74735c5ebff15db8891808068fe8569d8ccde9e7 67785Author: Colin Harrison <colin.harrison@virgin.net> 67786Date: Sat Feb 18 20:37:19 2012 +0000 67787 67788 hw/xwin: fmemopen is available on cygwin but not MinGW 67789 67790 This is a follow-up to commit 0659437f5ec0e3f646373394f5f9c5461e2170f3. 67791 67792 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 67793 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67794 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67795 67796commit 748be9da20a8bf3dd99701a95eac7ee3895a28de 67797Author: Jon TURNEY <jon.turney@dronecode.org.uk> 67798Date: Tue May 1 20:08:41 2012 +0100 67799 67800 doc: Update documentation about Windows platforms support a bit 67801 67802 Update what c-extensions says about Windows platforms support a bit: 67803 - Document that MinGW Win32 gets let off being POSIX-compilant 67804 - Document the minimum Windows version supported 67805 67806 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67807 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 67808 67809commit 760be785eb49fed4652950fb55dc07b0a41d87de 67810Author: Peter Hutterer <peter.hutterer@who-t.net> 67811Date: Wed Oct 24 11:20:54 2012 +1000 67812 67813 xfree86: remove unused variable sigstate 67814 67815 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67816 Reviewed-by: Keith Packard <keithp@keithp.com> 67817 67818commit e7cd5cce740e653000fb1192b600268dcf77dde2 67819Author: Peter Hutterer <peter.hutterer@who-t.net> 67820Date: Thu Oct 18 15:11:31 2012 +1000 67821 67822 dix: fix zaphod screen scrossing (#54654) 67823 67824 POINTER_SCREEN coordinates are screen-relative. For a Zaphod setup, the 67825 coordinates after a screen crossing are already relative to the new screen's 67826 origin. Add that offset to the coordinates before re-setting. 67827 67828 regression introduced by 67829 commit bafbd99080be49a17be97d2cc758fbe623369945 67830 Author: Peter Hutterer <peter.hutterer@who-t.net> 67831 Date: Wed Aug 8 11:34:32 2012 +1000 67832 67833 dix: work around scaling issues during WarpPointer (#53037) 67834 67835 X.Org Bug 54654 <http://bugs.freedesktop.org/show_bug.cgi?id=54654> 67836 67837 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67838 Reviewed-by: Keith Packard <keithp@keithp.com> 67839 67840commit 676447190190d8546165e21be242cf16dd69f5ae 67841Author: Peter Hutterer <peter.hutterer@who-t.net> 67842Date: Wed Oct 17 14:13:29 2012 +1000 67843 67844 Xi: don't deliver TouchEnd to a client waiting for TouchBegin (#55738) 67845 67846 If a client is still waiting for the TouchBegin, don't deliver a TouchEnd 67847 event. 67848 67849 X.Org Bug 55738 <http://bugs.freedesktop.org/show_bug.cgi?id=55738> 67850 67851 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67852 Tested-by: Thomas Jaeger <thjaeger@gmail.com> 67853 Reviewed-by: Keith Packard <keithp@keithp.com> 67854 67855commit 3018f9c1e5109680dcf69b8f2d7807696a473bde 67856Author: Peter Hutterer <peter.hutterer@who-t.net> 67857Date: Fri Oct 12 13:48:27 2012 +1000 67858 67859 Xi: set xChangeDeviceControlReply.status to Success by default 67860 67861 If the status is other than Success, the code will set it to the required 67862 value. 67863 67864 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67865 Reviewed-by: Keith Packard <keithp@keithp.com> 67866 67867commit 2decff6393a44b56d80d53570718f95354fde454 67868Author: Peter Hutterer <peter.hutterer@who-t.net> 67869Date: Thu Oct 11 16:03:33 2012 +1000 67870 67871 xkb: ProcesssPointerEvent must work on the VCP if it gets the VCP 67872 67873 For button release events, the current code picks the VCK. Because that has 67874 a XKB struct, it thinks this is a PointerKeys event and proceeds to send the 67875 release event through the XTest pointer. That has no effect in normal 67876 operation as the button is never down and an attempt is silently discarded 67877 (normal event processing continues with the VCP). 67878 67879 On server shutdown, the XTest device is already removed, leading to a 67880 null-pointer derefernce when the device is checked for whether buttons are 67881 down (XkbFakeDeviceButton → button_is_down(xtest pointer)). 67882 67883 The current state has only worked by accident, the right approach here is to 67884 handle the VCP's event as such and not switch to the keyboard. 67885 67886 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67887 Reviewed-by: Keith Packard <keithp@keithp.com> 67888 67889commit cc1d8fa8a4ef2e34e9837a44945094fd77e40a0b 67890Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67891Date: Sun Oct 14 23:38:20 2012 -0500 67892 67893 Configure fixes for MinGW 67894 67895 As for Cygwin, define FD_SETSIZE for uses of select() to be correct. 67896 Link with libpthread. 67897 67898 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 67899 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 67900 67901commit 53830281b4da096f9c13107d73ec9c76ff1d14cc 67902Merge: abc2ef590 9cbcb5bd6 67903Author: Keith Packard <keithp@keithp.com> 67904Date: Fri Oct 26 18:04:34 2012 -0700 67905 67906 Merge remote-tracking branch 'sandmann/for-keithp' 67907 67908commit abc2ef590c7cb33ebe21726f83b7347406975f95 67909Merge: 0e85e5e62 c5396ec05 67910Author: Keith Packard <keithp@keithp.com> 67911Date: Fri Oct 26 17:11:58 2012 -0700 67912 67913 Merge remote-tracking branch 'whot/for-keith' 67914 67915commit 9cbcb5bd6a5360a128d15b77a02d8d3351f74366 67916Author: Søren Sandmann Pedersen <ssp@redhat.com> 67917Date: Wed May 30 05:19:08 2012 -0400 67918 67919 Use new pixman_glyph_cache_t API that will be in pixman 0.28.0 67920 67921 This new API allows glyphs to be cached in a data structure in pixman, 67922 and entire glyph strings to be composited in one go. 67923 67924 Also bump pixman dependency to 0.27.2. 67925 67926 Results from the cairo peformance test suite running against Xvfb with 67927 a screen size of 1680x1050@32bpp: 67928 67929 Speedups 67930 ======== 67931 xlib firefox-talos-gfx 12416.63 -> 3603.93 3.45x speedup 67932 ██▌ 67933 xlib xfce4-terminal-a1 1727.57 -> 1048.85: 1.65x speedup 67934 ▋ 67935 xlib evolution 1370.49 -> 869.34: 1.58x speedup 67936 ▋ 67937 xlib gnome-terminal-vim 1832.83 -> 1251.94: 1.46x speedup 67938 ▌ 67939 xlib poppler 1519.70 -> 1204.05: 1.26x speedup 67940 ▎ 67941 xlib firefox-planet-gnome 6982.55 -> 5598.16: 1.25x speedup 67942 ▎ 67943 xlib ocitysmap 1142.77 -> 1071.53: 1.07x speedup 67944 ▏ 67945 67946 No slowdowns were reported. 67947 67948 Results of x11perf -aa10text: 67949 67950 Before: 67951 67952 8000000 reps @ 0.0007 msec (1450000.0/sec) 67953 8000000 reps @ 0.0007 msec (1460000.0/sec) 67954 8000000 reps @ 0.0007 msec (1460000.0/sec) 67955 8000000 reps @ 0.0007 msec (1470000.0/sec) 67956 8000000 reps @ 0.0007 msec (1480000.0/sec) 67957 40000000 trep @ 0.0007 msec (1460000.0/sec) 67958 67959 After: 67960 67961 32000000 reps @ 0.0002 msec (4910000.0/sec) 67962 32000000 reps @ 0.0002 msec (4830000.0/sec) 67963 32000000 reps @ 0.0002 msec (4890000.0/sec) 67964 32000000 reps @ 0.0002 msec (4830000.0/sec) 67965 32000000 reps @ 0.0002 msec (4900000.0/sec) 67966 160000000 trep @ 0.0002 msec (4870000.0/sec) 67967 67968 Version 2: Destroy the glyph cache at server regen time 67969 67970 Acked-by: Aaron Plattner <aplattner@nvidia.com> 67971 Reviewed-by: Keith Packard <keithp@keithp.com> 67972 Signed-off-by: Soren Sandmann <ssp@redhat.com> 67973 67974commit c5396ec05a5c6cab6608ba677f703c5227b1de13 67975Author: Thierry Reding <thierry.reding@avionic-design.de> 67976Date: Wed Oct 17 12:11:49 2012 +0200 67977 67978 xf86: Fix build against recent Linux kernel 67979 67980 Recent Linux kernels reworked the linux/input.h header file, which is 67981 now part of the "user-space API". The include guard therefore has an 67982 additional additional _UAPI prefix. 67983 67984 Instead of adding another case to the #ifdef, drop any include guard 67985 checks and instead always undefine the BUS_* definitions on Linux. 67986 67987 Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de> 67988 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 67989 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 67990 67991commit 7d89eb4978a6649be22747c3de2886781e311647 67992Author: Colin Walters <walters@verbum.org> 67993Date: Wed Jan 4 17:37:06 2012 -0500 67994 67995 autogen.sh: Honor NOCONFIGURE=1 67996 67997 (Also while we have the patient open, use exec instead of hanging 67998 around uselessly in waitpid) 67999 68000 http://people.gnome.org/~walters/docs/build-api.txt 68001 68002 Reviewed-by: Adam Jackson <ajax@redhat.com> 68003 Signed-off-by: Colin Walters <walters@verbum.org> 68004 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68005 68006commit c0a752d2864872023216005375a6a1973fadeffe 68007Author: Lionel Elie Mamane <lionel@mamane.lu> 68008Date: Tue Oct 9 03:17:49 2012 +0200 68009 68010 dix: fix Ungrab action #55785 68011 68012 UngrabAllDevices(Bool kill_client): 68013 If we are not going to kill the client (kill_clients false), 68014 we need to deactivate grabs of active clients, too. 68015 (If we are going to kill the client, 68016 no need to deactivate the grab, 68017 as this will be done as part of the client kill.) 68018 68019 Fixes: X.Org Bug 55785 <http://bugs.freedesktop.org/show_bug.cgi?id=55785> 68020 68021 Signed-off-by: Lionel Elie Mamane <lionel@mamane.lu> 68022 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 68023 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68024 68025commit 3b67cd26149eb991b5f015061a818af65369e668 68026Author: Chase Douglas <chase.douglas@canonical.com> 68027Date: Thu Jun 7 14:12:06 2012 -0700 68028 68029 End physically active touches when device is disabled 68030 68031 Otherwise: 68032 68033 * We can't end the touches while device is disabled 68034 * New touches after enabling the device may erroneously be mapped to old 68035 logical touches 68036 68037 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 68038 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 68039 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68040 68041commit 0e85e5e62819c9e516d6e19836f9f5d7a5d53e87 68042Author: Colin Harrison <colin.harrison@virgin.net> 68043Date: Thu Oct 11 18:05:47 2012 +0100 68044 68045 hw/xwin: Fix no return value warning in winClipboardProc() 68046 68047 winclipboardthread.c: In function ‘winClipboardProc’: 68048 winclipboardthread.c:415:9: warning: ‘return’ with no value, in function returning non-void 68049 winclipboardthread.c:424:13: warning: ‘return’ with no value, in function returning non-void 68050 68051 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 68052 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68053 68054commit 51ed6a7d46c6cc19eb9aaf838be41eb57fd8230a 68055Author: Colin Harrison <colin.harrison@virgin.net> 68056Date: Thu Oct 11 17:57:37 2012 +0100 68057 68058 hw/xwin: Remove unused variable in winmultiwindowwm.c 68059 68060 winmultiwindowwm.c: In function ‘UpdateName’: 68061 winmultiwindowwm.c:522:14: warning: unused variable ‘pszName’ [-Wunused-variable] 68062 68063 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 68064 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68065 68066commit 5886b1dc4d2ce12a238a2e3600fdb576abc6e419 68067Author: Colin Harrison <colin.harrison@virgin.net> 68068Date: Thu Oct 11 17:56:11 2012 +0100 68069 68070 hw/xwin: Remove unused variables in winwindow.c 68071 68072 winwindow.c: In function ‘winCreateWindowNativeGDI’: 68073 winwindow.c:62:19: warning: unused variable ‘pWinPriv’ [-Wunused-variable] 68074 winwindow.c: In function ‘winDestroyWindowNativeGDI’: 68075 winwindow.c:85:19: warning: unused variable ‘pWinPriv’ [-Wunused-variable] 68076 winwindow.c: In function ‘winPositionWindowNativeGDI’: 68077 winwindow.c:108:19: warning: unused variable ‘pWinPriv’ [-Wunused-variable] 68078 winwindow.c: In function ‘winChangeWindowAttributesNativeGDI’: 68079 winwindow.c:213:19: warning: unused variable ‘pWinPriv’ [-Wunused-variable] 68080 winwindow.c: In function ‘winUnmapWindowNativeGDI’: 68081 winwindow.c:241:19: warning: unused variable ‘pWinPriv’ [-Wunused-variable] 68082 winwindow.c: In function ‘winMapWindowNativeGDI’: 68083 winwindow.c:265:19: warning: unused variable ‘pWinPriv’ [-Wunused-variable] 68084 68085 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 68086 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68087 68088commit 08b33882f34fed0c23d54ff7361efc5833868a4b 68089Author: Colin Harrison <colin.harrison@virgin.net> 68090Date: Thu Oct 11 17:52:11 2012 +0100 68091 68092 hw/xwin: Remove unused variables in winvalargs.c 68093 68094 winvalargs.c: In function ‘winValidateArgs’: 68095 winvalargs.c:62:10: warning: unused variable ‘fImplicitScreenFound’ [-Wunused-variable] 68096 winvalargs.c:60:9: warning: unused variable ‘iMaxConsecutiveScreen’ [-Wunused-variable] 68097 68098 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 68099 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68100 68101commit 7e665f1a21468118dababdbf6d8573f1ce1ebd47 68102Author: Colin Harrison <colin.harrison@virgin.net> 68103Date: Thu Oct 11 17:50:42 2012 +0100 68104 68105 hw/xwin: Remove unused variables in winwindowswm.c 68106 68107 winwindowswm.c: In function ‘ProcWindowsWMQueryVersion’: 68108 winwindowswm.c:85:9: warning: unused variable ‘n’ [-Wunused-variable] 68109 winwindowswm.c: In function ‘SProcWindowsWMQueryVersion’: 68110 winwindowswm.c:565:9: warning: unused variable ‘n’ [-Wunused-variable] 68111 68112 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 68113 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68114 68115commit e893d570aa3e9ae8df6a30ac4adaf0949a2fc733 68116Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68117Date: Thu Oct 11 19:36:13 2012 +0100 68118 68119 Use X_ATTRIBUTE_PRINTF where suggested for os.h 68120 68121 InitOutput.c: In function ‘OsVendorInit’: 68122 InitOutput.c:630:29: warning: assignment left-hand side might be a candidate for a format attribute [-Wmissing-format-attribute] 68123 winprocarg.c: In function ‘ddxProcessArgument’: 68124 winprocarg.c:231:29: warning: assignment left-hand side might be a candidate for a format attribute [-Wmissing-format-attribute] 68125 68126 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68127 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68128 68129commit e59cc83bcdca6b64c1bfb2de395d741757f615f6 68130Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68131Date: Thu Oct 11 19:35:19 2012 +0100 68132 68133 hw/xwin: Use X_ATTRIBUTE_PRINTF where suggested for winmsg.c 68134 68135 winmsg.c: In function ‘winVMsg’: 68136 winmsg.c:47:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68137 winmsg.c: In function ‘winDrvMsg’: 68138 winmsg.c:56:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68139 winmsg.c: In function ‘winMsg’: 68140 winmsg.c:66:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68141 winmsg.c: In function ‘winDrvMsgVerb’: 68142 winmsg.c:77:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68143 winmsg.c: In function ‘winMsgVerb’: 68144 winmsg.c:87:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68145 winmsg.c: In function ‘winErrorFVerb’: 68146 winmsg.c:97:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68147 winmsg.c: In function ‘winDebug’: 68148 winmsg.c:107:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68149 winmsg.c: In function ‘winTrace’: 68150 winmsg.c:117:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68151 68152 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68153 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68154 68155commit 3d6e6a394ea2d5dcb77aa706f751423315d8a7ed 68156Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68157Date: Thu Oct 11 19:32:24 2012 +0100 68158 68159 hw/xwin: Use X_ATTRIBUTE_PRINTF where suggested for winerror.c 68160 68161 Also, fix the typoed name OsVenderVErrorF (sic), so it actually gets prototyped 68162 and remove redundant declarations of it's prototype. 68163 68164 winerror.c: In function ‘OsVendorVErrorF’: 68165 winerror.c:56:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68166 winerror.c: In function ‘winMessageBoxF’: 68167 winerror.c:104:5: warning: function might be possible candidate for ‘gnu_printf’ format attribute [-Wmissing-format-attribute] 68168 68169 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68170 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68171 68172commit 13fb6b36b8742a79b9768944eee6d1ad66d7e4d9 68173Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68174Date: Thu Oct 11 16:20:35 2012 +0100 68175 68176 hw/xwin: Use char strings in winClipboardUNIXtoDOS for consistency with the rest of the clipboard code 68177 68178 winclipboardxevents.c: In function ‘winClipboardFlushXEvents’: 68179 winclipboardxevents.c:575:13: warning: passing argument 1 of ‘winClipboardUNIXtoDOS’ from incompatible pointer type 68180 winclipboard.h:102:2: note: expected ‘unsigned char **’ but argument is of type ‘char **’ 68181 68182 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68183 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68184 68185commit fa36a7cd488e4f7fb91beed8a87764b8e0c1a72f 68186Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68187Date: Thu Oct 11 15:51:39 2012 +0100 68188 68189 hw/xwin: Fix warning about discarding const in initializing winKBLayouts 68190 68191 winlayouts.h:46:5: warning: initialization discards qualifiers from pointer target type 68192 winlayouts.h:46:5: warning: initialization discards qualifiers from pointer target type 68193 winlayouts.h:46:5: warning: initialization discards qualifiers from pointer target type 68194 [repeated for each layout] 68195 68196 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68197 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68198 68199commit 0ea65df8a5f97a6fe5225226493aa01380080d2f 68200Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68201Date: Thu Sep 27 17:23:43 2012 +0100 68202 68203 hw/xwin: Fix some of the warnings in generated gl wrapper code 68204 68205 Fix some of the warnings in generated gl wrapper code: 68206 - glWinResolveHelper takes a const char * argument 68207 - ensure formal parameter names don't collide with reserved names or shadow 68208 global declarations (e.g. near, far, index (from string.h), remainder (from 68209 math.h), pointer (from Xdefs.h) by postpending a '_'. 68210 68211 generated_gl_wrappers.c:13:3: warning: passing argument 2 of 'glWinResolveHelper' discards qualifiers from pointer target type 68212 glwrap.c:70:1: note: expected 'char *' but argument is of type 'const char *' 68213 [repeated many times] 68214 generated_gl_wrappers.c: In function 'glVertexAttrib1dARBWrapper': generated_gl_wrappers.c:6884:47: warning: declaration of 'index' shadows a global declaration [-Wshadow] 68215 generated_gl_wrappers.c:6886:3: warning: passing argument 2 of 'glWinResolveHelper' discards qualifiers from pointer target type 68216 [repeated many times] 68217 68218 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68219 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68220 68221commit f7f0739311927ad177dece7119272481b75b73b4 68222Author: Colin Harrison <colin.harrison@virgin.net> 68223Date: Tue Oct 19 23:42:53 2010 +0100 68224 68225 hw/xwin: Warning fix in ProcWindowsWMFrameSetTitle() 68226 68227 winwindowswm.c: In function ‘ProcWindowsWMFrameSetTitle’: 68228 winwindowswm.c:514: error: pointer targets in passing argument 2 of ‘strncpy’ differ in signedness 68229 68230 If you're going to stick random casts into your code, at least use the goddammed 68231 right ones. :-) 68232 68233 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 68234 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68235 68236commit 97c9ed026a7f7efe5fca02c188bcbdbcd7c594e9 68237Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68238Date: Wed Jan 25 19:01:55 2012 +0000 68239 68240 hw/xwin: Fix shadowed local variable i in HandleCustomWM_COMMAND() 68241 68242 Fix shadowed local variable i in HandleCustomWM_COMMAND() 68243 68244 Also, fds are meant to be representable as an int 68245 68246 winprefs.c: In function ‘HandleCustomWM_COMMAND’: 68247 winprefs.c:346:23: error: declaration of ‘i’ shadows a previous local 68248 winprefs.c:322:7: error: shadowed declaration is here 68249 68250 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68251 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68252 68253commit b55d0b92fc536793a877f6073754801a7c3c2dd5 68254Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68255Date: Wed Jan 25 18:56:22 2012 +0000 68256 68257 hw/xwin: Fix shadowed local variables in winLoadCursor() 68258 68259 Move shadowed local variables x and y to the places they are used 68260 68261 wincursor.c: In function ‘winLoadCursor’: 68262 wincursor.c:212:11: error: declaration of ‘x’ shadows a previous local 68263 wincursor.c:166:7: error: shadowed declaration is here 68264 wincursor.c:212:14: error: declaration of ‘y’ shadows a previous local 68265 wincursor.c:166:10: error: shadowed declaration is here 68266 wincursor.c:228:11: error: declaration of ‘x’ shadows a previous local 68267 wincursor.c:166:7: error: shadowed declaration is here 68268 wincursor.c:228:14: error: declaration of ‘y’ shadows a previous local 68269 wincursor.c:166:10: error: shadowed declaration is here 68270 68271 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68272 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68273 68274commit a8c9c3699e6dfde29eeec058d903740a499dcf3b 68275Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68276Date: Thu Oct 11 19:11:36 2012 +0100 68277 68278 hw/xwin/glx: Fix using Mask as a formal parameter shadows a global typedef of the same name 68279 68280 Fix using Mask as a formal parameter shadows the typedef of the same name from X.h 68281 68282 indirect.c: In function 'GetShift': 68283 indirect.c:1629:14: warning: declaration of 'Mask' shadows a global declaration [-Wshadow] 68284 68285 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68286 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68287 68288commit 47df98c7851fd9e97a54f0b4950340506d4a9244 68289Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68290Date: Wed Jan 25 19:17:16 2012 +0000 68291 68292 hw/xwin: Fix using screenInfo as a formal parameter to InitOutput() shadows a global declaration 68293 68294 Using screenInfo as a formal parameter to InitOutput() shadows a global declaration of screenInfo. 68295 Change the formal parameter name from screenInfo to pScreenInfo, as everywhere else uses. 68296 68297 InitOutput.c: In function ‘InitOutput’: 68298 InitOutput.c:891:25: error: declaration of ‘screenInfo’ shadows a global declaration 68299 ../../include/scrnintstr.h:570:19: error: shadowed declaration is here 68300 68301 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68302 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68303 68304commit 97e9c34687f7dc05ee0bdef097a19e7623fb59a3 68305Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68306Date: Wed Jan 25 19:07:49 2012 +0000 68307 68308 hw/xwin: Fix using menu as a formal parameter shadows a global variable of the same name 68309 68310 Using menu as a formal parameter shadows a global variable of the same name 68311 68312 winprefsyacc.y: In function ‘SetRootMenu’: 68313 winprefsyacc.y:286:20: error: declaration of ‘menu’ shadows a global declaration 68314 winprefsyacc.y:55:19: error: shadowed declaration is here 68315 winprefsyacc.y: In function ‘SetDefaultSysMenu’: 68316 winprefsyacc.y:293:26: error: declaration of ‘menu’ shadows a global declaration 68317 winprefsyacc.y:55:19: error: shadowed declaration is here 68318 68319 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68320 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68321 68322commit efe96a17bd741f222ccb226d306b5c1dc1f25ade 68323Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68324Date: Wed Jan 25 18:54:57 2012 +0000 68325 68326 hw/xwin: Fix using index as a formal parameter shadows index() 68327 68328 Using index as a formal parameter shadows index() from strings.h 68329 68330 winallpriv.c: In function ‘winInitCmapPrivates’: 68331 winallpriv.c:119:45: error: declaration of ‘index’ shadows a global declaration 68332 68333 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68334 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68335 68336commit 451c5d9175cdeb36fd614502ff0317f968490dfc 68337Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68338Date: Wed Jan 25 19:09:12 2012 +0000 68339 68340 hw/xwin: Fix using index as a formal parameter in winscrinit.c shadows index() 68341 68342 Using index as a local variable shadows index() from strings.h 68343 68344 winscrinit.c: In function ‘winFinishScreenInitFB’: 68345 winscrinit.c:274:28: error: declaration of ‘index’ shadows a global declaration 68346 winscrinit.c: In function ‘winFinishScreenInitNativeGDI’: 68347 winscrinit.c:623:35: error: declaration of ‘index’ shadows a global declaration 68348 68349 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68350 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68351 68352commit 21faee4b383a12bef58f300b4ad1a3bd34d6de2d 68353Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68354Date: Wed Jan 25 19:05:42 2012 +0000 68355 68356 hw/xwin: Fix using index as a local variable shadows index() 68357 68358 Using index as a local variable shadows index() from strings.h 68359 68360 winprefs.c: In function ‘LoadImageComma’: 68361 winprefs.c:574:7: error: declaration of ‘index’ shadows a global declaration 68362 68363 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68364 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68365 68366commit 2d9123fd0c89c1a72e7281e56985196bc53507bf 68367Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68368Date: Wed Jan 25 18:59:21 2012 +0000 68369 68370 hw/xwin: Fix using index as a formal parameter in winmonitors.c shadows index() 68371 68372 Fix using index as a local variable shadows global declaration of index() from strings.h 68373 68374 winmonitors.c: In function ‘QueryMonitor’: 68375 winmonitors.c:59:23: error: declaration of ‘index’ shadows a global declaration 68376 68377 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68378 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68379 68380commit 5b0435dbdaa47404629800e4e9c5a6952a7bb260 68381Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68382Date: Wed Jan 25 18:53:02 2012 +0000 68383 68384 hw/xwin: Fix using system as a local variable in winCheckMount() shadows system() 68385 68386 Using system as local variable in winCheckMount() shadows the global declaration of system() from stdlib.h 68387 68388 InitOutput.c: In function ‘winCheckMount’: 68389 InitOutput.c:296:10: error: declaration of ‘system’ shadows a global declaration 68390 68391 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68392 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68393 68394commit 17d84c743d6b52f32c4ca0341aebaff972522945 68395Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68396Date: Wed Jan 25 17:18:12 2012 +0000 68397 68398 hw/xwin: Remove obsolete redundant declarations of winPushPixels() 68399 68400 winPushPixels() and winpushpxl.c was removed in it's entirety in commit 68401 f31bd087e8a7f65cd588bd1d022bb18e72b2a60c "Death to mfb" 68402 68403 wingc.c:37:1: error: redundant redeclaration of ‘winPushPixels’ 68404 win.h:1104:1: note: previous declaration of ‘winPushPixels’ was here 68405 68406 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68407 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68408 68409commit f06b468d96afd323e28645dde44170ab367da4f1 68410Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68411Date: Wed Jan 25 17:09:18 2012 +0000 68412 68413 hw/xwin: Fix redundant declarations in winmultiwindowwm.c 68414 68415 winmultiwindowwm.c:139:14: error: redundant redeclaration of ‘display’ 68416 ../../include/opaque.h:52:14: note: previous declaration of ‘display’ was here 68417 winmultiwindowwm.c:140:13: error: redundant redeclaration of ‘ErrorF’ 68418 ../../include/os.h:558:13: note: previous declaration of ‘ErrorF’ was here 68419 68420 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68421 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68422 68423commit 3e8269177ee01fe107830032b60a8d311d4b4b0e 68424Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68425Date: Wed Jan 25 17:08:17 2012 +0000 68426 68427 hw/xwin: Fix redundant declarations in winclipboardwrappers.c 68428 68429 Fix redundant declarations in winclipboardwrappers.c. 68430 Also remove obsolete winProcQueryTree() prototype. 68431 68432 winclipboardwrappers.c:64:12: error: redundant redeclaration of ‘g_iNumScreens’ 68433 winglobals.h:33:12: note: previous declaration of ‘g_iNumScreens’ was here 68434 winclipboardwrappers.c:67:13: error: redundant redeclaration of ‘g_fXdmcpEnabled’ 68435 winglobals.h:45:13: note: previous declaration of ‘g_fXdmcpEnabled’ was here 68436 winclipboardwrappers.c:75:27: error: redundant redeclaration of ‘winProcEstablishConnectionOrig’ 68437 winglobals.h:66:27: note: previous declaration of ‘winProcEstablishConnectionOrig’ was here 68438 winclipboardwrappers.c:76:27: error: redundant redeclaration of ‘winProcQueryTreeOrig’ 68439 winglobals.h:67:27: note: previous declaration of ‘winProcQueryTreeOrig’ was here 68440 winclipboardwrappers.c:77:27: error: redundant redeclaration of ‘winProcSetSelectionOwnerOrig’ 68441 winglobals.h:68:27: note: previous declaration of ‘winProcSetSelectionOwnerOrig’ was here 68442 68443 winclipboardwrappers.c:56:5: warning: redundant redeclaration of ‘winProcEstablishConnection’ [-Wredundant-decls] 68444 winclipboardwrappers.c:52:5: note: previous declaration of ‘winProcEstablishConnection’ was here 68445 winclipboardwrappers.c:57:5: warning: redundant redeclaration of ‘winProcSetSelectionOwner’ [-Wredundant-decls] 68446 winclipboardwrappers.c:54:5: note: previous declaration of ‘winProcSetSelectionOwner’ was here 68447 68448 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68449 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68450 68451commit 0975fa0b391d7b8d7e8a127ebf7a863e53fcab7d 68452Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68453Date: Wed Jan 25 17:06:27 2012 +0000 68454 68455 hw/xwin: Fix redundant declaration in winclipboardthread.c 68456 68457 winclipboardthread.c:51:22: error: redundant redeclaration of ‘serverGeneration’ 68458 ../../include/misc.h:362:22: note: previous declaration of ‘serverGeneration’ was here 68459 winclipboardthread.c: In function ‘winClipboardProc’: 68460 68461 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68462 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68463 68464commit 4a1da00c495706e0fbcec2b30aef3ad7acf5b703 68465Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68466Date: Wed Jan 25 17:05:25 2012 +0000 68467 68468 hw/xwin: Fix redundant declaration in winclipboardinit.c 68469 68470 In file included from winclipboardinit.c:35:0: 68471 winclipboard.h:81:13: error: redundant redeclaration of ‘ErrorF’ 68472 ../../include/os.h:558:13: note: previous declaration of ‘ErrorF’ was here 68473 68474 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68475 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68476 68477commit 30f31b80f0105676359105c895bc195b9f3e704c 68478Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68479Date: Wed Jan 25 16:48:45 2012 +0000 68480 68481 hw/xwin: Fix various warnings in code generated by lex/yacc 68482 68483 Remove various redundant declarations. 68484 68485 Instruct lex not to generate code which isn't going to be used, thus avoiding 68486 some unused function warnings. 68487 68488 Note that bison <2.5 expects stdlib.h to define _STDLIB_H if it's been included, 68489 and avoids prototyping free/malloc in that case. Some stdlib.h (e.g. newlib) may 68490 not match precisely (e.g. defining _STDLIB_H_), so we define it after including 68491 stdlib.h to be sure, and avoid a redudant declaration warning for those symbols. 68492 68493 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68494 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68495 68496commit 4231b9541c6fd26f43b605ecc9dd76622dd786f7 68497Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68498Date: Wed Jan 25 16:38:49 2012 +0000 68499 68500 hw/xwin: Fix redundant declaration in winprefs.c 68501 68502 winprefs.c:63:14: error: redundant redeclaration of ‘display’ 68503 ../../include/opaque.h:52:14: note: previous declaration of ‘display’ was here 68504 68505 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68506 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68507 68508commit 70087c55fffc3cbc52f550418e6b1230bbb1b3b3 68509Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68510Date: Wed Jan 25 16:36:40 2012 +0000 68511 68512 hw/xwin: Fix redundant declaration warnings in winmonitors.c 68513 68514 Include xwin-config.h so HAVE_STRLCPY and HAVE_STRNDUP are defined, so 68515 duplicate declarations of strlcpy(), strlcat() and strndup() aren't made 68516 by os.h 68517 68518 In file included from ../../include/misc.h:111:0, 68519 from ../../include/screenint.h:51, 68520 from ../../include/scrnintstr.h:51, 68521 from win.h:157, 68522 from winmonitors.c:31: 68523 ../../include/os.h:491:15: error: redundant redeclaration of ‘strlcpy’ 68524 /usr/include/string.h:86:35: note: previous declaration of ‘strlcpy’ was here 68525 ../../include/os.h:492:15: error: redundant redeclaration of ‘strlcat’ 68526 /usr/include/string.h:85:35: note: previous declaration of ‘strlcat’ was here 68527 ../../include/os.h:496:15: error: redundant redeclaration of ‘strndup’ 68528 68529 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68530 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68531 68532commit 9904c8fe75c96cf84d8885aa95f2743c1b25ea37 68533Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68534Date: Wed Jan 25 16:09:59 2012 +0000 68535 68536 hw/xwin: Remove completely unused winregistry.c 68537 68538 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68539 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68540 68541commit a69429a17bf4630f6e26f61630a1c2b287202627 68542Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68543Date: Fri Sep 7 14:58:06 2012 +0100 68544 68545 Fix 'make distcheck' for hw/xwin 68546 68547 Add some missing header files to _SOURCES so they get distributed properly 68548 68549 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68550 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 68551 68552commit 1d9fd7ffb03ff6e44f3a8736a94e4558203ce607 68553Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68554Date: Thu Sep 6 13:51:38 2012 +0100 68555 68556 Fix compilation of Xorg DDX without XF86VIDMODE 68557 68558 Fix compilation of Xorg DDX without XF86VIDMODE since 6e74fdda, by putting 68559 xf86vmode.c back under the XF86VIDMODE automake conditional it was accidentally 68560 taken out of. 68561 68562 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68563 Tested-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 68564 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 68565 68566commit 25741b1aa3a0fb39cb842a43ab52935673cfdefa 68567Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68568Date: Sun Sep 2 15:17:05 2012 +0100 68569 68570 hw/xwin: Only add GLX extension once. 68571 68572 Using of LoadExtension() to add GLX onto to the end of the list of extensions on 68573 every server regeneration leads to hilarious consequences (i.e. crashing) after 68574 the first regeneration :-) 68575 68576 Possibly xf86ExtensionInit() and QuartzExtensionInit() need a similar fix. 68577 68578 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68579 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 68580 Tested-by: Colin Harrison <colin.harrison@virgin.net> 68581 68582commit 7ceb85481280fecbf033f147f0f7834cad3890c6 68583Author: Jon TURNEY <jon.turney@dronecode.org.uk> 68584Date: Sun Sep 2 14:55:25 2012 +0100 68585 68586 Correct description of -displayfd option in man page. 68587 68588 A display number, not a port number, is written to the specified fd. 68589 68590 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 68591 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 68592 68593commit 4b7f00346daed20c96f3e8ea13ae411858a5424b 68594Author: Peter Hutterer <peter.hutterer@who-t.net> 68595Date: Wed Oct 10 13:33:48 2012 +1000 68596 68597 dix: fix crash on shutdown if a disabled device is still grabbed (XI1 grab) 68598 68599 A disabled device doesn't have a sprite (less so a sprite->win) and triggers 68600 a NULL-pointer dereference on shutdown when all active grabs are released as 68601 part of the cleanup. 68602 68603 Fix this by checking for sprite being non-null and setting the focus window 68604 to the NullWindow if it is. The rest of the patch just attempts to make 68605 things more readable. 68606 68607 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68608 Reviewed-by: Keith Packard <keithp@keithp.com> 68609 68610commit 0c7109f3215378bb3ed6c00c71129a63ffb4e787 68611Author: Denys Vlasenko <dvlasenk@redhat.com> 68612Date: Mon Oct 8 14:18:02 2012 +1000 68613 68614 os: fix typo in OsSigHandler() error message 68615 68616 Recieved → Received 68617 68618 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 68619 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68620 68621commit 0a75bd640b3dc26b89d9e342999a7f4b7e98edbf 68622Author: Peter Hutterer <peter.hutterer@who-t.net> 68623Date: Wed Oct 3 13:12:27 2012 +1000 68624 68625 xfree86: add xf86UpdateDesktopDimensions() 68626 68627 This call is required for external drivers (specifically NVIDIA) that do 68628 not share the xfree86 infrastructure to update the desktop dimensions. 68629 Without it, the driver would update the ScreenRecs but not update the total 68630 dimensions the input code relies on for transformation. 68631 68632 This call is a thin wrapper around the already-existing internal call and 68633 should be backported to all stable series servers, with the minor ABI bump. 68634 68635 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68636 CC: Andy Ritger <aritger@nvidia.com> 68637 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 68638 68639commit 09f1e5b15b769e1122f0a8d7cae0820038992312 68640Author: Julien Cristau <jcristau@debian.org> 68641Date: Sun Oct 7 18:40:35 2012 +0200 68642 68643 Revert "kinput: allocate enough space for null character." 68644 68645 This reverts commit 531785dd746d64ef7f473a83ca73bb20e74b6fca. 68646 68647 The above commit breaks Xephyr option parsing. Andrzej writes: 68648 68649 Xephyr -retro -keybd evdev,,device=/dev/input/event2,xkbrules=evdev,xkbmodel=evdev,xkblayout=pl -mouse evdev,,device=/dev/input/event1 :3 68650 68651 results in: 68652 68653 <snip> 68654 Pointer option key (device=) of value (/dev/input/event1) not assigned! 68655 Kbd option key (device=) of value (/dev/input/event2) not assigned! 68656 Kbd option key (xkbrules=) of value (evdev) not assigned! 68657 Kbd option key (xkbmodel=) of value (evdev) not assigned! 68658 Kbd option key (xkblayout=) of value (pl) not assigned! 68659 <snip> 68660 68661 The effect of the patch is that the "key=value" pairs are parsed in such 68662 a way that the key is added an "equals" sign to it and we end up with 68663 keys like "device=" instead of "device". This in turn has effect on 68664 KdParsePointerOptions and KdParseKbdOptions: the key does not match 68665 any choice presented in the "switch" statement, and so "Pointer/Kbd 68666 option key (...) of value (...) not assigned!" happens, making all 68667 "key=value" options inaccessible to the user. Reverting the patch makes 68668 them available again. 68669 68670 Reference: http://bugs.debian.org/689246 68671 Reported-by: Andrzej Pietrasiewicz <andrzejtp2010@gmail.com> 68672 Signed-off-by: Julien Cristau <jcristau@debian.org> 68673 Cc: Dave Airlie <airlied@redhat.com> 68674 Reviewed-by: Søren Sandmann <ssp@redhat.com> 68675 Signed-off-by: Keith Packard <keithp@keithp.com> 68676 68677commit 7f9d78d8ffb1a312f9b3d69c05323f601571e162 68678Author: Keith Packard <keithp@keithp.com> 68679Date: Thu Oct 4 22:51:06 2012 -0700 68680 68681 os: Don't re-declare ConnectionInputPtr and ConnectionOutputPtr 68682 68683 They're declared in osdep.h, so don't redeclare them in io.c as 68684 well. Keeps the compiler happier. 68685 68686 Signed-off-by: Keith Packard <keithp@keithp.com> 68687 Reviewed-by: Dave Airlie <airlied@redhat.com> 68688 68689commit d5bf6f95f31037bd49b11348b500c3c13b7e0c99 68690Author: Keith Packard <keithp@keithp.com> 68691Date: Thu Oct 4 14:42:37 2012 -0700 68692 68693 Fix FlushClient to write extraBuf when provided (regression fix) 68694 68695 In commit: 68696 68697 commit 092c57ab173c8b71056f6feb3b9d04d063a46579 68698 Author: Adam Jackson <ajax@redhat.com> 68699 Date: Fri Jun 17 14:03:01 2011 -0400 68700 68701 os: Hide the Connection{In,Out}put implementation details 68702 68703 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 68704 Signed-off-by: Adam Jackson <ajax@redhat.com> 68705 68706 the check for an empty output buffer was moved from one calling 68707 location into the FlushClient implementation itself. However, this 68708 neglected the possibility that additional data, in the form of 68709 'extraBuf' would be passed to FlushClient from other code paths. If the 68710 output buffer happened to be empty at that time, the extra data would 68711 never be written to the client. 68712 68713 This is fixed by checking the total data to be written, which includes 68714 both pending and extra data, instead of just the pending data. 68715 68716 Signed-off-by: Keith Packard <keithp@keithp.com> 68717 Reviewed-by: Julien Cristau <jcristau@debian.org> 68718 68719commit 8367dd9736d74eca971da345c2bf559ce5bbf649 68720Merge: 36c18bb81 7998e2615 68721Author: Keith Packard <keithp@keithp.com> 68722Date: Thu Oct 4 13:08:35 2012 -0700 68723 68724 Merge remote-tracking branch 'whot/for-keith' 68725 68726commit 36c18bb81be619df712778bcb99dd6f1fa38701a 68727Author: Stephan Schreiber <info@fs-driver.org> 68728Date: Wed Aug 29 20:03:58 2012 +0200 68729 68730 int10: fix pci_device_read_rom usage 68731 68732 I noticed that the build-in int10 driver always reports 68733 "Unable to retrieve all of segment 0x0C0000." 68734 even though the entire BIOS data is retrieved with success. 68735 68736 The associated code is in hw/xfree86/int10/generic.c, in the function 68737 xf86ExtendedInitInt10(): 68738 68739 if (pci_device_read_rom(pInt->dev, vbiosMem) < V_BIOS_SIZE) { 68740 xf86DrvMsg(screen, X_WARNING, 68741 "Unable to retrieve all of segment 0x0C0000.\n"); 68742 } 68743 68744 The function pci_device_read_rom() is from libpciaccess; its return 68745 value is not a size but an error status code: 0 means success. 68746 If pci_device_read_rom() returns 0 for success, the warning is generated. 68747 68748 The proposed patch corrects the evaluation of the return value of 68749 pci_device_read_rom() and of the supplied BIOS size. 68750 68751 Debian bug#686153 68752 68753 Signed-off-by: Julien Cristau <jcristau@debian.org> 68754 Reviewed-by: Dave Airlie <airlied@redhat.com> 68755 Signed-off-by: Keith Packard <keithp@keithp.com> 68756 68757commit 4bf3eac5fe20fb203b917a486f69514c55be595a 68758Author: Adam Jackson <ajax@redhat.com> 68759Date: Thu Sep 27 17:12:19 2012 -0400 68760 68761 configure: Stop using AM_MAINTAINER_MODE 68762 68763 All this does is make it so editing configure.ac or Makefile.am doesn't 68764 rebuild the makefiles. Which is just stupid. 68765 68766 v2: Remove --enable-maintainer-mode from autogen.sh 68767 68768 Signed-off-by: Adam Jackson <ajax@redhat.com> 68769 Reviewed-by: Eric Anholt <eric@anholt.net> 68770 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 68771 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 68772 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 68773 Signed-off-by: Keith Packard <keithp@keithp.com> 68774 68775commit 262679d436941580485a10e30938974958f7d588 68776Author: Keith Packard <keithp@keithp.com> 68777Date: Thu Oct 4 11:42:22 2012 -0700 68778 68779 Set version to 1.14 development branch (1.13.99.0) 68780 68781 Signed-off-by: Keith Packard <keithp@keithp.com> 68782 68783commit f179b13b91c8891254d31345ee04e773839e6b87 68784Merge: 4dd5989d1 3d2b768ef 68785Author: Keith Packard <keithp@keithp.com> 68786Date: Thu Oct 4 11:39:46 2012 -0700 68787 68788 Merge remote-tracking branch 'daniels/master' 68789 68790commit 7998e26159893674be69183a73a89a53f5608d58 68791Author: Jason Gerecke <killertofu@gmail.com> 68792Date: Fri Sep 28 15:03:42 2012 -0700 68793 68794 Fix additional gcc -Wwrite-strings warning in xf86 ddx 68795 68796 Commit 09e4b78f missed a case. 68797 68798 Signed-off-by: Jason Gerecke <killertofu@gmail.com> 68799 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 68800 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68801 68802commit f64254d85e731d0b4369d871a9a735b03f283ba6 68803Author: Peter Hutterer <peter.hutterer@who-t.net> 68804Date: Thu Aug 23 15:00:24 2012 +1000 68805 68806 mi: drop two useless conditions in miPointerSetPosition 68807 68808 pDev cannot be NULL here since fill_pointer_events is the only caller. 68809 And if the screen is NULL, then the device tries to send events before it is 68810 fully initialised. That certainly shouldn't happen and would be a bug 68811 elsewhere. 68812 68813 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68814 Reviewed-by: Keith Packard <keithp@keithp.com> 68815 68816commit 314776eb369ca2e438907795ae030dd743c281fc 68817Author: Daniel Stone <daniel@fooishbar.org> 68818Date: Fri Sep 7 18:30:23 2012 +0100 68819 68820 Touch: Fix duplicate TouchBegin selection with virtual devices 68821 68822 Given the following scenario: 68823 1) client A selects for TouchBegin on window W for device D 68824 2) client B selects for TouchBegin on window W for XIAllDevices 68825 3) client C selects for TouchBegin on window W with device E 68826 68827 Step 3 will fail with BadImplementation, because attempting to look up 68828 XIAllDevices or XIAllMasterDevices with dixLookupDevices doesn't work. 68829 This should succeed (or, if it was selecting for device D, fail with 68830 BadAccess as it would be a duplicate selection). 68831 68832 Fix this by performing the appropriate lookup for virtual devices. 68833 68834 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 68835 Cc: Peter Hutterer <peter.hutterer@who-t.net> 68836 Cc: Chase Douglas <chase.douglas@ubuntu.com> 68837 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68838 68839commit 3e6358ee6c33979329b78fe2097a1fdf76fb69cd 68840Author: Daniel Drake <dsd@laptop.org> 68841Date: Fri Sep 7 21:48:35 2012 -0400 68842 68843 Xi: Don't check for TOUCH_END, it's never set 68844 68845 This flag is never set, so checking for it here means that we'll 68846 never release the simulated mouse button press after the user touches 68847 (and releases) the touchscreen for the first time. 68848 68849 Fixes a problem where the XO laptop touchpad became totally 68850 unusable after touching the screen for the first time (since X then 68851 behaved as if the mouse button was held down all the time). 68852 68853 Signed-off-by: Daniel Drake <dsd@laptop.org> 68854 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 68855 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68856 68857commit 9d6b8365702e4648e793fea21ad22f7174558680 68858Author: Peter Hutterer <peter.hutterer@who-t.net> 68859Date: Fri Sep 28 11:49:29 2012 +1000 68860 68861 dix: fix crash on XI 1.x grabs on disabled devices. (#54934) 68862 68863 If the device is disabled, the sprite window is NULL and dereferencing 68864 crashes the server. 68865 68866 This is only triggered for XI 1.x grabs (ProcXGrabDevice) as XI2 grabs would 68867 trigger another code path, creating a sprite for the disabled device as if 68868 detaching it (which is wrong and fixed with this patch too). 68869 68870 Grabbing a disabled device doesn't make sense as it won't send events 68871 anyway. However, the protocol specs do not prohibit it, so we need to keep 68872 it working. 68873 Luckily, oldWin is only used for focus out events, which aren't necessary 68874 given that the device is disabled. 68875 68876 X.Org Bug 54934 <http://bugs.freedesktop.org/show_bug.cgi?id=54934> 68877 68878 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 68879 Reviewed-by: Chase Douglas <chase.douglas@ubuntu.com> 68880 68881commit 3d2b768efae9936c6929c2bc13c7a1acc074ecd3 68882Author: Daniel Stone <daniel@fooishbar.org> 68883Date: Fri Sep 7 18:17:46 2012 +0100 68884 68885 Touch: Fix duplicate TouchBegin selection with virtual devices 68886 68887 Given the following scenario: 68888 1) client A selects for TouchBegin on window W for device D 68889 2) client B selects for TouchBegin on window W for XIAllDevices 68890 3) client C selects for TouchBegin on window W with device E 68891 68892 Step 3 will fail with BadImplementation, because attempting to look up 68893 XIAllDevices or XIAllMasterDevices with dixLookupDevices doesn't work. 68894 This should succeed (or, if it was selecting for device D, fail with 68895 BadAccess as it would be a duplicate selection). 68896 68897 Fix this by performing the appropriate lookup for virtual devices. 68898 68899 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 68900 Reviewed-by: Chase Douglas <chase.douglas@ubuntu.com> 68901 68902commit e3903a9383351b061b1a99dfc653ca50de764ec4 68903Author: Frederic Plourde <frederic.plourde@collabora.com> 68904Date: Wed Sep 12 10:48:32 2012 -0400 68905 68906 Add glXCreateNewContext support in Xephyr #54798 68907 68908 Similar to how we intercept and pass through CreateContext, also pass 68909 through newer CreateNewContext requests. 68910 68911 Fixes Clutter → Xephyr → VirtualBox. 68912 68913 Signed-off-by: Frederic Plourde <frederic.plourde@collabora.com> 68914 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 68915 68916commit 11afebc92ce1a7462ff2886286504425b1c8f0ba 68917Author: Daniel Stone <daniel@fooishbar.org> 68918Date: Fri Sep 7 18:10:35 2012 +0100 68919 68920 Xephyr: GLX: Support MakeContextCurrent and MakeCurrentReadSGI 68921 68922 We need to pass these requests through to the host server in the same 68923 way we do glXMakeCurrent. Generalise the existing MakeCurrent 68924 submission into once that will send MakeCurrent, MakeContextCurrent or 68925 MakeCurrentReadSGI, depending on the request and the host server's 68926 capabilities, and add decoding/hijack support for incoming 68927 MakeContextCurrent and MakeCurrentReadSGI requests. 68928 68929 Fixes Clutter → Xephyr → VirtualBox. 68930 68931 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 68932 Reviewed-by: Adam Jackson <ajax@redhat.com> 68933 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 68934 68935commit 4dd5989d15465f3f3480b521d4e36673972fa24a 68936Merge: 0b02150c2 506e3437c 68937Author: Keith Packard <keithp@keithp.com> 68938Date: Mon Sep 24 11:43:01 2012 -0700 68939 68940 Merge remote-tracking branch 'ajax/server-1.14-abi-churn' 68941 68942commit 0b02150c27e98f996e10d7489f9f67a30e4e3497 68943Author: Yufeng Shen <miletus@chromium.org> 68944Date: Mon Sep 24 14:03:31 2012 -0400 68945 68946 dix: fix scale_to_desktop for edge ABS events 68947 68948 Scale_to_desktop() converts ABS events from device coordinates 68949 to screen coordinates: 68950 [dev_X_min, dev_X_max] -> [screen_X_min, screen_X_max] 68951 [dev_Y_min, dev_Y_max] -> [screen_Y_min, screen_Y_max] 68952 68953 An edge ABS event with X = dev_X_max (e.g., generated from the 68954 edge of a touchscreen) will be converted to have screen X value 68955 = screen_X_max, which, however, will be filterd out when xserver 68956 tries to find proper Window to receive the event, because the 68957 range check for a Window to receive events is 68958 window_X_min <= event_screen_X < window_X_max 68959 Events with event_screen_X = screen_X_max will fail the test get 68960 and rejected by the Window. 68961 68962 To fix this, we change the device to screen coordinates mapping to 68963 [dev_X_min, dev_X_max] -> [screen_X_min, screen_X_max-1] 68964 [dev_Y_min, dev_Y_max] -> [screen_Y_min, screen_Y_max-1] 68965 68966 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 68967 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 68968 Signed-off-by: Yufeng Shen <miletus@chromium.org> 68969 Signed-off-by: Keith Packard <keithp@keithp.com> 68970 68971commit 7722bcbab2507d263c7685b15cccbfdd52fc3a24 68972Merge: 4dc2a7674 f54987de9 68973Author: Keith Packard <keithp@keithp.com> 68974Date: Sun Sep 23 16:04:13 2012 -0700 68975 68976 Merge remote-tracking branch 'jeremyhu/master' 68977 68978commit 4dc2a76740d921c824a4d8193f39dd373475f02a 68979Author: Adam Jackson <ajax@redhat.com> 68980Date: Thu Sep 20 17:57:00 2012 -0400 68981 68982 miext/damage: Only wrap into the GC ops chain if there's a listener (v3) 68983 68984 before after Operation 68985 -------- ----------------- ----------------- 68986 1148346.9 1191807.5 ( 1.04) PutImage 10x10 square 68987 2091666.1 2180983.0 ( 1.04) ShmPutImage 10x10 square 68988 68989 v3: In miDamage{R,Unr}egister, bump the serial number of the affected 68990 drawable (and all children if it's a window) so subsequent drawing 68991 against the damage will trigger another ValidateGC pass and we wrap 68992 in/out correctly. Spotted by Aaron Plattner. 68993 68994 Signed-off-by: Adam Jackson <ajax@redhat.com> 68995 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 68996 Signed-off-by: Keith Packard <keithp@keithp.com> 68997 68998commit ad0156c369a40762df6b70780358ce14e5f2c9b1 68999Author: Adam Jackson <ajax@redhat.com> 69000Date: Thu Sep 20 11:16:26 2012 -0400 69001 69002 dix: Remove MapUnmapEventsEnabled and friends 69003 69004 This hack was added to suppress events generated by Composite's internal 69005 unmap/map cycle on redirection state change. Since that cycle was 69006 removed in 193ecc8b4, these can go. 69007 69008 Signed-off-by: Adam Jackson <ajax@redhat.com> 69009 Reviewed-by: Ville Syrjälä <syrjala@sci.fi> 69010 Signed-off-by: Keith Packard <keithp@keithp.com> 69011 69012commit 387b1ac33ca63d27e42d4f61cf248bafb7c43c17 69013Author: Adam Jackson <ajax@redhat.com> 69014Date: Thu Sep 20 11:16:25 2012 -0400 69015 69016 dix: Factor out DeliverUnmapNotify 69017 69018 Signed-off-by: Adam Jackson <ajax@redhat.com> 69019 Reviewed-by: Keith Packard <keithp@keithp.com> 69020 Signed-off-by: Keith Packard <keithp@keithp.com> 69021 69022commit d20cc0fca4e36ad9bdbda8950c4466b1976e70a1 69023Author: Adam Jackson <ajax@redhat.com> 69024Date: Thu Sep 20 11:16:24 2012 -0400 69025 69026 dix: Factor out DeliverMapNotify 69027 69028 Signed-off-by: Adam Jackson <ajax@redhat.com> 69029 Reviewed-by: Keith Packard <keithp@keithp.com> 69030 Signed-off-by: Keith Packard <keithp@keithp.com> 69031 69032commit 63843cb70030812bb3a311b2e8dfb0d1b0a75176 69033Author: Adam Jackson <ajax@redhat.com> 69034Date: Thu Sep 20 11:16:23 2012 -0400 69035 69036 dix: Factor out MaybeDeliverMapRequest 69037 69038 Signed-off-by: Adam Jackson <ajax@redhat.com> 69039 Reviewed-by: Keith Packard <keithp@keithp.com> 69040 Signed-off-by: Keith Packard <keithp@keithp.com> 69041 69042commit dab7a1ec7fd97e07405e0b01dc1f1ec6e5fc2e7c 69043Author: Adam Jackson <ajax@redhat.com> 69044Date: Thu Sep 20 11:16:22 2012 -0400 69045 69046 dix: Fix some indentation 69047 69048 Signed-off-by: Adam Jackson <ajax@redhat.com> 69049 Reviewed-by: Keith Packard <keithp@keithp.com> 69050 Signed-off-by: Keith Packard <keithp@keithp.com> 69051 69052commit 3d1051aecbb1955084804133cacd12c7f696833a 69053Author: Peter Hutterer <peter.hutterer@who-t.net> 69054Date: Thu Sep 20 05:56:39 2012 +1000 69055 69056 dix: set the device transformation matrix 69057 69058 The property handler is registered after setting the property, so 69059 dev->transform remains as all-zeros. That causes pixman_f_transform_invert() 69060 to fail (in transformAbsolute()) and invert remains as garbage. This 69061 may then cause a cursor jump to 0,0. 69062 69063 Since the axes are not yet initialized here and we need to allow for drivers 69064 changing the matrix, we cannot use the property handler for matrix 69065 initialization, essentially duplicating the code. 69066 69067 Triggered by the fix to (#49347) in 749a593e49adccdf1225be28a521412ec85333f4 69068 69069 https://bugzilla.redhat.com/show_bug.cgi?id=852841 69070 69071 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69072 Reviewed-by: Chase Douglas <chase.douglas@ubuntu.com> 69073 Signed-off-by: Keith Packard <keithp@keithp.com> 69074 69075commit 1cb0261ef54b7dd6a7ef84e1c3959e424706228b 69076Author: Daniel Martin <consume.noise@gmail.com> 69077Date: Thu Sep 6 00:38:26 2012 +0200 69078 69079 dix: Delete mibstore.h 69080 69081 Since Nov 2010 (commit c4c4676) the only purpose of mibstore.h was to 69082 define an empty function (miInitializeBackingStore()) for backward 69083 compatibility. Time to say goodbye. 69084 69085 Reviewed-by: Adam Jackson <ajax@redhat.com> 69086 Signed-off-by: Keith Packard <keithp@keithp.com> 69087 69088commit 08a9ed2524679ed939fa0ad466ee3faf905b8ac4 69089Author: Daniel Martin <consume.noise@gmail.com> 69090Date: Thu Sep 6 00:38:25 2012 +0200 69091 69092 dix: Remove refs to mi backing store from docs 69093 69094 Remove any reference to mibstore.h and miInitializeBackingStore() from 69095 the documentation. 69096 69097 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 69098 Reviewed-by: Adam Jackson <ajax@redhat.com> 69099 Signed-off-by: Keith Packard <keithp@keithp.com> 69100 69101commit 76d8739b1f4ed07b874ae56a492d1e0aa0fd9b04 69102Author: Daniel Martin <consume.noise@gmail.com> 69103Date: Thu Sep 6 00:38:24 2012 +0200 69104 69105 dix: Remove #includes of mibstore.h 69106 69107 Remove more backing store leftovers. 69108 69109 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 69110 Reviewed-by: Adam Jackson <ajax@redhat.com> 69111 Signed-off-by: Keith Packard <keithp@keithp.com> 69112 69113commit 056fd15deaa85d3ed936df9a3b2e0cdcfa599da8 69114Author: Daniel Martin <consume.noise@gmail.com> 69115Date: Thu Sep 6 00:38:23 2012 +0200 69116 69117 dix: Delete mibstore.c 69118 69119 The only reference to mibstore.c was removed with commit c4c4676. Remove the 69120 file itself too. 69121 69122 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 69123 Reviewed-by: Adam Jackson <ajax@redhat.com> 69124 Signed-off-by: Keith Packard <keithp@keithp.com> 69125 69126commit f54987de97720200ee94eba1c7a737d4ad8c55c8 69127Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69128Date: Thu Sep 20 21:11:21 2012 -0700 69129 69130 XQuartz: Add some verbose logging to debug xp_lock_window being unbalanced 69131 69132 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69133 69134commit 25d26875bc9bd6fd23ae1b5280f015abf1b033b7 69135Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69136Date: Thu Sep 20 21:49:40 2012 -0700 69137 69138 XQuartz: Avoid a possible deadlock with DRI on OS X 10.7.5 and OS X 10.8.2 69139 69140 <rdar://problem/12338921> 69141 http://bugs.winehq.org/show_bug.cgi?id=31751 69142 69143 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69144 69145commit 506e3437c73e5ae935ff7c056d7808fbb0c7e614 69146Author: Adam Jackson <ajax@redhat.com> 69147Date: Thu Sep 20 14:22:48 2012 -0400 69148 69149 dix: Fix types in WindowOptRec 69150 69151 No reason for these to be 64 bits on LP64. 69152 69153 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 69154 Signed-off-by: Adam Jackson <ajax@redhat.com> 69155 69156commit 092c57ab173c8b71056f6feb3b9d04d063a46579 69157Author: Adam Jackson <ajax@redhat.com> 69158Date: Fri Jun 17 14:03:01 2011 -0400 69159 69160 os: Hide the Connection{In,Out}put implementation details 69161 69162 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 69163 Signed-off-by: Adam Jackson <ajax@redhat.com> 69164 69165commit e2c7d70e5ddb8b17676a13ceebfbb87d14d63243 69166Author: Adam Jackson <ajax@redhat.com> 69167Date: Fri Jun 17 13:43:38 2011 -0400 69168 69169 dix: Extend initial connection handshake for forwarding proxies 69170 69171 Forwarding proxies like sshd will appear to be local, even though they 69172 aren't really. This leads to weird behaviour for extensions that truly 69173 require running under the same OS services as the client, like MIT-SHM 69174 and DRI2. 69175 69176 Add two new legal values for the initial connection's byteOrder field, 69177 'r' and 'R'. These act like 'l' and 'B' respectively, but have the side 69178 effect of forcing the client to be treated as non-local. Forwarding 69179 proxies should attempt to munge the first packet of the connection 69180 accordingly; older servers will reject connections thusly munged, so the 69181 proxy should fall back to passthrough if the munged connection attempt 69182 fails. 69183 69184 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 69185 Signed-off-by: Adam Jackson <ajax@redhat.com> 69186 69187commit 31bf81772e146af79b0c456aae2159eba8b0280f 69188Author: Adam Jackson <ajax@redhat.com> 69189Date: Thu Sep 20 14:00:09 2012 -0400 69190 69191 dix: Repack ClientRec 69192 69193 Pick smaller types where possible, including bitfielding some Bools and 69194 small enums, then shuffle the result to be hole-free. 192 -> 128 bytes 69195 on LP64, 144 -> 96 bytes on ILP32. 69196 69197 Signed-off-by: Adam Jackson <ajax@redhat.com> 69198 69199commit ff8e3ad8074cd2c8bed49b39c40c2b4892118270 69200Author: Adam Jackson <ajax@redhat.com> 69201Date: Thu Sep 20 13:16:59 2012 -0400 69202 69203 dix: Pull client-is-local flag up to the ClientRec 69204 69205 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 69206 Signed-off-by: Adam Jackson <ajax@redhat.com> 69207 69208commit 3f7bc222638d5d38324ecbc8c2c4e39af17d110e 69209Author: Adam Jackson <ajax@redhat.com> 69210Date: Thu Jun 16 17:40:24 2011 -0400 69211 69212 os: Repack ConnectionOutput for LP64 69213 69214 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 69215 Signed-off-by: Adam Jackson <ajax@redhat.com> 69216 69217commit d01921ec18c21f21d377b60626cc2d3418b84a7c 69218Merge: 70e576687 245e7e036 69219Author: Keith Packard <keithp@keithp.com> 69220Date: Thu Sep 20 17:26:48 2012 +0200 69221 69222 Merge remote-tracking branch 'ajax/ioperm' 69223 69224commit 5109c7f6581b3e4e23700bfd03b4daf0ba97630e 69225Author: Adam Jackson <ajax@redhat.com> 69226Date: Thu Sep 20 11:19:39 2012 -0400 69227 69228 xfree86: Bump video ABI to 14 69229 69230 Signed-off-by: Adam Jackson <ajax@redhat.com> 69231 69232commit 245e7e0361b18766583ae391a2ac1231bb1a1f84 69233Author: Adam Jackson <ajax@redhat.com> 69234Date: Tue Jun 26 14:32:31 2012 -0400 69235 69236 xfree86: Change the semantics of driverFunc(GET_REQUIRED_HW_INTERFACES) 69237 69238 This is a really awkward interface, since we're calling it well before 69239 the driver knows what device it's going to drive. Drivers with both KMS 69240 and UMS support therefore don't know whether to say they need I/O port 69241 access or not, and have to assume they do. 69242 69243 With this change we now call it only to query whether port access might 69244 be needed; we don't use that to determine whether to call a driver's 69245 probe function or not, instead we call them unconditionally. If the 69246 driver doesn't check whether port access was enabled, they might crash 69247 ungracefully. To accomodate this, we move xorgHWAccess to be explicitly 69248 intentionally exported (sigh xf86Priv.h) so that drivers can check that 69249 before they attempt port access. 69250 69251 v2: Move initial xf86EnableIO() nearer the logic that determines whether 69252 to call it, suggested by Simon Farnsworth. 69253 69254 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 69255 Reviewed-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk> 69256 Signed-off-by: Adam Jackson <ajax@redhat.com> 69257 69258commit d88fb00d791c2b19cf9dd244276838aba3a6b442 69259Author: Adam Jackson <ajax@redhat.com> 69260Date: Tue Jun 26 13:15:45 2012 -0400 69261 69262 linux: Make failure to iopl non-fatal 69263 69264 We load the driver list, then enable I/O, then call driver probe based 69265 on whether I/O enable succeeded. That's bad, because the loaded 69266 security policy might forbid port access. We happen to treat that as 69267 fatal for some reason, which means even drivers that don't need I/O 69268 access (like kms and fbdev) don't get the chance to run. Facepalm. 69269 69270 How about we just make that non-fatal instead, that sounds like a much 69271 better plan. 69272 69273 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 69274 Reviewed-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk> 69275 Signed-off-by: Adam Jackson <ajax@redhat.com> 69276 69277commit 048674a6aeb61149a1b5f6b0bc3762ddf57f38ee 69278Author: Adam Jackson <ajax@redhat.com> 69279Date: Tue Jun 26 13:12:45 2012 -0400 69280 69281 linux: Refactor xf86{En,Dis}ableIO 69282 69283 Pull platform methods into their own sections for legibility, and 69284 rewrite the ifdefs to be more concise. 69285 69286 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 69287 Reviewed-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk> 69288 Signed-off-by: Adam Jackson <ajax@redhat.com> 69289 69290commit 70e5766874a919039678bb2ed75f2ccea0cb4345 69291Author: Dave Airlie <airlied@redhat.com> 69292Date: Mon Aug 27 15:22:44 2012 +1000 69293 69294 xf86: fix multi-seat video device support. (v2) 69295 69296 If we are not seat 0 the following apply: 69297 69298 don't probe any bus other than platform 69299 don't probe any drivers other than platform 69300 assume the first platform device we match on the bus is the primary GPU. 69301 69302 This just adds checks in the correct places to ensure this, and 69303 with this X can now start on a secondary seat for an output device. 69304 69305 v2: fix Seat0 macros 69306 Reviewed-by: Keith Packard <keithp@keithp.com> 69307 Signed-off-by: Dave Airlie <airlied@redhat.com> 69308 69309commit 37d956e3ac9513b74078882dff489f9b0a7a5a28 69310Author: Keith Packard <keithp@keithp.com> 69311Date: Mon Sep 10 11:14:20 2012 +1000 69312 69313 xf86: fix compat output selection for no output GPUs 69314 69315 This should work properly with dynamic outputs. 69316 69317 Reviewed-by: Dave Airlie <airlied@redhat.com> 69318 Signed-off-by: Dave Airlie <airlied@redhat.com> 69319 69320commit e0b92f5f86f8ce71ab4c80e9902e51e2525f9705 69321Author: Dave Airlie <airlied@gmail.com> 69322Date: Thu Sep 13 18:44:06 2012 +1000 69323 69324 config/udev: ignore change on drm devices 69325 69326 for input devices we handle change like remove/add, but for 69327 drm devices we get change events when we hotplug outputs, 69328 so lets just ignore change at this level, and let the drivers 69329 handle it. We may in the future want to route driver udev 69330 from here instead. 69331 69332 Reported-by: Timo Aaltonen <timo.aaltonen@canonical.com> 69333 Reviewed-by: Keith Packard <keithp@keithp.com> 69334 Signed-off-by: Dave Airlie <airlied@redhat.com> 69335 69336commit 10672a5abe7d4e924844d7e41c86c24d6367a52b 69337Author: Dave Airlie <airlied@redhat.com> 69338Date: Mon Sep 10 14:48:37 2012 +1000 69339 69340 xf86/platform: scan pci after probing devices 69341 69342 This solves a race if we are trying to dynamically power off 69343 secondary GPUs. Its not the greatest fix ever but it probably 69344 as good as we can do for now. 69345 69346 The GPU probing causes the devices to be powered up, then when 69347 we scan the PCI bus we get the correct information from the kernel, 69348 rather than a bunch of 0xff due to the device being powered off. 69349 69350 drop gratuitous '&'. 69351 69352 Reviewed-by: Keith Packard <keithp@keithp.com> 69353 Signed-off-by: Dave Airlie <airlied@redhat.com> 69354 69355commit e47ad8a0aec7662970e7f81e6ee487330bf094c1 69356Author: Dave Airlie <airlied@redhat.com> 69357Date: Mon Sep 17 11:48:14 2012 +1000 69358 69359 modesetting: remove alloca usage again 69360 69361 this slipped back in. 69362 69363 Signed-off-by: Dave Airlie <airlied@redhat.com> 69364 69365commit 22746df15b5f75dc85f5cf5b73932eb8a44cb89b 69366Author: Dave Airlie <airlied@redhat.com> 69367Date: Thu Sep 6 16:33:54 2012 +1000 69368 69369 dri2: invalidate drawable after sharing pixmap 69370 69371 After we share the pixmap, the backing storage may have changed, 69372 and we need to invalidate and buffers pointing at it. 69373 69374 This fixes GL compositors and prime windows lacking contents initially. 69375 69376 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 69377 Signed-off-by: Dave Airlie <airlied@redhat.com> 69378 69379commit f8eb8c1cb4f453bdbe6a81815be8ecefba2084aa 69380Author: Alon Levy <alevy@redhat.com> 69381Date: Tue Aug 28 11:46:47 2012 +1000 69382 69383 modesetting: add virtual connector support 69384 69385 [airlied: also make sure we don't crash in future.] 69386 69387 Signed-off-by: Alon Levy <alevy@redhat.com> 69388 Signed-off-by: Dave Airlie <airlied@redhat.com> 69389 69390commit 5eb697d88f0735e98fe9a97770ddda83202b1597 69391Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 69392Date: Thu Sep 13 08:45:14 2012 +0200 69393 69394 bump version to 0.5.0 69395 69396commit f0bad69edd57facd6cffde8cb0863d1a735e2492 69397Author: Keith Packard <keithp@keithp.com> 69398Date: Wed Sep 5 14:45:08 2012 -0700 69399 69400 Version bumped to 1.13 69401 69402 Signed-off-by: Keith Packard <keithp@keithp.com> 69403 69404commit 856f80c8d7f22b979c72d9c70b70187df6004a03 69405Merge: 0db936a5b b8ab93dfb 69406Author: Keith Packard <keithp@keithp.com> 69407Date: Wed Sep 5 11:02:58 2012 -0700 69408 69409 Merge remote-tracking branch 'jeremyhu/master' 69410 69411commit 0db936a5b7125d9b73633733cb34e5e4b72f01e8 69412Author: Dave Airlie <airlied@redhat.com> 69413Date: Mon Sep 3 15:09:36 2012 +1000 69414 69415 xf86: call enter/leave VT for gpu screens as well 69416 69417 Otherwise we can't do fast user switch properly for multiple GPUs. 69418 69419 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 69420 Signed-off-by: Dave Airlie <airlied@redhat.com> 69421 69422commit 18deac48db0d2387242461025904eb28a0f71fac 69423Author: Dave Airlie <airlied@redhat.com> 69424Date: Mon Aug 27 15:17:21 2012 +1000 69425 69426 config/udev: respect seat for hotplugged video devices. 69427 69428 This respects the seat tag for hotplugged video devices at X start. 69429 69430 Reviewed-by: Keith Packard <keithp@keithp.com> 69431 Signed-off-by: Dave Airlie <airlied@redhat.com> 69432 69433commit ec740cffeba59216f4293fba829097654cde3b11 69434Author: Dave Airlie <airlied@redhat.com> 69435Date: Mon Aug 27 15:15:19 2012 +1000 69436 69437 config/udev: add wrapper around check if server is not seat 0 69438 69439 this is a simple clean-up that is useful to stop further propogation 69440 of this construct. 69441 69442 Reviewed-by: Keith Packard <keithp@keithp.com> 69443 Signed-off-by: Dave Airlie <airlied@redhat.com> 69444 69445commit 49ec57d5094be0dd2b67435ac7bb04ead780f36d 69446Author: Dave Airlie <airlied@redhat.com> 69447Date: Fri Aug 24 13:23:01 2012 +1000 69448 69449 dix: free default colormap before screen deletion 69450 69451 If we don't free this here, it gets freed later in the resource 69452 cleanups, however it then looks up up pmap->pScreen, which we 69453 freed already in this function. So free the default colormap 69454 when we should. 69455 69456 This fixes a bug after a couple of hotplug cycles when you try 69457 to exit the X server and it crashes. 69458 69459 Reviewed-by: Keith Packard <keithp@keithp.com> 69460 Signed-off-by: Dave Airlie <airlied@redhat.com> 69461 69462commit 20f601a0fbc01a4f8ab85e54ee45f78f7817cdb5 69463Author: Dave Airlie <airlied@redhat.com> 69464Date: Fri Aug 24 12:56:18 2012 +1000 69465 69466 xf86/crtc: don't free config->name 69467 69468 This is set by pre_init not screen init, so if we free it here 69469 and then recycle the server, we lose all the providers. 69470 69471 I think we need to wrap FreeScreen here to do this properly, 69472 will investigate for 1.14 most likely, safer to just leak this 69473 on server exit for now. 69474 69475 Reviewed-by: Keith Packard <keithp@keithp.com> 69476 Signed-off-by: Dave Airlie <airlied@redhat.com> 69477 69478commit b8ab93dfbc7f292b5bfe7e9113e1af824ccbd1a8 69479Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69480Date: Tue Aug 28 10:06:51 2012 -0700 69481 69482 list: Use offsetof() and typeof() to determine member offsets within a structure 69483 69484 Some compilers have difficulty with the previous implementation which 69485 relies on undefined behavior according to the C standard. Using 69486 offsetof() from <stddef.h> (which most likely just uses 69487 __builtin_offsetof on modern compilers) allows us to accomplish this 69488 without ambiguity. 69489 69490 This fix also requires support for typeof(). If your compiler does not 69491 support typeof(), then the old implementation will be used. If you see 69492 failures in test/list, please try a more modern compiler. 69493 69494 v2: Added fallback if typeof() is not present. 69495 69496 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69497 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 69498 69499commit 148583d62b84832369e9df39b3e527b99ca96761 69500Author: Peter Hutterer <peter.hutterer@who-t.net> 69501Date: Wed Aug 29 09:11:35 2012 +1000 69502 69503 tests: move GCC diagnostics pragma outside of function 69504 69505 This is a a gcc 4.6+ feature. 69506 69507 signal-logging.c:210: error: #pragma GCC diagnostic not allowed inside 69508 functions 69509 69510 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69511 Reviewed-by: Dave Airlie <airlied@redhat.com> 69512 Signed-off-by: Keith Packard <keithp@keithp.com> 69513 69514commit c75c947b6e9bc725821b28835f3667c4aabef9ee 69515Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69516Date: Tue Aug 28 12:43:55 2012 -0700 69517 69518 test/list: Fix test_xorg_list_del test 69519 69520 We never use child[2], so it's state is undefined. 69521 69522 This issue seems to have existed since the test was first 69523 written: 92788e677be79bd04e5ef140f4ced50ad8b1bf8e 69524 69525 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69526 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 69527 69528commit 486038fe16957b39d39842365c5409c494c9a8df 69529Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69530Date: Mon Aug 27 22:35:51 2012 -0700 69531 69532 XQuartz: Bump Info.plist version to 2.7.4 69533 69534 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69535 69536commit b90b3f6eac34616c46ab693fafbf6b73a04f407b 69537Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69538Date: Mon Aug 27 17:15:19 2012 -0700 69539 69540 test: Make os test more compliant 69541 69542 sighandler_t is not UNIX. 69543 69544 Regression from: 7f09126e068015db54c56bb982b8f91065375700 69545 69546 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69547 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 69548 69549commit a557edca6152ad09c685a00e01534c35a1f00f2e 69550Merge: 6619f5c0e 24ffcfcde 69551Author: Keith Packard <keithp@keithp.com> 69552Date: Mon Aug 27 08:06:09 2012 -0700 69553 69554 Merge remote-tracking branch 'whot/for-keith' 69555 69556commit 24ffcfcded6b4b024958801e8a6cecad36d9a3e3 69557Author: Peter Hutterer <peter.hutterer@who-t.net> 69558Date: Mon Aug 20 10:28:26 2012 +1000 69559 69560 os: fix typo, fsync when WIN32 is _not_ defined 69561 69562 Introduced in 164b38c72fe9c69d13ea4f9c46d4ccc46566d826 69563 69564 Reported-by: Jon TURNEY <jon.turney@dronecode.org.uk> 69565 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69566 Reviewed-by: Keith Packard <keithp@keithp.com> 69567 69568commit d53e6e02a2595ced1882f5fcd34d08ea039b3b85 69569Author: Peter Hutterer <peter.hutterer@who-t.net> 69570Date: Thu Aug 16 13:54:42 2012 +1000 69571 69572 mi: don't check for core events in miPointerSetPosition (#53568) 69573 69574 As of 81cfe44b1ed0de84ad1941fe2ca74bebef3fc58d, miPointerSetPosition now 69575 returns the screen pointer of the device. This broke floating slave devices, 69576 as soon as a motion event was submitted, miPointerSetPosition returned NULL, 69577 crashing the server. 69578 69579 dev->coreEvents is only false if the device is a floating slave, in which 69580 case it has a sprite. 69581 69582 X.Org Bug 53568 <http://bugs.freedesktop.org/show_bug.cgi?id=53568> 69583 69584 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69585 Reviewed-by: Keith Packard <keithp@keithp.com> 69586 69587commit 6619f5c0e1086b57888ff7146e8ed5897b50d440 69588Author: Keith Packard <keithp@keithp.com> 69589Date: Tue Aug 21 13:46:34 2012 -0700 69590 69591 Relase 1.12.99.905 69592 69593 RC with updated ABI version numbers. 69594 69595 Signed-off-by: Keith Packard <keithp@keithp.com> 69596 69597commit bafbd99080be49a17be97d2cc758fbe623369945 69598Author: Peter Hutterer <peter.hutterer@who-t.net> 69599Date: Wed Aug 8 11:34:32 2012 +1000 69600 69601 dix: work around scaling issues during WarpPointer (#53037) 69602 69603 In WarpPointer calls, we get input in screen coordinates. They must be 69604 scaled to device coordinates, and then back to screen coordinates for screen 69605 crossing and root coordinates in events. 69606 69607 The rounding errors introduced (and clipping in core/XI 1.x events) can lead 69608 to the actual position being different to the requested input coordinates. 69609 e.g. 200 scales to 199.9999, truncated to 199 in the event. 69610 69611 Avoid this by simply overwriting the scaled screen coordinates with the 69612 input coordinates for the POINTER_SCREEN case. 69613 69614 X.Org Bug 53037 <http://bugs.freedesktop.org/show_bug.cgi?id=53037> 69615 69616 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69617 Reviewed-by: Keith Packard <keithp@keithp.com> 69618 69619commit 1ebba43052d68d874148e63c9ae38489ddfc5ec1 69620Author: Peter Hutterer <peter.hutterer@who-t.net> 69621Date: Wed Aug 15 14:49:04 2012 +1000 69622 69623 os: don't block signal-unsafe logging, merely warn about it. 69624 69625 Throw an error into the log file, but continue anyway. And after three 69626 warnings, stop complaining. Not all input drivers will be fixed in time (or 69627 ever) and our printf implementation is vastly inferior, so there is still a 69628 use-case for non-sigsafe logging. 69629 69630 This also adds more linebreaks to the message. 69631 69632 CC: Chase Douglas <chase.douglas@canonical.com> 69633 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69634 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 69635 69636commit 4912b4adb666dad96b832ab2d7caaae49808723e 69637Author: Peter Hutterer <peter.hutterer@who-t.net> 69638Date: Mon Aug 13 14:44:44 2012 +1000 69639 69640 os: add support for %d and %i to pnprintf 69641 69642 The mouse driver uses %i in some debug messages 69643 69644 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69645 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 69646 69647commit 7f8c39c8b5ef89153ecd84d16331e96d8feb18ef 69648Author: Peter Hutterer <peter.hutterer@who-t.net> 69649Date: Mon Aug 13 14:24:36 2012 +1000 69650 69651 Add FormatInt64 to convert signed integers in signal-safe manner 69652 69653 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69654 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 69655 69656commit 36c1d92ec0ef0f3927034a12d4cb79dcc22bd185 69657Author: Peter Hutterer <peter.hutterer@who-t.net> 69658Date: Mon Aug 13 12:24:39 2012 +1000 69659 69660 test: add a few tests for signal-safe logging 69661 69662 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69663 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 69664 69665commit b69536b475118a8787c1726355b504207bf83f8f 69666Author: Peter Hutterer <peter.hutterer@who-t.net> 69667Date: Mon Aug 13 11:15:54 2012 +1000 69668 69669 test: assert from signal-safe number conversion 69670 69671 Throw an assert when the conversion fails instead of just returning. Asserts 69672 are more informative. 69673 69674 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69675 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 69676 69677commit 3e091e1075dc29120e54ab2ef110b04ce7383ffd 69678Author: Aaron Plattner <aplattner@nvidia.com> 69679Date: Fri Aug 17 11:26:05 2012 -0700 69680 69681 xfree86: Bump extension ABI to 7.0 69682 69683 Commit 9d457f9c55f12106ba44c1c9db59d14f978f0ae8 added an array of 69684 DevPrivateSetRec structures in the middle of the ScreenRec, which throws off 69685 extension modules trying to call things like pScreen->DestroyPixmap. 69686 69687 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 69688 Signed-off-by: Keith Packard <keithp@keithp.com> 69689 69690commit ac616d8ed5a634c6b32775eb7b071a13c575fd97 69691Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69692Date: Thu Aug 16 19:09:42 2012 -0700 69693 69694 XQuartz: Use asl_log_descriptor for children as well 69695 69696 This change is #if'd out due to a bug in asl_log_descriptor, but 69697 it is left here as reference to be enabled in the future. 69698 69699 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69700 69701commit ad123071e54d890d36d1373f59b8a6eb4937e936 69702Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69703Date: Thu Aug 16 13:02:28 2012 -0700 69704 69705 XQuartz: Use asl_log_descriptor on Mountain Lion 69706 69707 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69708 69709commit 7c3d2e4828deb4e8ec38e8ef88d6f92b2d931033 69710Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69711Date: Thu Aug 16 19:42:54 2012 -0700 69712 69713 XQuartz: console_redirect: Properly zero-out the tail of the array on realloc() 69714 69715 We forgot to multiply by sizeof(), so it wasn't fully zeroed out. 69716 69717 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69718 69719commit a32e01802ff1c938c0afe0cc007a273b9ada8610 69720Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69721Date: Thu Aug 16 15:43:34 2012 -0700 69722 69723 XQuartz: console_redirect: Set the correct location for reading into the buffer 69724 69725 Prior to this change, it was possible that a large message would have some 69726 of its data prepended to subsequent messages due to our not incorrectly 69727 setting the location to write into the buffer. 69728 69729 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 69730 69731commit e3d7d5d18e3281677890064679af459e1af108cf 69732Author: Adam Jackson <ajax@redhat.com> 69733Date: Thu Jul 19 22:15:10 2012 +0000 69734 69735 Implement ->driverFunc 69736 69737 Copied from fbdev, makes it so we can run without iopl. 69738 69739 Signed-off-by: Adam Jackson <ajax@redhat.com> 69740 69741commit ad5fe2d9614959b68bf71e23abf7e5abac9c2734 69742Merge: 16d8da5ca 99b94af1f 69743Author: Keith Packard <keithp@keithp.com> 69744Date: Wed Aug 15 13:29:17 2012 -0700 69745 69746 Merge remote-tracking branch 'jeremyhu/master' 69747 69748commit 16d8da5ca99211bd180f532d78d7c5943d38c8ad 69749Author: Paul Berry <stereotype441@gmail.com> 69750Date: Tue Jul 31 15:15:56 2012 -0700 69751 69752 glx: Skip multisampled configs when matching pre-existing X visuals. 69753 69754 In __glXScreenInit() we generate the set of GLX visuals in two steps: 69755 first we match each pre-existing X visual with a corresponding 69756 FBConfig, then we generate a new X visual to correspond to all the 69757 remaining FBConfigs. 69758 69759 The first step is used for the two default 24-bit visuals (true color 69760 and direct color) and for the 32-bit visual. If windowsystem 69761 multisampling is enabled in Mesa, we need to ensure that none of these 69762 three visuals gets matched to a multisampled config. 69763 69764 Fixes a bug with windowsystem multisampling in gnome-shell. If the X 69765 server happens to match up a multisampled FBConfig to the 32-bit 69766 visual, gnome-shell will try to use it to read pixels from 69767 alpha-blended windows (such as gnome-terminal), resulting in no window 69768 appearing on screen. 69769 69770 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 69771 Reviewed-by: Chad Versace <chad.versace@linux.intel.com> 69772 Signed-off-by: Keith Packard <keithp@keithp.com> 69773 69774commit 99b94af1ffbef7047f88e7d6c594242d172bec02 69775Author: Jeremy Huddleston <jeremyhu@apple.com> 69776Date: Wed Aug 15 10:28:16 2012 -0700 69777 69778 XQuartz: Fix build regression for GlxExtensionInit 69779 69780 quartz.c:153:6: error: use of undeclared identifier 'GlxExtensionInit'; did you mean 'GEExtensionInit'? [Semantic Issue] 69781 {GlxExtensionInit, "GLX", &noGlxExtension}, 69782 69783 Regression-from: aad428b8e21c77397c623b78706eb64b1fea77c9 69784 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 69785 69786commit c0540b4c8db676d18a9a09828f1590b0b16f9ebc 69787Author: Keith Packard <keithp@keithp.com> 69788Date: Tue Aug 7 17:49:48 2012 -0700 69789 69790 Kludge -- Call RandR screen before cleaning up xf86 crtcs 69791 69792 The core RandR screen cleanup now involves cleaning up any GPU screen 69793 associations, and those call down into DDX to clean up the driver. If 69794 the pointers from the xf86 structures back to the core randr 69795 structures are set to NULL at that point, bad things happen. 69796 69797 This patch "knows" that the core RandR close screen is underneath the 69798 xf86 randr close screen function, and so makes sure it gets called 69799 first. 69800 69801 Signed-off-by: Keith Packard <keithp@keithp.com> 69802 Reviewed-by: Dave Airlie <airlied@redhat.com> 69803 69804commit 288b87e42cbb5db34cda848143655e84af8455da 69805Author: Keith Packard <keithp@keithp.com> 69806Date: Tue Aug 7 17:49:47 2012 -0700 69807 69808 Close GPU screens before core screens 69809 69810 This should make cleaning up the GPU screens easier as the core 69811 screens they are associated with will still be around. 69812 69813 Signed-off-by: Keith Packard <keithp@keithp.com> 69814 Reviewed-by: Dave Airlie <airlied@redhat.com> 69815 69816commit 19772670e3041fe1c7f5bbf32fa039a6d8245ccb 69817Author: Keith Packard <keithp@keithp.com> 69818Date: Tue Aug 7 17:49:46 2012 -0700 69819 69820 Only free Render filter names on last screen close 69821 69822 Hotplugging screens causes the render filter names to get freed while 69823 still in use; wait for the last core screen to be closed before 69824 freeing them. That only happens at server reset, when we want them to 69825 be freed. 69826 69827 Signed-off-by: Keith Packard <keithp@keithp.com> 69828 Reviewed-by: Dave Airlie <airlied@redhat.com> 69829 69830commit 863d528a9f76d0e8f122aebf19f8564a4c67a938 69831Author: Simon Schubert <2@0x2c.org> 69832Date: Sun Aug 12 09:40:16 2012 -0700 69833 69834 fb: reorder Bresenham error correction to avoid overshoot. 69835 69836 When fbBresSolid draws a line, it can happen that after the last 69837 pixel, the Bresenham error term overflows, and fbBresSolid paints 69838 another pixel before adjusting the error term. 69839 69840 However, if this happens on the last pixel (len=0), this extra pixel 69841 might overshoot the boundary, and, in rare cases, lead to a segfault. 69842 69843 Fix this issue by adjusting for the Bresenham error term before 69844 drawing the main pixel, not after. 69845 69846 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=24274 69847 Signed-off-by: Simon Schubert <2@0x2c.or> 69848 Tested-by: Mitch Davis <mjd+freedesktop.org@afork.com> 69849 Signed-off-by: Matt Turner <mattst88@gmail.com> 69850 Signed-off-by: Keith Packard <keithp@keithp.com> 69851 69852commit c22c936b35cee60609041ccb7ee52c00b0634d7b 69853Author: Keith Packard <keithp@keithp.com> 69854Date: Tue Aug 7 16:04:25 2012 -0700 69855 69856 Release 1.12.99.904 69857 69858 Closing the non-critical bug window for 1.13 69859 69860 Signed-off-by: Keith Packard <keithp@keithp.com> 69861 69862commit f94ec283efc922106e7ea1d85771950c66ee8f3b 69863Author: Keith Packard <keithp@keithp.com> 69864Date: Tue Aug 7 16:18:21 2012 -0700 69865 69866 Distribute include/glx_extinit.h 69867 69868 This changed should have been in aad428b8e21c77397c623b78706eb64b1fea77c9 69869 69870 Signed-off-by: Keith Packard <keithp@keithp.com> 69871 69872commit 9ca67de8feab59fcf7d6e3fdec8d8a7cfbb21651 69873Author: Julien Cristau <jcristau@debian.org> 69874Date: Sat Jul 14 16:07:51 2012 +0200 69875 69876 Bump glproto requirement to 1.4.16 69877 69878 Signed-off-by: Julien Cristau <jcristau@debian.org> 69879 Reviewed-by: Dave Airlie <airlied@redhat.com> 69880 Signed-off-by: Keith Packard <keithp@keithp.com> 69881 69882commit 02f94b2d446f6700f791f318c551ac835af8445a 69883Merge: 360fa7736 7f09126e0 69884Author: Keith Packard <keithp@keithp.com> 69885Date: Mon Aug 6 16:52:12 2012 -0700 69886 69887 Merge remote-tracking branch 'whot/for-keith' 69888 69889commit 360fa7736b214cae3ccd39feb4dc87b9e7e9734c 69890Merge: 2100e7238 ac09a4a09 69891Author: Keith Packard <keithp@keithp.com> 69892Date: Mon Aug 6 16:42:34 2012 -0700 69893 69894 Merge remote-tracking branch 'airlied/for-keithp' 69895 69896commit 2100e72388a3dc174093d80c9723f021b4614d17 69897Merge: 4a6f42dda c37c65052 69898Author: Keith Packard <keithp@keithp.com> 69899Date: Mon Aug 6 16:40:54 2012 -0700 69900 69901 Merge remote-tracking branch 'alanc/master' 69902 69903commit 7f09126e068015db54c56bb982b8f91065375700 69904Author: Peter Hutterer <peter.hutterer@who-t.net> 69905Date: Fri Aug 3 15:36:34 2012 +1000 69906 69907 os: don't unconditionally unblock SIGIO in OsReleaseSignals() 69908 69909 Calling OsReleaseSignal() inside the signal handler releases SIGIO, causing 69910 the signal handler to be called again from within the handler. 69911 69912 Practical use-case: when synaptics calls TimerSet in the signal handler, 69913 this causes the signals to be released, eventually hanging the server. 69914 69915 Regression introduced in 08962951de. 69916 69917 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69918 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 69919 69920commit cb306a8f174bec9ded95191b91797f59250e6808 69921Author: Peter Hutterer <peter.hutterer@who-t.net> 69922Date: Tue Jul 31 16:09:38 2012 +1000 69923 69924 dix: make sure the mask is set for emulated scroll events (#52508) 69925 69926 If a device has smooth scrolling axes, but submits scroll button events, we 69927 convert those to motion events and update the valuators. For legacy button 69928 events, the valuator mask is likely unset though, causing 69929 add_to_scroll_valuator() to return early, leaving us with an empty mask. 69930 That again skipped the rest of the code and no events were generated. 69931 69932 Fix it by making sure that the scroll valuator in the mask is at least 69933 initialized to 0. 69934 69935 Broke evdev wheel emulation, introduced by 69936 54476b5e4461ff523e935961affabcf0de12c556. 69937 69938 X.Org Bug 52508 <http://bugs.freedesktop.org/show_bug.cgi?id=52508> 69939 69940 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 69941 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 69942 69943commit 4a6f42dda00ba3b5616f8a86f0d4c9a652c7d9d4 69944Author: Adam Jackson <ajax@redhat.com> 69945Date: Mon Jul 23 16:34:28 2012 -0400 69946 69947 sync: Fix logic error from b55bf248581dc66321b24b29f199f6dc8d02db1b 69948 69949 That commit adds two hunks, and I _think_ they're backwards. It adds 69950 code to modify bracket_greater on NegativeTransition triggers, and 69951 bracket_less on PositiveTransition triggers. That breaks symmetry with 69952 the surrounding code; the code as of this commit could probably be 69953 simplified further. 69954 69955 I can't keep the sync trigger rules in my head for more than about five 69956 minutes at a time, so I'm sending this on for more eyes. RHEL 6.3's 69957 xserver is shipping with b55bf248 reverted: 69958 69959 https://bugzilla.redhat.com/show_bug.cgi?id=748704#c33 69960 69961 And there appear to be some upstream reports of the same issue: 69962 69963 https://bugzilla.gnome.org/show_bug.cgi?id=658955 69964 69965 So I'd like to get this sorted out. 69966 69967 Signed-off-by: Adam Jackson <ajax@redhat.com> 69968 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 69969 Signed-off-by: Keith Packard <keithp@keithp.com> 69970 69971commit 454d0e3a1bb14d7f2579ccb5e513cec5686160e7 69972Author: Dave Airlie <airlied@redhat.com> 69973Date: Thu Jul 26 12:56:51 2012 +1000 69974 69975 randr: fix xinerama output for output slaves 69976 69977 This fixes the xinerama geometry when output slaves are enabled. 69978 69979 Tested with xdpyinfo -ext XINERAMA before after slave added. 69980 69981 Signed-off-by: Dave Airlie <airlied@redhat.com> 69982 Signed-off-by: Keith Packard <keithp@keithp.com> 69983 69984commit ac09a4a091a43188241365152e1244fd8fd93ee7 69985Author: Dave Airlie <airlied@redhat.com> 69986Date: Fri Aug 3 16:43:22 2012 +1000 69987 69988 dri2: fix master pixmap free and reset pointer 69989 69990 These are two minor changes, one to reset the pointer to NULL, 69991 after freeing the pixmaps, one to make sure we use the right API for 69992 the master pixmap, though I doubt it'll ever really matter. 69993 69994 Signed-off-by: Dave Airlie <airlied@redhat.com> 69995 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 69996 69997commit 64623ef90db649e89345f71871affe7a9606f6ea 69998Author: Dave Airlie <airlied@redhat.com> 69999Date: Fri Aug 3 16:42:44 2012 +1000 70000 70001 dri2: free slave pixmap on app exit 70002 70003 When the drawable disappears we need to free the prime master/slave combos. 70004 70005 This fixes a leak after a prime app is run. 70006 70007 Signed-off-by: Dave Airlie <airlied@redhat.com> 70008 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 70009 70010commit 1a465fef9bc21142eecca3999f0761a3c0501a09 70011Author: Dave Airlie <airlied@redhat.com> 70012Date: Fri Aug 3 16:41:45 2012 +1000 70013 70014 pixmap: have slave pixmap take a reference on master pixmap 70015 70016 Since the free routines free the master pixmap then the slave, we should 70017 be taking a reference when we bind them together. 70018 70019 Fixes a use-after-free when resizing a primed gears. 70020 70021 Signed-off-by: Dave Airlie <airlied@redhat.com> 70022 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 70023 70024commit c37c65052f674cd58894ad0b9ec22928a62c624e 70025Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70026Date: Mon Jul 16 21:12:06 2012 -0700 70027 70028 Make indentation of dix/tables.c much more consistent and readable 70029 70030 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70031 Acked-by: Daniel Stone <daniel@fooishbar.org> 70032 70033commit 9f7ef7f7f0566f6319d8328ce0a1e6d0fa866720 70034Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70035Date: Mon Jul 16 20:53:00 2012 -0700 70036 70037 Fix up formatting of initializers for arrays of structs 70038 70039 The indenter seems to have gotten confused by initializing arrays of 70040 structs with the struct defined inline - for predefined structs it did 70041 a better job, so match that. 70042 70043 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70044 70045commit c7b7abfaa068042e396d19538215402cfbb4f1e4 70046Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70047Date: Sat Jul 14 11:21:15 2012 -0700 70048 70049 RRModeCreate: plug memory leak of newModes if AddResource fails 70050 70051 Reported by parfait 1.0: 70052 70053 Error: Memory leak (CWE 401) 70054 Memory leak of pointer 'newModes' allocated with realloc(((char*)modes), ((num_modes + 1) * 8)) 70055 at line 93 of randr/rrmode.c in function 'RRModeCreate'. 70056 pointer allocated at line 82 with realloc(((char*)modes), ((num_modes + 1) * 8)). 70057 Error: Memory leak (CWE 401) 70058 Memory leak of pointer 'newModes' allocated with malloc(8) 70059 at line 93 of randr/rrmode.c in function 'RRModeCreate'. 70060 pointer allocated at line 84 with malloc(8). 70061 70062 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70063 Reviewed-by: Keith Packard <keithp@keithp.com> 70064 70065commit 1eb7be863353be2cf3e83738253eb60c5fe49d19 70066Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70067Date: Sat Jul 14 09:29:52 2012 -0700 70068 70069 rrproperty.c: free newly allocated prop in more error paths 70070 70071 Reported by parfait 1.0: 70072 70073 Error: Memory leak (CWE 401) 70074 Memory leak of pointer 'prop' allocated with RRCreateOutputProperty(property) 70075 at line 220 of randr/rrproperty.c in function 'RRChangeOutputProperty'. 70076 'prop' allocated at line 154 with RRCreateOutputProperty(property). 70077 prop leaks when pending != 0 at line 160. 70078 Error: Memory leak (CWE 401) 70079 Memory leak of pointer 'prop' allocated with RRCreateOutputProperty(property) 70080 at line 346 of randr/rrproperty.c in function 'RRConfigureOutputProperty'. 70081 'prop' allocated at line 334 with RRCreateOutputProperty(property). 70082 at line 350 of randr/rrproperty.c in function 'RRConfigureOutputProperty'. 70083 'prop' allocated at line 334 with RRCreateOutputProperty(property). 70084 70085 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70086 Reviewed-by: Keith Packard <keithp@keithp.com> 70087 70088commit 42e655de4d95cb108aec50efec6bbdb709bb13d7 70089Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70090Date: Tue Jul 10 23:29:53 2012 -0700 70091 70092 rrproviderproperty.c: free newly allocated prop in more error paths 70093 70094 Reported by parfait 1.0: 70095 70096 Error: Memory leak (CWE 401) 70097 Memory leak of pointer 'prop' allocated with RRCreateProviderProperty(property) 70098 at line 221 of randr/rrproviderproperty.c in function 'RRChangeProviderProperty'. 70099 'prop' allocated at line 155 with RRCreateProviderProperty(property). 70100 prop leaks when pending != 0 at line 161. 70101 70102 Error: Memory leak (CWE 401) 70103 Memory leak of pointer 'prop' allocated with RRCreateProviderProperty(property) 70104 at line 345 of randr/rrproviderproperty.c in function 'RRConfigureProviderProperty'. 70105 'prop' allocated at line 333 with RRCreateProviderProperty(property). 70106 at line 349 of randr/rrproviderproperty.c in function 'RRConfigureProviderProperty'. 70107 'prop' allocated at line 333 with RRCreateProviderProperty(property). 70108 70109 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70110 Reviewed-by: Keith Packard <keithp@keithp.com> 70111 70112commit 94b514d5e4b376d05e106eb3853da511256e8545 70113Merge: 7d87545ba 732890004 70114Author: Keith Packard <keithp@keithp.com> 70115Date: Mon Aug 6 15:13:17 2012 -0700 70116 70117 Merge remote-tracking branch 'whot/for-keith' 70118 70119commit 7d87545ba7395ade507cca7bdca7052b26ed18d8 70120Merge: ad707a7dc 988d7ace1 70121Author: Keith Packard <keithp@keithp.com> 70122Date: Mon Aug 6 15:11:13 2012 -0700 70123 70124 Merge remote-tracking branch 'jturney/master' 70125 70126commit ad707a7dcc65c30e030c7600b54f734090e56a91 70127Merge: 5a51cb86f 884f51e97 70128Author: Keith Packard <keithp@keithp.com> 70129Date: Mon Aug 6 15:08:01 2012 -0700 70130 70131 Merge remote-tracking branch 'jeremyhu/master' 70132 70133commit 5a51cb86f39a6809305c403beea54c3625e36259 70134Author: Keith Packard <keithp@keithp.com> 70135Date: Tue Jul 10 23:09:47 2012 -0700 70136 70137 xfree86: When xf86CrtcCloseScreen is called, the randr CRTCs are gone 70138 70139 The RandR CRTC structures are freed when their resource IDs are 70140 destroyed during server shut down, which is before the screen is 70141 closed. Calling back into RandR with stale pointers just segfaults the 70142 server. 70143 70144 Signed-off-by: Keith Packard <keithp@keithp.com> 70145 Reviewed-by: Dave Airlie <airlied@redhat.com> 70146 Tested-by: Knut Petersen <knut_petersen@t-online.de> 70147 70148commit 1bf81af4a6be1113bcc3b940ab264d5c9e0f0c5d 70149Author: Rui Matos <tiagomatos@gmail.com> 70150Date: Mon Jul 30 14:32:12 2012 -0400 70151 70152 xf86RandR12: Don't call ConstrainCursorHarder() if panning is enabled 70153 70154 Panning is at odds with CRTC cursor confinement. This disables CRTC cursor 70155 confinement as long as panning is enabled. 70156 70157 Fixes regression introduced in 56c90e29f04727c903bd0f084d23bf44eb1a0a11. 70158 70159 Reviewed-by: Adam Jackson <ajax@redhat.com> 70160 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 70161 Signed-off-by: Keith Packard <keithp@keithp.com> 70162 70163commit ff56f88616aa63797384c2c484b2bd0f194df96a 70164Author: Adam Jackson <ajax@redhat.com> 70165Date: Mon Jul 30 14:32:11 2012 -0400 70166 70167 randr: Fix up yet another corner case in preferred mode selection 70168 70169 Let's say - purely for the sake of argument, mind you - that you had a 70170 server GPU with anemic memory bandwidth, and you walked up to it and 70171 plugged in a monitor that was 1920x1080 because that's what happened to 70172 be on the crash cart. Say the memory bandwidth is such that anything 70173 larger than 1280x1024 gets filtered away. Now you're in trouble, 70174 because the established timings section includes a 720x400 mode because 70175 that's what DOS 80x25 is, and that happens to just about match the 70176 physical aspect ratio. 70177 70178 Instead let's reuse the logic from the existing aspect-match path: pick 70179 the larger mode of either the physical aspect ratio or 4:3. 70180 70181 Reviewed-by: Julien Cristau <jcristau@debian.org> 70182 Signed-off-by: Adam Jackson <ajax@redhat.com> 70183 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 70184 Signed-off-by: Keith Packard <keithp@keithp.com> 70185 70186commit 531785dd746d64ef7f473a83ca73bb20e74b6fca 70187Author: Dave Airlie <airlied@redhat.com> 70188Date: Mon Jul 30 14:32:10 2012 -0400 70189 70190 kinput: allocate enough space for null character. 70191 70192 This code wasn't allocating enough space and was assigning the NULL 70193 one past the end. 70194 70195 Pointed out by coverity. 70196 70197 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 70198 Reviewed-by: Adam Jackson <ajax@redhat.com> 70199 Signed-off-by: Dave Airlie <airlied@redhat.com> 70200 Signed-off-by: Keith Packard <keithp@keithp.com> 70201 70202commit 8843aed82e7d69422e7763a35832a2be8f26723f 70203Author: Vic Lee <llyzs@163.com> 70204Date: Mon Jul 30 14:32:09 2012 -0400 70205 70206 ephyr: Resize screen automatically when parent window is resized 70207 70208 Bugzilla: https://bugs.freedesktop.org/25804 70209 Reviewed-by: Adam Jackson <ajax@redhat.com> 70210 Signed-off-by: Vic Lee <llyzs@163.com> 70211 Signed-off-by: Keith Packard <keithp@keithp.com> 70212 70213commit b46bbafae6d0a8b3f2f7853d5c1475fc223b1ed6 70214Author: Adam Jackson <ajax@redhat.com> 70215Date: Mon Jul 30 14:32:08 2012 -0400 70216 70217 ephyr: Fix up some bizarre formatting 70218 70219 Signed-off-by: Adam Jackson <ajax@redhat.com> 70220 Signed-off-by: Keith Packard <keithp@keithp.com> 70221 70222commit 7328900042b9c1312aed48753fd6054e64613e4c 70223Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70224Date: Mon Jul 30 23:37:06 2012 -0700 70225 70226 XIChangeDeviceProperty: free newly allocated prop when SetProperty fails 70227 70228 Reported by parfait 1.0: 70229 70230 Error: Memory leak (CWE 401) 70231 Memory leak of pointer 'prop' allocated with XICreateDeviceProperty(property) 70232 at line 774 of Xi/xiproperty.c in function 'XIChangeDeviceProperty'. 70233 'prop' allocated at line 700 with XICreateDeviceProperty(property). 70234 prop leaks when handler != NULL at line 768 70235 and handler->SetProperty != NULL at line 769 70236 and checkonly != 0 at line 772 70237 and rc != 0 at line 772. 70238 70239 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70240 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 70241 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 70242 70243commit 4dbbcdf64563cb95f83c04b2442cb7e868384264 70244Author: Daniel d'Andrada <daniel.dandrada@canonical.com> 70245Date: Thu Jul 26 17:31:57 2012 -0300 70246 70247 Do sent TouchEnd to listeners that don't own an accepted touch 70248 70249 When the owner of a touch accepts it, the other listeners must 70250 receive a TouchEnd. 70251 70252 Even though there's code implementing the logic above in 70253 ProcessTouchOwnershipEvent(), DeliverTouchEndEvent() was refusing to send 70254 those TouchEnd events in this situatuation. 70255 70256 Signed-off-by: Daniel d'Andrada <daniel.dandrada@canonical.com> 70257 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 70258 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 70259 70260commit 02d91ccb0955252153206061a44340f051077624 70261Author: Peter Hutterer <peter.hutterer@who-t.net> 70262Date: Thu Jun 21 15:42:17 2012 +1000 70263 70264 test: always add DIX_LIB and OS_LIB on XORG builds 70265 70266 With --disable-xorg, We also disabled a bunch of tests because of their 70267 perceived reliance on a DDX. The cause was libtool missing some object files 70268 that never ended up in libxservertest.la. Only the xfree86 test has a true 70269 dependency on XORG. 70270 70271 DIX_LIB was pointing to dix.O (instead of libdix.la) when 70272 DTRACE_SPECIAL_OBJECTS was defined. libdix.la should be part of XSERVER_LIBS 70273 but dix.O is not a recognised libtool object, so it got skipped for 70274 libxservertest.a. Only in the XORG case would we add DIX_LIB and OS_LIB 70275 manually, thus forcing linkage with the dtrace-generated objects. 70276 70277 Fixing this by packaging up the dtrace-generated files as part of 70278 libdix.la/libos.la doesn't work for Solaris (and possible others), so simply 70279 always force linkage against the DIX_LIB/OS_LIB in the case of dtrace objects. 70280 70281 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 70282 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70283 Tested-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70284 70285commit 988d7ace19a009991a4528e783d1a94c2444c66a 70286Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70287Date: Thu Jul 5 09:38:44 2012 +0100 70288 70289 glx: Do not report the GLX_INTEL_swap_event extension for indirect swrast 70290 70291 Commit 84956ca4 bogusly adds GLX_INTEL_swap_event to the extensions reported 70292 by swrast. 70293 70294 "DRI2 supports this now - and already enables it explicitly - but drisw does 70295 not and should not. Otherwise toolkits like clutter will only ever SwapBuffers 70296 once and wait forever for an event that's not coming." 70297 70298 (A similar bug for direct swrast is already fixed in mesa commit 25620eb1) 70299 70300 (Note that this may be papering over the cracks somewhat, as if we do report 70301 GLX_INTEL_swap_event, some clutter apps fail with GLXBadDrawable calling 70302 GLXChangeDrawableAttributes to change the setting of GLX_BUFFER_SWAP_COMPLETE_INTEL_MASK 70303 in the GLX_EVENT_MASK, apparently after the drawable is destroyed, which suggests 70304 a bug with GLXDrawable lifetimes) 70305 70306 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70307 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 70308 70309commit fd3d45c137bb849aa9030d732ea9277292e01d3d 70310Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70311Date: Thu Jul 5 09:34:24 2012 +0100 70312 70313 glx: Don't note GLX_INTEL_swap_event as being required by GLX 1.4, it isn't. 70314 70315 Don't note GLX_INTEL_swap_event as being required by GLX 1.4, it isn't. 70316 (This data is not currently used in the server) 70317 70318 (A similar change is made in mesa commit d3f7597bc9f6d5) 70319 70320 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70321 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 70322 70323commit a8464dfa28dea78201e3e4398eb3bcb745e10087 70324Author: Colin Harrison <colin.harrison@virgin.net> 70325Date: Sun Jul 22 13:15:02 2012 +0100 70326 70327 os: Fix TMP fall-back in Win32TempDir() 70328 70329 Fix Win32TempDir() in the case where we fell back to checking the TMP 70330 environment variable. It looks like this has been wrong since forever. 70331 70332 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 70333 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70334 70335commit f6e7b82acadfca8239edc0f7e72cd0f3f9cfc2c4 70336Author: Marc Haesen <marha@users.sourceforge.net> 70337Date: Sat Jul 21 21:18:44 2012 +0100 70338 70339 hw/xwin: Fixes to pixelFormat <-> fbConfig conversion in WGL mode 70340 70341 Fix FIXME in fbConfigToPixelFormat() to correctly populate RGBA-mask shift 70342 parameters. Also request colourindex pixelFormats correctly. 70343 70344 Now that they are requested correctly, don't skip colorindex visuals when 70345 converting pixelFormats to fbConfigs. 70346 70347 Populate transparent colour information when converting pixelFormat from 70348 DescribePixelFormats() to a fbConfig. 70349 70350 Signed-off-by: Marc Haesen <marha@users.sourceforge.net> 70351 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70352 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70353 70354commit 45c432871d6a244e9e558a6a4e7c36e90764135e 70355Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70356Date: Sat Jul 21 14:13:37 2012 +0100 70357 70358 hw/xwin: Introduce winProcessXEventsTimeout() to the concept of fractions of a second 70359 70360 Oh this is terrible. 70361 70362 Currently we only compute the select timeout in whole seconds. This means if we 70363 have less than 1 second remaining, we select with a timeout of 0 (i.e. poll) 70364 which causes the task to spin, burning 100% CPU for the remaining timeout (and 70365 possibly preventing the process we are waiting for from running :S) 70366 70367 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70368 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70369 70370commit 23cd4d0174194e10721d2e465fd1a1c52f001520 70371Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70372Date: Sat Jul 21 12:33:05 2012 +0100 70373 70374 hw/xwin: Fix winUpdateWindowPosition() not to assume WS_EX_APPWINDOW style 70375 70376 Also improve it's debug output a bit 70377 70378 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70379 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70380 70381commit 527cf13135cfd279733060e0028bbfbe02be5167 70382Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70383Date: Sat Jul 21 14:09:16 2012 +0100 70384 70385 hw/xwin: Make winOverrideIcon() thread-safe for icon data access 70386 70387 winOverrideIcon() is called from the internal WM client thread. 70388 70389 Accessing server-internal data structures to get icon data or window hints is 70390 not safe, as there is no lock to ensure we do not collide with these data 70391 structures being updated in the server thread. 70392 70393 Rewrite so the internal client thread uses X client calls to obtain this data 70394 safely 70395 70396 We used to also set the icon inside the server when the window was initially 70397 created. For simplicity, we simply send a message to the internal WM to update 70398 the icon when the window is created (rather than writing different icon update 70399 code which can work in the server thread for that one case...) 70400 70401 extwm mode used to do the icon update in the server. I'm not sure that actually 70402 made much sense. Let's assume the external WM client can do it instead... 70403 70404 v2 70405 Make sure that WM_WM_ICON_EVENT does nothing for override-redirect windows 70406 70407 v3 70408 Reinstate check that native window actually has expected properties for an X 70409 window before trying to update it's icon; some auxiliary windows owned by the 70410 XWin process don't, which would cause a crash 70411 70412 v4 70413 Various fixes to pixmap icon conversion: 70414 - remove left-over malloc in winScaleXimageToWindowsIcon causing a memory leak 70415 - don't recalculate DDBitmap stride in winScaleXimageToWindowsIcon, when we already have worked it out 70416 - properly check that XGetWindowProperty(NET_WM_ICON) returned some data 70417 - don't try to retrieve WM_HINTS icon_mask if it isn't set 70418 - restore accidentally dropped calculation of effBpp, stride, maskStride of output DDBitmap 70419 - make sure imageMask is zero-initalized before we use it to mask the DDBitmap 70420 70421 v5 70422 Remove a left-over unused variable 70423 70424 v6 70425 Avoid XDestroyImage(NULL) crash if XGetImage failed for icon_pixmap 70426 70427 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70428 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70429 70430commit 2677d89823b851fea31036f51589985bb86864b7 70431Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70432Date: Sun Jul 22 17:38:55 2012 +0100 70433 70434 hw/xwin: Also update icon when _NET_WM_ICON property changes 70435 70436 _NET_WM_ICON property is also considered to decide on the window icon, so also 70437 send a WM_WM_ICON_EVENT message to the WM if the PropertyNotify event is for 70438 _NET_WM_ICON property 70439 70440 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70441 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70442 70443commit a07541f1ffc28c57a762beefb96c7bf0ac40b0a4 70444Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70445Date: Sun Jul 22 16:15:12 2012 +0100 70446 70447 hw/xwin: Rename WM_WM_HINTS_EVENT to WM_WM_ICON_EVENT 70448 70449 WM_WM_HINTS_EVENT only updates the icon, so rename it to WM_WM_ICON_EVENT 70450 70451 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70452 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70453 70454commit 2a47c91eb382d1ff0fb009a39efa7dc9c6fd5112 70455Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70456Date: Fri Jul 20 13:51:35 2012 +0100 70457 70458 hw/xwin: Refactor Xutf8TextPropertyToString() from GetWindowName() as a separate utility function 70459 70460 Simplify GetWindowName() by moving UTF-8 to wchar conversion out to it's call 70461 site. This allows us to do extra processing on the window name in future. 70462 70463 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70464 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70465 70466commit 884f51e9770c9ccd3b38a14ea393ea2c7a5ef235 70467Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 70468Date: Wed Aug 1 18:46:08 2012 -0700 70469 70470 XQuartz: Fix xp_window_bring_all_to_front linking on OS versions with older libXplugin 70471 70472 Found-by: Tinderbox 70473 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 70474 70475commit b4c2358e0a388730f1beda5d3a12e79755208ce6 70476Author: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 70477Date: Wed Aug 1 18:45:43 2012 -0700 70478 70479 pbproxy: Fix a warning about a bad method prototype 70480 70481 x-selection.m:1502:1: warning: method has no return type specified; defaults to 'id' 70482 [-Wmissing-method-return-type,Semantic Issue] 70483 - init 70484 ^ 70485 (id) 70486 1 warning generated. 70487 70488 Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 70489 70490commit f193907f1b2f8d17dac9b59b7e93ddbb0f4e31ca 70491Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70492Date: Fri Jun 22 14:56:22 2012 +0100 70493 70494 Fix pseudoramiX.c compilation without darwin.h 70495 70496 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70497 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 70498 70499commit 067931ccce592e319baffa26d4f7380d8d924537 70500Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70501Date: Mon Jul 23 12:16:36 2012 +0100 70502 70503 hw/xquartz: Various fixes for pseudoramiX.c 70504 70505 Various fixes, applied to panoramiX.c in commit 2b266eda, also need applying to pseudoramiX.c: 70506 Fix panoramiX request and reply swapping 70507 Set window and screen values in panoramix replies 70508 Prevent buffer overrun in ProcPanoramiXGetScreenSize 70509 70510 These fixes seem to be necessary in order to compile pseudoramiX.c with gcc 70511 70512 pseudoramiX.c: In function 'ProcPseudoramiXGetState': 70513 pseudoramiX.c:221:56: error: call to 'wrong_size' declared with attribute error: wrong sized variable passed to swap 70514 pseudoramiX.c: In function 'ProcPseudoramiXGetScreenCount': 70515 pseudoramiX.c:250:62: error: call to 'wrong_size' declared with attribute error: wrong sized variable passed to swap 70516 pseudoramiX.c: In function 'ProcPseudoramiXGetScreenSize': 70517 pseudoramiX.c:283:56: error: call to 'wrong_size' declared with attribute error: wrong sized variable passed to swap 70518 pseudoramiX.c:284:57: error: call to 'wrong_size' declared with attribute error: wrong sized variable passed to swap 70519 70520 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70521 Reviewed-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com> 70522 70523commit aad428b8e21c77397c623b78706eb64b1fea77c9 70524Author: Dave Airlie <airlied@redhat.com> 70525Date: Thu Jul 26 09:54:08 2012 +1000 70526 70527 glx: drop GLX_LIBS from X server and workaround sdksyms. 70528 70529 We've had reports of two copies of the GLX bits, one in the server 70530 and one in libglx.so causing problems, I didn't understand why the 70531 X server needed a copy so drop it, however then we have to fix a missing 70532 GlxExtensionInit that comes from sdksyms, so work around it by moving 70533 that one declaration into a header that sdksyms doesn't scan. 70534 70535 Thanks to Jon Turney for debugging the actual problem. 70536 (copyright header from extinit.h that seems most appropriate put on top). 70537 70538 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=52402 70539 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 70540 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 70541 Tested-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70542 Signed-off-by: Dave Airlie <airlied@redhat.com> 70543 Signed-off-by: Keith Packard <keithp@keithp.com> 70544 70545commit afa53fe7cffd430cf11f25ca818cb955a78c0c1c 70546Author: Keith Packard <keithp@keithp.com> 70547Date: Wed Jul 25 14:29:48 2012 -0700 70548 70549 Version 1.12.99.903 (1.13 RC3) 70550 70551 Signed-off-by: Keith Packard <keithp@keithp.com> 70552 70553commit 98e3f3fde35ca805d44ee38394bfd86058f9f4b8 70554Author: Adam Jackson <ajax@redhat.com> 70555Date: Thu Jul 19 15:00:05 2012 -0400 70556 70557 doc: Drop XAA references from xorg.conf man page 70558 70559 Signed-off-by: Adam Jackson <ajax@redhat.com> 70560 Reviewed-by: Dave Airlie <airlied@redhat.com> 70561 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70562 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 70563 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 70564 Acked-by: Aaron Plattner <aplattner@nvidia.com> 70565 Signed-off-by: Keith Packard <keithp@keithp.com> 70566 70567commit dea928477bf835a13b2bd2b16c0dbaaf5468130c 70568Author: Adam Jackson <ajax@redhat.com> 70569Date: Thu Jul 19 15:00:04 2012 -0400 70570 70571 xfree86: Drop some dead XAA decls from SDK headers 70572 70573 Signed-off-by: Adam Jackson <ajax@redhat.com> 70574 Reviewed-by: Dave Airlie <airlied@redhat.com> 70575 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70576 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 70577 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 70578 Acked-by: Aaron Plattner <aplattner@nvidia.com> 70579 Signed-off-by: Keith Packard <keithp@keithp.com> 70580 70581commit a44a379a2a4c198dbd6ffd6cc0db376f8fb64327 70582Author: Adam Jackson <ajax@redhat.com> 70583Date: Thu Jul 19 15:00:03 2012 -0400 70584 70585 cw: Remove 70586 70587 XAA was the only consumer. 70588 70589 Signed-off-by: Adam Jackson <ajax@redhat.com> 70590 Reviewed-by: Dave Airlie <airlied@redhat.com> 70591 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70592 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 70593 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 70594 Acked-by: Aaron Plattner <aplattner@nvidia.com> 70595 Signed-off-by: Keith Packard <keithp@keithp.com> 70596 70597commit 20b4d90f32b3ca5def5be2fdf04a91ae6c47f125 70598Merge: 83df16953 fd228d318 70599Author: Keith Packard <keithp@keithp.com> 70600Date: Wed Jul 25 13:52:35 2012 -0700 70601 70602 Merge remote-tracking branch 'jeremyhu/master' 70603 70604commit fd228d318428d0d9dea52ca560a9b66660dd2d02 70605Author: Jeremy Huddleston <jeremyhu@apple.com> 70606Date: Wed Jul 25 11:40:33 2012 -0700 70607 70608 XQuartz: Call xp_window_bring_all_to_front if available in libXplugin 70609 70610 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 70611 70612commit 83df169538bddcf3a61e2c5e02a4f3e10af93b60 70613Author: Aaron Plattner <aplattner@nvidia.com> 70614Date: Fri Jul 20 00:59:20 2012 -0700 70615 70616 dix: don't use "new" as a parameter name 70617 70618 "new" is a reserved word in C++. 70619 70620 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 70621 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70622 Signed-off-by: Keith Packard <keithp@keithp.com> 70623 70624commit 02811f1a9cc35c5f82accb0459cb3910165e7e45 70625Author: Dave Airlie <airlied@redhat.com> 70626Date: Thu Jul 19 14:12:59 2012 +1000 70627 70628 modesetting: add output slave support. 70629 70630 This allows the driver to operate as an output slave. 70631 70632 It adds scan out pixmap, and the capability 70633 checks to make sure they available. 70634 70635 Signed-off-by: Dave Airlie <airlied@redhat.com> 70636 70637commit fa171c5a81b1c226b5da54f35e6726d9c8e13afb 70638Author: Dave Airlie <airlied@redhat.com> 70639Date: Tue Jun 5 14:43:21 2012 +0100 70640 70641 modesetting: add platform bus support 70642 70643commit 709dbc68cdea9d282ba318c1cbc76614e98f41c4 70644Author: Dave Airlie <airlied@redhat.com> 70645Date: Wed Jul 25 15:38:34 2012 +1000 70646 70647 modesetting: fix warning about close being undefined. 70648 70649 Signed-off-by: Dave Airlie <airlied@redhat.com> 70650 70651commit 71b86ea8dd69f21d090edd4f61698ad11b952d5e 70652Author: Dave Airlie <airlied@redhat.com> 70653Date: Wed Jul 25 15:33:52 2012 +1000 70654 70655 modesetting: drop useless xf86PciInfo include 70656 70657 Signed-off-by: Dave Airlie <airlied@redhat.com> 70658 70659commit 3b7f313b3d74f9df876152c55d55cbb339bf2d02 70660Author: Jeremy Huddleston <jeremyhu@apple.com> 70661Date: Tue Jul 24 00:32:30 2012 -0700 70662 70663 XQuartz: Fix runtime regressions introduced by extension loading changes 70664 70665 * GLX is now loaded 70666 * PseudoramiX loading is back in miinitext. It needs to be loaded 70667 before RandR. 70668 70669 Regression-from: 27a624bee66d2797eacf90317054207dcf248028 70670 Regression-from: 5f5bbbe543f65c48ecbb5cce80116a86ca3fbe86 70671 70672 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 70673 70674commit 0d64e881a31d89d28f2ded9b8e83e93632870a16 70675Author: Jeremy Huddleston <jeremyhu@apple.com> 70676Date: Mon Jul 23 23:16:29 2012 -0700 70677 70678 XQuartz: Build fix for __GLXscreen::createContext changes 70679 70680 Regression-from: 96d74138b1c0273e88933220d99a893858b649cd 70681 70682 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 70683 70684commit 77de2994774e7be24a0bad521180628feb1027c6 70685Author: Keith Packard <keithp@keithp.com> 70686Date: Tue Jul 17 15:47:39 2012 -0700 70687 70688 Bump to version 1.12.99.902 70689 70690 Signed-off-by: Keith Packard <keithp@keithp.com> 70691 70692commit beeea70495a8d5c4afc1b1eb2cc06a24ab1a986d 70693Merge: 6e438a0e1 691028029 70694Author: Keith Packard <keithp@keithp.com> 70695Date: Tue Jul 17 12:17:39 2012 -0700 70696 70697 Merge remote-tracking branch 'alanc/master' 70698 70699commit 6e438a0e183f5a6c22070746c038af53d5b935ca 70700Merge: 5884e7ded 75966a418 70701Author: Keith Packard <keithp@keithp.com> 70702Date: Tue Jul 17 12:12:53 2012 -0700 70703 70704 Merge remote-tracking branch 'jturney/master' 70705 70706commit 75966a4186955f3a4625e93796145036c5986d2e 70707Author: Keith Packard <keithp@keithp.com> 70708Date: Fri Jul 13 14:36:25 2012 -0700 70709 70710 xwin: Clean up os wrappers for System, Popen and Pclose on Windows 70711 70712 Popen and Pclose are never used on Windows, so don't bother to even 70713 try to define them. 70714 70715 System(s) was defined as system(s), but the two users of that 70716 function are in xkb, which carefully redefines that as 70717 Win32System. Move Win32System and Win32TempDir to os/utils.c, renaming 70718 Win32System to be just System, which simplifies the xkb code 70719 70720 Signed-off-by: Keith Packard <keithp@keithp.com> 70721 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70722 70723commit 6910280297ab610100dd8e7fbe5bb5cc08968f2f 70724Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70725Date: Sat Jul 14 10:54:36 2012 -0700 70726 70727 Use C99 designated initializers in DRI2GetParam replies 70728 70729 DRI2GetParam was going through review in parallel with main batch of 70730 C99 initialization changes - sync up now that both have landed. 70731 70732 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70733 Reviewed-by: Keith Packard <keithp@keithp.com> 70734 70735commit a6c5b8d3ee053cadf685b0e4de96407ad6df4dd4 70736Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70737Date: Sat Jul 14 10:50:41 2012 -0700 70738 70739 Use C99 designated initializers in RandR 1.4 extension Events 70740 70741 RandR 1.4 was going through review in parallel with main batch of 70742 C99 initialization changes - sync up now that both have landed. 70743 70744 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70745 Reviewed-by: Keith Packard <keithp@keithp.com> 70746 70747commit ada04ef0ca8a1213e865ba2c000ccd6ed6d0e371 70748Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70749Date: Sat Jul 14 10:45:29 2012 -0700 70750 70751 Use C99 designated initializers in RandR 1.4 extension Replies 70752 70753 RandR 1.4 was going through review in parallel with main batch of 70754 C99 initialization changes - sync up now that both have landed. 70755 70756 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70757 Reviewed-by: Keith Packard <keithp@keithp.com> 70758 70759commit fb73f7f40fa46458990038332017d4496caa0691 70760Author: Alan Coopersmith <alan.coopersmith@oracle.com> 70761Date: Sat Jul 14 10:02:52 2012 -0700 70762 70763 Fix two more C99 initialization mistakes using members of same struct 70764 70765 Similar to 34cf559bcf99dad, use temporary variables instead of 70766 referencing members of the struct being initialized in the middle 70767 of the initialization. 70768 70769 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 70770 Reviewed-by: Keith Packard <keithp@keithp.com> 70771 70772commit 5884e7dedecdd82ddbb037360cf9c85143e094b5 70773Author: Aaron Plattner <aplattner@nvidia.com> 70774Date: Mon Jul 16 17:40:31 2012 -0700 70775 70776 xf86: Re-export extension disable flags 70777 70778 These flags were unexported by commit a1d41e311c21eb6627caa0d168e070ceaf90806f, 70779 which moved the declarations around and lost the _X_EXPORT attributes in the 70780 process. Since drivers need these and it's late in the release cycle, just 70781 re-export them for now. 70782 70783 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 70784 Reviewed-by: Andy Ritger <aritger@nvidia.com> 70785 Signed-off-by: Keith Packard <keithp@keithp.com> 70786 70787commit 357ec54f46d31ff77c441c15c322152e04f89a11 70788Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70789Date: Wed Mar 14 13:58:32 2012 +0000 70790 70791 hw/xwin: Refer to xkeyboard-config man page for XKB configuration option values 70792 70793 Also polish the keyboard configuration text a bit 70794 Include man section number in references to setxkbmap 70795 Consistently refer to 'manual page' rather using both that and 'man page' 70796 70797 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70798 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70799 70800commit 5940580f8c12145c1ef4c3eee898676b48d95879 70801Author: Jörg Mensmann <joerg_ml@bitplanet.de> 70802Date: Tue Apr 17 10:48:19 2012 +0100 70803 70804 hw/xwin: Fix command line arguments for multiple monitors. 70805 70806 Moving Xwin to a certain monitor using "-screen 0 @2" would fail, 70807 printing "ddxProcessArgument - screen - Querying monitors failed". 70808 70809 This happened since commit 3ead1d810b0e157078db39712e02ea6dc85216d8, 70810 because EnumDisplayMonitor() returns FALSE if its callback function 70811 returns FALSE (which is not clearly documented), and QueryMonitor() 70812 would then also return FALSE. 70813 70814 Moving back to the old behaviour, where the return value of 70815 EnumDisplayMonitors() is ignored. 70816 70817 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70818 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70819 70820commit 18801f5e5a98d80f4811ade8c98df65175b1935a 70821Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70822Date: Fri Feb 10 09:27:16 2012 +0000 70823 70824 hw/xwin: Remove unused WindowExposuresProcPtr storage in screen private 70825 70826 We don't wrap the WindowExposures screen function 70827 70828 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70829 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70830 70831commit 8fd30859c293035fe505da76d8d7e64131bd5bc0 70832Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70833Date: Fri Feb 10 13:25:32 2012 +0000 70834 70835 hw/xwin: Remove long obsolete README for hw/xwin 70836 70837 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70838 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70839 70840commit 3568d4af4ff2490e79286d242916e81afed57b58 70841Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70842Date: Thu Feb 23 22:32:09 2012 +0000 70843 70844 hw/xwin: Remove a pointless OS version check in winInstallKeyboardHookLL() 70845 70846 Checking for OS later than NT4 SP3 (which we didn't do correctly anyhow, 70847 just checking for NT) is pointless when other code in hw/xwin assumes at 70848 least NT 5.0 70849 70850 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70851 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70852 70853commit 0ed0d3979050c4ee7bef64441d50c494d58d93b0 70854Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70855Date: Thu Jul 5 22:08:38 2012 +0100 70856 70857 hw/xwin: Remove winClipboardDetectUnicodeSupport() 70858 70859 This code for detecting if the Windows clipboard can support unicode 70860 is long obsolete. 70861 70862 All NT versions of Windows support unicode clipboard, so any version 70863 of Windows we can run on must support unicode clipboard. 70864 70865 The -nounicodeclipboard flag to disable use of unicode on the clipboard 70866 is retained. 70867 70868 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70869 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70870 70871commit c98471fad7d85fc9d9959b1c48edd90415f7bf15 70872Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70873Date: Thu Jul 5 18:00:06 2012 +0100 70874 70875 hw/xwin: Consolidate duplicated code as getHwnd() function 70876 70877 Consolidate code to find the native HWND of an X window, stored in the _WINDOWSWM_NATIVE_HWND 70878 property, duplicated in UpdateName() and PreserveWin32Stack() as getHwnd() 70879 70880 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70881 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70882 70883commit df7636dddb4981e14ee7d128c43bb08bdf6a8727 70884Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70885Date: Thu Jul 5 14:20:52 2012 +0100 70886 70887 hw/xwin: Correct function name in log message 70888 70889 Correct function name in log message winInitializeDefaultScreens -> winInitializeScreenDefaults 70890 70891 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70892 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70893 70894commit 9cdd3ed07425b222e8197b9a2f57ec1525c6fd54 70895Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70896Date: Fri Mar 16 12:56:45 2012 +0000 70897 70898 hw/xwin: Tidy up some more pointless output which is always emitted 70899 70900 It's not very useful to log lines like the below when the About.. dialog is used 70901 70902 winAboutDlgProc - WM_COMMAND - IDOK or IDCANCEL 70903 winAboutDlgProc - WM_CLOSE 70904 70905 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70906 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70907 70908commit eb5b4d8d809712053701f057a35bbf5348ac9059 70909Author: Colin Harrison <colin.harrison@virgin.net> 70910Date: Thu Jul 5 14:03:44 2012 +0100 70911 70912 hw/xwin: RANDR resize didn't occur when screen window was maximized or restored 70913 70914 In windowed mode, with option -resize, RANDR resize was missing when the screen 70915 window was maximized or restored. 70916 70917 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 70918 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70919 70920commit 7e07d61857c5d2ee6e3c36c81b51bb4fe378b5a0 70921Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70922Date: Thu Jul 5 14:01:03 2012 +0100 70923 70924 hw/xwin: Bug fixes for screen sizing when the screen window is on a non-primary monitor 70925 70926 There is a bug that when the -screen option is used to specify a monitor for 70927 the screen window to be located on, but no explicit size is specified 70928 (and the -multiplemonitors option isn't specified), the screen 70929 window size is always constrained to fit the work area of the primary 70930 monitor (rather than the work area of the specified monitor) 70931 70932 This gives incorrect results if you want a screen the same size as your 70933 non-primary monitor (e.g. by using -screen 0 @2) and your non-primary 70934 monitor is larger than your primary monitor. 70935 70936 (This can be worked around by specifying -multiplemonitors and an explicit 70937 screen size the same size as the monitor (e.g. -multiplemonitors -screen 0 70938 1600x1200@2)) 70939 70940 Fix to use work area for the monitor specified for the screen, rather than the 70941 primary monitor work area (unless -multiplemonitors is used, in which case we 70942 continue to use the virtual desktop work area instead) 70943 70944 Also fix the adjustment for an autohide taskbar, so that it is only done if the 70945 taskbar is on the same monitor as the screen (or -multiplemonitors is used) 70946 70947 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70948 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70949 70950commit 090856b55b5895ce07f36ddebe6d8f7cb86c1826 70951Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70952Date: Thu Jul 5 13:19:02 2012 +0100 70953 70954 hw/xwin: Tidy up some pointless output which is always emitted 70955 70956 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70957 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70958 70959commit 9fe9c5f868886abf671b3be3964af1965dd92472 70960Author: Jon TURNEY <jon.turney@dronecode.org.uk> 70961Date: Sun Feb 13 18:52:52 2011 +0000 70962 70963 hw/xwin: Don't bother reporting XORG_VERSION_CURRENT 70964 70965 Don't bother reporting XORG_VERSION_CURRENT, when we also 70966 report the version number broken down into it's components 70967 XORG_VERSION_MAJOR, XORG_VERSION_MINOR, etc. as well. 70968 70969 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 70970 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 70971 70972commit d5977e5bd2c8ee77db08d5f948f266c9661b9fb5 70973Author: Dave Airlie <airlied@gmail.com> 70974Date: Fri Jul 13 00:28:19 2012 +1000 70975 70976 xf86: include xf86platformBus.h in xf86AutoConfig.c 70977 70978 This fixes an implicit declaration, 70979 xf86AutoConfig.c:202:5: error: implicit declaration of function 'xf86PlatformMatchDriver' [-Werror=implicit-function-declaration] 70980 xf86AutoConfig.c:202:5: warning: nested extern declaration of 'xf86PlatformMatchDriver' [-Wnested-externs] 70981 70982 Signed-off-by: Dave Airlie <airlied@redhat.com> 70983 Reviewed-by: Keith Packard <keithp@keithp.com> 70984 70985commit afa3b9ccd27ff9c58f536cb3b727e85feb6a6f6a 70986Author: Bartosz Brachaczek <b.brachaczek@gmail.com> 70987Date: Fri Jul 6 02:52:40 2012 +0200 70988 70989 glx: Fix checking GL version 70990 70991 Signed-off-by: Bartosz Brachaczek <b.brachaczek@gmail.com> 70992 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 70993 Signed-off-by: Dave Airlie <airlied@gmail.com> 70994 70995commit 743ccf6fa0a9a4579d6de6d10e2e6040640772d9 70996Author: Dave Airlie <airlied@gmail.com> 70997Date: Thu Jul 12 19:09:15 2012 +1000 70998 70999 glx/dri2: don't return NULL in a function with no return. 71000 71001 Reported by gcc. 71002 71003 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 71004 Reviewed-by: Matt Turner <mattst88@gmail.com> 71005 Signed-off-by: Dave Airlie <airlied@redhat.com> 71006 71007commit d1a00c26becdc2490ab4660abf108eca09c30d23 71008Author: Dave Airlie <airlied@gmail.com> 71009Date: Thu Jul 12 19:05:41 2012 +1000 71010 71011 glx/dri2: use correct define from dri attribs. 71012 71013 Looks like idr renamed this and pushed the wrong one. 71014 71015 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 71016 Reviewed-by: Matt Turner <mattst88@gmail.com> 71017 Signed-off-by: Dave Airlie <airlied@redhat.com> 71018 71019commit c62205adcd48d497450655d0136f8ca5bd5cbae7 71020Author: Dave Airlie <airlied@gmail.com> 71021Date: Thu Jul 12 19:04:13 2012 +1000 71022 71023 glx/dri2: fix incorrect == vs = 71024 71025 glxdri2.c:486: warning: statement with no effect 71026 71027 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 71028 Reviewed-by: Matt Turner <mattst88@gmail.com> 71029 Signed-off-by: Dave Airlie <airlied@redhat.com> 71030 71031commit defca4c192bd7f4af2b273c7cb1fe8665888772a 71032Author: Dave Airlie <airlied@gmail.com> 71033Date: Thu Jul 12 18:52:34 2012 +1000 71034 71035 glx/dri2: fix typo in the notification reset for robustness. 71036 71037 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 71038 Reviewed-by: Matt Turner <mattst88@gmail.com> 71039 Signed-off-by: Dave Airlie <airlied@redhat.com> 71040 71041commit d1cae6bb804f7dfc889fd3c5d5eb116f4baf5149 71042Author: Dave Airlie <airlied@gmail.com> 71043Date: Thu Jul 12 18:51:52 2012 +1000 71044 71045 glx: pass screen into the convert function 71046 71047 when robustness is enabled, this is required. 71048 71049 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 71050 Signed-off-by: Dave Airlie <airlied@redhat.com> 71051 71052commit ea74e381ae91f45ec195a5505262e2e5a3c0e1c9 71053Author: Dave Airlie <airlied@gmail.com> 71054Date: Fri Jul 13 00:47:24 2012 +1000 71055 71056 xtest: add extinit.h for SyncExtensionInit 71057 71058 This adds the decl for SyncExtenionInit. 71059 71060 Signed-off-by: Dave Airlie <airlied@redhat.com> 71061 Tested-by: Keith Packard <keithp@keithp.com> 71062 Signed-off-by: Keith Packard <keithp@keithp.com> 71063 71064commit 746193a7f10f11cccae05c98b933aadf58fc7375 71065Merge: a2d082953 2b74949ad 71066Author: Keith Packard <keithp@keithp.com> 71067Date: Thu Jul 12 15:10:51 2012 -0700 71068 71069 Merge remote-tracking branch 'jturney/xwin-extmod-removal-fixes' 71070 71071commit a2d0829531249e24dbca25fc20ed30a2bb2d8ed8 71072Author: Michel Dänzer <michel.daenzer@amd.com> 71073Date: Thu Jul 12 13:16:37 2012 +0200 71074 71075 glx: Free DRI2 drawable reference to destroyed GLX drawable. 71076 71077 Otherwise the reference can lead to use after free in 71078 __glXDRIinvalidateBuffers(). 71079 71080 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=50019 71081 71082 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 71083 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 71084 Signed-off-by: Keith Packard <keithp@keithp.com> 71085 71086commit 8a87acc9e57be740e3bd252567622246051a0723 71087Author: Michel Dänzer <michel.daenzer@amd.com> 71088Date: Thu Jul 12 13:16:36 2012 +0200 71089 71090 dri2: Add DRI2CreateDrawable2. 71091 71092 Same as DRI2CreateDrawable, except it can return the DRI2 specific XID of the 71093 DRI2 drawable reference to the base drawable. 71094 71095 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 71096 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 71097 Signed-off-by: Keith Packard <keithp@keithp.com> 71098 71099commit 2b74949ad02bd916c3ac502db3f28057f12a8117 71100Author: Jon TURNEY <jon.turney@dronecode.org.uk> 71101Date: Thu Jul 12 00:36:43 2012 +0100 71102 71103 Restore GLX extension to XWin 71104 71105 - Add an XwinExtensionInit(), similar in spirit to QuartzExtensionInit() 71106 and xf86ExtensionInit() 71107 - Load the GLX extension (I'm guessing XQuartz needs this adding somewhere 71108 as well) 71109 - Also, since we now have a proper place to install the native GL provider 71110 where it will take priority over the the swrast provider, do so 71111 71112 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 71113 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71114 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 71115 71116commit 656e5fea6c9ca15f932889edcc590dd01c881945 71117Merge: 8b820f221 1720cd9ba 71118Author: Keith Packard <keithp@keithp.com> 71119Date: Thu Jul 12 08:26:35 2012 -0700 71120 71121 Merge remote-tracking branch 'jeremyhu/master' 71122 71123commit 8b820f221ad54eb96a16c7060048abc8824371b6 71124Author: Daniel Stone <daniel@fooishbar.org> 71125Date: Wed Jul 11 17:31:53 2012 +0100 71126 71127 sdksyms: Fix build with --disable-xv 71128 71129 I hate this [redacted] script. 71130 71131 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71132 Tested-by: Keith Packard <keithp@keithp.com> 71133 Signed-off-by: Keith Packard <keithp@keithp.com> 71134 71135commit 06ac7937fc37cdf073b77386bed47e14cff23628 71136Author: Jon TURNEY <jon.turney@dronecode.org.uk> 71137Date: Thu Jul 12 12:37:05 2012 +0100 71138 71139 Restore extern to the declaration of __glXDRISWRastProvider 71140 71141 b86aa74 dropped the 'extern' from the declaration of __glXDRISWRastProvider 71142 71143 This turns out to be important to me, as without it, the final link only gets 71144 the tentative definition of __glXDRISWRastProvider implied by the declaration, 71145 and not the proper one from glxdriswrast.c, presumably because nothing else 71146 references anything in the object that file generates. 71147 71148 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 71149 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71150 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 71151 71152commit bcbf95b1bafa6ffe724768b9309295e2fdb4b860 71153Author: Jon TURNEY <jon.turney@dronecode.org.uk> 71154Date: Thu Jul 12 00:36:10 2012 +0100 71155 71156 Revert bogus GlxPushProvider() in commit a1d41e3 71157 71158 a1d41e3 "Move extension initialisation prototypes into extinit.h" 71159 also includes a change to GlxExtensionInit to install the swrast GLX 71160 provider. 71161 71162 Since b86aa74 "GLX: Insert swrast provider from GlxExtensionInit" 71163 already does this (correctly, by installing the swrast provider 71164 at the end of the chain, rather than at the beginning), and since this 71165 would seem to have the effect of making the swrast provider the most 71166 preferred provider, I'm guessing this wasn't intended. 71167 71168 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 71169 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71170 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 71171 71172commit 1720cd9badf21e5d6b2c6c929e609ab3b0b92fd2 71173Author: Jeremy Huddleston <jeremyhu@apple.com> 71174Date: Wed Jul 11 23:54:27 2012 -0700 71175 71176 XQuartz: Re-add prototype for DRIExtensionInit 71177 71178 Regression from: a1d41e311c21eb6627caa0d168e070ceaf90806f 71179 71180 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 71181 CC: Daniel Stone <daniel@fooishbar.org> 71182 71183commit 1a64d3cc9c319d5e9ab2365b60469fb5b83407fc 71184Author: Jeremy Huddleston <jeremyhu@apple.com> 71185Date: Wed Jul 11 23:49:59 2012 -0700 71186 71187 XQuartz: Fix build failure due to incorrect PseudoramiXExtensionInit prototype 71188 71189 pseudoramiX.c:129:1: error: conflicting types for 'PseudoramiXExtensionInit' [Semantic Issue] 71190 PseudoramiXExtensionInit(int argc, char *argv[]) 71191 ^ 71192 ../../include/extinit.h:175:13: note: previous declaration is here 71193 extern void PseudoramiXExtensionInit(void); 71194 ^ 71195 71196 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 71197 71198commit 228f8bfe81890770f1ba109a0acca1c81a2ffee9 71199Author: Jeremy Huddleston <jeremyhu@apple.com> 71200Date: Wed Jul 11 23:49:33 2012 -0700 71201 71202 XQuartz: Silence an unused-variable warning 71203 71204 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 71205 71206commit deb08658e2a6b1647a7213a316c6f3019bcdce48 71207Author: Maarten Lankhorst <maarten.lankhorst@canonical.com> 71208Date: Wed Jul 11 16:27:46 2012 +0200 71209 71210 xfree86: Strip dangling pointers from desiredMode 71211 71212 Based on the original patch by Chris Wilson, which was a better fix than mine. 71213 71214 We stash a copy of the desiredMode on the crtc so that we can restore it 71215 after a vt switch. This copy is a simple memcpy and so also stashes a 71216 references to the pointers contained within the desiredMode. Those 71217 pointers are freed the next time the outputs are probed and mode list 71218 rebuilt, resulting in us chasing those dangling pointers on the next 71219 mode switch. 71220 71221 ==22787== Invalid read of size 1 71222 ==22787== at 0x40293C2: __GI_strlen (in 71223 /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 71224 ==22787== by 0x668F875: strdup (strdup.c:42) 71225 ==22787== by 0x5DBA00: XNFstrdup (utils.c:1124) 71226 ==22787== by 0x4D72ED: xf86DuplicateMode (xf86Modes.c:209) 71227 ==22787== by 0x4CA848: xf86CrtcSetModeTransform (xf86Crtc.c:276) 71228 ==22787== by 0x4D05B4: xf86SetDesiredModes (xf86Crtc.c:2677) 71229 ==22787== by 0xA7479D0: sna_create_screen_resources 71230 (sna_driver.c:220) 71231 ==22787== by 0x4CB914: xf86CrtcCreateScreenResources (xf86Crtc.c:725) 71232 ==22787== by 0x425498: main (main.c:216) 71233 ==22787== Address 0x72c60e0 is 0 bytes inside a block of size 9 free'd 71234 ==22787== at 0x4027AAE: free (in 71235 /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 71236 ==22787== by 0x4A547E: xf86DeleteMode (xf86Mode.c:1984) 71237 ==22787== by 0x4CD84F: xf86ProbeOutputModes (xf86Crtc.c:1578) 71238 ==22787== by 0x4DC405: xf86RandR12GetInfo12 (xf86RandR12.c:1537) 71239 ==22787== by 0x518119: RRGetInfo (rrinfo.c:202) 71240 ==22787== by 0x51D997: rrGetScreenResources (rrscreen.c:335) 71241 ==22787== by 0x51E0D0: ProcRRGetScreenResources (rrscreen.c:475) 71242 ==22787== by 0x513852: ProcRRDispatch (randr.c:493) 71243 ==22787== by 0x4346DB: Dispatch (dispatch.c:439) 71244 ==22787== by 0x4256E4: main (main.c:287) 71245 71246 Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> 71247 Reported-by: Zdenek Kabelac <zdenek.kabelac@gmail.com> 71248 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=36108 71249 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 71250 Signed-off-by: Keith Packard <keithp@keithp.com> 71251 71252commit 59c2c4f645b4d782599c274f4bc752de9623d308 71253Author: Daniel Stone <daniel@fooishbar.org> 71254Date: Tue Jul 10 09:10:22 2012 +0100 71255 71256 AllocDevicePair: Ensure XKB privates are initialised 71257 71258 Since we call directly into XKB and may be doing so before the extension 71259 has been initialised, make sure its privates are set up first. XTest 71260 had a hack to do this itself, but seems cleaner to just make sure we do 71261 it in AllocDevicePair. 71262 71263 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71264 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71265 Signed-off-by: Keith Packard <keithp@keithp.com> 71266 71267commit f937c3d50157a9a42d2f2f10590b6b139ce18a29 71268Author: Daniel Stone <daniel@fooishbar.org> 71269Date: Tue Jul 10 09:10:21 2012 +0100 71270 71271 Sync: Always initialise system counter list 71272 71273 Sync is designed to let you add system counters before the extension has 71274 been initialised, which means the system counter list may well be full 71275 of bees. Make sure it's initialised before we add to it, to avoid the 71276 risk of fatal injury. 71277 71278 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71279 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 71280 Signed-off-by: Keith Packard <keithp@keithp.com> 71281 71282commit 34cf559bcf99dad550527b5ff53f247f0e8e73ee 71283Author: Keith Packard <keithp@keithp.com> 71284Date: Tue Jul 10 15:58:48 2012 -0700 71285 71286 ProcGetPointerMapping uses rep.nElts before it is initialized 71287 71288 In: 71289 71290 commit d792ac125a0462a04a930af543cbc732f8cdab7d 71291 Author: Alan Coopersmith <alan.coopersmith@oracle.com> 71292 Date: Mon Jul 9 19:12:43 2012 -0700 71293 71294 Use C99 designated initializers in dix Replies 71295 71296 the initializer for the .length element of the xGetPointerMappingReply 71297 structure uses the value of rep.nElts, but that won't be set until 71298 after this initializer runs, so we get garbage in the length element 71299 and clients using it will generally wedge. 71300 71301 Easy to verify: 71302 71303 $ xmodmap -pp 71304 71305 Fixed by creating a local nElts variable and using that. 71306 71307 Signed-off-by: Keith Packard <keithp@keithp.com> 71308 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 71309 71310commit 44bd27cdd13b6e6c4e6fd3873361e3308ef25ffe 71311Author: Keith Packard <keithp@keithp.com> 71312Date: Tue Jul 10 01:14:13 2012 -0700 71313 71314 Version 1.12.99.901 (1.13 RC1) 71315 71316 Signed-off-by: Keith Packard <keithp@keithp.com> 71317 71318commit 6e12cb147da58bd12a158807b2cd5afac32ae70f 71319Merge: 532fbc29c 023127915 71320Author: Keith Packard <keithp@keithp.com> 71321Date: Tue Jul 10 00:52:11 2012 -0700 71322 71323 Merge branch 'local-fixes' 71324 71325commit 532fbc29c900469ef4867ccb7eabb953019dcab5 71326Author: Daniel Stone <daniel@fooishbar.org> 71327Date: Tue Jul 10 02:03:38 2012 +0100 71328 71329 GE: Remove unused RT_GECLIENT 71330 71331 Remove remnants of an earlier experiment which had the GE extension 71332 handling event delivery directly. Nothing's used the resource since, so 71333 purge it. 71334 71335 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71336 Reviewed-by: Adam Jackson <ajax@redhat.com> 71337 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71338 Signed-off-by: Keith Packard <keithp@keithp.com> 71339 71340commit 0e70b333d4846b166bf4a0a0c5d1a80623ba08dc 71341Author: Daniel Stone <daniel@fooishbar.org> 71342Date: Tue Jul 10 02:03:37 2012 +0100 71343 71344 XFree86: os-support: Remove unused xf86MakeNewMapping 71345 71346 No drivers used this, so it got unexported, and now it's so unused it 71347 got culled during the link. Take the poor function out behind the shed 71348 and put it out of its misery. 71349 71350 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71351 Reviewed-by: Adam Jackson <ajax@redhat.com> 71352 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 71353 Signed-off-by: Keith Packard <keithp@keithp.com> 71354 71355commit 43f2393746c44e9d35be6a2a3555205f8e1b4ddf 71356Author: Daniel Stone <daniel@fooishbar.org> 71357Date: Tue Jul 10 02:03:36 2012 +0100 71358 71359 Xinerama: Reset Composite request vector on close 71360 71361 Similar (identical) to how it interacts with Render and XFixes, also 71362 call PanoramiXCompositeReset() to restore the Composite dispatch table 71363 to how it was when it started, on reset. 71364 71365 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71366 Reviewed-by: Adam Jackson <ajax@redhat.com> 71367 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 71368 Signed-off-by: Keith Packard <keithp@keithp.com> 71369 71370commit 766a403bc7ff249fdc6b9049329b633e73738329 71371Author: Daniel Stone <daniel@fooishbar.org> 71372Date: Tue Jul 10 02:03:35 2012 +0100 71373 71374 XvMC: Remove unused XvMCGeneration 71375 71376 I wonder if this even works across multiple generations. 71377 71378 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71379 Reviewed-by: Adam Jackson <ajax@redhat.com> 71380 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 71381 Signed-off-by: Keith Packard <keithp@keithp.com> 71382 71383commit 7cef23a6dd7a5d76983cf20979e845ea6ae63a10 71384Author: Daniel Stone <daniel@fooishbar.org> 71385Date: Tue Jul 10 02:03:34 2012 +0100 71386 71387 Render: Remove unused glyphDepths 71388 71389 No-one has used this since 0a71e154. 71390 71391 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71392 Reviewed-by: Adam Jackson <ajax@redhat.com> 71393 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 71394 Signed-off-by: Keith Packard <keithp@keithp.com> 71395 71396commit 65011064d7e70aed59fb716f9ed3d76d196244ce 71397Author: Daniel Stone <daniel@fooishbar.org> 71398Date: Tue Jul 10 02:03:33 2012 +0100 71399 71400 Xi: Remove dead Device{Enter,Leave}WindowMask 71401 71402 These were an unused remnant of earlier MPX work; their only users got 71403 cleared out in dc153271, but the mask declarations remained. Remove 71404 them, and move DevicePropertyNotify's mask up to be contiguous with the 71405 rest of the range. 71406 71407 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71408 Reviewed-by: Adam Jackson <ajax@redhat.com> 71409 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71410 Signed-off-by: Keith Packard <keithp@keithp.com> 71411 71412commit fadfc2947075c832cd1107d2c0866ad13fa756c4 71413Author: Daniel Stone <daniel@fooishbar.org> 71414Date: Tue Jul 10 02:03:32 2012 +0100 71415 71416 GLX: Remove unused noGlxVisualInit 71417 71418 No-one ever did anything with this variable except assign its default 71419 value to it. 71420 71421 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71422 Reviewed-by: Adam Jackson <ajax@redhat.com> 71423 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 71424 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 71425 Signed-off-by: Keith Packard <keithp@keithp.com> 71426 71427commit 06e4ba8b26fb974bdd84144b3e002f0f7589e7f7 71428Author: Daniel Stone <daniel@fooishbar.org> 71429Date: Tue Jul 10 02:03:31 2012 +0100 71430 71431 XKB: Geom: Remove unused code 71432 71433 These codepaths were never called by anyone. Shame there weren't more 71434 of them. 71435 71436 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71437 Reviewed-by: Adam Jackson <ajax@redhat.com> 71438 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71439 Signed-off-by: Keith Packard <keithp@keithp.com> 71440 71441commit e191e296e6e7861978ea4a0ae9aa7b780e52732b 71442Author: Daniel Stone <daniel@fooishbar.org> 71443Date: Mon Jul 9 20:04:15 2012 +0100 71444 71445 Remove XAA 71446 71447 Commit 0c6987df in June 2008 disabled XAA offscreen pixmaps per default, 71448 as they were broken, leaving XAA only able to accelerate operations 71449 directly on the screen pixmap and nowhere else, eliminating acceleration 71450 for basically every modern toolkit, and any composited environment. 71451 71452 So, it hasn't worked for over four years. No-one's even come close to 71453 fixing it. 71454 71455 RIP. 71456 71457 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71458 Reviewed-by: Dave Airlie <airlied@redhat.com> 71459 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 71460 71461commit 66362441867f68dac97de33744bd0ca478fb65d3 71462Author: Daniel Stone <daniel@fooishbar.org> 71463Date: Tue Jul 10 02:03:29 2012 +0100 71464 71465 Remove last renmants of XFree86LOADER define 71466 71467 Not to be confused with XFree86Loader or XorgLoader. Which are both now 71468 dead too. 71469 71470 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71471 Reviewed-by: Keith Packard <keithp@keithp.com> 71472 Signed-off-by: Keith Packard <keithp@keithp.com> 71473 71474commit a15dac550918c55b08bbe0007f068cc5096ebaa1 71475Author: Daniel Stone <daniel@fooishbar.org> 71476Date: Tue Jul 10 02:03:28 2012 +0100 71477 71478 Remove unused setupFunc from extensions 71479 71480 setupFunc was used as an early callback for half-modular extensions such 71481 as Xv, XvMC and DGA to set up hooks between the core server and the 71482 modular component. Now we've rid ourselves of that, we can also bin 71483 setupFunc. 71484 71485 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71486 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71487 Signed-off-by: Keith Packard <keithp@keithp.com> 71488 71489commit 710318e803c6d5c6b8ae170d38c8253f5d780472 71490Author: Daniel Stone <daniel@fooishbar.org> 71491Date: Tue Jul 10 02:03:27 2012 +0100 71492 71493 Replace NO_HW_ONLY_EXTS with Xorg DDX test 71494 71495 Turns out the only thing we use NO_HW_ONLY_EXTS for is to check whether 71496 or not we're building inside the Xorg DDX. Replace it with an 71497 XorgLoader test instead, and remove all its users. 71498 71499 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71500 Reviewed-by: Keith Packard <keithp@keithp.com> 71501 Signed-off-by: Keith Packard <keithp@keithp.com> 71502 71503commit 5f5bbbe543f65c48ecbb5cce80116a86ca3fbe86 71504Author: Daniel Stone <daniel@fooishbar.org> 71505Date: Tue Jul 10 02:03:26 2012 +0100 71506 71507 Unify miinitext.c 71508 71509 Rather than having a non-Xorg and an Xorg-specific path which basically 71510 just duplicated each other for no reason, we could ... just have one. 71511 71512 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71513 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71514 Signed-off-by: Keith Packard <keithp@keithp.com> 71515 71516commit 147967f67d4976a87ed07fe11b59b174907f5d43 71517Author: Daniel Stone <daniel@fooishbar.org> 71518Date: Tue Jul 10 02:03:25 2012 +0100 71519 71520 Remove Xorg-specific extensions from non-Xorg miinitext 71521 71522 miinitext.c had a completely separate codepath for non-Xorg servers, 71523 which included tests for Xorg-specific extensions such as 71524 XFree86-VidMode, which were external even to the Xorg DDX. So we can 71525 just remove them, and the associated #undefs. 71526 71527 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71528 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71529 Signed-off-by: Keith Packard <keithp@keithp.com> 71530 71531commit 27a624bee66d2797eacf90317054207dcf248028 71532Author: Daniel Stone <daniel@fooishbar.org> 71533Date: Tue Jul 10 02:03:24 2012 +0100 71534 71535 Quartz: Move PseudoramiX initialisation to DDX 71536 71537 As PseudoramiX is a DDX-specific extension, move its loading and 71538 initialisation to hw/xquartz. This creates a QuartzExtensionInit() 71539 similar in spirit to xf86ExtensionInit. 71540 71541 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71542 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 71543 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 71544 Signed-off-by: Keith Packard <keithp@keithp.com> 71545 71546commit 79d14d9fc3d92ba866864d3a5910a8c6c8e9b64f 71547Author: Daniel Stone <daniel@fooishbar.org> 71548Date: Tue Jul 10 02:03:23 2012 +0100 71549 71550 Move the remnants of loadext.c to miinitext.c 71551 71552 There was nothing XFree86-specific or loader-specific about this, aside 71553 from using xf86MsgVerb instead of ErrorF. 71554 71555 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71556 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71557 Signed-off-by: Keith Packard <keithp@keithp.com> 71558 71559commit a089af3477dd01a831b89c06f982679c60e42975 71560Author: Daniel Stone <daniel@fooishbar.org> 71561Date: Tue Jul 10 02:03:22 2012 +0100 71562 71563 Loader: Move ExtensionModule types to DIX 71564 71565 In preparation for gutting loadext.c, move the ExtensionModule struct to 71566 the DIX, and unexport ExtensionModuleList (why, why, why, why was this 71567 ever exported in the first place, tbqh). 71568 71569 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71570 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71571 Signed-off-by: Keith Packard <keithp@keithp.com> 71572 71573commit 8171108602556446f12c1cfbc37c6abf3b0195eb 71574Author: Daniel Stone <daniel@fooishbar.org> 71575Date: Tue Jul 10 02:03:21 2012 +0100 71576 71577 Loader: Remove extension initialisation sorting 71578 71579 Extensions could previously declare initialisation dependencies on other 71580 extensions, which would then get nicely sorted by the loader. We only 71581 had one user for this, GLX, which had one pointless (Composite) and one 71582 possibly useful dependency (DBE). As DBE is now a built-in, it will 71583 always be sorted by GLX, so we no longer have any users for it. 71584 71585 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71586 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71587 Signed-off-by: Keith Packard <keithp@keithp.com> 71588 71589commit d52ab85c7ef5058dcbd8a75eb84058908dc22906 71590Author: Daniel Stone <daniel@fooishbar.org> 71591Date: Tue Jul 10 02:03:20 2012 +0100 71592 71593 GLX: Remove extension init dependencies 71594 71595 GLX was the only user of extension init order dependencies, using them 71596 to depend on Composite, which has always been built-in anyway, and DBE, 71597 which is now built-in. 71598 71599 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71600 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71601 Signed-off-by: Keith Packard <keithp@keithp.com> 71602 71603commit 2fba9445a0357f67641e41ac334b5529c37774a2 71604Author: Daniel Stone <daniel@fooishbar.org> 71605Date: Tue Jul 10 02:03:19 2012 +0100 71606 71607 Add static extensions before those in modules 71608 71609 Make sure we add static extensions before anything in a module. This is 71610 more or less a no-op at the moment, but will come in handy later when 71611 extension dependency sorting is removed. 71612 71613 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71614 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71615 Signed-off-by: Keith Packard <keithp@keithp.com> 71616 71617commit 9a953e0e9dcb8a8e43cc27ffaef460268fbe1916 71618Author: Daniel Stone <daniel@fooishbar.org> 71619Date: Tue Jul 10 02:03:18 2012 +0100 71620 71621 Move DRI2 from external module to built-in 71622 71623 Instead of keeping a tiny amount of code in an external module, just man 71624 up and build it into the core server. 71625 71626 v2: Fix test/Makefile.am to only link libdri2.la if DRI2 is set 71627 71628 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71629 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71630 Signed-off-by: Keith Packard <keithp@keithp.com> 71631 71632commit b8a3267c36e2e335b888bd4f2ef2f2c477cdfdce 71633Author: Daniel Stone <daniel@fooishbar.org> 71634Date: Tue Jul 10 02:03:17 2012 +0100 71635 71636 DRI2: Remove prototype for DRI2DestroyDrawable 71637 71638 DRI2DestroyDrawable() was still being _X_EXPORTed, but hasn't existed 71639 since 1da1f33f last year. 71640 71641 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71642 Reviewed-by: Cyril Brulebois <kibi@debian.org> 71643 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71644 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 71645 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71646 Signed-off-by: Keith Packard <keithp@keithp.com> 71647 71648commit 7025a909bfe4d1f5518ec6fc5f89e2c3fb6501fd 71649Author: Daniel Stone <daniel@fooishbar.org> 71650Date: Tue Jul 10 02:03:16 2012 +0100 71651 71652 XFree86: DRI: Don't use per-target CFLAGS 71653 71654 AM_CFLAGS will suffice, given we only have one target in this directory. 71655 71656 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71657 Reviewed-by: Cyril Brulebois <kibi@debian.org> 71658 Signed-off-by: Keith Packard <keithp@keithp.com> 71659 71660commit 459c6da0f907ba33d733c7e62a116184ba2f14e5 71661Author: Daniel Stone <daniel@fooishbar.org> 71662Date: Tue Jul 10 02:03:15 2012 +0100 71663 71664 Move DRI1 from external module to built-in 71665 71666 Rather than building the tiny amount of code required for XFree86-DRI as 71667 an external module, build it in if it's enabled at configure time. 71668 71669 v2: Fix test/Makefile.am to only link libdri.la if DRI is set 71670 71671 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71672 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71673 Signed-off-by: Keith Packard <keithp@keithp.com> 71674 71675 fixup for DRI1 move 71676 71677 Signed-off-by: Keith Packard <keithp@keithp.com> 71678 71679commit b905d87ba0a9b045ea41bb1ad36a4f9f1170a1b8 71680Author: Daniel Stone <daniel@fooishbar.org> 71681Date: Tue Jul 10 02:03:14 2012 +0100 71682 71683 Xext: Only build one library 71684 71685 Now that libXextmodule.la is both empty and unused, we can just build 71686 the one libXext.la for everyone, rather than having Xorg be special and 71687 unique. 71688 71689 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71690 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71691 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71692 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71693 Signed-off-by: Keith Packard <keithp@keithp.com> 71694 71695commit a7a2f9f66ddc27df96557093f3dd266e24eb653b 71696Author: Daniel Stone <daniel@fooishbar.org> 71697Date: Tue Jul 10 02:03:13 2012 +0100 71698 71699 Remove the last remnants of extmod 71700 71701 extmod was originally a big pointless module. Now it's an empty, 71702 pointless module. This commit makes it unexist. 71703 71704 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71705 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71706 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71707 Signed-off-by: Keith Packard <keithp@keithp.com> 71708 71709commit 6e74fdda42fb4bd80fd48e1ae6471bffbb837804 71710Author: Daniel Stone <daniel@fooishbar.org> 71711Date: Wed Jun 22 00:28:18 2011 +0100 71712 71713 Move XFree86-VidMode from extmod to built-in 71714 71715 As with DGA, move VidMode from being part of extmod to a built-in part 71716 of the server, if compiled as such. This is initialised from 71717 xf86ExtensionInit rather than miinitext because it's wholly dependent on 71718 the Xorg DDX. 71719 71720 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71721 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71722 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71723 71724commit 60f53e3012be795eee741ecd4be858552c5e86c8 71725Author: Tomas Carnecky <tom@dbservice.com> 71726Date: Tue Jul 10 02:03:11 2012 +0100 71727 71728 DGA: Remove excessive module-induced indirection 71729 71730 The DGA event base used to have to be passed through a function pointer, 71731 as the code was cleaved in two with half in a module, and half in the 71732 core server. Now that's not the case, just access DGAEventBase 71733 directly. 71734 71735 v2: Deal with Alan's event initialization cleanups 71736 71737 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 71738 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71739 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71740 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71741 Signed-off-by: Keith Packard <keithp@keithp.com> 71742 71743commit 7a5880bc3b47e1f90a638f056ad4d40be6931977 71744Author: Daniel Stone <daniel@fooishbar.org> 71745Date: Tue Jul 10 02:03:10 2012 +0100 71746 71747 Move DGA from extmod to built-in 71748 71749 Rather than leave DGA languishing in extmod, move it to be a built-in 71750 extension. As it's quite specific to the Xorg DDX, just move it 71751 sideways to the rest of the DGA code in hw/xfree86/common, and 71752 initialise it from xf86ExtensionInit, rather than miinitext. 71753 71754 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71755 Acked-by: Keith Packard <keithp@keithp.com> 71756 Signed-off-by: Keith Packard <keithp@keithp.com> 71757 71758commit 6fb481d1251626456914d0012d6ca5a342bfdf71 71759Author: Daniel Stone <daniel@fooishbar.org> 71760Date: Tue Jul 10 02:03:09 2012 +0100 71761 71762 Move SELinux from extmod to built-in 71763 71764 Instead of letting it languish in extmod just because we want to 71765 configure bits of it from xf86, move XSELinux to the builtin part of 71766 Xext, and do its configuration from xf86ExtensionInit. 71767 71768 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71769 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71770 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71771 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71772 Signed-off-by: Keith Packard <keithp@keithp.com> 71773 71774commit 5d92ee4081a23a6b4d1f9ba2a98985f539f99312 71775Author: Tomas Carnecky <tom@dbservice.com> 71776Date: Tue Jul 10 02:03:08 2012 +0100 71777 71778 Xv: Remove excessive module-induced indirection 71779 71780 Xv used to call XvScreenInit and co. through function pointers, as 71781 XvScreenInit may have been sitting on the other side of a module 71782 boundary from xf86XvScreenInit. Why this was so is a mystery, but make 71783 it not so any more. 71784 71785 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 71786 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71787 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 71788 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71789 Signed-off-by: Keith Packard <keithp@keithp.com> 71790 71791commit 2e6c5f959153cdcea4bcf719621ebec408250f54 71792Author: Daniel Stone <daniel@fooishbar.org> 71793Date: Tue Jul 10 02:03:07 2012 +0100 71794 71795 XFree86: sdksyms: Remove unused -DXorgLoader 71796 71797 We no longer have anything in the tree that checks for XorgLoader. This 71798 was a fairly monumental hack: xvdi.h used to hide all its functions 71799 behind #ifndef XorgLoader, solely to avoid sdksyms.sh picking up its 71800 symbols, as it was previously a module rather than built-in. 71801 71802 This is no longer the case, so we can remove the define. 71803 71804 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71805 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71806 Signed-off-by: Keith Packard <keithp@keithp.com> 71807 71808commit 7a11b817e770cd2196814f1ac6264c6d87b76c6a 71809Author: Tomas Carnecky <tom@dbservice.com> 71810Date: Tue Jul 10 02:03:06 2012 +0100 71811 71812 Move Xv and XvMC from extmod to built-in 71813 71814 Always build these extensions into the core server, rather than letting 71815 them languish in extmod. 71816 71817 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 71818 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71819 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71820 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71821 Signed-off-by: Keith Packard <keithp@keithp.com> 71822 71823commit 7d859bd87834dd79c7fa3792075496ece698c082 71824Author: Tomas Carnecky <tom@dbservice.com> 71825Date: Tue Jul 10 02:03:05 2012 +0100 71826 71827 Move XRes from extmod to built-in 71828 71829 Always build XRes support into the core server, rather than letting it 71830 languish in extmod. 71831 71832 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 71833 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71834 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71835 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71836 Signed-off-by: Keith Packard <keithp@keithp.com> 71837 71838commit ba21fc29587e4f11320b88a70433beb1fe22cdf3 71839Author: Tomas Carnecky <tom@dbservice.com> 71840Date: Tue Jul 10 02:03:04 2012 +0100 71841 71842 Move DPMS from extmod to built-in 71843 71844 Always build DPMS support into the core server, rather than letting it 71845 languish in extmod. 71846 71847 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 71848 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71849 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71850 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71851 Signed-off-by: Keith Packard <keithp@keithp.com> 71852 71853commit 3ed2c6e11298c4299042cfe4578f6cc02e441d58 71854Author: Tomas Carnecky <tom@dbservice.com> 71855Date: Tue Jul 10 02:03:03 2012 +0100 71856 71857 Move MIT-SCREEN-SAVER from extmod to built-in 71858 71859 If we've built MIT-SCREEN-SAVER support, then just build it into the 71860 main binary, rather than leaving it in extmod. 71861 71862 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 71863 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71864 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71865 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71866 Signed-off-by: Keith Packard <keithp@keithp.com> 71867 71868commit b8c9ab0fea0be299d9f48ac0c4ed0474104a7556 71869Author: Tomas Carnecky <tom@dbservice.com> 71870Date: Tue Jul 10 02:03:02 2012 +0100 71871 71872 Move RECORD from external module to built-in 71873 71874 Rather than languishing in its own special module, move RECORD into the 71875 core server. 71876 71877 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 71878 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71879 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71880 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71881 Signed-off-by: Keith Packard <keithp@keithp.com> 71882 71883commit bf61bf69b2b4f9e9993ca771de100079c7696bea 71884Author: Tomas Carnecky <tom@dbservice.com> 71885Date: Tue Jul 10 02:03:01 2012 +0100 71886 71887 Move DBE from an external module to built-in 71888 71889 If DBE support is compiled in the server, just man up and build it into 71890 the server, rather than having it as an external module. 71891 71892 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 71893 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 71894 Reviewed-by: Jamey Sharp <jamey@minilop.net> 71895 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71896 Signed-off-by: Keith Packard <keithp@keithp.com> 71897 71898commit d785368e0e171080fea5447c87e81c93f515ff21 71899Author: Daniel Stone <daniel@fooishbar.org> 71900Date: Tue Jul 10 02:03:00 2012 +0100 71901 71902 miinitext: Use ARRAY_SIZE rather than sentinel 71903 71904 NULL sentinels are totally lame. 71905 71906 v2: numExtensionModules isn't public, so keep using the provided 71907 sentinal - Keith Packard <keithp@keithp.com> 71908 71909 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71910 Reviewed-by: Stéphane Marchesin <stephane.marchesin@gmail.com> 71911 Signed-off-by: Keith Packard <keithp@keithp.com> 71912 71913commit f48d8f58b39af0036365368d6ac35319d6d856d5 71914Author: Daniel Stone <daniel@fooishbar.org> 71915Date: Tue Jul 10 02:02:59 2012 +0100 71916 71917 extmod: Use ARRAY_SIZE rather than sentinel 71918 71919 When the array gets down to size zero (which it does in later patches), 71920 gcc complains that the index is out of bounds. Avoid this by using 71921 ARRAY_SIZE on extensionModules instead. 71922 71923 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71924 Reviewed-by: Stéphane Marchesin <stephane.marchesin@gmail.com> 71925 Signed-off-by: Keith Packard <keithp@keithp.com> 71926 71927commit d35884da2f5a250ff6dd5131dc23ac629eccdd0c 71928Author: Daniel Stone <daniel@fooishbar.org> 71929Date: Tue Jul 10 02:02:58 2012 +0100 71930 71931 Add xf86ExtensionInit for DDX extension configuration 71932 71933 xf86ExtensionInit is called after configuration file parsing, so it can 71934 perform the two parts of extension initialisation currently done by 71935 extmod: enabling and disabling of extensions through an 'omit' option, 71936 and SELinux configuration. 71937 71938 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71939 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71940 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71941 Signed-off-by: Keith Packard <keithp@keithp.com> 71942 71943commit 4170e4e1f8e879394c896c490d9f08a38ecfdefd 71944Author: Daniel Stone <daniel@fooishbar.org> 71945Date: Tue Jul 10 02:02:57 2012 +0100 71946 71947 Loader: Drop EXTERN_MODULE flag 71948 71949 EXTERN_MODULE was used to specify that we shouldn't worry about modules 71950 lacking a ModuleData object. It was also completely unused. *shrug* 71951 71952 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71953 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71954 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 71955 Signed-off-by: Keith Packard <keithp@keithp.com> 71956 71957commit a1d41e311c21eb6627caa0d168e070ceaf90806f 71958Author: Daniel Stone <daniel@fooishbar.org> 71959Date: Tue Jul 10 02:02:56 2012 +0100 71960 71961 Move extension initialisation prototypes into extinit.h 71962 71963 Create extinit.h (and xf86Extensions.h, for Xorg-specific extensions) to 71964 hold all our extension initialisation prototypes, rather than 71965 duplicating them everywhere. 71966 71967 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71968 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 71969 Signed-off-by: Keith Packard <keithp@keithp.com> 71970 71971commit b86aa74cafddf9bca5addfeb458a23f46ddcf132 71972Author: Daniel Stone <daniel@fooishbar.org> 71973Date: Tue Jul 10 02:02:55 2012 +0100 71974 71975 GLX: Insert swrast provider from GlxExtensionInit 71976 71977 Rather than making poor old miinitext.c do it, including making DMX 71978 have fake symbols just to keep it happy. 71979 71980 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71981 Reviewed-by: Stéphane Marchesin <stephane.marchesin@gmail.com> 71982 Signed-off-by: Keith Packard <keithp@keithp.com> 71983 71984commit 13552a5c4701097a1507a99ed8366ab1438844b6 71985Author: Daniel Stone <daniel@fooishbar.org> 71986Date: Tue Jul 10 02:02:54 2012 +0100 71987 71988 Reorder extension initialisation for non-Xorg 71989 71990 Reorder static extension initialisation in miinitext for non-Xorg 71991 servers to match Xorg's order. 71992 71993 Tested with Xephyr; checked that the extension list was identical before 71994 and after. 71995 71996 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 71997 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 71998 Signed-off-by: Keith Packard <keithp@keithp.com> 71999 72000commit 5079db78aea8fa42e781dd876c1ee1c31571b3ae 72001Author: Tomas Carnecky <tom@dbservice.com> 72002Date: Tue Jul 10 02:02:53 2012 +0100 72003 72004 Replace INITARGS with void 72005 72006 INITARGS was a hardcoded define to void. Since knowing the function 72007 signature for your extensions is kinda useful, just replace it with a 72008 hardcoded void, but leave the define there for API compatibility. 72009 72010 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 72011 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 72012 Reviewed-by: Cyril Brulebois <kibi@debian.org> 72013 Reviewed-by: Jamey Sharp <jamey@minilop.net> 72014 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 72015 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 72016 Signed-off-by: Keith Packard <keithp@keithp.com> 72017 72018commit bddb8c6cbe52ba9923b1d36d01f5ac9391b0ec0e 72019Author: Daniel Stone <daniel@fooishbar.org> 72020Date: Tue Jul 10 02:02:52 2012 +0100 72021 72022 Xinerama: Fix ExtensionInit prototype 72023 72024 Huh, so I guess INITARGS used to be int argc, char *argv then. Either 72025 way, it's now void, so fix that ... 72026 72027 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 72028 Reviewed-by: Cyril Brulebois <kibi@debian.org> 72029 Reviewed-by: Jamey Sharp <jamey@minilop.net> 72030 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 72031 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 72032 Signed-off-by: Keith Packard <keithp@keithp.com> 72033 72034commit eb9661fef9f59587f898371f97a0952ac14d125f 72035Author: Daniel Stone <daniel@fooishbar.org> 72036Date: Tue Jul 10 02:02:51 2012 +0100 72037 72038 Make extension.h self-contained, remove C++ externs 72039 72040 externsion.h required bits from Xfuncproto.h and dixstruct.h, but 72041 included neither; fix that. 72042 72043 It also had _XFUNCPROTOBEGIN and _XFUNCPROTOEND wrappers, which is a bit 72044 pointless for a server-only library, as it's only needed for C++. 72045 72046 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 72047 Reviewed-by: Cyril Brulebois <kibi@debian.org> 72048 Reviewed-by: Jamey Sharp <jamey@minilop.net> 72049 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 72050 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 72051 Signed-off-by: Keith Packard <keithp@keithp.com> 72052 72053commit 854c1fa4a1e90288d1e2f3777fac3e1a9acb1235 72054Author: Daniel Stone <daniel@fooishbar.org> 72055Date: Tue Jul 10 02:02:50 2012 +0100 72056 72057 Add a common ARRAY_SIZE macro to dix.h 72058 72059 Does what it says on the box, replacing those from Xi/ and glx/. 72060 72061 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 72062 Reviewed-by: Cyril Brulebois <kibi@debian.org> 72063 Reviewed-by: Jamey Sharp <jamey@minilop.net> 72064 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 72065 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 72066 Signed-off-by: Keith Packard <keithp@keithp.com> 72067 72068commit 67953d6975ef2ee573b95e7641aaf3d72e9f8379 72069Author: Daniel Stone <daniel@fooishbar.org> 72070Date: Tue Jul 10 02:02:49 2012 +0100 72071 72072 Xorg: Link XKB DDX library after core server libs 72073 72074 libxorgxkb.a contains a number of libraries which are used by XKB action 72075 code to call back into the DDX, e.g. for VT switching, termination, grab 72076 breaking, et al. Make sure libxkb.a comes first in the link order, so 72077 it can mark XkbDDX* as used in order for the linker to not discard them. 72078 72079 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 72080 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 72081 Signed-off-by: Keith Packard <keithp@keithp.com> 72082 72083commit 656af2c7e73d948338683675e14cd827b46a97c4 72084Author: Daniel Stone <daniel@fooishbar.org> 72085Date: Tue Jul 10 02:02:46 2012 +0100 72086 72087 Don't make failure to -nolisten fatal 72088 72089 If failing to disable a protocol specified by -nolisten failed, we'd 72090 throw a FatalError and bomb startup entirely. From poking at xtrans, it 72091 looks like the only way we can get a failure here is because we've 72092 specified a protocol name which doesn't exist, which probably doesn't 72093 constitute a security risk. 72094 72095 And it makes it possible to start gdm even though you've built with 72096 --disable-tcp-transport. 72097 72098 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 72099 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72100 Signed-off-by: Keith Packard <keithp@keithp.com> 72101 72102commit dc0c2dd92d5eb314582ea10830678753824375d9 72103Author: Daniel Stone <daniel@fooishbar.org> 72104Date: Tue Jul 10 02:02:45 2012 +0100 72105 72106 Xext: Add hashtable.h to sources to fix distcheck 72107 72108 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 72109 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72110 Signed-off-by: Keith Packard <keithp@keithp.com> 72111 72112commit ad4092cf7d59a89b1b2922440eef65be5c0c5ebd 72113Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72114Date: Mon Jul 9 19:12:44 2012 -0700 72115 72116 Replace padlength tables with inline functions from misc.h 72117 72118 Adds new function padding_for_int32() and uses existing pad_to_int32() 72119 depending on required results. 72120 72121 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72122 Reviewed-by: Keith Packard <keithp@keithp.com> 72123 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 72124 Tested-by: Daniel Stone <daniel@fooishbar.org> 72125 72126commit 2b1c1300cc23912ee1c59f8dde938dd4d7287f4a 72127Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72128Date: Mon Jul 9 19:12:44 2012 -0700 72129 72130 ephyrGLXQueryServerString: Stop making an unused copy of server_string 72131 72132 ephyrGLXQueryServerString() carefully allocated a buffer padded to the 72133 word-aligned string length for sending to the client, copied the string 72134 to it, and then forgot to use it, potentially reading a few bytes of 72135 garbage past the end of the server_string buffer. 72136 72137 Since WriteToClient already handles the necessary padding, just send 72138 it the actual length of the original server_string, and don't bother 72139 making a padded copy. 72140 72141 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72142 Reviewed-by: Keith Packard <keithp@keithp.com> 72143 Tested-by: Daniel Stone <daniel@fooishbar.org> 72144 72145commit 7a29f6878284e2d65e69fcd157aa9ec01d21b3c0 72146Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72147Date: Mon Jul 9 19:12:44 2012 -0700 72148 72149 Initialize padding bits to 0 in ErrorConnMax() 72150 72151 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72152 Reviewed-by: Keith Packard <keithp@keithp.com> 72153 Tested-by: Daniel Stone <daniel@fooishbar.org> 72154 72155commit bed610fcae41ddfe21fa9acde599b17d1d15f5d1 72156Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72157Date: Mon Jul 9 19:12:44 2012 -0700 72158 72159 Set padding bytes to 0 in WriteToClient 72160 72161 Clear them out when needed instead of leaving whatever values were 72162 present in previously sent messages. 72163 72164 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72165 Reviewed-by: Keith Packard <keithp@keithp.com> 72166 Tested-by: Daniel Stone <daniel@fooishbar.org> 72167 72168commit 1622dd8ab24d2fe7b9908110eb237774d4e03568 72169Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72170Date: Mon Jul 9 19:12:44 2012 -0700 72171 72172 Use C99 designated initializers in dix registry 72173 72174 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72175 Reviewed-by: Keith Packard <keithp@keithp.com> 72176 Tested-by: Daniel Stone <daniel@fooishbar.org> 72177 72178commit 9805cedf7b0f76d3b75f94e956c4cc2dcf0d8b64 72179Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72180Date: Mon Jul 9 19:12:44 2012 -0700 72181 72182 Use C99 designated initializers in extension Events 72183 72184 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72185 Reviewed-by: Keith Packard <keithp@keithp.com> 72186 Tested-by: Daniel Stone <daniel@fooishbar.org> 72187 72188commit 0af79b124e1317c36d1613d28755c5a8ce612e2a 72189Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72190Date: Mon Jul 9 19:12:44 2012 -0700 72191 72192 Use C99 designated initializers in dix Events 72193 72194 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72195 Reviewed-by: Keith Packard <keithp@keithp.com> 72196 Tested-by: Daniel Stone <daniel@fooishbar.org> 72197 72198commit 483266a583b0fd8a552303f318a71502ce3ff1d2 72199Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72200Date: Mon Jul 9 19:12:43 2012 -0700 72201 72202 Use C99 designated initializers in xf86 extension Replies 72203 72204 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72205 Reviewed-by: Keith Packard <keithp@keithp.com> 72206 Tested-by: Daniel Stone <daniel@fooishbar.org> 72207 72208commit 14501fd33ee89acf9fc1d908003ed62f383d8bae 72209Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72210Date: Mon Jul 9 19:12:43 2012 -0700 72211 72212 Use C99 designated initializers in Xephyr Replies 72213 72214 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72215 Reviewed-by: Keith Packard <keithp@keithp.com> 72216 Tested-by: Daniel Stone <daniel@fooishbar.org> 72217 72218commit e4e827ec36fd20b7aea608db09790f76fb87e519 72219Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72220Date: Mon Jul 9 19:12:43 2012 -0700 72221 72222 Use C99 designated initializers in dmx Replies 72223 72224 v2: fix in __glXGetVisualConfigs to not re-declare local 'reply' variable 72225 72226 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72227 Reviewed-by: Keith Packard <keithp@keithp.com> 72228 Tested-by: Daniel Stone <daniel@fooishbar.org> 72229 72230 Fixup for Use C99 designated initializeres in dmx Replies 72231 72232 Signed-off-by: Keith Packard <keithp@keithp.com> 72233 72234commit bd6f948c41865c2c9d3fba1000bf5f7458d3afc1 72235Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72236Date: Mon Jul 9 19:12:43 2012 -0700 72237 72238 Use C99 designated initializers in randr Replies 72239 72240 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72241 Reviewed-by: Keith Packard <keithp@keithp.com> 72242 Tested-by: Daniel Stone <daniel@fooishbar.org> 72243 72244commit a406bd07593edb69285cf2fd91a6af4a5d956817 72245Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72246Date: Mon Jul 9 19:12:43 2012 -0700 72247 72248 Use C99 designated initializers in xkb Replies 72249 72250 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72251 Reviewed-by: Keith Packard <keithp@keithp.com> 72252 Tested-by: Daniel Stone <daniel@fooishbar.org> 72253 72254commit 26efa09d0c7ee2f6a18547205c84be75376e688b 72255Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72256Date: Mon Jul 9 19:12:43 2012 -0700 72257 72258 Use C99 designated initializers in glx Replies 72259 72260 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72261 Reviewed-by: Keith Packard <keithp@keithp.com> 72262 Tested-by: Daniel Stone <daniel@fooishbar.org> 72263 72264commit cc5f09c86f7bea23b7546c3491b2c52ce8100a71 72265Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72266Date: Mon Jul 9 19:12:43 2012 -0700 72267 72268 Use C99 designated initializers in various extension Replies 72269 72270 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72271 Reviewed-by: Keith Packard <keithp@keithp.com> 72272 Tested-by: Daniel Stone <daniel@fooishbar.org> 72273 72274commit 2f5caeaddb3616dc9ff57d784f7feba589c536e7 72275Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72276Date: Mon Jul 9 19:12:43 2012 -0700 72277 72278 Use C99 designated initializers in Xinput Replies 72279 72280 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72281 Reviewed-by: Keith Packard <keithp@keithp.com> 72282 Tested-by: Daniel Stone <daniel@fooishbar.org> 72283 72284commit 6a721e3af5e11bae1d59ca551a02470f35d9e844 72285Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72286Date: Mon Jul 9 19:12:43 2012 -0700 72287 72288 Use C99 designated initializers in Xext Replies 72289 72290 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72291 Reviewed-by: Keith Packard <keithp@keithp.com> 72292 Tested-by: Daniel Stone <daniel@fooishbar.org> 72293 72294commit d792ac125a0462a04a930af543cbc732f8cdab7d 72295Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72296Date: Mon Jul 9 19:12:43 2012 -0700 72297 72298 Use C99 designated initializers in dix Replies 72299 72300 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72301 Reviewed-by: Keith Packard <keithp@keithp.com> 72302 Tested-by: Daniel Stone <daniel@fooishbar.org> 72303 72304commit 69fa5630b5902aaad267fc67d0da4ca93625886a 72305Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72306Date: Mon Jul 9 19:12:43 2012 -0700 72307 72308 Use C99 designated initializers in SendErrorToClient 72309 72310 Let the compiler worry about 0-filling the rest of the fields, 72311 instead of memsetting the whole struct and then going back to 72312 overwrite some of the fields. 72313 72314 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72315 Reviewed-by: Keith Packard <keithp@keithp.com> 72316 Tested-by: Daniel Stone <daniel@fooishbar.org> 72317 72318commit cdf5bcd420e5bcf4a4a24a275d3133a4e16ce41e 72319Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72320Date: Mon Jul 9 19:12:42 2012 -0700 72321 72322 Use calloc to zero fill buffers being allocated for replies & events 72323 72324 Ensures padding bytes are zero-filled 72325 72326 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72327 Reviewed-by: Keith Packard <keithp@keithp.com> 72328 Tested-by: Daniel Stone <daniel@fooishbar.org> 72329 72330commit ef0f701c9245b0373eb0c7a84dd804af4738bb29 72331Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72332Date: Mon Jul 9 19:12:42 2012 -0700 72333 72334 xf86dga2.c & xf86vmode.c: Move REQUEST_SIZE_MATCH checks before using stuff 72335 72336 Seems silly waiting to check if the client failed to send us enough bytes 72337 until after we've already tried using them. 72338 72339 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72340 Reviewed-by: Keith Packard <keithp@keithp.com> 72341 Tested-by: Daniel Stone <daniel@fooishbar.org> 72342 72343commit 2e739a8870d91fbac7f536ffa8f6d2cee0f66cec 72344Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72345Date: Mon Jul 9 19:12:42 2012 -0700 72346 72347 ProcRRListOutputProperties: skip atom walk if the list is empty 72348 72349 pAtoms is only allocated if numProps was non-zero, so move the walk 72350 through the property list to copy atoms to it inside the if (numProps) 72351 72352 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72353 Reviewed-by: Keith Packard <keithp@keithp.com> 72354 Tested-by: Daniel Stone <daniel@fooishbar.org> 72355 72356commit 15bc13c8d088e05f14c7262348e0066929c29251 72357Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72358Date: Mon Jul 9 19:12:42 2012 -0700 72359 72360 ProcRRGetScreenInfo: swap configTimestamp as well 72361 72362 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72363 Reviewed-by: Keith Packard <keithp@keithp.com> 72364 Tested-by: Daniel Stone <daniel@fooishbar.org> 72365 72366commit 3a013b8816bc0bb765c99e8a2bf2ea536dc62c3b 72367Author: Jaroslav Šmíd <jardasmid@gmail.com> 72368Date: Mon Jul 9 19:12:42 2012 -0700 72369 72370 Bug 51375: Xorg doesn't set status for RRGetOutputInfo 72371 72372 https://bugs.freedesktop.org/show_bug.cgi?id=51375 72373 https://bugs.freedesktop.org/attachment.cgi?id=63397 72374 72375 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72376 Reviewed-by: Keith Packard <keithp@keithp.com> 72377 Tested-by: Daniel Stone <daniel@fooishbar.org> 72378 72379commit 5b86c072d1d586ce040d8831a05cf97ff8b17821 72380Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72381Date: Mon Jul 9 19:12:42 2012 -0700 72382 72383 Use temporary variables instead of parts of reply structures 72384 72385 When passing variable pointers to functions or otherwise doing long 72386 sequences to compute values for replies, create & use some new 72387 temporary variables, to allow for simpler initialization of reply 72388 structures in the following patches. 72389 72390 Move memsets & other initializations to group with the rest of the 72391 filling in of the reply structure, now that they're not needed so 72392 early in the code path. 72393 72394 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72395 Reviewed-by: Keith Packard <keithp@keithp.com> 72396 Tested-by: Daniel Stone <daniel@fooishbar.org> 72397 72398commit c2fb1a7b2ab58d70b38ee03ab2fdeb4e7183a356 72399Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72400Date: Mon Jul 9 19:12:42 2012 -0700 72401 72402 ProcQueryKeymap: rework logic around permission to copy key states 72403 72404 Always initialize to zero, and then if permission is granted, copy 72405 the current key state maps, instead of always copying and then 72406 zeroing out if permission was denied. 72407 72408 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72409 Reviewed-by: Keith Packard <keithp@keithp.com> 72410 Tested-by: Daniel Stone <daniel@fooishbar.org> 72411 72412commit dccb0858d71341e1258800b37d1bb78c0115838d 72413Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72414Date: Mon Jul 9 19:12:42 2012 -0700 72415 72416 Core events: invert check for permission to copy key states 72417 72418 Always initialize to zero, and then if permission is granted, copy 72419 the current key state maps. Use memcpy instead of memmove for the 72420 copy, since we're always copying to a newly allocated event on the 72421 stack, so guaranteed not to overlap with the device map structure. 72422 72423 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72424 Reviewed-by: Keith Packard <keithp@keithp.com> 72425 Tested-by: Daniel Stone <daniel@fooishbar.org> 72426 72427commit db69212df8a0bf09140368356d2d430c54afe346 72428Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72429Date: Mon Jul 9 19:12:42 2012 -0700 72430 72431 Rework reply initialization in ProcGetProperty & NullPropertyReply 72432 72433 Don't need to pass an empty reply to NullPropertyReply, let it make 72434 it's own. Move reply initialization code in remaining replies in 72435 ProcGetProperty to group with the rest of the fields. (Prepares 72436 for coming C99 designated initializer conversion.) 72437 72438 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72439 Reviewed-by: Keith Packard <keithp@keithp.com> 72440 Tested-by: Daniel Stone <daniel@fooishbar.org> 72441 72442commit 6be74a9080e30fc502421cd438cd0c73fb8eb0b0 72443Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72444Date: Mon Jul 9 19:12:42 2012 -0700 72445 72446 Fix more poorly indented/wrapped comments & code 72447 72448 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72449 Reviewed-by: Keith Packard <keithp@keithp.com> 72450 Tested-by: Daniel Stone <daniel@fooishbar.org> 72451 72452commit 789d64e19a3b3d98b88bc80f677e0c37bfb5c631 72453Author: Alan Coopersmith <alan.coopersmith@oracle.com> 72454Date: Sun May 13 00:03:35 2012 -0700 72455 72456 Remove unneccesary casts from WriteToClient calls 72457 72458 Casting return to (void) was used to tell lint that you intended 72459 to ignore the return value, so it didn't warn you about it. 72460 72461 Casting the third argument to (char *) was used as the most generic 72462 pointer type in the days before compilers supported C89 (void *) 72463 (except for a couple places it's used for byte-sized pointer math). 72464 72465 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 72466 Reviewed-by: Keith Packard <keithp@keithp.com> 72467 Tested-by: Daniel Stone <daniel@fooishbar.org> 72468 72469commit 023127915e6922bc53e4c768de760d8a4f25c07c 72470Author: Keith Packard <keithp@keithp.com> 72471Date: Mon Jul 9 16:34:39 2012 -0700 72472 72473 Reliably reset signals at server init time 72474 72475 Each DDX currently calls OsReleaseSIGIO in case it was suspended when 72476 the server regen started. This causes a BUG to occur if SIGIO was 72477 *not* blocked at that time. Instead of relying on each DDX, make the 72478 OS layer reliably reset all signal state at server init time, ensuring 72479 that signals are suitably unblocked and that the various signal state 72480 counting variables are set back to zero. 72481 72482 Signed-off-by: Keith Packard <keithp@keithp.com> 72483 72484commit d6756e0298e660a0aca58d75bcf79471ce07a634 72485Author: Keith Packard <keithp@keithp.com> 72486Date: Mon Jul 9 16:33:11 2012 -0700 72487 72488 xfree86: In InitOutput, only call OsReleaseSIGIO if OsBlockSIGIO was called 72489 72490 Otherwise, OsReleaseSIGIO will complain, or perhaps something worse 72491 will happen (if SIGIO actually needs to be blocked here). 72492 72493 Signed-off-by: Keith Packard <keithp@keithp.com> 72494 72495commit ea8b04507e5464e4817791db516a20cfed2a6724 72496Author: Keith Packard <keithp@keithp.com> 72497Date: Mon Jul 9 16:30:24 2012 -0700 72498 72499 privates: Resize GPU screen-specific privates too 72500 72501 When allocating new global privates, make sure the gpu 72502 screens get their private offsets updated. 72503 72504 This only affects GPU screens that enumerate before the non-GPU 72505 screens, which generally requires that the related device be present 72506 when the system boots so that it can get an earlier DRM filename. 72507 72508 Signed-off-by: Keith Packard <keithp@keithp.com> 72509 72510commit d1c639c006526b8cab14dac582508f3f54848967 72511Author: Keith Packard <keithp@keithp.com> 72512Date: Mon Jul 9 16:29:04 2012 -0700 72513 72514 Add 'install-headers' target in the top-level Makefile 72515 72516 This target recursively locates directories with sdk headers and 72517 installs them all. Useful when you want to build a complete new X 72518 server and drivers without having to install the X server before the 72519 drivers are actually working. 72520 72521 Signed-off-by: Keith Packard <keithp@keithp.com> 72522 72523commit 329db3292223cccd4887062956622285c45a1523 72524Author: Dave Airlie <airlied@redhat.com> 72525Date: Sat Jul 7 10:10:08 2012 +0100 72526 72527 dri2/prime: allocate prime id at screen allocation time 72528 72529 Add a static mask of prime id and allocate them at screen time, 72530 if the driver supports the prime interfaces and is a gpu screen. 72531 72532 This is instead of them changing due to user controlled randr commands, 72533 as suggested by Keith. 72534 72535 Reviewed-by: Keith Packard <keithp@keithp.com> 72536 Signed-off-by: Dave Airlie <airlied@redhat.com> 72537 72538commit 30298012162de7f76e8a4c7b0362e98703f80011 72539Author: Dave Airlie <airlied@redhat.com> 72540Date: Tue Jun 26 10:20:52 2012 +0100 72541 72542 dri2: add initial prime support. (v1.2) 72543 72544 This adds the initial prime support for dri2 offload. The main thing is 72545 when we get a connection from a prime client, we stored the information 72546 and mark all drawables from that client as prime. We then create all 72547 buffers for that drawable on the prime device dri2screen. 72548 72549 Then DRI2UpdatePrime is provided which drivers can call to get a shared 72550 pixmap which they can use as the front buffer. The driver is then 72551 responsible for doing the back->front copy to the shared buffer. 72552 72553 prime requires a compositing manager be run, but it handles the case where 72554 a window get un-redirected by allocating a new pixmap and pointing the crtc 72555 at it while the client is in that state. 72556 72557 Currently prime can't handle pageflipping, so always does straight copy swap, 72558 72559 v1.1: renumber on top of master. 72560 v1.2: fix auth on top of master. 72561 72562 Reviewed-by: Keith Packard <keithp@keithp.com> 72563 Signed-off-by: Dave Airlie <airlied@redhat.com> 72564 72565commit 234022cfb3ad2a1b16ab7981ce69f9cd5ba0fbeb 72566Author: Dave Airlie <airlied@redhat.com> 72567Date: Sat Jul 7 10:20:58 2012 +0100 72568 72569 configure: bump dri2proto to 2.8 72570 72571 This is required for the prime macros. 72572 72573 Reviewed-by: Keith Packard <keithp@keithp.com> 72574 Signed-off-by: Dave Airlie <airlied@redhat.com> 72575 72576commit c41922940adbc8891575b3321fadf01ff4cb5854 72577Author: Dave Airlie <airlied@redhat.com> 72578Date: Tue Jun 26 10:00:21 2012 +0100 72579 72580 dix/randr: add a hook into screen to replace scanout pixmap 72581 72582 For DRI2 in some offload cases we need to set a new pixmap on the crtc, 72583 this hook allows dri2 to call into randr to do the necessary work to set 72584 a pixmap as the scanout pixmap for the crtc the drawable is currently on. 72585 72586 This is really only to be used for unredirected full screen apps in composited 72587 environments. 72588 72589 Reviewed-by: Keith Packard <keithp@keithp.com> 72590 Signed-off-by: Dave Airlie <airlied@redhat.com> 72591 72592commit e2fd447e76c4a1233374c9d8fa9cae54a55dad50 72593Author: Dave Airlie <airlied@redhat.com> 72594Date: Thu Jul 5 15:50:56 2012 +0100 72595 72596 xf86: add callback for offloak sink setting support. 72597 72598 This adds support for setting the offload sink to the xf86 ddx. 72599 72600 Reviewed-by: Keith Packard <keithp@keithp.com> 72601 Signed-off-by: Dave Airlie <airlied@redhat.com> 72602 72603commit 22f02120eb4092e1c283f81c8040f178693ff0a4 72604Author: Dave Airlie <airlied@redhat.com> 72605Date: Tue Jun 26 09:53:54 2012 +0100 72606 72607 xf86: store scanout pixmap in the xf86 struct as well. 72608 72609 This is so we can tell the scanout pixmap has changed between calls 72610 to the crtc set function. 72611 72612 Reviewed-by: Keith Packard <keithp@keithp.com> 72613 Signed-off-by: Dave Airlie <airlied@redhat.com> 72614 72615commit 426bc0a28edbe0e9153f692a02dd25f744ffa034 72616Author: Dave Airlie <airlied@redhat.com> 72617Date: Thu Jul 5 15:41:38 2012 +0100 72618 72619 randr: add hooks for offload sink provider protocol 72620 72621 This adds the protocol handler and associated providers handling 72622 for the offload slaves, it allows two providers to be connected as 72623 offload sink/source. 72624 72625 Reviewed-by: Keith Packard <keithp@keithp.com> 72626 Reviewed-by: Adam Jackson <ajax@redhat.com> 72627 Signed-off-by: Dave Airlie <airlied@redhat.com> 72628 72629commit 12905dfaf01088a00f4a0a78cffba03329e7b724 72630Author: Dave Airlie <airlied@redhat.com> 72631Date: Sat May 26 13:11:01 2012 +0100 72632 72633 dix/xf86: initial offload slave tracking (v1.1) 72634 72635 add the linked list and provider hooks. 72636 72637 v1.1: add another assert in the add path. 72638 72639 Reviewed-by: Keith Packard <keithp@keithp.com> 72640 Reviewed-by: Adam Jackson <ajax@redhat.com> 72641 Signed-off-by: Dave Airlie <airlied@redhat.com> 72642 72643commit a7c01da54ab43f9d29333ccbd79dfc001f9dc5e3 72644Author: Dave Airlie <airlied@redhat.com> 72645Date: Thu Jul 5 20:34:54 2012 +0100 72646 72647 xf86: make sure rotate calcs are done on the right screen boundaries 72648 72649 This fixes a segfault where this code believes we are outside the screen 72650 boundaries on a slave device, but we aren't. 72651 72652 Reviewed-by: Keith Packard <keithp@keithp.com> 72653 Reviewed-by: Adam Jackson <ajax@redhat.com> 72654 Signed-off-by: Dave Airlie <airlied@redhat.com> 72655 72656commit cc02f4ef3e407be794132d004c947c9a4e14f4c2 72657Author: Dave Airlie <airlied@redhat.com> 72658Date: Tue Jun 19 16:01:16 2012 +0100 72659 72660 xf86/cursor: fallback to sw cursor if we have slaves present. 72661 72662 Current USB devices have no hw rendered cursors, so we need the 72663 master GPU to render the cursor, so whenever we plug in a 72664 slave device, fallback to sw rendered cursors. 72665 72666 Reviewed-by: Adam Jackson <ajax@redhat.com> 72667 Signed-off-by: Dave Airlie <airlied@redhat.com> 72668 72669commit 98686512cbe1affb75222a63136854af79f1411b 72670Author: Dave Airlie <airlied@redhat.com> 72671Date: Thu Jul 5 15:50:24 2012 +0100 72672 72673 xf86: add output source setting callback (v2) 72674 72675 This adds support for the randr callback for setting the output source 72676 for a device. 72677 72678 v2: drop root clip change on detach 72679 72680 Reviewed-by: Keith Packard <keithp@keithp.com> 72681 Reviewed-by: Adam Jackson <ajax@redhat.com> 72682 Signed-off-by: Dave Airlie <airlied@redhat.com> 72683 72684commit 4c92eb00f97f7b8258de8e366226880382cc9ce9 72685Author: Dave Airlie <airlied@redhat.com> 72686Date: Thu Jul 5 15:41:22 2012 +0100 72687 72688 randr: add output source setup 72689 72690 This adds the output sources to the associated list and adds the protocol 72691 handler for the randr SetProviderOutputSource. 72692 72693 Reviewed-by: Keith Packard <keithp@keithp.com> 72694 Reviewed-by: Adam Jackson <ajax@redhat.com> 72695 Signed-off-by: Dave Airlie <airlied@redhat.com> 72696 72697commit bec4cb72c55bb6dee09c65c0844af201067a090f 72698Author: Dave Airlie <airlied@redhat.com> 72699Date: Thu May 24 15:47:22 2012 +0100 72700 72701 randr: hook up output slave to screen resources return 72702 72703 When the client asks for the screen resources list, it will now 72704 get a list of crtc/outputs for the master + all attached slaves, 72705 this will let randr configure all attached slave devices properly. 72706 72707 Keith asked I merge the two functions, but not just yet, the current 72708 multi screen code doesn't handle primary yet properly, will fix it up later. 72709 72710 Reviewed-by: Keith Packard <keithp@keithp.com> 72711 Reviewed-by: Adam Jackson <ajax@redhat.com> 72712 Signed-off-by: Dave Airlie <airlied@redhat.com> 72713 72714commit 88bc02bfaae2c15a30c606d34cf7940e6ad5ea14 72715Author: Dave Airlie <airlied@redhat.com> 72716Date: Mon Jun 18 14:11:24 2012 +0100 72717 72718 xf86/crtc: add new interface to detach slave crtcs 72719 72720 This will detach any scanout pixmaps attached to slave crtcs. 72721 72722 Reviewed-by: Keith Packard <keithp@keithp.com> 72723 Reviewed-by: Adam Jackson <ajax@redhat.com> 72724 Signed-off-by: Dave Airlie <airlied@redhat.com> 72725 72726commit 9b8217f9ef6279fff6628631d18497bed0343ef9 72727Author: Dave Airlie <airlied@redhat.com> 72728Date: Thu Jun 14 15:24:46 2012 +0100 72729 72730 dix/pixmap: track dirty pixmaps in server. (v4) 72731 72732 This adds two functions for drivers to use directly to keep a 72733 linked list of slave pixmaps to do damage tracking on and keep 72734 updated. It also adds a helper function that drivers may optionally 72735 call to do a simple copy area damage update. 72736 72737 v2: use damage.h not damagestr.h, fixes ephyr build. 72738 72739 v3: address ajax review: use slave_dst, drop unused dst member. 72740 72741 v4: check DamageCreate return, add SourceValidate comment, 72742 add a comment addressing possible optimisation possibility 72743 72744 Reviewed-by: Keith Packard <keithp@keithp.com> 72745 Reviewed-by: Adam Jackson <ajax@redhat.com> 72746 Signed-off-by: Dave Airlie <airlied@redhat.com> 72747 72748commit c5cc2a8243c5e6bf454af989b7512ec7e20fc3f2 72749Author: Dave Airlie <airlied@redhat.com> 72750Date: Tue May 29 14:41:39 2012 +0100 72751 72752 randr: check the screen size bounds against the master 72753 72754 The master contains the complete screen size bounds, so check 72755 the width/height against it. 72756 72757 Reviewed-by: Keith Packard <keithp@keithp.com> 72758 Reviewed-by: Adam Jackson <ajax@redhat.com> 72759 Signed-off-by: Dave Airlie <airlied@redhat.com> 72760 72761commit 26d848cb916c8a85f2c69d2aeb004665ba07bf00 72762Author: Dave Airlie <airlied@redhat.com> 72763Date: Tue May 29 14:35:24 2012 +0100 72764 72765 randr: fix tellchanged to work with slave scanouts 72766 72767 Tell changed need to tell only for the master pixmap, 72768 however it gets called from various places for slave screens, 72769 so convert to telling clients about changes on the master screen. 72770 72771 Reviewed-by: Keith Packard <keithp@keithp.com> 72772 Reviewed-by: Adam Jackson <ajax@redhat.com> 72773 Signed-off-by: Dave Airlie <airlied@redhat.com> 72774 72775commit 4bf0192d810e01c89a1903cc4bc5e639fc13a547 72776Author: Dave Airlie <airlied@redhat.com> 72777Date: Tue May 29 14:37:11 2012 +0100 72778 72779 randr: fixup constrain to work with slave screens. 72780 72781 Current code constrains the cursor to the crtcs on the master 72782 device, for slave outputs to work we have to include their crtcs 72783 in the constrain calculations. 72784 72785 Reviewed-by: Keith Packard <keithp@keithp.com> 72786 Reviewed-by: Adam Jackson <ajax@redhat.com> 72787 Signed-off-by: Dave Airlie <airlied@redhat.com> 72788 72789commit 2ed3f64d9d453628ab5c03b8aee006093019c933 72790Author: Dave Airlie <airlied@redhat.com> 72791Date: Wed Jun 20 12:37:26 2012 +0100 72792 72793 xf86: add initial scanout pixmap support (v2) 72794 72795 Add the simple passthrough interface for drivers to use, 72796 so they can set scanout pixmaps. 72797 72798 v2: detach scanout pixmap properly. 72799 72800 Reviewed-by: Keith Packard <keithp@keithp.com> 72801 Reviewed-by: Adam Jackson <ajax@redhat.com> 72802 Signed-off-by: Dave Airlie <airlied@redhat.com> 72803 72804commit f2da2c12042d3447b1c954e7301632d174a08b4f 72805Author: Dave Airlie <airlied@redhat.com> 72806Date: Wed Jun 20 12:37:12 2012 +0100 72807 72808 randr: add initial scanout pixmap support (v3) 72809 72810 When randr notices a crtc configuration request for a slave device, 72811 it checks if the slave allocated pixmap exists and is suitable, 72812 if not it allocates a new shared pixmap from the master, shares 72813 it to the slave, and starts the master tracking damage to it, 72814 to keep it updated from the current front pixmap. 72815 72816 If the resize means the crtc is no longer used it will destroy 72817 the slave pixmap. 72818 72819 This adds the concept of a scanout_pixmap to the randr_crtc object, 72820 and also adds a master pixmap pointer to the pixmap object, along 72821 with defining some pixmap helper functions for getting pixmap box/regions. 72822 72823 v2: split out pixmap sharing to a separate function. 72824 72825 v3: update for void * 72826 72827 Reviewed-by: Keith Packard <keithp@keithp.com> 72828 Reviewed-by: Adam Jackson <ajax@redhat.com> 72829 Signed-off-by: Dave Airlie <airlied@redhat.com> 72830 72831commit 0b0e7148925e1f835d4f4f06e402a97023531828 72832Author: Dave Airlie <airlied@redhat.com> 72833Date: Thu May 24 16:03:31 2012 +0100 72834 72835 exa: pixmap sharing infrastructure (v3) 72836 72837 This just adds exa interfaces for mixed exa so drivers can 72838 share and set shared pixmaps up correctly. 72839 72840 v2: update for passing slave screen. 72841 v3: update for void * 72842 72843 Reviewed-by: Keith Packard <keithp@keithp.com> 72844 Reviewed-by: Adam Jackson <ajax@redhat.com> 72845 Signed-off-by: Dave Airlie <airlied@redhat.com> 72846 72847commit fd6c1bf0a3efced3ad4623dbd4716230a5f2419e 72848Author: Dave Airlie <airlied@redhat.com> 72849Date: Thu May 24 16:03:19 2012 +0100 72850 72851 dix: pixmap sharing infrastructure (v3) 72852 72853 This is a hooks for pixmap sharing and tracking. 72854 72855 The pixmap sharing ones get an integer handle for the pixmap 72856 and use a handle to be the backing for a pixmap. 72857 72858 The tracker interface is to be used when a GPU needs to 72859 track pixmaps to be updated for another GPU. 72860 72861 v2: pass slave to sharing so it can use it to work out driver. 72862 72863 v3: use void * as per keithp's suggestion. 72864 72865 Reviewed-by: Keith Packard <keithp@keithp.com> 72866 Reviewed-by: Adam Jackson <ajax@redhat.com> 72867 Signed-off-by: Dave Airlie <airlied@redhat.com> 72868 72869commit 382dd45bb88fa92764bf0ecd655cb9890eb84960 72870Author: Dave Airlie <airlied@redhat.com> 72871Date: Thu May 24 15:39:05 2012 +0100 72872 72873 dix: add ability to link output slave gpus to the current gpu (v1.1) 72874 72875 Just add the interfaces to attach/detach output slaves, and 72876 a linked list to keep track of them. Hook up the randr providers 72877 list to include these slaves. 72878 72879 v1.1: add another assert to the add path. 72880 72881 Reviewed-by: Keith Packard <keithp@keithp.com> 72882 Reviewed-by: Adam Jackson <ajax@redhat.com> 72883 Signed-off-by: Dave Airlie <airlied@redhat.com> 72884 72885commit 05d2472cd242293a5fa205efff5734a3380072b4 72886Author: Dave Airlie <airlied@redhat.com> 72887Date: Thu Jun 14 15:03:02 2012 +0100 72888 72889 xf86dga: handle DGAAvailable for gpu screens. (v2) 72890 72891 v2: Split out DGAAvailable into two interfaces, one for calls from protocol 72892 decoding and one for internal usage, after discussion with ajax and keithp. 72893 72894 Reviewed-by: Keith Packard <keithp@keithp.com> 72895 Reviewed-by: Adam Jackson <ajax@redhat.com> 72896 Signed-off-by: Dave Airlie <airlied@redhat.com> 72897 72898commit 3cbc4c10b52896324fe14d2ab56bd54577c9294c 72899Author: Dave Airlie <airlied@redhat.com> 72900Date: Wed Jun 13 14:35:49 2012 +0100 72901 72902 randr: expose unattached providers to the clients. 72903 72904 This provides the unattached provider list to the clients. 72905 72906 Reviewed-by: Keith Packard <keithp@keithp.com> 72907 Reviewed-by: Adam Jackson <ajax@redhat.com> 72908 Signed-off-by: Dave Airlie <airlied@redhat.com> 72909 72910commit d0e138c3f3e58ef127dad6c6aef48b1713b5ce58 72911Author: Dave Airlie <airlied@redhat.com> 72912Date: Wed Jun 13 14:32:05 2012 +0100 72913 72914 dix: attach unbound screens to protocol screen 0 (v2) 72915 72916 This is the default attachment, unbound gpu screens get 72917 attached to the 0 protocol screen. 72918 72919 detach on hotunplug. 72920 72921 v2: detach after tearing down crtc/outputs. 72922 72923 Reviewed-by: Keith Packard <keithp@keithp.com> 72924 Reviewed-by: Adam Jackson <ajax@redhat.com> 72925 Signed-off-by: Dave Airlie <airlied@redhat.com> 72926 72927commit 4caad34c932115dc0a9fd9c2995ebb2d38b9359c 72928Author: Dave Airlie <airlied@redhat.com> 72929Date: Wed Jun 13 14:27:42 2012 +0100 72930 72931 dix: add unattached list for attaching screens to initially. (v1.1) 72932 72933 This list is meant for attaching unbound gpu screens to initially, 72934 before the client side rebinds them. 72935 72936 v1.1: add another assert in the add path. 72937 72938 Reviewed-by: Keith Packard <keithp@keithp.com> 72939 Reviewed-by: Adam Jackson <ajax@redhat.com> 72940 Signed-off-by: Dave Airlie <airlied@redhat.com> 72941 72942commit 9b5cf2ed76b00937af12304e0e8dbd60df62067e 72943Author: Dave Airlie <airlied@redhat.com> 72944Date: Wed Jun 13 14:17:06 2012 +0100 72945 72946 xfree86: add framework for provider support in ddx. (v4) 72947 72948 This adds the framework for DDX provider support. 72949 72950 v2: as per keithp's suggestion remove the xf86 provider object 72951 and just store it in the toplevel object. 72952 72953 v3: update for new protocol 72954 72955 v4: drop devPrivate, free name 72956 72957 Reviewed-by: Keith Packard <keithp@keithp.com> 72958 Reviewed-by: Adam Jackson <ajax@redhat.com> 72959 Signed-off-by: Dave Airlie <airlied@redhat.com> 72960 72961commit 66d92afeaeed9f4a19267d95a1f81b9bf27162a5 72962Author: Dave Airlie <airlied@redhat.com> 72963Date: Wed Jun 13 14:03:04 2012 +0100 72964 72965 randr: add provider object and provider property support (v6) 72966 72967 This adds the initial provider object and provider property 72968 support to the randr dix code. 72969 72970 v2: destroy provider in screen close 72971 v2.1: fix whitespace 72972 72973 v3: update for latest rev of protocol + renumber after 1.4 tearout. 72974 72975 v4: fix logic issue, thanks Samsagax on irc 72976 72977 v5: keithp's review: fix current_role, fix copyrights, fix master 72978 reporting crtc/outputs. 72979 72980 v6: port to new randr interface, drop all set role bits for now 72981 72982 v7: drop devPrivate in provider, not needed, add BadMatch returns 72983 for NULL SetProviderOffloadSink and SetProviderOutputSource, drop 72984 the old typedef. 72985 72986 Reviewed-by: Keith Packard <keithp@keithp.com> 72987 Reviewed-by: Adam Jackson <ajax@redhat.com> 72988 Signed-off-by: Dave Airlie <airlied@redhat.com> 72989 72990commit 44eae69f1df2d00e0c9f1ea8d3c4fae06bcacfbc 72991Author: Dave Airlie <airlied@redhat.com> 72992Date: Thu Jul 5 21:11:16 2012 +0100 72993 72994 configure: bump randrproto min requirements to 1.4.0 72995 72996 Needed to build latest randr stuff. 72997 72998 Reviewed-by: Adam Jackson <ajax@redhat.com> 72999 Signed-off-by: Dave Airlie <airlied@redhat.com> 73000 73001commit 4d24192bd24fb20d16d0054d77bffda883faeb79 73002Merge: 7c9d8cbd3 74b786f7c 73003Author: Keith Packard <keithp@keithp.com> 73004Date: Fri Jul 6 12:17:17 2012 -0700 73005 73006 Merge remote-tracking branch 'airlied/for-keithp-stage1' 73007 73008commit 7c9d8cbd368cac53e7aed9308292614337db3be2 73009Author: Torsten Kaiser <just.for.lkml@googlemail.com> 73010Date: Fri Jul 6 12:10:25 2012 -0700 73011 73012 xfree86: EDID Est III parsing skips some modes 73013 73014 This loop needs to count from 7 to 0, not only from 7 to 1. 73015 The current code always skips the modes {1152, 864, 75, 0}, {1280, 1024, 85, 0}, 73016 {1400, 1050, 75, 0}, {1600, 1200, 70, 0} and {1920, 1200, 60, 0}. 73017 73018 Signed-off-by: Torsten Kaiser <x11@ariolc.dyndns.org> 73019 Reviewed-by: Adam Jackson <ajax@redhat.com> 73020 Signed-off-by: Keith Packard <keithp@keithp.com> 73021 73022commit 0b3abacb6430e6593c4358420ba457d8149bd32b 73023Author: Torsten Kaiser <just.for.lkml@googlemail.com> 73024Date: Fri Jul 6 12:06:13 2012 -0700 73025 73026 xfree86: EDID Est III parsing can walk off end of array 73027 73028 Using -O3 gcc notes that m could reach beyound the end of the EstIIIModes array, 73029 if the last bits of the 11s byte where set. 73030 Fix this, by extending the array to cover all possible bits from est. 73031 73032 https://bugs.freedesktop.org/show_bug.cgi?id=45623 73033 73034 Signed-off-by: Torsten Kaiser <x11@ariolc.dyndns.org> 73035 Reviewed-by: Adam Jackson <ajax@redhat.com> 73036 Signed-off-by: Keith Packard <keithp@keithp.com> 73037 73038commit 74b786f7cee5723468bba251a2155a65770b2ad4 73039Author: Dave Airlie <airlied@redhat.com> 73040Date: Tue Jun 5 17:45:03 2012 +0100 73041 73042 xfree86: add autoAddGPU option (v2) 73043 73044 This option is to stop the X server adding non-primary devices as 73045 gpu screens. 73046 73047 v2: fix per Keith's suggestion. 73048 73049 Reviewed-by: Keith Packard <keithp@keithp.com> 73050 Signed-off-by: Dave Airlie <airlied@redhat.com> 73051 73052commit ef6686480ad1de6582b9ef5d13e5d4ffb3b3918c 73053Author: Dave Airlie <airlied@redhat.com> 73054Date: Tue Jun 5 17:30:39 2012 +0100 73055 73056 xfree86: add platform bus hotplug support (v3) 73057 73058 This provides add/remove support for platform devices at xfree86 ddx level. 73059 73060 v2: cleanup properly if no driver found. 73061 73062 v3: load the modesetting driver before checking driver list. 73063 73064 Reviewed-by: Keith Packard <keithp@keithp.com> 73065 Signed-off-by: Dave Airlie <airlied@redhat.com> 73066 73067commit b27cf30995e05da11e827ec41060051673336a7b 73068Author: Dave Airlie <airlied@redhat.com> 73069Date: Tue Jun 5 17:29:56 2012 +0100 73070 73071 xserver/config: add udev/drm hotplug callbacks. (v2) 73072 73073 This adds callbacks into the ddx for udev gpu hotplug. 73074 73075 v2: fix some strncmp returns. 73076 73077 Reviewed-by: Keith Packard <keithp@keithp.com> 73078 Signed-off-by: Dave Airlie <airlied@redhat.com> 73079 73080commit 53b66c084fbf5bf47f8121632fabdc39b9fe8337 73081Author: Dave Airlie <airlied@redhat.com> 73082Date: Tue Jun 5 17:18:10 2012 +0100 73083 73084 xfree86: add DDX gpu screen support. (v3) 73085 73086 This just adds the structures and interfaces required for adding/deleteing 73087 gpu screens at the DDX level. The platform probe can pass a new flag 73088 to the driver, so they can call xf86AllocateScreen and pass back the new 73089 gpu screen flag. 73090 73091 It also calls the gpu screens preinit and screeninit routines at 73092 startup. 73093 73094 v2: fix delete screen use after free. 73095 73096 v3: split out pScrn into separate patch 73097 73098 Reviewed-by: Keith Packard <keithp@keithp.com> 73099 Signed-off-by: Dave Airlie <airlied@redhat.com> 73100 73101commit 726d467b534a30d8cbaafb858339c638c9f83195 73102Author: Dave Airlie <airlied@redhat.com> 73103Date: Mon Jun 25 11:51:26 2012 +0100 73104 73105 xf86: cleanup helper code to use a pointer. (v1.1) 73106 73107 This is in preparation for gpu screens in here, just use 73108 a pScrn pointer to point at the new screen. 73109 73110 suggested by Keith. 73111 v1.1: fix spacing as suggested by Aaron. 73112 73113 Reviewed-by: Keith Packard <keithp@keithp.com> 73114 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 73115 Signed-off-by: Dave Airlie <airlied@redhat.com> 73116 73117commit 9d179818293b466ec6f1777f0b792e1fbbeb318c 73118Author: Dave Airlie <airlied@redhat.com> 73119Date: Tue Jun 5 17:09:38 2012 +0100 73120 73121 dix: introduce gpu screens. (v5) 73122 73123 This patch introduces gpu screens into screenInfo. It adds interfaces 73124 for adding and removing gpu screens, along with adding private fixup, 73125 block handler support, and scratch pixmap init. 73126 73127 GPU screens have a myNum that is offset by GPU_SCREEN_OFFSET (256), 73128 this is used for logging etc. 73129 73130 RemoveGPUScreen isn't used until "xfree86: add platform bus hotplug support". 73131 73132 v2: no glyph pictures for GPU screens for now. 73133 v3: introduce MAXGPUSCREENS, fix return value check 73134 v4: fixup myNum when renumbering screens (ajax) 73135 v5: drop cursor privates for now. 73136 73137 Reviewed-by: Keith Packard <keithp@keithp.com> 73138 Signed-off-by: Dave Airlie <airlied@redhat.com> 73139 73140commit 3478af3374abffa0c226ee077fda1fcfc0751e74 73141Author: Dave Airlie <airlied@redhat.com> 73142Date: Tue Jun 5 17:01:21 2012 +0100 73143 73144 screen: split out screen init code. (v2) 73145 73146 This is a precursor for reusing this code to init gpu screens. 73147 73148 v2: fixup int check as per Keith's review. 73149 73150 Reviewed-by: Keith Packard <keithp@keithp.com> 73151 Signed-off-by: Dave Airlie <airlied@redhat.com> 73152 73153commit cf66471353ac5899383b573a3cfca407e90d501e 73154Author: Dave Airlie <airlied@redhat.com> 73155Date: Tue May 8 14:26:47 2012 +0100 73156 73157 xfree86: use udev to provide device enumeration for kms devices (v10) 73158 73159 On Linux in order for future hotplug work, we are required to interface 73160 to udev to detect device creation/removal. In order to try and get 73161 some earlier testing on this, this patch adds the ability to use 73162 udev for device enumeration on Linux. 73163 73164 At startup the list of drm/kms devices is probed and this info is 73165 used to load drivers. 73166 73167 A new driver probing method is introduced that passes the udev 73168 device info to the driver for probing. 73169 73170 The probing integrates with the pci probing code and will fallback 73171 to the pci probe and old school probe functions in turn. 73172 73173 The flags parameter to the probe function will be used later 73174 to provide hotplug and gpu screen flags for the driver to behave 73175 in a different way. 73176 73177 This patch changes the driver ABI, all drivers should at least 73178 be set with a NULL udev probe function after this commit. 73179 73180 v2: rename to platform bus, now with 100% less udev specific, 73181 73182 this version passes config_odev_attribs around which are an array 73183 of id/string pairs, then the udev code can attach the set of attribs 73184 it understands, the OS specific code can attach its attrib, and then 73185 the core/drivers can lookup the required attribs. 73186 73187 also add MATCH_PCI_DEVICES macro. 73188 73189 This version is mainly to address concerns raised by ajax. 73190 73191 v3: Address comments from Peter. 73192 fix whitespace that snuck in. 73193 rework to use a linked list with some core functions that 73194 xf86 wraps. 73195 73196 v4: add free list, fix struct whitespace. 73197 ajax this address most of your issues? 73198 73199 v5: drop probe ifdef, fix logic issue 73200 73201 v6: some overhaul after more testing. 73202 73203 Implement primaryBus for platform devices. 73204 document hotplug.h dev attribs - drop sysname attrib 73205 fix build with udev kms disabled 73206 make probing work like the PCI probe code, 73207 match against bus id if one exists, or primary device. 73208 73209 RFC: add new bus id support "PLAT:syspath". we probably 73210 want to match on this a bit different, or use a different 73211 property maybe. I was mainly wanting this for use with 73212 specifying usb devices in xorg.conf directly, but PLAT:path 73213 could also work I suppose. 73214 73215 v6.1: add missing noop platform function 73216 73217 v7: fix two interactions with pci probing and slot claiming, prevents 73218 pci and platform trying to load two drivers for same slot. 73219 73220 v8: test with zaphod mode on -ati driver, fixup resulting issue 73221 clean up common probe code into another function, change busid 73222 matching to allow dropping end of strings. 73223 73224 v9: fix platform probing logic so it actually works. 73225 v9.1: fix pdev init to NULL properly. 73226 73227 v10: address most of Keith's concerns. 73228 73229 v4 was thanks to Reviewed-by: Adam Jackson <ajax@redhat.com> 73230 v5 was Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73231 Reviewed-by: Keith Packard <keithp@keithp.com> 73232 Signed-off-by: Dave Airlie <airlied@redhat.com> 73233 73234commit 2c52d776a42a28bb3e1463edf0cfe0672d0c08a9 73235Author: Dave Airlie <airlied@redhat.com> 73236Date: Tue Jun 19 16:01:08 2012 +0100 73237 73238 xf86: cursor code got mangled by indenting 73239 73240 This fixes some really ugly code that got mangled by the indenting. 73241 73242 Reviewed-by: Keith Packard <keithp@keithp.com> 73243 Signed-off-by: Dave Airlie <airlied@redhat.com> 73244 73245commit 9e4b8b74b008dd72d1da7074bd7ecae677cefd20 73246Author: Keith Packard <keithp@keithp.com> 73247Date: Thu Jul 5 11:31:55 2012 -0700 73248 73249 privates: Allow device privates to be allocated after server start 73250 73251 This will permit midispcur to allocate its privates for hotplug outputs 73252 73253 Signed-off-by: Keith Packard <keithp@keithp.com> 73254 Reviewed-by: Dave Airlie <airlied@redhat.com> 73255 73256commit 493ad833233707b0a42e5fbdd957a24fc6e52917 73257Author: Keith Packard <keithp@keithp.com> 73258Date: Thu Jul 5 11:31:29 2012 -0700 73259 73260 dix: Allocate device privates separate from devices 73261 73262 This will allow device privates to be resized. 73263 73264 Signed-off-by: Keith Packard <keithp@keithp.com> 73265 Reviewed-by: Dave Airlie <airlied@redhat.com> 73266 73267commit f38b2b628322f4d763b0c7f8387d5ab0472a7d2b 73268Author: Keith Packard <keithp@keithp.com> 73269Date: Thu Jul 5 11:15:18 2012 -0700 73270 73271 mi: Don't save midispcur rendering resources for each cursor 73272 73273 Instead, only save them for the most recently drawn cursor. This saves 73274 a bunch of storage for idle cursors. 73275 73276 Signed-off-by: Keith Packard <keithp@keithp.com> 73277 Reviewed-by: Dave Airlie <airlied@redhat.com> 73278 73279commit 7651176b0093cf3a4cff2277f6918a7d0451af41 73280Author: Keith Packard <keithp@keithp.com> 73281Date: Tue Jul 3 16:12:59 2012 -0700 73282 73283 Use screen-specific privates for exa pixmap and gc privates 73284 73285 Signed-off-by: Keith Packard <keithp@keithp.com> 73286 Reviewed-by: Dave Airlie <airlied@redhat.com> 73287 73288commit 7ca32f1c282cb6d4d7255f118a1e2e6f54e6b3fa 73289Author: Keith Packard <keithp@keithp.com> 73290Date: Tue Jul 3 16:05:42 2012 -0700 73291 73292 Use new screen-specific privates for fb window and gc privates 73293 73294 This ensures that only screens using fb will have this space allocated. 73295 73296 Signed-off-by: Keith Packard <keithp@keithp.com> 73297 Reviewed-by: Dave Airlie <airlied@redhat.com> 73298 73299commit 9d457f9c55f12106ba44c1c9db59d14f978f0ae8 73300Author: Keith Packard <keithp@keithp.com> 73301Date: Tue Jul 3 15:45:09 2012 -0700 73302 73303 Add screen-specific privates. 73304 73305 Screen-specific privates areas are only allocated for objects related 73306 to the target screen; objects allocated for other screens will not 73307 have the private space reserved. This saves memory in these objects 73308 while also allowing hot-plug screens to have additional private 73309 allocation space beyond what the core screens are using. 73310 73311 Drivers are encouraged to switch to this mechanism as it will reduce 73312 memory usage in multi-GPU environments, but it is only required for 73313 drivers which will be loaded after the server starts, like 73314 modesetting. 73315 73316 Objects providing screen-specific privates *must* be managed by the 73317 screen-specific private API when allocating or initializing privates 73318 so that the per-screen area can be initialized properly. 73319 73320 The objects which support screen-specific privates are: 73321 73322 Windows 73323 Pixmaps 73324 GCs 73325 Pictures 73326 73327 Extending this list to include Colormaps would be possible, but 73328 require slightly more work as the default colormap is created before 73329 all colormap privates are allocated during server startup, and hence 73330 gets a bunch of special treatment. 73331 73332 Of particular note, glyphs are *not* capable of supporting 73333 screen-specific privates as they are global objects, not allocated on 73334 a screen-specific basis, and so each driver must be able to see their 73335 privates within the glyph. 73336 73337 Signed-off-by: Keith Packard <keithp@keithp.com> 73338 Reviewed-by: Dave Airlie <airlied@redhat.com> 73339 73340commit ed6daa15a7dcf8dba930f67401f4c1c8ca2e6fac 73341Author: Ian Romanick <ian.d.romanick@intel.com> 73342Date: Wed Jul 4 15:21:09 2012 -0700 73343 73344 glx/dri2: Enable GLX_ARB_create_context_robustness 73345 73346 If the driver supports __DRI2_ROBUSTNESS, then enable 73347 GLX_ARB_create_cotnext_robustness as well. If robustness values are 73348 passed to glXCreateContextAttribsARB and the driver doesn't support 73349 __DRI2_ROBUSTNESS, existing drivers will already generate the correct 73350 error values (so that the correct GLX errors are generated). 73351 73352 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 73353 Reviewed-by: Dave Airlie <airlied@redhat.com> 73354 Signed-off-by: Keith Packard <keithp@keithp.com> 73355 73356commit 90aa2486e394c0344aceb2a70432761665a79333 73357Author: Ian Romanick <ian.d.romanick@intel.com> 73358Date: Wed Jul 4 15:21:08 2012 -0700 73359 73360 glx: Allow attributes and flags from GLX_ARB_create_context_robustness 73361 73362 The assumtion is that if the underlying provider (e.g., glxdri2.c) doesn't 73363 support this extension, it will generate BadMatch for these flags and 73364 attributes. 73365 73366 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 73367 Reviewed-by: Dave Airlie <airlied@redhat.com> 73368 Signed-off-by: Keith Packard <keithp@keithp.com> 73369 73370commit aef0b4593e165634ca7fa0745882dad12a54ebed 73371Author: Ian Romanick <ian.d.romanick@intel.com> 73372Date: Wed Jul 4 15:21:07 2012 -0700 73373 73374 glx: Track the reset notification strategy for a context 73375 73376 Also require that the reset notification for a new context and the other 73377 context in the share group match. There isn't yet any way to specify a 73378 non-default reset notification strategy, but that will come. 73379 73380 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 73381 Reviewed-by: Dave Airlie <airlied@redhat.com> 73382 Signed-off-by: Keith Packard <keithp@keithp.com> 73383 73384commit f8198285cfc4704bc4d56922e5a2e99030a83778 73385Author: Ian Romanick <ian.d.romanick@intel.com> 73386Date: Wed Jul 4 15:21:06 2012 -0700 73387 73388 glx: Add extension string tracking for GLX_ARB_create_context_robustness 73389 73390 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 73391 Reviewed-by: Dave Airlie <airlied@redhat.com> 73392 Signed-off-by: Keith Packard <keithp@keithp.com> 73393 73394commit 4ffbc984eac8a15925094614021de30dc8084e65 73395Author: Ian Romanick <ian.d.romanick@intel.com> 73396Date: Wed Jul 4 15:21:05 2012 -0700 73397 73398 glx: Reject creations of indirect contexts with unsupportable versions. 73399 73400 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 73401 Reviewed-by: Dave Airlie <airlied@redhat.com> 73402 Signed-off-by: Keith Packard <keithp@keithp.com> 73403 73404commit 41d5120eaf2bdf308eb904cac9b5027b003fea8d 73405Author: Ian Romanick <ian.d.romanick@intel.com> 73406Date: Wed Jul 4 15:21:04 2012 -0700 73407 73408 glx: Implement GLX_EXT_create_context_es2_profile 73409 73410 This patch builds on the infrastucture put in place for 73411 GLX_ARB_create_context_profile. If GLX_CONTEXT_ES2_PROFILE_BIT_EXT is 73412 specified and the requested version is 2.0, create a context with the 73413 __DRI_API_GLES2 API. 73414 73415 This change assumes that any DRI2 driver can handle (possibly by saying "no 73416 seeing an API setting other than __DRI_API_OPENGL or __DRI_API_OPENGL_CORE. 73417 This allows enabling the extension any time GLX_ARB_create_context (and 73418 GLX_ARB_create_context_profile) is enabled. 73419 73420 v2: Clean up some comments. Note that our behavior for 73421 GLX_CONTEXT_ES2_PROFILE_BIT_EXT w/version != 2.0 matches NVIDIA's. 73422 73423 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 73424 Reviewed-by: Dave Airlie <airlied@redhat.com> 73425 Signed-off-by: Keith Packard <keithp@keithp.com> 73426 73427commit cef0b808d3152a20cd98d32457dde08bd5434e7c 73428Author: Ian Romanick <ian.d.romanick@intel.com> 73429Date: Wed Jul 4 15:21:03 2012 -0700 73430 73431 glx: Implement GLX_ARB_create_context_profile 73432 73433 Most of the infrastructure was already in place. This just adds: 73434 73435 * Validate values specified with the GLX_CONTEXT_PROFILE_MASK_ARB 73436 attribute. 73437 73438 * Select a DRI2 "api" based on the setting of 73439 GLX_CONTEXT_PROFILE_MASK_ARB. 73440 73441 * Enable GLX_ARB_create_context_profile extension. 73442 73443 This change assumes that any DRI2 driver can handle (possibly by saying "no") 73444 seeing an API setting other than __DRI_API_OPENGL. This allows enabling this 73445 extension any time GLX_ARB_create_context is enabled. 73446 73447 Also, the validation code in __glXDisp_CreateContextAttribsARB is structured 73448 in a very verbose manner (using a switch-statement) to ease the addition of 73449 GLX_EXT_create_context_es2_profile. 73450 73451 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 73452 Reviewed-by: Dave Airlie <airlied@redhat.com> 73453 Signed-off-by: Keith Packard <keithp@keithp.com> 73454 73455commit 09a8a169d5929665778ae33e58e345fe1192b64c 73456Author: Ian Romanick <ian.d.romanick@intel.com> 73457Date: Wed Jul 4 15:21:02 2012 -0700 73458 73459 glx: Implement GLX SetClientInfo2ARB protocol 73460 73461 The protocol is almost identical to SetClientInfoARB. The only 73462 difference is the GL versions include an extra 4 bytes for the supported 73463 profile. 73464 73465 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 73466 Reviewed-by: Dave Airlie <airlied@redhat.com> 73467 Signed-off-by: Keith Packard <keithp@keithp.com> 73468 73469commit 167993254a5cbe11a1f44fad1e8ae042089c1619 73470Author: Peter Hutterer <peter.hutterer@who-t.net> 73471Date: Thu Jul 5 16:17:56 2012 +1000 73472 73473 list.h: don't crash when removing an element from a NULL list 73474 73475 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73476 Signed-off-by: Keith Packard <keithp@keithp.com> 73477 73478commit 8aa6d49d46dacc31a85023d0159c1fb67401df73 73479Author: Ian Romanick <ian.d.romanick@intel.com> 73480Date: Wed Jul 4 15:21:01 2012 -0700 73481 73482 glx: Don't request prototypes for GL extension functions 73483 73484 The server does not want GL extension prototypes. It never links with 73485 anything that could possibly provide implementations of these functions. It 73486 *is* the provide, and it does not provde these symbols. All this does is 73487 create hundreds of warnings like: 73488 73489 incude/GL/glext.h:5570:45: warning: redundant redeclaration of 'glMultTransposeMatrixd' [-Wredundant-decls] 73490 include/GL/gl.h:1940:45: note: previous declaration of 'glMultTransposeMatrixd' was here 73491 73492 I believe this was necessary back before the AIGLX days, but that was a long, 73493 long time ago. The warnings recently showed up with the addition of 73494 -Wredundant-decls to the CWARNFLAGS. 73495 73496 v2: Update the commit message, no code changes. 73497 73498 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 73499 Cc: Adam Jackson <ajax@redhat.com> 73500 Cc: Dave Airlie <airlied@redhat.com> 73501 Acked-by: Keith Packard <keithp@keithp.com> 73502 Signed-off-by: Keith Packard <keithp@keithp.com> 73503 73504commit 258abbf823f753757e4dddc13ef495f4024680db 73505Author: Michal Srb <msrb@suse.com> 73506Date: Thu Jun 28 17:17:12 2012 +0200 73507 73508 Look for ModuleData only in appropriate library 73509 73510 LoaderSymbol calls dlsym with RTLD_DEFAULT pseudo handle making it search in 73511 every loaded library. In addition glibc adds NODELETE flag to the library 73512 containing the symbol. 73513 73514 It's used in doLoadModule to locate <modulename>ModuleData symbol, the 73515 module's library gets the flag and is kept in memory even after it is 73516 unloaded. 73517 73518 This patch adds LoaderSymbolFromModule function that looks for symbol only in 73519 library specified by handle. That way the NODELETE flag isn't added. 73520 73521 This glibc behavior doesn't seem to be documented, but even if other 73522 implementations differ, there is no reason to search ModuleData symbol outside 73523 the module's library. 73524 73525 Signed-off-by: Michal Srb <msrb@suse.com> 73526 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 73527 73528 v2: Switch LoaderSymbolFromModule arguments order. 73529 Correct description. 73530 Reviewed-by: Adam Jackson <ajax@redhat.com> 73531 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73532 73533commit d84f0f823eeeecdf0498aadd3fbb1d11dabc0837 73534Merge: 12bfb4cf1 d04dfe3f7 73535Author: Peter Hutterer <peter.hutterer@who-t.net> 73536Date: Wed Jul 4 21:23:48 2012 +1000 73537 73538 Merge branch 'sigio-vt-switch-issues' into for-keith 73539 73540 Conflicts: 73541 test/Makefile.am 73542 73543 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73544 73545commit 12bfb4cf1bebb66d2c2eb76b93c18a2915b865e5 73546Author: Peter Hutterer <peter.hutterer@who-t.net> 73547Date: Tue Jun 26 09:19:33 2012 +1000 73548 73549 Xi: extend PropagateMask to EMASKSIZE 73550 73551 Number of devices is 2 + MAXDEVICES, with index 0 and 1 reserved for 73552 XIAll{Master}Devices. At the current size, PropagateMask would be overrun in 73553 RecalculateDeviceDeliverableEvents(). 73554 73555 Found by Coverity. 73556 73557 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73558 Reviewed-by: Dave Airlie <airlied@redhat.com> 73559 73560commit a9c09f8f8e2a97e4bfe927bc4c7f29b04ee3403a 73561Author: Peter Hutterer <peter.hutterer@who-t.net> 73562Date: Mon Jun 25 14:24:11 2012 +1000 73563 73564 dix: fix memory leak in TouchEventHistoryReplay 73565 73566 Don't leak if ti->history is NULL. 73567 73568 Found by coverity. 73569 73570 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73571 Reviewed-by: Dave Airlie <airlied@redhat.com> 73572 73573commit 55ff20eb37f00a3ad07b5acff19d4497ab513a97 73574Author: Peter Hutterer <peter.hutterer@who-t.net> 73575Date: Mon Jun 25 14:22:31 2012 +1000 73576 73577 dix: fix dereference before null check 73578 73579 Found by Coverity. 73580 73581 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73582 Reviewed-by: Dave Airlie <airlied@redhat.com> 73583 73584commit e3f47be9fbc489e3eb5832445924810b2ff300f7 73585Author: Peter Hutterer <peter.hutterer@who-t.net> 73586Date: Mon Oct 24 10:01:26 2011 +1000 73587 73588 xfree86: fix use-after-free issue in checkInput 73589 73590 *dev is the condition of the while loop we're in, reset to NULL after 73591 freeing 73592 73593 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73594 Reviewed-by: Dave Airlie <airlied@redhat.com> 73595 73596commit 252a69b592001b34714d0fe7508e1898883f9d01 73597Author: Peter Hutterer <peter.hutterer@who-t.net> 73598Date: Wed Jun 20 14:25:54 2012 +1000 73599 73600 xkb: use local variable instead of casting arg 73601 73602 No functional changes. 73603 73604 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73605 73606commit d04dfe3f754ad3a5f158057175cbd44319c1ae51 73607Author: Peter Hutterer <peter.hutterer@who-t.net> 73608Date: Fri Jun 22 15:18:53 2012 +1000 73609 73610 Drop custom sigio block/unblock functions from kdrive, ephyr and dmx 73611 73612 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73613 Reviewed-by: Keith Packard <keithp@keithp.com> 73614 73615commit 5d309af2ed93e91c7d72f548a11052051efbb40f 73616Author: Peter Hutterer <peter.hutterer@who-t.net> 73617Date: Fri Jun 22 14:56:25 2012 +1000 73618 73619 xfree86: drop ddx-specific SIGIO blocking 73620 73621 The hooks are left for this cycle, we can drop it next cycle once the 73622 drivers that need it (e.g. wacom) have been updated. 73623 73624 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73625 Reviewed-by: Keith Packard <keithp@keithp.com> 73626 73627commit ff67135bec3ecea8cebeb46b42ece0a6671e231d 73628Author: Peter Hutterer <peter.hutterer@who-t.net> 73629Date: Fri Jun 22 13:09:35 2012 +1000 73630 73631 xfree86: use OsBlockSIGIO from the ddx 73632 73633 We can ignore the "wasset" argument now since the DIX will keep proper 73634 refcounting. 73635 73636 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73637 Reviewed-by: Keith Packard <keithp@keithp.com> 73638 73639commit 6bf356ef2831baeccd7a650ed3fde0831e33c6c0 73640Author: Peter Hutterer <peter.hutterer@who-t.net> 73641Date: Fri Jun 22 13:02:40 2012 +1000 73642 73643 os: add OsBlockSIGIO and OsReleaseSIGIO 73644 73645 Let the dix be in charge of changing the sigprocmask so we only have one 73646 entity that changes it. 73647 73648 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73649 Reviewed-by: Keith Packard <keithp@keithp.com> 73650 73651commit 24525d96a3b9dba67eb75042500b2f208a2cc246 73652Merge: 4cd91bd4c 35e3d2291 73653Author: Keith Packard <keithp@keithp.com> 73654Date: Mon Jul 2 22:35:39 2012 -0700 73655 73656 Merge branch 'sigsafe-logging-varargs' 73657 73658 This merge includes a minor fixup for '%p' arguments; must cast to 73659 uintptr_t instead of uint64_t as we use -Werror=pointer-to-int-cast 73660 which complains when doing a cast (even explicitly) from a pointer 73661 to an integer of different size. 73662 73663commit 35e3d229150395a222a0f53318daf5dbeb8f6eb6 73664Author: Peter Hutterer <peter.hutterer@who-t.net> 73665Date: Thu May 24 14:04:42 2012 +1000 73666 73667 Bump to ABI_XINPUT_VERSION 18 73668 73669 The input ABI hasn't changed, but input drivers need something to hook on if 73670 they want to log from within signal handlers and the input ABI is the 73671 simplest way of doing so. 73672 73673 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73674 73675commit 541934168dbeb17059542bb5a1da8eba7995fa05 73676Author: Peter Hutterer <peter.hutterer@who-t.net> 73677Date: Mon May 28 10:10:30 2012 +1000 73678 73679 xfree86: constify InputDriverPtr->driverName and default_options 73680 73681 Already treated as const anyway by all drivers. 73682 73683 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73684 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 73685 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73686 73687commit c66089d2206bafc01307a8327ff6089edcb4ed2d 73688Author: Peter Hutterer <peter.hutterer@who-t.net> 73689Date: Mon May 28 09:57:33 2012 +1000 73690 73691 xfree86: constify InputInfoPtr->type_name 73692 73693 This corresponds to XListInputDevice(3)'s "type" field (after being 73694 converted to an Atom). Input drivers use the XI_KEYBOARD and similar 73695 defines, even Wacom which falls out of the common defines uses constant 73696 strings here. The use-case for having this non-const is small. 73697 73698 Input ABI break technically, since we never freed this information anyway it 73699 is not a noticable change. 73700 73701 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73702 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 73703 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73704 73705commit 505c8a2b2cae0318db1148417ec850d54b38f7df 73706Author: Chase Douglas <chase.douglas@canonical.com> 73707Date: Mon Apr 9 09:41:38 2012 -0700 73708 73709 Log in OsVendorFatalError() in a signal safe manner 73710 73711 The function can be called from a fatal signal handler. 73712 73713 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73714 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73715 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73716 73717commit d51aebdbf99a9f240f7c318a70ba40e61cd43049 73718Author: Chase Douglas <chase.douglas@canonical.com> 73719Date: Mon Apr 9 08:30:50 2012 -0700 73720 73721 Log in LoaderUnload() in a signal safe manner 73722 73723 The function may be called from a fatal signal handler. 73724 73725 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73726 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73727 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73728 73729commit c3e1168778ec20beeac9979dc57e36400c70dd63 73730Author: Chase Douglas <chase.douglas@canonical.com> 73731Date: Mon Apr 9 08:28:17 2012 -0700 73732 73733 Log in UnloadModuleOrDriver() in a signal safe manner 73734 73735 The function may be called from a fatal signal handler. 73736 73737 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73738 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73739 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73740 73741commit 89e3ac07aca1def155299aff6f7a57ccafb68fd7 73742Author: Chase Douglas <chase.douglas@canonical.com> 73743Date: Mon Apr 9 08:23:32 2012 -0700 73744 73745 Log safely in fatal signal handler 73746 73747 While we probably don't need to be signal safe here since we will never 73748 return to the normal context, the logging signal context check will 73749 cause unsafe logging to be unhandled. Using signal safe logging here 73750 resolves the issue. 73751 73752 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73753 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73754 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73755 73756commit 6fd5add005d0660b591d808583d1a6c6a85f1277 73757Author: Chase Douglas <chase.douglas@canonical.com> 73758Date: Fri Apr 6 16:17:41 2012 -0700 73759 73760 Log mieq enqueue overflow in a signal safe manner 73761 73762 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73763 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73764 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73765 73766commit 7f4a69b628a6246855054a0b94d6d6dd14e8842c 73767Author: Chase Douglas <chase.douglas@canonical.com> 73768Date: Fri Apr 13 16:01:38 2012 -0700 73769 73770 Log messages in TouchBeginDDXTouch() in a signal-safe manner 73771 73772 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73773 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73774 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73775 73776commit f752226e40890643df213a62f0c96e6a0243e754 73777Author: Chase Douglas <chase.douglas@canonical.com> 73778Date: Fri Apr 6 10:21:14 2012 -0700 73779 73780 Log messages in GetTouchEvents() in a signal safe manner 73781 73782 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73783 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73784 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73785 73786commit 82d1c6b310eaa5095eed9ee4ea958261a46a78e1 73787Author: Chase Douglas <chase.douglas@canonical.com> 73788Date: Fri Apr 6 08:32:28 2012 -0700 73789 73790 Warn when attempting to log in a signal unsafe manner from signal context 73791 73792 Also, print out the offending message format. This will hopefully help 73793 developers track down unsafe logging. 73794 73795 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73796 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73797 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73798 73799commit 512bec06be6c79ca263da9de8f40430b8095b57b 73800Author: Chase Douglas <chase.douglas@canonical.com> 73801Date: Mon Apr 16 09:47:42 2012 -0700 73802 73803 Make BUG_WARN* signal safe 73804 73805 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73806 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73807 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73808 73809commit 0fa5217836cf7fd3872fccc9f3ff9ff32426c25b 73810Author: Chase Douglas <chase.douglas@canonical.com> 73811Date: Fri Apr 6 16:09:05 2012 -0700 73812 73813 Print backtrace in a signal-safe manner 73814 73815 Backtraces are often printed in signal context, such as when a segfault 73816 occurs. 73817 73818 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73819 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73820 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73821 73822 os: print offset as unsigned int, not long unsigned int 73823 73824 pnprintf() takes unsigned int for %u 73825 73826 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73827 73828commit ac20815d5235e7a8e7b331365aabf5a489fc5e34 73829Author: Chase Douglas <chase.douglas@canonical.com> 73830Date: Tue Jun 5 15:39:41 2012 +1000 73831 73832 Add ErrorFSigSafe() alternative to ErrorF() 73833 73834 ErrorF() is not signal safe. Use ErrorSigSafe() whenever an error 73835 message may be logged in signal context. 73836 73837 [whot: edited to "ErrorFSigSafe"] 73838 73839 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73840 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73841 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73842 73843commit 164b38c72fe9c69d13ea4f9c46d4ccc46566d826 73844Author: Chase Douglas <chase.douglas@canonical.com> 73845Date: Fri Apr 6 08:28:40 2012 -0700 73846 73847 Add LogMessageVerbSigSafe() for logging messages while in signal context 73848 73849 [whot: edited to use varargs, squashed commit below] 73850 73851 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 73852 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 73853 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73854 73855 os: fix vararg length calculation 73856 73857 Make %u and %x sizeof(unsigned int), %p sizeof(void*). This is printf 73858 behaviour and we can't guarantee that void* is uint64_t anyway. 73859 73860 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 73861 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 73862 73863commit 4cd91bd4c6325b2e617ca90a725f02ac124c2372 73864Author: Alan Coopersmith <alan.coopersmith@oracle.com> 73865Date: Sat May 12 10:11:33 2012 -0700 73866 73867 Enable no-undefined symbols mode in Solaris builds if -z parent is in ld 73868 73869 The Solaris linker recently added a -z parent flag for easier checking 73870 of symbol definitions in plugins against the program that loads them. 73871 If that's present, this enables it, along with -z defs to error on 73872 undefined symbols to alert us if any modules call symbols that won't 73873 be found at runtime. 73874 73875 This builds upon, and requires, the recent Cygwin work to build Xorg. 73876 It moves a couple more modules to be after the Xorg binary in the build 73877 order so that they can find the binary to check against, much as the 73878 Cygwin changes did (these would be modules built on Solaris but not 73879 Cygwin). 73880 73881 v2: This version only sets the flags for the xorg-server build itself, 73882 and does not yet export them in xorg-server.pc to the drivers, since 73883 most of the drivers are not ready to build with -no-undefined yet. 73884 73885 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 73886 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 73887 73888commit d31942e15c619e9355ef8183223f9165902a9d56 73889Author: Alan Coopersmith <alan.coopersmith@oracle.com> 73890Date: Sat May 12 10:10:49 2012 -0700 73891 73892 Add pixman to module dependencies for pixman_region_* functions 73893 73894 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 73895 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 73896 73897commit 339f38ef380d750c4c7afd60331a559845e4184e 73898Author: Alan Coopersmith <alan.coopersmith@oracle.com> 73899Date: Sat May 12 10:09:41 2012 -0700 73900 73901 Build tda9885.c functions into fi1236 to avoid libtool relinking problems 73902 73903 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 73904 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 73905 73906commit a00066d2916b6910429cc1c7feedafbaee0d4750 73907Author: Alan Coopersmith <alan.coopersmith@oracle.com> 73908Date: Thu Jun 28 15:15:53 2012 -0700 73909 73910 Add dixGetGlyphs to replace GetGlyphs from libXfont to simplify linking 73911 73912 No other Xfont consumer used it, and this saves us from having to link 73913 callers against libXfont for one simple function when doing 73914 -no-undefined symbols builds. 73915 73916 The function is given a new name to avoid clashing with existing libXfont 73917 binaries, but a #define is provided to preserve the API so we don't have 73918 to fix all the callers at the same time. 73919 73920 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 73921 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 73922 73923commit 212b9803238d2de2e77cbe5de62d3f616ae50daf 73924Author: Aaron Plattner <aplattner@nvidia.com> 73925Date: Fri Jun 29 13:33:58 2012 -0700 73926 73927 randr: Fix REQUEST vs. REQUEST_SIZE_MATCH mismatch 73928 73929 ProcRRGetScreenSizeRange uses REQUEST(xRRGetScreenSizeRangeReq) followed by 73930 REQUEST_SIZE_MATCH(xRRGetScreenInfoReq). This happens to work out because both 73931 requests have the same size, so this is not a functional change, just a cosmetic 73932 one. 73933 73934 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 73935 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 73936 Signed-off-by: Keith Packard <keithp@keithp.com> 73937 73938commit 957bf959fb577b292a3e4f6bb67740ca09e7aeb9 73939Author: Dave Airlie <airlied@redhat.com> 73940Date: Wed Jun 27 14:36:04 2012 +0100 73941 73942 dix/render: consolidate window format matching code. 73943 73944 This code existed in 3 different forms, perhaps it should be 73945 consolidated. 73946 73947 Reviewed-by: Keith Packard <keithp@keithp.com> 73948 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 73949 Signed-off-by: Dave Airlie <airlied@redhat.com> 73950 Signed-off-by: Keith Packard <keithp@keithp.com> 73951 73952commit a7b97b0fa85d695ae19d194cfa3267159d149e5d 73953Author: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 73954Date: Thu Jun 28 12:07:11 2012 +1000 73955 73956 dri2: Fix authentication for < v8 clients (v2). 73957 73958 The legacy logic was embarassingly wrong; AuthMagic should return errno, 73959 so returning FALSE only when AuthMagic returns nonzero is exactly wrong. 73960 73961 v2: Match drmAuthMagic by returning -EINVAL rather than EINVAL 73962 Fix trailing whitespace 73963 73964 https://bugs.freedesktop.org/show_bug.cgi?id=51400 73965 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 73966 Tested-by: Chris Wilson <chris@chris-wilson.co.uk> 73967 Reviewed-by: Michel Dänzer <michel.daenzer@amd.com> 73968 Tested-by: Knut Petersen <knut_petersen@t-online.de> 73969 Signed-off-by: Keith Packard <keithp@keithp.com> 73970 73971commit 8b4f0a4fdd6c15f49458bc3a8c3135b8c389240d 73972Merge: 59294a217 d642e7128 73973Author: Keith Packard <keithp@keithp.com> 73974Date: Thu Jun 28 10:38:26 2012 -0700 73975 73976 Merge remote-tracking branch 'jturney/master' 73977 73978commit 59294a2179bd5bb996693af004b470932df00ac6 73979Merge: 855003c33 191b63065 73980Author: Keith Packard <keithp@keithp.com> 73981Date: Thu Jun 28 10:11:44 2012 -0700 73982 73983 Merge remote-tracking branch 'whot/for-keith' 73984 73985commit 855003c333a0ead1db912695bc9705ef2b3144b4 73986Author: Keith Packard <keithp@keithp.com> 73987Date: Thu Jun 21 18:45:18 2012 -0700 73988 73989 randr: Catch two more potential unset rrScrPriv uses 73990 73991 Ricardo Salveti <ricardo.salveti@linaro.org> found one place where the 73992 randr code could use the randr screen private data without checking 73993 for null first. This happens when the X server is running with 73994 multiple screens, some of which are randr enabled and some of which 73995 are not. Applications making protocol requests to the non-randr 73996 screens can cause segfaults where the server touches the unset private 73997 structure. 73998 73999 I audited the code and found two more possible problem spots; the 74000 trick to auditing for this issue was to look for functions not taking 74001 a RandR data structure and where there was no null screen private 74002 check above them in the call graph. 74003 74004 Signed-off-by: Keith Packard <keithp@keithp.com> 74005 74006commit 32603f57ca03b6390b109960f8bb5ea53ac95ecb 74007Author: Ricardo Salveti de Araujo <ricardo.salveti@linaro.org> 74008Date: Thu Jun 21 00:55:53 2012 -0300 74009 74010 randr: first check pScrPriv before using the pointer at RRFirstOutput 74011 74012 Fix a seg fault in case pScrPriv is NULL at ProcRRGetScreenInfo, 74013 which later calls RRFirstOutput. 74014 74015 Signed-off-by: Ricardo Salveti de Araujo <ricardo.salveti@linaro.org> 74016 Reviewed-by: Keith Packard <keithp@keithp.com> 74017 Signed-off-by: Keith Packard <keithp@keithp.com> 74018 74019commit cf196efa05bbefdeb43f6e6bcc5f6d9080fea715 74020Author: Dave Airlie <airlied@redhat.com> 74021Date: Fri Jun 22 15:24:43 2012 +0100 74022 74023 modesetting 0.4.0: bump configure.ac 74024 74025commit 30ab80ef5421fea9d9dd0779aa138892ff62bd6e 74026Author: Dave Airlie <airlied@gmail.com> 74027Date: Fri Jun 22 15:26:28 2012 +0100 74028 74029 modesetting: pci probing requires interface version 1.4 74030 74031 Set the drm interface version to 1.4 so we get the bus id correctly. 74032 74033 Signed-off-by: Dave Airlie <airlied@redhat.com> 74034 74035commit d642e7128790ec6b709df8d1a91333ab87598077 74036Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 74037Date: Thu Jun 21 20:01:49 2012 +0100 74038 74039 configure.ac: link XWIN with GLX_SYS_LIBS 74040 74041 Link XWIN with GLX_SYS_LIBS, just like all the other DDXs 74042 74043 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 74044 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 74045 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 74046 74047commit 9f1edced9abc066f0ba47672d006fe50fb206371 74048Author: Peter Hutterer <peter.hutterer@who-t.net> 74049Date: Fri Jun 15 10:00:51 2012 +1000 74050 74051 xfree86: always enable SIGIO on OsVendorInit (#50957) 74052 74053 Drivers call xf86InstallSIGIOHandler() for their fd on DEVICE_ON. That 74054 function does not actually enable the signal if it was blocked to begin 74055 with. As a result, if one vt-switches away from the server (SIGIO is 74056 blocked) and then triggers a server regeneration, the signal remains 74057 blocked and input devices are dead. 74058 74059 Avoid this by always unblocking SIGIO when we start the server. 74060 74061 X.Org Bug 50957 <http://bugs.freedesktop.org/show_bug.cgi?id=50957> 74062 74063 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74064 Reviewed-by: Keith Packard <keithp@keithp.com> 74065 74066commit 4ba340cfaa8d430c808566495f8deda0ff1b4424 74067Author: Keith Packard <keithp@keithp.com> 74068Date: Thu Jun 21 18:42:46 2012 -0700 74069 74070 randr: Clean up compiler warnings about unused and shadowing variables 74071 74072 set but not used variables 74073 shadowing a previous local 74074 74075 A hidden problem was that the VERIFY_RR_* macros define local 'rc' 74076 variables, any other local definitions for those would be shadowed and 74077 generate warnings from gcc. I've renamed the other locals 'ret' 74078 instead of 'rc'. 74079 74080 Signed-off-by: Keith Packard <keithp@keithp.com> 74081 74082commit 191b630656ba206ad78b7803c3a76f7d9567d9d8 74083Author: Peter Hutterer <peter.hutterer@who-t.net> 74084Date: Thu Jun 21 11:25:24 2012 +1000 74085 74086 include: document _XkbErrCode2 macros 74087 74088 Why sending the number of the (implementation-dependent) error statement to 74089 the client is a good idea is a bit beyond me, but at least document it so we 74090 can all share the despair. 74091 74092 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74093 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 74094 74095commit 7100118c8d3314d73bfef0a262af56686b227055 74096Author: Peter Hutterer <peter.hutterer@who-t.net> 74097Date: Mon Jun 18 11:28:52 2012 +1000 74098 74099 AC_SUBST the GLX_SYS_LIBS 74100 74101 libxservertest needs -lpthread from glxapi.c's pthread_once() call. Usually 74102 this would be pulled in by the XORG_LIBS but not when building without Xorg. 74103 74104 This commit has no visible effect on the current tree, preparation for test 74105 cleanups. 74106 74107 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74108 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74109 74110commit 014ad46f1b353a95e2c4289443ee857cfbabb3ae 74111Author: Peter Hutterer <peter.hutterer@who-t.net> 74112Date: Thu Jun 21 14:26:15 2012 +1000 74113 74114 os: print newline after printing display name 74115 74116 Much easier for scripts that try to read the display value off the file 74117 descriptor. Plus, this restores the behaviour we had for this patch in 74118 Fedora since server 1.6 (April 2009). 74119 74120 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74121 Reviewed-by: Dave Airlie <airlied@redhat.com> 74122 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74123 74124commit e1f86a7c77ed49e001ea42ef76ab83e87dbdf126 74125Author: Sergei Trofimovich <slyfox@gentoo.org> 74126Date: Mon Jun 18 01:13:21 2012 +0300 74127 74128 test/.gitignore: add hashtabletest 74129 74130 Signed-off-by: Sergei Trofimovich <slyfox@gentoo.org> 74131 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 74132 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74133 74134commit c9bbf8fe574591c8e191eeef20f8994781718772 74135Author: Jon TURNEY <jon.turney@dronecode.org.uk> 74136Date: Thu Jun 21 19:38:55 2012 +0100 74137 74138 hw/xwin/glx: Fix glxWinCreateContext() function pointer mismatch 74139 74140 Fix glxWinCreateContext() function signature to align with commit 96d74138 74141 "glx: Extend __GLXscreen::createContext to take attributes", which added more 74142 parameters to the screen createContext function for implementing GLX_ARB_create_context 74143 74144 indirect.c: In function 'glxWinScreenProbe': 74145 indirect.c:683:36: warning: assignment from incompatible pointer type 74146 74147 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 74148 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 74149 74150commit 6e85660c7503c7e88e24003562444dffde6c88ff 74151Author: Jon TURNEY <jon.turney@dronecode.org.uk> 74152Date: Mon Mar 5 23:07:56 2012 +0000 74153 74154 hw/xwin/glx: Update gen_gl_wrappers.py for the current glx/dispatch.h 74155 74156 Fix the code generator for the dispatch table initialization after 74157 the whitespace/coding style changes to glx/dispatch.h 74158 74159 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 74160 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 74161 74162commit cb638ed9c81f07d8603b07db395e31a41fed8685 74163Author: Jon TURNEY <jon.turney@dronecode.org.uk> 74164Date: Thu Jun 21 18:55:57 2012 +0100 74165 74166 hw/xwin/glx: Blacklist 'GDI generic' GL renderer 74167 74168 If the native GL renderer is the GDI generic renderer (as can happen 74169 if we are in safe mode, or the video driver is VGA, or we have hybrid 74170 graphics which hasn't noticed that xwin requires 3d acceleration), don't 74171 use it. It's not accelerated and we will probably get better conformance 74172 and perfomance from swrast. 74173 74174 Fix so we don't install screen function wrappers in glxWinScreenProbe 74175 unless we are succesful. 74176 74177 Also, move fbConfig dumping to after GLX version has been determined 74178 from extensions 74179 74180 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 74181 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 74182 74183commit 8c24d20933ba9c2ef0e9a81815c3be23fa896aea 74184Author: Jon TURNEY <jon.turney@dronecode.org.uk> 74185Date: Thu Jun 21 18:05:24 2012 +0100 74186 74187 hw/xwin/glx: Fix glxLogExtensions to handle a null string without crashing 74188 74189 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 74190 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 74191 74192commit c08c7c8f655d7721c1e02bfeeb965b6379f72553 74193Author: Jon TURNEY <jon.turney@dronecode.org.uk> 74194Date: Fri Jun 24 13:32:10 2011 +0100 74195 74196 hw/xwin/glx: Create a new dispatch table rather than modifying the existing one 74197 74198 Create a new dispatch table rather than modifying the existing one 74199 74200 struct _glapi_table is not a complete type after including glapi.h, so we use 74201 glapi_get_dispatch_table_size() to determine it's size (alternatively, we could 74202 include glapitable.h, to complete the type) 74203 74204 This could possibly be written to use _glapi_create_table_from_handle() instead, but 74205 that requires making all the wrapper functions exports 74206 74207 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 74208 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 74209 74210commit 3ef3ce069d52dcfa932c90ccd30854a8d9daa15a 74211Merge: a0c8716cd ff541e0a1 74212Author: Keith Packard <keithp@keithp.com> 74213Date: Thu Jun 21 08:52:34 2012 -0700 74214 74215 Merge remote-tracking branch 'alanc/master' 74216 74217commit 704b847abfd29e9adde27127a15a963414f8bcf4 74218Author: Chase Douglas <chase.douglas@canonical.com> 74219Date: Fri Apr 6 10:13:45 2012 -0700 74220 74221 Add FormatUInt64{,Hex}() for formatting numbers in a signal safe manner 74222 74223 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 74224 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 74225 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74226 74227commit bc85c81687a24aea738094ff11f4448fb3b3afbb 74228Author: Chase Douglas <chase.douglas@canonical.com> 74229Date: Fri Apr 6 08:03:09 2012 -0700 74230 74231 Save log file file descriptor for signal context logging 74232 74233 None of the FILE based functions are signal safe, including fileno(), so 74234 we need to save the file descriptor for when we are in signal context. 74235 74236 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 74237 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 74238 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74239 74240commit d3725549f0276487fba1d419094209d18e86669f 74241Author: Chase Douglas <chase.douglas@canonical.com> 74242Date: Fri Apr 6 07:43:57 2012 -0700 74243 74244 Add global variable inSignalContext 74245 74246 This will be used for checking for proper logging when in signal 74247 context. 74248 74249 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 74250 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 74251 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74252 74253commit a0c8716cdca139059e7aec8a3cb75726dd37a4a1 74254Author: Peter Hutterer <peter.hutterer@who-t.net> 74255Date: Thu Jun 21 13:45:06 2012 +1000 74256 74257 kdrive: adjust BlockHandler/WakeupHandler to new API 74258 74259 Missing API update caused Xephyr to crash on startup, garbage memory 74260 dereference when accessing timeout. 74261 Introduced in 1f0e8bd5eb1a5539689cfc4f5a6b86b530907ec5 74262 74263 kdrive.c:868:27: warning: assignment from incompatible pointer type 74264 kdrive.c:869:28: warning: assignment from incompatible pointer type 74265 74266 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74267 Signed-off-by: Keith Packard <keithp@keithp.com> 74268 74269commit ff541e0a1f84bd98966148793f35afeaf71c271b 74270Author: Alan Coopersmith <alan.coopersmith@oracle.com> 74271Date: Wed Jun 13 21:20:51 2012 -0700 74272 74273 Remove obsolete tab stop comments from hw/xfree86/parser/*.c 74274 74275 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74276 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 74277 74278commit 687536b1044a3297f6b9a45f6a2d2987daa7be59 74279Author: Alan Coopersmith <alan.coopersmith@oracle.com> 74280Date: Tue Jun 12 22:38:46 2012 -0700 74281 74282 Fix some overly indented/poorly line wrapped comments in dix/events.c 74283 74284 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74285 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 74286 Reviewed-by: Keith Packard <keithp@keithp.com> 74287 74288commit 93a378aad4a4a03f37ddafccfedeb9d6e2be2a6a 74289Author: Alan Coopersmith <alan.coopersmith@oracle.com> 74290Date: Tue Jun 12 22:29:51 2012 -0700 74291 74292 OtherClientGone: Remove unreachable return statement 74293 74294 Now that FatalError is marked as _X_NORETURN, the compilers know we 74295 can't get here, and the return statement added to make them happy in 74296 the past now makes them unhappy. 74297 74298 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74299 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 74300 Reviewed-by: Keith Packard <keithp@keithp.com> 74301 74302commit 3f97284b10c250457888902debd4d793cb4544d4 74303Author: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 74304Date: Wed Jun 20 13:53:53 2012 +1000 74305 74306 dri2: Pass a ScreenPtr through to the driver's AuthMagic function. (v3) 74307 74308 xwayland drivers need access to their screen private data to authenticate. 74309 Now that drivers no longer have direct access to the global screen arrays, 74310 this needs to be passed in as function context. 74311 74312 v2: Don't break ABI 74313 v3: Paint the bikeshed blue; drop fd from AuthMagic2ProcPtr prototype 74314 74315 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 74316 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 74317 Signed-off-by: Keith Packard <keithp@keithp.com> 74318 74319commit dae317e7265007b38012244722e3b3a06e904ed5 74320Author: Andy Ritger <aritger@nvidia.com> 74321Date: Thu Jun 14 09:15:37 2012 -0700 74322 74323 randr: Don't recurse into mieqProcessInputEvents() from RRTellChanged(). 74324 74325 Call UpdateCurrentTimeIf(), not UpdateCurrentTime(), from RRTellChanged(). 74326 The latter calls ProcessInputEvents(), which can trigger a recursion 74327 into mieqProcessInputEvents(). The former omits the call to 74328 ProcessInputEvents(). 74329 74330 Signed-off-by: Andy Ritger <aritger@nvidia.com> 74331 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 74332 Signed-off-by: Keith Packard <keithp@keithp.com> 74333 74334commit 54476b5e4461ff523e935961affabcf0de12c556 74335Author: Peter Hutterer <peter.hutterer@who-t.net> 74336Date: Thu Jun 7 16:52:20 2012 +1000 74337 74338 dix: if the scroll valuator reaches INT_MAX, reset to 0 74339 74340 Too much scrolling down may eventually trigger an overflow of the valuator. 74341 If this happens, reset the valuator to 0 and skip this event for button 74342 emulation. Clients will have to figure out a way to deal with this, but a 74343 scroll event from (close to) INT_MAX to 0 is a hint of that it needs to be 74344 ignored. 74345 74346 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74347 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74348 74349commit 8dc70acbf3d82611ac9ec1ec2a52edcc01934850 74350Merge: ffb47a123 b59adc9c2 74351Author: Keith Packard <keithp@keithp.com> 74352Date: Thu Jun 14 13:04:29 2012 -0700 74353 74354 Merge remote-tracking branch 'alanc/master' 74355 74356commit ffb47a123ddd1233fb4229cf23483652065c5e82 74357Merge: db9d2b8a5 b840ba5f5 74358Author: Keith Packard <keithp@keithp.com> 74359Date: Thu Jun 14 13:01:17 2012 -0700 74360 74361 Merge remote-tracking branch 'idr/GLX_ARB_create_context' 74362 74363commit db9d2b8a508ab812e5c717a41310faad81879b09 74364Author: Dave Airlie <airlied@redhat.com> 74365Date: Thu Jun 7 15:01:59 2012 +0100 74366 74367 xserver: fix build on arm tinderbox 74368 74369 Rob Clark got an ARM tinderbox up and running, and this code is built there 74370 but not here, this should fix it, though I hope that code never gets executed. 74371 74372 Signed-off-by: Dave Airlie <airlied@redhat.com> 74373 Signed-off-by: Keith Packard <keithp@keithp.com> 74374 74375commit 6d86b64dbaef5a16712fd1fdc3157f716d238877 74376Merge: 812786f4d 4c68f5d39 74377Author: Keith Packard <keithp@keithp.com> 74378Date: Thu Jun 14 11:05:22 2012 -0700 74379 74380 Merge remote-tracking branch 'whot/for-keith' 74381 74382commit b59adc9c2427e5efac3513c20992723e23f92691 74383Author: Alan Coopersmith <alan.coopersmith@oracle.com> 74384Date: Thu Jun 7 20:58:01 2012 -0700 74385 74386 Make stub version of fbdevHWAdjustFrame match new prototype in fbdevhw.h 74387 74388 Removed flags argument to match api rework in commit 1f0e8bd5eb1a55 74389 74390 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74391 Reviewed-by: Dave Airlie <airlied@redhat.com> 74392 74393commit 4c68f5d395c66f28b56e488cb3cd12f36820357b 74394Author: Peter Hutterer <peter.hutterer@who-t.net> 74395Date: Wed May 9 09:21:28 2012 +1000 74396 74397 dix: disable all devices before shutdown 74398 74399 f3410b97cf9b48a47bee3d15d232f8a88e75f4ef introduced a regression on server 74400 shutdown. If any button or key was held on shutdown (ctrl, alt, backspace 74401 are usually still down) sending a raw event will segfault the server. The 74402 the root windows are set to NULL before calling CloseDownDevices(). 74403 74404 Avoid this by disabling all devices first when shutting down. Disabled 74405 devices won't send events anymore. 74406 74407 Master keyboards must be disabled first, otherwise disabling the pointer 74408 will trigger DisableDevice(keyboard) and the keyboard is removed from the 74409 inputInfo.devices list and moved to inputInfo.off_devices. A regular loop 74410 through inputInfo.devices would thus jump to off_devices and not recover. 74411 74412 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74413 Acked-by: Chase Douglas <chase.douglas@canonical.com> 74414 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74415 74416commit 9c0e820216cd1631f75b037b7908d55ac091692c 74417Author: Peter Hutterer <peter.hutterer@who-t.net> 74418Date: Thu May 10 12:55:44 2012 +1000 74419 74420 Xi: drop forced unpairing when changing the hierarchy 74421 74422 Devices are unpaired as needed on DisableDevice now. 74423 74424 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74425 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74426 74427commit e433d1046c222f9d969c2c28a4651ff9097614f4 74428Author: Peter Hutterer <peter.hutterer@who-t.net> 74429Date: Thu May 10 12:42:59 2012 +1000 74430 74431 dix: disable non-sprite-owners first when disabling paired devices 74432 74433 If a sprite-owner is to be disabled but still paired, disable the paired 74434 device first. i.e. disabling a master pointer will disable the master 74435 keyboard first. 74436 74437 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74438 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74439 74440commit df1704365e700d3cf1d36a241bdfc479159a8df7 74441Author: Peter Hutterer <peter.hutterer@who-t.net> 74442Date: Thu May 10 15:33:15 2012 +1000 74443 74444 dix: free the sprite when disabling the device 74445 74446 Disabled devices don't need sprites (they can't send events anyway) and the 74447 device init process is currently geared to check for whether sprite is 74448 present to check if the device should be paired/attached. 74449 74450 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74451 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74452 74453commit e57d6a89027c55fef987cdc259668c48a8b4ea1b 74454Author: Peter Hutterer <peter.hutterer@who-t.net> 74455Date: Thu May 10 15:32:20 2012 +1000 74456 74457 dix: move freeing the sprite into a function 74458 74459 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74460 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74461 74462commit 46adcefb0e08515195d8e49985a4e210395700b3 74463Author: Peter Hutterer <peter.hutterer@who-t.net> 74464Date: Thu May 10 12:10:12 2012 +1000 74465 74466 dix: return early from DisableDevice if the device is already disabled 74467 74468 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74469 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74470 74471commit 076f9d3a6660fa7fef4ae9decaffd2feebdd32c4 74472Author: Peter Hutterer <peter.hutterer@who-t.net> 74473Date: Thu May 10 12:20:37 2012 +1000 74474 74475 dix: drop client argument from PairDevices - unused 74476 74477 This is a leftover from early MPX days where any keyboard could be paired 74478 with any pointer (before the device hierarchy). 74479 74480 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74481 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74482 74483commit d034605f3914e41e0d636eafff9278be1c041b2c 74484Author: Peter Hutterer <peter.hutterer@who-t.net> 74485Date: Thu May 10 14:13:20 2012 +1000 74486 74487 dix: rename shadowing loop variable in TouchResizeQueue 74488 74489 second 'i' shadows the function-wide one, rename to 'j' 74490 74491 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74492 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74493 74494commit a86f3ef511db16be5ddc723bf14e770dad23ee45 74495Author: Peter Hutterer <peter.hutterer@who-t.net> 74496Date: Thu May 10 11:55:21 2012 +1000 74497 74498 dix: change "still paired warning" to use BUG_RETURN_VAL 74499 74500 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74501 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74502 74503commit 4bcf43919a14cd2cd20af4099cd213ce25792edb 74504Author: Peter Hutterer <peter.hutterer@who-t.net> 74505Date: Thu May 10 15:32:47 2012 +1000 74506 74507 dix: use BUG_RETURN_VAL for pairing errors 74508 74509 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74510 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74511 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74512 74513commit 642569fc79a1814acca1c8f529539b054bf36907 74514Author: Peter Hutterer <peter.hutterer@who-t.net> 74515Date: Thu May 10 12:11:57 2012 +1000 74516 74517 Replace a few BUG_WARN with BUG_RETURN_VAL 74518 74519 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74520 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74521 74522commit 11ecfb8884b7c4def0863a57868a6b9fa1b63ae5 74523Author: Peter Hutterer <peter.hutterer@who-t.net> 74524Date: Thu May 10 11:48:05 2012 +1000 74525 74526 include: add BUG_RETURN_* macros 74527 74528 Helper functions to avoid things like 74529 74530 if (foo) { 74531 BUG_WARN(foo); 74532 return 1; 74533 } 74534 74535 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74536 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 74537 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74538 74539commit ff41753b1bee414b16c0f1e8d183776e87c94065 74540Author: Peter Hutterer <peter.hutterer@who-t.net> 74541Date: Wed Jun 6 14:38:49 2012 +1000 74542 74543 xkb: warn if XKB SlowKeys have been automatically enabled 74544 74545 Slow keys are enabled when the XKB AccessX features are generally enabled 74546 (ctrls->enabled_ctrls & XkbAccessXKeysMask) and either shift key is held for 74547 8 seconds. For the unsuspecting user this appears as if the keyboard 74548 suddenly stops working. 74549 74550 Print a warning to the log, so we can later tell them "told you so". 74551 74552 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 74553 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 74554 74555commit f20a532ba3c7b7d74530f3906c7bacf016bb8faf 74556Author: Alan Coopersmith <alan.coopersmith@oracle.com> 74557Date: Sun May 20 00:00:59 2012 -0700 74558 74559 Fix statement not reached warning in _DMXXineramaActive 74560 74561 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74562 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 74563 74564commit 1be317c4e5d99d11a124ad4e7a90a5dcae591e20 74565Author: Alan Coopersmith <alan.coopersmith@oracle.com> 74566Date: Wed May 23 19:27:23 2012 -0700 74567 74568 Provide prototypes for Mmio functions for Solaris Studio on SPARC 74569 74570 Actual inline implementations are provided via external *.il files 74571 generated from *.S files in hw/xfree86/os-support/solaris 74572 74573 Fixes missing prototype warnings that xorg-macros has recently elevated 74574 to build-breaking errors. 74575 74576 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74577 Signed-off-by: Henry Zhao <henry.zhao@oracle.com> 74578 Reviewed-by: Dave Airlie <airlied@redhat.com> 74579 74580commit 812786f4d4306cb16f8ed57fa4a1a32bb1d13fe3 74581Author: Dave Airlie <airlied@redhat.com> 74582Date: Tue Jun 5 19:55:52 2012 +0100 74583 74584 xwin: fixup block/wakeup handlers 74585 74586 These got missed out in the api changeover. 74587 74588 Reported-by: Colin Harrison <colin.harrison@virgin.net> 74589 Signed-off-by: Dave Airlie <airlied@redhat.com> 74590 Signed-off-by: Keith Packard <keithp@keithp.com> 74591 74592commit 3cbaf621782fb5f5679acf9104571d3ae2b48b72 74593Author: Dave Airlie <airlied@redhat.com> 74594Date: Tue Jun 5 17:54:02 2012 +0100 74595 74596 xfree86: fix new warnings introduced in rebase. 74597 74598 Spotted these in tinderbox: 74599 xf86xv.c: In function 'xf86XVScreenInit': 74600 xf86xv.c:282: warning: assignment from incompatible pointer type 74601 xf86xv.c: In function 'xf86XVCloseScreen': 74602 xf86xv.c:1275: warning: assignment from incompatible pointer type 74603 xf86xv.c: In function 'xf86XVAdjustFrame': 74604 xf86xv.c:1366: warning: assignment from incompatible pointer type 74605 74606 Missed removing one flags. 74607 74608 Signed-off-by: Dave Airlie <airlied@redhat.com> 74609 Tested-by: Keith Packard <keithp@keithp.com> 74610 Signed-off-by: Keith Packard <keithp@keithp.com> 74611 74612commit ec79187d9a23ba30eafc7ba225049559e6bcb78d 74613Author: Dave Airlie <airlied@redhat.com> 74614Date: Tue Jun 5 14:41:18 2012 +0100 74615 74616 modesetting: workaround kernel bug reporting 0x0 as valid mins 74617 74618 It reports these but then you can't create a 0 sized bo. 74619 74620 Signed-off-by: Dave Airlie <airlied@redhat.com> 74621 74622commit 1f0e8bd5eb1a5539689cfc4f5a6b86b530907ec5 74623Author: Dave Airlie <airlied@redhat.com> 74624Date: Tue Jun 5 13:22:18 2012 +0100 74625 74626 api: rework the X server driver API to avoid global arrays. 74627 74628 This is a squash merge containing all the API changes, as 74629 well as the video ABI bump. 74630 74631 Its been squashed to make bisection easier. 74632 74633 Full patch log below: 74634 74635 commit b202738bbf0c5a1c1172767119c2c71f1e7f8070 74636 Author: Aaron Plattner <aplattner@nvidia.com> 74637 Date: Mon May 14 15:16:11 2012 -0700 74638 74639 xfree86: Bump video ABI to 13.0 74640 74641 The ABI was broken by changes to convert from screen index numbers to ScreenPtr 74642 / ScrnInfoPtr in various structures and function signatures. 74643 74644 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 74645 Reviewed-by: Dave Airlie <airlied@redhat.com> 74646 Signed-off-by: Dave Airlie <airlied@redhat.com> 74647 74648 commit 3d5f7d9f8d408bcad3f83277d255f25d3b0edbf3 74649 Author: Dave Airlie <airlied@redhat.com> 74650 Date: Thu May 24 10:56:57 2012 +0100 74651 74652 xf86: xf86ClearEntityListForScreen should take a pScrn 74653 74654 When adding GPU screens this make life easier. 74655 74656 (also fix comment, as pointed out by Alan) 74657 74658 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74659 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 74660 Signed-off-by: Dave Airlie <airlied@redhat.com> 74661 74662 commit afee8b5ab4501597ecc1ade34124d7ca227ab055 74663 Author: Dave Airlie <airlied@redhat.com> 74664 Date: Thu May 24 07:07:32 2012 +0100 74665 74666 xf86i2c: add pscrn for drivers to use 74667 74668 This just adds a pScrn pointer into the struct for the drivers to use 74669 instead of scrnIndex. Mostly scrnIndex is used for logging, but some 74670 drivers use it to lookup xf86Screens, so let them stash a pScrn instead. 74671 74672 Removing the scrnIndex is a bit more involved and I'm not sure its worth 74673 the effort. Doing i2c in the X server is legacy code as far as I'm concerned. 74674 74675 Reviewed-by: Alex Deucher <alexander.deucher@amd.com> 74676 Signed-off-by: Dave Airlie <airlied@redhat.com> 74677 74678 commit ea5092f1f679691d187f1eee9427e6057beec56e 74679 Author: Dave Airlie <airlied@redhat.com> 74680 Date: Wed May 23 19:25:20 2012 +0100 74681 74682 dix/gc: consolidate GC object creation in one place 74683 74684 The standard GC create and scratch GC create were 90% the same really, 74685 and I have a need in the future for creating GC objects without the 74686 other bits, so wanted to avoid a third copy. 74687 74688 Reviewed-by: Keith Packard <keithp@keithp.com> 74689 Signed-off-by: Dave Airlie <airlied@redhat.com> 74690 74691 commit 3d91482ea9b4883e64e496f2768168e0ffa21ba1 74692 Author: Dave Airlie <airlied@redhat.com> 74693 Date: Wed May 23 10:24:06 2012 +0100 74694 74695 xf86: add a define to denote the new non-index interfaces are being used 74696 74697 This can be used by drivers to provide compatible APIs. 74698 74699 Reviewed-by: Keith Packard <keithp@keithp.com> 74700 Signed-off-by: Dave Airlie <airlied@redhat.com> 74701 74702 commit 37c3ae3e6cd4f3dedc72f371096d6743f8f99df3 74703 Author: Dave Airlie <airlied@redhat.com> 74704 Date: Wed May 23 15:09:12 2012 +0100 74705 74706 dix: make Create/Free scratch pixmaps take a ScreenPtr 74707 74708 While technically an API/ABI change I doubt anyone uses it, 74709 but it helps in splitting screens up. 74710 74711 Reviewed-by: Keith Packard <keithp@keithp.com> 74712 Signed-off-by: Dave Airlie <airlied@redhat.com> 74713 74714 commit 75f2062a3fe94f04764ecc7d2ff2fbbeccb9da60 74715 Author: Dave Airlie <airlied@redhat.com> 74716 Date: Wed May 23 14:57:55 2012 +0100 74717 74718 xf86/xv: remove scrnIndexfrom xf86FindXvOptions. 74719 74720 Move this interface to taking an ScrnInfoPtr. 74721 74722 Reviewed-by: Keith Packard <keithp@keithp.com> 74723 Signed-off-by: Dave Airlie <airlied@redhat.com> 74724 74725 commit f80c2374f40ea7b2ee0556e2e76cc07406f3d843 74726 Author: Dave Airlie <airlied@redhat.com> 74727 Date: Wed May 23 14:53:59 2012 +0100 74728 74729 xf86: make xf86DeleteScreen take a ScrnInfoPtr (v2) 74730 74731 stop passing indices into this function. 74732 74733 v2: drop flags argument. 74734 74735 Reviewed-by: Keith Packard <keithp@keithp.com> 74736 Signed-off-by: Dave Airlie <airlied@redhat.com> 74737 74738 commit 58824e414f35682435f15bfe6c4b656bd90b9235 74739 Author: Dave Airlie <airlied@redhat.com> 74740 Date: Wed May 23 14:48:09 2012 +0100 74741 74742 xf86: fix xf86IsScreenPrimary interface to take a pScrn (API/ABI) 74743 74744 Reviewed-by: Keith Packard <keithp@keithp.com> 74745 Signed-off-by: Dave Airlie <airlied@redhat.com> 74746 74747 commit 6b4fc1f9d391bcdf7ca288766e49bce60f4635cd 74748 Author: Dave Airlie <airlied@gmail.com> 74749 Date: Tue Apr 10 14:18:59 2012 +0100 74750 74751 xserver: convert block/wakeup handlers to passing ScreenPtr (ABI/API) (v2) 74752 74753 Instead of passing an index, pass the actual ScreenPtr. This allows 74754 more moving towards not abusing xf86Screens + screenInfo. 74755 74756 v2: drop the blockData/wakeupData args as per ajax's suggestion., 74757 fix docs. 74758 74759 Reviewed-by: Keith Packard <keithp@keithp.com> 74760 Signed-off-by: Dave Airlie <airlied@redhat.com> 74761 74762 commit 790d003de20fb47674420a24dadd92412d78620d 74763 Author: Dave Airlie <airlied@gmail.com> 74764 Date: Wed Apr 11 09:53:14 2012 +0100 74765 74766 xf86/common: remove some more pScrn->pScreen uses 74767 74768 remove some more conversions that appeared after api cleanups. 74769 74770 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74771 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74772 Reviewed-by: Adam Jackson <ajax@redhat.com> 74773 Signed-off-by: Dave Airlie <airlied@redhat.com> 74774 74775 commit aac85e18d1dd093f2cad6bd29375e40bd7af0b8f 74776 Author: Dave Airlie <airlied@gmail.com> 74777 Date: Tue Apr 10 16:34:53 2012 +0100 74778 74779 ddc: change API to take ScrnInfoPtr (v2) 74780 74781 This removes all xf86Screens usage from ddc code, 74782 it modifies the API for some functions to avoid taking indices. 74783 74784 v2: address Alan's comments about dropping DDC2Init parameter. 74785 74786 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74787 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74788 Reviewed-by: Adam Jackson <ajax@redhat.com> 74789 Signed-off-by: Dave Airlie <airlied@redhat.com> 74790 74791 commit fe3f57b6eaf6860a33876a54f9439f69578f03a5 74792 Author: Dave Airlie <airlied@gmail.com> 74793 Date: Tue Apr 10 16:31:26 2012 +0100 74794 74795 vbe: don't use index for VBEInterpretPanelID (API) 74796 74797 Remove use of xf86screens from vbe module. 74798 74799 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74800 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74801 Reviewed-by: Adam Jackson <ajax@redhat.com> 74802 Signed-off-by: Dave Airlie <airlied@redhat.com> 74803 74804 commit abf1965f4ed91529036d3fdb470d6a3ce6f29675 74805 Author: Dave Airlie <airlied@gmail.com> 74806 Date: Tue Apr 10 16:25:11 2012 +0100 74807 74808 int10/vbe: don't use xf86Screens. (ABI) (v3) 74809 74810 Pass the ScrnInfoPtr instead of the index in the int10 struct. 74811 74812 This saves us using it to dereference xf86Screens. 74813 74814 v2: address Alan's comment to fix struct alignment. 74815 74816 v3: squash in all the int10 fixes, test the vm86 code builds, 74817 after comments by Keith. 74818 74819 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74820 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74821 Reviewed-by: Adam Jackson <ajax@redhat.com> 74822 Signed-off-by: Dave Airlie <airlied@redhat.com> 74823 74824 commit 23cca612b4fb5efc33683c7624b803b457387e3d 74825 Author: Dave Airlie <airlied@gmail.com> 74826 Date: Tue Apr 10 15:30:18 2012 +0100 74827 74828 xserver: drop index argument to ScreenInit (ABI/API) (v2) 74829 74830 This drops the index argument, its the same as pScreen->myNum, 74831 and its the last major index abuse I can find. 74832 74833 v2: address Alan's review - update docs, fix xwin/xnest/darwin 74834 74835 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74836 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74837 Reviewed-by: Adam Jackson <ajax@redhat.com> 74838 Signed-off-by: Dave Airlie <airlied@redhat.com> 74839 74840 commit 40d360e2d7e832407f3ed64e3a02c27ecc89a960 74841 Author: Dave Airlie <airlied@gmail.com> 74842 Date: Tue Apr 10 15:23:01 2012 +0100 74843 74844 xf86: migrate PointerMoved from index to ScrnInfoPtr (ABI/API) 74845 74846 This migrates PointerMoved from an index to ScrnInfoPtr. 74847 74848 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74849 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74850 Reviewed-by: Adam Jackson <ajax@redhat.com> 74851 Signed-off-by: Dave Airlie <airlied@redhat.com> 74852 74853 commit aa60a2f38679d0eeb979a9c2648c9bc771409bf9 74854 Author: Dave Airlie <airlied@gmail.com> 74855 Date: Tue Apr 10 15:20:46 2012 +0100 74856 74857 xf86: migrate PMEvent to a ScrnInfoPtr (ABI/API) 74858 74859 This migrates the PMEvent from index to ScrnInfoPtr. 74860 74861 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74862 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74863 Reviewed-by: Adam Jackson <ajax@redhat.com> 74864 Signed-off-by: Dave Airlie <airlied@redhat.com> 74865 74866 commit d3f28ef44371ed4a039ffc5dd7eb6408d1269ba2 74867 Author: Dave Airlie <airlied@gmail.com> 74868 Date: Tue Apr 10 15:18:30 2012 +0100 74869 74870 xf86: migrate SetDGAMode from index to ScrnInfoPtr (ABI/API) 74871 74872 This migrates the SetDGAMode callback from an index to ScrnInfoPtr. 74873 74874 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74875 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74876 Reviewed-by: Adam Jackson <ajax@redhat.com> 74877 Signed-off-by: Dave Airlie <airlied@redhat.com> 74878 74879 commit baf5e4818a74f2b68c3dfdcc56f54322351039a0 74880 Author: Dave Airlie <airlied@gmail.com> 74881 Date: Tue Apr 10 15:14:11 2012 +0100 74882 74883 xf86: migrate ChangeGamma from index to ScrnInfoPtr (ABI/API) (v2) 74884 74885 This migrates the ChangeGamma interface to avoid passing a index. 74886 74887 v2: fix xf86RandR12.c + xf86cmap.c call 74888 74889 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74890 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74891 Reviewed-by: Adam Jackson <ajax@redhat.com> 74892 Signed-off-by: Dave Airlie <airlied@redhat.com> 74893 74894 commit 51e5f90ada929d6b23176090badbb42fdb3fa550 74895 Author: Dave Airlie <airlied@gmail.com> 74896 Date: Tue Apr 10 15:11:09 2012 +0100 74897 74898 xf86/exa: migrate index to screen types for EnableDisableFBAccess (ABI/API) 74899 74900 The EXA interface migrates to ScreenPtr, 74901 and the xf86 interface migrated to ScrnInfoPtr. 74902 74903 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74904 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74905 Reviewed-by: Adam Jackson <ajax@redhat.com> 74906 Signed-off-by: Dave Airlie <airlied@redhat.com> 74907 74908 commit 94f1f21d17e86f96d4a54292a399160950087675 74909 Author: Dave Airlie <airlied@gmail.com> 74910 Date: Tue Apr 10 15:02:11 2012 +0100 74911 74912 xf86: migrate ValidMode callback to ScrnInfoPtr (ABI/API) 74913 74914 This migrates the ValidMode to passing a ScrnInfoPtr instead 74915 of an index. 74916 74917 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74918 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74919 Reviewed-by: Adam Jackson <ajax@redhat.com> 74920 Signed-off-by: Dave Airlie <airlied@redhat.com> 74921 74922 commit 3f8f18198fed4f39ec805b508a3482e91eea26b2 74923 Author: Dave Airlie <airlied@gmail.com> 74924 Date: Tue Apr 10 14:59:46 2012 +0100 74925 74926 xf86: migrate SwitchMode to taking ScrnInfoPtr (ABI/API) (v2) 74927 74928 This migrate the SwitchMode interface to take a ScrnInfoPtr 74929 instead of an index. 74930 74931 v2: drop flags. 74932 74933 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74934 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74935 Reviewed-by: Adam Jackson <ajax@redhat.com> 74936 Signed-off-by: Dave Airlie <airlied@redhat.com> 74937 74938 commit d06a038a5c49328ab3a8d969d24f9fcd22c63202 74939 Author: Dave Airlie <airlied@gmail.com> 74940 Date: Tue Apr 10 14:50:37 2012 +0100 74941 74942 xf86: move AdjustFrame to passing ScrnInfoPtr (ABI/API) (v2) 74943 74944 This converts AdjustFrame code paths to passing a ScrnInfoPtr 74945 instead of an integer index. 74946 74947 v2: drop flags args. 74948 74949 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74950 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74951 Reviewed-by: Adam Jackson <ajax@redhat.com> 74952 Signed-off-by: Dave Airlie <airlied@redhat.com> 74953 74954 commit 53d2f8608ffd4090d08e7d5cf2e92fb954959b90 74955 Author: Dave Airlie <airlied@gmail.com> 74956 Date: Tue Apr 10 14:41:27 2012 +0100 74957 74958 xf86: modify FreeScreen callback to take pScrn instead of index. (ABI/API) (v2) 74959 74960 Another index->pScrn conversion. 74961 74962 v2: drop flags arg. 74963 74964 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74965 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74966 Reviewed-by: Adam Jackson <ajax@redhat.com> 74967 Signed-off-by: Dave Airlie <airlied@redhat.com> 74968 74969 commit 60db37c0b247052e0f5c54b1921fe58a3609c2e3 74970 Author: Dave Airlie <airlied@gmail.com> 74971 Date: Tue Apr 10 14:35:41 2012 +0100 74972 74973 xf86: change EnterVT/LeaveVT to take a ScrnInfoPtr (ABI/API break) (v2) 74974 74975 This modifies the EnterVT/LeaveVT interfaces to take a ScrnInfoPtr 74976 instead of an index into xf86Screens. This allows dropping more 74977 public dereferences of the xf86Screens and screenInfo. 74978 74979 v2: drop flags args as suggested by Keith, fix docs. 74980 74981 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74982 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74983 Reviewed-by: Adam Jackson <ajax@redhat.com> 74984 Signed-off-by: Dave Airlie <airlied@redhat.com> 74985 74986 commit 06729dbbc804a20242e6499f446acb5d94023c3c 74987 Author: Dave Airlie <airlied@gmail.com> 74988 Date: Tue Apr 10 14:04:59 2012 +0100 74989 74990 xserver: remove index from CloseScreen (API/ABI breakage) 74991 74992 This drops the index from the CloseScreen callback, 74993 its always been useless really, since the pScreen contains it. 74994 74995 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 74996 Acked-by: Aaron Plattner <aplattner@nvidia.com> 74997 Reviewed-by: Adam Jackson <ajax@redhat.com> 74998 Signed-off-by: Dave Airlie <airlied@redhat.com> 74999 75000commit 12f9aaff38ac2cf528387ce610e41fbd448698e2 75001Author: Chase Douglas <chase.douglas@canonical.com> 75002Date: Fri Apr 6 12:44:07 2012 -0700 75003 75004 test/touch: Initialize device name 75005 75006 Without this change, the test will segfault when we switch to signal- 75007 safe logging. 75008 75009 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 75010 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 75011 75012commit 3476eb38063473a7a5fcd78e2095e284118de839 75013Merge: 78f0d9cdc 42ae2e819 75014Author: Keith Packard <keithp@keithp.com> 75015Date: Sun Jun 3 20:24:24 2012 -0700 75016 75017 Merge remote-tracking branch 'whot/for-keith' 75018 75019commit 42ae2e8199fe875319978249963de7499607988b 75020Author: Siddhesh Poyarekar <siddhesh.poyarekar@gmail.com> 75021Date: Tue May 29 10:17:50 2012 +0530 75022 75023 xkb: Allocate size_syms correctly when width of a type increases 75024 75025 The current code seems to skip syms with width less than 75026 type->num_levels when calculating the total size for the new 75027 size_syms. This leads to less space being allocated than necessary 75028 during the next phase, which is to copy over the syms to the new 75029 location. This results in an overflow leading to a crash. 75030 75031 Signed-off-by: Siddhesh Poyarekar <siddhesh.poyarekar@gmail.com> 75032 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75033 75034commit 88c767edb01ed7efb19ffe3a453e16107b27130b 75035Author: Michal Suchanek <hramrach@gmail.com> 75036Date: Wed May 23 11:22:05 2012 +0200 75037 75038 Fix crash for motion events from devices without valuators 75039 75040 A WarpPointer request may trigger a motion event on a device without 75041 valuators. That request is ignored by GetPointerEvents but during smooth 75042 scroll emulation we dereference dev->valuators to get the number of axes. 75043 75044 Break out early if the device doesn't have valuators. 75045 75046 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75047 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 75048 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75049 75050commit e07c945ab992b1b1acdd9f5554e0688f744e4769 75051Author: Dave Airlie <airlied@redhat.com> 75052Date: Fri Jun 1 12:34:42 2012 +0100 75053 75054 modesetting: bump to latest X server compat api. 75055 75056 Signed-off-by: Dave Airlie <airlied@redhat.com> 75057 75058commit 78f0d9cdc441a575e15e5f7200ce74750d2f7ad2 75059Author: Chad Versace <chad.versace@linux.intel.com> 75060Date: Wed May 2 16:18:49 2012 -0700 75061 75062 dri2: Add DRI2GetParam request (v2) 75063 75064 Bump dri2proto dependency to 2.7. 75065 Bump DRI2INFOREC_VERSION to 7. 75066 75067 This new protocol request effectively allows clients to perform feature 75068 detection on the DDX. The request was added in DRI2 protocol 1.4. 75069 75070 If I had DRI2GetParam in June 2011, when I was implementing support in the 75071 Intel DDX and Mesa for new hardware that required a new DRI2 attachment 75072 format, then I could have avoided a week of pain caused by the necessity 75073 to write a horrid feature detection hack [1] in Mesa. In the future, when 75074 the work begins to add MSAA support to the Intel DDX, having a clean way 75075 to do feature detection will allow us to avoid revisiting and expanding 75076 that hack. 75077 75078 [1] mesa, commit aea2236a, function intel_verify_dri2_has_hiz 75079 75080 v2: If driver doesn't define ds->GetParam, dont' crash. Fall back to 75081 default behavior, per keithp. 75082 75083 Reviewed-by: Keith Packard <keithp@keithp.com> 75084 Reviewed-by: Eric Anholt <eric@anholt.net> (v1) 75085 Reviewed-by: Ian Romanick <idr@freedesktop.org> (v1) 75086 Signed-off-by: Chad Versace <chad.versace@linux.intel.com> 75087 75088commit 594b4a4f0601066d09e49a84c5af8c2eb6f07240 75089Author: Peter Hutterer <peter.hutterer@who-t.net> 75090Date: Thu May 24 13:34:11 2012 +1000 75091 75092 Xext: include dix-config.h 75093 75094 silences 75095 In file included from ../include/misc.h:111:0, from hashtable.c:2: 75096 ./include/os.h:559:1: warning: redundant redeclaration of 'strndup' [-Wredundant-decls] 75097 75098 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75099 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 75100 75101commit d348ab06aae21c153ecbc3511aeafc8ab66d8303 75102Author: Julien Cristau <jcristau@debian.org> 75103Date: Fri May 11 21:31:50 2012 +0200 75104 75105 Xvfb: destroy the screen pixmap in CloseScreen 75106 75107 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75108 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 75109 75110commit e4153c1d9138ed40de1c615525066a0f5bb599dc 75111Author: Julien Cristau <jcristau@debian.org> 75112Date: Fri May 11 21:31:49 2012 +0200 75113 75114 Xi: make stub DeleteInputDeviceRequest call RemoveDevice 75115 75116 DeleteInputDeviceRequest is called from CloseDownDevices on reset, so 75117 call RemoveDevice to avoid leaking devices in Xvfb/Xnest/Xwin. 75118 75119 Signed-off-by: Julien Cristau <jcristau@debian.org> 75120 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75121 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 75122 75123commit b840ba5f54de5b00a7700ca49e51308b8cc66f92 75124Author: Ian Romanick <ian.d.romanick@intel.com> 75125Date: Fri Dec 16 14:44:55 2011 -0800 75126 75127 glx: Implement protocol for glXCreateContextAttribsARB 75128 75129 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75130 Reviewed-by: Dave Airlie <airlied@redhat.com> 75131 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75132 75133commit 23612a63fc12be13442a63193a19f883d7a87e5d 75134Author: Ian Romanick <ian.d.romanick@intel.com> 75135Date: Fri Dec 16 14:42:51 2011 -0800 75136 75137 glx: Make several functions available outside the glxcmds.c compilation unit 75138 75139 validGlxScreen, validGlxFBConfig, validGlxContext, and 75140 __glXdirectContextCreate will soon be used by createcontext.c. 75141 75142 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75143 Reviewed-by: Dave Airlie <airlied@redhat.com> 75144 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75145 75146commit cd5689cac5ed722e9dc1a3fab9a260b798062ee6 75147Author: Ian Romanick <ian.d.romanick@intel.com> 75148Date: Sat Dec 10 15:15:18 2011 -0800 75149 75150 glx: Use one function to add a context to all global tables 75151 75152 Instead of having separate __glXAddContextToList and AddResource 75153 functions, just have one function that does both steps. 75154 75155 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75156 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 75157 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75158 75159commit c1d91ab37025d3e0df7e8f647fb21816fe0e8420 75160Author: Ian Romanick <ian.d.romanick@intel.com> 75161Date: Sat Dec 10 15:11:57 2011 -0800 75162 75163 glx: Initialize remaining context fields 75164 75165 There is no reason to assume the screen's context allocated 75166 initialized these fields, so don't. 75167 75168 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75169 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 75170 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75171 75172commit 0db76e5f771fb964e67de716a254850a38e02f19 75173Author: Ian Romanick <ian.d.romanick@intel.com> 75174Date: Sat Dec 10 15:09:57 2011 -0800 75175 75176 glx: Initialize all context fields together 75177 75178 v2: Fix whitespace error noticed by Christopher James Halse Rogers. 75179 75180 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75181 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 75182 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75183 75184commit befa76d251eb5de85edb621056eed52784722bc7 75185Author: Ian Romanick <ian.d.romanick@intel.com> 75186Date: Fri Dec 9 17:28:21 2011 -0800 75187 75188 glx: Implement GLX SetClientInfoARB protocol 75189 75190 v2: Bump glproto version to 1.4.15. This patch uses structure names 75191 that only exist in that glproto version and later. Noticed by 75192 Christopher James Halse Rogers. 75193 75194 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75195 Reviewed-by: Dave Airlie <airlied@redhat.com> 75196 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75197 75198commit 8b8cd6f6f4dbf9d058ace638221f31801da7df9f 75199Author: Ian Romanick <ian.d.romanick@intel.com> 75200Date: Mon Dec 5 13:52:01 2011 -0800 75201 75202 glx: Optionally call DRI2 createContextAttribs from __glXDRIscreenCreateContext 75203 75204 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75205 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75206 75207commit 0a9226376c9a44c3b50baad25f2b7d2b9fecb708 75208Author: Ian Romanick <ian.d.romanick@intel.com> 75209Date: Mon Dec 5 13:48:19 2011 -0800 75210 75211 glx: Add tracking for GLX_ARB_create_context and GLX_ARB_create_context_profile 75212 75213 The formatting of the extension data table was mangled by indent. 75214 This patch also undoes that and adds approripate control comments so 75215 that indent won't do it again. 75216 75217 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75218 Reviewed-by: Dave Airlie <airlied@redhat.com> 75219 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75220 75221commit 96d74138b1c0273e88933220d99a893858b649cd 75222Author: Ian Romanick <ian.d.romanick@intel.com> 75223Date: Mon Dec 5 10:56:07 2011 -0800 75224 75225 glx: Extend __GLXscreen::createContext to take attributes 75226 75227 The attributes will be used for glXCreateContextAttribsARB additions 75228 in follow-on patches. 75229 75230 v2: Add missing 'int *error' parameters noticed by Christopher James 75231 Halse Rogers. 75232 75233 v3: Remove redundant 'int err;' declaration noticed by Christopher 75234 James Halse Rogers. This was supposed to be in v2, but I missed it. 75235 75236 v4: Add comma missing from additions in v2. Ugh. 75237 75238 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75239 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 75240 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75241 75242commit 62f06b0dcde13cb87bc3d0c5ff86d8db7e3ff0ef 75243Author: Ian Romanick <ian.d.romanick@intel.com> 75244Date: Fri Dec 2 21:14:46 2011 -0800 75245 75246 glx: Don't track GLClientmajorVersion or GLClientminorVersion 75247 75248 Nothing uses these fields anywhere in the server. 75249 75250 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75251 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 75252 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75253 75254commit 724176a87afdb9a79ce6307214bc2e869530aef3 75255Author: Ian Romanick <ian.d.romanick@intel.com> 75256Date: Mon Dec 5 10:55:20 2011 -0800 75257 75258 glx: Fix mishandling of shared contexts 75259 75260 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 75261 Reviewed-by: Eric Anholt <eric@anholt.net> 75262 Reviewed-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 75263 75264commit 1c01090f6485cb82c2511fb461e66ff15a097358 75265Author: Dave Airlie <airlied@redhat.com> 75266Date: Wed May 23 11:31:39 2012 +0100 75267 75268 modesetting: fix make distcheck 75269 75270 add missing compat header file. 75271 75272 Signed-off-by: Dave Airlie <airlied@redhat.com> 75273 75274commit ef47a1b3bfec33d4031af2e01bdd46d634be2f51 75275Author: Dave Airlie <airlied@redhat.com> 75276Date: Wed May 23 11:21:55 2012 +0100 75277 75278 modesetting: convert to new scrn conversion APIs. 75279 75280 Generated with util/modular/x-driver-screen-scrn-conv.sh 75281 75282 Signed-off-by: Dave Airlie <airlied@redhat.com> 75283 75284commit 3e46d578d8c40c95343278db0a66c931e85247c7 75285Author: Dave Airlie <airlied@redhat.com> 75286Date: Wed May 23 11:21:39 2012 +0100 75287 75288 modesetting: add compat header file 75289 75290 Signed-off-by: Dave Airlie <airlied@redhat.com> 75291 75292commit 2d4fda4b09e67e47d3e6fc4743fc6e81bfe40f28 75293Author: Marcin Slusarz <marcin.slusarz@gmail.com> 75294Date: Mon May 21 21:39:43 2012 +0200 75295 75296 xfree86: fix mouse wheel support for DGA clients 75297 75298 xf86-input-evdev (since "smooth scrolling" support was added) can send mouse 75299 motion and wheel events in one batch, so we need to handle it properly. 75300 Otherwise mouse wheel events which come with motion events are lost 75301 and separate mouse wheel events are handled through non-DGA path. 75302 75303 Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com> 75304 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 75305 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75306 75307commit 9bc53d8cb04af2be3feeebb1b10774c2d599a76b 75308Author: Michal Srb <msrb@suse.com> 75309Date: Tue May 22 21:19:40 2012 -0700 75310 75311 dri2: SProcDRI2Connect - send the response. 75312 75313 The swapped implementation of DRI2Connect is always responding with empty 75314 device and driver values. However the response was only prepared and never 75315 sent (also had undefined .type member), causing e.g. glxinfo get stuck waiting 75316 for response when started remotely from machine with different endianity. 75317 75318 Signed-off-by: Michal Srb <msrb@suse.com> 75319 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 75320 Tested-by: Michel Dänzer <michel@daenzer.net> 75321 Signed-off-by: Keith Packard <keithp@keithp.com> 75322 75323commit 457fc77691b2df66f6dd3ba10ce7c1b61b9b42a3 75324Merge: 2e237c838 5a3a98fcb 75325Author: Keith Packard <keithp@keithp.com> 75326Date: Tue May 22 10:32:29 2012 -0700 75327 75328 Merge remote-tracking branch 'alanc/master' 75329 75330commit 2e237c838fe1780c2fcbd2306f35fd15b6747831 75331Author: Dave Airlie <airlied@gmail.com> 75332Date: Wed Apr 11 10:02:25 2012 +0100 75333 75334 xf86: reimplement XF86SCRNINFO macro using new functions. 75335 75336 This macro did lookups via privates but we can just use the ScreenToScrn 75337 conversion instead. 75338 75339 This patch drops all in-server uses, we should drop the macro later, 75340 once drivers have been converted to not use it. 75341 75342 Reviewed-by: Adam Jackson <ajax@redhat.com> 75343 Reviewed-by: Keith Packard <keithp@keithp.com> 75344 Signed-off-by: Dave Airlie <airlied@redhat.com> 75345 75346commit 8da35202f390cd8a0410814911017d99fa9dddcb 75347Author: Dave Airlie <airlied@gmail.com> 75348Date: Wed Apr 11 09:53:35 2012 +0100 75349 75350 xf86/modes: drop more pScrn->pScreen usages 75351 75352 Just simple conversion to use the lookup function. 75353 75354 Reviewed-by: Adam Jackson <ajax@redhat.com> 75355 Reviewed-by: Keith Packard <keithp@keithp.com> 75356 Signed-off-by: Dave Airlie <airlied@redhat.com> 75357 75358commit e7a9e30fd20a7c2a526783726d0a76b8a04b053a 75359Author: Dave Airlie <airlied@gmail.com> 75360Date: Wed Apr 11 09:52:37 2012 +0100 75361 75362 xaa: convert pScrn->pScreen to use accessor. 75363 75364 This uses a standard conversion function to do the conversion. 75365 75366 Reviewed-by: Adam Jackson <ajax@redhat.com> 75367 Reviewed-by: Keith Packard <keithp@keithp.com> 75368 Signed-off-by: Dave Airlie <airlied@redhat.com> 75369 75370commit 738367ac9b6179593a38074d0105de206cbed296 75371Author: Dave Airlie <airlied@gmail.com> 75372Date: Wed Apr 11 09:46:39 2012 +0100 75373 75374 xf86/modes: drop two uses of screenInfo 75375 75376 Just use new macros to access scrn->screen. 75377 75378 Reviewed-by: Adam Jackson <ajax@redhat.com> 75379 Reviewed-by: Keith Packard <keithp@keithp.com> 75380 Signed-off-by: Dave Airlie <airlied@redhat.com> 75381 75382commit cc4fe613d0b0891ea9502340801b37e64301f573 75383Author: Dave Airlie <airlied@redhat.com> 75384Date: Sat May 5 09:47:24 2012 +0100 75385 75386 render/exa: use glyph picture accessors 75387 75388 use the glyph picture accessors in the X server, render and EXA code. 75389 75390 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 75391 Reviewed-by: Keith Packard <keithp@keithp.com> 75392 Signed-off-by: Dave Airlie <airlied@redhat.com> 75393 75394commit a45e1d6cc6f2a5f31ed737294351cfa251caa030 75395Author: Dave Airlie <airlied@redhat.com> 75396Date: Sat May 5 09:38:28 2012 +0100 75397 75398 render: add GetGlyphPicture accessor. 75399 75400 This is a new API to stop the drivers directly looking up the glyph pictures 75401 in a global array. It provides a define GLYPH_HAS_GLYPH_PICTURE_ACCESSOR for 75402 drivers to work in a compat way. 75403 75404 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 75405 Reviewed-by: Keith Packard <keithp@keithp.com> 75406 Signed-off-by: Dave Airlie <airlied@redhat.com> 75407 75408commit 41151f88a62f01b00c1912822b8b430df1eb307e 75409Author: Dave Airlie <airlied@gmail.com> 75410Date: Tue Apr 10 15:47:33 2012 +0100 75411 75412 xf86: migrate to using xf86ScreenToScrn wrapper (v2) 75413 75414 migrate to new helper API. 75415 75416 This just wraps all the obvious uses of xf86Screens[pScreen->myNum], 75417 and should be fairly simple to review. 75418 75419 v2: remove commented out lines. 75420 75421 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 75422 Reviewed-by: Keith Packard <keithp@keithp.com> 75423 Signed-off-by: Dave Airlie <airlied@redhat.com> 75424 75425commit 53932b3803fa2c02949fd7d4d0e433ea58fa89f1 75426Author: Dave Airlie <airlied@gmail.com> 75427Date: Tue Apr 10 15:47:32 2012 +0100 75428 75429 xf86: add helper functions to convert to from ScrnInfoPtr/ScreenPtr (v2) 75430 75431 These are just simple functions that we should start migrating drivers 75432 to using. 75433 75434 The end goal is to remove xf86Screens and screenInfo from the ABI. 75435 75436 This includes a define XF86_HAS_SCRN_CONV that drivers can ifdef to provide 75437 their own copies. I'll probably post a generic compat.h file for drivers later. 75438 75439 v2: add asserts. 75440 75441 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 75442 Reviewed-by: Keith Packard <keithp@keithp.com> 75443 Signed-off-by: Dave Airlie <airlied@redhat.com> 75444 75445commit 39f73e813f7d404498629f6104a9003d092af28d 75446Author: Dave Airlie <airlied@redhat.com> 75447Date: Tue May 8 13:01:12 2012 +0100 75448 75449 xf86/pci: fix slot claiming counting. 75450 75451 Currently if we claim a slot for a PCI driver, we never let it go properly, 75452 this prevents the fallback probe from reusing the slot, even though it 75453 isn't claimed for that pci slot. 75454 75455 So if you set the modesetting driver to point at a specific kms device, 75456 that isn't a PCI device (i.e. USB dongle), then the modesetting driver 75457 loads, the pci probe tries to bind the config slot to the primary PCI 75458 device, however we then check the kms device bus id to discover it 75459 isn't valid. However we don't remove the claim on the slot. Next the 75460 old probe function is called and there is no slots to claim. 75461 75462 This patch fixes that and converts the pciSlotClaimed boolean into 75463 a counter, and changes the unclaim api to take a device pointer 75464 to remove from the entity. 75465 75466 Reviewed-by: Adam Jackson <ajax@redhat.com> 75467 Signed-off-by: Dave Airlie <airlied@redhat.com> 75468 75469commit 07dcc3f1a98dff2ee374a15ac5beac778d2ccc97 75470Author: Dave Airlie <airlied@redhat.com> 75471Date: Sun May 6 17:13:02 2012 +0100 75472 75473 config/udev: add pre_init stage to config and udev. 75474 75475 In order to use udev for gpu enumeration, we need to init udev earlier 75476 than input initialisations. This splits the config init stuff so that udev 75477 pre init sets up before output initialisation. 75478 75479 this is just a prepatory patch, doesn't change anything major. 75480 75481 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 75482 Signed-off-by: Dave Airlie <airlied@redhat.com> 75483 75484commit a2a02882ab65133e6c0c69db1f38bc20b406236f 75485Author: Dave Airlie <airlied@redhat.com> 75486Date: Sun May 6 17:35:34 2012 +0100 75487 75488 xfree86: add modesetting driver to fallback list on Linux 75489 75490 Add the modesetting driver to the fallback list on Linux, after vesa 75491 before fbdev. 75492 75493 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 75494 Signed-off-by: Dave Airlie <airlied@redhat.com> 75495 75496commit 5a3a98fcb799c2ac8fa7494645ad9030f1cac837 75497Author: Alan Coopersmith <alan.coopersmith@oracle.com> 75498Date: Fri Apr 27 13:35:32 2012 -0700 75499 75500 Undocument Font Module loading 75501 75502 Code was deleted in commit affec10635343668839994ea2bac16c1d4524200 75503 75504 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 75505 Reviewed-by: Dave Airlie <airlied@redhat.com> 75506 Reviewed-by: Adam Jackson <ajax@redhat.com> 75507 75508commit afcb7ba24ebd9968d48a192df438e7f1c0ecb770 75509Author: Alan Coopersmith <alan.coopersmith@oracle.com> 75510Date: Fri Apr 27 13:31:56 2012 -0700 75511 75512 Undocument mandatory loadable modules 75513 75514 The code to implement was deleted when BaseModules[] was emptied by 75515 the replacement of the "pcidata" module with libpciaccess calls 75516 in commit 46f55f5dead5d70cdff30531d80a72f6be042315. 75517 75518 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 75519 Reviewed-by: Dave Airlie <airlied@redhat.com> 75520 Reviewed-by: Adam Jackson <ajax@redhat.com> 75521 75522commit c3180a74a4fc6304bcfe676e3285d707874a9fea 75523Author: Alan Coopersmith <alan.coopersmith@oracle.com> 75524Date: Fri Apr 6 21:51:37 2012 -0700 75525 75526 cvt man page should use Hz, not kHz, for vertical refresh rate 75527 75528 https://bugs.freedesktop.org/show_bug.cgi?id=48311 75529 75530 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 75531 Reviewed-by: Julien Cristau <jcristau@debian.org> 75532 75533commit 96e0ab54962ab59f03275a098ee02c1f0f86b94e 75534Author: Alan Coopersmith <alan.coopersmith@oracle.com> 75535Date: Thu Jan 26 15:44:40 2012 -0800 75536 75537 Convert sbusPaletteKey to latest DevPrivate API 75538 75539 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 75540 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 75541 Reviewed-by: Dave Airlie <airlied@redhat.com> 75542 75543commit ba883a0f3435d5da82a8134e696c4905eea70f23 75544Merge: ae1c48ebc acdc4f54e 75545Author: Keith Packard <keithp@keithp.com> 75546Date: Thu May 17 16:49:19 2012 -0700 75547 75548 Merge remote-tracking branch 'jeremyhu/master' 75549 75550commit acdc4f54ee5784d26c251ff8c941b4360ae80e9b 75551Author: Jeremy Huddleston <jeremyhu@apple.com> 75552Date: Thu May 17 10:08:02 2012 -0700 75553 75554 XQuartz: Avoid a race in initialization of darwinPointer 75555 75556 http://xquartz.macosforge.org/trac/ticket/579 75557 75558 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 75559 75560commit c495211373a118acee02a29c6eb25a332abc856c 75561Author: Jeremy Huddleston <jeremyhu@apple.com> 75562Date: Thu May 17 16:20:30 2012 -0700 75563 75564 XQuartz: Declare noPseudoramiXExtension for miinitext.c 75565 75566 Regression-from: 6699f54fe0b3f4952a75e46fb6c3a643988e65dd 75567 75568 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 75569 75570commit f5fc5334116ad2b0c8930ac68c52bf0f7a2f1417 75571Author: Jeremy Huddleston <jeremyhu@apple.com> 75572Date: Mon May 7 17:07:27 2012 -0700 75573 75574 XQuartz: Provide fls implementation for Tiger 75575 75576 Regression-from: 30623d6ff7bca223f9e1a825e86bd317eef8bf16 75577 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 75578 75579commit e0097ce9719f9b0aeccf7102630c0299f8acdf93 75580Author: Jeremy Huddleston <jeremyhu@apple.com> 75581Date: Mon May 7 16:31:38 2012 -0700 75582 75583 XQuartz: Tiger build fix 75584 75585 http://trac.macports.org/ticket/34364 75586 75587 Regression-from: 662d41acdde1dcb9774fbe4054e251c708acaffe 75588 75589 Reported-by: Peter Dyballa <Peter_Dyballa@Freenet.DE> 75590 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 75591 75592commit 4acbaa3027d89728ed932616f6ef66d93537b101 75593Author: Jeremy Huddleston <jeremyhu@apple.com> 75594Date: Mon May 7 16:16:20 2012 -0700 75595 75596 XQuartz: Workaround an SDK bug on Leopard/x86_64 75597 75598 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 75599 75600commit ae1c48ebc8a3c0b32b3d732e8a340bfe00bb9f79 75601Author: Peter Hutterer <peter.hutterer@who-t.net> 75602Date: Wed May 16 10:54:48 2012 +1000 75603 75604 Xi: fix "discards ''const' qualifier" warnings 75605 75606 extinit.c: In function 'XInputExtensionInit': 75607 extinit.c:1301:29: warning: assignment discards 'const' qualifier from 75608 pointer target type [enabled by default] 75609 extinit.c:1303:36: warning: assignment discards 'const' qualifier from 75610 pointer target type [enabled by default] 75611 75612 property.c: In function 'XIChangeDeviceProperty': 75613 xiproperty.c:757:39: warning: cast discards '__attribute__((const))' 75614 qualifier from pointer target type [-Wcast-qual] 75615 75616 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75617 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 75618 75619commit e42bf98283ddc4eac2213c96e8ebcd711583e6b1 75620Author: Peter Hutterer <peter.hutterer@who-t.net> 75621Date: Wed May 16 10:47:45 2012 +1000 75622 75623 Xi: fix "shadows previous local" warnings 75624 75625 exevents.c: In function 'DeepCopyFeedbackClasses': 75626 exevents.c:272:20: warning: declaration of 'classes' shadows a previous 75627 local [-Wshadow] 75628 exevents.c:245:16: warning: shadowed declaration is here [-Wshadow] 75629 (and a few more like this) 75630 75631 exevents.c: In function 'DeliverTouchEmulatedEvent': 75632 exevents.c:1442:27: warning: declaration of 'win' shadows a parameter 75633 [-Wshadow] 75634 exevents.c:1404:55: warning: shadowed declaration is here [-Wshadow] 75635 exevents.c:1475:28: warning: declaration of 'listener' shadows a parameter 75636 [-Wshadow] 75637 exevents.c:1403:62: warning: shadowed declaration is here [-Wshadow] 75638 75639 xiselectev.c: In function 'ProcXISelectEvents': 75640 xiselectev.c:178:34: warning: declaration of 'dummy' shadows a previous 75641 local [-Wshadow] 75642 xiselectev.c:91:18: warning: shadowed declaration is here [-Wshadow] 75643 75644 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75645 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 75646 75647commit ca6d25ad2d1fb4efae9d40c7e74c8494f9918625 75648Author: Peter Hutterer <peter.hutterer@who-t.net> 75649Date: Wed May 16 10:10:11 2012 +1000 75650 75651 dix: Move DeviceFocusEvent from Xi into enterleave.c 75652 75653 This is only called from the enterleave implementation, so move it and its 75654 helper functions to there. No functional changes. 75655 75656 Fixes build error introduced in 31174565ec0090b4c03c9334c82878be2455f938 if 75657 building with '-Werror=implicit-function-declaration' 75658 75659 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75660 Reviewed-by: Dave Airlie <airlied@redhat.com> 75661 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 75662 75663commit 749a593e49adccdf1225be28a521412ec85333f4 75664Author: Peter Hutterer <peter.hutterer@who-t.net> 75665Date: Wed May 9 11:30:46 2012 +1000 75666 75667 dix: undo transformation for missing valuators (#49347) 75668 75669 last.valuators contains the transformed valuators of the device. If the 75670 device submits events with x/y missing, we need to get that from 75671 last.valuators and undo the transformation to that axis. 75672 75673 X.Org Bug 49347 <http://bugs.freedesktop.org/show_bug.cgi?id=49347> 75674 75675 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75676 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 75677 75678commit 03318835a5a87a8a5cb3515130b5380d565c0d91 75679Author: Peter Hutterer <peter.hutterer@who-t.net> 75680Date: Mon May 14 15:06:56 2012 +1000 75681 75682 include: don't redeclare device events in xkbsrv.h 75683 75684 ../../include/xkbsrv.h:308:51: warning: redundant redeclaration of 75685 ‘DeviceKeyPress’ [-Wredundant-decls] 75686 75687 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75688 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 75689 75690commit 43754803c13c232fd5c37eb3d4b1ca2898d6a6cd 75691Author: Peter Hutterer <peter.hutterer@who-t.net> 75692Date: Mon May 14 15:07:57 2012 +1000 75693 75694 test: don't shadow parameter 'len' 75695 75696 protocol-xiquerydevice.c:226:25: warning: declaration of ‘len’ shadows a 75697 parameter [-Wshadow] 75698 75699 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75700 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 75701 75702commit 1bafc9f26bef0997c98b97efa888ed5e2a54f6f0 75703Author: Peter Hutterer <peter.hutterer@who-t.net> 75704Date: Mon May 14 15:01:09 2012 +1000 75705 75706 test: fix redundant declaration of "BadDevice" 75707 75708 In file included from protocol-common.c:36:0: 75709 protocol-common.h:36:12: warning: redundant redeclaration of ‘BadDevice’ 75710 [-Wredundant-decls] 75711 In file included from protocol-common.c:30:0: 75712 ../../Xi/exglobals.h:41:12: note: previous declaration of ‘BadDevice’ was 75713 here 75714 75715 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75716 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 75717 75718commit 47e1d0ea43eb7a6caf774f375f11849d0d515e14 75719Author: Peter Hutterer <peter.hutterer@who-t.net> 75720Date: Mon May 14 14:56:02 2012 +1000 75721 75722 test: fix "userdata shadows global declaration" warnings 75723 75724 protocol-xiquerypointer.c:124:72: warning: declaration of 75725 ‘userdata’ shadows a global declaration [-Wshadow] 75726 75727 and similar 75728 75729 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75730 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 75731 75732commit e6da635826bdda7f67bd492dda8dba8a97809213 75733Author: Peter Hutterer <peter.hutterer@who-t.net> 75734Date: Mon May 14 14:51:58 2012 +1000 75735 75736 test: fix "redundant declaration of devices" warning 75737 75738 In file included from protocol-xiwarppointer.c:41:0: 75739 protocol-common.h:91:23: warning: redundant redeclaration of ‘devices’ 75740 [-Wredundant-decls] 75741 protocol-common.h:86:3: note: previous declaration of ‘devices’ was here 75742 75743 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75744 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 75745 75746commit 693a31e704c0745b209e69711522417c4fa8541e 75747Author: Simon Thum <simon.thum@gmx.de> 75748Date: Mon Apr 2 18:49:53 2012 +0200 75749 75750 dix: indentation fixes for pointer acceleration 75751 75752 Signed-off-by: Simon Thum <simon.thum@gmx.de> 75753 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 75754 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 75755 75756commit 31174565ec0090b4c03c9334c82878be2455f938 75757Author: Michal Suchanek <hramrach@gmail.com> 75758Date: Wed Apr 25 15:03:31 2012 +0200 75759 75760 dix: Remove redundant declarations. 75761 75762 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75763 Reviewed-by: Dave Airlie <airlied@redhat.com> 75764 Signed-off-by: Dave Airlie <airlied@redhat.com> 75765 75766commit f1cec791d1def506b99b0e6c160e9e8be107bd13 75767Author: Michal Suchanek <hramrach@gmail.com> 75768Date: Tue Apr 24 20:08:02 2012 +0200 75769 75770 xfree86: Add ifdef wrapper to xvmodproc.h 75771 75772 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75773 Reviewed-by: Dave Airlie <airlied@redhat.com> 75774 Signed-off-by: Dave Airlie <airlied@redhat.com> 75775 75776commit d2f71b13946a14d9698d0355cc45b114b236e73b 75777Author: Michal Suchanek <hramrach@gmail.com> 75778Date: Tue Apr 24 18:44:48 2012 +0200 75779 75780 Xi: Remove redundant declaration. 75781 75782 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75783 Reviewed-by: Dave Airlie <airlied@redhat.com> 75784 Signed-off-by: Dave Airlie <airlied@redhat.com> 75785 75786commit a4f87f885a55373a6daf56fd540e10c5f1db5a5c 75787Author: Michal Suchanek <hramrach@gmail.com> 75788Date: Tue Apr 24 18:43:31 2012 +0200 75789 75790 kdrive: Remove redundant declarations. 75791 75792 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75793 Reviewed-by: Dave Airlie <airlied@redhat.com> 75794 Signed-off-by: Dave Airlie <airlied@redhat.com> 75795 75796commit 369edd78769fa1a174e4091d0fa0722350df5d68 75797Author: Michal Suchanek <hramrach@gmail.com> 75798Date: Tue Apr 24 18:41:07 2012 +0200 75799 75800 dix: don't duplicate DoFocusEvents declaration. 75801 75802 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75803 Reviewed-by: Dave Airlie <airlied@redhat.com> 75804 Signed-off-by: Dave Airlie <airlied@redhat.com> 75805 75806commit 36377fb0e88a6d237d98805e802ab3793582c976 75807Author: Michal Suchanek <hramrach@gmail.com> 75808Date: Tue Apr 24 18:37:30 2012 +0200 75809 75810 dix: Remove redundant declarations. 75811 75812 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75813 Reviewed-by: Dave Airlie <airlied@redhat.com> 75814 Signed-off-by: Dave Airlie <airlied@redhat.com> 75815 75816commit ea9ed83f8f0da241099d0a11c3460d46c5114dea 75817Author: Michal Suchanek <hramrach@gmail.com> 75818Date: Tue Apr 24 18:32:53 2012 +0200 75819 75820 xfree86: don't include xvmodproc.h when not needed 75821 75822 The functions are already declared in xf86xv.h 75823 75824 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75825 Reviewed-by: Dave Airlie <airlied@redhat.com> 75826 Signed-off-by: Dave Airlie <airlied@redhat.com> 75827 75828commit 1927e5ab25c9c9bf481e75ac50f4b140a96e634a 75829Author: Michal Suchanek <hramrach@gmail.com> 75830Date: Tue Apr 24 18:24:19 2012 +0200 75831 75832 xserver: Remove duplicate declaration in resource.h 75833 75834 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75835 Reviewed-by: Dave Airlie <airlied@redhat.com> 75836 Signed-off-by: Dave Airlie <airlied@redhat.com> 75837 75838commit 1339681e4fbd4f5981e6246eedb23e8a9c5d10f4 75839Author: Michal Suchanek <hramrach@gmail.com> 75840Date: Tue Apr 24 18:19:43 2012 +0200 75841 75842 miext: Remove redundant declaration. 75843 75844 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75845 Reviewed-by: Dave Airlie <airlied@redhat.com> 75846 Signed-off-by: Dave Airlie <airlied@redhat.com> 75847 75848commit 6699f54fe0b3f4952a75e46fb6c3a643988e65dd 75849Author: Michal Suchanek <hramrach@gmail.com> 75850Date: Tue Apr 24 18:18:43 2012 +0200 75851 75852 mi: Remove redundant declarations. 75853 75854 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75855 Reviewed-by: Dave Airlie <airlied@redhat.com> 75856 Signed-off-by: Dave Airlie <airlied@redhat.com> 75857 75858commit b4bf90f9fb20e0e56d2451096d4f62ecfc24a962 75859Author: Michal Suchanek <hramrach@gmail.com> 75860Date: Tue Apr 24 18:16:30 2012 +0200 75861 75862 exa: Remove redundant declaration. 75863 75864 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75865 Reviewed-by: Dave Airlie <airlied@redhat.com> 75866 Signed-off-by: Dave Airlie <airlied@redhat.com> 75867 75868commit 5595e7ce0735db6f82e1fe9df45434826019bf9e 75869Author: Michal Suchanek <hramrach@gmail.com> 75870Date: Tue Apr 24 18:15:17 2012 +0200 75871 75872 randr: Remove redundant declaration. 75873 75874 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75875 Reviewed-by: Dave Airlie <airlied@redhat.com> 75876 Signed-off-by: Dave Airlie <airlied@redhat.com> 75877 75878commit c59c9dac84301f8eb2a0e4a09a8839146aa5a069 75879Author: Michal Suchanek <hramrach@gmail.com> 75880Date: Tue Apr 24 18:13:22 2012 +0200 75881 75882 xkb: Remove redundant declarations. 75883 75884 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75885 Reviewed-by: Dave Airlie <airlied@redhat.com> 75886 Signed-off-by: Dave Airlie <airlied@redhat.com> 75887 75888commit d0a83460451df1a4c5667b2f414c35ee81b4ffa4 75889Author: Michal Suchanek <hramrach@gmail.com> 75890Date: Tue Apr 24 18:11:16 2012 +0200 75891 75892 xfixes: remove redundant declaration. 75893 75894 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 75895 Reviewed-by: Dave Airlie <airlied@redhat.com> 75896 Signed-off-by: Dave Airlie <airlied@redhat.com> 75897 75898commit 3a94b338aaa48b656aae1bb3a5a9d64300c0a093 75899Merge: afc153a5b c91d00e0f 75900Author: Keith Packard <keithp@keithp.com> 75901Date: Wed May 9 21:17:16 2012 -0700 75902 75903 Merge remote-tracking branch 'whot/for-keith' 75904 75905commit e8da587845978ad66c3b3e2e085fd99c16b5d3c8 75906Author: Dave Airlie <airlied@redhat.com> 75907Date: Wed May 9 09:33:54 2012 +0100 75908 75909 bump version to 0.3.0 75910 75911commit 62543946490e7decae4f6549d83072f004e850d4 75912Author: Matt Turner <mattst88@gmail.com> 75913Date: Wed May 9 00:26:29 2012 -0400 75914 75915 configure.ac: remove DRI and RENDER 75916 75917 Driver doesn't support either of these. 75918 75919 Signed-off-by: Matt Turner <mattst88@gmail.com> 75920 Signed-off-by: Dave Airlie <airlied@redhat.com> 75921 75922commit 7dc22b7911ac3d5c131075903e3fbf52c58eac15 75923Author: Dave Airlie <airlied@redhat.com> 75924Date: Wed May 9 09:32:05 2012 +0100 75925 75926 modesetting: make sure the pci device corresponds to the drm device 75927 75928 If we get asked to pci open a device with a kms path override, 75929 make sure they match, otherwise this driver can steal the primary 75930 device binding for a usb adaptor. 75931 75932 The driver should fallback to the old probe entry point in this case. 75933 75934 Signed-off-by: Dave Airlie <airlied@redhat.com> 75935 75936commit afc153a5b4fc58ae70dc214f61a71b1a8c855f06 75937Author: James Cloos <cloos@jhcloos.com> 75938Date: Tue May 8 17:55:10 2012 -0400 75939 75940 Fix RANDR’s gamma_to_ramp(). 75941 75942 In order to generate a 256-entry ramp in [0,65535] which covers the full 75943 range, one must mupliply eight-bit values not by 256 but rather by 257. 75944 75945 Many years back – well before the RANDR extension was written, and 75946 before xorg@fdo – a similar bug fix was made to the DIX for converting 75947 client-supplied eight-bit color values into sixteen-bit values. 75948 75949 Noticed by: Elle Stone and Graeme Gill. 75950 75951 Signed-off-by: James Cloos <cloos@jhcloos.com> 75952 75953commit e501c34d4937d5e6f19abd29f1ec7f95faa3bb55 75954Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 75955Date: Tue May 1 22:03:24 2012 -0500 75956 75957 hw/xwin/glx: Use silent rules 75958 75959 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 75960 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 75961 75962commit 026d402fef623a0bda0bedd9c0938ba787bf9829 75963Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 75964Date: Fri Mar 30 14:33:18 2012 -0500 75965 75966 xfree86: use silent rules with sdksyms generation 75967 75968 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 75969 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 75970 75971commit 3d98dac46c17bc6e03546b8a532bfb55f00e5084 75972Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 75973Date: Sun Apr 29 20:45:17 2012 -0500 75974 75975 xfree86: respect EXEEXT in relink target 75976 75977 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 75978 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 75979 75980commit e6461db6dcb437cfa7d4b23008c2c12f8169ff79 75981Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 75982Date: Thu Apr 26 01:49:06 2012 -0500 75983 75984 os: Add CryptoAPI as a choice of SHA1 implementation 75985 75986 Both Cygwin and MinGW can use Windows' native CryptoAPI for SHA1, 75987 saving a dependency on libgcrypt or OpenSSL. The necessary functions 75988 are in ADVAPI32.DLL, which is among the default lib flags and is 75989 already used in hw/xwin for accessing the registry. 75990 75991 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 75992 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 75993 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 75994 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 75995 Tested-by: Colin Harrison <colin.harrison@virgin.net> 75996 75997commit c91d00e0f330b9de604068e1bfcb0a307096434f 75998Author: Daniel Kurtz <djkurtz@chromium.org> 75999Date: Wed Apr 18 17:51:53 2012 +0800 76000 76001 os/log: refactor logging 76002 76003 It is not safe to ever use an arbitrary (possibly user supplied) string as 76004 part of the format for a *sprintf() call. 76005 76006 For example: 76007 1. Name a Bluetooth keyboard "%n%n%n%n%n%n%n%n" 76008 2. Pair it with a computer running X and try to use it 76009 3. X is not happy when trying to do the following in xf86-input-evdev: 76010 xf86IDrvMsg(pInfo, X_CONFIG, "Device: \"%s\"\n", device); 76011 because LogVHdrMessageVerb() has put the %n from the device name 76012 into a format string of the form: 76013 "evdev: %n%n%n%n%n%n%n%n: Device: \"%s\"\n" 76014 76015 Instead, build up a log message in place by appending successive formatted 76016 strings by sncprintf'ing to the end of the previous. 76017 76018 Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> 76019 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76020 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76021 76022commit 6ce0eac4f8a05f6d7401445cab95027709d3a479 76023Author: Daniel Kurtz <djkurtz@chromium.org> 76024Date: Wed Apr 18 17:51:52 2012 +0800 76025 76026 os/log: only write timestamp if a message is actually written to logfile 76027 76028 The current code will write a timestamps into the logFile whenever 76029 the last message ended with a '\n' - even if the verb for that timestamp 76030 is at too high a level. This timestamp will sit there with no matching 76031 message until the next call to LogVWrite with a valid verb. 76032 76033 In other words, in some cases, timestamps in the X.org.log are for some 76034 completely unrelated message that was previously ignored due to 76035 insufficient verbosity, and not for the message that appears next to it 76036 in the log file. 76037 76038 We keep the current policy which appears to be to only apply timestamps if 76039 a message is actually written to a log file. That is, no timestamps on 76040 stderr, or in the mem buffer. Therefore, the timestamp stringification 76041 is moved to the conditional where it is used. 76042 76043 Since logging uses a fixed length buffer, this patch also forces a '\n' 76044 whenever a buffer is terminated due to a too-long write request. This 76045 allows the newline detection to work even on overflow, and also cleans up 76046 the log a bit in the overflow case. 76047 76048 Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> 76049 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76050 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76051 76052commit 5c2e2a164d615ab06be28a663734e782614b5cc7 76053Author: Daniel Kurtz <djkurtz@chromium.org> 76054Date: Wed Apr 18 09:51:51 2012 +0000 76055 76056 os/xprintf: add Xvscnprintf and Xscnprintf 76057 76058 Normal snprintf() usually returns the number of bytes that would have been 76059 written into a buffer had the buffer been long enough. 76060 76061 The scnprintf() variants return the actual number of bytes written, 76062 excluding the trailing '\0'. 76063 76064 Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> 76065 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76066 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76067 76068commit c30862879d2c766519780bb7f353f35edf0daa9b 76069Author: Daniel Kurtz <djkurtz@chromium.org> 76070Date: Wed Apr 18 17:51:50 2012 +0800 76071 76072 os/log: trivial cleanups 76073 76074 * space->tab 76075 * remove comment that doesn't make any sense 76076 76077 Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> 76078 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76079 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76080 76081commit 97041364a6acb2b66b5cfd06757c90a006ad50e9 76082Merge: 19082726c ee542b855 76083Author: Keith Packard <keithp@keithp.com> 76084Date: Wed May 2 20:47:25 2012 -0700 76085 76086 Merge remote-tracking branch 'whot/for-keith' 76087 76088 No conflicts here 76089 76090commit ee542b85590814ee25369babce1ad14feeb137af 76091Author: Chase Douglas <chase.douglas@canonical.com> 76092Date: Tue May 1 10:21:12 2012 -0700 76093 76094 Report touch emulated buttons in XIQueryPointer for XI 2.1 and earlier 76095 76096 XInput 2.1 and earlier clients do not know about touches. We must report 76097 touch emulated button presses for these clients. For later clients, we 76098 only report true pointer button presses. 76099 76100 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76101 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76102 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76103 76104commit 19082726cb2c69f53e9720904521e3c98f788bff 76105Author: Ryan Pavlik <rpavlik@iastate.edu> 76106Date: Fri Oct 21 12:47:38 2011 -0500 76107 76108 configure.ac: on MinGW, link with ws2_32 instead of winsock2 76109 76110 ws2_32 is the correct name for the libary (even on 64 bit Windows :-)) 76111 76112 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 76113 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 76114 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 76115 76116commit ad9605a3e21b20fa8565dc6701610712ac6dd9e1 76117Author: Ryan Pavlik <rpavlik@iastate.edu> 76118Date: Fri Nov 4 13:26:14 2011 -0500 76119 76120 configure.ac: MinGW doesn't have setuid binaries either. 76121 76122 If the target platform isn't in a list of platforms we know don't use a 76123 setuid binary, we try to test if we can chown something to root. 76124 76125 This test possibly won't give the right answer if we are cross-compiling, 76126 which is common for the MinGW target. This patch adds MinGW to the list 76127 of platforms we know don't use a setuid binary. 76128 76129 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 76130 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 76131 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 76132 76133commit 2225208c2ef9857f14813376ce85d305f19263dd 76134Author: Ryan Pavlik <rpavlik@iastate.edu> 76135Date: Fri Nov 4 13:30:43 2011 -0500 76136 76137 configure.ac: auto-disable MITSHM if we lack IPC 76138 76139 The MITSHM extension uses SYSV IPC, but even if configure's test 76140 for IPC failed, MITSHM was still enabled by default, breaking 76141 MinGW builds by default. 76142 76143 Unfortunately, fixing this exposes the fact that the HAVE_SYSV_IPC 76144 test wasn't being used for anything before and so we hadn't noticed it 76145 was failing on Cygwin. 76146 76147 Change from using SHM_W|SHM_R flags (which aren't required by POSIX) to 76148 S_IRUSR|S_IWUSR flags (which are) 76149 76150 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 76151 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 76152 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 76153 76154commit 1e7b500a8e1d79b91a4e857a2da06194efe8cf69 76155Author: Chase Douglas <chase.douglas@canonical.com> 76156Date: Tue May 1 10:21:11 2012 -0700 76157 76158 Report logical button state in ProcXIQueryPointer 76159 76160 Physical button state is usually meaningless to an X client. 76161 76162 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76163 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76164 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76165 76166commit d063f64b5c8c8c27fca41163c93dbca06a774d1c 76167Author: Dave Airlie <airlied@redhat.com> 76168Date: Tue May 1 17:12:29 2012 +0100 76169 76170 modesetting: attempt to work out if we want 24 or 32bpp 76171 76172 the cirrus driver presents certain challenges, and this is a 76173 workaround, until we can possibly agree some sane interface 76174 for exposing this information. 76175 76176 Signed-off-by: Dave Airlie <airlied@redhat.com> 76177 76178commit 0426e6d65b6598edfbcb2fc66ee65fb08256469e 76179Author: Jon TURNEY <jon.turney@dronecode.org.uk> 76180Date: Tue Jan 3 00:14:32 2012 +0000 76181 76182 configure.ac: Make default configuration for MinGW disable unsupported extensions and DDXs 76183 76184 Same as the default configuration for Cygwin, --disable-xorg and --disable-dmx DDX by default, 76185 and force --disable-xv and other unsupported extensions 76186 76187 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 76188 Reviewed-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 76189 76190commit 610f532e6a3d934c473d2d16dff95f582285b2eb 76191Author: Dave Airlie <airlied@redhat.com> 76192Date: Tue May 1 16:52:18 2012 +0100 76193 76194 modesetting: move opening kernel device to before setting depth 76195 76196 due to interaction between option handling and set depth, we need 76197 to what fbdev does to get the device path early. 76198 76199 Signed-off-by: Dave Airlie <airlied@redhat.com> 76200 76201commit f3410b97cf9b48a47bee3d15d232f8a88e75f4ef 76202Author: Peter Hutterer <peter.hutterer@who-t.net> 76203Date: Mon Apr 30 10:01:48 2012 +1000 76204 76205 dix: when disabling a device, release all buttons and keys 76206 76207 A suspend-induced device disable may happen before the device gets to see 76208 the button release event. On resume, the server's internal state still has 76209 some buttons pressed, causing inconsistent behaviour. 76210 76211 Force the release and the matching events to be sent to the client. 76212 76213 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76214 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 76215 76216commit af88b43f9e604157b74270d609c08bdfa256a792 76217Author: Peter Hutterer <peter.hutterer@who-t.net> 76218Date: Fri Apr 27 16:31:17 2012 +1000 76219 76220 dix: don't emulate scroll events for non-existing axes (#47281) 76221 76222 Test case: 76223 - create a device with REL_HWHEEL and ABS_X and ABS_Y. evdev 2.7.0 will set 76224 that up as device with 1 relative axis 76225 - move pointer to VGA1 76226 - xrandr --output VGA1 --off 76227 76228 Warps the pointer to the new spot and calls GPE with the x/y mask bits set. 76229 When running through the loop to check for scroll event, this overruns the 76230 axes and may try to emulate scroll events based on random garbage in the 76231 memory. If that memory contained non-zero for the scroll type but near-zero 76232 for the increment field, the server would hang in an infinite loop. 76233 76234 This was the trigger for this suggested, never-merged, patch here: 76235 http://patchwork.freedesktop.org/patch/9543/ 76236 76237 X.Org Bug 47281 <http://bugs.freedesktop.org/show_bug.cgi?id=47281> 76238 76239 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76240 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 76241 76242commit 08962951de969b9d8c870af8b6e47303dc0decfd 76243Author: Peter Hutterer <peter.hutterer@who-t.net> 76244Date: Fri Apr 27 10:52:39 2012 +1000 76245 76246 os: make timers signal-safe 76247 76248 If TimerSet() is called from a signal handler (synaptics tap handling code) 76249 may result in list corruption if we're currently inside TimerSet(). 76250 76251 See backtrace in 76252 https://bugzilla.redhat.com/show_bug.cgi?id=814869 76253 76254 Block signals for all list manipulations in the timers. 76255 76256 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76257 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 76258 76259commit d662fa2450856777b59c4b62b912395a8bfd52fd 76260Author: Michal Suchanek <hramrach@gmail.com> 76261Date: Thu Apr 26 15:11:20 2012 +0200 76262 76263 dmx: Annotate dmxlog.c with _X_ATTRIBUTE_PRINTF and _X_NORETURN 76264 76265 and fix resulting printf warning in dmxLogVisual 76266 76267 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 76268 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76269 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76270 76271commit 5c361d59c5031d9b3f7f9093a52d2b1ff4d9ae5f 76272Author: Chase Douglas <chase.douglas@canonical.com> 76273Date: Fri Apr 20 11:08:15 2012 -0700 76274 76275 TouchListenerAcceptReject: Warn and return early on bad listener index 76276 76277 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76278 Reviewed-by: Bryce Harrington <bryce@canonical.com> 76279 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76280 76281commit 88bacc49f06da5927f716869f5a32672a8297ed0 76282Author: Chase Douglas <chase.douglas@canonical.com> 76283Date: Wed Apr 4 15:29:42 2012 -0700 76284 76285 os: Add -displayfd option 76286 76287 This option specifies a file descriptor in the launching process. X 76288 will scan for an available display number and write that number back to 76289 the launching process, at the same time as SIGUSR1 generation. This 76290 means display managers don't need to guess at available display numbers. 76291 As a consequence, if X fails to start when using -displayfd, it's not 76292 because the display was in use, so there's no point in retrying the X 76293 launch on a higher display number. 76294 76295 Signed-off-by: Adam Jackson <ajax@redhat.com> 76296 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76297 Reviewed-by: Julien Cristau <jcristau@debian.org> 76298 Tested-by: Julien Cristau <jcristau@debian.org> 76299 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 76300 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76301 76302commit 1d82ec95942b88dd01f0ac6b883368360a0b5fe6 76303Author: Michal Suchanek <hramrach@gmail.com> 76304Date: Mon Apr 23 13:52:40 2012 +0200 76305 76306 xserver: Fix out-of-tree build 76307 76308 Fixes regression caused by ccb3e78124fb05defd0c9b438746b79d84dfc3ae 76309 76310 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 76311 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76312 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76313 76314commit ea51e9b2877df60135edaf2a8f88d0f2a2b41060 76315Author: Peter Hutterer <peter.hutterer@who-t.net> 76316Date: Mon Apr 23 10:35:53 2012 +1000 76317 76318 Xi: return BadValue on XIQueryVersion if the version is less than first call 76319 76320 Clients that use plugin systems may require multiple calls to 76321 XIQueryVersion from different plugins. The current error handling requires 76322 client-side synchronisation of version numbers. 76323 76324 The first call to XIQueryVersion defines the server behaviour. Once cached, 76325 always return that version number to any clients. Unless a client requests a 76326 version lower than the first defined one, then a BadValue must be returned 76327 to be protocol-compatible. 76328 76329 Introduced in 2c23ef83b0e03e163aeeb06133538606886f4e9c 76330 76331 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76332 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 76333 76334commit 93d6ba5b711cbd3f502d83e54c9739856d2e6f2a 76335Author: Jeremy Huddleston <jeremyhu@apple.com> 76336Date: Thu Apr 26 21:17:54 2012 -0700 76337 76338 XQuartz: Make sure we tell startx the correct server binary to use in our fallback path 76339 76340 With multiple servers installed, we can't be certain if X is Xorg or Xquartz 76341 76342 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76343 76344commit fa6dddc6ce51c1b7a43fb379fb2a19550f6c8683 76345Author: Michal Suchanek <hramrach@gmail.com> 76346Date: Wed Mar 28 18:12:39 2012 -0700 76347 76348 xfree86: workaround crash on close 76349 76350 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=41653 76351 76352 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 76353 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 76354 Tested-by: Knut Petersen <Knut_Petersen@t-online.de> 76355 76356commit adcb48a29dcb3f32ff570551de4b014775c1e147 76357Author: Jonas Maebe <jonas.maebe@elis.ugent.be> 76358Date: Sun Apr 22 20:52:47 2012 -0700 76359 76360 glapi: Correct size of allocated _glapi_table struct 76361 76362 The __glapi_gentable_set_remaining_noop() routine treats the _glapi_struct 76363 as an array of _glapi_get_dispatch_table_size() pointers, so we have to 76364 allocate _glapi_get_dispatch_table_size()*sizeof(void*) bytes rather than 76365 sizeof(struct _glapi_struct) bytes. 76366 76367 Signed-off-by: Jonas Maebe <jonas.maebe@elis.ugent.be> 76368 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 76369 76370commit e34519e525559b01a63d26639f13f0487468de28 76371Author: Jeremy Huddleston <jeremyhu@apple.com> 76372Date: Thu Apr 19 16:38:06 2012 -0700 76373 76374 XQuartz: darwinPointer now sends both absolute and relative motion 76375 76376 This should hopefully help out wine clients that were continuing to 76377 have issues after the earlier changes. 76378 76379 http://xquartz.macosforge.org/trac/ticket/548 76380 76381 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76382 76383commit 662d41acdde1dcb9774fbe4054e251c708acaffe 76384Author: Jeremy Huddleston <jeremyhu@apple.com> 76385Date: Thu Apr 19 18:48:22 2012 -0700 76386 76387 XQuartz: Add a hack to better handle clicky wheel scroll mice 76388 76389 We loose information from AppKit being in our way. Before adopting 76390 smooth scrolling, we always rounded-up the number of scroll button 76391 clicks per NSEvent. Now, the scroll value is accumulated in the 76392 dix, and clicky scroll wheels with legacy X11 clients are seeing 76393 an accumulation of error due to so many translations (button press 76394 to smooth scrolling value in AppKit, passed to the dix, and then 76395 synthesized into a button press). This attempts to make the 76396 situation better. 76397 76398 http://xquartz.macosforge.org/trac/ticket/562 76399 76400 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76401 76402commit eda85290ae47d8dd497ef9335d10268bbe3e02a0 76403Author: Jeremy Huddleston <jeremyhu@apple.com> 76404Date: Thu Apr 19 16:20:30 2012 -0700 76405 76406 XQuartz: Use screenInfo.{width,height} instead of grabbing it from the first screen 76407 76408 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76409 76410commit b99586c9086ac29df144ef75df92f8c0c7554651 76411Author: Jeremy Huddleston <jeremyhu@apple.com> 76412Date: Wed Apr 18 17:50:55 2012 -0700 76413 76414 XQuartz: Separate out tablet and mouse event delivery into separate functions 76415 76416 This should have no immediate impact aside from fake mouse buttons no longer 76417 working with tablets (where they aren't needed or desired anyways). This 76418 prepares us for future changes. 76419 76420 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76421 76422commit 30623d6ff7bca223f9e1a825e86bd317eef8bf16 76423Author: Jeremy Huddleston <jeremyhu@apple.com> 76424Date: Thu Apr 19 16:06:45 2012 -0700 76425 76426 XQuartz: Correct calculation of the size of our file descriptor array in console_redirect 76427 76428 Reported-by: Joe Rohde <joer@valvesoftware.com> 76429 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76430 76431commit b4c4c65a35640e7274ccd4228f62110147878b72 76432Author: Jeremy Huddleston <jeremyhu@apple.com> 76433Date: Wed Apr 18 01:01:44 2012 -0700 76434 76435 XQuartz: Fix a deadlock in pre-dispatch code 76436 76437 The fact that this has been in place so long makes me really wonder if 76438 anybody cares about this running in Tiger or Leopard. 76439 76440 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76441 76442commit d1ef0d4964e43a4ab777750ad214d1092fd1ffda 76443Author: Jeremy Huddleston <jeremyhu@apple.com> 76444Date: Sat Apr 7 15:26:53 2012 -0700 76445 76446 test: Fix make dist 76447 76448 It seems like make dist should be doing te right thing without this commit, 76449 but it's not in some cases. Don't ask me to explain why. 76450 76451 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76452 76453commit d97868d934fd1a00f9e9ef23a1154a70c7bc0464 76454Author: Jeremy Huddleston <jeremyhu@apple.com> 76455Date: Thu Apr 19 16:39:23 2012 -0700 76456 76457 os: Annotate OsVendorFatalError as _X_ATTRIBUTE_PRINTF 76458 76459 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76460 Tested-By: Michal Suchanek <hramrach@gmail.com> 76461 76462commit 0ba1794fe2896395875f62e911720c7585687444 76463Author: Jeremy Huddleston <jeremyhu@apple.com> 76464Date: Thu Apr 19 16:51:46 2012 -0700 76465 76466 xres: Fix build without composite 76467 76468 Regression from: b8d0d19a6d410776b53a41e7cae90f68d4b22bb7 76469 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 76470 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 76471 Tested-By: Michal Suchanek <hramrach@gmail.com> 76472 76473commit d77eb7ee49ef19c2c4c7381d56e9d0f9c3fbc890 76474Merge: 31e3c0ff7 c7b162555 76475Author: Keith Packard <keithp@keithp.com> 76476Date: Thu Apr 19 15:48:34 2012 -0500 76477 76478 Merge remote-tracking branch 'yselkowitz/master' 76479 76480 Pull in Cygwin for XFree86 76481 76482commit 31e3c0ff7f5fb36421be8de06e5f6e675ec26ddc 76483Author: Keith Packard <keithp@keithp.com> 76484Date: Thu Apr 19 15:25:41 2012 -0500 76485 76486 Export CompositeClientWindowType 76487 76488 Make sure CompositeClientWindowType is visible for XResource v1.2 76489 76490 Signed-off-by: Keith Packard <keithp@keithp.com> 76491 76492commit e6308e32fe2b5f74133d4d238ffa512257f6327c 76493Merge: 3720aa33e 51a8d8dd1 76494Author: Keith Packard <keithp@keithp.com> 76495Date: Thu Apr 19 10:45:07 2012 -0500 76496 76497 Merge remote-tracking branch 'whot/for-keith' 76498 76499 Touch input changes from Chase 76500 76501commit 51a8d8dd19d7496fe84b37a1f0a7a03658120539 76502Merge: ebf214876 00cf1c40b 76503Author: Peter Hutterer <peter.hutterer@who-t.net> 76504Date: Thu Apr 19 17:03:54 2012 +1000 76505 76506 Merge branch 'input-fixes' of git://people.freedesktop.org/~cndougla/xserver into for-keith 76507 76508commit 00cf1c40b28417d7035c2917d048553eb720023c 76509Author: Chase Douglas <chase.douglas@canonical.com> 76510Date: Wed Apr 18 18:21:54 2012 -0700 76511 76512 Replay original touch begin event instead of generated begin event 76513 76514 The generated event does not have axes other than X and Y and has a 76515 newer timestamp. In particular, the newer timestamp may be newer than 76516 the real touch end event, which may be stuck in the syncEvents queue. If 76517 a client uses the timestamps for grabbing bad things may happen. 76518 76519 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76520 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76521 76522commit 312910b4e34215aaa50fc0c6092684d5878dc32f 76523Author: Chase Douglas <chase.douglas@canonical.com> 76524Date: Wed Apr 18 11:15:40 2012 -0700 76525 76526 Update currentTime in dispatch loop 76527 76528 A request, like input device grabs, may check a request timestamp 76529 against currentTime. It is possible for currentTime to lag a previously 76530 sent event timestamp. If the client makes a request based on such an 76531 event timestamp, the request may fail the validity check against 76532 currentTime unless we always update the time before processing the 76533 request. 76534 76535 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76536 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76537 76538commit a986f2f30cbe2a00e72ded7315c4951d7703e549 76539Author: Chase Douglas <chase.douglas@canonical.com> 76540Date: Tue Apr 17 11:40:15 2012 -0700 76541 76542 Update device state including when touch record does not exist 76543 76544 If a touch is physically active, the pointer core state should reflect 76545 that the first button is pressed. Currently, this only occurs when there 76546 are active listeners of the touch sequence. By moving the device state 76547 updating to the beginning of touch processing we ensure it is updated 76548 according to the processed physical state no matter what. 76549 76550 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76551 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76552 76553commit ec9c4295830c3de610e65aca17f4da4a7af3c4c5 76554Author: Chase Douglas <chase.douglas@canonical.com> 76555Date: Wed Apr 18 12:04:58 2012 -0700 76556 76557 Check other clients' core masks properly when adding touch listener 76558 76559 The current code checks the core event mask as though it were an XI 76560 mask. This change fixes the checks so the proper client and event masks 76561 are used. 76562 76563 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76564 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76565 76566commit e175971a6f44d94aa8306dc6b9228ccb2c8a0b4d 76567Author: Chase Douglas <chase.douglas@canonical.com> 76568Date: Tue Apr 17 11:39:10 2012 -0700 76569 76570 Ensure touch is ended when last listener is rejected 76571 76572 Currently, the touch is only logically ended if the touch has physically 76573 ended. If the touch hasn't physically ended, the touch record is never 76574 ended. If there aren't any more listeners, we don't need to keep the dix 76575 touch record around any more. 76576 76577 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76578 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76579 76580commit d0449851d1233543c3133d77e0ab7233319cdf5f 76581Author: Chase Douglas <chase.douglas@canonical.com> 76582Date: Tue Apr 17 10:49:14 2012 -0700 76583 76584 Create a new dix touch record for an emulated touch with no listeners 76585 76586 As a special case, if a still physically active pointer emulated touch 76587 has no listeners and the device is explicitly grabbed for pointer 76588 events, create a new dix touch record for the grab only. 76589 76590 This allows for clients to "hand off" grabs. For example, when dragging 76591 a window under compiz the window decorator sees the button press and 76592 then ungrabs the implicit grab. It then tells compiz to grab the device, 76593 and compiz then moves the window with the pointer motion. This is racy, 76594 but is allowed by the input protocol for pointer events when there are 76595 no other clients with a grab on the device. 76596 76597 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76598 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76599 76600commit 3d06bfe93d33cfe6150d8fb0058ee7bc8d80622b 76601Author: Chase Douglas <chase.douglas@canonical.com> 76602Date: Tue Apr 17 13:56:08 2012 -0700 76603 76604 Rename TouchEnsureSprite to TouchBuildSprite and event type checks 76605 76606 The function will be used for building a sprite for pointer emulation 76607 after an explicit device grab. This commit refactors the code so that 76608 TouchBuildSprite will function with any event type and moves the checks 76609 to the caller. 76610 76611 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76612 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76613 76614commit 2efbed23c29020f9994ab7c3155ce7386950dc7a 76615Author: Chase Douglas <chase.douglas@canonical.com> 76616Date: Tue Apr 17 08:39:38 2012 -0700 76617 76618 When activating an explicit grab, update owning listener 76619 76620 Pointer passive grabs may be changed by the grabbing client. This allows 76621 for a selecting client to change an implicit grab to an active grab, 76622 which is the mechanism used for pop-up windows like application menus. 76623 76624 We need to do the same thing with touches. If the grabbing client is the 76625 owner of a touch sequence, change the listener record to reflect the new 76626 grab. If the grabbing client is not the owner, nothing changes for the 76627 touch. 76628 76629 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76630 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76631 76632commit ef64b5ee97099618cf2e2cbbd3e471095695ae24 76633Author: Chase Douglas <chase.douglas@canonical.com> 76634Date: Mon Apr 16 15:53:51 2012 -0700 76635 76636 Don't deactivate implicit pointer grab on fake touch end event 76637 76638 Fake touch end events are generated by touch acceptance and rejection. 76639 These should not cause implicit pointer grabs to be deactivated. 76640 76641 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76642 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76643 76644commit fc518cd9f59060cc19bb90361767c0f47f0e25eb 76645Author: Chase Douglas <chase.douglas@canonical.com> 76646Date: Mon Apr 16 15:31:47 2012 -0700 76647 76648 End a pointer emulated touch event only on a "real" end event 76649 76650 Fake end events are generated by touch acceptance or rejection. These 76651 should not end the touch point. 76652 76653 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76654 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76655 76656commit 80d7d1ec6a9d61aa96e7d019dc1bee29d90cea34 76657Author: Chase Douglas <chase.douglas@canonical.com> 76658Date: Mon Apr 16 14:19:59 2012 -0700 76659 76660 On touch accept, only process end event for owner if it has seen the end 76661 76662 We still need to generate the touch ownership event to process the 76663 ending of the touch event in the case where the owner has the end 76664 already. 76665 76666 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76667 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76668 76669commit 8dfd98245d2c44a1eb4c8b7c275e6cfc10fe40f1 76670Author: Chase Douglas <chase.douglas@canonical.com> 76671Date: Wed Apr 11 11:14:51 2012 -0700 76672 76673 Fix copy/paste error from before git history in UpdateCurrentTimeIf() 76674 76675 See UpdateCurrentTime() for reference. I don't know what bug this might 76676 trigger, but it wouldn't hurt to fix this. 76677 76678 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76679 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76680 76681commit 6ca30cb33e829b4edd01822367e44ffe6f0951b0 76682Author: Chase Douglas <chase.douglas@canonical.com> 76683Date: Wed Apr 11 08:17:25 2012 -0700 76684 76685 When deactivating an explicit pointer grab, reject all grabs on touches 76686 76687 Explicit pointer grabs are placed at the head of the touch listener 76688 array for pointer emulated touches. If the grab is deactivated, we must 76689 remove it from all touches for the device. 76690 76691 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76692 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76693 76694commit cacdb9a74065ccba7d50a82e14abdf04b36c5309 76695Author: Chase Douglas <chase.douglas@canonical.com> 76696Date: Wed Apr 11 08:13:17 2012 -0700 76697 76698 Accept touch sequence for pointer listener after second event delivery 76699 76700 This is a bit of unimplemented code for touchscreen pointer emulation. A 76701 pointer grabbing client currently never accepts the touch sequence. The 76702 sequence must be accepted once any touch-derived event is irrevocably 76703 delivered to a client. 76704 76705 The first pointer event, derived from a touch begin event, may be caught 76706 in a sync grab and then replayed. This is essentially a revocable 76707 delivery of an event. Thus, we must wait till a non-begin event is 76708 delivered. 76709 76710 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76711 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76712 76713commit 447fe7a1a72513aa68145962c47894242c589cc9 76714Author: Chase Douglas <chase.douglas@canonical.com> 76715Date: Wed Apr 11 08:12:08 2012 -0700 76716 76717 Split out helper function TouchListenerAcceptReject() 76718 76719 This will be used for accepting and rejecting touches in the future. 76720 76721 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76722 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76723 76724commit 93c33403642a3de3c9d141ad7940a7b880846aad 76725Author: Chase Douglas <chase.douglas@canonical.com> 76726Date: Wed Apr 4 12:59:55 2012 -0700 76727 76728 Only set XI2 mask if pointer emulation is for XI2 client 76729 76730 The current code returns a reference to memory that may not actually be 76731 an XI2 mask. Instead, only return a value when an XI2 client has 76732 selected for events. 76733 76734 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76735 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 76736 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 76737 76738commit 4c1dfd21937efc6a85fb204a73dd7d7151d54daa 76739Author: Chase Douglas <chase.douglas@canonical.com> 76740Date: Wed Apr 4 12:57:40 2012 -0700 76741 76742 Check core event mask properly for pointer emulated touch events 76743 76744 The current code checks the core event mask as though it were an XI2 76745 mask. This change fixes the checks so the proper client and event masks 76746 are used. 76747 76748 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76749 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 76750 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 76751 76752commit 3720aa33ee50788dd3d4acc9bbf8dfcb72c8f5ce 76753Author: Erkki Seppälä <erkki.seppala@vincit.fi> 76754Date: Thu Apr 7 12:53:51 2011 +0300 76755 76756 Added resourceproto version dependency, >= 1.2.0 now required 76757 76758 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 76759 76760commit 233eab4d05cae1fdb4129a2e9905961b78693f74 76761Author: Erkki Seppälä <erkki.seppala@vincit.fi> 76762Date: Wed Dec 8 17:09:30 2010 +0200 76763 76764 dix: add reference count of the resource to ResourceSizeRec 76765 76766 The ResourceSizeRec now contains the number of references to the 76767 resource. For example a Pixmap knows this value and it can be useful 76768 for determining the "weight" of the resource. Typically this value 76769 is 1. 76770 76771 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 76772 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 76773 76774commit ccb3e78124fb05defd0c9b438746b79d84dfc3ae 76775Author: Erkki Seppälä <erkki.seppala@vincit.fi> 76776Date: Tue Dec 14 12:18:23 2010 +0200 76777 76778 Xext: add a generic hashtable implementation 76779 76780 The generic hashtable implementation adds a key-value container, that 76781 keeps the key and value inside the hashtable structure and manages 76782 their memory by itself. This data structure is best suited for 76783 fixed-length keys and values. 76784 76785 One creates a new hash table with ht_create and disposes it with 76786 ht_destroy. ht_create accepts the key and value sizes (in bytes) in 76787 addition to the hashing and comparison functions to use. When adding 76788 keys with ht_add, they will be copied into the hash and a pointer to 76789 the value will be returned: data may be put into this structure (or if 76790 the hash table is to be used as a set, one can just not put anything 76791 in). 76792 76793 The hash table comes also with one generic hashing function plus a 76794 comparison function to facilitate ease of use. It also has a custom 76795 hashing and comparison functions for hashing resource IDs with 76796 HashXID. 76797 76798 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 76799 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 76800 76801commit a2ac01a8ea8508ed35aa844a589672c1165e05e4 76802Author: Erkki Seppälä <erkki.seppala@vincit.fi> 76803Date: Wed Apr 6 10:16:53 2011 +0300 76804 76805 dix: don't use a local wrapper for calling HashResourceID 76806 76807 Calls to Hash(client, id) were replaced with calls directly to 76808 HashResourceID(id, clientTable[client].hashsize) and the Hash-function 76809 was removed. 76810 76811 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 76812 76813commit a0b0fb83f91bb82534a0d83fdd6c0222567b098d 76814Author: Erkki Seppälä <erkki.seppala@vincit.fi> 76815Date: Mon Dec 20 12:58:37 2010 +0200 76816 76817 dix: add hashing functions to resource.h for others to use. 76818 76819 The public hashing function HashResourceID uses the same hashing 76820 hashing algorithm as resource.c uses internally, but it provides an 76821 interface that will is usable by external modules. It provides a 76822 parameter for the number of bits for the hash, instead of finding the 76823 size from its internal hash table. 76824 76825 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 76826 76827commit 3ba0decb4b55a1fd122d269e15b2b2da8ced8624 76828Author: Erkki Seppälä <erkki.seppala@vincit.fi> 76829Date: Wed Dec 8 15:30:57 2010 +0200 76830 76831 dix: add a mechanism for iterating through all subresources 76832 76833 The mechanism allows iterating even through subresources that don't 76834 have specific XID's. When such 'resources' are iterated, the XID for 76835 them will be zero. A resource type can assign an iteration function 76836 for its subresources with SetResourceTypeFindSubResFunc; by default 76837 resources are assumed not to contain subresources. 76838 76839 The purpose of this extension is to enable accurate accounting of 76840 the resources a resource consumes or uses. 76841 76842 This patch provides the subresource iteration functions for Windows 76843 and GCs. 76844 76845 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 76846 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 76847 76848commit b8d0d19a6d410776b53a41e7cae90f68d4b22bb7 76849Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 76850Date: Wed Oct 27 17:25:50 2010 +0300 76851 76852 composite: Report pixmap usage of client windows to resource extension. 76853 76854 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 76855 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 76856 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 76857 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 76858 76859commit e83388cc70e21e7f377ed2e417d04469e23eb706 76860Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 76861Date: Wed Oct 27 16:59:06 2010 +0300 76862 76863 render: Report pixmap usage of pictures to resource extension. 76864 76865 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 76866 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 76867 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 76868 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 76869 76870commit 96864bfa951ea8bf4ab697753fc62c6a97598bc0 76871Author: Erkki Seppälä <erkki.seppala@vincit.fi> 76872Date: Mon Nov 29 12:40:56 2010 +0200 76873 76874 Implemented first part of XResource extension v1.2: X_XResQueryClientIds 76875 76876 This patch implements a part of the XResource extension v1.2 (as specified in 76877 http://patchwork.freedesktop.org/patch/2720/ ). The request implemented is 76878 X_XResQueryClientIds. 76879 76880 This patch depends on the feature introduced by 76881 1e933665bef26c74196bb7c59910e6a78bcacf0e "dix: Add facilities for 76882 client ID tracking." . 76883 76884 This latest version also adds Doxygen-formatted comments and takes a better 76885 notice of coding conventions (as in http://www.x.org/wiki/CodingStyle ). 76886 76887 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 76888 76889commit ebf214876a4885a98ded4f5525925b69005fae05 76890Author: Peter Hutterer <peter.hutterer@who-t.net> 76891Date: Wed Apr 18 15:56:37 2012 +1000 76892 76893 dix: indentation fix 76894 76895 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76896 76897commit 00c663a3bc76d567ad7edd6d6ef3f7f725b5df3a 76898Author: Dave Airlie <airlied@redhat.com> 76899Date: Tue Apr 17 11:50:40 2012 +0100 76900 76901 modesetting: add set/drop master around VT switch 76902 76903commit 001dec898098e11c569abc20a307b3c2af2e2184 76904Author: Dave Airlie <airlied@redhat.com> 76905Date: Tue Apr 17 11:48:03 2012 +0100 76906 76907 cursor: hide cursors on LeaveVT 76908 76909commit 80fefc42f5e67e6b4a4b440d8991bee7e5f38359 76910Merge: 9779b904c 12188c8a8 76911Author: Keith Packard <keithp@keithp.com> 76912Date: Sun Apr 15 21:05:30 2012 -0700 76913 76914 Merge remote-tracking branch 'whot/for-keith' 76915 76916commit 12188c8a8a537b38b1ca4cf8c0de5447e19c886a 76917Author: Chase Douglas <chase.douglas@canonical.com> 76918Date: Tue Apr 3 17:31:01 2012 -0700 76919 76920 Use touch state when querying pointer through core protocol 76921 76922 QueryPointer is part of the core protocol. As such, it knows nothing 76923 about touch devices. Touches are converted to button 1 press, pointer 76924 motion, and button 1 release for core clients, so we should ensure the 76925 pointer state mask has button 1 set when XQueryPointer is used. 76926 76927 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76928 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76929 76930commit 32ece7c09bf0ebc3d99b4078aacebbd44314776a 76931Author: Chase Douglas <chase.douglas@canonical.com> 76932Date: Tue Apr 10 17:12:42 2012 -0700 76933 76934 Ensure sequential touches are pointer emulated sequentially 76935 76936 Issue: 76937 * Two sequential touches (i.e. down, up, down, up) 76938 * Both are grabbed by a touch grab 76939 * Both have a second listener in the form of a pointer grab or selection 76940 * The second and first touches are rejected in that order 76941 76942 The first touch must be pointer emulated before the second touch, so the 76943 second touch must be paused until the first touch is rejected or 76944 accepted and all events are delivered to pointer clients. 76945 76946 This change ensures all pointer emulated events are emitted 76947 sequentially. It necessarily imposes a delay on further touch events 76948 when pointer grabs and selections are used, but there is no way around 76949 it. 76950 76951 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76952 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76953 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76954 76955commit 163b0f375d73c05873fb341652de3ed347337828 76956Author: Chase Douglas <chase.douglas@canonical.com> 76957Date: Tue Apr 10 17:12:41 2012 -0700 76958 76959 Update event type when delivering end event to a pointer listener 76960 76961 Just like when we deliver to a touch listener, we must convert a touch 76962 end event to an update event for further clients. This also ensures that 76963 the touch record is not deleted at the end of ProcessTouchEvent(). 76964 76965 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76966 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76967 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76968 76969commit 210cd12c47d063f97915ff23292b61d09abfd73a 76970Author: Chase Douglas <chase.douglas@canonical.com> 76971Date: Tue Apr 10 17:12:40 2012 -0700 76972 76973 Don't update listener after deactivating implicit pointer grab 76974 76975 After the pointer grab is deactivated, the touch listener record is 76976 updated at the end of DeliverTouchEmulatedEvent. However, the touch 76977 record is ended when the grab is deactivated, so the update to the 76978 listener record is in an array of memory that has been freed. 76979 76980 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 76981 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 76982 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76983 76984commit 82a1ae0af3b136371638659c3e909880a99f721c 76985Author: Peter Hutterer <peter.hutterer@who-t.net> 76986Date: Thu Apr 12 15:54:00 2012 +1000 76987 76988 xfree86: after VT switching back, only enable previously enabled devices 76989 76990 If a device was enabled before the VT switch, re-enabled it. Otherwise leave 76991 it as is, there was probably a reason why it was disabled. 76992 76993 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 76994 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 76995 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 76996 76997commit d5d5b407436dfdbffcd56bbefecf17d11750dddf 76998Author: Dave Airlie <airlied@redhat.com> 76999Date: Sat Apr 14 19:21:47 2012 +0100 77000 77001 dirty: check malloc return 77002 77003 Suggested by keithp. 77004 77005 Signed-off-by: Dave Airlie <airlied@redhat.com> 77006 77007commit 6c1b5cb903f2ff9cb915273835c301d6ac255f45 77008Author: Dave Airlie <airlied@redhat.com> 77009Date: Sat Apr 14 15:01:16 2012 +0100 77010 77011 drop use of alloca, just use malloc/free 77012 77013 Reported-by: Alan Coopersmith <alan.coopersmith@oracle.com> 77014 Signed-off-by: Dave Airlie <airlied@redhat.com> 77015 77016commit c5a45b0f7658c77725adce2b64a0fbd62f208328 77017Author: Daniel Kurtz <djkurtz@chromium.org> 77018Date: Thu Apr 12 10:11:10 2012 +1000 77019 77020 dix: don't BUG_WARN for button events from button-only device 77021 77022 Events from button-only devices still need coordinates, and they get them 77023 from scale_to_desktop(). Therefore, a dev without valuators is not a bug. 77024 However, a dev with valuators, but less than two of them still is a bug. 77025 77026 This was noticed when unplugging a "Creative Technology SB Arena Headset", 77027 which has some BTNs and some KEYs, but no REL or ABS valuators. 77028 It emits [BTN_3] = 0 on unplug, which would trigger the BUG_WARN. 77029 77030 Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> 77031 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 77032 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77033 77034commit 9779b904c7c0b49c74054c22c420012c40595cdc 77035Author: Keith Packard <keithp@keithp.com> 77036Date: Wed Apr 11 09:33:54 2012 -0700 77037 77038 hw/xfree86: Re-indent xf86vmode.c 77039 77040 This is the result of re-running the 'x-indent.sh' script over 77041 xf86vmode.c to clean up the disaster caused by broken syntax in the 77042 file. 77043 77044 Signed-off-by: Keith Packard <keithp@keithp.com> 77045 77046commit 592bd0ae2b60cd6f6afd3efc40f5f659b12900b4 77047Author: Keith Packard <keithp@keithp.com> 77048Date: Wed Apr 11 09:28:21 2012 -0700 77049 77050 hw/xfree86: Spurious ');' in xf86vmode.c messed up indentation badly 77051 77052 Inside the unfinished XF86VIDMODE_EVENTS #ifdef block the 77053 function definition for xf86VidModeNotifyEvent had an extra ');' 77054 before the prototype argument declarations. This was harmless for the 77055 compiler as the code never gets used, but completely messed up the 77056 file re-indentation. This patch removes the spurious characters in 77057 preparation for re-indenting the file. 77058 77059 Signed-off-by: Keith Packard <keithp@keithp.com> 77060 77061commit 72cfc1a097dc1e09d2cd9415ef7855a2cef92351 77062Author: Peter Hutterer <peter.hutterer@who-t.net> 77063Date: Wed Apr 11 09:43:23 2012 +1000 77064 77065 Xi: fix XITouchClass sourceid assignment 77066 77067 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77068 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 77069 77070commit 10cd6fdc431d2d0835c313b2f5bdda6dcea281fd 77071Author: Keith Packard <keithp@keithp.com> 77072Date: Sun Apr 8 00:16:54 2012 -0700 77073 77074 Revert "xfree86: workaround crash on close" 77075 77076 This reverts commit 55f552adb651715d2620db7248cd5b9b8187654a. 77077 77078 This appears to cause a crash at init time instead of close. 77079 77080 Reported-by: Knut Petersen <Knut_Petersen@t-online.de> 77081 Acked-by: Jeremy Huddleston <jeremyhu@apple.com> 77082 Signed-off-by: Keith Packard <keithp@keithp.com> 77083 77084commit c7b162555849ab2d4cb195d02d98a40494c5ce55 77085Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77086Date: Thu Mar 29 02:26:41 2012 -0500 77087 77088 xfree86: link modules against Xorg symbols on Cygwin 77089 77090 As a PE platform, all symbols in both EXEs and DLLs must be resolved 77091 at link time. As Xorg modules depend on symbols in the Xorg 77092 executable, we must build Xorg before its modules, creating an implib 77093 from the former which is used to link the latter. This implib must 77094 then be installed in order to build the drivers. 77095 77096 Currently only two drivers are supported on Cygwin: xf86-video-dummy 77097 (to replace Xvfb/Xfake) and xf86-video-nested (to replace Xnest/Xephyr). 77098 77099 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77100 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 77101 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77102 77103commit 2dffdcd60f70f534ad730830fe97f0bb1a192e2a 77104Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77105Date: Thu Mar 29 02:23:17 2012 -0500 77106 77107 xf86Config: load DIX libraries before drivers on Cygwin 77108 77109 Cygwin doesn't have ELF rpath capabilities, so these libraries need 77110 to be loaded before the drivers (namely dummy and nested) which 77111 depend on their symbols. 77112 77113 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77114 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 77115 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77116 77117commit 5dd3d2dbbafee28142552640ac6ef3eef2e6d9d3 77118Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77119Date: Thu Mar 29 02:21:01 2012 -0500 77120 77121 xf86Init: provide ddxBeforeReset ifdef DDXBEFORERESET 77122 77123 This is necessary when building Xorg and XWin simultaneously, otherwise 77124 undefined symbol errors result in sdksyms.c. 77125 77126 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77127 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 77128 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77129 77130commit 96186bc7219a88e549928707b3da4b8bfc16c54d 77131Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77132Date: Thu Mar 29 02:18:41 2012 -0500 77133 77134 loader: add Cygwin support 77135 77136 Cygwin libraries use the .dll extension and "cyg" prefix in place of "lib". 77137 77138 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77139 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 77140 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77141 77142commit 0ce48729d34ab610ff119303355883048b651500 77143Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77144Date: Thu Mar 29 02:17:22 2012 -0500 77145 77146 os-support: add Cygwin support 77147 77148 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77149 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 77150 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77151 77152commit fd115ee114e95b4e7b96f789d3ad67e2e15555de 77153Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77154Date: Thu Mar 29 02:13:27 2012 -0500 77155 77156 xfree86: allow modules to be built without undefined symbols 77157 77158 This will be necessary to port Xorg to Cygwin, but other platforms may 77159 find this useful as well. 77160 77161 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77162 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 77163 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77164 77165commit 12cd7ca83c35bc61fbda7e61535692e3ec78194c 77166Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77167Date: Thu Mar 29 02:06:36 2012 -0500 77168 77169 Cygwin/X: disable all unused or unapplicable extensions in configure 77170 77171 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77172 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 77173 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77174 77175commit cb5661e86d4a428ec0ca4efc30a1a165d94d7215 77176Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77177Date: Tue Mar 27 22:16:08 2012 -0500 77178 77179 xf86Crtc: include "xf86xv.h" only if XV 77180 77181 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 77182 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 77183 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 77184 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77185 77186commit 1110facdfeb95b1ad47d03c0ca3d73933b86dbd6 77187Author: Chase Douglas <chase.douglas@canonical.com> 77188Date: Thu Mar 29 18:06:03 2012 -0700 77189 77190 Implement passive touch ungrabbing 77191 77192 Whoops. Forgot to implement this. The code currently generates an error 77193 due to the unhandled grab type. 77194 77195 X.Org Bug 48069 <https://bugs.freedesktop.org/show_bug.cgi?id=48069> 77196 77197 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 77198 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 77199 Signed-off-by: Keith Packard <keithp@keithp.com> 77200 77201commit e9d3848d7bda7f7de4c0f497aee4482d0b30ad49 77202Author: Jeremy Huddleston <jeremyhu@apple.com> 77203Date: Thu Apr 5 14:03:35 2012 -0700 77204 77205 XQuartz: Automatically start our virtual tablet devices 77206 77207 Fixes: https://bugs.launchpad.net/inkscape/+bug/972914 77208 Regression introduced by: 7790dc86384cc451ac44663737fde84dd81ad4e1 77209 77210 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77211 77212commit 6cb83b78c47b80556cb4d573524d60f0cd1a64ac 77213Author: Jeremy Huddleston <jeremyhu@apple.com> 77214Date: Fri Mar 30 14:29:48 2012 -0700 77215 77216 XQuartz: Tiger build fix 77217 77218 Fixes: https://trac.macports.org/ticket/33818 77219 77220 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77221 77222commit eca58ee4db43d85e5940919be094c38ad166d5d0 77223Author: Jeremy Huddleston <jeremyhu@apple.com> 77224Date: Wed Mar 28 15:07:51 2012 -0700 77225 77226 os: Fix regression with FatalError not calling va_start 77227 77228 Regression From: a818b305989bbcde4e585112a7ee70cbc0b14a92 77229 77230 Found-by: Colin Harrison <colin.harrison@virgin.net> 77231 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77232 77233commit 55f552adb651715d2620db7248cd5b9b8187654a 77234Author: Michal Suchanek <hramrach@gmail.com> 77235Date: Wed Mar 28 18:12:39 2012 -0700 77236 77237 xfree86: workaround crash on close 77238 77239 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=41653 77240 77241 Signed-off-by: Michal Suchanek <hramrach@gmail.com> 77242 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77243 77244commit 641a1b9363d59808d2586f9e84847ccc69701482 77245Author: Peter Hutterer <peter.hutterer@who-t.net> 77246Date: Mon Mar 26 10:34:34 2012 +1000 77247 77248 test: reset the client before checking the swapped version 77249 77250 As of 2c23ef83b0e03e163aeeb06133538606886f4e9c, the server returns BadValue 77251 for the same client with multiple versions. Avoid this by resetting the 77252 client before we issue the same request as a fake swap client. 77253 77254 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77255 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77256 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 77257 77258commit 17f33ca0e5bff45a2d03e4d3b4753a714ce25baf 77259Author: Peter Hutterer <peter.hutterer@who-t.net> 77260Date: Mon Mar 26 10:28:32 2012 +1000 77261 77262 test: init the sync extensions for tests to pass again 77263 77264 Introduced in d645edd11e7482f98c8b7e0d6c8693285c484907 77265 77266 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77267 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77268 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 77269 77270commit 8053faa7b81520b0a46459ba6e3b98f2599d5bbf 77271Author: Peter Hutterer <peter.hutterer@who-t.net> 77272Date: Fri Mar 23 08:20:22 2012 +1000 77273 77274 Bump input ABI to 17 for per-device idlecounters 77275 77276 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77277 77278commit 7f3997b01a8813c1d7055317eb06111aed8572c9 77279Merge: 92d50c38b 31646d8fa 77280Author: Keith Packard <keithp@keithp.com> 77281Date: Mon Mar 26 16:41:52 2012 -0700 77282 77283 Merge remote-tracking branch 'jeremyhu/master' 77284 77285commit 92d50c38b2bfeb82e7374a1966a53235e3f8275e 77286Author: Gaetan Nadon <memsize@videotron.ca> 77287Date: Sun Feb 26 12:15:31 2012 -0500 77288 77289 man: s/__xservername__/Xorg/g - no longer required 77290 77291 Excerpt from http://lists.x.org/archives/xorg-devel/2011-March/020481.html: 77292 77293 The Xorg & xorg.conf substitutions are leftover from the transitional 77294 period where some distros were building our sources with the XFree86 77295 and XF86config names until they had time to adjust the rest of their 77296 packages/installer/config code to the new names. 77297 77298 This will fix inconsistencies and prevent the creation of new unneeded 77299 sed patterns. 77300 77301 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 77302 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 77303 Signed-off-by: Keith Packard <keithp@keithp.com> 77304 77305commit 9e880cd2e36236e466fda0adadee34a407ae3341 77306Author: Gaetan Nadon <memsize@videotron.ca> 77307Date: Sun Feb 26 12:15:30 2012 -0500 77308 77309 man: s/__xconfigfile__/xorg.conf/g - no longer required 77310 77311 Excerpt from http://lists.x.org/archives/xorg-devel/2011-March/020481.html: 77312 77313 The Xorg & xorg.conf substitutions are leftover from the transitional 77314 period where some distros were building our sources with the XFree86 77315 and XF86config names until they had time to adjust the rest of their 77316 packages/installer/config code to the new names. 77317 77318 This will fix inconsistencies and prevent the creation of new unneeded 77319 sed patterns. 77320 77321 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 77322 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 77323 Signed-off-by: Keith Packard <keithp@keithp.com> 77324 77325commit 31646d8fa9524adca1d7bc2cd2db90d47c2eb96e 77326Author: Jeremy Huddleston <jeremyhu@apple.com> 77327Date: Sun Mar 25 23:33:54 2012 -0700 77328 77329 XQuartz: Use scrolling valuators 77330 77331 This lets use send more accurate data to Xi clients and uses dix 77332 for legacy scroll buttons rather than reinventing the wheel. 77333 77334 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77335 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 77336 77337commit 7790dc86384cc451ac44663737fde84dd81ad4e1 77338Author: Jeremy Huddleston <jeremyhu@apple.com> 77339Date: Sun Mar 25 22:30:55 2012 -0700 77340 77341 XQuartz: Use AllocDevicePair to allocate our keyboard and mouse at the same time 77342 77343 Also correct isMaster to FALSE while we're here. 77344 77345 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77346 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 77347 77348commit 74fb534bf15687610367248071341fe4bfd1d8bf 77349Author: Jeremy Huddleston <jeremyhu@apple.com> 77350Date: Sun Mar 25 22:16:22 2012 -0700 77351 77352 XQuartz: Call update_desktop_dimensions() after updating our screen coordinates 77353 77354 Fixes: http://xquartz.macosforge.org/trac/ticket/555 77355 Regression from: 8bebb4b4896d8b6ba3309b5b28fce883bb9f8a96 77356 77357 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77358 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 77359 77360commit 41a46546e92b66fed323b5845dd71a6385feb418 77361Author: Jeremy Huddleston <jeremyhu@apple.com> 77362Date: Sun Mar 25 23:05:43 2012 -0700 77363 77364 XQuartz: Formatting cleanup for sendX11NSEvent 77365 77366 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77367 77368commit 5324557c7b12dcacddac830bdb65769cb622c6cf 77369Author: Jeremy Huddleston <jeremyhu@apple.com> 77370Date: Fri Mar 23 19:58:49 2012 -0700 77371 77372 XQuartz: Source formatting cleanup 77373 77374 indent butchered Objective-C formatting. This patch was created by: 77375 77376 1) Reverting the indent changes in hw/xquartz 77377 2) Editing X11Application.m and chaning some #ifdef logic to work 77378 with uncrustify 77379 3) Hand edited some (c) notifications 77380 4) Opened all XQuartz sources in XCode and re-indented (^I) 77381 5) Ran uncrustify with this configuration 77382 (as close to the indent rules as I could get): 77383 77384 tok_split_gte=false 77385 utf8_byte=true 77386 utf8_force=true 77387 indent_cmt_with_tabs=false 77388 indent_align_string=false 77389 indent_braces=false 77390 indent_braces_no_func=false 77391 indent_braces_no_class=false 77392 indent_braces_no_struct=false 77393 indent_brace_parent=false 77394 indent_namespace=false 77395 indent_extern=false 77396 indent_class=false 77397 indent_class_colon=false 77398 indent_else_if=false 77399 indent_var_def_cont=false 77400 indent_func_call_param=false 77401 indent_func_def_param=false 77402 indent_func_proto_param=false 77403 indent_func_class_param=false 77404 indent_func_ctor_var_param=false 77405 indent_template_param=false 77406 indent_func_param_double=false 77407 indent_relative_single_line_comments=false 77408 indent_col1_comment=false 77409 indent_access_spec_body=false 77410 indent_paren_nl=false 77411 indent_comma_paren=false 77412 indent_bool_paren=false 77413 indent_first_bool_expr=false 77414 indent_square_nl=false 77415 indent_preserve_sql=false 77416 indent_align_assign=true 77417 sp_balance_nested_parens=false 77418 align_keep_tabs=false 77419 align_with_tabs=false 77420 align_on_tabstop=false 77421 align_number_left=false 77422 align_func_params=false 77423 align_same_func_call_params=false 77424 align_var_def_colon=true 77425 align_var_def_attribute=true 77426 align_var_def_inline=true 77427 align_right_cmt_mix=false 77428 align_on_operator=false 77429 align_mix_var_proto=false 77430 align_single_line_func=false 77431 align_single_line_brace=false 77432 align_nl_cont=false 77433 align_left_shift=true 77434 align_oc_decl_colon=true 77435 nl_collapse_empty_body=true 77436 nl_assign_leave_one_liners=true 77437 nl_class_leave_one_liners=true 77438 nl_enum_leave_one_liners=true 77439 nl_getset_leave_one_liners=true 77440 nl_func_leave_one_liners=true 77441 nl_if_leave_one_liners=true 77442 nl_multi_line_cond=false 77443 nl_multi_line_define=false 77444 nl_before_case=true 77445 nl_after_case=true 77446 nl_after_return=false 77447 nl_after_semicolon=true 77448 nl_after_brace_open=true 77449 nl_after_brace_open_cmt=false 77450 nl_after_vbrace_open=false 77451 nl_after_vbrace_open_empty=false 77452 nl_after_brace_close=false 77453 nl_after_vbrace_close=false 77454 nl_define_macro=false 77455 nl_squeeze_ifdef=false 77456 nl_ds_struct_enum_cmt=false 77457 nl_ds_struct_enum_close_brace=false 77458 nl_create_if_one_liner=false 77459 nl_create_for_one_liner=false 77460 nl_create_while_one_liner=false 77461 ls_for_split_full=false 77462 ls_func_split_full=false 77463 nl_after_multiline_comment=false 77464 eat_blanks_after_open_brace=false 77465 eat_blanks_before_close_brace=false 77466 mod_full_brace_if_chain=false 77467 mod_pawn_semicolon=false 77468 mod_full_paren_if_bool=false 77469 mod_remove_extra_semicolon=false 77470 mod_sort_import=false 77471 mod_sort_using=false 77472 mod_sort_include=false 77473 mod_move_case_break=false 77474 mod_remove_empty_return=false 77475 cmt_indent_multi=true 77476 cmt_c_group=false 77477 cmt_c_nl_start=false 77478 cmt_c_nl_end=false 77479 cmt_cpp_group=false 77480 cmt_cpp_nl_start=false 77481 cmt_cpp_nl_end=false 77482 cmt_cpp_to_c=false 77483 cmt_star_cont=false 77484 cmt_multi_check_last=true 77485 cmt_insert_before_preproc=false 77486 pp_indent_at_level=false 77487 pp_region_indent_code=false 77488 pp_if_indent_code=false 77489 pp_define_at_level=false 77490 indent_columns=4 77491 indent_brace=0 77492 indent_switch_case=0 77493 align_struct_init_span=2 77494 align_pp_define_gap=0 77495 align_pp_define_span=2 77496 align_oc_msg_colon_span=16 77497 nl_end_of_file_min=1 77498 nl_func_var_def_blk=0 77499 code_width=78 77500 nl_max=2 77501 newlines=auto 77502 indent_with_tabs=0 77503 sp_arith=force 77504 sp_assign=force 77505 sp_assign_default=force 77506 sp_before_assign=force 77507 sp_after_assign=force 77508 sp_enum_assign=force 77509 sp_enum_before_assign=force 77510 sp_enum_after_assign=force 77511 sp_pp_stringify=add 77512 sp_bool=force 77513 sp_compare=force 77514 sp_inside_paren=remove 77515 sp_paren_paren=remove 77516 sp_paren_brace=force 77517 sp_before_ptr_star=ignore 77518 sp_before_unnamed_ptr_star=force 77519 sp_before_byref=force 77520 sp_before_unnamed_byref=force 77521 sp_after_byref=remove 77522 sp_after_type=force 77523 sp_before_sparen=force 77524 sp_inside_sparen=remove 77525 sp_inside_sparen_close=remove 77526 sp_after_sparen=force 77527 sp_sparen_brace=force 77528 sp_special_semi=force 77529 sp_before_semi=remove 77530 sp_after_semi=force 77531 sp_after_semi_for=force 77532 sp_after_semi_for_empty=remove 77533 sp_before_square=remove 77534 sp_inside_square=remove 77535 sp_after_comma=force 77536 sp_before_comma=remove 77537 sp_paren_comma=force 77538 sp_before_ellipsis=force 77539 sp_after_class_colon=force 77540 sp_before_class_colon=force 77541 sp_before_case_colon=remove 77542 sp_after_cast=remove 77543 sp_inside_paren_cast=remove 77544 sp_sizeof_paren=remove 77545 sp_inside_braces_enum=force 77546 sp_inside_braces_struct=force 77547 sp_inside_braces=force 77548 sp_inside_braces_empty=remove 77549 sp_func_proto_paren=remove 77550 sp_func_def_paren=remove 77551 sp_inside_fparens=remove 77552 sp_inside_fparen=remove 77553 sp_square_fparen=remove 77554 sp_fparen_brace=force 77555 sp_func_call_paren=remove 77556 sp_func_call_paren_empty=remove 77557 sp_return_paren=force 77558 sp_attribute_paren=remove 77559 sp_defined_paren=remove 77560 sp_macro=force 77561 sp_macro_func=force 77562 sp_else_brace=force 77563 sp_brace_else=force 77564 sp_brace_typedef=force 77565 sp_not=remove 77566 sp_inv=remove 77567 nl_start_of_file=remove 77568 nl_end_of_file=force 77569 nl_assign_square=remove 77570 nl_after_square_assign=remove 77571 nl_fcall_brace=remove 77572 nl_enum_brace=remove 77573 nl_struct_brace=remove 77574 nl_union_brace=remove 77575 nl_if_brace=remove 77576 nl_brace_else=force 77577 nl_elseif_brace=remove 77578 nl_else_brace=remove 77579 nl_else_if=remove 77580 nl_for_brace=remove 77581 nl_do_brace=remove 77582 nl_brace_while=remove 77583 nl_switch_brace=remove 77584 nl_case_colon_brace=force 77585 nl_func_type_name=force 77586 nl_func_type_name_class=force 77587 nl_func_proto_type_name=force 77588 nl_func_paren=remove 77589 nl_func_def_paren=remove 77590 nl_func_decl_start=remove 77591 nl_func_def_start=remove 77592 nl_func_decl_args=remove 77593 nl_func_decl_end=remove 77594 nl_func_def_end=remove 77595 nl_func_decl_end_single=remove 77596 nl_func_def_end_single=remove 77597 nl_func_decl_empty=remove 77598 nl_func_def_empty=remove 77599 nl_fdef_brace=force 77600 nl_return_expr=remove 77601 nl_before_if=ignore 77602 nl_after_if=ignore 77603 nl_before_for=ignore 77604 nl_after_for=ignore 77605 nl_before_while=ignore 77606 nl_after_while=ignore 77607 nl_before_switch=ignore 77608 nl_after_switch=ignore 77609 nl_before_do=ignore 77610 nl_after_do=ignore 77611 pp_space=remove 77612 77613 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77614 77615commit 3505e1faadddeeec85a0d3f823c877ea33f86e00 77616Author: Jeremy Huddleston <jeremyhu@apple.com> 77617Date: Fri Feb 17 13:15:12 2012 -0800 77618 77619 XQuartz: Detect FatalErrors on startup to prevent tight crash loops 77620 77621 If a FatalError occurs before the server finishes launching, it will 77622 not have drained the launchd-owned DISPLAY socket, so launchd will 77623 just relaunch it. This can cause the server to crash in a tight loop 77624 which will spam the user with CrashReporter windows that claim focus on 77625 appearance. 77626 77627 This allows users stuck in this loop to "deal" with the problem without 77628 popping up a crash report every 10 seconds. 77629 77630 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77631 77632commit a818b305989bbcde4e585112a7ee70cbc0b14a92 77633Author: Jeremy Huddleston <jeremyhu@apple.com> 77634Date: Fri Feb 17 12:35:02 2012 -0800 77635 77636 os: Pass the FatalError message to OsVendorFatalError 77637 77638 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77639 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 77640 77641commit 0bb8a2566debd6cd11164df9ddca534150ec440a 77642Author: Jeremy Huddleston <jeremyhu@apple.com> 77643Date: Tue Mar 13 00:15:55 2012 -0700 77644 77645 XQuartz: Add a defaults option to disable the RENDER extension 77646 77647 RENDER has some ugly issues on XQuartz, so add an option to disable RENDER. 77648 77649 Enables workaround for: https://bugs.freedesktop.org/show_bug.cgi?id=26124 77650 77651 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77652 77653commit 5ff46d651ac0cd396be2053b79675997ed7af765 77654Author: Jeremy Huddleston <jeremyhu@apple.com> 77655Date: Sat Mar 17 00:04:27 2012 -0700 77656 77657 XQuartz: Use doubles for input valuators 77658 77659 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77660 77661commit cd84c0949a9f397d8b803d3b87c9d02b260d4795 77662Author: Jeremy Huddleston <jeremyhu@apple.com> 77663Date: Sat Mar 17 00:08:19 2012 -0700 77664 77665 XQuartz: Xi: darwinPointer is now Relative 77666 77667 There is really no real reason why this should be necessary, but wine 77668 developers are stuborn, so doing this to try to work around this wine 77669 issue: 77670 77671 http://bugs.winehq.org/show_bug.cgi?id=29732 77672 77673 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77674 77675commit 87939bf8311c0e7d168741fe57bc8fac414811d8 77676Author: Jeremy Huddleston <jeremyhu@apple.com> 77677Date: Tue Mar 13 00:25:53 2012 -0700 77678 77679 XQuartz: Move our logs into an X11 subdirectory 77680 77681 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77682 77683commit df327f98f6b8e02fadcfd313a85f906722acc572 77684Author: Jeremy Huddleston <jeremyhu@apple.com> 77685Date: Mon Mar 5 00:05:48 2012 -0800 77686 77687 test: Fix 'make dist' for configurations that build the Xorg DDX 77688 77689 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 77690 77691commit 0e8ee1cf4f3377a2916c43527ae0dc6cb07a67ec 77692Author: Dave Airlie <airlied@redhat.com> 77693Date: Fri Mar 23 14:54:05 2012 +0000 77694 77695 modesetting: fix build against older Xext 77696 77697commit a7eac500e652f30deffd9dc5e623fab701077738 77698Merge: bf876c87a d645edd11 77699Author: Peter Hutterer <peter.hutterer@who-t.net> 77700Date: Thu Mar 22 13:13:07 2012 +1000 77701 77702 Merge branch 'per-device-sync-counters' into for-keith 77703 77704commit d645edd11e7482f98c8b7e0d6c8693285c484907 77705Author: Peter Hutterer <peter.hutterer@who-t.net> 77706Date: Mon Mar 12 16:36:31 2012 +1000 77707 77708 Xext: Add per-device SyncCounters 77709 77710 Previously, we only had one idle alarm that was triggered for all devices, 77711 whenever the user used any device, came back from suspend, etc. 77712 77713 Add system SyncCounters for each device (named "DEVICEIDLETIME x", with x 77714 being the device id) that trigger on that device only. This allows for 77715 enabling/disabling devices based on interaction with other devices. 77716 77717 Popular use-case: disable the touchpad when the keyboard just above the 77718 touchpad stops being idle. 77719 77720 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77721 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77722 Reviewed-by: Jamey Sharp <jamey@minilop.net> 77723 Reviewed-by: James Jones <jajones@nvidia.com> 77724 77725commit 6aef209ebc2e54f5465da505a780f7b4cc273ee0 77726Author: Peter Hutterer <peter.hutterer@who-t.net> 77727Date: Mon Mar 12 13:51:02 2012 +1000 77728 77729 Change lastDeviceIdleTime to be per-device 77730 77731 Preparation work for per-device idle counters. 77732 77733 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77734 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77735 Reviewed-by: Jamey Sharp <jamey@minilop.net> 77736 Reviewed-by: James Jones <jajones@nvidia.com> 77737 77738commit 20cf0ef825e3f14b0688b691691e0aeba0a4860a 77739Author: Peter Hutterer <peter.hutterer@who-t.net> 77740Date: Tue Mar 13 10:21:23 2012 +1000 77741 77742 Xext: strdup() the SystemSyncCounter name 77743 77744 Required for future dynamic names. 77745 77746 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77747 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77748 Reviewed-by: Jamey Sharp <jamey@minilop.net> 77749 Reviewed-by: James Jones <jajones@nvidia.com> 77750 77751commit f1b28aea4156f0381ea733ad2afbdd1f34f75599 77752Author: Peter Hutterer <peter.hutterer@who-t.net> 77753Date: Tue Mar 13 09:31:09 2012 +1000 77754 77755 Xext: store the bracket values for idle counters in the private 77756 77757 And drop the three global variables, we have a reference to the counter 77758 everywhere now. 77759 77760 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77761 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77762 Reviewed-by: Jamey Sharp <jamey@minilop.net> 77763 Reviewed-by: James Jones <jajones@nvidia.com> 77764 77765commit d9553b2bbe06fba0b209218ffed9465edd79b4d2 77766Author: Peter Hutterer <peter.hutterer@who-t.net> 77767Date: Tue Mar 13 09:29:39 2012 +1000 77768 77769 Xext: pass the counter into block/wakeup handlers 77770 77771 No functional changes, currently unused. Preparation work, we don't need a 77772 global variable if we can pass the counters around anyway. 77773 77774 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77775 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77776 Reviewed-by: Jamey Sharp <jamey@minilop.net> 77777 Reviewed-by: James Jones <jajones@nvidia.com> 77778 77779commit 90e6dc6de1ac339212989fd8e54131b196ebb369 77780Author: Peter Hutterer <peter.hutterer@who-t.net> 77781Date: Tue Mar 13 09:28:15 2012 +1000 77782 77783 Xext: add a private field to SyncSystemCounters 77784 77785 Will be used to store counter-specific data. 77786 77787 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77788 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77789 Reviewed-by: Jamey Sharp <jamey@minilop.net> 77790 Reviewed-by: James Jones <jajones@nvidia.com> 77791 77792commit 3ddae647c307005309daa2d5dfe4bc6acb8170ab 77793Author: Peter Hutterer <peter.hutterer@who-t.net> 77794Date: Mon Mar 12 15:31:39 2012 +1000 77795 77796 Xext: localise pIdleTimeValueLess/Greater 77797 77798 Cleanup for future features, no functional changes. 77799 77800 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77801 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77802 Reviewed-by: Jamey Sharp <jamey@minilop.net> 77803 Reviewed-by: James Jones <jajones@nvidia.com> 77804 77805commit 43eb2f2758dfc6ca5a49afce97cc5baea8caf9f5 77806Author: Peter Hutterer <peter.hutterer@who-t.net> 77807Date: Mon Mar 12 15:27:56 2012 +1000 77808 77809 Xext: localise use of IdleTimeCounter 77810 77811 Instead of referring to the global IdleTimeCounter everywhere, only do it 77812 once and use a local variable for the rest. 77813 77814 Cleanup for future features, no functional changes. 77815 77816 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77817 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77818 Reviewed-by: Jamey Sharp <jamey@minilop.net> 77819 Reviewed-by: James Jones <jajones@nvidia.com> 77820 77821commit 1f12f059ef994e0b9b68fbd1f1556d0285c96b8b 77822Author: Jamey Sharp <jamey@minilop.net> 77823Date: Wed Mar 14 17:22:18 2012 -0700 77824 77825 sync: Use a linked list instead of an array for SysCounterList. 77826 77827 Signed-off-by: Jamey Sharp <jamey@minilop.net> 77828 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 77829 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77830 77831commit bf876c87a9099fdfa63ed599f8ed9a954dd023d9 77832Merge: 908ab3d58 c0b0a9bce 77833Author: Peter Hutterer <peter.hutterer@who-t.net> 77834Date: Thu Mar 22 11:34:43 2012 +1000 77835 77836 Merge branch 'dtrace-input-abi' into for-keith 77837 77838commit c0b0a9bce9237b0abe150c1a7b54939affecc751 77839Author: Peter Hutterer <peter.hutterer@who-t.net> 77840Date: Mon Mar 5 14:12:52 2012 +1000 77841 77842 dix: add dtrace probes to input API 77843 77844 For driver debugging, it is helpful to know whether the driver has actually 77845 submitted an event to the server. dtrace hooks can help here. 77846 77847 Note that GetPointerEvents and friends may also be triggered by the server 77848 for other emulated devices, some care must be taken when analysing the 77849 results. 77850 77851 Additional difficulty: proximity events have a run-time assigned type, so 77852 this may make automatic detection a tad harder. If in doubt, go for any 77853 event > 64 since the only two that can have that value are ProximityIn and 77854 ProximityOut. 77855 77856 An example systemtap script is below: 77857 77858 # Compile+run with 77859 # stap -g xorg.stp /usr/bin/Xorg 77860 # 77861 77862 function print_valuators:string(nvaluators:long, mask_in:long, valuators_in:long) %{ 77863 int i; 77864 unsigned char *mask = (unsigned char*)THIS->mask_in; 77865 double *valuators = (double*)THIS->valuators_in; 77866 char str[128] = {0}; 77867 char *s = str; 77868 77869 #define BitIsSet(ptr, bit) (((unsigned char*)(ptr))[(bit)>>3] & (1 << ((bit) & 7))) 77870 77871 s += sprintf(s, "nval: %d ::", (int)THIS->nvaluators); 77872 for (i = 0; i < THIS->nvaluators; i++) 77873 { 77874 s += sprintf(s, " %d: ", i); 77875 if (BitIsSet(mask, i)) 77876 s += sprintf(s, "%d", (int)valuators[i]); 77877 } 77878 77879 sprintf(THIS->__retvalue, "%s", str); 77880 %} 77881 77882 probe process(@1).mark("input__event") 77883 { 77884 deviceid = $arg1 77885 type = $arg2 77886 detail = $arg3 77887 flags = $arg4 77888 nvaluators = $arg5 77889 77890 str = print_valuators(nvaluators, $arg6, $arg7) 77891 printf("Event: device %d type %d detail %d flags %#x %s\n", 77892 deviceid, type, detail, flags, str); 77893 } 77894 77895 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77896 Acked-by: Jeremy Huddleston <jeremyhu@apple.com> 77897 77898commit 61cb98da1c6199964825de158d9eee7682d9c983 77899Author: Peter Hutterer <peter.hutterer@who-t.net> 77900Date: Mon Mar 12 16:03:50 2012 +1000 77901 77902 Xext: SyncCreateSystemCounter returns a SyncCounter* 77903 77904 type safety++ 77905 77906 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77907 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77908 Reviewed-by: Bryce Harrington <bryce@canonical.com> 77909 77910commit 908ab3d580188533168c8cdfd2cab9dc689b4218 77911Author: Peter Hutterer <peter.hutterer@who-t.net> 77912Date: Wed Mar 21 14:05:29 2012 +1000 77913 77914 dix: set raw event values before adding up relative values (#46976) 77915 77916 Regression introduced in 4e52cc0ef48145134cd58d357fb7289e6f8bb709 77917 77918 Raw event values are values as-is from the driver, modified only be 77919 transformation or acceleration. 4e52cc caused the mask to be updated from 77920 relative to absolute coordinates which then got written into the raw events. 77921 77922 Move the raw event update into the respective branches for absolute/relative 77923 events. 77924 77925 X.Org Bug 46976 <http://bugs.freedesktop.org/show_bug.cgi?id=46976> 77926 77927 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77928 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 77929 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 77930 Tested-by: Sven Arvidsson <sa@whiz.se> 77931 Reviewed-by: Simon Thum <simon.thum@gmx.de> 77932 77933commit 45fe3085f8f45e529383025414fdd263d86dfd2b 77934Author: Peter Hutterer <peter.hutterer@who-t.net> 77935Date: Mon Mar 12 14:08:39 2012 +1000 77936 77937 Xext: remove needless /* parameter */ comments in declaration 77938 77939 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77940 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77941 Reviewed-by: Bryce Harrington <bryce@canonical.com> 77942 77943commit 9c3bd3ae652af386b6821b197d24528f20ba867d 77944Author: Peter Hutterer <peter.hutterer@who-t.net> 77945Date: Wed Mar 21 14:03:27 2012 +1000 77946 77947 dix: fix compiler warning "unused variable 'scr'" 77948 77949 getevents.c: In function 'updateSlaveDeviceCoords': 77950 getevents.c:326:15: warning: unused variable 'scr' [-Wunused-variable] 77951 77952 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77953 77954commit e21ffff4761d2b75815391c6947adcba425ab11e 77955Author: Peter Hutterer <peter.hutterer@who-t.net> 77956Date: Mon Mar 12 14:07:44 2012 +1000 77957 77958 Xext: typedef QueryValue and BracketValue prototypes 77959 77960 No functional changes, just for readability 77961 77962 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77963 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77964 Reviewed-by: Bryce Harrington <bryce@canonical.com> 77965 77966commit 14e3ea730eed344e6ed69c873f918d6f076d13dc 77967Author: Peter Hutterer <peter.hutterer@who-t.net> 77968Date: Wed Mar 14 13:48:56 2012 +1000 77969 77970 include: add an X_DEBUG message type 77971 77972 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77973 Reviewed-by: Jamey Sharp <jamey@minilop.net> 77974 77975commit 5910f2df58beaae2187438fef0b62c29a563e853 77976Author: Peter Hutterer <peter.hutterer@who-t.net> 77977Date: Mon Mar 12 16:26:29 2012 +1000 77978 77979 Xext: drop InitServertime() declaration. 77980 77981 Not implemented anywhere. 77982 77983 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 77984 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 77985 Reviewed-by: Bryce Harrington <bryce@canonical.com> 77986 77987commit 31df08a449cf9b6e1740e1c02257997490630d93 77988Author: Chase Douglas <chase.douglas@canonical.com> 77989Date: Wed Mar 7 16:06:27 2012 -0800 77990 77991 Use a new sprite trace for indirect touches when all touches have physically ended 77992 77993 All touches of an indirect device, such as a trackpad, are sent to the 77994 same window set. When there are no active touches, a new window set is 77995 created; otherwise, the window set of an existing touch is copied. 77996 77997 The current code checks for any logically active touches. This includes 77998 touches that have physically ended but are still logically active due to 77999 unhandled touch grabs. Instead, we want a new window set whenever there 78000 are no physically active touches. 78001 78002 This change skips over logically active but pending end touches, which 78003 are touches that have physically ended. 78004 78005 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78006 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78007 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78008 78009commit 58427e08a4a36ce9e213e4b4fe5249d5db2c764d 78010Author: Chase Douglas <chase.douglas@canonical.com> 78011Date: Wed Mar 7 16:06:26 2012 -0800 78012 78013 Xi: Fix TouchEnd to TouchUpdate change for one accepted grab 78014 78015 If there is only one listener of a touch, the listener is a grab, and is 78016 accepted before the touch has ended, the current code will not end the 78017 touch record when the touch does end. 78018 78019 This change adds a listener state for when a touch is accepted but has 78020 not yet ended. We now keep the touch record alive in this state, but end 78021 it when the touch ends. 78022 78023 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78024 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78025 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78026 78027commit e884ff8ad4df2b3272a3d8ece772906207af5142 78028Author: Peter Hutterer <peter.hutterer@who-t.net> 78029Date: Wed Mar 7 19:08:33 2012 +1000 78030 78031 xfree86: remove out-of-date comment 78032 78033 No idea what this was referring to but it goes past git history. 78034 78035 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78036 78037commit 5497ce3da442d27c2dc7796bfef6ccd670bbadc4 78038Author: Peter Hutterer <peter.hutterer@who-t.net> 78039Date: Mon Feb 20 12:09:33 2012 +1000 78040 78041 dix: IsFloating() on master devices is always false 78042 78043 There are a few subtle bugs during startup where IsFloating() returns true 78044 if the device is a master device that is not yet paired with its keyboard 78045 device. 78046 78047 Force IsFloating() to always return FALSE for master devices, that was the 78048 intent after all and any code that relies on the other behaviour should be 78049 fixed instead. 78050 78051 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78052 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 78053 Tested-by: Jon TURNEY <jon.turney@dronecode.org.uk> 78054 78055commit 2c23ef83b0e03e163aeeb06133538606886f4e9c 78056Author: Peter Hutterer <peter.hutterer@who-t.net> 78057Date: Mon Feb 27 08:01:07 2012 +1000 78058 78059 Xi: prohibit multiple XIQueryVersion requests with different versions 78060 78061 Return BadValue if major or minor differs on the second call. 78062 78063 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78064 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 78065 78066commit eb84c154ed38194c32651727b6dfe2d1bde4c599 78067Author: Peter Hutterer <peter.hutterer@who-t.net> 78068Date: Mon Feb 27 10:09:44 2012 +1000 78069 78070 dix: when rescaling from master, rescale from desktop dimensions (#46657) 78071 78072 master->last.valuators[] is in desktop dimensions, so use those as 78073 rescale axis ranges, not the screen. Otherwise, a rescale on any screen 78074 not the top-left will cause out-of-bounds coordinates which will always 78075 map to the bottom-right screen, causing the device to be stuck on that 78076 screen. 78077 78078 X.Org Bug 46657 <http://bugs.freedesktop.org/show_bug.cgi?id=46657> 78079 78080 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78081 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 78082 78083commit 6b6afd3d013e3f4910fae3520d1d786df2b0e47a 78084Author: Peter Hutterer <peter.hutterer@who-t.net> 78085Date: Thu Feb 16 15:11:40 2012 +1000 78086 78087 Xext: return BadAccess if PickPointer fails (#45796) 78088 78089 PickPointer or PickKeyboard return NULL, all MDs are currently disabled and 78090 we cannot emulate a core event. This wasn't anticipated by the protocol, so 78091 we don't really have an error code we may use here - BadAccess is simply the 78092 least bad of the possible ones. 78093 78094 And returning BadAccess beats crashing the server. 78095 78096 X.Org Bug 45796 <http://bugs.freedesktop.org/show_bug.cgi?id=45796> 78097 78098 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78099 78100commit 9e017cf0cf1f0c9d0d9c2cfeb82ea5dc0eb5905e 78101Author: Andreas Wettstein <wettstein509@solnet.ch> 78102Date: Sat Feb 25 20:48:17 2012 +0100 78103 78104 XKB: Redirect actions defunct with Gtk3 (XInput?) 78105 78106 When redirect actions are used with Gtk3, Gtk3 complained about 78107 events not holding a GdkDevice. This was caused by device IDs 78108 not being set for redirect actions. 78109 78110 More seriously, Gtk3 did not receive state changes redirect 78111 actions might specify. This was because event_set_state in 78112 dix/inpututils.c accesses the prev_state field, but the changes 78113 for the redirect action were only put into the state field. 78114 78115 Signed-off-by: Andreas Wettstein <wettstein509@solnet.ch> 78116 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78117 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78118 78119commit ab3a815a75ab5695753fa37a98b0ea5293d4cb91 78120Author: Daniel Stone <daniel@fooishbar.org> 78121Date: Thu Mar 15 15:18:29 2012 +0000 78122 78123 Indentation: Change '& stuff' to '&stuff' 78124 78125 If the typedef wasn't perfect, indent would get confused and change: 78126 foo = (SomePointlessTypedef *) &stuff[1]; 78127 to: 78128 foo = (SomePointlessTypedef *) & stuff[1]; 78129 78130 Fix this up with a really naïve sed script, plus some hand-editing to 78131 change some false positives in XKB back. 78132 78133 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 78134 78135commit 58b1f739d73b03ff7952ca986ed8746a7307fffe 78136Author: Daniel Stone <daniel@fooishbar.org> 78137Date: Thu Mar 15 15:13:51 2012 +0000 78138 78139 Xinerama: Fix up obnoxiously-indented code 78140 78141 Comments inside of conditionals confuse indent to death. Which is fair 78142 enough really. 78143 78144 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 78145 78146commit 9838b7032ea9792bec21af424c53c07078636d21 78147Author: Keith Packard <keithp@keithp.com> 78148Date: Wed Mar 21 12:55:09 2012 -0700 78149 78150 Introduce a consistent coding style 78151 78152 This is strictly the application of the script 'x-indent-all.sh' 78153 from util/modular. Compared to the patch that Daniel posted in 78154 January, I've added a few indent flags: 78155 78156 -bap 78157 -psl 78158 -T PrivatePtr 78159 -T pmWait 78160 -T _XFUNCPROTOBEGIN 78161 -T _XFUNCPROTOEND 78162 -T _X_EXPORT 78163 78164 The typedefs were needed to make the output of sdksyms.sh match the 78165 previous output, otherwise, the code is formatted badly enough that 78166 sdksyms.sh generates incorrect output. 78167 78168 The generated code was compared with the previous version and found to 78169 be essentially identical -- "assert" line numbers and BUILD_TIME were 78170 the only differences found. 78171 78172 The comparison was done with this script: 78173 78174 dir1=$1 78175 dir2=$2 78176 78177 for dir in $dir1 $dir2; do 78178 (cd $dir && find . -name '*.o' | while read file; do 78179 dir=`dirname $file` 78180 base=`basename $file .o` 78181 dump=$dir/$base.dump 78182 objdump -d $file > $dump 78183 done) 78184 done 78185 78186 find $dir1 -name '*.dump' | while read dump; do 78187 otherdump=`echo $dump | sed "s;$dir1;$dir2;"` 78188 diff -u $dump $otherdump 78189 done 78190 78191 Signed-off-by: Keith Packard <keithp@keithp.com> 78192 Acked-by: Daniel Stone <daniel@fooishbar.org> 78193 Acked-by: Alan Coopersmith <alan.coopersmith@oracle.com> 78194 78195commit 75199129c603fc8567185ac31866c9518193cb78 78196Author: Keith Packard <keithp@keithp.com> 78197Date: Wed Mar 21 12:33:19 2012 -0700 78198 78199 Handle blank betweeen type and name in sdksyms.sh 78200 78201 indent sometimes adds a blank line between the type and the name in a 78202 function declaration that includes _X_EXPORT, so handle that before 78203 the files are re-indented. 78204 78205 Signed-off-by: Keith Packard <keithp@keithp.com> 78206 78207commit a615b90cab7569fae9d123e4da1d3373c871d84b 78208Author: Keith Packard <keithp@keithp.com> 78209Date: Wed Mar 14 11:32:36 2012 -0700 78210 78211 Bump version number to 1.12.99.0 78212 78213 Now that 1.12 has branched, reset the version on master to a 78214 development number. 78215 78216 Signed-off-by: Keith Packard <keithp@keithp.com> 78217 78218commit 53204d5c8bd83c957fbdb7d3dc2891118ed7658b 78219Author: Sascha Hauer <s.hauer@pengutronix.de> 78220Date: Sat Mar 3 14:09:25 2012 +0100 78221 78222 make busID non mandatory 78223 78224 Currently the driver only probes a device when it has a 78225 busID. The busID is optional so don't depend on it. 78226 78227 Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> 78228 Signed-off-by: Dave Airlie <airlied@redhat.com> 78229 78230commit bb7e39c1f2112f5fb7e87baddb114ab164fbc749 78231Author: Sascha Hauer <s.hauer@pengutronix.de> 78232Date: Sat Mar 3 14:09:27 2012 +0100 78233 78234 do not bail out on non pci devices 78235 78236 To make the driver work on nin PCI devices we shouldn't bail 78237 out in this case. 78238 78239 Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> 78240 Signed-off-by: Dave Airlie <airlied@redhat.com> 78241 78242commit 2b6848fcb3f85f9ce18a6de1dc6c3ac047101aa8 78243Author: Sascha Hauer <s.hauer@pengutronix.de> 78244Date: Sat Mar 3 14:09:26 2012 +0100 78245 78246 Fix non PCI device probing 78247 78248 When no devicename is found in the option then the driver probes 78249 by PciInfo no matter if it's valid or not. Instead of doing this 78250 use PciInfo only when it's valid and fall back to the devicename 78251 otherwise. With devicename probing use open_hw() to fall back 78252 on the KMSDEVICE environment variable or to the default device. 78253 78254 Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> 78255 Signed-off-by: Dave Airlie <airlied@redhat.com> 78256 78257commit 456a001e3f7a24f443cf0bad8400d5d600e2ad46 78258Author: Sascha Hauer <s.hauer@pengutronix.de> 78259Date: Sat Mar 3 14:09:24 2012 +0100 78260 78261 introduce open_hw() function 78262 78263 probe_hw opens the hardware in the order we want it: 78264 first try devname, if this is NULL then try the KMSDEVICE 78265 environment variable and as a last fallback use "/dev/dri/card0". 78266 Instead of implementing the same code again when really opening 78267 the device move the code to a open_hw() function and let probe_hw 78268 use it. 78269 78270 Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> 78271 Signed-off-by: Dave Airlie <airlied@redhat.com> 78272 78273commit aa6ceaaa843525b2243569de162ed0b17faa7510 78274Author: Sascha Hauer <s.hauer@pengutronix.de> 78275Date: Sat Mar 3 14:09:23 2012 +0100 78276 78277 fix if() brackets in Probe function 78278 78279 in Probe() the indention shows what's meant but there are no 78280 brackets. Add them. 78281 78282 Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> 78283 Signed-off-by: Dave Airlie <airlied@redhat.com> 78284 78285commit b1be72c5ca6cb98ba64637990b142be0f1710a19 78286Author: Keith Packard <keithp@keithp.com> 78287Date: Sun Mar 4 20:26:18 2012 -0800 78288 78289 Version bumped to 1.12 78290 78291 Signed-off-by: Keith Packard <keithp@keithp.com> 78292 78293commit e08ed0b757b9b48344a301f612fabb3e39ffec78 78294Author: Gaetan Nadon <memsize@videotron.ca> 78295Date: Sun Feb 26 17:51:50 2012 -0500 78296 78297 test: add new test cases to .gitignore 78298 78299 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78300 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78301 Signed-off-by: Keith Packard <keithp@keithp.com> 78302 78303commit 2416ee4a015068359807a10f433e8c54192c78a9 78304Author: Peter Hutterer <peter.hutterer@who-t.net> 78305Date: Wed Feb 22 15:32:56 2012 +1000 78306 78307 dix: avoid NULL-pointer dereference on button-only devices (#38313) 78308 78309 And for such devices simply take the last.valuators[] which must be valid at 78310 all times anyway. UpdateSlaveDeviceCoords takes care of that. 78311 78312 X.Org Bug 38313 <http://bugs.freedesktop.org/show_bug.cgi?id=38313> 78313 78314 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78315 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 78316 78317commit 38000e7d1f958f5944e641de3e716944a5876d41 78318Author: Jeremy Huddleston <jeremyhu@apple.com> 78319Date: Tue Jan 31 11:06:21 2012 -0800 78320 78321 Revert "dix: don't XWarpPointer through the last slave anymore (#38313)" 78322 78323 This reverts commit 2bfb802839688ecf328119c4c6979390fc60348d. 78324 78325 This commit caused a regression. 78326 78327 See: http://xquartz.macosforge.org/trac/ticket/517#comment:10 78328 78329 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 78330 78331commit cb6a32da27f09261c4d561c19d1877c750d98cc3 78332Author: Jeremy Huddleston <jeremyhu@apple.com> 78333Date: Fri Feb 17 13:44:12 2012 -0800 78334 78335 XQuartz: Short-circuit activateX: 78336 78337 This also avoids a deadlock when calling activateX: before the server 78338 thread has initialized 78339 78340 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 78341 78342commit 1562c4a1ca748d2c4a8c69d57ba09ff7ca9cc453 78343Author: Jeremy Huddleston <jeremyhu@apple.com> 78344Date: Fri Feb 17 13:54:57 2012 -0800 78345 78346 XQuartz: Actually install the new locales 78347 78348 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 78349 78350commit 1349a2b7ab04553c8fddb60534400dea5cbd3cca 78351Author: Dave Airlie <airlied@redhat.com> 78352Date: Wed Feb 22 10:07:06 2012 +0000 78353 78354 xf86-video-modesetting 0.2.0 78355 78356commit e5648616d222966a95edd6c35807dcf5b91a0f96 78357Author: Dave Airlie <airlied@redhat.com> 78358Date: Wed Feb 22 09:59:12 2012 +0000 78359 78360 modesetting: fix warnings, remove dead code. 78361 78362 Signed-off-by: Dave Airlie <airlied@redhat.com> 78363 78364commit f7724167746d15c0ca32a7f5f95fb594addcddd1 78365Author: Gaetan Nadon <memsize@videotron.ca> 78366Date: Sun Feb 19 16:13:34 2012 -0500 78367 78368 config: layout and comment the top portion of configure.ac 78369 78370 Reorder statements to be consistent with other modules so things 78371 are easier to find. 78372 78373 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78374 Signed-off-by: Dave Airlie <airlied@redhat.com> 78375 78376commit a0bb835dc7db1ddd0472058a2a05a600ddea3930 78377Author: Gaetan Nadon <memsize@videotron.ca> 78378Date: Sun Feb 19 08:36:19 2012 -0500 78379 78380 config: replace deprecated AC_HELP_STRING with AS_HELP_STRING 78381 78382 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78383 Signed-off-by: Dave Airlie <airlied@redhat.com> 78384 78385commit d56293cae78323b8976859c4461f8809a76a0b03 78386Author: Gaetan Nadon <memsize@videotron.ca> 78387Date: Sun Feb 19 08:36:17 2012 -0500 78388 78389 make: add all warnings according to the platform 78390 78391 The current code only adds -Wall and only for gcc. 78392 Automake reserves the use of CPPFLAGS for the user to override 78393 on the command line. 78394 This also breaks the option --enable-strict-compilation 78395 78396 The variable CWARNFLAGS contains the complete set of warnings 78397 and is platform sensitive. 78398 78399 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78400 Signed-off-by: Dave Airlie <airlied@redhat.com> 78401 78402commit fa201fe8299099f7192a4399c9df245efcee3f8a 78403Author: Gaetan Nadon <memsize@videotron.ca> 78404Date: Sun Feb 19 08:36:16 2012 -0500 78405 78406 make: remove empty variable assignment 78407 78408 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78409 Signed-off-by: Dave Airlie <airlied@redhat.com> 78410 78411commit e159fbd8805005bb7a33bf2076fd2cef00e987c6 78412Author: Gaetan Nadon <memsize@videotron.ca> 78413Date: Sun Feb 19 08:36:15 2012 -0500 78414 78415 Remove unneeded AM_PROG_CC_C_CO 78416 78417 There are no objects in subdirs or compiled multiple times 78418 with different flags. 78419 78420 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78421 Signed-off-by: Dave Airlie <airlied@redhat.com> 78422 78423commit 70558492376e66b73063e51b128b5bbad14b84f4 78424Author: Gaetan Nadon <memsize@videotron.ca> 78425Date: Sun Feb 19 08:34:28 2012 -0500 78426 78427 Remove redundant AC_PROG_CC 78428 78429 Already covered by XORG_DEFAULT_OPTIONS 78430 78431 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78432 Signed-off-by: Dave Airlie <airlied@redhat.com> 78433 78434commit d4bb99ce754073de090221d98000f609a3891380 78435Author: Gaetan Nadon <memsize@videotron.ca> 78436Date: Sun Feb 19 08:34:27 2012 -0500 78437 78438 Replace obsolete AM_CONFIG_HEADER with AC_CONFIG_HEADERS 78439 78440 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78441 Signed-off-by: Dave Airlie <airlied@redhat.com> 78442 78443commit 5c2d152cad66c3f8ed6a4e82c3e6a45269bbb66d 78444Author: Gaetan Nadon <memsize@videotron.ca> 78445Date: Sun Feb 19 08:34:26 2012 -0500 78446 78447 Add missing targets for ChangeLog and INSTALL 78448 78449 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78450 Signed-off-by: Dave Airlie <airlied@redhat.com> 78451 78452commit 8673f545cdb465b6e74f1d2fcabe29bc7fa0243d 78453Author: Gaetan Nadon <memsize@videotron.ca> 78454Date: Sun Feb 19 08:34:25 2012 -0500 78455 78456 Remove redundant EXTRA_DIST for the README file. 78457 78458 Autotools know about it, it is always distributed. 78459 78460 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78461 Signed-off-by: Dave Airlie <airlied@redhat.com> 78462 78463commit 71594746c7da32e1c7986341a0da30f241a7a2be 78464Merge: d53235af8 eaba06a27 78465Author: Keith Packard <keithp@keithp.com> 78466Date: Wed Feb 22 18:07:20 2012 +1300 78467 78468 Merge remote-tracking branch 'whot/for-keith' 78469 78470commit 88b637e59a37af8ec6c4c0783504b8f1a49e5595 78471Author: Gaetan Nadon <memsize@videotron.ca> 78472Date: Sun Feb 19 08:34:24 2012 -0500 78473 78474 make: remove redundant AUTOMAKE_OPTIONS 78475 78476 Already covered by AM_INIT_AUTOMAKE([foreign dist-bzip2]) 78477 78478 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78479 Signed-off-by: Dave Airlie <airlied@redhat.com> 78480 78481commit a7ea959094e9a8d862637a0ea515f687aaf89d29 78482Author: Gaetan Nadon <memsize@videotron.ca> 78483Date: Sun Feb 19 08:34:23 2012 -0500 78484 78485 Fix typo in .gitignore 78486 78487 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78488 Signed-off-by: Dave Airlie <airlied@redhat.com> 78489 78490commit 861e2b444ffb24921e07fdc25ab2679c9b6d2f4d 78491Author: Gaetan Nadon <memsize@videotron.ca> 78492Date: Sun Feb 19 08:34:22 2012 -0500 78493 78494 man: missing AM_V_GEN and hard-coded sed command name 78495 78496 Let's use the common xorg makefile for all drivers. 78497 This ensures no new problems are introduced. 78498 Improvements are welcome and to be applied to all drivers. 78499 78500 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78501 Signed-off-by: Dave Airlie <airlied@redhat.com> 78502 78503commit 453bbe5b442a9c21d7c22437545e487e2993ceb7 78504Author: Gaetan Nadon <memsize@videotron.ca> 78505Date: Sun Feb 19 08:34:21 2012 -0500 78506 78507 Add contact information to the README file 78508 78509 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78510 Signed-off-by: Dave Airlie <airlied@redhat.com> 78511 78512commit dd1d7e3519ea2f148be83ef7ab0b228da95ff9c3 78513Author: Gaetan Nadon <memsize@videotron.ca> 78514Date: Sun Feb 19 08:34:20 2012 -0500 78515 78516 Fill the COPYING file with license text 78517 78518 This reflects the copyright license text in the source code 78519 78520 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 78521 Signed-off-by: Dave Airlie <airlied@redhat.com> 78522 78523commit b0e12e250d4b3438fb3306155a7bbff3e2d7f9b7 78524Author: Dave Airlie <airlied@redhat.com> 78525Date: Mon Feb 20 11:08:40 2012 +0000 78526 78527 modesetting: disable dirty updates for ENOSYS 78528 78529 the kernel can also return ENOSYS for this to say its not used. 78530 78531 Signed-off-by: Dave Airlie <airlied@redhat.com> 78532 78533commit c5529d68c5b01cf0f36d8f2ce3694a7a0f3333da 78534Author: Dave Airlie <airlied@redhat.com> 78535Date: Mon Feb 20 11:05:59 2012 +0000 78536 78537 modesetting: fix shadow resizing. 78538 78539 if we hotplugged and output, the shadow got disabled by accident. 78540 78541 Signed-off-by: Dave Airlie <airlied@redhat.com> 78542 78543commit f3b9e52b29e8b8c4b12bc2ce290ab44e88750ad9 78544Author: Dave Airlie <airlied@redhat.com> 78545Date: Mon Feb 20 11:00:56 2012 +0000 78546 78547 modesetting: move shadow stuff to other structure 78548 78549 we need this for resize to work properly. 78550 78551 Signed-off-by: Dave Airlie <airlied@redhat.com> 78552 78553commit eb44a004b7230321ce6837ef44610153b1d2017b 78554Author: Dave Airlie <airlied@redhat.com> 78555Date: Mon Feb 20 10:54:36 2012 +0000 78556 78557 modesetting: fix stride if kernel modifies it. 78558 78559 If the kernel aligns things use its values instead. 78560 78561 fixes output on nouveau here. 78562 78563 Signed-off-by: Dave Airlie <airlied@redhat.com> 78564 78565commit 2150cbda8c75d348f39926bbc23d3e4ca9b42c88 78566Author: Dave Airlie <airlied@redhat.com> 78567Date: Mon Feb 20 10:54:21 2012 +0000 78568 78569 shadowfb: dump shadowfb state at startup 78570 78571commit 86080a5f7a589e55a485c42bab47c6c8ff6428c1 78572Author: Dave Airlie <airlied@redhat.com> 78573Date: Thu Feb 16 19:41:40 2012 +0000 78574 78575 use a cap to decide if shadow is preferred or not. 78576 78577commit fc5f39acd49d023a82ab05fffde1b423b697753c 78578Author: Dave Airlie <airlied@redhat.com> 78579Date: Thu Feb 16 19:41:30 2012 +0000 78580 78581 cleanup dumb cap fetch 78582 78583commit d94b3eaba9a33d0d78ed2d9acbc33394f40fdc2e 78584Author: Dave Airlie <airlied@redhat.com> 78585Date: Mon Nov 14 11:22:44 2011 +0000 78586 78587 modesetting: add cursor fallback if kernel rejects cursor. 78588 78589 If the kernel rejects a cursor, cause a fallback, this isn't 100% as 78590 we can lose the initial cursor, but it works fine once wm starts. 78591 78592 Signed-off-by: Dave Airlie <airlied@redhat.com> 78593 78594commit 594b3c4d6f6f748e08baffc4bc1c3ffad9b0d9cf 78595Author: Dave Airlie <airlied@redhat.com> 78596Date: Thu Feb 16 19:31:33 2012 +0000 78597 78598 fix some whitespace 78599 78600commit 2e297978c928c571cf0a80e8d38423089014f07e 78601Author: Dave Airlie <airlied@redhat.com> 78602Date: Wed Oct 5 15:12:43 2011 +0100 78603 78604 fixup device open 78605 78606 Signed-off-by: Dave Airlie <airlied@redhat.com> 78607 78608commit eaba06a27c5520a02f08431ac1e4b0e0bdc22cd8 78609Author: Chase Douglas <chase.douglas@canonical.com> 78610Date: Mon Feb 13 16:00:48 2012 -0800 78611 78612 Keep virtual core pointer touch class around if new slave doesn't have one 78613 78614 The VCP may have active touch grabs. The touch records must be kept so 78615 these touch grabs may be accepted/rejected in the future. This means the 78616 touch class list will not represent the touch class of the attached 78617 slave device if it does not have a touch class, but we already were 78618 breaking that assumption by keeping a separate touches array for the 78619 VCP. 78620 78621 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78622 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78623 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78624 78625commit 8573b3519af138a3a12a2e77098718165f9fd8ff 78626Author: Peter Hutterer <peter.hutterer@who-t.net> 78627Date: Mon Feb 13 16:00:47 2012 -0800 78628 78629 Don't clobber virtual core pointer touches array length 78630 78631 The VCP has its own touches array, don't overwrite it when the class is 78632 copied from the SD to the master. 78633 78634 Reported-by: Chase Douglas <chase.douglas@canonical.com> 78635 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78636 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 78637 78638commit 03d32fe7a718d9016053cdb5d57f51a74ef99b59 78639Author: Chase Douglas <chase.douglas@ubuntu.com> 78640Date: Mon Feb 13 12:09:32 2012 -0800 78641 78642 Don't dereference a touch after it has been ended when punting to next owner 78643 78644 In this case, we have ended the touch because the last owner has 78645 rejected it. We need to return from the function right now so we don't 78646 attempt to dereference another touch client for early acceptance 78647 processing. 78648 78649 Signed-off-by: Chase Douglas <chase.douglas@ubuntu.com> 78650 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 78651 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78652 78653commit 1ecb7aaf2adedad1996cd26176ef5802113e3ad9 78654Author: Chase Douglas <chase.douglas@ubuntu.com> 78655Date: Sat Feb 11 07:58:38 2012 -0800 78656 78657 Focus event button state must show the logical buttons, not physical buttons 78658 78659 Similar to the fix in fcda98c48610fd507ca0b89c6006a5497d9dc1c9. This 78660 ensures we show the correct logical state of the buttons in device focus 78661 events too. 78662 78663 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78664 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78665 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78666 78667commit 6f28388187cffae9e5bc9bfc9425acff4f478b59 78668Author: Peter Hutterer <peter.hutterer@who-t.net> 78669Date: Sat Feb 11 01:29:26 2012 +1000 78670 78671 dix: reset last.scroll when resetting the valuator (#45611) 78672 78673 last.scroll remained on the last-submitted scrolling value but last.valuator 78674 was changed whenever the slave device changed. The first scrolling delta 78675 after a switch was then calculated as (last.scroll - new abs value), causing 78676 erroneous scrolling events. 78677 78678 Test case: 78679 - synaptics with a scrolling method enabled, other device with 3+ axes (e.g. 78680 wacom) 78681 - scroll on touchpad 78682 - use other device 78683 - scroll on touchpad 78684 78685 The second scroll caused erroneous button press/release events. 78686 78687 X.Org Bug 45611 <http://bugs.freedesktop.org/show_bug.cgi?id=45611> 78688 78689 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78690 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 78691 78692commit d53235af85d50774c68347720ce132daf9a5bc49 78693Author: Jeremy Huddleston <jeremyhu@apple.com> 78694Date: Mon Feb 13 23:06:07 2012 -0800 78695 78696 XQuartz: Build fix for possible conflict of BOOL type 78697 78698 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 78699 78700commit 62edd970f6f36058fcb31de4555eb7c1329cce74 78701Author: Keith Packard <keithp@keithp.com> 78702Date: Sat Feb 11 16:13:44 2012 +1300 78703 78704 Bump to version 1.11.99.903 (1.12 RC3) 78705 78706 Signed-off-by: Keith Packard <keithp@keithp.com> 78707 78708commit 42b6756463ee0476340656707f1088dc6c2fd220 78709Merge: 7674d00b0 ca64912c0 78710Author: Keith Packard <keithp@keithp.com> 78711Date: Sat Feb 11 15:36:43 2012 +1300 78712 78713 Merge remote-tracking branch 'alanc/master' 78714 78715commit 7674d00b04da5cf73cfa5c7ed1d3a9f42b59960e 78716Author: Peter Hutterer <peter.hutterer@who-t.net> 78717Date: Fri Jan 27 12:41:09 2012 +1000 78718 78719 Xi: handle new XIAllowEvents request in inputproto 2.1.99.6 78720 78721 grab_window and touchid were removed from the struct for ABI compatibility 78722 reasons, we need to pull in the new, XI 2.2-specific struct. 78723 78724 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78725 Reviewed-by: Keith Packard <keithp@keithp.com> 78726 Signed-off-by: Keith Packard <keithp@keithp.com> 78727 78728commit ceb026c6a6f1eea8b34e745f06f1ebcd652c0ba1 78729Author: Peter Hutterer <peter.hutterer@who-t.net> 78730Date: Fri Jan 27 12:41:09 2012 +1000 78731 78732 Xi: handle new XIAllowEvents request in inputproto 2.1.99.6 78733 78734 grab_window and touchid were removed from the struct for ABI compatibility 78735 reasons, we need to pull in the new, XI 2.2-specific struct. 78736 78737 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78738 78739commit b96275c4cdb164aa71f7aa9fbf88be18886d1936 78740Author: Benjamin Otte <otte@redhat.com> 78741Date: Tue Feb 7 18:01:25 2012 +1000 78742 78743 dix: fix an out-of-memory crash 78744 78745 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78746 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78747 78748commit 6241b5e4fdbdb08d30cc8787d858ac27122d2d49 78749Author: Chase Douglas <chase.douglas@canonical.com> 78750Date: Fri Feb 3 16:19:11 2012 -0800 78751 78752 Implement touch early accept 78753 78754 This doesn't really implement early accept as it should. Ideally, the 78755 server should send end events to all subsequent touch clients as soon as 78756 an early accept comes in. However, this implementation is still protocol 78757 compliant. We can always improve it later. 78758 78759 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78760 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78761 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78762 78763commit b0c54856df71f9cabf9dad176fdade960ef8c5d9 78764Author: Chase Douglas <chase.douglas@canonical.com> 78765Date: Fri Feb 3 16:19:10 2012 -0800 78766 78767 Implement early touch reject 78768 78769 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78770 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78771 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78772 78773commit 656ab879f20892975510723ce8fe78faf64aadb4 78774Author: Chase Douglas <chase.douglas@canonical.com> 78775Date: Fri Feb 3 16:19:09 2012 -0800 78776 78777 Check for proper window ID when processing touch allow requests 78778 78779 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78780 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78781 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78782 78783commit 9a260e9af83feb5c53ffd3b2da2dc3adf06240a1 78784Author: Chase Douglas <chase.douglas@canonical.com> 78785Date: Fri Feb 3 16:19:08 2012 -0800 78786 78787 Move AllowTouch to dix/touch.c, and rename to TouchAcceptReject 78788 78789 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78790 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78791 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78792 78793commit 192b2c9a2ec04522655675ddfe71de5cf974cc7b 78794Author: Chase Douglas <chase.douglas@canonical.com> 78795Date: Fri Feb 3 16:19:07 2012 -0800 78796 78797 Export TouchEventRejected as TouchRejected 78798 78799 This function is mostly correct for early reject usage. With a small 78800 change to pass the client resource explicitly and making the 78801 TouchOwnership event optional, it is usable for all rejection scenarios. 78802 This change exports it for use outside Xi/exevents.c and modifies the 78803 name accordingly. 78804 78805 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78806 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78807 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78808 78809commit 19073425e5df1317f73f74ce0b95ab24e0e0da14 78810Author: Chase Douglas <chase.douglas@canonical.com> 78811Date: Fri Feb 3 16:19:06 2012 -0800 78812 78813 Factor out TouchEnd generation and delivery 78814 78815 The server often needs to generate and deliver TouchEnd events for 78816 circumstances including touch grab acceptance and rejection. This change 78817 refactors the code so it can be used more easily. 78818 78819 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78820 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78821 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78822 78823commit ab60cadc2afb9d8882144e30fde11c3ef0f999cd 78824Author: Chase Douglas <chase.douglas@canonical.com> 78825Date: Fri Feb 3 16:19:05 2012 -0800 78826 78827 Store window pointer in touch listener record 78828 78829 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 78830 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78831 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78832 78833commit f92ba92adfce0e1975feb54ab0e1165b7c2a7bea 78834Author: Peter Hutterer <peter.hutterer@who-t.net> 78835Date: Sat Feb 4 19:27:21 2012 +1000 78836 78837 Revert "dix: deduplicate callers of DeliverDeviceEvents in DeliverGrabbedEvents" 78838 78839 This call was supposed to have no functional changes but in some cases 78840 DeliverDeviceEvents() was called with a uninitialised win variable. 78841 Revert, safer than trying to sort this out otherwise. 78842 78843 This reverts commit 6eff14a789341d366b3013c5aa020e959c954651. 78844 78845 Reported-by: Mathieu Taillefumier <mathieu.taillefumier@free.fr> 78846 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 78847 Reviewed-by: Keith Packard <keithp@keithp.com> 78848 78849commit ca64912c02bdff486fee420a49b11f54f8f5ba08 78850Author: Alan Coopersmith <alan.coopersmith@oracle.com> 78851Date: Thu Jan 26 16:41:25 2012 -0800 78852 78853 Namespace list api to reduce conflicts with similar system headers 78854 78855 Rename functions/macros from list_* to xorg_list_* 78856 Rename struct from struct list to struct xorg_list. 78857 78858 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 78859 Reviewed-by: Keith Packard <keithp@keithp.com> 78860 In-sed-I-trust: Peter Hutterer <peter.hutterer@who-t.net> 78861 78862commit 1541e242d1607d0db57b0cc12faca9b1c2850c67 78863Author: Alan Coopersmith <alan.coopersmith@oracle.com> 78864Date: Fri Dec 23 15:17:28 2011 -0800 78865 78866 Stop including <sys/proc.h> from xf86_OSlib.h on Solaris 78867 78868 We don't need anything from that header (which defines /proc & kernel 78869 structures for process information), and it causes some namespace conflicts. 78870 78871 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 78872 Reviewed-by: Keith Packard <keithp@keithp.com> 78873 78874commit 85cecd981191f9c3dab0fb13310d91eff643d423 78875Author: Jeremy Huddleston <jeremyhu@apple.com> 78876Date: Sat Jan 28 16:06:28 2012 -0800 78877 78878 XQuartz: Release all buttons and keys when deactivating 78879 78880 http://xquartz.macosforge.org/trac/ticket/486 78881 78882 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 78883 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 78884 78885commit 47b457541b33f00807fd495f5b0b24d5f143bf84 78886Author: Jeremy Huddleston <jeremyhu@apple.com> 78887Date: Sat Jan 28 23:11:44 2012 -0800 78888 78889 XQuartz: Toggle off fullscreen mode when XQuartz is hidden 78890 78891 http://xquartz.macosforge.org/trac/ticket/478 78892 78893 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 78894 78895commit c2ba3f339f089c595386c4c579935c7984a2a545 78896Author: Jeremy Huddleston <jeremyhu@apple.com> 78897Date: Sun Jan 29 14:26:47 2012 -0800 78898 78899 XQuartz: Localization Updates 78900 78901 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 78902 78903commit 052ca3f22eadd0aa60dd24ac7d5d76137273926f 78904Author: Keith Packard <keithp@keithp.com> 78905Date: Fri Jan 27 22:08:08 2012 -0800 78906 78907 Bump version to 1.11.99.902 (1.12 RC2) 78908 78909 Signed-off-by: Keith Packard <keithp@keithp.com> 78910 78911commit bafedb7e9bcff31e2963eeb54669b2492214fae7 78912Author: Jeremy Huddleston <jeremyhu@apple.com> 78913Date: Wed Jan 18 11:52:04 2012 -0800 78914 78915 XQuartz: Bump bundle version to 2.7.2 78916 78917 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 78918 78919commit 8e78bbb2d2dc5b18f127540b63b45ba293bbdd25 78920Author: Colin Harrison <colin.harrison@virgin.net> 78921Date: Thu Jan 26 13:28:24 2012 +0000 78922 78923 hw/xwin: Fix spelling of 'Canadian' in winkeybd.h 78924 78925 The Mounties always get their typo 78926 78927 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 78928 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 78929 78930commit 80c073352a276bad3722263629de3cd61df758ab 78931Author: Colin Harrison <colin.harrison@virgin.net> 78932Date: Wed Jan 4 17:16:36 2012 +0000 78933 78934 hw/xwin: Fix winEnqueueMotion() for change in miPointerSetPosition() 78935 78936 Commit 3b36fd1b49030ead44358945f62e5abe7f4609ce changed miPointerSetPosition() 78937 to take co-ordinates as doubles, not ints, so this code as it stands is now wrong 78938 (if it ever was correct in the first place :-)) 78939 78940 It's unclear that we can safely promote x,y to doubles, apply miPointerSetPosition() 78941 which potentially constrains the cursor, and then convert back to ints. 78942 78943 Fortunately, this whole dance seems to be unnecessary, and we can simply remove the 78944 call to miPointerSetPosition() entirely, and just QueuePointerEvents() like any other 78945 input driver. 78946 78947 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 78948 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 78949 78950commit ced9db65950e402d7ddc663225b888e8482b8c57 78951Author: Colin Harrison <colin.harrison@virgin.net> 78952Date: Wed Jan 25 01:37:16 2012 +0000 78953 78954 hw/xwin: Ignore WM_DISPLAYCHANGE messages with 0 bpp 78955 78956 Ignore WM_DISPLAYCHANGE messages which indicate bpp is changing to 0. 78957 That has no defined meaning I can find, but some graphics card drivers 78958 appear to generate it on suspend/resume or screensaver activate/deactivate. 78959 78960 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 78961 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 78962 78963commit bea6fb6c7af551778cfe1c0e8412ef4ccf560808 78964Author: Jon TURNEY <jon.turney@dronecode.org.uk> 78965Date: Sun Jan 22 19:31:51 2012 +0000 78966 78967 hw/xwin: Avoid WIN_WINDOW_PROP races during Windows window destruction 78968 78969 The WIN_WINDOW_PROP is removed during WM_DESTROY handling, so it is not neccessary to 78970 remove it in winDestroyWindowsWindow(), and doing so opens a race condition, as we may 78971 attempt to access that property in the wndproc before the WM_DESTROY has completed. 78972 78973 A specific example of that race is if a WM_KILLFOCUS occurs in the window between property 78974 removal and WM_DESTROY processing, where we will attempt to apply DeleteWindowFromAnyEvents() 78975 on an invalid (null) WindowPtr. 78976 78977 Also guard against null WindowPtr in the WM_KILLFOCUS handler 78978 78979 See http://cygwin.com/ml/cygwin-xfree/2012-01/msg00009.html 78980 78981 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 78982 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 78983 78984commit 9a709d5028094fcbeb7a49f47cb85e22cd772f36 78985Author: Jon TURNEY <jon.turney@dronecode.org.uk> 78986Date: Thu Sep 29 15:05:27 2011 +0100 78987 78988 hw/xwin: Handle more motif window decoration hinting 78989 78990 Handle the MWM_DECOR_MINIMIZE, MWM_DECOR_MAXIMIZE and MWM_DECOR_MENU 78991 decoration hints in a _MOTIF_WM_HINTS window property 78992 78993 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 78994 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 78995 78996commit b8b0b841a0e7cfcecaf0b6a5aa67e1b4499374d9 78997Author: Oliver Schmidt <oschmidt-mailinglists@gmx.de> 78998Date: Mon Sep 5 13:32:01 2011 +0100 78999 79000 hw/xwin: Fix AltGr key sometimes firing an additional Ctrl-L key 79001 79002 I also had problems with the AltGr key. These could reliably 79003 be reproduced by holding the AltGr for some seconds (causing 79004 Windows generating auto repeat events) 79005 79006 I discovered that the mechanism in winkeybd.c function 79007 winIsFakeCtrl_L had a problem if PeekMessage cannot obtain 79008 the next Alt_R message because it is not there yet. 79009 79010 I prepared a patch that remembers the last Ctrl_L event and 79011 reacts on a later Alt_R. 79012 79013 It was also necessary to alter the order in winWindowProc() in 79014 winwndproc.c: the invocation of winIsFakeCtrl_L had to be done 79015 before discarding auto-repeated key presses, as winIsFakeCtrl_L() 79016 now has an internal state which must be updated by all key events. 79017 79018 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79019 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79020 79021commit 3d3114d55a2a323f8d49c3549a0dfdf9d4acf89d 79022Author: Jon TURNEY <jon.turney@dronecode.org.uk> 79023Date: Thu Jun 30 14:19:01 2011 +0100 79024 79025 hw/xwin: Handle the virtual key code generated by the Fn key on IBM Lenovo laptops 79026 79027 Apparently, IBM Leonovo laptops can generate a key-press event for the Fn 79028 key, with virtual key code 0xFF and scan code extended 0x63 79029 79030 Handle this specially, rather than just passing on key code 0x63 (delete), 79031 so you don't delete what you just typed when you adjust the screen brightness, 79032 etc. :-) 79033 79034 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79035 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79036 79037commit 74af860f9a59332f6ed1ac9b3e7867fbbb3d7305 79038Author: Jon TURNEY <jon.turney@dronecode.org.uk> 79039Date: Wed Mar 24 22:41:22 2010 +0000 79040 79041 hw/xwin: turn on -emulate3buttons if less than 3 mouse buttons are reported 79042 79043 Try to be more intelligent with default options, turn on -emulate3buttons by 79044 default if less than 3 mouse buttons are reported by Windows 79045 79046 Also, add -noemulate3buttons option so this default setting can be reversed 79047 if desired 79048 79049 Also, correctly report the number of mouse buttons windows is reporting, rather 79050 than always at least 3 79051 79052 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79053 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79054 79055commit c0f3709501a0afd3bf77e783f11d2c2e5f489d3b 79056Author: Jon TURNEY <jon.turney@dronecode.org.uk> 79057Date: Thu Mar 24 20:19:47 2011 +0000 79058 79059 hw/xwin: In multiwindow mode, don't grab native input focus for new windows which hint they don't want it 79060 79061 In multiwindow mode, avoid grabbing the input focus for newly 79062 created windows which have InputHint FALSE 79063 79064 (this is used by e.g. glean to avoid every test window grabbing 79065 the focus) 79066 79067 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79068 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79069 79070commit 25caa8565d7d10f4c254bca5bb9efa05a77542ad 79071Author: Jon TURNEY <jon.turney@dronecode.org.uk> 79072Date: Wed Jan 25 13:33:39 2012 +0000 79073 79074 hw/xwin: Remove some redundant OS version reporting 79075 79076 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79077 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79078 79079commit a9aca218f557c723e637287272819a7c17174e1e 79080Author: Roland Cassard <roland.cassard@gmail.com> 79081Date: Sat Oct 23 18:12:36 2010 +0100 79082 79083 hw/xwin: Don't assume we'll always have converted the clipboard selection after 2 attempts 79084 79085 Rather than knowing we have to call winProcessXEventsTimeout() for up to 2 WIN_XEVENTS_CONVERT messages, process 79086 all messages in winProcessXEventsTimeout() until either: (i) the time out expired, (ii) an error occurred, or 79087 (iii) received a WIN_XEVENTS_NOTIFY messaage indicating the data has been to put on the clipboard. 79088 79089 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79090 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79091 79092commit 75fe336b6c903133ae386f5cb8d308a0e9e2768e 79093Author: Michel Hummel <hummel.michel@gmail.com> 79094Date: Fri Aug 6 15:23:52 2010 +0100 79095 79096 hw/xwin: Mitigate a race condition in clipboard thread initialization 79097 79098 Remove the variables g_fClipboardLaunched and g_fClipboardStarted from 79099 winInitializeGlobals(), as their re-initialization is handled in the 79100 file hw/xwin/InitOutput.c. 79101 79102 Re-initializing g_fClipboardLaunched and g_fClipboardStarted during 79103 the server reset procedure can lead to the clipboard thread being 79104 launched two times and sometimes leads to a crash of the X server... 79105 79106 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79107 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79108 79109commit 95b1391fe3d3192abdfbad4140513b2112cfa02a 79110Author: Jon TURNEY <jon.turney@dronecode.org.uk> 79111Date: Mon Apr 5 13:57:24 2010 +0100 79112 79113 hw/xwin: Chain IOError handlers to avoid longjmp across threads 79114 79115 Avoid crashes on shutdown due to the undefined behaviour of calling longjmp() on the 79116 result of setjmp() from a different thread, by chaining IOError handlers and only 79117 jumping back up to the frame for this thread 79118 79119 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79120 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79121 79122commit 0659437f5ec0e3f646373394f5f9c5461e2170f3 79123Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 79124Date: Sat Feb 20 23:40:53 2010 -0600 79125 79126 hw/xwin: Improve XWinrc loading and error recovery 79127 79128 If $HOME/.XWinrc is present but badly formed, ignore it and try 79129 system.XWinrc instead. If neither file is present or both are badly 79130 formed, provide a built-in default which gives the user the chance to 79131 load their new or fixed configuration without restarting. 79132 79133 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 79134 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79135 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79136 79137commit 47c7b6d3e626497747ae2780f259a15b8e6c846f 79138Author: Michel Hummel <hummel.michel@gmail.com> 79139Date: Sat Oct 23 18:35:57 2010 +0100 79140 79141 hw/xwin: Remove no-longer needed tricks used to prevent the clipboard client from being killed 79142 79143 Remove no-longer needed tricks used to hide the clipboard client from XDM to prevent 79144 it from being killed 79145 79146 - Delete XQuery wrapper used to hide clipboard client 79147 - Delete XDMCP mode heuristic which waits until some magic number of connections have 79148 been established before starting the clipboard 79149 79150 We still need the EstablishConnection wrapper to ensure that the clipboard client isn't 79151 the first client (causing a server restart if it disconnects) 79152 79153 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79154 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79155 79156commit c1bf3baa44fbd8af33a2b3ce045324485b85a7a7 79157Author: Michel Hummel <hummel.michel@gmail.com> 79158Date: Sun Oct 31 14:53:02 2010 +0000 79159 79160 hw/xwin: Automatically restart clipboard thread 79161 79162 Automatically restart clipboard thread on unexpected exit 79163 79164 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79165 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79166 79167commit ae981341a98f7a1ee5d6f3bc8ebde04e42042dc5 79168Author: Jon TURNEY <jon.turney@dronecode.org.uk> 79169Date: Mon Feb 8 22:37:30 2010 +0000 79170 79171 hw/xwin: Give the X window for the clipboard integration client a name 79172 79173 This makes it a bit easier to find when staring at the output of 79174 'xwininfo -tree -root' 79175 79176 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79177 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79178 79179commit b1093241f3e28223a139e7cb5c3ae85182bdf545 79180Author: Jon TURNEY <jon.turney@dronecode.org.uk> 79181Date: Tue Oct 19 20:47:21 2010 +0100 79182 79183 Xext: Warning fix for shm.c 79184 79185 shm.c: In function 'CheckForShmSyscall': 79186 shm.c:182:5: warning: function declaration isn't a prototype [-Wstrict-prototypes] 79187 79188 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79189 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 79190 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79191 79192commit 02775efb8930291cc62fc84086c97da75b912a55 79193Author: Adam Jackson <ajax@redhat.com> 79194Date: Tue Jan 24 17:35:04 2012 -0500 79195 79196 int10: Fix unmapping of the BIOS scratch area 79197 79198 342f3eac8460fc48cfad1f1d7be939d671e6e1cd introduced a bug, 'base' is 79199 incremented before use. The old code corrected this when unmapping, so 79200 the new code should too. 79201 79202 Signed-off-by: Adam Jackson <ajax@redhat.com> 79203 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 79204 Signed-off-by: Keith Packard <keithp@keithp.com> 79205 79206commit e1085a0da0b9299f48b3dc41dee5e33bf022bea5 79207Author: Chase Douglas <chase.douglas@canonical.com> 79208Date: Wed Jan 18 19:09:21 2012 -0800 79209 79210 Don't set X and Y valuators for indirect touch events 79211 79212 For expediency, it made sense to always have the X and Y axes set for 79213 direct touch device event propagation. The last X and Y values are 79214 stored internally. However, indirect device touch event propagation 79215 does not depend on the touch's X and Y values. Thus, we don't need to 79216 set the values for every indirect touch event. 79217 79218 On top of this, the previous X and Y values aren't stored for indirect 79219 touches, so without this change the axes get erroneously set to 0. 79220 79221 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 79222 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79223 79224commit 5201310559fe8708ba8278bdef77cdc1673fff71 79225Author: Daniel Stone <daniel@fooishbar.org> 79226Date: Thu Jan 19 22:40:32 2012 +1100 79227 79228 UngrabAllDevices: Don't kill clients if not told to 79229 79230 The kill_client argument to UngrabAllClients specifies if we want to 79231 kill the client holding the grab or just deactivate the grab. 79232 79233 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 79234 Reported-by: Julien Cristau <jcristau@debian.org> 79235 Reviewed-by: Cyril Brulebois <kibi@debian.org> 79236 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 79237 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79238 79239commit 9b1e18f42a05a79d9ebec372a37b47442e397ca9 79240Author: Chase Douglas <chase.douglas@canonical.com> 79241Date: Wed Jan 18 18:04:14 2012 -0800 79242 79243 Only update pointer motion data for pointer emulated touch events 79244 79245 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 79246 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 79247 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79248 79249commit f00e5b02f5d9928cd88961e761da91b5faff1b32 79250Author: Chase Douglas <chase.douglas@canonical.com> 79251Date: Wed Jan 18 18:04:13 2012 -0800 79252 79253 Only scale direct device touch coordinates 79254 79255 Indirect touch devices provide valuator values in pure device 79256 coordinates. They also don't need to be fixed up for screen crossings. 79257 79258 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 79259 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 79260 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79261 79262commit 24dc0389da3970e83e4908039957567e55f9ea05 79263Author: Peter Hutterer <peter.hutterer@who-t.net> 79264Date: Mon Jan 16 12:53:39 2012 +1000 79265 79266 include: Fix comment typo 79267 79268 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79269 79270commit 954bb994842aa43a0f272858e65036c016b729a4 79271Author: Jeremy Huddleston <jeremyhu@apple.com> 79272Date: Mon Jan 9 00:40:10 2012 -0800 79273 79274 configure.ac: Remove MIEXT_SHADOW_LIB from XORG_LIBS 79275 79276 This is provided by dixmods/libshadow.so and is not part of the main binary. 79277 79278 This addresses a build failure on darwin due to MIEXT_SHADOW_LIB having 79279 unsatisfied dependencies (FB_LIB) in XORG_LIBS. 79280 79281 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 79282 79283commit ba0f5cc1961bfcfb5e7d66ac5df19d429952631d 79284Author: Jeremy Huddleston <jeremyhu@apple.com> 79285Date: Mon Jan 9 00:14:29 2012 -0800 79286 79287 xfree86: Don't link libxorgxkb against libdix.la 79288 79289 libdix.a is already provided by XSERVER_LIBS. Including it in libxorgxkb 79290 results can result in duplicate symbols landing in the Xorg binary on some 79291 configurations (buggy glibtool on darwin). 79292 79293 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 79294 79295commit 2387fb23858d645f15061bc7bcbe4654386ba116 79296Author: Jeremy Huddleston <jeremyhu@apple.com> 79297Date: Fri Jan 13 12:00:14 2012 -0800 79298 79299 sdksyms.sh: Exit on error rather than building an empty symbol table 79300 79301 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 79302 Reviewed-by: Keith Packard <keithp@keithp.com> 79303 79304commit cd89482088f71ed517c2e88ed437e4752070c3f4 79305Author: Jeremy Huddleston <jeremyhu@apple.com> 79306Date: Wed Jan 11 12:17:06 2012 -0800 79307 79308 test: Fix linking issues when building unit tests without the Xorg DDX 79309 79310 This allows unit tests to build and run successfully on darwin when 79311 only the Xvfb or XQuartz DDX is built. 79312 79313 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 79314 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79315 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79316 79317commit 78d39b3222a4b0b92d840137f5455d3a20dd6906 79318Author: Jeremy Huddleston <jeremyhu@apple.com> 79319Date: Sun Jan 15 02:26:43 2012 -0800 79320 79321 Revert "glx: don't leak fbconfigs" 79322 79323 This reverts commit d26fae246d7c451b4d5ffe24fdb959d4bd00b107. 79324 79325 This patch free()s memory in __glXScreenDestroy which was allocated in 79326 glxdricommon (which isn't in every DDX). That breaks abstraction and 79327 causes a crash when XQuartz quits, because it results in freeing a pointer 79328 that was never allocated. 79329 79330 The correct fix is to do this cleanup in __glXDRIscreenDestroy. 79331 79332 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 79333 79334commit f9e6858d5c10be6a8439c0f18bfb2325fa0ee070 79335Author: Chase Douglas <chase.douglas@canonical.com> 79336Date: Wed Jan 11 07:38:12 2012 -0800 79337 79338 Use event time instead of CurrentTime for grab times 79339 79340 When {XI,X,}AllowEvents is called, the timestamp is compared against the 79341 grab time to ensure that the request pertains to the current grab in the 79342 server. While many clients may use CurrentTime (client-side), the 79343 timestamp of the event causing the grab is also valid. 79344 79345 This change ensures that the server's notion of the grab time is the 79346 time of the event that activated the grab rather than the time that the 79347 grab is actually activated. 79348 79349 This bug was exposed through nested touch then pointer grabs. 79350 79351 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 79352 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 79353 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79354 79355commit a6273cc85c01fc020643a68e49ca4e7a2d2ae898 79356Author: Peter Hutterer <peter.hutterer@who-t.net> 79357Date: Thu Jan 12 10:17:34 2012 +1000 79358 79359 xfree86: mention udev in the xorg.conf manpage AutoAddDevices section 79360 79361 And point out what "hotplugging" means. 79362 79363 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79364 Reviewed-by: Julien Cristau <jcristau@debian.org> 79365 79366commit a60d87ffe6d9a15fa830c8da2947c72487863c2b 79367Author: Peter Hutterer <peter.hutterer@who-t.net> 79368Date: Wed Jan 11 09:04:02 2012 +1000 79369 79370 os: prettify backtrace output 79371 79372 Changes to output: 79373 * "Backtrace:" now appears on a separate line _with_ a timestamp 79374 * A blank line is inserted after the last backtrace line 79375 79376 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79377 Reviewed-by: Julien Cristau <jcristau@debian.org> 79378 79379commit c495a839ace7fcc1f1fe414d3d3ba04f08885434 79380Author: Peter Hutterer <peter.hutterer@who-t.net> 79381Date: Wed Jan 11 09:01:05 2012 +1000 79382 79383 include: prettify BUG_WARN output 79384 79385 ErrorF output is prefixed with a timestamp, so the previous output would 79386 look like this: 79387 79388 [ 50.423] BUG: triggered 'if (dev->valuator->numAxes < 2)' 79389 BUG: getevents.c:842 in scale_to_desktop() 79390 79391 Change this to have the prefix on both lines: 79392 [ 50.423] BUG: triggered 'if (dev->valuator->numAxes < 2)' 79393 [ 50.423] BUG: getevents.c:842 in scale_to_desktop() 79394 79395 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79396 Reviewed-by: Julien Cristau <jcristau@debian.org> 79397 79398commit bbb6b8c834e0e1491ca14403b5d0840dd14380d3 79399Author: Peter Hutterer <peter.hutterer@who-t.net> 79400Date: Fri Jan 6 13:20:45 2012 +1000 79401 79402 render: don't bother with animated cursors on floating slaves (#39989) 79403 79404 X.Org Bug 39989 <http://bugs.freedesktop.org/show_bug.cgi?id=39989> 79405 79406 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79407 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 79408 79409commit 35bd77e9d0701daae87d681900d749604fc6471f 79410Author: Peter Hutterer <peter.hutterer@who-t.net> 79411Date: Mon Jan 9 15:04:10 2012 +1000 79412 79413 dix: Update pointer limits for floating devices too (#43635) 79414 79415 When the screen is restructured, the pointer limits need to be reset for 79416 floating slave devices as well, not just for master pointers. Only skip 79417 devices that don't have a cursor (attached slaves and keyboard) 79418 79419 Bug reproducer: float an absolute slave device, rotate the screen - the 79420 device is now confined to a section of the screen only. 79421 79422 X.Org Bug 43635 <http://bugs.freedesktop.org/show_bug.cgi?id=43635> 79423 79424 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79425 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 79426 79427commit 5de0c2582f9c80874b4f4a5b86d250059bfaa85b 79428Author: Keith Packard <keithp@keithp.com> 79429Date: Thu Jan 12 12:10:07 2012 -0800 79430 79431 Revert "os: Repack ConnectionOutput for LP64" 79432 79433 This reverts commit d5f724544afd2949cebfcf4f0b4510ec0c701bec. 79434 79435 ABI change pended for 1.13 79436 79437commit 11331305d4acd117b71502e45c0e86684fa9280b 79438Author: Keith Packard <keithp@keithp.com> 79439Date: Thu Jan 12 12:10:03 2012 -0800 79440 79441 Revert "dix: Repack ClientRec" 79442 79443 This reverts commit f702372822dadb1fef92cfc25086481f640147b3. 79444 79445 ABI change pended for 1.13 79446 79447commit d9eeede52f2d5ba9dd6368d988a5d2abb3b8b4e5 79448Author: Keith Packard <keithp@keithp.com> 79449Date: Thu Jan 12 12:09:59 2012 -0800 79450 79451 Revert "dix: Pull client-is-local flag up to the ClientRec" 79452 79453 This reverts commit 49d38b75c8f3276cfce33ffe6b8c4fbeb1081b96. 79454 79455 ABI change pended for 1.13 79456 79457commit 3be37375eed9eb9cfb9c42821deda4213af4057b 79458Author: Keith Packard <keithp@keithp.com> 79459Date: Thu Jan 12 12:09:55 2012 -0800 79460 79461 Revert "dix: Extend initial connection handshake for forwarding proxies" 79462 79463 This reverts commit 78fa121f4097d29458e5453c13473595df06e26e. 79464 79465 ABI change pended for 1.13 79466 79467commit 5b9f5c8a53aca03c3c73bc10bd362987621a9d72 79468Author: Keith Packard <keithp@keithp.com> 79469Date: Thu Jan 12 12:09:44 2012 -0800 79470 79471 Revert "os: Hide the Connection{In,Out}put implementation details" 79472 79473 This reverts commit 48e7a2ef574c8b38c4f8f07b45f54c8bfd02552b. 79474 79475 ABI change pended for 1.13 79476 79477commit 5867d453cebebec6b4f10409af86b603755fbf51 79478Author: Keith Packard <keithp@keithp.com> 79479Date: Thu Jan 12 12:09:34 2012 -0800 79480 79481 Revert "dix: Fix types in WindowOptRec" 79482 79483 This reverts commit a4553019a10b4e01cc06f3081db71a83338697b4. 79484 79485 ABI change pended for 1.13 79486 79487commit e722ad6c3efa57b806ca0f2dc13114bd3619a88c 79488Merge: e476af417 4e44580ef 79489Author: Keith Packard <keithp@keithp.com> 79490Date: Mon Jan 9 13:22:28 2012 -0800 79491 79492 Merge remote-tracking branch 'jturney/rpavlik-xwin-fixes' 79493 79494commit e476af417d83730b11054f4e5b127ab5540bb332 79495Merge: f4956faab 146008358 79496Author: Keith Packard <keithp@keithp.com> 79497Date: Mon Jan 9 13:17:37 2012 -0800 79498 79499 Merge remote-tracking branch 'whot/multitouch' 79500 79501commit f4956faab9ccf9aba6cf9603f4489f5dad19a347 79502Author: Zhigang Gong <zhigang.gong@linux.intel.com> 79503Date: Wed Jan 4 07:01:19 2012 +0000 79504 79505 mi/mibitblt: Fix an overflow bug of bit shift. 79506 79507 When depth equal to 32 and planeMask equal to 0, the overflow will 79508 occur and cause the pixmap can't be cleared. There are some test 79509 cases in XTS hit this bug, and this fix can eliminate the corresponding 79510 failures. 79511 79512 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 79513 Reviewed-by: Adam Jackson <ajax@redhat.com> 79514 Signed-off-by: Keith Packard <keithp@keithp.com> 79515 79516commit 6d6d4cb6043905d850834946e9bfc526ed5a9ef7 79517Author: Matthieu Herrb <matthieu.herrb@laas.fr> 79518Date: Mon Jan 2 13:23:59 2012 +0000 79519 79520 Add OpenBSD support to DetermineClientCmd() 79521 79522 Uses kvm_getargv() from libkvm. 79523 79524 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 79525 Reviewed-by: Adam Jackson <ajax@redhat.com> 79526 Signed-off-by: Keith Packard <keithp@keithp.com> 79527 79528commit a55214d11916b707b7c8c65c555cc0cbb59ac503 79529Author: Adam Jackson <ajax@redhat.com> 79530Date: Tue Jan 3 06:22:24 2012 +0000 79531 79532 Always install xaa sdk headers 79533 79534 Always install XAA SDK headers so drivers still build even with 79535 --disable-xaa 79536 79537 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79538 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 79539 Signed-off-by: Keith Packard <keithp@keithp.com> 79540 79541commit dafc327f3c75205cf7e5360e9ccd71c7457f61a5 79542Author: Matthieu Herrb <matthieu.herrb@laas.fr> 79543Date: Mon Jan 2 13:11:41 2012 +0000 79544 79545 UnloadSubModule(): accept pointer value '1' and ignore it. 79546 79547 Some driver modules try to unload submodules that are now built-in. 79548 79549 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 79550 Reviewed-by: Julien Cristau <jcristau@debian.org> 79551 Reviewed-by: Adam Jackson <ajax@redhat.com> 79552 Signed-off-by: Keith Packard <keithp@keithp.com> 79553 79554commit 8db029064bcbe378061e812bb8136608e3123226 79555Author: Adam Jackson <ajax@redhat.com> 79556Date: Wed Jan 4 16:39:33 2012 +0000 79557 79558 vgahw: Fix DACDelay() macro to use the driver's vtable 79559 79560 We don't want to unconditionally use I/O routines here, since if the 79561 driver is using mmap'd VGA ports then the I/O handle won't be set up. 79562 79563 Tested-by: Jeff Chua <jeff.chua.linux@gmail.com> 79564 Signed-off-by: Adam Jackson <ajax@redhat.com> 79565 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 79566 Signed-off-by: Keith Packard <keithp@keithp.com> 79567 79568commit 63a8012947fc44ccb3d661dec88dbda14e7f3c04 79569Author: Chris Halse Rogers <christopher.halse.rogers@canonical.com> 79570Date: Thu Jan 5 01:22:40 2012 +0000 79571 79572 Revert "dix: don't return BadMatch from GetProperty (#23562)" 79573 79574 This reverts commit f04fe06ae244b851b38be824b1a80f2f8a030591. 79575 79576 dixLookupWindow no longer returns BadMatch. No other caller was checking 79577 for it, so this problem is now fixed in the utility function. 79578 79579 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 79580 Reviewed-by: Adam Jackson <ajax@redhat.com> 79581 Signed-off-by: Keith Packard <keithp@keithp.com> 79582 79583commit ef492e9797b6d4f6bbc25e86bedc24477819fde7 79584Author: Chris Halse Rogers <christopher.halse.rogers@canonical.com> 79585Date: Thu Jan 5 01:22:39 2012 +0000 79586 79587 dix: Return BadWindow rather than BadMatch from dixLookupWindow 79588 79589 dixLookupWindow uses dixLookupDrawable internally, which returns 79590 BadMatch when the XID matches a non-Window drawable. Users 79591 of dixLookupWindow don't care about this, just that it's not 79592 a valid Window. 79593 79594 This is a generalised version of the fix for X.Org Bug 23562, 79595 where GetProperty was incorrectly returning BadMatch. Auditing other 79596 window requests, all that I checked would incorrectly return BadMatch 79597 in these circumstances. An incomplete list of calls that could 79598 incorrectly return BadMatch is: ListProperties, SetSelectionOwner, 79599 {Destroy,Map,Unmap}{,Sub}Window. 79600 79601 None of the callers of dixLookupWindow, except for GetProperty, check 79602 for BadMatch 79603 79604 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 79605 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 79606 Reviewed-by: Adam Jackson <ajax@redhat.com> 79607 Signed-off-by: Keith Packard <keithp@keithp.com> 79608 79609commit ff891bbf68caefc22cabb541b6b56af086ac2280 79610Author: Arthur Taylor <art@ified.ca> 79611Date: Fri Jan 6 07:21:37 2012 +0000 79612 79613 linux: Use K_OFF VT KB mode over K_RAW if available. 79614 79615 Linux kernels since 2.6.38 (March 2011) have an VT KB mode K_OFF in 79616 which special keys (like Ctrl+C) are not interpreted and input is not 79617 buffered. Use of this mode over K_RAW removes the need for a 79618 xf86ConsoleHandler to drain the VT input buffer, removing the grief it 79619 causes when it goes wrong or is (de)initialized out-of-order. (This 79620 also saves a few needless context switches per key event.) 79621 79622 If K_OFF is not defined or not understood by the kernel, K_RAW and the 79623 previous method is used as a fall-back. 79624 79625 Reviewed-by: Adam Jackson <ajax@redhat.com> 79626 Signed-off-by: Arthur Taylor <art@ified.ca> 79627 Signed-off-by: Keith Packard <keithp@keithp.com> 79628 79629commit 0b113f7cdf5228dccd51a749a339c8669e3f20ff 79630Merge: 0b2c6491c 777bf90ab 79631Author: Keith Packard <keithp@keithp.com> 79632Date: Mon Jan 9 13:07:25 2012 -0800 79633 79634 Merge commit '777bf90abeac37087a3d0538b847742523d5acf2' 79635 79636commit 0b2c6491c5e5e1c9ba067299f3de61f5acee263b 79637Merge: 1f5587e14 6b1943653 79638Author: Keith Packard <keithp@keithp.com> 79639Date: Mon Jan 9 11:40:23 2012 -0800 79640 79641 Merge remote-tracking branch 'whot/for-keith' 79642 79643commit 1f5587e14406c083687203030db6a11691ac9d9f 79644Merge: a97252db2 6269977c9 79645Author: Keith Packard <keithp@keithp.com> 79646Date: Mon Jan 9 11:37:59 2012 -0800 79647 79648 Merge remote-tracking branch 'kibi/master' 79649 79650commit a97252db24669a3e26a935024ea38b8a28d30586 79651Author: Jeremy Huddleston <jeremyhu@apple.com> 79652Date: Tue Dec 20 19:49:47 2011 -0800 79653 79654 XQuartz: GL: Buildfix for recent GLX changes 79655 79656 dispatch.h was leftover from an earlier implementation and is no longer 79657 needed, so remove it since including it causes a build failure due to 79658 conflicts between GL/gl.h and OpenGL/gl.h 79659 79660 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 79661 Reviewed-by: Jamey Sharp <jamey@minilop.net> 79662 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 79663 79664commit 6b194365364cb7a74c2b19f4a8c5e0e0404339da 79665Author: Andreas Wettstein <wettstein509@solnet.ch> 79666Date: Wed Nov 30 20:20:21 2011 +0100 79667 79668 xkb: Message actions suppress other key presses #28575 79669 79670 When a key to which a message action is mapped is held down, presses of 79671 other keys were not registered. 79672 79673 Signed-off-by: Andreas Wettstein <wettstein509@solnet.ch> 79674 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 79675 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79676 79677commit 4e44580efd4522ced644c698336d2f6ea54f3917 79678Author: Ryan Pavlik <rpavlik@iastate.edu> 79679Date: Fri Nov 4 14:29:01 2011 -0500 79680 79681 hw/xwin: Include manifest file in the dist tarball 79682 79683 Commit c02638fd added the manifest file, but didn't add it to EXTRA_DIST. 79684 79685 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79686 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79687 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79688 79689commit c763fe51b8ff18e204caab9cf97376a1b72324f0 79690Author: Ryan Pavlik <rpavlik@iastate.edu> 79691Date: Fri Nov 4 13:17:50 2011 -0500 79692 79693 hw/xwin: Fix duplicate definition of HAS_WINSOCK when building for MinGW 79694 79695 hw/xwin: Fix duplicate definition of HAS_WINSOCK when building for MinGW 79696 but still provide it if building for Win32 without autotools 79697 79698 xserver/hw/xwin/winclipboard.h:42:0: warning: "HAS_WINSOCK" redefined 79699 ../../include/xwin-config.h:11:0: note: this is the location of the previous definition 79700 79701 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79702 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79703 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79704 79705commit 3d80f202b06227f7fc03b674f5fbf809c2d1efb2 79706Author: Ryan Pavlik <rpavlik@iastate.edu> 79707Date: Fri Oct 28 09:56:19 2011 -0500 79708 79709 hw/xwin: Fix possible null ptr deref in winMWExtWMDecorateWindow() 79710 79711 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79712 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79713 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79714 79715commit c824004b4592b3d86b3514be7cab37b36642b13a 79716Author: Ryan Pavlik <rpavlik@iastate.edu> 79717Date: Fri Oct 28 09:56:11 2011 -0500 79718 79719 hw/xwin: Fix possible null ptr deref in winMWExtWMRestackFrame() 79720 79721 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79722 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79723 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79724 79725commit 3c501691a0a9fa17da4e2cc84f55010ef2a4790d 79726Author: Ryan Pavlik <rpavlik@iastate.edu> 79727Date: Fri Oct 28 09:52:34 2011 -0500 79728 79729 hw/xwin: Fix possible null ptr deref in winActivateAppPrimaryDD() 79730 79731 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79732 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79733 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79734 79735commit a492c02649de4c60ac21a7fcb6b7c730b558dca6 79736Author: Ryan Pavlik <rpavlik@iastate.edu> 79737Date: Fri Oct 28 09:49:00 2011 -0500 79738 79739 hw/xwin: Fix possible null ptr deref in winActivateAppNativeGDI() 79740 79741 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79742 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79743 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79744 79745commit b907079596bc8600d3420c189409053b0b5016f6 79746Author: Ryan Pavlik <rpavlik@iastate.edu> 79747Date: Fri Oct 28 09:46:56 2011 -0500 79748 79749 hw/xwin: Clarify an if statement mixed with ifdef in winSetEngine() 79750 79751 Use the same pattern as elsewhere so it's a bit clearer what we 79752 are checking 79753 79754 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79755 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79756 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79757 79758commit 5c35dd7be716a3464f012c92866919a3606f9beb 79759Author: Ryan Pavlik <rpavlik@iastate.edu> 79760Date: Thu Oct 27 15:49:59 2011 -0500 79761 79762 hw/xwin: Fix rrScreenSetSize function pointer mismatch 79763 79764 winrandr.c: In function ‘winRandRInit’: 79765 winrandr.c:218:31: warning: assignment from incompatible pointer type 79766 79767 Fix winRandRScreenSetSize() function signature to match RRScreenSetSizeProcPtr type, 79768 to align with commit fd9331f6 'Revert "Separate out screen size and screen pixmap 79769 sizes in RRScreenSizeSet"' 79770 79771 This is fall-out from the late revert of RANDR 1.4 in the 1.10 release cycle, it will 79772 probably need to be reverted if/when that goes back in again. 79773 79774 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79775 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79776 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79777 79778commit aa07d82908c28f4d083c0c55846a5b34f0e1ef31 79779Author: Ryan Pavlik <rpavlik@iastate.edu> 79780Date: Fri Oct 28 09:56:26 2011 -0500 79781 79782 hw/xwin: Remove an empty #if 0/#endif 79783 79784 Um... yeah 79785 79786 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79787 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79788 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79789 79790commit d459f42e64f2bafef5124a456d2ad6fa2cd92a7f 79791Author: Ryan Pavlik <rpavlik@iastate.edu> 79792Date: Fri Oct 28 09:52:37 2011 -0500 79793 79794 hw/xwin: Fix double-free in error path in winQueryRGBBitsAndMasks() 79795 79796 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79797 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79798 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79799 79800commit cc7dedd6b51a01996ddab10fe8b84e30700d1570 79801Author: Ryan Pavlik <rpavlik@iastate.edu> 79802Date: Fri Oct 28 09:46:45 2011 -0500 79803 79804 hw/xwin: Fix a memory leak in error path in winInitWM() 79805 79806 Signed-off-by: Ryan Pavlik <rpavlik@iastate.edu> 79807 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 79808 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 79809 79810commit 777bf90abeac37087a3d0538b847742523d5acf2 79811Author: Adam Jackson <ajax@redhat.com> 79812Date: Mon Jul 18 15:07:07 2011 -0400 79813 79814 xfree86: Remove the pretense of EDID v2 support 79815 79816 We don't do anything with EDID v2 blocks besides publish them on the 79817 root window. Worse, the check deleted by this patch would attempt to 79818 take a checksum of arbitrary memory if the rawData array isn't 256+ 79819 bytes long (and, for the monitors mentioned, it probably is only 128). 79820 79821 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 79822 Signed-off-by: Adam Jackson <ajax@redhat.com> 79823 79824commit 8f9bdfd293ad8e45755efe8d764b4dcc2a724f51 79825Author: Adam Jackson <ajax@redhat.com> 79826Date: Tue Dec 13 21:23:12 2011 -0500 79827 79828 dix: Tune dixLookupDrawable for success 79829 79830 The vast vast vast majority of resource lookups are successful. Move some 79831 work to the error paths so we don't punish success. 79832 79833 Before: 79834 40000000 trep @ 0.0009 msec (1109091.3/sec): PutImage 10x10 square 79835 60000000 trep @ 0.0005 msec (2072652.2/sec): ShmPutImage 10x10 square 79836 79837 After: 79838 40000000 trep @ 0.0009 msec (1148346.9/sec): PutImage 10x10 square 79839 60000000 trep @ 0.0005 msec (2091666.1/sec): ShmPutImage 10x10 square 79840 79841 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 79842 Signed-off-by: Adam Jackson <ajax@redhat.com> 79843 79844commit 83a98543b58c661a22574a6f8d8f9d777c0955b8 79845Author: Adam Jackson <ajax@redhat.com> 79846Date: Tue Dec 13 20:23:40 2011 -0500 79847 79848 dix: Tune dtrace hooks around Dispatch 79849 79850 Don't call LookupMajorName if the hooks aren't active, it's quite expensive. 79851 79852 Before: 79853 40000000 trep @ 0.0009 msec (1087458.5/sec): PutImage 10x10 square 79854 60000000 trep @ 0.0005 msec (2012238.6/sec): ShmPutImage 10x10 square 79855 79856 After: 79857 40000000 trep @ 0.0009 msec (1109091.3/sec): PutImage 10x10 square 79858 60000000 trep @ 0.0005 msec (2072652.2/sec): ShmPutImage 10x10 square 79859 79860 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 79861 Signed-off-by: Adam Jackson <ajax@redhat.com> 79862 79863commit 11977fab546da35d579ebe285e3c26864007805e 79864Author: Adam Jackson <ajax@redhat.com> 79865Date: Tue Dec 13 21:00:05 2011 -0500 79866 79867 xace: ricer tuning for XaceHook 79868 79869 gcc doesn't want to hoist the check for XaceHooks[hook] != NULL above the 79870 varargs code for some reason, so do it ourselves. 79871 79872 Before: 79873 40000000 trep @ 0.0010 msec (1050420.2/sec): PutImage 10x10 square 79874 60000000 trep @ 0.0005 msec (1921147.6/sec): ShmPutImage 10x10 square 79875 79876 After: 79877 40000000 trep @ 0.0009 msec (1087458.5/sec): PutImage 10x10 square 79878 60000000 trep @ 0.0005 msec (2012238.6/sec): ShmPutImage 10x10 square 79879 79880 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 79881 Signed-off-by: Adam Jackson <ajax@redhat.com> 79882 79883commit a4553019a10b4e01cc06f3081db71a83338697b4 79884Author: Adam Jackson <ajax@redhat.com> 79885Date: Fri Jun 17 14:30:06 2011 -0400 79886 79887 dix: Fix types in WindowOptRec 79888 79889 No reason for these to be 64 bits on LP64. 79890 79891 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 79892 Signed-off-by: Adam Jackson <ajax@redhat.com> 79893 79894commit c44ef2e1ff7bad168c348da63477b4636b3054fd 79895Author: Adam Jackson <ajax@redhat.com> 79896Date: Fri Jun 17 14:17:09 2011 -0400 79897 79898 os: Minor header cleanup 79899 79900 Move some constants near their only users, and remove some 79901 getdtablesize() logic that's second-guessing configure. 79902 79903 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 79904 Signed-off-by: Adam Jackson <ajax@redhat.com> 79905 79906commit 48e7a2ef574c8b38c4f8f07b45f54c8bfd02552b 79907Author: Adam Jackson <ajax@redhat.com> 79908Date: Fri Jun 17 14:03:01 2011 -0400 79909 79910 os: Hide the Connection{In,Out}put implementation details 79911 79912 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 79913 Signed-off-by: Adam Jackson <ajax@redhat.com> 79914 79915commit 78fa121f4097d29458e5453c13473595df06e26e 79916Author: Adam Jackson <ajax@redhat.com> 79917Date: Fri Jun 17 13:43:38 2011 -0400 79918 79919 dix: Extend initial connection handshake for forwarding proxies 79920 79921 Forwarding proxies like sshd will appear to be local, even though they 79922 aren't really. This leads to weird behaviour for extensions that truly 79923 require running under the same OS services as the client, like MIT-SHM 79924 and DRI2. 79925 79926 Add two new legal values for the initial connection's byteOrder field, 79927 'r' and 'R'. These act like 'l' and 'B' respectively, but have the side 79928 effect of forcing the client to be treated as non-local. Forwarding 79929 proxies should attempt to munge the first packet of the connection 79930 accordingly; older servers will reject connections thusly munged, so the 79931 proxy should fall back to passthrough if the munged connection attempt 79932 fails. 79933 79934 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 79935 Signed-off-by: Adam Jackson <ajax@redhat.com> 79936 79937commit 49d38b75c8f3276cfce33ffe6b8c4fbeb1081b96 79938Author: Adam Jackson <ajax@redhat.com> 79939Date: Fri Jun 17 13:22:41 2011 -0400 79940 79941 dix: Pull client-is-local flag up to the ClientRec 79942 79943 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 79944 Signed-off-by: Adam Jackson <ajax@redhat.com> 79945 79946commit f702372822dadb1fef92cfc25086481f640147b3 79947Author: Adam Jackson <ajax@redhat.com> 79948Date: Fri Jun 17 12:58:25 2011 -0400 79949 79950 dix: Repack ClientRec 79951 79952 sizeof(ClientRec) ILP32 LP64 79953 before 120 184 79954 after 104 136 79955 79956 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 79957 Signed-off-by: Adam Jackson <ajax@redhat.com> 79958 79959commit d5f724544afd2949cebfcf4f0b4510ec0c701bec 79960Author: Adam Jackson <ajax@redhat.com> 79961Date: Thu Jun 16 17:40:24 2011 -0400 79962 79963 os: Repack ConnectionOutput for LP64 79964 79965 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 79966 Signed-off-by: Adam Jackson <ajax@redhat.com> 79967 79968commit ed8f3c4bd17bddf1369d050ea8e63b9451d887ce 79969Author: Peter Hutterer <peter.hutterer@who-t.net> 79970Date: Fri Jan 6 10:10:16 2012 +1000 79971 79972 Xi: change XIChangeDeviceProperty from const pointer to const void * 79973 79974 According to Daniel Kurtz, a typedef void *pointer is a atomic type. So a 79975 'const pointer' is equivalent to 'void* const' instead of the intended 79976 'const void*'. 79977 79978 This technically changes the ABI, but we don't bump it for this. 79979 79980 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79981 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 79982 79983commit 146008358e9dd9b178bd5305bcc1680ca16c328d 79984Author: Chase Douglas <chase.douglas@ubuntu.com> 79985Date: Thu Jan 5 16:43:40 2012 -0800 79986 79987 Remove last listener on touch reject 79988 79989 The current code short-circuits around the block that removes the 79990 rejecting listener if it is the only listener left. It also does not 79991 delete the touchpoint record if the touch has not physically ended. 79992 79993 This change ensures the listener is removed under these circumstances. 79994 79995 Signed-off-by: Chase Douglas <chase.douglas@ubuntu.com> 79996 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 79997 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 79998 79999commit e30c3c5c59abba9713cf3be532a46220aa9aeeca 80000Author: Chase Douglas <chase.douglas@canonical.com> 80001Date: Wed Jan 4 15:50:40 2012 -0800 80002 80003 Don't end touchpoint if owning client hasn't accepted/rejected 80004 80005 A touchpoint is ended when no further processing will take place for it. 80006 This includes the situation where there is only one grabbing client, and 80007 the client receives a touch end before it has accepted/rejected the 80008 touchpoint. 80009 80010 This change ensures that a delivered touch end event is converted into a 80011 touch update event under the above scenario. If the event is left as a 80012 touch end event, the touchpoint will be ended in ProcessTouchEvent(). 80013 80014 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 80015 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 80016 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80017 80018commit 075a473e538efb771497b762101694b6c0c45c2b 80019Author: Carlos Garnacho <carlosg@gnome.org> 80020Date: Tue Jan 3 02:43:40 2012 +0100 80021 80022 dix: Set XITouchEmulatingPointer on events from the touch sequence emulating pointer events 80023 80024 The internal flag is kept around, merely translated to XITouchEmulatingPointer 80025 when creating the XI2 events that will be delivered to the client. 80026 80027 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 80028 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 80029 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80030 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80031 80032commit e8db10a033491059ddf2b4777c9fafed14c403e8 80033Author: Peter Hutterer <peter.hutterer@who-t.net> 80034Date: Fri Jan 6 13:49:04 2012 +1000 80035 80036 Require inputproto 2.1.99.5 80037 80038 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80039 80040commit 223ba8b46eacbc8e573bc5136a3d6677f3e39099 80041Author: Peter Hutterer <peter.hutterer@who-t.net> 80042Date: Thu Jan 5 15:55:04 2012 +1000 80043 80044 dix: fix wrong condition checking for attached slave (#44003) 80045 80046 We need to update the master if the device is not a master _and_ it is not 80047 floating. 80048 80049 X.Org Bug 44003 <http://bugs.freedesktop.org/show_bug.cgi?id=44003> 80050 80051 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80052 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80053 80054commit 3ce102c362cadcd7087bdcf48440d9498eaf77d0 80055Author: Peter Hutterer <peter.hutterer@who-t.net> 80056Date: Thu Jan 5 07:08:01 2012 -0500 80057 80058 dix: on PointerRootWin send a FocusIn to the sprite window too 80059 80060 XTS XSetDeviceFocus-7 80061 80062 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80063 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80064 80065commit a125aabda3a5cf27aa98cb61f16e49280b66f451 80066Author: Peter Hutterer <peter.hutterer@who-t.net> 80067Date: Thu Jan 5 07:02:51 2012 -0500 80068 80069 dix: send focus events to the immediate parent (#44079) 80070 80071 For a transition from windows A to B, A->parent did not receive an event. 80072 DeviceFocusOutEvents sends to windows ]from, to[, so start with the actual 80073 window, not it's parent. 80074 80075 X.Org Bug 44079 <http://bugs.freedesktop.org/show_bug.cgi?id=44079> 80076 80077 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80078 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80079 Reviewed-by: Keith Packard <keithp@keithp.com> 80080 80081commit 75953ccb9e9e5f1dbbaa44c714716a0049e22aae 80082Author: Peter Hutterer <peter.hutterer@who-t.net> 80083Date: Tue Jan 3 10:45:45 2012 +1000 80084 80085 xfree86: split warning about missing identifier or input driver 80086 80087 Check for identifier first and bail if it's missing (also remove the current 80088 identifier check after we've already bailed due to missing identifiers) 80089 80090 If a driver is missing, warn but also say that we may have added this device 80091 already. I see too many bugreports with incorrectly shortened log files. 80092 80093 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80094 Reviewed-by: Cyril Brulebois <kibi@debian.org> 80095 80096commit 6269977c91071e0ea16ca5b4b8e15fd6db0b6fcf 80097Author: Julien Cristau <jcristau@debian.org> 80098Date: Fri Dec 30 20:41:25 2011 +0100 80099 80100 os: don't ignore failure from dladdr 80101 80102 If dladdr returns 0, don't go and use the returned Dl_info, it may 80103 contain garbage. 80104 80105 X.Org bug#44315 <https://bugs.freedesktop.org/show_bug.cgi?id=44315> 80106 80107 Reported-and-tested-by: Cyril Brulebois <kibi@debian.org> 80108 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 80109 Reviewed-by: Cyril Brulebois <kibi@debian.org> 80110 Signed-off-by: Julien Cristau <jcristau@debian.org> 80111 80112commit 644efb43e0639af648dd46d9729c2cabd9511238 80113Author: Cyril Brulebois <kibi@debian.org> 80114Date: Fri Dec 30 02:34:36 2011 +0000 80115 80116 linux/ia64: Fix regression after domain I/O support code removal. 80117 80118 Side effect of aa0bfb0f133481c57762012e8e30c05ffa151423: 80119 | CCLD Xorg 80120 | sdksyms.o:(.data.rel+0x27d8): undefined reference to `outl' 80121 | collect2: ld returned 1 exit status 80122 80123 Since the linux/ia64 domain I/O support code got removed in that 80124 commit, there's no reason to keep on declaring those functions 80125 (inb, inl, inw, outb, outl, outw). 80126 80127 Bugzilla: https://bugs.freedesktop.org/43985 80128 80129 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 80130 Signed-off-by: Cyril Brulebois <kibi@debian.org> 80131 80132commit 98cde254acb9b98337ddecf64c138d38c14ec2bf 80133Author: Keith Packard <keithp@keithp.com> 80134Date: Tue Dec 27 13:42:21 2011 -0800 80135 80136 Bump version to 1.11.99.901 (1.12 RC1) 80137 80138 Signed-off-by: Keith Packard <keithp@keithp.com> 80139 80140commit 5037c9af78da6652189de7202e70e1b833395af5 80141Author: Keith Packard <keithp@keithp.com> 80142Date: Tue Dec 27 14:02:01 2011 -0800 80143 80144 glx/glapioffsets.h is no longer part of the build, remove it 80145 80146 Signed-off-by: Keith Packard <keithp@keithp.com> 80147 80148commit 8dedf9831bd80514d800f0085213296a3726dba7 80149Merge: cfc4c3d7f cf9618312 80150Author: Keith Packard <keithp@keithp.com> 80151Date: Tue Dec 27 13:13:48 2011 -0800 80152 80153 Merge remote-tracking branch 'kibi/master' 80154 80155commit cfc4c3d7fa8bd4da4c08b2ab8e6f85435f75353a 80156Author: Alan Coopersmith <alan.coopersmith@oracle.com> 80157Date: Sat Dec 24 10:00:56 2011 -0800 80158 80159 Add Solaris support to DetermineClientCmd 80160 80161 Uses /proc/pid/psinfo to read command & partial arguments. 80162 80163 Moves cmdsize & argsize variables into non-Solaris #else clause 80164 to avoid unused variable warnings. 80165 80166 Fixes format mismatch errors when building with DEBUG defined on 80167 a 64-bit platform (where Mask is defined as CARD32). 80168 80169 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 80170 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 80171 Signed-off-by: Keith Packard <keithp@keithp.com> 80172 80173commit a28ff2cf92c2b35e02eacca21af929afabbf6b83 80174Author: Chase Douglas <chase.douglas@canonical.com> 80175Date: Thu Dec 22 12:00:37 2011 -0800 80176 80177 test/xi2: Really fix infinite loop in test_convert_XITouchOwnershipEvent 80178 80179 long i; 80180 for (i = 1; ; i <<= 1) 80181 if (i == (1 << 31)) 80182 break; 80183 80184 (1 << 31) is compiled as an int, and thus is equal to -2147483648. We 80185 are trying to compare it against a long, which on 64-bit machines is 80186 2147483648. This results in an infinite loop. 80187 80188 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 80189 Signed-off-by: Keith Packard <keithp@keithp.com> 80190 80191commit e7df42ab68e30588a5e32ed543b0711821daf009 80192Author: Keith Packard <keithp@keithp.com> 80193Date: Thu Dec 22 09:35:51 2011 -0800 80194 80195 test/xi2: Fix infinite loop in test_convert_XITouchOwnershipEvent 80196 80197 The touchid test was using a loop like: 80198 80199 for(i = 1; i < 0xffffffff; i <<= 1) 80200 80201 When 'i' is a 32-bit variable, this infinite loops as it goes from 80202 0x80000000 to 0. 'i' is declared as 'long', which is 32-bit in 32-bit mode. 80203 80204 Signed-off-by: Keith Packard <keithp@keithp.com> 80205 80206commit cf96183122a85e3376e560f0360b38b5676e57a9 80207Author: Cyril Brulebois <kibi@debian.org> 80208Date: Wed Nov 2 12:41:20 2011 +0100 80209 80210 xorg.conf.man: Fix bad whatis entry. 80211 80212 Debian's QA tool “lintian” reported a bad whatis entry for the 80213 xorg.conf(.d) manpages. 80214 80215 It comes with the following pointers: 80216 For manual pages that document multiple programs, functions, files, or 80217 other things, the part before "\-" should list each separated by a 80218 comma and a space. […] 80219 80220 Refer to the lexgrog(1) manual page, the groff_man(7) manual page, and 80221 the groff_mdoc(7) manual page for details. 80222 80223 Indeed, the current situation is: 80224 $ whatis xorg.conf; whatis xorg.conf.d 80225 xorg.conf (5) - (unknown subject) 80226 xorg.conf.d (5) - (unknown subject) 80227 80228 With this patch: 80229 xorg.conf (5) - configuration files for Xorg X server 80230 xorg.conf.d (5) - configuration files for Xorg X server 80231 80232 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 80233 Signed-off-by: Cyril Brulebois <kibi@debian.org> 80234 80235commit f75bdf7fbe757f4603e39139acc3c90538a45e15 80236Author: Peter Hutterer <peter.hutterer@who-t.net> 80237Date: Thu Dec 22 10:00:15 2011 +1000 80238 80239 test: fix grab mode value tests for new XIGrabModeTouch 80240 80241 This test checks that last-valid-mode + 1 returns a BadValue. With the 80242 addition of XIGrabModeTouch, that value has changed - sync it up again. 80243 80244 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80245 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80246 80247commit e395efc25f9fd60377f32c2bf1a9153d70908236 80248Merge: 2bb282cd4 fcda98c48 80249Author: Peter Hutterer <peter.hutterer@who-t.net> 80250Date: Thu Dec 22 09:29:59 2011 +1000 80251 80252 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into multitouch 80253 80254 Conflicts: 80255 configure.ac 80256 dix/inpututils.c 80257 80258 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80259 80260commit 2bb282cd45cb8149b39d72397ef5bbcfebca11d2 80261Author: Peter Hutterer <peter.hutterer@who-t.net> 80262Date: Wed Dec 21 14:45:27 2011 +1000 80263 80264 Xi: only activate the device grab if we don't already have one 80265 80266 If the device is already grabbed, don't activate the passive grab, it screws 80267 with our event masks. Just deliver to the grabbing client instead. 80268 80269 Reported-by: Carlos Garnacho <carlosg@gnome.org> 80270 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80271 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80272 80273commit 475ef5cc8339fac3696fe654ef683f92f0b0cc2c 80274Author: Carlos Garnacho <carlosg@gnome.org> 80275Date: Wed Dec 21 01:08:40 2011 +0100 80276 80277 dix: fix xi2 mask/filter when delivering grabbed events 80278 80279 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 80280 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 80281 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80282 80283commit 4551510f752624592f9201f73db89ac1325a234a 80284Author: Peter Hutterer <peter.hutterer@who-t.net> 80285Date: Mon Dec 19 09:08:05 2011 +1000 80286 80287 test: add a basic test for ownership event values 80288 80289 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80290 80291commit 0c661dc478b4b436a3e61066d31bc510c7a08456 80292Author: Peter Hutterer <peter.hutterer@who-t.net> 80293Date: Mon Dec 19 08:31:37 2011 +1000 80294 80295 test: drop printfs from protocol-eventconvert.c 80296 80297 The test outputs are noisy enough, no need having these here too. 80298 80299 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80300 80301commit 8080d785b2c1e5791312ed0975efd1b8aea58d98 80302Author: Peter Hutterer <peter.hutterer@who-t.net> 80303Date: Wed Dec 21 11:11:48 2011 +1000 80304 80305 dix: remove requirement for client_id be the first element 80306 80307 Leftover code from an earlier version of GetTouchEvents. 80308 80309 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80310 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80311 80312commit 552de314bf0c8e4de5d002bdd1c12388c2f42283 80313Author: Peter Hutterer <peter.hutterer@who-t.net> 80314Date: Mon Dec 19 14:52:41 2011 +1000 80315 80316 dix: don't retrieve the syncEvents tail on an empty list 80317 80318 An empty list points to itself but syncEvents has the list head only and is 80319 of a different format than the elements. Thus, casting it to a QdEventPtr 80320 gives us garbage. 80321 80322 Segfaults with XTS test case Xlib13/XGrabKeyboard 80323 80324 Introduced in 7af23259d88f4c28ed21140f82cc03b3724c06bb. 80325 80326 Reported-by: Aaron Plattner <aplattner@nvidia.com> 80327 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80328 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80329 80330commit 76c11e0872dedcd360cbe02cf62bb9de3b212957 80331Author: Peter Hutterer <peter.hutterer@who-t.net> 80332Date: Fri Dec 16 12:41:08 2011 +1000 80333 80334 dix: button state must show the logical buttons, not physical buttons 80335 80336 If the device is mapped 3 2 1, a click on physical button 1 sends a button 3 80337 press, but the state was set for button 1. Fix this, the state must be set 80338 for that button's logical mapping. 80339 80340 https://bugzilla.gnome.org/show_bug.cgi?id=655928 80341 80342 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80343 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80344 80345commit 8a32c39c00789c16006de61112627722ce548be1 80346Author: Peter Hutterer <peter.hutterer@who-t.net> 80347Date: Thu Dec 15 07:59:06 2011 +1000 80348 80349 mi: handle screen switching on pointer emulated touch events 80350 80351 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80352 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80353 80354commit 48547068b115b1f5e5f46a70110454a2175fb9a4 80355Author: Peter Hutterer <peter.hutterer@who-t.net> 80356Date: Thu Dec 15 07:57:47 2011 +1000 80357 80358 mi: split move to new screen logic to a helper function 80359 80360 No functional changes. 80361 80362 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80363 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80364 80365commit 3b1e2035cc4740711360c845cfcdff07f7b60558 80366Author: Daniel Stone <daniel@fooishbar.org> 80367Date: Thu Dec 15 07:52:28 2011 +1000 80368 80369 dix: Remove touch grabs if the grab disappears 80370 80371 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80372 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80373 80374commit cd3de8324e8908955a2e4be3000c8ffee8684c68 80375Author: Peter Hutterer <peter.hutterer@who-t.net> 80376Date: Thu Dec 15 07:48:49 2011 +1000 80377 80378 dix: hook up passive grabs and pointer emulated passive grabs 80379 80380 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80381 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80382 80383commit 5b169cb695bd450d7f64e3800f00c9237ee67f96 80384Author: Carlos Garnacho <carlosg@gnome.org> 80385Date: Tue Dec 13 15:41:23 2011 +0100 80386 80387 Xi: assign correct grab_mode/other_device_mode in XI2 passive grabs 80388 80389 CreateGrab() expects the keyboard mode to be stored in grab_mode, and the 80390 pointer mode in other_device_mode, so respect this in passive XI2 grabs, 80391 and switch modes if needed. 80392 80393 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 80394 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80395 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80396 80397commit 4469430b109fa2da7ba3d2fadf66eca78b7156a0 80398Author: Peter Hutterer <peter.hutterer@who-t.net> 80399Date: Thu Dec 15 07:42:12 2011 +1000 80400 80401 Xi: add the hooks for passive touch grabs 80402 80403 Co-authored-by: Daniel Stone <daniel@fooishbar.org> 80404 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80405 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80406 80407commit bb0af002dc9ad5b464dc6793aedb6d1ff65d001d 80408Author: Peter Hutterer <peter.hutterer@who-t.net> 80409Date: Thu Dec 15 07:21:38 2011 +1000 80410 80411 Xi: handle grab accept/reject requests 80412 80413 Consists mostly of generating an ownership event and processing it. 80414 80415 Co-authored-by: Daniel Stone <daniel@fooishbar.org> 80416 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80417 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80418 80419commit 209b3d613a7bed126c81daedbad6461e4391e9e6 80420Author: Peter Hutterer <peter.hutterer@who-t.net> 80421Date: Tue Dec 20 16:34:43 2011 -0800 80422 80423 Xi: hook up touch ownership handling 80424 80425 Ownership changes don't get processed directly when they happen, instead the 80426 DIX submits an ownership event which goes through ProcessTouchEvents and 80427 ProcessTouchOwnershipEvents. 80428 Then on the required events are generated and sent to clients. 80429 80430 Co-authored-by: Daniel Stone <daniel@fooishbar.org> 80431 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80432 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80433 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80434 80435commit 1894468c2b9788efa70ef4fb5fc911711402f3a7 80436Author: Peter Hutterer <peter.hutterer@who-t.net> 80437Date: Tue Dec 20 16:33:33 2011 -0800 80438 80439 Xi: hook up pointer emulation for emulating touches 80440 80441 Includes a hack for implicit grab activation, because integrating this 80442 requires a larger rewrite and I'm not sleeping enough as it is. 80443 Right now, we deliver the event and check before/after if there is an 80444 implicit grab on. If one activated, then store the event in the grab and 80445 switch the listener type to a grab listener. 80446 80447 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80448 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80449 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80450 80451commit 753c4dca1e42035e29fb21458a746e484ed93ca3 80452Author: Peter Hutterer <peter.hutterer@who-t.net> 80453Date: Wed Dec 14 17:04:38 2011 +1000 80454 80455 dix: add touch to pointer event conversion helper 80456 80457 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80458 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80459 80460commit c3843f5498a849920ab68cca3557411ad8ab3454 80461Author: Peter Hutterer <peter.hutterer@who-t.net> 80462Date: Wed Dec 14 17:02:30 2011 +1000 80463 80464 dix: add real_event parameter to ActivatePassiveGrab 80465 80466 For touch events with pointer emulation, the event that triggers the grab 80467 (the pointer event) is not the same as the actual event (the touch event). 80468 For replaying, we need to store the real event then. 80469 80470 No effective changes at this point, for the current caller event and 80471 real_event are identical. 80472 80473 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80474 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80475 80476commit e389a19a730055d4550ff708da83c78099a516fd 80477Author: Peter Hutterer <peter.hutterer@who-t.net> 80478Date: Tue Dec 20 16:32:52 2011 -0800 80479 80480 Process and deliver touch events 80481 80482 Does not include pointer emulation handling. 80483 Does include partial ownership handling but not the actual processing of 80484 ownership events. 80485 80486 Note: this commit is a retroactive commit extracted from a series of ~50 80487 commits and may thus appear a bit more complicated than what you'd write out 80488 from scratch. 80489 80490 Pointer processing tree is roughly: 80491 - ProcessOtherEvents 80492 - ProcessTouchEvents 80493 - DeliverTouchEvents 80494 - DeliverTouchBeginEvent|DeliverTouchEndEvent|... 80495 - DeliverOneTouchEvent 80496 80497 Also hooks up the event history playing to the right function now. 80498 80499 Co-authored-by: Daniel Stone <daniel@fooishbar.org> 80500 Co-authored-by: Chase Douglas <chase.douglas@canonical.com> 80501 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80502 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80503 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80504 80505commit 634b0da9a83076d0e9e0fc44dc5dc77b0c368bc1 80506Author: Peter Hutterer <peter.hutterer@who-t.net> 80507Date: Thu Dec 15 07:53:19 2011 +1000 80508 80509 dix: set core and button state based on the touch state 80510 80511 If a pointer-emulating touch caused a button to be logically down, set that 80512 state in the input events. 80513 80514 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80515 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80516 80517commit dbfd7b37a0ba21899d8ebb7e0b324301bd466c49 80518Author: Peter Hutterer <peter.hutterer@who-t.net> 80519Date: Wed Dec 14 16:33:05 2011 +1000 80520 80521 Xi: make UpdateDeviceState aware of touch events 80522 80523 Update the logical button state for pointer-emulating events. Button state 80524 must be kept separate from the ButtonClassRec to avoid clearing the button 80525 state on a touch end if there is a physical button still down. 80526 80527 And obviously don't change the button state if we're currently replaying the 80528 event history for some client. 80529 80530 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80531 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80532 80533commit 593c3e2eb3da5c5fb957b68c8025dfdbe1139639 80534Author: Peter Hutterer <peter.hutterer@who-t.net> 80535Date: Wed Dec 14 16:30:05 2011 +1000 80536 80537 dix: add TouchResourceIsOwner helper function 80538 80539 The first listener in the sequence is the owner of the touch sequence. 80540 80541 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80542 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80543 80544commit 5d0260cc986e8c1f0907870b851d47fd73bb1b7d 80545Author: Peter Hutterer <peter.hutterer@who-t.net> 80546Date: Wed Dec 14 16:24:17 2011 +1000 80547 80548 dix: add helper functions for adding/removing touch listeners 80549 80550 The DIX will call TouchSetupListeners once for a new touch. After that 80551 the listener list remains static, with listeners only dropping out when they 80552 either reject the grab or disappear. 80553 80554 Exception: if grabs activate they are prefixed to the listeners. 80555 80556 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80557 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80558 80559commit 6a68f97ce6704dd2ff2a50f91acd2da72c9995bd 80560Author: Chase Douglas <chase.douglas@canonical.com> 80561Date: Wed Dec 14 16:26:04 2011 +1000 80562 80563 dix: add TouchGetPointerEventType helper function 80564 80565 Returns the respective pointer event type for a given touch event type. 80566 80567 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80568 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80569 80570commit 8e58ce73c8b79cb86e890369d0216cdd850c52a5 80571Author: Daniel Stone <daniel@fooishbar.org> 80572Date: Wed Dec 14 16:53:04 2011 +1000 80573 80574 dix: when a window disappears, remove it from the touch sprite trace 80575 80576 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80577 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80578 80579commit c18a173cf5f53778356c5e8f25ab132956f8c27d 80580Author: Daniel Stone <daniel@fooishbar.org> 80581Date: Wed Dec 14 16:03:01 2011 +1000 80582 80583 dix: add helper functions to build up/verify the sprite trace 80584 80585 Touch events' sprite trace stays the same for the duration of the touch 80586 sequence. 80587 80588 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80589 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80590 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80591 80592commit 40475261eacabb79a824db1138407a9d0f9fe37f 80593Author: Peter Hutterer <peter.hutterer@who-t.net> 80594Date: Wed Dec 14 15:56:05 2011 +1000 80595 80596 dix: add touch event history helper functions 80597 80598 If touch client has not registered for ownership events and a grab above 80599 that client is rejected, the client needs to receive the complete event 80600 history. 80601 80602 The history currently doesn't really do fancy overflow handling. We assume 80603 that the first TOUCH_HISTORY_SIZE events are the important ones and anything 80604 after that is dropped. If that is a problem, fix the client that takes > 100 80605 event to decide whether to accept or reject. 80606 80607 Events marked with TOUCH_CLIENT_ID or TOUCH_REPLAYING must not be stored in 80608 the history, they are events created by the DIX to comply with the protocol. 80609 Any such event should already be in the history anyway. 80610 80611 A fixme in this patch: we don't have a function to actually deliver the 80612 event yet. 80613 80614 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80615 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80616 80617commit b274c472a5ad700aea5fe651d677e8cb34ffc976 80618Author: Peter Hutterer <peter.hutterer@who-t.net> 80619Date: Wed Oct 26 09:08:17 2011 +1000 80620 80621 Support XI 2.2 officially 80622 80623 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80624 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80625 80626commit d2af968cb65873780a6e61342d5d3c23b9654e2a 80627Author: Daniel Stone <daniel@fooishbar.org> 80628Date: Wed Dec 14 15:45:19 2011 +1000 80629 80630 Xi: allow selecting for touch events 80631 80632 Selecting for any of XI_TouchBegin/Update/End/Ownership requires the three 80633 bits for begin/update/end to be set. 80634 80635 Only one client at a time may select for XI_TouchBegin event 80636 80637 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80638 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80639 80640commit f3df3ad668fcd417ffb5afa3bea79a73a348bc1a 80641Author: Peter Hutterer <peter.hutterer@who-t.net> 80642Date: Wed Dec 14 16:45:04 2011 +1000 80643 80644 dix: handle DIX-submitted touch events 80645 80646 The DIX needs to submit touch events for e.g. TouchEnd after an 80647 acceptance/rejection. These have the TOUCH_CLIENT_ID flag set. 80648 80649 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80650 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80651 80652commit fcda98c48610fd507ca0b89c6006a5497d9dc1c9 80653Author: Peter Hutterer <peter.hutterer@who-t.net> 80654Date: Fri Dec 16 12:41:08 2011 +1000 80655 80656 dix: button state must show the logical buttons, not physical buttons 80657 80658 If the device is mapped 3 2 1, a click on physical button 1 sends a button 3 80659 press, but the state was set for button 1. Fix this, the state must be set 80660 for that button's logical mapping. 80661 80662 https://bugzilla.gnome.org/show_bug.cgi?id=655928 80663 80664 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80665 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80666 80667commit 47da6b6b2c2e9245422655f94770146fb8cd430b 80668Author: Peter Hutterer <peter.hutterer@who-t.net> 80669Date: Mon Dec 19 14:52:41 2011 +1000 80670 80671 dix: don't retrieve the syncEvents tail on an empty list 80672 80673 An empty list points to itself but syncEvents has the list head only and is 80674 of a different format than the elements. Thus, casting it to a QdEventPtr 80675 gives us garbage. 80676 80677 Segfaults with XTS test case Xlib13/XGrabKeyboard 80678 80679 Introduced in 7af23259d88f4c28ed21140f82cc03b3724c06bb. 80680 80681 Reported-by: Aaron Plattner <aplattner@nvidia.com> 80682 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80683 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80684 80685commit 2d34b34ed7547b829979756ccfb2cff79368d0a4 80686Merge: e4b4d83fa 1deede3c4 80687Author: Keith Packard <keithp@keithp.com> 80688Date: Tue Dec 20 00:23:33 2011 -0800 80689 80690 Merge remote-tracking branch 'jeremyhu/master' 80691 80692commit e4b4d83fad37bb737e25f7226dbcd15b892f9528 80693Merge: 4df65d247 d26cb7027 80694Author: Keith Packard <keithp@keithp.com> 80695Date: Mon Dec 19 22:36:31 2011 -0800 80696 80697 Merge remote-tracking branch 'idr/glx-fixes' 80698 80699commit 4df65d247b56c6b1b06e17a61236d6637312cdbb 80700Author: Ville Syrjälä <syrjala@sci.fi> 80701Date: Sun Dec 18 18:29:41 2011 +0200 80702 80703 dri2: Invalidate window pixmaps 80704 80705 While a redirected window is flipped, its pixmap may still be used as 80706 and EGL image and should also get invalidated. When sending invalidate 80707 events for a window, also send the events for its pixmap. 80708 80709 Signed-off-by: Ville Syrjälä <syrjala@sci.fi> 80710 Reviewed-by: Keith Packard <keithp@keithp.com> 80711 Signed-off-by: Keith Packard <keithp@keithp.com> 80712 80713commit e8fd23fad046877ca0ee08c1e147557756debc8e 80714Author: Keith Packard <keithp@keithp.com> 80715Date: Sun Dec 18 18:29:40 2011 +0200 80716 80717 dri2: Invalidate DRI2 buffers for all windows with the same pixmap on swap 80718 80719 Without this, when a compositing manager unredirects a fullscreen window which 80720 uses DRI2 and page flipping, the DRI2 buffer information for the compositing 80721 manager's output window (typically the Composite Overlay Window or root window) 80722 may become stale, resulting in all kinds of hilarity. 80723 80724 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=35452 . 80725 80726 [Original patch by Michel Dänzer <michel@daenzer.net>] 80727 [Tree walk optimized version by Keith Packard <keithp@keithp.com>] 80728 80729 Signed-off-by: Ville Syrjälä <syrjala@sci.fi> 80730 Signed-off-by: Keith Packard <keithp@keithp.com> 80731 80732commit 6f916ffec7767eeab62132eb6575043969104c81 80733Author: Michel Dänzer <daenzer@vmware.com> 80734Date: Sun Dec 18 18:29:39 2011 +0200 80735 80736 dri2: Always re-generate front buffer information when asked for it. 80737 80738 Otherwise we might keep stale cached information, e.g. after the driver 80739 performed page flipping. 80740 80741 This is part of the fix for 80742 https://bugs.freedesktop.org/show_bug.cgi?id=35452 . 80743 80744 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 80745 Reviewed-by: Ville Syrjälä <ville.syrjala@nokia.com> 80746 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 80747 Signed-off-by: Keith Packard <keithp@keithp.com> 80748 80749commit 603fcb3abf155e91fd3250f4a7aa16f289506e8e 80750Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 80751Date: Sun Dec 18 18:29:38 2011 +0200 80752 80753 dri2: Initialize needInvalidate member of DRI2Drawable. 80754 80755 If the client is not behaving correctly and swaps buffers before 80756 getting them, Valgrind will complain about uninitialized memory being 80757 used in DRI2InvalidateDrawable. 80758 80759 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 80760 Reviewed-by: Ville Syrjälä <syrjala@sci.fi> 80761 Signed-off-by: Keith Packard <keithp@keithp.com> 80762 80763commit 3aca8199405cff5097ee5813605ec78b5ae44a43 80764Author: Peter Hutterer <peter.hutterer@who-t.net> 80765Date: Fri Dec 16 09:15:25 2011 -0500 80766 80767 dmx: force -fno-strict-aliasing for xinput example 80768 80769 Compiler warning: 80770 xinput.c:272: warning: dereferencing pointer 'e' does break strict-aliasing 80771 rules 80772 80773 The code itself is the usual XInput client-side code: 80774 XEvent event; 80775 XDeviceMotionEvent *e = (XDeviceMotionEvent *)&event; 80776 XNextEvent(display, &event); 80777 printf("%d\n", e->type); 80778 80779 Since XDeviceMotionEvent is not guaranteed the same size as XEvent, clients 80780 must use pointer aliasing as above when using the XNextEvent API. Disable 80781 strict aliasing for this example. 80782 80783 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 80784 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 80785 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80786 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 80787 Signed-off-by: Keith Packard <keithp@keithp.com> 80788 80789commit 1deede3c48d5ac91f50774c79f02afc6e036614e 80790Author: Jeremy Huddleston <jeremyhu@apple.com> 80791Date: Sat Dec 17 12:56:59 2011 -0800 80792 80793 kdrive/linux: Fix compilation with -Werror=int-to-pointer-cast -Werror=pointer-to-int-cast 80794 80795 Reported-by: Arkadiusz Miśkiewicz <arekm@maven.pl> 80796 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 80797 Reviewed-by: Jamey Sharp <jamey@minilop.net> 80798 80799commit d0170d54c71fc761b8f1ae7a0e555dc2ca23cf9c 80800Author: Jeremy Huddleston <jeremyhu@apple.com> 80801Date: Sat Dec 17 12:25:26 2011 -0800 80802 80803 configure.ac: Make Xephyr dependency error message more informative 80804 80805 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 80806 Reviewed-by: Jamey Sharp <jamey@minilop.net> 80807 80808commit 511beddffab3c4fd43c75556132159117b6ba884 80809Author: Jeremy Huddleston <jeremyhu@apple.com> 80810Date: Fri Nov 25 09:45:25 2011 -0800 80811 80812 XQuartz: Provide in-tree implementation of strndup(3) if needed 80813 80814 This function was added to Mac OS X 10.7, so we need to provide it 80815 for earlier releases. 80816 80817 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 80818 80819commit 9ce8463a3334f76f128bb35b541aa6e67b4e4374 80820Author: Daniel Stone <daniel@fooishbar.org> 80821Date: Wed Dec 14 15:41:07 2011 +1000 80822 80823 dix: generate touchpoints from driver-submitted data 80824 80825 The touchpoints are generated, enqueued but not processed since we don't 80826 handle them in the event processing yet. 80827 80828 Co-authored-by: Peter Hutterer <peter.hutterer@who-t.net> 80829 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80830 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80831 80832commit bec51f5ec6575a1971ae5371a4b6d8563933032c 80833Author: Daniel Stone <daniel@fooishbar.org> 80834Date: Wed Dec 14 15:29:48 2011 +1000 80835 80836 dix: add GetTouchOwnership event API 80837 80838 No callers yet. This API is not to be used by drivers, it's an API for the 80839 DIX which will create ownership events mainly on touch acceptance/rejection. 80840 80841 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80842 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80843 80844commit 5c63dc6dbcbebbb19d79575a9f1ec9878e6537f1 80845Author: Peter Hutterer <peter.hutterer@who-t.net> 80846Date: Wed Dec 14 14:57:46 2011 +1000 80847 80848 dix: add DIX API to create touchpoints 80849 80850 The DIX touchpoints are the ones used for event processing. 80851 80852 Co-authored-by: Daniel Stone <daniel@fooishbar.org> 80853 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80854 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80855 80856commit 7f8127d203394cae45c3ded0d063030d7c5fdb70 80857Author: Peter Hutterer <peter.hutterer@who-t.net> 80858Date: Wed Dec 14 14:53:52 2011 +1000 80859 80860 dix: if we run out of space for new touch events, resize the queue 80861 80862 The SIGIO handler forces us to drop the current touch and schedule the 80863 actual resize for later. Should not happen if the device sets up the 80864 TouchClassRec with the correct number of touchpoints. 80865 80866 Co-authored-by: Daniel Stone <daniel@fooishbar.org> 80867 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80868 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80869 80870commit d26cb70272a1245a7b1db639cebeddafd8ed4c40 80871Author: Ian Romanick <ian.d.romanick@intel.com> 80872Date: Thu Dec 8 12:07:38 2011 -0800 80873 80874 glx: use dispatch.h instead of glapioffsets.h 80875 80876 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 80877 Reviewed-by: Adam Jackson <ajax@redhat.com> 80878 80879commit 66e6037c334c2bcfe6e60d24cac83a8d6087fdd5 80880Author: Ian Romanick <ian.d.romanick@intel.com> 80881Date: Mon Dec 5 15:27:08 2011 -0800 80882 80883 glx: Regenerate GL API files from Mesa 80884 80885 Adam reviewed the underlying changes in the Mesa generated scripts. 80886 80887 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 80888 Reviewed-by: Adam Jackson <ajax@redhat.com> 80889 80890commit e8485768c1f1d9d9fb7ca9e52ff61175b7bb80e1 80891Author: Ian Romanick <ian.d.romanick@intel.com> 80892Date: Tue Dec 6 12:30:17 2011 -0800 80893 80894 glx: Add stub implementations of next GLX extension functions 80895 80896 Generated code from Mesa contains dispatch for GLX_ARB_create_context 80897 and GLX_ARB_create_context_profile. Provide stub implementations of 80898 these functions so that the GLX extension will link. 80899 80900 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 80901 Reviewed-by: Adam Jackson <ajax@redhat.com> 80902 80903commit 2df539c0bc3300ea858f8bc7d52e95e67ff379b8 80904Author: Ian Romanick <ian.d.romanick@intel.com> 80905Date: Mon Dec 5 14:42:35 2011 -0800 80906 80907 glx: Only declare GlxExtensionInit in one header file 80908 80909 This silences the GCC warning: 80910 80911 In file included from singlepixswap.c:36:0: 80912 glxext.h:47:13: warning: redundant redeclaration of 'GlxExtensionInit' [-Wredundant-decls] 80913 glxserver.h:80:6: note: previous declaration of 'GlxExtensionInit' was here 80914 80915 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 80916 Reviewed-by: Adam Jackson <ajax@redhat.com> 80917 80918commit 758bc57ba5a89f765d83f0b169aa09e79a89bf89 80919Author: Peter Hutterer <peter.hutterer@who-t.net> 80920Date: Wed Dec 14 14:48:56 2011 +1000 80921 80922 dix: add helper functions to create DDX touch recs 80923 80924 DDX touch points are the ones that keep records of the driver-submitted 80925 touchpoints. They're unaffected by the grab state and terminate on a 80926 TouchEnd submitted by the driver. 80927 80928 The client ID assigned is server-global. 80929 80930 Since drivers usually submit in the SIGIO handler, we cannot allocate in the 80931 these functions. 80932 80933 Co-authored-by: Daniel Stone <daniel@fooishbar.org> 80934 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80935 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80936 80937commit 1a133eb8b1ddbe0da7c2fbf7f6a686ec4512373e 80938Author: Peter Hutterer <peter.hutterer@who-t.net> 80939Date: Thu Dec 15 08:30:02 2011 +1000 80940 80941 xfree86: bump the input ABI for the touch changes 80942 80943 New additions to the API: 80944 - InitTouchClassDeviceStruct 80945 - xf86PostTouchEvent 80946 80947 Changes to the ABI: 80948 - DeviceIntRec now contains a TouchClassPtr 80949 80950 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80951 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80952 80953commit 3fb258ca28850c998097b55884774cb95f476f69 80954Author: Daniel Stone <daniel@fooishbar.org> 80955Date: Wed Dec 14 12:46:40 2011 +1000 80956 80957 input: add a TouchClassRec to the devices 80958 80959 These structs will be used to store touch-related data, events and 80960 information. 80961 80962 Drivers must call InitTouchClassDeviceStruct to set up a multi-touch capable 80963 device. 80964 80965 Touchpoints for the DDX and the DIX are handled separately - touchpoints 80966 submitted by the driver/DDX will be stored in the DDXTouchPointInfoRec. Once 80967 the touchpoints are processed by the DIX, new TouchPointInfoRecs are created 80968 and stored. This process is already used for pointer events with the 80969 last.valuators field. 80970 80971 Note that this patch does not actually add the generation of touch events, 80972 only the required structs. 80973 80974 TouchListeners are (future) recipients of touch or emulated pointer events. 80975 Each listener is in a state, depending which event they have already 80976 received. The type of listener defines how the listener got to be one. 80977 80978 Co-authored-by: Peter Hutterer <peter.hutterer@who-t.net> 80979 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80980 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80981 80982commit 098b837440e40bbc485368ec9658e12efd6ef581 80983Author: Daniel Stone <daniel@fooishbar.org> 80984Date: Wed Dec 14 14:41:48 2011 +1000 80985 80986 Add the touch input API stubs 80987 80988 xf86PostTouchEvent is the driver API to submit touch events to the server. 80989 This API doesn't do anything yet though but now we can at least bump the 80990 API. 80991 80992 For valuators, drivers should use the existing xf86InitValuatorAxisStruct 80993 function. 80994 80995 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 80996 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 80997 80998commit 93c2a1628a9f6e7480d834bf55c080997fd18911 80999Author: Peter Hutterer <peter.hutterer@who-t.net> 81000Date: Wed Dec 14 13:51:12 2011 +1000 81001 81002 Hook up the ownership events 81003 81004 This patch applies most of the protocol conversions and the internal event 81005 type for ownership events. 81006 81007 Note that ownership events are generated by the DIX only, they do not pass 81008 through the event queue. 81009 81010 Co-authored-by: Daniel Stone <daniel@fooishbar.org> 81011 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81012 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81013 81014commit c065f7b16329e30cb976359e0a807c2adaf50123 81015Author: Keith Packard <keithp@keithp.com> 81016Date: Sat Dec 17 17:22:45 2011 -0800 81017 81018 Version 1.11.99.2 (1.12 snapshot 2) 81019 81020 Signed-off-by: Keith Packard <keithp@keithp.com> 81021 81022commit ead968a4300c0adeff89b9886e888b6d284c75cc 81023Author: Antoine Martin <antoine@nagafix.co.uk> 81024Date: Sat Dec 17 01:36:51 2011 +0700 81025 81026 xserver: check for elevated privileges not uid=0 81027 81028 This allows us to run the server as a normal user whilst still 81029 being able to use the -modulepath, -logfile and -config switches 81030 We define a xf86PrivsElevated which will do the checks and cache 81031 the result in case it is called more than once. 81032 Also renamed the paths #defines to match their new meaning. 81033 Original discussion which led to this patch can be found here: 81034 http://lists.freedesktop.org/archives/xorg-devel/2011-September/025853.html 81035 81036 Signed-off-by: Antoine Martin <antoine@nagafix.co.uk> 81037 Tested-by: Michal Suchanek <hramrach at centrum.cz> 81038 Reviewed-by: Jamey Sharp <jamey at minilop.net> 81039 Reviewed-by: Adam Jackson <ajax@redhat.com> 81040 81041commit 3390d3fc0347947796aa245d70fefecf59b7bfe4 81042Author: Peter Hutterer <peter.hutterer@who-t.net> 81043Date: Wed Dec 14 13:21:59 2011 +1000 81044 81045 Xi: process raw touch events 81046 81047 No-one can generated them yet, but if they could, we'd be processing them 81048 like there was no tomorrow. 81049 81050 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81051 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81052 81053commit 84db813b9db34975b5fe288a8a551bb98f0cc1c1 81054Author: Peter Hutterer <peter.hutterer@who-t.net> 81055Date: Wed Dec 14 12:45:28 2011 +1000 81056 81057 Hook up TouchBegin/Update/End events 81058 81059 The are the same as device events internally but require the touch ID 81060 separately from the detail.button field (the protocol uses the detail field 81061 for the touch id). 81062 For simpler integration of pointer emulation we need to set the 81063 detail.button field while keeping the touchid around. 81064 81065 Add the three new touch event types to the various places in the server 81066 where they need to be handled. The actual handling of the events is somewhat 81067 more complicated in most places. 81068 81069 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81070 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81071 81072commit 92a5862d0c120b009a688237ec7142b7c21ae272 81073Author: Peter Hutterer <peter.hutterer@who-t.net> 81074Date: Wed Dec 14 15:00:33 2011 +1000 81075 81076 include: add a bunch of flags for GetTouchEvent processing 81077 81078 TOUCH_CLIENT_ID is set if the touch was generated from a client ID instead 81079 of a DDX/driver touch ID. i.e. submitted by the dix. 81080 81081 TOUCH_END is a special flag that's required to force the touch to end. 81082 Since the protocol with grab replaying and pointer emulation is rather 81083 complex, it's quite hard to know otherwise when a touch sequence should 81084 really die. 81085 81086 The others do what it says on the imaginary box. 81087 81088 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81089 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81090 81091commit 956a97487beada8cf5f354550ff779635cc64361 81092Author: Peter Hutterer <peter.hutterer@who-t.net> 81093Date: Wed Dec 14 12:19:27 2011 +1000 81094 81095 include: RawTouchEnd is the last event now 81096 81097 Plus, use the actual definition from the protocol instead of the numeric 81098 values. Turns out not everyone knows the protocol event IDs by heart. 81099 81100 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81101 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81102 81103commit 92c6c0c1d2d6a8882ed97b0e7f2952978ee5ce82 81104Author: Peter Hutterer <peter.hutterer@who-t.net> 81105Date: Wed Dec 14 12:15:29 2011 +1000 81106 81107 tests: update for touch support 81108 81109 Touch event mask must be set for all three event types. 81110 81111 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81112 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81113 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81114 81115commit 731b1db2fcc82c0a8afd28d7f28363c1db0a20ce 81116Author: Peter Hutterer <peter.hutterer@who-t.net> 81117Date: Wed Dec 14 10:41:17 2011 +1000 81118 81119 Require inputproto 2.1.99.3 81120 81121 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81122 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81123 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81124 81125commit 4c825eae89063fdc2241f92ab4101ba682b5b4b7 81126Author: Peter Hutterer <peter.hutterer@who-t.net> 81127Date: Wed Nov 2 10:38:26 2011 +1000 81128 81129 Xi: add a FIXME 81130 81131 All the DeepCopy stuff really needs to be shared between the init calls the 81132 drivers use and this code here. Too many bugs by not keeping the two in 81133 sync. 81134 81135 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81136 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81137 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81138 81139commit 0b9d2e426defabbf6df7e872e158b080a6a5e0be 81140Author: Peter Hutterer <peter.hutterer@who-t.net> 81141Date: Wed Dec 14 13:21:03 2011 +1000 81142 81143 Xi: split ProcessOtherEvent into ProcessDeviceEvent 81144 81145 No functional changes. 81146 81147 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81148 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81149 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81150 81151commit 190a6f45ee7412870b3d95c5db6c1079a285bc59 81152Author: Peter Hutterer <peter.hutterer@who-t.net> 81153Date: Fri Dec 16 09:52:43 2011 +1000 81154 81155 dix: split positionSprite into scale_to_desktop and positionSprite 81156 81157 For future touch points, we need positionSprite to calculate the coordinates 81158 but we don't want to actually change the cursor position for non-emulating 81159 touches. 81160 81161 No functional changes at this point. 81162 81163 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81164 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81165 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81166 81167commit 2c1d0a539c0fb1860ac89a00d1fb609f1959e1b6 81168Author: Gaetan Nadon <memsize@videotron.ca> 81169Date: Fri Dec 9 07:45:02 2011 -0500 81170 81171 dmx: fix distcheck failure, missing compsize.h in Makefile.am 81172 81173 which was added in commit: 81174 dmx: Build fix for -Werror=implicit-function-declaration 81175 81176 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81177 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 81178 Signed-off-by: Keith Packard <keithp@keithp.com> 81179 81180commit ff6d9c79c1aa502f1aef81d8063425e5fc002534 81181Author: Peter Hutterer <peter.hutterer@who-t.net> 81182Date: Thu Dec 15 10:19:26 2011 +1000 81183 81184 include: fix BUG_WARN_MSG for constant messages only 81185 81186 Previous declaration required the use of a message + printf varargs. We 81187 obviously want to allow the use of just a message. 81188 81189 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81190 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81191 81192commit f7ea7a324520844beaac8be7503ac50e20da5ab1 81193Author: Peter Hutterer <peter.hutterer@who-t.net> 81194Date: Wed Dec 14 10:34:19 2011 +1000 81195 81196 configure: split the required modules up 81197 81198 We do the same thing for libraries and optional modules already, and it's 81199 much easier to read when one of them changes 81200 81201 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81202 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81203 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 81204 Signed-off-by: Keith Packard <keithp@keithp.com> 81205 81206commit 7da7aa96a0acb895e1fe57a9d3f2d093479cf2a2 81207Merge: b79de3f42 7528a6b88 81208Author: Keith Packard <keithp@keithp.com> 81209Date: Wed Dec 14 11:40:10 2011 -0800 81210 81211 Merge remote-tracking branch 'whot/for-keith' 81212 81213commit 4c80dc1394344dc10cffe4f757c980944312dc3d 81214Author: Peter Hutterer <peter.hutterer@who-t.net> 81215Date: Wed Dec 14 10:31:23 2011 +1000 81216 81217 configure: split the required modules up 81218 81219 We do the same thing for libraries and optional modules already, and it's 81220 much easier to read when one of them changes 81221 81222 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81223 81224commit 7528a6b88eb32098af4369a8bd9d70a808fa6f1c 81225Merge: f0de5e399 898d97672 81226Author: Peter Hutterer <peter.hutterer@who-t.net> 81227Date: Tue Dec 13 14:19:54 2011 +1000 81228 81229 Merge branch 'for-whot' of git://people.freedesktop.org/~alanc/xserver into for-keith 81230 81231commit f0de5e3990b70df9899b7dd17d7ee6ff7c768183 81232Author: Peter Hutterer <peter.hutterer@who-t.net> 81233Date: Sun Dec 11 07:31:43 2011 +1000 81234 81235 dix: add a MAXEVENTS define for the number of core + extension events 81236 81237 Not including GenericEvents 81238 81239 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81240 Reviewed-by: Walter Harms <wharms@bfs.de> 81241 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81242 81243commit adf8a0d62c0643f4c44a3b6d2e3c6c9b02532915 81244Author: Peter Hutterer <peter.hutterer@who-t.net> 81245Date: Fri Dec 9 18:38:53 2011 +1000 81246 81247 dix: move event filter retrieval helpers to inpututils.c 81248 81249 No functional changes 81250 81251 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81252 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81253 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81254 81255commit a03fe819506fbcbfcb299d4919d16061603fd4ec 81256Author: Peter Hutterer <peter.hutterer@who-t.net> 81257Date: Fri Dec 9 18:02:49 2011 +1000 81258 81259 dix: switch EventIsDeliverable to take the event type only 81260 81261 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81262 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81263 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81264 81265commit 10a8c428fac0d4ff8f8a3e3487c13dd349d27ed7 81266Author: Peter Hutterer <peter.hutterer@who-t.net> 81267Date: Fri Dec 9 15:49:04 2011 +1000 81268 81269 dix: move storing last.valuators into a helper function 81270 81271 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81272 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81273 81274commit a15abf82ddb5a209dbc30c4d12e1dc23c4adf3e4 81275Author: Peter Hutterer <peter.hutterer@who-t.net> 81276Date: Thu Dec 8 15:38:44 2011 +1000 81277 81278 dix: always allocate and set the grab's sync.event 81279 81280 Allocate the memory at device creation time and always store the event, even 81281 if we're not frozen. This way we know which event triggered the grab. 81282 Since the event was never freed anyway except on device shutdown, this 81283 doesn't really change things much. 81284 81285 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81286 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81287 81288commit 218752bdc5d9323d1e6202e762573a925cf8a4eb 81289Author: Peter Hutterer <peter.hutterer@who-t.net> 81290Date: Thu Dec 8 14:27:01 2011 +1000 81291 81292 input: replace GRABTYPE_* with the InputLevel enums 81293 81294 They achieve the same thing, re-use the more generic InputLevel so we can 81295 convert to/fro easier. 81296 81297 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81298 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81299 81300commit cd56bd7b3ee34a4b10eb3a57a6e94cac7512167a 81301Author: Peter Hutterer <peter.hutterer@who-t.net> 81302Date: Thu Dec 8 14:59:06 2011 +1000 81303 81304 Add GrabIsPointerGrab and GrabIsKeyboardGrab helpers 81305 81306 No functional changes 81307 81308 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81309 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81310 81311commit 47734b2ba2e1f9246fd1a3b7059ca60b0e15acb2 81312Author: Peter Hutterer <peter.hutterer@who-t.net> 81313Date: Thu Dec 8 12:00:34 2011 +1000 81314 81315 dix: move delivery stop condition out of event mask 81316 81317 Previously, this was only called if there was a mask match, so even if we 81318 had a no-propagate flag set or a stopAt window specified, if no mask 81319 triggered on the window we would recurse up to the root window and 81320 eventually deliver. 81321 Move this, so that the stopAt and do-not-propagate mask is honoured. 81322 81323 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81324 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81325 81326commit 1d01e861b69d7ca8fe2335270b8286bc115fb6e9 81327Author: Peter Hutterer <peter.hutterer@who-t.net> 81328Date: Thu Dec 8 10:22:26 2011 +1000 81329 81330 dix: split out core state and event state setting into helper functions 81331 81332 no functional changes. 81333 81334 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81335 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81336 81337commit fb010c543696b781a83b945f72865b7d7b75bf95 81338Author: Peter Hutterer <peter.hutterer@who-t.net> 81339Date: Wed Dec 7 14:41:18 2011 +1000 81340 81341 dix: move grab matching code into a helper function 81342 81343 No functional changes. 81344 81345 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81346 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81347 81348commit d253a262c2c690357a4db7e235c48ab5dd0b77f8 81349Author: Peter Hutterer <peter.hutterer@who-t.net> 81350Date: Wed Dec 7 13:57:25 2011 +1000 81351 81352 dix: split core grab interference check into helper function 81353 81354 No functional changes. 81355 81356 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81357 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81358 81359commit 97d79dd740ed1ac4eb02b0d4135850f953716278 81360Author: Peter Hutterer <peter.hutterer@who-t.net> 81361Date: Wed Dec 7 13:54:28 2011 +1000 81362 81363 dix: compare the grab type, not the tempGrab type 81364 81365 No functional change. To get here, GrabMatchesSecond() needs to be TRUE and 81366 for that the two grab types must be identical. 81367 81368 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81369 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81370 81371commit b9d6ae42b686287d7d23c2eefe2b12da98fe6ff8 81372Author: Peter Hutterer <peter.hutterer@who-t.net> 81373Date: Wed Dec 7 15:43:31 2011 +1000 81374 81375 dix: remove event type check 81376 81377 Can't remember why this is there but we'll need to pass in XI2 events soon, 81378 so this check is obsolete. 81379 81380 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81381 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81382 81383commit 80816366aa77f1730a1b6ddabfa55a946d76d494 81384Author: Chase Douglas <chase.douglas@canonical.com> 81385Date: Fri Nov 25 13:57:03 2011 -0800 81386 81387 dix: Split ActivatePassiveGrab() from CheckPassiveGrab() 81388 81389 The changed logic means we don't require the explicit grab = NULL setting 81390 and early exit anymore. Not 100% of it, but if we see that message pop up in 81391 a log we know it's broken. 81392 81393 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 81394 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81395 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81396 81397commit 9ee62cd8ce3c3effc3663f3d56b322385ce12fdb 81398Author: Chase Douglas <chase.douglas@canonical.com> 81399Date: Thu Nov 17 17:40:24 2011 -0800 81400 81401 dix: Move grab check and activation logic to CheckPassiveGrab() 81402 81403 This is needed for future pointer emulation work. 81404 81405 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 81406 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81407 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81408 81409commit c53651dabc66aeb9882819a404a799a364f00a29 81410Author: Peter Hutterer <peter.hutterer@who-t.net> 81411Date: Tue Dec 6 12:57:38 2011 +1000 81412 81413 dix: move EventDeliveryState into a header file, we'll need it elsewhere 81414 81415 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81416 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81417 81418commit 86c3137c81e924bacd919e9d65809e07afeabbff 81419Author: Peter Hutterer <peter.hutterer@who-t.net> 81420Date: Tue Dec 6 12:04:11 2011 +1000 81421 81422 Xi: split updating button count and state into helper functions 81423 81424 Functional change: for a button mapped to 0, the motionHintWindow is not 81425 updated to the NullWindow anymore. Before it got updated unconditionally to 81426 the button mapping. I have no idea what the practical effect of this is, but 81427 I guess it's closer to the correct behaviour: pressing a button that's 81428 logically disabled now does not disrupt the motion hint delivery. 81429 81430 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81431 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81432 81433commit e0f37250ffff5dcb3bc0e8cad63439995ce01a20 81434Author: Peter Hutterer <peter.hutterer@who-t.net> 81435Date: Tue Dec 6 11:40:33 2011 +1000 81436 81437 Xi: deduplicate button motion mask setting 81438 81439 No functional changes 81440 81441 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81442 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81443 81444commit 6eff14a789341d366b3013c5aa020e959c954651 81445Author: Peter Hutterer <peter.hutterer@who-t.net> 81446Date: Mon Dec 5 18:54:30 2011 +1000 81447 81448 dix: deduplicate callers of DeliverDeviceEvents in DeliverGrabbedEvents 81449 81450 No functional change. 81451 81452 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81453 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81454 81455commit 93945b0a74aa8156a88f52b8ba77f1210042f396 81456Author: Peter Hutterer <peter.hutterer@who-t.net> 81457Date: Mon Dec 5 18:42:05 2011 +1000 81458 81459 dix: split grab event conversion and delivery into a helper function 81460 81461 Effective functional change: XI2 events are checked with XACE now. 81462 81463 DeliverOneGrabbedEvent is exported for future use by touch events. 81464 81465 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81466 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81467 81468commit c81cdb0862e2184f033b3933e6bacbe0809ef2c0 81469Author: Peter Hutterer <peter.hutterer@who-t.net> 81470Date: Mon Dec 5 18:29:27 2011 +1000 81471 81472 dix: replace conversion errors with BUG_WARN_MSG 81473 81474 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81475 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81476 81477commit 6368c2aa4613a7c7eb0e8afca8d41f1a9bc4fc4d 81478Author: Peter Hutterer <peter.hutterer@who-t.net> 81479Date: Mon Dec 5 16:46:18 2011 +1000 81480 81481 dix: deduplicate event delivery code 81482 81483 Move all the event delivery code into DeliverOneEvent, based on the 81484 InputLevel we're sending to. 81485 81486 Functional change: we now check XI2 events with XACE too. 81487 81488 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81489 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 81490 81491commit 7e2207548b3173afc9accb5ccd532c181a7b94ed 81492Author: Chase Douglas <chase.douglas@canonical.com> 81493Date: Wed Dec 7 20:32:28 2011 +1000 81494 81495 include: Add an InputLevel enum 81496 81497 Currently unused, but will be in the future. 81498 81499 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 81500 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81501 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 81502 81503commit 898d97672907a8074031d0d1038cd44bb5caf40d 81504Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81505Date: Mon Dec 12 16:49:33 2011 -0800 81506 81507 verify_internal_event: preserve constness of data pointer 81508 81509 All we're using it for is ErrorF calls, so make it a const char * 81510 to stop gcc from warning: 81511 81512 inpututils.c: In function 'verify_internal_event': 81513 inpututils.c:629:9: warning: cast discards qualifiers from pointer target type 81514 81515 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81516 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81517 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81518 81519commit 79d09647d878e66721a778979a9eb1f4bba5f8d6 81520Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81521Date: Mon Dec 12 16:49:33 2011 -0800 81522 81523 Use const cast in BitIsOn macro to avoid angering gcc 81524 81525 Fixes gcc warnings such as: 81526 inpututils.c: In function 'valuator_mask_isset': 81527 inpututils.c:498:5: warning: cast discards qualifiers from pointer target type 81528 inpututils.c: In function 'CountBits': 81529 inpututils.c:613:9: warning: cast discards qualifiers from pointer target type 81530 81531 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81532 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81533 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81534 81535commit 99dfe9b1de09cb481e95bd4c45dcf7493480798b 81536Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81537Date: Mon Dec 12 16:49:32 2011 -0800 81538 81539 Fix deconstifying cast warning in xi2_get_type 81540 81541 Since we're just comparing values in the struct, cast it to a 81542 const xGenericEvent * to clear gcc warning of: 81543 81544 events.c: In function 'xi2_get_type': 81545 events.c:193:5: warning: cast discards qualifiers from pointer target type 81546 81547 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81548 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81549 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81550 81551commit dd80156bf05c9eb4000d0981e00d0d9fb69a94f6 81552Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81553Date: Wed Dec 7 22:58:45 2011 -0800 81554 81555 Include client name if available in PrintDeviceGrabInfo 81556 81557 Also adds missing newline to first line of output. 81558 81559 Before patch: 81560 81561 [3581472.414] (II) Printing all currently active device grabs: 81562 [3581472.414] Active grab 0x1800000 (core) on device 'Virtual core pointer' (2): 81563 client pid 26174 uid 0 gid 10 81564 [3581472.415] at 3581469139 (from active grab) (device thawed, state 1) 81565 [3581472.415] core event mask 0x0 81566 [3581472.415] owner-events true, kb 1 ptr 1, confine 0, cursor 0x0 81567 [3581472.415] Active grab 0x1800000 (core) on device 'Virtual core keyboard' (3) 81568 : client pid 26174 uid 0 gid 10 81569 [3581472.415] at 3581469139 (from active grab) (device thawed, state 1) 81570 [3581472.415] core event mask 0x3 81571 [3581472.415] owner-events true, kb 1 ptr 1, confine 0, cursor 0x0 81572 [3581472.415] (II) End list of active device grabs 81573 81574 After patch: 81575 [3581736.601] (II) Printing all currently active device grabs: 81576 [3581736.601] Active grab 0x1600000 (core) on device 'Virtual core pointer' (2): 81577 [3581736.601] client pid 26741 /usr/bin/xscreensaver -nosplash 81578 [3581736.601] at 3581735000 (from active grab) (device thawed, state 1) 81579 [3581736.601] core event mask 0x0 81580 [3581736.601] owner-events true, kb 1 ptr 1, confine 0, cursor 0x0 81581 [3581736.601] Active grab 0x1600000 (core) on device 'Virtual core keyboard' (3) 81582 : 81583 [3581736.601] client pid 26741 /usr/bin/xscreensaver -nosplash 81584 [3581736.601] at 3581735000 (from active grab) (device thawed, state 1) 81585 [3581736.601] core event mask 0x3 81586 [3581736.601] owner-events true, kb 1 ptr 1, confine 0, cursor 0x0 81587 [3581736.601] (II) End list of active device grabs 81588 81589 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81590 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 81591 81592commit b79de3f42ff8412dcabf2f2a805cbcbd1c6907c3 81593Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81594Date: Mon Dec 12 16:49:35 2011 -0800 81595 81596 xf86 parser: convert Error to a varargs macro to clear gcc format warnings 81597 81598 Previously it always passed a format string with exactly one argument, 81599 using NULL when the format string needed none. Now pass the right number 81600 of arguments to clear gcc warnings of 'too many arguments for format'. 81601 81602 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81603 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81604 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81605 81606commit 7801b3dcd6501e3de93a6d2cee93b2593806e922 81607Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81608Date: Mon Dec 12 16:49:35 2011 -0800 81609 81610 Add some printf format attributes suggested by gcc 81611 81612 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81613 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81614 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81615 81616commit f68df9dfd2c7ca03c24aaa0a895b7191ecb81e85 81617Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81618Date: Mon Dec 12 16:49:34 2011 -0800 81619 81620 xf86Priv.h: Add some noreturn attributes suggested by gcc 81621 81622 Both functions call exit() at the end and have no other return path. 81623 Also correct comment/heading to reflect commit 6450f6ca7ee0 moving 81624 DoShowOptions into xf86Configure.c. 81625 81626 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81627 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81628 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81629 81630commit 3823eedf3c124d3a20360480ba349bea72de3069 81631Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81632Date: Mon Dec 12 16:49:34 2011 -0800 81633 81634 Remove duplicate declarations of KdAdd*Driver in kdrive.h 81635 81636 Clears gcc warnings in every file that includes kdrive.h: 81637 kdrive.h:507:1: warning: redundant redeclaration of 'KdAddPointerDriver' 81638 kdrive.h:225:6: note: previous declaration of 'KdAddPointerDriver' was here 81639 kdrive.h:510:1: warning: redundant redeclaration of 'KdAddKeyboardDriver' 81640 kdrive.h:284:6: note: previous declaration of 'KdAddKeyboardDriver' was here 81641 81642 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81643 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81644 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81645 81646commit 471e5373b6f87336dac3b9de9b3852f63a2b10dc 81647Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81648Date: Mon Dec 12 16:49:34 2011 -0800 81649 81650 Remove duplicate declaration of xf86ValidateModesFlags in xf86Modes.h 81651 81652 Clears gcc warning in every file that includes xf86Modes.h: 81653 xf86Modes.h:102:1: warning: redundant redeclaration of 'xf86ValidateModesFlags' 81654 xf86Modes.h:72:1: note: previous declaration of 'xf86ValidateModesFlags' was here 81655 81656 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81657 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81658 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81659 81660commit 41dd7cf673e47ab74c5065b8dbf268b2e5facb64 81661Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81662Date: Mon Dec 12 16:49:34 2011 -0800 81663 81664 _XkbFilterDeviceBtn: move variable declarations to match usage scope 81665 81666 The main body of this function is an if { } else if { } pair of blocks. 81667 Previously there was int button at the top level scope which is used 81668 only in the first block, and a redeclaration of int button inside the 81669 second block. Since there's no overlap in the code paths for the 81670 two uses of button, move the one from the outer block into the first 81671 block to help the programmer more quickly determine they are unrelated 81672 usages, and to silence the gcc warning of: 81673 81674 xkbActions.c: In function '_XkbFilterDeviceBtn': 81675 xkbActions.c:999:6: warning: declaration of 'button' shadows a previous local 81676 xkbActions.c:955:6: warning: shadowed declaration is here 81677 81678 For consistency, move DeviceIntPtr dev declarations as well that are 81679 used in the same way. 81680 81681 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81682 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81683 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81684 81685commit d8eb9b5faa2d2c9350bdaffef8749a9cac3c87b8 81686Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81687Date: Mon Dec 12 16:49:34 2011 -0800 81688 81689 XkbFindSrvLedInfo: remove extraneous name-clashing sli variable 81690 81691 Variable is already defined outside the outer if statement, and 81692 there's no need to redefine inside the if statement. 81693 81694 No point in setting sli before if (dev->kbdfeed->xkb_sli==NULL) 81695 check - if check is true, we immediately set it, if check is false, 81696 we immediately return without further reference or use of it. 81697 81698 The one thing we do with it inside the inner if statement is store 81699 an allocation in it for a brief moment before writing to the final 81700 destination, which is immediately returned to the caller. 81701 81702 In short, there's no benefit to the variable at all in this block, 81703 it just gives the optimizer more code to figure out how to omit. 81704 81705 Fixes gcc warning: 81706 xkbLEDs.c: In function 'XkbFindSrvLedInfo': 81707 xkbLEDs.c:683:19: warning: declaration of 'sli' shadows a previous local 81708 xkbLEDs.c:679:18: warning: shadowed declaration is here 81709 81710 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81711 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81712 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81713 81714commit 9edfa47bd58cd318d57d0f114714fd146be35a79 81715Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81716Date: Mon Dec 12 16:49:34 2011 -0800 81717 81718 KdParseFindNext: Constify delim argument 81719 81720 It's only used as input to strchr to find the delimiters, never 81721 written to. 81722 81723 Clears a bunch of gcc warnings of the form: 81724 kdrive.c:323:2: warning: passing argument 2 of 'KdParseFindNext' discards qualifiers from pointer target type 81725 kdrive.c:261:1: note: expected 'char *' but argument is of type 'const char *' 81726 81727 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81728 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81729 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81730 81731commit ff64ad6c74a3e14ca34bacb8866bdbd2262bddff 81732Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81733Date: Mon Dec 12 16:49:34 2011 -0800 81734 81735 Convert KdDoSwitchCmd to use asprintf instead of malloc/strcat/etc. 81736 81737 Also fix the reason argument to be const char * to clear several gcc 81738 warnings of: 81739 kdrive.c:151:2: warning: passing argument 1 of 'KdDoSwitchCmd' discards qualifiers from pointer target type 81740 kdrive.c:116:1: note: expected 'char *' but argument is of type 'const char *' 81741 81742 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81743 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81744 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81745 81746commit 5bc590bde23ce1c57015b1d9e1cc63189c37448e 81747Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81748Date: Mon Dec 12 16:49:34 2011 -0800 81749 81750 DoShowOptions: preserve constness of options list as we walk it 81751 81752 Since all we do with the option list is walk down the list printing 81753 the names, there's no need to cast away its constness. 81754 81755 Clears gcc warning: 81756 xf86Configure.c: In function 'DoShowOptions': 81757 xf86Configure.c:781:4: warning: cast discards qualifiers from pointer target type 81758 81759 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81760 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81761 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81762 81763commit 71efd868282d47a6db405a16de18ac322926962b 81764Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81765Date: Mon Dec 12 16:49:34 2011 -0800 81766 81767 x86emu: constify debug strings 81768 81769 Strings are all pointers to literal constants, just used as input 81770 to printf calls when debugging is enabled. 81771 81772 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81773 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81774 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81775 81776commit d5a5eece670dee963765eab1431c21525c16d9ee 81777Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81778Date: Mon Dec 12 16:49:34 2011 -0800 81779 81780 CompareISOLatin1Lowered: constify arguments 81781 81782 Allows callers to avoid deconstifying arguments when calling, fixing 81783 gcc warning: 81784 81785 filter.c: In function 'PictureGetFilterId': 81786 filter.c:59:2: warning: cast discards qualifiers from pointer target type 81787 81788 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81789 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81790 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81791 81792commit 83ac9502ea9f1dedf3a8002745668af16bb1f2af 81793Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81794Date: Mon Dec 12 16:49:33 2011 -0800 81795 81796 xdmcp.c: fix three small const warnings 81797 81798 xdmcp.c:63:36: warning: initialization discards qualifiers from pointer target type 81799 81800 xdmcp.c: In function 'XdmcpRegisterConnection': 81801 xdmcp.c:482:8: warning: cast discards qualifiers from pointer target type 81802 xdmcp.c:482:8: warning: cast discards qualifiers from pointer target type 81803 xdmcp.c:482:8: warning: cast discards qualifiers from pointer target type 81804 81805 xdmcp.c: In function 'get_mcast_options': 81806 xdmcp.c:1596:21: warning: initialization discards qualifiers from pointer target type 81807 81808 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81809 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81810 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81811 81812commit 3839d1480877818f3ccc04e25237d853bab210c3 81813Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81814Date: Mon Dec 12 16:49:33 2011 -0800 81815 81816 LockServer: store path to LOCKDIR literal string in a const char * 81817 81818 And instead of initializing to NULL, then resetting to LOCKDIR almost 81819 immediately (before ever using the NULL value), skip directly to setting 81820 it to LOCKDIR. 81821 81822 tmppath variable is only used as input for generating the path name 81823 via calls to strlen, sprintf, etc. 81824 81825 Fixes gcc warning of: 81826 utils.c: In function 'LockServer': 81827 utils.c:259:11: warning: assignment discards qualifiers from pointer target type 81828 81829 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81830 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81831 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81832 81833commit af4113bfeb1d155f5f037c3492e50513336fa4a7 81834Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81835Date: Mon Dec 12 16:49:33 2011 -0800 81836 81837 WriteToClient: preserve constness of buf while extracting length value 81838 81839 Fixes gcc warning: 81840 io.c: In function 'WriteToClient': 81841 io.c:826:6: warning: cast discards qualifiers from pointer target type 81842 81843 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81844 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81845 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81846 81847commit 2c9800f91559fbb12dd276cf0536631104092f67 81848Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81849Date: Mon Dec 12 16:49:33 2011 -0800 81850 81851 OsInit: store "/dev/null" in a const char * 81852 81853 It's only passed as the input side of a strcpy and as the filename to 81854 fopen, so doesn't need to be non-const. Fixes gcc warning: 81855 81856 osinit.c: In function 'OsInit': 81857 osinit.c:154:28: warning: initialization discards qualifiers from pointer target type 81858 81859 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81860 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81861 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81862 81863commit 50b1097643934c8caec9530e5eda6ed6534aaf61 81864Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81865Date: Mon Dec 12 16:49:33 2011 -0800 81866 81867 Constify the reason string throughout the authorization check framework 81868 81869 Almost all of the places the string is assigned point to a literal 81870 string constant, so use const char * for those, and const char ** 81871 for function calls that return it via an argument. Fortunately 81872 the top level function, ClientAuthorized, which returns the string 81873 as its return value is called from only one place, ProcEstablishConnection. 81874 81875 ProcEstablishConnection stores either that return value or a string literal 81876 in char *reason. It only uses reason as an argument to SendConnSetup. 81877 SendConnSetup passes the reason argument to strlen & WriteToClient, 81878 both of which already have const qualifiers on their args. 81879 Thus added const to the reason variable in ProcEstablishConnection 81880 and the reason argument to SendConnSetup. 81881 81882 Fixes gcc warnings: 81883 dispatch.c: In function 'ProcEstablishConnection': 81884 dispatch.c:3711:9: warning: assignment discards qualifiers from pointer target type 81885 auth.c: In function 'CheckAuthorization': 81886 auth.c:218:14: warning: assignment discards qualifiers from pointer target type 81887 auth.c:220:20: warning: assignment discards qualifiers from pointer target type 81888 connection.c: In function 'ClientAuthorized': 81889 connection.c:683:3: warning: return discards qualifiers from pointer target type 81890 mitauth.c: In function 'MitCheckCookie': 81891 mitauth.c:88:13: warning: assignment discards qualifiers from pointer target type 81892 xdmauth.c:259:14: warning: assignment discards qualifiers from pointer target type 81893 xdmauth.c:270:14: warning: assignment discards qualifiers from pointer target type 81894 xdmauth.c:277:11: warning: assignment discards qualifiers from pointer target type 81895 xdmauth.c:293:15: warning: assignment discards qualifiers from pointer target type 81896 xdmauth.c:313:14: warning: assignment discards qualifiers from pointer target type 81897 xdmauth.c:322:11: warning: assignment discards qualifiers from pointer target type 81898 rpcauth.c: In function 'SecureRPCCheck': 81899 rpcauth.c:136:10: warning: assignment discards qualifiers from pointer target type 81900 81901 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81902 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81903 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81904 81905commit 99998196017ef38ec88459b50605aa20f628788b 81906Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81907Date: Mon Dec 12 16:49:33 2011 -0800 81908 81909 Constify string for authorization protocol names 81910 81911 gcc was warning from storing string constants in a char *name field: 81912 auth.c:64:1: warning: initialization discards qualifiers from pointer target type 81913 auth.c:72:1: warning: initialization discards qualifiers from pointer target type 81914 auth.c:81:1: warning: initialization discards qualifiers from pointer target type 81915 81916 Making the field const requires changing AuthorizationFromID to take 81917 a const char ** pointer for the name argument which it sets to point 81918 to the matching name entry. 81919 81920 Changing that argument requires changing its sole caller in the security 81921 extension to pass the address of a const char * variable to it, which it 81922 can do, since the only thing it does with the returned name is to pass 81923 it back to the RemoveAuthorization function that already expects a const 81924 char *name. 81925 81926 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81927 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81928 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81929 81930commit 01834e99e461e2a8354f3b4aef7f14c5e83fa255 81931Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81932Date: Mon Dec 12 16:49:33 2011 -0800 81933 81934 os/access.c: replace acmp & acopy macros with memcmp & memcpy calls 81935 81936 No need to cast to char * now that all supported platforms use C89-standard 81937 void * argument types, so just drop the casts from acmp & acopy macros, 81938 which clears the gcc warnings for places const pointers were cast non-const: 81939 81940 access.c: In function 'DefineSelf': 81941 access.c:786:3: warning: cast discards qualifiers from pointer target type 81942 access.c:795:6: warning: cast discards qualifiers from pointer target type 81943 access.c: In function 'NewHost': 81944 access.c:1293:9: warning: cast discards qualifiers from pointer target type 81945 access.c:1298:6: warning: cast discards qualifiers from pointer target type 81946 access.c:1309:5: warning: cast discards qualifiers from pointer target type 81947 81948 Without the casts, acmp & acopy are just a funny way to write memcmp 81949 & memmove, so drop the macros and inline the calls, taking care to 81950 swap the first two arguments to memmove since it had swapped them. 81951 81952 Since all the calls to memmove end up being to non-overlapping memory 81953 (mostly copying from an existing pointer to a newly allocated one), 81954 replace those with memcpy. 81955 81956 And finally, don't actually call memcpy to copy 0 bytes from one place 81957 to another, since that's just a waste of a perfectly good function call. 81958 81959 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81960 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81961 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81962 81963commit 020d83d361b0ab54ac962e97b12935be785f9a67 81964Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81965Date: Mon Dec 12 16:49:33 2011 -0800 81966 81967 xres.c: Preserve constness of string returned by LookupResourceName 81968 81969 MakeAtom now accepts a const char * so we don't need to cast down to 81970 char * anymore. Fixes gcc warning of: 81971 81972 xres.c: In function 'ProcXResQueryClientResources': 81973 xres.c:155:6: warning: cast discards qualifiers from pointer target type 81974 81975 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 81976 Reviewed-by: Jamey Sharp <jamey@minilop.net> 81977 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 81978 81979commit 2ddae8f0bd2a9ce0cd15bf3848393af29e615acf 81980Author: Alan Coopersmith <alan.coopersmith@oracle.com> 81981Date: Mon Dec 12 16:49:33 2011 -0800 81982 81983 constify strings in resource name registry 81984 81985 LookupResourceName already returned a const char *, so just needed 81986 to change the variable we're storing the list in to be a const char ** 81987 and then add const to the name argument to RegisterResourceName 81988 (which just stores name in the array) and CreateNewResourceType 81989 (which just passes name to RegisterResourceName). 81990 81991 Clears a bunch of gcc warnings of the form: 81992 registry.c:319:5: warning: passing argument 2 of 'RegisterResourceName' discards qualifiers from pointer target type 81993 registry.c:200:1: note: expected 'char *' but argument is of type 'const char *' 81994 81995 and from all the extensions: 81996 damageext.c: In function 'DamageExtensionInit': 81997 damageext.c:490:5: warning: passing argument 2 of 'CreateNewResourceType' discards qualifiers from pointer target type 81998 ../include/resource.h:159:26: note: expected 'char *' but argument is of type 'const char *' 81999 82000 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82001 Reviewed-by: Jamey Sharp <jamey@minilop.net> 82002 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 82003 82004commit 424dbde891486ad6a6c00c61a334031ff18f5556 82005Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82006Date: Fri Dec 9 10:08:55 2011 -0800 82007 82008 CheckForEmptyMask does not need to declare int n twice 82009 82010 Just use the existing n variable again in the ARGB_CURSOR loop 82011 instead of creating another one. 82012 82013 Fixes gcc -Wshadow warning: 82014 cursor.c: In function 'CheckForEmptyMask': 82015 cursor.c:155:6: warning: declaration of 'n' shadows a previous local 82016 cursor.c:146:9: warning: shadowed declaration is here 82017 82018 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82019 Reviewed-by: Jamey Sharp <jamey@minilop.net> 82020 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 82021 82022commit b2bc38e4a553c29f49a0284333b34b4d6c8a8c12 82023Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82024Date: Sat Dec 10 10:01:18 2011 -0800 82025 82026 Even more correctly free config file names 82027 82028 If we didn't go into the if (!autoconfig) { } block, the filename, 82029 dirname, and sysdirname pointers were never initialized, but we 82030 freed them outside the block, leading to potential memory corruption. 82031 82032 Move the frees inside the block where they're initialized to avoid this. 82033 82034 To avoid similar problems, move the declarations of the variables that 82035 are only used in this block inside the block. 82036 82037 Regression introduced by commit 3d635fe84d6de53e2f74203b10e89f7851fe3fc1 82038 82039 Found by gcc warning: 82040 xf86Config.c: In function 'xf86HandleConfigFile': 82041 xf86Config.c:2303:11: warning: 'filename' may be used uninitialized in this function 82042 xf86Config.c:2303:22: warning: 'dirname' may be used uninitialized in this function 82043 xf86Config.c:2303:32: warning: 'sysdirname' may be used uninitialized in this function 82044 82045 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82046 Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 82047 82048commit 33d6e6743d86324c2078f156991b16ac4f2593fc 82049Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82050Date: Wed Dec 7 19:06:05 2011 -0800 82051 82052 xf86RegisterRootWindowProperty is confused about xnfcalloc 82053 82054 It will never return NULL, so don't try to handle a NULL condition, 82055 since that just confuses programmers and static analyzers. 82056 82057 It uses calloc, so all the allocated memory is cleared, so there's 82058 no point looping over the memory to manually initialize it NULL. 82059 82060 And just because it's annoying, it doesn't need to be the only 82061 place in this file to do if (NULL==...) instead of if (... == NULL). 82062 82063 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82064 Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 82065 82066commit e5aa00989cda9ebd18063c5e955235123ad37b88 82067Author: Peter Hutterer <peter.hutterer@who-t.net> 82068Date: Wed Dec 7 14:14:10 2011 +1000 82069 82070 Change GetXI2/XI/CoreType to just take a type argument 82071 82072 Avoids the dummy-event dance if we have an event type and need to get the 82073 matching XI2 type. 82074 82075 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82076 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82077 82078commit 6cc0e6a0afa6a87802767d484aa4e68fa935d3eb 82079Author: Peter Hutterer <peter.hutterer@who-t.net> 82080Date: Fri Oct 28 12:55:55 2011 +1000 82081 82082 include: add GetXI2MaskByte and GetXI2EventFilterMask to headers 82083 82084 This is needed for touch event processing. 82085 82086 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82087 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82088 82089commit 372a6f10dc2d74d2d179e8b92449e9b8636a99ef 82090Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82091Date: Thu Dec 8 21:52:07 2011 -0800 82092 82093 Change disable_clientpointer return type to void 82094 82095 It doesn't return anything, nor does it's caller expect it to. 82096 82097 Fixes Solaris Studio compiler error: 82098 "xichangehierarchy.c", line 214: Function has no return statement : disable_clientpointer 82099 82100 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82101 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 82102 Signed-off-by: Keith Packard <keithp@keithp.com> 82103 82104commit 522f8bcc0360d6a117e929a004bc956ab92037e9 82105Merge: 6369b5966 2abe83df6 82106Author: Keith Packard <keithp@keithp.com> 82107Date: Thu Dec 8 20:57:26 2011 -0800 82108 82109 Merge remote-tracking branch 'whot/for-keith' 82110 82111commit 2abe83df686ed64c4f4df711ac3c1fd12131c2e4 82112Author: Peter Hutterer <peter.hutterer@who-t.net> 82113Date: Mon Dec 5 14:02:51 2011 +1000 82114 82115 include: add BUG_WARN_MSG for custom error message on bug condition 82116 82117 __BUG_WARN_MSG is a simple helper to enable call with and without varargs. I 82118 couldn't find a way to otherwise do this without getting gcc warnings. 82119 82120 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82121 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82122 82123commit 4fc797f3756611a97767f407e1af0b6a7cf2f1d9 82124Author: Peter Hutterer <peter.hutterer@who-t.net> 82125Date: Fri Nov 11 16:25:30 2011 +1000 82126 82127 xfree86: include xorg-config.h from xaalocal.h 82128 82129 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82130 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82131 82132commit 8c9589c71d47f287588314d77ddbfcc22cd04c8a 82133Author: Peter Hutterer <peter.hutterer@who-t.net> 82134Date: Mon Dec 5 11:55:58 2011 +1000 82135 82136 Xi: rename "state" to "corestate" in ProcessDeviceEvents 82137 82138 'state' is shadowed by the XKB 'state' as well (which feeds into the event 82139 too), so rename this one to clarify that this is the core event state only. 82140 82141 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82142 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82143 82144commit 631516a4aa9858874ee197444cd93d91b97a1089 82145Author: Peter Hutterer <peter.hutterer@who-t.net> 82146Date: Fri Dec 2 15:47:58 2011 +1000 82147 82148 Xi: check button mapping value _before_ assigning it 82149 82150 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82151 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82152 82153commit a1304d6cb69301899c3c8450d6bf3e68573599df 82154Author: Peter Hutterer <peter.hutterer@who-t.net> 82155Date: Mon Dec 5 11:26:30 2011 +1000 82156 82157 Xi: skip superfluous cast 82158 82159 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82160 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82161 82162commit 1ab50be938524dcd4a9e56d27e3b96a27c2db2c0 82163Author: Peter Hutterer <peter.hutterer@who-t.net> 82164Date: Wed Nov 30 09:06:06 2011 +1000 82165 82166 xfixes: don't dereference a NULL cursor 82167 82168 If the new cursor is the NULL cursor, don't dereference it and use zeros 82169 instead. 82170 82171 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82172 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 82173 82174commit d2ebbcdaf6b13d70eee704b1764ff349e1be22a0 82175Author: Peter Hutterer <peter.hutterer@who-t.net> 82176Date: Tue Nov 29 16:15:37 2011 +1000 82177 82178 Xi: when removing a device, reset ClientPointers where needed 82179 82180 if a client had the to-be-removed device as ClientPointer, reset to NULL. 82181 82182 Fixes #43165 82183 82184 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82185 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 82186 82187commit 27425f07b29e0ddaa782345c1899273ca742891e 82188Author: Peter Hutterer <peter.hutterer@who-t.net> 82189Date: Wed Nov 9 14:45:02 2011 +1000 82190 82191 dix: use BUG_WARN for input mask size issues 82192 82193 Yes, we're likely corrupting memory here but really this is unlikely to be 82194 triggered other than a real bug in the server. In which case a stacktrace is 82195 going to be more useful than any silent error handling. 82196 82197 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82198 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82199 82200commit 9b570ecbda954227c89938ee6f94b9efd192d3c6 82201Author: Peter Hutterer <peter.hutterer@who-t.net> 82202Date: Fri Dec 9 10:48:05 2011 +1000 82203 82204 xfree86: bump the input ABI 82205 82206 The last few patches broke the ABI, bump it for convenience. 82207 82208 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82209 82210commit 86bb3781b336c09e4279136ed81974de5acdba7f 82211Author: Peter Hutterer <peter.hutterer@who-t.net> 82212Date: Fri Nov 4 11:29:01 2011 +1000 82213 82214 input: swap the server over to use the XI2mask struct 82215 82216 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82217 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82218 82219commit b8b90cd1610331ff12fa3f70bf372670af7795ec 82220Author: Peter Hutterer <peter.hutterer@who-t.net> 82221Date: Thu Nov 3 13:39:59 2011 +1000 82222 82223 Add a new XI2Mask struct and a few helper functions. 82224 82225 The current XI2 mask handling is handy for copying (fixed size arrays) but a 82226 pain to deal with otherwise. Add a struct for XI2 masks and the required 82227 accessors. 82228 82229 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82230 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82231 82232commit 4bc2761ad5ec2d0668aec639780ffb136605fbc8 82233Author: Peter Hutterer <peter.hutterer@who-t.net> 82234Date: Fri Nov 4 14:16:37 2011 +1000 82235 82236 dix: switch the dev->deviceGrab.activeGrab from GrabRec to GrabPtr 82237 82238 This breaks the input ABI. 82239 82240 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82241 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82242 82243commit b601ea769f1b8a4d7f19e9d4a13541c78e865fe5 82244Author: Peter Hutterer <peter.hutterer@who-t.net> 82245Date: Fri Nov 4 10:47:27 2011 +1000 82246 82247 dix: allocate temporary grabs on the heap 82248 82249 Once grabs start having nested memory locations, we can't just use the 82250 GrabRec on the stack anymore, we need to alloc/copy/free the grabs. 82251 82252 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82253 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82254 82255commit b0e9e2e32616d09c30a02b9d0ae9db0b13e150d1 82256Author: Peter Hutterer <peter.hutterer@who-t.net> 82257Date: Fri Nov 4 10:44:31 2011 +1000 82258 82259 dix: add CopyGrab() function 82260 82261 Not really needed at this point, but will be once touch support is added. 82262 Since grabs are now expected to be allocated/freed with AllocGrab and 82263 FreeGrab, CopyGrab must increase the refcount and duplicate the modifier 82264 masks. Until the callers are switched to use FreeGrab, this introduces 82265 memleaks. 82266 82267 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82268 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82269 82270commit 347f377f3b3f8c9d230d6309ec8ae92aa86d78b7 82271Author: Peter Hutterer <peter.hutterer@who-t.net> 82272Date: Thu Nov 3 16:12:09 2011 +1000 82273 82274 dix: add AllocGrab helper function 82275 82276 Not needed since the GrabRec is a self-contained struct but will be needed 82277 for the xi2 input mask rework. 82278 FreeGrab already exists, make it available to other callers. 82279 82280 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82281 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82282 82283commit 7af23259d88f4c28ed21140f82cc03b3724c06bb 82284Author: Peter Hutterer <peter.hutterer@who-t.net> 82285Date: Mon Nov 21 11:41:12 2011 -0800 82286 82287 dix: switch the syncEvent queue to a struct list 82288 82289 No effective functionality change, just cleanup to make this code slightly 82290 more sane. 82291 82292 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82293 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82294 82295commit 6369b59668613e51131857fbaeaa393a65a73048 82296Author: Gaetan Nadon <memsize@videotron.ca> 82297Date: Thu Dec 1 16:24:32 2011 -0500 82298 82299 test: update .gitignore with xfree86 and sort alphabetically 82300 82301 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82302 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 82303 Signed-off-by: Keith Packard <keithp@keithp.com> 82304 82305commit 3ab8ee32470f7cf1223e04238bb8c5f74ed52fc3 82306Merge: 22a666f99 98c4a888a 82307Author: Keith Packard <keithp@keithp.com> 82308Date: Wed Dec 7 12:42:17 2011 -0800 82309 82310 Merge remote-tracking branch 'airlied/reviewed-fixes' 82311 82312commit 22a666f9952feb7248e9bb2faf777edaaac8175f 82313Merge: 3824f558c e89b0324d 82314Author: Keith Packard <keithp@keithp.com> 82315Date: Wed Dec 7 12:27:23 2011 -0800 82316 82317 Merge remote-tracking branch 'alanc/master' 82318 82319commit 3824f558cc2ee051da8314c7bf08b8647b44e84a 82320Author: Keith Packard <keithp@keithp.com> 82321Date: Wed Dec 7 12:13:37 2011 -0800 82322 82323 hw/xfree86: fix segfault in config parser when config dir is missing 82324 82325 Treat a scandir error from a missing (or unusable) directory return as 82326 if it simply returned no files at all, which is what we want. 82327 82328 cc: Paulo Zanoni <przanoni@gmail.com> 82329 Signed-off-by: Keith Packard <keithp@keithp.com> 82330 82331commit 98c4a888a4428789386c7c47cecc81933b5999ba 82332Author: Dave Airlie <airlied@redhat.com> 82333Date: Mon Nov 28 16:37:59 2011 +0000 82334 82335 kdrive: drop screen crossing code. 82336 82337 The only kdrive server we probably care about anymore is Xephyr, 82338 and this screen enable/disable code totally breaks it in multi-screen mode. 82339 82340 When you are in one screen the other stops updating. 82341 82342 Fixes https://bugzilla.redhat.com/show_bug.cgi?id=757457 82343 82344 Signed-off-by: Dave Airlie <airlied@redhat.com> 82345 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 82346 82347commit 8d3731a811e33e263920dd7c8ec63d02968cb56e 82348Author: Dave Airlie <airlied@redhat.com> 82349Date: Thu Oct 20 10:48:26 2011 +0100 82350 82351 hal: free tmp_val in one missing case 82352 82353 Pointed out by coverity scan. 82354 82355 Signed-off-by: Dave Airlie <airlied@redhat.com> 82356 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 82357 82358commit 41229392b790f30a0f0ef1f4ed95647c5bca4001 82359Author: Dave Airlie <airlied@redhat.com> 82360Date: Thu Oct 20 11:00:43 2011 +0100 82361 82362 xv: test correct number of requests. (v2) 82363 82364 Pointed out by coverity. 82365 82366 v2: fix swapped as well, as pointed out by Alan 82367 82368 Signed-off-by: Dave Airlie <airlied@redhat.com> 82369 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82370 82371commit 1049139499d9132a20cd6d4d156fe9da9cddb6c2 82372Author: Dave Airlie <airlied@redhat.com> 82373Date: Wed Oct 19 16:57:13 2011 +0100 82374 82375 xaa: avoid possible freed pointer reuse in epilogue 82376 82377 If the pGCPriv->flags == 2, then we try to assign the freed pGCPriv->XAAOps 82378 avoid this by clearing the flags in to be destroyed pGCPriv. 82379 82380 Reported by coverity. 82381 82382 Signed-off-by: Dave Airlie <airlied@redhat.com> 82383 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 82384 82385commit 682c09a2cedd234b005334cc01247d859dd7f26a 82386Author: Dave Airlie <airlied@redhat.com> 82387Date: Wed Oct 19 16:22:31 2011 +0100 82388 82389 Xi: avoid overrun of callback array. 82390 82391 This code had an off-by-one and would allow writing one past the end of 82392 the callbacks array. 82393 82394 Pointed out by coverity. 82395 82396 Signed-off-by: Dave Airlie <airlied@redhat.com> 82397 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 82398 82399commit b62dc4fcbcffd10de16650bee284702c8608bb60 82400Author: Dave Airlie <airlied@redhat.com> 82401Date: Wed Oct 19 16:21:26 2011 +0100 82402 82403 xext: don't free uninitialised pointer when malloc fails. (v2) 82404 82405 Initialise the pAttr->values to values so if the values allocation 82406 fails it just ends up as free(NULL). 82407 82408 Pointed out by coverity. 82409 82410 v2: use Alan's suggestion. 82411 82412 Signed-off-by: Dave Airlie <airlied@redhat.com> 82413 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82414 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 82415 82416commit 22605effd188436629a0dbc688666549473741e4 82417Author: Adam Jackson <ajax@redhat.com> 82418Date: Thu Apr 28 13:34:28 2011 +1000 82419 82420 fbdevhw: iterate over all modes that match a mode. (v3) 82421 82422 So on RHEL5 anaconda sets an xorg.conf with a fixed 800x600 mode in it, 82423 we run radeonfb and fbdev since ati won't work in userspace due to domain 82424 issues in the older codebase. 82425 82426 On certain pseries blades the built-in KVM can't accept an 800x600-43 mode, 82427 it requires the 800x600-60 mode, so we have to have the kernel radeonfb 82428 driver reject the 800x600-43 mode when it sees it. However then fbdev 82429 doesn't try any of the other 800x600 modes in the modelist, and we end up 82430 getting a default 640x480 mode we don't want. 82431 82432 This patch changes the mode validation loop to continue on with the other modes 82433 that match to find one that works. 82434 82435 v2: move code around to avoid extra loop, after comment from Jamey. 82436 v3: move loop setup back into loop as per Jeremy's review. 82437 82438 Signed-off-by: Dave Airlie <airlied@redhat.com> 82439 Reviewed-by: Jamey Sharp <jamey@minilop.net> 82440 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 82441 82442commit fac464e310b82fadcedf790798d1016c4805640b 82443Author: Peter Hutterer <peter.hutterer@who-t.net> 82444Date: Fri Dec 2 08:52:53 2011 +1000 82445 82446 include: rename "foos" to "list_of_foos" in the struct list documentation 82447 82448 Makes things a little easier to read. 82449 82450 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82451 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82452 82453commit 7dfe8c32a96d3f96c8aaeb2802f5b122e381a1e4 82454Author: Peter Hutterer <peter.hutterer@who-t.net> 82455Date: Fri Dec 2 08:51:24 2011 +1000 82456 82457 include: update struct list documentation to use one set of structs only 82458 82459 The example at the top of the file used a struct bar and a list of struct 82460 foos. Use those two throughout instead of a different struct foo for the 82461 examples and for the API documentation. 82462 82463 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82464 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82465 82466commit 18539c89eca8f6e272ead2b631365da530065ae7 82467Author: Peter Hutterer <peter.hutterer@who-t.net> 82468Date: Fri Dec 2 08:51:04 2011 +1000 82469 82470 include: a new list element does not need initialization, state so 82471 82472 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82473 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82474 82475commit 58dc73314b6508121ca094bbcf00612fe19ed69f 82476Author: Peter Hutterer <peter.hutterer@who-t.net> 82477Date: Fri Dec 2 08:43:45 2011 +1000 82478 82479 include: point to the tests in list.c as examples 82480 82481 Even with the documentation, the list.c tests are the best examples. 82482 82483 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82484 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82485 82486commit c8e451a8e719ba432bcfa8976774c07307087809 82487Author: Peter Hutterer <peter.hutterer@who-t.net> 82488Date: Thu Dec 1 14:12:11 2011 +1000 82489 82490 include: add list_last_entry to get the tail of a list 82491 82492 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82493 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82494 82495commit 6acebf9e1298939593b942ec91ae9ec9e74faa19 82496Author: Peter Hutterer <peter.hutterer@who-t.net> 82497Date: Thu Dec 1 13:35:50 2011 +1000 82498 82499 include: add list_append() 82500 82501 The existing list_add() prepends to the list, but in some cases we need the 82502 list ordered in the way we append the elements. 82503 82504 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82505 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82506 82507commit e89b0324da89ba5c0ba64af1ef46a12b7f55f879 82508Author: Gaetan Nadon <memsize@videotron.ca> 82509Date: Sat Dec 3 21:23:51 2011 -0500 82510 82511 autoconf: the minimum required level for autoconf is 2.60 82512 82513 The toolchain requirements are documented here: 82514 http://www.x.org/wiki/ModularDevelopersGuide#Required_Tools 82515 82516 Note that autoconf features only found in versions later 82517 than 2.60 must not be used. 82518 82519 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 82520 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82521 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82522 82523commit 427cb53fddf0c517d1fd1025a87b1104735edd2a 82524Author: Gaetan Nadon <memsize@videotron.ca> 82525Date: Sat Dec 3 21:23:21 2011 -0500 82526 82527 autoconf: remove redundant AC_PROG_MAKE_SET 82528 82529 Already included during Automake initialization. 82530 82531 After the patch, no change: 82532 configure:3893: checking whether make sets $(MAKE) 82533 configure:3915: result: yes 82534 82535 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 82536 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82537 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82538 82539commit c19c55a93ab1bc6db56075ca34d6e1c3f522f622 82540Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82541Date: Fri Dec 2 00:09:07 2011 -0800 82542 82543 Fix builds of Xnest & Xephyr with Solaris Studio compilers 82544 82545 Required in order to build with Studio cc now that xorg-macros is 82546 setting -errwarn=E_FUNC_HAS_NO_RETURN_STMT since a bug in the Solaris 82547 system headers causes the noreturn attribute to not be correctly 82548 applied to the exit() prototype in <stdlib.h> when building with 82549 Studio instead of gcc. 82550 82551 Otherwise compiler exits with errors: 82552 "Display.c", line 65: Function has no return statement : x_io_error_handler 82553 "hostx.c", line 341: Function has no return statement : x_io_error_handler 82554 82555 Uses Studio-specific pragma instead of adding another exit() prototype 82556 with a noreturn attribute to avoid causing gcc to warn about having 82557 a redundant prototype for the exit() function. 82558 82559 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82560 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 82561 82562commit e4dcf580f007d6adcf5b0c639865d7aaab1a319d 82563Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82564Date: Thu Dec 1 23:45:42 2011 -0800 82565 82566 LoaderOpen returns either a valid pointer or NULL, so don't check for < 0 82567 82568 Fixes Sun cc warning that was recently elevated to error by the 82569 stricter default CFLAGS changes to xorg-macros: 82570 82571 "loadmod.c", line 914: improper pointer/integer combination: op "<" 82572 82573 Should have been changed when commit ab7f057ce9df4e905b12 changed the 82574 LoaderOpen return type from int to void *. 82575 82576 Changes log message when file is found but dlopen() fails from: 82577 (EE) LoadModule: Module dbe does not have a dbeModuleData data object. 82578 (EE) Failed to load module "dbe" (invalid module, 0) 82579 to: 82580 (EE) Failed to load module "dbe" (loader failed, 7) 82581 82582 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82583 Reviewed-by: Adam Jackson <ajax@redhat.com> 82584 82585commit 8b6a75009746db4b8e2e180210ac097e35136bcd 82586Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82587Date: Wed Nov 30 23:01:34 2011 -0800 82588 82589 Fix gcc warnings about redundant declarations of fallback functions 82590 82591 Ensure ffs, strndup, strlcat, etc. aren't defined by our headers 82592 if they're already defined in the system headers. 82593 82594 This does export the HAVE_FFS, HAVE_STRNDUP, etc. definitions to drivers, 82595 but if you built the Xserver with a libc that had those, and then build 82596 the drivers with a less capable libc, you're going to have problems anyway, 82597 and this should solve some reported problems with conflicts between our 82598 strndup definition and gcc magic for it. 82599 82600 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82601 Reviewed-by: Adam Jackson <ajax@redhat.com> 82602 82603commit d829a7c5cb42c979b58f3547136df5b05d906423 82604Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82605Date: Wed Nov 30 22:20:09 2011 -0800 82606 82607 Move to autoconf standard function name checks & defines 82608 82609 Replace multiple methods of checking for functions with AC_CHECK_FUNCS 82610 Replace multiple methods of selecting fallback funcs with AC_REPLACE_FUNCS 82611 Replace HAS_* and NEED_* #defines with autogenerated HAVE_* 82612 82613 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82614 Reviewed-by: Adam Jackson <ajax@redhat.com> 82615 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 82616 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 82617 82618commit 4be68b03128e958d2a6dc1b7feb3587329b9561b 82619Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82620Date: Wed Nov 30 19:39:51 2011 -0800 82621 82622 Remove unused function checks from configure.ac & include/*.h.in 82623 82624 The code that used getisax to check for MMX support was moved to pixman 82625 and removed from the X server by commit eb2d7fe02f9cbc. 82626 82627 The code that used HAVE_MKSTEMP was deleted by the Xprint removal in 82628 commit 1c8bd318fbaf. 82629 82630 All alloca calls were removed by the patch series end in commit 5e363500c8, 82631 and used custom X checks instead of the autoconf HAVE_ALLOCA anyway. 82632 82633 I can find no record of HAVE_GETUID, HAVE_GETEUID, HAVE_LINK, HAVE_MEMMOVE, 82634 HAVE_MEMSET, HAVE_STRCHR, HAVE_STRRCHR, HAVE_GETOPT, HAVE_GETOPT_LONG, 82635 HAVE_DOPRNT, or HAVE_VPRINTF ever being used, and the calls to those 82636 functions are not wrapped in #ifdefs. 82637 (Most of those are in our baseline requirements of C89 & Unix98 anyway.) 82638 82639 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82640 Reviewed-by: Adam Jackson <ajax@redhat.com> 82641 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 82642 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 82643 82644commit 84207def93e4489df8dca8e044ce544be8a3f845 82645Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82646Date: Mon Nov 28 20:38:58 2011 -0800 82647 82648 Disable building of tests requiring DDX functions when Xorg is not built 82649 82650 Some test cases require linking with some sort of DDX - ideally we'd 82651 have a stub ddx for testing, but for now, since we link with the Xorg 82652 ddx, disable those tests when configured with --disable-xorg 82653 82654 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=43320 82655 82656 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82657 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 82658 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 82659 82660commit fd976e4051e15ab2d01ec0bf89ff26926d4ed04e 82661Author: Gaetan Nadon <memsize@videotron.ca> 82662Date: Thu Dec 1 16:24:12 2011 -0500 82663 82664 config: conftest needs an include directive to locate X11/Xfuncproto.h 82665 82666 The compiler option -fvisibility=hidden is erroneously missing 82667 due to a faulty configuration test. The gcc command is unable to 82668 locate X11/Xfuncproto.h unless the build occurs on a 82669 system where X11 development headers are installed. 82670 82671 configure:21294: checking for symbol visibility support 82672 configure:21323: gcc -std=gnu99 -c -g -O2 -fvisibility=hidden conftest.c >&5 82673 conftest.c:144:28: fatal error: X11/Xfuncproto.h: No such file or directory 82674 82675 The solution is to add an include directive to obtain the location 82676 of X11/Xfuncproto.h which may or may not be the system installed headers. 82677 82678 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82679 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 82680 Signed-off-by: Keith Packard <keithp@keithp.com> 82681 82682commit feebf6746374aa04b12e9e3e51313a3a82c03530 82683Author: Alan Coopersmith <alan.coopersmith@oracle.com> 82684Date: Wed Nov 23 00:30:02 2011 -0800 82685 82686 Limit the number of screens Xvfb will attempt to allocate memory for 82687 82688 Commit f9e3a2955d2ca7 removing the MAXSCREEN limit left the screen 82689 number too unlimited, and allowed any positive int for a screen number: 82690 82691 Xvfb :1 -screen 2147483647 1024x1024x8 82692 82693 Fatal server error: 82694 Not enough memory for screen 2147483647 82695 82696 Found by Parfait 0.3.7: 82697 Error: Integer overflow (CWE 190) 82698 Integer parameter of memory allocation function realloc() may overflow due to multiplication with constant value 1112 82699 at line 293 of hw/vfb/InitOutput.c in function 'ddxProcessArgument'. 82700 82701 Since the X11 connection setup only has a CARD8 for number of SCREENS, 82702 limit to 255 screens, which is also low enough to avoid overflow on the 82703 sizeof(*vfbScreens) * (screenNum + 1) calculation for realloc. 82704 82705 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82706 Reviewed-by: Jamey Sharp <jamey@minilop.net> 82707 82708commit fb22a408c69a84f81905147de9e82cf66ffb6eb2 82709Author: Keith Packard <keithp@keithp.com> 82710Date: Tue Nov 8 10:13:15 2011 -0800 82711 82712 Save major/minor opcodes in ClientRec for RecordAReply 82713 82714 The record extension needs the major and minor opcodes in the reply 82715 hook, but the request buffer may have been freed by the time the hook 82716 is invoked. Saving the request major and minor codes as the request is 82717 executed avoids fetching from the defunct request buffer. 82718 82719 This patch also eliminates the public MinorOpcodeOfRequest function, 82720 inlining it into Dispatch. Usages of that function have been replaced 82721 with direct access to the new ClientRec field. 82722 82723 Signed-off-by: Keith Packard <keithp@keithp.com> 82724 Reviewed-by: Jamey Sharp <jamey@minilop.net> 82725 82726commit 05f09354a30a4f5edd421220e1aa97be754c71bb 82727Merge: 3197b773c 08ec4da6a 82728Author: Keith Packard <keithp@keithp.com> 82729Date: Thu Dec 1 14:44:52 2011 +0000 82730 82731 Merge remote-tracking branch 'jcristau/for-keith' 82732 82733commit 3197b773c81e58cd4aa412eccaee49526012ed6f 82734Merge: 2dc5ba4a1 dfcec1d3f 82735Author: Keith Packard <keithp@keithp.com> 82736Date: Thu Dec 1 14:36:58 2011 +0000 82737 82738 Merge remote-tracking branch 'whot/for-keith' 82739 82740commit 2dc5ba4a1b127e3ed286718ab0a6c20438361192 82741Author: Matt Turner <mattst88@gmail.com> 82742Date: Tue Nov 29 23:52:19 2011 -0500 82743 82744 Remove another if (E != NULL) check around free(E) 82745 82746 I wonder if there are any other patterns we haven't seen yet? 82747 82748 Reviewed-by: Jamey Sharp <jamey@minilop.net> 82749 Signed-off-by: Matt Turner <mattst88@gmail.com> 82750 Signed-off-by: Keith Packard <keithp@keithp.com> 82751 82752commit dfcec1d3f9d7bac5cde9eb034341428ee0ad3728 82753Author: Peter Hutterer <peter.hutterer@who-t.net> 82754Date: Fri Nov 11 08:55:14 2011 +1000 82755 82756 test: remove unneeded printf statements from misc.c 82757 82758 Leftover from debugging, is not really needeed in a test. 82759 82760 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82761 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82762 82763commit 9c38422fc4cf853c3104fada2a3851c79df2a66f 82764Author: Peter Hutterer <peter.hutterer@who-t.net> 82765Date: Wed Nov 9 14:37:26 2011 +1000 82766 82767 include: add BUG_WARN macro for internal bug cases. 82768 82769 There are plenty of cases that can only be triggered by a real bug in the 82770 server and doing the ErrorF dance manually everywhere is a tad painful and 82771 the error message is usually used only to find the spot in the file anyway. 82772 82773 Plus, reading BUG_WARN somewhere is a good indicator to the casual reader 82774 that this isn't intended behaviour. 82775 82776 Note that this is intentionally different to the BUG_ON behaviour on the 82777 kernel, we do not FatalError the server. It's just a warning + stacktrace. 82778 If the bug is really fatal, call FatalError. 82779 82780 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82781 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82782 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 82783 82784commit 61ef4daf6450da573b9de72ba7b200566821b916 82785Author: Peter Hutterer <peter.hutterer@who-t.net> 82786Date: Fri Nov 4 15:49:23 2011 +1000 82787 82788 Xi: add FreeInputMask function 82789 82790 Does what it says on the box, complements MakeInputMask. 82791 82792 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82793 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82794 82795commit ee9346bb31efce4036df1b8dd8f1a5dc43ae955a 82796Author: Peter Hutterer <peter.hutterer@who-t.net> 82797Date: Thu Nov 3 15:45:34 2011 +1000 82798 82799 Xi: add helper functions to alloc/free InputClientPtrs 82800 82801 Currently not needed since the InputClientRec is a self-contained struct. As 82802 part of the touch rework that won't be the case in the future and a function 82803 to allocate/free memory appropriately is required. 82804 82805 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82806 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82807 82808commit 4acf999c294868a44e559d212c6d88a4978256b2 82809Author: Peter Hutterer <peter.hutterer@who-t.net> 82810Date: Fri Nov 4 15:37:32 2011 +1000 82811 82812 dix: use a single return statement in CheckPassiveGrabsOnWindow 82813 82814 No functional change. 82815 82816 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82817 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82818 82819commit b371795f01c1d7fc338cfe18b8a567ed9f402846 82820Author: Peter Hutterer <peter.hutterer@who-t.net> 82821Date: Thu Nov 3 14:54:06 2011 +1000 82822 82823 dix: rename GetWindowXI2Mask to WindowXI2MaskIsset 82824 82825 And let it return a boolean value, that's all the callers need anyway. 82826 82827 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82828 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82829 82830commit bedb8fd90de8e2db33d5e3b1d859f24bf34efc9a 82831Author: Peter Hutterer <peter.hutterer@who-t.net> 82832Date: Thu Nov 3 14:25:51 2011 +1000 82833 82834 Xi: use single return code from XIPassiveGrabDevice 82835 82836 Some failures returned status but the actual return code was "ret". Use 82837 "ret" consistently and move status to the local block is used in. 82838 82839 [the goto isn't necessary yet, but for a future patch] 82840 82841 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82842 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82843 82844commit 2aad1a2b42b7def7812abfa2462b6bcc6382e03a 82845Author: Peter Hutterer <peter.hutterer@who-t.net> 82846Date: Wed Nov 2 14:07:19 2011 +1000 82847 82848 include: fix mask size calculation 82849 82850 Same bug as inputproto-2.0.1-9-gb1149ab, if the XI2LASTEVENT was a multiple 82851 of 8, the mask was one bit too short. 82852 82853 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 82854 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 82855 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 82856 82857commit b2015a2c01711646bb7ae23d684abee0cd55d4d0 82858Author: Jeremy Huddleston <jeremyhu@apple.com> 82859Date: Thu Nov 24 13:54:08 2011 -0800 82860 82861 dmx: Build fix for -Werror=implicit-function-declaration on linux 82862 82863 Fixes regression introduced by: 6e6d732bac3c21cb85f8e998908f9b393630e5f8 82864 Found-by: Tinderbox 82865 82866 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 82867 CC: Alan Coopersmith <alan.coopersmith@oracle.com> 82868 Signed-off-by: Keith Packard <keithp@keithp.com> 82869 82870commit 873a1ace3646994adf95961f48719e95dcade7a2 82871Author: Paulo Zanoni <paulo.r.zanoni@intel.com> 82872Date: Tue Nov 1 10:57:56 2011 -0200 82873 82874 parser: free val.str after xstrtokenize 82875 82876 After we tokenize val.str, we discard it. 82877 82878 This is just one example: 82879 6 bytes in 1 blocks are definitely lost in loss record 24 of 652 82880 at 0x4C2779D: malloc (in vgpreload_memcheck-amd64-linux.so) 82881 by 0x4D744D: xf86getToken (scan.c:400) 82882 by 0x4D75F1: xf86getSubToken (scan.c:462) 82883 by 0x4DB060: xf86parseInputClassSection (InputClass.c:145) 82884 by 0x4D664C: xf86readConfigFile (read.c:184) 82885 by 0x490556: xf86HandleConfigFile (xf86Config.c:2360) 82886 by 0x49AA77: InitOutput (xf86Init.c:365) 82887 by 0x425A7A: main (main.c:204) 82888 82889 Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 82890 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 82891 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 82892 82893commit d41987d77c903e00cca7bcf3e04ed07151e3bb4d 82894Author: Paulo Zanoni <paulo.r.zanoni@intel.com> 82895Date: Tue Nov 1 10:42:36 2011 -0200 82896 82897 parser: free val.str after xf86getBoolValue 82898 82899 After we convert the value to a boolean, we discard the string. 82900 82901 This is just one example: 82902 82903 3 bytes in 1 blocks are definitely lost in loss record 5 of 657 82904 at 0x4C2779D: malloc (vgpreload_memcheck-amd64-linux.so) 82905 by 0x4D744D: xf86getToken (scan.c:400) 82906 by 0x4D75F1: xf86getSubToken (scan.c:462) 82907 by 0x4DB3E0: xf86parseInputClassSection (InputClass.c:189) 82908 by 0x4D664C: xf86readConfigFile (read.c:184) 82909 by 0x490556: xf86HandleConfigFile (xf86Config.c:2360) 82910 by 0x49AA77: InitOutput (xf86Init.c:365) 82911 by 0x425A7A: main (main.c:204) 82912 82913 Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 82914 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 82915 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 82916 82917commit d5c7338b3eaea55177ade6fcba71a47ccd5547f5 82918Author: Paulo Zanoni <paulo.r.zanoni@intel.com> 82919Date: Mon Oct 31 17:54:03 2011 -0200 82920 82921 parser: free scandir's list 82922 82923 v2: move the free()s to the function that calls scandir 82924 82925 80 bytes in 1 blocks are definitely lost in loss record 411 of 631 82926 at 0x4C2779D: malloc (vgpreload_memcheck-amd64-linux.so) 82927 by 0x4C27927: realloc (vgpreload_memcheck-amd64-linux.so) 82928 by 0x696A80D: scandir (scandir.c:108) 82929 by 0x4D8828: OpenConfigDir (scan.c:854) 82930 by 0x4D8A43: xf86openConfigDirFiles (scan.c:952) 82931 by 0x49031F: xf86HandleConfigFile (xf86Config.c:2327) 82932 by 0x49A9E3: InitOutput (xf86Init.c:365) 82933 by 0x425A7A: main (main.c:204) 82934 82935 Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 82936 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 82937 82938commit 3d635fe84d6de53e2f74203b10e89f7851fe3fc1 82939Author: Paulo Zanoni <paulo.r.zanoni@intel.com> 82940Date: Sun Oct 30 18:04:59 2011 -0200 82941 82942 Correctly free config file names 82943 82944 We call xf86penConfigDirFiles twice, so we overwrite the configDirPath 82945 variable, losing the pointer. If we move the pointer management to the 82946 upper layer (the function callers), they will be able to call these 82947 functions as many times as they want, but they'll have to free those 82948 returned values. 82949 82950 v2: don't leak inside XWin 82951 82952 4,097 bytes in 1 blocks are definitely lost in loss record 625 of 632 82953 at 0x4C2779D: malloc (in vgpreload_memcheck-amd64-linux.so) 82954 by 0x4D7899: DoSubstitution (scan.c:615) 82955 by 0x4D87B0: OpenConfigDir (scan.c:845) 82956 by 0x4D8A2D: xf86openConfigDirFiles (scan.c:955) 82957 by 0x49031F: xf86HandleConfigFile (xf86Config.c:2327) 82958 by 0x49A9BF: InitOutput (xf86Init.c:365) 82959 by 0x425A7A: main (main.c:204) 82960 82961 Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 82962 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 82963 82964commit 0ae087e13192d9b498db782be5ba49ca91a81547 82965Author: Paulo Zanoni <paulo.r.zanoni@intel.com> 82966Date: Sat Oct 29 18:33:24 2011 -0200 82967 82968 glx: don't leak driConfigs 82969 82970 For dri, dri2 and driswrast. 82971 82972 12,968 (584 direct, 12,384 indirect) bytes in 1 blocks are definitely lost in loss record 569 of 570 82973 at 0x4C2779D: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 82974 by 0x7821E3B: driConcatConfigs (utils.c:560) 82975 by 0x7827CF2: dri_fill_in_modes (dri_screen.c:224) 82976 by 0x782831E: dri_init_screen_helper (dri_screen.c:405) 82977 by 0x7826C03: drisw_init_screen (drisw.c:266) 82978 by 0x782225F: driCreateNewScreen (drisw_util.c:69) 82979 by 0x4826E2: __glXDRIscreenProbe (glxdriswrast.c:451) 82980 by 0x4812FA: GlxExtensionInit (glxext.c:327) 82981 by 0x41FB14: InitExtensions (miinitext.c:471) 82982 by 0x568622: main (main.c:208) 82983 82984 Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 82985 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 82986 82987commit d26fae246d7c451b4d5ffe24fdb959d4bd00b107 82988Author: Paulo Zanoni <paulo.r.zanoni@intel.com> 82989Date: Tue Oct 25 14:56:35 2011 -0200 82990 82991 glx: don't leak fbconfigs 82992 82993 29,952 (208 direct, 29,744 indirect) bytes in 1 blocks are definitely lost in loss record 573 of 573 82994 at 0x4C2779D: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 82995 by 0x4829BC: createModeFromConfig (glxdricommon.c:131) 82996 by 0x482C09: glxConvertConfigs (glxdricommon.c:185) 82997 by 0x482788: __glXDRIscreenProbe (glxdriswrast.c:468) 82998 by 0x4812FA: GlxExtensionInit (glxext.c:327) 82999 by 0x41FB14: InitExtensions (miinitext.c:471) 83000 by 0x568636: main (main.c:208) 83001 83002 Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 83003 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83004 83005commit 305a8bcb2fc6f87f8d891fcb774198b6fd118ee1 83006Author: Paulo Zanoni <paulo.r.zanoni@intel.com> 83007Date: Thu Oct 20 17:39:29 2011 -0200 83008 83009 glx: fix memory leak when destroying screen 83010 83011 1,152 bytes in 1 blocks are definitely lost in loss record 536 of 575 83012 at 0x4C25E84: calloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) 83013 by 0x483820: __glXScreenInit (glxscreens.c:357) 83014 by 0x48271C: __glXDRIscreenProbe (glxdriswrast.c:469) 83015 by 0x4812BE: GlxExtensionInit (glxext.c:327) 83016 by 0x41FB14: InitExtensions (miinitext.c:471) 83017 by 0x5685AE: main (main.c:208) 83018 83019 Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 83020 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83021 83022commit f405dfffe7d5db81d398615a80acbeba7e014ada 83023Author: Jeremy Huddleston <jeremyhu@apple.com> 83024Date: Mon Nov 7 22:14:30 2011 +0000 83025 83026 dmx: Build fix for -Werror=implicit-function-declaration 83027 83028 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 83029 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83030 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83031 83032commit 632d205b309d0515b4ae0f9902059aa4b093a313 83033Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83034Date: Mon Nov 7 20:24:08 2011 -0800 83035 83036 Fix gcc -Wwrite-strings warnings in xf86Modes code 83037 83038 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83039 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83040 83041commit 09e4b78f7944234865f4648940453303af0c2663 83042Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83043Date: Mon Nov 7 19:19:47 2011 -0800 83044 83045 Fix gcc -Wwrite-strings warnings in xf86 ddx 83046 83047 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83048 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83049 83050commit 8e4556f560487f3d07812834d4785441d0a8323c 83051Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83052Date: Mon Nov 7 20:09:47 2011 -0800 83053 83054 FindModule: stop copying const char *dirname to char *dirpath 83055 83056 Not needed since 6cf844ab69926b split out the allocation/manipulation 83057 into the helper function, leaving FindModule just copying the pointer 83058 around, and causing gcc warnings and an unreachable call to free. 83059 83060 Also no longer need to store the combined strlen results in dirlen. 83061 83062 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83063 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83064 83065commit 05f589d464a961aea8d25632a390fb66a06cd186 83066Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83067Date: Fri Nov 4 23:21:34 2011 -0700 83068 83069 Fix gcc -Wwrite-strings warnings in various extensions 83070 83071 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83072 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83073 83074commit 232f1ddf3d060f3ce9d2ebd35f33b1294cac380e 83075Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83076Date: Thu Nov 3 21:21:19 2011 -0700 83077 83078 Fix gcc -Wwrite-strings warnings in XkbGetRulesDflts 83079 83080 Stop temporarily storing a pointer to a constant literal string 83081 in a char *, just to strdup it a few lines later. 83082 83083 Fixes gcc -Wwrite-strings warnings: 83084 83085 xkbInit.c: In function 'XkbGetRulesDflts': 83086 xkbInit.c:121:38: warning: assignment discards qualifiers from pointer target type 83087 xkbInit.c:123:23: warning: assignment discards qualifiers from pointer target type 83088 xkbInit.c:125:24: warning: assignment discards qualifiers from pointer target type 83089 xkbInit.c:127:25: warning: assignment discards qualifiers from pointer target type 83090 xkbInit.c:129:25: warning: assignment discards qualifiers from pointer target type 83091 83092 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83093 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83094 83095commit 0bc41d5f8d1a6192f57d9bf646cc46ed839b06b9 83096Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83097Date: Mon Nov 7 18:08:21 2011 -0800 83098 83099 Remove redundant redeclarations of functions in the same header file 83100 83101 Exposed by recent addition of -Wredundant-decls to default CWARNFLAGS 83102 83103 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83104 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83105 83106commit 285133a35eede03e37f41aeea6cbfd15a0800d98 83107Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83108Date: Mon Nov 7 15:44:26 2011 -0800 83109 83110 sun_agp: cast key to uintptr_t before casting to (int *) 83111 83112 Matches what linux_agp already does and prevents gcc from throwing up: 83113 83114 sun_agp.c: In function 'xf86DeallocateGARTMemory': 83115 sun_agp.c:236:40: error: cast to pointer from integer of different size 83116 83117 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83118 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83119 83120commit f8dd5efb673439e7e9f7d82e6f9cfb97afacc85a 83121Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83122Date: Thu Nov 3 21:18:56 2011 -0700 83123 83124 Mark XKB char * as const to clean up gcc -Wwrite-strings warnings 83125 83126 Cleans up around 120 warnings from this set 83127 83128 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83129 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83130 83131commit cccafabd56d0f8e5784ea0ad9fdc03224f952bfe 83132Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83133Date: Thu Nov 3 21:10:51 2011 -0700 83134 83135 Mark arguments to fopen/popen/system wrappers as const char * 83136 83137 Silencing more gcc -Wwrite-strings warnings 83138 83139 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83140 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83141 83142commit 3d0ece5e8467d823afb227b7a7f8b12b906b6ba9 83143Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83144Date: Thu Nov 3 20:44:43 2011 -0700 83145 83146 Reduce unnecessary string copying in xkbtext routines 83147 83148 Instead of using sprintf to copy a static string to a local buffer, 83149 just to pass it to TryCopyStr, pass the static string to TryCopyStr 83150 directly, as is already done in other parts of this code. 83151 83152 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83153 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83154 83155commit 05d8a7f7a785eff3292f0f0537bb3902930f1b5c 83156Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83157Date: Sat Oct 29 11:08:17 2011 -0700 83158 83159 Convert a bunch of sprintf to snprintf calls 83160 83161 This batch is the straightforward set - others are more complex and 83162 need more analysis to determine right size to pass. 83163 83164 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83165 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83166 83167commit b967bf2af264726042e2f6ffb9ca7d234f34b56b 83168Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83169Date: Mon Oct 31 23:01:35 2011 -0700 83170 83171 Remove xf86FormatPciBusNumber from API, inline the one place its used 83172 83173 Found no calls from current driver modules 83174 83175 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83176 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83177 83178commit 6450f6ca7ee070da3b6d70c2d3a6977f018ac421 83179Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83180Date: Thu Nov 3 22:13:05 2011 -0700 83181 83182 Move DoShowOptions to xf86Configure.c, delete xf86ShowOpts.c 83183 83184 Gets rid of duplicate static copy of optionTypeToString by putting 83185 both callers of that helper function in the same source file. 83186 83187 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83188 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 83189 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83190 83191commit 43fa1274263f76faaca995f6e498bc3179a857b8 83192Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83193Date: Mon Oct 31 22:17:35 2011 -0700 83194 83195 Remove bad code from DoShowOptions (Xorg -showopts handler) 83196 83197 When we want to print a string, it's okay to just print it. 83198 We don't need to first allocate a buffer 2 bytes bigger than the 83199 string, copy the entire string unmodified to the buffer, print the 83200 buffer, and then leak the buffer (though we AbortDDX 8 lines later, 83201 and then just in case we survived that, call exit as well, so the 83202 leak is short lived, just oh so pointless). 83203 83204 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83205 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 83206 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83207 83208commit 71a89d97332cc181becc5a5f73166e642f96c076 83209Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83210Date: Mon Oct 31 21:36:47 2011 -0700 83211 83212 Convert glx/single2.c:DoGetString() to use asprintf() 83213 83214 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83215 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83216 83217commit f3cb512dc4daaeed389bb4740e21b6e2330e01e1 83218Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83219Date: Sat Oct 29 11:19:44 2011 -0700 83220 83221 LogVMessageVerb: Fix const mismatch warning 83222 83223 "log.c", line 382: warning: assignment type mismatch: 83224 pointer to char "=" pointer to const char 83225 83226 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83227 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83228 83229commit 3d2d88029b29d6e1c53220ad275ba8ba2dedd89e 83230Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83231Date: Sat Oct 29 11:13:32 2011 -0700 83232 83233 AuthAudit: clean up string handling calls 83234 83235 The extra "out" pointer to redirect writes to the array isn't needed since 83236 the removal of LBX (commit a9ed5a87902a), and eliminating it allows more 83237 logical use of sizeof(addr) in length-checked strlcpy & snprintf calls to 83238 write to it. 83239 83240 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83241 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83242 83243commit 615f93a3d03d40924365061c6ae242240dd0ab7e 83244Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83245Date: Tue Nov 1 14:57:41 2011 -0700 83246 83247 Remove unnecessary variable rtrn in XkbKeysymText 83248 83249 Also removes even more unnecessary use of variable assignment inside 83250 function arguments. 83251 83252 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83253 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 83254 83255commit 5f285a30a1d6ffba82ebe5e08a0b68352bb51556 83256Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83257Date: Tue Nov 1 16:56:18 2011 -0700 83258 83259 Make XIGetKnownProperty take a const char * argument 83260 83261 Now that MakeAtom takes const char *, so can XIGetKnownProperty. 83262 Clears 71 warnings from gcc -Wwrite-strings of the form: 83263 devices.c:145:5: warning: passing argument 1 of 'XIGetKnownProperty' discards qualifiers from pointer target type 83264 ../include/exevents.h:128:23: note: expected 'char *' but argument is of type 'const char *' 83265 83266 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83267 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 83268 83269commit e0f3633632de609e60950aef07c82df534db7888 83270Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83271Date: Fri Oct 28 21:31:46 2011 -0700 83272 83273 Convert ProcRenderQueryFilters to use memcpy instead of strncpy 83274 83275 We just got the string length with strlen, might as well use it 83276 to copy the whole string quickly instead of checking each character 83277 a second time to see if it's 0 or not. 83278 83279 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83280 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83281 83282commit 780133f9ae7fada462714b47e79d26075bbd9abe 83283Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83284Date: Fri Oct 28 21:29:50 2011 -0700 83285 83286 Convert DetermineClientCmd to use strdup instead of malloc+strncpy 83287 83288 *cmdname is initialized to NULL earlier in the function, so it's 83289 okay to overwrite it with NULL if strdup fails, don't need that 83290 extra check. 83291 83292 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83293 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83294 83295commit 03ddca6f71339fad089c56484bf35c63642ae1be 83296Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83297Date: Fri Oct 28 21:25:20 2011 -0700 83298 83299 Convert dmxSetDefaultFontPath to use strdup instead of malloc+strncpy 83300 83301 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83302 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83303 83304commit d9243777c77d9b2992fbaf7f459430283837e323 83305Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83306Date: Sat Oct 29 10:32:52 2011 -0700 83307 83308 matchDriverFromFiles: use one snprintf instead of strncpy/cat series 83309 83310 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83311 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83312 83313commit 6e6d732bac3c21cb85f8e998908f9b393630e5f8 83314Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83315Date: Fri Oct 28 21:18:46 2011 -0700 83316 83317 Convert strncpy/strncat to strlcpy/strlcat 83318 83319 As long as we're carrying around a compatibility copy in os/strl*.c, 83320 might as well use them. 83321 83322 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83323 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83324 83325commit e189dbb3e57d30eb96034d4ce9544ce7a93a371e 83326Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83327Date: Fri Oct 28 20:19:44 2011 -0700 83328 83329 Convert AllocXTestDevice to use asprintf 83330 83331 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83332 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83333 83334commit 08093c25a91c07ab8af7cece9bba738b827cfd1b 83335Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83336Date: Mon Oct 24 23:16:30 2011 -0700 83337 83338 Convert some malloc + strncpy pairs into strndup calls 83339 83340 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83341 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83342 83343commit acde97a39d35bfb03af2614c68176ad9afb71f53 83344Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83345Date: Mon Oct 24 20:28:32 2011 -0700 83346 83347 Add fallback implementation of strndup() 83348 83349 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83350 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83351 Linux test code fixed by: Keith Packard <keithp@keithp.com> 83352 83353commit ed38c2648cf7cc04c1d03f8d14375815f6cf536e 83354Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83355Date: Fri Oct 28 14:32:56 2011 -0700 83356 83357 Fix Sun compiler check that got turned around 83358 83359 Since the check is for !(compilers that support __builtin_constant_p) 83360 it needs to be !(gcc or new enough Sun cc), but was written as 83361 !(gcc or too old Sun cc). 83362 83363 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83364 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83365 83366commit 7ee7fd1f4c72b2ab6dba0413e63dd2e8b95b2112 83367Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83368Date: Mon Oct 24 20:39:24 2011 -0700 83369 83370 Remove a couple Error() instances left behind by 09dbfcb0ad7b6c8 83371 83372 Two instances found in the SIOCGIFCONF code for listing network interfaces. 83373 83374 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83375 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83376 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 83377 83378commit e47aa9475027ed6a255daefec85561c6b15789bd 83379Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83380Date: Mon Oct 24 23:40:45 2011 -0700 83381 83382 Enable memory checking during unit testing 83383 83384 If configure is generated with xorg-macros 1.16.0 or newer, and 83385 an appropriate memory checking library is found for the platform, 83386 use it when running unit tests. If not, then no harm is done. 83387 83388 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83389 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83390 83391commit 36670065bd74b870d0da7c6a69a9c0d222b21706 83392Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83393Date: Tue Oct 25 00:12:48 2011 -0700 83394 83395 Don't require ld -wrap for tests that don't need it 83396 83397 If configure is generated with xorg-macros 1.16 or later, this allows 83398 builders to --enable-unit-test and run the tests other than the xi2 83399 tests which require ld -wrap (and are still wrapped in if HAVE_LD_WRAP 83400 in tests/xi2/Makefile). If an older xorg-macros is used, the existing 83401 behaviour is preserved of requiring ld -wrap for all unit tests, but 83402 no side effects occur, so the minimum xorg-macros version is not raised. 83403 83404 If unit testing is enabled without ld -wrap, then we create a bogus 83405 "xi2-tests" script just to report that we're skipping them. 83406 83407 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83408 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83409 83410commit 0e6b88db7f2e94ccc8153a0e002d176440914e01 83411Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83412Date: Mon Aug 1 14:09:33 2011 -0700 83413 83414 Don't fallback to wsfb or fbdev on Solaris 83415 83416 We don't ship either one, so don't waste time and make confusing log 83417 entries trying to load them. 83418 83419 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83420 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 83421 83422commit 7b0f53f0a5824fcd81ef505beb3e19ac7c77961f 83423Author: Alan Coopersmith <alan.coopersmith@oracle.com> 83424Date: Wed Nov 23 10:50:13 2011 -0800 83425 83426 Fix Xdmx build on Linux to work with strlcpy changes 83427 83428 Include strlcpy.c in the libdmxconfig.a library with the other functions 83429 shared among the xdmx configuration programs. 83430 83431 Also add a #include "os.h" to the scanner.l file that now calls strlcpy 83432 to include the prototype from $(top_srcdir)/include/os.h. 83433 83434 (To be squashed into 83435 http://cgit.freedesktop.org/~alanc/xserver/commit/?id=c19f0ff5223d428f8ad2ab3c563c974c96a521ba 83436 before next PULL request to avoid breaking bisection.) 83437 83438 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83439 Signed-off-by: Keith Packard <keithp@keithp.com> 83440 83441commit 922c1d817091c7eedfb701f4d9eb4324e22ccca4 83442Author: Gaetan Nadon <memsize@videotron.ca> 83443Date: Fri Nov 11 20:44:19 2011 -0500 83444 83445 docs: spell "X Server Version" consistently in titles. Add where missing. 83446 83447 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 83448 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 83449 Signed-off-by: Keith Packard <keithp@keithp.com> 83450 83451commit 5952efa903d57ac69070f1477ce46223a2a7893c 83452Author: Keith Packard <keithp@keithp.com> 83453Date: Sun Nov 20 13:07:33 2011 -0800 83454 83455 Bump version to 1.11.99.1 (1.12 development release 1) 83456 83457 Signed-off-by: Keith Packard <keithp@keithp.com> 83458 83459commit 795fbe3e6e440f023e52c452106f4c266c13d36c 83460Author: Keith Packard <keithp@keithp.com> 83461Date: Sun Nov 20 13:33:12 2011 -0800 83462 83463 Add 'optionstr.h' to include/Makefile.am 83464 83465 This makes sure it gets distributed 83466 83467 Signed-off-by: Keith Packard <keithp@keithp.com> 83468 83469commit 08ec4da6a7617637914bee1636858474213cf39d 83470Author: Julien Cristau <jcristau@debian.org> 83471Date: Sat Nov 5 13:00:07 2011 +0100 83472 83473 Fix ShmPutImage for XYBitmap 83474 83475 We can't call CopyArea in that case because the image has depth 1, which 83476 might not match the target drawable, so we might overrun the shm 83477 segment. Commit 11817a881cb93a89788105d1e575a468f2a8d27c apparently 83478 fixed a similar bug for XYPixmap, but missed the bitmap case. 83479 83480 Fixes: http://bugs.debian.org/629611 83481 83482 Thanks to Alan Curry for diagnosing this and providing a test case. 83483 83484 Reported-and-tested-by: Alan Curry <pacman@kosh.dhis.org> 83485 Reviewed-by: Peter Harris <pharris@opentext.com> 83486 Signed-off-by: Julien Cristau <jcristau@debian.org> 83487 83488commit 328074890eeb111950e984c6f618311983600b20 83489Author: dtakahashi42 <dtakahashi42@gmail.com> 83490Date: Fri Nov 18 11:30:22 2011 -0800 83491 83492 rootless: Fix a server crash when choosing a color with the gimp color wheel 83493 83494 https://trac.macports.org/ticket/30927 83495 83496 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 83497 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83498 83499commit 58864146fbdf1820d04825838691e84784ef91bc 83500Author: Ross Burton <ross@linux.intel.com> 83501Date: Wed Sep 28 11:46:02 2011 +0100 83502 83503 edid: Add quirk for Acer Aspire One 110 83504 83505 At least one revision of the AAO reports a 190x110mm maximum size but a 83506 451x113mm mode. 83507 83508 X.Org Bug 41141 <https://bugs.freedesktop.org/show_bug.cgi?id=41141> 83509 83510 Signed-off-by: Ross Burton <ross@linux.intel.com> 83511 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 83512 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83513 83514commit 34b0e4eee911f8b09a3682a7f1b4c8598ef48b8d 83515Author: Chris Wilson <chris@chris-wilson.co.uk> 83516Date: Thu Aug 25 16:04:04 2011 +0100 83517 83518 dri2: Register the DRI2DrawableType after server regeneration 83519 83520 The Resource database is reset upon regeneration and so the dri2 module 83521 needs to re-register its RESTYPE for the drawable or else it will 83522 clobber the next unsuspecting user of the database. Fortunately, DRI2 is 83523 loaded late in the initialisation sequence and was last up until 83524 xf86-video-intel started using the Resource database to track 83525 outstanding swaps... 83526 83527 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 83528 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83529 Tested-by: Paulo Zanoni <paulo.r.zanoni@intel.com> 83530 83531commit bfa1a0dd190ed88020d60eba3bb04681c8e83a68 83532Author: Chris Wilson <chris@chris-wilson.co.uk> 83533Date: Mon Jan 24 11:17:03 2011 +0000 83534 83535 DRI2: Avoid a NULL pointer dereference 83536 83537 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=41211 83538 83539 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 83540 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83541 83542commit eeb21a133b982f71de739baf62e53c8a68f5d495 83543Author: Chris Wilson <chris@chris-wilson.co.uk> 83544Date: Mon Jan 24 11:17:03 2011 +0000 83545 83546 VidMode: prevent crash with no modes 83547 83548 Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=17431 83549 83550 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 83551 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83552 83553commit c1bb8f43b9290c2b18a9f0ac59773ff8f1eb974f 83554Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 83555Date: Tue Oct 4 12:25:26 2011 +0300 83556 83557 record: Prevent out of bounds access when recording a reply. 83558 83559 Any pad bytes in replies are written to the client from a zeroed 83560 array. However, record extension tries to incorrectly access the pad 83561 bytes from the end of reply data. 83562 83563 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 83564 Reviewed-by: Erkki Seppälä <erkki.seppala@vincit.fi> 83565 83566commit 1f5baa924a5907978f564372265d9830ac72a077 83567Author: Jeremy Huddleston <jeremyhu@apple.com> 83568Date: Sun Oct 30 20:26:42 2011 -0700 83569 83570 xfree86: Deprecate the use of xf86PciInfo.h 83571 83572 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 83573 Reviewed-by: Tormod Volden <debian.tormod@gmail.com> 83574 83575commit eb3377ffb8a7baa26c9831e56ed782d48b28fa71 83576Author: Jeremy Huddleston <jeremyhu@apple.com> 83577Date: Tue Nov 1 14:59:15 2011 -0700 83578 83579 xfree86: Fix powerpc build with -Werror=int-to-pointer-cast -Werror=pointer-to-int-cast 83580 83581 memType is a uint64_t on powerpc. Using memType only really makes 83582 sense for *physical* addresses, which can be 64-bit for 32-bit 83583 systems running on 64-bit hardware. 83584 83585 However, unmapVidMem() only deals with *virtual* addresses, which 83586 are guaranteed to fit into an uintptr_t. 83587 83588 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 83589 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 83590 83591commit a551f126cc4af7a476d9c8fd0cb309d5aa600d54 83592Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 83593Date: Fri Nov 4 23:26:25 2011 -0700 83594 83595 xfree86: Fix RandR rotation across server generations 83596 83597 245cb8e94fd1599 fixed xf86RotateDestroy() to actually run its teardown 83598 code, causing the Damage object to properly be re-allocated after a 83599 server regeneration. However the block that does that still thinks the 83600 Rotate layer BlockHandler is wrapped from the last generation, meaning 83601 the shadow pixmap is never re-allocated and the Damage object is never 83602 re-registered, causing a blank screen, and potentially a driver crash 83603 on the next teardown after the server asks it to free a 0x0 Pixmap. 83604 83605 Signed-off-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 83606 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 83607 Signed-off-by: Keith Packard <keithp@keithp.com> 83608 83609commit bfa2a1857a1efda7f171f10e855d200ca0dbcc1a 83610Merge: f0d50cc66 e7cb8f802 83611Author: Keith Packard <keithp@keithp.com> 83612Date: Mon Nov 14 09:07:06 2011 -0800 83613 83614 Merge remote-tracking branch 'whot/for-keith' 83615 83616commit e7cb8f802adf6fc22e7a930c0dfc5c1076781c96 83617Author: Peter Hutterer <peter.hutterer@who-t.net> 83618Date: Wed Nov 9 15:31:57 2011 +1000 83619 83620 Xext: remove unused variable 'n' 83621 83622 xselinux_ext.c: In function 'SELinuxSendItemsToClient': 83623 xselinux_ext.c:340:16: warning: unused variable 'n' [-Wunused-variable] 83624 83625 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 83626 Reviewed-by: Julien Cristau <jcristau@debian.org> 83627 83628commit bb4aa1f263ad38c175bfda3b7e6c325260ce3f28 83629Author: Peter Hutterer <peter.hutterer@who-t.net> 83630Date: Wed Nov 9 15:31:10 2011 +1000 83631 83632 Xext: don't swap CARD8 in SProcSELinuxQueryVersion 83633 83634 xselinux_ext.c: In function 'SELinuxSendItemsToClient': 83635 xselinux_ext.c:340:16: warning: unused variable 'n' [-Wunused-variable] 83636 xselinux_ext.c: In function 'SProcSELinuxQueryVersion': 83637 xselinux_ext.c:532:62: error: call to 'wrong_size' declared with attribute 83638 error: wrong sized variable passed to swap 83639 xselinux_ext.c:533:62: error: call to 'wrong_size' declared with attribute 83640 error: wrong sized variable passed to swap 83641 83642 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 83643 Reviewed-by: Julien Cristau <jcristau@debian.org> 83644 83645commit 35ec24cf245e5ef676e98a0bf7c77296a3f1ff63 83646Author: Peter Hutterer <peter.hutterer@who-t.net> 83647Date: Mon Aug 1 14:14:02 2011 +1000 83648 83649 input: replace remaining GetPairedDevice() with GetMaster() 83650 83651 Wherever it's obvious which device we need (keyboard or pointer), use 83652 GetMaster() instead of GetPairedDevice(). It is more reliable in actually 83653 getting the device type we want. 83654 83655 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 83656 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 83657 83658commit fc16917ad6f0d2722bfb5c5eeca955dd3bc4125a 83659Author: Peter Hutterer <peter.hutterer@who-t.net> 83660Date: Fri Apr 15 13:32:10 2011 +1000 83661 83662 include: export GetProximityEvents and QueueProximityEvents 83663 83664 This is mainly needed for consistency with GetPointerEvents and friend. 83665 No-one seems to actually need this function from outside the usual DDXs. 83666 83667 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 83668 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83669 83670commit 11840595a1be9f2df7390bcc708cc176e60d3ef2 83671Author: Peter Hutterer <peter.hutterer@who-t.net> 83672Date: Thu Nov 3 08:58:58 2011 +1000 83673 83674 dix: Don't let a driver without a ProximityClassRec post events 83675 83676 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 83677 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83678 83679commit 22715e465b415b3351b83b8279a4f44157f63199 83680Author: Peter Hutterer <peter.hutterer@who-t.net> 83681Date: Thu Oct 27 11:03:39 2011 +1000 83682 83683 Xi: allow passive keygrabs on the XIAll(Master)Devices fake devices 83684 83685 They don't have a KeyClassRec, but we must still allow passive grabs on 83686 them. 83687 83688 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 83689 Tested-by: Bastien Nocera <hadess@hadess.net> 83690 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 83691 83692commit f0d50cc6651dce3a8a3cd3fb84210aa92b139763 83693Author: Derek Buitenhuis <derek.buitenhuis@gmail.com> 83694Date: Sun Nov 6 12:20:51 2011 -0500 83695 83696 Fix vesa's VBE PanelID interpretation 83697 83698 xserver's VESA driver's VBE (Vesa BIOS Extensions) code 83699 includes a PanelID probe, which can get a monitor's native 83700 resolution. From this, using CVT formulas, it derives 83701 horizontal sync rate and a vertical refresh rate ranges. 83702 83703 It however, only derives the upper bounds of the ranges, and 83704 the lower bounds cannot de derived. By default, they are set 83705 to hardcoded constants which represent the lowest supported 83706 resolution: 640x480. The constants in vbe.c however, were 83707 not actually derived from forulas, but carried over from 83708 other code from the bad old days, and are not relevant 83709 to flat panel displays. This caused, for example, EEEPC701's 83710 panel, with a native resolution of 800x480, to end up with 83711 a upper bound of the horizontal sync rate that was lower 83712 than the hardcoded lower bound, which of course broke things. 83713 83714 These numbers have been rederived using both my own CVT tool 83715 based on xf86CVTMode(), and using the provided 'cvt' tool 83716 that comes with xserver. 83717 83718 Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com> 83719 Reviewed-by: Adam Jackson <ajax@redhat.com> 83720 Signed-off-by: Keith Packard <keithp@keithp.com> 83721 83722commit c643c2b7bf480f5c27ff8606bda087f8bff3b154 83723Author: Peter Hutterer <peter.hutterer@who-t.net> 83724Date: Tue Nov 1 12:17:50 2011 +1000 83725 83726 xfree86: duplicate name and driver from pInfo for NewInputDeviceRequest 83727 83728 xorg.conf devices had the name and driver set in the DDX's InputInfoPtr list 83729 but not in the option list for those devices. That information was lost when 83730 passing the options into NewInputDeviceRequest. NIDR then refused to start 83731 the devices. 83732 83733 Introduced in xorg-server-1.11.0-250-ge4cd24e 83734 83735 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 83736 Tested-by: James Cloos <cloos@jhcloos.com> 83737 83738commit 9cc44b955b27de33348d6a20bebc9704930ee18e 83739Author: Rui Matos <tiagomatos@gmail.com> 83740Date: Tue Nov 1 21:32:36 2011 +0000 83741 83742 randr: Make the RRConstrainCursorHarder logic the same as miPointerSetPosition 83743 83744 The constraining logic in RRConstrainCursorHarder allows the cursor to reach 83745 crtc positions of x = width and y = height while the constraining code in 83746 miPointerSetPosition only allows it to reach x = width - 1 and y = height - 1 83747 for the analogous screen case. 83748 83749 This patch makes the former's logic equivalent to the latter's which allows 83750 applications to benefit from Fitts's law. E.g. a maximized application 83751 adjacent to a crtc border wouldn't get pointer events if the user moved the 83752 pointer all the way until it's contained. 83753 83754 Signed-off-by: Rui Matos <tiagomatos@gmail.com> 83755 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 83756 Signed-off-by: Keith Packard <keithp@keithp.com> 83757 83758commit f47f6b6a53d414f28374391964c2930427d28f46 83759Author: James Simmons <jsimmons@infradead.org> 83760Date: Thu Nov 3 13:20:18 2011 +0000 83761 83762 don't map cursors in sw cursor mode 83763 83764commit d4eebe5217a89de196c1933e1fe4d37e22d59c6b 83765Author: Dave Airlie <airlied@redhat.com> 83766Date: Thu Nov 3 13:18:46 2011 +0000 83767 83768 configure: don't require xvmc. 83769 83770 Not needed in here 83771 83772commit 548c6fe044068ffba9b5306dc6b11f2ba22782a4 83773Merge: 3881b0bf1 ffe8ec86d 83774Author: Keith Packard <keithp@keithp.com> 83775Date: Wed Nov 2 21:35:31 2011 -0700 83776 83777 Merge remote-tracking branch 'jturney/master' 83778 83779commit 3881b0bf1ce11a62473650bdf6ee01a1762f6f91 83780Author: Gaetan Nadon <memsize@videotron.ca> 83781Date: Tue Nov 1 07:17:07 2011 -0400 83782 83783 mi: remove deprecated miPointerAbsoluteCursor from design doc 83784 83785 Function was removed from the code by commit f5409aa026 83786 83787 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83788 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 83789 Signed-off-by: Keith Packard <keithp@keithp.com> 83790 83791commit d91aa0e6601bdf3e2cfa57b3412ab14ac486edc4 83792Merge: 8df3a9ca5 401150d7d 83793Author: Keith Packard <keithp@keithp.com> 83794Date: Wed Nov 2 21:20:07 2011 -0700 83795 83796 Merge remote-tracking branch 'whot/two-screen-coordinates' 83797 83798commit 8df3a9ca5abbd56eb2013fa65250d21a8f18865b 83799Merge: 8329afa59 ef895484c 83800Author: Keith Packard <keithp@keithp.com> 83801Date: Wed Nov 2 21:18:16 2011 -0700 83802 83803 Merge remote-tracking branch 'koba/reviewed' 83804 83805commit ffe8ec86db2655b2c83e5ae61521898140c855ac 83806Author: Colin Harrison <colin.harrison@virgin.net> 83807Date: Sat Oct 8 15:05:22 2011 +0100 83808 83809 hw/xwin: Fix a typo in ddraw.h 83810 83811 Fix a (fortunately benign) typo in ddraw.h 83812 83813 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 83814 83815commit 48fda3c52b41157313b7fd56d9f3fa3ad4166ed7 83816Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 83817Date: Thu Feb 18 20:37:00 2010 -0600 83818 83819 hw/xwin: Revert "Fix bug #5735, Serious flaw in CygwinX clipboard" 83820 83821 This commit wreaks havoc with other programs which manage the clipboard, 83822 such as MS Office Clipboard or Win32 VNC viewers: 83823 83824 http://sourceware.org/bugzilla/show_bug.cgi?id=9910 83825 83826 This reverts commit 70ddd0f39d5118db72a1a4e473cbfb502f1ed9ec. 83827 83828 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 83829 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 83830 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 83831 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 83832 83833commit 3ead1d810b0e157078db39712e02ea6dc85216d8 83834Author: Jon TURNEY <jon.turney@dronecode.org.uk> 83835Date: Sun Jul 3 19:10:52 2011 +0100 83836 83837 hw/xwin: Stop pretending we work on NT4 83838 83839 We already link directly to some functions not available in NT4, so stop 83840 pretending we will work on NT4 and link directly to EnumDisplayMonitors() 83841 and SHGetFolderPath() 83842 83843 Also remove mentions of NT4 & Win95 from error messages 83844 83845 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 83846 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 83847 83848commit f6529a05a2f087cb57a7fe84f226fef3780728d0 83849Author: Colin Harrison <colin.harrison@virgin.net> 83850Date: Sat Oct 8 14:39:07 2011 +0100 83851 83852 Xming: Always remove temporary file used when invoking xkbcomp on Win32 83853 83854 When built for native Win32, pipe() & fork() aren't available, so we 83855 use a tempoary file and system() to invoke xkbcomp 83856 83857 Ensure the temporary file is always removed. It was only being removed 83858 on most errors, not on success :S 83859 83860 Also fix a couple of warnings which occur when built with WIN32 defined 83861 83862 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 83863 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 83864 83865commit ee1985386700288af1a8b1789dcad0d7d9c97393 83866Author: Jon TURNEY <jon.turney@dronecode.org.uk> 83867Date: Sat Apr 16 23:28:45 2011 +0100 83868 83869 hw/xwin: Link directly to TrackMouseEvent() 83870 83871 TrackMouseEvent has existed in user32 since at least NT4, so 83872 don't bother with jumping through all the ancient compatibility hoops 83873 of finding if _TrackMouseEvent() exists in comctl32 so it can check 83874 if TrackMouseEvent() exists in user32 to see if it needs to emulate 83875 it... 83876 83877 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 83878 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 83879 83880commit e2e6fab1efa6895256ac0ed4d5b054a88ad18077 83881Author: Jon TURNEY <jon.turney@dronecode.org.uk> 83882Date: Tue Mar 23 04:45:37 2010 -0500 83883 83884 Cygwin/X: Make default DPI match native DPI 83885 83886 Make the default DPI match the current Windows DPI setting. If that 83887 setting can't be retrieved, change the fallback DPI value from 75 dpi 83888 to 96 dpi. 83889 83890 Mark the application as dpiAware in the manifest, which prevents 83891 dpi virtualization for high (>96) dpi values on Vista and later. 83892 83893 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 83894 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 83895 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 83896 83897commit 8329afa59dd5ea3adf7adebdb2111a9bccbb126b 83898Author: Aaron Plattner <aplattner@nvidia.com> 83899Date: Mon Oct 31 15:01:12 2011 -0700 83900 83901 dix: Reinstate GetImage window size check 83902 83903 Commit 587c3a2d1961834558193e8e14e8e381a077a253 fixed DoGetImage to 83904 check windows against their backing drawables, rather than against the 83905 screen dimensions, to prevent reading outside the bounds of redirected 83906 windows' backing pixmaps (see bug #22804). Unfortunately, while 83907 making that change I also removed the check that the rectangle is 83908 contained within the bounds of the source window, which is a violation 83909 of the specification: 83910 83911 If the drawable is a window, the window must be viewable, and it 83912 must be the case that, if there were no inferiors or overlapping 83913 windows, the specified rectangle of the window would be fully 83914 visible on the screen *and wholly contained within the outside 83915 edges of the window* (or a Match error results). Note that the 83916 borders of the window can be included and read with this request. 83917 83918 (emphasis mine) 83919 83920 Reinstate the window dimension check, to return BadMatch if the 83921 GetImage request falls outside the bounds of the window. 83922 83923 Fixes X Test Suite test XGetImage-15: 83924 83925 400|0 15 1 11:05:41|IC Start 83926 200|0 15 11:05:41|TP Start 83927 520|0 15 00005146 1 1|VSW5TESTSUITE PURPOSE 15 83928 520|0 15 00005146 1 2|Assertion XGetImage-15.(A) 83929 520|0 15 00005146 1 3|When the drawable is a window and the window is viewable 83930 520|0 15 00005146 1 4|and it is not the case that given there were no inferiors or 83931 520|0 15 00005146 1 5|overlapping windows the specified rectangle of the window 83932 520|0 15 00005146 1 6|would be fully visible on the screen and wholly contained 83933 520|0 15 00005146 1 7|within the outside edges of the window, then a BadMatch 83934 520|0 15 00005146 1 8|error occurs. 83935 520|0 15 00005146 1 9|METH: Create window which is not fully visible on the screen. 83936 520|0 15 00005146 1 10|METH: Call XMapWindow to make sure the window is viewable. 83937 520|0 15 00005146 1 11|METH: Call XGetImage with rectangle extending beyond edge of screen. 83938 520|0 15 00005146 1 12|METH: Verify XGetImage return value is null. 83939 520|0 15 00005146 1 13|METH: Verify that BadMatch error occurred. 83940 520|0 15 00005146 1 14|METH: Create window which is fully visible on the screen. 83941 520|0 15 00005146 1 15|METH: Call XMapWindow to make sure the window is viewable. 83942 520|0 15 00005146 1 16|METH: Call XGetImage with rectangle extending beyond edge of window. 83943 520|0 15 00005146 1 17|METH: Verify XGetImage return value is null. 83944 520|0 15 00005146 1 18|METH: Verify that BadMatch error occurred. 83945 520|0 15 00005146 1 19|REPORT: Got Success, Expecting BadMatch 83946 520|0 15 00005146 1 20|REPORT: Null image not returned. 83947 220|0 15 1 11:05:41|FAIL 83948 83949 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 83950 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 83951 Signed-off-by: Keith Packard <keithp@keithp.com> 83952 83953commit 132545ff576cc69ed63f5a08127151fe550de4c3 83954Merge: d0c6732a9 d7c44a7c9 83955Author: Keith Packard <keithp@keithp.com> 83956Date: Sun Oct 30 16:57:58 2011 -0700 83957 83958 Merge remote-tracking branch 'whot/for-keith' 83959 83960commit d7c44a7c9760449bef263413ad3b20f19b1dc95a 83961Author: Peter Hutterer <peter.hutterer@who-t.net> 83962Date: Mon Oct 24 12:00:32 2011 +1000 83963 83964 dix: block signals when closing all devices 83965 83966 When closing down all devices, we manually unset master for all attached 83967 devices, but the device's sprite info still points to the master's sprite 83968 info. This leaves us a window where the master is freed already but the 83969 device isn't yet. A signal during that window causes dereference of the 83970 already freed spriteInfo in mieqEnqueue's EnqueueScreen macro. 83971 83972 Simply block signals when removing all devices. It's not like we're really 83973 worrying about high-responsive input at this stage. 83974 83975 https://bugzilla.redhat.com/show_bug.cgi?id=737031 83976 83977 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 83978 Reviewed-by: Julien Cristau <jcristau@debian.org> 83979 83980commit 820d9040f50a8440741b3aefbc069a3ad81e824e 83981Author: Servaas Vandenberghe <vdb@picaros.org> 83982Date: Wed Aug 31 07:06:49 2011 +0200 83983 83984 xfree86: fix potential buffer overflow 83985 83986 The patch below fixes a potential buffer overflow in xf86addComment(). 83987 This occurs if curlen > 0 && eol_seen == 0 && iscomment == 0 , as 83988 follows from the code: 83989 83990 char *xf86addComment(char *cur, char *add) 83991 83992 <...> 83993 83994 len = strlen(add); 83995 endnewline = add[len - 1] == '\n'; 83996 len += 1 + iscomment + (!hasnewline) + (!endnewline) + eol_seen; 83997 83998 if ((str = realloc(cur, len + curlen)) == NULL) 83999 return cur; 84000 84001 cur = str; 84002 84003 if (eol_seen || (curlen && !hasnewline)) 84004 cur[curlen++] = '\n'; 84005 if (!iscomment) 84006 cur[curlen++] = '#'; 84007 strcpy(cur + curlen, add); 84008 if (!endnewline) 84009 strcat(cur, "\n"); 84010 84011 Signed-off-by: Servaas Vandenberghe <vdb@picaros.org> 84012 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84013 84014 [whot: added buffer overflow test case] 84015 84016 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84017 84018commit 63e87b8639eb8e0b4e32e5d3a09099d31a03bbcd 84019Author: Peter Hutterer <peter.hutterer@who-t.net> 84020Date: Tue Oct 25 11:49:26 2011 +1000 84021 84022 xfree86: reduce calls to input_option_get_key/value 84023 84024 No functional changes. 84025 84026 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84027 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 84028 84029commit d0c6732a99c9a7e40752b9ba7898a01c325103fa 84030Author: Anssi Hannula <anssi.hannula@iki.fi> 84031Date: Wed Apr 20 14:34:52 2011 +0300 84032 84033 xfree86: add nouveau as the first automatic driver for NVIDIA hardware 84034 84035 Add nouveau as the first driver on linux for NVIDIA hardware when 84036 driver autoconfiguration is done, as it is more capable than nv. 84037 84038 nv is also kept in the list as it is more widely supported and because 84039 some old cards are not supported by nouveau. 84040 84041 Signed-off-by: Anssi Hannula <anssi.hannula@iki.fi> 84042 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84043 Reviewed-by: Cyril Brulebois <kibi@debian.org> 84044 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 84045 Signed-off-by: Keith Packard <keithp@keithp.com> 84046 84047commit ef895484c8a43d447d9d3ea7d4226936cc5f92cc 84048Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84049Date: Tue Oct 4 10:46:00 2011 +0600 84050 84051 bsd: alpha_video: Remove unused variables 84052 84053 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84054 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84055 84056commit fea7c7a8c070ec17ec03e4cb193834b6ca72fb72 84057Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84058Date: Mon Oct 3 23:03:31 2011 +0600 84059 84060 bsd: alpha_video: Simplify #include 84061 84062 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84063 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84064 84065commit 93a3a28f2c1e95ea7fae6cf1c2fb50b30eae9c33 84066Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84067Date: Tue Oct 4 00:45:09 2011 +0600 84068 84069 bsd: alpha_video: Function sethae() need only for FreeBSD 84070 84071 Return value sethae() is becoming void because no caller used it. Also old 84072 msb_set static checked by each caller is replaced by the p.hae static checked 84073 in sethae() when it's called. 84074 84075 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84076 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84077 84078commit 05b41e2dc60a0aefc685a33260e5d8e09580908b 84079Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84080Date: Tue Oct 4 17:20:13 2011 +0600 84081 84082 Move check definition MAP_FAILED in xf86_OSlib.h 84083 84084 Also remove odd definition MAP_FAILED. 84085 84086 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84087 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84088 84089commit af56e502f5ddf1d2d495b633734cadcb023b7c36 84090Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84091Date: Tue Oct 4 16:49:22 2011 +0600 84092 84093 Remove odd definition DEV_MEM 84094 84095 DEV_MEM defined in xf86_OSlib.h 84096 84097 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84098 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84099 84100commit 0481e9d3d1e0dbc05099184128de5cd6f13445f7 84101Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84102Date: Tue Oct 4 15:09:45 2011 +0600 84103 84104 Remove BSDi support 84105 84106 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84107 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84108 84109commit 8838a86fd325cb7f262a582d22ccdaa8a8f1a534 84110Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84111Date: Tue Oct 4 14:57:20 2011 +0600 84112 84113 Remove unused VT_SYSREQ_DEFAULT 84114 84115 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84116 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84117 84118commit 91131037a2d2e07079de5d222d8a8c95300750cc 84119Author: Alexandr Shadchin <alexandr.shadchin@gmail.com> 84120Date: Sat Oct 29 00:14:32 2011 +0600 84121 84122 Fix position _X_EXPORT 84123 84124 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84125 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84126 Signed-off-by: Keith Packard <keithp@keithp.com> 84127 84128commit 219bcec73d3554c57b87f28812ab06a66b739b19 84129Author: Alexandr Shadchin <alexandr.shadchin@gmail.com> 84130Date: Sat Oct 29 00:14:31 2011 +0600 84131 84132 bsd: Remove odd message about -sharevts 84133 84134 This is missing in commit 'xfree86: move -novtswitch & -sharevts argument 84135 handling up to common layer' 84136 84137 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 84138 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84139 Signed-off-by: Keith Packard <keithp@keithp.com> 84140 84141commit 005ab41986b0bb6a4e626aee7a7a542247f422e7 84142Author: Dave Airlie <airlied@redhat.com> 84143Date: Thu Oct 27 08:38:45 2011 +1000 84144 84145 test: fix two more failing FP3232 tests 84146 84147 And put a comment in to explain why we're testing for a frac between .3 and 84148 .6. We can't directly compare the frac since the floating/fixed point 84149 conversion loses precision. 84150 84151 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84152 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84153 84154commit 286fa9bf9b4ffb0f9dbd3f1cae46ddcb27ee9318 84155Author: Jeremy Huddleston <jeremyhu@apple.com> 84156Date: Tue Oct 25 23:53:43 2011 -0700 84157 84158 XWin: windowswm: Correct byte swapping in event notifications 84159 84160 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84161 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 84162 84163commit dc054fefc586c5b404fe0f949dbd456e5267c647 84164Author: Jeremy Huddleston <jeremyhu@apple.com> 84165Date: Tue Oct 25 23:49:33 2011 -0700 84166 84167 XQuartz: appledri: Correct byte swapping in event notifications 84168 84169 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84170 84171commit d5fee2b638fabeb8bec31a24e57d5c2a232d0893 84172Author: Jeremy Huddleston <jeremyhu@apple.com> 84173Date: Tue Oct 25 23:48:16 2011 -0700 84174 84175 XQuartz: applewm: Correct byte swapping in event notifications 84176 84177 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84178 84179commit 5701ab4a441eb113abd0851b0d71b82d12112854 84180Merge: 7d50211ab a41214bc9 84181Author: Keith Packard <keithp@keithp.com> 84182Date: Mon Oct 24 22:09:00 2011 -0700 84183 84184 Merge remote-tracking branch 'whot/for-keith' 84185 84186commit a41214bc9a0f326c6dc129e4a6382efb8b826862 84187Author: Peter Hutterer <peter.hutterer@who-t.net> 84188Date: Tue Oct 25 12:57:07 2011 +1000 84189 84190 kdrive: check for null memory, fix OOB 84191 84192 If key/value allocation failed, don't bother adding another InputOption. And 84193 make sure the memory allocated is large enough for the trailing \0 84194 84195 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84196 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 84197 84198commit ffe20acedb3cdc4811eb52f8fc540ba6af7339fa 84199Author: Peter Hutterer <peter.hutterer@who-t.net> 84200Date: Thu Oct 20 16:42:20 2011 +1000 84201 84202 Use new FP1616/FP3232 conversion functions 84203 84204 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84205 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 84206 84207commit 7500d841f4c709ae0edc6420332096d3b9ef9fcc 84208Author: Peter Hutterer <peter.hutterer@who-t.net> 84209Date: Thu Oct 20 16:49:28 2011 +1000 84210 84211 test: fix test with new double -> fp3232 conversion functions 84212 84213 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84214 Tested-by: Dave Airlie <airlied@redhat.com> 84215 Reviewed-by: Dave Airlie <airlied@redhat.com> 84216 84217commit 4bb5d8fae4f9a70f12591315f0b267a2ea826a0c 84218Author: Peter Hutterer <peter.hutterer@who-t.net> 84219Date: Wed Oct 19 13:17:51 2011 +1000 84220 84221 Xi: send DeviceChangedEvents when the scroll valuators change value 84222 84223 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84224 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84225 84226commit 1f4af6c12fb5d4c19f4eac3df768517c9132cc88 84227Author: Peter Hutterer <peter.hutterer@who-t.net> 84228Date: Wed Oct 19 12:28:37 2011 +1000 84229 84230 dix: accept a NULL master for CreateClassesChangedEvent 84231 84232 If a floating device changes, the master is NULL but we must still create a 84233 DCE for it. 84234 84235 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84236 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84237 84238commit d77dec6971e4a0b306c8dbd5adf627908d7972cb 84239Author: Peter Hutterer <peter.hutterer@who-t.net> 84240Date: Wed Oct 19 12:11:16 2011 +1000 84241 84242 Xi: ensure the deviceid for DeviceChangedEvents is always the right one 84243 84244 If we're sending the event for a given device, make sure the deviceid is 84245 that of the device. 84246 This allows callers to use the same DCE for slave and master without having 84247 to fiddle the DCE's internal fields. 84248 84249 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84250 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84251 84252commit e9dee21fa3213bfe87b2b728a38eb41d3ba0e664 84253Author: Peter Hutterer <peter.hutterer@who-t.net> 84254Date: Tue Oct 18 17:47:59 2011 +1000 84255 84256 dix: pass the flags into the CreateClassesChangedEvent 84257 84258 No effective functional changes, prep work for future patches. 84259 84260 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84261 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84262 84263commit fcdd2587a17437b643b4592aa7f65d11c05a4cd8 84264Author: Peter Hutterer <peter.hutterer@who-t.net> 84265Date: Tue Oct 18 17:41:29 2011 +1000 84266 84267 dix: drop unused argument from XISendDeviceChangedEvent 84268 84269 Instead of device and master (and just using master), drop the master 84270 argument and let the callers pass in the device the event is to be sent for. 84271 84272 No effective functional changes. 84273 84274 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84275 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84276 84277commit 8473e441b0f832775153281bc3df5e2d4feb2b36 84278Author: Peter Hutterer <peter.hutterer@who-t.net> 84279Date: Tue Oct 18 17:11:27 2011 +1000 84280 84281 dix: add ScrollInfo to DeviceChangedEvents 84282 84283 3304bbff9b4ed63f1a47410a5320a136420ba2c6 added smooth scrolling support for 84284 pointer events and for XIQueryDevice but didn't add the matching parts to 84285 XIDeviceChangedEvents. 84286 84287 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84288 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84289 84290commit e3f6a76dd480717eae4b17ad8e2ff707de2ffe4c 84291Author: Andreas Wettstein <wettstein509@solnet.ch> 84292Date: Thu Aug 11 16:33:33 2011 +1000 84293 84294 xkb: Support noLock and noUnlock flags for LockMods 84295 84296 These flags are required by the XKB spec section 6.3. 84297 84298 Signed-off-by: Andreas Wettstein <wettstein509@solnet.ch> 84299 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84300 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84301 84302commit 6f33593dc087d367d899d1cb6e6e30b282d922e2 84303Author: Peter Hutterer <peter.hutterer@who-t.net> 84304Date: Fri Oct 21 09:09:55 2011 +1000 84305 84306 xfree86 doc: replace driver "keyboard" with "kbd" 84307 84308 We've deprecated keyboard a long time ago 84309 84310 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84311 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84312 84313commit e4cd24e717ef1059804b3f6bb483810b708cd56a 84314Author: Peter Hutterer <peter.hutterer@who-t.net> 84315Date: Fri Aug 5 14:48:02 2011 +1000 84316 84317 xfree86: use NewInputDeviceRequest for xorg.conf devices too 84318 84319 Only use one init path for input devices - through NIDR. 84320 84321 This requires that inp_driver and inp_identifier from the 84322 XF86ConfInputRec are copied over into the options for NIDR to see them. 84323 84324 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84325 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84326 84327commit 1ecc427a39d41e723912492b846512fd0ad9af2d 84328Author: Peter Hutterer <peter.hutterer@who-t.net> 84329Date: Wed Aug 10 14:37:41 2011 +1000 84330 84331 dix: move #if statement to stop compiler warning 84332 84333 ptrveloc.c: In function 'QueryTrackers': 84334 ptrveloc.c:598:34: warning: variable 'used_offset' set but not used 84335 [-Wunused-but-set-variable] 84336 84337 used_offset is used, but only in the debugging code. Move the #if statement 84338 to ignore that warning. 84339 84340 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84341 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84342 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84343 84344commit f9067c1dd8ce9058eb48a20dfae52bc8cf3a1e55 84345Author: Peter Hutterer <peter.hutterer@who-t.net> 84346Date: Fri Aug 5 14:05:58 2011 +1000 84347 84348 xfree86: Fix a comment, the old function doesn't exist anymore 84349 84350 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84351 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84352 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84353 84354commit aeab26e9e1751e1e3514798fa53e9bd604b0d254 84355Author: Peter Hutterer <peter.hutterer@who-t.net> 84356Date: Wed Aug 10 15:58:34 2011 +1000 84357 84358 xfree86: use xf86AddNewOption instead of xf86addNewOption 84359 84360 The former strdups for us. If the strdup fails we miss out on the 84361 CorePointer option (default on anyway) and we're likely to fall over soon 84362 anyway, so let's pretend this is the same behaviour. 84363 84364 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84365 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84366 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84367 84368commit c39c8d34282b82d73c3c69a16cf0c2816256d85b 84369Author: Peter Hutterer <peter.hutterer@who-t.net> 84370Date: Wed Aug 10 14:38:58 2011 +1000 84371 84372 input: switch InputOption to use XF86OptionRec storage. 84373 84374 Use the same struct for both InputOption and XF86OptionRec so we don't need 84375 to convert to and fro the two in the config backends. 84376 84377 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84378 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 84379 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84380 84381commit 16ac78a53c1edeae183db8672104587b306cfe13 84382Author: Peter Hutterer <peter.hutterer@who-t.net> 84383Date: Tue Oct 25 11:40:58 2011 +1000 84384 84385 kdrive: switch to new InputOption API. 84386 84387 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84388 Reviewed-by: Keith Packard <keithp@keithp.com> 84389 84390commit 7d50211ab57a35910d79fc3f67ae89aff91fa995 84391Author: Christopher Yeleighton <giecrilj@stegny.2a.pl> 84392Date: Mon Oct 24 18:47:06 2011 -0700 84393 84394 Bug 38420: Xvfb crashes in miInitVisuals() when started with depth=2 84395 84396 https://bugs.freedesktop.org/show_bug.cgi?id=38420 84397 84398 Exit with fatal error message, not segfault. 84399 84400 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84401 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 84402 Signed-off-by: Keith Packard <keithp@keithp.com> 84403 84404commit d9d3a01ffca5e2de650d098231e16205781804c5 84405Merge: 17416e88d fb55f8f79 84406Author: Keith Packard <keithp@keithp.com> 84407Date: Mon Oct 24 18:12:23 2011 -0700 84408 84409 Merge remote-tracking branch 'jeremyhu/master' 84410 84411commit 17416e88dcfcc584fe5f87580d5d2b719b3521c3 84412Author: Dave Airlie <airlied@redhat.com> 84413Date: Thu Oct 20 14:43:01 2011 +0100 84414 84415 xf86Crtc: handle no outputs with no modes harder. 84416 84417 If you started an X server with no connected outputs, we pick a default 84418 1024x768 mode, however if you then ran an xvidmode using app against that 84419 server it would segfault the server due to not finding any valid modes. 84420 84421 This was due to the no output mode set code, only adding the modes to the 84422 scrn->modes once, when something called randr 1.2 xf86SetScrnInfoModes would 84423 get called and remove all the modes and we'd end up with 0. 84424 84425 This change fixes xf86SetScrnInfoModes to always report a scrn mode of at 84426 least 1024x768, and pushes the initial configuration to just call it instead 84427 of setting up the mode itself. 84428 84429 Fixes: https://bugzilla.redhat.com/show_bug.cgi?id=746926 84430 84431 I've seen other bugs like this on other distros so it might also actually fix them. 84432 84433 Signed-off-by: Dave Airlie <airlied@redhat.com> 84434 Reviewed-by: Keith Packard <keithp@keithp.com> 84435 Signed-off-by: Keith Packard <keithp@keithp.com> 84436 84437commit fb55f8f790f69f30b0ab18648c337bde9d41f5ff 84438Author: Jeremy Huddleston <jeremyhu@apple.com> 84439Date: Mon Oct 24 16:45:36 2011 -0700 84440 84441 XQuartz: Fix the filename of our log file 84442 84443 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84444 84445commit 1c8bda798b4480bab0f0da6f531e4770c7a1f771 84446Author: Jeremy Huddleston <jeremyhu@apple.com> 84447Date: Fri Oct 21 10:27:16 2011 -0700 84448 84449 XQuartz: appledri: Allow byte swapped requests 84450 84451 Even though it's only valid when local, it is possible for a local 84452 client and the server to not match endianness, such as when running 84453 a ppc application under Rosetta. 84454 84455 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84456 84457commit 14205ade0c750191bf43fba8bd55c65dba912cf4 84458Author: Jeremy Huddleston <jeremyhu@apple.com> 84459Date: Thu Oct 20 22:34:44 2011 -0700 84460 84461 XQuartz: appledri: Fix byte swapping in replies 84462 84463 Even though it's only valid when local, it is possible for a local 84464 client and the server to not match endianness, such as when running 84465 a ppc application under Rosetta. 84466 84467 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84468 Reviewed-by: Jamey Sharp <jamey@minilop.net> 84469 84470commit 2ba0ac202ad64eb4a555715980ba538de19c9fd7 84471Author: Jeremy Huddleston <jeremyhu@apple.com> 84472Date: Thu Oct 20 22:23:07 2011 -0700 84473 84474 XQuartz: appledri: Set the correct reply length for XAppleDRICreatePixmap 84475 84476 http://xquartz.macosforge.org/trac/ticket/508 84477 84478 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84479 Reviewed-by: Jamey Sharp <jamey@minilop.net> 84480 84481commit e4f257748b51d041275a64b8be408729ea05fbb5 84482Author: Jeremy Huddleston <jeremyhu@apple.com> 84483Date: Thu Oct 20 11:35:54 2011 -0700 84484 84485 XQuartz: appledri: Change whitespace in appledristr.h to match mesa 84486 84487 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84488 Reviewed-by: Jamey Sharp <jamey@minilop.net> 84489 84490commit 83fef4235db86343477b4ec9858c6ba35e1aa7d9 84491Author: Jeremy Huddleston <jeremyhu@apple.com> 84492Date: Sun Oct 16 02:12:38 2011 -0700 84493 84494 Xnest: Match the host's keymap 84495 84496 This was a regression. 84497 84498 Introduced by: 08363c5830bdea34012dcd954b45ccfdc79a3a7e and 84499 32db27a7f867b503c2840ca7b815e96d10be9210 84500 Masked by: 1e69fd4a60147287b31e53bfc61543fb17bb82c8 84501 84502 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84503 Reviewed-by: Jamey Sharp <jamey@minilop.net> 84504 84505commit 4597ac6fe0580323860080dae2576733acc49757 84506Author: Jeremy Huddleston <jeremyhu@apple.com> 84507Date: Mon Oct 17 23:45:32 2011 -0700 84508 84509 test: Add unit test for mieq 84510 84511 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84512 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84513 84514commit ad450d76e468cd62406c4f09e71e131f7ff4555c 84515Author: Jeremy Huddleston <jeremyhu@apple.com> 84516Date: Mon Oct 17 21:16:37 2011 -0700 84517 84518 mieq: Reserve some space in EQ for release and other special events 84519 84520 The last 64 events in the event queue will be reserved for release 84521 events in order to help return the system to a cleaner state when 84522 it comes back from a soft wedge. 84523 84524 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84525 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84526 84527commit 7d153f25be678acf500236ae422b33d17e6cb7a6 84528Author: Jeremy Huddleston <jeremyhu@apple.com> 84529Date: Sat Oct 15 22:51:30 2011 -0700 84530 84531 mieq: Provide better adaptability and diagnostics during mieq overflow 84532 84533 This patch changes from a static length event queue (512) to one that 84534 starts at 128 and grows to 4096 as it overflows, logging each time it 84535 grows. 84536 84537 This change also allows for multiple backtraces to be printed when the 84538 server is wedged rather than just one. This increased sampling should 84539 help identify the true hog in cases where one backtrace might be 84540 insufficient. 84541 84542 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84543 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84544 84545commit 09eaa1f6f19f9a85a3cb74a18ac898162db451df 84546Author: Jeremy Huddleston <jeremyhu@apple.com> 84547Date: Sat Oct 30 14:55:06 2010 -0700 84548 84549 configure.ac: Add -fno-strict-aliasing to CFLAGS 84550 84551 This should force the server to have -fno-strict-aliasing even once it 84552 is removed from the warning flags. 84553 84554 See: https://bugs.freedesktop.org/show_bug.cgi?id=31238 84555 84556 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84557 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 84558 84559commit a046d649e4c7e4d28f350382dcdd293e92a59aad 84560Author: Jeremy Huddleston <jeremyhu@apple.com> 84561Date: Wed Oct 5 15:02:52 2011 -0700 84562 84563 test: Silence some debug lines from the input unit tests 84564 84565 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84566 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84567 84568commit 8db554d2d830953de3c6d0cbf58e23d53d89009f 84569Author: Jeremy Huddleston <jeremyhu@apple.com> 84570Date: Thu Oct 20 14:28:56 2011 -0700 84571 84572 Xnest: Fix DestroyNotify handler 84573 84574 This partially reverts the unwanted changes that crept into 84575 c13a48e74ec89eafa9b529a87253a631ef02f2e2 84576 84577 Reported-by: Julien Cristau <jcristau@debian.org> 84578 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84579 Reviewed-by: Julien Cristau <jcristau@debian.org> 84580 84581commit ff61592441916b83aeb778c74352bb5b26247f84 84582Merge: af3f64fb7 525d4172b 84583Author: Keith Packard <keithp@keithp.com> 84584Date: Wed Oct 19 19:14:32 2011 -0700 84585 84586 Merge remote-tracking branch 'aplattner/for-master' 84587 84588commit af3f64fb77c13180e513ee99d1fd9a1b624fd8ea 84589Merge: 15bbdc103 df0dd36de 84590Author: Keith Packard <keithp@keithp.com> 84591Date: Wed Oct 19 17:33:07 2011 -0700 84592 84593 Merge remote-tracking branch 'hramrach/pull' 84594 84595commit 15bbdc103b34b6b374815698946e6c409421a644 84596Merge: a5266dcb3 323869f32 84597Author: Keith Packard <keithp@keithp.com> 84598Date: Wed Oct 19 17:26:50 2011 -0700 84599 84600 Merge remote-tracking branch 'whot/for-keith' 84601 84602commit a5266dcb3a60587e1877f90c18552baf60b597a0 84603Author: Ville Syrjala <syrjala@sci.fi> 84604Date: Sun Oct 9 01:11:04 2011 +0300 84605 84606 composite: Update borderClip in compAllocPixmap() 84607 84608 Previously the parent constrained borderClip was copied over 84609 when compRedirectWindow() is called. That is insufficient eg. in 84610 case the window was already redirected, but not yet realized. So 84611 copy the borderClip over in compAllocPixmap() instead. 84612 84613 Example: 84614 Window 1 is below an automatically redirect window 2. Window 2 is 84615 unmapped and moved outside the extents of window 1. Window 2 is 84616 then mapped again, and MarkOverlappedWindows() uses the up to 84617 date borderSize of window 2 to mark windows, which leaves 84618 window 1 unmarked. Then exposures are calculated using the stale 84619 borderClip of window 2, which causes the window below window 2, 84620 to be exposed through an apparent hole in window 1. 84621 84622 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=22566 84623 84624 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 84625 Reviewed-by: Keith Packard <keithp@keithp.com> 84626 Signed-off-by: Keith Packard <keithp@keithp.com> 84627 84628commit e4787ec20b1b9d30fe00f17b60eb0898190bf77d 84629Author: Gaetan Nadon <memsize@videotron.ca> 84630Date: Mon Oct 17 15:08:14 2011 -0400 84631 84632 XWinrc.man: fix warning: tab character in unquoted macro argument #35054 84633 84634 Spaces or tabs do not affect the text output layout. 84635 84636 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 84637 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 84638 Signed-off-by: Keith Packard <keithp@keithp.com> 84639 84640commit 321873f8045139edd47b526a765f8b1c66f53cd4 84641Author: Gaetan Nadon <memsize@videotron.ca> 84642Date: Mon Oct 17 15:08:13 2011 -0400 84643 84644 xorg.conf.man: fix 382: warning: missing `)' (got `R') #35054 84645 84646 http://www.gnu.org/software/groff/manual/html_node/Man-usage.html 84647 84648 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84649 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 84650 Signed-off-by: Keith Packard <keithp@keithp.com> 84651 84652commit 3e145d3d67b472cb5d74cb8078b614160c4722b7 84653Author: Jesse Barnes <jbarnes@virtuousgeek.org> 84654Date: Thu Oct 13 12:01:10 2011 -0700 84655 84656 crtc: match full preferred modes if possible when choosing an initial config 84657 84658 It's fairly common to have multiple, identical monitors plugged in. In 84659 that case, it's preferable to run the monitor's preferred mode on each 84660 output, rather than just matching the width & height and end up with 84661 different timings or refresh rates. 84662 84663 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 84664 Reviewed-by: Keith Packard <keithp@keithp.com> 84665 Signed-off-by: Keith Packard <keithp@keithp.com> 84666 84667commit 525d4172b246e13b8122e059e3b22866e00561d9 84668Author: Chris Wilson <chris@chris-wilson.co.uk> 84669Date: Sun Sep 4 09:34:08 2011 -0700 84670 84671 render: export TriStrip and TriFan to the drivers 84672 84673 Rather than perform an intermediate copy and expand the strip and the 84674 fan into a triangle list (thereby tripling the number of edges that the 84675 driver needs to process), allow the backend to hook directly into the 84676 appropriate Composite function. 84677 84678 In order to extend the PictureScreen, without needlessly bumping the 84679 ABI, we move the existing copy implementations to mipict.c and assign 84680 those by default. To notify the ddx that the new entry points are 84681 available, we introduce PICTURE_SCREEN_VERSION. 84682 84683 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 84684 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 84685 84686commit 9b26e6bc8d2cdf5bac3025796855ccf05972358f 84687Author: Luc Verhaegen <libv@skynet.be> 84688Date: Tue Aug 23 15:19:59 2011 -0700 84689 84690 randr: stop clients from deleting immutable output properties 84691 84692 Immutable in randr means that clients are not able to alter the 84693 property itself, they are only allowed to alter the property value. 84694 This logically means that the property then should not be deleted 84695 by the client either. 84696 84697 Signed-off-by: Luc Verhaegen <libv@skynet.be> 84698 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 84699 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 84700 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 84701 84702commit c8413362049cee8c30e0a9d67f78f9ebefe8e71f 84703Merge: 64d2d1bef b67581cf8 84704Author: Keith Packard <keithp@keithp.com> 84705Date: Tue Oct 18 07:45:24 2011 -0700 84706 84707 Merge remote-tracking branch 'herrb/master' 84708 84709commit b67581cf825940fdf52bf2e0af4330e695d724a4 84710Author: Matthieu Herrb <matthieu.herrb@laas.fr> 84711Date: Mon Oct 17 22:27:35 2011 +0200 84712 84713 Fix CVE-2011-4029: File permission change vulnerability. 84714 84715 Use fchmod() to change permissions of the lock file instead 84716 of chmod(), thus avoid the race that can be exploited to set 84717 a symbolic link to any file or directory in the system. 84718 84719 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 84720 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84721 84722commit 6ba44b91e37622ef8c146d8f2ac92d708a18ed34 84723Author: Matthieu Herrb <matthieu.herrb@laas.fr> 84724Date: Mon Oct 17 22:26:12 2011 +0200 84725 84726 Fix CVE-2011-4028: File disclosure vulnerability. 84727 84728 use O_NOFOLLOW to open the existing lock file, so symbolic links 84729 aren't followed, thus avoid revealing if it point to an existing 84730 file. 84731 84732 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 84733 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84734 84735commit df0dd36deea0c756819825113e825059ddd19243 84736Author: Michal Suchanek <hramrach@centrum.cz> 84737Date: Sat Oct 8 14:26:24 2011 +0200 84738 84739 Do not uselessly reload modules in DuplicateModule 84740 84741 The function does not initialize the module so it has no business 84742 loading it. If some user of DuplicateModule expects a module actually 84743 loaded they should use LoadModule. 84744 84745 Signed-off-by: Michal Suchanek <hramrach@centrum.cz> 84746 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84747 84748commit 24d435163eb5fbd9b73cd8ba13a9b3cdbbe8a1df 84749Author: Michal Suchanek <hramrach@centrum.cz> 84750Date: Sat Oct 8 14:19:34 2011 +0200 84751 84752 Use UnloadModuleOrDriver for UnloadSubModule. 84753 84754 Signed-off-by: Michal Suchanek <hramrach@centrum.cz> 84755 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84756 84757commit 0d4bb5442ceb8e8e4a8de6cfc4203cae469eee72 84758Author: Michal Suchanek <hramrach@centrum.cz> 84759Date: Sat Oct 8 14:13:33 2011 +0200 84760 84761 Unload submodules. 84762 84763 Signed-off-by: Michal Suchanek <hramrach@centrum.cz> 84764 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84765 84766commit b04aff76ac2eb461c71b85525a00e25efb8bf267 84767Author: Michal Suchanek <hramrach@centrum.cz> 84768Date: Thu Oct 13 17:14:53 2011 +0200 84769 84770 Document -background none option 84771 84772 Document option introduced in commit 8976e97. 84773 84774 Signed-off-by: Michal Suchanek <hramrach@centrum.cz> 84775 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 84776 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 84777 84778commit 323869f3298cbbfe864af9404a8aed1bf7995d79 84779Author: Tomáš Trnka <tomastrnka@gmx.com> 84780Date: Tue Oct 11 09:11:18 2011 +0200 84781 84782 Fix drain_console unregistration 84783 84784 Bug introduced by 9dca441670d261a9a9fb6108960ed48f3d58fb7f 84785 xfree86: add a hook to replace the new console handler. 84786 84787 console_handler was not being set, making the server eat up CPU spinning 84788 in WaitForSomething selecting consoleFd over and over again, every time 84789 trying to unregister drain_console without success due to 84790 console_handler being NULL. 84791 84792 Let's just fix the unregistration in xf86SetConsoleHandler() and use that. 84793 84794 But wait, there could be a catch: If some driver replaced the handler using 84795 xf86SetConsoleHandler(), the unregistration in xf86CloseConsole will unregister 84796 that one. I don't understand Xorg well enough to know whether this poses a 84797 problem (could mess up driver deinit somehow or something like that). As it is, 84798 xf86SetConsoleHandler() doesn't offer any way to prevent this (i.e. check which 84799 handler is currently registered). 84800 84801 I had been using it for two days on my machine that previously hit 100% CPU 84802 several times a day. That has now gone away without any new problems appearing. 84803 84804 Signed-off-by: Tomas Trnka <tomastrnka@gmx.com> 84805 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 84806 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84807 84808commit 2cb63180fa9b54f763e7e92e433943e3e73741f3 84809Author: Peter Hutterer <peter.hutterer@who-t.net> 84810Date: Wed Oct 5 12:08:31 2011 +1000 84811 84812 dix: mark motion events as emulated if we're scrolling from button press 84813 84814 The protocol requires that the emulated event is marked as such. So if a 84815 driver with smooth scrolling axis sends legacy button events, the motion 84816 event must be marked as emulated. 84817 84818 Pass the real type to emulate_scroll_button_events and create the events 84819 accordingly. For real button press or relase events, only that event must be 84820 generated since a release event will follow or a press event has already 84821 occured, respectively. (This fixes a bug where we'd get two release events 84822 for each legacy button event) 84823 84824 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84825 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84826 84827commit 82c60232c07f50774ccc0198950f64c9338057a5 84828Author: Peter Hutterer <peter.hutterer@who-t.net> 84829Date: Fri Sep 30 10:59:47 2011 +1000 84830 84831 dix: add valuator_mask_fetch_double() 84832 84833 Using this call simplifies callers that don't know if the mask bit is set. 84834 84835 Before: 84836 if (valuator_mask_isset(mask, valnum)) 84837 value = valuator_mask_get_double(mask, valnum)); 84838 else 84839 value = someothervalue; 84840 84841 Now: 84842 if (!valuator_mask_fetch_double(mask, valnum, &value)) 84843 value = someothervalue; 84844 84845 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 84846 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 84847 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 84848 84849commit 64d2d1bef14c7ec7156bb3aae64eff69f9f7be54 84850Merge: fb84be47d 679c84bce 84851Author: Keith Packard <keithp@keithp.com> 84852Date: Mon Oct 17 14:36:34 2011 -0700 84853 84854 Merge remote-tracking branch 'jeremyhu/master' 84855 84856commit fb84be47db7cdaff406792c08e34670e8e0cbda9 84857Merge: fae7ed62a 32b289e46 84858Author: Keith Packard <keithp@keithp.com> 84859Date: Mon Oct 17 13:50:25 2011 -0700 84860 84861 Merge remote-tracking branch 'whot/next' 84862 84863commit 679c84bce9b03c17aa35fa69058e7e46267dbc7b 84864Author: Jeremy Huddleston <jeremyhu@apple.com> 84865Date: Sat Sep 24 00:01:11 2011 -0700 84866 84867 Bump ABI_VIDEODRV_VERSION to 12 84868 84869 The ABI changed in the previous series of changes, so bump the ABI version for 84870 the next release. 84871 84872 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84873 Reviewed-by: Jamey Sharp <jamey@minilop.net> 84874 84875commit a89cdcee4ebab6bbfa9ace94e1988a94cfb67087 84876Author: Jeremy Huddleston <jeremyhu@apple.com> 84877Date: Sun Oct 9 02:59:12 2011 -0700 84878 84879 xfree86: Deprecate xf86MapVidMem and friends 84880 84881 Drivers should transition over to using libpciaccess's instead. 84882 84883 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84884 84885commit 41d903918a5721e53ed694880eab44170f9717cc 84886Author: Jeremy Huddleston <jeremyhu@apple.com> 84887Date: Wed Sep 14 13:45:18 2011 -0500 84888 84889 xorg-server.pc.in: Remove libpciaccess and pixman-1 from Requires 84890 84891 Every module building against xorg-server does not *Require* pixman nor 84892 libpciaccess. If such modules need pixman or pciaccess, they should be 84893 depending on them directly rather than inheriting a dependency from 84894 xorg-server. To do this, they should use PKG_CHECK_MODULES in configure.ac 84895 to check for pixman-1 or pciaccess and include the apporpriate _LIBS variable 84896 to the appropriate _LDFLAGS variable in Makefile.am 84897 84898 This also moves pixman-1 to Requires.private, so CPPFLAGS is set right to 84899 to satisfy include dependencies but avoid linking needlessly. 84900 84901 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84902 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 84903 84904commit 40d338e9f723c4b3d63516aa62fdcded4d8af75e 84905Author: Jeremy Huddleston <jeremyhu@apple.com> 84906Date: Wed Sep 14 10:37:38 2011 -0500 84907 84908 darwin: configure: Force some irrelevant options to off on darwin 84909 84910 This removes the need to pass the following to configure on darwin: 84911 --disable-pciaccess -disable-int10-module --disable-vbe --disable-vgahw --disable-libdrm 84912 84913 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84914 Reviewed-by: Jamey Sharp <jamey@minilop.net> 84915 84916commit 43d730c0e4b1a00b29843c0660dcc7c60da6725d 84917Author: Jeremy Huddleston <jeremyhu@apple.com> 84918Date: Wed Sep 14 15:42:19 2011 -0500 84919 84920 xfree86: Link modules with -module 84921 84922 This makes a difference on darwin (and apparently nowhere else) 84923 84924 https://www.gnu.org/s/libtool/manual/libtool.html#Modules-for-libltdl 84925 84926 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84927 Reviewed-by: Jamey Sharp <jamey@minilop.net> 84928 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 84929 84930commit f7edc00a2af83764910ff89514a046dc5ce2ff9e 84931Author: Jeremy Huddleston <jeremyhu@apple.com> 84932Date: Wed Sep 14 12:14:56 2011 -0500 84933 84934 xfree86: fbdevhw: Remove unused include of pciaccess.h 84935 84936 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84937 Reviewed-by: Jamey Sharp <jamey@minilop.net> 84938 84939commit 6817050f3196d460ba03a62da23fb46ffd5a0f97 84940Author: Tiago Vignatti <tiago.vignatti@nokia.com> 84941Date: Tue Jun 1 13:05:47 2010 +0300 84942 84943 configure: wrap PCI code with macro and set it at build time 84944 84945 --disable-pciaccess, used together with --disable-module-int10, can be used to 84946 disable all pci code inside the server. 84947 84948 Note that XSERVER_LIBPCIACCESS was previously used only in the driver side and 84949 now it defines also whether the library is used inside the server. Also, 84950 XORG_BUS_PCI automake variable is introduced to track PCI code needs. 84951 84952 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 84953 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 84954 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 84955 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 84956 84957commit 5c12399b6c3a8def2df1bdde1d5d4d8d9e31fd84 84958Author: Tiago Vignatti <tiago.vignatti@nokia.com> 84959Date: Mon May 31 19:27:07 2010 +0300 84960 84961 configure: make PCI configuration more sane 84962 84963 No semantical changes. Just moved code around, grouping PCI related stuff in a 84964 single chunk. 84965 84966 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 84967 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 84968 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 84969 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 84970 84971commit a319e9e697faa0faa241e0e9baf0bc41efdcdf8f 84972Author: Tiago Vignatti <tiago.vignatti@nokia.com> 84973Date: Mon May 31 18:50:50 2010 +0300 84974 84975 configure: change PCI function checking by a meaningful version of the library 84976 84977 People that don't want VGA arbiter active can go to the library and enable the 84978 stubs there. 84979 84980 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 84981 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 84982 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 84983 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 84984 Reviewed-by: Jamey Sharp <jamey@minilop.net> 84985 84986commit b3ca84430d98f5338b9b1bc817c3656efb71cff1 84987Author: Jeremy Huddleston <jeremyhu@apple.com> 84988Date: Tue Sep 13 15:32:16 2011 -0500 84989 84990 xfree86: Work around issue where ar may be told to make an archive with no contents 84991 84992 Automake: 84993 84994 "Be careful when selecting library components conditionally. Because building 84995 an empty library is not portable, you should ensure that any library 84996 always contains at least one object." 84997 84998 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 84999 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 85000 85001commit e8bafb9d8c7d7a7077e163ecfac6174356989bbf 85002Author: Jeremy Huddleston <jeremyhu@apple.com> 85003Date: Tue Sep 13 15:38:45 2011 -0500 85004 85005 xfree86: Add stubs for os-support to help adding new architecture support 85006 85007 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 85008 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85009 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 85010 85011commit 342f3eac8460fc48cfad1f1d7be939d671e6e1cd 85012Author: Adam Jackson <ajax@redhat.com> 85013Date: Tue May 10 19:00:02 2011 -0400 85014 85015 int10: Port internal users off xf86MapVidMem 85016 85017 This API is apparently semi-deprecated even by XFree86 standards, and 85018 there are only four drivers left using it. Let's start chopping it off. 85019 85020 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85021 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85022 Signed-off-by: Adam Jackson <ajax@redhat.com> 85023 85024commit 7757b8092474c4f4ec04f4ebd1272236416154c3 85025Author: Adam Jackson <ajax@redhat.com> 85026Date: Tue May 10 17:34:23 2011 -0400 85027 85028 pci: Remove xf86MapDomainMemory 85029 85030 This is slightly draconian, but that API is just awful. In all but 85031 one case in the callers it's used to get a map of some legacy VGA 85032 memory, and it would be cleaner for the caller to just call 85033 pci_device_map_legacy. 85034 85035 The sole exception is in the vesa driver, which uses it to avoid having 85036 to look up which device the BAR belongs to. That's similarly trivial to 85037 fix. 85038 85039 Having done that, Linux's PCI layer is now very small indeed. 85040 85041 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85042 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85043 Signed-off-by: Adam Jackson <ajax@redhat.com> 85044 85045commit 6f5041d0b8e11e761d6403f1f8f8bb2ba2a9626d 85046Author: Adam Jackson <ajax@redhat.com> 85047Date: Tue May 10 18:04:10 2011 -0400 85048 85049 int10: Use pciaccess rom fetch for !PC machines 85050 85051 ... instead of rolling our own, badly. 85052 85053 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85054 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85055 Signed-off-by: Adam Jackson <ajax@redhat.com> 85056 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85057 85058commit 492ed3e53b0ee6746bd4b597978fd866d4a89ef6 85059Author: Adam Jackson <ajax@redhat.com> 85060Date: Tue May 10 16:41:08 2011 -0400 85061 85062 pci: Deprecate the PCITAG type 85063 85064 It is kept around to help drivers through the API transition and will be 85065 removed at some point in the future. 85066 85067 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85068 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85069 Signed-off-by: Adam Jackson <ajax@redhat.com> 85070 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85071 85072commit a248fa3a33cb015942386d52b7faa8558e592a51 85073Author: Adam Jackson <ajax@redhat.com> 85074Date: Fri Jan 7 17:20:15 2011 -0500 85075 85076 xfree86: Remove unused bios_devmem.c 85077 85078 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85079 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85080 Signed-off-by: Adam Jackson <ajax@redhat.com> 85081 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85082 85083commit aa0bfb0f133481c57762012e8e30c05ffa151423 85084Author: Adam Jackson <ajax@redhat.com> 85085Date: Fri Jan 7 16:57:15 2011 -0500 85086 85087 linux: Remove ia64 domain I/O support code 85088 85089 pciaccess handles this now. 85090 85091 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85092 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85093 Signed-off-by: Adam Jackson <ajax@redhat.com> 85094 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85095 85096commit 95b6935637628ffa511f39004e90c75aa75141c4 85097Author: Adam Jackson <ajax@redhat.com> 85098Date: Fri Oct 16 16:09:56 2009 -0400 85099 85100 linux: Remove pre-2.6 PCI interface support 85101 85102 If you haven't ported 2.6 to your architecture in the intervening seven 85103 years, you can keep running older servers. 85104 85105 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85106 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85107 Signed-off-by: Adam Jackson <ajax@redhat.com> 85108 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85109 85110commit 769f583783d17ffa3398bc8529f61e0f08b35530 85111Author: Adam Jackson <ajax@redhat.com> 85112Date: Fri Oct 16 14:05:54 2009 -0400 85113 85114 bus: remove some dead struct fields 85115 85116 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85117 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85118 Signed-off-by: Adam Jackson <ajax@redhat.com> 85119 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85120 85121commit dd72b3c1f1df57215329aa5ba9e9bb9ad0a65086 85122Author: Adam Jackson <ajax@redhat.com> 85123Date: Mon Sep 20 11:42:11 2010 -0400 85124 85125 linux: Use pci_device_get_parent_bridge instead of open-coding it 85126 85127 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85128 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85129 Signed-off-by: Adam Jackson <ajax@redhat.com> 85130 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85131 85132commit 858fbbb40d7c69540cd1fb5315cebf811c6e7b3f 85133Author: Adam Jackson <ajax@redhat.com> 85134Date: Fri Sep 16 13:33:04 2011 -0400 85135 85136 pci: Port xf86MapLegacyIO to pciaccess 85137 85138 Per-domain I/O is now something drivers must manually request, and must 85139 keep track of within their own state rather than in the ScrnInfoRec. 85140 It's not really possible to split that into two steps without an 85141 additional intermediate ABI break, so don't even try. Drivers that want 85142 source compatibility should ifdef on the presence of xf86UnmapLegacyIO. 85143 85144 As a fringe benefit, domain-aware I/O is now OS-independent, relying 85145 only on support in pciaccess. Simplify OS PCI setup to reflect this. 85146 85147 The IOADDRESS type is kept around to help drivers through the API 85148 transition and will be removed at some point in the future. 85149 85150 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85151 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85152 Signed-off-by: Adam Jackson <ajax@redhat.com> 85153 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85154 85155commit 6d9efdce0d06df6b85f0681bea306c0b1e851502 85156Author: Adam Jackson <ajax@redhat.com> 85157Date: Tue Sep 20 18:12:29 2011 -0400 85158 85159 vgahw: Port to pciaccess IO space routines 85160 85161 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85162 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85163 Signed-off-by: Adam Jackson <ajax@redhat.com> 85164 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85165 85166commit 4bd6579188e718654c35f95623fd4772f9e0ef06 85167Author: Adam Jackson <ajax@redhat.com> 85168Date: Thu Sep 22 13:45:57 2011 -0400 85169 85170 vgahw: Don't default to standard (port space) access routines 85171 85172 In fact, don't default to anything; drivers must explicitly say which 85173 kind they want, and they are strongly encouraged to do MMIO if possible. 85174 This is an ABI change in that drivers that don't will crash, but drivers 85175 that are explicit will work with both old and new servers. 85176 85177 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85178 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85179 Signed-off-by: Adam Jackson <ajax@redhat.com> 85180 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85181 85182commit 30fb334d218f8c1d809f88054b9fe8f5f556bb62 85183Author: Adam Jackson <ajax@redhat.com> 85184Date: Thu Sep 22 13:44:53 2011 -0400 85185 85186 vgahw: Remove IO domain setup 85187 85188 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85189 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85190 Signed-off-by: Adam Jackson <ajax@redhat.com> 85191 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85192 85193commit 51a5558beb71bddeff9352ef4f43269a8a22317d 85194Author: Adam Jackson <ajax@redhat.com> 85195Date: Sat Sep 18 08:24:19 2010 -0400 85196 85197 int10: Port to pciaccess' legacy IO API 85198 85199 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85200 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85201 Signed-off-by: Adam Jackson <ajax@redhat.com> 85202 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85203 85204commit c0b63ff88a6dc20f86141aa1a1ed5a9384aa63aa 85205Author: Adam Jackson <ajax@redhat.com> 85206Date: Sat Sep 18 08:05:11 2010 -0400 85207 85208 xfree86: Move xf86GetClocks to vgahw 85209 85210 This is really a vga-specific hack anyway. The only modern driver that 85211 uses it is trident, but it's already loaded vgahw by the time it would 85212 call xf86GetClocks. 85213 85214 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85215 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 85216 Signed-off-by: Adam Jackson <ajax@redhat.com> 85217 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85218 85219commit 1433103a49b97e356da0f20aa65046cdb2f5ece6 85220Author: Dave Airlie <airlied@redhat.com> 85221Date: Wed Oct 12 09:59:38 2011 +0100 85222 85223 test: fix input test 85224 85225 The test was memsetting the wrong thing, this fixes make check in my tinderbox. 85226 85227 Signed-off-by: Dave Airlie <airlied@redhat.com> 85228 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85229 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85230 85231commit f9c6903d4a90b59c328f4fa05d2be9e0ce1c5189 85232Author: Jeremy Huddleston <jeremyhu@apple.com> 85233Date: Wed Oct 5 15:02:52 2011 -0700 85234 85235 dix: add utility functions for double to/fro FP1616/FP3232 conversion 85236 85237 Co-authored-by: Jeremy Huddleston <jeremyhu@apple.com> 85238 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85239 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 85240 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85241 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 85242 85243commit 401150d7dcad08be7c1f07e076f810cd61e2105c 85244Author: Peter Hutterer <peter.hutterer@who-t.net> 85245Date: Tue Oct 4 11:41:17 2011 +1000 85246 85247 input: change pointer screen crossing behaviour for multiple ScreenRecs 85248 85249 miPointerSetPosition traditionally took coordinates on a per-screen basis, 85250 triggering a screen switch when these went out-of-bounds. For absolute 85251 devices, this prevented screen crossing in the negative x/y direction. 85252 85253 This patch changes the event generation patch to handle screen coordinates 85254 in a desktop range (i.e. all screens together). Screen switches are 85255 triggered when these coordinates are not on the current screen. 85256 85257 This unifies the pointer behaviour of single ScreenRec multihead and 85258 multiple ScreenRecs multihead in that the cursor by default moves about the 85259 whole screen rather than be confined to one single screen. The 85260 transformation matrix may then be used to actually confine the cursor to the 85261 screen again. 85262 85263 Note: fill_pointer_events has to deal with several different coordinate 85264 systems. Make sure you read the comment before trying to understand the code. 85265 85266 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85267 85268commit 6bd0eff40fae1e5d8fed28751851eb5b932d131b 85269Author: Peter Hutterer <peter.hutterer@who-t.net> 85270Date: Wed Oct 12 15:53:57 2011 +1000 85271 85272 dix: extend rescaleValuatorAxis to take a minimum default 85273 85274 Allow rescaling to non-zero based axis ranges as default (for when screen 85275 offsets are non-zero). Currently unused. 85276 85277 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85278 85279commit 8bebb4b4896d8b6ba3309b5b28fce883bb9f8a96 85280Author: Peter Hutterer <peter.hutterer@who-t.net> 85281Date: Wed Sep 7 14:04:10 2011 +1000 85282 85283 Store desktop dimensions in screenInfo. 85284 85285 For Zaphod mode screen crossing handling we need to know the size of all 85286 screens together (i.e. the whole desktop size). Store that in the screenInfo to 85287 have it readily available in events. 85288 85289 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85290 85291commit 9cbfa4739a51e4cc6b8094833928bf8678f63876 85292Author: Peter Hutterer <peter.hutterer@who-t.net> 85293Date: Tue Sep 6 15:55:19 2011 +1000 85294 85295 xfree86: remove xf86XInputSetScreen 85296 85297 Keeping track of which screen the pointer within the input driver is 85298 obsolete now. To bind to a screen, use the transformation matrix instead. 85299 85300 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85301 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85302 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85303 85304commit 09dbfcb0ad7b6c8bac94502f2801e82f2a2ef435 85305Author: Jeremy Huddleston <jeremyhu@apple.com> 85306Date: Tue Oct 11 17:37:44 2011 -0700 85307 85308 os: Remove Error() 85309 85310 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 85311 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 85312 85313commit db30615bcb3b872475e7d40eeee8cdda5b723078 85314Author: Jeremy Huddleston <jeremyhu@apple.com> 85315Date: Mon Oct 10 12:16:31 2011 -0700 85316 85317 Xephyr: Remove socket and its lock file on exit 85318 85319 https://bugs.freedesktop.org/show_bug.cgi?id=11484 85320 85321 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 85322 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 85323 85324commit c13a48e74ec89eafa9b529a87253a631ef02f2e2 85325Author: Jeremy Huddleston <jeremyhu@apple.com> 85326Date: Mon Oct 10 12:09:39 2011 -0700 85327 85328 Xnest: Remove socket and its lock file on exit 85329 85330 https://bugs.freedesktop.org/show_bug.cgi?id=11484 85331 85332 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 85333 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 85334 85335commit 32b289e46cc2d5ec32ff0c4ba5bbfbf602afb388 85336Author: Peter Hutterer <peter.hutterer@who-t.net> 85337Date: Mon Oct 3 13:58:01 2011 +1000 85338 85339 dix: move MD last.valuator update into fill_pointer_events 85340 85341 Don't update the MD where it's not expected, positionSprite should really 85342 just do that - position the sprite. 85343 85344 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85345 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85346 85347commit 3b36fd1b49030ead44358945f62e5abe7f4609ce 85348Author: Peter Hutterer <peter.hutterer@who-t.net> 85349Date: Mon Oct 3 13:10:53 2011 +1000 85350 85351 mi: switch miPointerSetPosition to take doubles 85352 85353 Don't switch between doubles and ints in the caller, instead take doubles in 85354 miPointerSetPosition and do the conversion there. For full feature we should 85355 change everything down from here for doubles too. 85356 85357 Functional change: previously we'd restore the remainder regardless of 85358 screen switching/confinement (despite what the comment said). Now, 85359 screen changing or cursor constraints will cause the remainder be clipped 85360 off. This should happen for cursor constraints but arguably not for screen 85361 crossing. 85362 85363 This also corrects a currently wrong comment about miPointerSetPosition's 85364 input coordinates. 85365 85366 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85367 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85368 85369commit 81cfe44b1ed0de84ad1941fe2ca74bebef3fc58d 85370Author: Peter Hutterer <peter.hutterer@who-t.net> 85371Date: Mon Oct 3 12:49:49 2011 +1000 85372 85373 mi: return the screen from miPointerSetPosition 85374 85375 miPointerSetPosition may switch screens. Always return the screen the sprite 85376 is on instead of relying on callers to call miPointerGetScreen(). 85377 85378 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85379 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85380 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85381 85382commit 88dfe5366d9855e0ebf8bbff74967b793ede57d1 85383Author: Peter Hutterer <peter.hutterer@who-t.net> 85384Date: Mon Oct 3 12:37:28 2011 +1000 85385 85386 dix: drop screen argument from positionSprite 85387 85388 We can just get this in the function, no effective functional changes. 85389 85390 Also return the screen to the caller. Though we don't use it yet, we will in 85391 a follow-up patch. 85392 85393 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85394 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85395 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85396 85397commit 967bc25da221a69c8fc390253465145ce534fcb9 85398Author: Peter Hutterer <peter.hutterer@who-t.net> 85399Date: Mon Oct 3 11:42:08 2011 +1000 85400 85401 dix: move screen- to device coordinate scaling to separate function 85402 85403 No functional changes. 85404 85405 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85406 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85407 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85408 85409commit b059e06e19ac9417ceeb8be58c1c91b159291865 85410Author: Peter Hutterer <peter.hutterer@who-t.net> 85411Date: Mon Aug 29 12:36:26 2011 +1000 85412 85413 dix: don't allow keyboard devices to submit motion or button events. 85414 85415 GPE unconditionally dereferences pDev->valuator if a mask is present. This 85416 shouldn't really happen but if it does, don't crash, just ignore the events 85417 with an error. 85418 85419 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85420 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85421 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85422 85423commit bccff533184a051b614a26304ce77ad30bede5e0 85424Author: Peter Hutterer <peter.hutterer@who-t.net> 85425Date: Mon Oct 3 12:19:21 2011 +1000 85426 85427 dix: moveRelative modifies parameter in-place, say so. 85428 85429 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85430 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85431 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85432 85433commit b966362ccf0fe6fdd44f4d778d47e3677f55f11b 85434Author: Peter Hutterer <peter.hutterer@who-t.net> 85435Date: Mon Oct 3 12:18:20 2011 +1000 85436 85437 dix: rename moveAbsolute to clipAbsolute 85438 85439 Let's be honest about what it does. 85440 85441 moveRelative accumulates delta _and_ clips in some cases, so that one can 85442 keep it's name. 85443 85444 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85445 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85446 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85447 85448commit 959d18c3765e447897a8cfd358e9ee645df595d9 85449Author: Peter Hutterer <peter.hutterer@who-t.net> 85450Date: Fri Sep 30 10:50:51 2011 +1000 85451 85452 dix: fix missing verb in comment 85453 85454 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85455 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85456 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85457 85458commit 7074ec87bdf81699df172619aea7aae1ad4ec3c6 85459Author: Peter Hutterer <peter.hutterer@who-t.net> 85460Date: Fri Sep 30 10:47:00 2011 +1000 85461 85462 dix: document transformAbsolute 85463 85464 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85465 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85466 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85467 85468commit f4ca19ce3ab91a9c8ad9de60f7dc95466f21f589 85469Author: Peter Hutterer <peter.hutterer@who-t.net> 85470Date: Fri Jul 29 10:56:44 2011 +1000 85471 85472 dix: copy the source ID into the RawEvent (#34240) 85473 85474 X.Org Bug 34240 <http://bugs.freedesktop.org/show_bug.cgi?id=34240> 85475 85476 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85477 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85478 85479commit 8b29addc30163bb7ba74c74351f312aac02f049a 85480Author: Jeremy Huddleston <jeremyhu@apple.com> 85481Date: Sun Oct 9 04:00:41 2011 -0700 85482 85483 loader: when creating sdksyms.c only include shmint.h if MITSHM is enabled #29109 85484 85485 https://bugs.freedesktop.org/show_bug.cgi?id=29109 85486 85487 When configured with --disable-mitshm the symbols declared in shmint.h 85488 do not exist. By guarding the include with '#ifdef MITSHM' these 85489 symbols are skipped when generating sdksyms.c with --disable-mitshm. 85490 85491 Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de> 85492 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85493 85494commit eabd5532fb589bd0550344ebd23c9765e591aadf 85495Author: Julien Cristau <jcristau@debian.org> 85496Date: Sun Oct 9 17:06:14 2011 -0700 85497 85498 xfree86: fix build with xv disabled 85499 85500 https://bugs.freedesktop.org/show_bug.cgi?id=29111 85501 85502 Signed-off-by: Julien Cristau <jcristau@debian.org> 85503 Tested-by: Cyril Brulebois <kibi@debian.org> 85504 Reviewed-by: Cyril Brulebois <kibi@debian.org> 85505 85506commit c53380be804852045722203c993379813d97f6d7 85507Author: Ville Skyttä <ville.skytta@iki.fi> 85508Date: Sun Mar 6 10:18:30 2011 +0200 85509 85510 Man page syntax and spelling fixes. 85511 85512 Signed-off-by: Ville Skyttä <ville.skytta@iki.fi> 85513 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85514 85515commit 765ef69295ddc473640c96f1b4f54e0b8bfc670e 85516Author: Max Schwarz <max.schwarz@online.de> 85517Date: Tue Oct 4 22:06:08 2011 +0200 85518 85519 dix: fix inverted handling of legacy scroll button events 85520 85521 This bug led to inverted scrolling axes with drivers that support smooth 85522 scrolling axes but send legacy button events. 85523 85524 Signed-off-by: Max Schwarz <Max@x-quadraht.de> 85525 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 85526 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85527 85528commit fae7ed62ad476114bd37b566202cf6b6546c0a1f 85529Author: Jeremy Huddleston <jeremyhu@apple.com> 85530Date: Tue Oct 4 21:40:03 2011 -0700 85531 85532 XQuartz: pbproxy: Add missing AM_OBJCFLAGS 85533 85534 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 85535 85536commit e4cddf509e1729e8ff40354275b65455111ad2bd 85537Author: vdb@picaros.org <vdb@picaros.org> 85538Date: Sat Sep 17 18:55:47 2011 +0200 85539 85540 Fix a rare memory leak 85541 85542 Signed-off-by: Servaas Vandenberghe <vdb@picaros.org> 85543 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85544 85545commit 53611213396479abfdce0f7752681572e0d26d78 85546Author: Kirill Elagin <kirelagin@gmail.com> 85547Date: Tue Oct 4 23:02:20 2011 +0400 85548 85549 Fix server crash due to invalid images 85550 85551 See https://bugs.freedesktop.org/show_bug.cgi?id=39383 85552 85553 Signed-off-by: Kirill Elagin <kirelagin@gmail.com> 85554 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85555 85556commit 7a33c5b934382b5554f41ab1502514e91c9bc52e 85557Author: Alan Hourihane <alanh@vmware.com> 85558Date: Tue Oct 4 19:42:46 2011 -0700 85559 85560 dixfonts: Don't overwrite local c variable until new_closure is safely initialized. 85561 85562 Signed-off-by: Alan Hourihane <alanh@vmware.com> 85563 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85564 85565commit 911e7368bf9c00d327e994a6f7a1d8d8f9b83c72 85566Author: Peter Hutterer <peter.hutterer@who-t.net> 85567Date: Tue Aug 30 16:58:04 2011 -0400 85568 85569 Move pointOnScreen to inpututils.c 85570 85571 We need this from other files too. 85572 85573 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85574 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85575 85576commit d0a7cd759d4741a1ae118d579c90704410cde244 85577Author: Peter Hutterer <peter.hutterer@who-t.net> 85578Date: Fri Sep 2 09:53:02 2011 +1000 85579 85580 dix: NewCurrentScreen must work on pointers where possible 85581 85582 When a screen switch is triggered by PointerKeys, the device for 85583 NewCurrentScreen is the keyboard. Submitting pointer events for this 85584 keyboard (without valuators) has no effect as GPE ignores the event. 85585 85586 Force the dequeuing through the XTest device attached to this device. 85587 85588 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85589 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85590 85591commit 535b3789be3a7b43b5d9026e2b5150521d91e32b 85592Author: Peter Hutterer <peter.hutterer@who-t.net> 85593Date: Wed Aug 31 14:15:02 2011 +1000 85594 85595 dix: warn about keyboard events with valuator masks 85596 85597 We don't actually handle the mask correctly. They're clipped and dropped 85598 into the event but that's about it. I don't think we did since 1.4, let's 85599 warn the user if this happens. 85600 85601 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85602 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85603 85604commit 9537afb13f2750d22350b7441570332ae60e4860 85605Author: Peter Hutterer <peter.hutterer@who-t.net> 85606Date: Tue Aug 30 13:37:31 2011 +1000 85607 85608 dix: fill out root_x/y for keyboard events 85609 85610 Switching screens relies on rootx/y to be set to the correct value. Note: 85611 though we technically take a mask for GetKeyboardEvents we don't actually 85612 handle it properly to move the pointer as required (and generate motion 85613 events if needed). 85614 85615 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85616 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85617 85618commit 6378d0233d21088b6429755627b4253859892c72 85619Merge: 6e965d8a1 98b230669 85620Author: Keith Packard <keithp@keithp.com> 85621Date: Mon Oct 3 13:56:06 2011 -0700 85622 85623 Merge remote-tracking branch 'herrb/master' 85624 85625commit 6e965d8a185087a55b2100b817e18f05b3ce5a00 85626Merge: 466e4b3f1 ac5881d6d 85627Author: Keith Packard <keithp@keithp.com> 85628Date: Mon Oct 3 13:47:49 2011 -0700 85629 85630 Merge remote-tracking branch 'koba/reviewed' 85631 85632commit 466e4b3f174bdd7a69a76eb06e6c3e3344b8adc2 85633Author: Gaetan Nadon <memsize@videotron.ca> 85634Date: Fri Sep 30 07:21:57 2011 -0400 85635 85636 Xserver-spec: use appropriate copyright markup 85637 85638 Use docbook copyright markup for year and holder. 85639 85640 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 85641 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 85642 Signed-off-by: Keith Packard <keithp@keithp.com> 85643 85644commit 62940bf38f70e3d9fb80521de40d6361548a0ea5 85645Author: Gaetan Nadon <memsize@videotron.ca> 85646Date: Fri Sep 30 07:21:56 2011 -0400 85647 85648 Xserver-spec: re-order markup to follow outpout order 85649 85650 No content change, no layout/ordering change. 85651 It is easier to find <title> at the top. 85652 85653 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 85654 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 85655 Signed-off-by: Keith Packard <keithp@keithp.com> 85656 85657commit 365bc2f5b132ac76a1f827923a99a75b33b158ac 85658Author: Gaetan Nadon <memsize@videotron.ca> 85659Date: Fri Sep 30 07:21:55 2011 -0400 85660 85661 Xserver-spec: add release information 85662 85663 It has been added in all docs (X Version 11, Release 7.6) 85664 85665 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 85666 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 85667 Signed-off-by: Keith Packard <keithp@keithp.com> 85668 85669commit ee3e26079e120bf9914cc27871d50501b35e75df 85670Author: Gaetan Nadon <memsize@videotron.ca> 85671Date: Fri Sep 30 07:16:34 2011 -0400 85672 85673 ddxDesign: drop the url in the coporate authors list 85674 85675 It was such an eyesore once rendered in html. 85676 Now it looks like other authors. 85677 85678 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 85679 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 85680 Signed-off-by: Keith Packard <keithp@keithp.com> 85681 85682commit 0ed71b48c0c7bd65cd05eda6c06b9c29b4e3f607 85683Author: Gaetan Nadon <memsize@videotron.ca> 85684Date: Fri Sep 30 07:16:33 2011 -0400 85685 85686 ddxDesign: remove server version from the document title 85687 85688 Such version information is already written in the appropriate location 85689 85690 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 85691 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 85692 Signed-off-by: Keith Packard <keithp@keithp.com> 85693 85694commit 83ec3bf1aa72848421660c44c4503fec655f5924 85695Author: Gaetan Nadon <memsize@videotron.ca> 85696Date: Fri Sep 30 07:16:32 2011 -0400 85697 85698 ddx-Design: add X Version and Release information 85699 85700 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 85701 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 85702 Signed-off-by: Keith Packard <keithp@keithp.com> 85703 85704commit 2f09f6e3f610e585aeb5bd1023323a909d1ba20d 85705Author: Gaetan Nadon <memsize@videotron.ca> 85706Date: Thu Sep 29 07:53:16 2011 -0400 85707 85708 Xinput: use appropriate copyright markup for a multi licensed doc 85709 85710 Removed a duplicate paragraph. 85711 85712 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 85713 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 85714 Signed-off-by: Keith Packard <keithp@keithp.com> 85715 85716commit 597348e969bf78dc2de4cf9bae7b76c110144266 85717Author: Gaetan Nadon <memsize@videotron.ca> 85718Date: Thu Sep 29 07:53:15 2011 -0400 85719 85720 Xinput: replace hard coded X11 version with X server version 85721 85722 Which is more useful. 85723 85724 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 85725 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 85726 Signed-off-by: Keith Packard <keithp@keithp.com> 85727 85728commit fad04dd9701aa76249a0bab31676103debb84307 85729Author: Lennart Poettering <lennart@poettering.net> 85730Date: Tue Sep 27 18:04:06 2011 +0100 85731 85732 udev: make use of udev_enumerate_add_match_tag() only when it is available 85733 85734 udev_enumerate_add_match_tag() and udev_monitor_filter_add_match_tag() 85735 are mostly optimizations, hence simply skip these calls if they are not 85736 available in the installed version of libudev. 85737 85738 This should fix the build on older versions of udev. 85739 85740 [airlied: fixes tinderbox failures on RHEL6] 85741 85742 Signed-off-by: Lennart Poettering <lennart@poettering.net> 85743 Reviewed-by: Dave Airlie <airlied@redhat.com> 85744 Tested-by: Dave Airlie <airlied@redhat.com> 85745 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 85746 Signed-off-by: Keith Packard <keithp@keithp.com> 85747 85748commit 77743f877d573cdf70fcb5cb5897250f668ebf9b 85749Author: Jamey Sharp <jamey@minilop.net> 85750Date: Mon Sep 26 22:38:44 2011 -0700 85751 85752 kdrive: Fix build for opaque InputOption structure. 85753 85754 Commit 05284a03f9002b03a66ae355b34790ec02b726f0 missed fixing up 85755 kdrive's use of the old non-opaque structure. 85756 85757 Signed-off-by: Jamey Sharp <jamey@minilop.net> 85758 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 85759 Signed-off-by: Keith Packard <keithp@keithp.com> 85760 85761commit 9a55b3661b4172f9a32cd6fefb254e06edb1a88a 85762Merge: cf11ca360 84bb0207f 85763Author: Keith Packard <keithp@keithp.com> 85764Date: Mon Oct 3 11:49:23 2011 -0700 85765 85766 Merge remote-tracking branch 'alanc/master' 85767 85768commit cf11ca360c2ded5ca309faa9d039160947387bb8 85769Merge: 463dd8706 6eae9fa28 85770Author: Keith Packard <keithp@keithp.com> 85771Date: Mon Oct 3 11:44:59 2011 -0700 85772 85773 Merge remote-tracking branch 'jeremyhu/for-keith' 85774 85775commit 463dd87062edaa974cb8c7163328a53197f237e1 85776Author: Pauli Nieminen <pauli.nieminen@linux.intel.com> 85777Date: Thu Sep 22 18:38:53 2011 +0300 85778 85779 xf86/modes: Fix shadow rotation crashing when screen pixmap changes 85780 85781 Driver may change screen pixmaps after page flipping that would then 85782 make damage lose track of the root pixmap. 85783 85784 Using root window for shadow damages fixes the problem because 85785 SetWindowPixmap is implemented in shadow code. 85786 85787 Signed-off-by: Pauli Nieminen <pauli.nieminen@linux.intel.com> 85788 Reviewed-by: Keith Packard <keithp@keithp.com> 85789 Signed-off-by: Keith Packard <keithp@keithp.com> 85790 85791commit f5d50b46ddeb039ab6564141f61261e94ee67637 85792Merge: 57cd32e93 524e5445c 85793Author: Keith Packard <keithp@keithp.com> 85794Date: Mon Oct 3 11:36:28 2011 -0700 85795 85796 Merge remote-tracking branch 'whot/next' 85797 85798commit 57cd32e93425597317b4b7722859155419836e4c 85799Author: Aaron Plattner <aplattner@nvidia.com> 85800Date: Thu Aug 25 15:41:55 2011 -0700 85801 85802 xfree86/modes: Make cursor position transform a helper function 85803 85804 When the driver can handle the crtc transform in hardware, it sets 85805 crtc->driverIsPerformingTransform, which turns off both the shadow 85806 layer and the cursor's position-transforming code. However, some 85807 drivers actually do require the cursor position to still be 85808 transformed in these cases. Move the cursor position transform into a 85809 helper function that can be called by such drivers. 85810 85811 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 85812 85813commit 245cb8e94fd15990e1b7d6622added460f104dba 85814Author: Aaron Plattner <aplattner@nvidia.com> 85815Date: Thu Aug 25 10:19:48 2011 -0700 85816 85817 xfree86/modes: Let the driver handle the transform 85818 85819 If a driver can use hardware to handle the crtc transform, then 85820 there's no need for the server's shadow layer to do it. Add a crtc 85821 flag that lets the driver indicate that it is handling the transform. 85822 If it's set, consider the transformed size of the screen but don't 85823 actually enable the shadow layer. Also stop adjusting the cursor 85824 image and position. 85825 85826 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 85827 85828commit e089737fb13868bd3a72b7ac4799d502d188f03e 85829Author: Keith Packard <keithp@keithp.com> 85830Date: Mon Oct 3 11:23:16 2011 -0700 85831 85832 Add AC_LANG_SOURCE wrappers around configure.ac code fragments 85833 85834 Current autoconf versions are very unhappy when code fragments are not 85835 wrapped in AC_LANG_SOURCE macros, generating errors like: 85836 85837 configure.ac:723: warning: AC_LANG_CONFTEST: no AC_LANG_SOURCE call detected in body 85838 ../../lib/autoconf/lang.m4:194: AC_LANG_CONFTEST is expanded from... 85839 ../../lib/autoconf/general.m4:2662: _AC_LINK_IFELSE is expanded from... 85840 ../../lib/autoconf/general.m4:2679: AC_LINK_IFELSE is expanded from... 85841 ../../lib/m4sugar/m4sh.m4:606: AS_IF is expanded from... 85842 ../../lib/autoconf/general.m4:2032: AC_CACHE_VAL is expanded from... 85843 ../../lib/autoconf/general.m4:2053: AC_CACHE_CHECK is expanded from... 85844 configure.ac:723: the top level 85845 85846 Signed-off-by: Keith Packard <keithp@keithp.com> 85847 85848commit 6eae9fa28445b6cfa09061b338242df5116aafe8 85849Author: Matt Turner <mattst88@gmail.com> 85850Date: Wed Sep 28 17:27:45 2011 -0400 85851 85852 Remove incorrect & in swap_uint32 85853 85854 Caused by commit 893e86a4, and hidden by the (char *) cast. 85855 85856 Signed-off-by: Matt Turner <mattst88@gmail.com> 85857 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 85858 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85859 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 85860 Tested-by: Aaron Plattner <aplattner@nvidia.com> 85861 85862commit 6f581e3fde7ddc943c04f274e01735beb358b7f9 85863Author: Dave Airlie <airlied@redhat.com> 85864Date: Mon Oct 3 16:09:12 2011 +0100 85865 85866 fix fb_id for dirty reporting 85867 85868commit 79959b734702486971447a11b2e4181463fda3bf 85869Author: Dave Airlie <airlied@redhat.com> 85870Date: Mon Oct 3 11:35:39 2011 +0100 85871 85872 fix make distcheck 85873 85874commit 524e5445c0b6df5247d4aac5368470bb89ef4080 85875Author: Alan Coopersmith <alan.coopersmith@oracle.com> 85876Date: Sat Oct 1 08:43:52 2011 -0700 85877 85878 Add #include "inpututils.h" to xkbAccessX.c for init_device_event 85879 85880 Fixes Sun compiler warning: 85881 "xkbAccessX.c", line 128: warning: implicit function declaration: init_device_event 85882 85883 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 85884 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85885 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85886 85887commit 98b230669f591bf5b8b8764db3b287d91857ee9e 85888Author: Matthieu Herrb <matthieu.herrb@laas.fr> 85889Date: Sun Jul 31 20:43:39 2011 +0200 85890 85891 sdksyms.sh may not be executable. 85892 85893 Use $(SHELL) to run it. Someone may want to build out of a source tree 85894 in a filesystem with the noexec mount flag set. 85895 85896 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 85897 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 85898 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 85899 85900commit f54852edc35a2831a4cd91906188101ea19904e1 85901Author: Matthieu Herrb <matthieu.herrb@laas.fr> 85902Date: Sun Sep 18 12:46:52 2011 +0200 85903 85904 Add a 'wscons' autoconf mechanism to configure input devices on BSD. 85905 85906 This does not really handle hotplug (it's handled inside the kernel, 85907 by the 'mux' devices), but uses the wscons console driver 85908 configuration to figure out the keyboard layout and the list of 85909 pointing devices found by the kernel. 85910 85911 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 85912 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 85913 85914commit f32c827d513c44f07e1d0fbcc0c96cef18c9a4d9 85915Author: Max Schwarz <Max@x-quadraht.de> 85916Date: Sun Sep 25 20:44:26 2011 +0200 85917 85918 Input: Fix frac calculation on [Raw]DeviceEvent conversion 85919 85920 (1UL << 32) evaluates to 0 (at least here), so do the 85921 fraction calculation in two steps as in libXi. Fractions on xXIRawEvent 85922 were not multiplied at all, which also gave 0 as result. 85923 85924 Signed-off-by: Max Schwarz <Max@x-quadraht.de> 85925 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85926 85927commit 4c6bc0e76599dbe5ede2e1f48c9936a0e996b638 85928Merge: 057cc92eb 3304bbff9 85929Author: Peter Hutterer <peter.hutterer@who-t.net> 85930Date: Fri Sep 30 09:24:56 2011 +1000 85931 85932 Merge branch 'smooth-scrolling' into next 85933 85934commit 3304bbff9b4ed63f1a47410a5320a136420ba2c6 85935Author: Peter Hutterer <peter.hutterer@who-t.net> 85936Date: Tue Feb 15 18:49:58 2011 +0000 85937 85938 Input: Add smooth-scrolling support to GetPointerEvents 85939 85940 For scroll wheel support, we used to send buttons 4/5 and 6/7 for 85941 horizontal/vertical positive/negative scroll events. For touchpads, we 85942 really want more fine-grained scroll values. GetPointerEvents now 85943 accepts both old-school scroll button presses, and new-style scroll axis 85944 events, while emitting both types of events to support both old and new 85945 clients. 85946 85947 This works with the new XIScrollClass to mark axes as scrolling axes. 85948 Drivers mark any valuators that send scroll events with SetScrollValuator. 85949 (Currently missing: the XIDeviceChangeEvent being sent when a driver changes 85950 a scroll axis at run-time. This can be added later.) 85951 85952 Note: the SCROLL_TYPE enums are intentionally different values to the XI2 85953 proto values to avoid copy/overlapping range bugs. 85954 85955 Co-authored-by: Daniel Stone <daniel@fooishbar.org> 85956 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 85957 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 85958 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 85959 85960commit ac5881d6d0bcf4bcc66a5d6ddde4eca950db1a5b 85961Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 85962Date: Thu Sep 15 23:05:24 2011 +0600 85963 85964 Remove unused vtSysreq 85965 85966 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 85967 Reviewed-by: Jamey Sharp <jamey@minilop.net> 85968 Tested-by: Matthieu Herrb <matthieu.herrb@laas.fr> 85969 85970commit 0be1640dbbdd072f26d937de2e49f5ed58d54ef9 85971Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 85972Date: Sat Aug 27 02:02:36 2011 +0600 85973 85974 bsd: ioctl KDENABIO/KDDISABIO do not matter for OpenBSD 85975 85976 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 85977 Tested-by: Matthieu Herrb <matthieu.herrb@laas.fr> 85978 85979commit 91042b98afdc6cecc746e101a88f79e56db1c0d0 85980Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 85981Date: Sat Aug 27 00:47:34 2011 +0600 85982 85983 bsd: OpenBSD and NetBSD not need extra headers in PCVT_SUPPORT 85984 85985 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 85986 Tested-by: Matthieu Herrb <matthieu.herrb@laas.fr> 85987 85988commit 86eaa9bbe4ed59f89d6509d5e13211ab7f5a8038 85989Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 85990Date: Sat Aug 27 01:03:28 2011 +0600 85991 85992 bsd: Some clean up 85993 85994 OpenBSD and NetBSD does not support syscons 85995 85996 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 85997 Tested-by: Matthieu Herrb <matthieu.herrb@laas.fr> 85998 85999commit dc864770d47fc1e4ec993e37a59e5aecb5b0a85e 86000Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 86001Date: Fri Aug 26 22:40:19 2011 +0600 86002 86003 bsd: Remove dead code 86004 86005 Since OsInit closes stdin before the xfree86 DDX opens the 86006 console, fstat on stdin will always fail, so it's safe to delete 86007 code that attempts it. 86008 86009 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 86010 Reviewed-by: Jamey Sharp <jamey@minilop.net> 86011 Tested-by: Matthieu Herrb <matthieu.herrb@laas.fr> 86012 86013commit cbb842666fa7ff26bbdd1e89a7f41b835a677c5e 86014Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 86015Date: Fri Aug 26 21:42:41 2011 +0600 86016 86017 bsd: Variable devConsoleFd need only if defined PCCONS_SUPPORT 86018 86019 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 86020 Reviewed-by: Jamey Sharp <jamey@minilop.net> 86021 Tested-by: Matthieu Herrb <matthieu.herrb@laas.fr> 86022 86023commit 38bd1e123d959e0a13514ec415afc8bfb5a170f2 86024Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 86025Date: Fri Aug 26 18:54:01 2011 +0600 86026 86027 bsd: Replacement screenFd on consoleFd because they are equivalent 86028 86029 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 86030 Reviewed-by: Jamey Sharp <jamey@minilop.net> 86031 Tested-by: Matthieu Herrb <matthieu.herrb@laas.fr> 86032 86033commit f015351c73bb24ce4172911c3e95d67922de42ab 86034Author: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 86035Date: Fri Aug 26 18:27:02 2011 +0600 86036 86037 bsd: Remove unused macros KBD_FD 86038 86039 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 86040 Reviewed-by: Jamey Sharp <jamey@minilop.net> 86041 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 86042 Tested-by: Matthieu Herrb <matthieu.herrb@laas.fr> 86043 86044commit cb65ec6b58daf9a1afc40c8693fa0632f8cf0585 86045Author: Dave Airlie <airlied@redhat.com> 86046Date: Thu Sep 29 16:06:52 2011 +0100 86047 86048 check drm support dumb buffer capability 86049 86050commit 6ccd13a2d7d78db2d5bd3fafb00f78d2ceb11741 86051Author: Dave Airlie <airlied@redhat.com> 86052Date: Thu Sep 29 16:00:50 2011 +0100 86053 86054 add initial man page 86055 86056commit 1f37991dec98073d5dfbf0c30213f838999bd9da 86057Author: Dave Airlie <airlied@redhat.com> 86058Date: Thu Sep 29 15:55:36 2011 +0100 86059 86060 add shadowfb support, default to on. 86061 86062 we should probably expose a bit from kernel to say if shadow is preferred 86063 or wasteful. 86064 86065commit 0b5e574dff230c7879eb9655bb7252553bbace50 86066Author: Dave Airlie <airlied@redhat.com> 86067Date: Thu Sep 29 15:55:24 2011 +0100 86068 86069 remove unused debug 86070 86071commit d686130c17126d86db55aae6b65b12d8bcf6a098 86072Author: Dave Airlie <airlied@redhat.com> 86073Date: Thu Sep 29 14:13:58 2011 +0100 86074 86075 add -Wall, cleanup warnings 86076 86077commit a4f6e661b5d6662c413c86eeefb4affddd65c562 86078Author: Dave Airlie <airlied@redhat.com> 86079Date: Thu Sep 29 14:05:43 2011 +0100 86080 86081 fix server recycling 86082 86083commit 7847321138be62b929123b25929a69f0a57aca3f 86084Author: Dave Airlie <airlied@redhat.com> 86085Date: Thu Sep 29 12:38:26 2011 +0100 86086 86087 port damage tracking code from st/xorg 86088 86089commit 4d3aadb22fc03f259a9d1717b55ce16d7a32e2be 86090Author: Dave Airlie <airlied@redhat.com> 86091Date: Thu Sep 29 12:34:27 2011 +0100 86092 86093 drop loading dri2 86094 86095commit 1b1382c9e1995dcfe5dfec1de17155fa627ac6e8 86096Author: Dave Airlie <airlied@redhat.com> 86097Date: Thu Sep 29 12:34:17 2011 +0100 86098 86099 update authors/copyright 86100 86101commit 263597dbb3a75d711b5ad2e1f51554e6b2601d9a 86102Author: Dave Airlie <airlied@redhat.com> 86103Date: Thu Sep 29 12:30:46 2011 +0100 86104 86105 gut some more unneeded code 86106 86107commit a72bdf170c8bc55eef3d2dbb8e316839552288a5 86108Author: Dave Airlie <airlied@redhat.com> 86109Date: Thu Sep 29 12:28:59 2011 +0100 86110 86111 modesetting: rewrite probing based on fbdev. 86112 86113 This isn't perfect, it should really do more with bus id matching 86114 and finding the write drm device using sysfs if it can. but it removes 86115 lots of hardcoding of pci ids. 86116 86117commit a77fae850511c4d73af2af2b88804586f950f28d 86118Author: Dave Airlie <airlied@redhat.com> 86119Date: Thu Sep 29 11:52:51 2011 +0100 86120 86121 cleanup some the drmmode_display header file 86122 86123commit 2a4618905445ce13fd272a4634ef5103dc804231 86124Author: Dave Airlie <airlied@redhat.com> 86125Date: Thu Sep 29 11:49:26 2011 +0100 86126 86127 modesetting: restart driver effort from other codebases 86128 86129 This starts a randr 1.2 compatible driver with cursors. 86130 86131 TODO: 86132 libkms 86133 dirty handling 86134 server recycling. 86135 86136commit 057cc92ebfeebe81b7d01ff2c6aa2c066c8d1a64 86137Merge: afb1fe695 52c9b59a9 86138Author: Peter Hutterer <peter.hutterer@who-t.net> 86139Date: Thu Sep 29 14:18:35 2011 +1000 86140 86141 Merge branch 'raw-events' into next 86142 86143 Conflicts: 86144 configure.ac 86145 dix/events.c 86146 86147commit 52c9b59a9fed6abfeca775a7a04133cee18eac0b 86148Author: Peter Hutterer <peter.hutterer@who-t.net> 86149Date: Mon May 30 09:47:06 2011 +1000 86150 86151 dix: use 'rc' for return code in DeliverRawEvent 86152 86153 Reported-by: Walter Harms <wharms@bfs.de> 86154 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86155 Reviewed-by: Cyril Brulebois <kibi@debian.org> 86156 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 86157 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86158 86159commit b6b1f1b514bf14ddf6b265b8d4551f892ded0dd2 86160Author: Peter Hutterer <peter.hutterer@who-t.net> 86161Date: Wed May 11 13:56:08 2011 +1000 86162 86163 input: deliver raw events unconditionally for XI 2.1 clients. 86164 86165 Deliver raw events regardless whether there is a grab on or not for clients 86166 supporting 2.1 or later. 86167 86168 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86169 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 86170 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86171 86172commit 16244fba001826190445302f56784f5de9c59c01 86173Author: Peter Hutterer <peter.hutterer@who-t.net> 86174Date: Thu Jun 2 13:50:13 2011 +1000 86175 86176 Support (and require) XI 2.1 86177 86178 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86179 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 86180 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86181 86182commit 54e05d80a122bac57920fce1704c0f57492b849c 86183Author: Peter Hutterer <peter.hutterer@who-t.net> 86184Date: Fri May 27 14:21:39 2011 +1000 86185 86186 dix: rename ProcessRawEvents to dix/events.c:DeliverRawEvent 86187 86188 No functional changes, prep work for future changes. 86189 86190 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86191 Reviewed-by: Cyril Brulebois <kibi@debian.org> 86192 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 86193 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86194 86195commit c48133f133c64d356a1208f185619bfdf7e9a5f2 86196Author: Peter Hutterer <peter.hutterer@who-t.net> 86197Date: Fri May 27 14:19:45 2011 +1000 86198 86199 Xi: use temporary variable for filter. 86200 86201 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86202 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 86203 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86204 86205commit ac0850e86f22191091a1eb07237cae9de49ee00d 86206Author: Peter Hutterer <peter.hutterer@who-t.net> 86207Date: Fri May 27 12:14:32 2011 +1000 86208 86209 dix: split DeliverEventToWindowMask up a bit more. 86210 86211 Move out the actual event delivery, it needs to be used from elsewhere. 86212 86213 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86214 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 86215 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86216 86217commit 5ea2fb389fce235366e9fce83d20abdc8874f4e2 86218Author: Peter Hutterer <peter.hutterer@who-t.net> 86219Date: Fri May 27 11:43:50 2011 +1000 86220 86221 dix: rename DeliverEventsToClients to DeliverEventsToWindowMask 86222 86223 To avoid confusion with a future patch and it better describes what this 86224 does anyway - delivering events to all clients that have the event mask on 86225 the window. 86226 86227 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86228 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 86229 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86230 86231commit 016413dae947fe4e8e918f728c87354fa4928275 86232Author: Peter Hutterer <peter.hutterer@who-t.net> 86233Date: Fri May 27 11:40:56 2011 +1000 86234 86235 dix: split client list retrieval out of DeliverEventToClients 86236 86237 No functional change, but "other" was renamed to "clients". 86238 86239 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86240 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 86241 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86242 86243commit 537c5f631d1f8d526e63355b7d88cb1fb2793492 86244Author: Cyril Brulebois <kibi@debian.org> 86245Date: Sun Jun 5 03:21:18 2011 +0200 86246 86247 xkb: Fix case checks for Latin 4. 86248 86249 That one was missing _XkbKSLower: 86250 XK_kra: U+0138 LATIN SMALL LETTER KRA 86251 86252 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86253 Signed-off-by: Cyril Brulebois <kibi@debian.org> 86254 86255commit 0b56a7f0ad01d3a86cdd7a14d5e567a00a3553f6 86256Author: Cyril Brulebois <kibi@debian.org> 86257Date: Sun Jun 5 03:12:44 2011 +0200 86258 86259 xkb: Fix case checks for Latin 2. 86260 86261 Those ones were getting _XkbKSLower for no reasons: 86262 XK_ogonek: U+02DB OGONEK 86263 XK_doubleacute: U+02DD DOUBLE ACUTE ACCENT 86264 86265 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86266 Signed-off-by: Cyril Brulebois <kibi@debian.org> 86267 86268commit 512a9750c197437889ff7f26a8d2bd242e030745 86269Author: Cyril Brulebois <kibi@debian.org> 86270Date: Sun Jun 5 03:03:47 2011 +0200 86271 86272 xkb: Fix case checks for Latin 1. 86273 86274 That one was missing _XkbKSLower: 86275 XK_ssharp: U+00DF LATIN SMALL LETTER SHARP S 86276 86277 That one was getting _XkbKSLower for no reasons: 86278 XK_division: U+00F7 DIVISION SIGN 86279 86280 For reference, XK_multiply was already excluded from the _XkbKSUpper 86281 check, it's no big surprise XK_division has to be excluded from the 86282 _XkbKSLower check. 86283 86284 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86285 Signed-off-by: Cyril Brulebois <kibi@debian.org> 86286 86287commit 635a1f50bc65512ec491fe71b9dfa8cf9118e2c2 86288Author: Peter Hutterer <peter.hutterer@who-t.net> 86289Date: Fri Sep 23 12:04:39 2011 +1000 86290 86291 input: allow for max < min for relative axes on InitValuatorAxisStruct 86292 86293 Relative axes are initialized with 0, -1 but so far this never had any 86294 effect as all users of this function (for relative axes) just set it to the 86295 defaults anyway. 86296 86297 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86298 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86299 86300commit 8d1a414cca51e7f8a93470c5aa2e998f5ca1bc9a 86301Author: Peter Hutterer <peter.hutterer@who-t.net> 86302Date: Fri Sep 23 12:03:48 2011 +1000 86303 86304 input: switch InitValuatorAxisStruct to return Bool 86305 86306 Return errors instead of silently ignoring them. 86307 86308 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86309 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 86310 86311commit 5ac1f885f5c106b970835de36e13a9a7ea413df4 86312Author: Daniel Stone <daniel@fooishbar.org> 86313Date: Tue Feb 15 19:44:53 2011 +0000 86314 86315 Input: Add POINTER_EMULATED flag to GetPointerEvents 86316 86317 POINTER_EMULATED merely sets XIPointerEmulated in the generated 86318 DeviceEvent. 86319 86320 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86321 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86322 86323commit 78d8d6dd7f31cb903d5c8baf64181795736f33ed 86324Author: Daniel Stone <daniel@fooishbar.org> 86325Date: Tue Feb 15 14:23:25 2011 +0000 86326 86327 Input: Add vertical and horizontal scroll axes 86328 86329 To be used for smooth scrolling with future driver APIs, replacing 86330 Rel Vert Wheel and Rel Horiz Wheel axes, which have not been used in any 86331 open driver to date. 86332 86333 Combined with double-granularity ValuatorMasks, these axes allow for 86334 fine-grained scroll data to be sent to clients. Future commits allow 86335 drivers to post these scroll axes to 86336 QueuePointerEvents/GetPointerEvents, which take care of emulating legacy 86337 scroll button events. 86338 86339 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86340 86341commit d8e42decbad4abe13265f4c546a0c561905d018f 86342Author: Daniel Stone <daniel@fooishbar.org> 86343Date: Tue Feb 15 18:54:14 2011 +0000 86344 86345 Input: Split GetPointerEvents body into a helper function 86346 86347 For smooth-scrolling support, we want GetPointerEvents to generate 86348 multiple events, so split the body of the function out into a helper 86349 function in order to call it multiple times. 86350 86351 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86352 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86353 86354commit bc8aad2376207b5ca9c74effae67fb8183222d2e 86355Author: Daniel Stone <daniel@fooishbar.org> 86356Date: Thu Mar 3 19:48:15 2011 +0000 86357 86358 Input: Set last valuators in GetPointerEvents only 86359 86360 Previously, various pieces of code, including acceleration, used to drop 86361 the values into DeviceIntRec::last.valuators. Remove all this and only 86362 do it in GetPointerEvents after all transformation, acceleration and 86363 clipping, so we're guaranteed to always have the correct values. 86364 86365 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86366 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86367 86368commit 94c19a0a72403fc522e3d05eeb57f35e111a2562 86369Author: Daniel Stone <daniel@fooishbar.org> 86370Date: Thu Mar 3 15:46:45 2011 +0000 86371 86372 Input: Convert DeviceIntRec::last to use doubles 86373 86374 Change the last real user of a split integer/fractional co-ordinate 86375 system, DeviceIntRec's last->{valuators,remainder} to just have one set 86376 of doubles. 86377 86378 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86379 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86380 86381commit 2d9beeb2174661d0a0732403ad4e49e9ff56718d 86382Author: Daniel Stone <daniel@fooishbar.org> 86383Date: Thu Mar 3 15:36:55 2011 +0000 86384 86385 Input: Make DeviceEvent use doubles internally 86386 86387 Change the DeviceEvent InternalEvent to use doubles for its valuators, 86388 instead of data and data_frac. 86389 86390 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86391 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86392 86393commit 626f487b97c99b48bf1d8c8ef4367a686639c3e6 86394Author: Daniel Stone <daniel@fooishbar.org> 86395Date: Thu Mar 3 15:06:45 2011 +0000 86396 86397 Input: Make RawDeviceEvent use doubles internally 86398 86399 Change RawDeviceEvent to use doubles for valuators internally, rather 86400 than data(_raw) and data(_raw)_frac. 86401 86402 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86403 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86404 86405commit 3b7fb0f68dc0d40c968c2cfc9bb74b1d0fb48bc8 86406Author: Daniel Stone <daniel@fooishbar.org> 86407Date: Wed Mar 2 17:18:48 2011 +0000 86408 86409 Input: Modify mask in-place in positionSprite 86410 86411 Instead of taking pointers to x and y values to modify in 86412 positionSprite, just modify the mask (as well as dev->last) in place. 86413 86414 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86415 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86416 86417commit 629a575261c08ca67324fea4c975636a1a95dc75 86418Author: Daniel Stone <daniel@fooishbar.org> 86419Date: Wed Mar 2 17:04:57 2011 +0000 86420 86421 Input: Convert positionSprite and GetPointerEvents to double 86422 86423 Use doubles internally in both of these functions, eliminating most of 86424 the remaining int co-ordinate usage. 86425 86426 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86427 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86428 86429commit 51437995a5041a8c53c33b508b1607c78a5fa463 86430Author: Daniel Stone <daniel@fooishbar.org> 86431Date: Wed Mar 2 16:50:55 2011 +0000 86432 86433 Input: Don't call positionSprite for non-pointer devices 86434 86435 If the device doesn't have any valuators, or if it has less than two of 86436 them, don't bother calling positionSprite. Users with one-dimensional 86437 pointing devices may be upset. 86438 86439 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86440 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86441 86442commit 4c364a312daf2b743a0a60b9907f671804a1b1b6 86443Author: Daniel Stone <daniel@fooishbar.org> 86444Date: Wed Mar 2 16:49:53 2011 +0000 86445 86446 Input: Convert rescaleValuatorAxis to double 86447 86448 Instead of passing fractional pointers around everywhere, just pass 86449 doubles instead. Much easier. 86450 86451 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86452 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86453 86454commit 5680fa41ea3373651f7017898a307e97cf29b0d3 86455Author: Daniel Stone <daniel@fooishbar.org> 86456Date: Wed Mar 2 16:30:30 2011 +0000 86457 86458 Input: Remove x and y from moveAbsolute/moveRelative 86459 86460 Both these functions modify the mask and 86461 pDev->last.{valuators,remainder} in-place now, so there's no need to 86462 pass in pointers to local x and y values. 86463 86464 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86465 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86466 86467commit 0882b788da97c75e464eb352dac1d83c938a148e 86468Author: Daniel Stone <daniel@fooishbar.org> 86469Date: Wed Mar 2 16:27:31 2011 +0000 86470 86471 Input: Convert acceleration code to using ValuatorMask 86472 86473 Instead of passing a set of int* to the acceleration code, pass it a 86474 mask instead, which avoids an unfortunate loss of precision. 86475 86476 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86477 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86478 Reviewed-by: Simon Thum <simon.thum@gmx.de> 86479 86480commit 2b8f1d07bd42c9d3db3dbacfe6a1335e47236a6c 86481Author: Daniel Stone <daniel@fooishbar.org> 86482Date: Wed Mar 2 15:37:53 2011 +0000 86483 86484 Input: Widen pointer acceleration types to double 86485 86486 This widens almost all of the float-using code in ptrveloc.[ch] to 86487 doubles, other than values coming from properties which are specified to 86488 be floats by the property API. 86489 86490 Bumps input API to v14 as this changes the AccelScheme signature, as 86491 used by xf86-input-synaptics. 86492 86493 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86494 86495commit 8a4a4e1b8a33899653d68a21bb7ff0557a576338 86496Author: Daniel Stone <daniel@fooishbar.org> 86497Date: Thu Jun 9 15:26:55 2011 +0100 86498 86499 Input: Use trunc instead of lrintf in acceleration code 86500 86501 For compatibility with all the rest of the input code. 86502 86503 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86504 Reviewed-by: Simon Thum <simon.thum@gmx.de> 86505 86506commit e1df51421bc392d57f88c831fe0661871a8aa769 86507Author: Daniel Stone <daniel@fooishbar.org> 86508Date: Wed Mar 2 13:34:32 2011 +0000 86509 86510 Input: Set fractional member in set_raw_valuators 86511 86512 RawDeviceEvents have space for fractional valuator members, so might as 86513 well start using them. 86514 86515 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86516 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86517 86518commit cdf202250ef616f3b261db70da30e421ea59a630 86519Author: Daniel Stone <daniel@fooishbar.org> 86520Date: Wed Mar 2 14:50:29 2011 +0000 86521 86522 Input: Convert transformAbsolute to work on doubles 86523 86524 Change transformAbsolute to use doubles internally. 86525 86526 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86527 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86528 86529commit 3463078f9697fad0ee11837d80e88889fc6a28a4 86530Author: Daniel Stone <daniel@fooishbar.org> 86531Date: Wed Mar 2 13:29:24 2011 +0000 86532 86533 Input: Convert clipAxis, moveAbsolute and moveRelative to double 86534 86535 Change all these three to use doubles internally, though the outputs of 86536 moveAbsolute and moveRelative are still truncated to int. 86537 86538 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86539 86540commit 4e52cc0ef48145134cd58d357fb7289e6f8bb709 86541Author: Daniel Stone <daniel@fooishbar.org> 86542Date: Wed Mar 2 13:20:37 2011 +0000 86543 86544 Input: Prepare moveRelative for conversion to double 86545 86546 Shuffle some code around in moveRelative to make the conversion to 86547 double easier later. 86548 86549 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86550 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86551 86552commit 80fdf9ca33a8f50f98a4b1a3279e66b040353b80 86553Author: Daniel Stone <daniel@fooishbar.org> 86554Date: Wed Mar 2 13:19:00 2011 +0000 86555 86556 Input: Prepare moveAbsolute for conversion to double 86557 86558 Shuffle some code around to make moving to double easier later. 86559 86560 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86561 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86562 86563commit 6a6b4eb05c7822860e2362fa9b8441fc67055e27 86564Author: Daniel Stone <daniel@fooishbar.org> 86565Date: Wed Mar 2 17:22:59 2011 +0000 86566 86567 Input: Store clipped absolute axes in the mask 86568 86569 Change moveAbsolute to be more symmetric with moveRelative by storing a 86570 clipped axis value back in the mask, rather than just in 86571 dev->last.valuators. 86572 86573 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86574 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86575 86576commit 7e919ef5bfa94d51a06eefb150ab947bdbfb6885 86577Author: Daniel Stone <daniel@fooishbar.org> 86578Date: Wed Feb 23 17:28:18 2011 +0000 86579 86580 Input: Add double-precision valuator_mask API 86581 86582 Add API for valuator_mask that accepts and returns doubles, rather than 86583 ints. No double API is provided for set_range at the moment. 86584 86585 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86586 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86587 86588commit 79d4deb76d5612b75ef26b4549b7ade3a656687d 86589Author: Daniel Stone <daniel@fooishbar.org> 86590Date: Wed Feb 23 14:31:01 2011 +0000 86591 86592 Input: Convert ValuatorMask to double-precision internally 86593 86594 Switch the ValuatorMask struct to using doubles instead of ints for the 86595 actual values. Preserve the old int API, and (attempt to) round towards 86596 zero for values we return. 86597 86598 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86599 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86600 86601commit 60a766a96f9ec97adf6d0273f70ec23f313a11c6 86602Author: Daniel Stone <daniel@fooishbar.org> 86603Date: Thu Jun 9 16:02:47 2011 +0100 86604 86605 Input: Add flags to RawDeviceEvent 86606 86607 Add a flags member which will be copied wholesale into the resultant 86608 xXIRawEvent. 86609 86610 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86611 86612commit 5d62c32981dbe9630bed32567b55008b536c81c5 86613Author: Daniel Stone <daniel@fooishbar.org> 86614Date: Mon Dec 27 13:24:35 2010 +0000 86615 86616 Input: Add flags to DeviceEvent 86617 86618 Add a flags member which will be copied wholesale into the resultant 86619 xXIDeviceEvent. 86620 86621 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 86622 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86623 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 86624 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86625 86626commit 38e9e28ba2fbffee52ad9889ef6d4e94c7af3e10 86627Author: Jeremy Huddleston <jeremyhu@apple.com> 86628Date: Tue Sep 20 20:39:06 2011 -0700 86629 86630 XQuartz: Use set_front_process rather than X11ApplicationSetFrontProcess since we're already in the AppKit thread 86631 86632 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 86633 86634commit 206b30ebc608d29a91cc18665d89e887cac4dba3 86635Author: Jeremy Huddleston <jeremyhu@apple.com> 86636Date: Thu Sep 22 18:55:19 2011 -0700 86637 86638 Address regressions from e8ff555b95ba and d206d52f657c to work with other compilers 86639 86640 1) The error attribute appeared in gcc-4.3 86641 2) The return type of __builtin_constant_p is int 86642 3) Sun Studio 12.0 and later builtin support for __builtin_constant_p 86643 86644 Found by Tinderbox. 86645 86646 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 86647 Reviewed-by: Matt Turner <mattst88@gmail.com> 86648 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 86649 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 86650 86651commit afb1fe695d197187a301c19863a128a65389b15c 86652Merge: 7fb4bef03 c7163fdd3 86653Author: Keith Packard <keithp@keithp.com> 86654Date: Mon Sep 26 20:24:15 2011 -0700 86655 86656 Merge remote-tracking branch 'whot/next' 86657 86658commit c7163fdd302f706a3d67f0fdf93eeb3396bb3332 86659Author: vdb@picaros.org <vdb@picaros.org> 86660Date: Wed Aug 31 07:23:01 2011 +0200 86661 86662 xfree86: .BI style: monitor section in xorg.conf man page 86663 86664 The xorg.conf manual uses the following convention in most of its 86665 sections: 86666 86667 bold = text to be copied literally to the config file, 86668 italic = a symbolic name to be substituted by a true value. 86669 86670 Some configuration keywords seem to have been changed into generic 86671 options. Prepending Option to the manual entry swapped the 86672 bold-italic logic. This patch restores the convention in the monitor 86673 section and consists of 86674 86675 -.BI "Option " "\*qPreferredMode\*q " \*qstring\*q 86676 +.BI "Option \*qPreferredMode\*q \*q" name \*q 86677 86678 modifications. 86679 86680 Plus a few minor changes (Modes → Mode) and a typo fix. 86681 86682 Signed-off-by: Servaas Vandenberghe 86683 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86684 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86685 86686commit 4b4caecb7d2c8e0b7e4fedc95fba2a728bbc25e6 86687Author: Peter Hutterer <peter.hutterer@who-t.net> 86688Date: Tue Sep 6 09:33:43 2011 +1000 86689 86690 xfree86: expose Option "TransformationMatrix" 86691 86692 Recent changes to the server change the default absolute input device 86693 behaviour on zaphods to span the whole desktop too. Since these setups 86694 usually use an xorg.conf, allow the transformation matrix to be specified in 86695 the config as well. 86696 86697 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86698 Reviewed-by: Simon Thum <simon.thum@gmx.de> 86699 86700commit cf51424a34fb2b567a867338ab44f83b5c43251c 86701Author: Peter Hutterer <peter.hutterer@who-t.net> 86702Date: Tue Sep 6 09:39:01 2011 +1000 86703 86704 xfree86: use subheader for Pointer Acceleration parts in xorg.conf(5) 86705 86706 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86707 86708commit 9125952b407090b040bffb3d752288293137cf77 86709Author: Peter Hutterer <peter.hutterer@who-t.net> 86710Date: Mon Sep 5 15:43:14 2011 +1000 86711 86712 xfree86: fix comment typo 86713 86714 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86715 86716commit 2d2dce558d24eeea0eb011ec9ebaa6c5c2273c39 86717Author: Sam Spilsbury <sam.spilsbury@canonical.com> 86718Date: Wed Sep 14 09:58:34 2011 +0800 86719 86720 Remove the SendEvent bit (0x80) before doing range checks on event type. 86721 86722 Some extension libraries may set this bit before converting the event to 86723 wire protocol and as such range checking the event will cause an invalid 86724 BadValue error to result. As the documentation suggests the the bit 86725 should be "forced on", remove it before doing range checks and continue 86726 to force it on in the server. 86727 86728 Reviewed-by: Jamey Sharp <jamey@minilop.net> 86729 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86730 86731commit f52d5cd374563544dafe29587411f345e31bbdf8 86732Author: Carlos Garnacho <carlosg@gnome.org> 86733Date: Wed Aug 31 00:46:52 2011 +0200 86734 86735 Xi: Fix passive XI2 ungrabs on XIAll[Master]Devices 86736 86737 The corresponding DeviceIntPtr wasn't being gotten properly, 86738 resulting in BadDevice from dixLookupDevice(). 86739 86740 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 86741 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 86742 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86743 86744commit 24823f1ab0f16e0815ef44340894537dc4babf6a 86745Author: Peter Hutterer <peter.hutterer@who-t.net> 86746Date: Wed Aug 3 16:01:28 2011 +1000 86747 86748 xfree86: switch options from pointer to XF86OptionPtr 86749 86750 In all cases, the pointer was simply type-cast anyway. Let's get some 86751 compile-time type safety going, how about that. 86752 86753 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86754 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 86755 86756 Squashed in: 86757 xfree86: Move definition of xf86OptionPtr into separate header file 86758 86759 The pile of spaghettis that is the xfree86 include dependencies make it 86760 rather hard to have a single typedef somewhere that's not interfering with 86761 everything else or drags in a whole bunch of other includes. 86762 86763 Move the xf86OptionRec and GenericListRec declarations into a separate 86764 header. 86765 86766 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 86767 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 86768 86769commit 284328a6beb0bd0ab6653973ce52b5a68b4e791f 86770Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86771Date: Mon Sep 26 17:04:56 2011 +0800 86772 86773 glamor-ddx: Rename glamor.c to glamor_ddx.c. 86774 86775 As glamor's dix module already has a glamor.c, rename this 86776 glamor.c to other name to avoid file name conflict which 86777 may confuse gdb. 86778 86779 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86780 86781commit 101b69dddff6cc538cb369c70377bbad825aa765 86782Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86783Date: Mon Sep 26 16:33:52 2011 +0800 86784 86785 glamor: Check libgbm if enable glamor ddx. 86786 86787 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86788 86789commit 6a7ad1c92934b35eb2431c3e0f11c0c56a097dc9 86790Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86791Date: Mon Sep 26 16:13:57 2011 +0800 86792 86793 glamor-ddx: Don't double free the cursor memory. 86794 86795 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86796 86797commit fe4c95ce843afffed981ac8bb89ca433bd20f1e2 86798Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86799Date: Mon Sep 26 16:08:29 2011 +0800 86800 86801 glamor: Refine the screen pixmap initialization for Xephyr. 86802 86803 The previous implementation is to override the CreatePixmap 86804 firstly and assume the first call to CreatePixmap must be 86805 screen pixmap. This is not clean. Now Refine it to normal 86806 way. Let the Xephyr to set texture 0 to screen pixmap 86807 during creating screen resources. 86808 86809 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86810 86811commit 1fb58860cc02d28545c66a916956a45918596870 86812Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86813Date: Mon Sep 26 16:05:52 2011 +0800 86814 86815 ephyr-glamor: Implment gl disptach function for host glx. 86816 86817 We should pass back the glXGetProcAddress to the underlying 86818 gl dispatch initialization function to get those gl function 86819 pointers. 86820 86821 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86822 86823commit df070205600cd736750d8a43426976a518b1fbae 86824Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86825Date: Mon Sep 26 16:01:03 2011 +0800 86826 86827 ephyr-glamor: Set texture 0 to screen pixmap. 86828 86829 In Xephyr, screen pixmap is different from other pixmap. 86830 It use fbo zero and texture zero. We initialize it during 86831 creating screen resources. 86832 86833 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86834 86835commit 937d3bcecf04305c01159f5c5084833f28014061 86836Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86837Date: Thu Sep 22 13:39:01 2011 +0800 86838 86839 glamor: Correct the version check for GLES2. 86840 86841 If we are using GLES2, then the version number should be no 86842 less than 2.0 rather than no less than 1.3 which is for GL 86843 API. 86844 86845commit 483717484afb4ccf5b48483b3a21a9fe91213dd4 86846Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86847Date: Tue Sep 20 17:15:12 2011 +0800 86848 86849 glamor-ddx: Destroy all the pixmap cursors when close current screen. 86850 86851 As now we use pixmap to handle cursor, and pixmaps are related to 86852 current screen. If we close the screen, then we have to destroy all 86853 the related cursors. 86854 86855 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86856 86857commit f1c36c611efaa56e49758d8e8b19d07c10a8277e 86858Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86859Date: Tue Sep 20 17:10:50 2011 +0800 86860 86861 glamor-ddx: Don't need to call preInit in screen_init. 86862 86863 The previous implementation call preInit again to reopen 86864 the device if the device is closed by eglTerminate. Now 86865 we don't need it anymore, as glamor dix doesn't call 86866 eglTerminate any more when close screen. 86867 86868 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86869 86870commit 09c63d320112fbb4995696d813c488e82915c2f5 86871Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86872Date: Tue Sep 20 15:19:13 2011 +0800 86873 86874 glamor-ddx: Reuse glamor_dix module to handle egl platform. 86875 86876 We already have a glamor egl module at glamor directory which 86877 could initialize and create egl context and also can attach a 86878 gbm buffer to a specified pixmap or to the screen pixmap. This 86879 commit remove all the duplicated code here and use the glamor_dix 86880 module directly. 86881 86882 As we don't want to handle egl stuffs in this layer, we have to 86883 change the way to handle hardware cursor. Previous method is to 86884 call egl functions to create a egl image and then bind the image 86885 to a texutre then call gl functions to load texture image to the 86886 cursor bo. Now we can bind the cursor bo to a cursor pixmap and 86887 then use putimage to load new image to the cursor bo. Then we can 86888 completely avoid to call egl or gl functions directly in this 86889 ddx driver. 86890 86891 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86892 86893commit 1590da001129ea61d6f7f8758d1d2842825be81b 86894Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86895Date: Tue Sep 20 15:14:49 2011 +0800 86896 86897 glamor-egl: Add gbm device support. 86898 86899 Latest mesa EGL implementation move to use gbm to manage/allocate buffers. 86900 To keep backward compatibility, we still try to use eglGetDRMDisplayMESA 86901 firstly, and if failed, then turn to use eglGetDisplay(gbm). 86902 86903 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86904 86905commit 2d4d40a4b5938249f3c73ed14fe5c5b926c2fefd 86906Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86907Date: Mon Sep 19 09:45:32 2011 +0800 86908 86909 glamor-ddx: Fix the broken switching between VTs. 86910 86911 Now we can allocate frame buffer during the screen initialization. 86912 And set mode major when enter VT. This is the correct logic. The 86913 previous implementation defer the set mode major to create screen 86914 resource which will cause the switching between VTs broken. 86915 86916 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86917 86918commit 4c8ed1756aa5121a559aa604926a730180a1403f 86919Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86920Date: Mon Sep 19 06:31:57 2011 +0800 86921 86922 glamor-ddx: Use gbm to allocate front buffer and cursor image. 86923 86924 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86925 86926commit 82a9b0c540dcf1064f5184c36a05d6216fb11034 86927Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86928Date: Fri Sep 16 13:37:13 2011 +0800 86929 86930 glamor-ddx: Change naming of glamor ddx module. 86931 86932 Use glamor_ddx prefix for all standalone glamor ddx functions. 86933 86934 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86935 86936commit 67dbe0ddf87a6dd602d3d4a6049072f6386d757e 86937Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86938Date: Thu Sep 8 15:16:01 2011 +0800 86939 86940 glamor-egl: Added new function to create image for a pixmap. 86941 86942 This function is used to support dri2. In the underlying 86943 driver, it will create a buffer object for a given pixmap. 86944 And then call this api to create a egl image from that 86945 buffer object, and then bind that image to a texture, and 86946 then bind that texture to the pixmap. 86947 86948 Normally, this pixmap's content is shared between a dri2 86949 client and the x server. 86950 86951 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86952 86953commit 2a9dfc963f610abfbf12d5efdf5ba5b3f55a4305 86954Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86955Date: Thu Sep 8 15:14:43 2011 +0800 86956 86957 glamor: Re-arrange some macros/definitions in header files. 86958 86959 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86960 86961commit 0dff23d65b8bcec615f4c7a49efa7a5bf220b299 86962Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86963Date: Thu Sep 8 14:05:46 2011 +0800 86964 86965 glamor: Don't direct call to any gl functions. 86966 86967 Create a new structure glamor_gl_dispatch to hold all the 86968 gl function's pointer and initialize them at run time , 86969 rather than use them directly. To do this is to avoid 86970 symbol conflicts. 86971 86972 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86973 86974commit 7daf9af086624133d6c4857ed92d57ad2a492d86 86975Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86976Date: Wed Aug 31 17:43:46 2011 +0800 86977 86978 glamor: Let GLAMOR_DDX implicit GLAMOR. 86979 86980 If user only enable GLAMOR_DDX, he must also want GLAMOR. 86981 86982 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86983 86984commit f08988455cd4f6339187e74df8edb5783b61f8fa 86985Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86986Date: Tue Aug 30 18:24:53 2011 +0800 86987 86988 glamor: Route UnrealizeGlyph to glamor_glyph_unrealize. 86989 86990 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 86991 86992commit 16659622708db82caaeb12527ba09545ad15b4c3 86993Author: Zhigang Gong <zhigang.gong@linux.intel.com> 86994Date: Tue Aug 30 17:45:07 2011 +0800 86995 86996 glamor: Improve glyphs cache mechanism. 86997 86998 This commit applying the latest uxa's glyphs cache mechanism 86999 and give up the old hash based cache algorithm. And the cache 87000 picture now is much larger than the previous one also. 87001 87002 This new algorithm can avoid the hash insert/remove and also 87003 the expensive sha1 checking. It could obtain about 10% 87004 performance gain when rendering glyphs. 87005 87006 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87007 87008commit bf4cbbd4e9c8ee104293a3c4eb202c880a139b73 87009Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87010Date: Tue Aug 30 17:37:47 2011 +0800 87011 87012 glamor: Move glamor_glyphs_init to create screen resources stage. 87013 87014 As in glamor_glyphs_init, we may need to create pixmap. Thus it must 87015 be called after the pixmap resources allocation. Just move it to 87016 screen resource creation stage is good enough for mow. 87017 87018 Also introduce a macro GLAMOR_FOR_XORG to glamor.h. As glamor may 87019 be used by Xephyr which doesn't don't have those xorg header files 87020 and also don't need the egl extension. 87021 87022 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87023 87024commit ae38bd7c459982d3e05f6f672ee68d187e97421c 87025Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87026Date: Thu Aug 25 10:05:11 2011 +0800 87027 87028 glamor: Change glamor_egl_init API. 87029 87030 Slightly change the API glamor_egl_init, 87031 as this initialization is to initialize the display not 87032 the screen, we should call it in xxx_preinit rather 87033 than xxxScreenInit(). we change the input parameter as 87034 below, as in preInit, the screen may not be allocated 87035 at all. And in glamor_egl_init, it will not call 87036 glamor_init. Driver should call glamor_init at 87037 screen_init stage. 87038 87039 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87040 87041commit d4f4d1272e1745752840eb7e8ec04eb155ac38d1 87042Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87043Date: Wed Aug 24 17:17:48 2011 +0800 87044 87045 glamor-egl: Fix direct call for glEGLImageTargetTexture2DOES. 87046 87047 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87048 87049commit 6cb4abca697c24bdc644502cce8a2ca9bbf131e3 87050Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87051Date: Wed Aug 24 17:13:35 2011 +0800 87052 87053 glamor: Remove the version check for EGL/gles. 87054 87055 As PVR's gles2 has different version number from mesa's, 87056 just simply remove the version check here to let both 87057 platform could build it. 87058 87059 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87060 87061commit 60f14d60f6169272951d41d8ca9e0a294c2ca3d0 87062Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87063Date: Wed Aug 24 16:54:32 2011 +0800 87064 87065 glamor-egl: Move EGL surfaceless related code to dix module. 87066 87067 The functions in glamor_egl.c could be shared by any egl back 87068 end drivers. 87069 87070 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87071 87072commit 3854409e9ad25418c6188f491410bb05eaf92fc6 87073Author: Zhenyu Wang <zhenyuw@linux.intel.com> 87074Date: Mon Aug 22 23:27:24 2011 -0700 87075 87076 glamor: Fix direct call for glEGLImageTargetTexture2DOES 87077 87078 Follow extension call rule to call glEGLImageTargetTexture2DOES. 87079 87080commit 54c91079d2bd21d3fd97b6d5fa3ba5c4a86b6337 87081Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87082Date: Tue Aug 23 14:04:07 2011 +0800 87083 87084 glamor-gles2: Add explicit precision qualifiers for gles2. 87085 87086 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87087 87088commit b4f265c7f9bcb7824df19ea915ab32fe84d9ec53 87089Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87090Date: Tue Aug 23 13:30:54 2011 +0800 87091 87092 glamor: Fallback to software fb when repeat is needed. 87093 87094 Need to be fixed latter. We should not need any fallback here. 87095 But currently, the implementation for repeating tiling is 87096 broken. 87097 87098 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87099 87100commit bd0ea43f3999e29ae5953ca5257b1795916dd5e7 87101Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87102Date: Tue Aug 23 13:27:48 2011 +0800 87103 87104 glamor: Change to use the original drawable in glamor_fill. 87105 87106 As glamor_fill may fallback to software rasterization, we'd 87107 better to use the original drawable as input paramter. 87108 87109 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87110 87111commit 65812b538f23bfccc32a7bbc0b2c4e93e1cd8fe8 87112Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87113Date: Tue Aug 23 10:31:50 2011 +0800 87114 87115 glamor-dix: Make a glamor dix module. 87116 87117 And modify the header file to export three APIs to the 87118 ddx driver. 87119 87120 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87121 87122commit ef3ea0f46bf1b688249cd37d54cfa67ec55041b0 87123Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87124Date: Fri Aug 19 11:06:56 2011 +0800 87125 87126 glamor: Don't need to check status when set the target. 87127 87128 As we already checked the status when create the fbo. 87129 87130 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87131 87132commit 52cce9333fed2035da9623f338428a5d220b577d 87133Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87134Date: Fri Aug 19 11:05:21 2011 +0800 87135 87136 glamor: egl version 7.11 should be ok for us. 87137 87138 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87139 87140commit 3648e35b5d0ffcb2ea5d40c842a7f8dd39b27b50 87141Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87142Date: Fri Aug 19 10:53:44 2011 +0800 87143 87144 glamor-ddx: Add code to check required egl extensions. 87145 87146 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87147 87148commit eaa07998c432f68485423c4d108220ce1d72e7b6 87149Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87150Date: Tue Aug 16 14:05:25 2011 +0800 87151 87152 glamor: Add device independent glamor to the installation package. 87153 87154 Other ddx driver may link libglamor_dix to utilize glamor. 87155 87156 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87157 87158commit 1658454dea4330144ebc0b02e1b42feed6db4f41 87159Author: Zhigang Gong <zhigang.gong@gmail.com> 87160Date: Tue Aug 16 00:33:35 2011 +0800 87161 87162 glamor: Fix the copy of overlaped region in one pixmap. 87163 87164 Originaly, we use fbo blit to handle overlaped region copy. 87165 But GLES2 doesn't support that, just simply copy the needed 87166 region to another texture can fix this problem. 87167 87168 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87169 87170commit 1f129c491bfc825465c3bc0e9b745b7f66dea720 87171Author: Zhigang Gong <zhigang.gong@gmail.com> 87172Date: Tue Aug 16 00:32:57 2011 +0800 87173 87174 glamor: Convert pixmap to supported format before getspans. 87175 87176 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87177 87178commit 67da52ec13a87846201a99d5a31db28668d9fdfd 87179Author: Zhigang Gong <zhigang.gong@gmail.com> 87180Date: Tue Aug 16 00:21:09 2011 +0800 87181 87182 glamor: Add color conversion support by using new shader. 87183 87184 There are two places we need to do color conversion. 87185 87186 1. When upload a image data to a texture. 87187 2. When download a texture to a memory buffer. 87188 87189 As the color format may not be supported in GLES2. We may 87190 need to do the following two operations to convert dat. 87191 87192 a. revert argb to bgra / abgr to rgba. 87193 b. swap argb to abgr / bgra to rgba. 87194 87195 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87196 87197commit 0eea084db54fb88b9933fd1326a2b0b059b3f9d7 87198Author: Zhigang Gong <zhigang.gong@gmail.com> 87199Date: Tue Aug 16 00:20:01 2011 +0800 87200 87201 glamor: GLES2 doesn't support glEnable/Disable Texture 2D. 87202 87203 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87204 87205commit 504e03c0b59f4dc3b8058abdce79706a72cd0dde 87206Author: Zhigang Gong <zhigang.gong@gmail.com> 87207Date: Tue Aug 16 00:17:58 2011 +0800 87208 87209 glamor: GLES2 doesn't support GL_CLAMP_TO_BORDER. 87210 87211 Simply comments it out. Need revisit latter. 87212 87213 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87214 87215commit f84cc8b84b2944b60eb2a8225e3c44a1cdfb228d 87216Author: Zhigang Gong <zhigang.gong@gmail.com> 87217Date: Tue Aug 16 00:09:38 2011 +0800 87218 87219 glamor-ddx: Hardware cursor's format should be GL_RGBA. 87220 87221 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87222 87223commit 23c1fba5e5a5fe8dd18c4c8a9b5a958ce4671160 87224Author: Zhigang Gong <zhigang.gong@gmail.com> 87225Date: Sat Aug 13 15:41:18 2011 -0400 87226 87227 glamor: Remove glu3 which is unnecessary. 87228 87229 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87230 87231commit a228effbeb339ae6f0b73c633022b3ec9981148f 87232Author: Zhigang Gong <zhigang.gong@gmail.com> 87233Date: Thu Aug 11 15:56:02 2011 -0400 87234 87235 glamor: Remove useless glVertexPointer related code. 87236 87237 As glVertexPointer is not supported by GLES2, I totally 87238 replaced it by VertexAttribArray. This commit remove those 87239 old code. 87240 87241 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87242 87243commit 5cbb2a4ca05f696d2af9a2e50bf106b6f2acd1c6 87244Author: Zhigang Gong <zhigang.gong@gmail.com> 87245Date: Thu Aug 11 15:35:26 2011 -0400 87246 87247 glamor: Only fallbac glamor_setspan when we are using gles2. 87248 87249 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87250 87251commit 667d65534d64df30146cfa0d32522bced51e0697 87252Author: Zhigang Gong <zhigang.gong@gmail.com> 87253Date: Thu Aug 11 15:27:12 2011 -0400 87254 87255 glamor: Unify the variable name which used to indicate no alpha. 87256 87257 The original code use different name and the name is vague. 87258 Now change it to no_alpha. 87259 87260 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87261 87262commit 5e7fdbb498c47b6e89280b8354f0ebaaf61d0646 87263Author: Zhigang Gong <zhigang.gong@gmail.com> 87264Date: Thu Aug 11 15:08:12 2011 -0400 87265 87266 glamor: Disable ALPHA8 fbo. 87267 87268 As some platform doesn't support to use ALPHA8 texture as 87269 draw target, we have to disable it. It seems there is no 87270 easy way to check that. 87271 87272 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87273 87274commit 172e8cfcd411b1abeaf8ede2e3882d6198cee5b8 87275Author: Zhigang Gong <zhigang.gong@gmail.com> 87276Date: Thu Aug 11 15:05:02 2011 -0400 87277 87278 glamor: Remove GLEW dependency. 87279 87280 Glamor doesn't need to use GLEW. We can parse the extension by 87281 ourself. This patch also fix the fbo size checking from a hard 87282 coded style to a dynamic checking style. 87283 87284 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87285 87286commit 2146a25bac4392c47ef06ecddbf12db184926b7e 87287Author: Zhigang Gong <zhigang.gong@gmail.com> 87288Date: Thu Aug 11 15:03:48 2011 -0400 87289 87290 glamor-ddx: Silence one warning. 87291 87292 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87293 87294commit d66bd9714e1e1af4267fb4eb208143a070e65ee3 87295Author: Zhigang Gong <zhigang.gong@gmail.com> 87296Date: Thu Aug 11 09:54:28 2011 -0400 87297 87298 glamor: Concentrate FBO's creation to one function. 87299 87300 And add status checking for it. 87301 87302 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87303 87304commit b2bff334ce4f36352497f1c1d4e9f0e5b3336136 87305Author: Zhenyu Wang <zhenyuw@linux.intel.com> 87306Date: Tue Aug 9 19:37:17 2011 -0700 87307 87308 glamor: Require libdrm CFLAGS for building 87309 87310 Need to depend on libdrm CFLAGS for glamor building. 87311 87312commit 41b13fbd264594f88905b1cf8dcb2f14e7ee5b69 87313Author: Zhenyu Wang <zhenyuw@linux.intel.com> 87314Date: Wed Aug 10 01:42:26 2011 -0700 87315 87316 glamor: Use function pointer to get extenstion calls 87317 87318 This is the formal usage for extension functions. 87319 87320commit 7aecfa245facf754ff8e09d6ef1ad0ab802156fc 87321Author: Zhigang Gong <zhigang.gong@gmail.com> 87322Date: Tue Aug 9 10:01:39 2011 -0400 87323 87324 glamor-es2: Add --enable-glamor-gles2 to build system. 87325 87326 Now, to build a gles2 version of glamor server, we could 87327 use ./autogen.sh --enable-glamor-ddx --enable-glamor-gles2 87328 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87329 87330commit 36a93f62c7830af088a54514eab6bf4d92423df6 87331Author: Zhigang Gong <zhigang.gong@gmail.com> 87332Date: Tue Aug 9 09:59:56 2011 -0400 87333 87334 glamor_gles2: Use gl_flavor to determine which version of GL. 87335 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87336 87337commit 29e358f6b3838a56b53ec809665f2e7be1731722 87338Author: Zhigang Gong <zhigang.gong@gmail.com> 87339Date: Tue Aug 9 09:59:08 2011 -0400 87340 87341 glamor: Fix one typo error in README. 87342 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87343 87344commit ee33c947aac10b2b1ec2c5b070fbd81cef1773c6 87345Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87346Date: Wed Aug 3 16:52:19 2011 +0800 87347 87348 glamor-es: Use glVertexAttribPointer to replace glVertexPointer. 87349 87350 As GLES2 doesn't support glVertexPointer. 87351 87352 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87353 87354commit 94d884a198b97ed2e78a38ce5a71955bdbf893f8 87355Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87356Date: Wed Aug 3 10:49:18 2011 +0800 87357 87358 glamor-es2: Fix some non-es2 functions. 87359 87360 ES2.0 doesn't support QUADS and also doesn't support 87361 some EXT APIs. Fix some of them in this commit. 87362 87363 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87364 87365commit 68c3c6eb0cafd5eb2c208dd76a3a65187256700c 87366Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87367Date: Tue Aug 2 15:54:15 2011 +0800 87368 87369 glamor-es2: Add gles2 option. 87370 87371 First commit to enable gles2 support. --enable-glamor-ddx 87372 --enable-glamor-gles2 will set thwo MACROs GLAMOR_DDX and 87373 GLAMOR_GLES2. 87374 87375 Currently, the gles2 support is still incomplete. 87376 87377 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87378 87379commit 98f8ef5f991e3c9d4b434af9003c466ad1edc23e 87380Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87381Date: Tue Aug 2 15:51:46 2011 +0800 87382 87383 glamor: Change to use official APIs rather than ARB version function. 87384 87385 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87386 87387commit 0ef1698be22750fb14418bd418114fc63039db0c 87388Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87389Date: Thu Jul 21 16:18:48 2011 +0800 87390 87391 glamor: Fix one bug for Xephyr. 87392 87393 Xephyr doesn't has a bounded valid texture. It seems that we can't 87394 load texture 0 directly sometimes. Especially in the copyarea, function 87395 if that is the case, we prefer to use fbo blit to read the screen pixmap 87396 rather than load the bound texture. 87397 87398 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87399 87400commit 9b667ffd564910ff7373547ce2f8d4a7c5d71137 87401Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87402Date: Wed Jul 20 14:33:00 2011 +0800 87403 87404 glamor: Use small internal texture format if possible. 87405 87406 Reduce some texture memory requirement and also save 87407 some bandwidth. 87408 87409 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87410 87411commit 01489f9da9aa357c59799d4d41c7ca96b4cbd1d8 87412Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87413Date: Wed Jul 20 14:02:15 2011 +0800 87414 87415 glamor: Add simple introduction of how to setup glamor xserver. 87416 87417 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87418 87419commit 1f3f3baf1469e74767474ddb383e92502384de80 87420Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87421Date: Wed Jul 20 13:35:25 2011 +0800 87422 87423 glamor: Avoid 2D bitblit if possible. 87424 87425 It turns out that the use of fbo blit is one of the root cause 87426 which lead to slow drawing, especially slow filling rects. 87427 87428 We guess there should be a performance bug in the mesa driver 87429 or even in the kernel drm driver. Currently, the only thing 87430 glamor can do is to avoid calling those functions. 87431 87432 We check whether the copy source and destination has overlapped 87433 region, if it has, we have to call fbo blit function. If it has 87434 not, we can load the source texture directly and draw it to the 87435 target texture. We totally don't need the glCopyPixels here, so 87436 remove it. 87437 87438 By apply this patch, the rendering time of firefox-planet-gnome 87439 decrease to 10.4 seconds. At the same platform, uxa driver get 13 87440 seconds. This is the first time we get better performance than 87441 uxa driver. 87442 87443 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87444 87445commit 5c4d53c5126bb9f603d04dac6874280b527319f5 87446Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87447Date: Wed Jul 20 13:06:23 2011 +0800 87448 87449 glamor: Implement delayed solid filling. 87450 87451 When we need to solid fill an entire pixmap with a specific color, 87452 we do not need to draw it immediately. We can defer it to the 87453 following occasions: 87454 87455 1. The pixmap will be used as source, then we can just use a shader 87456 to instead of one copyarea. 87457 2. The pixmap will be used as target, then we can do the filling 87458 just before drawing new pixel onto it. The filling and drawing 87459 will have the same target texture, we can save one time of 87460 fbo context switching. 87461 87462 Actually, for the 2nd case, we have opportunity to further optimize 87463 it. We can just fill the untouched region. 87464 87465 By applying this patch, the cairo-trace for the firefox-planet-gnome's 87466 rendering time decrease to 14seconds from 16 seconds. 87467 87468 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87469 87470commit 1caf741a4a41ef46a43980ba0be897770fea13db 87471Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87472Date: Wed Jul 20 13:02:28 2011 +0800 87473 87474 glamor: Fixed a bug when computing the bounds of boxes. 87475 87476 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87477 87478commit fe0a6a29300055f8288c9d1f470fd3e24f5bb3bb 87479Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87480Date: Thu Jul 7 16:54:18 2011 +0800 87481 87482 glamor: Should return when done gl drawing. 87483 87484 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87485 87486commit 7fbdc60fd4eb94c02b0b89d05cfe0243c3562ec1 87487Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87488Date: Thu Jul 7 15:32:39 2011 +0800 87489 87490 glamor: comment out the message when creating a system memory pixmap. 87491 87492 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87493 87494commit f961390db6fa0707895effb2e7672edda575f8ec 87495Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87496Date: Thu Jul 7 15:31:35 2011 +0800 87497 87498 glamor: Don't need to pad the tile image if pixmap is ni texture. 87499 87500 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87501 87502commit 77ecd366933bbe726a4a7e57e35b1510b675df8d 87503Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87504Date: Thu Jul 7 15:30:19 2011 +0800 87505 87506 glamor: We don't need to check format in compositing. 87507 87508 We already handle all format checking in pixmap uploading and 87509 converting, don't need to do that again. 87510 87511 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87512 87513commit 477a54bc9ed8f252b3a42af8ac01ef1da7b5ab6b 87514Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87515Date: Thu Jul 7 14:47:51 2011 +0800 87516 87517 glamor: Fixed two unintialized warnings. 87518 87519 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87520 87521commit d49d48257917a4a900c84dae111d3f5e761a8d89 87522Author: Zhigang Gong <zhigang.gong@gmail.com> 87523Date: Sat Jul 2 14:47:38 2011 +0800 87524 87525 glamor: Fix multiple crtc setup. 87526 87527 Now, support dual crtc configuration. 87528 87529 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87530 87531commit fc23334e54b0b90d2f00b017a03cf47b70dc71aa 87532Author: Zhigang Gong <zhigang.gong@gmail.com> 87533Date: Fri Jul 1 23:51:42 2011 +0800 87534 87535 glamor: Optimize fallback case for the polylines. 87536 87537 When fallback to cpu for the polylines procedure, we can just download 87538 required region to CPU rather than to download the whole pixmap. This 87539 significant improve the performance if we have to fallback, for example 87540 do non-solid filling in the game Mines. 87541 87542 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87543 87544commit da66a76f276eccee90855bc0cb28092c3755ed7b 87545Author: Zhigang Gong <zhigang.gong@gmail.com> 87546Date: Fri Jul 1 23:48:21 2011 +0800 87547 87548 Revert "glamor: Don't need to read dest if op is SRC or CLEAR." 87549 87550 This reverts commit eb16fe0b7c8ea27b5cf9122d02e48bf585495228. 87551 As currently glamor_prepare_access/finish_access will touch 87552 the whole pixmap, not just the request region, then write only 87553 mode will not work correctly. We may need to revisit all fallback 87554 case, and convert the image to the right size before do the 87555 prepare/finish processing. 87556 87557 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87558 87559commit cb3b34ec53e00c6ca212891b314f473ee2f80a40 87560Author: Zhigang Gong <zhigang.gong@gmail.com> 87561Date: Fri Jul 1 22:54:45 2011 +0800 87562 87563 glamor-ddx: Add missed drmmode_crtc_destroy function. 87564 87565 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87566 87567commit c68fb6ed723d7268a3c4950b40c20ddf9b4b9ca5 87568Author: Zhigang Gong <zhigang.gong@gmail.com> 87569Date: Fri Jul 1 22:36:31 2011 +0800 87570 87571 glamor-ddx: Reinit crtc if we need to reopen the card. 87572 87573 As the eglTerminate will close the card when close screen, we may 87574 need to reopen it at next time create a screen resource. and thus 87575 we need to re initialize the drmmode crtc also. Otherwise , the 87576 cursor handling will be broken as it has the wrong fd. 87577 87578commit ca614860fa2a56d469064b5e05070d8201597728 87579Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87580Date: Fri Jul 1 17:02:10 2011 +0800 87581 87582 glamor: convert if too large source or mask . 87583 87584 Some strange web page has 20000*1 png picture, and actually only use 87585 partial of it. We force to convert it to a actuall size rather than 87586 its original size,if it is the case. Then to avoid latter's failure 87587 uploading. 87588 87589 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87590 87591commit 1444fed4a8030ec59f4448b490f2f77db86f0b86 87592Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87593Date: Thu Jun 30 16:53:24 2011 +0800 87594 87595 glamor: silence compilation warnings. 87596 87597 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87598 87599commit 734b10e9f7e729fc300ed432bd3c81e1ea7483f3 87600Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87601Date: Thu Jun 30 16:50:23 2011 +0800 87602 87603 glamor: Fix one typo bug in glamor_tile. 87604 87605 It will return when the destination pixmap has a fbo but will continue 87606 when it doesn't have a fbo. 87607 87608 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87609 87610commit 2eb40a37920132d53e3b48830704f92d4c695f4c 87611Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87612Date: Thu Jun 30 16:49:40 2011 +0800 87613 87614 glamor: Remove one extra area copy in glamor_glyph. 87615 87616 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87617 87618commit 61e1ad39724e5baaed4f30ff3c86782e701f19fb 87619Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87620Date: Thu Jun 30 16:40:32 2011 +0800 87621 87622 glamor: Reduce source or mask picture size if possible. 87623 87624 If we only need a short part of the source or mask's drawable 87625 pixmap, we can convert it to a new small picture before 87626 call to the low level compositing function. Then it will only 87627 upload the smaller picture latter. 87628 87629 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87630 87631commit 14503fbb819d07fa2b58d9bb741b55f93aff2cf6 87632Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87633Date: Thu Jun 30 16:38:43 2011 +0800 87634 87635 glamor: Fix the coords calculation in glamor_fill. 87636 87637 glamor_fill is only called from internal functions 87638 glamor_fillspancs and glamor_polyfillrect. And both functions 87639 already add the offset to the coords, so the coords are already 87640 relative value, we can't add the offset once again. 87641 87642 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87643 87644commit 1dca5d7b91ed6fced34f389453678407bc223e9c 87645Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87646Date: Thu Jun 30 16:04:59 2011 +0800 87647 87648 glamor: Reduce source pixmap's size. 87649 87650 If the dest pixmap is in texture memory, but source pixmap is not. 87651 Then we need to upload the source pixmap to texture memory. Previous 87652 version will upload the whole source pixmap. This commit preprocess 87653 the source pixmap, and reduce it to a smaller tempory pixmap only 87654 contains the required region. 87655 87656 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87657 87658commit 33c6c78ae9c7e72a94cc27536f3b5f50cdcc9241 87659Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87660Date: Thu Jun 30 16:01:57 2011 +0800 87661 87662 glamor: Add one option to force create a cpu memory pixmap. 87663 87664 Some special case we want to get a cpu memory pixmap. For example 87665 to gather a large cpu memory pixmap's block to a small pixmap. 87666 87667 Add pixmap's priviate data's deallocation when destroy a pixmap. 87668 87669 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87670 87671commit 8890b38857128181cf506356046bf4a5bce53ab0 87672Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87673Date: Tue Jun 28 17:17:23 2011 +0800 87674 87675 glamor: Don't map the vbo to system memory. 87676 87677 Access mapped vbo address is too slow. And by use system memory 87678 directly, rgb10text/aa10text increases from 980K/1160K to 117K/140K. 87679 87680 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87681 87682commit c303949aabacb6a105ca0e0c521ca293dbda7d43 87683Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87684Date: Tue Jun 28 14:49:38 2011 +0800 87685 87686 glamor: Reduce one extra copy in glamor_trapezoids. 87687 87688 This reduce the time when running cairo-performance-trace with 87689 the firefox-planet-gnome.trace from 23.5 seconds to 21.5 seconds. 87690 87691 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87692 87693commit 9e4567afe6019ed7a330182d660880d8ea6c3685 87694Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87695Date: Tue Jun 28 14:37:29 2011 +0800 87696 87697 glamor: Change to use system memory for write only access. 87698 87699 If the pixmap is write-only, then use a pbo mapping will not 87700 get too much benefit. And even worse, when the software 87701 rendering is access this mapped data range, it's much slower 87702 than just using a system memory. From the glamor_prepare_access 87703 glamor_finish_access view, we have two options here: 87704 87705 option 1: 87706 1.0 create a pbo 87707 1.1 copy texture to the pbo 87708 1.2 map the pbo to va 87709 1.3 access the va directly in software rendering. 87710 1.4 bind the pbo as unpack buffer & draw it back to texture. 87711 87712 option 2: 87713 2.0 allocate a block memory in system memory space. 87714 2.1 read the texture memory to the system memory. 87715 2.2 access the system memory and do rendering. 87716 2.3 draw the system memory back to texture. 87717 87718 In general, 1.1 plush 1.2 is much faster than 2.1. 87719 And 1.3 is slower than 2.2. 1.4 is faster than 2.3. 87720 87721 If the access mode is read only or read write, option 1 87722 may be fater, but if the access mode is write only. Then 87723 most of the time option 1 is much faster. 87724 87725 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87726 87727commit 4afa9e4080eebbee0752e5f45b2ff16df75cb9b1 87728Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87729Date: Tue Jun 28 14:28:19 2011 +0800 87730 87731 glamor: Prepare/finish access once if src equal to dst. 87732 87733 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87734 87735commit cbedfe75135e3253c32ac8b89380f47d1b8a37c4 87736Author: Zhigang Gong <zhigang.gong@gmail.com> 87737Date: Sun Jun 26 20:00:26 2011 +0800 87738 87739 glamor: Don't need to read dest if op is SRC or CLEAR. 87740 87741 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87742 87743commit b8e692d94eb7075409e0d072fdf3066ca1d5d5f3 87744Author: Zhigang Gong <zhigang.gong@gmail.com> 87745Date: Sun Jun 26 15:54:24 2011 +0800 87746 87747 glamor: Move the blend set up after the pixmap uploading. 87748 87749 This is a bug, as if we do blend set up before do the pixmap 87750 dynamic uploading. We will have a incorrect blend env when 87751 doing the uploading. 87752 87753 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87754 87755commit bf782283371bc13c7b969a373c2a2eaaa3c81ec6 87756Author: Zhigang Gong <zhigang.gong@gmail.com> 87757Date: Sun Jun 26 15:51:47 2011 +0800 87758 87759 glamor: Fixed one bug when enable dynamic pixmap uploading. 87760 87761 When try to upload a pixmap without yInverted set, we must 87762 set up a fbo for it to do the y flip. Previous implementation 87763 only consider the ax bit. After fix this problem, we can 87764 enable the dynamic uploading feature in copyarea function when 87765 the yInverted is not set (from Xephyr). 87766 87767 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87768 87769commit 489e6c4e6f734d961ae8bbec67d78fccbeb9781f 87770Author: Zhigang Gong <zhigang.gong@gmail.com> 87771Date: Sun Jun 26 15:49:13 2011 +0800 87772 87773 glamor: Initialize pixmap private's container to correct value. 87774 87775 When calling from ephyr, we forgot to initialize it to the correct 87776 value. Will cause segfault when run Xephyr. 87777 87778 Signed-off-by: Zhigang Gong <zhigang.gong@gmail.com> 87779 87780commit ca36ada041558d1823b12b1a0141e43b9f9ad939 87781Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87782Date: Wed Jun 22 18:15:02 2011 +0800 87783 87784 glamor: Don't always fallback everything when change window attr. 87785 87786 Change the glamor_change_window_attributes's handling. We don't need 87787 to fallback every thing to cpu at the beginning. Only when there 87788 is a real need to change the pixmap's format, we need to do something. 87789 Otherwise, we need do nothing here. 87790 87791 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87792 87793commit ca1908e11dcb56cb952f6bce55503e932aa9a27c 87794Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87795Date: Wed Jun 22 14:33:38 2011 +0800 87796 87797 glamor: Concentrate and reduce some coords processing code. 87798 87799 Concentrate the verties and texture coords processing code to a new 87800 file glamor_utils.h. Change most of the code to macro. Will have some 87801 performance benefit on slow machine. And reduce most of the duplicate 87802 code when calculate the normalized coords. 87803 87804 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87805 87806commit 355334fcd99e4dce62e2be1e27290c9a74ea944f 87807Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87808Date: Tue Jun 21 18:31:11 2011 +0800 87809 87810 glamor : Add dynamic texture uploading feature. 87811 87812 Major refactoring. 87813 1. Rewrite the pixmap texture uploading and downloading functions. 87814 Add some new functions for both the prepare/finish access and 87815 the new performance feature dynamic texture uploading, which 87816 could download and upload the current image to/from a private 87817 texture/fbo. In the uploading or downloading phase, we need to 87818 handle two things: 87819 The first is the yInverted option, If it set, then we don't need 87820 to flip y. If not set, if it is from a dynamic texture uploading 87821 then we don't need to flip either if the current drawing process 87822 will flip it latter. If it is from finish_access, then we must 87823 flip the y axis. 87824 87825 The second thing is the alpha channel hanlding, if the pixmap's 87826 format is something like x8a8r8g8, x1r5g5b5 which means it doesn't 87827 has alpha channel, but it do has those extra bits. Then we need to 87828 wire those bits to 1. 87829 87830 2. Add almost all the required picture format support. 87831 This is not as trivial as it looks like. The previous implementation 87832 only support GL_a8,GL_a8r8g8b8,GL_x8r8g8b8. All the other format, 87833 we have to fallback to cpu. The reason why we can't simply add those 87834 other color format is because the exists of picture. one drawable 87835 pixmap may has one or even more container pictures. The drawable pixmap's 87836 depth can't map to a specified color format, for example depth 16 can 87837 mapped to r5g6b5, x1r5g5b5, a1r5g5b5, or even b5g6r5. So we can't get 87838 get the color format just from the depth value. But the pixmap do not 87839 has a pict_format element. We have to make a new one in the pixmap 87840 private data structure. Reroute the CreatePicture to glamor_create_picture 87841 and then store the picture's format to the pixmap's private structure. 87842 87843 This is not an ideal solution, as there may be more than one pictures 87844 refer to the same pixmap. Then we will have trouble. There is an example 87845 in glamor_composite_with_shader. The source and mask often share the 87846 same pixmap, but use different picture format. Our current solution is to 87847 combine those two different picture formats to one which will not lose any 87848 data. Then change the source's format to this new format and then upload 87849 the pixmap to texture once. It works. If we fail to find a matched new 87850 format then we fallback. 87851 87852 There still is a potential problem, if two pictures refer to the same 87853 pixmap, and one of them destroy the picture, but the other still remained 87854 to be used latter. We don't handle that situation currently. To be fixed. 87855 87856 3. Dynamic texture uploading. 87857 This is a performance feature. Although we don't like the client to hold 87858 a pixmap data to shared memory and we can't accelerate it. And even worse, 87859 we may need to fallback all the required pixmaps to cpu memory and then 87860 process them on CPU. This feature is to mitigate this penalty. When the 87861 target pixmap has a valid gl fbo attached to it. But the other pixmaps are 87862 not. Then it will be more efficient to upload the other pixmaps to GPU and 87863 then do the blitting or rendering on GPU than fallback all the pixmaps to CPU. 87864 To enable this feature, I experienced a significant performance improvement 87865 in the Game "Mines" :). 87866 87867 4. Debug facility. 87868 Modify the debug output mechanism. Now add a new macro: 87869 glamor_debug_output(_level_, _format_,...) to conditional output some messages 87870 according to the environment variable GLAMOR_DEBUG. We have the following 87871 levels currently. 87872 exports GLAMOR_DEBUG to 3 will enable all the above messages. 87873 87874 5. Changes in pixmap private data structure. 87875 Add some for the full color format supports and relate it to the pictures which 87876 already described. Also Add the following new elements: 87877 gl_fbo - to indicates whether this pixmap is on gpu only. 87878 gl_tex - to indicates whether the tex is valid and is containing the pixmap's 87879 image originally. 87880 As we bring the dynamic pixmap uploading feature, so a cpu memory pixmap may 87881 also has a valid fbo or tex attached to it. So we will have to use the above 87882 new element to check it true type. 87883 87884 After this commit, we can pass the rendercheck testing for all the picture formats. 87885 And is much much fater than fallback to cpu when doing rendercheck testing. 87886 87887 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87888 87889commit ba6dd8aa492d9d555d8b175bcf350e5db1821597 87890Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87891Date: Fri Jun 10 18:32:15 2011 +0800 87892 87893 glamor: Simplify fill acceleration for spans/polyfillrect by only clipping once. 87894 87895 This commit was borrowed from uxa driver contributed by Eric. 87896 commit number is e0066e77e026b0dd0daa0c3765473c7d63aa6753. commit log paste as 87897 below: 87898 We were clipping each span against the bounds of the clip, throwing 87899 out the span early if it was all clipped, and then walked the clip box 87900 clipping against each of the cliprects. We would expect spans to 87901 typically be clipped against one box, and not thrown out, so we were 87902 not saving any work there. For multiple cliprects, we were adding 87903 work. Only for many spans clipped entirely out of a complicated clip 87904 region would it have saved work, and it clearly didn't save bugs as 87905 evidenced by the many fix attempts here. 87906 87907 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87908 87909commit a0a52be73994c93521e94f1f10152ce851485ea3 87910Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87911Date: Fri Jun 10 16:51:27 2011 +0800 87912 87913 glamor: Fallback to fbPolylines for diagonal poly_line. 87914 87915 It's better to give a correct output when we haven't 87916 implement all the code path. 87917 87918 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87919 87920commit d7f8b888d0961af8d0ec6ad5c920ce758529b620 87921Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87922Date: Fri Jun 10 16:01:40 2011 +0800 87923 87924 glamor: For non-supported fill style fallback to fbFill. 87925 87926 The previous implementation will just skip the rendering 87927 which is not good. 87928 87929 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87930 87931commit b60e6cb66d3fba43d5403b2b5027537d09ee7c98 87932Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87933Date: Fri Jun 10 15:46:23 2011 +0800 87934 87935 glamor: Silence compilation warnings. 87936 87937 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87938 87939commit 229240e565f976de8e7b82dd606c2e862152b2b5 87940Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87941Date: Fri Jun 10 14:07:54 2011 +0800 87942 87943 glamor: Add render triangles support. 87944 87945 By default, fallback to frame buffer currently. This commit 87946 makes us pass the rendercheck's triangles testing. 87947 87948 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87949 87950commit ac0589c91699433bc9dbc25b7edff456dff742a4 87951Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87952Date: Fri Jun 10 14:02:19 2011 +0800 87953 87954 glamor: Use software fb for 1bpp pixmap. 87955 87956 For 1bpp pixmap, software fb get better performance than 87957 GL surface. The main reason is that fbo doesn't support 87958 1bpp texture as internal format, so we have to translate 87959 a 1bpp bitmap to a 8bit alpha format each time which is 87960 very inefficient. And the previous implementation is 87961 not supported by the latest OpenGL 4.0, the GL_BITMAP 87962 was deprecated. 87963 87964 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87965 87966commit 3c44e3e0ce1e286e0540298d5db547c43903629f 87967Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87968Date: Wed Jun 1 19:35:15 2011 +0800 87969 87970 glamor: Optimize composite when soure/mask is xrgb. 87971 87972 Added a new shader aswizlle_prog to wired the alpha to 1 when 87973 the image color depth is 24 (xrgb). Then we don't need to fallback 87974 the xrgb source/mask to software composite in render phase. Also 87975 don't wire the alpha bit to 1 in the render phase. This can get 87976 about 2x performance gain with the cairo performance trace's 87977 firefox-planet case. 87978 87979 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87980 87981commit 0e2af4d0c942405dd0869f1e6a1effee943de139 87982Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87983Date: Wed Jun 1 18:14:01 2011 +0800 87984 87985 glamor: Don't print those fallback messages by default. 87986 87987 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 87988 87989commit 925fc9724ff72d9e2a4940d7a1e39c23fb3ec9f0 87990Author: Zhigang Gong <zhigang.gong@linux.intel.com> 87991Date: Wed Jun 1 18:02:03 2011 +0800 87992 87993 glamor: Optimize glamor_finish_access. 87994 87995 use pbo if possible when we load texture to a temporary tex. 87996 And for the previous direct texture load function, it's not 87997 correct and get removed in this commit. 87998 87999 Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com> 88000 88001commit b8ce483f58112f200dec9853fcd7ff455479990f 88002Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88003Date: Wed Jun 1 17:54:01 2011 +0800 88004 88005 glamor: Don't use glamor_pixmap_type currently. 88006 88007 Added comments to glamor_pixmap_create. To be refined in the future. 88008 We need to identify whether a pixmap is a CPU memory pixmap or a 88009 GPU pixmap. Current implementation is not correct. There are three 88010 cases: 88011 88012 1. Too large pixmap, we direct it to CPU memory pixmap. 88013 2. w ==0 || h == 0 pixmap, this case has two possibilities: 88014 2.1 It will become a screen pixmap latter, then it should be 88015 GPU type. 88016 2.2 It's a scratch pixmap or created from a share memory, then 88017 it should belong to CPU memory. 88018 88019 XXX, need to be refined latter. 88020 88021commit 28835be1b8a05e510aa7fcc2f331771e3cf7dfec 88022Author: Zhigang Gong <zhigang.gong@gmail.com> 88023Date: Tue May 24 22:01:49 2011 +0800 88024 88025 glamor: improve glamor_finish_access. 88026 88027 When the platform's coordinate system is the same as X11's . We 88028 can load the texture to the fbo directly without one extra texture 88029 transformation. 88030 88031commit 1edf0cc6ab340c98fde201328631ca61e9d871a1 88032Author: Zhigang Gong <zhigang.gong@gmail.com> 88033Date: Sun May 22 23:45:51 2011 +0800 88034 88035 glamor: Don't write back read only pixmap to fbo. 88036 88037 For those pixmap which has valid fbo and opened as GLAMOR_ACCESS_RO 88038 mode, we don't need to upload the texture back when calling the 88039 glamor_finish_access(). This will get about 10% performance gain. 88040 88041commit 1495ba9e64633476508febb01ec2e9594a7b466b 88042Author: Zhigang Gong <zhigang.gong@gmail.com> 88043Date: Sun May 22 17:51:27 2011 +0800 88044 88045 glamor: Use buffer object as much as possible. 88046 88047 Change the row length of 1bit color depth pixmap to the actual stride. 88048 The previous implementation use the width as its stride which is not 88049 good. As it will waste 8 times of space and also bring some non-unify 88050 code path. With this commit, we can merge those 1bit or other color 88051 depth to almost one code path. And we will use pixel buffer object 88052 as much as possible due to performance issue. By default, some mesa 88053 hardware driver will fallback to software rasterization when use 88054 glReadPixels on a non-buffer-object frame buffer. This change will 88055 get about 4x times performance improvemention when we use y-inverted 88056 glamor or the driver support hardware y-flipped blitting. 88057 88058commit 529c38a4601e8f0712773bf29a48cddfa7a5edc4 88059Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88060Date: Thu May 19 18:03:12 2011 +0800 88061 88062 glamor-ddx: Remove debug message when moving cursor. 88063 88064commit cd43b1ea831eac11bc6ad0c3b1e040bb968b9861 88065Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88066Date: Thu May 19 18:02:19 2011 +0800 88067 88068 glamor: Add fallback code path for glamor_fill. 88069 88070commit 8593f22fb8fc5e0d0f406d9c94cca22347505fd6 88071Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88072Date: Thu May 19 17:59:19 2011 +0800 88073 88074 glamor: glamor_set_alu should enable GL_COLOR_LOGIC_OP. 88075 88076 GL_COLOR_OP seems not supported in current MESA. 88077 88078commit f871d174a861e7c3d2b8f4d9a3f10c38a5120606 88079Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88080Date: Thu May 19 11:40:38 2011 +0800 88081 88082 glamor: Switch to software fb for too large pixmap. 88083 88084 If pixmap's size exceeds the limitation of the MESA library, the 88085 rendering will fail. So we switch to software fb if it is the case. 88086 Add one new element for pixmap private structure to indicate whehter 88087 we are a software fb type or a opengl type. 88088 88089commit 74ca45e7d0ed6654204189793261ef65f213bb2e 88090Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88091Date: Wed May 18 16:48:06 2011 +0800 88092 88093 glamor-ddx: Fixed one bug when a client reset the connection. 88094 88095 This commit fixed two bugs when one client reset the connection. 88096 The first is that we should reopen the graphic device when the previous 88097 node was closed during the screen closing. The second one is we should 88098 call glamor_close_screen (not the ddx version) prior to call 88099 eglTerminate(). As eglTerminate will release the share library. And 88100 the glamor_close_screen may still need to call openGL APIs and thus 88101 will hit segfault. And renamed the ddx functions to avoid naming 88102 conflications with the glamor functions. 88103 88104commit c97d4533f29e77d80c076deff9ad0f218eb2e8f4 88105Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88106Date: Thu May 12 14:49:08 2011 +0800 88107 88108 glamor: Silent compilation warnings due to some deprecated APIs. 88109 88110 those xcalloc/xfree/xalloc/XNFprintf/... are deprecated. Replace 88111 then with the new one. And fix some other minor problems. 88112 88113commit e3295d4106ac5b62f63e32dcb24a6094194cb1a8 88114Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88115Date: Thu May 12 10:22:07 2011 +0800 88116 88117 glamor-ddx: Move the cursor EGL image to crtc strcture. 88118 88119 Cursor is a per crtc resource. And this commit also fix the cursor 88120 initialization regard to the latest mesa EGL code. Now hardware 88121 cursor works fine. 88122 88123commit 43280372686177603111b175e92e1c88ad4ccdee 88124Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88125Date: Thu May 12 01:42:05 2011 +0800 88126 88127 glamor-ddx: Fix a compiling error. Need revisit. 88128 88129 GC is redefined in the X11/Xlib.h and include/gcstruct.h which is 88130 a xorg header file. Just use a macro to simply avoid the conflict. 88131 Need revisit latter to find a correct way to fix this problem. 88132 88133commit 18a52e23882e4664fdb2de4958c4f9c97e16068c 88134Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88135Date: Thu May 12 01:37:52 2011 +0800 88136 88137 glamor-ddx: Migrate the code to latest mesa library. 88138 88139 Use eglCreateDRMImageMESA to create surfaceless image. And then 88140 export the drm buffer back which can be used to create the frame 88141 buffer. 88142 88143commit eb3487a448ff0efa46079323821d65d7c3e4d872 88144Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88145Date: Thu May 12 01:08:38 2011 +0800 88146 88147 glamor: Add new feature which is to flip output on y axis. 88148 88149 Due to the coordinate system on EGL is different from FBO 88150 object. To support EGL surface well, we add this new feature. 88151 When calling glamor_init from EGL ddx driver, it should use 88152 the new flag GLAMOR_INVERTED_Y_AXIS. 88153 88154commit 6dae8dc7ea6e7add3c6fda30773f264904ef8df5 88155Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88156Date: Wed May 11 15:57:05 2011 +0800 88157 88158 glamor: Add glamor-ddx driver to the build tree. 88159 88160 Correct the linking parameters and add dependency to the libglamor.a. 88161 88162commit 26ff612171d53baef08078a852875d5ddf1804ed 88163Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88164Date: Wed May 11 15:54:50 2011 +0800 88165 88166 glamor: Resolved merge conflictions with Kristian's glamor-ddx patch. 88167 88168commit 49bf0e301e7317a02e8b34ec2e290fcfda520e7e 88169Author: Kristian Høgsberg <krh@bitplanet.net> 88170Date: Fri Mar 5 14:53:35 2010 -0500 88171 88172 glamor ddx driver using EGL on KMS 88173 88174commit 47b6531273e3052b653194099a49ca11d80b12df 88175Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88176Date: Wed Apr 13 15:50:26 2011 +0800 88177 88178 glamor: Fixed one segfault bug when close screen. 88179 88180 move the original glamor_fini to glamor_close_screen. And wrap the CloseScreen 88181 with glamor_close_screen, Then we can do some thing before call the underlying 88182 CloseScreen(). 88183 The root cause is that glamor_fini will be called after the ->CloseScreen(). 88184 This may trigger a segmentation fault at 88185 glamor_unrealize_glyph_caches() at calling into FreePicture(). 88186 88187commit 0d9e8db8d031d22b02df2c9b226287b1ca489f5a 88188Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88189Date: Fri Apr 8 15:30:45 2011 +0800 88190 88191 glamor: Fixed one linking error. 88192 88193 As current glamor implementation depends on the glx library in the 88194 mesa package which is conflict with the version in xorg. We have to 88195 --disable-glx when build Xephyr. But this leads to the linking error 88196 here. We comment out the calling to ephyrHijackGLXExtension() now. 88197 Need revisit latter. 88198 88199commit b3577a1c853d870f4d583a35d27012acb9099ef4 88200Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88201Date: Fri Apr 8 15:24:55 2011 +0800 88202 88203 glamor: Merged with latest xserver. 88204 88205 Merged with latest xserver, the major change is due to the API change of 88206 the management of devPrivates. 88207 88208commit 3105fe9f6491f37a1f82a9e6f03f4efaed6c7e92 88209Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88210Date: Fri Apr 8 15:22:52 2011 +0800 88211 88212 glamor: Call glamor_composite_rects only when enable RENDER. 88213 88214commit f9843c7a35975b178eec9812d21e1f02a2120928 88215Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88216Date: Fri Apr 8 14:53:15 2011 +0800 88217 88218 glamor: Fixed one conflict data type(XID) bug. 88219 88220 We should include the dix-config.h for all the glamor files. Otherwise 88221 the XID type maybe inconsisitent in different files in 64bit machine. 88222 The root cause is this macro "#define _XSERVER64 1" should be included 88223 in all files refer to the data type "XID" which is originally defined 88224 in X.h. If _XSERVER64 is defined as 1, then XID is defined as CARD32 88225 which is a 32bit integer. If _XSERVER64 is not defined as 1 then XID 88226 is "unsigned long". In a 32bit machine, "unsigned long" should be 88227 identical to CARD32. But in a 64bit machine, they are different. 88228 88229commit e194740c9413bbfefee6506a734e0c82c091dfa3 88230Author: Zhigang Gong <zhigang.gong@linux.intel.com> 88231Date: Fri Apr 8 14:49:57 2011 +0800 88232 88233 glamor: Fixed some compiling and linking errors. 88234 88235 Change the inline functions to static type, otherwise when linking it 88236 complains multiple definitions of those functions. 88237 88238commit 8cfcc614032320bb6a2eca2c61baeebf388d22ea 88239Author: Eric Anholt <eric@anholt.net> 88240Date: Fri Feb 19 07:14:13 2010 -0800 88241 88242 glamor: Add support for using EXT_framebuffer_blit to do CopyArea. 88243 88244commit d2da9d1c22d25d1f86ecd83c15ca9909fab8610a 88245Author: Eric Anholt <eric@anholt.net> 88246Date: Fri Feb 19 07:36:28 2010 -0800 88247 88248 glamor: Reduce the noise from wide lines "fallback". 88249 88250commit 955ccfbc34f370c80f7189ac5b0240e859e00b5f 88251Author: Eric Anholt <eric@anholt.net> 88252Date: Fri Feb 19 07:52:50 2010 -0800 88253 88254 glamor: Fix the segfault at screen fini. 88255 88256commit 003dee4c82f8af9e7749876c736018410f347440 88257Author: Eric Anholt <eric@anholt.net> 88258Date: Fri Feb 19 06:10:11 2010 -0800 88259 88260 glamor: Replace the immediate mode in glamor_fill() with glDrawArrays(). 88261 88262commit 60775e21e3532452891e1b52589f7d5d7ef04b40 88263Author: Eric Anholt <eric@anholt.net> 88264Date: Thu Feb 18 18:11:54 2010 -0800 88265 88266 glamor: Use a VBO to accumulate multiple glyph quads at once. 88267 88268 This increases us from 23000 to 27000/sec on rgb24text. 88269 88270commit 8ce312e61952e9f2193c28ac6124eff30f3a122c 88271Author: Eric Anholt <eric@anholt.net> 88272Date: Thu Feb 18 14:49:06 2010 -0800 88273 88274 glamor: Use glamor_composite_with_shader once per group of glyphs. 88275 88276 This shaves CPU time in GL setup. Performance of rgb24text went from 88277 18400/sec to 23500/sec. 88278 88279commit 6ce05e0b28052e5206694228218322f0dd3a6b00 88280Author: Eric Anholt <eric@anholt.net> 88281Date: Thu Feb 18 14:44:20 2010 -0800 88282 88283 glamor: Convert the shaders path to handling glamor_composite_rect_t. 88284 88285commit 858ce0c1928c199435c9c2627f84c37c7ca7a38a 88286Author: Eric Anholt <eric@anholt.net> 88287Date: Thu Feb 18 14:02:48 2010 -0800 88288 88289 glamor: Add support for component alpha rendering. 88290 88291 Brings x11perf -rgb24text from 230/sec to 18400/sec 88292 88293commit d07fc66a056dd4eab2dac82b3784c482071d1915 88294Author: Eric Anholt <eric@anholt.net> 88295Date: Thu Feb 18 09:23:22 2010 -0800 88296 88297 glamor: Rework shader setup to make extending the IN types easier. 88298 88299commit 8cefa287ddb4ed4ad178e751b35bb93b4a44e0ab 88300Author: Eric Anholt <eric@anholt.net> 88301Date: Wed Feb 17 16:00:22 2010 -0800 88302 88303 glamor: Add a little mechanism for only printing fallbacks when they happen. 88304 88305 Sometimes we want to try a couple of different methods for 88306 accelerating. If one of them says "no" and the other says "yes", 88307 don't spam the log about the "no." 88308 88309commit 5f5c35b56d3568f72e2305e74ed8457e0a787a3b 88310Author: Eric Anholt <eric@anholt.net> 88311Date: Wed Feb 17 15:02:17 2010 -0800 88312 88313 glamor: Fix up the fallback message for no texture present on compositing. 88314 88315commit 2fa95725d845e5bf8a41ac776267be0d55d58004 88316Author: Eric Anholt <eric@anholt.net> 88317Date: Wed Feb 17 13:33:48 2010 -0800 88318 88319 glamor: Add support for a1 composite sources. 88320 88321 They're stored just like a8, but the values are set to either 0.0 or 88322 1.0. Because they're a8 with only two legal values, we can't use them 88323 as destinations, but nobody's rendering to a1 dests anyway (we hope). 88324 88325commit 9bcbcbf8c28e945fa5c4f4ad1a772b787618455f 88326Author: Eric Anholt <eric@anholt.net> 88327Date: Wed Feb 17 13:30:20 2010 -0800 88328 88329 glamor: Fix and enable ZPixmap PutImage acceleration. 88330 88331commit 2ba634fab9b03995629497efdbf5305394011e4a 88332Author: Eric Anholt <eric@anholt.net> 88333Date: Thu Feb 11 17:32:27 2010 -0800 88334 88335 glamor: Fix render source transforms. 88336 88337 Fixes (except for small bit differences) cairo source-pattern. 88338 88339commit b1f67a5082420bbff141733833905f8ac95fe983 88340Author: Eric Anholt <eric@anholt.net> 88341Date: Thu Feb 11 17:15:05 2010 -0800 88342 88343 glamor: Fix and enable glamor_get_spans(). 88344 88345 This makes running the cairo test suite almost tolerable. 88346 88347commit f88d76cf7e05cf885d80c0b91f37f1b30675e928 88348Author: Eric Anholt <eric@anholt.net> 88349Date: Thu Feb 11 16:51:13 2010 -0800 88350 88351 glamor: Enable glamor_fill_spans(). 88352 88353 x11perf -wline100 performance goes from 8.8/sec to 111/sec. 88354 88355commit a63df0c504be3be0e969f5080a6593a55cb87246 88356Author: Eric Anholt <eric@anholt.net> 88357Date: Thu Feb 11 15:00:32 2010 -0800 88358 88359 glamor: Fix up the wide/non-solid lines fallback. 88360 88361commit d61a13cf3f0a3028fbcf41310a4f5ecaa8546961 88362Author: Eric Anholt <eric@anholt.net> 88363Date: Thu Feb 11 14:52:04 2010 -0800 88364 88365 glamor: Fix and enable tile fill acceleration (aka the root weave). 88366 88367commit 86a206525350b7e82ab0e432ad0fb8858960857d 88368Author: Eric Anholt <eric@anholt.net> 88369Date: Thu Feb 11 13:40:28 2010 -0800 88370 88371 glamor: Fix off-by-one in CopyPixels CopyArea path. 88372 88373 Fixes window dragging in uncomposited metacity. 88374 88375commit be82a0624207b5f367b97c25ac7d7dbd8a518597 88376Author: Eric Anholt <eric@anholt.net> 88377Date: Thu Feb 11 13:21:47 2010 -0800 88378 88379 glamor: Fix screen_x/screen_y handling for compositing. 88380 88381 It's not an offset from pixmap coords to composited pixmap coords, 88382 it's an offset from screen-relative window drawable coords to 88383 composited pixmap coords. 88384 88385commit 5360b1e8bbae571b120999e7eec7cd2826601497 88386Author: Eric Anholt <eric@anholt.net> 88387Date: Thu Feb 11 09:47:38 2010 -0800 88388 88389 glamor: Implement glCopyPixels based src == dest CopyArea acceleration 88390 88391commit 0565c1d789ed82c533f860713d66b4941c299f9c 88392Author: Eric Anholt <eric@anholt.net> 88393Date: Tue Feb 9 22:35:04 2010 -0800 88394 88395 glamor: Add support for accel of x8r8g8b8 source pictures. 88396 88397 There's a limitation still for RepeatNone, but this fixes a bunch of 88398 fallbacks for gnome-terminal. 88399 88400commit e6bf50573650f03c8130b7783485b24e98e15c79 88401Author: Eric Anholt <eric@anholt.net> 88402Date: Thu Feb 11 09:23:09 2010 -0800 88403 88404 glamor: Set active texture on glamor_copy_n_to_n setup. 88405 88406 Fixes failure in rendercheck -t blend -o src 88407 88408commit be64167fea3c74447ed3c5116b97473676c25b29 88409Author: Eric Anholt <eric@anholt.net> 88410Date: Thu Feb 11 09:20:50 2010 -0800 88411 88412 glamor: Don't try to CopyArea from a Solid source picture. 88413 88414 Fixes failure with rendercheck. 88415 88416commit 126fc09cb5bf8c38b73f1c2cd2a0ef588c3265e8 88417Author: Eric Anholt <eric@anholt.net> 88418Date: Tue Feb 9 22:35:04 2010 -0800 88419 88420 glamor: Rework the Render shader setup to be easily modified, like cairo-gl. 88421 88422commit f4a3194837e640997c0c3a775f48a49800e213c6 88423Author: Eric Anholt <eric@anholt.net> 88424Date: Tue Feb 9 23:17:09 2010 -0600 88425 88426 glamor: Add the glyph cache from UXA (de-camelCased). 88427 88428 This doesn't yet have an optimized glamor_composite_rects() 88429 implementation, but it does triple the speed of x11perf -aa10text. 88430 88431commit 7e6432e7b94f7f99b257da8de53573108d30ae22 88432Author: Eric Anholt <eric@anholt.net> 88433Date: Tue Feb 9 22:54:40 2010 -0600 88434 88435 glamor: Fix up and enable accelerated composite. 88436 88437commit a0b589e90a885cf20b349b95df31bb0823a1514b 88438Author: Eric Anholt <eric@anholt.net> 88439Date: Tue Feb 9 22:49:27 2010 -0600 88440 88441 glamor: Restore planemask to all-on when finishing a fill. 88442 88443commit d8c2662bf452a82c8fcbdd1c95a40ee34d2c32ca 88444Author: Eric Anholt <eric@anholt.net> 88445Date: Tue Feb 9 22:48:28 2010 -0600 88446 88447 glamor: Set active texture in finishaccess drawing. 88448 88449commit 4f398b29dd42260d2c02e8e795546fd0623397b1 88450Author: Eric Anholt <eric@anholt.net> 88451Date: Tue Feb 9 21:56:39 2010 -0600 88452 88453 glamor: Align stride of fallback pixmap data to 32 bits. 88454 88455 fb/pixman demand this alignment, and all sorts of things go badly otherwise. 88456 Fixes piglit x11-8bpp-7x8-draw. 88457 88458commit 15e58b5ffb427c6c6f5172ccd72758047b98024a 88459Author: Eric Anholt <eric@anholt.net> 88460Date: Mon Feb 8 22:05:51 2010 +0100 88461 88462 glamor: Split the copy path out into its own function. 88463 88464commit ad67299fa2afc8b42432d71d14163a36e013fef7 88465Author: Eric Anholt <eric@anholt.net> 88466Date: Mon Feb 8 21:52:35 2010 +0100 88467 88468 glamor: Provide more information about the operands to fallback composites. 88469 88470commit 4811e428a9206ef59487fb0d3fab160a19845d46 88471Author: Eric Anholt <eric@anholt.net> 88472Date: Mon Feb 8 21:46:49 2010 +0100 88473 88474 glamor: Pull in UXA code for dumping some composites down to copy_n_to_n. 88475 88476 Window dragging with metacity+gnome-terminal+xcompmgr is almost credible. 88477 88478commit 35847c578e9c35a3c90da792c34c1cb0cc536261 88479Author: Eric Anholt <eric@anholt.net> 88480Date: Mon Feb 8 21:24:15 2010 +0100 88481 88482 glamor: Add acceleration for copyarea not from the screen. 88483 88484commit 647b9fb49a5bb636c9b0da6b708083328238543a 88485Author: Eric Anholt <eric@anholt.net> 88486Date: Mon Feb 8 20:12:44 2010 +0100 88487 88488 glamor: Add CopyWindow implementation so it doesn't crash. 88489 88490commit 95d4a5a6ab58625d8205461157263bfb635ccd1a 88491Author: Eric Anholt <eric@anholt.net> 88492Date: Mon Feb 8 19:38:00 2010 +0100 88493 88494 glamor: Enable fill acceleration and flip opposite for pixmaps too. 88495 88496 Here's my theory for the flipping: It doesn't really matter which 88497 orientation we store the pixmaps if we don't scan them out. We have 88498 to flip coordinates for the window system framebuffer. Doing so for 88499 everything else makes things consistent. I'm not sure how this will 88500 interact with future GLX integration, though. 88501 88502commit 5332547a0ac6ae6b9f1a6c6428e6fdbdf303108f 88503Author: Eric Anholt <eric@anholt.net> 88504Date: Mon Feb 8 17:14:34 2010 +0100 88505 88506 glamor: Fix prepare_access on 1bpp data. 88507 88508 Apparently I don't have anything hitting this path, so I'm not sure if 88509 it's good or not. 88510 88511commit c3c3a6349766b4fdee080ff203fb548725e42e50 88512Author: Eric Anholt <eric@anholt.net> 88513Date: Mon Feb 8 15:34:45 2010 +0100 88514 88515 glamor: Use the pixmap stride in fallbacks instead of trying to guess it. 88516 88517 Mostly fixes gnome-terminal text. 88518 88519commit 45de3d24b44aa3feabdda0e377d637f498e03eb0 88520Author: Eric Anholt <eric@anholt.net> 88521Date: Mon Feb 8 14:40:06 2010 +0100 88522 88523 glamor: Replace the glDrawPixels in glamor_finish_access with GLSL. 88524 88525 Root weave displays. \o/ 88526 88527commit d8d3fa10346302384b8e62967c5a67893ed1c67b 88528Author: Eric Anholt <eric@anholt.net> 88529Date: Mon Feb 8 14:53:41 2010 +0100 88530 88531 glamor: All the fallbacks in the world. 88532 88533 Bringup is really not flying when I can't see anything. So dump back 88534 to all software so I can turn on a bit at a time. 88535 88536commit 745502af962a2bb475c0c667c3091c2c080771dc 88537Author: Eric Anholt <eric@anholt.net> 88538Date: Mon Feb 8 13:27:47 2010 +0100 88539 88540 glamor: Add check for EXT_bgra which we rely on. 88541 88542commit 6ce378f11ff396f350cbaa280086f6f9600da29e 88543Author: Eric Anholt <eric@anholt.net> 88544Date: Mon Feb 8 13:15:38 2010 +0100 88545 88546 glamor: Add fallback support for glamor_get_spans(). 88547 88548commit ec526eab8f02e0bcce859a1b7ebaf6c5bccc5e87 88549Author: Eric Anholt <eric@anholt.net> 88550Date: Mon Feb 8 13:05:13 2010 +0100 88551 88552 ephyr: Hook the glamor into damage and draw into the backbuffer. 88553 88554 This should avoid a bunch of absurdity with GLX front buffer handling, 88555 fix exposes, and improve performance. For now we're copying the whole 88556 buffer while glamor is developed. 88557 88558commit b5087ff9b1e5495c742e581f586c055de878dd4b 88559Author: Eric Anholt <eric@anholt.net> 88560Date: Mon Feb 8 12:47:59 2010 +0100 88561 88562 glamor: Fix the row length of 1bpp prepare_access. 88563 88564commit e93070affdf64c47be02684e7902a61e7ae9fa33 88565Author: Eric Anholt <eric@anholt.net> 88566Date: Mon Feb 8 12:44:03 2010 +0100 88567 88568 glamor: Fix up the access flags for glamor_prepare_access(). 88569 88570commit 0e56c182c3b06f5b02d357e2b8e930a580bd0274 88571Author: Eric Anholt <eric@anholt.net> 88572Date: Mon Feb 8 12:16:30 2010 +0100 88573 88574 glamor: Add fallback support to glamor_put_image(). 88575 88576commit 22cad98975091ed724a37ec940946cd2899ebcbf 88577Author: Eric Anholt <eric@anholt.net> 88578Date: Mon Feb 8 12:13:15 2010 +0100 88579 88580 glamor: Add fallback support to glamor_set_spans(). 88581 88582commit 55dac9b42e18dcc1586770d31ee265178e68b744 88583Author: Eric Anholt <eric@anholt.net> 88584Date: Mon Feb 8 12:09:52 2010 +0100 88585 88586 glamor: Add glamor_prepare_access_gc() from UXA. 88587 88588commit 72a757ba00232fe8b42b977f5210fa119495b05e 88589Author: Eric Anholt <eric@anholt.net> 88590Date: Mon Feb 8 11:58:52 2010 +0100 88591 88592 glamor: Use GL_MESA_pack_invert to avoid complexity in prepare_access. 88593 88594commit e9d4794cd0bc6c8790204405fe09f968edfa409d 88595Author: Eric Anholt <eric@anholt.net> 88596Date: Mon Feb 8 11:23:14 2010 +0100 88597 88598 glamor: Add fallbacks for Render. 88599 88600commit de675893b7ce2880289cce13be6190d55f4fc29c 88601Author: Eric Anholt <eric@anholt.net> 88602Date: Mon Feb 8 10:16:21 2010 +0100 88603 88604 glamor: Add fallbacks for glamor_copy_area(). 88605 88606commit ca58607a44c310ef369a097a06be83aa2a9e1c5b 88607Author: Eric Anholt <eric@anholt.net> 88608Date: Mon Feb 8 09:58:55 2010 +0100 88609 88610 glamor: Move glamor_poly_lines to a separate file. 88611 88612commit 800fd4f8494ad6f0984073406b3dec584ba63648 88613Author: Eric Anholt <eric@anholt.net> 88614Date: Tue Sep 22 15:41:18 2009 -0700 88615 88616 glamor: Fix the type for copyarea. 88617 88618commit c4343dfa0a2b40d46d3feb14f1df9fae0fd4a214 88619Author: Eric Anholt <eric@anholt.net> 88620Date: Tue Sep 22 12:08:19 2009 -0700 88621 88622 glamor: Add prepare/finishaccess code based on UXA. 88623 88624commit f17473cdd5f93a2bf3a7b6a14bd5acd965e8e9f0 88625Author: Eric Anholt <eric@anholt.net> 88626Date: Tue Sep 22 10:55:04 2009 -0700 88627 88628 glamor: Fix pixmap private getter in copyarea. 88629 88630commit 5915b4c0cfcfb25a68b77b29a4d7b39b7f6b4822 88631Author: Eric Anholt <eric@anholt.net> 88632Date: Fri Sep 4 18:43:52 2009 -0700 88633 88634 glamor: Add support for solid source pictures. 88635 88636commit 1159ebb30b7530c2f4612109306cf85a594dd530 88637Author: Eric Anholt <eric@anholt.net> 88638Date: Fri Sep 4 18:43:02 2009 -0700 88639 88640 glamor: Add untested copyarea implementation 88641 88642commit 936385142c12d074dcb603490517ac095840fa3f 88643Author: Eric Anholt <eric@anholt.net> 88644Date: Tue Aug 25 17:24:30 2009 -0700 88645 88646 glamor: Accelerate PolyFillRect using glamor_fill instead of spans. 88647 88648commit 6dacc9b08c2af51c1826c20ccf12bb2c73856f25 88649Author: Eric Anholt <eric@anholt.net> 88650Date: Tue Aug 25 17:01:43 2009 -0700 88651 88652 glamor: Add untested PutImage 1bpp XYPixmap support. 88653 88654commit 854e9bd20a2d0f8ed636d4fba0ddfa4c71d54667 88655Author: Eric Anholt <eric@anholt.net> 88656Date: Tue Aug 25 16:56:50 2009 -0700 88657 88658 glamor: Add untested PutImage XYBitmap support. 88659 88660commit 8a53566acb5870816807425dc805df888fa42792 88661Author: Eric Anholt <eric@anholt.net> 88662Date: Tue Aug 25 13:06:11 2009 -0700 88663 88664 glamor: make the polylines complaint a little more useful. 88665 88666commit aa133069745fc59bb2f212b0816add9bae40c376 88667Author: Eric Anholt <eric@anholt.net> 88668Date: Tue Aug 25 12:10:32 2009 -0700 88669 88670 glamor: Start adding render acceleration support. 88671 88672 This brings in idr's glu3 code. We'll probably want to move to linking to 88673 it as a library, once an ABI-stable release is out. 88674 88675commit 4a51cc0440c8d026fd8c82b40cfb188c6ea4a9f0 88676Author: Eric Anholt <eric@anholt.net> 88677Date: Mon Aug 24 09:14:20 2009 -0700 88678 88679 glamor: Add clipping to setspans. 88680 88681commit 5fadea5d9c02670c35ddf3e0f12f9d355b412103 88682Author: Eric Anholt <eric@anholt.net> 88683Date: Mon Aug 24 09:10:14 2009 -0700 88684 88685 glamor: Fix some screen_xy offsets to be the right way around. 88686 88687commit 44e4599b83cc330a7c9edd7e18266f222a69418a 88688Author: Eric Anholt <eric@anholt.net> 88689Date: Mon Aug 24 08:59:52 2009 -0700 88690 88691 glamor: Move planemask to a function in case I decide to fill it in. 88692 88693commit 882411bef1091f97f61ae27a549970d87eef9cab 88694Author: Eric Anholt <eric@anholt.net> 88695Date: Mon Aug 24 08:29:30 2009 -0700 88696 88697 glamor: add alu support to setspans. 88698 88699commit c70ce72fd74354a0b9731dc8d4451f6aaf951117 88700Author: Eric Anholt <eric@anholt.net> 88701Date: Mon Aug 24 08:24:54 2009 -0700 88702 88703 glamor: Give setspans the same format support as getspans. 88704 88705commit d9eef95c74b7b04a3398b4cfc1a81b480838b25e 88706Author: Eric Anholt <eric@anholt.net> 88707Date: Mon Aug 24 08:17:06 2009 -0700 88708 88709 glamor: Fix bad fallthrough in getspans (accessing invalid memory). 88710 88711commit 0360ba361a67f6cfa5c9c65590ed8440a153c16e 88712Author: Eric Anholt <eric@anholt.net> 88713Date: Mon Aug 24 08:06:57 2009 -0700 88714 88715 glamor: Fix memory leak in getspans. 88716 88717commit cdb1fe6d9682616b1def52fa38b2dae623e2503f 88718Author: Eric Anholt <eric@anholt.net> 88719Date: Mon Aug 24 08:03:17 2009 -0700 88720 88721 glamor: Move setspans to a separate file. 88722 88723commit cfb8dea815086b9aafd27bd314e02739e04dfc7d 88724Author: Eric Anholt <eric@anholt.net> 88725Date: Mon Aug 24 07:22:14 2009 -0700 88726 88727 glamor: Hook up miGetImage to fix some invalid accesses. 88728 88729commit e3c02c09c125bb5def1a70a1264ef34b18182afa 88730Author: Eric Anholt <eric@anholt.net> 88731Date: Mon Aug 24 07:00:32 2009 -0700 88732 88733 glamor: Add ALU support to solid, tile, and putimage. 88734 88735commit 2b657d91d80a0a4d9fa2e390255ff87f51426b7c 88736Author: Eric Anholt <eric@anholt.net> 88737Date: Sat Aug 22 15:54:24 2009 -0700 88738 88739 glamor: Add untested putimage support. 88740 88741commit f66e5c4145d003247447ac1bf93f9737c6eaa27c 88742Author: Eric Anholt <eric@anholt.net> 88743Date: Fri Aug 21 10:05:32 2009 -0700 88744 88745 glamor: Add 8bpp to get/setspans. 88746 88747commit 93ad09144213fe6b9ce8bea7ebfcc2d60887810b 88748Author: Eric Anholt <eric@anholt.net> 88749Date: Fri Aug 21 10:02:02 2009 -0700 88750 88751 glamor: Add trapezoids code. 88752 88753 This fixes segfaults on starting gnome-terminal. 88754 88755commit 60ca0260f23cac6577aa85aacb0e5644d5f1b349 88756Author: Eric Anholt <eric@anholt.net> 88757Date: Thu Aug 20 17:17:43 2009 -0700 88758 88759 glamor: Add more solid_fail_regions for software fallbacks. 88760 88761commit 7500ee2b0215932ecb6688f301fb4e6f080f8a66 88762Author: Eric Anholt <eric@anholt.net> 88763Date: Thu Aug 20 17:05:00 2009 -0700 88764 88765 glamor: Fill in 1 and 24-bit getspans. 88766 88767commit f6cffbb2ba66948ad5cc1cbe20e12de204de175d 88768Author: Eric Anholt <eric@anholt.net> 88769Date: Thu Aug 20 16:43:22 2009 -0700 88770 88771 glamor: Add stub Composite support. 88772 88773commit 2fa26c4958168064da95361db78ccc518c8f3ed8 88774Author: Eric Anholt <eric@anholt.net> 88775Date: Thu Aug 20 16:43:03 2009 -0700 88776 88777 glamor: Free resources when destroying pixmaps. 88778 88779commit 16c3b929dd4cae6b40e847256db6ff32e63e9029 88780Author: Eric Anholt <eric@anholt.net> 88781Date: Thu Aug 20 16:06:42 2009 -0700 88782 88783 glamor: Add untested support for tile filling. 88784 88785commit 8016135ec7616530e64d0019ce0401e62b25e483 88786Author: Eric Anholt <eric@anholt.net> 88787Date: Thu Aug 20 15:37:22 2009 -0700 88788 88789 glamor: Fill out glamor_get_color_4f_from_pixel() a bit. 88790 88791commit 519103565c1b182014dea93c58917bf7e5a35fcb 88792Author: Eric Anholt <eric@anholt.net> 88793Date: Thu Aug 20 15:28:13 2009 -0700 88794 88795 glamor: Create FBOs for pixmaps. 88796 88797commit f92d1478e984d57806731886820f47e1258b5115 88798Author: Eric Anholt <eric@anholt.net> 88799Date: Thu Aug 20 14:45:47 2009 -0700 88800 88801 glamor: Maybe fix up the format/type for setspans. 88802 88803commit 4f139db92fb283792adae8d39a49a083e4382fb5 88804Author: Eric Anholt <eric@anholt.net> 88805Date: Thu Aug 20 14:43:43 2009 -0700 88806 88807 glamor: Add getspans implementation. 88808 88809commit f1dbed5456e5a608425530e0dde2dc617a7c6115 88810Author: Eric Anholt <eric@anholt.net> 88811Date: Thu Aug 20 12:32:53 2009 -0700 88812 88813 glamor: Move to using shader objects. 88814 88815commit fbbdd788cbc822fc8a2d2726435bfa39e69a0e85 88816Author: Eric Anholt <eric@anholt.net> 88817Date: Thu Aug 20 12:36:33 2009 -0700 88818 88819 glamor: Fix up DrawPixels arguments to bear some relation to reality. 88820 88821commit 74f262bcdb3d8ae769ef8975cf4a4b2ebaeb6b39 88822Author: Eric Anholt <eric@anholt.net> 88823Date: Thu Aug 20 11:39:05 2009 -0700 88824 88825 glamor: remove gratuitous flush. 88826 88827commit 399e00698976dd2e503b3a092bfe54ab24ea26d9 88828Author: Eric Anholt <eric@anholt.net> 88829Date: Thu Aug 20 11:37:37 2009 -0700 88830 88831 glamor: Fix up block/wakeup handler. 88832 88833 Prototypes are fixed, and we don't need to flush on wakeup. 88834 88835commit 2e35ceca65bbebf07c15ea2cff932a8749955c2b 88836Author: Eric Anholt <eric@anholt.net> 88837Date: Thu Aug 20 11:19:43 2009 -0700 88838 88839 ephyr: Use GLEW. 88840 88841 This will hide all sorts of extension stuff from us. 88842 88843commit 370df817ac0b346f1c7aa0b6e823c77797caaa9f 88844Author: Eric Anholt <eric@anholt.net> 88845Date: Thu Aug 20 09:35:24 2009 -0700 88846 88847 ephyr: Make sure a glamor-using window is created with a glx visual. 88848 88849commit 08097434ecc6479dc69a408523f398284168697e 88850Author: Eric Anholt <eric@anholt.net> 88851Date: Thu Aug 20 09:10:12 2009 -0700 88852 88853 add fill files, merge with first real attempt at rendering. 88854 88855commit e32efb02e5dff9124e8680f9a15d9d2b01db84c7 88856Author: Eric Anholt <eric@anholt.net> 88857Date: Sat Dec 20 14:48:18 2008 -0800 88858 88859 ephyr: merge the host glamor bits into hostx. 88860 88861commit df083fab84cf57fc1a2e3cd98572039bc9fe20de 88862Author: Eric Anholt <eric@anholt.net> 88863Date: Sat Dec 20 13:43:52 2008 -0800 88864 88865 glamor: first real attempt at rendering. 88866 88867commit 1b151f2e6aa581400b09220b87ae2d34a39aeb17 88868Author: Eric Anholt <eric@anholt.net> 88869Date: Sat Dec 20 13:48:39 2008 -0800 88870 88871 glamor: glFlush from the blockhandler so rendering happens. 88872 88873commit b530cdea4a8e05b1f487765b8dde42cde1a6fc98 88874Author: Eric Anholt <eric@anholt.net> 88875Date: Tue Dec 16 15:20:49 2008 -0800 88876 88877 glamor: Use a nicer struct initializer for gcops. 88878 88879commit 4d52ae7f2dca78d493e11375d764c978db5567ed 88880Author: Eric Anholt <eric@anholt.net> 88881Date: Tue Dec 16 15:14:31 2008 -0800 88882 88883 glamor: Start trying to hook up the rendering bits 88884 88885commit 41e3e580de95fca3095d015ec8febe1a5f47cca3 88886Author: Eric Anholt <eric@anholt.net> 88887Date: Thu Aug 20 10:49:41 2009 -0700 88888 88889 glamor: Use -lgl for Xephyr if GLX is unset. 88890 88891 I'm experimenting with that because of conflicts with GL dispatch tables. 88892 88893commit 8885b3b1ffb45ae16484151ce533683af526f509 88894Author: Eric Anholt <eric@anholt.net> 88895Date: Tue Dec 16 15:08:21 2008 -0800 88896 88897 Add exa/glamor to _DEPENDENCIES so the bin gets rebuilt with the libs. 88898 88899commit da6e0ffad2173d65f27feff36291f94e6aa06f88 88900Author: Eric Anholt <eric@anholt.net> 88901Date: Tue Dec 16 14:00:05 2008 -0800 88902 88903 More ephyr/glamor glue. 88904 88905commit 0ec823ce5b72bcf3be3923f2f23e5386feb0107b 88906Author: Eric Anholt <eric@anholt.net> 88907Date: Tue Dec 16 12:30:14 2008 -0800 88908 88909 Add build infrastructure for glamor. 88910 88911commit 84bb0207f6db433a8d387d933393357da87ca23e 88912Author: Gaetan Nadon <memsize@videotron.ca> 88913Date: Fri Sep 23 20:03:03 2011 -0400 88914 88915 dix and os: gitignore dix.O and os.O 88916 88917 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 88918 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 88919 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 88920 88921commit 3d41939d0349569dcfd2aef9b846c739d0ba0060 88922Author: Alan Coopersmith <alan.coopersmith@oracle.com> 88923Date: Thu Sep 22 17:41:38 2011 -0700 88924 88925 Unconditionally #include <stdint.h> 88926 88927 The more recent inclusions of this file haven't been checking for 88928 HAVE_STDINT_H, so might as well make the older ones consistent. 88929 88930 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 88931 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 88932 Reviewed-by: Jamey Sharp <jamey@minilop.net> 88933 88934commit 46d5ae5b551cc926331324073f2686b47e8aea07 88935Author: Gaetan Nadon <memsize@videotron.ca> 88936Date: Fri Sep 23 09:38:25 2011 -0400 88937 88938 dtrace: fix typo in title 88939 88940 The word provider should be capitalized. 88941 88942 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 88943 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 88944 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 88945 88946commit 6cae6224994e3252b3b7c7608f983ce73e0206fc 88947Author: Gaetan Nadon <memsize@videotron.ca> 88948Date: Fri Sep 23 09:38:24 2011 -0400 88949 88950 dtrace: use docbook copyright markup for copyright holder 88951 88952 No content change to copyright text. 88953 88954 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 88955 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 88956 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 88957 88958commit 58d3d46d6cc0d7a81db6bad61d158f3b9af8af71 88959Author: Alan Coopersmith <alan.coopersmith@oracle.com> 88960Date: Thu Sep 1 16:48:37 2011 -0700 88961 88962 Assign ids to more tags in Xserver-Dtrace.xml 88963 88964 Keeps xsltproc from adding random-number id tags to html output so that 88965 it's easier to see real changes vs. different random numbers. 88966 88967 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 88968 Reviewed-by: Matt Dew <marcoz@osource.org> 88969 88970commit 7fb4bef0394a5d09680985d34bce8252b61493cb 88971Merge: 98f494009 c90903b4f 88972Author: Keith Packard <keithp@keithp.com> 88973Date: Wed Sep 21 14:34:27 2011 -0700 88974 88975 Merge remote-tracking branch 'mattst88/for-keith' 88976 88977commit 98f4940093dc90c3366d3e4d3ddceacbaee299dc 88978Merge: b018b8153 c8eacae4f 88979Author: Keith Packard <keithp@keithp.com> 88980Date: Wed Sep 21 14:30:19 2011 -0700 88981 88982 Merge remote-tracking branch 'alanc/master' 88983 88984commit b018b815333b256b5e06dd6251b2e539f09f89fa 88985Merge: 03d032991 0f380a500 88986Author: Keith Packard <keithp@keithp.com> 88987Date: Wed Sep 21 14:17:14 2011 -0700 88988 88989 Merge remote-tracking branch 'jamey/reviewed' 88990 88991commit c90903b4f7a826ae6826a8dd0a901c8362500e46 88992Author: Peter Harris <pharris@opentext.com> 88993Date: Fri Sep 2 18:45:16 2011 -0400 88994 88995 xkb: add missing swaps for xkbGetDeviceInfoReply 88996 88997 Caught during review of e095369bf. 88998 88999 Signed-off-by: Peter Harris <pharris@opentext.com> 89000 Reviewed-by-by: Matt Turner <mattst88@gmail.com> 89001 Signed-off-by: Matt Turner <mattst88@gmail.com> 89002 89003commit d206d52f657cb63a0f6ba8b62d788c1812b57c81 89004Author: Matt Turner <mattst88@gmail.com> 89005Date: Tue Aug 16 19:19:24 2011 -0400 89006 89007 Use __builtin_constant_p to determine if we can use lswapl 89008 89009 If the address of the swapped memory location is known at compile time, 89010 we can check its alignment at no runtime cost and use lswapl instead. 89011 89012 text data bss dec hex filename 89013 before: 1872820 52136 78040 2002996 1e9034 hw/xfree86/Xorg 89014 after: 1864396 52136 78040 1994572 1e6f4c hw/xfree86/Xorg 89015 89016 bswap instructions: 131 -> 308 (used in lswapl) 89017 rol instructions: 943 -> 1174 (used in lswaps) 89018 89019 Reviewed-by: Peter Harris <pharris@opentext.com> 89020 Signed-off-by: Matt Turner <mattst88@gmail.com> 89021 89022commit 889b700e7760ced38bcf5f3aff2d31d3d9a058d7 89023Author: Matt Turner <mattst88@gmail.com> 89024Date: Tue Aug 16 19:12:21 2011 -0400 89025 89026 Use lswap{l,s} in cpswap{l,s} 89027 89028 Should be safe since cpswap isn't used on pointers. 89029 89030 text data bss dec hex filename 89031 before: 1875588 52136 78040 2005764 1e9b04 hw/xfree86/Xorg 89032 after: 1872820 52136 78040 2002996 1e9034 hw/xfree86/Xorg 89033 89034 bswap instructions: 5 -> 131 (used in lswapl) 89035 rol instructions: 811 -> 943 (used in lswaps) 89036 89037 Reviewed-by: Peter Harris <pharris@opentext.com> 89038 Signed-off-by: Matt Turner <mattst88@gmail.com> 89039 89040commit a2f0ff5f73db204a9d61e65148b28f6acc5121df 89041Author: Matt Turner <mattst88@gmail.com> 89042Date: Tue Aug 16 19:30:20 2011 -0400 89043 89044 Make lswap{l,s} inline functions 89045 89046 text data bss dec hex filename 89047 before: 1875668 52136 78040 2005844 1e9b54 hw/xfree86/Xorg 89048 after: 1875588 52136 78040 2005764 1e9b04 hw/xfree86/Xorg 89049 89050 Reviewed-by: Peter Harris <pharris@opentext.com> 89051 Signed-off-by: Matt Turner <mattst88@gmail.com> 89052 89053commit e8ff555b95baab66cc7d060c1e7f9fdd49d3802f 89054Author: Matt Turner <mattst88@gmail.com> 89055Date: Tue Aug 16 19:07:24 2011 -0400 89056 89057 Add type checking to swap macros 89058 89059 The original macros are retained (instead of replacing them with inline 89060 functions) because of implicit type promotion. That is, an int16 passed 89061 to an inline function taking int32 would be implicitly promoted to int32 89062 without a warning. 89063 89064 Reviewed-by: Peter Harris <pharris@opentext.com> 89065 Signed-off-by: Matt Turner <mattst88@gmail.com> 89066 89067commit 893e86a49e3e381cff48a9e86dc2d9b3d5431d95 89068Author: Matt Turner <mattst88@gmail.com> 89069Date: Tue Aug 16 19:03:26 2011 -0400 89070 89071 Introduce swap_uint{16,32} functions, used in swap{l,s} 89072 89073 Reviewed-by: Peter Harris <pharris@opentext.com> 89074 Signed-off-by: Matt Turner <mattst88@gmail.com> 89075 89076commit 54770c980cd2b91a8377f975a58ed69def5cfa42 89077Author: Matt Turner <mattst88@gmail.com> 89078Date: Tue Aug 16 16:59:07 2011 -0400 89079 89080 Cast char* buffers to swap functions 89081 89082 Reviewed-by: Peter Harris <pharris@opentext.com> 89083 Signed-off-by: Matt Turner <mattst88@gmail.com> 89084 89085commit 9edcae78c46286baff42e74bfe26f6ae4d00fe01 89086Author: Matt Turner <mattst88@gmail.com> 89087Date: Wed Sep 21 17:14:16 2011 -0400 89088 89089 Use correct swap{l,s} (or none at all for CARD8) 89090 89091 Swapping the wrong size was never caught because swap{l,s} are macros. 89092 89093 It's clear in the case of Xext/xres.c, that the author believed 89094 client_major/minor to be CARD16 from looking at the code in the first 89095 hunk. 89096 89097 v2: dmx.c fixes from Keith. 89098 89099 Reviewed-by: Peter Harris <pharris@opentext.com> 89100 Signed-off-by: Matt Turner <mattst88@gmail.com> 89101 89102commit 03d032991da21bf866ca30dcbb6b7f9f99df441a 89103Author: Aaron Plattner <aplattner@nvidia.com> 89104Date: Wed Sep 14 15:25:06 2011 -0700 89105 89106 fb: Rename wfbTriangles and wfbTrapezoids 89107 89108 These symbols were not renamed when they were added to libfb: 89109 89110 # nm -D libwfb.so | grep ' fb' 89111 0000000000028d00 T fbTrapezoids 89112 0000000000028d60 T fbTriangles 89113 89114 This causes corruption and/or crashes on wfb-ful drivers like nvidia: 89115 89116 Program received signal SIGABRT, Aborted. 89117 0x00007fd67f3a0405 in *__GI_raise (sig=<optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64 89118 64 ../nptl/sysdeps/unix/sysv/linux/raise.c: No such file or directory. 89119 in ../nptl/sysdeps/unix/sysv/linux/raise.c 89120 (gdb) bt 89121 #0 0x00007fd67f3a0405 in *__GI_raise (sig=<optimized out>) at ../nptl/sysdeps/unix/sysv/linux/raise.c:64 89122 #1 0x00007fd67f3a3680 in *__GI_abort () at abort.c:92 89123 #2 0x00007fd67f3995b1 in *__GI___assert_fail (assertion=0x7fd679ecb804 "key->initialized", file=<optimized out>, line=116, function=0x7fd679ecbbc0 "dixGetPrivateAddr") 89124 at assert.c:81 89125 #3 0x00007fd679ec55b6 in ?? () from /usr/lib/xorg/modules/libfb.so 89126 #4 0x00007fd679eca9ef in ?? () from /usr/lib/xorg/modules/libfb.so 89127 #5 0x00007fd679ecae20 in fbTriangles () from /usr/lib/xorg/modules/libfb.so 89128 #6 0x00007fd67a58fc55 in ?? () from /usr/lib/xorg/modules/drivers/nvidia_drv.so 89129 #7 0x00000000004f38d1 in ?? () 89130 #8 0x0000000000437ae9 in ?? () 89131 #9 0x0000000000426eaa in ?? () 89132 #10 0x00007fd67f38cead in __libc_start_main (main=<optimized out>, argc=<optimized out>, ubp_av=<optimized out>, init=<optimized out>, fini=<optimized out>, 89133 rtld_fini=<optimized out>, stack_end=0x7fff99860d78) at libc-start.c:228 89134 #11 0x000000000042719d in _start () 89135 89136 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 89137 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89138 Signed-off-by: Keith Packard <keithp@keithp.com> 89139 89140commit 2c7c520cfe0df30f4bc3adba59d9c62582823bf8 89141Author: Matt Turner <mattst88@gmail.com> 89142Date: Thu Aug 4 15:35:41 2011 -0400 89143 89144 Use internal temp variable for swap macros 89145 89146 Also, fix whitespace, mainly around 89147 swaps(&rep.sequenceNumber) 89148 89149 Reviewed-by: Peter Harris <pharris@opentext.com> 89150 Signed-off-by: Matt Turner <mattst88@gmail.com> 89151 89152commit c10bad3d3e8ff1b90014770fd470f9c67263e46f 89153Author: Matt Turner <mattst88@gmail.com> 89154Date: Wed Apr 27 19:29:06 2011 -0400 89155 89156 Silence printf format warnings in helper_exec.c 89157 89158 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89159 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89160 Signed-off-by: Matt Turner <mattst88@gmail.com> 89161 89162commit 9eab5b3443a1926a29385948acc6c5e0843465ea 89163Author: Matt Turner <mattst88@gmail.com> 89164Date: Wed Apr 27 18:25:27 2011 -0400 89165 89166 Replace Fabs() macro with fabs() function 89167 89168 gcc generates better code with fabs() anyway. 89169 89170 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89171 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 89172 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89173 Signed-off-by: Matt Turner <mattst88@gmail.com> 89174 89175commit 40a47bd628f525d2d8bd3ca76554089a6e9d2a1d 89176Author: Matt Turner <mattst88@gmail.com> 89177Date: Wed Apr 27 14:19:41 2011 -0400 89178 89179 Remove unnecessary #undefs of [f]abs and old comments 89180 89181 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89182 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 89183 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89184 Signed-off-by: Matt Turner <mattst88@gmail.com> 89185 89186commit 079a7585e44d25a0a9fb6662c6382dfbe8f6cd3c 89187Author: Gaetan Nadon <memsize@videotron.ca> 89188Date: Thu Sep 15 20:12:41 2011 -0400 89189 89190 devbook.am: maintenance update from docbook.am 89191 89192 The developer docs are generated from a subset of docbook.am 89193 which is sometimes updated. 89194 89195 The one difference for xserver is the embedded css style in the HEAD element. 89196 89197 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89198 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 89199 Signed-off-by: Keith Packard <keithp@keithp.com> 89200 89201commit 82c55b5a99a1bacf2bc11bbd9d58f6f2c6c07286 89202Author: Gaetan Nadon <memsize@videotron.ca> 89203Date: Thu Sep 15 20:12:40 2011 -0400 89204 89205 docbook.am: embed css styles inside the HTML HEAD element 89206 89207 Rather than referring to the external xorg.css stylesheet, embed the content 89208 of the file in the html output produced. This is accomplished by using 89209 version 1.10 of xorg-xhtml.xsl. 89210 89211 This makes the whole html docs tree much more relocatable. 89212 In addition, it eliminates xorg.css as a runtime file which makes 89213 xorg-sgml-doctools a build time only package. 89214 89215 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89216 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 89217 Signed-off-by: Keith Packard <keithp@keithp.com> 89218 89219commit 1602444e05cc59afe32c085c0cedb85e3268cd98 89220Author: Gaetan Nadon <memsize@videotron.ca> 89221Date: Thu Sep 15 20:12:39 2011 -0400 89222 89223 docbook.am: global maintenance update - entities, images and olinking 89224 89225 Adding support in libX11 for html chunking caused a reorg of docbook.am 89226 as well as the xorg-sgml-doctools masterdb for olinking. 89227 The parameter img.src.path is added for pdf images. 89228 A searchpath to the root builddir is added for local entities, if present. 89229 This feature was initiated by xserver module and made available to all. 89230 89231 The docbook.am makefile hides all the details and is identical for 89232 all 22 modules having DocBook documentation. It is included by a thin 89233 Makefile.am which requires no docbook knowledge. 89234 89235 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89236 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 89237 Signed-off-by: Keith Packard <keithp@keithp.com> 89238 89239commit 871d65790ec2ac0fd6d5105c2d599f63a86fcaf1 89240Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 89241Date: Mon Nov 1 16:02:11 2010 +0200 89242 89243 DRI2: Allow DDX to validate swap_limit changes 89244 89245 DDX can now implement validation for swap_limit changes to prevent 89246 configurations that are not support in driver. 89247 89248 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 89249 CC: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 89250 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 89251 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 89252 89253commit b435e2aac1b3fbb97d0275de73a1e36d16f170c0 89254Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 89255Date: Mon Nov 1 16:22:00 2010 +0200 89256 89257 DRI2: Expose API to set drawable swap limit. 89258 89259 This allows ddx to set swap_limit if there is more than one back 89260 buffer for drawable. Setting swap_limit has to also check if change 89261 affects a client that is blocked. 89262 89263 This can be used to implement N-buffering in driver with minimal 89264 logic in allocation and selecting next back. 89265 89266 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 89267 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 89268 Reviewed-by: Francisco Jerez <currojerez@riseup.net> 89269 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 89270 Signed-off-by: Keith Packard <keithp@keithp.com> 89271 89272commit 86f8da0aa7612558e6563f5de0d9f9793854053f 89273Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 89274Date: Mon Nov 1 16:21:59 2010 +0200 89275 89276 DRI2: Add ReuseBufferNotify hook 89277 89278 ReuseBufferNotify hook is called whenever old buffer is reused in DRI2 89279 code. 89280 89281 Driver can use this hook to rewrite the buffer name if hardware requires 89282 shared buffers. Shared buffer might be some hardware limited resources like 89283 framebuffer that is preallocated in boot. 89284 89285 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 89286 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 89287 Signed-off-by: Keith Packard <keithp@keithp.com> 89288 89289commit 2f47433fef2cf63063e069ead9003891d135e87f 89290Author: vdb@picaros.org <vdb@picaros.org> 89291Date: Wed Sep 7 17:47:56 2011 +0200 89292 89293 print DisplayMode type bits 89294 89295 Dear, 89296 89297 A patch I posted on xorg-devel was reviewed and is ready for 89298 inclusion in xserver. Would you be willing to apply the patch so that 89299 it finds its way into the master branch ? 89300 89301 Thank you, Servaas Vandenberghe. 89302 89303 http://lists.x.org/archives/xorg-devel/2011-August/024769.html 89304 http://lists.x.org/archives/xorg-devel/2011-August/024777.html 89305 89306 This patch adds printing of the DisplayMode type bits to 89307 xf86PrintModeline(). It helps to trace the modeline origin and to 89308 understand the initial configured modeline. 89309 89310 Reviewed-by: Alex Deucher <alexander.deucher at amd.com> 89311 Signed-off-by: Servaas Vandenberghe 89312 Signed-off-by: Keith Packard <keithp@keithp.com> 89313 89314commit c92c83523ede8bc361526ac93d09d089fbbdae08 89315Merge: 4ad271d06 24b2b43e8 89316Author: Keith Packard <keithp@keithp.com> 89317Date: Wed Sep 21 13:32:06 2011 -0700 89318 89319 Merge remote-tracking branch 'kibi/master' 89320 89321commit 4ad271d06c5aa42721c0e2e01e17e34a39825c65 89322Author: Aaron Plattner <aplattner@nvidia.com> 89323Date: Mon Aug 29 20:44:18 2011 -0700 89324 89325 xfree86: Bump extension ABI version to 6.0 89326 89327 The video driver ABI was bumped to 11.0 in commit 89328 0de7cec90738a7a5020150309866bb0e23b6f479 because of a change to the 89329 size of ATOM in commit 51f353d0a0d116af16d7d9590cadef6c56328746. This 89330 also affects extension modules, so the extension ABI version should 89331 have been bumped too. 89332 89333 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 89334 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 89335 Signed-off-by: Keith Packard <keithp@keithp.com> 89336 89337commit c8eacae4f833a0bd81907ba2b4cfaa80c08e037f 89338Author: Alan Coopersmith <alan.coopersmith@oracle.com> 89339Date: Thu Aug 25 21:46:16 2011 -0700 89340 89341 Space & style cleanup of hw/xfree86/i2c/fi1236.c 89342 89343 Fortunately, the massive decrease in the cost of whitespace in the past 89344 decade has allowed us to be much more generous with it, and much more 89345 consistent in its application, even for code like this that clearly no 89346 one has ever tried to read. 89347 89348 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89349 Reviewed-by: Jamey Sharp <jamey@minilop.net> 89350 89351commit c46215c100e964665580211a5d5893558ca9374f 89352Author: Alan Coopersmith <alan.coopersmith@oracle.com> 89353Date: Wed Aug 10 11:17:46 2011 -0700 89354 89355 Cross-reference cvt(1) & gtf(1) man pages 89356 89357 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89358 Reviewed-by: Julien Cristau <jcristau@debian.org> 89359 89360commit 1526f1caf396fefc5f09eaff884d5b92aac44997 89361Author: Alan Coopersmith <alan.coopersmith@oracle.com> 89362Date: Tue May 31 19:54:33 2011 -0700 89363 89364 Remove unused ClientStateAuthenticating 89365 89366 Appears to be leftover from the Kerberos code deleted in 2007 89367 (commit dfbe32b5b828cc4e3da36a0e2e6ad641164eaa5e). 89368 Nothing left ever set clientState to ClientStateAuthenticating 89369 89370 Skipped over 1 to preserve existing enum numbering. 89371 89372 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89373 Reviewed-by: Julien Cristau <jcristau@debian.org> 89374 89375commit 0f380a5005f800572773cd4667ce43c7459cc467 89376Author: Jamey Sharp <jamey@minilop.net> 89377Date: Tue Sep 14 18:35:21 2010 -0700 89378 89379 Fix pixmap double-frees on error paths. 89380 89381 If AddResource fails, it will automatically free the object that was 89382 passed to it by calling the appropriate deleteFunc; and of course 89383 FreeResource also calls the deleteFunc. In both cases it's wrong to call 89384 the destroy hook manually. 89385 89386 Commit by Jamey Sharp and Josh Triplett. 89387 89388 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89389 Signed-off-by: Josh Triplett <josh@joshtriplett.org> 89390 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 89391 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 89392 89393commit 8f69c935f6d06ad92fd0e8d9fcb3cde86cd258f5 89394Author: Jamey Sharp <jamey@minilop.net> 89395Date: Fri Sep 17 02:27:59 2010 +0200 89396 89397 Eliminate MAXSCREENS-sized CursorScreenKey array. 89398 89399 Use new per-screen privates API instead. 89400 89401 Commit by Jamey Sharp and Josh Triplett. 89402 89403 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89404 Signed-off-by: Josh Triplett <josh@joshtriplett.org> 89405 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 89406 89407commit c96a43888ceae72bf89c66de911120c0b6b7a71a 89408Author: Jamey Sharp <jamey@minilop.net> 89409Date: Sat Jul 10 09:28:47 2010 -0700 89410 89411 Revert "Bug #6924: Restore the ABI for DrawableRec and ColormapRec to the state" 89412 89413 This reverts commit bc0c56c407117d1545e20d21f7d30eb3472d618b since we're 89414 breaking ABI anyway. 89415 89416 Conflicts: 89417 89418 ChangeLog 89419 dix/colormap.c 89420 dix/pixmap.c 89421 dix/window.c 89422 include/pixmapstr.h 89423 89424 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89425 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 89426 89427commit 13ac1ba480ee3e89163825cd2777f9a6e9dcbc9f 89428Author: Alan Coopersmith <alan.coopersmith@oracle.com> 89429Date: Tue May 31 19:42:48 2011 -0700 89430 89431 Remove unused ClientStateCheckingSecurity & ClientStateCheckedSecurity 89432 89433 Appear to be leftovers from the XC-QUERY-SECURITY code deleted in 2007 89434 (commit 375864cb74cced40ae688078b1f7750998972535). 89435 Nothing left ever set clientState to ClientStateCheckingSecurity. 89436 89437 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89438 Reviewed-by: Julien Cristau <jcristau@debian.org> 89439 89440commit 78f946c297081d39d312e05928146c78f59fedcb 89441Author: Alan Coopersmith <alan.coopersmith@oracle.com> 89442Date: Thu May 19 23:26:31 2011 -0700 89443 89444 sun_init.c: Implement novtswitch & sharevts for Solaris 89445 89446 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89447 Reviewed-by: John Martin <john.m.martin@oracle.com> 89448 89449commit 87cc83474d9890d14d9bc60926a97bd0d5b059af 89450Author: Jamey Sharp <jamey@minilop.net> 89451Date: Sun Sep 18 20:47:04 2011 -0500 89452 89453 Replace XmuSnprintf with snprintf. 89454 89455 Alan Coopersmith explains: 89456 89457 XmuSnprintf() can be replaced by snprintf() now. (It was a 89458 implementation X provided for it's libraries to use in the days 89459 before all platforms we cared about had snprintf in libc.) 89460 89461 Reported-by: walter harms <wharms@bfs.de> 89462 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89463 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89464 89465commit df4eeb1256c7c5e7a3f88f10da2915c7a07a8db8 89466Author: Jamey Sharp <jamey@minilop.net> 89467Date: Tue Jun 8 13:58:04 2010 -0700 89468 89469 xnest: Delete unused nClipRects GC-private field. 89470 89471 This field was never read at any time in the git history. 89472 89473 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89474 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89475 89476commit 53a4744b69ab76ecabdde5dabfb56830570c0f4e 89477Author: Jamey Sharp <jamey@minilop.net> 89478Date: Sat Sep 18 01:23:12 2010 +0200 89479 89480 XineramaInitData ignores its argument. Quit passing one. 89481 89482 Also fix up XineramaInitData's caller, XineramaReinitData. 89483 89484 Commit by Jamey Sharp and Josh Triplett. 89485 89486 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89487 Signed-off-by: Josh Triplett <josh@joshtriplett.org> 89488 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89489 89490commit b85c10a35d44ad695797d0d19757e5f62eab4b98 89491Author: Jamey Sharp <jamey@minilop.net> 89492Date: Tue May 25 16:23:31 2010 -0700 89493 89494 dmx: Fix some "no previous prototype" warnings by making functions static. 89495 89496 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89497 Reviewed-by: Matt Turner <mattst88@gmail.com> 89498 89499commit d423012e776a81a197cde86be5c6bd471a593c39 89500Author: Jamey Sharp <jamey@minilop.net> 89501Date: Fri May 21 14:21:52 2010 -0700 89502 89503 dmx: Fix a non-constant printf format string warning. 89504 89505 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89506 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89507 89508commit 7e127f12a24acb8f43055af0a9be1ec585bbed36 89509Author: Alan Coopersmith <alan.coopersmith@oracle.com> 89510Date: Thu May 19 23:22:27 2011 -0700 89511 89512 sun_init.c: Move vt switches to a switch_to helper function like lnx_init.c 89513 89514 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89515 Reviewed-by: Julien Cristau <jcristau@debian.org> 89516 89517commit a00dc6af6e84d0bee3f69d8df89d7f68cb535d35 89518Author: Alan Coopersmith <alan.coopersmith@oracle.com> 89519Date: Thu May 19 22:43:56 2011 -0700 89520 89521 Xorg.man: Add -novtswitch and -sharevts options 89522 89523 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89524 Reviewed-by: Julien Cristau <jcristau@debian.org> 89525 89526commit 93abda6bddf770dbfc3d4d0ada61ea84a7b7b9be 89527Author: Alan Coopersmith <alan.coopersmith@oracle.com> 89528Date: Thu May 19 22:39:09 2011 -0700 89529 89530 xfree86: move -novtswitch & -sharevts argument handling up to common layer 89531 89532 Stop duplicating in each os-support variant before it gets replicated 89533 even further. 89534 89535 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89536 Reviewed-by: Julien Cristau <jcristau@debian.org> 89537 Reviewed-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 89538 89539commit d8d99e14b1241fd8c099d69b8d0ed0817585fe8d 89540Author: Jamey Sharp <jamey@minilop.net> 89541Date: Wed Sep 14 09:52:58 2011 -0500 89542 89543 configure.ac: Remove unreachable check for VM86 headers. 89544 89545 "configure --with-int10" is not a valid configuration, and the check for 89546 sys/vm86.h and sys/io.h is not used. Delete it. 89547 89548 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89549 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89550 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 89551 89552commit 94b2eea86319934316bdb013f6f3940685a88a2f 89553Author: Jamey Sharp <jamey@minilop.net> 89554Date: Wed Sep 14 11:23:03 2011 -0500 89555 89556 int10: Delete #if 0'd implementation that's older than our git history. 89557 89558 Throughout the xserver git history, the generic portion of the int10 89559 module has always used other methods for reading the video BIOS. For 89560 some time now it's been purely libpciaccess based. This commented-out 89561 use of xf86ReadBIOS is entirely superfluous. 89562 89563 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89564 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89565 89566commit 6ad7bb69eebd5e5b4068bd9aa0b9a7168f075006 89567Author: Jamey Sharp <jamey@minilop.net> 89568Date: Tue Sep 13 17:45:19 2011 -0500 89569 89570 Remove PC98 support. 89571 89572 Gaetan Nadon wrote: 89573 89574 Alan Coopersmith wrote: 89575 89576 "I think we recently dropped PC98 support from the X server, so I'd 89577 be okay with dropping the documentation now". 89578 89579 Let's make them be right, shall we? 89580 89581 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89582 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89583 Acked-by: Gaetan Nadon <memsize@videotron.ca> 89584 Acked-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89585 89586commit 9eb22ebf46b77c15e53017a37c7ef605521dc164 89587Author: Jamey Sharp <jamey@minilop.net> 89588Date: Tue Sep 13 16:30:06 2011 -0500 89589 89590 configure.ac: XORG_OS is not used, so delete it. 89591 89592 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89593 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89594 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 89595 89596commit 76636f30f7cf2344f4038c59bf292db19be33d53 89597Author: Jamey Sharp <jamey@minilop.net> 89598Date: Tue Sep 13 13:10:10 2011 -0500 89599 89600 configure.ac: Remove unused XORG_OS_PCI variable. 89601 89602 This is the last mention after ccfaf82367c9d057fd8314ce36b47f0a8eb696b6 89603 quit using the variable. 89604 89605 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89606 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89607 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 89608 89609 Tiago Vignatti posted an identical patch in June 2010, which I only 89610 noticed after getting the above reviews. His patch was: 89611 89612 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 89613 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 89614 89615commit 60b4bd181bbdc794c7f3547f67df916132aa111e 89616Author: Jamey Sharp <jamey@minilop.net> 89617Date: Tue Sep 13 14:31:24 2011 -0500 89618 89619 x86emu: There is no NO_INLINE implementation of unaligned access here. 89620 89621 Patch produced with: 89622 unifdef -UNO_INLINE -B 89623 89624 This change isn't relevant to the similar code in 89625 hw/xfree86/common/compiler.h, because x86emu is expected to someday move 89626 out of xserver entirely and so should not depend on any xserver headers. 89627 Also, some platforms apparently do have NO_INLINE versions of 89628 compiler.h. 89629 89630 Signed-off-by: Jamey Sharp <jamey@minilop.net> 89631 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 89632 Reviewed-by: Matt Turner <mattst88@gmail.com> 89633 89634commit 24b2b43e85c1e3f3d242ccab8793bf4ef83cfac4 89635Author: Cyril Brulebois <kibi@debian.org> 89636Date: Sun Jun 5 14:53:08 2011 +0200 89637 89638 render: Replace __inline with inline. 89639 89640 Also remove traces from the past. 89641 89642 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89643 Signed-off-by: Cyril Brulebois <kibi@debian.org> 89644 89645commit 9f01249d7f81514a38c44dff2a4a4612dbf97cae 89646Author: Cyril Brulebois <kibi@debian.org> 89647Date: Sun Jun 5 15:10:56 2011 +0200 89648 89649 configure.ac: Fix help string for mitshm. 89650 89651 Pasting from ./configure --help's output, one would get: 89652 | configure: WARNING: unrecognized options: --disable-shm 89653 89654 Fix the help string to include the previously missing “mit” bits. 89655 89656 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 89657 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89658 Signed-off-by: Cyril Brulebois <kibi@debian.org> 89659 89660commit 0caeef6146bee5fb1827ab25db191685dde9d4b4 89661Author: Keith Packard <keithp@keithp.com> 89662Date: Fri Aug 26 16:46:13 2011 -0700 89663 89664 Version bumped to 1.11 89665 89666 Signed-off-by: Keith Packard <keithp@keithp.com> 89667 89668commit 05284a03f9002b03a66ae355b34790ec02b726f0 89669Author: Peter Hutterer <peter.hutterer@who-t.net> 89670Date: Thu Aug 4 14:45:46 2011 +1000 89671 89672 input: make InputOption opaque, provide interface functions. 89673 89674 InputOptions is not switched to use struct list for a future patch to unify 89675 it with the XF86OptionRec. 89676 89677 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89678 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 89679 89680commit fcafe825751bef99f4c0b36250ca6f15f127502f 89681Author: Peter Hutterer <peter.hutterer@who-t.net> 89682Date: Mon Aug 8 15:21:46 2011 +1000 89683 89684 Add null-terminated list interface. 89685 89686 This is a set of macros to provide a struct list-alike interface for classic 89687 linked lists such as the XF86OptionRec or the DeviceIntRec. The typical 89688 format for these is to have a "struct foo *next" pointer in each struct foo 89689 and walk through those. These macros provide a few basic functions to add to, 89690 remove from and iterate through these lists. 89691 89692 While struct list is in some ways more flexible, switching legacy code to 89693 use struct list is not alway viable. These macros at least reduce the amount 89694 of open-coded lists. 89695 89696 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89697 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 89698 89699commit 79ca7c0b5786a02a80a1c40ed475e928da7c82b3 89700Author: Peter Hutterer <peter.hutterer@who-t.net> 89701Date: Fri Aug 12 15:55:48 2011 +1000 89702 89703 xfree86: comment typo fix 89704 89705 in synch → in sync 89706 89707 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89708 89709commit 2bfb802839688ecf328119c4c6979390fc60348d 89710Author: Peter Hutterer <peter.hutterer@who-t.net> 89711Date: Fri Jul 29 16:33:54 2011 +1000 89712 89713 dix: don't XWarpPointer through the last slave anymore (#38313) 89714 89715 This line was introduced pre-1.6 to fix Bug 19297. The effect of warping 89716 through the VCP then was that if a device had custom valuator ranges, the 89717 warp position would be wrong. The better device for this effect is the the 89718 XTest device. 89719 89720 This fixes a server crash where the lastSlave is a pointer device without 89721 valuators (Bug 38313#0). 89722 89723 And while we're at it, make sure the Xinerama code-path does the same. 89724 89725 X.Org Bug 38313 <http://bugs.freedesktop.org/show_bug.cgi?id=38313> 89726 89727 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89728 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89729 89730commit b29ce0726d55ec41cfbce0814e21b0217ef64efe 89731Author: Nobuhiro Iwamatsu <iwamatsu@nigauri.org> 89732Date: Tue Aug 9 11:00:34 2011 +0900 89733 89734 Disable check of double-aligned in test/input.c on Renesas SH 89735 89736 Renesas SH is not aligned at size of double. 89737 When structure has double value, It is aligned in 4byte (long). 89738 89739 Signed-off-by: Nobuhiro Iwamatsu <iwamatsu@nigauri.org> 89740 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 89741 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89742 89743commit 3be379f5076566edaf92c27df5a4d447bcf5d015 89744Author: Peter Hutterer <peter.hutterer@who-t.net> 89745Date: Thu Aug 11 10:59:07 2011 +1000 89746 89747 dix: use helper functions in EventIsDeliverable 89748 89749 Proximity events don't have an XI2 type and caused error messages in the 89750 log when trying to get the event filter. Use this opportunity to 89751 clean up the code, instead of manually setting the fields that 89752 GetEventFilter requires use EventTo(XI2|XI|Core) instead. 89753 89754 Co-Authored-by: Keith Packard <keithp@keithp.com> 89755 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89756 Reviewed-by: Keith Packard <keithp@keithp.com> 89757 89758commit 799879797505a5e891ccaec2bea73fd838c94b7a 89759Author: Peter Hutterer <peter.hutterer@who-t.net> 89760Date: Thu Aug 11 10:57:05 2011 +1000 89761 89762 dix: abstract XI2 filter mask lookup 89763 89764 Don't access the xi2mask bytes directly or calculate the offsets manually, 89765 use a few helper functions instead. XI2 masks are a bit weird in the event 89766 handling code since they slot onto the legacy code. For core/XI 1.x events, 89767 the event mask is a CARD32. That mask is used together with the event filter 89768 (also 32 bit) to determine if event delivery should be attempted. 89769 XI2 masks are of arbitrary size and their mask is simply the byte of the 89770 mask that contains the event mask. Likewise, the filter is a single byte 89771 matching that mask. Provide helper functions get these bytes and masks in 89772 the right order instead of accessing them manually. 89773 89774 EventIsDeliverable should be part of this cleanup patch but it will be 89775 gutted with the next patch. 89776 89777 Co-Authored-by: Keith Packard <keithp@keithp.com> 89778 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89779 Reviewed-by: Keith Packard <keithp@keithp.com> 89780 89781commit 8c5a4d6fbecf79f2dc4f2d836d741203b2d5e856 89782Author: Peter Hutterer <peter.hutterer@who-t.net> 89783Date: Mon Aug 1 14:12:41 2011 +1000 89784 89785 dix: don't use the pointer as modifier device in UngrabKey. 89786 89787 Modifier device is always the keyboard. 89788 89789 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89790 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89791 89792commit 3a077f246e9ac07a37c1b01c3d321e0f5ceb4153 89793Author: Peter Hutterer <peter.hutterer@who-t.net> 89794Date: Wed Aug 3 15:07:23 2011 +1000 89795 89796 input: provide a single function to init DeviceEvents to 0 89797 89798 getevents.c already had that function, but XKB was manually initializing it, 89799 causing bugs when the event structure was updated in one place but not the 89800 other. 89801 89802 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89803 89804commit 4b376ddeb4f3c9d9d279ffd2946d88edd5af4cfc 89805Author: Peter Hutterer <peter.hutterer@who-t.net> 89806Date: Mon Aug 1 14:20:53 2011 +1000 89807 89808 dix: fix compiler warnings ("foo" set but not used) 89809 89810 devices.c: In function 'AttachDevice': 89811 devices.c:2409:18: warning: variable 'oldmaster' set but not used 89812 [-Wunused-but-set-variable] 89813 89814 events.c: In function 'ConfineToShape': 89815 events.c:683:15: warning: variable 'pSprite' set but not used 89816 [-Wunused-but-set-variable] 89817 89818 events.c: In function 'ProcGrabPointer': 89819 events.c:4759:15: warning: variable 'time' set but not used 89820 [-Wunused-but-set-variable] 89821 89822 getevents.c: In function 'GetMotionHistory': 89823 getevents.c:425:9: warning: variable 'dflt' set but not used 89824 [-Wunused-but-set-variable] 89825 89826 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89827 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89828 89829commit 6fd2adc179141310e45a56ee90ef5b5f6115a1f6 89830Author: Peter Hutterer <peter.hutterer@who-t.net> 89831Date: Mon Aug 1 14:27:54 2011 +1000 89832 89833 mi: fix compiler warnings ("foo" set but not used) 89834 89835 misprite.c: In function 'miSpriteSaveUnderCursor': 89836 misprite.c:940:12: warning: variable 'y' set but not used 89837 [-Wunused-but-set-variable] 89838 misprite.c:940:9: warning: variable 'x' set but not used 89839 [-Wunused-but-set-variable] 89840 89841 mivaltree.c: In function 'miComputeClips': 89842 mivaltree.c:226:10: warning: variable 'resized' set but not used 89843 [-Wunused-but-set-variable] 89844 89845 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89846 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89847 89848commit 484cef5b29ef82402a15e155b3b8505b1e4a6830 89849Author: Peter Hutterer <peter.hutterer@who-t.net> 89850Date: Tue Aug 2 09:21:00 2011 +1000 89851 89852 Xi: silence compiler warnings (set but not used) 89853 89854 exevents.c: In function 'UpdateDeviceState': 89855 exevents.c:719:9: warning: variable 'bit' set but not used 89856 [-Wunused-but-set-variable] 89857 89858 exevents.c: In function 'ProcessOtherEvent': 89859 exevents.c:889:22: warning: variable 'v' set but not used 89860 [-Wunused-but-set-variable] 89861 exevents.c:888:17: warning: variable 'k' set but not used 89862 [-Wunused-but-set-variable] 89863 89864 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89865 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89866 89867commit 98fe735ea1d756711019c3d90ed6abd9c06abebf 89868Author: Peter Hutterer <peter.hutterer@who-t.net> 89869Date: Mon Aug 1 13:52:13 2011 +1000 89870 89871 dix: add KEYBOARD_OR_FLOAT and POINTER_OR_FLOAT to GetMaster() 89872 89873 GetMaster() currently requires an attached slave device as parameter, 89874 resuling in many calls being IsFloating(dev) ? dev : GetMaster(...); 89875 89876 Add two new parameters so GetMaster can be called unconditionally to get the 89877 right device. 89878 89879 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89880 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89881 89882commit dbbe5735d1451bb32f43bce90f0bcfeff46f9743 89883Author: Peter Hutterer <peter.hutterer@who-t.net> 89884Date: Tue Aug 2 08:54:00 2011 +1000 89885 89886 test: add a test for GetMaster() behaviour 89887 89888 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89889 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89890 89891commit b3c76b0c53ac42b70d12849da18465e8467e474c 89892Author: Peter Hutterer <peter.hutterer@who-t.net> 89893Date: Mon Aug 1 11:54:17 2011 +1000 89894 89895 mi: fix comment typo, whitespace in miPointerSetPosition 89896 89897 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89898 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89899 89900commit c9562bed0d5e26b7e3e55e26cf1ddc5086d61cc6 89901Author: Peter Hutterer <peter.hutterer@who-t.net> 89902Date: Mon Aug 1 09:52:38 2011 +1000 89903 89904 dix: rename mieqSwitchScreen argument fromDix → set_dequeue_screen, document 89905 89906 fromDIX is neither exactly true nor particularly helpful in understanding 89907 what this parameter triggers. Rename to set_dequeue_screen, because that's 89908 exactly what happens. 89909 89910 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89911 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89912 89913commit 09496996accfdaf7bc01097a25db400912004d97 89914Author: Peter Hutterer <peter.hutterer@who-t.net> 89915Date: Thu Jul 28 15:56:08 2011 +1000 89916 89917 dix: ignore devices when adding passive core grabs to list (#39545) 89918 89919 Passive core grabs are mostly device-independent. In an MPX scenario, they 89920 may change to reflect whichever master pair activated the grab last. For 89921 adding new grabs to the list, ignore the device for core grabs to return 89922 failures when trying to set the same grab combo twice on a window. 89923 89924 X.Org Bug 39545 <http://bugs.freedesktop.org/show_bug.cgi?id=39545> 89925 89926 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89927 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89928 89929commit 80c37048539daa1d257d127d66502bde45c97c85 89930Author: Peter Hutterer <peter.hutterer@who-t.net> 89931Date: Thu Jul 28 15:43:10 2011 +1000 89932 89933 dix: avoid using the VCP as modifier device 89934 89935 Core grabs may change device when they're activated to reflect the master 89936 they apply to. If the device is a keyboard, modifierDevice is erroneously 89937 set to the Virtual Core Pointer. 89938 89939 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89940 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89941 89942commit 20a61845d3c93c337bf3331a6bac30cf66c2a293 89943Author: Peter Hutterer <peter.hutterer@who-t.net> 89944Date: Tue Aug 9 11:21:06 2011 +1000 89945 89946 config: use add_option for '_source' too 89947 89948 _source was being allocated manually, with all other options added to that 89949 list through add_option. Skip the manual part, allocate the first option 89950 _source with add_option too. 89951 89952 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89953 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89954 89955commit d33652dad8838ab0a9175ca4613a3161ebc5676f 89956Author: Peter Hutterer <peter.hutterer@who-t.net> 89957Date: Tue Aug 9 11:20:31 2011 +1000 89958 89959 config: return the new InputOption from add_option. 89960 89961 Change add_option to return the new InputOption on success, or NULL 89962 failure. This way we can at least check for errors in callers. 89963 89964 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89965 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89966 89967commit e684e816acb617b4dc66a68e2b0ba8f80399170a 89968Author: Peter Hutterer <peter.hutterer@who-t.net> 89969Date: Wed Jul 20 16:21:28 2011 +1000 89970 89971 config: fix a log message 89972 89973 PRODUCT was taken from the parent, hence ppath. 89974 89975 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 89976 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 89977 89978commit 1357cd725143c1a35e32f15df658de111b151692 89979Author: Peter Hutterer <peter.hutterer@who-t.net> 89980Date: Fri Jul 8 16:13:45 2011 +1000 89981 89982 Revert "Attempt to add the 'mouse' driver in more situations." 89983 89984 This reverts commit 43d9edd31e31b33b9da4a50d8ab05004881c8d5a. 89985 89986 This commit was introduced in the 1.2 cycle when hotplugging was less than 89987 ideal (i.e. it didn't exist). From the commit message: 89988 89989 Always add a mouse driver instance configured to send core events, unless 89990 a core pointer already exists using either the mouse or void drivers. This 89991 handles the laptop case where the config file only specifies, say, 89992 synaptics, which causes the touchpad to work but not the pointing stick. 89993 We don't double-instantiate the mouse driver to avoid the mouse moving twice 89994 as fast, and we skip this logic when the user asked for a void core pointer 89995 since that probably means they want to run with no pointer at all. 89996 89997 To get this case above, a user would need to disable hotplugging _and_ have a 89998 xorg.conf that only references one device. This is possible, but not a use-case 89999 we should worry about too much now. 90000 90001 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90002 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90003 90004commit 159b03e13760920274b573a2bccdbf6a79f059e7 90005Author: Lennart Poettering <lennart@poettering.net> 90006Date: Mon Jul 18 21:19:23 2011 +0200 90007 90008 config: add udev/systemd multi-seat support 90009 90010 Add support for multi-seat-aware input device hotplugging. This 90011 implements the multi-seat scheme explained here: 90012 90013 http://www.freedesktop.org/wiki/Software/systemd/multiseat 90014 90015 This introduces a new X server switch "-seat" which allows configuration 90016 of the seat to enumerate hotplugging devices on. If specified the value 90017 of this parameter will also be exported as root window property 90018 Xorg_Seat. 90019 90020 To properly support input hotplugging devices need to be tagged in udev 90021 according to the seat they are on. Untagged devices are assumed to be on 90022 the default seat "seat0". If no "-seat" parameter is passed only devices 90023 on "seat0" are used. This means that the new scheme is perfectly 90024 compatible with existing setups which have no tagged input devices. 90025 90026 Note that the -seat switch takes a completely generic identifier, and 90027 that it has no effect on non-Linux systems. In fact, on other OSes a 90028 completely different identifier scheme for seats could be used but still 90029 be exposed with the Xorg_Seat and -seat. 90030 90031 I tried to follow the coding style of the surrounding code blocks if 90032 there was any one could follow. 90033 90034 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90035 90036commit 95772598b57f6054fbf88683fa0a492c77605790 90037Author: Peter Hutterer <peter.hutterer@who-t.net> 90038Date: Fri Jul 8 16:10:07 2011 +1000 90039 90040 xfree86: use xf86AllocateInput for implicit devices too 90041 90042 Slowly merging the vastly different code-paths. 90043 90044 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90045 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90046 90047commit fa8f4652819b692faaf2789cf32d7fa99fbb34aa 90048Author: Peter Hutterer <peter.hutterer@who-t.net> 90049Date: Mon Jul 4 15:02:20 2011 +1000 90050 90051 xfree86: factor out adding/removing a device from the input device array 90052 90053 No functional changes, just readability improvements. This also gets rid of 90054 the count variable. Count was just used for resizing the null-terminated 90055 list. Since we're not in a time-critical path here at all we can afford to 90056 loop the list multiple times instead of keeping an extra variable around. 90057 90058 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90059 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90060 90061commit 5b5477c05f691205064ca4d8034f8dd47ab975b7 90062Author: Peter Hutterer <peter.hutterer@who-t.net> 90063Date: Mon Jul 4 14:14:11 2011 +1000 90064 90065 xfree86: update comment for InitInput 90066 90067 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90068 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90069 90070commit 7354f607833c69626d8692bc5176b18ea1cf6263 90071Author: Peter Hutterer <peter.hutterer@who-t.net> 90072Date: Mon Jul 4 14:26:13 2011 +1000 90073 90074 xfree86: nest loops instead of 0x1 pointers. 90075 90076 If we find the core device, move all other device pointers forward right 90077 then and there. The break will jump out of the top loop. 90078 90079 They had a special on braces today, so I added some for readability (and 90080 fixed up tab vs space indentation. 90081 90082 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90083 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90084 90085commit 5669aa2d24dff9ab276e5f74a09f97ec77b90e75 90086Author: Peter Hutterer <peter.hutterer@who-t.net> 90087Date: Mon Jul 4 14:05:30 2011 +1000 90088 90089 xfree86: improve readability of synthesized device. 90090 90091 No functional changes. 90092 90093 The options we assign are the ones from the Pointer/Keyboard device so we 90094 might as well use those readable names instead of dev[count-1]->options. 90095 90096 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90097 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90098 90099commit 5aa826cdd1f2e768bedf23d399703a5d0b6302be 90100Author: Peter Hutterer <peter.hutterer@who-t.net> 90101Date: Mon Jul 4 13:52:11 2011 +1000 90102 90103 test: add a option duplication test 90104 90105 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90106 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90107 90108commit 4527e2b776cfcdac2b189b5439b9a3d0b6433077 90109Author: Peter Hutterer <peter.hutterer@who-t.net> 90110Date: Mon Jul 4 13:46:24 2011 +1000 90111 90112 xfree86: when implicitly choosing a core device, set the option to a value 90113 90114 Devices are core pointers/keyboards by default now anyway, but let's set the 90115 option to some value instead of just NULL. 90116 90117 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90118 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90119 90120commit 44d53728a6a533fc0a6e0a10269d1cc99e9dad32 90121Author: Peter Hutterer <peter.hutterer@who-t.net> 90122Date: Mon Jul 4 13:44:44 2011 +1000 90123 90124 xfree86: don't warn about duplicate core devices 90125 90126 It doesn't matter. All devices are core pointer devices by default now 90127 anyway. 90128 90129 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90130 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90131 90132commit 033f53c223dc12a91f00e10a69f87a4f2a7adb6b 90133Author: Daniel Kurtz <djkurtz@chromium.org> 90134Date: Mon Aug 8 15:09:48 2011 +0800 90135 90136 xf86Helper: use LogHdrMessageVerb in xf86VDrvMsgVerb 90137 90138 LogHdrMessageVerb allows passing a parameterized header to insert in a log 90139 message between MessageType and the formatted message body string. 90140 90141 Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> 90142 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 90143 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90144 90145commit cd8ee3e5cb29b9cd6402d2fbc71463c6b04b6077 90146Author: Daniel Kurtz <djkurtz@chromium.org> 90147Date: Mon Aug 8 15:09:47 2011 +0800 90148 90149 xf86Helper: use LogHdrMessageVerb in xf86VIDrvMsgVerb 90150 90151 LogHdrMessageVerb allows passing a parameterized header to insert in a log 90152 message between MessageType and the formatted message body string. 90153 90154 Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> 90155 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 90156 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90157 90158commit b31d104fc09a7241856ff8d226be11ec562beac3 90159Author: Daniel Kurtz <djkurtz@chromium.org> 90160Date: Mon Aug 8 15:09:46 2011 +0800 90161 90162 os/log: Add LogVHdrMessageVerb and friends 90163 90164 LogVHdrMessageVerb allows a custom header to be inserted in a log message, 90165 between the Log system's MessageType string, and a formatted variable 90166 message body. The custom header can itself be a formatted variable string. 90167 90168 These functions can be used, for example, by driver abstraction layers to 90169 format specific driver messages in a standard format, but do it in a way 90170 that is efficient, obeys the log-layers verbosity settings, and is safe 90171 to use in signal handlers (because they don't call malloc), even for 90172 types besides X_NONE. 90173 90174 Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> 90175 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 90176 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90177 90178commit b82f934db661c59d7397ce180d08cf8f8f7118b1 90179Author: Daniel Kurtz <djkurtz@chromium.org> 90180Date: Mon Aug 8 15:09:45 2011 +0800 90181 90182 os/log: Pull LogMessageTypeVerbString out of LogVMessageVerb 90183 90184 Also, optimize how the type and format strings are combined. 90185 90186 Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> 90187 Reviewed-by: Guillem Jover <guillem@hadrons.org> 90188 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90189 90190commit e32cc0b4c85c78cd8743a6e1680dcc79054b57ce 90191Author: Adam Jackson <ajax@redhat.com> 90192Date: Thu Apr 21 16:37:11 2011 -0400 90193 90194 fb: Fix memcpy abuse 90195 90196 The memcpy fast path implicitly assumes that the copy walks 90197 left-to-right. That's not something memcpy guarantees, and newer glibc 90198 on some processors will indeed break that assumption. Since we walk a 90199 line at a time, check the source and destination against the width of 90200 the blit to determine whether we can be sloppy enough to allow memcpy. 90201 (Having done this, we can remove the check for !reverse as well.) 90202 90203 On an Intel Core i7-2630QM with an NVIDIA GeForce GTX 460M running in 90204 NoAccel, the broken code and various fixes for -copywinwin{10,100,500} 90205 gives (edited to fit in 80 columns): 90206 90207 1: Disable the fastpath entirely 90208 2: Replace memcpy with memmove 90209 3: This fix 90210 4: The code before this fix 90211 90212 1 2 3 4 Operation 90213 ------ --------------- --------------- --------------- ------------ 90214 258000 269000 ( 1.04) 544000 ( 2.11) 552000 ( 2.14) Copy 10x10 90215 21300 23000 ( 1.08) 43700 ( 2.05) 47100 ( 2.21) Copy 100x100 90216 960 962 ( 1.00) 1990 ( 2.09) 1990 ( 2.07) Copy 500x500 90217 90218 So it's a modest performance hit, but correctness demands it, and it's 90219 probably worth keeping the 2x speedup from having the fast path in the 90220 first place. 90221 90222 Signed-off-by: Adam Jackson <ajax@redhat.com> 90223 Signed-off-by: Keith Packard <keithp@keithp.com> 90224 90225commit ac2c307f4716ebd3e955c004ceec9f4c029401a0 90226Author: Pelle Johansson <pelle@morth.org> 90227Date: Sun Aug 14 17:44:40 2011 -0700 90228 90229 XQuartz: Initialize darwin pointer valuators 90230 90231 This fixes a regression introduced by: 633b81e8ba09cc6a1ea8b43f323874fda2cf0bde 90232 90233 http://xquartz.macosforge.org/trac/ticket/498 90234 90235 Signed-off-by: Pelle Johansson <pelle@morth.org> 90236 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 90237 90238commit 4020cab88f5cf3164fc83cf912f94f288aa5a45d 90239Author: Michel Dänzer <michel.daenzer@amd.com> 90240Date: Wed Aug 10 11:36:16 2011 +0200 90241 90242 EXA/mixed: Update sys_pitch in MPH even when there's no system memory copy. 90243 90244 Otherwise sys_pitch will be stale when a system memory copy is allocated. 90245 90246 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=38322 and a crash when 90247 unlocking the screen with xscreensaver, reported by Janne Huttunen. 90248 90249 Signed-off-by: Michel Dänzer <michel.daenzer@amd.com> 90250 Tested-by: Janne Huttunen <jahuttun@gmail.com> 90251 Tested-by: Jan Kriho <Erbureth@gmail.com> 90252 Signed-off-by: Keith Packard <keithp@keithp.com> 90253 90254commit a1dec7cd6adabaf8770dc3b21b0df30ce2f0bc8d 90255Author: Julien Cristau <jcristau@debian.org> 90256Date: Thu Jul 7 19:18:03 2011 +0200 90257 90258 Xephyr/dri: register screen and window privates on init 90259 90260 Fixes assertion failure when calling dixSetPrivate 90261 Debian bug#632549 <http://bugs.debian.org/632549> 90262 90263 Reported-and-tested-by: Mohammed Sameer <msameer@foolab.org> 90264 Signed-off-by: Julien Cristau <jcristau@debian.org> 90265 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90266 Signed-off-by: Keith Packard <keithp@keithp.com> 90267 90268commit 39bc81b60ddc9e495ff3b9bb3c3a6c39e547b43a 90269Author: Keith Packard <keithp@keithp.com> 90270Date: Wed Aug 3 20:57:03 2011 -0700 90271 90272 Version bumped to 1.10.99.902 (1.11 RC2) 90273 90274 At the close of the 1.11 non-critical fixes window. 90275 90276 Signed-off-by: Keith Packard <keithp@keithp.com> 90277 90278commit 9504caf1c3243e3ab2eb7126bc2bb876a8f89918 90279Author: Ville Syrjälä <ville.syrjala@nokia.com> 90280Date: Fri May 6 18:19:34 2011 +0300 90281 90282 composite: Inhibit window background paint with manual subwindow redirection 90283 90284 The composite extension spec says that window background painting 90285 should be inhibited when the subwindow redirection mode is set to 90286 manual. 90287 90288 This eliminates the ugly flashing effect when compiz unredirects a 90289 fullscreen window. 90290 90291 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 90292 Reviewed-by: Owen Taylor <otaylor@fishsoup.net> 90293 Reviewed-by: Keith Packard <keithp@keithp.com> 90294 Signed-off-by: Keith Packard <keithp@keithp.com> 90295 90296commit f1d75f3b742231c4c09f13d7a22ed920014cb5ef 90297Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 90298Date: Thu Jul 28 15:17:59 2011 -0700 90299 90300 Revert "composite: Don't backfill non-MapWindow allocations" 90301 90302 This reverts commit db8840600e8e21356241eb87395031388d9b54d2. 90303 90304 It was an optimization for the resize case, but 193ecc8b453b22 made 90305 it so that no backfilling takes place on resize if left in. 90306 90307 Signed-off-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 90308 Reviewed-by: Adam Jackson <ajax@redhat.com> 90309 90310 Conflicts: 90311 90312 composite/compalloc.c 90313 Signed-off-by: Keith Packard <keithp@keithp.com> 90314 90315commit e87adcc9e024982fdad974de2876b00cf974dab8 90316Author: Julien Cristau <jcristau@debian.org> 90317Date: Sun Jul 31 21:14:59 2011 +0200 90318 90319 Xquartz: include new localization files in the tarball 90320 90321 Signed-off-by: Julien Cristau <jcristau@debian.org> 90322 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 90323 90324commit a1bb5062c72667b2f10b56925c61888acc89e3e4 90325Author: Jeremy Huddleston <jeremyhu@apple.com> 90326Date: Thu Jul 21 10:33:58 2011 -0700 90327 90328 XQuartz: xpr: Don't FatalError if xp_unlock_window fails 90329 90330 We added the FatalError in 5d1d9d9ae39fab2ee2ac085f9776f82768828dc8 but 90331 it caused a regression http://xquartz.macosforge.org/trac/ticket/482 90332 90333 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 90334 90335commit c319f7b5b3ee651636cdfd165588f0dc972a22a4 90336Author: Jeremy Huddleston <jeremyhu@apple.com> 90337Date: Wed Jul 20 00:16:42 2011 -0700 90338 90339 XQuartz: Use CFSTR to avoid implicit cast warning of NSString * to CFStringRef 90340 90341 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 90342 90343commit 08dfff92e8d0c2b1726634392c147f6634d1706d 90344Author: Aaron Plattner <aplattner@nvidia.com> 90345Date: Fri Jul 29 16:12:06 2011 -0700 90346 90347 randr: Compare all the bytes in RRPostPendingProperties 90348 90349 RRPostPendingProperties tries to compare the pending and current 90350 property values to decide whether they're actually changing. However, 90351 it does this using a memcmp that passes in pending_value->size as the 90352 number of bytes. This is actually the number of elements, where each 90353 element is (pending_value->format / 8) bytes long. This causes the 90354 pending value to not be propagated if the first pending_value->size 90355 bytes are the same and only the end of it is changing. 90356 90357 Fix this by computing the total number of bytes to compare in the 90358 memcmp. 90359 90360 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 90361 Reviewed-by: Keith Packard <keithp@keithp.com> 90362 Signed-off-by: Keith Packard <keithp@keithp.com> 90363 90364commit b8f61c11c94708d4f17720a55945dee4621315a4 90365Merge: 5596f10cc f51e42f58 90366Author: Keith Packard <keithp@keithp.com> 90367Date: Fri Jul 29 14:58:58 2011 -0700 90368 90369 Merge remote-tracking branch 'whot/for-keith' 90370 90371commit f51e42f583073bde0bc8131887cb7220636c8855 90372Author: Peter Hutterer <peter.hutterer@who-t.net> 90373Date: Wed Jul 20 13:09:05 2011 +1000 90374 90375 Terminate the log with one last message. 90376 90377 Instead of just closing the log when everything is done, put one more 90378 message in stating that we're actually terminating. Users or scripts that 90379 look at the Xorg.log will then know that a) the server has terminated 90380 properly and b) why the server terminated (to some degree, given that most 90381 real-world errors will be caused by AbortServer()). 90382 90383 Acked-by: Gaetan Nadon <memsize@videotron.ca> 90384 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90385 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 90386 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 90387 Tested-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90388 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90389 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90390 90391commit 01de08c7d2c00eef238adba6665896ea3cd7d511 90392Author: Julien Cristau <jcristau@debian.org> 90393Date: Tue Jul 26 20:40:38 2011 +0200 90394 90395 configure: set default xkb rules to evdev on Linux 90396 90397 If config/udev was enabled, this would default to base, which means that 90398 after regen the devices would get the wrong rules, and hilarity would 90399 ensue. 90400 90401 It's probably safe to default to evdev unconditionally on Linux by now. 90402 90403 Reported-by: Bastian Blank <waldi@debian.org> 90404 Signed-off-by: Julien Cristau <jcristau@debian.org> 90405 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 90406 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90407 90408commit 3798dd379c1ecf325f9907128fb66d20372f6876 90409Author: Peter Hutterer <peter.hutterer@who-t.net> 90410Date: Wed Jul 20 09:00:18 2011 +1000 90411 90412 Initialize the fd to -1 for xorg.conf input devices. 90413 90414 For hotplugged devices, xf86AllocateInput does that for us but the xorg.conf 90415 path is different. Since not all drivers reset the fd during PreInit but may 90416 still call close(pInfo->fd) in all cases, this can terminate the logging 90417 early. 90418 90419 Reproducible: add a wacom driver InputDevice section with no Option Device. 90420 90421 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90422 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90423 90424commit f2a6735cfc07789cca81852b24a85578f200d83d 90425Author: Peter Hutterer <peter.hutterer@who-t.net> 90426Date: Mon Jul 4 12:34:32 2011 +1000 90427 90428 xfree86: NULL option values are technically valid, don't strdup them 90429 90430 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90431 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90432 90433commit f0d7e9db28c374a3db359bcb63a7ce79fd84b541 90434Author: Peter Hutterer <peter.hutterer@who-t.net> 90435Date: Mon Jul 4 14:14:39 2011 +1000 90436 90437 xfree86: duplicate xorg.conf device information before xf86NewInputDevice 90438 90439 xf86ConfigLayout.inputs contains the information from the xorg.conf 90440 file. Passing this into xf86NewInputDevice means the device will get 90441 cleaned up on exit and the pointers in xf86ConfigLayout.inputs are left 90442 dangling. In the second server generation, this results in a server 90443 crash. 90444 90445 Also, rename pDev to pInfo. pDev is pretty much reserved for DeviceIntPtr 90446 types. 90447 90448 Reproducible: AutoAddDevices off and xorg.conf input sections, trigger 90449 server regeneration. 90450 90451 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90452 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90453 90454commit 8ffddbcf72170e246826ee0f39f18989a29fa218 90455Author: Peter Hutterer <peter.hutterer@who-t.net> 90456Date: Mon Jun 27 16:30:28 2011 +1000 90457 90458 xfree86: Remove devices that failed to enable on startup 90459 90460 Devices that succeeded during PreInit and DEVICE_INIT but failed in 90461 DEVICE_ON would be deleted through xf86DeleteInput but not removed from the 90462 list of input devices (and not turned off). The result was a double free on 90463 server shutdown. 90464 90465 Fix this by calling RemoveDevice if EnableDevice fails. 90466 90467 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90468 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90469 90470commit 6cea28fe4b7a4a22ad270d8c71403db84a9bfb2c 90471Author: Lennart Poettering <lennart@poettering.net> 90472Date: Mon Jul 18 21:18:27 2011 +0200 90473 90474 config: don't fail if a device vanished by the time we managed to look at it 90475 90476 The nature of hotplug is that a device we enumerated might already be 90477 gone by the time we look at it, so don't assume otherwise. 90478 90479 Signed-off-by: Lennart Poettering <lennart@poettering.net> 90480 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 90481 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90482 90483commit f13de9ca1b7c4dd0dd5c08037c6bd53f88ac30f5 90484Author: Lennart Poettering <lennart@poettering.net> 90485Date: Mon Jul 18 21:17:49 2011 +0200 90486 90487 config: limit the kernel subsystems we look for devices in 90488 90489 Don't enumerate/monitor all devices of the system (since that can be 90490 quite a few), but limit our search to devices from the "input" 90491 subsystem, as well as the "tty" subsystem (to cover Wacom tablets). 90492 90493 This should make X start up a bit faster and reduce the number of 90494 unnecessary wake-ups of the X server. 90495 90496 Signed-off-by: Lennart Poettering <lennart@poettering.net> 90497 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 90498 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90499 90500commit 882e3c2680c339ad7aa0d664e0b0f02b8a05b11d 90501Author: Lennart Poettering <lennart@poettering.net> 90502Date: Mon Jul 18 21:17:10 2011 +0200 90503 90504 config: process udev "changed" and "add" events in the same code paths 90505 90506 udev gives no guarantee that before each "changed" event for a device 90507 there's an "add" event, or that before each "remove" is an "add", or 90508 that before each "add" there was no "add" already and so on. Users can 90509 trigger these events at any time with "udevadm trigger", and netlink is 90510 a lossy transport, hence the events can come in unexpected ordering. 90511 90512 With other words: regardless which event is generated, the X server must 90513 not choke on it and make the best of it, hence make sure that if we get 90514 an "add" event for an existing device we don't add the device a second 90515 time. 90516 90517 Signed-off-by: Lennart Poettering <lennart@poettering.net> 90518 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 90519 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90520 90521commit 5596f10cce58295d8af73646352b314ecb3ddfca 90522Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90523Date: Sat Jul 9 13:21:04 2011 +0100 90524 90525 Remove hw/xwin/xlaunch 90526 90527 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90528 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90529 90530commit 73864a87aacbce85b520ccaa6e360b82c0e99716 90531Author: Jeremy Huddleston <jeremyhu@apple.com> 90532Date: Tue Jul 19 20:03:48 2011 -0700 90533 90534 XQuartz: Localization updates 90535 90536 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 90537 90538commit d2463df84d3a160b2dd1e706b02985e1817784e3 90539Author: Jeremy Huddleston <jeremyhu@apple.com> 90540Date: Tue Jul 19 20:02:33 2011 -0700 90541 90542 XQuartz: Add english NIB changes for scroll_in_device_direction 90543 90544 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 90545 90546commit e36e7ad6314ca5fcd3292b8022f0a618d567d72b 90547Author: Jeremy Huddleston <jeremyhu@apple.com> 90548Date: Tue Jul 19 20:01:11 2011 -0700 90549 90550 XQuartz: Add GUI controls to toggle scroll_in_device_direction 90551 90552 Also cleaned up the wording for fullscreen_menu. 90553 90554 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 90555 90556commit ac43984215faf464fd48bd006cac1c6539603cea 90557Author: Jeremy Huddleston <jeremyhu@apple.com> 90558Date: Tue Jul 19 20:00:27 2011 -0700 90559 90560 XQuartz: Add a scroll_in_device_direction preference 90561 90562 This preference allows users to override the related option in Mac OS X's 90563 Mouse/Trackpad preferences. This effectively lets the user determine 90564 which "context" all of X11 fits into for context-based scrolling until 90565 such API exists within X11 itself to pass along to X11 clients. 90566 90567 This is applicable to Mav OS X 10.7+ 90568 90569 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 90570 90571commit ec95a9c829b8e37529aa828c05fbaabc45cffe42 90572Author: George Staplin <gstaplin@apple.com> 90573Date: Wed Mar 4 02:03:52 2009 -0700 90574 90575 XQuartz: Add diagnostic error checking to xp_destroy_surface. 90576 90577 This occurred to me in hindsight after the last commit. If the 90578 original developer had done this, we would have noticed the 90579 problem sooner. 90580 (cherry picked from commit aa0a57996f3e7d16238f69976958c2526821388b) 90581 90582commit 4fe7df265324f63025686efe9d32342e3cef40d3 90583Author: George Staplin <gstaplin@apple.com> 90584Date: Wed Mar 4 01:39:58 2009 -0700 90585 90586 XQuartz: Fix a memory leak with surfaces that a new test found. 90587 90588 xp_destroy_surface was called with a surface id of 0, due to some 90589 premature cleanup that set it to 0. This means the surfaces 90590 weren't being destroyed until the window was. 90591 90592 The code that did that was: pDRIDrawablePriv->sid = 0; 90593 90594 In long running applications this leak may or may not have been 90595 harmful. With the old libGL the surfaces weren't destroyed until 90596 the context was destroyed or a new context created. In the new 90597 libGL they are reference counted, and released much sooner, so we 90598 ran into a resource leak more noticeably with some tests. 90599 90600 Make the Apple DRI code dispatch events to the client(s) for 90601 destroyed surfaces, when a resource is destroyed. This seems to 90602 work in my tests, however this clearly wasn't working for a while, 90603 so bugs may result in the future if it enables some new (unexpected) 90604 side effects. 90605 90606 Also add a few helpful comments to aid in understanding the code 90607 in the future. 90608 90609 Tested with the test suite, Pymol, and various Mesa demos. 90610 (cherry picked from commit bede83eb19a1629396fcd5a46441f8476a8fcd1b) 90611 90612commit 0ebe45a717faa6464d3b1ab73e30570518ee4798 90613Author: Jeremy Huddleston <jeremyhu@apple.com> 90614Date: Tue Jul 19 19:42:44 2011 -0700 90615 90616 XQuartz: DRI: Dead code removal 90617 90618 Also add some comments that weren't merged in from server-1.4-apple's 90619 99babae1326485c27eb9253db83afdd6aef9e362 90620 90621 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 90622 90623commit 88ad050f3c2f9f11bafb56a8cf777518795295a9 90624Author: Jeremy Huddleston <jeremyhu@apple.com> 90625Date: Mon Jul 18 22:54:21 2011 -0700 90626 90627 XQuartz: Add some sanity checking and a fallback for the bundle id. 90628 90629 This way we'll print an error and still mostly work rather than crashing if 90630 someone installs XQuartz.app incorrectly or tries running the server within 90631 the build system rather than the installed system. 90632 90633 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 90634 90635commit 89bfa0c633f3096cca6f983bca1507d3f4b902a8 90636Author: Jeremy Huddleston <jeremyhu@apple.com> 90637Date: Fri Jul 15 10:39:39 2011 -0700 90638 90639 XQuartz: DarwinEQFini doesn't return Bool 90640 90641 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 90642 90643commit 82f5521a6d91ebcd2a4400f6c221ad625edc99a1 90644Author: Daniel Stone <daniel@fooishbar.org> 90645Date: Wed Jul 13 12:08:04 2011 +0100 90646 90647 XKB: Work around broken interps from old xkbcomp 90648 90649 Bugfix for broken xkbcomp: if we encounter an XFree86Private action with 90650 Any+AnyOfOrNone(All), then we skip the interp as broken. Versions of 90651 xkbcomp below 1.2.2 had a bug where they would interpret a symbol that 90652 couldn't be found in an interpret as Any. So, an 90653 XF86LogWindowTree+AnyOfOrNone(All) interp that triggered the PrWins 90654 action would make every key without an action trigger PrWins if libX11 90655 didn't yet know about the XF86LogWindowTree keysym. None too useful. 90656 90657 We only do this for XFree86 actions, as the current XKB dataset relies 90658 on Any+AnyOfOrNone(All) -> SetMods for Ctrl in particular. 90659 90660 See xkbcomp commits 2a473b906943ffd807ad81960c47530ee7ae9a60 and 90661 3caab5aa37decb7b5dc1642a0452efc3e1f5100e for more details. 90662 90663 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 90664 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 90665 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90666 90667commit c177a5bcaa170d24a784540460906cc04ac5c752 90668Author: Daniel Stone <daniel@fooishbar.org> 90669Date: Wed Jul 13 12:08:03 2011 +0100 90670 90671 Fix non-Composite builds in PrintWindowTree 90672 90673 The previous patch accidentally introduced a hard dependency on 90674 Composite. Sorry, OS X. 90675 90676 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 90677 Reported-by: Jeremy Huddleston <jeremyhu@apple.com> 90678 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90679 90680commit 01dbf2514aa0cc8c40a876a24e3cb0737c41da19 90681Merge: 5a801af68 af0fbc37e 90682Author: Keith Packard <keithp@keithp.com> 90683Date: Wed Jul 13 11:44:15 2011 -0700 90684 90685 Merge remote-tracking branch 'jturney/master' 90686 90687commit 5a801af689fce9dfe84453bbc2b029cb4782de83 90688Author: Gaetan Nadon <memsize@videotron.ca> 90689Date: Tue Jul 12 19:50:05 2011 -0400 90690 90691 XWinrc: replace hard coded section number with __filemansuffix__ 90692 90693 Reviewed-by: Cyril Brulebois <kibi@debian.org> 90694 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 90695 Signed-off-by: Keith Packard <keithp@keithp.com> 90696 90697commit af0fbc37e35ddaabf12a8d7abc79ea11958ec8ac 90698Author: Gaetan Nadon <memsize@videotron.ca> 90699Date: Thu Jun 23 07:45:13 2011 -0400 90700 90701 XWinrc: replace hard coded section number with __filemansuffix__ 90702 90703 Reviewed-by: Cyril Brulebois <kibi@debian.org> 90704 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 90705 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90706 90707commit 42bc198775a78b328612427c4c8ed4d17e005ae9 90708Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90709Date: Mon Jan 24 18:29:41 2011 +0000 90710 90711 glx: Remove a few lingering traces of __GLXscreen.GLXVersion 90712 90713 The GLXversion member of the __GLXscreen struct 90714 is just cruft since commit ad5c0d9efa47476ed5cf75c82265c73919e468b4, 90715 when we started returning the minimum GLX version supported by all 90716 of the screens on the display, rather than the maximum version supported 90717 by the server. 90718 90719 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 90720 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90721 90722commit 005e68ba59782b60dd5c44a06c75b1fa5e8d70d8 90723Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90724Date: Thu Jan 20 16:21:04 2011 +0000 90725 90726 hw/xwin: Add -wgl option to XWin manpage 90727 90728 Document the -[no]wgl options in the XWin manpage 90729 90730 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90731 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90732 90733commit d62726019e4b7bfac8a23d0d733677b141047d66 90734Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90735Date: Tue Nov 2 11:38:53 2010 +0000 90736 90737 hw/xwin: Add items to WGL AIGLX todo 90738 90739 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90740 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90741 90742commit 276735ef5d12e79fe051996f100c5730c1e69334 90743Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90744Date: Mon Oct 11 16:22:49 2010 +0100 90745 90746 hw/xwin: Report Window XIDs in Window debug messages 90747 90748 Report Window XIDs in Window create/destroy/reparent debug messages 90749 90750 It's actually quite useful if you are trying to corrolate those events 90751 with what a client is doing... 90752 90753 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90754 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90755 90756commit b0b54338e2f8aed6e6a683ebddd2a2acbefabecd 90757Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90758Date: Mon Oct 11 16:25:03 2010 +0100 90759 90760 hw/xwin: Add a flag to track which windows have been drawn to using WGL. 90761 90762 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90763 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90764 90765commit 98c6d91a48f460fdbca93700ab637eadf946fb0e 90766Author: Colin Harrison <colin.harrison@virgin.net> 90767Date: Thu Mar 24 20:39:51 2011 +0000 90768 90769 hw/xwin/glx: Fix some warnings in generated wrapper code 90770 90771 Add a suitable cast to the generated code for glWinSetupDispatchTable() 90772 so it doesn't generate screeds of warnings 90773 90774 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90775 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90776 90777commit b67e80c84e4af54007aa1bd1e4a97a4d1b3d3765 90778Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90779Date: Mon Jan 24 18:41:05 2011 +0000 90780 90781 hw/xwin/glx: Handle failure to get any fbconfigs more gracefully. 90782 90783 Handle failure to get any useful pixel formats for GLX fbconfigs 90784 more gracefully: If we didn't get any useful pixel formats from 90785 wglGetPixelFormatAttribivARB(), fall back to using DescribePixelFormat(). 90786 If that doesn't give us any useful pixel formats, fallback to software 90787 rendering. 90788 90789 This works around a problem with Intel 845G drivers, where 90790 wglGetPixelFormatAttribivARB() doesn't seem to work as we expect it to... 90791 90792 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90793 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90794 90795commit 9915b169237e3b2d817bd3d4455f3c01dd7dd2b5 90796Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90797Date: Mon Nov 1 19:53:42 2010 +0000 90798 90799 hw/xwin/glx: Don't spam log with wglwrap symbol resolution status at startup 90800 90801 ... instead just log if an attempt is made to call a wrapper for 90802 a function which didn't resolve 90803 90804 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90805 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90806 90807commit 8216316867f07819b8e8b377f2633fbf8876abf7 90808Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90809Date: Sat Oct 30 17:23:52 2010 +0100 90810 90811 hw/xwin/glx: Better handling of SetPixelFormat() failure 90812 90813 Propagate and report the failure if SetPixelFormat() fails 90814 90815 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90816 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90817 90818commit eecc8e06eb4bdfd4354144206d990cd9ccb69721 90819Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90820Date: Sat Oct 30 17:22:33 2010 +0100 90821 90822 hw/xwin/glx: Improvements to glxWinErrorMessage() reporting in WGL GLX provider 90823 90824 Request the message using languageID 0 (best effort), rather than only using language neutral messages 90825 Always report the numeric error code. 90826 Trim any trailing \r from FormatMessage() output 90827 90828 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90829 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90830 90831commit 3873be16d050ab12fc78f1e7f0c36c2ac88d0a98 90832Author: Jon TURNEY <jon.turney@dronecode.org.uk> 90833Date: Mon Oct 11 16:21:40 2010 +0100 90834 90835 hw/xwin/glx: Fix fbconfig dumper formatting for 3 digit index numbers 90836 90837 Some graphics hardware supports hundreds of pixel formats, so adjust 90838 formatting in fbconfig dumper for 3 digit index numbers 90839 90840 Also report the PFD_SUPPORT_DIRECTDRAW, PFD_DIRECT3D_ACCELERATED and 90841 PFD_SUPPORT_COMPOSITION flags introduced with aero 90842 90843 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 90844 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90845 90846commit d4096abb591353405417e53816e5c46e904e7b25 90847Merge: 61f87a75f 87d4f90bf 90848Author: Keith Packard <keithp@keithp.com> 90849Date: Fri Jul 1 15:45:45 2011 -0700 90850 90851 Merge remote-tracking branch 'whot/for-keith' 90852 90853commit 87d4f90bfcb509471ac9e7886e14a92b33223fd7 90854Author: Peter Hutterer <peter.hutterer@who-t.net> 90855Date: Mon Jun 27 09:10:42 2011 +1000 90856 90857 input: free the EQ allocated memory on shutdown (#38634) 90858 90859 mieqFini() already does the right thing, but it needs to be called by the 90860 various DDXs and the XTest Extension. 90861 90862 X.Org Bug 38634 <http://bugs.freedesktop.org/show_bug.cgi?id=38634> 90863 90864 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90865 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 90866 Acked-by: Jeremy Huddleston <jeremyhu@apple.com> 90867 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90868 90869commit e084ac30b37c36016bb0dbb1b79e60a7799aee02 90870Author: Daniel Stone <daniel@fooishbar.org> 90871Date: Wed Mar 2 14:52:42 2011 +0000 90872 90873 Input: Reset SD remainder when copying co-ords from MD 90874 90875 In updateSlaveDeviceCoords, pDev->last.valuators was being copied from 90876 the master, but pDev->last.remainder wasn't. Make sure we copy both, to 90877 avoid minor inconsistencies. 90878 90879 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 90880 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 90881 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90882 90883commit 23a783278963919778d017965542e0f983814e37 90884Author: Peter Hutterer <peter.hutterer@who-t.net> 90885Date: Fri Jun 24 11:02:23 2011 +1000 90886 90887 input: add POINTER_NORAW to avoid generation of raw events (#30068) 90888 90889 RawEvents are supposed to be events coming from the driver. When warping the 90890 pointer, this should not generate a raw event. 90891 90892 X.Org Bug 30068 <http://bugs.freedesktop.org/show_bug.cgi?id=30068> 90893 90894 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90895 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 90896 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90897 90898commit 9444e40d77508d545e3c0a8123f61950c22d89fa 90899Author: Daniel Stone <daniel@fooishbar.org> 90900Date: Tue Mar 1 11:11:22 2011 +0000 90901 90902 Test: Input: Only ever set up to MAX_VALUATORS valuators 90903 90904 Previously, the input tests were working up to sizeof(mask) * 8, which 90905 could be more than the arrays; the latter only being sized as 90906 MAX_VALUATORS. 90907 90908 Hypothetically, if you were switching the stored valuator values to 90909 double instead of uint32_t, and you attempted to set the 39th member of 90910 a MAX_VALUATORS-sized (36) array, you'd probably end up smashing 90911 ev->key_repeat into oblivion and then tripping the check for invalid 90912 flags because you haven't yet put XIPointerEmulated into the valid flags 90913 for XI_Motion. Probably. 90914 90915 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 90916 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 90917 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90918 90919commit 1e65a01cf73f1fbdfe0d63eaecea412cf8472cb4 90920Author: Daniel Stone <daniel@fooishbar.org> 90921Date: Fri Jun 17 16:28:05 2011 +0100 90922 90923 Test: Ensure libxservertest gets relinked when necessary 90924 90925 Similar to how we link Xorg, make sure that whenever any of the 90926 component libraries changes, we relink libxservertest and the tests. 90927 Not much use testing anything other than the actual source in your tree. 90928 90929 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 90930 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 90931 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90932 90933commit 0f41d6c851df76d0423d9c37ad809d3440906944 90934Author: Daniel Stone <daniel@fooishbar.org> 90935Date: Fri Jun 17 13:20:19 2011 +0100 90936 90937 KDrive: Remove useless miPointerUpdateSprite call 90938 90939 miPointerUpdateSprite is already called from mieqProcessInputEvents, so 90940 calling it by hand immediately after isn't massively helpful. 90941 90942 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 90943 Reviewed-by: Cyril Brulebois <kibi@debian.org> 90944 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90945 90946commit 57b767d2927e5c2e561a431be4e604df65423422 90947Author: Daniel Stone <daniel@fooishbar.org> 90948Date: Fri Jun 17 13:19:46 2011 +0100 90949 90950 DMX: Remove useless miPointerUpdateSprite call 90951 90952 miPointerUpdateSprite is already called from mieqProcessInputEvents, so 90953 calling it by hand immediately after isn't massively helpful. 90954 90955 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 90956 Reviewed-by: Cyril Brulebois <kibi@debian.org> 90957 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90958 90959commit cf398755c0ca1e031514862b670956e7b9ace5ef 90960Author: Daniel Stone <daniel@fooishbar.org> 90961Date: Fri Jun 17 15:29:44 2011 +0100 90962 90963 XWin: Remove executable bit from headers 90964 90965 Headers don't really need to be mode 0755. 90966 90967 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 90968 Reviewed-by: Cyril Brulebois <kibi@debian.org> 90969 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 90970 90971commit 7d2543a3cb3089241982ce4f8984fd723d5312a1 90972Author: Daniel Stone <daniel@fooishbar.org> 90973Date: Wed Dec 29 12:03:01 2010 +0000 90974 90975 XKB: Add debug key actions for grabs & window tree 90976 90977 Add four new private XKB actions for debugging: 90978 * PrGrbs: print active grabs to the log file 90979 * Ungrab: ungrab all currently active grabs 90980 * ClsGrb: kill clients with active grabs 90981 * PrWins: dump the current window tree to the log file 90982 90983 To use these, you need to modify your XKB maps, e.g. the following to 90984 have Ctrl+Alt+(F9-F12) mapped to the above: 90985 - compat/xfree86: 90986 interpret XF86LogGrabInfo { 90987 action = Private(type=0x86, data="PrGrbs"); 90988 }; 90989 interpret XF86Ungrab { 90990 action = Private(type=0x86, data="Ungrab"); 90991 } 90992 interpret XF86ClearGrab { 90993 action = Private(type=0x86, data="ClsGrb"); 90994 } 90995 interpret XF86LogWindowTree { 90996 action = Private(type=0x86, data="PrWins"); 90997 } 90998 90999 - symbols/pc: 91000 key <FK09> { type="CTRL+ALT", [ Return, XF86LogGrabInfo ] }; 91001 key <FK10> { type="CTRL+ALT", [ Return, XF86Ungrab ] }; 91002 key <FK11> { type="CTRL+ALT", [ Return, XF86ClearGrab ] }; 91003 key <FK12> { type="CTRL+ALT", [ Return, XF86LogWindowTree ] }; 91004 91005 At the moment, this only works if the grabbing client continues to call 91006 AllowEvents, as the server does no event processing at all when a device 91007 is frozen. 91008 91009 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 91010 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 91011 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91012 91013commit ddf735fd4e9658bb5610f5e911f3b4055d5acf89 91014Author: Daniel Stone <daniel@fooishbar.org> 91015Date: Tue Oct 13 19:56:57 2009 +1100 91016 91017 DIX: Make PrintWindowTree actually useful 91018 91019 Rewrite PrintWindowTree to make it actually tell you what you want to 91020 know. 91021 91022 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 91023 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 91024 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91025 91026commit b27d61e4433a2a0140ed8a6128e4427451e37fec 91027Author: Matěj Cepl <mcepl@redhat.com> 91028Date: Fri Jun 17 15:26:17 2011 +0100 91029 91030 Fix UTF-8 encoding 91031 91032 Report to find out all non-UTF-8 files created by 91033 91034 cat extensions |xargs -I XXXX find . -name \*.XXXX |while read FILE ; do 91035 if ( iconv -f utf8 -t ucs2 $FILE >/dev/null 2>/dev/null ) ; then 91036 /bin/true 91037 else 91038 echo $FILE 91039 fi 91040 done >>report 91041 91042 Signed-off-by: Matěj Cepl <mcepl@redhat.com> 91043 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91044 91045 [Daniel: git am failed for me, so I redid it. The method listed in the 91046 commit message also failed, so I just used file/grep/iconv. The 91047 results are the same though.] 91048 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91049 91050commit 2ee85d954c8f3a6affbd9b1b745594ff12e8b670 91051Author: Scott James Remnant <scott@netsplit.com> 91052Date: Tue Jun 14 16:36:07 2011 -0700 91053 91054 dix: avoid calling deleted block and wakeup handlers 91055 91056 BlockHandler and WakeupHandlers may be removed within a different 91057 BlockHandler or WakeupHandler, especially since config/udev uses 91058 these and removes devices. 91059 91060 Calling the deleted handlers and passing potentially freed data 91061 can result in the X server segfaulting after device removal, or 91062 events that result in device removal such as undocking or suspend/ 91063 resume. 91064 91065 Signed-off-by: Scott James Remnant <scott@netsplit.com> 91066 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91067 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91068 91069commit b573cdd40ee49fae299fe2e7cbd02159ae8ae617 91070Author: Marcin Slusarz <marcin.slusarz@gmail.com> 91071Date: Tue Jun 7 21:22:15 2011 +0200 91072 91073 dri2: restore Screen->ConfigNotify on close 91074 91075 ConfigNotify is set by DRI2ScreenInit, but not restored to 91076 previous state on close. Fix it. 91077 (I'm preparing a patch for xf86-video-nouveau which detects GPU lockup 91078 after dri2 init and it needs to reinitialize dri2) 91079 91080 Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com> 91081 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91082 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91083 91084commit 61f87a75f2bcda939a1778d39be8cfa5c886e6d8 91085Merge: 0d2153a46 72d914335 91086Author: Keith Packard <keithp@keithp.com> 91087Date: Wed Jun 29 20:14:51 2011 -0700 91088 91089 Merge remote-tracking branch 'jturney/master' 91090 91091commit 0d2153a46cc72d5f2e6a9081a3cf153aa3eb7787 91092Merge: b631c39a2 932513e23 91093Author: Keith Packard <keithp@keithp.com> 91094Date: Wed Jun 29 20:08:32 2011 -0700 91095 91096 Merge remote-tracking branch 'jbarnes/master' 91097 91098commit b631c39a21c8d74fd00b8932df342d6921cce8f4 91099Author: Gaetan Nadon <memsize@videotron.ca> 91100Date: Thu Jun 23 13:37:45 2011 -0400 91101 91102 Remove unused check for PERL program 91103 91104 Unable to find any use of the PERL Automake variable. 91105 It was used in hw/xfree86/scanpci around 2005. 91106 Should it ever be needed, use XORG_WITH_PERL macro. 91107 91108 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91109 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 91110 Signed-off-by: Keith Packard <keithp@keithp.com> 91111 91112commit d5ca33ca2dd08e8436439da926d2dedaabc268fa 91113Merge: 38d55f06c 4edf49d03 91114Author: Keith Packard <keithp@keithp.com> 91115Date: Wed Jun 29 18:54:33 2011 -0700 91116 91117 Merge remote-tracking branch 'kibi/master' 91118 91119commit 72d914335c9ad0e0f4a34b9f7e2901b830c499bf 91120Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 91121Date: Tue Mar 23 04:34:22 2010 -0500 91122 91123 Cygwin/X: Left-justify website link in About box 91124 91125 Left-justify website link in About box. This is a cosmetic fix to make 91126 the About box display correctly when Windows is configured with a 91127 non-default DPI value 91128 91129 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 91130 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 91131 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 91132 Tested-by: Colin Harrison <colin.harrison@virgin.net> 91133 91134commit e78628406f0428454983db1c51a8c828bf5684e5 91135Author: Jon TURNEY <jon.turney@dronecode.org.uk> 91136Date: Thu Aug 12 15:00:01 2010 +0100 91137 91138 Cygwin/X: Preserve client area size and position on Windows window style change 91139 91140 When the style changes, adjust the window size so the client area remains the same. 91141 91142 Otherwise the window size may change when sizing is reflected from Windows to X, and 91143 some windows are drawn expecting them to be exactly the requested size (e.g. the 91144 gmplayer control window) 91145 91146 Use DeferWindowPos to delay the resize to preserve client area on WM_STYLECHANGING 91147 until after the style change has actually happened in WM_STYLECHANGED 91148 91149 As a consquence of this, we need to be more careful to create windows with exactly 91150 the requested placement and client area initially, so the client area matches what 91151 the X client requested 91152 91153 Also synchronize the X windows idea of the placement of a window which Windows is 91154 allowed to place 91155 91156 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 91157 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 91158 Tested-by: Colin Harrison <colin.harrison@virgin.net> 91159 91160commit 1d7f3a0031b4ae7d8aa984799ffa578788061ce4 91161Author: Jon TURNEY <jon.turney@dronecode.org.uk> 91162Date: Sat Nov 7 19:34:12 2009 +0000 91163 91164 glx: Move GLX extension string utility functions 91165 91166 Move GLX extension string utility functions from the DRI1 convenience library 91167 into the GLX extension convenience library, so other DDX which don't have DRI 91168 can use them. 91169 91170 This is probably also needed if anyone actually tries to build an Xorg DDX 91171 with only DRI2 support... 91172 91173 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 91174 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 91175 91176commit 932513e23b5fe1793cc38eabe02df82e1cc6e4d9 91177Author: Jesse Barnes <jbarnes@virtuousgeek.org> 91178Date: Thu May 5 13:11:46 2011 -0700 91179 91180 DRI2/GLX: use new swap event types 91181 91182 Use the new event types so we can pass a valid SBC value to clients. 91183 Fix up the completion calls to use CARD32 instead of CARD64 to match 91184 the new field size. 91185 91186 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91187 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 91188 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 91189 91190commit a3cb7d097a281252cebbc3c65c67149e106482ac 91191Author: Jesse Barnes <jbarnes@virtuousgeek.org> 91192Date: Tue May 3 10:36:44 2011 -0700 91193 91194 GLX/DRI2: pass drawable correctly for indirect swap events 91195 91196 Pass the right drawable pointer as data to the swap complete function. 91197 91198 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 91199 91200commit 3e8270b869ebf788778d5d56107f6d47c7c050d0 91201Author: Adam Jackson <ajax@redhat.com> 91202Date: Tue Apr 12 15:30:25 2011 -0400 91203 91204 glx: Fix fbconfigs with no corresponding visual 91205 91206 There are, in general, more fbconfig depths than visual depths. 91207 fbconfigs need not support Window rendering, however any that do must 91208 have an associated visual ID (which we got right), and any that do not 91209 must not claim GLX_WINDOW_BIT in GLX_DRAWABLE_TYPE (which we got 91210 wrong). 91211 91212 Fixes piglit/glx-fbconfig-sanity, assuming you have a sufficiently 91213 recent piglit, as that test formerly wrongly required pixmap-capable 91214 fbconfigs to have a visual. 91215 91216 v2: Additional check for fbconfigs that didn't have GLX_WINDOW_BIT in 91217 the first place, from previous patch by Jon TURNEY; also, also clear 91218 ->visualID. 91219 91220 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 91221 Signed-off-by: Adam Jackson <ajax@redhat.com> 91222 91223commit 38d55f06cfbf4b05fee0aef8f97f90e835beaacb 91224Author: Jon TURNEY <jon.turney@dronecode.org.uk> 91225Date: Mon Jun 27 15:06:16 2011 +0100 91226 91227 Guard use of backtrace() with HAVE_BACKTRACE 91228 91229 Guard the use of backtrace() with HAVE_BACKTRACE, since we already 91230 have the autoconf machinery for setting that. 91231 91232 For the moment, assume dladdr() is available when backtrace() is 91233 91234 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 91235 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 91236 91237commit bea57392c724f439987832b422941ae897953e4a 91238Author: Jeremy Huddleston <jeremyhu@apple.com> 91239Date: Sat Jun 25 11:29:53 2011 -0700 91240 91241 XQuartz: Fix Makefile dependencies 91242 91243 automake generates _DEPENDENCIES from _LIBADD, but it strips out variables. 91244 This resulted in not relinking if some components were rebuilt (like 91245 libdix, libos, etc). 91246 91247 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 91248 91249commit c8b80a82d97d92d445c1c0af6dba42c9de7c56cd 91250Author: Jeremy Huddleston <jeremyhu@apple.com> 91251Date: Tue Jun 21 21:14:46 2011 -0700 91252 91253 XQuartz: Make a call to activateIgnoringOtherApps in our NSApplicationActivatedEventType handler 91254 91255 In addition, this change will not call into the X11 activation unless an X11 91256 window was active when we deactivated. We can't rely on the event and current 91257 key windows because the key window will be nil until activated, and the event 91258 will only reference the window if the window was clicked (whereas it will be 91259 nil if we activated via dock or cmd-tab). 91260 91261 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 91262 91263commit 9d568450b1319e9057319ebb37e76003bcba447d 91264Author: Keith Packard <keithp@keithp.com> 91265Date: Tue Jun 28 09:33:25 2011 -0700 91266 91267 Create sdksyms.dep in the right place at configure time 91268 91269 sdksyms moved from hw/xfree86/loader to hw/xfree86, so the 91270 configure-time create of sdksyms.dep needs to reflect that 91271 change. Otherwise, make might be confused by a missing file and (more 91272 importantly to me) hw/xfree86/loader/sdksyms.dep will be left around 91273 after 'make clean' causing 'make distcheck' to fail. 91274 91275commit f968f4ace9410d827fb6b68c4e38ea9516641309 91276Merge: 4dbed2625 163441fe5 91277Author: Keith Packard <keithp@keithp.com> 91278Date: Tue Jun 28 09:10:14 2011 -0700 91279 91280 Merge remote-tracking branch 'dbn/no-libxorg' 91281 91282commit 4dbed2625431a92e20bede7b8cef847b5d5d99ba 91283Author: Peter Hutterer <peter.hutterer@who-t.net> 91284Date: Mon Jun 27 14:45:18 2011 +1000 91285 91286 xfree86: Only log the serial bytes if debug is on. 91287 91288 Introduced in e3f296d91dfe6b827195e1d387e1a04aa73b85c3, when the ifdef DEBUG 91289 around the whole block was removed, but only two of the three ErrorF 91290 switched to DebugF. 91291 91292 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91293 Signed-off-by: Keith Packard <keithp@keithp.com> 91294 91295commit 163441fe531a77f7fa54bfa59c5ab49c78f44306 91296Author: Dan Nicholson <dbn.lists@gmail.com> 91297Date: Tue Jun 14 20:03:06 2011 -0700 91298 91299 xfree86: Remove libxorg convenience library to speed up build 91300 91301 libxorg.la served to collect all the Xorg convenience libraries into one 91302 massive archive to link into Xorg. This made things easy for symbol 91303 resolution, but it tremendously slowed down the build since each change 91304 caused libxorg.la to be rebuilt. This is an extremely slow process of 91305 extracting all the objects from the sub-libraries and recombining them. 91306 91307 Instead, the archives are linked directly into Xorg. The order of the 91308 libraries had to be tweaked a bit to make symbols resolve correctly with 91309 the lower level code moving later in the link command. 91310 91311 As a side effect, since the dtrace objects are now being linked 91312 directly into Xorg, we don't need the SPECIAL_DTRACE_OBJECTS hack to 91313 add them twice. 91314 91315 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 91316 Tested-by: Jamey Sharp <jamey@minilop.net> 91317 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91318 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 91319 91320commit 6259b30111f568ec3d1b32dc6382bce46d8e6ccc 91321Author: Dan Nicholson <dbn.lists@gmail.com> 91322Date: Tue Jun 14 19:47:02 2011 -0700 91323 91324 xfree86: Move sdksyms generation to ddx toplevel 91325 91326 The symbols in sdksyms.c cover the entire source tree. In order to make 91327 them resolve when libxorg.la goes away, move the objects from libloader 91328 to Xorg. Unfortunately, this means sdksyms needs to get built again for 91329 the test code. 91330 91331 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 91332 Tested-by: Jamey Sharp <jamey@minilop.net> 91333 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91334 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 91335 91336commit 7d5c8a12cbc295b3e33e1b60bd7330e2bc93a966 91337Author: Dan Nicholson <dbn.lists@gmail.com> 91338Date: Tue Jun 14 17:21:48 2011 -0700 91339 91340 Don't use empty source files 91341 91342 When an empty _SOURCES variable is declared, automake will recognize that 91343 only linking is needed. 91344 91345 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 91346 Tested-by: Jamey Sharp <jamey@minilop.net> 91347 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91348 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 91349 91350commit 8ac651552bb70ba36238f430adab2f7a7f24db6e 91351Merge: 945b2ff81 47b6ba320 91352Author: Keith Packard <keithp@keithp.com> 91353Date: Wed Jun 22 11:03:37 2011 -0700 91354 91355 Merge remote-tracking branch 'alanc/master' 91356 91357commit 945b2ff8141c58344fc9558ee3853fad3e86134b 91358Author: Gaetan Nadon <memsize@videotron.ca> 91359Date: Wed Jun 8 15:18:12 2011 -0400 91360 91361 docbook.am: --path is also required for xsltproc 91362 91363 The XSLT processor also needs to know about the entities defined 91364 in xserver.ent. It removes error messages. 91365 91366 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91367 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 91368 Signed-off-by: Keith Packard <keithp@keithp.com> 91369 91370commit 6326b0f0746ded410672ad0270d1c8a0e9cb59f9 91371Author: Gaetan Nadon <memsize@videotron.ca> 91372Date: Wed Jun 8 15:17:59 2011 -0400 91373 91374 Update .gitignore as new tests have been added 91375 91376 Refactor to follow the pattern one Makefile, one .gitignore 91377 where needed. 91378 91379 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 91380 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 91381 Signed-off-by: Keith Packard <keithp@keithp.com> 91382 91383commit 47b6ba3204d650393a7255ce2af27b8c018bb586 91384Author: Alan Coopersmith <alan.coopersmith@oracle.com> 91385Date: Mon Jun 20 22:17:03 2011 -0700 91386 91387 Add xkeyboard-config to See Also of man pages referencing xkb 91388 91389 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91390 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 91391 91392commit cfbe471b586282aea89f7b4802c051f22301bf59 91393Author: Alan Coopersmith <alan.coopersmith@oracle.com> 91394Date: Mon Jun 20 22:06:00 2011 -0700 91395 91396 Xdmx.man: convert section references to substituted forms 91397 91398 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91399 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 91400 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 91401 91402commit c3278f3c1786e66019a39e2612761f74782a48ee 91403Author: Alan Coopersmith <alan.coopersmith@oracle.com> 91404Date: Sat Jun 11 10:56:08 2011 -0700 91405 91406 Xdmx.man: additional updates to modernize XKB information 91407 91408 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91409 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 91410 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 91411 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91412 91413commit 19e37067131175219eca7d2b11c2ee59af280575 91414Author: Alan Coopersmith <alan.coopersmith@oracle.com> 91415Date: Sat Jun 11 10:41:48 2011 -0700 91416 91417 Xdmx.man: Show actual configured XKB defaults instead of old hardcoded values 91418 91419 Passed through from configure.ac via manpages.am 91420 91421 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91422 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 91423 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 91424 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91425 91426commit acf2173b321749c42d29d858d3521a4d506c732e 91427Author: Gaetan Nadon <memsize@videotron.ca> 91428Date: Thu Jun 16 20:12:03 2011 -0400 91429 91430 Install xml versions of specs even if HAVE_XMLTO is false 91431 91432 DocBook/XML input source is also a useful output format that can be viewed 91433 with an XML viewer or editor and by some O/S help system. 91434 91435 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 91436 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91437 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 91438 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91439 91440commit 44890b3e266116ae71bfca2590e5fd9cdfbdb69b 91441Author: Gaetan Nadon <memsize@videotron.ca> 91442Date: Thu Jun 16 20:11:52 2011 -0400 91443 91444 Install target dbs alongside generated documents 91445 91446 This matches a change in xorg-sgml-docs whereby the masterdb will look for 91447 the target dbs into the same location as the generated documents. 91448 91449 The target dbs are now installed alongside the generated documents. 91450 Previously they are installed in $prefix/sgml/X11/dbs alongside masterdb which 91451 has the potential of installing outside the package prefix and cause 91452 distcheck to fail when user does not have write permission in this package. 91453 91454 Requires XORG_CHECK_SGML_DOCTOOLS(1.8) which was released 2011-06-11 91455 91456 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91457 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 91458 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91459 91460commit a42ef6c940604172e54de110b7d989cb2208f7ef 91461Author: Rob Clark <robdclark@gmail.com> 91462Date: Tue Jun 14 11:45:38 2011 -0500 91463 91464 EXA: fix typo 91465 91466 The incorrect drawable deltas were applied if dst was a redirected 91467 window. Resulting in a bogus region passed to prepare_access_reg(). 91468 91469 Signed-off-by: Rob Clark <rob@ti.com> 91470 Reviewed-by: Michel Dänzer <michel@daenzer.net> 91471 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91472 91473commit 4edf49d032028aa4b9887d02b926b0c372c4451d 91474Author: Cyril Brulebois <kibi@debian.org> 91475Date: Tue May 24 18:33:18 2011 +0200 91476 91477 dmx/examples: Fix missing key_click_percent assignment. 91478 91479 Thanks to gcc's -Wunused-but-set-variable, stop ignoring the percent 91480 parameter, and add it to the XKeyboardControl structure before the 91481 XChangeKeyboardControl call. 91482 91483 This warning goes away accordingly: 91484 | CC xbell-xbell.o 91485 | xbell.c: In function ‘main’: 91486 | xbell.c:74:22: warning: variable ‘percent’ set but not used [-Wunused-but-set-variable] 91487 91488 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91489 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91490 91491commit 33dad2b74e3489c8a6b5fa574172d6cd2b6de037 91492Author: Cyril Brulebois <kibi@debian.org> 91493Date: Tue May 24 18:33:17 2011 +0200 91494 91495 hw/dmx: Stop using variables for text widgets. 91496 91497 They are unused in the sense they're not getting any callback set up, so 91498 there's no point in storing them into variables. Keep the initial 91499 alignment of the parameters to try and reduce the diff noise. 91500 91501 Those warnings go away accordingly: 91502 | CC xdmxconfig-xdmxconfig.o 91503 | xdmxconfig.c: In function ‘main’: 91504 | xdmxconfig.c:881:29: warning: variable ‘quittext’ set but not used [-Wunused-but-set-variable] 91505 | xdmxconfig.c:880:53: warning: variable ‘abouttext’ set but not used [-Wunused-but-set-variable] 91506 91507 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91508 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91509 91510commit 34a9f2a90d34a2ab7ecf56d329e75e8da569f939 91511Author: Cyril Brulebois <kibi@debian.org> 91512Date: Tue May 24 18:33:17 2011 +0200 91513 91514 xfixes: Remove unused variable. 91515 91516 The last use of pScreen in ProcXFixesSetPictureClipRegion was removed in 91517 aa7096ca6f108e399d9916639cf20c57f9776305 so remove it entirely. 91518 91519 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91520 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91521 91522commit ad0d0833d45ea24d7af1f724b9905133fe3b065f 91523Author: Cyril Brulebois <kibi@debian.org> 91524Date: Tue May 24 18:33:17 2011 +0200 91525 91526 xfixes: Mark some variables as unused. 91527 91528 Calling Unwrap() is just a way of performing an assignment while keeping a 91529 backup of the original value. In the CursorCloseScreen function, the backup 91530 value doesn't seem to be useful, but Unwrap() is used anyway (probably to stay 91531 in line with other functions). As a consequence, mark those variables as unused. 91532 91533 The following warnings go away accordingly: 91534 | CC cursor.lo 91535 | cursor.c: In function 'CursorCloseScreen': 91536 | cursor.c:186:26: warning: variable 'display_proc' set but not used [-Wunused-but-set-variable] 91537 | cursor.c:185:24: warning: variable 'close_proc' set but not used [-Wunused-but-set-variable] 91538 91539 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91540 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91541 91542commit f41ac25d932ef34ab912201ab5ce65f85f2ff171 91543Author: Cyril Brulebois <kibi@debian.org> 91544Date: Tue May 24 18:33:17 2011 +0200 91545 91546 miext: Mark some variables as unused. 91547 91548 There's no use for the values set through the various macro calls 91549 (fbGetDrawable and fbGetDrawablePixmap), so mark those variables as unused. 91550 91551 The following warnings go away accordingly: 91552 | CC shpacked.lo 91553 | shpacked.c: In function 'shadowUpdatePacked': 91554 | shpacked.c:55:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91555 | shpacked.c:55:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91556 | CC shplanar8.lo 91557 | shplanar8.c: In function 'shadowUpdatePlanar4x8': 91558 | shplanar8.c:105:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91559 | shplanar8.c:105:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91560 | CC shplanar.lo 91561 | shplanar.c: In function 'shadowUpdatePlanar4': 91562 | shplanar.c:101:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91563 | shplanar.c:101:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91564 | CC shrot16pack_180.lo 91565 | In file included from shrot16pack_180.c:31:0: 91566 | shrotpack.h: In function 'shadowUpdateRotate16_180': 91567 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91568 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91569 | CC shrot16pack_270.lo 91570 | In file included from shrot16pack_270.c:31:0: 91571 | shrotpack.h: In function 'shadowUpdateRotate16_270': 91572 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91573 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91574 | CC shrot16pack_270YX.lo 91575 | In file included from shrot16pack_270YX.c:31:0: 91576 | shrotpackYX.h: In function 'shadowUpdateRotate16_270YX': 91577 | shrotpackYX.h:72:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91578 | shrotpackYX.h:72:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91579 | CC shrot16pack_90.lo 91580 | In file included from shrot16pack_90.c:31:0: 91581 | shrotpack.h: In function 'shadowUpdateRotate16_90': 91582 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91583 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91584 | CC shrot16pack_90YX.lo 91585 | In file included from shrot16pack_90YX.c:31:0: 91586 | shrotpackYX.h: In function 'shadowUpdateRotate16_90YX': 91587 | shrotpackYX.h:72:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91588 | shrotpackYX.h:72:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91589 | CC shrot16pack.lo 91590 | In file included from shrot16pack.c:30:0: 91591 | shrotpack.h: In function 'shadowUpdateRotate16': 91592 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91593 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91594 | CC shrot32pack_180.lo 91595 | In file included from shrot32pack_180.c:31:0: 91596 | shrotpack.h: In function 'shadowUpdateRotate32_180': 91597 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91598 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91599 | CC shrot32pack_270.lo 91600 | In file included from shrot32pack_270.c:31:0: 91601 | shrotpack.h: In function 'shadowUpdateRotate32_270': 91602 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91603 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91604 | CC shrot32pack_90.lo 91605 | In file included from shrot32pack_90.c:31:0: 91606 | shrotpack.h: In function 'shadowUpdateRotate32_90': 91607 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91608 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91609 | CC shrot32pack.lo 91610 | In file included from shrot32pack.c:30:0: 91611 | shrotpack.h: In function 'shadowUpdateRotate32': 91612 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91613 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91614 | CC shrot8pack_180.lo 91615 | In file included from shrot8pack_180.c:31:0: 91616 | shrotpack.h: In function 'shadowUpdateRotate8_180': 91617 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91618 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91619 | CC shrot8pack_270.lo 91620 | In file included from shrot8pack_270.c:31:0: 91621 | shrotpack.h: In function 'shadowUpdateRotate8_270': 91622 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91623 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91624 | CC shrot8pack_90.lo 91625 | In file included from shrot8pack_90.c:31:0: 91626 | shrotpack.h: In function 'shadowUpdateRotate8_90': 91627 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91628 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91629 | CC shrot8pack.lo 91630 | In file included from shrot8pack.c:30:0: 91631 | shrotpack.h: In function 'shadowUpdateRotate8': 91632 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91633 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91634 | CC shrotate.lo 91635 | shrotate.c: In function 'shadowUpdateRotatePacked': 91636 | shrotate.c:62:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91637 | shrotate.c:62:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91638 | CC shpacked.lo 91639 | shpacked.c: In function 'shadowUpdatePacked': 91640 | shpacked.c:55:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91641 | shpacked.c:55:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91642 | CC shplanar8.lo 91643 | shplanar8.c: In function 'shadowUpdatePlanar4x8': 91644 | shplanar8.c:105:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91645 | shplanar8.c:105:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91646 | CC shplanar.lo 91647 | shplanar.c: In function 'shadowUpdatePlanar4': 91648 | shplanar.c:101:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91649 | shplanar.c:101:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91650 | CC shrot16pack_180.lo 91651 | In file included from shrot16pack_180.c:31:0: 91652 | shrotpack.h: In function 'shadowUpdateRotate16_180': 91653 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91654 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91655 | CC shrot16pack_270.lo 91656 | In file included from shrot16pack_270.c:31:0: 91657 | shrotpack.h: In function 'shadowUpdateRotate16_270': 91658 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91659 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91660 | CC shrot16pack_270YX.lo 91661 | In file included from shrot16pack_270YX.c:31:0: 91662 | shrotpackYX.h: In function 'shadowUpdateRotate16_270YX': 91663 | shrotpackYX.h:72:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91664 | shrotpackYX.h:72:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91665 | CC shrot16pack_90.lo 91666 | In file included from shrot16pack_90.c:31:0: 91667 | shrotpack.h: In function 'shadowUpdateRotate16_90': 91668 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91669 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91670 | CC shrot16pack_90YX.lo 91671 | In file included from shrot16pack_90YX.c:31:0: 91672 | shrotpackYX.h: In function 'shadowUpdateRotate16_90YX': 91673 | shrotpackYX.h:72:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91674 | shrotpackYX.h:72:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91675 | CC shrot16pack.lo 91676 | In file included from shrot16pack.c:30:0: 91677 | shrotpack.h: In function 'shadowUpdateRotate16': 91678 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91679 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91680 | CC shrot32pack_180.lo 91681 | In file included from shrot32pack_180.c:31:0: 91682 | shrotpack.h: In function 'shadowUpdateRotate32_180': 91683 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91684 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91685 | CC shrot32pack_270.lo 91686 | In file included from shrot32pack_270.c:31:0: 91687 | shrotpack.h: In function 'shadowUpdateRotate32_270': 91688 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91689 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91690 | CC shrot32pack_90.lo 91691 | In file included from shrot32pack_90.c:31:0: 91692 | shrotpack.h: In function 'shadowUpdateRotate32_90': 91693 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91694 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91695 | CC shrot32pack.lo 91696 | In file included from shrot32pack.c:30:0: 91697 | shrotpack.h: In function 'shadowUpdateRotate32': 91698 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91699 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91700 | CC shrot8pack_180.lo 91701 | In file included from shrot8pack_180.c:31:0: 91702 | shrotpack.h: In function 'shadowUpdateRotate8_180': 91703 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91704 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91705 | CC shrot8pack_270.lo 91706 | In file included from shrot8pack_270.c:31:0: 91707 | shrotpack.h: In function 'shadowUpdateRotate8_270': 91708 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91709 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91710 | CC shrot8pack_90.lo 91711 | In file included from shrot8pack_90.c:31:0: 91712 | shrotpack.h: In function 'shadowUpdateRotate8_90': 91713 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91714 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91715 | CC shrot8pack.lo 91716 | In file included from shrot8pack.c:30:0: 91717 | shrotpack.h: In function 'shadowUpdateRotate8': 91718 | shrotpack.h:109:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91719 | shrotpack.h:109:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91720 | CC shrotate.lo 91721 | shrotate.c:62:19: warning: variable 'shaYoff' set but not used [-Wunused-but-set-variable] 91722 | shrotate.c:62:10: warning: variable 'shaXoff' set but not used [-Wunused-but-set-variable] 91723 91724 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91725 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91726 91727commit fe6f489d21d77aacf0a073883625619018d764b3 91728Author: Cyril Brulebois <kibi@debian.org> 91729Date: Tue May 24 18:33:17 2011 +0200 91730 91731 fb: Mark some variables as unused. 91732 91733 There's no use for the values set through the various macro calls 91734 (fbGetDrawable and fbGetDrawablePixmap), so mark those variables as unused. 91735 91736 The following warnings go away accordingly: 91737 | CC libfb_la-fb24_32.lo 91738 | fb24_32.c: In function 'fb24_32ReformatTile': 91739 | fb24_32.c:544:19: warning: variable 'newYoff' set but not used [-Wunused-but-set-variable] 91740 | fb24_32.c:544:10: warning: variable 'newXoff' set but not used [-Wunused-but-set-variable] 91741 | fb24_32.c:543:19: warning: variable 'oldYoff' set but not used [-Wunused-but-set-variable] 91742 | fb24_32.c:543:10: warning: variable 'oldXoff' set but not used [-Wunused-but-set-variable] 91743 | CC libfb_la-fbfill.lo 91744 | fbfill.c: In function 'fbFill': 91745 | fbfill.c:72:21: warning: variable 'stipYoff' set but not used [-Wunused-but-set-variable] 91746 | fbfill.c:72:11: warning: variable 'stipXoff' set but not used [-Wunused-but-set-variable] 91747 | fbfill.c:100:21: warning: variable 'stipYoff' set but not used [-Wunused-but-set-variable] 91748 | fbfill.c:100:11: warning: variable 'stipXoff' set but not used [-Wunused-but-set-variable] 91749 | fbfill.c:142:20: warning: variable 'tileYoff' set but not used [-Wunused-but-set-variable] 91750 | fbfill.c:142:10: warning: variable 'tileXoff' set but not used [-Wunused-but-set-variable] 91751 | CC libfb_la-fbgc.lo 91752 | fbgc.c: In function 'fbPadPixmap': 91753 | fbgc.c:92:19: warning: variable 'yOff' set but not used [-Wunused-but-set-variable] 91754 | fbgc.c:92:13: warning: variable 'xOff' set but not used [-Wunused-but-set-variable] 91755 | fbgc.c: In function 'fbCanEvenStipple': 91756 | fbgc.c:166:23: warning: variable 'stipYoff' set but not used [-Wunused-but-set-variable] 91757 | fbgc.c:166:13: warning: variable 'stipXoff' set but not used [-Wunused-but-set-variable] 91758 | CC libfb_la-fbpush.lo 91759 | fbpush.c: In function 'fbPushPixels': 91760 | fbpush.c:238:20: warning: variable 'stipYoff' set but not used [-Wunused-but-set-variable] 91761 | fbpush.c:238:10: warning: variable 'stipXoff' set but not used [-Wunused-but-set-variable] 91762 91763 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91764 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91765 91766commit 8da16898d9de5da1f80b789f0b371070968285f8 91767Author: Cyril Brulebois <kibi@debian.org> 91768Date: Tue May 24 18:33:17 2011 +0200 91769 91770 fb: Simplify logic, get rid of set but unused variable. 91771 91772 wrapped is only useful is FB_ACCESS_WRAPPER is set; simplify the logic 91773 accordingly, and only set it when that's defined. 91774 91775 The following warning goes away accordingly: 91776 | CC libfb_la-fbarc.lo 91777 | fbarc.c: In function 'fbPolyArc': 91778 | fbarc.c:71:11: warning: variable 'wrapped' set but not used [-Wunused-but-set-variable] 91779 91780 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91781 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91782 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91783 91784commit 9a8bfc2c06f0a9206b64ae01a783cf5513f021a6 91785Author: Cyril Brulebois <kibi@debian.org> 91786Date: Tue May 24 18:33:16 2011 +0200 91787 91788 configure: Bump dependency on xproto. 91789 91790 Bumping this dependency means some additional macros can be used, like 91791 _X_UNUSED. 91792 91793 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91794 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91795 91796commit 9da5aa09d70e26dd4d2ef878a21317dae7bf24de 91797Author: Cyril Brulebois <kibi@debian.org> 91798Date: Tue May 24 18:33:11 2011 +0200 91799 91800 Xext: Fix buggy checks. 91801 91802 XvMCScreenKey is defined as XvMCScreenKeyRec's address, so will never be 91803 NULL. Use the check proposed by Alan Coopersmith and Jeremy Huddleston 91804 instead. 91805 91806 Those warnings go away accordingly: 91807 | CC xvmc.lo 91808 | xvmc.c: In function 'ProcXvMCListSubpictureTypes': 91809 | xvmc.c:490:22: warning: the comparison will always evaluate as 'false' for the address of 'XvMCScreenKeyRec' will never be NULL [-Waddress] 91810 | xvmc.c: In function 'XvMCExtensionInit': 91811 | xvmc.c:671:21: warning: the comparison will always evaluate as 'false' for the address of 'XvMCScreenKeyRec' will never be NULL [-Waddress] 91812 | xvmc.c: In function 'XvMCFindXvImage': 91813 | xvmc.c:749:22: warning: the comparison will always evaluate as 'false' for the address of 'XvMCScreenKeyRec' will never be NULL [-Waddress] 91814 91815 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91816 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91817 91818commit 778309fd0ef6b586468c1a3b569e3548503459ee 91819Author: Jeremy Huddleston <jeremyhu@apple.com> 91820Date: Tue Jun 14 13:23:00 2011 -0700 91821 91822 XQuartz: GLX: Fix a -Wformat-security compilation warning 91823 91824 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 91825 91826commit 34e2598f0ad247071bd6a4312d9014d6e3b2305a 91827Author: Jeremy Huddleston <jeremyhu@apple.com> 91828Date: Tue Jun 14 11:18:50 2011 -0700 91829 91830 XQuartz: GLX: Use _glapi_create_table_from_handle to create out dispatch table 91831 91832 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 91833 91834commit ecec578e35f91a2cbc5d07bc8d45241af7bb585f 91835Author: Jeremy Huddleston <jeremyhu@apple.com> 91836Date: Tue Jun 14 12:51:04 2011 -0700 91837 91838 glx: Add _glapi_create_table_from_handle 91839 91840 _glapi_create_table_from_handle was recently added to mesa to ease creation of 91841 a _glapi_table. This commit brings this API over to xserver for use in 91842 XQuartz (next commit). 91843 91844 The API was generated with scripts from current mesa followed by various 91845 script-foo to remove entries that are not in the server and add those which 91846 are still in the server but not in mesa (CullParameterdvEXT and 91847 CullParameterfvEXT) 91848 91849 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 91850 Reviewed-by: Adam Jackson <ajax@redhat.com> 91851 91852commit c3df4df40761760b281a21f63336ca955670f7bc 91853Author: Jeremy Huddleston <jeremyhu@apple.com> 91854Date: Sun Jun 5 21:59:13 2011 -0400 91855 91856 XQuartz: Add missing include for <pthread.h> 91857 91858 BuildFix for Leopard and earlier. 91859 91860 https://trac.macports.org/ticket/29717 91861 91862 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 91863 91864commit 08a7246f43c9f18079219a1c982bed558722f545 91865Author: Peter Hutterer <peter.hutterer@who-t.net> 91866Date: Fri May 27 14:21:39 2011 +1000 91867 91868 dix: rename ProcessRawEvents to dix/events.c:DeliverRawEvent 91869 91870 No functional changes, prep work for future changes. 91871 91872 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91873 Reviewed-by: Cyril Brulebois <kibi@debian.org> 91874 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91875 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91876 91877commit 0aa45c5c53e2e7ba0723f5a12c1a5f5d4baab933 91878Author: Peter Hutterer <peter.hutterer@who-t.net> 91879Date: Fri May 27 14:19:45 2011 +1000 91880 91881 Xi: use temporary variable for filter. 91882 91883 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91884 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91885 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91886 91887commit a1beae635caf81ec31a8998687df0ae92b505fda 91888Author: Peter Hutterer <peter.hutterer@who-t.net> 91889Date: Fri May 27 12:14:32 2011 +1000 91890 91891 dix: split DeliverEventToWindowMask up a bit more. 91892 91893 Move out the actual event delivery, it needs to be used from elsewhere. 91894 91895 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91896 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91897 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91898 91899commit 58e4831ed68c2fbc219c7250e9837e76a73a6432 91900Author: Peter Hutterer <peter.hutterer@who-t.net> 91901Date: Fri May 27 11:43:50 2011 +1000 91902 91903 dix: rename DeliverEventsToClients to DeliverEventsToWindowMask 91904 91905 To avoid confusion with a future patch and it better describes what this 91906 does anyway - delivering events to all clients that have the event mask on 91907 the window. 91908 91909 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91910 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91911 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91912 91913commit 37a53f15b0dfa1b25f7e3e826c7c945453398ac5 91914Author: Peter Hutterer <peter.hutterer@who-t.net> 91915Date: Fri May 27 11:40:56 2011 +1000 91916 91917 dix: split client list retrieval out of DeliverEventToClients 91918 91919 No functional change, but "other" was renamed to "clients". 91920 91921 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91922 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91923 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91924 91925commit 02d11af01211da55e9d93fe0e1851a0c6fe57472 91926Merge: bf9fd0a83 ab0df72cd 91927Author: Keith Packard <keithp@keithp.com> 91928Date: Tue Jun 7 12:31:15 2011 -0700 91929 91930 Merge remote-tracking branch 'kibi/master' 91931 91932commit bf9fd0a83eef771ee018438f05e2d9f20b43539b 91933Merge: 21eec367d 0d140567b 91934Author: Keith Packard <keithp@keithp.com> 91935Date: Tue Jun 7 07:45:22 2011 -0700 91936 91937 Merge remote-tracking branch 'whot/for-keith' 91938 91939commit 0d140567ba2b5bf364265dba07de519f691c67a1 91940Author: Peter Hutterer <peter.hutterer@who-t.net> 91941Date: Tue Jun 7 10:42:50 2011 +1000 91942 91943 dix: fix crashers with floating device. 91944 91945 dc57f89959e549403f8488eb9f23425bd7118b22 accidentally reversed the 91946 conditions. 91947 91948 in dix/events.c we try to detach floating devices. This leads to a 91949 NULL-dereference on GetMaster()->id. 91950 91951 in dix/getevents.c we try to get the master device for the floating slave 91952 and dereference it. 91953 91954 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 91955 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91956 91957commit 21eec367d0d9c142458f973062e17add4f693033 91958Author: Nicolas Kaiser <nikai@nikai.net> 91959Date: Mon Jun 6 16:53:41 2011 +0200 91960 91961 randr: void function cannot return value 91962 91963 Providing an argument to return in a function with void return type 91964 is not allowed by the C standard, and makes the Sun compilers unhappy. 91965 91966 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 91967 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 91968 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 91969 Signed-off-by: Keith Packard <keithp@keithp.com> 91970 91971commit ab0df72cd33f011cc48736cfbfce9bbe9c359a80 91972Author: Cyril Brulebois <kibi@debian.org> 91973Date: Sun Jun 5 03:21:18 2011 +0200 91974 91975 xkb: Fix case checks for Latin 4. 91976 91977 That one was missing _XkbKSLower: 91978 XK_kra: U+0138 LATIN SMALL LETTER KRA 91979 91980 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91981 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91982 91983commit 9eb6e34c2dd09af714400ce6a1063477d7a64776 91984Author: Cyril Brulebois <kibi@debian.org> 91985Date: Sun Jun 5 03:12:44 2011 +0200 91986 91987 xkb: Fix case checks for Latin 2. 91988 91989 Those ones were getting _XkbKSLower for no reasons: 91990 XK_ogonek: U+02DB OGONEK 91991 XK_doubleacute: U+02DD DOUBLE ACUTE ACCENT 91992 91993 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 91994 Signed-off-by: Cyril Brulebois <kibi@debian.org> 91995 91996commit 2b88189863579eb377bca617465bb76b05fbdebd 91997Author: Cyril Brulebois <kibi@debian.org> 91998Date: Sun Jun 5 03:03:47 2011 +0200 91999 92000 xkb: Fix case checks for Latin 1. 92001 92002 That one was missing _XkbKSLower: 92003 XK_ssharp: U+00DF LATIN SMALL LETTER SHARP S 92004 92005 That one was getting _XkbKSLower for no reasons: 92006 XK_division: U+00F7 DIVISION SIGN 92007 92008 For reference, XK_multiply was already excluded from the _XkbKSUpper 92009 check, it's no big surprise XK_division has to be excluded from the 92010 _XkbKSLower check. 92011 92012 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 92013 Signed-off-by: Cyril Brulebois <kibi@debian.org> 92014 92015commit feab04397de2684568dedaaaa8f299cac9f44f8b 92016Author: Jeremy Huddleston <jeremyhu@apple.com> 92017Date: Fri Jun 3 02:50:40 2011 -0400 92018 92019 XQuartz: AIGLX: Remove unnecessary includes in indirect.c 92020 92021 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92022 92023commit 3fa5d27e46ea8af55dcaaad7a99973c38bd11967 92024Author: Jeremy Huddleston <jeremyhu@apple.com> 92025Date: Thu Jun 2 20:04:04 2011 -0700 92026 92027 XQuartz: AIGLX: Setup dispatch table based on runtime capabilities rather than buildtime 92028 92029 This will allow XQuartz built on older systems to pickup capabilities on 92030 newer systems and prevent runtime failures when building on newer systems and 92031 running on older ones. 92032 92033 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92034 92035commit dd5d91fc954450e99d53862900ef4fa6320ff382 92036Author: Jeremy Huddleston <jeremyhu@apple.com> 92037Date: Fri Jun 3 02:34:28 2011 -0400 92038 92039 XQuartz: Remove explicit link against libGL 92040 92041 This was indented to force a link against OpenGL.framework's libGL, but it 92042 actually resulted in linking against mesa's libGL due to the ordering of -L 92043 92044 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92045 92046commit 5115c2c7216fc10acb8a26deb904ad15ef18ce8a 92047Author: Cyril Brulebois <kibi@debian.org> 92048Date: Fri Jun 3 18:09:51 2011 +0200 92049 92050 Xephyr: Mention Xserver's options are also accepted. 92051 92052 Originally reported in the Debian BTS: 92053 http://bugs.debian.org/582650 92054 92055 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 92056 Signed-off-by: Jari Aalto <jari.aalto@cante.net> 92057 Signed-off-by: Cyril Brulebois <kibi@debian.org> 92058 92059commit 408ed1576dbcf93f4bd27bd3b66917b1bba2226b 92060Author: Cyril Brulebois <kibi@debian.org> 92061Date: Fri Jun 3 18:09:48 2011 +0200 92062 92063 xkb: Fix case checks for Latin 8. 92064 92065 Spotted by -Wlogical-op: 92066 | CC xkbfmisc.lo 92067 | xkbfmisc.c: In function '_XkbKSCheckCase': 92068 | xkbfmisc.c:104:3: warning: logical 'and' of mutually exclusive tests is always false [-Wlogical-op] 92069 | xkbfmisc.c:118:3: warning: logical 'and' of mutually exclusive tests is always false [-Wlogical-op] 92070 92071 A quick look at the keysymdef.h file (from xproto) suggests the 92072 implementor chose to use interval checks to determine the case, but 92073 since lines weren't sorted by codepoints, checks were quite wrong. 92074 92075 Implement _XkbKSUpper/_XkbKSLower checks based on a grep for 92076 CAPITAL/SMALL (respectively) on the Latin 8 part of the said file. 92077 92078 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 92079 Signed-off-by: Cyril Brulebois <kibi@debian.org> 92080 92081commit d9bda34d0df576d155e1d682d5e2a382b8e1ffda 92082Author: Peter Hutterer <peter.hutterer@who-t.net> 92083Date: Wed Jun 1 10:24:43 2011 +1000 92084 92085 test: fix memset size for WindowRec (#37801) 92086 92087 X.Org Bug 37801 <http://bugs.freedesktop.org/show_bug.cgi?id=37801> 92088 92089 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92090 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 92091 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 92092 Reviewed-by: Cyril Brulebois <kibi@debian.org> 92093 Signed-off-by: Keith Packard <keithp@keithp.com> 92094 92095commit bfd8422e8877acf02155ca0bd7fbd2416b029720 92096Author: Peter Hutterer <peter.hutterer@who-t.net> 92097Date: Fri May 27 12:24:21 2011 +1000 92098 92099 Xi: use __func__ instead of function name. 92100 92101 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92102 Reviewed-by: Cyril Brulebois <kibi@debian.org> 92103 92104commit c042a267c40d35b4fb947f4bafdfeb541dd9d772 92105Author: Peter Hutterer <peter.hutterer@who-t.net> 92106Date: Fri May 27 10:10:48 2011 +1000 92107 92108 dix: use xi2_get_type instead of manual typecast 92109 92110 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92111 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 92112 92113commit b75cdb5bf76303162e947173e8ee68253a861272 92114Author: Peter Hutterer <peter.hutterer@who-t.net> 92115Date: Tue May 31 10:51:19 2011 +1000 92116 92117 dix: drop x/y back into the right valuators after transformation. 92118 92119 If the matrix is used for rotation, the coordinates affected may change. 92120 e.g. a valuator mask of (x, nil) becomes [x, lasty] and is rotated to 92121 [lasty, x]. Since the second value was unset, we would not drop x back into 92122 the mask, resulting in a loss of movement. 92123 92124 Thus, drop any value that changed after applying the matrix into the 92125 valuators. Thus, the example above becomes 92126 (x, nil) → [x, lasty] → [lasty, x] → (lasty, x) 92127 92128 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92129 Reviewed-by: Simon Thum <simon.thum@gmx.de> 92130 92131commit fe4b818700453b76d9f46749dac5d12540b91e63 92132Author: Peter Hutterer <peter.hutterer@who-t.net> 92133Date: Tue May 31 10:23:57 2011 +1000 92134 92135 dix: don't pass x/y to transformAbsolute 92136 92137 We passed in the mask, but didn't do anything with it. Move the logic to 92138 take the axes out of the valuator masks into transformAbsolute. 92139 92140 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92141 Reviewed-by: Simon Thum <simon.thum@gmx.de> 92142 92143commit 62f1bf96e1ea27e1eff137cb9333f3dd3b014dd0 92144Author: Peter Hutterer <peter.hutterer@who-t.net> 92145Date: Wed Jun 1 10:22:44 2011 +1000 92146 92147 test: fix memset size for WindowRec (#37801) 92148 92149 X.Org Bug 37801 <http://bugs.freedesktop.org/show_bug.cgi?id=37801> 92150 92151 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92152 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 92153 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 92154 Reviewed-by: Cyril Brulebois <kibi@debian.org> 92155 92156commit f94898928dfaf8df648bc3eea7d304826b7ac1b8 92157Author: Peter Hutterer <peter.hutterer@who-t.net> 92158Date: Tue May 31 14:48:33 2011 +1000 92159 92160 dix: fix an error message. 92161 92162 event type, not device type. 92163 92164 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92165 92166commit 8ef0974874f322e3ce2f3e6be4ab3c7e73a7b380 92167Author: Peter Hutterer <peter.hutterer@who-t.net> 92168Date: Thu May 19 14:50:17 2011 +1000 92169 92170 test: don't test for double alignment on i386. (#36986) 92171 92172 i386 is one of the few architectures that doesn't need double alignment. 92173 92174 X.Org Bug 36986 <http://bugs.freedesktop.org/show_bug.cgi?id=36986> 92175 92176 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92177 92178commit 48b4b778f61daaa161483b510ab520f3fcb84925 92179Author: Jeremy Huddleston <jeremyhu@apple.com> 92180Date: Wed Jun 1 21:55:06 2011 -0700 92181 92182 XQuartz: xpbproxy: Correct NSUInteger format strings 92183 92184 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92185 92186commit 971193b2750b98b28597ba194ed388e79b3489b6 92187Author: Jeremy Huddleston <jeremyhu@apple.com> 92188Date: Wed Jun 1 21:18:03 2011 -0700 92189 92190 XQuartz: ASL: Use xpbproxy subsystem for logging xpbproxy messages 92191 92192 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92193 92194commit 7dfd65705ccd3c6e2458d6892282520fe5ed8a6b 92195Author: Jeremy Huddleston <jeremyhu@apple.com> 92196Date: Wed Jun 1 21:06:53 2011 -0700 92197 92198 XQuartz: ASL: Use GLXAqua subsystem for logging GLXAqua messages 92199 92200 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92201 92202commit 19020b23efa15d72349bd833aebddcc64780a615 92203Author: Jeremy Huddleston <jeremyhu@apple.com> 92204Date: Wed Jun 1 21:06:08 2011 -0700 92205 92206 XQuartz: ASL: Use xpr subsystem for logging xpr messages 92207 92208 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92209 92210commit a66400f8dba9b7c5b65f9d26996fff677c42d866 92211Author: Jeremy Huddleston <jeremyhu@apple.com> 92212Date: Wed Jun 1 20:45:58 2011 -0700 92213 92214 XQuartz: ASL: Update logging to support differnet levels and subsystems 92215 92216 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92217 92218commit c0077b41b4c8bb08129db97a2a01d0af55d13fa4 92219Author: Jeremy Huddleston <jeremyhu@apple.com> 92220Date: Wed Jun 1 20:54:27 2011 -0700 92221 92222 XQuartz: Add a semicolon after DEBUG_LOG 92223 92224 Don't rely on the macro to provide it 92225 92226 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92227 92228commit fbc8aae8e9b5c7b175073f02e79d7ff1d1363a31 92229Author: Jeremy Huddleston <jeremyhu@apple.com> 92230Date: Sun May 29 01:20:44 2011 -0700 92231 92232 XQuartz: Move -lXplugin to LDFLAGS from LDADD 92233 92234 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92235 92236commit 38df706c6b8a339c4267cae7be0881c4362f36da 92237Author: Jeremy Huddleston <jeremyhu@apple.com> 92238Date: Tue May 31 13:01:05 2011 -0700 92239 92240 XQuartz: GLX: Create a new dispatch table rather than modifying the existing one 92241 92242 Fixes regression introduced by b0c665ac0fe6840dda581e4d0d0b76c703d62a7b 92243 92244 0 X11.bin 0x0000000100118293 __glXAquaScreenCreateContext + 684 92245 1 X11.bin 0x00000001001315b0 DoCreateContext + 163 92246 2 X11.bin 0x000000010013509f __glXDispatch + 211 92247 3 X11.bin 0x00000001000c7dad Dispatch + 785 92248 4 X11.bin 0x00000001000b97e5 dix_main + 1022 92249 5 X11.bin 0x00000001000122bc server_thread + 50 92250 6 libSystem.B.dylib 0x00007fff836554f6 _pthread_start + 331 92251 7 libSystem.B.dylib 0x00007fff836553a9 thread_start + 13 92252 92253 http://lists.apple.com/archives/X11-users/2011/May/msg00045.html 92254 92255 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92256 Reviewed-by: Adam Jackson <ajax@redhat.com> 92257 92258commit c5b72fd350bbdfd1facd0ddd5085f238c4cf252a 92259Author: Marko Macek <Marko.Macek@gmx.net> 92260Date: Sat May 21 13:30:59 2011 +0100 92261 92262 DIX: Set backgroundState correctly for root window 92263 92264 When we change the root window's background to None, and we've run with 92265 -wr or -br for a forced solid background, make sure we also change the 92266 background state to BackgroundPixel, so we don't try to lookup either 92267 pScreen->whitePixel or pScreen->blackPixel as a pixmap. 92268 92269 Signed-off-by: Marko Macek <Marko.Macek@gmx.net> 92270 Reviewed-by: Walter Harms <wharms@bfs.de> 92271 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 92272 Signed-off-by: Keith Packard <keithp@keithp.com> 92273 92274commit db228d3d07a6d831f53e4c05c878327ad4d045eb 92275Author: Keith Packard <keithp@keithp.com> 92276Date: Wed Jun 1 11:20:10 2011 -0700 92277 92278 Version bumped to 1.10.99.901 (1.11 RC1) 92279 92280 Signed-off-by: Keith Packard <keithp@keithp.com> 92281 92282commit 0643c056512d10be8db223d18d6563292d57e916 92283Merge: a2e6cfc18 d45f5b249 92284Author: Keith Packard <keithp@keithp.com> 92285Date: Tue May 31 23:45:07 2011 -0700 92286 92287 Merge remote-tracking branch 'ajax/xserver-next' 92288 92289commit a2e6cfc18aec1c0027b51572b03ec9f2ab074b56 92290Merge: bc04065b5 8d84fd256 92291Author: Keith Packard <keithp@keithp.com> 92292Date: Tue May 31 23:42:52 2011 -0700 92293 92294 Merge remote-tracking branch 'sandmann/for-keithp' 92295 92296commit d45f5b2493bc0a2882bf972849b5c9c50cd533ca 92297Author: Adam Jackson <ajax@redhat.com> 92298Date: Wed May 25 05:54:35 2011 -0400 92299 92300 fixes: Add support for pointer barriers 92301 92302 Implements pointer barriers as specified by version 5 of the XFIXES 92303 protocol. Barriers are axis-aligned, zero-width lines that block pointer 92304 movement for relative input devices. Barriers may block motion in either 92305 the positive or negative direction, or both. 92306 92307 v3: 92308 - Fix off-by-one in version_requests array 92309 - Port to non-glib test harness 92310 - Fix review notes from Søren Sandmann Pedersen, add tests to match 92311 92312 Co-authored-by: Peter Hutterer <peter.hutterer@who-t.net> 92313 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 92314 Signed-off-by: Adam Jackson <ajax@redhat.com> 92315 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92316 92317commit bc04065b5ce277f3ac3491ff221a60ef3c7605cf 92318Author: Alan Coopersmith <alan.coopersmith@oracle.com> 92319Date: Fri May 20 19:24:34 2011 -0700 92320 92321 "privates.h", line 198: warning: void function cannot return value 92322 92323 Providing an argument to return in a function with void return type 92324 is not allowed by the C standard, and makes the Sun compilers unhappy. 92325 (They actually flag it as an error, unless using a new enough version 92326 to be able to downgrade it to a warning with "-features=extensions".) 92327 92328 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 92329 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 92330 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 92331 Reviewed-by: Cyril Brulebois <kibi@debian.org> 92332 92333commit eadf5021794782fde861d471ed408675f4926b89 92334Author: Alan Coopersmith <alan.coopersmith@oracle.com> 92335Date: Mon May 2 19:48:42 2011 -0700 92336 92337 Use XORG_STRICT_OPTION from util-macros 1.14 to set -Werror flags 92338 92339 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 92340 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 92341 92342commit 9275b1fb6f82a6971c4177ddd3d5a859a8f24119 92343Author: Alan Coopersmith <alan.coopersmith@oracle.com> 92344Date: Mon May 2 19:47:44 2011 -0700 92345 92346 Use XORG_COMPILER_BRAND from util-macros 1.14 to check for SUNCC 92347 92348 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 92349 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 92350 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 92351 92352commit 4621bb270a36d35d4ab67f1d7fb47674683dfc5b 92353Author: Peter Hutterer <peter.hutterer@who-t.net> 92354Date: Wed May 18 15:00:54 2011 +1000 92355 92356 Add a property for device/product ID. 92357 92358 In some cases, knowing about the device model number and the device's vendor 92359 is important to activate product-specific settings. Since this is 92360 nonetheless driver-specific, only provide the property but don't do anything 92361 with it. 92362 92363 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92364 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 92365 92366commit f26a5b9a994a980ab309bc7f3b4295a4992d81c2 92367Author: Peter Hutterer <peter.hutterer@who-t.net> 92368Date: Mon Jan 31 14:43:01 2011 +1000 92369 92370 Xi: add device node property to known properties. 92371 92372 Since the server has little choice (or even knowledge) of the actual device 92373 node used by the driver, this property is merely provided for 92374 standardisation. It is up to the driver to set it to the appropriate value, 92375 usually a device node in the form of /dev/input/event0 or similar. 92376 92377 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92378 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 92379 92380commit 8d84fd2566f8466b6152724df7eefc73780df093 92381Author: Søren Sandmann Pedersen <ssp@redhat.com> 92382Date: Wed May 25 12:14:05 2011 -0400 92383 92384 Don't call pixman_disable_out_of_bounds_workaround() anymore 92385 92386 Pixman used to have a workaround for a bug in old X servers, and this 92387 function was used to disable that workaround in servers known to be 92388 fixed. 92389 92390 Since 0.22, which the X server depends on, the workaround doesn't 92391 exist anymore, so there is no point disabling it. 92392 92393 Reviewed-by: Cyril Brulebois <kibi at debian.org> 92394 Signed-off-by: Soren Sandmann <sandmann@cs.au.dk> 92395 92396commit b6c7b9b2f39e970cedb6bc1e073f901e28cb0fa3 92397Author: Aaron Plattner <aplattner@nvidia.com> 92398Date: Tue May 24 16:02:42 2011 -0700 92399 92400 randr: check rotated virtual size limits correctly 92401 92402 Commit d1107918d4626268803b54033a07405122278e7f introduced checks to 92403 the RandR path that cause RRSetScreenConfig requests to fail if the 92404 size is too large. Unfortunately, when RandR 1.1 rotation is enabled 92405 it compares the rotated screen dimensions to the unrotated limits, 92406 which causes 90- and 270-degree rotation to fail unless your screen 92407 happens to be square: 92408 92409 X Error of failed request: BadValue (integer parameter out of range for operation) 92410 Major opcode of failed request: 153 (RANDR) 92411 Minor opcode of failed request: 2 (RRSetScreenConfig) 92412 Value in failed request: 0x780 92413 Serial number of failed request: 14 92414 Current serial number in output stream: 14 92415 92416 Fix this by moving the check above the code that swaps the dimensions 92417 based on the rotation. 92418 92419 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 92420 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 92421 Tested-by: Robert Hooker <robert.hooker@canonical.com> 92422 Tested-by: Kent Baxley <kent.baxley@canonical.com> 92423 Signed-off-by: Keith Packard <keithp@keithp.com> 92424 92425commit 4725d6b0dbb6371af4a1b2fbef851dcfe155514a 92426Merge: 1b2d17748 b5d828789 92427Author: Keith Packard <keithp@keithp.com> 92428Date: Wed May 25 09:40:51 2011 -0600 92429 92430 Merge remote-tracking branch 'whot/for-keith' 92431 92432commit 1b2d17748f0154da142e9b421d4f6a46e4e5a18c 92433Author: Alan Coopersmith <alan.coopersmith@oracle.com> 92434Date: Sun May 22 13:50:14 2011 -0700 92435 92436 fbbltone.c: Mark bitmasks as unsigned ints 92437 92438 Clears many Sun compiler warnings: 92439 "fbbltone.c", line 491: warning: integer overflow detected: op "<<" 92440 "fbbltone.c", line 491: warning: integer overflow detected: op "<<" 92441 "fbbltone.c", line 491: warning: integer overflow detected: op "<<" 92442 "fbbltone.c", line 491: warning: initializer will be sign-extended: -16777216 92443 "fbbltone.c", line 491: warning: integer overflow detected: op "<<" 92444 "fbbltone.c", line 491: warning: initializer will be sign-extended: -1 92445 "fbbltone.c", line 495: warning: integer overflow detected: op "<<" 92446 "fbbltone.c", line 495: warning: integer overflow detected: op "<<" 92447 "fbbltone.c", line 495: warning: integer overflow detected: op "<<" 92448 "fbbltone.c", line 495: warning: initializer will be sign-extended: -256 92449 "fbbltone.c", line 495: warning: integer overflow detected: op "<<" 92450 "fbbltone.c", line 495: warning: initializer will be sign-extended: -1 92451 "fbbltone.c", line 499: warning: integer overflow detected: op "<<" 92452 "fbbltone.c", line 499: warning: integer overflow detected: op "<<" 92453 "fbbltone.c", line 499: warning: integer overflow detected: op "<<" 92454 "fbbltone.c", line 499: warning: initializer will be sign-extended: -65536 92455 "fbbltone.c", line 499: warning: integer overflow detected: op "<<" 92456 "fbbltone.c", line 499: warning: initializer will be sign-extended: -1 92457 92458 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 92459 Reviewed-by: Jamey Sharp <jamey@minilop.net> 92460 Signed-off-by: Keith Packard <keithp@keithp.com> 92461 92462commit b5d828789c092c994722a896c252212e2b3614ab 92463Author: Simon Thum <simon.thum@gmx.de> 92464Date: Thu Mar 10 21:03:40 2011 +0100 92465 92466 xserver: remove AbsoluteClass, breaking the A(P|B)I 92467 92468 This struct was unused and has been effectively removed in 92469 commit 633b81e8ba09cc6a1ea8b43f323874fda2cf0bde 92470 Refs: xorg-server-1.10.0-133-g633b81e 92471 92472 Remove the remainder, with an ABI bump to 13.0. 92473 92474 Signed-off-by: Simon Thum <simon.thum@gmx.de> 92475 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 92476 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92477 92478commit c05c8640f13708384e77579e7714ca846fa93f5c 92479Author: Oleh Nykyforchyn <oleh.nyk@gmail.com> 92480Date: Thu May 19 09:39:52 2011 +0300 92481 92482 xfree86: Allow "MatchLayout" statements in config files 92483 92484 Usage example (tested on a dual-seat PC): 92485 Section "InputClass" 92486 Identifier "keyboard-all" 92487 MatchIsKeyboard "on" 92488 MatchDevicePath "/dev/input/event*" 92489 MatchLayout "!GeForce|!Matrox" 92490 Driver "evdev" 92491 Option "XkbLayout" "us" 92492 Option "XkbOptions" "terminate:ctrl_alt_bksp" 92493 EndSection 92494 92495 It disables auto keyboard configuration for layouts "GeForce" and "Matrox". 92496 Note that "" in patterns means "no Layout sections found", e.g. 92497 MatchLayout "GeForce|" 92498 is "in layout GeForce or without explicit layout at all". 92499 92500 Signed-off-by: Oleh Nykyforchyn <oleh.nyk@gmail.com> 92501 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 92502 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 92503 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92504 92505commit 83c059f03463f2d7f41e172afe510d1ca9bba8b0 92506Author: Tomas Frydrych <tomas@sleepfive.com> 92507Date: Wed May 18 20:49:52 2011 +0100 92508 92509 Xephyr: fix pointer coordinate translation when screen is rotated 92510 92511 In the Xephyr case the position of the pointer relative toward the 92512 Xephyr window is controlled by the host server without taking into 92513 account rotation of the Xephyr screen. Consequently the pointer coords 92514 must always be translated when the fb is rotated. 92515 92516 Signed-off-by: Tomas Frydrych <tomas@sleepfive.com> 92517 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92518 92519commit b387069fafbe549f2091f364e2aa92af6cc21261 92520Author: Tomas Frydrych <tomas@sleepfive.com> 92521Date: Wed May 18 20:47:52 2011 +0100 92522 92523 Xephyr: added dummy ephyrDeviceCursorCleanup() to avoid crashing 92524 92525 The DeviceCursorCleanup in miPointerSpriteFuncRec can no longer be 92526 NULL it seems. 92527 92528 Signed-off-by: Tomas Frydrych <tomas@sleepfive.com> 92529 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 92530 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92531 92532commit 53ccc54ac71c2655276678e13f0b3d879d691489 92533Author: Colin Harrison <colin.harrison@virgin.net> 92534Date: Mon May 16 19:31:57 2011 +0100 92535 92536 Fix XWin compilation after updates for input API changes 92537 92538 Fix XWin compilation after updates for input API changes in commits 92539 e7150db5 8670c46b and 20fb07f4 92540 92541 Also remove a no longer needed InternalEvent* variable 92542 92543 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 92544 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92545 92546commit 4c4df3ac0d8162c874b85fede12ffbe42f0062c3 92547Author: Jeremy Huddleston <jeremyhu@apple.com> 92548Date: Sat May 14 16:23:59 2011 -0700 92549 92550 input: Don't implicitly define verify_internal_event 92551 92552 Fixes regression introduced by 56901998020b6f443cbaa5eb303100d979e81b22 92553 92554 mieq.c:159:5: error: implicit declaration of function 'verify_internal_event' is invalid in C99 [-Wimplicit-function-declaration,Semantic Issue] 92555 verify_internal_event(e); 92556 ^ 92557 1 error generated. 92558 92559 Also includes some other warning cleanups in events.c we're there. 92560 92561 events.c:2198:24: warning: equality comparison with extraneous parentheses [-Wparentheses,Semantic Issue] 92562 else if ((type == MotionNotify)) 92563 ~~~~~^~~~~~~~~~~~~~~ 92564 events.c:2198:24: note: remove extraneous parentheses around the comparison to silence this warning [Semantic Issue] 92565 else if ((type == MotionNotify)) 92566 ~ ^ ~ 92567 events.c:2198:24: note: use '=' to turn this equality comparison into an assignment [Semantic Issue] 92568 else if ((type == MotionNotify)) 92569 ^~ 92570 = 92571 events.c:2487:5: error: implicit declaration of function 'verify_internal_event' is invalid in C99 [-Wimplicit-function-declaration,Semantic Issue] 92572 verify_internal_event(event); 92573 ^ 92574 events.c:5909:22: warning: declaration shadows a local variable [-Wshadow,Semantic Issue] 92575 DeviceIntPtr it = inputInfo.devices; 92576 ^ 92577 events.c:5893:18: note: previous declaration is here 92578 DeviceIntPtr it = inputInfo.devices; 92579 ^ 92580 3 warnings and 1 error generated. 92581 92582 events.c:2836:27: warning: incompatible pointer types passing 'DeviceEvent *' (aka 'struct _DeviceEvent *') to parameter of type 92583 'const InternalEvent *' (aka 'const union _InternalEvent *') 92584 verify_internal_event(ev); 92585 ^~ 92586 ../include/inpututils.h:40:56: note: passing argument to parameter 'ev' here 92587 extern void verify_internal_event(const InternalEvent *ev); 92588 ^ 92589 1 warning generated. 92590 92591 Found-by: yuffie tinderbox (-Werror=implicit) 92592 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92593 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 92594 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92595 92596commit ba7d1020ac0135ebc8acda4cd57ed48b331e0133 92597Author: Jeremy Huddleston <jeremyhu@apple.com> 92598Date: Sat May 14 16:31:10 2011 -0700 92599 92600 input: Fix format string for verify_internal_event 92601 92602 inpututils.c:577:25: warning: conversion specifies type 'unsigned short' but the argument has type 'unsigned char' [-Wformat,Format String Issue] 92603 ErrorF("%02hx ", *data); 92604 ~~~~^ ~~~~~ 92605 %02hhx 92606 1 warning generated. 92607 92608 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92609 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 92610 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92611 92612commit 0de7cec90738a7a5020150309866bb0e23b6f479 92613Author: Peter Hutterer <peter.hutterer@who-t.net> 92614Date: Thu May 19 15:36:29 2011 +1000 92615 92616 xfree86: bump to video ABI 11 92617 92618 We've broken the ABI with some commit and drivers built against ABI 10 92619 happily segfault now. 92620 92621 (The relevant patch is 51f353d0a0d116af16d7d9590cadef6c56328746 which 92622 changed the ATOM typedef from unsigned long to uint32_t, thanks to 92623 Cyril Brulebois <kibi@debian.org> for figuring this out) 92624 92625 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92626 Signed-off-by: Keith Packard <keithp@keithp.com> 92627 92628commit 1fb501ad1521cfedaa5cf3052d45a924ef1866cf 92629Author: Jeremy Huddleston <jeremyhu@apple.com> 92630Date: Sun May 22 10:18:36 2011 -0700 92631 92632 XQuartz: Don't crash if CG increases our display resolution 92633 92634 miPaintWindow would cause fbFill() to overwrite pScreen's pixmap which was 92635 sized for the old resolution. 92636 92637 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92638 92639commit 60af79e35ee8546a99d15a1358aac3deabfa22be 92640Author: Jeremy Huddleston <jeremyhu@apple.com> 92641Date: Sun May 22 09:32:57 2011 -0700 92642 92643 XQuartz: RandR: Don't crash if X11 is launched while there are no attached displays 92644 92645 If CG reports no displays when launching, we could crash in RandR. Instead, just 92646 provide a fake 800x600 display until we are notified about displays being attached. 92647 92648 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92649 92650commit f25ca898c54cb88c7886005fc75a53762c42710b 92651Author: Jeremy Huddleston <jeremyhu@apple.com> 92652Date: Thu May 19 14:39:22 2011 -0700 92653 92654 XQuartz: Mark functions _X_NORETURN 92655 92656 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92657 92658commit 7413886d650aef492ecbfdc4298c2d92d9af5f87 92659Author: Jeremy Huddleston <jeremyhu@apple.com> 92660Date: Thu May 19 14:35:56 2011 -0700 92661 92662 XQuartz: Silence clang warnings about shadow declarations 92663 92664 X11Application.m:1272:26: warning: declaration shadows a local variable [-Wshadow,Semantic Issue] 92665 xp_error e; 92666 ^ 92667 X11Application.m:1098:36: note: previous declaration is here 92668 - (void) sendX11NSEvent:(NSEvent *)e { 92669 ^ 92670 1 warning generated. 92671 92672 bundle-main.c:648:36: warning: declaration shadows a local variable [-Wshadow,Semantic Issue] 92673 int max_files, i; 92674 ^ 92675 bundle-main.c:594:9: note: previous declaration is here 92676 int i; 92677 ^ 92678 1 warning generated. 92679 92680 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92681 92682commit 96ac4e61f4618332d95d1fd0e4799dd82844f90f 92683Author: Jeremy Huddleston <jeremyhu@apple.com> 92684Date: Thu May 19 14:34:39 2011 -0700 92685 92686 XQuartz: Update DEBUG_LOG to report to ASL 92687 92688 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92689 92690commit 384eb45b944a4386eae74a5503423c13b5f2a659 92691Author: Jeremy Huddleston <jeremyhu@apple.com> 92692Date: Sat May 14 18:33:57 2011 -0700 92693 92694 XQuartz: RandR: Avoid over-releasing if we are unable to determine the current display mode. 92695 92696 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92697 92698commit 25191648b8db87735a99243697f73036255c1eb6 92699Author: Jeremy Huddleston <jeremyhu@apple.com> 92700Date: Sat May 14 16:11:32 2011 -0700 92701 92702 XQuartz: Don't call mieqEnqueue during server shutdown 92703 92704 Found-by: GuardMalloc 92705 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92706 92707commit fd5f630bc41de73be3b1d26bf9ce3e9ef1badc3b 92708Author: Jeremy Huddleston <jeremyhu@apple.com> 92709Date: Sat May 14 14:09:45 2011 -0700 92710 92711 XQuartz: Fix an array-index-out-of-bounds crasher 92712 92713 Found-by: GuardMalloc 92714 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92715 92716commit 0e7f61d72c4a929319e57c9b5b777e9413c23051 92717Author: Gaetan Nadon <memsize@videotron.ca> 92718Date: Sat May 14 13:30:20 2011 -0400 92719 92720 doc: use devbook.am for developers documentation 92721 92722 Relocate the docs under the doc subdir. 92723 Remove redundant xml subdir. 92724 The xmlrules set of makefiles are no longer used. 92725 92726 Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> 92727 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 92728 Signed-off-by: Keith Packard <keithp@keithp.com> 92729 92730commit 96c65be39b98cfe035be0a3988dce519549249f3 92731Author: Gaetan Nadon <memsize@videotron.ca> 92732Date: Sat May 14 13:30:19 2011 -0400 92733 92734 doc: relocate xserver.ent in the package root directory 92735 92736 This is the appropriate location for reusable configuration and/or 92737 makefile artifacts as opposed to picking one of the subdirs where 92738 it is used. 92739 92740 It shields them from future doc reorg as every subdir 92741 will refer to the root package which never changes location. 92742 92743 Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> 92744 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 92745 Signed-off-by: Keith Packard <keithp@keithp.com> 92746 92747commit dbe26634e4f12eba905540666783c9879aa3a49a 92748Author: Gaetan Nadon <memsize@videotron.ca> 92749Date: Sat May 14 13:30:18 2011 -0400 92750 92751 doc: add external doc references support to Xserver-DTrace 92752 92753 This makefile is used to build the 60+ docbooks in xorg. 92754 Dtrace is a user document and should be located under doc subdir. 92755 92756 This user document can now refer to external user/specs docs 92757 or can be referred to by such documnets. 92758 92759 Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> 92760 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 92761 Signed-off-by: Keith Packard <keithp@keithp.com> 92762 92763commit 221507e3bf6fcaa4a4c2e1bc264a1f806b4362c0 92764Author: Gaetan Nadon <memsize@videotron.ca> 92765Date: Thu Apr 28 21:16:13 2011 -0400 92766 92767 man: relocate manual pages in the man subdir outside doc 92768 92769 The convention is to have the manual pages in a man subdir 92770 which is not under a doc dir. The doc dir contains users docs. 92771 This will move man pages out of the way for upcoming DocBook patches. 92772 92773 Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> 92774 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 92775 Signed-off-by: Keith Packard <keithp@keithp.com> 92776 92777commit 9129beb507642e2414ef1f90d650572325d8c2dc 92778Author: Gaetan Nadon <memsize@videotron.ca> 92779Date: Thu Apr 28 21:16:12 2011 -0400 92780 92781 dmx: fix warning for doxygen explicit links 92782 92783 Explicit links to functions in another file are not supported. 92784 92785 Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> 92786 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 92787 Signed-off-by: Keith Packard <keithp@keithp.com> 92788 92789commit 0d8a5766a39320dedd34a215fc6655f9ac19e8bc 92790Author: Gaetan Nadon <memsize@videotron.ca> 92791Date: Thu Apr 28 21:16:11 2011 -0400 92792 92793 dmx: modernize doxygen generation. 92794 92795 The configuration and stylesheet were very old. 92796 The stylesheet is not checked-in, use the generated one. 92797 The header is not checked-in, use the generated one. 92798 Add datetime and projectname in default footer. 92799 Developer documentation is not installed and not included in tarball. 92800 92801 Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> 92802 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 92803 Signed-off-by: Keith Packard <keithp@keithp.com> 92804 92805commit 622ebb88157b0ed99355419767685dd70ce0f16c 92806Author: Gaetan Nadon <memsize@videotron.ca> 92807Date: Thu Apr 28 21:16:10 2011 -0400 92808 92809 dmx: split DocBook/XML and Doxygen makefile targets 92810 92811 Maintaining either requires full knowledge of both. 92812 It's not obvious one has to check the usage of global variables 92813 in devbook.am when maintaining doxygen target. Or vice-versa. 92814 92815 Being in their respective directory, one less thing to worry about. 92816 92817 Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> 92818 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 92819 Signed-off-by: Keith Packard <keithp@keithp.com> 92820 92821commit 595460c397a543fe512becd3c432de2a0a68d187 92822Author: Gaetan Nadon <memsize@videotron.ca> 92823Date: Thu Apr 28 21:16:09 2011 -0400 92824 92825 dmx/doc: use common makefile for developers documentation 92826 92827 The user/specs docs now have external references support. 92828 Developers doc are not installed so they do not participate. 92829 However, using a similar makefile shared amongst developers 92830 document reduces maintenance and is forward looking. 92831 92832 Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> 92833 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 92834 Signed-off-by: Keith Packard <keithp@keithp.com> 92835 92836commit 04011b0bc3b5b59efa0d981866349c4ad0868d8f 92837Author: Gaetan Nadon <memsize@videotron.ca> 92838Date: Thu Apr 28 21:16:08 2011 -0400 92839 92840 doc: use common makefile for developers documentation 92841 92842 The user/specs docs now have external references support. 92843 Developers doc are not installed so they do not participate. 92844 However, using a similar makefile shared amongst developers 92845 document reduces maintenance and is forward looking. 92846 92847 Man pages being out of here, reorg developers docs under the same roof. 92848 Drop the obsolete sgml subdir. 92849 92850 Reviewed-by Jeremy Huddleston <jeremyhu@apple.com> 92851 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 92852 Signed-off-by: Keith Packard <keithp@keithp.com> 92853 92854commit 531869448d07e00ae241120b59f3aaaa5709d59c 92855Author: Ville Syrjälä <ville.syrjala@nokia.com> 92856Date: Fri May 6 18:18:15 2011 +0300 92857 92858 dri2: Don't send so many needless invalidate events 92859 92860 Only send invalidate events for drawables if some client has requested 92861 some buffers. 92862 92863 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 92864 Reviewed-by: Michel Dänzer <michel@daenzer.net> 92865 Signed-off-by: Keith Packard <keithp@keithp.com> 92866 92867commit ba5540221f2a46133371b4ff0d527b1a0a1443b1 92868Merge: 6347a0b80 728d0bf20 92869Author: Keith Packard <keithp@keithp.com> 92870Date: Fri May 13 13:59:36 2011 -0700 92871 92872 Merge remote-tracking branch 'whot/for-keith' 92873 92874commit 6347a0b802812bb185ada1bf0951add306935184 92875Merge: 4d02c5397 f144fb771 92876Author: Keith Packard <keithp@keithp.com> 92877Date: Fri May 13 13:54:29 2011 -0700 92878 92879 Merge remote-tracking branch 'jeremyhu/master' 92880 92881commit 4d02c5397114ac4d15e794908f0708427e258261 92882Merge: 043c17586 eac37f32b 92883Author: Keith Packard <keithp@keithp.com> 92884Date: Fri May 13 13:52:18 2011 -0700 92885 92886 Merge remote-tracking branch 'vsyrjala/composite_validatetree_2' 92887 92888commit f144fb771f2e0665677aa88d08778899f666a1d5 92889Author: Jeremy Huddleston <jeremyhu@apple.com> 92890Date: Fri May 13 09:10:42 2011 -0700 92891 92892 XQuartz: Don't circumvent NDEBUG 92893 92894 If someone wants to turn off asserts with NDEBUG, let them. 92895 92896 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92897 92898commit f46835a09662f71f2cb90540609c34561d8af87f 92899Author: Jeremy Huddleston <jeremyhu@apple.com> 92900Date: Wed May 11 15:23:17 2011 -0700 92901 92902 XQuartz: Redirect stdout/stderr to asl 92903 92904 In order to improve logging in XQuartz, stdout and stderr should be redirected 92905 to asl (syslog). 92906 92907 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92908 92909commit 5bc05d96f883add324bab338ffb9b9409cf45b2c 92910Author: Jeremy Huddleston <jeremyhu@apple.com> 92911Date: Thu May 12 15:45:29 2011 -0700 92912 92913 XQuartz: Add a LOGGING section to our man page 92914 92915 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92916 92917commit c18b7165f971cbf6b634ccf670eb4aa9840eb9e4 92918Author: Jeremy Huddleston <jeremyhu@apple.com> 92919Date: Thu May 12 18:29:23 2011 -0700 92920 92921 XQuartz: stub: Log directly to ASL rather than stdout/stderr 92922 92923 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92924 92925commit d694601591d2a86a437c5150015d9889e2fe16fc 92926Author: Jeremy Huddleston <jeremyhu@apple.com> 92927Date: Thu May 12 15:15:40 2011 -0700 92928 92929 Fix a typo: laucnd instead of launchd 92930 92931 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 92932 92933commit 728d0bf20ed8e2612b100fca6526705fa6e1eef4 92934Author: Peter Hutterer <peter.hutterer@who-t.net> 92935Date: Wed May 11 13:35:39 2011 +1000 92936 92937 dix: replace CORE_EVENT and XI2_EVENT macros with inline functions. 92938 92939 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92940 Reviewed-by: Jamey Sharp <jamey@minilop.net> 92941 92942commit ffd4874798ba54f86acac75779a15b4babeaa5f3 92943Author: Peter Hutterer <peter.hutterer@who-t.net> 92944Date: Wed May 11 12:20:50 2011 +1000 92945 92946 include: add version_compare helper function 92947 92948 Compare two version numbers in the major.minor form. 92949 Switch the few users of manual version switching over to the new function. 92950 92951 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92952 Reviewed-by: Jamey Sharp <jamey@minilop.net> 92953 92954commit c4f9c3a07dbb05b81c8e2193a083102f710ebb27 92955Author: Peter Hutterer <peter.hutterer@who-t.net> 92956Date: Wed May 11 11:43:16 2011 +1000 92957 92958 dix: use a tmp variable instead of multiple rClient(other). 92959 92960 no functional changes. 92961 92962 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92963 Reviewed-by: Jamey Sharp <jamey@minilop.net> 92964 92965commit dc45d5816dd65168645f0017394eebfc5599d698 92966Author: Peter Hutterer <peter.hutterer@who-t.net> 92967Date: Thu Apr 14 15:59:08 2011 +1000 92968 92969 Xi: split DeviceStateNotify delivery into a separate function 92970 92971 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92972 Reviewed-by: Jamey Sharp <jamey@minilop.net> 92973 92974commit 5bcc22757e6e1f24ee2bfec65f68a5f567300532 92975Author: Peter Hutterer <peter.hutterer@who-t.net> 92976Date: Thu Apr 14 15:43:56 2011 +1000 92977 92978 dix: return deliveries from DeliverGrabbedEvent 92979 92980 This isn't currently used by any of the callers but it will likely be in the 92981 future. 92982 92983 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92984 Reviewed-by: Jamey Sharp <jamey@minilop.net> 92985 92986commit 2054ca73060a20b5a3025e8d5ef68182149484d3 92987Author: Peter Hutterer <peter.hutterer@who-t.net> 92988Date: Thu Apr 14 15:09:39 2011 +1000 92989 92990 dix: move the grab activation condition into a if block. 92991 92992 Rather than 3 conditions with if (deliveries && ...), have one block with 92993 the three in them. 92994 No functional changes. 92995 92996 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 92997 Reviewed-by: Jamey Sharp <jamey@minilop.net> 92998 92999commit 236ed6f50675dc0303a505ac6f0418c515438fe1 93000Author: Peter Hutterer <peter.hutterer@who-t.net> 93001Date: Tue Apr 12 14:04:37 2011 +1000 93002 93003 dix: split out client delivery from DeliverEventsToWindow 93004 93005 No real functional changes, this is just for improved readability. 93006 93007 DeliverEventsToWindow used to return an int to specify the number of 93008 deliveries (or rejected deliveries if negative). The number wasn't used by 93009 any caller other than for > 0 comparison. 93010 93011 This patch also changes the return value to be -1 or 1 even in case of 93012 multiple deliveries/rejections. The comment was updated accordingly. 93013 93014 A future patch should probably use the enum EventDeliveryState for 93015 DeliverEventsToWindow. 93016 93017 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93018 Reviewed-by: Jamey Sharp <jamey@minilop.net> 93019 93020commit 536ca28f1b0b4d8715a41b8acc5f30364c833f9b 93021Author: Peter Hutterer <peter.hutterer@who-t.net> 93022Date: Tue Apr 12 13:44:30 2011 +1000 93023 93024 dix: split out window owner event delivery from DeliverEventsToWindow 93025 93026 No functional changes, just for readability. 93027 93028 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93029 Reviewed-by: Jamey Sharp <jamey@minilop.net> 93030 93031commit a311a03135f9734aa4f7dc0f2fa0e4c91768343b 93032Author: Jamey Sharp <jamey@minilop.net> 93033Date: Wed May 11 14:08:28 2011 -0700 93034 93035 dix: split implicit grab activation into a separate function. 93036 93037 I'm not sure I like splitting the check for button-press event from the 93038 code which makes assumptions about that check. How about replacing 93039 patches 3 and 4 with this patch instead? 93040 93041 Signed-off-by: Jamey Sharp <jamey@minilop.net> 93042 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93043 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93044 93045commit a558a66d249cd51d3d675b20cc3a76651fd87930 93046Author: Jeremy Huddleston <jeremyhu@apple.com> 93047Date: Wed May 11 17:27:52 2011 -0700 93048 93049 configure.ac: XQuartz: Fix support for the deprecated --with-launchd-id-prefix 93050 93051 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93052 93053commit c89a6f824eaf647d2b182f79fbd78f5bd1c3a27f 93054Author: Jeremy Huddleston <jeremyhu@apple.com> 93055Date: Tue May 10 23:42:46 2011 -0700 93056 93057 XQuartz: Don't call into CoreFoundation after fork() and before exec() 93058 93059 After fork()ing, we should just limit ourselves to setting up 93060 the environment, file descriptors, and exec()ing. 93061 93062 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93063 93064commit c8674a328c68f03de6e4fad7790a595cdfc18736 93065Author: Peter Hutterer <peter.hutterer@who-t.net> 93066Date: Thu Apr 14 16:45:56 2011 +1000 93067 93068 dix: replace unneded goto with break. 93069 93070 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93071 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 93072 93073commit 20fb07f436f7d4a0f330b2067a93a5a4829fccf5 93074Author: Peter Hutterer <peter.hutterer@who-t.net> 93075Date: Fri Apr 15 10:07:10 2011 +1000 93076 93077 input: remove DDX event list handling 93078 93079 The current approach to event posting required the DDX to request the event 93080 list (allocated by the DIX) and then pass that list into QueuePointerEvent 93081 and friends. 93082 93083 Remove this step and use the DIX event list directly. This means that 93084 QueuePointerEvent is not reentrant but it wasn't before anyway. 93085 93086 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93087 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 93088 93089commit 8670c46bdfdade64e63119d2ebbd5ef63b6fa2c3 93090Author: Peter Hutterer <peter.hutterer@who-t.net> 93091Date: Thu Apr 14 22:05:41 2011 +1000 93092 93093 input: replace EventListPtr with InternalEvent array 93094 93095 EventListPtr is a relic from pre-1.6, when we had protocol events in the 93096 event queue and thus events of varying size. 93097 93098 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93099 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 93100 93101commit e7150db5350bc2113ff4126019b489847a4dc217 93102Author: Peter Hutterer <peter.hutterer@who-t.net> 93103Date: Mon Apr 11 15:48:15 2011 +1000 93104 93105 input: Provide Queue{Button|Keyboard|Proximity}Event helpers 93106 93107 Don't require every caller to use GPE + mieqEnqueue, provide matching 93108 Queue...Event functions instead. 93109 93110 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93111 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 93112 93113commit 1b8593a6c12315b1071a4fa586151e12f46458f5 93114Author: Peter Hutterer <peter.hutterer@who-t.net> 93115Date: Mon May 9 15:13:17 2011 +1000 93116 93117 xfree86: print the device ID to the log when adding a device. 93118 93119 Sometimes the name isn't enough, it's handy to see the device ID's from the 93120 log file. 93121 93122 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93123 93124commit 3231962db826f5efd431596a309c96e907a191d1 93125Author: Daniel Stone <daniel@fooishbar.org> 93126Date: Tue May 3 03:20:23 2011 +0100 93127 93128 XKB: Fix sense inversion for core MapNotify events 93129 93130 Due to an unfortunate sense inversion incident while switching from a 93131 if (foo) { ... } to if (!foo) continue; style in f06a9d, we punished any 93132 client who attempted to use XKB to restrict the MapNotify events they 93133 wanted by sending them exactly the events they _didn't_ want, and 93134 nothing else. 93135 93136 NewKeyboardNotifies (coming from a client setting the map with an XKB 93137 request, when switching between master devices, etc) weren't affected, 93138 but this would impact anyone using xmodmap-style core requests. Could 93139 explain a fair bit. 93140 93141 Clarified the comments while I was at it. 93142 93143 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93144 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93145 93146commit 460a377ef2d645d9ae38a8356fb248ccc47bed4a 93147Author: Daniel Stone <daniel@fooishbar.org> 93148Date: Tue May 3 03:07:50 2011 +0100 93149 93150 XKB: Send XKB events for all devices to all clients 93151 93152 We were using XIShouldNotify(client, device) as a test for whether or 93153 not to send XKB map/state/etc changed events, which limits it to only 93154 sending events for the current ClientPointer/ClientKeyboard for that 93155 client. While this makes perfect sense for core events (e.g. 93156 MappingNotify), XKB events carry a device ID, so are safe to send to all 93157 clients for all devices. 93158 93159 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93160 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93161 93162commit a79d4544fee806a25447d0147535ebc5a1cae6b9 93163Author: Daniel Stone <daniel@fooishbar.org> 93164Date: Tue May 3 03:03:06 2011 +0100 93165 93166 XKB: Send NewKeyboardNotify for dev before its master/slaves 93167 93168 When we change the keymap on a device, send the NewKeyboardNotify for 93169 that device before we copy the keymap to and notify for its attached 93170 master/slave devices. 93171 93172 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93173 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93174 93175commit c7634498d4cd42c8571805122224dc2d0e44a585 93176Author: Daniel Stone <daniel@fooishbar.org> 93177Date: Tue May 3 02:59:53 2011 +0100 93178 93179 XKB: Remove duplicate keymap-copying loop 93180 93181 Previously we had: 93182 foreach (device + slaves of device) { 93183 XkbCopyDeviceKeymap(i, device); 93184 [...] 93185 } 93186 if (device was last slave of its MD) { 93187 XkbCopyDeviceKeymap(master, device); 93188 } 93189 and now: 93190 foreach (device + slaves of device + MD if device was last slave) { 93191 XkbCopyDeviceKeymap(i, device); 93192 [...] 93193 } 93194 93195 As an extra bonus, when changing the keymap on a slave device, we now 93196 ensure the LED info on the master is kept in sync. 93197 93198 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93199 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93200 93201commit b8540d18c7d3a0f93f9e2565a410986eddedcecb 93202Author: Daniel Stone <daniel@fooishbar.org> 93203Date: Tue May 3 02:50:48 2011 +0100 93204 93205 XKB: Simplify a loop in ProcXkbGetKbdByName 93206 93207 Replace: 93208 for (stuff; things; etc) { 93209 if (misc || other) { 93210 [...] 93211 } 93212 } 93213 with: 93214 for (stuff; things; etc) { 93215 if (!misc && !other) 93216 continue; 93217 [...] 93218 } 93219 93220 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93221 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93222 93223commit 72b6639c83df74767094a5e0c2861fdc5ea03ecd 93224Author: Daniel Stone <daniel@fooishbar.org> 93225Date: Tue May 3 02:39:55 2011 +0100 93226 93227 XKB: Don't send unnecessary NewKeyboardNotifies 93228 93229 In the XKB GetKeyboardByName handler, we had the following pseudocode: 93230 if (device was last slave of its MD) { 93231 XkbCopyDeviceKeymap(master, slave); 93232 XkbSendNewKeyboardNotify(slave, ¬ify); 93233 } 93234 93235 Even if the SendNewKeyboardNotify line nominated the correct device, 93236 which it didn't, it's unnecessary as XkbCopyDeviceKeymap already sends a 93237 NewKeyboardNotify on the destination device. 93238 93239 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93240 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93241 93242commit 70cef8d8baf058bacaff87ef49e3851628269597 93243Author: Daniel Stone <daniel@fooishbar.org> 93244Date: Wed Jan 19 20:42:10 2011 +0000 93245 93246 Input: Simplify CheckPassiveGrabsOnWindow loop 93247 93248 Instead of a mega never-ending if branch with no else, just continue 93249 to the next iteration of the loop if the conditions aren't met - pretty 93250 much entirely reindentation. 93251 93252 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93253 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93254 93255commit 00ba884556c675b2b25e116f5ab4eb4590b6dd56 93256Author: Daniel Stone <daniel@fooishbar.org> 93257Date: Wed Jan 19 20:38:44 2011 +0000 93258 93259 Input: Make CheckPassiveGrabsOnWindow take InternalEvent 93260 93261 Previously, it only took DeviceEvents, but it would be much more useful 93262 if it took InternalEvents. Any event that activates a grab must still 93263 be a DeviceEvent, so put in a check to enforce this. 93264 93265 Change all callers to make the appropriate casts. 93266 93267 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93268 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93269 93270commit 56901998020b6f443cbaa5eb303100d979e81b22 93271Author: Peter Hutterer <peter.hutterer@who-t.net> 93272Date: Thu May 5 08:48:19 2011 +1000 93273 93274 input: change CHECKEVENT macro to verify_internal_event function 93275 93276 The macro is sufficient if called during a development cycle, but not 93277 sufficient information when triggered by a user (e.g. 93278 https://bugzilla.redhat.com/show_bug.cgi?id=688693). 93279 93280 Expand what this does to print the event content and a backtrace, so at 93281 least we know where we're coming from. Only the first 32 bytes are printed 93282 since if something goes wrong, the event we have is almost certainly an 93283 xEvent or xError, both restricted to 32 bytes. 93284 93285 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93286 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 93287 93288commit b47d2e43eb2cb3817c995d1f7e58500fb40efa2b 93289Author: Jeremy Huddleston <jeremyhu@apple.com> 93290Date: Thu May 5 09:04:41 2011 -0700 93291 93292 XQuartz: Make a copy of args for our crash reporter vsnprintf 93293 93294 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93295 93296commit bf2059b07a97e5e579c13c2c9d49707093427dc2 93297Author: Peter Hutterer <peter.hutterer@who-t.net> 93298Date: Fri Apr 15 14:06:20 2011 +1000 93299 93300 input: Only release SD buttons for explicit floating/reattachment (#36146) 93301 93302 Grabbing an SD device temporary floats the device but we must not release 93303 the buttons. Introduced in 93304 93305 commit 9d23459415b84606ee4f38bb2d19054c432c8552 93306 Author: Peter Hutterer <peter.hutterer@who-t.net> 93307 Date: Fri Feb 25 11:08:19 2011 +1000 93308 93309 dix: release all buttons and keys before reattaching a device (#34182) 93310 93311 X.Org Bug 36146 <http://bugs.freedesktop.org/show_bug.cgi?id=36146> 93312 93313 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 93314 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 93315 93316commit eac37f32b85b631d94ee3ba11fa65b9d2cb72c38 93317Author: Ville Syrjälä <ville.syrjala@nokia.com> 93318Date: Wed Apr 13 21:46:20 2011 +0300 93319 93320 composite: Recompute clipping when changing between manual and automatic redirection 93321 93322 Call compMarkWindows() when changing between manual and automatic 93323 redirection modes. Otherwise the window clipping won't be recomputed 93324 correctly. 93325 93326 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 93327 93328commit f2001b0f6dffa0a8f05def4a86ea37c4c91db724 93329Author: Ville Syrjälä <ville.syrjala@nokia.com> 93330Date: Wed Apr 13 21:45:43 2011 +0300 93331 93332 composite: Fix pWin->redirectDraw when changing between manual and automatic redirection 93333 93334 compAllowPixmap() is not called when changing between manual and 93335 automatic redirection modes. That means pWin->redirectDraw is left 93336 with an incorrect value, and miComputeClips() gets confused whether 93337 the window is supposed to be treated as transparent or not. Fix 93338 the issue by updating pWin->redirectDraw in compCheckRedirect() 93339 even when not calling compAllocPixmap(). 93340 93341 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 93342 93343commit 74663e61528346aeea9c11908b6980b51dcaeb68 93344Author: Ville Syrjälä <ville.syrjala@nokia.com> 93345Date: Mon Dec 20 16:37:24 2010 +0200 93346 93347 composite: Copy the window contents back from the pixmap 93348 93349 Since extra expose events are no longer generated during window 93350 unredirection, the window contents must be preserved by the server. 93351 So copy the window contents back from the pixmap. The copy can only 93352 be done after the clips have been recomputed, so delay the copy and 93353 the pixmap destruction until ValidateTree is done. Window borders are 93354 restored by HandleExposures and thus don't need to be copied back. 93355 93356 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 93357 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 93358 93359commit 193ecc8b453b22b3e60248b9354c768dbd405598 93360Author: Ville Syrjälä <ville.syrjala@nokia.com> 93361Date: Fri Dec 17 23:46:34 2010 +0200 93362 93363 composite: Get rid of the internal UnmapWindow+MapWindow cycle 93364 93365 Eliminate the internal MapWindow+UnmapWindow cycle around window 93366 redirection changes. Instead do the work in a single pass by marking 93367 the afected windows and calling ValidateTree and HandleExposures 93368 directly. This gets rid of unnecessary expose events, and invalid 93369 ClipNotify calls during rediredction changes. Now ClipNotify will only 93370 get called with the final clip values, and expose events are only sent 93371 to areas that actually got exposed. 93372 93373 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 93374 Reviewed-by: Adam Jackson <ajax@redhat.com> 93375 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 93376 93377commit a6ae91746212203a19450ac955fbb7abffff1ed3 93378Author: Ville Syrjälä <ville.syrjala@nokia.com> 93379Date: Mon Dec 20 16:30:52 2010 +0200 93380 93381 composite: Initialize borderClip with current values 93382 93383 ValidateTree needs a valid borderClip so initialize the parent 93384 constrained border clip with the window's current borderClip 93385 in compRedirectWindow. 93386 93387 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 93388 Reviewed-by: Adam Jackson <ajax@redhat.com> 93389 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 93390 93391commit ce9eff9e4c30deb16f059ed84b436d37da45d5d5 93392Author: Ville Syrjälä <ville.syrjala@nokia.com> 93393Date: Mon Dec 20 16:33:36 2010 +0200 93394 93395 composite: Call ValidateGC after ChangeGC 93396 93397 ChangeGC changes the GC, so ValidateGC should be called after it, not 93398 before. 93399 93400 Also pass NullClient instead of serverClient to ChangeGC() since we 93401 know the changed values to be valid, and setting 93402 serverClient->errorValue seems pointless anyway. 93403 93404 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 93405 Reviewed-by: Keith Packard <keithp@keithp.com> 93406 Reviewed-by: Adam Jackson <ajax@redhat.com> 93407 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 93408 93409commit 3ac220d6cc386c9ab9995e5b83463c40330e18c2 93410Author: Jeremy Huddleston <jeremyhu@apple.com> 93411Date: Tue May 3 10:54:09 2011 -0700 93412 93413 XQuartz: prefs_copy_url and prefs_get_copy return retained objects 93414 93415 No functional change. This just annotates the return policy. 93416 93417 Found-by: clang static analyzer 93418 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93419 93420commit bac8d12555dd44b2139be0696629910cd4d8c782 93421Author: Jeremy Huddleston <jeremyhu@apple.com> 93422Date: Tue May 3 10:36:19 2011 -0700 93423 93424 XQuartz: Ensure that {CF,NS}_RETURNS{,_NOT}_RETAINED are defined 93425 93426 These will be used in subsequent patches to denote proper retain counts in XQuartz 93427 93428 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93429 93430commit 043c1758652259fd12b88ae37720fe6e93eda76b 93431Author: Alan Coopersmith <alan.coopersmith@oracle.com> 93432Date: Tue Apr 19 19:02:54 2011 -0700 93433 93434 Clean up memory better when GetVisualInfo fails in ProcDbeGetVisualInfo 93435 93436 Use calloc to initialize pScrVisInfo array so we don't have to check 93437 which ones were already initialized when freeing them all. 93438 93439 On failure, set rc if necessary, and jump to code at end that already 93440 frees all the necessary allocations and return rc. 93441 93442 Fixes parfait reported error: 93443 Error: Memory leak (CWE 401) 93444 Memory leak of pointer 'pScrVisInfo' allocated with malloc((count * 16)) 93445 at line 724 of dbe/dbe.c in function 'ProcDbeGetVisualInfo'. 93446 'pScrVisInfo' allocated at line 693 with malloc((count * 16)). 93447 pScrVisInfo leaks when rc != 0 at line 710 93448 and j >= i at line 716. 93449 93450 [ This bug was found by the Parfait 0.3.7 bug checking tool. 93451 For more information see http://labs.oracle.com/projects/parfait/ ] 93452 93453 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 93454 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 93455 93456commit dadb0791ebfd05cd3bb82d4addf0fbc21aad6fbb 93457Author: Alan Coopersmith <alan.coopersmith@oracle.com> 93458Date: Tue Apr 19 18:53:22 2011 -0700 93459 93460 Only free pContext once when AddResource fails in ProcRecordCreateContext 93461 93462 Since RecordDeleteContext frees its argument, don't fall through to free 93463 it again. 93464 93465 Error: Double free (CWE 415) 93466 Double free of pointer 'malloc(1072)' defined by malloc 93467 at line 1964 of record/record.c in function 'ProcRecordCreateContext'. 93468 Previously freed at line 1960 with RecordDeleteContext. 93469 'malloc(1072)' was allocated at line 1926 with malloc. 93470 93471 [ This bug was found by the Parfait 0.3.7 bug checking tool. 93472 For more information see http://labs.oracle.com/projects/parfait/ ] 93473 93474 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 93475 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 93476 93477commit 73de54210446e9eca81b96ea6775ee5ea1a31d75 93478Author: Daniel Stone <daniel@fooishbar.org> 93479Date: Mon Jan 24 08:40:10 2011 +1000 93480 93481 Input: Fix event size confusion in CheckPassiveGrabsOnWindow 93482 93483 We were just storing a DeviceEvent, but allocating enough space for an 93484 InternalEvent. 93485 93486 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93487 Reviewed-by: Julien Cristau <jcristau@debian.org> 93488 Reviewed-by: Adam Jackson <ajax@redhat.com> 93489 Reviewed-by: Cyril Brulebois <kibi@debian.org> 93490 93491commit 706326491011be8cecb9b56c06f7241b7cbd425f 93492Author: Daniel Stone <daniel@fooishbar.org> 93493Date: Tue Jan 18 20:16:36 2011 +0000 93494 93495 Input: Add DeepestSpriteWin function 93496 93497 Does what it says on the box: returns the deepest child window in a 93498 given sprite's trace. 93499 93500 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 93501 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 93502 Reviewed-by: Keith Packard <keithp@keithp.com> 93503 93504commit 65b54548dce80c8e8ff5ff91fc4f0659e9b2d921 93505Author: Chase Douglas <chase.douglas@canonical.com> 93506Date: Tue Jan 18 20:08:09 2011 +0000 93507 93508 Input: Pass co-ordinates by reference to transformAbsolute 93509 93510 With the upcoming XI 2.1 touch work, the co-ordinate values will need to 93511 be passed by reference, rather than modified in-place. 93512 93513 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 93514 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 93515 93516commit a14a0c711397ff7ca0220946010300fc1b2a6e67 93517Author: Alan Coopersmith <alan.coopersmith@oracle.com> 93518Date: Fri Apr 22 22:19:39 2011 -0700 93519 93520 Move event filter initializer out of the structure itself 93521 93522 When kept in the structure, it causes the entire MAXDEVICES * 128 masks 93523 to be stored in the data segment and loaded from the file, and also leads 93524 to worries about later generations inheriting changes across server reset. 93525 93526 text data bss dec hex filename 93527 Before: 91837 20528 32 112397 1b70d .libs/events.o 93528 After: 92277 48 20512 112837 1b8c5 .libs/events.o 93529 Before: 3013384 122696 163156 3299236 3257a4 Xorg 93530 After: 3013832 102216 183636 3299684 325964 Xorg 93531 93532 File size before: 4337008 Xorg 93533 File size after: 4316568 Xorg 93534 93535 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 93536 Reviewed-by: Jamey Sharp <jamey@minilop.net> 93537 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 93538 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 93539 93540commit 8d229c4cf9e5bde78373ef3dd32708817ac97152 93541Author: Alan Coopersmith <alan.coopersmith@oracle.com> 93542Date: Fri Mar 11 15:43:14 2011 -0800 93543 93544 Make xorg.conf.example rule compatible with Solaris make 93545 93546 Solaris make won't substitute $< in explicit rules, only implicit ones 93547 93548 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 93549 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 93550 93551commit c6029246ddcf5ead3884dbf2ab4f57e0c3586e93 93552Author: Alan Coopersmith <alan.coopersmith@oracle.com> 93553Date: Wed Feb 2 17:00:25 2011 -0800 93554 93555 Move Xinput server API documentation from libXi to doc/xml 93556 93557 Documentation is quite out of date but still fits better here than 93558 in the client library. 93559 93560 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 93561 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 93562 93563commit 0fc7ec6dd504aa03e1a1b18c60942c0c8c8b701b 93564Author: Jeremy Huddleston <jeremyhu@apple.com> 93565Date: Fri Apr 29 11:06:18 2011 -0700 93566 93567 XQuartz: Fix incorrect typedefs with XPLUGIN_VERSION < 4 93568 93569 Ok, this time for sure... how many brown bags can I fit over my face? 93570 93571 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93572 93573commit 29d471663e4414a3cdf154fd032d74381a921ae0 93574Author: Jeremy Huddleston <jeremyhu@apple.com> 93575Date: Sat Apr 23 18:09:11 2011 -0700 93576 93577 XQuartz: Use a rwlock instead of a mutex to protect window_hash in the pthread case 93578 93579 Concurrent reads are acceptable, so using an rwlock should be better. 93580 93581 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93582 93583commit 5cb31cd0cbf83fff5f17a475e7b0e45246b19bf3 93584Merge: 0f284f0f4 4d8735d38 93585Author: Keith Packard <keithp@keithp.com> 93586Date: Fri Apr 29 09:59:49 2011 -0700 93587 93588 Merge remote-tracking branch 'jturney/remove-opengl-spec-download' 93589 93590commit 4d8735d388c3c6af27ef3d6a645e38e556f392f4 93591Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93592Date: Wed Apr 6 16:26:45 2011 +0100 93593 93594 hw/xwin: wglext.h should be provided by w32api, rather than downloaded 93595 93596 wglext.h should be provided by the w32api package, rather than downloaded. 93597 if it's not, do 'wget -P /usr/include/w32api/GL http://www.opengl.org/registry/api/wglext.h' 93598 93599 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93600 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93601 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 93602 93603commit a50878682cda541766c830e49ab406c78b38d903 93604Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93605Date: Wed Apr 6 16:23:45 2011 +0100 93606 93607 configure: Look for Khronos OpenGL spec files using pkg-config 93608 93609 Look for Khronos OpenGL spec files using pkg-config, rather than downloading them 93610 93611 Also add a --with-khronos-spec-dir=PATH configure option so XWin can be directed 93612 where to find these files without using the khronos-spec-files package 93613 93614 XWin with AIGLX requires OpenGL spec files in order to generate wrapper code which: 93615 (1) thunks from the glapi dispatch table which uses the default cdecl calling convention 93616 to native GL functions using the stdcall calling convention. 93617 (2) performs function address lookup for OpenGL 1.2+ functions, which are treated 93618 as extensions and so not directly linkable. 93619 93620 v2: KHRONOS_SPEC_DIR is only valid when XWIN_GLX_WINDOWS is defined. Avoid 'make dist' 93621 seeing invalid dependencies by only including rules using KHRONOS_SPEC_DIR if 93622 XWIN_GLX_WINDOWS is defined 93623 93624 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93625 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93626 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 93627 93628commit 0f284f0f42f26a242eb3153787f2e98b6b88dff8 93629Author: Jeremy Huddleston <jeremyhu@apple.com> 93630Date: Wed Apr 27 22:12:59 2011 -0700 93631 93632 XQuartz: BuildFix to build correctly with XPLUGIN_VERSION < 4 93633 93634 This fixes a regression introduced by d79cc14a51f3e8d4d2f66aad055092672cab1526 93635 93636 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93637 93638commit f6d4e75ec55ac6812f9dead42ecdffb9614578c7 93639Merge: ec9ea4017 4318e6a14 93640Author: Keith Packard <keithp@keithp.com> 93641Date: Wed Apr 27 12:08:51 2011 -0700 93642 93643 Merge remote-tracking branch 'jturney/master' 93644 93645commit ec9ea4017885cefe5519a4e890b0ff1a5518235a 93646Merge: c6cb70be1 6f29dbf5e 93647Author: Keith Packard <keithp@keithp.com> 93648Date: Wed Apr 27 12:01:56 2011 -0700 93649 93650 Merge remote-tracking branch 'jeremyhu/master' 93651 93652commit 4318e6a147e78b2663c5e0ea6ba0d351a1e87f98 93653Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93654Date: Sat Mar 5 17:34:42 2011 +0000 93655 93656 Cygwin/X: Handle failure during winScreenInit() 93657 93658 Handle failure during winScreenInit() a bit more cleanly, rather than crashing 93659 93660 This avoids a crash with 'XWin -fullscreen -screen 0 @2 -screen 1 @1' 93661 93662 Also document that fullscreen may only be applied to one screen. 93663 93664 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93665 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93666 Tested-by: Colin Harrison <colin.harrison@virgin.net> 93667 93668commit ce6136f8c553bbc6d3e3affa0faa2afbf8054f44 93669Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93670Date: Tue Mar 23 20:06:33 2010 +0000 93671 93672 Cygwin/X: Make winOverrrideStyle() thread-safe 93673 93674 Make winOverrrideStyle() thread-safe 93675 93676 winOverrideStyle() is called from the internal WM client thread. 93677 93678 Accessing server-internal data structures to get window name and 93679 class is not safe, as there is no lock to ensure we do not collide 93680 with these data structures being updated in the server thread. 93681 93682 Rewrite so the internal client thread uses X client calls to 93683 obtain this data safely 93684 93685 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93686 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93687 Tested-by: Colin Harrison <colin.harrison@virgin.net> 93688 93689commit 0c603509eb7f9c83baf4e00b4558dce78f897ebf 93690Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93691Date: Fri Feb 11 13:15:40 2011 +0000 93692 93693 Cygwin/X: Cosmetic fixes to logging of result from X*TextPropertyToTextList() 93694 93695 Report XLocaleNotSupported result from X*TextPropertyToTextList() 93696 Fix formatting for unknown results reported for X*TextPropertyToTextList() 93697 93698 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93699 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93700 Tested-by: Colin Harrison <colin.harrison@virgin.net> 93701 93702commit 71550a8665d861384332d81239ca0c1586a17137 93703Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93704Date: Fri Jan 28 20:17:22 2011 +0000 93705 93706 Cygwin/X: Decorate function pointers retrieved via GetProcAddress with WINAPI 93707 93708 Decorate function pointers retrieved via GetProcAddress which are currently 93709 missing it with WINAPI, to ensure stdcall convention is used when calling them. 93710 93711 This fixes a crash currently seen when compiled -O2 and the -screen option uses 93712 a size and monitor number e.g. -screen 0 1280x1000@2 93713 93714 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93715 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93716 Tested-by: Colin Harrison <colin.harrison@virgin.net> 93717 93718commit 38a1f5c613a48ef9fd6ba043bc3028f487750d3a 93719Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93720Date: Tue Oct 12 17:12:02 2010 +0100 93721 93722 Cygwin/X: Don't make InputOnly windows visible 93723 93724 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93725 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93726 Tested-by: Colin Harrison <colin.harrison@virgin.net> 93727 93728commit bd288c3458bc1ba2cbb4c8416e5b2dfd849581e6 93729Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93730Date: Thu Jul 22 18:36:51 2010 +0100 93731 93732 Cygwin/X: Fix a GDI bitmap resource leak of window icons 93733 93734 Ensure any icon created specially for a window is destroyed when 93735 the window is destroyed 93736 93737 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93738 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93739 Tested-by: Colin Harrison <colin.harrison@virgin.net> 93740 93741commit 19e764eee0c8b74d877fb2b1d6aedc933976660e 93742Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93743Date: Wed Jul 21 18:11:13 2010 +0100 93744 93745 Cygwin/X: Internal WM workaround for Java AWT bug 93746 93747 Java applications using AWT on JRE 1.6.0 break with non-reparenting WMs AWT 93748 doesn't explicitly know about (See sun bug #6434227) 93749 93750 XDecoratedPeer.handleConfigureNotifyEvent() only processes non-synthetic 93751 ConfigureNotify events to update window location if it's identified the 93752 WM as a non-reparenting WM it knows about (compiz or lookingglass) 93753 93754 Rather than tell all sorts of lies to get XWM to recognize us as one of 93755 those, simply send a synthetic ConfigureNotify for every non-synthetic one 93756 93757 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93758 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93759 Tested-by: Colin Harrison <colin.harrison@virgin.net> 93760 93761commit c5a612fc18a7f5c542fca580494e2c66ecfc2159 93762Author: Jon TURNEY <jon.turney@dronecode.org.uk> 93763Date: Fri Apr 16 21:55:01 2010 +0100 93764 93765 configure: Let configure --enable/disable-aiglx control building of AIGLX for all DDXs 93766 93767 Let configure --enable/disable-aiglx control building of AIGLX for all DDXs. Currently 93768 we can't use --enable/disable-aiglx to control if Xwin DDX is built with AIGLX enabled, 93769 as at the moment it's forced off if we aren't building the X.Org DDX DRI or DRI2 loader 93770 93771 Rearrange things a bit, introducing a new automake conditional, AIGLX_DRI_LOADER to 93772 specifically indicate if the X.Org DDX DRI/DRI2 loader convenience library should be 93773 built, and replace the previous X.Org DDX-specific uses of the AIGLX conditional with that 93774 93775 As before, AIGLX_DRI_LOADER is only enabled if --enable-glx, --enable-aiglx and at least one 93776 of --enable-dri or --enable-dri2 are enabled 93777 93778 This allows the general conditional AIGLX to control if AIGLX is built for the XWin DDX as 93779 well 93780 93781 The C #define AIGLX set by AC_DEFINE(AIGLX) seems to be obsolete, I can't find anything 93782 which checks it 93783 93784 Updated for ajax's "glx: Make --disable-dri not disable AIGLX" patch, which allows DRI2 93785 to be enabled independently of DRI1 93786 93787 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 93788 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 93789 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 93790 93791commit 6f29dbf5e36bef5ecb08f02d367988dee0f9f1cd 93792Author: Jeremy Huddleston <jeremyhu@apple.com> 93793Date: Mon Apr 25 22:00:41 2011 -0700 93794 93795 XQuartz: Use ErrorF rather than fprintf to log errors 93796 93797 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93798 93799commit cb083b05c469352ef80e1005a29ac320f2e4e096 93800Author: Jeremy Huddleston <jeremyhu@apple.com> 93801Date: Mon Apr 25 21:25:10 2011 -0700 93802 93803 XQuartz: stub: Dead code removal 93804 93805 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93806 93807commit 72ed7551f494c61283a7ac3d7b570eac39cc9786 93808Author: Emanuele Giaquinta <emanuele.giaquinta@gmail.com> 93809Date: Mon Apr 25 10:38:17 2011 -0700 93810 93811 XQuartz: pbproxy: LP64: Fix itteration through XGetWindowProperty where sizeof(long) != 4 93812 93813 http://xquartz.macosforge.org/trac/ticket/476 93814 93815 Signed-off-by: Emanuele Giaquinta <emanuele.giaquinta@gmail.com> 93816 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 93817 93818commit 0a60192a85ba9f64b522da181c2fe8a5b93b79df 93819Author: Jeremy Huddleston <jeremyhu@apple.com> 93820Date: Sun Apr 24 22:01:48 2011 -0700 93821 93822 XQuartz: Enable logging to a file for better debugging 93823 93824 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93825 93826commit de4023f194c561b74f8fa904bea3dd5298230cd1 93827Author: Jeremy Huddleston <jeremyhu@apple.com> 93828Date: Sun Apr 24 21:35:04 2011 -0700 93829 93830 XQuartz: Rename launchd-id-prefix to bundle-id-prefix 93831 93832 It's used many other places than just for launchd. 93833 93834 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93835 93836commit e466745109416a2fcdf5d7389c80b814a0363676 93837Author: Jeremy Huddleston <jeremyhu@apple.com> 93838Date: Sun Apr 24 20:35:06 2011 -0700 93839 93840 XQuartz: Dead code removal 93841 93842 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93843 93844commit f1d867c0a1577e2c46d5dd4e262444e0cf6f8e25 93845Author: Jeremy Huddleston <jeremyhu@apple.com> 93846Date: Sat Apr 23 23:22:27 2011 -0700 93847 93848 XQuartz: Silence clang static analyzer 93849 93850 Call to 'malloc' has an allocation size of 0 bytes. 93851 93852 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93853 93854commit 913223e9ddf2cb927ecad99aeadfa730df8561ac 93855Author: Jeremy Huddleston <jeremyhu@apple.com> 93856Date: Sat Apr 23 23:15:49 2011 -0700 93857 93858 XQuartz: xpr: Dead code removal 93859 93860 Assigned value is always the same as the existing value. 93861 93862 Found by clang static analyzer 93863 93864 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93865 93866commit d0caa0e4cd255bea315757be2ff387cf57e8bdf7 93867Author: Jeremy Huddleston <jeremyhu@apple.com> 93868Date: Sat Apr 23 20:12:38 2011 -0700 93869 93870 XQuartz: Silence warnings about deprecated functionality where it is an intended fallback 93871 93872 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93873 93874commit 9244a3a24f63239ec12675f3df091583e06c8314 93875Author: Jeremy Huddleston <jeremyhu@apple.com> 93876Date: Sat Apr 23 12:55:39 2011 -0700 93877 93878 XQuartz: xpr: Use a serial queue rather than pthread mutexes for window_hash 93879 93880 Additionally removes some dead code and fixes double-locking in 93881 xprIsX11Window. xprIsX11Window doesn't need to do any locking because 93882 those resources are protected by the called functions themselves. 93883 93884 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93885 Reviewed-by: Daniel A. Steffen <dsteffen@apple.com> 93886 93887commit bac34a54f7cb84f346b1b833df9917bd1ac70223 93888Author: Jeremy Huddleston <jeremyhu@apple.com> 93889Date: Sat Apr 23 12:11:39 2011 -0700 93890 93891 XQuartz: xpr: Initialize window_hash in xprInit 93892 93893 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93894 93895commit 1596ea72d66a03d9accb534679172ca6f63f78e1 93896Author: Jeremy Huddleston <jeremyhu@apple.com> 93897Date: Sat Apr 23 11:55:49 2011 -0700 93898 93899 XQuartz: Use a lighter spinlock instead of a pthread_mutex_t in QuartzScreenSaver 93900 93901 Currently, we only end up here through a call to QuartzShowFullscreen, and 93902 this is always on the same thread. Future changes (such as further 93903 incorporating libdispatch) may allow this to change, but contention will 93904 remain minimal since the call is infrequent and it is short held. 93905 93906 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93907 Reviewed-by: Daniel A. Steffen <dsteffen@apple.com> 93908 93909commit 3e253c603bc18f06fa48b611797eb5a7c8a96fe4 93910Author: Jeremy Huddleston <jeremyhu@apple.com> 93911Date: Sat Apr 23 01:48:25 2011 -0700 93912 93913 XQuartz: Remove the threadSafety dead-ish code 93914 93915 It's been a few years now since we've needed this to debug thread 93916 boundaries, so punt it out to clean up the namespace polution. 93917 93918 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93919 93920commit a52c8078c9cc83c84a8c6eb58810f49bdb90bcc1 93921Author: Jeremy Huddleston <jeremyhu@apple.com> 93922Date: Sat Apr 23 01:39:16 2011 -0700 93923 93924 XQuartz: Use xorg_backtrace() instead of spewCallStack() 93925 93926 xorg_backtrace() has been in os for two years now, we might as well 93927 start using it. 93928 93929 Ref: 94ed0ba1b5043ad9fc33b42756af447d5ab15bbd 93930 93931 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93932 93933commit d79cc14a51f3e8d4d2f66aad055092672cab1526 93934Author: Jeremy Huddleston <jeremyhu@apple.com> 93935Date: Fri Apr 22 12:05:32 2011 -0700 93936 93937 XQuartz: Fix compilation warnings with XPLUGIN_VERSION >= 4 93938 93939 xprAppleWM.c:143: warning: initialization from incompatible pointer type 93940 xprAppleWM.c:144: warning: initialization from incompatible pointer type 93941 93942 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93943 93944commit 6128544fd58ced0ed738b7150865294d214fb4eb 93945Author: Jeremy Huddleston <jeremyhu@apple.com> 93946Date: Mon Apr 25 12:38:07 2011 -0700 93947 93948 XQuartz: Bump bundle version to 2.7.0 93949 93950 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93951 93952commit 4944de24849a109c65f1b353bc12e44e90a1211d 93953Author: Jeremy Huddleston <jeremyhu@apple.com> 93954Date: Sat Apr 23 23:25:39 2011 -0700 93955 93956 rootless: Fix a typo in RootlessGlyphs which resulted in a garbage value 93957 93958 Found by clang static analyzer 93959 93960 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93961 93962commit 2098cb03c6b64bfca7694fc2b5213edb77bc12e4 93963Author: Jeremy Huddleston <jeremyhu@apple.com> 93964Date: Sat Apr 23 23:27:16 2011 -0700 93965 93966 rootless: Dead code removal 93967 93968 Found by clang static analyzer 93969 93970 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93971 93972commit bb4d145bd25e2aee988b100ecf1105ea3b6a40b8 93973Author: Jeremy Huddleston <jeremyhu@apple.com> 93974Date: Sat Apr 23 21:24:27 2011 -0700 93975 93976 glx: Silence warnings when building with clang 93977 93978 This replaces AX_TLS (GPL3) with XORG_TLS (MIT) 93979 93980 In file included from glapi.c:46: 93981 In file included from ./glapi.h:51: 93982 ./glthread.h:237:20: error: unknown attribute 'tls_model' ignored [-Werror,-Wunknown-attributes] 93983 __attribute__((tls_model("initial-exec"))); 93984 ^ 93985 In file included from glapi.c:46: 93986 ./glapi.h:92:20: error: unknown attribute 'tls_model' ignored [-Werror,-Wunknown-attributes] 93987 __attribute__((tls_model("initial-exec"))); 93988 ^ 93989 glapi.c:82:20: error: unknown attribute 'tls_model' ignored [-Werror,-Wunknown-attributes] 93990 __attribute__((tls_model("initial-exec"))) = NULL; 93991 ^ 93992 glapi.c:85:20: error: unknown attribute 'tls_model' ignored [-Werror,-Wunknown-attributes] 93993 __attribute__((tls_model("initial-exec"))); 93994 ^ 93995 4 errors generated. 93996 93997 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 93998 Tested-by: Jamey Sharp <jamey@minilop.net> 93999 Acked-by: Alan Coopersmith <alan.coopersmith@oracle.com> 94000 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 94001 94002commit 5c9eafc8e5f8575c06591d244c2cb5ea07691cb4 94003Author: Jeremy Huddleston <jeremyhu@apple.com> 94004Date: Sat Apr 23 20:55:53 2011 -0700 94005 94006 render: Silence warnings when building with clang 94007 94008 picture.c:351:37: error: implicit conversion from 'unsigned int' to 'CARD16' (aka 'unsigned short') changes value from 4294967295 to 65535 94009 [-Werror,-Wconstant-conversion] 94010 pFormats[f].direct.alphaMask = Mask(PICT_FORMAT_A(format)); 94011 ^~~~~~~~~~~~~~~~~~~~~~~~~~~ 94012 ... 94013 fatal error: too many errors emitted, stopping now [-ferror-limit=] 94014 94015 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94016 Suggested-by: Jamey Sharp <jamey@minilop.net> 94017 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94018 94019commit c524f8bb768f886d413839bc22184098394c2559 94020Author: Jeremy Huddleston <jeremyhu@apple.com> 94021Date: Sat Apr 23 20:49:27 2011 -0700 94022 94023 os: Silence warnings when building with clang 94024 94025 access.c:1492:20: error: equality comparison with extraneous parentheses [-Werror,-Wparentheses] 94026 if ((host->family == FamilyServerInterpreted)) { 94027 ~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~ 94028 access.c:1492:20: note: use '=' to turn this equality comparison into an assignment 94029 if ((host->family == FamilyServerInterpreted)) { 94030 ^~ 94031 = 94032 access.c:1492:20: note: remove extraneous parentheses around the comparison to silence this warning 94033 if ((host->family == FamilyServerInterpreted)) { 94034 ~ ^ ~ 94035 94036 In file included from xstrans.c:8: 94037 In file included from /usr/X11/include/X11/Xtrans/transport.c:62: 94038 /usr/X11/include/X11/Xtrans/Xtranssock.c:262:5: error: implicit declaration of function 'ErrorF' is invalid in C99 94039 [-Werror,-Wimplicit-function-declaration] 94040 PRMSG (3,"SocketSelectFamily(%s)\n", family, 0, 0); 94041 ^ 94042 94043 log.c:180:29: error: format string is not a string literal [-Werror,-Wformat-nonliteral] 94044 if (asprintf(&logFileName, fname, display) == -1) 94045 ^~~~~ 94046 log.c:190:26: error: format string is not a string literal [-Werror,-Wformat-nonliteral] 94047 if ((asprintf(&suffix, backup, display) == -1) || 94048 ^~~~~~ 94049 log.c:382:25: error: format string is not a string literal [-Werror,-Wformat-nonliteral] 94050 LogVWrite(verb, tmpBuf, args); 94051 ^~~~~~ 94052 94053 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94054 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94055 94056commit ab81aa91404ca0fa6843ce8021cbd9de42255a8f 94057Author: Jeremy Huddleston <jeremyhu@apple.com> 94058Date: Sat Apr 23 20:39:25 2011 -0700 94059 94060 Xext: Silence warnings when building with clang 94061 94062 xvmain.c:1113:22: warning: equality comparison with extraneous parentheses [-Wparentheses] 94063 if ((pf->depth == pDraw->depth) 94064 ~~~~~~~~~~^~~~~~~~~~~~~~~ 94065 xvmain.c:1113:22: note: use '=' to turn this equality comparison into an assignment 94066 if ((pf->depth == pDraw->depth) 94067 ^~ 94068 = 94069 xvmain.c:1113:22: note: remove extraneous parentheses around the comparison to silence this warning 94070 if ((pf->depth == pDraw->depth) 94071 ~ ^ ~ 94072 1 warning generated. 94073 94074 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94075 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94076 94077commit fd086f87cd1bab53c9e918cb687009e7ac8718d8 94078Author: Jeremy Huddleston <jeremyhu@apple.com> 94079Date: Sat Apr 23 20:36:53 2011 -0700 94080 94081 fb: Silence warnings when building with clang 94082 94083 fbpict.c:163:8: warning: implicit conversion from enumeration type 'PictFormatShort' (aka 'enum _PictFormatShort') to different enumeration 94084 type 'pixman_format_code_t' [-Wconversion] 94085 pict->format, 94086 ~~~~~~^~~~~~ 94087 94088 fbbltone.c:486:2: warning: shift result (281474959933440) requires 49 bits to represent, but 'int' only has 32 bits [-Wshift-overflow] 94089 C2_24( 0, 0), C2_24 ( 1, 0), C2_24 ( 2, 0), C2_24 ( 3, 0), 94090 ^~~~~~~~~~~~ 94091 fbbltone.c:474:6: note: instantiated from: 94092 SelMask24(b,1,r)) 94093 ^ 94094 fbbltone.c:429:46: note: instantiated from: 94095 ^ 94096 fbbltone.c:427:18: note: instantiated from: 94097 0xffffff << Mask24Check(x,r)) : 0) 94098 ~~~~~~~~ ^ 94099 94100 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94101 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94102 94103commit aad7b324aefc906f28ac9d10c64650b3445c2ae0 94104Author: Jeremy Huddleston <jeremyhu@apple.com> 94105Date: Sun Apr 24 19:57:22 2011 -0700 94106 94107 os: Add missing _X_ATTRIBUTE_PRINTF to va_list variants 94108 94109 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94110 94111commit 5423da9fb2ec0dfeff866f12cf26fbba04ea673d 94112Author: Nicolas Kaiser <nikai@nikai.net> 94113Date: Sun Apr 24 12:44:15 2011 +0200 94114 94115 xkb: remove duplicated include 94116 94117 Remove duplicated include. 94118 94119 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94120 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 94121 94122commit 79b3a7f83d7cbee397eecebef1af31c048d295e8 94123Author: Nicolas Kaiser <nikai@nikai.net> 94124Date: Sun Apr 24 12:43:05 2011 +0200 94125 94126 test/xi2: remove duplicated include 94127 94128 Remove duplicated include. 94129 94130 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94131 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 94132 94133commit a17c30d43e0db6c65cc73cd2be5193c02f138bf5 94134Author: Nicolas Kaiser <nikai@nikai.net> 94135Date: Sun Apr 24 12:42:07 2011 +0200 94136 94137 hw/xwin: remove duplicated includes 94138 94139 Remove duplicated includes. 94140 94141 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94142 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 94143 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 94144 94145commit c270cfc30c547e28017533f490904013f73cf6f0 94146Author: Nicolas Kaiser <nikai@nikai.net> 94147Date: Sun Apr 24 12:41:11 2011 +0200 94148 94149 hw/xquartz: remove duplicated includes 94150 94151 Remove duplicated includes. 94152 94153 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94154 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 94155 94156commit 50ced6cfa0aef472e713ffccfdba003829df68df 94157Author: Nicolas Kaiser <nikai@nikai.net> 94158Date: Sun Apr 24 12:40:11 2011 +0200 94159 94160 hw/xnest: remove duplicated include 94161 94162 Remove duplicated include. 94163 94164 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94165 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 94166 94167commit 387f45a7077faadf232edc3f608c4a081a7fd290 94168Author: Nicolas Kaiser <nikai@nikai.net> 94169Date: Sun Apr 24 12:39:11 2011 +0200 94170 94171 hw/xfree86/modes: remove duplicated include 94172 94173 Remove duplicated include. 94174 94175 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94176 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 94177 94178commit 78a9ec125dcb1546ca7e7e18aedf2b323eebb1fc 94179Author: Nicolas Kaiser <nikai@nikai.net> 94180Date: Sun Apr 24 12:38:08 2011 +0200 94181 94182 hw/xfree86/fbdevhw: remove duplicated include 94183 94184 Remove duplicated include. 94185 94186 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94187 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 94188 94189commit 0320db25f917980621107b790c4fc5171cb0b6ac 94190Author: Nicolas Kaiser <nikai@nikai.net> 94191Date: Sun Apr 24 12:36:48 2011 +0200 94192 94193 hw/xfree86/dri: remove duplicated includes 94194 94195 Remove duplicated includes. 94196 94197 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94198 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 94199 94200commit b56271b954d6ab8eeb320f314925b77403458b9c 94201Author: Nicolas Kaiser <nikai@nikai.net> 94202Date: Sun Apr 24 12:35:45 2011 +0200 94203 94204 hw/xfree86/ddc: remove duplicated include 94205 94206 Remove duplicated include. 94207 94208 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94209 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 94210 94211commit 86c0c8b9c567f57fe7477d5302fa22dc7557ce90 94212Author: Nicolas Kaiser <nikai@nikai.net> 94213Date: Sun Apr 24 12:34:46 2011 +0200 94214 94215 dix: remove duplicated includes 94216 94217 Remove duplicated includes. 94218 94219 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94220 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 94221 94222commit 9b5046d2130aead79956019faf7103b5c676fd02 94223Author: Nicolas Kaiser <nikai@nikai.net> 94224Date: Sun Apr 24 12:33:32 2011 +0200 94225 94226 Xi: remove duplicated includes 94227 94228 Remove duplicated includes. 94229 94230 Signed-off-by: Nicolas Kaiser <nikai@nikai.net> 94231 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 94232 94233commit c6cb70be1ed7cf73bd3411b8d66ec05a9efcfeb9 94234Author: Søren Sandmann Pedersen <ssp@redhat.com> 94235Date: Mon Mar 28 13:30:52 2011 -0400 94236 94237 Fix trapezoid and triangle rendering to windows 94238 94239 For fbAdd{Traps,Triangles}() and fbRasterizeTrapezoid() this is just a 94240 matter of adding the image offsets to the trap offsets. 94241 94242 For fbShapes, the story is more complicated: 94243 94244 The recently added pixman API did not allow offsetting 94245 trapezoids. Instead, it would use x_dst and y_dst in such a way that 94246 the effect was to only offset the source image. 94247 94248 In pixman 0.21.8, this API has changed such that all the traps are 94249 conceptually rendered to an infinitely big image, and the source and 94250 destination coordinates are then aligned with (0, 0) of that 94251 image. This means offsetting dst_x and dst_y will now offset the 94252 entire drawing, which is similar to how other composite functions 94253 work. 94254 94255 This patch then changes fbComposite{Triangles,Traps} such that the 94256 source image is aligned with the shapes, and the destination 94257 coordinates offset according to drawable->{x, y}. 94258 94259 Reviewed-by: Keith Packard <keithp@keithp.com> 94260 Signed-off-by: Soren Sandmann <ssp@redhat.com> 94261 94262commit 1b96a99d8edd9016bc4a35348f9d5ddb45832f14 94263Author: Søren Sandmann Pedersen <ssp@redhat.com> 94264Date: Thu Mar 10 08:52:41 2011 -0500 94265 94266 fb: Call miCompositeSourceValidate() on the source in fbShapes() 94267 94268 Reviewed-by: Keith Packard <keithp@keithp.com> 94269 Signed-off-by: Soren Sandmann <ssp@redhat.com> 94270 94271commit 04635069554859ec67003b89f56965421cba7f52 94272Author: Søren Sandmann Pedersen <ssp@redhat.com> 94273Date: Tue Mar 29 00:11:00 2011 -0400 94274 94275 render: Remove unused TriStrip and TriFan typedefs 94276 94277 Reviewed-by: Keith Packard <keithp@keithp.com> 94278 Signed-off-by: Soren Sandmann <ssp@redhat.com> 94279 94280commit 2b0cabec620f3a2a5e431052441b092ef979bf94 94281Author: Søren Sandmann Pedersen <ssp@redhat.com> 94282Date: Tue Mar 29 00:07:44 2011 -0400 94283 94284 render: Remove unused fields in the source picture structs 94285 94286 The fields class, stopRange, colorTable and colorTableSize are not 94287 used by any current code. 94288 94289 Reviewed-by: Keith Packard <keithp@keithp.com> 94290 Signed-off-by: Soren Sandmann <ssp@redhat.com> 94291 94292commit b0d84f94393edab395d65d2b2cb983fc9fec3d36 94293Author: Søren Sandmann Pedersen <ssp@redhat.com> 94294Date: Mon Mar 28 20:59:34 2011 -0400 94295 94296 render: Delete PictureGradientColor() 94297 94298 PictureGradientColor(), INTERPOLATE_PIXEL_256() and premultiply() are 94299 not used by anything. 94300 94301 Reviewed-by: Keith Packard <keithp@keithp.com> 94302 Signed-off-by: Soren Sandmann <ssp@redhat.com> 94303 94304commit 61a92a78cd49969f74a046fa26c3199e06365814 94305Author: Søren Sandmann Pedersen <ssp@redhat.com> 94306Date: Tue Mar 29 13:06:36 2011 -0400 94307 94308 Add RegionInitBoxes(), and fix some buggy callers of RegionInit(). 94309 94310 The interface to RegionInit(): 94311 94312 RegionInit (RegionPtr pReg, BoxPtr rect, int size); 94313 94314 is very confusing because it doesn't take a list of boxes, it takes 94315 *one* box, but if that box is NULL, it initializes an empty region 94316 with 'size' rectangles preallocated. 94317 94318 Most callers of this function were correctly passing either NULL or 94319 just one box, but there were three confused cases, where the code 94320 seems to expect a region to be created from a list of boxes. 94321 94322 This patch adds a new function RegionInitBoxes() and fixes those 94323 instances to call that instead. 94324 94325 And yes, the pixman function to initialize a region from a list of 94326 boxes is called init_rects() because pixman is also awesome. 94327 94328 V2: Make RegionInitBoxes() return a Bool indicating whether the call 94329 succeeded, and fix the callers to check this return value. 94330 94331 Reviewed-by: Keith Packard <keithp@keithp.com> 94332 Signed-off-by: Søren Sandmann <ssp@redhat.com> 94333 94334commit c7bce22b58530239e583d91ae56312bad1630da4 94335Author: Søren Sandmann Pedersen <ssp@redhat.com> 94336Date: Mon Mar 28 11:38:11 2011 -0400 94337 94338 Track damage for fbTrapezoids() and fbTriangles(). 94339 94340 These calls no longer go through the CompositePicture() hook, so 94341 damage was no longer generated for them. This patch simply damages the 94342 entire destination clip region. 94343 94344 It would be possible to generate tighter damage for certain operators 94345 such as Over and Add, where blank source pixels have no effect on the 94346 destination, but given that virtually all trapezoid rendering takes 94347 place on pixmaps, it's unlikely that anybody would actually benefit 94348 from this optimization, and the miTrapezoidBounds function did 94349 sometimes show up on profiles, probably because it does several 94350 divisions per trapezoid. 94351 94352 V2: Call DamageRegionProcessPending() - pointed out by Michel Dänzer. 94353 V3: Call DamageRegionProcessPending() *after* rendering - 94354 pointed out by Maarten Maathuis 94355 94356 Reviewed-by: Michel Dänzer <daenzer@vmware.com> 94357 Signed-off-by: Søren Sandmann <ssp@redhat.com> 94358 94359commit 918a9c99cf2ebc73acb34b95f597904b93c690d9 94360Merge: 88c4622b5 5f496bc91 94361Author: Keith Packard <keithp@keithp.com> 94362Date: Fri Apr 22 11:20:16 2011 -0700 94363 94364 Merge remote-tracking branch 'jeremyhu/master' 94365 94366commit 88c4622b594a1725d0cee86bc82ad640d241c520 94367Author: Aaron Plattner <aplattner@nvidia.com> 94368Date: Mon Apr 18 08:23:48 2011 -0700 94369 94370 linux: Retry VT ioctls while errno == EINTR 94371 94372 When the smart scheduler is enabled, the VT ioctls (particularly 94373 VT_WAITACTIVE) can be interrupted by the smart scheduler's SIGALRMs. 94374 Previously, this caused the server to immediately continue on to 94375 ScreenInit, almost certainly causing a crash or failure because the X 94376 server that owned the VT hadn't finished cleaning up. As of commit 94377 7ee965a300c9eddcc1acacf9414cfe3e589222a8, it causes a FatalError 94378 instead. 94379 94380 Retrying the ioctl as long as it fails with errno == EINTR fixes the 94381 problem and allows server regenerations to trigger VT switches that 94382 actually succeed. 94383 94384 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 94385 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 94386 Reviewed-by: Cyril Brulebois <kibi@debian.org> 94387 Signed-off-by: Keith Packard <keithp@keithp.com> 94388 94389commit 302bdc3c9678c028bed71dbe9c5ba04997626b6b 94390Merge: 001b6b8b7 6f97fe5df 94391Author: Keith Packard <keithp@keithp.com> 94392Date: Fri Apr 22 10:48:37 2011 -0700 94393 94394 Merge remote-tracking branch 'whot/for-keith' 94395 94396commit 5f496bc91941140ceaa3061472fda95bd812f2d2 94397Author: Jeremy Huddleston <jeremyhu@apple.com> 94398Date: Fri Apr 22 01:23:09 2011 -0700 94399 94400 XQuartz: Do translation and handoff of NSEvent to X11 in a separate serial queue 94401 94402 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94403 94404commit 72bd232b117b2867282e0ae1855d779e126f912b 94405Author: Jeremy Huddleston <jeremyhu@apple.com> 94406Date: Fri Apr 22 00:39:12 2011 -0700 94407 94408 XQuartz: Send tablet proximity events with tilt and pressure 94409 94410 <rdar://problem/6257569> 94411 94412 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94413 94414commit 034538ea9b4770025e3573bc708039cabbe1e10d 94415Author: Jeremy Huddleston <jeremyhu@apple.com> 94416Date: Thu Apr 21 16:31:58 2011 -0700 94417 94418 XQuartz: Use dispatch_async to handoff the FD 94419 94420 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94421 94422commit ca7b9e6c817681b9cec738e43cf020ac19b5e732 94423Author: Jeremy Huddleston <jeremyhu@apple.com> 94424Date: Thu Apr 21 15:51:32 2011 -0700 94425 94426 configure.ac: Add check for libdispatch when building for darwin 94427 94428 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94429 94430commit 3960115dbc83ec1eb8d9c8e90466af3fa0b32abd 94431Author: Jeremy Huddleston <jeremyhu@apple.com> 94432Date: Thu Apr 21 15:19:12 2011 -0700 94433 94434 XQuartz: Fix prototypes for thread functions 94435 94436 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94437 94438commit 7524dbd06113ec081eaa882aa54e03553ccf96aa 94439Author: Jeremy Huddleston <jeremyhu@apple.com> 94440Date: Thu Apr 21 15:11:52 2011 -0700 94441 94442 XQuartz: Make the DarwinProcessFDAdditionQueue_thread wait 3 seconds to allow xinitrc to catch up 94443 94444 Previously, we weren't always waiting the full three seconds. This should 94445 be better, but is still sub-optimal. We really want to start processing 94446 these once a WM has been started. 94447 94448 http://xquartz.macosforge.org/trac/ticket/416 94449 94450 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 94451 94452commit 6f97fe5df16f947e52875187ff4f9e25fab33ad7 94453Author: Peter Hutterer <peter.hutterer@who-t.net> 94454Date: Thu Apr 21 14:55:22 2011 +1000 94455 94456 dix: improve control flow in QueryTrackers 94457 94458 If the velocity is 0, skip the remainder. 94459 If we're not in range, skip the remainder. 94460 94461 No functional change. 94462 94463 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94464 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94465 94466commit a0eda8cd36ffba2ecc79e0d7a1908c89b45e76c4 94467Author: Peter Hutterer <peter.hutterer@who-t.net> 94468Date: Thu Apr 21 14:45:09 2011 +1000 94469 94470 dix: rename a bunch of variables to be more self-explanatory 94471 94472 i → used_offset 94473 iveloc → initial_velocity 94474 res → result 94475 vdiff → velocity_diff 94476 vfac → velocity_factor 94477 tmp → tracker_velocity 94478 94479 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94480 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94481 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94482 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94483 94484commit e4901905903c5630a1092c69e42f313abd05187b 94485Author: Peter Hutterer <peter.hutterer@who-t.net> 94486Date: Wed Apr 20 15:41:45 2011 +1000 94487 94488 dix: reduce scope of tmp and mult. 94489 94490 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94491 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94492 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94493 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94494 94495commit 81c6e79f42cc263349e04c0d82b3a85d6b557885 94496Author: Peter Hutterer <peter.hutterer@who-t.net> 94497Date: Wed Apr 20 15:39:27 2011 +1000 94498 94499 dix: reduce the work done by ApplySoftening 94500 94501 We can modify fdx/fdy in-place rather than requiring dx/dy as well. And the 94502 decision to soften can be made in the caller (unless decided by the velocity 94503 state). 94504 94505 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94506 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94507 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94508 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94509 94510commit 5870d507cee8547c0993791195fe04d96b271289 94511Author: Peter Hutterer <peter.hutterer@who-t.net> 94512Date: Wed Apr 20 15:32:33 2011 +1000 94513 94514 dix: split softening and constant deceleration into two functions 94515 94516 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94517 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94518 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94519 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94520 94521commit 8ae90552f95f025ad09a17258d56603a424ca076 94522Author: Peter Hutterer <peter.hutterer@who-t.net> 94523Date: Wed Apr 20 15:30:19 2011 +1000 94524 94525 dix: use single return statement in ApplySimpleSoftening 94526 94527 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94528 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94529 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94530 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94531 94532commit c45f4cc6bd885855ecb54d9113010906d9ee58e2 94533Author: Peter Hutterer <peter.hutterer@who-t.net> 94534Date: Wed Apr 20 15:28:32 2011 +1000 94535 94536 dix: rename od, d to prev_delta, delta 94537 94538 And res to result. 94539 94540 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94541 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94542 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94543 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94544 94545commit 5ffe06bada85e98d46a071dad578d72f5b96cb51 94546Author: Peter Hutterer <peter.hutterer@who-t.net> 94547Date: Wed Apr 20 15:26:11 2011 +1000 94548 94549 dix: Don't use short as bool 94550 94551 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94552 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94553 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94554 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94555 94556commit 97c1967689d7990be86ee490690ac5a91935c875 94557Author: Peter Hutterer <peter.hutterer@who-t.net> 94558Date: Wed Apr 20 15:02:46 2011 +1000 94559 94560 dix: rename "res" to "result" for improved readability 94561 94562 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94563 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94564 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94565 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94566 94567commit 1c162ebcaba78930f34639145a8ac01b87f8e6fd 94568Author: Peter Hutterer <peter.hutterer@who-t.net> 94569Date: Wed Apr 20 14:56:44 2011 +1000 94570 94571 dix: change ProcessVelocityData2D to BOOL. 94572 94573 Don't confuse users with a return type of short, that's even less indicative 94574 that it returns 0/non-0 than "int". 94575 94576 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94577 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94578 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94579 94580commit 70b76286dbef99daabbafccd23f3ba436fe07b09 94581Author: Peter Hutterer <peter.hutterer@who-t.net> 94582Date: Wed Apr 20 16:28:21 2011 +1000 94583 94584 dix: add some more documentation to ptraccel code 94585 94586 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94587 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94588 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94589 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94590 94591commit 6b3a4cc6af240c12d05e27250b61e954eafe9f29 94592Author: Peter Hutterer <peter.hutterer@who-t.net> 94593Date: Wed Apr 20 14:44:51 2011 +1000 94594 94595 dix: don't pass the index for a tracker around, pass the tracker 94596 94597 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94598 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94599 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94600 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94601 94602commit 9e8645cacf5f1c398d7a6e837a36c66cf6cf0c9c 94603Author: Peter Hutterer <peter.hutterer@who-t.net> 94604Date: Wed Apr 20 14:23:29 2011 +1000 94605 94606 dix: CalcTracker only uses the tracker, thus only pass the tracker. 94607 94608 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94609 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94610 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94611 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94612 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94613 94614commit 62f4bef82d1252515689420f8aac198e2f29c4f1 94615Author: Peter Hutterer <peter.hutterer@who-t.net> 94616Date: Thu Apr 21 13:58:55 2011 +1000 94617 94618 dix: use single return statement in DoGetDirection 94619 94620 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94621 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94622 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94623 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94624 94625commit 5f8edbe47e3915f71a20f063a9e8b49591239600 94626Author: Peter Hutterer <peter.hutterer@who-t.net> 94627Date: Wed Apr 20 13:12:53 2011 +1000 94628 94629 dix: use single return value in GetDirection 94630 94631 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94632 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94633 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94634 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94635 94636commit afaa17812513232785c8c629294190629197dfd8 94637Author: Peter Hutterer <peter.hutterer@who-t.net> 94638Date: Wed Apr 20 13:12:35 2011 +1000 94639 94640 dix: document GetDirection 94641 94642 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94643 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94644 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94645 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94646 94647commit df90db892d16143742eb3c09e048f5b76457a62e 94648Author: Peter Hutterer <peter.hutterer@who-t.net> 94649Date: Wed Apr 20 11:49:57 2011 +1000 94650 94651 dix: document DoGetDirection's maths 94652 94653 This is the best explanation I can come up with, but it seems to hold true 94654 for my example values. 94655 94656 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94657 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94658 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94659 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94660 94661commit 7513190ec7bc08db1629b04d3f88f86ea2b1588e 94662Author: Peter Hutterer <peter.hutterer@who-t.net> 94663Date: Wed Apr 20 11:47:21 2011 +1000 94664 94665 dix: improve comment, directions flagged are 45° each 94666 94667 The two directions returned by this calculation are always the two boundary 94668 conditions. Since we don't do quadrants but octants, the flagged ones are 94669 45° each. e.g. an angle of 35° flags E and NE. 94670 94671 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94672 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94673 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94674 94675commit e804ae85fec290894af69c2e65df55e73b417287 94676Author: Peter Hutterer <peter.hutterer@who-t.net> 94677Date: Wed Apr 20 11:12:35 2011 +1000 94678 94679 dix: improve readbility of DoGetDirection. 94680 94681 Use enums for the direction bits, not hardcoded bitfield values that are 94682 added up. 94683 94684 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94685 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94686 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94687 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94688 94689commit 709b291972f8c65f4427f36b45ee78c3895d1e6e 94690Author: Peter Hutterer <peter.hutterer@who-t.net> 94691Date: Wed Apr 20 10:40:28 2011 +1000 94692 94693 dix: only use a single return where only one is needed. 94694 94695 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94696 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94697 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94698 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94699 94700commit 8c528c667da7c7fba0ba07973a06b2d4613fbf70 94701Author: Peter Hutterer <peter.hutterer@who-t.net> 94702Date: Wed Apr 20 13:14:11 2011 +1000 94703 94704 dix: fix typo in direction calculation 94705 94706 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94707 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94708 Reviewed-by: Jamey Sharp <jamey@minilop.net> 94709 Reviewed-by: Simon Thum <simon.thum@gmx.de> 94710 94711commit 47f8cba6f3ae24e5dcdc4e348bdaaaf00d25a91c 94712Author: Peter Hutterer <peter.hutterer@who-t.net> 94713Date: Mon Apr 11 14:50:16 2011 +1000 94714 94715 xfree86: removed unused "event" variable in xf86PostMotionEventM 94716 94717 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94718 94719commit c6212a3e5dc43bd3b5263b392cb75ec96cdf4e4f 94720Author: Peter Hutterer <peter.hutterer@who-t.net> 94721Date: Fri Apr 15 14:13:51 2011 +1000 94722 94723 dix: silence compiler warning 94724 94725 resource.c: In function 'AddResource': 94726 resource.c:493:3: warning: format '%lx' expects type 'long unsigned int', 94727 but argument 3 has type 'RESTYPE' 94728 94729 RESTYPE is uint32_t, not long. 94730 94731 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94732 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94733 94734commit 196d679bed2e9959d2fca89d4ee4bbc551681d0b 94735Author: Peter Hutterer <peter.hutterer@who-t.net> 94736Date: Fri Apr 15 14:51:06 2011 +1000 94737 94738 test: remove glib dependency 94739 94740 The few features from the glib test suite we used can be replaced with 94741 assert and printf. This patch is a simple replacement for these two 94742 g_assert → assert 94743 g_test_message → printf 94744 94745 g_test_init is removed and so is g_test_bug_base. g_test_run replaced with a 94746 simple return 0. 94747 94748 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94749 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 94750 Acked-by: Gaetan Nadon <memsize@videotron.ca> 94751 94752commit 071a6ac4d0c347aa7fc6efe37f4f6992524d7ef1 94753Author: Peter Hutterer <peter.hutterer@who-t.net> 94754Date: Mon Apr 11 15:22:52 2011 +1000 94755 94756 input: remove GetKeyboardValuatorEvents, this is now unnecessary. 94757 94758 GetKeyboardValuatorEvents handles NULL valuator masks already, so the 94759 GetKeyboardEvents wrapper is not needed. Rename GKVE to GKE. 94760 94761 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94762 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 94763 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94764 94765commit 91a735328c59db006f0ed52b49e9c59895796c0c 94766Author: Peter Hutterer <peter.hutterer@who-t.net> 94767Date: Wed Mar 9 10:45:55 2011 +1000 94768 94769 test: add some XIPassiveGrab protocol testing. 94770 94771 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94772 Acked-by: Daniel Stone <daniel@fooishbar.org> 94773 94774commit 60b08e013dd1e971f82e5bc8708d3f120c217497 94775Author: Peter Hutterer <peter.hutterer@who-t.net> 94776Date: Wed Mar 9 10:44:48 2011 +1000 94777 94778 Xi: don't swap the status byte in the XIPassiveGrab replies 94779 94780 Reported-by: Julien Cristau <jcristau@debian.org> 94781 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94782 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94783 94784commit a3f37f3698880aec508b5ecfb88daf39360610aa 94785Author: Peter Hutterer <peter.hutterer@who-t.net> 94786Date: Wed Mar 9 10:07:19 2011 +1000 94787 94788 Xi: exit with error value if CheckGrabValues failed. 94789 94790 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94791 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 94792 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94793 94794commit f4d9ff73b161d84d3fcacc5834fa714c113c7a10 94795Author: Peter Hutterer <peter.hutterer@who-t.net> 94796Date: Wed Mar 9 09:54:12 2011 +1000 94797 94798 Xi: fix reply swapping function check for XIPassiveGrabDevice 94799 94800 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94801 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 94802 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94803 94804commit 2a49ffa3c94819e9f28dd939f23ee8a675258172 94805Author: Peter Hutterer <peter.hutterer@who-t.net> 94806Date: Wed Mar 9 09:53:47 2011 +1000 94807 94808 Xi: return the bad device ID if a passive grab fails with BadDevice. 94809 94810 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94811 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 94812 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 94813 94814commit 315aa8444b13f827237674535eb2ad55a2066808 94815Author: Peter Hutterer <peter.hutterer@who-t.net> 94816Date: Wed Mar 23 10:32:35 2011 +1000 94817 94818 configure: actually require macros 1.13 94819 94820 Introduced in 6a5bf15fa99cf5b2358b3b3e2f29e5044aa8724a 94821 94822 Reported-by: Dave Airlie <airlied@redhat.com> 94823 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94824 Reviewed-by: Dave Airlie <airlied@redhat.com> 94825 Reviewed-by: Julien Cristau <jcristau@debian.org> 94826 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 94827 94828commit f0f0eec869d932a2d9228ff3e41e9e7cda155e68 94829Author: Peter Hutterer <peter.hutterer@who-t.net> 94830Date: Mon Apr 11 11:24:29 2011 +1000 94831 94832 Xi: silence valgrind warning. (#36120) 94833 94834 Conditional jump or move depends on uninitialised value(s) 94835 at 0x4357A1: GetEventMask (events.c:454) 94836 by 0x43B9E8: DeliverEventsToWindow (events.c:2029) 94837 by 0x4E0C59: SendEventToAllWindows (exevents.c:2125) 94838 by 0x4E8301: XISendDeviceHierarchyEvent (xichangehierarchy.c:118) 94839 by 0x426F99: DisableDevice (devices.c:507) 94840 by 0x46BF72: xf86Wakeup (xf86Events.c:457) 94841 by 0x432ABA: WakeupHandler (dixutils.c:419) 94842 by 0x45B708: WaitForSomething (WaitFor.c:235) 94843 by 0x42E8D9: Dispatch (dispatch.c:367) 94844 by 0x422DC9: main (main.c:287) 94845 Uninitialised value was created by a stack allocation 94846 at 0x4E8190: XISendDeviceHierarchyEvent (xichangehierarchy.c:61) 94847 94848 Conditional jump or move depends on uninitialised value(s) 94849 at 0x43BB78: DeliverEventsToWindow (events.c:2010) 94850 by 0x4DDEEA: FindInterestedChildren (exevents.c:2103) 94851 by 0x4DDEFF: FindInterestedChildren (exevents.c:2104) 94852 by 0x4DDEFF: FindInterestedChildren (exevents.c:2104) 94853 by 0x4DDEFF: FindInterestedChildren (exevents.c:2104) 94854 by 0x4DDEFF: FindInterestedChildren (exevents.c:2104) 94855 by 0x4E0C6F: SendEventToAllWindows (exevents.c:2127) 94856 by 0x4E8301: XISendDeviceHierarchyEvent (xichangehierarchy.c:118) 94857 by 0x426F99: DisableDevice (devices.c:507) 94858 by 0x46BF72: xf86Wakeup (xf86Events.c:457) 94859 by 0x432ABA: WakeupHandler (dixutils.c:419) 94860 by 0x45B708: WaitForSomething (WaitFor.c:235) 94861 Uninitialised value was created by a stack allocation 94862 at 0x4E8190: XISendDeviceHierarchyEvent (xichangehierarchy.c:61) 94863 94864 Set the type of dummyDev to SLAVE. The jump listed above comes from a check 94865 to IsMaster() in GetEventMask() that would then set the 94866 XIAllMasterDevices mask. 94867 Hierarchy events can only be set for XIAllDevices so the above IsMaster() 94868 check had no effect and the device type doesn't really matter anyway beyond 94869 shuting up valgrind. 94870 94871 Also initialize dummyDev to 0 to ease future debugging. 94872 94873 X.Org Bug 36120 <http://bugs.freedesktop.org/show_bug.cgi?id=36120> 94874 94875 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94876 94877commit 419a27b5219a739f2fbd50cc96a1b54c469e4a88 94878Author: Peter Hutterer <peter.hutterer@who-t.net> 94879Date: Tue Apr 12 10:17:07 2011 +1000 94880 94881 Xi: fix valuator alignment in DeepCopyDeviceClasses (#36119) 94882 94883 commit 678f5396c91b3d0c7572ed579b0a4fb62b2b4655 only fixed the 94884 initialization, not the copy. After a slave device change, the valuator 94885 were out of alignment again. 94886 94887 X.Org Bug 36119 <http://bugs.freedesktop.org/show_bug.cgi?id=36119> 94888 94889 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94890 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 94891 94892commit 7762de65e153790ee9f63903964d168a6680d815 94893Author: Alexandr Shadchin <alexandr.shadchin@gmail.com> 94894Date: Sat Apr 9 17:45:10 2011 +0600 94895 94896 Simplify auto-detect mouse for WSCONS_SUPPORT 94897 94898 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 94899 Reviewed-by: Matthieu Herrbb <matthieu.herrb@laas.fr> 94900 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 94901 94902commit 001b6b8b70734db1fa2f68e45c1db8337ba9f662 94903Author: Adam Jackson <ajax@redhat.com> 94904Date: Fri Apr 15 13:01:37 2011 -0400 94905 94906 glx: Make --disable-dri not disable AIGLX 94907 94908 Either the DRI1 or DRI2 loaders are sufficient. 94909 94910 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 94911 Signed-off-by: Adam Jackson <ajax@redhat.com> 94912 94913commit f8540b9dcc1fba886be5c4ce7ea0d74952cb48c4 94914Author: Adam Jackson <ajax@redhat.com> 94915Date: Fri Apr 15 12:57:05 2011 -0400 94916 94917 glx: Use 0 rather than garbage for unknown INTEL_swap_event types 94918 94919 Otherwise the garbage you return could well be numerically identical to 94920 one of the swap type tokens, and apps which rely on us to tell the truth 94921 would be in trouble. 94922 94923 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 94924 Signed-off-by: Adam Jackson <ajax@redhat.com> 94925 94926commit dc0cf7596782087bdda0e7f9cd2f60907c45b2c4 94927Author: Adam Jackson <ajax@redhat.com> 94928Date: Tue Mar 29 10:09:46 2011 -0400 94929 94930 Revert "composite: Don't backfill non-bg-None windows" 94931 94932 This reverts commit 6dd775f57d2f94f0ddaee324aeec33b9b66ed5bc. 94933 94934 Bugzilla: https://bugs.freedesktop.org/34427 94935 94936 Acked-by: Alex Deucher <alexdeucher@gmail.com> 94937 Signed-off-by: Adam Jackson <ajax@redhat.com> 94938 94939commit 6a433b67ca15fd1ea58334e607f867554f227451 94940Author: Adam Jackson <ajax@redhat.com> 94941Date: Mon Mar 28 12:30:09 2011 -0400 94942 94943 glx: Fix lifetime tracking for pixmaps 94944 94945 GLX pixmaps take a reference on the underlying pixmap; X and GLX pixmap 94946 IDs can be destroyed in either order with no error. Only windows need 94947 to be tracked under both XIDs. 94948 94949 Fixes piglit/glx-pixmap-life. 94950 94951 Reviewed-by: Michel Dänzer <michel@daenzer.net> 94952 Signed-off-by: Adam Jackson <ajax@redhat.com> 94953 94954commit b3d2164a0361f636bfe77b51456bee9213af4f13 94955Author: Ville Syrjälä <ville.syrjala@nokia.com> 94956Date: Tue Apr 12 17:16:50 2011 +0300 94957 94958 dri2: Pass out_count by value to update_dri2_drawable_buffers() 94959 94960 update_dri2_drawable_buffers() doesn't modify out_count, so pass it 94961 by value. 94962 94963 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 94964 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 94965 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 94966 94967commit 93c833ee84a3465ec5d251e622ba26434cb532f8 94968Author: Ville Syrjälä <ville.syrjala@nokia.com> 94969Date: Tue Apr 12 17:13:28 2011 +0300 94970 94971 dri2: Handle calloc() failure 94972 94973 Don't access invalid memory if calloc() fails to allocate the buffers 94974 array. 94975 94976 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 94977 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 94978 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 94979 94980commit b2997431fd426ab318bc5dfd2cd43956d733ebec 94981Author: Jeremy Huddleston <jeremyhu@apple.com> 94982Date: Wed Apr 13 11:51:30 2011 -0700 94983 94984 Send events that were missing from RRSelectInput 94985 94986 The RANDR spec (randrproto.txt) specifies that RRSelectInput will send out 94987 events corresponding to the event mask, if there have been changes to 94988 CRTCs or outputs. Only screen events were being generated, however. 94989 94990 Fixes http://bugs.freedesktop.org/21760 94991 94992 Signed-off-by: Federico Mena Quintero <federico@novell.com> 94993 Reviewd-by: Keith Packard <keithp@keithp.com> 94994 Signed-off-by: Keith Packard <keithp@keithp.com> 94995 94996commit e409fb32b97033718f270a273f29f24c0b562b84 94997Author: Erkki Seppälä <erkki.seppala@vincit.fi> 94998Date: Tue Apr 12 12:55:56 2011 +0300 94999 95000 damage: use DamageReportDamage for the initial borderClip damage report 95001 95002 Instead of using DamageDamageRegion for reporting the first (virtual) 95003 damage in ProcDamageCreate that covers the borderClip of the drawable 95004 window, use a function DamageReportDamage directly (previously called 95005 damageReportDamage). This avoids sending all other damage listeners a 95006 full window update when a new damage object is created. 95007 95008 As this patch makes DamageReportDamage a public interface, the 95009 function has been moved into the part of the file that contains all 95010 the other public functions. The function has not been otherwise 95011 modified. 95012 95013 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 95014 Reviewed-by: Keith Packard <keithp@keithp.com> 95015 Signed-off-by: Keith Packard <keithp@keithp.com> 95016 95017commit 274dca8f2c6707121d45df8015fe7eddb129dec9 95018Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95019Date: Mon Apr 4 22:31:42 2011 +0300 95020 95021 dix: don't free stranger pointers inside AllocARGBCursor 95022 95023 This seems a good convention to follow: if pointers are allocate outside a 95024 given function, then free there as well when a failure occurs. 95025 95026 AllocARGBCursor and its callers were mixing up the freeing of resources and 95027 causing a particular double free inside TileScreenSaver (srcbits and mskbits). 95028 95029 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95030 Reviewed-by: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 95031 95032commit f603061e9482ad5caf1975ba5395b3294852d072 95033Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95034Date: Mon Apr 4 21:40:06 2011 +0300 95035 95036 os: fix use after free in EstablishNewConnections 95037 95038 In the case of failure on AllocNewConnection, new_trans_conn cannot be 95039 dereferenced because it's already freed. Swapping the order of this logic fix 95040 the changes introduced in 04956b80431169e0ae713a3e6ba4cdc157ce3a66. 95041 95042 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95043 CC: Jeremy Huddleston <jeremyhu@freedesktop.org> 95044 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 95045 95046commit 82498e3c2cce6f515063ecb4b6ae9303e828da00 95047Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95048Date: Mon Apr 4 20:25:32 2011 +0300 95049 95050 xfree86: xv: set pointers to NULL in xf86XVFreeAdaptor 95051 95052 As a good practice and for eventual double frees. 95053 95054 The reason of this patch is due the resilience of xf86XVInitAdaptors, where 95055 for any adaptor failure it's able to keep trying registering the following 95056 ones. 95057 95058 I discussed briefly with Pauli and Ville about a bigger refactoring of such 95059 function, doing it in a way to return instantly when a failure happens; after 95060 all that's how mostly of the other driver functions work. Instead, we just 95061 thought that xf86XVInitAdaptors is wise and cool, and eventually other driver 95062 functions should be even following the main idea of resilience. 95063 95064 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95065 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 95066 95067commit 81414c1c836ae30628606545edbf7392d9b3d009 95068Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95069Date: Thu Mar 31 23:44:03 2011 +0300 95070 95071 xfree86: xv: fix double free in xf86XVFreeAdaptor 95072 95073 When xf86XVFreeAdaptor is called more than once in xf86XVInitAdaptors (it may, 95074 but not often), the conditional being changed in this patch will always take 95075 true path and will keep freeing pAdaptor->pAttributes, thus letting the system 95076 error-prone. 95077 95078 This patch fix such problem checking for a pointer instead the number of 95079 attributes. Such pointer will be deallocated when xf86XVFreeAdaptor is called 95080 first and will not let the code re-run in the following calls. This is a bit 95081 similar how the surroundings code is already doing. 95082 95083 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95084 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 95085 95086commit 74476b700f1e499a731ba2ddbba87b12b9b5139b 95087Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95088Date: Thu Mar 31 17:46:42 2011 +0300 95089 95090 xfree86: loader: use one exit code only for readability 95091 95092 No functional changes. Spaghetti code for the win! \o/ 95093 95094 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95095 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95096 95097commit 7aad7dbf409eaa16a2c80d76ba2a42dc84501919 95098Author: Peter Hutterer <peter.hutterer@who-t.net> 95099Date: Thu Apr 7 08:43:13 2011 +1000 95100 95101 Xext: use EXT_MASK macro instead of manual & 0x7f 95102 95103 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95104 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 95105 95106commit b28a1af55cf1ad2a882cc8cd73b77341dec7ff5a 95107Author: Chase Douglas <chase.douglas@canonical.com> 95108Date: Wed Apr 6 14:51:45 2011 -0400 95109 95110 Fix unset valuator handling for XI 1.x valuator events again 95111 95112 Set the valuator values for unset masked absolute valuators in the 95113 internal device event. This ensures the values will always be correct in 95114 getValuatorEvents even if the device has been removed. 95115 95116 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 95117 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95118 95119commit 3085b178626c957854385c7a88a05ec3c7eb74f3 95120Merge: c9d89cec1 3c45b59e6 95121Author: Keith Packard <keithp@keithp.com> 95122Date: Wed Apr 6 08:34:10 2011 -0700 95123 95124 Merge remote-tracking branch 'jturney/master' 95125 95126commit c9d89cec1407550cb2568f4cef146c93607bbae6 95127Merge: a52049de2 bc61787a2 95128Author: Keith Packard <keithp@keithp.com> 95129Date: Mon Apr 4 11:57:39 2011 -0700 95130 95131 Merge remote-tracking branch 'vignatti/for-keith' 95132 95133commit bc61787a20e7683cbc4dfa45fe855da98a8c0cd0 95134Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95135Date: Tue Mar 29 18:22:50 2011 +0300 95136 95137 render: fix memory leaks in ProcRenderCompositeGlyphs 95138 95139 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95140 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95141 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95142 Reviewed-by: Soren Sandmann <ssp@redhat.com> 95143 95144commit 45b6667b651a0a26b17f64c9e99d70784045e4bf 95145Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95146Date: Tue Mar 29 17:17:40 2011 +0300 95147 95148 xkb: fix fd leak in XkbDDXListComponent 95149 95150 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95151 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95152 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95153 95154commit 0b288c8738a97cf6aa3f36aa5c05e7ac2a5cbca8 95155Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95156Date: Tue Mar 29 16:54:14 2011 +0300 95157 95158 dix: fix memory leak in ProcListExtensions 95159 95160 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95161 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95162 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95163 95164commit 90e03ab026cd8ffdc6202e6b8bae119717ee8528 95165Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95166Date: Tue Mar 29 16:48:21 2011 +0300 95167 95168 mi: fix memory leak in miZeroLine 95169 95170 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95171 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95172 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95173 95174commit daae5e5de194757f7084f9b2b24353c34b961f19 95175Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95176Date: Mon Mar 28 20:13:11 2011 +0300 95177 95178 xi: fix memory leak in AddExtensionClient 95179 95180 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95181 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95182 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95183 95184commit c1875ac25b7b730d464c98c4c151c35efd64a562 95185Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95186Date: Mon Mar 28 20:09:04 2011 +0300 95187 95188 dix: fix memory leak in AllocShared 95189 95190 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95191 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95192 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95193 95194commit 6e5020f46165862ff09211d5425c8b6a13966303 95195Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95196Date: Mon Mar 28 20:07:06 2011 +0300 95197 95198 dix: fix memory leak in AllocPseudo 95199 95200 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95201 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95202 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95203 95204commit a6c71ce5d2d2fe89e07a2ef5041c915acc3dc686 95205Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95206Date: Mon Mar 28 19:21:28 2011 +0300 95207 95208 os: fix memory and fd leaks in Popen 95209 95210 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95211 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95212 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95213 95214commit bafec9a25efa902bef2a3730dc44dc50f0e45877 95215Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95216Date: Mon Mar 28 19:19:57 2011 +0300 95217 95218 os: use DebugF for debugging 95219 95220 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95221 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95222 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95223 95224commit 719b37c33a89d6ad86c79097e5c67b9b3cea5aac 95225Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95226Date: Tue Mar 29 20:04:24 2011 +0300 95227 95228 xfree86: fix memory leak in xf86LoadModules 95229 95230 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95231 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 95232 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95233 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95234 95235commit af054db005b48160b117dfd5bec5f821ee614ea9 95236Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95237Date: Tue Mar 29 19:51:23 2011 +0300 95238 95239 xfree86: fix bad free configInputDevices 95240 95241 introduced in 93ca526892c0d22afa05cce6496198c652043a19. 95242 95243 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95244 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 95245 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95246 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95247 95248commit ac2fac24d8baa56f369f4e54908bc00c2b73a60a 95249Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95250Date: Tue Mar 29 19:49:32 2011 +0300 95251 95252 xfree86: fix memory leaks in configLayout 95253 95254 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95255 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95256 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95257 95258commit bc123319206930638ef1217922824d586d4ada6d 95259Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95260Date: Fri Mar 25 22:21:21 2011 +0200 95261 95262 xfree86: fix memory leak in xf86ConfigFbEntity 95263 95264 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95265 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95266 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95267 95268commit 2aa935bc5cc1e2d5365a97b8c5bb3d33eb5fc758 95269Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95270Date: Fri Mar 25 22:10:55 2011 +0200 95271 95272 fb: fix memory leak in fbOverlayFinishScreenInit 95273 95274 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95275 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95276 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95277 95278commit 4d87606a0d0eb63458098028c300c39c6f1bd2bf 95279Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95280Date: Fri Mar 25 22:07:44 2011 +0200 95281 95282 Xi: fix memory leak in ProcXGetSelectedExtensionEvents 95283 95284 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95285 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95286 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95287 95288commit 12e46e83733b47d2704e1509960192365102af46 95289Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95290Date: Fri Mar 25 22:07:31 2011 +0200 95291 95292 dix: fix memory leak in SetDefaultFontPath 95293 95294 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95295 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95296 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95297 95298commit 7ae46b69ba3f05f46529131e6a864904967cde3a 95299Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95300Date: Fri Mar 25 21:33:15 2011 +0200 95301 95302 mi: fix memory leak in miFillUniqueSpanGroup 95303 95304 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95305 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95306 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95307 95308commit 623bb34342641811e2151232f1ab4e2ee3d6b871 95309Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95310Date: Fri Mar 25 19:27:17 2011 +0200 95311 95312 xi: fix memory leak in ProcXIQueryDevice 95313 95314 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95315 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95316 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95317 95318commit dce2f10cf7bef9929cefadb5088d5b66df43a865 95319Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95320Date: Fri Mar 25 19:21:03 2011 +0200 95321 95322 mi: fix memory leak in miInitVisuals 95323 95324 Free the pointers inside miInitVisuals, so the callers of this function 95325 (fboverlay.c and fbscreen.c) don't need to worry with deallocation in the case 95326 of failure. 95327 95328 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95329 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95330 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95331 95332commit 8ab92cd9822510f426d179a636ef34bb0ace3bb3 95333Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95334Date: Fri Mar 25 18:56:02 2011 +0200 95335 95336 xfree86: dri2: fix memory leak and free resources properly 95337 95338 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95339 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95340 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95341 95342commit aa7096ca6f108e399d9916639cf20c57f9776305 95343Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95344Date: Fri Mar 25 18:24:34 2011 +0200 95345 95346 xorg: remove unused pointer values all over the server 95347 95348 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95349 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95350 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95351 95352commit 49051355d47849b3dd73631e3e2287e319a7f4e5 95353Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95354Date: Wed Mar 23 18:08:29 2011 +0200 95355 95356 dix: remove unused macro 95357 95358 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95359 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95360 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95361 95362commit f1a80e1268fa4d235a411a8cdf20b4f13eafcde0 95363Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95364Date: Wed Mar 23 18:06:51 2011 +0200 95365 95366 dix: remove unused debug code 95367 95368 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95369 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 95370 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95371 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95372 95373commit a52049de2f846fe984d4db5ac8d2c1826c7b2d0b 95374Merge: d044d3675 266ea63bc 95375Author: Peter Hutterer <peter.hutterer@who-t.net> 95376Date: Mon Apr 4 09:58:53 2011 +1000 95377 95378 Merge branch 'master' of git://people.freedesktop.org/~herrb/xserver into for-keith 95379 95380commit d044d3675635f037bf0eb30e47f82460f78227d1 95381Author: Tiago Vignatti <tiago.vignatti@nokia.com> 95382Date: Thu Mar 31 16:26:06 2011 +0300 95383 95384 xfree86: loader: fix memory leaks in LoaderListDirs 95385 95386 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 95387 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95388 Reviewed-by: Nicolas Peninguy <nico@lostgeeks.org> 95389 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95390 95391commit 9c4aae2141161e4bf69313a771db91c0acc4cc83 95392Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 95393Date: Wed Mar 30 16:47:31 2011 +0300 95394 95395 xkb: Prevent leaking of XKB geometry information on copy. 95396 95397 Currently shapes, sections and doodads may leak on copy. 95398 95399 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95400 Reviewed-by: Erkki Seppälä <erkki.seppala@vincit.fi> 95401 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 95402 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95403 95404commit 29d63ba175ff1ef1587c390b18ce61c8f1c150f3 95405Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 95406Date: Wed Mar 30 16:47:30 2011 +0300 95407 95408 xkb: Introduce helper function to handle similar reallocations. 95409 95410 This is preparation for a memory leak fix and doesn't contain any 95411 functional changes. 95412 95413 Note that two variables are generally used for reallocation and 95414 clearing of arrays: geom->sz_elems (reallocation) and geom->num_elems 95415 (clearing). The interface of XkbGeomRealloc is deliberately kept 95416 simple and it only accepts geom->sz_elems as argument, because that is 95417 needed to determine whether the array needs to be resized. When the 95418 array is cleared, we just assume that either geom->sz_elems and 95419 geom->num_elems are synchronized to be equal or that unused elements 95420 are cleared whenever geom->num_elems is set to be less than 95421 geom->sz_elems without reallocation. 95422 95423 Reviewed-by: Erkki Seppälä <erkki.seppala@vincit.fi> 95424 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 95425 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95426 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95427 95428commit f40103cee1d591387359f401a5a7c21f4105aeb4 95429Author: Chase Douglas <chase.douglas@canonical.com> 95430Date: Thu Mar 31 11:29:01 2011 -0400 95431 95432 Don't report old relative values in getValuatorEvents 95433 95434 Relative valuator values should not be reported in any future events. If 95435 a relative valuator value is not set in an internal event, set the value 95436 to 0 for XI 1.x valuator events sent over the wire. 95437 95438 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 95439 Reviewed-by: Simon Thum <simon.thum@gmx.de> 95440 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95441 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95442 95443commit 8199eac443d2c22d313cb23e39d5e607a8cc7f99 95444Author: Chase Douglas <chase.douglas@canonical.com> 95445Date: Mon Mar 28 16:04:48 2011 -0400 95446 95447 Handle non continuous valuator data in getValuatorEvents 95448 95449 This allows for masked valuators to be handled properly in XI 1.x 95450 events. Any unset valuators in the device event are set to the last 95451 known value when transmitted on the wire through XI 1.x valuator events. 95452 95453 Fixes https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/736500 95454 95455 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 95456 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95457 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95458 95459commit 266ea63bc3e052b753c5484fa08dcc4fb67c8952 95460Author: Matthieu Herrb <matthieu.herrb@laas.fr> 95461Date: Fri Apr 1 22:55:57 2011 +0200 95462 95463 Xi: fix querydevice request swapping 95464 95465 WriteReplyToClient() swaps rep.length, so it can't be used 95466 on return of WriteReplyToClient(). So save it's value for later 95467 use. 95468 95469 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 95470 Reviewed-by: Julien Cristau <jcristau@debian.org> 95471 95472commit 552c78592c5e5f52fe16e2429f77c63adf398247 95473Author: Matthieu Herrb <matthieu.herrb@laas.fr> 95474Date: Fri Apr 1 22:54:54 2011 +0200 95475 95476 Xext: fix test on extension number for the swapped case. 95477 95478 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 95479 Reviewed-by: Julien Cristau <jcristau@debian.org> 95480 95481commit a074e6b6754d6c2706b2b5de54e22cbffc5b1a17 95482Author: Matthieu Herrb <matthieu.herrb@laas.fr> 95483Date: Fri Apr 1 22:53:40 2011 +0200 95484 95485 Xi: add XI_Focus{In,Out} to swapped events. 95486 95487 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 95488 Reviewed-by: Julien Cristau <jcristau@debian.org> 95489 95490commit 81257377a2d011ce47fba5822df0f7918dea1d72 95491Author: Matthieu Herrb <matthieu.herrb@laas.fr> 95492Date: Fri Apr 1 22:51:48 2011 +0200 95493 95494 Xi: take XI2 requests into account also for the swapping case. 95495 95496 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 95497 Reviewed-by: Julien Cristau <jcristau@debian.org> 95498 95499commit f9834d312e3059073e8ad77d9f9d57cb9d96e1e5 95500Author: Dave Airlie <airlied@redhat.com> 95501Date: Thu Mar 31 15:33:46 2011 +1000 95502 95503 fb: cleanup fbChangeWindowAttributes 95504 95505 This cleans up the duplication in fbChangeWindowAttributes, 95506 and fixes a bug if the fb24_32ReformatTile ever failed, 95507 since the old code would happily dereference it in the fbEvenTile 95508 call a few lines later. 95509 95510 Signed-off-by: Dave Airlie <airlied@redhat.com> 95511 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 95512 95513commit 36c7158133660520034d645b124c2c973d2971bb 95514Author: Dave Airlie <airlied@redhat.com> 95515Date: Thu Mar 31 13:21:44 2011 +1000 95516 95517 fb: drop comments around 24-bit support. 95518 95519 Signed-off-by: Dave Airlie <airlied@redhat.com> 95520 Reviewed-by: Keith Packard <keithp@keithp.com> 95521 95522commit ac2ae2ed378cc877444bec17b353dda83b620186 95523Author: Dave Airlie <airlied@redhat.com> 95524Date: Tue Mar 29 14:24:52 2011 +1000 95525 95526 fb: add back fb defines for driver compat. 95527 95528 we could drop these really and just fixup the drivers, but 95529 since they'll build fine but fail to work this seems safer. 95530 95531 Signed-off-by: Dave Airlie <airlied@redhat.com> 95532 Reviewed-by: Keith Packard <keithp@keithp.com> 95533 95534commit be9cfb249da18d134c1d65bbb75161aff65f2942 95535Author: Dave Airlie <airlied@redhat.com> 95536Date: Tue Mar 29 14:27:41 2011 +1000 95537 95538 fb: drop defines that aren't used/necessary anymore. 95539 95540 This was generated by: 95541 cd fb 95542 coan source --replace -DFB_SCREEN_PRIVATE -DFB_24BIT -DFB_24_32BIT -DFB_SCREEN_PRIVATE -UFBNOPIXADDR -UFBNO24BIT -UFBNO24_32 *.[ch] 95543 95544 A follow up patch readds the FB_24_32BIT define for Intel UXA. 95545 95546 Signed-off-by: Dave Airlie <airlied@redhat.com> 95547 Reviewed-by: Keith Packard <keithp@keithp.com> 95548 95549commit e0a2ad51dfb7373aa602335490d9666d6101b5ea 95550Merge: 327e1d880 3d688316a 95551Author: Keith Packard <keithp@keithp.com> 95552Date: Wed Mar 30 10:51:27 2011 -0700 95553 95554 Merge remote-tracking branch 'ajax/xserver-next' 95555 95556commit ac00ab77d5a00cfd198958aa1afaa4c3ccc6d7bc 95557Author: Chase Douglas <chase.douglas@canonical.com> 95558Date: Mon Mar 28 16:04:47 2011 -0400 95559 95560 Clean up getValuatorEvents using array loop logic 95561 95562 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 95563 Reviewed-by: Jamey Sharp <jamey@minilop.net> 95564 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95565 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95566 95567commit 3d688316af0e4ff3c44acfc168bd9c76c892b629 95568Author: Adam Jackson <ajax@redhat.com> 95569Date: Tue Mar 15 11:03:26 2011 -0400 95570 95571 xfree86: warning fix 95572 95573 xf86RandR12.c: In function 'xf86RandR12EnterVT': 95574 xf86RandR12.c:1769:5: warning: ISO C90 forbids mixed declarations and code 95575 95576 Reviewed-by: Keith Packard <keithp@keithp.com> 95577 Signed-off-by: Adam Jackson <ajax@redhat.com> 95578 95579commit 2762eef8c34b48be803226f48fc71f790286f968 95580Author: Adam Jackson <ajax@redhat.com> 95581Date: Tue Mar 15 11:01:50 2011 -0400 95582 95583 xfree86: warning fix 95584 95585 Pointer.c: In function 'xf86parsePointerSection': 95586 Pointer.c:192:5: warning: format '%u' expects type 'unsigned int', but argument 3 has type 'long unsigned int' 95587 95588 Reviewed-by: Keith Packard <keithp@keithp.com> 95589 Signed-off-by: Adam Jackson <ajax@redhat.com> 95590 95591commit 5275fbe2e226e1c2a46d26288390c8ef6e1a6a52 95592Author: Adam Jackson <ajax@redhat.com> 95593Date: Tue Mar 15 10:58:36 2011 -0400 95594 95595 render: warning fixes 95596 95597 picture.c: In function 'CompositeTriStrip': 95598 picture.c:1777:25: warning: unused variable 'ps' 95599 picture.c: In function 'CompositeTriFan': 95600 picture.c:1807:16: warning: unused variable 'pScreen' 95601 95602 Reviewed-by: Keith Packard <keithp@keithp.com> 95603 Signed-off-by: Adam Jackson <ajax@redhat.com> 95604 95605commit f1c2c01ac9ea7381cb4cd21ac1b2185ff28da3ec 95606Author: Adam Jackson <ajax@redhat.com> 95607Date: Thu Mar 24 16:51:54 2011 -0400 95608 95609 dix: Small refactor to resource deletion 95610 95611 Reviewed-by: Matt Turner <mattst88@gmail.com> 95612 Signed-off-by: Adam Jackson <ajax@redhat.com> 95613 95614commit 3e0b311aba54895d806174526511ef11b31432b1 95615Author: Adam Jackson <ajax@redhat.com> 95616Date: Mon Mar 21 12:12:53 2011 -0400 95617 95618 glx: minor cast cleanup 95619 95620 Reviewed-by: Eric Anholt <eric@anholt.net> 95621 Signed-off-by: Adam Jackson <ajax@redhat.com> 95622 95623commit 501506cf9df3d81556d44d074d8321fe6c171576 95624Author: Adam Jackson <ajax@redhat.com> 95625Date: Fri Mar 18 17:48:51 2011 -0400 95626 95627 glx: dead variable removal 95628 95629 Reviewed-by: Eric Anholt <eric@anholt.net> 95630 Signed-off-by: Adam Jackson <ajax@redhat.com> 95631 95632commit 10317682e2fbb95f1d4caa1bdbbb8116272c0f86 95633Author: Adam Jackson <ajax@redhat.com> 95634Date: Fri Mar 18 17:44:11 2011 -0400 95635 95636 glx: Remove some obfuscatory macros 95637 95638 Reviewed-by: Eric Anholt <eric@anholt.net> 95639 Signed-off-by: Adam Jackson <ajax@redhat.com> 95640 95641commit c327d07cc68802256c679a65d78f98f17327abda 95642Author: Adam Jackson <ajax@redhat.com> 95643Date: Fri Mar 18 17:03:09 2011 -0400 95644 95645 glx: Remove ->forceCurrent from the context vtable 95646 95647 All the implementations of makeCurrent and forceCurrent are identical, 95648 so just use makeCurrent everywhere. 95649 95650 Reviewed-by: Eric Anholt <eric@anholt.net> 95651 Signed-off-by: Adam Jackson <ajax@redhat.com> 95652 95653commit a48dadc98a28c969741979b70b7a639f24f4cbbd 95654Author: Adam Jackson <ajax@redhat.com> 95655Date: Mon Mar 21 11:59:29 2011 -0400 95656 95657 glx: Reimplement context tags 95658 95659 This would let you do a constant-time context lookup, but if that's your 95660 performance problem you have two problems. Just use the context's XID 95661 as the tag value instead. 95662 95663 In order to do this, we have to defer destroying a context until it 95664 actually goes unreferenced, as you're allowed to mention a context tag 95665 after you've (ostensibly) destroyed the context, as long as it's still 95666 your current context. Thus, change DestroyContext to merely mark the 95667 context as dead if it's a current context, and call down to actual 95668 resource destruction (and XID reclamation) in StopUsingContext. 95669 95670 Also, stop trying to delete context state from DrawableGone. This was 95671 always broken, as GLX does not say that contexts are destroyed when 95672 their drawables are destroyed. But with the above change to defer 95673 context destruction, this would trigger a server crash on client exit as 95674 we'd free the context state twice. 95675 95676 Reviewed-by: Eric Anholt <eric@anholt.net> 95677 Signed-off-by: Adam Jackson <ajax@redhat.com> 95678 95679commit 30d6947cee714385cf62a40ea6fa6d2e68388c78 95680Author: Adam Jackson <ajax@redhat.com> 95681Date: Mon Mar 7 14:53:28 2011 -0500 95682 95683 glx: Fix _glapi_add_dispatch 95684 95685 We never need to generate stubs, because those conditions can't happen 95686 in the server. Yank that code out, but keep the bookkeeping for which 95687 extension functions are registered so the DRI driver doesn't get 95688 confused. 95689 95690 As a pleasant bonus, we're now friendlier for environments like selinux 95691 that make runtime code generation difficult, and we're portable to more 95692 arches since we don't have to port the assembly stubs. 95693 95694 Fixes the following clutter conformance tests (indirect rendering, 95695 llvmpipe driver): 95696 95697 test-cogl-backface-culling 95698 test-cogl-materials 95699 test-cogl-readpixels 95700 test-cogl-texture-mipmaps 95701 test-cogl-texture-get-set-data 95702 test-cogl-viewport 95703 test-cogl-offscreen 95704 95705 Reviewed-by: Dave Airlie <airlied@redhat.com> 95706 Signed-off-by: Adam Jackson <ajax@redhat.com> 95707 95708commit 0cb44cec7b9fbd75285f7380f162cf6140a5bf3f 95709Author: Adam Jackson <ajax@redhat.com> 95710Date: Thu Mar 3 10:42:41 2011 -0500 95711 95712 glx: Flatten -DXFree86Server 95713 95714 Always defined by the makefile, so, just get rid of it. 95715 95716 Reviewed-by: Dave Airlie <airlied@redhat.com> 95717 Signed-off-by: Adam Jackson <ajax@redhat.com> 95718 95719commit ed6c13c01e017f9f40c89247ab6652895fb1ea14 95720Author: Adam Jackson <ajax@redhat.com> 95721Date: Wed Mar 2 13:37:43 2011 -0500 95722 95723 glx: Update some glapi comments 95724 95725 Reviewed-by: Dave Airlie <airlied@redhat.com> 95726 Reviewed-by: Julien Cristau <jcristau@debian.org> 95727 Signed-off-by: Adam Jackson <ajax@redhat.com> 95728 95729commit 17d9e374721d6c8ee3f7f9cdc882f80127bdb57f 95730Author: Adam Jackson <ajax@redhat.com> 95731Date: Wed Mar 2 13:21:39 2011 -0500 95732 95733 glx: Lobotomize _glapi_get_proc_address 95734 95735 This isn't a meaningful thing in the indirect glx loader, so just warn 95736 if it ever happens and move on. 95737 95738 But also, mark it PUBLIC, so if the driver does ever call it we merely 95739 warn instead of aborting because ld.so can't find the symbol. 95740 95741 Reviewed-by: Dave Airlie <airlied@redhat.com> 95742 Signed-off-by: Adam Jackson <ajax@redhat.com> 95743 95744commit b0c665ac0fe6840dda581e4d0d0b76c703d62a7b 95745Author: Adam Jackson <ajax@redhat.com> 95746Date: Wed Mar 2 12:30:22 2011 -0500 95747 95748 glx: Remove noop dispatch table 95749 95750 We can never hit this, because the indirect GLX dispatch code always 95751 forces a current context and checks that it's non-NULL before calling 95752 into the dispatch table. If it's _not_ null, then _glapi_set_context 95753 will call into the driver, which is responsible for calling 95754 _glapi_set_dispatch to make sure the dispatch table is non-NULL. 95755 95756 Also remove _glapi_set_warning_func and friends, since we can no longer 95757 call them even from dead code. 95758 95759 Reviewed-by: Dave Airlie <airlied@redhat.com> 95760 Signed-off-by: Adam Jackson <ajax@redhat.com> 95761 95762commit 327e1d88012102af6aca6c6840aa0ed3c7041a77 95763Author: Michel Dänzer <daenzer@vmware.com> 95764Date: Mon Mar 28 17:18:39 2011 +0200 95765 95766 EXA: Use dixGetPrivate(Addr) instead of dixLookupPrivate. 95767 95768 The latter calls the former, let's cut the middle man and eliminate a branch 95769 in a hot path. 95770 95771 According to Git history, ExaSetPixmapPriv was never used anywhere, just drop 95772 it. 95773 95774 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 95775 Reviewed-by: Cyril Brulebois <kibi@debian.org> 95776 Signed-off-by: Keith Packard <keithp@keithp.com> 95777 95778commit a095a6d4e8f5090907e8d3d66018636216300846 95779Merge: a22486f84 ef9d04f8a 95780Author: Keith Packard <keithp@keithp.com> 95781Date: Sun Mar 27 20:06:29 2011 -0700 95782 95783 Merge remote-tracking branch 'airlied/pwin-cleanup' 95784 95785commit a22486f848014000dc13dda470f77d4d8ea9e9f6 95786Merge: 2ef4ff45e 633b81e8b 95787Author: Keith Packard <keithp@keithp.com> 95788Date: Sun Mar 27 18:27:10 2011 -0700 95789 95790 Merge remote-tracking branch 'whot/for-keith' 95791 95792commit 2ef4ff45ef1fcfc4967ebe3d550408769e5f6500 95793Author: Erkki Seppälä <erkki.seppala@vincit.fi> 95794Date: Fri Mar 25 10:38:23 2011 +0200 95795 95796 os/client: Prevent rare fd leak in DetermineClientPid 95797 95798 DetermineClientPid didn't close file descriptor if read on 95799 /proc/pid/cmdline failed. Adjusted the code to disregard the close 95800 return value and perform the return after that, if the read failed or 95801 returned EOF. 95802 95803 Signed-off-by: Mark Kettenis <mark.kettenis@xs4all.nl> 95804 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 95805 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 95806 Signed-off-by: Keith Packard <keithp@keithp.com> 95807 95808commit ef9d04f8ad4239ac30984769c9c8c101bb8e78bf 95809Author: Dave Airlie <airlied@redhat.com> 95810Date: Thu Mar 24 14:07:34 2011 +1000 95811 95812 dri: kill if 0 out code. 95813 95814 This refers to pWin->winSize in some #if 0 code remove it. 95815 95816 Signed-off-by: Dave Airlie <airlied@redhat.com> 95817 Reviewed-by: Cyril Brulebois <kibi@debian.org> 95818 95819commit eb9266c7176eb240a85e72beda9b6033338294f1 95820Author: Dave Airlie <airlied@redhat.com> 95821Date: Thu Mar 24 14:04:25 2011 +1000 95822 95823 consolidate SetRootClip (v2) 95824 95825 each DDX has its own copy, I've taken the darwin one, 95826 though I'm not sure why it needs the pOldClip piece that nobody 95827 else has and the commit msg is like an "Updates from magic land" 95828 type message. 95829 95830 This removes the main uses of pWin->winSize from the DDXen. 95831 95832 v2: drop old clip like ajax suggests. 95833 95834 Signed-off-by: Dave Airlie <airlied@redhat.com> 95835 Reviewed-by: Adam Jackson <ajax@redhat.com> 95836 95837commit 3c45b59e675ba926ec95842ac3d49e2091e41eb8 95838Author: Jon TURNEY <jon.turney@dronecode.org.uk> 95839Date: Fri Mar 18 15:00:14 2011 +0000 95840 95841 Fix XWin compilation after commit 769531b9 95842 95843 commit 769531b9 "Add mode field to pointer movement hooks" changes the 95844 function signature of miPointerSetPosition() to include the movement mode 95845 which resulted in the pointer position 95846 95847 Update use of miPointerSetPosition() in winEnqueueMotion() appropriately 95848 95849 (See http://tinderbox.freedesktop.org/builds/2011-03-16-0008/logs/xserver/#build) 95850 95851 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 95852 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95853 95854commit 633b81e8ba09cc6a1ea8b43f323874fda2cf0bde 95855Author: Simon Thum <simon.thum@gmx.de> 95856Date: Wed Mar 9 20:48:22 2011 +0100 95857 95858 xserver: remove AbsoluteClassRec keeping the ABI 95859 95860 This removes the struct, but keeps InitAbsoluteClassDeviceStruct as 95861 a no-op and preserves related struct layout. 95862 95863 Signed-off-by: Simon Thum <simon.thum@gmx.de> 95864 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95865 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95866 95867commit 2103d61909c7e6bd345622962df7784a19df72c5 95868Author: Simon Thum <simon.thum@gmx.de> 95869Date: Sat Mar 5 05:02:54 2011 +0100 95870 95871 simplify ChangeDeviceControl in stubs 95872 95873 Signed-off-by: Simon Thum <simon.thum@gmx.de> 95874 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95875 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95876 95877commit 118ef6f806f629a8e2ad61e4f2a43820b0fa621c 95878Author: Simon Thum <simon.thum@gmx.de> 95879Date: Sat Mar 5 05:01:22 2011 +0100 95880 95881 xf86: don't pretend to support DEVICE_ABS_* in ChangeDeviceControl 95882 95883 Signed-off-by: Simon Thum <simon.thum@gmx.de> 95884 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95885 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95886 95887commit dee83dff4b8fbf1351263beaf2d531d033b6d0c4 95888Author: Simon Thum <simon.thum@gmx.de> 95889Date: Sat Mar 5 04:59:33 2011 +0100 95890 95891 kdrive: don't pretent to support DEVICE_ABS_* in ChangeDeviceControl 95892 95893 Signed-off-by: Simon Thum <simon.thum@gmx.de> 95894 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95895 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95896 95897commit 9eaecb1bfe358a0c5f24645a29d6c377c0c4b37c 95898Author: Simon Thum <simon.thum@gmx.de> 95899Date: Sun Mar 20 16:28:22 2011 +0100 95900 95901 xquartz: simplify ChangeDeviceControl 95902 95903 Signed-off-by: Simon Thum <simon.thum@gmx.de> 95904 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95905 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95906 95907commit 33e257ab43ebf35e838b80c416762e9b743e9385 95908Author: Peter Hutterer <peter.hutterer@who-t.net> 95909Date: Tue Mar 22 11:53:00 2011 +1000 95910 95911 test: when unit tests are enabled, build them during "make" 95912 95913 Catch compiler errors that were otherwise only spotted on make check. 95914 95915 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95916 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 95917 95918commit 6243332782f2d286d49f1dec8a628b0a403665fd 95919Author: Erkki Seppälä <erkki.seppala@vincit.fi> 95920Date: Fri Mar 18 16:35:36 2011 +0200 95921 95922 config: handle device change event properly 95923 95924 wakeup_handler in udev.c wasn't dealing with udev change events. 95925 There are situations when a device can gain its input capabilities 95926 after it has been added to the system and therefore the change events 95927 must be handled as well. 95928 95929 The change is handled as a consecutive device removal and addition. 95930 95931 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 95932 Signed-off-by: Stefan Kost <Stefan.Kost@nokia.com> 95933 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 95934 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95935 95936commit fad10cb38ef2433c82f8a43d3b7d64e323728060 95937Author: Peter Hutterer <peter.hutterer@who-t.net> 95938Date: Fri Mar 18 11:06:57 2011 +1000 95939 95940 xfree86: print out which driver is about to be used. 95941 95942 Makes reading the log file a lot easier for those that don't magically 95943 recognise the log spew by the individual drivers. 95944 95945 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 95946 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 95947 95948commit cb5d4b416a0fc850a1c119755a9cc3c4f0d7a545 95949Author: Jeremy Huddleston <jeremyhu@apple.com> 95950Date: Wed Mar 23 12:10:15 2011 -0700 95951 95952 XQuartz: applewm: Don't check if requested window level is < 0 because it is unsigned (-Wtautological-compare) 95953 95954 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 95955 95956commit c116e32dc2799c4abcf38d22bc773b755387c89a 95957Author: Jeremy Huddleston <jeremyhu@apple.com> 95958Date: Wed Mar 23 12:07:36 2011 -0700 95959 95960 XQuartz: Put ifdef guards around have_depth label (-Wunused-label) 95961 95962 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 95963 95964commit 1a583f7940e18a8ef28ca14e74e85e0785d39067 95965Author: Jeremy Huddleston <jeremyhu@apple.com> 95966Date: Wed Mar 23 12:05:05 2011 -0700 95967 95968 XQuartz: Properly comment extra tokens (-Wextra-tokens) after endif 95969 95970 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 95971 95972commit c55baebf4ebf1887262cc16899eb297b9f284f6e 95973Author: Jeremy Huddleston <jeremyhu@apple.com> 95974Date: Tue Mar 22 19:01:48 2011 -0700 95975 95976 GLX: Support TLS with better portability 95977 95978 AX_TLS detects when toolchains support __thread or __declspec(thread), 95979 but existing code assumed __thread. 95980 95981 This also adds a check to configure.ac to error out if TLS is requested 95982 but unsupported. 95983 95984 Found-by: Tinderbox 95985 http://tinderbox.x.org/builds/2011-03-22-0007 95986 95987 Regression-from: 82b1eaa6cad20f39dbf15573bdb3d62acbcd91f9 95988 95989 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 95990 Reviewed-by: Tom Fogal <tfogal@alumni.unh.edu> 95991 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 95992 95993commit 03f45df93469f6aef391e97007b9614e0770cc4c 95994Merge: efcb7275c 5fb329a04 95995Author: Keith Packard <keithp@keithp.com> 95996Date: Wed Mar 23 13:38:37 2011 +0900 95997 95998 Merge remote-tracking branch 'airlied/xinerama-cleanup' 95999 96000commit efcb7275ce5de651f91ba4ff8bb227dfb68bb154 96001Author: Gaetan Nadon <memsize@videotron.ca> 96002Date: Thu Mar 17 19:26:37 2011 -0400 96003 96004 test: git ignore the list test executable 96005 96006 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96007 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 96008 Signed-off-by: Keith Packard <keithp@keithp.com> 96009 96010commit 6a5bf15fa99cf5b2358b3b3e2f29e5044aa8724a 96011Author: Gaetan Nadon <memsize@videotron.ca> 96012Date: Thu Mar 17 19:26:36 2011 -0400 96013 96014 Add generalized unit test support using util-macros. 96015 96016 A handful of modules have begun adding unit test programs. 96017 These macros will help providing a consistent interface which will 96018 help package builders and developers to manage the functionality. 96019 96020 XORG_ENABLE_UNIT_TESTS will turn on/off unit testing, regardless 96021 of how it is implemented. The default (yes/no) can be specified by each 96022 module. It can be used by itself if glib or -wrap support is not needed. 96023 96024 XORG_WITH_GLIB will probe the system for glib-2.0. A different version 96025 can be specified in each module. It will consult XORG_ENABLE_UNIT_TESTS 96026 but can be used by itself in contexts other then unit testing. 96027 The default (yes/no) can be specified by each module. 96028 96029 XORG_LD_WRAP will probe the linker for -wrap support. It will consult 96030 XORG_ENABLE_UNIT_TESTS but can be used by itself in contexts 96031 other then unit testing. 96032 96033 configure options: 96034 --enable-unit-tests Enable building unit test cases (default: auto) 96035 --with-glib Use GLib library for unit testing (default: auto) 96036 96037 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96038 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 96039 Signed-off-by: Keith Packard <keithp@keithp.com> 96040 96041commit ee583cb33423fa79beb22db20b30e10a677f9b5a 96042Author: Gaetan Nadon <memsize@videotron.ca> 96043Date: Thu Mar 17 19:26:35 2011 -0400 96044 96045 config: group document related XORG_ macros together 96046 96047 No functional changes. 96048 96049 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96050 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 96051 Signed-off-by: Keith Packard <keithp@keithp.com> 96052 96053commit d5b16b037b8fe12ba85c68c8289b6a8cc5e3a09d 96054Author: Keith Packard <keithp@keithp.com> 96055Date: Thu Mar 17 23:48:52 2011 -0700 96056 96057 Revert "dix: Remove usage_hint from pixmaps, store it in ->drawable.class" 96058 96059 This reverts commit 1564c82417d201de5b9a5ec5e7aa4ef14c45fbad. 96060 96061 The drivers used the top bits of the usage_hint to store driver 96062 private flags (intel, radeon, nouveau). 96063 96064 With EXA we need to get at this data so if we migrate the pixmap we 96065 can create the correct type of pixmap in the driver, however this 96066 commit truncates the usage_hint into 8-bit class and loses all the 96067 good stuff. 96068 96069 Signed-off-by: Dave Airlie <airlied@gmail.com> 96070 Reviewed-by: Keith Packard <keithp@keithp.com> 96071 96072commit dc9ce695a69ca0787f58f8d160212a7a41acb703 96073Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 96074Date: Wed Mar 9 15:45:40 2011 +0200 96075 96076 xkb: Initialize pad bytes sent in replies of geometry requests. 96077 96078 Valgrind complains about uninitialized data being written to clients. 96079 96080 Reviewed-by: Erkki Seppälä <erkki.seppala@vincit.fi> 96081 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 96082 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96083 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96084 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96085 96086commit ee3a4951a49312a2b68025db00c71472eb84293f 96087Author: Peter Hutterer <peter.hutterer@who-t.net> 96088Date: Thu Mar 10 09:30:02 2011 +1000 96089 96090 xkb: Document XkbWriteCountedString. 96091 96092 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96093 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96094 96095commit 5c47f8beac7f87680c6f7331483b9cf94a1dbc86 96096Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 96097Date: Fri Mar 11 14:30:49 2011 +0200 96098 96099 xkb: Release XKB component names when compiling keymap. 96100 96101 Reviewed-by: Erkki Seppälä <erkki.seppala@vincit.fi> 96102 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 96103 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96104 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96105 96106commit 6f46ae3c69c48f389560af1491ed1729ba45866f 96107Author: Peter Hutterer <peter.hutterer@who-t.net> 96108Date: Tue Mar 15 13:20:03 2011 +1000 96109 96110 man: list the drivers that are ignored when hotplugging (#35209) 96111 96112 X.Org Bug 35209 <http://bugs.freedesktop.org/show_bug.cgi?id=35209> 96113 96114 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96115 Reviewed-by: Cyril Brulebois <kibi@debian.org> 96116 Reviewed-by: Julien Cristau <jcristau@debian.org> 96117 96118commit 642569591a73806d038f00f2f9a5c8738f764c0b 96119Author: Gaetan Nadon <memsize@videotron.ca> 96120Date: Sat Mar 12 07:29:35 2011 -0500 96121 96122 dmx/doc: remove dead code in the makefile 96123 96124 This was leftover from some older ways of building dmx/scale docbook. 96125 96126 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 96127 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 96128 Signed-off-by: Keith Packard <keithp@keithp.com> 96129 96130commit 21f70cad19474a6f3ba419e03c0df12881ed1092 96131Author: Gaetan Nadon <memsize@videotron.ca> 96132Date: Sat Mar 12 07:29:34 2011 -0500 96133 96134 man: relocate manual pages in the man subdir outside doc 96135 96136 The convention is to have the manual pages in a man subdir 96137 which is not under a doc dir. The doc dir contains users docs. 96138 This will move man pages out of the way for upcoming DocBook patches. 96139 96140 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 96141 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 96142 Signed-off-by: Keith Packard <keithp@keithp.com> 96143 96144commit d3adf2d9350bee4125107e2ea1ed0c51bb736562 96145Author: Erkki Seppälä <erkki.seppala@vincit.fi> 96146Date: Thu Mar 10 11:40:40 2011 +0200 96147 96148 xfree86/modes: Fixed memory leak in xf86InitialConfiguration 96149 96150 There were two memory leaks in the function: one was the lack of free 96151 for "enabled", the other was the full lack of releasing anything when 96152 configuration was too small. The first issue was fixed by adding the 96153 missing free, the other was addressed by replacing the duplicate 96154 memory releasing sequences with one that is gotoed into. 96155 96156 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96157 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 96158 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 96159 Signed-off-by: Keith Packard <keithp@keithp.com> 96160 96161commit a713832cbe494ff468a627e88b11164074bbd14e 96162Author: Jon TURNEY <jon.turney@dronecode.org.uk> 96163Date: Sun Mar 13 20:01:30 2011 +0000 96164 96165 When XDMCP -from is specified, only register the requested address 96166 96167 When XDMCP -from is specified, only register the requested address, 96168 rather than the requested address, and any others we have of different 96169 address families to the requested address. 96170 96171 e.g. if we have 4 interfaces with both IPv4 and IPv6 addresses (which 96172 are not IPv6 mapped IPV4 addresses), using -from with one of those IPv4 96173 addresses currently means only that IPv4 address, and all IPv6 addresses 96174 are used in the connection data in XDMCP REQUEST packet. 96175 96176 (See http://cygwin.com/ml/cygwin-xfree/2011-02/msg00000.html) 96177 96178 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 96179 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 96180 Signed-off-by: Keith Packard <keithp@keithp.com> 96181 96182commit 0ac4931753a5d5925fc844c8cbec08585aea57a7 96183Merge: a8146f6be d7f801141 96184Author: Keith Packard <keithp@keithp.com> 96185Date: Mon Mar 14 13:06:41 2011 -0700 96186 96187 Merge remote-tracking branch 'ajax/xserver-next' 96188 96189commit d7f8011418f9da06631f27c66c29bcb226d0dffe 96190Author: Søren Sandmann Pedersen <ssp@redhat.com> 96191Date: Tue Mar 8 10:14:28 2011 -0500 96192 96193 Remove TriStrip and TriFan from the picture screen 96194 96195 These functions no longer go through the screen vtable, so remove 96196 them and fix up the various wrappers. 96197 96198 Reviewed-by: Adam Jackson <ajax@redhat.com> 96199 Acked-by: Keith Packard <keithp@keithp.com> 96200 Signed-off-by: Soren Sandmann <ssp@redhat.com> 96201 96202commit 0eb5b0fbcf1233a93f285ff1e1609fcbd01e7c79 96203Author: Søren Sandmann Pedersen <ssp@redhat.com> 96204Date: Tue Mar 8 10:14:27 2011 -0500 96205 96206 Absorb miTriStrip() into CompositeTriStrip() 96207 96208 There is no need to virtualize this function that nobody cares about. 96209 96210 Reviewed-by: Adam Jackson <ajax@redhat.com> 96211 Acked-by: Keith Packard <keithp@keithp.com> 96212 Signed-off-by: Soren Sandmann <ssp@redhat.com> 96213 96214commit c2af0cea02bd85f4d5954c16e34b4a8fb0fe2243 96215Author: Søren Sandmann Pedersen <ssp@redhat.com> 96216Date: Tue Mar 8 10:14:26 2011 -0500 96217 96218 Absorb miTriFan() into CompositeTriFan() 96219 96220 There is no need to virtualize this function that nobody cares about. 96221 96222 Reviewed-by: Adam Jackson <ajax@redhat.com> 96223 Acked-by: Keith Packard <keithp@keithp.com> 96224 Signed-off-by: Soren Sandmann <ssp@redhat.com> 96225 96226commit a8146f6becc44bf9ad611d33bded17df07e6af21 96227Author: Erkki Seppälä <erkki.seppala@vincit.fi> 96228Date: Wed Mar 9 17:29:14 2011 +0200 96229 96230 mi/misprite: use memory management provided by dixRegisterPrivateKey 96231 96232 The record allocated by miSpriteDeviceCursorInitialize was not being 96233 released. 96234 96235 This patch makes misprite use dixRegisterPrivateKey with the record 96236 size argument, which handles the memory management 96237 issues. miSpriteDeviceCursorInitialize is restructured to initialize 96238 pCursorInfo only if miDCDeviceInitialize succeeds. The record itself 96239 is zeroed on cleanup to ensure that the assumptions in the code still 96240 hold. 96241 96242 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 96243 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 96244 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96245 Signed-off-by: Keith Packard <keithp@keithp.com> 96246 96247commit f985a7319ef80b9b613eeaf24581000827cb220f 96248Author: Erkki Seppälä <erkki.seppala@vincit.fi> 96249Date: Wed Mar 9 17:29:14 2011 +0200 96250 96251 mi/misprite: use memory management provided by dixRegisterPrivateKey 96252 96253 The record allocated by miSpriteDeviceCursorInitialize was not being 96254 released. 96255 96256 This patch makes misprite use dixRegisterPrivateKey with the record 96257 size argument, which handles the memory management issues. 96258 miSpriteDeviceCursorInitialize is restructured to initialize pCursorInfo 96259 only if miDCDeviceInitialize succeeds. The record itself is zeroed on 96260 cleanup to ensure that the assumptions in the code still hold. 96261 96262 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96263 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 96264 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 96265 96266commit 57b35adaed112520c3b3b2fbad13cf5a91cd6652 96267Author: Søren Sandmann Pedersen <ssp@redhat.com> 96268Date: Wed Mar 9 13:57:25 2011 -0500 96269 96270 Remove geometry arguments from miSourceValidate() 96271 96272 The only user of the geometry coordinates is the software sprite code, 96273 which uses them to remove the pointer whenever the window beneath is 96274 being used as a source. However, using Window pictures as a source is 96275 extremely rare (let alone *partial* windows), so there is no harm done 96276 in just validating all of the drawable. 96277 96278 Additionally, the miSourceValidate() function was buggy in at least 96279 three respects: 96280 96281 (a) It added drawable->{x,y} before calling down, which is wrong since 96282 the misprite code already adds them in its check. (Alternatively, 96283 the misprite code is wrong, but there are actual users who would 96284 notice if that code was broken). 96285 96286 (b) It didn't account for the width of the interpolation filter, so if 96287 the Picture had a bilinear or convolution filter, the edges 96288 surrounding the source area would not be validated. 96289 96290 (c) It didn't validate alpha maps. 96291 96292 Finally, computing the bounding box of the transform on every 96293 composite request was a real performance issue in pixman, so 96294 presumably it could be one here as well. 96295 96296 This patch changes miSourceValidate() to simply validate all of the 96297 underlying drawable. 96298 96299 Reviewed-by: Adam Jackson <ajax@redhat.com> 96300 Reviewed-by: Keith Packard <keithp@keithp.com> 96301 Signed-off-by: Soren Sandmann <ssp@redhat.com> 96302 96303commit 016edc17512ba966d60edede8cf947996bae0b3c 96304Author: Adam Jackson <ajax@redhat.com> 96305Date: Mon Feb 28 17:12:26 2011 -0500 96306 96307 dix: Define RESTYPE as uint32_t 96308 96309 long is needlessly long on LP64. 96310 96311 Reviewed-by: Dave Airlie <airlied@redhat.com> 96312 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96313 Signed-off-by: Adam Jackson <ajax@redhat.com> 96314 96315commit 1f2bc777f96fd41feb55a4799ece939652130ef4 96316Author: Adam Jackson <ajax@redhat.com> 96317Date: Mon Feb 28 13:11:12 2011 -0500 96318 96319 dix: Shrink PropertyRec on LP64 96320 96321 size needn't be a long. No change on ILP32 but, combined with the 96322 previous change, 56 -> 40 bytes on LP64. 96323 96324 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96325 Signed-off-by: Adam Jackson <ajax@redhat.com> 96326 96327commit 51f353d0a0d116af16d7d9590cadef6c56328746 96328Author: Adam Jackson <ajax@redhat.com> 96329Date: Mon Feb 28 13:10:20 2011 -0500 96330 96331 dix: Fix ATOM typedef 96332 96333 unsigned long is needlessly large on LP64. Use uint32_t instead. 96334 96335 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96336 Signed-off-by: Adam Jackson <ajax@redhat.com> 96337 96338commit 7ca75abbbdd2a1211e52a4f43ac4ed24d3c8ab34 96339Author: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 96340Date: Wed Mar 9 11:17:28 2011 +1100 96341 96342 glx: Use PATH_MAX as size of filename buffer 96343 96344 Reviewed-by: Adam Jackson <ajax@redhat.com> 96345 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 96346 96347commit d17a9fb8414becf6a8998041df68f209f9222b2b 96348Author: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 96349Date: Wed Mar 9 11:17:27 2011 +1100 96350 96351 Consolidate all the PATH_MAX handling into misc.h 96352 96353 Reviewed-by: Adam Jackson <ajax@redhat.com> 96354 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 96355 96356commit 021393d1b8bcc9ff2ff5deb2306360e6b0afa1c6 96357Author: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 96358Date: Wed Mar 9 11:15:07 2011 +1100 96359 96360 glx: Factor out glxProbeDriver function. 96361 96362 DRI, DRI2 and swrast all had near-identical driver probing logic. 96363 Pull it into glxdricommon. 96364 96365 [ajax: warning fix] 96366 Reviewed-by: Adam Jackson <ajax@redhat.com> 96367 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 96368 96369commit 56c90e29f04727c903bd0f084d23bf44eb1a0a11 96370Author: Adam Jackson <ajax@redhat.com> 96371Date: Mon Nov 15 14:29:14 2010 -0500 96372 96373 randr: Add RRConstrainCursorHarder 96374 96375 Confine cursor motion to within the bounds of a single CRTC, iff all the 96376 CRTCs within a ScreenRec are reachable from each other. If not you get 96377 the same "cursor floats within the bounding rect" behaviour you get now. 96378 96379 v3: 96380 - Incorporate review feedback from Christopher James Halse Rogers 96381 v4: 96382 - Add mode field. 96383 96384 Signed-off-by: Adam Jackson <ajax@redhat.com> 96385 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96386 96387commit 810fbfa44626bff9f443ab17c0ad27ff7ae121d7 96388Author: Adam Jackson <ajax@redhat.com> 96389Date: Wed Feb 9 17:32:16 2011 -0500 96390 96391 mi: Call pScreen->ConstrainCursorHarder from the position update path 96392 96393 v2: Cover more paths, spotted by Daniel Stone. 96394 v3: pass down the mode field for movement mode. 96395 96396 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96397 Signed-off-by: Adam Jackson <ajax@redhat.com> 96398 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96399 96400commit 769531b9ccade723a56498b0888af58d085fec9e 96401Author: Peter Hutterer <peter.hutterer@who-t.net> 96402Date: Fri Feb 18 14:19:18 2011 +1000 96403 96404 Add mode field to pointer movement hooks. 96405 96406 Preparation work for pointer barriers. 96407 96408 Reviewed-by: Adam Jackson <ajax@redhat.com> 96409 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96410 96411commit 92788e677be79bd04e5ef140f4ced50ad8b1bf8e 96412Author: Peter Hutterer <peter.hutterer@who-t.net> 96413Date: Tue Feb 22 12:32:01 2011 +1000 96414 96415 test: add some tests for basic list manipulation. 96416 96417 This has less purpose as a test but more as documentation on how to actually 96418 use the differnent list calls. 96419 96420 Reviewed-by: Adam Jackson <ajax@redhat.com> 96421 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96422 96423commit 1ad9f01c31742157934a791e6141d10520d13e8a 96424Author: Adam Jackson <ajax@redhat.com> 96425Date: Thu Mar 10 10:47:40 2011 -0500 96426 96427 glx: Add texbuffer2 support to swrast 96428 96429 Reviewed-by: Julien Cristau <jcristau@debian.org> 96430 Signed-off-by: Adam Jackson <ajax@redhat.com> 96431 96432commit 5fb329a04a18835ce864d0563f6dfeb3d3c78d69 96433Author: Dave Airlie <airlied@redhat.com> 96434Date: Wed Mar 9 15:05:26 2011 +1000 96435 96436 panoramiX: convert 1->panoramiXNumScreens loops to use macro (v2) 96437 96438 This converts all the remaining 1->num loops to the macro, 96439 this removes nearly all the panoramiXNumScreens usage in 96440 loops, and is a step to replacing it. 96441 96442 v2: move some from the other patch. 96443 96444 Signed-off-by: Dave Airlie <airlied@redhat.com> 96445 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96446 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 96447 96448commit 0bccfcc97b2300b83aa4693454c27ac87f23f221 96449Author: Dave Airlie <airlied@redhat.com> 96450Date: Wed Mar 9 15:01:20 2011 +1000 96451 96452 panoramiX: convert 0->panoramiXNumScreens loops to macro (v3) 96453 96454 This just uses the FOR_NSCREENS macro instead. 96455 96456 v2: remove some of the 1->x loops. 96457 v3: drop the 1->0 loop, will rework later. 96458 96459 Signed-off-by: Dave Airlie <airlied@redhat.com> 96460 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96461 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 96462 96463commit 5a94934487ea477947e24dcd4720b7cde77d3d2f 96464Author: Dave Airlie <airlied@redhat.com> 96465Date: Wed Mar 9 14:53:26 2011 +1000 96466 96467 panoramiX: consolidate common id assignment code. 96468 96469 This adds a new FOR_NSCREENS_FORWARD_SKIP, which skips the first 96470 element and is a common idiom throughout panoramiX code. 96471 96472 It then adds a new inline function to hide id assignment to a 96473 panoramiX resource and cleans up lots of common repeated code. 96474 96475 Signed-off-by: Dave Airlie <airlied@redhat.com> 96476 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96477 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 96478 96479commit e2f8315daefc6e56ac8395a9f58f87bbf0809039 96480Author: Dave Airlie <airlied@redhat.com> 96481Date: Thu Mar 10 09:13:50 2011 +1000 96482 96483 drop exa 96484 96485 don't provide accel in this framework 96486 96487commit c48c530e541e201ea1afe2512d85967659ceef84 96488Author: Dave Airlie <airlied@redhat.com> 96489Date: Wed Mar 9 15:11:42 2011 +1000 96490 96491 panoramiX: macro checking if drawable is root (v2) 96492 96493 this code appears in quite a few places, consolidate it into 96494 a macro in a header. 96495 96496 v2: align braces with macro just above it, and with 96497 lines removed 96498 96499 Signed-off-by: Dave Airlie <airlied@redhat.com> 96500 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96501 96502commit c3c0e2fdd326214e271ce144e8fc2725cbd738ef 96503Merge: a19771e43 73555555a 96504Author: Keith Packard <keithp@keithp.com> 96505Date: Wed Mar 9 14:25:54 2011 -0800 96506 96507 Merge remote branch 'whot/for-keith' 96508 96509commit 73555555a440855f9ae64c3367c5c7dca98c8741 96510Author: Adam Jackson <ajax@redhat.com> 96511Date: Tue Mar 8 15:33:13 2011 -0500 96512 96513 record: warning fix 96514 96515 record.c:810:9: warning: unused variable 'count' 96516 96517 Scope-shadowed by a later variable of the same name, safe to just 96518 delete. 96519 96520 Signed-off-by: Adam Jackson <ajax@redhat.com> 96521 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96522 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96523 96524commit 4d114cc5467a514faa437ce7f4c5e772e2f6a21d 96525Author: Adam Jackson <ajax@redhat.com> 96526Date: Tue Mar 8 15:33:12 2011 -0500 96527 96528 input: warning fixes 96529 96530 eventconvert.c:287:9: warning: enumeration value 'ET_Enter' not handled in switch 96531 eventconvert.c:287:9: warning: enumeration value 'ET_Leave' not handled in switch 96532 eventconvert.c:287:9: warning: enumeration value 'ET_FocusIn' not handled in switch 96533 eventconvert.c:287:9: warning: enumeration value 'ET_FocusOut' not handled in switch 96534 eventconvert.c:287:9: warning: enumeration value 'ET_DeviceChanged' not handled in switch 96535 eventconvert.c:287:9: warning: enumeration value 'ET_Hierarchy' not handled in switch 96536 eventconvert.c:287:9: warning: enumeration value 'ET_DGAEvent' not handled in switch 96537 eventconvert.c:287:9: warning: enumeration value 'ET_RawKeyPress' not handled in switch 96538 eventconvert.c:287:9: warning: enumeration value 'ET_RawKeyRelease' not handled in switch 96539 eventconvert.c:287:9: warning: enumeration value 'ET_RawButtonPress' not handled in switch 96540 eventconvert.c:287:9: warning: enumeration value 'ET_RawButtonRelease' not handled in switch 96541 eventconvert.c:287:9: warning: enumeration value 'ET_RawMotion' not handled in switch 96542 eventconvert.c:287:9: warning: enumeration value 'ET_XQuartz' not handled in switch 96543 eventconvert.c:287:9: warning: enumeration value 'ET_Internal' not handled in switch 96544 96545 From the code it appears these are can't happens, so if they ever do, 96546 BadImplementation seems entirely appropriate. 96547 96548 Signed-off-by: Adam Jackson <ajax@redhat.com> 96549 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96550 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96551 96552commit 7b5e562ea74039832116ee13db910f290f074782 96553Author: Adam Jackson <ajax@redhat.com> 96554Date: Tue Mar 8 15:33:11 2011 -0500 96555 96556 input: warning fix 96557 96558 getevents.c:770:5: warning: suggest parentheses around '&&' within '||' 96559 96560 Introduced with dc57f89959e549403f8488eb9f23425bd7118b22: 96561 96562 - if(dev->u.master && dev->valuator) { 96563 + if(dev->valuator && IsMaster(dev) || !IsFloating(dev)) { 96564 96565 So I'm assuming the two terms around the || are meant to be a unit. 96566 96567 Signed-off-by: Adam Jackson <ajax@redhat.com> 96568 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96569 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96570 96571commit 33fee13361e745e1db29e250b08622c83046d488 96572Author: Peter Hutterer <peter.hutterer@who-t.net> 96573Date: Tue Mar 8 14:41:21 2011 +1000 96574 96575 Xi: fix XI2 passive grab reply length calculation 96576 96577 If modifiers failed, the reply length was 4 bytes too short. 96578 96579 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96580 Reviewed-by: Julien Cristau <jcristau@debian.org> 96581 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96582 96583commit eb8141b6edd8b477c0ba796be71e985c35520a9b 96584Author: Julien Cristau <jcristau@debian.org> 96585Date: Mon Mar 7 18:55:19 2011 +0100 96586 96587 Xi: fix length checks for swapped clients 96588 96589 ChangeDeviceProperty and XIChangeProperty are followed by some data, so 96590 use REQUEST_AT_LEAST_SIZE instead of REQUEST_SIZE_MATCH. 96591 96592 X.Org bug#35082 <https://bugs.freedesktop.org/show_bug.cgi?id=35082> 96593 96594 Reported-by: Markus Fleschutz <markus.fleschutz@x-software.com> 96595 Signed-off-by: Julien Cristau <jcristau@debian.org> 96596 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96597 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96598 96599commit 4114533db6704324fc26f28a444415e325ace8e0 96600Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 96601Date: Fri Mar 4 17:55:33 2011 +0200 96602 96603 config: Ensure that stolen option list elements are released. 96604 96605 NewInputDeviceRequest steals the contents of option list elements but 96606 doesn't use the elements themselves for anything. Therefore the list 96607 elements need to be released always. 96608 96609 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 96610 Reviewed-by: Erkki Seppälä <erkki.seppala@vincit.fi> 96611 Reviewed-by: Adam Jackson <ajax@redhat.com> 96612 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96613 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96614 96615commit 8d30aff4aa708b9b885d492602ced7493a96a4df 96616Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 96617Date: Fri Mar 4 17:55:32 2011 +0200 96618 96619 dix: Release input device config info when the device disconnects. 96620 96621 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 96622 Reviewed-by: Erkki Seppälä <erkki.seppala@vincit.fi> 96623 Reviewed-by: Adam Jackson <ajax@redhat.com> 96624 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96625 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96626 96627commit 40e56d34538f4663426db50893c231a2b5d760dc 96628Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 96629Date: Fri Mar 4 17:55:31 2011 +0200 96630 96631 xkb: Ensure that XKB device private won't leak on device disconnect. 96632 96633 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 96634 Reviewed-by: Erkki Seppälä <erkki.seppala@vincit.fi> 96635 Reviewed-by: Adam Jackson <ajax@redhat.com> 96636 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96637 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96638 96639commit 18413f55089623123537c1499b02aa95ca2014d2 96640Author: Peter Hutterer <peter.hutterer@who-t.net> 96641Date: Thu Mar 3 14:15:55 2011 +1000 96642 96643 xfree86: block signals between EnableDevice and first CheckMotion() 96644 96645 Devices usually enable SIGIO processing in EnableDevice. CheckMotion 96646 initialises the pointer sprite, sends Enter/Leave events, etc. This leaves 96647 us with a small window where events may be processed without the sprite or 96648 pointer position (as seen from the protocol) is valid. 96649 Block signals during this window. 96650 96651 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96652 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96653 96654commit a19771e4337d1c4600550314bbc42a1495a023ff 96655Author: Erkki Seppälä <erkki.seppala@vincit.fi> 96656Date: Tue Mar 8 13:29:41 2011 -0500 96657 96658 xfree86/common: Remove a configScreen leak when conf_screen is NULL 96659 96660 configScreen used a dynamically allocated buffer for XF86ConfScreenRec 96661 when conf_screen argument was NULL. This pointer was never stored 96662 anywhere, nor was it released, so this patch makes the function use 96663 automatically allocated storage in that situation. 96664 96665 [ajax: minor grammar fix] 96666 Reviewed-by: Adam Jackson <ajax@redhat.com> 96667 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 96668 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 96669 96670commit b73ebe65cd071a063417ac9c38513a6742e32428 96671Author: Adam Jackson <ajax@redhat.com> 96672Date: Wed Mar 2 12:12:35 2011 -0500 96673 96674 glx: Remove unused _glapi_get_proc_name 96675 96676 Reviewed-by: Julien Cristau <jcristau@debian.org> 96677 Signed-off-by: Adam Jackson <ajax@redhat.com> 96678 96679commit 4377288abdbaac8506d8643b440c300ba03182fe 96680Author: Adam Jackson <ajax@redhat.com> 96681Date: Wed Mar 2 11:52:27 2011 -0500 96682 96683 glx: Simplify _glapi_check_multithread 96684 96685 We can guarantee that the X server does not call DRI driver services 96686 from multiple threads, so _glapi_check_multithread need never do 96687 anything special. As a result, ThreadSafe is always false, so remove it 96688 and simplify expressions where it appeared to match. 96689 96690 Reviewed-by: Julien Cristau <jcristau@debian.org> 96691 Signed-off-by: Adam Jackson <ajax@redhat.com> 96692 96693commit d5e8bb2150ce9d8a4132d61f76621a2c7a7f0acf 96694Author: Adam Jackson <ajax@redhat.com> 96695Date: Wed Mar 2 11:26:05 2011 -0500 96696 96697 glx: unifdef BEOS_THREADS 96698 96699 We don't support BeOS, so, no. 96700 96701 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 96702 Signed-off-by: Adam Jackson <ajax@redhat.com> 96703 96704commit 3692ff0b35e774a6f6cbedf634d53c4df2e53a24 96705Author: Adam Jackson <ajax@redhat.com> 96706Date: Wed Mar 2 11:22:45 2011 -0500 96707 96708 glx: unifdef USE_XTHREADS 96709 96710 xthreads is an ancient wrapper around a half dozen or so common thread 96711 APIs, including pthreads. Just use pthreads directly if you have it; if 96712 you don't, get with the times. 96713 96714 Reviewed-by: Julien Cristau <jcristau@debian.org> 96715 Signed-off-by: Adam Jackson <ajax@redhat.com> 96716 96717commit 2371b44f9eb9181ec2c081ce66befee51d18bf6e 96718Author: Adam Jackson <ajax@redhat.com> 96719Date: Wed Mar 2 11:17:07 2011 -0500 96720 96721 glx: unifdef SOLARIS_THREADS 96722 96723 Solaris 7 and older are not supported; Solaris 8 and later have (and 96724 use) pthreads. 96725 96726 Acked-by: Alan Coopersmith <alan.coopersmith@oracle.com> 96727 Signed-off-by: Adam Jackson <ajax@redhat.com> 96728 96729commit a5fdd1aa06c7c8ecefeafafc4a778fc816ad41a4 96730Author: Adam Jackson <ajax@redhat.com> 96731Date: Wed Mar 2 10:20:02 2011 -0500 96732 96733 glx: Remove unused _glapi_get_proc_offset and friends 96734 96735 Reviewed-by: Julien Cristau <jcristau@debian.org> 96736 Signed-off-by: Adam Jackson <ajax@redhat.com> 96737 96738commit 2cb3dc020cbec1ba64848dac532b3bb2d84d0635 96739Author: Adam Jackson <ajax@redhat.com> 96740Date: Mon Feb 28 17:11:35 2011 -0500 96741 96742 xi: Use RESTYPE consistently 96743 96744 No functional change 96745 96746 Reviewed-by: Soren Sandmann <ssp@redhat.com> 96747 Signed-off-by: Adam Jackson <ajax@redhat.com> 96748 96749commit 2413702aa0b488f1a4f828a017d5b508c4f219f1 96750Author: Adam Jackson <ajax@redhat.com> 96751Date: Mon Feb 28 16:54:17 2011 -0500 96752 96753 xvmc: Use RESTYPE consistently 96754 96755 Also mark these static since they're not used externally. 96756 96757 Reviewed-by: Soren Sandmann <ssp@redhat.com> 96758 Signed-off-by: Adam Jackson <ajax@redhat.com> 96759 96760commit cd58924d294de2abddbd2f03c2db021894fe9759 96761Author: Adam Jackson <ajax@redhat.com> 96762Date: Mon Feb 28 16:28:26 2011 -0500 96763 96764 xinerama: Use RESTYPE consistently 96765 96766 No functional change 96767 96768 Reviewed-by: Soren Sandmann <ssp@redhat.com> 96769 Signed-off-by: Adam Jackson <ajax@redhat.com> 96770 96771commit 9a6a9971b03381b7c44257284edecdfa0135957f 96772Author: Adam Jackson <ajax@redhat.com> 96773Date: Mon Feb 28 16:25:23 2011 -0500 96774 96775 xv: Use RESTYPE where appropriate 96776 96777 No functional change. 96778 96779 Reviewed-by: Soren Sandmann <ssp@redhat.com> 96780 Signed-off-by: Adam Jackson <ajax@redhat.com> 96781 96782commit fddfd026a0e4a8190dc01fb0f38af4b2c29b63a3 96783Author: Adam Jackson <ajax@redhat.com> 96784Date: Mon Feb 28 13:01:15 2011 -0500 96785 96786 dix: Remove PIXEL typedef 96787 96788 Doesn't appear to be used anywhere. 96789 96790 Reviewed-by: Soren Sandmann <ssp@redhat.com> 96791 Signed-off-by: Adam Jackson <ajax@redhat.com> 96792 96793commit 82a8677d9175732a61df4116a396b76a7704efb4 96794Author: Adam Jackson <ajax@redhat.com> 96795Date: Thu Jan 20 22:49:32 2011 -0500 96796 96797 dix: Remove deprecated.c 96798 96799 No more internal users, this can be dropped now. 96800 96801 Reviewed-by: Soren Sandmann <ssp@redhat.com> 96802 Signed-off-by: Adam Jackson <ajax@redhat.com> 96803 96804commit 1564c82417d201de5b9a5ec5e7aa4ef14c45fbad 96805Author: Adam Jackson <ajax@redhat.com> 96806Date: Thu Jan 20 18:37:45 2011 -0500 96807 96808 dix: Remove usage_hint from pixmaps, store it in ->drawable.class 96809 96810 The class field was unused for pixmaps, and we don't have enough classes 96811 to justify a whole uint32 anyway. 96812 96813 Reviewed-by: Soren Sandmann <ssp@redhat.com> 96814 Signed-off-by: Adam Jackson <ajax@redhat.com> 96815 96816commit d8caa782009abf4dc17b945e325e83fda299a534 96817Author: Adam Jackson <ajax@redhat.com> 96818Date: Thu Feb 24 16:06:34 2011 -0500 96819 96820 vbe: Fix malloc size bug 96821 96822 v2: Slightly more obvious sizing math. 96823 96824 ==14882== Invalid write of size 2 96825 ==14882== at 0x6750267: VBEGetVBEInfo (vbe.c:400) 96826 ==14882== by 0x6142064: ??? (in /usr/lib64/xorg/modules/drivers/vesa_drv.so) 96827 ==14882== by 0x471895: InitOutput (xf86Init.c:519) 96828 ==14882== by 0x422778: main (main.c:205) 96829 ==14882== Address 0x4f32fa8 is 72 bytes inside a block of size 73 alloc'd 96830 ==14882== at 0x4A0640D: malloc (vg_replace_malloc.c:236) 96831 ==14882== by 0x675024B: VBEGetVBEInfo (vbe.c:398) 96832 ==14882== by 0x6142064: ??? (in /usr/lib64/xorg/modules/drivers/vesa_drv.so) 96833 ==14882== by 0x471895: InitOutput (xf86Init.c:519) 96834 ==14882== by 0x422778: main (main.c:205) 96835 96836 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 96837 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 96838 Signed-off-by: Adam Jackson <ajax@redhat.com> 96839 96840commit 628d16a92a7fa556fbb70bf4a4adf57ec05c190b 96841Author: Keith Packard <keithp@keithp.com> 96842Date: Sat Feb 26 10:31:50 2011 -0800 96843 96844 loader: Don't distribute sdksyms.c and make it depend on the config 96845 96846 sdksyms.c is constructed by processing header files with the C 96847 preprocessor. Its contents will vary depending on the precise 96848 configuration options, and so must depend on the config header 96849 files. 96850 96851 We have one header file which is always changed when any config option 96852 is modified called do-not-use-config.h (which may want a different 96853 name at some point), so make sdksyms.c depend on that file. 96854 96855 Also, we don't want to ship this file; it always needs to be 96856 built. So, include it in the nodist_libloader_la_SOURCES list to 96857 prevent it from being added to the tarball. 96858 96859 Signed-off-by: Keith Packard <keithp@keithp.com> 96860 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 96861 96862commit 705f047876f78bbdfb368c3643925af1d1d17f71 96863Author: Julien Cristau <jcristau@debian.org> 96864Date: Fri Mar 4 01:12:01 2011 +0100 96865 96866 configure.ac: bump version for 1.11 development 96867 96868 Signed-off-by: Julien Cristau <jcristau@debian.org> 96869 Signed-off-by: Keith Packard <keithp@keithp.com> 96870 96871commit 6c90e839d9872a37d371578c9c423e8213922044 96872Merge: 0bc95d5b0 edcceedbd 96873Author: Keith Packard <keithp@keithp.com> 96874Date: Thu Mar 3 14:41:44 2011 -0800 96875 96876 Merge remote branch 'whot/for-keith' 96877 96878commit 0bc95d5b06dcea65a1aa193ea907b50f7dd168b5 96879Merge: 3f41f4ade 69a9171db 96880Author: Keith Packard <keithp@keithp.com> 96881Date: Thu Mar 3 14:33:08 2011 -0800 96882 96883 Merge remote branch 'jeremyhu/master' 96884 96885commit 3f41f4adea4bbb90d4bda4dab600595b655e3ed8 96886Merge: 8e4c3ce55 dae24abcd 96887Author: Keith Packard <keithp@keithp.com> 96888Date: Thu Mar 3 14:12:36 2011 -0800 96889 96890 Merge remote branch 'sandmann/for-keithp' 96891 96892commit 8e4c3ce55b0f186bc6ba4039e30629669b6087b7 96893Merge: f3d19c0cf 296561506 96894Author: Keith Packard <keithp@keithp.com> 96895Date: Thu Mar 3 13:42:07 2011 -0800 96896 96897 Merge remote branch 'rjy/clientids' 96898 96899commit 1c008e7e7865b405b8033f625333cd64ece4499e 96900Author: Simon Thum <simon.thum@gmx.de> 96901Date: Wed Feb 23 02:29:35 2011 +0100 96902 96903 dix: change all timestamps in pointer acceleration to CARD32 96904 96905 CARD32 is being returned by GetTimeInMilis(), so use it consistently. 96906 96907 Signed-off-by: Simon Thum <simon.thum@gmx.de> 96908 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96909 96910commit a4b85261859b17dba9ad8f7f1ce650133f0235d4 96911Author: Simon Thum <simon.thum@gmx.de> 96912Date: Wed Feb 23 02:29:34 2011 +0100 96913 96914 dix: update pointer acceleration code to use ValuatorMask 96915 96916 Signed-off-by: Simon Thum <simon.thum@gmx.de> 96917 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96918 96919commit 8128846e16b406c7d459ae7556f7cd09bdc24c91 96920Author: Simon Thum <simon.thum@gmx.de> 96921Date: Wed Feb 23 02:29:33 2011 +0100 96922 96923 dix: refactor predictable scheme initialization 96924 96925 This intends to clean up the predictable accel struct 96926 from purely scheme-related things like input properties, 96927 as they would be useless in other use cases such 96928 as wheel acceleration. 96929 96930 Signed-off-by: Simon Thum <simon.thum@gmx.de> 96931 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96932 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96933 96934commit 69a9171dbbafd6a7db702d48770d28fb54717545 96935Author: Jeremy Huddleston <jeremyhu@apple.com> 96936Date: Tue Mar 1 20:02:46 2011 -0800 96937 96938 XQuartz: pbproxy: Protect against possible collision between Cocoa and X11 Cursor types 96939 96940 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 96941 96942commit dae24abcd44f3bb1966faa88222f851c784b37dc 96943Author: Søren Sandmann Pedersen <ssp@redhat.com> 96944Date: Mon Feb 28 11:10:38 2011 -0500 96945 96946 Delete RegionClipSpans() 96947 96948 Nothing uses it. 96949 96950 Reviewed-by: Adam Jackson <ajax@redhat.com> 96951 Signed-off-by: Søren Sandmann Pedersen <ssp@redhat.com> 96952 96953commit edcceedbd35df576929685767d0b619659e5b020 96954Author: Adam Jackson <ajax@redhat.com> 96955Date: Mon Feb 28 12:32:53 2011 -0500 96956 96957 xext: Remove XCALIBRATE extension 96958 96959 Nobody can have been using this, it's never called from extension init. 96960 XI2 device properties should now be a functional replacement. 96961 96962 Signed-off-by: Adam Jackson <ajax@redhat.com> 96963 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 96964 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96965 96966commit c2207d11f243900bc2f641e08d80da63d84e97a8 96967Merge: 00779932d 579ee8f5d 96968Author: Peter Hutterer <peter.hutterer@who-t.net> 96969Date: Mon Feb 28 11:26:09 2011 +1000 96970 96971 Merge branch 'next' into for-keith 96972 96973 Conflicts: 96974 dix/devices.c 96975 96976 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 96977 96978commit 0343aed1f082baf4bfbd730c1c3582f1410b8741 96979Author: Jeremy Huddleston <jeremyhu@apple.com> 96980Date: Sun Feb 27 16:18:27 2011 -0800 96981 96982 XQuartz: Don't use deprecated CoreGraphics API on SL and Lion 96983 96984 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 96985 96986commit de32d4dcf57c10fc100c1b33630127947a8c8460 96987Author: Jeremy Huddleston <jeremyhu@apple.com> 96988Date: Sun Feb 27 15:49:10 2011 -0800 96989 96990 Revert "rootless: Remove ROOTLESS_WORKAROUND" 96991 96992 Christof Wolf has reported a regression that seems to be caused by 96993 this change, so reverting the change in the 1.9 branch. We'll 96994 investigate a proper fix in master for 1.10. 96995 96996 This reverts commit c89f0521044083a11d538ebfeaabee6fc7fb9a03. 96997 96998 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 96999 97000commit 00779932de861bf7ac8c625802f1afed75d01b6f 97001Author: Matthieu Herrb <matthieu.herrb@laas.fr> 97002Date: Sun Feb 27 20:16:03 2011 +0100 97003 97004 Don't clobber input device options from xorg.conf 97005 97006 Since commit b8d9c5ff removed commonOptions, we now 97007 need to append the "Core{Keyboard,Pointer}" options to 97008 the existing list. 97009 97010 Fixes passing options to devices confirured in xorg.conf 97011 on systems where autoaddevices is false. 97012 97013 Signed-off-by: Matthieu Herrb <matthieu.herrb@laas.fr> 97014 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 97015 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97016 97017commit 8cf3348e90846f5b04236479042228fb98ac8f70 97018Author: Jeremy Huddleston <jeremyhu@apple.com> 97019Date: Tue Feb 8 11:38:35 2011 -0800 97020 97021 XQuartz: RandR: Add RandR modes for the primary display in multi-monitor configs 97022 97023 We now support using RandR to set the resolution of the primary display (and 97024 place a shielding window on other displays) in multi-monitor configurations. 97025 97026 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 97027 97028commit 968652983f8e6ae6889b48e15f4098ff6ad4a15a 97029Author: Jeremy Huddleston <jeremyhu@apple.com> 97030Date: Tue Feb 8 12:26:35 2011 -0800 97031 97032 XQuartz: RandR: Provide an alert box when entering a RandR mode for the first time. 97033 97034 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 97035 97036commit 13578b852b7631f99cf1fd5e2e5469edc5aae369 97037Author: Jeremy Huddleston <jeremyhu@apple.com> 97038Date: Tue Feb 8 11:37:52 2011 -0800 97039 97040 XQuartz: RandR: Capture the display when switching modes with RandR 97041 97042 This will prevent native windows from resizing as we change resolutions. 97043 97044 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 97045 97046commit 7dff79e39564b403c3afbc5e7bacffa0df190a23 97047Author: Søren Sandmann Pedersen <ssp@redhat.com> 97048Date: Fri Feb 11 09:00:10 2011 -0500 97049 97050 Implement fbAddTriangles() in terms of pixman_add_triangles() 97051 97052 This allows the remaining triangle-to-trap conversion code to be 97053 deleted. 97054 97055 Reviewed-by: Adam Jackson <ajax@redhat.com> 97056 Signed-off-by: Søren Sandmann <ssp@redhat.com> 97057 97058commit 566f1931ee2916269e164e114bffaf2da1d039d1 97059Author: Søren Sandmann Pedersen <ssp@redhat.com> 97060Date: Tue Jan 11 10:33:57 2011 -0500 97061 97062 Move miTriangles to fb as fbTriangles() 97063 97064 The fb version simply calls the new pixman_composite_triangles(). This 97065 allows us to get rid of miCreateAlphaPicture(). 97066 97067 Reviewed-by: Adam Jackson <ajax@redhat.com> 97068 Signed-off-by: Søren Sandmann <ssp@redhat.com> 97069 97070commit 788ccb9a8bcf6a4fb4054c507111eec3338fb969 97071Author: Søren Sandmann Pedersen <ssp@redhat.com> 97072Date: Tue Jan 11 09:46:46 2011 -0500 97073 97074 Move miTrapezoids() into fb as fbTrapezoids() 97075 97076 The main consumer of trapezoids, cairo, is using the Trapezoids 97077 request, which is currently implemented in the miTrapezoids() 97078 function. That function splits the request into smaller bits and calls 97079 lower level functions such as AddTrap. 97080 97081 By moving the implementation of the whole request into fb, we can 97082 instead call pixman_composite_trapezoids() to do the whole request in 97083 one step. 97084 97085 There are no callers of miTrapezoids in any of the open source 97086 drivers, although exa and uxa have their own copies of the function. 97087 97088 Reviewed-by: Adam Jackson <ajax@redhat.com> 97089 Signed-off-by: Søren Sandmann <ssp@redhat.com> 97090 97091commit 197df069a4037d6faa2723c31ffba09c95d71166 97092Author: Søren Sandmann Pedersen <ssp@redhat.com> 97093Date: Wed Feb 23 10:36:57 2011 -0500 97094 97095 Require pixman 0.21.6 97096 97097 The following patches need pixman_composite_trapezoids() and 97098 pixman_add_triangles(). 97099 97100 Signed-off-by: Soren Sandmann <ssp@redhat.com> 97101 97102commit f3d19c0cf3327cceb90f7378f5d8fc0c3e327400 97103Author: Keith Packard <keithp@keithp.com> 97104Date: Fri Feb 25 21:22:23 2011 -0800 97105 97106 Version bumped to 1.10 97107 97108 Signed-off-by: Keith Packard <keithp@keithp.com> 97109 97110commit 678f5396c91b3d0c7572ed579b0a4fb62b2b4655 97111Author: Keith Packard <keithp@keithp.com> 97112Date: Fri Feb 25 21:10:21 2011 -0800 97113 97114 input: Ensure Valuator axes are aligned as needed 97115 97116 Let the compiler figure out the correct alignment for the axes data 97117 for a valuator by using a union to force double alignment of the 97118 initial ValuatorClassRec structure in the allocation. 97119 97120 Signed-off-by: Keith Packard <keithp@keithp.com> 97121 Tested-by: Julien Cristau <jcristau@debian.org> 97122 Reviewed-by: Julien Cristau <jcristau@debian.org> 97123 97124commit dc8f52e77f51b6fa8908d9611c3f7e3cfbbaf2d1 97125Author: Keith Packard <keithp@keithp.com> 97126Date: Fri Feb 25 16:38:27 2011 -0800 97127 97128 hw/dmx/doc: Add explicit dependency for all doxygen output files 97129 97130 Instead of listing one of the doxygen output files and depending on 97131 sequential execution to ensure that the other files were present 97132 before make checked for them, create explicit dependencies so that 97133 make will not check for the additional files until after doxygen has 97134 been run. 97135 97136 This allows parallel make to work correctly in this directory. 97137 97138 Signed-off-by: Keith Packard <keithp@keithp.com> 97139 Tested-by: Kristian Høgsberg <krh@bitplanet.net> 97140 97141commit fd4d9c75c265add8d6cc74afb341c9c4f2ec1073 97142Author: Keith Packard <keithp@keithp.com> 97143Date: Fri Feb 25 10:54:50 2011 -0800 97144 97145 hw/xwin: Look for gl spec files in $(srcdir) or . 97146 97147 Tarballs include the downloaded gl spec files, which will end up in 97148 $(srcdir). But, git-based builds will not have them at all and will 97149 need to download them from opengl.org. They'll land in in the build 97150 directory instead of $(srcdir), and so we need to allow them to be in 97151 either place. 97152 97153 This change checks for the files in $(srcdir), linking them to . if 97154 present. Otherwise, it downloads them from opengl.org. 97155 97156 A suggested better solution is to have Mesa install these files somewhere. 97157 97158 Signed-off-by: Keith Packard <keithp@keithp.com> 97159 Tested-by: Kristian Høgsberg <krh@bitplanet.net> 97160 97161commit 6b951dec69103ca76d6d4a61b1a2f67a52d013b9 97162Author: Dan Nicholson <dbn.lists@gmail.com> 97163Date: Sun Feb 20 09:51:15 2011 -0800 97164 97165 dmx: Construct paths in doxygen.conf to fix VPATH builds 97166 97167 The paths in doxygen.conf assumed that srcdir=builddir and broke 97168 otherwise. Use autoconf to fill in the paths to the srcdir so that the 97169 files can be found when users have a separate build directory (as with 97170 distcheck). 97171 97172 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 97173 Tested-by: Cyril Brulebois <kibi@debian.org> 97174 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 97175 Tested-by: Sedat Dilek <sedat.dilek@gmail.com> 97176 97177commit 61ce915bf78570b7c8d53c118700ef7274bbdfb7 97178Author: Dan Nicholson <dbn.lists@gmail.com> 97179Date: Tue Feb 15 06:03:28 2011 -0800 97180 97181 xfree86: Allow sdksyms.dep to be included portably 97182 97183 Non-GNU makes don't deal with the sinclude or -include variants that 97184 allow Makefile stubs to be created and then included during the build. 97185 Instead, create an empty file at the end of configure so that the 97186 regular include statement can be included. This is how automake handles 97187 automatic source dependencies. 97188 97189 In order to trick automake into not processing the include statement, a 97190 variable is used. 97191 97192 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 97193 Tested-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97194 97195commit 780a77acce1dd369549ece802b3e2c4006058dfe 97196Author: Keith Packard <keithp@keithp.com> 97197Date: Thu Feb 24 20:45:46 2011 -0800 97198 97199 Version bumped to 1.9.99.903 (1.10 RC3) 97200 97201 Signed-off-by: Keith Packard <keithp@keithp.com> 97202 97203commit 9d23459415b84606ee4f38bb2d19054c432c8552 97204Author: Peter Hutterer <peter.hutterer@who-t.net> 97205Date: Fri Feb 25 11:08:19 2011 +1000 97206 97207 dix: release all buttons and keys before reattaching a device (#34182) 97208 97209 Testcase: 97210 xinput float <keyboard name> 97211 97212 results in the keyboard's enter key being repeated as the device is detached 97213 while the key is still physically down. To avoid this, release all keys and 97214 buttons before reattaching the device. 97215 97216 X.Org Bug 34182 <http://bugs.freedesktop.org/show_bug.cgi?id=34182> 97217 97218 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97219 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 97220 97221commit 229b055bdbc53370944401649234e8a38183f4fb 97222Merge: 6178959e3 b17fc99cb 97223Author: Keith Packard <keithp@keithp.com> 97224Date: Thu Feb 24 19:44:35 2011 -0800 97225 97226 Merge remote branch 'jeremyhu/master' 97227 97228commit 6178959e3dd6482a4317de6eb14eb19ca7329b9c 97229Author: Keith Packard <keithp@keithp.com> 97230Date: Thu Feb 24 19:42:02 2011 -0800 97231 97232 xfree86: Bump video ABI to 10.0 97233 97234 RandR 1.4 revert changed things 97235 97236 Signed-off-by: Keith Packard <keithp@keithp.com> 97237 97238commit 31704510f448706524b7b7085cc4ff0ada7bfe7e 97239Author: Michel Dänzer <daenzer@vmware.com> 97240Date: Thu Feb 24 12:17:57 2011 +0100 97241 97242 EXA/mixed: ModifyPixmapHeader pitch fixes. (bug #33929) 97243 97244 If there's a GPU copy and a non-zero devKind was passed in, set the GPU copy 97245 pitch to that instead of to a possibly bogus value derived from the new width. 97246 This is e.g. used by the radeon driver's drmmode_xf86crtc_resize hook, fixes 97247 https://bugs.freedesktop.org/show_bug.cgi?id=33929 . 97248 97249 On the other hand, the system memory copy doesn't need the pitch to be aligned 97250 beyond the PixmapBytePad of the width. 97251 97252 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 97253 Acked-by: Cyril Brulebois <kibi@debian.org> 97254 Tested-by: Cyril Brulebois <kibi@debian.org> 97255 Reported-by: Thierry Vignaud <thierry.vignaud@gmail.com> 97256 Tested-by: Thierry Vignaud <thierry.vignaud@gmail.com> 97257 Signed-off-by: Keith Packard <keithp@keithp.com> 97258 97259commit b17fc99cb9ca9ff0a3592f783a906cdcc35da748 97260Author: Jeremy Huddleston <jeremyhu@apple.com> 97261Date: Thu Feb 24 19:24:35 2011 -0800 97262 97263 XQuartz: Localization Updates 97264 97265 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 97266 97267commit 59850630fefe5e4f0430d7e2e106937fdf1e7ffb 97268Author: Jeremy Huddleston <jeremyhu@apple.com> 97269Date: Thu Feb 24 19:18:35 2011 -0800 97270 97271 XQuartz: Add LSApplicationCategoryType key to Info.plist 97272 97273 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 97274 97275commit 83861595782aaa05907f9cf7b236d50261d404d9 97276Merge: a1cc0e52b 4102a0096 97277Author: Keith Packard <keithp@keithp.com> 97278Date: Thu Feb 24 18:59:07 2011 -0800 97279 97280 Merge remote branch 'kibi/master' 97281 97282commit a1cc0e52b0e8ca40bc1218de553424b2bcb760ef 97283Author: Alexandr Shadchin <alexandr.shadchin@gmail.com> 97284Date: Tue Nov 30 21:17:19 2010 +0500 97285 97286 Removing unused code 97287 97288 In OpenBSD removed support PCCONS in 2002 year 97289 http://marc.info/?l=openbsd-cvs&m=102435816424294&w=2 97290 97291 Signed-off-by: Alexandr Shadchin <Alexandr.Shadchin@gmail.com> 97292 Reviewed-by: Matthieu Herrb <matthieu.herrb@laas.fr> 97293 Signed-off-by: Keith Packard <keithp@keithp.com> 97294 97295commit 365ad68fb9f7029550505b7c276a808050cada9c 97296Merge: 0801afbd7 720c895db 97297Author: Keith Packard <keithp@keithp.com> 97298Date: Thu Feb 24 18:49:40 2011 -0800 97299 97300 Merge remote branch 'ajax/for-keithp' 97301 97302commit 0801afbd7c2c644c672b37f8463f1a0cbadebd2e 97303Author: Erkki Seppälä <erkki.seppala@vincit.fi> 97304Date: Thu Feb 10 15:35:14 2011 +0200 97305 97306 record: avoid crash when calling RecordFlushReplyBuffer recursively 97307 97308 RecordFlushReplyBuffer can call itself recursively through 97309 WriteClient->CallCallbacks->_CallCallbacks->RecordFlushAllContexts 97310 when the recording client's buffer cannot be completely emptied in one 97311 WriteClient. When a such a recursion occurs, it will not be broken out 97312 of which results in segmentation fault when the stack is exhausted. 97313 97314 This patch adds a counter (a flag, really) that guards against this 97315 situation, to break out of the recursion. 97316 97317 One alternative to this change would be to change _CallCallbacks to 97318 check the corresponding counter before the callback loop, but that 97319 might affect existing behavior, which may be relied upon. 97320 97321 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 97322 Signed-off-by: Erkki Seppälä <erkki.seppala@vincit.fi> 97323 Signed-off-by: Keith Packard <keithp@keithp.com> 97324 97325commit 816f1f8ffafbfbf0dd31ea86f295987f84151feb 97326Author: Peter Hutterer <peter.hutterer@who-t.net> 97327Date: Tue Feb 22 12:48:04 2011 +1000 97328 97329 include: document list interface. 97330 97331 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97332 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97333 97334commit b833f9b924ddc043244c0a026547c438f1c4d4df 97335Author: Keith Packard <keithp@keithp.com> 97336Date: Wed Feb 23 11:18:35 2011 -0800 97337 97338 Revert "Replace huge argument list in xf86CrtcSetModeTransform with struct" 97339 97340 This reverts commit 8b35118c03590a7ad3786d3284bafb3f40fcb8cc. 97341 97342commit 7030c82648930cea7ef8a2ea49d8138078920d4d 97343Author: Keith Packard <keithp@keithp.com> 97344Date: Wed Feb 23 11:18:16 2011 -0800 97345 97346 Revert "Require RandR protocol version 1.4 or newer" 97347 97348 This reverts commit c8bc25fd7629df10f2825b7cc713b031ae78f223. 97349 97350commit a91d33917befd6719bb7f01007bc253a2aefb0bb 97351Author: Keith Packard <keithp@keithp.com> 97352Date: Wed Feb 23 11:18:12 2011 -0800 97353 97354 Revert "randr: Add sprite position transforms" 97355 97356 This reverts commit 66294afcab7b7a82f7dd897767e46c48a94b8ee8. 97357 97358commit 3d4ee25a1e2e0ed2463ceaabea8f5a284a6ddd60 97359Author: Keith Packard <keithp@keithp.com> 97360Date: Wed Feb 23 11:18:08 2011 -0800 97361 97362 Revert "randr: Implement RRSetCrtcConfigs" 97363 97364 This reverts commit d94a035ea9eb3167fc4f35b2d9f0d53f8807014c. 97365 97366commit 8b0e651f7d09a7f35c87a4d0a6e2053cae93b516 97367Author: Keith Packard <keithp@keithp.com> 97368Date: Wed Feb 23 11:18:05 2011 -0800 97369 97370 Revert "hw/xfree86/modes: Add optional driver API for RRSetCrtcConfigs" 97371 97372 This reverts commit 86c489c319c705f710bee3897fe27600ce15008e. 97373 97374commit 9e8c20b0d49783d6b9334d4c11c4e6e3ba273524 97375Author: Keith Packard <keithp@keithp.com> 97376Date: Wed Feb 23 11:17:42 2011 -0800 97377 97378 Revert "randr: Add per-crtc pixmaps" 97379 97380 This reverts commit 82612045e11f2b882ae132e184a9629f43f1c424. 97381 97382commit 9f47780ecdc7693f756587a758ec0141e75cb1eb 97383Author: Keith Packard <keithp@keithp.com> 97384Date: Wed Feb 23 11:17:36 2011 -0800 97385 97386 Revert "randr: Hook up the new RandR 1.4 functionality" 97387 97388 This reverts commit afb6ebf1d5829346c40fe1053c9f50afe926e6c6. 97389 97390commit 793a242c89fccdbf1a2557ba33da1da2cb4ff8c4 97391Author: Keith Packard <keithp@keithp.com> 97392Date: Wed Feb 23 11:17:33 2011 -0800 97393 97394 Revert "DIX is responsible for ref counting scanout pixmaps." 97395 97396 This reverts commit 96b4d4787bf82edd9d06eb9a6e94bc45412c7df2. 97397 97398commit 7b7cd3f121b3904ad851155f853ce87a0e00d002 97399Author: Keith Packard <keithp@keithp.com> 97400Date: Wed Feb 23 11:17:29 2011 -0800 97401 97402 Revert "Set sprite transforms from RRSetCrtcConfigs" 97403 97404 This reverts commit a88d70fb20a2bc3152b84adff4380857e6cfadf5. 97405 97406commit fd9331f6eb39a28da7de4867b2e3a460f667d514 97407Author: Keith Packard <keithp@keithp.com> 97408Date: Wed Feb 23 11:17:24 2011 -0800 97409 97410 Revert "Separate out screen size and screen pixmap sizes in RRScreenSizeSet" 97411 97412 This reverts commit 752c368421c1c824752cf467fba9318d75d2ca2c. 97413 97414commit 4d91e7a63161cdde9af53aedb3fe3e53dbe3049e 97415Author: Keith Packard <keithp@keithp.com> 97416Date: Wed Feb 23 11:17:20 2011 -0800 97417 97418 Revert "ProcRRSetCrtcConfigs uses 'configs' without being initialized" 97419 97420 This reverts commit b0f4bd61f0caf80f3be9a176f1f7a707bc6628d8. 97421 97422commit 00d0b235cfbfb162ddd58c088d2ed03e55a9c5cb 97423Author: Keith Packard <keithp@keithp.com> 97424Date: Wed Feb 23 11:16:48 2011 -0800 97425 97426 Revert "randr: handle RRSetCrtcConfigs request with zero configs" 97427 97428 This reverts commit 0d01b66df9081ef48843b3bad81c56bb2cd1ae69. 97429 97430commit 720c895db9327bbdc32215a67236303d2ceafad4 97431Author: Peter Hutterer <peter.hutterer@who-t.net> 97432Date: Wed Feb 23 13:36:20 2011 -0500 97433 97434 Add mode field to ConstrainCursorHarder 97435 97436 For Pointer Barriers, the movement mode is important and must be passed 97437 through. 97438 97439 Reviewed-by: Adam Jackson <ajax@redhat.com> 97440 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97441 97442commit a16e282200f84233041e5a565c6363a5a78be525 97443Author: Adam Jackson <ajax@redhat.com> 97444Date: Mon Feb 8 15:40:22 2010 -0500 97445 97446 xfree86: Fix the sdk headers to be multilib-safe 97447 97448 Use _LP64 (verified on gcc and sun compilers) instead of _XSERVER64 in 97449 internal header usage, and always define _XSERVER64 for modules if _LP64 97450 is defined. Prevents differing xorg-server.h between 32 and 64 bit 97451 packages. 97452 97453 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97454 Signed-off-by: Adam Jackson <ajax@redhat.com> 97455 97456commit dc498b433f36af5d2de3065e7c64cdb575385d81 97457Author: Adam Jackson <ajax@redhat.com> 97458Date: Mon Feb 8 15:23:04 2010 -0500 97459 97460 xfree86: If the driver found modes on an output, don't add more 97461 97462 Inferring modes from sync ranges is only valid if the monitor says it's 97463 valid. If the monitor says it's valid, then we'll have already added 97464 those modes during EDID block parse. If it doesn't, then we should 97465 believe it. 97466 97467 If there's no EDID for an output, but sync ranges from the config, we'll 97468 still add default modes as normal. 97469 97470 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 97471 Signed-off-by: Adam Jackson <ajax@redhat.com> 97472 97473commit 303977fbcf9c641b7e19dfde192cef585f5b455c 97474Author: Adam Jackson <ajax@redhat.com> 97475Date: Tue Feb 8 14:52:49 2011 -0500 97476 97477 glxproxy: warning fix 97478 97479 glxvendor.c: In function ‘__glXVForwardPipe0WithReply’: 97480 glxvendor.c:205:10: warning: ‘be_buf’ may be used uninitialized in this function 97481 97482 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97483 Signed-off-by: Adam Jackson <ajax@redhat.com> 97484 97485commit 405ad0610d6e2f4606768294b655ac52aee9ae58 97486Author: Adam Jackson <ajax@redhat.com> 97487Date: Tue Feb 8 14:51:47 2011 -0500 97488 97489 glxproxy: warning fix 97490 97491 glxvendor.c: In function ‘__glXVForwardAllWithReply’: 97492 glxvendor.c:284:10: warning: ‘be_buf’ may be used uninitialized in this function 97493 glxvendor.c:285:10: warning: ‘be_buf_size’ may be used uninitialized in this function 97494 97495 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97496 Signed-off-by: Adam Jackson <ajax@redhat.com> 97497 97498commit a4cd2e0da6415ec1b3b182579bebbe2a41f29d30 97499Author: Adam Jackson <ajax@redhat.com> 97500Date: Tue Feb 8 14:47:05 2011 -0500 97501 97502 glxproxy: warning fix 97503 97504 glxsingle.c: In function ‘__glXForwardPipe0WithReply’: 97505 glxsingle.c:218:10: warning: ‘be_buf’ may be used uninitialized in this function 97506 97507 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97508 Signed-off-by: Adam Jackson <ajax@redhat.com> 97509 97510commit f928caca6e60215bc9ab423acae5542dfabc9bec 97511Author: Adam Jackson <ajax@redhat.com> 97512Date: Tue Feb 8 14:45:31 2011 -0500 97513 97514 glxproxy: warning fix 97515 97516 glxsingle.c: In function ‘__glXForwardAllWithReply’: 97517 glxsingle.c:300:10: warning: ‘be_buf’ may be used uninitialized in this function 97518 glxsingle.c:301:10: warning: ‘be_buf_size’ may be used uninitialized in this function 97519 97520 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97521 Signed-off-by: Adam Jackson <ajax@redhat.com> 97522 97523commit dad2712c9328e113db4de768a12a8dafa6c177e9 97524Author: Adam Jackson <ajax@redhat.com> 97525Date: Tue Feb 8 14:44:03 2011 -0500 97526 97527 glxproxy: warning fix 97528 97529 glxsingle.c: In function ‘__glXDisp_ReadPixels’: 97530 glxsingle.c:760:11: warning: ‘buf’ may be used uninitialized in this function 97531 97532 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97533 Signed-off-by: Adam Jackson <ajax@redhat.com> 97534 97535commit 4270157bac645550e2c0afe89479c0bfe9d53447 97536Author: Adam Jackson <ajax@redhat.com> 97537Date: Tue Feb 8 14:38:23 2011 -0500 97538 97539 glxproxy: warning fix 97540 97541 render2swap.c:264:13: warning: ‘swapArray’ defined but not used 97542 97543 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97544 Signed-off-by: Adam Jackson <ajax@redhat.com> 97545 97546commit 5d0910b4fdf99e3239635a01eb2709c32d0e5bb9 97547Author: Adam Jackson <ajax@redhat.com> 97548Date: Tue Feb 8 14:32:54 2011 -0500 97549 97550 glxproxy: warning fix 97551 97552 glxcmds.c: In function ‘CreateGLXPixmap’: 97553 glxcmds.c:1663:20: warning: comparison between pointer and integer 97554 glxcmds.c:1663:38: warning: comparison between pointer and integer 97555 97556 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97557 Signed-off-by: Adam Jackson <ajax@redhat.com> 97558 97559commit e1b5d3e5e7f157ab769c40ad2efdc1281c25a03a 97560Author: Adam Jackson <ajax@redhat.com> 97561Date: Tue Feb 8 14:18:53 2011 -0500 97562 97563 glxproxy: warning fix 97564 97565 glxcmds.c: In function ‘CreateGLXPixmap’: 97566 glxcmds.c:1641:22: warning: ‘pGlxScreen’ may be used uninitialized in this function 97567 97568 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97569 Signed-off-by: Adam Jackson <ajax@redhat.com> 97570 97571commit 3b9de273a94fb68de51238c20c3182396aa41b84 97572Author: Adam Jackson <ajax@redhat.com> 97573Date: Tue Feb 8 14:07:20 2011 -0500 97574 97575 glxproxy: warning fix 97576 97577 glxcmds.c: In function ‘CreateContext.clone.6’: 97578 glxcmds.c:105:19: warning: ‘be_fbconfigId’ may be used uninitialized in this function 97579 glxcmds.c:104:14: warning: ‘be_vid’ may be used uninitialized in this function 97580 97581 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97582 Signed-off-by: Adam Jackson <ajax@redhat.com> 97583 97584commit 767b93e783a981b4fc926299a85a1a18387e693a 97585Author: Adam Jackson <ajax@redhat.com> 97586Date: Tue Feb 8 14:02:37 2011 -0500 97587 97588 glxproxy: warning fix 97589 97590 glxcmds.c: In function ‘__glXGetDrawableAttributes’: 97591 glxcmds.c:3295:8: warning: ‘screen’ may be used uninitialized in this function 97592 glxcmds.c:3298:8: warning: ‘attribs_size’ may be used uninitialized in this function 97593 97594 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97595 Signed-off-by: Adam Jackson <ajax@redhat.com> 97596 97597commit a391089186cd1063d807bf14a9651e6861b248de 97598Author: Adam Jackson <ajax@redhat.com> 97599Date: Tue Feb 8 13:55:23 2011 -0500 97600 97601 glxproxy: warning fix 97602 97603 glxcmds.c: In function ‘__glXChangeDrawableAttributes’: 97604 glxcmds.c:3464:8: warning: ‘screen’ may be used uninitialized in this function 97605 97606 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97607 Signed-off-by: Adam Jackson <ajax@redhat.com> 97608 97609commit 4102a00962f568b1d34cf72ad524e1cc705b0b1c 97610Author: Cyril Brulebois <kibi@debian.org> 97611Date: Wed Feb 23 14:44:23 2011 +0100 97612 97613 xfree86: Fix undefined reference to `XNFsprintf' on sparc. 97614 97615 Build failure on sparc: 97616 | copying selected object files to avoid basename conflicts... 97617 | CCLD Xorg 97618 | ./.libs/libxorg.a(xf86sbusBus.o): In function `xf86SbusConfigureNewDev': 97619 | […]/hw/xfree86/common/xf86sbusBus.c:712: undefined reference to `XNFsprintf' 97620 | collect2: ld returned 1 exit status 97621 97622 Fix typo introduced in: 97623 3a9bb93dd178084f4ff1abcea331ca5a62f88ce6 97624 97625 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97626 Apologized-for-by: Alan Coopersmith <alan.coopersmith@oracle.com> 97627 Signed-off-by: Cyril Brulebois <kibi@debian.org> 97628 97629commit 579ee8f5d84c3a523b7b3e3941eabb226d1d19e2 97630Merge: b63689313 17265ccb0 97631Author: Peter Hutterer <peter.hutterer@who-t.net> 97632Date: Wed Feb 23 08:44:42 2011 +1000 97633 97634 Merge branch 'mi-cleanup' into next 97635 97636commit b636893137da1695e235e3a9354bfd9243fdddc2 97637Author: Daniel Stone <daniel@fooishbar.org> 97638Date: Tue Feb 22 13:43:28 2011 +0000 97639 97640 Input: Don't freeze unrelated devices in DeliverGrabbedEvent 97641 97642 When delivering an event to a device grabbed with SyncBoth, 97643 DeliverGrabbedEvent walks the device tree looking for associated devices 97644 to freeze them. Unfortunately, it froze all devices instead of just the 97645 paired device, and the previous fix in 4fbadc8b17237f3c would still break 97646 if the same client had a non-SyncBoth grab on another unrelated master 97647 device. 97648 97649 Fix this by completely ignoring devices that aren't our paired device. 97650 97651 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 97652 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 97653 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97654 97655commit 17265ccb027e3f956bf7409106174f44621d1cb8 97656Author: Peter Hutterer <peter.hutterer@who-t.net> 97657Date: Fri Feb 18 12:28:43 2011 +1000 97658 97659 Move master/lastSlave out of the union into separate fields. 97660 97661 The removal of the double-use will cause some suble bugs as some conditions 97662 to check for the dev->u.master case were broken and also evaluated as true 97663 if lastSlave was set (instead of master). 97664 97665 Also breaks the input ABI. 97666 97667 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97668 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97669 97670commit 77113dd3eef03dd65b556b672d976817b3f4542e 97671Author: Peter Hutterer <peter.hutterer@who-t.net> 97672Date: Fri Feb 18 11:45:29 2011 +1000 97673 97674 input: Change a bunch of direct dev->u.master accesses to use GetMaster() 97675 97676 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97677 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97678 97679commit 2936635698619271a790004480a14285149277cb 97680Author: Peter Hutterer <peter.hutterer@who-t.net> 97681Date: Fri Feb 18 12:06:56 2011 +1000 97682 97683 xkb: Fix a check for MASTER_KEYBOARD 97684 97685 And copy into the master keyboard, not just the directly attached device. 97686 97687 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97688 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 97689 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97690 97691commit 0ba526deeb4d54ee1a75400291ee7a3673ed4d96 97692Author: Peter Hutterer <peter.hutterer@who-t.net> 97693Date: Fri Feb 18 12:04:46 2011 +1000 97694 97695 mi: check for MASTER_POINTER instead of manual checks. 97696 97697 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97698 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97699 97700commit 52c13896ce9a1e178d4eeed15f68020947cc20ed 97701Author: Peter Hutterer <peter.hutterer@who-t.net> 97702Date: Fri Feb 18 11:57:06 2011 +1000 97703 97704 mi: update macro to search for MASTER_POINTER 97705 97706 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97707 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97708 97709commit d270f12a3e44f4ea01f176a86783e8cd4c59ddf7 97710Author: Peter Hutterer <peter.hutterer@who-t.net> 97711Date: Fri Feb 18 11:55:51 2011 +1000 97712 97713 xfree86: update GetMaster check for the VCP. 97714 97715 Same result, but now also triggers on slave keyboards that send pointer 97716 events. 97717 97718 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97719 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97720 97721commit ce7f79efd588899f01c74f95c2f58d6dd6d816cd 97722Author: Peter Hutterer <peter.hutterer@who-t.net> 97723Date: Fri Feb 18 11:53:06 2011 +1000 97724 97725 dix: fix a master device access in change_modmap. 97726 97727 We need to check if our master keyboard is the given device since we may be 97728 a pointer with keys and thus need to change the modmap too. 97729 97730 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97731 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97732 97733commit e1ac704185dee31b427a46cd41a00ef7a28b4129 97734Author: Peter Hutterer <peter.hutterer@who-t.net> 97735Date: Fri Feb 18 11:42:09 2011 +1000 97736 97737 dix: fix up a master check in ChangeKeyboardMapping handling. 97738 97739 We don't just care about the directly attached master, we care about the 97740 master keyboard. 97741 97742 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97743 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97744 97745commit febce8cb814df46018f6ae1e6a9daea019b8ad0a 97746Author: Peter Hutterer <peter.hutterer@who-t.net> 97747Date: Fri Feb 18 11:32:28 2011 +1000 97748 97749 Xi: replace a direct master access with GetMaster() 97750 97751 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97752 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97753 97754commit df6559237a2d641b2fc38d14975beab9bae0d971 97755Author: Peter Hutterer <peter.hutterer@who-t.net> 97756Date: Fri Feb 18 11:30:10 2011 +1000 97757 97758 dix: add MASTER_ATTACHED as allowed type for GetMaster(). 97759 97760 In some cases, we don't know/care whether we want the master pointer or keyboard 97761 for a device. Add a new type MASTER_ATTACHED to return the master this 97762 device is attached to. 97763 97764 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97765 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97766 97767commit dc57f89959e549403f8488eb9f23425bd7118b22 97768Author: Peter Hutterer <peter.hutterer@who-t.net> 97769Date: Fri Feb 11 12:47:37 2011 +1000 97770 97771 Switch to use IsFloating() 97772 97773 This is not a straightforward search/replacement due to a long-standing 97774 issue. 97775 97776 dev->u.master is the same field as dev->u.lastSlave. Thus, if dev is a master 97777 device, a check for dev->u.master may give us false positives and false 97778 negatives. 97779 The switch to IsFloating() spells out these cases and modifies the 97780 conditions accordingly to cover both cases. 97781 97782 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97783 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97784 97785commit e48bf3b6403dde33586ca0e421db61e402525453 97786Author: Peter Hutterer <peter.hutterer@who-t.net> 97787Date: Fri Feb 11 11:11:57 2011 +1000 97788 97789 xfree86: replace two inputInfo.pointer uses with device loops. 97790 97791 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97792 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97793 97794commit 703baece7ebd128a6742d0523d5b3ebe65126fa5 97795Author: Peter Hutterer <peter.hutterer@who-t.net> 97796Date: Fri Feb 11 12:19:26 2011 +1000 97797 97798 dix: Add IsFloating(device) wrapper. 97799 97800 Simplifies check for floating devices. 97801 97802 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97803 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97804 97805commit 77af45ebc3eda32dc534968cab8d5adfb01bd9e3 97806Author: Peter Hutterer <peter.hutterer@who-t.net> 97807Date: Fri Feb 11 12:48:30 2011 +1000 97808 97809 dix: Simplify retrieving the master device. 97810 97811 GetMaster() returns NULL for floating slaves. 97812 97813 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97814 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97815 97816commit dc32a23890776edf575bf18b3f3c079da6214340 97817Author: Peter Hutterer <peter.hutterer@who-t.net> 97818Date: Fri Feb 11 15:21:51 2011 +1000 97819 97820 Fix two incorrect checks for master devices. 97821 97822 These two were sideeffects of lastSlave being in the same field as the 97823 master. For devices generated by the master device directly, lastSlave was 0 97824 and the device would (with the old checks) be interpreted as floating. 97825 97826 Add the required checks to safeguard against master devices. 97827 97828 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97829 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97830 97831commit d63c979c7fe0f2b114b27e73ebe0a706be8840ae 97832Author: Peter Hutterer <peter.hutterer@who-t.net> 97833Date: Fri Feb 11 15:21:31 2011 +1000 97834 97835 dix: replace direct master access with GetMaster and temp. device. 97836 97837 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97838 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97839 97840commit c2a9e0a96983f22f721196513e75a53ea0b86d7e 97841Author: Peter Hutterer <peter.hutterer@who-t.net> 97842Date: Fri Feb 11 11:41:47 2011 +1000 97843 97844 mi: simplify a "check for keyboard" condition 97845 97846 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97847 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97848 97849commit 3a6160b408447ce620140849f962683a5d139cb9 97850Author: Peter Hutterer <peter.hutterer@who-t.net> 97851Date: Fri Feb 11 11:40:01 2011 +1000 97852 97853 mi: remove if 0 code. 97854 97855 Note sure why this is in if 0 but it's been that way since 2007. 97856 97857 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97858 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97859 97860commit d9987c8c45236bc9cfcaf716f4bfcac2a9d0e7e6 97861Author: Peter Hutterer <peter.hutterer@who-t.net> 97862Date: Fri Feb 11 11:22:45 2011 +1000 97863 97864 mi: Add a few comments explaining various cursor move functions. 97865 97866 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97867 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97868 97869commit 15fe86e69fc256342881112cd07565527e32435b 97870Author: Peter Hutterer <peter.hutterer@who-t.net> 97871Date: Fri Feb 11 11:01:19 2011 +1000 97872 97873 mi: remove static GenerateEvent variable. 97874 97875 Push into the respective devices. This should have no functional changes 97876 since we never warp more than one device at a time. In the glorious future 97877 with true multithreading, still the better thing to do. 97878 97879 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97880 Reviewed-by: Benjamin Tissoires <tissoire@cena.fr> 97881 97882commit 93a73993708b1345c86ec3ec06b02ed236595673 97883Author: Peter Hutterer <peter.hutterer@who-t.net> 97884Date: Mon Feb 21 15:40:40 2011 +1000 97885 97886 test: write some event → XI1 conversion tests. 97887 97888 Don't test everything, but hey, life is short and I'm trying to have one. 97889 97890 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97891 Reviewed-by: Keith Packard <keithp@keithp.com> 97892 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 97893 97894commit 4cdf1013771bc86fe2f6d9223bc4a46753bc918f 97895Author: Peter Hutterer <peter.hutterer@who-t.net> 97896Date: Mon Feb 21 15:32:57 2011 +1000 97897 97898 dix: a valuator number of 0 is valid (#34510) 97899 97900 For all but motion and proximity events, having no valuators is ok. 97901 Regression from 1.9, keyboard events are not converted to protocol events. 97902 97903 X.Org Bug 34510 <http://bugs.freedesktop.org/show_bug.cgi?id=34510> 97904 97905 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97906 Reviewed-by: Keith Packard <keithp@keithp.com> 97907 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 97908 Tested-by: Timo Aaltonen <timo.aaltonen@canonical.com> 97909 97910commit b4ef34d4664e0eaac7211f7a22a2025958aa1527 97911Author: Maarten Maathuis <madman2003@gmail.com> 97912Date: Sun Feb 20 11:59:41 2011 +0100 97913 97914 Revert "exa/mixed: Exclude frontbuffer from deferred pixmap handling." 97915 97916 This reverts commit 541b25038a5de74411a094570b407c5ae018c2ba. 97917 97918 - It turns out that the high latency was a driver problem. 97919 - catting a large amount of text turns out to look prettier when the 97920 throughput is lower, but it's not worth the loss for a minor 97921 improvement that may not even exist on someone else's computer. 97922 97923 Reviewed-by: Michel Dänzer <michel@daenzer.net> 97924 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 97925 Signed-off-by: Keith Packard <keithp@keithp.com> 97926 97927commit 443d75446bcbe7d97a23860a1e2d46c0b7e7cb26 97928Author: Keith Packard <keithp@keithp.com> 97929Date: Fri Feb 18 14:46:33 2011 -0800 97930 97931 Version bumped to 1.9.99.902 (1.10 RC2) 97932 97933 Signed-off-by: Keith Packard <keithp@keithp.com> 97934 97935commit 2c77aeb39f59f1a94cc603a2e6256a62ce785c36 97936Merge: 816d67de2 402b329c3 97937Author: Keith Packard <keithp@keithp.com> 97938Date: Fri Feb 18 14:29:32 2011 -0800 97939 97940 Merge remote branch 'jcristau/for-keith' 97941 97942commit 816d67de2ba9bdfe652da32cb6447a3056342b98 97943Merge: a73c28f0b 3bbb70a1a 97944Author: Keith Packard <keithp@keithp.com> 97945Date: Fri Feb 18 12:20:26 2011 -0800 97946 97947 Merge remote branch 'whot/for-keith' 97948 97949commit a73c28f0bdafb1c5cb8129179188a99c0ca052e2 97950Author: Justin Dou <Justin.Dou@intel.com> 97951Date: Thu Feb 10 16:27:29 2011 -0500 97952 97953 Replace malloc with calloc to initialize the buffers[] as NULL in do_get_buffers function 97954 97955 The calling for allocate_or_reuse_buffer may fail due to some reason, e.g. out of memory. 97956 If the buffers[] were not initialized to be NULL, the following err_out may try to access an illegal memory, which will cause X crash afterward. 97957 97958 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 97959 Signed-off-by: Justin Dou <Justin.Dou@intel.com> 97960 Signed-off-by: Keith Packard <keithp@keithp.com> 97961 97962commit 0bc9b15a622377c57e617411ccd26c4b5bf8cba4 97963Merge: d1123b66e 93cd4b103 97964Author: Keith Packard <keithp@keithp.com> 97965Date: Fri Feb 18 12:04:41 2011 -0800 97966 97967 Merge remote branch 'ajax/for-keithp' 97968 97969commit d1123b66ebbcf96dd816236f54befc568a5f7c68 97970Merge: 5d020c313 f6e4ace9e 97971Author: Keith Packard <keithp@keithp.com> 97972Date: Fri Feb 18 11:59:25 2011 -0800 97973 97974 Merge remote branch 'jturney/master' 97975 97976commit 649269d40667cfb387cb5286dd3519dd68f7dd80 97977Author: Bryce Harrington <bryce@canonical.com> 97978Date: Wed Feb 16 16:55:57 2011 -0800 97979 97980 Check for OOM condition in XISendDeviceHierarchyEvent 97981 97982 When system is out of memory, calloc can fail returning a NULL pointer. 97983 Check for this before dereferencing it, and bail out if it fails. 97984 97985 Ref.: https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/720445 97986 97987 Signed-off-by: Bryce Harrington <bryce@canonical.com> 97988 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 97989 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 97990 97991commit 31ddb7ef4f5bac2c13c199e74a0716e43b8dc38e 97992Author: Simon Thum <simon.thum@gmx.de> 97993Date: Wed Feb 2 00:03:44 2011 +0100 97994 97995 dix: avoid FP promotion during pointer acceleration 97996 97997 Signed-off-by: Simon Thum <simon.thum@gmx.de> 97998 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 97999 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98000 98001commit 006157f203f8493bb3d18e333a3fd3d6cb10f8ea 98002Author: Simon Thum <simon.thum@gmx.de> 98003Date: Sat Sep 4 16:31:24 2010 +0200 98004 98005 dix: refactor scheme init 98006 98007 This makes it possible to init a scheme in one init call, so we 98008 get rid of the tightly coupled two-phase init used before. 98009 98010 Signed-off-by: Simon Thum <simon.thum@gmx.de> 98011 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98012 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98013 98014commit 38ffeec0c89e83afc62579dec221c325d667cc1e 98015Author: Simon Thum <simon.thum@gmx.de> 98016Date: Sun Feb 6 19:13:00 2011 +0100 98017 98018 xfree86: allow to check for options without warnings in the log 98019 98020 This allows set_percent_option in synaptics to work as described, 98021 and should generally enable to check option syntax without log spam. 98022 98023 Signed-off-by: Simon Thum <simon.thum@gmx.de> 98024 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98025 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98026 98027commit 2fbccc881bfab4bdc5b97f74a7e5bcec7fdeae20 98028Author: Simon Thum <simon.thum@gmx.de> 98029Date: Sun Feb 6 19:07:19 2011 +0100 98030 98031 fix percent options parsing 98032 98033 Signed-off-by: Simon Thum <simon.thum@gmx.de> 98034 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98035 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98036 98037commit 46b49796015762fd131bec96eff6caaa890bfd5f 98038Author: Daniel Stone <daniel@fooishbar.org> 98039Date: Tue Feb 15 11:28:02 2011 +0000 98040 98041 Input: Allow EventToCore to return multiple events 98042 98043 Some event types (notably Expose and GraphicsExpose) require multiple 98044 events, a la XI 1.x. Bring the EventToCore API in line with EventToXI's 98045 and allow it to generate multiple events. 98046 98047 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 98048 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98049 98050commit ea71495adabca52df44241c0eba39e6bceb58202 98051Author: Daniel Stone <daniel@fooishbar.org> 98052Date: Tue Feb 15 11:28:01 2011 +0000 98053 98054 Input: Reduce use of filter in DeliverEvents 98055 98056 Instead of switching on the event filter to determine delivery, use the 98057 event type instead. 98058 98059 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 98060 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98061 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98062 98063commit 8a0241656cecb6ab2fb377f02b14238d18e65fc1 98064Author: Daniel Stone <daniel@fooishbar.org> 98065Date: Tue Feb 15 11:28:00 2011 +0000 98066 98067 Input: Rename EventIsDeliverable return masks 98068 98069 Rename the return mask values for EventIsDeliverable: 98070 * CORE_MASK -> EVENT_CORE_MASK 98071 * XI_MASK -> EVENT_XI1_MASK 98072 * XI2_MASK -> EVENT_XI2_MASK 98073 * DONT_PROPAGATE_MASK -> EVENT_DONT_PROPAGATE_MASK 98074 98075 And don't undef them in dix/events.c, since they're supposed to be 98076 global. 98077 98078 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 98079 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98080 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98081 98082commit 737562257e8ee30b1c438e5160a61fbb26ca609e 98083Author: Daniel Stone <daniel@fooishbar.org> 98084Date: Tue Feb 15 11:27:59 2011 +0000 98085 98086 Input: Actually send Xi 1.x DeviceStateNotify events 98087 98088 When a client has selected for Xi 1.x DeviceStateNotify events, they 98089 should receive them when a DeviceFocusIn event is generated. The code 98090 to do this was there, but an incorrect test meant they were never being 98091 sent. 98092 98093 The "type" passed in is the XI2 type, the XI1 type is in event.type. 98094 98095 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 98096 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98097 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98098 98099commit da39d57a20d4281fc4099e356645fef980290030 98100Author: Daniel Stone <daniel@fooishbar.org> 98101Date: Tue Feb 15 11:27:57 2011 +0000 98102 98103 Input: Store passive-activating key in CheckDeviceGrabs 98104 98105 CheckDeviceGrabs will activate a passive grab for KeyPress and 98106 ButtonPress events. GrabInfoRec::activatingKey contains the keycode 98107 which activated the passive grab, so we can deactivate it later in 98108 ProcessOtherEvents. 98109 98110 Previously, CheckDeviceGrabs relied on its callers to set 98111 activatingKey, which not all callers were doing (I'm looking at you, 98112 ComputeFreezes). Just set it in CheckDeviceGrabs instead. 98113 98114 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 98115 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98116 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98117 98118commit 3bbb70a1a7b24d3d1375b20a13db7011cf961c86 98119Author: Peter Hutterer <peter.hutterer@who-t.net> 98120Date: Wed Feb 16 07:56:58 2011 +1000 98121 98122 xfree86: fix up an out-of-date comment. 98123 98124 InitInput simply initialises all input devices now. 98125 98126 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98127 Reviewed-by: Keith Packard <keithp@keithp.com> 98128 98129commit 402b329c3aa8ddbebaa1f593306a02d4cd6fed26 98130Author: Julien Cristau <jcristau@debian.org> 98131Date: Sun Jan 23 13:35:54 2011 +0100 98132 98133 glx: Work around wrong request lengths sent by mesa 98134 98135 mesa used to send too long requests for GLXDestroyPixmap, 98136 GLXDestroyWindow, GLXChangeDrawableAttributes, GLXGetDrawableAttributes 98137 and GLXGetFBConfigsSGIX. 98138 98139 Fixes a regression introduced in ec9c97c6bf70b523bc500bd3adf62176f1bb33a4 98140 X.Org bug#33324 <https://bugs.freedesktop.org/show_bug.cgi?id=33324> 98141 98142 Reported-by: xunx.fang@intel.com 98143 Signed-off-by: Julien Cristau <jcristau@debian.org> 98144 Reviewed-by: Adam Jackson <ajax@redhat.com> 98145 98146commit 1137c11be0f82049d28024eaf963c6f76e0d4334 98147Author: Julien Cristau <jcristau@debian.org> 98148Date: Wed Jan 26 13:06:53 2011 +0100 98149 98150 glx: fix BindTexImageEXT length check 98151 98152 The request is followed by a list of attributes. 98153 98154 X.Org bug#33449 98155 98156 Reported-and-tested-by: meng <mengmeng.meng@intel.com> 98157 Signed-off-by: Julien Cristau <jcristau@debian.org> 98158 Reviewed-by: Adam Jackson <ajax@redhat.com> 98159 98160commit a883cf1545abd89bb2cadfa659718884b56fd234 98161Author: Julien Cristau <jcristau@debian.org> 98162Date: Sun Jan 23 17:05:26 2011 +0100 98163 98164 glx: fix request length check for CreateGLXPbufferSGIX 98165 98166 The request is followed by an attribute list. 98167 98168 Signed-off-by: Julien Cristau <jcristau@debian.org> 98169 Reviewed-by: Adam Jackson <ajax@redhat.com> 98170 98171commit 0f9c6f2f822ff53b9d12ff4fa0b26cbeb7394ba5 98172Author: Alan Coopersmith <alan.coopersmith@oracle.com> 98173Date: Sun Feb 13 21:36:05 2011 -0800 98174 98175 xkb: Replace malloc(strlen) + strcpy with strdup 98176 98177 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 98178 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98179 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 98180 98181commit 682865c460945e1299f943561140f46439e2b4cb 98182Author: Alan Coopersmith <alan.coopersmith@oracle.com> 98183Date: Sun Feb 13 21:36:04 2011 -0800 98184 98185 XkbAddGeomProperty: Fix checks for malloc failure 98186 98187 Check the variable we just tried to malloc, not the string we're copying 98188 and already checked for NULL at the beginning of the function. 98189 98190 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 98191 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98192 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 98193 98194commit aac1b435664819008989ed19f73e9c89920602c5 98195Author: Alan Coopersmith <alan.coopersmith@oracle.com> 98196Date: Sun Feb 13 21:36:03 2011 -0800 98197 98198 Replace _XkbDupString with Xstrdup 98199 98200 The two functions have identical semantics, including safely returning 98201 NULL when NULL is passed in (which POSIX strdup does not guarantee). 98202 98203 Some callers could probably be adjusted to call libc strdup directly, 98204 when we know the input is non-NULL. 98205 98206 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 98207 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98208 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 98209 98210commit a4a2e814d5d0e6152307a301eda1d6fc1c555aaa 98211Author: Alan Coopersmith <alan.coopersmith@oracle.com> 98212Date: Sun Feb 13 21:36:02 2011 -0800 98213 98214 xkb: Use snprintf to measure string lengths instead of manual strlen math 98215 98216 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 98217 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98218 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 98219 98220commit c9f7b303a36ca501c6ecf1196c266ee8e8f49d2d 98221Author: Peter Hutterer <peter.hutterer@who-t.net> 98222Date: Fri Feb 11 13:50:10 2011 +1000 98223 98224 xfixes: calloc, not malloc the cursorScreenRec 98225 98226 Debugging NULL pointers is significantly easier than random memory. 98227 Plus, if new fields (such as pointer barriers) are added they may just be 98228 properly initialised. 98229 98230 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98231 Reviewed-by: Keith Packard <keithp@keithp.com> 98232 98233commit d3499556d8d83396fa2585bd00371a81e086be36 98234Author: Peter Hutterer <peter.hutterer@who-t.net> 98235Date: Thu Feb 10 15:12:14 2011 +1000 98236 98237 xkb: if the keymap failed to compile, load the default keymap instead. 98238 98239 We really need symbols, compat, keynames, vmods and types for a sensible keymap. 98240 98241 Try this in your xorg.conf.d snippets for all keyboards: 98242 Option "XkbLayout" "us" 98243 Option "XkbVariant" "nodeadkeys" 98244 98245 us(nodeadkeys) doesn't exist so xkbcomp provides everything but the symbols 98246 map. We say we want everything but don't _need_ anything, the server happily 98247 gives us a keymap with every key mapped to NoSymbol. This in turn isn't what 98248 we want after all. 98249 98250 So instead, require symbols, compat, keynames, vmods and types from the 98251 keymap and if that fails, load the default keymap instead. If that fails 98252 too, all bets are off. 98253 98254 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98255 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 98256 98257commit 47d1d2fed656c3a3b2600491078da90962c46934 98258Author: Peter Hutterer <peter.hutterer@who-t.net> 98259Date: Thu Feb 10 15:11:34 2011 +1000 98260 98261 xkb: split out keymap compilation. 98262 98263 Refactoring for simpler double-use in the next patch. No functional changes. 98264 98265 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98266 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 98267 98268commit 787ba25a8a3af52b38448a1a6f8c9704ea8b7905 98269Author: Carlos Garnacho <carlosg@gnome.org> 98270Date: Mon Feb 7 18:21:31 2011 +0100 98271 98272 Xi: make XIQueryPointer return the current modifiers/group as documented. 98273 98274 The previous XKB info was being returned instead of the current 98275 one, producing inconsistent results between the latest events 98276 and the modifiers/group returned by this call. 98277 98278 Signed-off-by: Carlos Garnacho <carlosg@gnome.org> 98279 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 98280 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net>` 98281 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98282 98283commit 85f9017393c9bb19553e9afcf554673a44a09993 98284Author: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 98285Date: Tue Feb 8 11:10:10 2011 +0200 98286 98287 ProcXkbGetXkbByName: fix use of uninitialised bytes valgrind error. 98288 98289 ==9999== Syscall param writev(vector[...]) points to uninitialised byte(s) 98290 ==9999== at 0x4AB5154: writev (writev.c:51) 98291 ==9999== by 0x7C7C3: _XSERVTransWritev (Xtrans.c:912) 98292 ==9999== by 0x61C8B: FlushClient (io.c:924) 98293 ==9999== by 0x62423: WriteToClient (io.c:846) 98294 ==9999== by 0xCE39B: XkbSendMap (xkb.c:1408) 98295 ==9999== by 0xD247B: ProcXkbGetKbdByName (xkb.c:5814) 98296 ==9999== by 0x4AB53: Dispatch (dispatch.c:432) 98297 ==9999== by 0x205BF: main (main.c:291) 98298 ==9999== Address 0x557eb68 is 40 bytes inside a block of size 4,096 alloc'd 98299 ==9999== at 0x48334A4: calloc (vg_replace_malloc.c:467) 98300 ==9999== by 0x62567: WriteToClient (io.c:1065) 98301 ==9999== by 0x452EB: ProcEstablishConnection (dispatch.c:3685) 98302 ==9999== by 0x4AB53: Dispatch (dispatch.c:432) 98303 ==9999== by 0x205BF: main (main.c:291) 98304 ==9999== Uninitialised value was created by a stack allocation 98305 ==9999== at 0xD1910: ProcXkbGetKbdByName (xkb.c:5559) 98306 98307 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98308 Reviewed-by: Oliver McFadden <oliver.mcfadden@nokia.com> 98309 Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 98310 98311commit 8a34d7a8532c7ca013e67307f3baf200167abb92 98312Author: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 98313Date: Tue Feb 8 11:10:11 2011 +0200 98314 98315 XkbSendNames: fix use of uninitialised bytes valgrind error. 98316 98317 ==537== Syscall param writev(vector[...]) points to uninitialised byte(s) 98318 ==537== at 0x4AB7154: writev (writev.c:51) 98319 ==537== by 0x8935B: _XSERVTransWritev (Xtrans.c:912) 98320 ==537== by 0x6C55F: FlushClient (io.c:924) 98321 ==537== by 0x6CCF3: WriteToClient (io.c:846) 98322 ==537== by 0xD51D3: XkbSendNames (xkb.c:3765) 98323 ==537== by 0xD8183: ProcXkbGetKbdByName (xkb.c:5825) 98324 ==537== by 0x27B7B: Dispatch (dispatch.c:432) 98325 ==537== by 0x205B7: main (main.c:291) 98326 ==537== Address 0x55899f2 is 154 bytes inside a block of size 1,896 alloc'd 98327 ==537== at 0x4834C48: malloc (vg_replace_malloc.c:236) 98328 ==537== by 0xD47AF: XkbSendNames (xkb.c:3642) 98329 ==537== by 0xD8183: ProcXkbGetKbdByName (xkb.c:5825) 98330 ==537== by 0x27B7B: Dispatch (dispatch.c:432) 98331 ==537== by 0x205B7: main (main.c:291) 98332 ==537== Uninitialised value was created by a heap allocation 98333 ==537== at 0x4834C48: malloc (vg_replace_malloc.c:236) 98334 ==537== by 0xD47AF: XkbSendNames (xkb.c:3642) 98335 ==537== by 0xD8183: ProcXkbGetKbdByName (xkb.c:5825) 98336 ==537== by 0x27B7B: Dispatch (dispatch.c:432) 98337 ==537== by 0x205B7: main (main.c:291) 98338 98339 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98340 Reviewed-by: Oliver McFadden <oliver.mcfadden@nokia.com> 98341 Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 98342 98343commit 87fbef9157a6f1e1318382e368d27942d7ad72ab 98344Author: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 98345Date: Tue Feb 8 11:10:09 2011 +0200 98346 98347 ProcRRCreateMode: fix use of uninitialised bytes valgrind error. 98348 98349 ==543== Syscall param writev(vector[...]) points to uninitialised byte(s) 98350 ==543== at 0x4AB7154: writev (writev.c:51) 98351 ==543== by 0x8935B: _XSERVTransWritev (Xtrans.c:912) 98352 ==543== by 0x6C55F: FlushClient (io.c:924) 98353 ==543== by 0x6D013: FlushAllOutput (io.c:668) 98354 ==543== by 0x27A83: Dispatch (dispatch.c:453) 98355 ==543== by 0x205B7: main (main.c:291) 98356 ==543== Address 0x556dc8c is 12 bytes inside a block of size 4,096 alloc'd 98357 ==543== at 0x48334A4: calloc (vg_replace_malloc.c:467) 98358 ==543== by 0x6CE37: WriteToClient (io.c:1065) 98359 ==543== by 0x223A7: ProcEstablishConnection (dispatch.c:3685) 98360 ==543== by 0x27B7B: Dispatch (dispatch.c:432) 98361 ==543== by 0x205B7: main (main.c:291) 98362 ==543== Uninitialised value was created by a stack allocation 98363 ==543== at 0xA3350: ProcRRCreateMode (rrmode.c:289) 98364 98365 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98366 Reviewed-by: Oliver McFadden <oliver.mcfadden@nokia.com> 98367 Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 98368 98369commit 0ef5973860e17c5edc996c923610f7ad88b4dfbe 98370Author: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 98371Date: Tue Feb 8 11:10:08 2011 +0200 98372 98373 ProcRRQueryVersion: fix use of uninitialised bytes valgrind error. 98374 98375 ==9999== Syscall param writev(vector[...]) points to uninitialised byte(s) 98376 ==9999== at 0x4AB5154: writev (writev.c:51) 98377 ==9999== by 0x7C7C3: _XSERVTransWritev (Xtrans.c:912) 98378 ==9999== by 0x61C8B: FlushClient (io.c:924) 98379 ==9999== by 0x62743: FlushAllOutput (io.c:668) 98380 ==9999== by 0x4AA5B: Dispatch (dispatch.c:453) 98381 ==9999== by 0x205BF: main (main.c:291) 98382 ==9999== Address 0x55711b9 is 1 bytes inside a block of size 4,096 alloc'd 98383 ==9999== at 0x48334A4: calloc (vg_replace_malloc.c:467) 98384 ==9999== by 0x62567: WriteToClient (io.c:1065) 98385 ==9999== by 0x452EB: ProcEstablishConnection (dispatch.c:3685) 98386 ==9999== by 0x4AB53: Dispatch (dispatch.c:432) 98387 ==9999== by 0x205BF: main (main.c:291) 98388 ==9999== Uninitialised value was created by a stack allocation 98389 ==9999== at 0x160E78: ProcRRQueryVersion (rrdispatch.c:37) 98390 98391 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98392 Reviewed-by: Oliver McFadden <oliver.mcfadden@nokia.com> 98393 Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 98394 98395commit 81daba8ce906bfbbe44cd71d0ff269ad34e2f6b5 98396Author: Peter Hutterer <peter.hutterer@who-t.net> 98397Date: Mon Jan 31 13:53:08 2011 +1000 98398 98399 Xi: constify XIChangeDeviceProperty() 98400 98401 We don't modify "value", make it official. 98402 98403 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98404 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 98405 Signed-off-by: Ander Conselvan de Oliveira <ander.conselvan-de-oliveira@nokia.com> 98406 98407commit 93cd4b1034e3c8bed32c5acd391eadd7605e10f0 98408Author: Adam Jackson <ajax@redhat.com> 98409Date: Thu Jan 20 00:27:00 2011 -0500 98410 98411 shadow: Remove hw/xfree86/ from includes 98412 98413 Reviewed-by: Matt Turner <mattst88@gmail.com> 98414 Reviewed-by: Julien Cristau <jcristau@debian.org> 98415 Signed-off-by: Adam Jackson <ajax@redhat.com> 98416 98417commit 0c230728f3af2c4250e93a070c0851a597069447 98418Author: Adam Jackson <ajax@redhat.com> 98419Date: Thu Jan 20 00:25:57 2011 -0500 98420 98421 rootless: Remove hw/xfree86/ from includes 98422 98423 Reviewed-by: Matt Turner <mattst88@gmail.com> 98424 Reviewed-by: Julien Cristau <jcristau@debian.org> 98425 Signed-off-by: Adam Jackson <ajax@redhat.com> 98426 98427commit 505defd270e49cfbcbe6a04d1de817d305edb3c2 98428Author: Adam Jackson <ajax@redhat.com> 98429Date: Thu Jan 20 00:24:20 2011 -0500 98430 98431 damage: Remove hw/xfree86/ from includes 98432 98433 Reviewed-by: Matt Turner <mattst88@gmail.com> 98434 Reviewed-by: Julien Cristau <jcristau@debian.org> 98435 Signed-off-by: Adam Jackson <ajax@redhat.com> 98436 98437commit 59d8e3cbdba032c7354c28c86cbd155e9da6447a 98438Author: Adam Jackson <ajax@redhat.com> 98439Date: Thu Jan 20 00:23:07 2011 -0500 98440 98441 fb: Remove hw/xfree86/ from includes 98442 98443 Reviewed-by: Matt Turner <mattst88@gmail.com> 98444 Reviewed-by: Julien Cristau <jcristau@debian.org> 98445 Signed-off-by: Adam Jackson <ajax@redhat.com> 98446 98447commit 994b7b3dc97f9afa713fe636af45a78002a4366f 98448Author: Adam Jackson <ajax@redhat.com> 98449Date: Thu Jan 20 00:17:52 2011 -0500 98450 98451 xfree86: warning fix 98452 98453 Did you know that anonymous enums with function scope will not only 98454 override the enum values from global scope, but will be treated as 98455 entirely different types? C's type system just rules. 98456 98457 xf86Crtc.c: In function 'handle_detailed_monrec': 98458 xf86Crtc.c:1555:33: warning: comparison between 'enum det_monrec_source' and 'enum <anonymous>' 98459 xf86Crtc.c:1562:33: warning: comparison between 'enum det_monrec_source' and 'enum <anonymous>' 98460 98461 Reviewed-by: Matt Turner <mattst88@gmail.com> 98462 Reviewed-by: Julien Cristau <jcristau@debian.org> 98463 Signed-off-by: Adam Jackson <ajax@redhat.com> 98464 98465commit ae16c5796fdb22fb6b2f680fe5123bfd2c89a825 98466Author: Adam Jackson <ajax@redhat.com> 98467Date: Thu Jan 20 00:14:11 2011 -0500 98468 98469 fbdevhw: Remove the non-sysfs hack for mapping from PCI to fbdev 98470 98471 It's broken for devices with BARs above 4G, and the sysfs method should 98472 work everywhere anyway. As a pleasant side effect, this fixes some 98473 warnings: 98474 98475 fbdevhw.c: In function 'fbdev_open_pci': 98476 fbdevhw.c:333:4: warning: cast from pointer to integer of different size 98477 fbdevhw.c:334:4: warning: cast from pointer to integer of different size 98478 fbdevhw.c:336:4: warning: cast from pointer to integer of different size 98479 fbdevhw.c:337:4: warning: cast from pointer to integer of different size 98480 98481 Reviewed-by: Matt Turner <mattst88@gmail.com> 98482 Reviewed-by: Julien Cristau <jcristau@debian.org> 98483 Signed-off-by: Adam Jackson <ajax@redhat.com> 98484 98485commit f1b7c9cead94b520e6b96774d605d63d3f341b50 98486Author: Adam Jackson <ajax@redhat.com> 98487Date: Thu Jan 20 00:09:19 2011 -0500 98488 98489 int10: warning fix 98490 98491 helper_exec.c: In function 'pciCfg1in': 98492 helper_exec.c:507:4: warning: passing argument 2 of 'pci_device_cfg_read_u32' from incompatible pointer type 98493 /usr/include/pciaccess.h:153:5: note: expected 'uint32_t *' but argument is of type 'CARD32 *' 98494 98495 Reviewed-by: Matt Turner <mattst88@gmail.com> 98496 Reviewed-by: Julien Cristau <jcristau@debian.org> 98497 Signed-off-by: Adam Jackson <ajax@redhat.com> 98498 98499commit be186586e504197623c69637b8122fb814f59429 98500Author: Adam Jackson <ajax@redhat.com> 98501Date: Thu Jan 20 00:06:48 2011 -0500 98502 98503 xf86vidmode: warning fix 98504 98505 xf86VidMode.c: In function 'VidModeGetMonitorValue': 98506 xf86VidMode.c:637:19: warning: 'ret.i' may be used uninitialized in this function 98507 98508 Reviewed-by: Matt Turner <mattst88@gmail.com> 98509 Reviewed-by: Julien Cristau <jcristau@debian.org> 98510 Signed-off-by: Adam Jackson <ajax@redhat.com> 98511 98512commit f6e4ace9eabea1bb2a06dd86b6ffb954917a74ce 98513Author: Jon TURNEY <jon.turney@dronecode.org.uk> 98514Date: Sun Jul 18 13:24:48 2010 +0100 98515 98516 Cygwin/X: Fix for Canadian Multilingual Standard keyboard layout handling 98517 98518 Add a keycode mapping for VK_OEM_8 as RCtrl, which is issued by Canadian 98519 Multilingual Standard layout 98520 98521 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 98522 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 98523 98524commit b421f5e5122dbe125a5629969cc657bd966b4261 98525Author: Jon TURNEY <jon.turney@dronecode.org.uk> 98526Date: Thu Apr 1 23:10:58 2010 +0100 98527 98528 Cygwin/X: Ignore MappingNotify events sent to clipboard integration client 98529 98530 Ignore MappingNotify events sent to clipboard integration client, 98531 xmodmap changes aren't of interest to it, but there is no mechanism 98532 to express that disinterest. 98533 98534 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 98535 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 98536 98537commit e930aa7bbc7f04a128df8e6fd441aed5123a2eb6 98538Author: Jon TURNEY <jon.turney@dronecode.org.uk> 98539Date: Tue Sep 14 16:02:53 2010 +0100 98540 98541 Cygwin/X: Add Turkish keyboard layouts to keyboard layout mapping table 98542 98543 0x0000041f "Turkish Q" => layout tr 98544 0x0001041f "Turkish F" => layout tr variant f 98545 98546 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 98547 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 98548 98549commit ae99a6f895a3a5b6f71e2dd4874cd4bb0235aff0 98550Author: Jon TURNEY <jon.turney@dronecode.org.uk> 98551Date: Sun Jul 18 13:42:29 2010 +0100 98552 98553 Cygwin/X: Make the keyboard layout detection logging a bit clearer 98554 98555 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 98556 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 98557 98558commit 57ca09441ee5ff2a8deb95eaa17a2625a756d912 98559Author: Jon TURNEY <jon.turney@dronecode.org.uk> 98560Date: Thu Jul 1 14:54:48 2010 +0100 98561 98562 Cygwin/X: Update mapping for Canadian keyboard layouts 98563 98564 0x00000c0c "Canadian French (legacy)" => layout ca variant fr-legacy 98565 0x00001009 "Canadian French" => layout ca variant fr 98566 0x00011009 "Canadian Multilingual Standard" => layout ca variant multix 98567 98568 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 98569 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 98570 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 98571 98572commit be3e5bb50178406e46368be00860331e1b6c4093 98573Author: Jon TURNEY <jon.turney@dronecode.org.uk> 98574Date: Wed Apr 14 21:45:29 2010 +0100 98575 98576 Cygwin/X: Better keycode debugging output 98577 98578 Replace useless #if 0/ErrorF/#endif with winDebug 98579 98580 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 98581 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 98582 98583commit 81892bf6b7e3730ebd19318183734f55dbaa5d4b 98584Author: Jon TURNEY <jon.turney@dronecode.org.uk> 98585Date: Wed Apr 14 21:34:39 2010 +0100 98586 98587 Cygwin/X: Generate X keycodes for multimedia keys 98588 98589 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 98590 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 98591 98592commit 5d020c31351594c104fb46615ce41f256cd8e8b9 98593Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98594Date: Fri Jan 14 15:03:28 2011 +0200 98595 98596 xf86/xv: Only register PostValidateTree hook when there is work to do 98597 98598 If none of Xv ports were affected by window tree modifications we don't 98599 want scan the port list. To avoid useless scanning of port list 98600 PostValidateTree hook is only registered when ClipNotify was called for 98601 any port. 98602 98603 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98604 Reviewed-by: Ville Syrjälä <ville.syrjala@nokia.com> 98605 98606commit ddf32df5cc2e73b833f6e414276d938fdc38869e 98607Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98608Date: Thu Jan 13 20:50:16 2011 +0200 98609 98610 xf86/xv: Use PostValidateTree to do reput 98611 98612 ValidateTree calls first ClipNotify and later might call 98613 WindowExposures. To avoid useless double reput ClipNotify delays reput 98614 to WindowExposures or PostValidateTree. 98615 98616 PostValidatTree checks all ports if there is clip changes. On clip 98617 changes reput is done to move or scale the overlay. 98618 98619 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98620 Reviewed-by: Ville Syrjälä <ville.syrjala@nokia.com> 98621 98622commit 5678a41f6bc0ef8965bd50a88c5ef01ba7eb8b24 98623Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98624Date: Mon Jan 17 16:58:00 2011 +0200 98625 98626 xf86/xv: Fill color key on expose 98627 98628 If window gets exposed but clipboxes doesn't change drivers would avoid 98629 color key fill. This makes XResizeWindo&co to lose colorkey if 98630 background is painted. 98631 98632 To help drivers to avoid filling colorkey for each put server can 98633 provide helper function if there is exposed areas. Server can subtract 98634 exposed areas from filled region. 98635 98636 As a side effect we can avoid useless color key fills if window only 98637 moves in screen without background fills. 98638 98639 v3: 98640 * Change tracking to filled area to account for client initiated clip 98641 changes 98642 * Make overlaid XvPutImage behavior like textured XvPutImage or PutImage 98643 * Make region dynamically allocated only when required. 98644 98645 v4: 98646 * Simplify new driver interface to reduce duplicate code 98647 98648 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98649 Reviewed-by: Ville Syrjälä <ville.syrjala@nokia.com> 98650 98651commit 9b048ca420864e19862f63f4a491a4da0ff9b54f 98652Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98653Date: Mon Jan 17 15:07:28 2011 +0200 98654 98655 xf86/xv: Remove unused variable from XvPortRecPrivate 98656 98657 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98658 Reviewed-by: Ville Syrjälä <ville.syrjala@nokia.com> 98659 98660commit 4f8f803280894a8c41691e350b7a34bfd719fe05 98661Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98662Date: Thu Jan 13 20:21:09 2011 +0200 98663 98664 xf86/xv: Remove unused GC pointers 98665 98666 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98667 Reviewed-by: Ville Syrjälä <ville.syrjala@nokia.com> 98668 98669commit e012f2312b4bcdf31bb5d622ddd34571f3c12429 98670Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98671Date: Thu Jan 13 20:19:18 2011 +0200 98672 98673 xf86/xv: Remove copy paste code. 98674 98675 xf86XVFillKeyHelperDrawable can be used to implement 98676 xf86XVFillKeyHelper. 98677 98678 V2: 98679 * Remove RegionTranslate that clobbered parameter region. 98680 98681 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 98682 Reviewed-by: Ville Syrjälä <ville.syrjala@nokia.com> 98683 98684commit ea1ffd3e60bdcedbec5a6f28929f8677bf45d450 98685Merge: 12b0f7df2 53602c362 98686Author: Keith Packard <keithp@keithp.com> 98687Date: Wed Feb 2 15:19:55 2011 -0800 98688 98689 Merge remote branch 'whot/for-keith' 98690 98691commit 12b0f7df2caa78d68b453aaa91248bb3577fd724 98692Author: Evan Broder <ebroder@mokafive.com> 98693Date: Tue Feb 1 16:09:50 2011 -0800 98694 98695 xfree86/modes: Be sure to only use new EDID for physical output dimensions 98696 98697 The EDID processing regards physical dimensions of 0mm x 0mm as 98698 invalid. Previously the old values for height and width would be 98699 preserved if none of the physical dimension specifications in the new 98700 EDID were considered valid. 98701 98702 This will come up in particular if first a monitor is connected to an 98703 output, and then a projector is connected. Since projectors generally 98704 report physical dimensions of 0mm x 0mm, this would result in the 98705 projector claiming to have the physical dimensions of the monitor. 98706 98707 Signed-off-by: Evan Broder <ebroder@mokafive.com> 98708 Reviewed-by: Adam Jackson <ajax@redhat.com> 98709 Signed-off-by: Keith Packard <keithp@keithp.com> 98710 98711commit 0b15033031d78836582bd1b641c3160c3803a708 98712Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 98713Date: Mon Jan 31 23:25:10 2011 +0100 98714 98715 xserver: enable TLS even if AIGLX is not enabled 98716 98717 This aligns the xorg server build with the mesa build, which is needed on 98718 systems where aiglx with dri support is not enabled. Else the following error is 98719 obtained when trying to load the software raster: 98720 98721 (EE) AIGLX error: dlopen of /usr/lib/dri/swrast_dri.so failed (/usr/lib/dri/swrast_dri.so: undefined symbol: _glapi_tls_Context) 98722 (EE) GLX: could not load software renderer 98723 (II) GLX: no usable GL providers found for screen 0 98724 98725 because mesa always enables TLS use in GLX, even if dri is not available. 98726 98727 Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 98728 Signed-off-by: Julien Cristau <jcristau@debian.org> 98729 Signed-off-by: Keith Packard <keithp@keithp.com> 98730 98731commit 541b25038a5de74411a094570b407c5ae018c2ba 98732Author: Maarten Maathuis <madman2003@gmail.com> 98733Date: Sun Jan 30 14:06:16 2011 +0100 98734 98735 exa/mixed: Exclude frontbuffer from deferred pixmap handling. 98736 98737 - Apps like xterm can trigger a lot of fallback rendering. 98738 - This can lead to (annoyingly) high latencies, because you 98739 have to wait for the block handler. 98740 - You need a driver that doesn't directly access the front 98741 buffer to trigger this (NV50+ nouveau for example). 98742 - Repeatingly doing dmesg on an xterm with a bitmap font 98743 will reveal that you never see part of the text. 98744 - I have recieved at least one complaint in the past of slow 98745 terminal performance, which was related to core font 98746 rendering. 98747 - This does sacrifice some throughput, roughly 33% slower. 98748 98749 Reviewed-by: Michel Dänzer <michel@daenzer.net> 98750 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 98751 Signed-off-by: Keith Packard <keithp@keithp.com> 98752 98753commit af87f6367ef733d1a4f3cfca4eeb92bfd84c2b6f 98754Author: Maarten Maathuis <madman2003@gmail.com> 98755Date: Sun Jan 30 14:06:15 2011 +0100 98756 98757 exa: Only call driver FinishAccess hook if PrepareAccess hook succeeded. 98758 98759 Reviewed-by: Michel Dänzer <michel@daenzer.net> 98760 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 98761 Signed-off-by: Keith Packard <keithp@keithp.com> 98762 98763commit 648d4fe5172cad33c6f09e05dd3d8d7171ef1a7f 98764Author: Maarten Maathuis <madman2003@gmail.com> 98765Date: Sun Jan 30 14:06:14 2011 +0100 98766 98767 exa/driver: set pExaPixmap->use_gpu_copy to the right value 98768 98769 - Not sure if it was causing problems, but you never know. 98770 98771 Reviewed-by: Michel Dänzer <michel@daenzer.net> 98772 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 98773 Signed-off-by: Keith Packard <keithp@keithp.com> 98774 98775commit 246d40b685292fa6393ba22127152137494af2af 98776Merge: 70f2d39fb 8c1f75a1d 98777Author: Keith Packard <keithp@keithp.com> 98778Date: Wed Feb 2 14:51:33 2011 -0800 98779 98780 Merge remote branch 'jeremyhu/master' 98781 98782commit 70f2d39fbde38d0baa687f1755b6c4c686f23052 98783Merge: 3a91c2d21 a9d3c43a5 98784Author: Keith Packard <keithp@keithp.com> 98785Date: Wed Feb 2 14:43:19 2011 -0800 98786 98787 Merge remote branch 'jturney/master' 98788 98789commit 3a91c2d21c3231e2a011d95c3cfc01785430c1cf 98790Merge: ac0a00a84 82b1eaa6c 98791Author: Keith Packard <keithp@keithp.com> 98792Date: Wed Feb 2 14:37:00 2011 -0800 98793 98794 Merge remote branch 'dbn/build-fixes' 98795 98796commit ac0a00a840a5d8d8374e5db2b849e5959550ca4e 98797Author: Adam Jackson <ajax@redhat.com> 98798Date: Thu Jan 20 23:01:03 2011 -0500 98799 98800 xselinux: Fix GetDrawableContext 98801 98802 M_DRAWABLE_PIXMAP is the lookup mask to dixLookupDrawable, and _not_ the 98803 type value in the drawable itself. 98804 98805 Signed-off-by: Adam Jackson <ajax@redhat.com> 98806 Reviewed-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 98807 Signed-off-by: Keith Packard <keithp@keithp.com> 98808 98809commit 8c1f75a1d4cb5a34b4d10f3f5068bccc4356f31a 98810Author: Jeremy Huddleston <jeremyhu@apple.com> 98811Date: Fri Jan 28 21:13:38 2011 -0800 98812 98813 XQuartz: Localization Updates 98814 98815 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 98816 98817commit 53602c36217487e0a80880e8328f89ab302d85c8 98818Author: Ville Syrjala <syrjala@sci.fi> 98819Date: Mon Jan 24 01:06:49 2011 +0200 98820 98821 xfree86/dga: Remove DGAIsDgaEvent() 98822 98823 DGAIsDgaEvent() is not used anymore. 98824 98825 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98826 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98827 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98828 98829commit 2081728b7432e949c274e0dbbabc296fa4db17f9 98830Author: Ville Syrjala <syrjala@sci.fi> 98831Date: Mon Jan 24 01:06:48 2011 +0200 98832 98833 xfree86/dga: Only send DGA events for master devices 98834 98835 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98836 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98837 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98838 98839commit 9f09b62e594321a8e45a33f4f13559827d660ed3 98840Author: Ville Syrjala <syrjala@sci.fi> 98841Date: Mon Jan 24 01:06:47 2011 +0200 98842 98843 xfree86/dga: Remove useless ifdefs 98844 98845 Move some variables to the scope where they are used. 98846 98847 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98848 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98849 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98850 98851commit 5c941c04ca68cdc2b236c524970bc23f91906af4 98852Author: Ville Syrjala <syrjala@sci.fi> 98853Date: Mon Jan 24 01:06:46 2011 +0200 98854 98855 xfree86/dga: DGAProcessKeyboardEvent() forgot ET_Internal 98856 98857 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98858 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98859 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98860 98861commit ce4e0f660beb097e90110c97d039981702f0b0bd 98862Author: Ville Syrjala <syrjala@sci.fi> 98863Date: Mon Jan 24 01:06:45 2011 +0200 98864 98865 xfree86/dga: Remove DGAMouseX and DGAMouseY 98866 98867 Previously some sort of absolute coordinates were sent out in 98868 the padding of the DGA2 Motion and Button events. DGAMouseX 98869 and DGAMouseY were used to keep track of said coordinates. 98870 libXxf86dga doesn't use that data for anything, and at least 98871 git history didn't show any past usage either. So let's just 98872 remove the last remnants of of this mess. 98873 98874 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98875 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98876 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98877 98878commit ee8faeadb60f51a36edce19e568600dab7c8b357 98879Author: Ville Syrjala <syrjala@sci.fi> 98880Date: Mon Jan 24 01:06:44 2011 +0200 98881 98882 xfree86/dga: DGA2 events are missing the dx/dy information 98883 98884 Copy dx/dy from the internal event to the DGA2 Motion/Button events. 98885 Do the same for Key events for the sake of keeping the code consistent. 98886 98887 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98888 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98889 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98890 98891commit e3fef815cf85cf59256f3edeaa21c1267898b42c 98892Author: Ville Syrjala <syrjala@sci.fi> 98893Date: Mon Jan 24 01:06:43 2011 +0200 98894 98895 xfree86/dga: Make mieq_install Bool 98896 98897 mieq_installed is used as a boolean, so why not make it such. Also 98898 it's a static variable, so the the explicit zero initialization can 98899 be removed. 98900 98901 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98902 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98903 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98904 98905commit f4cb25c2ed832bf4cfcd4d25198fe5a895430fed 98906Author: Ville Syrjala <syrjala@sci.fi> 98907Date: Mon Jan 24 01:06:42 2011 +0200 98908 98909 xfree86/dga: Remove the ET_DGAEvent handler only if it was installed 98910 98911 Remove the handler only if it was installed. Also mark it as 98912 uninstalled, otherwise it wouldn't get reinstalled after a 98913 server reset. 98914 98915 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98916 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98917 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98918 98919commit e1547a9f1f6d7991219b7711f4cacc2852dbf88c 98920Author: Ville Syrjala <syrjala@sci.fi> 98921Date: Mon Jan 24 01:06:41 2011 +0200 98922 98923 xfree86/dga: Install the DGA event handler when DGA2 is used 98924 98925 The ET_DGAEvent handler is only installed when a client 98926 requests relative events via DGA1. Do it also when a client 98927 requests DGA2 events. 98928 98929 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98930 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98931 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98932 98933commit 7b16e8633d5a95167aac0c19e43fcfe38bd4a6a4 98934Author: Ville Syrjala <syrjala@sci.fi> 98935Date: Mon Jan 24 01:06:40 2011 +0200 98936 98937 xfree86/input: Restore DGAStealKeyEvent() 98938 98939 DGA key event support was lost in commit 98940 8da0ff2d51086666d10ca7330d428e8610a4a0e3. Bring it back. 98941 98942 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98943 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98944 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98945 98946commit a2c481821997ae14aa5816895ae36db2ca2e4dec 98947Author: Ville Syrjala <syrjala@sci.fi> 98948Date: Mon Jan 24 01:06:39 2011 +0200 98949 98950 include: Fix a copy/paste error in a comment 98951 98952 Signed-off-by: Ville Syrjala <syrjala@sci.fi> 98953 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98954 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 98955 98956commit 77af0dc5217fc9c446097d6d487c8ddaf00bdd2a 98957Author: Peter Hutterer <peter.hutterer@who-t.net> 98958Date: Mon Jan 24 10:06:34 2011 +1000 98959 98960 xfree86: bump input minor ABI 98961 98962 The new valuator_mask_free() call requires a minor bump. 98963 98964 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98965 98966commit 5927e070b45fbe3fea4b394920ffd6f7d036f1ed 98967Author: Erkki Seppälä <erkki.seppala@vincit.fi> 98968Date: Fri Jan 21 15:56:22 2011 +0200 98969 98970 xkb: Cancel a key's repetition when its autorepeat is disabled. 98971 98972 When XkbChangeEnabledControls is called to disable key repetition of a 98973 certain key (or keys), currently ongoing repetition of that key was 98974 not cancelled. It was cancelled if ChangeKeyboardControl was used to 98975 disable key repetition globally. 98976 98977 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 98978 Reviewed-by: Dirk Wallenstein <halsmit@t-online.de> 98979 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 98980 98981commit a9d3c43a57cbaf42d9e96ad49cf12bbf1d275fe5 98982Author: Jon TURNEY <jon.turney@dronecode.org.uk> 98983Date: Wed Jan 19 15:57:59 2011 +0000 98984 98985 Fix compilation of xf86bigfont.c 98986 98987 xf86bigfont.c: In function 'XFree86BigfontExtensionInit': 98988 xf86bigfont.c:146: error: 'ProcXF86BigfontDispatch' undeclared (first use in this function) 98989 xf86bigfont.c:147: error: 'SProcXF86BigfontDispatch' undeclared (first use in this function) 98990 98991 It seems this has been broken since commit cbd4d5dbb70db62ba1cb79c7b904e6fa11f62d7e 98992 "delete pervasively use of DISPATCH_PROC" (2010-09-28), which is a bit worrying as 98993 that presumably indicates that no tinderbox is configuring with --enable-xf86bigfont. 98994 98995 In a similar fashion to that commit, fix by moving XFree86BigfontExtensionInit() 98996 below the definitions of the static dispatch functions it references. 98997 98998 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 98999 Reviewed-by: Julien Cristau <jcristau@debian.org> 99000 99001commit effaf9459c450f13e35bf263835d3e37d48ff29d 99002Author: Jeremy Huddleston <jeremyhu@apple.com> 99003Date: Sun Jan 23 22:35:34 2011 -0800 99004 99005 XQuartz: Use the default signal handler 99006 99007 This allows better interaction with CrashTracer 99008 99009 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 99010 99011commit 418bb57a3923311338baa8a85c0794a72e15ceeb 99012Author: Jeremy Huddleston <jeremyhu@apple.com> 99013Date: Thu Jan 13 17:12:17 2011 -0800 99014 99015 XQuartz: Call RRScreenSizeNotify after handling externally-generated XP_EVENT_DISPLAY_CHANGED 99016 99017 We get an XP_EVENT_DISPLAY_CHANGED event when our display configuration is 99018 changed. If this change was caused by hotplugging a monitor or Mac Display 99019 Preferences changes by the user, we need to call RRScreenSizeNotify in order 99020 to ensure new connections get the correct screen size. 99021 99022 http://xquartz.macosforge.org/trac/ticket/460 99023 99024 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 99025 99026commit 6f52b10b9e8068e3deafb2bbb64b62175a208c49 99027Author: Jeremy Huddleston <jeremyhu@apple.com> 99028Date: Thu Jan 13 16:08:46 2011 -0800 99029 99030 XQuartz: Bump version string to 2.6.1 99031 99032 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 99033 99034commit 1cfb96dd6151ebe84a9a26a428005db97e5d5577 99035Author: Jeremy Huddleston <jeremyhu@apple.com> 99036Date: Thu Jan 13 16:08:29 2011 -0800 99037 99038 XQuartz: Update copyright dates in bundle's plist 99039 99040 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 99041 99042commit be3be7580b6f6fd2f7fa4d4abfe5e1ab19470223 99043Merge: 57a1d9b85 a1d885fdd 99044Author: Keith Packard <keithp@keithp.com> 99045Date: Thu Jan 20 21:21:21 2011 -0800 99046 99047 Merge remote branch 'ajax/for-keithp' 99048 99049commit 57a1d9b85331f6fe19d5111f5163139572ffbf02 99050Merge: 24ce650cf 642376979 99051Author: Keith Packard <keithp@keithp.com> 99052Date: Thu Jan 20 21:16:24 2011 -0800 99053 99054 Merge remote branch 'whot/for-keith' 99055 99056commit 24ce650cf4f0c6fa72faecd38c53d40703e6c959 99057Merge: bbdf81a05 09fd01090 99058Author: Keith Packard <keithp@keithp.com> 99059Date: Thu Jan 20 21:11:53 2011 -0800 99060 99061 Merge remote branch 'jturney/jturney-framebuffer-resize-for-master' 99062 99063commit bbdf81a056be0ea645da17a642dad5eadef3a906 99064Author: Alan Coopersmith <alan.coopersmith@oracle.com> 99065Date: Tue Jan 18 19:26:20 2011 -0800 99066 99067 Add xorg.conf.d shadow man page pointing to xorg.conf man page 99068 99069 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99070 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 99071 Tested-by: Gaetan Nadon <memsize@videotron.ca> 99072 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 99073 Signed-off-by: Keith Packard <keithp@keithp.com> 99074 99075commit a1d885fdd67503a442b348626d2eddf6d22419e8 99076Author: Adam Jackson <ajax@redhat.com> 99077Date: Wed Jun 23 13:28:55 2010 -0400 99078 99079 resource: Fix indentation 99080 99081 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99082 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99083 Signed-off-by: Adam Jackson <ajax@redhat.com> 99084 99085commit cb61cf5c99004ba3c76b504220c6728b5f2d2de6 99086Author: Adam Jackson <ajax@redhat.com> 99087Date: Thu Apr 15 18:36:55 2010 -0400 99088 99089 resource: Remove expectID hack 99090 99091 This is clearly meant to short-circuit the (modestly) expensive resource 99092 lookup in LegalNewID. The problem is that long-lived clients will 99093 eventually run completely through their XID space and start asking 99094 XC-MISC for IDs to reuse. Once that happens, the comparison against 99095 expectID will always be true, and we'll no longer catch XID collisions 99096 at all. 99097 99098 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99099 Signed-off-by: Adam Jackson <ajax@redhat.com> 99100 99101commit 3282e3c627f97f079e3a9af756a6b13bd9a5f227 99102Author: Adam Jackson <ajax@redhat.com> 99103Date: Thu Apr 15 08:46:28 2010 -0400 99104 99105 resource: s/NullResource/NULL/g 99106 99107 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99108 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99109 Signed-off-by: Adam Jackson <ajax@redhat.com> 99110 99111commit f953ae7d8a578d135a6faaf69d9c06eae7c85ede 99112Author: Adam Jackson <ajax@redhat.com> 99113Date: Mon Dec 14 14:38:10 2009 -0500 99114 99115 os: Reduce smart scheduler setup calls 99116 99117 We can return from WaitForSomething with no clients ready for any number 99118 of reasons. There's no reason to set up the scheduler timer when this 99119 happens. 99120 99121 Reviewed-by: Keith Packard <keithp@keithp.com> 99122 Signed-off-by: Adam Jackson <ajax@redhat.com> 99123 99124commit d127075da06239852c1cc745abfe63d0d180d984 99125Author: Adam Jackson <ajax@redhat.com> 99126Date: Thu Jan 20 19:03:33 2011 -0500 99127 99128 xdmxconfig: warning fix 99129 99130 xdmxconfig.c: In function ‘dmxConfigCanvasDraw’: 99131 xdmxconfig.c:299:23: warning: ‘maxHeight’ may be used uninitialized in this function 99132 99133 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 99134 Signed-off-by: Adam Jackson <ajax@redhat.com> 99135 99136commit 7a08f9abef7219fabdab8d1d49e8d3afb042e36a 99137Author: Adam Jackson <ajax@redhat.com> 99138Date: Thu Jan 20 18:59:39 2011 -0500 99139 99140 dmx: warning fixes 99141 99142 dmxinputinit.c: In function ‘dmxBlockHandler’: 99143 dmxinputinit.c:610:44: warning: cast from pointer to integer of different size 99144 dmxinputinit.c: In function ‘dmxWakeupHandler’: 99145 dmxinputinit.c:637:41: warning: cast from pointer to integer of different size 99146 dmxinputinit.c: In function ‘dmxInputInit’: 99147 dmxinputinit.c:1041:36: warning: cast to pointer from integer of different size 99148 99149 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 99150 Signed-off-by: Adam Jackson <ajax@redhat.com> 99151 99152commit ffd323b7c0212ed9b348e51cd9b36363d7c4d1f2 99153Author: Adam Jackson <ajax@redhat.com> 99154Date: Thu Jan 20 18:55:20 2011 -0500 99155 99156 dmx: warning fix 99157 99158 dmxinputinit.c: At top level: 99159 dmxinputinit.c:135:29: warning: ‘DMXCommonOth’ defined but not used 99160 99161 DMXCommonOth is actually mentioned in a #if 0 block, so delete it and 99162 the block that references it. If anyone needs it, git remembers. 99163 99164 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 99165 Signed-off-by: Adam Jackson <ajax@redhat.com> 99166 99167commit c1fe0b155d0567440228aa5d9e36036f37670e3b 99168Author: Adam Jackson <ajax@redhat.com> 99169Date: Thu Jan 20 18:52:57 2011 -0500 99170 99171 dmx: warning fix 99172 99173 dmxgc.c: In function ‘dmxChangeClip’: 99174 dmxgc.c:386:5: warning: case label value exceeds maximum value for type 99175 dmxgc.c:387:5: warning: case label value exceeds maximum value for type 99176 dmxgc.c:388:5: warning: case label value exceeds maximum value for type 99177 dmxgc.c:389:5: warning: case label value exceeds maximum value for type 99178 99179 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 99180 Signed-off-by: Adam Jackson <ajax@redhat.com> 99181 99182commit b02e006b2733ea457df41791f6054309e4edf7f6 99183Author: Adam Jackson <ajax@redhat.com> 99184Date: Thu Jan 20 18:46:00 2011 -0500 99185 99186 dmx: warning fixes 99187 99188 Dear gcc: I do not care about machines where sizeof(void *) < 99189 sizeof(int), and neither should you. 99190 99191 dmxextension.c: In function ‘dmxBECreateResources’: 99192 dmxextension.c:858:26: warning: cast from pointer to integer of different size 99193 dmxextension.c: In function ‘dmxBERestoreRenderPict’: 99194 dmxextension.c:1062:29: warning: cast from pointer to integer of different size 99195 dmxextension.c: In function ‘dmxBERestoreRenderGlyph’: 99196 dmxextension.c:1084:35: warning: cast from pointer to integer of different size 99197 dmxextension.c: In function ‘dmxAttachScreen’: 99198 dmxextension.c:1277:8: warning: cast to pointer from integer of different size 99199 dmxextension.c:1286:34: warning: cast to pointer from integer of different size 99200 dmxextension.c:1292:35: warning: cast to pointer from integer of different size 99201 dmxextension.c: In function ‘dmxBEDestroyResources’: 99202 dmxextension.c:1456:26: warning: cast from pointer to integer of different size 99203 dmxextension.c: In function ‘dmxDetachScreen’: 99204 dmxextension.c:1599:8: warning: cast to pointer from integer of different size 99205 99206 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 99207 Signed-off-by: Adam Jackson <ajax@redhat.com> 99208 99209commit 64237697994871adfcf4905b5784e75cd7281579 99210Author: Peter Hutterer <peter.hutterer@who-t.net> 99211Date: Thu Jan 13 12:20:38 2011 +1000 99212 99213 xkb: after making changes to the xkb ctrls, copy them back into kbdfeed. 99214 99215 enabled_ctrls_changes nowhere near the usual event or config paths. So this 99216 condition always evaluated to false and the memcpy would thus never been 99217 hit. As a result, any modification to the XKB struct during 99218 XkbUpdateDescActions was not reflected in the kbdfeed ctrls. 99219 The flag that is set by XkbUpdateDescActions() if ctrls were changed are in 99220 enabled_ctrls. 99221 99222 This mainly affected keyboard repeat control as XKB uses the kbdfeed ctrls, 99223 not XKB's per_key_repeats, to determine if a key needs to be repeated. Thus, 99224 adding a "repeat= False" to the XKB map of any action did not have any 99225 effect. 99226 99227 Test case: 99228 assign Mode_switch to any key that by default repeats, e.g. the menu key. 99229 99230 key <COMP> { [ Mode_switch ] }; 99231 99232 Then modify the Mode_switch action to not repeat the key. 99233 99234 interpret Mode_switch+AnyOfOrNone(all) { 99235 virtualModifier= AltGr; 99236 useModMapMods=level1; 99237 action= SetGroup(group=+1); 99238 // Add this line 99239 repeat= False; 99240 }; 99241 99242 Though the flags are correctly reflected in the description loaded in the 99243 server, the change is not handed back to the kbdfeed struct and XKB will 99244 trigger softrepeats of this key. 99245 99246 This patch also adds two explanatory comments and an extra check, as this 99247 path may be hit before the CtrlProc for the kbdfeed struct is set. 99248 99249 Red Hat Bug 537708 <https://bugzilla.redhat.com/show_bug.cgi?id=537708> 99250 99251 Also fixes broken auto-repeat of the backspace key in the colemak layout 99252 (mapped to CapsLock). 99253 99254 X.Org Bug 16318 <http://bugs.freedesktop.org/show_bug.cgi?id=16318> 99255 99256 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 99257 Tested-by: Dirk Wallenstein <halsmit@t-online.de> 99258 Reviewed-by: Dirk Wallenstein <halsmit@t-online.de> 99259 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99260 99261commit f1326ed2910bd985fafdb48714b1d6f38116b083 99262Author: Peter Hutterer <peter.hutterer@who-t.net> 99263Date: Thu Jan 13 09:19:28 2011 +1000 99264 99265 xkb: Replace a few manual bitflips with SetBit & friends. 99266 99267 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 99268 Reviewed-by: Dirk Wallenstein <halsmit@t-online.de> 99269 99270commit 16eb4f2a7b67ee169785973f9e0a3773e13fdea6 99271Author: Peter Hutterer <peter.hutterer@who-t.net> 99272Date: Fri Jan 7 09:53:35 2011 +1000 99273 99274 Xi: reset remainders when warping the device. 99275 99276 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 99277 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99278 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 99279 Reviewed-by: Simon Thum <simon.thum@gmx.de> 99280 99281commit 09fd010902fad56735b8069b1becb80d85bd6a35 99282Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99283Date: Wed Sep 29 22:54:22 2010 +0100 99284 99285 Cygwin/X: DirectDraw engines shouldn't try to blit if the surface wasn't allocated 99286 99287 Fix winShadowUpdateDD(|NL) so we don't try to blit to primary surface if it didn't get allocated 99288 99289 (Intel drivers, in particular, seem to like to issue a WM_DISPLAYCHANGE during a suspend/resume 99290 cycle, but not allow surface to be allocated right then) 99291 99292 Also: 99293 Use winReleasePrimarySurfaceShadowDD(|NL) in winFreeFBShadowDD(|NL) rather than open coding it 99294 Don't mess about recreating surface if we're going to resize it anyhow 99295 99296 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99297 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99298 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99299 99300commit 625ab9701fd75b879c1dafc05fa979591eea87c0 99301Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99302Date: Tue Aug 31 13:45:43 2010 +0100 99303 99304 Cygwin/X: Simplify and consolidate reporting of the bpp value we are going to use 99305 99306 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99307 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99308 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99309 99310commit 3f7339a7c5d3dcd05909b041865125f4cb6fa29e 99311Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99312Date: Tue Aug 31 10:13:01 2010 +0100 99313 99314 Cygwin/X: Deal with RANDR depth changes correctly in ShadowGDI drawing engine 99315 99316 Make ShadowGDI drawing engine only change the size of the screen 99317 pixmap/shadow framebuffer on an RANDR change, not the bpp/depth 99318 as well. 99319 99320 The server requires the screen pixmap's depth to be invariant. 99321 99322 Other drawing engines aren't quite as affected by this issue as 99323 they won't draw to the display, if it has changed colour depth, 99324 but probably still need some attention. 99325 99326 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99327 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99328 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99329 99330commit 8b22f83113fbdc09b932b5ad7e44f629fc15e3b5 99331Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99332Date: Mon Aug 30 17:17:57 2010 +0100 99333 99334 Cygwin/X: Use winUpdateFBPointer() in winshaddd.c rather than duplicating it inline 99335 99336 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99337 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99338 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99339 99340commit 8385c426f86e9955e9e570a46f75bddd3c10ca01 99341Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99342Date: Tue Aug 10 21:55:15 2010 +0100 99343 99344 Cygwin/X: Remove WIN_DIB_MAXIMUM_SIZE check 99345 99346 This is only relevant to pre-NT versions of Windows, which are all EOL. 99347 99348 Also, it's in the wrong place now as framebuffer can get resized. 99349 99350 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99351 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99352 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99353 99354commit 981ad1f364cf4fe8008c0f3592eb0f73dd14a118 99355Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99356Date: Tue Aug 10 16:24:57 2010 +0100 99357 99358 Cygwin/X: Fix a typo in command line argument validation code 99359 99360 Fortunately, these swapped constants are benign as they have the same 99361 value, 0 99362 99363 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99364 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99365 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99366 99367commit 5390b494672393506466d8afdb9b146b0e585cc0 99368Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99369Date: Tue Aug 10 13:16:13 2010 +0100 99370 99371 Cygwin/X: Don't turn off -multiplemonitors when all monitors don't have the same pixel format when using shadow GDI engine 99372 99373 Don't turn off -multiplemonitors when all monitors don't have the 99374 same pixel format and when using shadow GDI engine, just warn that 99375 performance may be degraded 99376 99377 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99378 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99379 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99380 99381commit 33106e1e807a828208b306512e78c5e3e93960d3 99382Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99383Date: Tue Mar 30 19:49:41 2010 +0100 99384 99385 Cygwin/X: Generate RANDR change on WM_DISPLAYCHANGE for rootless modes 99386 99387 When RANDR resizing is enabled, generate an internal RANDR change when 99388 WM_DISPLAYCHANGE occurs in rootless modes for screens which occupy an 99389 entire monitor or the virtual desktop. 99390 99391 Store the monitor number and use that to handle WM_DISPLAYCHANGE for a 99392 screen specified with '-screen @monitor' 99393 99394 In rooted mode, WM_DISPLAYCHANGE isn't relevant (except where display 99395 depth changes may cause problems). (A maximized screen window will get 99396 WM_SIZE to adjust it to the new monitor size) 99397 99398 In rooted fullscreen mode, WM_DISPLAYCHANGE shouldn't be seen, as we 99399 have the resolution we have selected for the fullscreen session) 99400 (Could client randr requests be handled in fullscreen to cause a change 99401 of the fullscreen resolution? ) 99402 99403 Don't bother do a RANDR resize if the dimensions aren't actually changing 99404 when WM_DISPLAYCHANGE is sent (should handle WM_DISPLAYCHANGE to size 0x0 99405 that the intel driver seems to like to send) 99406 99407 Various debug output improvements 99408 99409 Also, remove the note that XWin can't handle display mode changes from 99410 the man page 99411 99412 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99413 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99414 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99415 99416commit bbc511e80b2a9365f6a1528bc1595772f83be654 99417Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99418Date: Mon Feb 15 13:42:04 2010 +0000 99419 99420 Cygwin/X: Make WM_SIZE use RandR resizing when -resize=randr 99421 99422 To avoid recursion, WM_SIZE requests shouldn't generate XRANDR requests 99423 when no change is neeeded. 99424 99425 We do the actual resize on WM_EXITSIZEMOVE, as resizing occurs in 99426 a modal loop, to avoid a backlog of resize events building up as 99427 the X server doesn't get a change to process anything until the resize 99428 is completed. 99429 99430 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99431 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99432 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99433 99434commit 85c497a8b6c488ef9ea2c6d7b49e6f9b992fb4a2 99435Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99436Date: Tue Mar 30 20:48:10 2010 +0100 99437 99438 Cygwin/X: Implement framebuffer resizing in RANDR extension 99439 99440 Implement framebuffer resizing in RANDR extension: 99441 99442 Resize the frame buffer, the screen's root window and the native window 99443 containing the root window image. 99444 99445 Correctly allow for decorations in new native window size when resizing native window 99446 to fit the new framebuffer size with AdjustWindowRectEx() 99447 99448 Update physical size info for a screen when it is changed by RANDR 99449 99450 Forbid client-requested RANDR changes in fullscreen and rootless modes 99451 99452 Only resize window on an external RandR request, to avoid recursing on 99453 a WM_SIZE requested resize. 99454 99455 Also, add prototypes for winRandRInit() and winDoRandRScreenSetSize() to header file 99456 99457 Also, update the author list and copyright for winrandr.c 99458 99459 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99460 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99461 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99462 99463commit 873abef315f5d947b864428891381bff539c5869 99464Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99465Date: Tue Mar 30 20:07:19 2010 +0100 99466 99467 Cygwin/X: Add -resize command line option 99468 99469 Add -resize command line option to configure how native window sizing frame 99470 is used. 99471 99472 In additions to the existing fixed and scrollbars modes, add a new mode to 99473 allow framebuffer to be resized using native window frame 99474 99475 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99476 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99477 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99478 99479commit 46c57788539d8a5f0246528b4f88ad4ed6d867d1 99480Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99481Date: Tue Feb 23 15:44:36 2010 +0000 99482 99483 Cygwin/X: Move QueryMonitor() out of windprocarg.c 99484 99485 Move QueryMonitor() out of windprocarg.c into a new file, winmonitors.c, 99486 as we use to use it from other places as well 99487 99488 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99489 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99490 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99491 99492commit 73b02e964787ea9fb2d139dc781c9b2495ccb031 99493Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99494Date: Fri Feb 12 12:29:30 2010 +0000 99495 99496 Cygwin/X: Remove an attempt at detecting if WM_DISPLAYCHANGE affects the X screen 99497 99498 Remove an attempt at the rather difficult optimization of detecting 99499 if WM_DISPLAYCHANGE affects any of the monitors which intersect the 99500 native window for the X screen. 99501 99502 We'll always act as if it does, which it probably usually the case. 99503 99504 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99505 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99506 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99507 99508commit a46146af5f6c2d96728c43f0df369ea00bcae017 99509Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99510Date: Thu Apr 1 21:00:43 2010 +0100 99511 99512 Cygwin/X: Rather than storing calculated physical display sizes, calculate them when needed 99513 99514 (This stored calculation was wrong if -dpi came after -screen on the command 99515 line, anyhow) 99516 99517 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99518 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99519 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99520 99521commit df518001271dbc0c49935350b755f0c0562cdb92 99522Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99523Date: Tue Oct 13 14:47:33 2009 +0100 99524 99525 Cygwin/X: Refactor framebuffer allocation/release in drawing engines 99526 99527 Refactor the drawing engines so we can explicitly allocate and release the framebuffer for a screen 99528 99529 Move the setting of dwPaddedWidth into the DDNL engine, so it is updated when the framebuffer changes size 99530 99531 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99532 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99533 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99534 99535commit 708d0b9781c6f3f0c0ae7052f45223b1c2524809 99536Author: Jon TURNEY <jon.turney@dronecode.org.uk> 99537Date: Tue Oct 13 14:48:17 2009 +0100 99538 99539 Cygwin/X: Enable building of nativegdi and primaryfb engines to avoid further rusting 99540 99541 Some trivial build fixes required 99542 99543 Also fill out all function pointers for primaryfb engine 99544 99545 Also tidy up the man page section describing drawing engines. 99546 99547 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 99548 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 99549 Tested-by: Colin Harrison <colin.harrison@virgin.net> 99550 99551commit c6aa4755ec355101a62bef86dbb090262fe806f6 99552Author: Alan Coopersmith <alan.coopersmith@oracle.com> 99553Date: Sat Jan 15 12:13:39 2011 -0800 99554 99555 xkb/ddxLoad.c doesn't need <paths.h> any more 99556 99557 Was previously used for _PATH_VARTMP, but that was removed in 99558 534fc5140b039a8c98ab715d0a6740d513b41209 99559 99560 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99561 Reviewed-by: Julien Cristau <jcristau@debian.org> 99562 Signed-off-by: Keith Packard <keithp@keithp.com> 99563 99564commit abfea17342da774aa4e3a351a73c8f2af6e6fd28 99565Merge: 361128389 e65c3f8bc 99566Author: Keith Packard <keithp@keithp.com> 99567Date: Tue Jan 18 15:19:34 2011 -0800 99568 99569 Merge remote branch 'ajax/for-keithp' 99570 99571commit 361128389e5cb0101cbd091ff8de77cf34608f6c 99572Merge: 65ceaadbf d9225b960 99573Author: Keith Packard <keithp@keithp.com> 99574Date: Tue Jan 18 15:18:08 2011 -0800 99575 99576 Merge remote branch 'jcristau/for-keith' 99577 99578commit 65ceaadbf0fb65f1962cb96d5b87fd1175f85a3b 99579Author: Gaetan Nadon <memsize@videotron.ca> 99580Date: Sun Jan 9 20:28:31 2011 -0500 99581 99582 Xquartz: rename man page directory from doc to man 99583 99584 Following the convention makes it easier to locate man pages, 99585 user's or developer's documentation and specifications. 99586 99587 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99588 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99589 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99590 Signed-off-by: Keith Packard <keithp@keithp.com> 99591 99592commit 7ed0d426f616fe9e44ac7091622a798f2043dfe4 99593Author: Gaetan Nadon <memsize@videotron.ca> 99594Date: Sun Jan 9 20:28:30 2011 -0500 99595 99596 xfree86: use sed rather than cpp to perform string substitutions 99597 99598 Makefile.am: there are only 4 string substitutions to be done in this file. 99599 SED is much simpler than the C pre processor which adds its own 99600 strings which must be substituted by sed, still. 99601 99602 xorgconf.cpp: replaced __xconfigfile__ with xorg.conf as this file name 99603 is hard coded in the xserver configuration and cannot change. 99604 Replace XCOMM with # permanently. 99605 99606 Delete cpprules.in as it isn't used anywhere else. Should one need 99607 cpprules for real cpp work, there is one in Xquartz from which the 99608 the old man pages code have been stripped. 99609 99610 Fix trailing spaces. 99611 99612 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99613 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99614 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99615 Signed-off-by: Keith Packard <keithp@keithp.com> 99616 99617commit 885a3330092c13786baa3aaabd728e71962486c4 99618Author: Gaetan Nadon <memsize@videotron.ca> 99619Date: Sun Jan 9 20:28:29 2011 -0500 99620 99621 man: refactor common code in the man pages makefiles 99622 99623 Create a manpages.am makefile snippet in the root dir. 99624 Each man page makefile includes manpages.am. 99625 99626 Now all man pages in xserver are generated the same way 99627 using the same method as all of other xorg modules. 99628 All ".man.pre" files in git are ".man" now. 99629 Links are no longer created between different file types. 99630 99631 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99632 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99633 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99634 Signed-off-by: Keith Packard <keithp@keithp.com> 99635 99636commit c86f7c2b8eedd3126066a3aff44bc8767eb5e277 99637Author: Gaetan Nadon <memsize@videotron.ca> 99638Date: Sun Jan 9 20:28:28 2011 -0500 99639 99640 Xdmx: build all man pages using XORG_MANPAGE_SECTIONS 99641 99642 Use standard directory and makefile. 99643 99644 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99645 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99646 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99647 Signed-off-by: Keith Packard <keithp@keithp.com> 99648 99649commit 9bd9d550a0d6fb4fe3daacb604658de92e69d99c 99650Author: Gaetan Nadon <memsize@videotron.ca> 99651Date: Sun Jan 9 20:28:27 2011 -0500 99652 99653 exa: build man pages using XORG_MANPAGE_SECTIONS 99654 99655 Nothing requires the use of a C preprocessor 99656 Using standard file extensions (.man) means no need for .gitignore 99657 Use standard directory and makefile 99658 Fix trailing whitespaces 99659 99660 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99661 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99662 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99663 Signed-off-by: Keith Packard <keithp@keithp.com> 99664 99665commit d4b1fd3ff66d1a3e44a85aed34d9a01eee0d37fd 99666Author: Gaetan Nadon <memsize@videotron.ca> 99667Date: Sun Jan 9 20:28:26 2011 -0500 99668 99669 fbdevhw: build man pages using XORG_MANPAGE_SECTIONS 99670 99671 Nothing requires the use of a C preprocessor 99672 Using standard file extensions (.man) means no need for .gitignore 99673 Use standard directory and makefile 99674 Fix trailing whitespaces 99675 99676 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99677 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99678 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99679 Signed-off-by: Keith Packard <keithp@keithp.com> 99680 99681commit b9276cb86b433bc740fe26c2e624d3d5399732f6 99682Author: Gaetan Nadon <memsize@videotron.ca> 99683Date: Sun Jan 9 20:27:53 2011 -0500 99684 99685 XFree86 utils: build utils man pages using XORG_MANPAGE_SECTIONS 99686 99687 Nothing requires the use of a C preprocessor 99688 Using standard file extensions (.man) means no need for .gitignore 99689 Use standard directory and makefile 99690 Fix trailing whitespaces 99691 99692 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99693 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99694 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99695 Signed-off-by: Keith Packard <keithp@keithp.com> 99696 99697commit ac5ffef60e05eb05275913b133786709881e7e7e 99698Author: Gaetan Nadon <memsize@videotron.ca> 99699Date: Sun Jan 9 20:27:52 2011 -0500 99700 99701 Xephyr: build Xephyr man pages using XORG_MANPAGE_SECTIONS 99702 99703 Nothing requires the use of a C preprocessor 99704 Using standard file extensions (.man) means no need for .gitignore 99705 Use standard directory and makefile 99706 Fix trailing whitespaces 99707 99708 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99709 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99710 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99711 Signed-off-by: Keith Packard <keithp@keithp.com> 99712 99713commit 9240cffd659b1c13d66162200c6e384ea1cdbc44 99714Author: Gaetan Nadon <memsize@videotron.ca> 99715Date: Sun Jan 9 20:27:51 2011 -0500 99716 99717 Xvfb: build Xvfb man pages using XORG_MANPAGE_SECTIONS 99718 99719 Nothing requires the use of a C preprocessor 99720 Using standard file extensions (.man) means no need for .gitignore 99721 Use standard directory and makefile 99722 Fix trailing whitespaces 99723 99724 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99725 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99726 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99727 Signed-off-by: Keith Packard <keithp@keithp.com> 99728 99729commit 3f4c71799b3164440624d09b7b631d3d3c0a29ef 99730Author: Gaetan Nadon <memsize@videotron.ca> 99731Date: Sun Jan 9 20:27:50 2011 -0500 99732 99733 XWin: build XWin man pages using XORG_MANPAGE_SECTIONS 99734 99735 Nothing requires the use of a C preprocessor 99736 Using standard file extensions (.man) means no need for .gitignore 99737 Use standard directory and makefile 99738 Now can be built easily on any platform in the man directory 99739 99740 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99741 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99742 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99743 Signed-off-by: Keith Packard <keithp@keithp.com> 99744 99745commit b8492e18a7e5322c4a492f6cf6cfeb5b0221d994 99746Author: Gaetan Nadon <memsize@videotron.ca> 99747Date: Sun Jan 9 20:27:49 2011 -0500 99748 99749 Xnest: build Xnest man pages using XORG_MANPAGE_SECTIONS 99750 99751 Nothing requires the use of a C preprocessor 99752 Using standard file extensions (.man) means no need for .gitignore 99753 Use standard directory and makefile 99754 99755 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99756 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99757 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99758 Signed-off-by: Keith Packard <keithp@keithp.com> 99759 99760commit 57ca20d26614a0787cdd0db91a2eb8cdb301f2dc 99761Author: Gaetan Nadon <memsize@videotron.ca> 99762Date: Sun Jan 9 20:27:48 2011 -0500 99763 99764 Xquartz: strip the man page code from cpprules.in 99765 99766 This is dead code which will not be used in the future. 99767 99768 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99769 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99770 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99771 Signed-off-by: Keith Packard <keithp@keithp.com> 99772 99773commit c06c46acdc2ead20a6030118fad6092b3da6709f 99774Author: Gaetan Nadon <memsize@videotron.ca> 99775Date: Sun Jan 9 20:27:47 2011 -0500 99776 99777 Xquartz: build Xquartz man pages using XORG_MANPAGE_SECTIONS 99778 99779 Nothing requires the use of a C preprocessor 99780 Using standard file extensions (.man) 99781 Use standard directory and makefile 99782 99783 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99784 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99785 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99786 Signed-off-by: Keith Packard <keithp@keithp.com> 99787 99788commit a9e85013b1045c7e02891a019bfa8042a219b7f1 99789Author: Gaetan Nadon <memsize@videotron.ca> 99790Date: Sun Jan 9 20:27:46 2011 -0500 99791 99792 man: build Xorg and xorg.conf man pages using XORG_MANPAGE_SECTIONS 99793 99794 Nothing requires the use of a C preprocessor 99795 Using standard file extensions (.man) means no need for .gitignore 99796 Use standard directory and makefile 99797 99798 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99799 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99800 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99801 Signed-off-by: Keith Packard <keithp@keithp.com> 99802 99803commit dcdd8ea82c0e2f657abdc80a22aa0c01ead07e54 99804Author: Gaetan Nadon <memsize@videotron.ca> 99805Date: Sun Jan 9 20:27:45 2011 -0500 99806 99807 Xorg.man: fix whitespace issues 99808 99809 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 99810 Reviewed-by: Rémi Cardona <remi@gentoo.org> 99811 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 99812 Signed-off-by: Keith Packard <keithp@keithp.com> 99813 99814commit d9225b9602c85603ae616a7381c784f5cf5e811c 99815Author: Julien Cristau <jcristau@debian.org> 99816Date: Wed Nov 10 22:39:54 2010 +0100 99817 99818 glx: validate numAttribs field before using it 99819 99820 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 99821 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99822 Signed-off-by: Julien Cristau <jcristau@debian.org> 99823 99824commit 62319e8381ebd645ae36b25e5fc3c0e9b098387b 99825Author: Julien Cristau <jcristau@debian.org> 99826Date: Sun Aug 22 16:20:45 2010 +0100 99827 99828 glx: swap the request arrays entirely, not just half of them 99829 99830 Various glx requests include a list of pairs of attributes. We were 99831 only swapping the first half. 99832 99833 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 99834 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99835 Signed-off-by: Julien Cristau <jcristau@debian.org> 99836 99837commit 6c69235a9dfc52e4b4e47630ff4bab1a820eb543 99838Author: Julien Cristau <jcristau@debian.org> 99839Date: Sun Aug 22 00:50:05 2010 +0100 99840 99841 glx: check request length before swapping 99842 99843 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 99844 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99845 Signed-off-by: Julien Cristau <jcristau@debian.org> 99846 99847commit ec9c97c6bf70b523bc500bd3adf62176f1bb33a4 99848Author: Julien Cristau <jcristau@debian.org> 99849Date: Sat Jul 3 19:47:55 2010 +0100 99850 99851 glx: validate request lengths 99852 99853 Reviewed-by: Adam Jackson <ajax@redhat.com> 99854 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 99855 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99856 Signed-off-by: Julien Cristau <jcristau@debian.org> 99857 99858commit 3f0d3f4d97bce75c1828635c322b6560a45a037f 99859Author: Julien Cristau <jcristau@debian.org> 99860Date: Sat Jul 3 19:42:26 2010 +0100 99861 99862 glx: make sure screen is non-negative in validGlxScreen 99863 99864 Reviewed-by: Adam Jackson <ajax@redhat.com> 99865 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 99866 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99867 Signed-off-by: Julien Cristau <jcristau@debian.org> 99868 99869commit 82b1eaa6cad20f39dbf15573bdb3d62acbcd91f9 99870Author: Tom Fogal <tfogal@alumni.unh.edu> 99871Date: Wed Dec 22 14:39:48 2010 -0700 99872 99873 Autodetect TLS support instead of defaulting to disabled. 99874 99875 Uses the AX_TLS macro to figure out if the current platform 99876 supports TLS. If it does, enable TLS automagically. Still 99877 respects --enable option, regardless. 99878 99879 Signed-off-by: Tom Fogal <tfogal@alumni.unh.edu> 99880 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 99881 99882commit 93393411e39fbb0f72cf6c2f79ebde177aa86932 99883Author: Tom Fogal <tfogal@alumni.unh.edu> 99884Date: Wed Dec 22 16:03:03 2010 -0700 99885 99886 Add macro for detecting thread local storage support. 99887 99888 This adds an AX_TLS macro which attempts to identify if the 99889 underlying compiler/platform supports thread local storage (TLS). 99890 The macro comes from the autoconf macro archive. 99891 99892 Signed-off-by: Tom Fogal <tfogal@alumni.unh.edu> 99893 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 99894 99895commit 60801ff8703ec36995139a6dd7b1fc26ae8854a6 99896Author: Dan Nicholson <dbn.lists@gmail.com> 99897Date: Tue Nov 23 11:38:50 2010 -0800 99898 99899 xfree86: Convert libxf86config to static libtool library 99900 99901 In order to use libxf86config in a shared library, all the code must be 99902 compiled with -fPIC. Add proper PIC support for libxf86config by turning 99903 it into a libtool library. However, since we don't want to guarantee API 99904 or ABI stability, make sure it's only built static. 99905 99906 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 99907 Acked-by: Gaetan Nadon <memsize@videotron.ca> 99908 99909commit 8456625d64bb0013a496f3b56ea3aa20cf7a6142 99910Merge: aa30a8658 ce74e7562 99911Author: Peter Hutterer <peter.hutterer@who-t.net> 99912Date: Fri Jan 7 09:35:07 2011 +1000 99913 99914 Merge branch 'for-peter' of git://people.freedesktop.org/~daniels/xserver into for-keith 99915 99916commit aa30a86583c3fc1a83c46e8c7583182d750edcde 99917Author: Tomas Carnecky <tom@dbservice.com> 99918Date: Sun Aug 1 21:28:18 2010 +0200 99919 99920 NO_PANORAMIX preprocessor symbol is nowhere defined 99921 99922 Which makes it useless, so remove it. 99923 99924 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 99925 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 99926 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 99927 99928commit fc6cbee772bafabf7ddd7a75043bd55bb78ad09e 99929Author: Peter Hutterer <peter.hutterer@who-t.net> 99930Date: Thu Jan 6 12:58:00 2011 +1000 99931 99932 input: add valuator_mask_free() to free a valuator mask. 99933 99934 Expecting the caller to free the mask requires us to keep it in a single 99935 memory block (which may be an issue lateron), aside from leaving the API 99936 asymetrical. Provide valuator_mask_free() to free the memory and reset the 99937 mask pointer to NULL. 99938 99939 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 99940 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 99941 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 99942 99943commit 4fbadc8b17237f3c7f8c0bf56003d1139c86655e 99944Author: Daniel Stone <daniel@fooishbar.org> 99945Date: Thu Jan 6 11:32:26 2011 +1000 99946 99947 Input: Prevent MD grabs from wandering on to other MDs 99948 99949 The code to set sync.other in DeliverGrabbedEvents is supposed to reset 99950 sync.other for a paired MD to the grab under consideration, but was 99951 rather optimistic in resetting sync.other for _all_ devices. 99952 99953 This would fall apart given two sets of MDs (A paired with B, Y paired 99954 with Z), where both MDs were in FREEZE_BOTH_NEXT_EVENT due to being 99955 called with SyncBoth, where no event had yet triggered the grab. An 99956 event being processed on MD A would result in B, Y and Z all having 99957 sync.other set to A's grab, rather than just B. 99958 99959 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 99960 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 99961 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 99962 99963commit e65c3f8bcc86845f21ac575e2bfb4b21b67d5ebf 99964Author: Adam Jackson <ajax@redhat.com> 99965Date: Wed Jan 5 12:04:19 2011 -0500 99966 99967 dix: Add a Screen method for additional cursor confinement 99968 99969 This just reserves the slot in the ABI. Confining cursors to CRTCs will 99970 come soon. 99971 99972 v2: Just reserve the slot. 99973 99974 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 99975 Signed-off-by: Adam Jackson <ajax@redhat.com> 99976 99977commit ce74e7562dcc800306ff92b6d3c5ee885cd4763d 99978Author: Daniel Stone <daniel@fooishbar.org> 99979Date: Tue Dec 7 11:48:14 2010 +0000 99980 99981 Input: Make CheckPassiveGrabsOnWindow return grab, export 99982 99983 Change CheckPassiveGrabsOnWindow to return the GrabPtr it used (or NULL 99984 if none) rather than a boolean, and export it. Also add an additional 99985 boolean 'activate' parameter; use TRUE for existing behaviour, or FALSE 99986 to only find the grab and then return it. 99987 99988 This will be used in forthcoming touch patches to find the grabs, rather 99989 than open-coding same. 99990 99991 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 99992 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 99993 99994commit eb5aaf5eb819f1287ce40da79edbe07bafb2b5c0 99995Author: Daniel Stone <daniel@fooishbar.org> 99996Date: Tue Dec 28 12:15:28 2010 +0000 99997 99998 Input: Export GetEventMask 99999 100000 Make it non-static, add to headers. 100001 100002 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100003 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100004 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100005 100006commit 30c8c7a86310139e79c9698caee21b3bd51936bd 100007Author: Daniel Stone <daniel@fooishbar.org> 100008Date: Tue Dec 28 12:13:29 2010 +0000 100009 100010 Input: Handle grabs with no Xi 1.x equivalent 100011 100012 Don't try to search for an Xi 1.x grab in CheckPassiveGrabsOnWindow for 100013 events with no Xi 1.x equivalent. 100014 100015 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100016 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100017 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100018 100019commit 690476250f58616d5156317e01ba547497f4331a 100020Author: Daniel Stone <daniel@fooishbar.org> 100021Date: Mon Sep 20 15:03:05 2010 +1000 100022 100023 Input: Pass sprite instead of device to XYToWindow, make non-static 100024 100025 XYToWindow calculates the position of the cursor and updates the sprite 100026 trace, but does nothing else with the device. Pass a SpritePtr instead 100027 so we can update an alternate focus instead of hardcoding the device's 100028 sprite. Also make this function non-static, so we can use it elsewhere. 100029 100030 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100031 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100032 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100033 100034commit 07a892cd8276f8092de95f1909f2ebec6f5fdf08 100035Author: Daniel Stone <daniel@fooishbar.org> 100036Date: Mon Sep 20 15:03:04 2010 +1000 100037 100038 Input: Pass sprite instead of device to FixUpEventFromWindow 100039 100040 Since FixUpEventFromWindow only uses the sprite trace to determine the 100041 window stack, pass in a sprite instead of hardcoding the device sprite, 100042 so we can deliver to windows other than the one currently containing the 100043 sprite. 100044 100045 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100046 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100047 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100048 100049commit 10d6c3ca103ea155488cd2f29052334b5691e680 100050Author: Daniel Stone <daniel@fooishbar.org> 100051Date: Mon Dec 27 13:06:03 2010 +0000 100052 100053 Input: Make RootWindow() take a sprite instead of device 100054 100055 GetCurrentRootWindow already works for the device case, although not as 100056 an lvalue. 100057 100058 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100059 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100060 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100061 100062commit 23c37ce1fe42e3da0609a566478e947843b05023 100063Author: Daniel Stone <daniel@fooishbar.org> 100064Date: Mon Dec 27 12:48:42 2010 +0000 100065 100066 Input: Export SpriteRec for use in public API 100067 100068 Change its definition to be more in line with other structs, so we can 100069 use it from input.h. 100070 100071 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100072 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100073 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100074 100075commit 06a103ad1d6a0ef6146c97a8c81dc7de1c0a0083 100076Author: Daniel Stone <daniel@fooishbar.org> 100077Date: Tue Dec 28 16:37:48 2010 +0000 100078 100079 Test: Input: Test up to supported server, not protocol, events 100080 100081 When building against a newer inputproto, the server may support fewer 100082 than XI_LASTEVENT events. We already have XI2LASTEVENT for the highest 100083 event number supported by the server, so use that instead. 100084 100085 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100086 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100087 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100088 100089commit a083efe8715e8b29c9bd7f4e7bb429a94620bfc4 100090Author: Daniel Stone <daniel@fooishbar.org> 100091Date: Tue Dec 28 16:06:29 2010 +0000 100092 100093 Test: Input: Check flags on DeviceEvent 100094 100095 Add initial validation of acceptable flags for XI2 device events, and 100096 make sure they're swapped. 100097 100098 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100099 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100100 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100101 100102commit 2592effef5f171af3f01a2b5130d9747403140f6 100103Author: Daniel Stone <daniel@fooishbar.org> 100104Date: Tue Dec 28 13:42:06 2010 +0000 100105 100106 Test: Input: Add helper function for failing EventToCore 100107 100108 We have quite a few tests which involve checking that EventToCore fails 100109 for specific events, so refactor them into a separate function. 100110 100111 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100112 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100113 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100114 100115commit e1aed88be92c27d76bff1743f35f9915dea9264f 100116Author: Daniel Stone <daniel@fooishbar.org> 100117Date: Tue Dec 28 17:00:29 2010 +0000 100118 100119 Input: Swap flags in DeviceEvents 100120 100121 Swap flags for different-endian clients when delivering XI2 100122 DeviceEvents. 100123 100124 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100125 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100126 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 100127 100128commit 6358a60065eef167d4e5f4afd981ff26deeba80d 100129Author: Michel Dänzer <daenzer@vmware.com> 100130Date: Thu Jan 6 09:55:41 2011 +0100 100131 100132 EXA: Pad size of system memory copy for 1x1 pixmaps (bug #32803). 100133 100134 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=32803 . 100135 100136 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 100137 Reviewed-by: Dave Airlie <airlied@redhat.com> 100138 Signed-off-by: Keith Packard <keithp@keithp.com> 100139 100140commit f3480286aeb3009623d8d4b0202eadda0049552d 100141Author: Ville Syrjälä <ville.syrjala@nokia.com> 100142Date: Wed Jan 5 20:41:09 2011 +0200 100143 100144 composite: Support updating an arbitrary subtree 100145 100146 Rename compUpdateWindow to compPaintWindowToParent and split the child 100147 walk to compPaintChildrenToWindow. Calling compPaintChildrenToWindow 100148 allows an arbitrary subtree to be updated, instead of having to update 100149 all the windows. This will be used to make sure all the descendants are 100150 copied to the parent when the parent window contents need to be accessed 100151 in IncludeInferios sub-window mode. 100152 100153 WindowRec has a new member 'damagedDescendants' that is used to keep 100154 track of which subtrees need updating. When a window is damaged, 100155 'damagedDescendants' will be set for all the ancestors, and when a 100156 subtree is updated, the tree walk can be stopped early if no damaged 100157 descendants are present. 100158 100159 CompScreenRec no longer needs the 'damaged' member since the root 100160 window's 'damagedDescendants' provides the same information. 100161 100162 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 100163 Signed-off-by: Keith Packard <keithp@keithp.com> 100164 100165commit b89e6dbdfbb50e3b5bc7fcb7eccc397c467c92f8 100166Author: Ville Syrjälä <ville.syrjala@nokia.com> 100167Date: Wed Jan 5 20:41:08 2011 +0200 100168 100169 composite: Add SourceValidate wrapper 100170 100171 When SourceValidate is performed on a window with IncludeInferiors 100172 sub-window mode, force an immediate update of all the automatically 100173 redirected windows, so that the current window contents will be up 100174 to date. 100175 100176 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 100177 Reviewed-by: Keith Packard <keithp@keithp.com> 100178 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100179 Signed-off-by: Keith Packard <keithp@keithp.com> 100180 100181commit 84154954db54696d4661eb8d0a6014cdbff3c91f 100182Author: Ville Syrjälä <ville.syrjala@nokia.com> 100183Date: Wed Jan 5 20:41:07 2011 +0200 100184 100185 composite: Add GetImage wrapper 100186 100187 When GetImage is performed on a window, force an immediate update of 100188 all the automatically redirected windows, so that the current window 100189 contents will be up to date. 100190 100191 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 100192 Reviewed-by: Keith Packard <keithp@keithp.com> 100193 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100194 Signed-off-by: Keith Packard <keithp@keithp.com> 100195 100196commit a5dc3531e14589ac473cea482944d2d67517aabd 100197Author: Ville Syrjälä <ville.syrjala@nokia.com> 100198Date: Wed Jan 5 20:41:06 2011 +0200 100199 100200 Revert "composite: Convert compWindowUpdate to use TraverseTree" 100201 100202 TraverseTree visits the parent before the children. When performing 100203 the automatic redirection updates, the children must be visited before 100204 the parent. 100205 100206 If there are automatically redirected windows on multiple levels of the 100207 tree, updating the parents before the children would cause the parent 100208 updates to use stale data for areas covered by the children. Also 100209 updating the damaged children would re-damage the parent, which would 100210 cause additional walks over the tree. 100211 100212 In the worst case with an unbroken chain of automatically redirected 100213 subwindows, all of which are damaged, only the leaf window would be 100214 properly updated on the first round. Then it's parent would be properly 100215 updated on the second round, and so on. And on every round all of the 100216 ancestor windows would be updated as well, but with stale data. 100217 So with N damaged windows you would end up with (N^2+N)/2 updates, 100218 instead of the expected N. 100219 100220 This reverts commit 648c8871c92727d7b6b16859f27f12266a06a16e. 100221 100222 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 100223 Reviewed-by: Keith Packard <keithp@keithp.com> 100224 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100225 Signed-off-by: Keith Packard <keithp@keithp.com> 100226 100227commit 0998574699502e6ab14fd8899c2e42961d4df7d0 100228Author: Ville Syrjälä <ville.syrjala@nokia.com> 100229Date: Wed Jan 5 20:41:05 2011 +0200 100230 100231 Call SourceValidate even if src == dst 100232 100233 The extra SourceValidate calls from damageCopyArea and damageCopyPlane 100234 can be removed. 100235 100236 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 100237 Reviewed-by: Keith Packard <keithp@keithp.com> 100238 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100239 Signed-off-by: Keith Packard <keithp@keithp.com> 100240 100241commit e41e907b3c19908f5316346fa587ced3115478cd 100242Author: Ville Syrjälä <ville.syrjala@nokia.com> 100243Date: Wed Jan 5 20:41:04 2011 +0200 100244 100245 Add subWindowMode parameter to SourceValidate 100246 100247 Pass the subWindowMode from the GC/source Picture to SourceValidate. 100248 100249 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 100250 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100251 Reviewed-by: Keith Packard <keithp@keithp.com> 100252 Signed-off-by: Keith Packard <keithp@keithp.com> 100253 100254commit 0dede200c9ac7adbe8b8c16efacc3edc1f183cd9 100255Merge: 7714357f5 2e781457d 100256Author: Keith Packard <keithp@keithp.com> 100257Date: Wed Jan 5 08:51:46 2011 -0800 100258 100259 Merge remote branch 'vsyrjala/misc_fixes' 100260 100261commit 7714357f506782973d25e270f85140b42507ed35 100262Merge: 02e18c9fb b01dd9d33 100263Author: Keith Packard <keithp@keithp.com> 100264Date: Wed Jan 5 08:43:18 2011 -0800 100265 100266 Merge remote branch 'whot/for-keith' 100267 100268commit 02e18c9fb58c33d340af4573551fb9c7c59e0f43 100269Author: James Jones <jajones@nvidia.com> 100270Date: Mon Dec 20 11:05:57 2010 -0800 100271 100272 X Sync Cleanups 100273 100274 Various cleanups identified during review of the 100275 X Sync Fence Object patches. 100276 100277 -Correctly handle failure of AddResource() 100278 100279 -Don't assert when data structures are corrupt. Instead, 100280 use a new helper function to check for counter sync 100281 objects when they're expected, and warn if the type is 100282 wrong. 100283 100284 -Use the default switch label rather than reimplementing 100285 it. 100286 100287 -Re-introduce cast of result of dixAllocateObjectWithPrivate() 100288 to kill an incompatible pointer type warning. 100289 100290 -Remove comments claiming protocol updates are needed. One 100291 wasn't true and the other was addressed with a xextproto 100292 change. 100293 100294 -Return BadFence, not BadCounter from XSyncAwaitFence() 100295 100296 Signed-off-by: James Jones <jajones@nvidia.com> 100297 Reviewed-by: Keith Packard <keithp@keithp.com> 100298 Signed-off-by: Keith Packard <keithp@keithp.com> 100299 100300commit 86ca434a1ab766c1519ee7146b48dbfead843524 100301Author: Zhao Yakui <yakui.zhao@intel.com> 100302Date: Tue Jan 4 14:08:16 2011 -0500 100303 100304 edid: Fix incorrect timings for VIC61 100305 100306 Reviewed-by: Adam Jackson <ajax@redhat.com> 100307 Signed-off-by: Anssi Hannula <anssi.hannula@iki.fi> 100308 Signed-off-by: Zhao Yakui <yakui.zhao@intel.com> 100309 100310commit bb18f277156c08be028a6e12d8987fb1593e9168 100311Author: Adam Jackson <ajax@redhat.com> 100312Date: Fri Dec 10 14:24:02 2010 -0500 100313 100314 x86emu: Fix more mis-decoding of the data prefix 100315 100316 cc2c73ddcb4370a7c3ad439cda4da825156c26c9's three-cent titanium tax 100317 doesn't go too far enough. Fix the rest of the call and jmp 100318 instructions to handle the data prefix correctly. 100319 100320 Reference: Intel 64 and IA-32 Architectures Software Developer's Manual 100321 Volume 2A: Instruction Set Reference, A-M 100322 100323 http://www.intel.com/Assets/PDF/manual/253666.pdf 100324 100325 Reviewed-by: Julien Cristau <jcristau@debian.org> 100326 Signed-off-by: Adam Jackson <ajax@redhat.com> 100327 100328commit 261d0d16af797bb52d4c778e220296d7f2b28e14 100329Author: Alan Coopersmith <alan.coopersmith@oracle.com> 100330Date: Thu Dec 30 16:08:48 2010 -0800 100331 100332 Fix asprintf typo in commit d2c42b10 100333 100334 Reported-by: Jerome Carretero <cJ@zougloub.eu> 100335 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 100336 Reviewed-by: Cyril Brulebois <kibi@debian.org> 100337 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100338 100339commit c6afe64bad7a27ff8828de499d56776c113df60f 100340Author: Daniel Stone <daniel@fooishbar.org> 100341Date: Fri Dec 31 03:15:27 2010 +0000 100342 100343 XFree86: Linux: AGP: Fix void-pointer-arithmetic warning 100344 100345 Cast it to a char *, mimicking the return immediately below it. 100346 100347 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100348 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 100349 100350commit 2d67ada3c4079a11c52024a9c3d4138becca5171 100351Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100352Date: Thu Dec 30 19:19:43 2010 +0200 100353 100354 os: always check if client is local when connection is accepted 100355 100356 LocalClient is used for all DRI2 requests that makes it frequently 100357 called function. Querying if connection is local or not takes 10-15us 100358 (on ARM) depending on malloc speed. 100359 100360 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100361 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100362 100363commit 617b7d22115ccaaaa7ec69c99885054d33a3bc37 100364Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100365Date: Thu Dec 30 19:19:42 2010 +0200 100366 100367 os: Fix a memory leak 100368 100369 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100370 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 100371 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100372 100373commit 3e1455505addc7e52178fa04399aef2a8522c921 100374Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100375Date: Thu Dec 30 19:19:41 2010 +0200 100376 100377 udev: Fix removing of the wake up handler 100378 100379 RemoveBlockAndWakeupHandlers requires caller to pass same block data 100380 parameter as for RegisterBlockAndWakeupHandlers. 100381 100382 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100383 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100384 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100385 100386commit 13ac3deedb532b39f41dcd45d70f9eb4a85c1f58 100387Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100388Date: Thu Dec 30 19:19:40 2010 +0200 100389 100390 mi: Only register sprite block handler when needed 100391 100392 miSpriteBlockHandler takes about 10us in arm each time BlockHandlers are 100393 called. To eliminate that overhead from xserver side only register the 100394 BlockHandler when there actually is any cursor down. 100395 100396 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100397 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100398 100399commit 4f6e3b0f378d7306dbd8c00ef9a7df81e24e5769 100400Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100401Date: Thu Dec 30 19:19:39 2010 +0200 100402 100403 mi: Register sprite damage handler only when required 100404 100405 There is no need to have damage handler registered when there isn't any 100406 pointer in the screen. This avoids some overhead from damage handling 100407 which takes tens of microseconds on arm. 100408 100409 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100410 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100411 100412commit bf48082f30818b96bc623834be3022600371d4fc 100413Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100414Date: Thu Dec 30 19:19:38 2010 +0200 100415 100416 mi: Fix the debug message 100417 100418 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100419 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100420 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100421 100422commit 8fb43b8bf9fcbe015d4e98c7e09889184d136a1e 100423Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100424Date: Thu Dec 30 19:19:37 2010 +0200 100425 100426 mi: Fix wrapping for sprite screen 100427 100428 SCREEN_EPILOGUE should read the current function pointer from screen 100429 after calling wrapped function in case the pointer changes. 100430 100431 Passing pPriv to macros instead of asking dix each time makes sense when 100432 both macros need same pointer. 100433 100434 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100435 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100436 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100437 100438commit 4be2a6d1b33fac7ceb50ad1de249a1acd5a46ab0 100439Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100440Date: Thu Dec 30 19:19:36 2010 +0200 100441 100442 mi: Add GetSpriteScreen macro to hide dixLookupPrivate 100443 100444 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100445 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100446 100447commit ed8db09b4bff5c4a8d8056808368650e1876547b 100448Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100449Date: Thu Dec 30 19:19:35 2010 +0200 100450 100451 composite: Remove unnecessary variable. 100452 100453 We know that there is damage if the dynamic block handler is registered. 100454 100455 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100456 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100457 100458commit c038b8b28e5c436cd31bea7a3bef2c8c9d818b58 100459Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100460Date: Thu Dec 30 19:19:34 2010 +0200 100461 100462 composite: Only register the block handler when it is required 100463 100464 Even calling block handler that doesn't do much is costly in arm. It 100465 takes a few microseconds each time which adds up to relative high CPU 100466 time because it is done 500+ times per second. 100467 100468 Simple optimization is to register the block handler only when it is 100469 required. 100470 100471 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100472 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100473 100474commit 6d0e9e5d6e1b847961ab52a11aae96981a1cf1c0 100475Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100476Date: Thu Dec 30 19:19:33 2010 +0200 100477 100478 render: Enable animated cursor block handler only when needed 100479 100480 Calling BlockHandlers takes some time for each iteration in main loop 100481 which adds up quickly over multiple request. To reduce the round-trip 100482 costs to xserver BlockHandlers should be registered only when required. 100483 100484 AnimCurScreenBlockHandler is the first victim for this optimization. 100485 100486 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100487 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100488 100489commit aa8cea953dc66bcf4cb4d08f2681f9e6cb1bc8c5 100490Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100491Date: Thu Dec 30 19:19:32 2010 +0200 100492 100493 render: Don't filter 0 return from GetTimeInMillis 100494 100495 In animate cursor block handler code assumes GetTimeInMillis returns 100496 always nonzero value. This isn't true when time wraps around. 100497 100498 To prevent any problems in case GetTimeInMillis would return zero use 100499 activeDevice variable to track if we have received time. 100500 100501 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 100502 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100503 100504commit 91beeee05f88eed10ab0fd97dc625e96cb7763ba 100505Author: Daniel Stone <daniel@fooishbar.org> 100506Date: Tue Oct 13 16:27:30 2009 +1100 100507 100508 DIX: Reset window properties when freeing them 100509 100510 This enables us to reliably inspect properties when destroying windows. 100511 100512 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100513 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 100514 100515commit 36ebdd361616eedbe4919deae2a4e6f6606ae6b4 100516Author: Adam Jackson <ajax@redhat.com> 100517Date: Tue Oct 19 11:59:24 2010 -0400 100518 100519 linux: warning fix 100520 100521 lnx_agp.c: In function ‘xf86DeallocateGARTMemory’: 100522 lnx_agp.c:267: warning: cast to pointer from integer of different size 100523 100524 Signed-off-by: Adam Jackson <ajax@redhat.com> 100525 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100526 100527commit 63ccaec2c39f4b5742383472c951ee2cd35c9e14 100528Author: Adam Jackson <ajax@redhat.com> 100529Date: Tue Oct 19 11:59:23 2010 -0400 100530 100531 glx: warning fixes 100532 100533 render2.c: In function ‘__glXDisp_Map2d’: 100534 render2.c:127: warning: ‘u1’ may be used uninitialized in this function 100535 render2.c: In function ‘__glXDisp_Map1d’: 100536 render2.c:90: warning: ‘u1’ may be used uninitialized in this function 100537 100538 Remove unnecessary test, and change memcpy to memmove as all users were 100539 doing overlapping copies. 100540 100541 Signed-off-by: Adam Jackson <ajax@redhat.com> 100542 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100543 100544commit d057e265c862f90e9b6fddff89f40cb82b2a59b1 100545Author: Adam Jackson <ajax@redhat.com> 100546Date: Tue Oct 19 11:59:22 2010 -0400 100547 100548 xcmisc: warning fix 100549 100550 xcmisc.c:202: warning: no previous prototype for ‘XCMiscExtensionInit’ 100551 100552 Signed-off-by: Adam Jackson <ajax@redhat.com> 100553 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100554 100555commit 67e0bcba4307de5a330e027f4504c9aed4987e5a 100556Author: Adam Jackson <ajax@redhat.com> 100557Date: Tue Oct 19 11:59:21 2010 -0400 100558 100559 dri1: warning fix 100560 100561 dri.c: In function ‘DRIScreenInit’: 100562 dri.c:434: warning: cast from pointer to integer of different size 100563 100564 Signed-off-by: Adam Jackson <ajax@redhat.com> 100565 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100566 100567commit a298d044f965e5ba91f178c6b599c1df26a958ba 100568Author: Adam Jackson <ajax@redhat.com> 100569Date: Tue Oct 19 11:59:20 2010 -0400 100570 100571 loader: warning fix 100572 100573 loadmod.c: In function ‘FreeSubdirs’: 100574 loadmod.c:377: warning: passing argument 1 of ‘free’ discards qualifiers 100575 from pointer target type 100576 /usr/include/stdlib.h:488: note: expected ‘void *’ but argument is of 100577 type ‘const char *’ 100578 100579 Signed-off-by: Adam Jackson <ajax@redhat.com> 100580 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100581 100582commit ee0b1b564413327c73bb8cdbabb76667ca4a744c 100583Author: Adam Jackson <ajax@redhat.com> 100584Date: Tue Oct 19 11:59:18 2010 -0400 100585 100586 config: warning fixes 100587 100588 xf86AutoConfig.c: In function ‘FreeList’: 100589 xf86AutoConfig.c:123: warning: passing argument 1 of ‘free’ discards 100590 qualifiers from pointer target type 100591 /usr/include/stdlib.h:488: note: expected ‘void *’ but argument is of 100592 type ‘const char *’ 100593 100594 Signed-off-by: Adam Jackson <ajax@redhat.com> 100595 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100596 100597commit 32c85ad4b8fe27fed0c494f69c39e3902ce57b1b 100598Author: Daniel Stone <daniel@fooishbar.org> 100599Date: Tue Dec 28 12:40:31 2010 +0000 100600 100601 GLX: DRI2: Fix mismatched-types warning 100602 100603 All the DRI extension types have a base extension type as their first 100604 member to avoid exactly these types of warning. 100605 100606 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100607 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 100608 Reviewed-by: Cyril Brulebois <kibi@debian.org> 100609 100610commit 22796cfa4805cc9551e1b3fa1d3e2e1bfae5bad1 100611Author: Daniel Stone <daniel@fooishbar.org> 100612Date: Tue Dec 28 12:29:53 2010 +0000 100613 100614 udev: Add strdups to kill const warnings 100615 100616 InputAttributes wants non-const members, and while it appears safe to 100617 cast it, just leave it be for the moment. 100618 100619 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100620 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 100621 100622commit 03f2eb1e156796afb70118d7f7f60ac61beed026 100623Author: Daniel Stone <daniel@fooishbar.org> 100624Date: Fri Dec 17 17:13:34 2010 +0000 100625 100626 Input: Set client error value for invalid mask bits 100627 100628 When we send BadValue back to the client for having invalid mask bits, 100629 at least tell them what the (first) invalid bit was. 100630 100631 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100632 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 100633 100634commit 47818287983b04625661b08ae362e0966cddf3e1 100635Author: Chase Douglas <chase.douglas@canonical.com> 100636Date: Fri Dec 17 17:13:30 2010 +0000 100637 100638 Make EventIsDeliverable non-static 100639 100640 Will be used outside dix/events.c in proceeding XI 2.1 MT changes. 100641 100642 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 100643 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100644 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 100645 100646commit f7d8ade3c5ac44faf7f3ea0d846e35f75b31c9ec 100647Author: Daniel Stone <daniel@fooishbar.org> 100648Date: Fri Dec 17 17:13:26 2010 +0000 100649 100650 Resources: Move rClient to resource.h 100651 100652 The definition of rClient was duplicated across three source files, so 100653 move it to resource.h. 100654 100655 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100656 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 100657 100658commit 05e437c2ce230b588362f012c06ead5c6d89c006 100659Author: Chase Douglas <chase.douglas@canonical.com> 100660Date: Fri Dec 17 17:13:27 2010 +0000 100661 100662 Export all valuator_mask_* functions 100663 100664 Input drivers may use valuator masks for internal state. Having all the 100665 valuator_mask_* functions available will help. 100666 100667 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 100668 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100669 100670 Bump ABI_XINPUT_VERSION minor. 100671 100672 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 100673 100674commit 2e781457d43ec4bf0d633257ac6852cde3b00541 100675Author: Ville Syrjälä <ville.syrjala@nokia.com> 100676Date: Thu Dec 16 18:22:13 2010 +0200 100677 100678 xfree86/modes: Take rotation into account when checking mode size 100679 100680 Assume that a mode can be used in either landscape or portrait 100681 orientation. I suppose the correct thing to do would be to 100682 collect all the supported rotations from the CRTCs that can be used 100683 with a specific output, but that information doesn't seem to be 100684 readily available when these checks are done. So just assume that 100685 either orientation is fine. 100686 100687 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 100688 Reviewed-by: Adam Jackson <ajax@redhat.com> 100689 100690commit 0ce25fd7904c792924c3e0ee6fc21a5f1bec1a68 100691Author: Ville Syrjälä <ville.syrjala@nokia.com> 100692Date: Fri Dec 3 17:42:16 2010 +0200 100693 100694 dri2: Don't page flip when the window size doesn't match the pixmap size 100695 100696 If the drawable size doesn't match the pixmap size page flipping should 100697 not be allowed. 100698 100699 If the window is larger than the pixmap, page flipping might need to 100700 reposition the CRTC somewhere in the middle of the pixmap. I didn't 100701 spot any code that would handle that at least in the intel driver. 100702 100703 Also the root pixmap could then move to some negative screen 100704 coordinates. Not sure if all bits of code could handle that. Perhaps 100705 when composite is enabled screen_x/y would make it work, but without 100706 composite there's no way that it would work AFAICS. 100707 100708 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 100709 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 100710 100711commit efcb63d0ce43f96d0ac02b6f4a480dfd2374fc84 100712Author: Daniel Stone <daniel@fooishbar.org> 100713Date: Thu Dec 23 13:33:00 2010 +0000 100714 100715 Render: Fix 'comparing between distinct pointer types' warning 100716 100717 Add the appropriate casts so that gcc shuts up, even if it doesn't 100718 matter. 100719 100720 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100721 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 100722 100723commit 469d5bf8b75038631c27edbb0f9cdf7d737fa233 100724Author: Daniel Stone <daniel@fooishbar.org> 100725Date: Fri Dec 17 16:48:45 2010 +0000 100726 100727 mi: Sync: Don't free managed screen private 100728 100729 misync allocates space for its screen private with 100730 dixRegisterPrivateKey, which means it doesn't have to free it at 100731 CloseScreen time; doing so will, in fact, result in a crash. 100732 100733 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100734 Reviewed-by: James Jones <jajones@nvidia.com> 100735 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 100736 100737commit 44adb31bfece29260a9bbd9075c9212ebf00d24d 100738Author: Daniel Stone <daniel@fooishbar.org> 100739Date: Fri Aug 27 20:36:37 2010 +1000 100740 100741 GetTimeInMillis: Use CLOCK_MONOTONIC_COARSE where available 100742 100743 On some systems, using CLOCK_MONOTONIC forces a readback of HPET or some 100744 similarly expensive timer. CLOCK_MONOTONIC_COARSE can alleviate this, 100745 at the cost of negligibly-reduced resolution, so prefer that where we 100746 can. 100747 100748 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100749 Reviewed-by: Julien Cristau <jcristau@debian.org> 100750 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 100751 100752commit 296561506a91742cc150a0fb6fc0df5dbe98c780 100753Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 100754Date: Wed Dec 22 16:57:17 2010 +0200 100755 100756 Xext: Use general OS functions to determine client command string in SELinux. 100757 100758 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 100759 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 100760 100761commit 1e933665bef26c74196bb7c59910e6a78bcacf0e 100762Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 100763Date: Wed Dec 22 16:51:09 2010 +0200 100764 100765 os: Add facilities for client ID tracking. 100766 100767 An interface is provided for figuring out the PID and process name of 100768 a client. Make some existing functionality from SELinux and IA 100769 extensions available for general use. 100770 100771 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 100772 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 100773 100774commit 671b2a1823a1c90d0b6254e2e2af8865151fff8c 100775Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 100776Date: Wed Dec 22 16:38:54 2010 +0200 100777 100778 config: Fix linking order of Xnest libraries. 100779 100780 MAIN depends on DIX and not vice versa. 100781 100782 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 100783 100784commit 2b364bf970b2ce6829af656990c33afd0d365f3c 100785Author: Rami Ylimäki <rami.ylimaki@vincit.fi> 100786Date: Tue Jun 15 14:44:38 2010 +0200 100787 100788 Revert "os: Prevent backtrace from being stopped in noreturn functions." 100789 100790 This reverts commit 579715f830fbbca9e1ecb17dc18176132f5969e7. 100791 100792 The patch is not needed anymore. I haven't encountered backtrace 100793 problems with GCC 4.3.3. Even if the problems still persisted, this 100794 commit should be removed and instead the definition of _X_NORETURN 100795 should be modified to be empty if GCC/ARM is used. However, currently 100796 it seems that ARM backtraces are OK even if _X_NORETURN is used and 100797 -mapcs-frame is not defined in CFLAGS. 100798 100799 Signed-off-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 100800 Reviewed-by: Jamey Sharp <jamey@minilop.net> 100801 100802commit b01dd9d33651999b0390bf8ed412b84f6929a13e 100803Author: Daniel Stone <daniel@fooishbar.org> 100804Date: Fri Dec 17 17:13:34 2010 +0000 100805 100806 Input: Set client error value for invalid mask bits 100807 100808 When we send BadValue back to the client for having invalid mask bits, 100809 at least tell them what the (first) invalid bit was. 100810 100811 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100812 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 100813 100814commit 495e0142a77152c4cde5c88cdba4103ddc94df2c 100815Author: Chase Douglas <chase.douglas@canonical.com> 100816Date: Fri Dec 17 17:13:30 2010 +0000 100817 100818 Make EventIsDeliverable non-static 100819 100820 Will be used outside dix/events.c in proceeding XI 2.1 MT changes. 100821 100822 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 100823 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100824 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 100825 100826commit 4fbe67beb56e5600c1df789c08725c45b9bef0d4 100827Author: Daniel Stone <daniel@fooishbar.org> 100828Date: Fri Dec 17 17:13:26 2010 +0000 100829 100830 Resources: Move rClient to resource.h 100831 100832 The definition of rClient was duplicated across three source files, so 100833 move it to resource.h. 100834 100835 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 100836 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 100837 100838commit 871c1e17e527857cd86b5e517d9a18e5041f00fd 100839Author: Chase Douglas <chase.douglas@canonical.com> 100840Date: Fri Dec 17 17:13:27 2010 +0000 100841 100842 Export all valuator_mask_* functions 100843 100844 Input drivers may use valuator masks for internal state. Having all the 100845 valuator_mask_* functions available will help. 100846 100847 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 100848 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100849 100850 Bump ABI_XINPUT_VERSION minor. 100851 100852 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 100853 100854commit e06fa804009798ea95efa8babaabb0228dfdfe65 100855Author: Michel Dänzer <daenzer@vmware.com> 100856Date: Wed Dec 22 11:45:36 2010 +0100 100857 100858 EXA: Fix crash with fill using 1x1 tile of depth < 8 (bug #24703). 100859 100860 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=24703 . 100861 100862 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 100863 Reviewed-by: Dave Airlie <airlied@redhat.com> 100864 Signed-off-by: Keith Packard <keithp@keithp.com> 100865 100866commit 56d9c736f0c450bffd43300107689eb6dd1ea70b 100867Author: Adam Jackson <ajax@redhat.com> 100868Date: Thu Dec 16 14:26:10 2010 -0500 100869 100870 xfree86: Convert some xf86BlockSIGIO to OsBlockSignals 100871 100872 The DDC1 and int10 code are blocking SIGIO to get some assurance that 100873 their usleep() calls take as long as they expect. That's a good start 100874 but you really want to be blocking more than just SIGIO, SIGALRM too at 100875 minimum. 100876 100877 At this point, except for SIGIO handler setup itself, BlockSIGIO really 100878 means "block input events". 100879 100880 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 100881 Signed-off-by: Adam Jackson <ajax@redhat.com> 100882 100883commit ce13a1dbbfbeee3a3d51d0385817df50944cfdee 100884Author: Adam Jackson <ajax@redhat.com> 100885Date: Thu Dec 16 14:03:13 2010 -0500 100886 100887 xfree86: Remove xf86EnterServerState 100888 100889 Back when we had RAC this was a vaguely meaningful thing. Since then 100890 it's been a glorified (and confusing) wrapper around xf86BlockSIGIO. 100891 100892 Note that the APM and VT switch code are unusual relative to other code 100893 that cares about SIGIO state. Most callers push a SIGIO disable to 100894 create a critical section for the duration of the caller's stack frame, 100895 but those two effectively disable SIGIO after their return and re-enable 100896 on their next entry. 100897 100898 Reviewed-by: Tiago Vignatti <tigo.vignatti@nokia.com> 100899 Signed-off-by: Adam Jackson <ajax@redhat.com> 100900 100901commit a04e74cb480f98bc3dd4def2da6f6752640d9e38 100902Author: Adam Jackson <ajax@redhat.com> 100903Date: Wed Dec 15 15:16:33 2010 -0500 100904 100905 xfree86: Symmetrize xf86AccessEnter and xf86AccessLeave 100906 100907 Enter was changing server operating state, Leave wasn't. Which was 100908 wholly redundant, since all callers of Enter would immediately change 100909 the operating state to exactly what Enter had just done. 100910 100911 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 100912 Signed-off-by: Adam Jackson <ajax@redhat.com> 100913 100914commit 7a9062f2f029b4f911ba56f291375fbf5a98ca73 100915Author: Adam Jackson <ajax@redhat.com> 100916Date: Tue Dec 14 15:44:16 2010 -0500 100917 100918 fonts: Fix typo in async ListFonts logic 100919 100920 This was introduced in 3ab6cd31cbdf8095b2948034fce5fb645422d8da. Mea 100921 culpa. This logic is still incorrect [1], but at least it's less 100922 incorrect. 100923 100924 [1] - https://bugzilla.redhat.com/658587 100925 100926 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100927 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100928 Signed-off-by: Adam Jackson <ajax@redhat.com> 100929 100930commit f28e48834e40c7901c2efc72f962c9724e74a531 100931Author: Adam Jackson <ajax@redhat.com> 100932Date: Thu Dec 16 13:13:54 2010 -0500 100933 100934 Remove SCO support 100935 100936 This has never been buildable in any modular server release. 100937 100938 Reviewed-by: Julien Cristau <jcristau@debian.org> 100939 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 100940 Signed-off-by: Adam Jackson <ajax@redhat.com> 100941 100942commit 040d54fb88f3970e19d0fbf911d6447f592496bd 100943Author: Adam Jackson <ajax@redhat.com> 100944Date: Wed Dec 15 14:59:13 2010 -0500 100945 100946 xfree86: Remove unused xf86inSuspend 100947 100948 What is this, I don't even. 100949 100950 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 100951 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100952 Signed-off-by: Adam Jackson <ajax@redhat.com> 100953 100954commit e7dc253452a1ba64718a08fdc070405b494f53cd 100955Merge: d1107918d d5b0d5857 100956Author: Keith Packard <keithp@keithp.com> 100957Date: Mon Dec 20 00:48:45 2010 -0800 100958 100959 Merge remote branch 'alanc/master' 100960 100961commit d1107918d4626268803b54033a07405122278e7f 100962Author: Tiago Vignatti <tiago.vignatti@nokia.com> 100963Date: Fri Dec 17 16:09:35 2010 +0200 100964 100965 randr: check for virtual size limits before set crtc 100966 100967 Return a error if the screen is configured to an invalid size. 100968 100969 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 100970 Reviewed-by: Adam Jackson <ajax@redhat.com> 100971 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 100972 Signed-off-by: Keith Packard <keithp@keithp.com> 100973 100974commit 4b88c7be8de4149fe3d166bf115775f9e81a1373 100975Author: Jörn Horstmann <launchpad@planetxml.de> 100976Date: Mon Dec 6 11:24:02 2010 +1100 100977 100978 Add EDID quirk for HP Compaq nc8430. 100979 100980 Like some other LPL panels, this one reports the vertical size in cm rather 100981 than mm. 100982 Patch taken from Launchpad bug #380009 <https://launchpad.net/bugs/380009> 100983 100984 X.Org Bug 28414 <https://bugs.freedesktop.org/show_bug.cgi?id=28414> 100985 100986 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 100987 Reviewed-by: Adam Jackson <ajax@redhat.com> 100988 Signed-off-by: Keith Packard <keithp@keithp.com> 100989 100990commit a2e67a6412386782cb8b644b86a5744591397d45 100991Author: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 100992Date: Mon Dec 6 11:24:01 2010 +1100 100993 100994 IDLETIME: Fix edge-case in IdleTimeBlockHandler 100995 100996 Ensure that if we're called exactly on the threshold of a 100997 NegativeTransition trigger that we reshedule to pick up 100998 an idle time over the threshold. 100999 101000 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 101001 Reviewed-by: Adam Jackson <ajax@redhat.com> 101002 Signed-off-by: Keith Packard <keithp@keithp.com> 101003 101004commit b55bf248581dc66321b24b29f199f6dc8d02db1b 101005Author: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 101006Date: Mon Dec 6 11:24:00 2010 +1100 101007 101008 Xext: Fix edge case with {Positive, Negative}Transition triggers. 101009 101010 The {Positive,Negative}Transition triggers only fire when the counter 101011 goes from strictly {below,above} the threshold. If 101012 SyncComputeBracketValues gets called exactly at this threshold we may update 101013 the bracket values so that the counter is not updated past the threshold. 101014 101015 Signed-off-by: Christopher James Halse Rogers <christopher.halse.rogers@canonical.com> 101016 Reviewed-by: Adam Jackson <ajax@redhat.com> 101017 Signed-off-by: Keith Packard <keithp@keithp.com> 101018 101019commit d5b0d58573ffb90ea70273cd013102168fd2a18a 101020Author: Alan Coopersmith <alan.coopersmith@oracle.com> 101021Date: Thu Dec 16 22:29:37 2010 -0800 101022 101023 Bug 32436 - hw/xfree86/parser/Makefile.am TOP_SRCDIR should be top_srcdir 101024 101025 https://bugs.freedesktop.org/show_bug.cgi?id=32436 101026 Fix typo introduced in 2416255f7e3fd9190a9 that breaks builds when 101027 configured --enable-install-libxf86config 101028 101029 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101030 Reviewed-by: Julien Cristau <jcristau@debian.org> 101031 Tested-by: Simon Thum <simon.thum@gmx.de> 101032 101033commit 058b889fde47b4c32534f11ed651bf1749d1dbb2 101034Author: David Barksdale <david.barksdale@adcedosolutions.com> 101035Date: Mon Dec 13 18:35:20 2010 -0800 101036 101037 Bug 21827: faulty sdksyms.c generated by the awk script in sdksyms.sh 101038 101039 https://bugs.freedesktop.org/show_bug.cgi?id=21827 101040 101041 Tested-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101042 Signed-off-by: David Barksdale <david.barksdale@adcedosolutions.com> 101043 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101044 101045commit 1324b0ca9f8a7fdaf03b374c75eb3c9df407c2f1 101046Author: Colin Harrison <colin.harrison@virgin.net> 101047Date: Fri Dec 10 00:08:24 2010 +0000 101048 101049 Typo in xserver Xvasprintf() 101050 101051 I needed this patch in the wrapper around vsnprintf() in os/xprintf.c 101052 (MinGW for Windows build) to correct various crashes. 101053 101054 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101055 101056commit f1542f1d716723cba7c323849086585635121893 101057Merge: 9716d3124 8a8fdd762 101058Author: Keith Packard <keithp@keithp.com> 101059Date: Tue Dec 14 15:04:12 2010 -0800 101060 101061 Merge remote branch 'whot/for-keith' 101062 101063commit 9716d3124799c6db0d1c782aa72c72f972d5a158 101064Merge: 03ea0b772 f641e4b34 101065Author: Keith Packard <keithp@keithp.com> 101066Date: Tue Dec 14 15:01:36 2010 -0800 101067 101068 Merge remote branch 'jeremyhu/master' 101069 101070commit 03ea0b7726487b96df73936518da73d023e51334 101071Author: Gaetan Nadon <memsize@videotron.ca> 101072Date: Tue Dec 7 16:31:55 2010 -0500 101073 101074 docs: replace BUILLDOCS with existing ENABLE_DEVEL_DOCS 101075 101076 Use the util-macros AM Conditionals to control generation of developers 101077 documents. This is used throughout xorg modules. 101078 101079 The doxygen generated docs are now also managed by --enable-devel-docs. 101080 101081 Remove --enable-builddocs as this was last use for BUILDDOCS 101082 101083 *** From the RELEASE NOTES *** 101084 101085 New configure options for documentation in modules 101086 -------------------------------------------------- 101087 As many more modules now contain documentation to be converted from DocBook XML to text, 101088 HTML, PostScript, and/or PDF formats, new standard options have been added to the configure 101089 macros to control the build of these in the modules. 101090 --with-xmlto=yes|no 101091 Enables or disables use of the xmlto [https://fedorahosted.org/ 101092 xmlto/] command to translate DocBook XML to other formats. 101093 All DocBook XML conversions require use of this command. 101094 --with-fop=yes|no 101095 Enables or disables use of the Apache fop [http:// 101096 xmlgraphics.apache.org/fop/] command to translate DocBook 101097 XML to PostScript and PDF formats. 101098 --enable-docs=yes|no 101099 Enables or disables the build and installation of all 101100 documentation except traditional man pages or those covered 101101 by the --enable-devel-docs and --enable-specs options. 101102 --enable-devel-docs=yes|no 101103 Enables or disables the build and installation of documentation 101104 for developers of the X.Org software modules. 101105 --enable-specs=yes|no 101106 Enables or disables the build and installation of the formal 101107 specification documents for protocols and APIs. 101108 101109 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 101110 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 101111 Signed-off-by: Keith Packard <keithp@keithp.com> 101112 101113commit c78c71c3a872665e426a0a4b82f17007bd8c12cd 101114Author: Gaetan Nadon <memsize@videotron.ca> 101115Date: Tue Dec 7 16:31:54 2010 -0500 101116 101117 hw/dmx/doc: DIST_SUBDIRS not required when using AM_CONDITIONAL 101118 101119 The makefile can be simplyfied as Automake handle this automatically. 101120 All directories in SUBDIRS are visited for dist purposes. 101121 101122 www.gnu.org/software/automake/manual/automake.html#SUBDIRS-vs-DIST_005fSUBDIRS 101123 101124 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 101125 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 101126 Signed-off-by: Keith Packard <keithp@keithp.com> 101127 101128commit 8a8fdd762ad89c350854943311ec4aadc50245fa 101129Author: Peter Hutterer <peter.hutterer@who-t.net> 101130Date: Mon Dec 13 11:28:30 2010 +1000 101131 101132 xfree86: always report the input options before initialising the device. 101133 101134 After collecting the driver's default options, report the list of options 101135 set for the device before calling PreInit(). This helps with debugging those 101136 cases where options are not merged correctly. 101137 101138 xf86OptionListReport reports with verbosity 5, higher than the default 101139 verbosity so this won't generate logspam in the default case. 101140 101141 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 101142 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 101143 Reviewed-by: Simon Thum <simon.thum@gmx.de> 101144 101145commit 9db9e964f6ca553dcbd3b7b037745d9581eaa065 101146Author: Peter Hutterer <peter.hutterer@who-t.net> 101147Date: Fri Dec 10 13:23:13 2010 +1000 101148 101149 xfree86: swap the order to-be-merged lists in xf86CollectInputOptions. 101150 101151 Current order causes the user-configured option list to be overwritten with 101152 the default list supplied by the driver. Swap around so we overwrite the 101153 driver's default values instead. 101154 101155 This only affected options supplied by the driver such as XkbLayout in the 101156 case of evdev. 101157 101158 Reported-by: Sebastian Glita <glseba@yahoo.com> 101159 Reported-by: Simon Thum <simon.thum@gmx.de> 101160 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 101161 Reviewed-by: Simon Thum <simon.thum@gmx.de> 101162 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 101163 Tested-by: Sebastian Glita <glseba@yahoo.com> 101164 101165commit 0d440a1c6e219cd39dbddd2b7e813c6431aac6ea 101166Author: Peter Hutterer <peter.hutterer@who-t.net> 101167Date: Wed Dec 8 14:52:19 2010 +1000 101168 101169 dix: allow for button-only input devices (#21457) 101170 101171 Add a few checks for the existence of a valuator class on the device to 101172 avoid null-pointer dereferences for button events from devices without a 101173 valuator class. 101174 101175 X.Org Bug 21457 <http://bugs.freedesktop.org/show_bug.cgi?id=21457> 101176 101177 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 101178 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 101179 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 101180 101181commit f641e4b34aa91ecda29e546b8b975e72ce037ed0 101182Author: Jeremy Huddleston <jeremyhu@apple.com> 101183Date: Wed Dec 8 22:58:12 2010 -0800 101184 101185 XQuartz GL: Add $(GL_CFLAGS) to CFLAGS 101186 101187 This fixes a build failure I found on tinderbox. 101188 101189 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 101190 101191commit 9cf055892dd413932e54b43cc2dfea70bafd525f 101192Author: Peter Hutterer <peter.hutterer@who-t.net> 101193Date: Wed Dec 8 14:43:51 2010 +1000 101194 101195 xfree86: don't set movement flags for non-valuator events. 101196 101197 If a device doesn't send valuators, don't try to move its position. 101198 101199 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 101200 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 101201 101202commit aba8133c9c5a50753c388d76407868ac69f4134b 101203Author: Peter Hutterer <peter.hutterer@who-t.net> 101204Date: Wed Dec 8 14:02:17 2010 +1000 101205 101206 dix: clear up an overly convoluted if statement. 101207 101208 No functional changes, just improves readability. This statement had things 101209 added to/removed from it for a few server releases while the input event 101210 queue was revamped. What made sense once is now mainly confusing. 101211 101212 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 101213 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 101214 101215commit 883039e07b0c0ef23c7fc4ba3455b41173c7a7fc 101216Author: Jeremy Huddleston <jeremyhu@apple.com> 101217Date: Wed Dec 8 13:27:41 2010 -0800 101218 101219 XQuartz: Just NSBeep() for XBell() 101220 101221 The CoreAudio path uses deprecated API and has reported crashes that 101222 aren't worth fixing (4e8bf12b13690afa6d9fee0e339d3819ef16fb3f fixed 101223 one and introduced another). NSBeep() does the job just fine. 101224 101225 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 101226 101227commit 2c70b650b342378898064cf27e2f95c4b2e53d24 101228Author: Jon TURNEY <jon.turney@dronecode.org.uk> 101229Date: Wed Dec 8 18:54:30 2010 +0000 101230 101231 Fix Xwin build after af0f9f91 101232 101233 Fix damage to XWIN_LIBS in commit af0f9f913398d34a885c3fb4e8d40c1a7e2b3ee9 101234 "Move some sync code to miext", which broke the XWin build 101235 101236 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 101237 Reviewed-by: Colin Harrison <colin.harrison at virgin.net> 101238 Signed-off-by: Keith Packard <keithp@keithp.com> 101239 101240commit 519d243f0c8e3c80226701f71d2cfa62e42dbff7 101241Merge: 9f9c73231 446482efa 101242Author: Keith Packard <keithp@keithp.com> 101243Date: Tue Dec 7 13:39:58 2010 -0800 101244 101245 Merge remote branch 'alanc/master' 101246 101247commit 446482efaa3d266266e2a143492a3ec0523622bd 101248Author: Alan Coopersmith <alan.coopersmith@oracle.com> 101249Date: Sun Nov 28 09:41:17 2010 -0800 101250 101251 Convert cvt code to use XNFasprintf() 101252 101253 Requires linking xprintf.c into standalone cvt utility 101254 101255 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101256 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 101257 Reviewed-by: Julien Cristau <jcristau@debian.org> 101258 101259commit d2c42b102707db6c1ca9dccc4de9210a43650cb1 101260Author: Alan Coopersmith <alan.coopersmith@oracle.com> 101261Date: Sat Nov 27 23:10:46 2010 -0800 101262 101263 Replace alloc+strcpy+strcat with asprintf() & XNFasprintf() calls 101264 101265 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101266 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 101267 101268commit 3a9bb93dd178084f4ff1abcea331ca5a62f88ce6 101269Author: Alan Coopersmith <alan.coopersmith@oracle.com> 101270Date: Sat Nov 27 22:38:27 2010 -0800 101271 101272 Convert alloc+sprintf pairs into asprintf() & XNFasprintf() calls 101273 101274 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101275 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 101276 101277commit 2416255f7e3fd9190a9e01bda57c992932de4bd9 101278Author: Alan Coopersmith <alan.coopersmith@oracle.com> 101279Date: Sat Nov 27 22:36:44 2010 -0800 101280 101281 Convert hw/xfree86/parser code to use asprintf() calls 101282 101283 Requires linking xprintf.c into libxf86config for those who build it. 101284 101285 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101286 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 101287 101288commit 03e8bfa1d122f7dea905d48c93cfd54afd991dfd 101289Author: Alan Coopersmith <alan.coopersmith@oracle.com> 101290Date: Sat Nov 27 20:09:04 2010 -0800 101291 101292 Convert existing Xprintf style calls to asprintf style 101293 101294 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101295 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 101296 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 101297 101298commit c95c1d338fdb62dbe3dba934b97324fa778b7fce 101299Author: Alan Coopersmith <alan.coopersmith@oracle.com> 101300Date: Sat Nov 27 18:43:12 2010 -0800 101301 101302 Add asprintf() implementation for platforms without it 101303 101304 Provides a portable implementation of this common allocating sprintf() 101305 API found in many, but not yet all, of the platforms we support. 101306 If the platform provides vasprintf() we simply wrap it, otherwise we 101307 implement it - either way callers can use it regardless of platform. 101308 101309 Since not all platforms guarantee to NULL out the return pointer on 101310 failure, we don't either, and require callers to check the return 101311 value for -1. 101312 101313 The old Xprintf() API is deprecated, but left for compatibility for now. 101314 101315 The new API is added in a new header so that it can be used in parts of 101316 the server such as hw/xfree86/parser that don't include all the server 101317 headers. 101318 101319 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101320 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 101321 101322commit 2db6951763115a4a5dfac53d1a8f78ac33f91235 101323Author: Alan Coopersmith <alan.coopersmith@oracle.com> 101324Date: Fri Dec 3 16:32:01 2010 -0800 101325 101326 Sun's copyrights now belong to Oracle 101327 101328 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101329 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 101330 101331commit f06e8a61d6ed8f3054878167bffee619f852aad9 101332Author: Alan Coopersmith <alan.coopersmith@oracle.com> 101333Date: Tue Nov 30 16:34:38 2010 -0800 101334 101335 Move inclusion of panoramiX headers to top of render.c 101336 101337 Fixes compiler error from Sun compilers due to _X_EXPORT declaration 101338 being included after the unlabeled version: 101339 101340 "../Xext/panoramiXsrv.h", line 29: redeclaration must have the same or more restrictive linker scoping: XRT_PICTURE 101341 101342 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101343 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 101344 101345commit 9f9c732311ef0cc198e5ec44d02412eb7e8c3f3a 101346Author: Cyril Brulebois <kibi@debian.org> 101347Date: Fri Nov 12 21:35:25 2010 +0100 101348 101349 dix: Fix logic for displaying deprecation warning only once. 101350 101351 As pointed out by Jamey Sharp (again), the logic is faulty: --warn is 101352 always going to be false. Replace it with warn-- accordingly, so that 101353 there's (at least, but also only) one warning showing up. 101354 101355 Reviewed-by: Jamey Sharp <jamey@minilop.net> 101356 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101357 101358commit a937803c1f671ef29332e5fe8c190d8b48239912 101359Author: Cyril Brulebois <kibi@debian.org> 101360Date: Fri Nov 12 21:29:26 2010 +0100 101361 101362 dix: Simplify deprecated *Lookup* wrappers around dixLookup*. 101363 101364 As pointed out by Jamey Sharp: “the result pointer is already guaranteed 101365 to be NULL if the return value is not Success”, so get rid of the 101366 variable used to catch the return value, and used in a ternary operation 101367 to decide whether to return the pointer or NULL. Always return the 101368 result pointer instead. 101369 101370 Reviewed-by: Jamey Sharp <jamey@minilop.net> 101371 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101372 101373commit 73fbc4a4a7997b3ee1c779d8f394114270bcb20d 101374Author: Cyril Brulebois <kibi@debian.org> 101375Date: Fri Nov 12 12:04:11 2010 +0100 101376 101377 Replace SecurityLookupIDByClass() with dixLookupResourceByClass(). 101378 101379 This patch has been prepared with the following Coccinelle semantic patch: 101380 101381 @@ 101382 expression a, b, c, d; 101383 identifier r; 101384 type t; 101385 @@ 101386 -r = (t) SecurityLookupIDByClass(a, b, c, d); 101387 +dixLookupResourceByClass((pointer*) &r, b, c, a, d); 101388 101389 The only occurrence not matching directly was processed after separating 101390 declaration of pXinDraw and call to SecurityLookupIDByClass(). 101391 101392 Reviewed-by: Jamey Sharp <jamey@minilop.net> 101393 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101394 101395commit 6e8b34b7048bc9e49ef52b81e2de1b518a05890e 101396Author: Cyril Brulebois <kibi@debian.org> 101397Date: Fri Nov 12 11:36:36 2010 +0100 101398 101399 Replace SecurityLookupIDByType() with dixLookupResourceByType(), take 2. 101400 101401 These occurrences are a bit harder to catch through a semantic patch, 101402 so process them “manually”. 101403 101404 Reviewed-by: Jamey Sharp <jamey@minilop.net> 101405 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101406 101407commit ac04c1377a42ab6803adf551756653fd8e1a694b 101408Author: Cyril Brulebois <kibi@debian.org> 101409Date: Fri Nov 12 11:36:31 2010 +0100 101410 101411 Replace SecurityLookupIDByType() with dixLookupResourceByType(), take 1. 101412 101413 This patch has been prepared with the following Coccinelle semantic patch: 101414 101415 @@ 101416 expression a, b, c, d; 101417 identifier r; 101418 @@ 101419 -r = SecurityLookupIDByType(a, b, c, d); 101420 +dixLookupResourceByType((pointer*) &r, b, c, a, d); 101421 101422 Reviewed-by: Jamey Sharp <jamey@minilop.net> 101423 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101424 101425commit 28b4c02d7463ffe5c29f803f5656c5ecfc040185 101426Author: Cyril Brulebois <kibi@debian.org> 101427Date: Thu Nov 11 18:36:45 2010 +0100 101428 101429 Enable silent rules for cat and echo. 101430 101431 One gets this: 101432 | GEN lf3-xaaBitmap.c 101433 | GEN lf3-xaaBitmap.c 101434 | GEN lf3-xaaBitmap.c 101435 | GEN lf3-xaaBitmap.c 101436 | CC lf3-xaaBitmap.lo 101437 101438 instead of this: 101439 | echo "#define LSBFIRST" > l3-xaaBitmap.c 101440 | echo "#define TRIPLE_BITS" >> l3-xaaBitmap.c 101441 | echo '#include "../../../../hw/xfree86/xaa/xaaBitmap.c"' >> l3-xaaBitmap.c 101442 | CC l3-xaaBitmap.lo 101443 101444 Occurrences found using: 101445 | git grep -E '(cat|echo)' -- '*Makefile.am' 101446 101447 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 101448 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101449 101450commit 838b9582280030a159878af566126cb0a523ec29 101451Author: Cyril Brulebois <kibi@debian.org> 101452Date: Thu Nov 11 15:40:13 2010 +0100 101453 101454 Replace LookupIDByType() with dixLookupResourceByType(), take 2. 101455 101456 These occurrences are a bit harder to catch through a semantic patch, 101457 so process them “manually”. 101458 101459 Reviewed-by: Jamey Sharp <jamey@minilop.net> 101460 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101461 101462commit ffb93533fd8067c7383aef5e0018a197af59b46c 101463Author: Cyril Brulebois <kibi@debian.org> 101464Date: Thu Nov 11 15:30:15 2010 +0100 101465 101466 Replace LookupIDByType() with dixLookupResourceByType(), take 1. 101467 101468 This patch has been prepared with the following Coccinelle semantic patch: 101469 101470 @@ 101471 expression x, y; 101472 type t; 101473 identifier r; 101474 @@ 101475 -r = (t) LookupIDByType(x, y); 101476 +dixLookupResourceByType((pointer*)&r, x, y, NullClient, DixUnknownAccess); 101477 101478 @@ 101479 expression x, y; 101480 type t; 101481 identifier r; 101482 @@ 101483 -t r = (t) LookupIDByType(x, y); 101484 +t r; 101485 +dixLookupResourceByType((pointer*)&r, x, y, NullClient, DixUnknownAccess); 101486 101487 Reviewed-by: Jamey Sharp <jamey@minilop.net> 101488 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101489 101490commit d1b45b0fd54efb952f9ff435a516c1bfd85ca186 101491Author: Cyril Brulebois <kibi@debian.org> 101492Date: Thu Nov 11 11:35:47 2010 +0100 101493 101494 Fix missing <string.h> include. 101495 101496 The following happens otherwise (with -Wall -Werror): 101497 | In file included from /usr/include/X11/Xfuncs.h:47, 101498 | from ../../include/misc.h:112, 101499 | from ../../include/screenint.h:52, 101500 | from ../../include/scrnintstr.h:52, 101501 | from ../../dix/cursor.c:58: 101502 | /usr/include/string.h:534: error: conflicting types for ‘xstrcasecmp’ 101503 | ../../include/os.h:488: note: previous declaration of ‘xstrcasecmp’ was here 101504 | /usr/include/string.h:538: error: conflicting types for ‘xstrncasecmp’ 101505 | ../../include/os.h:493: note: previous declaration of ‘xstrncasecmp’ was here 101506 101507 Reviewed-by: Julien Cristau <jcristau@debian.org> 101508 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101509 101510commit 780754050bc9cb1489f92a2a890ab5665e3e6358 101511Author: Cyril Brulebois <kibi@debian.org> 101512Date: Fri Nov 12 12:18:59 2010 +0100 101513 101514 Fix GLX_LIBS vs. DMX_LIBS order. 101515 101516 If glxproxy needs symbols which aren't pulled in by dmx itself, glxproxy 101517 fails to link because of undefined references. 101518 101519 Signed-off-by: Cyril Brulebois <kibi@debian.org> 101520 Reviewed-by: Julien Cristau <jcristau@debian.org> 101521 Signed-off-by: Keith Packard <keithp@keithp.com> 101522 101523commit 14983286c489ea1ec6ba4fadbeaec5c2d2ee34e1 101524Author: Keith Packard <keithp@keithp.com> 101525Date: Mon Dec 6 20:45:35 2010 -0800 101526 101527 Version bumped to 1.9.99.901 (1.10 RC1) 101528 101529 Signed-off-by: Keith Packard <keithp@keithp.com> 101530 101531commit d96d5f5443358e33e47d1a61f9bd2afd8064a8dd 101532Author: James Jones <jajones@nvidia.com> 101533Date: Sun Dec 5 19:42:41 2010 -0800 101534 101535 Bump extension ABI to 5 101536 101537 Commit 606e079cc4d9a9db3197652ca51683c36f74efb8 moved the visual 101538 field in WindowOptRec, breaking the extension module ABI. 101539 101540 Signed-off-by: James Jones <jajones@nvidia.com> 101541 Reviewed-by: Keith Packard <keithp@keithp.com> 101542 Signed-off-by: Keith Packard <keithp@keithp.com> 101543 101544commit 435361bd73b9fc733f093d81af6b839953e35176 101545Author: Keith Packard <keithp@keithp.com> 101546Date: Mon Dec 6 20:38:14 2010 -0800 101547 101548 sync: syncObject may be None in SyncInitTrigger 101549 101550 And often is, especially when called from ProcSyncCreateAlarm. 101551 Crashing in this case seems unwise. 101552 101553 Signed-off-by: Keith Packard <keithp@keithp.com> 101554 101555commit 0d01b66df9081ef48843b3bad81c56bb2cd1ae69 101556Author: Keith Packard <keithp@keithp.com> 101557Date: Sun Dec 5 23:35:28 2010 -0800 101558 101559 randr: handle RRSetCrtcConfigs request with zero configs 101560 101561 Need to actually return a reply in this case. 101562 101563 Signed-off-by: Keith Packard <keithp@keithp.com> 101564 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101565 101566commit b0f4bd61f0caf80f3be9a176f1f7a707bc6628d8 101567Author: Keith Packard <keithp@keithp.com> 101568Date: Sun Dec 5 21:53:25 2010 -0800 101569 101570 ProcRRSetCrtcConfigs uses 'configs' without being initialized 101571 101572 If the client sends invalid data for this request, the server 101573 will jump to 'sendReply' and call RRFreeCrtcConfigs, passing it the 101574 uninitialized 'configs' and 'num_configs' values. 101575 101576 Signed-off-by: Keith Packard <keithp@keithp.com> 101577 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101578 101579commit 752c368421c1c824752cf467fba9318d75d2ca2c 101580Author: Keith Packard <keithp@keithp.com> 101581Date: Sun Dec 5 20:57:47 2010 -0800 101582 101583 Separate out screen size and screen pixmap sizes in RRScreenSizeSet 101584 101585 This provides for separate sizes for the screen scanout and rendering 101586 buffer and the application-visible screen size. 101587 101588 Signed-off-by: Keith Packard <keithp@keithp.com> 101589 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101590 101591commit a88d70fb20a2bc3152b84adff4380857e6cfadf5 101592Author: Keith Packard <keithp@keithp.com> 101593Date: Sun Dec 5 20:55:46 2010 -0800 101594 101595 Set sprite transforms from RRSetCrtcConfigs 101596 101597 These were getting ignored. 101598 101599 Signed-off-by: Keith Packard <keithp@keithp.com> 101600 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101601 101602commit 96b4d4787bf82edd9d06eb9a6e94bc45412c7df2 101603Author: Keith Packard <keithp@keithp.com> 101604Date: Sun Dec 5 20:49:19 2010 -0800 101605 101606 DIX is responsible for ref counting scanout pixmaps. 101607 101608 Remove some extra ref counting inside hw/xfree86/modes 101609 101610 Signed-off-by: Keith Packard <keithp@keithp.com> 101611 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101612 101613commit afb6ebf1d5829346c40fe1053c9f50afe926e6c6 101614Author: Keith Packard <keithp@keithp.com> 101615Date: Fri Dec 3 13:04:37 2010 -0800 101616 101617 randr: Hook up the new RandR 1.4 functionality 101618 101619 This bumps the supported RandR protocol version and adds the dispatch 101620 hooks needed to call the new functions 101621 101622 Signed-off-by: Keith Packard <keithp@keithp.com> 101623 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101624 101625commit 82612045e11f2b882ae132e184a9629f43f1c424 101626Author: Keith Packard <keithp@keithp.com> 101627Date: Fri Dec 3 13:00:46 2010 -0800 101628 101629 randr: Add per-crtc pixmaps 101630 101631 This adds new driver hooks to allocate scanout pixmaps and 101632 changes the mode setting APIs to pass the new scanout pixmaps 101633 along from DIX. DIX is responsible for reference counting the pixmaps 101634 by tracking them through RRCrtcNotify. 101635 101636 Signed-off-by: Keith Packard <keithp@keithp.com> 101637 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101638 101639commit 86c489c319c705f710bee3897fe27600ce15008e 101640Author: Keith Packard <keithp@keithp.com> 101641Date: Sat Dec 4 20:12:26 2010 -0800 101642 101643 hw/xfree86/modes: Add optional driver API for RRSetCrtcConfigs 101644 101645 This provides a driver hook which can either completely replace, or 101646 just validate the parameters for, the RRSetCrtcConfigs request. 101647 101648 Signed-off-by: Keith Packard <keithp@keithp.com> 101649 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101650 101651commit d94a035ea9eb3167fc4f35b2d9f0d53f8807014c 101652Author: Keith Packard <keithp@keithp.com> 101653Date: Sat Dec 4 19:22:11 2010 -0800 101654 101655 randr: Implement RRSetCrtcConfigs 101656 101657 This provides a driver-independent implementation of the 101658 RRSetCrtcConfigs API by simply using the existing interfaces. 101659 101660 Signed-off-by: Keith Packard <keithp@keithp.com> 101661 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101662 101663commit 66294afcab7b7a82f7dd897767e46c48a94b8ee8 101664Author: Keith Packard <keithp@keithp.com> 101665Date: Fri Dec 3 22:08:06 2010 -0800 101666 101667 randr: Add sprite position transforms 101668 101669 This implements sprite position transformations. Sprite image 101670 transforms are passed all the way to the DDX layer, but the images are 101671 not yet manipulated before being passed to the drivers. 101672 101673 Signed-off-by: Keith Packard <keithp@keithp.com> 101674 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101675 101676commit c8bc25fd7629df10f2825b7cc713b031ae78f223 101677Author: Keith Packard <keithp@keithp.com> 101678Date: Mon Dec 6 19:40:30 2010 -0800 101679 101680 Require RandR protocol version 1.4 or newer 101681 101682 Signed-off-by: Keith Packard <keithp@keithp.com> 101683 101684commit 8b35118c03590a7ad3786d3284bafb3f40fcb8cc 101685Author: Keith Packard <keithp@keithp.com> 101686Date: Sun Jun 6 15:23:37 2010 -0700 101687 101688 Replace huge argument list in xf86CrtcSetModeTransform with struct 101689 101690 xf86CrtcSetModeTransform was starting to get ridiculous with 6 101691 arguments, this change has it take a single structure that contains 101692 all of those values along with a set of flags that says which have 101693 changed. 101694 101695 Signed-off-by: Keith Packard <keithp@keithp.com> 101696 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 101697 101698commit f5b8bd620f91214c0b87e9b04d57015655792352 101699Merge: 01e9fa7da 8127c83c8 101700Author: Keith Packard <keithp@keithp.com> 101701Date: Mon Dec 6 19:51:06 2010 -0800 101702 101703 Merge remote branch 'jajones/for-keith' 101704 101705commit 01e9fa7da389fc7ab834b4234b8484514144b7f4 101706Merge: 79870dbf7 68a1b0de9 101707Author: Keith Packard <keithp@keithp.com> 101708Date: Mon Dec 6 19:45:11 2010 -0800 101709 101710 Merge remote branch 'whot/for-keith' 101711 101712commit 79870dbf72d463cf9efea1b3e61d1d5759df27b5 101713Author: Gaetan Nadon <memsize@videotron.ca> 101714Date: Sat Dec 4 16:13:29 2010 -0500 101715 101716 config: build Xserver man pages using XORG_MANPAGE_SECTIONS 101717 101718 Nothing requires the use of a C preprocessor 101719 Using standard file exentions (.man) means no need for .gitignore 101720 Use standard directory and makefile 101721 Fix man page whitespace issues 101722 101723 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 101724 Reviewed-by: Matt Dew <matt@osource.org> 101725 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 101726 Signed-off-by: Keith Packard <keithp@keithp.com> 101727 101728commit c4a7a5917ab828d4a0bd825a98e4d641bcb378f1 101729Merge: 5de312a60 714b68d9e 101730Author: Keith Packard <keithp@keithp.com> 101731Date: Mon Dec 6 19:28:07 2010 -0800 101732 101733 Merge remote branch 'jeremyhu/master' 101734 101735commit 5de312a60d8227ed670849ecf888ea878aa81430 101736Merge: 311cad331 8f42b2b69 101737Author: Keith Packard <keithp@keithp.com> 101738Date: Mon Dec 6 19:22:52 2010 -0800 101739 101740 Merge remote branch 'alanc/master' 101741 101742commit 8127c83c81bf64369a8ba2999088226d14e0b128 101743Author: James Jones <jajones@nvidia.com> 101744Date: Mon Dec 6 12:14:01 2010 -0800 101745 101746 Expose Sync Fence Object protocol 101747 101748 Add the new protocol handlers for XSync 3.1 to the 101749 dispatch tables and report support for Sync protocol 101750 version 3.1. 101751 101752 Signed-off-by: James Jones <jajones@nvidia.com> 101753 Reviewed-by: Keith Packard <keithp@keithp.com> 101754 101755commit 9c0c7cc9a7adf230b5eee0b4166f9e17e07a088b 101756Author: James Jones <jajones@nvidia.com> 101757Date: Mon Dec 6 13:10:42 2010 -0800 101758 101759 Add XSyncAwaitFence() handler 101760 101761 -Add the actual ProcSyncAwaitFence() dispatch func 101762 101763 -Add support for fence sync triggers. 101764 101765 Signed-off-by: James Jones <jajones@nvidia.com> 101766 Reviewed-by: Keith Packard <keithp@keithp.com> 101767 101768commit 397dfd9f87e1cdf105d10a789a97230f4f0d204e 101769Author: James Jones <jajones@nvidia.com> 101770Date: Fri Jun 25 17:59:14 2010 -0700 101771 101772 Create/Destroy/Trigger/Reset/Query Fence Sync objs 101773 101774 Initial server side implementation of fence sync 101775 objects. Allows creation, management, and state 101776 queries of binary state objects. Currently they 101777 are not very useful as there is no way to wait for 101778 them efficiently. 101779 101780 The basic trigger operation added here triggers 101781 relative to a given X screen's rendering operations. 101782 To perform this operation, fence sync objects must 101783 be tied to a screen. As Aaron Plattner pointed out, 101784 screens are identified but a drawable in X protocol, 101785 so a drawable argument is included in 101786 XSyncCreateFence(). The screen also could have been 101787 specified as part of the trigger operation. However, 101788 it is also desireable to associate a screen with 101789 fence sync objects at creation time so that the 101790 associated screen's driver can allocate any HW- 101791 specific resources needed by the fence object up 101792 front. 101793 101794 Signed-off-by: James Jones <jajones@nvidia.com> 101795 Reviewed-by: Keith Packard <keithp@keithp.com> 101796 101797commit 53ea965d64bd9563e6bfcc81d04f283d2b85f09d 101798Author: James Jones <jajones@nvidia.com> 101799Date: Mon Nov 29 10:09:36 2010 -0800 101800 101801 Require xextproto 7.1.99 101802 101803 Subsequent changes rely on fence sync protocol 101804 in the sync extension. This protocol is only 101805 complete in xextproto version 7.1.99 and 101806 above. 101807 101808 Signed-off-by: James Jones <jajones@nvidia.com> 101809 Reviewed-by: Keith Packard <keithp@keithp.com> 101810 101811commit 1c4a0db2c6bf0320cb630b84ab87bcfd3801a53d 101812Author: James Jones <jajones@nvidia.com> 101813Date: Mon Jun 28 16:10:13 2010 -0700 101814 101815 Add fence sync driver interface 101816 101817 -Add fence sync objects 101818 101819 -Add fence sync devPrivates 101820 101821 -Add a X sync module screen private 101822 101823 -Add wrappable functions to create and destroy 101824 fence sync objects 101825 101826 -Give fence sync objects wrappable functions to 101827 trigger, test, and reset their 'triggered' value. 101828 101829 -Give fence sync objects wrappable functions to 101830 notify driver when adding/removing triggers to/ 101831 from the sync object. 101832 101833 Signed-off-by: James Jones <jajones@nvidia.com> 101834 Reviewed-by: Keith Packard <keithp@keithp.com> 101835 101836commit af0f9f913398d34a885c3fb4e8d40c1a7e2b3ee9 101837Author: James Jones <jajones@nvidia.com> 101838Date: Mon Jun 28 15:39:04 2010 -0700 101839 101840 Move some sync code to miext 101841 101842 As a precursor to the fence sync object video driver 101843 and extension API, move some code from Xext to 101844 miext/sync. Most of this is just code to set up the 101845 build system to include the new directory. No 101846 functional code is added in this change. 101847 101848 Signed-off-by: James Jones <jajones@nvidia.com> 101849 Reviewed-by: Keith Packard <keithp@keithp.com> 101850 101851commit 12b65de7db6e3e8bf831914d247da269d01c5fbe 101852Author: James Jones <jajones@nvidia.com> 101853Date: Mon Dec 6 10:11:45 2010 -0800 101854 101855 Factor out generic code from ProcSyncAwait() 101856 101857 In preparation for adding more sync object types 101858 that will need Await requests of their own, factor 101859 out some setup and finalization code from 101860 ProcSyncAwait() into SyncAwaitPrologue() and 101861 SyncAwaitEpilogue() 101862 101863 Signed-off-by: James Jones <jajones@nvidia.com> 101864 Reviewed-by: Keith Packard <keithp@keithp.com> 101865 101866commit c66a410d378090f350beb398649e9d9262933785 101867Author: James Jones <jajones@nvidia.com> 101868Date: Mon Jun 28 14:59:01 2010 -0700 101869 101870 Make Await SyncTrigger functions generic 101871 101872 Update all the functions dealing with Await 101873 sync triggers handle generic sync objects 101874 instead of just counters. This will 101875 facilitate code sharing between the counter 101876 sync waits and the fence sync waits. 101877 101878 Signed-off-by: James Jones <jajones@nvidia.com> 101879 Reviewed-by: Keith Packard <keithp@keithp.com> 101880 101881commit 99daf419a3070c347c0cd29b4a9bae07eca7bd6d 101882Author: James Jones <jajones@nvidia.com> 101883Date: Mon Jun 28 14:51:54 2010 -0700 101884 101885 Create SyncObject base type. 101886 101887 SyncObject is now the base type for SyncCounter. 101888 Data to be used by all sync types is stored in 101889 the base object. SyncCounter can be safely cast 101890 to SyncObject, and a SyncObject can be cast to 101891 the correct type based on SyncObject::type. 101892 101893 Signed-off-by: James Jones <jajones@nvidia.com> 101894 Reviewed-by: Keith Packard <keithp@keithp.com> 101895 101896commit 27593eea7efcbed8de0c6e8233cbd1a1b8a50459 101897Author: James Jones <jajones@nvidia.com> 101898Date: Mon Dec 6 13:36:14 2010 -0800 101899 101900 Add and use SERVER_SYNC_*_VERSION 101901 101902 Most extensions have a version defined 101903 in the protocol headers, and also in the 101904 server's protocol-versions.h. The latter 101905 defines which version the server advertises 101906 support for. Sync wasn't included in 101907 protocol-versions.h, and was advertising 101908 support for whatever was in the protocol 101909 headers the server was built against. 101910 101911 Signed-off-by: James Jones <jajones@nvidia.com> 101912 Reviewed-by: Keith Packard <keithp@keithp.com> 101913 101914commit 68a1b0de95f71f74835c6c0f002699fcdccbb268 101915Author: Peter Hutterer <peter.hutterer@who-t.net> 101916Date: Mon Dec 6 14:58:37 2010 +1000 101917 101918 xfree86: use xf86AllocateInput() for xorg.conf devices too. 101919 101920 Single allocation point for input devices, most notably a single point to 101921 reset default values. 101922 Without this patch, the file descriptor default was -1 for hotplugged 101923 devices and 0 for config devices. Drivers that don't overwrite the default 101924 themselves would thus fail if configured in the xorg.conf. 101925 101926 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 101927 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 101928 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 101929 101930commit beea2378f142556471c62290e275935af848e137 101931Author: Peter Hutterer <peter.hutterer@who-t.net> 101932Date: Mon Dec 6 14:33:43 2010 +1000 101933 101934 xfree86: don't overwrite option list (#32115) 101935 101936 Options set in the configuration file were unconditionally overwritten by 101937 the server. Merge the already existing options and the new options together 101938 instead of just overwriting ones. 101939 101940 Introduced in commit 2199842ed50b3eb40d54146827fc58cae7e873ec 101941 Author: Peter Hutterer <peter.hutterer@who-t.net> 101942 Date: Thu Sep 2 10:52:54 2010 +1000 101943 101944 xfree86: remove extraOptions field from IDevRec. 101945 101946 X.Org Bug 32115 <http://bugs.freedesktop.org/show_bug.cgi?id=32115> 101947 101948 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 101949 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 101950 Tested-by: David Ronis <ronis@ronispc.chem.mcgill.ca> 101951 101952commit 8f3fa8fb0b0a75dac714fc213c034b20595898d3 101953Author: Peter Hutterer <peter.hutterer@who-t.net> 101954Date: Mon Dec 6 15:44:50 2010 +1000 101955 101956 xfree86: add option "Floating", deprecate SendCoreEvents and friends. 101957 101958 Some devices should be initialised as floating from the start (e.g. 101959 Joysticks and accelerometers benefit from this). Currently users use the 101960 "SendCoreEvents" "off" flag for this, which isn't the most appropriate 101961 naming. 101962 101963 Add an option "Floating", deprecate the others. Still parsed and handled by 101964 the server. 101965 101966 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 101967 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 101968 101969commit 36b614dedf4ddc428e43ad1542d4f9314f73f60a 101970Author: Eoghan Sherry <ejsherry@gmail.com> 101971Date: Sun Nov 28 16:15:51 2010 -0500 101972 101973 Xi: Fix master button update when slave buttons are mapped. #24887 101974 101975 It is currently assumed that an event button delieved to a master device 101976 corresponds to the slave button states. However, the event button is a 101977 logical (mapped) slave button and slave button states correspond to 101978 physical (unmapped) slave buttons. This leads to incorrect update of the 101979 master button state and incorrect events devlivered to clients. Fix the 101980 situation by taking the slave button map into account when querying a 101981 slave button state. 101982 101983 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=24887 101984 101985 Signed-off-by: Eoghan Sherry <ejsherry@gmail.com> 101986 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 101987 101988commit 31ab9f8860848504df18a8be9d19b817b191e0df 101989Author: Peter Hutterer <peter.hutterer@who-t.net> 101990Date: Fri Sep 3 11:54:41 2010 +1000 101991 101992 mi: handle DGA subtypes when determining the master device. 101993 101994 The subtype in the DGA event is the core type and all ET_ event types (where 101995 applicable) are identical to the core types. Thus the switch statement below 101996 will work as required and assign the right master device. 101997 101998 Fixes a crasher bug on keyboard devices with valuators. If a device sends a 101999 motion event while grabbed and a DGA client is active (but has not selected 102000 input through DGA), the valuator event is posted through the VCK and 102001 eventually results in a NULL-pointer dereference on dev->valuator. 102002 102003 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 102004 102005commit 5d31c3e705dfd9f38f0fffcd07a6d8d06644735c 102006Author: Peter Hutterer <peter.hutterer@who-t.net> 102007Date: Fri Nov 26 09:37:37 2010 +1000 102008 102009 dix: don't stop processing valuators when the mode changes. 102010 102011 XI 1.x events still contain absolute coordinates anyway. By the time we get 102012 to the InternalEvent to XI event conversion, the valuators are already 102013 absolute. 102014 102015 Stopping because of a different mode on a valuator is not necessary. 102016 102017 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 102018 Reviewed-by: Adam Jackson <ajax@redhat.com> 102019 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 102020 102021commit 714b68d9e5bf624a6703f168e0f7dc980e88e8c0 102022Author: Jeremy Huddleston <jeremyhu@apple.com> 102023Date: Fri Dec 3 17:27:44 2010 -0800 102024 102025 XQuartz: RandR: Fix mode changing for multi-monitor configurations. 102026 102027 This just fixes the regression whereby we couldn't switch between the legacy 102028 fullscreen mode and rootless on multi-monitor configurations. This was 102029 happening because ref wasn't being set in these cases (since we don't ever 102030 actually change CG modes), so we failed a CFEqual. Setting the references 102031 fixes this regression and places us one step closer to more mode RandR 102032 mode switching in multi-monitor configurations. 102033 102034 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 102035 102036commit 14f00449eb81771c01fffcdaf3dd697cdf4e41de 102037Author: Jeremy Huddleston <jeremyhu@apple.com> 102038Date: Fri Dec 3 16:46:11 2010 -0800 102039 102040 XQuartz: Cleanup some compiler warnings 102041 102042 Mark __crashreporter_info__ as __attribute__((__used__)) 102043 102044 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 102045 102046commit 14aec8b82a316c6df51ac2a81985ebf1990abd2b 102047Author: Jeremy Huddleston <jeremyhu@apple.com> 102048Date: Fri Dec 3 16:45:37 2010 -0800 102049 102050 XQuartz: Remove one more backing store leftover 102051 102052 See: c4c4676e6874b42c2371eee96faa2c2dbb59a704 102053 102054 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 102055 102056commit ca431371a23a2b9ad36c1d64e11ea41d5e4e5f04 102057Author: Jeremy Huddleston <jeremyhu@apple.com> 102058Date: Fri Nov 26 15:31:22 2010 -0500 102059 102060 XQuartz: Disable the Mac OS X screensaver when in full screen mode 102061 102062 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 102063 102064commit 311cad33155c64ed996418808727fc417168592e 102065Author: Nicolas Peninguy <nico@lostgeeks.org> 102066Date: Wed Nov 24 00:11:11 2010 +0100 102067 102068 Fix Xdmx and Xephyr build when DTrace support is enabled 102069 102070 This fixes the following build errors when DTrace is enabled 102071 (--with-dtrace): 102072 102073 CCLD Xdmx 102074 /usr/bin/ld: ../../os/os.O: undefined reference to symbol 'dladdr@@GLIBC_2.2.5' 102075 /usr/bin/ld: note: 'dladdr@@GLIBC_2.2.5' is defined in DSO /lib64/libdl.so.2 so try adding it to the linker command line 102076 102077 CCLD Xephyr 102078 ../../../os/os.O: In function `TimerForce': 102079 /home/nico/work/xserver/os/WaitFor.c:481: multiple definition of `TimerForce' 102080 ../../../os/os.O:/home/nico/work/xserver/os/WaitFor.c:481: first defined here 102081 102082 Signed-off-by: Nicolas Peninguy <nico@lostgeeks.org> 102083 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102084 Signed-off-by: Keith Packard <keithp@keithp.com> 102085 102086commit 3824417335ec00a87b51f8d05e592099b8507680 102087Merge: dab064fa5 f4f41c812 102088Author: Keith Packard <keithp@keithp.com> 102089Date: Thu Dec 2 08:39:33 2010 -0800 102090 102091 Merge remote branch 'vsyrjala/xv_reput' 102092 102093commit dab064fa5e0b1f5c67222562ad5367005832cba1 102094Author: Andrea Canciani <ranma42@gmail.com> 102095Date: Tue Nov 2 20:10:32 2010 +0100 102096 102097 render: Fix byteswapping of gradient stops 102098 102099 The function swapStops repeatedly swaps the color components as 102100 CARD16, but incorrectly steps over them as if they were CARD32. 102101 102102 This causes half of the stops not to be swapped at all and some 102103 unrelated data be swapped instead. 102104 102105 Signed-off-by: Andrea Canciani <ranma42@gmail.com> 102106 Reviewed-by: Soren Sandmann <sandmann@daimi.au.dk> 102107 Reviewed-by: Julien Cristau <jcristau@debian.org> 102108 Signed-off-by: Keith Packard <keithp@keithp.com> 102109 102110commit 279ef1ffd787dba2f0d5056849b9cb15d36aa3eb 102111Author: Cyril Brulebois <kibi@debian.org> 102112Date: Wed Dec 1 14:12:55 2010 +0100 102113 102114 Fix screen number checks. 102115 102116 screenInfo.numScreens is not a valid screen number, they go from 0 to 102117 numScreens - 1. 102118 102119 Signed-off-by: Cyril Brulebois <kibi@debian.org> 102120 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 102121 Signed-off-by: Keith Packard <keithp@keithp.com> 102122 102123commit 8684543021b9b1aa165b1bc69bc58685cb5942c1 102124Author: Julien Cristau <jcristau@debian.org> 102125Date: Tue Nov 30 17:23:10 2010 +0100 102126 102127 DGA: fix screen number check 102128 102129 screenInfo.numScreens is not a valid screen number, they go from 0 to 102130 numScreens - 1. 102131 102132 Signed-off-by: Julien Cristau <jcristau@debian.org> 102133 Reviewed-by: Keith Packard <keithp@keithp.com> 102134 Signed-off-by: Keith Packard <keithp@keithp.com> 102135 102136commit 8f42b2b69387b006bfcd373c3d023ebea9035db2 102137Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102138Date: Sat Nov 27 22:34:57 2010 -0800 102139 102140 Simplify Error() - don't allocate temporary copy of error string 102141 102142 Doesn't seem to be any reason to just not pass the error string 102143 as another argument directly to LogVWrite() 102144 102145 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102146 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 102147 Reviewed-by: Julien Cristau <jcristau@debian.org> 102148 102149commit 685286b17d30335d799a9da11914943e466ea955 102150Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102151Date: Sat Nov 27 20:43:28 2010 -0800 102152 102153 FindModuleInSubdir: Stop allocating one more byte than needed 102154 102155 15ac25627e7239629be59 removed the "/" from the sprintf strings, 102156 but failed to remove the extra byte allocated for the '/'. 102157 102158 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102159 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 102160 Reviewed-by: Julien Cristau <jcristau@debian.org> 102161 102162commit 40d5a019352fa8f12230c863e11cbb1f6258a93e 102163Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102164Date: Sat Nov 27 19:50:38 2010 -0800 102165 102166 xf86VIDrvMsgVerb: print args, not format string 102167 102168 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102169 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 102170 Reviewed-by: Julien Cristau <jcristau@debian.org> 102171 102172commit 4bbc90cd8b7e749fd8072ce7cd8dd998f4396981 102173Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102174Date: Sat Nov 27 19:06:56 2010 -0800 102175 102176 xf86AutoConfig: make copyScreen memory allocation & error handling more sane 102177 102178 No point calling the no-fail-alloc if you check for failure and your 102179 only caller checks for failure. 102180 102181 No point calling calloc to zero fill memory you're about to memcpy over. 102182 102183 In the unlikely event of a loss of memory allocation, drop your previous 102184 allocations before returning to others. 102185 102186 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102187 Reviewed-by: Julien Cristau <jcristau@debian.org> 102188 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 102189 102190commit 2c8e534c8e9334562485aeaaef374871cf14d5fe 102191Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102192Date: Sat Nov 27 23:49:55 2010 -0800 102193 102194 xf86ValidateModes: xnfalloc(strlen) + strcpy => xnfstrdup 102195 102196 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102197 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 102198 Reviewed-by: Julien Cristau <jcristau@debian.org> 102199 102200commit 29e467a1f1548a826ee2793244e3ff416aa1a0f2 102201Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102202Date: Fri Nov 26 16:53:40 2010 -0800 102203 102204 xf86OutputRename: Replace another strlen/malloc/strcpy set with strdup 102205 102206 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102207 Reviewed-by: Matt Turner <mattst88@gmail.com> 102208 102209commit d346bc3083c6d4bea59b77f634c7c5ec6c1d8cc9 102210Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102211Date: Sun Nov 28 13:48:16 2010 -0800 102212 102213 Fix compiler warnings in hw/xfree86/os-support/solaris 102214 102215 sun_init.c: In function `xf86OpenConsole': 102216 sun_init.c:99: warning: cast does not match function type 102217 sun_init.c:74: warning: unused variable `FreeVTslot' 102218 sun_init.c: In function `xf86UseMsg': 102219 sun_init.c:417: warning: old-style parameter declaration 102220 102221 sun_vid.c: In function `solUnMapVidMem': 102222 sun_vid.c:162: warning: long unsigned int format, pointer arg (arg 6) 102223 sun_vid.c: In function `xf86ReadBIOS': 102224 sun_vid.c:217: warning: long unsigned int format, pointer arg (arg 5) 102225 sun_vid.c:217: warning: long unsigned int format, int arg (arg 6) 102226 102227 sun_agp.c: In function `xf86EnableAGP': 102228 sun_agp.c:321: warning: unsigned int format, CARD32 arg (arg 4) 102229 102230 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102231 Reviewed-by: Julien Cristau <jcristau@debian.org> 102232 102233commit d75777d54c2107163305f50e8ee4306da202b95e 102234Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102235Date: Sun Nov 28 13:45:40 2010 -0800 102236 102237 Move xchomp inside #ifdef __linux__ 102238 102239 static function only called from the matchDriverFromFiles function 102240 that's inside #ifdef __linux__ section 102241 102242 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102243 Reviewed-by: Matt Turner <mattst88@gmail.com> 102244 102245commit ccbba444b7b8e1ba555532a847377600bea43d03 102246Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102247Date: Sun Nov 28 10:45:17 2010 -0800 102248 102249 config: Remove AC_PROG_CC, SED & INSTALL that XORG_DEFAULT_OPTIONS provide 102250 102251 Most importantly removes AC_PROG_CC call that resets compiler flags back 102252 to C89 mode, breaking use of C99 isfinite() on Solaris in dix/devices.c. 102253 102254 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102255 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 102256 102257commit 92cde0b84cdec164cd698dc74a7512d0791d7708 102258Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102259Date: Sat Nov 27 08:15:52 2010 -0800 102260 102261 Xserver-spec: Update discussion of font library 102262 102263 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102264 Reviewed-by: Julien Cristau <jcristau@debian.org> 102265 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102266 102267commit c1e769250078cbc74d7e6e6ddc5323c4f420ab14 102268Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102269Date: Sat Nov 27 08:06:40 2010 -0800 102270 102271 Xserver-spec: Update location of log functions 102272 102273 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102274 Reviewed-by: Julien Cristau <jcristau@debian.org> 102275 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102276 102277commit deae18f3cad94959110761ef2ecd9665690e3db5 102278Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102279Date: Sat Nov 27 00:45:48 2010 -0800 102280 102281 Xserver-spec: Fix assorted typos 102282 102283 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102284 Reviewed-by: Julien Cristau <jcristau@debian.org> 102285 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102286 102287commit a410bf53798bdca43f99476a01ef27cabdf73e01 102288Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102289Date: Sat Nov 27 00:35:28 2010 -0800 102290 102291 Xserver-spec: Update lists of macros 102292 102293 LOOKUP_DRAWABLE & VERIFY_GC are no longer in dix.h, but 102294 WriteReplyToClient & WriteSwappedDataToClient are. 102295 102296 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102297 Reviewed-by: Julien Cristau <jcristau@debian.org> 102298 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102299 102300commit de518c8f378ea31345c946693d58a26a493af603 102301Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102302Date: Sat Nov 27 00:30:37 2010 -0800 102303 102304 Xserver-spec: Remove CreateCallbackList 102305 102306 The function is defined as a static, so can't be called by anyone 102307 but AddCallback. 102308 102309 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102310 Reviewed-by: Julien Cristau <jcristau@debian.org> 102311 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102312 102313commit f6c880b257a21a574cf1a47095cb39f32252802e 102314Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102315Date: Sat Nov 27 00:27:46 2010 -0800 102316 102317 Xserver-spec: Replace deprecated resource id lookup functions 102318 102319 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102320 Reviewed-by: Julien Cristau <jcristau@debian.org> 102321 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102322 102323commit 922806a5aa6eafc432d6787495b475aaa3f1790d 102324Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102325Date: Sat Nov 27 00:14:51 2010 -0800 102326 102327 Xserver-spec: Update Memory Management functions 102328 102329 Xalloc, Xrealloc, & Xfree are deprecated now 102330 ALLOCATE_LOCAL is removed due to stack overflow issues 102331 102332 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102333 Reviewed-by: Julien Cristau <jcristau@debian.org> 102334 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102335 102336commit 02449ee24b484c9fea501df5274d95a9f87cab23 102337Merge: afd6eb66d b85f9063c 102338Author: Keith Packard <keithp@keithp.com> 102339Date: Tue Nov 30 13:30:59 2010 -0800 102340 102341 Merge remote branch 'ajax/for-keithp' 102342 102343commit afd6eb66d5ef6a59be4c03da4588aaa2ee47b9ad 102344Merge: b16964910 ffcbfa006 102345Author: Keith Packard <keithp@keithp.com> 102346Date: Tue Nov 30 13:28:13 2010 -0800 102347 102348 Merge remote branch 'aplattner/for/keith' 102349 102350commit b16964910d29c0bd039e8bb48bcf1199d709fe3e 102351Author: Ferry Huberts <ferry.huberts@pelagic.nl> 102352Date: Tue Nov 30 19:06:55 2010 +0100 102353 102354 dix: do not use bit-wise operators on the boolean result of BitIsOn 102355 102356 Performing bit-wise operations on a boolean amounts to mixing types, 102357 is confusing and basically incorrect; one should only perform 102358 logical operations on booleans. 102359 102360 Performing such operations relies on the implementation detail 102361 that a boolean is in fact an integer and that its value FALSE 102362 is implemented as zero. 102363 102364 Signed-off-by: Ferry Huberts <ferry.huberts@pelagic.nl> 102365 Reviewed-by: Matt Turner <mattst88@gmail.com> 102366 Reviewed-by: Keith Packard <keithp@keithp.com> 102367 Signed-off-by: Keith Packard <keithp@keithp.com> 102368 102369commit 5f34853d6f936e8a1a317a59972b07e3f191d1ab 102370Author: Adam Jackson <ajax@redhat.com> 102371Date: Mon Nov 15 11:25:13 2010 +1000 102372 102373 xfree86: Bump classic driver default to 1024x768 102374 102375 Signed-off-by: Adam Jackson <ajax@redhat.com> 102376 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 102377 Reviewed-by: Keith Packard <keithp@keithp.com> 102378 Signed-off-by: Keith Packard <keithp@keithp.com> 102379 102380commit 903e0f6f0f0ec54151640fc459ff03dcba0522af 102381Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102382Date: Sat Nov 27 10:46:44 2010 -0800 102383 102384 Xserver-spec: Delete DBE Idioms section 102385 102386 The code has been gone for a while 102387 102388 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102389 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102390 Reviewed-by: Keith Packard <keithp@keithp.com> 102391 Signed-off-by: Keith Packard <keithp@keithp.com> 102392 102393commit f4ba75a494cf9f4f5b0a979e9ceb519f4a6fcacc 102394Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102395Date: Sat Nov 27 10:46:43 2010 -0800 102396 102397 Remove unused ReqLen & CastxReq macros 102398 102399 According to Xserver-spec, they were part of the now-deleted DBE "Idioms" 102400 code. The last callers of them were removed in commits fe616f9230b6 & 102401 3d642905477f. 102402 102403 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102404 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102405 Reviewed-by: Keith Packard <keithp@keithp.com> 102406 Signed-off-by: Keith Packard <keithp@keithp.com> 102407 102408commit b85f9063c2370d7710523bccb8bf030a0710425b 102409Author: Adam Jackson <ajax@redhat.com> 102410Date: Mon Nov 22 16:42:04 2010 -0500 102411 102412 xfree86: Remove the xf8_16bpp overlay module 102413 102414 Only one (marginal) driver was using it, and it's been fixed to just 102415 implement it directly. 102416 102417 v2: Also fix sdksyms.sh (spotted by Jesse Adkins) 102418 v3: Also fix DESIGN.xml (spotted bu Julien Cristau) 102419 102420 Reviewed-by: Julien Cristau <jcristau@debian.org> 102421 Signed-off-by: Adam Jackson <ajax@redhat.com> 102422 102423commit a861fe6a1cf56df76928d21707551602289963c1 102424Author: Adam Jackson <ajax@redhat.com> 102425Date: Tue Nov 30 13:50:42 2010 -0500 102426 102427 vbe: Refuse to believe tiny (or negative) panel sizes from PanelID 102428 102429 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 102430 Signed-off-by: Adam Jackson <ajax@redhat.com> 102431 102432commit 84a14fab8f930ef1855444ae4e9e3e14ee008328 102433Author: David Reveman <davidr@novell.com> 102434Date: Wed Jan 6 14:55:48 2010 -0500 102435 102436 composite: add panoramix support 102437 102438 Taken from: 102439 50d2d8c8969c165582d215c6e85c4be9eac02b6a 102440 dbffd0d44a33dcc84898c7a891d7ba212f65cbb8 102441 9b5b102163b4eaa1b70647354fcab4f6e461c94c 102442 75f9b98af31abf537ac6616c99f3797deb7ba017 102443 07fba8b1f77a6bca44ea6568b346a18ce9d1e61d 102444 102445 With minor style fixes, ported to dixLookupResourceByType, and ported 102446 away from client->noClientException and xalloc/xfree. 102447 102448 v2: Fix a memory leak in PanoramiXCompositeNameWindowPixmap, spotted by 102449 James Jones. 102450 v3: Fix a buglet in PanoramiXCompositeUnredirectSubwindows, spotted by 102451 Dave Airlie. 102452 v4: Fix a style issue with resource lookup noted by Jamey Sharp. 102453 102454 Reviewed-by: Dave Airlie <airlied@gmail.com> 102455 Signed-off-by: Adam Jackson <ajax@redhat.com> 102456 102457commit 6dea617e0f71a6fc7937e3a1e10474fa87488284 102458Author: David Reveman <davidr@novell.com> 102459Date: Tue Jan 5 15:35:10 2010 -0500 102460 102461 fixes: Add panoramix support 102462 102463 Taken from: 102464 583f4dde81d8d5e9101b0289946e1914ea1ee124 102465 102466 With minor style fixes and ported to dixLookupResourceByType. 102467 102468 Reviewed-by: Dave Airlie <airlied@gmail.com> 102469 Signed-off-by: Adam Jackson <ajax@redhat.com> 102470 102471commit 1b3c57f075049250c6d059412af3d6954468a5b7 102472Author: Adam Jackson <ajax@redhat.com> 102473Date: Fri Aug 20 12:25:23 2010 -0400 102474 102475 xfree86: Remove unused xf86PixmapKeyRec 102476 102477 Reviewed-by: Matt Turner <mattst88@gmail.com> 102478 Signed-off-by: Adam Jackson <ajax@redhat.com> 102479 102480commit 63b9d67d7f4f1fa108370c619e1d0d06baedbd4c 102481Author: Adam Jackson <ajax@redhat.com> 102482Date: Thu Aug 19 18:40:24 2010 -0400 102483 102484 mi: Remove unused miAllocateGCPrivateIndex 102485 102486 Reviewed-by: Matt Turner <mattst88@gmail.com> 102487 Signed-off-by: Adam Jackson <ajax@redhat.com> 102488 102489commit 87f5b843a08fd92b4b31c04e3e6d947b8bc0d160 102490Author: Adam Jackson <ajax@redhat.com> 102491Date: Fri Nov 19 16:25:51 2010 -0500 102492 102493 render: Reshuffle and cull dead field from GlyphSetRec 102494 102495 Reviewed-by: Keith Packard <keithp@keithp.com> 102496 Reviewed-by: Matt Turner <mattst88@gmail.com> 102497 Signed-off-by: Adam Jackson <ajax@redhat.com> 102498 102499commit 606e079cc4d9a9db3197652ca51683c36f74efb8 102500Author: Adam Jackson <ajax@redhat.com> 102501Date: Fri Nov 19 16:17:48 2010 -0500 102502 102503 dix: reshuffle WindowOptRec to fill a hole on LP64 102504 102505 Reviewed-by: Keith Packard <keithp@keithp.com> 102506 Signed-off-by: Adam Jackson <ajax@redhat.com> 102507 102508commit fba5c8154dbc8f8c33801a5a43d7361cd6b54ffe 102509Author: Adam Jackson <ajax@redhat.com> 102510Date: Fri Nov 19 15:21:31 2010 -0500 102511 102512 dix: Remove unused ChangeSaveUnder hooks 102513 102514 Reviewed-by: Keith Packard <keithp@keithp.com> 102515 Reviewed-by: Matt Turner <mattst88@gmail.com> 102516 Signed-off-by: Adam Jackson <ajax@redhat.com> 102517 102518commit c4c4676e6874b42c2371eee96faa2c2dbb59a704 102519Author: Adam Jackson <ajax@redhat.com> 102520Date: Fri Nov 19 15:04:09 2010 -0500 102521 102522 dix: Remove the backing store leftovers 102523 102524 Reviewed-by: Keith Packard <keithp@keithp.com> 102525 Reviewed-by: Matt Turner <mattst88@gmail.com> 102526 Signed-off-by: Adam Jackson <ajax@redhat.com> 102527 102528commit 48bc8d0dd99284f509bc6de01a8fd24a7bfc69d5 102529Author: Adam Jackson <ajax@redhat.com> 102530Date: Wed Oct 6 09:56:56 2010 -0400 102531 102532 render: repack PictureRec 102533 102534 Eliminate the unused dither field, move filter and stateChanges into the 102535 bitfield, and reorder elements to pack holes on LP64. 102536 102537 sizeof(PictureRec) ILP32 LP64 102538 before: 84 152 102539 after: 72 120 102540 102541 Reviewed-by: Dave Airlie <airlied@redhat.com> 102542 Reviewed-by: Matt Turner <mattst88@gmail.com> 102543 Signed-off-by: Adam Jackson <ajax@redhat.com> 102544 102545commit ffcbfa0063bdc7c9ad5ac724285c7b6e67044c18 102546Author: Aaron Plattner <aplattner@nvidia.com> 102547Date: Mon Nov 15 20:43:18 2010 -0800 102548 102549 xfree86: Fix rotation of 2-color non-interleaved cursor images 102550 102551 When RandR 1.2's transformation code is enabled, it rotates the cursor 102552 image so that it appears upright on a rotated screen. This code 102553 completely mangles 2-color cursors on hardware where the the mask and 102554 source images are not interleaved due to two problems: 102555 102556 1. stride is calculated as (width / 4) rather than (width / 8), so the 102557 expression (y * stride) skips two lines instead of one for every 102558 time y is incremented. 102559 2. cursor_bitpos ignores the 'mask' parameter if the hardware doesn't 102560 specify any of the HARDWARE_CURSOR_SOURCE_MASK_INTERLEAVE_* flags. 102561 102562 To fix this, refactor the code to pass the whole xf86CursorInfoPtr 102563 through to cursor_bitpos and compute the correct stride there based on 102564 the flags. If none of the SOURCE_MASK_INTERLEAVE flags are set, use 102565 the total cursor size to move the 'image' variable into the mask part 102566 of the image before computing the desired byte pointer. 102567 102568 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 102569 Reviewed-by: Robert Morell <rmorell@nvidia.com> 102570 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 102571 Tested-by: Cyril Brulebois <kibi@debian.org> 102572 102573commit 7ee965a300c9eddcc1acacf9414cfe3e589222a8 102574Author: Adam Jackson <ajax@redhat.com> 102575Date: Tue May 11 09:16:19 2010 -0400 102576 102577 linux: Refactor VT switch on {re,}generation and shutdown 102578 102579 This makes more things fatal than were fatal before, but that's correct; 102580 if you need the VT, then failing to get it on regeneration means things 102581 are about to go very very badly. 102582 102583 Reviewed-by: Julien Cristau <jcristau@debian.org> 102584 Signed-off-by: Adam Jackson <ajax@redhat.com> 102585 102586commit 0def735c6a9fab19acc4b806def6921a702f617f 102587Author: Adam Jackson <ajax@redhat.com> 102588Date: Tue May 11 08:40:33 2010 -0400 102589 102590 xfree86: Remove dead ifdefs from VT switching 102591 102592 Reviewed-by: Julien Cristau <jcristau@debian.org> 102593 Signed-off-by: Adam Jackson <ajax@redhat.com> 102594 102595commit 96a7a7d197d5f96e7aa51928155d4b8d5a115875 102596Author: Adam Jackson <ajax@redhat.com> 102597Date: Mon May 10 18:40:09 2010 -0400 102598 102599 linux: Don't muck about with tty permissions 102600 102601 This is not X's job. 102602 102603 Reviewed-by: Julien Cristau <jcristau@debian.org> 102604 Signed-off-by: Adam Jackson <ajax@redhat.com> 102605 102606commit 6ce1908ba4959aaa0a48171acb62cf6a5d5cbf18 102607Author: Adam Jackson <ajax@redhat.com> 102608Date: Wed May 12 08:50:54 2010 -0400 102609 102610 xfree86: Remove os-support/sysv 102611 102612 Nothing's using it, the SysV derivatives we support have their own 102613 custom versions. 102614 102615 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102616 Signed-off-by: Adam Jackson <ajax@redhat.com> 102617 102618commit 01159ffc4718ba48ab7b16acaffd8008f9803811 102619Author: Adam Jackson <ajax@redhat.com> 102620Date: Mon May 10 18:14:49 2010 -0400 102621 102622 linux: Remove redundant variable 102623 102624 Reviewed-by: Julien Cristau <jcristau@debian.org> 102625 Signed-off-by: Adam Jackson <ajax@redhat.com> 102626 102627commit a80780a7638f847c3be20e5e0c7fe85e83d9bdd1 102628Author: Adam Jackson <ajax@redhat.com> 102629Date: Wed Nov 17 09:03:06 2010 -0500 102630 102631 glx: Remove swap barrier and hyperpipe support 102632 102633 Never implemented in any open source driver. The implementation 102634 assumed explicit DDX driver knowledge of how the client-side driver 102635 worked, since at the time the server's GL renderer was not a DRI driver. 102636 But now, it is, so any implementation of these should be done with 102637 additional DRI driver API, like the swap control extension. 102638 102639 Reviewed-by: Julien Cristau <jcristau@debian.org> 102640 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 102641 Signed-off-by: Adam Jackson <ajax@redhat.com> 102642 102643commit f4f41c812d2ba5edba7e0e0dbaab4bbf0af834b3 102644Author: Ville Syrjälä <ville.syrjala@nokia.com> 102645Date: Fri Oct 29 17:34:56 2010 +0300 102646 102647 xfree86/xv: Provide a ModeSet hook 102648 102649 Reput all ports when the display mode or panning has been changed by 102650 RandR code. This makes the overlays appear in the correct position 102651 on the screen. 102652 102653 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102654 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102655 102656commit 640c8716e0400e35afd7e91efc826fc447ea6745 102657Author: Ville Syrjälä <ville.syrjala@nokia.com> 102658Date: Thu Oct 28 18:13:02 2010 +0300 102659 102660 xfree86/modes: Provide a ModeSet hook in ScrnInfo 102661 102662 Add a new hook ModeSet that will be called after display mode is 102663 changed, or after the display has been panned. 102664 102665 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102666 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102667 102668commit a942534ca3908418c407115c6393263c2fe05931 102669Author: Ville Syrjälä <ville.syrjala@nokia.com> 102670Date: Tue Nov 2 16:05:13 2010 +0200 102671 102672 xfree86/xv: Pass all coordinate arguments to ReputImage 102673 102674 Pass all of the src/dst coordinates to ReputImage so that drivers 102675 don't necessarily have to do double bookkeeping. 102676 102677 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102678 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102679 102680commit 0fef4e94480f2bf1157ce5f92fcb0c7dd1585371 102681Author: Ville Syrjälä <ville.syrjala@nokia.com> 102682Date: Tue Nov 2 16:23:02 2010 +0200 102683 102684 xfree86/xv: Document VIDEO_CLIP_TO_VIEWPORT incompatibility with reput 102685 102686 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102687 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102688 102689commit 5f8ec1ade8b485f48de8c72011409219afad9dd7 102690Author: Ville Syrjälä <ville.syrjala@nokia.com> 102691Date: Tue Nov 2 16:04:36 2010 +0200 102692 102693 xfree86/xv: Add some helpful comments about ReputImage 102694 102695 Document the fact that ReputImage is used for stills as well as images. 102696 102697 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102698 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102699 102700commit b4ebde23d25bef1b891902d75b2db3aad92685b7 102701Author: Ville Syrjälä <ville.syrjala@nokia.com> 102702Date: Mon Nov 1 20:27:25 2010 +0200 102703 102704 xfree86/xv: Fix ReputImage clipping 102705 102706 PutImage/PutStill respect the GC clip, however ReputImage does not. 102707 PutImage/PutStill are supposed to be oneshot operations so ReputImage 102708 should never expand the area covered by the clip, instead it should 102709 only shrink if the window clip shrinks. So commandeer clientClip 102710 into use by ReputImage and initially make it a copy of the original 102711 GC composite clip. Whenever ReputImage needs reclipping update 102712 clientClip with the newly calculated composite clip. 102713 102714 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102715 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102716 102717commit 7294236bdb29b4fa7a7bc27aff9c786c5a33c544 102718Author: Ville Syrjälä <ville.syrjala@nokia.com> 102719Date: Mon Nov 1 20:02:40 2010 +0200 102720 102721 xfree86/xv: Remove clipOrg from XvPortRecPrivate 102722 102723 clipOrg never changes except when clientClip changes, so instead of 102724 keeping copies of both originals translate clientClip by clipOrg 102725 immediately and just keep the translated clientClip. 102726 102727 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102728 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102729 102730commit 47d8bd0a999693f47fd244dc876ef7f8fcd52854 102731Author: Ville Syrjälä <ville.syrjala@nokia.com> 102732Date: Fri Oct 29 17:31:14 2010 +0300 102733 102734 xfree86/xv: Factor out the meat of xf86XVAdjustFrame for later reuse 102735 102736 No functional change. 102737 102738 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102739 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102740 102741commit 3d4d0237a3981820a9b7290f69c7172a48659a10 102742Author: Ville Syrjälä <ville.syrjala@nokia.com> 102743Date: Fri Oct 29 17:14:34 2010 +0300 102744 102745 xfree86/xv: Change the behaviour of AdjustFrame to reput everything 102746 102747 Also reput PutVideo/GetVideo ports in AdjustFrame. This makes the 102748 overlay track the screen panning instead of staying stationary in the 102749 wrong place. 102750 102751 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102752 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102753 102754commit d794be81218934ad4580211c7a0a27cb71eb5d60 102755Author: Ville Syrjälä <ville.syrjala@nokia.com> 102756Date: Fri Oct 29 18:07:32 2010 +0300 102757 102758 xfree86/xv: No need to free composite clip in AdjustFrame 102759 102760 Nothing should change in AdjustFrame that would need the composite clip 102761 to be recomputed. 102762 102763 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102764 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102765 102766commit 6051c7e940767b7120ad9c7e9188aee6a6f62b94 102767Author: Ville Syrjälä <ville.syrjala@nokia.com> 102768Date: Fri Oct 29 17:05:02 2010 +0300 102769 102770 xfree86/xv: Change the behaviour of ClipNotify to reput instead of stop 102771 102772 When ClipNotify gets called for a visible window, reput instead of 102773 stopping the port. This eliminates nasty overlay flickering that 102774 happens during clip changes. 102775 102776 If the window is invisible or if ReputImage isn't supported stop 102777 and remove the port from the window as was done before. 102778 102779 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102780 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102781 102782commit 27707ac5bdf6831630944f02f2128951a529faab 102783Author: Ville Syrjälä <ville.syrjala@nokia.com> 102784Date: Fri Oct 29 17:00:24 2010 +0300 102785 102786 xfree86/xv: Allow xf86XVReputOrStopPort() to stop PutVideo/GetVideo ports 102787 102788 Modify xf86XVReputOrStopPort() to allow stopping of all types of ports. 102789 Will be useful later. 102790 102791 No functional change. 102792 102793 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102794 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102795 102796commit 38987aae199982f17eae664f3a28ff5941d3ee73 102797Author: Ville Syrjälä <ville.syrjala@nokia.com> 102798Date: Fri Oct 29 16:51:24 2010 +0300 102799 102800 xfree86/xv: Factor out the meat of xf86XVWindowExposures for later reuse 102801 102802 No fucntional change. 102803 102804 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102805 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102806 102807commit 0178173f0579c4b9dd0cf60b9cd3e5d264f86ae4 102808Author: Ville Syrjälä <ville.syrjala@nokia.com> 102809Date: Fri Oct 29 16:39:13 2010 +0300 102810 102811 xfree86/xv: Eliminate two open coded copies of xf86XVRemovePortFromWindow 102812 102813 No functional change. 102814 102815 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102816 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102817 102818commit b02858b5af47d524001ec571737cc92f6a6f06d3 102819Author: Ville Syrjälä <ville.syrjala@nokia.com> 102820Date: Fri Oct 29 16:54:42 2010 +0300 102821 102822 xfree86/xv: Remove useless NULL check from ClipNotify 102823 102824 WinPriv->PortRec should never be NULL as WinPriv itself would be removed 102825 from the list when the port is removed from the window. 102826 102827 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 102828 Reviewed-by: Luc Verhaegen <luc.verhaegen@basyskom.de> 102829 102830commit 23e3d1f23318ce69623f91908f888a09f8b74ac2 102831Author: Peter Hutterer <peter.hutterer@who-t.net> 102832Date: Fri Nov 26 10:00:49 2010 +1000 102833 102834 dix: remove now unnecessary !! before BitIsOn() 102835 102836 The macro has been changed to do this already, no need for double 102837 not-not-ing. 102838 102839 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 102840 Reviewed-by: Keith Packard <keithp@keithp.com> 102841 102842commit 42dc91e32a02b6b21ff5c45f465f3349e5822615 102843Author: Peter Hutterer <peter.hutterer@who-t.net> 102844Date: Wed Nov 24 14:20:30 2010 +1000 102845 102846 include: let BitIsOn() return a boolean value. 102847 102848 Simply returning the mask bit breaks checks like 102849 BitIsOn(mask, 0) != BitIsOn(mask, 1); 102850 as used in 048e93593e3f7a99a7d2a219e1ce2bdc9d407807. 102851 102852 The naming of this macro suggests that it should return boolean values 102853 anyway. This patch also adds a few simple tests for these macros to make 102854 sure they don't accidentally break in the future. 102855 102856 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 102857 Reviewed-by: Pat Kane <pekane52@gmail.com> 102858 Reviewed-by: Julien Cristau <jcristau@debian.org> 102859 Reviewed-by: Keith Packard <keithp@keithp.com> 102860 102861commit 4e0f8f666e61390206c42ad2087477a912525bc7 102862Merge: 400ddf2f5 639600fa7 102863Author: Keith Packard <keithp@keithp.com> 102864Date: Wed Nov 24 11:47:33 2010 -0800 102865 102866 Merge remote branch 'whot/for-keith' 102867 102868commit 400ddf2f58ce673c74d4e8ee829729cd3c75f8c9 102869Merge: 7250f078c e074f745a 102870Author: Keith Packard <keithp@keithp.com> 102871Date: Wed Nov 24 11:46:18 2010 -0800 102872 102873 Merge remote branch 'jeremyhu/master' 102874 102875commit 7250f078c12fd20d5ac9150f54495926e5121461 102876Author: Gaetan Nadon <memsize@videotron.ca> 102877Date: Sun Nov 21 14:27:58 2010 -0500 102878 102879 doc: refactor Makefile and xmlrules.in code for reusability 102880 102881 A different approach which requires less variables setting 102882 and internal knowledge of the reused code. 102883 Changing from "install" to "not install" is very easy now. 102884 102885 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102886 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 102887 Signed-off-by: Keith Packard <keithp@keithp.com> 102888 102889commit f33512b70cf0350a62df2ee2287a870d5fc991a8 102890Author: Gaetan Nadon <memsize@videotron.ca> 102891Date: Sun Nov 21 14:27:57 2010 -0500 102892 102893 xmlrules.in: use $(top_srcdir) rather than ../../../ [...] 102894 102895 Relative paths don't always work in distcheck when srcdir not = builddir 102896 include $(top_srcdir)/doc/xml/xmlrules.in 102897 102898 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102899 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 102900 Signed-off-by: Keith Packard <keithp@keithp.com> 102901 102902commit c25b407f22456f50eef90d9bc5e026c05415c021 102903Author: Gaetan Nadon <memsize@videotron.ca> 102904Date: Sun Nov 21 14:27:56 2010 -0500 102905 102906 xmlrules.in: specify the xserver entities depedencies on the target 102907 102908 The generated docs will rebuild when the xserver.ent file changes. 102909 102910 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102911 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 102912 Signed-off-by: Keith Packard <keithp@keithp.com> 102913 102914commit 73841074eb856e6be3b9f9e3fc2d2721443166ab 102915Author: Gaetan Nadon <memsize@videotron.ca> 102916Date: Sun Nov 21 14:27:55 2010 -0500 102917 102918 xmlrules.in: use pattern rules to enable dependencies 102919 102920 This will allow a dependency to be specified as done in libX11: 102921 102922 %.html: %.xml $(dist_spec_DATA) 102923 $(AM_V_GEN)$(XMLTO) $(XMLTO_FLAGS) xhtml-nochunks $< 102924 102925 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102926 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 102927 Signed-off-by: Keith Packard <keithp@keithp.com> 102928 102929commit 97e307dda201d6ba28a723c21a95e484921d1b08 102930Author: Gaetan Nadon <memsize@videotron.ca> 102931Date: Sun Nov 21 14:27:54 2010 -0500 102932 102933 xmlrules.in: no need to setup xmlto flags when configuring --without-xmlto 102934 102935 The AM conditional HAVE_XMLTO should wrap more statements. 102936 102937 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102938 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 102939 Signed-off-by: Keith Packard <keithp@keithp.com> 102940 102941commit 1a7b14d1183057fc69a5da446e898104b65c8347 102942Author: Gaetan Nadon <memsize@videotron.ca> 102943Date: Sun Nov 21 14:27:53 2010 -0500 102944 102945 xmlrules.in: remove unrequired "@rm -f $@" from doc targets 102946 102947 Unable to find a purpose for this, not used anywhere else 102948 but in the font module. 102949 102950 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102951 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 102952 Signed-off-by: Keith Packard <keithp@keithp.com> 102953 102954commit d3c523bd124dc1a3f4539f7f6cd4868adbfc25e1 102955Author: Gaetan Nadon <memsize@videotron.ca> 102956Date: Sun Nov 21 14:27:52 2010 -0500 102957 102958 doc: HTML file generation: use the installed copy of xorg.css 102959 102960 Currently the xorg.css file is copied in each location 102961 where a DocBook/XML file resides. This produces about 102962 70 copies in the $(docdir) install tree for all of xorg. 102963 102964 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102965 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 102966 Signed-off-by: Keith Packard <keithp@keithp.com> 102967 102968commit 504e3010e976ccc008ff48417850b594beaee2ec 102969Author: Alan Coopersmith <alan.coopersmith@oracle.com> 102970Date: Fri Nov 19 22:09:05 2010 -0800 102971 102972 Xorg.man: Replace XDarwin reference with Xquartz 102973 102974 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 102975 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 102976 Signed-off-by: Keith Packard <keithp@keithp.com> 102977 102978commit 566d09a5cd6452ee2e05e23d2205e7c3aa31f0da 102979Author: Bill Nottingham <notting@redhat.com> 102980Date: Mon Nov 15 11:25:14 2010 +1000 102981 102982 xfree86: store the screen's gamma information on init. 102983 102984 This fixes a gamma issue on vt switch observed with KDM. VT switching away 102985 and back would result in a black screen. Avoid this by storing the current 102986 gamma information on init. 102987 102988 https://bugzilla.redhat.com/show_bug.cgi?id=533217 102989 102990 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 102991 Reviewed-by: Keith Packard <keithp@keithp.com> 102992 Signed-off-by: Keith Packard <keithp@keithp.com> 102993 102994commit c050aa2f5fa6a7ff8ce4f91def14ca44f799f067 102995Author: Adam Jackson <ajax@redhat.com> 102996Date: Mon Nov 15 11:25:12 2010 +1000 102997 102998 xfree86: apply gamma settings on EnterVT. 102999 103000 When entering the VT, re-apply the saved gamma settings for each screen. 103001 103002 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103003 Reviewed-by: Keith Packard <keithp@keithp.com> 103004 Signed-off-by: Keith Packard <keithp@keithp.com> 103005 103006commit e074f745a80b87d1d82fb39489d7f35163d7552a 103007Author: Jeremy Huddleston <jeremyhu@apple.com> 103008Date: Tue Nov 23 16:03:48 2010 -0800 103009 103010 XQuartz: Remove unused CloseInputDevice stub 103011 103012 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 103013 103014commit 639600fa7ebad7a4953d0b632abeff2d339a49c6 103015Author: Peter Hutterer <peter.hutterer@who-t.net> 103016Date: Tue Nov 23 13:30:05 2010 +1000 103017 103018 dix: add a fixme about a corner-case that should probably be fixed. 103019 103020 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103021 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 103022 103023commit 0aca9e8424d8b978053a29c4468f800fc2e7238c 103024Author: Peter Hutterer <peter.hutterer@who-t.net> 103025Date: Tue Nov 23 13:04:19 2010 +1000 103026 103027 dix: fix typo, set the second valuator with the y-axis data. 103028 103029 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103030 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 103031 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 103032 103033commit 45131bb67f9eacb3b55de4bcd3730fb9bb028e8a 103034Author: Peter Hutterer <peter.hutterer@who-t.net> 103035Date: Tue Nov 23 12:55:50 2010 +1000 103036 103037 dix: GetProximityEvents needs to check up to the last valuator 103038 103039 valuator_mask_size() returns the highest valuator set as opposed to the 103040 number of set bits (which obviously changes as we unset valuators). 103041 103042 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103043 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 103044 103045commit 048e93593e3f7a99a7d2a219e1ce2bdc9d407807 103046Author: Peter Hutterer <peter.hutterer@who-t.net> 103047Date: Tue Nov 23 11:15:52 2010 +1000 103048 103049 dix: when converting to XI 1.x events, use the first_valuator's device mode 103050 103051 Don't stop processing events when a valuator's mode doesn't match the 0th 103052 valuator's mode. Instead, start with the first_valuator in the event and 103053 keep stacking them on until the first valuator with a different mode is hit. 103054 103055 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103056 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 103057 103058commit 4f3e670fd28fceec384ff0d4ca7c1ff8d7a872ad 103059Author: Peter Hutterer <peter.hutterer@who-t.net> 103060Date: Tue Nov 23 10:59:52 2010 +1000 103061 103062 dix: when comparing axis modes, compare the bits, not the bytes. 103063 103064 The DeviceEvent's mode field is a set of bits for each valuator, not bytes. 103065 103066 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103067 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 103068 103069commit 6f12934d4ea52f17ac7b21e92f4f929db165f83d 103070Author: Peter Hutterer <peter.hutterer@who-t.net> 103071Date: Tue Nov 23 10:58:32 2010 +1000 103072 103073 dix: replace a manual valuator check with valuator_get_mode(). 103074 103075 This check was missing the OutOfProximity mask and resulted in the wrong 103076 bits being set in InternalEvents. 103077 103078 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103079 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 103080 103081commit ebe3ddaf284ad388835da9cbaa6ad61fa21ce6dd 103082Author: Simon Thum <simon.thum@gmx.de> 103083Date: Fri Nov 19 11:09:21 2010 +1000 103084 103085 dix: fix up valuators passed to acceleration code. 103086 103087 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103088 Reviewed-by: Simon Thum <simon.thum@gmx.de> 103089 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 103090 103091commit d435e1ecb86e2fe6292b5945262956644f979bbb 103092Author: Peter Hutterer <peter.hutterer@who-t.net> 103093Date: Fri Feb 5 11:08:27 2010 +1000 103094 103095 test: reduce range of byte-padding macro tests. 103096 103097 Byte padding and conversion is interesting for the rage of 0-8 bytes, and 103098 then interesting towards the end of the valid range (INT_MAX - 7 and INT_MAX 103099 - 3). 103100 103101 Note: this changes the upper range for pad_to_int32() and bytes_to_int32() 103102 from the previous (INT_MAX - 4) to (INT_MAX - 3). 103103 103104 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103105 Reviewed-by: Julien Cristau <jcristau@debian.org> 103106 103107commit f49ee9074a66883a3c525b0d6e71589123288465 103108Author: Peter Hutterer <peter.hutterer@who-t.net> 103109Date: Fri Feb 5 10:59:52 2010 +1000 103110 103111 test: compare byte padding macros against the expected bytes. 103112 103113 We calculate the expected bytes for each value, let's use it. 103114 103115 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103116 Reviewed-by: Julien Cristau <jcristau@debian.org> 103117 103118commit b31df0439fe336a43a2355e2f1fb223d86045a05 103119Author: Peter Hutterer <peter.hutterer@who-t.net> 103120Date: Mon Nov 22 15:14:02 2010 +1000 103121 103122 xfree86: add missing linebreak in error message. 103123 103124 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103125 Reviewed-by: Magnus Kessler <Magnus.Kessler@gmx.net> 103126 103127commit 97b9bb1e2cebbb769ef347455d9c7979d19de99f 103128Author: Jeremy Huddleston <jeremyhu@apple.com> 103129Date: Mon Nov 22 20:43:14 2010 -0800 103130 103131 XQuartz: Localization Updates 103132 103133 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 103134 103135commit f9fbafa1a07f951e89f208bf4928a0bf37b67cfe 103136Author: Jeremy Huddleston <jeremyhu@apple.com> 103137Date: Mon Nov 22 20:41:39 2010 -0800 103138 103139 XQuartz: Remove existing localization files on install to avoid conflicts 103140 103141 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 103142 103143commit 6eed13ad5998ba80296bba2bb3f89928ee51194a 103144Author: Jeremy Huddleston <jeremyhu@apple.com> 103145Date: Mon Nov 22 20:25:33 2010 -0800 103146 103147 XQuartz dri: Don't check CoreGraphics version 103148 103149 The version check is hackish, and I highly doubt anyone using this code is on 103150 a version of OS X that predates this support (10.2 has support for it). 103151 103152 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 103153 103154commit a4698754c1da1cf2ca64add5276d0c6012960504 103155Author: Jeremy Huddleston <jeremyhu@apple.com> 103156Date: Sun Nov 21 10:49:24 2010 -0800 103157 103158 XQuartz: Add pointer debugging 103159 103160 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 103161 103162commit 31737fff08ec19b394837341d5e358ec401f5cd8 103163Author: Chase Douglas <chase.douglas@ubuntu.com> 103164Date: Mon Nov 15 11:42:12 2010 -0500 103165 103166 Fix transformAbsolute 103167 103168 transformAbsolute must use old values if valuator mask doesn't have new 103169 ones, and it must only set new values if there was a change. 103170 103171 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 103172 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 103173 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103174 103175commit 463841f45aafd1227bac652965c9ea337dd527fa 103176Author: Chase Douglas <chase.douglas@ubuntu.com> 103177Date: Mon Nov 15 11:42:11 2010 -0500 103178 103179 Fix GPE Y axis scaling 103180 103181 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 103182 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 103183 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103184 103185commit a6b9e8f1e5d5d0b3b0f121a6f677eeca7aab1950 103186Author: Adam Jackson <ajax@redhat.com> 103187Date: Wed Aug 25 11:06:38 2010 -0400 103188 103189 linux: Fix CPU usage bug in console fd flushing 103190 103191 If the vt gets a vhangup from under us, then the tty will appear ready 103192 in select(), but trying to tcflush() it will return -EIO, so we'll spin 103193 around at 100% CPU for no reason. Notice this condition and unregister 103194 the handler if it happens. 103195 103196 Signed-off-by: Adam Jackson <ajax@redhat.com> 103197 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 103198 Reviewed-by: Julien Cristau <jcristau@debian.org> 103199 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103200 103201commit ecdbe817fd7eb53fd1e7485c0492355f7f002d2a 103202Author: Dan Horák <dan@danny.cz> 103203Date: Tue Jun 22 19:45:26 2010 +0200 103204 103205 XF86CONFIGDIR is undefined when built with --disable-xorg (#28672) 103206 103207 The 10-evdev.conf file gets installed as /usr/share/X11/10-evdev.conf on 103208 platforms that built the server with --disable-xorg like s390/s390x. The 103209 definition/installation should be guarded with "if XORG" because it makes 103210 sense only when built with xorg. 103211 103212 X.Org Bug 28672 <http://bugs.freedesktop.org/show_bug.cgi?id=28672> 103213 103214 Signed-off-by: Dan Horák <dan@danny.cz> 103215 Acked-by: Julien Cristau <jcristau@debian.org> 103216 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103217 103218commit 38c46ccf796462d068d95acb9022de2acfc3a1f7 103219Merge: 3051dd513 c8998af3e 103220Author: Keith Packard <keithp@keithp.com> 103221Date: Mon Nov 15 09:06:33 2010 +0800 103222 103223 Merge remote branch 'whot/for-keith' 103224 103225commit 3051dd5135962283fc6c7de377a7c819d8a04e7f 103226Merge: ed547a9fc c5975722a 103227Author: Keith Packard <keithp@keithp.com> 103228Date: Mon Nov 15 08:58:02 2010 +0800 103229 103230 Merge remote branch 'jeremyhu/master' 103231 103232commit ed547a9fc8a4f5cd95ef882b656a376c163eda6d 103233Merge: 78f94f19a 84e8de127 103234Author: Keith Packard <keithp@keithp.com> 103235Date: Mon Nov 15 08:51:27 2010 +0800 103236 103237 Merge remote branch 'adkins/master' 103238 103239commit 78f94f19aab66a1e5331df0ce29f36e310b4195d 103240Author: Linus Arver <linusarver@gmail.com> 103241Date: Sun Oct 17 12:26:01 2010 -0700 103242 103243 Xext: panoramiXprocs: fix typo 103244 103245 This fixes a typo introduced in commit 103246 80b5d3a3264d2c5167e5ac85a3b04af0f89cece1. The pointer pDst was changed 103247 unintentionally to pWin from a copy/paste error. This resulted in all 103248 QT-based apps and some tcl/tk ones (like fontforge) to crash X 1.9 on 103249 starting up, when Xinerama was enabled. 103250 103251 Bug report: https://bbs.archlinux.org/viewtopic.php?id=106125 103252 103253 Signed-off-by: Elie Bleton <drozofil@gmail.com> 103254 Reviewed-by: Adam Jackson <ajax@redhat.com> 103255 Reviewed-by: Matt Turner <mattst88@gmail.com> 103256 Tested-by: Linus Arver <linusarver@gmail.com> 103257 103258commit c8998af3eba49a542625912d512087e4eaaf5ede 103259Author: Cyril Brulebois <kibi@debian.org> 103260Date: Sat Nov 13 02:05:16 2010 +0100 103261 103262 xfree86: Initialize Pointer and Keyboard. 103263 103264 That helps us get rid of: 103265 | CC xf86Config.lo 103266 | xf86Config.c: In function ‘T.174’: 103267 | xf86Config.c:1228: warning: ‘Pointer$module’ may be used uninitialized in this function 103268 | xf86Config.c:1097: note: ‘Pointer$module’ was declared here 103269 | xf86Config.c:1228: warning: ‘Pointer$drv’ may be used uninitialized in this function 103270 | xf86Config.c:1097: note: ‘Pointer$drv’ was declared here 103271 | xf86Config.c:1228: warning: ‘Pointer$type_name’ may be used uninitialized in this function 103272 | xf86Config.c:1097: note: ‘Pointer$type_name’ was declared here 103273 | xf86Config.c:1228: warning: ‘Pointer$private’ may be used uninitialized in this function 103274 | xf86Config.c:1097: note: ‘Pointer$private’ was declared here 103275 | xf86Config.c:1228: warning: ‘Pointer$dev’ may be used uninitialized in this function 103276 | xf86Config.c:1097: note: ‘Pointer$dev’ was declared here 103277 | xf86Config.c:1228: warning: ‘Pointer$fd’ may be used uninitialized in this function 103278 | xf86Config.c:1097: note: ‘Pointer$fd’ was declared here 103279 | xf86Config.c:1228: warning: ‘Pointer$set_device_valuators’ may be used uninitialized in this function 103280 | xf86Config.c:1097: note: ‘Pointer$set_device_valuators’ was declared here 103281 | xf86Config.c:1228: warning: ‘Pointer$switch_mode’ may be used uninitialized in this function 103282 | xf86Config.c:1097: note: ‘Pointer$switch_mode’ was declared here 103283 | xf86Config.c:1228: warning: ‘Pointer$control_proc’ may be used uninitialized in this function 103284 | xf86Config.c:1097: note: ‘Pointer$control_proc’ was declared here 103285 | xf86Config.c:1228: warning: ‘Pointer$read_input’ may be used uninitialized in this function 103286 | xf86Config.c:1097: note: ‘Pointer$read_input’ was declared here 103287 | xf86Config.c:1228: warning: ‘Pointer$device_control’ may be used uninitialized in this function 103288 | xf86Config.c:1097: note: ‘Pointer$device_control’ was declared here 103289 | xf86Config.c:1228: warning: ‘Pointer$flags’ may be used uninitialized in this function 103290 | xf86Config.c:1097: note: ‘Pointer$flags’ was declared here 103291 | xf86Config.c:1228: warning: ‘Pointer$next’ may be used uninitialized in this function 103292 | xf86Config.c:1097: note: ‘Pointer$next’ was declared here 103293 103294 Signed-off-by: Cyril Brulebois <kibi@debian.org> 103295 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103296 103297commit 98abac809a024ac3ed9dd0514ca96367f668c05d 103298Author: Jon TURNEY <jon.turney@dronecode.org.uk> 103299Date: Thu Nov 11 23:06:13 2010 +0000 103300 103301 Fix compilation after input API change 103302 103303 Fix compilation after commit 675f4a8525d29ebad783351e17be785b2f32b2e8 103304 converts winmouse.c to the new ValuatorMask input API, but doesn't 103305 include the required header inpututil.h 103306 103307 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 103308 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103309 103310commit c5975722a95c41b8715a4612fc1c3275f8e9b59e 103311Author: Jeremy Huddleston <jeremyhu@apple.com> 103312Date: Sat Nov 13 16:04:10 2010 -0800 103313 103314 XQuartz: Fix make distcheck 103315 103316 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 103317 Reviewed-by: Cyril Brulebois <kibi@debian.org> 103318 103319commit 7b33ef74738d6a557cb25aeb14693b2a340c0e12 103320Author: Jeremy Huddleston <jeremyhu@apple.com> 103321Date: Sat Nov 13 15:46:16 2010 -0800 103322 103323 dix: Fix make distcheck 103324 103325 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 103326 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 103327 103328commit 965e709f2b0b17f1e59e5aeb5e7717fede51ef97 103329Author: Jeremy Huddleston <jeremyhu@apple.com> 103330Date: Thu Nov 11 17:23:33 2010 -0800 103331 103332 XQuartz: Build-fixes for XQuartz for recent input API changes 103333 103334 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103335 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 103336 103337commit fa50670c32637faff6bb91f1206a0e3c17a9bada 103338Author: Peter Hutterer <peter.hutterer@who-t.net> 103339Date: Fri Nov 12 11:48:08 2010 +1000 103340 103341 xfree86: rename allowEmptyInput to forceInputDevices. 103342 103343 This is a more accurate name for the actual functionality than 103344 allowEmptyInput. Historically, allowEmptyInput has allowed the server to 103345 start with no input devices. Since 1.4 and the introduction of VCP and VCK, 103346 there are always two input devices present. 103347 103348 allowEmptyInput was changed in behaviour to essentially "ignore xorg.conf 103349 devices or not", auto-adding the built-in devices if disabled. 103350 103351 Rename to forceInputDevices, because that's essentially what it does. When 103352 disabled (i.e. when hotplugging is enabled), it disables all 103353 mouse/kbd/vmmouse devices configured in the xorg.conf file. 103354 When enabled, it forces the traditional behaviour for input devices: 103355 - use input devices configured in the server layout 103356 - if none are configured, use the first pointer and the first keyboard 103357 device in the xorg.conf 103358 - if none are configured, create the default pointer/keyboard devices. 103359 103360 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103361 Reviewed-by: Julien Cristau <jcristau@debian.org> 103362 103363commit c481dae0dc89d8d50cac23210023d75bfaf6012b 103364Author: Cyril Brulebois <kibi@debian.org> 103365Date: Thu Nov 11 23:55:55 2010 +0100 103366 103367 test: Fix missing xkbsrv.h include. 103368 103369 Otherwise, building fails with CFLAGS="-Wall -Werror" this way: 103370 | protocol-common.c: In function ‘init_simple’: 103371 | protocol-common.c:159: error: implicit declaration of function ‘XkbInitPrivates’ 103372 | protocol-common.c:159: error: nested extern declaration of ‘XkbInitPrivates’ 103373 103374 Signed-off-by: Cyril Brulebois <kibi@debian.org> 103375 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103376 103377commit b8114f25b266624c8f73d03c710349f98693d877 103378Author: Peter Hutterer <peter.hutterer@who-t.net> 103379Date: Fri Nov 12 08:32:05 2010 +1000 103380 103381 dix: fix typo in setting valuator modes (#21548) 103382 103383 X.Org Bug 31548 <http://bugs.freedesktop.org/show_bug.cgi?id=31548> 103384 103385 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103386 Tested-by: Cyril Brulebois <kibi@debian.org> 103387 103388commit 5cd11d2356d153840f1b429bdb8284367f8dc468 103389Author: Peter Hutterer <peter.hutterer@who-t.net> 103390Date: Wed Jun 9 17:10:57 2010 +1000 103391 103392 Xi: move property reset from extension shutdown to init. 103393 103394 If any part of the stack calls XIGetKnownProperty during device shutdown 103395 the property is re-initialized before the server generation resets, leaving 103396 the value invalid again. 103397 103398 Move the reset to the extension init which happens before input devices are 103399 initialized before the first property is requested. 103400 103401 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103402 Reviewed-by: Julien Cristau <jcristau@debian.org> 103403 103404commit da66119593b63ef82ae6fa2e034ccd0856b03a80 103405Author: Tiago Vignatti <tiago.vignatti@nokia.com> 103406Date: Wed Nov 10 21:48:04 2010 +0200 103407 103408 dix: fix root window background behaviour for protocol calls 103409 103410 Instead always paint root tiled (-retro like), protocol calls 103411 (XSetWindowBackgroundPixmap and related) should behave accordingly when None 103412 and ParentRelative is set as background pixmap. 103413 103414 It follow what the protocol states: "changing the background of a root window 103415 to None or ParentRelative restores the default background pixmap". 103416 103417 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 103418 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 103419 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 103420 103421commit 8976e9766edfb33f5cf3f6d54f09e46d29bc4ec5 103422Author: Tiago Vignatti <tiago.vignatti@nokia.com> 103423Date: Wed Nov 10 21:48:03 2010 +0200 103424 103425 dix: adds support for none root window background 103426 103427 It lets the driver notify the server whether it can draw a background when 103428 '-background none' option is used by the system platform. Use cases for that 103429 could be video drivers performing mode-setting in kernel time, before X is up, 103430 so a seamless transition would happen until X clients start to show up. 103431 103432 If the driver can copy the framebuffer cleanly then it can set the flag 103433 (canDoBGNoneRoot), otherwise the server will fallback to the normal behaviour. 103434 The system must explicit indicates willingness of doing so through 103435 '-background none'. We could do this option as default; in such case, 103436 malicious users would be able to steal the framebuffer with a bit of tricks. 103437 103438 For instance, I can see the content of my nVidia Quadro FX 580 framebuffer 103439 old X session modifying a bit nv driver: 103440 103441 xf86DPMSInit(pScreen, xf86DPMSSet, 0); 103442 103443 - /* Clear the screen */ 103444 - if(pNv->xaa) { 103445 - /* Use the acceleration engine */ 103446 - pNv->xaa->SetupForSolidFill(pScrn, 0, GXcopy, ~0); 103447 - pNv->xaa->SubsequentSolidFillRect(pScrn, 103448 - 0, 0, pScrn->displayWidth, pNv->offscreenHeight); 103449 - G80DmaKickoff(pNv); 103450 - } else { 103451 - /* Use a slow software clear path */ 103452 - memset(pNv->mem, 0, pitch * pNv->offscreenHeight); 103453 - } 103454 + pScreen->canDoBGNoneRoot = TRUE; 103455 103456 The commit is originally based on discussions happened on xorg-devel: 103457 http://lists.freedesktop.org/archives/xorg-devel/2010-June/009755.html 103458 103459 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 103460 Reviewed-by: Rami Ylimäki <rami.ylimaki@vincit.fi> 103461 Acked-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 103462 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 103463 103464commit bfe9cdd11bcb60cf33dc48136ebee028001a737e 103465Author: Tiago Vignatti <tiago.vignatti@nokia.com> 103466Date: Wed Oct 20 17:54:58 2010 -0200 103467 103468 dix: delete logo hack screen saver 103469 103470 Protocol doesn't mention about screen saver with logo being required and 103471 people are already using more intelligent ways to draw screen saver themes. So 103472 consider -logo as deprecated option, deleting its code. 103473 103474 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 103475 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 103476 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 103477 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 103478 103479commit 84e8de1271bb11b5b4b9747ae4647f47333a8ab7 103480Author: Jesse Adkins <jesserayadkins@gmail.com> 103481Date: Wed Nov 10 21:37:24 2010 -0800 103482 103483 fb: Remove fbcmap.c (bug 5436) 103484 103485 This is obsolete since a240c039c47c0be22ea5e100692307b26d938747. 103486 Updated fb.h to mention that the functions come from fbcmap_mi.c now. 103487 Dropped fbcmap.c from the makefile. 103488 103489 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 103490 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 103491 103492commit 23e329b5641ef044c34ac712c9193171eaf47198 103493Author: Jesse Adkins <jesserayadkins@gmail.com> 103494Date: Wed Nov 10 21:35:58 2010 -0800 103495 103496 xfree86: parser: Remove 'CUSTOM' flag option in Monitor section. 103497 103498 Not used in the initial import, and also not documented. 103499 103500 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 103501 Reviewed-by: Adam Jackson <ajax@redhat.com> 103502 103503commit 88cb61e1e55c54982b90e2a77465faaac6a0ba89 103504Merge: 19f43836d 4ed4915bc 103505Author: Peter Hutterer <peter.hutterer@who-t.net> 103506Date: Thu Nov 11 12:54:46 2010 +1000 103507 103508 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-api 103509 103510 Conflicts: 103511 dix/getevents.c 103512 hw/xfree86/common/xf86Xinput.h 103513 103514 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103515 103516commit 4ed4915bc0fbe9b4419ca4a0d4e43c2fee032ae9 103517Merge: 291c39dfe ec1bfbc66 103518Author: Keith Packard <keithp@keithp.com> 103519Date: Wed Nov 10 16:58:21 2010 -0800 103520 103521 Merge remote branch 'whot/for-keith' 103522 103523commit ec1bfbc66926130e1153facc3b92ee175f1cb6b6 103524Author: Peter Hutterer <peter.hutterer@who-t.net> 103525Date: Tue Nov 9 15:27:26 2010 +1000 103526 103527 xfree86: remove user-configured AllowEmptyInput 103528 103529 An estimated 100% (rounded down to the nearest percent) of the people who 103530 have this in their configuration don't actually know what this option does. 103531 Protect the users from themselves. 103532 103533 IIRC, AEI on was useful for some time between 1.4 and 1.5 and never since. 103534 103535 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103536 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 103537 103538commit cbaa6a66e0f9e3b7e305606924ecda0147b59e96 103539Author: Peter Hutterer <peter.hutterer@who-t.net> 103540Date: Tue Nov 9 15:21:23 2010 +1000 103541 103542 config: remove mention of AllowEmptyInput 103543 103544 "You will probably want to add the following option to the ServerFlags of 103545 your xorg.conf: 103546 Option "AllowEmptyInput" "True"" 103547 103548 I can't imagine why you would want to do that. My life is painful enough 103549 already. 103550 103551 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103552 Reviewed-by: Julien Cristau <jcristau@debian.org> 103553 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 103554 103555commit 291c39dfef84d1b204d7f5e63e925dee2dbdaca7 103556Author: Jeremy Huddleston <jeremyhu@apple.com> 103557Date: Tue Nov 9 13:33:55 2010 -0800 103558 103559 DOC: Only build dtrace documentation with --enable-docs 103560 103561 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 103562 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 103563 Signed-off-by: Keith Packard <keithp@keithp.com> 103564 103565commit f0b4df99486066ad0e638b3b5debc905bc8ef1c4 103566Author: Adam Jackson <ajax@redhat.com> 103567Date: Sat Sep 11 17:55:57 2010 -0400 103568 103569 shadow: Optimize shadowUpdatePacked(). (#26973) 103570 103571 Signed-off-by: Matt Turner <mattst88@gmail.com> 103572 Reviewed-by: Soren Sandmann <sandmann@daimi.au.dk> 103573 Reviewed-by: Matt Turner <mattst88@gmail.com> 103574 Signed-off-by: Keith Packard <keithp@keithp.com> 103575 103576commit c9c0f93f8ed42413a05193ee71c69fa6b41f6a03 103577Merge: 73f6de1ad 4132b1c59 103578Author: Keith Packard <keithp@keithp.com> 103579Date: Wed Nov 10 16:11:29 2010 -0800 103580 103581 Merge remote branch 'whot/for-keith' 103582 103583commit 73f6de1ad8b5e1641102b54ff48a36ebb99ccacb 103584Merge: 333b6ed26 3dc3aefb8 103585Author: Keith Packard <keithp@keithp.com> 103586Date: Wed Nov 10 16:02:22 2010 -0800 103587 103588 Merge remote branch 'sandmann/fbdelete' 103589 103590commit 333b6ed26e36735ef72f314a3f8c4ec565af223f 103591Author: Kristian Høgsberg <krh@bitplanet.net> 103592Date: Thu Oct 28 20:46:23 2010 -0700 103593 103594 Set DamageSetReportAfterOp to true for the damage extension 103595 103596 Change the damage extension reporter to queue up events after we chain 103597 to the wrapped functions. Damage events are typically sent out after 103598 the rendering happens anyway, since we submit batch buffers from the 103599 flush callback chain and then flush client io buffers. Compositing 103600 managers relie on this order, and there is no way we could reliably 103601 provide damage events to clients before the rendering happens anyway. 103602 103603 By queueing up the damage events before the rendering happens, there's 103604 a risk that the client io buffer may overflow and send the damage 103605 events to the client before the driver has even seen the rendering 103606 request. Reporting damage events after the rendering fixes this 103607 corner case and better corresponds with how we expect this to work. 103608 103609 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 103610 Reviewed-by: Keith Packard <keithp@keithp.com> 103611 (cherry picked from commit 8d7b7a0d71e0b89321b3341b781bc8845386def6) 103612 [anholt: re-applied to revert the revert, now that the cause of the 103613 revert is fixed] 103614 Signed-off-by: Keith Packard <keithp@keithp.com> 103615 103616commit f36153e3ef6b13a87d016caab09cc9be274b0dd5 103617Author: Eric Anholt <eric@anholt.net> 103618Date: Thu Oct 28 20:46:22 2010 -0700 103619 103620 Replace usage of DamageRegionAppend with DamageDamageRegion to fix reportAfter. 103621 103622 In all these cases, any rendering implied by this damage has already 103623 occurred, and we want to get the damage out to the client. Some of 103624 the DamageRegionAppend calls were explicitly telling damage to flush 103625 the reportAfter damage out, but not all. 103626 103627 Bug #30260. Fixes the compiz wallpaper plugin with client damage 103628 changed to reportAfter. 103629 103630 Signed-off-by: Eric Anholt <eric@anholt.net> 103631 Reviewed-by: Keith Packard <keithp@keithp.com> 103632 Signed-off-by: Keith Packard <keithp@keithp.com> 103633 103634commit c80c41767eb101e9dbd8393d8cca7764b4e248a4 103635Author: Aaron Plattner <aplattner@nvidia.com> 103636Date: Mon Oct 25 22:01:32 2010 -0700 103637 103638 os: Fix BigReq ignoring when another request is pending 103639 103640 Commit cf88363db0ebb42df7cc286b85d30d7898aea840 fixed the handling of 103641 BigReq requests that are way too large and handles the case where the 103642 read() syscall returns a short read. However, it neglected to handle 103643 the case where it returns a long read, which happens when the client 103644 has another request in the queue after the bogus large one. 103645 103646 Handle the long read case by subtracting the smaller of 'needed' and 103647 'gotnow' from oci->ignoreBytes. If needed < gotnow, simply subtract 103648 the two, leaving gotnow equal to the number of extra bytes read. 103649 Since the code immediately following the (oci->ignoreBytes > 0) block 103650 tries to handle the next request, advance oci->bufptr immediately 103651 instead of setting oci->lenLastReq and letting the next call to 103652 ReadRequestFromClient do it. 103653 103654 Fixes the XTS pChangeKeyboardMapping-3 test. 103655 103656 CASES TESTS PASS UNSUP UNTST NOTIU WARN FIP FAIL UNRES UNIN ABORT 103657 -Xproto 122 389 367 2 19 0 0 0 1 0 0 0 103658 +Xproto 122 389 368 2 19 0 0 0 0 0 0 0 103659 103660 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 103661 Reviewed-by: Adam Jackson <ajax@redhat.com> 103662 Signed-off-by: Keith Packard <keithp@keithp.com> 103663 103664commit cc2c73ddcb4370a7c3ad439cda4da825156c26c9 103665Author: Luc Verhaegen <libv@skynet.be> 103666Date: Sun Oct 24 23:57:06 2010 +0200 103667 103668 x86emu: fix jump_near_IMM to handle DATA: flag correctly. 103669 103670 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=24348 103671 103672 Before (data flag ignored -> broken): 103673 66 DATA: 103674 e944f1 JMP 1ff6 103675 103676 After (fixed): 103677 66 DATA: 103678 e944f1ffff JMP 00001ff8 103679 103680 This subtle difference in the length of decoded instruction meant 103681 that the VBE call jumped to the routine setting AX=0x14F (VBE Failed) 103682 instead of the routine that set AX=0x4F (VBE success). 103683 103684 The ability to run the same code in vm86 significantly aided the 103685 debugging of this issue. Those X.org developers who would like to drop 103686 vm86 better take special care towards _all_ vesa bugs, as those will 103687 expose further issues. 103688 103689 Patch applies easily to even xserver 1.4.2. 103690 103691 Signed-off-by: Luc Verhaegen <libv@skynet.be> 103692 Tested-by: Luc Verhaegen <libv@skynet.be> 103693 Reviewed-by: Adam Jackson <ajax@redhat.com> 103694 Signed-off-by: Keith Packard <keithp@keithp.com> 103695 103696commit be7cf14c365c8ee0d69c4335e01316bcfcba69a4 103697Author: Siddhesh Poyarekar <siddhesh.poyarekar@gmail.com> 103698Date: Tue Oct 26 09:11:53 2010 +0530 103699 103700 mi: Integer overflow for dashed lines longer than 46340. #31093 103701 103702 Lines of length greater than 46340 can be drawn with one of the 103703 coordinates being negative. However for dashed lines, miPolyBuildPoly 103704 overflows the int type when setting up edges for a section of the 103705 dashed line. This results in the dashed segments not being drawn at 103706 all. 103707 103708 Signed-off-by: Siddhesh Poyarekar <siddhesh.poyarekar@gmail.com> 103709 Reviewed-by: Keith Packard <keithp@keithp.com> 103710 Signed-off-by: Keith Packard <keithp@keithp.com> 103711 103712commit a6c64d96456d794b2f3ce40915922a3fba636b95 103713Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 103714Date: Sun Oct 24 15:18:28 2010 +0200 103715 103716 Do not trap access to timer and keyboard 103717 103718 Disable timer/keyboard trapping on GNU/Hurd for now 103719 103720 Trapping disabled for now, as some VBIOSes (mga-g450 notably) use these 103721 ports, and the int10 wrapper is not emulating them. 103722 103723 It's effectively what happens in the Linux variant too, as iopl() is used there, 103724 making the ioperm() meaningless. 103725 103726 Signed-off-by: Olaf Buddenhagen <antrik@users.sf.net> 103727 Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 103728 Reviewed-by: Adam Jackson <ajax@redhat.com> 103729 Signed-off-by: Keith Packard <keithp@keithp.com> 103730 103731commit 69e8e1b0b95a325da3e3a2d76d092e7131baa9ad 103732Author: Julien Cristau <jcristau@debian.org> 103733Date: Mon May 17 19:39:54 2010 +0200 103734 103735 os: include dix-config.h, not xorg-config.h 103736 103737 os/strlc{at,py}.c were trying to include xorg-config.h, which is not 103738 available in dix. 103739 103740 Signed-off-by: Julien Cristau <jcristau@debian.org> 103741 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 103742 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103743 103744commit acba00c0c1129fe4ef7655f018cd21dbb1ee9a1f 103745Merge: 383dfe23f 44e816541 103746Author: Keith Packard <keithp@keithp.com> 103747Date: Wed Nov 10 14:41:08 2010 -0800 103748 103749 Merge remote branch 'dottedmag/master' 103750 103751commit 383dfe23f1641435c0713f8a56424814a156af8e 103752Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 103753Date: Mon Oct 25 17:13:57 2010 +0300 103754 103755 DRI2: Avoid call to NULL pointer 103756 103757 DDX driver may implement schedule swap without GetMSC. In that case we 103758 can't call GetMSC in DRI2SwapBuffers. 103759 103760 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 103761 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 103762 CC: Keith Packard <keithp@keithp.com> 103763 CC: Kristian Høgsberg <krh@bitplanet.net> 103764 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 103765 Signed-off-by: Keith Packard <keithp@keithp.com> 103766 103767commit 68e4a628d65312df93cc71f3e76241584c4bbd23 103768Author: Cyril Brulebois <kibi@debian.org> 103769Date: Wed Nov 10 15:28:29 2010 +0100 103770 103771 Remove more superfluous if(p!=NULL) checks around free(p). 103772 103773 This patch has been generated by the following Coccinelle semantic patch: 103774 103775 @@ 103776 expression E; 103777 @@ 103778 - if (E != NULL) { 103779 - free(E); 103780 - } 103781 + free(E); 103782 103783 Signed-off-by: Cyril Brulebois <kibi@debian.org> 103784 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 103785 Reviewed-by: Matt Turner <mattst88@gmail.com> 103786 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 103787 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103788 103789commit b142b0d27442191d628762604f4eb1f65263d717 103790Author: Cyril Brulebois <kibi@debian.org> 103791Date: Mon Nov 8 23:35:33 2010 +0100 103792 103793 Remove more superfluous if(p!=NULL) checks around free(p). 103794 103795 This patch has been generated by the following Coccinelle semantic patch: 103796 103797 @@ 103798 expression E; 103799 @@ 103800 - if (E != NULL) 103801 - free(E); 103802 + free(E); 103803 103804 Signed-off-by: Cyril Brulebois <kibi@debian.org> 103805 Reviewed-by: Matt Turner <mattst88@gmail.com> 103806 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 103807 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103808 103809commit 0649ac0afdb3f8ed11c2634563d6b9df161cb9ec 103810Author: Cyril Brulebois <kibi@debian.org> 103811Date: Mon Nov 8 23:35:32 2010 +0100 103812 103813 Remove more superfluous if(p) checks around free(p). 103814 103815 This patch has been generated by the following Coccinelle semantic patch: 103816 103817 @@ 103818 expression E; 103819 @@ 103820 - if (E) 103821 - free(E); 103822 + free(E); 103823 103824 Signed-off-by: Cyril Brulebois <kibi@debian.org> 103825 Reviewed-by: Matt Turner <mattst88@gmail.com> 103826 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 103827 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103828 103829commit 99275ad2fa99778afaefc54b62c8638afc59e755 103830Author: Cyril Brulebois <kibi@debian.org> 103831Date: Wed Nov 10 16:06:10 2010 +0100 103832 103833 Remove superfluous if(p!=NULL) checks around free(p); p=NULL; 103834 103835 This patch has been generated by the following Coccinelle semantic patch: 103836 103837 @@ 103838 expression E; 103839 @@ 103840 - if (E != NULL) { 103841 - free(E); 103842 ( 103843 - E = NULL; 103844 | 103845 - E = 0; 103846 ) 103847 - } 103848 + free(E); 103849 + E = NULL; 103850 103851 Signed-off-by: Cyril Brulebois <kibi@debian.org> 103852 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 103853 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 103854 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103855 103856commit 9e999d18b004b8ead9c6c5d79b4a3d4bbf0e3152 103857Author: Peter Hutterer <peter.hutterer@who-t.net> 103858Date: Fri Nov 5 12:08:11 2010 +1000 103859 103860 Xi: if XTEST device creation fails, fail the master devices. 103861 103862 When getting close to the MAXDEVICES limit, the creation of XTEST devices 103863 may fail due to device id exhaustion. In that case, fail the creation of 103864 master devices too and return an error to the client. 103865 103866 Theoretically, we could alloc the MDs without the XTEST devices but that 103867 will get interesting when a client starts sending XTEST events through those 103868 devices. 103869 103870 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103871 Reviewed-by: Julien Cristau <jcristau@debian.org> 103872 103873commit 5a455e0c80d433adc4109ebf313fd92afa194545 103874Author: Peter Hutterer <peter.hutterer@who-t.net> 103875Date: Fri Nov 5 11:49:12 2010 +1000 103876 103877 Xi: rename two variables from ptr to dev. 103878 103879 They were named ptr when everything was in one function to save one more 103880 variable. Now that the stuff is split out, "dev" makes more sense. 103881 103882 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103883 Reviewed-by: Julien Cristau <jcristau@debian.org> 103884 103885commit 9b89b91c685426c9944f7fc8890f436c18b7583c 103886Author: Peter Hutterer <peter.hutterer@who-t.net> 103887Date: Fri Nov 5 11:47:43 2010 +1000 103888 103889 Xi: split hierarchy manipulation into static functions. 103890 103891 No functional changes, just code cleanup to improve readability. 103892 103893 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103894 Reviewed-by: Julien Cristau <jcristau@debian.org> 103895 103896commit 4132b1c591bd01c29df91b88afbd144647da9ed8 103897Author: Peter Hutterer <peter.hutterer@who-t.net> 103898Date: Thu Oct 21 14:04:42 2010 +1000 103899 103900 xfree86: fix compiler warnings - unused variable i 103901 103902 ../../../../hw/xfree86/common/xf86Config.c: In function 'configDRI': 103903 ../../../../hw/xfree86/common/xf86Config.c:2213:9: warning: unused variable 103904 'i' 103905 103906 Introduced in 788bfbf18ac19923604b0d676933fdc0d78526a1 103907 103908 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103909 Reviewed-by: Matt Turner <mattst88@gmail.com> 103910 103911commit 6b982d73bb8878238ed46665b220a94a591e11a9 103912Author: Peter Hutterer <peter.hutterer@who-t.net> 103913Date: Thu Oct 21 13:53:35 2010 +1000 103914 103915 xfree86: fix two compiler warnings - unused variable ati, atimisc 103916 103917 ../../../../hw/xfree86/common/xf86Config.c: In function 103918 'fixup_video_driver_list': 103919 ../../../../hw/xfree86/common/xf86Config.c:507:19: warning: unused variable 103920 'atimisc' 103921 ../../../../hw/xfree86/common/xf86Config.c:507:12: warning: unused variable 103922 'ati' 103923 103924 Introduced in 52577ae8eeda64601db8dd425027cf8e4271b873. 103925 103926 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103927 Reviewed-by: Matt Turner <mattst88@gmail.com> 103928 103929commit 0799f0771cca335d6400a8494ba236a0d10e122b 103930Author: Peter Hutterer <peter.hutterer@who-t.net> 103931Date: Thu Nov 4 12:53:18 2010 +1000 103932 103933 xfree86: remove two unused variables from Makefile.am 103934 103935 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 103936 Reviewed-by: Julien Cristau <jcristau@debian.org> 103937 103938commit a52efb096e166e325deb3d6b502671f339a4fa15 103939Merge: fe8d122b1 f72aadd38 103940Author: Keith Packard <keithp@keithp.com> 103941Date: Wed Nov 3 06:50:45 2010 -0700 103942 103943 Merge remote branch 'sthibaul/master-iopl' 103944 103945commit 3dc3aefb88c167d365795dcd44017742c52e4967 103946Author: Søren Sandmann Pedersen <ssp@redhat.com> 103947Date: Mon Nov 1 22:23:05 2010 -0400 103948 103949 fb: Delete a bunch more left-overs 103950 103951 All of these definitions were unused since compositing moved to pixman. 103952 103953 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 103954 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 103955 Signed-off-by: Søren Sandmann <ssp@redhat.com> 103956 103957commit 565b5e2b0261d1148cc2707ffcbb915414dd3e02 103958Author: Søren Sandmann Pedersen <ssp@redhat.com> 103959Date: Mon Nov 1 22:14:29 2010 -0400 103960 103961 wfb: Delete unused renamings 103962 103963 These functions haven't existed in a while. 103964 103965 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 103966 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 103967 Signed-off-by: Søren Sandmann <ssp@redhat.com> 103968 103969commit 0dae479e16590e767df35985150c5acf696e5df2 103970Author: Søren Sandmann Pedersen <ssp@redhat.com> 103971Date: Mon Nov 1 22:17:08 2010 -0400 103972 103973 fb: Delete fbCompositeGeneral() 103974 103975 This function was an unused and trivial wrapper around fbComposite(). 103976 103977 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 103978 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 103979 Signed-off-by: Søren Sandmann <ssp@redhat.com> 103980 103981commit 74bc2d8c6bafc5373ff9c14eed12e26b5b965953 103982Author: Søren Sandmann Pedersen <ssp@redhat.com> 103983Date: Mon Nov 1 22:12:03 2010 -0400 103984 103985 render: Delete renderedge.[ch] 103986 103987 The functions in these files have not been used since trap 103988 rasterization was moved to pixman. They survived until now to preserve 103989 the server abi. 103990 103991 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 103992 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 103993 Signed-off-by: Søren Sandmann <ssp@redhat.com> 103994 103995commit 6118346d64e3c2fbe1fe2f041ea773dd2a3c0438 103996Author: Søren Sandmann Pedersen <ssp@redhat.com> 103997Date: Mon Nov 1 22:04:49 2010 -0400 103998 103999 Delete unused fbWalkCompositeRegion() 104000 104001 This function has not been used since most of the compositing was 104002 moved to pixman. The only reason it has survived until now is that it 104003 was part of the server ABI. 104004 104005 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 104006 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 104007 Signed-off-by: Søren Sandmann <ssp@redhat.com> 104008 104009commit fe8d122b133a43125cc961d2c3c811fa776f29c8 104010Author: Peter Hutterer <peter.hutterer@who-t.net> 104011Date: Thu Oct 28 16:13:09 2010 +1000 104012 104013 Xi: reshuffle conditions for labeling a device as IsXExtensionKeyboard (#29046) 104014 104015 From the original bug reporter Ezra Reeves: 104016 104017 "I did some more digging on this today, and I found that an HP branded 104018 wireless USB mouse has the same issue. With this mouse (as well as the 104019 logitech wireless mouse), the return from: 104020 104021 xdev = XListInputDevices(GDK_WINDOW_XDISPLAY(rootwin), &ndevices_return); 104022 104023 lists the USB device twice, but both have xdev[num].use == 3 104024 (IsXExtensionKeyboard as defined in X11/XI.h). 104025 104026 [...] 104027 104028 Swapping the order of the test in Xi/listdev.c that determines whether a 104029 device is a pointer or a keyboard properly detects my devices (OEM USB 104030 wireless mouse/kb combo) -- one as a keyboard and one as a pointer." 104031 104032 X.Org Bug 29046 <http://bugs.freedesktop.org/show_bug.cgi?id=29046> 104033 104034 Reported-by: Erik Kilfoil <ekilfoil@gmail.com> 104035 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 104036 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104037 104038commit a210068c5222acbdb04db17cb8cef314735bf977 104039Author: Tiago Vignatti <tiago.vignatti@nokia.com> 104040Date: Tue Oct 26 20:50:13 2010 +0300 104041 104042 dix: advance parent window pointer when no node is found 104043 104044 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 104045 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 104046 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 104047 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104048 104049commit 19f43836d1006d253391926e880b14ed2e7a0cf8 104050Author: Peter Hutterer <peter.hutterer@who-t.net> 104051Date: Mon Oct 25 15:10:19 2010 +1000 104052 104053 Re-export xf86CollectInputOptions. 104054 104055 Some drivers, most notably the mouse driver need this and reimplementing on 104056 the driver side doesn't make sense. 104057 104058 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104059 104060commit 4613ce3f78a32f2a6f422e719150d55b035c758e 104061Author: Peter Hutterer <peter.hutterer@who-t.net> 104062Date: Fri Oct 22 15:17:50 2010 +1000 104063 104064 test: valuator_mode tests. 104065 104066 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104067 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104068 104069commit 58554f1c6429535e7ab432c920aabca71a77edb8 104070Author: Peter Hutterer <peter.hutterer@who-t.net> 104071Date: Fri Oct 22 15:10:32 2010 +1000 104072 104073 Convert some leftover axes->mode access to valuator_get_mode() 104074 104075 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104076 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104077 104078commit 44e81654147065b7b907ce0d14fdc2e74a638f79 104079Author: Mikhail Gusarov <dottedmag@dottedmag.net> 104080Date: Mon Oct 25 01:30:33 2010 +0400 104081 104082 Remove now-misleading comment 104083 104084 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 104085 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 104086 104087commit 433dddcf29d257310e95e4215dceef8b0b976e2d 104088Author: Mikhail Gusarov <dottedmag@dottedmag.net> 104089Date: Mon Oct 25 01:28:38 2010 +0400 104090 104091 Replace "if(buf) realloc(buf, size) else malloc(size)" with realloc() 104092 104093 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 104094 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 104095 104096commit 04a04e162886ca9bc3977d3a7509def876487921 104097Author: Mikhail Gusarov <dottedmag@dottedmag.net> 104098Date: Mon Oct 25 01:27:46 2010 +0400 104099 104100 Remove now-misleading comment 104101 104102 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 104103 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 104104 104105commit f72aadd38b730c46c0a027dacbbb57f03068383d 104106Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 104107Date: Sat Oct 23 22:42:33 2010 +0200 104108 104109 hurd: Fix use of deprecated iopl device 104110 104111 Fix Xserver on GNU/Hurd into using the "mem" device instead of 104112 the deprecated "iopl" device. 104113 104114 Reviewed-by: Olaf Buddenhagen <antrik@users.sf.net> 104115 Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org> 104116 104117commit dd11f734a9a73bf34357c149d2809842938d57b5 104118Author: Peter Hutterer <peter.hutterer@who-t.net> 104119Date: Fri Oct 22 16:11:24 2010 +1000 104120 104121 input: remove "mode" field from ValuatorClassRec. 104122 104123 We have per-axis mode now. For those bits that still need it (XI 1.x), 104124 assume that the first axis holds the device's mode. 104125 104126 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104127 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104128 104129commit 5cf3b654fccf09de63a3f983c85096cda78fb693 104130Author: Peter Hutterer <peter.hutterer@who-t.net> 104131Date: Fri Oct 22 15:49:40 2010 +1000 104132 104133 input: move proximity state into ProximityClassRec. 104134 104135 Previously the OutOfProximity bit in the valuator mode. 104136 104137 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104138 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104139 104140commit e909af88bf17dbadcc00d50c13a5bdb6001367be 104141Author: Peter Hutterer <peter.hutterer@who-t.net> 104142Date: Fri Oct 22 14:22:44 2010 +1000 104143 104144 dmx: clarify a valuator mode condition. 104145 104146 Relative is defined as 0, so change the condition to be more obvious. 104147 104148 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104149 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104150 104151commit b5ef88c911f697530ddc2c7302fce15ba2ead4b2 104152Author: Peter Hutterer <peter.hutterer@who-t.net> 104153Date: Fri Oct 22 14:02:17 2010 +1000 104154 104155 dix: clip absolute axes depending on their mode. 104156 104157 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104158 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104159 104160commit ea567b675f814ac41e75f1ed8ded0ac3e9d552a8 104161Author: Peter Hutterer <peter.hutterer@who-t.net> 104162Date: Fri Oct 22 14:01:11 2010 +1000 104163 104164 dix: populate motion history only if the mode matches the first axis. 104165 104166 XI1 doesn't cater for mixed mode devices, so bail out on the first valuator 104167 that has a different mode. 104168 104169 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104170 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104171 104172commit 4381b70f5aa293400735e9b4e5a792408afa25a2 104173Author: Peter Hutterer <peter.hutterer@who-t.net> 104174Date: Fri Oct 22 13:57:27 2010 +1000 104175 104176 input: add valuator_get_mode() helper. 104177 104178 Returns the mode of the specified valuator. 104179 104180 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104181 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104182 104183commit 6f6f460c2485d16d126a1ab5d939083c322dc3dd 104184Author: Peter Hutterer <peter.hutterer@who-t.net> 104185Date: Fri Oct 22 13:48:06 2010 +1000 104186 104187 dix: send proximity events if one or more axes are Absolute. 104188 104189 We only skip relative events for proximity, not absolute ones. Now with 104190 mixed mode, just unset those axes that are relative. 104191 104192 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104193 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104194 104195commit 65c0fc81eb920085e650b8c9e874c9dd26c7ec98 104196Author: Chase Douglas <chase.douglas@canonical.com> 104197Date: Fri Jul 16 09:21:19 2010 -0400 104198 104199 Add support for per-axis valuator modes (Relative/Absolute) 104200 104201 The XI2 protocol supports per-axis modes, but the server so far does 104202 not. This change adds support in the server. 104203 104204 A complication is the fact that XI1 does not support per-axis modes. 104205 The solution provided here is to set a per-device mode that defines the 104206 mode of at least the first two valuators (X and Y). Note that initializing 104207 the first two axes to a different mode than the device mode will fail. 104208 104209 For XI1 events, any axes following the first two that have the same mode 104210 will be sent to clients, up to the first axis that has a different mode. 104211 Thus, if a device has relative, then absolute, then relative mode axes, 104212 only the first block of relative axes will be sent over XI1. 104213 104214 Since the XI2 protocol supports per-axis modes, all axes are sent to the 104215 client. 104216 104217 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 104218 104219commit 290af0418f8c94355e42ad81aa7779b90e173fc9 104220Author: Chase Douglas <chase.douglas@ubuntu.com> 104221Date: Thu Oct 21 22:41:09 2010 -0400 104222 104223 test: input - set valuators mask for event to core conversion 104224 104225 Commit de8be07cc0a8163b6ef04455706fd5ca2cebe587 adds a requirement to 104226 event to core conversion that at least one of the X or Y valuators are 104227 set in the valuator mask. This commit fixes the event conversion test to 104228 be compliant. 104229 104230 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 104231 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 104232 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104233 104234commit 8990b31214bcbc08090604147287455bfde91c11 104235Author: Paulius Zaleckas <paulius.zaleckas@gmail.com> 104236Date: Mon Oct 18 00:02:01 2010 +0300 104237 104238 KDrive: Fix error handlig in tslib driver 104239 104240 If ts_open() fails and return NULL, then next call to ts_fd() 104241 segfaults because of NULL dereference. There is no need to 104242 check output of ts_fd() as ts_open() did this internally. 104243 104244 Signed-off-by: Paulius Zaleckas <paulius.zaleckas@gmail.com> 104245 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 104246 Reviewed-by: Keith Packard <keithp@keithp.com> 104247 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104248 104249commit 9696c782c8cb86b06d12949899582533a2e04cfe 104250Author: Peter Hutterer <peter.hutterer@who-t.net> 104251Date: Tue Oct 19 12:52:57 2010 +1000 104252 104253 test: valuator_mask tests. 104254 104255 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104256 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104257 104258commit 2362adc2da9745e01c79ed4905aa81926355c4e2 104259Author: Peter Hutterer <peter.hutterer@who-t.net> 104260Date: Tue Oct 19 10:59:09 2010 +1000 104261 104262 dix: remove valuator_mask_copy_valuators, not needed anymore. 104263 104264 With the switch to masks internally, this isn't needed anymore. 104265 104266 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104267 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104268 104269commit 2b04a3ef739a7ba5708400999e5d81418236a0a4 104270Author: Peter Hutterer <peter.hutterer@who-t.net> 104271Date: Tue Oct 19 09:58:58 2010 +1000 104272 104273 xfree86: convert xf86Post{Motion|Key|Button|Proximity}Event to masks 104274 104275 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104276 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104277 104278commit 0418a39e71e50b88e050b0d8a758265b17c4bf0e 104279Author: Peter Hutterer <peter.hutterer@who-t.net> 104280Date: Mon Oct 18 15:55:53 2010 +1000 104281 104282 dix: get rid of the now-superfluous valuator arrays in GPE and friends. 104283 104284 The valuators are stored inside the mask, use it from there. are stored 104285 inside the mask, use it from there. are stored inside the mask, use it from 104286 there. are stored inside the mask, use it from there. 104287 104288 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104289 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104290 104291commit 675f4a8525d29ebad783351e17be785b2f32b2e8 104292Author: Peter Hutterer <peter.hutterer@who-t.net> 104293Date: Tue Oct 19 13:37:46 2010 +1000 104294 104295 Abstract valuator masks through a set of APIs. 104296 104297 This commit introduces an abstraction API for handling masked valuators. The 104298 intent is that drivers just allocate a mask, set the data and pass the mask 104299 to the server. The actual storage type of the mask is hidden from the 104300 drivers. 104301 104302 The new calls for drivers are: 104303 valuator_mask_new() /* to allocate a valuator mask */ 104304 valuator_mask_zero() /* to reset a mask to zero */ 104305 valuator_mask_set() /* to set a valuator value */ 104306 104307 The new interface to the server is 104308 xf86PostMotionEventM() 104309 xf86PostButtonEventM() 104310 xf86PostKeyboardEventM() 104311 xf86PostProximityEventM() 104312 104313 all taking a mask instead of the valuator array. 104314 104315 The ValuatorMask is currently defined for MAX_VALUATORS fixed size due to 104316 memory allocation restrictions in SIGIO handlers. 104317 104318 For easier review, a lot of the code still uses separate valuator arrays. 104319 This will be fixed in a later patch. 104320 104321 This patch was initially written by Chase Douglas. 104322 104323 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 104324 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104325 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104326 104327commit 1a0d9324b3d9fd93e685066e0e5cea0611878c0d 104328Author: Aaron Plattner <aplattner@nvidia.com> 104329Date: Sun Oct 17 09:58:50 2010 -0700 104330 104331 Revert "Set DamageSetReportAfterOp to true for the damage extension" (#30260) 104332 104333 This commit breaks the Compiz "Wallpaper" plugin. 104334 104335 This reverts commit 8d7b7a0d71e0b89321b3341b781bc8845386def6. 104336 104337 Reviewed-by: Keith Packard <keithp@keithp.com> 104338 Signed-off-by: Keith Packard <keithp@keithp.com> 104339 104340commit d738175eaf1098e29b8afb6de8e99b5098e366a7 104341Merge: 8806a04f7 9872e0f01 104342Author: Keith Packard <keithp@keithp.com> 104343Date: Tue Oct 19 22:50:15 2010 -0700 104344 104345 Merge remote branch 'whot/for-keith' 104346 104347commit fc48a8f9f5f66e591b3e39211d44ce68267303f8 104348Author: Chase Douglas <chase.douglas@canonical.com> 104349Date: Tue Oct 19 13:37:38 2010 +1000 104350 104351 Add CountBits() to the server. 104352 104353 Function to count the number of bits set in the given array. 104354 104355 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 104356 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104357 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104358 104359commit 9872e0f0115cd7d0543e55a1831facddf3aff90c 104360Author: Peter Hutterer <peter.hutterer@who-t.net> 104361Date: Wed Oct 13 10:31:51 2010 +1000 104362 104363 xfree86: set mask for valuators 0/1 when emulating core events (#30267) 104364 104365 EventToCore as of the commit below won't generate core motion events if the 104366 valuator mask for x/y isn't set. For DGA, we work around this check by 104367 forcibly setting the mask in the event we pass down. 104368 104369 commit de8be07cc0a8163b6ef04455706fd5ca2cebe587 104370 Author: Peter Hutterer <peter.hutterer@who-t.net> 104371 Date: Tue Aug 17 12:08:52 2010 +1000 104372 104373 dix: don't create core motion events for non-x/y valuators. 104374 104375 X.Org Bug 30267 <http://bugs.freedesktop.org/show_bug.cgi?id=30267> 104376 104377 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104378 Tested-by: Andrew Randrianasulu <randrik@mail.ru> 104379 Tested-by: Andy Furniss 104380 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 104381 104382commit 649293f6b634e6305b6737a841d6e9d0f0065d6c 104383Author: Peter Hutterer <peter.hutterer@who-t.net> 104384Date: Mon Oct 11 16:09:18 2010 +1000 104385 104386 xkb: always fill the symsPerKey array, regardless of client flags (#30527) 104387 104388 Even if a client does not modify the symbols, symsPerKey and mapWidths must 104389 be filled from the current configuration. Both arrays are then passed into 104390 other functions (pending the right flag), thus they must contain valid 104391 values regardless of the XkbKeySymsMask flag in req->present. 104392 104393 X.Org Bug 30527 <http://bugs.freedesktop.org/show_bug.cgi?id=30527> 104394 104395 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104396 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 104397 104398commit ca21a266224b6eff4fd817c2082d2e144f1ea58c 104399Author: Peter Hutterer <peter.hutterer@who-t.net> 104400Date: Mon Oct 11 15:38:38 2010 +1000 104401 104402 xkb: init mapWidth and symsPerKey arrays to 0. 104403 104404 Helps debugging greatly, random 8 or 16 bit values can sometimes look like 104405 valid values, causing much excitement on the client front. 104406 104407 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104408 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 104409 104410commit 8806a04f7f028dd9f6a7ff5c36e460d3895c7deb 104411Author: Michał Górny <mgorny@gentoo.org> 104412Date: Thu Oct 14 15:49:51 2010 +0200 104413 104414 kdrive: Fix linking with tslib, with -Wl, --as-needed. #30600 104415 104416 Change the library order in TSLIB_LIBS variable, placing ${TSLIB_LIBS} 104417 after the local static libraries. This fixes linking with 104418 -Wl,--as-needed. 104419 104420 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=30600 104421 Signed-off-by: Michał Górny <mgorny@gentoo.org> 104422 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 104423 Signed-off-by: Keith Packard <keithp@keithp.com> 104424 104425commit c7e436e9987a6fb0876c63e96f59ff4c20b4bef0 104426Author: Michał Górny <mgorny@gentoo.org> 104427Date: Thu Oct 14 15:49:50 2010 +0200 104428 104429 kdrive: Fix tslib check fallback to set TSLIB_LIBS. #30599 104430 104431 If pkg-config is unable to find tslib but the fallback check does find 104432 it, the compilation continues with tslib support enabled though 104433 TSLIB_LIBS are unset. Thus, the compilation fails with a linking error 104434 on tslib functions. 104435 104436 This patch sets TSLIB_LIBS to '-lts' whenever the tslib fallback check 104437 succeeds. 104438 104439 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=30599 104440 Signed-off-by: Michał Górny <mgorny@gentoo.org> 104441 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 104442 Signed-off-by: Keith Packard <keithp@keithp.com> 104443 104444commit 25801a2556ca9b42646023e5a95a3cbf7309e852 104445Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104446Date: Thu Oct 14 22:38:17 2010 +0100 104447 104448 Cygwin/X: Move duplicate extern variable declarations from various .c files to a new header file 104449 104450 For the global variables defined in winglobals.c, remove duplicate extern 104451 declarations from the beginning of various .c files, and move most of them 104452 into a new header file, winglobals.h 104453 104454 Leave some clipboard related variables alone for the moment, they need treating 104455 more carefully, to avoid mixing client and server type definitions. 104456 104457 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104458 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104459 104460commit 04d11abcf2ea2f36d899428e9c885b69e4f779ad 104461Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104462Date: Thu Oct 14 22:45:31 2010 +0100 104463 104464 Cygwin/X: Remove g_hwndKeyboardFocus 104465 104466 It's set but it's value is never used 104467 104468 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104469 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104470 104471commit 5e896e425fe14022b0714301531bd54151c763f0 104472Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104473Date: Thu Oct 14 22:43:51 2010 +0100 104474 104475 Cygwin/X: Make g_hmodCommonControls static 104476 104477 Make g_hmodCommonControls static, it's only used in InitOutput.c 104478 104479 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104480 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104481 104482commit 49b996faadf0fedcee25d4a8f96f4c4007c29f5f 104483Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104484Date: Fri Oct 15 13:00:46 2010 +0100 104485 104486 Cygwin/X: Make g_hmodDirectDraw static 104487 104488 Add winReleaseDDProcAddresses() for releasing the direct draw module 104489 handle, so g_hmodDirectDraw can be made static 104490 104491 Remove unneeded duplicate extern definitions of g_fpDirectDrawCreate 104492 and g_fpDirectDrawCreateClipper, and move the definition from 104493 winglobals.c to winengine.c 104494 104495 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104496 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104497 104498commit 6f12277e37d16a7de9d3f3079688c43b07921b0a 104499Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104500Date: Thu Oct 14 22:44:53 2010 +0100 104501 104502 Cygwin/X: Make g_hhookKeyboardLL static 104503 104504 Make g_hhookKeyboardLL static, it's only used in winkeyhook.c 104505 104506 Also remove unused externs 104507 104508 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104509 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104510 104511commit 9952b51f6822aeaa19c27caf85698741eb4c27d8 104512Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104513Date: Fri Oct 15 13:01:45 2010 +0100 104514 104515 Cygwin/X: Fix typos in comments 104516 104517 Fix a few typos in comments 104518 104519 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104520 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104521 104522commit 1de5db73d31411798ae93619d9e626fa32e0c4bd 104523Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104524Date: Thu Oct 14 22:46:09 2010 +0100 104525 104526 Cygwin/X: Drop several unneeded includes of winprefs.h 104527 104528 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104529 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104530 104531commit 8e723108151136f115e8e6d4b97ed4e2fa685dbf 104532Author: Colin Harrison <colin.harrison@virgin.net> 104533Date: Wed Sep 29 14:18:27 2010 +0100 104534 104535 Xming: Remove uses of register keyword 104536 104537 I'm pretty sure the compiler has a better idea how to optimize this 104538 104539 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104540 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104541 104542commit d11761c6a6dd2a9963dfd3bcb313f1b2525387fc 104543Author: Colin Harrison <colin.harrison@virgin.net> 104544Date: Fri Feb 26 13:31:18 2010 +0000 104545 104546 Xming: Casts to remove warnings on event pointers 104547 104548 winkeybd.c: In function ‘winSendKeyEvent’: 104549 winkeybd.c:489: warning: passing argument 2 of ‘mieqEnqueue’ from incompatible pointer type 104550 winmouse.c: In function ‘winMouseButtonsSendEvent’: 104551 winmouse.c:247: warning: passing argument 2 of ‘mieqEnqueue’ from incompatible pointer type 104552 winmouse.c: In function ‘winEnqueueMotion’: 104553 winmouse.c:380: warning: passing argument 2 of ‘mieqEnqueue’ from incompatible pointer type 104554 104555 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104556 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104557 104558commit 10bf8345cf15c6030147a9edb7f6a093c8db8099 104559Author: Colin Harrison <colin.harrison@virgin.net> 104560Date: Mon Oct 12 13:50:00 2009 +0100 104561 104562 Xming: Warning fix in winDeviceCursorCleanup() 104563 104564 return statement with an expression in a function whose return-type is void 104565 104566 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104567 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104568 104569commit cab837a60a7aa653ccea77ff714335186ce38a4b 104570Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104571Date: Sun Jul 18 13:22:46 2010 +0100 104572 104573 Cygwin/X: Reformat the keyboard layout mapping table 104574 104575 Also remove the out-of-date listing of keyboard layout IDs and which ones 104576 have known mappings 104577 104578 XXX: layout zh_TW doesn't exist anymore 104579 104580 This patch brought to you by C-u M-x align-regexp 104581 104582 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104583 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104584 104585commit ee2867649d9485f6559f88a973eae6393971ff35 104586Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104587Date: Sun Jul 18 13:20:01 2010 +0100 104588 104589 Cygwin/X: Clarify XKB options in XWin man page 104590 104591 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104592 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104593 104594commit 1ba1b12d24d8f825babd8cf38710c7fa9a94821e 104595Author: Colin Harrison <colin.harrison@virgin.net> 104596Date: Fri Oct 15 13:46:09 2010 +0100 104597 104598 Xming: Warning fix in winApplyHints() 104599 104600 winmultiwindowwm.c: In function ‘winApplyHints’: 104601 winmultiwindowwm.c:1587: warning: ISO C90 forbids mixed declarations and code 104602 104603 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104604 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104605 104606commit fb64414548c437bf52c3dcd0613aab5079faffba 104607Author: Colin Harrison <colin.harrison@virgin.net> 104608Date: Fri Oct 15 14:19:57 2010 +0100 104609 104610 Xming: Fix warnings in ProcWindowsWMFrameSetTitle() 104611 104612 winwindowswm.c: In function ‘ProcWindowsWMFrameSetTitle’: 104613 winwindowswm.c:516: warning: pointer targets in passing argument 1 of ‘strncpy’ differ in signedness 104614 winwindowswm.c:516: warning: pointer targets in passing argument 2 of ‘strncpy’ differ in signedness 104615 winwindowswm.c:528: warning: pointer targets in passing argument 2 of ‘SetWindowTextA’ differ in signedness 104616 104617 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104618 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104619 104620commit da5955cc023ae980fdc0397f6fd541e788040c29 104621Author: Colin Harrison <colin.harrison@virgin.net> 104622Date: Fri Oct 15 14:18:53 2010 +0100 104623 104624 Xming: Fix warnings in windialogs.c 104625 104626 windialogs.c: In function ‘winDisplayExitDialog’: 104627 windialogs.c:327: warning: passing argument 3 of ‘PostMessageA’ makes integer from pointer without a cast 104628 windialogs.c: In function ‘winDisplayAboutDialog’: 104629 windialogs.c:597: warning: passing argument 3 of ‘PostMessageA’ makes integer from pointer without a cast 104630 windialogs.c: In function ‘winAboutDlgProc’: 104631 windialogs.c:697: warning: comparison between pointer and integer 104632 windialogs.c:701: warning: format ‘%d’ expects type ‘int’, but argument 2 has type ‘HINSTANCE’ 104633 windialogs.c:716: warning: assignment makes integer from pointer without a cast 104634 windialogs.c:736: warning: assignment makes integer from pointer without a cast 104635 windialogs.c:756: warning: assignment makes integer from pointer without a cast 104636 104637 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104638 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104639 104640commit 6d9fb07db210bcfdcdde632fa341c4ee08d44b35 104641Author: Colin Harrison <colin.harrison@virgin.net> 104642Date: Fri Oct 15 14:16:54 2010 +0100 104643 104644 Xming: Fix warnings in winNameCompare(), winNormalizeName() 104645 104646 winconfig.c: In function ‘winNameCompare’: 104647 winconfig.c:715: warning: array subscript has type ‘char’ 104648 winconfig.c:715: warning: array subscript has type ‘char’ 104649 winconfig.c:716: warning: array subscript has type ‘char’ 104650 winconfig.c:716: warning: array subscript has type ‘char’ 104651 winconfig.c:730: warning: array subscript has type ‘char’ 104652 winconfig.c:730: warning: array subscript has type ‘char’ 104653 winconfig.c:731: warning: array subscript has type ‘char’ 104654 winconfig.c:731: warning: array subscript has type ‘char’ 104655 winconfig.c: In function ‘winNormalizeName’: 104656 winconfig.c:1092: warning: array subscript has type ‘char’ 104657 winconfig.c:1093: warning: array subscript has type ‘char’ 104658 104659 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104660 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104661 104662commit 20510e51001d1659000c8d8607c0de16022d1c74 104663Author: Colin Harrison <colin.harrison@virgin.net> 104664Date: Fri Oct 15 14:15:51 2010 +0100 104665 104666 Xming: Fix warnings in winClipboardFlushXEvents() 104667 104668 winclipboardxevents.c: In function ‘winClipboardFlushXEvents’: 104669 winclipboardxevents.c:225: warning: format ‘%08x’ expects type ‘unsigned int’, but argument 2 has type ‘DWORD’ 104670 winclipboardxevents.c:266: warning: format ‘%08x’ expects type ‘unsigned int’, but argument 2 has type ‘DWORD’ 104671 winclipboardxevents.c:619: warning: pointer targets in passing argument 1 of ‘strcat’ differ in signedness 104672 winclipboardxevents.c:659: warning: pointer targets in passing argument 1 of ‘strlen’ differ in signedness 104673 winclipboardxevents.c:669: warning: pointer targets in passing argument 3 of ‘MultiByteToWideChar’ differ in signedness 104674 winclipboardxevents.c:690: warning: pointer targets in passing argument 3 of ‘MultiByteToWideChar’ differ in signedness 104675 winclipboardxevents.c:698: warning: pointer targets in passing argument 1 of ‘strdup’ differ in signedness 104676 104677 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104678 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104679 104680commit 44c8b449d281c4d984d65f66bac96a65806b2efa 104681Author: Colin Harrison <colin.harrison@virgin.net> 104682Date: Fri Oct 15 14:13:57 2010 +0100 104683 104684 Xming: Fix warning in winClipboardErrorHandler() 104685 104686 winclipboardthread.c: In function ‘winClipboardErrorHandler’: 104687 winclipboardthread.c:444: warning: format ‘%d’ expects type ‘int’, but argument 3 has type ‘long unsigned int’ 104688 104689 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104690 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104691 104692commit 1f2a57d9e46424a2e0861f55253b91d930d79083 104693Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104694Date: Fri Oct 15 14:12:22 2010 +0100 104695 104696 Cygwin/X: Turn off the bad-function-cast warning for XWin code 104697 104698 Turn off the bad-function-cast warning for the XWin code, it's near impossible 104699 to write code which uses Win32 API calls that doesn't trigger this warning 104700 104701 For example, SendMessage(WM_SETICON) returns an LRESULT, but we are supposed to 104702 know this is safe to cast this result to a HICON, which gcc considers a non-matching 104703 type. 104704 104705 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104706 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104707 104708commit 55e528b6cc6451dbb1e65baec199d4df0441e86d 104709Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104710Date: Thu Mar 25 23:39:37 2010 +0000 104711 104712 Cygwin/X: Remove execute permission from source files 104713 104714 Remove execute permission from source files 104715 104716 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104717 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104718 104719commit b2b685e2693ff5f089006f5b3882d6b0c0853862 104720Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104721Date: Thu Mar 25 23:00:48 2010 +0000 104722 104723 Cygwin/X: unifdef XFree86Server 104724 104725 Remove XFree86Server define, which was always on anyhow, and the 104726 code which was guarded by !XFree86Server 104727 104728 This completes the process of removal started in 2006 :-) 104729 104730 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104731 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104732 104733commit 9ed70f15eff6a85222b46db1fe7af25154314464 104734Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104735Date: Mon Nov 2 17:37:14 2009 +0000 104736 104737 Cygwin/X: Use normal apostrophe in XWin DDX help text 104738 104739 Use U+0027 APOSTROPHE in XWin DDX help text, rather than 104740 U+2019 RIGHT SINGLE QUOTATION MARK 104741 104742 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104743 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104744 104745commit 9f8f056a21b9480fd9fd43a22c9d6f45a2c1ef87 104746Author: Jon TURNEY <jon.turney@dronecode.org.uk> 104747Date: Tue Mar 23 16:34:26 2010 +0000 104748 104749 Cygwin/X: Cleanup some VENDOR_STRING/VENDOR_CONTACT cruft 104750 104751 Cleanup some VENDOR_STRING/VENDOR_CONTACT cruft 104752 104753 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 104754 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 104755 104756commit 5aff712a8d2eb9f965ecbb93216cc0bcdc327ae6 104757Merge: 5b98c6267 e354ccac3 104758Author: Keith Packard <keithp@keithp.com> 104759Date: Mon Oct 18 17:16:23 2010 -0700 104760 104761 Merge remote branch 'whot/for-keith' 104762 104763commit 5b98c6267f575ec4ff498ecb651e34dbff030b83 104764Author: Adam Jackson <ajax@redhat.com> 104765Date: Mon Oct 18 16:01:40 2010 -0400 104766 104767 randr: Remove mirandr 104768 104769 This isn't used anywhere, not least because it's completely 104770 nonfunctional. 104771 104772 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 104773 Signed-off-by: Adam Jackson <ajax@redhat.com> 104774 104775commit 7bb653bedceb6180a0361ead1c612839e776ce98 104776Author: Olivier Fourdan <ofourdan@redhat.com> 104777Date: Mon Oct 18 15:59:35 2010 -0400 104778 104779 modes: improve aspect ratio match for classic drivers 104780 104781 After we infer the aspect ratio for the screen, we pick the largest 104782 mode matching that aspect ratio from the best mode pool available. 104783 We then clamp virtual size to that mode, and run the resulting mode 104784 list through the driver's ValidMode hook. In doing so we might filter 104785 away our initial guess. If this happens we shrink the default mode 104786 to the next largest mode from _any_ mode pool. This is usually wrong, 104787 and we should instead pick the next aspect-matched mode from the best 104788 available mode pool (as always, user then driver then default). 104789 104790 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 104791 Signed-off-by: Adam Jackson <ajax@redhat.com> 104792 104793commit d6d90a4fccef086e7c7934adea901b2cd9a595bf 104794Author: Adam Jackson <ajax@redhat.com> 104795Date: Mon Oct 18 15:53:43 2010 -0400 104796 104797 composite: Move the backfill comment to a more appropriate indent level 104798 104799 Reviewed-by: Keith Packard <keithp@keithp.com> 104800 Signed-off-by: Adam Jackson <ajax@redhat.com> 104801 104802commit db8840600e8e21356241eb87395031388d9b54d2 104803Author: Adam Jackson <ajax@redhat.com> 104804Date: Mon Oct 18 15:52:52 2010 -0400 104805 104806 composite: Don't backfill non-MapWindow allocations 104807 104808 Reviewed-by: Keith Packard <keithp@keithp.com> 104809 Signed-off-by: Adam Jackson <ajax@redhat.com> 104810 104811commit 6dd775f57d2f94f0ddaee324aeec33b9b66ed5bc 104812Author: Adam Jackson <ajax@redhat.com> 104813Date: Mon Oct 18 15:50:37 2010 -0400 104814 104815 composite: Don't backfill non-bg-None windows 104816 104817 If there's a defined background then backfilling is a waste of effort, 104818 since exposure processing will paint that in for us. But note that we 104819 have to backfill if any children are bg=None to preserve semantics with 104820 non-composited servers. 104821 104822 Reviewed-by: Keith Packard <keithp@keithp.com> 104823 Signed-off-by: Adam Jackson <ajax@redhat.com> 104824 104825commit d2064fbb687839c297a851a5d85f32dfbbe4a0d5 104826Author: Adam Jackson <ajax@redhat.com> 104827Date: Mon Oct 18 15:42:54 2010 -0400 104828 104829 ddc: Don't probe for DDC/CI or EEPROM 104830 104831 For whatever reason, some (broken) monitors will crash if you do this. 104832 We're not actually using this information for anything, so let's just 104833 not do it. 104834 104835 Originally reported as http://bugzilla.redhat.com/620333 104836 104837 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 104838 Signed-off-by: Adam Jackson <ajax@redhat.com> 104839 104840commit 47c91dca8d8eecb429123e8370302831bcd57938 104841Author: Adam Jackson <ajax@redhat.com> 104842Date: Mon Oct 18 15:40:28 2010 -0400 104843 104844 xfree86: Drop linux libc5 support from the SIGIO code 104845 104846 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 104847 Reviewed-by: Matt Turner <mattst88@gmail.com> 104848 Signed-off-by: Adam Jackson <ajax@redhat.com> 104849 104850commit 788bfbf18ac19923604b0d676933fdc0d78526a1 104851Author: Adam Jackson <ajax@redhat.com> 104852Date: Thu Oct 14 15:10:43 2010 -0400 104853 104854 dri1: Remove "buffers" from the config logic 104855 104856 This was only ever used from the glint driver, which has since lost its 104857 DRI support. 104858 104859 Reviewed-by: Matt Turner <mattst88@gmail.com> 104860 Signed-off-by: Adam Jackson <ajax@redhat.com> 104861 104862commit a77458486a7b9d78dcd284e2bcc96c1fda0e862a 104863Author: Adam Jackson <ajax@redhat.com> 104864Date: Thu Oct 14 14:00:53 2010 -0400 104865 104866 xfree86: Remove %M expansion from config parser 104867 104868 This was to distinguish XFree86 3.x files from XFree86 4.x files. It 104869 never really made sense to be looking for xorg.conf-4. 104870 104871 Reviewed-by: Matt Turner <mattst88@gmail.com> 104872 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 104873 Reviewed-by: Jesse Adkins <jesserayadkins@gmail.com> 104874 Signed-off-by: Adam Jackson <ajax@redhat.com> 104875 104876commit f8ec71603c796f3b272a0592cbb934c2a6e8d633 104877Author: Adam Jackson <ajax@redhat.com> 104878Date: Thu Oct 14 13:45:31 2010 -0400 104879 104880 xfree86: Remove an open-coded strtoul() 104881 104882 Reviewed-by: Matt Turner <mattst88@gmail.com> 104883 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 104884 Signed-off-by: Adam Jackson <ajax@redhat.com> 104885 104886commit 52577ae8eeda64601db8dd425027cf8e4271b873 104887Author: Adam Jackson <ajax@redhat.com> 104888Date: Thu Oct 14 13:22:00 2010 -0400 104889 104890 config: Remove atimisc workaround 104891 104892 Reviewed-by: Matt Turner <mattst88@gmail.com> 104893 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 104894 Reviewed-by: Jesse Adkins <jesserayadkins@gmail.com> 104895 Signed-off-by: Adam Jackson <ajax@redhat.com> 104896 104897commit 8bb1983a5f2dbf72366de9fb793a46e35912ab61 104898Author: Adam Jackson <ajax@redhat.com> 104899Date: Mon Aug 2 15:22:18 2010 -0400 104900 104901 config: Don't look for .o drivers anymore 104902 104903 This hasn't worked since we switched to dlloader. 104904 104905 Reviewed-by: Jesse Adkins <jesserayadkins@gmail.com> 104906 Reviewed-by: Matt Turner <mattst88@gmail.com> 104907 Signed-off-by: Adam Jackson <ajax@redhat.com> 104908 104909commit e354ccac36a8ee3a23bdc845833c16a5646cc200 104910Author: Joe Shaw <joeshaw@litl.com> 104911Date: Thu Oct 14 15:09:20 2010 -0400 104912 104913 fix a sign problem with valuator data. 104914 104915 Without this patch, any negative valuator value is wrong when returned 104916 from XQueryDeviceState(). This is a regression from at least xserver 104917 1.4. 104918 104919 Valuator data is set in dix/getevents.c:set_valuators() by copying 104920 signed int values into an unsigned int field 104921 DeviceEvent.valuators.data. 104922 104923 That data is converted into a double with an implicit cast by 104924 assignment to axisVal[i] in Xi/exevents.c:UpdateDeviceState(). 104925 104926 That double is converted back to a signed int in 104927 queryst.c:ProcXQueryDeviceState(). If the original value in 104928 set_valuators() is negative, the double value will be > 2^31 and the 104929 conversion back to a signed int is undefined. (Although I 104930 consistently see the value -2^31.) 104931 104932 Fix this by changing the definition of DeviceEvent.valuators.data from 104933 uint32_t to int32_t. 104934 104935 Signed-off-by: Joe Shaw <joeshaw@litl.com> 104936 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104937 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 104938 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104939 104940commit 424b856e8e19f35c24bfc0a9fced9464d2f17c90 104941Author: Peter Hutterer <peter.hutterer@who-t.net> 104942Date: Fri Oct 15 14:35:22 2010 +1000 104943 104944 dix: update comments for GetPointerEvents and friends 104945 104946 All these now generate InternalEvents, point this out. Remove XKB/XI 104947 references, that's just confusing. This comment referred to the old-style 104948 event generation code from server 1.4 to including 1.6 but is now just 104949 confusing to newcomers. 104950 104951 Remove comment about SwitchCoreKeyboard() for the same reason. 104952 104953 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104954 Reviewed-by: Chase Douglas <chase.douglas@canonical.com> 104955 104956commit e2ada55de3e7ec8724b98c4e4adfec741d415499 104957Author: Adam Jackson <ajax@redhat.com> 104958Date: Mon Sep 27 13:58:40 2010 -0400 104959 104960 os: Clean up various xtrans bits that we've never supported 104961 104962 Or at least, not supported since xserver 1.0. 104963 104964 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 104965 Signed-off-by: Adam Jackson <ajax@redhat.com> 104966 104967commit eaf0b6a4d83d49930d21d5191f335fcac962632e 104968Merge: 693e92d40 23229c7ce 104969Author: Peter Hutterer <peter.hutterer@who-t.net> 104970Date: Fri Oct 15 13:46:59 2010 +1000 104971 104972 Merge branch 'master' into input-api 104973 104974 Conflicts: 104975 config/udev.c 104976 hw/xfree86/common/xf86Helper.c 104977 hw/xfree86/common/xf86Module.h 104978 hw/xfree86/common/xf86Xinput.h 104979 hw/xfree86/os-support/linux/lnx_init.c 104980 104981 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104982 104983commit 3c28a29e132d6f73c36d4b64818d112b1c6e9a40 104984Author: Peter Hutterer <peter.hutterer@who-t.net> 104985Date: Wed Oct 13 15:53:59 2010 +1000 104986 104987 xfree86: rename parameter names to xf86ScaleAxis. 104988 104989 Maybe it's just me but every time I look at it I get confused again and need 104990 to work it out from scratch. Rename the parameters to something 104991 self-explanatory, to/from and min/max. 104992 104993 No functional change. 104994 104995 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 104996 Reviewed-by: Julien Cristau <jcristau@debian.org> 104997 104998commit 23229c7ce5970516b97a1cc12e65257ff65ab9a7 104999Author: Jeremy Huddleston <jeremyhu@apple.com> 105000Date: Wed Oct 13 16:28:14 2010 -0700 105001 105002 test: Fix make distcheck when not building unit tests 105003 105004 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 105005 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 105006 Signed-off-by: Keith Packard <keithp@keithp.com> 105007 105008commit ff82f724e4ed24d73f15fa9c35a1d0c2c7d8cef7 105009Author: Colin Harrison <colin.harrison@virgin.net> 105010Date: Wed Oct 13 10:20:06 2010 -0700 105011 105012 Comparison between pointer and integer in latest dix/dispatch.c 105013 105014 Signed-off-by: Keith Packard <keithp@keithp.com> 105015 105016commit da4eebe58ec81e2297cc1773af1fb622fb392bbd 105017Author: Jon TURNEY <jon.turney@dronecode.org.uk> 105018Date: Wed Oct 13 17:34:31 2010 +0100 105019 105020 event_size is currently never assigned to in mieqProcessInputEvents(). 105021 105022 event_size is never assigned to in mieqProcessInputEvents(), so realloc() 105023 is always called. This is benign, but I'm guessing not intended. 105024 105025 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 105026 Reviewed-by: Keith Packard <keithp@keithp.com> 105027 Signed-off-by: Keith Packard <keithp@keithp.com> 105028 105029commit 6274dca9d984ad3c553b4901edc3151e770e6c40 105030Author: Adam Jackson <ajax@redhat.com> 105031Date: Mon Oct 11 16:37:16 2010 -0400 105032 105033 dix: optimize CallCallbacks 105034 105035 Move the basic sanity checking to an inline wrapper, which avoids the 105036 function call overhead if the callback list is empty. On an XACEful 105037 server on a 2.4GHz Core 2 Duo: 105038 105039 1 2 Operation 105040 -------- ----------------- ----------------- 105041 20000000.0 25100000.0 ( 1.25) X protocol NoOperation 105042 105043 Signed-off-by: Adam Jackson <ajax@redhat.com> 105044 Reviewed-by: Keith Packard <keithp@keithp.com> 105045 Signed-off-by: Keith Packard <keithp@keithp.com> 105046 105047commit e418cd332c1f458d028df3fdf684011109d0c183 105048Author: Adam Jackson <ajax@redhat.com> 105049Date: Mon Oct 11 10:49:57 2010 -0400 105050 105051 xwin: Remove some undue paranoia about UNDRAWABLE_WINDOW 105052 105053 mi will throw away rendering to input-only windows. 105054 105055 Signed-off-by: Adam Jackson <ajax@redhat.com> 105056 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 105057 Signed-off-by: Keith Packard <keithp@keithp.com> 105058 105059commit febf3e7ead8dbb355622c896ff1744b2ca52a49a 105060Author: Adam Jackson <ajax@redhat.com> 105061Date: Mon Oct 11 10:49:56 2010 -0400 105062 105063 dix: Remove the memory of the multibuffer extension 105064 105065 Drop DRAWABLE_BUFFER and related checks, mbuf was the only thing that 105066 used them and it was killed in 0ba82562. 105067 105068 Signed-off-by: Adam Jackson <ajax@redhat.com> 105069 Reviewed-by: Keith Packard <keithp@keithp.com> 105070 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 105071 Signed-off-by: Keith Packard <keithp@keithp.com> 105072 105073commit 1333e101b4f6ecca72568b7f462884f8b102a5c2 105074Author: Alan Coopersmith <alan.coopersmith@oracle.com> 105075Date: Thu Oct 7 18:17:02 2010 -0700 105076 105077 Remove leftover fbSaveAreas & fbRestoreAreas defines from wfbrename.h 105078 105079 The fb functions they try to rename were deleted in 2007 by 105080 commit ae7f71a8b3d6756161e55d998d6eec37d2695c98 105081 105082 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105083 Reviewed-by: Matt Turner <mattst88@gmail.com> 105084 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 105085 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 105086 Signed-off-by: Keith Packard <keithp@keithp.com> 105087 105088commit 8efa3623f32939ef525926b57fd533f4c01ec4b3 105089Author: Jamey Sharp <jamey@minilop.net> 105090Date: Wed Aug 4 13:48:33 2010 -0700 105091 105092 Use GetCurrentRootWindow or equivalent instead of spriteTrace[0]. 105093 105094 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105095 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 105096 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 105097 105098commit 32de0c19075b886c2fc03f6ad73d9c15df1b2309 105099Author: Jamey Sharp <jamey@minilop.net> 105100Date: Wed Aug 4 11:44:05 2010 -0700 105101 105102 Move replay-window check from ComputeFreezes to CheckDeviceGrabs. 105103 105104 This just simplifies ComputeFreezes, eliminating some duplicated code 105105 and a goto. 105106 105107 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105108 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 105109 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 105110 105111commit 1c1811ecafc144151630b3588394e2c345c53aa4 105112Author: Jamey Sharp <jamey@minilop.net> 105113Date: Tue Aug 3 19:49:59 2010 -0700 105114 105115 CheckDeviceGrabs: Delete redundant pWin->optional test. 105116 105117 CheckPassiveGrabsOnWindow returns FALSE if pWin->optional is NULL, 105118 because wPassiveGrabs uses wUseDefault, so don't bother checking at the 105119 caller. 105120 105121 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105122 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 105123 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 105124 105125commit c360b3eb1deeac51913e7b1286b2767711c541fc 105126Author: Jamey Sharp <jamey@minilop.net> 105127Date: Tue Aug 3 19:40:26 2010 -0700 105128 105129 Let CheckPassiveGrabsOnWindow callers check if device is already grabbed. 105130 105131 CheckDeviceGrabs checked all the ancestors of the window containing this 105132 device's pointer even if no new grabs could possibly apply due to the 105133 device already being grabbed. 105134 105135 ActivateFocusInGrab and ActivateEnterGrab already checked whether they 105136 should break an existing grab, and then set up an event that was 105137 completely ignored if they didn't actually break the grab. 105138 105139 In both cases, just do what we would have done eventually anyway--return 105140 FALSE from CheckPassiveGrabsOnWindow's caller--but do it sooner. 105141 105142 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105143 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 105144 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 105145 105146commit 853d7ebfa3e2d281d92890a39010ff5787a00ffd 105147Author: Adam Jackson <ajax@redhat.com> 105148Date: Tue Oct 5 11:24:28 2010 -0400 105149 105150 vbe: Fix copying the mode info block 105151 105152 Just use memcpy, seriously. 105153 105154 Reviewed-by: Rémi Cardona <remi@gentoo.org> 105155 Signed-off-by: Adam Jackson <ajax@redhat.com> 105156 Signed-off-by: Keith Packard <keithp@keithp.com> 105157 105158commit b2f9ce201cc2a14d5d6ad055b46c9317b040ec2e 105159Author: Adam Jackson <ajax@redhat.com> 105160Date: Tue Jul 27 17:34:36 2010 -0400 105161 105162 edid: Fix the HDTV sync pulse adjustment 105163 105164 Simple typo, should have been adjusting the horizontal timings 105165 consistently since we're not trying to mangle vertical at all. 105166 105167 Signed-off-by: Adam Jackson <ajax@redhat.com> 105168 Reviewed-by: Patrick E. Kane <pekane52@gmail.com> 105169 Signed-off-by: Keith Packard <keithp@keithp.com> 105170 105171commit 9a893fe52f1bfd415e9ffd9d2f5aff5c4da7a756 105172Author: Chris Wilson <chris@chris-wilson.co.uk> 105173Date: Sun Oct 3 16:56:03 2010 +0100 105174 105175 xfree86: Do not call dlclose(NULL) [regression after ab7f057] 105176 105177 During unwind following an error when attempting to a load a module, we 105178 attempt to call dlclose on a potentially NULL handle. This is a 105179 side-effect of removing the abstraction layer in ab7f057. 105180 105181 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 105182 Cc: Adam Jackson <ajax@redhat.com> 105183 Reviewed-by: Adam Jackson <ajax@redhat.com> 105184 Signed-off-by: Keith Packard <keithp@keithp.com> 105185 105186commit 533faf7f43cb4b5155a74d78d6144b702e227efa 105187Author: Keith Packard <keithp@keithp.com> 105188Date: Fri Oct 1 10:57:48 2010 -0700 105189 105190 Bump version to 1.9.99.1 (1.10 development series) 105191 105192 Make the 1.10 development series distinguishable from the 1.9 branch. 105193 105194 Signed-off-by: Keith Packard <keithp@keithp.com> 105195 105196commit a769f4c22a9cfb5ba248c924a66c31ec966bd8a0 105197Author: Adrian Bunk <bunk@stusta.de> 105198Date: Thu Sep 30 13:18:20 2010 +0300 105199 105200 remove dolt 105201 105202 With libtool 2.2.10 the difference in build time is so small 105203 (< 5% with -j8 builds) that it isn't worth having this hack. 105204 105205 Signed-off-by: Adrian Bunk <bunk@stusta.de> 105206 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 105207 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 105208 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 105209 Signed-off-by: Keith Packard <keithp@keithp.com> 105210 105211commit c7e4222c9a27094ce4fc2831ac92acbb7b21fb1a 105212Author: Tobias Droste <tdroste@gmx.de> 105213Date: Wed Sep 29 22:51:48 2010 +0200 105214 105215 randr: set error numbers of resource types in RRExtenstionInit() (V2) 105216 105217 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=30367 105218 105219 Currently the ddx calls xf86RandR12Init() 105220 (-> RRScreenInit() 105221 -> RRInit() 105222 -> RRModeInit() 105223 -> RRCrtcInit() 105224 -> RROutputInit()) 105225 before RRExtensionInit() is called. This causes RRErrorBase 105226 being 0 while setting resource type error values (resource types: 105227 RROutput, RRMode and RRCrtc). The fix moves the setting of error 105228 values to own functions which are called in RRExtensionInit() 105229 to get the right RRErrorBase. 105230 105231 V2: With header file 105232 105233 Signed-off-by: Tobias Droste <tdroste@gmx.de> 105234 Reviewed-by: Keith Packard <keithp@keithp.com> 105235 Signed-off-by: Keith Packard <keithp@keithp.com> 105236 105237commit 72a9c686d7ecaa68d27687b316dbba92902809b4 105238Author: Jon TURNEY <jon.turney@dronecode.org.uk> 105239Date: Wed Sep 29 17:05:12 2010 +0100 105240 105241 Cygwin/X: Fix compilation after "delete pervasively use of DISPATCH_PROC" 105242 105243 commit cbd4d5dbb70db62ba1cb79c7b904e6fa11f62d7e 105244 removes the static declarations of ProcWindowsWMDispatch and 105245 SProcWindowsWMDispatch which precede their first use in 105246 winWindowsWMExtensionInit() 105247 105248 Move winWindowsWMExtensionInit() to after the definition of those 105249 two functions to fix compilation. 105250 105251 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 105252 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105253 Signed-off-by: Keith Packard <keithp@keithp.com> 105254 105255commit 1a9022d3289ff74043112808cc8aa315fd9c40e8 105256Author: Aaron Plattner <aplattner@nvidia.com> 105257Date: Tue Sep 28 13:59:19 2010 -0700 105258 105259 Bump video driver ABI version to 9.0 105260 105261 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 105262 Reviewed-by: Keith Packard <keithp@keithp.com> 105263 Signed-off-by: Keith Packard <keithp@keithp.com> 105264 105265commit c45bea0c044ad37bedb42209f7e6ea8b587999f0 105266Author: Jeremy Huddleston <jeremyhu@apple.com> 105267Date: Tue Aug 3 12:51:51 2010 -0700 105268 105269 XQuartz: RandR: Refactor legacy mode-switching to be better integrated with RandR 105270 105271 Adds three new functions 105272 void QuartzRandRSetFakeRootless (void); 105273 void QuartzRandRSetFakeFullscreen (void); 105274 void QuartzRandRToggleFullscreen (void); 105275 105276 The first two are identical to requesting the fake modes from a RandR client 105277 The third responds to cmd-alt-a to leave fullscreen or RandR. 105278 105279 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 105280 105281commit 229323a19b06f80d9b03f487e598b933b9b31d87 105282Author: Jeremy Huddleston <jeremyhu@apple.com> 105283Date: Mon Jul 26 23:34:49 2010 -0700 105284 105285 XQuartz: RandR: Don't change the rootless preference when changing RandR mode 105286 105287 Also renames a bunch of other variables for better consistency. 105288 105289 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 105290 105291commit 721edc69c3597578d3e7f711769ff0195af6068a 105292Author: Jeremy Huddleston <jeremyhu@apple.com> 105293Date: Mon Jul 26 11:45:59 2010 -0700 105294 105295 XQuartz: RandR: Better handle switching betwen RandR modes that share CG modes 105296 105297 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 105298 105299commit 026a47e21226b64eb9a7624ada3d74c3fe5fa763 105300Author: Jeremy Huddleston <jeremyhu@apple.com> 105301Date: Mon Jul 26 01:17:12 2010 -0700 105302 105303 XQuartz: RandR: Respond better to resolution changes made outside X 105304 105305 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 105306 105307commit 59cd93f48a3a0a34781d4a94a3cd05a575e2d4ba 105308Author: Jeremy Huddleston <jeremyhu@apple.com> 105309Date: Sun Jul 25 22:25:49 2010 -0700 105310 105311 XQuartz: RandR: Use deprecated CG APIs only on Leopard and earlier 105312 105313 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 105314 105315commit 629c3f473b7415069a87ef4717ec3d701978dc48 105316Author: Jan Hauffa <hauffa@in.tum.de> 105317Date: Fri Jul 16 17:56:44 2010 +0200 105318 105319 XQuartz: RandR: Remove FAKE_RANDR code. 105320 105321 Signed-off-by: Jan Hauffa <hauffa@in.tum.de> 105322 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 105323 105324commit 97b5f5306437bfd13390485fc7a58a363c261ec9 105325Author: Jan Hauffa <hauffa@in.tum.de> 105326Date: Fri Jul 16 17:54:55 2010 +0200 105327 105328 XQuartz: RandR: Toggle rootless mode on XRandR mode switch. 105329 105330 Report a fake screen mode that corresponds to the screen mode at startup of the 105331 server excluding the height of the menu bar. If a client requests this mode, 105332 rootless mode is enabled. In all other modes, the root window is shown. 105333 105334 Signed-off-by: Jan Hauffa <hauffa@in.tum.de> 105335 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 105336 105337commit 2d411472c2915614bdeb8fdc15d19dae4621444d 105338Author: Jan Hauffa <hauffa@in.tum.de> 105339Date: Fri Jul 16 17:18:03 2010 +0200 105340 105341 XQuartz: RandR: Implement basic RandR functionality. 105342 105343 Querying and changing of resolution and refresh rate is supported, 105344 rotation is not implemented yet. 105345 105346 Signed-off-by: Jan Hauffa <hauffa@in.tum.de> 105347 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 105348 105349commit f4926709485c0605f0b1171c84c803b68eebfc38 105350Author: Jeremy Huddleston <jeremyhu@apple.com> 105351Date: Tue Sep 28 09:48:41 2010 -0700 105352 105353 XQuartz: Move use of static dispatch procs after declaration 105354 105355 This fixes fallout from cbd4d5dbb70db62ba1cb79c7b904e6fa11f62d7e 105356 105357 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 105358 105359commit c36ca7b1862ecd86df089f4e1933379a113928ea 105360Author: Gaetan Nadon <memsize@videotron.ca> 105361Date: Sat Sep 25 10:05:10 2010 -0400 105362 105363 config: remove --with-dri-driver-path option, use dri.pc #29740 105364 105365 The trusted location of the dri driver directory can be obtained 105366 from the mesa dri.pc pkg-config file. 105367 105368 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 105369 Acked-by: Matt Turner <mattst88@gmail.com> 105370 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 105371 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 105372 Signed-off-by: Keith Packard <keithp@keithp.com> 105373 105374commit 5532687a929426c4b1c4667f4591ed362f097c9b 105375Merge: 9df4fb0ad 682a3ee60 105376Author: Keith Packard <keithp@keithp.com> 105377Date: Tue Sep 28 08:15:07 2010 -0700 105378 105379 Merge remote branch 'ajax/less-loader' 105380 105381commit 9df4fb0adf66fb82f0c007897d79af0f54a6dad9 105382Author: Adam Jackson <ajax@redhat.com> 105383Date: Mon Sep 27 17:06:46 2010 -0400 105384 105385 xfree86: Add 18bpp support 105386 105387 Signed-off-by: Adam Jackson <ajax@redhat.com> 105388 Reviewed-by: Keith Packard <keithp@keithp.com> 105389 Signed-off-by: Keith Packard <keithp@keithp.com> 105390 105391commit 5c6a2f93ebc16a78093782b442306de23ae94e78 105392Author: Chris Wilson <chris@chris-wilson.co.uk> 105393Date: Mon Sep 27 19:39:23 2010 +0100 105394 105395 xfree86: Kill pixmapPrivate with a vengeance (v2) 105396 105397 ScrnInfo->pixmapPrivate only existed in order to catch invalid access to 105398 the framebuffer by making the backing data NULL across the VT switch. 105399 This was causing more confusion in the higher layers during mode setting 105400 without any real benefit, so remove it. 105401 105402 v2: Kill ShadowModifyPixmapHeader() as well. 105403 105404 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 105405 Cc: Julien Cristau <jcristau@debian.org> 105406 Cc: Andrew Guertin <lists@dolphinling.net> 105407 Reviewed-by: Keith Packard <keithp@keithp.com> 105408 Signed-off-by: Keith Packard <keithp@keithp.com> 105409 105410commit fdabcec57ae0fdc9910060609bb0848552c1db4d 105411Author: Chris Wilson <chris@chris-wilson.co.uk> 105412Date: Mon Sep 27 17:30:13 2010 +0100 105413 105414 modes: Beware the driver switching root pixmaps 105415 105416 Program received signal SIGSEGV, Segmentation fault. 105417 0x080d4a2d in xf86RandR12ScreenSetSize (pScreen=0x8dca3a0, width=800, 105418 height=600, mmWidth=210, mmHeight=157) 105419 at ../../../../hw/xfree86/modes/xf86RandR12.c:731 105420 731 ../../../../hw/xfree86/modes/xf86RandR12.c: No such file or directory. 105421 in ../../../../hw/xfree86/modes/xf86RandR12.c 105422 (gdb) bt full 105423 height=600, mmWidth=210, mmHeight=157) 105424 at ../../../../hw/xfree86/modes/xf86RandR12.c:731 105425 randrp = 0x8dcae68 105426 pScrn = 0x8dbeb28 105427 config = <value optimized out> 105428 pRoot = 0x8e08e30 105429 pScrnPix = 0xb6d12008 105430 ret = 1 105431 c = <value optimized out> 105432 mmWidth=210, mmHeight=157) at ../../randr/rrscreen.c:185 105433 No locals. 105434 at ../../randr/rrscreen.c:307 105435 pWin = 0x8e08e30 105436 pScreen = 0x8dca3a0 105437 i = <value optimized out> 105438 rc = 0 105439 ../../randr/randr.c:485 105440 105441 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 105442 Cc: Julien Cristau <jcristau@debian.org> 105443 Tested-by: Julien Cristau <jcristau@debian.org> 105444 Cc: Keith Packard <keithp@keithp.com> 105445 Reviewed-by: Keith Packard <keithp@keithp.com> 105446 Signed-off-by: Keith Packard <keithp@keithp.com> 105447 105448commit 2f0abd7d690c3bcce2a80c093a0880f43e56d8cb 105449Author: Kristian Høgsberg <krh@bitplanet.net> 105450Date: Thu Sep 23 09:04:11 2010 -0400 105451 105452 glx: Fix use after free in DrawableGone 105453 105454 (note that nearly identical patches were proposed by Kristian and Chris) 105455 105456 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 105457 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 105458 Reported-by: Julien Cristau <jcristau@debian.org> 105459 Reviewed-by: Adam Jackson <ajax@redhat.com> 105460 Reviewed-by: Jamey Sharp <jamey@minilop.net> 105461 Tested-by: Chris Wilson <chris@chris-wilson.co.uk> 105462 Signed-off-by: Keith Packard <keithp@keithp.com> 105463 105464commit e416965bfd08dfae5d2b7932b5118efac20546ad 105465Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105466Date: Sun Jul 18 18:56:02 2010 +0300 105467 105468 dix: remove proc tables initialization 105469 105470 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105471 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105472 105473commit cbd4d5dbb70db62ba1cb79c7b904e6fa11f62d7e 105474Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105475Date: Fri Jul 16 20:19:50 2010 +0300 105476 105477 xserver: delete pervasively use of DISPATCH_PROC 105478 105479 Some functions had to be moved around due some missing static definitions. 105480 Another minor clean up like inexistent function declarations and etc were made 105481 also. 105482 105483 Part of this patch was cooked using: 105484 sed -i -e '/static DISPATCH_PROC*.*;/d' `git ls-files` 105485 105486 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105487 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105488 105489commit 58bd317e29f4abf7f950891339d2a6a78ddf7903 105490Author: Macpaul Lin <macpaul@andestech.com> 105491Date: Tue Sep 21 15:13:34 2010 +0800 105492 105493 xfree86: nds32: add nds32 support for compiler related mmio codes 105494 105495 Add nds32 support for compiler related mmio codes. 105496 It includes byte-swap or non-swap operations. 105497 105498 Signed-off-by: Macpaul Lin <macpaul@andestech.com> 105499 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105500 Signed-off-by: Keith Packard <keithp@keithp.com> 105501 105502commit 28e6de66b42062a885ba38416387f2dbc15707fd 105503Author: Macpaul Lin <macpaul@andestech.com> 105504Date: Tue Sep 21 15:13:33 2010 +0800 105505 105506 xfree86: nds32: add nds32 support for compiler specific codes 105507 105508 Add nds32 definitions and related assembly codes to compiler header files. 105509 105510 Signed-off-by: Macpaul Lin <macpaul@andestech.com> 105511 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105512 Signed-off-by: Keith Packard <keithp@keithp.com> 105513 105514commit 2b24b2bd859a4b4b8ccaf3b7d1b529dcda2d3b94 105515Author: Macpaul Lin <macpaul@andestech.com> 105516Date: Tue Sep 21 15:13:32 2010 +0800 105517 105518 xfree86: nds32: add nds32 definition for vgaHW support. 105519 105520 Add __nds32__ definitions for vgaHW support. 105521 105522 Signed-off-by: Macpaul Lin <macpaul@andestech.com> 105523 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105524 Signed-off-by: Keith Packard <keithp@keithp.com> 105525 105526commit 22376aa7aa44464636b12ee1d61b94e52ca02b0f 105527Author: Macpaul Lin <macpaul@andestech.com> 105528Date: Tue Sep 21 15:13:31 2010 +0800 105529 105530 xfree86: nds32: add nds32 definition for support linux video related io. 105531 105532 Add __nds32__ definitions for supporing correct io method for lnx_video.c 105533 105534 Signed-off-by: Macpaul Lin <macpaul@andestech.com> 105535 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105536 Signed-off-by: Keith Packard <keithp@keithp.com> 105537 105538commit d52799f67f590b0f91b521cd2bdf5a5a8a229c96 105539Author: Macpaul Lin <macpaul@andestech.com> 105540Date: Tue Sep 21 15:13:30 2010 +0800 105541 105542 xfree86: nds32: add nds32 related definitions into include headers. 105543 105544 Add MSB/LSB related definitions into include/servermd.h 105545 105546 Signed-off-by: Macpaul Lin <macpaul@andestech.com> 105547 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105548 Signed-off-by: Keith Packard <keithp@keithp.com> 105549 105550commit 98197d931b266674557b52b4a7099c6470114e55 105551Merge: 01ad3725a 402942cdb 105552Author: Keith Packard <keithp@keithp.com> 105553Date: Thu Sep 23 15:30:37 2010 -0700 105554 105555 Merge remote branch 'jamey/for-keith' 105556 105557commit 682a3ee60867da027fe1bcda0c8587dd3db9d58e 105558Author: Adam Jackson <ajax@redhat.com> 105559Date: Sat Sep 18 07:27:13 2010 -0400 105560 105561 xfree86: Remove useless module setup functions where appropriate 105562 105563 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105564 Reviewed-by: Julien Cristau <jcristau@debian.org> 105565 Signed-off-by: Adam Jackson <ajax@redhat.com> 105566 105567commit 6130170e7e9b64c611ee942ec3455dd1a185193d 105568Author: Adam Jackson <ajax@redhat.com> 105569Date: Sat Sep 18 06:41:35 2010 -0400 105570 105571 xfree86: Remove unused refcounting from input drivers 105572 105573 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105574 Reviewed-by: Julien Cristau <jcristau@debian.org> 105575 Signed-off-by: Adam Jackson <ajax@redhat.com> 105576 105577commit ab7f057ce9df4e905b12cebc1e587b9a7f200418 105578Author: Adam Jackson <ajax@redhat.com> 105579Date: Fri Sep 17 11:19:17 2010 -0400 105580 105581 loader: Remove a silly layer of reference counting 105582 105583 libdl will refcount objects for us just fine, thanks. 105584 105585 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105586 Reviewed-by: Julien Cristau <jcristau@debian.org> 105587 Signed-off-by: Adam Jackson <ajax@redhat.com> 105588 105589commit 0438002cd238011f33dbf176c555de5e380ba77e 105590Author: Adam Jackson <ajax@redhat.com> 105591Date: Fri Sep 17 09:13:04 2010 -0400 105592 105593 loader: include cleanup 105594 105595 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105596 Reviewed-by: Julien Cristau <jcristau@debian.org> 105597 Signed-off-by: Adam Jackson <ajax@redhat.com> 105598 105599commit 2f003fe49697f8343c052dff2b3752eace814ea7 105600Author: Adam Jackson <ajax@redhat.com> 105601Date: Fri Sep 17 09:02:13 2010 -0400 105602 105603 loader: Remove the handle field from LoaderOpen 105604 105605 This was always 0 from all the callers. 105606 105607 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105608 Reviewed-by: Julien Cristau <jcristau@debian.org> 105609 Signed-off-by: Adam Jackson <ajax@redhat.com> 105610 105611commit 09929da50503b559459f5b722c11647c47857fdb 105612Author: Adam Jackson <ajax@redhat.com> 105613Date: Fri Sep 17 08:16:51 2010 -0400 105614 105615 loader: Remove unused canonical name field 105616 105617 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105618 Reviewed-by: Julien Cristau <jcristau@debian.org> 105619 Signed-off-by: Adam Jackson <ajax@redhat.com> 105620 105621commit 3a26e7f459764d4aee71b2d7e25b113b729b94ac 105622Author: Adam Jackson <ajax@redhat.com> 105623Date: Fri Sep 17 07:55:48 2010 -0400 105624 105625 loader: Remove unused module serial number 105626 105627 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105628 Reviewed-by: Julien Cristau <jcristau@debian.org> 105629 Signed-off-by: Adam Jackson <ajax@redhat.com> 105630 105631commit 2a24a013bf3f554bb03c0f5af155d23dbb27b599 105632Author: Adam Jackson <ajax@redhat.com> 105633Date: Fri Sep 17 07:32:17 2010 -0400 105634 105635 loader: Merge dlloader directly into the loader 105636 105637 This lets us drop some double-tracking of loaded modules too. If your 105638 OS is too lame to have libdl, fix that first. 105639 105640 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105641 Reviewed-by: Julien Cristau <jcristau@debian.org> 105642 Signed-off-by: Adam Jackson <ajax@redhat.com> 105643 105644commit 693e92d4049ee76d074737480f3fc1bb236f4278 105645Author: Peter Hutterer <peter.hutterer@who-t.net> 105646Date: Fri Sep 10 09:54:33 2010 +1000 105647 105648 xfree86: add xf86OptionListDuplicate() 105649 105650 Does what it says on the box. 105651 105652 Some drivers need to duplicate option lists from the original device to 105653 ensure that devices created by the driver (driver-internal hotplugging) have 105654 the same list of options as the original device. 105655 105656 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 105657 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 105658 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105659 105660commit 402942cdbc518395a2943fd226b9f3071f24d39f 105661Author: Jamey Sharp <jamey@minilop.net> 105662Date: Fri Sep 17 02:11:44 2010 +0200 105663 105664 Introduce per-object per-screen privates. 105665 105666 This replaces dixCreatePrivateKey and the only uses, which were in 105667 midispcur. 105668 105669 Commit by Jamey Sharp and Josh Triplett. 105670 105671 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105672 Signed-off-by: Josh Triplett <josh@joshtriplett.org> 105673 Reviewed-by: Keith Packard <keithp@keithp.com> 105674 105675commit e4d4d6ddd52801cf0b0b253d9ba3bdabfa9a9d8d 105676Author: Jamey Sharp <jamey@minilop.net> 105677Date: Fri Sep 17 02:18:10 2010 +0200 105678 105679 Xserver need not be compatible with old versions of xserver. 105680 105681 Delete time-traveling multiple personality disorder from the server. 105682 105683 Gaetan notes: 105684 105685 There were a couple of drivers containing an unknown version of the 105686 modes/parser code. This was done in server 1.2 time frame because it 105687 was released without mode code. It was barely or not maintained 105688 afterwards. There are currently no video drivers with a copy of the 105689 modes code. 105690 105691 Most of these ifdefs were introduced in commit 105692 a8d760f567b19268329c4682495caa591f08a854, where Aaron wrote, 105693 105694 This change uses XORG_VERSION_CURRENT < 7.0 to mean "server newer 105695 than 1.2" since XORG_VERSION current went backwards at some point. 105696 105697 Alan explains that: 105698 105699 In Xorg 1.3, when we first released an Xorg server release decoupled 105700 from the katamari release schedule. (1.0 through 1.2 were released 105701 as part of X11R7.0 through 7.2, while 1.3 came out between X11R7.2 & 105702 7.3.) 105703 105704 Commit by Jamey Sharp and Josh Triplett. 105705 105706 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105707 Signed-off-by: Josh Triplett <josh@joshtriplett.org> 105708 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 105709 105710commit 01ad3725a8d0a8958720b81d19598a08c072ccee 105711Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105712Date: Tue Jun 1 20:37:54 2010 +0300 105713 105714 xfree86: delete useless "Primary device is not PCI" message 105715 105716 The primary device being PCI or not has no effect on the server working. This 105717 message is superfluous. 105718 105719 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105720 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105721 105722commit 7360235ecca25af29bb2d99a8e46d5b259ad1b1e 105723Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105724Date: Mon May 31 18:20:57 2010 +0300 105725 105726 xfree86: bus: assign PCI_SOURCES with the actual source files 105727 105728 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105729 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105730 105731commit 18b2d2e8bd056266410226d779752e7933847b6b 105732Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105733Date: Mon May 31 18:09:50 2010 +0300 105734 105735 xfree86: bus: move macros from common PCI header to private file 105736 105737 Only int10/helper_exec.c is using them. 105738 105739 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105740 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105741 105742commit 17a23f9ba8fa95f3e15afba7de22a12fe60bacfb 105743Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105744Date: Fri Sep 10 19:18:42 2010 +0300 105745 105746 xfree86: dri: remove unused PCI macro 105747 105748 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105749 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105750 105751commit fc3ab84de7b5692f0db2b282ab0ed8e5a61d1fce 105752Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105753Date: Fri Sep 10 19:16:30 2010 +0300 105754 105755 xfree86: configure: move buses references to their own location 105756 105757 This patch makes xf86Configure.c free of PCI and SBUS code, moving to a more 105758 meaningful location. 105759 105760 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105761 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105762 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 105763 105764commit 49b817501f97d55480063c0b62544b3af75b4b7c 105765Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105766Date: Tue Jun 1 16:16:53 2010 +0300 105767 105768 xfree86: bus: remove xfree86 and few other references from OS files 105769 105770 Conceptually, os-support should have only a basic set of OS helpers which 105771 wouldn't mix with any DDX common structure (e.g. windowing structures, etc) 105772 105773 This patch removes some xfree86-only and a few other unused references from 105774 os-support/bus. 105775 105776 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105777 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105778 105779commit d9c682a5e250677883c4d375ca84cf184ee19d77 105780Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105781Date: Sun May 30 21:35:23 2010 +0300 105782 105783 xfree86: bus: remove useless automake variable 105784 105785 Leftover that was missed in f7abe05b. 105786 105787 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105788 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 105789 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105790 105791commit a7efeda8a56a540542e056a426e89cfc053a4162 105792Author: Tiago Vignatti <tiago.vignatti@nokia.com> 105793Date: Sun May 30 19:12:31 2010 +0300 105794 105795 xfree86: use xf86PciIsolateDevice to get PCI config information 105796 105797 Make xf86IsolateDevice private on PCI common file. 105798 105799 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105800 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105801 105802commit c768cdda92696b636c10bb2df64167d5274b4b99 105803Merge: 5a725385f 207985363 105804Author: Keith Packard <keithp@keithp.com> 105805Date: Mon Sep 13 16:48:33 2010 -0700 105806 105807 Merge remote branch 'jamey/reviewed' 105808 105809commit 20798536326f94ac9c97b4ec7a51329f7c912991 105810Author: Jamey Sharp <jamey@minilop.net> 105811Date: Wed Jul 14 13:14:28 2010 -0700 105812 105813 rootless: Delete dead remnants of rootless acceleration. 105814 105815 Jeremy deleted rootlessAccelInit's implementation in 2008 in commit 105816 587c010a1cd733fded4d49dc339df0634bda8be6. Delete its prototype and the 105817 remaining commented-out call to it. 105818 105819 It still makes sense for the rootless GC ops to relax the planemask, but 105820 that's independent of the size of the operation, so quit checking the 105821 thresholds there. 105822 105823 FillBytes and CompositePixels are not called anywhere, so delete 105824 everything related to both. 105825 105826 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105827 Cc: Jeremy Huddleston <jeremyhu@apple.com> 105828 Cc: Jon TURNEY <jon.turney@dronecode.org.uk> 105829 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 105830 105831commit 58d1ec0fe84412c66f1ea1de15e3901c0b120164 105832Author: Jamey Sharp <jamey@minilop.net> 105833Date: Wed Jul 14 11:30:17 2010 -0700 105834 105835 rootless: ROOTLESS_GLOBAL_COORDS is always set, so unifdef it. 105836 105837 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105838 Cc: Jeremy Huddleston <jeremyhu@apple.com> 105839 Cc: Jon TURNEY <jon.turney@dronecode.org.uk> 105840 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 105841 105842commit 35ce892a705f9377a36f19414e5a25dcbe44a354 105843Author: Jamey Sharp <jamey@minilop.net> 105844Date: Wed Jul 14 11:19:41 2010 -0700 105845 105846 rootless: ROOTLESS_TRACK_DAMAGE is never set, so unifdef it. 105847 105848 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105849 Cc: Jeremy Huddleston <jeremyhu@apple.com> 105850 Cc: Jon TURNEY <jon.turney@dronecode.org.uk> 105851 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 105852 105853commit ccdf15689050a0f988721c91fab75861ec4d339c 105854Author: Jamey Sharp <jamey@minilop.net> 105855Date: Mon May 31 16:01:18 2010 -0700 105856 105857 miPolyGlyphBlt: Delete redundant ValidateGC. 105858 105859 The caller is required to have validated the GC for the drawable before 105860 calling any ops, including PolyGlyphBlt. 105861 105862 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105863 Reviewed-by: Keith Packard <keithp@keithp.com> 105864 105865commit 3fd963de27f3c4452c320e617ad33b91b0967737 105866Author: Jamey Sharp <jamey@minilop.net> 105867Date: Wed May 19 14:16:20 2010 -0700 105868 105869 Un-export CreateScratchGC now that nothing external uses it. 105870 105871 The server and drivers sometimes use GetScratchGC, but never 105872 CreateScratchGC. 105873 105874 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105875 Reviewed-by: Keith Packard <keithp@keithp.com> 105876 105877commit 0f48e362c5baee2bc281a4b88cfb68de24172d15 105878Author: Jamey Sharp <jamey@minilop.net> 105879Date: Wed May 19 14:03:19 2010 -0700 105880 105881 glxdriswrast: Use CreateGC, not CreateScratchGC. 105882 105883 Since this code was using CreateScratchGC and not GetScratchGC, 105884 FreeScratchGC would always call FreeGC, so just call it directly. 105885 105886 As long as the drawable provided to CreateGC has the same screen and 105887 depth as were passed to CreateScratchGC, these functions are basically 105888 identical. There are two differences: 105889 105890 - CreateGC gives you a non-null stipple. You probably don't care. 105891 105892 - CreateGC passes the gcid and client to XACE. There are several 105893 internal GCs allocated in the server, and they all pass 0 and 105894 serverClient. I expect XACE will never reject that combination. 105895 105896 The callers of createDrawable all verify that pDraw has the same screen 105897 that driScreen does. In short, this should have no behavioral change. 105898 105899 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105900 Cc: Kristian Høgsberg <krh@bitplanet.net> 105901 Reviewed-by: Keith Packard <keithp@keithp.com> 105902 105903commit af2a94e12fe4e426dfca461a0d14526193c01cef 105904Author: Jamey Sharp <jamey@minilop.net> 105905Date: Tue Jul 13 13:33:59 2010 -0700 105906 105907 VGA arbiter: No need for arbitration around CreateGC. 105908 105909 None of XAA, EXA, or UXA do any hardware access during CreateGC, so they 105910 don't need VGA arbitration. I haven't found any open source drivers that 105911 hook CreateGC, so they're safe. I'd be surprised if any driver directly 105912 hooks CreateGC and does hardware access from it and needs VGA 105913 arbitration. 105914 105915 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105916 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105917 105918commit 7a63b2db636f01694cb54726bbb04287718e286a 105919Author: Jamey Sharp <jamey@minilop.net> 105920Date: Sun Jul 11 20:36:34 2010 -0700 105921 105922 Delete Colormap->devPriv. It isn't used any more. 105923 105924 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105925 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105926 105927commit 2e8069dd36c2f5caf868937405fc03328004a85d 105928Author: Jamey Sharp <jamey@minilop.net> 105929Date: Sun Jul 11 20:35:43 2010 -0700 105930 105931 xnest: Replace use of Colormap->devPriv with standard devPrivates. 105932 105933 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105934 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 105935 105936commit 34d6492436c0f4a81822769b24d8c7816d0226ad 105937Author: Jamey Sharp <jamey@minilop.net> 105938Date: Sat Jul 10 06:55:58 2010 -0700 105939 105940 xquartz: Delete compile-time-constant DRIGCRec.driOps field. 105941 105942 The driOps field of this structure is never set to anything but 105943 &driGCOps, and this structure is not visible outside this source file. 105944 Just use the constant in the one place the field was used. 105945 105946 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105947 Cc: Jeremy Huddleston <jeremyhu@apple.com> 105948 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 105949 105950commit 2051e0f371de892b2e598b30c18dae96b58d68de 105951Author: Jamey Sharp <jamey@minilop.net> 105952Date: Thu Jul 8 02:38:13 2010 -0700 105953 105954 dixfonts: Deobfuscate GC ops calls. 105955 105956 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105957 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 105958 105959commit bbbf795e6b1ac048ea17164ba1089055cfe4306d 105960Author: Jamey Sharp <jamey@minilop.net> 105961Date: Sun May 23 17:15:43 2010 -0700 105962 105963 Delete redundant GC initializations. 105964 105965 When a GC is allocated, it is zeroed, including all storage requested 105966 with dixRegisterPrivateKey. So CreateGC hooks don't need to initialize 105967 anything to zero. 105968 105969 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105970 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 105971 105972commit 3288232f3d2a0239944bd856101c2d16c713ef40 105973Author: Jamey Sharp <jamey@minilop.net> 105974Date: Sun Jun 13 18:10:38 2010 -0700 105975 105976 Delete xaaWrapper. 105977 105978 This was part of "An experimental pseudocolor emulation layer. Not fully 105979 completed, currently only works for 16bpp." Only neomagic tried to use 105980 it, and that was neutered by the removal of the fbpseudocolor portion of 105981 that emulation layer; the rest is easily removed. 105982 105983 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105984 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 105985 105986commit f856dcdc3a7b635dd94d3dac387ddec952e4880b 105987Author: Jamey Sharp <jamey@minilop.net> 105988Date: Sat Jun 12 08:42:31 2010 -0700 105989 105990 Delete unused lastWinOrg field from GCs. 105991 105992 Signed-off-by: Jamey Sharp <jamey@minilop.net> 105993 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 105994 105995commit b2184024361717358e35a56b084c4a71cbc29234 105996Author: Jamey Sharp <jamey@minilop.net> 105997Date: Sat Jun 12 16:20:00 2010 -0700 105998 105999 Delete unused devPrivate field from GCFuncs and GCOps. 106000 106001 Signed-off-by: Jamey Sharp <jamey@minilop.net> 106002 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106003 106004commit 139aa6e74227f0fe52cd8dccf1d563664103aaba 106005Author: Jamey Sharp <jamey@minilop.net> 106006Date: Sun Jun 13 18:57:08 2010 -0700 106007 106008 damage: Delete NOTUSED block--it was never not NOTUSED. 106009 106010 Signed-off-by: Jamey Sharp <jamey@minilop.net> 106011 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106012 106013commit 5ff9e58bef668c1279c5e253795711a1e118aaa1 106014Author: Jamey Sharp <jamey@minilop.net> 106015Date: Sat Jun 12 08:51:04 2010 -0700 106016 106017 fb: Delete unused oneRect private field. 106018 106019 Signed-off-by: Jamey Sharp <jamey@minilop.net> 106020 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 106021 106022commit e99bbfa0a24ec038b88e81881133a676bf63580e 106023Author: Jamey Sharp <jamey@minilop.net> 106024Date: Tue May 25 13:50:11 2010 -0700 106025 106026 dmx: Delete '#undef Xmalloc' and friends. 106027 106028 Nothing in dmx uses these names any more, and it builds cleanly without 106029 trying to undef them. 106030 106031 Signed-off-by: Jamey Sharp <jamey@minilop.net> 106032 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 106033 106034commit b5217bfdca67ef2c5323838afcfcd89eba8899d5 106035Author: Jamey Sharp <jamey@minilop.net> 106036Date: Tue May 25 13:28:32 2010 -0700 106037 106038 dmx: __glXMalloc -> malloc, etc. 106039 106040 Kristian made equivalent edits to the reference GLX implementation in 106041 2006, with commit 2d2d38d17cc2558f8a41166a4a1578bc4c663c37. 106042 106043 Signed-off-by: Jamey Sharp <jamey@minilop.net> 106044 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 106045 106046commit a715de7f11afeda7798f0882148a94d4db2291e6 106047Author: Jamey Sharp <jamey@minilop.net> 106048Date: Tue May 25 13:10:05 2010 -0700 106049 106050 dmx: Delete unused GLX visual matching code. 106051 106052 Kristian deleted this API from the rest of the server in 2007, in commit 106053 7d74690536b64f7b8e8036507ab7790807349c50. 106054 106055 Signed-off-by: Jamey Sharp <jamey@minilop.net> 106056 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 106057 106058commit 5a725385fbb93553931ca0217e79271ab2a2c613 106059Author: Tiago Vignatti <tiago.vignatti@nokia.com> 106060Date: Mon Sep 13 12:33:53 2010 +0300 106061 106062 xfree86: vgaarb: fix decoding stub API 106063 106064 This fix a problem introduced in commit 65466652. It closes also: 106065 https://bugs.freedesktop.org/show_bug.cgi?id=30160 106066 106067 Reported-by: <xunx.fang@intel.com> 106068 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106069 Signed-off-by: Keith Packard <keithp@keithp.com> 106070 106071commit 0a3e6b0345a96aae278dd776c00b5a7c7ccf65af 106072Author: Jeremy Huddleston <jeremyhu@apple.com> 106073Date: Sat Sep 11 14:25:21 2010 -0700 106074 106075 XQuartz: Localization updates 106076 106077 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 106078 106079commit 4e09a69c5b1f3c8d6f2f92ec24aaab3a1119940c 106080Author: Jeremy Huddleston <jeremyhu@apple.com> 106081Date: Sun Aug 15 20:07:28 2010 -0700 106082 106083 rootless: Remove ROOTLESS_WORKAROUND 106084 106085 This was already removed for XWin (20701522be803fe47e921fcf059dadf64c7f287d) 106086 with no reported side effects. XQuartz seems to be behaving ok without it as 106087 well. While this possibly brings back bug #1168, we don't have any reproduction 106088 steps for that issue, and if it crops up again, we should fix it a real way 106089 rather than this hokey workaround which doesn't even work for COMPOSITE. 106090 106091 This effectively reverts the following two changes: 106092 b2135e589baeb2ea26da50b9167feaea23bcce3c 106093 d7fef52254126aa5897a5c58faeda1f61d5b13d8 106094 106095 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 106096 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 106097 106098commit 63b7d9448abbded1cb91e82a13319d5e4313ca0c 106099Author: Tiago Vignatti <tiago.vignatti@nokia.com> 106100Date: Sun May 30 18:44:06 2010 +0300 106101 106102 xfree86: fbdevhw: remove unnecessary debug code 106103 106104 It very likely no one will want to print all functions of this file for 106105 debugging purposes. If this is the case, then a mix of ctags + cpp + gdb 106106 can do the same job. 106107 106108 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106109 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 106110 106111commit 66105d8281f8a6442cf7fee5d10e8b8a34bef952 106112Author: Tiago Vignatti <tiago.vignatti@nokia.com> 106113Date: Sun May 30 18:16:32 2010 +0300 106114 106115 xfree86: fbdevhw: remove unused cmap fields from main structure 106116 106117 It was never used since first git revision and probably no one cares for it. 106118 106119 ABI break. 106120 106121 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106122 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 106123 106124commit 317e491e067e31ea7cf8cb7301e47de0c10f119e 106125Author: Tiago Vignatti <tiago.vignatti@nokia.com> 106126Date: Mon Jul 19 15:44:39 2010 +0300 106127 106128 xext: remove unused header 106129 106130 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106131 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 106132 106133commit 5098801b3a570d0bb3386d7d210f21a1cbf3bf12 106134Author: Tiago Vignatti <tiago.vignatti@nokia.com> 106135Date: Sat Jun 12 16:57:50 2010 +0300 106136 106137 mi: delete unused code 106138 106139 pBSReg is always NULL, so the statement after the conditional will never be 106140 reached. 106141 106142 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106143 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 106144 106145commit 7d8cabd027cfc55e9921307e5b3dd7052de8496f 106146Author: Tiago Vignatti <tiago.vignatti@nokia.com> 106147Date: Wed Aug 4 16:12:59 2010 +0300 106148 106149 os: simplify smart scheduler init process 106150 106151 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106152 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 106153 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 106154 106155commit d9c18c3b9badf38d499e3f53291bd23fda3c4340 106156Author: Tiago Vignatti <tiago.vignatti@nokia.com> 106157Date: Wed Aug 4 15:54:48 2010 +0300 106158 106159 os: remove useless smart scheduler macros 106160 106161 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106162 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 106163 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 106164 106165commit a9e6080dc94633a4f497535f619a14fd8b566645 106166Author: Tiago Vignatti <tiago.vignatti@nokia.com> 106167Date: Fri Sep 10 15:01:29 2010 +0300 106168 106169 os/xfree86: remove macro checking for POSIX symbols 106170 106171 We assume already that our X implementation is POSIX compliant anyway. So 106172 remove those redundant checking. 106173 106174 SA_SIGINFO is left there. 106175 106176 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106177 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 106178 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 106179 106180commit 648c8871c92727d7b6b16859f27f12266a06a16e 106181Author: Adam Jackson <ajax@redhat.com> 106182Date: Wed Sep 8 16:08:24 2010 -0400 106183 106184 composite: Convert compWindowUpdate to use TraverseTree 106185 106186 v2: Use != RedirectDrawNone, since we aren't called for manual windows. 106187 106188 Signed-off-by: Adam Jackson <ajax@redhat.com> 106189 Reviewed-by: Keith Packard <keithp@keithp.com> 106190 Signed-off-by: Keith Packard <keithp@keithp.com> 106191 106192commit bbffb85461eb63bfb1e01e2cb9674607b9221604 106193Merge: 9a1a4ccbe 71972c253 106194Author: Keith Packard <keithp@keithp.com> 106195Date: Fri Sep 10 11:55:34 2010 -0700 106196 106197 Merge remote branch 'whot/for-keith' 106198 106199commit 9a1a4ccbe026f34beca5db3b2243874906961b49 106200Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 106201Date: Mon Aug 30 16:37:05 2010 -0400 106202 106203 Add screens to the PRIVATE_XSELINUX set. 106204 106205 The SELinux extension does store a security label in the screen 106206 devPrivates. Fixes crash caused by overwriting another private. 106207 106208 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 106209 Reported-by: Justin Mattock <justinmattock@gmail.com> 106210 Signed-off-by: Keith Packard <keithp@keithp.com> 106211 106212commit 5b680f2c23fb3d35c721647d0334f63297589371 106213Author: Vignatti Tiago (Nokia-MS/Helsinki) <tiago.vignatti@nokia.com> 106214Date: Mon Aug 30 21:24:30 2010 +0300 106215 106216 os: add -sigstop option for Upstart (or equivalent) startup 106217 106218 This is very similar to the RunFromSmartParent (implicit) option, except 106219 we do not send the signal to our parent process, but our own process 106220 instead, and that signal is SIGSTOP, not SIGUSR1. 106221 106222 Upstart or a similar equivalent program will detect this, realize that 106223 we are ready to accept clients now, send us SIGCONT and move our job 106224 status from SPAWNED to RUNNING. 106225 106226 Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com> 106227 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 106228 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106229 Signed-off-by: Keith Packard <keithp@keithp.com> 106230 106231commit ca0d578d29f37b61e62556fff59732741885625b 106232Merge: cf88363db 08adf41f6 106233Author: Keith Packard <keithp@keithp.com> 106234Date: Fri Sep 10 11:50:27 2010 -0700 106235 106236 Merge remote branch 'mattst88/master' 106237 106238commit cf88363db0ebb42df7cc286b85d30d7898aea840 106239Author: Aaron Plattner <aplattner@nvidia.com> 106240Date: Fri Aug 27 10:20:29 2010 -0700 106241 106242 os: Return BadLength instead of disconnecting BigReq clients (#4565) 106243 106244 If a client sends a big request that's too big (i.e. bigger than 106245 maxBigRequestSize << 2 bytes), the server just disconnects it. This makes the 106246 client receive SIGPIPE the next time it tries to send something. 106247 106248 The X Test Suite sends requests that are too big when the test specifies the 106249 TOO_LONG test type. When the client receives SIGPIPE, XTS marks it as 106250 UNRESOLVED, which counts as a failure. 106251 106252 Instead, remember how long the request is supposed to be and then return that 106253 size. Dispatch() checks the length and sends BadLength to the client. Then, 106254 whenever oci->ignoreBytes is nonzero, ignore the data read instead of trying to 106255 process it as a request. 106256 106257 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 106258 Reviewed-by: Keith Packard <keithp@keithp.com> 106259 Signed-off-by: Keith Packard <keithp@keithp.com> 106260 106261commit fa22f97af65efc2a147e9be0b7f288848965ff60 106262Merge: 6edbdb7f9 654666529 106263Author: Keith Packard <keithp@keithp.com> 106264Date: Fri Sep 10 11:47:41 2010 -0700 106265 106266 Merge remote branch 'vignatti/vgaarb-fixes' 106267 106268commit 6edbdb7f9e02ecb069c6fe6c0600fc389553ec5d 106269Author: Aaron Plattner <aplattner@nvidia.com> 106270Date: Mon Aug 23 11:37:46 2010 -0700 106271 106272 Tag sdksyms.c entries with source file and line number 106273 106274 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 106275 Reviewed-by: Adam Jackson <ajax@redhat.com> 106276 Signed-off-by: Keith Packard <keithp@keithp.com> 106277 106278commit 9ef6241c2382bfc555284a4985f6d1e37d750d6f 106279Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 106280Date: Mon Aug 23 16:04:03 2010 -0400 106281 106282 Fix property and selection devPrivate allocation. 106283 106284 Selection objects were not being allocated with privates, and both 106285 objects had a stray statement that zeroed out the devPrivates field. 106286 106287 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 106288 Reported-by: Justin Mattock <justinmattock@gmail.com> 106289 Reviewed-by: Adam Jackson <ajax@redhat.com> 106290 Signed-off-by: Keith Packard <keithp@keithp.com> 106291 106292commit 9dca441670d261a9a9fb6108960ed48f3d58fb7f 106293Author: Peter Hutterer <peter.hutterer@who-t.net> 106294Date: Mon Sep 6 11:32:38 2010 +1000 106295 106296 xfree86: add a hook to replace the new console handler. 106297 106298 This hook is only necessary for the keyboard driver to remove the race 106299 condition between drain_console() and the driver's ReadInput (Bug 29969). 106300 106301 The idea is that a driver that needs to handle events from the console 106302 calls xf86ReplaceConsoleHandler() with it's own ReadInput (or NULL) and thus 106303 removes the drain_console call. It's the driver's responsibility to restore 106304 the previous behaviour when the driver is unloaded. 106305 106306 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106307 CC: Thomas Hellstrom <thellstrom@vmware.com> 106308 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106309 Reviewed-by: Adam Jackson <ajax@redhat.com> 106310 106311commit da31ca747f8e3993b2aca5d31b7d09b9907ad1d9 106312Author: Thomas Hellstrom <thellstrom@vmware.com> 106313Date: Thu Sep 2 10:53:28 2010 +0200 106314 106315 linux: Don't lose console events on non-evdev drivers (#29969) 106316 106317 The drain_console() function will race with new keyboard events being added 106318 by the hardware causing the server to lose keyboard events if the console fd 106319 is used for input. 106320 106321 Only use the drain_console() when AllowEmptyInput is off which is the best 106322 indicator we have for whether the keyboard driver will be used. This patch 106323 will only fix the bug when hotplugging is disabled. 106324 What we really need is a way to figure out either whether we're _not_ using 106325 the keyboard driver (not predictable) or a way for the keyboard driver to 106326 disable drain_console(). 106327 106328 X.Org Bug 29969 <http://bugs.freedesktop.org/show_bug.cgi?id=29969> 106329 106330 Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> 106331 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 106332 Reviewed-by: Adam Jackson <ajax@redhat.com> 106333 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106334 (cherry picked from commit 71972c2534d490284d3d42b456c2f34b964b2894) 106335 106336commit 52df92a563980a7af0fb61248da3654bd90e8232 106337Author: Peter Hutterer <peter.hutterer@who-t.net> 106338Date: Thu Sep 2 15:08:01 2010 +1000 106339 106340 xfree86: move XI_VERIFY_VALUATORS to the source file it's used in. 106341 106342 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106343 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106344 106345commit 824e970c5cc94fc7d5fe5f78a24d0d3057d87a84 106346Author: Peter Hutterer <peter.hutterer@who-t.net> 106347Date: Thu Sep 2 15:03:39 2010 +1000 106348 106349 xfree86: remove TS_Raw and TS_Scaled defines. 106350 106351 There are no references to it other than the commit that added them. But 106352 since we're re-doing the API anyway, now is a good time to break things. 106353 106354 commit 9398d62f27ee1b287e4458fd8b011c10f7b59efd 106355 Author: Daniel Stone <daniel@fooishbar.org> 106356 Date: Wed Mar 21 00:18:24 2007 +0200 106357 106358 XFree86 input: Add backwards compatibility for motion history 106359 Add the old motion history API back, as a shim around the new mi 106360 API. 106361 106362 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106363 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106364 106365commit ac3f88e604212fc3668d623798d50aa218c13995 106366Author: Peter Hutterer <peter.hutterer@who-t.net> 106367Date: Thu Sep 2 15:01:26 2010 +1000 106368 106369 xfree86: remove {Dont}SendCoreEvents defines 106370 106371 The input drivers that use it only do so with ABI 0 and we're long past this 106372 one now. Input driver don't have a say in whether they send core events now 106373 anyway. 106374 106375 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106376 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106377 106378commit 3f264149ff226d7c0bfcfcfc2c8845fa6326d0bd 106379Author: Peter Hutterer <peter.hutterer@who-t.net> 106380Date: Thu Sep 2 14:52:30 2010 +1000 106381 106382 xfree86: don't export xf86InputDevs. 106383 106384 Use xf86FirstLocalDevice() instead (but don't get me started on the naming 106385 of that one...) 106386 106387 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106388 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106389 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 106390 106391commit 8f0531bddac8e774aa25b51155445309219da0f6 106392Author: Peter Hutterer <peter.hutterer@who-t.net> 106393Date: Thu Sep 2 14:46:59 2010 +1000 106394 106395 xfree86: remove XI_PRIVATE macro, unused. 106396 106397 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106398 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106399 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 106400 106401commit c08f7afdbcf16d2ba30b67e27162c5aaa1ed3f61 106402Author: Peter Hutterer <peter.hutterer@who-t.net> 106403Date: Thu Sep 2 14:45:03 2010 +1000 106404 106405 xfree86: update comment for xf86DeleteInput 106406 106407 We have a driver hook - it's UnInit. 106408 106409 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106410 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106411 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 106412 106413commit b8d9c5ff0003f15720737949e65e3159c4f4576e 106414Author: Peter Hutterer <peter.hutterer@who-t.net> 106415Date: Thu Sep 2 14:30:11 2010 +1000 106416 106417 xfree86: remove IDevRec, replace with InputInfoRec. 106418 106419 This struct is superfluous, maintaining the same info as the InputInfoRec 106420 (with the exception of the driver name). 106421 106422 This is a rather large commit with the majority of changes being a rename 106423 from the fields of the IDevRec (idev, commonOptions) to the InputInfoRec 106424 (pInfo, options). 106425 106426 The actual changes affect the initialization process of the input device: 106427 In NewInputDeviceRequest, the InputInfoRec is now always allocated and just 106428 added to the internal list in xf86NewInputDevice() if the init process 106429 succeeded. 106430 106431 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106432 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106433 106434commit c5da32a6105b3cf57b5996971149c39103cd64b3 106435Author: Peter Hutterer <peter.hutterer@who-t.net> 106436Date: Thu Sep 2 14:26:20 2010 +1000 106437 106438 xfree86: plug minor memory leak. 106439 106440 When no identifier for the device was specified, the allocated IDevRec (and 106441 its associated fields) need to be freed. 106442 106443 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106444 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106445 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 106446 106447commit 34ade08174bfd652cc79bf26d2ad6a10ca73d2c4 106448Author: Peter Hutterer <peter.hutterer@who-t.net> 106449Date: Thu Sep 2 14:04:50 2010 +1000 106450 106451 xfree86: Remove extraOptions parameter from xf86CollectInputOptions. 106452 106453 And unexport it, drivers don't need to call this in the new init process. 106454 106455 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106456 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106457 106458commit 2199842ed50b3eb40d54146827fc58cae7e873ec 106459Author: Peter Hutterer <peter.hutterer@who-t.net> 106460Date: Thu Sep 2 10:52:54 2010 +1000 106461 106462 xfree86: remove extraOptions field from IDevRec. 106463 106464 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106465 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106466 106467commit 5ab8ca0807f9d2f8c8de2f9a1421e59e2a8e976c 106468Author: Peter Hutterer <peter.hutterer@who-t.net> 106469Date: Thu Sep 2 10:41:27 2010 +1000 106470 106471 xfree86: remove some ifdef 0 code. 106472 106473 InputInfoRec hasn't had a free function pointer since the git import. 106474 106475 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106476 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106477 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 106478 106479commit b0da1bc509446a714eb2b29cbd35d43b1e5187ed 106480Author: Peter Hutterer <peter.hutterer@who-t.net> 106481Date: Thu Sep 2 10:39:24 2010 +1000 106482 106483 xfree86: move a declaration down to the block it is used in. 106484 106485 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106486 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106487 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 106488 106489commit cd7059db4a766bc9d03d7968aebeba8542e6c6c0 106490Author: Peter Hutterer <peter.hutterer@who-t.net> 106491Date: Thu Sep 2 10:36:47 2010 +1000 106492 106493 xfree86: minor comment fix. HAL → config backend 106494 106495 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106496 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106497 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 106498 106499commit f3861522fe7d148fd4b0802f2f9ea5d98ab2db93 106500Author: Peter Hutterer <peter.hutterer@who-t.net> 106501Date: Thu Sep 2 11:47:30 2010 +1000 106502 106503 xfree86: skip a few NULL initialization, calloc does it for us. 106504 106505 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106506 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106507 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 106508 106509commit 12170978617c6862fb6aa286b18aaacd0a4e2737 106510Author: Peter Hutterer <peter.hutterer@who-t.net> 106511Date: Wed Sep 1 15:44:44 2010 +1000 106512 106513 xfree86: remove conversion procs and close proc from InputInfoRec. 106514 106515 None of them are called by the server. 106516 106517 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106518 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106519 Reviewed-by: Fernando Carrijo <fcarrijo@freedesktop.org> 106520 106521commit 71972c2534d490284d3d42b456c2f34b964b2894 106522Author: Thomas Hellstrom <thellstrom@vmware.com> 106523Date: Thu Sep 2 10:53:28 2010 +0200 106524 106525 linux: Don't lose console events on non-evdev drivers (#29969) 106526 106527 The drain_console() function will race with new keyboard events being added 106528 by the hardware causing the server to lose keyboard events if the console fd 106529 is used for input. 106530 106531 Only use the drain_console() when AllowEmptyInput is off which is the best 106532 indicator we have for whether the keyboard driver will be used. This patch 106533 will only fix the bug when hotplugging is disabled. 106534 What we really need is a way to figure out either whether we're _not_ using 106535 the keyboard driver (not predictable) or a way for the keyboard driver to 106536 disable drain_console(). 106537 106538 X.Org Bug 29969 <http://bugs.freedesktop.org/show_bug.cgi?id=29969> 106539 106540 Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> 106541 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 106542 Reviewed-by: Adam Jackson <ajax@redhat.com> 106543 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106544 106545commit e00e2e7b68fbc932269d607ac5dc2c441d07ad9d 106546Author: Peter Hutterer <peter.hutterer@who-t.net> 106547Date: Mon Aug 23 08:35:01 2010 +1000 106548 106549 xfree86: Check for existence of button class before dereferencing it. 106550 106551 The Irxon Super Mini Bluetooth Wireless Keyboard for PC/PDA/Cell Phones 106552 keyboards have axes but not buttons. The evdev driver doesn't set up a 106553 button class for these keyboards and a motion event handled by 106554 DGAProcessPointerEvent dereferences the dev->button NULL pointer, causing a 106555 server crash. 106556 106557 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106558 Reviewed-by: Keith Packard <keithp@keithp.com> 106559 106560commit 9802cca816884e7f055f054ac1ba9194e0eba10b 106561Author: Peter Hutterer <peter.hutterer@who-t.net> 106562Date: Wed Sep 1 15:42:44 2010 +1000 106563 106564 Remove atom field from InputInfoRec. 106565 106566 This field was only used in one location where we can use a local variable. 106567 106568 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106569 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106570 106571commit 79ea9ef39971d008d199b18f34d1aef2bab6e33f 106572Author: Peter Hutterer <peter.hutterer@who-t.net> 106573Date: Tue Aug 10 15:19:20 2010 +1000 106574 106575 input: constify valuators passed in by input drivers. 106576 106577 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106578 Reviewed-by: Keith Packard <keithp@keithp.com> 106579 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 106580 106581commit 86560b5d05f14bdf04d21b3457a66c0d5045db9c 106582Author: Peter Hutterer <peter.hutterer@who-t.net> 106583Date: Tue Aug 24 13:46:35 2010 +1000 106584 106585 dix: don't set time to CurrentTime in DeviceChangedEvents. 106586 106587 CurrentTime is used by clients to skip setting the time, but not by the 106588 server. 106589 106590 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106591 Reviewed-by: Keith Packard <keithp@keithp.com> 106592 106593commit ff055506f0cbb852bed17acb9f9bbf1d715a854e 106594Author: Peter Hutterer <peter.hutterer@who-t.net> 106595Date: Mon Aug 23 13:20:09 2010 +1000 106596 106597 dix: fix crash when removing devices on a buttonless MD pointer (#29669) 106598 106599 If the master does not have a button class, recalculating the number of 106600 buttons required for this master dereferences a NULL pointer. Guard against 106601 this, if the master pointer doesn't have a button class, it doesn't need to 106602 update it's number of buttons. 106603 106604 Reproducible: 106605 Two devices on the same master, device NB with axes but no buttons, device 106606 A+B with axes and button . 106607 If NB was the last one to send an event through the master when A+B is 106608 removed from the server, master->button is NULL and leads to the above 106609 NULL-pointer dereference. 106610 106611 X.Org Bug 29669 <http://bugs.freedesktop.org/show_bug.cgi?id=29669> 106612 106613 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106614 Reviewed-by: Keith Packard <keithp@keithp.com> 106615 106616commit be978c7fc97e98ef1bc7b20c6c0740d8a81b609a 106617Author: Jesse Adkins <jesserayadkins@gmail.com> 106618Date: Tue Aug 3 18:21:28 2010 -0700 106619 106620 xfree86: Purge kbdCustomKeycodes from xf86Info. 106621 106622 This was obsolete after 3eeb62e8f587732e6b433c2b9c6879eb26a3f1b4 "bug #890: completely remove deprecated keyboard driver". 106623 106624 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 106625 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 106626 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106627 106628commit 6922b31a136b2b0ac185d61785969a11f84c7943 106629Author: Peter Hutterer <peter.hutterer@who-t.net> 106630Date: Wed Sep 1 14:35:40 2010 +1000 106631 106632 xfree86: remove history_size 106633 106634 For a couple of ABIs now the history size was essentially static anyway. 106635 106636 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106637 106638commit 892e39025b45109ba38c0b5d0b5f4535e092f58c 106639Author: Peter Hutterer <peter.hutterer@who-t.net> 106640Date: Wed Sep 1 14:33:52 2010 +1000 106641 106642 xfree86: remove always_core_feedback from InputInfoRec. 106643 106644 Unused 106645 106646 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106647 106648commit 5e2912bd1d2bc8ff1508e5f726e77aa5f8677cbd 106649Author: Peter Hutterer <peter.hutterer@who-t.net> 106650Date: Wed Sep 1 14:33:26 2010 +1000 106651 106652 xfree86: remove first/last from InputInfoRec. 106653 106654 Unused. 106655 106656 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106657 106658commit 974bc2322b3667db38d28063786fbc76fb303d00 106659Author: Peter Hutterer <peter.hutterer@who-t.net> 106660Date: Wed Sep 1 14:32:51 2010 +1000 106661 106662 xfree86: remove private_flags from InputInfoRec. 106663 106664 Unused field. 106665 106666 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106667 106668commit 23b361b25b7b48a820a4ef851fb89706f7ec2cd0 106669Author: Peter Hutterer <peter.hutterer@who-t.net> 106670Date: Wed Sep 1 14:14:20 2010 +1000 106671 106672 xfree86: Purge old_x/old_y from InputInfoRec. 106673 106674 Unused field. 106675 106676 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106677 106678commit 4ac3be29bc1af19f15e84ee0ea609de5c54ca0e0 106679Author: Peter Hutterer <peter.hutterer@who-t.net> 106680Date: Fri Jul 30 14:26:06 2010 +1000 106681 106682 input: Purge AddOtherInputDevices DDX hook. 106683 106684 This hook wasn't used by any DDX. Device addition and removal is handled by 106685 the config backend, so we don't need to do anything special that during the 106686 ListInputDevices request processing. 106687 106688 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106689 Reviewed-by: Adam Jackson <ajax@redhat.com> 106690 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106691 106692commit cb672a461cc6cd668ab7e61994b94e9ff46b3ef1 106693Author: Peter Hutterer <peter.hutterer@who-t.net> 106694Date: Fri Jul 30 14:21:14 2010 +1000 106695 106696 input: remove OpenInputDevice and CloseInputDevice DDX hooks. 106697 106698 In theory, these hooks were to be used for DDX-specific device enablement. 106699 None of the DDXs however did anything here. Now we call DEVICE_INIT on all 106700 devices when they are added, so the xfree86 DDX as the only one with real 106701 code didn't do anything here. 106702 106703 kdrive checked for device validity but that's already handled in 106704 ProcXOpenDevice. 106705 106706 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106707 Reviewed-by: Adam Jackson <ajax@redhat.com> 106708 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106709 106710commit a46d2bb344e822bbd9d69cb59829b85f9d8f0213 106711Author: Simon Thum <simon.thum@gmx.de> 106712Date: Sun Jul 18 12:31:26 2010 +0200 106713 106714 xfree86: Fix xf86 backend-specific input initialization 106715 106716 Instead of shoving it in rather unrelated places, move acceleration init 106717 into xf86NewInputDevice. 106718 106719 Caveat: It's not clear atm how relevant other callers of ActivateDevice 106720 (like OpenDevice) actually are. 106721 106722 Signed-off-by: Simon Thum <simon.thum@gmx.de> 106723 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 106724 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106725 Reviewed-by: Adam Jackson <ajax@redhat.com> 106726 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106727 106728commit e930710ae5579752785d6b96ace4b44bf0199a6e 106729Author: Peter Hutterer <peter.hutterer@who-t.net> 106730Date: Fri Jul 30 13:59:29 2010 +1000 106731 106732 xfree86: purge superfluous includes from xf86Xinput.c 106733 106734 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106735 Reviewed-by: Adam Jackson <ajax@redhat.com> 106736 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106737 106738commit 2b7840b63da6bc00aa40e8a427d9fd719ba5e6d0 106739Author: Peter Hutterer <peter.hutterer@who-t.net> 106740Date: Fri Jul 30 13:44:20 2010 +1000 106741 106742 xfree86: remove unused DeviceAssocRec struct. 106743 106744 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106745 Reviewed-by: Adam Jackson <ajax@redhat.com> 106746 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106747 106748commit cdb5863c9f68548afa0c69e08c697c83f9a5e671 106749Author: Peter Hutterer <peter.hutterer@who-t.net> 106750Date: Fri Jul 23 14:16:38 2010 +1000 106751 106752 xfree86: remove LocalDeviceRec/Ptr definition. 106753 106754 Two names pointing to the same struct for over 7 years now. Remove the 106755 define, if drivers don't want to change they can always do the typedef 106756 themselves. 106757 106758 Rename all "LocalDevicePtr local" to "InputInfoPtr pInfo". 106759 106760 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106761 Reviewed-by: Adam Jackson <ajax@redhat.com> 106762 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106763 106764commit 8764782f6de56a9dc5e9d5a8e9fb616a8ddb2f7c 106765Author: Peter Hutterer <peter.hutterer@who-t.net> 106766Date: Mon Jun 7 11:17:10 2010 +1000 106767 106768 xfree86: add xf86IDrvMsg and friends for input driver logging. 106769 106770 Input driver messages are only standardised by convention, with the drivers 106771 prefixing the device name to most messages. This makes it rather hard to 106772 grep on "evdev" for example when looking for the evdev ouput. 106773 106774 This patch adds three new logging functions, modeled after xf86DrvMsg(), the 106775 logging function for output drivers. New functions are 106776 xf86IDrvMsg() - input driver log message in default verbosity. 106777 xf86IDrvMsgVerb() - input driver log message in specified verbosity. 106778 xf86VIDrvMsgVerb() - same as xf86IDrvMsgVerb() but takes a varargs 106779 argument. 106780 106781 Default log format is <driver name>: <device name>: <message>. 106782 106783 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106784 Reviewed-by: Adam Jackson <ajax@redhat.com> 106785 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106786 106787commit b55ac354afa0f16f71bb4f2d3d4277c887082520 106788Author: Peter Hutterer <peter.hutterer@who-t.net> 106789Date: Thu Jul 22 12:00:07 2010 +1000 106790 106791 xfree86: move xf86AllocateInput and xf86DeleteInput to xf86Xinput.c 106792 106793 Make xf86AllocateInput static in the process, this function is only called 106794 from one location. 106795 106796 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106797 Reviewed-by: Adam Jackson <ajax@redhat.com> 106798 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106799 106800commit d2a5f4166417b9d4a02a219b28470e41b6cfe012 106801Author: Peter Hutterer <peter.hutterer@who-t.net> 106802Date: Thu Jul 22 10:30:56 2010 +1000 106803 106804 xfree86: return the device from xf86ActivateDevice. 106805 106806 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106807 Reviewed-by: Adam Jackson <ajax@redhat.com> 106808 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106809 106810commit fbf35e27f3005cf009f2dd1e112260bb294342c7 106811Author: Peter Hutterer <peter.hutterer@who-t.net> 106812Date: Thu Jul 22 10:28:14 2010 +1000 106813 106814 input: set XKB extension for all new devices, not just xfree86 ones. 106815 106816 Right now, Xephyr and others don't get to use XKB on the slave devices. 106817 Which works given that no-one cares about SDs just yet but event processing 106818 is different if the ProcessInputProc isn't wrapped properly. 106819 106820 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106821 Reviewed-by: Adam Jackson <ajax@redhat.com> 106822 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106823 106824commit 11ed32b62c8793ecc1c68e1e4ba91e2eb45eef3e 106825Author: Peter Hutterer <peter.hutterer@who-t.net> 106826Date: Thu Jul 22 10:24:02 2010 +1000 106827 106828 xfree86: purge SendDragEvents support. 106829 106830 From the documentation: 106831 "This is mainly to allow a touch screen to be used with netscape and other 106832 browsers which do strange things if the mouse moves between button down and 106833 button up." 106834 106835 CLOSED - NOTOURBUG 106836 106837 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106838 Reviewed-by: Adam Jackson <ajax@redhat.com> 106839 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106840 106841commit 7defd282beb57e8880980416ed579f62d561d1ac 106842Author: Peter Hutterer <peter.hutterer@who-t.net> 106843Date: Thu Jul 22 10:16:15 2010 +1000 106844 106845 xfree86: remove XI86_CONFIGURED flag. 106846 106847 PreInit returns a status code. Let's use that instead of having it report 106848 Success in some cases but not set the XI86_CONFIGURED flag and thus signal 106849 an init failure. 106850 106851 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106852 Reviewed-by: Adam Jackson <ajax@redhat.com> 106853 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106854 106855commit 97a4acdac23b8ff33da43917c10b65d2fee5c07f 106856Author: Peter Hutterer <peter.hutterer@who-t.net> 106857Date: Thu Jul 22 10:04:53 2010 +1000 106858 106859 xfree86: purge some unused defines. 106860 106861 These defines have been write-only for a while now. 106862 106863 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106864 Reviewed-by: Adam Jackson <ajax@redhat.com> 106865 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106866 106867commit 0fb7a5c261aa0d87d6596d72b70696bffe0c0aff 106868Author: Peter Hutterer <peter.hutterer@who-t.net> 106869Date: Thu Jul 22 09:53:35 2010 +1000 106870 106871 input: Purge Register*Device() functions. 106872 106873 RegisterPointerDevice() and RegisterKeyboardDevice() were already mapped to 106874 RegisterOtherDevice() and obsolete. 106875 106876 RegisterOtherDevice() was called for all devices and the two assignments can 106877 simply be moved into AddInputDevice(). Purge RegisterOtherDevice() and 106878 pretend it never happened. 106879 106880 *lalalalala* 106881 106882 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106883 Reviewed-by: Adam Jackson <ajax@redhat.com> 106884 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106885 106886commit 67ffbcc14cbc61474520d4531599edca24965543 106887Author: Peter Hutterer <peter.hutterer@who-t.net> 106888Date: Thu Jul 22 09:24:08 2010 +1000 106889 106890 xfree86: remove superflous assignments. 106891 106892 ActivateGrab and DeactivateGrab are set in AddInputDevice() already. 106893 106894 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106895 Reviewed-by: Adam Jackson <ajax@redhat.com> 106896 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106897 106898commit 4cd54d9ed9e87074734789a9a7708c2218f87f1e 106899Author: Peter Hutterer <peter.hutterer@who-t.net> 106900Date: Thu Jul 22 09:22:05 2010 +1000 106901 106902 xfree86: make xf86ActivateDevice static. 106903 106904 No-one but the joystick driver uses it and that one should be using NIDR 106905 instead. 106906 106907 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106908 Reviewed-by: Adam Jackson <ajax@redhat.com> 106909 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106910 106911commit 31c71425ac13a7f554316356691a79175ea82a67 106912Author: Peter Hutterer <peter.hutterer@who-t.net> 106913Date: Thu Jul 22 09:09:10 2010 +1000 106914 106915 config: expose config_info as an input option. 106916 106917 config_info is the only reliable indicator we have in the server for 106918 duplicate devices (drivers can test for maj/min on fds as well). Don't set 106919 this after the device has been initialized but assume it's important enough 106920 to set during NIDR. 106921 106922 This makes the option "config_info" available to the drivers as well. 106923 106924 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106925 Reviewed-by: Adam Jackson <ajax@redhat.com> 106926 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106927 106928commit de0cc5a72deb7c477e368aa4fe9a713788d7ae4c 106929Author: Peter Hutterer <peter.hutterer@who-t.net> 106930Date: Wed Jul 21 16:00:26 2010 +1000 106931 106932 xfree86: rework driver PreInit API - XInput ABI 12 106933 106934 The main change introduced in this patch is the removal of the 106935 back-and-forth between DDX and the driver. 106936 The DDX now allocates the InputInfoRec and fills it with default values. The 106937 DDX processes common options (and module-specific default options, if 106938 appropriate) before passing the initialised struct to the driver. 106939 106940 The driver may do module-specific initializations and return Success or an 106941 error code in the case of a failure. 106942 106943 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106944 Reviewed-by: Adam Jackson <ajax@redhat.com> 106945 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 106946 106947commit 7925e8945649d4af237e6c3c5593b895a461bd1e 106948Author: Chase Douglas <chase.douglas@ubuntu.com> 106949Date: Wed Sep 1 14:45:34 2010 +1000 106950 106951 Fix udev population of Bluetooth input device product IDs 106952 106953 The udev device_added function takes the vendor and model IDs of added 106954 devices and converts them into an attribute that can be matched for by 106955 an InputClass configuration using MatchUSBID. Currently, the udev 106956 mechanism works for USB devices, but fails to work properly for 106957 Bluetooth devices. The product IDs of the event node are actually the 106958 IDs of the Bluetooth receiver instead of the device. 106959 106960 This patch reads the product ID from the PRODUCT property of the parent 106961 of the added device. This tag is set correctly for both USB and 106962 Bluetooth input devices. The following devices have been tested by 106963 specifying individual InputClass sections in xorg.conf: 106964 106965 * Apple Keyboard (Bluetooth) 106966 * Apple Magic Trackpad (Bluetooth) 106967 * Apple Magic Mouse (Bluetooth) 106968 * Microsoft Bluetooth Notebook Mouse 5000 (Bluetooth) 106969 * Microsoft IntelliMouse Optical (USB) 106970 * N-Trig Touchscreen (USB) 106971 * Wacom Bamboo Touch (USB) 106972 106973 Signed-off-by: Chase Douglas <chase.douglas@canonical.com> 106974 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 106975 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106976 106977commit 05e616767e5b7e60b92d31c4042ded5892dce6d4 106978Author: Adam Tkac <atkac@redhat.com> 106979Date: Wed Aug 25 10:38:40 2010 +0200 106980 106981 Return Success from generate_modkeymap() when max_keys_per_mod is zero 106982 106983 max_keys_per_mod equal to zero is a valid situation so generate_modkeymap 106984 should not return BadAlloc in this case. 106985 106986 Signed-off-by: Adam Tkac <atkac@redhat.com> 106987 Reviewed-by: Patrick E. Kane <pekane52 at gmail.com> 106988 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 106989 106990commit 4a12aecac670debd0dafb17c245fccb93eea2d60 106991Author: Jesse Adkins <jesserayadkins@gmail.com> 106992Date: Wed Aug 25 13:48:29 2010 -0700 106993 106994 xfree86: Document terminate not mapped by default (bug 25083) 106995 106996 Document that terminate is not mapped to Ctrl+Alt+Backspace by default, to help alleviate some confusion. 106997 106998 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 106999 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107000 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107001 107002commit a90052ba8697e217b0dc68057d7b9202ae8797db 107003Author: David Ge <davidqge@gmail.com> 107004Date: Thu Aug 19 00:33:57 2010 -0500 107005 107006 xkb: Fix RedirectKey didn't send any event. 107007 107008 Xorg.log shows error: Valuators reported for non-valuator device. 107009 This is caused by uninitialized valuators.mask in _XkbFilterRedirectKey(), 107010 which trigger the error in UpdateDeviceState(). 107011 107012 Signed-off-by: David Ge <davidqge@gmail.com> 107013 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107014 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107015 107016commit ff109bf84401a451380eb7f3f94a6e0aa2776e3e 107017Author: Alan Coopersmith <alan.coopersmith@oracle.com> 107018Date: Thu Aug 12 00:09:02 2010 -0700 107019 107020 Use GetMaster instead of direct u.master access in core procs 107021 107022 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107023 Acked-by: Daniel Stone <daniel@fooishbar.org> 107024 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107025 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107026 107027commit 3cc5e4422430e9ca44615f3e63feccd2e5729046 107028Author: Peter Hutterer <peter.hutterer@who-t.net> 107029Date: Fri Jul 23 14:48:32 2010 +1000 107030 107031 xfree86: fix compiler warning about implicied decl of DuplicateModule. 107032 107033 ../../../../hw/xfree86/common/xf86Xinput.c: In function ‘xf86AllocateInput’: 107034 ../../../../hw/xfree86/common/xf86Xinput.c:722: warning: implicit 107035 declaration of function ‘DuplicateModule’ 107036 ../../../../hw/xfree86/common/xf86Xinput.c:722: warning: nested extern 107037 declaration of ‘DuplicateModule’ 107038 ../../../../hw/xfree86/common/xf86Xinput.c:722: warning: assignment makes 107039 pointer from integer without a cast 107040 107041 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107042 Reviewed-by: Adam Jackson <ajax@redhat.com> 107043 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 107044 107045commit b5c9953bbf4ffd11f1a70d058c6d3feb2bd1bca8 107046Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107047Date: Mon Jul 26 15:31:03 2010 +0300 107048 107049 xkb: Check if AddResource failed 107050 107051 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107052 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107053 107054commit 2e6d7174042cc8007e947b7d9fb54acc0ebe29d2 107055Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107056Date: Mon Jul 26 15:13:34 2010 +0300 107057 107058 xkb: Fix possible NULL pointer dereference 107059 107060 sli is null before allocation assigment so deference t osli has to be 107061 protected. 107062 107063 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107064 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107065 107066commit d6642de7ebdda16e0056600a86a7802bd4c393b7 107067Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107068Date: Mon Jul 26 14:50:30 2010 +0300 107069 107070 xkb: Fix possible NULL pointer dereference 107071 107072 If search for device failed sli is NULL. In that case we have to protect 107073 dereference to prevent server crash. 107074 107075 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107076 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107077 107078commit adc0697cfcfba295a15d7a307125093cbccd637f 107079Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107080Date: Thu Jul 22 15:11:27 2010 +0300 107081 107082 xkb: Fix memory leak in error path 107083 107084 map is allocated but not freed if reply length and data don't match. 107085 107086 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107087 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107088 107089commit 67cfb66562cd9d39f30fec6fbc38eb1eb5e5b030 107090Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107091Date: Thu Jul 22 15:05:57 2010 +0300 107092 107093 xkb: Remove redurant intialization code 107094 107095 calloc already initializes allocated memory to zero. 107096 107097 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107098 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107099 107100commit 2475ef60977f6813dec74ef0837a5915b8a48bbc 107101Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107102Date: Thu Jul 22 12:48:55 2010 +0300 107103 107104 xkb: Fix NULL pointer dereference 107105 107106 xkb->names is dereferenced in else path too. 107107 107108 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107109 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107110 107111commit 184ef0d35612d6ed0619283d376f04d9a904f47c 107112Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107113Date: Thu Jul 22 11:34:54 2010 +0300 107114 107115 xkb: Don't check for NULL before calling free 107116 107117 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107118 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107119 107120commit 1223340644744c0b38aa85f5956eb5ab7c696517 107121Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107122Date: Thu Jul 22 11:33:33 2010 +0300 107123 107124 xkb: Fix memory leak if opening file fails 107125 107126 If fopen fails pointer in buf would be overwriten with a new pointer. 107127 107128 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107129 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107130 107131commit 20cb9c923efa4edc348eba30f956a66413a8208f 107132Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107133Date: Thu Jul 22 11:13:10 2010 +0300 107134 107135 xkb: Use memcpy for copy that has known length 107136 107137 Fixes warning that strncpy is not able to append NULL to the end 107138 of destination. 107139 107140 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 107141 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107142 107143commit de8be07cc0a8163b6ef04455706fd5ca2cebe587 107144Author: Peter Hutterer <peter.hutterer@who-t.net> 107145Date: Tue Aug 17 12:08:52 2010 +1000 107146 107147 dix: don't create core motion events for non-x/y valuators. 107148 107149 Devices that send motion events with valuators other than x/y get core 107150 motion events with unchanged x/y coordinates. This confuses some 107151 applications. 107152 107153 If the DeviceEvent does not have the x/y valuators set, return BadMatch on 107154 core conversion, thus skipping the event altogether. 107155 107156 Reported-by: Bartosz Brachaczek <b.brachaczek@gmail.com> 107157 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107158 Tested-by: Bartosz Brachaczek <b.brachaczek@gmail.com> 107159 107160commit 08adf41f6315663cbac33d010214d98f3e1c8814 107161Author: Matt Turner <mattst88@gmail.com> 107162Date: Fri Aug 27 18:34:49 2010 -0400 107163 107164 Replace malloc/strlen/strcpy with strdup. 107165 107166 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 107167 Signed-off-by: Matt Turner <mattst88@gmail.com> 107168 107169commit ea239112b030588ed3cdd113643ba5f6207a5bd9 107170Author: Jesse Adkins <jesserayadkins@gmail.com> 107171Date: Wed Aug 4 09:21:33 2010 +0000 107172 107173 xfree86: Purge parsePrologueVoid. 107174 107175 This was included in the original commit, and then never used. 107176 107177 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 107178 Signed-off-by: Matt Turner <mattst88@gmail.com> 107179 107180commit b25fb9fe998209e79b6917d12d5765c5f6719a7e 107181Author: Jesse Adkins <jesserayadkins@gmail.com> 107182Date: Wed Aug 4 09:21:32 2010 +0000 107183 107184 xfree86: Removed unused messages from Configint.h 107185 107186 AUTOREPEAT_MSG, MOVED_TO_FLAGS_MSG, and XLEDS_MSG made obsolete by 107187 81913a12910e39d7ea6af8657c1c66cc6791cd65 Jul 21 2006 (remove undead files from master) 107188 UNDEFINED_DEVICE_MSG made obsolete by 107189 6033d8150be3a115b90226eaa42f237bb0cf3369 Oct 9 2007 (first pass at video driver autoloading) 107190 107191 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 107192 Signed-off-by: Matt Turner <mattst88@gmail.com> 107193 107194commit 18b62e0479f15e965611880ada6e0195367df025 107195Author: Jesse Adkins <jesserayadkins@gmail.com> 107196Date: Wed Aug 4 09:21:31 2010 +0000 107197 107198 xfree86: Fix leaks in OpenConfigFile and OpenConfigDir 107199 107200 [mattst88: fixed whitespace and a missing semicolon] 107201 107202 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 107203 Signed-off-by: Matt Turner <mattst88@gmail.com> 107204 107205commit 4f04fd595e82226f1d91226a41bb98ed3d940b37 107206Author: Jesse Adkins <jesserayadkins@gmail.com> 107207Date: Wed Aug 4 09:21:30 2010 +0000 107208 107209 xfree86: Simplify xf86Msg{,Verb} 107210 107211 Previously, the functions would call xf86VDrvMsgVerb with a screen of -1 107212 despite their comments saying they were for "non-driver messages". 107213 They now call LogVMessageVerb, which is what xf86VDrvMsgVerb does anyway 107214 when it has a screen == -1. 107215 107216 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107217 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 107218 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 107219 Signed-off-by: Matt Turner <mattst88@gmail.com> 107220 107221commit f49473abfd6034e68576b2dddd30ba8d8dd0838f 107222Author: Jesse Adkins <jesserayadkins@gmail.com> 107223Date: Wed Aug 4 09:21:29 2010 +0000 107224 107225 xfree86: Remove prototypes for non-existant functions. 107226 107227 This was obsolete from 9a0f25de7ca3c68af867b38936103d17daa92ac6 "Static cleanups, dead code deletion." (server 1.3). 107228 107229 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 107230 Signed-off-by: Matt Turner <mattst88@gmail.com> 107231 107232commit 747bf5fe80f51554205c5a50b8ed9b89065c8a54 107233Author: Jesse Adkins <jesserayadkins@gmail.com> 107234Date: Sun Aug 1 16:01:23 2010 -0700 107235 107236 xfree86: Remove comments about unable to use malloc. 107237 107238 These are leftovers from when X still used Xmalloc and friends for allocation. 107239 Now that those are gone, these comments are just confusing. 107240 107241 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 107242 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107243 Signed-off-by: Matt Turner <mattst88@gmail.com> 107244 107245commit 65466652936bbb2706be455d0a416bcf08e88f66 107246Author: Tiago Vignatti <tiago.vignatti@nokia.com> 107247Date: Wed May 12 13:03:14 2010 +0300 107248 107249 xfree86: vgaarb: fix device decoding interface to send resources type properly 107250 107251 Right now, when there is more than one vide card on the machine, we're 107252 adopting a pessimistic approach and setting all cards to decode VGA legacy 107253 address. 107254 107255 Some cards may want to skip the arbitration and the only way to do so is 107256 through pci_device_vgaarb_decodes. Therefore, send the desired kind of 107257 resource instead force the worst case. 107258 107259 Note that xf86VGAarbiterDeviceDecodes is not being used so far by any 107260 open-source driver. Even so, API break. 107261 107262 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 107263 Reviewed-by: Mark Kettenis <kettenis@openbsd.org> 107264 107265commit 7fbf3e7cf4b261194faad747e0bf59d528df6d91 107266Author: Tiago Vignatti <tiago.vignatti@nokia.com> 107267Date: Mon May 10 21:05:14 2010 +0300 107268 107269 xfree86: vgaarb: remove useless macro 107270 107271 This was inherited from RAC and was never used there either. 107272 107273 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 107274 Reviewed-by: Jamey Sharp <jamey@minilop.net> 107275 107276commit ad698dd5bcc41fcec1e7ce1117c7ad1052710132 107277Author: Tiago Vignatti <tiago.vignatti@nokia.com> 107278Date: Mon May 10 21:03:30 2010 +0300 107279 107280 xfree86: vgaarb: remove superfluous and confusing VGAGet_GC and VGAPut_GC 107281 107282 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 107283 Reviewed-by: Jamey Sharp <jamey@minilop.net> 107284 107285commit c0aed4c99bf553bd7b8bbc79d0ed8f26d0ab3f94 107286Author: Tiago Vignatti <tiago.vignatti@nokia.com> 107287Date: Mon May 10 20:58:30 2010 +0300 107288 107289 xfree86: vgaarb: change macros by inline functions to ease debug 107290 107291 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 107292 Reviewed-by: Jamey Sharp <jamey@minilop.net> 107293 107294commit 79ee78de9de49d0cab03401662baa476a18e53b8 107295Author: Keith Packard <keithp@keithp.com> 107296Date: Fri Aug 20 17:32:27 2010 -0700 107297 107298 xserver 1.9.0 107299 107300 Signed-off-by: Keith Packard <keithp@keithp.com> 107301 107302commit 3e56efcfb63677cd8574e1e435e61d96f79ea536 107303Author: Keith Packard <keithp@keithp.com> 107304Date: Fri Aug 20 10:01:48 2010 -0700 107305 107306 fb: make isClipped always reject negative coordinates (bug 11503) 107307 107308 A window with either dimension > 32767 can be positioned such that 107309 coordinates > 32767 are visible on the screen. Attempts to draw to 107310 those pixels will generate coordinates wrapped around to negative 107311 values. 107312 107313 The optimized clipping macro, 'isClipped', in fbbits.h, computes 107314 clipping in window space rather than screen space using int16 values, 107315 and so it too has coordinates wrapped around to negative values and 107316 hence ends up accepting the wrapped drawing coordinates. 107317 107318 Two possible fixes for this problem 107319 107320 1) Detect wrapped region coordinates and clip those to 32767. 107321 2) Detect negative incoming coordinates and reject those 107322 107323 This patch takes the second approach as it is much shorter, simply 107324 detecting when either X or Y incoming coordinate is negative, which 107325 can never be 'within' any drawable. 107326 107327 Signed-off-by: Keith Packard <keithp@keithp.com> 107328 Reviewed-by: Adam Jackson <ajax@redhat.com> 107329 107330commit 951605b4660290044fb238bcf1d6d9e498567e8c 107331Author: Chris Wilson <chris@chris-wilson.co.uk> 107332Date: Fri Aug 20 13:51:04 2010 +0100 107333 107334 edid: Adjust rounding of max_clock 107335 107336 A simple hack to accommodate various EDID who have detailed modes that 107337 exceed the EDID's max pixel clock. The pixel clock is only defined in 107338 units of 10MHz and often appears as the maximum pixel code of the 107339 detailed modes, rounded to the nearest 10MHz. Adjusting the max_clock to 107340 include an extra 5MHz prevents the parser from rejecting the detailed 107341 modes. 107342 107343 The kernel uses the same fuzz and by including it in X we can use the 107344 same modes in X as for the console. 107345 107346 Fixes: 107347 107348 Bug 23833 - X uses different refresh rate to that set by kernel module 107349 https://bugs.freedesktop.org/show_bug.cgi?id=23833 107350 107351 In the future, we will want to try harder to keep the KMS modes but at 107352 the same time we need to apply the restrictions as specified by the 107353 user's configuration, and need to fill in modes for fullscreen games on 107354 fixed-mode panels. 107355 107356 Reported-and-tested-by: Fabio Pedretti <fabio.ped@libero.it> 107357 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 107358 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 107359 Signed-off-by: Keith Packard <keithp@keithp.com> 107360 107361commit 7e581780603d6b15291d032efdeeca77f969e0ba 107362Author: Chris Wilson <chris@chris-wilson.co.uk> 107363Date: Thu Jun 24 12:24:58 2010 +0100 107364 107365 glx: Prevent NULL context deref in __glXGetDrawable() (bug 29184) 107366 107367 During a SwapBuffers request, we may end up querying an unknown drawable 107368 outside of an active context, and so need to report this error prior to 107369 attempting to dereference the NULL context. 107370 107371 Also fixes: 107372 107373 [Bug 29184] glXSwapBuffers with no GLX context crashes X. 107374 https://bugs.freedesktop.org/show_bug.cgi?id=29184 107375 107376 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 107377 Cc: Kristian Høgsberg <krh@bitplanet.net> 107378 Reviewed-by: Adam Jackson <ajax@redhat.com> 107379 Signed-off-by: Keith Packard <keithp@keithp.com> 107380 107381commit 5725849a1b427cd4a72b84e57f211edb35838718 107382Author: Adam Jackson <ajax@redhat.com> 107383Date: Mon Jun 28 18:08:50 2010 -0400 107384 107385 render: Bounds check for nglyphs in ProcRenderAddGlyphs (#28801) 107386 107387 Signed-off-by: Adam Jackson <ajax@redhat.com> 107388 Reviewed-by: Julien Cristau <jcristau@debian.org> 107389 Signed-off-by: Keith Packard <keithp@keithp.com> 107390 107391commit fc091936e2bddbbab9c9a501edc5a5f08388617e 107392Author: Peter Hutterer <peter.hutterer@who-t.net> 107393Date: Mon Aug 16 14:18:45 2010 +1000 107394 107395 dix: copy the valuators passed into GPE/GKVE/GProxE. 107396 107397 GPE and friends modify the valuators array passed in. Which means any driver 107398 using e.g. xf86PostButtonEventP(..., valuators) twice to emulate a button 107399 click will provide garbage data on the second run. 107400 107401 This is currently affecting the wacom driver, xf86PostButtonEventP() with 107402 valuators is required to have input events with device-specific axis values. 107403 Passing the same valuators in twice, once with press, once with release, 107404 will see the valuators modified in the first call and garbage submitted in 107405 the next one. 107406 107407 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107408 Reviewed-by: Keith Packard <keithp@keithp.com> 107409 Signed-off-by: Keith Packard <keithp@keithp.com> 107410 107411commit 6e3e559e9fa63069a10eb834a6dab9a4cfc140ee 107412Author: Keith Packard <keithp@keithp.com> 107413Date: Sun Aug 15 20:53:20 2010 -0700 107414 107415 dix: reset pScreen->root to NULL when root window is deleted. 107416 107417 From: Dave Airlie <airlied@linux.ie> 107418 107419 We were seeing a crash in the FreeAllResources codepath, 107420 running valgrind revealed this, 107421 107422 ==12536== Invalid read of size 4 107423 ==12536== at 0x810BCAB: DeliverPropertyEvent (rrproperty.c:33) 107424 ==12536== by 0x80958A4: TraverseTree (window.c:227) 107425 ==12536== by 0x809593E: WalkTree (window.c:255) 107426 ==12536== by 0x810BC66: RRDeliverPropertyEvent (rrproperty.c:53) 107427 ==12536== by 0x810BD5D: RRDeleteProperty.clone.0 (rrproperty.c:76) 107428 ==12536== by 0x810BD98: RRDeleteAllOutputProperties (rrproperty.c:88) 107429 ==12536== by 0x810A36E: RROutputDestroyResource (rroutput.c:407) 107430 ==12536== by 0x808DF4E: FreeClientResources (resource.c:859) 107431 ==12536== by 0x808E005: FreeAllResources (resource.c:876) 107432 ==12536== by 0x8062300: main (main.c:305) 107433 ==12536== Address 0x46ba8ac is 4 bytes inside a block of size 164 free'd 107434 ==12536== at 0x40057F6: free (vg_replace_malloc.c:325) 107435 ==12536== by 0x8087F1F: _dixFreeObjectWithPrivates (privates.c:357) 107436 ==12536== by 0x809832A: DeleteWindow (window.c:926) 107437 ==12536== by 0x808DF4E: FreeClientResources (resource.c:859) 107438 ==12536== by 0x808E005: FreeAllResources (resource.c:876) 107439 ==12536== by 0x8062300: main (main.c:305) 107440 107441 Its a use after free on the root window, since we have already deleted it 107442 at this point. This patch checks if the window we are destroying is the root 107443 window and resets the pointer to NULL if it is. 107444 107445 Signed-off-by: Keith Packard <keithp@keithp.com> 107446 Reviewed-by: Dave Airlie <airlied@redhat.com> 107447 Tested-by: Dave Airlie <airlied@redhat.com> 107448 107449commit 5d1d9d9ae39fab2ee2ac085f9776f82768828dc8 107450Author: Jeremy Huddleston <jeremyhu@apple.com> 107451Date: Sun Aug 1 11:41:58 2010 -0700 107452 107453 XQuartz: xpr: Bail on errors during unlock and destroy 107454 107455 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 107456 107457commit ee7fd8fc58d9fadfbb92302ddea224537f068538 107458Author: Jeremy Huddleston <jeremyhu@apple.com> 107459Date: Sun Aug 1 11:39:14 2010 -0700 107460 107461 XQuartz: UpdateScreen at the end of SetRootless 107462 107463 This will ensure that pRoot is unlocked after the miPaintWindow 107464 107465 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 107466 107467commit 4fc4cab98d454afbfd0d2f48548b5b481e8e7c82 107468Author: Jeremy Huddleston <jeremyhu@apple.com> 107469Date: Thu Jul 29 14:49:10 2010 -0700 107470 107471 XQuartz: Make application switching work better for the no-spaces case 107472 107473 We still have the issue with not raising the frontmost window for the case 107474 when spaces is enabled, and the AppleSpacesSwitchOnActivate preference is 107475 disabled. 107476 107477 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 107478 107479commit e5bc62a03289f956c54c4699edf47f7ff237b5be 107480Author: Jeremy Huddleston <jeremyhu@apple.com> 107481Date: Sun Jul 25 22:29:11 2010 -0700 107482 107483 XQuartz: Ignore kXquartzToggleFullscreen when rootless 107484 107485 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 107486 107487commit 71af1f71c0492c365707c6b3810f94642ff39352 107488Author: Keith Packard <keithp@keithp.com> 107489Date: Thu Aug 12 23:01:59 2010 -0700 107490 107491 Bump to version 1.8.99.906 (1.9 RC6) 107492 107493 Signed-off-by: Keith Packard <keithp@keithp.com> 107494 107495commit 0af322858e86665ee43f065741318e69c2755510 107496Author: Keith Packard <keithp@keithp.com> 107497Date: Thu Aug 12 22:56:36 2010 -0700 107498 107499 Silence GCC warning about uninitialized lastSlave variable 107500 107501 Not an actual bug, but gcc can't tell that this variable cannot be 107502 used without being initialized 107503 107504 Signed-off-by: Keith Packard <keithp@keithp.com> 107505 107506commit b5cf9c5090d15a50b105470900823f2d398d4bd2 107507Author: Alan Coopersmith <alan.coopersmith@oracle.com> 107508Date: Thu Aug 12 00:09:01 2010 -0700 107509 107510 Stop checking or calling PtrCtrlProcs 107511 107512 None of them do anything useful now that pointer acceleration is 107513 entirely handled in the server. (Does not completely nuke yet, 107514 since that would be an API/ABI break.) 107515 107516 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107517 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107518 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107519 107520commit bce12f2956f23c0ee53f7f6485dba631293a0931 107521Author: Jesse Adkins <jesserayadkins@gmail.com> 107522Date: Wed Aug 4 23:39:14 2010 -0700 107523 107524 xfree86: parser: Never use constant strings for driver names (fixes #17438) 107525 107526 When the parser sees the "keyboard" driver, it automatically (and 107527 silently) replaces it with the constant string "kbd". 107528 Everybody else uses malloc'd memory for the driver name, so input 107529 device closure assumes it can use free. 107530 Free val.str, so this crash doesn't turn into a memory leak. Whew. 107531 107532 Signed-off-by: Jesse Adkins <jesserayadkins@gmail.com> 107533 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 107534 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107535 107536commit 619ca32202cd22f2a408586cbc906b8bbaeb9358 107537Author: Peter Hutterer <peter.hutterer@who-t.net> 107538Date: Wed Jul 28 15:08:27 2010 +1000 107539 107540 Xi: reset the unused classes pointer after copying 107541 107542 After copying the unused_classes into the device, reset the original 107543 pointer. Otherwise we have two pointers pointing to the same field and both 107544 get freed on device removal. 107545 107546 Some classes already have this behaviour since 51c8fd69. 107547 107548 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107549 Reviewed-by: Keith Packard <keithp@keithp.com> 107550 107551commit 1a172f3297369a72865232c382abfc14281102a4 107552Author: Peter Hutterer <peter.hutterer@who-t.net> 107553Date: Fri Jul 23 13:24:34 2010 +1000 107554 107555 xkb: if the button isn't down, don't fake an event. 107556 107557 If the button we're about to fake isn't down (or up), don't fake a release 107558 (or press) event for it. Behaviour is the same as before, this just saves 107559 a few cycles. 107560 107561 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107562 Reviewed-by: Keith Packard <keithp@keithp.com> 107563 107564commit 651c36e95ec0ac60d3fb98966df4218712ae78c2 107565Author: Peter Hutterer <peter.hutterer@who-t.net> 107566Date: Fri Jul 23 11:46:30 2010 +1000 107567 107568 xkb: post-fix PointerKeys button events with a DeviceChangedEvent. 107569 107570 commit 14327858391ebe929b806efb53ad79e789361883 107571 xkb: release XTEST pointer buttons on physical releases. (#28808) 107572 revealed a bug with the XTEST/PointerKeys interaction. 107573 107574 Events resulting from PointerKeys are injected into the event processing 107575 stream, not appended to the event queue. The events generated for the fake 107576 button press include a DeviceChangedEvent (DCE), a raw button event and the 107577 button event itself. The DCE causes the master to switch classes to the 107578 attached XTEST pointer device. 107579 107580 Once the fake button is processed, normal event processing continues with 107581 events in the EQ. The master still contains the XTEST classes, causing some 107582 events to be dropped if e.g. the number of valuators of the event in the 107583 queue exceeds the XTEST device's number of valuators. 107584 107585 Example: the EQ contains the following events, processed one-by-one, left to 107586 right. 107587 107588 [DCE (dev)][Btn down][Btn up][Motion][Motion][...] 107589 ^ XkbFakeDeviceButton injects [DCE (XTEST)][Btn up] 107590 107591 Thus the event sequence processed looks like this: 107592 107593 [DCE (dev)][Btn down][Btn up][DCE (XTEST)][Btn up][Motion][Motion][...] 107594 107595 The first DCE causes the master to switch to the device. The button up event 107596 injects a DCE to the XTEST device, causing the following Motion events to be 107597 processed with the master still being on XTEST classes. 107598 107599 This patch post-fixes the injected event sequence with a DCE to restore the 107600 classes of the original slave device, resulting in an event sequence like 107601 this: 107602 [DCE (dev)][Btn down][Btn up][DCE (XTEST)][Btn up][DCE (dev)][Motion][Motion] 107603 107604 Note that this is a simplified description. The event sequence injected by 107605 the PointerKeys code is injected for the master device only and the matching 107606 slave device that caused the injection has already finished processing on 107607 the slave. Furthermore, the injection happens as part of the the XKB layer, 107608 before the unwrapping of the processInputProc takes us into the DIX where 107609 the DCE is actually handled. 107610 107611 Bug reproducible with a device that reports more than 2 valuators. Simply 107612 cause button releases on the device and wait for a "too many valuators" 107613 warning message. 107614 107615 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107616 Acked-by: Daniel Stone <daniel@fooishbar.org> 107617 Reviewed-by: Keith Packard <keithp@keithp.com> 107618 107619commit 6dae7f3792611aace1df0cca63bf50c50d93de43 107620Author: Chris Wilson <chris@chris-wilson.co.uk> 107621Date: Tue Aug 10 19:30:20 2010 +0100 107622 107623 xace: Invalid reference to out-of-scope data. 107624 107625 The callback data passed by reference to the hook was allocated on stack 107626 within the scope of the case statement. The compiler is free to reuse 107627 any of that stack space whilst making the function call so we may end up 107628 passing garbage into the callback. 107629 107630 References: 107631 107632 Bug 18451 - Xorg server 1.5.2 SEGV during XFixesGetCursorImage() 107633 https://bugs.freedesktop.org/show_bug.cgi?id=18451 107634 107635 v2: Drop the unrelated hunk that snuck in when ammending the commit 107636 message. 107637 107638 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 107639 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107640 Signed-off-by: Keith Packard <keithp@keithp.com> 107641 107642commit fbd02046797185715e1a120d52e410ec78fc365f 107643Author: Jon TURNEY <jon.turney@dronecode.org.uk> 107644Date: Thu Aug 5 15:19:20 2010 +0100 107645 107646 Cygwin/X: Fix glxWinCreateDrawable() for API change 107647 107648 Commit 9de0e31746d5f0d9d39d11c94ec3cbc04a9935fc changed the signature 107649 of __GLXScreen's createDrawable method. 107650 107651 Update the glxWinCreateDrawable() function in XWin's GLX provider 107652 appropriately. 107653 107654 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 107655 Tested-by: Colin Harrison <colin.harrison@virgin.net> 107656 Signed-off-by: Keith Packard <keithp@keithp.com> 107657 107658commit ff70848e623920779d20f35d47e9e1f34157de47 107659Author: Keith Packard <keithp@keithp.com> 107660Date: Fri May 21 09:01:43 2010 -0700 107661 107662 Don't let alpha maps recurse in fb. Bug 23581. 107663 107664 Recursive alpha maps (where one picture's alpha map is set to a 107665 picture with an external alpha map) would be all fine and dandy, 107666 except for the case where the client constructs a loop. Detecting this 107667 case when setting the alpha map values would be difficult as any time 107668 an alpha map is set, the server would have to check for the looping 107669 case. 107670 107671 Instead, a far simpler fix is to simply disallow recursive alpha maps 107672 in the rendering code, the Render spec is ambiguous in this area and 107673 allows us to to ignore the recursive case. 107674 107675 Signed-off-by: Keith Packard <keithp@keithp.com> 107676 Reviewed-by: Adam Jackson <ajax@redhat.com> 107677 107678commit 70a94c5b7a42adc0995bf774c44587a0778be0d0 107679Author: Jon TURNEY <jon.turney@dronecode.org.uk> 107680Date: Tue Aug 3 19:49:10 2010 +0100 107681 107682 rootless: fix uninitialized private key assert in non-rootless modes in Cygwin/X 107683 107684 IsFramedWindow() is called from miPaintWindow() if the server has 107685 been built with ROOTLESS defined, irrespective of if RootlessInit() 107686 has ever been called, or not. 107687 107688 Add a check to IsFramedWindow() to check if rootlessWindowPrivateKey 107689 has been registered (as a proxy for checking if the rootless extension 107690 has been initialized) so we don't go on to try to use that key, 107691 triggering an assert. 107692 107693 This bug exposes what appears to be a difference in opinion about 107694 the rootless extension between XQuartz and XWin. XQuartz always 107695 initializes the rootless extension, whereas XWin offers several modes 107696 of operation, and the rootless extension is only used for one of them 107697 107698 That probably means that the all code under compile time guard for 107699 ROOTLESS should be carefully checked that it doesn't also need to be 107700 under a run-time guard 107701 107702 (I've reviewed the other ROOTLESS blocks in dix/events.c and 107703 dix/window.c and they look ok -- keithp) 107704 107705 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 107706 Reviewed-by: Keith Packard <keithp@keithp.com> 107707 Signed-off-by: Keith Packard <keithp@keithp.com> 107708 107709commit 3ab6cd31cbdf8095b2948034fce5fb645422d8da 107710Author: Adam Jackson <ajax@redhat.com> 107711Date: Mon Aug 9 15:20:20 2010 -0400 107712 107713 fonts: Fix refcounting for asynchronous font operations (#3040) 107714 107715 When doing Xinerama, we'll dispatch font ops across all backend screens. 107716 If using a font server (such that some operations can sleep), we'll put 107717 the client to sleep once for each screen, but only wake up once, because 107718 we're trying to keep track of the sleep count in _each_ screen's 107719 closure. 107720 107721 Instead, just ask the core whether the client is already asleep. 107722 107723 Signed-off-by: Adam Jackson <ajax@redhat.com> 107724 Reviewed-by: Keith Packard <keithp@keithp.com> 107725 Signed-off-by: Keith Packard <keithp@keithp.com> 107726 107727commit 35c0dbe4b0a6ab790f4271325b8a0b16894daa8b 107728Author: Gaetan Nadon <memsize@videotron.ca> 107729Date: Sun Aug 8 10:03:08 2010 -0400 107730 107731 doc: add missing .gitignore for Xserver-DTrace 107732 107733 The dtrace doc was recently added in commit 107734 9c171d4aee695ab66e6db1ab92539557bd368cfa 107735 107736 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107737 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 107738 107739commit 05c4fe83fdc20b838fd59658478278dc31a55eb9 107740Author: Alan Coopersmith <alan.coopersmith@oracle.com> 107741Date: Fri Jul 30 17:43:24 2010 -0700 107742 107743 Check HAVE_XMLTO_TEXT before trying to use xmlto to make text files 107744 107745 Reported-by: Matt Turner <mattst88@gmail.com> 107746 Tested-by: Gaetan Nadon <memsize@videotron.ca> 107747 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107748 107749commit a6fb7829ed9bf26c4c2a02c6ed075fb1b17f7b2a 107750Author: Alan Coopersmith <alan.coopersmith@oracle.com> 107751Date: Mon Jul 26 19:55:27 2010 -0700 107752 107753 Xserver-spec: Update ChangeGC prototype, add ChangeGCXIDs 107754 107755 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107756 Reviewed-by: Jamey Sharp <jamey@minilop.net> 107757 107758commit a817271d461e2f95dd7dc62cd1c7d123ce92f555 107759Author: Alan Coopersmith <alan.coopersmith@oracle.com> 107760Date: Thu Jul 22 23:57:57 2010 -0700 107761 107762 Update Xserver-spec for new devPrivates API 107763 107764 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107765 Reviewed-by: Patrick E. Kane <pekane52@gmail.com> 107766 107767commit 40d598a4f84091db743ceef4d60752bb910c3e56 107768Author: Alan Coopersmith <alan.coopersmith@oracle.com> 107769Date: Thu Jul 22 23:57:02 2010 -0700 107770 107771 Correct function name in dixRegisterPrivateKey comments 107772 107773 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107774 Reviewed-by: Patrick E. Kane <pekane52@gmail.com> 107775 107776commit 8d7b7a0d71e0b89321b3341b781bc8845386def6 107777Author: Kristian Høgsberg <krh@bitplanet.net> 107778Date: Thu Jul 29 20:36:25 2010 -0400 107779 107780 Set DamageSetReportAfterOp to true for the damage extension 107781 107782 Change the damage extension reporter to queue up events after we chain 107783 to the wrapped functions. Damage events are typically sent out after 107784 the rendering happens anyway, since we submit batch buffers from the 107785 flush callback chain and then flush client io buffers. Compositing 107786 managers relie on this order, and there is no way we could reliably 107787 provide damage events to clients before the rendering happens anyway. 107788 107789 By queueing up the damage events before the rendering happens, there's 107790 a risk that the client io buffer may overflow and send the damage 107791 events to the client before the driver has even seen the rendering 107792 request. Reporting damage events after the rendering fixes this 107793 corner case and better corresponds with how we expect this to work. 107794 107795 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 107796 Reviewed-by: Keith Packard <keithp@keithp.com> 107797 107798commit c65f610e12f9df168d5639534ed3c2bd40afffc8 107799Author: Kristian Høgsberg <krh@bitplanet.net> 107800Date: Thu Jul 29 18:52:35 2010 -0400 107801 107802 Always call the flush callback chain when we flush client buffers 107803 107804 We were missing the callback in a couple of places. Drivers may use 107805 the flush callback to submit batched up rendering before events (for 107806 example, damage events) are sent out, to ensure that the rendering 107807 has been queued when the client receives the event. 107808 107809 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 107810 Reviewed-by: Keith Packard <keithp@keithp.com> 107811 107812commit 7e0575baf14ec4a89492fd2780f9ab5b9244afbd 107813Author: Matt Turner <mattst88@gmail.com> 107814Date: Tue Jul 27 23:32:36 2010 -0400 107815 107816 ddc: Fix memory leak in GetEDID_DDC1 107817 107818 Mark argument to DDC_checksum as const too. 107819 107820 Signed-off-by: Matt Turner <mattst88@gmail.com> 107821 Reviewed-by: Adam Jackson <ajax@redhat.com> 107822 Signed-off-by: Keith Packard <keithp@keithp.com> 107823 107824commit 750d4e82a0c1161292d24216bcff200cd6647611 107825Author: Jeremy Huddleston <jeremyhu@apple.com> 107826Date: Thu Jul 29 11:39:40 2010 -0700 107827 107828 XQuartz: xpbproxy: Don't take down the whole server on an IO error 107829 107830 Calls pthread_exit to prevent _XIOError from calling exit() 107831 107832 This fixes http://xquartz.macosforge.org/trac/ticket/421 107833 107834 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 107835 107836commit 98f90145d786695ecbc02a667c6ffe7c619dc67e 107837Author: Jeremy Huddleston <jeremyhu@apple.com> 107838Date: Wed Jul 28 18:08:02 2010 -0700 107839 107840 XQuartz: GLX: Don't mangle __GLXDrawable's pDraw 107841 107842 We were incorrectly NULLing out pDraw in __GLXDrawable instead of ours in 107843 __GLXAquaDrawable. (we should refactor to eliminate this redundancy later) 107844 107845 This was causing http://xquartz.macosforge.org/trac/ticket/426 107846 This was benign until commit f0006aa58f6cf7552a239e169ff6e7e4fda532f4 107847 The root cause of this change was fed7ccc481ad1caaa518cafe944c2327a5d0b6c65 107848 107849 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 107850 107851commit 9c171d4aee695ab66e6db1ab92539557bd368cfa 107852Author: Alan Coopersmith <alan.coopersmith@oracle.com> 107853Date: Fri Jun 25 16:52:42 2010 -0700 107854 107855 Add documentation of the Xserver DTrace probes 107856 107857 Mostly pulled together from posts to my blog and the docs posted 107858 at http://people.freedesktop.org/~alanc/dtrace/ and converted to 107859 DocBook. 107860 107861 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107862 Reviewed-by: Patrick E. Kane <pekane52@gmail.com> 107863 107864commit 6b912b08ce65072b2401167fbf2150b99c0ca6b0 107865Author: Alan Coopersmith <alan.coopersmith@oracle.com> 107866Date: Sat Jun 19 23:56:20 2010 -0700 107867 107868 Add name argument to CreateNewResourceType documentation 107869 107870 Reflects API change made in commit 895f40792a during Xorg 1.8 development 107871 107872 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107873 Reviewed-by: Patrick E. Kane <pekane52@gmail.com> 107874 107875commit 44c9350d72b3eda982c642feb45d6648afc626cf 107876Author: Alan Coopersmith <alan.coopersmith@oracle.com> 107877Date: Sat Jun 19 23:40:37 2010 -0700 107878 107879 Use DocBook stylesheets from xorg-sgml-doctools if they're available 107880 107881 Bumps minimum xorg-macros requirement from 1.6 to 1.10 107882 107883 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 107884 Reviewed-by: Patrick E. Kane <pekane52@gmail.com> 107885 107886commit 9fbbff3c0456f1969d45cc957d3260723caf62d7 107887Merge: 95756f410 9ac8e206f 107888Author: Keith Packard <keithp@keithp.com> 107889Date: Wed Jul 21 11:56:39 2010 -0700 107890 107891 Merge remote branch 'whot/for-keith' 107892 107893commit 9ac8e206ffe1016a8bc203261ade6c763a8a4f86 107894Author: Peter Hutterer <peter.hutterer@who-t.net> 107895Date: Thu Jul 15 15:46:15 2010 +1000 107896 107897 xkb: use GetMaster instead of dev->u.master. 107898 107899 Devices that are both pointers and keyboards are not affected by keyboard 107900 changes as their master device is a master pointer, not a master keyboard. 107901 Use GetMaster() instead to ensure devices that are attached to the paired 107902 master pointer device will still be update. 107903 107904 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107905 Reviewed-by: Keith Packard <keithp@keithp.com> 107906 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 107907 107908commit c54f81ba7a58faf37a612bd9a45276bb2922b5d8 107909Author: Peter Hutterer <peter.hutterer@who-t.net> 107910Date: Thu Jul 15 13:24:14 2010 +1000 107911 107912 dix: hack around enter/leave event issues for grabbed devices (#27804) 107913 107914 The current core enter/leave does not cater for device grabs during 107915 enter/leave events. If a window W contains a pointer P1 and a client grabs a 107916 pointer P2, this pointer will not generate enter/leave events inside this 107917 window. 107918 107919 Hack around this by forcing grabbed devices to always send enter/leave 107920 events. 107921 107922 X.Org Bug 27804 <http://bugs.freedesktop.org/show_bug.cgi?id=27804> 107923 107924 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 107925 Reviewed-by: Keith Packard <keithp@keithp.com> 107926 107927commit 95756f410c65a6510a797e94a792b959d45cdb9e 107928Author: Jan Hauffa <hauffa@in.tum.de> 107929Date: Sat Jul 17 12:18:39 2010 -0700 107930 107931 rootless: Adjust the frame size of the native root window in RootlessResizeWindow 107932 107933 If the native root window isn't resized as well, we will likely crash the 107934 next time we draw to the root. On OS X, this can be seen by: 107935 107936 1) Put the display preferences in the menu bar and set X11's preferences so you 107937 can access the menu bar in fullscreen mode 107938 2) Set the resolution of your screen lower than normal. 107939 3) Start X11 in fullscreen mode. The root window will cover the screen as 107940 expected. 107941 4) Use the menu bar to increase the resolution of the display. The root 107942 window will now cover the old area and not the full screen, but 107943 'xwininfo -root' will report the full width. 107944 5) Run 'xsetroot -solid red', and we have the crash you mention above. 107945 107946 Leaving/entering fullscreen after #4 will fix the problem. This is because the 107947 WINREC is erased when we leave fullscreen mode and it is recreated upon 107948 re-entry: 107949 107950 RootlessUpdateRooted(FALSE) 107951 RootlessDisableRoot(screenInfo.screens[0]) 107952 RootlessDestroyFrame (pRoot, winRec); 107953 RootlessUpdateRooted(TRUE) 107954 RootlessEnableRoot(screenInfo.screens[0]) 107955 RootlessEnsureFrame(screenInfo.screens[0]->pRoot) 107956 creates a new WINREC... 107957 107958 Signed-off-by: Jan Hauffa <hauffa@in.tum.de> 107959 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 107960 Acked-By: Jon TURNEY <jon.turney@dronecode.org.uk> 107961 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 107962 107963commit a2c13f0d6548310e3cd115cf486d3e43edf23dcc 107964Author: Keith Packard <keithp@keithp.com> 107965Date: Wed Jul 14 12:57:29 2010 -0700 107966 107967 Bump to version 1.8.99.905 (1.9 RC5) 107968 107969 Signed-off-by: Keith Packard <keithp@keithp.com> 107970 107971commit 0fc02c0bf92f694889589e3648acc08d4684de37 107972Merge: 0540c4606 9d8ec712a 107973Author: Keith Packard <keithp@keithp.com> 107974Date: Tue Jul 13 15:05:36 2010 -0700 107975 107976 Merge remote branch 'jeremyhu/master' 107977 107978commit 0540c46066f938ad5611c56081cfcd8457a9b718 107979Author: Michel Dänzer <daenzer@vmware.com> 107980Date: Tue Jul 13 14:56:53 2010 +0200 107981 107982 EXA: Finish access to pixmap if it's prepared at destruction time. 107983 107984 Previously we assumed every pixmap destroyed during a software fallback was 107985 also created during a software fallback and had access prepared, but that's 107986 not always true. 107987 107988 Fixes a server abort 107989 Reported-by: 邓逸昕 <bupt.dengyixin@gmail.com> 107990 107991 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 107992 Acked-by: Maarten Maathuis <madman2003@gmail.com> 107993 Signed-off-by: Keith Packard <keithp@keithp.com> 107994 107995commit d75e8146c414bfd512ba5dbd4a83acb334bbe19b 107996Author: Keith Packard <keithp@keithp.com> 107997Date: Mon Jul 12 16:01:34 2010 -0700 107998 107999 Unwrap/rewrap EnterVT/LeaveVT completely, Fixes 28998 108000 108001 Because some EnterVT code needs to remove it self from the 108002 call chain, we need to fix all of the wrappers to correctly 108003 unwrap/rewrap during the call chain. This is a follow-on to the fix 108004 for bug 27114 in commit 68a9ee8370e6f9b38218376ac92d5130a5b0ef1e. 108005 108006 Signed-off-by: Keith Packard <keithp@keithp.com> 108007 Tested-by: Jesse Barnes <jesse.barnes@intel.com> 108008 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 108009 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 108010 108011commit 9d8ec712a67ce71ea7408f0626cda7e0fa7c3bac 108012Author: Jeremy Huddleston <jeremyhu@apple.com> 108013Date: Tue Jul 13 08:56:36 2010 -0700 108014 108015 XQuartz: Bump bundle version to 2.6.0 for xorg-server-1.9.0 and X11R7.6 108016 108017 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 108018 108019commit b2b9c458a46e9a41c3c76ffe83a2b580a41d0e90 108020Author: Jan Hauffa <hauffa@in.tum.de> 108021Date: Wed Jun 16 09:25:41 2010 -0700 108022 108023 XQuartz: Remove some dead code. 108024 108025 Signed-off-by: Jan Hauffa <hauffa@in.tum.de> 108026 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 108027 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 108028 108029commit 648d189548530fa23d97d1e8737f89d297f1c443 108030Author: Jeremy Huddleston <jeremyhu@apple.com> 108031Date: Tue Jul 13 08:25:27 2010 -0700 108032 108033 XQuartz: Avoid a crash when mistakenly free()ing in QuartzSetCursor on some configs 108034 108035 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 108036 108037commit 3209b094a3b1466b579e8020e12a4f3fa78a5f3f 108038Author: Jesse Barnes <jbarnes@virtuousgeek.org> 108039Date: Fri Jul 9 10:36:56 2010 -0700 108040 108041 DRI2: re-allocate DRI2 drawable if pixmap serial changes 108042 108043 If a pixmap header is modified or the drawable serial changes, some 108044 aspects of the drawable are likely to have changed so we should 108045 re-allocate the corresponding DRI2 drawable in that case. This is one 108046 way of catching when the root window pixmap changes through xrandr. 108047 108048 Fixes bug https://bugs.freedesktop.org/show_bug.cgi?id=28365. 108049 108050 Reviewed-by: Keith Packard <keithp@keithp.com> 108051 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 108052 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 108053 Signed-off-by: Keith Packard <keithp@keithp.com> 108054 108055commit 02b11509b25686ff7bd567ecb78a435701edc4c2 108056Author: Jesse Barnes <jbarnes@virtuousgeek.org> 108057Date: Fri Jul 9 10:36:12 2010 -0700 108058 108059 miModifyPixmapHeader: always update serialNumber 108060 108061 We should update the serial number even if we just change a single field. 108062 108063 Reviewed-by: Keith Packard <keithp@keithp.com> 108064 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 108065 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 108066 Signed-off-by: Keith Packard <keithp@keithp.com> 108067 108068commit 2307ab5bc9365ebbe04568edb7c7620a23689b70 108069Merge: c65280ce8 fd4f5059f 108070Author: Keith Packard <keithp@keithp.com> 108071Date: Tue Jul 6 23:54:54 2010 -0400 108072 108073 Merge remote branch 'whot/for-keith' 108074 108075commit fd4f5059f08165a726071dc9f1ca877038292f6f 108076Author: Peter Hutterer <peter.hutterer@who-t.net> 108077Date: Tue Jul 6 09:19:09 2010 +1000 108078 108079 dix: purge leftover manual key down bit setting. 108080 108081 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108082 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 108083 108084commit a1afe172559aff010e886cfc2a7a922d4a06c697 108085Author: Peter Hutterer <peter.hutterer@who-t.net> 108086Date: Tue Jul 6 09:16:42 2010 +1000 108087 108088 dix: add aux. functions for button_is_down, set_button_down, set_button_up. 108089 108090 Same as the matching key functions. Buttons, like keys, can have two states 108091 for down/up - one posted, one processed. Posted is set during event 108092 generation (usually in the signal handler). Processed is set during event 108093 processing when the event queue is emptied and events are being delivered to 108094 the client. 108095 108096 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108097 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 108098 108099commit 32473d6bf38c95b2d6d5ddbf583a1e801c6605e4 108100Author: Peter Hutterer <peter.hutterer@who-t.net> 108101Date: Tue Jul 6 08:56:12 2010 +1000 108102 108103 dix: use BitIsOn/SetBit/ClearBit macros for set_key_down helpers. 108104 108105 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108106 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 108107 108108commit c18442908080c9833dfd6bb2ff367945d1892421 108109Author: Peter Hutterer <peter.hutterer@who-t.net> 108110Date: Mon Jul 5 16:47:57 2010 +1000 108111 108112 Xi: use set_key_up/down instead of manual bit handling. 108113 108114 We have the wrappers, use them. 108115 108116 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108117 108118commit 10442ce02b5be7f82b373bee1939e2b523e291d9 108119Author: Peter Hutterer <peter.hutterer@who-t.net> 108120Date: Mon Jul 5 16:54:48 2010 +1000 108121 108122 dix: treat flags as flags, not as value in key_is_down. 108123 108124 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108125 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 108126 108127commit b46ffd25d3f211e91c67bc618ecbd58257939388 108128Author: Peter Hutterer <peter.hutterer@who-t.net> 108129Date: Fri Jul 2 15:25:14 2010 +1000 108130 108131 mi: rename miPointerMoved to miPointerMoveNoEvent. 108132 108133 Having miPointerMove and miPointerMoved is confusing, especially since both 108134 do the same thing bar the event delivery. Also, miPointerMove calls 108135 miPointerMoved which indicates some confusion in the temporal alignment of 108136 cause and effect. 108137 108138 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108139 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 108140 Reviewed-by: Keith Packard <keithp@keithp.com> 108141 108142commit dbd621705a5211540b353af81c4af83c297b74dc 108143Author: Peter Hutterer <peter.hutterer@who-t.net> 108144Date: Fri Jul 2 14:22:03 2010 +1000 108145 108146 mi: De-duplicate some code in mipointer.c 108147 108148 miPointerMoved already has the same code. 108149 108150 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108151 Reviewed-by: Keith Packard <keithp@keithp.com> 108152 108153commit c65280ce8df4836bd7424a90482e8aa00ab6f447 108154Author: Robert Hooker <sarvatt@ubuntu.com> 108155Date: Sat Jun 12 20:23:09 2010 -0400 108156 108157 Increase advertised RENDER protocol minor version to 11 108158 108159 Support for the blend mode operators was added in 108160 0ce42adbf4cff9e7f049d9fc79d588ece5936177 108161 and the requirement was bumped but when things were split off into 108162 include/protocol-versions.h it defined it to 10. render uses 108163 the lower of the client and server advertised versions so it's not 108164 using the new blend mode operators. 108165 108166 Signed-off-by: Robert Hooker <sarvatt@ubuntu.com> 108167 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 108168 Signed-off-by: Keith Packard <keithp@keithp.com> 108169 108170commit 9f0b193acdc29e491b6245390cf9f53b5222e6d3 108171Author: Keith Packard <keithp@keithp.com> 108172Date: Thu Jul 1 09:13:43 2010 -0400 108173 108174 miDbe window priv priv is pre-allocated, don't use dixSetPrivate (bug 28639) 108175 108176 miDbeInit pre-allocates space in each DBE window private private for a 108177 MiDbeWindowPrivPrivRec. miDbeAllocBackBufferName used the 108178 pre-allocated space correctly (simply fetching it instead of 108179 allocating a new piece of memory). However, it then called 108180 dixSetPrivate anyways, which isn't necessary, and (in the new 108181 dixPrivate world) causes an assert failure. 108182 108183 Signed-off-by: Keith Packard <keithp@keithp.com> 108184 Tested-by: Magnus Kessler <Magnus.Kessler@gmx.net> 108185 Reviewed-by: Magnus Kessler <Magnus.Kessler@gmx.net> 108186 108187commit a94cb400d15b8c78dc04148cbd8db8e5ec8364b5 108188Author: Keith Packard <keithp@keithp.com> 108189Date: Thu Jul 1 09:11:36 2010 -0400 108190 108191 Delete unused miDbe screen private private datatype 108192 108193 MiDbeScreenPrivPrivRec is not used in the server. Remove it, along 108194 with the MI_DBE_SCREEN_PRIV_PRIV macro that tried to use it. 108195 108196 Signed-off-by: Keith Packard <keithp@keithp.com> 108197 Reviewed-by: Magnus.Kessler <Magnus.Kessler@gmx.net> 108198 108199commit f0fcffe55f280add5e4db2f5e9198a48c6f1b015 108200Author: Alan Coopersmith <alan.coopersmith@oracle.com> 108201Date: Thu Jun 3 19:00:54 2010 -0700 108202 108203 Update the sprite immediately when moving it with MouseKeys 108204 108205 Fix for OpenSolaris bug 6949755: Mouse Keys are ununusable 108206 and possibly https://bugs.freedesktop.org/show_bug.cgi?id=24856 108207 108208 Ensures waitForUpdate is False before calling SetCursorPosition. 108209 Normally waitForUpdate is False when SilkenMouse is active, True 108210 when it's not. When it's True, the mouse cursor position on 108211 screen is not updated immediately. 108212 108213 This is more critical on Solaris, since we disabled SigIO, thus in turn 108214 disable SilkenMouse, due to the SSE2 vs. signal handler issues described in 108215 Sun bugs 6849925, 6859428, and 6879897. 108216 108217 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 108218 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 108219 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108220 108221commit 3d9079b898c432a87f9b95c1f39a85f660bf0858 108222Author: Alan Coopersmith <alan.coopersmith@oracle.com> 108223Date: Thu Jun 3 19:00:53 2010 -0700 108224 108225 Add API to update setting of waitForUpdate screen private in miPointer 108226 108227 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 108228 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 108229 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108230 108231commit 69b2b5c85ec079ef49f84722daa5f148cedc2e1b 108232Merge: b90faa715 143278583 108233Author: Keith Packard <keithp@keithp.com> 108234Date: Thu Jul 1 23:46:53 2010 -0400 108235 108236 Merge remote branch 'whot/for-keith' 108237 108238commit b90faa71567c4461b28515756ba5c1e6286dda16 108239Author: Keith Packard <keithp@keithp.com> 108240Date: Thu Jul 1 23:46:27 2010 -0400 108241 108242 Revert "xkb: merge lockedPtrButtons state from all attached SDs." 108243 108244 Preparing to merge Peter's branch. 108245 108246 This reverts commit 6052710670953b43b4fff5d101b727163fcb1187. 108247 108248commit 9fb0785449b287ba1998e08613b3c2102ec24842 108249Author: Keith Packard <keithp@keithp.com> 108250Date: Thu Jul 1 23:45:50 2010 -0400 108251 108252 Revert "Revert "dix: use the event mask of the grab for TryClientEvents."" 108253 108254 Preparing to merge Peter's branch. 108255 108256 This reverts commit 018c878e9495b21146c8f38617fdd1bf6d8cc73b. 108257 108258commit 48cac27870992f6bde2c48429ff03c0a7606d5c1 108259Author: James Jones <jajones@nvidia.com> 108260Date: Fri Jun 18 17:28:15 2010 -0700 108261 108262 Cast void* to pointer* to appease some compilers. 108263 108264 When this privates.h is included in C++ builds, the compiler 108265 complains about implicitly casting void* to void**. This small 108266 patch fixes that up. 108267 108268 Signed-off-by: James Jones <jajones@nvidia.com> 108269 Reviewed-by: Keith Packard <keithp@keithp.com> 108270 Signed-off-by: Keith Packard <keithp@keithp.com> 108271 108272commit 14327858391ebe929b806efb53ad79e789361883 108273Author: Peter Hutterer <peter.hutterer@who-t.net> 108274Date: Thu Jul 1 12:44:57 2010 +1000 108275 108276 xkb: release XTEST pointer buttons on physical releases. (#28808) 108277 108278 If a button release event is posted for the MD pointer, post a release event 108279 through the matching XTEST device. This way, a client who posts a button 108280 press through the XTEST extension cannot inadvertedly lock the button. 108281 108282 This behaviour is required for historical reasons, until server 1.7 the core 108283 pointer would release a button press on physical events, regardless of the 108284 XTEST state. Clients seem to rely on this behaviour, causing seemingly stuck 108285 grabs. 108286 108287 The merged behaviour is kept for multiple keyboard PointerKey events, if two 108288 physical keyboards hold the button down as a result of PointerKey actions, 108289 the button is not released until the last keyboard releases the button. 108290 108291 X.Org Bug 28808 <http://bugs.freedesktop.org/show_bug.cgi?id=28808> 108292 108293 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108294 108295commit a71dbc03e65cf7b0654a6eca93ce0bf6a1711ffa 108296Author: Keith Packard <keithp@keithp.com> 108297Date: Thu Jul 1 08:27:05 2010 -0400 108298 108299 Bump to version 1.8.99.904 (1.9 RC4) 108300 108301 Signed-off-by: Keith Packard <keithp@keithp.com> 108302 108303commit 5ea497fc32202ff19c2b7d8ef4eba08889108a87 108304Author: Tiago Vignatti <tiago.vignatti@nokia.com> 108305Date: Tue Jun 29 16:08:02 2010 +0300 108306 108307 xfree86: configure: remove vendor and card name matching rules 108308 108309 Although vendor and board naming are used to create the configure file, the 108310 server doesn't actually use it when fetching such file and probing devices. 108311 108312 Reported-by: Richard Barnette <jrbarnette@chromium.org> 108313 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 108314 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 108315 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 108316 Tested-by: Richard Barnette <jrbarnette@chromium.org> 108317 Signed-off-by: Keith Packard <keithp@keithp.com> 108318 108319commit 90fd9e40b59e69333d250a795998f44dc3a5dc0c 108320Author: Tiago Vignatti <tiago.vignatti@nokia.com> 108321Date: Tue Jun 29 16:08:01 2010 +0300 108322 108323 xfree86: pci: remove superfluous vendor and card name logging 108324 108325 X server suffers in startup time when relying on the pciaccess's linear search 108326 to fetch vendor and video device name from PCI ID file (when existent). Such 108327 names are only used to write the log, which may be superfluous. This 108328 information often is provided by the drivers or likewise users can get the it 108329 using external tools like lspci or scanpci. 108330 108331 This patch remove the references of those functions from X start up. 108332 108333 Reported-by: Richard Barnette <jrbarnette@chromium.org> 108334 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 108335 Tested-by: Richard Barnette <jrbarnette@chromium.org> 108336 Reviewed-by: James Cloos <cloos@jhcloos.com> 108337 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 108338 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 108339 Signed-off-by: Keith Packard <keithp@keithp.com> 108340 108341commit 339f62b1bfadb0ee77d67e351f4e30f5d5e9625f 108342Author: Peter Hutterer <peter.hutterer@who-t.net> 108343Date: Tue Jun 29 15:24:51 2010 +1000 108344 108345 xkb: emulate PointerKeys events only on the master device. 108346 108347 This patch replicates the behaviour for button events. Only generate a 108348 PointerKeys motion event on the master device, not on the slave device. 108349 Fixes the current issue of PointerKey motion events generating key events as 108350 well. 108351 108352 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108353 108354commit 69ac909878ef80bb74c4a9ca4150eda66debd754 108355Author: Peter Hutterer <peter.hutterer@who-t.net> 108356Date: Tue Jun 29 12:12:53 2010 +1000 108357 108358 xkb: merge lockedPtrButtons state from all attached SDs. 108359 108360 Problem: 108361 lockedPtrButtons keeps the state of the buttons locked by a PointerKeys button 108362 press. Unconditionally clearing the bits may cause stuck buttons in this 108363 sequence of events: 108364 108365 1. type Shift + NumLock to enable PointerKeys 108366 2. type 0/Ins on keypad to emulate Button 1 press 108367 → button1 press event to client 108368 3. press and release button 1 on physical mouse 108369 → button1 release event to client 108370 108371 Button 1 on the MD is now stuck and cannot be released. 108372 108373 Cause: 108374 XKB PointerKeys button events are posted through the XTEST pointer device. 108375 Once a press is generated, the XTEST device's button is down. The DIX merges 108376 the button state of all attached SDs, hence the MD will have a button down 108377 while the XTEST device has a button down. 108378 108379 PointerKey button events are only generated on the master device to avoid 108380 duplicate events (see XkbFakeDeviceButton()). If the MD has the 108381 lockedPtrButtons bit cleared by a release event on a physical device, no 108382 such event is generated when a keyboard device triggers the PointerKey 108383 ButtonRelease trigger. Since the event - if generated - is posted through 108384 the XTEST pointer device, lack of a generated ButtonRelease event on the 108385 XTEST pointer device means the button is never released, resulting in the 108386 stuck button observed above. 108387 108388 Solution: 108389 This patch merges the MD's lockedPtrButtons with the one of all attached 108390 slave devices on release events. Thus, as long as one attached keyboard has 108391 a lockedPtrButtons bit set, this bit is kept in the MD. Once a PointerKey 108392 button is released on all keyboards, the matching release event is emulated 108393 from the MD through the XTEST pointer device, thus also releasing the button 108394 in the DIX. 108395 108396 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108397 108398commit 09645864f5a52882eee51c801b3e610d683e7147 108399Author: Peter Hutterer <peter.hutterer@who-t.net> 108400Date: Tue Jun 29 13:49:27 2010 +1000 108401 108402 xkb: Mark switch case fallthrough with comment. 108403 108404 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108405 108406commit c7330ecb5d28d7a92d24feb289f7f1812ce055a4 108407Author: Peter Hutterer <peter.hutterer@who-t.net> 108408Date: Wed Jun 30 13:23:14 2010 +1000 108409 108410 dix: fix up erroneous error message. 108411 108412 (WW) Device 'device name' has 36 axes, only using first 36. 108413 does seem a bit silly. 108414 108415 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108416 108417commit dbf249ec6638f0a8dfa4c2286099845aafc8ac88 108418Author: Peter Hutterer <peter.hutterer@who-t.net> 108419Date: Tue Jun 29 10:43:51 2010 +1000 108420 108421 xkb: remove now obsolete comment. 108422 108423 Looks like nothing broke from removing the hardcoded CoreProcessPointerEvent 108424 call. Whoop. Di. Doo. 108425 108426 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108427 108428commit f028e70ca714f6956e41754f132cb9b8a1e8db63 108429Author: Keith Packard <keithp@keithp.com> 108430Date: Wed Jun 30 08:33:55 2010 -0700 108431 108432 Initialize dev privates before using any 108433 108434 Initializing the dev privates code after allocating the server client 108435 dev privates would cause the memory leak check to fire at server 108436 startup or reset. 108437 108438 Signed-off-by: Keith Packard <keithp@keithp.com> 108439 Acked-by: Daniel Stone <daniel@fooishbar.org> 108440 108441commit 433d0851cd3e61d841ff374ee0a0f052d5907029 108442Author: Julien Cristau <jcristau@debian.org> 108443Date: Fri Jun 25 11:34:44 2010 +0100 108444 108445 Xephyr: fix Xv adaptor capability tests 108446 108447 Signed-off-by: Julien Cristau <jcristau@debian.org> 108448 Reviewed-by: Keith Packard <keithp@keithp.com> 108449 Signed-off-by: Keith Packard <keithp@keithp.com> 108450 108451commit 9626eedebf620559652ffb1fefa82b5d659e57be 108452Author: Julien Cristau <jcristau@debian.org> 108453Date: Thu Jun 24 18:45:11 2010 +0100 108454 108455 configure: bail if Xephyr was requested but its dependencies are missing 108456 108457 Signed-off-by: Julien Cristau <jcristau@debian.org> 108458 Reviewed-by: Jamey Sharp <jamey@minilop.net> 108459 Signed-off-by: Keith Packard <keithp@keithp.com> 108460 108461commit e6531ae9d5bdd37775e921268366fb76056e22d5 108462Author: Julien Cristau <jcristau@debian.org> 108463Date: Thu Jun 24 18:45:10 2010 +0100 108464 108465 Xephyr: fix memory leak in XF86DRIOpenConnection 108466 108467 The allocated bus id string was not being freed. 108468 108469 Signed-off-by: Julien Cristau <jcristau@debian.org> 108470 Reviewed-by: Jamey Sharp <jamey@minilop.net> 108471 Signed-off-by: Keith Packard <keithp@keithp.com> 108472 108473commit 390a8466dd1914f4786b811ff8454f6e0c4b6b04 108474Author: Ville Syrjälä <ville.syrjala@nokia.com> 108475Date: Mon Jun 28 23:26:48 2010 +0300 108476 108477 xfree86/modes: Allow the driver to specify initial rotation 108478 108479 When the "Rotate" option isn't specified allow the driver to specify 108480 the initial rotation mode. This way the driver can choose to retain 108481 the same settings that were used by software that was used prior to 108482 starting X. 108483 108484 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 108485 Reviewed-by: Keith Packard <keithp@keithp.com> 108486 Signed-off-by: Keith Packard <keithp@keithp.com> 108487 108488commit 6052710670953b43b4fff5d101b727163fcb1187 108489Author: Keith Packard <keithp@keithp.com> 108490Date: Wed Jun 30 08:21:04 2010 -0700 108491 108492 xkb: merge lockedPtrButtons state from all attached SDs. 108493 108494 Problem: 108495 lockedPtrButtons keeps the state of the buttons locked by a PointerKeys button 108496 press. Unconditionally clearing the bits may cause stuck buttons in this 108497 sequence of events: 108498 108499 1. type Shift + NumLock to enable PointerKeys 108500 2. type 0/Ins on keypad to emulate Button 1 press 108501 → button1 press event to client 108502 3. press and release button 1 on physical mouse 108503 → button1 release event to client 108504 108505 Button 1 on the MD is now stuck and cannot be released. 108506 108507 Cause: 108508 XKB PointerKeys button events are posted through the XTEST pointer device. 108509 Once a press is generated, the XTEST device's button is down. The DIX merges 108510 the button state of all attached SDs, hence the MD will have a button down 108511 while the XTEST device has a button down. 108512 108513 PointerKey button events are only generated on the master device to avoid 108514 duplicate events (see XkbFakeDeviceButton()). If the MD has the 108515 lockedPtrButtons bit cleared by a release event on a physical device, no 108516 such event is generated when a keyboard device triggers the PointerKey 108517 ButtonRelease trigger. Since the event - if generated - is posted through 108518 the XTEST pointer device, lack of a generated ButtonRelease event on the 108519 XTEST pointer device means the button is never released, resulting in the 108520 stuck button observed above. 108521 108522 Solution: 108523 This patch merges the MD's lockedPtrButtons with the one of all attached 108524 slave devices on release events. Thus, as long as one attached keyboard has 108525 a lockedPtrButtons bit set, this bit is kept in the MD. Once a PointerKey 108526 button is released on all keyboards, the matching release event is emulated 108527 from the MD through the XTEST pointer device, thus also releasing the button 108528 in the DIX. 108529 108530 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108531 Reviewed-by: Keith Packard <keithp@keithp.com> 108532 Signed-off-by: Keith Packard <keithp@keithp.com> 108533 108534commit 28e33ae6f69f716ece5d68e63fc52557236c5f6e 108535Author: Jesse Barnes <jbarnes@virtuousgeek.org> 108536Date: Wed Jun 30 07:59:04 2010 -0700 108537 108538 OS support: fix writeable client vs IgnoreClient behavior 108539 108540 When ResetCurrentRequest is called, or IgnoreClient is called when a 108541 client has input pending, IgnoredClientsWithInput will be set. However, 108542 a subsequent IgnoreClient request will clear the client fd from that fd 108543 set, potentially causing the client to hang. 108544 108545 So add an Ignore/Attend count, and only apply the ignore logic on the 108546 first ignore and the attend logic on the last attend. This is 108547 consistent with the comments for these functions; callers must pair 108548 them. 108549 108550 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=27035. 108551 108552 Reviewed-by: Keith Packard <keithp@keithp.com> 108553 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 108554 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 108555 Signed-off-by: Keith Packard <keithp@keithp.com> 108556 108557commit 018c878e9495b21146c8f38617fdd1bf6d8cc73b 108558Author: Peter Hutterer <peter.hutterer@who-t.net> 108559Date: Fri Jun 25 09:48:10 2010 +1000 108560 108561 Revert "dix: use the event mask of the grab for TryClientEvents." 108562 108563 Behaviour of earlier X servers was to deliver the ButtonPress event 108564 unconditionally, regardless of the actual event mask being set. Thus, a 108565 GrabButton event will always deliver the button press event, a GrabKey 108566 always the key press event, etc. Same goes for XI and XI2. 108567 108568 Reproducible with a simple client requesting a button grab in the form of: 108569 XGrabButton(dpy, AnyButton, AnyModifier, win, True, ButtonReleaseMask, 108570 GrabModeAsync, GrabModeAsync, None, None); 108571 108572 On servers before MPX/XI2, the client will receive a button press and 108573 release event. On current servers, the client receives only the release. 108574 Clients that expect the press event to be delivered unconditionally may 108575 leave the user with a stuck grab. 108576 108577 XTS test results for XGrabButton are identical with and without this patch. 108578 108579 This reverts commit 48585bd1e3e98db0f3df1ecc68022510216e00cc. 108580 108581 Conflicts: 108582 108583 dix/events.c 108584 108585 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108586 Reviewed-by: Keith Packard <keithp@keithp.com> 108587 Signed-off-by: Keith Packard <keithp@keithp.com> 108588 108589commit 1884db430a5680e37e94726dff46686e2218d525 108590Author: Peter Hutterer <peter.hutterer@who-t.net> 108591Date: Thu Jun 24 12:52:53 2010 +1000 108592 108593 Revert "dix: use the event mask of the grab for TryClientEvents." 108594 108595 Behaviour of earlier X servers was to deliver the ButtonPress event 108596 unconditionally, regardless of the actual event mask being set. This is 108597 documented in the protocol: 108598 "This request establishes a passive grab. In the future, the pointer is 108599 actively grabbed as described in GrabPointer, the last-pointer-grab time is 108600 set to the time at which the button was pressed (as transmitted in the 108601 ButtonPress event), and the ButtonPress event is reported if all of the 108602 following conditions are true: 108603 <list of conditions, event mask is not one of them>" 108604 108605 Thus, a GrabButton event will always deliver the button press event, a 108606 GrabKey always the key press event, etc. Same goes for XI and XI2. 108607 108608 Reproducible with a simple client requesting a button grab in the form of: 108609 XGrabButton(dpy, AnyButton, AnyModifier, win, True, ButtonReleaseMask, 108610 GrabModeAsync, GrabModeAsync, None, None); 108611 108612 On servers before MPX/XI2, the client will receive a button press and 108613 release event. On current servers, the client receives only the release. 108614 Clients that expect the press event to be delivered unconditionally. 108615 108616 XTS Xlib13 XGrabButton 5/39 now passes. 108617 108618 This reverts commit 48585bd1e3e98db0f3df1ecc68022510216e00cc. 108619 Effectively reverts commit 1c612acca8568fcdf9761d23f112adaf4d496f1b as well, 108620 the code introduced with 1c612 is not needed anymore. 108621 108622 Conflicts: 108623 108624 dix/events.c 108625 108626 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108627 Acked-by: Daniel Stone <daniel@fooishbar.org> 108628 Reviewed-by: Keith Packard <keithp@keithp.com> 108629 108630commit 3b3c77b87070ddcdbb2acb114a81628485e7a129 108631Author: Tiago Vignatti <tiago.vignatti@Nokia.com> 108632Date: Wed Jun 23 15:18:04 2010 +0300 108633 108634 configure: don't check xfont always for each server 108635 108636 It's already defined inside REQUIRED_LIBS and all DDX are getting it already. 108637 No semantical changes. 108638 108639 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 108640 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 108641 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 108642 Signed-off-by: Keith Packard <keithp@keithp.com> 108643 108644commit 75536ee80595c79bba95a1fb6844126ee08486d4 108645Author: Tiago Vignatti <tiago.vignatti@nokia.com> 108646Date: Wed Jun 23 15:18:03 2010 +0300 108647 108648 dix: use one single function to register fpe fonts 108649 108650 X server doesn't need to understand fpe internals, so use 108651 register_fpe_functions from libXfont. 108652 108653 It's required to get new version of libXfont, therefore adjust it to be passed 108654 to autoconf. 108655 108656 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 108657 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 108658 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 108659 Signed-off-by: Keith Packard <keithp@keithp.com> 108660 108661commit a68a101a730df7f73421555e381a50b074607c5e 108662Author: Keith Packard <keithp@keithp.com> 108663Date: Tue Jun 22 12:13:26 2010 -0700 108664 108665 Bump to 1.8.99.903 -- 1.9 RC3 108666 108667 Signed-off-by: Keith Packard <keithp@keithp.com> 108668 108669commit 2c0159d21788b16607e6ebe4571b57fe78545c2a 108670Author: Kristian Høgsberg <krh@bitplanet.net> 108671Date: Mon Jun 14 09:25:22 2010 -0400 108672 108673 list.h: Fix list_for_each_entry_safe() 108674 108675 Can't use next as a macro argument since we're accessing the .next field 108676 of struct list. 108677 108678 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 108679 Reviewed-by: Keith Packard <keithp@keithp.com> 108680 Signed-off-by: Keith Packard <keithp@keithp.com> 108681 108682commit e2918b67395a48397bf7db545584cc2d4f4e7ccf 108683Author: Keith Packard <keithp@keithp.com> 108684Date: Tue Jun 22 12:11:20 2010 -0700 108685 108686 Revert "mi: Remove unused overlay support" 108687 108688 This reverts commit 9120e58556014e514eef9dfe43cdea4b74dc66ae. 108689 108690 Whoops, please revert this patch -- overlay is in use in nvidia drivers, 108691 and it's too late in release cycle to remove it. 108692 108693 I feel really sorry that I kept this patch in my tree. I will submit 108694 another one, removing only XAA overlay hooks which are not used at all. 108695 108696commit 7673451ade46ebacfdfc02a5190149210198fdd8 108697Author: Keith Packard <keithp@keithp.com> 108698Date: Tue Jun 22 11:49:30 2010 -0700 108699 108700 Bump to 1.8.99.902 -- 1.9 RC2 108701 108702 Signed-off-by: Keith Packard <keithp@keithp.com> 108703 108704commit 8b65f2edb407521e0ac0e207e5d5358986bd46da 108705Author: Rami Ylimäki <ext-rami.ylimaki@nokia.com> 108706Date: Tue Jun 22 14:57:20 2010 +0300 108707 108708 record: Prevent a crash on recording client disconnect. 108709 108710 Execute the following steps to reproduce the issue. 108711 108712 1. Run at least two recording clients simultaneously. 108713 $ cnee --record --request-range 1-127 & 108714 $ cnee --record --request-range 1-127 & 108715 2. Kill the recording clients. 108716 $ killall cnee 108717 3. Give X server something to do so that the clients are closed. 108718 $ xinput list 108719 $ xinput list 108720 108721 As a result RecordUninstallHooks accesses NullClient, because 108722 RecordAClientStateChange doesn't clean the recording clients up 108723 properly. 108724 108725 Fix RecordUninstallHooks to fail locally on an assertion instead of 108726 much later in privates code, if NullClient is still accessed because 108727 of some other bug. Fix RecordAClientStateChange to iterate through all 108728 contexts so that modifications of the iterated array during iteration 108729 don't cause contexts to be skipped. 108730 108731 Signed-off-by: Rami Ylimäki <ext-rami.ylimaki@nokia.com> 108732 Reviewed-by: Keith Packard <keithp@keithp.com> 108733 Signed-off-by: Keith Packard <keithp@keithp.com> 108734 108735commit 660f6ab5494a728c3ca7ba00c305e9ff06c8ecb2 108736Author: Simon Farnsworth <simon.farnsworth@onelan.com> 108737Date: Tue Jun 22 10:13:30 2010 +0100 108738 108739 Don't crash when asked if a client that has disconnected was local 108740 108741 ProcDRI2Dispatch uses LocalClient to determine if it's safe to respond 108742 to a client that has made DRI2 requests which aren't sensible for 108743 remote clients (anything but version). When the client has disappeared 108744 mid-request stream (e.g. as a result of a kill -9, or a client-side 108745 bug), LocalClient causes the X server to follow suit, as 108746 ((OsCommPtr)client->osPrivate)->trans_conn is NULL at this point. 108747 108748 The simple and obvious fix is to just return "not local" when 108749 trans_conn is NULL, which fixes the crash I was seeing; however Keith 108750 Packard pointed out that just checking trans_conn isn't enough; 108751 quoting Keith: 108752 108753 "This looks almost right to me -- I reviewed the os code to see when 108754 _XSERVTransClose is called (which is what frees the trans_conn data) and 108755 found that every place which called that immediately set trans_conn to 108756 NULL, except for the call in CloseDownFileDescriptor which is only 108757 called from CloseDownConnection and which is immediately followed by 108758 freeing the OsCommRec and setting client->osPrivate to NULL. So, I'd 108759 suggest checking client->osPrivate in addition to the above check." 108760 108761 Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.com> 108762 Reviewed-by: Keith Packard <keithp@keithp.com> 108763 Signed-off-by: Keith Packard <keithp@keithp.com> 108764 108765commit 82d41ada993d8cbdcdfea878d1a5b031afe4e593 108766Author: Dave Airlie <airlied@redhat.com> 108767Date: Mon Jun 21 16:33:55 2010 +1000 108768 108769 composite: fix freeing of old pixmap until after move/resize/cbw (bug 28345) 108770 108771 The fixes for the composite reallocation failure, were freeing the oldpixmap 108772 straight after reallocating the new one, however this led to some wierd 108773 effects in e16 compositing, and I think in a few other places. This patch 108774 moves the freeing of the pixmap into the post wrapped stage. I'm not sure if 108775 we are actually better off breaking ABI and adding another callback from the 108776 ConfigureWindow to composite to make sure the old pixmap is always freed, 108777 but this should be satisfactory as we should always hit one of the freeing 108778 paths or else its a bug in the DIX. 108779 108780 bug: https://bugs.freedesktop.org/show_bug.cgi?id=28435 108781 108782 Reported-by: Andrew Randrianasulu <randrik@mail.ru> 108783 Signed-off-by: Dave Airlie <airlied@redhat.com> 108784 Reviewed-by: Keith Packard <keithp@keithp.com> 108785 Signed-off-by: Keith Packard <keithp@keithp.com> 108786 108787commit c6bc52cb6663138d1273447cb7661fa6e958f539 108788Author: Dave Airlie <airlied@redhat.com> 108789Date: Mon Jun 21 14:08:50 2010 +1000 108790 108791 dri2: avoid crash with old dri drivers. 108792 108793 I built latest server and copied over a mesa 7.8 dri2 driver to use with it, 108794 and it crashed in here starting compiz. Looks like we need to validate the 108795 flush extension is registered before calling invalidate. 108796 108797 Signed-off-by: Dave Airlie <airlied@redhat.com> 108798 Reviewed-by: Michel Dänzer <michel@daenzer.net> 108799 Reviewed-by: Keith Packard <keithp@keithp.com> 108800 Signed-off-by: Keith Packard <keithp@keithp.com> 108801 108802commit 80d1a548d6ce73c2ff097536c1bc7044bf74965d 108803Author: Dave Airlie <airlied@redhat.com> 108804Date: Mon Jun 21 10:05:08 2010 +1000 108805 108806 rotation: fix cursor and overlap of one pixel. 108807 108808 Commit 77c7a64e8885696665556c9fbcb3cffb552e367a was introduced to fix 108809 a cursor off by one on Intel hw, however it also move the whole crtc 108810 into an off by one position and you could see gnom-eshell overlapping. 108811 108812 This commit reverts that and instead fixes the cursor hotspot 108813 translation to work like pixman does. We add 0.5 to the cursor vector 108814 before translating, and floor the value afterwards. 108815 108816 Thanks to Soeren (ssp) for pointing out where the real problem was 108817 after explaning how pixman translates points. 108818 108819 Signed-off-by: Dave Airlie <airlied@redhat.com> 108820 Reviewed-by: Keith Packard <keithp@keithp.com> 108821 Signed-off-by: Keith Packard <keithp@keithp.com> 108822 108823commit 32c706c4ffd7433dbfc79dba8785b1510d2f053f 108824Author: Gaetan Nadon <memsize@videotron.ca> 108825Date: Sun Jun 20 20:37:26 2010 -0400 108826 108827 config: declare xserver private dependencies in xorg-server.pc 108828 108829 Any module (drivers) depending on xserver also depends on some of the 108830 server private dependencies. Any driver including xf86.h depends on 108831 xext, kbproto, inputproto and randr. 108832 108833 These dependencies are in separate packages, so anything can happen, 108834 removal, wrong version, etc... and the driver fails during compilation. 108835 Having the private dependencies declared will ensure all packages the 108836 server depends on are present and at the correct version. 108837 108838 Currently each module attempts to check for server dependencies with 108839 various degrees of accuracy. With this patch, the driver will only need 108840 to check for its own explicit dependencies. 108841 108842 Now that xproto is included in Requires.private it is removed from Requires. 108843 All the cflags from both Requires and Requires.private are returned 108844 to caller to pkg-config. 108845 108846 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 108847 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 108848 Signed-off-by: Keith Packard <keithp@keithp.com> 108849 108850commit 37734ba79ca3a9bad1c1e29d89710c8f9c299210 108851Author: Colin Harrison <colin.harrison@virgin.net> 108852Date: Wed Jun 16 18:07:04 2010 +0100 108853 108854 XWin: Fixes for devPrivates API change 108855 108856 Fix some typos in devPrivates API changes in XWin code. 108857 Move allocation of private keys as it's no longer valid to do them during OsVendorInit() 108858 108859 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 108860 Reviewed-by: Jamey Sharp <jamey@minilop.net> 108861 Signed-off-by: Keith Packard <keithp@keithp.com> 108862 108863commit f7dae972aac28c5792566e997c40254e787d246a 108864Author: Ville Syrjälä <ville.syrjala@nokia.com> 108865Date: Tue Jun 15 23:25:34 2010 +0300 108866 108867 xv: Don't send port notify when SetPortAttribute fails 108868 108869 Currently a port notify event is sent even if SetPortAttribute 108870 fails. Furthermore the value field in the event will contain the 108871 value that was specified in the failed request. So any client 108872 interested in the actual value of the attribute will have to 108873 double check the current value with GetPortAttribute after 108874 receiving a port notify event. 108875 108876 Fix the problem by sending port notifications only when 108877 SetPortAttribute succeeds. 108878 108879 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 108880 Reviewed-by: Keith Packard <keithp@keithp.com> 108881 Signed-off-by: Keith Packard <keithp@keithp.com> 108882 108883commit c20166739603f86888f28ca2c65f5cf474d0d8c9 108884Author: Mikhail Gusarov <dottedmag@dottedmag.net> 108885Date: Sat Jun 12 01:05:07 2010 +0700 108886 108887 os: Remove unused dev_tty_from_init variable 108888 108889 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 108890 Reviewed-by: Jamey Sharp <jamey@minilop.net> 108891 108892commit 2e15f1903dac8473285ad2937c079a08cf7c5d31 108893Author: Mikhail Gusarov <dottedmag@dottedmag.net> 108894Date: Sat Jun 12 00:53:38 2010 +0700 108895 108896 os: Remove unused pread/pwrite/lockit functions 108897 108898 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 108899 Reviewed-by: Jamey Sharp <jamey@minilop.net> 108900 108901commit 9120e58556014e514eef9dfe43cdea4b74dc66ae 108902Author: Mikhail Gusarov <dottedmag@dottedmag.net> 108903Date: Sat Jun 12 00:30:27 2010 +0700 108904 108905 mi: Remove unused overlay support 108906 108907 The only reference to it in server and drivers is in XAA overlay code which 108908 would segfault as no miInitOverlay is called ever. No segfaults were observed 108909 "in wild", so XAA overlay is probably also unused. 108910 108911 XAA code is modified to act as if miOverlayCopyUnderlay always returned false, 108912 because XAACopyWindow8_32 could only set doUnderlay to true if it's called from 108913 miOverlayMoveWindow or miOverlayResizeWindow, which can only be called if 108914 miInitOverlay has hooked those functions, and no driver (on fd.o) or server code 108915 calls that. 108916 108917 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 108918 Reviewed-by: Jamey Sharp <jamey@minilop.net> 108919 108920commit 91b5aadbdfd9d05ca1ffdeb443e602ecdba1e04d 108921Author: Mikhail Gusarov <dottedmag@dottedmag.net> 108922Date: Fri Jun 11 20:20:00 2010 +0700 108923 108924 mi: do not use X11/extensions/shape.h header from libXext 108925 108926 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 108927 Tested-by: Dan Nicholson <dbn.lists@gmail.com> 108928 108929commit b8615d592700b7be319c04cc0563fdeb5a266534 108930Author: Alan Coopersmith <alan.coopersmith@oracle.com> 108931Date: Sat Jun 12 08:19:16 2010 -0700 108932 108933 Don't coredump on "X -showopts" (bug 25874) 108934 108935 Don't try walking the xf86ConfigLayout.screens table if it's empty 108936 https://bugs.freedesktop.org/show_bug.cgi?id=25874 108937 108938 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 108939 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 108940 Signed-off-by: Keith Packard <keithp@keithp.com> 108941 108942commit 32fd57509c3e06f63ac6b0ad607767fd2b688e5a 108943Author: Rami Ylimäki <ext-rami.ylimaki@nokia.com> 108944Date: Fri Jun 18 12:40:58 2010 +0300 108945 108946 record: Register client private to prevent crash. 108947 108948 Running the following command causes an assertion to fail: 108949 cnee --record --request-range 1-127 108950 108951 3 in dixGetPrivateAddr at ../../include/privates.h:122 108952 4 in dixGetPrivate at ../../include/privates.h:137 108953 5 in dixLookupPrivate at ../../include/privates.h:167 108954 6 in RecordInstallHooks at ../../record/record.c:911 108955 7 in ProcRecordEnableContext at ../../record/record.c:2377 108956 8 in ProcRecordDispatch at ../../record/record.c:2571 108957 9 in Dispatch at ../../dix/dispatch.c:432 108958 10 in main at ../../dix/main.c:289 108959 108960 Signed-off-by: Rami Ylimäki <ext-rami.ylimaki@nokia.com> 108961 Reviewed-by: Oliver McFadden <oliver.mcfadden@nokia.com> 108962 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 108963 Signed-off-by: Keith Packard <keithp@keithp.com> 108964 108965commit 75beadd766fed7b12a76e59e57c244e297c2d2cb 108966Author: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 108967Date: Sun Jun 13 18:05:26 2010 +0200 108968 108969 DRI2/xserver: Don't hang in glXSwapBuffers if drawable moves between crtc's (bug 28383) 108970 108971 Detect if a drawable has been moved from an original crtc to a new crtc 108972 with a lower current vblank count than the original crtc inbetween 108973 glXSwapBuffers() calls. Reinitialize drawable's last_swap_target 108974 before scheduling next swap if such a move has taken place. 108975 108976 last_swap_target defines the baseline for scheduling the next swap. 108977 If a movement between crtc's is not taken into account, the swap may 108978 schedule for a vblank count on the new crtc far in the future, resulting 108979 in a apparent "hang" of the drawable for a long time. 108980 108981 Fixes Bugzilla bug #28383. 108982 108983 Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 108984 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 108985 Signed-off-by: Keith Packard <keithp@keithp.com> 108986 108987commit 868bf3de478fac0080d525137fc728c551c78ea6 108988Author: Peter Hutterer <peter.hutterer@who-t.net> 108989Date: Tue Jun 15 13:22:20 2010 +1000 108990 108991 dix: the default axis mode is Relative, not "0". 108992 108993 Yay for readability. 108994 108995 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 108996 Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 108997 108998commit db0d3d4e573c93a8af43499682ed788c7bd0d70d 108999Author: Xiaoyang Yu (Max) <max.a.yu@intel.com> 109000Date: Fri Jun 4 17:17:53 2010 +0800 109001 109002 Re-enabled Xnest fix for focus in + modifier bug. 109003 109004 * See https://bugs.freedesktop.org/show_bug.cgi?id=3030 109005 109006 Signed-off-by: Xiaoyang Yu (Max) <max.a.yu@intel.com> 109007 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109008 109009 This takes the xnest way of working around this (see 109010 5904ef2ccd6056b187ca76f104c21e2d686bfc1d "xnest: restore 109011 xnestUpdateModifierState") and copies it to Xephyr. 109012 109013 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109014 109015commit 505724c5f3e4fc1475fa54f20d08c0c06fb32443 109016Author: Keith Packard <keithp@keithp.com> 109017Date: Mon Jun 7 16:28:34 2010 -0700 109018 109019 In DRI2Connect, check to see if DRI2 has been initialized (bug 28424) 109020 109021 This prevents DRI2GetScreen from being invoked with an uninitialized 109022 private key which would cause an assert failure. 109023 109024 Signed-off-by: Keith Packard <keithp@keithp.com> 109025 Tested-by: David Ronis <David.Ronis@McGill.CA> 109026 109027commit 68a9ee8370e6f9b38218376ac92d5130a5b0ef1e 109028Author: Keith Packard <keithp@keithp.com> 109029Date: Fri May 7 22:56:04 2010 -0700 109030 109031 Clean up RandR12 bits on screen close (bug 27114) 109032 109033 When resetting the server, pScrn->EnterVT must be unwrapped or the 109034 next server generation will end up wrapping the wrapper and causing an 109035 infinite recursion on EnterVT. 109036 109037 Signed-off-by: Keith Packard <keithp@keithp.com> 109038 Tested-by: Michael Stapelberg <michael+freedesktop@stapelberg.de> 109039 109040commit d5ab717768524c8552b81607aaeffb447ab268ee 109041Author: Keith Packard <keithp@keithp.com> 109042Date: Thu Jun 10 19:21:18 2010 -0700 109043 109044 Bump to 1.8.99.901 -- 1.9 RC1 109045 109046 Signed-off-by: Keith Packard <keithp@keithp.com> 109047 109048commit a41d6e9bffbe56cfa1c3b84388a3d9f5a982f1a9 109049Merge: 7e8f10012 f4190feb2 109050Author: Keith Packard <keithp@keithp.com> 109051Date: Fri Jun 11 10:08:13 2010 -0700 109052 109053 Merge remote branch 'dottedmag/for-keithp' 109054 109055commit 7e8f1001217326cc451974bacf25275420c4bb4e 109056Author: Éric Piel <E.A.B.Piel@tudelft.nl> 109057Date: Fri Jun 11 09:16:32 2010 -0700 109058 109059 exa: fix ExaCheckCopyNtoN for exa_classic when source = dest 109060 109061 In case you want to copy a region with source = dest, you have the same pixmap 109062 as source and dest. 109063 109064 At the end of exaPixmapIsOffscreen_classic() the devPrivate.ptr is reset to 109065 NULL (look at the sources). 109066 109067 Now this is what happens in ExaCheckCopyNtoN: 109068 109069 exaPrepareAccess( pDst ); 109070 Calls IsOffscreen() 109071 sets devPrivate.ptr to NULL 109072 sets up devPrivate.ptr to real pointer 109073 Everything OK 109074 exaPrepareAccess( pSrc ); 109075 Calls IsOffscreen() 109076 sets devPrivate.ptr to NULL 109077 BAILS OUT CAUSE OF NESTED OPERATION SINCE DST EQUALS SRC 109078 109079 We end up with devPrivate.ptr as NULL, and that is clearly wrong. 109080 109081 In particular this fixes a segfault when using the psb driver (bug 28077) 109082 109083 Signed-off-by: Éric Piel <eric.piel@tremplin-utc.net> 109084 Reviewed-by: Michel Dänzer <michel@daenzer.net> 109085 Signed-off-by: Keith Packard <keithp@keithp.com> 109086 109087commit f4190feb25ecc3d8278decc8bf28a5ef0e568942 109088Author: Matt Turner <mattst88@gmail.com> 109089Date: Thu Jun 10 04:08:54 2010 +0700 109090 109091 Remove more superfluous if(p) checks around free(p) 109092 109093 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109094 Signed-off-by: Matt Turner <mattst88@gmail.com> 109095 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109096 109097commit 89bd05106e5823fc5cfca9abf082729f2444363b 109098Author: Matt Turner <mattst88@gmail.com> 109099Date: Fri Jun 4 21:09:35 2010 -0400 109100 109101 record: move free() to after last use of pContext 109102 109103 No functional change, since free doesn't change the value of the pointer 109104 passed to it, but it makes this code less confusing. 109105 109106 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109107 Signed-off-by: Matt Turner <mattst88@gmail.com> 109108 109109commit a54a766dfb39fb3df671045878ac706215d83cef 109110Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109111Date: Mon Jun 7 03:19:03 2010 +0700 109112 109113 xkb: replace xstrdup with strdup in Win32System 109114 109115 The only caller of Win32System is XkbDDXCompileKeymapByNames. Add allocation 109116 check there to avoid passing NULL pointers to various functions down the code. 109117 109118 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109119 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109120 109121commit 6e7417c342b0624e3f3c5686bb43026786423692 109122Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109123Date: Fri Jun 4 17:05:35 2010 +0700 109124 109125 config: Replace xstrdup with strdup in add_option() 109126 109127 All callers of add_option pass string literal as "key" argument 109128 except one, where non-NULL condition is guarded by if(). 109129 109130 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109131 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109132 109133commit e4570f5db5157f4233454c938733a2a0d6a1cb8f 109134Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109135Date: Fri Jun 4 17:01:42 2010 +0700 109136 109137 xfree86: Get rid of xstrdup when argument is definitely non-NULL 109138 109139 ditto for XFree86 109140 109141 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109142 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109143 109144commit 416d228481d71204cf9bfad3ab4773abc4757f79 109145Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109146Date: Fri Jun 4 17:01:25 2010 +0700 109147 109148 xquartz: Get rid of xstrdup when argument is definitely non-NULL 109149 109150 ditto for XQuartz 109151 109152 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109153 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109154 109155commit eea286f2b9e4b1acf2b1c9800f5a4bc7c3fa968d 109156Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109157Date: Fri Jun 4 17:01:05 2010 +0700 109158 109159 xwin: Get rid of xstrdup when argument is definitely non-NULL 109160 109161 ditto for XWin 109162 109163 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109164 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109165 109166commit dd45b7d74696cd4fe9545852640a3c2e66a808fd 109167Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109168Date: Fri Jun 4 17:00:46 2010 +0700 109169 109170 kdrive: Get rid of xstrdup when argument is definitely non-NULL 109171 109172 ditto for Kdrive 109173 109174 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109175 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109176 109177commit 16158366954d945db6263f6de505fc02ee03c6cd 109178Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109179Date: Fri Jun 4 17:00:15 2010 +0700 109180 109181 dmx: Get rid of xstrdup when argument is definitely non-NULL 109182 109183 ditto for DMX 109184 109185 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109186 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109187 109188commit 6592db6bb526f0c43b4c7b55859c629709e039b4 109189Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109190Date: Fri Jun 4 16:58:58 2010 +0700 109191 109192 Get rid of xstrdup when argument is definitely non-NULL 109193 109194 Replace xstrdup with strdup when either constant string is 109195 being duplicated or argument is guarded by conditionals and 109196 obviously can't be NULL 109197 109198 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109199 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109200 109201commit 620ca54aaa0b363fcf68cec1bd6c37e68c988352 109202Merge: fd69234db 8ff9b502c 109203Author: Keith Packard <keithp@keithp.com> 109204Date: Thu Jun 10 19:18:53 2010 -0700 109205 109206 Merge remote branch 'alanc/master' 109207 109208commit fd69234db004e98c663f5c6c4360d2cecaa8a4df 109209Author: Keith Packard <keithp@keithp.com> 109210Date: Sun Jun 6 15:21:28 2010 -0700 109211 109212 Ignore RandR timestamps harder 109213 109214 Checking timestamps in post 1.1 randr requests was never a good idea, 109215 let's ignore them and just make the configuration changes. 109216 109217 Signed-off-by: Keith Packard <keithp@keithp.com> 109218 Acked-by: Chase Douglas <chase.douglas@canonical.com> 109219 109220commit a8ec9eca850f2a7ad4c5cf31c1c011c120688496 109221Author: Keith Packard <keithp@keithp.com> 109222Date: Mon Jun 7 13:39:11 2010 -0700 109223 109224 Fix a couple more possible errors with input-only windows 109225 109226 Using type == DRAWABLE_WINDOW to differentiate between pixmaps and 109227 windows isn't sufficient as input-only windows will end up in the 109228 pixmap case. This patch changes a few more code paths to use 109229 WindowDrawable instead. 109230 109231 Signed-off-by: Keith Packard <keithp@keithp.com> 109232 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109233 109234commit 8ff9b502cfce3828f7855ffba7949d6ebee34031 109235Author: Alan Coopersmith <alan.coopersmith@oracle.com> 109236Date: Fri Jun 4 08:44:02 2010 -0700 109237 109238 Solaris: avoid memory leak if AGPIOC_INFO ioctl fails 109239 109240 Move malloc after ioctl, so we don't have to worry about free'ing the 109241 memory if the ioctl fails. 109242 109243 [ This bug was found by the Parfait bug checking tool. 109244 For more information see http://research.sun.com/projects/parfait ] 109245 109246 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109247 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109248 109249commit c5eb5d69e5183860185a05cfcce16af635cab9aa 109250Author: Alan Coopersmith <alan.coopersmith@oracle.com> 109251Date: Thu Jun 3 19:28:43 2010 -0700 109252 109253 Record some additional library dependencies in xf86 modules 109254 109255 Helps with symbol resolution when building with -z defs 109256 109257 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109258 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 109259 109260commit 1c08a37e0eb4746e8974eb7a70ca4b7b84712963 109261Author: Sam Lau <sam.lau@oracle.com> 109262Date: Thu Jun 3 19:17:14 2010 -0700 109263 109264 SecurityResource should not segfault when client owning resource has exited 109265 109266 Fixes OpenSolaris bug 6949754: 109267 Xorg crashes when the magnifier is enabled at gdm login greeter window. 109268 http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6949754 109269 109270 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109271 Reviewed-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 109272 109273commit 07a093add0b7e40c4d9b9b59273e3ff9e14a88a7 109274Merge: 84190d209 dc614484f 109275Author: Keith Packard <keithp@keithp.com> 109276Date: Thu Jun 10 18:39:10 2010 -0700 109277 109278 Merge remote branch 'whot/for-keith' 109279 109280commit dc614484f93b67e8b62dbb1bb2fd247fe5a4c850 109281Author: Peter Hutterer <peter.hutterer@who-t.net> 109282Date: Thu Jun 10 12:21:36 2010 +1000 109283 109284 Xi: don't copy the modifier key count when copying device classes (#25480) 109285 109286 The modifier key count is maintained by the XKB layer and 109287 increased/decreased for all modifiers that set state. 109288 109289 Test case, MD/SD modifier key count in comment: 109290 1. keyboard 1: press and hold Shift_L # SD:1 MD:1 109291 2. keyboard 2: press and release Shift_L # SD:1,0 MD:1,0 109292 <class copy happens> # SD:1 MD:1 109293 3. keyboard 1: release Shift_L # SD:0 MD:1 109294 4. keyboard 1: press and release Shift_L # SD:1,0 MD:2,1 109295 109296 The modifier is now logically down on the MD but not on keyboard 1 or 109297 keyboard 2. 109298 109299 XKB is layered in before the DIX, it increases/decreases the modifier key 109300 count accordingly. In the above example, during (2), the MD gets the key 109301 release and thus clears the modifier bit. (3) doesn't forward the release to 109302 the MD because it is already cleared. The copy of modifierKeysDown when the 109303 lastSlave changes however increases the counter for the held key. On (4), 109304 the press and release are both forwarded to the MD, causing a offset by 1 109305 and thus do not clear the logical modifier state. 109306 109307 X.Org Bug 25480 <http://bugs.freedesktop.org/show_bug.cgi?id=25480> 109308 109309 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109310 Acked-by: Daniel Stone <daniel@fooishbar.org> 109311 109312commit 7805e45284264b4cd286eece02e85426896e9f7b 109313Author: Peter Hutterer <peter.hutterer@who-t.net> 109314Date: Thu Jun 10 14:31:48 2010 +1000 109315 109316 config: remove redundant DBUS API define, require dbus-1 >= 1.0 109317 109318 It's still being pulled in by the HAL CFLAGS but the requirement to define 109319 this was dropped from DBus pre 1.0 (November 2006). 109320 This means we require dbus 1.0 now. 109321 109322 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109323 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 109324 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 109325 109326commit 66b21b2f455a1dfbc92f7caa571dcff3f3765808 109327Author: Dan Nicholson <dbn.lists@gmail.com> 109328Date: Mon Jun 7 20:39:58 2010 -0700 109329 109330 xfree86: Match devices based on current driver setting 109331 109332 Often we want to apply a driver specific option to a set of devices and 109333 don't care how the driver was selected for that device. The MatchDriver 109334 entry can be used to match the current driver string: 109335 109336 MatchDriver "evdev|mouse" 109337 Option "Emulate3Buttons" "yes" 109338 109339 The driver string is a case sensitive match. 109340 109341 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109342 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109343 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109344 109345commit a71bdff47d4cc80da6ceeb548db1dcc8e8b59702 109346Author: Dan Nicholson <dbn.lists@gmail.com> 109347Date: Mon Jun 7 20:39:57 2010 -0700 109348 109349 xfree86: Allow multiple InputClass Match* entries for && matching 109350 109351 Currently when there multiple InputClass entries of the same type, only 109352 the last entry is used and the previous ones are ignored. Instead, 109353 multiple entries are used to create multiple matching conditions. 109354 109355 For instance, an InputClass with 109356 109357 MatchProduct "foo" 109358 MatchProduct "bar" 109359 109360 will require that the device's product name contain both foo and bar. 109361 This provides a complement to the || style matching when an entry is 109362 split using the "|" token. 109363 109364 The xorg.conf man page has added an example to hopefully clarify the two 109365 types of compound matches. 109366 109367 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109368 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109369 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109370 109371commit 842625928632ae767b0689fcdb5103c1779a4c91 109372Author: Dan Nicholson <dbn.lists@gmail.com> 109373Date: Mon Jun 7 20:39:56 2010 -0700 109374 109375 config: Script to convert HAL fdi settings to InputClass sections 109376 109377 In the new world of udev and InputClass, x11_* settings from HAL fdi 109378 files will not be honored. This script converts those settings into 109379 valid InputClass sections that can be dropped into xorg.conf.d. 109380 109381 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109382 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109383 109384commit 87a1507da7e7788232d74285ef377b67b70e0fa4 109385Author: Dan Nicholson <dbn.lists@gmail.com> 109386Date: Thu Jun 10 06:15:41 2010 -0700 109387 109388 xfree86: Match devices based on USB ID 109389 109390 Sometimes the vendor and product names aren't specific enough to target 109391 a USB device, so expose the numeric codes in the ID. A MatchUSBID entry 109392 has been added that supports shell pattern matching when fnmatch(3) is 109393 available. For example: 109394 109395 MatchUSBID "046d:*" 109396 109397 The IDs are stored in lowercase hex separated by a ':' like "lsusb" or 109398 "lspci -n". 109399 109400 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109401 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109402 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109403 109404commit 645679c1523eee7028f3244cee57936b93326a2a 109405Author: Dan Nicholson <dbn.lists@gmail.com> 109406Date: Thu Jun 10 06:11:10 2010 -0700 109407 109408 xfree86: Match devices based on PnP ID 109409 109410 Serial input devices lack properties such as product or vendor name. This 109411 makes matching InputClass sections difficult. Add a MatchPnPID entry to 109412 test against the PnP ID of the device. The entry supports a shell pattern 109413 match on platforms that support fnmatch(3). For example: 109414 109415 MatchPnPID "WACf*" 109416 109417 A match type for non-path pattern matching, match_pattern, has been added. 109418 The difference between this and match_path_pattern is the FNM_PATHNAME 109419 flag in fnmatch(3). 109420 109421 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109422 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109423 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109424 109425commit 84190d20954ef7888e4d96832c2a4b3225b4dfa2 109426Merge: 4172aa137 5e1ef1f4d 109427Author: Keith Packard <keithp@keithp.com> 109428Date: Thu Jun 10 13:36:35 2010 -0700 109429 109430 Merge remote branch 'jeremyhu/master' 109431 109432commit 4172aa137c1b9b6f2a25c320d847af1f5ac56fba 109433Author: Keith Packard <keithp@keithp.com> 109434Date: Thu Jun 10 07:15:49 2010 -0700 109435 109436 dri2: Only deal with output windows and pixmaps. 109437 109438 This reverts commit fdb081b430ddffb495aa5b05bcc4cf10882ff4b2 109439 "dri2: Deal with input-only windows by using WindowDrawable()" 109440 and replaces it as follows: 109441 109442 Reject the creation of a DRI2 drawable for UNDRAWABLE_WINDOW (input-only 109443 windows) and DRAWABLE_BUFFER (whatever those are) drawables and only look up 109444 privates for the supported drawable types. 109445 109446 The rest of the the code can continue pretending there's only output windows 109447 and pixmaps, which are the only kinds of drawables relevant for DRI2. 109448 109449 Fixes server crash with GLX compositing managers such as compiz or kwin, due 109450 to looking up a window private for a pixmap and getting a bogus pointer. 109451 109452 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 109453 Reviewed-by: Keith Packard <keithp@keithp.com> 109454 Signed-off-by: Keith Packard <keithp@keithp.com> 109455 109456commit 353e32d3712f3a883a796ba562ec9fb5a8354837 109457Author: Jamey Sharp <jamey@minilop.net> 109458Date: Sun Jun 6 06:42:11 2010 -0700 109459 109460 glxdriswrast bugfix: configure swapgc, not gc. 109461 109462 I got this wrong in e2929db7b737413cf93fbebdf4d15abdfebff05c. 109463 109464 Signed-off-by: Jamey Sharp <jamey@minilop.net> 109465 Cc: Kristian Høgsberg <krh@bitplanet.net> 109466 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 109467 Signed-off-by: Keith Packard <keithp@keithp.com> 109468 109469commit b3548612c7943011f79a910f9a59bb975984d8a6 109470Author: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 109471Date: Sat Jun 5 13:09:28 2010 +0200 109472 109473 DRI2: Don't return junk reply instead of blocking in glXWaitForSbcOML() 109474 109475 DRI2WaitSBC() didn't block if requested targetSBC wasn't yet reached. 109476 109477 Instead it returned a xreply with uninitialized junk return values, then 109478 blocked the connection until targetSBC was reached. 109479 109480 Therefore the client didn't block, but continued with bogus return 109481 values from glXWaitForSbcOML. 109482 109483 This patch fixes the problem by implementing DRI2WaitSBC similar 109484 to the clean and proven DRI2WaitMSC implementation. 109485 109486 Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 109487 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 109488 Signed-off-by: Keith Packard <keithp@keithp.com> 109489 109490commit 7287ef9e6cf953066e4a092cca9d0e4a279172bf 109491Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109492Date: Sun Jun 6 21:18:40 2010 +0700 109493 109494 Remove unnecessary parentheses around return values in functions 109495 109496 This patch was generated by the following Perl code: 109497 109498 perl -i -pe 's/([^_])return\s*\(\s*([^(]+?)\s*\)s*;(\s+(\n))?/$1return $2;$4/g;' 109499 109500 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109501 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109502 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 109503 Signed-off-by: Keith Packard <keithp@keithp.com> 109504 109505commit d1b4beecbc16448282dcc825dd5c354e96e48eca 109506Author: Dan Nicholson <dbn.lists@gmail.com> 109507Date: Mon Jun 7 20:39:53 2010 -0700 109508 109509 xfree86: Add MatchOS InputClass entry for operating system matching 109510 109511 Allow InputClass sections to match against the running operating system 109512 to narrow the application of rules. An example where this could be used 109513 is to specify that the default input driver on Linux is evdev while it's 109514 mouse/kbd everywhere else. 109515 109516 The operating system name is the same as `uname -s`, and matching is 109517 case-insensitive. 109518 109519 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109520 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109521 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109522 109523commit 9b30fa9f8fedb7ddb5672f93ed1a154d13578c47 109524Author: Dan Nicholson <dbn.lists@gmail.com> 109525Date: Mon Jun 7 20:39:52 2010 -0700 109526 109527 xfree86: Refactor InputClass matching code 109528 109529 InputClassMatches was starting to get a little hairy with all the loops 109530 over the tokenized match strings. This adds code, but makes it easier to 109531 read and add new matches. 109532 109533 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109534 Reviewed-by: Jamey Sharp <jamey at minilop.net> 109535 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109536 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109537 109538commit d09bc8f8e5bb92899f7c90eab6b0907920c7d643 109539Author: Dan Nicholson <dbn.lists@gmail.com> 109540Date: Mon Jun 7 20:39:51 2010 -0700 109541 109542 xfree86: Constify InputClass functions 109543 109544 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109545 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109546 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109547 109548commit 61c35d8ac4b6423caf2dcecb4ff25baaaaa14b1c 109549Author: Dan Nicholson <dbn.lists@gmail.com> 109550Date: Mon Jun 7 20:39:50 2010 -0700 109551 109552 config/hal: Missing newline in log message 109553 109554 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109555 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109556 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109557 109558commit 697c5b010d7f6df0ec9cae6352b125c4cbdbaad3 109559Author: Dan Nicholson <dbn.lists@gmail.com> 109560Date: Mon Jun 7 20:39:49 2010 -0700 109561 109562 config/udev: Add verbose info when fetching udev properties 109563 109564 Give the user a chance to see why their input devices are being ignored, 109565 even if they have to start the server with -logverbose. 109566 109567 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 109568 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109569 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109570 109571commit 94547b2d1448232ae316c228d65482cc0b646ac2 109572Author: Dan Nicholson <dbn.lists@gmail.com> 109573Date: Mon Jun 7 20:39:48 2010 -0700 109574 109575 config/hal: don't leak the input.tags property 109576 109577 Signed-off-by: Julien Cristau <jcristau@debian.org> 109578 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 109579 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109580 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 109581 109582commit 5e1ef1f4dde4615498c1c5e7a4e422c3b3432fe8 109583Author: Jeremy Huddleston <jeremyhu@apple.com> 109584Date: Tue Jun 8 13:01:10 2010 -0700 109585 109586 XQuartz: Add more sanity checking around string manipulation for xmodmap 109587 109588 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 109589 109590commit 0be81f9a163f2838448caee41b2dda274d73391a 109591Author: Jeremy Huddleston <jeremyhu@apple.com> 109592Date: Sun Jun 6 16:31:19 2010 -0700 109593 109594 XQuartz: Fix build regression from faeebead7bfcc78535757ca7acc1faf7554c03b7 109595 109596 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 109597 109598commit b3a7b229e1e1f212bdd185af5443311091824005 109599Author: Keith Packard <keithp@keithp.com> 109600Date: Mon Jun 7 21:49:43 2010 -0700 109601 109602 Make DRI_SCREEN_PRIV just return NULL when DRI isn't initialized 109603 109604 The code relied upon the previous devPrivate implementation which 109605 offered this convenience. 109606 109607 Signed-off-by: Keith Packard <keithp@keithp.com> 109608 109609commit f7af00e9f0e0e1d854b0e882378c032518ab71ca 109610Author: Ben Skeggs <bskeggs@redhat.com> 109611Date: Mon Jun 7 10:21:18 2010 +1000 109612 109613 randr: prevent an unnecessary screen resize with multiple displays 109614 109615 crtc->{x,y} is always 0 when xf86DefaultScreenLimits() is called, so we 109616 calculate too small an area for the initial framebuffer and force a resize 109617 to happen. 109618 109619 This commit fixes the code to use desired{X,Y} instead, which contains the 109620 initial output positions. 109621 109622 Signed-off-by: Ben Skeggs <bskeggs@redhat.com> 109623 Reviewed-by: Dave Airlie <airlied@redhat.com> 109624 Reviewed-by: Keith Packard <keithp@keithp.com> 109625 Signed-off-by: Keith Packard <keithp@keithp.com> 109626 109627commit 8e97e5f9425639ad0a084150d0b232cad417595d 109628Author: Keith Packard <keithp@keithp.com> 109629Date: Mon Jun 7 14:23:47 2010 -0700 109630 109631 If XTest is always required, then eliminate the XTest devPrivate 109632 109633 The internals of XTest are used by Xi and Xkb, and both Xi and Xkb are 109634 always required, so it makes little sense to have XTest place data in 109635 a devPrivate, especially a devPrivate which is only available when the 109636 XTest extension is enabled. 109637 109638 Signed-off-by: Keith Packard <keithp@keithp.com> 109639 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 109640 109641commit fdb081b430ddffb495aa5b05bcc4cf10882ff4b2 109642Author: Keith Packard <keithp@keithp.com> 109643Date: Mon Jun 7 00:54:18 2010 -0700 109644 109645 dri2: Deal with input-only windows by using WindowDrawable() 109646 109647 Input only windows aren't DRAWABLE_WINDOW, but casting them to a 109648 PixmapPtr is a bit harsh, and unlikely to get the appropriate privates 109649 structure. use WindowDrawable instead which checks for both 109650 input-output and input-only windows. 109651 109652 Signed-off-by: Keith Packard <keithp@keithp.com> 109653 Tested-by: Eric Anholt <eric@anholt.net> 109654 109655commit bc26665661565918af484ccd17caad951010df60 109656Author: Keith Packard <keithp@keithp.com> 109657Date: Sun Jun 6 20:45:53 2010 -0700 109658 109659 Initialize private keys in test suite 109660 109661 Make sure all of the private keys used by the test code are 109662 initialized before being used. 109663 109664 Signed-off-by: Keith Packard <keithp@keithp.com> 109665 Tested-by: Robert Hooker <sarvatt@ubuntu.com> 109666 109667commit f03be727d647183a2176355ad0ac9a6735067be9 109668Author: Keith Packard <keithp@keithp.com> 109669Date: Sun Jun 6 16:04:42 2010 -0700 109670 109671 Place glyph privates at correct location within the allocate storage 109672 109673 A glyph allocation consists of : 109674 109675 GlyphRec 109676 numScreens * PicturePtr 109677 glyph privates 109678 109679 Tell the dix private bits to start past the picture pointers. 109680 109681 Signed-off-by: Keith Packard <keithp@keithp.com> 109682 Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> 109683 109684commit 67b824a81bb8254827023de351a4a094031f0e94 109685Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109686Date: Sun Jun 6 20:37:07 2010 +0700 109687 109688 randr: Make deletion of output properties more robust 109689 109690 Previously there was two branches of code with small discrepancies between them 109691 (especially prop->valid_values field was not free(3)ed). Extract the common 109692 routine and fix double-free prop->valid_values in RRDestroyOutputProperty by 109693 the way. 109694 109695 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109696 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109697 109698commit 0a4d8cbdcd7b25313fb29ccdb498044af95f7de6 109699Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109700Date: Sun Jun 6 15:28:30 2010 +0700 109701 109702 Remove more superfluous if(p) checks around free(p) 109703 109704 This patch has been generated by the following Coccinelle semantic patch: 109705 109706 @@ 109707 expression E; 109708 @@ 109709 109710 -if(E) { free(E); } 109711 +free(E); 109712 109713 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109714 Reviewed-by: Julien Cristau <jcristau@debian.org> 109715 Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 109716 Reviewed-by: Matt Turner <mattst88@gmail.com> 109717 109718commit 5a0fc0ad21d562612676ef88ef2d533b2391810a 109719Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109720Date: Fri Jun 4 15:35:31 2010 +0700 109721 109722 Replace deprecated bzero with memset 109723 109724 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109725 Reviewed-by: Marcin Baczyński <marbacz@gmail.com> 109726 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 109727 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109728 Reviewed-by: Matt Turner <mattst88@gmail.com> 109729 109730commit f9810ba914877b379cb36f1b9755f7923ceca14c 109731Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109732Date: Fri Jun 4 14:26:31 2010 +0700 109733 109734 xkb: Remove superfluous if(x) x = realloc(x, sz); else x = malloc(sz); logic 109735 109736 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109737 Reviewed-by: Marcin Baczyński <marbacz@gmail.com> 109738 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109739 Reviewed-by: Matt Turner <mattst88@gmail.com> 109740 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 109741 109742commit 77ea20895ca1d163e14c6b767d8c287c94ba3114 109743Author: Mikhail Gusarov <dottedmag@dottedmag.net> 109744Date: Fri Jun 4 14:22:06 2010 +0700 109745 109746 xkb: Remove superfluous if(p) checks around free(p) 109747 109748 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 109749 Reviewed-by: Julien Cristau <jcristau@debian.org> 109750 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 109751 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 109752 109753commit 5d4e2c594059ffb536c8e506c2623320d3c6a787 109754Merge: eacc42132 793dd3994 109755Author: Keith Packard <keithp@keithp.com> 109756Date: Sat Jun 5 22:20:28 2010 -0700 109757 109758 Merge remote branch 'jamey/for-keith' 109759 109760commit 793dd399468ff3ba22ffa9c62fa54fc87261fbe7 109761Author: Jamey Sharp <jamey@minilop.net> 109762Date: Tue May 25 15:08:38 2010 -0700 109763 109764 dmx: Split glx scalar swap variables from array swap variables. 109765 109766 This makes hw/dmx/glxProxy/unpack.h more closely resemble glx/unpack.h, 109767 and fixes the "unused variable 'swapEnd'" and "unused variable 'swapPC'" 109768 warnings. 109769 109770 Signed-off-by: Jamey Sharp <jamey@minilop.net> 109771 Reviewed-by: Keith Packard <keithp@keithp.com> 109772 109773commit eacc42132ed0f1e8f4a4d65a4b6637ab06f7e2aa 109774Author: Kim Woelders <kim@woelders.dk> 109775Date: Fri Jun 4 20:20:47 2010 +0200 109776 109777 dix: Remove test for obsolete NO_XINERAMA_PORT macro. 109778 109779 Signed-off-by: Kim Woelders <kim@woelders.dk> 109780 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109781 Signed-off-by: Keith Packard <keithp@keithp.com> 109782 109783commit a0fe6987b55b5326157b50e169b8d6b9ae26a1ae 109784Author: Jamey Sharp <jamey@minilop.net> 109785Date: Sat Jun 5 20:55:29 2010 -0700 109786 109787 Clean up after removal of screen parameters from region macros. 109788 109789 Signed-off-by: Jamey Sharp <jamey@minilop.net> 109790 Reviewed-by: Keith Packard <keithp@keithp.com> 109791 109792commit 424b593c7872a703d3d0c942b7e8acc0770eb023 109793Author: Keith Packard <keithp@keithp.com> 109794Date: Sat Jun 5 21:34:51 2010 -0700 109795 109796 Damage pixmap private key was being mis-registered as a picture key 109797 109798 This would cause all kinds of fun, in particular Xnest would crash at startup. 109799 109800 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109801 Signed-off-by: Keith Packard <keithp@keithp.com> 109802 109803commit 9c121ae45ae1d2ebd892c45e1631716ef5e7d649 109804Author: Keith Packard <keithp@keithp.com> 109805Date: Sat Jun 5 21:34:27 2010 -0700 109806 109807 Register Xnest colormap screen private key 109808 109809 This key is used to store the currently installed colormap and must be 109810 registered before we can use it. 109811 109812 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109813 Signed-off-by: Keith Packard <keithp@keithp.com> 109814 109815commit df534d05208436e9ca8c9b42acff3fbe240e3ec4 109816Author: Keith Packard <keithp@keithp.com> 109817Date: Sat Jun 5 21:24:54 2010 -0700 109818 109819 Make GetPictureScreenIfSet check if Render is running 109820 109821 This macro originally checked to see if the Render screen private 109822 index had been allocated. When the privates were changed the first 109823 time, there wasn't any need to check as dixLookupPrivate would simply 109824 return NULL in that case. Now that we care, check to see if the key 109825 has been initialized before asking for the value. 109826 109827 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109828 Signed-off-by: Keith Packard <keithp@keithp.com> 109829 109830commit e7fc8b32e41e10c057d2787fcc377296be67f2e9 109831Author: Keith Packard <keithp@keithp.com> 109832Date: Sat Jun 5 21:08:01 2010 -0700 109833 109834 Move the shadow screen private key initialization to shadowSetup 109835 109836 Some users of the shadow code don't call shadowInit, just shadowSetup 109837 and so make sure the key is initialized there. 109838 109839 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109840 Signed-off-by: Keith Packard <keithp@keithp.com> 109841 109842commit 6bd5f0d75bca727c4686b20eee166c8cae472ba2 109843Author: Keith Packard <keithp@keithp.com> 109844Date: Sat Jun 5 21:07:15 2010 -0700 109845 109846 Fix exa_priv.h declarations of privates 109847 109848 exa_priv.h exposes the privates to the rest of the exa code, and yet 109849 the declarations of the privates wasn't fixed when the exa.c versions were. 109850 109851 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109852 Signed-off-by: Keith Packard <keithp@keithp.com> 109853 109854commit 6b306f43384e5c2143197e746a5a39c4ebb2583c 109855Author: Keith Packard <keithp@keithp.com> 109856Date: Sat Jun 5 20:58:04 2010 -0700 109857 109858 kdrive: Xv code uses shared screen private instead of kdrive-specific private 109859 109860 When doing the devPrivate API conversion, I missed this as there was 109861 also a local private key which wasn't ever used. 109862 109863 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109864 Signed-off-by: Keith Packard <keithp@keithp.com> 109865 109866commit ab07e2b8ededaa2193fc199a8c09623d84032280 109867Author: Keith Packard <keithp@keithp.com> 109868Date: Fri Apr 30 19:38:48 2010 -0700 109869 109870 Allocate per-screen device/cursor-bits private keys in midispcur 109871 109872 midispcur was abusing the CursorScreenKey to index the cursor_bits 109873 privates, it also had a MAXSCREENS array of keys to index device 109874 privates. Switch both of these to the new dixCreatePrivateKey API and 109875 store a pointer to that in the screen private. 109876 109877 Signed-off-by: Keith Packard <keithp@keithp.com> 109878 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109879 109880commit 34db537907c6cb2635dbefdce7dcfcae90f7c902 109881Author: Keith Packard <keithp@keithp.com> 109882Date: Fri Apr 30 19:36:33 2010 -0700 109883 109884 Add dixCreatePrivateKey API 109885 109886 Keys need to persist through server reset so that the private system 109887 can be cleaned up in dixResetPrivates. In particular, this means that 109888 keys cannot live in objects freed at reset time. This API provides 109889 suitable object lifetime by having the privates code free the key in 109890 the reset path. 109891 109892 Signed-off-by: Keith Packard <keithp@keithp.com> 109893 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109894 109895commit 495fc3eb2d6c98bde82ae1278f89fcf131fd9bf8 109896Author: Keith Packard <keithp@keithp.com> 109897Date: Fri Apr 30 19:38:38 2010 -0700 109898 109899 Change devPrivates implementation. 109900 109901 Each key now declares which object type it belongs to, this permits 109902 keys for different types to share the same offset within the allocated 109903 privates. As a special case for XSELinux, a key may be allocated as 109904 PRIVATE_XSELINUX which will allow it to continue to be used across the 109905 relevant object types. 109906 109907 Signed-off-by: Keith Packard <keithp@keithp.com> 109908 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109909 109910commit faeebead7bfcc78535757ca7acc1faf7554c03b7 109911Author: Keith Packard <keithp@keithp.com> 109912Date: Mon Apr 26 17:22:21 2010 -0700 109913 109914 Change the devPrivates API to require dixRegisterPrivateKey 109915 109916 This patch only changes the API, not the implementation of the 109917 devPrivates infrastructure. This will permit a new devPrivates 109918 implementation to be layed into the server without requiring 109919 simultaneous changes in every devPrivates user. 109920 109921 Signed-off-by: Keith Packard <keithp@keithp.com> 109922 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> 109923 109924commit c865a24401f06bcf1347d8b41f736a066ab25693 109925Author: Keith Packard <keithp@keithp.com> 109926Date: Fri Apr 30 22:05:27 2010 -0700 109927 109928 Create separate private key for midispcur cursor bits 109929 109930 The CursorScreenKey array is supposed to be used by the DDX for cursor 109931 private data, but midispcur was abusing it to hold cursor bits private 109932 information. Create a separate set of privates for the dispcur cursor 109933 bits information. 109934 109935 This also renames the device private index and macros to better 109936 reflect their usage: 109937 109938 miDCSpriteKey -> miDCDeviceKey 109939 MIDCBUFFER -> miGetDCDevice 109940 109941 Signed-off-by: Keith Packard <keithp@keithp.com> 109942 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109943 109944commit 431781a921251d54782f0a4f194bbef1fabd1380 109945Author: Keith Packard <keithp@keithp.com> 109946Date: Sat May 15 14:52:39 2010 -0700 109947 109948 Remove dixRegisterPrivateOffset; hard-code devPrivates offsets instead 109949 109950 For predefined resource types, the offset of the devPrivates field was 109951 already kept in a constant table. The only non-predefined type needing 109952 this treatment was dbeDrawableResType, which is just a magic alias for 109953 RT_PIXMAP. 109954 109955 This patch special-cases looking up RC_DRAWABLE offsets and uses the 109956 table directly for everything else. 109957 109958 Signed-off-by: Keith Packard <keithp@keithp.com> 109959 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109960 109961commit 7ef612de784daaed09ba13f4615c10714614033f 109962Author: Keith Packard <keithp@keithp.com> 109963Date: Fri May 21 12:35:29 2010 -0700 109964 109965 Add REGION_ macros for source compatibility with existing drivers. 109966 109967 This makes the region code changes source compatible with existing 109968 code, although none of them are used within the server source itself. 109969 109970 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109971 Signed-off-by: Keith Packard <keithp@keithp.com> 109972 109973commit 965cade6ee7c226f850e806cb2e910793d1536f9 109974Author: Keith Packard <keithp@keithp.com> 109975Date: Fri May 21 12:18:22 2010 -0700 109976 109977 Change region implementation from macros to inline functions. 109978 109979 This makes all of the previous macros into inline functions and also 109980 turns all of the direct calls to pixman region code into inline 109981 functions as well. 109982 109983 Reviewed-by: Jamey Sharp <jamey@minilop.net> 109984 Signed-off-by: Keith Packard <keithp@keithp.com> 109985 109986commit 2dc138922b7588515d5f2447e4b9dcdc0bef15e0 109987Author: Keith Packard <keithp@keithp.com> 109988Date: Fri May 21 15:05:48 2010 -0700 109989 109990 Rename region macros to eliminate screen argument 109991 109992 This is a combination of a huge mechanical patch and a few small 109993 fixups required to finish the job. They were reviewed separately, but 109994 because the server does not build without both pieces, I've merged 109995 them together at this time. 109996 109997 The mechanical changes were performed by running the included 109998 'fix-region' script over the whole tree: 109999 110000 $ git ls-files | grep -v '^fix-' | xargs ./fix-region 110001 110002 And then, the white space errors in the resulting patch were fixed 110003 using the provided fix-patch-whitespace script. 110004 110005 $ sh ./fix-patch-whitespace 110006 110007 Thanks to Jamey Sharp for the mighty fine sed-generating sed script. 110008 110009 The hand-done changes involve removing functions from dix/region.c 110010 that duplicate inline functions in include/regionstr.h, along with 110011 their declarations in regionstr.h, mi.h and mispans.h. 110012 110013 Reviewed-by: Jamey Sharp <jamey@minilop.net> 110014 Signed-off-by: Keith Packard <keithp@keithp.com> 110015 110016commit d17e726e89ef644310de77b960b715c2d11088da 110017Author: Keith Packard <keithp@keithp.com> 110018Date: Sat May 22 00:09:47 2010 -0700 110019 110020 Move mi/miregion.c to dix/region.c 110021 110022 Reviewed-by: Jamey Sharp <jamey@minilop.net> 110023 Signed-off-by: Keith Packard <keithp@keithp.com> 110024 110025commit 9b6f5f549dfe4c20e3cb3c2ef7095ab47abaca3b 110026Author: Keith Packard <keithp@keithp.com> 110027Date: Fri May 21 11:14:59 2010 -0700 110028 110029 Change region implementation names to eliminate the 'mi' prefix 110030 110031 This prepares the file to be moved from mi to dix. This patch 110032 was done mechanically with the included scripts 'fix-miregion' run over 110033 the entire X server and 'fix-miregion-private' run over 110034 include/regionstr.h and mi/miregion.c. 110035 110036 Reviewed-by: Jamey Sharp <jamey@minilop.net> 110037 Signed-off-by: Keith Packard <keithp@keithp.com> 110038 110039commit 1768e51976ae8949d999309bd2384ccad72deecd 110040Author: Jeremy Huddleston <jeremyhu@apple.com> 110041Date: Sat Jun 5 15:58:21 2010 -0700 110042 110043 rootless: Fix SetShape regression 110044 110045 This fixes a regression in miext/rootless from 643cb6e87c10ab554c03ada81930001a8ebcc909 110046 110047 Found-by: tinderbox 110048 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 110049 Reviewed-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 110050 Signed-off-by: Keith Packard <keithp@keithp.com> 110051 110052commit 6b4af3b7925978cd79f717761f1b6f33bd8dfbaf 110053Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 110054Date: Fri Jun 4 16:27:28 2010 +0300 110055 110056 configure: Check for libsha1.pc 110057 110058 xserver fails to detect libsha1. Problem is that configure checks for 110059 sha1.pc when libsha1 provides libsha1.pc. 110060 110061 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 110062 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 110063 Signed-off-by: Keith Packard <keithp@keithp.com> 110064 110065commit d163266692235261e1e0cfa6b900f54e6ec63de5 110066Author: Alan Coopersmith <alan.coopersmith@oracle.com> 110067Date: Thu Jun 3 18:59:04 2010 -0700 110068 110069 Stop searching for XF86Config files 110070 110071 xorg.conf has been used since the X11R6.7 release in April 2004. 110072 6 years has been a generous transition period for users to 110073 "mv XF86Config xorg.conf" and for distros to update their 110074 configuration tools and packages. 110075 110076 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 110077 Reviewed-by: James Cloos <cloos@jhcloos.com> 110078 Signed-off-by: Keith Packard <keithp@keithp.com> 110079 110080commit 01c75522b6ed15e979347acd38e234e2cb9006c3 110081Author: Alan Coopersmith <alan.coopersmith@oracle.com> 110082Date: Thu Jun 3 18:59:03 2010 -0700 110083 110084 Stop searching for XF86Config files 110085 110086 xorg.conf has been used since the X11R6.7 release in April 2004. 110087 6 years has been a generous transition period for users to 110088 "mv XF86Config xorg.conf" and for distros to update their 110089 configuration tools and packages. 110090 110091 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 110092 Reviewed-by: James Cloos <cloos@jhcloos.com> 110093 Signed-off-by: Keith Packard <keithp@keithp.com> 110094 110095commit 7f19a7a6e90a4fd7b7ec0256974f62e575218541 110096Author: Peter Hutterer <peter.hutterer@who-t.net> 110097Date: Fri Jun 4 14:15:04 2010 +1000 110098 110099 xkb: fix invalid memory writes in _XkbCopyGeom. 110100 110101 Classic strlen/strcpy mistake of 110102 foo = malloc(strlen(bar)); 110103 strcpy(foo, bar); 110104 110105 Testcase: valgrind Xephyr :1 110106 110107 ==8591== Invalid write of size 1 110108 ==8591== at 0x4A0638F: strcpy (mc_replace_strmem.c:311) 110109 ==8591== by 0x605593: _XkbCopyGeom (xkbUtils.c:1994) 110110 ==8591== by 0x605973: XkbCopyKeymap (xkbUtils.c:2118) 110111 ==8591== by 0x6122B3: InitKeyboardDeviceStruct (xkbInit.c:560) 110112 ==8591== by 0x4472E2: CoreKeyboardProc (devices.c:577) 110113 ==8591== by 0x447162: ActivateDevice (devices.c:530) 110114 ==8591== by 0x4475D6: InitCoreDevices (devices.c:672) 110115 ==8591== by 0x4449EE: main (main.c:254) 110116 ==8591== Address 0x6f96505 is 0 bytes after a block of size 53 alloc'd 110117 ==8591== at 0x4A0515D: malloc (vg_replace_malloc.c:195) 110118 ==8591== by 0x6054B7: _XkbCopyGeom (xkbUtils.c:1980) 110119 ==8591== by 0x605973: XkbCopyKeymap (xkbUtils.c:2118) 110120 ==8591== by 0x6122B3: InitKeyboardDeviceStruct (xkbInit.c:560) 110121 ==8591== by 0x4472E2: CoreKeyboardProc (devices.c:577) 110122 ==8591== by 0x447162: ActivateDevice (devices.c:530) 110123 ==8591== by 0x4475D6: InitCoreDevices (devices.c:672) 110124 ==8591== by 0x4449EE: main (main.c:254) 110125 110126 Reported-by: Dave Airlie <airlied@redhat.com> 110127 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110128 Reviewed-by-and-apologised-for: Daniel Stone <daniel@fooishbar.org> 110129 Signed-off-by: Keith Packard <keithp@keithp.com> 110130 110131commit 959a1eaf1c15a691141f1b0dc53757fe9b6e9b13 110132Author: Dave Airlie <airlied@redhat.com> 110133Date: Fri Jun 4 11:09:46 2010 +1000 110134 110135 composite: use config notify hook to do pixmap resize. 110136 110137 Since reallocating the backing pixmap can fail, we need to try and do 110138 it before any other side effects of reconfiguring the window happen. 110139 110140 This changes the ConfigNotify hook to return status, and moves the 110141 composite window reconfiguration wrappers to ConfigNotify. They all 110142 basically did the same thing, so we can drop the MoveWindow, 110143 ResizeWindow, ChangeBorderWidth wrappers, and allow ConfigNotify to do 110144 all the work. If reallocation fails we fail before we send any 110145 confiureNotify events, or enter the area we can't recover from. 110146 110147 The only place we now enforce 32k limits are in EXA/UXA/fb, so drivers 110148 that don't use this should probably deal with it in their pixmap 110149 allocate if they don't already. 110150 110151 This also breaks ABI, so we need an alternate fix for older servers, 110152 working on the X server makes me realise why I'm a kernel hacker. 110153 110154 Signed-off-by: Dave Airlie <airlied@redhat.com> 110155 Reviewed-by: Keith Packard <keithp@keithp.com> 110156 Signed-off-by: Keith Packard <keithp@keithp.com> 110157 110158commit 0f12e86e600522768f5f64eafc1230526e700ab7 110159Merge: aa7c09f7d f114f5498 110160Author: Keith Packard <keithp@keithp.com> 110161Date: Thu Jun 3 15:09:32 2010 -0700 110162 110163 Merge remote branch 'jamey/for-keith' 110164 110165commit f114f54986aca1add4c8808a05d9692e289547e7 110166Author: Jamey Sharp <jamey@minilop.net> 110167Date: Fri Apr 23 19:33:49 2010 -0700 110168 110169 Accumulate graphics exposures incrementally in PanoramiXCopyArea/Plane. 110170 110171 This fuses two loops in each function, eliminating an intermediate 110172 MAXSCREENS-sized array from each. 110173 110174 Aside from being more efficient, I believe this is equivalent to the 110175 previous implementation, since 110176 - each per-screen GC has the graphicsExposures flag set the same way, 110177 and 110178 - the REGION_* macros ignore their screen argument. 110179 110180 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110181 Reviewed-by: Keith Packard <keithp@keithp.com> 110182 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110183 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> (i686 GNU/Linux) 110184 110185commit a7c7ebe4b3a0df1a4d1ffe3c690cfb21470f103d 110186Author: Jamey Sharp <jamey@minilop.net> 110187Date: Sat May 22 21:13:48 2010 -0700 110188 110189 Delete XineramaScreenRegions cache. 110190 110191 Every screen region consists of a single rectangle, so initializing a 110192 stack-allocated region for each screen on-demand does no heap allocation 110193 and is fast. 110194 110195 This eliminates a MAXSCREENS-sized array. 110196 110197 The REGION_UNINIT calls are no-ops since no boxes are actually allocated 110198 for a single-rectangle region, but it seemed wiser to include them. 110199 110200 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110201 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110202 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> (i686 GNU/Linux) 110203 110204commit a0456da339fa9b3c3e6fead060338ee7348d6460 110205Author: Jamey Sharp <jamey@minilop.net> 110206Date: Sun May 23 12:57:52 2010 -0700 110207 110208 XineramaSetCursorPosition: use screen bounds directly, not POINT_IN_REGION. 110209 110210 This hides a MAXSCREENS-sized array as an implementation detail of 110211 panoramiX.c rather than an exported global. 110212 110213 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110214 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110215 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> (i686 GNU/Linux) 110216 110217commit a83cff9f4d622b069c96a68b4e87a669bf1f6446 110218Author: Jamey Sharp <jamey@minilop.net> 110219Date: Sun May 23 10:11:47 2010 -0700 110220 110221 Move each screen's x/y origin into ScreenRec. 110222 110223 Many references to the dixScreenOrigins array already had the 110224 corresponding screen pointer handy, which meant they usually looked like 110225 "dixScreenOrigins[pScreen->myNum]". Adding a field to ScreenRec instead 110226 of keeping this information in a parallel array simplifies those 110227 expressions, and eliminates a MAXSCREENS-sized array. 110228 110229 Since dix declared the dixScreenOrigins array, I figure allocating a 110230 screen private for these values is overkill. 110231 110232 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110233 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110234 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> (i686 GNU/Linux) 110235 110236commit 217ccaa5a341018457f468a774c035c0df47d918 110237Author: Jamey Sharp <jamey@minilop.net> 110238Date: Sun May 23 11:05:01 2010 -0700 110239 110240 Delete panoramiXdataPtr: it's redundant. 110241 110242 This eliminates a dynamically-allocated MAXSCREENS-sized array. 110243 110244 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110245 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110246 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> (i686 GNU/Linux) 110247 110248commit e7fae9ecc42ab5e73b89117722dbf4117d928f9a 110249Author: Jamey Sharp <jamey@minilop.net> 110250Date: Sat May 22 00:26:28 2010 -0700 110251 110252 Move each screen's root-window pointer into ScreenRec. 110253 110254 Many references to the WindowTable array already had the corresponding 110255 screen pointer handy, which meant they usually looked like 110256 "WindowTable[pScreen->myNum]". Adding a field to ScreenRec instead of 110257 keeping this information in a parallel array simplifies those 110258 expressions, and eliminates a MAXSCREENS-sized array. 110259 110260 Since dix uses this data, a screen private entry isn't appropriate. 110261 110262 xf86-video-dummy currently uses WindowTable, so it needs to be updated 110263 to reflect this change. 110264 110265 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110266 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110267 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> (i686 GNU/Linux) 110268 110269commit 80b5d3a3264d2c5167e5ac85a3b04af0f89cece1 110270Author: Jamey Sharp <jamey@minilop.net> 110271Date: Fri May 21 23:13:51 2010 -0700 110272 110273 Move each screen's screensaver data into ScreenRec. 110274 110275 Most references to the savedScreenInfo array already had the 110276 corresponding screen pointer handy, which meant they usually looked like 110277 "savedScreenInfo[pScreen->myNum]". Adding a field to ScreenRec instead 110278 of keeping this information in a parallel array simplifies those 110279 expressions, and eliminates a MAXSCREENS-sized array. 110280 110281 Since dix uses this data, a screen private entry isn't appropriate. 110282 110283 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110284 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110285 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> (i686 GNU/Linux) 110286 110287commit 5030540d6cd2aa9bbd4ecb7f4487616f9ef0140a 110288Author: Jamey Sharp <jamey@minilop.net> 110289Date: Thu May 27 11:50:52 2010 -0700 110290 110291 midispcur: Add comment explaining why pRootPicture must not be freed. 110292 110293 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110294 Cc: Peter Hutterer <peter.hutterer@who-t.net> 110295 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 110296 110297commit f308cbea8c9fa5c5c6705bc301178b167de450d5 110298Author: Jamey Sharp <jamey@minilop.net> 110299Date: Mon May 24 18:41:54 2010 -0700 110300 110301 mi: Delete unused flicker-free MoveCursor code. 110302 110303 It's been commented-out for three and a half years and nobody seems to 110304 be missing it enough to resurrect it. 110305 110306 Besides deleting code that is untested and therefore buggy, this saves a 110307 little memory for each pointer device on each screen. 110308 110309 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110310 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 110311 110312commit d1bad1b59f24ab4dcf2d1ce1c5d4f42b5fa40bc1 110313Author: Jamey Sharp <jamey@minilop.net> 110314Date: Tue May 25 14:25:07 2010 -0700 110315 110316 dmx: Fix XSetExtensionErrorHandler calls to agree on constness. 110317 110318 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110319 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 110320 110321commit aa7c09f7d04d59852b850b2b6993f4cbdc7ec940 110322Author: Alan Coopersmith <alan.coopersmith@oracle.com> 110323Date: Wed Jun 2 14:31:56 2010 -0700 110324 110325 Install fbdevhw module in normal module directory instead of linux/ 110326 110327 Since we no longer support OS-independent custom elfloader modules, 110328 we don't need to put the OS-dependent modules into os-specific subdirs 110329 any more. 110330 110331 We do however still need to install the stubs version of this module 110332 on non-Linux platforms, since a number of drivers link to functions 110333 in it, even when built on non-Linux platforms. 110334 110335 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 110336 Reviewed-by: Michel Dänzer <michel@daenzer.net> 110337 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110338 Signed-off-by: Keith Packard <keithp@keithp.com> 110339 110340commit b13d7a8e09c8e3eb5003420c3c0484431e5a052b 110341Merge: f0ab726d8 dcceb90b1 110342Author: Keith Packard <keithp@keithp.com> 110343Date: Thu Jun 3 07:01:26 2010 -0700 110344 110345 Merge remote branch 'vignatti/for-keith' 110346 110347commit dcceb90b1dccbbf5793f6ba1ed857013b71dafbc 110348Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110349Date: Wed May 19 11:45:46 2010 +0200 110350 110351 xext: delete wrong source reference from fontcache 110352 110353 This should go away in 0b45ba48. 110354 110355 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110356 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 110357 110358commit c043de3261b6e14f2446b2cd9d62dde5860d4720 110359Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110360Date: Mon May 3 12:37:25 2010 +0200 110361 110362 configure: sha1: check libsha1 using pkg-config instead 110363 110364 Previously the code was using AC_CHECK_LIB, guaranteeing whether the library 110365 is correct by tracking sha1_begin function. This paranoic checking is not 110366 necessary given there's only one libsha1 in the market, which surely contains 110367 such function. 110368 110369 Moreover, this patch now improves a bit the sha1 implementation checking 110370 behavior using pkg-config to find the right flags that needs to link against. 110371 110372 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110373 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 110374 110375commit f0ab726d8966cab4e50154c216d577db79328d77 110376Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 110377Date: Wed Apr 21 11:07:47 2010 +0300 110378 110379 DRI2: Use single error path in initialization 110380 110381 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 110382 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 110383 110384commit 6eef70dc56bcc1e3047e4e488bcd6ae62c8ffac6 110385Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110386Date: Fri Jan 22 18:34:36 2010 +0200 110387 110388 DRI2: Allow building without libdrm 110389 110390 Some drivers use DRI protocol but implement their own kernel rendering 110391 manager. For these drivers, libdrm becomes useless. --disable-libdrm 110392 configure parameter can be used to disable libdrm support in dri2. 110393 110394 To provide ABI/API compatibility for libdrm based drivers, libdrm call 110395 is wrapped in ifdef. 110396 110397 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110398 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 110399 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 110400 110401commit cdcb575664d3d60b662c542e782de83a047165c9 110402Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110403Date: Fri Jan 22 17:58:17 2010 +0200 110404 110405 DRI2: add AuthMagic hook for driver side support 110406 110407 With this new hook, drmAuthMagic becomes useless and should be deprecated. 110408 You might want to implement AuthMagic on driver side instead. 110409 110410 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110411 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 110412 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 110413 110414commit 643cb6e87c10ab554c03ada81930001a8ebcc909 110415Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 110416Date: Thu May 27 09:11:50 2010 -0700 110417 110418 Only deal with input code when changing the input shape. 110419 110420 Propagate the shape kind all the way to SetShape to avoid performing non-input 110421 operations such as revalidating the tree and generating exposures when only 110422 changing a window's input shape. 110423 110424 Signed-off-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 110425 Acked-by: Aaron Plattner<aplattner@nvidia.com> 110426 Reviewed-by: Daniel Stone<daniel@fooishbar.org> 110427 Signed-off-by: Keith Packard <keithp@keithp.com> 110428 110429commit d90f2cd98a97e6534792d3867b3fba70d850b706 110430Author: Chris Wilson <chris@chris-wilson.co.uk> 110431Date: Thu May 27 13:26:23 2010 +0100 110432 110433 xfree86: Unbreak autoconfig following 0abf065e38c4 110434 110435 The move of the PCI device id probing into a separate file neglected to 110436 return the number of found devices, and so the PCI devices were being 110437 overwritten by the default entries for vesa and fbdev. 110438 110439 Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> 110440 Cc: Tiago Vignatti <tiago.vignatti@nokia.com> 110441 Cc: Alex Deucher <alexdeucher@gmail.com> 110442 Reviewed-by: Julien Cristau <jcristau@debian.org> 110443 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110444 Signed-off-by: Keith Packard <keithp@keithp.com> 110445 110446commit 1304b8b27cb12c803c4f51f04cb6f9d508b82c69 110447Author: Aaron Plattner <aplattner@nvidia.com> 110448Date: Fri May 28 09:35:54 2010 -0700 110449 110450 Fix pixmap validation in miDbePositionWindow. 110451 110452 miDbePositionWindow allocates two pixmaps: a front buffer, and a back buffer. 110453 If the buffers are supposed to be initialized, it validates a GC against the 110454 front buffer, then uses it to fill and/or copy both the front buffer *and* the 110455 back buffer, without revalidating. If the acceleration architecture needs 110456 different GC funcs for the two pixmaps -- for example if allocation of the front 110457 buffer exhausted video memory -- then this can cause crashes because the GC is 110458 not validated for the back buffer pixmap. 110459 110460 Fix this by performing the rendering for the front buffer first, then 110461 revalidating against the back buffer before performing the back buffer 110462 rendering. 110463 110464 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 110465 Reviewed-by: Jamey Sharp <jamey@minilop.net> 110466 Signed-off-by: Keith Packard <keithp@keithp.com> 110467 110468commit 91a6359caf24d94343ff76f43ea7b7fc3223203d 110469Author: Dave Airlie <airlied@redhat.com> 110470Date: Thu Jun 3 12:16:40 2010 +1000 110471 110472 composite: initialise pOldPixmap to NullPixmap at alloc time. 110473 110474 We just never initialised the malloced value. 110475 110476 Signed-off-by: Dave Airlie <airlied@redhat.com> 110477 Reviewed-by: Keith Packard <keithp@keithp.com> 110478 Signed-off-by: Keith Packard <keithp@keithp.com> 110479 110480commit 968a79dcf5e17ac3963953ef56b8f94dbd75323b 110481Author: Nicolas George <nicolas.george@normalesup.org> 110482Date: Wed Jun 2 13:40:51 2010 +0200 110483 110484 Change keyboard controls on slave keyboards (#27926) 110485 110486 Makes the use of IsMaster in ProcChangeKeyboardControl consistent with other 110487 similar loops. 110488 110489 Signed-off-by: Nicolas George <nicolas.george@normalesup.org> 110490 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 110491 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110492 110493commit 108b766c31b57fb1955d34d85673a235c7f743aa 110494Author: Peter Hutterer <peter.hutterer@who-t.net> 110495Date: Thu May 20 11:04:53 2010 +1000 110496 110497 xfree86: initialize InputAttributes to NULL in the autoconfig code. 110498 110499 Reported-by: Jamey Sharp <jamey@minilop.net> 110500 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110501 Reviewed-by: Jamey Sharp <jamey@minilop.net> 110502 110503commit 151659f9636088fd70bc5586de97bf43ee706180 110504Author: Peter Hutterer <peter.hutterer@who-t.net> 110505Date: Mon May 24 16:24:03 2010 +1000 110506 110507 test: fix up InputAttributes helper function test. 110508 110509 Just some extra clarification as pointed out by Dan Nicholson, 110510 and that memcpy should have been a memcmp. 110511 110512 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110513 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 110514 110515commit 6cccf0131c8464d8838cae2200730873d7dd9e45 110516Author: Peter Korsgaard <peter.korsgaard@barco.com> 110517Date: Tue May 25 11:03:28 2010 +0200 110518 110519 dix: add 3x3 transformation matrix xinput property for multi-head handling 110520 110521 For absolute input devices (E.G. touchscreens) in multi-head setups, 110522 we need a way to bind the device to an randr output. This adds the 110523 infrastructure to the server to allow us to do so. 110524 110525 positionSprite() scales input coordinates to the dimensions of the shared 110526 (total) screen frame buffer, so to restrict motion to an output we need to 110527 scale/rotate/translate device coordinates to a subset of the frame buffer 110528 before passing them on to positionSprite. 110529 110530 This is done here using a 3x3 transformation matrix, which is applied to 110531 the device coordinates using homogeneous coordinates, E.G.: 110532 110533 [ c0 c1 c2 ] [ x ] 110534 [ c3 c4 c5 ] * [ y ] 110535 [ c6 c7 c8 ] [ 1 ] 110536 110537 Notice: As input devices have varying input ranges, the coordinates are 110538 first scaled to the [0..1] range for generality, and afterwards scaled 110539 back up. 110540 110541 E.G. for a dual head setup (using same resolution) next to each other, you 110542 would want to scale the X coordinates of the touchscreen connected to the 110543 both heads by 50%, and translate (offset) the coordinates of the rightmost 110544 head by 50%, or in matrix form: 110545 110546 left: right: 110547 [ 0.5 0 0 ] [ 0.5 0 0.5 ] 110548 [ 0 1 0 ] [ 0 1 0 ] 110549 [ 0 0 1 ] [ 0 0 0 ] 110550 110551 Which can be done using xinput: 110552 110553 xinput set-prop <left> --type=float "Coordinate Transformation Matrix" \ 110554 0.5 0 0 0 1 0 0 0 1 110555 110556 xinput set-prop <right> --type=float "Coordinate Transformation Matrix" \ 110557 0.5 0 0.5 0 1 0 0 0 1 110558 110559 Likewise more complication setups involving more heads, rotation or 110560 different resolution can be handled. 110561 110562 Signed-off-by: Peter Korsgaard <peter.korsgaard@barco.com> 110563 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110564 110565commit e4582d9e5c8649347742a13eae68cf27005296fc 110566Author: Peter Hutterer <peter.hutterer@who-t.net> 110567Date: Thu May 27 10:42:54 2010 -0400 110568 110569 xfree86: bump ABI_XINPUT_VERSION to 11. 110570 110571 commit c2d0b3b437b7ce6ce975f2ead4d8bb8295ef0ddc 110572 "xfree86: store the InputAttributes in the input device." 110573 introduced the new API. Bump the input version so drivers can handle this 110574 appropriately. 110575 110576 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110577 Signed-off-by: Keith Packard <keithp@keithp.com> 110578 110579commit cd0ef0b6a2f9c6bfeda98684569da922792d3feb 110580Author: Jamey Sharp <jamey@minilop.net> 110581Date: Tue May 25 14:30:22 2010 -0700 110582 110583 dmx: Delete unused ChangeKeyboardDevice/ChangePointerDevice. 110584 110585 Daniel Stone deleted the API for these in 2006, in commit 110586 96e32805d12fc36f0fa0926dbfb0dd8a5cadb739. 110587 110588 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110589 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 110590 110591commit 70454864917f534483fa049e067b5646df61c88a 110592Author: Jamey Sharp <jamey@minilop.net> 110593Date: Fri May 14 12:31:44 2010 -0700 110594 110595 Ignore build products in the server's test suite. 110596 110597 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110598 Cc: Peter Hutterer <peter.hutterer@who-t.net> 110599 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 110600 110601commit 8bb88379d95356384e233225bfda6abfd71389b4 110602Author: Jamey Sharp <jamey@minilop.net> 110603Date: Tue May 25 15:52:18 2010 -0700 110604 110605 dmx: Remove dead __glXDispSwap_DrawArraysEXT definition. 110606 110607 Eric Anholt made the corresponding fix in glx/render2swap.c in commit 110608 49d38ab2328c409b2a98465b52677af057121513. 110609 110610 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110611 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110612 110613commit c97a3ade5a8718b063292c73a31aac2196fe2452 110614Author: Jamey Sharp <jamey@minilop.net> 110615Date: Tue May 25 15:17:30 2010 -0700 110616 110617 dmx: Delete unused local variables. 110618 110619 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110620 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 110621 110622commit 3c3d099fc10c63d746f90c08043d45180d6a7226 110623Author: Jamey Sharp <jamey@minilop.net> 110624Date: Tue May 25 13:44:32 2010 -0700 110625 110626 dmx: Xfree -> free 110627 110628 Mikhail fixed the corresponding Xallocs, but missed these uses of Xfree 110629 in commit 3f3ff971ecff9936cebafc813af9193b97bba89c. 110630 110631 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110632 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 110633 110634commit 01ca6400d949a873c0bb5de7e0dbf138e8a7949b 110635Author: Jamey Sharp <jamey@minilop.net> 110636Date: Tue May 25 14:02:34 2010 -0700 110637 110638 Xephyr: Fix Xcalloc deprecation warnings. 110639 110640 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110641 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 110642 110643commit 34df659687ad3936ab03d19ce8fdb21f385c2e8a 110644Author: Jamey Sharp <jamey@minilop.net> 110645Date: Tue May 25 11:51:32 2010 -0700 110646 110647 xfree86/int10/helper_mem: Fix log message. 110648 110649 Three years ago in commit f62beb6f3609e8b6e61325ac89017590811bbd07 ajax 110650 deleted the code that could have set this format string to anything 110651 else, so just use the format string literal. This makes GCC happy since 110652 it can check the argument types, which, by the way, weren't correct 110653 since this format string doesn't need any arguments. 110654 110655 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110656 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110657 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110658 110659commit 0a98d9e6cec7c611a3c56f97d4ddc0c546975c55 110660Author: Jamey Sharp <jamey@minilop.net> 110661Date: Tue May 25 10:30:58 2010 -0700 110662 110663 xf86bigfont: Silence compiler warning by initializing pDesc. 110664 110665 Access to pDesc was always guarded by (nCharInfos > 0), so the code 110666 wasn't actually buggy, but this makes it clear that it's correct. 110667 110668 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110669 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110670 110671commit 4e9d3e413245e7c01c2ff3c0f130b8967754131b 110672Author: Keith Packard <keithp@keithp.com> 110673Date: Wed May 26 07:54:35 2010 -0700 110674 110675 Revert "Add a "flags" field to DeleteInputDeviceRequest." 110676 110677 Peter wants to get a larger patch sequence put together and I didn't 110678 read past the commit message to see the 'don't take this patch 110679 please'. 110680 110681 This reverts commit 531ff40301975519af7b20109c17d296312d3f2b. 110682 110683commit 531ff40301975519af7b20109c17d296312d3f2b 110684Author: Peter Hutterer <peter.hutterer@who-t.net> 110685Date: Thu May 20 18:07:12 2010 +1000 110686 110687 Add a "flags" field to DeleteInputDeviceRequest. 110688 110689 Some input drivers need to implement an internal hotplugging scheme for 110690 dependent devices to provide multiple X devices off one kernel device file. 110691 Such dependent devices can be added with NewInputDeviceRequest() but they are 110692 not removed when the config backend calls DeleteInputDeviceRequest(), 110693 leaving the original device to clean up. 110694 110695 Example of the wacom driver: 110696 110697 config/udev calls NewInputDeviceRequest("stylus") 110698 110699 wacom PreInit calls 110700 NewInputDeviceRequest("eraser") 110701 NewInputDeviceRequest("pad") 110702 NewInputDeviceRequest("cursor") 110703 PreInit finishes. 110704 110705 When the device is removed, the config backend only calls 110706 DeleteInputDeviceRequest for "stylus". The driver needs to call 110707 DeleteInputDeviceRequest for the dependent devices eraser, pad and cursor to 110708 clean up properly. 110709 However, when the server terminates, DeleteInputDeviceRequest is called for 110710 all devices - the driver must not remove the dependent devices to avoid 110711 double-frees. There is no method for the driver to detect why a device is 110712 being removed, leading to elaborate guesswork and some amount of wishful 110713 thinking. 110714 110715 Though the input driver's UnInit already supports flags, they are unused. 110716 This patch uses the flags to supply information where the 110717 DeleteInputDeviceRequest request originates from, allowing a driver to 110718 selectively call DeleteInputDeviceRequest when necessary. 110719 110720 Also bumps XINPUT ABI. 110721 110722 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110723 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 110724 Signed-off-by: Keith Packard <keithp@keithp.com> 110725 110726commit c2d0b3b437b7ce6ce975f2ead4d8bb8295ef0ddc 110727Author: Peter Hutterer <peter.hutterer@who-t.net> 110728Date: Tue May 25 17:15:32 2010 +1000 110729 110730 xfree86: store the InputAttributes in the input device. 110731 110732 InputAttributes largely decide which configuration values get merged from 110733 the xorg.conf.d snippets. While they are available in the config backend, 110734 they are not available for any other callers of NewInputDeviceRequest(). 110735 110736 Drivers implementing driver-side hotplugging do not have access to these 110737 attributes and cannot have xorg.conf.d snippets specific to dependent 110738 devices. For example, the following case cannot work right now: 110739 110740 Section "InputClass" 110741 MatchProduct "Wacom" 110742 Option "PressCurve" "0 0 100 100" 110743 ... 110744 EndSection 110745 110746 Section "InputClass" 110747 MatchProduct "Wacom" 110748 MatchProduct "eraser" 110749 Option "PressCurve" "10 10 50 50" 110750 ... 110751 EndSection 110752 110753 The second section is not triggered, as the wacom driver cannot supply the 110754 InputAttributes to NewInputDeviceRequest(). 110755 110756 Add the attributes to the IDevRec and merge them into the InputInfoRec to 110757 make them accessible in the driver. This changes the ABI for input drivers. 110758 110759 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110760 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 110761 Signed-off-by: Keith Packard <keithp@keithp.com> 110762 110763commit 86303a338ad563d0b986a3c052104301c310c4ac 110764Author: Peter Hutterer <peter.hutterer@who-t.net> 110765Date: Tue May 25 17:12:34 2010 +1000 110766 110767 dix: add helper functions to duplicate and free InputAttributes. 110768 110769 No special memory handling is used to give drivers the maximum flexibility 110770 with the data. Drivers should be able to call realloc on the product string 110771 if needed and perform similar operations. 110772 110773 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110774 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 110775 Signed-off-by: Keith Packard <keithp@keithp.com> 110776 110777commit bf78e11839f8278020b604672ff7c3d194232be9 110778Author: Jamey Sharp <jamey@minilop.net> 110779Date: Mon May 24 09:15:32 2010 -0700 110780 110781 midispcur: Construct Picture objects lazily in case Render is disabled. 110782 110783 Reverts part of the effects of 518f3b189b6c8aa28b62837d14309fd06163ccbb, 110784 "mi: don't thrash resources when displaying the software cursor across 110785 screens". The per-screen cache is preserved, and the GCs are still 110786 allocated eagerly, but now it doesn't construct pRootPicture until 110787 somebody attempts to draw an ARGB cursor. 110788 110789 I noticed crashes in Xnest, which doesn't support the RENDER extension, 110790 but I suspect other DDXes that support disabling that extension would 110791 have had issues as well. 110792 110793 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110794 Reviewed-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 110795 110796commit bf9b64eefbadbf299d3b841c1890bfcaef5aab1b 110797Author: Jamey Sharp <jamey@minilop.net> 110798Date: Mon May 24 09:14:27 2010 -0700 110799 110800 xnest: Don't ignore errors from DeviceCursorInitialize. 110801 110802 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110803 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 110804 110805commit b11465a6ecdc2b8373e6fc8af427edc4602bcaa1 110806Merge: 7c085aebf 0abf065e3 110807Author: Keith Packard <keithp@keithp.com> 110808Date: Mon May 24 22:18:31 2010 -0700 110809 110810 Merge remote branch 'vignatti/bus-cleanup-take2' 110811 110812commit 7c085aebfedeb621a6fbeb3f09f4fcc640452044 110813Author: Jamey Sharp <jamey@minilop.net> 110814Date: Mon May 24 11:45:42 2010 -0700 110815 110816 xfixes: Remove the invisible cursor sprite, using NullCursor instead. 110817 110818 Oliver McFadden reports that the invisible cursor sprite caused damage 110819 events and thus unnecessary redrawing, so removing it improves 110820 performance when using software cursor sprites, especially on those 110821 devices where you do not want a visible cursor: touchscreen tablets, 110822 embedded devices, etc. 110823 110824 For the xfree86 DDX, if hardware cursors are used, the driver is 110825 required to provide a HideCursor function, which will be called instead 110826 of trying to set a null cursor. I think software cursors are already 110827 safe. The other DDXes also look safe. 110828 110829 As far as I can tell, there's no reason to realize a null cursor. I 110830 think everything that handles null cursors doesn't rely on any setup in 110831 RealizeCursor, and treats them as empty cursors. 110832 110833 Xnest assumes that if a cursor is created, it will be realized before it 110834 is freed, which didn't happen if the invisible cursor was never used in 110835 a server generation. So this fixes a segfault in Xnest as well. 110836 110837 Signed-off-by: Jamey Sharp <jamey@minilop.net> 110838 Cc: Oliver McFadden <oliver.mcfadden@nokia.com> 110839 Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com> 110840 Tested-by: Oliver McFadden <oliver.mcfadden@nokia.com> 110841 Signed-off-by: Keith Packard <keithp@keithp.com> 110842 110843commit 0abf065e38c4c48917054eb3e39e12bd20e6e8ec 110844Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110845Date: Thu May 20 19:24:47 2010 +0300 110846 110847 xfree86: move all pci code from auto configurator into a meaningful location 110848 110849 This patch creates the private xf86PciMatchDriver hook, which goes inside pci 110850 code to match the drivers found in the system. 110851 110852 Now there's no direct references to PCI inside xf86AutoConfig.c anymore. 110853 110854 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110855 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110856 110857commit 0ceac6f64f5ad9bc2ac4b19be2dd245ffba78b05 110858Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110859Date: Thu May 20 18:23:24 2010 +0300 110860 110861 xfree86: no need to check for the configuration case when matching devices 110862 110863 xf86MatchDevice will never be called in configuration time. 110864 110865 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110866 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110867 110868commit b0eef1101815b08da2d375af7bec2be3136cf713 110869Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110870Date: Thu May 20 18:19:14 2010 +0300 110871 110872 xfree86: organize and group all pci related stuff inside xf86.h 110873 110874 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110875 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110876 110877commit 47df5a489ea69a68e753367423bfbe8830521f4e 110878Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110879Date: Thu May 20 17:39:59 2010 +0300 110880 110881 xfree86: remove BUS_ISA type given we don't support anymore 110882 110883 Should go together within commit df14682a. 110884 110885 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110886 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110887 110888commit 9d000a5509e0a356ee33ad08e4dc967a2ef8cbe0 110889Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110890Date: Thu May 20 16:52:37 2010 +0300 110891 110892 xfree86: remove all kind of bus and PCI dependency from the common helper file 110893 110894 Move all PCI procedures from xf86Helper.c to a more meaningful place (namely 110895 xf86pciBus.c). xf86Helper.c is free of PCI code now. 110896 110897 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110898 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110899 110900commit 610009cf39984f2f5e818a221b626f7af10bd90e 110901Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110902Date: Tue May 18 20:54:28 2010 +0300 110903 110904 xfree86: bus: remove useless field from EntityRec 110905 110906 RAC is the champion of remaining trash for sure! 110907 110908 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110909 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110910 Reviewed-by: Keith Packard <keithp@keithp.com> 110911 110912commit 95889a569feabd3f803afb92950d067bd76e7397 110913Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110914Date: Tue May 18 20:48:35 2010 +0300 110915 110916 xfree86: bus: delete useless xf86FindPrimaryDevice 110917 110918 This function had a wrong name and was just logging the primary device. No one 110919 cares about it honestly. 110920 110921 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110922 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110923 110924commit f6865909485d7d282c8131a2ae306143ad51ba2b 110925Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110926Date: Tue May 18 19:39:31 2010 +0300 110927 110928 xfree86: bus: remove unused headers 110929 110930 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110931 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110932 110933commit 111b1fff1ca69724a8350c4375089c312329c378 110934Author: Tiago Vignatti <tiago.vignatti@nokia.com> 110935Date: Tue May 18 19:35:10 2010 +0300 110936 110937 xfree86: bus: remove unused pci macros 110938 110939 Should be gone in commits 3c03d9f1 and a9d7d659a respectively. 110940 110941 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 110942 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110943 110944commit 2ffffb4daf6161e6a22d81442ecf6209acc9e975 110945Merge: b5e0f6d8f d5306084b 110946Author: Keith Packard <keithp@keithp.com> 110947Date: Sun May 23 23:22:08 2010 -0700 110948 110949 Merge remote branch 'alanc/docs' 110950 110951commit b5e0f6d8f45c5b24eb50b305c66fa80c783ef488 110952Author: Peter Hutterer <peter.hutterer@who-t.net> 110953Date: Mon May 24 12:25:15 2010 +1000 110954 110955 xfree86: fix multiple InputAttributes tag matching. 110956 110957 attr->tags is an array of strings (null-terminated). When matching, match 110958 against each string instead of each [i,end] substring in the first tag. 110959 110960 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 110961 Reviewed-by: Jamey Sharp <jamey@minilop.net> 110962 Signed-off-by: Keith Packard <keithp@keithp.com> 110963 110964commit d5306084b57583c670c56ce9e7d3c78cca7aa07b 110965Author: Brice Goglin <bgoglin@debian.org> 110966Date: Sat May 22 20:29:02 2010 +0200 110967 110968 Remove obsolete reference to README.DRI in xorg.conf.man 110969 110970 Signed-off-by: Brice Goglin <Brice.Goglin@ens-lyon.org> 110971 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 110972 110973commit 89dd7d2b6fe18f32c2f9336c3d8cd9f35490e5ae 110974Author: Alan Coopersmith <alan.coopersmith@oracle.com> 110975Date: Sat May 22 10:33:04 2010 -0700 110976 110977 Remove ancient documentation of IBM RapidAccess keyboard hack 110978 110979 Keyboard setup belongs in drivers, not in a document no one sees 110980 because we don't even install it. 110981 110982 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 110983 Reviewed-by: Julien Cristau <jcristau@debian.org> 110984 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 110985 110986commit 480fcdf0335abae0ee544b022c7985211eee462f 110987Author: Alan Coopersmith <alan.coopersmith@oracle.com> 110988Date: Sat May 22 10:29:47 2010 -0700 110989 110990 Remove completely out-of-date README.DRI 110991 110992 The license only allows distribution of verbatim copies, so we can't 110993 update it, even to correct the incorrect address to send updates to. 110994 110995 The Mesa & DRI web pages are much better sources of current information 110996 anyway. 110997 110998 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 110999 Reviewed-by: Julien Cristau <jcristau@debian.org> 111000 111001commit 1805c74d9b0ed46802481bcd6beea5feeb60212c 111002Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111003Date: Sat May 22 10:29:00 2010 -0700 111004 111005 Add RandR 1.2 README.modes doc to EXTRA_DIST 111006 111007 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111008 Reviewed-by: Julien Cristau <jcristau@debian.org> 111009 111010commit 5939e39a641773a36c22104e1184143678dca7a2 111011Author: Dave Airlie <airlied@redhat.com> 111012Date: Mon May 24 09:31:51 2010 +1000 111013 111014 xf86: allow for no outputs connected at startup operation. 111015 111016 When nothing is connected at startup and we canGrow, allow the server to start with a 1024x768 framebuffer, and when the drivers send hotplug events this will expand to the correct size dynamically. 111017 111018 Reviewed-by: Keith Packard <keithp@keithp.com> 111019 Signed-off-by: Dave Airlie <airlied@redhat.com> 111020 Signed-off-by: Keith Packard <keithp@keithp.com> 111021 111022commit b9f48d60bc0c839bd323c582231e8e7e2b810af6 111023Author: Jamey Sharp <jamey@minilop.net> 111024Date: Wed May 19 10:44:33 2010 -0700 111025 111026 Device init: Don't crash when CreateGC fails. 111027 111028 ActivateDevice was ignoring errors from DeviceCursorInitialize, so 111029 cursor-related calls failed later. Jeremy Huddleston saw that crash in 111030 miPointerConstrainCursor, while with Xvfb I saw it in 111031 miSpriteRealizeCursor. 111032 111033 miDCDeviceCleanup frees any non-NULL GCs. miDCDeviceInitialize calls 111034 Cleanup on any failure, but if it failed early then some of the pointers 111035 in the miDCBufferPtr were garbage. Switch from malloc to calloc to 111036 ensure everything's initialized safely first. 111037 111038 With these two fixes, if CreateGC fails then the server gracefully fails 111039 in FatalError instead of segfaulting. 111040 111041 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111042 Cc: Peter Hutterer <peter.hutterer@who-t.net> 111043 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 111044 Signed-off-by: Keith Packard <keithp@keithp.com> 111045 111046commit c51534961e1bb4c68ff881758862d2f85f572ce7 111047Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111048Date: Sat May 22 11:11:48 2010 -0700 111049 111050 xorg-server.pc Requires: xproto >= 7.0.17 111051 111052 Since the headers we export in the SDK use the _X_NORETURN #define 111053 introduced in xproto 7.0.17, drivers built with the SDK will also 111054 need to have at least that version of xproto installed. 111055 111056 I considered exporting the version from configure.ac, but decided not 111057 to since the minimum required to build the X server may not always 111058 be the same as the minimum required to build drivers (for instance, 111059 if the X server used a new #define in its .c files or internal headers, 111060 but didn't export it in any of the SDK headers). 111061 111062 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111063 Reviewed-by: Julien Cristau <jcristau@debian.org> 111064 Signed-off-by: Keith Packard <keithp@keithp.com> 111065 111066commit e90f20eed394de2200aba42bd61fd31c1067ac33 111067Author: Jamey Sharp <jamey@minilop.net> 111068Date: Fri May 21 22:16:50 2010 -0700 111069 111070 Declare functions that unconditionally call FatalError as _X_NORETURN. 111071 111072 For AtomError, this should fix a clang warning; in the other cases it's 111073 just good documentation. 111074 111075 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111076 Cc: Jeremy Huddleston <jeremyhu@apple.com> 111077 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 111078 Signed-off-by: Keith Packard <keithp@keithp.com> 111079 111080commit 9d31d5b639c2a6ab8b1a80a0e2f0ee2c4f8cddd3 111081Author: Adam Jackson <ajax@redhat.com> 111082Date: Mon May 10 11:48:31 2010 -0400 111083 111084 vfb: Remove dead variable and header file 111085 111086 Signed-off-by: Adam Jackson <ajax@redhat.com> 111087 Reviewed-by: Jamey Sharp <jamey@minilop.net> 111088 Signed-off-by: Keith Packard <keithp@keithp.com> 111089 111090commit d8cbcbc01d1355d55c1c0af64f7f7f61629f6c61 111091Author: Jamey Sharp <jamey@minilop.net> 111092Date: Fri May 21 11:12:26 2010 -0700 111093 111094 doPolyText: forget about FontChange's XID after looking up pFont. 111095 111096 As of e2929db7b737413cf93fbebdf4d15abdfebff05c, doPolyText uses pFont 111097 consistently rather than looking it up again from the saved XID. 111098 111099 clang noticed that "oldfid = fid" could run when fid hadn't been 111100 initialized yet. 111101 111102 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111103 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 111104 Tested-by: Jeremy Huddleston <jeremyhu@apple.com> 111105 Signed-off-by: Keith Packard <keithp@keithp.com> 111106 111107commit 989db930d739483759087b13b8d9a043299feafb 111108Author: Aaron Zang <Aaron.Zang@Sun.COM> 111109Date: Thu May 20 17:56:28 2010 -0700 111110 111111 Solaris: Use VT_SET_CONSUSER ioctl to set Console User rights profile 111112 111113 When Xorg is started on display :0, this ioctl is called to grant the 111114 user the rights traditionally associated with /dev/console (before VT 111115 support was added), such as access to local peripheral devices. 111116 111117 Also adds a Solaris-specific -C flag to force starting on /dev/console 111118 instead of /dev/vt*, allowing programs like xterm -C to access the 111119 console device. 111120 111121 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111122 Reviewed-by: Adam Jackson <ajax@redhat.com> 111123 Signed-off-by: Keith Packard <keithp@keithp.com> 111124 111125commit 98553e52a14b97a03aef8dc4fc0300b3f4c2c4b5 111126Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111127Date: Thu May 20 17:56:27 2010 -0700 111128 111129 vbe.h: Use __attribute__((packed)) on Sun cc 5.9 & later as well as gcc 111130 111131 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111132 Reviewed-by: Adam Jackson <ajax@redhat.com> 111133 Signed-off-by: Keith Packard <keithp@keithp.com> 111134 111135commit 7b09335a46f9428141811230c69eef7968531359 111136Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111137Date: Thu May 20 17:56:26 2010 -0700 111138 111139 Xserver(1) man page updates 111140 111141 - Note that -br is now default. 111142 - Move -bs after -br for alphabetical ordering. 111143 - Remove -config option that's been hidden in "ignore" section, 111144 since ajax removed the -config code a couple years back. 111145 - Add -nocursor option. 111146 - Add xinput & xrandr to list of runtime server control programs 111147 - Replace XDarwin with Xquartz in list of Xservers 111148 111149 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111150 Reviewed-by: Adam Jackson <ajax@redhat.com> 111151 Signed-off-by: Keith Packard <keithp@keithp.com> 111152 111153commit cf4f3d051858aadedd6e333bb317a1daa3987ad7 111154Author: Peter Hutterer <peter.hutterer@who-t.net> 111155Date: Fri May 21 13:30:25 2010 +1000 111156 111157 dix: remove obsolete comment.from EnableDevice. 111158 111159 The code this comment was referring to was removed in 111160 8b5086250aa5dae8de8b763408ff480d7beac819 "Eliminate bogus event resizing." 111161 111162 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 111163 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 111164 Signed-off-by: Keith Packard <keithp@keithp.com> 111165 111166commit 5754e66044571d549c295b7c9e02ce3348dbe3c7 111167Author: Jamey Sharp <jamey@minilop.net> 111168Date: Thu May 20 10:46:43 2010 -0700 111169 111170 Replace screen->rgf scratch GC flags with a bit in each GC. 111171 111172 This eliminates a poorly-named, poorly-documented field from the 111173 ScreenRec, using a previously-unused flag bit in each GC instead. 111174 111175 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111176 Cc: Keith Packard <keithp@keithp.com> 111177 Reviewed-by: Keith Packard <keithp@keithp.com> 111178 Signed-off-by: Keith Packard <keithp@keithp.com> 111179 111180commit 5a7275d78a2f1c20ed5bb7b228cf370c4ada22c9 111181Author: Keith Packard <keithp@keithp.com> 111182Date: Thu May 20 13:57:02 2010 -0700 111183 111184 Find windows GL wrapper files in $(srcdir) instead of . 111185 111186 This fixes 'make distcheck' for me. 111187 111188 Signed-off-by: Keith Packard <keithp@keithp.com> 111189 Reviewed-by: Julien Cristau <jcristau@debian.org> 111190 111191commit 1197a87b5638037cc48084fba4eec9d3b599e900 111192Author: Keith Packard <keithp@keithp.com> 111193Date: Thu May 20 13:40:24 2010 -0700 111194 111195 winpriv.h was moved from hw/xwin to hw/xwin/glx in November 111196 111197 hw/xwin/Makefile.am was referencing this instead of hw/xwin/glx/Makefile.am 111198 111199 Signed-off-by: Keith Packard <keithp@keithp.com> 111200 Reviewed-by: Julien Cristau <jcristau@debian.org> 111201 111202commit 1cad520f3f1d9d520422e3e4d5d8f4ea85a212a4 111203Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111204Date: Mon May 17 20:35:10 2010 -0700 111205 111206 XFree86 Design doc: Convert LinuxDoc ``quotes'' to DocBook <quote> tags 111207 111208 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111209 Tested-by: Gaetan Nadon <memsize@videotron.ca> 111210 111211commit 8367913be3305be4f03d095615d3885107094427 111212Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111213Date: Fri May 14 22:05:50 2010 -0700 111214 111215 XFree86 Design doc: Explain this version covers the current Xorg release 111216 111217 Update the title & preface to explain that while this was originally 111218 the XFree86 4.0 design, we've changed a lot since forking. 111219 111220 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111221 Tested-by: Gaetan Nadon <memsize@videotron.ca> 111222 111223commit 78fe4be4c43db1f07a30e8a6c9a17f81c1082d76 111224Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111225Date: Mon May 17 20:27:10 2010 -0700 111226 111227 Show Xserver release/version date in DIX & DDX docs 111228 111229 Uses a fake absolute path to the entity definition files so that 111230 the xmlto --searchpath will work for finding the actual path 111231 111232 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111233 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 111234 Tested-by: Gaetan Nadon <memsize@videotron.ca> 111235 111236commit 2bdc73bcf41ff0c10e70ecc15fc8b8333af6db65 111237Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111238Date: Fri May 14 15:33:46 2010 -0700 111239 111240 Move Xserver-spec.xml from xorg-docs to doc/xml 111241 111242 Allows keeping it in sync with the sources it documents, and to be 111243 released with them 111244 111245 Requires the previous patch to convert the Xserver tree from 111246 LinuxDoc to DocBook for SGML/XML documents. 111247 111248 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111249 Reviewed-by: Rémi Cardona <remi@gentoo.org> 111250 Tested-by: Gaetan Nadon <memsize@videotron.ca> 111251 111252commit fc6ebe1e1d3057378f61f992549a98e67a04dc6c 111253Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111254Date: Fri May 14 14:56:09 2010 -0700 111255 111256 Convert LinuxDoc documents to DocBook/XML 111257 111258 Only the markup/formatting is changed - the contents should still 111259 be wildly out of date for now. 111260 111261 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111262 Reviewed-by: Rémi Cardona <remi@gentoo.org> 111263 Tested-by: Gaetan Nadon <memsize@videotron.ca> 111264 111265commit ebd745ced89b2a2d9f6b4dcbd9f5f7e7f3b35451 111266Author: Simon Farnsworth <simon.farnsworth@onelan.com> 111267Date: Thu May 20 09:58:38 2010 +0100 111268 111269 Add documentation for the new DefaultModes option 111270 111271 This patch adds documentation for the DefaultModes monitor option 111272 added in the previous patch. 111273 111274 Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.com> 111275 Reviewed-by: Adam Jackson <ajax@redhat.com> 111276 Signed-off-by: Keith Packard <keithp@keithp.com> 111277 111278commit cfc629148891be317ba60f2495abcda485780191 111279Author: Simon Farnsworth <simon.farnsworth@onelan.com> 111280Date: Thu May 20 09:58:37 2010 +0100 111281 111282 Add configuration option to disable default modes on an output 111283 111284 Setting 'Option "DefaultModes" "No"' in an output will guarantee that 111285 X will omit the default modes, just picking up modes from the 111286 configuration file and modes provided by the output. 111287 111288 Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.com> 111289 Reviewed-by: Adam Jackson <ajax@redhat.com> 111290 Signed-off-by: Keith Packard <keithp@keithp.com> 111291 111292commit a1c2bdb6ab7a4f374844f80e83620b155991e817 111293Author: Michel Dänzer <daenzer@vmware.com> 111294Date: Thu May 20 10:13:06 2010 +0200 111295 111296 EXA: Wrap Glyphs even without Composite acceleration. 111297 111298 In order to avoid migration ping-pong when accumulating glyphs in a mask 111299 picture. 111300 111301 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 111302 Reviewed-by: Adam Jackson <ajax@redhat.com> 111303 Signed-off-by: Keith Packard <keithp@keithp.com> 111304 111305commit 00581d4a42e2a0e7652834973645a4a9ed59cf93 111306Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111307Date: Fri May 14 19:31:56 2010 +0300 111308 111309 configure: force new xproto version dependency 111310 111311 commit bca85e2e127a8a23e3a2debcfeb3ae07cd3c66ac introduced it. 111312 111313 Use the version that includes _X_NORETURN 111314 111315 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111316 Reviewed-by: Julien Cristau <jcristau@debian.org> 111317 Signed-off-by: Keith Packard <keithp@keithp.com> 111318 111319commit 103507af0c0ce7d7482a67163249864af36a2374 111320Merge: 8bd8d81dc 99fcf655b 111321Author: Keith Packard <keithp@keithp.com> 111322Date: Wed May 19 22:27:53 2010 -0700 111323 111324 Merge remote branch 'vignatti/bus-cleanup' 111325 111326commit 8bd8d81dc473bf72ea108b1896a55e32defa02e2 111327Merge: c620a1c0a 66d5ecc5f 111328Author: Keith Packard <keithp@keithp.com> 111329Date: Wed May 19 22:27:20 2010 -0700 111330 111331 Merge remote branch 'vignatti/animcursor-state-fix' 111332 111333commit c620a1c0a4ca564e778cce019dd30046c6969956 111334Author: Peter Hutterer <peter.hutterer@who-t.net> 111335Date: Thu May 20 10:26:56 2010 +1000 111336 111337 xfree86: fix typo in optionTypeToSting. 111338 111339 Because we don't want anyone to get hurt. 111340 111341 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 111342 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 111343 Reviewed-by: Matt Turner <mattst88@gmail.com> 111344 Signed-off-by: Keith Packard <keithp@keithp.com> 111345 111346commit 8b6c1809c09f832051327d86e1a25dc0ec5cc878 111347Merge: d88ba7721 c38552d11 111348Author: Keith Packard <keithp@keithp.com> 111349Date: Wed May 19 12:58:02 2010 -0700 111350 111351 Merge remote branch 'jamey/for-keith' 111352 111353commit c38552d115e3bc71ad6179a8ad0d68778e943793 111354Author: Jamey Sharp <jamey@minilop.net> 111355Date: Sat Apr 24 23:56:36 2010 -0700 111356 111357 Add typed resource-lookup errors for non-core resource types. 111358 111359 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111360 Reviewed-by: Adam Jackson <ajax@redhat.com> 111361 111362commit e291c561821ae86b7dd74269d5cd29bc31703962 111363Author: Jamey Sharp <jamey@minilop.net> 111364Date: Sat Apr 24 23:26:40 2010 -0700 111365 111366 Return an appropriately-typed error from dixLookupResourceByType. 111367 111368 Rather than always returning BadValue, associate an error status like 111369 BadWindow with a resource type like RT_WINDOW, and return the 111370 appropriate one for the requested type. 111371 111372 This patch only touches the core protocol resource types. Others still 111373 return BadValue and need to be mapped appropriately. 111374 111375 dixLookupResourceByType can now return BadImplementation, if the caller 111376 asked for a resource type that has not been allocated in the server. 111377 111378 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111379 Reviewed-by: Adam Jackson <ajax@redhat.com> 111380 111381commit 90e612dcbe370da095d317fac62c80ac2447fa0b 111382Author: Jamey Sharp <jamey@minilop.net> 111383Date: Sat May 15 15:51:32 2010 -0700 111384 111385 Use WriteEventsToClient rather than TryClientEvents where possible. 111386 111387 If filter is NoEventMask (aka CantBeFiltered), grab is null, and the 111388 first event is not in the set of "critical events", then TryClientEvents 111389 simply calls WriteEventsToClient. In that case, it returns 0 for fake or 111390 dead clients, and 1 otherwise. Inline for this special case. 111391 111392 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111393 Reviewed-by: Julien Cristau <jcristau@debian.org> 111394 Reviewed-by: Adam Jackson <ajax@redhat.com> 111395 111396commit 8033fb6c9792820a82fbdff6a14ff8a7a141ba74 111397Author: Jamey Sharp <jamey@minilop.net> 111398Date: Sat May 15 13:12:44 2010 -0700 111399 111400 Set event sequence number in WriteEventsToClient instead of at callers. 111401 111402 TryClientEvents already did this; this commit just moves the assignment 111403 one level down so that no event source has to worry about sequence 111404 numbers. 111405 111406 ...No event source, that is, except XKB, which inexplicably calls 111407 WriteToClient directly for several events. 111408 111409 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111410 Reviewed-by: Julien Cristau <jcristau@debian.org> 111411 Reviewed-by: Adam Jackson <ajax@redhat.com> 111412 111413commit 4b9600a4167427a8fe88bca9b738c9a99cac9469 111414Author: Jamey Sharp <jamey@minilop.net> 111415Date: Sat May 15 12:31:34 2010 -0700 111416 111417 Make WriteEventsToClient/WriteToClient no-op on fake or dead clients. 111418 111419 This matches the test in TryClientEvents, and is a superset of tests 111420 done by the callers of these functions. The consequence of forgetting 111421 these tests is a server crash, so they're always desirable. In my 111422 opinion, it's better to not require the callers to remember to do these 111423 checks. 111424 111425 For callers that don't do very much work before calling WriteToClient or 111426 WriteEventsToClient, I've removed the redundant checks. 111427 111428 hw/xquartz/xpr/appledri.c has an interesting case: While its check for 111429 "client == NULL" appears redundant with the test in WriteEventsToClient, 111430 it dereferences client to get the sequence number. 111431 111432 Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=27497 111433 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111434 Cc: Chris Wilson <chris@chris-wilson.co.uk> 111435 Reviewed-by: Julien Cristau <jcristau@debian.org> 111436 Reviewed-by: Adam Jackson <ajax@redhat.com> 111437 111438commit ee9cd2df4a5af66c6b7073563785b8aef61fcdb8 111439Author: Jamey Sharp <jamey@minilop.net> 111440Date: Sat May 15 13:37:34 2010 -0700 111441 111442 xv: TryClientEvents already checks client and sets sequenceNumber. 111443 111444 So don't bother doing those steps again. 111445 111446 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111447 Reviewed-by: Julien Cristau <jcristau@debian.org> 111448 Reviewed-by: Adam Jackson <ajax@redhat.com> 111449 111450commit e1e7deefb59cfdd82c9de46a7ca241d42436b2f8 111451Author: Jamey Sharp <jamey@minilop.net> 111452Date: Tue May 18 16:49:21 2010 -0700 111453 111454 mi/misprite.c: Delete unused private lookups. 111455 111456 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111457 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 111458 111459commit 936b09f2a3e31afe41050c00736bbb6b812c7003 111460Author: Jamey Sharp <jamey@minilop.net> 111461Date: Tue May 18 16:48:08 2010 -0700 111462 111463 Replace miSpriteCursorFuncRec with direct calls to midispcur.c. 111464 111465 Nobody wraps the mi software-cursor sprite rendering implementations. 111466 111467 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111468 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 111469 111470commit 66d5ecc5fd36e21715531a39ac83dc2b6b988a7e 111471Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111472Date: Tue May 18 13:53:29 2010 +0300 111473 111474 render: set anim cursor state for pointer enabled devices only 111475 111476 The structure containing the state of animated cursor was amended within 111477 SpriteInfoRec, removing all previously privates logic to keep such state. 111478 111479 API change: It was removed MAXDEVICES dependency \o/ 111480 111481 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111482 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 111483 Reviewed-by: Keith Packard <keithp@keithp.com> 111484 111485commit 99fcf655bdc96f7e5b2249b05da51ba4fc9190c7 111486Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111487Date: Sun May 2 22:34:44 2010 +0300 111488 111489 xfree86: remove PCI dependency from InitOutput 111490 111491 All functions that touch PCI and BUS were moved to their own files, organizing 111492 the mess inside the InitOutput. Now, inside InitOutput, mostly accesses to 111493 buses are coordinated by the new xf86BusConfig. 111494 111495 Two PCI probe functions just changed the name and a procedure to receive the 111496 isolate devices parameters was created also, named xf86PciIsolateDevice. 111497 111498 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111499 Reviewed-by: Adam Jackson <ajax@redhat.com> 111500 111501commit 0dd299864a133d478e4635bd1cd305b67fc3964d 111502Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111503Date: Sun May 2 21:33:26 2010 +0300 111504 111505 xfree86: remove xf86EnableAccess 111506 111507 It's RAC remnant. This was substituted by xf86VGAarbiter{Lock, Unlock} 111508 mechanism. 111509 111510 It's an API break, but the few drivers using it were covered already with 111511 macros to avoid problems. 111512 111513 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111514 Reviewed-by: Adam Jackson <ajax@redhat.com> 111515 111516commit f28515b5f774b83e0481acbcdda2f682738079b9 111517Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111518Date: Sun May 2 21:24:46 2010 +0300 111519 111520 xfree86: bus: remove SetSIGIOForState and simplify the code 111521 111522 Also removed some dumb debug messages. 111523 111524 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111525 Reviewed-by: Adam Jackson <ajax@redhat.com> 111526 111527commit 211ca67e4d8d00b20a74a78304a61d4dbea62706 111528Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111529Date: Sun May 2 21:16:23 2010 +0300 111530 111531 xfree86: bus: fb drivers might want to use vga arbitration either 111532 111533 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111534 Reviewed-by: Adam Jackson <ajax@redhat.com> 111535 111536commit 2464eb7d56407422e033ab3b94634ba802a119d5 111537Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111538Date: Sun May 2 21:14:53 2010 +0300 111539 111540 xfree86: bus: simplify entity related hooks 111541 111542 Remove some out dated commentaries either. 111543 111544 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111545 Reviewed-by: Adam Jackson <ajax@redhat.com> 111546 111547commit 971768e6884f20981652da09a1617c51a760fb86 111548Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111549Date: Fri May 7 15:25:48 2010 +0300 111550 111551 xfree86: bus: enable declaration of sparc function as its code usage 111552 111553 The code in xf86sbusBus.c seems too OS-specific to be usable on OpenBSD. 111554 111555 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111556 Reviewed-by: Adam Jackson <ajax@redhat.com> 111557 111558commit a70ec0d136bbb9efb6f4e683d7ebcea00375083b 111559Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111560Date: Sun May 2 20:48:01 2010 +0300 111561 111562 xfree86: bus: rework xf86PostProbe logic and remove useless log info 111563 111564 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111565 Reviewed-by: Adam Jackson <ajax@redhat.com> 111566 111567commit 2f7d630a1ee446711288af69711d57ea3054b594 111568Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111569Date: Sun May 2 20:29:11 2010 +0300 111570 111571 xfree86: bus: reuse already assigned variable when fb driver claimed 111572 111573 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111574 Reviewed-by: Adam Jackson <ajax@redhat.com> 111575 111576commit 964f29bb80c1bb05508a27969c3ac34cbf072ee9 111577Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111578Date: Sun May 2 20:14:40 2010 +0300 111579 111580 xfree86: bus: fix Enter/Leave accesses behaviour 111581 111582 Okay, seems we're not using extensively such hooks in fact. But fix the 111583 expected behaviour at least. 111584 111585 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111586 Reviewed-by: Adam Jackson <ajax@redhat.com> 111587 111588commit b37462b4da7c389ec7bbcfd800e5f88bded79330 111589Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111590Date: Sun May 2 19:53:37 2010 +0300 111591 111592 xfree86: remove unused xf86AccessInit() 111593 111594 The function was only initializing the boolean xf86ResAccessEnter, which 111595 couldn't get any other value in the life of the server. 111596 111597 The only possible, though suspicious, code was in xf86AccessLeave(), which 111598 could be triggered if AbortDDX is called before xf86AccessInit(). Even so, 111599 such change is safety because no driver would have configured any entity leave 111600 procedure at this point. 111601 111602 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111603 Reviewed-by: Adam Jackson <ajax@redhat.com> 111604 111605commit d88ba7721d2d3b58cdc664fd4c23a3c5e2a5f909 111606Author: Peter Hutterer <peter.hutterer@who-t.net> 111607Date: Tue May 18 11:12:49 2010 +1000 111608 111609 xfree86: Add option parsing for percent options. 111610 111611 In some cases, an option of "50%" would be preferable over fixed value 111612 configuration - especially if the actual values are autoprobed. 111613 Add a new set of functions to parse percent values from configurations. 111614 111615 The percent value parsing differs slightly - if the option is not to marked 111616 as used (e.g. xf86CheckPercentOption()), no warning is emitted to the log 111617 file if the value is not a percent value. This allows double-options (either 111618 as % or as absolute number) without warnings. 111619 111620 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 111621 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 111622 Signed-off-by: Keith Packard <keithp@keithp.com> 111623 111624commit 673eb707ce6737284c4886265ba149c5587a74e2 111625Author: Jamey Sharp <jamey@minilop.net> 111626Date: Sat May 15 12:06:25 2010 -0700 111627 111628 SyncSendAlarmNotifyEvents: check the correct client's clientGone flag. 111629 111630 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111631 Reviewed-by: Keith Packard <keithp@keithp.com> 111632 Signed-off-by: Keith Packard <keithp@keithp.com> 111633 111634commit c3ba199aa643e59abd30170c53fbb4e03077c377 111635Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111636Date: Tue Apr 27 16:27:55 2010 +0300 111637 111638 render: remove unused animcursor fields from private rec 111639 111640 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111641 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 111642 111643commit 9db63e4766bde07ebc375c4390bcc5c854b46ac3 111644Author: Tiago Vignatti <tiago.vignatti@nokia.com> 111645Date: Tue Apr 27 16:18:50 2010 +0300 111646 111647 render: removed unused macro from animcursor 111648 111649 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 111650 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 111651 111652commit cba9942f7716b12c329616a9994c5bce857617fa 111653Author: Kristian Høgsberg <krh@bitplanet.net> 111654Date: Mon May 17 17:40:30 2010 -0400 111655 111656 dri2: Fix NULL check in allocation error handling 111657 111658 1461b31e40a3e44f6327199d23cf2ab70d8f5176 tests to see if the allocation 111659 of buffer[i] failed after i has been incremented, causing it to look 111660 at the wrong pointer. 111661 111662 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 111663 Tested-by: Eric Anholt <eric@anholt.net> 111664 Signed-off-by: Keith Packard <keithp@keithp.com> 111665 111666commit 27cec0a0501369a68031c048e691bda4c7095f72 111667Author: Julien Cristau <jcristau@debian.org> 111668Date: Mon May 17 19:23:44 2010 +0200 111669 111670 Fix build without XACE 111671 111672 dix/colormap.c and dix/gc.c now dereference a ClientPtr, so they need to 111673 include dixstruct.h. Regression introduced by commit 111674 11c69880c7c48ef9e755c4e09fadef7a629d7bc7. 111675 111676 Reported-by: Robert Hooker <sarvatt@ubuntu.com> 111677 Signed-off-by: Julien Cristau <jcristau@debian.org> 111678 Reviewed-by: Jamey Sharp <jamey@minilop.net> 111679 Signed-off-by: Keith Packard <keithp@keithp.com> 111680 111681commit 5a387cf6d248e3cb3337a938200ed5a09fc1d8cb 111682Author: Julien Cristau <jcristau@debian.org> 111683Date: Sun May 16 21:29:02 2010 +0200 111684 111685 Only link Xorg with libconfig.la 111686 111687 Other DDXs don't use input hotplugging since config_init was moved to 111688 the DDX in commit d33adcdf03c69407d151e732fa0cf9947151eb19, so there's 111689 no need to link this in. 111690 111691 Signed-off-by: Julien Cristau <jcristau@debian.org> 111692 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 111693 Signed-off-by: Keith Packard <keithp@keithp.com> 111694 111695commit 37f0b45dbe4ea52bc127720207c7157b2b803ac0 111696Author: Jeremy Huddleston <jeremyhu@apple.com> 111697Date: Sun May 16 10:14:02 2010 -0700 111698 111699 XQuartz: Fix a build failure on Tiger 111700 111701 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 111702 111703commit ecfeabec8d0dcfe286fb893047f1fe1a7ea9f8f5 111704Author: Jeremy Huddleston <jeremyhu@apple.com> 111705Date: Sun May 16 10:03:13 2010 -0700 111706 111707 XQuartz: Don't use deltaXY for determining pointer location on scroll events 111708 111709 <rdar://problem/7989690> 111710 111711 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 111712 Reviewed-by: Edward Moy <emoy@apple.com> 111713 111714commit a911292c85f7069d2caabcb677ed716a04227526 111715Author: Jeremy Huddleston <jeremyhu@apple.com> 111716Date: Sat May 15 10:53:09 2010 -0700 111717 111718 XQuartz: Don't trust deltaXY for middle mouse clicks. 111719 111720 The middle mouse clicks return erroneous values after returning from 111721 Fast User Switching. 111722 111723 <rdar://problem/7979468> 111724 http://xquartz.macosforge.org/trac/ticket/389 111725 111726 Signed-off-by: Martin Otte <otte@duke.edu> 111727 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 111728 Reviewed-by: Edward Moy <emoy@apple.com> 111729 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 111730 111731commit 345eb171264325d73ea2c50ba8c692cf589c2a9b 111732Author: Keith Packard <keithp@keithp.com> 111733Date: Sat May 1 01:41:19 2010 -0700 111734 111735 Remove dbe screen private privates. 111736 111737 Each DBE Screen private structure could have nested privates. Oddly, 111738 no code ever used them. 111739 111740 Signed-off-by: Keith Packard <keithp@keithp.com> 111741 Reviewed-by: Jamey Sharp <jamey@minilop.net> 111742 111743commit 87ea5760f86eb60840e6e2c10012915952df5377 111744Author: Keith Packard <keithp@keithp.com> 111745Date: Mon May 10 11:46:25 2010 -0700 111746 111747 Check pixmap allocation return value when creating glyphs 111748 111749 The driver may decide that the pixmap is too large or something and 111750 fail to allocate a pixmap; not checking would lead to a segfault. 111751 111752 Signed-off-by: Keith Packard <keithp@keithp.com> 111753 Reviewed-by: Jamey Sharp <jamey@minilop.net> 111754 111755commit 03f7e92a6cbfe58db08ce840283f9809695c70e7 111756Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111757Date: Mon May 10 13:44:30 2010 -0700 111758 111759 Update list of supported platforms in configure.ac error message 111760 111761 Adds DragonFly BSD, OpenSolaris, & GNU Hurd. 111762 Drops MacOS X, since this is in the section specific to the Xorg/XFree86 DDX. 111763 (Matches the OS patterns the configure script checks for.) 111764 111765 Also uses m4 macros to fix the spacing/formatting of the resulting message. 111766 111767 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111768 Acked-by: Gaetan Nadon <memsize@videotron.ca> 111769 111770commit e42a29d269fadc11e065c63ee369e3165196f2d7 111771Author: Alan Coopersmith <alan.coopersmith@oracle.com> 111772Date: Thu Apr 29 18:45:34 2010 -0700 111773 111774 Fix compiler issues with getifaddrs() call on OpenSolaris 111775 111776 OpenSolaris recently added support for the getifaddrs() API. 111777 111778 Building with that uncovered two compiler issues (one warning, one error) 111779 in the code that was now being built for the first time in our builds: 111780 111781 "access.c", line 768: warning: argument #1 is incompatible with prototype: 111782 prototype: pointer to struct sockaddr {unsigned short sa_family, array[14] of char sa_data} : "access.c", line 213 111783 argument : pointer to struct sockaddr_storage {unsigned short ss_family, array[6] of char _ss_pad1, double _ss_align, array[240] of char _ss_pad2} 111784 111785 "access.c", line 838: assignment type mismatch: 111786 struct sockaddr {unsigned short sa_family, array[14] of char sa_data} "=" struct sockaddr_storage {unsigned short ss_family, array[6] of char _ss_pad1, double _ss_align, array[240] of char _ss_pad2} 111787 111788 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 111789 Reviewed-by: Jamey Sharp <jamey@minilop.net> 111790 111791commit c6613cfc26e76b296e620640bf33e756eb4ae178 111792Author: Jamey Sharp <jamey@minilop.net> 111793Date: Fri May 14 10:36:13 2010 -0700 111794 111795 test/xi2: Initialize predefined atoms before XInputExtensionInit. 111796 111797 XInputExtensionInit calls MakeAtom, which doesn't work without the atoms 111798 table initialized. 111799 111800 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111801 Cc: Peter Hutterer <peter.hutterer@who-t.net> 111802 Tested-by: Keith Packard <keithp@keithp.com> 111803 Signed-off-by: Keith Packard <keithp@keithp.com> 111804 111805commit 1461b31e40a3e44f6327199d23cf2ab70d8f5176 111806Author: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 111807Date: Fri May 14 16:25:43 2010 +0300 111808 111809 dri2: Fix xserver crash if dri2 buffer allocation fails. 111810 111811 If driver fails to allocate memory for dri2 buffer server would crash 111812 in send_buffers_reply. 111813 111814 Solution is to handle the allocation failure and return BadAlloc to 111815 client. 111816 111817 Signed-off-by: Pauli Nieminen <ext-pauli.nieminen@nokia.com> 111818 Reviewed-by: Michel Dänzer <michel@daenzer.net> 111819 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 111820 Signed-off-by: Keith Packard <keithp@keithp.com> 111821 111822commit f87002cb7d8729b1da4cbefe7653f4cfd3a2d956 111823Merge: a4fbc0fed 890f536f8 111824Author: Keith Packard <keithp@keithp.com> 111825Date: Fri May 14 11:46:29 2010 -0700 111826 111827 Merge remote branch 'jamey/for-keith' 111828 111829commit 890f536f8d66f08e7d6a5685659eef4001a8757d 111830Author: Jamey Sharp <jamey@minilop.net> 111831Date: Fri May 14 00:18:14 2010 -0700 111832 111833 Bugfix for "Pre-validate ChangeGC XIDs": off-by-one in loop index. 111834 111835 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111836 Tested-by: Colin Harrison <colin.harrison@virgin.net> 111837 111838commit a4fbc0fedafda6df98f647173a2deccb86ef0479 111839Author: Oliver McFadden <oliver.mcfadden@nokia.com> 111840Date: Fri May 14 07:49:31 2010 +0300 111841 111842 scrnintstr.h: removed unused PaintWindow function pointers. 111843 111844 Please note this patch breaks the ABI. 111845 111846 Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com> 111847 Reviewed-by: Keith Packard <keithp@keithp.com> 111848 111849commit 71296885b0bb946e7ab986169f5b872952c221a0 111850Author: Oliver McFadden <oliver.mcfadden@nokia.com> 111851Date: Thu May 13 07:12:43 2010 +0300 111852 111853 mi: removed unused PointerNonInterestBox functions. 111854 111855 Please note this patch breaks the ABI. 111856 111857 Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com> 111858 Reviewed-by: Keith Packard <keithp@keithp.com> 111859 111860commit 795432d4a92ed5b9fa4d9163e73c8e4fe4c74534 111861Merge: f2a0c324e 92ed75ac5 111862Author: Keith Packard <keithp@keithp.com> 111863Date: Thu May 13 18:22:49 2010 -0700 111864 111865 Merge remote branch 'jamey/cleanups' 111866 111867commit 92ed75ac59e2d3af149cddb962efd05fc8487750 111868Author: Jamey Sharp <jamey@minilop.net> 111869Date: Mon May 10 20:22:05 2010 -0700 111870 111871 Eliminate boilerplate around client->noClientException. 111872 111873 Just let Dispatch() check for a noClientException, rather than making 111874 every single dispatch procedure take care of it. 111875 111876 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111877 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 111878 111879commit 11c69880c7c48ef9e755c4e09fadef7a629d7bc7 111880Author: Jamey Sharp <jamey@minilop.net> 111881Date: Sat May 8 22:16:32 2010 -0700 111882 111883 Quit using clientErrorValue in dix/colormap.c. 111884 111885 And that's it! No more clientErrorValue kludge. 111886 111887 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111888 Reviewed-by: Keith Packard <keithp@keithp.com> 111889 111890commit a3d948ddbb54b9e831e67f22d5031922a3c44107 111891Author: Jamey Sharp <jamey@minilop.net> 111892Date: Thu May 6 12:35:52 2010 -0700 111893 111894 clientErrorValue is never used outside dix. Stop importing it. 111895 111896 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111897 Reviewed-by: Keith Packard <keithp@keithp.com> 111898 111899commit 653e4878c4cc03613172a93ad4800e1eacc98f17 111900Author: Jamey Sharp <jamey@minilop.net> 111901Date: Sat May 8 19:18:11 2010 -0700 111902 111903 Quit using clientErrorValue in dix/gc.c. 111904 111905 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111906 Reviewed-by: Keith Packard <keithp@keithp.com> 111907 111908commit 5193f25ea33eed31d6a75cdc1a86427c23d8033c 111909Author: Jamey Sharp <jamey@minilop.net> 111910Date: Sat May 8 19:08:47 2010 -0700 111911 111912 Define GCAllBits as the union of all valid CreateGC masks. 111913 111914 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111915 Reviewed-by: Keith Packard <keithp@keithp.com> 111916 111917commit 6a84cd943430cfc9df55c83aef6a7f8dea6dbb94 111918Author: Jamey Sharp <jamey@minilop.net> 111919Date: Fri May 7 19:38:05 2010 -0700 111920 111921 Replace dixChangeGC with calls directly to the right variant. 111922 111923 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111924 Reviewed-by: Keith Packard <keithp@keithp.com> 111925 111926commit 2d7eb4a19b773d0406c0c2e018a7da97f3565fd5 111927Author: Jamey Sharp <jamey@minilop.net> 111928Date: Fri May 7 18:11:36 2010 -0700 111929 111930 Pre-validate ChangeGC XIDs. 111931 111932 In order to execute a wire-level ChangeGC request, we need to look up 111933 the resources named by any XIDs in the value-list. Various places in the 111934 server already have pointers to the resources they want to set into the 111935 GC, though, so over time the interface has evolved to accept either XIDs 111936 or pointers, with several different function call signatures used in 111937 different eras. 111938 111939 This patch makes the existing code require pointers to resources rather 111940 than XIDs, and adds a simple wrapper that looks up any XIDs. The old 111941 dixChangeGC API is preserved by delegating to whichever implementation 111942 is appropriate. 111943 111944 This affects error-handling: If any of the XIDs are invalid, then the GC 111945 is unchanged, and its ChangeGC callback is not invoked. This change is 111946 allowed by the protocol spec, which says, "The order in which components 111947 are verified and altered is server-dependent. If an error is generated, 111948 a subset of the components may have been altered." 111949 111950 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111951 Reviewed-by: Keith Packard <keithp@keithp.com> 111952 111953commit e2929db7b737413cf93fbebdf4d15abdfebff05c 111954Author: Jamey Sharp <jamey@minilop.net> 111955Date: Sat May 8 14:10:51 2010 -0700 111956 111957 dixChangeGC callers: Use ChangeGCVal instead of XID almost everywhere. 111958 111959 The exceptions are ProcChangeGC and CreateGC. 111960 111961 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111962 Reviewed-by: Keith Packard <keithp@keithp.com> 111963 111964commit 95728ca09d45afc84c8d1828c09c6b6725f1a58d 111965Author: Jamey Sharp <jamey@minilop.net> 111966Date: Sat May 8 09:25:34 2010 -0700 111967 111968 Don't statically allocate the ChangeGC parameter array. 111969 111970 Because that's insane. 111971 111972 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111973 Reviewed-by: Keith Packard <keithp@keithp.com> 111974 111975commit bff8525f8483304d5f93e83e36b47209381da721 111976Author: Jamey Sharp <jamey@minilop.net> 111977Date: Sat May 8 12:06:50 2010 -0700 111978 111979 Simplify miFillPolyHelper and miLineArc. 111980 111981 Both functions compute a set of spans and either fill them immediately 111982 or accumulate them into a caller-provided buffer. 111983 111984 Computing the spans used only the miTranslate and lineWidth fields of 111985 pGC, and neither could have been changed by the initial 111986 ChangeGC/ValidateGC pair, so it's safe to compute the spans first. 111987 111988 Then both functions consume the spans the same way, so factor that into 111989 a new fillSpans function. 111990 111991 Signed-off-by: Jamey Sharp <jamey@minilop.net> 111992 Reviewed-by: Keith Packard <keithp@keithp.com> 111993 111994commit 83f7ec97279405958aed86c6a57704a460c3bfba 111995Author: Jamey Sharp <jamey@minilop.net> 111996Date: Sat May 8 13:38:00 2010 -0700 111997 111998 miwideline: Factor out span buffer allocation. 111999 112000 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112001 Reviewed-by: Keith Packard <keithp@keithp.com> 112002 Reviewed-by: Adam Jackson <ajax@redhat.com> 112003 112004commit f2a0c324e37c9fa3eb9087adbf963addb7f21d88 112005Author: Keith Packard <keithp@keithp.com> 112006Date: Thu May 13 10:33:39 2010 -0700 112007 112008 Remove devPrivates init and delete callback lists. 112009 112010 XSELinux was the only consumer of these interfaces and it no longer 112011 needs them. 112012 112013 Signed-off-by: Keith Packard <keithp@keithp.com> 112014 Reviewed-by: Jamey Sharp <jamey@minilop.net> 112015 112016commit c9e7ca4404803fe44d4684e0bb2ca2ee10fd4bb3 112017Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 112018Date: Wed May 12 20:25:30 2010 -0400 112019 112020 xselinux: Remove use of devPrivates init/free callbacks. 112021 112022 Commit eb9210097efea81065c301e5b6b4da7a566deb4a removed the 112023 sidget/sidput calls which were the major reason for using the 112024 callbacks. The remaining operations can be skipped or worked 112025 around. 112026 112027 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 112028 Reviewed-by: Keith Packard <keithp@keithp.com> 112029 Signed-off-by: Keith Packard <keithp@keithp.com> 112030 112031commit bca85e2e127a8a23e3a2debcfeb3ae07cd3c66ac 112032Author: Jeremy Huddleston <jeremyhu@apple.com> 112033Date: Thu May 13 08:32:00 2010 -0700 112034 112035 Use _X_ATTRIBUTE_PRINTF _X_DEPRECATED _X_NORETURN 112036 112037 Use the values from xproto rather than duplicating the effort 112038 112039 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 112040 Reviewed-by: Keith Packard <keithp@keithp.com> 112041 Signed-off-by: Keith Packard <keithp@keithp.com> 112042 112043commit 421606a8ef447d10c2ee0986f20e752056a47675 112044Author: Kristian Høgsberg <krh@bitplanet.net> 112045Date: Tue May 11 10:52:18 2010 -0400 112046 112047 dri2: Send out event when auxillary buffers are invalidated 112048 112049 This lets the DRI2 clients rely on the server to notify them when they 112050 need to get new buffers. Without this, OpenGL clients poll the server 112051 in glViewport() which can be a performance problems and also isn't 112052 completely correct behaviour. 112053 112054 We bump the DRI2 protocol minor to indicate the availability of the 112055 event, which the DRI2 clients can use to avoid polling. This speeds up 112056 various piglit and oglc test cases as well as real applications. 112057 112058 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 112059 Reviewed-by: Adam Jackson <ajax@redhat.com> 112060 Signed-off-by: Keith Packard <keithp@keithp.com> 112061 112062commit f281db9a5e400c822e03a19937247baa20ecc213 112063Author: Alan Coopersmith <alan.coopersmith@oracle.com> 112064Date: Wed May 12 23:57:58 2010 -0700 112065 112066 Remove ResNoAvoid definition, missed in the RAC removal 112067 112068 ResNoAvoid is #defined to ResBios, but ResBios was removed 112069 from xf86str.h in 4b42448a2388d40f257774fbffdccaea87bd0347 112070 112071 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 112072 Reviewed-by: Julien Cristau <jcristau@debian.org> 112073 Signed-off-by: Keith Packard <keithp@keithp.com> 112074 112075commit 315041762313598aad90df84226e2d2def4a0fc9 112076Author: Jamey Sharp <jamey@minilop.net> 112077Date: Sat May 8 18:55:55 2010 -0700 112078 112079 CreateGC: Pass current client on to ChangeGC. 112080 112081 Otherwise we can't check that the XIDs this GC is being initialized with 112082 are accessible to this client. 112083 112084 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112085 Cc: Eamon Walsh <ewalsh@tycho.nsa.gov> 112086 Reviewed-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 112087 112088commit 04bad1b8a118a83064316d6e4327234aee19361e 112089Author: Jamey Sharp <jamey@minilop.net> 112090Date: Thu May 6 21:56:10 2010 -0700 112091 112092 Kill ChangeGC in favor of dixChangeGC. 112093 112094 This doesn't change any behavior, but it isn't clear whether NullClient 112095 is correct in all cases. As ajax says, 112096 112097 > For most of these changes, I think it's correct to use NullClient, 112098 > since they are server-initiated changes and should not fail for (eg) 112099 > xace reasons. ... At any rate, you're certainly not changing any 112100 > semantics by leaving them all as NullClient, so this patch can't be 112101 > more wrong than before. 112102 112103 The call in CreateGC is particularly questionable. 112104 112105 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112106 Reviewed-by: Adam Jackson <ajax@redhat.com> 112107 112108commit 65cedf39054cf3a1e695e84ac228cce9e8d48097 112109Author: Jamey Sharp <jamey@minilop.net> 112110Date: Thu May 6 21:39:43 2010 -0700 112111 112112 Kill DoChangeGC in favor of dixChangeGC. 112113 112114 This doesn't change any behavior, but it isn't clear whether NullClient 112115 is correct in all cases. As ajax says, 112116 112117 > For most of these changes, I think it's correct to use NullClient, 112118 > since they are server-initiated changes and should not fail for (eg) 112119 > xace reasons. ... At any rate, you're certainly not changing any 112120 > semantics by leaving them all as NullClient, so this patch can't be 112121 > more wrong than before. 112122 112123 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112124 Reviewed-by: Keith Packard <keithp@keithp.com> 112125 112126commit afcbbd6dfea51c5b0adca0d720edc02ba0c2dc16 112127Author: Jamey Sharp <jamey@minilop.net> 112128Date: Sat May 8 15:19:03 2010 -0700 112129 112130 doPolyText needs UseAccess to each font, not ReadAccess. 112131 112132 In commit 42d6112ec21949a336ee8b34469f2695273ee2d6, Eamon changed 112133 dixChangeGC to require DixUseAccess on any GCFont XID. I think 112134 doPolyText needs to require the same level of access. Otherwise 112135 dixChangeGC could fail when it does the same lookup, which doPolyText 112136 doesn't check for. 112137 112138 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112139 Cc: Eamon Walsh <ewalsh@tycho.nsa.gov> 112140 Reviewed-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 112141 112142commit 59857ee5da5f1f3f4900292581b9586477513211 112143Merge: 21ceae900 432cbbec1 112144Author: Keith Packard <keithp@keithp.com> 112145Date: Wed May 12 16:48:08 2010 -0700 112146 112147 Merge remote branch 'dottedmag/for-keithp' 112148 112149commit 432cbbec194e47bf2a117c9302146e786c8a4ee1 112150Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112151Date: Thu May 13 03:51:00 2010 +0700 112152 112153 Misc coding style cleanup 112154 112155 Use a[b] instead of *(a+b), fix whitespace. 112156 112157 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112158 Reviewed-by: Keith Packard <keithp@keithp.com> 112159 112160commit 816b79dd061e9839cec94a4986a7820b70ca8a7f 112161Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112162Date: Thu May 13 03:45:21 2010 +0700 112163 112164 Remove useless casts 112165 112166 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112167 Reviewed-by: Keith Packard <keithp@keithp.com> 112168 112169commit 63a647abd51f44226cbd16aa04ebc57d07463c6d 112170Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112171Date: Thu May 13 03:44:12 2010 +0700 112172 112173 Fix code style: extra whitespace before () 112174 112175 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112176 Reviewed-by: Keith Packard <keithp@keithp.com> 112177 112178commit 28211c443c693a1ca3db5740d0128274a3eef723 112179Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112180Date: Thu May 13 03:43:04 2010 +0700 112181 112182 Fix warning: it's safe to pass atom strings > XA_LAST_PREDEFINED to free(3) 112183 112184 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112185 Reviewed-by: Keith Packard <keithp@keithp.com> 112186 112187commit 8b5326aa98eba201dd78aea3dd7114e1a084489b 112188Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112189Date: Wed May 12 20:27:02 2010 +0000 112190 112191 Mark OsAbort as noreturn function to make gcc happier. 112192 112193 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112194 Reviewed-by: Matt Turner <mattst88@gmail.com> 112195 112196commit 868e372a73b377705217e0379bc6e00f36c4d8e5 112197Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112198Date: Thu May 13 01:59:06 2010 +0700 112199 112200 Introduce X_NORETURN macro defined as __attribute__((noreturn)) for gcc 112201 112202 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112203 Reviewed-by: Matt Turner <mattst88@gmail.com> 112204 112205commit 5a8e2f2745ae1f74501cd3f42614a1ed2cf974f2 112206Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112207Date: Wed May 12 18:54:51 2010 +0000 112208 112209 Do not jump through the hoops to deallocate xkbbasedirflag variable 112210 112211 Fixes gcc warning as well. 112212 112213 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112214 Reviewed-by: Jamey Sharp <jamey@minilop.net> 112215 Reviewed-by: Matt Turner <mattst88@gmail.com> 112216 112217commit ff2b4cf8329b1678adafcda02e5d47a072550d47 112218Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112219Date: Thu May 13 01:51:37 2010 +0700 112220 112221 Turn sprintf argument into literaral string, shutting up gcc warning 112222 112223 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112224 Reviewed-by: Matt Turner <mattst88@gmail.com> 112225 112226commit f62ba192c285b1e49bf299f03fc0b763680afaaf 112227Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112228Date: Thu May 13 01:47:26 2010 +0700 112229 112230 Do not use deprecated Xalloc function 112231 112232 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112233 Reviewed-by: Matt Turner <mattst88@gmail.com> 112234 112235commit 21ceae9002c6364deb3d074cf2da7d3864cf6879 112236Author: Jamey Sharp <jamey@minilop.net> 112237Date: Tue May 11 10:24:00 2010 -0700 112238 112239 SetFontPath: set client->errorValue on failure. 112240 112241 Previously the callers were only setting errorValue on Success, when 112242 it's ignored, and leaving it alone on failure, when it's sent to the 112243 client. 112244 112245 Since SetFontPath takes the ClientPtr, let it set client->errorValue 112246 instead of letting the callers continue to get it wrong. 112247 112248 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112249 Reviewed-by: Julien Cristau <jcristau@debian.org> 112250 Signed-off-by: Keith Packard <keithp@keithp.com> 112251 112252commit 801162919d1c625d950a5d105ae4b3487ebc30a7 112253Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112254Date: Thu May 6 01:55:36 2010 +0700 112255 112256 Removed outdated comments obsoleted by documentation in headers 112257 112258 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112259 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 112260 112261commit 4d55c8c4ffcb5d6c6500b462a41ace8e2359151d 112262Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112263Date: Thu May 6 01:54:13 2010 +0700 112264 112265 Use lowercase variant of XNFalloc and Xstrdup 112266 112267 Using one variant of function/macro makes it easier to fix the code 112268 later. 112269 112270 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112271 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 112272 112273commit 3f3ff971ecff9936cebafc813af9193b97bba89c 112274Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112275Date: Thu May 6 01:44:06 2010 +0700 112276 112277 Replace X-allocation functions with their C89 counterparts 112278 112279 The only remaining X-functions used in server are XNF*, the rest is converted to 112280 plain alloc/calloc/realloc/free/strdup. 112281 112282 X* functions are still exported from server and x* macros are still defined in 112283 header file, so both ABI and API are not affected by this change. 112284 112285 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112286 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 112287 112288commit 96c7ab27c383ec767f62a7a11e5fd76f86363fbc 112289Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112290Date: Thu May 6 00:28:25 2010 +0700 112291 112292 Deprecate allocation functions which are plain wrappers for C stdlib 112293 112294 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112295 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 112296 112297commit a5dba0f5caefce236ebd5f71948e5a659bea58e4 112298Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112299Date: Thu May 6 00:27:47 2010 +0700 112300 112301 Document allocation functions, noting deviations from C library 112302 112303 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112304 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 112305 112306commit e983848ab44b0769f97f6207f1aa8b4f127be6a9 112307Author: Mikhail Gusarov <dottedmag@dottedmag.net> 112308Date: Thu May 6 00:16:24 2010 +0700 112309 112310 Clean {X,XNF}{alloc,calloc,realloc,free,strdup} from pre-C89 baggage 112311 112312 C89 guarantees alignment of pointers returned from malloc/calloc/realloc, so 112313 stop fiddling with alignment manually and just pass the arguments to library 112314 functions. 112315 112316 Also convert silent error when negative size is passed into function into 112317 warning in log file. 112318 112319 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 112320 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 112321 112322commit 4f0006c2203abe0b7660c5068d6afe236f9bd2a4 112323Author: Kristian Høgsberg <krh@bitplanet.net> 112324Date: Mon May 10 08:36:02 2010 -0400 112325 112326 dri2: Copy front to fake front in swapbuffer 112327 112328 If we don't have a fake front, nothing will happen. The fix was extracted 112329 from a bigger patch from Francisco Jerez. 112330 112331 https://bugs.freedesktop.org/show_bug.cgi?id=27305 112332 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 112333 Reviewed-by: Pauli Nieminen <suokkos@gmail.com> 112334 Signed-off-by: Keith Packard <keithp@keithp.com> 112335 112336commit ba1f1f9d9e48226f1ef6b3222a8d92fc969d6560 112337Author: Jeremy Huddleston <jeremyhu@apple.com> 112338Date: Tue May 11 09:49:34 2010 -0700 112339 112340 XQuartz GL: Fix __glXAquaScreenCreateDrawable prototype 112341 112342 Fixes regression introduced in 9de0e31746d5f0d9d39d11c94ec3cbc04a9935fc 112343 112344 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 112345 Reviewed-by: Keith Packard <keithp@keithp.com> 112346 Signed-off-by: Keith Packard <keithp@keithp.com> 112347 112348commit 286935c6f4965a6b09de1fb9e579a00afa14082a 112349Author: Keith Packard <keithp@keithp.com> 112350Date: Tue May 11 09:08:40 2010 -0700 112351 112352 Remember drawable ID used to find damage target for use in events. Bug 5730. 112353 112354 Using Composite, window pixmaps are given names in the client resource 112355 namespace and yet may not have any XID recorded in the drawable 112356 structure. As such, we need to remember the XID used to lookup the 112357 pixmap in the resource database so that we can report the correct XID 112358 back to the client in damage events. 112359 112360 Signed-off-by: Keith Packard <keithp@keithp.com> 112361 Reviewed-by: Adam Jackson <ajax@redhat.com> 112362 112363commit 3df22b293c1f1d27e7ce16662744c578d35b5aba 112364Merge: fa6c70125 9de0e3174 112365Author: Keith Packard <keithp@keithp.com> 112366Date: Tue May 11 08:38:44 2010 -0700 112367 112368 Merge remote branch 'krh/dri2-resource-fixes' 112369 112370commit fa6c7012572093a82c9389682977efff85590719 112371Author: Cyril Brulebois <kibi@debian.org> 112372Date: Mon Mar 1 02:11:36 2010 +0100 112373 112374 Fix null pointer dereference in xf86_reload_cursors(). 112375 112376 Upon resume, X may try to dereference a null pointer, which has been 112377 reported in Debian bug #507916 (http://bugs.debian.org/507916). 112378 112379 Jim Paris came up with a patch which solves the problem for him. Here's 112380 a (hopefully) fixed version of his patch (without the typo). 112381 112382 Cc: Jim Paris <jim@jtan.com> 112383 Signed-off-by: Cyril Brulebois <kibi@debian.org> 112384 Reviewed-By: Matthias Hopf <mhopf@suse.de> 112385 Signed-off-by: Keith Packard <keithp@keithp.com> 112386 112387commit 9de0e31746d5f0d9d39d11c94ec3cbc04a9935fc 112388Author: Kristian Høgsberg <krh@bitplanet.net> 112389Date: Thu Apr 29 16:36:10 2010 -0400 112390 112391 dri2: Take an XID for tracking the DRI2 drawable 112392 112393 Some pixmaps (window pixmaps and scratch pixmaps) don't have the 112394 drawable->id set and thus DRI2 gets confused when using that field 112395 for looking up the DRI2 drawable. Go back to using privates for getting 112396 at the DRI2 drawable from a DrawablePtr. We need to keep the resource 112397 tracking in place so we can remove the DRI2 drawable when the X resource 112398 it was created for goes away. Additionally, we also now track the DRI2 112399 drawable using a client XID so we can reclaim the DRI2 drawable even if 112400 the client goes before the drawable and doesn't destroy the DRI2 drawable. 112401 112402 Tested-by: Owen W. Taylor <otaylor@fishsoup.net> 112403 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 112404 112405commit 32381363cd8f43aeb741bad70bcf96a287dac0c9 112406Author: Kristian Høgsberg <krh@bitplanet.net> 112407Date: Sat May 1 13:15:00 2010 -0400 112408 112409 list.h: Add list_for_each_entry_safe() 112410 112411 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 112412 Reviewed-by: Adam Jackson <ajax@redhat.com> 112413 112414commit 4a8a615d01b9ed18c272414bd11dc2fc661727e5 112415Author: Kristian Høgsberg <krh@bitplanet.net> 112416Date: Sat May 1 13:13:54 2010 -0400 112417 112418 glxdri2: Hard-code the extension version we need 112419 112420 If we use the #define'd version from dri_interface.h, the server will 112421 require at least that version of the extension. If we're compiling against 112422 a dri_interface.h with a newer version we don't really require, glxdri2 112423 will require a too high version of the extension. 112424 112425 The right approach is to just hard-code the version we need instead of 112426 using the #defines. 112427 112428 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 112429 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 112430 Reviewed-by: Adam Jackson <ajax@redhat.com> 112431 112432commit 6d7ba5e0fcb5d1bce6bb213dec009f3a0f802d26 112433Author: Kristian Høgsberg <krh@bitplanet.net> 112434Date: Sat May 1 13:07:46 2010 -0400 112435 112436 dix: Update element count in FreeResource*() 112437 112438 FreeResource() keeps clientTable[cid].elements up to date with the 112439 number of resources allocated to the client. The other free 112440 resource functions (FreeResourceByType(), 112441 FreeClientNeverRetainResources() and FreeClientResources()) don't 112442 maintain this invariant. 112443 112444 Typically, the only consequence is that the element count is too high 112445 and we end up allocating the hash table bigger than necessary. However, 112446 FreeResource() also relies on the element count to restart the search if 112447 the list of resources has been changed during a resource destruction 112448 callback. Since FreeResourceByType() doesn't update the count, if we call 112449 that from a resource destruction callback from FreeResource(), the 112450 loop isn't restarted and we end up following an invalid next pointer. 112451 112452 Furthermore, LookupClientResourceComplex() and 112453 FreeClientNeverRetainResources() don't use the element count to detect 112454 if a callback deleted a resource and may end up following an invalid 112455 next pointer if the resource system is called into recursively. 112456 112457 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 112458 Reviewed-by: Keith Packard <keithp@keithp.com> 112459 112460commit 10de9e8ee37265a35ceeceb2007d711da70d4f2d 112461Author: Peter Hutterer <peter.hutterer@who-t.net> 112462Date: Fri Apr 16 16:35:22 2010 +1000 112463 112464 xfree86: dga needs to use the master keyboard state (#27573) 112465 112466 GetPairedDevice() may not always return the keyboard, resulting in a 112467 null-pointer dereference when accessing the XKB state. 112468 For floating devices, the GetMaster() returns the device itself. 112469 112470 X.Org Bug 27573 <http://bugs.freedesktop.org/show_bug.cgi?id=27573> 112471 112472 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 112473 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 112474 Tested-by: Ben Hutchings <ben@decadent.org.uk> 112475 112476commit 758b8614477b53dc3de2b884fec5ccaf8a736432 112477Author: David James <davidjames@google.com> 112478Date: Mon May 10 14:00:49 2010 -0700 112479 112480 Fix wrong bracket values when startOver = FALSE. 112481 112482 Currently, SyncComputeBracketValues reuses old values of bracket_greater 112483 and bracket_less when startOver = FALSE. This can result in incorrect bracket 112484 values. To fix this issue, the startOver parameter is removed, and we do not 112485 reuse old values of bracket_greater and bracket_less. 112486 112487 X.Org Bug 27023 <http://bugs.freedesktop.org/show_bug.cgi?id=27023> 112488 112489 Signed-off-by: David James <davidjames@google.com> 112490 Reviewed-by: Adam Jackson <ajax@redhat.com> 112491 Signed-off-by: Keith Packard <keithp@keithp.com> 112492 112493commit 95074538cce93abc1739a4d9dd78b0500fb8fec9 112494Merge: c0799779d 2160ff524 112495Author: Keith Packard <keithp@keithp.com> 112496Date: Mon May 10 13:49:33 2010 -0700 112497 112498 Merge remote branch 'jamey/for-keith' 112499 112500commit 2160ff5240d7e11bdc6cec8859fd1f10395c8edd 112501Author: Jamey Sharp <jamey@minilop.net> 112502Date: Fri May 7 20:00:40 2010 -0700 112503 112504 Simplify error-handling in dixChangeGC. 112505 112506 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112507 Reviewed-by: Keith Packard <keithp@keithp.com> 112508 112509commit 4080cd42fde8b4c693eab7557581af2622479109 112510Author: Jamey Sharp <jamey@minilop.net> 112511Date: Sat May 8 09:41:46 2010 -0700 112512 112513 miPolyPoint bugfix: Check memory allocation before changing the GC. 112514 112515 miPolyPoint ought to leave the GC unchanged even if it fails. ajax says: 112516 112517 > We have a new winner for the oldest-bug competition! It's actually 112518 > been like that since X11R1: 112519 > 112520 > -rw-r--r--. 1 ajax ajax 2817 1987-09-12 01:20 ddx/mi/mipolypnt.c 112521 112522 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112523 Reviewed-by: Adam Jackson <ajax@redhat.com> 112524 112525commit ccdaecc5471cc2d56b1702eeed4280bb4d58c515 112526Author: Jamey Sharp <jamey@minilop.net> 112527Date: Sat May 8 12:26:47 2010 -0700 112528 112529 Move AppendSpanGroup from miwideline.h to miwideline.c. 112530 112531 Also make it stop being a macro. 112532 112533 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112534 Reviewed-by: Keith Packard <keithp@keithp.com> 112535 112536commit 4272d6722c0f65bdfafc95e91cc2fdb8f6bd072e 112537Author: Jamey Sharp <jamey@minilop.net> 112538Date: Sat May 8 12:18:37 2010 -0700 112539 112540 Expand macros that are used only in miFillPolyHelper. 112541 112542 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112543 Reviewed-by: Keith Packard <keithp@keithp.com> 112544 112545commit c0799779dffeef8cef228bdc57aa6c8b323e7485 112546Author: Keith Packard <keithp@keithp.com> 112547Date: Mon Apr 26 17:22:12 2010 -0700 112548 112549 dmxBERestoreRenderGlyph was passing wrong value to free 112550 112551 Signed-off-by: Keith Packard <keithp@keithp.com> 112552 Reviewed-by: Jamey Sharp <jamey@minilop.net> 112553 Reviewed-by: Matt Turner <mattst88@gmail.com> 112554 112555commit 946f664b6ab421f69df3dc3213b14b085424c9e4 112556Author: Adam Jackson <ajax@redhat.com> 112557Date: Thu May 6 17:58:29 2010 -0400 112558 112559 mi: Delete wide ellipse arc cache 112560 112561 This does make wide ellipses slower, by a factor of twoish: 112562 112563 946000.0 521000.0 ( 0.55) 10-pixel wide ellipse 112564 98300.0 49900.0 ( 0.51) 100-pixel wide ellipse 112565 17700.0 9310.0 ( 0.53) 500-pixel wide ellipse 112566 16900.0 7980.0 ( 0.47) 100-pixel wide dashed ellipse 112567 16100.0 5370.0 ( 0.33) 100-pixel wide double-dashed ellipse 112568 112569 But no one cares, and it's a modest size win: 112570 112571 text data bss dec hex filename 112572 1773824 69552 59288 1902664 1d0848 hw/vfb/Xvfb.before 112573 1773112 69552 58648 1901312 1d0300 hw/vfb/Xvfb 112574 112575 Signed-off-by: Adam Jackson <ajax@redhat.com> 112576 Reviewed-by: Keith Packard <keithp@keithp.com> 112577 Signed-off-by: Keith Packard <keithp@keithp.com> 112578 112579commit 580b99bb8af4ea093953c2b885bd4272aaf02baf 112580Author: Jamey Sharp <jamey@minilop.net> 112581Date: Thu May 6 11:00:39 2010 -0700 112582 112583 Return BadPicture, not BadPixmap, if alpha-map is invalid. 112584 112585 I believe this is what was intended. 112586 112587 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112588 Reviewed-by: Adam Jackson <ajax@redhat.com> 112589 Signed-off-by: Keith Packard <keithp@keithp.com> 112590 112591commit c677fc611b124c44adfcfb95718798105f26357b 112592Author: Jamey Sharp <jamey@minilop.net> 112593Date: Thu May 6 11:00:38 2010 -0700 112594 112595 VERIFY_PICTURE always returns BadPicture. Don't bother specifying. 112596 112597 Same goes for VERIFY_ALPHA, VERIFY_XIN_PICTURE, and VERIFY_XIN_ALPHA. 112598 112599 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112600 Reviewed-by: Julien Cristau <jcristau@debian.org> 112601 Reviewed-by: Adam Jackson <ajax@redhat.com> 112602 Signed-off-by: Keith Packard <keithp@keithp.com> 112603 112604commit 35761d5f811406bc0b6a68c1b02bdb699142745c 112605Author: Jamey Sharp <jamey@minilop.net> 112606Date: Thu May 6 11:00:37 2010 -0700 112607 112608 Introduce dixLookupFontable for "FONT or GC" parameters. 112609 112610 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112611 Reviewed-by: Julien Cristau <jcristau@debian.org> 112612 Reviewed-by: Adam Jackson <ajax@redhat.com> 112613 Signed-off-by: Keith Packard <keithp@keithp.com> 112614 112615commit 2eab697adba4b1858a530750e9a35fba79a7bf26 112616Author: Alan Coopersmith <alan.coopersmith@oracle.com> 112617Date: Fri May 7 09:32:34 2010 -0700 112618 112619 Constify function prototypes in auth & xdmcp code 112620 112621 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 112622 Tested-by: Gaetan Nadon <memsize@videotron.ca> 112623 Signed-off-by: Keith Packard <keithp@keithp.com> 112624 112625commit e2e2747f40c4b9e23dba38ea656d4c082e83a794 112626Author: Matthias Hopf <mhopf@suse.de> 112627Date: Fri May 7 18:46:30 2010 +0200 112628 112629 Fix mod() definition. 112630 112631 Presumably no implications, especially security-wise. 112632 112633 Signed-off-by: Matthias Hopf <mhopf@suse.de> 112634 Reviewed-by: Matt Turner <mattst88@gmail.com> 112635 Signed-off-by: Keith Packard <keithp@keithp.com> 112636 112637commit 626f97688a89dbc5da80390d822d21b1a737de0e 112638Author: Adam Tkac <atkac@redhat.com> 112639Date: Fri May 7 17:02:03 2010 +1000 112640 112641 Export XkbCopyDeviceKeymap from Xorg, it is needed by VNC. 112642 112643 Signed-off-by: Adam Tkac <atkac@redhat.com> 112644 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 112645 Signed-off-by: Keith Packard <keithp@keithp.com> 112646 112647commit 21ed660f30a3f96c787ab00a16499e0fb034b2ad 112648Author: Chris Humbert <freedesktop@mahadri.com> 112649Date: Fri May 7 17:02:43 2010 +1000 112650 112651 dix: make DeviceEvent coordinates signed for Xinerama. #24986 112652 112653 With Xinerama enabled, event coordinates are relative to Screen 0, so 112654 they can be negative. The new DeviceEvent's coordinates are of type 112655 uint16_t, making screens above and to the left of Screen 0 unusable. 112656 112657 X.Org Bug 24986 <https://bugs.freedesktop.org/show_bug.cgi?id=24986> 112658 112659 Signed-off-by: Chris Humbert <freedesktop@mahadri.com> 112660 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 112661 Signed-off-by: Keith Packard <keithp@keithp.com> 112662 112663commit 9802839d35aaf788790f1d0e8300db4693a70096 112664Author: Peter Hutterer <peter.hutterer@who-t.net> 112665Date: Fri May 7 16:22:12 2010 +1000 112666 112667 Xi: reset the known properties at the end of the server generation. 112668 112669 Properties allocated through XIGetKnownProperty() aren't reset on the second 112670 server generation but keep the old value. As a result, wrong Atoms are 112671 supplied to the driver, resulting in potential data corruption or weird 112672 error message. 112673 112674 Reproducible by running "xlsatom | grep FLOAT" twice on a plain X server. 112675 The second X server generation won't have the FLOAT atom defined anymore, 112676 despite the users of this atom not noticing any errors. 112677 112678 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 112679 Reviewed-by: Julien Cristau <jcristau@debian.org> 112680 Signed-off-by: Keith Packard <keithp@keithp.com> 112681 112682commit 72758287f79a4f1aa8fa388f20947042e3e14693 112683Author: Julien Cristau <jcristau@debian.org> 112684Date: Tue May 4 15:15:02 2010 +0200 112685 112686 XQuartz: add new localization files to EXTRA_DIST 112687 112688 commit 206531f75cd41c034e89fdfbc75ab0910682eef8 added localization files 112689 for ar, add them to the Makefile. 112690 112691 Signed-off-by: Julien Cristau <jcristau@debian.org> 112692 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 112693 112694commit a3f5d30ba61e70b3de8b48754dea32715ba93c40 112695Author: Keith Packard <keithp@keithp.com> 112696Date: Fri Apr 30 16:10:32 2010 -0700 112697 112698 Make sure XFixes invisible cursor gets freed on server reset 112699 112700 This uses the same hack that dix uses for the rootCursor -- allocate 112701 a resource ID for the invisible cursor so that it gets freed at reset 112702 time. This also allows us to unconditionally create it during 112703 extension initialization; necessary as the privates layout may well be 112704 different on subsequent generations. 112705 112706 Reviewed-by: Adam Jackson <ajax@redhat.com> 112707 Signed-off-by: Keith Packard <keithp@keithp.com> 112708 112709commit cdeb2c23f80ced961fe205125547e16905885534 112710Author: Keith Packard <keithp@keithp.com> 112711Date: Fri Apr 30 12:18:09 2010 -0700 112712 112713 Fix cursor ref counting mistakes with sprites and xf86Cursor.c 112714 112715 A few cursor value assignments weren't getting correctly ref counted, 112716 causing leaks of cursor objects. 112717 112718 Reviewed-by: Adam Jackson <ajax@redhat.com> 112719 Signed-off-by: Keith Packard <keithp@keithp.com> 112720 112721commit a80b1f888f2f7674a715d512ab950eeadee20448 112722Merge: 497109986 0efd7b8d8 112723Author: Keith Packard <keithp@keithp.com> 112724Date: Sun May 2 16:35:17 2010 -0700 112725 112726 Merge remote branch 'jeremyhu/master' 112727 112728commit 0efd7b8d82a48793341e27d4c224986e0971687c 112729Author: Jeremy Huddleston <jeremyhu@apple.com> 112730Date: Tue Apr 27 13:14:47 2010 -0700 112731 112732 XQuartz GLX: Change around includes for better compatability with different OS versions 112733 112734 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 112735 112736commit 7b506fdc840aebed6b5acb91437a2cb620b5bddc 112737Author: Jeremy Huddleston <jeremyhu@apple.com> 112738Date: Wed Apr 21 08:38:53 2010 -0700 112739 112740 rootless: Add some sanity checking to miPaintWindow 112741 112742 This avoids painting the root window when it isn't actually drawable. 112743 112744 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 112745 112746commit 4d3789c9b3fbe6aad32fcacd964353b612640f27 112747Author: Jeremy Huddleston <jeremyhu@apple.com> 112748Date: Sun Apr 18 01:22:27 2010 -0700 112749 112750 rootless: Remove an unneeded comment 112751 112752 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 112753 112754commit 048d15192af39ba5111882d18a957b6ff4e34cad 112755Author: Oliver McFadden <oliver.mcfadden@nokia.com> 112756Date: Thu Apr 22 10:29:44 2010 +0300 112757 112758 x-list.c: null-returning function malloc() was called without checking 112759 112760 Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com> 112761 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 112762 112763commit 4971099860e8e6d7817ada92e2636cb435fd002d 112764Author: Tiago Vignatti <tiago.vignatti@nokia.com> 112765Date: Sun May 2 19:14:39 2010 +0300 112766 112767 dix: wrap variables with #ifdef when panoramix is not used 112768 112769 this shut up some warnings. 112770 112771 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 112772 Reviewed-by: Jamey Sharp <jamey@minilop.net> 112773 Signed-off-by: Keith Packard <keithp@keithp.com> 112774 112775commit 40858960c02a7694b0e1b72e2a986805580ff6f1 112776Author: Keith Packard <keithp@keithp.com> 112777Date: Thu Apr 29 12:24:51 2010 -0700 112778 112779 Delete loop with no effect from GlyphUninit 112780 112781 Signed-off-by: Keith Packard <keithp@keithp.com> 112782 Reviewed-by: Jamey Sharp <jamey@minilop.net> 112783 112784commit 65e961fcc1da0f1e6c1c477701b2eeb36955ed24 112785Author: Keith Packard <keithp@keithp.com> 112786Date: Thu Apr 29 12:24:49 2010 -0700 112787 112788 Replace some input devPrivates with regular struct fields 112789 112790 In the process, fixes a memory leak in CloseDevice, and an unchecked 112791 memory allocation in InitializePredictableAccelerationProperties. 112792 112793 Signed-off-by: Keith Packard <keithp@keithp.com> 112794 Reviewed-by: Jamey Sharp <jamey@minilop.net> 112795 112796commit 998e982b775073da59412dce8ce609ea2ad7967c 112797Author: Keith Packard <keithp@keithp.com> 112798Date: Thu Apr 29 12:24:48 2010 -0700 112799 112800 Move callback definitions to their own file 112801 112802 Needed as the privates stuff uses the callbacks before they would have 112803 been defined in dix.h 112804 112805 Signed-off-by: Keith Packard <keithp@keithp.com> 112806 Reviewed-by: Jamey Sharp <jamey@minilop.net> 112807 112808commit 7dde50c649089b03650d6497079a690feea65e22 112809Author: Keith Packard <keithp@keithp.com> 112810Date: Thu Apr 29 12:24:47 2010 -0700 112811 112812 Move X_DEPRECATED to misc.h so it can be used outside resource.h 112813 112814 Signed-off-by: Keith Packard <keithp@keithp.com> 112815 Reviewed-by: Jamey Sharp <jamey@minilop.net> 112816 112817commit 5623908aeef70e5083f3b49986c7547ed044fedd 112818Author: Aaron Plattner <aplattner@nvidia.com> 112819Date: Wed Apr 28 12:37:08 2010 -0700 112820 112821 dix: Improve documentation of the DIX private data functions. 112822 112823 The functions exported by the devPrivates code were poorly documented. I tried 112824 to spruce it up a little. 112825 112826 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 112827 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 112828 Signed-off-by: Keith Packard <keithp@keithp.com> 112829 112830commit 986d46144b183a36b4e98aed95eca0c55a8b4251 112831Merge: a974c8e7c b5b8f91b8 112832Author: Keith Packard <keithp@keithp.com> 112833Date: Fri Apr 30 12:40:53 2010 -0700 112834 112835 Merge remote branch 'jamey/for-keith' 112836 112837 Conflicts: 112838 hw/xfree86/common/xf86xv.c 112839 112840commit a974c8e7cba40c8d7d1b91e07de8c414627b71a2 112841Merge: 6581bc459 02e86221b 112842Author: Keith Packard <keithp@keithp.com> 112843Date: Fri Apr 30 12:33:00 2010 -0700 112844 112845 Merge remote branch 'whot/for-keith' 112846 112847commit 6581bc4591746c906d97f8b868f946c47bc6d756 112848Merge: 0e91e19f7 d073e51be 112849Author: Keith Packard <keithp@keithp.com> 112850Date: Fri Apr 30 12:29:52 2010 -0700 112851 112852 Merge remote branch 'yselkowitz/master' 112853 112854commit 0e91e19f783d21198fc2e210203e0f10040a9cd0 112855Merge: 81a081c1f edbc56c08 112856Author: Keith Packard <keithp@keithp.com> 112857Date: Fri Apr 30 12:27:51 2010 -0700 112858 112859 Merge remote branch 'vignatti/for-keith' 112860 112861commit 02e86221b851e5423a95782aa7e297ea051022ca 112862Author: Peter Hutterer <peter.hutterer@who-t.net> 112863Date: Thu Apr 29 09:48:37 2010 +1000 112864 112865 xfree86: a missing input driver is not an error. 112866 112867 We call NIDR on all devices that make it through the config backend. 112868 Including some that have no driver assigned to them (/dev/input/mouse0 for 112869 example). Those ones then simply get ignored by NIDR, but this should not be 112870 noted as an error in the log file. 112871 112872 X_INFO is sufficient, and it may just prevent some bugreports. 112873 112874 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 112875 Acked-by: Julien Cristau <jcristau@debian.org> 112876 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 112877 112878commit 81a081c1f0cb55df94cb10495aa7ad71cd5a9afb 112879Author: Jon TURNEY <jon.turney@dronecode.org.uk> 112880Date: Sat Nov 7 19:28:18 2009 +0000 112881 112882 Cygwin/X: AIGLX using native WGL 112883 112884 A rewrite of the XWin DDX AIGLX code to actually make it do something useful again 112885 112886 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 112887 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 112888 112889commit 262c2c9031ea60cbe738ed80ddb7faa1b888d137 112890Author: Rami Ylimäki <ext-rami.ylimaki@nokia.com> 112891Date: Tue Apr 27 14:04:47 2010 +0300 112892 112893 test: Fix linking of wrapped functions in XI2 tests. 112894 112895 Running "make check" will lead to build problems in 112896 scratchbox. Building the first test that wraps dixLookupWindow fails 112897 because symbol __real_dixLookupWindow can't be resolved. Defining 112898 wrapping options as linker options instead of compiler options makes 112899 everything build nicely in scratchbox. 112900 112901 Signed-off-by: Rami Ylimäki <ext-rami.ylimaki@nokia.com> 112902 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 112903 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 112904 112905commit ffaae7c0c69a51a53a76146c79f3630ae197a443 112906Author: Adam Jackson <ajax@redhat.com> 112907Date: Mon Apr 26 20:11:31 2010 -0700 112908 112909 Remove mibank support 112910 112911 Banked framebuffers are so 1990. As of 7.4 the only drivers remaining 112912 that used this were chips, neomagic, trident, and vesa. vesa only used 112913 it when not using shadowfb, which is broadly undesirable anyway, and no 112914 longer uses it at all as of 2.3.0. neomagic never used it by default, 112915 and support for it is gone in git master. The other two effectively 112916 only ever used it for ISA chips; since ISA support is now gone from 112917 the server, they have been modified to only compile mibank support when 112918 ISA support is available. 112919 112920 Signed-off-by: Adam Jackson <ajax@redhat.com> 112921 Reviewed-by: Keith Packard <keithp@keithp.com> 112922 Signed-off-by: Keith Packard <keithp@keithp.com> 112923 112924commit b5b8f91b82d7b150c926dd3fecee6c3aafff6e39 112925Author: Jamey Sharp <jamey@minilop.net> 112926Date: Thu Apr 22 21:35:17 2010 -0700 112927 112928 xfree86: use screen privates for Xv offscreen images. 112929 112930 This replaces a globally-allocated array that depended on MAXSCREENS. 112931 112932 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112933 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 112934 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 112935 112936commit a1c2acfe798c57e5be7e5f6c111a6ce91400487a 112937Author: Jamey Sharp <jamey@minilop.net> 112938Date: Wed Apr 21 18:05:45 2010 -0700 112939 112940 xfree86: use screen privates for exclusive DGA clients. 112941 112942 Most DGA requests allow at most one client to be using DGA on each 112943 screen. Instead of keeping track of the current client in a 112944 MAXSCREEN-sized array, track it in a per-screen private. 112945 112946 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112947 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 112948 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 112949 112950commit f9e3a2955d2ca73604c68fc9d51405581b832edb 112951Author: Jamey Sharp <jamey@minilop.net> 112952Date: Mon Apr 26 18:23:27 2010 -0700 112953 112954 Make Xvfb independent of MAXSCREENS. 112955 112956 If a -screen option specifies a screen number higher than any previously 112957 specified, reallocate the vfb-private array of screen-info structs. 112958 112959 If built with a DIX that still has a MAXSCREENS limit, asking for too 112960 many screens won't be detected until InitOutput calls AddScreen. 112961 112962 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112963 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 112964 112965commit 20e84b0b44e8b3b40a3ecab5b2e64a27de247b16 112966Author: Jamey Sharp <jamey@minilop.net> 112967Date: Mon Apr 26 18:09:23 2010 -0700 112968 112969 Xvfb: Simplify screen option processing. 112970 112971 Inspired by Jon Turney's "Xwin: Simplify screen option processing" 112972 patch, which does something similar for the Xwin server. 112973 112974 Besides making the code more readable, this eliminates most of Xvfb's 112975 references to MAXSCREENS. 112976 112977 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112978 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 112979 112980commit eeb84547556b943af2acff207e034823205c7dfe 112981Author: Jamey Sharp <jamey@minilop.net> 112982Date: Mon Apr 26 18:04:25 2010 -0700 112983 112984 Delete redundant scrnum field from Xvfb private screen-info struct. 112985 112986 The screen number can be inferred from the position in the vfbScreens 112987 array, and it was only used in two places, so it was hardly important. 112988 112989 Signed-off-by: Jamey Sharp <jamey@minilop.net> 112990 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 112991 112992commit da0217891904bc48d5f0b7ea5c62c8ea0e9b95f9 112993Author: Jamey Sharp <jamey@minilop.net> 112994Date: Wed Apr 21 22:26:28 2010 -0700 112995 112996 Track screens' installed colormaps as screen privates. 112997 112998 Several DDXes allow each screen to have at most one (or in some cases, 112999 exactly one) installed colormap. These all use the same pattern: Declare 113000 a global-lifetime array of MAXSCREENS ColormapPtrs, and index it by 113001 screen number. This patch converts most of those to use screen privates 113002 instead. 113003 113004 Signed-off-by: Jamey Sharp <jamey@minilop.net> 113005 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113006 113007commit 22c4300ee25a20e1f815e46225bf0de9cfd6748f 113008Author: Jamey Sharp <jamey@minilop.net> 113009Date: Wed Oct 7 09:00:02 2009 -0700 113010 113011 Simplify XineramifyXv. 113012 113013 The structure of the adaptor-matching algorithm was obscured by trying 113014 to use "continue" and "break" as the only control-flow primitives. It's 113015 a lot more clear if you add "return" to that set. 113016 113017 Signed-off-by: Jamey Sharp <jamey@minilop.net> 113018 Reviewed-by: Adam Jackson <ajax@redhat.com> 113019 113020commit af170a4ab76d33254dd4e54a33bb410fa1624739 113021Author: Peter Hutterer <peter.hutterer@who-t.net> 113022Date: Wed Apr 21 15:51:20 2010 +1000 113023 113024 Xext: only update the sprite on pointer events. 113025 113026 A call to miPointerUpdateSprite for the XTEST keyboard may result in a 113027 NULL pointer dereference in miDCPutUpCursor() when the save buffer is NULL. 113028 113029 XTS test case: Xlib 11 KeymapNotify 113030 113031 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113032 113033commit bac1c5f1be4588b2b1eb646ee98a5442e1b767d5 113034Author: Dirk Wallenstein <halsmit@t-online.de> 113035Date: Sat Apr 17 21:36:23 2010 +0200 113036 113037 xkb: Fix omissions in geometry initialization #27679 113038 113039 _XkbCopyGeom did not copy all of the data from the source geometry. This 113040 resulted in failures when trying to obtain the keymap from a server 113041 where the default geometry has not been replaced by a custom 113042 configuration. 113043 113044 Signed-off-by: Dirk Wallenstein <halsmit@t-online.de> 113045 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 113046 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113047 113048commit 3f63db89191701139d45332ca80b9d6eb327c773 113049Merge: 41bdb6c00 54e51de8c 113050Author: Keith Packard <keithp@keithp.com> 113051Date: Mon Apr 26 20:18:13 2010 -0700 113052 113053 Merge remote branch 'jamey/for-keith' 113054 113055commit 41bdb6c003cca3ef0ff88d9c7de318115bab1ba2 113056Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 113057Date: Wed Apr 21 18:11:05 2010 -0700 113058 113059 xf86: Don't crash when switching modes through RandR without owning the VT. 113060 113061 While VT-switched, FB access is disabled and should remain so. Trying to switch 113062 modes in that state would re-enable it, potentially causing crashes if trying 113063 to access it before the driver has recovered from the mode switch. 113064 113065 Signed-off-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 113066 Reviewed-by: Adam Jackson <ajax@redhat.com> 113067 Signed-off-by: Keith Packard <keithp@keithp.com> 113068 113069commit 2a763c9023b8bc978b32eaa56d2c625b6f2badee 113070Author: Adam Jackson <ajax@redhat.com> 113071Date: Mon Feb 15 16:54:35 2010 -0500 113072 113073 dix: Default DPMS timeout values to match screensaver values 113074 113075 These have the same default, but if you specify something different with 113076 -s on the command line, only the screensaver time is changed. As DPMS 113077 is usually what's desired, change it to match. 113078 113079 Signed-off-by: Adam Jackson <ajax@redhat.com> 113080 Reviewed-by: Keith Packard <keithp@keithp.com> 113081 Signed-off-by: Keith Packard <keithp@keithp.com> 113082 113083commit 518f3b189b6c8aa28b62837d14309fd06163ccbb 113084Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 113085Date: Wed Apr 21 16:46:17 2010 -0700 113086 113087 mi: don't thrash resources when displaying the software cursor across screens 113088 113089 This changes the DC layer to maintain a persistent set of GCs/pixmaps/pictures 113090 for each pScreen instead of failing to thrash between them when changing 113091 screens. 113092 113093 Signed-off-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 113094 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113095 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113096 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 113097 113098commit 54e51de8cd950b3f5b30e72cf1d604f54f619f0e 113099Author: Jamey Sharp <jamey@minilop.net> 113100Date: Sat Apr 24 23:07:47 2010 -0700 113101 113102 ProcGrabButton: remove redundant error check. 113103 113104 If dixLookupResourceByType did not return Success, it will have set the 113105 pointer to NULL, so the second if will always be true. 113106 113107 Signed-off-by: Jamey Sharp <jamey@minilop.net> 113108 Reviewed-by: Alan Coopersmith <alan.coopersmith@oracle.com> 113109 113110commit 7e573b5ccd2c759f74ebb84d3a5a23da6b88ec6d 113111Author: Jamey Sharp <jamey@minilop.net> 113112Date: Mon Apr 26 15:05:06 2010 -0700 113113 113114 Kill unused/useless macros in panoramiX.h: BREAK_IF, IF_RETURN, FORCE_ROOT 113115 113116 This header doesn't seem to be used outside the server tree, so I don't 113117 think anything was depending on these except what I've fixed in this 113118 patch. 113119 113120 FORCE_ROOT is not used anywhere, and never has been in all the history 113121 that's in git. ajax removed its FORCE_WIN and FORCE_CMAP companions 113122 three years ago, in 04b87d6dfae02e4ecdb5216d12c6cdafd1e8c2b4. 113123 113124 BREAK_IF and IF_RETURN expanded to exactly the keywords in their names, 113125 which is silly. They were rarely used and served only to make callers 113126 inconsistent with the idioms used elsewhere in the server. 113127 113128 Signed-off-by: Jamey Sharp <jamey@minilop.net> 113129 Reviewed-by: Matt Turner <mattst88@gmail.com> 113130 113131commit e372cc58e02391c3e8e82ac03f73fbe8b1f7d49c 113132Author: Jamey Sharp <jamey@minilop.net> 113133Date: Wed Oct 7 11:06:34 2009 -0700 113134 113135 Boolean-or short-circuits, so (a || (!a && ...)) is redundant. Simplify. 113136 113137 Signed-off-by: Jamey Sharp <jamey@minilop.net> 113138 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113139 Reviewed-by: Matt Turner <mattst88@gmail.com> 113140 113141commit 25d1c6cd37cc4079adf7c4bad8d10678fb08b409 113142Author: Keith Packard <keithp@keithp.com> 113143Date: Mon Apr 26 11:04:38 2010 -0700 113144 113145 Exa screen private has SavedSourceValidate, not SourceValidate 113146 113147 Need to check the right field to see if there is a wrapped 113148 SourceValidate function. 113149 113150 Signed-off-by: Keith Packard <keithp@keithp.com> 113151 113152commit d073e51beeed2e702fa5c7534b1345360bc302f2 113153Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113154Date: Mon Apr 26 11:39:19 2010 -0500 113155 113156 Clarify help strings for disabled features in configure 113157 113158 Features which are disabled by default use the --enable-* syntax to show 113159 how they need to be passed to override the default. 113160 113161 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113162 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 113163 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113164 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 113165 113166commit 57409af1267f48457f93134922f2450518182e09 113167Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113168Date: Wed Apr 14 13:18:28 2010 -0500 113169 113170 Catch when requested SHA1 implementation is missing 113171 113172 The other SHA1 implementation detections already error out if 113173 specifically requested but were not found. 113174 113175 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113176 Reviewed-by: Julien Cristau <jcristau@debian.org> 113177 113178commit 0d8fe2fe5c37659147cbecac793183c62fc6818a 113179Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113180Date: Tue Apr 13 13:32:51 2010 -0500 113181 113182 Prefer libsha1 over libgcrypt 113183 113184 When no SHA1 implementation is specified, we should first prefer 113185 system-builtin solutions (libc/libmd/CommonCrypto), then smaller 113186 implementations over the larger ones. libsha1 is much smaller than 113187 libgcrypt, so it should be first. 113188 113189 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113190 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 113191 Acked-by: Julien Cristau <jcristau@debian.org> 113192 113193commit 65a2c09549d2372c23b155addff428177708c910 113194Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113195Date: Thu Apr 8 21:55:45 2010 -0500 113196 113197 Revert "Disable Record by default." 113198 113199 Record was broken during the pre-1.6 development cycle and was not fixed 113200 until 1.7.6. Now that it is fixed, re-enable it by default. 113201 113202 This reverts commit 3eaecdd66e791e0f3d86b23ce10be057ca44c044. 113203 113204 Conflicts: 113205 113206 configure.ac 113207 113208 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113209 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113210 113211commit 6e0d847c273ab89323883a05f806f449180e3669 113212Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113213Date: Thu Mar 25 03:37:09 2010 -0500 113214 113215 xfree86: Fix reference to SGML entities 113216 113217 XORG_CHECK_LINUXDOC sets the SGML search path to the parent of X11/defs.ent. 113218 113219 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113220 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 113221 Tested-by: Gaetan Nadon <memsize@videotron.ca> 113222 113223commit 04b7d529c09af2c842cf020d5b26080e1ffd88d6 113224Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113225Date: Thu Mar 25 03:36:25 2010 -0500 113226 113227 xfree86: Ignore linuxdoc generated docs 113228 113229 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113230 Acked-by: Gaetan Nadon <memsize@videotron.ca> 113231 113232commit 39e5108c78e0802833a8e9ea7f29d042738625ba 113233Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113234Date: Thu Mar 25 03:35:36 2010 -0500 113235 113236 dmx: use silent rules in doc 113237 113238 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113239 Acked-by: Gaetan Nadon <memsize@videotron.ca> 113240 113241commit 3ba2ce5d1054c99354686aeb373247f1b9debf79 113242Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113243Date: Thu Mar 25 03:33:55 2010 -0500 113244 113245 dmx: Ignore linuxdoc generated docs 113246 113247 dmx.txt and scaled.txt are generated from SGML, so they probably never 113248 should have been in version control in the first place. 113249 113250 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113251 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 113252 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 113253 113254commit 40972576b606237d5d42abc13d846163e264a4ac 113255Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113256Date: Tue Mar 23 04:47:27 2010 -0500 113257 113258 Use AC_PROG_SED and respect its result 113259 113260 AC_PROG_SED sets SED as the path to a fully-functional 'sed' (which may 113261 also be called 'gsed' if GNU sed is installed alongside a proprietary 113262 version). 113263 113264 This is a follow up to commit 9be4157391edf0c5fc4ee36adfb1eb1c3bdb8e3b. 113265 113266 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 113267 Reviewed-by: Rémi Cardona <remi@gentoo.org> 113268 113269commit b557571dbff3d9a615e03cad904d46b8af9153a0 113270Author: Jerome Glisse <jglisse@redhat.com> 113271Date: Mon Apr 26 16:01:48 2010 +0200 113272 113273 exa: avoid calling non existent SourceValidate callback 113274 113275 In unaccel path we were unconditionaly calling the SourceValidate 113276 callback but in some case it could be NULL. Check if we have a 113277 valid callback before calling it. 113278 113279 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 113280 Acked-by: Michel Dänzer <michel@daenzer.net> 113281 Signed-off-by: Keith Packard <keithp@keithp.com> 113282 113283commit 0e6e75b04eaca311962390533acc03d0ac5ece83 113284Author: Keith Packard <keithp@keithp.com> 113285Date: Mon Apr 26 09:14:22 2010 -0700 113286 113287 Revert "exa: avoid calling non existent SourceValidate callback" 113288 113289 This reverts commit daa6f9308fa57e41210f78d92dbdb3f1c8eae6ad. 113290 113291 This one checked pScreen->SourceValidate instead of 113292 pExaScr->SourceValidate which would cause a segfault when 113293 pExaScr->SourceValidate was NULL. 113294 113295commit daa6f9308fa57e41210f78d92dbdb3f1c8eae6ad 113296Author: Jerome Glisse <jglisse@redhat.com> 113297Date: Mon Apr 26 15:55:26 2010 +0200 113298 113299 exa: avoid calling non existent SourceValidate callback 113300 113301 In unaccel path we were unconditionaly calling the SourceValidate 113302 callback but in some case it could be NULL. Check if we have a 113303 valid callback before calling it. 113304 113305 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 113306 Acked-by: Michel Dänzer <michel@daenzer.net> 113307 Signed-off-by: Keith Packard <keithp@keithp.com> 113308 113309commit edbc56c088462844f6445f5e6d7cbcfc58a207dc 113310Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113311Date: Mon Apr 26 14:49:25 2010 +0300 113312 113313 include: remove couple of unused structures fields and bump ABI 113314 113315 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113316 113317commit 4f8dc1109dcdfa973466a038c16c07da5f6c16b4 113318Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113319Date: Thu Apr 22 15:43:11 2010 +0300 113320 113321 dix and others: remove unused arraySize field from ScreenInfo 113322 113323 Bizarre. This seems to never be used before. I left the field in ScreenInfo, 113324 with another name. So, stop looking at it. 113325 113326 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113327 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 113328 113329commit 64fd39f2f0ffbbdc2d5929047d260263d0962141 113330Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113331Date: Thu Apr 22 15:48:48 2010 +0300 113332 113333 xfree86: no need to assign numScreens again 113334 113335 numScreens is always being assigned to 0 in dix for any server generation. 113336 113337 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113338 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113339 113340commit a7d398e545a4be5491248d5ccb303aa03ee1594f 113341Author: Jon TURNEY <jon.turney@dronecode.org.uk> 113342Date: Fri Apr 16 18:13:50 2010 +0100 113343 113344 Xwin: make screens structures run-time adjustable 113345 113346 Change g_ScreenInfo, an array of winScreenInfo elements, from a 113347 static array of MAXSCREENS elements, to a dynamically allocated one 113348 113349 Fix up the validation that -screen option screen numbers are 113350 contiguous from zero (which possibly didn't work correctly before 113351 anyhow) 113352 113353 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 113354 Reviewed-by: Jamey Sharp<jamey@minilop.net> 113355 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113356 113357commit d8454ae488cfc073cd6010c9a08d53855a0c2612 113358Author: Jon TURNEY <jon.turney@dronecode.org.uk> 113359Date: Mon Apr 12 20:18:13 2010 +0100 113360 113361 Xwin: Simplify screen option processing 113362 113363 Use an instance of the XWin DDX-specific screen info structure to hold 113364 the current default values, to simplify greatly the code for applying 113365 options to all screens and remove all those loops over MAXSCREENS screens 113366 in the command line option processing 113367 113368 Use g_iNumScreens for tracking the current initialized screen count 113369 113370 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 113371 Reviewed-by: Jamey Sharp <jamey@minilop.net> 113372 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113373 113374commit b61870595ba4df06006d24ed8c07cfe781bce1b7 113375Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113376Date: Thu Apr 22 20:20:48 2010 +0300 113377 113378 xfree86: track screens' installed colormaps as screen privates 113379 113380 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113381 Reviewed-by: Jamey Sharp <jamey@minilop.net> 113382 113383commit a94f5455c71363d5047668ee093901b9e1645cf1 113384Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113385Date: Thu Apr 22 20:20:43 2010 +0300 113386 113387 mi: track screens' installed colormaps as screen privates 113388 113389 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113390 Reviewed-by: Jamey Sharp <jamey@minilop.net> 113391 113392commit a2c716eaf6b3a4ce75382394636a0a890b5dcfe0 113393Author: Jamey Sharp <jamey@minilop.net> 113394Date: Thu Apr 22 07:34:04 2010 +0200 113395 113396 fb: track screens' installed colormaps as screen privates. 113397 113398 Several DDXes allow each screen to have at most one (or in some cases, 113399 exactly one) installed colormap. These all use the same pattern: Declare 113400 a global-lifetime array of MAXSCREENS ColormapPtrs, and index it by 113401 screen number. This patch converts most of those to use screen privates 113402 instead. 113403 113404 Signed-off-by: Jamey Sharp <jamey@minilop.net> 113405 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113406 113407commit e055bef055b6c726e9f3ef91a83585d13c80651d 113408Author: Jamey Sharp <jamey@minilop.net> 113409Date: Thu Apr 22 06:08:29 2010 +0200 113410 113411 xfree86: use screen privates for exclusive DGA clients 113412 113413 Most DGA requests allow at most one client to be using DGA on each screen. 113414 Instead of keeping track of the current client in a MAXSCREEN-sized array, 113415 track it in a per-screen private. 113416 113417 Signed-off-by: Jamey Sharp <jamey@minilop.net> 113418 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113419 113420commit 7c9733d063acee04735a0e75d62e508fe4ea1163 113421Author: Jamey Sharp <jamey@minilop.net> 113422Date: Fri Apr 23 06:41:00 2010 +0200 113423 113424 xfree86: use screen privates for Xv offscreen images. 113425 113426 This replaces a globally-allocated array that depended on MAXSCREENS. 113427 113428 Signed-off-by: Jamey Sharp <jamey@minilop.net> 113429 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113430 113431commit 35d208125fa0ec78e8d694adc52886b977e789cf 113432Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113433Date: Wed Apr 21 18:28:13 2010 +0300 113434 113435 xfree86: fix not reached code in fi1236 driver from i2c 113436 113437 This issue was introduced in the first dump of the code in 2004. I haven't 113438 check what's the correct fix for it so I simply kept the behaviour of someone 113439 calling this and removed the unreachable code. 113440 113441 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113442 Reviewed-by: Keith Packard <keithp@keithp.com> 113443 Signed-off-by: Keith Packard <keithp@keithp.com> 113444 113445commit 7ac6a6b7d0dfc5e021270b2898accd3425aa008a 113446Merge: 0782894b5 f7ef6fd9a 113447Author: Keith Packard <keithp@keithp.com> 113448Date: Wed Apr 21 16:07:38 2010 -0700 113449 113450 Merge remote branch 'vignatti/for-keith' 113451 113452 Conflicts: 113453 exa/exa.c 113454 113455 Signed-off-by: Keith Packard <keithp@keithp.com> 113456 113457commit 0782894b5702adcf6f4a90861793b717f3856fa5 113458Author: Keith Packard <keithp@keithp.com> 113459Date: Wed Apr 21 16:00:28 2010 -0700 113460 113461 Xvfb: Usage message typo fix. 113462 113463 Signed-off-by: Keith Packard <keithp@keithp.com> 113464 113465commit ee3412b8702072c3a0b006bd20dd3bc7071d721c 113466Author: Keith Packard <keithp@keithp.com> 113467Date: Wed Apr 21 15:58:23 2010 -0700 113468 113469 Restore some unused win32 debug code accidentally removed with unifdef 113470 113471 unifdef found some code covered by 113472 113473 which it decided to remove. This patch simply restores that in case 113474 someone wants it back. 113475 113476 Signed-off-by: Keith Packard <keithp@keithp.com> 113477 113478commit e3d2a7a613366c26e5316cf582d9a8c6c6692b0a 113479Author: Peter Hutterer <peter.hutterer@who-t.net> 113480Date: Wed Apr 21 11:47:24 2010 +1000 113481 113482 Revert "mi: don't thrash resources when displaying the software cursor across screens" 113483 113484 This commit leads to a segfault on the very first XTS test case. 113485 113486 Backtrace: 113487 0: /opt/xorg/bin/Xorg (xorg_backtrace+0x3b) [0x80a33db] 113488 1: /opt/xorg/bin/Xorg (0x8048000+0x62a75) [0x80aaa75] 113489 2: (vdso) (__kernel_rt_sigreturn+0x0) [0x5d140c] 113490 3: /lib/libc.so.6 (0x9bb000+0x73579) [0xa2e579] 113491 4: /lib/libc.so.6 (realloc+0xe0) [0xa2e830] 113492 5: /opt/xorg/bin/Xorg (Xrealloc+0x33) [0x80a3f33] 113493 6: /opt/xorg/bin/Xorg (0x8048000+0x1ab79) [0x8062b79] 113494 7: /opt/xorg/bin/Xorg (0x8048000+0x1ac4e) [0x8062c4e] 113495 8: /opt/xorg/bin/Xorg (RegisterExtensionNames+0x2ce) [0x8062fbe] 113496 9: /opt/xorg/bin/Xorg (AddExtension+0x19a) [0x807bd7a] 113497 10: /opt/xorg//lib/xorg/modules/extensions/libextmod.so (0x728000+0x1169a) 113498 [0x73969a] 113499 11: /opt/xorg/bin/Xorg (InitExtensions+0x85) [0x80c0eb5] 113500 12: /opt/xorg/bin/Xorg (0x8048000+0x1a51d) [0x806251d] 113501 13: /lib/libc.so.6 (__libc_start_main+0xe6) [0x9d1bb6] 113502 14: /opt/xorg/bin/Xorg (0x8048000+0x1a2a1) [0x80622a1] 113503 Segmentation fault at address 0x10b2d5f8 113504 113505 valgrind output: 113506 ==5069== Invalid read of size 4 113507 ==5069== at 0x80F928D: FreePicture (picture.c:1531) 113508 ==5069== by 0x818DDEF: miDCDeviceCleanup (midispcur.c:867) 113509 ==5069== by 0x81B97F0: miSpriteDeviceCursorCleanup (misprite.c:968) 113510 ==5069== by 0x80995FA: miPointerDeviceCleanup (mipointer.c:292) 113511 ==5069== by 0x807973E: CloseDevice (devices.c:840) 113512 ==5069== by 0x80799B6: CloseDownDevices (devices.c:933) 113513 ==5069== by 0x8062705: main (main.c:309) 113514 ==5069== Address 0x4cce844 is 12 bytes inside a block of size 84 free'd 113515 ==5069== at 0x40057F6: free (vg_replace_malloc.c:325) 113516 ==5069== by 0x80A3DE0: Xfree (utils.c:1154) 113517 ==5069== by 0x80F9332: FreePicture (picture.c:1576) 113518 ==5069== by 0x80FBB4B: PictureDestroyWindow (picture.c:69) 113519 ==5069== by 0x810B1A3: damageDestroyWindow (damage.c:1840) 113520 ==5069== by 0x80864F1: FreeWindowResources (window.c:846) 113521 ==5069== by 0x8086812: DeleteWindow (window.c:925) 113522 ==5069== by 0x806B53E: FreeClientResources (resource.c:806) 113523 ==5069== by 0x806B60F: FreeAllResources (resource.c:823) 113524 ==5069== by 0x80626E4: main (main.c:299) 113525 ==5069== 113526 ==5069== Invalid write of size 4 113527 ==5069== at 0x80F9295: FreePicture (picture.c:1531) 113528 ==5069== by 0x818DDEF: miDCDeviceCleanup (midispcur.c:867) 113529 ==5069== by 0x81B97F0: miSpriteDeviceCursorCleanup (misprite.c:968) 113530 ==5069== by 0x80995FA: miPointerDeviceCleanup (mipointer.c:292) 113531 ==5069== by 0x807973E: CloseDevice (devices.c:840) 113532 ==5069== by 0x80799B6: CloseDownDevices (devices.c:933) 113533 ==5069== by 0x8062705: main (main.c:309) 113534 ==5069== Address 0x4cce844 is 12 bytes inside a block of size 84 free'd 113535 ==5069== at 0x40057F6: free (vg_replace_malloc.c:325) 113536 ==5069== by 0x80A3DE0: Xfree (utils.c:1154) 113537 ==5069== by 0x80F9332: FreePicture (picture.c:1576) 113538 ==5069== by 0x80FBB4B: PictureDestroyWindow (picture.c:69) 113539 ==5069== by 0x810B1A3: damageDestroyWindow (damage.c:1840) 113540 ==5069== by 0x80864F1: FreeWindowResources (window.c:846) 113541 ==5069== by 0x8086812: DeleteWindow (window.c:925) 113542 ==5069== by 0x806B53E: FreeClientResources (resource.c:806) 113543 ==5069== by 0x806B60F: FreeAllResources (resource.c:823) 113544 ==5069== by 0x80626E4: main (main.c:299) 113545 113546 XTS test case: Xproto pAllocColor 113547 113548 This reverts commit 00b8b7ad61b6f818271fb4d1e383113170309d72. 113549 113550 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113551 Signed-off-by: Keith Packard <keithp@keithp.com> 113552 113553commit f7ef6fd9a1a1f48603df2796fe7c904c25a81077 113554Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113555Date: Tue Apr 20 15:56:58 2010 +0300 113556 113557 xfree86: fix not reached code in tty code 113558 113559 CLEARRTS_SUPPORT cannot be triggered at all. Notice that mouse driver manual 113560 page states the support for it though. 113561 113562 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113563 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113564 113565commit b36eeb713a349961a6be8a6dd183ded6d27f9974 113566Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113567Date: Tue Apr 20 16:26:02 2010 +0300 113568 113569 xkb: check for NULL pointer before dereferences it in XkbWriteXKBSymbols 113570 113571 move srv assignment to before it's being used. Also, check for xkb being nil. 113572 113573 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113574 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113575 113576commit 471f350c99b2e7c5a9045eef5ebff65cd2ddb7c5 113577Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113578Date: Mon Apr 19 20:34:30 2010 +0300 113579 113580 exa: don't need to check for NULL pointer if we already assumed it has a value 113581 113582 the alternative would be to check ps in the beginning of the function. 113583 113584 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113585 Acked-by: Michel Dänzer <michel@daenzer.net> 113586 113587commit 96784f4fcb6c2fb82f6d1abbd28ea1e189e4e6f5 113588Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113589Date: Mon Apr 19 20:46:12 2010 +0300 113590 113591 xkb: check for NULL pointer before dereferences it in XkbAddClientResource 113592 113593 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113594 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113595 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113596 113597commit 057c147541bde6f0cbe22ca069b43a97ddc95baf 113598Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113599Date: Fri Apr 16 18:01:41 2010 +0300 113600 113601 Xi: check for NULL pointer before dereferences it in ListButtonInfo 113602 113603 Both dev and dev->button are already used before their checking were being 113604 performed. So check on the beginning. 113605 113606 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113607 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113608 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113609 113610commit 7f457351d2a09013cd57fcb2f95c0d6f56bfaccb 113611Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113612Date: Mon Apr 19 20:18:51 2010 +0300 113613 113614 xfree86: check for NULL pointer before dereferences it in parser code 113615 113616 Seems to be harmless. Meh. 113617 113618 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113619 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113620 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113621 113622commit d948dcd9555c81ccb378054383ef8da464202bdf 113623Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113624Date: Fri Apr 16 17:08:09 2010 +0300 113625 113626 Xi: fix not reached code in XSendExtensionEvent 113627 113628 Error was introduced in 31a7994a. I.e., broken since 2007. I guess nobody uses 113629 XSendExtensionEvent. 113630 113631 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113632 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113633 113634commit f491b0aa5b86a5242fc4e04218202bbb52d0af39 113635Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113636Date: Fri Apr 16 16:47:33 2010 +0300 113637 113638 xfree86: fix not reached code in parser 113639 113640 ...because Error is a macro that returns NULL. 113641 113642 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113643 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113644 113645commit 099946a3ac94ff5e575e9edcecd26cf9f346b241 113646Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113647Date: Fri Apr 16 18:35:55 2010 +0300 113648 113649 mi: check for NULL pointer before dereferences it in miPointerSetPosition 113650 113651 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113652 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113653 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113654 113655commit 49835eec0c996ad95a01f0fe340336b6b60e51aa 113656Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113657Date: Fri Apr 16 17:50:43 2010 +0300 113658 113659 exa: check for NULL pointer before dereferences it 113660 113661 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113662 Acked-by: Michel Dänzer <michel@daenzer.net> 113663 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113664 113665commit 996c115deb558e8e9490ea773528b8e96650df53 113666Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113667Date: Thu Mar 25 18:17:54 2010 +0200 113668 113669 configure: remove unused builtin font macro from autoconf file 113670 113671 This should be removed together with 49b93df8. 113672 113673 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113674 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113675 113676commit 0ba82562eeba8bf3bcd00b6e3ff28ce5b2c8df3c 113677Author: Tiago Vignatti <tiago.vignatti@nokia.com> 113678Date: Tue Mar 30 19:10:36 2010 +0300 113679 113680 Death to Multibuffer extension 113681 113682 The rationale behind is because no sane application will use this when we have 113683 modern APIs such DRI2. Besides, as a fact, xfree86 server has already 113684 deprecated this extension in 1998: 113685 113686 http://www.xfree86.org/3.3.6/isc7.html 113687 113688 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 113689 Reviewed-by: Keith Packard <keithp@keithp.com> 113690 113691commit 28b7b2b8d02d975480080865f0dddebcaa2f7968 113692Author: Keith Packard <keithp@keithp.com> 113693Date: Mon Apr 19 09:26:10 2010 -0700 113694 113695 unifdef -B -DRENDER to always include RENDER code 113696 113697 This patch was created with: 113698 113699 git ls-files '*.[ch]' | while read f; do unifdef -B -DRENDER -o $f $f; done 113700 113701 Signed-off-by: Keith Packard <keithp@keithp.com> 113702 113703commit b3ab978df861c08298f57529e3db980489055c35 113704Merge: a92b2c2c8 0ad022a72 113705Author: Keith Packard <keithp@keithp.com> 113706Date: Sun Apr 18 22:01:40 2010 -0700 113707 113708 Merge remote branch 'whot/for-keith' 113709 113710commit 0ad022a729bafa56cc7b5d241f567444a34514ad 113711Author: Peter Hutterer <peter.hutterer@who-t.net> 113712Date: Wed Apr 14 10:54:29 2010 +1000 113713 113714 xkb: rename XkbFakeDeviceButton and XkbFakeDeviceMotion, move into xkbActions.c 113715 113716 The name XkbDDXFakeDeviceButton and XkbDDXFakeDeviceMotion is somewhat 113717 misleading, there's no DDX involved in the game at all anymore. 113718 113719 This removes XkbFakeDeviceMotion and XkbFakeDeviceButton from the API where 113720 it arguably shouldn't have been in the first place. 113721 113722 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113723 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 113724 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113725 113726commit da4e2e382828d7ba460766709368ec6214b286dd 113727Author: Peter Hutterer <peter.hutterer@who-t.net> 113728Date: Wed Apr 14 09:48:53 2010 +1000 113729 113730 xkb: purge unneeded includes from ddxDevBtn.c 113731 113732 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113733 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 113734 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 113735 113736commit a92b2c2c8dd1e86ee852168146f01bdf72bfe2d0 113737Author: Kristian Høgsberg <krh@bitplanet.net> 113738Date: Fri Apr 16 05:55:35 2010 -0400 113739 113740 glx: Drop DestroyWindow hook 113741 113742 Now that glx doesn't call DRI2DestroyDrawable anymore, we don't need to 113743 force a specific resource destruction order in the DestroyWindow hook. 113744 113745 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 113746 Reviewed-by: Michel Dänzer <michel@daenzer.net> 113747 113748 https://bugs.freedesktop.org/show_bug.cgi?id=26394 113749 Signed-off-by: Keith Packard <keithp@keithp.com> 113750 113751commit 1da1f33f2dd5b437dd56cd9f5d6782de4ad5a1bc 113752Author: Kristian Høgsberg <krh@bitplanet.net> 113753Date: Fri Apr 16 05:55:34 2010 -0400 113754 113755 DRI2: Track DRI2 drawables as resources, not privates 113756 113757 The main motivation here is to have the resource system clean up the 113758 DRI2 drawable automatically so glx doesn't have to. Right now, the 113759 glx drawable resource must be destroyed before the X drawable, so that 113760 calling DRI2DestroyDrawable doesn't crash. By making the DRI2 113761 drawable a resource, GLX doesn't have to worry about that and the 113762 resource destruction order becomes irrelevant. 113763 113764 https://bugs.freedesktop.org/show_bug.cgi?id=26394 113765 113766 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 113767 Signed-off-by: Keith Packard <keithp@keithp.com> 113768 113769commit 22da7aa9d743deee198aaf6df5d370a446db9763 113770Author: Kristian Høgsberg <krh@bitplanet.net> 113771Date: Fri Apr 16 05:55:33 2010 -0400 113772 113773 glx: Let the resource system destroy pixmaps 113774 113775 GLX pbuffers are implemented using a pixmap allocated by the server. 113776 With the change to DRI2 to track DRI2 drawables as resources, we need to make 113777 sure that every drawable we create a DRI2 drawable for has an XID. By 113778 using the XID of the pbuffer, the resource system will automatically 113779 reclaim the hidden pixmap and the DRI2 drawable when the pbuffer is 113780 destroyed or the client exits. 113781 113782 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 113783 Signed-off-by: Keith Packard <keithp@keithp.com> 113784 113785commit f0006aa58f6cf7552a239e169ff6e7e4fda532f4 113786Author: Kristian Høgsberg <krh@bitplanet.net> 113787Date: Fri Apr 16 05:55:32 2010 -0400 113788 113789 glx: Track GLX 1.3 style GLX drawables under their X drawable ID as well 113790 113791 This ensures that the DrawableGone callback gets called as necessary 113792 when the X drawable goes away. Otherwise, using a GLX drawable 113793 (say, glXSwapBuffers) in indirect mode after the X drawable has been 113794 destroyed will crash the server. 113795 113796 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 113797 Reviewed-by: Michel Dänzer <michel@daenzer.net> 113798 Signed-off-by: Keith Packard <keithp@keithp.com> 113799 113800commit f4106c02318fcc4b534224df5b95a58aff555fb4 113801Author: Peter Hutterer <peter.hutterer@who-t.net> 113802Date: Tue Apr 13 14:44:59 2010 +1000 113803 113804 xkb: use GPE for XKB fake motion events. 113805 113806 Section 4.6.1 of the XKB spec says that "the initial event always moves the 113807 cursor the distance specified in the action [...]", so skip the 113808 POINTER_ACCELERATE flag for GPE, it would cause double-acceleration. 113809 113810 Potential regression - GPE expects the coordinates to be either relative or 113811 both. XKB in theory allows for x to be relative and y to be absolute (or 113812 vice versa). Let's pretend that scenario has no users. 113813 113814 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113815 Reviewed-by: Simon Thum <simon.thum@gmx.de> 113816 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 113817 113818commit 6c42c8c356be305dc7f3f92ad8d58675da8c2f07 113819Author: Peter Hutterer <peter.hutterer@who-t.net> 113820Date: Wed Apr 14 10:51:41 2010 +1000 113821 113822 xkb: Guard against SIGIO updates during PointerKeys. 113823 113824 In theory, an event coming in during GPE could reset our lastSlave, leading 113825 to rather interesting events lateron. 113826 113827 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113828 Reviewed-by: Simon Thum <simon.thum@gmx.de> 113829 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 113830 113831commit 108457dff816569453a2d88cd72595fa7eb02479 113832Author: Peter Hutterer <peter.hutterer@who-t.net> 113833Date: Tue Apr 13 14:41:07 2010 +1000 113834 113835 xkb: Post PointerKeys through the XTEST device. 113836 113837 Posting an event through a master device may cause pointer jumps once 113838 lastSlave == master, caused by double scaling. To avoid this, post the fake 113839 event generated by XKB through the XTEST device instead. 113840 113841 Fedora bug #560356 <https://bugzilla.redhat.com/560356> 113842 Tested-by: Andrew McNabb 113843 113844 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113845 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 113846 113847commit a780e5b3638a0ff81301fc68aca15b47ba0befb7 113848Author: Benjamin Tissoires <tissoire@cena.fr> 113849Date: Wed Apr 14 17:27:51 2010 +0200 113850 113851 xf86ScaleAxis: support for high resolution devices 113852 113853 High resolution devices was generating integer overflow. 113854 For instance the wacom Cintiq 21UX has an axis value up to 113855 87000. Thus the term (dSx * (Cx - Rxlow)) is greater than 113856 MAX_INT32. 113857 113858 Using 64bits integer avoids such problem. 113859 113860 Signed-off-by: Philippe Ribet <ribet@cena.fr> 113861 Signed-off-by: Benjamin Tissoires <tissoire@cena.fr> 113862 Reviewed-by: Keith Packard <keithp@keithp.com> 113863 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113864 113865commit e424d5812300e82de375d83dc0b490a76d865016 113866Merge: 5b0faf355 0e7703f9b 113867Author: Keith Packard <keithp@keithp.com> 113868Date: Thu Apr 15 15:01:34 2010 -0700 113869 113870 Merge remote branch 'whot/for-keith' 113871 113872commit 0e7703f9b1927328954a2fc87aac6be244819329 113873Author: Julien Cristau <jcristau@debian.org> 113874Date: Tue Apr 13 22:28:36 2010 +0200 113875 113876 vfb: add a name and type to the pointer and keyboard 113877 113878 Fixes a crash in XIQueryDevice which calls strlen on a NULL pointer. 113879 113880 #0 strlen () at ../sysdeps/x86_64/strlen.S:31 113881 #1 0x00000000004c16ed in SizeDeviceInfo (dev=0x969bd0) 113882 at ../../Xi/xiquerydevice.c:204 113883 #2 0x00000000004c1a01 in ProcXIQueryDevice (client=0xa57510) 113884 at ../../Xi/xiquerydevice.c:98 113885 113886 Debian bug#575905 <http://bugs.debian.org/575905> 113887 113888 Reported-by: "Bernhard R. Link" <brlink@debian.org> 113889 Signed-off-by: Julien Cristau <jcristau@debian.org> 113890 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113891 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113892 113893commit 2f29b163bbdebe384c24f781bb97b446959e8f4c 113894Author: Adam Tkac <atkac@redhat.com> 113895Date: Mon Mar 22 14:52:29 2010 +0100 113896 113897 dix: Export AllocDevicePair GetPointerEvents, GetKeyboardEvents and generate_modkeymap functions from Xorg. 113898 113899 Those functions are used by TigerVNC libvnc.so module which doesn't 113900 use standard XInput infrastructure but uses same functions like, 113901 for example, XTest devices. 113902 113903 Signed-off-by: Adam Tkac <atkac@redhat.com> 113904 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113905 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113906 113907commit 241b53b77750b5eea6759e79b23be4ff270a3d1f 113908Author: Peter Hutterer <peter.hutterer@who-t.net> 113909Date: Tue Mar 23 14:35:56 2010 +1000 113910 113911 mi: remove log-spamming bogus error message (#26843) 113912 113913 miSpriteRealizeCursor is called whenever a device is set floating and it's 113914 fine to do so, no need for an error message. 113915 Same goes for the other miSprite messages. 113916 113917 X.Org Bug 26843 <http://bugs.freedesktop.org/show_bug.cgi?id=26843> 113918 113919 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113920 113921commit 00b8b7ad61b6f818271fb4d1e383113170309d72 113922Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 113923Date: Wed Apr 7 13:52:47 2010 -0700 113924 113925 mi: don't thrash resources when displaying the software cursor across screens 113926 113927 This changes the DC layer to maintain a persistent set of GCs/pixmaps/pictures 113928 for each pScreen instead of failing to thrash between them when changing 113929 screens. 113930 113931 Signed-off-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 113932 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113933 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113934 113935commit 5f31e2196179f8db3170d65a17d8ad40da1acb0d 113936Author: Tim Yamin <plasm@roo.me.uk> 113937Date: Mon Mar 8 12:45:15 2010 +1000 113938 113939 dix: fix cursor screen check for xinerama setups. 113940 113941 The de-duplication of CheckPhysLimits 942eae6868b8b0f343b6a added a 113942 condition that is invalid for a Xinerama setup. pScreen is invalid for the 113943 Xinerama case, so comparing it to anything is a bad idea. 113944 113945 Signed-off-by: Tim Yamin <plasm@roo.me.uk> 113946 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 113947 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113948 113949commit 5b0faf355465c9f21ad96e0ed266fbdbc29efb5b 113950Author: Jeremy Huddleston <jeremyhu@apple.com> 113951Date: Thu Apr 15 10:55:53 2010 -0700 113952 113953 XQuartz GLX: Don't let garbage enter our pixel request 113954 113955 https://bugs.freedesktop.org/show_bug.cgi?id=27654 113956 113957 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 113958 113959commit fe7778e58e099d353689755ed2f5aa440569ebe3 113960Author: Jeremy Huddleston <jeremyhu@apple.com> 113961Date: Wed Apr 14 09:06:04 2010 -0700 113962 113963 XQuartz: Fix possible NULL dereference in ListenOnOpenFD 113964 113965 <rdar://problem/7862319> 113966 113967 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 113968 Reviewed-by: Marc Majka <majka@apple.com> 113969 113970commit 206531f75cd41c034e89fdfbc75ab0910682eef8 113971Author: Jeremy Huddleston <jeremyhu@apple.com> 113972Date: Tue Apr 13 14:33:49 2010 -0700 113973 113974 XQuartz: Localization update 113975 113976 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 113977 113978commit 9ddbb03fa56aa73c3f417d8ee6433e45b94445b3 113979Author: Peter Hutterer <peter.hutterer@who-t.net> 113980Date: Wed Apr 14 17:43:22 2010 +1000 113981 113982 dix: Fix crash in DeliverGrabbedEvents. 113983 113984 If both devices are synchronously grabbed, first with a GrabPointer, then 113985 with a GrabKeyboard (GrabModeSync on both), sync.other of each device points 113986 to the grab of the respective other device. 113987 113988 If the keyboard is then thawed through a AllowSome request, the VCK's 113989 sync.other is reset to NULL. Subsequently, an event on the VCP would crash 113990 the server when dereferencing sync.other on the VCP. 113991 113992 The check's purpose is to compare if the other device is grabbed by the same 113993 client, which should be checked by accessing (dev->deviceGrab->grab->resource). 113994 A check of the server-1.3 sources confirms that. 113995 113996 XTS test case: Xlib13 XAllowEvents 20. 113997 113998 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 113999 Reviewed-by: Keith Packard <keithp@keithp.com> 114000 Signed-off-by: Keith Packard <keithp@keithp.com> 114001 114002commit 7b6517526631a65891b806bca30be8f49955d0a8 114003Author: Michel Dänzer <daenzer@vmware.com> 114004Date: Wed Apr 14 19:28:37 2010 +0200 114005 114006 EXA: Check sys_ptr isn't NULL before passing it to the UploadToScreen hook. 114007 114008 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=27510 . 114009 114010 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 114011 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 114012 Signed-off-by: Keith Packard <keithp@keithp.com> 114013 114014commit fc5d76740851725e3788c68e14474a012a205892 114015Author: Julien Cristau <jcristau@debian.org> 114016Date: Tue Apr 6 11:51:53 2010 +0200 114017 114018 XGE: don't register an extension event 114019 114020 The GenericEvent is a core event, we never send an extension event, so 114021 don't reserve an id for one. 114022 114023 Signed-off-by: Julien Cristau <jcristau@debian.org> 114024 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 114025 Signed-off-by: Keith Packard <keithp@keithp.com> 114026 114027commit 0b676acd47e0d6a675df374b42ddb200deab5d06 114028Author: Julien Cristau <jcristau@debian.org> 114029Date: Tue Apr 13 11:45:41 2010 +0200 114030 114031 Bump version in configure.ac 114032 114033 Signed-off-by: Keith Packard <keithp@keithp.com> 114034 114035commit 1760d2bef9f5b248cb2332f6ebf0220eb02bab42 114036Author: Kristian Høgsberg <krh@bitplanet.net> 114037Date: Fri Apr 9 10:58:48 2010 -0400 114038 114039 glx: Set the pbuffer bit for dri2 fbconfigs 114040 114041 They've been implemented for a while, but we never advertised them. All we 114042 need to do is set the GLX_PBUFFER_BIT in the drawable type fbconfig 114043 field when we're using DRI2. 114044 114045 https://bugs.freedesktop.org/show_bug.cgi?id=26581 114046 114047 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 114048 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 114049 Signed-off-by: Keith Packard <keithp@keithp.com> 114050 114051commit ab60975fe28cb135b4183c57b7f50d3615921c1f 114052Merge: adbbc6610 59edde7c2 114053Author: Keith Packard <keithp@keithp.com> 114054Date: Mon Apr 12 21:40:41 2010 -0700 114055 114056 Merge remote branch 'jeremyhu/master' 114057 114058commit 59edde7c28db2d3174a5813a0af334e4ead2082f 114059Author: Jeremy Huddleston <jeremyhu@apple.com> 114060Date: Fri Apr 9 16:19:43 2010 -0700 114061 114062 XQuartz: Add a GUI preference for the Alt / Mode_switch toggle 114063 114064 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114065 Reviewed-by: Kevin Van Vechten <kvv@apple.com> 114066 114067commit 5600f7f001529b3afdee95546aec212a70d6a5b2 114068Author: Jeremy Huddleston <jeremyhu@apple.com> 114069Date: Fri Apr 9 13:33:06 2010 -0700 114070 114071 XQuartz: Customize the NSDefaults id in the man file. 114072 114073 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114074 Reviewed-by: Kevin Van Vechten <kvv@apple.com> 114075 114076commit 840d12c7a6ac6a974da185045215ce944a61fab6 114077Author: Jeremy Huddleston <jeremyhu@apple.com> 114078Date: Fri Apr 9 13:29:34 2010 -0700 114079 114080 XQuartz: Add a defaults option to toggle Alt / Mode_switch 114081 114082 See option_sends_alt in Xquartz(1) 114083 114084 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114085 Reviewed-by: Kevin Van Vechten <kvv@apple.com> 114086 114087commit d6f160510a50d4c4eaa48c9c69a5ddda0d50052c 114088Author: Jeremy Huddleston <jeremyhu@apple.com> 114089Date: Mon Apr 5 16:54:22 2010 -0700 114090 114091 XQuartz: Blacklist some oddball legacy Mac keycodes that break wine 114092 114093 http://xquartz.macosforge.org/trac/ticket/295 114094 114095 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114096 Reviewed-by: Kevin Van Vechten <kvv@apple.com> 114097 114098commit adbbc661080ba4efdb764c154d40c4d2fe718e16 114099Author: Dan Nicholson <dbn.lists@gmail.com> 114100Date: Tue Feb 16 08:11:41 2010 -0800 114101 114102 xfree86: Fix priority ordering for ignoring input classes 114103 114104 Commit 8736d112afb0dd61dfdaadd6378eafd200b2ef5f changed the priority 114105 ordering of the InputClass option merging to be "last match wins". This 114106 fixes the handling of Option "Ignore" to follow that logic. 114107 114108 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 114109 Reviewed-by: Keith Packard <keithp@keithp.com> 114110 114111commit d7c98c1c81ae272f66edb05fde20f4c616604add 114112Merge: 82cf3a4ae 03ccbd257 114113Author: Keith Packard <keithp@keithp.com> 114114Date: Wed Apr 7 22:25:51 2010 -0700 114115 114116 Merge remote branch 'whot/for-keith' 114117 114118commit 03ccbd2579948d11e5f16f88cdf68a55f57c9d26 114119Author: Peter Hutterer <peter.hutterer@who-t.net> 114120Date: Thu Apr 8 13:47:26 2010 +1000 114121 114122 xfree86: remove dead input drivers from xorg.conf man page. 114123 114124 These drivers have been deactivated for over a year now, let's not refer 114125 potential users to them. 114126 114127 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114128 114129commit d4dd3d0085f8eaca8c989aaeace1a1a24c5ebf49 114130Author: Dan Nicholson <dbn.lists@gmail.com> 114131Date: Sat Apr 3 09:33:51 2010 -0700 114132 114133 Move 10-evdev.conf to system config dir $datadir/X11/xorg.conf.d 114134 114135 Since the server searches in a vendor specific path now, we can install 114136 the evdev catchall there without disturbing local administration files. 114137 114138 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 114139 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114140 114141commit 95f01bdfee8241371675f0089170fa6b2908d815 114142Author: Dan Nicholson <dbn.lists@gmail.com> 114143Date: Sat Apr 3 10:08:25 2010 -0700 114144 114145 xfree86: Search for a system xorg.conf.d 114146 114147 In addition to the conf files found in /etc/X11 or $sysconfdir/X11 used 114148 for local administration, we also reserve a system directory for vendor 114149 and package usage. The simple search path is: 114150 114151 /usr/share/X11/xorg.conf.d 114152 $datadir/X11/xorg.conf.d 114153 114154 Files from these directories will have the lowest config priority. The 114155 directory $datadir/X11/xorg.conf.d is exported from xorg-server.pc in 114156 the variable "sysconfigdir". Packages should install their .conf files 114157 to the directory specified by: 114158 114159 `pkg-config --variable=sysconfigdir xorg-server` 114160 114161 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 114162 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114163 114164commit a1bae63dc634a5a79861eab128c63506081f7ed9 114165Author: Dan Nicholson <dbn.lists@gmail.com> 114166Date: Sat Apr 3 09:33:49 2010 -0700 114167 114168 xfree86: Set a saner search path for xorg.conf.d 114169 114170 There's no reason to carry all the oddities from xorg.conf like appended 114171 hostname to the search path for xorg.conf.d. This changes it to something 114172 very simple: 114173 114174 /etc/X11/<cmdline> 114175 $sysconfdir/X11/<cmdline> 114176 /etc/X11/xorg.conf.d 114177 $sysconfdir/X11/xorg.conf.d 114178 114179 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 114180 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114181 114182commit 2ac33888a9a5766aaf88bc92a8bd1d489873524c 114183Author: Dan Nicholson <dbn.lists@gmail.com> 114184Date: Sat Apr 3 09:33:48 2010 -0700 114185 114186 xfree86: Document how -configdir affects the xorg.conf.d search path 114187 114188 Explain the "safe" path dance for -configdir, too. 114189 114190 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 114191 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114192 114193commit 2460e921d18fd264e6f6374be0908f4100442650 114194Author: Dan Nicholson <dbn.lists@gmail.com> 114195Date: Sat Apr 3 09:33:47 2010 -0700 114196 114197 xfree86: Allow adding sysconfdir and datadir to config search paths 114198 114199 We could just use $projectroot/etc and $projectroot/share, but the user 114200 might have other plans for them. 114201 114202 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 114203 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114204 114205commit c8a608cb6ce8f9c86258c1ab49084f691fa9cc51 114206Author: Peter Hutterer <peter.hutterer@who-t.net> 114207Date: Wed Apr 7 10:04:38 2010 +1000 114208 114209 config: only match sane devices in 10-evdev.conf 114210 114211 Having a generic catchall also adds devices like accelerometers. These 114212 devices make X unusable, hence restrict matching to "known sane" devices 114213 like pointers, touchpads, keyboards, tablets and touchscreens. 114214 114215 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114216 Acked-by: Julien Cristau <jcristau@debian.org> 114217 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 114218 Acked-by: James Cloos <cloos@jhcloos.com> 114219 114220commit 6b09f66d8c6ebcee70382b5cca1ba82b68f20afa 114221Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 114222Date: Fri Apr 2 12:48:21 2010 -0700 114223 114224 Don't keep a pointer to a possibly freed cursor when changing screens, preventing a crash in xf86CursorEnableDisableFBAccess() trying to restore it. 114225 114226 Signed-off-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 114227 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114228 114229commit 82cf3a4ae01811917f7903d6f62ba9b7132adf7e 114230Author: Alan Coopersmith <alan.coopersmith@oracle.com> 114231Date: Tue Apr 6 15:52:18 2010 -0700 114232 114233 Convert x86emu fixed size int typedefs to use stdint types 114234 114235 Fixes x86emu builds when using non-gnu compilers now that u64 is required 114236 114237 Before this fix, the u64 type would not be defined, causing 114238 x86emu/sys.c to fail to build: 114239 "sys.c", line 102: syntax error before or at: ldq_u 114240 "sys.c", line 102: syntax error before or at: * 114241 114242 Since Keith requested using <stdint.h>, converted all the x86emu 114243 typedefs to use the stdint types. 114244 114245 Signed-off-by: Alan Coopersmith <alan.coopersmith@oracle.com> 114246 Acked-by: Matt Turner <mattst88@gmail.com> 114247 Acked-by: Tiago Vignatti <tiago.vignatti@nokia.com> 114248 Signed-off-by: Keith Packard <keithp@keithp.com> 114249 114250commit a9fe7cfa77ccee64d68732dc3f37d35cbfc27a65 114251Author: Paulo Ricardo Zanoni <pzanoni@mandriva.com> 114252Date: Thu Mar 11 14:28:18 2010 -0300 114253 114254 dix: be more verbose when we run out of opcodes 114255 114256 If we run out of opcodes, nothing is print on the log, making the 114257 problem hard to debug. In the current Xserver, if you enable some 114258 extensions like multibuffer (+2 events) and use nvidia binary driver (+5 114259 events) you can run out of opcode numbers. 114260 114261 Signed-off-by: Paulo Ricardo Zanoni <pzanoni@mandriva.com> 114262 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 114263 Reviewed-by: Julien Cristau <jcristau@debian.org> 114264 Signed-off-by: Keith Packard <keithp@keithp.com> 114265 114266commit 6150595bdbacde13bb4f9aba81b15435089bee94 114267Author: Gaetan Nadon <memsize@videotron.ca> 114268Date: Mon Apr 5 19:12:22 2010 -0400 114269 114270 doc: specify 1.6.1 as the minimum version for doxygen. 114271 114272 Older versions generate filenames that are different from 114273 the ones listed in the Makefile. 114274 114275 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 114276 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 114277 Signed-off-by: Keith Packard <keithp@keithp.com> 114278 114279commit cbda58c963af5bb9c4643e9017dd4eec589ceff3 114280Merge: a7698a677 c983f40fe 114281Author: Keith Packard <keithp@keithp.com> 114282Date: Wed Apr 7 15:51:38 2010 -0700 114283 114284 Merge remote branch 'yselkowitz/master' 114285 114286commit a7698a677682a32960b885c22fdba2add70f4658 114287Merge: 495cec794 165a4a9c7 114288Author: Keith Packard <keithp@keithp.com> 114289Date: Tue Apr 6 12:36:15 2010 -0700 114290 114291 Merge remote branch 'jbarnes/master' 114292 114293commit c983f40fe4f30311f31a9df699cdc888709251d0 114294Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114295Date: Wed Feb 24 15:17:42 2010 -0600 114296 114297 Cygwin/X: Add configure option for WindowsWM 114298 114299 WindowsWM support is still experimental, and uses the Rootless extension 114300 which currently breaks the simultaneous build of the other DDXs (see 114301 commit b3415187e92960cbff784108b5a3a8d130dc34c5). So we disable it by 114302 default for now; once the latter issue is fixed we can make this 'auto'. 114303 114304 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114305 114306commit 20701522be803fe47e921fcf059dadf64c7f287d 114307Author: Jon TURNEY <jon.turney@dronecode.org.uk> 114308Date: Thu Feb 18 14:19:21 2010 -0600 114309 114310 Don't enable ROOTLESS_WORKAROUND, it breaks composite 114311 114312 This possibly brings back whatever the bug is in 114313 http://bugs.freedesktop.org/show_bug.cgi?id=1168 for 114314 -rootless mode, but since we don't have reproduction 114315 steps for that, I can't test that... 114316 114317 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 114318 Tested-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114319 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114320 114321commit 95c9947f6244d37592958b70760bd1ac885b625a 114322Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114323Date: Wed Mar 24 00:08:23 2010 -0500 114324 114325 Cygwin/X: Disable unsupported extensions in configure 114326 114327 Several extensions are not supported by XWin, some of which are enabled 114328 by default in configure. We forcefully disable these early on so that 114329 configure will succeed without arguments and without the corresponding 114330 proto installed. 114331 114332 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114333 114334commit 24c9d48cd6999f1e90f9f0e2bbfebd8e95d314d8 114335Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114336Date: Mon Mar 22 12:47:52 2010 -0500 114337 114338 Disable setuid configure test on Cygwin 114339 114340 Only Xorg is installed setuid, so there is no need to run this configure 114341 test on Cygwin. 114342 114343 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114344 114345commit 495cec794dad95ed0c79048f3c410ad23e7d5ea4 114346Author: Keith Packard <keithp@keithp.com> 114347Date: Fri Apr 2 00:14:23 2010 -0700 114348 114349 xserver 1.8.0 114350 114351 Signed-off-by: Keith Packard <keithp@keithp.com> 114352 114353commit b9ad452ec92a7dcbed680acb3f3b8ec29fa660df 114354Author: Tiago Vignatti <tiago.vignatti@nokia.com> 114355Date: Thu Apr 1 18:57:50 2010 +0300 114356 114357 xfree86: die gracefully in the vga arbiter if AddScreen fails 114358 114359 vga arbiter will be locked in one device while AbortDDX will call LeaveVT 114360 routines from the other device. Fail! 114361 114362 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 114363 Reviewed-by: Keith Packard <keithp@keithp.com> 114364 Signed-off-by: Keith Packard <keithp@keithp.com> 114365 114366commit e75f202e7768b7cc0eb2ad0e0f3f51ea81cd7275 114367Author: Peter Hutterer <peter.hutterer@who-t.net> 114368Date: Wed Mar 24 11:08:55 2010 +1000 114369 114370 configure: enable udev backend as "auto" 114371 114372 Due to the checks in configure, this means it gets priority over HAL if 114373 libudev is found. 114374 114375 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114376 Signed-off-by: Keith Packard <keithp@keithp.com> 114377 114378commit 24952b786625fef808cb26d539dd5109d623c053 114379Author: Keith Packard <keithp@keithp.com> 114380Date: Thu Apr 1 23:44:24 2010 -0700 114381 114382 Install 10-evdev.conf in $(prefix)/etc/X11/xorg.conf.d under udev 114383 114384 udev needs some xorg.conf file to tell it to load a suitable input 114385 driver, 10-evdev.conf is as simple as they come, mapping all evdev 114386 devices to the evdev driver. 114387 114388 Signed-off-by: Keith Packard <keithp@keithp.com> 114389 114390commit 02328f190f893cb4ad0f3af7a2ecd39ed77ae5ad 114391Author: Jeremy Huddleston <jeremyhu@apple.com> 114392Date: Thu Apr 1 18:21:51 2010 -0700 114393 114394 Bump bundle version to 2.5.1 114395 114396 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114397 114398commit 0e342add7768df55a1917e27f9b85feddd444e82 114399Author: Jeremy Huddleston <jeremyhu@apple.com> 114400Date: Thu Apr 1 15:30:46 2010 -0700 114401 114402 darwin: Correct inline assembly for ___crashreporter_info__ 114403 114404 It was missing an underscore. 114405 114406 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114407 114408commit 67b814d9b2baea6beccfb1625a1e3f0b2ba7218b 114409Author: Ruediger Oertel <ro@suse.de> 114410Date: Mon Mar 29 00:23:40 2010 +0200 114411 114412 Remove now obsolete function chooseVideoDriver 114413 114414 Signed-off-by: Keith Packard <keithp@keithp.com> 114415 114416commit 1dd5fbc5a4d860af1c102b33e04dedd8c72617a1 114417Author: Ruediger Oertel <ro@suse.de> 114418Date: Fri Mar 26 09:33:20 2010 -0700 114419 114420 xfree86: Handle driver autoconfiguration when .conf files exist 114421 114422 When doing driver autoconfiguration with some parts of the config file 114423 present but no driver set (e.g. only input configuration) fix the case 114424 that we may have multiple drivers to try. 114425 114426 Create a screen section for each driver and let them be tried in a row. 114427 114428 Signed-off-by: Ruediger Oertel <ro@suse.de> 114429 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 114430 Tested-by: Timo Aaltonen <timo.aaltonen@aalto.fi> 114431 Signed-off-by: Keith Packard <keithp@keithp.com> 114432 114433commit 57cb1a854a1b6d91d214b3fa250a7df2ed20d433 114434Merge: 4b3d67a76 28a5f14b4 114435Author: Keith Packard <keithp@keithp.com> 114436Date: Mon Mar 29 11:46:00 2010 -0700 114437 114438 Merge remote branch 'jeremyhu/master' 114439 114440commit 4b3d67a7647f696957727948f9757a261e15d14e 114441Merge: 579715f83 aa9150835 114442Author: Keith Packard <keithp@keithp.com> 114443Date: Mon Mar 29 11:44:07 2010 -0700 114444 114445 Merge remote branch 'whot/for-keith' 114446 114447commit 165a4a9c7de0fcc6ef6a6421736b412ccb35965e 114448Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114449Date: Tue Mar 23 09:47:08 2010 -0700 114450 114451 GLX/DRI2: expose swap control extensions if DDX support is present 114452 114453 Export DDX swap control status from the DRI2 module and check for it in 114454 GLX when initializing extensions. 114455 114456 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114457 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114458 114459commit 5933b0abc6a76aaea84aa534df89900cd795c888 114460Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114461Date: Mon Mar 8 15:10:47 2010 -0800 114462 114463 DRI2: prevent swap wakes from waking MSC waiters 114464 114465 If a few swaps were queued leading to a throttle related block on the 114466 client, and then the client submitted an MSC wait, one of the previous 114467 swap wakeups could have caused the MSC wait to complete early. Add a 114468 flag for this to prevent a swap wake from prematurely waking an MSC 114469 waiter. 114470 114471 Reported-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114472 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114473 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114474 114475commit b00d435ddf2e9817e33bfd5f7e9b905442dc23c7 114476Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114477Date: Mon Mar 8 12:41:25 2010 -0800 114478 114479 DRI2: handle swapsPending better 114480 114481 Avoid a potential swapsPending underflow by incrementing it before 114482 ScheduleSwap, which may complete it immediately. And be sure to 114483 decrement it again in case the schedule failed. 114484 114485 Reported-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114486 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114487 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114488 114489commit 0294ff2a5cadddc8fcc77ba9a851f979f0b91fc3 114490Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114491Date: Mon Mar 8 12:39:54 2010 -0800 114492 114493 DRI2: throttle swaps at submission time too 114494 114495 We need to throttle swaps here in addition to when the context is made 114496 current to avoid causing problems with clients that just swap. 114497 Throttling here also ensures our swaps get ordered as long as we block 114498 the client occasionally. 114499 114500 Reported-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114501 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114502 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114503 114504commit db1c7cb604167baf49e61be4c09ccf7b592c4af3 114505Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114506Date: Mon Mar 8 12:38:37 2010 -0800 114507 114508 DRI2: advertise lowest supported DRI2 protocol version 114509 114510 Update our supported DRI2 protocol version as each driver does 114511 DRI2ScreenInit, since depending on available kernel features, each DDX 114512 may support different callbacks and therefore protocol. 114513 114514 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114515 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114516 114517commit 87ca6320f26eb3129e3c19056e1d8fa5c1784723 114518Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114519Date: Fri Mar 5 09:49:03 2010 -0800 114520 114521 DRI2: handle swap_interval of 0 correctly 114522 114523 A 0 swap interval means that swaps shouldn't be sync'd to vblank, so 114524 just complete the swap immediately in that case. 114525 114526 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114527 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114528 114529commit 8476d99231cb725c090305d60f1c1c889d25c8dc 114530Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114531Date: Fri Mar 5 09:15:24 2010 -0800 114532 114533 DRI2: drawable lifetime fixes 114534 114535 Handle drawable destruction and lifetime correctly. 114536 114537 Check whether the drawable priv is valid in DRI2SwapInterval(), 114538 DRI2WaitSBC() and DRI2WaitMSC(); it may have gone away, so be sure to 114539 check it before using it. 114540 114541 If more than 1 outstanding swap is queued, we may complete several after 114542 an app has exited. If we free it after the first one completes and the 114543 refcount reaches 0, we'll crash the server on subsequent completions. 114544 So delay freeing until all swaps complete and remove the error message 114545 as this is a normal occurence. To do this properly, we must also avoid 114546 destroying drawables in DRI2DestroyDrawable() if a swap or wait event is 114547 pending. 114548 114549 And finally, make sure we free drawables in DRI2WaitMSCComplete() if 114550 necessary (i.e. if the refcount has reached 0 and this MSC was the last 114551 pending event on the object). 114552 114553 Reported-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114554 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114555 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114556 114557commit b180e43977710b56ccfd6780f204ddcc952987a1 114558Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114559Date: Thu Mar 4 10:31:59 2010 -0800 114560 114561 DRI2: fix swapbuffers handling of SBC and target MSC 114562 114563 Returns expected SBC after completion of swap to caller, as required by 114564 OML_sync_control spec, instead of the last_swap_target value. 114565 114566 Passes target_msc, divisor, remainder, correctly for 114567 glXSwapBuffersMscOML() call, while retaining old behaviour for simple 114568 glXSwapBuffers() call. 114569 114570 An OML swap can have a 0 target_msc, which just means it needs to 114571 satisfy the divisor/remainder equation. Pass this down to the driver as 114572 needed so we can support it. 114573 114574 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114575 Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114576 114577commit 751e8c09d34df4b41e8d8384a3ec1bf5cb8ca028 114578Author: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114579Date: Sun Feb 21 05:26:00 2010 +0100 114580 114581 DRI2WaitSbc(): Fixes for correct semantic of glXWaitForSbcOML() 114582 114583 Added implementation for case target_sbc == 0. In that case, the 114584 function shall schedule a wait until all pending swaps for the drawable 114585 have completed. 114586 114587 Fix for non-blocking case. Old implementation returned random, 114588 uninitialized values for (ust,msc,sbc) if it returned immediately 114589 without scheduling a wait due to sbc >= target_sbc. 114590 114591 Now if function doesn't schedule a wait, but returns immediately, 114592 it returns the (ust,msc,sbc) of the most recently completed swap, 114593 i.e., the UST and MSC corresponding to the time when the returned 114594 current SBC was reached. 114595 114596 Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114597 114598commit 0de4974b90b10fa6a447cdf980b4a114c6c9e5a8 114599Author: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114600Date: Sun Feb 21 05:25:59 2010 +0100 114601 114602 DRI2: Fix glitches in DRI2SwapComplete() and DRI2WakeupClient() 114603 114604 DRI2SwapComplete(): Increment pPriv->swap_count++; before calling 114605 into callback for INTEL_swap_events extension, so the swap event 114606 contains the current SBC after swap completion instead of the 114607 previous one. 114608 114609 DRI2WakeupClient: Check for pPriv->target_sbc <= pPriv->swap_count, 114610 had wrong comparison pPriv->target_sbc >= pPriv->swap_count for 114611 unblocking of clients of DRI2WaitSBC(). 114612 114613 Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114614 114615commit 4c8ec49826a46eb3b36c69d2ad3f82320c179c38 114616Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114617Date: Thu Mar 4 09:54:15 2010 -0800 114618 114619 DRI2: make target_sbc signed 114620 114621 We need to track invalid targets as well as 0 targets, so just make it 114622 signed so our comparisons work like they should. 114623 114624 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114625 Reported-by: Kristian Høgsberg <krh@bitplanet.net> 114626 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114627 114628commit c4d54816f2ee4883d8f9bcf4595474fb58c95146 114629Author: Jesse Barnes <jbarnes@virtuousgeek.org> 114630Date: Thu Mar 4 09:19:13 2010 -0800 114631 114632 DRI2: fixup handling of last_swap_target 114633 114634 We need to initialize the swap target, which is passed to the driver to 114635 schedule events. Rather than using -1 to indicate that the field is 114636 uninitialized, just make sure we initialize it at drawable creation 114637 time. 114638 114639 Reviewed-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de> 114640 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 114641 114642commit 28a5f14b4089dccb8045cc4fdc923542a73dd22d 114643Author: Jeremy Huddleston <jeremyhu@apple.com> 114644Date: Sun Mar 28 13:46:23 2010 -0700 114645 114646 XQuartz: Re-query dixScreenOrigins as the value could've changed. 114647 114648 Fix a regression in 9c9c3a85b094a3c7b2763a572715d710325091aa 114649 114650 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114651 Acked-by: Adam Jackson <ajax@redhat.com> 114652 114653commit 436d0bb9cca122bfdde32902b683d2499f61e6fc 114654Author: Jeremy Huddleston <jeremyhu@apple.com> 114655Date: Sat Mar 27 17:35:10 2010 -0700 114656 114657 darwin: Generate crash reports on FatalError() 114658 114659 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114660 Acked-by: Adam Jackson <ajax@redhat.com> 114661 114662commit 73b3b67aac9f3938a96cb8822b9c270bd82ded5c 114663Author: Jeremy Huddleston <jeremyhu@apple.com> 114664Date: Mon Mar 22 09:30:51 2010 -0700 114665 114666 GLX: Remove a redundant initialization 114667 114668 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114669 Acked-by: Adam Jackson <ajax@redhat.com> 114670 114671commit 339207be6f184cc783076fc7e2cc12f92f57f2ba 114672Author: Jeremy Huddleston <jeremyhu@apple.com> 114673Date: Thu Mar 25 22:15:58 2010 -0700 114674 114675 XQuartz: Workaround weird key data reported on some layouts 114676 114677 This should make 'Unicode Hex Input' work as an input layout. 114678 114679 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114680 Acked-by: Adam Jackson <ajax@redhat.com> 114681 114682commit aa91508356f243edc3b11795b1481edcfe0d39c2 114683Author: Dan Nicholson <dbn.lists@gmail.com> 114684Date: Wed Mar 24 23:18:35 2010 -0700 114685 114686 config/udev: Prefer product name from attribute rather than uevent 114687 114688 The input device product name for evdev devices in the kernel uevent has 114689 embedded quotes that aren't expected here. Use the sysfs name attribute 114690 instead, which does not suffer this problem. The uevent name will be 114691 used as a fallback if no name attribute is found. 114692 114693 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 114694 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114695 114696commit d00c3298a698ddddc8967215cef7dd941fa15804 114697Author: Fernando Carrijo <fcarrijo@yahoo.com.br> 114698Date: Wed Mar 24 20:11:34 2010 -0300 114699 114700 Cleanup some comments in SpriteRec 114701 114702 Signed-off-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 114703 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114704 114705commit 70bd02f2ea3a91c0340df97ca3b8a4f9400f6f46 114706Author: Peter Hutterer <peter.hutterer@who-t.net> 114707Date: Tue Mar 23 12:13:56 2010 +1000 114708 114709 xfree86: merge driver from the input class into the options. 114710 114711 A driver that is assigned by an input class is only present as idev->driver. 114712 The driver itself has no access to this information once PreInit is called. 114713 For devices that rely on chain-hotplugging (wacom), this means that for the 114714 second device the driver information is lost and the second device cannot be 114715 initialized through NewInputDeviceRequest. Although this could be worked 114716 around by hardcoding the driver name in the wacom driver, having the 114717 assigned driver in the options seems like the better solution. 114718 114719 This issue only manifests itself with the udev backend. With HAL, the driver 114720 is assigned by HAL and the option is duplicated in config/hal.c. 114721 114722 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114723 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 114724 114725commit 579715f830fbbca9e1ecb17dc18176132f5969e7 114726Author: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 114727Date: Wed Mar 17 12:16:57 2010 +0200 114728 114729 os: Prevent backtrace from being stopped in noreturn functions. 114730 114731 There are two noreturn functions in the X server: FatalError and 114732 AbortServer. Having any of those two functions in the middle of a call 114733 stack will prevent unwinding the program properly and stops the 114734 backtrace at those functions in gdb. 114735 114736 The file containing FatalError and AbortServer, os/log.c, has to be 114737 compiled with the -mapcs-frame option on ARM to get proper 114738 backtraces. Automake imposes its own restrictions on compiling 114739 individual source files with different options. The recommended way to 114740 do this is to put os/log.c into a convenience library and add this 114741 library inside os/libos.la. See the documentation of GNU Automake 114742 manual, version 1.11.1, section 27.8 Per-Object Flags Emulation, for 114743 details. 114744 114745 Signed-off-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 114746 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 114747 Signed-off-by: Keith Packard <keithp@keithp.com> 114748 114749commit e086b99c1f5ce351b578de7cd9f616bc79d6cf64 114750Author: Mikhail Gusarov <dottedmag@dottedmag.net> 114751Date: Tue Mar 23 01:03:53 2010 +0600 114752 114753 kdrive: Bump evdev maxKeycode 114754 114755 There are keycodes > 193 in evdev, e.g. KEY_WIMAX which is 246 . 114756 114757 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 114758 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 114759 Acked-by: Adam Jackson <ajax@nwnk.net> 114760 Signed-off-by: Keith Packard <keithp@keithp.com> 114761 114762commit 185185eeb44a277c324be0f58a4b4a469b56b69b 114763Author: Peter Harris <pharris@opentext.com> 114764Date: Tue Mar 23 12:08:19 2010 -0400 114765 114766 Fix crash when all glyphs of a given depth are freed, but not all glyphsets 114767 114768 This is how the crash can be triggered with only two clients on the system: 114769 Client A: (already running) 114770 Client B: Connect 114771 Client B: CreateGlyphSet(depthN) 114772 Client A: Disconnect 114773 Server: free globalGlyphs(depthN) 114774 Client B: AddGlyphs(depthN) 114775 Server: SEGV 114776 114777 This crash was introduced with the FindGlyphsByHash function 114778 in 516b96387b0e57b524a37a96da22dbeeeb041712. Before that revision, 114779 ResizeGlyphSet was always called before FindGlyphRef, which would 114780 re-create globalGlyphs(depthN) if necessary. 114781 114782 X.Org Bug 20718 <http://bugs.freedesktop.org/show_bug.cgi?id=20718> 114783 114784 Reviewed-by: Adam Jackson <ajax@redhat.com> 114785 Signed-off-by: Peter Harris <pharris@opentext.com> 114786 Signed-off-by: Keith Packard <keithp@keithp.com> 114787 114788commit 8311cd5f89ca6781842bb24671b8122cdf8be148 114789Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 114790Date: Tue Mar 23 10:37:36 2010 -0700 114791 114792 XKB: Fix garbage initialization 114793 114794 XkbEnableDisableControls set extra garbage bits on the xkbControlsNotify 114795 changedControls mask because it was uninitialized on the stack. 114796 114797 Found by clang 114798 114799 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 114800 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114801 114802commit 0ecfab89d1754b0b7e32b3964b40663d1810cd62 114803Author: Peter Hutterer <peter.hutterer@who-t.net> 114804Date: Mon Mar 22 14:04:09 2010 +1000 114805 114806 xfree86: remove if 1 from the dawn of time. 114807 114808 7+ years for an if 1 should be enough to just admit that there's no other 114809 option. 114810 114811 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114812 Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 114813 Reviewed-by: Corbin Simpson <MostAwesomeDude@gmail.com> 114814 114815commit 6389405b56d2221d8a7b80c4bb3bd3b4e1a29bab 114816Author: Peter Hutterer <peter.hutterer@who-t.net> 114817Date: Mon Mar 22 09:20:49 2010 +1000 114818 114819 configure: Always define XINPUT. 114820 114821 This define was removed in e251c9e75afdc "Remove all traces of #ifdef XINPUT 114822 and the matching bits from the configure.ac". Drivers that support multiple 114823 server versions and still check for ifdef XINPUT now always build without 114824 XINPUT support if they're calling XORG_DRIVER_CHECK_EXT(XINPUT, inputproto). 114825 114826 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 114827 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 114828 114829commit 0c2fde5c8ad6e94b4ed1588aa93256a2b64f74d9 114830Author: Tomas Carnecky <tom@dbservice.com> 114831Date: Mon Mar 22 11:20:15 2010 -0700 114832 114833 Fix typos in the swap functions 114834 114835 This should fix bug #3539. 114836 114837 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 114838 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 114839 Reviewed-by: Keith Packard <keithp@keithp.com> 114840 Signed-off-by: Keith Packard <keithp@keithp.com> 114841 114842commit 3083c5d0c4386cdd7083b7a83ac72fdad2f1e61e 114843Author: Michel Dänzer <daenzer@vmware.com> 114844Date: Mon Mar 22 18:01:17 2010 +0100 114845 114846 Xext: Fix cursor reference counting hazard. 114847 114848 Make sure the reference count of the new cursor is increased before the old 114849 one is decreased, otherwise bad things will happen if they're one and the 114850 same and the reference count is 1 initially. Not sure this can actually happen 114851 here, but better safe than sorry. 114852 114853 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 114854 Reviewed-by: Roland Scheidegger <sroland@vmware.com> 114855 Signed-off-by: Keith Packard <keithp@keithp.com> 114856 114857commit b29220dc765cb6f878c5466e00e4bd21f3bd803d 114858Merge: c3da76643 ed31d50b5 114859Author: Keith Packard <keithp@keithp.com> 114860Date: Sun Mar 21 23:01:58 2010 -0700 114861 114862 Merge remote branch 'jeremyhu/master' 114863 114864commit c3da76643aeebdd24047e3e14a860507eac37ec3 114865Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114866Date: Thu Mar 18 23:19:20 2010 -0500 114867 114868 Cygwin/X: Make X -> XWin symlink during install 114869 114870 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114871 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 114872 114873commit abf4e0b7e349b0238b2b97a16b8a308b5026af31 114874Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114875Date: Mon Mar 22 00:44:28 2010 -0500 114876 114877 New header for XF86Bigfont server functions 114878 114879 Xext/xf86bigfont.c contains three non-static functions which are called 114880 elsewhere in the server. This creates a new header containing these 114881 declarations in order to fix several warnings: 114882 114883 xf86bigfont.c:285: warning: no previous prototype for `XF86BigfontFreeFontShm' 114884 dixfonts.c:502: warning: implicit declaration of function `XF86BigfontFreeFontS$ 114885 dixfonts.c:502: warning: nested extern declaration of `XF86BigfontFreeFontShm' 114886 log.c:436: warning: implicit declaration of function `XF86BigfontCleanup' 114887 log.c:436: warning: nested extern declaration of `XF86BigfontCleanup' 114888 114889 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114890 Reviewed-by: Julien Cristau <jcristau@debian.org> 114891 114892commit 57a049ea89b008c0b60316c3b6e6ff5c8fbd4cad 114893Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114894Date: Thu Mar 18 04:46:20 2010 -0500 114895 114896 Xext: fix old-style function definitions in xf86bigfont.c 114897 114898 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114899 Reviewed-by: Julien Cristau <jcristau@debian.org> 114900 114901commit 822b9f9a3e822df8848995eda246f83864d1366f 114902Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114903Date: Sun Feb 21 01:24:44 2010 -0600 114904 114905 mi: remove deprecated #include <X11/extensions/xf86bigfstr.h> in miinitext.c 114906 114907 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114908 Reviewed-by: Julien Cristau <jcristau@debian.org> 114909 114910commit 946b49ebcbf47e030a6e00ce1a699f3b055f638a 114911Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114912Date: Wed Mar 17 17:37:05 2010 -0500 114913 114914 Catch errors in recursive relink targets 114915 114916 If make relink fails in a subdirectory, we need to catch the error 114917 otherwise make will continue iterating the 'for' loop. 114918 114919 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114920 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 114921 114922commit 0820a6e2fb005b5f856dda25e1644f34fc994129 114923Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114924Date: Wed Mar 17 13:40:41 2010 -0500 114925 114926 Fix .man.N targets for AM_SILENT_RULES 114927 114928 Add $(AM_V_GEN) for sed-based rules so they appear as expected with 114929 automake silent rules, and $(AM_V_at) to completely hide cp/ln/rm 114930 commands which are not prone to fail. 114931 114932 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114933 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 114934 Reviewed-by: Julien Cristau <jcristau@debian.org> 114935 114936commit 9be4157391edf0c5fc4ee36adfb1eb1c3bdb8e3b 114937Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114938Date: Wed Mar 17 13:33:39 2010 -0500 114939 114940 Respect value of SED from configure 114941 114942 We now use libtool, which calls AC_PROG_SED and sets SED as the path to 114943 a fully-functional 'sed' (which may also be called 'gsed' if GNU sed is 114944 installed alongside a proprietary version). Therefore we should respect 114945 the value of SED so we are sure to use the correct one. 114946 114947 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114948 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 114949 Reviewed-by: Julien Cristau <jcristau@debian.org> 114950 114951commit 5e00f464c546debf9164f24c6d648623f7ee5946 114952Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114953Date: Wed Mar 17 14:56:24 2010 -0500 114954 114955 Fix relink targets for silent rules 114956 114957 Add $(AM_V_at) to all relink make targets to silence them when automake 114958 silent rules are in use. 114959 114960 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114961 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 114962 Reviewed-by: Julien Cristau <jcristau@debian.org> 114963 114964commit fa7e062962d17862032bc600bfcffd486addb063 114965Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114966Date: Wed Mar 17 13:52:38 2010 -0500 114967 114968 Use EXEEXT in relink rules for portable DDXs 114969 114970 On Cygwin and MinGW, executables use the .exe suffix. Autoconf and 114971 automake set EXEEXT on these platforms, and leave it empty on others 114972 where no suffix is used. $(EXEEXT) must be appended to executable names 114973 in custom rules for portability: 114974 114975 http://www.gnu.org/software/automake/manual/html_node/EXEEXT.html 114976 114977 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114978 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 114979 Reviewed-by: Julien Cristau <jcristau@debian.org> 114980 114981commit 037869d7b876f43158df88d63788b93a5d6187c1 114982Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114983Date: Wed Mar 17 13:48:49 2010 -0500 114984 114985 kdrive: Use $(MAKE) in relink rules 114986 114987 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114988 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 114989 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 114990 Reviewed-by: Julien Cristau <jcristau@debian.org> 114991 114992commit 8e5de45513577c63a33833931f2afd0cc59d42a7 114993Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 114994Date: Wed Feb 24 02:56:59 2010 -0600 114995 114996 Use libtool -export-dynamic flag for portability 114997 114998 The linker flag required for exporting symbols in executables varies 114999 by platform. libtool handles this with a single -export-dynamic 115000 flag (not to be confused with the similarly-named ELF linker flag) 115001 which tells it to use the correct platform-specific flag at link time. 115002 115003 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 115004 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 115005 Reviewed-by: Keith Packard <keithp@keithp.com> 115006 115007commit 2a11ffa977d2267d8b2f27b76490a98e9c73b8c7 115008Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 115009Date: Fri Feb 19 14:31:28 2010 -0600 115010 115011 Rename xdmx client to dmxinfo 115012 115013 The DMX Xdmx server and xdmx client cannot both be installed on 115014 case-insensitive file systems. The client is undocumented and 115015 so renaming it is the best option. 115016 115017 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 115018 Reviewed-by: Keith Packard <keithp@keithp.com> 115019 115020commit fbb4903eaa23174d21ab8c640d9de1fe3eb17539 115021Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 115022Date: Sun Feb 21 02:08:08 2010 -0600 115023 115024 Cygwin/X: Fix windres rule for automake silent rules 115025 115026 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 115027 Reviewed-by: Rémi Cardona <rem@gentoo.org> 115028 Reviewed-by: Keith Packard <keithp@keithp.com> 115029 115030commit 8db894690988f2bac9c0acb96fd39811d2b42f47 115031Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 115032Date: Fri Feb 19 00:05:28 2010 -0600 115033 115034 Cygwin/X: Fix make dist after 11252ed82e1f361b99e86521ac9314f868bd1a3a 115035 115036 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 115037 Acked-by: Adam Jackson <ajax@redhat.com> 115038 Reviewed-by: Keith Packard <keithp@keithp.com> 115039 115040commit ed31d50b5f7f25e4db986711699704e615b8afcb 115041Author: Jeremy Huddleston <jeremyhu@apple.com> 115042Date: Fri Mar 19 17:24:27 2010 -0700 115043 115044 XQuartz: Constrain the pointer to the updated display bounds on display reconfigure. 115045 115046 http://xquartz.macosforge.org/trac/ticket/346 115047 115048commit 927480be12ff61deb4e8120d2f90ddef3a1a6beb 115049Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 115050Date: Sat Mar 20 11:03:28 2010 -0700 115051 115052 XQuartz: pbproxy: Make standalone xpbproxy respect the launchd prefix 115053 115054 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 115055 115056commit d16bc8a3cdbe0e8b9cba836c7aa2ff7c0458049e 115057Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 115058Date: Sat Mar 20 03:28:57 2010 -0700 115059 115060 XQuartz: xpbproxy: Cleanup xpbproxy threading 115061 115062 Confine xpbproxy to a single thread 115063 Runs inside its own CFRunLoop 115064 115065 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 115066 115067commit 9c9c3a85b094a3c7b2763a572715d710325091aa 115068Author: Jeremy Huddleston <jeremyhu@apple.com> 115069Date: Fri Mar 19 17:23:12 2010 -0700 115070 115071 XQuartz: Minor cleanup 115072 115073 Move RandRInit to where it will need to be (not yet implemented) 115074 115075 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 115076 115077commit bb75d0df8b5238bfe8b011bb5737fae2a3584290 115078Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 115079Date: Fri Mar 19 15:26:50 2010 -0700 115080 115081 XQuartz: GLX: Fix Availability for Tiger ppc workaround 115082 115083 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 115084 115085commit eac7cdabecafb7c505795207182ab2578d672c06 115086Author: Jeremy Huddleston <jeremyhu@apple.com> 115087Date: Tue Mar 16 11:29:08 2010 -0700 115088 115089 Revert "XQuartz: Explicitly pass a bellProc to make XBell() work again." 115090 115091 I'm not quite sure why this was necessary, but DDXRingBell is being called 115092 from CoreKeyboardBell, so we don't need a separate bellProc which would 115093 result in multiple rings. 115094 115095 This reverts commit 9071b0d69748cfa7ecca17b4cb0e431bbb0ef2a4. 115096 115097 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 115098 115099commit 13c007f9224ad871193e40466e64df1477ed26b0 115100Author: Keith Packard <keithp@keithp.com> 115101Date: Sun Mar 21 17:23:46 2010 -0700 115102 115103 Bump to 1.7.99.902 -- 1.8 RC2 115104 115105commit 235fa5030428084368e5be57fca695647b7b79c4 115106Merge: 1c612acca fa5103a02 115107Author: Keith Packard <keithp@keithp.com> 115108Date: Sun Mar 21 15:38:40 2010 -0700 115109 115110 Merge commit 'fa5103a02bd509e4a102afdad2ab26cb22210367' 115111 115112commit 1c612acca8568fcdf9761d23f112adaf4d496f1b 115113Author: Peter Hutterer <peter.hutterer@who-t.net> 115114Date: Wed Mar 17 14:32:38 2010 +1000 115115 115116 dix: if owner-events is true for passive grabs, add the window mask (#25400) 115117 115118 A client requesting a GrabModeSync button grab, owner-events true, with only 115119 the ButtonRelease mask set would never receive the press event even if the 115120 grab window had the ButtonPress mask set. 115121 115122 The protocol requires that if owner-events is true, then the delivery mask 115123 is the combination of the grab mask + the window event mask. 115124 115125 X.Org Bug 25400 <http://bugs.freedesktop.org/show_bug.cgi?id=25400> 115126 115127 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115128 Tested-by: Jim Ramsay <i.am@jimramsay.com> 115129 Signed-off-by: Keith Packard <keithp@keithp.com> 115130 115131commit 95ca39b981624df5338a16b506289014c431ae82 115132Merge: e7ff95663 bf181915e 115133Author: Keith Packard <keithp@keithp.com> 115134Date: Sun Mar 21 15:24:06 2010 -0700 115135 115136 Merge remote branch 'jturney/master' 115137 115138commit e7ff956638780dc1c6f160a1fd041b92cebff9ac 115139Author: Oliver McFadden <oliver.mcfadden@nokia.com> 115140Date: Mon Mar 15 23:46:36 2010 +0200 115141 115142 common: xf86Configure: alloc_strlen: Allocated memory does not have space for the terminating NUL of the string 115143 115144 buffer_alloc: Called allocating function "realloc" which allocated memory dictated by parameter "len + strlen(displaySize_string)" 115145 alloc_strlen: Allocated memory does not have space for the terminating NUL of the string 115146 var_assign: Assigned "ptr->mon_comment" to storage allocated by "realloc(ptr->mon_comment, len + strlen(displaySize_string))" 115147 115148 Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com> 115149 Reviewed-by: Matt Turner <mattst88@gmail.com> 115150 Signed-off-by: Keith Packard <keithp@keithp.com> 115151 115152commit aeae6637b52b042cf8b07c6943abcd5d31968d49 115153Author: Oliver McFadden <oliver.mcfadden@nokia.com> 115154Date: Mon Mar 15 23:46:34 2010 +0200 115155 115156 fb: fbFinishScreenInit: leaked_storage: Variable "(visuals|depths)" goes out of scope 115157 115158 Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com> 115159 Reviewed-by: Matt Turner <mattst88@gmail.com> 115160 Signed-off-by: Keith Packard <keithp@keithp.com> 115161 115162commit d9ee31ab05efbd2d598cc93a41a8834f84d30803 115163Author: Oliver McFadden <oliver.mcfadden@nokia.com> 115164Date: Mon Mar 15 23:46:33 2010 +0200 115165 115166 exa: exaFinishAccess: Overrun of static array "pExaScr->access" of size 6 at position 6 with index variable "i" 115167 115168 Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com> 115169 Reviewed-by: Keith Packard <keithp@keithp.com> 115170 Signed-off-by: Keith Packard <keithp@keithp.com> 115171 115172commit bf181915e103b0659c5a0bc3b2f6bccb18c9ec36 115173Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 115174Date: Mon Feb 1 17:11:51 2010 +0000 115175 115176 Cygwin/X: Allow the default log location to be configurable 115177 115178 Allow the default log location to be configurable (e.g. /var/log), 115179 and use separate logs for each display instance (e.g. XWin.0.log). 115180 115181 Make the type of g_pszLogFile const char*, per os/log.c:LogInit(). 115182 115183 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 115184 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 115185 115186commit cca4952d750779a4c58a11fe08a53336042930bb 115187Author: Colin Harrison <colin.harrison@virgin.net> 115188Date: Fri Feb 26 14:40:30 2010 +0000 115189 115190 Xming: Warning fixes 115191 115192 Fix warnings due to prototypes not specifying function arguments 115193 Fix warning with RegQueryValueEx() 115194 Tidy up an include 115195 115196 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 115197 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 115198 115199commit b8cf4153f84404b2a1b62a247d5d36d24bf481da 115200Author: Colin Harrison <colin.harrison@virgin.net> 115201Date: Fri Feb 26 14:41:00 2010 +0000 115202 115203 Xming: Dead code removal 115204 115205 Remove some dead code, mostly code made obsolete by mandatory XKB 115206 115207 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 115208 Reviewed-by: Jon TURNEY <jon.turney@dronecode.org.uk> 115209 115210commit 9b18f7ac0ba9aadb8c86bc2717cbdff75e951b69 115211Author: Jon TURNEY <jon.turney@dronecode.org.uk> 115212Date: Fri Feb 26 14:22:57 2010 +0000 115213 115214 Cygwin/X: Tidy up some cosmetic issues in log strings 115215 115216 Tidy up some cosmetic issues in log strings: 115217 - Add missing '\n' 115218 - Fix some strings starting with '\n' 115219 - Remove '\f' from some log strings 115220 115221 These all just look daft in a log with timestamps. 115222 115223 Also clarify log message about screen origin coordinates 115224 115225 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 115226 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 115227 115228commit 178e830378c3514b9c53a5b1c6d2d5f930c3779b 115229Author: Jon TURNEY <jon.turney@dronecode.org.uk> 115230Date: Mon Feb 22 19:13:30 2010 +0000 115231 115232 Cygwin/X: Fix thinko in mount option checking 115233 115234 Fix a thinko in mount option checking. 115235 Use symbolic names for values assigned to binary flag for clarity. 115236 115237 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 115238 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 115239 115240commit fa5103a02bd509e4a102afdad2ab26cb22210367 115241Author: Francisco Jerez <currojerez@riseup.net> 115242Date: Wed Feb 24 23:18:01 2010 +0100 115243 115244 dri2: No need to blit from front on DRI2GetBuffers if they're just being reused. 115245 115246 It can be quite an expensive operation, so we're better off not doing 115247 it unless it's totally required. 115248 115249 Signed-off-by: Francisco Jerez <currojerez@riseup.net> 115250 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 115251 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 115252 115253commit e38e01081af42825bb7b44d18aa35845318f8556 115254Author: Francisco Jerez <currojerez@riseup.net> 115255Date: Wed Feb 24 23:18:00 2010 +0100 115256 115257 Import linked list helpers from the intel DDX. 115258 115259 Borrowed from i830.h, except for list_for_each_entry(). 115260 115261 Signed-off-by: Francisco Jerez <currojerez@riseup.net> 115262 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 115263 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 115264 115265commit b8a3be5f34016b745e38cd53825a4e398c6127bc 115266Author: Francisco Jerez <currojerez@riseup.net> 115267Date: Wed Feb 24 23:17:59 2010 +0100 115268 115269 Add a ConfigNotify hook. 115270 115271 Executed from the ConfigureWindow request, right before sending 115272 ConfigureNotify to the clients. 115273 115274 This commit breaks the ScreenRec ABI. 115275 115276 Signed-off-by: Francisco Jerez <currojerez@riseup.net> 115277 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 115278 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 115279 115280commit 67a8c659f25218904bae64aac6e98e326c90330b 115281Author: Roland Scheidegger <sroland@vmware.com> 115282Date: Fri Mar 12 16:53:30 2010 +0100 115283 115284 hw/xfree86: move reference counting out of the UseHWCursor[ARGB] functions 115285 115286 The problem is that the xf86_use_hw_cursor(_argb) functions may get this 115287 correctly now, some drivers will replace these generic versions with their 115288 own functions. It is pretty insane to expect them to do reference counting 115289 of the cursor (as an example, look at driver/xf86-video-vmware to see how 115290 that looks like as a workaround). There are even places in xserver itself 115291 which replace these two functions. 115292 The segfaults if no reference counting is done are caused because the 115293 reference count of the cursor reached zero, hence the cursor was freed, 115294 however xf86CursorEnableDisableFBAccess() brought it back to life from 115295 the dead (from the SavedCursor). 115296 This patch hence adds reference counting in xf86CursorSetCursor. As per Michel 115297 Daenzer's suggestion, also free the cursor upon xf86CursorCloseScreen. 115298 In theory with this it should be possible to remove the reference 115299 counting in the UseHwCursor functions I think, though it should also be 115300 safe to keep them. 115301 115302 Signed-off-by: Roland Scheidegger <sroland@vmware.com> 115303 Reviewed-by: Michel Dänzer <michel@daenzer.net> 115304 Signed-off-by: Keith Packard <keithp@keithp.com> 115305 115306commit 75efb46a14fe45ffe73faff637b1fa6d017e1e52 115307Author: Roland Scheidegger <sroland@vmware.com> 115308Date: Fri Mar 12 16:52:56 2010 +0100 115309 115310 hw/xfree86: fix refcounting in xf86_use_hw_cursor 115311 115312 This is the same fix as was done in 115313 fcdc1d78cca3b8bb6b77d53eda7e21d649df6943 for xf86_use_hw_cursor_argb. 115314 115315 Signed-off-by: Roland Scheidegger <sroland@vmware.com> 115316 Reviewed-by: Michel Dänzer <michel@daenzer.net> 115317 Signed-off-by: Keith Packard <keithp@keithp.com> 115318 115319commit df9b6f16b27398545cd4cff8a56dd59a3813351d 115320Merge: 5f169f549 5172253ba 115321Author: Keith Packard <keithp@keithp.com> 115322Date: Mon Mar 15 08:26:58 2010 -0700 115323 115324 Merge remote branch 'jeremyhu/master' 115325 115326commit 5f169f54936c9868ad0f3778cb95c1f35eef41ea 115327Author: Gaetan Nadon <memsize@videotron.ca> 115328Date: Fri Mar 5 10:35:54 2010 -0500 115329 115330 XQuartz: remove undefined XSERVER_CFLAGS variable 115331 115332 This is a variable local to configure.ac which is not AC_SUBST() 115333 It is undefined in any generated Makefile. 115334 115335 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 115336 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 115337 Signed-off-by: Keith Packard <keithp@keithp.com> 115338 115339commit 5172253bae3b9867118c6717434e73c173acd5e9 115340Author: Jeremy Huddleston <jeremyhu@apple.com> 115341Date: Wed Feb 10 15:52:14 2010 -0800 115342 115343 XQuartz: Use an empty xkb keymap by default 115344 115345 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 115346 115347commit b117bc7a441bec8f61610fb384d747112f73d236 115348Author: Jeremy Huddleston <jeremyhu@apple.com> 115349Date: Thu Mar 11 00:06:01 2010 -0800 115350 115351 XQuartz: Include os.h for OsAbort() 115352 115353 Fixes regression from 5b9a52be7e975e59e0bbc6b43539ecaff96b2ecd 115354 115355 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 115356 115357commit b7a16117c6d87a9d33a5f682b592b4507f2c065e 115358Author: Jeremy Huddleston <jeremyhu@apple.com> 115359Date: Thu Mar 11 00:03:08 2010 -0800 115360 115361 XQuartz: GLX: Fix prototype for swapBuffers 115362 115363 This was a regression introduced by 04a54f69a8085ab3fe11a8713bd8b6b16ed1db27 115364 115365 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 115366 115367commit 178da6534fe7ab4b99fb87925c04e6f963c88583 115368Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 115369Date: Mon Mar 8 16:33:37 2010 -0500 115370 115371 Xext: Link to external libraries when necessary. 115372 115373 Although the DDX should be linked to the necessary libraries, we may 115374 also need to pull them in directly to the module to ensure the symbols 115375 are resolved at runtime. Should fix this bug with XSELINUX: 115376 115377 /usr/bin/X: symbol lookup error: 115378 /usr/lib64/xorg/modules/extensions/libextmod.so: undefined symbol: 115379 is_selinux_enabled 115380 115381 -v2: use _LIBADD instead of _LIBS; remove SELINUX_LIBS from 115382 XSERVER_SYS_LIBS as it should only be needed in extmod. 115383 115384 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 115385 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115386 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115387 115388commit a2ea8c2f2cc53607d57d60f41c879380ea70cd02 115389Author: Daniel Stone <daniel@fooishbar.org> 115390Date: Thu Mar 11 14:19:04 2010 +0200 115391 115392 Record: Avoid duplicates from replaying frozen events 115393 115394 Reintroduce a check which used to be there in the old 115395 ProcessKeyboardEvent/ProcessPointerEvent codepath, which avoids us 115396 recording events subject to a grab twice: once when it's first processed 115397 in EnqueueEvent, and then again when it's thawed and being replayed. 115398 115399 This required a tiny amount of code motion to expose syncEvents. 115400 115401 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 115402 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 115403 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115404 115405commit f2eacb4646beb25d055de22868f93e6b24f229b6 115406Author: Peter Hutterer <peter.hutterer@who-t.net> 115407Date: Wed Mar 10 09:39:05 2010 +1000 115408 115409 Revert "dix: Use DeliverGrabbedEvent for implicit passive grabs (#25400)" 115410 115411 Several users have pointed out that this commit introduces regressions, most 115412 notably perhaps fluxbox which essentially stops working after a few clicks. 115413 115414 This reverts commit cf72b5437d2d620521279077a29c5df6d0fbb576. 115415 115416 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115417 115418commit df9f3273041c6c3e0da2d2254e8c156cd582e296 115419Author: Peter Hutterer <peter.hutterer@who-t.net> 115420Date: Wed Mar 10 09:24:32 2010 +1000 115421 115422 xfree86: fix xf86Config.c build error in --enable-debug mode. (#26971) 115423 115424 xf86Config.c: In function 'configInputDevices': 115425 xf86Config.c:1514: error: request for member 'lay_identifier' in something 115426 not a structure or union 115427 make[5]: *** [xf86Config.lo] Error 1 115428 115429 Introduced with e1165632bdfbd720889ed1adf5f7ab338032c0ee. 115430 115431 X.Org Bug 26971 <http://bugs.freedesktop.org/show_bug.cgi?id=26971> 115432 115433 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115434 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 115435 115436commit 116068103282924ef1476231d13d54719a797252 115437Author: Peter Hutterer <peter.hutterer@who-t.net> 115438Date: Tue Mar 9 09:35:23 2010 +1000 115439 115440 xfree86: don't warn about nonexisting core pointer/keyboard in config. 115441 115442 In the vast majority of cases there is no xorg.conf that specifies a core 115443 pointer/keyboard. Skip this warning, since we'll get another notification 115444 about how the server relies on the config backend for input devices anyway. 115445 115446 Leave the warning in for the error case (AEI off). 115447 115448 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115449 Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 115450 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 115451 115452commit c39c8ece9bd3ce025549f4bac7794f5a62266f4a 115453Author: Peter Hutterer <peter.hutterer@who-t.net> 115454Date: Wed Feb 10 15:06:52 2010 +1000 115455 115456 os: remove INTERNAL_MALLOC define. 115457 115458 There doesn't seem to be anything that defines it and given that the 115459 counterpart (the X internal malloc) was removed in 115460 01cfba75229f4b9bf1e4fe80814931acdacde14c it's unlikely to work anyway. 115461 115462 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115463 115464commit 9f462ff9082634719e64d2d8d4dbd09ec7deaf2e 115465Author: Peter Hutterer <peter.hutterer@who-t.net> 115466Date: Fri Feb 26 14:02:09 2010 +1000 115467 115468 dix: Clip only into axis ranges if we're in absolute mode. (#26543) 115469 115470 An absolute device in relative mode may provide valuators outside of the 115471 axis range. Clipping back into the range prevents screen crossings in a 115472 multi-screen (Xinerama) setup as the required screen edge for crossing is 115473 never met: miPointerSetPosition crosses the screen conditional to the X 115474 coordinate being equal to the screen width or _less than_ 0. While the 115475 former can be met when clipping into the coordinate range and scaling, the 115476 latter cannot, resulting in a mouse pointer that gets stuck on the rightmost 115477 screen. 115478 115479 This patch only applies axis clipping for valuators in mode Absolute. If 115480 relative, we allow the values to get above/below the axis ranges. Doesn't 115481 matter, miPointerSetPosition will reset the values to the allowed range even 115482 if no screen was crossed. 115483 This leads to interesting values provided to clients, the valuator range of 115484 the device resets once a screen is crossed and essentially reflects 115485 the position of the cursor on the screen - scaled into the valuator range. 115486 The values themselves are valid given the range though. 115487 115488 In theory, the XI1 specs require that a relative device has a min/max range 115489 of 0/0. This doesn't really go well with devices that actually can switch 115490 mode between relative and absolute since they would have to reset their axis 115491 range when switching. If multiple XI clients are in use, we have no method 115492 of notifying them about the changes, so other clients may continue to use 115493 the wrong axis ranges (note: XI1 wasn't really designed to have multiple 115494 clients use a device). Expecting all relative devices to have this min/max 115495 of 0 is unrealistic at this point. 115496 115497 So pick what is possibly the lesser of all evils, pass the beer and despair. 115498 115499 X.Org Bug 26543 <http://bugs.freedesktop.org/show_bug.cgi?id=26543> 115500 115501 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115502 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 115503 115504commit 1f0e69b3a13258b79bffdec76b46f7960eb334b3 115505Author: Matt Turner <mattst88@gmail.com> 115506Date: Mon Mar 8 10:54:21 2010 -0500 115507 115508 Don't check for Xinerama. 115509 115510 It doesn't seem to be used anywhere, so don't require it. 115511 115512 CC: Jerome Glisse <jglisse@redhat.com> 115513 CC: Alex Deucher <alexdeucher@gmail.com> 115514 CC: Dave Airlie <airlied@redhat.com> 115515 Signed-off-by: Matt Turner <mattst88@gmail.com> 115516 115517commit bbae92795c7eab062e6722c42fa7915e0cee5d69 115518Author: Matt Turner <mattst88@gmail.com> 115519Date: Mon Feb 15 20:08:09 2010 -0500 115520 115521 Replace assembly with generic unaligned access code 115522 115523 Removes Alpha assembly, and probably works around unaligned accesses on 115524 other sensitive platforms. 115525 115526 Signed-off-by: Matt Turner <mattst88@gmail.com> 115527 Acked-by: Adam Jackson <ajax@redhat.com> 115528 Compiled-by: Tiago Vignatti <tiago.vignatti@nokia.com> 115529 Signed-off-by: Keith Packard <keithp@keithp.com> 115530 115531commit 1c2abec479cfbac9aeadc121af0825d368317658 115532Author: Peter Hutterer <peter.hutterer@who-t.net> 115533Date: Fri Feb 26 11:44:17 2010 +1000 115534 115535 configure: move SELINUX_LIBS to XSERVER_SYS_LIBS 115536 115537 All ddxs require linking against selinux if enabled. 115538 115539 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115540 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 115541 Reviewed-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115542 Signed-off-by: Keith Packard <keithp@keithp.com> 115543 115544commit cab99b947c66a42dbf8a906bdb02e23aa7cab519 115545Author: Peter Hutterer <peter.hutterer@who-t.net> 115546Date: Fri Feb 26 11:41:38 2010 +1000 115547 115548 configure: move libselinux requirement up to the common section. 115549 115550 Let's have all version-specific requirements in one block. 115551 115552 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115553 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 115554 Reviewed-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115555 Signed-off-by: Keith Packard <keithp@keithp.com> 115556 115557commit 9e7b6e565f6e69f20002bf03ed05fe8b63f86924 115558Author: Peter Hutterer <peter.hutterer@who-t.net> 115559Date: Fri Feb 26 10:14:00 2010 +1000 115560 115561 dmx: fix OsAbort()-related build error in examples/xinput.c 115562 115563 5b9a52be7e975e59e0bbc6b43539ecaff96b2ecd changed the server to use OsAbort() 115564 instead of abort(). xinput in dmx is a client program though and fails to 115565 link if it tries to use OsAbort(). Switch it back to using abort(). 115566 115567 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115568 Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 115569 Reviewed-by: Julien Cristau <jcristau@debian.org> 115570 Reviewed-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 115571 Signed-off-by: Keith Packard <keithp@keithp.com> 115572 115573commit cf72b5437d2d620521279077a29c5df6d0fbb576 115574Author: Peter Hutterer <peter.hutterer@who-t.net> 115575Date: Thu Feb 25 12:49:21 2010 +1000 115576 115577 dix: Use DeliverGrabbedEvent for implicit passive grabs (#25400) 115578 115579 A client requesting a GrabModeSync button grab, owner-events true, with only 115580 the ButtonRelease mask set would never receive the press event even if the 115581 grab window had the ButtonPress mask set. 115582 115583 The protocol requires that if owner-events is true, then the delivery mask 115584 is the combination of the grab mask + the window event mask. 115585 115586 DeliverGrabbedEvents does this already for us, checking first the delivery 115587 based on owner_events and then based on the grab mask. AFAICT, the device 115588 cannot enter the states FREEZE_BOTH_NEXT_EVENT or FREEZE_NEXT_EVENT that 115589 would be handled by DGE in any possible path here. 115590 115591 Bonus point - CheckPassiveGrabsOnWindows suddenly becomes a lot lesss 115592 complicated. 115593 115594 X.Org Bug 25400 <http://bugs.freedesktop.org/show_bug.cgi?id=25400> 115595 115596 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115597 Reviewed-by: Keith Packard <keithp@keithp.com> 115598 Signed-off-by: Keith Packard <keithp@keithp.com> 115599 115600commit de86a3a3448f0a55c1cd99aee9ea80070a589877 115601Author: Keith Packard <keithp@keithp.com> 115602Date: Thu Feb 25 11:37:05 2010 -0800 115603 115604 Allow for missing or disabled compat_output 115605 115606 When the compat output is missing (I don't think this is actually 115607 possible), or is disabled (and hence has no crtc), we would like to 115608 avoid dereferencing NULL pointers. This patch creates inline functions 115609 to extract the current compat output, crtc or associated RandR crtc 115610 structure, carefully checking for NULL pointers everywhere. 115611 115612 Reviewed-by: Adam Jackson <ajax@redhat.com> 115613 Signed-off-by: Keith Packard <keithp@keithp.com> 115614 115615commit fbbadca7e88391e81ab0f470290f5eec36aa9ce7 115616Author: Keith Packard <keithp@keithp.com> 115617Date: Thu Feb 25 11:35:03 2010 -0800 115618 115619 Share enum definition for det_monrec_parameter sync_source 115620 115621 There were two separate enum definitions, one inside 115622 det_monrec_parameter struct and one for a local variable (which was then 115623 stored inside the struct). Sharing a single definition makes the 115624 code more obviously correct while making the compiler happier. 115625 115626 Reviewed-by: Adam Jackson <ajax@redhat.com> 115627 Signed-off-by: Keith Packard <keithp@keithp.com> 115628 115629commit 4b55b2cf8a52c39b53bae11cd1bc7314481d4c86 115630Author: Robert Bragg <bob@o-hand.com> 115631Date: Wed Jan 27 20:38:38 2010 +0000 115632 115633 DRI2: initialize event->drawable in DRI2SwapEvent 115634 115635 We weren't initialising the drawable in the event structure so the 115636 client side DRI2WireToEvent used for translating the event into a GLX 115637 event wouldn't be able to lookup up the corresponding GLXDrawable before 115638 passing the event on. 115639 115640 Signed-off-by: Robert Bragg <robert@linux.intel.com> 115641 Reviewed-by: Eric Anholt <eric@anholt.net> 115642 Signed-off-by: Keith Packard <keithp@keithp.com> 115643 115644commit 780c95caf9888fa4548dfe4c1c78a7e7ce99a9ed 115645Merge: 018b17759 758f69717 115646Author: Keith Packard <keithp@keithp.com> 115647Date: Wed Feb 24 09:59:19 2010 -0800 115648 115649 Merge remote branch 'whot/for-keith' 115650 115651commit 018b177591c9fade6d065e31858cc6e054d33eff 115652Author: Kristian Høgsberg <krh@bitplanet.net> 115653Date: Wed Feb 24 10:37:26 2010 -0500 115654 115655 glx: Compile fix to let server compile with new and old mesa 115656 115657 We broke the __DRI2_FLUSH API since it was never released, but since it's 115658 taking a little longer than expected to get the X server side of the changes 115659 ready, fix things up so it compiles. 115660 115661 Signed-off-by: Kristian Høgsberg <krh@bitplanet.net> 115662 Reviewed-by: Keith Packard <keithp@keithp.com> 115663 Signed-off-by: Keith Packard <keithp@keithp.com> 115664 115665commit 758f6971750ed507e64eee817d720a77181439f2 115666Author: Peter Hutterer <peter.hutterer@who-t.net> 115667Date: Mon Feb 15 16:32:16 2010 +1000 115668 115669 dix: try to ring the bell even if the current device doesn't have one. (#24503) 115670 115671 Evdev devices do not have the bell proc set, but XTEST devices do. By 115672 exiting early, the bell only rings if the last keyboard used was the XTEST 115673 keyboard and hence the bell proc is still set on the master but not if an 115674 evdev keyboard was used last. 115675 115676 The better approach here is to try to ring the bell on all devices attached 115677 to this master device in case one or more actually do produce an audible 115678 sound. That's also XKB's behaviour if XkbUseCoreKbd is specified as device 115679 identifier. 115680 115681 X.Org Bug 24503 <http://bugs.freedesktop.org/show_bug.cgi?id=24503> 115682 115683 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115684 Reviewed-by: Adam Jackson <ajax@redhat.com> 115685 115686commit c0d941913199ed251fceab62fcf65437f6f9bd31 115687Author: Peter Hutterer <peter.hutterer@who-t.net> 115688Date: Thu Feb 18 16:00:15 2010 +1000 115689 115690 dix: remove now-erroneous comment about frozen slave devices. 115691 115692 A direct grab on a slave device through XI2 detaches it, regardless of 115693 whether the grab is sync or async. So this comment doesn't apply to XI2 115694 anyway. 115695 115696 For XI1, aside from your life being miserable already, it doesn't matter as 115697 XI1 does not have a concept of attachment. You can freeze a device and if 115698 you don't freeze _all_ other devices at the same time, the master device can 115699 still happily send events to the client. 115700 115701 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115702 Acked-by: Adam Jackson <ajax@redhat.com> 115703 115704commit db4f676f25c6d8e58263d5151942be730592d444 115705Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 115706Date: Wed Feb 3 15:40:47 2010 -0500 115707 115708 xselinux: Bump extension minor version. 115709 115710 Changes introduced in this version: 115711 - 3 window-related requests now handle pixmaps also. 115712 115713 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115714 Reviewed-by: Keith Packard <keithp@keithp.com> 115715 115716commit a6cbcf6e782e65f3d4d14202ed0b7a8f65b33c9d 115717Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 115718Date: Wed Feb 3 15:38:57 2010 -0500 115719 115720 xselinux: Rename window-related requests that now support pixmaps. 115721 115722 Renamed requests: 115723 SetWindowCreateContext -> SetDrawableCreateContext 115724 GetWindowCreateContext -> GetDrawableCreateContext 115725 GetWindowContext -> GetDrawableContext 115726 115727 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115728 Reviewed-by: Keith Packard <keithp@keithp.com> 115729 115730commit ab68c707fd02613d9359120ba6402639dd8bd89e 115731Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 115732Date: Wed Feb 3 15:08:40 2010 -0500 115733 115734 xselinux: Allow GetWindowContext to be used for pixmaps as well. 115735 115736 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115737 Reviewed-by: Keith Packard <keithp@keithp.com> 115738 115739commit eb9210097efea81065c301e5b6b4da7a566deb4a 115740Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 115741Date: Wed Jan 6 12:52:51 2010 -0500 115742 115743 xselinux: Remove reference counting calls for SID objects. 115744 115745 Starting with libselinux 2.0.86, SID objects are no longer 115746 reference counted and the sidput() and sidget() calls are no-ops. 115747 115748 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115749 Reviewed-by: Keith Packard <keithp@keithp.com> 115750 115751commit 0c4b75a08922416050bb73d2e279ce938d3823e7 115752Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 115753Date: Wed Oct 21 21:59:01 2009 -0400 115754 115755 Don't print a failure message when XACE denies an input event delivery. 115756 115757 A denial is normal and the behavior should be to drop the event. 115758 Having the log message creates excessive log spam. 115759 115760 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115761 Reviewed-by: Keith Packard <keithp@keithp.com> 115762 115763commit e687e11b1248494a45b9a4e4c321675526808703 115764Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 115765Date: Wed Oct 21 21:46:50 2009 -0400 115766 115767 Revert "Remove some debug messages that trigger on XACE event delivery failure." 115768 115769 The log messages still need to be there for non-XACE failures. 115770 115771 This reverts commit 4be354c4c2da5168b302601b91bd80cfaca7e193. 115772 115773 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115774 Reviewed-by: Keith Packard <keithp@keithp.com> 115775 115776commit 2ecb9f7d86117f212d7845741a3e155553954a85 115777Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 115778Date: Wed Oct 21 19:28:26 2009 -0400 115779 115780 libselinux now has a pkgconfig file. Use it. 115781 115782 Also remove HAVE_NETLINK_AVC_ACQUIRE_FD tests, because we now 115783 require a version of libselinux that has it. 115784 115785 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115786 Reviewed-by: Keith Packard <keithp@keithp.com> 115787 115788commit 9fc6b59d46382e949526dc10c3f5bfddbe2acb92 115789Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 115790Date: Thu Oct 15 17:32:21 2009 -0400 115791 115792 xselinux: Allow SetWindowCreateContext to be used for pixmaps as well. 115793 115794 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 115795 Reviewed-by: Keith Packard <keithp@keithp.com> 115796 115797commit 93ff1bc5884cdfca1132af7fbc06c48e06f218c8 115798Author: Simon Farnsworth <simon.farnsworth@onelan.co.uk> 115799Date: Mon Feb 15 14:18:00 2010 +0200 115800 115801 Always enable outputs that have been forced on in the configuration file 115802 115803 If the user has gone to the effort of manually enabling an output in 115804 the configuration file assume that they know what they're doing. 115805 115806 X.org Bug 14611 <http://bugs.freedesktop.org/show_bug.cgi?id=14611> 115807 115808 Signed-off-by: Simon Farnsworth <simon.farnsworth@onelan.co.uk> 115809 Reviewed-by: Alex Deucher <alexdeucher@gmail.com> 115810 Signed-off-by: Keith Packard <keithp@keithp.com> 115811 115812commit ca9c2472d74c7107ccc117e2c9c723c168d2eccf 115813Merge: e40ba5798 02d1116e7 115814Author: Keith Packard <keithp@keithp.com> 115815Date: Mon Feb 22 13:26:06 2010 -0800 115816 115817 Merge remote branch 'whot/for-keith' 115818 115819commit e40ba5798c367bec584f6437b23d2c5f801fd013 115820Author: Peter Hutterer <peter.hutterer@who-t.net> 115821Date: Thu Feb 18 15:49:09 2010 +1000 115822 115823 Xi: assume BadMode for non-BadMatch errors returned from SetDeviceMode. 115824 115825 The XI protocol spec only allows for two errors on the SetDeviceMode 115826 requests: BadMatch or BadMode. BadMode however is a dynamically assigned 115827 extension error and the driver doesn't have access to the actual error 115828 number. Hence, if a SetDeviceMode driver returns an error other than 115829 BadMatch, assume BadMode. 115830 115831 The two exceptions are BadAlloc and BadImplementations, pass these on to the 115832 client (any request is allowed to return either of those). 115833 115834 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115835 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 115836 Signed-off-by: Keith Packard <keithp@keithp.com> 115837 115838commit 02d1116e7eb84068fa32477f1640d29b0a81b638 115839Author: Peter Hutterer <peter.hutterer@who-t.net> 115840Date: Mon Jan 11 16:35:36 2010 +1000 115841 115842 Xi: assume BadMode for non-BadMatch errors returned from SetDeviceMode. 115843 115844 The XI protocol spec only allows for two errors on the SetDeviceMode 115845 requests: BadMatch or BadMode. BadMode however is a dynamically assigned 115846 extension error and the driver doesn't have access to the actual error 115847 number. Hence, if a SetDeviceMode driver returns an error other than 115848 BadMatch, assume BadMode. 115849 115850 The two exceptions are BadAlloc and BadImplementations, pass these on to the 115851 client (any request is allowed to return either of those). 115852 115853 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115854 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 115855 115856commit 122fc0e7a0712df4ec2bd9ca6773f90bcd2095cf 115857Author: Chris Dekter <cdekter@gmail.com> 115858Date: Thu Feb 11 16:34:09 2010 +1000 115859 115860 Re-enable RECORD extension. 115861 115862 RECORD was disabled during the switch to internal events. This patch 115863 modifies the record callback to work with internal events instead of 115864 xEvents. The InternalEvents are converted to core/Xi events as needed. 115865 115866 Since record is a loadable extension, the EventTo* calls must be externed. 115867 115868 Signed-off-by: Chris Dekter <cdekter@gmail.com> 115869 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 115870 Reviewed-by: Adam Jackson <ajax@redhat.com> 115871 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 115872 115873commit 03ef00ae79687046dc9f9987857062a16ff2c2e0 115874Author: Michel Dänzer <daenzer@vmware.com> 115875Date: Wed Feb 17 11:40:44 2010 +0100 115876 115877 EXA/mixed: Clean up exaPrepareAccessReg_mixed() a little. 115878 115879 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 115880 Acked-by: Maarten Maathuis <madman2003@gmail.com> 115881 Signed-off-by: Keith Packard <keithp@keithp.com> 115882 115883commit 9623daad7c864d46b0dfcbb6e689820b96cc8115 115884Author: Maarten Maathuis <madman2003@gmail.com> 115885Date: Wed Feb 17 11:40:43 2010 +0100 115886 115887 exa/mixed: fix gnome-panel corruption 115888 115889 - A mapped pixmap can't be used for acceleration, any decent memory manager 115890 will refuse this. 115891 - Source pixmaps migrated with a bounding region are incomplete (from the 115892 gpu point of view), so do the upload unconditionally, instead of just for 115893 deferred destination pixmaps. 115894 - Fixes fd.o bug #26076. 115895 115896 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 115897 Acked-by: Michel Dänzer <michel@daenzer.net> 115898 Signed-off-by: Keith Packard <keithp@keithp.com> 115899 115900commit 5b9a52be7e975e59e0bbc6b43539ecaff96b2ecd 115901Author: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 115902Date: Thu Jan 28 11:08:34 2010 +0200 115903 115904 os: Prevent core dump from being truncated. 115905 115906 The problem fixed by this patch can be reproduced on Linux with the 115907 following steps. 115908 - Access NULL pointer intentionally in ProcessOtherEvent on key press. 115909 - Instead of saving core dump to a file, write it into a pipe. 115910 echo "|/usr/sbin/my-core-dumper" > /proc/sys/kernel/core_pattern 115911 - Dump the core by pressing a key. 115912 115913 While the core is being dumped into the pipe, the smart schedule timer 115914 will cause a pending SIGALRM. Linux kernel stops writing data to the 115915 pipe when there are pending signals. This causes the core dump to be 115916 truncated. On my system I'm expecting a 6 MB dump but the size will be 115917 60 kB instead. The problem is solved if we block the SIGALRM caused by 115918 expired smart schedule timer. 115919 115920 I haven't been able to reproduce this problem in the following cases. 115921 - Save core dump to a file instead of a pipe. 115922 - kill -SEGV `pidof Xorg` 115923 - Press a key to dump core while gdb is attached to Xorg. 115924 - Give option -dumbSched to Xorg. 115925 115926 Also note that the fix works only when NoTrapSignals has the default 115927 value FALSE. The problem can still be reproduced if error signals 115928 aren't trapped. In addition to pending SIGALRM, there is a similar 115929 problem with pending SIGIO from the keyboard driver during core dump. 115930 115931 Signed-off-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 115932 Reviewed-by: Keith Packard <keithp@keithp.com> 115933 Signed-off-by: Keith Packard <keithp@keithp.com> 115934 115935commit ca364ca82a760d8e5347a6f9f79636c9a5e4e03f 115936Author: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 115937Date: Thu Jan 28 11:08:33 2010 +0200 115938 115939 os: Introduce OsAbort for proper core dumps. 115940 115941 Signed-off-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 115942 Reviewed-by: Keith Packard <keithp@keithp.com> 115943 Signed-off-by: Keith Packard <keithp@keithp.com> 115944 115945commit 031f92bf9ab15226df410012a0d1c9c390efc36d 115946Author: Oliver McFadden <oliver.mcfadden@nokia.com> 115947Date: Wed Feb 3 10:05:20 2010 +0200 115948 115949 parser: corrected xf86getBoolValue to use case insensitive compare 115950 115951 commit c6e8637e29e0ca11dfb35c02da7ca6002ac8c597 introduced this 115952 regression; it can cause existing config files to be parsed incorrectly. 115953 115954 Acked-by: Julien Cristau <jcristau@debian.org> 115955 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 115956 Signed-off-by: Oliver McFadden <oliver.mcfadden@nokia.com> 115957 Signed-off-by: Keith Packard <keithp@keithp.com> 115958 115959commit be96fb2f02c13a6ee8aba40f7d4c3f9141f06cea 115960Author: Alan Coopersmith <alan.coopersmith@sun.com> 115961Date: Mon Feb 15 17:42:11 2010 -0800 115962 115963 Solaris xf86OSRingBell() off-by-one error in filling iov[] array 115964 115965 When generating sound buffers for /dev/audio bells, insert waveform 115966 for beep *or* silence, but not both, so we don't write one entry past 115967 the end of the iov buffer when the final bit of soundwave ends up in 115968 the final entry allocated in the iov array. 115969 115970 Fixes OpenSolaris bug 6894890: 115971 http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6894890 115972 115973 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 115974 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 115975 Acked-by: Adam Jackson <ajax@redhat.com> 115976 Signed-off-by: Keith Packard <keithp@keithp.com> 115977 115978commit 2d40f22d1ec970a84b23aa42b1feca4feedeb4bb 115979Author: Matt Turner <mattst88@gmail.com> 115980Date: Mon Feb 15 20:05:24 2010 -0500 115981 115982 Use C-style comments in x86emu 115983 115984 Signed-off-by: Matt Turner <mattst88@gmail.com> 115985 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 115986 Signed-off-by: Keith Packard <keithp@keithp.com> 115987 115988commit 4e8bf12b13690afa6d9fee0e339d3819ef16fb3f 115989Author: Jeremy Huddleston <jeremyhu@apple.com> 115990Date: Mon Feb 15 16:46:21 2010 -0800 115991 115992 XQuartz: Fix a possible buffer overrun in quartzAudio 115993 115994 Also dropped deprecated API while there 115995 115996 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 115997 115998commit 4127e8e5be5c366eb752bbb8d7f0ff5b519e641b 115999Author: Jeremy Huddleston <jeremyhu@apple.com> 116000Date: Wed Feb 10 17:56:30 2010 -0800 116001 116002 XQuartz: clang static analysis fixes 116003 116004 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 116005 116006commit c686c5cbdc5f35744407af5f9aabbd18cc17714b 116007Author: Jeremy Huddleston <jeremyhu@apple.com> 116008Date: Mon Feb 15 15:03:01 2010 -0800 116009 116010 XQuartz: Fix linking (CloseInput()) 116011 116012 Fixes regresison from d33adcdf03c69407d151e732fa0cf9947151eb19 116013 116014 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 116015 116016commit 84905007702da2c05a4f7446b3fc5ff52be49655 116017Author: Thomas Jaeger <ThJaeger@gmail.com> 116018Date: Mon Jan 4 15:00:49 2010 -0500 116019 116020 udev: Don't filter subsystem "input" 116021 116022 This allows serial wacom devices to work, whose subsystem is "tty". 116023 116024 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 116025 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116026 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116027 116028commit 8736d112afb0dd61dfdaadd6378eafd200b2ef5f 116029Author: Dan Nicholson <dbn.lists@gmail.com> 116030Date: Sat Feb 13 10:40:40 2010 -0800 116031 116032 xfree86: Reorder InputClass option priorities 116033 116034 Currently the config and InputClasses are merged together so that the 116035 options from the config backend have the highest priority. This is bad 116036 since it means options such as a default XKB layout set by the backend 116037 cannot be changed by the user. 116038 116039 This patch changes order of precedence to be: 116040 116041 1. xorg.conf 116042 2. xorg.conf.d (later files have higher priority) 116043 3. config backend 116044 116045 In order to allow this ordering, the config parsing has been changed to 116046 read the xorg.conf.d files before xorg.conf. This has the consequence 116047 that the core device picking which looks for the first InputDevice may 116048 not find it in xorg.conf. 116049 116050 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 116051 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116052 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116053 116054commit 48f7298657f91843db36566b8d66d6c4c18dbd4c 116055Author: Peter Hutterer <peter.hutterer@who-t.net> 116056Date: Wed Jan 13 15:03:45 2010 +1000 116057 116058 Xi: reset the sli pointers after copying device classes. (#25640) 116059 116060 If the indicator flags have the XkbSLI_IsDefault bit set, the indicator map 116061 and names aren't their own bit of memory but rather point into the 116062 device->key->xkbInfo->desc structure. XkbCopySrvLedInfo knows about this and 116063 leaves the pointers alone. 116064 116065 When copying the classes from the slave to the master, these pointers are 116066 copied and still point to the dev->key class of the slave device. If the 116067 slave device is removed, the memory becomes invalid and a call to modify 116068 this data (e.g. XkbSetIndicators) may cause a deadlock. 116069 116070 The copying of dev->key relies on dev->kbdfeed to be already set up. Hence 116071 the pointers need to be reset once _both_ kbdfeed and key have been copied 116072 into the master device. 116073 116074 X.Org Bug 25640 <http://bugs.freedesktop.org/show_bug.cgi?id=25640> 116075 Fedora Bug 540584 <https://bugzilla.redhat.com/show_bug.cgi?id=540584> 116076 116077 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116078 Reviewed-by: Keith Packard <keithp@keithp.com> 116079 116080commit d33adcdf03c69407d151e732fa0cf9947151eb19 116081Author: Peter Hutterer <peter.hutterer@who-t.net> 116082Date: Fri Jan 15 15:22:02 2010 +1000 116083 116084 dix: move config_init into the DDX. 116085 116086 The only DDX currently using hotplugging is the xfree86 one and it looks 116087 like it'll stay that way for a bit. Move the initialization to the DDX, 116088 since Xephyr, Xnest, and friends don't need HAL or udev notifications. 116089 116090 Add CloseInput (counterpart to InitInput) to be able to clean up the config 116091 initialization from the DDX as well. 116092 116093 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116094 Reviewed-by: Julien Cristau <jcristau@debian.org> 116095 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 116096 116097commit 001ce71dc11287dc94cc2fbc5d35677c046e6c04 116098Author: Julien Cristau <jcristau@debian.org> 116099Date: Fri Feb 12 23:34:57 2010 +0100 116100 116101 dix: restore lastDeviceEventTime update in dixSaveScreens 116102 116103 This was removed in 6b5978dcf1f7ac3ecc2f22df06f7000f360e2066 (Do not 116104 reset lastDeviceEventTime when we do dixSaveScreens), but caused a 116105 regression for XResetScreenSaver. Add the lastDeviceEventTime update 116106 back, but restrict it to that case. 116107 116108 X.Org bug#25855 <http://bugs.freedesktop.org/25855> 116109 116110 Reported-by: Lubos Lunak <l.lunak@suse.cz> 116111 Tested-by: Lubos Lunak <l.lunak@suse.cz> 116112 Signed-off-by: Julien Cristau <jcristau@debian.org> 116113 Signed-off-by: Keith Packard <keithp@keithp.com> 116114 116115commit 97b03037f4d99fcebc7603011f41c3aff9871ce2 116116Author: Peter Harris <pharris@opentext.com> 116117Date: Fri Feb 12 15:36:30 2010 -0500 116118 116119 Don't double-swap the RandR PropertyNotify event 116120 116121 The event is already swapped in randr.c/SRROutputPropertyNotifyEvent, so 116122 it should not be swapped here. 116123 116124 X.Org Bugzilla #26511: http://bugs.freedesktop.org/show_bug.cgi?id=26511 116125 116126 Tested-by: Leonardo Chiquitto <leonardo@ngdn.org> 116127 Acked-by: Adam Jackson <ajax at redhat.com> 116128 Reviewed-by: Julien Cristau <jcristau at debian.org> 116129 Signed-off-by: Peter Harris <pharris@opentext.com> 116130 Signed-off-by: Keith Packard <keithp@keithp.com> 116131 116132commit 0b73f98cbd1803b82c2faf1a40f1a0037673d712 116133Author: Soeren Sandmann <sandmann@daimi.au.dk> 116134Date: Fri Feb 12 20:04:53 2010 +0100 116135 116136 xfree86: Add qxl driver to the autoconfig logic 116137 116138 The qxl driver is for the QXL virtualized graphics device. 116139 116140 Signed-off-by: Søren Sandmann Pedersen <ssp@redhat.com> 116141 Reviewed-by: Adam Jackson <ajax@redhat.com> 116142 Signed-off-by: Keith Packard <keithp@keithp.com> 116143 116144commit 9c5bb550a2234e4bee2a144417d74c6bdf160e89 116145Author: Keith Packard <keithp@keithp.com> 116146Date: Fri Feb 12 13:46:03 2010 -0800 116147 116148 Bump to 1.7.99.901 -- 1.8 RC1 116149 116150commit 57ffeb3f2b3313dcef92a396f1f55fdbc064b2c5 116151Merge: c6d9bc092 c76248fda 116152Author: Keith Packard <keithp@keithp.com> 116153Date: Fri Feb 12 13:36:32 2010 -0800 116154 116155 Merge remote branch 'jturney/master' 116156 116157commit c6d9bc092c84ad5c68083a126aa7577baa42cef7 116158Author: Peter Hutterer <peter.hutterer@who-t.net> 116159Date: Wed Feb 10 15:36:50 2010 +1000 116160 116161 Add tag matching to input attributes. 116162 116163 Tags may be a list of comma-separated strings that match against a MatchTag 116164 InputClass section. If any of the tags specified for a device match against 116165 the MatchTag of the section, this match is evaluated true and passed on to 116166 the next match condition. 116167 116168 Tags are specified as "input.tags" (hal) or "ID_INPUT.tags" (udev), the 116169 value of the tags is case-sensitive and require an exact match (not a 116170 substring match). 116171 116172 i.e. "quirk" will not match "QUIRK", "need_quirk" or "quirk_needed". 116173 116174 Example configuration: 116175 udev: 116176 ENV{ID_INPUT.tags}="foo,bar" 116177 116178 hal: 116179 <merge key="input.tags" type="string">foo,bar</merge> 116180 116181 xorg.conf: 116182 Section "InputClass" 116183 Identifier "foobar quirks" 116184 MatchTag "foo|foobar" 116185 Option "Foobar" "on" 116186 EndSection 116187 116188 Where the xorg.conf section matches against any device with the tag "foo" 116189 or tag "foobar" set. 116190 116191 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116192 Tested-by: Dan Nicholson <dbn.lists@gmail.com> 116193 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 116194 116195commit 3ac43df5d4a25d6e0058b327fa05a1c1436b4794 116196Author: Peter Hutterer <peter.hutterer@who-t.net> 116197Date: Wed Feb 10 15:36:49 2010 +1000 116198 116199 Add xstrtokenize to the dix. 116200 116201 Move tokenize out of the parser, make it a dix util function instead. 116202 Splitting a string into multiple substrings is useful by other places, so 116203 let's use it across the line. Future users include config/hal, config/udev 116204 and of course the parser. 116205 116206 Example usage: 116207 char **substrings = xstrtokenize(my_string, "\n"); 116208 116209 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116210 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 116211 116212commit 27d1b86d1b858f931b4cb1b6ddf857c76d92a6d9 116213Author: Dan Nicholson <dbn.lists@gmail.com> 116214Date: Wed Feb 10 15:36:48 2010 +1000 116215 116216 xfree86: Set fnmatch pathname flag for InputClass device matching 116217 116218 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 116219 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116220 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116221 116222commit 9b369f71273fb117c982e6ce16cd4462f206d365 116223Author: Dan Nicholson <dbn.lists@gmail.com> 116224Date: Wed Feb 10 15:36:47 2010 +1000 116225 116226 xfree86: Allow multiple arguments to InputClass matches 116227 116228 In order to keep the number of InputClass sections manageable, allow 116229 matches to contain multiple arguments. The arguments will be separated 116230 by the '|' character. This allows a policy to apply to multiple types of 116231 devices. For example: 116232 116233 Section "InputClass" 116234 Identifier "Inverted Mice" 116235 MatchProduct "Crazy Mouse|Silly Mouse" 116236 Option "InvertX" "yes" 116237 EndSection 116238 116239 This applies to the MatchProduct, MatchVendor and MatchDevicePath 116240 entries. Currently there is no way to escape characters, so names or 116241 patterns cannot contain '|'. 116242 116243 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 116244 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116245 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116246 116247commit a378e361a5de89f0be8b68ebc3e854f56cefe666 116248Author: Dan Nicholson <dbn.lists@gmail.com> 116249Date: Wed Feb 10 15:36:46 2010 +1000 116250 116251 xfree86: Use "Ignore" option in InputClass to skip devices 116252 116253 Sometimes it is desirable to skip adding specific input devices to the 116254 server. The "Ignore" option is used similarly to Monitor sections so 116255 that matched devices will not be added. BadIDChoice is returned to the 116256 config backend so that it will clean up all resources. 116257 116258 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 116259 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116260 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116261 116262commit 8b1a685f00ae76be864cc188943a0874f48b8d64 116263Author: Dan Nicholson <dbn.lists@gmail.com> 116264Date: Wed Feb 10 15:36:45 2010 +1000 116265 116266 xfree86: Handle config files ending without newline 116267 116268 The config parser expects to find a newline at the end of each line, so 116269 files ending without one would confuse it. A newline is inserted at the 116270 end of the buffer in these situations. Additionally, switching to the 116271 next config file is moved to the higher level to allow parsing of the 116272 last line of the previous file to complete before shifting the index and 116273 resetting the line number. 116274 116275 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 116276 Tested-by: Stephan Raue<stephan.raue@gmx.net> 116277 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116278 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116279 116280commit 67bc278a511ca6ec42e1f8d2d5897c0109e94e2c 116281Author: Dan Nicholson <dbn.lists@gmail.com> 116282Date: Wed Feb 10 15:36:44 2010 +1000 116283 116284 xfree86: Make InputClass docs and comments match reality 116285 116286 Drivers and options specified in InputClass sections work on a "first 116287 match wins" strategy. Let's be consistent when documenting it. 116288 116289 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 116290 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116291 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116292 116293commit a34812b09000db2ff2a1dc6182602839123edd4e 116294Author: Benjamin Tissoires <tissoire@cena.fr> 116295Date: Mon Feb 8 13:50:43 2010 +0100 116296 116297 Add labels for multitouch valuators 116298 116299 Thoses definitions have been included in the kernel but the X server is not updated accordingly. 116300 Without these definitions, the multitouch axes are not correctly labelled. 116301 116302 Signed-off-by: Benjamin Tissoires <tissoire@cena.fr> 116303 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116304 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116305 116306commit c76248fda99c38aef0ccf0ed6b58fbe95f6fe497 116307Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116308Date: Tue Feb 2 20:53:44 2010 +0000 116309 116310 Cygwin/X: Avoid a collision between DEBUG and a token name 116311 116312 Rename a token to avoid a collision between DEBUG defined via AC_DEFINE 116313 if --enable-debug is configured, and the token for the 'debug' instruction 116314 in the XWin preferences file 116315 116316 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116317 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116318 116319commit b7dbbd3cd4d28cd4939706b3bf6394ba8ecafaff 116320Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116321Date: Sun Nov 22 23:35:09 2009 +0000 116322 116323 Cygwin/X: Avoid cursor size log spam 116324 116325 Fedora 12 likes to use a 39x26 animated wait cursor. Avoid spamming 116326 the log with warnings that each frame can't be completely contained 116327 in the 32x32 native cursor 116328 116329 Also reformat log message so it doesn't contain a '\n\t'. I mean, 116330 it's not like we might want to grep the log or something... 116331 116332 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116333 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116334 116335commit c9cbbd5d1cfa58a2d9f08e25534ea8439284322d 116336Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116337Date: Sat Nov 28 17:51:33 2009 +0000 116338 116339 Repair '-nolock' 116340 116341 commit 446fe9eecddd1337f9d5164dd7c301e1ba3dfe32 removes the AC_DEFINE for 116342 SERVER_LOCK and conditional compilation checking it, making it always on 116343 everywhere, except in os/utils.c where code is left under SERVER_LOCK, which 116344 now never gets built, making the '-nolock' option non-functional... 116345 116346 This seems to have been broken since Xserver 1.7.0, but this option is 116347 actually of some slight use on cygwin, as if /tmp resides on a FAT filesystem 116348 (yes, I know...), hard links aren't supported. 116349 116350 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116351 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116352 116353commit 7f54ccafadf99c0a1a3e788734199b306b4fa51d 116354Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116355Date: Fri Jan 16 13:12:12 2009 +0000 116356 116357 Cygwin/X: Make transient windows resizable again 116358 116359 Reverts the change which makes parented windows non-resizeable 116360 116361 Because this was trying to do something which we should be doing, as an alternative we 116362 try checking WM_NORMAL_HINTS for windows which shouldn't be resizable 116363 116364 If a window has a maximum size specified, no maximize box 116365 If a window has a fixed size (max size == min size, per EWMH 1.3 Implementation Notes), no resizing frame 116366 116367 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116368 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116369 116370commit 654d2e372dc2978ce379ab9f02137333ec224f0c 116371Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116372Date: Tue Nov 10 13:24:40 2009 +0000 116373 116374 Cygwin/X: Discourage other WMs in multiwindow mode 116375 116376 Tidying up of other WM detection code when ading SWT/Motif 116377 fix in commit 71519a572fe15b85c0eb2b02636c9e871f2c858f 116378 was rather over-agressive and now allows other WMs to think 116379 they can start when the internal WM is running. 116380 116381 Revert to the behaviour of selecting ButtonPressMask events 116382 as well on the root window, so other WMs will be dissuaded 116383 from starting 116384 116385 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116386 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116387 116388commit 26857b1c2003797b02e258247f63064aa1e37c10 116389Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116390Date: Sat Oct 31 17:19:17 2009 +0000 116391 116392 Cygwin/X: Update XWin man page 116393 116394 Restructure to group similar options and offer some commentary on those groups 116395 Review option descriptions, clarify and bring up-to-date 116396 116397 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116398 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116399 116400commit 6f1836bfbd80e88d4c57a32757d0a5b398504c35 116401Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116402Date: Mon Nov 2 17:37:14 2009 +0000 116403 116404 Cygwin/X: Update DDX specific -help text 116405 116406 Alphabetize options 116407 Review option descriptions, clarify and bring up to date 116408 Add missing option descriptions 116409 116410 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116411 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116412 116413commit 22982b9e95a2339d5ba60d66263e42a0331ee41f 116414Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116415Date: Thu Oct 29 18:50:23 2009 +0000 116416 116417 Cygwin/X: XSupportsLocale() failure is non-critical 116418 116419 Treat XSupportsLocale() returning false as non-critical to internal client 116420 theads startup, and issue a warning, not an error 116421 116422 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116423 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116424 116425commit f60b7712b3451649f138b158ee282be89a66b9ef 116426Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 116427Date: Mon Feb 1 17:13:00 2010 +0000 116428 116429 Cygwin/X: Report BUILDERSTRING with version information 116430 116431 Report BUILDERSTRING with version information 116432 116433 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116434 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116435 116436commit ed2c2a68e1e6b409c198a52d99d300eb6d517e89 116437Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 116438Date: Mon Feb 1 17:12:57 2010 +0000 116439 116440 Define FD_SETSIZE on Cygwin regardless of XWin DDX 116441 116442 All DDXs segfault on Cygwin unless -DFD_SETSIZE=256 is set, so make sure 116443 we do so whether or not we are building XWin. 116444 116445 FD_SETSIZE must be at least XFD_SETSIZE for uses of select() to be correct. 116446 The Cygwin default is only 64, so it must be increased to 256 116447 116448 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 116449 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116450 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116451 116452commit 2f2f3da080629d410dd99e281c382b54f0dbbf5d 116453Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116454Date: Fri Jun 19 21:14:47 2009 +0100 116455 116456 Cygwin/X: Copy the state of the Windows keyboard device to the Virtual Core Keyboard at startup. 116457 116458 Otherwise, this happens lazily after the first keypress, which can lead 116459 to applications which are started from a shell window and inspect the 116460 keyboard state before a character is typed getting the wrong idea about 116461 the desired keymap (e.g. xemacs shows this behaviour) 116462 116463 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116464 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116465 116466commit 7a440e5b7a416e582b6c3cc4c33822854ce73aed 116467Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116468Date: Sun Oct 4 15:55:40 2009 +0100 116469 116470 Cygwin/X: Tidy up system.Xwinrc 116471 116472 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116473 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116474 116475commit 11252ed82e1f361b99e86521ac9314f868bd1a3a 116476Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 116477Date: Mon Feb 1 17:11:54 2010 +0000 116478 116479 Cygwin/X: Look for system.Xwinrc in SYSCONFDIR/X11 116480 116481 Look for system.Xwinrc in SYSCONFDIR/X11 (usually /etc/X11) 116482 Rename sample system.Xwinrc file not to have a X11R6 path in it's name 116483 Add makefile install rule for system.XWinrc 116484 116485 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116486 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116487 116488commit 34269a90ea2087f883f5dc8805894fc4998e4b81 116489Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 116490Date: Tue Feb 2 16:21:44 2010 +0000 116491 116492 Cygwin/X: Update resources file and About dialog 116493 116494 Use the configured vendor web address in the About dialog 116495 116496 Update resources file: rework About dialog, use 'MS Shell Dlg 2' 116497 logical font for all dialogs, add ellipsis to exit option in tray 116498 menu as it (may) trigger another dialog. 116499 116500 Get short vendor name from xwin-config.h, like other configuration 116501 data presented in the About dialog box, rather than creating the 116502 PROJECT_NAME define 116503 116504 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116505 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116506 116507commit 881812438b430d2856d0494707e028e5f30d6e9a 116508Author: Colin Harrison <colin.harrison@virgin.net> 116509Date: Tue Feb 2 16:22:39 2010 +0000 116510 116511 Xming: Use 8 point font for URL in About dialog 116512 116513 Use an 8 point font for URL in About dialog, to match the rest 116514 of the text in that dialog 116515 116516 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116517 116518commit c02638fd68440513b6046315547c554a910bd7e2 116519Author: Colin Harrison <colin.harrison@virgin.net> 116520Date: Tue Feb 2 16:08:45 2010 +0000 116521 116522 Xming: Add a manifest file 116523 116524 Use manifests to enable XP style controls 116525 (only effective for XP and later and when themes are enabled). 116526 The addition of manifests shouldn't cause compatibility problems with older Windows versions. 116527 116528 Manifest must have execute permissions, otherwise attempts to 116529 execute XWin.exe in the same directory will fail... 116530 116531 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116532 116533commit fd8a32baba77aba8d124658a19f4f6eda79e49c6 116534Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116535Date: Fri Jun 19 21:14:39 2009 +0100 116536 116537 Cygwin/X: Window placement refinement for multiwindow mode 116538 116539 Window placement refinement for multiwindow mode, ensure a window actually 116540 ends up somewhere visible if it tries to create itself offscreen (which 116541 can happen for e.g. if it has a stored position from a different sized 116542 display) 116543 116544 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116545 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116546 116547commit 7eb840fd42bc62d88fb4fcf6600546c9ff6e56be 116548Author: Jon TURNEY <jon.turney@dronecode.org.uk> 116549Date: Tue Feb 2 15:42:12 2010 +0000 116550 116551 Cygwin/X: Respect the system's ownership of the clipping region 116552 116553 Respect the system's ownership of the clipping region used for shaped windows 116554 116555 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116556 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 116557 116558commit bad41bdfd4a12f0c92fd221bffd9f3c5e40d0fe1 116559Author: Colin Harrison <colin.harrison@virgin.net> 116560Date: Mon Feb 1 17:11:22 2010 +0000 116561 116562 Xming: Always update the Windows title Unicode (wide-character) in -multiwindow mode. 116563 116564 Apply the Windows title wide-character in -multiwindow mode. Windows should now 116565 display correct client X Window titles for locales with wide characters. 116566 116567 Copyright (C) Colin Harrison 2005-2008 116568 http://www.straightrunning.com/XmingNotes/ 116569 http://sourceforge.net/projects/xming/ 116570 116571 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 116572 116573commit 901c4ebf620579c5bb951a8c003430d967ecf2ff 116574Author: Julien Cristau <jcristau@debian.org> 116575Date: Mon Jan 11 00:31:27 2010 +0000 116576 116577 main: move config_init() after InitInput() 116578 116579 With the udev backend, config_init() calls NewInputDeviceRequest(), 116580 which enables devices. They can then start sending events, even though 116581 the event queue is only initialized later in InitInput(). Oops. 116582 116583 Debian bug#564256 <http://bugs.debian.org/564256> 116584 116585 Reported-by: Sedat Dilek <sedat.dilek@googlemail.com> 116586 Signed-off-by: Julien Cristau <jcristau@debian.org> 116587 Tested-by: Sedat Dilek <sedat.dilek@googlemail.com> 116588 Reviewed-by: Adam Jackson <ajax@redhat.com> 116589 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116590 116591commit 7f648777680b9dea31c4adb657e205fc6880d98b 116592Author: Dan Nicholson <dbn.lists@gmail.com> 116593Date: Thu Jan 14 16:42:16 2010 +1000 116594 116595 Don't use AC_CHECK_FILE for fontpath checks when cross compiling 116596 116597 AC_CHECK_FILE chokes when cross compiling, so instead we set the default 116598 to the standard FONTROOTDIR directories in that case. 116599 116600 Signed-off-by: Dan Nicholson<dbn.lists@gmail.com> 116601 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116602 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116603 116604commit db687f718f760ba254ab51994769db101dc9ca3a 116605Author: Peter Hutterer <peter.hutterer@who-t.net> 116606Date: Fri Jan 15 10:07:38 2010 +1000 116607 116608 xkb: sed True -> TRUE and False -> FALSE 116609 116610 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116611 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 116612 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 116613 116614commit 0ea2b0bd02f8683998e8b9ebc2b96d606ce45f45 116615Author: Peter Hutterer <peter.hutterer@who-t.net> 116616Date: Fri Dec 18 20:07:30 2009 +1000 116617 116618 xkb: Add XKM file format description. 116619 116620 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116621 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 116622 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 116623 116624commit 08b22c7faf97217ea4d497eec6624fc3dd916d9b 116625Author: Oldřich Jedlička <oldium.pro@seznam.cz> 116626Date: Sun Jan 17 17:59:03 2010 +0100 116627 116628 Allow driver to call DeleteInputDeviceRequest during UnInit 116629 116630 When the input driver (like xf86-input-wacom) removes it's devices 116631 during a call to UnInit, the CloseDownDevices() cannot handle it. The 116632 "next" variable can become a pointer to freed memory. 116633 116634 The patch introduces order-independent device freeing mechanism by 116635 remembering the already freed device ids. The devices can reorder any 116636 time during freeing. No device will be double-freed - if the removing 116637 failed for any reason; some implementations of DeleteInputDeviceRequest 116638 don't free the devices already. 116639 116640 Signed-off-by: Oldřich Jedlička <oldium.pro@seznam.cz> 116641 Reviewed-by: Simon Thum <simon.thum@gmx.de> 116642 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116643 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116644 116645commit b91cec26de367f75388e620772747b08eee34294 116646Author: Horst Wente <horst.wente@acm.org> 116647Date: Wed Dec 30 19:35:20 2009 +0100 116648 116649 xkb: make ctrl+alt+keypad + / ctrl+alt+keypad - work again (#25743) 116650 116651 Video mode switching via keypad keys did not work 116652 116653 Signed-off-by: Horst Wente <horst.wente@acm.org> 116654 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116655 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116656 116657commit 993e78d6c433f65bd8a87890a6bb6da480b2cdc4 116658Author: Oldřich Jedlička <oldium.pro@seznam.cz> 116659Date: Tue Jan 12 19:15:06 2010 +0100 116660 116661 Fix typo in updateSlaveDeviceCoords 116662 116663 The index [0] for the second valuator looks bogus; fix it. 116664 116665 Signed-off-by: Oldřich Jedlička <oldium.pro@seznam.cz> 116666 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116667 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116668 116669commit 52456c602c3cdd7d5eac677889a18fad37dfb7ae 116670Author: Jeremy Huddleston <jeremyhu@apple.com> 116671Date: Sat Jan 30 14:49:02 2010 -0800 116672 116673 XQuartz: Attatch a stub display when CoreGraphics reports no displays. 116674 116675 This is half of the required changes to address the "stuck mouse pointer" 116676 bug that occurs when X11 launches while the displays are asleep. The 116677 remainder of the fix is part of libXplugin which needs to be updated to 116678 send XP_EVENT_DISPLAY_CHANGED when the display wakes up. If you don't 116679 have a recent enough libXplugin (expected in 2.5.0_beta2 or later), you 116680 can cause this event to be sent by changing your display resolution (or 116681 you could just start X11.app with your screens awake). 116682 116683 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 116684 116685commit 15f4bb6dc64313ff100aa5444a94c60922a498df 116686Author: Jeremy Huddleston <jeremyhu@apple.com> 116687Date: Wed Jan 27 17:12:12 2010 -0800 116688 116689 XQuartz: Dead code removal 116690 116691 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 116692 116693commit 7ba5de6e9e1e8efda9f79af601fb6269bae841ce 116694Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 116695Date: Sat Jan 30 10:55:38 2010 -0800 116696 116697 XQuartz: Add some .gitignore magic 116698 116699 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 116700 116701commit 4d575b0559817258f7a0ce6c4d2d0f9e7e5bba63 116702Author: Robert Morell <rmorell@nvidia.com> 116703Date: Fri Jan 29 19:07:03 2010 -0800 116704 116705 RENDER: Fix gradient and solid fill pictures with Xinerama, and misc cleanup 116706 116707 If these aren't wrapped, then procs that are wrapped (such as 116708 RenderChangePicture) will fail in Xinerama when they see the resource 116709 type of a picture created through one of these interfaces is PictureType 116710 and not XRT_PICTURE like those allocated via RenderCreatePicture. 116711 116712 Signed-off-by: Robert Morell <rmorell@nvidia.com> 116713 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 116714 Signed-off-by: Keith Packard <keithp@keithp.com> 116715 116716commit 8d6b1838330625b758ebc55bb0df1b425c5cefaf 116717Author: Robert Morell <rmorell@nvidia.com> 116718Date: Thu Jan 21 12:38:47 2010 -0800 116719 116720 Render: Fix request size verification 116721 116722 RenderSetPictureClipRectangles and the Xinerama version of 116723 RenderChangePicture were using the wrong structure types for request 116724 size verification. 116725 116726 Signed-off-by: Robert Morell <rmorell@nvidia.com> 116727 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 116728 Signed-off-by: Keith Packard <keithp@keithp.com> 116729 116730commit 5e81078cf56aabbf6551d96d312c7840ba9370ae 116731Author: Tiago Vignatti <tiago.vignatti@nokia.com> 116732Date: Mon Jan 18 14:31:12 2010 +0200 116733 116734 xfree86: vgaarb: remove useless debug 116735 116736 This is RAC's remnant. Any sane person would use a more wise method of 116737 debugging instead. 116738 116739 X.Org Bug 26074 <http://bugs.freedesktop.org/show_bug.cgi?id=26074> 116740 116741 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 116742 Reviewed-by: Keith Packard <keithp@keithp.com> 116743 Signed-off-by: Keith Packard <keithp@keithp.com> 116744 116745commit e38df06772965cd5099646f1e7032b7c944d16ed 116746Author: Aaron Plattner <aplattner@nvidia.com> 116747Date: Thu Jan 28 14:33:25 2010 -0800 116748 116749 DRI2: Bump the dri2 module version to 1.2.0 to reflect recent API changes. 116750 116751 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 116752 Acked-by: Jesse Barnes <jbarnes@virtuousgeek.org> 116753 Signed-off-by: Keith Packard <keithp@keithp.com> 116754 116755commit 60b6477f6dc005a1b052be8c5e5a549550a4aa88 116756Author: Tiago Vignatti <tiago.vignatti@nokia.com> 116757Date: Thu Jan 14 17:18:28 2010 +0200 116758 116759 dix/configure: remove null root cursor option 116760 116761 The default behavior of the server is to start with an invisible root cursor. 116762 Be such cursor invisible or inexistent (null), in the end it doesn't matter - 116763 for the user. The content on screen will be the same. Besides, there's no 116764 difference, in terms of performance, whether such cursor is invisible or 116765 simply null. The paths that both take inside the server are roughly the same. 116766 116767 Therefore create a null root cursor becomes irrelevant. 116768 116769 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 116770 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 116771 Acked-by: Daniel Stone <daniel@fooishbar.org> 116772 Signed-off-by: Keith Packard <keithp@keithp.com> 116773 116774commit b4baab90c0d98bef98d485682d4a69a327a380d6 116775Author: Peter Hutterer <peter.hutterer@who-t.net> 116776Date: Fri Jan 8 15:18:41 2010 +1000 116777 116778 dix: EventToCore needs to copy the root window too. 116779 116780 This value isn't actually set for normal events but it saves us some work 116781 for the record extension support. 116782 116783 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 116784 Reviewed-by: Keith Packard <keithp@keithp.com> 116785 Signed-off-by: Keith Packard <keithp@keithp.com> 116786 116787commit 27fc98ecfed99fefd5c14388b4c6242f87281895 116788Author: Keith Packard <keithp@keithp.com> 116789Date: Wed Jan 27 14:37:32 2010 -0800 116790 116791 Revert "dix: EventToCore needs to copy the root window too." 116792 116793 This reverts commit 0f2d297dedeff8bd227df4c498cc668b0e902344. 116794 116795 Peter provided an updated patch which fixes the tests. 116796 116797commit 711e26466ae04ae93ff4c48d377d83d68a6320e9 116798Author: Jesse Barnes <jbarnes@virtuousgeek.org> 116799Date: Mon Jan 25 09:21:51 2010 -0800 116800 116801 DRI2: handle drawable destruction properly at DRI2SwapComplete time 116802 116803 Simon reported an issue with kwin that turned out to be a general problem. If 116804 a drawable goes away before its swap completes, we'll try to free it up. 116805 However, we free it improperly, which causes a server crash in 116806 DRI2DestroyDrawable. Fix that up by splitting the free code out and calling 116807 it from DRI2SwapComplete. 116808 116809 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 116810 Reviewed-by: Michel Dänzer <michel@daenzer.net> 116811 Signed-off-by: Keith Packard <keithp@keithp.com> 116812 116813commit b68f0204a2e4fa9d8884cbdd84b6a5df21d6b36e 116814Author: Jesse Barnes <jbarnes@virtuousgeek.org> 116815Date: Wed Jan 27 09:10:41 2010 -0800 116816 116817 xserver: require libdri 7.8.0 to build 116818 116819 We depend on new DRI interfaces now, so require them in configure.ac. 116820 116821 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 116822 Signed-off-by: Keith Packard <keithp@keithp.com> 116823 116824commit 003829072853546abd973266fe9b24d803f4f5cb 116825Author: Alan Coopersmith <alan.coopersmith@sun.com> 116826Date: Tue Jan 26 22:25:04 2010 -0800 116827 116828 Avoid segfaults in XF86VidMode GammaRamp functions if randr_crtc is NULL 116829 116830 Fixes crash when xscreensaver tries to use GammaRamp calls to fade out 116831 http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6915712 116832 116833 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 116834 Reviewed-by: Keith Packard <keithp@keithp.com> 116835 Signed-off-by: Keith Packard <keithp@keithp.com> 116836 116837commit 19d03d4f49e08442f58cf02240e3e6bab04633d2 116838Author: Gaetan Nadon <memsize@videotron.ca> 116839Date: Tue Jan 26 20:15:49 2010 -0500 116840 116841 doc: finish the removal of SecurityPolicy file man pages 116842 116843 The variable was unassigned but all the code was left in. 116844 116845 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 116846 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 116847 Signed-off-by: Keith Packard <keithp@keithp.com> 116848 116849commit 7962c8f78964d460c76f76dda2795b971a8c2a94 116850Author: Gaetan Nadon <memsize@videotron.ca> 116851Date: Tue Jan 26 20:02:13 2010 -0500 116852 116853 man: add missing __datadir__ for Xserver XWin man pages 116854 116855 The man pages display __datadir__/fonts/X11/... because __datadir__ 116856 is missing in cpprules.in. Problem was introduced in commit: 116857 b54bc14ce0ae38c4863794bc3096ca86cdb23908 116858 when replacing __projectroot__ in the previous font path. 116859 116860 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 116861 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 116862 Signed-off-by: Keith Packard <keithp@keithp.com> 116863 116864commit d2322b6309bf15a45002b42e7e6ba3d6b5bfa932 116865Author: Kok, Auke <auke-jan.h.kok@intel.com> 116866Date: Wed Jan 27 11:34:45 2010 -0800 116867 116868 xserver: Add timestamps to logfile output. 116869 116870 Add timestamps in seconds derived from clock_monotonic to the log 116871 file. 116872 116873 Signed-off-by: Auke Kok <auke-jan.h.kok@intel.com> 116874 Reviewed-by: Keith Packard <keithp@keithp.com> 116875 Signed-off-by: Keith Packard <keithp@keithp.com> 116876 116877commit 0b21a0416b4cb2c32da5e3fda05a0682eb97d56e 116878Merge: 1e6fd65d0 837bd2bbc 116879Author: Keith Packard <keithp@keithp.com> 116880Date: Wed Jan 27 14:14:15 2010 -0800 116881 116882 Merge remote branch 'alanc/master' 116883 116884commit 1e6fd65d0b95260253828678131885a4ec21c594 116885Author: Michael Vogt <mvo@ubuntu.com> 116886Date: Mon Jan 25 18:41:20 2010 +0100 116887 116888 xfree86/modes: only call gamma_set if its non NULL 116889 116890 I ran accross a crash with xf86-video-nv-2.1.15 [1] and xserver 116891 1.7.3.901. It looks like the problem is that gamma_set is called even 116892 if that is NULL. 116893 116894 [1] https://launchpad.net/bugs/494627 116895 116896 Reviewed-By: Matthias Hopf <mhopf@suse.de> 116897 Signed-off-by: Julien Cristau <jcristau@debian.org> 116898 Signed-off-by: Keith Packard <keithp@keithp.com> 116899 116900commit 0688dca044f966abc3da667f6d4e79e7cf47f996 116901Author: Eric Anholt <eric@anholt.net> 116902Date: Thu Jan 21 10:31:04 2010 -0800 116903 116904 dri2: Fix order of operations issue in __glXdriSwapEvent test. 116905 116906 Clients would have received swap events regardless of asking for it. 116907 116908 Signed-off-by: Eric Anholt <eric@anholt.net> 116909 Reviewed-by: Keith Packard <keithp@keithp.com> 116910 Signed-off-by: Keith Packard <keithp@keithp.com> 116911 116912commit a6bd5d2e482a5aa84acb3d4932e2a166d8670ef1 116913Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 116914Date: Wed Jan 27 14:03:03 2010 -0800 116915 116916 Fix source pictures getting random transforms after 2d6a8f668342a5190cdf43b5. 116917 116918 *xoff and *yoff were uninitialized for source-only pictures.x 116919 116920 Signed-off-by: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 116921 Reviewed-by: Aaron Plattner <aplattner@nvidia.com> 116922 Signed-off-by: Keith Packard <keithp@keithp.com> 116923 116924commit c3395158678aaab9dca5fc6a812cbe715ddc0e1a 116925Author: Gaetan Nadon <memsize@videotron.ca> 116926Date: Fri Jan 15 14:13:18 2010 -0500 116927 116928 packaging: provide a default README file #24206 116929 116930 All modules should have a README file. 116931 The bottom URL section is found in all X.Org README files. 116932 116933 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 116934 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 116935 Signed-off-by: Keith Packard <keithp@keithp.com> 116936 116937commit f311f2d047120fb816897444d2101465ff5189db 116938Author: Aaron Plattner <aplattner@nvidia.com> 116939Date: Wed Jan 13 11:35:52 2010 -0800 116940 116941 DRI2: Allow multiple driver names. 116942 116943 Each driver type (e.g. DRI2DriverDRI or DRI2DriverVDPAU) can have a name in the 116944 driverNames array in DRI2InfoRec. DRI2Connect returns the name for the driver 116945 specified by driverType. Also print names of supported drivers in 116946 DRI2ScreenInit. 116947 116948 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 116949 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 116950 Signed-off-by: Keith Packard <keithp@keithp.com> 116951 116952commit f57bc0ede8e018c7e264b917927c42a018cd1d5a 116953Author: Christian Zander <chzander@nvidia.com> 116954Date: Mon Jan 11 12:29:07 2010 -0800 116955 116956 x86emu: Respect the LEA 67h address size prefix. 116957 116958 Signed-off-by: Christian Zander <chzander@nvidia.com> 116959 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 116960 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> 116961 Signed-off-by: Keith Packard <keithp@keithp.com> 116962 116963commit 837bd2bbc02b893f96861b48c1f02b7b8e7e3e48 116964Author: Alan Coopersmith <alan.coopersmith@sun.com> 116965Date: Fri Oct 16 22:32:15 2009 -0700 116966 116967 Remove unbalanced ( from failure to move log error 116968 116969 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 116970 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 116971 116972commit 138d4c1670ebab435bf00627c97098a3a54b81a6 116973Author: Alan Coopersmith <alan.coopersmith@sun.com> 116974Date: Sat Jan 16 21:03:00 2010 -0800 116975 116976 glx: Sun compilers now support some gcc __attribute__ values 116977 116978 Sun cc 5.9 and later (__SUNPRO_C >= 0x590) support __attribute__ 116979 calls for aligned, always_inline, noinline, pure, const, and malloc. 116980 116981 This commit consists of the related updates to files that were 116982 regenerated by gl_XML.py in mesa after adding the __SUNPRO_C checks to it 116983 116984 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 116985 Signed-off-by: Brian Paul <brianp@vmware.com> 116986 116987commit 79e7b0b875634d0f9e1a95232a4e38adf617bc14 116988Author: Alan Coopersmith <alan.coopersmith@sun.com> 116989Date: Sat Jan 16 10:45:32 2010 -0800 116990 116991 Only enable kdrive input drivers on Linux by default 116992 116993 Fixes build on non-Linux platforms by restoring the defaults to 116994 where they were before commit 6c2b3a4247d10a50699ffa6abb643c5e959eefa8, 116995 to only enable the Linux kbd, mouse & evdev drivers when building 116996 on Linux platforms. 116997 116998 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 116999 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117000 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117001 117002commit 39ab474197bdad7d8e9ef496df2d61cbea39d370 117003Author: Alan Coopersmith <alan.coopersmith@sun.com> 117004Date: Tue Dec 15 19:07:38 2009 -0800 117005 117006 Move OS-specific VT key handler code from common to os-support 117007 117008 Adds new function xf86Activate to the OS-specific *VTsw*.c files 117009 and calls it from xf86ProcessActionEvent 117010 117011 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 117012 Tested-by: Tiago Vignatti <tiago.vignatti@nokia.com> (GNU/Linux) 117013 117014commit 15ca3312c069526b7f2207de9dfb9b9e851caf95 117015Author: Aaron Zang <Aaron.Zang@Sun.COM> 117016Date: Mon Dec 14 17:55:46 2009 -0800 117017 117018 Solaris: Avoid switching to inactive VT's 117019 117020 Fix for OpenSolaris bug 6876992: "[vconsole] Ctrl+Alt+F12 switchs to blank 117021 console screen with hotkeys property turned-off" 117022 http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6876992 117023 117024 Xorg needs to do sanity test for the VT it is commanded to switch to. 117025 If the VT is not opened by any process, discard the switching request. 117026 117027 The changes also contain the fix for some flaws discovered when 117028 getting the new gdm to run. 117029 117030 Signed-off-by: Aaron Zang <Aaron.Zang@Sun.COM> 117031 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 117032 117033commit 2984c18eb994696927a7f3b94d86fd47907334a0 117034Author: Gaetan Nadon <memsize@videotron.ca> 117035Date: Thu Jan 21 14:21:07 2010 -0500 117036 117037 config: replace custom code with reusable macro XORG_WITH_DOXYGEN 117038 117039 XORG_WITH_DOXYGEN provides additional functions like a configure 117040 option which allow platform builders to control the usage of 117041 the doxygen program. 117042 117043 This is a requirement from platforms that do not have such doc tool. 117044 A platform with a back level doxygen may use --without-doxygen 117045 to get the rest of the documentation built. 117046 117047 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 117048 Signed-off-by: Keith Packard <keithp@keithp.com> 117049 117050commit d9c20ee4191de7276a08288adffc24dff48aff8f 117051Author: Tiago Vignatti <tiago.vignatti@nokia.com> 117052Date: Fri Jan 15 19:34:29 2010 +0200 117053 117054 dix: move cursor realize code to its own function 117055 117056 The semantic remains, only code was moved: reuse chunk of code to realize 117057 cursor on both AllocARGBCursor and AllocGlyphCursor. 117058 117059 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 117060 Reviewed-by: Keith Packard <keithp@keithp.com> 117061 Signed-off-by: Keith Packard <keithp@keithp.com> 117062 117063commit f818f222362c93bcc859d997d96d8f3b447729f4 117064Author: Peter Hutterer <peter.hutterer@who-t.net> 117065Date: Fri Jan 15 15:32:19 2010 +1000 117066 117067 xkb: unexport xkbDevicePrivateKey and xkbUnwrapProc. 117068 117069 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117070 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 117071 Signed-off-by: Keith Packard <keithp@keithp.com> 117072 117073commit 6850ea8fb95417db9ce3a70fc17f90d6fdea1389 117074Author: Peter Hutterer <peter.hutterer@who-t.net> 117075Date: Fri Jan 15 15:32:16 2010 +1000 117076 117077 xfree86: replace True/False with TRUE/FALSE. 117078 117079 xf86Xinput.c relied on xkbsrv.h's definition of True/False which seems odd 117080 at first and weird on second glance. 117081 117082 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117083 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 117084 Signed-off-by: Keith Packard <keithp@keithp.com> 117085 117086commit 64b1372c15c3ede2696346e153bca9bd89016a20 117087Author: Peter Hutterer <peter.hutterer@who-t.net> 117088Date: Fri Jan 15 15:32:15 2010 +1000 117089 117090 xkb: remove unused _XkbIsPressEvent and _XkbIsReleaseEvent defines 117091 117092 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117093 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 117094 Signed-off-by: Keith Packard <keithp@keithp.com> 117095 117096commit c8bba14a390fe3fa16027e6b2433a314b3ea00c3 117097Author: Peter Hutterer <peter.hutterer@who-t.net> 117098Date: Fri Jan 15 15:32:13 2010 +1000 117099 117100 xkb: remove XkbAtomGetString, replace with NameForAtom. 117101 117102 XKB really XKBdoes not XKBneed its own XKBdefines for XKBeverything. 117103 117104 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117105 Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 117106 Signed-off-by: Keith Packard <keithp@keithp.com> 117107 117108commit f37799c9712afb7769f369162b904d9ea2dbd608 117109Author: Peter Hutterer <peter.hutterer@who-t.net> 117110Date: Fri Jan 15 15:32:12 2010 +1000 117111 117112 xkb: remove IsKeypadKey define, only used in two places. 117113 117114 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117115 Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 117116 Signed-off-by: Keith Packard <keithp@keithp.com> 117117 117118commit c8076f317ee5044ed3cc21b097f8b1741af8e0f1 117119Author: Peter Hutterer <peter.hutterer@who-t.net> 117120Date: Fri Jan 15 15:32:11 2010 +1000 117121 117122 xkb: remove XConvertCase. 117123 117124 Since it's typedef'd to XkbConvertCase anyway and the headers are now split 117125 from the client headers, simply get rid of it altogether. 117126 117127 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117128 Reviewed-by: Keith Packard <keithp@keithp.com> 117129 Signed-off-by: Keith Packard <keithp@keithp.com> 117130 117131commit d627dd9d1ef436b01f0581e40a6736ffc6b5d2c8 117132Author: Peter Hutterer <peter.hutterer@who-t.net> 117133Date: Fri Jan 15 15:32:10 2010 +1000 117134 117135 xkb: remove _XkbClearElems, a memset will do. 117136 117137 Bonus point - it's easier to understand what's actually being done with the 117138 memory. 117139 117140 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117141 Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 117142 Signed-off-by: Keith Packard <keithp@keithp.com> 117143 117144commit ea1de3fcdc40d2060bb6d13775005eadd624e767 117145Author: Peter Hutterer <peter.hutterer@who-t.net> 117146Date: Fri Jan 15 15:32:09 2010 +1000 117147 117148 xkb: remove _XkbTyped*alloc 117149 117150 Please no extension-specific macros for memory allocation. 117151 117152 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117153 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 117154 Signed-off-by: Keith Packard <keithp@keithp.com> 117155 117156commit 0f2d297dedeff8bd227df4c498cc668b0e902344 117157Author: Peter Hutterer <peter.hutterer@who-t.net> 117158Date: Tue Jan 5 16:31:37 2010 +1000 117159 117160 dix: EventToCore needs to copy the root window too. 117161 117162 This value isn't actually set for normal events but it saves us some work 117163 for the record extension support. 117164 117165 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117166 Reviewed-by: Keith Packard <keithp@keithp.com> 117167 Signed-off-by: Keith Packard <keithp@keithp.com> 117168 117169commit 6f6a99abc12ddee82898fdabfb50c17e90e094b9 117170Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 117171Date: Wed Jan 13 14:39:54 2010 -0800 117172 117173 XQuartz: Don't FatalError in x_hook_run if the list is empty 117174 117175 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 117176 117177commit 6bde306f7f6b9bbabeaa8bb910ea549be906cd8b 117178Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 117179Date: Wed Jan 13 11:20:29 2010 -0800 117180 117181 XQuartz: Update copyright in bundle for 2010 117182 117183 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 117184 117185commit 6008cc116493cb2825ad0bda0b407b7aefabb3f4 117186Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 117187Date: Mon Jan 11 18:02:55 2010 -0800 117188 117189 XQuartz: Setup the modifier map in the quartz thread 117190 117191 This avoids possible doing it twice which could result in incorrect 117192 keycodes for alt due to our loss of information about its side. 117193 117194 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 117195 117196commit 44f9c3d16c9c9b3362306a9ba26ee52e7baafeca 117197Merge: 032f97808 84956ca43 117198Author: Keith Packard <keithp@keithp.com> 117199Date: Wed Jan 13 10:19:21 2010 -0800 117200 117201 Merge remote branch 'jbarnes/master' 117202 117203commit 84956ca43b087600d9db297cffd62e960c516d9e 117204Author: Jesse Barnes <jbarnes@virtuousgeek.org> 117205Date: Mon Jan 11 14:56:24 2010 -0500 117206 117207 GLX/DRI2: add INTEL_swap_event support 117208 117209 This allows clients to easily check for swap completion status in their 117210 main loop. 117211 117212 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 117213 Reviewed-by: Adam Jackson <ajax@nwnk.net> 117214 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 117215 117216commit 04a54f69a8085ab3fe11a8713bd8b6b16ed1db27 117217Author: Jesse Barnes <jbarnes@virtuousgeek.org> 117218Date: Thu Jul 16 09:01:17 2009 -0400 117219 117220 DRI2: add support for new DRI2 protocol requests 117221 117222 Support the new DRI2 2.2 protocol requests: DRI2SwapBuffers, DRI2GetMSC, 117223 DRI2WaitMSC, DRI2WaitSBC and DRI2SwapInterval. 117224 117225 These requests allow the server to support the SGI_video_sync, 117226 SGI_swap_interval, and OML_sync_control GLX extensions if DDX support is 117227 present. The new DDX APIs are documented in dri2.h. 117228 117229 Reviewed-by: Kristian Høgsberg <krh@bitplanet.net> 117230 Reviewed-by: Adam Jackson <ajax@nwnk.net> 117231 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 117232 117233commit 032f97808c65771a07bac748212cf6457a5d1660 117234Author: Simon Thum <simon.thum@gmx.de> 117235Date: Wed Jan 6 17:32:24 2010 +0100 117236 117237 xfree86: init pointer feedback controls from options 117238 117239 With InputClass support, it makes more sense to cover all 117240 aspects of acceleration in options. Previously, one could only set the 117241 default on the command line. 117242 117243 Signed-off-by: Simon Thum <simon.thum@gmx.de> 117244 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117245 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117246 117247commit 0722c287a4c8a6cdedca9756192547bfcf77ade5 117248Author: Simon Thum <simon.thum@gmx.de> 117249Date: Wed Jan 6 19:43:59 2010 +0100 117250 117251 xfree86: document pointer acceleration in xorg.conf.man 117252 117253 Signed-off-by: Simon Thum <simon.thum@gmx.de> 117254 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117255 Reviewed-by: Fernando Carrijo <fcarrijo@yahoo.com.br> 117256 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117257 117258commit 1aca2d757534879eb55faaee421cb972a9546712 117259Author: Simon Thum <simon.thum@gmx.de> 117260Date: Wed Jan 6 19:13:36 2010 +0100 117261 117262 whitespace fixes 117263 117264 Signed-off-by: Simon Thum <simon.thum@gmx.de> 117265 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117266 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117267 117268commit 14039b5a7982fbf8130501bb00766176a4e9bccb 117269Author: Simon Thum <simon.thum@gmx.de> 117270Date: Wed Jan 6 18:16:20 2010 +0100 117271 117272 doc: actually document SendDragEvents 117273 117274 Signed-off-by: Simon Thum <simon.thum@gmx.de> 117275 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117276 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117277 117278commit 14e4e4a294e648e0bdcb70c34748e1b81c5bb64f 117279Author: Simon Thum <simon.thum@gmx.de> 117280Date: Wed Jan 6 18:13:27 2010 +0100 117281 117282 os: state effect of -a and -t options more precisely 117283 117284 Signed-off-by: Simon Thum <simon.thum@gmx.de> 117285 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117286 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117287 117288commit d727ddb91a549d23f6bf323e91db4b8916c1b7a0 117289Author: Gaetan Nadon <memsize@videotron.ca> 117290Date: Sat Jan 9 16:28:32 2010 -0500 117291 117292 kdrive: klinux.h breaks make dist 117293 117294 This file is no longer part of the source code 117295 and must be removed from distribution. 117296 117297 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 117298 Reviewed-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117299 Signed-off-by: Keith Packard <keithp@keithp.com> 117300 117301commit 6313d2da6c6910827d68cf31fe00b46a34c5bfc7 117302Author: Gaetan Nadon <memsize@videotron.ca> 117303Date: Fri Jan 8 19:04:25 2010 -0500 117304 117305 configure: use backticks rather than $() for commands 117306 117307 This patch to xserver configure.ac is to increase code portability to 117308 non POSIX system by using backticks rather than $() for command 117309 substitution for BUILD_DATE and BUILD_TIME. 117310 117311 Reviewed-by: Rémi Cardona <remi@gentoo.org> 117312 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 117313 Signed-off-by: Keith Packard <keithp@keithp.com> 117314 117315commit 74ab27f64a747ee45810f99cdc6e29b82cf8035e 117316Author: Tiago Vignatti <tiago.vignatti@nokia.com> 117317Date: Mon Dec 28 18:00:28 2009 +0200 117318 117319 Revert "Make sys.c use unaligned access functions provided in compiler." 117320 117321 This reverts commit da923d0bc15e99a8ed1986bd6f5df37f7af7284b. 117322 117323 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 117324 Signed-off-by: Keith Packard <keithp@keithp.com> 117325 117326commit 6a2a57832709798d99d19ff4c587e22f3b3c150d 117327Author: Gaetan Nadon <memsize@videotron.ca> 117328Date: Thu Jan 7 15:23:34 2010 -0500 117329 117330 macros: use PKG_CONFIG variable rather than executable name 117331 117332 User can defined alternate location for pkg-config. 117333 Once option in place, all instances of pkg-config must be converted. 117334 117335 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 117336 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 117337 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 117338 Signed-off-by: Keith Packard <keithp@keithp.com> 117339 117340commit 1b3132f5e981b3398e08263742332cc966bebc07 117341Author: Gaetan Nadon <memsize@videotron.ca> 117342Date: Thu Jan 7 14:12:34 2010 -0500 117343 117344 xwin: ignore intermediate .man file 117345 117346 Only detectable on a Cygwin build 117347 117348 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net 117349 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 117350 Signed-off-by: Keith Packard <keithp@keithp.com> 117351 117352commit 8ab4749aaefb3727b9fc58bb37b50a2d07eb1531 117353Merge: 8d53d8448 9437504b2 117354Author: Keith Packard <keithp@keithp.com> 117355Date: Thu Jan 7 10:32:21 2010 -0800 117356 117357 Merge remote branch 'dottedmag/for-keithp' 117358 117359commit 8d53d84485fdce8ea9686e6f300a69f7ddebd467 117360Author: Adam Tkac <atkac@redhat.com> 117361Date: Thu Jan 7 15:34:52 2010 +0100 117362 117363 Do not define members of include/eventstr.h:EventType enum conditionally. 117364 117365 Main problem is that EventType enumeration members can be different in 117366 module and in server, which obviously causes problems. 117367 117368 Signed-off-by: Adam Tkac <atkac@redhat.com> 117369 Reviewed-by: Adam Jackson <ajax@redhat.com> 117370 Signed-off-by: Keith Packard <keithp@keithp.com> 117371 117372commit 9437504b21dfdb509ab8b3a00f64403e1e4026dd 117373Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117374Date: Mon Jan 4 05:42:02 2010 +0600 117375 117376 kdrive: Remove unused kdNoopOps external variable declaration 117377 117378 Reviewed-by: Adam Jackson <ajax@redhat.com> 117379 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 117380 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117381 117382commit aa6e0936be15488455e6a4236806a58496da2979 117383Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117384Date: Mon Jan 4 05:30:22 2010 +0600 117385 117386 kdrive: Remove unused kmap.c 117387 117388 Reviewed-by: Adam Jackson <ajax@redhat.com> 117389 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 117390 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117391 117392commit f78ac5c6e273f575af29746e08e7c182ff7b93b3 117393Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117394Date: Mon Jan 4 02:36:30 2010 +0600 117395 117396 kdrive: Make internal functions static 117397 117398 Reviewed-by: Adam Jackson <ajax@redhat.com> 117399 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 117400 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117401 117402commit 37a8ddc0dbf3fb7f059b689038979bace206dccf 117403Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117404Date: Mon Jan 4 02:28:37 2010 +0600 117405 117406 kdrive: Remove unused 'k' variable 117407 117408 Reviewed-by: Adam Jackson <ajax@redhat.com> 117409 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 117410 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117411 117412commit af1f1a05e1aa9ec921f1288818a66766c301f8b0 117413Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117414Date: Mon Jan 4 01:50:04 2010 +0600 117415 117416 kdrive: Remove bus mouse driver 117417 117418 Bus mice aren't used anymore, do not keep dead code around. 117419 117420 Reviewed-by: Adam Jackson <ajax@redhat.com> 117421 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 117422 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117423 117424commit a3cc3af186b55ac80abd6859ffe491f4ae936596 117425Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117426Date: Sat Jan 2 23:31:28 2010 +0600 117427 117428 kdrive: Adjust kdrive usage message as shadow overlay support has been dropped 117429 117430 Reviewed-by: Adam Jackson <ajax@redhat.com> 117431 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 117432 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117433 117434commit c40b9f959175057fac726fcdfff9641ae0681cf6 117435Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117436Date: Mon Jan 4 14:07:37 2010 +0600 117437 117438 kdrive: Remove unused VxWorks* variable declarations 117439 117440 Reviewed-by: Adam Jackson <ajax@redhat.com> 117441 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 117442 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117443 117444commit 4e0d580d5ac00d8d87c8e0080a02868c87a31caa 117445Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117446Date: Sat Jan 2 15:04:03 2010 +0600 117447 117448 kdrive: Remove unused KdCardInfo::lastMarker field 117449 117450 Reviewed-by: Adam Jackson <ajax@redhat.com> 117451 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117452 117453commit dbaab9759ba31edf63cf73eda86ba931d88e8cce 117454Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117455Date: Sat Jan 2 16:18:29 2010 +0600 117456 117457 kdrive: Extract common part of fbdevPutColors and fbdevEnable 117458 117459 Put framebuffer colormap updating code in separate function 117460 for brevity. 117461 117462 Reviewed-by: Adam Jackson <ajax@redhat.com> 117463 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117464 117465commit 7ee14154b230b931d294288e1291c4fc617b1354 117466Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117467Date: Sat Jan 2 15:42:53 2010 +0600 117468 117469 kdrive: Remove unused overlay fb support 117470 117471 Xfbdev, Xephyr and Xfake all use only one framebuffer, so simplify 117472 implementation by removing overlay support. 117473 117474 Reviewed-by: Adam Jackson <ajax@redhat.com> 117475 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117476 117477commit 730f7d1c4fae8ca582a4a9998a4dc4b325f98896 117478Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117479Date: Sat Jan 2 15:06:18 2010 +0600 117480 117481 kdrive: Remove unused KD_MAX_CARD_ADDRESS macro 117482 117483 Reviewed-by: Adam Jackson <ajax@redhat.com> 117484 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117485 117486commit 5337ddcfd9dc2a5b823da39df7adaad6692214f0 117487Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117488Date: Sat Jan 2 14:59:07 2010 +0600 117489 117490 kdrive: Move Xephyr-specific fields out of KdScreenInfo 117491 117492 memory_base, memory_size, off_screen_base fields in 117493 KdScreenInfo are used only by fake EXA in Xephyr. Move 117494 them into Xephyr, cleanup Xfake and Xfbdev. 117495 117496 Reviewed-by: Adam Jackson <ajax@redhat.com> 117497 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117498 117499commit 62883b499b72f678ab055de4d0370227ae441370 117500Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117501Date: Sat Jan 2 13:17:21 2010 +0600 117502 117503 kdrive: Remove unused KdCardAttr from KdCardInfo 117504 117505 Card attrs are unused in all current kdrive servers, so 117506 remove it completely to avoid allocating and passing dummy 117507 values to KdCardInfoAdd. 117508 117509 Reviewed-by: Adam Jackson <ajax@redhat.com> 117510 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117511 117512commit 6c2b3a4247d10a50699ffa6abb643c5e959eefa8 117513Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117514Date: Mon Jan 4 14:07:04 2010 +0600 117515 117516 kdrive: Add option to compile out input drivers 117517 117518 Add --without-kdrive-{kbd,mouse,evdev} configure options disabling 117519 Linux keyboard driver, Linux mouse drivers (ps2, bus,ms), 117520 and Linux evdev driver. 117521 117522 Build all drivers by default as before. 117523 117524 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 117525 Reviewed-by: Adam Jackson <ajax@redhat.com> 117526 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117527 117528commit 72ba717b1d639e94fa9ab56ca026131edc020a30 117529Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117530Date: Sat Jan 2 15:03:24 2010 +0600 117531 117532 kdrive: Remove unused KdCardInfo::needSync field 117533 117534 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117535 Reviewed-by: Adam Jackson <ajax@redhat.com> 117536 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117537 117538commit d21b41cdb5e89eb428f36b19e965f0ab49e051d9 117539Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117540Date: Sat Jan 2 13:21:25 2010 +0600 117541 117542 kdrive: Remove .gitignore left from SDL server 117543 117544 SDL server is gone, no need to keep its .gitignore anymore. 117545 117546 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 117547 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117548 Reviewed-by: Adam Jackson <ajax@redhat.com> 117549 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117550 117551commit 6f265d55a61f9be323583b8acacae783be72bda9 117552Author: Peter Hutterer <peter.hutterer@who-t.net> 117553Date: Tue Jan 5 13:15:25 2010 +1000 117554 117555 dix: don't update the slave coordinates from the VCK. 117556 117557 A keyboard event from a device with both valuators and keys will be posted 117558 through the VCK. In this case, do not update the slave device coordinates 117559 from the VCK - they're always 0/0. Leave them as-is, for the next pointer 117560 event will continue where it left. 117561 117562 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117563 Reviewed-by: Keith Packard <keithp@keithp.com> 117564 Signed-off-by: Keith Packard <keithp@keithp.com> 117565 117566commit 326429badfc76885e4652ddc72860810c0e8d102 117567Author: Adam Jackson <ajax@redhat.com> 117568Date: Tue Nov 10 14:56:36 2009 -0500 117569 117570 modes: Remove the ClockRanges type 117571 117572 ba2d39dd5428cb5922b797a1d4ea45b859412b40 introduced warnings: 117573 117574 xf86Mode.c: In function ‘xf86CheckModeForDriver’: 117575 xf86Mode.c:986: warning: passing argument 1 of ‘modeInClockRange’ from incompatible pointer type 117576 xf86Mode.c:253: note: expected ‘ClockRangePtr’ but argument is of type ‘ClockRangesPtr’ 117577 xf86Mode.c:1002: warning: passing argument 1 of ‘modeInClockRange’ from incompatible pointer type 117578 xf86Mode.c:253: note: expected ‘ClockRangePtr’ but argument is of type ‘ClockRangesPtr’ 117579 117580 Because I foolishly didn't notice that we had types with nearly 117581 identical members named ClockRange and ClockRanges. The latter 117582 contained an extra 'strategy' member at the end, which claimed to be 117583 needed by the vidmode extension. Of course, this was a lie: the only time 117584 we'd use it was in mode validation, for drivers using LOOKUP_CLKDIV2 with 117585 non-programmable clocks. The only driver using LOOKUP_CLKDIV2 is 117586 rendition, which has a programmable clock. The only driver using the 117587 ClockRanges type was smi, which did not use it for its 'strategy' member, 117588 so has been fixed to use ClockRange instead. 117589 117590 Signed-off-by: Adam Jackson <ajax@redhat.com> 117591 Reviewed-by: Keith Packard <keithp@keithp.com> 117592 Signed-off-by: Keith Packard <keithp@keithp.com> 117593 117594commit 13c8bd3fde3b0831921e59f84936022a16379d63 117595Author: Alan Coopersmith <alan.coopersmith@sun.com> 117596Date: Mon Jan 4 18:21:54 2010 -0800 117597 117598 CloseDevice: call XkbRemoveResourceClient before freeing key class struct 117599 117600 XkbRemoveResourceClient() returns immediately if dev->key is NULL. 117601 CloseDevice calls XkbRemoveResourceClient until it removes all resources. 117602 117603 If we free dev->key and NULL it before XkbRemoveResourceClient, then 117604 infinite loop ensues, and the server appears to hang on exit or crash. 117605 117606 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 117607 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117608 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 117609 Signed-off-by: Keith Packard <keithp@keithp.com> 117610 117611commit e707612e9de4e5e1d91dc1a8152810912d7e18bb 117612Author: Michel Dänzer <daenzer@vmware.com> 117613Date: Mon Jan 4 09:07:48 2010 +0100 117614 117615 EXA: Restore migration call in exaDoPutImage(). 117616 117617 Turns out this is still necessary if the driver PrepareAccess hook succeeds. 117618 117619 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117620 Tested-by: Maarten Maathuis <madman2003@gmail.com> 117621 Tested-by: Andrew Chant <andrew.chant+debian@gmail.com> 117622 Signed-off-by: Keith Packard <keithp@keithp.com> 117623 117624commit e10072b7c74a6f5e24f6dfa37e73688d1f3425e3 117625Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117626Date: Sat Jan 2 05:43:28 2010 +0600 117627 117628 Add Xephyr.man to .gitignore 117629 117630 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117631 Reviewed-by: Gaetan Nadon <memsize@videotron.ca> 117632 Signed-off-by: Keith Packard <keithp@keithp.com> 117633 117634commit fb26cb793c6eba189050662c566f7ea0559bd486 117635Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117636Date: Sat Jan 2 05:43:27 2010 +0600 117637 117638 Remove unused pShadow field from drivers' private structures 117639 117640 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117641 Signed-off-by: Keith Packard <keithp@keithp.com> 117642 117643commit 2c85d72fc7bd013af327321ad5d626dc73c3caf0 117644Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117645Date: Sat Jan 2 05:43:26 2010 +0600 117646 117647 Do not check xfree argument for NULL 117648 117649 xfree itself checks for NULL, and even this is not necessary 117650 as passing NULL to free(3) is safe. 117651 117652 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117653 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117654 Signed-off-by: Keith Packard <keithp@keithp.com> 117655 117656commit 15bc35c6e44c65812de5c3e85ec8bbf78b032ded 117657Author: Mikhail Gusarov <dottedmag@dottedmag.net> 117658Date: Sat Jan 2 05:43:25 2010 +0600 117659 117660 Remove trailing whitespace 117661 117662 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 117663 Signed-off-by: Keith Packard <keithp@keithp.com> 117664 117665commit 1763550d0181ac1c775b9ddf490114eff2fbe67e 117666Author: Simon Thum <simon.thum@gmx.de> 117667Date: Fri Jan 1 19:58:05 2010 +0100 117668 117669 dix: add smooth limited pointer acceleration profile 117670 117671 This profile is inspired by the accel code removed from the wacom driver. 117672 It ascends from zero to acceleration, maxing out at threshold. This means you 117673 can control the slope using threshold, which wasn't possible in wacom. 117674 For sanity's sake, threshold should grow with acceleration. 117675 117676 Works best with adaptive deceleration, since otherwise it only generates 117677 acceleration above 1, causing seldom pixel skips. 117678 117679 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117680 Signed-off-by: Keith Packard <keithp@keithp.com> 117681 117682commit 435f27667f84269768efecde34de4af2b2d43376 117683Author: Julien Cristau <jcristau@debian.org> 117684Date: Sun Sep 27 14:45:47 2009 +0200 117685 117686 config: add libudev input-hotplug backend 117687 117688 Add a backend using libudev for input hotplug, and disable the hal and 117689 dbus backends if this one is enabled. 117690 117691 XKB configuration happens using xkb{rules,model,layout,variant,options} 117692 properties (case-insensitive) on the device. We fill in InputAttributes 117693 to allow configuration through InputClass in Xorg. 117694 117695 Requires udev 148 for the input_id helper and ID_INPUT* properties. 117696 117697 Signed-off-by: Julien Cristau <jcristau@debian.org> 117698 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 117699 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117700 117701commit b8b12e41c453c3bf94b11e7a18934d3b6e1869bf 117702Author: Peter Hutterer <peter.hutterer@who-t.net> 117703Date: Wed Dec 23 16:05:16 2009 +1000 117704 117705 xfree86: move sanity checks below option and input classes merges. 117706 117707 While the identifier is likely set before the input classes are merged, the 117708 driver may not be. Hence don't check for a driver before we've completed 117709 configuration for this device. 117710 117711 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117712 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 117713 Signed-off-by: Julien Cristau <jcristau@debian.org> 117714 117715commit 9fad8f06fb89ac2ae05bea0fa24cab3df7677297 117716Merge: 871bbe1d8 42e8c9224 117717Author: Keith Packard <keithp@keithp.com> 117718Date: Wed Dec 30 09:28:19 2009 -0800 117719 117720 Merge remote branch 'dbn/inputclass' 117721 117722commit 871bbe1d87fa3c7ebd075e1d1eec33e45b08493d 117723Merge: db2c6f7c9 e1165632b 117724Author: Keith Packard <keithp@keithp.com> 117725Date: Wed Dec 30 09:16:45 2009 -0800 117726 117727 Merge remote branch 'dbn/xorg.conf.d' 117728 117729commit db2c6f7c91289b5d49978974093a1002b3b53a56 117730Author: Michel Dänzer <daenzer@vmware.com> 117731Date: Tue Dec 29 15:48:44 2009 +0100 117732 117733 EXA: Don't use UTS/DFS directly for Put/GetImage when there's a system copy. 117734 117735 We want to save the result in the system memory copy, in case we'll need it 117736 again for subsequent software fallbacks. 117737 117738 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117739 Acked-By: Maarten Maathuis <madman2003@gmail.com> 117740 Signed-off-by: Keith Packard <keithp@keithp.com> 117741 117742commit f28ca728e90ccbb901e8e7215a842525d8e786d3 117743Author: Thomas Hellstrom <thellstrom@vmware.com> 117744Date: Tue Dec 29 12:51:30 2009 +0100 117745 117746 EXA: Restrict the regions that need to be migrated for composite fallback for src / mask pictures. 117747 117748 [ Michel: Minor fixups to address compiler warnings ] 117749 117750 Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> 117751 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117752 Acked-by: Maarten Maathuis <madman2003@gmail.com> 117753 Signed-off-by: Keith Packard <keithp@keithp.com> 117754 117755commit 0c1f43c0f3e888172f11f62a2f208af70e0183cc 117756Author: Michel Dänzer <daenzer@vmware.com> 117757Date: Tue Dec 29 12:51:29 2009 +0100 117758 117759 EXA/mixed: Handle results of software fallbacks in DamageReport hook. 117760 117761 This is more elegant and probably also slightly more correct than doing it 117762 at FinishAccess time. 117763 117764 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117765 Acked-by: Maarten Maathuis <madman2003@gmail.com> 117766 Signed-off-by: Keith Packard <keithp@keithp.com> 117767 117768commit 08bf26c28fc8147b64f2fe8345b9581e0101571c 117769Author: Michel Dänzer <daenzer@vmware.com> 117770Date: Tue Dec 29 12:51:28 2009 +0100 117771 117772 EXA: Fix migration avoidance for 1x1 pixmaps. 117773 117774 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117775 Acked-by: Maarten Maathuis <madman2003@gmail.com> 117776 Signed-off-by: Keith Packard <keithp@keithp.com> 117777 117778commit 0bd8f0cd7f3823ee17ae8c88fd3e004bfff9982d 117779Author: Michel Dänzer <daenzer@vmware.com> 117780Date: Tue Dec 29 12:51:27 2009 +0100 117781 117782 EXA/classic: Fix crash with migration heuristic "smart". 117783 117784 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117785 Acked-by: Maarten Maathuis <madman2003@gmail.com> 117786 Signed-off-by: Keith Packard <keithp@keithp.com> 117787 117788commit 40453bf7718f1c3d672b87e9d84991032cbef859 117789Author: Thomas Hellstrom <thellstrom@vmware.com> 117790Date: Tue Dec 29 09:34:44 2009 +0100 117791 117792 EXA: Limit src prepareAccess regions for a number of unaccelerated operations. 117793 117794 When we can trivially calculate the affected source regions, 117795 do that before calling region bounded prepareAccess. 117796 117797 [ Michel: Minor fixups to address compiler warnings ] 117798 117799 Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> 117800 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117801 Acked-by: Maarten Maathuis <madman2003@gmail.com> 117802 Signed-off-by: Keith Packard <keithp@keithp.com> 117803 117804commit aa9ce8ab343d51a67924757c5a43c9572248bb24 117805Author: Michel Dänzer <daenzer@vmware.com> 117806Date: Tue Dec 29 09:34:43 2009 +0100 117807 117808 EXA: Use relevant source region to minimize migration on CopyWindow fallbacks. 117809 117810 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117811 Acked-by: Maarten Maathuis <madman2003@gmail.com> 117812 Signed-off-by: Keith Packard <keithp@keithp.com> 117813 117814commit 39cc110caa1f6481a7213ae39f82669333ec1645 117815Author: Thomas Hellstrom <thellstrom@vmware.com> 117816Date: Tue Dec 29 09:34:42 2009 +0100 117817 117818 EXA: Fix bugs in exaGetImage / ExaCheckGetImage migration. 117819 117820 Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> 117821 Acked-by: Michel Dänzer <michel@daenzer.net> 117822 Acked-by: Maarten Maathuis <madman2003@gmail.com> 117823 Signed-off-by: Keith Packard <keithp@keithp.com> 117824 117825commit d42f3960514f1c410863ac6c09c5638b5ddeae20 117826Author: Michel Dänzer <daenzer@vmware.com> 117827Date: Mon Dec 28 16:48:18 2009 +0100 117828 117829 EXA: Allow optimized migration to be enabled with mixed pixmaps. 117830 117831 This was always the intention, I only recently realized it wasn't the case 117832 yet... 117833 117834 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117835 Acked-by: Maarten Maathuis <madman2003@gmail.com> 117836 Signed-off-by: Keith Packard <keithp@keithp.com> 117837 117838commit 661630f9099f254e54c8da513aaed12f6532cc35 117839Author: Michel Dänzer <daenzer@vmware.com> 117840Date: Mon Dec 28 16:48:17 2009 +0100 117841 117842 EXA/mixed: Don't clear deferred status of pixmaps if migration is limited. 117843 117844 * With optimized migration, only the pending damage region is synchronized for 117845 destination pixmaps. 117846 * Migration of source pixmaps can be limited to a bounding region. 117847 117848 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 117849 Acked-by: Maarten Maathuis <madman2003@gmail.com> 117850 Signed-off-by: Keith Packard <keithp@keithp.com> 117851 117852commit 42e8c9224e6c54655c45f87999d37d0d67b3f7f5 117853Author: Dan Nicholson <dbn.lists@gmail.com> 117854Date: Fri Oct 2 06:29:28 2009 -0700 117855 117856 xfree86: Introduce InputClass configuration 117857 117858 Currently Xorg uses hal's fdi files to decide what configuration options 117859 are applied to automatically added input devices. This is sub-optimal 117860 since it requires users to use a new and different configuration store 117861 than xorg.conf. 117862 117863 The InputClass section attempts to provide a system similar to hal where 117864 configuration can be applied to all devices with certain attributes. For 117865 now, devices can be matched to: 117866 117867 * A substring of the product name via a MatchProduct entry 117868 * A substring of the vendir name via a MatchVendor entry 117869 * A pathname pattern of the device file via a MatchDevicePath entry 117870 * A device type via boolean entries for MatchIsKeyboard, MatchIsPointer, 117871 MatchIsJoystick, MatchIsTablet, MatchIsTouchpad and MatchIsTouchscreen 117872 117873 See the INPUTCLASS section in xorg.conf(5) for more details. 117874 117875 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 117876 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 117877 117878commit 0711598dd3e8366217676f462f1af7d0899656d9 117879Author: Dan Nicholson <dbn.lists@gmail.com> 117880Date: Wed Sep 2 06:47:13 2009 -0700 117881 117882 config: Introduce InputAttributes in NewInputDeviceRequest 117883 117884 In order to give NewInputDeviceRequest more information, a new 117885 InputAttributes type is introduced. Currently, this collects the product 117886 and vendor name, device path, and sets booleans for attributes such as 117887 having keys and/or a pointer. Only the HAL backend fills in the 117888 attributes, though. 117889 117890 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 117891 Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net> 117892 117893commit c6e8637e29e0ca11dfb35c02da7ca6002ac8c597 117894Author: Dan Nicholson <dbn.lists@gmail.com> 117895Date: Fri Oct 2 06:28:03 2009 -0700 117896 117897 xfree86: Support non-Option boolean entries in configuration 117898 117899 Refactored code into the parser to allow the freeform boolean types used 117900 in Option entries to be used in other configuration entries. This isn't 117901 as powerful as allowing "No" to precede the option names, but it atleast 117902 gives a common handling of "yes", "no", etc. 117903 117904 A type xf86TriState has been added to support an optional boolean. This 117905 allows the boolean sense of the value to be kept while providing a means 117906 to signal that it is unset. 117907 117908 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 117909 Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net> 117910 117911commit e1165632bdfbd720889ed1adf5f7ab338032c0ee 117912Author: Peter Hutterer <peter.hutterer@who-t.net> 117913Date: Mon Dec 21 15:56:35 2009 +1000 117914 117915 xfree86: Add Option AutoServerLayout for input devices. 117916 117917 Any input device with this option will be automatically added to whichever 117918 server layout is selected at startup. This removes the need to reference a 117919 device from the ServerLayout section. The two following configuration are 117920 identical: 117921 117922 CONFIG 1: 117923 Section "ServerLayout" 117924 InputDevice "foo" 117925 EndSection 117926 117927 Section "InputDevice" 117928 Identifier "foo" 117929 ... 117930 EndSection 117931 117932 CONFIG 2: 117933 Section "InputDevice" 117934 Identifier "foo" 117935 Option "AutoServerLayout" "on" 117936 ... 117937 EndSection 117938 117939 The selection of the server layout affects both explicitly specified 117940 layouts and the implicit layout. 117941 117942 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 117943 Reviewed-by: Keith Packard <keithp at keithp.com> 117944 117945commit 592b20c517461d32daf44a940386ffcc11c434f8 117946Author: Dan Nicholson <dbn.lists@gmail.com> 117947Date: Mon Dec 21 01:13:44 2009 -0800 117948 117949 xfree86: Allow config directory to be specified on command line 117950 117951 Add a new command line parameter, -configdir, to specify the config 117952 directory to be used. Rules are the same as -config for root vs. user 117953 privileges. 117954 117955 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 117956 Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net> 117957 117958commit efa5269f23c2237eb5368bf5245ffbbf35714153 117959Author: Dan Nicholson <dbn.lists@gmail.com> 117960Date: Mon Dec 21 00:04:16 2009 -0800 117961 117962 xfree86: Use xorg.conf.d directory for multiple config files 117963 117964 Currently there is a single file, xorg.conf, for configuring the server. 117965 This works fine most of the time, but it becomes a problem when packages 117966 or system services need to adjust the configuration. Instead, allow 117967 multiple configuration files to live in a directory. Typically this will 117968 be /etc/X11/xorg.conf.d. 117969 117970 Files with a suffix of .conf will be read and added to the server 117971 configuration after xorg.conf. The server won't fall back to using the 117972 auto configuration unless there is no config file and there are no files 117973 in the config directory. 117974 117975 Right now this uses a simpler search template than the config file 117976 search path by not using the command line or environment variable 117977 parameters. The matching code was refactored a bit to make this more 117978 coherent. Any DDX wanting to read the config files will need to call 117979 xf86initConfigFiles before opening/reading them. This is to allow 117980 xf86openConfigFile without xf86openConfigDirFiles and vice-versa. 117981 117982 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 117983 Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net> 117984 117985commit f1e869aca176e9113f9d234b0ea59b4cac295cc4 117986Author: Dan Nicholson <dbn.lists@gmail.com> 117987Date: Mon Dec 21 00:03:11 2009 -0800 117988 117989 xfree86: Unexport configuration file symbols 117990 117991 These functions should not be used outside of DDXs, so no need to put 117992 them in the ABI. 117993 117994 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 117995 Reviewed-by: Peter Hutterer <peter.hutterer at who-t.net> 117996 117997commit 90e6d93cf9bfafd63d7849dc16ce194d6f9c9d5f 117998Author: Peter Hutterer <peter.hutterer@who-t.net> 117999Date: Wed Dec 23 12:54:14 2009 +1000 118000 118001 test/xi2: fail if xi2 class type is garbage. (#25492) 118002 118003 If the keycode range exceeds the allowable length, memory gets overwritten. 118004 Catch this case by making sure that only allowed class types are 118005 present. 118006 118007 X.Org Bug 25492 <http://bugs.freedesktop.org/show_bug.cgi?id=25492> 118008 118009 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118010 Reviewed-by: Dave Airlie <airlied@redhat.com> 118011 Signed-off-by: Keith Packard <keithp@keithp.com> 118012 118013commit 72286eae9f4a2f2ca6c46919ff8aa1011429e03a 118014Author: Peter Hutterer <peter.hutterer@who-t.net> 118015Date: Tue Dec 22 15:42:59 2009 +1000 118016 118017 xfree86: remove LAYOUT_DEBUG section. 118018 118019 I don't think this one has been in use since 2003. 118020 118021 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118022 Reviewed-by: Dave Airlie <airlied@redhat.com> 118023 Signed-off-by: Keith Packard <keithp@keithp.com> 118024 118025commit 38b5afb0b5b46b61effa6f0e5a377b343e357e7e 118026Author: Peter Hutterer <peter.hutterer@who-t.net> 118027Date: Tue Dec 22 15:40:40 2009 +1000 118028 118029 xfree86: remove some 'enable this later' and if 0 ifdefs 118030 118031 2003 called, they want their ifdefs back. 118032 118033 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118034 Reviewed-by: Dave Airlie <airlied@redhat.com> 118035 Signed-off-by: Keith Packard <keithp@keithp.com> 118036 118037commit 7dc78db79edd15b971110acaf16bbe7f69e40bdb 118038Merge: 81a623f03 b44c9be24 118039Author: Keith Packard <keithp@keithp.com> 118040Date: Tue Dec 22 21:25:18 2009 -0800 118041 118042 Merge remote branch 'whot/master' 118043 118044commit b44c9be244cee286835855483a69c69e80b095c0 118045Author: Julien Cristau <jcristau@debian.org> 118046Date: Tue Dec 22 17:14:09 2009 +0100 118047 118048 test/xi2: fix maximum max_keycode (bug#25492) 118049 118050 The number of keycodes needs to be lower than 0xFFFD so that the length 118051 field of xXIKeyInfo doesn't overflow. 118052 118053 Signed-off-by: Julien Cristau <jcristau@debian.org> 118054 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118055 118056commit 190610e0c62170a27ab3e40c6c6210a583ae1ad4 118057Author: Peter Hutterer <peter.hutterer@who-t.net> 118058Date: Mon Dec 21 11:08:26 2009 +1000 118059 118060 xfree86: remove HistorySize from the xorg.conf man page. 118061 118062 This option isn't parsed by anything anymore. 118063 118064 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118065 Reviewed-by: Keith Packard <keithp@keithp.com> 118066 118067commit 094c6b9f97a9f92e5a0ef3cf5be24c09ed4d6063 118068Author: Peter Hutterer <peter.hutterer@who-t.net> 118069Date: Mon Dec 21 10:42:32 2009 +1000 118070 118071 xfree86: reword InputDevice man sections, deprecate CorePointer/CoreKeyboard 118072 118073 Reshuffle and reword - InputDevice sections are only necessary if 118074 hotplugging is disabled. Put more emphasis on hotplugging and less on HAL 118075 since we'll switch backends eventually. 118076 118077 CorePointer, CoreKeyboard, and AlwaysCore should be listed as deprecated 118078 since they don't do what they used to since 1.4. These days, only 118079 SendCoreEvents matters and it's enabled for any driver calling 118080 xf86ProcessCommonOptions (== every driver). 118081 It only controls the startup behavior too, so document this. 118082 118083 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118084 Acked-by: Keith Packard <keithp@keithp.com> 118085 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 118086 118087commit 81a623f036fe56d53c4e3bdafb3eaf945502525f 118088Author: Keith Packard <keithp@keithp.com> 118089Date: Sat Dec 19 20:11:41 2009 -0800 118090 118091 Bump to 1.7.99.3 (unreleased) 118092 118093 Signed-off-by: Keith Packard <keithp@keithp.com> 118094 118095commit 3c30c5b6d321f34736c442c9cd982308d9b8b93a 118096Author: Keith Packard <keithp@keithp.com> 118097Date: Fri Dec 18 22:58:59 2009 -0800 118098 118099 Set release date for 1.7.99.2 118100 118101 Signed-off-by: Keith Packard <keithp@keithp.com> 118102 118103commit 895f40792a14d8b88923bf3b428d31ae3bb31e46 118104Author: Alan Coopersmith <alan.coopersmith@sun.com> 118105Date: Wed Dec 2 17:43:01 2009 -0800 118106 118107 Add type name argument to CreateNewResourceType 118108 118109 Convert all calls of CreateNewResourceType to pass name argument 118110 118111 Breaks DIX ABI. 118112 118113 ABI versions bumped: 118114 118115 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118116 Reviewed-by: Keith Packard <keithp@keithp.com> 118117 Signed-off-by: Keith Packard <keithp@keithp.com> 118118 118119commit a11c58fa0c5809f918b36b84be53385cb2d8ea59 118120Author: Alan Coopersmith <alan.coopersmith@sun.com> 118121Date: Fri Dec 11 00:58:25 2009 -0800 118122 118123 Ensure all resource types created have names registered 118124 118125 Calls RegisterResourceName to record the type name for 118126 use by X-Resource, XACE/SELinux/XTsol, and DTrace. 118127 118128 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118129 Reviewed-by: Keith Packard <keithp@keithp.com> 118130 118131commit eb750f8b5e14751d4c40b50499baec5d2ba79db9 118132Author: Alan Coopersmith <alan.coopersmith@sun.com> 118133Date: Fri Dec 11 00:46:22 2009 -0800 118134 118135 Check for failures from CreateNewResourceType 118136 118137 Make sure to check return value before setting bitmask flags. 118138 For most calls, just fails to init the extension. Since Xinput 118139 already calls FatalError() on initialization failure, so does 118140 failure to allocate Xinput's resource type. 118141 118142 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118143 Reviewed-by: Keith Packard <keithp@keithp.com> 118144 118145commit 1df4bd6011e110dcf0649b15bfffd4ab9e6961d6 118146Author: Alan Coopersmith <alan.coopersmith@sun.com> 118147Date: Thu Dec 17 18:24:40 2009 -0800 118148 118149 Change default xkb model from pc104 to pc105 118150 118151 PC105 is a more useful default for non-American keyboard users, 118152 not harmful for American PC101/PC104 keyboard users. 118153 118154 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118155 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 118156 Signed-off-by: Keith Packard <keithp@keithp.com> 118157 118158commit 1d2e189cdc1d65c9ca2aa0a950780dc18e6c50f1 118159Author: Alan Coopersmith <alan.coopersmith@sun.com> 118160Date: Thu Dec 17 18:24:39 2009 -0800 118161 118162 Change sysconfdir brackets to avoid some shells trying to run it as a command 118163 118164 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118165 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 118166 Signed-off-by: Keith Packard <keithp@keithp.com> 118167 118168commit 48749cc21b074ee8e68c3854bd3977ec7c408225 118169Author: Alan Coopersmith <alan.coopersmith@sun.com> 118170Date: Thu Dec 17 18:24:38 2009 -0800 118171 118172 Add platform compatibility defines for Sun Studio compilers 118173 118174 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118175 Signed-off-by: Keith Packard <keithp@keithp.com> 118176 118177commit b63912ed4c69fedd1bea92274d6cae0429a79677 118178Author: Alan Coopersmith <alan.coopersmith@sun.com> 118179Date: Thu Dec 17 18:24:37 2009 -0800 118180 118181 Convert checks for PC98 support from platform #ifdefs to configure flag 118182 118183 Default remains the same - on for most OS'es on i386 (except Solaris), 118184 off for everyone else. Can be manually toggled via --enable-pc98 or 118185 --disable-pc98. 118186 118187 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118188 Signed-off-by: Keith Packard <keithp@keithp.com> 118189 118190commit 801bc8075aee664bd4c6b6ff842ab737c143a1de 118191Author: Peter Hutterer <peter.hutterer@who-t.net> 118192Date: Mon Dec 14 08:43:19 2009 +1000 118193 118194 xfree86: update man page for special keys handling. 118195 118196 SpecialKeyHandling was removed from the kbd driver with version 1.4.0. Since 118197 this is the only version that will build against server 1.7+ it's not 118198 reasonable to mention it in the man page. Reword, point to XKB instead and 118199 make clear that some key combinations _may_ not be available in any given 118200 config. 118201 118202 Reported-by: Derek Fawcus 118203 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118204 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 118205 118206commit 753310837cd3812882d8de67f063bb61813db675 118207Author: Peter Hutterer <peter.hutterer@who-t.net> 118208Date: Mon Dec 14 08:41:18 2009 +1000 118209 118210 xfree86: DontZap has been disabled for a while now, say so in the man page. 118211 118212 1.7 always shipped with DontZap disabled, it's just the default keymaps that 118213 may not include the symbol to trigger it. 118214 118215 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118216 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 118217 118218commit 0cb638dc6822e54567a1731ea1cf588475a226e9 118219Merge: 6a6a041c2 fbdf493a3 118220Author: Keith Packard <keithp@keithp.com> 118221Date: Wed Dec 16 21:58:44 2009 -0800 118222 118223 Merge remote branch 'alanc/master' 118224 118225commit 6a6a041c2d8d32f6355db77cd59cff371038c683 118226Merge: 8127465f4 d50319550 118227Author: Keith Packard <keithp@keithp.com> 118228Date: Wed Dec 16 21:57:32 2009 -0800 118229 118230 Merge remote branch 'jeremyhu/master' 118231 118232commit 8127465f44fee99181f37a5e55dafc90bfafe3d8 118233Author: Ville Syrjälä <ville.syrjala@nokia.com> 118234Date: Mon Dec 14 19:19:01 2009 +0200 118235 118236 xf86xv: Fix off-by-one in viewport clipping 118237 118238 Most of the Xv Put/Get operations have an off by one error in the 118239 viewport clipping. 118240 118241 Apparently PutImage was fixed at some point but the same code was 118242 already copy-pasted all over the place, and so the other operations 118243 still suffer from the bug. 118244 118245 Signed-off-by: Ville Syrjälä <ville.syrjala@nokia.com> 118246 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 118247 Signed-off-by: Keith Packard <keithp@keithp.com> 118248 118249commit faca1bc582e374d32ee9d63d10e072fbef4940a3 118250Author: Peter Hutterer <peter.hutterer@who-t.net> 118251Date: Mon Dec 14 11:05:20 2009 +1000 118252 118253 xfree86: belately init RandR12 if xinerama fails. (#24627) 118254 118255 On Fri, Dec 11, 2009 at 10:19:01AM -0800, Keith Packard wrote: 118256 > On Wed, 9 Dec 2009 11:55:14 +1000, Peter Hutterer <peter.hutterer@who-t.net> wrote: 118257 > > On Tue, Dec 08, 2009 at 05:24:06PM -0800, Aaron Plattner wrote: 118258 > > > On Tue, Dec 08, 2009 at 03:52:27PM -0800, Peter Hutterer wrote: 118259 > > > > Xorg +xinerama crashes immediately due to whacky dependency between Xinerama 118260 > > > > and RandR12. The latter doesn't initialize if Xinerama is enabled, but if 118261 > > > > only one screen is found, Xinerama is disabled again and RandR12 tries to 118262 > > > > access data it never initialized. 118263 > 118264 > I'd sure like to have RandR get enabled when xinerama doesn't; is there 118265 > an easy way of making that happen here? Perhaps having the RandR12 code 118266 > disable Xinerama when only one screen is found? Or some other kludge? 118267 118268 you know the dependency better than I do so any hints are apreciated. 118269 afaict, the screenInfo.numScreens (the check used by Xinerama) isn't 118270 necessarily initialized at this point so we can't use the same check. 118271 The following seems to work though: 118272 118273 From 670b3ebdb7312a6433a8f093d0820785db2aea20 Mon Sep 17 00:00:00 2001 118274 From: Peter Hutterer <peter.hutterer@who-t.net> 118275 Date: Mon, 14 Dec 2009 11:00:58 +1000 118276 Subject: [PATCH] xfree86: if only one screen was found, disable Xinerama (#24627) 118277 118278 Xorg +xinerama crashes immediately due to whacky dependency between Xinerama 118279 and RandR12. The latter doesn't initialize if Xinerama is enabled, but if 118280 only one screen is found, Xinerama is disabled again and RandR12 tries to 118281 access data it never initialized. 118282 118283 Dependency chain is: 118284 - ProcessCommandLine sets noPanoramiXExtension to FALSE 118285 - xf86RandR12Init() is a noop 118286 - PanoramiXExtensionInit sets noPanoramiXExtension to TRUE 118287 - xf86RandR12CreateScreenResources tries to use the devPrivates key it never 118288 initialized. 118289 118290 This hack checks if there's only one screen at the time RandR12 is 118291 initialized. If so, we expect Xinerama to fail anyhow so we disable it 118292 ourselves and proceed as planned. 118293 118294 X.Org Bug 24627 <http://bugs.freedesktop.org/show_bug.cgi?id=24627> 118295 118296 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118297 Signed-off-by: Keith Packard <keithp@keithp.com> 118298 118299commit 5f898ddbaa7df01d2479e40238d8c7954fc4b67a 118300Author: Tiago Vignatti <tiago.vignatti@nokia.com> 118301Date: Fri Dec 11 17:04:14 2009 +0200 118302 118303 xfree86: fix -quiet option behaviour 118304 118305 Previously it was trying to set the same value as the default one. Sigh. 118306 118307 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 118308 Acked-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 118309 Signed-off-by: Keith Packard <keithp@keithp.com> 118310 118311commit 7284e198619cdacb38d0ad715e932acff86c8367 118312Author: Tiago Vignatti <tiago.vignatti@nokia.com> 118313Date: Fri Dec 11 17:04:13 2009 +0200 118314 118315 os: print log markers only if log level is >= 0 118316 118317 FWIW default log verbosity is 0, so this will affect only if one start the 118318 server with a different -verbose argument. 118319 118320 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 118321 Acked-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 118322 Signed-off-by: Keith Packard <keithp@keithp.com> 118323 118324commit f57240a0a8c05e0a2589d0127f8888db140d6974 118325Author: Tiago Vignatti <tiago.vignatti@nokia.com> 118326Date: Fri Dec 11 17:04:12 2009 +0200 118327 118328 xfree86: spam output but with verbose level checked instead 118329 118330 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 118331 Acked-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 118332 Signed-off-by: Keith Packard <keithp@keithp.com> 118333 118334commit fbdf493a3e7e48039a0826531e11aa33b04bf391 118335Author: Alan Coopersmith <alan.coopersmith@sun.com> 118336Date: Mon Dec 14 18:46:00 2009 -0800 118337 118338 Add freetype & type1 to the LoadModules ignore list 118339 118340 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118341 Acked-by: Adam Jackson <ajax@redhat.com> 118342 118343commit c9726bbe3148c7a8adb4dfde483741545d59c3dc 118344Author: Alan Coopersmith <alan.coopersmith@sun.com> 118345Date: Tue Dec 15 15:54:22 2009 -0800 118346 118347 Update Sun license notices to current X.Org standard form 118348 118349 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118350 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 118351 118352commit d50319550458f8127298cf8672b47f914b74ab1f 118353Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 118354Date: Fri Dec 11 16:42:43 2009 -0800 118355 118356 Xfake: Nuke -Wl,-undefined=InitExtensions from LDFLAGS 118357 118358 This is not portable, and accodring to Paulo should not be required any more. 118359 118360 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 118361 118362commit ed7d2527190ec56aed38db19d9e958abe88441b9 118363Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 118364Date: Fri Dec 11 11:53:28 2009 -0800 118365 118366 dtrace: Add Xserver-dtrace.h to CLEANFILES 118367 118368 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 118369 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 118370 118371commit 44f102ed64e2552a0e04714c7574f74ca5ee66e2 118372Author: Alan Coopersmith <alan.coopersmith@sun.com> 118373Date: Mon Nov 23 22:24:44 2009 -0800 118374 118375 COPYING: Update license preamble 118376 118377 Update the comment about "a modification is currently under discussion" 118378 since there hasn't been any such discussion for some time and this is 118379 the consensus agreed upon preferred format. 118380 118381 Also throw in a pointer to the OSI discussion page about the MIT license. 118382 118383 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118384 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 118385 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 118386 118387commit 00a7252fe3e519e3e4d62cdd8746833ca59bce77 118388Author: Alan Coopersmith <alan.coopersmith@sun.com> 118389Date: Fri Nov 20 13:38:30 2009 -0800 118390 118391 Fix dtrace object builds of libos to link with SHA1_LIBS 118392 118393 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118394 Acked-by: Adam Jackson <ajax@redhat.com> 118395 118396commit dd2c83d03ecedce8b546d583e436c12f92510754 118397Author: Alan Coopersmith <alan.coopersmith@sun.com> 118398Date: Wed Nov 11 14:42:20 2009 -0800 118399 118400 Fix builds with --with-int10=stub 118401 118402 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 118403 Acked-by: Adam Jackson <ajax@redhat.com> 118404 118405commit 77221c9155a3c96cb63b210a048db1ea71a5e359 118406Author: Adam Jackson <ajax@redhat.com> 118407Date: Thu Oct 29 14:01:29 2009 -0400 118408 118409 glx: swrast can do GLX 1.4 too 118410 118411 Reviewed-by: Ian Romanick <ian.d.romanick@intel.com> 118412 Signed-off-by: Adam Jackson <ajax@redhat.com> 118413 Signed-off-by: Keith Packard <keithp@keithp.com> 118414 118415commit a72c65e9176c51de95db2fdbf4c5d946a4911695 118416Author: Keith Packard <keithp@keithp.com> 118417Date: Wed Dec 2 08:37:20 2009 -0800 118418 118419 fb: Adjust transform or composite coordinates for pixman operations 118420 118421 Windows (or even pixmaps, in some cases) may not sit at the origin of 118422 the containing pixmap, so any coordinates relative to the drawable 118423 must be adjusted. For destinations and untransformed sources, the 118424 operation coordinates are adjusted. For transformed sources, the 118425 transform matrix is adjusted. 118426 118427 Signed-off-by: Keith Packard <keithp@keithp.com> 118428 Acked-by: Soeren Sandmann <sandmann@daimi.au.dk> 118429 118430commit bd567061c8b84b268d9bbb01bc4d8981feefb862 118431Author: Keith Packard <keithp@keithp.com> 118432Date: Wed Dec 2 15:51:22 2009 -0800 118433 118434 Split fbGetDrawable into fbGetDrawablePixmap and fbGetPixmapBitsData 118435 118436 These two sub-macros each perform half of the original macro work and 118437 the old macro is now implemented in terms of the new ones. This makes 118438 way for new code which wants to know the underlying pixmap for a 118439 window instead of just getting a pointer to the bits. 118440 118441 Signed-off-by: Keith Packard <keithp@keithp.com> 118442 Acked-by: Soeren Sandmann <sandmann@daimi.au.dk> 118443 118444commit 61335052972a78d67c0ba74f902273b34c63a198 118445Author: Keith Packard <keithp@keithp.com> 118446Date: Tue Dec 1 11:09:32 2009 -0800 118447 118448 Revert "Fix clipping when windows are used as sources" 118449 118450 This reverts commit e9aa61e9f0d663d5b34a397b943b4d1df44e873d. 118451 118452 Conflicts: 118453 118454 fb/fbpict.c 118455 118456 Signed-off-by: Keith Packard <keithp@keithp.com> 118457 Acked-by: Soeren Sandmann <sandmann@daimi.au.dk> 118458 118459commit 071b3c1810d9f2602173acc8578caac20e0b771e 118460Author: Keith Packard <keithp@keithp.com> 118461Date: Tue Dec 1 11:05:04 2009 -0800 118462 118463 Revert "Use IncludeInferiors when copying windows before compositing." 118464 118465 This reverts commit 7c7f0c2c6a04f7044d5ce69e97a615735e5831f1. 118466 118467 Signed-off-by: Keith Packard <keithp@keithp.com> 118468 Acked-by: Soeren Sandmann <sandmann@daimi.au.dk> 118469 118470commit 8e640d6b131d2865a9725d8997023865b0ef3d69 118471Author: Keith Packard <keithp@keithp.com> 118472Date: Tue Dec 1 11:04:49 2009 -0800 118473 118474 Revert "Reserve space for two GC values in copy_drawable()." 118475 118476 This reverts commit 08df24555cb432eb0d90a3f63275e9485e777c4c. 118477 118478 Signed-off-by: Keith Packard <keithp@keithp.com> 118479 Acked-by: Soeren Sandmann <sandmann@daimi.au.dk> 118480 118481commit 0816c5482c25251da8af5978cc348e1c547bf731 118482Author: Keith Packard <keithp@keithp.com> 118483Date: Tue Dec 1 11:04:06 2009 -0800 118484 118485 Revert "fb: Don't crash if copy_drawable() returns NULL." 118486 118487 This reverts commit 66a9616d645f5a23225251d197e00b94c79274f6. 118488 118489 Signed-off-by: Keith Packard <keithp@keithp.com> 118490 Acked-by: Soeren Sandmann <sandmann@daimi.au.dk> 118491 118492commit d3e054ac07dae12a82ad764b0622fddbef4b9ec5 118493Merge: 12fb31815 97a6454ea 118494Author: Keith Packard <keithp@keithp.com> 118495Date: Fri Dec 11 09:22:22 2009 -0800 118496 118497 Merge remote branch 'jeremyhu/master' 118498 118499commit 12fb31815db9de9c01f2d4155a2b74531777c0bf 118500Author: Peter Hutterer <peter.hutterer@who-t.net> 118501Date: Mon Dec 7 12:58:57 2009 +1000 118502 118503 xkb: don't assign garbage value to led_return. 118504 118505 As the comment for the function states, led_return is undefined if map is 118506 NULL. We might as well skip writing to it then. 118507 118508 Found by clang. 118509 118510 Reported-by: Tomas Carnecky <tom@dbservice.com> 118511 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118512 Reviewed-by: Jamey Sharp <jamey@minilop.net> 118513 118514commit ea14a418214ba4c58910f5650d8b25b3fc3a7099 118515Author: Peter Hutterer <peter.hutterer@who-t.net> 118516Date: Thu Dec 3 11:21:31 2009 +1000 118517 118518 Xi: reset device properties to NULL after deleting them. (#25374) 118519 118520 Commit 0e6cee853d8e5bef3274e632ef034d37f14674a9 introduced cleanup code to 118521 remove the accel properties when switching schemes. The same code is 118522 triggered by the default closedown code but only after unconditionally 118523 removing all device properties (as part of the cleanup). The properties, 118524 although deleted never got reset to NULL. 118525 118526 X.Org Bug 25374 <http://bugs.freedesktop.org/show_bug.cgi?id=25374> 118527 118528 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118529 Tested-by: Andy Furniss <lists@andyfurniss.entadsl.com> 118530 Reviewed-by: Dave Airlie <airlied@redhat.com> 118531 118532commit 9ac6a048efaec51f158bbc1a4a3f448cbe5658ad 118533Author: Peter Hutterer <peter.hutterer@who-t.net> 118534Date: Wed Dec 9 11:43:34 2009 +1000 118535 118536 xfree86: remove unused variable in configureDDMonitorSection. 118537 118538 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118539 Reviewed-by: Dave Airlie <airlied@redhat.com> 118540 118541commit 3fad969a90ccab604359ec4de05e745bb4b909e1 118542Author: Julien Cristau <jcristau@debian.org> 118543Date: Sun Dec 6 20:27:06 2009 +0100 118544 118545 Move config_init() after CreateWellKnownSockets() and InitCoreDevices() 118546 118547 config_init() can now add devices directly instead of scheduling a 118548 timer. 118549 118550 Signed-off-by: Julien Cristau <jcristau@debian.org> 118551 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 118552 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118553 118554commit 886140787852c660978a9bb3abc7cc2702cbaed1 118555Author: Tomas Carnecky <tom@dbservice.com> 118556Date: Mon Dec 7 05:54:57 2009 +0100 118557 118558 Fix possible NULL dereference in XkbFlushLedEvents() 118559 118560 Through some code paths it is possible that NULL is being passed in the 118561 'ed' parameter to XkbFlushLedEvents(). Make sure we don't pass it along 118562 to bzero(). 118563 118564 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 118565 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 118566 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118567 118568commit 92d9cb7e1340b259294ba317ea3a7d18487498b5 118569Author: Tomas Carnecky <tom@dbservice.com> 118570Date: Sun Dec 6 23:52:54 2009 +0100 118571 118572 XkbWriteCountedString(): return early if str is NULL 118573 118574 This avoids NULL from being passed to memcpy() later in the code. While 118575 that wasn't an issue before - that value being NULL implied 'size == 0' 118576 so memcpy() wouldn't try to dereference it - it made the code harder 118577 to read and also confused clang. 118578 118579 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 118580 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 118581 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118582 118583commit df2635d4644d1fa2772ff58d8d34cdf788ece412 118584Author: Michel Dänzer <daenzer@vmware.com> 118585Date: Thu Dec 10 00:04:53 2009 +0100 118586 118587 EXA/mixed: Don't consider devKind for detecting dimension change. 118588 118589 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=25536 . 118590 118591 There should be no reason for the pitch to actually change when none of the 118592 other dimensions do, and this could lead to incorrectly freeing the system 118593 memory copy when allocating a GPU copy, resulting in loss of valid pixmap 118594 contents. 118595 118596 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 118597 Acked-by: Maarten Maathuis <madman2003@gmail.com> 118598 Signed-off-by: Keith Packard <keithp@keithp.com> 118599 118600commit 97a6454ea57587db490873fee8ff0f899882972d 118601Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 118602Date: Wed Dec 9 18:27:31 2009 -0800 118603 118604 XQuartz: pbproxy: Fix building of standalone xpbproxy executable 118605 118606 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 118607 118608commit 1b8f90aa8b1017c74b254db2f398131b857323d4 118609Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 118610Date: Tue Dec 8 15:50:28 2009 -0800 118611 118612 Miscellaneous compilation warning fixes 118613 118614 main.c:134: warning: no previous prototype for 'dix_main' 118615 rootlessScreen.c: In function 'RootlessMarkOverlappedWindows': 118616 rootlessScreen.c:434: warning: function declaration isn't a prototype 118617 backtrace.c:51: warning: format '%lx' expects type 'long unsigned int', but argument 5 has type 'int' 118618 backtrace.c:54: warning: format '%lx' expects type 'long unsigned int', but argument 5 has type 'int' 118619 set.c: In function 'RecordSetMemoryRequirements': 118620 set.c:413: warning: old-style function definition 118621 set.c: In function 'RecordCreateSet': 118622 set.c:425: warning: old-style function definition 118623 stub.c: In function ‘main’: 118624 stub.c:236: warning: ISO C90 forbids mixed declarations and code 118625 118626 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 118627 118628commit 955b9f23a34cc79a5cd9676b45b3df4ffcc7302b 118629Author: Michel Dänzer <daenzer@vmware.com> 118630Date: Tue Dec 8 20:25:02 2009 +0100 118631 118632 EXA: ModifyPixmapHeader_mixed fixes. 118633 118634 * Better detection of dimension changes. 118635 * Make sure to re-create the system memory copy when the pixmap dimensions 118636 change (e.g. the screen pixmap on screen resize). 118637 * Clear the valid regions. 118638 118639 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 118640 Acked-by: Maarten Maathuis <madman2003@gmail.com> 118641 Signed-off-by: Keith Packard <keithp@keithp.com> 118642 118643commit c1503861cf75654d4f7b22e6f7f6487c47a0a395 118644Author: Michel Dänzer <daenzer@vmware.com> 118645Date: Tue Dec 8 20:25:01 2009 +0100 118646 118647 EXA: ExaDoPrepareAccess return value fixes. 118648 118649 Only return TRUE if the GPU copy is being accessed, and preserve the return 118650 value on repeated / nested calls for the same pixmap. 118651 118652 exaPrepareAccessReg_mixed could get inconsistent return values e.g. when the 118653 same pixmap is both the destination and source of an operation, potentially 118654 resulting in a crash. 118655 118656 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 118657 Acked-by: Maarten Maathuis <madman2003@gmail.com> 118658 Signed-off-by: Keith Packard <keithp@keithp.com> 118659 118660commit fd867387335b6175d76bbe93118bbe5e1e45ce88 118661Author: Maarten Maathuis <madman2003@gmail.com> 118662Date: Sat Dec 5 03:18:31 2009 +0100 118663 118664 exa/mixed: pixmaps that succeed prepare access have no need for a cpu copy 118665 118666 - When they have a gpu copy ofcource. 118667 - Use the presence of a cpu copy as a hint to fall back instead of UTS'ing in 118668 exaHWCopyNtoN. 118669 118670 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 118671 Acked-by: Michel Dänzer <michel@daenzer.net> 118672 Signed-off-by: Keith Packard <keithp@keithp.com> 118673 118674commit bb7acfbcfbc37869c2215c26791c6175a5a6c526 118675Author: Michel Dänzer <daenzer@vmware.com> 118676Date: Fri Dec 4 08:37:14 2009 +0100 118677 118678 EXA: Use correct coordinate system for calculating Composite fallback region. 118679 118680 Fixes incorrectly skipped rendering of some Composite operations to windows. 118681 118682 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 118683 Reviewed-by: Keith Packard <keithp@keithp.com> 118684 Signed-off-by: Keith Packard <keithp@keithp.com> 118685 118686commit 0f4ef7123d3e6e09e04dc55e8edb47aecf017648 118687Author: Maarten Maathuis <madman2003@gmail.com> 118688Date: Thu Dec 3 19:28:13 2009 +0100 118689 118690 exa/mixed: setting devKind before exaCopyDirty* is not needed 118691 118692 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 118693 Acked-by: Michel Dänzer <michel@daenzer.net> 118694 Signed-off-by: Keith Packard <keithp@keithp.com> 118695 118696commit 8ea415d417b3ef6b8a288d10da76ff4bc334e08b 118697Author: Maarten Maathuis <madman2003@gmail.com> 118698Date: Wed Dec 2 20:24:02 2009 +0100 118699 118700 Revert "exa: a few small pitch related changes" 118701 118702 This reverts commit 99d88ef69d5f7dbf99ca605eceb92f42230a89f4. 118703 118704 - Some pixmaps under classic have a sys_pitch which is 0, no idea why. This is 118705 causing rendering corruption. 118706 118707 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 118708 Acked-by: Michel Dänzer <michel@daenzer.net> 118709 Signed-off-by: Keith Packard <keithp@keithp.com> 118710 118711commit 98c8b752254a27ab1aaf881b36bfda0f74929d0a 118712Merge: 0e084d8c7 91c1bd78f 118713Author: Keith Packard <keithp@keithp.com> 118714Date: Wed Dec 2 15:28:07 2009 -0800 118715 118716 Merge remote branch 'whot/master' 118717 118718commit 91c1bd78f7240c92702828f8e5a6b6ce944b9e36 118719Author: Gaetan Nadon <memsize@videotron.ca> 118720Date: Sat Nov 28 21:32:47 2009 -0500 118721 118722 configure.ac: error while checking for XDMXCONFIG_DEP 118723 118724 Introduced in commit 9998105a387e0294054502331a56e1e020cd93e4 118725 The replacement third parameters to PKG_CHECK_MODULES([DMXMODULES] 118726 was not quoted. 118727 118728 Signed-off-by: Gaetan Nadon <memsize@videotron.ca> 118729 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 118730 Tested-by: Julien Cristau <jcristau@debian.org> 118731 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118732 118733commit 761ae22f880bd79550ccf93d321b8a28b3755956 118734Author: Peter Hutterer <peter.hutterer@who-t.net> 118735Date: Tue Dec 1 14:38:19 2009 +1000 118736 118737 xfree86: tell users to disable AutoAddDevices, not AllowEmptyInput. 118738 118739 Technically, disabling AEI is the right suggestion. AEI off forces the 118740 server to init the built-in defaults for input devices (or pick the first 118741 one from the config file). At the same time, hotplugging is still available 118742 with AEI off. 118743 118744 Unfortunatly, in the vast majority of cases users want to simply disable 118745 hotplugging or have a working server while the local HAL configuration is 118746 broken or missing. Disabling AEI will lead to duplicate events, triple 118747 keystrokes, etc. once the configuration works again. 118748 It's not actually required to remove AEI once hotplugging works again, 118749 though it will in many cases lead to a setup that appears broken. 118750 118751 Asking users to disable AutoAddDevices instead means those users disable 118752 hotplugging, can then fix the HAL setup and they _must_ remove the config 118753 line again to test if hotplugging works again. Which doesn't leave them with 118754 a broken config once everything is working nice and dandy. Less bugreports, 118755 everybody wins. 118756 118757 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118758 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 118759 Acked-by: Daniel Stone <daniel@fooishbar.org> 118760 Acked-by: Rémi Cardona <remi@gentoo.org> 118761 Acked-by: James Cloos <cloos@jhcloos.com> 118762 118763commit b584c224a888c9e7f92d7e49021f74232a727c7f 118764Author: Peter Hutterer <peter.hutterer@who-t.net> 118765Date: Fri Oct 30 12:11:41 2009 +1000 118766 118767 Set the source and deviceid for key repeat events (#24785) 118768 118769 X.Org Bug 24785 <http://bugs.freedesktop.org/show_bug.cgi?id=24785> 118770 118771 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118772 Acked-by: Adam Jackson <ajax@redhat.com> 118773 118774commit 66bb8c6fbdfc0fc0d971aac4ec6f949bb9288c1b 118775Author: Peter Hutterer <peter.hutterer@who-t.net> 118776Date: Fri Nov 27 16:20:13 2009 +1000 118777 118778 dix: remove core devices when shutting down. (#25028) 118779 118780 NewInputDeviceRequest (and RemoveDevice) have checks in place to not allow 118781 removal of the VCP/VCK. When shutting down, they need to be cleaned up 118782 nonetheless to free the memory associated. 118783 118784 X.Org Bug 25028 <http://bugs.freedesktop.org/show_bug.cgi?id=25028> 118785 118786 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118787 Reviewed-by: Keith Packard <keithp@keithp.com> 118788 118789commit 83d90b90bcb71c89750f92a177361e53dd261414 118790Author: Peter Hutterer <peter.hutterer@who-t.net> 118791Date: Fri Nov 27 16:08:44 2009 +1000 118792 118793 dix: remove some obsolete comment. 118794 118795 The "counterpart to biggest hack" included checking for the motion history 118796 function - which is unified in 1.7. Hence the check (which is already 118797 removed) would evaluate to true anyway, and this comment isn't needed. 118798 118799 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118800 Reviewed-by: Keith Packard <keithp@keithp.com> 118801 118802commit c20c8897272427cb3f755a3e28e80a9ad46f08a1 118803Author: Peter Hutterer <peter.hutterer@who-t.net> 118804Date: Fri Nov 27 16:01:53 2009 +1000 118805 118806 dix: fix memory leak, free event list on shutdown. (#25028) 118807 118808 X.Org Bug 25028 <http://bugs.freedesktop.org/show_bug.cgi?id=25028> 118809 118810 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118811 Reviewed-by: Keith Packard <keithp@keithp.com> 118812 118813commit 0e084d8c71e697a5cf5d5d7c749455ae14bd6eb7 118814Author: Luc Verhaegen <libv@skynet.be> 118815Date: Wed Nov 11 15:59:27 2009 +0100 118816 118817 Xv: Fix AdjustFrame when driver implements ReputImage. 118818 118819 Should probably also be applied to stabler xserver branches too. 118820 118821 Luc Verhaegen. 118822 118823 From a22bc20721bad506d8fa9772b1258568cbffe7d2 Mon Sep 17 00:00:00 2001 118824 From: Luc Verhaegen <libv@skynet.be> 118825 Date: Wed, 11 Nov 2009 15:52:39 +0100 118826 Subject: [PATCH] Xv: Fix AdjustFrame when driver implements ReputImage. 118827 118828 Finally fixes fd.o #4653, filed more than 4 years ago. 118829 118830 Patch can be happily applied to all modular Xorg versions. 118831 118832 Signed-off-by: Luc Verhaegen <libv@skynet.be> 118833 Reviewed-by: Keith Packard <keithp@keithp.com> 118834 Signed-off-by: Keith Packard <keithp@keithp.com> 118835 118836commit f4fc3406720410e37a2bce1b782cba0f0b734e42 118837Author: Jeremy Huddleston <jeremyhu@apple.com> 118838Date: Mon Nov 30 11:03:59 2009 -0800 118839 118840 XQuartz: Drop calls to alloca 118841 118842 This makes us more consistent with the rest of the codebase, using xalloc/xfree 118843 118844 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 118845 Signed-off-by: Keith Packard <keithp@keithp.com> 118846 118847commit 66a9616d645f5a23225251d197e00b94c79274f6 118848Author: Michel Dänzer <daenzer@vmware.com> 118849Date: Mon Nov 30 13:17:53 2009 +0100 118850 118851 fb: Don't crash if copy_drawable() returns NULL. 118852 118853 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=24634 . 118854 118855 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 118856 Signed-off-by: Keith Packard <keithp@keithp.com> 118857 118858commit 8754db77d8169e5ea506a963cebee1a651bcf094 118859Author: Michel Dänzer <daenzer@vmware.com> 118860Date: Mon Nov 30 13:17:52 2009 +0100 118861 118862 EXA: Don't defragment offscreen memory at allocation time. 118863 118864 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=24300 . 118865 118866 Offscreen memory allocation can occur from various places, and apparently 118867 doing defragmentation from at least some of them can confuse some driver 118868 acceleration code. 118869 118870 There's still the regular background defragmentation in the WakeupHandler, 118871 which should manage to keep fragmentation at a reasonable level. 118872 118873 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 118874 Signed-off-by: Keith Packard <keithp@keithp.com> 118875 118876commit 0e555a1033e5deed1db8582ca075455a6a2d4228 118877Author: Michel Dänzer <daenzer@vmware.com> 118878Date: Mon Nov 30 13:17:51 2009 +0100 118879 118880 Revert "EXA: Accumulate arbitrary number of glyphs without flushing." 118881 118882 This reverts commit c11678cc189551f2a01eaa7a63969c16950739b4. 118883 118884 Not sure what I was thinking, turns out alloca() of a size derived from client 118885 input is a bad idea. 118886 118887 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 118888 Signed-off-by: Keith Packard <keithp@keithp.com> 118889 118890commit 99d88ef69d5f7dbf99ca605eceb92f42230a89f4 118891Author: Maarten Maathuis <madman2003@gmail.com> 118892Date: Sat Nov 28 10:34:45 2009 +0100 118893 118894 exa: a few small pitch related changes 118895 118896 - Setting pitch before exaCopyDirty* is not needed. 118897 118898 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 118899 Acked-by: Michel Dänzer <michel@daenzer.net> 118900 Signed-off-by: Keith Packard <keithp@keithp.com> 118901 118902commit a54c23fe647cb4d610d871094193ae5959606008 118903Author: Maarten Maathuis <madman2003@gmail.com> 118904Date: Sat Nov 28 10:34:44 2009 +0100 118905 118906 exa: s/PixmapIsOffscreen/PixmapHasGpuCopy and s/pExaPixmap->offscreen/pExaPixmap->use_gpu_copy 118907 118908 - Fixup some variable names as well. 118909 118910 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 118911 Acked-by: Michel Dänzer <michel@daenzer.net> 118912 Signed-off-by: Keith Packard <keithp@keithp.com> 118913 118914commit 342f3689d17256c92cbfee079d24501d27aa1153 118915Author: Maarten Maathuis <madman2003@gmail.com> 118916Date: Sat Nov 28 10:34:43 2009 +0100 118917 118918 exa: handle pixmap create/destroy in lower layers 118919 118920 - Pixmaps that are created during a fallback are automatically prepared access. 118921 - During the fallback accelerated ops are blocked to prevent new/scratch gc's 118922 from triggering accelerated ops on mapped pixmaps. 118923 - A few cases of incorrect wrapping (on the top level pointer instead of 118924 between damage and (w)fb) have been removed. 118925 118926 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 118927 Acked-by: Michel Dänzer <michel@daenzer.net> 118928 Signed-off-by: Keith Packard <keithp@keithp.com> 118929 118930commit b54bc14ce0ae38c4863794bc3096ca86cdb23908 118931Author: Ingmar Vanhassel <ingmar@exherbo.org> 118932Date: Wed Nov 25 20:31:48 2009 +0100 118933 118934 Update man-pages for new default font paths 118935 118936 Signed-off-by: Ingmar Vanhassel <ingmar@exherbo.org> 118937 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 118938 Signed-off-by: Keith Packard <keithp@keithp.com> 118939 118940commit c0367028ac02b3ccfbc245863dc6b7295a4014d6 118941Author: Ingmar Vanhassel <ingmar@exherbo.org> 118942Date: Wed Nov 25 20:31:47 2009 +0100 118943 118944 Update COMPILEDDEFAULTFONTPATH to match the new default font path 118945 118946 Signed-off-by: Ingmar Vanhassel <ingmar@exherbo.org> 118947 Reviewed-by: Alan Coopersmith <alan.coopersmith@sun.com> 118948 Signed-off-by: Keith Packard <keithp@keithp.com> 118949 118950commit 0e6cee853d8e5bef3274e632ef034d37f14674a9 118951Author: Peter Hutterer <peter.hutterer@who-t.net> 118952Date: Wed Nov 18 14:39:25 2009 +1000 118953 118954 dix: clean up accel old scheme data when switching schemes. 118955 118956 InitValuatorClassDeviceStruct always initializes with the default profile. 118957 The default profile allocs data and adds a few properties which become 118958 obsolete if the profile is changed lateron by the driver. 118959 118960 The property handlers are stored in the device's devPrivates and cleaned up. 118961 Ideally, the property handler ID's could be stored somewhere more obvious, 118962 but that seems to require breaking the ABI. 118963 118964 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118965 Acked-by: Simon Thum <simon.thum@gmx.de> 118966 118967commit 1b127ab8429616adf9ec31ba4d8bdd9af6e104a9 118968Author: Peter Hutterer <peter.hutterer@who-t.net> 118969Date: Tue Nov 24 15:31:48 2009 +1000 118970 118971 Xi: when deleting all properties, reset property handler to NULL. 118972 118973 Trying to unregister property handlers during the device closure process 118974 leads to invalid memory accesses. 118975 118976 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118977 Reviewed-by: Keith Packard <keithp@keithp.com> 118978 118979commit 8806375ed72a3cd465fe0a49ead079a334accd6b 118980Author: Peter Hutterer <peter.hutterer@who-t.net> 118981Date: Tue Nov 17 09:25:47 2009 +1000 118982 118983 Move xdmxconfig modules into DMX conditionals (#25102) 118984 118985 xdmxconfig requires additional modules not checked for if Xdmx build is set 118986 to auto (the default). This may lead to build errors if the Xdmx modules are 118987 installed, but not the extra ones required for xdmxconfig. 118988 118989 X.Org Bug 25102 <http://bugs.freedesktop.org/show_bug.cgi?id=25102> 118990 118991 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 118992 Reviewed-by: Remi Cardona <remi@gentoo.org> 118993 Acked-by: Julien Cristau <jcristau@debian.org> 118994 118995commit a2adda7ab3defd953cf0f48f5372efb037786a9e 118996Author: Peter Hutterer <peter.hutterer@who-t.net> 118997Date: Thu Nov 19 16:35:55 2009 +1000 118998 118999 Revert "Move xdmxconfig modules into DMX conditionals (#25102)" 119000 119001 New package dependency unnecessarily links in a few libraries that Xdmx 119002 doesn't need. This can be fixed more elegantly. 119003 119004 This reverts commit 0ef15ca9d2d9c78c79a2771c550563bc6931b365. 119005 119006 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 119007 119008commit eb967ca36cfe0409972ac987a74d498cd6f6bafb 119009Author: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 119010Date: Tue Nov 17 17:13:43 2009 +0200 119011 119012 Xext: Fix a memory leak on GE client disconnect. 119013 119014 Add a call to dixRequestPrivate to inform dixFreePrivates that memory 119015 allocated in GEClientCallback should be released when client 119016 disconnects. Otherwise there is a leak of sizeof(GEClientInfoRec) for 119017 every client connect/disconnect. 119018 119019 Also remove the explicit allocation and let GEGetClient / 119020 dixLookupPrivate do it. This makes GEClientCallback similar to the 119021 other extension callbacks. 119022 119023 Signed-off-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 119024 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 119025 119026commit 55f60e192318132157d3dfdd6732f141bd9dc352 119027Author: Maarten Maathuis <madman2003@gmail.com> 119028Date: Mon Nov 23 22:17:43 2009 +0100 119029 119030 Revert "exa/mixed: be more thorough about setting fb_pitch when needed" 119031 119032 This reverts commit d4fc245115eb2cb323e06a82f9dd52518d9b6a16. 119033 119034 - This is causing crashes/problems for some. 119035 119036 Acked-by: Michel Dänzer <michel@daenzer.net> 119037 Signed-off-by: Keith Packard <keithp@keithp.com> 119038 119039commit 4677b5a80025b50ba2a3e953fd487a549586ae9f 119040Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119041Date: Mon Nov 23 16:33:00 2009 -0800 119042 119043 XQuartz: Allow better compatability with older versions of xinit 119044 119045 If we are id="org.x" and the launchd socket is ":0", we will claim 119046 the socket to match the old behavior before we prefixed the 119047 socket name with our id. 119048 119049 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119050 119051commit ec16357edaba23694b12fd234ffa12fd58a73dfa 119052Author: Tiago Vignatti <tiago.vignatti@nokia.com> 119053Date: Mon Nov 23 17:00:06 2009 +0200 119054 119055 configure: remove unused kdrive Xvesa config variable 119056 119057 Xvesa was gone in commit 6d21fbf0 and this should be there as well. 119058 119059 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 119060 Signed-off-by: Keith Packard <keithp@keithp.com> 119061 119062commit a30e739a144912a68adcaa9f426d600c6ecbd529 119063Author: Peter Hutterer <peter.hutterer@who-t.net> 119064Date: Mon Nov 23 13:07:56 2009 +1000 119065 119066 Xi: don't crash when deleting invalid device properties. 119067 119068 Deleting a property that was not set on a device leads to a null-pointer 119069 reference. The protocol allows deleting those properties - it has to be a 119070 noop. 119071 119072 Reproducible: 119073 xinput --set-prop "My device" --type=int --format=8 "my property" 1 119074 xinput --delete-prop "My other device" "my property" 119075 119076 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 119077 Signed-off-by: Keith Packard <keithp@keithp.com> 119078 119079commit 19f7c15e2008dab3c46ba3e14dfa353d01c74f72 119080Author: Zhao Yakui <yakui.zhao@intel.com> 119081Date: Fri Nov 20 14:43:35 2009 +0800 119082 119083 xfree86: Edid quirk for Philips LCD LP154W01 119084 119085 v1->v2: Make one condition case for one quirk instead of merging them 119086 together. This is based on the Keithp's suggestion. 119087 119088 Move the EDID quirk for Philips LCD LP154W01 as the panel reports the vertical 119089 size in cm. 119090 119091 https://bugs.freedesktop.org/show_bug.cgi?id=24482 119092 119093 Signed-off-by: Zhao Yakui <yakui.zhao@intel.com> 119094 Signed-off-by: Keith Packard <keithp@keithp.com> 119095 119096commit b8b8db98353760738eead612fe846c2ce1c174fa 119097Author: Michel Dänzer <daenzer@vmware.com> 119098Date: Thu Nov 19 10:46:30 2009 +0100 119099 119100 EXA: Don't use UploadToScreen for CopyNtoN with mixed pixmaps. 119101 119102 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 119103 119104commit d4fc245115eb2cb323e06a82f9dd52518d9b6a16 119105Author: Maarten Maathuis <madman2003@gmail.com> 119106Date: Wed Nov 18 21:23:09 2009 +0100 119107 119108 exa/mixed: be more thorough about setting fb_pitch when needed 119109 119110 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 119111 Acked-by: Michel Dänzer <michel@daenzer.net> 119112 119113commit 647b79f87a9891225678dc6fc2fbda3bdef8fa9d 119114Author: Maarten Maathuis <madman2003@gmail.com> 119115Date: Wed Nov 18 21:23:08 2009 +0100 119116 119117 exa/mixed: avoid copying back pixmap data when no migration took place 119118 119119 - When the driver handles the prepare access no copying is needed. 119120 - Delayed pixmap creation should be fine, because it's handled by the 119121 first prepare access, but the exaPixmapIsOffscreen check in finish access 119122 will return FALSE without a driver pixmap. 119123 119124 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 119125 Acked-by: Michel Dänzer <michel@daenzer.net> 119126 119127commit 30be7ceaf228497ac1ff0a1123c1b35e3aa1fc73 119128Author: Julien Cristau <jcristau@debian.org> 119129Date: Sat Nov 14 18:39:00 2009 +0100 119130 119131 xfree86: set a sane umask before opening the log 119132 119133 Xorg creates its log file following the umask of the user running 119134 startx, which may result in a world-writable log. Set umask to 022 to 119135 prevent this. 119136 119137 Debian bug#555308 <http://bugs.debian.org/555308> 119138 See also http://thread.gmane.org/gmane.comp.security.oss.general/2299 119139 119140 Signed-off-by: Julien Cristau <jcristau@debian.org> 119141 Reviewed-by: Adam Jackson <ajax@redhat.com> 119142 Signed-off-by: Keith Packard <keithp@keithp.com> 119143 119144commit fb95090730360b6b7f5429c40937e8fc9dfe5c14 119145Author: Matt Turner <mattst88@gmail.com> 119146Date: Mon Nov 9 03:33:45 2009 +0000 119147 119148 Use glibc's in/out routines 119149 119150 Let's let glibc do the right thing for dense/sparse selection. 119151 119152 The _alpha_iobase code has been unused since the switch to libpciaccess. It 119153 really should have been killed by fba700f1f6a8976. 119154 119155 Signed-off-by: Matt Turner <mattst88@gmail.com> 119156 Tested-by: Michael Cree <mcree@orcon.net.nz> 119157 Signed-off-by: Keith Packard <keithp@keithp.com> 119158 119159commit 0ef15ca9d2d9c78c79a2771c550563bc6931b365 119160Author: Peter Hutterer <peter.hutterer@who-t.net> 119161Date: Tue Nov 17 11:48:35 2009 +1000 119162 119163 Move xdmxconfig modules into DMX conditionals (#25102) 119164 119165 xdmxconfig requires additional modules not checked for if Xdmx build is set 119166 to auto (the default). This may lead to build errors if the Xdmx modules are 119167 installed, but not the extra ones required for xdmxconfig. 119168 119169 X.Org Bug 25102 <http://bugs.freedesktop.org/show_bug.cgi?id=25102> 119170 119171 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 119172 Tested-by: Adrian Bunk <bunk@stusta.de> 119173 Signed-off-by: Keith Packard <keithp@keithp.com> 119174 119175commit df95be4d3d9a744720e37a16fd89d9569bf6e188 119176Author: Paulo Ricardo Zanoni <pzanoni@mandriva.com> 119177Date: Thu Nov 12 11:27:34 2009 -0200 119178 119179 configure: change 'sigio-default' to 'use-sigio-by-default' 119180 119181 This was discussed in this thread: 119182 http://lists.x.org/archives/xorg-devel/2009-September/002025.html 119183 The patch sent fixed the help string but not the configure option. 119184 119185 Signed-off-by: Keith Packard <keithp@keithp.com> 119186 119187commit 801f0e59800ea4a3f6c85b22dea6f2c400bda703 119188Author: Gaetan Nadon <memsize@videotron.ca> 119189Date: Wed Oct 28 14:09:11 2009 -0400 119190 119191 INSTALL, NEWS, README or AUTHORS files are missing/incorrect #24206 119192 119193 Add missing INSTALL file. Use standard GNU file on building tarball 119194 README may have been updated 119195 Remove AUTHORS file as it is empty and no content available yet. 119196 Remove NEWS file as it is empty and no content available yet. 119197 119198 Reviewed-By: Jeremy Huddleston <jeremyhu@apple.com> 119199 Signed-off-by: Keith Packard <keithp@keithp.com> 119200 119201commit c93f516e850c2ef2b47cc2e282f9081ae7165ac7 119202Author: Gaetan Nadon <memsize@videotron.ca> 119203Date: Tue Oct 27 15:07:26 2009 -0400 119204 119205 Deploy the new XORG_DEFAULT_OPTIONS #24242 119206 119207 This macro aggregate a number of existing macros that sets commmon 119208 X.Org components configuration options. It shields the configuration file from 119209 future changes. 119210 119211 Reviewed-By: Jeremy Huddleston <jeremyhu@apple.com> 119212 Signed-off-by: Keith Packard <keithp@keithp.com> 119213 119214commit 0f38159cdd154a29caab0342aee41378e9921929 119215Author: Gaetan Nadon <memsize@videotron.ca> 119216Date: Tue Nov 10 09:12:27 2009 -0500 119217 119218 Makefile.am: ChangeLog not required: EXTRA_DIST or *CLEANFILES #24432 119219 119220 ChangeLog filename is known to Automake and requires no further 119221 coding in the makefile. 119222 119223 Reviewed-By: Jeremy Huddleston <jeremyhu@apple.com> 119224 Signed-off-by: Keith Packard <keithp@keithp.com> 119225 119226commit c739beb439bb813e4c82b3216daee7e89aa96f2f 119227Author: Gaetan Nadon <memsize@videotron.ca> 119228Date: Tue Nov 10 08:56:21 2009 -0500 119229 119230 .gitignore: use common defaults with custom section #24239 119231 119232 Using common defaults will reduce errors and maintenance. 119233 Only the very small or inexistent custom section need periodic maintenance 119234 when the structure of the component changes. Do not edit defaults. 119235 119236 Reviewed-By: Jeremy Huddleston <jeremyhu@apple.com> 119237 Signed-off-by: Keith Packard <keithp@keithp.com> 119238 119239commit 9071b0d69748cfa7ecca17b4cb0e431bbb0ef2a4 119240Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119241Date: Mon Nov 9 13:09:55 2009 -0800 119242 119243 XQuartz: Explicitly pass a bellProc to make XBell() work again. 119244 119245 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119246 119247commit e87a03fd531ce7974877688680d3bf9b85c2d835 119248Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119249Date: Tue Oct 27 17:16:57 2009 -0700 119250 119251 XQuartz: pbproxy: Wait for the server to finish starting up, so display is valid. 119252 119253 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119254 119255commit 15fc56addcc906592af7c0f4c0a5ac906fa5c389 119256Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119257Date: Sun Nov 8 20:25:42 2009 -0800 119258 119259 XQuartz: Buildfix for Leopard and older 119260 119261 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119262 119263commit b5aa2e0a5fe233dc883084a5026013472e85bca4 119264Author: Chase Douglas <chasedouglas@gmail.com> 119265Date: Mon Nov 9 22:54:39 2009 -0500 119266 119267 Move FD_CLR above pInfo->read_input 119268 119269 The event fd may be invalidated by the pInfo->read_input call. If it is 119270 invalidated, the subsequent FD_CLR call will segfault. Thus, the FD_CLR 119271 call must precede the pInfo->read_input call. 119272 119273 Signed-off-by: Chase Douglas <chasedouglas@gmail.com> 119274 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 119275 Signed-off-by: Keith Packard <keithp@keithp.com> 119276 119277commit 982f6648fd29d085265bf6035c1bf4d1b2499316 119278Author: Peter Hutterer <peter.hutterer@who-t.net> 119279Date: Tue Nov 10 09:50:50 2009 +1000 119280 119281 dix: increase default number of buttons to 10. 119282 119283 Currently the XTEST device is limited to the same number of buttons the core 119284 device has. This breaks if a user has a mouse with more than 3 buttons 119285 connected and is using a core client to fake button 8+ presses. 119286 119287 Rather than expecting all clients to fix themselves, just increase the 119288 default number of buttons to 10, which is somewhat a compromise. Ideally, 119289 the XTEST devices should adjust themselves to the highest number of buttons 119290 available on the slave devices (like the master pointers already do), but 119291 that's a taks for another day. 119292 119293 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 119294 Signed-off-by: Keith Packard <keithp@keithp.com> 119295 119296commit ec5417b965616e60b51466fe9fd2b44cb5abb349 119297Author: Jon TURNEY <jon.turney@dronecode.org.uk> 119298Date: Sun Jul 12 23:58:32 2009 +0100 119299 119300 Cygwin/X: Ensure WM_STATE atom exists in multiwindow mode 119301 119302 Workaround a bug in iiimxcf (assuming the WM_STATE atom exists), 119303 which can cause many Solaris clients to simply fail with a BadAtom 119304 error 119305 119306 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119307 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 119308 119309commit aa860552fd7e2888258a7b48b8c3bd4af527dc6c 119310Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 119311Date: Sun Nov 1 18:19:08 2009 +0000 119312 119313 Cygwin/X: Enable clipboard integration by default 119314 119315 Enable clipboard integration by default, can be turned off with -noclipboard. 119316 We still accept -clipboard for backwards compatibility. If both are passed, 119317 the last one is accepted (just as other arguments are handled). 119318 119319 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119320 119321commit 87b00ced3d308a9168828c0e38ecffa0640621a0 119322Author: Jon TURNEY <jon.turney@dronecode.org.uk> 119323Date: Sun Nov 1 18:18:48 2009 +0000 119324 119325 Cygwin/X: Setup screen layout in Xinerama mode 119326 119327 Setup screen layout according to the screen window native window 119328 positions in Xinerama mode 119329 119330 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119331 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 119332 119333commit 0fe2e8c64e7a5a9140c77b88b266d7ae58b0bca2 119334Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 119335Date: Sun Nov 1 18:18:31 2009 +0000 119336 119337 Cygwin/X: Allow composite to be enabled for Xwin 119338 119339 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119340 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 119341 119342commit 26a69bbd9b6f415c443d6d6f8f450329f348edc3 119343Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 119344Date: Sun Nov 1 18:19:04 2009 +0000 119345 119346 Cygwin/X: Mount options have changed in cygwin-1.7 119347 119348 Mount options have changed in cygwin-1.7 119349 119350 Also fix a typo in the warning issued if /tmp is a textmode mount 119351 119352 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119353 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 119354 119355commit 35901ece6a49e5f8e245364c27723da2f9009a1d 119356Author: Paul Loewenstein <paul.loewenstein@gmail.com> 119357Date: Sun Nov 1 18:18:45 2009 +0000 119358 119359 Cygwin/X: Handle fake keypresses generated by speech recognizers 119360 119361 Apparently, fake keypresses generated by speech recognizers may not bother 119362 with a scan code, so look up what scan code corresponds to the virtual key 119363 code if this occurs. 119364 119365 Patch by Paul Loewenstein <paul.loewenstein@gmail.com> 119366 119367 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119368 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 119369 119370commit 0866322b574b1f3695467535ed5fa8f9e629ad1d 119371Author: Jon TURNEY <jon.turney@dronecode.org.uk> 119372Date: Mon Jul 13 13:46:45 2009 +0100 119373 119374 Cygwin/X: Always use an authorization cookie for internal clients 119375 119376 Don't conditionalize use of an authorization cookie for internal client 119377 threads on XCSECURITY, always use one (this avoids certain problems 119378 with XDMCP setups where the XDMCP host removes localhost from the access 119379 list etc.) 119380 119381 Conditionalize the use of a XCSECURITY authorization descriptor on XCSECURITY 119382 119383 Consolidate the various places where the authorization cookie is set 119384 for internal threads into a new function, winSetAuthorization() 119385 119386 Use authorization cookie for multiwindow WM X message thread as well 119387 119388 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119389 119390commit 72f81f4e449defb0654e9bdb2c9ec014475a3977 119391Author: Colin Harrison <colin.harrison@virgin.net> 119392Date: Fri Nov 6 17:46:58 2009 +0000 119393 119394 Xming: Make -auth option work in with -multiwindow 119395 119396 Use an internally generated cookie for authentication of the internal 119397 window manager client when using the -auth option in -multiwindow mode. 119398 119399 Copyright (C) Colin Harrison 2005-2008 119400 http://www.straightrunning.com/XmingNotes/ 119401 http://sourceforge.net/projects/xming/ 119402 119403 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119404 119405commit 6dbf8f27c885ea3a761548183fb00ba3df2df553 119406Author: Colin Harrison <colin.harrison@virgin.net> 119407Date: Tue Jul 21 01:15:25 2009 +0100 119408 119409 Xming: Tidy up code for initial native window positioning 119410 119411 Tidy up code for initial native window positioning and avoid a 119412 duplicate call to winMultiWindowGetTransientFor() 119413 119414 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119415 119416commit 28eb61fc04811bb9bbb523e0a252933313bb16ce 119417Author: Jon TURNEY <jon.turney@dronecode.org.uk> 119418Date: Sun Nov 1 18:19:14 2009 +0000 119419 119420 Avoid a null dereference if IFF_BROADCAST is set but there is no broadcast address 119421 119422 It seems that the getifaddrs() function can return interfaces with 119423 IFF_BROADCAST & IFF_UP set, but no broadcast address (at least 119424 under Cygwin 1.7, this seems to happen for v6 mapped v4 addresses) 119425 119426 Avoid a null dereference if this ever happens 119427 119428 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119429 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 119430 119431commit 091cbbaed7aafab6f9839a40fe977ab067d1ace1 119432Author: Colin Harrison <colin.harrison@virgin.net> 119433Date: Mon Oct 12 13:40:11 2009 +0100 119434 119435 Xming: Fix UT8String and CompoundText clipboard text sharing with windows clipboard 119436 119437 XConvertSelection() in libX11 always returns 1, so there is no point in 119438 testing it incorrectly against Success. This is possibly a bug in 119439 XConvertSelection() 119440 119441 This should fix UTF8String and CompoundText selection via the clipboard. 119442 119443 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119444 119445commit 9657eae5d65eff289089afc46dfb629758ebbdf7 119446Author: Jon TURNEY <jon.turney@dronecode.org.uk> 119447Date: Sun Nov 1 18:18:51 2009 +0000 119448 119449 Cygwin/X: Clearly diagnose a timeout while waiting for SelectionNotify event 119450 119451 Clearly diagnose a timeout while waiting for SelectionNotify event 119452 in the clipboard integration internal client. 119453 (which seems to be behind some of the reported failures) 119454 119455 Turn useless #if 0/ErrorF()/#endif into useful winDebug() 119456 119457 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119458 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 119459 119460commit 062f49a8e0d1afb4dd32e9451a47ab2792639e7f 119461Author: Jon TURNEY <jon.turney@dronecode.org.uk> 119462Date: Wed Nov 4 17:36:54 2009 +0000 119463 119464 Cygwin/X: Add a workaround for a SWT/Motif bug to internal window manager 119465 119466 SWT/Motif expects all top-level windows to get reparented, and waits until they 119467 do. So workaround that in our internal WM by forcing a reparent event to 119468 occur, even though we don't actually need to reparent the window to 119469 frame it (as the frame is a native window, not an X window) 119470 119471 http://sourceware.org/bugzilla/show_bug.cgi?id=9848 119472 https://bugs.eclipse.org/bugs/show_bug.cgi?id=36806 119473 119474 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119475 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 119476 119477commit 71519a572fe15b85c0eb2b02636c9e871f2c858f 119478Author: Jon TURNEY <jon.turney@dronecode.org.uk> 119479Date: Wed Nov 4 17:34:40 2009 +0000 119480 119481 Cygwin/X: Fix typo in g_fAnotherWMRunning and tidy up WM detection code 119482 119483 Tidy up code for detecting another WM is already running 119484 Fix typo g_fAnotherWMRunnig -> g_fAnotherWMRunning 119485 Remove some unused event mask macros 119486 119487 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119488 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 119489 119490commit f3fad371cce0f3836514ad5b29e59fa1ca0627a7 119491Author: Colin Harrison <colin.harrison@virgin.net> 119492Date: Wed Nov 4 15:33:53 2009 +0000 119493 119494 Xming: Add FORCEEXIT option to configuration file 119495 119496 Add a new option to configuration file: FORCEEXIT, like SILENTEXIT 119497 but ignores the client count. Unsaved client work may be lost with 119498 this option but it is useful if you want no dialogs. 119499 119500 Add description of this new keyword to XWinrc man page 119501 119502 Also fix grammar of the exit confirmation dialog warning to be correct 119503 when there is only one(1) client connected. 119504 119505 Also rearrange yacc tokens to one per line to make future merges 119506 easier 119507 119508 Also amend default system.XWinrc so that SILENTEXIT is on by default 119509 119510 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119511 119512commit 019a601de6a8ae083852c2d384c814dcc7aeeb72 119513Author: Colin Harrison <colin.harrison@virgin.net> 119514Date: Wed Nov 4 15:16:03 2009 +0000 119515 119516 Xming: Add mouse motion events with flag POINTER_SCREEN in winEnqueueMotion() 119517 119518 Add mouse motion events with flag POINTER_SCREEN in winEnqueueMotion(), as 119519 they are in screen coordindates and may need to be scaled to the axis range 119520 appropriately 119521 119522 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119523 119524commit 7af1240b57daa4ec55512276a9fda499d923eaa0 119525Author: Colin Harrison <colin.harrison@virgin.net> 119526Date: Wed Nov 4 14:20:17 2009 +0000 119527 119528 Xming: Replace all the uses of deprecated functions in hw/xwin with current ones 119529 119530 Replace uses of LookupIDByType() and SecurityLookupIDByType() 119531 with dixLookupResourceByType() 119532 119533 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119534 119535commit 2a38f7c0dbc890e6408eee143d77719f265d583d 119536Author: Jon TURNEY <jon.turney@dronecode.org.uk> 119537Date: Fri Jun 19 21:12:55 2009 +0100 119538 119539 Cygwin/X: Remove a couple of extraneous '\n' in logged version info 119540 119541 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119542 119543commit 56658fc06cb773ed20ef1b23a3fe918c0d3b36bb 119544Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 119545Date: Thu Oct 15 22:40:02 2009 -0500 119546 119547 Set default font path with fontutil 1.1 and fontpath.d 119548 119549 Look for $sysconfdir/X11/fontpath.d and, if it exists, set it to the 119550 default font path as a catalogue: entry. Based on app/xfs commit 119551 c66a46e35ae40a23ad9acee838ab42300eddbd67. 119552 119553 Otherwise, use the font-util 1.1 macros to determine the system 119554 fontrootdir and standard subdirectories, or let the user configure it 119555 and the subdirectories themselves. 119556 119557 This adds a build-time (or at least an autoconf/autogen time) dependency 119558 on font-util, and changes the --with-fontdir argument to --with-fontrootdir. 119559 119560 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 119561 Reviewed-by: Alan Coopersmith <alan.coopersmith-xsfywfwIY+M@public.gmane.org> 119562 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 119563 119564commit c7d1319872e0ae0b7d945e9af8c4027ee5aa2055 119565Author: Adam Jackson <ajax@redhat.com> 119566Date: Mon Nov 9 11:12:22 2009 -0500 119567 119568 randr: Fill in errorValue when verifying outputs/crtcs/modes 119569 119570 Signed-off-by: Adam Jackson <ajax@redhat.com> 119571 Signed-off-by: Keith Packard <keithp@keithp.com> 119572 119573commit d85ab6b6483d6ca7d9d22298d05ed1fa3076b042 119574Author: Dave Airlie <airlied@redhat.com> 119575Date: Tue Oct 13 10:15:00 2009 +1000 119576 119577 loader: actually stat something that has some chance of existing. (v2) 119578 119579 FindModuleInSubdir seems to expect a / at the end of the subdir its 119580 finding for, so we add the / early, the stat will fail if its 119581 not a subdir, I'm leaving the S_ISDIR in just in case there is another 119582 reason it could return 0. This does look a bit silly in strace 119583 but it seems to work fine. 119584 119585 I have a very intermittent issue where drivers loses its / that 119586 I've been seeing on/off for a while, this may or may not fix it. 119587 119588 Signed-off-by: Dave Airlie <airlied@redhat.com> 119589 Reviewed-by: Julien Cristau <jcristau@debian.org> 119590 Signed-off-by: Keith Packard <keithp@keithp.com> 119591 119592commit a5e59230de3abafda9cd9d571cea192897155a14 119593Author: Adam Jackson <ajax@redhat.com> 119594Date: Thu Nov 5 13:44:30 2009 -0800 119595 119596 randr: Turn on ModeDebug during server setup 119597 119598 Signed-off-by: Adam Jackson <ajax@redhat.com> 119599 Signed-off-by: Keith Packard <keithp@keithp.com> 119600 119601commit c61e77c77ca309f3676370992f41867807ed0c67 119602Author: Matt Turner <mattst88@gmail.com> 119603Date: Thu Nov 5 22:40:03 2009 -0800 119604 119605 Remove lnx_font.c and lnx.h 119606 119607 I couldn't find any version of the X xserver that ever used lnx_font.c 119608 so let's delete it. I tried contacting its author, Egbert, multiple 119609 times on IRC and email [*] but never got any response. It also hasn't 119610 been seriously touched since January 2005. 119611 119612 [*] http://lists.x.org/archives/xorg-devel/2009-October/002855.html 119613 119614 Signed-off-by: Matt Turner <mattst88@gmail.com> 119615 Reviewed-by: Tiago Vignatti <tiago.vignatti@nokia.com> 119616 Signed-off-by: Keith Packard <keithp@keithp.com> 119617 119618commit 15ffe9f51b122494a2b292b3ab1f199d3e81600c 119619Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119620Date: Thu Nov 5 18:28:28 2009 -0800 119621 119622 configure.ac: Notify user about which SHA1 implementation is being used 119623 119624 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119625 119626commit 6b109919f6e1593b27b0760bb56a65b43fb86ea4 119627Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119628Date: Thu Nov 5 18:13:07 2009 -0800 119629 119630 SHA1: Add support for Common Crypto 119631 119632 libSystem on darwin can handle SHA1 computation without needing to pull in 119633 OpenSSL. See CC_crypto(3) 119634 119635 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119636 119637commit 840a68dc5e3b4d285894f86df2a8c41fca5a4bec 119638Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119639Date: Wed Nov 4 18:39:16 2009 -0800 119640 119641 XQuartz: Cleanup X11Controller.m compilation warnings. 119642 119643 Declare X11Controller as implementing NSTableViewDataSource. 119644 Use selectRowIndexes:byExtendingSelection instead of selectRow:byExtendingSelection 119645 119646 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119647 119648commit 0e3ad44c3b05cbe7b1d8f5ce7949fec3c7ae2e7c 119649Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119650Date: Wed Nov 4 18:15:08 2009 -0800 119651 119652 XQuartz: Use dixLookupResourceByType instead of LookupIDByType 119653 119654 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119655 119656commit 069fc6ce0a0e7f2c418e11941568ffcc52b6b331 119657Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119658Date: Wed Nov 4 16:18:06 2009 -0800 119659 119660 XQuartz: Don't weed out duplicates in generated keymap 119661 119662 There seems to be an issue in the 1.5+ server where shift-space does not 119663 produce a space when 'keycode 57 = space' but it does when 'keycode 57 = space 119664 space' 119665 119666 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119667 119668commit 6d6e8fb27f00f0c3128cef624b39a60aa754fdc8 119669Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119670Date: Wed Nov 4 13:34:20 2009 -0800 119671 119672 XQuartz: Controller thread launches clients 119673 119674 This avoids a memory leak due to no active auto-release pool on the server thread. 119675 119676 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119677 119678commit f7b375bd141d0cf1e3add5443a5838dd8f554ef6 119679Author: Alan Coopersmith <alan.coopersmith@sun.com> 119680Date: Thu Nov 5 11:42:34 2009 -0800 119681 119682 Enable XF86PM on all Solaris platforms, not just x86/x64 119683 119684 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 119685 Signed-off-by: Keith Packard <keithp@keithp.com> 119686 119687commit f77262513ea17401092479e17ad20fc0eb91ffb5 119688Author: Alan Coopersmith <alan.coopersmith@sun.com> 119689Date: Thu Nov 5 11:42:33 2009 -0800 119690 119691 Use $(MAKE) instead of "make" to build Solaris inline assembly 119692 119693 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 119694 Signed-off-by: Keith Packard <keithp@keithp.com> 119695 119696commit 964040764387d89ef64324cfbee31872ee6ce41c 119697Author: Tiago Vignatti <tiago.vignatti@nokia.com> 119698Date: Wed Nov 4 15:35:37 2009 +0200 119699 119700 EXA: Preserve pPixmap->devPrivate.ptr in exaPixmapIsOffscreen_driver 119701 119702 It crash the server when the drawable is 32 bit and the framebuffer is 16. 119703 119704 This is pretty much a copy-past from commit 8e873185. 119705 119706 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 119707 Signed-off-by: Keith Packard <keithp@keithp.com> 119708 119709commit 7442f3355ab8f0bb2b1a270da18c65d8d315d4dd 119710Merge: 0573042cd 7897b6c2d 119711Author: Keith Packard <keithp@keithp.com> 119712Date: Wed Nov 4 08:55:20 2009 -0800 119713 119714 Merge remote branch 'jeremyhu/master' 119715 119716commit 0573042cddb6f9942e408687a16c6842e62a8bfa 119717Author: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 119718Date: Wed Nov 4 15:59:30 2009 +0200 119719 119720 dix: Fixes a memory leak when a cursor resource is released. 119721 119722 Just open and close a client that creates cursors in order to 119723 reproduce. In the problem case bits->refcnt is -1 and therefore 119724 bits->devPrivates is never released. 119725 119726 Signed-off-by: Rami Ylimaki <ext-rami.ylimaki@nokia.com> 119727 Signed-off-by: Keith Packard <keithp@keithp.com> 119728 119729commit 7897b6c2d41bccb72c19418674c3526ecce29515 119730Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119731Date: Tue Oct 27 18:00:48 2009 -0700 119732 119733 dix: Properly detect if the other device is frozen 119734 119735 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119736 119737commit 5e79976c13c5b94b12392b493846ca26be11750b 119738Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 119739Date: Tue Nov 3 16:35:27 2009 -0800 119740 119741 XQuartz: Run xmodmap after programatically updating the keymap. 119742 119743 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 119744 Signed-off-by: Martin Otte <otte@duke.edu> 119745 119746commit 412c56ef3332d09efbd861e41c3e985f44729729 119747Author: Maarten Maathuis <madman2003@gmail.com> 119748Date: Sun Nov 1 22:08:40 2009 +0100 119749 119750 exa: remove some outdated comment 119751 119752 - This comment is still in exa_driver.c and there it makes sense. 119753 119754 Signed-off-by: Maarten Maathuis <madman2003@gmail.com> 119755 Acked-by: Michel Dänzer <michel@daenzer.net> 119756 Signed-off-by: Keith Packard <keithp@keithp.com> 119757 119758commit 48b8c076a7b7dbbddfc472e8c6e5251a0b174cbc 119759Author: Matthieu Herrb <matthieu@laas.fr> 119760Date: Mon Nov 2 11:11:24 2009 +0100 119761 119762 add libc as a choice for SHA1 implementation 119763 119764 On Sun, Nov 01, 2009 at 02:54:13PM -0800, Keith Packard wrote: 119765 > Excerpts from Matthieu Herrb's message of Sun Nov 01 09:34:35 -0800 2009: 119766 > 119767 > > +AC_CHECK_FUNCS([SHA1Init], [HAVE_LIBC_SHA1=yes]) 119768 > 119769 > I'd suggest AC_CHECK_FUNC instead; as far as I can tell, AC_CHECK_FUNCS 119770 > will also define HAVE_SHA1INIT. Also, can you use HAVE_LIBC_SHA1 119771 > consistently rather than having two separate names (HAVE_LIBC_SHA1 and 119772 > HAVE_SHA1_IN_LIBC)? Yes, I know one is a preprocessor symbol and the 119773 > other is a cpp shell variable, but I think that will work anyway. 119774 > 119775 New version taking you comments into account. 119776 119777 From: Matthieu Herrb <matthieu.herrb@laas.fr> 119778 Date: Sun, 1 Nov 2009 18:19:27 +0100 119779 Subject: [PATCH] Add a probe for SHA1 functions in libc in *BSD. 119780 119781 The interface is the same as the one in libmd. 119782 119783 Signed-off-by: Keith Packard <keithp@keithp.com> 119784 119785commit da923d0bc15e99a8ed1986bd6f5df37f7af7284b 119786Author: Matt Turner <mattst88@gmail.com> 119787Date: Sun Nov 1 14:30:48 2009 -0500 119788 119789 Make sys.c use unaligned access functions provided in compiler. 119790 119791 Favorite deleted line was definitely 119792 /* to cope with broken egcs-1.1.2 :-(((( */ 119793 119794 Signed-off-by: Matt Turner <mattst88@gmail.com> 119795 Signed-off-by: Keith Packard <keithp@keithp.com> 119796 119797commit d30637339963950910e5f5fb755b8465ac7dddb4 119798Author: Mikhail Gusarov <dottedmag@dottedmag.net> 119799Date: Thu Oct 29 11:46:22 2009 -0700 119800 119801 Supply all code using dl*() with DLOPEN_LIBS 119802 119803 Previously DLOPEN_LIBS was managed in top-level configure.ac. 119804 Instead bundle it with the code using dl*() functions to 119805 avoid breakages in uncommon configurations. 119806 119807 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 119808 Reviewed-by: Dan Nicholson <dbn.lists@gmail.com> 119809 Signed-off-by: Keith Packard <keithp@keithp.com> 119810 119811commit ec98d7fc78efefcf9fc61492529157c0d289c3f2 119812Author: Adam Jackson <ajax@redhat.com> 119813Date: Mon Sep 21 10:23:16 2009 -0400 119814 119815 EDID: Extend the HDTV hack to handle "1368x769" 119816 119817 Hate televisions so much. 119818 119819 Signed-off-by: Adam Jackson <ajax@redhat.com> 119820 119821commit 7c0803f555782dbf451b7c79112d7deae02e5c9f 119822Author: Adam Jackson <ajax@redhat.com> 119823Date: Wed Oct 28 15:44:37 2009 -0400 119824 119825 modes: Fix duplicate detection, and do it more consistently 119826 119827 Signed-off-by: Adam Jackson <ajax@redhat.com> 119828 119829commit ba2d39dd5428cb5922b797a1d4ea45b859412b40 119830Author: Adam Jackson <ajax@redhat.com> 119831Date: Mon Oct 26 15:10:30 2009 -0400 119832 119833 modes: De-duplicate a clock range check. 119834 119835 Signed-off-by: Adam Jackson <ajax@redhat.com> 119836 119837commit 25236d19e6ef07fcb2c71569f1b7b0c12810834a 119838Author: Adam Jackson <ajax@redhat.com> 119839Date: Mon Oct 26 14:49:57 2009 -0400 119840 119841 EDID: Fix interlaced detailed timings to be frame size, not field size 119842 119843 Signed-off-by: Adam Jackson <ajax@redhat.com> 119844 119845commit fb86433d897c116315cc7994390d11ac2f577511 119846Author: Adam Jackson <ajax@redhat.com> 119847Date: Mon Oct 26 14:04:23 2009 -0400 119848 119849 modes: Decorate interlaced mode names with a trailing 'i' 119850 119851 Signed-off-by: Adam Jackson <ajax@redhat.com> 119852 119853commit fc2ec95664d55f45f77f1ebb039a7c17a1fcdaa3 119854Author: Ma Ling <ling.ma@intel.com> 119855Date: Wed Feb 18 17:41:26 2009 +0800 119856 119857 EDID: CEA extension support 119858 119859 Reviewed-by: Adam Jackson <ajax@redhat.com> 119860 119861commit fab74d1081270fb8f1d231e6e10d10aa33e164da 119862Author: Jamey Sharp <jamey@minilop.net> 119863Date: Wed Oct 28 17:48:19 2009 -0700 119864 119865 Suppress certain GCC warnings in auto-generated code. 119866 119867 - Don't warn for references to deprecated functions in xorg_symbols. 119868 - Ignore functions generated by gl_apitemp.py that are never used. 119869 119870 Signed-off-by: Jamey Sharp <jamey@minilop.net> 119871 Signed-off-by: Keith Packard <keithp@keithp.com> 119872 119873commit 239435875d6a92ed31731b500a992a3af0943594 119874Author: Jamey Sharp <jamey@minilop.net> 119875Date: Wed Oct 28 17:48:18 2009 -0700 119876 119877 Don't cast double to int: use default conversions or explicitly round. 119878 119879 GCC warns about casting a double return value to int. 119880 119881 Signed-off-by: Jamey Sharp <jamey@minilop.net> 119882 Signed-off-by: Keith Packard <keithp@keithp.com> 119883 119884commit e8c48fd8f7aab54327b0091cd17c60235ae27168 119885Author: Jamey Sharp <jamey@minilop.net> 119886Date: Wed Oct 28 17:48:17 2009 -0700 119887 119888 Suppress GCC warnings like "the address of `u1' will always evaluate as `true'". 119889 119890 Signed-off-by: Jamey Sharp <jamey@minilop.net> 119891 Signed-off-by: Keith Packard <keithp@keithp.com> 119892 119893commit 662594aeff9d1767316f08600949c73ac5060d18 119894Author: Mikhail Gusarov <dottedmag@dottedmag.net> 119895Date: Thu Oct 29 00:40:48 2009 +0600 119896 119897 kdrive: Grab evdev mouse/keyboard devices when X server is active 119898 119899 Input events are directed to both vt and input devices by default. 119900 Unless input devices are grabbed, keyboard events fill it vt buffers 119901 and cause spontaneous wakeups in kernel tty layer when buffers are full. 119902 119903 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 119904 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 119905 Signed-off-by: Keith Packard <keithp@keithp.com> 119906 119907commit 83d520d86c888f2c8a4abb18b7a8858d568aa18f 119908Author: Jamey Sharp <jamey@minilop.net> 119909Date: Wed Oct 28 16:45:44 2009 -0700 119910 119911 Add video driver flag to indicate that console access is not needed. 119912 119913 Existing video drivers will get the console enabled by default. 119914 119915 Signed-off-by: Jamey Sharp <jamey@minilop.net> 119916 Signed-off-by: Keith Packard <keithp@keithp.com> 119917 119918commit 25979c46b467847ccb54f5c86a1be6b9c303c99a 119919Author: Jamey Sharp <jamey@minilop.net> 119920Date: Wed Oct 28 16:35:28 2009 -0700 119921 119922 Alloc/free drawables array for each ProcPanoramiXShmGetImage call. 119923 119924 Updates my previous patch, b422b532f3dcab54c53f61a66f2ad76059d1874a. 119925 keithp recommended against allocating the drawables array globally, but my 119926 updated patch with that fixed isn't the patch that landed. 119927 119928 Signed-off-by: Jamey Sharp <jamey@minilop.net> 119929 Signed-off-by: Keith Packard <keithp@keithp.com> 119930 119931commit 757c11630d5999ad13dcac79191429badc92a3a6 119932Author: Tormod Volden <lists.tormod@gmail.com> 119933Date: Wed Oct 28 23:10:43 2009 +0100 119934 119935 xfree86: Fix description of DDC_QUIRK_DETAILED_USE_MAXIMUM_SIZE 119936 119937 The message ending up in the log is misleading as to what the quirk 119938 actually does: It ignores the sizes in the detailed timings and 119939 replaces them with the display "Max Image Size". 119940 119941 Signed-off-by: Tormod Volden <debian.tormod@gmail.com> 119942 Signed-off-by: Keith Packard <keithp@keithp.com> 119943 119944commit ccf4a69db747b000aee09072aac0a2891bde139a 119945Author: Mikhail Gusarov <dottedmag@dottedmag.net> 119946Date: Thu Oct 29 01:54:00 2009 +0600 119947 119948 os: Add libsha1 as a choice of SHA1 implementation 119949 119950 There are small systems which don't need OpenSSL or gcrypt. 119951 Add libsha1 (http://github.com/dottedmag/libsha1) as an alternative 119952 small SHA1 implementation. 119953 119954 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 119955 Signed-off-by: Keith Packard <keithp@keithp.com> 119956 119957commit 8613e4b0eb04150b1e377871f02b164be5d001e9 119958Author: Mikhail Gusarov <dottedmag@dottedmag.net> 119959Date: Wed Oct 28 11:44:27 2009 -0700 119960 119961 Add missing DLOPEN_LIBS to kdrive compilation flags 119962 119963 Xfbdev failed to build due to dladdr being used by xorg_backtrace. 119964 Explicitly add DLOPEN_LIBS to KDRIVE_LIBS as there does not 119965 seem to be a better place for it. 119966 119967 Signed-off-by: Mikhail Gusarov <dottedmag@dottedmag.net> 119968 Signed-off-by: Keith Packard <keithp@keithp.com> 119969 119970commit deb72fc61464250af8185dab2da8ee09f13c55d8 119971Merge: 55f4c80a4 a60e676f1 119972Author: Keith Packard <keithp@keithp.com> 119973Date: Wed Oct 28 10:54:13 2009 -0700 119974 119975 Merge remote branch 'jcristau/sha1' 119976 119977commit 55f4c80a4c891b355a99e6a05978ca945397c5cc 119978Author: Hans Nieser <hnsr@xs4all.nl> 119979Date: Wed Oct 28 06:59:40 2009 +0100 119980 119981 Xinput: allow non-integer values again for Constant- and AdaptiveDeceleration 119982 119983 This was initially fixed by commit 3932a848572f4eaf8b7f1d91d9b74aeafab069a2 119984 but then (presumably not intentionally) undone by commit 119985 1d54479cb3c8b4f75b7564f8b5e1c5da940b20f4 . 119986 119987 Signed-off-by: Hans Nieser <hnsr@xs4all.nl> 119988 Signed-off-by: Keith Packard <keithp@keithp.com> 119989 119990commit 50a5c32430a5267f2a05656d2417f9a8a44d8b97 119991Author: Jon TURNEY <jon.turney@dronecode.org.uk> 119992Date: Fri Oct 23 19:32:42 2009 +0100 119993 119994 dmx: Correctly compute DMXGetScreenAttributes reply length 119995 119996 Correctly allow for excess length of DMXGetScreenAttributes reply 119997 over standard 32 byte reply in addition to the displayName string 119998 when computing the length of reply 119999 120000 http://bugs.freedesktop.org/show_bug.cgi?id=24685 120001 120002 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 120003 Reviewed-by: Julien Cristau <jcristau@debian.org> 120004 Signed-off-by: Keith Packard <keithp@keithp.com> 120005 120006commit 9a2f6135bfb0f12ec28f304c97917d2f7c64db05 120007Author: Keith Packard <keithp@keithp.com> 120008Date: Fri Oct 23 10:04:57 2009 +0900 120009 120010 DRI2: Report the correct extension minor version 120011 120012 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 120013 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 120014 Signed-off-by: Keith Packard <keithp@keithp.com> 120015 120016commit d886008c96cd16f735f54dace070cb00b23c6f44 120017Author: Marcin Baczyński <marbacz@gmail.com> 120018Date: Thu Oct 22 14:43:25 2009 +0200 120019 120020 Kill compilation warnings. 120021 120022 Signed-off-by: Marcin Baczyński <marbacz@gmail.com> 120023 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 120024 Signed-off-by: Keith Packard <keithp@keithp.com> 120025 120026commit 15b30fde179cba3877182cd51b0f870ef29ffaee 120027Author: Adam Jackson <ajax@redhat.com> 120028Date: Thu Oct 22 17:29:01 2009 -0400 120029 120030 dix: Fix up colormap fixup. 120031 120032 FindClientResourcesByType() will walk all colormaps on all screens; we 120033 only want to fix up the current screen. Otherwise, screens > 0 will 120034 have the visual pointers for their colormaps pointing off into space. 120035 120036 Signed-off-by: Adam Jackson <ajax@redhat.com> 120037 120038commit 3785475a78636eb6547ef9e46be9e009c7cf7800 120039Author: Keith Packard <keithp@keithp.com> 120040Date: Fri Oct 23 09:03:39 2009 +0900 120041 120042 Bump to 1.7.99.2 (unreleased) 120043 120044commit 909df9beb3ddd02632f36ae682537280a6a8e5b4 120045Author: Jon TURNEY <jon.turney@dronecode.org.uk> 120046Date: Tue Oct 20 12:57:05 2009 +0100 120047 120048 Resolve an inconsistency between libX11 and Xserver over GetModifierMapping 120049 120050 libX11 ModMap.c believes that GetModifierMapping can never return an error 120051 120052 Xserver devices.c believes that GetModifierMapping can return an error if 120053 the ModMap couldn't be generated 120054 120055 According to the protocol document I have, libX11 is right, so adjust the 120056 server to send back an empty modmap if one couldn't be made... 120057 120058 http://bugs.freedesktop.org/show_bug.cgi?id=24621 120059 120060 Reviewed-by: Peter Hutterer <peter.hutterer@who-t.net> 120061 Signed-off-by: Keith Packard <keithp@keithp.com> 120062 120063commit 9c48862ac1ac119b6cfb7e376533f53af6a857f4 120064Author: Keith Packard <keithp@keithp.com> 120065Date: Thu Oct 22 01:12:03 2009 +0900 120066 120067 Bump version to 1.7.99.1 120068 120069 This is a development snapshot for 1.8. 120070 120071 Signed-off-by: Keith Packard <keithp@keithp.com> 120072 120073commit 26f4d8a2c12ff06c4beb06842cd8d9c0d9ddd4aa 120074Author: Keith Packard <keithp@keithp.com> 120075Date: Thu Oct 22 01:04:42 2009 +0900 120076 120077 Make sure dmx docs are built for distribution. 120078 120079 Ok, dmx docs are driving me slightly nuts. We probably shouldn't 120080 include the built versions in the tarball, but we do, so this is an 120081 attempt to make that work by having both the 'all' and 'dist' targets 120082 depends on the doxygen output. 120083 120084 Signed-off-by: Keith Packard <keithp@keithp.com> 120085 120086commit 4114854893ff57a1be525ec099a4b33e5f9963b1 120087Author: Keith Packard <keithp@keithp.com> 120088Date: Thu Oct 22 00:17:37 2009 +0900 120089 120090 Don't attempt to build hw/kdrive/sdl/Makefile anymore 120091 120092 With sdl removed, there's no way to build the sdl Makefile. 120093 120094 Signed-off-by: Keith Packard <keithp@keithp.com> 120095 120096commit 4d333c5121818754356853724333eadec2dcd18c 120097Author: Keith Packard <keithp@keithp.com> 120098Date: Thu Oct 22 00:16:12 2009 +0900 120099 120100 Always check for doxygen so that distcheck gets dmx docs 120101 120102 make distcheck wants the built dmx documentation so that users don't 120103 have to install doxygen. This means that even if dmx isn't built, the 120104 docs need to be so that the tarball can include them. 120105 120106 Signed-off-by: Keith Packard <keithp@keithp.com> 120107 120108commit 7e92bac5f769aca99bd20e21fe2811f0480b647b 120109Author: Keith Packard <keithp@keithp.com> 120110Date: Thu Oct 22 00:00:35 2009 +0900 120111 120112 Make sure HAVE_DOXYGEN is defined when not building dmx 120113 120114 The DMX docs are build using doxygen if present, so configure.ac 120115 checks to see if that is available. However, when not building dmx 120116 (the default), this conditional must still be defined to make automake 120117 happy. 120118 120119 Signed-off-by: Keith Packard <keithp@keithp.com> 120120 120121commit 1228e2d052f0bb98175c55c194340773b5fedb40 120122Merge: 08e7f62fa 52bc6d944 120123Author: Keith Packard <keithp@keithp.com> 120124Date: Wed Oct 21 22:46:53 2009 +0900 120125 120126 Merge remote branch 'whot/master' 120127 120128commit 08e7f62faf72540cb3a6f1023024c145f7fa1a23 120129Author: Keith Packard <keithp@keithp.com> 120130Date: Wed Oct 21 16:46:55 2009 +0900 120131 120132 Fix 'distcheck' to use host xkb files but install to build dir 120133 120134 'make distcheck' needs to read xkb files and write out compiled 120135 versions as a part of the 'make check' phase. This patch passes 120136 suitable options to the configure stage of the distcheck process to 120137 read xkb files from the system location and write them to the 120138 distcheck _inst directory. 120139 120140 Signed-off-by: Keith Packard <keithp@keithp.com> 120141 120142commit 52bc6d944946e66ea2cc685feaeea40bb496ea83 120143Author: Peter Hutterer <peter.hutterer@who-t.net> 120144Date: Fri Oct 16 11:03:09 2009 +1000 120145 120146 kdrive: Purge Xsdl 120147 120148 From the original Xsdl commit: 120149 "sdl x server so that we can x-on-x the fb stuff for ease of debugging. if 120150 anyone uses this in production, a big scary monster will eat them. 120151 hrm, perhaps i should make it have a --i-know-what-i'm-doing 120152 param that it doens't start without, heh" 120153 120154 That should be reason enough to not spend time maintaing it. Also, no more 120155 elephants. 120156 120157 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120158 120159commit f713f447a2110718dfc091380699362d76f0cd6c 120160Author: Lee Leahu <freedesktop-bugs@dyweni.com> 120161Date: Mon Oct 19 15:43:59 2009 -0500 120162 120163 dmxDestroyWindow() - must call the X's native DetroyWindow() 120164 120165 Don't really know why this section was disabled, but without it, 120166 certain pPicture resources do not get free'd until later in the 120167 FreeClientResources() process after the screen has been free'd - 120168 resulting in seg fault. 120169 120170 With this patch, all resources normally free'd using vanilla X are 120171 now also being freed correctly by Xdmx. 120172 120173 https://bugs.freedesktop.org/show_bug.cgi?id=24576 120174 120175 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120176 120177commit 664a8e37fd83141974b772980f680b94e48b4f87 120178Author: Lee Leahu <freedesktop-bugs@dyweni.com> 120179Date: Sat Oct 17 00:45:44 2009 -0500 120180 120181 dmx: when setting up device axis, use the correct counter number 120182 120183 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120184 120185commit 010d5e28655d619693632938e9e4325eef938295 120186Author: Peter Hutterer <peter.hutterer@who-t.net> 120187Date: Fri Oct 16 10:01:13 2009 +1000 120188 120189 dmx: remove doxygen-generated files. 120190 120191 These can be recreated by simply running 'doxygen doxygen.conf' in 120192 hw/dmx/doc. Some of the files do not exist anymore, these have been removed. 120193 Some other files have a different naming scheme. 120194 Doxygen warnings about missing links fixed, two warnings remain: 120195 120196 /home/whot/xorg/xserver/hw/dmx/dmxwindow.c:142: Warning: explicit link 120197 request to 'dmxConfigureRootWindow' could not be resolved 120198 /home/whot/xorg/xserver/hw/dmx/dmxwindow.c:119: Warning: explicit link 120199 request to 'dmxConfigureScreenWindow()' could not be resolved 120200 120201 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120202 120203commit ff3e171568c8b604fc5eacd5f7e9b7ee9417f172 120204Author: Peter Hutterer <peter.hutterer@who-t.net> 120205Date: Fri Oct 16 10:26:21 2009 +1000 120206 120207 xkb: don't conditionally include xkb-config.h. 120208 120209 If HAVE_XKB_CONFIG_H is ever undefined, we fail to build anyway. 120210 120211 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120212 120213commit 0b9dbd4bf2d1061514bebd28648538a5fc0b60e6 120214Author: Peter Hutterer <peter.hutterer@who-t.net> 120215Date: Fri Oct 16 08:24:37 2009 +1000 120216 120217 Use the default XKB settings for dmx and kdrive. 120218 120219 Drop the dmx-specific defines, there's no reason to have separate ones 120220 considering they're about as hardcoded as the default rules anyway. 120221 120222 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120223 120224commit e08d8a2b043e5be821c5e8b2dbf37b09102b5d01 120225Author: Peter Hutterer <peter.hutterer@who-t.net> 120226Date: Thu Sep 24 11:25:28 2009 +1000 120227 120228 kdrive: silence tslib compiler warnings 120229 120230 tslib.c: In function 'TslibInit': 120231 tslib.c:157: warning: unused variable 'tsDev' 120232 tslib.c:156: warning: unused variable 'inputent' 120233 tslib.c:155: warning: unused variable 'inputdir' 120234 tslib.c:154: warning: unused variable 'i' 120235 tslib.c:154: warning: unused variable 'fd' 120236 120237 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120238 120239commit 522ca8179ad2b2b935993f08c9382cd5ad1d8b2f 120240Author: Mikhail Gusarov <dottedmag@dottedmag.net> 120241Date: Tue Oct 20 04:18:01 2009 +0700 120242 120243 Remove unused LinuxFindPci and LinuxGetPciCfg 120244 120245 These two functions are not referenced from inside xserver. 120246 Remove now-empty klinux.h too. 120247 120248 Signed-off-by: Keith Packard <keithp@keithp.com> 120249 120250commit 63f4bf39170eb2262617ef2dc95fd6d337b9dad5 120251Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120252Date: Mon Oct 19 13:07:26 2009 +1000 120253 120254 Fix make dist after 78c87bdad1feab91c2a39e01513b4b0826665f06 120255 120256 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120257 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120258 Signed-off-by: Keith Packard <keithp@keithp.com> 120259 120260commit 2bd71a6db241924bd17ac9dd4aa38e7e563ee5c1 120261Author: Alan Coopersmith <alan.coopersmith@sun.com> 120262Date: Fri Oct 16 22:29:26 2009 -0700 120263 120264 Fix segfault when -extension XKEYBOARD is passed on the command line 120265 120266 Users should be told they can't disable XKB or XInput via error messages, 120267 not core dumps. 120268 120269 Reported by T`2 on #xorg irc 120270 120271 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 120272 Signed-off-by: Keith Packard <keithp@keithp.com> 120273 120274commit a32b2420d85f076282721afe005e85fff7d9837e 120275Author: Peter Hutterer <peter.hutterer@who-t.net> 120276Date: Thu Oct 15 13:50:36 2009 +1000 120277 120278 kdrive: fix Xfake build by removing the old keysym stuff 120279 120280 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120281 Acked-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120282 Signed-off-by: Keith Packard <keithp@keithp.com> 120283 120284commit 7c31dd5db8b43c7796bf97a07e08213af5afd2ae 120285Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120286Date: Thu Oct 15 04:24:25 2009 -0500 120287 120288 Remove CopyISOLatin1Lowered 120289 120290 This function was moved verbatim into libXfont-1.4, and it is not used 120291 by the server or any drivers. Exporting it in both places leads to 120292 multiple definition linking errors on Cygwin, where we need to use a 120293 static libXfont due to poor weak-symbol handling. 120294 120295 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120296 Signed-off-by: Keith Packard <keithp@keithp.com> 120297 120298commit 9bc4e88d84daf0f4faf0599b575675e74c75f4b8 120299Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120300Date: Thu Oct 15 01:53:43 2009 -0500 120301 120302 Define ddxBeforeReset stubs in platform-neutral DDXs 120303 120304 XWin uses ddxBeforeReset, which is called in DIX. Other DDXs need to 120305 define these in order to avoid an undefined symbol error at link time 120306 when building alongside XWin. Xnest and Xvfb already provide empty stubs; 120307 this does the same for Xdmx and the platform-neutral KDrive servers. 120308 120309 Also add a prototype to avoid a warning in all DDXs. 120310 120311 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120312 Signed-off-by: Keith Packard <keithp@keithp.com> 120313 120314commit a2f27b97900f335cd5f6a3e5bf8fa1d0ec9eb9ac 120315Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120316Date: Thu Oct 15 01:53:42 2009 -0500 120317 120318 Remove duplicates from Xfake_LDADD 120319 120320 KDRIVE_LIBS already contains the libs in XSERVER_LIBS, so linking against 120321 both leads to multiple-definition errors when linking on Cygwin. 120322 120323 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120324 Signed-off-by: Keith Packard <keithp@keithp.com> 120325 120326commit ed4c6bc79a797830895ad6c3601318b2d1c4bdcb 120327Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120328Date: Thu Oct 15 01:53:41 2009 -0500 120329 120330 Fix XWIN_LIBS definition 120331 120332 Move XWIN_LIBS from XWIN=auto conditional to XWIN=yes conditional, 120333 otherwise the build breaks if an explicit --enable-xwin is passed to 120334 configure. 120335 120336 Also remove obsolete library values from the definition. 120337 120338 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120339 Signed-off-by: Keith Packard <keithp@keithp.com> 120340 120341commit bcd0c02cfd0707b5f8b1ca91dfbe4412df2c643e 120342Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120343Date: Thu Oct 15 01:53:40 2009 -0500 120344 120345 Install SDK pkg-config and aclocal files only if XORG 120346 120347 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120348 Signed-off-by: Keith Packard <keithp@keithp.com> 120349 120350commit 264ce9e8360374b3a43442c8bdea08abde705446 120351Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 120352Date: Wed Oct 14 20:33:53 2009 -0400 120353 120354 xselinux: Use the now-exported IsPointerDevice() instead of a copy. 120355 120356 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120357 120358commit 0ae1632be045bfbb288bb57190c830f94247460f 120359Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 120360Date: Wed Oct 14 20:32:28 2009 -0400 120361 120362 dix: Export IsPointerDevice() and IsKeyboardDevice(). 120363 120364 Makes the functions available to extmod for extensions to call. 120365 120366 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120367 120368commit 6c2ae5fec552366e11ad64a27626eb5dec4becf0 120369Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 120370Date: Tue Sep 22 13:31:49 2009 -0700 120371 120372 xselinux: switch from x_device to separate x_pointer and x_keyboard classes. 120373 120374 This will allow separate controls over pointer and keyboard without having 120375 to relabel the devices to separate types. 120376 120377 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120378 120379commit 4be354c4c2da5168b302601b91bd80cfaca7e193 120380Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 120381Date: Tue Sep 22 13:18:44 2009 -0700 120382 120383 Remove some debug messages that trigger on XACE event delivery failure. 120384 120385 It is normal for XACE to deny an event delivery, so these log messages 120386 shouldn't trigger when that happens. Just drop them for now. 120387 120388 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120389 120390commit c4ffce4dc84a0a9d134a59b7e7765c99ed767e53 120391Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 120392Date: Tue Sep 22 13:13:03 2009 -0700 120393 120394 xace: Relax permissions on XkbGetState from Read to Getattr. 120395 120396 This request is used to get the current keyboard group and is called from 120397 GTK. It does not return an actual keymap (aside from modifiers) so it 120398 should be safe to relax the permission on it. However it does return 120399 button state information which should be controlled through a separate 120400 pointer Read check. 120401 120402 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120403 120404commit 0ff28319906eeb3f236acd72201c416ce01f2c6e 120405Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 120406Date: Tue Sep 15 19:41:04 2009 -0400 120407 120408 xselinux: Stop special-casing QueryPointer access checks. 120409 120410 XACE has been changed to not return BadAccess on device read failures. 120411 Thus, no need for this workaround code. 120412 120413 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120414 120415commit 8502c06e19a4c00bf1311f54f9a365ee9e026e97 120416Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 120417Date: Tue Sep 15 19:29:34 2009 -0400 120418 120419 xace: Fake return values on denials in input polling requests. 120420 120421 Instead of returning BadAccess when "read" permission is denied 120422 on a device, falsify the device state (buttons down, keys pressed). 120423 This is nicer to applications, but may still have undesired side 120424 effects. The long-term solution is not to use these requests in 120425 event-driven code! 120426 120427 Requests affected: QueryPointer, QueryKeymap, XiQueryDevice. 120428 120429 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120430 120431commit 0493935691e925ae137af7636fa15befa76c8b45 120432Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 120433Date: Mon Aug 31 21:20:53 2009 -0400 120434 120435 xselinux: Factor out some dynamic array code into common helpers. 120436 120437 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120438 120439commit ae8891ba0b63bfe6941a324e201d9ab7c645c0f3 120440Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 120441Date: Mon Aug 31 18:46:23 2009 -0400 120442 120443 xselinux: refactor extension code into smaller files. 120444 120445 New files: 120446 xselinux_ext.c: Extension init and request handlers. 120447 xselinux_hooks.c: XACE hook functions and other callbacks. 120448 xselinux_label.c: Object security-labeling code. 120449 xselinuxint.h: Shared internal functions. 120450 120451 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120452 120453commit a60e676f1fd243c78859440b87652f523d3f2ec1 120454Author: Julien Cristau <jcristau@debian.org> 120455Date: Wed Oct 14 23:51:22 2009 +0200 120456 120457 Add libgcrypt as an option for SHA1 120458 120459 Signed-off-by: Julien Cristau <jcristau@debian.org> 120460 Reviewed-by: Rémi Cardona <remi@gentoo.org> 120461 120462commit d2a6a395435919aff8943285f9cbfe6569a9728f 120463Author: Julien Cristau <jcristau@debian.org> 120464Date: Wed Oct 14 23:30:55 2009 +0200 120465 120466 configure: add --with-sha1={libmd,libcrypto} option 120467 120468 Signed-off-by: Julien Cristau <jcristau@debian.org> 120469 Reviewed-by: Rémi Cardona <remi@gentoo.org> 120470 120471commit 55516094947dd78ad2734bb784a2fb109b64c990 120472Author: Julien Cristau <jcristau@debian.org> 120473Date: Wed Oct 14 23:20:44 2009 +0200 120474 120475 Move SHA1 computation from render/glyph.c to os/ 120476 120477 Signed-off-by: Julien Cristau <jcristau@debian.org> 120478 Reviewed-by: Rémi Cardona <remi@gentoo.org> 120479 120480commit 22b38f513c93c2f2aea5909878c3c9acae7d35c1 120481Author: Jon TURNEY <jon.turney@dronecode.org.uk> 120482Date: Sat Oct 3 12:44:50 2009 +0100 120483 120484 Cygwin/X: Avoid a potential null pointer dereference before input initialization 120485 120486 Avoid a null pointer dereference if WM_MOUSEMOVE occurred before 120487 the input device had been initialized (a timing sensitive bug 120488 occassionally seen during initialization) 120489 120490 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 120491 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 120492 120493commit 1b0dfd8dee639870725d3bd9b70c3bd589d09e5a 120494Author: Jon TURNEY <jon.turney@dronecode.org.uk> 120495Date: Tue Jul 21 02:23:21 2009 +0100 120496 120497 Cygwin/X: Make -logverbose affect the verbosity of logging to the log file 120498 120499 Make -logverbose affect the verbosity of logging to the log file, not just the 120500 verbosity of logging to the console 120501 120502 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 120503 Reviewed-by: Colin Harrison <colin.harrison@virgin.net> 120504 120505commit 83d120b90482d356be730f63aead0f8e44a4e846 120506Author: Colin Harrison <colin.harrison@virgin.net> 120507Date: Thu Oct 1 14:47:22 2009 +0100 120508 120509 Xming: Fix various 'ISO C90 forbids mixed declarations and code' warnings 120510 120511 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 120512 120513commit 78c87bdad1feab91c2a39e01513b4b0826665f06 120514Author: Colin Harrison <colin.harrison@virgin.net> 120515Date: Fri Sep 25 17:49:59 2009 +0100 120516 120517 Xming: Remove unused X-boxed icon resource 120518 120519 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 120520 120521commit fdf7f7e95c852734c620f58df408a324902172cf 120522Author: Colin Harrison <colin.harrison@virgin.net> 120523Date: Fri Sep 25 17:46:40 2009 +0100 120524 120525 Xming: update .rc file 120526 120527 Remove the obsolete DISCARDABLE flag in the .rc file 120528 Replace the obsolete DIALOG resource with DIALOGEX in the .rc file 120529 120530 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 120531 120532commit 5ddfd9d0898629627e8ab10e65ea2e082be3af37 120533Author: Jon TURNEY <jon.turney@dronecode.org.uk> 120534Date: Sat Oct 3 13:07:23 2009 +0100 120535 120536 Cygwin/X: Don't build rootless extension code unless we need it 120537 120538 Only try to build rootless extension code if we really need it (i.e. mwextwm mode is enabled) 120539 120540 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 120541 120542commit 25344ba7f7845654364d62bf15322b3b79465bd9 120543Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120544Date: Tue Oct 13 20:14:58 2009 -0500 120545 120546 dolt: add Cygwin to supported platforms 120547 120548 Signed-off-by: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 120549 Signed-off-by: Keith Packard <keithp@keithp.com> 120550 120551commit 6e158003e80534ce007290f75c89d698aec1d00b 120552Merge: 2b14e1420 5e762f0e2 120553Author: Keith Packard <keithp@keithp.com> 120554Date: Tue Oct 13 18:40:42 2009 -0700 120555 120556 Merge remote branch 'mattst88/master' 120557 120558 * mattst88/master: 120559 [alpha] assume we have __NR_pciconfig_iobase 120560 [alpha] don't return from void functions 120561 Fix undefined symbols on alpha 120562 Fix breakage on alpha caused by c7680befe5ae 120563 Revert "alpha: kill xf86SlowBCopyToBus and xf86SlowBCopyFromBus" 120564 120565commit 5e762f0e2f203b5121a3de3b9af3c8981b31a77f 120566Author: Matt Turner <mattst88@gmail.com> 120567Date: Tue Oct 13 20:42:14 2009 -0400 120568 120569 [alpha] assume we have __NR_pciconfig_iobase 120570 120571 The code path if we didn't have support has been broken since before we 120572 switched to git. 120573 120574 The pciconfig_iobase syscall has been supported since 2000. 120575 120576 Signed-off-by: Matt Turner <mattst88@gmail.com> 120577 120578commit 46785c04bca16f495af3ed8d685aee939a1a8f39 120579Author: Matt Turner <mattst88@gmail.com> 120580Date: Tue Oct 13 20:40:59 2009 -0400 120581 120582 [alpha] don't return from void functions 120583 120584 Signed-off-by: Matt Turner <mattst88@gmail.com> 120585 120586commit 2b14e142039193ea854a02706662204f4f5f6db8 120587Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 120588Date: Tue Oct 13 12:54:56 2009 -0700 120589 120590 Rootless: src drawable window can now be NULL 120591 120592 Fix a possible crash when pSrc->pDrawable is NULL. 120593 120594 Signed-off-by: Colin Harrison <colin.harrison@virgin.net> 120595 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 120596 Signed-off-by: Keith Packard <keithp@keithp.com> 120597 120598commit 5b91dfac6fbdf35288a4558638d4923e230ab8d3 120599Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 120600Date: Tue Oct 13 12:54:11 2009 -0700 120601 120602 Rootless: Abstract some of the Xplugin specific stuff which has crept into rootlessWindow.c 120603 120604 The rootless extension now directly calls some Xplugin functions, and relies 120605 on types defined in Xplugin.h, which isn't very abstracted :-) 120606 120607 This patch is a start at abstracting some of the Xplugin specific stuff which 120608 has crept into rootlessWindow.c. This has been done in a pretty mindless fashion, 120609 without much thought as to if the additions to the generic rootless interface are 120610 the correct ones 120611 120612 There is some confusion as to if RootlesscolormapCallback() returns a Bool or 120613 xp_error_enum value (not so abstact), but I have no way of checking, of finding 120614 out if Xplugin actually checks the result :-) 120615 120616 Based on patches from Colin Harrison, Jon Turney and Yaakov Selkowitz 120617 120618 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 120619 Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org> 120620 Signed-off-by: Keith Packard <keithp@keithp.com> 120621 120622commit 72f5874434c0c015b671c492c1318f35f1793668 120623Author: Peter Hutterer <peter.hutterer@who-t.net> 120624Date: Tue Oct 13 13:15:05 2009 +1000 120625 120626 dix: extend IsPointerDevice check to valuator-only devices. 120627 120628 A device with valuators but no keys is definitely a pointer device and needs 120629 to be attached to the VCP. Otherwise, the class copying happens on the VCK 120630 and the VCP isn't updated with the events that are to be sent through it. 120631 This addresses the trigger for #24441, not the actual issue. 120632 Jury is still out on valuator+key devices. 120633 120634 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120635 Signed-off-by: Keith Packard <keithp@keithp.com> 120636 120637commit 1088073b11ed488c0df45af3867b900ef93c6fe1 120638Author: Michel Dänzer <daenzer@vmware.com> 120639Date: Fri Oct 9 11:31:44 2009 +0200 120640 120641 EXA: Fix exaTryDriverSolidFill() for solid source pictures. 120642 120643 Solid pictures have a NULL pFormat field, but their format is always 120644 PICT_a8r8g8b8. 120645 120646 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 120647 Signed-off-by: Keith Packard <keithp@keithp.com> 120648 120649commit 55305cf8db7787883bc80b7348eb626e609626f8 120650Author: Ben Skeggs <bskeggs@redhat.com> 120651Date: Fri Oct 9 16:08:15 2009 -0700 120652 120653 EXA: fix exaGetRGBAFromPixel to not loop forever on PICT_a8 picture 120654 120655 Easily reproducible by running "rendercheck -t fill". 120656 120657 It should be safe to just test against rbits for all colour components 120658 as we should always have values for r/g/bbits for PICT_FORMAT_COLOR 120659 formats. 120660 120661 Signed-off-by: Ben Skeggs <bskeggs@redhat.com> 120662 Signed-off-by: Michel Dänzer <daenzer@vmware 120663 Signed-off-by: Keith Packard <keithp@keithp.com> 120664 120665commit b375be9285c40467578fac2f1360c29a75306ab4 120666Author: Michel Dänzer <daenzer@vmware.com> 120667Date: Fri Oct 9 11:31:46 2009 +0200 120668 120669 composite: Revert changes from adding support for BGRA picture formats. 120670 120671 They were aimed towards a since abandoned approach for making radeon KMS work 120672 on big endian machines, and Aaron Plattner pointed out that they break the 120673 Composite extension when the X server runs in 16bpp. 120674 120675 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 120676 Tested-by: Aaron Plattner <aplattner@nvidia.com> 120677 Signed-off-by: Keith Packard <keithp@keithp.com> 120678 120679commit c0a1bb511a4629bf5683d8a710dc4a1c577a5d44 120680Author: Alan Coopersmith <alan.coopersmith@sun.com> 120681Date: Wed Oct 7 22:10:55 2009 -0700 120682 120683 Fix make warning: overriding commands for target `dix.O' 120684 120685 Not only does automake generate unnecessary rules for dix.O on platforms 120686 for which SPECIAL_DTRACE_OBJECTS is false, it generates duplicate sets 120687 when "if SPECIAL_DTRACE_OBJECTS" is nested inside "if XSERVER_DTRACE" 120688 120689 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 120690 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 120691 Signed-off-by: Keith Packard <keithp@keithp.com> 120692 120693commit b0dd6be2c8703f7062d45ac9fd646550c7d54e3b 120694Author: Jamey Sharp <jamey@minilop.net> 120695Date: Thu Oct 8 13:38:44 2009 +1100 120696 120697 Cast small-int values through intptr_t when passed as pointers 120698 120699 On 64-bit systems, int and pointers don't have the same size, so GCC gives 120700 warnings about casts between int and pointer types. However, in the cases 120701 covered by this patch, it's always a value that fits in int being stored 120702 temporarily as a pointer and then converted back later, which is safe. 120703 Casting through the pointer-sized integer type intptr_t convinces the 120704 compiler that this is OK. 120705 120706 Signed-off-by: Jamey Sharp <jamey@minilop.net> 120707 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 120708 120709commit b680a89262efcfef4644adb4a61ae42ea0db0c93 120710Author: Jamey Sharp <jamey@minilop.net> 120711Date: Thu Oct 8 13:38:01 2009 +1100 120712 120713 Fix GCC warnings in xorg_backtrace 120714 120715 Signed-off-by: Jamey Sharp <jamey@minilop.net> 120716 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 120717 120718commit 9bf2ff4faf730913de3073f346646a8727be41d4 120719Author: Jamey Sharp <jamey@minilop.net> 120720Date: Thu Oct 8 13:36:44 2009 +1100 120721 120722 Fix "possibly uninitialized" warnings in glx 120723 120724 In both functions, "answer" was uninitialized if "compsize" was 0, but in 120725 that case __GLX_SEND_VOID_ARRAY(compsize) results in a call to 120726 WriteToClient for 0 bytes, which returns immediately without examining the 120727 "answer" argument. So initializing to a null pointer is as good as 120728 anything else. 120729 120730 Signed-off-by: Jamey Sharp <jamey@minilop.net> 120731 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 120732 120733commit b422b532f3dcab54c53f61a66f2ad76059d1874a 120734Author: Jamey Sharp <jamey@minilop.net> 120735Date: Thu Oct 8 13:29:27 2009 +1100 120736 120737 Remove static MAXSCREENS limit from Xext/shm.c 120738 120739 Dynamically allocate per-screen data in the SHM extension, instead of 120740 having a single static-sized array. 120741 120742 Signed-off-by: Jamey Sharp <jamey@minilop.net> 120743 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 120744 120745commit 4df3e8c8053d9d4b3043f339a73d0de97020d884 120746Author: Martin Ettl <ettl.martin@gmx.de> 120747Date: Thu Oct 8 13:27:30 2009 +1100 120748 120749 Resource leakage: 0 is a valid file descriptor 120750 120751 When testing if an fd is valid, the required construct is >= 0, not > 0. 120752 120753 [Daniel: Fixed up the Linux MTRR case as well.] 120754 120755 Signed-off-by: Martin Ettl <ettl.martin@gmx.de> 120756 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 120757 120758commit 4151a13c80f3afa43f88afcf19a7aeb16dace93a 120759Author: Francisco Jerez <currojerez@riseup.net> 120760Date: Mon Oct 5 02:39:03 2009 +0200 120761 120762 dix: Fix a double free in dixFreePrivates. 120763 120764 It can be reproduced when the server is regenerated and for some 120765 reason the private keys are reassigned in a different order: a 120766 manually allocated private may get an index formerly used by a 120767 preallocated private. In that case it will first be manually freed and 120768 then again by dixFreePrivates, as items[i].size was never zeroed 120769 out. Do it in dixResetPrivates. 120770 120771 Signed-off-by: Francisco Jerez <currojerez@riseup.net> 120772 Acked-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 120773 Signed-off-by: Keith Packard <keithp@keithp.com> 120774 120775commit 34eddbbb73bb16395dba0818247909c1b4bee4c2 120776Author: Michael Cree <mcree@orcon.net.nz> 120777Date: Wed Oct 7 18:33:29 2009 -0400 120778 120779 Fix undefined symbols on alpha 120780 120781 Signed-off-by: Matt Turner <mattst88@gmail.com> 120782 120783commit 44efcdde501d54ca9c3e33ab5b1f699956fc3f9f 120784Author: Jamey Sharp <jamey@minilop.net> 120785Date: Wed Oct 7 09:07:03 2009 -0700 120786 120787 Fix overlay detection when matching Xv adaptors across screens. 120788 120789 Signed-off-by: Jamey Sharp <jamey@minilop.net> 120790 Acked-by: Adam Jackson <ajax@redhat.com> 120791 Signed-off-by: Keith Packard <keithp@keithp.com> 120792 120793commit 68304215e25876ee639015969b4f07e1c9c515e0 120794Merge: 6676f49e3 315aaef55 120795Author: Keith Packard <keithp@keithp.com> 120796Date: Tue Oct 6 22:41:42 2009 -0700 120797 120798 Merge remote branch 'alanc/master' 120799 120800 Conflicts: 120801 configure.ac 120802 120803 Signed-off-by: Keith Packard <keithp@keithp.com> 120804 120805commit 315aaef55750a863c08a16ad9120ffb76f9b48e0 120806Author: Alan Coopersmith <alan.coopersmith@sun.com> 120807Date: Fri Oct 2 21:54:53 2009 -0700 120808 120809 Use $(AM_V_GEN) to silence more commands when AM_SILENT_RULES is active 120810 120811 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 120812 Reviewed-by: Eric Anholt <eric@anholt.net> 120813 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 120814 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 120815 120816commit 2e294380b64b6362173d0590c4292f290e50af50 120817Author: Alan Coopersmith <alan.coopersmith@sun.com> 120818Date: Fri Oct 2 16:51:24 2009 -0700 120819 120820 Migrate to xorg macros 1.3 & XORG_DEFAULT_OPTIONS 120821 120822 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 120823 Reviewed-by: Eric Anholt <eric@anholt.net> 120824 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 120825 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 120826 120827commit 4b0911565d18b1aefca564315d0f1b9b4a17aa65 120828Author: Alan Coopersmith <alan.coopersmith@sun.com> 120829Date: Mon Oct 5 16:52:19 2009 -0700 120830 120831 Add platform tests for Dtrace linker magic 120832 120833 Replaces special handling for Xquartz DDX and scales better to handling 120834 the multiple platforms that now have some level of Dtrace support available. 120835 120836 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 120837 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 120838 120839commit 8c51b886400c5962b31ff565771be1b01a3ca8fb 120840Author: Julien Cristau <jcristau@debian.org> 120841Date: Tue Oct 6 16:08:42 2009 +0200 120842 120843 configure: quote argument to m4_pattern_forbid 120844 120845 Without this, configure spits out 120846 ../configure: line 15460: ac_fn_c_check_member: command not found 120847 120848 Also anchor the pattern to make it stricter. 120849 120850 Signed-off-by: Julien Cristau <jcristau@debian.org> 120851 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 120852 Signed-off-by: Keith Packard <keithp@keithp.com> 120853 120854commit 6ee61578ddb9f880ef12bbe9c3671d7dd53f2809 120855Author: Alan Coopersmith <alan.coopersmith@sun.com> 120856Date: Mon Oct 5 17:27:27 2009 -0700 120857 120858 Remove shave now that automake-1.11 has AM_SILENT_RULES 120859 120860 Revert "Add shave so that we can see the steaming piles of warnings generated." 120861 This reverts commit 181cc08c8908a119fc403f970dea8cc98d3e0b9b. 120862 120863 shave was left in the xserver 1.7 release/branch to allow developers & 120864 distros time to transition to automake-1.11. xserver 1.8 series will 120865 require automake-1.11 for silencing build noise. 120866 120867 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 120868 Signed-off-by: Keith Packard <keithp@keithp.com> 120869 120870commit d5d960f57834f16164457dab2e80290717dbf2cd 120871Author: Kim Woelders <kim@woelders.dk> 120872Date: Fri Oct 2 19:31:15 2009 +0200 120873 120874 render: Fix clip region translation in miClipPictureSrc(). 120875 120876 Signed-off-by: Kim Woelders <kim@woelders.dk> 120877 Reviewed-by: Soren Sandmann Pedersen <sandmann@redhat.com> 120878 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120879 120880commit d2118c8ca94fc7bcb8e0de547e63cf022f53a8a9 120881Author: Peter Hutterer <peter.hutterer@who-t.net> 120882Date: Mon Oct 5 15:31:54 2009 +1000 120883 120884 xfree86: remove log-spamming DebugF 120885 120886 All input drivers use xf86PostKeyEventP indirectly now and have been since 120887 it exists. I guess that qualifies it as tested - no need to spam the logs. 120888 120889 Reported-by: Felix Wenk 120890 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120891 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 120892 120893commit bf116057c7efd43c247c93f2f5d733a5db857cc9 120894Author: Peter Hutterer <peter.hutterer@who-t.net> 120895Date: Mon Oct 5 10:44:43 2009 +1000 120896 120897 Require libXtst >= 1.0.99.2 and libdmx >= 1.0.99.1 120898 120899 libXtst requirement is already implicit since we require xextproto 7.1 and 120900 that doesn't go well with pre 1.0.99.2 versions of libXtst. Nonetheless, 120901 list it explicitly. 120902 120903 Since d0440275108920f5cb5d630f55fc9a3320c496d3 we require dmxproto 2.2.99.1. 120904 Complementing that is libdmx 1.0.99.1 with the dmxext.h header file. 120905 120906 Reported-by: Mark Rosenstand 120907 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120908 120909commit 9625f6d328d6f516520930227b218979309938bc 120910Author: Matt Turner <mattst88@gmail.com> 120911Date: Tue Oct 6 20:58:30 2009 -0400 120912 120913 Fix breakage on alpha caused by c7680befe5ae 120914 120915 Pinpointed by by Michael Cree. 120916 120917 Commit c7680befe5ae removed Jensen support, but at the same time broke 120918 support for dense memory systems. 120919 120920 Signed-off-by: Matt Turner <mattst88@gmail.com> 120921 120922commit aa07957373fd7cbe67458a001e4afd6a7f1ea37f 120923Author: Peter Hutterer <peter.hutterer@who-t.net> 120924Date: Mon Sep 28 12:20:03 2009 +1000 120925 120926 Revert "alpha: kill xf86SlowBCopyToBus and xf86SlowBCopyFromBus" 120927 120928 The vesa driver still uses slowbcopy_frombus and slowbcopy_tobus. 120929 120930 This reverts commit 5ef53a94ce4e48e11de26290cd677266308640c8. 120931 120932 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120933 120934commit 6676f49e34f8db7ef5767a6dbf6c2a4d2087f79c 120935Author: Julien Cristau <jcristau@debian.org> 120936Date: Tue Oct 6 16:08:42 2009 +0200 120937 120938 configure: quote argument to m4_pattern_forbid 120939 120940 Without this, configure spits out 120941 ../configure: line 15460: ac_fn_c_check_member: command not found 120942 120943 Also anchor the pattern to make it stricter. 120944 120945 Signed-off-by: Julien Cristau <jcristau@debian.org> 120946 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 120947 Signed-off-by: Keith Packard <keithp@keithp.com> 120948 120949commit a0363bf7b135d816e73849c91fa53d98fd6072d1 120950Author: Alan Coopersmith <alan.coopersmith@sun.com> 120951Date: Mon Oct 5 17:27:27 2009 -0700 120952 120953 Remove shave now that automake-1.11 has AM_SILENT_RULES 120954 120955 Revert "Add shave so that we can see the steaming piles of warnings generated." 120956 This reverts commit 181cc08c8908a119fc403f970dea8cc98d3e0b9b. 120957 120958 shave was left in the xserver 1.7 release/branch to allow developers & 120959 distros time to transition to automake-1.11. xserver 1.8 series will 120960 require automake-1.11 for silencing build noise. 120961 120962 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 120963 Signed-off-by: Keith Packard <keithp@keithp.com> 120964 120965commit cbc886a3513079c084fb7ce47b87c3e5bba19f3f 120966Author: Kim Woelders <kim@woelders.dk> 120967Date: Fri Oct 2 19:31:15 2009 +0200 120968 120969 render: Fix clip region translation in miClipPictureSrc(). 120970 120971 Signed-off-by: Kim Woelders <kim@woelders.dk> 120972 Reviewed-by: Soren Sandmann Pedersen <sandmann@redhat.com> 120973 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120974 120975commit bd7430a32e63df8cd60352764744076448ee623f 120976Author: Peter Hutterer <peter.hutterer@who-t.net> 120977Date: Mon Oct 5 15:31:54 2009 +1000 120978 120979 xfree86: remove log-spamming DebugF 120980 120981 All input drivers use xf86PostKeyEventP indirectly now and have been since 120982 it exists. I guess that qualifies it as tested - no need to spam the logs. 120983 120984 Reported-by: Felix Wenk 120985 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 120986 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 120987 120988commit 1a2dde3335864abfbbc133dbc709f564272dd540 120989Author: Alan Coopersmith <alan.coopersmith@sun.com> 120990Date: Fri Oct 2 20:44:19 2009 -0700 120991 120992 Set XQUARTZ to no on non-Darwin OS'es 120993 120994 Fixes build with dtrace probes on Solaris after efacd7bfd08ffc0725de6f639c6afbf3b2f6c9fe 120995 120996 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 120997 Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> 120998 120999commit c629e0fc50dfaffaa40a56709da7f035c289fed2 121000Author: Peter Hutterer <peter.hutterer@who-t.net> 121001Date: Mon Oct 5 10:44:43 2009 +1000 121002 121003 Require libXtst >= 1.0.99.2 and libdmx >= 1.0.99.1 121004 121005 libXtst requirement is already implicit since we require xextproto 7.1 and 121006 that doesn't go well with pre 1.0.99.2 versions of libXtst. Nonetheless, 121007 list it explicitly. 121008 121009 Since d0440275108920f5cb5d630f55fc9a3320c496d3 we require dmxproto 2.2.99.1. 121010 Complementing that is libdmx 1.0.99.1 with the dmxext.h header file. 121011 121012 Reported-by: Mark Rosenstand 121013 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121014 121015commit 24e640e1fa6cfb3917ba90dcdabf1fadddff20aa 121016Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121017Date: Fri Oct 2 19:20:24 2009 -0700 121018 121019 XQuartz: Fix a possible minor memory leak 121020 121021commit c4886fbabc1d8b4054654b227fcad83f58e8e798 121022Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121023Date: Fri Oct 2 18:58:35 2009 -0700 121024 121025 XQuartz: Send mouse location with scroll events. 121026 121027 This fixes the problem where (0, 0) was sent as the mouse location with scroll 121028 button events causing the event to not reach the client. 121029 121030commit 4c6bfa2c09ae2b0cffdf9211a6dfbcaefe0366b5 121031Author: Ian Romanick <ian.d.romanick@intel.com> 121032Date: Thu Oct 1 10:39:19 2009 -0700 121033 121034 GLX: More clearly document the GLX protocol version handling 121035 121036 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 121037 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 121038 121039commit 45f447dafded5adfe11b7df3325c2d8f6ae0639b 121040Author: Peter Hutterer <peter.hutterer@who-t.net> 121041Date: Thu Oct 1 21:19:11 2009 +1000 121042 121043 dix: force a minimum of 0 for screen coordinates. 121044 121045 Currently the root coordinates may fall into ]-1..0] if the subpixel 121046 remainder is less than 0. Screen coordinates mustn't go below 0, so use 121047 miPointerSetPosition to cap off the remainder if the coordinates are below 121048 0. 121049 121050 This is cheating a bit, a more comprehensive solution to deal with subpixels 121051 correctly when crossing screens is needed. For now, this'll do. 121052 121053 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121054 Acked-by: Simon Thum <simon.thum@gmx.de> 121055 121056commit 64fe5784b49347e1fd27b0c463be5c16557594c9 121057Author: Peter Hutterer <peter.hutterer@who-t.net> 121058Date: Fri Oct 2 09:34:24 2009 +1000 121059 121060 configure: if xnest was requested but modules weren't found, fail. 121061 121062 Tested-by: Xavier Chantry <shiningxc@gmail.com> 121063 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121064 Acked-by: Julien Cristau <jcristau@debian.org> 121065 121066commit 3ebb82d61c2b56e8f7145443a552a4e913bbfc80 121067Author: Dave Airlie <airlied@redhat.com> 121068Date: Thu Oct 1 15:25:55 2009 +1000 121069 121070 rotate: drop unwrapping inside block handler. 121071 121072 Keith has shown half the block handlers wrappers are wrong, also 121073 dynamic wrapping/unwrapping from what I can see will happen after 121074 the drivers, so its really accidental ABI, that we can't change 121075 now without modifing drivers. So be safe for 1.7. 121076 121077 Signed-off-by: Dave Airlie <airlied@redhat.com> 121078 Declared-as-sane-by: Keith Packard <keithp@keithp.com> 121079 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121080 121081commit cb54cf1b3e8c4109541cfb698542c00f2473e731 121082Author: Dave Airlie <airlied@linux.ie> 121083Date: Wed Sep 30 11:33:16 2009 +1000 121084 121085 glx: fixup deref of null pointer when glx screen init fails. 121086 121087 I think this is what the original author wanted. 121088 121089 Signed-off-by: Dave Airlie <airlied@redhat.com> 121090 Acked-by: Ian Romanick <ian.d.romanick@intel.com> 121091 121092commit 3d7cf468df96c0130a862f0d93fec990b9110c2f 121093Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121094Date: Thu Oct 1 11:58:18 2009 -0700 121095 121096 XQuartz: Update version strings to be X11R7.5 and the bundle 2.5.0 121097 121098 Is fink really _still_ relying on this X11R7.x version string? 121099 121100commit 85b831f701d085e514afcf8a5ce9b11f8a57984e 121101Author: Julien Cristau <jcristau@debian.org> 121102Date: Thu Oct 1 17:27:11 2009 +0200 121103 121104 xfree86: fix xorg.conf manpage formatting error 121105 121106commit 9bc7cbf9c02656982c2525836b5498993f708e02 121107Author: Nirbheek Chauhan <nirbheek@gentoo.org> 121108Date: Tue Sep 29 22:35:06 2009 +0530 121109 121110 xkb: check permissions on XKM_OUTPUT_DIR 121111 121112 Checking just for root is insufficient since that does not guarantee write/read 121113 permissions in XKM_OUTPUT_DIR (for example with sandbox). 121114 121115 Check if we can write a file, as well as read it later. Otherwise, invoke the 121116 fallback to /tmp 121117 121118 Signed-off-by: Nirbheek Chauhan <nirbheek@gentoo.org> 121119 Signed-off-by: Rémi Cardona <remi@gentoo.org> 121120 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121121 121122commit 622fc98fd08aba98369e6933c3ab8c9ff85385d5 121123Author: Peter Hutterer <peter.hutterer@who-t.net> 121124Date: Thu Oct 1 10:03:42 2009 +1000 121125 121126 render: Fix crash in RenderAddGlyphs (#23645) 121127 121128 This patch fixes two bugs: 121129 size is calculated as glyph height * padded_width. If the client submits 121130 garbage, this may get above INT_MAX, resulting in a negative size if size is 121131 unsigned. The sanity checks don't trigger for negative sizes and the server 121132 goes and writes into random memory locations. 121133 121134 If the client submits glyphs with a width or height 0, the destination 121135 pixmap is NULL, causing a null-pointer dereference. Since there's nothing to 121136 composite if the width/height is 0, we might as well skip the whole thing 121137 anyway. 121138 121139 Tested with Xvfb, Xephyr and Xorg. 121140 121141 X.Org Bug 23645 <http://bugs.freedesktop.org/show_bug.cgi?id=23645> 121142 121143 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121144 Reviewed-by: Keith Packard <keithp@keithp.com> 121145 121146commit 758ab55d2defc78d0169fd61a7036eb9f889e9e7 121147Author: Peter Hutterer <peter.hutterer@who-t.net> 121148Date: Thu Oct 1 15:22:19 2009 +1000 121149 121150 render: set the glyph picture to NULL by default. 121151 121152 In a follow-up patch we may have glyphs with a NULL picture. To cope with 121153 that, always set the pictures for glyphs to NULL at creation time and cope 121154 with cleaning up such glyphs. Also, since compositing a NULL source doesn't 121155 do a lot anyway, skip trying to do so. 121156 121157 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121158 Reviewed-by: Keith Packard <keithp@keithp.com> 121159 121160commit b5fcc5553eb784c9f4826936e839079c0cdee55a 121161Author: Michel Dänzer <daenzer@vmware.com> 121162Date: Thu Oct 1 15:17:11 2009 +1000 121163 121164 exa: avoid infinite loops if UTS sw fallbacks. 121165 121166 The upload in finish access can cause an infinite loop if 121167 UTS returns FALSE in here. 121168 121169 Fixes fd.o bug #24246. 121170 121171 Signed-off-by: Dave Airlie <airlied@redhat.com> 121172 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121173 121174commit db98b26ee145f70e732e2cf4a6ac3de77fdf4adc 121175Author: Keith Packard <keithp@keithp.com> 121176Date: Wed Sep 30 11:40:19 2009 -0700 121177 121178 Re-fix DGA removal. 121179 121180 Removing DGA ended up breaking any drivers calling into the old 121181 xf86DiDGAInit function as it tried to see if DGA was already enabled 121182 and ended up crashing if the VT wasn't completely initialized. Oops. 121183 121184 Also, if the driver initializes DGA itself, have the DiDGA 121185 initialization overwrite that information as the DiDGA code will call 121186 ReInit on mode detect. 121187 121188 Signed-off-by: Keith Packard <keithp@keithp.com> 121189 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121190 121191commit 873467adad479be02cd9cc6b43685919f5612d91 121192Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121193Date: Wed Sep 30 00:22:02 2009 -0700 121194 121195 XQuartz: Set the proper bitmap for key repeats... 121196 121197 XkbSetRepeatKeys lies and doesn't do what it says it will... 121198 121199commit 7bef78e199a4e4f7916be506807513efcd8fd4cb 121200Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121201Date: Tue Sep 29 23:40:47 2009 -0700 121202 121203 xkb: Use XkbPerKeyBitArraySize instead of hardcoded value of 32 121204 121205commit 4970666827e65424ee17ccf6341ff84aac974383 121206Author: Kim Woelders <kim@woelders.dk> 121207Date: Tue Sep 29 20:31:45 2009 +0200 121208 121209 dix: Fix potential memory corruption in doListFontsWithInfo. 121210 121211 Signed-off-by: Kim Woelders <kim@woelders.dk> 121212 Reviewed-by: Keith Packard <keithp@keithp.com> 121213 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121214 121215commit 6ffda5aae75272fabdc27d6f693ae827be119e95 121216Author: Dave Airlie <airlied@redhat.com> 121217Date: Tue Sep 29 11:49:09 2009 +1000 121218 121219 dix/glx/composite: consolidate visual resize in one place. 121220 121221 The previous code was copied and in both cases incorrectly fixed 121222 up the colormaps after resizing the visuals, this patch consolidates 121223 the visual resize + colormaps fixups in one place. This version 121224 also consolidates the vid allocation for the DepthPtr inside the 121225 function. 121226 121227 I'm not 100% sure colormap.[ch] is the correct place for this but 121228 visuals are mostly created in fb and I know thats not the place to 121229 be resizing them. 121230 121231 Fixes fd.o bug #19470. 121232 121233 Signed-off-by: Dave Airlie <airlied@redhat.com> 121234 Reviewed-by: Keith Packard <keithp@keithp.com> 121235 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121236 121237commit ad5c0d9efa47476ed5cf75c82265c73919e468b4 121238Author: Ian Romanick <ian.d.romanick@intel.com> 121239Date: Tue Sep 29 16:43:43 2009 -0700 121240 121241 GLX: Enable GLX 1.4 on DRI2 121242 121243 Return the minimum GLX version supported by all screens. Assume that 121244 DRI2 screens have all the required features for GLX 1.4. Assume that 121245 everyone else can only support GLX 1.2. 121246 121247 Reviewed-by: Kristian Høgsberg <krh@redhat.com> 121248 Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org> 121249 121250commit f11a356bcef1bc0a6440325019d5967b745a42dd 121251Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121252Date: Mon Sep 28 17:05:29 2009 -0700 121253 121254 XQuartz: Cleaned up keymap setting for easier maintenance 121255 (cherry picked from commit b9dfed9e88389cbd29406a20d38ee4297638649b) 121256 121257commit f3223c71cfc638e695981e527517d48ea00d124d 121258Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121259Date: Mon Sep 28 23:01:59 2009 -0700 121260 121261 XQuartz: Remove the redundant xquartz_resetenv_display 121262 unsetenv(DISPLAY) takes care of this for us anyway 121263 (cherry picked from commit d2263645d839c9edeedea0835d26f1f41b37f70e) 121264 121265commit 9b98b883227ed23d5470e8de689afeec4a0fd742 121266Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121267Date: Mon Sep 28 17:47:31 2009 -0700 121268 121269 XQuartz: Query the BundleIdentifier from the bundle in X11.bin rather than using the configure option. 121270 121271 This lets X11.bin drop into any .app ... the Info.plist and Xquartz binary need to have it hardcoded still. 121272 (cherry picked from commit 9ad16b8e50b13eb6d0cd20386d07aa8d7320f671) 121273 121274commit 11817a881cb93a89788105d1e575a468f2a8d27c 121275Author: Michel Dänzer <daenzer@vmware.com> 121276Date: Tue Sep 29 08:56:59 2009 +0200 121277 121278 Fix ShmPutImage non-ZPixmap case. 121279 121280 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=23298 . 121281 121282commit 19be992d9dc542b61fa3f4fd32a09071c9e64880 121283Author: Peter Hutterer <peter.hutterer@who-t.net> 121284Date: Mon Sep 28 14:18:45 2009 +1000 121285 121286 ephyr: if -parent is given, check for a trailing -screen. (#24144) 121287 121288 If -parent is given, don't open up a new window if -screen is given as well. 121289 The commandline option -screen allows to set the depth of the embedded 121290 Xephry instance, even though width and height are autoscaled on -parent. 121291 121292 This patch checks for a -screen parameter after -parent and - if one is 121293 found - delays initializing the screen. The parent window id is stored 121294 temporarily but re-set after a -screen argument. 121295 The following command is thus valid: 121296 121297 Xephyr -parent 1234 -screen 640x480@8 -screen 1024x768 121298 121299 It embeds the first 8-bit screen into window 1234 and opens up a new window 121300 for the second screen. Multiple parent arguments are possible, the screens 121301 are embedded in-order. 121302 121303 X.Org Bug 24144 <http://bugs.freedesktop.org/show_bug.cgi?id=24144> 121304 121305 Tested-by: Vic Lee 121306 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121307 121308commit efacd7bfd08ffc0725de6f639c6afbf3b2f6c9fe 121309Author: Ben Byer <bbyer@freedesktop.org> 121310Date: Sun Sep 27 10:35:53 2009 -0700 121311 121312 Add (ok, fix) support for DTrace under OS X 121313 (cherry picked from commit 8428a57184f542941d2c8c90e97d18e111a69dd2) 121314 121315commit b3415187e92960cbff784108b5a3a8d130dc34c5 121316Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121317Date: Sun Sep 27 23:09:51 2009 -0700 121318 121319 Rootless: Correct border rendering on parent-relative windows 121320 121321 Resurected code from the punted RootlessPaintBackground/Border and added it conditionally to miPaintWindow 121322 (cherry picked from commit cf2e3312cff3f341e9edba8c321a4ca7ffd8748e) 121323 121324commit 6df00917cab5c1096070625385fd76ee6c52e0f1 121325Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121326Date: Sun Sep 27 23:00:59 2009 -0700 121327 121328 XQuartz: Fix QuartzSetCursor to match the expected prototype. 121329 (cherry picked from commit dadab5a2279a19dcf709402d7f22f0cd48670db0) 121330 121331commit 67a51cd9ef60b4c9f25fdde84f9eb352936c8c67 121332Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121333Date: Sun Sep 27 22:57:25 2009 -0700 121334 121335 XQuartz: Fix a bunch of compilation warnings about style 121336 (cherry picked from commit 54000bdcbca52a2de31f7c1a1147de6d8e9dbbb8) 121337 121338commit 96780eaf32636c94a9cf33f22eb9f01d984ff754 121339Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121340Date: Sun Sep 27 22:29:49 2009 -0700 121341 121342 XQuartz: Nuke TSM 121343 121344 It's deprecated in SnowLeopard. Ben and I both have no idea what it is for. It says something about unicode input, but urxvt seems fine taking in unicode, so /shrug... bye. 121345 (cherry picked from commit 29cb904e4de2411a9b6dbe68694954788f0525f7) 121346 121347commit 15e15816a2f011d0aeeaff9e394d30a147c973ce 121348Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121349Date: Sun Sep 27 22:26:22 2009 -0700 121350 121351 XQuartz: Fix inverse map from mode_switch to alt 121352 (cherry picked from commit de6cee11e1c335a0e5f708e7641e81d3cfe52529) 121353 121354commit 558d803b2966c2e44345a80e635e091dc3e49f02 121355Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121356Date: Sun Sep 27 18:25:57 2009 -0700 121357 121358 XQuartz: Force a keymap resync on the first keypress to workaround XKB mucking with our keymap. 121359 121360 We need to find a better way to work with XKB on this. 121361 (cherry picked from commit ceaa5c779ceed3de5ea53727649613be3133b24e) 121362 121363commit 226b1033b48807fd6871ba626e20ef1411904939 121364Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121365Date: Sun Sep 27 17:08:12 2009 -0700 121366 121367 XQuartz: Transition from xEvent based mieq to InternalEvent 121368 (cherry picked from commit a3dbde2de87ee4f577748a8c447501a3ea462559) 121369 121370commit 0c2731596f27f2cdf5000ba41de37e7eb86ad6f9 121371Author: Peter Hutterer <peter.hutterer@who-t.net> 121372Date: Mon Sep 28 15:07:48 2009 +1000 121373 121374 Put tests for zero-sized strings in quotes (#24060) 121375 121376 X.Org Bug 24060 <http://bugs.freedesktop.org/show_bug.cgi?id=24060> 121377 121378 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121379 121380commit 83023ffd09a84ff48e6b99f57ebad101a00478db 121381Author: Peter Hutterer <peter.hutterer@who-t.net> 121382Date: Mon Sep 28 13:59:25 2009 +1000 121383 121384 xfree86: use the DDC size if either width or height of DisplaySize is bogus. 121385 121386 If either width or height of DisplaySize is invalid, assume that the 121387 configuration is invalid and use the DDC-reported values instead. 121388 121389 See Comment 9, Bug 9758. 121390 http://bugs.freedesktop.org/show_bug.cgi?id=9758#c9 121391 121392 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121393 Acked-by: Dave Airlie <airlied@redhat.com> 121394 121395commit f772014c435f56db56520ca13ffa39431684f122 121396Author: Peter Hutterer <peter.hutterer@who-t.net> 121397Date: Wed Sep 23 10:53:51 2009 +1000 121398 121399 render: Plug a memory leak in AddGlyph. (#23286) 121400 121401 AddGlyph was missing the FreePicture() call that DeleteGlyph used, resulting 121402 in a memory leak when more than one Glyph was added in a RenderAddGlyphs 121403 request. 121404 121405 Since the code in AddGlyph and DeleteGlyph is identical, move into a static 121406 function to avoid such mistakes in the future. 121407 121408 X.Org Bug 23286 <http://bugs.freedesktop.org/show_bug.cgi?id=23286> 121409 121410commit 3db28f92b0c810b452506abbed299a204c90ba0b 121411Author: Tilman Sauerbeck <tilman@code-monkey.de> 121412Date: Thu Sep 24 12:57:02 2009 +0200 121413 121414 configure: make XNEST default to auto. 121415 121416 When we're checking whether to build Xnest, we're comparing the 121417 variable to auto but before it never was assigned that. 121418 121419 Signed-off-by: Tilman Sauerbeck <tilman@code-monkey.de> 121420 121421 [Xnest was enabled to yes to increase build exposure and catch compiler 121422 errors early. The requirements to Xnest are quite low and I expect most 121423 developers have them, so Xnext will be enabled on most boxes. Anyone 121424 missing those requires probably doesn't want to build Xnest anyway.] 121425 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121426 121427commit 78ad6ca9a97440b74019c00a28144ea7d1e03431 121428Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 121429Date: Sun Sep 27 17:30:23 2009 +0200 121430 121431 xfree86: Hurd fix 121432 121433 I hadn't paid attention that the parameters order had changed, here is a 121434 trivial patch, please apply. 121435 121436 Signed-off-by: Julien Cristau <jcristau@debian.org> 121437 121438commit fb7938315bed9d4cb5641ee9acebf78a13bc109e 121439Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121440Date: Sun Sep 27 09:48:43 2009 -0700 121441 121442 Xi: Make CopyKeyClass X_HIDDEN to avoid ugly ifdef-fu 121443 (cherry picked from commit 6d436e17a9ae7f4ce8537f3fabc052d4f07ca75f) 121444 121445commit a2cd21177859eb45320a94c96dde7d9058ab7650 121446Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121447Date: Sat Sep 26 23:41:45 2009 -0700 121448 121449 XQuartz: Stop checking version numbers of the bundle because CFBundleGetVersionNumber is gimpish 121450 (cherry picked from commit 9f5bdd89608ec12012592ff395b82e954fbb4da8) 121451 121452commit fbfbf93405222322571e15e78dea57c8f2755eae 121453Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121454Date: Sat Sep 26 22:52:39 2009 -0700 121455 121456 XQuartz: Add pressure/tilt property labels 121457 (cherry picked from commit 84ea67130ef6b4086042aad6036ce66f93ea3e56) 121458 121459commit 65f14840898dd7e9d61d8c966f3ff11ae3e82d4f 121460Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121461Date: Sat Sep 26 22:48:49 2009 -0700 121462 121463 XQuartz: Fix a brain-o array indexing problem 121464 121465 /bop Peter 121466 (cherry picked from commit 494a6b046a258ad83dc98eb92b7c3d8f1d2626bb) 121467 121468commit 7159381881e5da33a49e8735811b93cbfe4f681e 121469Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121470Date: Sat Sep 26 22:17:09 2009 -0700 121471 121472 XQuartz: Nuke duplicate locks that make painful headaches 121473 (cherry picked from commit 1dd56322bd1722f2427fb2d833c5608248b60cf0) 121474 121475commit fba8c702ba6aa8fac27682e0838fd744e197d54f 121476Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121477Date: Sat Sep 26 22:13:49 2009 -0700 121478 121479 Xi: CopyKeyClass is not static for XQuartz 121480 (cherry picked from commit 85d6402354cdf143c6490f2725744c2f08b5605b) 121481 121482commit dc1e1bebff8b9d1eb6196dfd4b2f52f1e4200efa 121483Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121484Date: Fri Aug 7 00:02:21 2009 -0700 121485 121486 XQuartz: Use internal xshm header for new xextproto 121487 (cherry picked from commit 1755239330c0bdac820d88a3e06ff391d65f29be) 121488 (cherry picked from commit 697be460d0e555e2c75eed6889293650e02d423c) 121489 121490commit 73ae547d5e687ef10dea45801fc627e10ac4b659 121491Author: Michel Dänzer <daenzer@vmware.com> 121492Date: Sun Sep 27 02:08:10 2009 +0200 121493 121494 EXA: Fix mixed pixmaps crash with missing / failing UploadToScreen hook. 121495 121496 For the recent mixed pixmaps changes, I failed to realize (or hit in my 121497 testing) a problem which can occur if the driver doesn't provide an 121498 UploadToScreen hook or provides one which can fail: There can be a crash 121499 in exaMemcpyBox() because exaCopyDirtyToFb() passes pExaPixmap->fb_ptr to 121500 exaCopyDirty(), but that's normally NULL with driver allocated pixmaps. 121501 121502 The solution is to make exaCopyDirty*() no longer rely on pExaPixmap->fb_ptr 121503 but use pPixmap->devPrivate.ptr after PrepareAccess instead. 121504 121505 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=24167 . 121506 121507commit c9ec2bab2f258798fd6e6676698c732f09571a60 121508Author: Kevin E Martin <kem@redhat.com> 121509Date: Sat Sep 26 13:10:54 2009 +1000 121510 121511 dmx: undefine MITSHM, move undefs to miinitext.c. 121512 121513 This patch undefines MITSHM for dmx - we don't support the required 121514 screen->ModifyPixmapHeaders. All undefines are moved from dmx-config to 121515 miinitext.c, where they belong. 121516 121517 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121518 121519commit fc9d733bab3ff0e4e51b19c73b66196dca563a70 121520Author: Kevin E Martin <kem@redhat.com> 121521Date: Sat Sep 26 13:09:52 2009 +1000 121522 121523 dmx: reshuffle linker order to avoid errors when MITSHM is undefined. 121524 121525 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121526 121527commit e7c2598f565e8252dd66ee3e6212b310856476cb 121528Author: Peter Hutterer <peter.hutterer@who-t.net> 121529Date: Fri Sep 25 14:53:33 2009 +1000 121530 121531 dmx: core events are always in screen coordinates when passed to GPE. 121532 121533 This fixes input in dmx, the pointer appears at the right positions to the 121534 clients now. 121535 121536 Also mark the spot where we pass in the button state as valuator to GPE 121537 with a FIXME. (??) 121538 121539 Tested-by: Kevin Martin 121540 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121541 121542commit 43a2eb794f19a2ba56d653f465fc5f6b2ff0d3d3 121543Author: Peter Hutterer <peter.hutterer@who-t.net> 121544Date: Wed Sep 23 21:49:11 2009 +1000 121545 121546 configure: Unify all library defines that require a specific version. 121547 121548 This patch moves all libraries that require a specific version into a single 121549 location instead or duplicating them across the configure.ac file. 121550 Libraries that do not require specific versions are left where they are. 121551 121552 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121553 121554commit 9bd08c690fc687c4d69bb70536f3079a9184476d 121555Author: Peter Hutterer <peter.hutterer@who-t.net> 121556Date: Thu Sep 24 14:07:26 2009 +1000 121557 121558 Xi: update axisVals with the right subpixel data. 121559 121560 Subpixel data in data_frac is stored as FP32.32, hence we need to get that 121561 down again before adding it to the current value. 121562 121563 Reported-by: Thomas Jaeger 121564 Tested-by: Thomas Jaeger 121565 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121566 121567commit abb8108fb44cc9b08fe4ae7d805a0a22564ca6bb 121568Author: Michel Dänzer <daenzer@vmware.com> 121569Date: Sat Sep 26 02:05:24 2009 +0200 121570 121571 EXA: Only calculate cache position once for each glyph. 121572 121573commit c11678cc189551f2a01eaa7a63969c16950739b4 121574Author: Michel Dänzer <daenzer@vmware.com> 121575Date: Sat Sep 26 02:05:06 2009 +0200 121576 121577 EXA: Accumulate arbitrary number of glyphs without flushing. 121578 121579commit 0369eeeb6bf8a808fa2df503fc8b8df81e6e07b8 121580Author: Michel Dänzer <daenzer@vmware.com> 121581Date: Sat Sep 26 02:04:48 2009 +0200 121582 121583 EXA: Try to minimize UploadToScreen calls for mixed pixmaps. 121584 121585 If there are several software fallbacks affecting the system memory copy of the 121586 same pixmap, only copy the results back to the driver pixmap when it's used for 121587 acceleration again, or in the BlockHandler, whichever happens first. 121588 121589commit 1818cbd70fc1f2e1487b4c678e67e28f1265c0ef 121590Author: Michel Dänzer <daenzer@vmware.com> 121591Date: Sat Sep 26 01:59:39 2009 +0200 121592 121593 EXA: Extend mixed pixmaps scheme to allow driver PrepareAccess hook to fail. 121594 121595 If the PrepareAccess hook fails, use the DownloadFromScreen hook to retrieve 121596 driver pixmap contents to a system RAM copy, perform software rendering on that 121597 and copy the results back using the UploadToScreen hook. Use the classic 121598 migration logic to minimize transfers (which as a bonus allows slightly 121599 cleaning up some of the existing mixed pixmap code). 121600 121601 This enables things that weren't possible before with driver-allocated pixmap 121602 storage: If some (or all) GPU pixmap storage can't be mapped directly by the 121603 CPU, this can be handled between the PrepareAccess and 121604 DownloadFrom/UploadToScreen hooks, e.g.: 121605 121606 * Radeon KMS on big endian machines can fail PrepareAccess if the pixmap 121607 requires byte-swapping and swap bytes in DownloadFrom/UploadToScreen. 121608 * Environments where GPU and CPU don't have a shared address space at all. 121609 Here the driver PrepareAccess hook will always fail and leave all transfers 121610 between GPU / CPU storage to the Download/From/UploadToScreen hooks. 121611 121612 Drivers which can handle all pixmaps in the PrepareAccess hook should notice 121613 little if any difference. 121614 121615commit e23bffc41b007f1bc2b8f5cd4ac54213062c95cc 121616Author: Alan Coopersmith <alan.coopersmith@sun.com> 121617Date: Tue Sep 15 19:17:48 2009 -0700 121618 121619 Fix build of unit tests when dtrace probes are enabled 121620 121621 ar loses the dtrace probe magic when building static libraries, so we 121622 have to link with the .O files in order to resolve the dtrace probe symbols. 121623 121624 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 121625 121626commit 9fa73be9fa543a686ea35c861084f5af37d44caa 121627Author: Peter Hutterer <peter.hutterer@who-t.net> 121628Date: Tue Sep 22 20:34:54 2009 +1000 121629 121630 Require libXext >= 1.0.99.4 121631 121632 Reported-by: Tilman Sauerbeck 121633 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121634 121635commit a9c274df5c37cb4ece6449e934342d8ff8e61705 121636Author: Peter Hutterer <peter.hutterer@who-t.net> 121637Date: Wed Sep 23 11:43:27 2009 +1000 121638 121639 kdrive: plug two memory leaks when freeing the KdKeyboard/Pointer. 121640 121641 xkbRules, xkbModel and xkbLayout are strdup'd in KdNewKeyboard, need to be 121642 freed. 121643 121644 The ephyr driver strdups the name on top of the already allocated 121645 kdrive-assigned name. Memory must be freed beforehand. 121646 121647 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121648 121649commit fd913136732ff14a0484ca28f60ac1fbf49be81d 121650Author: Peter Hutterer <peter.hutterer@who-t.net> 121651Date: Wed Sep 23 11:44:12 2009 +1000 121652 121653 dix: plug memory leak in DeviceEnterLeaveEvents. 121654 121655 'event' must be freed before exiting. 121656 121657 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121658 121659commit 6ee796e9bb4e46782b50a69c7b4fa5b49576f139 121660Author: Peter Hutterer <peter.hutterer@who-t.net> 121661Date: Thu Sep 24 14:05:52 2009 +1000 121662 121663 Xi: fix length calculation for ValuatorState in QueryDeviceState reply. 121664 121665 The length field needs to include the bytes required for the valuators 121666 (INT32) as well. 121667 121668 The reply length has the right value and since the valuator state is always 121669 last, clients didn't notice the wrong offset. 121670 121671 Tested-by: Thomas Jaeger 121672 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121673 121674commit 3b5bbb149d4c932d9624336f5cbe9fe71c87bea3 121675Author: Peter Hutterer <peter.hutterer@who-t.net> 121676Date: Wed Sep 23 12:32:44 2009 +1000 121677 121678 configure: fix up tslib check once again. 121679 121680 This patch addresses two issues: 121681 The check for HAVE_TSLIB = xauto can never be true, the check has been 121682 corrected to TSLIB = xauto. 121683 121684 Pre-pkgconfig versions of tslib fail to be found, this patch restores the 121685 additional AC_CHECK_LIB. However, the pgk-config check must happen before 121686 AC_CHECK_LIB, as AC_CHECK_LIB does not seem to honour the LD_LIBRARY_PATH. 121687 Thus, if tslib is installed outside of the default paths, AC_CHECK_LIB 121688 fails. 121689 121690 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121691 Acked--by: Daniel Stone <daniel@fooishbar.org> 121692 121693commit 5402f18d9c3f7ba19cc05b3a814e3a9e94c8d551 121694Author: Thomas Jaeger <thjaeger@gmail.com> 121695Date: Tue Sep 22 20:16:21 2009 -0400 121696 121697 dix: report XI1 axis values correctly if first_valuator != 0 121698 121699 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 121700 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121701 121702commit 33bf9cb69dccbb6889b2f83e6db61f40dc644e17 121703Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 121704Date: Wed Sep 23 17:41:28 2009 -0700 121705 121706 XQuartz: GLX capabilities: Allow 16bit accumulation buffers 121707 121708 http://xquartz.macosforge.org/trac/ticket/308 121709 (cherry picked from commit e9e63a2118b76b6c31c4081fec08a99e4d796e22) 121710 121711commit ce1fe8ddb4a4dbe6cfd909e5b1b73b459d742bec 121712Author: Michel Dänzer <daenzer@vmware.com> 121713Date: Wed Sep 23 13:10:05 2009 +0200 121714 121715 render: Don't add b8g8r8x8 format for depth 24. 121716 121717 The components are required to be packed in the bottom of the pixel, so this 121718 format can't fit in depth 24. 121719 121720 Also fix up a comment for the addition of BGRA formats. 121721 121722commit 096f21bb7a1217443d8a03529b1a2938518eb24f 121723Author: Michel Dänzer <daenzer@vmware.com> 121724Date: Wed Sep 23 08:24:06 2009 +0200 121725 121726 EXA: Fix some issues pointed out by clang. 121727 121728 Remove dead variables, fix use of uninitialized values, that kind of thing. 121729 121730commit 824a09d856a5f750694e11d2fd2faaa3de705eaa 121731Author: Simon Thum <simon.thum@gmx.de> 121732Date: Mon Sep 21 15:23:27 2009 +0200 121733 121734 dix: move bounds check before access 121735 121736 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121737 121738commit 9edb9e9b4dde6f73dc5241d078425a7a70699ec9 121739Author: Peter Hutterer <peter.hutterer@who-t.net> 121740Date: Tue Sep 22 12:56:17 2009 +1000 121741 121742 Bump to 1.7.99.1 121743 121744 X Server 1.7 has branched off. 121745 121746 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121747 121748commit 66ece3bfb19642bf0f1ef73c3caa8a99c3dfb5a9 121749Author: Kevin E Martin <kem@redhat.com> 121750Date: Tue Sep 22 11:42:24 2009 +1000 121751 121752 dmx: only free the default pixmaps that we actually allocated. 121753 121754 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121755 121756commit 20fb8c2e2fccb28c1b89e6e2a5c685e1287baedb 121757Author: Kevin E Martin <kem@redhat.com> 121758Date: Tue Sep 22 11:41:33 2009 +1000 121759 121760 dmx: disable Get/SetWindowPixmap, we don't support composite/redirected windows 121761 121762 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121763 121764commit fcdc1d78cca3b8bb6b77d53eda7e21d649df6943 121765Author: Keith Packard <keithp@keithp.com> 121766Date: Mon Sep 21 19:16:59 2009 +0200 121767 121768 Fix sporadic segfault on resume due to accidentally freeing cursor. 121769 121770 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121771 121772commit 91e1fe5863a180b2d70a6943e83b98bbd4c9ebce 121773Author: Matthias Hopf <mhopf@suse.de> 121774Date: Mon Sep 21 19:13:31 2009 +0200 121775 121776 Revert fe31f9c + 977953b to fix issue for good. 121777 121778 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121779 121780commit 90aa0e4a49cdd637178a771365e1e8dab8bceb87 121781Author: Peter Hutterer <peter.hutterer@who-t.net> 121782Date: Fri Sep 18 10:03:02 2009 +1000 121783 121784 input: don't use typecasts to access members of InternalEvent. 121785 121786 To avoid confusion, the member names are now postfixed with _event. 121787 121788 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121789 121790commit 693d4fdb1c7b7c789a812790122454e718602449 121791Author: Peter Hutterer <peter.hutterer@who-t.net> 121792Date: Tue Sep 22 11:13:58 2009 +1000 121793 121794 Xext: switch mbuf.c to dixLookupResourceByType 121795 121796 Resolves a linker error caused by LookupIDByType. 121797 121798 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121799 121800commit 493d2f9c2fc56dd3b4c03451c07cd25e3012142e 121801Author: Peter Hutterer <peter.hutterer@who-t.net> 121802Date: Tue Sep 22 11:05:07 2009 +1000 121803 121804 Xext: remove DisplayImageBuffers from mbuf.c 121805 121806 Not referenced by anything. 121807 121808 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121809 121810commit fff40b3353c7025c1d1e3dd14b5c78f7c9f5de7c 121811Author: Peter Hutterer <peter.hutterer@who-t.net> 121812Date: Tue Sep 22 10:53:47 2009 +1000 121813 121814 Xext: fix up multibuffer compiler errors. 121815 121816 Triggered by the xextproto 7.1 change, fixed by moving the matching 121817 declarations from the header file to here. 121818 121819 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121820 121821commit 4c8f834da6c8c84849313453fe223a8165c6afc0 121822Author: Peter Hutterer <peter.hutterer@who-t.net> 121823Date: Mon Sep 21 14:52:31 2009 +1000 121824 121825 configure: fix up check for tslib. 121826 121827 Reduce the tslib-check to the pkg-config check only instead of the previous 121828 library symbol check followd by a pkg-config check. 121829 121830 This patch also reduces the required version of tslib back down to 121831 tslib-0.0. Unfortunately, the 1.0 tarball available through 121832 http://tslib.berlios.de/ still announces itself as 0.0.2. 121833 121834 Reported-by: Werner Landgraf 121835 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121836 121837commit 977953bf14858d17c22208c848854ddbe7e86527 121838Author: Matthias Hopf <mhopf@suse.de> 121839Date: Mon Sep 21 15:44:00 2009 +0200 121840 121841 Less intrusive workaround for sporadic segfault on resume. 121842 121843 Hopefully fixes fdo #24010 (memleak). 121844 121845commit 55747d256d759850141e4a9c4dec965616a31dc8 121846Author: Peter Hutterer <peter.hutterer@who-t.net> 121847Date: Fri Sep 18 16:27:54 2009 +1000 121848 121849 input: define server-supported protocol versions in one single file. 121850 121851 include/protocol-versions.h specifies each extension version as supported by 121852 the server and sent back on the wire to the client. 121853 121854 This fixes up several issues with the server potentially reporting a higher 121855 version of the protocol if recompiled against a newer version of the 121856 protocol. 121857 121858 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121859 Acked-by: Rémi Cardona <remi@gentoo.org> 121860 Acked-by: Julien Cristau <jcristau@debian.org> 121861 121862commit 0b7c6c728c2e2d8433a188315cc591308a89cd85 121863Author: Keith Packard <keithp@keithp.com> 121864Date: Fri Sep 18 21:12:17 2009 -0700 121865 121866 xfree86/modes: Remove all framebuffer support from DGA 121867 121868 This removes all rendering and mapping code from xf86DiDGA, leaving 121869 just mode setting and raw input device access. The mapping code didn't 121870 have the offset within /dev/mem for the frame buffer and the pixmap 121871 support assumed that the framebuffer was never reallocated. 121872 121873 Signed-off-by: Keith Packard <keithp@keithp.com> 121874 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121875 121876commit d3a3c904e5e3b5e8bb3c51ca6a563880ff589bf3 121877Author: Peter Hutterer <peter.hutterer@who-t.net> 121878Date: Mon Sep 21 09:41:46 2009 +1000 121879 121880 configure: resourceproto is optional. 121881 121882 Although a --disable-xres flag is provided, resourceproto was listed in 121883 REQUIRED_MODULES since 2005 (5b218617fa8ba52bf65aef35da39e06c662495e6). 121884 Remove, make conditional on the xres flag only. 121885 121886 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121887 121888commit b9ae1b91f316a98eb3bbefe1fc61f3e0faa861e0 121889Author: Peter Hutterer <peter.hutterer@who-t.net> 121890Date: Mon Sep 21 09:22:53 2009 +1000 121891 121892 dmx: silence 'implicit declaration of function ‘XFixesSetPictureClipRegion’ 121893 121894 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121895 121896commit aa71e268f55ac39351814ab6840196c4f58bad39 121897Author: Peter Hutterer <peter.hutterer@who-t.net> 121898Date: Fri Sep 18 11:35:47 2009 +1000 121899 121900 Xext: silence 'warning: no previous prototype' for sync, bigreq and xcmisc. 121901 121902 xcmisc and bigreq don't have their own header so just declare it here to 121903 shut up the compiler. 121904 121905 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121906 121907commit 00dfe69aab786ea650b41dfefc031f0e0fc72b0c 121908Author: Peter Hutterer <peter.hutterer@who-t.net> 121909Date: Thu Sep 17 14:48:51 2009 +1000 121910 121911 randr: reply with the server or client version, whichever is lower. 121912 121913 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121914 121915commit 8b75d0f9d347d7c3aad1e53809e0a2bed03b92fc 121916Author: Peter Hutterer <peter.hutterer@who-t.net> 121917Date: Thu Sep 17 14:17:54 2009 +1000 121918 121919 render: reply with the server or client version, whichever is lower. 121920 121921 Protocol requires that the lower of [server version, client version] is 121922 returned to the client. 121923 121924 The other part of the issue discussed in reply to [1] remains. 121925 [1] http://lists.freedesktop.org/archives/xorg-devel/2009-September/001990.html 121926 121927 Reported-by: Julien Cristau 121928 121929 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121930 121931commit 20ccc66708bcd55372bff03fa0d78a7846e8c82d 121932Author: Peter Hutterer <peter.hutterer@who-t.net> 121933Date: Sun Sep 20 21:12:31 2009 +1000 121934 121935 xfree86: silence warning: ‘ramplen’ may be used uninitialized 121936 121937 'ramplen' is only set and used if stuff->size > 0 but the compiler doesn't 121938 pick that up. 121939 121940 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121941 121942commit 2df10a49867635c341b1d1a9faadf4dbae043a0f 121943Author: Keith Packard <keithp@keithp.com> 121944Date: Thu Sep 17 18:14:38 2009 -0700 121945 121946 RRDestroyOutputProperty: Free randr property valid values 121947 121948 These were leaked when the property was destroyed. 121949 121950 Signed-off-by: Keith Packard <keithp@keithp.com> 121951 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121952 121953commit d3a84906c00b8fe4a4e33fa89bff3ed66ceafdec 121954Author: Keith Packard <keithp@keithp.com> 121955Date: Thu Sep 17 18:14:37 2009 -0700 121956 121957 Free randr crtc and output pointer arrays 121958 121959 All of the crts and outputs were freed, but not the arrays full of 121960 pointers to them. 121961 121962 Signed-off-by: Keith Packard <keithp@keithp.com> 121963 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121964 121965commit 6086a6065666fb9ded5d99919ddfa51c66b76246 121966Author: Keith Packard <keithp@keithp.com> 121967Date: Thu Sep 17 18:14:36 2009 -0700 121968 121969 xf8CrtcSetModeTransform: free adjusted_mode name too 121970 121971 The adjusted mode was freed, but any name allocated for that was leaked. 121972 121973 Signed-off-by: Keith Packard <keithp@keithp.com> 121974 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121975 121976commit 088107827347cee05a6e5317bcebdbc282b9a3bc 121977Author: Keith Packard <keithp@keithp.com> 121978Date: Thu Sep 17 18:14:35 2009 -0700 121979 121980 LoaderSortExtensions: free graph nodes 121981 121982 Signed-off-by: Keith Packard <keithp@keithp.com> 121983 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121984 121985commit d5959d37cd36895898786e4ea5bf72c480b79832 121986Author: Keith Packard <keithp@keithp.com> 121987Date: Thu Sep 17 18:14:34 2009 -0700 121988 121989 Free libpciaccess iterator in xf86PciProbe 121990 121991 Signed-off-by: Keith Packard <keithp@keithp.com> 121992 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 121993 121994commit ded35b7bf7a5ad27de152476c657a56a4f232800 121995Author: Keith Packard <keithp@keithp.com> 121996Date: Thu Sep 17 18:14:33 2009 -0700 121997 121998 xf86Xinput.c: get DIX event queue pointer once at InitInput time 121999 122000 The DIX event queue is allocated before InitInput is called, so fetch 122001 the pointer there and not randomly at other times. This avoids failing 122002 to fetch the pointer sometimes during server regen and then smashing 122003 memory through the stale pointer from the previous server generation. 122004 122005 Signed-off-by: Keith Packard <keithp@keithp.com> 122006 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122007 122008commit 54f15a414130f89832b5c2604b0b44e7e4c6709f 122009Author: Keith Packard <keithp@keithp.com> 122010Date: Thu Sep 17 18:14:32 2009 -0700 122011 122012 probe_devices_from_device_sections: don't leak list of devices 122013 122014 xf86MatchDevice returned malloc'd storage containing the list of 122015 devices to look at; make sure that gets freed. 122016 122017 Signed-off-by: Keith Packard <keithp@keithp.com> 122018 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122019 122020commit 8b5086250aa5dae8de8b763408ff480d7beac819 122021Author: Keith Packard <keithp@keithp.com> 122022Date: Thu Sep 17 18:14:31 2009 -0700 122023 122024 Eliminate bogus event resizing. 122025 122026 Now that all event queues hold internal events only, they never need 122027 to be resized. Resizing them led to memory corruption as they would 122028 get sized for an appropriate xEvent, not an internal event. 122029 122030 Signed-off-by: Keith Packard <keithp@keithp.com> 122031 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122032 122033commit c7987660fd665f32406b14db7078a3f5f7b76e9c 122034Author: Peter Hutterer <peter.hutterer@who-t.net> 122035Date: Thu Sep 17 15:33:19 2009 +1000 122036 122037 xfree86: fix up wrong use of OptionRec for AutoAddDevices andAutoEnableDevices 122038 122039 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122040 122041commit 971d4eac1274db78efa546ea50ca5e01a80bbd81 122042Author: Thomas Jaeger <ThJaeger@gmail.com> 122043Date: Sat Sep 19 15:48:30 2009 -0400 122044 122045 Xext: include security protocol header instead of client header 122046 122047 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 122048 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122049 122050commit a62664bc07b35aedb6b11b984c94341deb4486ca 122051Author: Thomas Jaeger <ThJaeger@gmail.com> 122052Date: Sat Sep 19 16:01:52 2009 -0400 122053 122054 Bump fixesproto requirement to 4.1 122055 122056 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 122057 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122058 122059commit 2a890d85e923e1da0331b02e58e95f98b11f5de0 122060Author: Thomas Jaeger <ThJaeger@gmail.com> 122061Date: Sun Sep 20 01:31:59 2009 -0400 122062 122063 fixes: Make sure invisible cursor is inialized with zero src/mask 122064 122065 This fixes a regression introduced in 9040dab76182d1a019ca7fef7b29733d2c199e61. 122066 122067 Signed-off-by: Thomas Jaeger <ThJaeger@gmail.com> 122068 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122069 122070commit 2bb3f0c87119e73157141431d7c00ab104d4b34b 122071Author: Peter Hutterer <peter.hutterer@who-t.net> 122072Date: Fri Sep 18 10:27:14 2009 +1000 122073 122074 xfree86: remove now unused variable 122075 122076 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122077 122078commit 1888d7d3e3ae81b5cb23e3a0852317362c8e215a 122079Author: Carl Worth <cworth@cworth.org> 122080Date: Fri Sep 18 10:39:19 2009 -0700 122081 122082 configure: Fix broken quoting for initialization of REQUIRED_MODULES 122083 122084 A recent change switched to C-style automatic string concatenation, 122085 (presumably for better readability and maintainability), but this 122086 doesn't actually work for the shell. So we revert to the previously 122087 working initialization. 122088 122089 This commit reverts one hunk of c947605eda082fbca0821efc0cc14fec8f958fd0 122090 122091commit 4982bbc3b49b1b61e63a8461ea803e41318c8187 122092Author: Tiago Vignatti <tiago.vignatti@nokia.com> 122093Date: Fri Sep 18 12:36:11 2009 +0300 122094 122095 xfree86: vgaarb: missing wrapping macro 122096 122097 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 122098 122099commit af1af50de621bc451e6413fa5602ae5db2dc9809 122100Author: Martin-Éric Racine <q-funk@iki.fi> 122101Date: Fri Sep 18 09:33:37 2009 +0300 122102 122103 xserver: added more Geode oddities to xf86AutoConfig.c 122104 122105commit ec0ad408ef8d5a2cf7a2bd65d13de3ef3147308c 122106Author: Peter Hutterer <peter.hutterer@who-t.net> 122107Date: Thu Sep 17 16:01:29 2009 +1000 122108 122109 xfree86: use SendDevicePresenceEvents instead of manual event handling. 122110 122111 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122112 122113commit d0440275108920f5cb5d630f55fc9a3320c496d3 122114Author: Peter Hutterer <peter.hutterer@who-t.net> 122115Date: Wed Sep 16 10:58:52 2009 +1000 122116 122117 Require dmxproto 2.2.99.1 122118 122119 This requirement was introduced in code with commit 122120 e85775aa64d1e9322c82090a26256af5e1fede0a. 122121 122122 Reported-by: Marvin Schmidt 122123 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122124 122125commit c947605eda082fbca0821efc0cc14fec8f958fd0 122126Author: Peter Hutterer <peter.hutterer@who-t.net> 122127Date: Tue Sep 15 15:05:27 2009 +1000 122128 122129 configure.ac: move all protocol requirements to a single location. 122130 122131 Protocol requirements were spread throughout configure.ac, leaving us with a 122132 few locations where one DDX asked for a different version of the package 122133 than another. 122134 For each protocol, define <NAME>PROTO="[<name>proto >= version]" and 122135 assemble REQUIRED_MODULES based on these. 122136 122137 Future changes to protocol versions need to only be done in a single place. 122138 122139 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122140 122141commit 4d6b20c25af5f590f19530b6c007e3648a8037c3 122142Author: Shunichi Fuji <palglowr@gmail.com> 122143Date: Wed Sep 16 12:02:00 2009 +1000 122144 122145 dri: use noPanoramiXExtension directly. 122146 122147 xf86LoaderCheckSymbol() is never useful if using externed variable directly. 122148 noPanoramiXExtension will be just used through dlopen() like other extension modules. 122149 122150 Signed-off-by: Shunichi Fuji <palglowr@gmail.com> 122151 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122152 122153commit 3853314d13b503684c9922715323f8033febe5de 122154Author: Alan Coopersmith <alan.coopersmith@sun.com> 122155Date: Thu Sep 17 08:11:27 2009 -0700 122156 122157 Fix typo in Xephyr man page 122158 122159 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122160 122161commit 2edf967b00963559552d061bb98b15243d978279 122162Author: Dave Airlie <airlied@redhat.com> 122163Date: Thu Sep 17 15:24:01 2009 +1000 122164 122165 xserver: SIGIO option handling was incorrect. 122166 122167 reported by a user on #radeon when DRI1 broke. 122168 122169 Signed-off-by: Dave Airlie <airlied@redhat.com> 122170 122171commit 715953bf5c00b4605dd6ab45c92140c0e81174ee 122172Author: Alan Coopersmith <alan.coopersmith@sun.com> 122173Date: Tue Sep 15 21:13:00 2009 -0700 122174 122175 Check for -wrap support in the linker 122176 122177 Allows unit tests to be built with non-gnu-linkers that 122178 also have -wrap support 122179 122180 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122181 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122182 122183commit e2c64551808988657069006f74c6780973ec0557 122184Author: Alan Coopersmith <alan.coopersmith@sun.com> 122185Date: Tue Sep 15 21:13:48 2009 -0700 122186 122187 Add configuration option for use of SIGIO handlers for input events 122188 122189 Boolean option to enable/disable SIGIO handlers is set by the first 122190 of these found: 122191 - UseSIGIO option is set in xorg.conf ServerFlags 122192 - Default set at build time by ./configure --enable-use-sigio-by-default 122193 - Platform default value: Solaris = no, all others = yes 122194 122195 This matches the current settings on all platforms except Solaris. 122196 This reverts Solaris (for now) to the settings used in Xorg 1.6, before 122197 SIGIO support for Solaris was added, due to some system level bugs that 122198 won't be resolved in time for Xorg 1.7 release, but allows us to enable 122199 when those are resolved (or when we need to test if they're resolved). 122200 See http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6879897 122201 122202 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122203 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122204 122205commit 3bdf36476179257561b9409c3f49a77640fbb9f9 122206Author: Matthias Hopf <mhopf@suse.de> 122207Date: Wed Sep 16 16:03:46 2009 +0200 122208 122209 Work around sporadic segfault on resume with intel/KMS due to cursor->bits == NULL. 122210 122211 Apparently SavedCursor is sometime tried to be set while already being set. 122212 122213commit fe31f9c646bc26163497f29c3b833c090df0b3f9 122214Author: Alan Coopersmith <alan.coopersmith@sun.com> 122215Date: Mon Sep 14 19:51:23 2009 -0700 122216 122217 Change xf86dristr.h includes to use xf86driproto.h instead 122218 122219 Clears warnings about obsolete headers, but raises minimum 122220 required version of xf86driproto to 2.1.0 122221 122222 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122223 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122224 122225commit 159e03204093991ccd4133133c50c0cae88c828f 122226Author: Peter Hutterer <peter.hutterer@who-t.net> 122227Date: Tue Sep 15 11:39:09 2009 +1000 122228 122229 dmx: don't include panoramiXsrv.h if building w/o Xinerama 122230 122231 Reported-by: Marvin Schmidt 122232 Tested-by: Marvin Schmidt 122233 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122234 122235commit f56cbe1ef24415d0142b9a7d0ab0a031069ccb52 122236Author: Rémi Cardona <remi@gentoo.org> 122237Date: Mon Sep 14 17:09:59 2009 +0200 122238 122239 dix: append "built-ins" to the font path in SetDefaultFontPath 122240 122241 49b93df8a3002db7196aa3fc1fd8dca1c12a55d6 made the hard dependency on 122242 a "fixed" font go away but only Xorg could use the built-ins fonts by 122243 default. 122244 122245 With this commit, all DDXs get "built-ins" appended to their FontPath, not 122246 just Xorg. 122247 122248 Tested with Xorg, Xvfb and Xnest. 122249 122250 Signed-off-by: Rémi Cardona <remi@gentoo.org> 122251 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 122252 Tested-by: Jon TURNEY <jon.turney@dronecode.org.uk> 122253 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122254 122255commit 139368f7ae192b592e24d013e8ca5ce4175effe1 122256Author: Alan Coopersmith <alan.coopersmith@sun.com> 122257Date: Tue Sep 15 17:53:50 2009 -0700 122258 122259 Fix num_masks/length overflow test for XiSelectEvents 122260 122261 Have to set windowid to a valid value first, since that check 122262 appears earlier in the code than the masks/length check. 122263 122264 Also have to have data[] set large enough so that reading mask 122265 data for 0xFFFF masks doesn't overflow past the end of the array 122266 into uninitialized data. 122267 122268 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122269 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122270 122271commit 13decf5efe76e735a003227c951e58a41e36a8b5 122272Author: Alan Coopersmith <alan.coopersmith@sun.com> 122273Date: Tue Sep 15 13:54:51 2009 -0700 122274 122275 test_convert_XIRawEvent: Don't write more data than the struct holds 122276 122277 sizeof(in.valuators.mask) * 8 == 40, but the valuators arrays only 122278 holds 36 entries, so the test was smashing the stack when 36 < i < 40 122279 (leading to core dumps on Solaris x86, since the return address was 122280 overwritten with an invalid pointer). 122281 122282 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122283 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122284 122285commit 38a61e9a7cf616e26a79dba815bf67737317d1eb 122286Author: Alan Coopersmith <alan.coopersmith@sun.com> 122287Date: Tue Sep 15 13:54:50 2009 -0700 122288 122289 Remove stray semi-colon in protocol-common.c 122290 122291 Caused Sun compilers to issue warning: 122292 "protocol-common.c", line 141: warning: syntax error: empty declaration 122293 122294 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122295 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122296 122297commit e320736c457ea7b2f19a4c7c2dfc0c461d0dd405 122298Author: Alan Coopersmith <alan.coopersmith@sun.com> 122299Date: Mon Sep 14 19:51:22 2009 -0700 122300 122301 Fix module path in xorg.conf man page 122302 122303 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122304 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122305 122306commit 49b98d092ec3d032e657eeff62c22fce243b1e8b 122307Author: Tiago Vignatti <tiago.vignatti@nokia.com> 122308Date: Thu Sep 10 17:52:33 2009 +0300 122309 122310 xfree86: vgaarb: rework wrapping function 122311 122312 No semantical changes. 122313 122314 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 122315 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122316 122317commit 422cd7a674e6831f40be14ee853bd946cb62eadc 122318Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 122319Date: Mon Sep 14 13:52:53 2009 -0700 122320 122321 XQuartz: pbproxy: Remove debugging XBell() 122322 (cherry picked from commit 1fd7c1fd4740393a32e261af33e64c19be48ca9a) 122323 122324commit 663807ce0e65b51537540bb95fbf3c18988c04c7 122325Author: Peter Hutterer <peter.hutterer@who-t.net> 122326Date: Mon Sep 14 16:44:52 2009 +1000 122327 122328 xserver 1.6.99.901 (xserver 1.7 RC1) 122329 122330 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122331 122332commit f450be999268d118b384c04ea070c87ad48c4c59 122333Author: Soeren Sandmann <sandmann@daimi.au.dk> 122334Date: Fri Sep 11 04:03:06 2009 +0200 122335 122336 Add xoff and yoff to drawable->x and drawable->y in create_bits_picture(). 122337 122338 These are 0, except when the drawable is a redirected window. 122339 122340 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122341 122342commit 2b00afec4b5b51851aee4cb0fba7baff3e10ef3f 122343Author: Alan Coopersmith <alan.coopersmith@sun.com> 122344Date: Mon Sep 14 15:59:41 2009 +1000 122345 122346 Update xorg.conf man page & sample for changes in Xorg 1.7 122347 122348 AllowMouseOpenFail description changed to reflect actual behaviour 122349 and point to AllowEmptyInput for previously described behaviour. 122350 122351 Update default DPMS mode timeouts to match new defaults set 122352 in April 2009 by commit d52fddefaeb43f15b677eefbea4a288a9948373b 122353 122354 Update autoloaded module list to match ModuleDefaults in xf86Config.c 122355 Update module subdir list to match stdSubdirs in loadmod.c 122356 122357 Add xorg.conf options that were added to the code: 122358 - XkbDir option added in February 2009 122359 by commit 76f18b94bd2719a8199334742d021c4d0806187d 122360 - DRI2 option added in April 2008 122361 by 35982bc109d424c464551ab22ec90af69908c884 122362 122363 Remove xorg.conf options that were removed from the code: 122364 - XkbDisable option was removed in January 2009 122365 by commit 40877c6680863bd6a8475f2bb0c54df55bcf0b0e 122366 - PciProbe/Config options were removed in August 2008 122367 by commit fdf7c747a8e1bc59dfb31b7f90dd5eab2687315b 122368 - EstimateSizesAggressively was removed in August 2008 122369 by commit cd1e8f26147919227e7624ac4c6b313d972a4d35 122370 - loadable font modules were removed in July 2008 122371 by commit affec10635343668839994ea2bac16c1d4524200 122372 - ModInDev options were removed in December 2008 122373 by commit 6de6ffff35ac03d49fa61de195d4a0605e0ef8bf 122374 122375 (Also strips some trailing whitespaces to make git happier.) 122376 122377 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122378 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122379 122380commit b6f5c54fc61588c638b987bde914b0e5ff9b655e 122381Author: Alan Coopersmith <alan.coopersmith@sun.com> 122382Date: Mon Sep 14 15:49:01 2009 +1000 122383 122384 xfree86: Change default font path to match configure.ac's --default-font-path. 122385 122386 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122387 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122388 122389commit 23dc4678e9333ba0ee1a5374521905fa3f4f309c 122390Author: Peter Hutterer <peter.hutterer@who-t.net> 122391Date: Mon Sep 14 14:14:27 2009 +1000 122392 122393 dmx: use top_builddir, not top_srcdir to get libxfixes.la 122394 122395 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122396 Tested-by: Byeong-ryeol Kim <brofkims@gmail.com> 122397 122398commit 4650e6ebe6e3ea48a026ace60b25daa165580467 122399Author: Peter Hutterer <peter.hutterer@who-t.net> 122400Date: Mon Sep 14 12:00:18 2009 +1000 122401 122402 xkb: drop key presses for already repeating keys. (#23889) 122403 122404 The event sequence for continuously pressed keys with the keyboard driver is 122405 PRESS - PRESS - PRESS - ... - RELEASE. 122406 The first press sets the repeatKey to the keycode and the matching timer. 122407 The second press (on the same keycode) can be silently dropped instead of 122408 overwriting the timer again. 122409 122410 X.Org Bug 23889 <http://bugs.freedesktop.org/show_bug.cgi?id=23889> 122411 122412 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122413 Acked-by: Daniel Stone <daniel@fooishbar.org> 122414 122415commit c626a5578aa16c17ef68c1b15ed53828080d9d90 122416Author: Peter Hutterer <peter.hutterer@who-t.net> 122417Date: Sun Sep 13 17:35:51 2009 +1000 122418 122419 xfree86: fix VT_WAITACTIVE control flow (#11477) 122420 122421 Move misplaced } to get the flow of 122422 122423 if (!ShareVTs) { 122424 VT_ACTIVATE 122425 VT_WAITACTIVE 122426 } 122427 122428 X.Org Bug 11477 <http://bugs.freedesktop.org/show_bug.cgi?id=11477> 122429 122430 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122431 Acked-by: Daniel Stone <daniel@fooishbar.org> 122432 122433commit d3c66d6b69c04466b8388586090dc48a0aad7684 122434Author: Peter Hutterer <peter.hutterer@who-t.net> 122435Date: Mon Sep 14 08:03:17 2009 +1000 122436 122437 dix: FindChildForEvent needs init child to None. 122438 122439 silences compiler warning: 122440 events.c: In function 'FixUpEventFromWindow': 122441 events.c:2262: warning: 'child' may be used uninitialized in this function 122442 122443 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122444 122445commit e85775aa64d1e9322c82090a26256af5e1fede0a 122446Author: Kevin E Martin <kem@redhat.com> 122447Date: Mon Sep 14 08:01:56 2009 +1000 122448 122449 Include <X11/extensions/dmx.h> and remove _DMX_SERVER_ define. 122450 122451 Fixes build with newer dmx header files that have been split between dmx.h 122452 and dmxext.h. _DMX_SERVER_ is no longer needed with the split headers. 122453 122454 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122455 122456commit 232d43510e0f9f780a9f8eb857366e03acf2fcdc 122457Author: Kevin E Martin <kem@redhat.com> 122458Date: Mon Sep 14 08:00:42 2009 +1000 122459 122460 dmx: Make description match default option. 122461 122462 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122463 122464commit 6389a190faa7c06bdedfd1cc3acb280ebb5cc04a 122465Author: Tiago Vignatti <tiago.vignatti@nokia.com> 122466Date: Sat Sep 12 21:47:51 2009 +0300 122467 122468 shave: use CC tag variable on libtool 122469 122470 This fix scratchbox compilation. 122471 122472 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 122473 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122474 122475commit dfb106c846c880a06fa5caada92f45576eab7b00 122476Author: Peter Hutterer <peter.hutterer@who-t.net> 122477Date: Tue Sep 8 18:09:12 2009 +1000 122478 122479 configure: re-enable dmx build (default=auto) 122480 122481 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122482 122483commit 72607960c136894c4533a54cef7a9be64246f1a3 122484Author: Peter Hutterer <peter.hutterer@who-t.net> 122485Date: Wed Sep 9 11:02:22 2009 +1000 122486 122487 dmx: Remove some dead code. 122488 122489 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122490 122491commit 88187ad889ab707fcae3f64890cdb291b567ec23 122492Author: Peter Hutterer <peter.hutterer@who-t.net> 122493Date: Wed Sep 9 11:02:18 2009 +1000 122494 122495 dmx: purge DMX EQ leftovers 122496 122497 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122498 122499commit f3aaa46f27f802233e54b3c70ba827e0fcca2605 122500Author: Peter Hutterer <peter.hutterer@who-t.net> 122501Date: Wed Sep 9 09:51:45 2009 +1000 122502 122503 dmx: formatting fix to improve readability 122504 122505 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122506 122507commit 7a588c833e87dd07f167b9885d5ec43dde271ebe 122508Author: Peter Hutterer <peter.hutterer@who-t.net> 122509Date: Tue Sep 8 18:14:43 2009 +1000 122510 122511 dmx: Default rules are now base, not xfree86 122512 122513 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122514 122515commit 33243ae971db90477a29dfb0e51cc00c95e561bc 122516Author: Peter Hutterer <peter.hutterer@who-t.net> 122517Date: Tue Aug 4 17:04:08 2009 +1000 122518 122519 dmx: include xfixes lib, reshuffle include order to resolve linker errors. 122520 122521 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122522 122523commit 2d46678762b12c9746fc8f89821c15705680ba35 122524Author: Peter Hutterer <peter.hutterer@who-t.net> 122525Date: Tue Sep 8 17:46:06 2009 +1000 122526 122527 dmx: remove dmx-internal event queue. 122528 122529 The EQ is in the DIX now. 122530 122531 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122532 122533commit 545f11139d600febf1c5fd46b11bea4352b82c32 122534Author: Peter Hutterer <peter.hutterer@who-t.net> 122535Date: Tue Sep 8 17:14:53 2009 +1000 122536 122537 dmx: compiler warning fix (mixed declarations + code) 122538 122539 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122540 122541commit 7d703af74ced36968f4480507a7eb54d249b5edc 122542Author: Peter Hutterer <peter.hutterer@who-t.net> 122543Date: Tue Sep 8 17:13:38 2009 +1000 122544 122545 dmx: get the state from XKB for dmxCheckSpecialKeys 122546 122547 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122548 122549commit 6e52f9233176986508356f6eee4fce86ba740e71 122550Author: Peter Hutterer <peter.hutterer@who-t.net> 122551Date: Tue Sep 8 17:10:22 2009 +1000 122552 122553 dmx: fix dmxKeySymToKeyCode to work with mandatory XKB. 122554 122555 Approach taken is inefficient, it converts the xkb symbol table to a core 122556 symbol table first and then extracts the keycode from there. 122557 Consider this a todo for a rainy afternoon when the beer fridge demands 122558 emptying. 122559 122560 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122561 122562commit 61ae0a5ca5de24590f444c405445124404ba9753 122563Author: Peter Hutterer <peter.hutterer@who-t.net> 122564Date: Tue Sep 8 17:05:12 2009 +1000 122565 122566 dmx: fix up dmxKeyCodeToKeySym for XKB-only. 122567 122568 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122569 122570commit 9caba3e2abceda0a1ff4ad4bf05708f0166c2218 122571Author: Peter Hutterer <peter.hutterer@who-t.net> 122572Date: Tue Aug 4 15:12:25 2009 +1000 122573 122574 dmx: fix two calls to InitKeyboardDeviceStruct. 122575 122576 Takes RMLVO now instead of the keysyms directly. 122577 122578 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122579 122580commit 29b9a10dad1ca780dc90f4f1f63fd11fd7d1221b 122581Author: Peter Hutterer <peter.hutterer@who-t.net> 122582Date: Tue Aug 4 15:04:45 2009 +1000 122583 122584 dmx: XkbComponentNamesRec doesn't have a keymap field anymore. 122585 122586 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122587 122588commit a6f3f6a63ae76afbe01eb9049a63609476d2ed4d 122589Author: Peter Hutterer <peter.hutterer@who-t.net> 122590Date: Tue Sep 8 16:43:04 2009 +1000 122591 122592 dmx: switch lnx-input and usb-input over to xkb-only. 122593 122594 kbd*Convert are nearly identical, it just asks for merging. Not today 122595 though. 122596 122597 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122598 122599commit 5d81aee5f75e80ae9d3ac7cb32d293ade62d6c5c 122600Author: Peter Hutterer <peter.hutterer@who-t.net> 122601Date: Tue Aug 4 15:24:02 2009 +1000 122602 122603 dmx: include xkbsrv.h to silence compiler. 122604 122605 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122606 122607commit 4da59f478686fa7e80a3837bf9fa61672c13c50b 122608Author: Peter Hutterer <peter.hutterer@who-t.net> 122609Date: Tue Sep 8 16:30:36 2009 +1000 122610 122611 xkb: split effectiveGroup calculation into separate utility function. 122612 122613 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122614 122615commit 8fb3fa28a5a1b36cdaad38055a607400828b9e1c 122616Author: Alan Coopersmith <alan.coopersmith@sun.com> 122617Date: Fri Sep 11 11:21:26 2009 -0700 122618 122619 compiler.h: include <string.h> before using memmove() 122620 122621 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 122622 122623commit d908ee6e549fd8ff653e2c58e9637df347540ef9 122624Author: Tiago Vignatti <tiago.vignatti@nokia.com> 122625Date: Thu Sep 10 20:56:18 2009 +0300 122626 122627 Revert "Make sys.c use compiler.h unaligned access functions" 122628 122629 This reverts commit 1b30545c04a51bfa3ff95a26d64962907a62ff15. 122630 122631 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 122632 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122633 122634commit 281ebb453d2f5254e057a24e580dafa9abf82305 122635Author: Tiago Vignatti <tiago.vignatti@nokia.com> 122636Date: Thu Sep 10 17:52:32 2009 +0300 122637 122638 xfree86: vgaarb: close and cleanup vga arbiter 122639 122640 So far there are no apparently issues on not closing the fd. But let's do the 122641 right job here. 122642 122643 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 122644 Acked-by: Dave Airlie <airlied@redhat.com> 122645 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122646 122647commit e62549b8d61cd739adbbd04487d96fe706da1370 122648Author: Peter Hutterer <peter.hutterer@who-t.net> 122649Date: Wed Sep 9 16:03:38 2009 +1000 122650 122651 Xext: don't try to initialize XTEST device properties if they failed. 122652 122653 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122654 122655commit 744cdc897746dca86f5ab9e57be982621c2f0986 122656Author: Peter Hutterer <peter.hutterer@who-t.net> 122657Date: Wed Sep 9 15:05:15 2009 +1000 122658 122659 dix: Remove two _X_EXPORT defines from the function definition. 122660 122661 These two are defined _X_EXPORT in their declaration anyway. 122662 122663 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122664 122665commit 9871f9be4e5b184b151ce93458b2578cc8c7b962 122666Author: Martin-Éric Racine <q-funk@iki.fi> 122667Date: Wed Sep 9 13:43:45 2009 +0300 122668 122669 Use lowercase PCI vendor ID for NSC. 122670 122671commit f4350c66b493d63fa06be87caa958d7033232ea4 122672Author: Daniel Stone <daniel@fooishbar.org> 122673Date: Wed Sep 9 15:48:33 2009 +1000 122674 122675 fbdevhw: Test for graphics:fb%d as well as graphics/fb%d 122676 122677 Apparently the kernel can't decide on an API to expose to userspace, so 122678 let's just try both in the hope that one will work. 122679 122680 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 122681 Acked-by: Michel Dänzer <michel@daenzer.net> 122682 122683commit 72f6be540840b46aa4b158b98fac11a6f90dc9d1 122684Author: Peter Hutterer <peter.hutterer@who-t.net> 122685Date: Fri Sep 4 11:13:55 2009 +1000 122686 122687 xfree86: remove now unused xf86CrtcShadowClear 122688 122689 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122690 122691commit 1f24638985ce053b696ca3359cc50b5bc26ccb03 122692Author: Peter Hutterer <peter.hutterer@who-t.net> 122693Date: Wed Sep 9 09:28:46 2009 +1000 122694 122695 kdrive: output meaningful error message for HAL devices. 122696 122697 kdrive ignores all devices from hal as they don't have the 'type' option 122698 set. Instead of "Unrecognised device identifier!" print out "Ignoring 122699 device from HAL." to indicate that the errors surrounding the device don't 122700 really matter. 122701 122702 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122703 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 122704 122705commit ae812bb79668e59cda1f851fb1fabe9db9ea705f 122706Author: Peter Hutterer <peter.hutterer@who-t.net> 122707Date: Tue Sep 8 13:53:26 2009 +1000 122708 122709 Xnest: set the default pointer accel to the host's values (#10013) 122710 122711 The defaultPointerControl holds compile-time defaults for pointer 122712 acceleration. If an Xnest instance is started, it resets the hosts pointer 122713 accel values to these built-in defaults instead of using the host values. 122714 122715 This patch queries the host for the values before initializing the device, 122716 thus leaving the host values untouched. 122717 122718 X.Org Bug 10013 <http://bugs.freedesktop.org/show_bug.cgi?id=10013> 122719 122720 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122721 122722commit 97e3f4316b1c917f93d324f05d3a4551c1d051a3 122723Author: Peter Hutterer <peter.hutterer@who-t.net> 122724Date: Tue Sep 8 09:55:22 2009 +1000 122725 122726 dix: auto-float SD's with SendCoreEvents "false" 122727 122728 AlwaysCore and SendCoreEvents specify whether a device is to send core 122729 events. A device that has either disabled is not supposed to send core 122730 events. 122731 122732 With MPX/XI2, a device that is attached automatically sends core events when 122733 the event is routed through the master device. Floating a slave device 122734 disables core events by breaking the route. 122735 122736 This patch automatically floats devices that have coreEvents disabled in the 122737 xorg.conf/HAL. This replicates the behaviour of a SendCoreEvents "false" 122738 device in server 1.6 and earlier. 122739 122740 The devices may still be reattached to a master at runtime. 122741 122742 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122743 122744commit 447bd4ce90113037a489ce55d9878523fe5294c7 122745Author: Peter Hutterer <peter.hutterer@who-t.net> 122746Date: Mon Sep 7 16:02:50 2009 +1000 122747 122748 kdrive: fix sdl build by removing old keySymsRec references. 122749 122750 These are set in kinput now through InitKeyboardDeviceStruct. 122751 122752 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122753 122754commit d8aadfa5af85ce6613289efe119e592aa687ab03 122755Author: Peter Hutterer <peter.hutterer@who-t.net> 122756Date: Mon Sep 7 15:04:33 2009 +1000 122757 122758 dix: remove unused and half-broken code to restore original classes. 122759 122760 In theory, the MD should change back to its old, original classes when the 122761 last SD is detached. Thanks to the XTEST devices, we'll always have an SD 122762 attached until the MD is removed. So let's not worry about that and do 122763 nothing instead of having some code that's essentially untested. 122764 122765 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122766 122767commit e9a507acca01234de189f0f64da63bac32dc13fd 122768Author: Peter Hutterer <peter.hutterer@who-t.net> 122769Date: Mon Sep 7 15:59:37 2009 +1000 122770 122771 kdrive: remove xkb member from KdKeyboardInfo 122772 122773 Wasn't being used by anyone. 122774 122775 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122776 122777commit cafa7c5f9017efdc33c348b6a55d80d44df0fe17 122778Author: Martin-Éric Racine <q-funk@iki.fi> 122779Date: Tue Sep 8 23:52:52 2009 +0300 122780 122781 Added comment with the PCI device ID of Geode SCx200 series. 122782 122783commit d3ad1804a5216487b5837a80f3e53b81212dcf84 122784Author: Michel Dänzer <daenzer@vmware.com> 122785Date: Tue Sep 8 15:25:19 2009 +0200 122786 122787 EXA: Only take special code path for 1x1 fill for pixmaps. 122788 122789 It doesn't make sense for windows. 122790 122791 Also double-check that the data pointer is valid. 122792 122793 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=23461 (not sure that could 122794 happen on master even without this, but just in case) 122795 122796commit a9d30f6a038166052a4684308db205bd4e60799e 122797Author: Peter Hutterer <peter.hutterer@who-t.net> 122798Date: Tue Sep 8 17:43:24 2009 +1000 122799 122800 dix: GetKeyboardValuatorEvents doesn't recurse anymore - fix comment. 122801 122802 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122803 122804commit 1bdc9ec617d357b076c9e69296018bc212d91c7d 122805Author: Eamon Walsh <efw@eamonwalsh.com> 122806Date: Fri Sep 4 21:49:29 2009 -0400 122807 122808 Fix "warning: left-hand operand of comma expression has no effect" 122809 122810 Signed-off-by: Eamon Walsh <efw@eamonwalsh.com> 122811 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122812 122813commit 9005ec4ce19d704d5055232d2a36acee3247be67 122814Author: Peter Hutterer <peter.hutterer@who-t.net> 122815Date: Fri Sep 4 11:38:03 2009 +1000 122816 122817 dix: remove dixLookupResource - we don't have any users left. 122818 122819 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122820 Acked-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 122821 Acked-by: Aaron Plattner <aplattner@nvidia.com> 122822 122823commit 1e0caf9a2ca983f171976822d9c04bacc14e9af7 122824Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 122825Date: Sun Sep 6 14:02:53 2009 -0700 122826 122827 XQuartz: launchd: Fallback on :0 if prefix:0 gives an error for the socket name 122828 (cherry picked from commit 647c871dc9f2d0adc172b401cde89ffbdfcc4d7a) 122829 122830commit 081f72390a25e2244561e238f772f98289a6a581 122831Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 122832Date: Sat Sep 5 14:04:53 2009 -0700 122833 122834 XQuartz: Fix "warning: function declaration isn’t a prototype" noise 122835 (cherry picked from commit 9a77905a975e562daa4230739937bbb0b4caf087) 122836 122837commit 7ece2cb220e4fe2c9a8c4c62361522a45adf22c1 122838Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 122839Date: Sat Sep 5 14:03:19 2009 -0700 122840 122841 XQuartz: pbproxy: 64bit fixes: Properly process an array of Atoms 122842 (cherry picked from commit 1b659cda1af02762a31cc7875e457b08c8dc68b5) 122843 122844commit 1747120043cc5b5d201b7efd06b75ef08b032922 122845Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 122846Date: Sat Sep 5 02:51:09 2009 -0700 122847 122848 XQuartz: Fix a strcpy/strcmp typo 122849 (cherry picked from commit 701c24da3ef76a05ccb2d7e84ccfa1b7c5d38e15) 122850 122851commit be37f339a9ed90b9b5e44aeb396ca73347078027 122852Author: Michel Dänzer <daenzer@vmware.com> 122853Date: Fri Sep 4 23:11:13 2009 +0200 122854 122855 EXA: Stop hiding solid or gradient pictures from the driver. 122856 122857 Add support for solid pictures in exaTryDriverSolidFill(), but otherwise just 122858 pass solid and gradient pictures to the driver Composite hook. 122859 122860 While we're at it, clean up the logic to detect Composite operations which are 122861 effectively solid fills or copies. This should also fix some false negatives 122862 and positives. 122863 122864commit 1f61bbeffdf2ad5a2c54121c75f3b049baf01708 122865Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 122866Date: Fri Sep 4 00:38:58 2009 -0700 122867 122868 XQuartz: Check the DISPLAY environment variable to see if the socket at startup is ours. If not, ignore it and fork/exec startx. 122869 (cherry picked from commit 3d5e10cce360a04bf917227615fb9b825675124f) 122870 122871commit 0a361d0e5f896351b98dc9765b4a33c79efa6f02 122872Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 122873Date: Thu Sep 3 19:40:11 2009 -0700 122874 122875 XQuartz: Use --with-launchd-id-prefix for consistency with xinit 122876 122877 Also actually honor LAUNCHD_ID_PREFIX and APPLE_APPLICATION_NAME 122878 (cherry picked from commit 990038ab006b2f5e03dcef385514ba4e4584bd25) 122879 122880commit e6e83d81ee7366779d600cbb128034e248815339 122881Author: Peter Hutterer <peter.hutterer@who-t.net> 122882Date: Fri Sep 4 14:51:08 2009 +1000 122883 122884 xserver 1.6.99.900 122885 122886 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122887 122888commit bc0841b6e87cdd752b44b17ccb021c831a3e02a5 122889Author: Peter Hutterer <peter.hutterer@who-t.net> 122890Date: Fri Sep 4 15:58:36 2009 +1000 122891 122892 shave: hack around the configure ld check for make distcheck. 122893 122894 configure calls $(CC) -print-prog-name=ld to get the binary for the linker. 122895 The shave script prints " CC \nld" which can't be parsed by configure, 122896 resulting in a fatal error. 122897 122898 This patch special-cases -print-prog-name and passes it though directly 122899 without any additional output. 122900 122901 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122902 122903commit 58d9a4c6356c98c32be2826410df027ece70d44d 122904Author: Peter Hutterer <peter.hutterer@who-t.net> 122905Date: Fri Sep 4 15:06:45 2009 +1000 122906 122907 xfree86: fix make distcheck after removal of Domain.note and RAC.Notes. 122908 122909 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122910 122911commit 8af2793a90e8e644a6a598adce3f752c15b3e37a 122912Author: Peter Hutterer <peter.hutterer@who-t.net> 122913Date: Thu Sep 3 11:30:48 2009 +1000 122914 122915 xkb: xkbGetKbdByName on the lastSlave needs to change the master (#21859) 122916 122917 If the layout is changed on a master's lastSlave, the master needs to change 122918 layout immediately. Otherwise, the master stays on the same layout until the 122919 lastSlave changes - which may not happen if only a single keyboard is 122920 available. 122921 122922 X.Org Bug 21859 <http://bugs.freedesktop.org/show_bug.cgi?id=21859> 122923 122924 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122925 122926commit 219ae5b628f9ffe2be0876d5911cc04e52bef5f8 122927Author: Peter Hutterer <peter.hutterer@who-t.net> 122928Date: Fri Aug 21 16:38:53 2009 +1000 122929 122930 dix: hide the sequenceNumber key repeat hack behind two functions. 122931 122932 Just in case we figure out a better way to do it in the future. 122933 122934 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122935 122936commit bfb219f532f3c78ba905424365ee7c5f7b5f21a2 122937Author: Daniel Stone <daniel@fooishbar.org> 122938Date: Fri Aug 21 16:03:36 2009 +1000 122939 122940 input: allow for detectable autorepeat. 122941 122942 For core and XI1 events, store the key_repeat flag in the sequence number 122943 until TryClientEvents. The sequenceNumber is unset until TryClientEvents. 122944 122945 [Also thrown in, some random indentation changes. Thanks] 122946 122947 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122948 122949commit 3d988e01e41e98fc5160f825a250522ba274d09f 122950Author: Daniel Stone <daniel@fooishbar.org> 122951Date: Fri Aug 21 15:15:41 2009 +1000 122952 122953 dix: improve code flow in TryClientEvents, better debugging messages. 122954 122955 Instead of a massive if (blah && blah), return early where possible. 122956 122957 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122958 122959commit 6fb01c8286c16968fd07e4a5b78da89cd8768d79 122960Author: Peter Hutterer <peter.hutterer@who-t.net> 122961Date: Fri Sep 4 10:57:33 2009 +1000 122962 122963 require xf86vidmodeproto 2.2.99.1 or newer. 122964 122965commit 2122fedb2c058dc27f9df0e856d2193fabe725e2 122966Author: Peter Hutterer <peter.hutterer@who-t.net> 122967Date: Fri Sep 4 11:37:58 2009 +1000 122968 122969 os: silence bigreqsproto compiler warning 122970 122971 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122972 122973commit 59e731ef66095723e74a16b2e5548aed627089ff 122974Author: Peter Hutterer <peter.hutterer@who-t.net> 122975Date: Fri Sep 4 11:06:37 2009 +1000 122976 122977 xfree86: silence some xf86dgaproto compiler warnings 122978 122979 We already require xf86dgaproto > 2.0.99 since 6fffcd582 122980 122981 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122982 122983commit ead57a13531aa14033839112fa66e33e142f67da 122984Author: Peter Hutterer <peter.hutterer@who-t.net> 122985Date: Fri Sep 4 11:03:55 2009 +1000 122986 122987 xfree86: silence 'unused variable' compiler warning 122988 122989 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122990 122991commit 21f04f008e6b1ab66a63de3de668b1f261a6928a 122992Author: Peter Hutterer <peter.hutterer@who-t.net> 122993Date: Fri Sep 4 11:01:09 2009 +1000 122994 122995 os: don't redefine GNU_SOURCE 122996 122997 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 122998 122999commit 09f0ee1324b046f2830bdd260f6e2a67ee8f2973 123000Author: Martin-Éric Racine <q-funk@iki.fi> 123001Date: Fri Sep 4 04:17:09 2009 +0300 123002 123003 Formatted recent Geode PCI ID changes to fit a 80-column coding style. 123004 123005commit a17b50b130fe01c058a74f2696a8b97c756acb49 123006Author: Martin-Éric Racine <q-funk@iki.fi> 123007Date: Fri Sep 4 03:51:01 2009 +0300 123008 123009 Further explained what the Geode PCI ID is all about as comments. 123010 123011commit afd800dae4f7491baee08509094a30160bb9c849 123012Author: Martin-Éric Racine <q-funk@iki.fi> 123013Date: Fri Sep 4 03:41:08 2009 +0300 123014 123015 Removed unnecessary curly braces for Geode LX PCI ID. 123016 123017commit 842b7a47d583a7729099de3031ae20977dddf774 123018Author: Martin-Éric Racine <q-funk@iki.fi> 123019Date: Fri Sep 4 03:25:32 2009 +0300 123020 123021 Updated the PCI Vendor and Device IDs for all Geode variants. 123022 123023 In practice, some of the native drivers for older Geode products 123024 have become deprecated due to lack of e.g. libpciaccess upgrade, 123025 but that's OK, since most distributions don't ship them anymore. 123026 In that case, we'll let X server fall back to good old VESA. 123027 123028commit 8e873185f4cbfb2a36e1f43fe7da47fd9fd5aeea 123029Author: Michel Dänzer <daenzer@vmware.com> 123030Date: Thu Sep 3 19:11:02 2009 +0200 123031 123032 EXA: Preserve pPixmap->devPrivate.ptr in exaPixmapIsOffscreen_mixed. 123033 123034 Otherwise we may incorrectly clobber it to NULL on repeated PrepareAccess 123035 calls. 123036 123037commit 120286aef59dabdb7c9fa762e08457e5cc8ec3a6 123038Author: Michel Dänzer <daenzer@vmware.com> 123039Date: Thu Sep 3 08:05:59 2009 +0200 123040 123041 glx: Add screen DestroyWindow wrapper to destroy the GLX drawable. 123042 123043 Fixes crashes exitting MacSlow's rgba-glx demo. 123044 123045commit f04fe06ae244b851b38be824b1a80f2f8a030591 123046Author: Peter Hutterer <peter.hutterer@who-t.net> 123047Date: Thu Sep 3 14:17:14 2009 +1000 123048 123049 dix: don't return BadMatch from GetProperty (#23562) 123050 123051 dixLookupWindow may return BadMatch if the window in question isn't actually 123052 a window. In this case, GetProperty needs to return BadWindow - not 123053 BadMatch. 123054 123055 X.Org Bug 23562 <http://bugs.freedesktop.org/show_bug.cgi?id=23562> 123056 123057 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123058 123059commit 84eb4c66a4a09c360cef260fb2f35dfb6d8a93c6 123060Author: Peter Hutterer <peter.hutterer@who-t.net> 123061Date: Thu Sep 3 10:00:42 2009 +1000 123062 123063 test: add protocol testing for XIWarpPointer. 123064 123065 TODO: some way to check src_x/y coordinates would be good. 123066 123067 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123068 123069commit 1b7858e8469aea6d2031039ba41d7191a4b80f28 123070Author: Peter Hutterer <peter.hutterer@who-t.net> 123071Date: Thu Sep 3 10:58:00 2009 +1000 123072 123073 test: expose the default screen to tests, some cleanup work. 123074 123075 Provide common #define for invalid window IDs. 123076 Init the sprite's hotPhys, provide a common #define for the initial sprite 123077 position. 123078 123079 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123080 123081commit 61a6e1f074d9ff75d61446b946aab6c04019c287 123082Author: Peter Hutterer <peter.hutterer@who-t.net> 123083Date: Thu Sep 3 10:00:00 2009 +1000 123084 123085 Xi: return BadDevice for master kbds and attached slaves in XIWarpPointer 123086 123087 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123088 123089commit 8939ad2b2aa0385f072d3e1169eaf99289ed737a 123090Author: Peter Hutterer <peter.hutterer@who-t.net> 123091Date: Thu Sep 3 09:44:11 2009 +1000 123092 123093 Xi: return error values to client from XIWarpPointer. 123094 123095 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123096 123097commit d481720c57f558c47273bf89854fc69254319772 123098Author: Peter Hutterer <peter.hutterer@who-t.net> 123099Date: Wed Sep 2 15:28:28 2009 +1000 123100 123101 Xi: standardise XI2 headers. 123102 123103 Adding missing dix-config.h include, adding one missing header guard. 123104 123105 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123106 123107commit 377a5f655ca88cd836da24ab42361df9f3ff51a2 123108Author: Peter Hutterer <peter.hutterer@who-t.net> 123109Date: Wed Sep 2 11:16:15 2009 +1000 123110 123111 Xi: if XISetEventMask fails, return this to the client. 123112 123113 The only failure point can be a BadAlloc. 123114 123115 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123116 123117commit 993ca5acbdccef9e43cd6580e7cc22cfda1a2578 123118Author: Peter Hutterer <peter.hutterer@who-t.net> 123119Date: Wed Sep 2 10:53:40 2009 +1000 123120 123121 dix: when unsetting a cursor, update the sprite immediately (#23608) 123122 123123 Removing the device cursor while the cursor was within the window did not 123124 update the visible sprite until the next enter/leave event. 123125 123126 X.Org Bug 23608 <http://bugs.freedesktop.org/show_bug.cgi?id=23608> 123127 123128 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123129 123130commit b41db8fe460fc0956922b79261752d1bb104684f 123131Author: Peter Hutterer <peter.hutterer@who-t.net> 123132Date: Wed Sep 2 08:57:40 2009 +1000 123133 123134 Xi: fix broken swap code in XISelectEvents request processing. 123135 123136 The pointer advanced 12 bytes too short. Rather unfortunate if both the code 123137 and the test have the same bug. 123138 123139 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123140 123141commit 727de7c90de4198222e0dc58d7503b4a1672f642 123142Author: Peter Hutterer <peter.hutterer@who-t.net> 123143Date: Tue Sep 1 16:33:56 2009 +1000 123144 123145 Xi: Unify checking for invalid bits in grab masks. 123146 123147 Bits above XI2LASTEVENT are invalid and cause in BadValues. These checks 123148 must be performed anywhere where a mask_len parameter is given. 123149 123150 This patch also adds the missing checks to grab masks. 123151 123152 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123153 123154commit 59a6d7d478903a8bc9c5d4cc8b2e62e2ad102dba 123155Author: Peter Hutterer <peter.hutterer@who-t.net> 123156Date: Tue Sep 1 16:30:52 2009 +1000 123157 123158 Xi: don't overrun memory for grab masks. 123159 123160 A grab mask provided in the request may be larger than the one used in the 123161 server. Cut down to size before memcopying. 123162 123163 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123164 123165commit 83db2b126e8623824e1303e74070375994984599 123166Author: Peter Hutterer <peter.hutterer@who-t.net> 123167Date: Tue Sep 1 15:58:02 2009 +1000 123168 123169 Xi: don't try to set oversized or non-existing masks. 123170 123171 Fixes crash if the first XISelectEvents has a zero sized event mask. 123172 Fixes crash if the mask provided is larger than others->xi2mask[]. 123173 123174 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123175 123176commit 6fe235d3635aaf995dfa198c60ded1a43fe9dd45 123177Author: Peter Hutterer <peter.hutterer@who-t.net> 123178Date: Tue Sep 1 15:46:34 2009 +1000 123179 123180 test: fix request length calculation, add length tests for XISelectEvents 123181 123182 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123183 123184commit 7ab1b50c545ae1a62384bedee60168e227bb2a94 123185Author: Peter Hutterer <peter.hutterer@who-t.net> 123186Date: Tue Sep 1 15:27:29 2009 +1000 123187 123188 test: fix build after changing libxf86config.la 123189 123190 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123191 123192commit 58c298acc1045927e0d90be73b8dbc8837252589 123193Author: Peter Hutterer <peter.hutterer@who-t.net> 123194Date: Tue Sep 1 15:16:17 2009 +1000 123195 123196 Xi: extra length checking for requests providing masks. 123197 123198 masks can be of arbitrary length. If the client did not initialize mask_len, 123199 some sort of boundary check is needed to avoid running over memory. 123200 123201 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123202 123203commit 0e4dd3b2d28d3dbbfc152d6f5030901ec063a7ae 123204Author: Peter Hutterer <peter.hutterer@who-t.net> 123205Date: Mon Aug 31 09:50:36 2009 +1000 123206 123207 Un-export ApplyPointerMapping. This is an in-server function. 123208 123209 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123210 123211commit a470e8426ee290046a927a17fc969235c4b1e2e7 123212Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 123213Date: Wed Sep 2 17:56:52 2009 -0400 123214 123215 Correct outdated e-mail address in "Author" statements. 123216 123217 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 123218 123219commit 17545ed80ee3bb6a058a2748fc7bfb0fc5139fc1 123220Author: Adam Jackson <ajax@redhat.com> 123221Date: Wed Sep 2 14:40:07 2009 -0400 123222 123223 randr: Fix crtcs using set_mode_major() 123224 123225 We'd never mark the crtc as active, meaning (among other things) gamma 123226 upload wouldn't work. 123227 123228commit 291408980f33b1e541c89d958535e6fad55fdac9 123229Author: Dave Airlie <airlied@linux.ie> 123230Date: Wed Sep 2 20:14:30 2009 +1000 123231 123232 vgaarb: protect fini as well just in case 123233 123234commit 8aeadbff0840efc0c9a073e60d6fec0dba313060 123235Author: Dave Airlie <airlied@linux.ie> 123236Date: Wed Sep 2 19:07:40 2009 +1000 123237 123238 vgaarb: if arb init fails, make sure locking doesn't occur. 123239 123240 pointed out by Martin Jansa on irc. 123241 123242 Signed-off-by: Dave Airlie <airlied@redhat.com> 123243 123244commit 0e46066efea4e385b2874d4a62a786957e572fe5 123245Author: Jerome Glisse <jglisse@redhat.com> 123246Date: Tue Sep 1 13:59:47 2009 +0200 123247 123248 Xext: fix a typo for bigreqsproto.h header file 123249 123250 Signed-off-by: Jerome Glisse <jglisse@redhat.com> 123251 123252commit c61d8ce1507318476b75b81bf871da19cb1bf747 123253Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 123254Date: Mon Aug 31 17:57:50 2009 -0700 123255 123256 Xext: Require newer versions of xcmiscproto, bigreqsproto, and xf86bigfontproto for new *proto.h header file names 123257 123258 bigreqsproto >= 1.1.0 123259 xcmiscproto >= 1.2.0 123260 xf86bigfontproto >= 1.2.0 123261 123262commit 1c37be80982151a5acd74fe4b960695f604f22ca 123263Author: Michael Witrant <mike@lepton.fr> 123264Date: Tue Sep 1 10:46:50 2009 +1000 123265 123266 XFree86: Linux: Fix 100% CPU usage with ShareVTs and kbd 123267 123268 Leave consoleFd open over the course of the server, even though any use 123269 of it in this context is likely to be disastrous. 123270 123271 Signed-off-by: Michael Witrant <mike@lepton.fr> 123272 Reviewed-by: Daniel Stone <daniel@fooishbar.org> 123273 123274commit 65183dc3153232943dff3cfbaacab141e7ed0e9c 123275Author: Daniel Stone <daniel@fooishbar.org> 123276Date: Sun Jul 5 19:53:55 2009 +0300 123277 123278 Input: Mark Xi input events as critical 123279 123280 Note that the Xi events are critical and should thus cause a flush to 123281 the client when an input event is pending. 123282 123283 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 123284 123285commit a4e614d301cfa2577de1ee6a15abf023590586d1 123286Author: Daniel Stone <daniel@fooishbar.org> 123287Date: Sat Apr 25 16:46:10 2009 +1000 123288 123289 Xi: Fix harmless ButtonPress/ButtonRelease confusion 123290 123291 A missing break meant that ButtonPress would fall through into 123292 ButtonRelease, but luckily it appears to have been completely harmless. 123293 123294 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 123295 123296commit 5ef53a94ce4e48e11de26290cd677266308640c8 123297Author: Matt Turner <mattst88@gmail.com> 123298Date: Mon Aug 31 20:23:26 2009 -0400 123299 123300 alpha: kill xf86SlowBCopyToBus and xf86SlowBCopyFromBus 123301 123302 xf86SlowBCopyToBus and xf86SlowBCopyFromBus cause segfaults on my 123303 system. 123304 123305 Also remove associated slowbcopy_tobus/slowbcopy_frombus macros. 123306 123307 Signed-off-by: Matt Turner <mattst88@gmail.com> 123308 123309commit a8e634b86d47353ef354717904fa66ebe35e1303 123310Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 123311Date: Mon Aug 31 15:53:58 2009 -0700 123312 123313 XQuartz: GLX: Drawable does not contain resize anymore. 123314 (cherry picked from commit 1e642f22fbf029bb3917091cb5fa2f78b4a92f62) 123315 123316commit 1b30545c04a51bfa3ff95a26d64962907a62ff15 123317Author: Matt Turner <mattst88@gmail.com> 123318Date: Mon Aug 31 16:16:02 2009 -0400 123319 123320 Make sys.c use compiler.h unaligned access functions 123321 123322 Signed-off-by: Matt Turner <mattst88@gmail.com> 123323 123324commit 9522ab9a98e98bc394703038e34b1c12aa39a179 123325Author: Matt Turner <mattst88@gmail.com> 123326Date: Mon Aug 31 16:12:29 2009 -0400 123327 123328 Remove unused Delay.c 123329 123330 Signed-off-by: Matt Turner <mattst88@gmail.com> 123331 123332commit 4fac13ff8848bc38654b5567170fbc0d7b2c96d6 123333Author: Matt Turner <mattst88@gmail.com> 123334Date: Mon Aug 31 16:11:59 2009 -0400 123335 123336 Remove unused BUSmemcpy.c 123337 123338 BUSmemcpy.c provides xf86BusToMem and xf86MemToBus, which are are memcpy 123339 wrappers written to avoid glibc's memcpy on Alpha. glibc'c memcpy on 123340 Alpha has improved much since this was written, so it's no longer 123341 needed. Neither function is used inside the xserver, and no module on 123342 my machine uses either as well. 123343 123344 Signed-off-by: Matt Turner <mattst88@gmail.com> 123345 123346commit b65146fc8418b53b7e43dd34cc20000369475671 123347Author: Matt Turner <mattst88@gmail.com> 123348Date: Mon Aug 31 16:11:10 2009 -0400 123349 123350 Remove unused ia64_flush_cache function 123351 123352 Signed-off-by: Matt Turner <mattst88@gmail.com> 123353 123354commit f201705e29bdc3309dd611b461ec67740cb706ef 123355Author: Matt Turner <mattst88@gmail.com> 123356Date: Mon Aug 31 16:09:33 2009 -0400 123357 123358 Simplify unaligned access code 123359 123360 All architectures should be able to use the same unaligned access code, 123361 regardless of whether they need special unaligned access instructions. 123362 Let's let gcc do the heavy lifting. 123363 123364 In the case that we're not using a gcc-compatible compiler, use memmove. 123365 123366 The xserver already requires pixman, so include pixman.h for its uint*_t 123367 types. 123368 123369 Signed-off-by: Matt Turner <mattst88@gmail.com> 123370 123371commit 3718beb69cc5b4cdb79dd97242fce9231a26a39a 123372Author: Matt Turner <mattst88@gmail.com> 123373Date: Mon Aug 31 16:08:52 2009 -0400 123374 123375 Clean up unaligned access functions on alpha 123376 123377 Checks for __GNUC__ are superfluous since the only other compiler for 123378 the platform is Compaq C, and it doesn't support GCC style inline 123379 assembly. 123380 123381 Signed-off-by: Matt Turner <mattst88@gmail.com> 123382 123383commit e1fdd7671346ce374a2b77fe94b8c1a25c0520da 123384Author: Matt Turner <mattst88@gmail.com> 123385Date: Mon Aug 31 16:08:20 2009 -0400 123386 123387 Move unaligned access functions to common section 123388 123389 Signed-off-by: Matt Turner <mattst88@gmail.com> 123390 123391commit 628b863bcc6a2a14fd3c0e6c2be4394833cc314c 123392Author: Matt Turner <mattst88@gmail.com> 123393Date: Mon Aug 31 15:53:41 2009 -0400 123394 123395 Remove unused arm_flush_cache function 123396 123397 Signed-off-by: Matt Turner <mattst88@gmail.com> 123398 123399commit a544139196ac7f5bc89bc642c6c5c0ea1b0bb60e 123400Author: Matt Turner <mattst88@gmail.com> 123401Date: Mon Aug 31 15:53:19 2009 -0400 123402 123403 Remove unused ppc_flush_icache function 123404 123405 Signed-off-by: Matt Turner <mattst88@gmail.com> 123406 123407commit 52aa0495218dc7821a23b045b2c74b2a66e34616 123408Author: Matt Turner <mattst88@gmail.com> 123409Date: Mon Aug 31 15:52:39 2009 -0400 123410 123411 Replace 8 nops with proper sync instruction on mips 123412 123413 Cc: Ralf Baechle <ralf@linux-mips.org> 123414 Acked-by: David Daney <ddaney@caviumnetworks.com> 123415 Signed-off-by: Matt Turner <mattst88@gmail.com> 123416 123417commit 563fa1c5d762173cd6fb78f9e33cb960ef3153bb 123418Author: Matt Turner <mattst88@gmail.com> 123419Date: Mon Aug 31 15:52:10 2009 -0400 123420 123421 Use sfence and mfence instructions on amd64 123422 123423 Signed-off-by: Matt Turner <mattst88@gmail.com> 123424 123425commit 430c4af0d6e2cad03ca5f49ec6e87f3717d0d9e2 123426Author: Matt Turner <mattst88@gmail.com> 123427Date: Mon Aug 31 15:51:41 2009 -0400 123428 123429 Add x86 barrier macros 123430 123431 Signed-off-by: Matt Turner <mattst88@gmail.com> 123432 123433commit 0d87f77a3ff4b22729220b7edcb98c6f06d0787c 123434Author: Matt Turner <mattst88@gmail.com> 123435Date: Mon Aug 31 15:51:05 2009 -0400 123436 123437 define barrier macros as nops if not otherwise defined 123438 123439 Signed-off-by: Matt Turner <mattst88@gmail.com> 123440 123441commit 1b43ee13b30d9e3f828e94cdf686d103ce1cac07 123442Author: Matt Turner <mattst88@gmail.com> 123443Date: Mon Aug 31 15:53:58 2009 -0400 123444 123445 Move memory barrier macros into common section 123446 123447 alphabetize by architecture also. 123448 123449 Signed-off-by: Matt Turner <mattst88@gmail.com> 123450 123451commit 295e598d9a83ff7162a093810ca9f546d00c0968 123452Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123453Date: Mon Aug 31 20:04:51 2009 +0300 123454 123455 xfree86: remove _more_ RAC junk 123456 123457 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123458 123459commit 38627b2eaf64aa8fe694fc3e1f5e4fee0c9c11c3 123460Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123461Date: Mon Aug 31 19:36:02 2009 +0300 123462 123463 xfree86: shut up vgaarb warnings when server doesn't support it 123464 123465 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123466 123467commit 79182538b43ecd35bf2b47eccfb2dd3bbcdc12eb 123468Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123469Date: Mon Aug 31 19:25:13 2009 +0300 123470 123471 xfree86: removal of some dead code due VGA arbiter's inclusion 123472 123473 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123474 123475commit 6f3cc08ec01609c7efe424ac3e30fe4ff9391007 123476Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 123477Date: Thu Aug 27 16:21:11 2009 -0700 123478 123479 XQuartz: GL: Unset GL_EXT_gpu_program_parameters for Tiger/ppc 123480 123481 See http://trac.macports.org/ticket/20638 123482 (cherry picked from commit 0f3a89d306838b3c75a73cd1e9e2928737222b70) 123483 123484commit fdb29ebeed143fa05630966b847b05399a446ddc 123485Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 123486Date: Thu Aug 27 11:04:00 2009 -0700 123487 123488 XQuartz: GL: Explicitly set GL_EXT symbols to 0 if they are not in OpenGL.framework to prevent X11's glext.h from setting them to 1. 123489 (cherry picked from commit 7fe37137d826d1b698e87a5b35050dd02f0a4d4b) 123490 123491commit 3be80bd98357cf362aa9a004d1292e5167d83a17 123492Author: Peter Hutterer <peter.hutterer@who-t.net> 123493Date: Fri Aug 28 16:20:59 2009 +1000 123494 123495 Xi: silence compiler warning "type may be used uninitialized" 123496 123497 GrabKey and GrabButton are only called from XI/XI2 code. Set type to -1, 123498 just in case. 123499 123500 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123501 123502commit 24716b9254fa7d609792596723a192bb044a7d3f 123503Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 123504Date: Fri Aug 28 22:52:48 2009 -0400 123505 123506 xace: fix up access modes in dixLookupDrawable calls from dri2. 123507 123508 Referencing a screen through a drawable only requires GetAttr access. 123509 Treat dri2 drawables as child windows (Add/Remove access). 123510 Treat getting buffers as intent to read/write the drawable. 123511 123512 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 123513 123514commit 0f3a64c3a5bce6cf3f00fe42cc7d1d3e69822945 123515Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 123516Date: Thu Aug 27 15:11:12 2009 -0400 123517 123518 Add DRI2 requests to protocol.txt 123519 123520 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 123521 123522commit 3d17c4f6e1538986fe7b1f17614d76e28f2d1e11 123523Author: Adam Jackson <ajax@redhat.com> 123524Date: Fri Aug 28 12:12:52 2009 -0400 123525 123526 EDID: Vendor detailed blocks aren't worth X_WARNING about 123527 123528commit 2c1a845521a31f08938b0f8d6e38835ce127c94a 123529Author: Adam Jackson <ajax@redhat.com> 123530Date: Fri Aug 28 12:09:57 2009 -0400 123531 123532 EDID: Print 1152x864 in established timings, not x870 123533 123534 The spec says x870, but we actually use x864 because that's a real DMT 123535 mode and x870 isn't. This might or might not be wrong, but we should at 123536 least tell the truth. 123537 123538commit 2a806d7fa372e28d039761c9b4087cf812e8e46b 123539Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123540Date: Fri Aug 28 16:05:04 2009 +0300 123541 123542 os: remove unused -cursor option 123543 123544 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123545 123546commit 9040dab76182d1a019ca7fef7b29733d2c199e61 123547Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123548Date: Fri Aug 28 15:50:50 2009 +0300 123549 123550 xfixes: minor clean ups on createInvisibleCursor() 123551 123552 - unused return value 123553 - no reason to declare static variable given the function is just called 123554 once 123555 - no reason to declare different type and cast it after. 123556 123557 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123558 123559commit 80ed8096f9b8e974b556a6c9f3b600fb71b994c3 123560Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123561Date: Fri Aug 28 15:33:19 2009 +0300 123562 123563 render: delete unused headers declaration 123564 123565 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123566 123567commit f959b1e5485b93700c7da449a98182f5ce067ca2 123568Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123569Date: Fri Aug 28 15:30:21 2009 +0300 123570 123571 render: AnimCurInit and AnimCursorCreate shouldn't be _X_EXPORT 123572 123573 Pointed by Peter Hutterer on xorg-devel ml. 123574 123575 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123576 123577commit e454f106dc65ecfacc154a1fa0810935022a8fee 123578Author: Dave Airlie <airlied@redhat.com> 123579Date: Fri Aug 28 11:39:42 2009 +1000 123580 123581 xf86 ddx: add vga arbiter support. 123582 123583 This adds support for using the libpciaccess interface for 123584 vga arbitration support on top of a kernel which supports it. 123585 123586 Currently patches are queued for kernel 2.6.32 in jbarnes 123587 pci tree, and shipping in Fedora kernel. 123588 123589 Co-authors: 123590 Tiago Vignatti <tiago.vignatti@nokia.com> 123591 123592 Signed-off-by: Dave Airlie <airlied@redhat.com> 123593 123594commit 6fffcd5825454a7fe58ffbcfb219f007cf38e731 123595Author: Peter Hutterer <peter.hutterer@who-t.net> 123596Date: Thu Aug 27 13:43:15 2009 +1000 123597 123598 xfree86: require xf86dgaproto 2.0.99.1. 123599 123600 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123601 123602commit 2fba2eac0b4c8d07bdf7bea20ef75ff579621728 123603Author: Peter Hutterer <peter.hutterer@who-t.net> 123604Date: Mon Aug 24 11:23:23 2009 +1000 123605 123606 test: add a few tests for xtest device initialization. 123607 123608 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123609 123610commit 903c3db1d1685bd855dceed9e7b92890743663e1 123611Author: Peter Hutterer <peter.hutterer@who-t.net> 123612Date: Mon Aug 24 09:48:00 2009 +1000 123613 123614 Xext: rename Xtst* to XTest* 123615 123616 This patch corrects a misnaming of XTest-related functions. 123617 123618 The extension itself announces itself as XTEST. Xtst is the library name 123619 itself, but all library functions are prefixed by XTest. Same with the 123620 naming in the server. 123621 123622 - Rename all *Xtst* functions to *XTest* for consistency with the library 123623 and in-server API. 123624 - Rename the "Xtst device" property to "XTEST device" for consistency with 123625 the extension naming. 123626 - Rename the device naming to "<master device name> XTEST device". The 123627 default xtest devices become "Virtual core XTEST pointer" and "Virtual 123628 core XTEST keyboard". 123629 123630 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123631 123632commit 8bfd23e144e51401e3756de9260a4811fcc59e91 123633Author: Peter Hutterer <peter.hutterer@who-t.net> 123634Date: Mon Aug 24 09:38:15 2009 +1000 123635 123636 input: move XTest device initialization into Xext/xtest.c 123637 123638 XTest devices are non-optional but nonetheless specific to the XTEST 123639 extension. 123640 123641 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123642 123643commit a95f80fa914678d360f6dfd2b58926193df6ea4a 123644Author: Peter Hutterer <peter.hutterer@who-t.net> 123645Date: Mon Aug 24 09:26:56 2009 +1000 123646 123647 dix: use IsXtstDevice instead of the direct key lookup. 123648 123649 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123650 123651commit 16b7ebd7d876034edfe8f74562bd06e747879d79 123652Author: Peter Hutterer <peter.hutterer@who-t.net> 123653Date: Mon Aug 24 14:34:23 2009 +1000 123654 123655 xfree86: Remove xf86GetMotionEvents from public API. 123656 123657 This function was used as the default motion event queue API until 123658 including XINPUT_ABI 2 (server 1.5). 123659 123660 This API was broken with 1883485 in May 2008 (wrong casting of parameters) 123661 and isn't in use by input drivers past ABI 3. 123662 123663 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123664 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123665 123666commit af76bddd7a47092e2f0f85b2547c1cc74eb5cebc 123667Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 123668Date: Sun Aug 9 16:13:32 2009 -0700 123669 123670 XQuartz: Check NSINTEGER_DEFINED to make sure the NSInteger and NSUInteger types are defined. 123671 (cherry picked from commit 36e02a8649c0a545084cce5fb2c5717277b9273c) 123672 123673commit 07c935b72fb14cb12f1cb5afbf8cc81f97086bac 123674Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 123675Date: Tue Aug 25 20:11:38 2009 -0700 123676 123677 XQuartz: Use applewmproto 1.4 updated headers. 123678 (cherry picked from commit 203df06c0eb2dcd5adfa788a1ba9569650c15f9c) 123679 123680commit 0f613007cd43545eda40825bd74ada3e6e5f36f6 123681Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 123682Date: Wed Aug 5 17:43:29 2009 -0700 123683 123684 CheckWindowOptionalNeed: Ensure w->optional is set to avoid SEGFAULT 123685 123686 If CheckWindowOptionalNeed is called consecutively (and the first removes optional), then 123687 we will SEGFAULT. This can happen in XQuartz because ReparentWindow will call 123688 pScreen->ReparentWindow which can DeleteProperty which will CheckWindowOptionalNeed... then 123689 ReparentWindow will call it again later indiscriminantly. 123690 (cherry picked from commit b608c864ccc59628bd33c033393121b61580460e) 123691 123692commit e7dd1efef408effe52d0bd3d3aa0b5d4ee10ed90 123693Author: Keith Packard <keithp@keithp.com> 123694Date: Tue Aug 25 18:07:00 2009 -0700 123695 123696 Ensure that rotation updates happen frequently 123697 123698 The smart scheduler is designed to minimize scheduler overhead by 123699 increasing the interval between WaitForSomething calls when a single 123700 client is running. However, the software rotation code depends on 123701 its BlockHandler being invoked for screen updates; the long delays 123702 caused by the smart scheduler optimizations means that screen updates 123703 can be delayed a long time as well. 123704 123705 The change is simple -- prevent the smart scheduler from increasing 123706 the scheduling interval while any screen is using software rotation. 123707 123708 Signed-off-by: Keith Packard <keithp@keithp.com> 123709 123710commit 1740cda7a37abc7d0a169ab4555b446adaa62211 123711Author: Keith Packard <keithp@keithp.com> 123712Date: Tue Aug 25 16:58:07 2009 -0700 123713 123714 Perform rotation redisplay before calling driver block handler (which may flush rendering) 123715 123716 The rotation block handler uses regular driver rendering functions to 123717 repaint the screen, if those functions queue commands in the driver, 123718 it's important that the driver block handler be invoked after the 123719 rotated image is drawn. 123720 123721 Signed-off-by: Keith Packard <keithp@keithp.com> 123722 123723commit 4aab05e3b3231f1ec9795a66a075d17a722634a7 123724Author: Keith Packard <keithp@keithp.com> 123725Date: Tue Aug 25 16:54:16 2009 -0700 123726 123727 xf86_reload_cursors: fix cursor position to eliminate jumping after mode set 123728 123729 xf86_reload_cursors restores the cursor to the correct position, but 123730 that must adjust for cursor hot spot and frame before calling down to 123731 the hardware function, otherwise the cursor jumps to the wrong 123732 position until it is repositioned by the user. 123733 123734 Signed-off-by: Keith Packard <keithp@keithp.com> 123735 123736commit 0f9ffc887ca1471e98df746253d9300e03e46a15 123737Author: Peter Hutterer <peter.hutterer@who-t.net> 123738Date: Mon Aug 24 15:10:52 2009 +1000 123739 123740 Xi: fix XIWarpPointer up for FP3232 as input coordinates. 123741 123742 requires inputproto 1.9.99.902 123743 123744 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123745 123746commit 5e96945cf54136afdb80cc17f67611251d59205d 123747Author: Peter Hutterer <peter.hutterer@who-t.net> 123748Date: Fri Aug 21 12:15:54 2009 +1000 123749 123750 Xi: fix swapping for XIWarpPointer and XIChangeCursor requests. 123751 123752 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123753 123754commit 990d204ef8074665763ad129d0b015b7ab3ae9ef 123755Author: Peter Hutterer <peter.hutterer@who-t.net> 123756Date: Mon Aug 24 09:22:16 2009 +1000 123757 123758 Xext: remove un-used extern of DeviceMotionNotify. 123759 123760 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123761 123762commit 80f18a73267bd3090e459de609a7006a21e6c649 123763Author: Peter Hutterer <peter.hutterer@who-t.net> 123764Date: Mon Aug 24 09:21:03 2009 +1000 123765 123766 input: move CorePointer/KeyboardProc declarations into header. 123767 123768 The extern declaration in xichangehierarchy.c was broken anyway. 123769 This fixes a crash on creating a new master device. 123770 123771 Reported-by: Maxim Levitsky 123772 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123773 123774commit b96823713129f641ee99306e3cadba8dfe21afa2 123775Author: Peter Hutterer <peter.hutterer@who-t.net> 123776Date: Mon Aug 24 09:06:29 2009 +1000 123777 123778 mi: include shmint.h if needed, silence compiler warning. 123779 123780 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123781 123782commit 1b3859a49a3861517c9b34f6e8750f7ae0ff2b5d 123783Author: Peter Hutterer <peter.hutterer@who-t.net> 123784Date: Fri Aug 21 12:29:39 2009 +1000 123785 123786 Xi: add swapping hook for XIGetFocus reply. 123787 123788 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123789 123790commit fe430d7d3b0a85fbea9e93171423688f76dc9e1e 123791Author: Peter Hutterer <peter.hutterer@who-t.net> 123792Date: Fri Aug 21 12:09:32 2009 +1000 123793 123794 test: add XIQueryPointer protocol tests. 123795 123796 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123797 123798commit 285c88018ba4b3819500027c9b6a05b8c27ff2ee 123799Author: Peter Hutterer <peter.hutterer@who-t.net> 123800Date: Fri Aug 21 11:38:42 2009 +1000 123801 123802 Xi: correct length field for XIQueryPointer reply. 123803 123804 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123805 123806commit e4ea91a02d6fac9dad844150c5d2fbe7dc9629b1 123807Author: Peter Hutterer <peter.hutterer@who-t.net> 123808Date: Fri Aug 21 12:12:05 2009 +1000 123809 123810 Xi: allow XIQueryPointer requests for master pointers and floating slaves. 123811 123812 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123813 123814commit 8e396f2b806496cdcac233cc731b0322735020c3 123815Author: Peter Hutterer <peter.hutterer@who-t.net> 123816Date: Fri Aug 21 12:11:19 2009 +1000 123817 123818 Xi: fix swapping for XIQueryPointer request/reply handling. 123819 123820 buttons_size is necessary as WriteToClient swaps the buttons_len field, 123821 resulting in the wrong number of bytes being written later. 123822 123823 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123824 123825commit f1c856ef9073d9d02bc84ce1472883c8441a9758 123826Author: Peter Hutterer <peter.hutterer@who-t.net> 123827Date: Fri Aug 21 10:42:15 2009 +1000 123828 123829 test: fake initialization of a sprite trace and window locations 123830 123831 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123832 123833commit 58fd28f6b6bebab7f31c54e696f96d64ebb50f5a 123834Author: Peter Hutterer <peter.hutterer@who-t.net> 123835Date: Fri Aug 21 09:44:37 2009 +1000 123836 123837 Xi: return deviceid as error value in XIQueryPointer. 123838 123839 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123840 123841commit b41dd4328bfb8a0981d299bb60f986a2bc55fd24 123842Author: Peter Hutterer <peter.hutterer@who-t.net> 123843Date: Fri Aug 21 09:43:46 2009 +1000 123844 123845 record: include recordproto.h, require recordproto 1.13.99.1 123846 123847 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123848 123849commit 65354e5a698a5b527db09afc431110afba0e14b2 123850Author: Peter Hutterer <peter.hutterer@who-t.net> 123851Date: Thu Aug 20 10:52:52 2009 +1000 123852 123853 Xi: return BadDevice for changing device cursors on non-master pointers. 123854 123855 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123856 123857commit ef75301854d41c6b11dd2742fa9d17a5de8a4cfa 123858Author: Peter Hutterer <peter.hutterer@who-t.net> 123859Date: Thu Aug 20 10:52:24 2009 +1000 123860 123861 dix: FixUpEventFromWindow mustn't scramble non-device events. 123862 123863 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 123864 123865commit c73cd3b265c301b8a54ffe484d6c696f2abefb46 123866Author: Dan Nicholson <dbn.lists@gmail.com> 123867Date: Sun Aug 16 18:07:49 2009 -0700 123868 123869 xfree86: Link libselinux with Xorg system libraries 123870 123871 Put the reference to -lselinux in XORG_SYS_LIBS so it isn't added as a 123872 make dependency for libxorg.la. Otherwise, make goes looking for a file 123873 -lselinux in the current directory, which it obviously won't find. 123874 123875 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 123876 Tested-by: Justin P. Mattock <justinmattock@gmail.com> 123877 123878commit 9d8fc33cae2fee7777edd20658dbf6a3ffb3d557 123879Author: Adam Jackson <ajax@redhat.com> 123880Date: Thu Aug 20 16:56:04 2009 -0400 123881 123882 composite: Move screen init before extension registration. 123883 123884 Otherwise, you'd still advertise the extension even if no screens 123885 actually supported it, and the first Composite protocol request would 123886 probably crash. 123887 123888commit ae68708ebe24c4cdf8c5265e206f6913d9298f37 123889Author: Adam Jackson <ajax@redhat.com> 123890Date: Thu Aug 20 15:45:57 2009 -0400 123891 123892 xfree86: dump /proc/cmdline in the log on Linux 123893 123894commit 0b131a5cd91cea54240777c66a9cd385029e8cb2 123895Author: Adam Jackson <ajax@redhat.com> 123896Date: Thu Aug 20 15:28:57 2009 -0400 123897 123898 linux: Yet more malloc() avoidance for backtrace() 123899 123900 Turns out, there's an initializer at the top of backtrace() that (on 123901 some arches) calls dlopen(). dlopen(), unsurprisingly, calls malloc(). 123902 So, call backtrace() early in signal handler setup so we can later 123903 safely call it from the signal handler itself. 123904 123905commit 792dee3854cbede64e17fdc4736831edad295706 123906Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123907Date: Wed Aug 19 18:20:54 2009 +0300 123908 123909 xfree86: remove bus state change notification callback 123910 123911 No one is using bus notifications now. We hope that the kernel take care of 123912 this properly. 123913 123914 For other not-so-urgent-notifications (ACPI wakeups, etc) we can just register 123915 a handler on server's scheduler (using xf86AddGeneralHandler). And for 123916 external applications, the "trend" is to use HAL to kick notifications. So 123917 we're already provided of enough notification schemes. 123918 123919 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123920 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 123921 123922commit eb031d4013d36f6aef4aba45840762ae8635cc13 123923Author: Adam Jackson <ajax@redhat.com> 123924Date: Wed Aug 19 15:42:34 2009 -0400 123925 123926 linux: hand-roll a backtrace printer instead of using backtrace_symbols 123927 123928 Why? Because backtrace_symbols calls malloc, which you can't do from a 123929 signal handler. Face? Palm. 123930 123931commit 75b9383d8a4c113ab3c6cfc1d5efcb5d9982a1bf 123932Author: Jesse Adkins <jesse_adkins7@yahoo.com> 123933Date: Tue Aug 18 21:13:40 2009 -0700 123934 123935 xserver doesn't stop all connections to localhost 123936 123937 X.Org Bugzilla #23329: http://bugs.freedesktop.org/show_bug.cgi?id=23329 123938 Patch #28648: http://bugs.freedesktop.org/attachment.cgi?id=28648 123939 123940 I noticed in xserver at os/access.c that xorg tries to stop connections 123941 to localhost by checking against the address 127.0.0.1. However, RFC 123942 3330 defines the localhost network as 127.0.0.0/8. This means that any 123943 IPv4 address that starts with 127 is just another name for localhost. 123944 123945commit 2d34eace13c2016048c627c4e96c3b2399901078 123946Author: Alan Coopersmith <alan.coopersmith@sun.com> 123947Date: Tue Aug 18 20:11:23 2009 -0700 123948 123949 Sun bug 6872917: Xorg not querying /dev/fb when no xorg.conf exists 123950 123951 commit 48ee5558333bd324463b6994735cabb23de262ec (OpenSolaris VT support) 123952 broke the autoconfiguration code in xf86AutoConfig.c that uses the 123953 Solaris-specific VIS_GETIDENTIFIER ioctl on a frame buffer device like 123954 /dev/fb by changing xf86Info.consoleFd from /dev/fb to a /dev/vt/* 123955 device. 123956 123957 This fixes it by reworking the code to split the console device 123958 (/dev/vt/*, the vtXX CLI option) from the frame buffer device 123959 (/dev/fb, -dev option) to allow both VT and autoconfig to work. 123960 123961 It also fixes the console device to use /dev/fb when VT's are not 123962 supported instead of throwing a Fatal Error because it can't open 123963 /dev/vt/0. 123964 123965 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 123966 123967commit d225230148b3cdab3b32fce3abee26b818d4bab0 123968Author: Jesse Barnes <jbarnes@virtuousgeek.org> 123969Date: Tue Aug 18 12:15:32 2009 -0700 123970 123971 xfree86/linux: don't change VT perms unless we're running as root 123972 123973 In non-setuid root installations, we shouldn't try to adjust VT/tty 123974 ownership. It will fail, and shouldn't be necessary anyway (since 123975 startup scripts or PAM should be handling perms for us in that case). 123976 123977 Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org> 123978 123979commit 41884b8f4dfe4d931c64a6046adfefcc2b7646c2 123980Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123981Date: Tue Aug 18 15:47:56 2009 +0300 123982 123983 xfree86: remove unused code (xf86Configure.c) 123984 123985 The xorg.conf generator was not assigning correctly the primary device 123986 ("bootable") as screen zero. So just skip this kind of routines for now. 123987 123988 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123989 123990commit 097376266f256e3ae37a16b343ed1e3441a8f003 123991Author: Tiago Vignatti <tiago.vignatti@nokia.com> 123992Date: Tue Aug 18 13:13:57 2009 +0300 123993 123994 xfree86: remove unaffected code (xf86Configure.c) 123995 123996 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 123997 123998commit dd8960a9f37fbef65cb99e058c3772edfda7b6db 123999Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124000Date: Tue Aug 18 12:53:47 2009 +0300 124001 124002 xfree86: remove stupid macro (xf86Configure.c) 124003 124004 It was confusing more instead helping. 124005 124006 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124007 124008commit 4ec7667bb17839bd4ec11091c75ecfff317fc96c 124009Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124010Date: Tue Aug 18 12:33:35 2009 +0300 124011 124012 xfree86: remove wrong commentary 124013 124014 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124015 124016commit 2e8c2b8ae6a092871838b7cceaaf4a7590bce1d3 124017Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124018Date: Tue Aug 18 12:28:15 2009 +0300 124019 124020 xfree86: header clean up (xf86Configure.c) 124021 124022 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124023 124024commit 2df414a33321ae772af1e7e52c36aec222bb3c8f 124025Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124026Date: Tue Aug 18 12:14:27 2009 +0300 124027 124028 xfree86: reorganize xf86Configure.c 124029 124030 No semantical changes. Only code moved around. 124031 124032 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124033 124034commit 113e66a1ffe2a43c4d1c92a7d78ca43b4db58822 124035Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124036Date: Sun Aug 16 21:02:11 2009 +0300 124037 124038 xfree86: remove unused PCI header 124039 124040 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124041 124042commit 6b5978dcf1f7ac3ecc2f22df06f7000f360e2066 124043Author: Richard Hughes <richard@hughsie.com> 124044Date: Mon Aug 17 09:15:32 2009 +0100 124045 124046 Do not reset lastDeviceEventTime when we do dixSaveScreens 124047 124048 When we turn off DPMS with DPMSModeOff and do dixSaveScreens, don't reset the 124049 event time else session clients using IDLETIME will be reset. 124050 124051 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124052 124053commit 4098ad72d00e65d142fb9fe8a1194be35338508b 124054Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 124055Date: Sun Aug 16 00:06:10 2009 -0400 124056 124057 xace: fix access mode in dixLookupWindow within several RandR calls. 124058 124059 Referencing a screen using a window only requires GetAttr access. 124060 124061 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 124062 124063commit d4b8f7602b5e266a0ebd3b1ba23724362cc7de3a 124064Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 124065Date: Mon Aug 10 18:39:04 2009 -0400 124066 124067 xace: fix access mode in dixLookupWindow call within ProcUngrabKey. 124068 124069 Referencing a window (as grab-window) only requires GetAttr access. 124070 124071 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 124072 124073commit 6d2f4e487869f10de4a62365b4d6de036c752ab8 124074Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 124075Date: Mon Aug 10 18:36:53 2009 -0400 124076 124077 Add XI2 requests to protocol.txt 124078 124079 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 124080 124081commit 73975ef3a39ce522c6206ca800ed175fbf851dcf 124082Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 124083Date: Mon Aug 10 18:34:07 2009 -0400 124084 124085 xselinux: Allow per-client device create contexts. 124086 124087 The previous behavior was to set the serverClient's value which was used globally. 124088 This is in support of XI2, where clients can create device pairs directly. 124089 124090 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 124091 124092commit 66e32d252cffcd4fe7d505f1c211253f23c5002c 124093Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 124094Date: Mon Aug 10 18:33:05 2009 -0400 124095 124096 xselinux: Add more new device permissions for XI2. 124097 124098 Reflects the ability of clients to create/destroy device objects. 124099 124100 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 124101 124102commit 1f4ea22a20fdd30412a259a87eee133155e1163d 124103Author: Rémi Cardona <remi@gentoo.org> 124104Date: Mon Aug 17 11:33:24 2009 +0200 124105 124106 configure: fix help string after Xnest default build was changed 124107 124108 Signed-off-by: Rémi Cardona <remi@gentoo.org> 124109 124110commit 7e37e7c50e4fdcb53296a99d90af3d90081ce9bb 124111Author: Peter Hutterer <peter.hutterer@who-t.net> 124112Date: Tue Aug 11 14:39:50 2009 +1000 124113 124114 test: add protocol tests for DeviceChangedEvents 124115 124116 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124117 124118commit 2851f04cb2f6e5c30267f733d867c86d4e69a485 124119Author: Peter Hutterer <peter.hutterer@who-t.net> 124120Date: Tue Aug 11 16:27:40 2009 +1000 124121 124122 dix: rework DeviceChangedEvents a bit. 124123 124124 DCEs are now processed when sent throught the master device, not when sent 124125 through the slave device. This includes a removal of some un-used (or partly 124126 used) fields in the DCE itself to something more self-explanatory. 124127 124128 TODO: if a device has events queued and its attachment is changed, the DCE 124129 is silently dropped now. Instead, it should be generated as soon as the 124130 first event after the attachment is sent. 124131 124132 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124133 124134commit c1d901d723c3bee523736eacc15b44a7dff484fe 124135Author: Richard Hughes <richard@hughsie.com> 124136Date: Fri Aug 14 11:44:35 2009 +0100 124137 124138 Don't reset the lastDeviceEventTime when doing DPMS actions 124139 124140 When we change the DPMS mode, don't play games with the last event time as 124141 this breaks applications using IDLETIME to turn the backlight off after a 124142 preset time. 124143 124144 This patch fixes gnome-power-manager and xfce-power-manager 124145 124146 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124147 124148commit dca4de72d375c4b9bcdd25b151e291a77a5b06cc 124149Author: Peter Hutterer <peter.hutterer@who-t.net> 124150Date: Sat Aug 15 20:17:20 2009 +1000 124151 124152 dix: fix potential use of unused variable 'mask'. 124153 124154 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124155 124156commit 978b65bce14598f2d42ca0177ea58fef71fc12c5 124157Author: Peter Hutterer <peter.hutterer@who-t.net> 124158Date: Fri Aug 14 15:05:42 2009 +1000 124159 124160 include: XInputExtensionInit doesn't need to be exported. 124161 124162 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124163 124164commit 3989dc1d34f116f30915632cc5286937392e180a 124165Author: Peter Hutterer <peter.hutterer@who-t.net> 124166Date: Tue Aug 11 16:20:20 2009 +1000 124167 124168 dix: use the XI2 defines for class types. 124169 124170 Doesn't matter really, they have the same values anyway. 124171 124172 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124173 124174commit fe045820f1fb33991e8bff5c6e192097caa85727 124175Author: Peter Hutterer <peter.hutterer@who-t.net> 124176Date: Mon Aug 10 15:34:00 2009 +1000 124177 124178 input: move DeviceChangedEvent conversion into eventconvert.c 124179 124180 The version in eventconvert.c was half broken and for some reason we ended 124181 up with a second version in exevents.c (which works). Move it over to where 124182 it belongs and call EventToXI2 instad of having a custom function for it. 124183 124184 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124185 124186commit 33eb6f70816921abc5da62e434f40e78d672274e 124187Author: Peter Hutterer <peter.hutterer@who-t.net> 124188Date: Fri Aug 14 13:49:58 2009 +1000 124189 124190 include: Unexport most symbols from exevents.h. 124191 124192 And shuffle them around so that the part used by drivers is up the top and 124193 commented. 124194 124195 Also, woo, the sdksyms script doesn't like declarations with return type and 124196 function name on the same line... 124197 124198 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124199 124200commit 1373c2ea877b562076a4a3c97e8e4f999c977217 124201Author: Dave Airlie <airlied@redhat.com> 124202Date: Sat Aug 15 12:14:26 2009 +1000 124203 124204 exa: clarify createpixmap2 new pitch return 124205 124206commit 48ee5558333bd324463b6994735cabb23de262ec 124207Author: Aaron Zang <Aaron.Zang@Sun.COM> 124208Date: Mon Aug 3 23:21:39 2009 -0700 124209 124210 Add new VT support for OpenSolaris & future Solaris releases 124211 124212 Signed-off-by: Aaron Zang <Aaron.Zang@Sun.COM> 124213 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 124214 124215commit 613e76ff9055d8ac2b1af1130668180646a9e14c 124216Author: Alan Coopersmith <alan.coopersmith@sun.com> 124217Date: Mon Aug 3 21:38:51 2009 -0700 124218 124219 Remove support for Solaris x86 releases older than Solaris 8 124220 124221 If you want to run a pre-1999 kernel, you'll need a pre-2009 X server 124222 124223 [Some pre-Solaris 8 VT support is left by this patch to allow reuse by 124224 the new Solaris VT support that follows in the next patch.] 124225 124226 Signed-off-by: Aaron Zang <Aaron.Zang@Sun.COM> 124227 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 124228 124229commit 6c292d17053eb2a7e7054e51210f423dbc0cb7e8 124230Author: Peter Hutterer <peter.hutterer@who-t.net> 124231Date: Fri Aug 14 11:48:54 2009 +1000 124232 124233 dix: update the sprite trace for all masters && floating slaves (#23257) 124234 124235 When the windows are restructured, CheckMotion needs to be called for all 124236 masters and floating slaves to update the spriteTrace. 124237 124238 X.Org Bug 23257 <http://bugs.freedesktop.org/show_bug.cgi?id=23257> 124239 124240 Tested-by: Thomas Jaeger 124241 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124242 124243commit 49046088f10cceaea7da97401d742d3fb59371f5 124244Author: Peter Hutterer <peter.hutterer@who-t.net> 124245Date: Fri Aug 14 09:48:45 2009 +1000 124246 124247 config: don't shutdown the libhal ctx if it failed to initialize (#23213) 124248 124249 Regression introduced by b1c3dc6ae226db178420e3b5f297b94afc87c94c. 124250 Shutting down the libhal_ctx if the init failed may cause an abort. 124251 This can happen if hald is not yet running at server startup. 124252 124253 X.Org Bug 23213 <http://bugs.freedesktop.org/show_bug.cgi?id=23213> 124254 124255 Tested-by: Stefan Dirsch 124256 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124257 124258commit 1545a120df6dffb5b84fe96c5a992357520b7c8d 124259Author: Dave Airlie <airlied@redhat.com> 124260Date: Tue Aug 11 15:00:36 2009 +1000 124261 124262 exa: fix CreatePixmap2 to be useful for tiling. 124263 124264 This adds a pitch return so that the driver can align the pitch to any 124265 value it wishes and not just the one it gave to EXA at startup. 124266 124267commit db568f9eabf3450d8a023597ff007df355b13ea8 124268Author: Peter Hutterer <peter.hutterer@who-t.net> 124269Date: Thu Aug 13 16:37:00 2009 +1000 124270 124271 Xext: fix up wrong conditions for negative sync transitions. 124272 124273 If the counter had a value higher than the trigger value for a negative 124274 transition, the trigger value did not get set. 124275 124276 The correct sequence of checks is: 124277 if (positive transition) 124278 if (counter value < trigger value) 124279 set up trigger 124280 if (negative transition) 124281 if (counter value > trigger value) 124282 set up trigger 124283 124284 Red Hat Bug 501601 <https://bugzilla.redhat.com/show_bug.cgi?id=501601> 124285 124286 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124287 124288commit a3e50b05747cab92090ae7d7f4475cd61d3fcadf 124289Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 124290Date: Mon Jul 27 20:58:44 2009 +0200 124291 124292 Add 4 missing 10bpc picture formats to the server format list. 124293 124294 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 124295 124296commit 36e24a6d93bd5aced4e566b80bf2d03555fab9ca 124297Author: Peter Hutterer <peter.hutterer@who-t.net> 124298Date: Thu Aug 13 10:50:26 2009 +1000 124299 124300 Xext: add missing return code check to ProcSyncDestroyAlarm 124301 124302 Introduced with 57aff88c7d0761e590806d07bee1c9410680c89f. 124303 124304 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124305 124306commit ce69a06aff934b2dcded8606cab079ac6465007c 124307Author: Peter Hutterer <peter.hutterer@who-t.net> 124308Date: Tue Aug 11 16:16:54 2009 +1000 124309 124310 Xi: fix up broken DeviceChangedEvent swapping code 124311 124312commit b44d34d5fd0d8aaacb89121e8b4afba04f1dcc80 124313Author: Peter Hutterer <peter.hutterer@who-t.net> 124314Date: Mon Aug 10 15:32:59 2009 +1000 124315 124316 include: correct a copy/paste error in a comment. 124317 124318 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124319 124320commit c477ac8a151a28ba23f5c87e8f7affd6e82cfaa9 124321Author: Peter Hutterer <peter.hutterer@who-t.net> 124322Date: Mon Aug 10 15:15:02 2009 +1000 124323 124324 dix: re-name internal eventToClassesChanged to eventToDeviceChanged 124325 124326 DeviceClassesChangedEvents (where this name comes from) have been replaced 124327 with DeviceChangedEvents. 124328 124329 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124330 124331commit af38f6790c4c8ba1e70f5c5ba3530ded85d6e372 124332Author: Peter Hutterer <peter.hutterer@who-t.net> 124333Date: Mon Aug 10 14:08:59 2009 +1000 124334 124335 test: add event conversion tests for XIDeviceEvents 124336 124337commit 3f2e4b9867b6877ee7be32b151fcaf221ef0812f 124338Author: Peter Hutterer <peter.hutterer@who-t.net> 124339Date: Mon Aug 10 14:07:54 2009 +1000 124340 124341 Xi: add event swapping for XIRawEvents. 124342 124343 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124344 124345commit f3b2f9fb734ecfff6db9ae85b0d247856ede8112 124346Author: Peter Hutterer <peter.hutterer@who-t.net> 124347Date: Mon Aug 10 13:12:40 2009 +1000 124348 124349 Xi: fix event swapping for XIDeviceEvents. 124350 124351 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124352 124353commit e46f02fa2de79261221b42ab73f9daa2ce8ac650 124354Author: Peter Hutterer <peter.hutterer@who-t.net> 124355Date: Mon Aug 10 09:50:16 2009 +1000 124356 124357 Xext: allocate a separate event list for XTest events (#23100) 124358 124359 XTest event processing may be interrupted by a SIGIO. If Xtest uses the same 124360 event list as the rest of the server, this list may be overwritten 124361 in-flight. 124362 124363 X.Org Bug 23100 <http://bugs.freedesktop.org/show_bug.cgi?id=23100> 124364 124365 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124366 124367commit 8483b08152104dc619d9e44f877d2687d866355e 124368Author: Peter Hutterer <peter.hutterer@who-t.net> 124369Date: Fri Aug 7 16:11:37 2009 +1000 124370 124371 test: add focus and enter conversion testing. 124372 124373 Doesn't actually convert anything, but verify the expected behaviour. 124374 124375commit 59dc59a72ffd4cbc4df207bc688c92bb4863e8a9 124376Author: Peter Hutterer <peter.hutterer@who-t.net> 124377Date: Fri Aug 7 11:23:33 2009 +1000 124378 124379 test: add XI2 eventconversion test for raw events. 124380 124381 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124382 124383commit 10d7948e0360860e1e9633dca39f646d492e73bf 124384Author: Peter Hutterer <peter.hutterer@who-t.net> 124385Date: Fri Aug 7 10:17:14 2009 +1000 124386 124387 test: fix build error introduced by new AllocDevicePair API 124388 124389 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124390 124391commit 3f161a0aac39fbdeef393a17269486b8dace4672 124392Author: Peter Hutterer <peter.hutterer@who-t.net> 124393Date: Mon Aug 10 14:14:35 2009 +1000 124394 124395 Xi: un-statify XI2EventSwap, it is needed for tests. 124396 124397 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124398 124399commit 55e1ea08d03d89ecc0f2db7652a4d15567204696 124400Author: Peter Hutterer <peter.hutterer@who-t.net> 124401Date: Mon Aug 3 16:46:11 2009 +1000 124402 124403 Xi: get device changed data from DeviceChangedEvents, not the device (#23100) 124404 124405 If a new device posts an event while the DCE is in the queue, getting the 124406 data from the device may result in invalid memory access. 124407 124408 X.Org Bug 23100 <http://bugs.freedesktop.org/show_bug.cgi?id=23100> 124409 124410commit 8ce4fde76fdf747fd3e00836c9c9c542e516ae9a 124411Author: Matt Turner <mattst88@gmail.com> 124412Date: Wed Aug 12 14:45:13 2009 -0400 124413 124414 Fix build on Alpha 124415 124416commit a400dbb38f93030d51afe806b4b20d5ef501c855 124417Author: Colin Harrison <colin.harrison@virgin.net> 124418Date: Wed Aug 12 17:48:36 2009 +0100 124419 124420 Xming: Add NET_WM_ICON to native icon conversion 124421 124422 fd.o bugzilla #4491 124423 originally from a patch by Joe Krahn <jkrahn@nc.rr.com> 124424 124425 Convert a NET_WM_ICON to a native icon by converting to a native 124426 bitmap and then using CreateIconIndirect() 124427 124428 Don't use icon alpha on Windows 2000 or if display isn't 32-bit, convert 124429 alpha channel to a 1-bit transparency mask using a threshold value 124430 124431 Fix warning in winScaleXBitmapToWindows() about signedness of *iconData 124432 124433 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 124434 124435commit 211511f150ce189681aff176c311fa8f312517e9 124436Author: Soren Sandmann Pedersen <ssp@redhat.com> 124437Date: Tue Aug 11 14:40:42 2009 -0400 124438 124439 Disable the out-of-bounds workaround in pixman. 124440 124441 This workaround was necessary for older X servers, but now fb should 124442 be using correct coordinates. 124443 124444 Also bump pixman requirement to 0.15.20. 124445 124446commit 93d9646c713336e03c135204c061f561d3654e23 124447Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124448Date: Sat Aug 8 20:18:10 2009 +0300 124449 124450 mi: fix indentation 124451 124452 mieq.c looks indented-wise now. Let's see how long it will take to someone 124453 mess it again. 124454 124455 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124456 124457commit f9a2fff2248d7254958857677cabfea914ed4853 124458Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124459Date: Wed Aug 5 21:02:29 2009 +0300 124460 124461 mi: fix cursor warping screens 124462 124463 The server was processing ET_RawMotion type when the cursor was wrapping to 124464 another screen and getting wrong valuator values. This fix such issue 124465 considering only ET_Motion, ET_KeyPress, ET_KeyRelease, ET_ButtonPress and 124466 ET_ButtonRelease types when the cursor detects a new screen, keeping the 124467 "normal" processing of device events. 124468 124469 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124470 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124471 124472commit d573cc46d3106824902ab4f926617bd9959af57c 124473Author: Maarten Maathuis <madman2003@gmail.com> 124474Date: Sat Aug 8 11:32:05 2009 +0200 124475 124476 exa: more safety 124477 124478commit 8b652435cd42929e2d187b353b3b20e798569356 124479Author: Maarten Maathuis <madman2003@gmail.com> 124480Date: Sat Aug 8 10:35:01 2009 +0200 124481 124482 exa: minor cleanup 124483 124484commit a73f95aa93634cf7ba2c7c9274a046b2df2adca7 124485Author: Maarten Maathuis <madman2003@gmail.com> 124486Date: Sat Aug 8 02:31:23 2009 +0200 124487 124488 exa: Fix the broken upload fallback for "mixed" 124489 124490 - Replace it with something wfb friendly while i'm at it. 124491 124492commit c029678bf7883b91f3c8095e764abbee246c9bb6 124493Author: Julien Cristau <jcristau@debian.org> 124494Date: Fri Aug 7 20:42:30 2009 +0200 124495 124496 configure.ac: drop dependency on fontenc 124497 124498 Nothing in the server uses this anymore. 124499 124500commit 3943df7f76c1b4930fb7370d9d145cee96dd562b 124501Author: Maarten Maathuis <madman2003@gmail.com> 124502Date: Fri Aug 7 20:36:03 2009 +0200 124503 124504 exa: one can never be too careful 124505 124506commit 1548e8ae5d28ed0eb6057a1a19cfc84a78ef34dc 124507Author: Maarten Maathuis <madman2003@gmail.com> 124508Date: Fri Aug 7 20:04:53 2009 +0200 124509 124510 exa: fix a potential loophole in "mixed" 124511 124512 - Always free sys_ptr before setting the pixmap to pinned. 124513 124514commit e94c7c42ce9d6a194b76ba7e8b3904b180f93a1b 124515Author: Dave Airlie <airlied@redhat.com> 124516Date: Fri Aug 7 13:36:52 2009 +1000 124517 124518 parser: make libxf86config_internal.la not installed. 124519 124520commit 5fb188b547e0ef4e2f2c59b66a3183b004e4a1e9 124521Author: Dave Airlie <airlied@redhat.com> 124522Date: Fri Aug 7 12:05:51 2009 +1000 124523 124524 ddx: fix xf86Config.a generation 124525 124526 We were generating a shared library, but this lib is foobar, the parser 124527 requires some symbols from the X server or from the program its being linked 124528 into. If the program its being linked into (say a python .so) has symbol 124529 visibility enabled then it will fail to dynamic link, also if this .so has 124530 symbol visiblity enabled it will fail to dynamic link. 124531 124532 Screw it go back to a .a file really unless someone cleans it up properly. 124533 124534 Signed-off-by: Dave Airlie <airlied@redhat.com> 124535 124536commit 3047bd067464efb9857960d3fa6324b947faa970 124537Author: Maarten Maathuis <madman2003@gmail.com> 124538Date: Wed Aug 5 18:39:47 2009 +0200 124539 124540 exa: delay malloc for "mixed" 124541 124542commit e8ac2ed5dc4c2ac0a5e1e1f371f94c15b1c729dd 124543Author: Maarten Maathuis <madman2003@gmail.com> 124544Date: Wed Aug 5 16:12:16 2009 +0200 124545 124546 exa: implement exaMoveInPixmap for "mixed" 124547 124548 - This can be used to force creation of driver pixmap. 124549 - Not for 1 or 4 bpp. 124550 - Driver can still fail (driver) pixmap creation. 124551 124552commit 9d2a7128d3e66b8c076a714d69f84bcad49391b9 124553Author: Maarten Maathuis <madman2003@gmail.com> 124554Date: Sun Aug 2 02:35:46 2009 +0200 124555 124556 exa: Use damage to optimise away useless copies. 124557 124558commit 03ecb164f2592c954aa408bf121e0c67b604d854 124559Author: Maarten Maathuis <madman2003@gmail.com> 124560Date: Sat Aug 1 19:19:19 2009 +0200 124561 124562 exa: A simple 3rd backend implementation. 124563 124564 - Based on driver pixmaps with some changes (completely transparent to driver). 124565 - It helps with the problem of known software fallbacks, such as trapezoids. 124566 - exaDoMigration is now called for all cases that provide a do_migration hook. 124567 - exa_migration.c is renamed to exa_migration_classic.c 124568 124569commit ac7ac913fd98ea359c05c89968ab53a3223615b4 124570Author: Maarten Maathuis <madman2003@gmail.com> 124571Date: Wed Jul 22 21:35:41 2009 +0200 124572 124573 exa: Split out some classic and driver allocated pixmap code into seperate files 124574 124575 - Create a few seperate functions and a few private function pointers. 124576 - Replace a few if conditions with a check for pExaPix->pDamage instead. 124577 - This is in preperation of a third scheme that lies somewhere in between. 124578 - Code clarity would have suffered (i started working on it and didn't like the mess). 124579 124580commit 3abbd327f4a732408119de1f8e9ecba4812772a5 124581Author: Peter Hutterer <peter.hutterer@who-t.net> 124582Date: Wed Aug 5 15:47:46 2009 +1000 124583 124584 dix: detach SD during XI2 grabs only. 124585 124586 XI1 grabs on slave devices leave the device attached - just like in earlier 124587 versions of XI. 124588 124589 Tested-by: Thomas Jaeger 124590 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124591 124592commit c29aa7da220661532b05972cacd3dbaff29408b5 124593Author: Peter Hutterer <peter.hutterer@who-t.net> 124594Date: Wed Aug 5 15:33:07 2009 +1000 124595 124596 Revert "dix: Remove temporary detachment of slave devices." 124597 124598 This reverts commit 0c0ef42292f4c910c73b308cd75d77637312da53. 124599 124600 Tested-by: Thomas Jaeger 124601 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124602 124603commit 931160fcf378120e58849d801dbc62bed6e65a4b 124604Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 124605Date: Tue Aug 4 02:53:17 2009 -0700 124606 124607 XQuartz: AIGLX: Provide empty __glXAquaDrawableResize to avoid crashing in DoMakeCurrent 124608 (cherry picked from commit 60a1d2c2764f2f02c0751940a264588717afce79) 124609 124610commit 0fe639a314de397516cf00d8100da8086abdd97a 124611Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 124612Date: Sat Aug 1 22:53:45 2009 -0700 124613 124614 XQuartz: Only save lastpt on mouse/tablet events 124615 (cherry picked from commit 552be074e5be6492df3e290e8b7d9daff1a2cb34) 124616 124617commit 491c2d74c12d7f9bc9f31b427ece04e049ebff5a 124618Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 124619Date: Sat Aug 1 12:43:14 2009 -0700 124620 124621 XQuartz: Use mouseLocation rather than locationInWindow when setting lastpt 124622 124623 I don't understand the *why* ... I just see that it works better this way for games like Quake2 through wine. It *should* be better the other way, but somehow it's not. 124624 124625 I guess this will go in my list of puzzles to unravel. 124626 (cherry picked from commit 65ae2d00e1a53f97f2ff9522406ab69d50bf3199) 124627 124628commit df6c01352471c766e9f71ceac03cab8c3911faa1 124629Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 124630Date: Sat Aug 1 02:32:52 2009 -0700 124631 124632 XQuartz: Don't use location delta for tablets since NSEvent does not give a precise delta. 124633 (cherry picked from commit 6c5bf756a7f5389cdfe2e43a339d7c31a3e522e9) 124634 124635commit e360104880e6e2e666aa05dfd56e2ef3880f38ef 124636Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 124637Date: Wed Aug 5 16:17:01 2009 -0700 124638 124639 XQuartz: Purge redundant QuartzBell 124640 (cherry picked from commit de14a63d20095e1537fd74352850c734d900031d) 124641 124642commit 8c0085c715effdc450d78eec14bc32e6214c78af 124643Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124644Date: Wed Aug 5 22:50:44 2009 +0300 124645 124646 xfree86: remove more RAC junky 124647 124648 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124649 124650commit 444723273ae82fc3f6707ed1a461aaaa8fdae39c 124651Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124652Date: Wed Aug 5 22:48:02 2009 +0300 124653 124654 xfree86: reorganize pci code 124655 124656 - xf86PciVideoInfo doesn't need to be global 124657 - remove unused macros 124658 124659 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124660 124661commit 20169414e1afd5d1d02cb1b57866b1c158b2fc6c 124662Author: Tiago Vignatti <tiago.vignatti@nokia.com> 124663Date: Wed Aug 5 22:44:19 2009 +0300 124664 124665 xfree86: remove unused functions 124666 124667 RAC trash. 124668 124669 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 124670 124671commit 95b678e6dc41f2524ada4eb11289687fafce7588 124672Author: Alan Coopersmith <alan.coopersmith@sun.com> 124673Date: Mon Aug 3 23:49:56 2009 -0700 124674 124675 Correct modifier map built when ProcSetModifierMapping is called 124676 124677 Fixes xmodmap changes to modifiers to stop corrupting modifier maps 124678 124679 Previous code had two bugs: 124680 - the code to increment mod was after the code to continue if no 124681 modifier was set, so mod wouldn't be incremented for modifiers 124682 with no keys mapped to them (such as if you called 124683 xmodmap -e 'clear Lock') 124684 - the value it set in the modifier map was the raw modifier number, 124685 not the bitmask value for that modifier 124686 124687 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 124688 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124689 124690commit 4ca305956e5ea6f606b22ef62aa462186a7b95f0 124691Author: Peter Hutterer <peter.hutterer@who-t.net> 124692Date: Wed Aug 5 09:40:56 2009 +1000 124693 124694 Re-enable Xnest by default. 124695 124696 Xnest was disabled in 82fc102568b3d6b0daeb6c5f5b3a1310a7f14fcd due do build 124697 errors. These errors have since been fixed. 124698 124699 Re-enable Xnest by default to increase coverage by default builds. 124700 124701 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124702 124703commit 5904ef2ccd6056b187ca76f104c21e2d686bfc1d 124704Author: Peter Hutterer <peter.hutterer@who-t.net> 124705Date: Wed Aug 5 10:40:20 2009 +1000 124706 124707 xnest: restore xnestUpdateModifierState 124708 124709 The meat of xnestUpdateModifierState was ifdef'd out in 124710 6ef46c40e62def4841a4cff4e0b443516a2ed782. This resulted in stuck modifiers 124711 when a modifier key release event wasn't sent to Xnest (e.g. Alt-Tab away). 124712 124713 See X.Org Bug 3664 <https://bugs.freedesktop.org/show_bug.cgi?id=3664> for 124714 the original bug report. 124715 124716 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124717 124718commit 44afc7577ad3339f4555438388f6a4d8e220c991 124719Author: Peter Hutterer <peter.hutterer@who-t.net> 124720Date: Wed Aug 5 09:51:05 2009 +1000 124721 124722 xnest: include exevents.h for XIGetKnownProperty. 124723 124724 Silences compiler warning. 124725 Pointer.c: In function ‘xnestPointerProc’: 124726 Pointer.c:64: warning: implicit declaration of function ‘XIGetKnownProperty’ 124727 Pointer.c:64: warning: nested extern declaration of ‘XIGetKnownProperty’ 124728 124729 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124730 124731commit 32ce5c83a504d2cafd2b19ed628b40032ef347b3 124732Author: Peter Hutterer <peter.hutterer@who-t.net> 124733Date: Wed Aug 5 09:39:47 2009 +1000 124734 124735 xnest: use AllocDevicePair for Xnest device initalization 124736 124737 All Xnest needs is a single pointer+keyboard pair. AllocDevicePair sets them 124738 up nicely with the name assigned etc. 124739 124740 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124741 124742commit 26b83ad4a29dc180b336a19d9e97589814e93e37 124743Author: Peter Hutterer <peter.hutterer@who-t.net> 124744Date: Wed Aug 5 09:32:50 2009 +1000 124745 124746 dix: require PointerProc and KeyboardProc to be passed into AllocDevicePair. 124747 124748 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124749 124750commit 6a500fdd4d717947b86f6d6844ebf9a8603eb8e2 124751Author: Peter Hutterer <peter.hutterer@who-t.net> 124752Date: Tue Aug 4 15:29:55 2009 +1000 124753 124754 dmx: include exglobals.h for DeviceKeyPress and friends. 124755 124756 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124757 124758commit 591c2f8758ef36a9bbf1625f9c207a1804f58ef4 124759Author: Peter Hutterer <peter.hutterer@who-t.net> 124760Date: Tue Aug 4 15:30:52 2009 +1000 124761 124762 dmx: typecast to InternalEvent* before mieqEnqueue. 124763 124764 Silences a few compiler warnings. 124765 124766 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124767 124768commit 15b425b5a6630d2ac95b490f4f4dbb9240ac9c5d 124769Author: Peter Hutterer <peter.hutterer@who-t.net> 124770Date: Tue Aug 4 15:31:42 2009 +1000 124771 124772 dmx: move 'state' around to silence compiler warning. 124773 124774 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124775 124776commit e10f802c8260e034fd6c835d3f2622d9cd5085f8 124777Author: Peter Hutterer <peter.hutterer@who-t.net> 124778Date: Tue Aug 4 15:40:17 2009 +1000 124779 124780 dmx: include inputstr.h for inputInfo.pointer 124781 124782 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124783 124784commit 9ffc3f280b3907a672f99101548a5eb52027b37f 124785Author: Peter Hutterer <peter.hutterer@who-t.net> 124786Date: Tue Aug 4 15:40:31 2009 +1000 124787 124788 dmx: don't call InitKbdFeedbackClassDeviceStruct - doesn't exist anymore. 124789 124790 InitKeyboardDeviceStruct does the same task now. 124791 124792 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124793 124794commit 63155cf985ee15bf6aad95066b076ab680cf5a31 124795Author: Peter Hutterer <peter.hutterer@who-t.net> 124796Date: Tue Aug 4 15:28:30 2009 +1000 124797 124798 dmx: pass button/axis labels into the init functions (currently unset) 124799 124800 Currently only None labels are passed in, in the future these labels should 124801 be whatever the respective buttions/axes are. 124802 124803 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124804 124805commit af85973a0a80fb0ba82330ae3ec897f65d618003 124806Author: Peter Hutterer <peter.hutterer@who-t.net> 124807Date: Tue Aug 4 15:25:06 2009 +1000 124808 124809 dmx: remove now-useless defines. 124810 124811 XI is mandatory so we can expect both to be defined. 124812 124813commit 654d7a02600bbafd890f555a081aa0a2fcb5e253 124814Author: Peter Hutterer <peter.hutterer@who-t.net> 124815Date: Tue Aug 4 15:23:01 2009 +1000 124816 124817 Revert "dmx: claim we support XI 2." 124818 124819 XQueryInputVersion doesn't exist anymore and for now XI 1 support is good 124820 enough. 124821 124822 This reverts commit 8da8a0fec4b1b9d9208635dedb2f449dc99e0004. 124823 124824commit e7c950c6273ff93f25c091c93d86da6e332a2277 124825Author: Peter Hutterer <peter.hutterer@who-t.net> 124826Date: Tue Aug 4 15:15:24 2009 +1000 124827 124828 dmx: fix a call to XkbSetRulesDflts, takes XkbRMLVOSet now. 124829 124830commit 2cde9208ff756e33d162e2324f4b99540230d743 124831Author: Peter Hutterer <peter.hutterer@who-t.net> 124832Date: Tue Aug 4 14:58:25 2009 +1000 124833 124834 dmx: Enable/DisableDevice take a boolean variable now. 124835 124836 We want to send events here, so pass in TRUE. 124837 124838 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124839 124840commit 1e210d6d10ec63d22247e5801890e024826fc861 124841Author: Peter Hutterer <peter.hutterer@who-t.net> 124842Date: Tue Aug 4 14:41:46 2009 +1000 124843 124844 xkb: remove now-unused XkbGetKeysym. 124845 124846 XkbGetKeysyms was only used by the now-removed Keysym grabs. 124847 124848 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124849 124850commit a6ce6c70cff5108f0751b662b8e52c83daab0722 124851Author: Michel Dänzer <daenzer@vmware.com> 124852Date: Tue Aug 4 23:23:21 2009 +0200 124853 124854 EXA: Simplify exaGetPixmapFirstPixel using GetImage. 124855 124856commit 842373104d08d47efc863cecbe30431d3faebef1 124857Author: Michel Dänzer <daenzer@vmware.com> 124858Date: Tue Aug 4 23:23:21 2009 +0200 124859 124860 Add support for RENDER BGRA formats. 124861 124862commit 0bf7eaf3052ce24066b0a7c14860b4762fb81364 124863Author: Alan Coopersmith <alan.coopersmith@sun.com> 124864Date: Mon Aug 3 23:14:50 2009 -0700 124865 124866 Ansify function arguments in VTsw_noop.c 124867 124868 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 124869 124870commit 736f2d64725c6df8413e627bd40ce7ecb011acc7 124871Author: Paul Bender <pebender@gmail.com> 124872Date: Tue Jun 30 15:50:46 2009 -0800 124873 124874 Bug 16832: XDMCP related build error when --disable-xdmcp is used 124875 124876 X.Org Bugzilla #16832: http://bugs.freedesktop.org/show_bug.cgi?id=16832 124877 Patch #27279: http://bugs.freedesktop.org/attachment.cgi?id=27279 124878 124879commit b159a98f7fbdb3f8b337cd1bfd2ff27312856619 124880Author: Lee Leahu <freedesktop-bugs@dyweni.com> 124881Date: Thu Oct 23 21:03:57 2008 -0500 124882 124883 seg fault when initializing DMX screens 124884 124885 X.Org Bugzilla #18086: http://bugs.freedesktop.org/show_bug.cgi?id=18086 124886 Patch #19837: http://bugs.freedesktop.org/attachment.cgi?id=19837 124887 124888commit 5da9b255a64bec7dbf5ddb392d54dac9be5b43c0 124889Author: Alan Coopersmith <alan.coopersmith@sun.com> 124890Date: Mon Aug 3 19:53:27 2009 -0700 124891 124892 Fix blddir != srcdir builds of solaris-*.il files 124893 124894 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=17509 124895 124896 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 124897 124898commit 1e69fd4a60147287b31e53bfc61543fb17bb82c8 124899Author: Peter Hutterer <peter.hutterer@who-t.net> 124900Date: Mon Aug 3 12:03:26 2009 +1000 124901 124902 xnest: fix up parameters to InitKeyboardDeviceStruct. 124903 124904 IKDS takes a DeviceIntPtr as first argument, and an RMVLO struct as second. 124905 The keysyms stuff is long gone now. 124906 124907 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124908 124909commit 008658049cbeea35a9f76f98037aa2f4173f3573 124910Author: Peter Hutterer <peter.hutterer@who-t.net> 124911Date: Mon Aug 3 12:00:45 2009 +1000 124912 124913 xnest: silence compiler warnings by typecasing properly. 124914 124915 Events.c: In function ‘xnestQueueKeyEvent’: 124916 Events.c:112: warning: passing argument 2 of ‘mieqEnqueue’ from incompatible 124917 pointer type 124918 ../../mi/mi.h:203: note: expected ‘union InternalEvent *’ but argument is of 124919 type ‘struct xEvent *’ 124920 Events.c: In function ‘xnestCollectEvents’: 124921 Events.c:141: warning: passing argument 2 of ‘mieqEnqueue’ from incompatible 124922 pointer type 124923 ../../mi/mi.h:203: note: expected ‘union InternalEvent *’ but argument is of 124924 type ‘struct xEvent *’ 124925 Events.c:150: warning: passing argument 2 of ‘mieqEnqueue’ from incompatible 124926 pointer type 124927 ../../mi/mi.h:203: note: expected ‘union InternalEvent *’ but argument is of 124928 type ‘struct xEvent *’ 124929 Events.c:160: warning: passing argument 2 of ‘mieqEnqueue’ from incompatible 124930 pointer type 124931 ../../mi/mi.h:203: note: expected ‘union InternalEvent *’ but argument is of 124932 type ‘struct xEvent *’ 124933 Events.c:193: warning: passing argument 2 of ‘mieqEnqueue’ from incompatible 124934 pointer type 124935 ../../mi/mi.h:203: note: expected ‘union InternalEvent *’ but argument is of 124936 type ‘struct xEvent *’ 124937 124938 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124939 124940commit 664ac92d8bbe956dd6fd80fac5dc3161028803b2 124941Author: Peter Hutterer <peter.hutterer@who-t.net> 124942Date: Fri Jul 31 14:38:35 2009 +1000 124943 124944 xfixes: backup the DisplayCursor/CloseScreen proc before restoring it (#23034) 124945 124946 The screen's DisplayCursor func is wrapped as 124947 AnimCurDisplayCursor -> CursorDisplayCursor -> miPointerDisplayCursor. 124948 124949 Calling CursorDisplayCursor while an animated cursor was currently displayed 124950 would remove AnimCurDisplayCursor from the wrap stack. Thus, the next call 124951 to ChangeToCursor wouldn't update the animated cursor state. The block 124952 handler for animated cursors would then continuously overwrite the actual 124953 cursor, leaving an animated cursor everywhere on the screen. 124954 124955 X.Org Bug 23034 <http://bugs.freedesktop.org/show_bug.cgi?id=23034> 124956 124957commit f48dfcc1b772a09428e328c72124ea0d46a73416 124958Author: Peter Hutterer <peter.hutterer@who-t.net> 124959Date: Fri Jul 31 10:44:28 2009 +1000 124960 124961 xfixes: allocate CurrentCursor for all devices. 124962 124963 XFixes requires cursor notifies to clients when the cursor changes. This 124964 should work on the ClientPointer and then on all master pointers. Hence 124965 change CurrentCursor to a MAXDEVICES array. 124966 124967 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124968 124969commit d3e5629fac224d53a972df5e2a87db02534f9b17 124970Author: Peter Hutterer <peter.hutterer@who-t.net> 124971Date: Thu Jul 30 16:36:56 2009 +1000 124972 124973 dix: call SetFocusOut and LeaveWindow when disabling a device. 124974 124975 PointerWindows[x] would be set after removing a master pointer. Destroying 124976 this window then crashed the server. 124977 124978 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124979 124980commit 97ed946cfccec5979c47b5fabf1ad56cc7d33ef4 124981Author: Peter Hutterer <peter.hutterer@who-t.net> 124982Date: Thu Jul 30 16:36:30 2009 +1000 124983 124984 dix: SetFocusOut and LeaveWindow don't need parameters other than dev. 124985 124986 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124987 124988commit d040940efe041b57e6323921c380ceb2bb43f41e 124989Author: Peter Hutterer <peter.hutterer@who-t.net> 124990Date: Thu Jul 30 16:16:51 2009 +1000 124991 124992 xfixes: FixesHideCursor should work on all master pointers. 124993 124994 Presumably, a client calling HideCursor doesn't want any cursor displayed. 124995 124996 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 124997 124998commit b83c682e7ea4e10757eca6363de28281108591a8 124999Author: Oliver McFadden <oliver.mcfadden@nokia.com> 125000Date: Mon Aug 3 02:46:57 2009 +0300 125001 125002 xf86Xinput: xf86PostButtonEventP must set POINTER_ACCELERATE for relative events. 125003 125004 Thanks to Peter Hutterer for spotting this bug. 125005 125006commit ae65daab080c6269f0eadc6be1a48bf27866c542 125007Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125008Date: Fri Jul 31 15:32:07 2009 -0700 125009 125010 XQuartz: Unify how we set our bitmasks for visuals 125011 (cherry picked from commit c230b52c27ce50ac6c27011ec8e88f1b263b25f9) 125012 125013commit 4a53fed41d711f270c7d8a7fbeaafb4fb619e879 125014Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125015Date: Fri Jul 31 14:19:50 2009 -0700 125016 125017 XQuartz: Dead code removal for StaticColor visual 125018 (cherry picked from commit e457a44e87950207f6c16bf82bf5af11c777ecf1) 125019 125020commit da0a2747803dbecb1308181176a784e14cf02809 125021Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125022Date: Fri Jul 31 14:14:21 2009 -0700 125023 125024 XQuartz: Cleanup the bitmask setting for GLX visuals. 125025 (cherry picked from commit d32c3df258e748958ef997c675dc4fae118c0d7b) 125026 125027commit 70ac671af2f092471022590cb7a19ef9155c51b4 125028Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125029Date: Thu Jul 30 01:30:52 2009 -0700 125030 125031 XQuartz: no DirectColor 125032 (cherry picked from commit df2fbc410f2c484612f65a6539a6cb069ef4a468) 125033 125034commit c0c72a866a237d3c2d9e69e1c69181ef5446e3f8 125035Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125036Date: Thu Jul 30 00:51:55 2009 -0700 125037 125038 XQuartz: Define DDXRingBell 125039 (cherry picked from commit 113347381289497cb2a79994d0ef5f427ae63ac5) 125040 125041commit 949811313989809f5b58424af04b9fd8e1d0bec4 125042Author: Peter Hutterer <peter.hutterer@who-t.net> 125043Date: Thu Jul 30 21:59:40 2009 +1000 125044 125045 test: fix build by including eventstr.h 125046 125047 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125048 125049commit 0565f4ed4519962bed40a0bbcf0b409471f4de40 125050Author: Peter Hutterer <peter.hutterer@who-t.net> 125051Date: Wed Jul 29 16:46:45 2009 +1000 125052 125053 Xi: set the sourceid for focus devices to the device id. 125054 125055 Unlike Enter/Leave events generated by a device pushing the pointer around, 125056 a device doesn't change focus all by itself. It's a result of a 125057 SetInputFocus call, a window becoming unviewable or a grab activating. As 125058 such, the sourceid for focus events is always the deviceid itself. 125059 125060 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125061 125062commit de4dd5848cab90b0f8b8243ca0b49985ef047124 125063Author: Peter Hutterer <peter.hutterer@who-t.net> 125064Date: Wed Jul 29 16:45:34 2009 +1000 125065 125066 include: DeviceFocusEvent is not to be exported. 125067 125068 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125069 125070commit 46ac9f92416f3cb99b5d84a9d200237dc33a3bb7 125071Author: Peter Hutterer <peter.hutterer@who-t.net> 125072Date: Wed Jul 29 16:34:28 2009 +1000 125073 125074 dix: pass the sourceid around for enter/leave events. 125075 125076 The sourceid for enter/leave events as a result of pointer motion is the ID 125077 of the slave device. The sourceid for those as a result of a grab activating 125078 is the device itself. 125079 125080 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125081 125082commit c299b2228fb63c192b72851c90e14ceaceb67bbc 125083Author: Peter Hutterer <peter.hutterer@who-t.net> 125084Date: Wed Jul 29 16:00:19 2009 +1000 125085 125086 dix: remove obsolete comment, parameter described doesn't exist. 125087 125088 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125089 125090commit 5085ac09a50721d87196bd9f2607dc76200ca399 125091Author: Peter Hutterer <peter.hutterer@who-t.net> 125092Date: Wed Jul 29 13:45:32 2009 +1000 125093 125094 input: switch internal event types to enums. 125095 125096 Use enum EventType instead of ints. This requires a load of default 125097 cases in various switch statements to silence compiler warnings. 125098 125099 Reported-by: Aaron Plattner 125100 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125101 125102commit 1ae8332d643299a3ee9a9f45a8e25b8c87c751e1 125103Author: Peter Hutterer <peter.hutterer@who-t.net> 125104Date: Wed Jul 29 13:39:38 2009 +1000 125105 125106 include: fix enum EventType declaration. 125107 125108 Having EventType after the enum declares a variable. silly me. 125109 125110 Reported-by: Aaron Plattner 125111 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125112 125113commit 994f7a1c814a89e90f710dac5bf6b2445fb64712 125114Author: Peter Hutterer <peter.hutterer@who-t.net> 125115Date: Wed Jul 29 12:11:13 2009 +1000 125116 125117 record: silence some compiler warnings. 125118 125119 warning: passing argument 4 of ‘RecordAProtocolElement’ discards qualifiers 125120 from pointer target type 125121 note: expected ‘pointer’ but argument is of type ‘const void *’ 125122 125123 record.c:2745: warning: passing argument 1 of ‘SwapConnSetupInfo’ from 125124 incompatible pointer type 125125 ../include/swaprep.h:243: note: expected ‘char *’ but argument is of type 125126 ‘struct xConnSetup *’ 125127 125128 record.c:2745: warning: passing argument 1 of ‘SwapConnSetupInfo’ from 125129 incompatible pointer type 125130 ../include/swaprep.h:243: note: expected ‘char *’ but argument is of type 125131 ‘struct xConnSetup *’ 125132 125133 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125134 125135commit a863d636293cd7361639c1a8cf9c4f7f15da1e1d 125136Author: Peter Hutterer <peter.hutterer@who-t.net> 125137Date: Wed Jul 29 12:09:34 2009 +1000 125138 125139 Xi: remove FIXME and obsolete include. 125140 125141 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125142 125143commit fac49df08f173f091cbb77feaf373d7d465358af 125144Author: Peter Hutterer <peter.hutterer@who-t.net> 125145Date: Wed Jul 29 12:07:22 2009 +1000 125146 125147 Xi: remove obsolete comment. 125148 125149 XI1 only uses 7 bits for deviceids, bit 8 is used for the MORE_EVENTS flag 125150 on the wire (when DeviceValuator events are required). 125151 125152 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125153 125154commit 845e65f08059e8f4bfd37356e99b48bba9416c0c 125155Author: Peter Hutterer <peter.hutterer@who-t.net> 125156Date: Wed Jul 29 11:54:14 2009 +1000 125157 125158 xkb: move XkbFilterEvents to xkbsrv.h 125159 125160 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125161 125162commit 8da0ff2d51086666d10ca7330d428e8610a4a0e3 125163Author: Oliver McFadden <oliver.mcfadden@nokia.com> 125164Date: Tue Jul 28 08:20:37 2009 +0300 125165 125166 xf86Xinput: Add the xf86Post(Proximity|Button|Key)EventP helper functions. 125167 125168 xf86PostKeyboardEvent also makes use of xf86PostKeyEventP to avoid code 125169 duplication, and the valuator verification has been split into the 125170 XI_VERIFY_VALUATORS macro. 125171 125172 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125173 125174commit a148d407429c7d13136b3fcafd2d279c5438df73 125175Author: Peter Hutterer <peter.hutterer@who-t.net> 125176Date: Tue Jul 28 16:53:51 2009 +1000 125177 125178 xkb: restore XKB PtrBtn actions. 125179 125180 Ifdef'd out since the switch to internal events. PtrBtn actions now work 125181 again. Instead of generating the event directly, GPE generates the event and 125182 it is then posted through the usual event processing routines 125183 (mieqProcessDeviceEvent). 125184 125185 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125186 125187commit 3d3b8babd1a5407082f1a40875ed69f62ba2153f 125188Author: Peter Hutterer <peter.hutterer@who-t.net> 125189Date: Tue Jul 28 16:51:14 2009 +1000 125190 125191 mi: update master event copying to InternalEvents. 125192 125193 This is long overdue. The device events are InternalEvents now (and only one 125194 at a time), diminishing the need for an EventList for the master event. 125195 Furthermore, don't make masterEvent a static since this will interfere if 125196 mieqProcessDeviceEvent is called from somewhere else (e.g. XKB actions). 125197 125198 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125199 125200commit f85619b14d130ec54d42cabfaee15e55ced0c665 125201Author: Peter Hutterer <peter.hutterer@who-t.net> 125202Date: Tue Jul 28 16:48:56 2009 +1000 125203 125204 dix: update GetMaximumEventsNum() to real value (3). 125205 125206 GPE and friends now use internal events so they may generate up to 3 events. 125207 One (optional) DeviceChanged event and one raw event plus a device event. 125208 125209 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125210 125211commit 0217d0370c0b0bce66a9c09092eda8e820274e2e 125212Author: Peter Hutterer <peter.hutterer@who-t.net> 125213Date: Tue Jul 28 14:54:30 2009 +1000 125214 125215 record: ifdef out RecordExtensionInit and print a warning to the log. 125216 125217 The RECORD extension is currently broken. By ifdef'ing out the content of 125218 RecordExtensionInit the extension isn't added to the server's internal list 125219 and it does not get advertised to the client. Clients can thus fail 125220 gracefully with a "extension not supported" instead of waiting forever for 125221 events that never arrive. 125222 125223 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125224 125225commit 4e9b2938cd8637a5d3b0a4c9f69d6ee75faab3a0 125226Author: Peter Hutterer <peter.hutterer@who-t.net> 125227Date: Mon Jul 27 16:54:33 2009 +1000 125228 125229 include: untangle events.h from the SDK headers. 125230 125231 InternalEvents shouldn't be used anywhere outside the X server itself. Split 125232 up into events.h for opaque typedefs for the events needed by various 125233 headers and eventstr.h for the actual struct definitions. 125234 125235 eventstr.h must only be included by code that requires internal events and 125236 is not part of the SDK. 125237 125238 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125239 125240commit fad5f96c01811af7490a071719e4017c019a8d9b 125241Author: Jon TURNEY <jon.turney@dronecode.org.uk> 125242Date: Wed Jul 29 13:50:48 2009 +0100 125243 125244 Cygwin/X: Only try to build rootless extension if multiwindow extwm mode is being built 125245 125246 Rootless extension still needs a bit more work to build successfully for Cygwin/X 125247 125248commit de2ae521abde445daaf025a07aa01563ca5ddd41 125249Author: Colin Harrison <colin.harrison@virgin.net> 125250Date: Sun Jul 26 20:48:19 2009 +0100 125251 125252 Xming: Use RegisterClassEx() instead of superseded RegisterClass() 125253 125254 RegisterClass is supserseded by RegisterClassEx, so change to using that everywhere 125255 125256 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 125257 125258commit 764ce6ee683db342264bbca4df6379eb6093fb85 125259Author: Colin Harrison <colin.harrison@virgin.net> 125260Date: Tue Jul 28 16:16:57 2009 +0100 125261 125262 Xming: Use GetClassLongPtr() instead of superseded GetClassLong() 125263 125264 GetClassLong() is superseded by GetClassLongPtr(), so change to using that 125265 125266commit a85523dc50f392a33a1c00302a0946828bc9249d 125267Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125268Date: Tue Jul 28 22:52:33 2009 -0700 125269 125270 XQuartz: Avoid a possible spinlock in applicationWillTerminate 125271 (cherry picked from commit f430cda0fdcc1a8fc5f4795743b40f09ff0bd869) 125272 125273commit 1e49c8d340ee8d8b6f90abcc2e2e9c390cbcd622 125274Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125275Date: Tue Jul 28 13:40:15 2009 -0700 125276 125277 XQuartz: Avoid namespace collission for BOOL in Sparkle 125278 (cherry picked from commit 227c6e01a641b72158201553273299283cdb5599) 125279 125280commit 442967c90dd9d8483a56bdc9237c49e33d619126 125281Author: Alan Coopersmith <alan.coopersmith@sun.com> 125282Date: Tue Jul 28 15:02:37 2009 -0700 125283 125284 Remove hardcoded gcc -Wall option from configure.ac 125285 125286 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 125287 125288commit 52e9ef5664a697a31102e8761eaa03cff01d14d8 125289Author: Tiago Vignatti <tiago.vignatti@nokia.com> 125290Date: Tue Jul 28 19:04:59 2009 +0300 125291 125292 xfree86: remove some RAC junk 125293 125294 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 125295 125296commit d57361bb9ac08bec470f76ca8ca602d60c339502 125297Author: Tiago Vignatti <tiago.vignatti@nokia.com> 125298Date: Tue Jul 28 18:55:32 2009 +0300 125299 125300 doc: remove outdated PCI/RAC/Domain notes 125301 125302 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 125303 125304commit 7b3d05ebd5a55f88098f4a763d1fa7ca110bb780 125305Author: Tiago Vignatti <tiago.vignatti@nokia.com> 125306Date: Tue Jul 28 18:45:16 2009 +0300 125307 125308 xfree86: remove pci debug macros 125309 125310 there's no effect when their are enabled. 125311 125312 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 125313 125314commit 6089d9cfde533c6a7aec7dfc08aec9c8f2f18e27 125315Author: Tiago Vignatti <tiago.vignatti@nokia.com> 125316Date: Tue Jul 28 18:29:35 2009 +0300 125317 125318 xfree86: remove RAC/resource doc 125319 125320 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 125321 125322commit c553161e17ebc577ecb91ec6c81f0bdd0ae85e13 125323Author: Tiago Vignatti <tiago.vignatti@nokia.com> 125324Date: Fri Jul 24 13:47:31 2009 +0300 125325 125326 xfree86: delete devices probe code (-probe and -probeonly options) 125327 125328 Inside a windowing system, it's not the place to probe for devices. Goodbye 125329 -probe and -probeonly. 125330 125331 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 125332 125333commit c09779f95a9772c0556760222dfc570dbaf8a28e 125334Author: Tiago Vignatti <tiago.vignatti@nokia.com> 125335Date: Fri Jul 24 13:37:31 2009 +0300 125336 125337 xfree86: delete stupid video driver dump (-modalias option) 125338 125339 Such stupid and ugly way to dump PCI information! Oh boy... Anyway, this 125340 doesn't belong to the X server at all. Go away! 125341 125342 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 125343 125344commit 0524420622df6cbdb3872917906f7b2a6ec02958 125345Author: Colin Harrison <colin.harrison@virgin.net> 125346Date: Mon Jul 27 15:49:41 2009 +0100 125347 125348 Cygwin/X: use GWLP_WNDPROC, GWLP_USERDATA with Get/SetWindowLongPtr 125349 125350 Missed from commit b3751454cbe02ee952bab213e8c3684d429c41b3 125351 125352 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 125353 125354commit b1c3dc6ae226db178420e3b5f297b94afc87c94c 125355Author: Rémi Cardona <remi@gentoo.org> 125356Date: Mon Jul 27 12:07:51 2009 +0200 125357 125358 config: add HAL error checks 125359 125360 This patch simplifies error handling in the HAL code and fixes a 125361 segfault if libhal_find_device_by_capability() failed. 125362 125363 Fixes http://bugs.gentoo.org/278760 125364 125365 Based on a patch by Martin von Gagern <Martin.vGagern@gmx.net> 125366 125367 Signed-off-by: Rémi Cardona <remi@gentoo.org> 125368 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 125369 125370commit 8898203b0d0e9fa03453b2bcd9b88843cccc3230 125371Author: Dave Airlie <airlied@redhat.com> 125372Date: Tue Jul 28 19:35:37 2009 +1000 125373 125374 sbus: fixup for rac removal 125375 125376commit 4b42448a2388d40f257774fbffdccaea87bd0347 125377Author: Dave Airlie <airlied@redhat.com> 125378Date: Tue Jul 28 14:47:42 2009 +1000 125379 125380 xserver: remove RAC/resource handling code. 125381 125382 This changes the ABI, but since the video ABI is at 6 already 125383 it should be fine. 125384 125385 driver changes are in the pipeline after this. 125386 125387 Signed-off-by: Dave Airlie <airlied@redhat.com> 125388 125389commit 0a168401c401727e49a12cae43a6a387b1f2928d 125390Author: Dave Airlie <airlied@redhat.com> 125391Date: Tue Jul 28 14:19:24 2009 +1000 125392 125393 pci: add support for pci is boot vga call. 125394 125395 This allows us to ask the kernel for the boot VGA device 125396 instead of nasty guessing. 125397 125398commit b3751454cbe02ee952bab213e8c3684d429c41b3 125399Author: Colin Harrison <colin.harrison@virgin.net> 125400Date: Tue Jul 21 16:07:56 2009 +0100 125401 125402 Cygwin/X: Update Get/SetWindowLong() to Get/SetWindowLongPtr() everywhere 125403 125404 Get/SetWindowLong() is superseded by Get/SetWindowLongPtr(), so change to 125405 using that everywhere it remains 125406 125407 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 125408 125409commit c81595e23b48368dafc054f023c1dd16bbad9494 125410Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125411Date: Sun Jul 26 02:32:17 2009 -0700 125412 125413 XQuartz: Overhaul setting up visuals 125414 125415 The main change is cleanup of the visualConfigs and setting up alpha correctly there to match the visuals being added earlier (so the default visual has a corresponding GLX visual) 125416 (cherry picked from commit 7351db5c8746be30a4047469ee9b50bc19e62a89) 125417 125418commit 7f28c555b80dda2ed4f518efdb79733647dfea80 125419Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125420Date: Sat Jul 25 20:19:05 2009 -0700 125421 125422 XQuartz: Use CopyKeyClass to copy the keymap to the virtual core keyboard. 125423 (cherry picked from commit 9a801d1716f005c30be076fcc9ac8dbb3e74d989) 125424 125425commit 1031ac3a7306e7a82169c79c64607696c826c47f 125426Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125427Date: Sat Jul 25 20:18:38 2009 -0700 125428 125429 Revert "XQuartz: Copy the keyboard map to the core keyboard" 125430 125431 This reverts commit 795de791cf18c658421d701af645718493eac51e. 125432 125433commit 48703083a1cf308306f254691d7c2ecda09b3812 125434Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125435Date: Sat Jul 25 19:34:17 2009 -0700 125436 125437 XQuartz: Use pDev=NULL for DarwinSendDDXEvent 125438 125439 These events aren't really related to physical input devices anyways, so it doesn't make sense to use the pointer. 125440 (cherry picked from commit bfe0b9cfa7af4a48dba849cab1eb152c409b4e08) 125441 125442commit b8e0f740829d0c81324aeb59222fc8e3d22493cc 125443Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125444Date: Sat Jul 25 20:17:25 2009 -0700 125445 125446 mieq: Protect from pDev=NULL in mieqEnqueue and mieqProcessInputEvents 125447 (cherry picked from commit bf60ffb49700da367f7f88983b042a88fef7219b) 125448 125449commit dab8de036808e363a8ec9be826cff7fbcd92b953 125450Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125451Date: Sat Jul 25 15:21:12 2009 -0700 125452 125453 XQuartz: Bump the reported version to X11R7.4 125454 (cherry picked from commit c83f701aa75c75433b8745f5d79bca3a7516df91) 125455 125456commit 9e74bb97ab6cb83dd99ec36bd3842197263b8a8c 125457Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125458Date: Sat Jul 25 10:39:26 2009 -0700 125459 125460 XQuartz: xpr: Added missing include for RootlessHideAllWindows 125461 (cherry picked from commit 75e104730810acbb11bf8503bac24ea25243b2ac) 125462 125463commit 09b024db636f5a306c9cec3806fffde9acd3c408 125464Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125465Date: Sat Jul 25 15:22:01 2009 -0700 125466 125467 XQuartz: Use the master device in DarwinSendDDXEvent to avoid duplicate events. 125468 (cherry picked from commit 6fa62192af937aa9656f64b516050bc099231c7a) 125469 125470commit 795de791cf18c658421d701af645718493eac51e 125471Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125472Date: Sat Jul 25 15:25:06 2009 -0700 125473 125474 XQuartz: Copy the keyboard map to the core keyboard 125475 125476 This still doesn't handle the modifier map... gotta figure out what to do now that SwitchCoreKeyboard is gone 125477 (cherry picked from commit 427e1aab41dabb54354bfd30f2baae98ac8202c0) 125478 125479commit af8047f7816b39be7015e8cbbe085c57bb354592 125480Author: Aaron Plattner <aplattner@nvidia.com> 125481Date: Fri Jul 17 15:10:46 2009 -0700 125482 125483 Damage: Add devPrivates to DamageRec 125484 125485 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 125486 125487commit c80ed4945ef6dd50473fab871211b71b3d20dc3c 125488Author: Aaron Plattner <aplattner@nvidia.com> 125489Date: Fri Jul 17 15:10:32 2009 -0700 125490 125491 Damage: Add wrappable hooks for damage create, destroy, register, & unregister. 125492 125493 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 125494 125495commit fb46474feb69d52d3dfdd61452ced5cfc38f6651 125496Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125497Date: Fri Jul 24 09:16:37 2009 -0700 125498 125499 XQuartz: Rever the "Set can_quit to true during a Sparkle-initiated relaunch." change 125500 125501 This was in place to work around the issue that was correctly solved with the 125502 previous commit (changing the Windows menu behavior). Reverting this change 125503 no longer causes crashing, so it's safe to show the dialog now. 125504 (cherry picked from commit 023cef31bbb2ab80a241098d82fcfd35ada75fc0) 125505 125506commit 0d73893a5795d5c1d28fd0287f2f1965f2ff1bda 125507Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125508Date: Thu Jul 23 20:03:29 2009 -0700 125509 125510 XQuartz: Change handling of Windows menu to workaround a bug triggered by mixing Sparkle and X11 windows 125511 125512 <rdar://problem/7088335> NSApplication releases the separator in the Windows menu even though it's an IBOutlet 125513 (cherry picked from commit 27ac5135267be9cb221329ae68461117dd43a4bf) 125514 125515commit 171409cecbd848e5fec6334fc61a20e882e80fa9 125516Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125517Date: Tue Jul 21 00:22:12 2009 -0700 125518 125519 XQuartz: Set can_quit to true during a Sparkle-initiated relaunch. 125520 (cherry picked from commit b2e9a77111d4572402d8ca95e3368db97ba7d598) 125521 125522commit 7ff84d350b44fa40669c1d0d48a715a0bf056ece 125523Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125524Date: Mon Jul 20 22:38:25 2009 -0700 125525 125526 XQuartz: Added a "Check for X11 Updates..." menu item. 125527 (cherry picked from commit 305144bfa4b59791123c44e869fab93a084792d6) 125528 125529commit 1e1dbd1e462f571dad2f9684fcf4cd8ae17eedd5 125530Author: Tiago Vignatti <tiago.vignatti@nokia.com> 125531Date: Mon Jun 29 11:46:17 2009 +0300 125532 125533 xfree86: "Staticize" functions in xf86AutoConfig.c 125534 125535 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 125536 125537commit 7c6b5458de9bc7f6cd972a36b56888aaa3d201ee 125538Author: Aaron Plattner <aplattner@nvidia.com> 125539Date: Thu Jul 23 15:16:44 2009 -0700 125540 125541 Fix dist. 125542 125543 * Makefile.am: Include the test/ subdirectory in the dist tarball. 125544 125545 * include/Makefile.am: Move events.h to sdk_HEADERS and eventconvert.h to 125546 EXTRA_DIST so they're included in the tarball. events.h shouldn't be 125547 included in the SDK either, but for now it's needed by input.h. 125548 125549commit 20daa145c437c3ba67970146f6182849f87a1b43 125550Author: Peter Hutterer <peter.hutterer@who-t.net> 125551Date: Thu Jul 23 08:33:00 2009 +1000 125552 125553 mi: fix build error caused by missing xtest.h 125554 125555 xtest.h was renamed to xtestconst.h in xextproto. 125556 Requires xextproto 7.0.99.3 125557 125558 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125559 125560commit 587c3a2d1961834558193e8e14e8e381a077a253 125561Author: Aaron Plattner <aplattner@nvidia.com> 125562Date: Wed Jul 22 10:55:46 2009 -0700 125563 125564 Bug #22804: Reject out of bounds XGetImage requests 125565 125566 The XGetImage man page states: 125567 125568 If the drawable is a window, the window must be viewable, and it must be the 125569 case that if there were no inferiors or overlapping windows, the specified 125570 rectangle of the window would be fully visible on the screen and wholly 125571 contained within the outside edges of the window, or a BadMatch error 125572 results. Note that the borders of the window can be included and read with 125573 this request. 125574 125575 However, the server was only checking the requested region against the screen 125576 bounds, allowing XGetImage requests to read pixels outside the bounds of a 125577 window's ancestors. Normally, this would just read other pixels from the 125578 screen, but if one of the ancestor windows is redirected, the window's backing 125579 pixmap may be smaller than the window itself. 125580 125581 This change checks the region against the window's bounding drawable, which is 125582 either the screen pixmap, a redirected window's backing pixmap, or the root 125583 window for servers that don't support GetWindowPixmap. 125584 125585 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 125586 Reviewed-by: Keith Packard <keithp@keithp.com> 125587 125588commit ecd618957ebf01cb4137f98efec3faed35f8a9f5 125589Author: Peter Hutterer <peter.hutterer@who-t.net> 125590Date: Wed Jul 22 15:19:34 2009 +1000 125591 125592 dix: fix null-pointer dereference on activating enter/focus grabs. 125593 125594 EventToXI2 returns a NULL event for enter/focus events since these events 125595 aren't yet wrapped into internal events. This is a quickfix only, it works 125596 thanks to the alignment of internal and XI2 event types. 125597 125598 Eventually, enter/leave events should be wrapped into internal events. 125599 125600 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125601 125602commit 48585bd1e3e98db0f3df1ecc68022510216e00cc 125603Author: Peter Hutterer <peter.hutterer@who-t.net> 125604Date: Wed Jul 22 12:09:24 2009 +1000 125605 125606 dix: use the event mask of the grab for TryClientEvents. 125607 125608 A client that grabs for button/key events may not have the 125609 ButtonPress/KeyPress mask set and should not receive an event in that case. 125610 125611 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125612 125613commit 4dc91b3e54503a1be555bae5b18f3e52f58be307 125614Author: Pierre Willenbrock <pierre@pirsoft.de> 125615Date: Tue Jul 21 17:21:28 2009 +0200 125616 125617 Check if new space was actually allocated before freeing. 125618 125619 There will be no new space allocated, if mode != PropModeReplace and 125620 len == 0, or if mode is not one of the handled modes. 125621 This fixes freeing data that is still in use, leading to double frees and 125622 other memory corruption. 125623 125624 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125625 125626commit 9a1bfa5664a80f03cedeb89b9f8a86115a08e7af 125627Author: Peter Hutterer <peter.hutterer@who-t.net> 125628Date: Mon Jul 20 16:39:16 2009 +1000 125629 125630 input: remove XI2 keysym grabs, use keycode grabs instead. 125631 125632 Keysym grabs are tricky in the details, keycode grabs are known to work. So 125633 for now, provide keycode grabs only. 125634 125635 Requires inputproto 1.9.99.15. 125636 125637 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125638 125639commit afc3e3b5955ea4a49308399820cc4c499f4312da 125640Author: Jon TURNEY <jon.turney@dronecode.org.uk> 125641Date: Tue Jul 21 16:14:21 2009 +0100 125642 125643 Cygwin/X: winInitMultiWindowClass() should be static 125644 125645 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 125646 125647commit 9dc8106ba095474fea1474622b0f0162c8fd5aa4 125648Author: Michel Dänzer <daenzer@vmware.com> 125649Date: Tue Jul 21 14:34:47 2009 +0200 125650 125651 randr12: Add compatibility for XF86VidMode gamma ramps. 125652 125653 Fixes screensaver fadeout effects. 125654 125655 Also make all RandR 1.2 compatibility code for XF86VidMode operate only on the 125656 CRTC associated with the compatibility output, not all CRTCs at once. 125657 125658commit 268e227ba06c027f5c56b1aaee5dcc6a2034403f 125659Author: Michel Dänzer <daenzer@vmware.com> 125660Date: Tue Jul 21 14:34:13 2009 +0200 125661 125662 EXA: Make Prepare/FinishAccess tracking resilient to repeated / nested calls. 125663 125664 Use reference counting and do nothing unless the reference count transitions 125665 to/from 0. 125666 125667 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=22822 . 125668 125669 As a bonus, this avoids calling the driver Prepare/FinishAccess hooks more than 125670 once per pixmap and operation. 125671 125672 Also update the Doxygen documentation for the PrepareAccess driver hook to 125673 better match current reality. 125674 125675commit de7a14ca92f99ff03c8ad204aab5be8203c86a72 125676Author: Michel Dänzer <daenzer@vmware.com> 125677Date: Tue Jul 21 12:55:27 2009 +0200 125678 125679 EXA: Fix up some issues introduced by 00fe4a297744c81b40f0243fb56ad848a9be6a2b. 125680 125681 * Check all pixmaps involved for damage records, fixes visual corruption due to 125682 the screen pixmap never having one. 125683 * Fix an array size and remove a now superfluous assignment. 125684 125685commit be4dd35ffecbf49aff13aa9f604a44c9b665ae92 125686Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125687Date: Sat Jul 18 15:51:29 2009 -0700 125688 125689 XQuartz: Initial support for automatic updates through Sparkle 125690 (cherry picked from commit c45f1be36426bceeef9af67c26351114f14f5277) 125691 125692commit 2415e2dce918efd49d5e6d71f705255a234a866b 125693Author: Michel Dänzer <daenzer@vmware.com> 125694Date: Mon Jul 20 02:08:31 2009 +0200 125695 125696 EXA: Bail earlier from exaDoPutImage if the driver has no UploadToScreen hook. 125697 125698 Also remove dead code associated with access_prepared local variable. 125699 125700commit 00fe4a297744c81b40f0243fb56ad848a9be6a2b 125701Author: Michel Dänzer <daenzer@vmware.com> 125702Date: Mon Jul 20 02:04:40 2009 +0200 125703 125704 EXA: Completely eliminate exaDoMigration calls for drivers that manage pixmaps. 125705 125706commit 7b9915b11ed9eedd0698b4563328504d686ac4ec 125707Author: Gaetan Nadon <memsize@videotron.ca> 125708Date: Sat Jul 18 20:52:20 2009 -0400 125709 125710 dix: xserver "make dist" fails due to eventconvert.h (#27825) 125711 125712 This header file is not in the /dix dir, but in the /include dir. 125713 The makefile aborted and the xserver distribution files were not created. 125714 The fix is to remove this header file from the libdix_la_SOURCES 125715 in the dix/Makefile.am. 125716 125717 X.Org Bug 27825 <http://bugs.freedesktop.org/show_bug.cgi?id=27825> 125718 125719 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125720 125721commit a98acb290737ac6b0776f2ea8128a4613c0f778b 125722Author: Adam Jackson <ajax@redhat.com> 125723Date: Wed Jul 15 13:56:19 2009 -0400 125724 125725 s/xf86EnableOutputs/xf86CollectEnabledOutputs/ for clarity 125726 125727commit 8fab7f72f2cc4ac5ca415c95ccbd05d3084f10c4 125728Author: Adam Jackson <ajax@redhat.com> 125729Date: Wed Jul 15 14:09:08 2009 -0400 125730 125731 randr: Add Option "Primary" to Monitor sections 125732 125733commit 053bb92145045e06f8d72c3bddb75806432faa1c 125734Author: Adam Jackson <ajax@redhat.com> 125735Date: Thu Jul 16 14:29:05 2009 -0400 125736 125737 xfree86: Remove xf86SetPriority 125738 125739 GetClocks is pretty doomed to begin with, this really isn't going to 125740 make it worse. 125741 125742commit 3637945a45d55385335833197b18703379892884 125743Author: Adam Jackson <ajax@redhat.com> 125744Date: Thu Jul 16 14:10:45 2009 -0400 125745 125746 loader: Remove a useless check. 125747 125748commit ef2bb08c76f0d1eed0c8b3e4937648f0b6ae01e5 125749Author: Adam Jackson <ajax@redhat.com> 125750Date: Thu Jul 16 13:53:00 2009 -0400 125751 125752 loader: remove dead LoaderCheckUnresolved 125753 125754commit 06a2784232b5d0e2440ed365d272abdfa1f5d35f 125755Author: Adam Jackson <ajax@redhat.com> 125756Date: Thu Jul 16 13:40:48 2009 -0400 125757 125758 loader: Remove useless TestFree() macro 125759 125760commit 087745b809b9a834751e5c719b9d87967194369b 125761Author: Adam Jackson <ajax@redhat.com> 125762Date: Thu Jul 16 10:51:53 2009 -0400 125763 125764 fbdevhw: Remove pointless OS check, this never gets built on non-Linux 125765 125766commit 892c93553aa0b03aeb3f4c27d952cc4bb7120aff 125767Author: Adam Jackson <ajax@redhat.com> 125768Date: Thu Jul 16 10:51:27 2009 -0400 125769 125770 fbdevhw: Remove a #if 0 125771 125772commit e5712f2926abf33b146a4ccfcf6e89c914dad37b 125773Author: Adam Jackson <ajax@redhat.com> 125774Date: Thu Jul 16 10:49:22 2009 -0400 125775 125776 fbdevhw: simplify some #if 1 125777 125778commit 8868bb4131be25340bf65ec61b998d353965bab6 125779Author: Adam Jackson <ajax@redhat.com> 125780Date: Thu Jul 16 10:44:24 2009 -0400 125781 125782 randr: Un-duplicate the reduced blanking check. 125783 125784commit 453ee39bc6a6b46e6cda11512cfcd431ba32a4b7 125785Author: Adam Jackson <ajax@redhat.com> 125786Date: Thu Jul 16 10:39:37 2009 -0400 125787 125788 xfree86: Fix some misleading comments 125789 125790commit 2c57a7aa07fdf52be315ecb498341776268c1a10 125791Author: Adam Jackson <ajax@redhat.com> 125792Date: Thu Jul 16 10:32:53 2009 -0400 125793 125794 xfree86: Remove some #if 0 125795 125796 This code all lives in xf86Modes.c now anyway 125797 125798commit 1a8c89683ff40cdd1d33da89ab47e7bd240b9bb7 125799Author: Adam Jackson <ajax@redhat.com> 125800Date: Thu Jul 16 10:15:41 2009 -0400 125801 125802 xfree86: Remove TargetRefresh option 125803 125804 This was a vestige from the days before we'd make the mode list from the 125805 EDID data, and from CRT technology when you could reasonably assume that 125806 higher refresh rates were better. Also it did not function as advertised, 125807 acting as a high-pass filter instead of a band-pass. 125808 125809commit b4ee3bf700e04ae1b3c4ec021373424ed6e5338f 125810Author: Adam Jackson <ajax@redhat.com> 125811Date: Fri Jul 17 14:57:50 2009 -0400 125812 125813 dbe: Fix indentation 125814 125815commit 0bb9a7e1650180a24246d14493a8168487cf8914 125816Author: Adam Jackson <ajax@redhat.com> 125817Date: Tue Jun 9 11:49:41 2009 -0400 125818 125819 displayid: Implement mode decoding. 125820 125821commit 2f1a9c5baa367818bf017bdb72f20a2f6fa7ac21 125822Author: Adam Jackson <ajax@redhat.com> 125823Date: Tue Jun 9 10:13:47 2009 -0400 125824 125825 ddc: s/xf86DDCMonitorSet/xf86EdidMonitorSet/ 125826 125827 Since we need a second path for DisplayID. 125828 125829commit d0cb4f5a91932e901d10cac5f2a4ba12bb8a0e6f 125830Author: Adam Jackson <ajax@redhat.com> 125831Date: Tue Jun 9 10:10:18 2009 -0400 125832 125833 ddc: Refactor root window property code 125834 125835commit 14103b781201bc36896cbe9112a2e0d991fb785d 125836Author: Adam Jackson <ajax@redhat.com> 125837Date: Tue Jun 9 10:05:01 2009 -0400 125838 125839 ddc: Don't try to publish a root window property for DisplayID 125840 125841commit c302a5ff49146bff24df196fc36ed38745d42911 125842Author: Adam Jackson <ajax@redhat.com> 125843Date: Mon Jun 8 17:42:10 2009 -0400 125844 125845 ddc: Skeleton for xf86DoDisplayID() 125846 125847commit 0b36f68efb1171fcdbe53e93064394f5609b7fb5 125848Author: Adam Jackson <ajax@redhat.com> 125849Date: Mon Jun 8 15:36:15 2009 -0400 125850 125851 ddc: mv xf86DDC.c ddc.c 125852 125853commit 8eb82168fc5c7ea958a4f63676738510647dd203 125854Author: Adam Jackson <ajax@redhat.com> 125855Date: Mon Jun 8 15:34:42 2009 -0400 125856 125857 ddc: Give DisplayID a place to hang its hat 125858 125859commit ace0fe09aee48d57cd0079260cd8d20d041e8eb6 125860Author: Adam Jackson <ajax@redhat.com> 125861Date: Mon Jun 8 14:45:47 2009 -0400 125862 125863 ddc: Yet more code motion 125864 125865commit 3a350688bbe1257feaf1502b4009f5f701d2640e 125866Author: Adam Jackson <ajax@redhat.com> 125867Date: Mon Jun 8 14:37:38 2009 -0400 125868 125869 ddc: Code motion to eliminate forward decls 125870 125871commit 8797831f82637b0e65a08e1262d1ec57c075cc12 125872Author: Adam Jackson <ajax@redhat.com> 125873Date: Mon Jun 8 14:30:53 2009 -0400 125874 125875 ddc: Update a comment. 125876 125877commit bb6fa39eb3aa044ffc18632288fd59909ff06344 125878Author: Adam Jackson <ajax@redhat.com> 125879Date: Thu Jul 16 17:43:31 2009 -0400 125880 125881 dbe: Adapt to new headers 125882 125883commit 9965174f4f09a7de00a9569607e96226208bc2ce 125884Author: Pierre Willenbrock <pierre@pirsoft.de> 125885Date: Thu Jul 16 12:12:48 2009 +0200 125886 125887 Check dev->u.master if there is a custom event handler, too 125888 125889 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125890 125891commit c806162c133603a99d9cd844bb04485bb663707d 125892Author: Pierre Willenbrock <pierre@pirsoft.de> 125893Date: Tue Jul 14 13:42:19 2009 +0200 125894 125895 Always update u.lastSlave 125896 125897 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125898 125899commit 4db8950b1dfc3150ee2e9f1f975e9ecb4eabb1f2 125900Author: Peter Hutterer <peter.hutterer@who-t.net> 125901Date: Fri Jul 17 10:40:59 2009 +1000 125902 125903 Xext: include securproto.h instead of securstr.h 125904 125905 Reported-by: Byeong-ryeol Kim 125906 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125907 125908commit b5f6da1e9b37f52484948185dcf17950657cd65b 125909Author: Oliver McFadden <oliver.mcfadden@nokia.com> 125910Date: Fri Jul 17 06:34:00 2009 +0300 125911 125912 Revert "Coverity Prevent: RESOURCE_LEAK in AccelSetProfileProperty:" 125913 125914 This reverts commit 7333dc2969f60af0abcfb28e7182a5fff9918223. 125915 125916 False positive. 125917 125918commit f517fca31d8c341ad36f3de4863adb0bc5206176 125919Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125920Date: Thu Jul 16 17:28:56 2009 -0700 125921 125922 XQuartz: Localization updates 125923 (cherry picked from commit 4b797fc1edf2bd963410a3133e3d2182ccfda4c3) 125924 125925commit c8a35165147a9b2f747b25da80d0fdb4279a8331 125926Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125927Date: Thu Jul 16 17:20:16 2009 -0700 125928 125929 XQuartz: Cleanup getGlCapabilities to avoid hardcoding the number of displays 125930 (cherry picked from commit 12f7365f1f58d648217b16f09c2152fa47dcd7a1) 125931 125932commit e7bc9ff816c1848c700a376908a1411f1e20d29f 125933Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125934Date: Thu Jul 16 16:59:47 2009 -0700 125935 125936 XQuartz: Allow more than 3 OSX displays 125937 (cherry picked from commit 45045eb3968069a8d74ce9188890a5537dc60fec) 125938 125939commit 7333dc2969f60af0abcfb28e7182a5fff9918223 125940Author: Oliver McFadden <oliver.mcfadden@nokia.com> 125941Date: Thu Jul 16 18:57:49 2009 +0300 125942 125943 Coverity Prevent: RESOURCE_LEAK in AccelSetProfileProperty: 125944 125945 Event alloc_arg: Called allocation function "XIPropToInt" on "ptr" [details] 125946 167 rc = XIPropToInt(val, &nelem, &ptr); 125947 125948 Event leaked_storage: Variable "ptr" goes out of scope 125949 125950commit 66eabbebaf1b40fed7670b7c05337ed6226dd81e 125951Author: Oliver McFadden <oliver.mcfadden@nokia.com> 125952Date: Thu Jul 16 18:50:22 2009 +0300 125953 125954 Coverity Prevent: RESOURCE_LEAK in xf86CrtcSetInitialGamma 125955 125956commit 464e8ad733fa6afee028607d6e7d4663b4c273cc 125957Author: Oliver McFadden <oliver.mcfadden@nokia.com> 125958Date: Thu Jul 16 17:51:08 2009 +0300 125959 125960 Coverity Prevent: NEGATIVE_RETURNS in fbdev_open_pci: 125961 125962 Event var_tested_neg: Variable "fd" tested NEGATIVE 125963 At conditional (1): "fd != -1" taking false path 125964 335 if (fd != -1) { 125965 125966 Event negative_returns: Tracked variable "fd" was passed to a negative sink. 125967 347 close(fd); 125968 125969commit 53ae6b63387e04324b23b6f8311cc22a154c1fb8 125970Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 125971Date: Wed Jul 15 23:18:35 2009 -0700 125972 125973 XQuartz: Cleanup keymap locking, fix a possible synchro bug 125974 (cherry picked from commit 33e7437a4984ee7c1b04b87d23dee7d4739d5f12) 125975 125976commit 6a90c7b93724a2d26eae70b5806ca06c91e7df4c 125977Author: Peter Hutterer <peter.hutterer@who-t.net> 125978Date: Thu Jul 16 09:28:42 2009 +1000 125979 125980 xkb: cosmetic fix, use TRUE instead of True. 125981 125982 Rest of InitKeyboardDeviceStruct uses TRUE and FALSE. 125983 125984 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125985 125986commit 52088d3c2d53a91b2690f9cc402f839d77769e37 125987Author: Éric Piel <eric.piel@trempin-utc.net> 125988Date: Sun Jun 21 18:11:28 2009 +0200 125989 125990 xserver: remove unused code in clipValuators 125991 125992 The axes variables was never used, remove it. 125993 125994 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 125995 125996commit d3f6b43a240eb763025b3cbf546cb7ae502c94fa 125997Author: Peter Hutterer <peter.hutterer@who-t.net> 125998Date: Wed Jul 15 16:51:05 2009 +1000 125999 126000 Update to xextproto 7.0.99.1. 126001 126002 xextproto had Xlib client headers moved into libXext. 126003 Protocol header files are named fooproto.h, header files with constants 126004 foo.h or fooconst.h where foo.h was already in use for client-side headers. 126005 126006commit 693babbf12cc7969c3e211c4037c7af0d41c13e9 126007Author: Peter Hutterer <peter.hutterer@who-t.net> 126008Date: Wed Jul 15 12:30:02 2009 +1000 126009 126010 xkb: Remove XKMformat.h include from xkbsrv.h into the files that need it. 126011 126012 xkbsrv.h is used by drivers, they don't need the XKM format and shouldn't 126013 require it. 126014 126015 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126016 126017commit 35ff5cd26eb7564fefebf238e30b8d43cbb4dc25 126018Author: Peter Hutterer <peter.hutterer@who-t.net> 126019Date: Wed Jul 15 10:29:46 2009 +1000 126020 126021 dix: fix wrong raw valuator copy 126022 126023 internal events keep valuator data at the index for the valuator, not like 126024 the wire events that start with first_valuator. 126025 126026 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126027 126028commit 975bf60a82f863e3bdc36fc3f8201b48a4d1e6f7 126029Author: Peter Hutterer <peter.hutterer@who-t.net> 126030Date: Wed Jul 15 10:29:19 2009 +1000 126031 126032 dix: use sizeof(FP3232) instead of 2 * sizeof(int32_t). 126033 126034 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126035 126036commit d040af7fa3c7314917414a28d723bdda3c4289c3 126037Author: Peter Hutterer <peter.hutterer@who-t.net> 126038Date: Mon Jul 13 15:09:38 2009 +1000 126039 126040 Update to type-specific raw events - require inputproto 1.9.99.14. 126041 126042 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126043 126044commit 81b64f668541217c1c9518ac4a7e3fdf0a6e4002 126045Author: Fredrik Höglund <fredrik@kde.org> 126046Date: Wed Jul 15 00:38:01 2009 +0200 126047 126048 Require renderproto >= 0.11 and pixman-1 >= 0.15.14 for the new blend modes. 126049 126050commit 0ce42adbf4cff9e7f049d9fc79d588ece5936177 126051Author: Fredrik Höglund <fredrik@kde.org> 126052Date: Wed Jul 15 00:37:05 2009 +0200 126053 126054 Render: Add support for the PDF blend mode operators. 126055 126056commit 686e4867302a741f3029c4105b997d0f0ac7c13c 126057Author: Tiago Vignatti <tiago.vignatti@nokia.com> 126058Date: Tue Jul 14 17:11:13 2009 +0300 126059 126060 xfree86: remove a bunch of unused pci headers 126061 126062 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 126063 126064commit cc575a3ba4a52265e410b325c2291fe900a54f33 126065Author: Peter Hutterer <peter.hutterer@who-t.net> 126066Date: Fri Jun 12 11:38:46 2009 +1000 126067 126068 test: add tests for ProcXIGetClientPointer. 126069 126070 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126071 126072commit 92b0065a19dcde50494d2528e19a4ee76723b3da 126073Author: Peter Hutterer <peter.hutterer@who-t.net> 126074Date: Fri Jun 12 11:38:00 2009 +1000 126075 126076 test: add ProcXISetClientPointer tests. 126077 126078 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126079 126080commit 019bc9d70c5b3beb69ca20b6bca3551e00415992 126081Author: Peter Hutterer <peter.hutterer@who-t.net> 126082Date: Fri Jun 12 11:37:17 2009 +1000 126083 126084 test: add ProcXIGetSelectedEvents tests. 126085 126086 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126087 126088commit 8579d39c12b065d8e82fe75af4b254c8ad4f4ab8 126089Author: Peter Hutterer <peter.hutterer@who-t.net> 126090Date: Fri Jun 12 11:36:22 2009 +1000 126091 126092 test: add ProcXISelectEvents tests. 126093 126094 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126095 126096commit 60b7a7990003a22be5cc26ebe4d8d2d0803b97de 126097Author: Peter Hutterer <peter.hutterer@who-t.net> 126098Date: Fri Jun 12 11:33:22 2009 +1000 126099 126100 test: add ProcXIQueryDevice tests. 126101 126102 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126103 126104commit 441c91be05912e2665cc1d3ee72fb70dcd32e7a2 126105Author: Peter Hutterer <peter.hutterer@who-t.net> 126106Date: Fri Jun 12 11:32:15 2009 +1000 126107 126108 test: Add ProcXIQueryVersion tests. 126109 126110 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126111 126112commit 1d635ae65b217e124b3bd5a8844e73f447eb1b79 126113Author: Peter Hutterer <peter.hutterer@who-t.net> 126114Date: Fri Jun 12 11:30:53 2009 +1000 126115 126116 test: add xi2 directory, prepare for protocol testing. 126117 126118 These two files provide a couple of common defines, functions and variables 126119 that will be used in a number of protocol tests. 126120 126121 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126122 126123commit 4b93413b2a490a74de2e27b37268332b609ac872 126124Author: Peter Hutterer <peter.hutterer@who-t.net> 126125Date: Tue Jul 14 11:20:01 2009 +1000 126126 126127 Add test subdir to base Makefile.am 126128 126129 If unittests are enabled, make will build those as well - spotting potential 126130 build errors in the tests faster. Furthermore, this allows for the tests to 126131 be run from the top-level directory. 126132 This patch removes the "run make check to build the test suite" message 126133 since that'd pop up after every build now. 126134 126135 If unittests are disabled, this change has no effect. 126136 126137 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126138 126139commit e2226f3cd7d1bef598f9657b756a171b02f1a299 126140Author: Peter Hutterer <peter.hutterer@who-t.net> 126141Date: Mon Jul 6 14:57:04 2009 +1000 126142 126143 dbe: switch to byte counting functions 126144 126145 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126146 126147commit 3e1a054423c22f0e35f25127dde1ea8263892480 126148Author: Peter Hutterer <peter.hutterer@who-t.net> 126149Date: Mon Jul 6 14:24:45 2009 +1000 126150 126151 record: switch to byte counting functions 126152 126153 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126154 126155commit 089c460058c4e6814d98d40179eb4c731f37fa80 126156Author: Peter Hutterer <peter.hutterer@who-t.net> 126157Date: Mon Jul 6 14:22:40 2009 +1000 126158 126159 xkb: switch to byte counting functions 126160 126161 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126162 126163commit 0dc11da57562a7bbb6830ab657a2b818c596a537 126164Author: Peter Hutterer <peter.hutterer@who-t.net> 126165Date: Mon Jul 6 12:41:29 2009 +1000 126166 126167 ephyr: switch to byte counting functions 126168 126169 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126170 126171commit 34bfaa9d9ecd90cfe8413bc275179fdcc193eab3 126172Author: Peter Hutterer <peter.hutterer@who-t.net> 126173Date: Mon Jul 6 12:29:26 2009 +1000 126174 126175 xfixes: switch to byte counting functions 126176 126177 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126178 126179commit e216527107fda470b92b7e526f3db22465962a43 126180Author: Peter Hutterer <peter.hutterer@who-t.net> 126181Date: Mon Jul 6 12:27:23 2009 +1000 126182 126183 render: switch to byte counting functions 126184 126185 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126186 126187commit 1d6b71b8d0dc24355e84391ba413170a03ccdbf4 126188Author: Peter Hutterer <peter.hutterer@who-t.net> 126189Date: Mon Jul 6 12:25:39 2009 +1000 126190 126191 os: switch to byte counting functions 126192 126193 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126194 126195commit 86b239ff9c4d01685c357ca2b1ef761d167e3224 126196Author: Peter Hutterer <peter.hutterer@who-t.net> 126197Date: Mon Jul 6 12:24:27 2009 +1000 126198 126199 randr: switch to byte counting functions 126200 126201 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126202 126203commit 7b9e84e320a6f6449fe7bc58a8d6a094ae37b86c 126204Author: Peter Hutterer <peter.hutterer@who-t.net> 126205Date: Mon Jul 6 12:20:21 2009 +1000 126206 126207 Xext: switch to byte counting functions 126208 126209 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126210 126211commit aa19d355125a10b1a385c8f134d68e79d3d609c7 126212Author: Peter Hutterer <peter.hutterer@who-t.net> 126213Date: Fri Jul 10 14:25:22 2009 +1000 126214 126215 xwin: switch to byte counting functions 126216 126217 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126218 126219commit 0b4e6af857bcc5513e4c19912a54656d4696e56d 126220Author: Peter Hutterer <peter.hutterer@who-t.net> 126221Date: Mon Jul 6 12:12:57 2009 +1000 126222 126223 xquartz: switch to byte counting functions 126224 126225 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126226 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 126227 126228commit ad508c93c239a5ba8381000c031e96caf2769265 126229Author: Peter Hutterer <peter.hutterer@who-t.net> 126230Date: Mon Jul 6 12:12:20 2009 +1000 126231 126232 xfree86: switch to byte counting functions 126233 126234 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126235 126236commit c20304226b3ca2f8d0a4f4866480b0d71913941c 126237Author: Peter Hutterer <peter.hutterer@who-t.net> 126238Date: Mon Jul 6 12:07:00 2009 +1000 126239 126240 glx: switch to byte counting functions 126241 126242 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126243 126244commit 87ce93c9973067255b4197d82772f83cd4ea5d27 126245Author: Peter Hutterer <peter.hutterer@who-t.net> 126246Date: Mon Jul 6 12:06:47 2009 +1000 126247 126248 dmx: switch to byte counting functions 126249 126250 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126251 126252commit dc82e11e509ecf586d77c3e7c1325d515509ce51 126253Author: Peter Hutterer <peter.hutterer@who-t.net> 126254Date: Mon Jul 6 12:03:34 2009 +1000 126255 126256 exa: switch to byte counting functions. 126257 126258 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126259 Acked-by: Michel Dänzer <michel@daenzer.net> 126260 126261commit 2d35ea8d957a955e1200ba2b14424bddfe1f4148 126262Author: Peter Hutterer <peter.hutterer@who-t.net> 126263Date: Fri Jul 3 16:50:03 2009 +1000 126264 126265 dix: switch to byte-counting functions. 126266 126267 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126268 126269commit 7dd415aa6a3959f15276741db168ba264948ecfe 126270Author: Peter Hutterer <peter.hutterer@who-t.net> 126271Date: Fri Jul 3 13:57:14 2009 +1000 126272 126273 Xi: use byte-counting macros instead of manual calculation. 126274 126275 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126276 126277commit 912402fd71144bcee255141efe12a78abad39240 126278Author: Peter Hutterer <peter.hutterer@who-t.net> 126279Date: Mon Jun 29 13:09:57 2009 +1000 126280 126281 include: introduce byte counting functions. 126282 126283 This patch adds the following three functions: 126284 bits_to_bytes(bits) - the number of bytes needed to hold 'bits' 126285 bytes_to_int32(bytes) - the number of 4-byte units to hold 'bytes' 126286 pad_to_int32(bytes) - the closest multiple of 4 equal to or larger than 126287 'bytes'. 126288 126289 All three operations are common in protocol processing and currently the 126290 server has ((foo + 7)/8 + 3)/4 operations all over the place. A common set 126291 of functions reduce the error rate of these (albeit simple) calculations and 126292 improve readability of the code. 126293 126294 The functions do not check for overflow. 126295 126296 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126297 126298commit 2c535b6f13ffbf2c4ac59834dae39bb8e172c003 126299Author: Peter Hutterer <peter.hutterer@who-t.net> 126300Date: Tue Jul 14 08:58:23 2009 +1000 126301 126302 dix: don't send presence events for attaching/detaching slave devices. 126303 126304 The code that didn't list attached slave devices for XI1 clients doesn't 126305 exist anymore, so there's no need for these presence events. 126306 126307 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126308 126309commit 0c0ef42292f4c910c73b308cd75d77637312da53 126310Author: Peter Hutterer <peter.hutterer@who-t.net> 126311Date: Sun Jul 12 21:43:06 2009 +1000 126312 126313 dix: Remove temporary detachment of slave devices. 126314 126315 Previously, an active grab on an attached slave device would send the device 126316 floating for the duration of the grab. This breaks existing XI applications 126317 (e.g. the GIMP) since they grab all devices automatically - resulting in the 126318 loss of control over the VCP. 126319 126320 The behaviour of extended input devices during a grab in relation to the 126321 core pointer is not specified in the XI protocol specification. 126322 The removal of the temporary detachment restores the behaviour of extended 126323 input devices as present in currently released servers - even if a device is 126324 grabbed, an event from this device will result in an event from the core 126325 pointer. 126326 126327 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126328 126329commit 9f1570c8f4f549cdd2fbae1234011290fcc73e18 126330Author: Peter Hutterer <peter.hutterer@who-t.net> 126331Date: Fri Jun 26 09:12:47 2009 +1000 126332 126333 input: include effective modifiers in device events. 126334 126335 Require inputproto 1.9.99.13 126336 126337 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126338 126339commit 78aedbe609e309446e71835ed15e4557d3ccf723 126340Author: Alan Coopersmith <alan.coopersmith@sun.com> 126341Date: Fri Jul 10 17:35:57 2009 -0700 126342 126343 Fix int10 module build after typo in commit 2638e9899 126344 126345 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 126346 126347commit 35758544813f156eaac28844e693b2a28f6de316 126348Author: Michel Dänzer <daenzer@vmware.com> 126349Date: Fri Jul 10 20:06:02 2009 +0200 126350 126351 EXA: Only pass CT_YXBANDED to RECTS_TO_REGION() if that is really true. 126352 126353 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=22642 . 126354 126355commit 2638e9899e57d9b25a2cade3247083a6e101dee2 126356Author: Tiago Vignatti <tiago.vignatti@nokia.com> 126357Date: Fri Jul 10 16:09:08 2009 +0300 126358 126359 configure: introduce --{enable,disable}-int10-module 126360 126361 Not all chipsets need to rely on the int10 scheme to do its daily work. 126362 126363 Well, the ideal would be to remove all int10 module from the server. I'll try 126364 to provide some patches "soon" for this. Something like: 126365 http://cgit.freedesktop.org/~vignatti/libx86/ 126366 126367 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 126368 Acked-by: Oliver McFadden <oliver.mcfadden@nokia.com> 126369 126370commit 42d95a38535f382bfb2ddfaef93839ab9bbf584f 126371Author: Tiago Vignatti <tiago.vignatti@nokia.com> 126372Date: Fri Jul 10 15:54:07 2009 +0300 126373 126374 configure: introduce --{enable,disable}-vbe 126375 126376 Not all drivers need this kind of access as well. 126377 126378 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 126379 Acked-by: Oliver McFadden <oliver.mcfadden@nokia.com> 126380 126381commit 53d64930513fecaa417bb5a922966b45c9ff8679 126382Author: Tiago Vignatti <tiago.vignatti@nokia.com> 126383Date: Fri Jul 10 15:38:51 2009 +0300 126384 126385 configure: introduce --{enable,disable}-vgahw 126386 126387 Not all drivers need this kind of access. 126388 126389 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 126390 Acked-by: Oliver McFadden <oliver.mcfadden@nokia.com> 126391 126392commit aafcae79d17c1f802bc880d2142af7171fed75d8 126393Author: Tiago Vignatti <tiago.vignatti@nokia.com> 126394Date: Fri Jul 10 14:56:31 2009 +0300 126395 126396 xfree86: remove unused vbe header file 126397 126398 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 126399 Acked-by: Oliver McFadden <oliver.mcfadden@nokia.com> 126400 126401commit 3711d68f657c77b947cc4670cc4eac4f62de3af8 126402Author: Peter Hutterer <peter.hutterer@who-t.net> 126403Date: Fri Jul 10 11:10:30 2009 +1000 126404 126405 Revert "XKB: Sanitise * actions" commits (#19602) 126406 126407 Reverts the following four patches: 126408 126409 feb757f384382c7782ceac55 "XKB: Sanitise vmods for redirected keys" 126410 b5f49382fe48f0a762d9a15f "XKB: Sanitise ctrls action" 126411 1bd7fd195d85681e722161f8 "XKB: Sanitise pointer actions" 126412 61c508fa78aa08ea2666fde9 "XKB: Sanitise vmods in actions" 126413 126414 Strictly speaking, the structs used in the server are not part of the client 126415 ABI. Practically, they are as we copy from the wire straight into the 126416 structs. Changing the struct sizes breaks various wire/server conversions. 126417 126418 Even when the structs have the same size, some internal magic causes 126419 conversions to fail. Visible by diffing the output files of: 126420 setxkbmap -layout de; xkbcomp -xkb :0 busted.xkb 126421 setxkbmap -layout de -print | xkbcomp -xkb - correct.xkb 126422 126423 Interestingly enough, busted.xkb is the working one although the output is 126424 incorrect. Revert the four offending patches until the exact cause of this 126425 breakage can be determined. 126426 126427 This patch restores functionality to Level3 modifiers. 126428 126429 X.Org Bug 19602 <http://bugs.freedesktop.org/show_bug.cgi?id=19602> 126430 126431 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126432 126433commit 08df24555cb432eb0d90a3f63275e9485e777c4c 126434Author: Soren Sandmann Pedersen <ssp@redhat.com> 126435Date: Thu Jul 9 17:23:04 2009 -0400 126436 126437 Reserve space for two GC values in copy_drawable(). 126438 126439 Pointed out by Pierre Willenbrock. 126440 126441commit 3c53f3241319d96affad2f4539f7661e02e80521 126442Author: Adam Jackson <ajax@redhat.com> 126443Date: Thu Jul 9 17:21:07 2009 -0400 126444 126445 Since font modules are dead, don't mention them in xorg-server.pc 126446 126447commit e812103382a7bc3f7ba8b2f44ab97960dec265a9 126448Author: Adam Jackson <ajax@redhat.com> 126449Date: Thu Jul 9 10:58:17 2009 -0400 126450 126451 config: drop i810. 126452 126453 The driver assuredly no longer builds against 1.7. Please don't feel 126454 compelled to fix it. 126455 126456commit 55ac155d42268009eaecbc183329a8744d6ae48c 126457Author: Alan Coopersmith <alan.coopersmith@sun.com> 126458Date: Tue Jul 7 16:59:39 2009 -0700 126459 126460 Xephyr & Xserver man page fixes 126461 126462 Xephyr(1): Fix quote formatting, add missing ' to contraction 126463 Xserver(1): Add Xephyr(1) & startx(1) to SEE ALSO section 126464 126465 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 126466 126467commit 36dc66ae79b6e1cbe0a23587e32aebb68365b0d0 126468Author: Michael Lorenz <macallan@netbsd.org> 126469Date: Tue Jul 7 20:21:39 2009 -0400 126470 126471 The way XaceHook() mixes struct initializers and va_arg() is not portable and 126472 gives bogus data on sparc and probably others leading to a crash. 126473 126474 Fix: 126475 Don't use initializers, instead set each member directly to enforce order. 126476 126477 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 126478 126479commit 48a9d65b88f56d1f8ab3bf824a4fe48c2f68725f 126480Author: Jon TURNEY <jon.turney@dronecode.org.uk> 126481Date: Sun Jul 5 16:06:32 2009 +0100 126482 126483 Cygwin/X: Window positioning improvements for multiwindow mode 126484 126485 A few tweaks to winUpdateWindowPosition(): 126486 126487 * Don't allow window decoration to disappear off to top or left edge of the 126488 display as a result of adjustments to add decoration for the window style 126489 126490 * Honour the position requested by window geometry for the client area (so 126491 windows which save their position don't get moved by the decoration width 126492 every time they are created) (unless we need to bump it away from top or left edges) 126493 126494 * Fix an issue with initial window placement being offscreen on multimonitor 126495 systems when some monitors have negative coordinates (are to the left of or 126496 above the primary monitor) 126497 126498 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 126499 126500commit 5aec72745232dd61e60cfbf3acc4628d4fcd0315 126501Author: Keith Packard <keithp@keithp.com> 126502Date: Mon Jul 6 11:53:02 2009 -0700 126503 126504 Remove old DRI2 buffer alloc/free interfaces 126505 126506 These old interfaces are no longer supported by the server, removing them 126507 requires bumping the video driver ABI. Note that this is not guaranteed to 126508 be the last change in ABI version 6. 126509 126510 Signed-off-by: Keith Packard <keithp@keithp.com> 126511 Reviewed-by: Eric Anholt <eric@anholt.net> 126512 126513commit 2e2c5b216cc1c7a9bc26bd2c68226aaed5fc52ca 126514Author: Keith Packard <keithp@keithp.com> 126515Date: Wed Jul 1 14:01:57 2009 -0700 126516 126517 dri2: Preserve compatibility with 1.6 DRI2 API/ABI 126518 126519 The old DRI2 buffer allocation API wasn't great, but there's no reason to 126520 make the server stop working with those drivers. This patch has the 126521 X server adapting to the API provided by the driver, using the new API where 126522 available and falling back to the old API as necessary. A warning will be 126523 placed in the log file when the old API is in use. 126524 126525 Signed-off-by: Keith Packard <keithp@keithp.com> 126526 126527commit 7c7f0c2c6a04f7044d5ce69e97a615735e5831f1 126528Author: Søren Sandmann Pedersen <sandmann@redhat.com> 126529Date: Fri Jul 3 14:36:54 2009 -0400 126530 126531 Use IncludeInferiors when copying windows before compositing. 126532 126533 Part of bug 22484. 126534 126535commit 43ee8d2ead862f84a4526a472519663ef27a8d6a 126536Author: Matthias Hopf <mhopf@suse.de> 126537Date: Wed Jun 24 18:26:23 2009 +0200 126538 126539 Unclaim PCI slot if driver probing fails. 126540 126541 Otherwise no subsequent driver will be able to claim this pci slot. 126542 Example for this: fbdev tries to claim, but framebuffer device is not 126543 available. Later on VESA cannot claim the device. 126544 126545commit 73abdc94c3cceadeda26a9b6bd3cdfecf0df8db2 126546Author: RALOVICH, Kristóf <tade60@freemail.hu> 126547Date: Sun Jul 5 16:42:54 2009 +0200 126548 126549 glx: damage is only used with DRI 126550 126551 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 126552 126553commit 7208a0f032d487bd3bc9809800f4557f1da1e2db 126554Author: RALOVICH, Kristóf <tade60@freemail.hu> 126555Date: Mon Jun 29 15:18:56 2009 +0200 126556 126557 glx: remove Xgl leftover 126558 126559 GlxSetRenderTables was only used by the long gone Xgl. 126560 126561 Signed-off-by: Tiago Vignatti <tiago.vignatti@nokia.com> 126562 126563commit 89cf81cd85919e3dbb5adff5e6c6056c7990b60f 126564Author: Peter Hutterer <peter.hutterer@who-t.net> 126565Date: Fri Jul 3 10:29:40 2009 +1000 126566 126567 Xext: return BadValue for XTestFakeInput on unsupported capabilities. 126568 126569 Calling XTestFakeDevice*Event on a device that doesn't allow the matching 126570 event returns BadValue. 126571 126572 Reported-by: Florian Echtler 126573 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126574 126575commit b773b4e8e0087993406d1d8c2df895db4d6301ff 126576Author: Peter Hutterer <peter.hutterer@who-t.net> 126577Date: Mon Jul 6 11:04:47 2009 +1000 126578 126579 s/MAX_DEVICES/MAXDEVICES/ updates. 126580 126581 The number of input devices is MAXDEVICES, not MAX_DEVICES (f781a752e6) 126582 Two comments updated to refer to MAXDEVICES. 126583 126584 MAX_FUNCS in sigio.c was set to 16 if MAX_DEVICES was undefined. If more 126585 than 15 physical input devices were present, this could result in a 126586 failure to install the SIGIO handler for any device above 15. 126587 126588 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126589 126590commit 183c075d2f9d5f6effa1ce7ab135fb0c2e46085a 126591Author: Peter Hutterer <peter.hutterer@who-t.net> 126592Date: Fri Jul 3 13:26:48 2009 +1000 126593 126594 dix: always init the full button map to default values (#22594) 126595 126596 Master devices must have the standard button map applied for all buttons to 126597 ensure buttons larger than 7 (the default for MDs) are mapped appropriately. 126598 126599 We can't copy the button map from SDs to MDs since that breaks the chained 126600 button mapping. However, by ensuring all buttons (even non-existing ones) 126601 are mapped, devices that send such buttons continue to work. 126602 126603 X.Org Bug 22594 <http://bugs.freedesktop.org/show_bug.cgi?id=22594> 126604 126605 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126606 126607commit 744bb559826ede37a77e9000b6c620eaa6a3c837 126608Author: Peter Hutterer <peter.hutterer@who-t.net> 126609Date: Fri Jul 3 13:58:38 2009 +1000 126610 126611 Xext: remove unused variable 'it'. 126612 126613 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126614 126615commit fd4eed69e40cf090aa16e79a8596eb1442461829 126616Author: Eric Anholt <eric@anholt.net> 126617Date: Thu Jul 2 15:24:54 2009 -0700 126618 126619 dri2: Enable GLX_SGI_make_current_read when the DRI driver supports it. 126620 126621 This matches idr's 82f150d73cc9a7d8eaf7241a51b03af05bcec159 for DRI1. 126622 126623commit d6b8205e699c0c62af76c4a9cbff1402337927b3 126624Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 126625Date: Fri Jul 3 19:25:33 2009 -0700 126626 126627 XQuartz: ProcAppleWMAttachTransient to play nice with the new Dock in SL 126628 (cherry picked from commit ddc0242d8f291d0f961ba9aa8c883e39861ce2ad) 126629 126630commit a49ae50370ec94e08e7dec3c742d33e20e79ef36 126631Author: Colin Harrison <colin.harrison@virgin.net> 126632Date: Mon Jun 29 22:55:52 2009 +0100 126633 126634 Xming: Apply window style hints in -multiwindow mode 126635 126636 Remove frames from 'dock' windows and make them topmost in -multiwindow mode. 126637 Remove frames from windows with MOTIF_WM_HINTS of no decorations in -multiwindow mode. 126638 Apply some _NET_WM_STATE hints in -multiwindow mode. 126639 Apply window styles overrides from .rc file 126640 Correctly update region of shaped windows when applying styles 126641 126642 Copyright (C) Colin Harrison 2005-2009 126643 http://www.straightrunning.com/XmingNotes/ 126644 http://sourceforge.net/projects/xming/ 126645 126646commit 17e67c407d130c325d3899c18d68b8eef6a88bea 126647Author: Joe Krahn <jkrahn@nc.rr.com> 126648Date: Wed Jun 24 23:32:09 2009 +0100 126649 126650 Cygwin/X: Change to a single native window class for all X windows 126651 126652 from fd.o Bugzilla #4491: 126653 126654 There is no point in having one class for every window, aside from trying to 126655 set custom icons via the class, which we no longer do, so converted to using 126656 a single class for all client windows. 126657 126658 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 126659 126660commit 2c69deb92e11542f615df0f24fdc03e3b4415475 126661Author: Rémi Cardona <remi@gentoo.org> 126662Date: Fri Jul 3 10:51:50 2009 +0200 126663 126664 configure: libXinerama isn't needed anymore 126665 126666 since libXinerama commit 90d4d23bf2e94721149ddc0a80093b10a82e8845 and 126667 xineramaproto commit 21477147613c28c968b5e1eb9d8aea7017dd399d, the 126668 server no longer needs libXinerama. 126669 126670 Signed-off-by: Rémi Cardona <remi@gentoo.org> 126671 126672commit 97e29ffb5bf04a9d8280f1fa32ceced148503492 126673Author: Peter Hutterer <peter.hutterer@who-t.net> 126674Date: Fri Jul 3 09:52:04 2009 +1000 126675 126676 xfree86: fix wrong IsMaster() check causing crashes. 126677 126678 Crashes caused by dereferencing NULL if the path was executed for a floating 126679 slave device. 126680 126681 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126682 126683commit 50a2a8dc76645d8736f7d712f0ef05f23089407e 126684Author: Peter Hutterer <peter.hutterer@who-t.net> 126685Date: Fri Jul 3 09:18:57 2009 +1000 126686 126687 Fix IsXtstDevice - returns false positives since 0814f511d5. 126688 126689 Missing check for the value of 'mid' returned false positives if master was 126690 NULL. 126691 126692 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126693 126694commit 3525d140567e0ad5f0184e4b37893c47239e1628 126695Author: Felix Kuehling <felix.kuehling@amd.com> 126696Date: Tue Jun 30 20:07:30 2009 -0400 126697 126698 Export symbols needed by the RandR implementation in fglrx 126699 126700 GIT change 126701 http://cgit.freedesktop.org/xorg/xserver/commit/?id=45c8bd0fe54273039fdaa1eeeafb81b5774f2c75 126702 changed the default symbol visibility of the Xserver. As a result 2 symbols 126703 that are needed by the RandR 1.2/1.3 implementation in the fglrx driver are no 126704 longer visible: 126705 126706 xf86configptr 126707 xf86CursorScreenKey 126708 126709 We would like to get these two symbols _X_EXPORT'ed before Xserver 1.7 is 126710 released. Otherwise it will be problematic for fglrx to support RandR 1.3 on 126711 Xserver 1.7. 126712 126713 In the future, we may want to sync our RandR implementation to later versions 126714 of the RandR implementation in hw/xfree86/modes. Therefore it would be nice if 126715 all symbols used by the Xserver RandR implementation were _X_EXPORT'ed in the 126716 future. 126717 126718commit e13605ea40cfc671314a0086c75b917564298b55 126719Author: Peter Hutterer <peter.hutterer@who-t.net> 126720Date: Tue Jun 23 14:09:22 2009 +1000 126721 126722 dix: introduce "Xtst Device" label property. 126723 126724 Xtst devices get this property assigned automatically so they can be 126725 detected easily by a client. 126726 The property is read-only. 126727 126728 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126729 126730commit 0814f511d56a89c7b1868b17eba7a89f990b9ab1 126731Author: Peter Hutterer <peter.hutterer@who-t.net> 126732Date: Tue Jun 23 10:50:52 2009 +1000 126733 126734 input: store the master device's ID in the devPrivate for XTest devices. 126735 126736 Rather than storing a simple boolean in the devPrivate for XTest devices, 126737 store the actual master device's id (since it is constant for the life of 126738 the device anyway). 126739 126740 Callers should use GetXtstDevice now instead of digging around in the 126741 devPrivates themselves. 126742 126743 This patch allows for a cleanup in the creation of new master devices since 126744 GetMaster and GetXtstDevice spare the need for loops, IsPointer checks and 126745 similar. 126746 126747 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126748 Acked-by: Benjamin Close <Benjamin.Close@clearchain.com> 126749 126750commit 1bcc0d3c244ce7d9f5cbab626aa5fd5784b41a1c 126751Author: Peter Hutterer <peter.hutterer@who-t.net> 126752Date: Mon Jun 22 16:55:12 2009 +1000 126753 126754 input: abstract Xtst device lookup 126755 126756 The callers should need to use the dev privates key to look up xtest 126757 devices. 126758 126759 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126760 Acked-by: Benjamin Close <Benjamin.Close@clearchain.com> 126761 126762commit 34424fab9abd7a4ca11036be25414129980db0e0 126763Author: Peter Hutterer <peter.hutterer@who-t.net> 126764Date: Tue Jun 23 11:50:29 2009 +1000 126765 126766 dix: set the client's error value to the bad deviceid in check_butmap_change. 126767 126768 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126769 126770commit da04e8f1354fa9d253443489dc002b16f94d6fab 126771Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 126772Date: Tue Jun 30 17:52:42 2009 -0400 126773 126774 dix/property.c: use memcpy where appropriate. 126775 126776 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 126777 126778commit 65d74d93145d22b68bad5728a7ebe38dc662cb21 126779Author: Jon TURNEY <jon.turney@dronecode.org.uk> 126780Date: Wed May 20 21:32:54 2009 +0100 126781 126782 Cygwin/X: Fix multiwindow extwm mode to build again 126783 126784 Build and link with rootless extension 126785 Update Xwin code for removal of RootlessAccelInit() 126786 Fix Xwin code which now has a collision with the type name EventType 126787 126788 Based on patches from Colin Harrison, Jon Turney and Yaakov Selkowitz 126789 126790 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 126791 126792commit 638ca9a7a2363757dc5b5d456e10d34f6f158885 126793Author: Jon TURNEY <jon.turney@dronecode.org.uk> 126794Date: Mon May 18 18:14:41 2009 +0100 126795 126796 Cygwin/X: Avoid a visual glitch on window move in rootless modes 126797 126798 Handle and ignore WM_ERASEBKGND since we repaint the entire invalidated region anyhow 126799 (this avoids a white flickering on window resize) 126800 126801 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 126802 126803commit b718b2e0880cf2b969675da98d5ef8a4a01ca5d6 126804Author: Jon TURNEY <jon.turney@dronecode.org.uk> 126805Date: Mon May 18 18:14:09 2009 +0100 126806 126807 Cygwin/X: Allow pointer warping to work in rootless modes 126808 126809 Mouse pointer warping in multiwindow/rootless mode was never implemented, 126810 due to concerns that moving the mouse pointer without asking might be rude 126811 126812 This patch allows X applications to move the mouse pointer in rootless modes, 126813 Let's hope they don't abuse this privilege ;-) 126814 126815 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 126816 126817commit 85614946ba3d5a233eece612afc7f09572a347c2 126818Author: Jon TURNEY <jon.turney@dronecode.org.uk> 126819Date: Mon May 18 18:13:08 2009 +0100 126820 126821 Cygwin/X: Improve mouse tracking for moving/resizing undecorated windows 126822 126823 In -multiwindow mode, tell Windows we wish to capture the mouse when a button 126824 is down. This causes Windows to continue to send movement events for the mouse 126825 even if the mouse pointer moves outside the window frame. 126826 126827 This helps greatly with undecorated windows which have regions you can grab 126828 to move (e.g. gmplayer, xine control panels) or resize (e.g. Songbird) the 126829 window, as it means the window continues to receive the mouse motion even if the 126830 mouse pointer moves out of the window (which presumably happens if we don't 126831 manage to update the window fast enough to track the mouse pointer) 126832 126833 Consolidate the multiple instances of the code to start the mouse position 126834 polling timer into a new function winStartMousePolling(), and use that to 126835 restart the polling timer when we release the mouse. 126836 126837 Also, start the timer on WM_SHOW, so that xeyes will track the mouse position 126838 when it is first shown, even if the mouse doesn't enter it's window 126839 (You probably need focus-stealing turned off to see this problem) 126840 126841 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 126842 126843commit f351c10a9774cc0ea2cbb58f00f07ece7f7c6e73 126844Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 126845Date: Mon Jan 5 16:17:30 2009 +0000 126846 126847 Cygwin/X: Man page fixes 126848 126849 Correct path names in man pages, using cpprules.in 126850 Install XWinrc man page into section 5 126851 126852 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 126853 126854commit 5c1afac5eea1d8327c74342d12d082b75f0cebde 126855Author: Adam Jackson <ajax@redhat.com> 126856Date: Mon Jun 29 11:16:16 2009 -0400 126857 126858 EDID: Fix timing class names to match the spec 126859 126860commit bf0f3b8f2bc830c8bd6f8f9410b89394b8d96257 126861Author: Peter Hutterer <peter.hutterer@who-t.net> 126862Date: Mon Jun 29 13:31:02 2009 +1000 126863 126864 test: fix build error introduced by XINPUT_ABI 7 126865 126866 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126867 126868commit cbeb6a73c44143674a700d36a7e9804d41003a4f 126869Author: Thomas Jaeger <ThJaeger@gmail.com> 126870Date: Sat Jun 20 20:17:41 2009 -0400 126871 126872 dix: report subpixel coordinates for high-resolution devices 126873 126874 Acked-by: Simon Thum <simon.thum@gmx.de> 126875 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126876 126877commit 069b4c5f0a38649de73d1e0f70ee81bf862399dd 126878Author: Simon Thum <simon.thum@gmx.de> 126879Date: Sat Jun 20 18:57:22 2009 +0200 126880 126881 dix: make part of ptrveloc.h internal 126882 126883 Though this is a SDK header, some functions are intended solely 126884 for the server. 126885 126886 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126887 126888commit 2830e8493757e0da1253fe5ab280562b84730e77 126889Author: Simon Thum <simon.thum@gmx.de> 126890Date: Wed Jun 24 11:33:19 2009 +0200 126891 126892 dix: rename pDev->dev, pVel->vel for consistency 126893 126894 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126895 126896commit 373e8c960d00e2b8c2250dd3f66859b081e14854 126897Author: Simon Thum <simon.thum@gmx.de> 126898Date: Wed Jun 24 11:16:24 2009 +0200 126899 126900 dix: improve pointer acceleration API 126901 126902 This makes the ptr accel api actually sensible from a driver 126903 perspective, since it avoids superfluous device lookups. 126904 Also, makes independent accel contexts possible. 126905 126906 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 126907 126908commit b8050bb6deebdb1ee60731f63884ffca575c09ce 126909Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 126910Date: Sat Jun 27 15:28:34 2009 -0700 126911 126912 XQuartz: Don't leave zombied processes at startup 126913 (cherry picked from commit 40c1406830588fa85d880e9f4e9ca570db1db306) 126914 126915commit aaff92c8c22a47804a21010d023ef76d82e7ec7e 126916Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 126917Date: Sat Jun 27 14:40:03 2009 -0700 126918 126919 XQuartz: 64bit fix for screen origin in AppleWM 126920 (cherry picked from commit 60a757d2802a5c34acd91ca9a052937b5a169ede) 126921 126922commit cabcc1a7ad697dde915794ddcea6a300f66b0a28 126923Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 126924Date: Sat Jun 27 10:40:59 2009 -0700 126925 126926 XQuartz: More localization updates 126927 (cherry picked from commit 5925c1f48ad05bf94195b986c1fdefc52a20ae42) 126928 126929commit 31166c2ae0ce898c96995a8b16b58b127dc85a2f 126930Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 126931Date: Fri Jun 26 16:51:22 2009 -0400 126932 126933 xace: add a new hook for checking property content after it has been set. 126934 126935 Allows security modules to enforce what property contents can be set by 126936 clients. Uses the new DixPostAccess bit to distinguish between the 126937 existing call made during the lookup (with the old property data) and 126938 this new call. Note that this only applies to writes, prepends, or 126939 appends to existing properties; for new properties the existing 126940 DixCreateAccess hook call may be used since it includes the new data. 126941 126942 Refer to the XACE-Spec document in xorg-docs, section "Property Access." 126943 126944 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 126945 126946commit 51105de9b0d865c4b5e5a7d9ab23c89d808d1cfa 126947Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 126948Date: Fri Jun 26 16:50:12 2009 -0400 126949 126950 xselinux: ignore property hook calls with the new Post access mode bit set. 126951 126952 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 126953 126954commit 10812204b3415c969bcebd3215e84d758a0b4dd8 126955Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 126956Date: Fri Jun 26 16:48:24 2009 -0400 126957 126958 dix: add a new DixAccess bit, "DixPostAccess". 126959 126960 This will be used for follow-up checks after a client has written something, 126961 for security modules that enforce a set of valid values a client can set. 126962 126963 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 126964 126965commit 9480725af0275b72b24b9e1cb3b68d5b97288d01 126966Author: Rémi Cardona <remi@gentoo.org> 126967Date: Wed Jun 24 10:12:50 2009 +0200 126968 126969 xfree86: DIST_SUBDIRS should not contain variables 126970 126971 This patch reverts a change done in commit 126972 800d4ceb946c93724661ebfd28afb658ca63b6b1 to add conditional XAA support 126973 126974 Signed-off-by: Rémi Cardona <remi@gentoo.org> 126975 126976commit 84662e40c3d4141ebb298a1ad714f75056a4ab74 126977Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 126978Date: Thu Jun 18 23:19:21 2009 -0400 126979 126980 Xi: check for GetAttr permission when listing or querying devices. 126981 126982 If the check fails, leave the device off the returned list of info 126983 structures. Under XI2, this may cause inconsistent views of the device 126984 topology after a change (for example, devices disappearing from view, 126985 or showing as attached to a master that cannot be seen). More work is 126986 needed to deal with topology changes and device relabeling. 126987 126988 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 126989 126990commit 00bc8d34c68dab6c818cd1c7e03e9992d1d0cbfc 126991Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 126992Date: Thu Jun 18 21:41:17 2009 -0400 126993 126994 Xi: check for Use permission on the device in SetClientPointer(). 126995 126996 Presumably, some intelligent, XI2-aware management app will be calling 126997 XISetClientPointer on behalf of other clients; this check makes sure 126998 the target client has permission on the device. 126999 127000 Requires changing the prototype to return status code instead of Bool. 127001 127002 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 127003 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 127004 127005commit 119b96667778391436998c76a68bf64e746c9e08 127006Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 127007Date: Thu Jun 18 21:02:28 2009 -0400 127008 127009 Xi: fix up access modes for calls to dixLookupDevice(). 127010 127011 New access modes are being passed to the device access hook for XI2: 127012 DixCreateAccess for creating a new master device; 127013 DixAdd/RemoveAccess for attaching/removing slave devices to a master; and 127014 DixListProp/GetProp/SetPropAccess for device properties. 127015 127016 Refer to the XACE-Spec document in xorg-docs, section "Device Access." 127017 127018 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 127019 127020commit 07c36e4fdcd93df3d33bdab6cca4780ebc9c1f54 127021Author: Dave Airlie <airlied@redhat.com> 127022Date: Wed Jun 10 13:33:47 2009 +1000 127023 127024 dix/resource: fix use after free in resource code with DRI 127025 127026 LookupClientResourceComplex is used by DRI1 code to find and free a DRI 127027 drawable in a callback, however when the DRI code returns this->value 127028 is now pointing at freed memory. It seemed easiest to store the value 127029 to a temporary and return it afterwards. 127030 127031 Another option might be a new FreeClientResourceComplex or one that 127032 also returns the id, so we can free it using an alternative means. 127033 127034 found using valgrind. 127035 127036 amended along ajax's suggestions 127037 127038commit 184deb9bc325eb7aa7eb7b7d4f98aa917f0269cb 127039Author: Dave Airlie <airlied@redhat.com> 127040Date: Wed Jun 24 10:40:05 2009 +1000 127041 127042 GLX: make function static. 127043 127044 This function isn't called from anywhere else and I don't think it shuold be. 127045 127046commit 9d85b56078ec05da1369ca22930d8eb214c389db 127047Author: Dave Airlie <airlied@redhat.com> 127048Date: Wed Jun 24 10:38:49 2009 +1000 127049 127050 GLX: note the implicit flushes with ReadPixels in indirect contexts. 127051 127052 This just notes the flush has occured when readpixels returns, and 127053 fixes the glean test. 127054 127055commit e341512bfa40dd98853a20596dc65dcac4dcaa37 127056Author: Thomas Jaeger <ThJaeger@gmail.com> 127057Date: Sat Jun 20 21:37:59 2009 -0400 127058 127059 dix: update a comment 127060 127061 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127062 127063commit 5cbd4d3d6e54d202ecdbb527b57aaefeb8435600 127064Author: Thomas Jaeger <ThJaeger@gmail.com> 127065Date: Sat Jun 20 20:17:04 2009 -0400 127066 127067 dix: do away with an instance of temporary in-place modification 127068 127069 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127070 127071commit 94cdc1ef0a72802573a11ba292e2f5cc8474aa8e 127072Author: Thomas Jaeger <ThJaeger@gmail.com> 127073Date: Mon Jun 22 13:00:37 2009 -0400 127074 127075 dix: deal with first_valuator > 0 correctly if POINTER_SCREEN is set 127076 127077 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127078 127079commit 21cbb4c9de44e2629dc0fa6f647ce2d139f2cef0 127080Author: Oliver McFadden <oliver.mcfadden@nokia.com> 127081Date: Wed Jun 24 00:24:44 2009 +0300 127082 127083 xorg-server.h.in: Export the X Access Control Extension (XACE), too. 127084 127085 Acked-by: Daniel Stone <daniel@fooishbar.org> 127086 127087commit 800d4ceb946c93724661ebfd28afb658ca63b6b1 127088Author: Tiago Vignatti <tiago.vignatti@nokia.com> 127089Date: Wed Jun 24 00:13:32 2009 +0300 127090 127091 configure: Provide the --enable/disable-xaa option. 127092 127093 Acked-by: Daniel Stone <daniel@fooishbar.org> 127094 127095commit ce3d539ac6ecab3e343cef7c4dc112bc82ea3b02 127096Author: Oliver McFadden <oliver.mcfadden@nokia.com> 127097Date: Wed Jun 24 00:08:51 2009 +0300 127098 127099 xf86Config: Avoid attempting to load non-compiled modules. 127100 127101 Acked-by: Daniel Stone <daniel@fooishbar.org> 127102 127103commit 33e25143361d63ea776dfafeeb6d4ffb2f6fd8d4 127104Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 127105Date: Fri May 15 22:42:30 2009 +0100 127106 127107 Cygwin/X: add hebrew to autodetected keyboard layouts 127108 127109 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 127110 127111commit b079945c39e2a72220b46953352e24c3a74ef39a 127112Author: Jon TURNEY <jon.turney@dronecode.org.uk> 127113Date: Fri Feb 20 15:21:35 2009 +0000 127114 127115 Cygwin/X: more warnings fixes 127116 127117 More warning fixes, mainly removing casts on function return values 127118 127119 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 127120 127121commit e79db6a97d02c8a256a4a7e145ea0b48b5a084ab 127122Author: Jon TURNEY <jon.turney@dronecode.org.uk> 127123Date: Wed Feb 11 23:00:58 2009 +0000 127124 127125 Cygwin/X: Remove an obsolete mention of xf86Config 127126 127127 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 127128 127129commit 3020b1d43e34fca08cd51f7c7c8ed51497d49ef3 127130Author: Michel Dänzer <daenzer@vmware.com> 127131Date: Tue Jun 23 16:45:40 2009 +0200 127132 127133 glx: Clean up more thoroughly if the drawable of a current context goes away. 127134 127135 Fixes crash when restarting compiz, due to cl->currentContexts[x] being stale. 127136 127137commit df597709d71f47b8516e27c6fb1bfffd59de5e48 127138Author: Michel Dänzer <daenzer@vmware.com> 127139Date: Tue Jun 23 16:45:39 2009 +0200 127140 127141 dri2: Don't crash if pPriv is NULL. 127142 127143commit 048697ccfa31cf7f7a29afa90a2f702d43efb7d4 127144Author: Ben Skeggs <bskeggs@redhat.com> 127145Date: Wed May 27 15:24:34 2009 +1000 127146 127147 quirk: use first detailed timing as preferred for PEA prod 9003 (rh#492359) 127148 127149 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127150 127151commit 07154db4a8b96467785fd6be93d20379acacffa8 127152Author: Ben Gamari <bgamari.foss@gmail.com> 127153Date: Mon Jun 22 08:59:45 2009 -0400 127154 127155 Xi: Remove redundant and incorrect butmap range check 127156 127157 Maps are CARD8s, therefore checking for values above 255 is completely 127158 unnecessary. Moreover, 0 is a valid value for maps, so the check wasn't 127159 even correct to begin with. This fixes bug #22392, which was uncovered 127160 by commit 280b7f92d729ec910ffa3d18dce7bbc215be7a3c. 127161 127162 Signed-off-by: Ben Gamari <bgamari.foss@gmail.com> 127163 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127164 127165commit 96706c24bd57fbd9b11e5bd5e38d05d81b90aebe 127166Author: Peter Hutterer <peter.hutterer@who-t.net> 127167Date: Thu Jun 18 15:40:08 2009 +1000 127168 127169 dix: fix wrong indices in set_valuator. 127170 127171 Reported-by: Thomas Jaeger 127172 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127173 127174commit ae20e748cd3a656173e1f50109bfd4af0712bb87 127175Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 127176Date: Fri Jun 19 21:16:10 2009 -0700 127177 127178 XQuartz: More localization updates 127179 (cherry picked from commit 7fbe974246f54535c545861a57c043d80e127ee7) 127180 127181commit 30df49f54945e75f033a0eb6445c26d37eb33c5e 127182Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 127183Date: Thu Jun 18 23:15:47 2009 -0400 127184 127185 Xi: fix 2 memory leaks. 127186 127187 In ProcXIQueryDevice() and XISendDeviceHierarchyEvent(). 127188 127189 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 127190 127191commit eb35402d0a5290e8a73d7d1e92f173294c364cc2 127192Author: Adam Jackson <ajax@redhat.com> 127193Date: Fri Jun 19 12:42:07 2009 -0400 127194 127195 pci: Dump vendor/devices ids in the printed device list 127196 127197commit c733660428c0a7c1d11f7bd21e23e1bb934d352e 127198Author: Søren Sandmann Pedersen <sandmann@redhat.com> 127199Date: Fri Jun 19 08:37:18 2009 -0400 127200 127201 Use pixman_version_string() instead of PIXMAN_VERSION_STRING 127202 127203 Pointed out by Julien Cristau. 127204 127205commit aef6b904ebf0d7de6259058606c7c04ea177bda3 127206Author: Dave Airlie <airlied@panoply-rh.(none)> 127207Date: Thu Mar 13 16:16:46 2008 +1000 127208 127209 fbdev: make entity fail if PCI claimed already. 127210 127211 bad kitty fbdev, been shipping this in Fedora for a while now 127212 127213commit 3efb23a6c40dc3583d083d25ada3853ecc56000d 127214Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 127215Date: Thu Jun 18 19:35:22 2009 -0400 127216 127217 xselinux: Add new device permissions for XI2. 127218 127219 Refects the ability of clients to add/remove devices and device properties. 127220 127221 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 127222 127223commit 1e060c3d8b13d352a58fc65980cb9a3c6cb5718f 127224Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 127225Date: Thu Jun 18 18:48:24 2009 -0400 127226 127227 xselinux: Move the security class mapping to the header file. 127228 127229 Take the mapping of DixAccess bits to Flask permissions, move it 127230 into the header file, break up the extremely long lines, and 127231 annotate the permission names with the bit being referenced. 127232 127233 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 127234 127235commit 75c51c67b340548286efd41a53882e2acaf74ab5 127236Author: Alan Coopersmith <alan.coopersmith@sun.com> 127237Date: Thu Jun 18 09:49:12 2009 -0700 127238 127239 Clarify use of and need for mffs vs. ffs 127240 127241 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 127242 127243commit 128cd03eecacc6d5c5903d59a11966dcf3697bf1 127244Author: Søren Sandmann Pedersen <sandmann@redhat.com> 127245Date: Sat Jun 13 10:55:04 2009 -0400 127246 127247 Fix miComputeCompositeRegion() to follow new clip rules. 127248 127249 Ignore the hierarchy clip, and always apply any client clip after 127250 transformation and repeating. 127251 127252commit e9aa61e9f0d663d5b34a397b943b4d1df44e873d 127253Author: Søren Sandmann Pedersen <sandmann@redhat.com> 127254Date: Sat Jun 13 10:28:21 2009 -0400 127255 127256 Fix clipping when windows are used as sources 127257 127258 The new clipping rules: 127259 127260 - client clips happen after transformation 127261 - pixels unavailable due to the hierarchy are undefined 127262 127263 The first one is implemented in pixman; the second one is realized by 127264 making a copy of window sources (to prevent out-of-bounds access). 127265 127266commit d9b5e77a0e48a16c53653b56bc61a0b8dc4122a1 127267Author: Søren Sandmann Pedersen <sandmann@redhat.com> 127268Date: Tue Jun 9 14:36:21 2009 -0400 127269 127270 Print the current version of pixman. 127271 127272commit 1e9907499c27321a2aa5dc8a75a375b7a82c999a 127273Author: Peter Hutterer <peter.hutterer@who-t.net> 127274Date: Thu Jun 18 15:14:00 2009 +1000 127275 127276 record: use dixLookupResourceByClass instead of LookupIDByClass. 127277 127278 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127279 127280commit 66089e9129a821cfb1983d3d35f41b975a52de5e 127281Author: Peter Hutterer <peter.hutterer@who-t.net> 127282Date: Wed Mar 25 12:55:42 2009 +1000 127283 127284 xfree86: fix SWCursor check in xf86CursorSetCursor. 127285 127286 Wrong check for inputInfo.pointer resulted in a SW cursor being rendered when 127287 the pointer left the screen (in a Xinerama setup). 127288 We must call the sprite rendering function if 127289 - SW cursors are enabled, or 127290 - The current device is not the VCP and not attached to the VCP. 127291 127292 Reported-by: Gordon Yuan <GordonYuan@viatech.com.cn> 127293 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127294 127295commit afa680e495622f521cae80563511c0d284f57551 127296Author: David Miller <davem@davemloft.net> 127297Date: Thu Jun 11 05:15:05 2009 -0700 127298 127299 mi: ignore DGA events in ChangeDeviceID 127300 127301 DGA events don't have a deviceid, so they don't need changing. 127302 127303 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127304 127305commit cd8abd17abcc8ae9da6704d03b97a9f134f36aab 127306Author: Jon TURNEY <jon.turney@dronecode.org.uk> 127307Date: Wed Jun 17 14:17:07 2009 +0100 127308 127309 Cygwin/X: Fix permuted args to InitPointerDeviceStruct() 127310 127311 Fix a minor error in commit a30fef9956b296f59ea18a9ee38d0abafeb15a4e, 127312 new btn_labels argument to InitPointerDeviceStruct() wasn't added in 127313 the right place 127314 127315commit 14581afb474552716c02ca15220ca7050123c375 127316Author: Benjamin Close <Benjamin.Close@clearchain.com> 127317Date: Thu Feb 26 17:32:10 2009 +1030 127318 127319 xfree86: correctly define barriers for FreeBSD amd64 127320 127321 Previously when compiling on freebsd amd64 we'd end up at xi86 127322 block (line 1315) which would define mem_barrier and write_mem_barrier 127323 to be NOP's. Instead they should be valid, as per the linux amd64 setup. 127324 127325 This stops the hangs experienced by many when using the nv driver 127326 which would hang due to out of order dma requests as noticed in 127327 http://bugs.freedesktop.org/show_bug.cgi?id=3168 127328 127329 Signed-off-by: Benjamin Close <Benjamin.Close@clearchain.com> 127330 127331commit e92dcb6ce07aa3cfb53e8bad5701481c106c4094 127332Author: Peter Hutterer <peter.hutterer@who-t.net> 127333Date: Thu Jun 18 14:40:20 2009 +1000 127334 127335 input: unify button numbers on master devices. 127336 127337 Master devices provide the union of all attached slave devices' buttons, 127338 i.e. the number of buttons on the master device is always the number of 127339 buttons of the slave device with the highest number of buttons. When slaves 127340 are attached or detached, the master device adjusts the button number to 127341 reflect the new buttons. 127342 127343 On a slave switch, this slave's button labels are copied into the master (up 127344 to slave->num_buttons). The remaining button labels (if any) stay as they 127345 are. Thus, if any of the higher buttons is still pressed, it reflects the 127346 label of the last pressed device that provided this button. 127347 127348 If two devices press the same button and it is differently labelled the last 127349 pressed one will be reflected in the master device. 127350 127351 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127352 127353commit 280b7f92d729ec910ffa3d18dce7bbc215be7a3c 127354Author: Peter Hutterer <peter.hutterer@who-t.net> 127355Date: Wed Jun 17 22:38:24 2009 +1000 127356 127357 dix: reduce MDs and xtest pointers to 7 buttons by default. 127358 127359 MD's will soon be the union of all devices anyway. XTest pointers are only 127360 for the core protocol XTest stuff, so 7 buttons (lmr + 4 wheel buttons) 127361 should do. 127362 127363 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127364 127365commit 6c7d992735eebbd7a20247926e7725896348b865 127366Author: Peter Hutterer <peter.hutterer@who-t.net> 127367Date: Wed Jun 17 09:11:52 2009 +1000 127368 127369 Require inputproto 1.9.99.12 127370 127371 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127372 127373commit 09cef7573938e5c08007e578e1b638bc5e1796a8 127374Author: Peter Hutterer <peter.hutterer@who-t.net> 127375Date: Wed Jun 17 09:04:08 2009 +1000 127376 127377 Xi: valuator/button labels are called labels now, not 'names' 127378 127379 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127380 127381commit bc2ff5365030ad8bc11efde430b1064080dd7098 127382Author: Peter Hutterer <peter.hutterer@who-t.net> 127383Date: Tue Jun 16 19:36:57 2009 +1000 127384 127385 Xi: copy the valuator mode from SD to MD. 127386 127387 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127388 127389commit 87d1f44bad608507e3995e17eb84fa0a0119796c 127390Author: Peter Hutterer <peter.hutterer@who-t.net> 127391Date: Tue Jun 16 14:18:45 2009 +1000 127392 127393 Xi: copy the button and axes labels into the XIQueryDevice reply. 127394 127395 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127396 127397commit a30fef9956b296f59ea18a9ee38d0abafeb15a4e 127398Author: Peter Hutterer <peter.hutterer@who-t.net> 127399Date: Tue Jun 16 16:38:11 2009 +1000 127400 127401 input: Add labels to buttons and valuators - ABI_XINPUT_VERSION 7 127402 127403 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127404 127405commit 17f9723f488d0470e3879c6b0dfdba61544cdd7b 127406Author: Peter Hutterer <peter.hutterer@who-t.net> 127407Date: Wed Jun 17 08:37:44 2009 +1000 127408 127409 input: bump to ints for deviceids - XI2 requires 16-bit deviceids. 127410 127411 Note: ABI break, but ABI_XINPUT_VERSION has NOT been bumped. Recompile input 127412 drivers. 127413 127414 Revert "Xi: return BadImplementation for deviceids 256 and above" 127415 This reverts commit 2b459f44f3edaea137df9a28bc7adfeb1b9f1df7. 127416 127417 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127418 127419commit cbeb73e2055f6c013c8fe6325851f2631170137d 127420Author: Peter Hutterer <peter.hutterer@who-t.net> 127421Date: Tue Jun 16 10:51:52 2009 +1000 127422 127423 Xi: return current valuator values in XIQueryDevice. 127424 127425 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127426 127427commit 80837dbefd9d5e96ab5c1f4b4c2d5c66ce17ce67 127428Author: Peter Hutterer <peter.hutterer@who-t.net> 127429Date: Tue Jun 16 10:51:17 2009 +1000 127430 127431 input: change axisVal from uint to double. 127432 127433 With subpixel support, uint just doesn't cut it. 127434 127435 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127436 127437commit 25b6fc4a42f7698e6ae0b16becb316bfd7835d05 127438Author: Peter Hutterer <peter.hutterer@who-t.net> 127439Date: Tue Jun 16 10:48:48 2009 +1000 127440 127441 Xi: last_valuator is used like an index, so range it accordingly. 127442 127443 The previous code would always skip the last valuator due to a wrong 127444 upper boundary in the loop. last_valuator is the index of the last set 127445 valuator - which also means it must be initialized to -1, not 0. 127446 127447 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127448 127449commit 3fc762ea9474cf92f6b7ef60dd0bb68f59385272 127450Author: Peter Hutterer <peter.hutterer@who-t.net> 127451Date: Tue Jun 16 10:48:03 2009 +1000 127452 127453 Xi: Motion events update the device state too. 127454 127455 Without this line the device's axis values don't get updated properly for 127456 pure motion events. 127457 127458 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127459 127460commit ad6347afead500fefb34131841f4beb5b03ad4ec 127461Author: Peter Hutterer <peter.hutterer@who-t.net> 127462Date: Tue Jun 16 16:35:40 2009 +1000 127463 127464 Xi: don't use a constant number for class sizes - use sizeof instead 127465 127466 The protocol is still changing, and having to debug crap like this is 127467 annoying. 127468 127469 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127470 127471commit d5a15e65b7d2fdda14fb719e3ebf076cce896386 127472Author: Peter Hutterer <peter.hutterer@who-t.net> 127473Date: Sun Jun 14 09:15:42 2009 +1000 127474 127475 Xi: include button state in XIButtonInfo. 127476 127477 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127478 127479commit 01241b4247a7ad0974d48412cc0d88e1a09a3c49 127480Author: Peter Hutterer <peter.hutterer@who-t.net> 127481Date: Sun Jun 7 19:43:11 2009 +1000 127482 127483 Xi: Add support for sourceid in the device classes. 127484 127485commit d230742ea820a21a3f1ed0c58b5e6d8680b2f2aa 127486Author: Peter Hutterer <peter.hutterer@who-t.net> 127487Date: Tue Jun 16 12:13:16 2009 +1000 127488 127489 Xi: namespace XI2 files. 127490 127491 Some files (notably those merged with MPX before XI2 came along) didn't use 127492 a 'xi' prefix. This patch changes all of them to meaningful names. 127493 127494 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127495 127496commit 4ddb002b6847e8a88e6f13ae0453a35ee79946ae 127497Author: Peter Hutterer <peter.hutterer@who-t.net> 127498Date: Mon Jun 15 15:14:47 2009 +1000 127499 127500 xfree86: fix two compiler warnings in xf86vmode.c 127501 127502 xf86vmode.c:1578: warning: pointer targets in passing argument 1 of 127503 ‘SwapShorts’ differ in signedness 127504 ../../../../include/misc.h:231: note: expected ‘short int *’ but argument is 127505 of type ‘CARD16 *’ 127506 127507 xf86vmode.c:1543: warning: unused variable ‘i’ 127508 127509 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127510 127511commit 9d039ef5047af2dedff0c583dda11032f71bed01 127512Author: Peter Hutterer <peter.hutterer@who-t.net> 127513Date: Mon Jun 15 15:11:55 2009 +1000 127514 127515 xfree86: silence a few int10 compiler warnings. 127516 127517 'format ‘%#x’ expects type ‘unsigned int’, but argument 2 has type ‘CARD32’' 127518 and similar. 127519 127520 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127521 127522commit fb882b3da189847bc98b646faddac9e8b42aa435 127523Author: Adam Jackson <ajax@redhat.com> 127524Date: Tue Jun 16 14:58:10 2009 -0400 127525 127526 EDID: Carp about 1.4 monitors with no preferred refresh rate 127527 127528 You would think, having finally tightened down the spec, that 127529 monitor vendors would bother to implement what the spec actually 127530 mandates. You would be wrong. 127531 127532commit b2d2af76b43d62c9be07fd736652f904632fd663 127533Author: Adam Jackson <ajax@redhat.com> 127534Date: Mon Jun 15 10:10:26 2009 -0400 127535 127536 randr: fix typo in swapped dispatch 127537 127538commit 1d20b9021e8aecb942619f64017cc2fc31f02d29 127539Author: Peter Hutterer <peter.hutterer@who-t.net> 127540Date: Sun Jun 14 18:21:43 2009 +1000 127541 127542 Xi: fix wrong bit->byte conversion in ProcXIQueryPointer 127543 127544 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127545 127546commit 9afc3241c157e5c19454f913c7105b3fdfa97ec4 127547Author: Peter Hutterer <peter.hutterer@who-t.net> 127548Date: Sun Jun 14 18:24:22 2009 +1000 127549 127550 Xi: remove un-used IsOn macro. 127551 127552 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127553 127554commit b67ff1d6e0d01170c6261593ca36d0f51d917393 127555Author: Peter Hutterer <peter.hutterer@who-t.net> 127556Date: Fri Jun 12 16:45:36 2009 +1000 127557 127558 Xi: XISelectEvents/XIGetSelectedEvents use 'win' instead of 'window' now. 127559 127560 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127561 127562commit b2bf67b61c564a4b92a429ca9ad455403161f33a 127563Author: Benjamin Defnet <benjamin.r.defnet@intel.com> 127564Date: Wed Jun 10 21:35:48 2009 -0700 127565 127566 randr: fix operation order so that rotation+transform works 127567 127568 The matrix multiply to combine rotation and projective transforms was being 127569 done in the wrong order. 127570 127571 Signed-off-by: Keith Packard <keithp@keithp.com> 127572 127573commit 3627215e97734cc367fff273a3358c2a95a53481 127574Author: Peter Hutterer <peter.hutterer@who-t.net> 127575Date: Fri Jun 5 16:38:46 2009 +1000 127576 127577 Add check for GNU ld in the test suite. 127578 127579 The GNU linker supports a -wrap option to wrap function calls at link-time. 127580 This allows for easy overriding of functions in the X server with stubs in 127581 the test suite. This functionality is only supported on the GNU linker and 127582 will be used extensively in the tests. Disable the tests if GNU ld is not 127583 available. 127584 127585commit 5e0ca6fabd12401de6bd6039484be2079e28851e 127586Author: Peter Hutterer <peter.hutterer@who-t.net> 127587Date: Thu Jun 11 16:26:23 2009 +1000 127588 127589 input: remove un-used "setter" argument from SetClientPointer. 127590 127591 It's obsolete, not likely to come back, let's drop it. 127592 127593 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127594 127595commit e6a18762ef113296c6a09833be70cb4b45aa3940 127596Author: Peter Hutterer <peter.hutterer@who-t.net> 127597Date: Thu Jun 11 15:40:56 2009 +1000 127598 127599 Xi: fix XISetClientPointer return values. 127600 127601 If SetClientPointer fails, the only reason may be that the device is not a 127602 pointer or that the device is an SD. Return BadDevice instead of BadAccess. 127603 (BadAccess is a leftover from the early times of the ClientPointer 127604 implementation when only one client was allowed to set it). 127605 127606 If the window parameter doesn't name a valid window or client, return 127607 BadWindow. 127608 127609 Finally, allow both master keyboards and master pointers as deviceid. 127610 127611 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127612 127613commit ae7dab2a136d15b976b956f68feec53886951dd6 127614Author: Peter Hutterer <peter.hutterer@who-t.net> 127615Date: Thu Jun 11 15:40:38 2009 +1000 127616 127617 Xi: Fix XISetClientPointer swapping. 127618 127619 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127620 127621commit 98e8ec8deb09db1c56e06afed267e6f517042fa3 127622Author: Peter Hutterer <peter.hutterer@who-t.net> 127623Date: Fri Jun 12 10:54:55 2009 +1000 127624 127625 Xi: sanitize ProcXIGetClientPointer. 127626 127627 This was quite old code and can be streamlined a bit. The new code is 127628 essentially the same as in ProcXISetClientPointer. 127629 127630 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127631 127632commit 96ea82fdac7c28feb2748cd4ff7faa7c00dbbdcd 127633Author: Peter Hutterer <peter.hutterer@who-t.net> 127634Date: Fri Jun 12 10:54:25 2009 +1000 127635 127636 Xi: fix reply swapping in XIGetClientPointer. 127637 127638 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127639 127640commit 76f25086a9ab6f0c27d9e2cd745f8463c6b3df5f 127641Author: Adam Jackson <ajax@redhat.com> 127642Date: Thu Jun 11 16:34:06 2009 -0400 127643 127644 vfb: Re-enable 30bpp support 127645 127646commit 0de58c88aba7ddd69b04f24ab5b2967c359aa69e 127647Author: Dave Airlie <airlied@redhat.com> 127648Date: Thu Jun 11 14:21:53 2009 +1000 127649 127650 xfree86: move didLock assignment down to where the function pointer is valid. 127651 127652 crtc->funcs->lock is NULL, so it's no use calling it here. Move it down so 127653 it's actually defined before we use it. 127654 127655 Introduced with 6f59a8160042ea145514fdcb410f17f33fd437c2. 127656 127657 Tested-by: Peter Hutterer <peter.hutterer@who-t.net> 127658 127659commit a85d210b34506cb39e7f08d81d4586fe4448b47f 127660Author: Peter Hutterer <peter.hutterer@who-t.net> 127661Date: Thu Jun 11 12:32:39 2009 +1000 127662 127663 Xi: store mask_len before swapping in ProcXIGetSelectedEvents. 127664 127665 Swapping the mask_len and then advancing the pointer by the swapped length 127666 is just a bad idea. 127667 127668 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127669 127670commit 5043f42f36b49e7702d0c496dd4fe157788d44ee 127671Author: Peter Hutterer <peter.hutterer@who-t.net> 127672Date: Thu Jun 11 12:31:40 2009 +1000 127673 127674 Xi: correct return buffer size for XIGetSelectedEvents. 127675 127676 The maximum number of bytes is calculated by the mask len, and the mask len 127677 is always in 4-byte units. XI2MASKSIZE however is in bytes. 127678 127679 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127680 127681commit fcf0c0b8f332cbde03c1a479376b7913197962eb 127682Author: Peter Hutterer <peter.hutterer@who-t.net> 127683Date: Thu Jun 11 13:23:40 2009 +1000 127684 127685 Xi: XISetEventMask needs to clear the mask if len is 0 127686 127687 zero-length masks are supposed to clear the device's mask. 127688 ProcXISelectEvents passes these masks through directly, so we need to clear 127689 the bits here if such a mask is supplied. 127690 127691 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127692 127693commit 22b4ac44e4fdbe8f36f2b570cdbdaf237062e33e 127694Author: Peter Hutterer <peter.hutterer@who-t.net> 127695Date: Thu Jun 11 11:08:07 2009 +1000 127696 127697 Xi: XIGetSelectedEvents mustn't returned masks from non-existing devices. 127698 127699 Or devices the client doesn't have XACE permissions for. 127700 127701 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127702 127703commit 7868956b0fb50a9d77506f01413612e429f540a3 127704Author: Peter Hutterer <peter.hutterer@who-t.net> 127705Date: Wed Jun 10 17:09:39 2009 +1000 127706 127707 Xi: ProcXIGetSelectedEvents must use WriteToClient for swapped data. 127708 127709 The data is already swapped before, so we just post it to the client as-is, 127710 without attempting to swap it again. 127711 127712 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127713 127714commit aa2babf11c30be4d289b58212dc330b84efb4053 127715Author: Peter Hutterer <peter.hutterer@who-t.net> 127716Date: Wed Jun 10 14:41:11 2009 +1000 127717 127718 input: remove dependency on XI2 protocol for XI_LASTEVENT. 127719 127720 inputstr includes XI2proto.h for the sole purpose of XI_LASTEVENT. 127721 However, using XI_LASTEVENT in the server is prone to errors, if the server 127722 is recompiled against a newer version of the protocol it would bump this 127723 variable and associates bits, including potential ABI. 127724 127725 This patch defines an XI2LASTEVENT for use in the server and removes the 127726 XI2proto.h require. XI2LASTEVENT is the current value of XI_LASTEVENT. 127727 127728 This patch is required by components that require access to inputInfo 127729 (currently xf86-video-geode and xf86-video-cirrus) but should not have a 127730 require for the XI2 protocol. 127731 127732 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127733 127734commit 77cc816da4a2777110182ed01c22c0f5e6ac3b65 127735Author: Peter Hutterer <peter.hutterer@who-t.net> 127736Date: Wed Jun 10 13:54:42 2009 +1000 127737 127738 Xi: rename ProcXISelectEvent to ProcXISelectEvents. 127739 127740 The request name has the plural, so let's do it here too. Purely cosmetic 127741 change. 127742 127743 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127744 127745commit 9563feeeb5309de1971c4e04419fed6fd31c9cea 127746Author: Peter Hutterer <peter.hutterer@who-t.net> 127747Date: Tue Jun 9 16:57:09 2009 +1000 127748 127749 Xi: start checking for invalid mask bits _after_ LASTEVENT. 127750 127751 Two issues that combined to false positives and false negatives. 127752 - The checking for invalid bits must be performed when there are enough bits 127753 that an event outside of LASTEVENT may be selected. 127754 - The first invalid bit is LASTEVENT + 1, not LASTEVENT. 127755 127756 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127757 127758commit e2fbaebb87e18198143a4b8e6be7e650563cd819 127759Author: Peter Hutterer <peter.hutterer@who-t.net> 127760Date: Tue Jun 9 14:38:25 2009 +1000 127761 127762 Xi: XISelectEvents returns BadValue for num_masks == 0. 127763 127764 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127765 127766commit cfeb65ac45b194dead5e8b9b26192df2619d6f9c 127767Author: Peter Hutterer <peter.hutterer@who-t.net> 127768Date: Tue Jun 9 14:37:42 2009 +1000 127769 127770 Xi: XISelectEvents needs to be at least size 3, not exactly. 127771 127772 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127773 127774commit f3c26034eca98924d93a004aaa1f61ffe5d344ca 127775Author: Peter Hutterer <peter.hutterer@who-t.net> 127776Date: Tue Jun 9 13:04:56 2009 +1000 127777 127778 Xi: XIQueryDevice should use XIFooClass instead of the old FooClass defines. 127779 127780 This is merely a cosmetic change, the actual values are the same anyway. 127781 127782 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127783 127784commit 482cc72aa790bd2f23c7ebe12c5549b47c737115 127785Author: Peter Hutterer <peter.hutterer@who-t.net> 127786Date: Tue Jun 9 13:03:38 2009 +1000 127787 127788 Xi: get the class length before swapping. 127789 127790 Advancing by the already-swapped length lets our pointers point into 127791 nirvana. 127792 127793 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127794 127795commit 9974249980894f74f3ead466655da87958a43670 127796Author: Peter Hutterer <peter.hutterer@who-t.net> 127797Date: Tue Jun 9 13:02:52 2009 +1000 127798 127799 Xi: Swapping 32 bit keycodes requires swapl, not swaps. 127800 127801 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127802 127803commit 918923e285f4e269a257bb5be4d3c8a50174aad0 127804Author: Dave Airlie <airlied@redhat.com> 127805Date: Wed Jun 10 15:13:45 2009 +1000 127806 127807 glx: fix open-coded linked list removal function 127808 127809 OMG stab stab stab, YALL. 127810 127811 removal function was made of crack, actually truncated the list from 127812 the one after the find point. 127813 127814 However fixing this makes glean makecurrent fail with a GLX error. 127815 127816commit 3ea747c0dbbec0db6761d66d4f6c680d2e9ddeaf 127817Author: Daniel Stone <daniel@fooishbar.org> 127818Date: Sat Apr 25 16:53:18 2009 +1000 127819 127820 KDrive: Warning fixes 127821 127822 xEvent vs. InternalEvent confusion still reigns though. 127823 127824 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 127825 127826commit 754be1e2ec9be2486bf45000d7244d217556de07 127827Author: Daniel Stone <daniel@fooishbar.org> 127828Date: Sat Apr 25 16:51:48 2009 +1000 127829 127830 KDrive: Xephyr: DRI: Warning fixes 127831 127832 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 127833 127834commit f534e6bea17746db952feb563ffea7320846b49d 127835Author: Daniel Stone <daniel@fooishbar.org> 127836Date: Sat Apr 25 15:38:54 2009 +1000 127837 127838 OS: Fix compile warnings 127839 127840 It's a marvel the sigaction() ever actually worked. 127841 127842 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 127843 127844commit 305ab237f666936cd812c464bf43f86f6079838e 127845Author: Daniel Stone <daniel@fooishbar.org> 127846Date: Sat Apr 25 22:11:19 2009 +1000 127847 127848 KDrive: Warning fixes and cleanups 127849 127850 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 127851 127852commit 4c8812b544ce5f319a7ac59a131c626c8509ef99 127853Author: Adam Jackson <ajax@redhat.com> 127854Date: Wed Jun 10 12:46:19 2009 -0400 127855 127856 kdrive: undef PSEUDO8 127857 127858 Whatever that was, it no longer is. 127859 127860commit 973ef5a7086aedb88a4888e73630c96ae68078e4 127861Author: Adam Jackson <ajax@redhat.com> 127862Date: Wed Jun 10 12:44:45 2009 -0400 127863 127864 kdrive: Remove a lie about PCMCIA support. 127865 127866commit e3c65cf1df9bbfb126f07b4b8e1254a855fb70c7 127867Author: Adam Jackson <ajax@redhat.com> 127868Date: Wed Jun 10 12:05:09 2009 -0400 127869 127870 xephyr: Add -title option. 127871 127872commit 803522300344265f11ea9f869ee6747730413bae 127873Author: Colin Harrison <colin.harrison@virgin.net> 127874Date: Wed Jun 10 13:32:54 2009 +0100 127875 127876 Xming: Simplify logic for ownership release of native clipboard 127877 127878 Make the logic simpler for the ownership release of the Windows 127879 clipboard in winclipboardwrappers.c - We've already marked a selection 127880 as unowned if it is owned by our clipboard window 127881 127882 Copyright (C) Colin Harrison 2005-2008 127883 http://www.straightrunning.com/XmingNotes/ 127884 http://sourceforge.net/projects/xming/ 127885 127886 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 127887 127888commit c7d3965bee0edb72468a0135aaa8a081348c2b23 127889Author: Jon TURNEY <jon.turney@dronecode.org.uk> 127890Date: Wed Jun 10 13:09:44 2009 +0100 127891 127892 Cygwin/X: Add a needed inputstr.h 127893 127894 Add a needed inputstr.h to fix build after commit 6d4ffcc9e066bc7c49dcff1a2ff0111801c7286d 127895 127896 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 127897 127898commit 6d4ffcc9e066bc7c49dcff1a2ff0111801c7286d 127899Author: Dave Airlie <airlied@redhat.com> 127900Date: Wed Jun 10 11:15:24 2009 +1000 127901 127902 input: move inputstr.h to where its needed. 127903 127904 This stops inputstr.h being needed to be included by output drivers. 127905 127906 Signed-off-by: Dave Airlie <airlied@redhat.com> 127907 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127908 127909commit 6f59a8160042ea145514fdcb410f17f33fd437c2 127910Author: Benjamin Defnet <benjamin.r.defnet@intel.com> 127911Date: Mon Jun 8 21:45:42 2009 -0700 127912 127913 hw/xf86/modes: Set crtc mode/rotation/transform before calling set_mode_major 127914 127915 This moves code out of each implementation of set_mode_major and back into 127916 the X server. The real feature here is that the transform is now available 127917 in the crtc for use by either xf86CrtcRotate or whatever the driver wants to 127918 do. Without this change, the transform was lost for drivers providing the 127919 set_mode_major interface. 127920 127921 Note that users of this API will want to stop smashing the transformPresent 127922 field, and could also stop setting mode/x/y/rotation for new enough X servers, 127923 but there's no reason to make that change as it will break things when 127924 running against older X servers. 127925 127926 Signed-off-by: Keith Packard <keithp@keithp.com> 127927 Acked-by: Daniel Stone <daniel@fooishbar.org> 127928 127929commit e244a5991e2cc55f5aa2f6e5255f1dabf56f0235 127930Author: Federico Mena Quintero <federico@novell.com> 127931Date: Tue Jun 9 13:28:28 2009 -0700 127932 127933 dix/randr: Add missing fields to SRR*NotifyEvent() 127934 127935 Also, remove redundant field swaps and make others match the order in which 127936 they are declared in the xRR*NotifyEvent structs. 127937 127938 Signed-off-by: Federico Mena Quintero <federico@novell.com> 127939 Signed-off-by: Keith Packard <keithp@keithp.com> 127940 127941commit 746e7b22e1bdfbf2363040367209b4c46b1c2fec 127942Author: Adam Jackson <ajax@redhat.com> 127943Date: Tue Jun 9 16:20:11 2009 -0400 127944 127945 cvt: Allow multiple-of-60Hz refresh rates for reduced blanking. 127946 127947 Doing so generates the same timings as given in the DMT spec for 127948 120Hz RB, so we should be set there. Other rates might be legal 127949 too but why push our luck. 127950 127951commit 2a035600e603117a4dbe668cad50bae2a5609094 127952Author: Peter Hutterer <peter.hutterer@who-t.net> 127953Date: Tue Jun 9 10:14:18 2009 +1000 127954 127955 require inputproto 1.9.99.11 127956 127957commit c5bebca46f63bcd7279f3c73ed1d50e0c1b6051f 127958Author: Peter Hutterer <peter.hutterer@who-t.net> 127959Date: Mon Jun 8 17:54:05 2009 +1000 127960 127961 Xi: hierarchy events have a num_info now instead of num_devices. 127962 127963 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127964 127965commit 810b74dbbcc1db3e68e0f95ff5223ddb2032e46d 127966Author: Peter Hutterer <peter.hutterer@who-t.net> 127967Date: Mon Jun 8 13:41:22 2009 +1000 127968 127969 Xi: change from XICreateMaster to XIAddMaster for consistency. 127970 127971 add/remove is used for slave devices and hierarchy flags. 127972 127973 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127974 127975commit 0e66a443a09176b8a875195d5473b953c07aa1d7 127976Author: Peter Hutterer <peter.hutterer@who-t.net> 127977Date: Mon Jun 8 12:24:24 2009 +1000 127978 127979 Xi: return BadValue for XI_HierarchyChangd mask on devices. 127980 127981 This mask may only be selected for XIAllDevices. 127982 127983 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 127984 127985commit 69a9545d1f8110841538410818df19fd960412c5 127986Author: Keith Packard <keithp@keithp.com> 127987Date: Thu May 28 14:43:27 2009 -0700 127988 127989 Make RANDR 'set' timestamps follow client specified time. Bug 21987. 127990 127991 The lastSetTime value which indicates when the configuration within the 127992 server was last changed was not getting set in the appropriate RandR 127993 requests. 127994 127995 Signed-off-by: Keith Packard <keithp@keithp.com> 127996 127997commit fa18c569ed26d2fdd442af6b39723cb7e30d18df 127998Author: Ben Skeggs <bskeggs@redhat.com> 127999Date: Tue Jun 9 09:14:50 2009 +1000 128000 128001 exa: driver pixmaps enabled if either CreatePixmap or CreatePixmap2 present 128002 128003commit 9390b7a133b0e17bc4d303effb7a99ff7017b086 128004Author: Peter Hutterer <peter.hutterer@who-t.net> 128005Date: Sun Jun 7 19:26:45 2009 +1000 128006 128007 dix: protect against missing ButtonClasses in GetPointerMapping. 128008 128009 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128010 128011commit 21ee53060b4dcb5c12074cf6682ff0067021ff60 128012Author: Simon Thum <simon.thum@gmx.de> 128013Date: Fri Jun 5 19:14:48 2009 +0200 128014 128015 config: fix build after XI2 API changes to RemoveDevice. 128016 128017 This patch fixes the build with --enable-config-dbus is enabled 128018 (default disabled). 128019 128020 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128021 128022commit 07c59411817c10fbcbf1bc9efa6a0818b6e0493a 128023Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128024Date: Sat Jun 6 22:17:44 2009 -0700 128025 128026 GLX: Make sure the types match for ALIAS in indirect_reqsize.c 128027 (cherry picked from commit 31a20a573b867fb7f02da15425e3b92eccb82dec) 128028 128029commit 077556ec39ca7e0d14486005cf71fcdc70a350cf 128030Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128031Date: Sat Jun 6 21:55:51 2009 -0700 128032 128033 XQuartz: AIGLX dispatch table cleanup 128034 128035 This should fix some issues when building on different versions of OSX. We only use extensions that OpenGL.framework's glext.h header tells us are supported. 128036 (cherry picked from commit e10c53421f1e780573c8adcd8ea89d6ffa28a819) 128037 128038commit fc24a97a8d022e986082b65b5133701e7273805c 128039Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128040Date: Tue Jun 2 19:12:41 2009 -0700 128041 128042 Rootless: Use serverClient instead of NullClient 128043 128044 This seems to stop the crash when moving windows. 128045 (cherry picked from commit fcd857c8e295ac7a2b0e58d49dc71251ed9d1266) 128046 128047commit 2429daa7e6b35a4a8032e1c3736e417a5a9baa07 128048Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128049Date: Tue Jun 2 10:19:10 2009 -0700 128050 128051 XQuartz: Default to forcing the dialog box before quitting. 128052 128053 This will show the quit dialog for users of other WMs. 128054 (cherry picked from commit 639ee0c913f892776fee40bcfb856e0d7c2c7db5) 128055 (cherry picked from commit 09432fa5b1cc8e9a30fb8aa4f8dfcd08c4a91e30) 128056 128057commit 6d75606a9e7f6ebc4d79a5261a48f5915c71cbcd 128058Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128059Date: Mon Jun 1 22:58:40 2009 -0700 128060 128061 Rootless: Despite its name, we still want to allow rootless to mess with the root window 128062 128063 This fixes the bug whereby we weren't drawing the contents of the root window since RootlessPaintWindowBackground got nuked. 128064 (cherry picked from commit bb0e208787739fbed2a0c8d1af6b6504319fdf88) 128065 128066commit 7348bf4935c8887bca33fe340ac2548f551bee3d 128067Author: Peter Hutterer <peter.hutterer@who-t.net> 128068Date: Sun Jun 7 09:25:51 2009 +1000 128069 128070 dix: set the generic event type for implicit XI2 grabs. 128071 128072 event->type is always GenericEvent for XI2 events. Instead, XI_ButtonPress 128073 (the generic event's evtype must be stored. 128074 128075 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128076 128077commit bb1c131b785ca30cf40041db3513b7669beebd34 128078Author: Simon Thum <simon.thum@gmx.de> 128079Date: Thu May 28 16:43:42 2009 +0200 128080 128081 dix: suppress pointer acceleration on xtest devices 128082 128083 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128084 128085commit 707b1241681c33401d5e0eb6edeace33f5970928 128086Author: Simon Thum <simon.thum@gmx.de> 128087Date: Sat Mar 21 18:19:19 2009 +0100 128088 128089 dix: prefer lroundf() over roundf() in axis scaling 128090 128091 it's unclear whether there actually is a problem, but in a very similar 128092 case there is (bug#21456). Also, integer addition is generally faster. 128093 128094 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128095 128096commit ef0e435052b2fda598cd8670c8e892aea685a102 128097Author: Jon TURNEY <jon.turney@dronecode.org.uk> 128098Date: Fri Jun 5 19:29:02 2009 +0100 128099 128100 Cygwin/X: Remove obsolete ChangeLog files 128101 128102 Remove obsolete ChangeLog files, this information is preserved in 128103 the automatically generated top-level ChangeLog 128104 128105commit 752508577aa7f0a1d80c8b7f4d3fe2b098e5e5fa 128106Author: Jon TURNEY <jon.turney@dronecode.org.uk> 128107Date: Tue Jun 2 21:19:46 2009 +0100 128108 128109 Cygwin/X: Tidy up a bodge to avoid collision between X header and native Win32 API definitions of ATOM 128110 128111 Remove a bodge to avoid collision between X header and the native 128112 Win32 API definitions of ATOM, use X11/Xwindows.h which exists to 128113 address these issues 128114 128115 Consequently, include misc.h in winclipboardthread.c, to provide the 128116 max() macro, as previously we were using the the one from windows.h 128117 128118 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128119 128120commit 7f3456a4080d1bf21010170e126128cb653bd2ac 128121Author: Peter Hutterer <peter.hutterer@who-t.net> 128122Date: Fri Jun 5 16:44:46 2009 +1000 128123 128124 Revert "test: check from INT_MIN to INT_MAX for core type conversion" 128125 128126 This patch requires extra special casing to check if the linker supports the 128127 -wrap option. Patches to do so will follow, in the meantime, revert this 128128 commit. It shouldn't have been pushed in the first place anyway. 128129 128130 This reverts commit d979f443946011158b6a183582728a6899c33b85. 128131 128132 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128133 128134commit d770d57f1701c7c8272b4ca3714f64c46c47adf1 128135Author: Peter Hutterer <peter.hutterer@who-t.net> 128136Date: Fri Jun 5 11:38:22 2009 +1000 128137 128138 Xi: fix XIQueryVersion major/minor swapping. 128139 128140 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128141 128142commit 17d62306b602ec61e8c5b44a79c9e1edc6b63a21 128143Author: Peter Hutterer <peter.hutterer@who-t.net> 128144Date: Fri Jun 5 11:31:39 2009 +1000 128145 128146 Xi: add XIQueryVersion reply swapping hook. 128147 128148 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128149 128150commit 2b459f44f3edaea137df9a28bc7adfeb1b9f1df7 128151Author: Peter Hutterer <peter.hutterer@who-t.net> 128152Date: Fri Jun 5 10:15:04 2009 +1000 128153 128154 Xi: return BadImplementation for deviceids 256 and above 128155 128156 The protocol allows for 16 bit device ids, but the implementation doesn't 128157 yet. We need to break the input ABI once more to shift the DeviceIntRec's 128158 CARD8 to a CARD16, along with some changes in the privates. 128159 128160 Once that is done, revert this patch. 128161 128162 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128163 128164commit c77f42049f77417c68bbce3dd2e995d916fc5af5 128165Author: Peter Hutterer <peter.hutterer@who-t.net> 128166Date: Fri Jun 5 09:29:04 2009 +1000 128167 128168 Xi: fix flawed logic in XIQueryVersion return value. 128169 128170 The server returned the minimum of major/minor each instead of the lower of 128171 the combined major.minor version. As a result, a client reporting 128172 3.0 and a server supporting 2.7, the return value would be 2.0 (the minimum 128173 of both). 128174 128175 Reported-by: Rémi Cardona 128176 Reported-by: Simon Thum 128177 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128178 128179commit b6c16fc7dad5a4ffcaaab647c0e3fabed372efd5 128180Author: Dave Airlie <airlied@redhat.com> 128181Date: Fri Jun 5 11:57:39 2009 +1000 128182 128183 crtc/gamma: check xf86_config is valid before using it. 128184 128185 If you have multiple cards, some that support randr 1.2 and some that don't 128186 you can get a null dereference in here. 128187 128188 Signed-off-by: Dave Airlie <airlied@redhat.com> 128189 128190commit da682abc78c9a5b2fde09679061fe08a3a1c665b 128191Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 128192Date: Thu Jun 4 17:33:25 2009 -0400 128193 128194 xace: fix a bad send access hook call. 128195 128196commit 993daf06497c85bb6a1e70592df380503d721dfb 128197Merge: 2bda50ee1 e1e8c7ddd 128198Author: Peter Hutterer <peter.hutterer@who-t.net> 128199Date: Thu Jun 4 15:11:36 2009 +1000 128200 128201 Merge branch 'master' into xi2 128202 128203commit 2bda50ee14b7873f28081bf0f1bfb7b79a40de43 128204Author: Peter Hutterer <peter.hutterer@who-t.net> 128205Date: Thu Jun 4 14:51:32 2009 +1000 128206 128207 Xi: memset the device name padding to 0. 128208 128209 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128210 128211commit 09266d6cd1bb6e09fe8bcb875c63fb4011c05b55 128212Author: Peter Hutterer <peter.hutterer@who-t.net> 128213Date: Thu Jun 4 13:36:29 2009 +1000 128214 128215 Xi: XIQueryVersion requires major_version of 2 or higher. 128216 128217 Otherwise, a BadValue error is sent. 128218 128219 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128220 128221commit 6ea76c550486bd17eb31f43757030fb44bdf1eab 128222Author: Peter Hutterer <peter.hutterer@who-t.net> 128223Date: Thu Jun 4 12:55:59 2009 +1000 128224 128225 Xi: reply with the XI2 version that is supported by both client and server. 128226 128227 XIQueryVersion must return the client's version if the client's version is 128228 lower than or equal to the server's version, or the server's version 128229 otherwise. 128230 128231 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128232 128233commit d979f443946011158b6a183582728a6899c33b85 128234Author: Peter Hutterer <peter.hutterer@who-t.net> 128235Date: Thu Jun 4 11:40:14 2009 +1000 128236 128237 test: check from INT_MIN to INT_MAX for core type conversion 128238 128239 ErrorF is link-wrapped to reduce the run-time of the test. 128240 128241 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128242 128243commit d8471bc7a19cd248fa3e0845496c86cd9251c013 128244Author: Peter Hutterer <peter.hutterer@who-t.net> 128245Date: Thu Jun 4 11:32:15 2009 +1000 128246 128247 test: adjust event_to_core_conversion test for new ProximityIn/Out behaviour 128248 128249 Both types now return BadMatch. 128250 128251 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128252 128253commit e1e8c7ddd7fdbfd674361364295fb1cbd5f28b45 128254Author: Adam Jackson <ajax@redhat.com> 128255Date: Wed Jun 3 13:37:32 2009 -0400 128256 128257 s/MIN/min/, s/MAX/max/ (#2968) 128258 128259commit 6de67d3206f0a4e307070714564c19efc84da2ec 128260Merge: af4e65840 7f027d9dc 128261Author: Peter Hutterer <peter.hutterer@who-t.net> 128262Date: Wed Jun 3 16:11:33 2009 +1000 128263 128264 Merge branch 'master' into xi2 128265 128266commit 7f027d9dc0146e229802aeac342ea2dbab63264a 128267Author: Olivier Blin <blino@mandriva.com> 128268Date: Thu May 28 20:57:36 2009 +0200 128269 128270 kdrive: add protocol mouse option 128271 128272 kdrive probes a lot of PS/2 protocols for the mouse device, which 128273 makes the mouse unusable for some seconds after X startup. 128274 This new "protocol" option allows forcing the mouse protocol. 128275 It can be used this way: 128276 Xfbdev -mouse mouse,,protocol=ps/2 -keybd keyboard 128277 128278 Signed-off-by: Olivier Blin <blino@mandriva.com> 128279 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128280 128281commit af4e658401ecbd2f2cc515b4c5abe5aee361cd64 128282Author: Peter Hutterer <peter.hutterer@who-t.net> 128283Date: Wed Jun 3 12:30:49 2009 +1000 128284 128285 mi: only process master if the SD is still attached. 128286 128287 SDs may be detached during event processing (e.g. if a passive grab 128288 activates). In this case, the event must not be processed through the master 128289 device. 128290 128291 Reported-by: Thomas Jaeger 128292 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128293 128294commit eb3fe72cbcd10c3953d713e07c321e695194a328 128295Author: Peter Hutterer <peter.hutterer@who-t.net> 128296Date: Wed Jun 3 12:19:14 2009 +1000 128297 128298 dix: set the device state for XI valuator events. 128299 128300 Valuator events need to include the device's state, while other device 128301 events need to include the state of the core devices. 128302 128303 Reported-by: Thomas Jaeger 128304 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128305 128306commit 08fd2ae652883393ecff6f3d5177e63408d58472 128307Author: Peter Hutterer <peter.hutterer@who-t.net> 128308Date: Wed Jun 3 12:18:17 2009 +1000 128309 128310 dix: use GetMaster in PickPointer and PickKeyboard. 128311 128312 GetMaster is more reliable than GetPairedDevice, it always returns the 128313 keyboard/pointer if desired, even if the wrong device was passed in. 128314 128315 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128316 128317commit a1d1dd06f8f91c1d1802f7581a4ba7568b85e69c 128318Author: Peter Hutterer <peter.hutterer@who-t.net> 128319Date: Wed Jun 3 12:15:32 2009 +1000 128320 128321 Xi: fix wrong grabtype from ProcXGrabDevice. 128322 128323 86077f0058ce88ee9b3df5d1ab854eeca43 switched from a boolean to a grabtype 128324 enum. ProcXGrabDevice didn't switch with it. PickPointer during an XI grab 128325 on a slave device would thus return a wrong (or NULL) device and crash the 128326 server. 128327 128328 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128329 128330commit a371a47205fc58a587e1aecaa4a031184481c228 128331Author: Peter Hutterer <peter.hutterer@who-t.net> 128332Date: Wed Jun 3 11:47:16 2009 +1000 128333 128334 dix: EventToCore must return BadMatch for proximity events. 128335 128336 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128337 128338commit 2bd5ea80652358565346071bed829070ddf5de85 128339Author: Peter Hutterer <peter.hutterer@who-t.net> 128340Date: Wed Jun 3 10:45:28 2009 +1000 128341 128342 dix: rework event conversion before delivery. 128343 128344 Let EventToXI, EventToCore and EventToXI2 return BadMatch if there's no 128345 matching event for this protocol spec. 128346 Adjust the delivery paths to cope with BadMatch errors (and clean them up on 128347 the way). 128348 128349 As a side-effect, this fixes server crashes on proximity events for a 128350 grabbed device. 128351 128352 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128353 128354commit 20d9f11abfd6ce3630d9640ff7383c7e27196956 128355Author: Adam Jackson <ajax@redhat.com> 128356Date: Tue Jun 2 13:49:09 2009 -0400 128357 128358 Typo fix 128359 128360commit ab5d1ae82ebd467d0a5b39821b286097865ba0c7 128361Author: Adam Jackson <ajax@redhat.com> 128362Date: Tue Jun 2 13:39:55 2009 -0400 128363 128364 Remove some libXfont leftovers 128365 128366commit 75795637c7160f1579dbe81c2d7600e85b1d141f 128367Author: Matthias Hopf <mhopf@suse.de> 128368Date: Tue Jun 2 18:41:36 2009 +0200 128369 128370 Revert "randr: Setting gamma: inverse logic looks more sane" 128371 128372 This reverts commit 79138eec1b49cbaca6a16f2bdd8579b5828aeb28. 128373 128374commit fc8cfc3a055d8af4ac809799c71f3db0d5246433 128375Author: Peter Hutterer <peter.hutterer@who-t.net> 128376Date: Tue Jun 2 16:53:41 2009 +1000 128377 128378 dix: ensure EventIsDeliverable has inputMasks set at all times. 128379 128380 For proximity events, the XI2 type is 0 and inputMasks never got set in the 128381 preceding condition. As a result, proximity events got never delivered. 128382 128383 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128384 128385commit ef9f851057a0ae214be02cb919f1e6634cead20b 128386Author: Peter Hutterer <peter.hutterer@who-t.net> 128387Date: Tue Jun 2 16:38:32 2009 +1000 128388 128389 dix: protect against floating slaves in DeviceFocusEvent. 128390 128391 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128392 128393commit b40289c8766c5e7543b7288009b4d965f57c04ce 128394Author: Peter Hutterer <peter.hutterer@who-t.net> 128395Date: Tue Jun 2 16:22:58 2009 +1000 128396 128397 xkb: allow pointer events to pass through for floating SDs without a key class. 128398 128399 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128400 128401commit 38df8351b36195070232b42a6d3369a9a08df8df 128402Author: Peter Hutterer <peter.hutterer@who-t.net> 128403Date: Mon Jun 1 17:09:03 2009 +1000 128404 128405 test: add grab matching tests. 128406 128407commit a66686a83edd6093ed540414dda2b8700e4696fe 128408Author: Peter Hutterer <peter.hutterer@who-t.net> 128409Date: Mon Jun 1 15:37:53 2009 +1000 128410 128411 input: add support for XIAllDevices and XIAllMasterDevices passive grabs. 128412 128413 These grabs are suported through two fake devices inputInfo.all_devices and 128414 inputInfo.all_master_devices. These devices are not part of the device list 128415 and are only initialised for their device id, nothing else. 128416 128417 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128418 128419commit f00cf7675164bb984ef310412f9e09582813adb1 128420Author: Peter Hutterer <peter.hutterer@who-t.net> 128421Date: Tue Jun 2 10:23:21 2009 +1000 128422 128423 dix: get the current MD keyboard for the grab modifier state. 128424 128425 A passive XI2 grab always uses the paired master device as a modifier 128426 device. After issuing a passive grab, the slave may be reattached to a 128427 different master and hence the modifier device may change. 128428 128429commit 1cd7ed75d903b5b0ea4e1070d2c26f4f931bdec7 128430Author: Peter Hutterer <peter.hutterer@who-t.net> 128431Date: Mon Jun 1 17:14:06 2009 +1000 128432 128433 dix: float slave devices during passive grabs. 128434 128435 Exception - implicit passive grabs. 128436 128437 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128438 128439commit f56ad22e8fabdca6d9fd1e408a539696e30599fc 128440Author: Peter Hutterer <peter.hutterer@who-t.net> 128441Date: Tue Jun 2 15:07:53 2009 +1000 128442 128443 dix: rework temporary slave detachment. 128444 128445 Rework addresses two issues: 128446 - storing the master device's pointer is a bad idea, we need to store the ID 128447 of the device in case it disappears during the grab. 128448 - restoring the old master did not actually reattach the device. Fixed now. 128449 128450commit b3e4810a283d369e1a2c6c3288177fe79fd1e9d6 128451Author: Peter Hutterer <peter.hutterer@who-t.net> 128452Date: Mon Jun 1 16:22:51 2009 +1000 128453 128454 dix: move grab device assignment into GRABTYPE_CORE block. 128455 128456 grab->type is the device type and XI2 types overlap with core events (being 128457 less than GenericEvent). Thus, for passive grabs the grab device would be 128458 overwritten with whatever device was activating it. 128459 128460 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128461 128462commit 18ae48670859d05788776e164c65985e3b35a723 128463Author: Peter Hutterer <peter.hutterer@who-t.net> 128464Date: Mon Jun 1 16:06:52 2009 +1000 128465 128466 dix: don't try to activate a passive grab on top of an active one. 128467 128468 If we already have an active grab on the device, don't attempt to activate 128469 another one. 128470 128471 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128472 128473commit ff1d6244eba8367a8d50601ca393e951c60c94b6 128474Author: Peter Hutterer <peter.hutterer@who-t.net> 128475Date: Mon Jun 1 12:14:52 2009 +1000 128476 128477 input: add support for XIAnyModifier in passive XI2 grabs. 128478 128479commit eb97d125fb6adeff0b23c2460108e0587ba253e2 128480Author: Adam Jackson <ajax@redhat.com> 128481Date: Mon Jun 1 15:50:50 2009 -0400 128482 128483 doc: Remove mention of Kerberos replay cache from Xserver.1 128484 128485 The krb5 auth code is gone, so this is no longer relevant. 128486 128487commit 80ea32ad4f1440d068e18db65830f45498703b09 128488Author: Peter Hutterer <peter.hutterer@who-t.net> 128489Date: Mon Jun 1 12:06:55 2009 +1000 128490 128491 dix: switch grab checking from unsigned shorts to unsigned ints 128492 128493 XIAnyModifier is outside of the unsigned short range. 128494 128495 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128496 128497commit 080b0331b3d1de1bab5ecdc9df175b287616e358 128498Author: Peter Hutterer <peter.hutterer@who-t.net> 128499Date: Mon Jun 1 13:33:29 2009 +1000 128500 128501 input: Add grabtype to GrabParameters. 128502 128503 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128504 128505commit bee02a154145c0e4ed97980f7857dc27a96b1be8 128506Author: Peter Hutterer <peter.hutterer@who-t.net> 128507Date: Mon Jun 1 10:49:22 2009 +1000 128508 128509 dix: ensure implicit passive grabs have the xi2mask zeroed. 128510 128511 This fixes XI2 enter/leave events being sent to non-xi2 clients. 128512 128513 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128514 128515commit cc71dfb4f5cf6113053951070dec9556e941aa24 128516Author: Peter Hutterer <peter.hutterer@who-t.net> 128517Date: Mon Jun 1 13:19:30 2009 +1000 128518 128519 test: Add GetSelectedEvents request to struct size tests. 128520 128521 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128522 128523commit de5c5b1db083a9c1d23e1461b54ca646c5a5c445 128524Author: Peter Hutterer <peter.hutterer@who-t.net> 128525Date: Mon Jun 1 13:21:36 2009 +1000 128526 128527 test: fix compile error introduced with the removal of isMaster. 128528 128529 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128530 128531commit 45530d16097459a756696e255ab4e72d6e51fbc4 128532Author: Adam Jackson <ajax@redhat.com> 128533Date: Fri May 29 18:07:48 2009 -0400 128534 128535 vfb: Fix depth setup. 128536 128537 Initialize the depth corresponding to the root window before the 128538 pixmap-only depths. Otherwise you end up with the root window depth in 128539 the depth list twice, which is mildly confusing for clients and 128540 catastrophically confusing for PanoramiXConsolidate(). 128541 128542commit 5c8540d8cf9ebaafa7572cb2879d057cd3d15190 128543Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128544Date: Fri May 29 14:49:05 2009 -0700 128545 128546 GLX: Purge glxint.h usage 128547 128548 Change (__GLXvisualConfig *) to (void *) in ABI compat stubs. 128549 (cherry picked from commit b4adec886c65e49cfbd30d1e3a41b613f75ed3f5) 128550 128551commit da2aaee24a77e1a9e5c623af4a78ca6acc94b787 128552Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128553Date: Fri May 29 14:27:46 2009 -0700 128554 128555 GLX: Purge some glxint.h usage in glxcmds.c 128556 128557 This is related to d0b249f1c5df81f3941d638f9625fe1e16c31807 in the 1.4 branch and was ok'd by krh 128558 (cherry picked from commit e587436cf2e1237813c924699ff0662aea3034e1) 128559 128560commit f2151f1b0dcbfec23488a29a7ea887fbfd53b760 128561Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128562Date: Fri May 29 14:09:41 2009 -0700 128563 128564 XQuartz: Fix alpha to be 1 (screenshots bug, etc) 128565 128566 This is related to: bc964ff1e3f3d8ab590eab40d767c6241e9480d4 128567 XQuartz: Stab at fixing the alpha 0/1 bug (screenshots, etc) by pulling in some old code that got gutted from rootless. 128568 128569 which was on the 1.4 branch and implemented in fbPaintWindow. Now that fbPaintWindow is gone, this is now in miPaintWindow(). 128570 (cherry picked from commit 032173f693f75385a40e144ffe2b7b5378bb6f2e) 128571 128572commit 7c158ffa766d32f5e8ec459693da5ef6ba0892e9 128573Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128574Date: Fri May 29 14:02:18 2009 -0700 128575 128576 Rootless: Use miPaintWindow since PaintWindowBackground is dead 128577 (cherry picked from commit 4975c087257b52189a5c21d93d121f183e60e4c8) 128578 128579commit ec9e9c8bcabda3d8b4ea07037fb247f7408dde8b 128580Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128581Date: Thu May 28 15:30:20 2009 -0700 128582 128583 XQuartz: English localization update for HIG compliance 128584 (cherry picked from commit 86b319d7cbef39d75b54c9d321e5dbf097b552be) 128585 128586commit 0f47203a9a690498c18b1c04e2ef5deed23a0e3e 128587Author: Peter Hutterer <peter.hutterer@who-t.net> 128588Date: Fri May 29 14:18:37 2009 +1000 128589 128590 dix: Set the implicit passive grab flag for keyboards too. 128591 128592 Extension devices have ActivateKeyboardGrab as their grab activation 128593 function, hence we need to ensure the implicit passive grab flag is set 128594 accordingly in the grab for further event delivery. 128595 128596 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128597 128598commit 75c73ddbfc9236bb80a3709d467674c4a1018fa3 128599Author: Peter Hutterer <peter.hutterer@who-t.net> 128600Date: Fri May 29 14:01:21 2009 +1000 128601 128602 dix: fix broken XI event delivery during grabs. 128603 128604 If an implicit passive grab is active, the XI event mask is in 128605 grab->deviceMask. Otherwise, for explicit grabs, the XI event mask is in 128606 grab->eventMask. 128607 128608 Reported-by: Thomas Jaeger 128609 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128610 128611commit 29ec14625111a8ae600d245ccbf905f059074d7c 128612Author: Peter Hutterer <peter.hutterer@who-t.net> 128613Date: Fri May 29 13:26:22 2009 +1000 128614 128615 Xi: passively grabbing slave devices must be possible. 128616 128617 There's use-cases where this is useful, so take the check out preventing 128618 that. 128619 128620 Reported-by: Thomas Jaeger 128621 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128622 128623commit 235cf5713a8fe535c439628cce9ea5b42432acb1 128624Author: Peter Hutterer <peter.hutterer@who-t.net> 128625Date: Fri May 29 12:09:06 2009 +1000 128626 128627 dix: init inputMasks to NULL to avoid random value dereference. 128628 128629 If the XI2 type is 0 (e.g. proximity events), inputMasks is never set and 128630 may not be NULL. 128631 128632 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128633 128634commit 68d86adbe8d84099812c37a04098f2ab5ccc45a4 128635Author: Peter Hutterer <peter.hutterer@who-t.net> 128636Date: Fri May 29 12:03:07 2009 +1000 128637 128638 dix: GetProximityEvents doesn't need extra events for valuators. 128639 128640 With internal events, we only have one event for all the data, no need to 128641 calculate for extra events. 128642 128643 Reported-by: Thomas Jaeger 128644 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128645 128646commit 71a83d68ab07536c0778d13cd7da974c40e77457 128647Author: Adam Jackson <ajax@redhat.com> 128648Date: Thu May 28 15:32:03 2009 -0400 128649 128650 xfree86: Remove loader symbol list ABI stubs 128651 128652 These have been nops since 2006 and functionally void since 7.0. 128653 128654commit d7aef3f663f4b5d861799f8615dcd301d5ce2906 128655Merge: a25f248fc e08c6a075 128656Author: Peter Hutterer <peter.hutterer@who-t.net> 128657Date: Thu May 28 17:20:58 2009 +1000 128658 128659 Merge branch 'master' into xi2 128660 128661 Conflicts: 128662 Xext/geext.c 128663 Xi/chdevcur.c 128664 Xi/extgrbdev.c 128665 Xi/xiproperty.c 128666 configure.ac 128667 dix/ptrveloc.c 128668 hw/xfree86/common/xf86Config.c 128669 mi/mipointer.h 128670 test/input.c 128671 xkb/xkb.c 128672 128673commit a25f248fc3bd0375d91ca8a44320200d445ecfbb 128674Author: Peter Hutterer <peter.hutterer@who-t.net> 128675Date: Wed May 27 15:47:12 2009 +1000 128676 128677 Xi: Send Enter or Leave events with XIPassive(Un)grabNotify 128678 128679 If a passive enter or focus in grab activates, send additional enter or 128680 focus events with mode XIPassiveGrabNotify to the grabbing client. 128681 Likewise, if the grab deactivates, send additional leave or focus out 128682 events. 128683 128684 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128685 128686commit 2107becb0ce2ffda001be65728c53563496d8d50 128687Author: Adam Jackson <ajax@redhat.com> 128688Date: Wed May 27 17:13:19 2009 -0400 128689 128690 xfree86: Remove unused (and useless) xf86{Add,Delete}ModuleInfo 128691 128692commit e08c6a0752772745f35f7afcf6c2b1c927b91cc0 128693Author: Federico Mena Quintero <federico@novell.com> 128694Date: Mon May 4 19:24:47 2009 -0500 128695 128696 randr: bug #21554 - re-probe outputs when coming back from laptop unsuspend 128697 128698 Make xf86 RANDR wrap the EnterVT call chain, so it can re-probe the 128699 outputs when a laptop comes back from suspend/unsuspend (actually, any 128700 time that we enter our VT again). The X server should then send RR* 128701 events to clients, so they can cope with a monitor that was unplugged 128702 while the laptop was suspended. 128703 128704 Signed-off-by: Federico Mena Quintero <federico@novell.com> 128705 128706commit b12010e10f38951358b48ff1076c026f943b7cc3 128707Merge: 6e6927247 f1441b835 128708Author: Søren Sandmann Pedersen <sandmann@redhat.com> 128709Date: Wed May 27 15:56:30 2009 -0400 128710 128711 Merge branch 'master' of git+ssh://sandmann@git.freedesktop.org/git/xorg/xserver 128712 128713commit 6e6927247359cc0db47501fb9d2dbb7856aa5985 128714Author: Søren Sandmann Pedersen <sandmann@redhat.com> 128715Date: Wed May 27 15:51:19 2009 -0400 128716 128717 Fix alpha map computation in miComputeCompositeRegion() 128718 128719 According to the RENDER spec, the origin of the alpha map is 128720 interpreted relative to the origin of the drawable of the image, not 128721 the origin of the drawable of the alpha map. 128722 128723 This commit fixes that and adds an alpha-test.c test program. 128724 128725 The only use of alpha maps I have been able to find is in Qt and they 128726 don't use a non-zero alpha origin. 128727 128728commit f1441b83538e6ba8a1557c4f5ea50c53a1837087 128729Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 128730Date: Wed May 27 10:06:10 2009 -0700 128731 128732 XQuartz: More localization updates 128733 (cherry picked from commit b107390c9f30571902556b4d053e62d345a06254) 128734 128735commit 996357e905c1082479bb238110b93bc170b8cb84 128736Author: Joe Krahn <jkrahn@nc.rr.com> 128737Date: Tue Jan 20 17:22:37 2009 +0000 128738 128739 Cygwin/X: Update icons directly, rather than modifying the window's class 128740 128741 Update icons directly to windows rather than modifying 128742 the window's class. Respect custom icons overriden via 128743 the configuration file. 128744 128745 fd.o bugzilla #4491 128746 128747 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128748 128749commit 9955252564922e7e8702dfb6eab52c75605a4a22 128750Author: Jon TURNEY <jon.turney@dronecode.org.uk> 128751Date: Mon Jan 5 17:55:58 2009 +0000 128752 128753 Cygwin/X: Remove unused TimeSinceLastInputEvent() 128754 128755 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128756 128757commit 966c8fccf16e77c8f459b4b175718323f2ec5d58 128758Author: Colin Harrison <colin.harrison@virgin.net> 128759Date: Mon Jan 5 18:04:42 2009 +0000 128760 128761 Xming: Fix some log spam 128762 128763 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128764 128765commit d190a27a33ab8c198568baf184fad14df57e24ad 128766Author: Colin Harrison <colin.harrison@virgin.net> 128767Date: Thu Jan 29 11:00:49 2009 +0000 128768 128769 Xming: Various tidy ups in winClipboardFlushXEvents() 128770 128771 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128772 128773commit fbe9ed27b1c2f932a74aa6ced134ecec021a3fc1 128774Author: Colin Harrison <colin.harrison@virgin.net> 128775Date: Thu Jan 29 10:55:38 2009 +0000 128776 128777 Xming: Avoid log spam if a windows application won't provide CF_UNICODETEXT clipboard format 128778 128779 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128780 128781commit 71ba9856a4f01aa7a42f1178c8da98a2e5ac23ae 128782Author: Colin Harrison <colin.harrison@virgin.net> 128783Date: Tue Feb 3 15:52:11 2009 +0000 128784 128785 Xming: Cache atom lookups in clipboard integration code 128786 128787 Cache the CLIPBOARD atom lookups in winClipboardWindowProc() 128788 Cache atom lookups in winClipboardFlushXEvents() 128789 Recache on server regeneration 128790 128791 Copyright (C) Colin Harrison 2005-2008 128792 http://www.straightrunning.com/XmingNotes/ 128793 http://sourceforge.net/projects/xming/ 128794 128795 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128796 128797commit 4ec110327bdc4f2617cd6116539812d7ef96b24e 128798Author: Colin Harrison <colin.harrison@virgin.net> 128799Date: Sun Jan 4 18:29:49 2009 +0000 128800 128801 Xming: Prevent the mouse wheel from stalling when another window is minimized. 128802 128803 Fix internal WM to prevent the mouse wheel from stalling when another window is minimized. 128804 128805 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128806 128807commit 888e6961a4ee76d05d212cfb946f089caafb3f69 128808Author: Jon TURNEY <jon.turney@dronecode.org.uk> 128809Date: Wed Jan 7 18:29:16 2009 +0000 128810 128811 Cygwin/X: Correctly allow for the native window frame width in ValidateSizing() 128812 128813 Fix internal WM to correctly calculate the native window border when validating window sizing 128814 128815 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128816 128817commit a72865868f03b675f86990476fcee601822894b3 128818Author: Joe Krahn <jkrahn@nc.rr.com> 128819Date: Tue Jan 20 17:09:43 2009 +0000 128820 128821 Cygwin/X: Consolidate dialog initialization in winInitDialog() 128822 128823 Changed windialogs.c to set icons via window properties rather than class 128824 properties, and use LoadImage() for small icons, because LoadIcon() can only open 128825 large icons. Since this code is redundant across the dialogs, I put it in the 128826 winCenterDialog function, along with a few other redundant instructions, and 128827 renamed in winInitDialog(). 128828 128829 Also, don't bogusly put our dialogs at the center of the virtual desktop if we 128830 are on a multimonitor system (this causes the dialog to end up split across two 128831 monitors in a dual-monitor side-by-side setup) 128832 128833 Corrections to use HWND_TOPMOST instead of HWND_TOP and not to use SWP_NOZORDER 128834 from Colin Harrison 128835 128836 fd.o bugzilla #4491 128837 128838 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128839 128840commit d053c3b35f730fdaddd22f841dbd8ed6c9a07ba0 128841Author: Colin Harrison <colin.harrison@virgin.net> 128842Date: Sun Nov 2 20:46:40 2008 +0000 128843 128844 Xming: Slightly adjust the formatting of the logged command line 128845 128846 Copyright (C) Colin Harrison 2005-2008 128847 http://www.straightrunning.com/XmingNotes/ 128848 http://sourceforge.net/projects/xming/ 128849 128850 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 128851 128852commit a7e23a79c1fc429aedbf9b6c0e78b1c8d7e02238 128853Author: Peter Hutterer <peter.hutterer@who-t.net> 128854Date: Tue May 26 14:42:25 2009 +1000 128855 128856 Xi: Add support for Enter and FocusIn grabs. 128857 128858 Enter grabs are checked for in CheckMotion(), each time the sprite window 128859 changes the current grab is deactivated (if applicable) and the new grab is 128860 activated (if applicable). Exception - if the grab is on a parent window of 128861 the current window since we keep the grab across descendants. 128862 128863 Since CheckMotion() may change the grab status of a device, we mustn't get 128864 "dev->deviceGrab.grab" in ProcessOtherEvents until after CheckMotion(). 128865 FocusIn grabs are checked in much the same manner. 128866 128867 The event delivery for grabs replaces the NotifyNormal on window change with 128868 a NotifyGrab on window change. Note that this happens before the grab 128869 activates, so the EnterNotify(NotifyGrab) is still delivered to the window, 128870 not to the grabbing client. This is in line with the core protocol semantics 128871 for NotifyGrab events. 128872 128873 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128874 128875commit ec2fe9660dbc0c16cdaca33b3b878011857e0fe2 128876Author: Peter Hutterer <peter.hutterer@who-t.net> 128877Date: Wed May 27 14:18:51 2009 +1000 128878 128879 Require inputproto 1.9.99.10 128880 128881commit 0ef6ba9f3b9e68307bfe623e12dc36d44440e97f 128882Author: Peter Hutterer <peter.hutterer@who-t.net> 128883Date: Wed May 27 11:15:54 2009 +1000 128884 128885 Xi: set the passive grab's resource mask on ungrabbing. 128886 128887 Not having the resource mask set means we never match an existing grab, 128888 hence we never actually ungrab. 128889 128890 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128891 128892commit 1f61d6647f25add487c5cae9739643a6e33ef9f7 128893Author: Peter Hutterer <peter.hutterer@who-t.net> 128894Date: Wed May 27 11:15:23 2009 +1000 128895 128896 Xi: use GetMaster(MASTER_KEYBOARD) to get the modifier device for XI2 pgrabs 128897 128898 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128899 128900commit bedfd561fa534b65baea45c4a05b1752ef172c45 128901Author: Peter Hutterer <peter.hutterer@who-t.net> 128902Date: Tue May 26 15:28:33 2009 +1000 128903 128904 dix: fix wrong cast of eventMask into CreateGrab. 128905 128906commit 023f959e6812e63f9d38b078158275bc13d238ae 128907Author: Peter Hutterer <peter.hutterer@who-t.net> 128908Date: Tue May 26 14:56:40 2009 +1000 128909 128910 Xi: fix an inputInfo.keyboard usage. 128911 128912 In the case of a RevertToFollowKeyboard, the master device should be used 128913 (since this is the closest equivalent to the VCK as before). Only if the 128914 master keyboard is the same as the device, revert to the VCK itself. 128915 128916commit b3463fd6a89d50a7562c357aeb5314ce0fd6493d 128917Author: Peter Hutterer <peter.hutterer@who-t.net> 128918Date: Tue May 26 13:58:04 2009 +1000 128919 128920 dix: convert window coordinates to FP1616 before calculating event_x/y 128921 128922 root_x/y are already in FP1616 at this point. 128923 128924 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128925 128926commit 10b5e981016132b4c07ad9ce68646d0a385e9c24 128927Author: Peter Hutterer <peter.hutterer@who-t.net> 128928Date: Tue May 26 11:12:08 2009 +1000 128929 128930 dix: take grabs into account for XI_Enter/Leave events. 128931 128932commit 1815defdb2f19e79f4ec0354d188ea763ff1f15e 128933Author: Peter Hutterer <peter.hutterer@who-t.net> 128934Date: Tue May 26 10:08:46 2009 +1000 128935 128936 Xi: only deactivate passive grabs if the event type matches. 128937 128938 This didn't use to be a problem when devices could only be pointers or 128939 keyboards, not both. Nowadays, slave devices may have both buttons and 128940 keyboards, and in this case we don't want to deactivate a passive keyboard 128941 grab when a button release is detected. 128942 128943commit 9dc3810a034f067096d71b33b0b47f1f52092fe6 128944Author: Peter Hutterer <peter.hutterer@who-t.net> 128945Date: Wed May 27 13:53:46 2009 +1000 128946 128947 dix: set the right grab event type for implicit passive grabs. 128948 128949 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128950 128951commit 6583477035234e23ead2fad9db7a07e5862447a4 128952Author: Nicolai Hähnle <nhaehnle@gmail.com> 128953Date: Sat May 23 13:35:24 2009 +0200 128954 128955 Remove reference to non-existing requestLog and requestLogIndex 128956 128957 These fields were removed in 252ec504817e05b185e4896a2d899e9c00b8aeef. 128958 128959 Signed-off-by: Nicolai Haehnle <nhaehnle@gmail.com> 128960 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128961 128962commit 53e821ab4a41cbc85932b9ef150311bc2e57c108 128963Author: Peter Hutterer <peter.hutterer@who-t.net> 128964Date: Mon May 25 12:20:37 2009 +1000 128965 128966 Xi: add request processing for XIGetSelectedEvents. 128967 128968 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128969 128970commit b377994336c978875dd786f4a4a9ef2e7b20e32c 128971Author: Peter Hutterer <peter.hutterer@who-t.net> 128972Date: Mon May 25 13:55:17 2009 +1000 128973 128974 Xi: advance by the right number of bytes when trawling XI2 event masks. 128975 128976 The wire layout is [struct xXIEventMask][mask bytes]. So the pointer needs 128977 to not only be advanced by the mask bytes, but also by the size of the 128978 struct. 128979 128980 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 128981 128982commit 12e725d08b4cf7dbb7f09b9ec09fa1b621156ea9 128983Author: Julien Cristau <jcristau@debian.org> 128984Date: Fri May 22 09:54:38 2009 +0200 128985 128986 randr: fix server crash in RRGetScreenInfo 128987 128988 We don't return rates to randr < 1.1 clients, so don't allocate space 128989 for them. This fixes a FatalError due to not all allocated space being 128990 used. 128991 128992 X.Org bug#21861 <http://bugs.freedesktop.org/show_bug.cgi?id=21861> 128993 128994 Reported-by: Guillaume Quintin <coincoin169g@gmail.com> 128995 Signed-off-by: Julien Cristau <jcristau@debian.org> 128996 128997commit c70511931ca1441878daed50e710fa514f6ab077 128998Author: Matthieu Herrb <matthieu.herrb@laas.fr> 128999Date: Sun May 24 10:32:38 2009 +0200 129000 129001 fix typo in cabff9007 which led to an unintialized memory read and a crash. 129002 129003commit 99e22b86c5f1a3653f3caaf01368a777d2b208d0 129004Author: Adam Jackson <ajax@redhat.com> 129005Date: Fri May 22 12:01:55 2009 -0400 129006 129007 EDID: Add modes from Established Timings III descriptor to mode pool 129008 129009 EDID 1.4, section 3.10.3.9 129010 129011commit b395da91c531d633ec47bb5a51e361d321663a3a 129012Author: Peter Hutterer <peter.hutterer@who-t.net> 129013Date: Fri May 22 15:24:46 2009 +1000 129014 129015 Xi: silence two compiler warnings 129016 129017 chdevcur.c:97: warning: ‘SecurityLookupIDByType’ is deprecated (declared at 129018 ../include/resource.h:269) 129019 129020 xiproperty.c:200: warning: passing argument 2 of ‘GetEventFilter’ from 129021 incompatible pointer type 129022 129023 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129024 129025commit 4a36db60774640ea92a2133a3b08fbf1a48d2cb6 129026Author: Peter Hutterer <peter.hutterer@who-t.net> 129027Date: Fri May 22 15:16:43 2009 +1000 129028 129029 os: fix compiler warning "too few arguments to format" 129030 129031 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129032 129033commit da0d3baf71b34657cc235d6c6b37d548541f9449 129034Author: Peter Hutterer <peter.hutterer@who-t.net> 129035Date: Thu May 21 14:11:27 2009 +1000 129036 129037 dix: protect event generation against single-valuator devices. 129038 129039 If we have a single-axis device and it sends events it should not access 129040 non-existant memory. 129041 129042commit b25e29e8014927815d4fdc9c416bfeb2af0ecba6 129043Author: Peter Hutterer <peter.hutterer@who-t.net> 129044Date: Thu May 21 12:07:03 2009 +1000 129045 129046 mi: use GetMaster() from MIPOINTER and MISPRITE. 129047 129048 Both may in some cases be called for a SD attached to a master keyboard. In 129049 this case, we need to get the right master device (i.e. the pointer). 129050 129051 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129052 129053commit c9df51b070377ca33e48644dbc842b6e3dfb2975 129054Author: Peter Hutterer <peter.hutterer@who-t.net> 129055Date: Thu May 21 11:16:37 2009 +1000 129056 129057 input: allow for master pointers to not have a button class. 129058 129059 There's devices (e.g. some barcode readers) that have axes but no buttons. 129060 When such a device sends a motion event, the valuator and button class is 129061 copied into the master pointer (i.e. removing the button class). 129062 So we need a couple of extra sanity checks for the button class to exist. 129063 129064 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129065 129066commit ebe45e1a72a81ad22413e5d0514869b3e45a69b7 129067Author: Peter Hutterer <peter.hutterer@who-t.net> 129068Date: Wed May 20 10:55:03 2009 +1000 129069 129070 input: introduce partial class copying depending on the event. 129071 129072 Copying all classes into the master device has drawbacks for hybrid devices 129073 (devices that are both mice and keyboards). If such a device posts an event, 129074 it's key classes are moved into the VCP. The key event itself is unaffected 129075 by keyboard grabs and the like. 129076 129077 Partial class copying copies depending on the event and copies the classes 129078 into the right master device (i.e. the VCK for key events, the VCP for 129079 pointer events). 129080 129081 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129082 129083commit bc63c8a4570c989f19a036965854bceb9800ce19 129084Author: Peter Hutterer <peter.hutterer@who-t.net> 129085Date: Wed May 20 12:33:49 2009 +1000 129086 129087 dix: introduce GetMaster() 129088 129089 For hybrid devices (keys + buttons/axes) the attached master device is 129090 generally the wrong one. One shouldn't post a button event through a 129091 keyboard and vice versa. 129092 129093 GetMaster(dev) returns the right master device for the given type needed. 129094 This may be the MD paired with this device's MD. 129095 129096 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129097 129098commit d79318f269d959d566ec66239b4c985afd61b259 129099Author: Peter Hutterer <peter.hutterer@who-t.net> 129100Date: Thu May 21 08:29:14 2009 +1000 129101 129102 dix: Add a deviceid to the DeviceChangedEvent. 129103 129104 ChangeDeviceId would actually overwrite the flags field if deviceid wasn't 129105 present. Aside from the event of course not telling which device generated 129106 it in the first place. 129107 129108 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129109 129110commit eeffb83fea2d87070beafd0c692fb0ad54d8335b 129111Author: Peter Hutterer <peter.hutterer@who-t.net> 129112Date: Mon May 18 20:03:07 2009 +1000 129113 129114 ABI_XINPUT_VERSION 6 129115 129116commit 438a4eafa5d14ab676827f7a5d58782c02db4cdf 129117Author: Peter Hutterer <peter.hutterer@who-t.net> 129118Date: Wed May 20 17:02:50 2009 +1000 129119 129120 input: remove nested union from InternalEvent. 129121 129122 There's no need for internal events to be a struct with a single nested 129123 union, we might as well make the union itself the InternalEvent. 129124 129125 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129126 129127commit b12d302df8283186ce87882c29b2b0294adb2770 129128Author: Peter Hutterer <peter.hutterer@who-t.net> 129129Date: Wed May 20 15:28:16 2009 +1000 129130 129131 Input: rename DeviceIntRec->isMaster to ->type. 129132 129133 isMaster is not enough as long as we differ between master pointers and 129134 keyboard. With flexible device classes, the usual checks for whether a 129135 master device is a pointer (currently check for ->button, ->valuators or 129136 ->key) do not work as an SD may post an event through a master and mess this 129137 check up. 129138 129139 Example, a device with valuators but no buttons would remove the button 129140 class from the VCP and thus result in the 129141 IsPointerDevice(inputInfo.pointer) == FALSE. 129142 129143 This will become worse in the future when new device classes are introduced 129144 that aren't provided in the current system (e.g. a switch class). 129145 129146 This patch replaces isMaster with "type", one of SLAVE, MASTER_POINTER and 129147 MASTER_KEYBOARD. All checks for dev->isMaster are replaced with an 129148 IsMaster(dev). 129149 129150commit add2defac7e59dcb355e5ba12b6de60861741d57 129151Author: Peter Hutterer <peter.hutterer@who-t.net> 129152Date: Wed May 20 16:23:38 2009 +1000 129153 129154 Split the signal-handler's lastSlave out into a separate variable. 129155 129156 dev->u.lastSlave was not signal safe since it was accessed by the DIX and 129157 during signal handling. 129158 Replaced with: 129159 'dev->last.slave' for the signal handler's lastSlave (used to generate 129160 DeviceChangedEvents), . 129161 'dev->u.lastSlave' for the DIX lastSlave (currently only used in 129162 change_modmap) 129163 129164 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129165 129166commit 525aa17f804d37d1cfcbbf6b8e6cddb45e999b20 129167Author: Tomas Janousek <tomi@nomi.cz> 129168Date: Wed May 20 15:03:01 2009 +0200 129169 129170 Bug #6428, #16458, #21464: Fix crash due to uninitialized VModMap fields. 129171 129172 In ProcXkbGetKbdByName, mrep.firstVModMapKey, .nVModMapKeys and 129173 .totalVModMapKeys were not initialized, contained random values and caused 129174 accesses to unallocated and later modified memory, causing 129175 XkbSizeVirtualModMap and XkbWriteVirtualModMap to see different number of 129176 nonzero values, resulting in writes past the end of an array in XkbSendMap. 129177 129178 This patch initializes those values sensibly and reverts commits 5c0a2088 and 129179 6dd4fc46, which have been plain non-sense. 129180 129181 Signed-off-by: Tomas Janousek <tomi@nomi.cz> 129182 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129183 129184commit 7db55a0806c82bd4143c8bf1b8eb2b62e456ad9a 129185Author: Peter Hutterer <peter.hutterer@who-t.net> 129186Date: Wed May 20 14:38:25 2009 +1000 129187 129188 dix: remove superfluous loop in change_modmap. 129189 129190 A device can only be attached to a single master device. So instead of 129191 looping and searching for the master device, we can just use dev->u.master 129192 directly. 129193 129194 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129195 129196commit cabff9007a4abad026b450a4aab155c7bcd94326 129197Author: Peter Hutterer <peter.hutterer@who-t.net> 129198Date: Fri May 15 08:33:07 2009 +1000 129199 129200 xfree86: treat other drivers as mouse drivers in the config. 129201 129202 Historically, if no input device was referenced in the ServerLayout, 129203 the server would pick the first "mouse" device found in the xorg.conf. 129204 This patch gives evdev, synaptics, vmmouse and void the same status. If 129205 there is a section in the config file using this driver - use it as the core 129206 pointer. 129207 129208 Device selection is in driver-order, not in config-order. If a "mouse" 129209 device is listed after a "synaptics" device, the "mouse" device gets 129210 preference. This replicates the original behaviour. 129211 129212 This code only takes effect if AllowEmptyInput is off and there is no core 129213 pointer in the server layout. 129214 129215 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129216 129217commit a8bd1e1f96d8e5380972a7dce8d9940cd912aa09 129218Author: Søren Sandmann Pedersen <ssp@l3000.localdomain> 129219Date: Thu May 21 18:21:28 2009 -0400 129220 129221 Fix alpha map computation in miComputeCompositeRegion() 129222 129223 According to the RENDER spec, the origin of the alpha map is 129224 interpreted relative to the origin of the drawable of the image, not 129225 the origin of drawable of the alpha map. 129226 129227 The only use of alpha maps I have been able to find is in Qt and they 129228 don't use a non-zero alpha origin. 129229 129230commit a2c5ee36b21c2ee5c0468f1b251e74c1412dbecb 129231Author: Adam Jackson <ajax@redhat.com> 129232Date: Thu May 21 10:20:48 2009 -0400 129233 129234 EDID: Be more cautious about finding vendor blocks. 129235 129236 Many old monitors zero-fill the detailed descriptors, so check for that 129237 to avoid a useless warning like: 129238 129239 (WW) RADEON(0): Unknown vendor-specific block 0 129240 129241commit 73c7398aafa050950ad614483fb0d5bdd2da97ca 129242Author: Peter Hutterer <peter.hutterer@who-t.net> 129243Date: Wed May 20 10:40:50 2009 +1000 129244 129245 dix: 'namespace' HAS_OLD_SLAVE and HAS_NEW_SLAVE. 129246 129247 We need more flags for this in the near future, so let's namespace them now. 129248 129249commit 5b5e3fa2771383a85afff679be34df19d3a4e290 129250Author: Peter Hutterer <peter.hutterer@who-t.net> 129251Date: Fri May 15 08:33:07 2009 +1000 129252 129253 xfree86: treat other drivers as mouse drivers in the config. 129254 129255 Historically, if no input device was referenced in the ServerLayout, 129256 the server would pick the first "mouse" device found in the xorg.conf. 129257 This patch gives evdev, synaptics, vmmouse and void the same status. If 129258 there is a section in the config file using this driver - use it as the core 129259 pointer. 129260 129261 Device selection is in driver-order, not in config-order. If a "mouse" 129262 device is listed after a "synaptics" device, the "mouse" device gets 129263 preference. This replicates the original behaviour. 129264 129265 This code only takes effect if AllowEmptyInput is off and there is no core 129266 pointer in the server layout. 129267 129268 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129269 129270commit 1cce55cc0333a026474100cf2f784e220fd96473 129271Author: Peter Hutterer <peter.hutterer@who-t.net> 129272Date: Wed May 20 14:07:03 2009 +1000 129273 129274 input: rename device->type to device->xinput_type. 129275 129276 This type is only used in XI to give a hint of what type this device may be. 129277 Call it xinput_type for clarity. 129278 129279 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129280 129281commit 81b3b0cce088866dc3cda099d7c8d6655849fd43 129282Author: Tomas Janousek <tomi@nomi.cz> 129283Date: Wed May 20 15:03:01 2009 +0200 129284 129285 Bug #6428, #16458, #21464: Fix crash due to uninitialized VModMap fields. 129286 129287 In ProcXkbGetKbdByName, mrep.firstVModMapKey, .nVModMapKeys and 129288 .totalVModMapKeys were not initialized, contained random values and caused 129289 accesses to unallocated and later modified memory, causing 129290 XkbSizeVirtualModMap and XkbWriteVirtualModMap to see different number of 129291 nonzero values, resulting in writes past the end of an array in XkbSendMap. 129292 129293 This patch initializes those values sensibly and reverts commits 5c0a2088 and 129294 6dd4fc46, which have been plain non-sense. 129295 129296 Signed-off-by: Tomas Janousek <tomi@nomi.cz> 129297 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129298 129299commit 79138eec1b49cbaca6a16f2bdd8579b5828aeb28 129300Author: Matthias Hopf <mhopf@suse.de> 129301Date: Tue Mar 10 16:03:26 2009 +0100 129302 129303 randr: Setting gamma: inverse logic looks more sane 129304 129305commit c2785ae7eb6197bbfc75e92e99fffbb8ad8064da 129306Author: Peter Hutterer <peter.hutterer@who-t.net> 129307Date: Wed May 20 14:29:13 2009 +1000 129308 129309 dix: refuse events from disabled devices. 129310 129311 If the device is disabled ("off"), it must not send events to a client. 129312 The driver shouldn't send events in that case anyway, but just to make sure 129313 we simply drop events coming while the device is disabled. 129314 129315 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129316 129317commit 47a89b1cba67a43ddc225f7e1b5d2b3b217c8d93 129318Author: Peter Hutterer <peter.hutterer@who-t.net> 129319Date: Wed May 20 13:59:22 2009 +1000 129320 129321 Xi: remove DeviceIsPointerType 129322 129323 This approach is broken anyway. DIPT only checked for the XInput type 129324 "MOUSE" and the only user of this is xf86ActivateDevice when it sets the 129325 Activate/DeactivateGrab functions. 129326 Since synaptics and wacom set their own types, evdev only sets MOUSE for, 129327 well, mice half the devices didn't have this set correctly anyway. 129328 129329 Instead, ActivatePointerGrab should be merged together with 129330 ActivateKeyboardGrab. 129331 129332 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129333 129334commit 7b891e450f005688e41e61016e6d3e394c67f6be 129335Author: Peter Hutterer <peter.hutterer@who-t.net> 129336Date: Wed May 20 14:23:16 2009 +1000 129337 129338 dix: Remove obsolete comment 129339 129340 CorePointerProc doesn't back up the device classes anymore. 129341 129342commit 42719ce5c715ac3a0c74c3c7df872d35e0449de9 129343Author: Peter Hutterer <peter.hutterer@who-t.net> 129344Date: Wed May 20 14:38:25 2009 +1000 129345 129346 dix: remove superfluous loop in change_modmap. 129347 129348 A device can only be attached to a single master device. So instead of 129349 looping and searching for the master device, we can just use dev->u.master 129350 directly. 129351 129352 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129353 129354commit 22a33a7250f879b8f3eb84176a1dc57027d8cfe4 129355Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 129356Date: Tue May 19 22:45:26 2009 -0400 129357 129358 glx: Register names for the GLX resource types. 129359 129360 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 129361 129362commit 77e90261d5f91e1170c6e02deb0e3739c24fcfd6 129363Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 129364Date: Tue May 19 19:36:03 2009 -0400 129365 129366 dix/events.c: Use wClient where appropriate. 129367 129368 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 129369 129370commit 3cea176d5abcb0f14eefbdcbe17fed0847524dd4 129371Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 129372Date: Tue May 19 19:30:33 2009 -0400 129373 129374 xace: Fix a bad device access hook call. 129375 129376 Add a proper access mode, and reverse the logic of the return value. 129377 Zero ("Success") is returned on success from the hook calls. 129378 129379 Signed-off-by: Eamon Walsh <ewalsh@tycho.nsa.gov> 129380 129381commit 4addfcd633484b7f53b5df1b92bfca8bfc1dc48d 129382Author: Michel Dänzer <daenzer@vmware.com> 129383Date: Mon May 18 17:53:35 2009 +0200 129384 129385 EXA: Allocate from the end of free offscreen memory rather than from the start. 129386 129387 This way we don't always need to scan over previously allocated areas when 129388 looking for an available one, and there might be less fragmentation. 129389 129390 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 129391 129392commit 510cbd43cd4e34bd459e8f74ab2855714b4ca95d 129393Author: Michel Dänzer <daenzer@vmware.com> 129394Date: Mon May 18 17:48:57 2009 +0200 129395 129396 EXA: Defragment offscreen memory. 129397 129398 At most once per second, under the following circumstances: 129399 129400 * We can't satisfy an offscreen memory allocation, but there seems to be enough 129401 offscreen memory available in total. 129402 129403 or 129404 129405 * The server has been idle for at least 100ms, and there is more than one 129406 available offscreen area. 129407 129408 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 129409 129410commit 8331bde0adeccefb275c4d707e7b2cb1d95b1581 129411Author: Jon TURNEY <jon.turney@dronecode.org.uk> 129412Date: Fri May 15 21:36:28 2009 +0100 129413 129414 Cygwin/X: link with libmain.a to provide main() 129415 129416 Commit 987579c930bda803427a28cb82773c389f5110d6 moves main.c into 129417 into libmain.a to be linked with separately, and updates the various 129418 DDXs to link with libmain.a, except Xwin, which also needs this change. 129419 129420 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 129421 129422commit b89dcfbfbd38f8d19f1d4f24f81820b1048572c8 129423Author: Peter Hutterer <peter.hutterer@who-t.net> 129424Date: Mon May 18 16:07:08 2009 +1000 129425 129426 Xi: fix length field in XIQueryPointer. 129427 129428 Buttons append 4-byte units to the end of the reply, they need to be 129429 included too. 129430 129431 Reported-by: Mark Dokter 129432 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129433 129434commit 33fcaaaea5ea30be39156ecfdbcca891b47ab465 129435Author: Peter Hutterer <peter.hutterer@who-t.net> 129436Date: Mon May 18 16:05:35 2009 +1000 129437 129438 Fix missing parentheses in FP1616 macro. 129439 129440 Missing parens led to interesting results if an expression instead of a 129441 constant was passed in (ProcXIQueryPointer for example). 129442 129443commit 6258bb1d6bb78ea4ba1d33470bc74e9b711b25f2 129444Author: Peter Hutterer <peter.hutterer@who-t.net> 129445Date: Fri May 15 20:24:45 2009 +1000 129446 129447 require inputproto 1.9.99.9 129448 129449commit e7e94ff5fd3205a9ee01fe9342d9f3a426b0a068 129450Author: Peter Hutterer <peter.hutterer@who-t.net> 129451Date: Thu May 14 21:45:35 2009 +1000 129452 129453 Xi: set the right length for the XIQueryPointer reply. 129454 129455commit 00ab04d73c612ac85e024105015bdb7c79e8874b 129456Author: Peter Hutterer <peter.hutterer@who-t.net> 129457Date: Thu May 14 16:24:35 2009 +1000 129458 129459 Include full modifier + button state in XIQueryPointer. 129460 129461commit de78aab0ea3a2d8c61222493f25911258dea65a2 129462Author: Peter Hutterer <peter.hutterer@who-t.net> 129463Date: Thu May 14 15:53:00 2009 +1000 129464 129465 Xi: take the paired device as modifier device only for MD pointers. 129466 129467commit 0e089f9ceb0edec532078c6e1df7f642f7220fb5 129468Author: Peter Hutterer <peter.hutterer@who-t.net> 129469Date: Thu May 14 15:29:20 2009 +1000 129470 129471 Xi: check cursor and grab_window before attempting the passive grab. 129472 129473 Both values need to fail the request, so we need to check them before 129474 looping through the modifier masks. Otherwise, a wrong grab_window will 129475 show up in each failed modifier return but not actually cause an error. 129476 129477commit 5c3e7289801c958e2c70e2629d3ebee122449be8 129478Author: Peter Hutterer <peter.hutterer@who-t.net> 129479Date: Wed May 13 16:54:01 2009 +1000 129480 129481 dix: count must be 1 if we have an XI2 match. 129482 129483 This fixes events that activated a grab not getting delivered to the client. 129484 129485 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129486 129487commit e083a0d0692baa5ad29002a38293b36dd3b4c06f 129488Author: Peter Hutterer <peter.hutterer@who-t.net> 129489Date: Wed May 13 15:23:42 2009 +1000 129490 129491 Xi: set the correct length for XIPassiveGrab replies. 129492 129493commit 8ff1bff8f462b15e05fc298704e1e40cb244587f 129494Author: Peter Hutterer <peter.hutterer@who-t.net> 129495Date: Tue May 12 21:33:54 2009 +1000 129496 129497 Update to new XI2 names 129498 129499commit bbf266cca3de726e3eb932d66dd5b8981ef40fa8 129500Author: Peter Hutterer <peter.hutterer@who-t.net> 129501Date: Tue May 12 21:24:07 2009 +1000 129502 129503 mi: un-deprecate miPointerWarpCursor 129504 129505 This function was deprecated in ef68273f5bdb27a492ec0b69548ec4fbede46c08 129506 because it didn't take a device argument. The device argument was added in 129507 1c7568b8a1417257fa67c7fca69aa253099b9461 though, so the deprecation is 129508 obsolete. 129509 129510 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129511 129512commit 273890924b8ed6f8b7949c0322c8258b9e6f8630 129513Author: Peter Hutterer <peter.hutterer@who-t.net> 129514Date: Tue May 12 16:09:35 2009 +1000 129515 129516 input: reduce the number of superfluous hierarchy events 129517 129518 There's only two reasons for hierarchy events: 129519 - device is added, removed, etc. In this case we want to send the event as 129520 it happens. 129521 - devices are added in a XIChangeDeviceHierarchy request. In this case we 129522 only want one event cumulating all changes. 129523 129524commit 033a2b12fcd02fa9a2c2f20a352bec0a43074512 129525Author: Peter Hutterer <peter.hutterer@who-t.net> 129526Date: Tue May 12 14:50:57 2009 +1000 129527 129528 Xi: set per-device hierarchy changed flags. 129529 129530 Rather than have one field per hierarchy change, XI2 has two fields - one 129531 generic one and one per-device that include the device-specific flags. 129532 This requires some funky handling for removed devices, but oh well. 129533 129534commit 8fb51feae222ff4f4aa1c440b6973ce7383bbc09 129535Author: Peter Hutterer <peter.hutterer@who-t.net> 129536Date: Mon May 11 15:45:46 2009 +1000 129537 129538 xkb: if kbd init failed, NULL out the pointers after freeing them (#21278) 129539 129540 Reproducible: 129541 Configure a server that uses the keyboard driver with an invalid ruleset, 129542 e.g. (Option "XkbRules" "foobar"). Ensure that Option "AllowEmptyInput" is 129543 "off" in the ServerFlags or ServerLayout section. Start the server. 129544 After failing to init the keymap, the server will try to clean up after the 129545 device, double-freeing some xkb structs that have not been reset properly. 129546 129547 X.Org Bug 21278 <http://bugs.freedesktop.org/show_bug.cgi?id=21278> 129548 129549 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129550 129551commit 0cea199eda772a5ddd5ce72e4823a00ec14405d4 129552Author: Peter Hutterer <peter.hutterer@who-t.net> 129553Date: Mon May 11 12:53:56 2009 +1000 129554 129555 kdrive: set Activate/Deactivate grab for input devices (#21591) 129556 129557 X.Org Bug 21591 <http://bugs.freedesktop.org/show_bug.cgi?id=21591> 129558 129559commit fdce58ca822f4e43e03e84076190f52d7104e01f 129560Author: Peter Hutterer <peter.hutterer@who-t.net> 129561Date: Mon May 11 12:51:40 2009 +1000 129562 129563 dix: ensure Activate/DeactivateGrab has a valid value. 129564 129565 Xephyr doesn't manually set Activate/DeactivateGrab for new devices, 129566 resulting in a NULL-pointer dereference later when a grab is activated. 129567 Avoid the segfault by ensuring that the pointer is always valid. 129568 129569 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129570 129571commit e3f296d91dfe6b827195e1d387e1a04aa73b85c3 129572Author: Peter Hutterer <peter.hutterer@who-t.net> 129573Date: Sat May 9 10:33:20 2009 +1000 129574 129575 xfree86: Remove superfluous ifdef DEBUG checks. 129576 129577 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129578 129579commit 2f6253376df45e508284ff7766b186cfc7fb56d6 129580Author: Peter Hutterer <peter.hutterer@who-t.net> 129581Date: Sat May 9 10:33:20 2009 +1000 129582 129583 xfree86: Remove superfluous ifdef DEBUG checks. 129584 129585 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129586 129587commit b1b5ec45c1cb650ccb8c659218f9481379c777d9 129588Author: Peter Hutterer <peter.hutterer@who-t.net> 129589Date: Mon May 11 12:53:56 2009 +1000 129590 129591 kdrive: set Activate/Deactivate grab for input devices (#21591) 129592 129593 X.Org Bug 21591 <http://bugs.freedesktop.org/show_bug.cgi?id=21591> 129594 129595commit 6f9e22049862ee9ac7f604411d005d8bb1b2dd1c 129596Author: Peter Hutterer <peter.hutterer@who-t.net> 129597Date: Mon May 11 12:51:40 2009 +1000 129598 129599 dix: ensure Activate/DeactivateGrab has a valid value. 129600 129601 Xephyr doesn't manually set Activate/DeactivateGrab for new devices, 129602 resulting in a NULL-pointer dereference later when a grab is activated. 129603 Avoid the segfault by ensuring that the pointer is always valid. 129604 129605 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129606 129607commit c1d7deca9281ba1df0df1abcea1fe7f841b10ff9 129608Author: Peter Hutterer <peter.hutterer@who-t.net> 129609Date: Tue May 12 21:24:07 2009 +1000 129610 129611 mi: un-deprecate miPointerWarpCursor 129612 129613 This function was deprecated in ef68273f5bdb27a492ec0b69548ec4fbede46c08 129614 because it didn't take a device argument. The device argument was added in 129615 1c7568b8a1417257fa67c7fca69aa253099b9461 though, so the deprecation is 129616 obsolete. 129617 129618 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129619 129620commit 8609a4e883e917d1652fbe810108de99bc89713f 129621Author: Alan Coopersmith <alan.coopersmith@sun.com> 129622Date: Fri May 15 11:22:32 2009 -0700 129623 129624 Correct some Sun license notices to Sun's standard X11 license format 129625 129626 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 129627 129628commit 94c5093c8bc803d0b14429b5a4d07ae46da865cc 129629Author: David Marx <David.Marx@Sun.COM> 129630Date: Wed May 13 17:53:30 2009 -0700 129631 129632 Solaris: Make sure non-inline versions of asm routines end with ret statements 129633 129634 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 129635 129636commit 2be5eecb4b0f30b9c55d12521415edbb88c1f8fd 129637Author: Julien Cristau <jcristau@debian.org> 129638Date: Fri May 15 17:29:32 2009 +0200 129639 129640 xfree86: add edid quirk for Samsung Syncmaster 2333HD 129641 129642 It reports vertical size in cm in the detailed mode. 129643 X.Org bug#21750 <http://bugs.freedesktop.org/show_bug.cgi?id=21750> 129644 129645 Reported-by: Peter Poklop <Peter.Poklop@gmx.at> 129646 Signed-off-by: Julien Cristau <jcristau@debian.org> 129647 129648commit 7c8327f0a75087a85864256a9cea80dd4b86def5 129649Author: Michel Dänzer <daenzer@vmware.com> 129650Date: Fri May 15 15:48:37 2009 +0200 129651 129652 EXA: Always damage glyph cache pixmap manually after uploading a glyph. 129653 129654 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 129655 129656commit 850675d4de4373e5df95507dbf2cd9affaaf54bc 129657Author: Michel Dänzer <daenzer@vmware.com> 129658Date: Fri May 15 15:48:37 2009 +0200 129659 129660 EXA: Take GC client clip type into account for migration. 129661 129662 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=18950 . 129663 129664 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 129665 129666commit c1bbac2708bbccbec01cecba39cdf2b8160c1dc3 129667Author: Matt Turner <mattst88@gmail.com> 129668Date: Thu May 14 16:00:41 2009 -0400 129669 129670 Fix mem_barrier() on Alpha 129671 129672 Some clean up (da086901c6579e41b28b1418fa12ebb82744fc87) went a bit 129673 too far and removed the needed mem_barrier macro on Alpha. 129674 129675 This re-adds it. 129676 129677commit 4cc33190ef1a4207c0f8fc74201cdf2568a239c2 129678Author: Julien Cristau <jcristau@debian.org> 129679Date: Thu May 14 17:01:57 2009 +0200 129680 129681 xfree86: Add two __FreeBSD_kernel__ checks 129682 129683 Debian bug#525475 <http://bugs.debian.org/525475> 129684 129685 Reported-by: Petr Salinger <Petr.Salinger@seznam.cz> 129686 Signed-off-by: Julien Cristau <jcristau@debian.org> 129687 129688commit ed9aecf8517bf74e943e3e5dd3ee407524e8f122 129689Author: Adam Jackson <ajax@redhat.com> 129690Date: Thu May 14 06:55:21 2009 -0400 129691 129692 COPYING: SGI FreeB 1.1 -> SGI FreeB 2.0 129693 129694commit c66cf7b64c7bcb31c81b9fde27cc40c1440753b1 129695Author: Adam Jackson <ajax@redhat.com> 129696Date: Thu May 14 06:53:08 2009 -0400 129697 129698 Update several of my and/or Red Hat's licenses to standard form. 129699 129700commit 2075d4bf9e53b8baef0b919da6c44771220cd4a5 129701Author: Michel Dänzer <daenzer@vmware.com> 129702Date: Thu May 14 11:46:41 2009 +0200 129703 129704 glx: If a destroyed window is bound to the current context, make it not current. 129705 129706 Avoids subsequent crashes due to stale pointers to the DrawableRec, see 129707 https://bugs.freedesktop.org/show_bug.cgi?id=21132#c15 and previous comments. 129708 129709 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 129710 129711commit 2c1190f888515292de01e60fe74657c34b99fd9e 129712Author: Michel Dänzer <daenzer@vmware.com> 129713Date: Thu May 14 11:39:16 2009 +0200 129714 129715 randr12: Initialize and keep track of updates to VidMode extension gamma value. 129716 129717 This way clients querying the gamma value via the VidMode extension at least 129718 get the last value set via the same, rather than always something bogus. 129719 129720 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 129721 129722commit fc3ce861cdab8606610726ce7c53f57d950c2407 129723Author: Michel Dänzer <daenzer@vmware.com> 129724Date: Thu May 14 11:35:25 2009 +0200 129725 129726 randr12: Fix calculation of gamma ramp values. 129727 129728 The reciprocal gamma value was missed in the first copy and this mistake was 129729 propagated to the second one. 129730 129731 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 129732 129733commit f3c61377d0fb1f09a18833995556fc9ffd11e097 129734Author: Alan Coopersmith <alan.coopersmith@sun.com> 129735Date: Tue May 12 18:27:09 2009 -0700 129736 129737 COPYING file updates for git master changes since 1.6 branch 129738 129739 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 129740 129741commit b746a00cffca5c553b607a8e9c1074294a23b443 129742Author: Alan Coopersmith <alan.coopersmith@sun.com> 129743Date: Tue May 12 16:11:01 2009 -0700 129744 129745 Resync COPYING file with notices in code base as of xorg-server-1.6.1 129746 129747 Remove notices for code no longer in tree (Xprint, Xgl, kdrive, cfb, etc.) 129748 Add/update notices for new/changed code in tree 129749 129750 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 129751 129752commit f250eea2e90fc50bec5214c2f41132b95edc2c46 129753Author: Jerome Glisse <glisse@freedesktop.org> 129754Date: Mon May 11 22:52:46 2009 +0200 129755 129756 DRI2: update DRI2 private drawable width & height according to X drawable 129757 129758commit 0952d12717031e9dda9e48123bb922d0f4e81834 129759Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 129760Date: Mon May 11 15:27:46 2009 -0400 129761 129762 xselinux: Relax ownership restriction on SetSelectionUseContext. 129763 129764 Instead, clients should keep track of the selection instances they use. 129765 129766commit ac13145dbcb284293582435409d8a90f276785c5 129767Author: Peter Hutterer <peter.hutterer@who-t.net> 129768Date: Mon May 11 15:45:46 2009 +1000 129769 129770 xkb: if kbd init failed, NULL out the pointers after freeing them (#21278) 129771 129772 Reproducible: 129773 Configure a server that uses the keyboard driver with an invalid ruleset, 129774 e.g. (Option "XkbRules" "foobar"). Ensure that Option "AllowEmptyInput" is 129775 "off" in the ServerFlags or ServerLayout section. Start the server. 129776 After failing to init the keymap, the server will try to clean up after the 129777 device, double-freeing some xkb structs that have not been reset properly. 129778 129779 X.Org Bug 21278 <http://bugs.freedesktop.org/show_bug.cgi?id=21278> 129780 129781 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129782 129783commit ebfd6688d1927288155221e7a78fbca9f9293952 129784Author: Søren Sandmann Pedersen <sandmann@redhat.com> 129785Date: Sun May 10 23:27:56 2009 -0400 129786 129787 Make compositing with transformed windows work again. 129788 129789 The coordinate translation was broken in pretty much every way 129790 imaginable. 129791 129792commit c2b668e6fff11c52f2a3f0dc3f4d1fc2c5bcf3e6 129793Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 129794Date: Tue May 5 20:27:53 2009 -0700 129795 129796 XQuartz: pbproxy standalone: added missing variable declarations. 129797 (cherry picked from commit 2d947d17a3ef5fb8cef959bff0eb08b8bb5b6548) 129798 129799commit 9c5b761c8c86cca56f4b4af272f42ba89c9c3ba5 129800Author: Peter Hutterer <peter.hutterer@who-t.net> 129801Date: Sat May 9 16:19:13 2009 +1000 129802 129803 Revert "xkb: write the _XKB_RF_RULES_PROP to each device." 129804 129805 This commit shouldn't have been pushed, we're still sorting out the API we 129806 want to use. 129807 129808 This reverts commit 876910a951053f0bd31e30098de3da15a1c1f5d6. 129809 129810 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129811 129812commit 1e816065e5ec3b9394dc1fa5815457a664e15fd9 129813Author: Alan Coopersmith <alan.coopersmith@sun.com> 129814Date: Fri May 8 21:31:01 2009 -0700 129815 129816 Don't printf NULL pointers on HAL connection error 129817 129818 Fixes Solaris bug 6801386 Xorg core dumps on startup if hald not running 129819 http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6801386 129820 129821 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 129822 129823commit b680bda34da130ce408783f04214771471e41e8d 129824Author: Alan Coopersmith <alan.coopersmith@sun.com> 129825Date: Thu Apr 30 18:49:06 2009 -0700 129826 129827 Fix a couple off-by-one array boundary checks. 129828 129829 Error: Write outside array bounds at Xext/geext.c:406 129830 in function 'GEWindowSetMask' [Symbolic analysis] 129831 In array dereference of cli->nextSib[extension] with index 'extension' 129832 Array size is 128 elements (of 4 bytes each), index <= 128 129833 129834 Error: Buffer overflow at dix/events.c:592 129835 in function 'SetMaskForEvent' [Symbolic analysis] 129836 In array dereference of filters[deviceid] with index 'deviceid' 129837 Array size is 20 elements (of 512 bytes each), index >= 0 and index <= 20 129838 129839 Error: Read buffer overflow at hw/xfree86/loader/loader.c:226 129840 in function 'LoaderOpen' [Symbolic analysis] 129841 In array dereference of refCount[new_handle] with index 'new_handle' 129842 Array size is 256 elements (of 4 bytes each), index >= 1 and index <= 256 129843 129844 These bugs were found using the Parfait source code analysis tool. 129845 For more information see http://research.sun.com/projects/parfait 129846 129847 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 129848 Signed-off-by: Adam Jackson <ajax@redhat.com> 129849 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 129850 129851commit 00bc043fa0398a1d14d46b87da2ff3031a9535dc 129852Author: Peter Hutterer <peter.hutterer@who-t.net> 129853Date: Fri May 8 10:10:25 2009 +1000 129854 129855 dix: export subpixel precision in XI2 events for root/event coordinates. 129856 129857 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129858 129859commit bae070914fc27db122e6131ae4838559c4a72f65 129860Author: Peter Hutterer <peter.hutterer@who-t.net> 129861Date: Thu May 7 16:52:31 2009 +1000 129862 129863 input: update to inputproto 1.9.99.8 XI2 defines. 129864 129865 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129866 129867commit 83f32d3972b8bfb0a87069dfb3fcd64b6b7c6424 129868Author: Peter Hutterer <peter.hutterer@who-t.net> 129869Date: Thu May 7 10:05:29 2009 +1000 129870 129871 Xi: Add XI2 property requests. 129872 129873commit 9935bec6e860cba9a3cc5baadd372ddb89d72ef0 129874Author: Peter Hutterer <peter.hutterer@who-t.net> 129875Date: Thu May 7 09:43:56 2009 +1000 129876 129877 Xi: split some code out of the XI 1.5 property request processing. 129878 129879 This is in preparation for the XI2 property requests that can re-use much of 129880 this code. 129881 129882commit f6f1e417063d2d61d65731a7e3ebca3ec2bdaacc 129883Author: Peter Hutterer <peter.hutterer@who-t.net> 129884Date: Wed May 6 22:25:28 2009 +1000 129885 129886 Xi: fix copy/paste error causing sizeof against wrong struct. 129887 129888 This wrong check may cause BadLength to be returned to the client even if the 129889 length is correct. 129890 129891 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129892 129893commit fc0013d744a345199f013ba5b6ef0e44201d0e68 129894Author: Peter Hutterer <peter.hutterer@who-t.net> 129895Date: Wed May 6 16:55:32 2009 +1000 129896 129897 Xi: don't double-swap the XListDeviceProperties reply. 129898 129899 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129900 129901commit 409012061b7d361a172b20455d4aaa5e47527c42 129902Author: Peter Hutterer <peter.hutterer@who-t.net> 129903Date: Wed May 6 16:42:46 2009 +1000 129904 129905 Xi: add missing break in XI event swapping function 129906 129907 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129908 129909commit c3c64978c4a231a3a8c18211d0716df875c75efa 129910Author: Peter Hutterer <peter.hutterer@who-t.net> 129911Date: Wed May 6 15:05:06 2009 +1000 129912 129913 Xi: fix a couple of wrong dixLookupDevice permission tags. 129914 129915 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129916 129917commit 0d947aa8e87c5d92b702c60190c8bc5d32c9ba9c 129918Author: Peter Hutterer <peter.hutterer@who-t.net> 129919Date: Wed May 6 12:07:07 2009 +1000 129920 129921 xfree86: fix xf86PostMotionEventP type checking 129922 129923 We only put internal events into the queue now, so let's check for ET_Motion 129924 rather than the MotionNotify. 129925 129926 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129927 129928commit 9457ce448988dacea9a3a1c14b6955767118ff31 129929Author: Simon Thum <simon.thum@gmx.de> 129930Date: Thu Apr 30 13:01:17 2009 +0200 129931 129932 dix: remove superfluous includes from ptrveloc.c 129933 129934 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129935 129936commit ee30e1b102d779403eeda97e007b81e85f006a37 129937Author: Simon Thum <simon.thum@gmx.de> 129938Date: Thu Apr 30 12:58:48 2009 +0200 129939 129940 dix: fix warning in pointer acceleration 129941 129942 newer gcc's warn against how this cast is done (though it eludes me why), 129943 and lrintf() is also faster especially on insane processors like the P4 129944 (http://www.mega-nerd.com/FPcast). 129945 129946 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129947 129948commit 5cf70183812541b33a6e83c7e1e3bc6198730cbe 129949Author: Peter Hutterer <peter.hutterer@who-t.net> 129950Date: Tue May 5 21:00:31 2009 +1000 129951 129952 xkb: remove _XkbAlloc, _XkbCalloc, _XkbRealloc and _XkbFree 129953 129954 We all agree that wrapping is fun, but seriously. One of these days someone 129955 will get hurt. 129956 129957 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129958 129959commit 0e31d3906deaee5d9ada66e538b9e93574a4d610 129960Author: Peter Hutterer <peter.hutterer@who-t.net> 129961Date: Tue May 5 21:07:07 2009 +1000 129962 129963 xkb: remove some now-useless XFUNCPROTOBEGIN 129964 129965 We bring them back if we start rewriting the server in C++, promise. 129966 129967 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129968 129969commit 8b583ca2b21155359c6255f406c96599b277c762 129970Author: Peter Hutterer <peter.hutterer@who-t.net> 129971Date: Wed May 6 22:25:28 2009 +1000 129972 129973 Xi: fix copy/paste error causing sizeof against wrong struct. 129974 129975 This wrong check may cause BadLength to be returned to the client even if the 129976 length is correct. 129977 129978 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129979 129980commit 7ab5e9b97c300bba793a23fa13506b0c77c50ddf 129981Author: Peter Hutterer <peter.hutterer@who-t.net> 129982Date: Wed May 6 12:07:07 2009 +1000 129983 129984 xfree86: fix xf86PostMotionEventP type checking 129985 129986 We only put internal events into the queue now, so let's check for ET_Motion 129987 rather than the MotionNotify. 129988 129989 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129990 129991commit 50cc8adafca4ba3838d468278d6eb8a4692d2488 129992Author: Peter Hutterer <peter.hutterer@who-t.net> 129993Date: Wed May 6 16:55:32 2009 +1000 129994 129995 Xi: don't double-swap the XListDeviceProperties reply. 129996 129997 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 129998 129999commit 3fc6fcfb267888d65a22ed7c3eda175a2530a3d5 130000Author: Simon Thum <simon.thum@gmx.de> 130001Date: Wed May 6 10:39:16 2009 +0200 130002 130003 dix: add 'none' pointer acceleration profile with number -1 130004 130005 This is a shorthand for disabling acceleration, while retaining the 130006 possiblity to use constant deceleration. If constant deceleration is 130007 also unused, it will optimize motion processing. 130008 130009 Other possiblities to deactivate acceleration were quite hidden, 130010 and didn't always work as expected. E.g. xset m 1 1 would retain 130011 adaptive deceleration, while xset m 1 0 would not (in the default 130012 profile). 130013 130014 Also removes the 'reserved' profile; it was unused and it's trivial 130015 to add new ones anyway. 130016 130017 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130018 130019commit 9d1597cbefea6a5e7959f2099c46f1a284def7ad 130020Author: Simon Thum <simon.thum@gmx.de> 130021Date: Thu Apr 30 13:01:17 2009 +0200 130022 130023 dix: remove superfluous includes from ptrveloc.c 130024 130025 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130026 130027commit 35fce4e5c7fa34f98b3e4010c6cb09ce38a9205c 130028Author: Simon Thum <simon.thum@gmx.de> 130029Date: Thu Apr 30 12:58:48 2009 +0200 130030 130031 dix: fix warning in pointer acceleration 130032 130033 newer gcc's warn against how this cast is done (though it eludes me why), 130034 and lrintf() is also faster especially on insane processors like the P4 130035 (http://www.mega-nerd.com/FPcast). 130036 130037 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130038 130039commit d220d6907d1d5138d1528c48b739e77f65616225 130040Author: Peter Hutterer <peter.hutterer@who-t.net> 130041Date: Mon May 4 17:38:02 2009 +1000 130042 130043 Xi: add GrabButton and GrabKeysym code. 130044 130045 We don't do keycode grabs in XI2, they're pointless. 130046 130047commit 1b593ced171d02f1d00034f0f733060706bb6d41 130048Author: Peter Hutterer <peter.hutterer@who-t.net> 130049Date: Tue May 5 16:57:46 2009 +1000 130050 130051 Xi: access the prev_state to get group/modifier state for key events. 130052 130053 Key events may change the modifier state, so we need to get the prev_state for 130054 those (i.e. without the changes by the event already applied). 130055 130056 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130057 130058commit 35a4b8e7f4526a92d44cb16a783f21030cd1f6df 130059Author: Peter Hutterer <peter.hutterer@who-t.net> 130060Date: Tue May 5 16:51:59 2009 +1000 130061 130062 xkb: remove oldState from XkbHandleActions. 130063 130064 I really don't know what the purpose of this variable is or was, aside from 130065 potentially clobbering up our key state since there's a path where it may be 130066 used uninitialised. 130067 130068 Also, this means that xkbi->prev_state is now accessible from the DIX with 130069 meaningful data. 130070 130071 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130072 130073commit d523fbe428e4513d85402caa83e9349256b680cc 130074Author: Peter Hutterer <peter.hutterer@who-t.net> 130075Date: Mon May 4 17:00:57 2009 +1000 130076 130077 test: add test for xi2 struct sizes. 130078 130079commit 1b1b20d6e3e696e4437a9ef56128cde70a485f66 130080Author: Peter Hutterer <peter.hutterer@who-t.net> 130081Date: Tue May 5 13:47:55 2009 +1000 130082 130083 Change glib require for tests to auto. 130084 130085 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130086 130087commit 749cae82204f347248c815b1068ffcae9825fe70 130088Author: Peter Hutterer <peter.hutterer@who-t.net> 130089Date: Thu Apr 30 14:58:00 2009 +1000 130090 130091 mi: don't crash if we're trying to update the pointer for a keyboard. 130092 130093 If a device doesn't have a pointer, just return. 130094 130095 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130096 130097commit 7ecedb0f2ee5b53513205d1f0aa381451893b9c5 130098Author: Peter Hutterer <peter.hutterer@who-t.net> 130099Date: Fri May 1 14:35:12 2009 +1000 130100 130101 include: up the number of max. input devices to 40. 130102 130103 With the Xtest virtual slave devices we have 4 devices for each MD 130104 pointer/keyboard pair, plus the AllDevices and AllMasterDevices reserved 130105 deviceids. It's quite easy to hit the current limit. 130106 130107 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130108 130109commit a38d33f6ac765bc5a3296b325e2c9b4a6a10ebf1 130110Author: Peter Hutterer <peter.hutterer@who-t.net> 130111Date: Fri May 1 15:24:02 2009 +1000 130112 130113 Xext: shut up compiler warnings in xtest.c 130114 130115 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130116 130117commit 14d5471981151b27678de826daa078977217477a 130118Author: Peter Hutterer <peter.hutterer@who-t.net> 130119Date: Fri May 1 15:22:37 2009 +1000 130120 130121 Xext: return BadDevice from XTest if we don't have keys/buttons/valuators. 130122 130123 BadDevice is an XI error, but this cannot happen for core XTest fake input 130124 anyway since the device will be the matching virtual XTest slave device. 130125 130126 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130127 130128commit f1bed0959e803784af142a0f45177ee3b675d78a 130129Author: Peter Hutterer <peter.hutterer@who-t.net> 130130Date: Fri May 1 15:15:04 2009 +1000 130131 130132 Xext: fix core Xtest button presses, don't call PickPointer. 130133 130134 We already did the device selection before, so dev should be the XTest virtual 130135 pointer. 130136 130137 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130138 130139commit 94648bb797d94b025746c60679c584e5be2fae28 130140Author: Jesse Barnes <jbarnes@virtuousgeek.org> 130141Date: Mon May 4 15:38:22 2009 -0700 130142 130143 Don't prepare outputs & crtcs if set_mode_major is present 130144 130145 A driver with this hook will take care of preparing the outputs & crtcs, 130146 so calling the prepare functions will just cause unnecessary flicker. 130147 130148 Fixes bug #21077 130149 130150commit 7d85169c7ab1d05c21fdbb2877ffd79f344dbb6b 130151Author: Michel Dänzer <daenzer@vmware.com> 130152Date: Mon May 4 10:14:43 2009 +0200 130153 130154 Fix typo in ProcRenderCreateAnimCursor. 130155 130156commit 59358de4b68a01452bf68593beb83605087ebc8f 130157Author: Peter Hutterer <peter.hutterer@who-t.net> 130158Date: Mon May 4 17:30:19 2009 +1000 130159 130160 input: ensure various ProcUngrabKey/Buttons have the right grabtype set. 130161 130162commit 834ea071b619506e0bef5bdbf0e3c59831cf6e99 130163Author: Peter Hutterer <peter.hutterer@who-t.net> 130164Date: Mon May 4 15:58:15 2009 +1000 130165 130166 dix: if grab types differ, two grabs cannot be the same. 130167 130168commit 63a3c727b35d1ffd8a850a5177bd6623cd78b179 130169Author: Peter Hutterer <peter.hutterer@who-t.net> 130170Date: Fri May 1 15:24:02 2009 +1000 130171 130172 Xext: shut up compiler warnings in xtest.c 130173 130174 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130175 130176commit 226dd90597ef29998ac5db15c5457aad3fe9729e 130177Author: Peter Hutterer <peter.hutterer@who-t.net> 130178Date: Fri May 1 15:22:37 2009 +1000 130179 130180 Xext: return BadDevice from XTest if we don't have keys/buttons/valuators. 130181 130182 BadDevice is an XI error, but this cannot happen for core XTest fake input 130183 anyway since the device will be the matching virtual XTest slave device. 130184 130185 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130186 130187commit 0fdff0a47a57a0096974ee83e07fefbad03075dd 130188Author: Peter Hutterer <peter.hutterer@who-t.net> 130189Date: Fri May 1 15:15:04 2009 +1000 130190 130191 Xext: fix core Xtest button presses, don't call PickPointer. 130192 130193 We already did the device selection before, so dev should be the XTest virtual 130194 pointer. 130195 130196 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130197 130198commit f0124ed93cdcee8d22a6690e639a3f60b9a63c54 130199Author: Peter Hutterer <peter.hutterer@who-t.net> 130200Date: Fri May 1 14:35:12 2009 +1000 130201 130202 include: up the number of max. input devices to 40. 130203 130204 With the Xtest virtual slave devices we have 4 devices for each MD 130205 pointer/keyboard pair, plus the AllDevices and AllMasterDevices reserved 130206 deviceids. It's quite easy to hit the current limit. 130207 130208 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130209 130210commit 08cd846000286eb5e07cc8d8cb313aff2d6e7faf 130211Author: Peter Hutterer <peter.hutterer@who-t.net> 130212Date: Fri May 1 11:06:06 2009 +1000 130213 130214 dix: only free the old cursor if the grab was successful. 130215 130216 Testcase: start thunderbird and move a message around, crashes after two or 130217 three moves. 130218 130219 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130220 130221commit 557dbadf3be273255e8fdb12d9321f4e88bf2b65 130222Author: Alan Coopersmith <alan.coopersmith@sun.com> 130223Date: Sat May 2 00:26:18 2009 -0700 130224 130225 XkbSetNamedIndicator should ignore SD's without LED's 130226 130227 When ProcXkbSetNamedIndicator is called on a core device, and we 130228 walk the slaves to set the LED's on each of them, ignore any slaves 130229 that do not have either a KbdFeedbackCtrl or LedCtrl structure. 130230 130231 (This is much more critical in xserver-1.5-branch, where we walk 130232 *all* devices, not just the slaves of the specified master, and 130233 thus return failure when setting an LED on the Core Keyboard and 130234 hit a xf86-input-mouse device with no LED's to set.) 130235 130236 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 130237 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 130238 130239commit 0d9d3f3e361f769822caedccf4c2a58cc9930ecc 130240Author: Ian Romanick <ian.d.romanick@intel.com> 130241Date: Mon Apr 27 15:11:10 2009 -0700 130242 130243 DRI2: Force allocation of real-front buffer for non-windows as well 130244 130245 For redirected rendering we end up with pixmaps (which the app thinks are 130246 windows) that are double buffered. 130247 130248 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 130249 Tested-by: Pierre Willenbrock <pierre@pirsoft.de> 130250 130251commit 3d30789a05a730a03faa6058c73a5eda36ef3779 130252Author: Peter Hutterer <peter.hutterer@who-t.net> 130253Date: Thu Apr 30 14:58:00 2009 +1000 130254 130255 mi: don't crash if we're trying to update the pointer for a keyboard. 130256 130257 If a device doesn't have a pointer, just return. 130258 130259 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130260 130261commit 876910a951053f0bd31e30098de3da15a1c1f5d6 130262Author: Peter Hutterer <peter.hutterer@who-t.net> 130263Date: Thu Apr 30 14:46:45 2009 +1000 130264 130265 xkb: write the _XKB_RF_RULES_PROP to each device. 130266 130267 We only have one root window and writing the rules used to the same property 130268 for each device is quite pointless if you don't have the same RMLVO on all 130269 devices. So let's be sensible and write the same property to the device too, 130270 so at least we know which device got loaded with which RMLVO. 130271 130272commit 44044adc92e6bb6537f48c84727523d5cb23528c 130273Author: Peter Hutterer <peter.hutterer@who-t.net> 130274Date: Thu Apr 30 11:39:50 2009 +1000 130275 130276 xfree86: print a message if NIDR fails due to AutoAddDevices off. 130277 130278 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130279 130280commit 43dbbc30327f826068bc7cd2cdaf870ed61b4522 130281Author: Peter Hutterer <peter.hutterer@who-t.net> 130282Date: Tue Apr 28 21:01:36 2009 +1000 130283 130284 Xi: fix typo in ProcXIGrabDevice 130285 130286 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130287 130288commit 6a618929a0c06ba0d6dac13d7e644cd9658d98ed 130289Author: Peter Hutterer <peter.hutterer@who-t.net> 130290Date: Tue Apr 28 16:49:45 2009 +1000 130291 130292 input: reshuffle CreateGrab and friends to take a GrabParameters param. 130293 130294 This is cleaning up work in preparation for XI2 passive grabs. 130295 130296commit e8e26f700c9c70d3f1bb53bdb71d1100f5c43a69 130297Author: Peter Hutterer <peter.hutterer@who-t.net> 130298Date: Sun Apr 26 21:43:25 2009 +1000 130299 130300 Xi: split some grab parameter checking out of GrabButton and GrabKey. 130301 130302commit 57aff88c7d0761e590806d07bee1c9410680c89f 130303Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 130304Date: Wed Apr 29 01:04:37 2009 -0400 130305 130306 Fix most remaining deprecated resource lookups. 130307 130308 Callsites updated to use dixLookupResourceBy{Type,Class}. 130309 TODO: Audit access modes to make sure they reflect the usage. 130310 130311commit 1abe0ee3da5e1268c7315f841d31337ea6524cf0 130312Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 130313Date: Tue Apr 28 23:51:40 2009 -0400 130314 130315 kdrive: fix Xvfb build with separate libmain. 130316 130317commit 737b49199a05299486064e6e762cf2a2f6f95be6 130318Author: Peter Hutterer <peter.hutterer@who-t.net> 130319Date: Thu Apr 9 15:40:24 2009 +1000 130320 130321 xfree86: restore default off for DontZap 130322 130323 Zapping is triggered by xkb these days, so note in the man page that it's the 130324 Terminate_Server action. Since it's XKB, personal preferences towards or 130325 against zapping should be achieved through xkb rulesets. 130326 If Terminate_Server is not in the xkb actions, then we can't zap anyway and we 130327 don't need a default of DontZap "on". 130328 130329 This patch restores the old meaning of DontZap - disallow zapping altogether, 130330 regardless of XKB's current keymap. 130331 Ideally, this patch should be accompanied by b0f64bdab00db652e in 130332 xkeyboard-config. 130333 130334 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130335 130336commit 98a1fa994158feca8bc80a10296a7675465c0e56 130337Author: Peter Hutterer <peter.hutterer@who-t.net> 130338Date: Tue Apr 21 22:18:23 2009 +1000 130339 130340 test: add InternalEvent to core event conversion tests. 130341 130342 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130343 130344commit 990e8362a76fadc173e890d9052a2e5274b6eac5 130345Author: Peter Hutterer <peter.hutterer@who-t.net> 130346Date: Thu Apr 16 11:41:58 2009 +1000 130347 130348 test: add a simple test to verify device axis intialization. 130349 130350 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130351 130352commit 4eac0df060b897470fb01cfd3b70f500da757c74 130353Author: Peter Hutterer <peter.hutterer@who-t.net> 130354Date: Wed Apr 15 17:21:08 2009 +1000 130355 130356 Add a test-suite for in-server unit-testing. 130357 130358 This patch adds a test/ directory that contains the setup for a unit-testing 130359 suite designed for in-server unit-testing. All functions available to the X 130360 server are available to the test binaries through static linking. 130361 130362 This test suite uses the glib testing framework. 130363 Do not use glib calls outside of the test/ directory. 130364 130365 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130366 130367commit c2c515ead38d9a6c9eae0b83aa7a82208f177b7e 130368Author: Peter Hutterer <peter.hutterer@who-t.net> 130369Date: Wed Apr 22 13:26:40 2009 +1000 130370 130371 dix: remove all but main() from main.c 130372 130373 All other functions are pushed into where they seemed to fit. 130374 main.c is now linked separately into libmain.a and linked in by the various 130375 DDXs. 130376 130377 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130378 130379commit dc2767d1c5db60385867c76ba2de507fe0cb8a90 130380Author: Peter Hutterer <peter.hutterer@who-t.net> 130381Date: Tue Apr 21 22:18:23 2009 +1000 130382 130383 test: add InternalEvent to core event conversion tests. 130384 130385 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130386 130387commit 1127ca097cb75450bcccfc5f5d82e435de2fb5b7 130388Author: Peter Hutterer <peter.hutterer@who-t.net> 130389Date: Thu Apr 16 11:41:58 2009 +1000 130390 130391 test: add a simple test to verify device axis intialization. 130392 130393 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130394 130395commit 4124c465a85713fe44843a16c5e2b13ece17e9d2 130396Author: Peter Hutterer <peter.hutterer@who-t.net> 130397Date: Wed Apr 15 17:21:08 2009 +1000 130398 130399 Add a test-suite for in-server unit-testing. 130400 130401 This patch adds a test/ directory that contains the setup for a unit-testing 130402 suite designed for in-server unit-testing. All functions available to the X 130403 server are available to the test binaries through static linking. 130404 130405 This test suite uses the glib testing framework. 130406 Do not use glib calls outside of the test/ directory. 130407 130408 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130409 130410commit 987579c930bda803427a28cb82773c389f5110d6 130411Author: Peter Hutterer <peter.hutterer@who-t.net> 130412Date: Wed Apr 22 13:26:40 2009 +1000 130413 130414 dix: remove all but main() from main.c 130415 130416 All other functions are pushed into where they seemed to fit. 130417 main.c is now linked separately into libmain.a and linked in by the various 130418 DDXs. 130419 130420 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130421 130422commit 2a2a7fa6856b2e1954e904577535118f20cd9133 130423Author: Alan Coopersmith <alan.coopersmith@sun.com> 130424Date: Mon Apr 27 20:23:53 2009 -0700 130425 130426 Solaris: use <sys/agpgart.h> instead of stale copy in Xorg sources 130427 130428 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 130429 130430commit 707af5f8c5fae3b6c4c5125a91917667bb0ab908 130431Author: Simon Farnsworth <simon.farnsworth@onelan.co.uk> 130432Date: Thu Apr 23 16:15:01 2009 +0100 130433 130434 Make the cursor completely optional 130435 130436 For embedded use, it's convenient to be able to disable the cursor 130437 completely, without having to audit and fix up all your third-party 130438 code (e.g. Mozilla Firefox). 130439 130440 Add -nocursor and -cursor server options to enable and disable the 130441 cursor. The default is still -cursor, but embedded users can run the 130442 server with -nocursor to hide the cursor regardless of what 130443 application developers do. 130444 130445 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130446 130447commit 40a8f2f4088aa353de733a11327b628ef13a52e6 130448Author: Pierre Willenbrock <pierre@pirsoft.de> 130449Date: Sun Apr 19 21:15:22 2009 +0200 130450 130451 Fix obvious copypasta 130452 130453 Reviewed-by: Ian Romanick <idr@freedesktop.org> 130454 130455commit a8cf63fd18b0ddadbadf7b47d620159854920050 130456Author: Aaron Plattner <aplattner@nvidia.com> 130457Date: Thu Apr 23 14:57:04 2009 -0700 130458 130459 Export CompositeRegisterAlternateVisuals. 130460 130461 Drivers need this if they add visuals that require implicit redirection. Add a 130462 new SDK header, compositeext.h, to contain the prototype for this function. 130463 130464commit 48573e7ea13e279593249036e6d30f1d7507882f 130465Author: Peter Hutterer <peter.hutterer@who-t.net> 130466Date: Sun Apr 26 21:32:37 2009 +1000 130467 130468 Xi: swap stuff->cursor in SProcXIGrabDevice 130469 130470commit ff6c7764c2909e4126403b7211faa6c58556b341 130471Author: Ian Romanick <ian.d.romanick@intel.com> 130472Date: Fri Apr 24 12:49:19 2009 -0700 130473 130474 DRI2: Implement protocol for DRI2GetBuffersWithFormat 130475 130476 This change implements the protocol for DRI2GetBuffersWithFormat, but 130477 the bulk of the differences are the changes to the extension / driver 130478 interface to make this function work. The old CreateBuffers and 130479 DeleteBuffers routines are replaced with CreateBuffer and DeleteBuffer 130480 (both singular). 130481 130482 This allows drivers to allocate buffers for a drawable one at a time. 130483 As a result, 3D drivers can now allocate the (fake) front-buffer for a 130484 window only when it is needed. Since 3D drivers only ask for the 130485 front-buffer on demand, the real front-buffer is always created. This 130486 allows CopyRegion impelemenations of SwapBuffers to continue working. 130487 As with previous version of this code, if the client asks for the 130488 front-buffer for a window, we instead give it the fake front-buffer. 130489 130490 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 130491 Reviewed-by: Kristian Høgsberg <krh@redhat.com> 130492 130493commit 28ddfc88d8d547941c7f4713db527a3c2f9ec35a 130494Author: Ian Romanick <ian.d.romanick@intel.com> 130495Date: Fri Apr 24 12:09:21 2009 -0700 130496 130497 DRI2: Add interface for drivers to query DRI2 extension version 130498 130499 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 130500 130501commit a1c98f7fa9aa0cd5075ef2054c5d2be4d1e740f0 130502Author: Joel Bosveld <Joel.Bosveld@gmail.com> 130503Date: Fri Apr 24 10:12:26 2009 -0700 130504 130505 Correct fcntl(F_SETOWN) error check in non-Solaris sigio code 130506 130507 commit 64b7f96dca accidentally inverted the comparison, could 130508 result in crashes on some systems later on. 130509 130510 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 130511 130512commit 64b7f96dca23d2b33e0b8a8785d628d9a2f50e7b 130513Author: Alan Coopersmith <alan.coopersmith@sun.com> 130514Date: Fri Apr 24 00:14:12 2009 -0700 130515 130516 Add SIGIO/SIGPOLL support for Solaris 130517 130518 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 130519 130520commit 0c467b1f990f156a34d5ef117bd825df6b633114 130521Author: Alan Coopersmith <alan.coopersmith@sun.com> 130522Date: Fri Apr 24 00:05:28 2009 -0700 130523 130524 Add casts to generated entries in sdksyms.c to silence type mismatch warnings 130525 130526 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 130527 130528commit 7d0f7518c2235a9dc783029971259ddaada2db20 130529Author: Alan Coopersmith <alan.coopersmith@sun.com> 130530Date: Thu Apr 23 08:42:38 2009 -0700 130531 130532 Fix byte swapping of XF86VidMode{Get,Set}GammaRamp 130533 130534 Fixes OpenSolaris Bug 8315: 130535 Xorg segfaults when screensaver fades in cross-endian xdmcp session 130536 <http://defect.opensolaris.org/bz/show_bug.cgi?id=8315> 130537 130538 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 130539 130540commit 057fc9a4f89282e440b5e11458f1dd8655879dd8 130541Merge: 95628b797 932d6bcbb 130542Author: Peter Hutterer <peter.hutterer@who-t.net> 130543Date: Fri Apr 24 16:15:47 2009 +1000 130544 130545 Merge branch 'master' into xi2 130546 130547 Conflicts: 130548 Xi/chdevhier.c 130549 include/input.h 130550 130551commit 932d6bcbb68194c5bdfeb336f700dc8b31529223 130552Author: Peter Hutterer <peter.hutterer@who-t.net> 130553Date: Fri Apr 24 15:28:45 2009 +1000 130554 130555 mi: remove superfluous check. 130556 130557 mieqProcessInputEvents doesn't process events from MDs anymore, so we don't 130558 need to check for pDev->isMaster. 130559 130560 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130561 130562commit fab563bf8f6b63906ce9d5a568c467425843265b 130563Author: Benjamin Close <Benjamin.Close@clearchain.com> 130564Date: Fri Mar 27 16:44:15 2009 +1030 130565 130566 input: propagate XTst events through virtual slave devices. 130567 130568 A XTest virtual slave device pair (kbd/ptr) exists for every master 130569 device pair. This is so XTest events are correctly propogated via slave 130570 devices up to Master devices and the classes are correctly changed along 130571 the way. We add the XTest slave device pair to the Virtual Core pointer 130572 and provide a simple way of creating the devices. 130573 130574 A XTest Slave Device is identified by the XTstDevicePrivateKey property 130575 being set in the devices devProperties 130576 130577 XI events are still propagated through the matching device, in the hope the 130578 client knows what it is doing. 130579 130580 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130581 130582commit ccd28ca2ad40aecf951ab058d89062828e3e5d8a 130583Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 130584Date: Wed Apr 22 19:23:17 2009 -0700 130585 130586 XQuartz: Localization update 130587 (cherry picked from commit 9435fc5e20e25ce66ec85bde033daff51f39f69c) 130588 130589commit a362d0f4fb79150f7b8e0eb3d1dc1cdee2f98d95 130590Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 130591Date: Sat Apr 11 19:06:45 2009 -0700 130592 130593 XQuartz: Make sure the bgMouseLocationUpdated state is consistent 130594 (cherry picked from commit 8fa8a8e1db60c8ac60d10f1f15517471d2291f77) 130595 130596commit 8dc8812c226222fc64f8e489f9f123687a95d3d1 130597Author: Dave Airlie <airlied@redhat.com> 130598Date: Thu Apr 23 12:13:35 2009 +1000 130599 130600 exa: add missing exa.h header include 130601 130602commit 3ec6a121e110ba3cae964ee201d4d5bf818b6dff 130603Author: Dave Airlie <airlied@redhat.com> 130604Date: Thu Apr 23 12:13:05 2009 +1000 130605 130606 exa: missed exa.h change 130607 130608commit 02ae85c4c93bbd0013c063ce0014b365a8ba061f 130609Author: Dave Airlie <airlied@redhat.com> 130610Date: Thu Apr 23 12:04:17 2009 +1000 130611 130612 exa: add CreatePixmap2 hook for driver pixmaps. 130613 130614 This adds a revised pixmap hook for driver pixmaps, which is 130615 required to support tiling on various chips. 130616 130617 Signed-off-by: Dave Airlie <airlied@redhat.com> 130618 130619commit 9d684ba0bcab53841a17ebd521507b6df8bc4b45 130620Author: Dave Airlie <airlied@redhat.com> 130621Date: Thu Apr 16 10:28:36 2009 +1000 130622 130623 exa: avoid offscreen pixmap swapped out flag for driver pixmaps 130624 130625commit 0dfb97f15f591f85e079f5829c77d0c328d00464 130626Author: Julien Cristau <jcristau@debian.org> 130627Date: Wed Apr 22 20:01:49 2009 +0200 130628 130629 Bug#21324: Add quirk for Iiyama Vision Master 450 130630 130631 Reported-by: Jeremy Henty <onepoint@starurchin.org> 130632 Signed-off-by: Julien Cristau <jcristau@debian.org> 130633 130634commit 44227ef1b77467c76147b9bf79bdd0e6305a522a 130635Author: Ian Romanick <ian.d.romanick@intel.com> 130636Date: Mon Apr 20 18:20:52 2009 -0700 130637 130638 DRI2: Send the version the code actually supports 130639 130640 This prevents building an older server with a new dri2proto.h from 130641 resulting in a DRI2 extension module that lies about the version it 130642 supports. 130643 130644 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 130645 130646commit d1e916d29be8b470cbc8cadcf6e83991fdbc5a9f 130647Author: Ian Romanick <ian.d.romanick@intel.com> 130648Date: Thu Apr 16 12:10:34 2009 -0700 130649 130650 DRI2: Add missing front-buffer flush callback. 130651 130652 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 130653 130654commit 826a5bff0136b2b4d55a9e6e6bc3a7a64da9031e 130655Author: Benjamin Close <benjamin.close@clearchain.com> 130656Date: Thu Apr 16 15:33:30 2009 +1000 130657 130658 dix: Change AllocMaster into AllocDevicePair, allow creation of SDs too. 130659 130660 Allocating a slave device is essentially the same as allocating a master device. 130661 Hence we rename AllocMaster to AllocDevicePair and provided the ability to 130662 indicate if a master or slave device pair is required. 130663 130664 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130665 130666commit e2e5932bda3f473629d4be6f3ca4dcab18993eb6 130667Author: Peter Hutterer <peter.hutterer@who-t.net> 130668Date: Thu Apr 16 11:06:52 2009 +1000 130669 130670 dix: don't allow more than MAX_VALUATORS on one device. 130671 130672 Some keyboards (?) advertise more than MAX_VALUATORS axes. Parts of the 130673 internal event delivery relies on not having more than MAX_VALUATOR axes, so 130674 let's cap it down. 130675 If there's real devices that require more than the current 36, I'm sure we can 130676 bump this up. 130677 130678 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130679 130680commit 063833f3a6d9f8f657e3de309c8d6d5c3d606513 130681Author: Julien Cristau <jcristau@debian.org> 130682Date: Wed Dec 24 14:23:37 2008 +0100 130683 130684 Add XI 1.5 event and requests to protocol.txt 130685 130686commit 4f86ee61a4abf7a29e565d095aa08abd0ca9dc66 130687Author: Julien Cristau <jcristau@debian.org> 130688Date: Wed Dec 24 14:17:27 2008 +0100 130689 130690 Add RandR 1.3 requests to protocol.txt 130691 130692commit 6559f02ef89cb0ee365110d6a928ebd29632ad53 130693Author: Colin Harrison <colin.harrison@virgin.net> 130694Date: Fri Apr 17 15:14:45 2009 +0100 130695 130696 xkb: set bell_func in InitKeyboardDeviceStruct. 130697 130698 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130699 130700commit 95628b797e2b326ae15406df4f736e25f5d2f12f 130701Author: Peter Hutterer <peter.hutterer@who-t.net> 130702Date: Mon Apr 20 14:04:54 2009 +1000 130703 130704 Xi: use the XI2 defines for AsyncPair, SyncDevice, etc, not the XI ones. 130705 130706 And don't allow some values on SDs either. 130707 130708commit 335c63fcd6a53c1d3dffd6e00c0acd400c61e3d3 130709Author: Werner LEMBERG <wl@gnu.org> 130710Date: Fri Apr 17 15:22:57 2009 +0200 130711 130712 Add newline to some LogMessage strings. 130713 130714 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130715 130716commit d5ad14c8ed4d8360e1df8cd0bacf6a7c9c31df91 130717Merge: 54716fd3d 0e0642ee9 130718Author: Peter Hutterer <peter.hutterer@who-t.net> 130719Date: Sun Apr 19 22:28:22 2009 +1000 130720 130721 Merge branch 'master' into xi2 130722 130723commit 54716fd3dbc251db9d251d1d0435942efaa63259 130724Author: Peter Hutterer <peter.hutterer@who-t.net> 130725Date: Sun Apr 19 22:12:11 2009 +1000 130726 130727 Convert to using int32_t fixed point values on the wire. 130728 130729 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130730 130731commit 129ac9a9145323e3f126590b491e718f976f80ce 130732Author: Peter Hutterer <peter.hutterer@who-t.net> 130733Date: Sun Apr 19 21:31:56 2009 +1000 130734 130735 dix: set root_x/y for device events. 130736 130737 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130738 130739commit eeb1e4cd254c047539933c3b886037e67ef30762 130740Author: Peter Hutterer <peter.hutterer@who-t.net> 130741Date: Wed Apr 15 22:17:15 2009 +1000 130742 130743 Xi: add support for XIAllowEvents. 130744 130745 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130746 130747commit 74d0fc3aee05bc4f505274d81dd44c0d2b4368ff 130748Author: Peter Hutterer <peter.hutterer@who-t.net> 130749Date: Thu Apr 16 11:06:52 2009 +1000 130750 130751 dix: don't allow more than MAX_VALUATORS on one device. 130752 130753 Some keyboards (?) advertise more than MAX_VALUATORS axes. Parts of the 130754 internal event delivery relies on not having more than MAX_VALUATOR axes, so 130755 let's cap it down. 130756 If there's real devices that require more than the current 36, I'm sure we can 130757 bump this up. 130758 130759 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130760 130761commit edb70caf21fa5da3c336c47c62218a3f05c635b8 130762Author: Peter Hutterer <peter.hutterer@who-t.net> 130763Date: Wed Apr 15 22:16:49 2009 +1000 130764 130765 dix: remove un-used parameter "core" from AllowSome 130766 130767 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130768 130769commit bb5418d4901017c657031181d3839f58b387a2a3 130770Author: Peter Hutterer <peter.hutterer@who-t.net> 130771Date: Sun Apr 12 22:22:21 2009 +1000 130772 130773 Xi: Add support for XI2 active grabs and ungrabs. 130774 130775 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130776 130777commit c11ef87931f920ba782ba4e9b47d9c31ad7c1cf7 130778Author: Peter Hutterer <peter.hutterer@who-t.net> 130779Date: Sun Apr 12 22:16:14 2009 +1000 130780 130781 Xi: take XI2 requests into account when checking opcodes. 130782 130783 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130784 130785commit c94ea5bc055e4efc323e84b7a8266e8b8a4af48e 130786Author: Peter Hutterer <peter.hutterer@who-t.net> 130787Date: Sun Apr 12 17:38:28 2009 +1000 130788 130789 input: use a GrabMask union in GrabDevice to allow for XI2 masks. 130790 130791 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130792 130793commit 6bb4b5b93701535402f65ea828348ed7747c7dbf 130794Author: Peter Hutterer <peter.hutterer@who-t.net> 130795Date: Sun Apr 12 16:20:10 2009 +1000 130796 130797 Xi: add XI2 grab protocol request handling. 130798 130799 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130800 130801commit 09f9a86077f0058ce88ee9b3df5d1ab854eeca43 130802Author: Peter Hutterer <peter.hutterer@who-t.net> 130803Date: Sun Apr 12 16:19:45 2009 +1000 130804 130805 input: replace GrabRec's coreGrab field with grabtype. 130806 130807 Don't allow grabs of different types to override each other. 130808 130809 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130810 130811commit 7fbe1b7d63c94b344e8ff9d09c0bb162ca104909 130812Author: Peter Hutterer <peter.hutterer@who-t.net> 130813Date: Sat Apr 11 13:10:31 2009 +1000 130814 130815 Xi: remove the GetExtensionVersion hack. 130816 130817 Squashing data into a request that's defined to be padding is probably not the 130818 best idea. 130819 130820 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130821 130822commit 0e0642ee9466d3268476d0084a83a9d93a4aa555 130823Author: Peter Hutterer <peter.hutterer@who-t.net> 130824Date: Thu Apr 16 16:17:07 2009 +1000 130825 130826 os: don't malloc memory in LogVMessageVerb. 130827 130828 LogVWrite is limited to a buffer size of 1024, so we don't loose anything here 130829 by truncating. This way we can use LogVMessageVerb (and xf86Msg and friends) 130830 during signal handlers with the normal message types. 130831 130832 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130833 Acked-by: Alan Coopersmith <alan.coopersmith@sun.com> 130834 130835commit 62d2fb68638e9f2aa3c1d72027619c4d38f5b812 130836Author: Peter Hutterer <peter.hutterer@who-t.net> 130837Date: Thu Apr 16 17:06:33 2009 +1000 130838 130839 xkb: Add XkbFreeRMLVOSet helper function. 130840 130841 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130842 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 130843 130844commit 73e1055b1275470dcc2d53b15898c866d9721d3e 130845Author: Peter Hutterer <peter.hutterer@who-t.net> 130846Date: Sat Apr 11 11:19:29 2009 +1000 130847 130848 Xi: fix a typo in a #ifdef 130849 130850 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130851 130852commit cbbc7fed994448852961a987a8b0eadf718b24bf 130853Author: Simon Thum <simon.thum@gmx.de> 130854Date: Wed Apr 1 10:53:10 2009 +0200 130855 130856 dix: correctly utilize tracker buffer and protect from timer overruns 130857 130858 two small related fixes hard to split up 130859 130860 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130861 130862commit ca4918ac2ff003142aefbd8ac7ec2bc398ccc6eb 130863Author: Peter Hutterer <peter.hutterer@who-t.net> 130864Date: Thu Apr 9 19:31:04 2009 +1000 130865 130866 Xi: remove fail from SProcXISelectEvent. 130867 130868 Leftover from rebasing once too often, I guess. 130869 130870 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130871 130872commit 463676f016a105a790cbb386d20eb5b188c04652 130873Author: Peter Hutterer <peter.hutterer@who-t.net> 130874Date: Thu Apr 9 19:11:59 2009 +1000 130875 130876 Xi: remove un-used variable pXIClient 130877 130878 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130879 130880commit 32b13ec4cd42842a133aa6428f49638b0448efc9 130881Author: Peter Hutterer <peter.hutterer@who-t.net> 130882Date: Thu Apr 9 19:11:27 2009 +1000 130883 130884 Xi: silence compiler warnings about "wrong" event types. 130885 130886 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130887 130888commit f2e957c5342aefc8329e9fe282b7072f767fcd28 130889Author: Peter Hutterer <peter.hutterer@who-t.net> 130890Date: Thu Apr 9 19:07:36 2009 +1000 130891 130892 mi: add prototype for CopyGetMasterEvent. 130893 130894 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130895 130896commit d80798ba2ae018086aaf230fcd8bc71d3bf06b11 130897Author: Peter Hutterer <peter.hutterer@who-t.net> 130898Date: Thu Apr 9 20:27:00 2009 +1000 130899 130900 mi: fix wrong (*EnqueueEvent) declaration in miPointerScreenFuncRec. 130901 130902 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130903 130904commit 4dedb900f4fbd3bdc9538b1e82cab2044438a770 130905Author: Peter Hutterer <peter.hutterer@who-t.net> 130906Date: Thu Apr 9 08:42:06 2009 +1000 130907 130908 xfree86: shut up compiler warnings - typecast to InternalEvent 130909 130910 Reported-by: Eric Anholt <eric@anholt.net> 130911 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130912 130913commit 4fa7fcf610d2ab91c1158f8034bbc0531060a86f 130914Author: Peter Hutterer <peter.hutterer@who-t.net> 130915Date: Thu Apr 9 08:40:20 2009 +1000 130916 130917 mi: fix compiler warning - explicitly typecast to InternalEvent. 130918 130919 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130920 130921commit 8bf2f562678107c346e8664f18c10f315c2768f5 130922Author: Peter Hutterer <peter.hutterer@who-t.net> 130923Date: Thu Apr 9 08:14:39 2009 +1000 130924 130925 dix: fix dev/keybd variable mixup. 130926 130927 Reported-by: Eric Anhold <eric@anholt.net> 130928 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130929 130930commit 012cb1956a69f6cf2a878c5331e792fb12c73e9d 130931Author: Peter Hutterer <peter.hutterer@who-t.net> 130932Date: Thu Apr 9 08:06:32 2009 +1000 130933 130934 xfree86: fix use of uninitialized variable in DGAProcessPointerEvent. 130935 130936 Reported-by: Eric Anholt <eric@anholt.net> 130937 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130938 130939commit 73b01a9aac5fde46ccd90c08a070e9d2496092b9 130940Author: Simon Thum <simon.thum@gmx.de> 130941Date: Wed Apr 8 14:35:01 2009 +0200 130942 130943 dix: fix pointer accelerations remainder handling 130944 130945 This didn't really work as intended, but did amazingly well thanks 130946 to roundf() hiding the defect. Cheers! 130947 130948 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130949 130950commit 4318075140cc287871d3c3b9f777289ea4ffa23a 130951Author: Peter Hutterer <peter.hutterer@who-t.net> 130952Date: Tue Apr 7 21:51:02 2009 +1000 130953 130954 dix: store subpixel precision and send it down the wire to the client. 130955 130956 For the valuator data, not yet for root x/y and event x/y. 130957 130958 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130959 130960commit 9212948461c5e053abb61fff73ced4a00f138544 130961Author: Simon Thum <simon.thum@gmx.de> 130962Date: Sat Mar 21 18:19:19 2009 +0100 130963 130964 dix: allow relative motion buffer to accumulate in a natural way 130965 130966 Since with XI2 we pass sub-pixel motion, anything else is broken. 130967 130968 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 130969 130970commit ad76656f8869e2065f0c4e66cfbeef0b42c61769 130971Author: Adam Jackson <ajax@redhat.com> 130972Date: Fri Apr 17 17:46:58 2009 -0400 130973 130974 randr: Accept gamma set requests from XF86VidMode clients too 130975 130976commit d9bf52b4abd29a3c206cd1e765b680659ddac1c6 130977Author: David Jander <david.jander@protonic.nl> 130978Date: Fri Apr 17 01:34:18 2009 -0400 130979 130980 [kdrive] Fix rotation of pointer 130981 130982 Rotation matrix for pointer coordinates was incomplete and pointers with 130983 absolute coordinates did not work correctly in xserver (kdrive) when the 130984 sceen was rotated other than by 0 degrees. 130985 130986 Signed-off-by: David Jander <david.jander@protonic.nl> 130987 Signed-off-by: James Cloos <cloos@jhcloos.com> 130988 130989commit 4559d2ace6ac55fe361f572ded0769cdd1f3b545 130990Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 130991Date: Thu Apr 16 22:48:11 2009 -0400 130992 130993 security: Grant untrusted windows remove access on all windows. 130994 130995 This allows untrusted clients to destroy their own windows when they 130996 have been reparented by a trusted window manager. 130997 130998commit 3481b32ab971c41cb972f6819ae049f3e9f7033b 130999Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 131000Date: Thu Apr 16 22:39:54 2009 -0400 131001 131002 security: Fix a crash caused by wrong ordering of format arguments. 131003 131004commit 6045506be0cebca4ebbe943ae77f020aafa703d4 131005Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 131006Date: Thu Apr 16 22:33:12 2009 -0400 131007 131008 security: Revert behavior of extension access for compatibility. 131009 131010 Previously, three extensions were defined as "trusted" by the extension: 131011 BIG-REQUESTS, XC-MISC, and XPrint. No other extensions were permitted 131012 to be used by untrusted clients. 131013 131014 In commit 8b5d21cc1d1f4e9d20e5d5eca44cb1e60a419763 this was changed for 131015 some reason. Return to the old, compatible behavior. 131016 131017commit 56a5955c8cd87137248edb2cbc65d384376d72ad 131018Author: Peter Hutterer <peter.hutterer@who-t.net> 131019Date: Tue Apr 14 17:05:04 2009 +1000 131020 131021 xkb: strdup the values returned by XkbGetRulesDflts 131022 131023 XkbGetRulesDftls may get a copy of what will later be freed when passed into 131024 XkbSetRulesDftls. 131025 131026 On the second run of XkbGet/SetRulesDflts: 131027 XkbGetRulesDflts(rmlvo) 131028 rmlvo->rules = current-rules 131029 131030 XkbSetRulesDflts(rmlvo) 131031 free(current-rules) 131032 current-rules = strdup(rmlvo->rules) 131033 131034 Leaving us with garbage in current-rules. 131035 131036 This patch requires callers of XkbGetRulesDflts to free the associated memory. 131037 131038 See also 131039 http://lists.freedesktop.org/archives/xorg-devel/2009-February/000305.html 131040 131041 Reported-by: Benjamin Close <Benjamin.Close@clearchain.com> 131042 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131043 Acked-by: Benjamin Close <Benjamin.Close@clearchain.com> 131044 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 131045 131046commit b406886bbffadaa52864a99f2a0520999eadc15d 131047Author: Peter Hutterer <peter.hutterer@who-t.net> 131048Date: Tue Apr 14 16:57:29 2009 +1000 131049 131050 input: allow NULL as XkbRMVLOSet in InitKeyboardDeviceStruct. 131051 131052 Virtually all callers use 131053 XkbGetRulesDefault(&rmlvo); 131054 InitKeyboardDeviceStruct(..., rmlvo); 131055 131056 Let's save them the trouble and accept NULL as a hint to take the 131057 default RMLVO. 131058 131059 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131060 Acked-by: Benjamin Close <Benjamin.Close@clearchain.com> 131061 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 131062 131063commit 4e4e263bc073bf452f19c932b937c4881ae71f64 131064Author: Peter Hutterer <peter.hutterer@who-t.net> 131065Date: Wed Apr 15 22:16:49 2009 +1000 131066 131067 dix: remove un-used parameter "core" from AllowSome 131068 131069 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131070 131071commit dd6e8a14ec1c8f4ed9c51ca2764261e6e48d13b3 131072Author: Peter Hutterer <peter.hutterer@who-t.net> 131073Date: Sat Apr 11 11:19:29 2009 +1000 131074 131075 Xi: fix a typo in a #ifdef 131076 131077 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131078 131079commit de1e43181bd670877b994db221ad8a04b5d63324 131080Author: Ian Romanick <ian.d.romanick@intel.com> 131081Date: Wed Apr 15 11:13:48 2009 -0700 131082 131083 DRI2: Don't leave empty entries in private->buffers 131084 131085 This should fix bug #21130. 131086 131087 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 131088 131089commit b3e3154cce47add97f5561088036ce3b9e7dc937 131090Author: Robert Noland <rnoland@2hip.net> 131091Date: Wed Apr 15 12:06:19 2009 -0500 131092 131093 One = is more than adequate here. Make is sh safe. 131094 131095commit 74d27c8b5bac7c8d2ed02ba86e09bf09924ce05c 131096Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 131097Date: Wed Apr 15 15:44:17 2009 +0200 131098 131099 Fix build on hurd-i386 131100 131101 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=523970 131102 131103commit faf7dfa099f5b42a703313fbd1bf8afdad07a179 131104Author: Dave Airlie <airlied@linux.ie> 131105Date: Wed Apr 15 18:26:06 2009 +1000 131106 131107 randr12: looking up these bits if randr isn't initialised is bad. 131108 131109 When xinerama is enabled we don't get randr protocol, but the 131110 driver might still want randr internals 131111 131112commit efa31092d6703397121a0ada4f7205a8ecad3d3d 131113Author: Ander Conselvan de Oliveira <ander@mandriva.com.br> 131114Date: Mon Apr 6 16:01:20 2009 -0300 131115 131116 xfree86: Remove device from inputInfo.devices if ActivateDevice failed. 131117 131118 After the call to xf86ActivateDevice, the new device will be added to 131119 inputInfo.devices. However, if the subsequent call to ActivateDevice 131120 fails, the correponding InputInfoRec for the device is deleted but an 131121 entry still remains in inputInfo.devices. This might lead to a server 131122 crash later on (on InitAndStartDevices for instance) when the device 131123 control proc would be called for an invalid device. 131124 131125 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131126 131127commit d79bad0aa70403ead8ec87bac8463a6e2005802c 131128Author: Peter Hutterer <peter.hutterer@who-t.net> 131129Date: Mon Apr 13 17:49:00 2009 +1000 131130 131131 xfree86: don't synthesise a mouse section if synaptics devices are found. 131132 131133 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131134 131135commit 4a27618565f3151ab17f0ca9ecbde12fa7ba13c0 131136Author: Kristian Høgsberg <krh@redhat.com> 131137Date: Tue Apr 14 16:21:04 2009 -0400 131138 131139 composite: Fix resource lookups 131140 131141commit e7785e8af3e34f9d1089d8499d16802984ab9823 131142Author: Kristian Høgsberg <krh@redhat.com> 131143Date: Tue Apr 14 14:24:31 2009 -0400 131144 131145 xfixes: Fix a couple of resource lookups 131146 131147commit 0eb19f9437b7d8c19592e49eedb028771d300d80 131148Author: Adam Jackson <ajax@redhat.com> 131149Date: Tue Apr 14 10:54:25 2009 -0400 131150 131151 xdmcp: Don't crash on X -query with more than 255 IP addresses. (#20675) 131152 131153 You could be more clever than this, but the wire protocol says this 131154 really is an array of not more than 255 ARRAY8, so it's not just a 131155 matter of changing the types. 131156 131157commit f0543ae4ec0fcb5d696e7b2983653bd779f1eddc 131158Author: Eric Anholt <eric@anholt.net> 131159Date: Wed Apr 8 14:17:41 2009 -0700 131160 131161 Fix unused var warning from pci cleanups. 131162 131163commit 181cc08c8908a119fc403f970dea8cc98d3e0b9b 131164Author: Eric Anholt <eric@anholt.net> 131165Date: Wed Apr 8 14:17:40 2009 -0700 131166 131167 Add shave so that we can see the steaming piles of warnings generated. 131168 131169 The old style output can be reenabled for build system debugging using 131170 "make V=1", or --disable-shave at configure time. 131171 131172commit 4474c200a102feda72f9572a96cb588009aa0147 131173Author: Eric Anholt <eric@anholt.net> 131174Date: Wed Apr 8 14:17:39 2009 -0700 131175 131176 Move VENDOR_* defines from AC_SUBST to a header to avoid angering shave. 131177 131178 This is more sane anyway, as it ensures a rebuild when changing them. 131179 131180commit e72c85547b405fbd0117dc1236e5ca5a2126063c 131181Author: Eric Anholt <eric@anholt.net> 131182Date: Wed Apr 8 14:17:38 2009 -0700 131183 131184 Move contributed m4 to a subdir so we can more easily update contributions. 131185 131186commit bef1cfb2395df47458159a0b7ae27b9db15025ef 131187Author: Peter Hutterer <peter.hutterer@who-t.net> 131188Date: Tue Apr 14 16:54:42 2009 +1000 131189 131190 xnest: remove unused variable 'names'. 131191 131192 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131193 131194commit 7b6400a1b8d2f228fcbedf17c30a7e3924e4dd2a 131195Author: Kristian Høgsberg <krh@redhat.com> 131196Date: Thu Apr 9 13:16:37 2009 -0400 131197 131198 glx: Fix drawable private leak on destroy 131199 131200 When a drawable goes away, we don't destroy the GLX drawable in full, 131201 since it may be current for a context. This means that when the drawable 131202 is destroyed in full later, the backend doesn't get a chance to 131203 destroy resources associated with the drawable (the DRI2Drawable). 131204 131205 With this patch, we destroy the GLX drawable in full when it goes away 131206 and then track down all contexts that reference it and NULL their 131207 pointers. 131208 131209commit 140463a197fb93d0a4bfad924efc35b860e8cc54 131210Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131211Date: Sat Apr 11 16:53:59 2009 -0700 131212 131213 XQuartz: Use correct values for ProximityIn and ProximityOut 131214 131215 This was the other underlying cause of teh 100% CPU tablet issue. 131216 (cherry picked from commit a9cecf34c23fbcd59b56b380c51d31a9fabc3eb7) 131217 131218commit 80a7bb2605f9b439d6221c0495a629a39177a018 131219Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131220Date: Sat Apr 11 17:26:07 2009 -0700 131221 131222 XQuartz: Solve the tablet 100% CPU bug 131223 131224 This happened because we put a byte in the fd to wake up dispatch, but we never actually enqueued anything in mieq because the num_events was 0. 131225 (cherry picked from commit c21ca7558d2faf93c61f5feaafd7c878e9e21942) 131226 131227commit 6f8f7c78f1b722bc70a0ea8f6340116a1e09e858 131228Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131229Date: Sat Apr 11 16:13:42 2009 -0700 131230 131231 XQuartz: Re-enable background window checking 131232 131233 since that code was not the culprit for the wacom tablet, background 100% CPU bug 131234 (cherry picked from commit fc1dc5d71b2a488a8a94d953dd8e67353161a590) 131235 131236commit 8522a759c9e78478bb399b91d3a0af2d23ea6766 131237Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131238Date: Sat Apr 11 16:12:23 2009 -0700 131239 131240 XQuartz: Only set MotionNotify on activation if it is updated. 131241 (cherry picked from commit ae8077a251ef27381a755d57ff974767bda16148) 131242 131243commit 5e55becddccc376ab7338789562ad6d6cd750de9 131244Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131245Date: Sat Apr 11 14:23:41 2009 -0700 131246 131247 XQuartz: Make sure the Fn doesn't trigger unneccessary calls to DarwinUpdateModKeys() 131248 (cherry picked from commit 70a18558c6b0a02b633fd8974f002cdf3cdc713e) 131249 131250commit bd1d9179094657865b0606ed0ac835a8b8df3be1 131251Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131252Date: Sat Apr 11 13:53:38 2009 -0700 131253 131254 XQuartz: Re-enable Fn as an option for 3button mouse simulation. 131255 131256 Patch from Martin Otte <otte@duke.edu> 131257 (cherry picked from commit b5ec3be6b5449b5d575bc1472fdd1c9cb15cb8be) 131258 131259commit 1b5758bef0840c6614244e321790231b3c9477c9 131260Author: Maarten Maathuis <madman2003@gmail.com> 131261Date: Thu Apr 9 15:45:57 2009 +0200 131262 131263 exa: implement UTS based upload through CopyArea 131264 131265 - Some image viewers (eog, gqview) trigger the CopyArea path of Xext/shm.c 131266 - I'm not aware of any code path that wouldn't like UTS and trigger this code. 131267 - miDoCopy should handle src coordinate clipping. 131268 - Overlapping blits are obviously not an issue (both would have to be offscreen or not). 131269 131270commit 567cf67959b30432ae30f4851ec17b3a375ab838 131271Author: Ian Romanick <ian.d.romanick@intel.com> 131272Date: Thu Apr 9 14:38:24 2009 -0700 131273 131274 DRI2: Synchronize the contents of the real and fake front-buffers 131275 131276 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 131277 131278commit f1a995d1496d73741731e32f475097c44a8da972 131279Author: Ian Romanick <ian.d.romanick@intel.com> 131280Date: Thu Apr 9 14:31:01 2009 -0700 131281 131282 DRI2: Do not send the real front buffer of a window to the client 131283 131284 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 131285 131286commit aa2928325fe51d94a636dde9c090e8f54a311a12 131287Author: Ian Romanick <ian.d.romanick@intel.com> 131288Date: Wed Apr 8 15:44:34 2009 -0700 131289 131290 DRI2: Add fake front-buffer to request list for windows 131291 131292 If a front-buffer is requested for a window, add the fake front-buffer 131293 to the list of requested buffers. 131294 131295 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 131296 131297commit 808fd2c67f303cb721769375b11ce8b90ffc1909 131298Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131299Date: Thu Apr 9 20:00:37 2009 -0700 131300 131301 XQuartz: xprSetWindowLevel updated to store the level requested by the WM 131302 (cherry picked from commit c28c2ddc3a8f3c5b9beec396953bb3ac9ee4714b) 131303 131304commit dfb0d7aefbbdfc0db966e3a84d52f638135d9138 131305Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131306Date: Thu Apr 9 18:51:22 2009 -0700 131307 131308 XQuartz: Update window levels when changing rootless state 131309 (cherry picked from commit 1359ded5bfc14a80fb998b01a54ecacb96c4ff88) 131310 131311commit d5ef88d7543ed787093099ab18db766c446b47d1 131312Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131313Date: Thu Apr 9 18:27:22 2009 -0700 131314 131315 XQuartz: Fix window levels for rooted mode to allow showing the menu bar. 131316 (cherry picked from commit 80759a4186bf0335edc85aecea2faf11fe09f491) 131317 131318commit 15146b863759640e7a73fab2301fb28ef4dead84 131319Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131320Date: Thu Apr 9 17:47:41 2009 -0700 131321 131322 XQuartz: Properly set the window level for the root window 131323 (cherry picked from commit bdf9286d1cbfeaaf8eaf03d28091e91ee587ee25) 131324 131325commit 8d5dcfe2154f217bd8fde5509d78f3383add8725 131326Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131327Date: Thu Apr 9 04:36:26 2009 -0700 131328 131329 XQuartz: Properly set the menu bar and hotkey state when changing rootless mode. 131330 131331 Currently no code path exhibits the broken behavior since we only toggle into rootless if we don't have the root. 131332 (cherry picked from commit 970f100ca3c5fc0662ae7658d49d118fbd9de943) 131333 131334commit 82d7cf5cdcbc5e451a87f8c2f64cd5d2e3627d54 131335Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131336Date: Thu Apr 9 03:55:13 2009 -0700 131337 131338 XQuartz: In rooted mode, make sure we start in the hidden state. 131339 (cherry picked from commit 5ecc497f71c2133f773f6c56ad76cb778862ddd6) 131340 131341commit e86f4e93020d56385418850a9eebae8076dcb9ac 131342Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131343Date: Mon Apr 6 21:34:14 2009 -0700 131344 131345 XQuartz: Send MotionNotify before button presses when X11 is in the background 131346 (cherry picked from commit c80d0ec18ef5b842447d31360406d0b5b9424222) 131347 131348commit 5ccfad8df099e8ebc4bf2dd53c3db1460903b135 131349Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131350Date: Mon Apr 6 19:21:46 2009 -0700 131351 131352 XQuartz: Revert most of the previous override redirect patch 131353 131354 The changes actually caused all windows to move to the current space. Instead, we're going with a fix entirely within Xplugin that depends on quartz-wm being the window-manager for now. 131355 (cherry picked from commit 997b6f3142c622541bb5bac98652abae75d1101d) 131356 131357commit ff7f019bbcbc52618cc478db7baed57aa5b7c3d3 131358Author: Simon Thum <simon.thum@gmx.de> 131359Date: Wed Apr 1 10:53:10 2009 +0200 131360 131361 dix: correctly utilize tracker buffer and protect from timer overruns 131362 131363 two small related fixes hard to split up 131364 131365 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131366 131367commit 06aebecb19dd9d90d73b742a09b6068b862f1d05 131368Author: Simon Thum <simon.thum@gmx.de> 131369Date: Wed Apr 8 14:35:01 2009 +0200 131370 131371 dix: fix pointer accelerations remainder handling 131372 131373 This didn't really work as intended, but did amazingly well thanks 131374 to roundf() hiding the defect. Cheers! 131375 131376 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131377 131378commit 98f4179156391752e6688339487458ad7828abf4 131379Author: Alan Coopersmith <alan.coopersmith@sun.com> 131380Date: Thu Mar 26 23:04:24 2009 -0700 131381 131382 Use RTLD_DI_SETSIGNAL to catch runtime dynamic loader errors and clean up 131383 131384 Based on fix for Sun bug 6813925: Xorg needs to catch ld.so.1 failure 131385 so it can close down devices cleanly 131386 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6813925> 131387 131388 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 131389 131390commit a0b6a363dca8ce0dc6f4eb79333e48496153cd67 131391Author: Alan Coopersmith <alan.coopersmith@sun.com> 131392Date: Mon Mar 9 13:22:57 2009 -0700 131393 131394 Lift fatal signal handlers from DDX'es up to a common DIX implementation 131395 131396 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 131397 131398commit fcc19e673e3ef33d64916dd933853f8aa667c4d7 131399Author: Peter Hutterer <peter.hutterer@who-t.net> 131400Date: Thu Apr 9 20:27:00 2009 +1000 131401 131402 mi: fix wrong (*EnqueueEvent) declaration in miPointerScreenFuncRec. 131403 131404 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131405 131406commit 4fee979d0632751d3d54d2115e84e9654edf0622 131407Author: Peter Hutterer <peter.hutterer@who-t.net> 131408Date: Thu Apr 9 19:11:27 2009 +1000 131409 131410 Xi: silence compiler warnings about "wrong" event types. 131411 131412 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131413 131414commit 6b467bf879eeb77d167ef321e6dda97ca9d7010a 131415Author: Peter Hutterer <peter.hutterer@who-t.net> 131416Date: Thu Apr 9 08:42:06 2009 +1000 131417 131418 xfree86: shut up compiler warnings - typecast to InternalEvent 131419 131420 Reported-by: Eric Anholt <eric@anholt.net> 131421 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131422 131423commit 8a2a184da78a3e9cbeae8290431f40d5ec7f3636 131424Author: Peter Hutterer <peter.hutterer@who-t.net> 131425Date: Thu Apr 9 08:06:32 2009 +1000 131426 131427 xfree86: fix use of uninitialized variable in DGAProcessPointerEvent. 131428 131429 Reported-by: Eric Anholt <eric@anholt.net> 131430 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131431 131432commit 011cee3103c146c8096b7098a27993f99e07a824 131433Author: Peter Hutterer <peter.hutterer@who-t.net> 131434Date: Thu Apr 9 08:14:39 2009 +1000 131435 131436 dix: fix dev/keybd variable mixup. 131437 131438 Reported-by: Eric Anhold <eric@anholt.net> 131439 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131440 131441commit 630a6e9d14ffbf036fa72f580c72c0172d7c20bd 131442Author: Peter Hutterer <peter.hutterer@who-t.net> 131443Date: Thu Apr 9 08:40:20 2009 +1000 131444 131445 mi: fix compiler warning - explicitly typecast to InternalEvent. 131446 131447 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131448 131449commit e3bb7dea06b9ab0e30d801bf6c3a59f94290aaed 131450Author: Peter Hutterer <peter.hutterer@who-t.net> 131451Date: Thu Apr 9 19:07:36 2009 +1000 131452 131453 mi: add prototype for CopyGetMasterEvent. 131454 131455 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131456 131457commit 346e71525fc545c6ca4ad79425722282d1544459 131458Author: Michel Dänzer <daenzer@vmware.com> 131459Date: Thu Apr 9 09:36:41 2009 +0200 131460 131461 EXA: If the driver can't composite to an a8 mask, try an argb mask for glyphs. 131462 131463 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 131464 131465commit 682d7b55699cacbb2dbcd84a5e816bf6e2d2f02a 131466Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 131467Date: Thu Apr 9 02:48:04 2009 -0400 131468 131469 xselinux: Don't BadAlloc in List* requests if there are no items to list. 131470 131471commit 3a0ee199dcec39596756a995996eac388acf6315 131472Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 131473Date: Thu Apr 9 02:26:24 2009 -0400 131474 131475 config: fix crash caused by strdup(NULL) 131476 131477commit 7b3982eb6518da33ab01c2fbf7ceb45b89f841df 131478Author: Michel Dänzer <daenzer@vmware.com> 131479Date: Thu Apr 9 08:21:09 2009 +0200 131480 131481 glx: Test the error value, not its address... 131482 131483commit 03aebed519986c4dd03e02b3b3d4af1f64595ca7 131484Author: Ian Romanick <ian.d.romanick@intel.com> 131485Date: Wed Apr 8 14:54:30 2009 -0700 131486 131487 Use a #define instead of a magic number 131488 131489 The number of buffers is likely to change in the future, so having 131490 this as a define is the right way to go. 131491 131492 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 131493 131494commit 0d9f3ca7eabd4c514808114d30627f682c8bd030 131495Author: Ian Romanick <ian.d.romanick@intel.com> 131496Date: Wed Apr 8 14:53:46 2009 -0700 131497 131498 Allow GLX sources to build against Mesa 7.4 sources 131499 131500 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 131501 131502commit e8b324102f6e21ae2b8292a6f50d016dd6254dd6 131503Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 131504Date: Wed Apr 8 15:10:16 2009 -0400 131505 131506 xselinux: Don't require incoming context strings to be null-terminated. 131507 131508commit df27b870a8db7a5153b18a556fe77efa590f9eee 131509Author: Kristian Høgsberg <krh@redhat.com> 131510Date: Tue Apr 7 16:28:08 2009 -0400 131511 131512 Convert remaining GLX LookupIDByType() calls 131513 131514commit 92562747a0fdbef1dbedf734cb55dd6a9e1d2994 131515Author: Kristian Høgsberg <krh@redhat.com> 131516Date: Tue Apr 7 13:58:53 2009 -0400 131517 131518 Add validGlxDrawable() and use dixLookupResourceByType(). 131519 131520 Fixes deprecation warnings, and fixes a couple of GLX error codes 131521 for failing drawable lookups. 131522 131523commit f70cfc8f90091ef0f5ed0a5b2e023e7fd6369b36 131524Author: Kristian Høgsberg <krh@redhat.com> 131525Date: Thu Apr 2 17:30:19 2009 -0400 131526 131527 Don't stomp on dixLookupDrawable() return value in DoCreateGLXPixmap(). 131528 131529commit 30d81ad72e870cc37754bd8c8aadf605450ec16e 131530Author: Kristian Høgsberg <krh@redhat.com> 131531Date: Thu Apr 2 17:24:12 2009 -0400 131532 131533 Make GLX context lookup use dixLookupResourceByType() 131534 131535commit 66539cc05d0b017b9feb4a038499907810140623 131536Author: Alan Coopersmith <alan.coopersmith@sun.com> 131537Date: Thu Mar 26 22:22:32 2009 -0700 131538 131539 Don't leak default font path when appending built-ins 131540 131541 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 131542 131543commit 6c3b633299f12051fcf37fb8439f358de876cf03 131544Merge: adf21dba7 0cfd48130 131545Author: Peter Hutterer <peter.hutterer@who-t.net> 131546Date: Tue Apr 7 19:36:27 2009 +1000 131547 131548 Merge branch 'master' into xi2 131549 131550commit 0cfd48130842685828e0662a27272a3a65facd23 131551Author: Peter Hutterer <peter.hutterer@who-t.net> 131552Date: Mon Mar 23 11:41:16 2009 +1000 131553 131554 dix: Dont change the keyboard mapping on non-keyboard devices. 131555 131556 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131557 131558commit 69e73e5ce0e3971eceed71216d2d7e7d9cd57cc2 131559Author: Adam Jackson <ajax@redhat.com> 131560Date: Mon Apr 6 17:56:49 2009 -0400 131561 131562 config: Remove useless xf86conf{{m,c,re}alloc},free} macros 131563 131564commit 485946120a6522db9298111752de9be3d675e0b3 131565Author: Adam Jackson <ajax@redhat.com> 131566Date: Mon Apr 6 17:51:35 2009 -0400 131567 131568 config: s/xf86configStrdup/strdup/ 131569 131570commit 5f769b73441cf29c05a05c728ab81bdde1fc00b6 131571Author: David Jander <david.jander@protonic.nl> 131572Date: Mon Apr 6 13:01:08 2009 -0400 131573 131574 kdrive: Fix segfault in tslib support 131575 131576commit 843166b033dc0544b9f3cf3c91fc3ae3650bda14 131577Author: Adam Jackson <ajax@redhat.com> 131578Date: Mon Apr 6 11:05:17 2009 -0400 131579 131580 os: signal handlers return void. 131581 131582commit aa6fbc2a3679896181610aee7ce2844d4322a5ce 131583Author: Adam Jackson <ajax@redhat.com> 131584Date: Mon Apr 6 10:58:47 2009 -0400 131585 131586 Remove some OS/2 leftovers. 131587 131588commit 2c03a63ff6855272d587b4972a67f64115a1ce4c 131589Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131590Date: Sun Apr 5 03:20:40 2009 -0700 131591 131592 XQuartz: Use updated Xplugin API to send overide-redirect windows to the current space when they're ordered in 131593 131594 This fixes the annoying "using a menu moves me to another space" bug 131595 (cherry picked from commit 9e6dab89e08a26d764ba1aeaeb804c3d25c667da) 131596 131597commit 67621f3b39c387b3f7bfc65781051aea51c0b83f 131598Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131599Date: Sun Apr 5 00:06:57 2009 -0700 131600 131601 XQuartz: Send a MotionNotify event for the mouse cursor when activating X11.app 131602 131603 Otherwise if X11.app was activated with a mouse click, the location of the even is the last location of the cursor before X11 was deactivated 131604 (cherry picked from commit c7457d7b31ddd2ddfd04dd6294bed4532664e1ab) 131605 131606commit e1338d27141125af61634597c5dea880692c4017 131607Author: Julien Cristau <jcristau@debian.org> 131608Date: Sat Apr 4 23:43:28 2009 +0200 131609 131610 xfree86: Remove unused DEFAULT_UNRESOLVED and DEFAULT_BEST_REFRESH macros 131611 131612commit e25d55ad546905c4921f0a26385bdb9f22038429 131613Author: Adam Jackson <ajax@redhat.com> 131614Date: Fri Apr 3 22:38:00 2009 -0400 131615 131616 Revert accidental Makefile change from previous commit 131617 131618commit 472811dc4ba741ab09db184216878700f4adf745 131619Author: Adam Jackson <ajax@redhat.com> 131620Date: Fri Apr 3 22:27:51 2009 -0400 131621 131622 DPMS: Re-export the various DPMS variables. 131623 131624 The drivers might not need them but extmod does. Should move it to 131625 builtin though. 131626 131627commit cbb165ab88cb0810268001e84d87671440baf837 131628Author: Adam Jackson <ajax@redhat.com> 131629Date: Fri Apr 3 18:34:45 2009 -0400 131630 131631 os: Remove the useless -x option 131632 131633commit 6574ab092fb80b7731e06ede653b6af92fa7684e 131634Author: Adam Jackson <ajax@redhat.com> 131635Date: Fri Apr 3 18:03:43 2009 -0400 131636 131637 DPMS: Remove the defaultDPMS* variables 131638 131639commit c1cf36ee29e41901ecd382f0258110ebb4a0124e 131640Author: Adam Jackson <ajax@redhat.com> 131641Date: Fri Apr 3 17:51:55 2009 -0400 131642 131643 DPMS: Simplify command line parsing 131644 131645commit 2d0a4fff7985f58c21023cc4757090ee67067593 131646Author: Adam Jackson <ajax@redhat.com> 131647Date: Fri Apr 3 13:44:37 2009 -0400 131648 131649 DPMS: Remove unused DPMSGet() 131650 131651commit d52fddefaeb43f15b677eefbea4a288a9948373b 131652Author: Adam Jackson <ajax@redhat.com> 131653Date: Fri Apr 3 13:33:03 2009 -0400 131654 131655 DPMS: Align the default timeouts with the default screensaver timeout. 131656 131657 On a typical LCD, a black screensaver is actually worse for power 131658 consumption than a normal screen, because it takes more energy to turn 131659 the crystals opaque. Also, the intermediate DPMS states are essentially 131660 useless and most monitors alias them to the 'off' state, so we may as 131661 well do the same. 131662 131663 As a pleasant side effect, this brings the default DPMS timeouts in line 131664 with the EnergyStar Program Requirements for Computers: 131665 131666 http://www.energystar.gov/index.cfm?c=revisions.computer_spec 131667 131668 which state that products must be "shipped with the display's Sleep mode 131669 set to activate within 15 minutes of user inactivity". 131670 131671commit 38576dc3ae8ddec2baedd2f7e64abdfc56bedf93 131672Author: Adam Jackson <ajax@redhat.com> 131673Date: Fri Apr 3 11:55:41 2009 -0400 131674 131675 DPMS: Simplify some macro silliness. 131676 131677commit 1c01127a214731a21c4784249a429c74f17bd424 131678Author: Adam Jackson <ajax@redhat.com> 131679Date: Fri Apr 3 11:39:18 2009 -0400 131680 131681 DPMS: Unexport all the various setup variables. 131682 131683 The drivers don't need to know any of this. 131684 131685commit a99b0ab918998c9a1b9f8692591b8cb59a12eb9e 131686Author: Adam Jackson <ajax@redhat.com> 131687Date: Fri Apr 3 11:32:00 2009 -0400 131688 131689 DPMS: Fix cast abuse 131690 131691commit 4da347adb1fcc39e8f4fb2cadd98e887d7ac877a 131692Author: Adam Jackson <ajax@redhat.com> 131693Date: Fri Apr 3 11:18:21 2009 -0400 131694 131695 DPMS: Code motion. 131696 131697commit 1731882341c191b4ffd78e0c1fd9297c636e0401 131698Author: Havoc Pennington <hp@pobox.com> 131699Date: Fri Mar 27 12:30:37 2009 -0400 131700 131701 Set bg pixmap of composite overlay window to None (#20912) 131702 131703 Otherwise it's impossible to get the COW without a white 131704 flash on the screen, because it's on top, mapped immediately, 131705 and unaffected by composite redirection. This makes 131706 initial login ugly when it doesn't need to be. 131707 131708commit d2690375dfd994817f004cda133ca2a492c0b956 131709Author: Adam Jackson <ajax@redhat.com> 131710Date: Fri Apr 3 11:06:50 2009 -0400 131711 131712 kdrive: s/KdSaveString/strdup/g 131713 131714 This isn't even funny anymore. 131715 131716commit 91b697efdefba125348dbcaf584ee51a7f8c9bf6 131717Author: Kristian Høgsberg <krh@redhat.com> 131718Date: Wed Apr 1 17:42:33 2009 -0400 131719 131720 Support setTexBuffer2 in AIGLX. 131721 131722 Fixes broken GLX_tfp, specifically, lets compositors ignore un-defined 131723 alpha channel for pixmaps. 131724 131725commit 1c101d75d4855b2698e3fc8d2dd662f20585812f 131726Author: Alan Coopersmith <alan.coopersmith@sun.com> 131727Date: Thu Mar 26 22:25:08 2009 -0700 131728 131729 Don't leak canonical module name and patterns if module is built-in 131730 131731 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 131732 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 131733 131734commit 9180081ec32b6c655d193de4c89b59d222a72579 131735Author: Jeremy Huddleston <jeremy@yuffie.local> 131736Date: Wed Apr 1 13:45:02 2009 -0700 131737 131738 XQuartz: Still send mouse events while X11 is in the background if we have test extensions enabled 131739 (cherry picked from commit f393fb686cfbedb11a1539d4eaf5c491af8b3a64) 131740 131741commit 029307e79a14cad5bde8f1f4a5d0a2f740271f06 131742Author: Jeremy Huddleston <jeremy@yuffie.local> 131743Date: Wed Apr 1 13:07:15 2009 -0700 131744 131745 XQuartz: Return BadRequest when SendPSN isn't implemented rather than success 131746 (cherry picked from commit 49871e99179d5176a1ec516c8f449e2dc219faea) 131747 131748commit 0603fbc1d3b2db5c25faabbfb5e04914110b4d0e 131749Author: Jeremy Huddleston <jeremy@yuffie.local> 131750Date: Wed Apr 1 12:46:51 2009 -0700 131751 131752 XQuartz: Pad xAppleDRINotifyEvent to 32bytes to match sizeof(xEvent) 131753 (cherry picked from commit 276ee3bb087c772f57a987519aa4d94afeceadaa) 131754 131755commit b1dab580bdfb4acfe3feddeda6e760098ec4922a 131756Author: Tormod Volden <debian.tormod@gmail.com> 131757Date: Wed Apr 1 19:32:21 2009 +0200 131758 131759 xfree86: edid quirk for Philips LCD LP154W01-TLAJ 131760 131761 This panel reports its vertical size in cm. 131762 131763 X.Org bug#21000 <http://bugs.freedesktop.org/show_bug.cgi?id=21000> 131764 131765 Signed-off-by: Tormod Volden <debian.tormod@gmail.com> 131766 Signed-off-by: Julien Cristau <jcristau@debian.org> 131767 131768commit 7d94414cf4b16501ce4c4291846498f393f150a4 131769Author: Adam Jackson <ajax@redhat.com> 131770Date: Wed Apr 1 09:42:06 2009 -0400 131771 131772 input: Remove xf86ReloadInputDevs hack 131773 131774 We have input hotplug now, no need to fake it. 131775 131776commit 3311ef6ec93df4325d198715e0fce73af19f0781 131777Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131778Date: Tue Mar 31 18:11:19 2009 -0700 131779 131780 XQuartz: Don't report mouse events while X11 is not the foreground application 131781 (cherry picked from commit 5d1dd35096334b59564d77caef7a97bd58845c5d) 131782 131783commit 89887fadc97f6fd5059886396a6d7cf61de1982f 131784Author: Adam Jackson <ajax@redhat.com> 131785Date: Tue Mar 31 15:09:57 2009 -0400 131786 131787 APM: Don't carp if not available. 131788 131789commit 2e145989d10ef454581f8aa60358ed3782fd6442 131790Author: Adam Jackson <ajax@redhat.com> 131791Date: Mon Mar 30 17:22:28 2009 -0400 131792 131793 dix: Simplify InitClient() 131794 131795commit 252ec504817e05b185e4896a2d899e9c00b8aeef 131796Author: Adam Jackson <ajax@redhat.com> 131797Date: Mon Mar 30 15:18:30 2009 -0400 131798 131799 Document which bits of ClientRec are currently unused 131800 131801commit 0b6e14d46feaf8c46fd3768ee5353b844391b79b 131802Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 131803Date: Mon Mar 30 20:37:25 2009 -0700 131804 131805 XQuartz: Fix mouse tracking for quake, et. al. in wine 131806 131807 Patch courtesy of Codeweavers 131808 131809 Fix mouse movement tracking. For a non-window-related mouse-move event, 131810 calculate the new position by adding the event's delta-x and delta-y values 131811 to the previous mouse position. Do not rely on the current mouse position 131812 because it may have been changed by a XWarpPointer call. 131813 (cherry picked from commit 7a67935b05a475215b9bdbb959e4f7e15f32416f) 131814 131815commit c74220b853cd99d03e33cc62d32431ca3a879483 131816Author: Adam Jackson <ajax@redhat.com> 131817Date: Mon Mar 30 14:56:15 2009 -0400 131818 131819 APM: Use general handlers, not input handlers. 131820 131821 Otherwise APM events get treated as input events, which messes up idle 131822 time accounting and screensavers and such. Not, we hope, that anyone 131823 is using APM anymore. 131824 131825commit 283a081572d8db787c77d09e5ba6bcadebf4f7fe 131826Author: Adam Jackson <ajax@redhat.com> 131827Date: Fri Mar 27 15:56:15 2009 -0400 131828 131829 selinux: Only activate if policy says to be an object manager 131830 131831commit 8e7facfe3013abda12a0c39ad8b4d025618077a7 131832Author: Adam Jackson <ajax@redhat.com> 131833Date: Fri Mar 27 15:48:17 2009 -0400 131834 131835 config: Fall back to vesa for Intel Poulsbo. 131836 131837commit 17239c8e3163fef0159835d43c0dde20e78192a1 131838Author: Adam Jackson <ajax@redhat.com> 131839Date: Fri Mar 27 12:44:20 2009 -0400 131840 131841 misprite: Remove private header, fold into misprite.c 131842 131843commit bb89fc2771d2e54d7e0cd0371db68ddb754a64e5 131844Author: Adam Jackson <ajax@redhat.com> 131845Date: Fri Mar 27 12:36:42 2009 -0400 131846 131847 misprite: Fix software cursor on multiple ScreenRecs 131848 131849commit 2c000f4980d30015e6da95de3c77c0fdda0090e4 131850Author: Adam Jackson <ajax@redhat.com> 131851Date: Fri Mar 27 11:54:52 2009 -0400 131852 131853 misprite: Do window check first for Get{Image,Spans} too 131854 131855commit 8beced6c02610444105f3d0b8ee3a49e1072096d 131856Author: Adam Jackson <ajax@redhat.com> 131857Date: Fri Mar 27 11:43:33 2009 -0400 131858 131859 misprite: RECT_IN_REGION -> miRectIn 131860 131861commit a232116b9ecd41ce55f8eeee9dfe2b67d76479d6 131862Author: Adam Jackson <ajax@redhat.com> 131863Date: Fri Mar 27 11:41:40 2009 -0400 131864 131865 misprite: Fix a typo 131866 131867commit 1f0eaf886d88f3935e43985d2c889b374f412e43 131868Author: Adam Jackson <ajax@redhat.com> 131869Date: Fri Mar 27 11:29:40 2009 -0400 131870 131871 misprite: Fix cast abuse 131872 131873commit e3c1096273446bb7eb993b179ceb989617d149cc 131874Author: Peter Hutterer <peter.hutterer@who-t.net> 131875Date: Mon Mar 23 13:20:22 2009 +1000 131876 131877 dix: build_modmap_from_modkeymap needs to bounds-check its argument. 131878 131879 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131880 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 131881 131882commit 45baef0b77f6f0e5e6230eca045dfcbb1e48c647 131883Author: Peter Hutterer <peter.hutterer@who-t.net> 131884Date: Mon Mar 23 13:32:32 2009 +1000 131885 131886 dix: return BadValue as error in SetModifierMapping. 131887 131888 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131889 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 131890 131891commit 603db34337a61754e0c5f71525011d10eab78411 131892Author: Peter Hutterer <peter.hutterer@who-t.net> 131893Date: Wed Mar 25 15:51:43 2009 +1000 131894 131895 Xext: set POINTER_SCREEN flag in XTestFakeInput if necessary. (RH #490984) 131896 131897 The POINTER_SCREEN flag must be set explicitly for XTest core events to avoid 131898 out-of-range events when the lastSlave was an SD with an explicit axis range. 131899 Device events sent through XTest don't need this flag, they are expected to be 131900 in the valuator range of the device anyway. 131901 131902 Red Hat Bug 490984 <https://bugzilla.redhat.com/show_bug.cgi?id=490984> 131903 131904 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131905 131906commit 4ab93f05664890ff6738ef1089bcd956ce3f06e9 131907Author: Adam Jackson <ajax@redhat.com> 131908Date: Mon Mar 23 15:19:47 2009 -0400 131909 131910 misprite: In SourceValidate, check that it's a Window first 131911 131912 It almost never is. However, you have _lots_ of input devices now, and 131913 walking them all on every Composite operation is not the cheapest thing 131914 in the world. 131915 131916commit d698e62690e0a52fa537f1cd792cab7a5b1bafdf 131917Author: Maarten Maathuis <madman2003@gmail.com> 131918Date: Thu Mar 19 21:35:25 2009 +0100 131919 131920 xkb: plug a memory leak in XkbCopySrvLedInfo (#20756) 131921 131922 X.Org Bug 20756 <http://bugs.freedesktop.org/show_bug.cgi?id=20756> 131923 131924 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131925 131926commit e26f79335bb7a9ad34fe9b23c283a214202f89ed 131927Author: Peter Hutterer <peter.hutterer@who-t.net> 131928Date: Tue Mar 17 16:22:00 2009 +1000 131929 131930 xkb: put a few extra checks in against non-keyboards 131931 131932commit 844bb629fe786ecf536039df3f5f4b3b214c9d40 131933Author: Peter Hutterer <peter.hutterer@who-t.net> 131934Date: Thu Mar 19 15:44:34 2009 +1000 131935 131936 dix: remove a truly useless a = (foo) ? bar : bar; statement. 131937 131938 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131939 131940commit 62183a06e5f72a18abc3d58c4bf5658a20ece217 131941Author: Tomas Carnecky <tom@dbservice.com> 131942Date: Fri Mar 20 16:19:33 2009 +0100 131943 131944 Remove two unused defines in C files 131945 131946 These two defines were defined in C files but not used anywhere: 131947 131948 dix/window.c #define DeviceEventMasks (KeyPressMask | [...] 131949 os/connection.c #define MAXFD 500 131950 131951 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 131952 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131953 131954commit 4e0d7cc506b661b5e5cedf8be6e17aa4c8075aa8 131955Author: Tomas Carnecky <tom@dbservice.com> 131956Date: Fri Mar 20 16:19:34 2009 +0100 131957 131958 FID, whatever that was, isn't anymore 131959 131960 No traces of FID in the xserver nor in the modules listed in 131961 util/modular/xorg.modules 131962 131963 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 131964 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131965 131966commit adf21dba7617542f08309415e315d4b2699c10e0 131967Author: Peter Hutterer <peter.hutterer@who-t.net> 131968Date: Thu Mar 19 20:40:43 2009 +1000 131969 131970 include: un-export a bunch of server-only functions. 131971 131972 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131973 131974commit 111ef10375204a6981a5be5c045860854c4da4e6 131975Author: Peter Hutterer <peter.hutterer@who-t.net> 131976Date: Thu Mar 19 09:33:02 2009 +1000 131977 131978 dix: move ProcGrabPointer guts into GrabDevice. 131979 131980 Yes, this means we have even more arguments to GrabDevice. But it beats having 131981 a copy of most but not all of GrabDevice in ProcGrabPointer. 131982 Also, reshuffle the order of parameters, the CARD* status is a return value 131983 and should be last. 131984 131985 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 131986 131987commit dc153271b6458cacd63a4bc0208594011c6460b0 131988Author: Peter Hutterer <peter.hutterer@who-t.net> 131989Date: Wed Mar 18 15:13:00 2009 +1000 131990 131991 Xi: purge old device enter/leave masks. 131992 131993commit 15a969c0fcd4f19f3c560a7037c1a6a58bce51f0 131994Author: Peter Hutterer <peter.hutterer@who-t.net> 131995Date: Wed Mar 18 10:21:59 2009 +1000 131996 131997 dix: remove now obsolete mskidx parameter from DeliverEventsToWindow. 131998 131999 mskidx would always be dev->id anyway, so if we're already passing in the 132000 device, mskidx is superfluous. 132001 132002commit f5409aa026b2cb501170867d04c6e220f3fe0a44 132003Author: Peter Hutterer <peter.hutterer@who-t.net> 132004Date: Fri Mar 13 15:51:50 2009 +1000 132005 132006 mi: remove deprecated miPointerAbsoluteCursor 132007 132008 Functions that don't specify the device are so last year. 132009 132010commit 46145a9312eaf5dcd0e8b6788ed8219e086f790e 132011Author: Peter Hutterer <peter.hutterer@who-t.net> 132012Date: Fri Mar 13 15:12:55 2009 +1000 132013 132014 dix: remove coreMods field from GrabRec. 132015 132016 Nobody uses it anyway, and it's taking up a whole bit! 132017 132018commit 40e2a61e37109cd9c2ab57954ee40dcf4a062e31 132019Author: Peter Hutterer <peter.hutterer@who-t.net> 132020Date: Thu Mar 12 16:53:56 2009 +1000 132021 132022 Add Get/SetDeviceFocus handling. 132023 132024commit d9b7343eac1157490442cc0fb3b5b6d7d6e9a705 132025Author: Peter Hutterer <peter.hutterer@who-t.net> 132026Date: Thu Mar 12 12:56:18 2009 +1000 132027 132028 dix: even if we don't get a XI1 event, continue processing. 132029 132030commit e11dc10f01603b181e224c90d1dd4abe940f2ce7 132031Author: Peter Hutterer <peter.hutterer@who-t.net> 132032Date: Thu Mar 12 10:21:00 2009 +1000 132033 132034 dix: Send HierarchyEvents when devices are added/removed/enabled/disabled. 132035 132036commit 8634e1bcbae9317fc22132f7c9bde1a4a881b9a7 132037Author: Peter Hutterer <peter.hutterer@who-t.net> 132038Date: Thu Mar 12 10:42:05 2009 +1000 132039 132040 dix: send presence events when floating and unfloating SDs 132041 132042commit c9483a53bea1f721d2902ac11467f67317e7de14 132043Author: Peter Hutterer <peter.hutterer@who-t.net> 132044Date: Wed Mar 11 16:20:22 2009 +1000 132045 132046 include: add a few prototypes to silence compiler warnings. 132047 132048commit a4b61cfc78d81de02a71fa6ad8a44a06616f6794 132049Author: Peter Hutterer <peter.hutterer@who-t.net> 132050Date: Wed Mar 11 16:08:24 2009 +1000 132051 132052 dix: un-static FixUpEventFromWindow and deal with focus events too. 132053 132054commit 4cc6a96d7171e567a9bd5a10f552bb953077aafb 132055Author: Peter Hutterer <peter.hutterer@who-t.net> 132056Date: Tue Mar 10 16:08:14 2009 +1000 132057 132058 input: add support for RawDeviceEvents. 132059 132060commit a668d91e28d5a3042a8ce0d087474883b046869a 132061Author: Peter Hutterer <peter.hutterer@who-t.net> 132062Date: Mon Mar 9 16:36:07 2009 +1000 132063 132064 dix: store the xi2mask on grabs and pass it around as needed. 132065 132066 This enables passive and implicit passive grabs for XI2 events, except that we 132067 don't have the protocol spec yet to request them. 132068 132069commit 32f338263ff7de1a2e76d570c98f5be979c18d4e 132070Author: Peter Hutterer <peter.hutterer@who-t.net> 132071Date: Sun Mar 8 21:32:31 2009 +1000 132072 132073 Xi: Deliver XI2 HierarchyEvents when the hierarchy changes. 132074 132075 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132076 132077commit 97e89a59572a4be6757510a317c142ec1d82e8f8 132078Author: Peter Hutterer <peter.hutterer@who-t.net> 132079Date: Fri Mar 6 23:22:00 2009 +1000 132080 132081 Allow XI2 event selection for AllDevices and AllMasterDevices. 132082 132083commit 808a158182b0a3b985385899252c190947b4dbbd 132084Author: Peter Hutterer <peter.hutterer@who-t.net> 132085Date: Fri Mar 6 14:45:26 2009 +1000 132086 132087 dix: Enable XI2 delivery for events and focused events. 132088 132089 No support for grabbed events yet. 132090 132091commit 87ff1159b497c66554a4036ae619f92d5631f00f 132092Author: Peter Hutterer <peter.hutterer@who-t.net> 132093Date: Wed Mar 11 09:01:23 2009 +1000 132094 132095 dix: deliver device enter/leave events. 132096 132097commit 04ed0bcb2530866f7248b412974ecd15f0fbf6ac 132098Author: Peter Hutterer <peter.hutterer@who-t.net> 132099Date: Mon Feb 23 17:07:55 2009 +1000 132100 132101 xkb: remove Device/Enter leave handling - XI2 enter/leave don't have compat state. 132102 132103commit 3f37923a727d5fc4b1880a53dac1145884ba7140 132104Author: Peter Hutterer <peter.hutterer@who-t.net> 132105Date: Mon Feb 23 16:12:25 2009 +1000 132106 132107 Xi: send XI2 focus events. 132108 132109commit daf7dd3bfecc9029bdd10c7a4ea5d2c875a89c41 132110Author: Peter Hutterer <peter.hutterer@who-t.net> 132111Date: Wed Mar 11 09:01:12 2009 +1000 132112 132113 dix: Add GetWindowXI2Mask helper 132114 132115commit 8b6a370058ad5a20e0a0e49ec9443daf03775de8 132116Author: Peter Hutterer <peter.hutterer@who-t.net> 132117Date: Mon Feb 23 15:58:07 2009 +1000 132118 132119 Add XI2 masks and XISelectEvent() request handling. 132120 132121 XI2 event masks are simply stored in the OtherEventMasks as a separate field. 132122 This replaces the XiSelectEvent code. 132123 132124commit 38bba0c1b75b84e8bbdfa7975cf701a9414a3afd 132125Author: Peter Hutterer <peter.hutterer@who-t.net> 132126Date: Fri Feb 20 16:07:33 2009 +1000 132127 132128 Xi: Change ChangeMasterDeviceClasses to new XI2 events. 132129 132130 Split ChangeMasterDeviceClasses into an extra XISendDeviceChangedEvent that 132131 assembles the XI2 wire event for the DeviceChanged event. Re-use this when 132132 detaching the last SD. 132133 132134 Not quite perfect yet, we still copy the device classes from the slave now 132135 rather than from the data we had when the event occured. But it's a start. 132136 132137 (We can now unexport SizeDeviceInfo and CopySwapDevices, not needed anymore) 132138 132139 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132140 132141commit 181e41511d5772962b8bc998a1940d2a9ba62d5e 132142Author: Peter Hutterer <peter.hutterer@who-t.net> 132143Date: Mon Feb 23 16:14:51 2009 +1000 132144 132145 dix: fix up device enter/leave for XI_Enter/XI_Leave. 132146 132147commit 0befeb36c175b14eac8d5416851f1556af90dd86 132148Author: Peter Hutterer <peter.hutterer@who-t.net> 132149Date: Tue Feb 24 15:27:07 2009 +1000 132150 132151 dix: Add device info to DeviceChangedEvent, and fill in CCCE. 132152 132153 We need to fill the info here, as the device may change until we get a chance 132154 to process it. 132155 132156 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132157 132158commit b2ba77bac4fc3edab39372dfdc529e745ae74f77 132159Author: Peter Hutterer <peter.hutterer@who-t.net> 132160Date: Thu Mar 5 14:23:47 2009 +1000 132161 132162 dix: add EventToXI2 and GetXI2Type. 132163 132164 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132165 132166commit dfa2e8e966b699f9d292628119d5a9cbd2dbdf5a 132167Author: Peter Hutterer <peter.hutterer@who-t.net> 132168Date: Mon Feb 23 15:12:32 2009 +1000 132169 132170 Xi: add XIQueryDevice request handling. 132171 132172commit 560c58b53e4e60ebd26e0c66dd00399c438bd619 132173Author: Peter Hutterer <peter.hutterer@who-t.net> 132174Date: Fri Feb 20 15:17:52 2009 +1000 132175 132176 Xi: name-space and fix XI2 requests 132177 132178 All XI2 requests have XI prefixes. 132179 Requests affected: 132180 ChangeDeviceCursor 132181 QueryDevicePointer 132182 WarpDevicePointer 132183 SetClientPointer 132184 GetClientPointer 132185 XIChangeDeviceHierarchy. 132186 132187commit 903449615b1a1188955c0287a552aec8ac5b3fbb 132188Author: Peter Hutterer <peter.hutterer@who-t.net> 132189Date: Fri Feb 20 15:03:35 2009 +1000 132190 132191 Xi: add XIQueryVersion request handling. 132192 132193commit 77d51b94bd264be5a8625a2bdd98f2bfb2aa33d3 132194Author: Peter Hutterer <peter.hutterer@who-t.net> 132195Date: Sun Mar 8 11:47:26 2009 +1000 132196 132197 dix: FixUpEventForWindow needs to handle XI2 events. 132198 132199commit 95ed2ab715fa881d36d508adf4ae5dadd92ae7b2 132200Author: Peter Hutterer <peter.hutterer@who-t.net> 132201Date: Thu Mar 12 11:46:07 2009 +1000 132202 132203 Add XI2_EVENT check. 132204 132205commit 92e0a324b5c738166ed4046cb00fe85ab8aea5a0 132206Author: Peter Hutterer <peter.hutterer@who-t.net> 132207Date: Thu Mar 12 14:15:02 2009 +1000 132208 132209 require inputproto 1.9.99.7 132210 132211commit 906fe9547957917ac463d646935473ceca10a72b 132212Author: Peter Hutterer <peter.hutterer@who-t.net> 132213Date: Fri Mar 20 09:36:09 2009 +1000 132214 132215 Xi: don't allow OpenDevice on any MD. 132216 132217 This restores the original behaviour of XI 1. Any slave device may be opened, 132218 but none of the MDs. 132219 132220 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132221 132222commit c7eb27f3927ba9a36f9904938996f346cca33021 132223Author: Peter Hutterer <peter.hutterer@who-t.net> 132224Date: Wed Mar 4 18:45:09 2009 +1000 132225 132226 Xi: always only list VCP/VCK and the SD's in XListInputDevices. 132227 132228 If a client wants other MDs, it should use XI2's XQueryDevice() instead. 132229 132230 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132231 132232commit 801f14d4b742e506201b619d472087a7e82bd018 132233Author: Peter Hutterer <peter.hutterer@who-t.net> 132234Date: Tue Mar 3 17:04:08 2009 +1000 132235 132236 Xi: don't send attachment info down with ListInputDevices. 132237 132238commit 66dd8879b3527f30309fb747cad4a161d0709d91 132239Author: Peter Hutterer <peter.hutterer@who-t.net> 132240Date: Fri Feb 20 15:01:05 2009 +1000 132241 132242 Xi: purge ExtendedGrabDevice request handling. 132243 132244commit 43e48900fccd245e98d6d697e4cdbb6fa866a0dd 132245Author: Peter Hutterer <peter.hutterer@who-t.net> 132246Date: Fri Mar 6 22:27:54 2009 +1000 132247 132248 dix: Force deviced ids of 2 and higher for actual devices. 132249 132250 0 is now reserved for the "AllDevices" virtual device. 132251 1 is now reserved for the "AllMasterDevices" virtual device. 132252 132253 This also means that wherever we passed in (mskidx = 0), we now need to pass 132254 in the deviceid. 132255 132256commit 70896f6127e2c12c13479a027fdd6e67853d1f8b 132257Author: Peter Hutterer <peter.hutterer@who-t.net> 132258Date: Fri Feb 13 16:42:02 2009 +1000 132259 132260 dix: abstract event filters through GetEventFilters. 132261 132262 Don't let everyone acces the filters[] array directly. This is necessary once 132263 we start dealing with GenericEvents, where the filters are a bit more 132264 complicated. 132265 132266 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132267 132268commit 8a8763f1502c9ddd1961f96f2cc7c73fbc0b49b5 132269Author: Peter Hutterer <peter.hutterer@who-t.net> 132270Date: Thu Feb 12 16:07:15 2009 +1000 132271 132272 dix: restructure DeliverEventsToWindow, a little bit anyway. 132273 132274 Desperate attempt of making the code easy to read and understand. A failed 132275 desparate attempt, possibly. 132276 132277commit 891cfc802028c3e008ee6c2d3e2438a31f57f39f 132278Author: Peter Hutterer <peter.hutterer@who-t.net> 132279Date: Thu Feb 12 10:12:45 2009 +1000 132280 132281 dix: clean up DeliverDeviceEvents. 132282 132283 Move mask checking into external function and streamline the code. 132284 132285 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132286 132287commit 6b6b660c3633af75a94302725137549df0dbe446 132288Author: Peter Hutterer <peter.hutterer@who-t.net> 132289Date: Tue Mar 10 10:15:45 2009 +1000 132290 132291 Xext: remove ev_fill from GEExtensions. 132292 132293commit 445daa62e720fe58fa2d6c3cdc26663671b2bc48 132294Author: Peter Hutterer <peter.hutterer@who-t.net> 132295Date: Thu Jan 15 17:16:22 2009 +1000 132296 132297 Xext: purge XGE event masks. 132298 132299 The masks were originally designed to generically handle event masks for 132300 extensions. Since all that is in-server anyway, it's much better writing 132301 custom event masks for those extensions that need it and not providing a 132302 unified mechanism. 132303 XI2 needs more than the current implementation, which is already too complex 132304 for most other extensions. good riddance. 132305 132306 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132307 132308commit 6fe4c28bf4f3a9f7c341b9403213470e9ded5cb3 132309Author: Peter Hutterer <peter.hutterer@who-t.net> 132310Date: Mon Feb 16 16:34:05 2009 +1000 132311 132312 Add agressive event type checking. 132313 132314 Best to FatalError if a wrong event comes in. At least that forces me to fix 132315 it really quickly. 132316 132317 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132318 132319commit 836864b65794dc0954a01245e418e714cead8125 132320Author: Peter Hutterer <peter.hutterer@who-t.net> 132321Date: Fri Mar 13 12:43:17 2009 +1000 132322 132323 xkb: don't overrun the map index when accessing symbols. 132324 132325 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132326 132327commit 497a12533905b98f388775b6ba49adf21017cc75 132328Author: Peter Hutterer <peter.hutterer@who-t.net> 132329Date: Fri Mar 13 11:52:33 2009 +1000 132330 132331 dix: remove duplicate PickKeyboard() command. 132332 132333 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132334 132335commit 7d4df0ee4414779ad5e519e7bd297c5dfec409a3 132336Author: Peter Hutterer <peter.hutterer@who-t.net> 132337Date: Thu Mar 19 20:24:06 2009 +1000 132338 132339 dix: fix device sync state when calling SyncBoth during AllowEvents. 132340 132341 This did access the wrong device's sync state, potentially freezing or not 132342 thawing the actual device that was supposed to be thawed. 132343 132344 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132345 132346commit f1c7b95d83948160a0d5796ef6c16fc0d1bf5c5d 132347Author: Peter Hutterer <peter.hutterer@who-t.net> 132348Date: Fri Mar 20 13:55:00 2009 +1000 132349 132350 dix: do percentage check before device check in ProcBell 132351 132352 This is just for correctness. The server should return BadValue for anything 132353 not in [-100, 100]. 132354 132355 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132356 132357commit d60391d8ca9918d2089c23c0baef5c91177325f0 132358Author: Peter Hutterer <peter.hutterer@who-t.net> 132359Date: Fri Mar 20 14:01:30 2009 +1000 132360 132361 dix: fix uncredible fail in PostSyntheticMotion. 132362 132363 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132364 132365commit 1a71862d333282e2d251ff0036866cec22bcce85 132366Author: Simon Thum <simon.thum@gmx.de> 132367Date: Sat Feb 28 22:17:47 2009 +0100 132368 132369 dix/xfree86: simplified velocity approximation algorithm 132370 132371 Replace multi-stage filtering with simple linear velocity, 132372 tracked several instances backwards. A heuristic ensures 132373 only approximately linear motion is considered, so velocity 132374 remains valid in any case. Numerical stability is much 132375 better, and nothing changes to people who didn't tune the 132376 advanced features of the previous algorithm. 132377 132378 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132379 132380commit 5ae129baef85b47590c02e4cf61b23904d8f7aa9 132381Author: Peter Hutterer <peter.hutterer@who-t.net> 132382Date: Tue Mar 17 15:47:57 2009 +1000 132383 132384 Xi: check for existence of the button class before accessing it 132385 132386 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132387 132388commit e8094d8f3f69f45850af031efd79a3928e584638 132389Author: Peter Hutterer <peter.hutterer@who-t.net> 132390Date: Thu Mar 19 10:09:04 2009 +1000 132391 132392 dix: ProcSendEvent shouldn't use inputInfo.keyboard directly. 132393 132394 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132395 132396commit 12aeddf5ad41902a180f8108623f356642b3e911 132397Author: Maarten Maathuis <madman2003@gmail.com> 132398Date: Thu Mar 19 19:07:59 2009 +0100 132399 132400 exa: Accept scratch pixmaps with offscreen memory as such. 132401 132402commit a635bb5357648779815ec6729d55f6b478493213 132403Author: Maarten Maathuis <madman2003@gmail.com> 132404Date: Tue Mar 17 19:34:45 2009 +0100 132405 132406 exa: round of fb_pitch to the next byte 132407 132408commit ec328783593339e33772301a8c5fea35fd63c0a2 132409Author: Maarten Maathuis <madman2003@gmail.com> 132410Date: Tue Mar 17 10:45:21 2009 +0100 132411 132412 exa: check if the pixmap is pinned for unsupported AUX indices. 132413 132414commit 3992dd38caf33b343affd8d732c94880d1099dcf 132415Author: Adam Jackson <ajax@redhat.com> 132416Date: Mon Mar 16 13:24:48 2009 -0400 132417 132418 selinux: Add support for avc_acquire_netlink_fd() 132419 132420 Requires libselinux 2.0.79 or newer. Without this, libselinux will 132421 check for policy updates on the netlink socket on basically every policy 132422 lookup. Statistically speaking, they never happen, and the check 132423 translates to at least one more syscall on basically every operation. 132424 132425 Instead, take control of the fd from the library, and check it in 132426 WakeupHandler if it polls readable. 132427 132428commit be6dc9023b5fb3995a9ce56d607627e247918aef 132429Author: Peter Hutterer <peter.hutterer@who-t.net> 132430Date: Fri Mar 13 16:29:36 2009 +1000 132431 132432 xfree86: remove a superfluous assignment. 132433 132434 If dev is NULL, we don't have a VCP and that means we have other issues 132435 anyway. 132436 132437commit bfab422dcdbb9c009b2a91fe0dba288fac9bc859 132438Author: Peter Hutterer <peter.hutterer@who-t.net> 132439Date: Fri Mar 13 16:27:49 2009 +1000 132440 132441 xfixes: useless (void) typecast removal 132442 132443 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132444 132445commit 4eeaee1e5a05ea973cbeb653abf99f2a352edc70 132446Author: Peter Hutterer <peter.hutterer@who-t.net> 132447Date: Fri Mar 13 16:25:12 2009 +1000 132448 132449 xkb: xkbi has a pointer to the device - use this instead of inputInfo.pointer. 132450 132451commit 3f801ba62ac493d4aeb664a5ae638e3fb0c9a262 132452Author: Peter Hutterer <peter.hutterer@who-t.net> 132453Date: Fri Mar 13 16:23:50 2009 +1000 132454 132455 xkb: the VCP can post device events, don't stop xkb filtering on it. 132456 132457commit 23686e5680d565c36e0b6320ce62c38c7830d06b 132458Author: Peter Hutterer <peter.hutterer@who-t.net> 132459Date: Fri Mar 13 16:20:56 2009 +1000 132460 132461 xkb: Fix a mis-use of inputInfo.keyboard. 132462 132463commit c97c6c3de337f247406eaffd5845a847630ee6c1 132464Author: Peter Hutterer <peter.hutterer@who-t.net> 132465Date: Fri Mar 13 16:17:20 2009 +1000 132466 132467 xkb: fix a couple of device checks when looping through all devices. 132468 132469 Generally, we want to apply stuff to the device and to all attached slave 132470 devices. 132471 132472commit d21a546e9604ce2d3c914900f6e6ff2df9f5ca8b 132473Author: Peter Hutterer <peter.hutterer@who-t.net> 132474Date: Fri Mar 13 16:02:22 2009 +1000 132475 132476 Xi: VCP and VCK may register for ext. events, so delete them if necessary. 132477 132478commit 6ec49098f7ef259b43870139511006627b308bb6 132479Author: Peter Hutterer <peter.hutterer@who-t.net> 132480Date: Fri Mar 13 15:57:27 2009 +1000 132481 132482 render: replace a wrong inputInfo.pointer with pDev 132483 132484 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132485 132486commit 0b2b330f6fbf8633ae682a17fbd95f6b5db2c112 132487Author: Peter Hutterer <peter.hutterer@who-t.net> 132488Date: Fri Mar 13 15:54:27 2009 +1000 132489 132490 mi: remove a bunch of useless inputInfo.pointer assignments. 132491 132492 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132493 132494commit 31301861c1dd87a2f55c5c4300ac7c8f54154bf8 132495Author: Peter Hutterer <peter.hutterer@who-t.net> 132496Date: Fri Mar 13 15:49:22 2009 +1000 132497 132498 dix: remove wrong InputInfo.pointer usage - should be pDev instead. 132499 132500commit 639658b2eab480a5bfb943dff51e4bdfc1e16e3b 132501Author: Peter Hutterer <peter.hutterer@who-t.net> 132502Date: Fri Mar 13 15:34:29 2009 +1000 132503 132504 dix: use GetPairedDevice in ProcQueryPointer rather than inputInfo.keyboard. 132505 132506commit f01ee198ff0cbd7ca418217a84248d1c6131a0c6 132507Author: Peter Hutterer <peter.hutterer@who-t.net> 132508Date: Fri Mar 13 15:22:23 2009 +1000 132509 132510 dix: don't use inputInfo.keyboard to get the focus window in ActivateKbdGrab 132511 132512 I'm not sure if that's the right solution, but the other one is wronger. 132513 132514commit de415743f4c1878df2de79f84f5fe087a88d5241 132515Author: Peter Hutterer <peter.hutterer@who-t.net> 132516Date: Fri Mar 13 15:19:23 2009 +1000 132517 132518 dix: fix XACE checks in ProcWarpPointer 132519 132520commit 18430616cfd5d0bc5a3efe61fcc2982b1726b0de 132521Author: Peter Hutterer <peter.hutterer@who-t.net> 132522Date: Fri Mar 13 14:59:18 2009 +1000 132523 132524 dix: use PickPointer() and PickKeyboard instead of inputInfo.pointer/keyboard. 132525 132526 Affected calls: 132527 ProcChangeKeyboardControl 132528 ProcBell 132529 ProcUngrabKey 132530 ProcUngrabButton 132531 132532commit e31727158cda7729283233cfdc9bc5bcb59a52c1 132533Author: Peter Hutterer <peter.hutterer@who-t.net> 132534Date: Fri Mar 13 14:51:55 2009 +1000 132535 132536 dix: fix ProcChangePointerControl's wrong inputInfo.pointer usage. 132537 132538 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132539 132540commit fc5e8dc4a24c2b5bed4e6b5d00670c4401ec92d9 132541Author: Adam Jackson <ajax@redhat.com> 132542Date: Fri Mar 13 14:00:37 2009 -0400 132543 132544 Unexport xf86SetPriority 132545 132546commit 89d7b88f32f9227fa72a043675367b94c4c9283c 132547Author: Maarten Maathuis <madman2003@gmail.com> 132548Date: Fri Mar 13 00:40:43 2009 +0100 132549 132550 exa: allow exaModifyPixmapHeader to set sys_ptr for EXA_HANDLES_PIXMAPS 132551 132552 - exaModifyPixmapHeader would save sys_ptr if needed, but it would be NULL'ed afterwards. 132553 - This is needed to support pixmaps that are not offscreen. 132554 132555commit 544cd9e7b50cd9905afc52404de1a5a2bcff91b5 132556Author: Ander Conselvan de Oliveira <ander@mandriva.com.br> 132557Date: Thu Mar 5 12:57:28 2009 -0500 132558 132559 Honour Option "DPMS" "off" on xorg.conf 132560 132561 Signed-off-by: James Cloos <cloos@jhcloos.com> 132562 132563commit 7abb72c5c711a26a27a82604bd2398bcb6df0a99 132564Author: Simon Thum <simon.thum@gmx.de> 132565Date: Wed Feb 25 10:55:36 2009 +0100 132566 132567 xfree86: remove obsolete code 132568 132569 remove a few lines which redo part of the pointer acceleration 132570 init. Properties is the way to go for them. 132571 132572 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132573 132574commit 97cb3cc28e4a5694d17f156276efd84555a96d18 132575Author: Peter Hutterer <peter.hutterer@who-t.net> 132576Date: Thu Mar 12 10:17:16 2009 +1000 132577 132578 dix: Merge DevicePresence notify events generation into a single function. 132579 132580 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132581 132582commit 7f1ba804a1c1aef725e20a04e5f6ef74478ecf44 132583Author: Peter Hutterer <peter.hutterer@who-t.net> 132584Date: Tue Mar 10 11:55:04 2009 +1000 132585 132586 dix: s/numEvents/num_events/ in GetKeyboardValuatorEvents 132587 132588 GPE and GProxE use the same. 132589 132590 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132591 132592commit 067f26f7ba4c8cf37af7ddf0eb204b7e1d06096d 132593Author: Peter Hutterer <peter.hutterer@who-t.net> 132594Date: Fri Mar 6 19:42:27 2009 +1000 132595 132596 Xi: set the modifier + group state during POE. 132597 132598 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132599 132600commit da724e8f1d1867c36c9550a153c58829939800fc 132601Author: Peter Hutterer <peter.hutterer@who-t.net> 132602Date: Fri Mar 6 15:50:05 2009 +1000 132603 132604 Xi: Set the button state on the event during POE. 132605 132606 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132607 132608commit b368dd1720c708aaec492b4a67fb7e4f87569c11 132609Author: Peter Hutterer <peter.hutterer@who-t.net> 132610Date: Thu Mar 5 15:22:41 2009 +1000 132611 132612 dix: remove a few pointless (void) casts of return values. 132613 132614 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132615 132616commit cadf65a6e190a8952ad3cc216dc9ea55241de91a 132617Author: Matthias Hopf <mhopf@suse.de> 132618Date: Tue Mar 10 15:49:49 2009 +0100 132619 132620 randr: Nuke broken set_origin shortcut 132621 132622 Shortcut is impossible to implement this way, because we don't know for sure 132623 whether the crtc of an output has changed or not. 132624 132625commit 6544490700051b3b5e88ac1890d71b35634c9100 132626Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 132627Date: Mon Mar 9 17:28:40 2009 -0400 132628 132629 Correct access mode in call to dixLookupWindow() within RRSelectInput. 132630 132631 Reported by Alan Coopersmith. 132632 132633commit f8dd80d13bb5313a11b38b280f8ad3e22f0a6300 132634Author: Keith Packard <keithp@keithp.com> 132635Date: Wed Aug 6 15:26:24 2008 -0700 132636 132637 Replace dixLookupResource by dixLookupResourceBy{Type,Class} 132638 132639 dixLookupResource attempted to automatically detect whether the caller 132640 wanted a lookup by-type or by-class, unfortunately, it guessed wrong for 132641 RT_NONE. Instead of trying to make the guess better, this patch just reverts 132642 the unification and creates separate functions for each operation. 132643 132644commit 0d9a42dc0380d1583889b6b6521bd5a2451735d4 132645Author: Peter Hutterer <peter.hutterer@who-t.net> 132646Date: Mon Mar 9 08:59:01 2009 +1000 132647 132648 Xi: don't crash on a NULL property name, just return None. 132649 132650 strcmp doesn't like NULL-pointers. 132651 132652 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132653 132654commit 432a95032eb10f82fb59caa29e52d3f56dfb363c 132655Author: Peter Hutterer <peter.hutterer@who-t.net> 132656Date: Sun Mar 8 21:34:02 2009 +1000 132657 132658 Remove useless (void) typecasts. 132659 132660 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132661 132662commit f508446a2bcd887df9cf5db786b15fc3cbc435a0 132663Author: Peter Åstrand <astrand@cendio.se> 132664Date: Fri Mar 6 19:41:35 2009 -0800 132665 132666 Composite: Uninitialized borderClipX/borderClipY 132667 132668 Valgrind warned me about two other uninitialized variables, which are not 132669 padding. Valgrind output at the end. I'm a bit unsure of what these should 132670 be initialized to, is zero fine? 132671 132672commit d1338a94805cc774fe0a5d00c2225a3ee9673a9f 132673Author: Winfried Grünewald <winfried.gruenewald.ext@eizo-dt.de> 132674Date: Fri Mar 6 12:59:35 2009 -0500 132675 132676 [hw/xfree86] Fix StaticGray cmap. 132677 132678 Fix this bug report: 132679 132680 ,----< from http://bugzilla.freedesktop.org/show_bug.cgi?id=20504 > 132681 | Using the Visual StaticGray (8 bit depth) is missing one gray level. 132682 | The gray level of index zero and index one are the same and all 132683 | other levels are shifted by one. The max level (255) cannot be used. 132684 `---- 132685 132686 Signed-off-by: James Cloos <cloos@jhcloos.com> 132687 132688commit eba3bab71ff26d58bb0e49e4a9726fc21ff07258 132689Author: Alan Coopersmith <alan.coopersmith@sun.com> 132690Date: Thu Mar 5 15:09:28 2009 -0800 132691 132692 Remove #ifdef macII code left over from ancient A/UX 3.0 support 132693 132694 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 132695 132696commit daa9727a02701cfac7b92f3e61d752be6fcc005e 132697Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 132698Date: Wed Mar 4 11:33:00 2009 -0800 132699 132700 XQuartz: Implement SendPSN for AppleWM 132701 (cherry picked from commit 779e9fdc2de9bfebc5f0a4c9d5f42e389fdd8593) 132702 132703commit ce6e1771be5f2c21af6f72a9705795df26210413 132704Author: Maarten Maathuis <madman2003@gmail.com> 132705Date: Mon Mar 2 17:05:28 2009 +0100 132706 132707 exa: fix a serious issue in exaChangeWindowAttributes (and some more related things) 132708 132709 - fbChangeWindowAttributes can create pixmaps (and access them) without use preparing access. 132710 - Also handle the destroyed pixmaps by finishing them first. 132711 - Switch to DEST indices again in exaCreatePixmapWithPrepare, because they are obviously being rendered to. 132712 - Also avoid calling FinishAccess on pixmaps that are destroyed (and their memory potentially invalid). 132713 132714commit 7fb68e8b31d259325ce311ad67befc43a933f009 132715Author: Maarten Maathuis <madman2003@gmail.com> 132716Date: Sat Feb 28 22:29:42 2009 +0100 132717 132718 exa: remove a few pExaPixmap checks. 132719 132720 - And make some fatal for a debug build. 132721 132722commit bd2f35ef0d4b8c56eb205b8694b4fe5d2fa2c049 132723Author: Maarten Maathuis <madman2003@gmail.com> 132724Date: Sat Feb 28 22:15:04 2009 +0100 132725 132726 exa: fix unwrapping of ModifyPixmapHeader upon CloseScreen. 132727 132728 - Cleanup wrapping too. 132729 132730commit da8ea41a542787691ea1120e5c8c7dc3182cbea5 132731Author: Maarten Maathuis <madman2003@gmail.com> 132732Date: Sat Feb 28 21:59:09 2009 +0100 132733 132734 exa: increase/rework safety checks in Prepare/FinishAccess. 132735 132736commit 3ea3d505e8128ab3e878edcef697fd5656b0c917 132737Author: Maarten Maathuis <madman2003@gmail.com> 132738Date: Sun Mar 1 01:15:28 2009 +0100 132739 132740 exa: avoid a potential Prepare/FinishAccess inbalance. 132741 132742commit 10334cf7e64e2e633cd507f2aa216027e840c8e5 132743Author: Maarten Maathuis <madman2003@gmail.com> 132744Date: Sat Feb 28 21:46:00 2009 +0100 132745 132746 exa: simplify exaPixmapIsOffscreen 132747 132748 - This includes properly handling the framebuffer. 132749 132750commit 57bed76e218471b60d340b18c346d492f37798c2 132751Author: Tomas Carnecky <tom@dbservice.com> 132752Date: Wed Mar 4 10:29:34 2009 -0500 132753 132754 Remove dead EXTENSION_PROC_ARGS define 132755 132756 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 132757 132758commit a6fb9f4ccfc69dc090b152e29142e3159f812166 132759Author: Tomas Carnecky <tom@dbservice.com> 132760Date: Wed Mar 4 10:27:30 2009 -0500 132761 132762 Remove dead XTest code from extmod 132763 132764 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 132765 132766commit c4081fee0995dae691665a00ea9bfdb8b8043b8c 132767Author: Topi Kanerva <topi.kanerva@nokia.com> 132768Date: Wed Mar 4 10:23:10 2009 -0500 132769 132770 Make DGA optional. 132771 132772commit c7ebb4bef152b47880abb6ea4819416e89e6522c 132773Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 132774Date: Tue Mar 3 14:02:36 2009 -0500 132775 132776 Fix 2 const warnings. 132777 132778commit c7cf926d25523d5d0d8e5c83671c901660070d71 132779Author: Eric Paris <eparis@redhat.com> 132780Date: Tue Mar 3 13:15:39 2009 -0500 132781 132782 This patch changes all places in the X code to use _raw functions. The 132783 X server should never see, translate, or deal with a munged context. 132784 Display managers which show contexts to the user should take care of 132785 translating these to human readable form. 132786 132787commit d84a2f855da9fd293e984084168c1e27f5a5115c 132788Author: Peter Hutterer <peter.hutterer@who-t.net> 132789Date: Sun Mar 1 16:59:14 2009 +1000 132790 132791 xfree86: Add linebreak to two debug statements. 132792 132793 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132794 132795commit e0ed9f16d0ab871f3209837f54055446503c4108 132796Author: Jason Vas Dias <jason.vas.dias@gmail.com> 132797Date: Tue Mar 3 13:24:54 2009 +1000 132798 132799 Fix build with --enable-debug. 132800 132801 [amended by Peter Hutterer] 132802 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132803 132804commit 6093d3eb1d7d1e9cbacc2c2f4d376352f787dd53 132805Author: Peter Hutterer <peter.hutterer@who-t.net> 132806Date: Thu Feb 26 15:38:45 2009 +1000 132807 132808 dix: fix two compiler warnings (old-style function definition). 132809 132810 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132811 132812commit b874a5c0aafc766891c9e567af48b891ea14be4f 132813Author: Peter Hutterer <peter.hutterer@who-t.net> 132814Date: Thu Feb 26 14:40:22 2009 +1000 132815 132816 mi: only print the "EQ overflowing" error once. 132817 132818 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132819 132820commit a706dd8771f87cf0a73578755552124d162baa5d 132821Author: Peter Hutterer <peter.hutterer@who-t.net> 132822Date: Thu Feb 26 10:04:14 2009 +1000 132823 132824 include: add a range of button labels. 132825 132826 Mostly the same buttons as defined by linux/input.h, with five exceptions: 132827 "Button Unknown" for a button that cannot be labelled. 132828 "Button Wheel Up", "Button Wheel Down" for buttons 4/5, traditionally the 132829 wheel buttons. 132830 "Button Horiz Wheel Up", "Button Horiz Wheel Down" for buttons 6/7, 132831 traditionally the horiz. wheel buttons. 132832 132833 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 132834 Acked-by: Daniel Stone <daniel@fooishbar.org> 132835 132836commit 8a6ed44a8b2fc5f14729dc54fec17607ced03859 132837Author: Adam Jackson <ajax@redhat.com> 132838Date: Mon Mar 2 14:26:23 2009 -0500 132839 132840 randr: Fix thinko in xf86TargetPreferred 132841 132842 The only-one-output case would only work right if that also happened to 132843 be the zeroth output. Oops. 132844 132845commit 7bfad37f7706da5a76174c0da78dce4d6c6a585b 132846Author: Maarten Maathuis <madman2003@gmail.com> 132847Date: Fri Feb 27 22:26:52 2009 +0100 132848 132849 exa: fixup aux indices and ensure that the indices are used as they should be. 132850 132851 - In a previous patch i forgot to add a FALSE somewhere it seems. 132852 - Rename AUX indices so the driver (think of driver managed pixmaps) can do optimisations based upon them. 132853 - Fix one abuse of DEST index now that we have the AUX indices (same reason as above). 132854 132855commit 8c6b4a827624aa70aea94b449fee6d5df8504105 132856Author: Adam Jackson <ajax@redhat.com> 132857Date: Fri Feb 27 12:42:43 2009 -0500 132858 132859 DDC: Redo extended device probe slightly. 132860 132861 We'll now only mention the E-EDID segment register if the device is 132862 actually E-EDID-capable. While we're here, check for DDC/CI and 132863 standard EEPROM support too. 132864 132865commit b030f858f2f1ce1fd27a73ebf7f9ec5db541a668 132866Author: Adam Jackson <ajax@redhat.com> 132867Date: Sat Feb 21 19:56:20 2009 -0500 132868 132869 selinux: Don't bother relabeling resources that are being destroyed 132870 132871 Makes window destroy about 40x faster in Xvfb. 132872 132873commit f028b14876dc536b575d4b6e1df7f37ee525acec 132874Author: Maarten Maathuis <madman2003@gmail.com> 132875Date: Fri Feb 27 13:06:28 2009 +0100 132876 132877 exa: whitespace 132878 132879commit ed00515ec54b294f304b8ae2857e3073f97860d2 132880Author: Maarten Maathuis <madman2003@gmail.com> 132881Date: Thu Feb 26 22:18:03 2009 +0100 132882 132883 exa: fixup exaAssertNotDirty. 132884 132885 - Do the right thing based on prepare access. 132886 132887commit 2e88b6004f09dbcb888abf05d4d5554231b2bd1a 132888Author: Maarten Maathuis <madman2003@gmail.com> 132889Date: Thu Feb 26 21:14:23 2009 +0100 132890 132891 exa: minor glyphs cleanup. 132892 132893 - This should fix subtle coordinate bugs and make the code a bit cleaner to read. 132894 132895commit 265d20068af5434489752b6dba0bf0065b3cc3ec 132896Author: Michel Dänzer <daenzer@vmware.com> 132897Date: Fri Feb 27 16:41:39 2009 +0100 132898 132899 EXA: Fix check for whether the glyph we're evicting from the cache is in use. 132900 132901 Since commit f07f18231a921d3ae9dd9b75881c9e58e9e2e235 ('EXA: Allow using 132902 exaCompositeRects also when we can't use a mask in exaGlyphs.') we were 132903 checking the wrong set of coordinates in the buffer where glyphs to be rendered 132904 are accumulated when no mask is used in exaGlyphs. 132905 132906 This fixes occasional glyph corruption which can be corrected with redraws, in 132907 particular with Qt4. 132908 132909 Thanks to Maarten Maathuis for asking the right question: 'where do we protect 132910 against evicting glyphs that are still needed?' 132911 132912 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 132913 132914commit 9d5141f7bcfed36417873b47f850332506af83ac 132915Author: Michel Dänzer <daenzer@vmware.com> 132916Date: Fri Feb 27 16:37:28 2009 +0100 132917 132918 EXA: No longer use the driver UploadToScratch hook. 132919 132920 See http://bugs.freedesktop.org/show_bug.cgi?id=20213 . 132921 132922 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 132923 132924commit 4cfb36f6ad2df01215028fec48d99239a0e4496b 132925Author: Michel Dänzer <daenzer@vmware.com> 132926Date: Fri Feb 27 16:37:28 2009 +0100 132927 132928 EXA: Handle separate alpha maps properly in Composite fallback, take two. 132929 132930 Preserve the EXA ABI by introducing a new driver flag EXA_SUPPORTS_PREPARE_AUX. 132931 If the driver doesn't set this flag, we have to assume any Prepare/FinishAccess 132932 driver hooks can't handle the EXA_PREPARE_AUX* indices, so we move out such 132933 pixmaps at PrepareAccess time. 132934 132935 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=18710 . 132936 132937 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 132938 132939commit 4bf707f01822abe99286909fd561da7e7a4211d6 132940Author: Michel Dänzer <daenzer@vmware.com> 132941Date: Fri Feb 27 16:37:27 2009 +0100 132942 132943 Revert "Fix up Xephyr build for recent EXA changes." 132944 132945 This reverts commit 4e8d98b61e1f763c187e7994c683b543cca1a33c. 132946 132947commit de022f8e634baf5a7acd186934b370647df58355 132948Author: Michel Dänzer <daenzer@vmware.com> 132949Date: Fri Feb 27 16:37:27 2009 +0100 132950 132951 Revert "EXA: Handle separate alpha maps properly in Composite fallback." 132952 132953 This reverts commit 170cf1270dff38d3cce7f5ba5b940d1c0d70eff5. 132954 132955 Conflicts: 132956 132957 exa/exa_render.c 132958 132959commit d0dd649035fc3698c5b436f9d9d248116aa106a3 132960Author: Jon TURNEY <jon.turney@dronecode.org.uk> 132961Date: Thu Feb 12 16:03:16 2009 +0000 132962 132963 Remove references to rgb.txt from files section of Xserver and Xorg man pages 132964 132965 The references to this file in the server code were removed in commit dda10c9066a660b647384179f82e1da8e063264f 132966 132967 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 132968 132969commit b0ad9e1ced9619f37acf77764c395c57b86cf463 132970Author: Jon TURNEY <jon.turney@dronecode.org.uk> 132971Date: Thu Feb 12 16:02:17 2009 +0000 132972 132973 Remove long-gone '-co' option from Xserver man page 132974 132975 The '-co' option was removed in commit dda10c9066a660b647384179f82e1da8e063264f 132976 132977 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 132978 132979commit 7907b37d96dc81f333351a6fedd6a9f4e31ae66a 132980Author: Jeremy Huddleston <jeremy@yuffie.local> 132981Date: Thu Feb 26 21:22:48 2009 -0800 132982 132983 XQuartz: Don't erase the mode_switch+keysym if it's the same as shift+keysym 132984 (cherry picked from commit 3e2427e6e9675dd655c8ad328b82e12191b064f2) 132985 132986commit f07f18231a921d3ae9dd9b75881c9e58e9e2e235 132987Author: Michel Dänzer <daenzer@vmware.com> 132988Date: Thu Feb 26 10:35:44 2009 +0100 132989 132990 EXA: Allow using exaCompositeRects also when we can't use a mask in exaGlyphs. 132991 132992 This should give the full benefits of the glyph cache even when we can't use a 132993 mask. 132994 132995 This also means we no longer need to scan the glyphs to see if they overlap, 132996 we can just use a mask or not as the client asks. 132997 132998 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 132999 133000commit b26c136ee9bf7015c583136af53d0c9e9da67ace 133001Author: Michel Dänzer <daenzer@vmware.com> 133002Date: Thu Feb 26 10:35:44 2009 +0100 133003 133004 EXA: Stop tracking damage for pixmaps subject to ModifyPixmapHeader. 133005 133006 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 133007 133008commit 812e8d681a49b33d8748983843799ef692b8b148 133009Author: Keith Packard <keithp@keithp.com> 133010Date: Wed Feb 25 11:08:01 2009 -0800 133011 133012 Oops. 62fc98c had crtc offsets applied backwards (fix from server-1.6-branch) 133013 133014 Crtc offsets are in screen space, not crtc space, and hence should be 133015 applied relative to the screen->crtc transform, not the crtc->transform. 133016 133017 This fix was found in the 'cherry pick' of 62fc98c on server-1.6-branch, 133018 clearly some new definition of 'cherry pick' that I am unaware of. 133019 133020 Signed-off-by: Keith Packard <keithp@keithp.com> 133021 133022commit c090f5514d28e1602a6ebbe7c909e98a0e3374d7 133023Author: Keith Packard <keithp@keithp.com> 133024Date: Wed Feb 25 10:14:36 2009 -0800 133025 133026 Pre-clip panning coordinates to keep crtc within panning region 133027 133028 There is a separate panning region check, but that doesn't work under 133029 transformation, so just pre-clip the mouse coordinates when computing the 133030 panning offsets. This leaves the case where panning constants are changing 133031 unresolved. 133032 133033 Signed-off-by: Keith Packard <keithp@keithp.com> 133034 133035commit 77c7a64e8885696665556c9fbcb3cffb552e367a 133036Author: Keith Packard <keithp@keithp.com> 133037Date: Wed Feb 25 10:12:23 2009 -0800 133038 133039 RandR rotations and reflections offset by one pixel 133040 133041 The matrix computation for rotation and reflection resulted in dropping a 133042 row or column of pixels as the offsets used in the matrix computations used 133043 width and height rather than width-1 and height-1. 133044 133045 Signed-off-by: Keith Packard <keithp@keithp.com> 133046 133047commit 4e8d98b61e1f763c187e7994c683b543cca1a33c 133048Author: Michel Dänzer <daenzer@vmware.com> 133049Date: Wed Feb 25 10:26:04 2009 +0100 133050 133051 Fix up Xephyr build for recent EXA changes. 133052 133053 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 133054 133055commit d6976e1436c035b8e4a6d1e10971b083b1b4184b 133056Author: Jeremy Huddleston <jeremy@yuffie.local> 133057Date: Tue Feb 24 23:12:57 2009 -0800 133058 133059 XQuartz: Re-enable support for capslock 133060 (cherry picked from commit 4901b8147e593d26d7a31a9b73a201254b948916) 133061 133062commit 4b63f88b0ad5f6614fee5b7af4f97ce4581ab935 133063Author: Jeremy Huddleston <jeremy@yuffie.local> 133064Date: Tue Feb 24 18:54:16 2009 -0800 133065 133066 XQuartz: Comment out the background pointer interaction that seems to be causing CPU spinning on some configurations 133067 (cherry picked from commit 1d479fa4f3bca9cbd81808a6c87582422fd2a257) 133068 133069commit 5672e2d66444be74125c71b81373b82f8e1b7dd2 133070Author: Peter Hutterer <peter.hutterer@who-t.net> 133071Date: Fri Feb 20 14:55:51 2009 +1000 133072 133073 Xi: remove AllExtensionVersions. 133074 133075 Keeping an AllExtensionVersions array to save all versions of 133076 all extension is rather pointless if only one extension uses it. 133077 Rename to XIVersion, reduce to a single struct. 133078 133079 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133080 133081commit 59bc615ed2e60c595887dcc11fe04935b49bcaff 133082Author: Peter Hutterer <peter.hutterer@who-t.net> 133083Date: Tue Feb 24 15:26:20 2009 +1000 133084 133085 dix: set the valuator mode in set_valuators 133086 133087 We only have per-device mode for now anyway. 133088 133089 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133090 133091commit cffa3c499a0495e132557bd6b08a8c0b858cdf3b 133092Author: Peter Hutterer <peter.hutterer@who-t.net> 133093Date: Tue Feb 24 15:25:49 2009 +1000 133094 133095 dix: remove unused variable 'tmp' 133096 133097 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133098 133099commit 8d6dd83a38020080808705a156d32ec2a135633c 133100Author: Peter Hutterer <peter.hutterer@who-t.net> 133101Date: Fri Feb 20 13:13:58 2009 +1000 133102 133103 include: fix indentation for lastSlave/master. 133104 133105 Yeah! 133106 133107 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133108 133109commit b3412adca8b2ba3fff348bceb4f01b9e968c0afc 133110Author: Peter Hutterer <peter.hutterer@who-t.net> 133111Date: Sun Feb 22 20:53:46 2009 +1000 133112 133113 Xi: don't need to set the XKB settings for new core devices. 133114 133115 This is done by the XKB code these days anyway, so we might as well ignore it 133116 and keep using the stanard stuff. 133117 133118 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133119 133120commit a180c2367974abe2ccdc88bb5c6f138c7df4d2e9 133121Author: Peter Hutterer <peter.hutterer@who-t.net> 133122Date: Mon Feb 23 16:38:41 2009 +1000 133123 133124 dix: remove ConvertBackToXI. 133125 133126 This was a temporary solution for the internal-events rework and is not needed 133127 anymore. 133128 133129 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133130 133131commit 26cf709e712573548990d165e71137f28271a8bb 133132Author: Peter Hutterer <peter.hutterer@who-t.net> 133133Date: Mon Feb 23 22:39:08 2009 +1000 133134 133135 Doxygenify events.h and eventconvert.c 133136 133137 Should have done that before pushing, but oh well. 133138 133139 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133140 133141commit 170cf1270dff38d3cce7f5ba5b940d1c0d70eff5 133142Author: Michel Dänzer <daenzer@vmware.com> 133143Date: Tue Feb 24 09:22:09 2009 +0100 133144 133145 EXA: Handle separate alpha maps properly in Composite fallback. 133146 133147 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=18710 . 133148 133149 As this can't work without new EXA_PREPARE_AUX* indices, this requires a major 133150 version bump, so we can also drop the UploadToScratch driver hook and 133151 ExaOffscreenSwap*(). So this also fixes 133152 http://bugs.freedesktop.org/show_bug.cgi?id=20213 . 133153 133154 Moreover, introduce EXA_DRIVER_KNOWN_MAJOR to break compilation of drivers 133155 which may not be able to handle EXA_PREPARE_AUX*, giving instructions how to 133156 make them build again in the #error message. 133157 133158 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 133159 133160commit d3b355875ac46104a174966e1974e6af99e40fd6 133161Author: Peter Hutterer <peter.hutterer@who-t.net> 133162Date: Tue Feb 24 11:14:14 2009 +1000 133163 133164 dix: fix wrong condition for setting valuators on the event. 133165 133166 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133167 133168commit ef320bdd5ec3419abba77041d3a4d96a3ff87563 133169Author: Michel Dänzer <daenzer@vmware.com> 133170Date: Mon Feb 23 16:49:51 2009 +0100 133171 133172 DRI1: Make DRICreateDrawable return TRUE for pixmaps. 133173 133174 GLX_EXT_texture_from_pixmap was broken since commit 133175 a26c77ff432d2e85a2665fc36fca25143460c476 ('glx: fix retval checks when failures 133176 occur for drawable creation.') 133177 133178 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 133179 133180commit 1df6716281579e2937743d840ab1079343c503ac 133181Author: Adam Tkac <vonsch@gmail.com> 133182Date: Mon Feb 23 09:28:26 2009 +1000 133183 133184 xfree86: don't try to UnInit virtual devices in DIDR. (#20087) 133185 133186 DeleteInputDeviceRequest function doesn't handle "virtual" devices well. 133187 TightVNC libvnc.so module to X (which makes bare Xorg VNC capable) uses such 133188 kind of devices. 133189 133190 Bare Xvnc (it is something like Xvfb) simply uses AddInputDevice & 133191 RegisterDevice functions. Xvnc uses DeleteInputDeviceRequest from Xi/stubs.c 133192 so everything works fine (now I see that DeleteInputDeviceRequest in 133193 Xi/stubs.c should call RemoveDevice function, shouldn't it? :) ) 133194 133195 Situation is quite different when you use libvnc.so module. It uses same 133196 schema as Xvnc, so it simply calls AddInputDevice & RegisterDevice. Thus 133197 device is created correctly. When server is terminated it calls 133198 DeleteInputDeviceRequest (now from hw/xfree86/common/xf86Xinput.c) for each 133199 device. Here is the difference - Xvnc calls DeleteInputDeviceRequest from 133200 Xi/stubs.c as I wrote above. Thus Xorg gets sigsegv because "VNC" devices 133201 don't have real input driver. 133202 133203 X.Org Bug 20087 <http://bugs.freedesktop.org/show_bug.cgi?id=20087> 133204 133205 [This isn't really a fix (libVNC should behave correctly) but not crashing the 133206 server sounds like an improvement.] 133207 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133208 133209commit a50c40be1f28467bcef7dc71fab54ebdddacffdf 133210Author: Peter Hutterer <peter.hutterer@who-t.net> 133211Date: Sat Feb 21 08:09:40 2009 +1000 133212 133213 dix: don't apply button mappings to SDs, or the device's MD. (#20122) 133214 133215 Only ever change the button map on the device we actually care about, not the 133216 attached SDs, not the current MD of the device. 133217 133218 X.Org Bug 20122 <http://bugs.freedesktop.org/show_bug.cgi?id=20122> 133219 133220 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133221 Acked-by: Daniel Stone <daniel@fooishbar.org> 133222 133223commit 36583a49965c0bb40a84284939b1539b3cb9fc9c 133224Author: Peter Hutterer <peter.hutterer@who-t.net> 133225Date: Tue Feb 10 12:45:49 2009 +1000 133226 133227 mi: split EQ popping and event processing into two functions. 133228 133229 mieqProcessInputEvents() - pop an event off the EQ and pass it to 133230 mieqProcessDeviceEvent() - process the event according to the MD/SD hierarchy. 133231 133232 This way, we can use mieqPDE() from Xtest, xkb, and others to post an event. 133233 133234 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133235 133236commit 47f136ed6fd80310f715a2555501d1b271dd084c 133237Author: Peter Hutterer <peter.hutterer@who-t.net> 133238Date: Fri Feb 6 12:08:43 2009 +1000 133239 133240 mi: change custom handlers to internal events 133241 133242 This should re-enable DGA, but XQuartz needs to be changed to internal events 133243 too now. 133244 133245 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133246 133247commit bdc262701a37a0c12ead810d63fa99a26cbb82ec 133248Author: Peter Hutterer <peter.hutterer@who-t.net> 133249Date: Mon Feb 2 22:39:37 2009 +1000 133250 133251 dix: remove un-used getValuatorEvents and countValuatorEvents from getevents.c 133252 133253 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133254 133255commit 763848d3ab84b4e33a0254120c44b5a1877c819d 133256Author: Peter Hutterer <peter.hutterer@who-t.net> 133257Date: Mon Feb 2 17:20:17 2009 +1000 133258 133259 Input: change processing API to InternalEvents. 133260 133261 Don't pass xEvent* and count through to processing, pass a single 133262 InternalEvent. 133263 133264 Custom handlers are disabled for the time being. And for extra fun, 133265 XKB's pointer motion emulation is disabled. But stick an error in there so 133266 that we get reminded should we forget about it. 133267 133268 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133269 133270commit 0b4066c116e07918a13dc6b4159df7ac9eb92b4b 133271Author: Peter Hutterer <peter.hutterer@who-t.net> 133272Date: Mon Feb 2 17:18:16 2009 +1000 133273 133274 xkb: _XkbFilterRedirectKey needs to pass InternalEvents down. 133275 133276 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133277 133278commit 085d50360863ccc8280cd3eccea2bcb4f3dd9a14 133279Author: Peter Hutterer <peter.hutterer@who-t.net> 133280Date: Mon Feb 2 16:56:52 2009 +1000 133281 133282 dix: Fix PostSyntheticMotion to use a DeviceEvent for posting. 133283 133284 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133285 133286commit 155986a93dc78d6aa060ca3038d5fafa3d8753c2 133287Author: Peter Hutterer <peter.hutterer@who-t.net> 133288Date: Mon Feb 2 16:06:15 2009 +1000 133289 133290 dix: CheckPassiveGrabsOnWindow moved to internal events. 133291 133292 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133293 133294commit 1c38abd4b136301008ba77d4e68aea34508f4c22 133295Author: Peter Hutterer <peter.hutterer@who-t.net> 133296Date: Mon Feb 2 14:54:34 2009 +1000 133297 133298 dix: switch DeliverGrabbedEvent to internal events. 133299 133300 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133301 133302commit fb858774b80eb75c2f8e81fe893bbbdd37065fbd 133303Author: Peter Hutterer <peter.hutterer@who-t.net> 133304Date: Mon Feb 2 14:44:13 2009 +1000 133305 133306 dix: switch DeliverDeviceEvents to internal events. 133307 133308 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133309 133310commit 3aa9404fcd161e94a80b057a77ef47afe428a56d 133311Author: Peter Hutterer <peter.hutterer@who-t.net> 133312Date: Mon Feb 2 14:00:41 2009 +1000 133313 133314 dix: switch DeliverFocusedEvent to internal events. 133315 133316 And because of xfree() macro hilarity, rename "pointer" to "ptr". Oh, how we 133317 laughed. 133318 133319 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133320 133321commit 8c873e7f514844d1056f2b20e653f1dd75f4c327 133322Author: Peter Hutterer <peter.hutterer@who-t.net> 133323Date: Sun Feb 1 20:41:16 2009 +1000 133324 133325 dix: Switch DeliverGrabbedEvents to use internal events. 133326 133327 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133328 133329commit a3718536989fa7d3358e0b2d859c25fde0a2d93e 133330Author: Peter Hutterer <peter.hutterer@who-t.net> 133331Date: Sun Feb 1 20:18:33 2009 +1000 133332 133333 dix: Deliver{Grabbed|Focused|Device}Events API changed to InternalEvents. 133334 133335 With the API change, we can now purge the XI conversion from POE. 133336 133337 Note: this commit breaks DGA even more. 133338 133339 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133340 133341commit 8f94ec6f788565474931cc7d5e3d4672f0f31670 133342Author: Peter Hutterer <peter.hutterer@who-t.net> 133343Date: Sun Feb 1 09:58:15 2009 +1000 133344 133345 dix: convert passive grabs to use internal events. 133346 133347 deviceGrab.sync.event is now an internal event, and CheckDeviceGrabs and 133348 friends is changed over. 133349 133350 Note that this currently breaks some frozen grabs. See towards the end of 133351 ComputeFreezes(). 133352 133353 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133354 133355commit daa3245c479b19d445a070b5b76ee005915b5335 133356Author: Peter Hutterer <peter.hutterer@who-t.net> 133357Date: Sat Jan 31 08:46:43 2009 +1000 133358 133359 dix: fix EnqueueEvent to work with internal events. 133360 133361 Note that we're only partially switched to internal events. The event in the 133362 devices' event queue (dev->deviceGrab.sync.event) is still an XI event. The 133363 events in syncEvents are InternalEvents only now. 133364 This also implies fixing CheckVirtualMotion to work with internal events. 133365 133366 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133367 133368commit 75595ba4aa9c3823ffe3b3388ce088929824816f 133369Author: Peter Hutterer <peter.hutterer@who-t.net> 133370Date: Fri Jan 30 15:14:46 2009 +1000 133371 133372 Xi: make ProcessOtherEvents more InternalEvent aware. 133373 133374 Get rid of the deviceValuator processing and a few other things, but still 133375 drop back into XI before checking device grabs or doing anything else. 133376 133377 NoticeEventTime now needs to take InternalEvents, and while we're at it, 133378 change NoticeTime from a macro to a function. 133379 133380 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133381 133382commit 3a02e538dbdb3cd482e01baeaf2aba2ddb7731df 133383Author: Peter Hutterer <peter.hutterer@who-t.net> 133384Date: Fri Jan 30 14:59:22 2009 +1000 133385 133386 dix: update CheckMotion to deal with DeviceEvents. 133387 133388 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133389 133390commit 8829d966a6bacb05d322b60531c59366b58f4514 133391Author: Peter Hutterer <peter.hutterer@who-t.net> 133392Date: Fri Jan 30 13:33:55 2009 +1000 133393 133394 Xi: support InternalEvents in UpdateDeviceState, parts of POE and EnqueueEvent 133395 133396 Note that this breaks DGA. Life is tough. 133397 133398 EnqueueEvent is a somewhat half-baked solution, we immediately drop back into 133399 XI and store them. But it should in theory work. 133400 133401 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133402 133403 Don't let the dcce be random data. 133404 133405commit 007e93c869325cafcc29d975b356dbb8e7cd2ac1 133406Author: Peter Hutterer <peter.hutterer@who-t.net> 133407Date: Fri Jan 30 14:35:22 2009 +1000 133408 133409 xkb: Switch the xkb event processing path over to InternalEvents. 133410 133411 Before dropping down into the DIX, convert back into XI events. This is a 133412 temporary solution only, until the DIX is capable of handling InternalEvents 133413 anyway. 133414 133415 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133416 133417commit 64ea6078105f73d1b727619fc123920bc7e4a06c 133418Author: Peter Hutterer <peter.hutterer@who-t.net> 133419Date: Mon Feb 2 14:15:52 2009 +1000 133420 133421 dix: change eventconvert to always return an array of xEvents 133422 133423 Just alloc the memory on demand rather than doing things with EventListPtrs 133424 etc. 133425 133426 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133427 133428commit 4026c63e4eb16481bafc1a41ad67cd2556728d40 133429Author: Peter Hutterer <peter.hutterer@who-t.net> 133430Date: Fri Jan 30 14:34:02 2009 +1000 133431 133432 mi: switch the EQ to contain InternalEvents only. 133433 133434 This gets rid of the nevents parameter, InternalEvents are always a single 133435 item per event. Also remove the special DeviceValuator handling in both 133436 enqueueing and dequeueing. 133437 133438 Custom callback handlers are now broken until fixed. 133439 133440 For bisectability, we copy the InternalEvent back into the XI required during 133441 POE and friends. Consider this a temporary solution. 133442 133443 Note: Because of misc linker bonghits, Xvfb won't link in this revision. 133444 133445 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133446 133447commit 5a827593f9517fd5593751dd8bd90c611de06c5d 133448Author: Peter Hutterer <peter.hutterer@who-t.net> 133449Date: Thu Jan 29 15:26:57 2009 +1000 133450 133451 dix: switch event generation to InternalEvents. 133452 133453 GPE, GKVE, GProxE generate InternalEvents now. 133454 DeviceClassesChangedEvents generates an InternalEvent now, but incomplete! We 133455 need to tack on the information about the new SD in the ClassesChanged events. 133456 133457 Note: To make the progress bisectable, we drop back into XI events at the end of the 133458 Get*Events functions. So the rest of the server still uses XI events. 133459 133460 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133461 133462commit 00b03683d0e5cda40fa23b9fe6a83d7227f86f5d 133463Author: Peter Hutterer <peter.hutterer@who-t.net> 133464Date: Fri Jan 30 14:31:08 2009 +1000 133465 133466 dix: Add temporary conversion function ConvertBackToXI. 133467 133468 Until the InternalEvents are used throughout the server, we can use this one 133469 to drop us back into XI la-la land where every event is the wire format. 133470 133471 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133472 133473commit 269d4d9f2e86fde8c6a28ef0293fb0e44b577886 133474Author: Peter Hutterer <peter.hutterer@who-t.net> 133475Date: Sun Feb 1 09:57:27 2009 +1000 133476 133477 dix: add GetCoreType and GetXIType. 133478 133479 Convert from an InternalEvent type to the matching core/XI type. Currently 133480 only for a few events, those we actually need in the server. 133481 133482 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133483 133484commit 656491921e17b2371057041f4551ad6165067551 133485Author: Peter Hutterer <peter.hutterer@who-t.net> 133486Date: Wed Jan 28 15:27:38 2009 +1000 133487 133488 dix: add InternalEvent -> core/xi event conversion routines. 133489 133490 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133491 133492commit e7867d12541ef9683d5d7fc766e918c13a742981 133493Author: Peter Hutterer <peter.hutterer@who-t.net> 133494Date: Fri Jan 23 15:37:23 2009 +1100 133495 133496 include: add XInternalEvent. 133497 133498 This is the event we want to feed into the EQ and process on the way through. 133499 Only applies for input events for now. 133500 133501 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133502 133503commit 10dcf3ed9c543a6811340567e586ec0e476fcf61 133504Author: Jon TURNEY <jon.turney@dronecode.org.uk> 133505Date: Fri Feb 20 15:22:35 2009 +0000 133506 133507 Cygwin/X: Fix keyboard layout mapping for Latin American keyboards 133508 133509 Fix keyboard layout mapping table entry for Windows keyboard layout 0x80a 133510 (Latin American), layout should (now) be 'latam', not 'la' which is Laos 133511 133512 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 133513 133514commit 9531b96470a2fd812eb3fa0d1e10419331842496 133515Author: Jeremy Huddleston <jeremy@yuffie.local> 133516Date: Sun Feb 22 00:41:52 2009 -0800 133517 133518 XQuartz: DefineInitialRootWindow is gone 133519 133520commit c7c50157b5137a147e69abb70b43a28fac4a9ffa 133521Author: Jeremy Huddleston <jeremy@yuffie.local> 133522Date: Sun Feb 22 00:29:48 2009 -0800 133523 133524 XQuartz: quartzKeyboard compile fixes for recent XKB Changes 133525 133526 Seems we're less divorced from XKB than we'd like for now... hopefully not having any rules installed will still let us override things the way we want to... 133527 133528commit 61abf3189f3f97ec5228d45c035dbfd0c249f844 133529Author: George Staplin <gstaplin@apple.com> 133530Date: Sat Feb 21 22:33:10 2009 -0700 133531 133532 XQuartz: GL: Change from xalloc to xcalloc to potentially avoid 133533 uninitialized data in the __GLXAquaContext. 133534 (cherry picked from commit d3120241f9f48d21f9a6ecfb848434a4a0270855) 133535 133536commit 5587f9b771f5b5427a81d4d657f403667d20f310 133537Author: George Staplin <gstaplin@apple.com> 133538Date: Sat Feb 21 22:22:52 2009 -0700 133539 133540 XQuartz: GL: Fix a bug with an uninitialized GLX data structure. 133541 133542 This was causing a crash randomly, due to random memory contents. 133543 Use xcalloc to prevent this in the future, due to future changes or mistakes. 133544 133545 Set the drawableType to include GLX_PIXMAP_BIT and GLX_PBUFFER_BIT. 133546 The new libGL supports these. 133547 133548 Set the max Pbuffer width/height, based on the results of a test program. 133549 We may someday want to revisit this depending on what users need, so that 133550 we create a CGLContextObj, make it current, and call glGetIntegerv to 133551 gather the information at runtime. 133552 (cherry picked from commit c7e338330943e0d03a99328c740540d03f018d20) 133553 133554commit ab61033700b5383a7a15370dd054eaa80e72e811 133555Author: Alan Coopersmith <alan.coopersmith@sun.com> 133556Date: Thu Feb 19 13:57:28 2009 -0800 133557 133558 Add Extensions section to xorg.conf man page 133559 133560 Extensions section was added in X11R6.8.0 and documented in the release notes: 133561 http://www.x.org/archive/X11R6.8.0/doc/RELNOTES2.html#3 133562 but never made it into the man page. 133563 133564 Also fix a bonus typo. 133565 133566 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 133567 133568commit 372977354c3a09ca77ca13ba0aec00a61046f60d 133569Author: George Staplin <gstaplin@apple.com> 133570Date: Wed Feb 18 21:43:19 2009 -0700 133571 133572 XQuartz: Fix the new ProcAppleDRIDestroyPixmap code REQUEST_SIZE_MATCH. 133573 133574 It had a copy and paste mistake that I didn't notice. :/ 133575 It was using the CreatePixmapReq. 133576 133577 Also add a missing B16 to the end of the length for the DestroyPixmapReq struct. 133578 133579 Now the AppleDRIDestroyPixmap request seem to work. 133580 (cherry picked from commit 295fe25bd8fa2d141291a9d9b6ef7b75fcccb4dd) 133581 133582commit 225853d51d1fb610261ab0c295b1b5a96ce177d5 133583Author: Dan Nicholson <dbn.lists@gmail.com> 133584Date: Thu Feb 19 06:45:05 2009 -0800 133585 133586 xkb: Use cached XKB keymap when rules haven't changed 133587 133588 Rather than compiling a new keymap every time InitKeyboardDeviceStruct 133589 is called, cache the previous keymap and reuse it if the rules have not 133590 changed. 133591 133592 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 133593 Acked-by: Daniel Stone <daniel@fooishbar.org> 133594 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133595 133596commit 64e595d12e05c4df56b0230cc57924b9beb274d3 133597Author: Peter Hutterer <peter.hutterer@who-t.net> 133598Date: Thu Feb 19 09:18:53 2009 +1000 133599 133600 dix: check if keybd->key is valid before getting the keyboard state. 133601 133602 For floating SDs, GetPairedDevice() returns the SD itself, hence keybd->key 133603 may be NULL. 133604 133605 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133606 133607commit e0a451eb7cc812f11db3614b730ba0f07d6ae18c 133608Author: Alan Coopersmith <alan.coopersmith@sun.com> 133609Date: Tue Feb 17 19:37:22 2009 -0800 133610 133611 Obsolete InputDevices keyword in xorg.conf Files section 133612 133613 Was only used to provide a list of input devices that XF86-Misc could use, 133614 now that XF86-Misc is gone, was parsed and logged, then completely ignored. 133615 133616 (Depends on previous patch that introduces OBSOLETE_TOKEN in parser to 133617 make obsolete keywords like InputDevices & RgbPath be non-fatal errors.) 133618 133619 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 133620 Acked-by: Adam Jackson <ajax@redhat.com> 133621 133622commit d2cf562bbad553d7f09b70202134f5b6ada0114e 133623Author: Alan Coopersmith <alan.coopersmith@sun.com> 133624Date: Tue Feb 17 18:48:52 2009 -0800 133625 133626 Make RgbPath keyword in xorg.conf a non-fatal error 133627 133628 Xorg shouldn't refuse to run just because the user has an xorg.conf that 133629 had the previously-used RgbPath keyword in it. 133630 133631 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 133632 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 133633 133634commit 344b1f2b25d627bdf7d802c5831b6a72d22ffe34 133635Author: Jeremy Huddleston <jeremy@yuffie.local> 133636Date: Tue Feb 17 23:25:14 2009 -0800 133637 133638 configure.ac: Support version strings like W.X.Y.Z-XXXXXX 133639 (cherry picked from commit 43967514cd57ad836d7fb85c8c9e58ada07e0232) 133640 133641commit 67faf41b3f3e0cd66e35e3c176a9adb37ff1a759 133642Author: Jeremy Huddleston <jeremy@yuffie.local> 133643Date: Tue Feb 17 23:24:25 2009 -0800 133644 133645 XQuartz: Don't need GlxSetVisualConfig any more 133646 133647 Fixes build failure resulting from 516f8e2cad1311a09764e2633644188d1e3c31bb 133648 (cherry picked from commit 066b17028a35956a089815716e38571f305469c5) 133649 133650commit b349a764e98f0d8f221190157ffa0904b91beca5 133651Author: Eric Anholt <eric@anholt.net> 133652Date: Tue Feb 17 09:59:17 2009 -0800 133653 133654 xinerama: Put the proto version in the code instead using proto headers. 133655 133656 Proto headers updating resulting in the server advertising new versions is 133657 broken. This should be applied to every extension. 133658 133659 This fixes the build against slightly-older xineramaproto. 133660 133661commit 5394b7e66224d20888dd4020f5cb8ca930720fb4 133662Author: Keith Packard <keithp@keithp.com> 133663Date: Fri Feb 6 06:34:35 2009 +0100 133664 133665 Eliminate the shadow clear on transform change 133666 133667 When the crtc transformation changes, the entire crtc must be repainted. 133668 This was being done by clearing the shadow and then painting the rectangle 133669 containing the screen image; the clear being required as the screen image 133670 may not fill the crtc. When changing the transform rapidly, this leads to 133671 flashing. Eliminate the clear by painting the entire crtc instead of just 133672 the screen rectangle. 133673 133674 Signed-off-by: Keith Packard <keithp@keithp.com> 133675 133676commit 76f18b94bd2719a8199334742d021c4d0806187d 133677Author: Keith Packard <keithp@keithp.com> 133678Date: Mon Feb 9 13:09:11 2009 +0100 133679 133680 Add XkbDir to Files config file section 133681 133682 The XKB base directory was not configuable through the config file. 133683 133684 Signed-off-by: Keith Packard <keithp@keithp.com> 133685 133686commit 123093996507c4d3b6dc457240ce00f8ac42f410 133687Author: Adam Jackson <ajax@redhat.com> 133688Date: Tue Feb 17 12:59:16 2009 -0500 133689 133690 RANDR: Fail softly on GetPanning if the screen can't do it. 133691 133692 Just return a zeroed-out reply in that case. This is unambiguous, and 133693 distinguishes "you didn't name a CRTC" from "you named a CRTC that can't 133694 do panning". 133695 133696commit b2ceea3635ec05dca9d4aa2f823b96ae9fce7fe8 133697Author: Maarten Maathuis <madman2003@gmail.com> 133698Date: Tue Feb 17 19:40:59 2009 +0100 133699 133700 Revert "exa: the extent of the valid region is probably much larger than that of the pending damage." 133701 133702 This reverts commit 97c1cbc70216366e92b9371de608ce94e60aa874. 133703 133704 - Sorry for the thinko, pending damage is often not fragmentated. 133705 - Should the dst region become fragmentated, you actually want to copy more to unfragmentate it. 133706 133707commit 3175646b10c602d17d5dd37bdace7c1c7ee92b3d 133708Author: Maarten Maathuis <madman2003@gmail.com> 133709Date: Tue Feb 17 19:06:35 2009 +0100 133710 133711 exa: exaPixmapDirty should use official damage functions. 133712 133713 - Otherwise other users will not be notified of damage. 133714 133715commit 97c1cbc70216366e92b9371de608ce94e60aa874 133716Author: Maarten Maathuis <madman2003@gmail.com> 133717Date: Tue Feb 17 19:01:44 2009 +0100 133718 133719 exa: the extent of the valid region is probably much larger than that of the pending damage. 133720 133721 - Since the default has been changed from nothing to everything. 133722 133723commit 4039603413f9f46d7f725463a70b4a51838e0049 133724Author: Ian Romanick <ian.d.romanick@intel.com> 133725Date: Tue Feb 17 08:27:32 2009 -0800 133726 133727 glx: Inialize best_score before calculating visual scores 133728 133729 This bug was pointed out by Peter Åstrand. 133730 133731 Signed-off-by: Ian Romanick <ian.d.romanick@intel.com> 133732 133733commit bac13e54b164ceff7b921a2d17c4a198b42aeccf 133734Author: George Staplin <gstaplin@apple.com> 133735Date: Mon Feb 16 17:33:19 2009 -0700 133736 133737 XQuartz: Add driWrap.h that I missed in the last commit. 133738 133739 This should complete the GLXPixmap commits for now. 133740 (cherry picked from commit 7e99b55651bf7f6275bb227d02f203fb885a67a2) 133741 133742commit 95052caa2c8a79981cdc7936888a8a47e09415e0 133743Author: George Staplin <gstaplin@apple.com> 133744Date: Mon Feb 16 17:22:18 2009 -0700 133745 133746 XQuartz: Add support for GLXPixmaps to the AppleDRI. 133747 133748 This involved wrapping some GCOps to get the proper behavior 133749 when using X11 raster ops mixed with OpenGL (see driWrap.c). 133750 133751 This extends the AppleDRI protocol with create and destroy pixmap 133752 functions. 133753 133754 The dri.c code has been extended quite a bit to enable this, and 133755 to initialize the wrapping of CreateGC for GCOps. 133756 133757 This has been tested with tests/glxpixmap and proven to work with 133758 the new libGL. Existing applications seem to work fine too. Redraws 133759 all appear to be correct. 133760 133761 There may be some bugs lurking that I haven't found yet. I plan 133762 to drive them out by extending the libGL test suite. 133763 (cherry picked from commit 630518766b01022c49fe3a9e7d501808f71b06e2) 133764 133765commit 970725d23e83285d41e3f7cab678a825d9cc25cf 133766Author: Adam Jackson <ajax@redhat.com> 133767Date: Mon Feb 16 18:44:00 2009 -0500 133768 133769 kdrive: Nuke dead AGP and VGA code. (#19921) 133770 133771commit a932744d983845f81791045cf0dfb6a359790158 133772Author: Chris Ball <cjb@laptop.org> 133773Date: Mon Feb 16 18:27:26 2009 -0500 133774 133775 Build fix, remove export of pciBusAddrToHostAddr 133776 133777 a9d7d659.. (PCI: Remove pciBusAddrToHostAddr and associated nonsense) 133778 removes pciBusAddrToHostAddr(), but not its prototype, resulting in: 133779 133780 ./.libs/libxorg.a(sdksyms.o):(.data.rel+0xe64): undefined reference to 133781 `pciBusAddrToHostAddr' 133782 133783 Signed-off-by: Chris Ball <cjb@laptop.org> 133784 133785commit ab331aa76f0bc6a85a519b2deaafa4ce8a070c53 133786Author: Julien Cristau <jcristau@debian.org> 133787Date: Mon Feb 16 21:56:03 2009 +0100 133788 133789 xfree86: ModeDebug is a boolean, not a string 133790 133791 Signed-off-by: Julien Cristau <jcristau@debian.org> 133792 133793commit f5bf1fdaf36163d5c2f1b9b51df96326ebbb0e9c 133794Author: Peter Hutterer <peter.hutterer@who-t.net> 133795Date: Fri Feb 13 09:56:22 2009 +1000 133796 133797 xkb: Fix wrong colour reference in XKB geometry copying. #20081 133798 133799 base_color and label_color need to reference the color in the destination, not 133800 in the source. 133801 133802 X.Org Bug 20081 <http://bugs.freedesktop.org/show_bug.cgi?id=20081> 133803 133804 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133805 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 133806 133807commit a9d7d659a05e4b6b25e6acd2ce59926911aac0f7 133808Author: Adam Jackson <ajax@redhat.com> 133809Date: Fri Feb 13 18:36:27 2009 -0500 133810 133811 PCI: Remove pciBusAddrToHostAddr and associated nonsense 133812 133813 This was all a glorified no-op. We rely on pciaccess to create device 133814 maps anyway, so we should have no reason to care about what the host 133815 address is. 133816 133817 Acked-by: Ian Romanick <ian.d.romanick at intel.com> 133818 Signed-off-by: Adam Jackson <ajax@redhat.com> 133819 133820commit be8c3d4022faa7d169778ea10d1f11e8f90e1c0c 133821Author: Adam Jackson <ajax@redhat.com> 133822Date: Fri Feb 13 17:54:33 2009 -0500 133823 133824 Linux: Remove dead USE_DEV_FB #ifdefs 133825 133826commit 51ae4d6bbfd95cac8a9cdefcd6f22fbe21a6cb9d 133827Author: Ian Romanick <ian.d.romanick@intel.com> 133828Date: Mon Feb 16 12:08:28 2009 -0800 133829 133830 glx: Add comments around some extension string weirdness 133831 133832commit caa5310e163b040689383f6d25ccab5ef1b2e215 133833Author: Maarten Maathuis <madman2003@gmail.com> 133834Date: Mon Feb 16 20:41:14 2009 +0100 133835 133836 exa: reintroduce src rect optimisation, with a slightly higher threshold. 133837 133838 - I got some feedback on gtkperf line test regression. 133839 - The increased threshold should ensure the destination optimisation is dominant. 133840 133841commit 46eeaf82e228df1c5971a24dc815566516e19be9 133842Author: Maarten Maathuis <madman2003@gmail.com> 133843Date: Mon Feb 16 17:17:14 2009 +0100 133844 133845 exa: fix performance regression from 736b6fbd2c941b6276066cd1503523edebe7bf3d 133846 133847 - The src optimisation is more aggressive and possibly harmful in light of the new initial state of pixmaps. 133848 - There is now actually a performance improvement by almost always keeping the number of rects low. 133849 133850commit 6198373ff2a8e36113bb1dcaebe975530ff01b86 133851Author: David Nusinow <dnusinow@debian.org> 133852Date: Sun Feb 15 18:36:01 2009 -0500 133853 133854 xorg.conf (5) refer to mousedrv (4). Debian #394058 133855 133856 Signed-off-by: Julien Cristau <jcristau@debian.org> 133857 133858commit 2a8b8077d8f6001eb57deba60e1009fc99c28668 133859Author: Alan Hourihane <alanh@vmware.com> 133860Date: Mon Feb 16 11:39:34 2009 +0000 133861 133862 dri2: support glXWaitGL & glXWaitX by copying fake front to front and 133863 vice-versa. 133864 133865commit ddb8d8945d1f44d16adc366b6612eef20ae813f7 133866Author: Peter Åstrand <astrand@cendio.se> 133867Date: Fri Feb 13 10:23:28 2009 +0100 133868 133869 xserver: Avoid sending uninitialized padding data over the network 133870 133871 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133872 133873commit b735a4b4951b607e614682836f24d5fd86c1f7fb 133874Author: Peter Hutterer <peter.hutterer@who-t.net> 133875Date: Thu Feb 12 14:29:23 2009 +1000 133876 133877 dix: Remove lastEventMask variable, substitute with a #define. 133878 133879 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133880 133881commit 772e0f9159ca3ab8fb4e03ed6feee3cc93697724 133882Author: Peter Hutterer <peter.hutterer@who-t.net> 133883Date: Thu Feb 12 13:38:34 2009 +1000 133884 133885 dix: Don't set core events in SetMaskForEvent. 133886 133887 Rather, modify the two callers to call separately for the two different. 133888 events. Unexport SetMaskForEvent too. 133889 And while we're at it, get rid of the MotionFilter macro, because it's one 133890 half confusing and one half pointless. 133891 133892 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133893 133894commit 58f3127919ba5fcb3bb467b6913c28ee1127c82b 133895Author: Peter Hutterer <peter.hutterer@who-t.net> 133896Date: Thu Feb 12 08:37:49 2009 +1000 133897 133898 Xi: remove dynamic mask assignment for XI events. 133899 133900 They end up being the same anyway on startup, so let's not have a dynamic mask 133901 assignment mechanism and instead just hardcode them already. 133902 Also unexport SelectForWindow and remove the valid_masks parameter. We can 133903 check that before calling, since there's only one caller anyway. 133904 133905 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133906 133907commit 7057a9a97b9316b1de954d05451e04223fdd9fd8 133908Author: Peter Hutterer <peter.hutterer@who-t.net> 133909Date: Thu Feb 12 22:14:48 2009 +1000 133910 133911 Remove two more define XKB leftovers. 133912 133913 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 133914 133915commit 736b6fbd2c941b6276066cd1503523edebe7bf3d 133916Author: Michel Dänzer <daenzer@vmware.com> 133917Date: Sun Feb 15 17:44:51 2009 +0100 133918 133919 EXA: Try to prevent the valid regions from growing too many rects. 133920 133921 This helps contain region management overhead in some pathological cases, see 133922 e.g. http://bugs.freedesktop.org/show_bug.cgi?id=16647 . 133923 133924 Signed-off-by: Michel Dänzer <daenzer@vmware.com> 133925 133926commit 45e14f515337159e600a47ebf349692751552af8 133927Author: Chris Ball <cjb@laptop.org> 133928Date: Fri Feb 13 20:32:02 2009 -0500 133929 133930 Build fix for xf86EdidModes.c 133931 133932 Commit 20ac314.. forgets to declare hsize/vsize/refresh. 133933 133934commit ae68be312619ad3bcb2d550a6267d38cce3284f5 133935Author: Adam Jackson <ajax@redhat.com> 133936Date: Fri Feb 13 17:25:32 2009 -0500 133937 133938 Input: Don't re-block SIGIO in xf86SigioReadInput() 133939 133940 It's already blocked on the way in, because we asked for it to be 133941 blocked at sigaction() time. 133942 133943commit c18f454e4a9cab03e5d55b6d761a4a735c943b6b 133944Author: Aaron Plattner <aplattner@nvidia.com> 133945Date: Thu Feb 12 18:27:48 2009 -0800 133946 133947 Query the DDX in ProcRRGetScreenInfo. 133948 133949 This is the RandR 1.1 version of GetScreenResources and needs to re-query the 133950 DDX to see if the mode pool changed. 133951 133952 Fixes Launchpad bug #325115. 133953 Signed-off-by: Adam Jackson <ajax@redhat.com> 133954 (cherry picked from commit 660c2a7d4c6cb52fd0992e9d2008a24805dc8c02) 133955 133956commit 20ac3140ce16af460992b83aa8aeff4afe28c2ee 133957Author: Adam Jackson <ajax@redhat.com> 133958Date: Mon Feb 9 13:36:05 2009 -0500 133959 133960 EDID: Hack for 1366x768 in standard timing descriptors 133961 133962 All you get for standard timing descriptors is horizontal size in 133963 multiples of 8 pixels (which means you can't say 1366) and height in 133964 terms of aspect ratio (which means you can't say 768). You'd like to 133965 just fuzzy-match this by walking the DMT list for sufficiently close 133966 modes, but you can't because DMT is useless and only defines a 1360x768 133967 mode, because it's _also_ specified in terms of character cells despite 133968 providing pixel exact timings. Neither can you use CVT or GTF to 133969 generate the timings, because they _also_ believe that modes have to be 133970 a multiple of 8 pixels. 133971 133972 You'd also hope you could find a timing definition for this in CEA, but 133973 you can't because CEA only defines transmission formats that actually 133974 exist. So there's 480p, 720p, and 1080p, but no 768p. And why would 133975 there be, after all, the encoded signal is never 768p so obviously no 133976 one would ever make a display in that format. 133977 133978 So instead, make a CVT mode since that's likely to be handled well by 133979 just about everything, smash the horizontal active down by 2, and shift 133980 the sync pulse by 1. Underscanning the hard way. 133981 133982 Pass the suicide. 133983 133984commit bcafdfbed6f1e1f901c2459f60f94a0da506bd90 133985Author: Adam Jackson <ajax@redhat.com> 133986Date: Fri Feb 13 14:06:07 2009 -0500 133987 133988 RANDR: Validate entire mode list for interlace and doublescan 133989 133990 Otherwise drivers have to refuse interlace twice: once in the output 133991 config, and once in ->valid_mode() to catch output and config modes. 133992 If you can't do interlaced modes, asking nicely for it in the config 133993 isn't going to suddenly make it work. 133994 133995commit a26c77ff432d2e85a2665fc36fca25143460c476 133996Author: Alan Hourihane <alanh@vmware.com> 133997Date: Fri Feb 13 13:45:08 2009 +0000 133998 133999 glx: fix retval checks when failures occur for drawable creation. 134000 134001commit a38ca0063c91ef221f20bf0c4ac0dc84ce07d557 134002Author: Benjamin Close <Benjamin.Close@clearchain.com> 134003Date: Fri Feb 13 14:44:53 2009 +1030 134004 134005 xext: Use proto header rather than the Xext include file, this prevents userspace being pulled in causing issues 134006 134007 Signed-off-by: Benjamin Close <Benjamin.Close@clearchain.com> 134008 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 134009 Acked-by: Daniel Stone <daniel@fooishbar.org> 134010 134011commit d62c085f36892c327785e85571bd2c7c70513ac6 134012Author: Benjamin Close <Benjamin.Close@clearchain.com> 134013Date: Fri Feb 13 14:44:09 2009 +1030 134014 134015 dix: Remove includes which are not actually used directly 134016 134017 Signed-off-by: Benjamin Close <Benjamin.Close@clearchain.com> 134018 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 134019 Acked-by: Daniel Stone <daniel@fooishbar.org> 134020 134021commit 340f1576afcdaf883d185da356e5d6282aa65e19 134022Author: Peter Hutterer <peter.hutterer@who-t.net> 134023Date: Thu Feb 12 10:44:27 2009 +1000 134024 134025 dix: remove DefineInitialRootWindow() 134026 134027 Obsolete. 134028 134029 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134030 134031commit c178c3f814f7666f43cbf47d24c7543b289f9b1a 134032Author: Peter Hutterer <peter.hutterer@who-t.net> 134033Date: Thu Feb 12 15:29:38 2009 +1000 134034 134035 dix: doxygen-ify enterleave.c 134036 134037 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134038 134039commit cc696a2b7928ec497b6a0df3602fc70dec83a629 134040Author: Peter Hutterer <peter.hutterer@who-t.net> 134041Date: Thu Feb 12 15:05:57 2009 +1000 134042 134043 Doxygen-ify bits of events.c 134044 134045commit 8364bf7374aa5f6d991700f0c02921dc6c638c9f 134046Author: Peter Hutterer <peter.hutterer@who-t.net> 134047Date: Wed Feb 11 16:06:36 2009 +1000 134048 134049 Document the event masks. 134050 134051 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134052 134053commit ed9d58c3c25ee1b3dedbc4c116823c263ccf164d 134054Author: Simon Thum <simon.thum@gmx.de> 134055Date: Sun Feb 8 17:21:09 2009 +0100 134056 134057 dix: refactor pointer acceleration 134058 134059 The algorithm is split in a 2D-specific and a general part. 134060 This potentially allows to accelerate more than just screen motion. 134061 A state machine is intoduced to make code more explicit and readable. 134062 It also improves handling of 'phase 1' mickeys when axial correction 134063 kicks in (corner case). 134064 134065 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134066 134067commit ac470dfb4fadaa0b28b6f8b57f4f13a20842b897 134068Author: Alan Coopersmith <alan.coopersmith@sun.com> 134069Date: Thu Feb 12 18:34:22 2009 -0800 134070 134071 Check for and report errors writing xorg.conf.new from Xorg -configure 134072 134073 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 134074 134075commit bd713794ceaa1b2890522554562103c0a2d50f04 134076Author: Alan Coopersmith <alan.coopersmith@sun.com> 134077Date: Thu Feb 12 18:23:59 2009 -0800 134078 134079 Correct error message if specified config file is not found 134080 134081 By making the "Unable to open config file" header a warning, it was 134082 not appearing with the filename when a config file was specified and 134083 not found. Now we make it an error message again, but only issue 134084 the error if a filename was specified - if none was specified, then 134085 we don't even issue a warning, just the "Using autoconfig" info message. 134086 134087 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 134088 134089commit d229ba7068f0adf1bc8b30b7ccf2fce2c7e8b97b 134090Author: George Staplin <gstaplin@apple.com> 134091Date: Fri Feb 6 12:55:09 2009 -0700 134092 134093 XQuartz: xpr: The dri.c code for pixmaps was wrong in several ways. They weren't 134094 being exported correctly by Xplugin. 134095 134096 This should fix a bug with the surface for a window, when an export fails. 134097 Before the export could fail and leave behind an invalid (freed) pointer in the dix privates. 134098 134099 I have an idea of how to fix the GLXPixmaps now without using CGLSetOffScreen. 134100 134101 This work is a step towards that. The Xplugin will need a small patch to fix an 134102 issue that this change brought forth. 134103 (cherry picked from commit 58c4116c47543b5e30c2232e7bee8efc0b9be176) 134104 134105commit b17d6bed97eddd0773d0845f93f842444b7de0ae 134106Author: Jeremy Huddleston <jeremy@yuffie.local> 134107Date: Thu Feb 12 15:33:47 2009 -0800 134108 134109 XQuartz: Fix copyright info in Info.plist to be prettier when viewed in Finder 134110 (cherry picked from commit 8b607c9c507a68a6abda63831d227d0f8947028c) 134111 134112commit d1ad43a8bc3a93a9cddcb4b0be1d722993bf5dd6 134113Author: Ian Romanick <ian.d.romanick@intel.com> 134114Date: Thu Feb 12 14:05:10 2009 -0800 134115 134116 Enable pbuffers 134117 134118 If a driver does not actually support pbuffer rendering, it can just 134119 not enable any pbuffer fbconfigs. 134120 134121commit 0065896e1ae6d419a42b318b66d6e4d5f17a40ca 134122Author: Colin Harrison <colin.harrison@virgin.net> 134123Date: Sun Nov 2 20:46:37 2008 +0000 134124 134125 Xming: Update keyboard layouts mapping table 134126 134127 An update to the keyboard layout mapping table which updates some no longer valid keyboard layout names 134128 (e.g. de_CH -> ch -variant de) and adds a few more reported layouts 134129 134130 Copyright (C) Colin Harrison 2005-2008 134131 http://www.straightrunning.com/XmingNotes/ 134132 http://sourceforge.net/projects/xming/ 134133 134134 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134135 134136commit d7c27bfe553235e55eb48799129ff9f8e7e0d4f6 134137Author: Colin Harrison <colin.harrison@virgin.net> 134138Date: Tue Feb 3 15:57:29 2009 +0000 134139 134140 Xming: Check clipboard selection ownership after taking it 134141 134142 When you take ownership of a selection you should always check you got 134143 it (X11 ICCCM gospel). 134144 134145 Copyright (C) Colin Harrison 2005-2008 134146 http://www.straightrunning.com/XmingNotes/ 134147 http://sourceforge.net/projects/xming/ 134148 134149 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134150 134151commit cb6315072ff23740bdc5fc2f95dd960601df5664 134152Author: Colin Harrison <colin.harrison@virgin.net> 134153Date: Tue Feb 3 15:54:31 2009 +0000 134154 134155 Xming: Fix a memory leak in clipboard integration code 134156 134157 Free the memory allocated for ReturnData in FlushXEvents(). 134158 134159 Copyright (C) Colin Harrison 2005-2008 134160 http://www.straightrunning.com/XmingNotes/ 134161 http://sourceforge.net/projects/xming/ 134162 134163 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134164 134165commit d54d24dd5f25188b951b82b1c95eb9d36b5be970 134166Author: Colin Harrison <colin.harrison@virgin.net> 134167Date: Tue Feb 3 15:53:43 2009 +0000 134168 134169 Xming: Fix prototype for winProcessXEventsTimeout() in clipboard integration code 134170 134171 Function 'winProcessXEventsTimeout()' is declared 'static Bool' and 134172 defined 'static int' (correct) in 'winclipboardwndproc.c'. This has been 134173 wrong since dot, but luckily was of no significance 134174 134175 Copyright (C) Colin Harrison 2005-2008 134176 http://www.straightrunning.com/XmingNotes/ 134177 http://sourceforge.net/projects/xming/ 134178 134179 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134180 134181commit d8bae80305057acfb8e82341f68090eac0dccb90 134182Author: Colin Harrison <colin.harrison@virgin.net> 134183Date: Tue Feb 3 15:52:57 2009 +0000 134184 134185 Xming: Improve clipboard integration startup in XDMCP mode 134186 134187 Improve clipboard integration startup in XDMCP mode 134188 Stop the "Clipboard not yet started, aborting" message from spamming the log. 134189 134190 Copyright (C) Colin Harrison 2005-2008 134191 http://www.straightrunning.com/XmingNotes/ 134192 http://sourceforge.net/projects/xming/ 134193 134194 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134195 134196commit 7cb1622ce44754d9dcd6f405fb8403bd79c3e7e4 134197Author: Colin Harrison <colin.harrison@virgin.net> 134198Date: Tue Feb 3 15:51:02 2009 +0000 134199 134200 Xming: Select events correctly in clipboard integration code 134201 134202 The clipboard should only select PropertyChange events and do this on 134203 the correct window, the messaging window for the clipboard not the root 134204 window! 134205 134206 Copyright (C) Colin Harrison 2005-2008 134207 http://www.straightrunning.com/XmingNotes/ 134208 http://sourceforge.net/projects/xming/ 134209 134210 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134211 134212commit b997b3849d902894dd5b38673b73a23a47edf502 134213Author: Colin Harrison <colin.harrison@virgin.net> 134214Date: Tue Feb 3 15:48:04 2009 +0000 134215 134216 Xming: Add Colin Harrison to copyright and authors list for clipboard integration files 134217 134218 Add Colin Harrison to copyright and authors list 134219 Adjust license text for multiple authors 134220 134221 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134222 134223commit 5ff289834d7445cc6f60dbb1cf18423e7236311a 134224Author: Jon TURNEY <jon.turney@dronecode.org.uk> 134225Date: Mon Jan 19 17:42:54 2009 +0000 134226 134227 Cygwin/X: Place prototype for winSelectIcons() in a header file 134228 134229 Fixes warning as prototype is now in scope for the definition 134230 134231 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134232 134233commit ab2c98b86f1c3968581dfdad8b67432082feec78 134234Author: Jon TURNEY <jon.turney@dronecode.org.uk> 134235Date: Mon Jan 19 17:36:58 2009 +0000 134236 134237 Cygwin/X: Place prototypes for winCheckKeyPressed() and winFixShiftKeys() in a header file 134238 134239 Fixes warnings as they are now in scope at the definition 134240 134241 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134242 134243commit cc7a4c7c6e051fceacb32997740d83fc0655414e 134244Author: Jon TURNEY <jon.turney@dronecode.org.uk> 134245Date: Mon Jan 19 17:29:28 2009 +0000 134246 134247 cygwin/X: Remove an unused variable to fix a warning 134248 134249 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134250 134251commit 7c2fb098b206c0fb97c18cce8f88387cd0c0e82a 134252Author: Jon TURNEY <jon.turney@dronecode.org.uk> 134253Date: Mon Jan 19 17:22:59 2009 +0000 134254 134255 Cygwin/X: Fix several prototypes to return HICON 134256 134257 Fix prototypes of winOverrideIcon(), winTaskbarIcon() and winOverrideDefaultIcon() to return HICON 134258 Also use HICON type in WINPREFS stucture 134259 Remove various casts these changes make unnecessary 134260 134261 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134262 134263commit 19b3e44ce51cc2f74ba388cb2cd3a26467848e9a 134264Author: Jon TURNEY <jon.turney@dronecode.org.uk> 134265Date: Mon Jan 19 17:03:23 2009 +0000 134266 134267 Cygwin/X: Fix -Wold-style-definition warnings 134268 134269 Fix old-style definition warnings caused by definitions with empty 134270 parameter lists "()", which should be "(void)" to indicate the function 134271 takes no parameters, rather than unspecified parameters 134272 134273 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134274 134275commit 734b23e5982e171031077a2d5d6b5dc2a12e1a70 134276Author: Maarten Maathuis <madman2003@gmail.com> 134277Date: Wed Feb 11 14:31:17 2009 +0100 134278 134279 fb: add fbDoCopy and fbCopyRegion compatibility wrappers. 134280 134281commit 619c4d60eb3a8a51a8300ccf8ab5b98a688e6e12 134282Author: Eric Anholt <eric@anholt.net> 134283Date: Sun Feb 8 12:37:50 2009 +0100 134284 134285 glx: Replace broken GLX visual setup with a fixed "all" mode. 134286 134287 With trying to match depths so that you didn't end up with a depth 24 134288 fbconfig for the 32-bit composite visual, I broke the alpha bits on the depth 134289 24 X visual, which angered other applications. But in fixing that, the 134290 pickFBconfigs code for "minimal" also could end up breaking GLX visuals if 134291 the same FBconfig was chosen for more than one X visual. 134292 We have no reason to not expose as many visuals as possible, but the old 134293 "all" mode didn't match any existing X visuals to GLX visuals, so normal 134294 GL apps didn't work at all. 134295 134296 Instead, replace it with a simple combination of the two modes: Create GLX 134297 visuals by picking unique FBconfigs with as many features as possible for 134298 each X visual in order. Then, for all remaining FBconfigs that are 134299 appropriate for display, add a corresponding X and GLX visual. 134300 134301 This gets all applications (even ones that aren't smart enough to do FBconfigs) 134302 get all the options to get the visual configuration they want. The only 134303 potential downside is that the composite ARGB visual is unique and gets a 134304 nearly full-featured GLX visual (except that the root visual might have taken 134305 the tastiest FBconfig), which means that a dumb compositing manager could 134306 waste resources. Write compositing managers using FBconfigs instead, please. 134307 134308commit 7bbf05d9aa75b536a5321abf51f37852b8f109ed 134309Author: Shunichi Fuji <palglowr@gmail.com> 134310Date: Tue Feb 10 14:44:18 2009 +0900 134311 134312 dix: build fix if PANORAMIX is not defined. 134313 134314 Signed-off-by: Shunichi Fuji <palglowr@gmail.com> 134315 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134316 134317commit 4844bff58f296b2851be4e6b955c3a68d02437a9 134318Author: Peter Hutterer <peter.hutterer@who-t.net> 134319Date: Wed Feb 4 11:50:18 2009 +1000 134320 134321 config: if we can't connect to HAL, listen for a startup notification. 134322 134323 If HAL isn't available when we try to connect, the registered NameOwnerChanged 134324 signal handler waits until HAL is available. Once we connected to HAL, we 134325 unregister the signal handler again. 134326 This allows HAL to be started in parallel or after the server has started. 134327 134328 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134329 134330commit 5b2bbffc12cd7d29e4200943f8736bd6336ae4aa 134331Author: Peter Hutterer <peter.hutterer@who-t.net> 134332Date: Fri Jan 30 16:46:27 2009 +1000 134333 134334 dix: deduplicate SyntheticMotion. 134335 134336 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134337 134338commit 34820f9c1304891bebb4bff6a492288aec4fc0d1 134339Author: Peter Hutterer <peter.hutterer@who-t.net> 134340Date: Fri Jan 30 16:37:19 2009 +1000 134341 134342 dix: remove XineramaChangeToCursor. 134343 134344 It's nearly identical to ChangeToCursor now anyway. 134345 134346 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134347 134348commit 942eae6868b8b0f343b6aa921ddf77e8bb70798a 134349Author: Peter Hutterer <peter.hutterer@who-t.net> 134350Date: Fri Jan 30 16:25:22 2009 +1000 134351 134352 dix: Remove XineramaCheckPhysLimits. 134353 134354 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134355 134356commit bf044f6d10ad9f2baca91a2f04e82dc1af2657cd 134357Author: Peter Hutterer <peter.hutterer@who-t.net> 134358Date: Fri Jan 30 16:14:54 2009 +1000 134359 134360 dix: reduce redunancy in XineramaConfineCursorToWindow. 134361 134362 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134363 134364commit cdce5699f085e717ee1a2ed8390746dfcb9ddadf 134365Author: Peter Hutterer <peter.hutterer@who-t.net> 134366Date: Fri Jan 30 16:09:23 2009 +1000 134367 134368 dix: Get rid of XineramaCheckVirtualMotion. 134369 134370 Yes, this is an ugly piece mess of #ifdefs, but it beats having two nearly 134371 identical functions. 134372 134373 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134374 134375commit 1cb5afdf1ef53e72e0eef28ff6197bf5e86ef758 134376Author: Maarten Maathuis <madman2003@gmail.com> 134377Date: Sun Feb 8 18:59:17 2009 +0100 134378 134379 dix: always NULL pGC->tile.pixmap in Create{Scratch}GC. 134380 134381commit 6fd8fe9cb993f9b733cf7d878bc324acd09aa48d 134382Author: Maarten Maathuis <madman2003@gmail.com> 134383Date: Sun Feb 8 18:43:15 2009 +0100 134384 134385 exa: fix exaValidateGC. 134386 134387 - use DEST in the createPixmap wrapper, because stipple already takes MASK (in case someone uses swappers). 134388 - Anticipate some of the less common situations when fbValidateDrawable will access tile related pixmaps. 134389 134390commit 30daba6903378dd47f386c0012a0667fd2368f3e 134391Author: Maarten Maathuis <madman2003@gmail.com> 134392Date: Sun Feb 8 02:08:13 2009 +0100 134393 134394 exa: All fallbacks should have a GC, remove some code. 134395 134396 - I did some testing with full fallbacks forced by the driver. 134397 - I ran rendercheck, expedite and the (full) x11perf test suite. 134398 - Thanks to ajax for pointing out this should be unneeded. 134399 134400commit d591c7d1b87c61ee8add2fce689e35cf097eed4f 134401Author: Maarten Maathuis <madman2003@gmail.com> 134402Date: Sat Feb 7 20:47:54 2009 +0100 134403 134404 mi: kill a few warnings 134405 134406commit 00226d0b589595cdd45c75e7e28237334a8883b1 134407Author: Maarten Maathuis <madman2003@gmail.com> 134408Date: Fri Feb 6 20:56:49 2009 +0100 134409 134410 mi: clip exposures to pGC->clientClip. 134411 134412commit ffa72793e25f19a388a1369708f472921f721284 134413Author: Maarten Maathuis <madman2003@gmail.com> 134414Date: Fri Feb 6 20:47:01 2009 +0100 134415 134416 exa: Calling exaMarkSync after UTS is the drivers responsibility. 134417 134418commit da086901c6579e41b28b1418fa12ebb82744fc87 134419Author: Matt Turner <mattst88@gmail.com> 134420Date: Fri Feb 6 18:31:12 2009 -0500 134421 134422 Remove unused barrier macros on alpha. 134423 134424commit 68665d78e72c4c564073d66a0f59ac8d3f5fd756 134425Author: Maarten Maathuis <madman2003@gmail.com> 134426Date: Thu Feb 5 17:07:31 2009 +0100 134427 134428 exa: don't use fbCopyNtoN 134429 134430commit 2e76958d304a3c4080d62f32449724eeb9b95d93 134431Author: Maarten Maathuis <madman2003@gmail.com> 134432Date: Wed Feb 4 19:48:03 2009 +0100 134433 134434 fb: move some code to mi 134435 134436commit f06bde3d4bf9a0f245dfa6c65cd98f8e0c7dfdfc 134437Author: Maarten Maathuis <madman2003@gmail.com> 134438Date: Sat Jan 31 19:57:07 2009 +0100 134439 134440 exa: create ExaCheckGetImage 134441 134442commit 258fc4b106dfd7b583999f97caf4d4a7a11569b6 134443Author: Maarten Maathuis <madman2003@gmail.com> 134444Date: Sat Jan 31 19:40:48 2009 +0100 134445 134446 exa: wrap the remainder of exa_unaccel.c 134447 134448commit 6fabf249494bf38b69e3c9c6c9718ee8b12fc6e1 134449Author: Maarten Maathuis <madman2003@gmail.com> 134450Date: Sat Jan 31 19:25:20 2009 +0100 134451 134452 exa: use proper wrapping in exa.c 134453 134454commit 015c99a4ad572f5a1b68054efb070e1ffee1fed0 134455Author: Maarten Maathuis <madman2003@gmail.com> 134456Date: Sat Jan 31 18:53:52 2009 +0100 134457 134458 exa: properly wrap GC functions 134459 134460commit 5e6a06fe691db63ccf2dc7161a1f6c71c09741b7 134461Author: Maarten Maathuis <madman2003@gmail.com> 134462Date: Sat Jan 31 16:30:31 2009 +0100 134463 134464 exa: add GC private 134465 134466commit 5cc67ae94c066dcac78072ad8a819c3b602d8bab 134467Author: Maarten Maathuis <madman2003@gmail.com> 134468Date: Fri Jan 30 19:32:33 2009 +0100 134469 134470 exa: kill of exaImageGlyphBlt 134471 134472 - It serves no obvious purpose, yet it directly accesses many fb 134473 symbols. 134474 134475commit 6483834200fab90164c27c72d4710014edeb4479 134476Author: Maarten Maathuis <madman2003@gmail.com> 134477Date: Fri Jan 30 18:40:46 2009 +0100 134478 134479 exa: Remove one of the many calls directly into the fb layer. 134480 134481commit f112d6bf59964f317ab236221f7bce7fe2603ab8 134482Author: Michel Dänzer <daenzer@vmware.com> 134483Date: Fri Feb 6 18:30:43 2009 +0100 134484 134485 EXA: Try harder to keep current pixmap copy up to date in exaMigrateToward*. 134486 134487 This is probably required, but apparently not sufficient, for making migration 134488 heuristics other than "always" work correctly again. Not that I really care 134489 about them... 134490 134491commit 3948b523893d3d44b6a088340c4252e969613769 134492Author: Michel Dänzer <daenzer@vmware.com> 134493Date: Fri Feb 6 11:37:53 2009 +0100 134494 134495 EXA: Guard empty pending region warning by DEBUG_MIGRATE. 134496 134497 It isn't very useful yet while the damage layer calls us for empty operations, 134498 mostly confuses users. 134499 134500commit 9a1d07ecb74b7c3267a6910af66ada917a525110 134501Author: Tomas Carnecky <tom@dbservice.com> 134502Date: Wed Feb 4 20:20:24 2009 +0100 134503 134504 Fix "warning: cast from pointer to integer of different size" 134505 134506 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 134507 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134508 134509commit 0e15697b53c9448ce9911aa6499b2ea0bda92af6 134510Author: Peter Hutterer <peter.hutterer@who-t.net> 134511Date: Wed Feb 4 10:11:13 2009 +1000 134512 134513 dix: die if we can't activate or init the VCP/VCK. 134514 134515 If we have a busted xkb setup, the XKB initialization on the core devices 134516 fails and leaves us with dev->key->xkbInfo == NULL. This in turn causes 134517 segfaults lateron. 134518 134519 Return BadValue when the XKB configuration for a master device failed, and if 134520 that happens for the VCP/VCK, die semi-gracefully. 134521 The VCP init can only fail on OOM. 134522 134523 Reported by Aaron Plattner. 134524 134525 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134526 Acked-by: Daniel Stone <daniel@fooishbar.org> 134527 Acked-by: Dan Nicholson <dbn.lists@gmail.com> 134528 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 134529 134530commit db836715198a445dc3b32c3df0610f700930483f 134531Author: Adam Jackson <ajax@redhat.com> 134532Date: Wed Feb 4 23:27:32 2009 -0500 134533 134534 Revert "Bus: Remove xf86SetAccessFuncs() and related machinery" 134535 134536 This reverts commit 2fe79990e31ac684637ca3c47198ee0cead10abe. 134537 134538 Reminder: nvidia uses this, because other parts of RAC are broken in 134539 worse ways, and it needs to disable that. 134540 134541commit 2fe79990e31ac684637ca3c47198ee0cead10abe 134542Author: Adam Jackson <ajax@redhat.com> 134543Date: Mon Jan 26 07:11:49 2009 -0500 134544 134545 Bus: Remove xf86SetAccessFuncs() and related machinery 134546 134547 Only mga was using this, and even then only behind an ifdef that no one 134548 ever built. Not a great idea in the first place. 134549 134550commit 9ff6180161267e71ae750498da0ec4889cd5c260 134551Author: Adam Jackson <ajax@redhat.com> 134552Date: Mon Jan 26 06:23:09 2009 -0500 134553 134554 Bus: Delete some obfuscatory macros 134555 134556commit 8eca37cb8934e79f02291460eb7b49a5d528f391 134557Author: Adam Jackson <ajax@redhat.com> 134558Date: Wed Feb 4 22:48:52 2009 -0500 134559 134560 Bus: Remove unused xf86ExtractTypeFromList 134561 134562commit b013111e98d50f5962e9b275eb28d969d5c9576e 134563Author: Adam Jackson <ajax@redhat.com> 134564Date: Mon Jan 26 03:14:18 2009 -0500 134565 134566 Bus: make the resource list static. 134567 134568commit bde028dd8d856f8d47c39e3c6d3731c34c398c74 134569Author: Adam Jackson <ajax@redhat.com> 134570Date: Mon Jan 26 22:58:58 2009 -0500 134571 134572 loader: Remove icache flushes on alpha. 134573 134574 If libdl doesn't already get this right for you, you have two problems. 134575 134576commit 910a1e88d23e0b7e4ca4679ab50a883a918f2590 134577Author: Adam Jackson <ajax@redhat.com> 134578Date: Mon Jan 26 07:54:07 2009 -0500 134579 134580 loader: Remove a vestige of archive support 134581 134582commit 366f23c6eb504fc23112f121769bcb719948474f 134583Author: Adam Jackson <ajax@redhat.com> 134584Date: Mon Jan 26 07:53:05 2009 -0500 134585 134586 loader: Remove useless call to LoaderGetOS 134587 134588 We know what OS you're running on anyway, it's printed in the uname 134589 string at the top of the log. 134590 134591commit 05e415a5a3f376e28955be3bf976390e8631e25f 134592Author: Adam Jackson <ajax@redhat.com> 134593Date: Mon Jan 26 07:52:07 2009 -0500 134594 134595 loader: Simplify loader magic 134596 134597commit bf1ca06a1693cbe5415a152936939ce3b088a022 134598Author: Adam Jackson <ajax@redhat.com> 134599Date: Mon Jan 26 07:51:35 2009 -0500 134600 134601 loader: Make a comment slightly less stupid 134602 134603commit 499908aeca9a31692c787070e23bfaa92d1b32a4 134604Author: Adam Jackson <ajax@redhat.com> 134605Date: Mon Jan 26 07:50:00 2009 -0500 134606 134607 loader: code motion. 134608 134609commit 28b6b1519c4c526662302fac4ed9b7f540bafd4a 134610Author: Adam Jackson <ajax@redhat.com> 134611Date: Mon Jan 26 07:47:46 2009 -0500 134612 134613 loader: Simplify handle allocation and refcounting. 134614 134615commit 0bad0552bfd4d10529ab51d618af8d08c5aeea71 134616Author: Adam Jackson <ajax@redhat.com> 134617Date: Wed Feb 4 17:20:24 2009 -0500 134618 134619 XKB: Remove -kb and +kb from -help text and man page. 134620 134621commit 7968823cbc02615e1080c0d3f34dcebe14ea1771 134622Author: Keith Packard <keithp@keithp.com> 134623Date: Tue Feb 3 20:18:37 2009 -0800 134624 134625 Handle the combination of panning and crtc transforms 134626 134627 This patch gets the shadow scanout buffer repainted on panning area changes. 134628 It does not, however, track the mouse correctly. 134629 134630 Signed-off-by: Keith Packard <keithp@keithp.com> 134631 134632commit 763df9eec79a867978efc5138d8635a46c1a0d17 134633Author: Keith Packard <keithp@keithp.com> 134634Date: Tue Feb 3 20:15:41 2009 -0800 134635 134636 Damage re-used shadow scanout buffer using new transforms. 134637 134638 When the shadow scanout buffer can be re-used, the underlying framebuffer 134639 area must be damaged so that the scanout will be repainted. This patch 134640 delays the addition of that damaged area until after the transform in the 134641 crtc has been updated, otherwise the old transform would have been used and 134642 the wrong area repainted. 134643 134644 Signed-off-by: Keith Packard <keithp@keithp.com> 134645 134646commit 62fc98cb88e4e8b636f343453fc1168a87c58972 134647Author: Keith Packard <keithp@keithp.com> 134648Date: Tue Feb 3 20:12:42 2009 -0800 134649 134650 Handle matrix computation overflow in RRTransformCompute 134651 134652 If the computation of the composite fixed-point transform for RandR 134653 overflows at any point, take the resulting floating point transform and 134654 scale that back to fit in a fixed point matrix. This ensures that a matrix 134655 will always be available, although perhaps at reduced precision. Someday we 134656 should add floating point matrices to Render. 134657 134658 Signed-off-by: Keith Packard <keithp@keithp.com> 134659 134660commit 7a3ddef3bc27982f3558183c74fc8e365cf2e7fd 134661Author: Tomas Carnecky <tom@dbservice.com> 134662Date: Wed Feb 4 00:07:02 2009 +0100 134663 134664 fbdevHWGetRec() doesn't return the private data 134665 134666 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 134667 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134668 134669commit e04d9ab999978995b06f4b702293b96f69b0e8c3 134670Author: Tomas Carnecky <tom@dbservice.com> 134671Date: Wed Feb 4 00:08:21 2009 +0100 134672 134673 DBUS_API_SUBJECT_TO_CHANGE is already defined on the command line 134674 134675 dbus-core.c:30:1: warning: "DBUS_API_SUBJECT_TO_CHANGE" redefined 134676 <command-line>: warning: this is the location of the previous definition 134677 134678 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 134679 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134680 134681commit 9f60d127b5371f2853dcc53903c6ff8c7670ad1b 134682Author: Tomas Carnecky <tom@dbservice.com> 134683Date: Wed Feb 4 00:07:40 2009 +0100 134684 134685 Fix "warning: braces around scalar initializer" 134686 134687 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 134688 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134689 134690commit a88995dfb88719ae4a03483dc571dbbe83565285 134691Author: Tomas Carnecky <tom@dbservice.com> 134692Date: Wed Feb 4 00:03:58 2009 +0100 134693 134694 Make gcc happy: correct third argument of CopyGetMasterEvent() 134695 134696 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 134697 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134698 134699commit dd3f4e829286e62f4c5c07ee93f9f576876acbf9 134700Author: Tomas Carnecky <tom@dbservice.com> 134701Date: Wed Feb 4 00:02:24 2009 +0100 134702 134703 Fix "warning: cast to pointer from integer of different size" 134704 134705 Add parenthesis around the whole expression. 134706 134707 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 134708 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134709 134710commit 731a907ac21a753f306b53706f689469cefeb953 134711Author: Tomas Carnecky <tom@dbservice.com> 134712Date: Tue Feb 3 23:58:25 2009 +0100 134713 134714 Fix "warning: XXX defined but not used" 134715 134716 generic.c:80: warning: ‘read_legacy_video_BIOS’ defined but not used 134717 134718 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 134719 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134720 134721commit bc57efffe654c493ceb62555cccb7926f4a646f1 134722Author: Tomas Carnecky <tom@dbservice.com> 134723Date: Tue Feb 3 23:54:53 2009 +0100 134724 134725 Fix "warning: unused variable XXX" 134726 134727 events.c:4614: warning: unused variable ‘kbd’ 134728 xkbUtils.c:361: warning: unused variable ‘maxKeysPerMod’ 134729 xf86Events.c:409: warning: unused variable ‘ke’ 134730 generic.c:131: warning: unused variable ‘cs’ 134731 generic.c:130: warning: unused variable ‘size’ 134732 xf86RandR12.c:591: warning: unused variable ‘crtc’ 134733 134734 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 134735 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134736 134737commit 111fdef74dfc7e288cef675b171b673bf64fe056 134738Author: Tomas Carnecky <tom@dbservice.com> 134739Date: Tue Feb 3 21:18:53 2009 +0100 134740 134741 ISO C90: Move declaration to the beginning of the function 134742 134743 devices.c: In function ‘DoChangeKeyboardControl’: 134744 devices.c:1768: warning: ISO C90 forbids mixed declarations and code 134745 134746 Signed-off-by: Tomas Carnecky <tom@dbservice.com> 134747 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134748 134749commit 9fe9b6e4ef669b192ee349e3290db5d2aeea273c 134750Author: Peter Hutterer <peter.hutterer@who-t.net> 134751Date: Mon Feb 2 10:20:13 2009 +1000 134752 134753 mi: don't call UpdateSpriteForScreen if we have Xinerama enabled. #18668 134754 134755 In Xinerama all windows hang off the first root window. Crossing the screens 134756 must not reset the spriteTrace, otherwise picking fails and events are sent to 134757 the root window. 134758 134759 X.Org Bug 18668 <http://bugs.freedesktop.org/show_bug.cgi?id=18668> 134760 134761 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134762 134763commit f80332a04d0714a74f16b2552cccf3457e6494af 134764Author: Peter Hutterer <peter.hutterer@who-t.net> 134765Date: Mon Feb 2 14:28:16 2009 +1000 134766 134767 dix: for core events, pass the core filter into DeliverEventsToWindow 134768 134769 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134770 134771commit 01879e583b0ce0d8fddc37e36d9b69fe4f1d1e72 134772Author: Aaron Plattner <aplattner@nvidia.com> 134773Date: Tue Feb 3 13:49:19 2009 -0800 134774 134775 Don't enable XAA offscreen pixmaps with Option "XaaOffscreenPixmaps" "no". 134776 134777 xf86IsOptionSet is rarely the right function to use for boolean options because 134778 it returns TRUE whenever the option is present, even if its value is "no", 134779 "off", "0", etc. 134780 134781commit 5623c27700b7b23a8dbbd8c8f45e5d4fa0c667e3 134782Author: Alan Coopersmith <alan.coopersmith@sun.com> 134783Date: Mon Feb 2 19:25:14 2009 -0800 134784 134785 Constify atom name strings 134786 134787 Changes MakeAtom to take a const char * and NameForAtom to return them, 134788 since many callers pass pointers to constant strings stored in read-only 134789 ELF sections. Updates in-tree callers as necessary to clear const 134790 mismatch warnings introduced by this change. 134791 134792 Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com> 134793 Acked-by: Peter Hutterer <peter.hutterer@who-t.net> 134794 134795commit 6869efae74381e5305b2d6517bf286e3ef7fdcb7 134796Author: Jeremy Huddleston <jeremy@yuffie.local> 134797Date: Tue Feb 3 09:58:51 2009 -0800 134798 134799 XQuartz: Accept bundle version 2.1.6 to work with our startup method 134800 (cherry picked from commit 7f2966557374cb5ea22bc7b73d5ca04891eb5e78) 134801 134802commit 43a1c91a3dac4f0c8cea5de1b9785e349e09978a 134803Author: Peter Hutterer <peter.hutterer@who-t.net> 134804Date: Sun Feb 1 20:44:27 2009 +1000 134805 134806 Xext: fix typo in GEEventFill macro 134807 134808 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134809 134810commit 412e09e1e37b74421abd4dd9606536d2785da440 134811Author: Peter Hutterer <peter.hutterer@who-t.net> 134812Date: Thu Jan 15 11:03:03 2009 +1000 134813 134814 dix: Use GenericEvent instead of LASTEvent to check for core events. 134815 134816 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134817 134818commit 9f3c5d46f81887b2b413b3e6cc9df8282f06522c 134819Author: Peter Hutterer <peter.hutterer@who-t.net> 134820Date: Mon Feb 2 12:55:39 2009 +1000 134821 134822 Xi: initialize the device before applying button map changes. 134823 134824 This part got lost in bc909f71367a0. 134825 Reported by Tomas Carnecky. 134826 134827 Acked-by: Daniel Stone <daniel@fooishbar.org> 134828 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134829 134830commit 6bf6a4c47e94780bf0eef47702f21a505c336556 134831Author: Simon Thum <simon.thum@gmx.de> 134832Date: Fri Jan 23 12:06:16 2009 +0100 134833 134834 Xi: create well-known atoms on demand, rather than preinit them 134835 134836 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134837 134838commit 32adf8d34fb7db9b78859241f2670349b75c3f7e 134839Author: Peter Hutterer <peter.hutterer@redhat.com> 134840Date: Wed Jan 28 20:52:45 2009 +1000 134841 134842 Xi: If the MD doesn't have a key/kbdfeed class, init the keyboard. 134843 134844 This fixes crashes on mouse/keyboard combos where the mouse is both a keyboard 134845 and a mouse. Upon copying the SD's key info into the VCP, the NULL xkbInfo of 134846 the VCP would crash the server. 134847 134848 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134849 134850commit caade11b8465a743b2ecc890a4d180277e9ca716 134851Author: Peter Hutterer <peter.hutterer@redhat.com> 134852Date: Wed Jan 28 20:44:47 2009 +1000 134853 134854 Xi: remove now obsolete oldXkbInfo. 134855 134856 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 134857 134858commit 5e0967f5fc2e0fcf2a55128eb03931cf6f5709e9 134859Author: Peter Hutterer <peter.hutterer@who-t.net> 134860Date: Fri Dec 12 10:49:53 2008 +1000 134861 134862 dix: add SetBit(arr, bit) and ClearBit(arr, bit) to include/inputstr.h 134863 134864 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134865 134866commit faeb18eb7eb20f15528f9c8004d843146ebc2a50 134867Author: Peter Hutterer <peter.hutterer@who-t.net> 134868Date: Tue Jan 27 15:40:31 2009 +1000 134869 134870 Xext: rename shape's EventType to ShapeEventType to avoid name clashing. 134871 134872 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134873 134874commit cfd3443fe81685e91a53063dee58e24a5684dc29 134875Author: Peter Hutterer <peter.hutterer@who-t.net> 134876Date: Wed Jan 28 13:34:10 2009 +1000 134877 134878 Xext: rename saver's EventType to SaverEventType. 134879 134880 Avoid namespace clashing with the internal events. 134881 134882 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134883 134884commit 8460425740c72e58bcb897557f033e2836251160 134885Author: Peter Hutterer <peter.hutterer@who-t.net> 134886Date: Wed Dec 10 12:31:59 2008 +0000 134887 134888 render: rename SetBit to RenderSetBit. 134889 134890 Avoiding namespace collision with the SetBit macro soon to be used in the 134891 input code. 134892 134893 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134894 134895commit 940a7aeebc9b042b0208ee475ef162c485ac2274 134896Author: Peter Hutterer <peter.hutterer@who-t.net> 134897Date: Tue Jan 27 15:52:49 2009 +1000 134898 134899 include: remove now-unused sempahore macros. 134900 134901 Obsolete with the new enter/leave model. 134902 134903 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 134904 134905commit f5c7f8a3ab5723b6c1c3cd21b90a87fd7f37cb7f 134906Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 134907Date: Mon Feb 2 12:30:57 2009 -0800 134908 134909 XQuartz: GLX: Fix accidental exclusion of glBlitFramebufferEXT on Leopard 134910 (cherry picked from commit 8620579483789ae9ac1b68dbefe32b70011386de) 134911 134912commit 5100d829a4d71ce4a9fbc2b81694a1fb90066ccf 134913Author: Eric Anholt <eric@anholt.net> 134914Date: Mon Feb 2 10:13:46 2009 -0800 134915 134916 glx: Don't match fbconfigs to visuals with mismatched channel masks. 134917 134918 This fixes at least one known bug, where the depth 32 visual would end up 134919 with a depth 24 fbconfig attached, angering compiz. 134920 134921commit 4475ae036c3d5a06a6ce88a359ce2a9a8ae9b4f1 134922Author: Jon TURNEY <jon.turney@dronecode.org.uk> 134923Date: Mon Jan 26 13:15:07 2009 +0000 134924 134925 Cygwin/X: Fix compilation for mandatory XKB 134926 134927 Fix a stray '}' 134928 Update to use RMLVO interface 134929 Remove g_winInfo.xkb.disable, can never be set since noXkbExtension has been removed 134930 Change to retrieve modifier key state using XkbStateFieldFromRec() from Colin Harrison 134931 Update to use XKB defaults from xkb-config.h 134932 134933 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 134934 134935commit e2b4d3d4a1994347979508fbb483cd55c2232502 134936Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 134937Date: Sun Feb 1 15:08:02 2009 -0800 134938 134939 XQuartz: GLX: OpenGL.framework on Tiger doesn't have glBlitFramebufferEXT 134940 (cherry picked from commit e01662cadcaa4052e3a5aa82cbaed00a2d0220e0) 134941 134942commit 66a6fbbfa3a1fd0d91fe61cf2ab299d0e3df46b9 134943Author: Jim Huang <jserv@kaffe.org> 134944Date: Sun Feb 1 09:19:20 2009 -0800 134945 134946 kdrive: Move a close() to the right place so we don't close(-1) normally. 134947 134948 Bug #18343 134949 134950commit 0055fe66d5f73742cafab868ccdb7a6f36ea1dd5 134951Author: Manuel Bouyer <bouyer@netbsd.org> 134952Date: Sun Feb 1 09:14:19 2009 -0800 134953 134954 netbsd: Force the use of ev56 instructions for register access on ev56. 134955 134956 This avoids 32-bit access which might affect other registers. The linux code 134957 uses gcc flags to get this to happen, but this seems like more of a sure thing. 134958 134959commit 1375c1b07df114d864a0a06786bc75a3a4f65177 134960Author: Manuel Bouyer <bouyer@netbsd.org> 134961Date: Sun Feb 1 09:12:21 2009 -0800 134962 134963 netbsd: Fix alpha ev6 support. 134964 134965 Bug #19721 134966 134967commit ad8ffbe7cb50387e6a9cc4a24fea0d9c13b7bfc8 134968Author: John Hein <jhein@timing.com> 134969Date: Fri Jan 30 21:30:35 2009 -0800 134970 134971 Actually require bigfontproto when it's enabled. Bug #13710. 134972 134973commit 29b3b88dc744f4919c6709747ddb7baac47486c5 134974Author: Pierre Willenbrock <pierre@pirsoft.de> 134975Date: Fri Jan 30 21:16:48 2009 -0800 134976 134977 Prevent double unref of glxdrawables 134978 134979 Found by valgrind. Bug #18917. 134980 134981commit 60bcdd687040db76490851d4b459284ce37020e0 134982Author: Alan Coopersmith <alan.coopersmith@sun.com> 134983Date: Fri Jan 30 21:44:08 2009 -0800 134984 134985 x11-input.fdi: Add options needed to handle adding USB devices on Solaris 134986 134987commit 4de5705545792d77c9990bbec3e263a517757a2a 134988Author: Keith Packard <keithp@keithp.com> 134989Date: Fri Jan 30 20:17:10 2009 -0800 134990 134991 RandR crtcs not fetched correctly when primary output is set. 134992 134993 Primary outputs may not have a CRTC. 134994 Loops fetching CRTCs respecting primary output were broken. 134995 134996 Signed-off-by: Keith Packard <keithp@keithp.com> 134997 134998commit f716e3f3445d443cbc6507d27f806e9ad387120a 134999Author: Eric Anholt <eric@anholt.net> 135000Date: Fri Jan 30 20:10:21 2009 -0800 135001 135002 modes: Protect xf86_crtc_supports_gamma() from non-RandR 1.2 drivers. 135003 135004commit 317f2b4a9fe4b606975711bc332166a82db5087d 135005Author: Eric Anholt <eric@anholt.net> 135006Date: Fri Jan 30 19:06:17 2009 -0800 135007 135008 randr: Avoid re-querying the configuration on everything but GetScreenResources. 135009 135010 The new path should only re-query on the other requests when we haven't 135011 gathered the information from the DDX yet (such as with a non-RandR 1.2 DDX). 135012 135013 Bug #19037. 135014 135015commit c1f2be1f3fd0c80cb4c85b98140b98aa9311242b 135016Author: Keith Packard <keithp@keithp.com> 135017Date: Fri Jan 30 19:14:36 2009 -0800 135018 135019 RandR SetOutputPrimary should work with or without a crtc connected 135020 135021 The test was inverted from the protocol spec, and besides, the test is bogus 135022 anyways. 135023 135024 Signed-off-by: Keith Packard <keithp@keithp.com> 135025 135026commit dd098501d800571e71d06ffc936635a6c9d44e05 135027Author: Eric Anholt <eric@anholt.net> 135028Date: Fri Jan 30 16:17:58 2009 -0800 135029 135030 Move the apple fat binary hacks back to a header file, and make it apple-only. 135031 135032commit 6e49fdd2c839b6244c23ce95c3ae76a1a265b986 135033Author: Keith Packard <keithp@keithp.com> 135034Date: Fri Jan 30 15:43:04 2009 -0800 135035 135036 Make RandR CRTC info report panning area instead of just crtc area 135037 135038 This makes the RandR info consistent with the Xinerama info. 135039 135040 Signed-off-by: Keith Packard <keithp@keithp.com> 135041 135042commit ea309e47457156b60aadbf113f04e5b6851029c8 135043Author: Keith Packard <keithp@keithp.com> 135044Date: Fri Jan 30 14:19:41 2009 -0800 135045 135046 Make crtc_notify wrap/unwrap code do nothing unless mode code is inuse 135047 135048 Drivers not using the new hw/xfree86/modes code would crash in DRI due to 135049 that code trying to monitor CRTC changes. 135050 135051 Signed-off-by: Keith Packard <keithp@keithp.com> 135052 135053commit fff00df94d7ebd18a8e24537ec96073717375a3f 135054Author: Keith Packard <keithp@keithp.com> 135055Date: Fri Jan 30 14:16:12 2009 -0800 135056 135057 Ignore EDID-supplied monitor physical sizes for core screen size 135058 135059 We report the EDID values in RandR, and we let people configure whatever 135060 they like for the screen in xorg.conf. Reporting the EDID values in the core 135061 means applications get inconsistent font sizes in the default configuration. 135062 135063 Signed-off-by: Keith Packard <keithp@keithp.com> 135064 135065commit 4e7bc2d7a9879297aff0924febeca9bd3b6c336d 135066Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 135067Date: Fri Jan 30 03:56:39 2009 -0200 135068 135069 Work around inclusion of <X11/extensions/panoramiXext.h> 135070 135071 The X Server build only needs the macros PANORAMIX_MAJOR_VERSION 135072 and PANORAMIX_MINOR_VERSION from that header. 135073 Addition of extra prototypes to <X11/extensions/panoramiXext.h> 135074 caused a X Server build failure. 135075 135076commit 15ac25627e7239629be59adedda5d8a9deb330ef 135077Author: Matt Turner <mattst88@gmail.com> 135078Date: Thu Jan 29 14:40:27 2009 -0500 135079 135080 Remove extra / from Loading module line 135081 135082 In the X log, upon module load, it prints a line similar to the following. 135083 135084 > (II) Loading /usr/lib64/xorg/modules/extensions//libdbe.so 135085 135086 The attached patch removes the extra / before the module name. 135087 135088 Code already exists in hw/xfree86/loader/loadmod.c:InitPathList to add a 135089 trailing slash if needed, removing the one added by sprintf is harmless. 135090 135091 Signed-off-by: James Cloos <cloos@jhcloos.com> 135092 135093commit 3903e23153a0d47704050944fa83a7b31bc2eebc 135094Author: Maarten Maathuis <madman2003@gmail.com> 135095Date: Thu Jan 29 12:42:57 2009 +0100 135096 135097 XKB: ifdef XKB is dead, yet it was still present in a few places. 135098 135099commit 756a2c848305aa06bfe636314e28ad7f97c1168a 135100Author: Pierre Willenbrock <pierre@pirsoft.de> 135101Date: Wed Jan 28 22:18:50 2009 +0100 135102 135103 Fix duplicate code, off-by one in space calculation, not initialized members 135104 135105commit b55db937fee5ddf34518a54f352c0ee9b8cc2c2c 135106Author: Søren Sandmann Pedersen <sandmann@redhat.com> 135107Date: Wed Jan 28 17:33:25 2009 -0500 135108 135109 [shadow] Store the closure in the buffer 135110 135111commit 82fc102568b3d6b0daeb6c5f5b3a1310a7f14fcd 135112Author: Peter Hutterer <peter.hutterer@redhat.com> 135113Date: Wed Jan 28 07:51:32 2009 +1000 135114 135115 Disable Xnest by default. 135116 135117 Xneest doesn't build right now and is in the process of being removed. So 135118 let's not build it by default. 135119 135120commit f7585d020593345e7a61b1fe1a517c38da8f6d79 135121Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 135122Date: Tue Jan 27 03:13:28 2009 -0200 135123 135124 Convert kdrive libraries to libtool convenience libraries. 135125 135126 Also correct a link failure due to unresolved symbols. This 135127 is arguably a libtool/ranlib/ld bug, that "may" be corrected 135128 by linking all convenience libraries in a single one. But in 135129 this case, it was preferred to just add a linker option to 135130 Xfake_LDFLAGS, to force linkage of all libraries. 135131 135132 This corrects #19725. 135133 135134commit b43a4e2d38adb9ff16204218f24df93cfd820d1b 135135Author: Magnus Kessler <Magnus.Kessler@gmx.net> 135136Date: Tue Jan 27 09:30:37 2009 +1000 135137 135138 Xvfb: add missing include for new xkbsrv.h 135139 135140 Commit 08363c5830bdea34012dcd954b45ccfdc79a3a7e added call to 135141 XkbGetRulesDflts defined in xkbsrv.h 135142 135143 Signed-off-by: Magnus Kessler <Magnus.Kessler@gmx.net> 135144 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135145 135146commit a7e6424d992151c504a75bb807fa34c57ad52cad 135147Author: Peter Hutterer <peter.hutterer@who-t.net> 135148Date: Fri Jan 23 17:11:57 2009 +1100 135149 135150 dix: remove obsolete comment. Event lists should not be allocated by the DDX 135151 135152commit b137a0cdb1d517a807d081fbde8e8dc0539fb3ca 135153Author: George Staplin <gstaplin@apple.com> 135154Date: Mon Jan 26 03:02:38 2009 -0700 135155 135156 XQuartz: GL: indirect.c changes to fix the build with newer OpenGL frameworks. 135157 (cherry picked from commit 143224405ba74929c702a95de52b56df140b0d1b) 135158 135159commit 34b896e50fa154ad890a458c60932964421e739f 135160Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 135161Date: Fri Jan 16 19:26:07 2009 -0800 135162 135163 XQuartz: SnowLeopard: Help system book name changed in 10.6 135164 (cherry picked from commit b57cb05c69acbedb00a97234099ea104309aa2cb) 135165 135166commit c7d62d74bbd6e0067f511fd910b9e633576befb7 135167Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 135168Date: Sun Jan 25 19:19:01 2009 -0200 135169 135170 Explicitly add libxorg_la_LIBADD to libxorg_la_DEPENDENCIES 135171 135172 Otherwise, for example, when hacking config/*.c, it is required to 135173 run make clean on that directory, to ensure the proper libconfig.a 135174 will be linked in the generated Xorg binary. 135175 135176commit 0eec7348875035792097f606487a558422774e2c 135177Author: Magnus Kessler <Magnus.Kessler@gmx.net> 135178Date: Fri Jan 23 12:11:00 2009 -0200 135179 135180 .gitignore: ignore sdksyms.dep 135181 135182 Signed-off-by: Magnus Kessler <Magnus.Kessler@gmx.net> 135183 135184commit 88efe9c76f85423d952ff38e0cfd8b690439902a 135185Author: Magnus Kessler <Magnus.Kessler@gmx.net> 135186Date: Fri Jan 23 11:25:00 2009 -0200 135187 135188 Xvfb: Remove unused function GetLK201Mappings in InitInput.c 135189 135190 Silence a gcc warning. After commit 08363c5830bdea34012dcd954b45ccfdc79a3a7e 135191 GetLK201Mappings is no longer needed. 135192 135193 Signed-off-by: Magnus Kessler <Magnus.Kessler@gmx.net> 135194 135195commit 6ef46c40e62def4841a4cff4e0b443516a2ed782 135196Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 135197Date: Sun Jan 25 19:01:10 2009 -0200 135198 135199 Update xnest keyboard code to match xephyr/kdrive. 135200 135201 Xnest was not updated in the last batch of xkb changes. This 135202 patch is basically cut&paste from hw/kdrive/src/kinput.c and 135203 hw/kdrive/ephyr/ephyr.c, and appears to generate a Xnest as 135204 functional as before the xkb changes. 135205 135206commit 759348f13ffa02fb1804238189ffce1c92d33e2a 135207Author: Adam Jackson <ajax@redhat.com> 135208Date: Thu Jan 22 22:09:11 2009 -0500 135209 135210 Xephyr: Add some rudimentary Xnest command line compat 135211 135212commit 01ffaf6834fd6a3b1ddbd33d0f0b8d086be0737e 135213Author: Peter Hutterer <peter.hutterer@who-t.net> 135214Date: Mon Jan 19 21:11:05 2009 +1100 135215 135216 Xi: define a range of axis labels. 135217 135218 This is copied from linux/input.h, presumably that's the ones at least the 135219 Linux kernel can give us for any device. 135220 135221 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135222 135223commit 132b464d734b077038e19b21e46d3a6258f4b998 135224Author: Adam Jackson <ajax@redhat.com> 135225Date: Thu Jan 22 02:11:16 2009 -0500 135226 135227 Remove a bunch of useless casts. 135228 135229 We've had void * for twenty years now people let's try to act like we 135230 know how it works. 135231 135232commit 0fb4390526bb829ab17ff4635d41a3012f63c1b2 135233Author: Adam Jackson <ajax@redhat.com> 135234Date: Thu Jan 22 01:13:17 2009 -0500 135235 135236 Remove xorgcfg 'Options' list. 135237 135238commit b5ece46b76c9182a077141505cc0e788c0fca3ea 135239Author: Daniel Stone <daniel@fooishbar.org> 135240Date: Sat Jan 17 20:07:40 2009 +0200 135241 135242 Input: Clarify valuator FatalError messages 135243 135244 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135245 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135246 135247commit ddb46c1421a98190313b79621c0f5ae43b97d3f6 135248Author: Daniel Stone <daniel@fooishbar.org> 135249Date: Sat Jan 17 20:03:10 2009 +0200 135250 135251 Input: Use previous state in valuator events 135252 135253 Make valuator event state match other events by using the device state 135254 from before processing the event, not after. Also, we already check the 135255 number of valuators in UpdateDeviceState, so no need to do it again. 135256 135257 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135258 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135259 135260commit 18e51911f5748f9a27a18cd6bb421c899f5b048f 135261Author: Daniel Stone <daniel@fooishbar.org> 135262Date: Sat Jan 17 13:44:47 2009 +0200 135263 135264 Input: Add postdown to ButtonClassRec 135265 135266 This mirrors that in KeyClassRec: the state of the buttons as posted to 135267 GetPointerEvents, rather than the state of the buttons as processed by 135268 ProcessOtherEvent and friends. 135269 135270 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135271 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135272 135273commit c05f3bc935bef4d98216d38c0e98bd57ced382df 135274Author: Daniel Stone <daniel@fooishbar.org> 135275Date: Sat Jan 17 19:35:15 2009 +0200 135276 135277 Input: Don't allow DDX to generate repeat events 135278 135279 We always use soft-repeat at the moment; XKB posts a release/press sequence, 135280 which admittedly needs cleaning up, but that's for another day. 135281 135282 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135283 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135284 135285commit 6aef4e96affcc26b9415c6bc9c9bebb5af05a1a7 135286Author: Daniel Stone <daniel@fooishbar.org> 135287Date: Sat Jan 17 13:42:46 2009 +0200 135288 135289 XKB: Fix logic error 135290 135291 Use logical or instead of bitwise or. 135292 135293 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135294 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135295 135296commit 0ec9b1069aeeaf66fd104c09f08202f7ebccd170 135297Author: Daniel Stone <daniel@fooishbar.org> 135298Date: Sat Jan 10 14:44:28 2009 +0200 135299 135300 XKB: Remove unused XkbProcessOtherEvent 135301 135302 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135303 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135304 135305commit 7af53799ccefe27b996bdc6073075b7f12353d3a 135306Author: Daniel Stone <daniel@fooishbar.org> 135307Date: Sun Jan 4 16:39:18 2009 +1100 135308 135309 Input: Remove unused CoreProcess{Keyboard,Pointer}Event 135310 135311 Everything goes through XKB's Process{Keyboard,Pointer}Event on its way 135312 through to ProcessOtherEvent now, so get rid of the old, useless functions. 135313 135314 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135315 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135316 135317commit 7c4c00649cae855e141ec9a3667bfe399e3156d0 135318Author: Daniel Stone <daniel@fooishbar.org> 135319Date: Thu Jan 1 05:25:52 2009 +1100 135320 135321 XKB: Remove unused DDX functions 135322 135323 They were complete no-ops anyway. 135324 135325 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135326 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135327 135328commit 4fa3872dc2bcfd6d1fc88d0a94c7071683eea899 135329Author: Daniel Stone <daniel@fooishbar.org> 135330Date: Sat Oct 18 20:26:00 2008 +0100 135331 135332 Input: Remove core keysyms from KeyClassRec 135333 135334 Instead of always keeping two copies of the keymap, only generate the 135335 core keymap from the XKB keymap when we really need to, and use the XKB 135336 keymap as the canonical keymap. 135337 135338 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135339 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135340 135341commit bc909f71367a02297e725bb5769c2bcadab22395 135342Author: Daniel Stone <daniel@fooishbar.org> 135343Date: Tue Aug 12 00:35:56 2008 +0300 135344 135345 Input: Centralise pointer map changing 135346 135347 Replace both core and Xi functions with one function that validates the 135348 proposed map, and sends out both kinds of notification. 135349 135350 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135351 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135352 135353commit f06a9d2e05e13466c115fc706966a90b1fb0518e 135354Author: Daniel Stone <daniel@fooishbar.org> 135355Date: Sat Oct 18 20:59:30 2008 +0100 135356 135357 Input: Clean up keymap change notifications 135358 135359 Keyboard map notifications are always generated from within XKB code, 135360 which also takes care of copying the keysyms, etc. If you need to 135361 mangle the keymap yourself, generate a new core keymap/modmap, and pass 135362 it to XkbApplyMappingChange. 135363 135364 SendMappingNotify is renamed to SendPointerMappingNotify (and ditto its 135365 Device variants), which still only _sends_ the notifications, as opposed 135366 to also doing the copying a la XkbApplyMappingChange. 135367 135368 Also have the modmap change code traverse the device hierachy, rather 135369 than just going off the core keyboard. 135370 135371 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135372 135373commit 1d1a0f67eee330a286fbdef17e967ce8ea201548 135374Author: Daniel Stone <daniel@fooishbar.org> 135375Date: Tue Oct 28 16:27:37 2008 +1100 135376 135377 Xi: Introduce XIShouldNotify 135378 135379 XIShouldNotify just lets you know if you should send an event for a 135380 keymap change (or similar) concerning a given device to a given client; 135381 at the moment, this is only for devices which are sending events to that 135382 client. 135383 135384 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135385 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135386 135387commit b5242789edb726bb539e4d42fa35bfdaca0025c9 135388Author: Daniel Stone <daniel@fooishbar.org> 135389Date: Sat Aug 9 17:02:02 2008 +0300 135390 135391 XKB: Simplify keymap writing a bit 135392 135393 We don't need no temporary variable. 135394 135395 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135396 135397commit feb757f384382c7782ceac55f99d54c7caadbd9d 135398Author: Daniel Stone <daniel@fooishbar.org> 135399Date: Fri Apr 25 19:54:29 2008 +0300 135400 135401 XKB: Sanitise vmods for redirected keys 135402 135403 Turn two unsigned chars into one unsigned int for both vmods and the 135404 vmod mask. As a bonus, remove broken unused accessor macro for setting 135405 the vmods. 135406 135407 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135408 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135409 135410commit b5f49382fe48f0a762d9a15fb10a7d7e1183fc8d 135411Author: Daniel Stone <daniel@fooishbar.org> 135412Date: Fri Apr 25 19:52:11 2008 +0300 135413 135414 XKB: Sanitise ctrls action 135415 135416 Turn four unsigned chars into one unsigned long. 135417 135418 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135419 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135420 135421commit 1bd7fd195d85681e722161f8c636a28f11b40abb 135422Author: Daniel Stone <daniel@fooishbar.org> 135423Date: Fri Apr 25 19:50:12 2008 +0300 135424 135425 XKB: Sanitise pointer actions 135426 135427 Turn two unsigned chars into one int. 135428 135429 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135430 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135431 135432commit 61c508fa78aa08ea2666fde950fbafad95d65056 135433Author: Daniel Stone <daniel@fooishbar.org> 135434Date: Fri Apr 25 19:47:16 2008 +0300 135435 135436 XKB: Sanitise vmods in actions 135437 135438 Turn vmods from two unsigned chars into one int. 135439 135440 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135441 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135442 135443commit 4f3078df2acf07364a17f738bc5c32f9a5962723 135444Author: Daniel Stone <daniel@fooishbar.org> 135445Date: Sat Aug 9 16:50:22 2008 +0300 135446 135447 XKB: Add a hell of a lot more FIXMEs 135448 135449 For some reason, we insist on having daft internal representations that 135450 make no sense, that always have to be converted to be used. We should 135451 really sort this one out. 135452 135453 Also, comment the hojillion members of XkbStateRec. 135454 135455 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135456 135457commit edeb033f29ae2d52e86c543ccbf22fb7ae1fdb7f 135458Author: Daniel Stone <daniel@fooishbar.org> 135459Date: Wed Oct 15 18:07:52 2008 +0100 135460 135461 XKB: Explicitly decode action data 135462 135463 Rather than requiring a one-to-one correspondence between XKM and struct 135464 formats in action data, explicitly fill the action data, so we can break 135465 API. 135466 135467 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135468 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135469 135470commit 699824a35791735b4b4378adf8648e3aaa599ace 135471Author: Daniel Stone <daniel@fooishbar.org> 135472Date: Fri Jul 18 16:41:29 2008 +0300 135473 135474 XKB: Remove descriptions from maprules 135475 135476 We don't use them, as they're not up to the task. We'll get a better 135477 solution someday, promise. 135478 135479 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135480 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135481 135482commit 1ad80678d88525159e55bb814205fd4eaca060a3 135483Author: Daniel Stone <daniel@fooishbar.org> 135484Date: Fri Jul 18 16:29:35 2008 +0300 135485 135486 XKB: Remove support for setting combined keymaps 135487 135488 We don't do full keymaps anymore. Deal. 135489 135490 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135491 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135492 135493commit 133e0bd6f1cc147a2ddbbe1519d10300e7dbe669 135494Author: Daniel Stone <daniel@fooishbar.org> 135495Date: Fri Jul 18 16:27:57 2008 +0300 135496 135497 XKB: Remove 'extra' functionality from rules parsing 135498 135499 When we find something weird in the rules, don't stash it as an extra 135500 freeform component, just state that the rules file is likely broken and 135501 move on with our lives. 135502 135503 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135504 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135505 135506commit cc5c6d628aa0dc87c2cc9d063972e09ad747a596 135507Author: Daniel Stone <daniel@fooishbar.org> 135508Date: Fri Apr 25 18:45:37 2008 +0300 135509 135510 XKB: Remove unsupported Xi operation flags 135511 135512 We support every XKB operation on Xi devices, so always report that we 135513 support everything, and that nothing is ever unsupported. 135514 135515 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135516 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135517 135518commit 32db27a7f867b503c2840ca7b815e96d10be9210 135519Author: Daniel Stone <daniel@fooishbar.org> 135520Date: Thu Apr 17 12:42:01 2008 -0700 135521 135522 Input: Remove modifierMap from core 135523 135524 We already have modmap (in the exact same format!) in XKB, so just use 135525 that all the time, instead of duplicating the information. 135526 135527 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135528 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135529 135530commit f062e90a95f9b7ae5458ef2100615e8ace9b66a7 135531Author: Daniel Stone <daniel@fooishbar.org> 135532Date: Wed Apr 16 19:15:30 2008 +0300 135533 135534 Input: Remove modifierKeyMap 135535 135536 Since modifierKeyMap is generated from modifierMap, just remove it, and 135537 only generate it when we need to send the modifier map to the client. 135538 135539 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135540 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135541 135542commit cf6a2fc2bd94b392cfea120444d5e032d26f1d37 135543Author: Daniel Stone <daniel@fooishbar.org> 135544Date: Sun Mar 30 17:25:39 2008 +0300 135545 135546 Input: Ignore modifiers in core input processing 135547 135548 Modifiers get cleared by the XKB code when we drop down into core input 135549 processing, so just delete the dead code path to simplify things a bit. 135550 135551 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135552 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135553 135554commit 6727ee94086d238f9f99379280d4e4827f388cce 135555Author: Daniel Stone <daniel@fooishbar.org> 135556Date: Sat Aug 9 16:45:45 2008 +0300 135557 135558 Input: Remove state from KeyClassRec 135559 135560 We already have state fully stored within XKB, so instead of duplicating it, 135561 just generate the values to send to clients when required. 135562 135563 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135564 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135565 135566commit 08363c5830bdea34012dcd954b45ccfdc79a3a7e 135567Author: Daniel Stone <daniel@fooishbar.org> 135568Date: Fri Dec 26 18:38:25 2008 +1100 135569 135570 Input: Overhaul keyboard initialisation process 135571 135572 XkbInitKeyboardDeviceStruct is now the only valid keyboard 135573 initialisation: all the details are hidden behind here. This now makes 135574 it impossible to supply a core keymap at startup. 135575 135576 If dev->key is valid, dev->key->xkbInfo->desc is also valid. 135577 135578 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135579 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135580 135581commit 40877c6680863bd6a8475f2bb0c54df55bcf0b0e 135582Author: Daniel Stone <daniel@fooishbar.org> 135583Date: Tue Mar 18 10:05:57 2008 +0200 135584 135585 XKB: Make XKB mandatory 135586 135587 No more #ifdef XKB, because you can't disable the build, and no more 135588 noXkbExtension either. 135589 135590 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135591 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135592 135593commit f9da417163b6b2d6234d2542c1f375e33db7159a 135594Author: Daniel Stone <daniel@fooishbar.org> 135595Date: Fri Mar 28 19:26:42 2008 +0200 135596 135597 Xephyr: Disable state inheritance hack 135598 135599 Prepare for the impending removal of the state field by disabling this hack 135600 for a while: it's hell of nasty and I'm amazed it ever really worked. 135601 135602 Basically, on focus out, it should do as current DDXes do and fake releases 135603 for all keys (not just mangle the core state) that are currently down; 135604 buttons too. When focus comes back in, we already have a KeymapNotify that 135605 lets us know what's currently down, so we can use this to fake the 135606 appropriate keypresses, and send it through the event routing layer. 135607 135608 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135609 135610commit 5c281446d2a8c38511d45baee6f0d25a640f12e1 135611Author: Daniel Stone <daniel@fooishbar.org> 135612Date: Fri Mar 28 11:56:34 2008 +0200 135613 135614 XKB: Remove lock actions disabling 135615 135616 For some reason, XKB allows clients to set a global (!) flag that simply 135617 turns lock keys into state no-ops. Ignore this flag. 135618 135619 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135620 135621commit 2f7cb7306709d5266688e05a066701d309323035 135622Author: Daniel Stone <daniel@fooishbar.org> 135623Date: Fri Dec 26 18:36:30 2008 +1100 135624 135625 XKB: Add XkbRMLVOSet 135626 135627 XkbRMLVOSet is just a set of strings for rules, model, layout, variant 135628 and options; use that in preference to XkbRF_VarDefsRec, which is a 135629 hideously complicated monster that somehow managed to not include the 135630 actual rules. 135631 135632 While we're at it, clean up xkbrules.h so it doesn't require xkbstr.h. 135633 135634 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135635 135636commit 3208eff2d7b856a4ac0d650498ccb36e7e62e034 135637Author: Alan Coopersmith <alan.coopersmith@sun.com> 135638Date: Wed Jan 21 14:45:06 2009 -0800 135639 135640 Update xorg-server.h.in & Xnest to use new XKB default #defines 135641 135642commit cdf216aa17b776130a91fa6792c321620e810a18 135643Author: Alan Coopersmith <alan.coopersmith@sun.com> 135644Date: Tue Jan 20 16:50:43 2009 -0800 135645 135646 Fix XKB default configure options to use the right variables 135647 135648commit 0d3c3dad21857ef502fe0d0235620a1228248e13 135649Author: Adam Jackson <ajax@redhat.com> 135650Date: Tue Jan 20 23:46:48 2009 -0500 135651 135652 Don't include fontmod.h 135653 135654commit 4f004ab9a83e209943760e1faa39f06c8eeb27af 135655Author: Benjamin Close <Benjamin.Close@clearchain.com> 135656Date: Thu Jan 22 01:36:31 2009 +1030 135657 135658 XNest: Fix the build caused by __XKBDEFRULES__ -> XKB_DFLT_RULES in 23862ede59a9ce11a06ec5151bde460fb836c603 135659 135660commit e45a8a498d1ee8e40fc334805c38c05cb04b49d0 135661Author: Adam Jackson <ajax@redhat.com> 135662Date: Tue Jan 20 22:07:07 2009 -0500 135663 135664 Remove CreateUnclippedWinSize from window.h 135665 135666 This is utterly wrong, but then, so is sdksyms.sh 135667 135668commit 20d2117eb82fb7ce91afd4f01fe1b9d4b1bf0459 135669Author: Adam Jackson <ajax@redhat.com> 135670Date: Tue Jan 20 00:17:42 2009 -0500 135671 135672 Move CreateUnclippedWinSize to mbuf 135673 135674 It's the only user, so. 135675 135676commit e1611d8d55ade35bf8c2e1da7bc2dbf05b011301 135677Author: Daniel Stone <daniel@fooishbar.org> 135678Date: Thu Nov 27 13:34:05 2008 +1100 135679 135680 XKB: Be more verbose about XkbCopyKeymap failure in debug mode 135681 135682 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135683 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135684 135685commit 5ee504cae5f17d16ee7599beaa4b1944a9a8aee6 135686Author: Daniel Stone <daniel@fooishbar.org> 135687Date: Thu Nov 27 13:31:18 2008 +1100 135688 135689 XKB: Trying to copy to the same keymap is not fatal 135690 135691 Humour the user if they run XkbCopyKeymap(foo, foo). 135692 135693 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135694 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135695 135696commit d929ba6886c6f121b101c0bf0f7c63ef9d851a3e 135697Author: Daniel Stone <daniel@fooishbar.org> 135698Date: Fri Dec 26 18:32:41 2008 +1100 135699 135700 XFree86: Input: Fail NIDR if we can't enable devices 135701 135702 If we can't enable a device, bail out of NewInputDeviceRequest rather than 135703 blithely continuing. Also, be more verbose when initialization failed. Also, 135704 be more verbose when initialization failed. Also, be more verbose when 135705 initialization failed. Also, be more verbose when initialization failed. 135706 135707 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135708 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135709 135710commit 534669b376a6a703fbc97269d279a5418cf60c98 135711Author: Daniel Stone <daniel@fooishbar.org> 135712Date: Sun Jan 4 16:24:44 2009 +1100 135713 135714 XKB: Remove unnecessary prototype 135715 135716 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135717 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135718 135719commit 27ea1a7e4e7eca394e052791c64bc6be515e075e 135720Author: Daniel Stone <daniel@fooishbar.org> 135721Date: Sun Jan 4 16:40:46 2009 +1100 135722 135723 XKB: Only Xi events are processed 135724 135725 Core events aren't run through these functions, so don't bother testing 135726 for them. 135727 135728 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135729 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135730 135731commit 2762cafc321166ac48f80146720709b1011e6894 135732Author: Daniel Stone <daniel@fooishbar.org> 135733Date: Thu Nov 27 13:35:42 2008 +1100 135734 135735 XkbCopyKeymap: inputInfo.keyboard is not a special case 135736 135737 The device-walking code is still depressing, though. 135738 135739 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135740 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135741 135742commit 07c3bb922b8c1ae8fe28198f6c451430b836f21e 135743Author: Daniel Stone <daniel@fooishbar.org> 135744Date: Sun Jan 4 14:12:19 2009 +1100 135745 135746 XKB: Move XkbCopyKeymap definition to xkbsrv.h 135747 135748 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135749 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135750 135751commit 4f35a34f3cd3df3c8af129936d1808dcc063d8a7 135752Author: Peter Hutterer <peter.hutterer@redhat.com> 135753Date: Thu Aug 14 17:26:02 2008 +0930 135754 135755 dix: Remove traces of CoreDevicePrivateKey 135756 135757 Obsolete with MPX. 135758 135759 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 135760 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135761 135762commit 58a6b5b9eff25b1b9c6de239d6f5a952a31dfd15 135763Author: Peter Hutterer <peter.hutterer@redhat.com> 135764Date: Thu Aug 14 17:21:28 2008 +0930 135765 135766 Xi: don't care about CoreDevicePrivateKey when copying keys 135767 135768 If we get here, we must copy. 135769 135770 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 135771 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135772 135773commit 23862ede59a9ce11a06ec5151bde460fb836c603 135774Author: Daniel Stone <daniel@fooishbar.org> 135775Date: Sat Jan 10 14:36:16 2009 +0200 135776 135777 XKB: Allow build-time configuration of XKB defaults 135778 135779 Instead of hardcoding base/pc105/us, allow users to change the defaults at 135780 ./configure time. Change the default model to be evdev on Linux. 135781 135782 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 135783 135784commit 57a5eb2583bdf1069861fe2184de9739c0d0c7c1 135785Author: Colin Harrison <colin.harrison@virgin.net> 135786Date: Sun Jan 4 17:40:54 2009 +0000 135787 135788 Xming: Update .rc file to tidy About and Exit dialogs 135789 135790 Tidy up About and Exit dialogs, add keyboard accelerators 135791 135792 Copyright (C) Colin Harrison 2005-2008 135793 http://www.straightrunning.com/XmingNotes/ 135794 http://sourceforge.net/projects/xming/ 135795 135796 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 135797 135798commit df3f701ef2d41ff4be4839f473ae71d33777bea3 135799Author: Colin Harrison <colin.harrison@virgin.net> 135800Date: Sun Nov 2 20:17:23 2008 +0000 135801 135802 Xming: Ensure we dont try to print a null value when displaying xkb configuration 135803 135804 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 135805 135806commit c612f0df45d388c98108e1e0c80afdfae05e338e 135807Author: Colin Harrison <colin.harrison@virgin.net> 135808Date: Sun Nov 2 20:46:41 2008 +0000 135809 135810 Xming: Add styles keyword and attributes. 135811 135812 Add handling for style keyword and atttributes in .XWinrc 135813 Update man page to document these additions 135814 135815 Copyright (C) Colin Harrison 2005-2008 135816 http://www.straightrunning.com/XmingNotes/ 135817 http://sourceforge.net/projects/xming/ 135818 135819 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 135820 135821commit 847a91ad2e8c9b0992c09f0ec38466885d6d25d4 135822Author: Colin Harrison <colin.harrison@virgin.net> 135823Date: Sun Jan 4 19:57:56 2009 +0000 135824 135825 Xming: Notify X when the keyboard focus is lost to a pure Windows window in -multiwindow mode. 135826 135827 Fix internal WM to notify X when the keyboard focus is lost to a pure Windows window in -multiwindow mode. 135828 135829 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 135830 135831commit 6c3e7e4d0568ecb7d10b814593bbbd817b7d2aab 135832Author: Colin Harrison <colin.harrison@virgin.net> 135833Date: Sun Jan 4 18:28:34 2009 +0000 135834 135835 Xming: Only allow WM_MOUSEWHEEL messages to act on the client area of a focused window. 135836 135837 Fix internal WM so it only allows WM_MOUSEWHEEL messages to act on the client area of a focused window. 135838 135839 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 135840 135841commit d84e0e387764236c7e97509baf4606d96cfc19d4 135842Author: Colin Harrison <colin.harrison@virgin.net> 135843Date: Sun Jan 4 18:20:44 2009 +0000 135844 135845 Xming: Correctly parent XA_WM_TRANSIENT_FOR windows in -multiwindow mode when a windows window is created. 135846 135847 Fix internal WM to correctly parent XA_WM_TRANSIENT_FOR windows in -multiwindow mode when a windows window is created, 135848 and to de-iconize parent windows when a child window acquires focus. 135849 135850 XXX: Perhaps we should also shuffle parent(s) forward through Z-order when a child acquires focus? 135851 135852 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 135853 135854commit 8d82d806ee38d45ec7cdbf8b1f671ec1e4437567 135855Author: Colin Harrison <colin.harrison@virgin.net> 135856Date: Sun Nov 2 20:46:38 2008 +0000 135857 135858 Xming: Correct the way display and screen number is reported in window titles and tooltips 135859 135860 Correct the tooltip text for the toolbar X icon to be strictly correct, 'display-number:screen' should be ':display-number.screen'. 135861 Also for the default window title. 135862 Adjust the style of the Windows title in XDMCP mode from 'Xming - hostname' to 'hostname:display-number.screen'. 135863 135864 Copyright (C) Colin Harrison 2005-2008 135865 http://www.straightrunning.com/XmingNotes/ 135866 http://sourceforge.net/projects/xming/ 135867 135868 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 135869 135870commit 656f92e247f2f257105daf1e0e6998e1b3d10eba 135871Author: Jon TURNEY <jon.turney@dronecode.org.uk> 135872Date: Sun Nov 2 20:17:02 2008 +0000 135873 135874 Cygwin/X: should also use GetTickCount(), just like Xming 135875 135876 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 135877 135878commit 7f781e780ee2b261c5cb688a8e6c5dc1a53bdb6f 135879Author: Yaakov Selkowitz <yselkowitz@users.sourceforge.net> 135880Date: Wed Dec 17 19:13:30 2008 +0000 135881 135882 Cygwin/X: Cygwin doesn't have RTLD_LOCAL 135883 135884 RTLD_LOCAL is not defined on Cygwin 135885 135886 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 135887 135888commit 5630ff809676ab6b4833757060e48e201e8768a5 135889Author: Dan Nicholson <dbn.lists@gmail.com> 135890Date: Mon Dec 22 13:53:43 2008 -0800 135891 135892 Ignore generated dmx manpages 135893 135894 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 135895 135896commit 64f924fbf513b5f91b3f2bcf5767f6fe927d3f7a 135897Author: Dan Nicholson <dbn.lists@gmail.com> 135898Date: Mon Dec 22 13:43:30 2008 -0800 135899 135900 Kill off --with-mesa-source completely 135901 135902 There were a few spots left in the source that were using the 135903 --with-mesa-source defined headers or the now removed $(top_srcdir)/GL 135904 directory. These aren't needed anymore as all the necessary source for 135905 GLX is in $(top_srcdir)/glx. 135906 135907 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 135908 135909commit e1c8ee215797fa8c6a3a5421bb6f99b58064248b 135910Author: Dan Nicholson <dbn.lists@gmail.com> 135911Date: Sun Dec 14 14:18:58 2008 -0800 135912 135913 Don't reuse PKG_CHECK_MODULES identifiers 135914 135915 Using GL for the PKG_CHECK_MODULES identifier multiple times means only 135916 the first call will actually be used. Later calls will be skipped due to 135917 GL_CFLAGS and GL_LIBS already being set. This changes DRI to using a 135918 different identifier and DMX to just reusing GL_CFLAGS. 135919 135920 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 135921 135922commit ca46c01869768a8661a9d8a71493ed5f0760a8ab 135923Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 135924Date: Fri Jan 16 13:58:20 2009 -0800 135925 135926 XQuartz: mieq: Wait for the server to finish initializing before letting other threads mieqEnqueue 135927 135928 Avoid possible race condition whereby one thread might call mieqEnqueue before InitAndStartDevices finishes 135929 (cherry picked from commit 94e417ac87a98cd5c6bf2d7c495d702748398931) 135930 135931commit b33905234025f005819c7e2acd653a3a0ecfeb82 135932Author: Peter Hutterer <peter.hutterer@who-t.net> 135933Date: Fri Jan 9 14:32:14 2009 +1000 135934 135935 xfree86: always force RAW mode under linux. 135936 135937 The previous check for AEI on left us with the possibility that AEI is forced 135938 off in the config, but devices are added through evdev nonetheless. A keyboard 135939 added this way can CTRL+C the server. Even when we use kbd, we can set the 135940 mode to RAW, so it's safer alround to to so. 135941 135942 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 135943 135944commit 646add4eb47c01c045ef64405510a08ecfb5bfd6 135945Author: Peter Hutterer <peter.hutterer@who-t.net> 135946Date: Fri Jan 16 16:04:07 2009 +1000 135947 135948 dix: remove XineramaCheckMotion, merge into CheckMotion. 135949 135950 For 5 lines difference, we don't need to duplicate the whole function. really. 135951 135952commit e76f904abe530df9ddaa09510088c545298b0b91 135953Author: Peter Hutterer <peter.hutterer@who-t.net> 135954Date: Fri Jan 16 15:53:50 2009 +1000 135955 135956 dix: remove "notyet" conditional compilation. 135957 135958 If it's "notyet" for 3 years, then "yet" won't arive anytime soon. 135959 135960commit 5c41b3f9c8b72f0115019237f812d74610b82c24 135961Author: George Staplin <gstaplin@apple.com> 135962Date: Mon Jan 12 19:30:43 2009 -0700 135963 135964 XQuartz: xpr: Cleanup some of the code and possibly fix part of the GLX Pixmap problem. 135965 135966 Split DRICreateSurface into 3 functions. 135967 135968 Make CreateSurfaceForPixmap use xp_configure_surface. I suspect this is partly why 135969 GLXPixmaps never worked. 135970 135971 It will require some more work and thoughts for pbuffers, unless we fake those with 135972 pixmaps and surfaces. 135973 (cherry picked from commit 9cf264e67744262b9f45079e6cd752eb3e3b0e08) 135974 135975commit c5d52d4c37ebec92cb668b0653bf3de2bdfc5782 135976Author: George Staplin <gstaplin@apple.com> 135977Date: Mon Jan 12 14:14:44 2009 -0700 135978 135979 XQuartz: GL: Make indirect.c build and work in the 1.6 branch. 135980 135981 Reorder some header files and provide some types earlier on. 135982 135983 Remove the static __GLXextensionInfo __glDDXExtensionInfo; that isn't 135984 used in 1.4 or 1.5, and seems to have been removed from 1.6. Remove 135985 the data structures associated with that too. 135986 135987 Remove __glXAquaDrawableResize. The GLX structure doesn't use 135988 it anymore, and the Apple version did nothing useful before. 135989 135990 __glXAquaDrawableSwapBuffers(): base no longer contains a drawGlxc 135991 member. Now provide the the Apple/Aqua context in the __GLXAquaDrawable 135992 struct. Add the context member to the __GLXAquaDrawable struct. 135993 135994 Remove the fallback configs. They aren't used in 1.4 or 1.5 either. 135995 135996 Remove init_visuals(). It's not used in 1.4 or 1.5 either. 135997 135998 In the drawable constructor initialize the state properly, including 135999 the new context member. 136000 136001 Remove glAquaResetExtension() -- it's not used anymore. 136002 136003 This has been tested remotely and proven to work with glxgears, 136004 fire, various texture programs I wrote, and various Mesa demos. 136005 (cherry picked from commit d514152195452ae11ec7769c76489651620ad380) 136006 136007commit 1c504dff8a84804888f18ee9b184f0c4584185d7 136008Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136009Date: Thu Jan 15 21:03:27 2009 -0800 136010 136011 XQuartz: Fix builddir != srcdir issues and undef _XSERVER64 where appropriate on fat binary compilation 136012 136013commit 717a961528ec69a6e630d536e15568670e0b398a 136014Author: Thomas Jaeger <ThJaeger@gmail.com> 136015Date: Sat Dec 20 16:17:02 2008 +0100 136016 136017 Don't release grabs unless all buttons are up 136018 136019 Previously, only buttons <= 5 would count here, but the core protocol 136020 allows for 255 buttons. 136021 136022 http://lists.freedesktop.org/archives/xorg/2009-January/042092.html 136023 136024 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136025 136026commit f7f85f696570541e2dd43462675de9e6ee46f545 136027Author: Thomas Jaeger <ThJaeger@gmail.com> 136028Date: Mon Dec 22 00:55:09 2008 +0100 136029 136030 Count the number of logically down buttons in buttonsDown 136031 136032 This fixes the following bug. Assuming your window manager grabs 136033 Alt+Button1 to move windows, map Button3 to 0 via XSetPointerMapping, 136034 then press the physical button 3 (this shouldn't have any effect), press 136035 Alt and then button 1. The press event is delivered to the application 136036 instead of firing the grab. 136037 136038 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136039 136040commit 6f065a26db62aaf1d7aab9f9a3d0e6cd09d1c1dd 136041Author: Colin Harrison <colin.harrison@virgin.net> 136042Date: Thu Jan 15 21:42:26 2009 +0000 136043 136044 Xming: Reduce MOUSE_POLLING_INTERVAL 136045 136046 Reduce MOUSE_POLLING_INTERVAL from 500ms to 50ms 136047 136048 In -multiwindow mode, this determines how quickly X windows can track the mouse 136049 when it is outside any X window... Move the cursor around an xeyes, and you will 136050 see what this timing effects. 136051 136052commit c745db1674c3cb55249c9eb6e74939b74c42409c 136053Author: Jon TURNEY <jon.turney@dronecode.org.uk> 136054Date: Mon Jan 5 13:52:45 2009 +0000 136055 136056 GLX: Avoid a crash when we have an uninitialized GL context 136057 136058 If the GL dispatch table pointer points to glapi_noop_table, 136059 (due to some kind of GL initialization failure), DoGetString(GL_VERSION) 136060 (for example as invoked by glxinfo) will crash as it tries to 136061 do atof() on the null pointer returned by the noop dispatch function 136062 136063 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 136064 136065commit e30032d0bbbf0bf71e1b53a4ac388f3bd6f68e53 136066Author: Peter Hutterer <peter.hutterer@redhat.com> 136067Date: Mon Dec 1 21:14:01 2008 +1000 136068 136069 Xext: clean up XGE macros. 136070 136071 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 136072 136073commit 1ae529bef50a8ccc4a8a264ed120aeae544b21e7 136074Author: Peter Hutterer <peter.hutterer@who-t.net> 136075Date: Wed Jan 14 14:56:10 2009 +1000 136076 136077 dix: replace manual check for event types with IsPointerEvent() 136078 136079commit 1d54479cb3c8b4f75b7564f8b5e1c5da940b20f4 136080Author: Simon Thum <simon.thum@gmx.de> 136081Date: Tue Nov 25 14:12:02 2008 +1000 136082 136083 xfree86: init properties when ptraccel configuration is found. 136084 136085 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 136086 Signed-off-by: Simon Thum <simon.thum@gmx.de> 136087 136088commit 19275ea8e9dd93d5e61906943706dfe49003dd9c 136089Author: Simon Thum <simon.thum@gmx.de> 136090Date: Wed Nov 19 16:01:21 2008 +1000 136091 136092 dix: add property support for pointer acceleration. 136093 136094 Note: properties don't need to be cleaned up, the DIX does it for us anyway. 136095 Data that is stored in properties is cleaned up by the property system. 136096 Handlers, etc. don't need to be unregistered while cleaning up, as they get 136097 deleted when the device is removed anyway. 136098 136099 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 136100 Signed-off-by: Simon Thum <simon.thum@gmx.de> 136101 136102commit d645721170b1196e5064b397cfbffd1da8c79bb1 136103Author: Peter Hutterer <peter.hutterer@who-t.net> 136104Date: Thu Jan 8 11:53:30 2009 +1000 136105 136106 mi: ensure chained button mappings from SD -> MD (#19282) 136107 136108 After copying the master event, flip the detail field to the mapped button of 136109 the SD, not the physical button. This way if the SD has a mapping 1:3 and the 136110 MD has a mapping of 3:4, a press on button 1 on the SD results in a core event 136111 on button 4. 136112 136113 X.Org Bug 19282 <http://bugs.freedesktop.org/show_bug.cgi?id=19282> 136114 136115 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136116 136117commit 639f289dcdbe00a516820f573c01a8339e120ed4 136118Author: Michel Dänzer <daenzer@vmware.com> 136119Date: Tue Jan 13 13:05:32 2009 +0100 136120 136121 EXA: Declare glyph cache picture as component-alpha when necessary. 136122 136123 Without this, rendering component-alpha glyphs may break without a mask. 136124 136125 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=19233 . 136126 136127commit f4e00b783d3f824058a5f140e6462142688f5116 136128Author: Michel Dänzer <daenzer@vmware.com> 136129Date: Tue Jan 13 11:03:27 2009 +0100 136130 136131 Fix build with --enable-visibility --disable-xdmcp. 136132 136133commit d36adf52a2b2711d22b11105f7bd907d4493fb9b 136134Author: Peter Hutterer <peter.hutterer@who-t.net> 136135Date: Mon Jan 12 15:29:36 2009 +1000 136136 136137 dix: fix WarpPointer calls for devices with custom valuator ranges (#19297) 136138 136139 If the MD's lastSlave was a devices with custom axes ranges, then a 136140 WarpPointer would position the cursor at the wrong location. A WarpPointer 136141 request provides screen coordinates and these coordinates were scaled to the 136142 device range before warping. 136143 136144 This patch consists of two parts: 136145 1) in the WarpPointer handling, get the lastSlave and post the event through 136146 this device. 136147 2) assume that WarpPointer coordinates are always in screen coordinates and 136148 scale them to device coordinates in GPE before continuing. Note that this 136149 breaks device-coordinate based XWarpDevicePointer calls (for which the spec 136150 isn't nailed down yet anyway) until a better solution is found. 136151 136152 X.Org Bug 19297 <http://bugs.freedesktop.org/show_bug.cgi?id=19297> 136153 136154 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136155 136156commit 488d45295105daf10ccd17ca93ae6a6f4d0104f1 136157Author: Peter Hutterer <peter.hutterer@who-t.net> 136158Date: Mon Jan 12 16:16:24 2009 +1000 136159 136160 dix: EnqueueEvent and PlayReleasedEvent need to handle DeviceMotionNotifies 136161 136162 No MotionNotify events in the processing anymore, so let's have them treat DMN 136163 instead. 136164 136165 Reported by Thomas Jaeger. 136166 136167commit 1e470645983ad18fa38c0885d674bc9d1dc1e560 136168Author: Julien Cristau <jcristau@debian.org> 136169Date: Tue Jan 13 00:12:56 2009 +0100 136170 136171 require randrproto 1.2.99.3 136172 136173commit a48c81dcdf569a3f634ac23e08d2491354de6a36 136174Author: Peter Hutterer <peter.hutterer@who-t.net> 136175Date: Fri Dec 5 16:24:57 2008 +1000 136176 136177 Xi: add XATOM_FLOAT to server-defined properties. 136178 136179 This property is used to denote type float for input properties. Such 136180 properties can be accessed easily through the XIPropToFloat() function. 136181 136182 Code originally written by Simon Thum. 136183 136184 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 136185 136186commit 669f6810af9a89187d6149841925fe765f3988ff 136187Author: Peter Hutterer <peter.hutterer@redhat.com> 136188Date: Wed Nov 19 15:50:57 2008 +1000 136189 136190 Xi: add XIPropToInt() auxiliary function. 136191 136192 Converts an XIPropertyValuePtr to an integer, provided that type and format is 136193 right. 136194 136195 Code originally written by Simon Thum. 136196 136197 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 136198 136199commit 56efbc0986e782da45addb05ece9f456d41d7a90 136200Author: Peter Hutterer <peter.hutterer@who-t.net> 136201Date: Fri Jan 9 13:46:20 2009 +1000 136202 136203 dix: drop x/y back into last.valuators before updating the history (#19285) 136204 136205 positionSprite needs to scale to screen coordinates and in the process of 136206 doing so alters dev->last.valuators[0:1]. Drop the real coordinates back after 136207 finishing and before updating the motion history. This way, we don't push the 136208 screen coordinates into the motion history. 136209 136210 X.Org Bug 19285 <http://bugs.freedesktop.org/show_bug.cgi?id=19285> 136211 136212commit 67710ad1d172cc4d53e28e5cc7dab2efb57a9a85 136213Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136214Date: Sun Jan 11 16:34:36 2009 -0800 136215 136216 XQuartz: Don't need explicit Activate/EnableDevice in InitInput 136217 (cherry picked from commit b1d29784410b3b93037e5636f336ba608d8ad6e3) 136218 136219commit dde5c591be74139ec6f2a90a3fdc7904d7342554 136220Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136221Date: Sun Jan 11 16:10:45 2009 -0800 136222 136223 mieq: Avoid possible race condition whereby one thread might call mieqEnqueue before InitAndStartDevices finishes 136224 136225 This is more of a hack around the problem. This is something that will need to be addressed in a 136226 more structured manner with the multi threaded input efforts. 136227 (cherry picked from commit 3c596c061e75848cfa76dd9259c23a3f3a67444c) 136228 136229commit 0720816aea20971ee94b5df3faa9cb9f16e93f82 136230Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136231Date: Sun Jan 11 14:49:39 2009 -0800 136232 136233 XQuartz: Only call DarwinUpdateModKeys when needed 136234 136235 Previously, we were calling it on almost every itteration through sendX11Event 136236 (cherry picked from commit 6461729647ff4441d80811e73f0c0d2f108f2700) 136237 136238commit c8bf582db8111f71b9250a602a9d5aa7bd828136 136239Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136240Date: Thu Oct 23 11:41:22 2008 -0700 136241 136242 XQuartz: Comment explaining the 0x10 flag for [e data2] 136243 (cherry picked from commit 3c695280641c1205b97d3bb9f1d5e15a19cfa45f) 136244 136245commit 5339c22a8ca8bce6cd2b51e97c904292e7c40e74 136246Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136247Date: Sun Jan 11 01:56:45 2009 -0800 136248 136249 XQuartz: Add locking to make mieq thread safe on OSX 136250 (cherry picked from commit 7a8d2266861e74176b5310b83652a9c10a170494) 136251 136252commit 1beff89810cd193e45d437846929b3938cee6d96 136253Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136254Date: Sun Jan 11 01:55:26 2009 -0800 136255 136256 XQuartz: misc 1.6 updates (still --disable-glx) 136257 rlAccel is not longer compatable, and it's not worth fixing 136258 Don't override DeviceCursorInitialize with a noop 136259 Don't do a SwitchCoreKeyboard (which wasn't even needed in the first place) 136260 (cherry picked from commit c137f681680e1d04b1513a8be68aeda4d1c56fd5) 136261 136262commit 2d917cfe56af1e74205f80ddbaddfb5c6ce82f0a 136263Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136264Date: Sat Jan 10 01:39:08 2009 -0800 136265 136266 Apple: Don't use DRI2 136267 (cherry picked from commit a1d35cee5907a76977ee43c49cb55c8f411c9794) 136268 136269commit 7a8abb9c8554d7967d912dcd427804f65906dc5c 136270Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136271Date: Tue Jan 6 15:32:44 2009 -0800 136272 136273 XQuartz: Bundle version 2.4.0 136274 (cherry picked from commit 3a3ccf5354e4275abe9d01f41a92602df5d690dc) 136275 136276commit 63ff7d236f67e159943312e1a551fcf5bb3f463a 136277Author: Julien Cristau <jcristau@debian.org> 136278Date: Sun Jan 11 08:45:24 2009 +0100 136279 136280 os: don't mix declarations and code 136281 136282commit f43a3c61ab1eab8aecdc0e4bdee9f3bae2cab234 136283Author: Julien Cristau <jcristau@debian.org> 136284Date: Sun Jan 11 08:44:53 2009 +0100 136285 136286 os: backtrace() returns int, not size_t 136287 136288commit 5321ae76cffc5fcfe4bebcfb10e08eecaf640503 136289Author: Julien Cristau <jcristau@debian.org> 136290Date: Sun Jan 11 08:44:29 2009 +0100 136291 136292 os: ANSI cleanups 136293 136294commit 47438a2161a92fec08081d2de28966cf8d31654a 136295Author: Julien Cristau <jcristau@debian.org> 136296Date: Sun Jan 11 08:43:48 2009 +0100 136297 136298 randr: RRSetPrimaryOutput can be static 136299 136300commit 7f82114b1dc738f98b9f22f6e61d6d5191383b67 136301Author: Julien Cristau <jcristau@debian.org> 136302Date: Sun Jan 11 08:43:35 2009 +0100 136303 136304 xkb: ANSI cleanup 136305 136306commit 46f3c1b937002f9a860878b359dcf5756a82dcf2 136307Author: Julien Cristau <jcristau@debian.org> 136308Date: Sun Jan 11 08:42:57 2009 +0100 136309 136310 mi: don't mix declarations and code 136311 136312commit 5aa9483d3ee306a79a9235610ce6b2754d697585 136313Author: Julien Cristau <jcristau@debian.org> 136314Date: Sun Jan 11 08:42:28 2009 +0100 136315 136316 mi: ANSI cleanups 136317 136318commit 359ed32c9b013ed88e14c99aa7fd7a617a85e2c5 136319Author: Julien Cristau <jcristau@debian.org> 136320Date: Sun Jan 11 08:40:58 2009 +0100 136321 136322 xnest: only define ddxBeforeReset if needed 136323 136324commit 9651cd8d7ac26635b6fb1826806ac9f4e2bddb63 136325Author: Julien Cristau <jcristau@debian.org> 136326Date: Sun Jan 11 08:40:47 2009 +0100 136327 136328 xnest: ANSI cleanups 136329 136330commit 31439195c51f183f5a8989a8d4f95b3e6326eda0 136331Author: Julien Cristau <jcristau@debian.org> 136332Date: Sun Jan 11 08:37:06 2009 +0100 136333 136334 xfree86: linuxPci.c needs a prototype for xf86AccResFromOS 136335 136336commit a00dbd0015845e08ef2af9480cc3833819cf86f5 136337Author: Julien Cristau <jcristau@debian.org> 136338Date: Sun Jan 11 08:36:16 2009 +0100 136339 136340 xfree86: use %zu to format size_t arguments 136341 136342commit 17c8cb50990a9ad20bbd3f8ce14c149dac9d72ad 136343Author: Julien Cristau <jcristau@debian.org> 136344Date: Sun Jan 11 08:34:56 2009 +0100 136345 136346 xfree86: don't mix declarations and code 136347 136348commit aec4c0caca41a06832536ba82ba5bcdccf4a0aea 136349Author: Julien Cristau <jcristau@debian.org> 136350Date: Sun Jan 11 08:33:18 2009 +0100 136351 136352 xfree86: ANSI cleanups 136353 136354commit dba3b1944d04232f88bb310277639f2c1adcbe08 136355Author: Julien Cristau <jcristau@debian.org> 136356Date: Sun Jan 11 08:28:23 2009 +0100 136357 136358 Xvfb: ANSI cleanups 136359 136360commit a2a760e33c17e211650b1f53c05a89c013916cf3 136361Author: Julien Cristau <jcristau@debian.org> 136362Date: Sun Jan 11 08:25:39 2009 +0100 136363 136364 dix: don't mix declarations and code 136365 136366commit 466dddbb83aa37ae93fb32976d51ecaa947c3616 136367Author: Julien Cristau <jcristau@debian.org> 136368Date: Sun Jan 11 08:25:17 2009 +0100 136369 136370 dix: fix cast from pointer to integer 136371 136372commit 54332e4732577dc49e440d9c9f8dd7d791f03f8e 136373Author: Julien Cristau <jcristau@debian.org> 136374Date: Sun Jan 11 08:20:06 2009 +0100 136375 136376 dix: ANSI cleanups 136377 136378commit ef9902652b8a1c716ea3851565d6e2a98bd30eb0 136379Author: Julien Cristau <jcristau@debian.org> 136380Date: Sun Jan 11 08:18:56 2009 +0100 136381 136382 config: ANSI cleanups 136383 136384commit e64473b14e091e4b75e6ac2bf9a35e9aaa907258 136385Author: Julien Cristau <jcristau@debian.org> 136386Date: Sun Jan 11 08:17:41 2009 +0100 136387 136388 Xi: fix missing declaration of XkbSetRulesDflts 136389 136390commit 6e20a4ae731b7bc72266e5809faa0ecfde2bdfe4 136391Author: Julien Cristau <jcristau@debian.org> 136392Date: Sun Jan 11 08:16:55 2009 +0100 136393 136394 xsync: make SyncAlarmCounterDestroyed static 136395 136396commit 292311b421e53ead53c7df67d735a3d6d3e5b515 136397Author: Julien Cristau <jcristau@debian.org> 136398Date: Sun Jan 11 08:16:12 2009 +0100 136399 136400 Xext: ANSI cleanups 136401 136402commit 7a05c8b1e70680ddd3b3e09ad448788f8d70a428 136403Author: Julien Cristau <jcristau@debian.org> 136404Date: Tue Dec 23 13:08:49 2008 +0100 136405 136406 xfree86/linux: fix log flood on acpid open error 136407 136408 When we can't open the acpid socket, warn once, not once every second 136409 136410commit 0d8184de591c75794083c8a8ca3fb8459e9394bc 136411Author: George Staplin <gstaplin@apple.com> 136412Date: Fri Jan 9 16:08:26 2009 -0700 136413 136414 XQuartz: GL: Set the __GLXconfig renderType to GLX_RGBA_BIT. 136415 (cherry picked from commit 3c14546f58f8a138fe67c9cacc3bd0b7fa90c29a) 136416 136417commit 9c3975d63e129ad7230cae585f08a962d45fb861 136418Author: George Staplin <gstaplin@apple.com> 136419Date: Fri Jan 9 15:53:39 2009 -0700 136420 136421 XQuartz: GL: Make many more fbconfigs and visuals available for the 1.5 branch. 136422 136423 Use a __GLXconfig linked list struct to store the configurations for the 136424 fbconfigs and visuals in a pGlxScreen. 136425 136426 Also, remove the __GLXvisualConfig/GlxSetVisualConfigs code that isn't used 136427 anymore. There is more code we can remove later, but I want to do that in 136428 separate commits. 136429 (cherry picked from commit 94162b0f8a25267aca280d25e216cc6bde47da6e) 136430 136431commit a68373994f4957a73f4eab52c90ebcfb96f285f8 136432Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136433Date: Wed Jan 7 18:27:21 2009 -0800 136434 136435 XQuartz: Update padding in appledristr. This shouldn't change how the struct ends up in memory, but should make it more obvious for human eyes 136436 (cherry picked from commit f7b6c1907c326a1befd8b5c8a1d311d551e8670f) 136437 136438commit 5f3188228eb988bd8f08b62c84f98a8ff66ee283 136439Author: Xiang, Haihao <haihao.xiang@intel.com> 136440Date: Tue Jan 6 10:03:19 2009 +0800 136441 136442 avoid a potential endless loop. 136443 136444 Previously it is possible that creating rotation data, then cleaning 136445 up and creating again so that pScreen->BlockHandler and 136446 xf86_config->BlockHandler all point to xf86RotateBlockHandler. 136447 See bug #19343. 136448 136449commit b2756a71a432f7cf7c870a48676c98625512558d 136450Author: Thomas Jaeger <ThJaeger@gmail.com> 136451Date: Fri Jan 9 02:02:38 2009 -0500 136452 136453 Xext: Send out correct events in ProcXTestFakeInput 136454 136455 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136456 136457commit 9bf77569cc8f9b0cc25c99487672d375d7c32f99 136458Author: Peter Hutterer <peter.hutterer@who-t.net> 136459Date: Thu Jan 8 16:32:24 2009 +1000 136460 136461 Xi: call CheckMotion for floating SDs too. 136462 136463 CheckMotion updates the sprite position so we need to call it for all devices 136464 that have a sprite - including floating SDs (which have an invisible sprite). 136465 136466commit 83eb863e3cfcd62549b836f3195107d8908b36d0 136467Author: Peter Hutterer <peter.hutterer@who-t.net> 136468Date: Fri Jan 9 11:39:02 2009 +1000 136469 136470 Xext: don't accept DeviceValuator if the dev doesn't have valuators (in xtest) 136471 136472 Reported by Chris Ball. 136473 136474commit a3f9e887d9ff5fdee6c35ccc0f7b324723a83a46 136475Author: Peter Hutterer <peter.hutterer@who-t.net> 136476Date: Fri Jan 9 11:40:18 2009 +1000 136477 136478 dix: don't accept Button 0 presses in GPE. 136479 136480commit 0c6987df3b9b3a37d201d740d8248c326449835e 136481Author: Adam Jackson <ajax@redhat.com> 136482Date: Tue Jun 17 16:10:51 2008 -0400 136483 136484 XAA: Disable offscreen pixmaps by default. 136485 136486 Say Option "XaaOffscreenPixmaps" to turn them back on. 136487 136488 Apropos of bugs #13795 and #15098. But this still isn't correct. 136489 136490commit eb2d7b3d700952ba88c77deacf687b251300e660 136491Author: Peter Hutterer <peter.hutterer@who-t.net> 136492Date: Fri Dec 19 21:05:47 2008 +1000 136493 136494 dix: move focus handling into enterleave.c. 136495 136496 This commit moves the focus handling from events.c into enterleave.c and 136497 implements a model similar to the core enter/leave model. 136498 For a full description of the model, see: 136499 http://lists.freedesktop.org/archives/xorg/2008-December/041740.html 136500 136501 This commit also gets rid of the focusinout array in the WindowRec, ditching 136502 it in favour of a local array that keeps the current focus window for each 136503 device. 136504 136505 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136506 136507commit 673eb23aac578dcdc04e2a99d1fa5c2987eb58b8 136508Author: Peter Hutterer <peter.hutterer@who-t.net> 136509Date: Fri Dec 19 21:05:09 2008 +1000 136510 136511 dix: add a few auxiliary functions for the updated focus model. 136512 136513 SetFocusIn and SetFocusOut, including the static array to keep all focus 136514 windows. 136515 136516 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136517 136518commit 38b28dcadd0990cb43f50db4300eebb8f044db96 136519Author: Peter Hutterer <peter.hutterer@who-t.net> 136520Date: Thu Dec 18 15:55:11 2008 +1000 136521 136522 dix: reduce FirstPointerChild complexity 136523 136524 Instead of keeping a flag on each window for the devices that are in this 136525 window, keep a local array that holds the current pointer window for each 136526 device. Benefit: searching for the first descendant of a pointer is a simple 136527 run through the array. 136528 136529 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136530 136531commit 72ad4a85cc0ffe60a90011d65ef718d5852beae4 136532Author: Peter Hutterer <peter.hutterer@who-t.net> 136533Date: Thu Dec 18 15:43:18 2008 +1000 136534 136535 dix: remove now unused "exclude" parameter from FirstPointerChild 136536 136537 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136538 136539commit 0aa4460c3b4f9bb17d5412d76fa8c4c501132429 136540Author: Peter Hutterer <peter.hutterer@who-t.net> 136541Date: Wed Dec 17 16:14:41 2008 +1000 136542 136543 dix: re-implement enter/leave model. 136544 136545 The old model was implemented based on a misunderstanding of NotifyVirtual and 136546 NotifyNonlinearVirtual events. It became complicated and was broken in some 136547 places [1]. This patch wipes this model completely. 136548 136549 A much simplified implementation is provided instead. Rather than a top-down 136550 approach ("we have a tree of windows, which ones need to get which event") 136551 this one uses a step-by-step approach. For each window W between A and B 136552 determine the pointer window P as perceived by this window and determine the 136553 event type based on this information. This is in-line with the model described 136554 by Owen Taylor [2]. 136555 136556 [1] http://lists.freedesktop.org/archives/xorg/2008-December/041559.html 136557 [2] http://lists.freedesktop.org/archives/xorg/2008-August/037606.html 136558 136559commit f8ecc2f08fc375b14d60a74e2fdd7830bfdcd74f 136560Author: Christian Beier <beier@informatik.hu-berlin.de> 136561Date: Tue Jan 6 11:31:03 2009 +1000 136562 136563 Xi: fix typo in WarpDevicePointer handling. 136564 136565 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136566 136567commit 3d549438c29004d78032ecc50ab45ca0e3f49623 136568Author: Thomas Jaeger <thjaeger@gmail.com> 136569Date: Mon Jan 5 01:26:42 2009 -0500 136570 136571 Don't alter device button maps in DoSetPointerMapping 136572 136573 Currently, if a device map differs from the core pointer map, then the 136574 request may return MappingBusy, even though all the affected core 136575 buttons are in the up state. 136576 136577 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136578 136579commit b72c6b083baeadfd7b18b6025df054be502d6e28 136580Author: Peter Hutterer <peter.hutterer@who-t.net> 136581Date: Tue Jan 6 09:01:48 2009 +1000 136582 136583 dix: remove inputInfo.keyboard reference in QueryPointer handling. 136584 136585 PickKeyboard() is to be used instead. 136586 136587commit 30d2cfcd3851870178d62e5067211aa36f87fbd2 136588Author: Kim Woelders <kim@woelders.dk> 136589Date: Mon Jan 5 19:35:18 2009 +0100 136590 136591 dix: Fix handling of do_not_propagate_mask window attribute. 136592 136593 This was broken in 32aa252e988be8cbfd4f7e373fb7b7736ef1f5f2. 136594 136595 Signed-off-by: Kim Woelders <kim@woelders.dk> 136596 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136597 136598commit 515ce3e4ba42605a1ee9979e8bb5acd3cf6470a3 136599Author: Peter Hutterer <peter.hutterer@who-t.net> 136600Date: Mon Jan 5 15:48:45 2009 +1000 136601 136602 xkb: fix typo - missing negation when checking button state. 136603 136604 Introduced with a85f0d6b98237d8a196de624207acf1983a1859a. 136605 136606 Reported by Thomas Jaeger. 136607 136608 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136609 136610commit 9a313fb03153eb764aab3e9cb9988b6603541637 136611Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 136612Date: Wed Jan 7 23:07:01 2009 -0200 136613 136614 Don't call free, use xfree macro. 136615 136616 There are several other direct calls to free, check 136617 % egrep '\<free\(' `find . -name \*.c` 136618 but they are free'ing memory from explicit malloc calls. 136619 This one was not intended, and corrected, so that it would 136620 both, follow the conventions everywhere (and work on some 136621 libc that doesn't like free(0)), and make it easier to use 136622 malloc wrappers. 136623 136624commit 49b93df8a3002db7196aa3fc1fd8dca1c12a55d6 136625Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 136626Date: Wed Jan 7 19:37:03 2009 -0200 136627 136628 Default to use standard bitmap fonts, with builtins as fallback 136629 136630 The builtin-fonts configure option was removed, as it at best should 136631 have been a runtime option. Instead, now it always register all "font 136632 path element" backends, and adds built-ins fonts at the end of the 136633 default font path. 136634 This should be a more reasonable solution, to "correct" the most 136635 common Xorg FAQ (could not open default font 'fixed'), and also don't 136636 break by default applications that use only the standard/historical 136637 X Font rendering. 136638 136639commit 4f004231871c830e16c4525cfaab183fad0aee86 136640Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136641Date: Wed Jan 7 10:55:08 2009 -0800 136642 136643 XQuartz: Fat binary buildfixes 136644 136645 Let Xarch.h set X_BYTE_ORDER for us. 136646 Determine _XSERVER64 at build time. 136647 (cherry picked from commit f54736c2b0b537c8322ecb2db8e9cee5eea68165) 136648 136649commit fbf44d85bd59bfa629193d6494b16812a7343cab 136650Author: Jon TURNEY <jon.turney@dronecode.org.uk> 136651Date: Mon Jan 5 23:18:11 2009 +0000 136652 136653 Cygwin/X: Tidy up an unused #define 136654 136655 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 136656 136657commit 1badeb78b0b974a381e563e3d4f4cf37debd4266 136658Author: Rémi Cardona <remi@gentoo.org> 136659Date: Wed Jan 7 09:31:17 2009 -0500 136660 136661 Include Xmd.h from edid.h 136662 136663commit fc14d52297f8f408693ed867ab649ea5b210b0b3 136664Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136665Date: Sun Nov 16 15:35:17 2008 -0800 136666 136667 XQuartz: Use AvailabilityMacros.h instead of availability.h for Tiger 136668 (cherry picked from commit 91f3c3435716195e6d606c8e2f6b0de52eb5972d) 136669 136670commit 4c326c4d342bede0de57e71c1431f56188f812ce 136671Author: Keith Packard <keithp@keithp.com> 136672Date: Tue Jan 6 08:36:39 2009 -0800 136673 136674 Make cvt complain about invalid arguments more often. 136675 136676 cvt doesn't correctly parse 'cvt -vr 1920 1080'. This patch doesn't fix 136677 that, but it does at least generate an error message now. 136678 136679 Signed-off-by: Keith Packard <keithp@keithp.com> 136680 136681commit 1da4a41b0881b605c7cfa4c4bf637efa132c58fc 136682Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 136683Date: Tue Jan 6 05:01:32 2009 -0200 136684 136685 Add sdksyms.sh to EXTRA_DIST to correct make distcheck 136686 136687 Thanks to <vehemens at verizon.net> for noticing the problem. 136688 136689commit 200230535fa80c04faaa562bbbd8272713e2763b 136690Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 136691Date: Mon Jan 5 16:49:57 2009 -0200 136692 136693 Update sdk headers to export new symbols. 136694 136695 All symbols in installed sdk headers should be explicitly tagged 136696 as exported symbols. Otherwise, to ensure it is not a mistake, one 136697 could write it as something like: 136698 extern /* NOEXPORT */ type name ...; 136699 but the proper procedure really should be to use a non sdk header 136700 (or a "noinst_" one). 136701 This patch also removes prototypes to some functions that existed 136702 only temporarily. 136703 136704commit af6152499654a7146250d4ad0df89f611345434e 136705Author: Jon TURNEY <jon.turney@dronecode.org.uk> 136706Date: Wed Dec 17 14:15:34 2008 +0000 136707 136708 Cygwin/X: Add *.exe pattern to .gitignore 136709 136710 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 136711 136712commit 8428b92fcb478b041f778c0e8c66fdeb96c807f3 136713Author: Colin Harrison <colin.harrison@virgin.net> 136714Date: Sun Jan 4 20:12:16 2009 +0000 136715 136716 Xming: Fix a couple of warnings 136717 136718 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 136719 136720commit f6f97419bd2db651384ecaec0b296840810463a4 136721Author: Jon TURNEY <jon.turney@dronecode.org.uk> 136722Date: Mon Jan 5 18:35:52 2009 +0000 136723 136724 Cygwin/X: Add -static back to linker flags 136725 136726 Add -static back to linker flags to avoid screwing things up for 136727 people who don't know that they need to avoid dynamically linking 136728 with libXfont, as weak symbol handling requires static linking to 136729 work correctly on Cygwin... 136730 136731commit 2ec0ac08addb938a3d5aa5b9888ce7c79a0daada 136732Author: Jon TURNEY <jon.turney@dronecode.org.uk> 136733Date: Wed Dec 17 19:18:38 2008 +0000 136734 136735 Cygwin/X: Fix some remaining build issues 136736 136737 Fixes for a few remaining build issues, omitted from commmit 6e85a8304b50c0bfd16a400a05d257f0168fb99b in error 136738 136739 Provide VENDOR_RELEASE in XORG_VERSION_CURRENT for XWin 136740 Remove -static from linker flags for XWin 136741 136742 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 136743 136744commit e351d10773742d1570e1111976b7c8b6adda5ae6 136745Author: Daniel Stone <daniel@fooishbar.org> 136746Date: Mon Jan 5 17:41:38 2009 +0200 136747 136748 RandR: Only export Xinerama symbols when building Xinerama support 136749 136750 Otherwise compilation fails due to sdksyms. 136751 136752 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 136753 136754commit 3387cbb7c13636eb62b22640d28df7011c2268f0 136755Author: Daniel Stone <daniel@fooishbar.org> 136756Date: Mon Jan 5 11:25:57 2009 +0200 136757 136758 XFree86: Xv: Add ClipNotify helper 136759 136760 Add a ClipNotify helper that lets the driver know about changes in the 136761 clipping of an Xv backing drawable. 136762 136763 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 136764 136765commit 55c4ec0a1ef78481b4d82153c19c20fbf88f1624 136766Author: Daniel Stone <daniel@fooishbar.org> 136767Date: Mon Jan 5 11:24:24 2009 +0200 136768 136769 XFree86: fbdevhw: Add helper function to get fd 136770 136771 Most devices need to do custom weird ioctls, so let them know what the 136772 fd is. 136773 136774 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 136775 136776commit 7da4e901e078e5597b8dbcb7ee39b949edf7ef03 136777Author: Daniel Stone <daniel@fooishbar.org> 136778Date: Mon Jan 5 11:23:39 2009 +0200 136779 136780 XFree86: Fix build with DGA disabled 136781 136782 It's optional, so we might as well work when it's disabled. 136783 136784 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 136785 136786commit 13ce7b98d48261b061da5768a6405bdbe9caa1fb 136787Author: Dan Nicholson <dbn.lists@gmail.com> 136788Date: Mon Dec 22 14:31:03 2008 -0800 136789 136790 dmx: Require newer libXi for XQueryInputVersion 136791 136792 Signed-off-by: Dan Nicholson <dbn.lists@gmail.com> 136793 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136794 136795commit 27011254c4de4e573a0851bf46892fb488db6522 136796Author: Peter Hutterer <peter.hutterer@who-t.net> 136797Date: Tue Dec 23 09:59:28 2008 +1000 136798 136799 xfree86: If an input device failed to activate, return immediately. 136800 136801 Devices are only activated once - right after they've been added to the 136802 server. If a device failes activation, it's dead. There's no reason to 136803 continue. Return the error code from ActivateDevice() without setting up 136804 sprite information or even sending a event to the client. 136805 136806 Then - in the DDX - just remove the device again. 136807 136808 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136809 136810commit e1a3a1a0d85c9971aea65c2228b5fd4dbf3bf57a 136811Author: Peter Hutterer <peter.hutterer@who-t.net> 136812Date: Tue Dec 23 10:00:34 2008 +1000 136813 136814 xfree86: don't call CheckMotion if a device hasn't been enabled. #19176 136815 136816 X.Org Bug <http://bugs.freedesktop.org/show_bug.cgi?id=19176> 136817 136818commit af820259a06a6e8dbd299fe22347aa83b0c985b0 136819Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136820Date: Fri Jan 2 11:06:12 2009 -0800 136821 136822 XQuartz: Copyright Update... happy new year 136823 (cherry picked from commit 9f1c26ae8d9ccf81cba0ab946326968a701f8c36) 136824 136825commit d8c9ba37abd0d63ea2c5394af434b46ec3b647fd 136826Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136827Date: Fri Jan 2 10:58:40 2009 -0800 136828 136829 XQuartz: pbproxy: Push dpy init and CFRunLoop hook setup into the pbproxy thread to avoid possible deadlock 136830 (cherry picked from commit 49e59d32b88e4fad070f230b5efaa261b47f78db) 136831 136832commit 2f361186f93a791e446c4a9471bb0c3b9b818d98 136833Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136834Date: Wed Dec 31 12:40:02 2008 -0800 136835 136836 XQuartz: Don't use NX_SECONDARYFNMASK, NX_NUMERICPADMASK, NX_HELPMASK 136837 136838 We don't have keycodes for them, so don't try to use them 136839 (cherry picked from commit cb912aca3a2834c4ad8e386c8a0d05c1bb31b0e7) 136840 136841commit a2abaa9fd3a5ad713c1f946e9d7f598825ad3a84 136842Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136843Date: Wed Dec 31 11:57:49 2008 -0800 136844 136845 XQuartz: Make sure to reset the saved key state when deactivating X11.app 136846 (cherry picked from commit 3eef78eb321f4f7dbca5a10c80666c621e28a1e0) 136847 136848commit 3986b683adc4f6f967cc5f643e973d89764234c7 136849Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136850Date: Wed Dec 31 12:44:43 2008 -0800 136851 136852 XQuartz: Honor system key repeat rate 136853 (cherry picked from commit 4303c9be39a86f5a21de108f72b90a989435905e) 136854 136855commit 1a69d56afda082298edc76271d2cb71af00526ce 136856Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136857Date: Mon Dec 29 19:54:12 2008 -0800 136858 136859 XQuartz: Better avoid stuck keys on context switches 136860 (cherry picked from commit e9963f1a4f4f12f253eae9d4f01694b6cabe35ad) 136861 136862commit f464b5a9164b24f74adea198bcabe9affaf5552c 136863Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136864Date: Mon Dec 29 19:38:31 2008 -0800 136865 136866 XQuartz: Workaround OSX VNC server bug for modifier key state 136867 136868 A better approach which ensures we have a L modifier key down if we are told neither are down and atleast one is down... =/ 136869 (cherry picked from commit bc13dda345f716bf4de9bfe4e1d85969263b60c2) 136870 136871commit 3ead9aa2602cab9c883aa7bcbd2a8d82c5c88f5a 136872Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136873Date: Mon Dec 29 00:30:31 2008 -0800 136874 136875 XQuartz: Updated man page fullscreen_hotkeys fullscreen_menu 136876 (cherry picked from commit 00f3a2e33c8e91da5f855821313a04ea97445656) 136877 136878commit 48dbaf173a82693fd72953983da9fc556cd1c6ed 136879Author: Daniel Stone <daniel@fooishbar.org> 136880Date: Tue Dec 30 12:17:14 2008 +1100 136881 136882 XKB: Also copy keyboard feedback when copying the keymap 136883 136884 When updating the XKB keymap, make sure the keyboard feedback is also 136885 copied, to preserve autorepeat settings etc. 136886 136887 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 136888 136889commit cf85a32b5f80694f3c2f434e91cf77edb33a7c17 136890Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136891Date: Sun Dec 28 15:57:59 2008 -0800 136892 136893 XQuartz: pbproxy: We explicitly need libX11 for pbproxy 136894 (cherry picked from commit 32824120da0749b8369c2592f851bb1030dea9f1) 136895 136896commit 997e8c321bb11067591785ba38103a3a5029301d 136897Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136898Date: Sun Dec 28 01:53:38 2008 -0800 136899 136900 XQuartz: Try harder to get the user's login environment 136901 (cherry picked from commit 6bedaddd78aa04bd303df434a4c49bb87bd7deaa) 136902 136903commit ae8dc47fd910ec4b9edd64bf3b7ba33e212c5942 136904Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136905Date: Wed Dec 24 00:39:06 2008 -0800 136906 136907 XQuartz: Reposition windows when we enter fullscreen to ensure our root window 136908 (cherry picked from commit 7617d3659b5481ef85aecc1f936e7dd2662bdfbd) 136909 136910commit 895186e026881196ac9383aa165e123055d9cf09 136911Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136912Date: Tue Dec 23 20:33:27 2008 -0800 136913 136914 XQuartz: Re-enable rlAccel 136915 136916 It was incorrectly blamed for some crashes a few months ago, and it should be 136917 safe to use until we get an EXA driver to replace it. 136918 (cherry picked from commit 8121f30bd29591fc74fbc680fbbf20210271fa58) 136919 136920commit dc4498ec5cb7f25f710bdec8793746da2d2324ed 136921Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136922Date: Sun Dec 28 22:16:52 2008 -0800 136923 136924 Updated .gitignore 136925 (cherry picked from commit 06964c322e4da81f3146022049a2de13f6b2892d) 136926 136927commit 3534a5e5d9c5af85149c799f324257f89507fa23 136928Author: Maarten Maathuis <madman2003@gmail.com> 136929Date: Fri Dec 26 16:38:27 2008 +0100 136930 136931 exa: Allow drivers to set non-NULL devPrivate.ptr for !offscreen pixmaps. 136932 136933commit 027b440d4f9f0cdd46addff46fd2d5c44cd5c847 136934Author: Maarten Maathuis <madman2003@gmail.com> 136935Date: Fri Dec 26 15:51:55 2008 +0100 136936 136937 exa: preparing as source and finishing access as mask is a bad idea 136938 136939commit 89bbd07e9ddd688df81405151f9c6284fa3367a8 136940Author: Alan Coopersmith <alan.coopersmith@sun.com> 136941Date: Tue Dec 23 21:40:21 2008 -0800 136942 136943 Remove obsolete XF86_DATE & xf86Date.h 136944 136945commit 8dc3a7d6817b7371bbfece1108ef856788256853 136946Author: Alan Coopersmith <alan.coopersmith@sun.com> 136947Date: Tue Dec 23 14:59:46 2008 -0800 136948 136949 Remove unused CLOG_DATE (leftover from pre-git ChangeLog) 136950 136951commit fe05a46f191ebd36251dda7a9703d90c25efec11 136952Author: Alan Coopersmith <alan.coopersmith@sun.com> 136953Date: Tue Dec 23 14:55:26 2008 -0800 136954 136955 Remove powerpc load/store functions from export list on other platforms 136956 136957commit a1b201bcb01e702f8f0a5889a3d49317503b9475 136958Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 136959Date: Tue Dec 23 12:39:38 2008 -0800 136960 136961 XQuartz: Use depth=24 instead of FatalError if we can't figure out our depth 136962 (cherry picked from commit 80b65c5b78d125c4ad3620b87b121c9e666299c3) 136963 136964commit 86dc660588a615baefb1799d78a501c95a931d77 136965Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 136966Date: Tue Dec 23 18:07:54 2008 -0200 136967 136968 Improve sdksyms.c automatic generation (Fix #19245). 136969 136970 Since it is already parsing cpp output, create a dependency file 136971 in the same process. This will cause sdksyms.c to be regenerated 136972 whenever a sdk header is modified. 136973 This also uses the gmake 'sinclude' directive (don't fail if 136974 included file doesn't exist). This should not cause any problems 136975 given that gmake only constructs are used in several other Makefiles. 136976 136977commit 477c32708718f4470ac2e786b2446d7a44e45b9c 136978Author: Peter Hutterer <peter.hutterer@who-t.net> 136979Date: Fri Dec 19 08:56:35 2008 +1000 136980 136981 dix: don't disable uninitialized devices. 136982 136983 If a device hasn't been initialized, it doesn't have a cursor yet. So don't 136984 set the cursor to the NullCursor, and don't try to DisableDevice either. 136985 136986 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 136987 136988commit 25aac8b5793caa92116b766d5a9e12dbecdb0290 136989Author: Peter Hutterer <peter.hutterer@redhat.com> 136990Date: Tue Dec 2 16:16:58 2008 +1000 136991 136992 xkb: don't treat groups with different no of symbols as identical. 136993 136994 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 136995 136996commit f141c1b4bb9482f188783dd6f161d3c7960a3329 136997Author: Peter Hutterer <peter.hutterer@redhat.com> 136998Date: Tue Dec 2 14:54:50 2008 +1000 136999 137000 xkb: explicitly check for group replication in the core representation. 137001 137002 Single-group keys may get replicated amongst all groups. Check explicitly for 137003 this case and squash it down to one group. 137004 137005 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 137006 137007commit 70a977c021e107c4fabe46ec2f619be9fb55abeb 137008Author: Peter Hutterer <peter.hutterer@redhat.com> 137009Date: Tue Dec 2 15:36:51 2008 +1000 137010 137011 xkb: don't replicate past the number of groups we have. 137012 137013 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137014 137015commit a157575eeeb523cd43197c5caeb00cb3d56f9988 137016Author: Peter Hutterer <peter.hutterer@redhat.com> 137017Date: Tue Dec 2 13:31:58 2008 +1000 137018 137019 xkb: ensure enough symbols for core Group1 replication. 137020 137021 A single-group key on a multi-group keyboard has to be replicated across all 137022 three groups (see Section 12.4 of the XKB protocol spec). Ensure that there's 137023 enough symbols available to actually do that. 137024 137025 e.g. a key ABCD on a 3 group keyboard needs to be replicated as ABABCDCDABCD, 137026 hence requiring space for 12 symbols, even if maxSymsPerKey is less than that. 137027 137028 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 137029 137030commit b5736d237a21d5e65d839c4d213dd3bda5a11e9d 137031Author: Maarten Maathuis <madman2003@gmail.com> 137032Date: Mon Dec 22 19:53:14 2008 +0100 137033 137034 randr/xfree86: Fix initial gamma computation. 137035 137036 - The previous version overflowed sometimes. 137037 137038commit ba9dc353a60e9a6bba64e854b23271011bff7015 137039Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137040Date: Sun Dec 21 21:03:59 2008 -0800 137041 137042 XQuartz: Don't use keycode 0 to determine !swallow since our most common key to swallow is actual keycode 0 (a) 137043 (cherry picked from commit 33f43a7f03023bfbab25a957cb81fc25b4afa4ca) 137044 137045commit f0cdccd1e22198f4f5e1aa0d0617441b79ffbc44 137046Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137047Date: Sun Dec 21 20:40:36 2008 -0800 137048 137049 XQuartz: Update our "screens" when we toggle rootless rather than when we toggle fullscreen 137050 137051 This old behavior was used as a workaround for the menubar behavior in the older server, 137052 but we handle it better now and need to update our screens when we toggle the rootless 137053 state instead. 137054 (cherry picked from commit 508aa95bc2cd3fdc3dff448ec090919bf807d153) 137055 137056commit 8f04a5237384fa97ff960b4607a79b688ea03fc1 137057Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137058Date: Sun Dec 21 14:54:11 2008 -0800 137059 137060 XQuartz: Run applications via '/bin/sh -c ...' to support users who expect shell parsing 137061 (cherry picked from commit 67455e716e3ecffd528930479192785958d37988) 137062 137063commit 029d255a654eca6401c374d145f63bbb923554b5 137064Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137065Date: Sun Dec 21 14:43:12 2008 -0800 137066 137067 XQuartz: pbproxy: Release display notification lock when not needed to avoid deadlock 137068 (cherry picked from commit 22f664ab95a0cae981e9cefad6f075d051583ca5) 137069 137070commit b1f166f2981595aea6f8e2a336d6d25436f8a07d 137071Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137072Date: Sun Dec 21 14:42:26 2008 -0800 137073 137074 XQuartz: update quoting in case X11.app is moved to a directory with a space. 137075 (cherry picked from commit cc805dc799efa37c8dcefa3db04d87e9b835ffbd) 137076 137077commit 2f47bda645eb43c60b71315e8d265190acce481d 137078Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137079Date: Thu Dec 18 09:14:16 2008 -0800 137080 137081 XQuartz: Changed X11.sh to allow use of a ~/.x11run as requested by users of alternate shells 137082 (cherry picked from commit b62ed1f8eaf041a946bb591165bb18ee481dedbf) 137083 137084commit 83edcc08de4c815c5cbe815f521b3b728441cad6 137085Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137086Date: Wed Dec 17 15:09:35 2008 -0800 137087 137088 XQuartz: Get rid of white rectangle bug 137089 (cherry picked from commit 3269959033ed0c675a3a906666454df34086896a) 137090 137091commit 1335a300b9cd03208a583acfdb7f7153eaef4ea6 137092Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137093Date: Wed Dec 17 14:43:02 2008 -0800 137094 137095 rootless: Make expose_1 static 137096 (cherry picked from commit 60c8d2697036a125ca5381df8e2eaedabad4d242) 137097 137098commit 22b2811e6aa0bdbfa6a0682ac862063e52ec0c62 137099Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137100Date: Wed Dec 17 13:54:53 2008 -0800 137101 137102 XQuartz: Name the startup shell script X11 for better compatability 137103 (cherry picked from commit f84613d6fcbbb5f769ee34cff0900353a13693cf) 137104 137105commit 82fba7697fbbb6f66cc22424ba99755b99ff482e 137106Author: Maarten Maathuis <madman2003@gmail.com> 137107Date: Mon Dec 22 00:33:55 2008 +0100 137108 137109 xfree86/randr: Avoid a crash when xf86CrtcConfigPrivateIndex is -1. 137110 137111commit 1567b7243f4799808ab93fbd962df14ce3af2d49 137112Author: Maarten Maathuis <madman2003@gmail.com> 137113Date: Sun Dec 21 00:08:56 2008 +0100 137114 137115 exa: A more correct fix. 137116 137117commit 2db7b66863ae6055c3ce13c88b36d620de8a4d75 137118Author: Maarten Maathuis <madman2003@gmail.com> 137119Date: Fri Dec 19 23:12:37 2008 +0100 137120 137121 exa: a few cleanups 137122 137123 - Some warnings silenced. 137124 - Some whitespace cleanup. 137125 137126commit aedd2f566df585db7a1614f302cc8d3feda54275 137127Author: Maarten Maathuis <madman2003@gmail.com> 137128Date: Fri Dec 19 19:10:23 2008 +0100 137129 137130 randr/xfree86: Fix a one off error in the panning calculations. 137131 137132 - Example: mode 1280x1024, panned area 1281x1024 137133 panned_area.x2 = 1281 137134 mode.width = 1280 137135 If you substract 1280 from 1281, then that leaves you with one. 137136 Which is the one pixel that you need to move to actually see the last pixel collumn. 137137 Substracting 1 from this will consistently prevent you from seeing the right and bottom edge. 137138 137139commit 332d65ec7a6e94d75efe95d53742f137835274de 137140Author: Maarten Maathuis <madman2003@gmail.com> 137141Date: Fri Dec 19 18:59:27 2008 +0100 137142 137143 randr: Consider panned crtc's when calculating xinerama screen sizes. 137144 137145 - This will allow window managers and applications to actually use the panned area. 137146 137147commit f2d1de6c2a3e6e50db6d9714c293eb26324fbd3d 137148Author: Adam Jackson <ajax@redhat.com> 137149Date: Fri Dec 19 10:27:37 2008 -0500 137150 137151 xv: remove useless XVCALL macro 137152 137153commit 5a072c55350f4b73d911ea6a2aeddad844924834 137154Author: Adam Jackson <ajax@redhat.com> 137155Date: Fri Dec 19 09:51:52 2008 -0500 137156 137157 dmx: Fix calloc macro confusion. 137158 137159commit 8c488ac3b3990cd203baed7f2127b9bed8aab534 137160Author: Alan Coopersmith <alan.coopersmith@sun.com> 137161Date: Thu Dec 18 19:47:49 2008 -0800 137162 137163 Fix linking of Xorg with dtrace probes on Solaris 137164 137165commit 396433d0da721951e6e7abc02dc3b2f682495154 137166Author: Alan Coopersmith <alan.coopersmith@sun.com> 137167Date: Thu Dec 18 18:49:47 2008 -0800 137168 137169 Fix sdksyms.sh to work with Solaris/Sun compiler builds 137170 137171 - Pass $(CPP) & $(AWK) settings from configure to sdksyms.sh 137172 - Only reset sdk variable (tracks if header is part of sdk) if 137173 a filename is included on the cpp # <line-no> <filename> line, 137174 since Sun compilers omit filename when it is unchanged from 137175 previous line. 137176 137177commit aea6f19f25e13768b1d09fac4991d6a5e6c2cdac 137178Author: Peter Hutterer <peter.hutterer@who-t.net> 137179Date: Tue Dec 16 17:10:03 2008 +1000 137180 137181 xfree86: don't restore the TTY mode if we didn't initialize it ourselves 137182 137183 Restoring it unconditionally means we restore to whatever tty_mode has as 137184 default value (i.e. 0). K_RAW happens to be 0x00, so we always restore to raw 137185 mode if allowEmptyInput is off. 137186 137187 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137188 Signed-off-by: Adam Jackson <ajax@redhat.com> 137189 Signed-off-by: Julien Cristau <jcristau@debian.org> 137190 137191commit 5d065a889074558634216eebd4bba35a60db9a09 137192Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 137193Date: Thu Dec 18 13:59:36 2008 -0500 137194 137195 xselinux: Use xace Xtrans wrappers instead of the now-inaccessible wrapees. 137196 137197commit f87e66486c3610c56888915b66ae5ab0af03da8b 137198Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 137199Date: Thu Dec 18 13:58:35 2008 -0500 137200 137201 xace: Export wrappers around two Xtrans functions used by modules. 137202 137203 Don't know a better way to do this, since Xtrans isn't a library that 137204 can be linked into modules. 137205 137206commit 777408914dbf395e9969e3f34720286e6f740369 137207Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 137208Date: Thu Dec 18 12:35:57 2008 -0500 137209 137210 Add xace headers to the SDK when enabled and export the XaceHooks symbol 137211 to modules. 137212 137213commit 9e4ef3cfe5ee178c266edd64a8edd767827aa0d7 137214Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 137215Date: Thu Dec 18 15:00:25 2008 -0200 137216 137217 Use regex pattern understood by all known awk variants. 137218 137219 Ubuntu uses mawk by default, but it doesn't understand posix character 137220 classes (which are locale dependent, and this patch uses only valid C 137221 identifiers). 137222 Also make sure awk runs with LC_ALL=C to match the regex patterns. 137223 137224commit 3d3c234b434a3443a00c3be9f32c698bcced111c 137225Author: Maarten Maathuis <madman2003@gmail.com> 137226Date: Thu Dec 18 16:26:38 2008 +0100 137227 137228 randr: add some notes about the stuff that changed in driver ABI 2 137229 137230commit 1b244feb4cf11d1cdba79fbead50035965535552 137231Author: Maarten Maathuis <madman2003@gmail.com> 137232Date: Thu Dec 18 16:11:21 2008 +0100 137233 137234 randr: check for malloc failure 137235 137236commit 21a2df9f359584e634dc12d8602401d54defd7db 137237Author: Colin Harrison <colin.harrison@virgin.net> 137238Date: Sun Nov 2 20:46:44 2008 +0000 137239 137240 Xming,Cygwin/X: Fix crashes when using the DirectDraw '-refresh rate-in-Hz' option in -fullscreen (#11128) 137241 137242 fd.o bugzilla #11128 137243 137244 Fix crashes when using the '-refresh rate-in-Hz' option in -fullscreen mode and using a DirectDraw engine. 137245 On failure (typically 0x80004001 DDERR_UNSUPPORTED), use the default refresh rate for the driver, instead of the 137246 entered value. 137247 137248 Also fix some nearby error reporting 137249 137250 Copyright (C) Colin Harrison 2005-2008 137251 http://www.straightrunning.com/XmingNotes/ 137252 http://sourceforge.net/projects/xming/ 137253 137254 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137255 137256commit b0ed1567c1782041cf3d14ec6dd4d622915c7d0c 137257Author: Colin Harrison <colin.harrison@virgin.net> 137258Date: Wed Dec 17 17:16:02 2008 +0000 137259 137260 Xming, Cygwin/X: Tidy up warnings, headers and unused variables (#11132) 137261 137262 Tidy up various warnings, include headers and unused variables, etc. 137263 137264 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137265 137266commit 15c4a6e088039e14b4f2387ff204cb5448916b39 137267Author: Colin Harrison <colin.harrison@virgin.net> 137268Date: Wed Dec 17 16:44:07 2008 +0000 137269 137270 Xming, Cygwin/X: Fix crash in function winMultiWindowGetTransientFor() (#11147) 137271 137272 The function winMultiWindowGetTransientFor() crashes, due to a bogus pointer 137273 dereference, but this fact was previously hidden by the function not being called! 137274 137275 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137276 137277commit 08a3d6928c87032998b6113d67f69b79b09eecee 137278Author: Peter Hutterer <peter.hutterer@who-t.net> 137279Date: Tue Dec 16 14:19:12 2008 +1000 137280 137281 dix: Allocate the space for the DCCE when the last SD is removed. 137282 137283 This still doesn't fix reset the MD's classes (a TODO that's been here for 137284 ages), but at least we don't segfault anymore when detaching the last SD. 137285 137286 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137287 137288commit 689be0cb4b259840208c17f974066a942196793a 137289Author: Peter Hutterer <peter.hutterer@who-t.net> 137290Date: Tue Dec 16 14:19:38 2008 +1000 137291 137292 dix: set the correct length in CreateClassesChangedEvent. 137293 137294 Fallout from aeff14d5f208d02f211b8b2a1a98999624e8c2cf. Yes, we don't malloc 137295 anymore because we are inside a SIGIO and the memory is already there anyway. 137296 But we still need to set the event length correctly, otherwise 137297 mieqEnqueue/mieqProcessInputEvent don't know how much memory to copy. 137298 137299 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137300 137301commit 1001403301001fcc84ce6833938590956df40d2a 137302Author: Maarten Maathuis <madman2003@gmail.com> 137303Date: Thu Dec 18 00:19:06 2008 +0100 137304 137305 Forgot to fix a comment. 137306 137307commit 9fa15bef59881bdcf087889f16ab3c8d953da8f1 137308Author: Maarten Maathuis <madman2003@gmail.com> 137309Date: Wed Dec 17 23:42:30 2008 +0100 137310 137311 randr: some improvements, fixes and crtc abi bump 137312 137313 - Add active field to crtc. 137314 - Set gamma (only) whenever a crtc becomes active. 137315 - Check for xf86_config being NULL. 137316 - Increase crtc abi to 3. 137317 - A few other fixes. 137318 137319commit bf65523ab0b39774f07a7ae478ff3f5653fad469 137320Author: Joe Krahn <jkrahn@nc.rr.com> 137321Date: Sun Nov 2 20:17:03 2008 +0000 137322 137323 Cygwin/X: Fix for mis-aligned icon data creates bad background masks (#4491) 137324 137325 fd.o bugzilla #4491 137326 137327 Transparent icon backgrounds appear with black stripes as mask data is incorrectly aligned. 137328 Modified the icon DDB bitmap data alignment to 16 bits 137329 137330 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137331 137332commit 91f73b79b7ae64e5b846d1efeb470bb61a913720 137333Author: Maarten Maathuis <madman2003@gmail.com> 137334Date: Wed Dec 17 16:56:26 2008 +0100 137335 137336 randr: Improve per-crtc gamma support. 137337 137338 - The Gamma values from the monitor section are now used during initial config. 137339 - The old colormap system is disabled when gamma set hook is available. 137340 - Gamma values are now persistent for the lifetime of the xserver. 137341 - This requires no driver changes and should be driver ABI compatible. 137342 137343commit 1556815d34cecb4b4b62d2a4ce813b1435a937ec 137344Author: Jon TURNEY <jon.turney@dronecode.org.uk> 137345Date: Sun Nov 2 18:27:18 2008 +0000 137346 137347 Cygwin/X: Initialize native HWND atom when built !XWIN_MULTIWINDOWEXTWM 137348 137349 Fix the native HWND atom identifier to be initialized correctly, when built without external WM support 137350 137351 This is neccessary for native window naming to work correctly in multiwindow internal WM mode. 137352 137353 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137354 137355commit 92082a0a00f05ecdc90814a8104828853c9c8f29 137356Author: Jon TURNEY <jon.turney@dronecode.org.uk> 137357Date: Sun Nov 2 18:27:15 2008 +0000 137358 137359 Cygwin/X: update to use standard DPMS stubs 137360 137361 Bring Cygwin/X up to date with DPMS API changes by using the standard stubs 137362 137363 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137364 137365commit 2850f26bb2999ef3fb2a10adb80fe358463aecb5 137366Author: Jon TURNEY <jon.turney@dronecode.org.uk> 137367Date: Sun Nov 2 18:27:14 2008 +0000 137368 137369 Cygwin/X: update for changes to shadow framebuffer 137370 137371 Bring Cygwin/X up to date with the change to damage-based shadow framebuffer. 137372 137373 Register the screen bitmap as the shadow framebuffer after screen resources 137374 are created. 137375 137376 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137377 137378commit 35d76028304f910edeaf642376df71b8134ec02c 137379Author: Jon TURNEY <jon.turney@dronecode.org.uk> 137380Date: Sun Nov 2 18:27:13 2008 +0000 137381 137382 Cygwin/X: enqueue a pointer motion event on mouse movement 137383 137384 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137385 137386commit 816e0d243df068bc06d328874b5655d4530fe16a 137387Author: Jon TURNEY <jon.turney@dronecode.org.uk> 137388Date: Sun Nov 2 18:27:12 2008 +0000 137389 137390 Cygwin/X: update for changes in mieq API 137391 137392 Bring Cygwin/X up to date with mieq event API changes 137393 137394 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137395 137396commit bca7483c4e95ad450fb09048a3b85bb3056a0518 137397Author: Jon TURNEY <jon.turney@dronecode.org.uk> 137398Date: Sun Nov 2 18:27:11 2008 +0000 137399 137400 Cygwin/X: update for MPX device changes 137401 137402 Bring Cygwin/X up to date with MPX device API changes 137403 137404 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137405 137406commit 829a8a970dd17623ef3cb5022d34190744f74741 137407Author: Jon TURNEY <jon.turney@dronecode.org.uk> 137408Date: Sun Nov 2 18:27:10 2008 +0000 137409 137410 Cygwin/X: update for MPX cursor API changes 137411 137412 Bring Cygwin/X up to date with cursor API changes 137413 137414 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137415 137416commit d61e902aab92c262e6c8ee9cd70aec4493cf6cae 137417Author: Francis Giraldeau <francis.giraldeau@revolutionlinux.com> 137418Date: Wed Dec 17 09:49:39 2008 +0200 137419 137420 Don't log audit messages when -audit 0 specified 137421 137422 make the auditTrailLevel check more general and don't log when not 137423 using DTRACE and -audit 0 is set. 137424 137425commit 295a3fa721ee2f2226891de308c0c4b0ee26fe72 137426Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 137427Date: Wed Dec 17 02:04:12 2008 -0200 137428 137429 Ensure symbols required by swrast_dri.so are visible. 137430 137431commit 6d8ea5104cf97dbf64612f58fc06f94f869ed5ec 137432Author: Colin Guthrie <cguthrie@mandriva.org> 137433Date: Tue Dec 16 17:44:27 2008 -0500 137434 137435 Fix compilation with -Werror=format-security 137436 137437commit 1bb5a919f4bf38ac96c73077021b0cdc82965f31 137438Author: Adam Jackson <ajax@redhat.com> 137439Date: Tue Dec 16 13:46:59 2008 -0500 137440 137441 config: Add vboxvideo and tga to the magic driver list 137442 137443 Also sort the list while I'm here. 137444 137445commit 8511a964f81b3b06a526f0fca4232afb2152f405 137446Author: Matthias Hopf <mhopf@suse.de> 137447Date: Tue Dec 16 18:21:20 2008 +0100 137448 137449 randr: Update EDID_ATOM_NAME to reflect RandR 1.3 name change 137450 137451commit a058192d970f7417b06334aa666c17e9891795c6 137452Author: Owen W. Taylor <otaylor@redhat.com> 137453Date: Wed Oct 22 16:12:53 2008 -0400 137454 137455 Xephyr: Fix crash with control-shift and GLX (#18185) 137456 137457 When we are looking up the screen for an event, we need to take 137458 into account the fact that the event may have been delivered to the 137459 "peer window" that we create when implementing GLX. Since we only 137460 ever create one such window per screen, just add a single peer_win 137461 field to EphyrHostScreen. 137462 137463commit 6e85a8304b50c0bfd16a400a05d257f0168fb99b 137464Author: Jon TURNEY <jon.turney@dronecode.org.uk> 137465Date: Sun Nov 2 18:27:09 2008 +0000 137466 137467 Cygwin/X: build machinery fixes 137468 137469 The Cygwin/X build has been broken for a long time, probably since modular X 137470 This patch fixes up configure/makefile and some general build issues (#18568) 137471 137472 Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk> 137473 137474commit 7aa29b9d66c3cd0f8af4fafbe92efd0c0556d225 137475Author: Robert Noland <rnoland@2hip.net> 137476Date: Tue Dec 16 00:07:42 2008 -0500 137477 137478 Support -sharevts on FreeBSD 137479 137480 Bugzilla #11179 137481 137482 Submitted-by: Nathan Whitehorn 137483 137484 Signed-off-by: Robert Noland <rnoland@2hip.net> 137485 137486commit 063eb6743cd0d98dd52d1a9559b804381ee5144d 137487Author: Keith Packard <keithp@keithp.com> 137488Date: Mon Dec 15 19:53:45 2008 -0800 137489 137490 Patch brown-bag fix for bug 19017 (scrn->virtualX/virtualY 0 before PreInit) 137491 137492 When a driver uses a crtc during device detection, the scrn has not yet been 137493 configured and virtualX/virtualY are still zero. This caused the X server 137494 to try and allocate a shadow frame buffer, which couldn't work. 137495 137496 Detect this by checking for zero virtualX/virtualY values. 137497 137498 Signed-off-by: Keith Packard <keithp@keithp.com> 137499 137500commit fde2f961035609cfff8761ef1e0f23e6a9761be2 137501Author: Keith Packard <keithp@keithp.com> 137502Date: Mon Dec 15 11:39:24 2008 -0800 137503 137504 Use scrn->virtualX/virtualY in xf86CrtcFitsScreen. Fix bug 19017. 137505 137506 pScreen->width/height are not initialized when doing initial mode setting, 137507 which makes this function incorrectly fail. Using scrn->virtualX should work 137508 in all cases though. 137509 137510 Bug 19017 reports a crash in xf86CrtcSetModeTransform when doing a modeset 137511 for output probing, long before the screen array is initialized; that was 137512 caused by a work-around to set pScreen->width/height so that xf86CrtcFitsScreen 137513 could find the right values. 137514 137515 Signed-off-by: Keith Packard <keithp@keithp.com> 137516 137517commit 29a5b0596b396d3e4a8a014cacd3e3ef77467ab7 137518Author: Keith Packard <keithp@keithp.com> 137519Date: Mon Dec 15 11:38:00 2008 -0800 137520 137521 When disabling SIGIO tracking, use SIG_IGN instead of SIG_DFL. 137522 137523 This avoids a race condition for drivers which mis-order the fd close and 137524 signal disable. 137525 137526 Signed-off-by: Keith Packard <keithp@keithp.com> 137527 137528commit 1ba4cbb15919759aadd71960c5c057af9ba94fe3 137529Author: Keith Packard <keithp@keithp.com> 137530Date: Mon Dec 15 11:36:43 2008 -0800 137531 137532 Clean up rotation data when crtc is turned off 137533 137534 The shadow frame buffer and other data used for rotation need to be freed 137535 when the crtc is disabled, not just when rotation is disabled. 137536 137537 Signed-off-by: Keith Packard <keithp@keithp.com> 137538 137539commit 2a61397d17339113b9e37995b06ca543589814ce 137540Author: Cooper Yuan <Cooper.Yuan@amd.com> 137541Date: Mon Dec 15 10:29:34 2008 -0500 137542 137543 Fix typo in xf86PickCrtcs() 137544 137545commit 3db60add6671f766f9360e7d03492dfd51eba225 137546Author: Jon TURNEY <jon.turney@dronecode.org.uk> 137547Date: Mon Dec 15 12:42:10 2008 +0000 137548 137549 Cygwin/X: Update the native icon to one based on the X.Org logo 137550 137551 Update the native icon for the X server to one created from the X.Org logo by Colin Harrison. 137552 137553commit 450739efa904a0fce5024372da7236bb7b84b67a 137554Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137555Date: Sat Dec 13 00:45:27 2008 -0800 137556 137557 XQuartz: Updated menu item ordering for better HIG compliance 137558 (cherry picked from commit 4c9bb241ce5fb856fab20ae96fa89bd2b71ef242) 137559 137560commit aee27be2785d8b09ac10f5be040b4d573190de0e 137561Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137562Date: Wed Dec 10 23:04:30 2008 -0800 137563 137564 XQuartz: Tiger fix, don't call Xplugin code in the Appkit thread if Xplugin isn't threadsafe. 137565 (cherry picked from commit 748d9e5bd756513d42c4046f3b31e1fdc55bccb6) 137566 137567commit a939368ab8140d48c1da4ba0bb229d13b221189c 137568Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137569Date: Mon Dec 15 03:18:13 2008 -0800 137570 137571 mi: Reuse memory in mieqProcessInputEvents rather than making excessive calls to calloc() 137572 137573 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 137574 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 137575 137576commit f1c9b5ab230cbb4124d8d476ae4886d05022adcb 137577Author: Ian Romanick <ian.d.romanick@intel.com> 137578Date: Sun Dec 14 18:58:33 2008 -0800 137579 137580 GLX: Changes resulting from changes to Mesa generator scripts / data 137581 137582 Several recent Mesa commits (listed below) make modifications to the 137583 protocol generator data and scripts. This commit represents the 137584 changes to the generated files resulting from the previous changes. 137585 137586 - 0f73302d24f4201813da2939742c5bcb6964b3b1 137587 GLX: Fix protocol for glTexSubImage#D 137588 137589 - 1709ab01ef24279c782e420568e9257b4b92b224 137590 Return 0 as the request size when the pixels parameter is NULL 137591 137592 - 63cca2ba10ce7dcc8481cfa4be3872dfc269dded 137593 GLX: Include glapi.h before glapitable.h 137594 137595 This is the server-side part of the fix for bugzilla #11003. 137596 137597commit 6383bc93b2ae76361a58653bc7e291df0d6ddfec 137598Author: Peter Hutterer <peter.hutterer@who-t.net> 137599Date: Sat Dec 13 15:44:52 2008 +1000 137600 137601 dix: Init DeviceEnterLeave event's type before FixUpEventFromWindow (#19064) 137602 137603 X.Org Bug 19064 <http://bugs.freedesktop.org/show_bug.cgi?id=19064> 137604 137605commit 49d38ab2328c409b2a98465b52677af057121513 137606Author: Eric Anholt <eric@anholt.net> 137607Date: Tue Dec 2 15:09:57 2008 -0800 137608 137609 Warning fix: Remove dead glXDisp{,Swap}_DrawArraysEXT definitions. 137610 137611 Signed-off-by: Eric Anholt <eric@anholt.net> 137612 Acked-by: Ian Romanick <ian.d.romanick@intel.com> 137613 137614commit 2c5bfffc832e3818bbf0c0a96522865ce1ef2653 137615Author: Eric Anholt <eric@anholt.net> 137616Date: Tue Dec 2 14:59:04 2008 -0800 137617 137618 Warning fix (GL likes to call strings GLubyte * instead of char *). 137619 137620 Signed-off-by: Eric Anholt <eric@anholt.net> 137621 Acked-by: Ian Romanick <ian.d.romanick@intel.com> 137622 137623commit 7be6520d94df874c6bbd46d06a1830a12d0967f2 137624Author: Stuart Bennett <sb476@cam.ac.uk> 137625Date: Fri Dec 12 00:08:44 2008 +0000 137626 137627 dolt: allow older versions of bash to compile the xserver (#19031) 137628 137629 Signed-off-by: James Cloos <cloos@jhcloos.com> 137630 137631commit aeff14d5f208d02f211b8b2a1a98999624e8c2cf 137632Author: Peter Hutterer <peter.hutterer@redhat.com> 137633Date: Tue Nov 25 20:23:50 2008 +1000 137634 137635 dix: don't alloc in ChangeMasterDeviceClasses. 137636 137637 We mustn't realloc as we are inside a signal handler. With 137638 SetMinimumEventSize, this code should never be hit anyway, as the event list 137639 should have the required memory before this code is hit. 137640 137641 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 137642 137643commit fb2a8d0e59a3d187255538f6add22ec67551507a 137644Author: Peter Hutterer <peter.hutterer@who-t.net> 137645Date: Mon Dec 8 15:58:15 2008 +1000 137646 137647 Xi: silence compiler warning 137648 137649 Don't mix declarations and statements. 137650 137651commit d961abb59f0a8bee4bd11b1540aa43ff83cddbb8 137652Author: Peter Hutterer <peter.hutterer@who-t.net> 137653Date: Thu Dec 11 15:54:15 2008 +1000 137654 137655 dix: fix compiler warning (mixing declarations + statements) 137656 137657 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137658 137659commit cb95642dc8edebb2935dd471f8b339cb98aa8481 137660Author: Peter Hutterer <peter.hutterer@redhat.com> 137661Date: Fri Nov 28 22:28:32 2008 +1000 137662 137663 Remove #define NEED_EVENTS and NEED_REPLIES 137664 137665 A grep on xorg/* revealed there's no consumer of this define. 137666 137667 Quote Alan Coopersmith: 137668 "The consumer was in past versions of the headers now located 137669 in proto/x11proto - for instance, in X11R6.0's xc/include/Xproto.h, 137670 all the event definitions were only available if NEED_EVENTS were 137671 defined, and all the reply definitions required NEED_REPLIES. 137672 137673 Looks like Xproto.h dropped them by X11R6.3, which didn't have 137674 the #ifdef's anymore, so these are truly ancient now." 137675 137676 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 137677 Signed-off-by: Adam Jackson <ajax@redhat.com> 137678 137679commit 920b7622be4a68ebb5d8432f55c5362fbe964fce 137680Author: Simon Thum <simon.thum@gmx.de> 137681Date: Tue Nov 25 14:28:26 2008 +1000 137682 137683 xfree86: dump the ptraccel filter setup to the log on init. 137684 137685 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 137686 137687commit fbb57a2c54b141ccbb50833143b0b7cb2c4cb903 137688Author: Simon Thum <simon.thum@gmx.de> 137689Date: Tue Nov 25 13:43:28 2008 +1000 137690 137691 dix: ptraccel - Add GetAccelerationProfile() 137692 137693 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 137694 137695commit beb749c87f383a62172a9ffa36373438e7a12d8e 137696Author: Adam Jackson <ajax@redhat.com> 137697Date: Thu Dec 11 17:04:37 2008 -0500 137698 137699 Fix syncsrv.h guard define 137700 137701commit 9fba808b4f43f5dc69d85a48fa40cb83dd4d9f9d 137702Author: Colin Harrison <colin.harrison@virgin.net> 137703Date: Thu Dec 11 13:22:53 2008 -0500 137704 137705 xsync: Prototype fix. 137706 137707commit dce887ff53222652c397bc96a1dd3566f5847745 137708Author: Adam Jackson <ajax@redhat.com> 137709Date: Thu Dec 11 13:16:02 2008 -0500 137710 137711 xsync: build fix 137712 137713 argh protocol header disaster 137714 137715commit d41b8960be8ad316504ef5657c0abfe6d76d3b5e 137716Author: Matthias Hopf <mhopf@suse.de> 137717Date: Thu Dec 11 18:48:02 2008 +0100 137718 137719 randr: Fix initial panning border copy 137720 137721 Thanks to Julien Cristau for finding this. 137722 137723commit fd77ce9f884f5ac4d36736f3a99ba86101f133ea 137724Author: Matthias Hopf <mhopf@suse.de> 137725Date: Thu Dec 11 17:51:20 2008 +0100 137726 137727 randr: Oops, miscalculated panning rectangle's coordinates 137728 137729commit f8e52f1b6d7c59d007de99a1c9c69c053d4f3cbe 137730Author: Matthias Hopf <mhopf@suse.de> 137731Date: Thu Dec 11 16:48:40 2008 +0100 137732 137733 randr: Update initial screen size if panning information is present 137734 137735commit 24d6b7d1c5fc5d07c2eb06a9fc4406e393d59b17 137736Author: Matthias Hopf <mhopf@suse.de> 137737Date: Thu Dec 11 16:48:11 2008 +0100 137738 137739 randr: Fix error message for bad panning config 137740 137741commit fdbbe65a7e777b7777bfae5a161efb89d4fb9a8d 137742Author: Matthias Hopf <mhopf@suse.de> 137743Date: Thu Dec 11 14:56:51 2008 +0100 137744 137745 randr: Add monitor option "Panning" for initial panning configuration 137746 137747commit 27261a950d91c352eac25a3036656c3e4f81fb12 137748Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 137749Date: Thu Dec 11 14:43:04 2008 -0200 137750 137751 Modify sdksyms.sh to receive $top_srcdir as first argument. 137752 137753 If the basename of header file processed by cpp matches $top_srcdir, 137754 check for extern symbols in the output, and add to the xorg_symbols 137755 vector. 137756 Possibly a better solution then using this script would be to somehow 137757 tell the linker to not drop any symbols from the binary being generated. 137758 137759commit 1f4fb0225b278d1cf4145aebeb0bdd23dc8f62d5 137760Author: Adam Jackson <ajax@redhat.com> 137761Date: Wed Dec 10 16:13:20 2008 -0500 137762 137763 xsync: Fix wakeup storm in idletime counter. 137764 137765 Wakeup scheduling only considered the threshold values, and not whether 137766 the trigger was edge or level. 137767 137768 See also: 137769 https://bugzilla.redhat.com/show_bug.cgi?id=474586 137770 http://svn.gnome.org/viewvc/gnome-screensaver/trunk/src/test-idle-ext.c?view=markup 137771 137772commit 1a99110f0c221b79045ea26d61c4a1ec1e0d7341 137773Author: Adam Jackson <ajax@redhat.com> 137774Date: Wed Dec 10 14:18:59 2008 -0500 137775 137776 xsync: ANSI cleanups 137777 137778commit e0d8f6a8084a3fe6c549c1dee11e4502e316c382 137779Author: Adam Jackson <ajax@redhat.com> 137780Date: Wed Dec 10 13:32:04 2008 -0500 137781 137782 xsync: Use a local header for server API definitions 137783 137784commit 1208a1dbcaecbc218a77bf51068f9c0c768e16c5 137785Author: Adam Jackson <ajax@redhat.com> 137786Date: Wed Dec 10 11:47:35 2008 -0500 137787 137788 xsync: remove cast abuse. 137789 137790commit 119d9c46e841f5fa35610f557e6fa1ec58587c24 137791Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137792Date: Wed Dec 10 01:56:01 2008 -0800 137793 137794 XQuartz: Fix path to executable 137795 (cherry picked from commit 7e9d3698e070a0c63dd2556651373c3aca58e2fe) 137796 137797commit c46b564c47fde3474e948aa1b188a975836cba47 137798Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137799Date: Wed Dec 10 00:00:29 2008 -0800 137800 137801 XQuartz: fixed make dist 137802 (cherry picked from commit 2d52367ab92429e58596d1a1ed3ef52a0a38a7aa) 137803 137804commit 23156a21d8e2f1be7d19d69cb91dae1b885ef671 137805Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137806Date: Tue Dec 9 23:48:18 2008 -0800 137807 137808 XQuartz: Make debugging output for invalid depths a bit more detailed 137809 (cherry picked from commit 609fb166b7062c76f0561df12ffe893811f6ac8f) 137810 137811commit 76351d2faf9ceb79aaa00b374c203b8b279c58f3 137812Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137813Date: Tue Dec 9 23:47:32 2008 -0800 137814 137815 XQuartz: Avoid using login /bin/sh blech. Just use a bash script to start the app, so it will inherit the right environment 137816 (cherry picked from commit f4b7ad9cc6b0c99fc7ee8516c4bf858ece938148) 137817 137818commit e5ce6e198f4c245f4dc840c5e9e90a7ef80fb6e9 137819Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137820Date: Mon Dec 8 21:59:49 2008 -0800 137821 137822 XQuartz: unset DISPLAY if we didn't get a launchd socket handoff 137823 (cherry picked from commit b959727f38733009c6381cc8ca06b5984257bdac) 137824 137825commit ecaaf0462d6c70cabfcaf6868a526e5415505e2c 137826Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137827Date: Mon Dec 8 21:00:11 2008 -0800 137828 137829 XQuartz: unsetenv(DISPLAY) if we're not org.x.X11 137830 Also some prefix related fixes. 137831 (cherry picked from commit fd459e96581a883e30323c840b71004aa0747169) 137832 137833commit ea94995982d1cc585768d52fec75cae018457fa5 137834Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 137835Date: Sat Dec 6 12:01:28 2008 -0800 137836 137837 XQuartz: Removed some debug spew 137838 (cherry picked from commit 370dcf11b360b87aa24cace35eb615419057fda0) 137839 137840commit 86c64ddf21763972aa7fc8c5770259123c9907b3 137841Author: Adam Jackson <ajax@redhat.com> 137842Date: Mon Dec 8 17:42:47 2008 -0500 137843 137844 randr: clear primaryOutput when the output is deleted 137845 137846commit fe65f400ed16cb39db8c9518b9446f590c34db1a 137847Author: Adam Jackson <ajax@redhat.com> 137848Date: Mon Dec 8 17:40:07 2008 -0500 137849 137850 randr: use primary output for RRFirstOutput() 137851 137852commit cdcb516e561e2f65eb2fa523ca001c57674d5caf 137853Author: Adam Jackson <ajax@redhat.com> 137854Date: Mon Dec 8 17:37:17 2008 -0500 137855 137856 randr: Mangle GetScreenResources sort order based on primary output 137857 137858commit d7b316e82bc7051f8829b4f4a640f50ae91c2db9 137859Author: Adam Jackson <ajax@redhat.com> 137860Date: Mon Dec 8 17:28:55 2008 -0500 137861 137862 randr: Mangle compat Xinerama reply based on primary output 137863 137864commit 9d58d2a319059989ccdfa758f586149ccdc16df6 137865Author: Adam Jackson <ajax@redhat.com> 137866Date: Tue Dec 9 10:51:37 2008 -0500 137867 137868 randr: Add [GS]etOutputPrimary 137869 137870commit d281866b74f7067f2704c278fe9720eafc0ee5ef 137871Author: Peter Hutterer <peter.hutterer@redhat.com> 137872Date: Mon Dec 1 21:20:48 2008 +1000 137873 137874 mi: Clean up CopyGetMasterEvent, re-use the memory. 137875 137876 Alloc an EventList once and then re-use instead of allocing a new event each 137877 time we need a master event. 137878 There's a trick included: because all the event processing handlers only take 137879 an xEvent, init a size 1 EventList and squash the events into this one. 137880 137881 Events that have count > 1 must be squished into an xEvent array anyway before 137882 passing into the event handlers, so we don't lose anything here. 137883 137884 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137885 Acked-by: Daniel Stone <daniel@fooishbar.org> 137886 137887commit 0b4fef6337d88ae8ef05b8b73941350a9007565c 137888Author: Peter Hutterer <peter.hutterer@who-t.net> 137889Date: Wed Dec 10 11:35:09 2008 +1000 137890 137891 dix: move MAX_VALUATOR_EVENTS into include/input.h 137892 137893 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137894 137895commit b36c398b11321a908cfe217108b26a32ffc1d850 137896Author: Peter Hutterer <peter.hutterer@who-t.net> 137897Date: Mon Dec 8 16:41:20 2008 +1000 137898 137899 dix: purge dead device-based window access code. 137900 137901 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137902 137903commit aa71ac1dd4697738752c73fe4c106b665ea3ec25 137904Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 137905Date: Wed Dec 10 00:32:34 2008 -0200 137906 137907 Add dependency tracking to sdksyms.c and export composite wrapper. 137908 137909 Thanks to David Miller for noticing a make problem with sdksyms.c 137910 not being regenerated when sdksyms.sh is updated. This is not yet the 137911 best solution; automake generates dependency for sdksyms.o, but the 137912 build really should also regenerate sdksyms.c when sdksyms.o needs to 137913 be regenerated. 137914 Export the symbols in miext/cw/cw.h. These symbols are in libxaa, and 137915 at least the nvidia driver uses them. Maybe cw.h should be installed 137916 in the sdk. 137917 137918commit 090f63123975220819d531f569df6e5787607ec6 137919Author: Peter Hutterer <peter.hutterer@who-t.net> 137920Date: Mon Dec 8 11:53:20 2008 +1000 137921 137922 mi: always update the sprite for master devices. 137923 137924 Follow-up to 9ce995373e4a. This re-enables cursor rendering if the MD is 137925 controlled through software (e.g. synergy). 137926 137927 Reported by John Tapsell: "I use Xorg with no mouse attached, but use synergy 137928 to control the mouse. The commit means that I no longer have a visible mouse 137929 cursor. The mouse cursor is still 'there' in terms that I can click buttons 137930 etc with it, but it's just not visible." 137931 137932 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137933 137934commit 7c8720c1433d2c3b85bbf4b811cc54c2df4c0080 137935Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 137936Date: Mon Dec 8 22:11:50 2008 -0200 137937 137938 Correct wrong symbol reference on sparc. 137939 137940 The awk script was incorrectly referencing the struct name, and 137941 not the struct variable. 137942 Also added some comments to sdksyms.sh, for the reason it generates 137943 the "symbol table" and add a message to the generated file, telling 137944 is was automatically generated. 137945 137946commit 58a27d2932164e43c0db42b1286ec2f95250b420 137947Author: Adam Jackson <ajax@redhat.com> 137948Date: Mon Dec 8 16:28:00 2008 -0500 137949 137950 Default to x86emu even on i386 linux 137951 137952 vm86 mode is a bad idea anyway, and using the emulator everywhere 137953 means we get a consistent set of bugs. 137954 137955commit bbf811514d3cdf84790bad5b852942a4e636902b 137956Author: Sascha Hlusiak <saschahlusiak@arcor.de> 137957Date: Mon Dec 8 12:24:39 2008 +0100 137958 137959 ddxCtrls.c: XkbDDXUsesSoftRepeat always returns 1 now 137960 137961 We'd like to do soft repeat in the server for all keys. Remove obscure check, that'd 137962 prevent the server from autorepeating when delay is set to exactly 660ms and rate is 137963 set to exactly 25 (interval=40). 137964 137965 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 137966 137967commit 8c1dd40a04693f09f4fcea6e7f905af34c7589de 137968Author: Benjamin Close <Benjamin.Close@clearchain.com> 137969Date: Mon Dec 8 14:49:38 2008 +1030 137970 137971 Don't use gnu specific extensions to awk when builing symbols 137972 137973 Traditional posix awk doesn't know about \W and whilst we check that 137974 awk exists in configure.ac we don't check which awk we are using. 137975 This corrects symbol generation for posix only awk. 137976 137977commit 39db182b637041255ed6dac739ff77c8e4e07c30 137978Author: Peter Hutterer <peter.hutterer@who-t.net> 137979Date: Mon Dec 8 12:12:39 2008 +1000 137980 137981 xfree86: init EQ before trying to initialise the devices (#18890) 137982 137983 The kbd driver may send events during device initialisation, and these events 137984 need the EQ set up already. 137985 137986 X.Org Bug 18890 <http://bugs.freedesktop.org/show_bug.cgi?id=18890> 137987 137988 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 137989 137990commit 78a62d7713c708d067d8824ec41b0a0225c1997f 137991Author: Tom Jaeger <ThJaeger@gmail.com> 137992Date: Mon Dec 8 11:38:12 2008 +1000 137993 137994 Xi: XIGetDevice needs to ignore the MORE_EVENTS flag. 137995 137996 Reported in X.Org Bug 18882, Comment 5. 137997 <http://bugs.freedesktop.org/show_bug.cgi?id=18882> 137998 137999 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 138000 138001commit ee1a6c28418a6dad6c89f79a994f27bfbaa77368 138002Author: Peter Hutterer <peter.hutterer@redhat.com> 138003Date: Fri Dec 5 10:12:57 2008 +1000 138004 138005 dix: fix calculation of valuator events. 138006 138007 Follow-up to 4971315296cb. countValuatorEvents was copied from GKVE where it 138008 was obviously broken but nobody noticed. GPE had the correct version, but that 138009 one got lost during de-duplication. Restoring the correct calculation - if we 138010 have 6 valuators, we want 1 valuator event, not 2. 138011 138012 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138013 138014commit 13de7511b17b57a28668e1a60b196ccfe61dbcbe 138015Author: Peter Hutterer <peter.hutterer@redhat.com> 138016Date: Thu Dec 4 10:30:02 2008 +1000 138017 138018 xfree86: Only use the evdev ruleset on linux. 138019 138020 As suggested by Julien Cristau 138021 138022 This is an follow-up to 138023 commit 9c5dd7337fa93fb1650cc017e523b939dcbf482a 138024 Author: Peter Hutterer <peter.hutterer@redhat.com> 138025 Date: Wed Dec 3 14:24:25 2008 +1000 138026 138027 Let the DDX decide on the XkbRulesDefaults. 138028 138029 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138030 Acked-by: Julien Cristau <jcristau@debian.org> 138031 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 138032 138033commit 8e3279134987a45f2a89c963ef2d33bc3d3c8179 138034Author: Peter Hutterer <peter.hutterer@redhat.com> 138035Date: Thu Dec 4 09:47:25 2008 +1000 138036 138037 xfree86: fix compiler warning (use of uninitialized variable) 138038 138039 drv and idev are only set for SDs, but are only dereferenced for SDs too, so 138040 initializing them to NULL is safe. 138041 138042 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138043 138044commit 22d4ddcc3d63b7fbf2a23f5162075e4ee06db781 138045Merge: d2dad384f c3bf15ba8 138046Author: Matthieu Herrb <matthieu@crux.(none)> 138047Date: Sun Dec 7 23:56:15 2008 +0100 138048 138049 Merge branch 'master' of git://anongit.freedesktop.org/git/xorg/xserver 138050 138051commit d2dad384f5725a36e593fd75f03d712f53b20620 138052Author: Matthieu Herrb <matthieu@crux.(none)> 138053Date: Sun Dec 7 23:52:26 2008 +0100 138054 138055 Fix build in separate build directory. 138056 138057commit c3bf15ba85fd3ab2e1fe809428b75bf34db25cc8 138058Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138059Date: Sun Dec 7 18:52:23 2008 -0200 138060 138061 Require macros 1.2.0 or newer for XORG_CHANGELOG and XORG_CWARNFLAGS. 138062 138063commit 5a8068b390b7d30eb526e954443d8e43a199b971 138064Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138065Date: Sun Dec 7 18:31:32 2008 -0200 138066 138067 Convert libx86emu.a to a "libtool convenience library". 138068 138069commit ed4a17274015ecd8040ae85bd5cd9d1dbcc9460a 138070Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138071Date: Sun Dec 7 14:59:25 2008 -0200 138072 138073 Export some symbols from libxf86config when installing it. 138074 138075 These are private symbols, but used by the X Server. 138076 The newly exported symbols were not added to the sdk headers. 138077 Optionally, libxf86config could be compiled without hidden symbols 138078 when being installed. 138079 Thanks to Maarten Maathuis for noticing the problem. 138080 138081commit bf4c29ab48a166eb158cf4be7c597982d65ef214 138082Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138083Date: Sun Dec 7 05:12:41 2008 -0200 138084 138085 Correct xf86acpiDisableFlag symbol. 138086 138087 It is declared as 138088 <hash>ifdef HAVE_ACPI 138089 Bool xf86acpiDisableFlag = FALSE; 138090 <hash>endif 138091 in hw/xfree86/common/xf86Globals.c 138092 but not protected by the ifdef in the sdk header xf86Priv.h, what 138093 caused a build failure in the tinderbox, due to the address of the 138094 symbol being taken (to ensure it is available) in sdksyms.c. 138095 138096commit b1dac41fb3853ca8182048ea57b88b6e84ecceb3 138097Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138098Date: Sun Dec 7 02:22:19 2008 -0200 138099 138100 Use libtool convenience libraries and better "symbol" table. 138101 138102 All .a libraries were converted to .la, and instead of linking the 138103 Xorg binary with a mix of .a and .la, and adding some libraries more 138104 then once in the command line, etc, now it generates a single libxorg.la 138105 from all the required convenience libraries, and links with a dummy 138106 xorg.c (that should usually be the file with the main function...). 138107 This removes the requirement of some things like libosandcommon and 138108 libinit, that existed to circumvent problems when linking multiple 138109 .a and .la in the final Xorg binary. 138110 138111 The "symbol table" is now generated dynamically, by a shell script, 138112 with an embedded gawk parser that parses cpp output. The new file 138113 sdksyms.sh is generated by hand by analyzing all Makefile.am's and 138114 making it create a sdksyms.c file, that includes all sdk headers that 138115 will add symbols for the Xorg binary. Module headers aren't read, and 138116 a in 2 files it was required to add a "<hash>ifndef XorgLoader" around 138117 declarations shared between the Xorg binary and libextmod. A few 138118 other changes were added to other sdk headers, like preventing 138119 multiple inclusion, or including other headers to satisfy dependencies. 138120 138121 This should be a lot more portable, and better (hopefully properly) 138122 using libtool to generate convenience libraries. 138123 138124commit ccd2c668c13863ed704f86b29fc6b3ee628e56a2 138125Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138126Date: Sat Dec 6 14:11:52 2008 -0800 138127 138128 XQuartz: darwinPointer reports the actual pixel position now rather than a relative position 138129 (cherry picked from commit a41e7f75decd340d064fdc0d2c4fe6c88d7dbc82) 138130 138131commit 7ebd0c7e8d42a13079957d9bacf5cb30d7855a59 138132Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138133Date: Fri Dec 5 12:42:31 2008 -0800 138134 138135 XQuartz: Corrected name/command labels in the customization widget 138136 (cherry picked from commit a689c23f17eb445c36b97eb617ef4b8bd157985f) 138137 138138commit 027ff97a1354ab4c83fecb615f6bc2a6b739b871 138139Author: Alan Coopersmith <alan.coopersmith@sun.com> 138140Date: Fri Dec 5 12:00:08 2008 -0800 138141 138142 Make sure _X_EXPORT is defined in edid.h 138143 138144 Fixes build error in xf86-video-intel utilities 138145 138146commit 466b0fca9ba5d5e7fb36e47a6d1bd60218f51c75 138147Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138148Date: Fri Dec 5 17:01:09 2008 -0200 138149 138150 Add back a simplified version of the loader static address tables. 138151 138152 If not taking the symbol addresses, linkage will break badly, as not 138153 all symbols will be present, and it also requires changing library order, 138154 and/or making some changes like the "libosandcommon". 138155 138156 This table should be modified to be generated automatically, as 138157 it is required to "fool" the compiler/loader into adding all required 138158 symbols to the X Server. 138159 138160commit 44bef8b850c5a78a3d3eca5f0d92b71bdd0a87e2 138161Author: Matthias Hopf <mhopf@suse.de> 138162Date: Fri Dec 5 15:37:15 2008 +0100 138163 138164 randr: Update SProcRandrVector for panning 138165 138166commit 44c8c3cf7de589fb8e987f4ab931294eaf0b405f 138167Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138168Date: Fri Dec 5 04:26:57 2008 -0200 138169 138170 Remove static symbol address tables in hw/xfree86/loader/*sym.{c,h} 138171 138172 Those tables were once used to decide what symbols are visible to 138173 modules, but they were outdated. The only real usage was that, since 138174 it was taking the address of symbols, linkage should fail if the 138175 symbols were not available. 138176 Now the proper way to make symbols available to modules should 138177 be to use the _X_EXPORT macro, or not compile with hidden symbols, 138178 so that all symbols would be available. 138179 All symbols in the tables were revised to ensure they are exported, 138180 and only symbols that were not exported are ClientSleepUntil() and 138181 DuplicateModule(), that were not in the sdk for quite some time 138182 already, and should not have any users outside of the X Server 138183 (and/or builtin modules). 138184 138185commit 091a50d1260c70055aba1fa6a2d1b1a36de1114e 138186Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138187Date: Thu Dec 4 18:23:25 2008 -0200 138188 138189 Move _X_EXPORT attribute to header file. 138190 138191 The attribute should be set on header files to make it easier to 138192 manage what symbols are available to modules. 138193 _X_EXPORT should be used in sources only for special cases, like 138194 symbols that must be visible by non video/input driver/modules but 138195 should not be "advertised" in the sdk. 138196 138197commit e5ab9e66628cde081757cf2a1013a78e927a622e 138198Author: Matthias Hopf <mhopf@suse.de> 138199Date: Thu Dec 4 18:13:40 2008 +0100 138200 138201 randr: Allow panning to be disabled per axis 138202 138203commit bad118ace6c5bae5a5ed8a35129c90c38f1c1932 138204Author: Matthias Hopf <mhopf@suse.de> 138205Date: Thu Dec 4 16:55:14 2008 +0100 138206 138207 randr: Rework panning area verification 138208 138209commit 219c26ce0c65625d55cfd943ec66fe94a1a0ddfd 138210Author: Matthias Hopf <mhopf@suse.de> 138211Date: Thu Dec 4 16:28:40 2008 +0100 138212 138213 randr: Don't change panning parameters if verification fails. 138214 138215commit 18a8bac1a1567b6215928f96870554ea63f39aab 138216Author: Matthias Hopf <mhopf@suse.de> 138217Date: Thu Dec 4 16:30:38 2008 +0100 138218 138219 randr: Rename pan() to set_origin(), and xf86CrtcPan() to xf86CrtcSetOrigin() 138220 138221commit 825b2c2f4a59ac4852f90bbbddf18ab832297fdd 138222Author: Matthias Hopf <mhopf@suse.de> 138223Date: Thu Dec 4 16:11:21 2008 +0100 138224 138225 randr: Nuke config-timestamp for panning 138226 138227commit eeeb98d1df59baaaec954b6318d788a37e388d11 138228Author: Matthias Hopf <mhopf@suse.de> 138229Date: Fri Nov 28 17:51:20 2008 +0100 138230 138231 randr: Protocol bits for panning support 138232 138233commit b929d721efdb17bcc94b9984c4f34d0df3d267d5 138234Author: Matthias Hopf <mhopf@suse.de> 138235Date: Fri Nov 28 17:49:31 2008 +0100 138236 138237 randr: Panning support 138238 138239commit 834cbc16f3eb539704faade7bff347b161ce69d9 138240Author: Matthias Hopf <mhopf@suse.de> 138241Date: Fri Nov 28 17:39:23 2008 +0100 138242 138243 randr: Crtc interface update for panning support. 138244 138245commit a475eb9feec75e9ce1e316da0f1679acd7dd3aa8 138246Author: Matthias Hopf <mhopf@suse.de> 138247Date: Fri Nov 28 17:38:52 2008 +0100 138248 138249 randr: Weird enough, crtc->version was never set upon creation. Fix that. 138250 138251commit 9db84fac0cc767e23986223d22bc085cde0cc86f 138252Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138253Date: Tue Dec 2 14:54:10 2008 -0800 138254 138255 Removed unised option from configure.ac for launchd 138256 (cherry picked from commit 87e4de0a9a8af6e5e0cbba74e585761f1160a6ab) 138257 138258commit fa4f2527a5002711fe47c02b14097032fd845dc4 138259Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138260Date: Tue Dec 2 14:53:21 2008 -0800 138261 138262 XQuartz: pbproxy: Simplify linking 138263 (cherry picked from commit 909cc5c4dca0f63e90505575bbd454b46a4670cc) 138264 138265commit 126d8a4c8a72b039903023dce0da64b251080e1d 138266Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138267Date: Thu Dec 4 02:43:29 2008 -0200 138268 138269 Update .gitignore. 138270 138271 Add files left by make distcheck and remove files that no longer 138272 are generated. 138273 138274commit 9826b95c198e74f89680cc247b9bd9dd884d72be 138275Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138276Date: Thu Dec 4 02:05:05 2008 -0200 138277 138278 Correct make distcheck by removing dolt files. 138279 138280commit 5dbe70dd52fd8daf8d0797951cc5a758d2c3b44c 138281Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138282Date: Thu Dec 4 01:57:46 2008 -0200 138283 138284 Remove dummylib. 138285 138286 Only cvt required it, and only used XNF{,c}alloc 138287 via xnf{,c}alloc macros. 138288 Based on patch by Eric Anholt. 138289 138290commit f841d4e3cccbde02e91c948f5ffb9e32c8c3b3cc 138291Author: Julien Cristau <jcristau@debian.org> 138292Date: Wed Dec 3 23:02:03 2008 +0100 138293 138294 Move RELEASE_DATE closer to AC_INIT so it's more likely to be updated 138295 138296commit 75b02dd0be2ef64c0f8e3138c90b5c5e48b17913 138297Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138298Date: Wed Dec 3 14:46:30 2008 -0200 138299 138300 Include <X11/Xfuncproto.h> if _X_EXPORT is not defined. 138301 138302commit dd128ddcdcbe254a9cdd973590f6a979a7f0427e 138303Author: Timo Aaltonen <tjaalton@cc.hut.fi> 138304Date: Wed Dec 3 18:40:29 2008 +0200 138305 138306 If AEI is on, disable 'vmmouse' in addition to 'kbd' and 'mouse'. 138307 138308 Signed-off-by: Timo Aaltonen <tjaalton@cc.hut.fi> 138309 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138310 138311commit 0b5ecabfb803cd820338fb0364521fe39b05578b 138312Author: Julien Cristau <jcristau@debian.org> 138313Date: Wed Nov 26 21:19:55 2008 +0100 138314 138315 randr: add swapped dispatch for RR[GS]etCrtcTransform 138316 138317 Fix a memory leak in ProcRRGetCrtcTransform() while I'm at it. 138318 138319 Signed-off-by: Julien Cristau <jcristau@debian.org> 138320 Cc: Keith Packard <keithp@keithp.com> 138321 138322commit 110a71d11ab7a1a55a6a24d792457fdef0b0746d 138323Author: Kristian Høgsberg <krh@redhat.com> 138324Date: Wed Dec 3 11:22:38 2008 -0500 138325 138326 Test for DRI2 extension in dri_internal.h and only enable AIGLX DRI2 if found. 138327 138328commit fd2d40b7ec5d685dac55453eb1f2da672dc83126 138329Author: Peter Hutterer <peter.hutterer@redhat.com> 138330Date: Wed Dec 3 11:31:48 2008 +1000 138331 138332 Xi: change XIUnRegisterPropertyHandler to XIUnregisterPropertyHandler 138333 138334 CamelCase can be taken too far, and AFAICT there's no consumers of that 138335 function yet anyway. 138336 138337 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138338 138339commit 49f77fff1495c0a2050fb18f9b1fc627839bbfc2 138340Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138341Date: Wed Dec 3 05:43:34 2008 -0200 138342 138343 Rework symbol visibility for easier maintenance 138344 138345 Save in a few special cases, _X_EXPORT should not be used in C source 138346 files. Instead, it should be used in headers, and the proper C source 138347 include that header. Some special cases are symbols that need to be 138348 shared between modules, but not expected to be used by external drivers, 138349 and symbols that are accessible via LoaderSymbol/dlopen. 138350 138351 This patch also adds conditionally some new sdk header files, depending 138352 on extensions enabled. These files were added to match pattern for 138353 other extensions/modules, that is, have the headers "deciding" symbol 138354 visibility in the sdk. These headers are: 138355 o Xext/panoramiXsrv.h, Xext/panoramiX.h 138356 o fbpict.h (unconditionally) 138357 o vidmodeproc.h 138358 o mioverlay.h (unconditionally, used only by xaa) 138359 o xfixes.h (unconditionally, symbols required by dri2) 138360 138361 LoaderSymbol and similar functions now don't have different prototypes, 138362 in loaderProcs.h and xf86Module.h, so that both headers can be included, 138363 without the need of defining IN_LOADER. 138364 138365 xf86NewInputDevice() device prototype readded to xf86Xinput.h, but 138366 not exported (and with a comment about it). 138367 138368commit 0b8f8b24f718820a72ebdc52423c2e6a44e848c5 138369Author: Stuart Bennett <sb476@cam.ac.uk> 138370Date: Tue Dec 2 22:52:53 2008 -0800 138371 138372 xf86Cursors: xf86_reload_cursors shouldn't unconditionally show hwcursor (#14820) 138373 138374 Also, no need to call ShowCursor when SetCursorPosition already does it 138375 Based on a previous patch by Maarten Maathuis 138376 138377 Signed-off-by: Keith Packard <keithp@keithp.com> 138378 138379commit 9c5dd7337fa93fb1650cc017e523b939dcbf482a 138380Author: Peter Hutterer <peter.hutterer@redhat.com> 138381Date: Wed Dec 3 14:24:25 2008 +1000 138382 138383 Let the DDX decide on the XkbRulesDefaults. 138384 138385 Rather than assuming rules in the CoreKeyboardProc, init the default rules in 138386 InitCoreDevices, then re-use them later. 138387 138388 In the xfree86 DDX, set the rules to "base" or "evdev", depending on whether 138389 we'll load kbd or evdev. 138390 138391 If we create a new MD, use pc105,us as default and re-use the rules file used 138392 previously. 138393 138394 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138395 138396commit 463e02e7de5da3e582a3a049110a476713c7210e 138397Author: Peter Hutterer <peter.hutterer@redhat.com> 138398Date: Wed Dec 3 15:06:37 2008 +1000 138399 138400 xkb: Allow NULL as rulesFile in XkbSetRulesDflts. 138401 138402 If no rules file is given, simply re-use the previous one. If no RF is given 138403 the first time this function is called, use the built-in default. 138404 This includes fixing the built-in default to something that actually exists. 138405 138406 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138407 138408commit e670fd889607fa712876218882cd4a9b46937661 138409Author: Peter Hutterer <peter.hutterer@redhat.com> 138410Date: Wed Dec 3 11:55:13 2008 +1000 138411 138412 dix: fix GetMaximumEventsNum(), may return a DCCE event too. 138413 138414 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138415 138416commit 85d84c7cf2d368d1803dc2cdd03d6c9df8cc0430 138417Author: Eric Anholt <eric@anholt.net> 138418Date: Tue Dec 2 17:33:43 2008 -0800 138419 138420 Fix GLX after 180bad84774493d48f2793a6281d825560944863. 138421 138422 Sigh. 138423 138424commit 4a61ad427c634bcc38a31ce0f14ff5c2d3f706c7 138425Author: Alan Coopersmith <alan.coopersmith@sun.com> 138426Date: Tue Dec 2 14:35:45 2008 -0800 138427 138428 Update sample xorg.conf file 138429 138430commit c1db925d10fd37077bed90612ed95c3fd20cd2e2 138431Author: Alan Coopersmith <alan.coopersmith@sun.com> 138432Date: Tue Dec 2 14:32:48 2008 -0800 138433 138434 Add atKeynames.h to libdmxinput_a_SOURCES so it's included in tarballs 138435 138436 Reported by geearu in #xorg-devel 138437 138438commit 8561514574b3540c729bcc3acca9c943adcdc778 138439Author: Alan Coopersmith <alan.coopersmith@sun.com> 138440Date: Tue Dec 2 13:32:34 2008 -0800 138441 138442 Don't need to check uid/euid for every commandline argument 138443 138444 Check uid/euid only when handling the arguments that are restricted 138445 to root/non-setuid users 138446 138447commit 6de6ffff35ac03d49fa61de195d4a0605e0ef8bf 138448Author: Alan Coopersmith <alan.coopersmith@sun.com> 138449Date: Tue Dec 2 13:19:08 2008 -0800 138450 138451 Remove unused config flags from FlagValues & FlagOptions 138452 138453 FLAG_DISABLEMODINDEV, FLAG_MODINDEVALLOWNONLOCAL, & FLAG_HANDLE_SPECIAL_KEYS 138454 138455commit bb072019fa8dd292a50ef433d05caeefd1304a73 138456Author: Peter Hutterer <peter.hutterer@redhat.com> 138457Date: Mon Dec 1 16:57:06 2008 +1000 138458 138459 xfree86: don't render SW cursors for devices attached to VCP (#16805) 138460 138461 When leaving 3D games such as quake3 or sauerbraten, a cursor may stay on the 138462 screen. This is caused by one run of SW rendering for the SD, even though the 138463 SD was attached to the VCP and thus has HW rendering capabilities. 138464 138465 Check for the SD's attachment (like in all other functions) before deciding on 138466 SW or HW rendering. 138467 138468 X.Org Bug 16805 <http://bugs.freedesktop.org/show_bug.cgi?id=16805> 138469 138470 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138471 138472commit 3a690598cf18c4cdc6aadd10a1ecf0772cacd34b 138473Author: Alan Coopersmith <alan.coopersmith@sun.com> 138474Date: Tue Dec 2 12:59:57 2008 -0800 138475 138476 Remove unused HandleSpecialKeys config option 138477 138478 Was only used by "keyboard" driver - the last place the value was checked 138479 was in xf86PostKbdEvent, which was removed in commit 138480 60ea7b51fe2b8a19a08b63db48504971a7a50ee6 138481 138482commit 1dfed222e93f4684c2a450944a9a0ea9e085c43f 138483Author: Bernhard Rosenkraenzer <bero@arklinux.org> 138484Date: Tue Dec 2 09:01:56 2008 +1000 138485 138486 Xext: fix MultiBuffer compilation error with TryClientEvents. (#18835) 138487 138488 X.Org Bug 18835 <http://bugs.freedesktop.org/show_bug.cgi?id=18835> 138489 138490 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138491 138492commit 9ce995373e4aa4b1bf51b4adafa2324f781ec1cd 138493Author: Peter Hutterer <peter.hutterer@redhat.com> 138494Date: Mon Dec 1 21:32:39 2008 +1000 138495 138496 mi: UpdateSprite only if the device is attached. 138497 138498commit d507f60689f4e14383b0d24e63afc8cf836360d5 138499Author: Peter Hutterer <peter.hutterer@redhat.com> 138500Date: Wed Nov 26 14:15:04 2008 +1000 138501 138502 xfree86: don't FatalError on "too many input devices". 138503 138504 Just ignore devices after MAXDEVICES has been reached, but warn the user that 138505 the devices are ignored. 138506 138507 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138508 138509commit 9bf761468fa3baf4c5fa40efa717b4b581d920f9 138510Author: Peter Hutterer <peter.hutterer@redhat.com> 138511Date: Fri Nov 28 11:44:58 2008 +1000 138512 138513 Xi: fix xi_filters size. 138514 138515commit 260e48c252c95fd1b1ba9ad7478791f6b9a67e2d 138516Author: Peter Hutterer <peter.hutterer@redhat.com> 138517Date: Fri Nov 28 11:29:06 2008 +1000 138518 138519 dix: remove confusing (and wrong) comment. VCP is not the only pointer. 138520 138521 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138522 138523commit 5d02e580d7c4144ea6f0984240fc913e05fec877 138524Author: Peter Hutterer <peter.hutterer@redhat.com> 138525Date: Fri Nov 28 11:46:15 2008 +1000 138526 138527 dix: use UpdateFromMaster in GetProximityEvents. 138528 138529 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138530 138531commit 95fc59a199f99bf167fbb09297a9bb0e33e31869 138532Author: Peter Hutterer <peter.hutterer@redhat.com> 138533Date: Fri Nov 28 14:55:15 2008 +1000 138534 138535 xkb: Extra sanity checks to prevent dev->key == NULL dereferencing. 138536 138537commit a425abf0eaa61ee5ccb2f32a1af24edd190f7889 138538Author: Peter Hutterer <peter.hutterer@redhat.com> 138539Date: Wed Nov 26 12:20:00 2008 +1000 138540 138541 xkb: don't attempt to filter events for devices without key classes. 138542 138543 Reported by Magnus Kessler. 138544 138545 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138546 138547commit a85f0d6b98237d8a196de624207acf1983a1859a 138548Author: Peter Hutterer <peter.hutterer@redhat.com> 138549Date: Wed Nov 26 11:15:05 2008 +1000 138550 138551 Xi: fix use of button->down - bitflags instead of int arrays. 138552 138553 The device's button down state array was changed to use DOWN_LENGTH and thus 138554 bitflags for each button in cfcb3da7. 138555 138556 Update the DBSN events to copy this bit-wise state. 138557 Update xkb and Xi to check for the bit flag instead of the array value. 138558 138559 Reported by ajax. 138560 138561 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 138562 138563commit 180bad84774493d48f2793a6281d825560944863 138564Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138565Date: Tue Dec 2 02:50:45 2008 -0200 138566 138567 Add visibility flags to XSERVER_CFLAGS. 138568 138569 This is done to actually change DIX_CFLAGS, as not all "modules" use 138570 XORG_CFLAGS. 138571 Also export the symbols that are required by other modules after 138572 the change. 138573 138574commit 8c560422b44e012053612754430d2b87dc44ed59 138575Author: Alan Coopersmith <alan.coopersmith@sun.com> 138576Date: Mon Dec 1 20:02:34 2008 -0800 138577 138578 More man page updates for 1.6 release for Xorg, xorg.conf & exa man pages. 138579 138580 - Remove remaining references to XFree86-Misc options AllowNonLocalModInDev 138581 and DisableModInDev. 138582 - Remove remaining references to grab-breaking keys & associated options. 138583 - Update description of Ctrl-Alt-Backspace to new -retro/DontZap defaults. 138584 - Add description of new options -modalias and -showopts. 138585 - Update list of modules loaded by default. 138586 - Update input driver references from keyboard to evdev & kbd. 138587 - Update list of driver man pages to match xf86-*-* drivers with man pages. 138588 - Add See Also section to exa man page. 138589 138590 and various formatting/typo/etc. fixes. 138591 138592 The Xorg/xorg.conf sections on input device selection could use further 138593 updates to better match the current state of HAL-enabled configuration. 138594 138595commit 6a1850b8c677e2a2993f6f6b731ee3d35aa55d09 138596Author: Alan Coopersmith <alan.coopersmith@sun.com> 138597Date: Mon Dec 1 19:19:04 2008 -0800 138598 138599 Correct warning for unknown GlxVisuals option in conf file 138600 138601commit afeacb8e74b2a1e366e5ca9f0e58805d8d96c457 138602Author: Kristian Høgsberg <krh@redhat.com> 138603Date: Mon Dec 1 20:53:17 2008 -0500 138604 138605 Bump dri2proto requirement to 1.99.3. 138606 138607commit ad01e86b5c7c528adec8a1f95ecaa294f58a8922 138608Author: Kristian Høgsberg <krh@redhat.com> 138609Date: Mon Dec 1 20:52:41 2008 -0500 138610 138611 Drop unused DRI2 vblank infrastructure. 138612 138613 For this first iteration of DRI2 we don't have any vsync functionality 138614 in place yet, so back out the support in DRI2 and the protocol for now. 138615 138616commit 24c562f04b41d219c34f5fa3f963564accf329f2 138617Author: Alan Coopersmith <alan.coopersmith@sun.com> 138618Date: Mon Dec 1 16:28:45 2008 -0800 138619 138620 Update See Also lists in Xorg & xorg.conf man pages 138621 138622 Remove xorgconfig & xorgcfg 138623 Update driver lists: 138624 - Remove vga 138625 - i810->intel 138626 - via->openchrome 138627 138628commit bd147e6f4b03f344c967c88fd433877b14fe1912 138629Author: Alan Coopersmith <alan.coopersmith@sun.com> 138630Date: Mon Dec 1 16:08:55 2008 -0800 138631 138632 Remove hack to only load font modules if magic strings appear in font path 138633 138634commit 5f3f14179edf48aad518f6f707bfdc37c27267c6 138635Author: Jay Cotton <jay.cotton@sun.com> 138636Date: Mon Dec 1 16:06:09 2008 -0800 138637 138638 Sun bug 6618220: Xorg server core dump in xf86RandRModeRefresh(NULL) 138639 138640 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6618220> 138641 138642commit 09df7cc5ad7b72d8a23c3e22fc718aad8c16f4d3 138643Author: Kristian Høgsberg <krh@redhat.com> 138644Date: Mon Dec 1 12:41:10 2008 -0500 138645 138646 Avoid dereferencing NULL pScreen in xf86CrtcSetModeTransform(). 138647 138648 We can get there during PreInit as we set a mode for load detection. 138649 At that time there's no pScreen anywhere, so just skip the optimization 138650 then. 138651 138652commit b0d371ab0a6efd4956c3677faa20b2ac15c33765 138653Author: Adam Jackson <ajax@redhat.com> 138654Date: Mon Dec 1 11:36:06 2008 -0500 138655 138656 randr: Don't send output property events on server exit 138657 138658 If the Window resource type is already gone, there's no point in trying 138659 to send events, all it can do is access already-freed memory. 138660 138661 Relevant thread: 138662 138663 http://lists.freedesktop.org/archives/xorg/2008-November/040443.html 138664 138665commit 16b11cd03d8c5def07f0e598f237f71a37883a46 138666Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138667Date: Sun Nov 30 02:59:34 2008 -0200 138668 138669 Correct static symbol XkmReadTOC and first pass on compile warning fixes. 138670 138671 The warnings corrected were only the ones that should correct 138672 real problems. The most common one is 64 bit integers as 138673 "printf %l" arguments. 138674 Note that there is a patch related to this at: 138675 http://bugs.freedesktop.org/show_bug.cgi?id=18204 138676 138677commit fb22d4d928bc5d1a893494a059359da2ed2ac639 138678Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138679Date: Sun Nov 30 01:33:20 2008 -0200 138680 138681 Remove declarations of symbols that are never defined. 138682 138683 These symbols were removed from the X Server, or never declared. 138684 One symbol that may need special attention is XkbBuildCoreState(), 138685 that doesn't have a prototype anywhere, but is called from 138686 xkb/xkbEvents.c:XkbFilterEvents(), and also used by the macros 138687 XkbStateFieldFromRec() and XkbGrabStateFromRec() defined in 138688 include/xkbstr.h. 138689 fb/wfbrename.h also may need some cleanup, as it makes several 138690 "renames" of non existing symbols. 138691 138692commit d6cbd4511e35a89a0353f11834c6fdb8d4d2189f 138693Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138694Date: Sat Nov 29 23:56:06 2008 -0200 138695 138696 Export symbols defined in the sdk. 138697 138698 This is the biggest "visibility" patch. Instead of doing a "export" 138699 symbol on demand, export everything in the sdk, so that if some module 138700 fails due to an unresolved symbol, it is because it is using a symbol 138701 not in the sdk. 138702 138703 Most exported symbols shouldn't really be made visible, neither 138704 advertised in the sdk, as they are only used by a single shared object. 138705 138706 Symbols in the sdk (or referenced in sdk macros), but not defined 138707 anywhere include: 138708 XkbBuildCoreState() 138709 XkbInitialMap 138710 XkbXIUnsupported 138711 XkbCheckActionVMods() 138712 XkbSendCompatNotify() 138713 XkbDDXFakePointerButton() 138714 XkbDDXApplyConfig() 138715 _XkbStrCaseCmp() 138716 _XkbErrMessages[] 138717 _XkbErrCode 138718 _XkbErrLocation 138719 _XkbErrData 138720 XkbAccessXDetailText() 138721 XkbNKNDetailMaskText() 138722 XkbLookupGroupAndLevel() 138723 XkbInitAtoms() 138724 XkbGetOrderedDrawables() 138725 XkbFreeOrderedDrawables() 138726 XkbConvertXkbComponents() 138727 XkbWriteXKBSemantics() 138728 XkbWriteXKBLayout() 138729 XkbWriteXKBKeymap() 138730 XkbWriteXKBFile() 138731 XkbWriteCFile() 138732 XkbWriteXKMFile() 138733 XkbWriteToServer() 138734 XkbMergeFile() 138735 XkmFindTOCEntry() 138736 XkmReadFileSection() 138737 XkmReadFileSectionName() 138738 InitExtInput() 138739 xf86CheckButton() 138740 xf86SwitchCoreDevice() 138741 RamDacSetGamma() 138742 RamDacRestoreDACValues() 138743 xf86Bpp 138744 xf86ConfigPix24 138745 xf86MouseCflags[] 138746 xf86SupportedMouseTypes[] 138747 xf86NumMouseTypes 138748 xf86ChangeBusIndex() 138749 xf86EntityEnter() 138750 xf86EntityLeave() 138751 xf86WrapperInit() 138752 xf86RingBell() 138753 xf86findOptionBoolean() 138754 xf86debugListOptions() 138755 LoadSubModuleLocal() 138756 LoaderSymbolLocal() 138757 getInt10Rec() 138758 xf86CurrentScreen 138759 xf86ReallocatePciResources() 138760 xf86NewSerialNumber() 138761 xf86RandRSetInitialMode() 138762 fbCompositeSolidMask_nx1xn 138763 fbCompositeSolidMask_nx8888x0565C 138764 fbCompositeSolidMask_nx8888x8888C 138765 fbCompositeSolidMask_nx8x0565 138766 fbCompositeSolidMask_nx8x0888 138767 fbCompositeSolidMask_nx8x8888 138768 fbCompositeSrc_0565x0565 138769 fbCompositeSrc_8888x0565 138770 fbCompositeSrc_8888x0888 138771 fbCompositeSrc_8888x8888 138772 fbCompositeSrcAdd_1000x1000 138773 fbCompositeSrcAdd_8000x8000 138774 fbCompositeSrcAdd_8888x8888 138775 fbGeneration 138776 fbIn 138777 fbOver 138778 fbOver24 138779 fbOverlayGeneration 138780 fbRasterizeEdges 138781 fbRestoreAreas 138782 fbSaveAreas 138783 composeFunctions 138784 VBEBuildVbeModeList() 138785 VBECalcVbeModeIndex() 138786 TIramdac3030CalculateMNPForClock() 138787 shadowBufPtr 138788 shadowFindBuf() 138789 miRRGetScreenInfo() 138790 RRSetScreenConfig() 138791 RRModePruneUnused() 138792 PixmanImageFromPicture() 138793 extern int miPointerGetMotionEvents() 138794 miClipPicture() 138795 miRasterizeTriangle() 138796 fbPush1toN() 138797 fbInitializeBackingStore() 138798 ddxBeforeReset() 138799 SetupSprite() 138800 InitSprite() 138801 DGADeliverEvent() 138802 138803 SPECIAL CASES 138804 o defined as _X_INTERNAL 138805 xf86NewInputDevice() 138806 o defined as static 138807 fbGCPrivateKey 138808 fbOverlayScreenPrivateKey 138809 fbScreenPrivateKey 138810 fbWinPrivateKey 138811 o defined in libXfont.so, but declared in xorg/dixfont.h 138812 GetGlyphs() 138813 QueryGlyphExtents() 138814 QueryTextExtents() 138815 ParseGlyphCachingMode() 138816 InitGlyphCaching() 138817 SetGlyphCachingMode() 138818 138819commit ffb484f7ef84099019b196ef97bfb2355eb6d52a 138820Author: Maarten Maathuis <madman2003@gmail.com> 138821Date: Sat Nov 29 14:30:55 2008 +0100 138822 138823 randr: Avoid needlessly creating a shadow framebuffer. 138824 138825 - pScreen->width and height were zero, so it didn't "fit" the screen. 138826 138827commit 01eaebdc98f30fdf543af6337cdf012d3ff16f09 138828Author: George Staplin <gstaplin@apple.com> 138829Date: Fri Nov 28 13:57:45 2008 -0700 138830 138831 XQuartz: GL: Make various changes to makeFormat, so that it works better. Now glxgears looks normal with the old libGL. 138832 138833 Add various GLX extensions to the list of supported extensions. 138834 138835 Reformat the oddly formatted code in some areas. 138836 138837 Use xalloc and xfree instead of malloc and free. 138838 (cherry picked from commit 755f53eb40c4329d8c82a31cb910c31fdd4ea12e) 138839 138840commit 3d527b91b4bfa31e78d5fc7a1447a4026876f14e 138841Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138842Date: Fri Nov 28 12:59:09 2008 -0800 138843 138844 XQuartz: Added option to enable/disable test extensions 138845 138846 defaults write org.x.X11 enable_test_extensions -boolean true 138847 (cherry picked from commit fd4710aff3723d5f3422cf6a6530172eafc556d9) 138848 (cherry picked from commit 635019ad18db921fec99256294debd8571074abf) 138849 138850commit 71d3df7cc37403b5842227a035d9b995fd920e0b 138851Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138852Date: Fri Nov 28 11:18:14 2008 -0800 138853 138854 XQuartz: Avoid some warning messages being spewed to system.log by AppKit 138855 (cherry picked from commit 780eff230ce41ad785e54d61a82c731269ae0446) 138856 138857commit a9796c7bc43223df44f12a7be08e361aea963ec1 138858Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138859Date: Fri Nov 28 01:45:37 2008 -0800 138860 138861 XQuarz: Setup our PATH and PWD earlier, so our initial client benefits from it as well... 138862 (cherry picked from commit cbae2b447357b3fed6ff19414c60dd3792045600) 138863 138864commit ef4179f43a84d90f867b95ee5974ad26884253b5 138865Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138866Date: Wed Nov 26 12:15:58 2008 -0800 138867 138868 XQuartz: Fix dead-acute on Greek keyboards 138869 (cherry picked from commit 807f2ec35c5f95b56b3e2b5eac51aec300fe5cb6) 138870 138871commit 800f5c681bed300d8caf99935e8f80ea6c1993ba 138872Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138873Date: Wed Nov 26 12:07:52 2008 -0800 138874 138875 XQuartz: Fix Czech keyboard dead-acute 138876 (cherry picked from commit 771df2786bc60389489f0967e705c7c95bdda085) 138877 138878commit ac57bb36d56a7a4d41add8d5a206ff37544a1819 138879Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 138880Date: Tue Nov 25 20:14:37 2008 -0800 138881 138882 XQuartz: Fix an uninitialized keyboard_type on Tiger 138883 (cherry picked from commit 27b1a5eb3482052253ebdce1a09aedf05ac1b099) 138884 138885commit 45c8bd0fe54273039fdaa1eeeafb81b5774f2c75 138886Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138887Date: Fri Nov 28 02:01:57 2008 -0200 138888 138889 Enable compiling the X Server and modules with hidden symbols by default. 138890 138891 Note that it checks if support for symbol visibility is available by 138892 the compiler, not if it is functional. It may have problems on non x86 138893 architectures. 138894 To disable the feature, just pass the option "--disable-visibility" 138895 to the X Server configure script. 138896 Unless using an alternate build schema, drivers built from a git 138897 checkout will use the same "visibility" related CFLAGS used to compile 138898 the X Server. 138899 138900commit 31285d063ec4623cb0764437d6d57e244f20460d 138901Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138902Date: Fri Nov 28 01:55:11 2008 -0200 138903 138904 Make visible symbols required by xorg modules. 138905 138906 This patch exports all symbols required by the compilable 138907 (in a x86 linux computer) xorg/driver/* modules. 138908 Still missing symbols worth mentioning are: 138909 138910 sunleo 138911 miFindMaxBand no longer available 138912 138913 intel (uxa/uxa-accel.c) 138914 fbShmPutImage no longer available (and should have been static) 138915 138916 mga 138917 MGAGetClientPointer (should come from matrox's libhal) 138918 138919 This is not a definitive "visibility" patch, as all it does is to 138920 export missing symbols, but the modules that current don't compile, 138921 may require more symbols once fixed, and third party drivers should 138922 also require more symbols exported. 138923 A "definitive" patch should export symbols defined in the sdk. 138924 138925commit 6f18ea7e656b79e58f66f51067334414767b413f 138926Author: Julien Cristau <jcristau@debian.org> 138927Date: Wed Nov 26 19:07:04 2008 +0100 138928 138929 Typo fix 138930 138931commit efefc03f9b26738b69abb570c1af2167cc99c4d8 138932Author: Julien Cristau <jcristau@debian.org> 138933Date: Sun Oct 26 13:19:48 2008 +0100 138934 138935 Add missing include 138936 138937 ../../../../hw/xfree86/xaa/xaaTables.c:9:14: warning: symbol 'byte_expand3' was not declared. Should it be static? 138938 ../../../../hw/xfree86/xaa/xaaTables.c:53:14: warning: symbol 'byte_reversed_expand3' was not declared. Should it be static? 138939 138940commit 74bc792e1814849b8eee9e448e36c3568b821e6c 138941Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138942Date: Thu Nov 27 05:34:14 2008 -0200 138943 138944 Export symbols required by the vesa and fbdev drivers. 138945 138946 This is a minimal set of patches, to have an usable X Server when 138947 compiling it with symbols hidden by default. 138948 138949commit 74e2669103b59d51b24ddc48f378b1d04a3effd4 138950Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138951Date: Thu Nov 27 04:33:44 2008 -0200 138952 138953 Export symbols also defined in libXfont. 138954 138955 libXfont has stubs for these symbols, so, when compiling with hidden 138956 symbols by default, these symbols must be visible in the X Server, or 138957 the stubs in libXfont will be used. 138958 138959commit 87a7fb7438b70d07ebf240c530b8548a01021d48 138960Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 138961Date: Thu Nov 27 00:12:59 2008 -0200 138962 138963 Rework code using return value of LoaderSymbol as a function pointer. 138964 138965 The patch removes all macros in the format 138966 define xf86_sym ((type (*)(argument-list))LoaderSymbol("sym")) 138967 creates a new macro in the format 138968 define xf86_sym sym 138969 and ensures "sym" is a "visible" symbol. 138970 The patch doesn't add or remove features, and is source and binary 138971 compatible with previous shared objects (with the difference that it 138972 requires the dlloader). 138973 These symbols are a special case, as, due to the fact that LoaderSymbol 138974 was being used to reference them, they are not easily found by "automated" 138975 tools that check for missing symbols. And now it also have the benefit 138976 that the compiler/loader "knows what is going on". 138977 138978commit fc708394318ed92c612e2938b335c08c1ffebb28 138979Author: Keith Packard <keithp@keithp.com> 138980Date: Wed Nov 26 15:49:02 2008 -0800 138981 138982 Add server support for RRGetScreenResourcesCurrent 138983 138984 This depends on randrproto 1.2.99.1 138985 138986 Signed-off-by: Keith Packard <keithp@keithp.com> 138987 138988commit f710ce269c020a39f58b2bfbd0fe5192a3279c72 138989Author: Alan Hourihane <alanh@tungstengraphics.com> 138990Date: Wed Nov 26 20:02:44 2008 +0000 138991 138992 bump master to 1.6.99.1 (now the 1.6 branch is created) 138993 138994commit f3edc1fb0210149f35eab4e413700b5c4ac48214 138995Author: Keith Packard <keithp@keithp.com> 138996Date: Tue Nov 25 23:15:35 2008 -0800 138997 138998 New version of dolt 138999 139000commit 4715f079b9c61362755c95c1ebf89c97cc6fff2b 139001Merge: 9ffc67193 ed597f19f 139002Author: Keith Packard <keithp@keithp.com> 139003Date: Tue Nov 25 22:19:08 2008 -0800 139004 139005 Merge commit 'origin/master' 139006 139007commit 9ffc6719390df8fdd0a5295a7a7a0eaea792be45 139008Author: Keith Packard <keithp@keithp.com> 139009Date: Mon Nov 24 13:08:48 2008 -0800 139010 139011 Move matrix operations from X server to pixman 0.13.2 139012 139013 pixman 0.13.2 now holds all of the matrix operations. This leaves 139014 the protocol conversion routines and some ABI stubs in place 139015 139016 Signed-off-by: Keith Packard <keithp@keithp.com> 139017 139018commit ed597f19fdc3017dde6d1452b5cdf8ddcd69a5b1 139019Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 139020Date: Tue Nov 25 19:36:31 2008 -0500 139021 139022 xselinux: use "raw context" variants of getpeercon() and getcon(). 139023 139024commit ec1d08442f69353cb0e73ac4eaf0346ebb975594 139025Author: Peter Hutterer <peter.hutterer@who-t.net> 139026Date: Fri Nov 21 15:13:00 2008 +1000 139027 139028 dix: Enable core devices in InitCoreDevices already. 139029 139030 Updated patch, see 139031 http://lists.freedesktop.org/archives/xorg/2008-November/040540.html 139032 139033 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139034 Signed-off-by: Adam Jackson <ajax@redhat.com> 139035 139036commit 2b45602e828a07a0817691b2838cd34ffee531bd 139037Author: Peter Hutterer <peter.hutterer@redhat.com> 139038Date: Wed Nov 26 10:42:52 2008 +1000 139039 139040 Revert "dix: Enable core devices in InitCoreDevices already." 139041 139042 I merged the wrong patch. See correct patch at: 139043 http://lists.freedesktop.org/archives/xorg/2008-November/040540.html 139044 139045 Not activating the device before attempting to enable it would leave the 139046 sprite unset, crashing the server when enabling the real devices. 139047 139048 This reverts commit e078901a4eca02bd3e7a80d9462dafbca939a187. 139049 139050 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139051 139052commit 416685c295353b5816689994c7c58ae7db3e878d 139053Author: Jeremy Uejio <jeremy.uejio@sun.com> 139054Date: Tue Nov 25 16:26:44 2008 -0800 139055 139056 Refix Sun bug #6685465: Xephyr uses wrong or bad colortable in 8-bit mode 139057 139058 <http://bugs.opensolaris.org/view_bug.do?bug_id=6685465> 139059 139060 This is a refix of the previous fix for CR 6685465. In the first fix 139061 I was shifting the colors to match the mask by the bits_per_rgb amount 139062 in the visual structure. That field has nothing to do with the # of 139063 bits to shift by. I should just instead shift the bits to match the mask. 139064 139065commit d5f9a131a2d5bd33f82fdd4e809880b0ff792b45 139066Author: Alan Coopersmith <alan.coopersmith@sun.com> 139067Date: Tue Nov 25 15:46:39 2008 -0800 139068 139069 Fix const-mismatch warnings for DisplayModePtr's 139070 139071 Includes fixes for: 139072 "xf86Config.c", line 2434: warning: argument #1 is incompatible with prototype: 139073 prototype: pointer to struct _DisplayModeRec: "xf86.h", line 351 139074 argument : pointer to const struct _DisplayModeRec 139075 139076 "xf86EdidModes.c", line 312: warning: argument #1 is incompatible with prototype: 139077 prototype: pointer to struct _DisplayModeRec: "../../../hw/xfree86/common/xf86.h", line 351 139078 argument : pointer to const struct _DisplayModeRec 139079 139080 "xf86EdidModes.c", line 438: warning: assignment type mismatch: 139081 pointer to struct _DisplayModeRec "=" pointer to const struct _DisplayModeRec 139082 139083 "xf86Modes.c", line 701: warning: assignment type mismatch: 139084 pointer to struct _DisplayModeRec "=" pointer to const struct _DisplayModeRec 139085 139086commit d5ad296869c38ab30136b5a293a0125b76aad994 139087Author: Alan Coopersmith <alan.coopersmith@sun.com> 139088Date: Tue Nov 25 14:12:26 2008 -0800 139089 139090 Remove duplication from code paths in XkbDDXCompileKeymapByNames 139091 139092commit 2538fc0d893a150e978355d281750f0a989728a7 139093Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 139094Date: Tue Nov 25 18:18:46 2008 -0500 139095 139096 xselinux: don't pass a NULL key string to selabel_lookup(). 139097 139098commit c8472a74441838e16d0d3414db1fa7fe996868a9 139099Author: Michael Vogt <michael.vogt@ubuntu.com> 139100Date: Tue Nov 25 19:35:17 2008 +0200 139101 139102 Do not send VisibilityNotify events when MapUnmapEvents are disabled 139103 139104 This prevents a protocol visible side-effect (XVisibilityEvent) on 139105 XCompositeRedirectWindow() followed by a XCompositeUnredirectWindow(). 139106 139107 The problem shows up in gnome-screensaver with compiz and "unredirect 139108 fullscreen windows" enable. A VisibilityNotify event is generated (first 139109 with obscured and than with unobscured) when the window swithces from 139110 redirected to unredirected. 139111 139112 https://bugs.freedesktop.org/show_bug.cgi?id=18133 139113 http://launchpad.net/bugs/278112 139114 139115commit 88297558aada44bc714ad57adbeed3740aaadee5 139116Author: Adam Jackson <ajax@redhat.com> 139117Date: Tue Nov 25 14:20:58 2008 -0500 139118 139119 Dead code cleanup over DBE. 139120 139121 No DDXes have explicit DBE support anymore, so the init registration 139122 table never got used. Just nuke it all. 139123 139124commit e5eaea599ab16428c69912b6b3427ebe46707d7c 139125Author: Adam Jackson <ajax@redhat.com> 139126Date: Tue Nov 25 14:05:51 2008 -0500 139127 139128 Warning fix 139129 139130 xf86info.c:11: warning: initialization makes integer from pointer 139131 without a cast 139132 139133commit 37072500f7bcf39e0d6aa2ceb5d1f2aeeab0b26b 139134Author: Adam Jackson <ajax@redhat.com> 139135Date: Tue Nov 25 14:04:44 2008 -0500 139136 139137 Warning fix 139138 139139 helper_exec.c: In function ‘port_rep_inb’: 139140 helper_exec.c:219: warning: implicit declaration of function 139141 ‘DEBUG_IO_TRACE’ 139142 helper_exec.c:219: warning: nested extern declaration of 139143 ‘DEBUG_IO_TRACE’ 139144 139145commit d96bffce2dcf209e76be9b36ca1ede7e0c976d77 139146Author: Adam Jackson <ajax@redhat.com> 139147Date: Tue Nov 25 13:59:00 2008 -0500 139148 139149 Warning fix 139150 139151 Init.c:139: warning: no previous prototype for ‘ddxBeforeReset’ 139152 139153 Just declare the prototype always, seriously. 139154 139155commit 09ea671cbff605fd2b2af71619e7db5002108bf8 139156Author: Adam Jackson <ajax@redhat.com> 139157Date: Tue Nov 25 13:40:47 2008 -0500 139158 139159 Warning fix 139160 139161 lnx_bell.c:37: warning: no previous prototype for ‘xf86OSRingBell’ 139162 139163commit a9853c7d337b3b1ad49793e9b4b90e313b6fa536 139164Author: Adam Jackson <ajax@redhat.com> 139165Date: Tue Nov 25 00:53:55 2008 -0500 139166 139167 Warning fix 139168 139169 Pixmap.c: In function ‘xnestPixmapToRegion’: 139170 Pixmap.c:93: warning: ‘Box.x1’ may be used uninitialized in this 139171 function 139172 139173commit 81eafe9f93a272b06aa9f9235ec5676b9aa3ee3e 139174Author: Adam Jackson <ajax@redhat.com> 139175Date: Tue Nov 25 00:45:07 2008 -0500 139176 139177 Warning fix. 139178 139179 vbe.c: In function ‘VBEReadPanelID’: 139180 vbe.c:1145: warning: return from incompatible pointer type 139181 139182commit 09bfb25e031772611a2f0902d4ba77b587e4bdb2 139183Author: Adam Jackson <ajax@redhat.com> 139184Date: Tue Nov 25 00:43:36 2008 -0500 139185 139186 Remove unused XAAAvailableOptions 139187 139188commit 8b9253f6383df3fefe38bde43a5f892b158a77c4 139189Author: Adam Jackson <ajax@redhat.com> 139190Date: Tue Nov 25 00:38:47 2008 -0500 139191 139192 Code motion: subsume xf86DoProbe.c into xf86Init.c 139193 139194commit fbabb1c5c243cfd8c954dec4c060dff1a0b81015 139195Author: Adam Jackson <ajax@redhat.com> 139196Date: Tue Nov 25 00:34:28 2008 -0500 139197 139198 Warning fix. 139199 139200 In file included from l3-xaaStipple.c:4: 139201 ./xaaStipple.c:35: warning: no previous prototype for 139202 ‘XAAGetStippleScanlineFunc3LSBFirst’ 139203 139204 etc 139205 139206commit cbaca6ec666d7349c4680b8affc13b5c9cae1fa5 139207Author: Adam Jackson <ajax@redhat.com> 139208Date: Tue Nov 25 00:14:24 2008 -0500 139209 139210 Remove dead FreeModuleDesc. 139211 139212commit eb474adf98229a43bbe17ab98ff084371cb9fa09 139213Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139214Date: Tue Nov 25 11:25:58 2008 -0800 139215 139216 XQuartz: Simplify the xquartz_resetenv_display path 139217 (cherry picked from commit d2e0624dd30eb234bb25595ceedfa51d48ca1724) 139218 139219commit 94df1ab7f09a64f57c1e1453e3640462e984619c 139220Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139221Date: Tue Nov 25 01:30:03 2008 -0800 139222 139223 XQuartz: Force X11Controller to reset a broken DISPLAY envvar. 139224 (cherry picked from commit f1a52b5b5ac31702497937efe3ac578be9a6c54f) 139225 139226commit 0b314c50a2a0ca1afbdc06663c3b719b05ebb851 139227Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139228Date: Tue Nov 25 01:13:35 2008 -0800 139229 139230 XQuartz: Removed hardcoded org.x.X11 from MachIPC as well 139231 (cherry picked from commit b4add7826d485600a13eba6a9c7be533f2c02d51) 139232 139233commit 065d2afb0ca34f89806e0936c51cd27805bc5123 139234Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139235Date: Tue Nov 25 00:51:01 2008 -0800 139236 139237 XQuartz: Add fallback for xpbproxy's display for Tiger or no-launchd-Leopard 139238 (cherry picked from commit 7a8c6665949d7804a97ef2539a74ec4aa682e1cc) 139239 139240commit 13df49dca28cf680a4d104630cd675de25d3e944 139241Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139242Date: Tue Nov 25 00:39:52 2008 -0800 139243 139244 XQuartz: Use the environment to pass the bundle's prefs domain on to xinit/quartz-wm for Tiger or no-launchd-LEOPARD 139245 (cherry picked from commit fbf4b0d33fa5dc618c3191a4e823232dfa33cd95) 139246 139247commit 40187f782beae4ae751824ef511c9f56a80357c7 139248Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139249Date: Tue Nov 25 00:20:57 2008 -0800 139250 139251 XQuartz: Dead code removal 139252 (cherry picked from commit eeb323612e0adbea37befed31bbaa1d295728385) 139253 139254commit b55cad4569e34e3c10e9a327e20b91ea87d9dd98 139255Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139256Date: Tue Nov 25 00:15:53 2008 -0800 139257 139258 XQuartz: Don't hardcode values of org.x.X11 for the preferences domain 139259 (cherry picked from commit 3a500d9247cf34686ec17b4a88c34d51ecd38ecd) 139260 139261commit d508a3dcca2f160021aced872715e1ded23cef97 139262Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139263Date: Mon Nov 24 23:33:54 2008 -0800 139264 139265 XQuartz: More dead code removal 139266 (cherry picked from commit dcb0f6a2e62823a671051874d14a33ce59505892) 139267 139268commit 41fbdf72f2154a3fca8cf484a611501e3c174fbe 139269Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139270Date: Mon Nov 24 12:33:20 2008 -0800 139271 139272 XQuartz: Disable some error spew on Tiger (where it wouldn't be an error) 139273 (cherry picked from commit 73ec6d3dfe0086d352f4eca25f1df5ae1884bb18) 139274 139275commit 09c3f6e04c273ffafcb547c252137fb17c8ce016 139276Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139277Date: Sat Nov 22 20:23:46 2008 -0800 139278 139279 XQuartz: Dead code removal 139280 (cherry picked from commit 46c077d9b4a883fc809c32077ce40f33a70d268b) 139281 139282commit ab12c7516207908f3e063a78904d68e2db14208e 139283Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139284Date: Sat Nov 22 15:47:14 2008 -0800 139285 139286 XQuartz: More Tiger cleanup: bootstrap_strerror 139287 (cherry picked from commit 37f535aff3e9a7a02711daa98152cdff97745622) 139288 139289commit 208f091bfc657e9ee57b988f035d3aac7e9e173a 139290Author: Adam Jackson <ajax@redhat.com> 139291Date: Sun Oct 12 00:07:56 2008 -0400 139292 139293 Remove xf86IODelay 139294 139295commit 47478aa10adb20b6e48335ac8c4ef31f3fdfe68c 139296Author: Adam Jackson <ajax@redhat.com> 139297Date: Sat Oct 11 23:59:24 2008 -0400 139298 139299 Remove xf86{En,Dis}ableInterrupts entirely 139300 139301commit c7680befe5aebd0f4277d11ff3984d8a7deb9d5b 139302Author: Adam Jackson <ajax@redhat.com> 139303Date: Sat Oct 11 23:35:24 2008 -0400 139304 139305 Remove the remnants of Jensen support 139306 139307 As being an EISA-only machine, and as ISA support is gone now... 139308 139309commit df14682a31b92751091571ed82f6095f55f19cca 139310Author: Adam Jackson <ajax@redhat.com> 139311Date: Sat Oct 11 22:48:51 2008 -0400 139312 139313 Bus: Remove ISA support. 139314 139315 No, really. PCI is old enough to drive now. If you want this, get the kernel 139316 to expose a framebuffer device. 139317 139318commit a8bcab2d3b224e4d4d5b6a097ea530beee920213 139319Author: Adam Jackson <ajax@redhat.com> 139320Date: Sat Oct 11 22:14:23 2008 -0400 139321 139322 Bus: Remove yet more unused overlap processing. 139323 139324commit 994b7c034fc20d76651cf7f6a285526d9aff8770 139325Author: Adam Jackson <ajax@redhat.com> 139326Date: Sat Oct 11 22:11:12 2008 -0400 139327 139328 Bus: Don't pretend to care about IRQs, DMA, or PCI config space 139329 139330commit eb5ae45127fa9f08f0badec7e21f8c26c9c7c969 139331Author: Adam Jackson <ajax@redhat.com> 139332Date: Sat Oct 11 21:44:16 2008 -0400 139333 139334 Bus: Simplify a failure case (that pretty much never happens) 139335 139336commit ee8b5cf94c2a9f8526a3bb5011ebb10f3246a4d9 139337Author: Adam Jackson <ajax@redhat.com> 139338Date: Sat Oct 11 21:41:47 2008 -0400 139339 139340 Bus: Trust the kernel when registering driver resources 139341 139342 ... everywhere, not just (linux && (ia64 || alpha)). 139343 139344commit 8397df89456558e3c85b05e0acfccb9f6af6b695 139345Author: Adam Jackson <ajax@redhat.com> 139346Date: Sat Oct 11 21:36:14 2008 -0400 139347 139348 Remove unused MIN macro 139349 139350commit c251c0baae59714a6ac83b69cd106c08baa3613e 139351Author: Adam Jackson <ajax@redhat.com> 139352Date: Sat Oct 11 21:34:27 2008 -0400 139353 139354 Bus: remove special handling for init-only resources 139355 139356 This isn't used by any driver, nor has it ever been as far as I can 139357 tell. 139358 139359commit 3e5281af17841cf50d0e52a728b12c6ab56e61df 139360Author: Adam Jackson <ajax@redhat.com> 139361Date: Sat Oct 11 21:16:45 2008 -0400 139362 139363 PCI: Unexport xf86scanpci 139364 139365commit 6b198daa46f2f609aff7900761cf82cc2fb4e0b4 139366Author: Adam Jackson <ajax@redhat.com> 139367Date: Sat Oct 11 20:51:39 2008 -0400 139368 139369 Bus: remove useless isaConvertRange2Host 139370 139371commit a96db74c2a95bb1dce132cf47ea720ae939dfad7 139372Author: Adam Jackson <ajax@redhat.com> 139373Date: Thu Oct 9 00:43:26 2008 -0400 139374 139375 Bus: Don't try to find an ISA bus just for fun. 139376 139377commit b21311a99d58997cd1fc68726d0848242e9c34fc 139378Author: Adam Jackson <ajax@redhat.com> 139379Date: Thu Oct 9 00:34:42 2008 -0400 139380 139381 Bus: Remove unused RemoveOverlaps 139382 139383commit 41be6b3f0dc0baa1c6ae8d2b41a6be73ca0e7268 139384Author: Adam Jackson <ajax@redhat.com> 139385Date: Thu Oct 9 00:33:28 2008 -0400 139386 139387 Bus: Remove the notion of estimated resources. 139388 139389commit 095ba1435501776c8c8a34e767b89f89e5dc949a 139390Author: Adam Jackson <ajax@redhat.com> 139391Date: Thu Oct 9 00:27:33 2008 -0400 139392 139393 Bus: remove the "reducer" 139394 139395 This code effectively didn't do anything anymore. 139396 139397commit 4457e31710af90f9ac295bb686c841e9473fb767 139398Author: Adam Jackson <ajax@redhat.com> 139399Date: Thu Oct 9 00:14:54 2008 -0400 139400 139401 PCI: Remove unused ia64 platform code. 139402 139403commit 86cfe0ee236bfd3613e5f9ba589211db42d009eb 139404Author: Adam Jackson <ajax@redhat.com> 139405Date: Wed Oct 8 23:45:40 2008 -0400 139406 139407 PCI: Simplify OS PCI function registration a bit. 139408 139409commit 2d427b9cb1594f8f2f66b463033fff5b459962fd 139410Author: Adam Jackson <ajax@redhat.com> 139411Date: Wed Oct 8 23:38:23 2008 -0400 139412 139413 PCI: Remove config mechanism details. 139414 139415 pciaccess handles this for us now, no need to remember PC arcana. 139416 139417commit 5bb86bafd6fda296011cbcd5d15a85a6d770ae29 139418Author: Adam Jackson <ajax@redhat.com> 139419Date: Wed Oct 8 23:34:41 2008 -0400 139420 139421 PCI: Remove non-pciaccess path for x86. 139422 139423commit a67360e79fa7e17c3d907771694009c57c1cd195 139424Author: Adam Jackson <ajax@redhat.com> 139425Date: Wed Oct 8 23:31:38 2008 -0400 139426 139427 PCI: Always build domain support. 139428 139429commit 51e105ccc3d0ac8c0fe74efd029ffbddb80b140e 139430Author: Adam Jackson <ajax@redhat.com> 139431Date: Wed Oct 8 23:12:31 2008 -0400 139432 139433 Remove xf86GetResourcesImplicitly 139434 139435commit 599a0f3f1e3ae92676e3648471576c0001cfd9ae 139436Author: Alan Coopersmith <alan.coopersmith@sun.com> 139437Date: Mon Nov 24 20:37:58 2008 -0800 139438 139439 Fix typo in 5bf2c88d2317230b95b2904cb975167d03ee13a2 139440 139441 Amazing how these things hide until you see the diff come back from 139442 the commit list. 139443 139444commit 5bf2c88d2317230b95b2904cb975167d03ee13a2 139445Author: Alan Coopersmith <alan.coopersmith@sun.com> 139446Date: Mon Nov 24 20:34:46 2008 -0800 139447 139448 Simplify filename generation code for Xorg -configure 139449 139450commit 75784e1e53ad78e21518696dd9d297bc08c17d54 139451Author: Alan Coopersmith <alan.coopersmith@sun.com> 139452Date: Mon Nov 24 20:32:20 2008 -0800 139453 139454 Solaris: Make KDSETMODE failure non fatal, and retry it on interrupts 139455 139456commit 3e6da1636093d7dc98baac40544c0b0fb7fd8aec 139457Author: Peter Hutterer <peter.hutterer@redhat.com> 139458Date: Fri Nov 14 15:55:57 2008 +1000 139459 139460 include: remove ENTER_LEAVE_SEMAPHORE macros. 139461 139462commit b292a7a2d7e259177e1cc37346c2bee27a018630 139463Author: Peter Hutterer <peter.hutterer@redhat.com> 139464Date: Fri Nov 14 15:44:29 2008 +1000 139465 139466 dix: updated enter/leave core event model. 139467 139468 As proposed by Owen Taylor [1], the enter-leave event model needs to adjust 139469 the events sent to each window depending on the presence of pointers in a 139470 window, or in a subwindow. 139471 139472 The new model can be summarised as: 139473 - if the pointer moves into or out of a window that has a pointer in a child 139474 window, the events are modified to appear as if the pointer was moved out of 139475 or into this child window. 139476 - if the pointer moves into or out of a window that has a pointer in a parent 139477 window, the events are modified to appear as if the pointer was moved out of 139478 or into this parent window. 139479 139480 Note that this model requires CoreEnterLeaveEvent and DeviceEnterLeaveEvent to 139481 be split and treated separately. 139482 139483 [1] http://lists.freedesktop.org/archives/xorg/2008-August/037606.html 139484 139485 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139486 139487commit 724f83b87bb16472d4c328e35d2a477384b29f84 139488Author: Peter Hutterer <peter.hutterer@redhat.com> 139489Date: Fri Nov 14 15:29:01 2008 +1000 139490 139491 dix: add FirstPointerChild, FirstPointerAncestor auxiliary functions. 139492 139493 FirstPointerChild: Return the first child that has a pointer within its 139494 boundaries. 139495 139496 FirstPointerAncestor: return the first ancestor with a child within its 139497 boundaries. 139498 139499 These are required for the updated enter/leave model. 139500 139501 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139502 139503commit 7d3e595f93dcd3d334e766a9dea602c05affdbaf 139504Author: Peter Hutterer <peter.hutterer@redhat.com> 139505Date: Fri Nov 14 15:27:19 2008 +1000 139506 139507 dix: Add EnterWindow, LeaveWindow, HasPointer auxiliary functions. 139508 139509 These replace the ENTER_LEAVE_SEMAPHORE_* macros. Unused currently. 139510 139511 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139512 139513commit 6bdc963cdabb4a2e77de7f00a1d062aa2b873f9b 139514Author: Peter Hutterer <peter.hutterer@redhat.com> 139515Date: Fri Nov 14 15:37:35 2008 +1000 139516 139517 dix: split enter/leave event handling into core and device handling. 139518 139519 Device events always need to be delivered, core events only in some cases. 139520 Let's keep them completely separate so we can adjust core event delivery. 139521 139522 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139523 139524commit 5e48f5e2dd2dec7cfd1fa40b61e25123dfca515e 139525Author: Peter Hutterer <peter.hutterer@redhat.com> 139526Date: Fri Nov 14 15:41:59 2008 +1000 139527 139528 dix: remove unused EnterLeaveSemaphoresIsset. 139529 139530 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139531 139532commit 868fd503b7aeab31dba72046b59061008d8b7501 139533Author: Peter Hutterer <peter.hutterer@redhat.com> 139534Date: Fri Nov 14 15:18:56 2008 +1000 139535 139536 dix: move Enter-Leave related functions into new enterleave.c 139537 139538 Preparation for the new core enter/leave model. 139539 139540 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139541 139542commit c4b9ab6bf56139fdd8c7c584a6f523c6766cddd6 139543Author: Keith Packard <keithp@keithp.com> 139544Date: Tue Apr 29 08:29:42 2008 -0700 139545 139546 Handle transform failure when computing shadow damage area. 139547 139548 PictureTransformBounds can fail, when this happens, damage the entire screen 139549 so that the shadow gets repainted correctly. 139550 139551commit 9c7679240ad90367693a462e288308b3fdc08f26 139552Author: Keith Packard <keithp@keithp.com> 139553Date: Sun Mar 30 19:05:07 2008 -0700 139554 139555 [randr] don't try to compute crtc transform when no mode is set. 139556 139557 Dereferencing the NULL mode pointer would cause a crash. As these transform 139558 matrices won't be used while the CRTC is disabled, just leave their values 139559 alone. 139560 139561commit 315b6d0a425d0257f226ee7032aca3ca31665e59 139562Author: Keith Packard <keithp@keithp.com> 139563Date: Thu Mar 27 11:46:39 2008 -0700 139564 139565 Update RandR global transform when driver notifies of transform change. 139566 139567 Need to compute and save the global transform when the driver changes it. 139568 139569commit 5d9282fde919c936816a85c2a9c5734dc9d57cc2 139570Author: Keith Packard <keithp@keithp.com> 139571Date: Thu Mar 27 11:45:53 2008 -0700 139572 139573 Export rrtransform.h as needed by randrstr.h 139574 139575 When the transform management was moved from randrstr.h, the associated 139576 header file became necessary to build drivers. Include it as a part of the 139577 sdk headers. 139578 139579commit 7236fd9dd4e9f5c4cb5bafa5c2a18cff12a0444c 139580Author: Keith Packard <keithp@keithp.com> 139581Date: Fri Mar 21 03:16:09 2008 -0700 139582 139583 Be careful about copying transforms around; they have allocated memory. 139584 139585commit 7e69e364d2ef146d8ec3651d04bdd6d641017449 139586Author: Keith Packard <keithp@keithp.com> 139587Date: Fri Mar 21 03:15:00 2008 -0700 139588 139589 Fix rotated/reflected cursor positions. 139590 139591 Doing projective transforms required repositioning the cursor using the 139592 hotspot, but that requires relocating the upper left corner in terms of said 139593 hotspot. 139594 139595commit 93179c214fc6ed88f72955a11c69ae0a47316d8c 139596Author: Keith Packard <keithp@keithp.com> 139597Date: Fri Mar 21 03:14:47 2008 -0700 139598 139599 rrtransform needs randrstr to get RANDR_INTERFACE defines 139600 139601commit 7c61db66a470a8306e346ed8bf8934f014dada42 139602Author: Keith Packard <keithp@keithp.com> 139603Date: Fri Mar 21 02:39:49 2008 -0700 139604 139605 Create rrtransform.[ch]. Add RRTransform argument to RRCrtcNotify. 139606 139607 Instead of using a separate function to notify DIX about transform changes, 139608 add the transform to RRCrtcNotify so that the whole Crtc state changes 139609 atomically. 139610 139611commit fa6a1df209bd74da1d545982cca437afc2198cc1 139612Author: Keith Packard <keithp@keithp.com> 139613Date: Fri Mar 21 02:35:28 2008 -0700 139614 139615 Avoid overflow in PictureTransformPoint. Fix PictureTransformIsIdentity. 139616 139617 PictureTransformPoint computes homogeneous coordinates internally, but fails 139618 to handle intermediate values larger than 16.16. Use 64 bit intermediate 139619 values while computing the final result at 16.16 and only complain if that 139620 result is too large. 139621 139622 PictureTransformIsIdentity was completely wrong -- it was not checking for 139623 identity transforms at all. 139624 139625commit 6fe9c15731be5f5afabacb3aa3ed71b840f4238b 139626Author: Keith Packard <keithp@keithp.com> 139627Date: Wed Mar 19 12:15:39 2008 -0700 139628 139629 Allow drivers to set crtc transforms. 139630 139631 Track curent transform down in the mode setting code so that it may be set 139632 separately from RandR. 139633 139634commit 49db14e4ac26070ed86088419483888dda18b603 139635Author: Keith Packard <keithp@keithp.com> 139636Date: Wed Mar 19 00:46:35 2008 -0700 139637 139638 Handle RandR transform matrices in floating point. 139639 139640 RandR matrix computations lose too much precision in fixed point; 139641 computations using the inverted matrix can be as much as 10 pixels off. 139642 Convert them to double precision values and pass those around. These API 139643 changes are fairly heavyweight; the official Render interface remains fixed 139644 point, so the fixed point matrix comes along for the ride everywhere. 139645 139646commit 6f734aecaec4f5c6152c3ebca197ef65bb4523da 139647Author: Keith Packard <keithp@keithp.com> 139648Date: Tue Mar 18 16:06:06 2008 -0700 139649 139650 Eliminate inverse matrix from randr transform protocol 139651 139652 It is easier, and potentially more precise, to compute the inverse in the 139653 server where everything can eventually be kept in floating point form. 139654 139655commit 197aa784694992f6bcfd194932309e440334c237 139656Author: Keith Packard <keithp@keithp.com> 139657Date: Tue Mar 18 15:35:44 2008 -0700 139658 139659 Report whether transforms are support from driver through extension to client 139660 139661 Add APIs to xf86RandR12 support and randr extension to record whether the 139662 driver supports transforms, report that value in the RRGetCrtcTransform 139663 reply. 139664 139665commit e86c34663ef97e946a1129450105efa89a123af6 139666Author: Keith Packard <keithp@keithp.com> 139667Date: Tue Mar 18 15:25:31 2008 -0700 139668 139669 Clear shadow pixmaps before using them. 139670 139671 This eliminates some ugly flashing, as well as clearing the borders when the 139672 shadow will not be completely painted. 139673 139674commit 97ab0c6eff870b52c0383b63a78cec49059b2545 139675Author: Keith Packard <keithp@keithp.com> 139676Date: Tue Mar 18 15:15:40 2008 -0700 139677 139678 When converting from double to fixed, round carefully. 139679 139680 This reduces the matrix representation error after inverting a 139681 transformation matrix (although it doesn't eliminate it entirely). 139682 139683 Perhaps we should extend Render to include 64-bit floating point transforms... 139684 139685commit 6d3a9e40a4b9ec455af11cce31e4aa616c93db32 139686Author: Keith Packard <keithp@keithp.com> 139687Date: Mon Mar 17 23:04:49 2008 -0700 139688 139689 Compute matrix inversion instead of using wire version in RRCrtcTransformSet 139690 139691 It doesn't make sense to have the client invert this matrix when the server 139692 can do so reasonably efficiently. This avoids weird fixed point rounding 139693 errors when testing the transform against its inverse. Now to fix the 139694 protocol. 139695 139696commit 160252d94f04acc95f0a4e0f884ff565a5aa0744 139697Author: Keith Packard <keithp@keithp.com> 139698Date: Mon Mar 17 23:03:56 2008 -0700 139699 139700 Add matrix inversion function (uses doubles) 139701 139702 The obvious matrix inversion function, coded using doubles to avoid fiddling 139703 with fixed point precision adventures. 139704 139705commit 8fd82c88e3f2060fda4ba30376900ece77668c63 139706Author: Keith Packard <keithp@keithp.com> 139707Date: Mon Mar 17 16:14:43 2008 -0700 139708 139709 Adjust transformed cursor position to account for hotspot 139710 139711commit eb222e64128034df8361d5a82d4f4aa1318923ce 139712Author: Keith Packard <keithp@keithp.com> 139713Date: Mon Mar 17 16:14:15 2008 -0700 139714 139715 Actually use filter kernel size to expand transform redisplay box 139716 139717commit 3fdb963f6e8287edeb4c5bc7bbadbc02eb8bb910 139718Author: Keith Packard <keithp@keithp.com> 139719Date: Mon Mar 17 16:13:25 2008 -0700 139720 139721 Correct bilinear filter kernel size (should be 2x2) 139722 139723commit ee6326a6b89b2d223f6e1eaf02aac1fe8aebfaf6 139724Author: Keith Packard <keithp@keithp.com> 139725Date: Mon Mar 17 15:22:06 2008 -0700 139726 139727 Handle filter widths in xf86Rotate 139728 139729commit ddc8466137d229fdc100cc403f492c7d61a1ba89 139730Author: Keith Packard <keithp@keithp.com> 139731Date: Mon Mar 17 15:21:26 2008 -0700 139732 139733 Initialize and cleanup new filter fields in xf86Crtc. 139734 139735commit 16c093afd4f6b30d889cacdc994b0024f9bd83d5 139736Author: Keith Packard <keithp@keithp.com> 139737Date: Mon Mar 17 15:20:52 2008 -0700 139738 139739 Pass filter kernel size through transforms 139740 139741commit 1df02d7ddd44f84bcaefd6583af1a9141cb3c78b 139742Author: Keith Packard <keithp@keithp.com> 139743Date: Mon Mar 17 15:19:17 2008 -0700 139744 139745 Add kernel size to Render filters. 139746 139747 This width/height value lets filter users know how far the filter spreads 139748 into the source image. 139749 139750commit 40f3dff6b350fe0be55ebf7dbca88a0ef8f6380f 139751Author: Keith Packard <keithp@keithp.com> 139752Date: Mon Mar 17 13:57:47 2008 -0700 139753 139754 Use transform when computing scanout size of modes 139755 139756 Report transformed crtc sizes through RandR and Xinerama. Test screen size 139757 against transformed mode sizes when configuring the Crtc. 139758 139759commit f50349e1930e620cacdf27a6f8585fcb9cb5199a 139760Author: Keith Packard <keithp@keithp.com> 139761Date: Sat Mar 15 00:36:45 2008 -0700 139762 139763 [RANDR] Support filters in CRTC transforms. 139764 139765 Create new RRTransform datatype to hold all of the transform related 139766 information, use that in lots of places to pass filters around. 139767 139768commit acda790e430b2a18c7c35379f6e538f3d01ff221 139769Author: Keith Packard <keithp@keithp.com> 139770Date: Fri Mar 14 13:46:30 2008 -0700 139771 139772 [render] Split out filter finding from filter setting. 139773 139774 To prepare for RandR using filters in transforms, split out 139775 code paths so that the RandR code can validate the filter name and 139776 parameters during the transform set operation so that use of the filter 139777 later will not have unreportable errors. 139778 139779commit e3d6f279d5c305dfb81ca109fbfb665870712d2f 139780Author: Keith Packard <keithp@keithp.com> 139781Date: Thu Mar 13 21:31:12 2008 -0700 139782 139783 Wire up RandR CRTC transform protocol, bump server to RandR 1.3 139784 139785 This involved removing a pile of matrix code from the DDX, 139786 as well as moving a bit of transform logic from DDX to DIX. 139787 139788commit ff9d1cd843a9b0aba69a3d788b21d5f6af702590 139789Author: Keith Packard <keithp@keithp.com> 139790Date: Thu Mar 13 21:30:18 2008 -0700 139791 139792 Add funcs to convert between protocol and pixman matrices 139793 139794commit c2f254da694731ea573aa8bbc2707c083743f2d4 139795Author: Keith Packard <keithp@keithp.com> 139796Date: Thu Mar 13 21:29:19 2008 -0700 139797 139798 Randr now depends on Render for matrices 139799 139800commit f547650328287545a7a4d96df8d6a6c606dd95a9 139801Author: Keith Packard <keithp@keithp.com> 139802Date: Thu Mar 13 14:50:13 2008 -0700 139803 139804 Export a bunch of matrix operations from render. 139805 139806 The render extension uses many matrix operations internally, this change 139807 exposes those functions to other parts of the server, drivers and 139808 extensions. The change is motivated by the 'transform' additions to the 139809 RandR extension but will likely be useful elsewhere. 139810 139811commit e063162e80e51ed4368874b3af7ba690ea280d9e 139812Author: Keith Packard <keithp@keithp.com> 139813Date: Thu Mar 13 14:26:01 2008 -0700 139814 139815 Add projective transforms to RandR DIX/DDX API. 139816 139817 New RRCrtcGetTransform function in DIX that DDX can use to get the pending 139818 transform. The DDX code should be complete; the DIX code is just a stub at 139819 this point. 139820 139821commit 9c7ac47b871a71e42d2f6933749ca462f1a65b40 139822Author: Keith Packard <keithp@keithp.com> 139823Date: Tue Nov 4 00:10:08 2008 -0800 139824 139825 Notify DRI when crtc regions change 139826 139827 Drivers that care about crtc positions on the screen to ensure that vblank 139828 works correctly need to be notified when crtcs are changed. 139829 139830 Provide a hook in the mode setting code that is invoked whenever any 139831 configuration is done to the screen. 139832 139833 Use this new hook in the DRI code so that DRI clients are notified and 139834 receive updated information. 139835 139836 Signed-off-by: Keith Packard <keithp@keithp.com> 139837 139838commit 554592cd70543f87cd8bee5ff47cc8281511e041 139839Author: Peter Hutterer <peter.hutterer@redhat.com> 139840Date: Mon Nov 24 11:50:38 2008 +1000 139841 139842 DGA: silence compiler warning. 139843 139844 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139845 139846commit 0dbcbd35f63db14734199a0beb2a91b6937f3838 139847Author: Peter Hutterer <peter.hutterer@redhat.com> 139848Date: Fri Nov 21 12:02:02 2008 +1000 139849 139850 xfree86: Split the working code of NIDR into new xf86NewInputDevice. 139851 139852 The xfree86 server previously hat NewInputDeviceRequest and InitInput, and 139853 both basically did the same thing. Reduce NIDR to parameter checking and use 139854 xf86NewInputDevice from both InitInput and NIDR to actually create the device. 139855 139856 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139857 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 139858 Signed-off-by: Adam Jackson <ajax@redhat.com> 139859 139860commit e078901a4eca02bd3e7a80d9462dafbca939a187 139861Author: Peter Hutterer <peter.hutterer@redhat.com> 139862Date: Fri Nov 21 12:39:55 2008 +1000 139863 139864 dix: Enable core devices in InitCoreDevices already. 139865 139866 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139867 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 139868 Signed-off-by: Adam Jackson <ajax@redhat.com> 139869 139870commit d939f2482e71ad20bac28b7aa4b2e8c8196e9d65 139871Author: Peter Hutterer <peter.hutterer@redhat.com> 139872Date: Fri Nov 21 12:21:45 2008 +1000 139873 139874 dix: fix false comment. 139875 139876 VCP and VCK are always present, this comment is a leftover from earlier MPX 139877 days. 139878 139879 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 139880 139881commit 387563b77743d92be83420d982eaf57f5459a883 139882Merge: 85f650c9b f6e01fa1b 139883Author: Matthieu Herrb <matthieu.herrb@laas.fr> 139884Date: Sun Nov 23 13:15:46 2008 +0100 139885 139886 Merge branch 'master' of git+ssh://herrb@git.freedesktop.org/git/xorg/xserver 139887 139888commit 85f650c9b76ab2e244f01908a7941c320635becd 139889Author: Matthieu Herrb <matthieu.herrb@laas.fr> 139890Date: Sun Nov 23 13:14:27 2008 +0100 139891 139892 missing action in if statement. 139893 139894commit f6e01fa1b87ea190ea5ad723ce46893784ea1de4 139895Author: Stefan Dirsch <sndirsch@suse.de> 139896Date: Sun Nov 23 11:16:03 2008 +0100 139897 139898 Added '-showopts' option to print available driver options (#5564). 139899 139900commit 8964b8d0ec2b7b3b6bf540cd647b14a20e8f64a5 139901Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139902Date: Sat Nov 22 14:23:23 2008 -0800 139903 139904 XQuartz: More 1.6 server API updates 139905 139906commit ad0f232165fe1a25ca4fb6da817da02b6ce31779 139907Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139908Date: Sat Nov 22 14:04:28 2008 -0800 139909 139910 XQuartz: Fixed --disable-glx 139911 139912commit 8e2287c220694953e972cd5119c2b0cd256b7a30 139913Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139914Date: Sat Nov 22 13:57:45 2008 -0800 139915 139916 XQuartz: Updated some code to use newer server API 139917 139918commit 6eb33bc0cb2e62339d323e1f1894015d7e3142f3 139919Author: George Staplin <gstaplin@apple.com> 139920Date: Sat Nov 22 11:37:08 2008 -0700 139921 139922 XQuartz: GL: Remove the inclusion of glcontextmodes.h. 139923 139924 Add some commentary about future directions needed for the GLX drawable 139925 creation and destruction code. 139926 139927 Match xalloc with xfree. 139928 139929 I made some minor formatting improvements. 139930 (cherry picked from commit b772d64fce31d16b498c621096e39d5203994d6e) 139931 139932commit 429b4b20d5708d608fd55f91dd5bcd4ac0b51a12 139933Author: George Staplin <gstaplin@apple.com> 139934Date: Sat Nov 22 10:57:58 2008 -0700 139935 139936 XQuartz: GL: Add a branch to prevent a NULL DrawablePtr structure access. 139937 139938 In attach() check for pDraw being NULL, and also print an ErrorF message, 139939 because we eventually want to track down why this is occuring. 139940 139941 It's unclear how this occurs, but as I noted in the 1.4 branch, I believe that 139942 the DrawablePtr/struct _Drawable -> id is the member being accessed that causes 139943 KERN_PROTECTION_FAILURE at 0x0000000000000004 139944 139945 This passes my tests using: env LIBGL_ALWAYS_INDIRECT=1 ./sometest. 139946 139947 I fixed a warning: caused by initializing the screen->base.visuals with the 139948 configs. It is a ** not a *. It seems that some other part of GLX will 139949 initialize this for us. 139950 (cherry picked from commit 17f6a261fca6d5856069dce28bb4838261afc6bc) 139951 139952commit bc0c7075e29b1719409cf7de8a4ab9ae1315aa67 139953Author: Matthieu Herrb <matthieu.herrb@laas.fr> 139954Date: Sat Nov 22 19:10:09 2008 +0100 139955 139956 Remove the old libusbhid copy from os-support/bsd. 139957 139958 This has been replaced by in-kernel support for USB mices several 139959 years ago and hasn't been enabled in build since X.Org 6.8. 139960 139961commit 76fcfc480133726112049e5bd9c3082d46825918 139962Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139963Date: Fri Nov 21 20:25:55 2008 -0800 139964 139965 XQuartz: GLX: sync up indirect.c to match George's updates in the xorg-server-1.5-apple branch 139966 139967commit 1a717779b67b9c1cdf734eef64135ac38829f726 139968Author: George Staplin <gstaplin@apple.com> 139969Date: Fri Nov 21 17:28:11 2008 -0700 139970 139971 XQuartz: GL: remove glcontextmodes.* from the Makefile.am. It has been replaced. 139972 139973 The new replacement is __GLXconfig. 139974 (cherry picked from commit 3bed9b65c807a1aeb662b8042826cbb54280181d) 139975 139976commit dacad9ae0bdaedff9403fbe84a5c5a4b4eef3542 139977Author: Alan Coopersmith <alan.coopersmith@sun.com> 139978Date: Fri Nov 21 17:16:22 2008 -0800 139979 139980 Xephyr man page typo fixes. 139981 139982 Also trailing white space removal to clear git warnings. 139983 139984commit ca64b5949d229a60a3c62b6d29d51873b7694bd6 139985Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139986Date: Fri Nov 21 11:20:31 2008 -0800 139987 139988 XQuartz: Don't use LS to find X11.app on Tiger. 139989 (cherry picked from commit e62107e55261ef252a2a24dd26a60e5dd295d560) 139990 139991commit 701f8e3a94d2b659ebcfffd09acd1d66f9e797b8 139992Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 139993Date: Fri Nov 21 11:08:00 2008 -0800 139994 139995 XQuartz: Update applewm deps 139996 (cherry picked from commit 0ad91c59be8759a9c9e9d4f639056d8c689a3bc5) 139997 139998commit 393325f908f8bd6648e466183f3683a2d80b264e 139999Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140000Date: Fri Nov 21 10:57:31 2008 -0800 140001 140002 XQuartz: spewCallStack is noop on Tiger now... 140003 (cherry picked from commit 56c469a68b92c6cf003802f418ea00bd596a41c3) 140004 140005commit 5d47a5d6526b7e52387647a15a580dfcafd1733f 140006Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140007Date: Fri Nov 21 10:54:55 2008 -0800 140008 140009 XQuartz: pbproxy: Added some typedefs for types not available on Tiger 140010 (cherry picked from commit 0947aa7911f1de44bfe16e505a757b659c5ab2a8) 140011 140012commit b262788401715787c68cea4ca79f4b270307afee 140013Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140014Date: Fri Nov 21 10:53:39 2008 -0800 140015 140016 XQuartz: More Tiger ifdefs 140017 (cherry picked from commit 803509072f2c039e5cd555c4ac14d672f0e698c5) 140018 140019commit 510744444329170f702d31d43997a6bbc1591b44 140020Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140021Date: Fri Nov 21 10:44:11 2008 -0800 140022 140023 XQuartz: Resurrect the old audio path for Tiger 140024 (cherry picked from commit a61a8d9047e8765faf3892f4f2148c8553fd192c) 140025 140026commit d7ee76f9dd84da05b59591a971b96bf990136767 140027Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140028Date: Wed Nov 19 11:42:59 2008 -0800 140029 140030 XQuartz: pbproxy: Push the pbproxy Xevent processing into its own thread 140031 and just have the AppKit thread wake it up. 140032 (cherry picked from commit 799715b8f3327c8da59ab45706e85af2d2c438e4) 140033 140034commit a9e20306fbe3262602f21b876a52a1ef38cdf20a 140035Author: Egbert Eich <eich@ovid.suse.de> 140036Date: Fri Nov 21 18:50:01 2008 +0100 140037 140038 int10: Do an mprotect(..,PROT_EXEC) on shmat()ed memory ranges. 140039 140040 When the linux kernel sets the NX bit vm86 segfaults when it tries to execute 140041 code in memory that is not marked EXEC. Such code gets called whenever 140042 we return from a VBIOS call to signal the calling program that the call 140043 is actually finished and that we are not trapping for other reasons (like 140044 IO accesses). 140045 Use mprotect(2) to set these memory ranges PROT_EXEC. 140046 140047commit a54153e669fd293a47f0077bf25505dd545ddce2 140048Author: Rémi Cardona <remi@gentoo.org> 140049Date: Fri Nov 21 16:39:00 2008 +0100 140050 140051 xfree86: don't reset Auto(Add|Enable)Devices, use defaults from xf86Globals 140052 140053 Without this, commit ace38fafb062372dcd3d56378b5b8f86525c6241 is useless 140054 when HAL support is disabled. 140055 140056 Signed-off-by: Julien Cristau <jcristau@debian.org> 140057 140058commit 6a1b2fed7e465a9c652e8ee250b8eab440945862 140059Author: Matthieu Herrb <matthieu.herrb@laas.fr> 140060Date: Sun Sep 28 12:18:16 2008 +0200 140061 140062 Add a man page for Xephyr, with information from the README file. 140063 140064commit 10d472bf9d61bef32d0a383b2e3783e7a1d7621f 140065Author: Matthieu Herrb <matthieu.herrb@laas.fr> 140066Date: Mon May 19 05:15:30 2008 +0200 140067 140068 Support for BSD apm 140069 XXX check amd64 pci init 140070 140071commit f4e725f248870a2de10449cc8f53de32e7840d45 140072Author: Peter Hutterer <peter.hutterer@redhat.com> 140073Date: Thu Nov 20 13:51:15 2008 +1000 140074 140075 dix: memset DeviceVelocityPtr to zero. 140076 140077 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140078 Signed-off-by: Adam Jackson <ajax@redhat.com> 140079 140080commit 7f818776b7a19bd6100596d327ecaa69be317bc6 140081Author: Simon Thum <simon.thum@gmx.de> 140082Date: Wed Nov 19 15:05:50 2008 +1000 140083 140084 dix: fix typos in comments, one formatting fix. 140085 140086 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140087 140088commit 037ec5a3c496b07f885a128828ef804b3aa8eee2 140089Author: Peter Hutterer <peter.hutterer@redhat.com> 140090Date: Wed Nov 19 15:12:04 2008 +1000 140091 140092 gitignore: ignore .patch files 140093 140094commit ace38fafb062372dcd3d56378b5b8f86525c6241 140095Author: Peter Hutterer <peter.hutterer@redhat.com> 140096Date: Tue Nov 18 08:19:45 2008 +1000 140097 140098 xfree86: without CONFIG_HAL, Auto{Add|Enable}Devices and AEI is false. 140099 140100 There's little chance that we'll get the input devices at runtime without HAL, 140101 we might as well force the server to add mouse/kbd devices automatically - 140102 just like in the olden days. 140103 140104 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140105 140106commit 1cd894173ee7fa3d1a78d330c12453ea53e7b908 140107Author: Alan Coopersmith <alan.coopersmith@sun.com> 140108Date: Wed Nov 19 13:44:26 2008 -0800 140109 140110 Always use server-<display>.xkm to avoid races when multiple servers start 140111 140112 Previously each server starting ran xkbcomp with the output set to 140113 <keymapname>.xkm, read it, then deleted it - which led to races if 140114 two servers were starting at the same time with the same keymap. 140115 140116 Sun bug #6773816 Xorg uses the same xkm output file for compiled keymap file 140117 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6773816> 140118 140119commit 1e4412abc499b9938bbac8a3ecd82f7f12d448d9 140120Author: Rémi Cardona <remi.cardona@lri.fr> 140121Date: Mon Nov 17 09:56:49 2008 +0100 140122 140123 drop unused HAVE_LIBDRM_2_2 140124 140125 Signed-off-by: Julien Cristau <jcristau@debian.org> 140126 140127commit c4ec71b405fcb5132d5089e3f8dd7b00c235265a 140128Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140129Date: Sun Nov 16 17:57:46 2008 -0800 140130 140131 XQuartz: pbproxy: nuke RCS Id, update header license/copyright 140132 (cherry picked from commit 477c6968a88429bb9c10222a5836bdc936d10ab1) 140133 140134commit 42f330cf9ef44aea7428a67d282dfb1a60196e55 140135Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140136Date: Sun Nov 16 10:42:03 2008 -0800 140137 140138 XQuartz: pbproxy: Remove thread-main.m 140139 140140 it got accidently added and isn't used. 140141 (cherry picked from commit cb6ffce68a2d10022fa92725d2d1837da4b3b89f) 140142 140143commit 19aac4274bbac55ce944acfaf3274a4403fe392e 140144Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140145Date: Wed Nov 19 10:55:29 2008 -0800 140146 140147 mi: Push screen-saver/DPMS handling to after the EQ pop operation. 140148 140149 This way we on't need to hold the mutex during the dixSaveScreens() call. 140150 140151 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 140152 Signed-off-by: Tiago Vignatti <vignatti@c3sl.ufpr.br> 140153 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140154 140155commit 56d1793c435abbaababab031860f1160f09fff06 140156Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140157Date: Wed Nov 19 10:54:57 2008 -0800 140158 140159 mi: Only increment tail (push) when the event data is actually in the queue 140160 140161 Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com> 140162 Signed-off-by: Tiago Vignatti <vignatti@c3sl.ufpr.br> 140163 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140164 140165commit 570835b9a89187a36e08defbccc56051bb349b52 140166Author: Alan Coopersmith <alan.coopersmith@sun.com> 140167Date: Wed Nov 19 10:19:20 2008 -0800 140168 140169 Use bsd_pci stub routines on Solaris too, since we use pciaccess for real work 140170 140171 Depends on commit 70e18a3b6b9d52169bca650f6cd4ef5d8ab40d95 to get 140172 definition of xf86InitVidMem on Solaris. 140173 140174commit 3932a848572f4eaf8b7f1d91d9b74aeafab069a2 140175Author: Keith Packard <keithp@keithp.com> 140176Date: Tue Nov 18 14:21:01 2008 -0500 140177 140178 Xinput: use floats for ConstantDeceleration and AdaptiveDeceleration 140179 140180 These values need not be constrained to integer values. 140181 140182 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140183 Signed-off-by: Adam Jackson <ajax@redhat.com> 140184 140185commit 08cd361234ed0410f67342f46ae01120c4fe3331 140186Author: Dave Airlie <airlied@redhat.com> 140187Date: Mon Nov 17 10:28:48 2008 +1000 140188 140189 exa: avoid doing prepare/done without intervening copies in exaFillRegionTiled 140190 140191 This does a precursor check to make sure the copies are required before 140192 entering the prepare/done code. 140193 140194commit 8f8a9c19ad58768b07461a3f4bccea98f7c4f958 140195Author: Dave Airlie <airlied@redhat.com> 140196Date: Mon Nov 17 10:24:39 2008 +1000 140197 140198 EXA: avoid copy operations if no boxes in use 140199 140200 Simple fix for now, I'm sure damage shouldn't be calling us with nbox = 0. 140201 140202commit 03f7a66a1ae3f522a49bcc783066d53d69c97262 140203Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140204Date: Sat Nov 15 19:45:49 2008 -0800 140205 140206 XQuartz: Don't use TIS on Tiger... 140207 (cherry picked from commit be8ac84c15e7e76c581b9e75cda74139394aa4c7) 140208 140209commit e8112784e8ef80a96ed54eca7e148b424001a38f 140210Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140211Date: Sat Nov 15 19:52:34 2008 -0800 140212 140213 XQuartz: Integrated xpbproxy into the server. 140214 140215 Added a configure option called --enable-standalone-xpbproxy which is useful for deveoping xpbproxy. 140216 The 'active' switch in preferences just disables the in-server xpbproxy (not this standalone). 140217 (cherry picked from commit 42944936326ef8732f622db9f75b79a92980550d) 140218 140219commit 24928b85dc996c8d0a052ddf383df25718ee5305 140220Author: Michel Dänzer <michel@tungstengraphics.com> 140221Date: Sat Nov 15 13:28:53 2008 +0100 140222 140223 DRI2: Don't crash in DRI2ClipNotify when DRI2DrawablePtr is NULL. 140224 140225commit e5271c374715dd4e9a724ab3e5f77108d4ce8c23 140226Author: Michel Dänzer <michel@tungstengraphics.com> 140227Date: Sat Nov 15 13:27:42 2008 +0100 140228 140229 Set libdrm flags correctly even when only DRI2 is enabled. 140230 140231commit bc4f804cd86497655f6765b0954e64dc959939e3 140232Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140233Date: Fri Nov 14 15:13:22 2008 -0800 140234 140235 XQuartz: Actually add glcontextmodes.[hc] 140236 (cherry picked from commit 763195d338345e3637d112279bdbe002b6e7791a) 140237 140238commit aedafbea0e1c7bc780ebe7ae878f62525769aca8 140239Author: Ben Byer <bbyer@apple.com> 140240Date: Mon Mar 31 20:21:24 2008 -0700 140241 140242 fix compilation on systems that don't have 140243 XP_EVENT_SPACE_CHANGED defined 140244 (cherry picked from commit 6e833e41e2f3a84f7e5164b918737b3975517c12) 140245 140246commit b0789365719f48c0b477b4703f2072f932246b76 140247Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140248Date: Fri Nov 14 14:57:09 2008 -0800 140249 140250 XQuartz: Initial work on readding GLX to the 1.5 server 140251 (cherry picked from commit a186cab6ae82340770c13b37ff5df731ce2a0ceb) 140252 140253commit 1d7049f8fd3c0798250b9213149d3ce0b37ab77f 140254Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140255Date: Thu Nov 13 12:45:35 2008 -0800 140256 140257 Xquartz: Force sending mouse clicks to AppKit if we're over the menu bar 140258 (cherry picked from commit 7dff93ec1ac92aeaf3b70d5cfe787fa4a28c0dba) 140259 140260commit 3c124832642f1ec3228a57bea3d1eda68a188ff3 140261Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140262Date: Wed Aug 20 10:37:17 2008 -0700 140263 140264 XQuartz: Made X11Application.m a little more tidy. 140265 (cherry picked from commit 3520386261b838196a8918e8bee16bdccbc9781d) 140266 140267commit e917806fc556c2054b2835c8f85770b554c78092 140268Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140269Date: Fri Nov 14 10:56:21 2008 -0800 140270 140271 os/connection: TRANS_NOXAUTH bit comparison brainfart fix. 140272 140273 https://bugs.freedesktop.org/show_bug.cgi?id=18524 140274 140275commit 17bccff4ce905f46c074fd8ee7c6567b1371a0a1 140276Author: Fernando Carrijo <fcarrijo@yahoo.com.br> 140277Date: Fri Nov 14 12:36:20 2008 -0500 140278 140279 [dix] Remove a duplicate statement. 140280 140281 Signed-off-by: James Cloos <cloos@jhcloos.com> 140282 140283commit ba4e08244ed3923eecf26842dfc1df17c696e053 140284Merge: 81fd17f5f 5bad5d2ab 140285Author: Alex Deucher <alexdeucher@gmail.com> 140286Date: Thu Nov 13 15:04:18 2008 -0500 140287 140288 Merge branch 'master' of git+ssh://agd5f@git.freedesktop.org/git/xorg/./xserver 140289 140290commit 81fd17f5f49cdd2c10d0bf3b7ddeb8b5953886a5 140291Author: Alex Deucher <alexdeucher@gmail.com> 140292Date: Thu Nov 13 15:00:30 2008 -0500 140293 140294 Only add default modes if EDID supports continuous-frequency 140295 140296 When an EDID is present, only add the default mode pool if the 140297 continuous-frequency bit is set in the EDID. Should fix bugs 140298 like 18512. 140299 140300commit c232f3d673fb00d7fceb8e82741349d64e5ac0ad 140301Author: Adam Jackson <ajax@nwnk.net> 140302Date: Thu Nov 13 14:58:21 2008 -0500 140303 140304 xf86TargetExact should try harder if there's only one monitor attached. 140305 140306 If there's no preferred mode, but only one monitor, pick the 140307 biggest mode for its aspect ratio, assuming one exists. 140308 140309commit 5bad5d2abea2f51cbf53dc8dd016e90ccb086220 140310Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140311Date: Thu Nov 13 11:32:40 2008 -0800 140312 140313 XQuartz: Fixed XBell() when not using system alert. 140314 (cherry picked from commit aa0e9ab1c7a226cdcb3c5e62be159355a290faf8) 140315 140316commit 0fca8588a238c91580b305c3671567e910150a02 140317Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140318Date: Thu Nov 13 11:10:15 2008 -0800 140319 140320 XQuartz: Added more debugging to handoff spew 140321 (cherry picked from commit acefa7e3c8706ffedc052effd50b36ce10e72c22) 140322 140323commit 47d38218706564d1437dc3ff9d1bba3511db063c 140324Author: George Staplin <gstaplin@apple.com> 140325Date: Wed Nov 12 17:17:25 2008 -0700 140326 140327 XQuartz: GL: Handle the alpha differently when the alpha is equal to GLCAPS_COLOR_BUF_INVALID_VALUE. 140328 140329 This prevents visuals with odd sizes. The machine I use didn't have 140330 this problem, but it shows up on some others. 140331 (cherry picked from commit ed181382ddeb77019577d39b9c06b1cd839e18e4) 140332 140333commit 6b501d7cc55f1d06966cdf3caf36bcfc6cf18939 140334Author: Alex Deucher <alexdeucher@gmail.com> 140335Date: Thu Nov 13 13:57:30 2008 -0500 140336 140337 Add missing newline when printing edid info 140338 140339commit 33edd32702d963802e64d382124eabb9d0352fa8 140340Author: George Staplin <gstaplin@apple.com> 140341Date: Sun Nov 9 18:35:10 2008 -0700 140342 140343 XQuartz: Disable 2 calls to miSetVisualTypesAndMasks that seem to be unnecessary, and possibly wrong (with regard to masks and bits per RGB). 140344 140345 Use the settings queried from the system in xprScreen.c, rather than those 2 calls. 140346 140347 The 2 calls increased the total number of visuals a great deal (when using GLXEXT), 140348 and not all of the visuals were usable with GLX. Some of the visuals aren't usable 140349 with GLX still, such as DirectColor, but that seems to be acceptable based on my 140350 understanding of the manual that states "a subset of visuals are made available 140351 for OpenGL rendering." 140352 (cherry picked from commit 373b8a5f32509722e06d8776109e6c3f06645ee0) 140353 140354commit 1523476afa6657c20eba112e3ee051d45e721164 140355Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140356Date: Sat Nov 8 23:32:40 2008 -0800 140357 140358 XQuartz: CFBundleShortVersionString needs to conform to X.Y.Z as well... sigh 140359 (cherry picked from commit 54d20f6d33c347bcb37a0a8287c2d92fed2b9092) 140360 140361commit 1f301a0051d65296189248b2529616d9b33276b3 140362Author: George Staplin <gstaplin@apple.com> 140363Date: Thu Nov 13 01:27:33 2008 -0700 140364 140365 XQuartz: Restore the logic of AppleWMDoReorderWindow to the same as the 1.4 branch. 140366 140367 It was returning inverted values in comparison to the 1.4 branch. This resulted in 140368 the windows not drawing due to a deep path of: RootlessReorderWindow -> 140369 SCREENREC(pScreen)->imp->DoReorderWindow(winRec) - > xprDoReorderWindow -> 140370 AppleWMDoReorderWindow. 140371 (cherry picked from commit d1d398db762fe7d5720f9b5fafa2bbe8372fceac) 140372 140373commit f5841e96487234df5ead5f5c0fb3c587c418cb46 140374Author: Peter Hutterer <peter.hutterer@redhat.com> 140375Date: Tue Nov 11 22:50:35 2008 +1000 140376 140377 dix: don't store enter/leave and focus semaphores in a devPrivate. 140378 140379 We need them for each window, every time a window is allocated. Storing them 140380 in a devPrivate is the wrong thing to do. 140381 140382 This also removes the unused ENTER_LEAVE_SEMAPHORE_ISSET macro. 140383 140384 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140385 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 140386 140387commit 26f701704b4e536cd91bd8a9f7d2194793471998 140388Author: Peter Hutterer <peter.hutterer@redhat.com> 140389Date: Wed Nov 12 16:53:29 2008 +1000 140390 140391 dix: remove duplicate line in EnterLeaveEvent. 140392 140393 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140394 140395commit 8ff62ea5ba1bc331c886840b2b1e101cb4f2ebed 140396Author: Kristian Høgsberg <krh@redhat.com> 140397Date: Wed Nov 12 15:19:09 2008 -0500 140398 140399 CopySubBuffer expects GL style coordinates. 140400 140401commit d91243beb63008fff2ce6affa7db091786e8c49a 140402Author: Peter Hutterer <peter.hutterer@redhat.com> 140403Date: Fri Nov 7 08:39:38 2008 +1030 140404 140405 dix: formatting fix. 140406 140407commit dde09076c5c0c36a4a9321cd23a2c7922c4c755b 140408Author: Peter Hutterer <peter.hutterer@redhat.com> 140409Date: Wed Nov 5 12:52:03 2008 +1030 140410 140411 dix: remove superfluous variable devgrab 140412 140413 grab == devgrab anyway, this is a leftover from the time when we had two 140414 different grabs per device (core and XI grab). 140415 140416 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140417 140418commit fc3e8bfb82323477b54181c6d546c90f53f1ebd8 140419Author: Peter Hutterer <peter.hutterer@redhat.com> 140420Date: Tue Nov 4 15:27:30 2008 +1030 140421 140422 mi: clean up mieqProcessInputEvents, copy all events before processing. 140423 140424 Copy the EventRec's information into local variables before processing them, 140425 this should make it safer for upcoming threading and also makes it easier to 140426 read. 140427 140428 Simplify the event allocation code from the abyss it was before. 140429 140430 This also fixes a potential bug where a custom handler could scramble the 140431 event before the same -now scrambled- event was then passed through the 140432 master's custom event handler. 140433 140434 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140435 140436commit 31005efb1bb6be5a892ff905f754109dc3fa8fe6 140437Author: Fernando Carrijo <fcarrijo@yahoo.com.br> 140438Date: Mon Nov 10 13:23:45 2008 -0500 140439 140440 input: Remove useless conditional. 140441 140442commit 24383056d0513cc588cee5cee78f1cba2c0fbe89 140443Author: James Cloos <cloos@jhcloos.com> 140444Date: Sat Nov 8 13:04:20 2008 -0500 140445 140446 Don’t recommend fixing HAL if HAL support is disabled. 140447 140448 As reported in http://bugs.freedesktop.org/show_bug.cgi?id=18438 140449 the server suggests reconfiguring HAL if AllowEmptyInput is enabled 140450 and no input devices are known. 140451 140452 Instead of that notice, if HAL is disabled at configure time, 140453 AllowEmptyInput is enabled in the config and no input devices are 140454 found report those facts and recommend disabling AllowEmptyInput. 140455 140456commit b3c7e62664457a3802ebc23785446051624a8c2f 140457Author: James Cloos <cloos@jhcloos.com> 140458Date: Sat Nov 8 12:18:13 2008 -0500 140459 140460 Remove some null statements. 140461 140462 Remove several doubled statement-terminal semicolons. 140463 140464 Reported by Fernando Carrijo. 140465 140466commit 21c116219cd5c6845a0955f2d88fdb5fab5c17cf 140467Author: Michel Dänzer <michel@tungstengraphics.com> 140468Date: Sat Nov 8 01:25:14 2008 +0100 140469 140470 Fix typos which caused exaCompositeRects() to use an incorrect damage region. 140471 140472 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=18065 . 140473 140474 Also, it can bail if either width or height of the bounding box is <= 0, not 140475 only if both of them are. 140476 140477commit 59f9fb4b8c031df69b3592a26b77e744ff4a556e 140478Author: Luc Verhaegen <libv@skynet.be> 140479Date: Fri Nov 7 19:11:11 2008 +0100 140480 140481 XAA PixmapOps: Sync before accessing unwrapped callbacks. 140482 140483 When using any XAAPixmapOps, we call into unknown but freshly 140484 unwrapped callbacks (like fb ones). Unlike the XAA*Fallback calls, 140485 we did so without syncing first, exposing us to all kinds of 140486 synchronisation issues. 140487 140488 I believe that the rendering errors appeared now because *PaintWindow 140489 vanished (e4d11e58), and we just use miPaintWindow instead. This 140490 takes a less direct route to the hw and ends up at 140491 PolyFillRectPixmap, which very often left drawing artifacts. 140492 140493 We now sync accordingly, and no longer get the rendering artifacts i 140494 was methodically reproducing on radeonhd, radeon, unichrome... 140495 140496 Also, in order to allow driver authors to remove extensive syncing 140497 or flushing to hide this issue, create XAA_VERSION_ defines, put 140498 them in xaa.h and bump the patchlevel. 140499 140500 (novell bug #435791) 140501 140502commit 991c88b7542164194be73573e7644164416ea90c 140503Author: Julien Cristau <jcristau@debian.org> 140504Date: Fri Nov 7 17:36:38 2008 +0100 140505 140506 xfree86: xf86SetDepthBpp needs to respect the driver's depth24flags 140507 140508 When setting the depth to 24, leave bpp unset so the logic to pick 140509 a supported value is used instead of ignoring the driver's preference 140510 and forcing 32 bpp. 140511 140512commit d3d6be4948fa19947fd3b03e6694247109cc0ffb 140513Author: Julien Cristau <jcristau@debian.org> 140514Date: Fri Nov 7 18:36:00 2008 +0100 140515 140516 mi: Fix infinite loop on regen when swrast_dri.so is missing 140517 140518 The swrast DRI provider gets pushed on the glx provider stack at every 140519 server generation, so the stack turns into a circular list on regen. 140520 140521 X.Org bug#18388 <https://bugs.freedesktop.org/show_bug.cgi?id=18388> 140522 140523commit f59352881f315a634f60c9aac885b2764b28b167 140524Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140525Date: Thu Nov 6 14:41:55 2008 -0800 140526 140527 XQuartz: Make LS parse our version properly... it didn't like our CFBundleVersion 140528 (cherry picked from commit 980df1b4ff7d4dab9dc1462dc086542265b1fb1b) 140529 140530commit b4ca9dc2393ace2415163094b916f0d17ecba9ee 140531Author: Alan Coopersmith <alan.coopersmith@sun.com> 140532Date: Wed Nov 5 21:58:04 2008 -0800 140533 140534 When HAL returns a NULL property, print "(null)" instead of a NULL pointer 140535 140536 They've promised to fix Solaris printf soon to check for NULL pointers 140537 instead of segfaulting, but that won't help people on existing releases. 140538 140539commit 13d06f5aaf6120c902a323649615c1ce3d1b5359 140540Author: George Peter Staplin <gps@Georges-Workstation.local> 140541Date: Tue Nov 4 12:53:12 2008 -0700 140542 140543 XQuartz: GL: Add capability detection for depth buffers, and multisampling. 140544 140545 GL/capabilities.c: Add handleDepthModes(), and extend 140546 handleRendererDescription() for the various depth and multisampling flags. 140547 140548 Add initialization of the new config options to initConfig(). 140549 140550 GL/capabilities.h: Add depth and multisample config members. 140551 140552 GL/visualConfigs.c: Add depth and multisampling support to the visual config 140553 setup. 140554 (cherry picked from commit f527381eea6a8ae6cd791475b2060d21fcf8efb2) 140555 140556commit 34bb06b292dc8b07d8602941aab3e69a73811314 140557Author: George Peter Staplin <gps@Georges-Workstation.local> 140558Date: Mon Nov 3 19:52:48 2008 -0700 140559 140560 XQuartz: GL: Work around problems with really deep visuals. 140561 140562 GL/capabilities.c: #if 0 (for now) any capabilities above 8 bits per channel, 140563 because they introduce drawing problems. 140564 140565 GL/indirect.c: Comment out some visual setup code that shouldn't be running, and 140566 actually seemed to cause some problems. The current visualConfigs.c code seems 140567 to do a reasonable job of setting up visuals for XQuartz. 140568 140569 GL/visualConfigs.c: Make use of the proper visual .class. Eliminate depth 0. 140570 It seems we really just want 24 for now, and 0 I think was a flaw in the original 140571 code. 140572 (cherry picked from commit 1e5f63f15e13a40a6e69a1505934d10d6990b6a2) 140573 140574commit 39e82a9b1250b23768136f3c397377a97f8b908c 140575Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140576Date: Sun Nov 2 15:52:22 2008 -0800 140577 140578 XQuartz: Replace deprecated API usage thanks to Robery Murphy 140579 (cherry picked from commit 71dd052412400362793f2f0c7c02bf4f4309738c) 140580 140581commit b4db46320a437ba8894cb8cfd4ee3abdcc894b01 140582Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140583Date: Fri Oct 31 20:18:27 2008 -0700 140584 140585 XQuartz: version string updated for 2.3.2_beta3 140586 (cherry picked from commit 81bb4128570b44d779126a8dffcd3c5620747383) 140587 140588commit c9051b684b524549eab6d5b88ee3e195a6f6fbe8 140589Author: Alan Coopersmith <alan.coopersmith@sun.com> 140590Date: Wed Nov 5 18:25:57 2008 -0800 140591 140592 Use OsSignal in Popen/Pclose to avoid SysV signal() stupidity 140593 140594commit d63ea510138c8b6de66184c78cda39ed9981fc1f 140595Author: Alan Coopersmith <alan.coopersmith@sun.com> 140596Date: Wed Nov 5 14:52:29 2008 -0800 140597 140598 Non-Linux OS'es should default to kbd driver, not now-dead keyboard driver 140599 140600commit 446d9443cea31e493d05c939d0128a8116788468 140601Author: Adam Jackson <ajax@redhat.com> 140602Date: Wed Nov 5 11:51:06 2008 -0500 140603 140604 linux: Drain the console fd of data when using evdev for keyboards 140605 140606 Works around a silly bug in the kernel that causes wakeup storms after 140607 too many keypresses. Should fix the kernel bug too, but this at least 140608 keeps the idle wakeup count below 1000/sec. 140609 140610commit 6d21fbf00648307208146aca0837ec63ea490659 140611Author: Adam Jackson <ajax@redhat.com> 140612Date: Wed Nov 5 11:39:46 2008 -0500 140613 140614 kdrive: Bye bye Xvesa 140615 140616commit 9a874a71a791c6110fd57b8a5c083f777a446d0f 140617Author: Egbert Eich <eich@freedesktop.org> 140618Date: Tue Nov 4 21:31:27 2008 +0100 140619 140620 DDX/RandR: When positioning outputs assume width/height = 0 if no modes exist. 140621 140622 When a user specifies the position of an output for which no modes exist 140623 (for whatever reason) assume that the width and height of this output 140624 is 0. The result will be the same as if this output isn't taken into 140625 consideration at all and thus should be sane. It will prevent a segfault 140626 when trying to determine the width and height of a non-existent mode. 140627 140628commit a4d62bbf215894bad8e19d99f7330c637d3d49e3 140629Author: Michel Dänzer <michel@tungstengraphics.com> 140630Date: Tue Nov 4 11:27:53 2008 +0100 140631 140632 AIGLX: Reinstate call to driver texOffsetFinish hook. 140633 140634 It was accidentally lost when factoring out __glXDRIdoReleaseTexImage, so this 140635 is a regression fix and should probably be backported to server-1.5-branch. 140636 140637commit 27f1ad466cd9662d638f4a8e59beafc32a8b2a8f 140638Author: Michel Dänzer <michel@tungstengraphics.com> 140639Date: Tue Nov 4 11:26:03 2008 +0100 140640 140641 AIGLX: Don't truncate offset returned by driver texOffsetStart hook on 32 bit. 140642 140643commit de1a8b68eb10c39a0d4a197107d68d7cc009edb2 140644Author: Peter Hutterer <peter.hutterer@redhat.com> 140645Date: Mon Nov 3 11:30:48 2008 +1030 140646 140647 xkb: extract the correct device in XkbFilterEvents. 140648 140649 If the event is an XI event, we need to work on the correct device, not on 140650 the VCK. 140651 140652 Adds XIGetDevice(event) function to extract the device from an event. 140653 140654commit d97239956667f8181f30271759573b75bf455fbb 140655Author: Peter Hutterer <peter.hutterer@redhat.com> 140656Date: Fri Oct 31 16:40:41 2008 +1030 140657 140658 Purge device-based WindowAccess code. 140659 140660 Really, this was a bad idea. It's not security, the UI features that would 140661 have been cool (e.g. clicking through windows) aren't implemented anyway, and 140662 there's nothing you can't achieve just by using plain XI anyway. 140663 140664 Requires inputproto 1.9.99.6. 140665 140666commit cbc6f983959595aa21c9dd72fac6a7070a650ef7 140667Author: Peter Hutterer <peter.hutterer@redhat.com> 140668Date: Fri Oct 31 14:36:48 2008 +1030 140669 140670 xkb: when faking mouse button events, fake them on the correct devices. 140671 140672 When MouseKeys are activated, keyboard devices may generate fake mouse button 140673 events through XKB. Let's get then running through the appropriate paths, i.e. 140674 as XI events on the correct device. 140675 140676 To make matters more fun, ProcessOtherEvents drops events if the DIX device 140677 state cannot be updated accordingly, i.e. all button events from keyboard 140678 devices. 140679 Hence we need to get the paired MD for the device in XkbDDXFakeDeviceButton, 140680 and post the event through the paired MD (usually the VCP). 140681 140682 Removes now-unused ddxFakeBtn.c. 140683 140684 Note: this patch only half-arsedly fixed button events, motion events are a 140685 more complicated matter. 140686 140687commit 4971315296cbf868dd738c1c0c1c504fcfe1b619 140688Author: Peter Hutterer <peter.hutterer@redhat.com> 140689Date: Mon Nov 3 17:47:36 2008 +1030 140690 140691 dix: clean up GetPointerEvents and GetKeyboardValuatorEvents. 140692 140693 Split into several functions, remove some stale comments. 140694 140695 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140696 140697commit 6bb0e0a53656db6168a053fb51b242a8640c1461 140698Author: Peter Hutterer <peter.hutterer@redhat.com> 140699Date: Fri Oct 24 16:22:09 2008 +1030 140700 140701 xnest: I'm sure xfree works fine, even if we're 64 bit. 140702 140703 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140704 140705commit 67b7e06c4fa4e8dafb64d9f3eb976836d0cc5889 140706Author: Daniel Stone <daniel@fooishbar.org> 140707Date: Fri Apr 25 19:39:48 2008 +0300 140708 140709 XKB: Remove pointless macros 140710 140711 These weren't even being used, which isn't overly surprising, given that 140712 they were already in the struct. 140713 140714 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140715 140716commit d9ff20e323ba9fb309629ac21946e85709625589 140717Author: Daniel Stone <daniel@fooishbar.org> 140718Date: Thu Oct 23 16:48:15 2008 +1030 140719 140720 DIX: Remove unnecessary prototype 140721 140722 We don't call XkbCopyKeymap from getevents.c anymore, so axe it. 140723 140724 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140725 140726commit 064ee458c7183eb91ef2e0f262b312ba1e3ad6e5 140727Author: Daniel Stone <daniel@fooishbar.org> 140728Date: Sat Aug 9 15:54:56 2008 +0300 140729 140730 XKB: Tiny cleanups to _XkbLookupAnyDevice 140731 140732 A couple of coding style cleanups, a warning fix via removing a 140733 now-unused label, and also put an else so we don't spuriously trip a 140734 condition that should admittedly never occur anyway. 140735 140736 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140737 140738commit 97c9e6a7139e07a0761be05034fae42fcd65ba36 140739Author: Daniel Stone <daniel@fooishbar.org> 140740Date: Fri Mar 28 17:10:28 2008 +0200 140741 140742 XKB: Fix thinko, causing warning (erroneously fixed in 5544c51447) 140743 140744 newTypes is a local variable which always has an address. newTypesIn, 140745 on the other hand, might be sus. 140746 140747 See also 5544c51447f551dfc6df64438873a7ce64743976. 140748 140749 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140750 140751commit caf1e6a1c92a2e70de9159f93d6f5b48c202dd81 140752Author: Daniel Stone <daniel@fooishbar.org> 140753Date: Tue Mar 18 10:05:00 2008 +0200 140754 140755 XKB: Move XKBrules.h to the server too 140756 140757 Get our own local copy of this, so we don't have to define XKB_IN_SERVER 140758 anymore. 140759 140760 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140761 140762commit f4036f6ace5f770f0fe6a6e3dc3749051a81325a 140763Author: Daniel Stone <daniel@fooishbar.org> 140764Date: Wed Oct 22 22:40:44 2008 +1030 140765 140766 Remove XEvIE 140767 140768 It's unmaintained and has been broken for quite a while; MPX finally 140769 smashed it completely. 140770 140771 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140772 140773commit 8c3753a0b711b5708b61f64ab4010ca01608d679 140774Author: Daniel Stone <daniel@fooishbar.org> 140775Date: Wed Oct 22 22:29:54 2008 +1030 140776 140777 A bit of whitespace cleanup in include directives. 140778 140779 It's #include <foo>, not # include <foo> 140780 140781 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140782 140783commit 889a8dc02bbd3926eaf16dce8f844dc58558e982 140784Author: Daniel Stone <daniel@fooishbar.org> 140785Date: Wed Oct 22 22:29:54 2008 +1030 140786 140787 Xi: XINPUT has been mandatory for ages 140788 140789 Ditch all #ifdef XINPUT, since it's always built. 140790 140791 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140792 140793commit 4cf2f4172abc4bf848c1726e6dac535713e08f0b 140794Author: Adam Jackson <ajax@redhat.com> 140795Date: Mon Nov 3 12:50:28 2008 -0500 140796 140797 Add xf86MonitorIsHDMI() helper. 140798 140799commit a7951a4dad902edea76a5cd68f833f0e48ae804b 140800Author: Michel Dänzer <michel@tungstengraphics.com> 140801Date: Mon Nov 3 10:00:54 2008 +0100 140802 140803 AIGLX: Allow 2D driver to prevent zero-copy texturing of a pixmap. 140804 140805 The driver can return ~0ULL to achieve this, e.g. if the pixmap doesn't fit 140806 into offscreen storage or if its pixel format isn't supported by the 3D engine 140807 for texturing. 140808 140809 See http://bugs.freedesktop.org/show_bug.cgi?id=17723 or 140810 http://bugs.freedesktop.org/show_bug.cgi?id=12385 . 140811 140812commit f8af66ddb21b4fb1ae5b4f7f852418ca7b0e7aec 140813Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140814Date: Fri Oct 31 20:03:34 2008 -0700 140815 140816 XQuartz: Add option to allow access to the menu in fullscreen mode 140817 (cherry picked from commit 637a9f4bd1ff2b55c870a08ead4940df0f9818e5) 140818 140819commit c5086badf0387d4d5af10882de90a1faa180550f 140820Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140821Date: Fri Oct 31 18:58:57 2008 -0700 140822 140823 XQuartz: LP64 related casting fixes from Bob Murphy 140824 (cherry picked from commit ea71710aaa7166ab510abe70f2dc47942de0ead7) 140825 140826commit 20239a4deebccc9f86586ef6b8ef61311a6fa6a0 140827Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140828Date: Fri Oct 31 18:58:25 2008 -0700 140829 140830 Rootless: LP64 related casting fixes from Bob Murphy 140831 (cherry picked from commit 40441c4e31167932721a79d7ca572caaa1d36e72) 140832 140833commit ecd55bc859eb5d7c639d5a6d531d9df524124247 140834Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140835Date: Fri Oct 31 15:10:34 2008 -0700 140836 140837 XQuartz: displayScreenBounds uses quartzEnableRootless rather than !quartzHasRoot 140838 (cherry picked from commit c883a78ef0c2e7e29437881db85d3815a89ab874) 140839 140840commit 4e762b8aa37b588e6cffcdd4a0152090190005e2 140841Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140842Date: Fri Oct 31 14:39:54 2008 -0700 140843 140844 Rootless: Force initialization of static variable before use. 140845 (cherry picked from commit cece29317269385aa7a0d3047e6f90e6a17d01eb) 140846 140847commit 9f49f5172d10070ebd4a9e72ac3c5227c6f35853 140848Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140849Date: Fri Oct 31 13:00:01 2008 -0700 140850 140851 XQuartz: Silence pointer debug spam 140852 (cherry picked from commit 9ca099e928653b5a0fb745186e17bfcc3d74ccc0) 140853 140854commit 93298c7f8f73bdb9892da0e41d2c2a67719dc736 140855Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140856Date: Fri Oct 31 12:55:46 2008 -0700 140857 140858 XQuartz: Don't map the menu bar to the screen area. 140859 (cherry picked from commit 1115917ada127a1cd6a8b61efe7fd134e2ea85fd) 140860 140861commit 31907986a6d4dc726d8ce0071f5884ce6cfea33b 140862Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140863Date: Fri Oct 31 12:37:32 2008 -0700 140864 140865 XQuartz: Fixed Xplugin thread safety issues, so rea-enable background window check 140866 (cherry picked from commit 5fc73b1141eca8b678e54412ca33afa31a76d088) 140867 140868commit 227d782a1bddfa1393cddd0c73c0288e18645c47 140869Author: George Peter Staplin <gps@Georges-Workstation.local> 140870Date: Thu Oct 30 13:44:55 2008 -0600 140871 140872 XQuartz: GL: Mark visuals that are slower than others, due to lack of acceleration. 140873 140874 This is used in libGL to determine a preferred visual. glxinfo -v will also 140875 report "visualCaveat=Slow" for such visuals. 140876 (cherry picked from commit d5c17f585a240be9f9135975738e62951cde3353) 140877 140878commit 5287602a4161cd036e3125cda09ba750190a0b47 140879Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140880Date: Wed Oct 29 17:20:11 2008 -0700 140881 140882 XQuartz: if 0 out a block to avoid thread-unsafe lockups 140883 (cherry picked from commit 60d7359a67900ad4601dbc73e07bcd8c803100a5) 140884 140885commit 93d7cd31c527afbf26f4e039afb6adbb6a60fab7 140886Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 140887Date: Wed Oct 29 17:18:05 2008 -0700 140888 140889 XQuartz: Purge the old XQUARTZ_EXPORTS_LAUNCHD_FD option 140890 (cherry picked from commit 490756c7fbe0792c4bd6ae434e7a48e0beffa5c7) 140891 140892commit 0a076b6a7f3c97ff712412eee32e013d3bfd7364 140893Author: Peter Hutterer <peter.hutterer@redhat.com> 140894Date: Fri Oct 31 18:32:26 2008 +1030 140895 140896 xkb: remove unused label "out", clean up program flow. 140897 140898commit 971aef64416058665bc903fee39fea1619cd3727 140899Author: Peter Hutterer <peter.hutterer@redhat.com> 140900Date: Fri Oct 31 17:05:40 2008 +1030 140901 140902 dix: FreeDeviceClass() and friends can be static. 140903 140904commit 3c3a47d0f5ce7fb9548a3eefee9e72488d3ac279 140905Author: Peter Hutterer <peter.hutterer@redhat.com> 140906Date: Fri Oct 31 17:00:21 2008 +1030 140907 140908 include: remove unused declaration for RegisterPairingClient(). 140909 140910 Matching functions got removed in 004876355b43fb4d3c2bc5653a7dc1bfd3f985ee. 140911 140912commit 7c5dec9f716962a6e67dba99418061cebe10e1d2 140913Author: Peter Hutterer <peter.hutterer@redhat.com> 140914Date: Fri Oct 31 16:56:49 2008 +1030 140915 140916 dix: remove unused GuessFreePointerDevice(). 140917 140918 Not used since the MD/SD hierarchy was introduced many moons ago. 140919 140920commit 12599af285a610fb429245798330f2317c75ac2c 140921Author: Peter Hutterer <peter.hutterer@redhat.com> 140922Date: Fri Oct 31 16:50:28 2008 +1030 140923 140924 dix: NextFreePointerDevice() can be static. 140925 140926commit 5d0a945a030cd7626046e9f6c8d5c6bd170649d9 140927Author: Peter Hutterer <peter.hutterer@redhat.com> 140928Date: Fri Oct 31 16:29:42 2008 +1030 140929 140930 dix: SwitchCorePointer is long gone, bury it. 140931 140932commit eb014021c2b3604fc67daebd6860b5dc959fabb6 140933Author: Peter Hutterer <peter.hutterer@redhat.com> 140934Date: Fri Oct 31 16:27:58 2008 +1030 140935 140936 dix: PairDevices() can be static, remove from input. 140937 140938 PairDevices() is only ever called in EnableDevice(). 140939 140940commit b6b26560d65d9468c89dd046396d9aeb14e1b227 140941Author: Peter Hutterer <peter.hutterer@redhat.com> 140942Date: Fri Oct 31 16:23:34 2008 +1030 140943 140944 Move EXTENSION_BASE and EXTENSION_EVENT_BASE to misc.h. 140945 140946commit f781a752e642cfe7d08f841b332c3ae507315598 140947Author: Peter Hutterer <peter.hutterer@redhat.com> 140948Date: Fri Oct 31 16:16:18 2008 +1030 140949 140950 Move MAX_DEVICES to misc.h, rename to MAXDEVICES for consistency. 140951 140952commit 245d1c162c7b13c98a9a28dc7ad441366d7e8006 140953Author: Peter Hutterer <peter.hutterer@redhat.com> 140954Date: Fri Oct 31 11:50:47 2008 +1030 140955 140956 xkb: ProcXkbSetCompatMap should do dry-runs, then normal runs. 140957 140958 Was doing only dry-runs, which kinda explains why changing the compat map 140959 didn't really have any effect. 140960 Fallout from e8c2a3d7c996cb41c4c44ba67acae5ff9438fc06. 140961 140962 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 140963 140964commit 70e18a3b6b9d52169bca650f6cd4ef5d8ab40d95 140965Author: Alan Coopersmith <alan.coopersmith@sun.com> 140966Date: Thu Oct 30 20:38:34 2008 -0700 140967 140968 Refactor Solaris mapping code to work with standard shared/vidmem.c 140969 140970 Also merge sun_bios.c into sun_vid.c and upstream Solaris patch to 140971 keep aperture device open, to allow mappings to occur after X server 140972 has given up uid 0. 140973 140974commit 2809a56b4bc7f09f132c70dc39c6695495351c4d 140975Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 140976Date: Thu Oct 30 16:05:40 2008 -0400 140977 140978 xace: Fix an incorrect call to the RECEIVE_ACCESS hook. 140979 140980commit 0f2fd0577fe2740a32d3a2749e2c3582f891417c 140981Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 140982Date: Thu Oct 30 15:35:33 2008 -0400 140983 140984 xselinux: send more specific message types to libaudit. 140985 140986commit 4ce19b4477057a724b548e342d4c6da2b6721824 140987Author: Peter Hutterer <peter.hutterer@redhat.com> 140988Date: Thu Oct 30 16:02:13 2008 +1030 140989 140990 xfree86: fix xf86ScaleAxis once again. 140991 140992 Maybe one day I stop doing stupid patches like 140993 a3a7c12fcf8e4ac1418f9ea53f76091f309a721b. 140994 140995 So, if X < low, reset to low, and _not_ to high. 140996 If X > high, reset to high, and _not_ to low. 140997 140998commit f12d7ad1647d12ccf92291ebc0607defe6a634a1 140999Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141000Date: Tue Oct 28 12:38:05 2008 -0700 141001 141002 XQuartz: Pass along SIGINT and SIGTERM from the stub to X11.app 141003 (cherry picked from commit 183ca5d68b4f34e248749f304ce140de11bd451b) 141004 141005commit 273df54685768f74ed29926e88a63dc9f86744f1 141006Author: George Peter Staplin <gps@Georges-Workstation.local> 141007Date: Mon Oct 27 22:07:00 2008 -0600 141008 141009 XQuartz: pbproxy: Fix a bug that prevented nedit CLIPBOARD from being proxied. 141010 141011 It seems that we were calling copy_completed: too soon. Now we return instead 141012 of falling through to the free_propdata, and copy_completed:. 141013 (cherry picked from commit 5797d5b662817a3369a74a673f4a7c995cde75cf) 141014 141015commit 313d87341aff2396645d2c315086d487c0073d27 141016Author: George Peter Staplin <gps@Georges-Workstation.local> 141017Date: Mon Oct 27 20:44:02 2008 -0600 141018 141019 XQuartz: pbproxy: Work around bugs in some apps that aren't ICCCM compliant. 141020 141021 Eterm ends up setting the type of the property to TARGETS, instead of ATOM. 141022 That's why it wasn't working. 141023 (cherry picked from commit 9be5998ede7427f14dd9597e117b3d6b427ba4e5) 141024 141025commit 9007d3beea2e2d8b271f696d433cb30e32cda401 141026Author: George Peter Staplin <gps@Georges-Workstation.local> 141027Date: Mon Oct 27 16:34:24 2008 -0600 141028 141029 XQuartz: pbproxy: Add code to handle PICT conversion to PNG and JPEG. 141030 141031 This may work, unfortunately I don't have test apps that fail. 141032 141033 The way it works is by using an NSImage class initWithPasteboard: 141034 method, which we then get the TIFFRepresentation of, and convert 141035 to PNG or JPEG. 141036 141037 The TIFFRepresentation uses NSTIFFCompressionNone; which should be 141038 lossless. 141039 (cherry picked from commit 8d048cfa956f4a0860250cc836a6748912b37ad8) 141040 141041commit b742da0b719198a42a086049771792af84f470ca 141042Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141043Date: Sun Oct 26 13:17:49 2008 -0700 141044 141045 XQuartz: pbproxy: fixed blocking 141046 (cherry picked from commit bb330a1b1c480e89727fb75ff5aeb71f4afc5ee3) 141047 141048commit c885110aedc19e551de22a526397d77013c3dfcb 141049Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141050Date: Sun Oct 26 10:22:12 2008 -0700 141051 141052 Xquartz: pbproxy: Some debugging about unhandled mime types... this will spam up quick, but will be useful for a while... 141053 (cherry picked from commit 3647215310f5a52126661b9ec473dcfeed25b40f) 141054 141055commit e7c26964eef6732c9603b26b7396f95443c5ba8a 141056Author: George Peter Staplin <gps@Georges-Workstation.local> 141057Date: Fri Oct 24 19:12:05 2008 -0600 141058 141059 XQuartz: GL: Fix several error handling mistakes in handleRenderDescriptions, that were leftover before the final code evolved. 141060 (cherry picked from commit fe35a3639081dc78ac6cc4123cde37770f6a39fd) 141061 141062commit 21ea9d82aab8ff0e0d1904d188f96f9cfb2b051d 141063Author: George Peter Staplin <gps@Georges-Workstation.local> 141064Date: Fri Oct 24 18:11:46 2008 -0600 141065 141066 XQuartz: GL: Add feature detection of color sizes, and accum sizes. 141067 141068 This gives us a *lot* more visuals, and they should be more accurate. 141069 Some of the visuals may have been wrong before. This may enable more 141070 OpenGL programs to behave correctly. 141071 141072 I also suspect libGL needs a little work to handle some of these correctly. 141073 (cherry picked from commit 98a084ce9c455260ecdb5f8eecb67f33dbb53b69) 141074 141075commit c264826da96ad1859dd112b17eb8aa9e5278478f 141076Author: Peter Hutterer <peter.hutterer@redhat.com> 141077Date: Thu Oct 16 11:22:29 2008 +1030 141078 141079 xfree86: If AEI is on, disable "kbd" and "mouse" devices. 141080 141081 This consists of two parts: 141082 In the implicit server layout, ignore those drivers when looking for a core 141083 device. 141084 141085 And after finishing the server layout, run through the list of devices and 141086 remove any that use mouse or kbd. 141087 141088 AEI is mutually exclusive with the kbd and mouse drivers, so pick either - or. 141089 141090 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 141091 141092commit 6c451859552e1fc78f6589617482f9ff96d7ed8a 141093Author: Peter Hutterer <peter.hutterer@redhat.com> 141094Date: Mon Oct 20 12:19:55 2008 +1030 141095 141096 config: don't add duplicate devices through HAL. 141097 141098 If HAL is restarted, the device list is sent to the server again, leading 141099 first to duplicate devices (and thus duplicate events), and later to a 141100 FatalError "Too many input devices." 141101 141102 dev->config_info contains the UDI for the device. If the UDI of a new devices 141103 is equal to one we already have in the device list, just ignore it. 141104 141105 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 141106 141107commit 102c4dac7c521941f52652152b1660cd7f559d56 141108Author: Julien Cristau <jcristau@debian.org> 141109Date: Sun Oct 26 13:17:31 2008 +0100 141110 141111 x86emu: fix sparse warnings 141112 141113 ops2.c:113:18: warning: Using plain integer as NULL pointer 141114 ops2.c:207:18: warning: Using plain integer as NULL pointer 141115 141116commit ffaaa1a198a77eb6800c08d4613ee1cc0b068ba0 141117Author: Julien Cristau <jcristau@debian.org> 141118Date: Sun Oct 26 13:13:21 2008 +0100 141119 141120 xfree86: fix compiler warnings in DoModalias() 141121 141122 The precedence of == is higher than that of &, so that code was 141123 probably buggy. 141124 141125 xf86Init.c: In function 'DoModalias': 141126 xf86Init.c:300: warning: suggest parentheses around comparison in operand of & 141127 xf86Init.c:304: warning: suggest parentheses around comparison in operand of & 141128 xf86Init.c:308: warning: suggest parentheses around comparison in operand of & 141129 141130commit 66fd05acd6118102aaaebc230d5c03e0758c6089 141131Author: Julien Cristau <jcristau@debian.org> 141132Date: Sun Oct 26 13:11:41 2008 +0100 141133 141134 xfree86: Silence compiler warnings about function prototypes 141135 141136 xf86Init.c:136: warning: function declaration isn't a prototype 141137 xf86Init.c:243: warning: function declaration isn't a prototype 141138 xf86Init.c:249: warning: function declaration isn't a prototype 141139 141140commit 4b96ba8ddf36f84d4b8c0f2964535bd90477a742 141141Author: Julien Cristau <jcristau@debian.org> 141142Date: Sun Oct 26 11:56:35 2008 +0100 141143 141144 Nuke unused miBankNewSerialNumber 141145 141146commit e85d5d707d51d6a1558dfcf0ebd50183b84be39d 141147Author: Julien Cristau <jcristau@debian.org> 141148Date: Sun Oct 26 11:55:50 2008 +0100 141149 141150 Ansify declaration of FontToXError() 141151 141152commit 30a7a06a9e91a91b8517f67897831c1407713e57 141153Author: Julien Cristau <jcristau@debian.org> 141154Date: Sun Oct 26 11:38:44 2008 +0100 141155 141156 Nuke prototype for nonexistent function 141157 141158 miDoInitVisuals was removed in commit 141159 94825ad3c72a1c3f6a61199c302841f63241cf87, but its prototype was left. 141160 141161commit 5244690a5e0285b6243ee1f28ae6c8287a32b73d 141162Author: Julien Cristau <jcristau@debian.org> 141163Date: Sun Oct 26 11:37:11 2008 +0100 141164 141165 Add prototypes for strlcpy/strlcat 141166 141167commit eadf90aa65d1ea63ddcd74e60e66a79ad567c7e8 141168Author: Julien Cristau <jcristau@debian.org> 141169Date: Sun Oct 26 11:20:33 2008 +0100 141170 141171 xaa: Fix compiler warning (implicit declaration) 141172 141173 xaaInit.c: In function 'XAAInit': 141174 xaaInit.c:201: warning: implicit declaration of function 'miInitializeCompositeWrapper' 141175 xaaInit.c:201: warning: nested extern declaration of 'miInitializeCompositeWrapper' 141176 141177commit aee514cfb5e5a59b7770699765ccdad8fe42f904 141178Author: Julien Cristau <jcristau@debian.org> 141179Date: Sun Oct 26 11:16:05 2008 +0100 141180 141181 xfree86: Fix compiler warnings (implicit declarations) 141182 141183 Add missing includes to fix the following warnings: 141184 xf86DGA.c: In function 'DGAProcessKeyboardEvent': 141185 xf86DGA.c:1050: warning: implicit declaration of function 'UpdateDeviceState' 141186 xf86DGA.c:1050: warning: nested extern declaration of 'UpdateDeviceState' 141187 xf86Xinput.c: In function 'xf86ActivateDevice': 141188 xf86Xinput.c:303: warning: implicit declaration of function 'AssignTypeAndName' 141189 xf86Xinput.c:303: warning: nested extern declaration of 'AssignTypeAndName' 141190 xf86Xinput.c:311: warning: implicit declaration of function 'DeviceIsPointerType' 141191 xf86Xinput.c:311: warning: nested extern declaration of 'DeviceIsPointerType' 141192 xf86Xinput.c:324: warning: implicit declaration of function 'XkbSetExtension' 141193 xf86Xinput.c:324: warning: nested extern declaration of 'XkbSetExtension' 141194 141195commit add946daed531e29935593fb2e7ee70286185baf 141196Author: Julien Cristau <jcristau@debian.org> 141197Date: Sun Oct 26 11:14:40 2008 +0100 141198 141199 Nuke unused variables 141200 141201commit f2bda61c03f64133ef68dd1557b26ae7dc5beb0e 141202Author: Julien Cristau <jcristau@debian.org> 141203Date: Sun Oct 26 11:05:13 2008 +0100 141204 141205 Xext: Fix compiler warnings 141206 141207 Make GEClientGone static and include registry.h, to fix the following 141208 warnings: 141209 geext.c:225: warning: no previous prototype for 'GEClientGone' 141210 geext.c: In function 'GEExtensionInit': 141211 geext.c:280: warning: implicit declaration of function 'RegisterResourceName' 141212 geext.c:280: warning: nested extern declaration of 'RegisterResourceName' 141213 141214commit 24e863b0eb6ff11010a14cfd252a39df87a09d0e 141215Author: Bernhard R. Link <brlink@debian.org> 141216Date: Sun Oct 26 00:19:34 2008 +0200 141217 141218 Bug#16501: autodetection of sbus graphic cards 141219 141220 Add automatic detection of the graphic driver to load for sbus devices. 141221 This allows xorg to work on those devices without a "Device" section. 141222 141223 Debian bug#483942. 141224 141225 Signed-off-by: Julien Cristau <jcristau@debian.org> 141226 141227commit fd97ff1bdd30c358525068a4c833f681e00719a2 141228Author: Julien Cristau <jcristau@debian.org> 141229Date: Sat Oct 25 22:54:15 2008 +0200 141230 141231 Bug#5851: Document the "Log" option in xorg.conf(5) 141232 141233commit 80b6768536fa439bfc5ce948bc8845a6f4af857b 141234Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141235Date: Thu Oct 23 17:32:32 2008 -0700 141236 141237 XQuartz: Do check_xinitrc() later to avoid possible deadlock. 141238 (cherry picked from commit af95174a088dc787d124ec548b1005f5b67c9e8d) 141239 141240commit d837d4e2af35a83c7932e4d0422097ebd1f6ba2c 141241Author: George Peter Staplin <gps@Georges-Workstation.local> 141242Date: Thu Oct 23 17:56:46 2008 -0600 141243 141244 XQuartz: pbproxy: pbproxy.h had invalid macro usage with 2 #else statements, so it wasn't building. Fix the build. 141245 (cherry picked from commit 6bcbb9ba85d95d7df63811a482abc39c4d595366) 141246 141247commit 6a2693b2f9cfdec6f30e2ca4c4176f986f97249e 141248Author: George Peter Staplin <gps@Georges-Workstation.local> 141249Date: Thu Oct 23 17:40:23 2008 -0600 141250 141251 XQuartz: Remove the enable_stereo default. We now use feature detection to enable stereo visuals. 141252 141253 We need to revisit the GL/indirect.c changes. I think indirect.c should 141254 be using the same feature detection and GLX visual config code. The 141255 indirect changes will require testing... 141256 (cherry picked from commit 49b7a7c6f20ca85d4d624444eb83db2f91344591) 141257 141258commit b015a021b28623b7b734833dc9b8269c41db4e8a 141259Author: Alan Coopersmith <alan.coopersmith@sun.com> 141260Date: Tue Oct 21 20:30:35 2008 -0700 141261 141262 Allow non-gcc compilers to build xf86Xinput.c 141263 141264commit b7cc6ca22a4ff7a94abee6962a07fdd0b563d26a 141265Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141266Date: Thu Oct 23 16:33:11 2008 -0700 141267 141268 miext/rootless: Silence some warnings by using proper C prototypes. 141269 (cherry picked from commit 050e46e09efcc7e6f090fa4a749e9076c972b760) 141270 141271commit b397cc727787a854417b11b482cce4807fc8ab58 141272Author: George Peter Staplin <gps@Georges-Workstation.local> 141273Date: Thu Oct 23 16:19:58 2008 -0600 141274 141275 miext/rootless: Fix the usage of DeleteProperty so that this builds again. 141276 (cherry picked from commit 3de250e28a9a86bcae4464a15c2025805877ce1e) 141277 141278commit 28d4f71193d2be26d1dcae724067539ac9db1123 141279Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141280Date: Thu Oct 23 15:01:03 2008 -0700 141281 141282 XQuartz: pbproxy: Fix compilation on case-sensitive file systems... 141283 (cherry picked from commit 6622efca216663c99d112c7a226bde691d1f3215) 141284 141285commit 40f9b6bf0dc62098680235a958e0733770902cca 141286Author: Adam Jackson <ajax@redhat.com> 141287Date: Thu Oct 23 17:33:52 2008 -0400 141288 141289 Argh, macros. 141290 141291commit 8b3c278fe7502f9343355f09213837622b0d9ca0 141292Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141293Date: Thu Oct 23 11:39:04 2008 -0700 141294 141295 Rootless: Properly typed expose_1 141296 (cherry picked from commit 78e874fe17174ac8f61586573edaf7f16ece7b21) 141297 141298commit 3e999d35c26b0ffb894c257a3057d44b949744bd 141299Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141300Date: Thu Oct 23 11:37:13 2008 -0700 141301 141302 XQuartz: Honor the Spaces preference for "When switching to an application, switch to a space with open windows for the application" 141303 (cherry picked from commit bf561a06b2bf60395d26026ac7627d0cb254bc07) 141304 141305commit 75067d12e4f4daa080dfb0209a9abf9c5654271f 141306Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141307Date: Tue Oct 21 19:36:48 2008 -0700 141308 141309 XQuartz: Added missing semicolons to Localizable.strings 141310 (cherry picked from commit e9fe3f36d9529f00daeefa1379cdd6f01a88f410) 141311 141312commit a003932dccc438c4500d383982cee2376e808bd5 141313Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141314Date: Tue Oct 21 11:23:42 2008 -0700 141315 141316 XQuartz: Default dpi to 96 instead of 75 141317 (cherry picked from commit 08a5a333d7fea68146cd9cf50682b811378ab3ce) 141318 141319commit 2d764e21f278ea26bae135db6dae07b94651951c 141320Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141321Date: Tue Oct 21 11:19:06 2008 -0700 141322 141323 XQuartz: pbproxy: SIGHUP reloads preferences. 141324 (cherry picked from commit cbc0935ef74ebd76ed3db16d6ab86043eeed42a3) 141325 141326commit 0b695262176e96f508921905ee7f9cc086c7daf0 141327Author: Adam Jackson <ajax@redhat.com> 141328Date: Thu Oct 23 15:01:58 2008 -0400 141329 141330 Change some #define foo() -> static void foo() 141331 141332commit aab73d719a8de6b830e47c7f47860b7990ea8d03 141333Author: Adam Jackson <ajax@redhat.com> 141334Date: Thu Oct 23 13:20:59 2008 -0400 141335 141336 Fix previous cherry pick for API changes 141337 141338commit 8767fc8d47276c7489ae50dd556a446a462776ea 141339Author: Keith Packard <keithp@keithp.com> 141340Date: Fri Apr 4 12:11:14 2008 -0700 141341 141342 Wrap AddTraps in exa and damage. 141343 141344 This fine (and unused) function wasn't ever wrapped which made it not work 141345 under exa. 141346 141347 (cherry picked from commit 06e7e1d0486e8c516a9b3219a2c86026f88825fc) 141348 141349commit 4c926dbac65ceb699b91bb0d3fd0005a5854572d 141350Author: Peter Hutterer <peter.hutterer@redhat.com> 141351Date: Thu Oct 23 22:07:37 2008 +1030 141352 141353 dix: fix two more endian issues, correct initial "enabled" value. #18111 141354 141355 EnableDevice and DisableDevice both change the property too. 141356 141357 And enabled must be set to FALSE in AddInputDevice, the device is not enabled 141358 yet. 141359 141360 X.Org Bug 18111 <https://bugs.freedesktop.org/show_bug.cgi?id=18111> 141361 141362commit 084ae9e38812f19751a086eea7ce555761ae5a17 141363Author: Jerome Glisse <glisse@freedesktop.org> 141364Date: Thu Oct 23 11:28:53 2008 +0200 141365 141366 dri2: fail at context creation if driver fail to create it's context 141367 141368commit ff60e592f003d8bc45789a1a14e86be22f5242b2 141369Author: Peter Hutterer <peter.hutterer@redhat.com> 141370Date: Thu Oct 23 09:19:41 2008 +1030 141371 141372 dix: don't allow VCP/VCK be disabled through properties. 141373 141374commit 98f01c2abe4771d76febf8fe70111b2bddfab776 141375Author: Peter Hutterer <peter.hutterer@redhat.com> 141376Date: Thu Oct 23 09:13:41 2008 +1030 141377 141378 dix: fix endianess issue in AddInputDevice. #18111 141379 141380 dev->enabled is a Bool. Bool is two bytes. 141381 BOOL on the other hand is a protocol type and always 1 byte. So copy the value 141382 into the one-byte type before passing it into XIChangeDeviceProperty. 141383 141384 Found by Michel Dänzer. 141385 141386 X.Org Bug 18111 <http://bugs.freedesktop.org/show_bug.cgi?id=18111> 141387 141388commit f6cbe0326c57e3c9897f17baef78199b06537c44 141389Author: Peter Hutterer <peter.hutterer@redhat.com> 141390Date: Wed Oct 22 18:16:12 2008 +1030 141391 141392 dix: init remaining three xkb fields to zero in a new client. 141393 141394commit 0b56b44addc323a00eb7cd86240cb0dd4275bcf8 141395Author: Peter Hutterer <peter.hutterer@redhat.com> 141396Date: Mon Oct 20 17:06:40 2008 +1030 141397 141398 xfree86: AllowEmptyInput is true by default - update the xf86Info defaults. 141399 141400 Also set AutoAddDevices and AutoEnableDevices to their defaults. 141401 141402 And in doing so, switch the rest of the defaults over to named intializers. 141403 141404 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 141405 141406commit d72cd753b99fae147ef4c189700fc697f1ea7fb0 141407Author: Matthieu Herrb <matthieu.herrb@laas.fr> 141408Date: Tue Oct 21 22:32:57 2008 +0200 141409 141410 Close well known connections in ServerAbort() 141411 141412commit 81e197b2a5457bb9f7ed72f82c2d33fd8dbfc202 141413Author: Adam Jackson <ajax@redhat.com> 141414Date: Tue Oct 21 16:00:32 2008 -0400 141415 141416 Bug #18159: Spell "anisotropic" correctly 141417 141418commit 3891dd892449fcdb7a514e3c5e7e763ba7e74003 141419Author: Adam Jackson <ajax@redhat.com> 141420Date: Tue Oct 21 13:10:44 2008 -0400 141421 141422 exa: Add exaDrawableIsOffscreen() to the driver API. 141423 141424commit c837d182c4674c01296bb7efd7d834d9dfd55f83 141425Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141426Date: Tue Oct 21 09:11:59 2008 -0700 141427 141428 XQuartz: Only send mouse events while in the background if we're over a visible window. 141429 141430 This "breaks" xeyes but gets rid of annoying tooltips. 141431 (cherry picked from commit 10090a132a0110fe24c8c886d9f53e6e947f8038) 141432 141433commit 2a9613980f93959eedf9396735785d4c50546e2d 141434Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141435Date: Tue Oct 21 09:07:08 2008 -0700 141436 141437 XQuartz: More fullscreen fixes 141438 Start in rooted mode when enabled 141439 Make the hotkey cmd-alt-a behave 141440 (cherry picked from commit 22a54e46176c9cbdfaaa59ef33a6ae6a9c4969ef) 141441 141442commit e4ea1494debda4d3da03864c2640921fe0430694 141443Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141444Date: Mon Oct 20 12:23:04 2008 -0700 141445 141446 XQuartz: Cleaned up keyboard init and map reload. 141447 (cherry picked from commit f78c9fc06cac2cc6ddfd6e9ba435dd26a57d1f51) 141448 141449commit ebb2e1449cfb382a806e04c52d42240670717069 141450Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141451Date: Mon Oct 20 09:46:34 2008 -0700 141452 141453 XQuartz: Fixed follow-keyboard-layout 141454 ≈ 141455 (cherry picked from commit a9f9fbf512c9fd5a773fce402182486edc71d5ab) 141456 141457commit f67cd31d2cf1b8ad850bd21dbed5a975d77b4c54 141458Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141459Date: Tue Oct 21 09:29:37 2008 -0700 141460 141461 XQuartz: More progress towards fullscreen... 141462 (cherry picked from commit 628cae86b4ad02c49d20df1e1afdcb7abcabb649) 141463 (cherry picked from commit b47b7ab6d6d0a7f6cbf678a5755124d6b7516d3e) 141464 141465commit d434ed400986653771bbd60c224a2275bd6147ef 141466Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141467Date: Fri Oct 17 09:25:26 2008 -0700 141468 141469 XQuartz: Fixed missing key-equiv for Hide-Others. 141470 (cherry picked from commit 2c4b0f9615b646712609ebef3c0851c9a22e781b) 141471 141472commit 3695f62d072e7864fdcee385c8b5ca40f2018101 141473Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141474Date: Thu Oct 16 22:36:00 2008 -0700 141475 141476 XQuartz: Require applewmproto-1.1.1 141477 (cherry picked from commit ed235e3406e603a8a92c4b72f9f9018bae72ad65) 141478 141479commit f935c11398b93860461672e88814026d75b71689 141480Author: George Peter Staplin <gps@Georges-Workstation.local> 141481Date: Thu Oct 16 13:09:21 2008 -0600 141482 141483 XQuartz: pbproxy: Fix 2 bugs with the preferences. 141484 141485 Add dump_prefs() for instrospection. 141486 141487 Make prefs_get_bool a little more clear. 141488 141489 In get_property don't assume that the Mac realloc(NULL, 0); returns 141490 non-NULL. We shouldn't depend on implementation-defined behavior too 141491 much. 141492 141493 Add more commentary in various spots. 141494 141495 Add TARGETS in the TARGETS response. I don't know why some X11 apps 141496 do this, but it could be something that some weird toolkit/app tests 141497 for. 141498 141499 In reload_preferences add preferences synchronization, so that xpbproxy 141500 reloads the preferences, and picks up any changes. 141501 141502 Fix a typo/misspelling of the sync_clipboard_to_pasteboard preference. 141503 (cherry picked from commit 165cbbb90c793d9335ea76f0274b1f60ad295903) 141504 141505commit 4d978af5c71430f0043b1272f8dc6344b8f03f12 141506Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141507Date: Wed Oct 15 21:26:58 2008 -0700 141508 141509 XQuartz: Fixed window levels. 141510 (cherry picked from commit ef4d37c73f31048dd74efee2c63fa08b92608f87) 141511 141512commit 58eacf0b7beb8c6f1971570ede0bfd3976a84665 141513Author: Rémi Cardona <remi@gentoo.org> 141514Date: Tue Oct 21 10:07:24 2008 -0400 141515 141516 Force LC_ALL=C when building xf86DefModeSet.c 141517 141518 Fixes build in not-quite-latin locales. 141519 141520commit 2188582e5ea90edb432a2f421d0a267439ba08f9 141521Author: Michel Dänzer <michel@tungstengraphics.com> 141522Date: Mon Oct 20 09:55:24 2008 -0400 141523 141524 EXA: Avoid excessive syncing in PutImage 141525 141526commit 8d0cb507ae568c8da92f10ff1188d797bcaa3d1a 141527Author: Peter Hutterer <peter.hutterer@redhat.com> 141528Date: Fri Oct 17 09:59:29 2008 +1030 141529 141530 dmx: avoid hilarious #define Xcalloc xcalloc game. 141531 141532commit 8a23707b866c11807e01056db97afec56e9b19c8 141533Author: Peter Hutterer <peter.hutterer@redhat.com> 141534Date: Fri Oct 17 09:15:39 2008 +1030 141535 141536 dmx: change remainder of manual EQ allocation to GetEventList(). 141537 141538commit 8de26770a41ec87c46eed2eddfde6f867d71fe1f 141539Author: Pierre Willenbrock <pierre@pirsoft.de> 141540Date: Thu Oct 16 14:28:14 2008 -0400 141541 141542 RANDR: Fix output property event delivery. 141543 141544commit d936a4235c9625bd41569cef3452dd086284e0d7 141545Author: Peter Hutterer <peter.hutterer@redhat.com> 141546Date: Sun Oct 12 21:58:30 2008 +1030 141547 141548 xfree86: if AllowEmptyInput is true, enable RAW mode on the console. 141549 141550 Usually, the console is set to RAW in the kbd driver. If we hotplug all input 141551 devices (i.e. the evdev driver for keyboards) and the console is left as-is. 141552 As a result, the evdev driver must put an EVIOCGRAB on the device to avoid 141553 characters leaking onto the console. This again breaks many things, amongst 141554 them lirc, in-kernel mouse button emulation and HAL. 141555 141556 This patch sets the console to RAW if AllowEmptyInput is on. 141557 141558 Use-cases: 141559 1. AEI is off 141560 1.1. Only kbd driver is used - behaviour as-is. 141561 1.2. kbd and evdev driver is used: if evdev does not grab the device, 141562 duplicate events are generated. 141563 2. AEI is on 141564 2.1. Only evdev driver is used - behaviour as-is, but evdev does not need 141565 to grab the device anymore. 141566 2.2. evdev and kbd are used: duplicate key events are generated if evdev 141567 does not grab the device. 141568 141569 1.2 is a marginal use-case that can be fixed by adding a "grab" option to the 141570 evdev driver (update of xorg.conf is needed). 141571 141572 2.2 is an issue. If we have no ServerLayout section, AEI is on, but devices 141573 specified in the xorg.conf are still added [1], resulting in duplicate events. 141574 This is a common configuration and needs sorting out. 141575 141576 [1] 2eaed4a10fe5bf727579bca4ab8d4a47c8763a7d 141577 141578 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 141579 Signed-off-by: Adam Jackson <ajax@redhat.com> 141580 141581commit 84ef8ed6fbefd8d6c0aaa3c862879f9804299bd8 141582Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141583Date: Wed Oct 15 14:11:18 2008 -0700 141584 141585 XQuartz: implemented primary-on-grab and fixed clipboard-to-pasteboard 141586 (cherry picked from commit bcb83eea729a01026d99d1cfc2b77385b5b275fd) 141587 141588commit 0195d318465d5a6a9039091bcb252202471df5a4 141589Author: Luc Verhaegen <libv@skynet.be> 141590Date: Wed Oct 15 21:46:27 2008 +0200 141591 141592 GLX: fix build when dri2 _is_ available. 141593 141594 Caught by Julien Cristau. 141595 141596commit 398cdf959afadfa964aece13cc9d89b2dd589930 141597Author: Luc Verhaegen <libv@skynet.be> 141598Date: Wed Oct 15 18:24:10 2008 +0200 141599 141600 GLX: fix build when dri2 is not available. 141601 141602commit 2e60ebbd254290e19492b97c960d9678513c94bf 141603Author: Daniel Stone <daniel@fooishbar.org> 141604Date: Wed Oct 15 14:10:51 2008 +0100 141605 141606 XFree86: Simplify DontZap/-retro interaction 141607 141608 How hard could it possibly be, right? 141609 141610commit 66a87517bd80b21e107df9d57968d81a92f91fd5 141611Author: Alan Hourihane <alanh@tungstengraphics.com> 141612Date: Wed Oct 15 12:00:16 2008 +0100 141613 141614 re-enable DRI2 141615 141616commit ced6690284fd334f225bbc94685c114ad3ab96d8 141617Author: Kristian Høgsberg <krh@redhat.com> 141618Date: Tue Oct 7 13:49:28 2008 -0400 141619 141620 dri2: Update to latest protocol draft. 141621 141622 Mainly rename SwapBuffers to CopyRegion, which adds the xfixes region 141623 argument and the bitmask argument to let us extend it in the future. 141624 141625commit 87a016ae00feac3fbaa7e7a518076a3852d49554 141626Author: Maarten Maathuis <madman2003@gmail.com> 141627Date: Wed Oct 15 00:02:13 2008 +0200 141628 141629 exa: restore {x,y}{Src,Dst} to their original values when !done 141630 141631commit b677bff97b62c3931df853bae25ab485892ae173 141632Author: Adam Jackson <ajax@redhat.com> 141633Date: Tue Oct 14 14:30:19 2008 -0400 141634 141635 int10: Degrade the checksum warning to X_INFO. 141636 141637 I've seen about one case in three years where this has actually been 141638 correlated with the real cause of failure, and we've trained people to 141639 freak out about X_WARNING, so let's be less alarmist. 141640 141641commit bd9c6b3a4d726a3f83ac6d8cf7211eddbc28f25a 141642Author: Adam Jackson <ajax@redhat.com> 141643Date: Tue Oct 14 14:04:01 2008 -0400 141644 141645 EDID: Catch even more cases of encoding aspect as size. 141646 141647 Very cute, Samsung, not only do you claim to be 16cm by 9cm in the 141648 global size record, you also claim to be 160mm by 90mm in the detailed 141649 timings. Grrr. 141650 141651commit ad677238bc96a8578113bbe76d605d7a87aca44c 141652Author: Adam Jackson <ajax@redhat.com> 141653Date: Tue Oct 14 13:00:50 2008 -0400 141654 141655 Add backtrace definitions to dix-config.h.in 141656 141657commit 8e368cf5b964f1d29fda0a463f9510457619b14d 141658Author: Nathaniel McCallum <nathaniel@natemccallum.com> 141659Date: Tue Oct 14 09:50:29 2008 -0400 141660 141661 Xorg: add -modalias option 141662 141663 This scans the installed video drivers and prints a Linux-style modalias 141664 listing of the devices each driver claims to support. 141665 141666commit 3eb52de7f28b0050582f9ac4c28bc894d3f06f4b 141667Author: Adam Jackson <ajax@redhat.com> 141668Date: Tue Oct 14 09:45:19 2008 -0400 141669 141670 kdrive: Hardcode a different wrong value for number of mouse buttons 141671 141672 This should really come from the input driver directly, but, kdrive. 141673 141674commit f3f6ea89aa9e0ffe9e37bc059e5e6bf75be4ee9f 141675Author: Peter Hutterer <peter.hutterer@redhat.com> 141676Date: Wed Oct 8 14:12:21 2008 +1030 141677 141678 Xi: check all handlers before applying property changes. 141679 141680 The current code exposes to inconsistent updates, i.e. if handler N succeeds 141681 but handler N+1 fails in setting the property, an error is returned to the 141682 client although parts of the server now behave as if the property change 141683 succeeded. 141684 141685 This patch adds a "checkonly" parameter to the SetProperty handler. The 141686 handlers are then called twice, once with checkonly set to TRUE. 141687 On the checkonly run, handlers _MUST_ return error codes if the property 141688 cannot be applied. Handlers are not permitted to actually apply the changes. 141689 On the second run, handlers are permitted to apply property changes. 141690 Errors codes returned on the second run are ignored. 141691 141692commit ad67e3f063aa79247270f29e989bbfe5f62c9ed7 141693Author: Peter Hutterer <peter.hutterer@redhat.com> 141694Date: Sun Oct 12 10:19:36 2008 +1030 141695 141696 dix: remove duplicate code in ReleaseActiveGrabs 141697 141698 Spotted by Colin Harrison. 141699 141700commit 3fc4f40b6c6cb416c9dc4bdb35c91b4f32c03ccc 141701Author: Aaron Plattner <aplattner@nvidia.com> 141702Date: Sun Oct 12 16:08:26 2008 -0700 141703 141704 Restore xf86{Enable, Disable}GeneralHandler. 141705 141706 These were useful as part of the generic handler ABI, and are used by the NVIDIA 141707 driver. 141708 141709 This reverts part of commit 50081d2dfb79878cb931a15c265f0d60698dfd39. 141710 141711commit 2217d22a76cdb2460f9683a6bf74c7248612889d 141712Author: Aaron Plattner <aplattner@nvidia.com> 141713Date: Sun Oct 12 16:07:24 2008 -0700 141714 141715 Revert "xfree86: xf86{Enable, Disable}InputHandler can be static." 141716 141717 These were potentially useful as part of the input handler ABI, even if nobody 141718 currently uses them. 141719 141720 This reverts commit 278c11f01fbc6d6bd91c5a7127928c9ef5d29fca. 141721 141722commit 5b336585a4cdf11d20831a9536ad581e959ea7f1 141723Author: Maarten Maathuis <madman2003@gmail.com> 141724Date: Fri Oct 10 23:23:02 2008 +0200 141725 141726 dri: don't set the dixPrivate key to NULL, as this is a staticly set variable. 141727 141728 - This breaks on the 2nd server generation. 141729 - No other subsystem seems to NULL their key. 141730 - This should fix bug 17982. 141731 141732commit b736f477f5324f79af30fc0f941ba0714a34ccda 141733Author: Adam Jackson <ajax@redhat.com> 141734Date: Fri Oct 10 16:33:24 2008 -0400 141735 141736 mieq: Backtrace when the queue overflows. 141737 141738 Since we're probably stuck down in a driver somewhere, let's at least 141739 try to point out where. This will need to be rethought when the input 141740 thread work lands though. 141741 141742commit 94ed0ba1b5043ad9fc33b42756af447d5ab15bbd 141743Author: Adam Jackson <ajax@redhat.com> 141744Date: Fri Oct 10 15:53:48 2008 -0400 141745 141746 Move xorg_backtrace() up to the OS level so we can call it from DIX. 141747 141748commit f650c1aeb06e957626bb0a74eddbb133f59dbecb 141749Author: Adam Jackson <ajax@redhat.com> 141750Date: Fri Oct 10 14:07:28 2008 -0400 141751 141752 Revert "Don't leak EDID_block" 141753 141754 This reverts commit 05cdef18ca668ffc86b71033f71cfbd262cdd994. 141755 141756 http://lists.freedesktop.org/archives/xorg/2008-October/039305.html 141757 141758commit 0660dd9d7009147c395b9ea904539f76f55b9a7f 141759Author: Adam Jackson <ajax@redhat.com> 141760Date: Fri Oct 10 13:41:50 2008 -0400 141761 141762 EDID: Catch monitors that encode aspect ratio for physical size. 141763 141764 This is not legal in either EDID 1.3 or 1.4, but hey, when did a little 141765 thing like legality stop anyone. 141766 141767commit 956a4ba40b28251ab2d88b5d5b0796e01afd2bdd 141768Author: Maarten Maathuis <madman2003@gmail.com> 141769Date: Fri Oct 10 00:28:34 2008 +0200 141770 141771 pixmap: Add some notes to devKind and devPrivate. 141772 141773commit 4808bdec45775342eb9a6352b41e4919e1a69279 141774Author: Peter Hutterer <peter.hutterer@redhat.com> 141775Date: Thu Oct 9 16:23:24 2008 +1030 141776 141777 Xi: don't memcpy the KeyClassRec from SD to MD. #16167 141778 141779 Most of its component get copied during CopyKeyClass anyway. 141780 The ones that aren't: 141781 postdown - never changed for virtual devices anyway. 141782 down - shouldn't change that without sending events. 141783 141784 memcpy'ing the struct also copied mapWidth, which means we didn't realloc 141785 during SetKeySymsMap lateron, overwriting the memory assigned to us. 141786 141787 X.Org Bug 16167 <http://bugs.freedesktop.org/show_bug.cgi?id=16167> 141788 141789commit 18cdd733ad3dd36fa0a23dfbf18c7d99b86c0276 141790Author: Peter Hutterer <peter.hutterer@redhat.com> 141791Date: Thu Oct 9 17:26:00 2008 +1030 141792 141793 kdrive: fix dixflags so they're always defined before enqueuing events #17734 141794 141795 If absolute events were posted, dixflags got set conditionally on whether the 141796 valuators are different from the last posted set of values. 141797 If dixflags are undefined however, the DIX interprets them as relative 141798 valuators. Fix this by making sure defining dixflags is always defined. 141799 141800 X.Org Bug 17724 <http://bugs.freedesktop.org/show_bug.cgi?id=17734> 141801 141802commit d58f2c30c64e8b2d8179dac06e0a54be06099da6 141803Author: Peter Hutterer <peter.hutterer@redhat.com> 141804Date: Thu Oct 9 16:50:54 2008 +1030 141805 141806 kdrive: lastx/y needs to be set to the event coordinates #17728 141807 141808 X.Org Bug 17728 <http://bugs.freedesktop.org/show_bug.cgi?id=17728> 141809 141810commit decec14219bcd992ec426e202ff3c8681b520b74 141811Author: Peter Hutterer <peter.hutterer@redhat.com> 141812Date: Thu Oct 9 15:10:21 2008 +1030 141813 141814 dix: silence "unused variable" compiler warning. 141815 141816commit 0b4b683f3eeec43536e73be302a4c396fe4d9894 141817Author: Peter Hutterer <peter.hutterer@redhat.com> 141818Date: Tue Oct 7 18:44:42 2008 +1030 141819 141820 xfree86: export NewInputDeviceRequest/DeleteInputDeviceRequest 141821 141822 NIDR should be used to create a new SD from e.g. within a driver. 141823 DIDR should be used to remove a device from the server. 141824 141825 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 141826 141827commit 5e213fe1022bf96747834a3ac227929b23136ee6 141828Author: Peter Hutterer <peter.hutterer@redhat.com> 141829Date: Thu Oct 9 11:29:11 2008 +1030 141830 141831 xfree86: cache dev->isMaster before freeing the device. 141832 141833commit 05cdef18ca668ffc86b71033f71cfbd262cdd994 141834Author: Zhenyu Wang <zhenyu.z.wang@intel.com> 141835Date: Fri Oct 10 13:45:51 2008 +0800 141836 141837 Don't leak EDID_block 141838 141839commit 2d9da7a5f384d5f38b2be79b1ea0df5a3deb52d1 141840Author: Luc Verhaegen <libv@skynet.be> 141841Date: Thu Oct 9 22:22:53 2008 +0200 141842 141843 DGA: Track client state even when using old style DGA. 141844 141845 This fixes the issue that a badly killed DGA will keep on hogging 141846 mode/framebuffer/mouse/keyboard. 141847 141848commit 322335d5b5b6f155f56fe3c1cbe372f13dc20932 141849Author: Luc Verhaegen <libv@skynet.be> 141850Date: Thu Oct 9 22:21:05 2008 +0200 141851 141852 DGA: Mash together xf86dga.c and xf86dga2.c. 141853 141854 This in preparation for an upcoming client state tracking fix. 141855 141856commit d13c3cbd43bc5e6b459c7df822292cf02ad2c9c4 141857Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141858Date: Wed Oct 8 18:39:41 2008 -0700 141859 141860 XQuartz: Some motion made towards supporting fullscreen. 141861 (cherry picked from commit 99be3d68b64059caada739a373e5e01844c776e0) 141862 141863commit dc166bf6423f4243ea05f17340ff9ff99ecad74b 141864Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141865Date: Tue Oct 7 11:39:08 2008 -0700 141866 141867 XQuartz: Fixed EXTRA_DIST for new GL files. 141868 (cherry picked from commit a11c9052030b44045f037050a5825256b539ed50) 141869 141870commit 803327929445256e159c10894e6305fbaa389982 141871Author: George Staplin <georgestaplin@george-staplins-mac-pro.local> 141872Date: Tue Oct 7 12:12:38 2008 -0600 141873 141874 XQuartz: GL: setVisualConfigs() copied the behavior of indirect.c which is wrong, with regard to the handling of xcalloc failure. 141875 141876 Use ErrorF for an error message after an xcalloc failure, and return instead of falling through to GlxSetVisualConfigs, or abort()ing at the test branch. 141877 (cherry picked from commit 1056700971fd5c034396ed6dbea15e092f0c6332) 141878 141879commit f1fec6f323b27a0165325d8fcf6af7ae163bde14 141880Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141881Date: Wed Oct 8 18:45:04 2008 -0700 141882 141883 XQuartz: Call setVisualConfigs in InitOutput, and only when GLXEXT is defined. 141884 (cherry picked from commit 97173d4eda142c73bb975cc05225b791778f85af) 141885 141886commit 52accec7dbc613a521e40449b8bb5f0c92ddb105 141887Author: George Staplin <georgestaplin@george-staplins-mac-pro.local> 141888Date: Tue Oct 7 11:31:48 2008 -0600 141889 141890 XQuartz: GL: Add the proper license information to the new files, and add more commentary to setVisualConfigs(). 141891 (cherry picked from commit 190a3d5e5de9915931928fb8e6da88bb9644cda4) 141892 141893commit 63abdf61151cc90f769ed1708ee19b6c434c27c9 141894Author: George Staplin <georgestaplin@george-staplins-mac-pro.local> 141895Date: Mon Oct 6 18:05:12 2008 -0600 141896 141897 XQuartz: GL: Add the new C code to the Makefile.am. 141898 141899 Commit the darwin.c changes I missed in the last commit, for calling 141900 setVisualConfigs(). 141901 (cherry picked from commit eb3c014e1710bf0b93bda10ddb9b795cd150d02d) 141902 141903commit a8f5d422c9c0a39f55e80bbd180439b6ec3a805c 141904Author: George Staplin <georgestaplin@george-staplins-mac-pro.local> 141905Date: Mon Oct 6 18:01:23 2008 -0600 141906 141907 XQuartz: GL: Provide code for getting the capabilities of the underlying system's CGL. 141908 141909 Add a setVisualConfigs that is called before the fbScreenInit, to setup the __GLXvisualConfigs. 141910 (cherry picked from commit fc86f9e4482043eca76d9d7a96e166be1aabf674) 141911 141912commit 2998e48be343ab2a11d6d328fc961ab5b8eb9292 141913Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 141914Date: Sat Oct 4 21:25:50 2008 -0700 141915 141916 Xquartz: Fix parallel builds 141917 (cherry picked from commit 1fe7298a3b9a2ed82c612292f9e547cf78f99ae5) 141918 141919commit 7aa535708f1f33173bf05abc17b75e0841ffc385 141920Author: Adam Jackson <ajax@redhat.com> 141921Date: Wed Oct 8 16:25:59 2008 -0400 141922 141923 Fix more xcalloc stupidity. 141924 141925 Did I mention the hate? Pretty sure I mentioned the hate. 141926 141927commit 4fe80aa14a24df6d0611c47fc5654c3c4f56fdb1 141928Author: Adam Jackson <ajax@redhat.com> 141929Date: Wed Oct 8 16:22:35 2008 -0400 141930 141931 I hate you xlib. Please die in a fire. 141932 141933commit 2c5332872f808e7a646a63ae7ec63a2b80f83ee7 141934Author: Adam Jackson <ajax@redhat.com> 141935Date: Wed Oct 8 13:39:49 2008 -0400 141936 141937 s/kdDontZap/kdAllowZap/ 141938 141939commit 1feb69eb63e6739ff5db255ad529e84adf941a10 141940Author: Luc Verhaegen <libv@skynet.be> 141941Date: Wed Oct 8 14:55:29 2008 +0200 141942 141943 DGA: Fix ProcXF86DGASetViewPort for missing support in driver. 141944 141945 Fixes a segfault when trying to activate a DGA mode without checking 141946 whether DGA modesetting is at all possible. 141947 141948commit d1bb5e3b47ab71aa2f18f69daaf2a4c5ce06d800 141949Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 141950Date: Wed Oct 8 03:19:45 2008 -0300 141951 141952 mi: functions here are carrying 'time' without need. 141953 141954commit 4964a9ca6c3f26376469b6d8217815a47231f0a9 141955Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 141956Date: Wed Oct 8 03:06:32 2008 -0300 141957 141958 mi: remove deprecated and unused functions. 141959 141960 miPointerDeltaCursor, miPointerGetMotionBufferSize and 141961 miPointerSetNewScreen. 141962 141963commit adf3e296e2b00164fb22e89ad4f44acbd2e3deae 141964Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 141965Date: Wed Oct 8 02:12:49 2008 -0300 141966 141967 Remove unused declarations. 141968 141969 AvailableClientInput (LBX debris) and xf86XinputFinalizeInit. 141970 141971commit bca9fa2efe97d2816a2e9412cdff08a0145c8f96 141972Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 141973Date: Wed Oct 8 01:38:27 2008 -0300 141974 141975 Everyone knows that the loader is present. Silence it. 141976 141977commit 56c615368c5a8e7acb0398434c2c68578626aa38 141978Author: Zhenyu Wang <zhenyu.z.wang@intel.com> 141979Date: Wed Oct 8 13:33:55 2008 +0800 141980 141981 Check nextEnabledOutput()'s return in bestModeForAspect() 141982 141983 In case no enabled outputs, we will reference wrong index of 141984 output array. 141985 141986commit 278c11f01fbc6d6bd91c5a7127928c9ef5d29fca 141987Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 141988Date: Wed Oct 8 01:18:49 2008 -0300 141989 141990 xfree86: xf86{Enable, Disable}InputHandler can be static. 141991 141992commit 50081d2dfb79878cb931a15c265f0d60698dfd39 141993Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 141994Date: Wed Oct 8 01:11:38 2008 -0300 141995 141996 xfree86: remove xf86{Enable, Disable}GeneralHandler and others unused. 141997 141998commit 54294e15ecf4a8ec0fd9fa8b6ed1a1b2d912ec7e 141999Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 142000Date: Wed Oct 8 00:43:22 2008 -0300 142001 142002 xfree86: fix wrong function name message errors. 142003 142004commit 57dbb90f198116dcaf17c8111a7f80644d7ebe7f 142005Author: Adam Jackson <ajax@redhat.com> 142006Date: Tue Oct 7 20:36:45 2008 -0400 142007 142008 Document zap/-retro interaction more. 142009 142010commit 617933e8c259ad12388c0dd13421846a873975e3 142011Author: Daniel Stone <daniel@fooishbar.org> 142012Date: Wed Oct 8 02:45:59 2008 +0300 142013 142014 XFree86: Fix DontZap interaction with -retro 142015 142016 -retro will now _not_ override (No)DontZap, if it's set in the config 142017 file. 142018 142019commit d80bae2237e555025465d4d761a5cc537cc2bcdd 142020Author: Daniel Stone <daniel@fooishbar.org> 142021Date: Wed Oct 8 02:43:27 2008 +0300 142022 142023 Update docs to reflect new DontZap default 142024 142025commit 42e7b64bf90604e0099702f5447163be80a2efb6 142026Author: Adam Jackson <ajax@redhat.com> 142027Date: Tue Oct 7 20:32:02 2008 -0400 142028 142029 -retro enables zapping 142030 142031commit 9d135ac10a7374c7ccda705f1eeb02cc53076c34 142032Author: Daniel Stone <daniel@fooishbar.org> 142033Date: Wed Oct 8 02:30:05 2008 +0300 142034 142035 Input: Make DontZap the default 142036 142037 If you need to bail out the server, use Ctrl-Alt-Fx, or enable zapping 142038 if it bothers you that much. If Ctrl-Alt-Fx is broken, nag me until 142039 it's permanently fixed. 142040 142041commit b574365fca954a999ebaeb7203d1e0ea6d99b3ea 142042Author: Aaron Plattner <aplattner@nvidia.com> 142043Date: Tue Oct 7 13:09:11 2008 -0700 142044 142045 Revert "Unexport (and unimplement) xf86SetAccessFuncs" for now. 142046 142047 The nvidia driver currently uses these hooks to work around problems where RAC 142048 will disable access to the hardware at unexpected times. This change restores 142049 these hooks until we can come up with a better API for working around RAC. 142050 142051 This reverts commit c1df4fbede8058c15ce3a5759a7758fecafbb9e7. 142052 142053commit fe85ec34ec2b58188633dac4464a1e9e9977652b 142054Author: Aaron Plattner <aplattner@nvidia.com> 142055Date: Tue Oct 7 13:08:38 2008 -0700 142056 142057 Revert "Remove unused server state change callbackery" for now. 142058 142059 The nvidia driver currently uses these callbacks to work around problems where 142060 RAC will disable access to the hardware at unexpected times. This change 142061 restores these hooks until we can come up with a better API for working around 142062 RAC. 142063 142064 This reverts commit d7c0ba2e9eae7044ef4c31d9c3cbb9a71ee0f9f9. 142065 142066 Conflicts: 142067 142068 hw/xfree86/loader/xf86sym.c 142069 142070commit 94919480d8bb66e1807b4fe87b8f326ef6e012c6 142071Author: Adam Jackson <ajax@redhat.com> 142072Date: Tue Oct 7 13:41:25 2008 -0400 142073 142074 int10: Fix a nasty memory leak. 142075 142076commit a57b2f172c1291f22f7ba2780c1b2f55e353c3e9 142077Author: Adam Jackson <ajax@redhat.com> 142078Date: Tue Oct 7 13:39:10 2008 -0400 142079 142080 int10: Don't warn when scanning for devices we don't have. 142081 142082 Some BIOSes (hi XGI!) will attempt to enumerate the PCI bus by asking 142083 for the config space of every possible device number. This despite 142084 perfectly functional BIOS methods to enumerate the bus exactly. 142085 142086commit a65e36a873cd1ba9896cd0f9a3e94dd933666005 142087Author: Adam Jackson <ajax@redhat.com> 142088Date: Tue Oct 7 13:38:12 2008 -0400 142089 142090 int10: Remove useless check. 142091 142092 If you have more than one PCI device with the same d/b/d/f, you're 142093 already in trouble. 142094 142095commit 6dbfca571f6d214dc456710a60ebbbbd204cdbbd 142096Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142097Date: Tue Oct 7 10:07:03 2008 -0700 142098 142099 Depend on xtrans >= 1.2.2 for TRANS_NOXAUTH 142100 142101commit 30193fbff9d9284fbbd81fe5f7b6c4d102daf1d1 142102Author: Adam Jackson <ajax@redhat.com> 142103Date: Tue Oct 7 10:36:12 2008 -0400 142104 142105 Reinstate cursorScreenDevPriv 142106 142107commit e7cd5d84ed1fb3637dce6e9e6ef58c8611818c4e 142108Author: Dave Airlie <airlied@redhat.com> 142109Date: Tue Oct 7 16:09:17 2008 +1000 142110 142111 fixup another calloc build issue 142112 142113commit 60cfaa45a532bd3cfcfb4bd57b2f657c0a8b6a6b 142114Author: Adam Jackson <ajax@redhat.com> 142115Date: Mon Oct 6 18:47:06 2008 -0400 142116 142117 Build fix. 142118 142119commit c6d6d3e87a472b641d883a6c000f96e5bdf48532 142120Author: Adam Jackson <ajax@redhat.com> 142121Date: Mon Oct 6 16:18:21 2008 -0400 142122 142123 Loader: Unexport a handful of consumerless symbols from dixsym. 142124 142125commit 8a5b89e8e184f4cbf33c6dee0b06e61d8f857576 142126Author: Adam Jackson <ajax@redhat.com> 142127Date: Mon Oct 6 15:36:51 2008 -0400 142128 142129 xalloc+memset(0) -> xcalloc 142130 142131commit 0b7b89fbac0b3865b2cf51295c68a5f4c7523f28 142132Author: Adam Jackson <ajax@redhat.com> 142133Date: Mon Oct 6 14:19:30 2008 -0400 142134 142135 xalloc+bzero -> xcalloc 142136 142137commit 9187f6ad9ec7ba9569a93d92561aac17eaa83491 142138Author: Keith Packard <keithp@keithp.com> 142139Date: Mon Oct 6 12:21:20 2008 -0400 142140 142141 RANDR: Delivery output property events. 142142 142143commit 717c7492a0f6ba3fb3eabda33515881eef314155 142144Author: Dave Airlie <airlied@linux.ie> 142145Date: Tue Oct 7 02:38:44 2008 +1000 142146 142147 exa: don't call composite routines with no buffer. 142148 142149 We can get a case with gnome-terminal + links, where we get two arrays 142150 of glyphs all with 0 width and 0 heights in them. If this happens 142151 we manage to get to this case without any buffer setup and segfault. 142152 142153commit aacac3edac7c6b82a0a0e3bc500f563d187e5df3 142154Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142155Date: Sat Oct 4 21:07:31 2008 -0700 142156 142157 XQuartz: Bumped bundle version to 2.3.2 for beta1 142158 (cherry picked from commit 2021c4c4e726a6b8f27f19f16ed3d86e37dbe448) 142159 142160commit ed3134918dcbbd58a5ef8359a77e8024c98396e6 142161Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142162Date: Sat Oct 4 21:06:58 2008 -0700 142163 142164 XQuartz: pbcopy: Re-enabled some code I commented out for testing 142165 (cherry picked from commit 9b5d11409264080e394c8f18e06237122de80014) 142166 142167commit b3a971d8b5df5ab4af0be49105ef32f67d010901 142168Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142169Date: Sat Oct 4 18:54:15 2008 -0700 142170 142171 XQuartz: xpbproxy: Support some of the preference toggles from X11.app, cleaned up CLIPBOARD_MANAGER atom management. 142172 (cherry picked from commit f7673bb4de3c1f71eb390a3279eed3589efc3df4) 142173 142174commit fc35c7a2e2888cb767413f5e85a79c7f422a6284 142175Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142176Date: Sat Oct 4 21:16:54 2008 -0700 142177 142178 XQuartz: GLX - enable_stereo defaults option should work now 142179 (cherry picked from commit 3a6898f48756fdb2898292679301ef3dc109b4f6) 142180 142181commit 7818aba1d1e74f4ffe68b7d6806622b94ae8a230 142182Author: Roland Scheidegger <sroland@tungstengraphics.com> 142183Date: Sat Oct 4 15:50:35 2008 +0200 142184 142185 dri2: fix compilation (remove no longer existing include file) 142186 142187commit a29af6465883796a6acf04b72ce812a0632b2186 142188Author: Maarten Maathuis <madman2003@gmail.com> 142189Date: Fri Oct 3 23:54:10 2008 +0200 142190 142191 exa: remove "direct" case for exa{Trapezoids,Triangles} 142192 142193 - By adding a small hack to the xserver i was able to easily test the performance of the normally rare direct case (using cairo). 142194 - It turned out to be 70% slower for me (large test on an otherwise idle computer), which seems enough of a reason to remove it. 142195 - AddTraps could also use a 2nd look, but since noone is using that it's a bit hard and less useful to test. 142196 142197commit 416870d8c9e7e383b87c8513c2296a0cb66e7ef4 142198Author: Maarten Maathuis <madman2003@gmail.com> 142199Date: Fri Oct 3 22:59:40 2008 +0200 142200 142201 exa: remove some excessive whitespace 142202 142203commit b72a217ad59cc527327d7b1fc81898677b977d8f 142204Author: Adam Jackson <ajax@redhat.com> 142205Date: Fri Oct 3 18:30:37 2008 -0400 142206 142207 Dangling use of miDestroyGCOps 142208 142209commit f5bbc54a93f53cd4ba124712159bb44c93282a8e 142210Author: Adam Jackson <ajax@redhat.com> 142211Date: Fri Oct 3 18:18:02 2008 -0400 142212 142213 Fix compile for miSetScreenPixmap prototype. 142214 142215commit 9dbfcd89214241626ac2704d1ffffff1cc0c67ae 142216Author: Adam Jackson <ajax@redhat.com> 142217Date: Fri Oct 3 17:40:30 2008 -0400 142218 142219 Delete unused miModifyBanking() 142220 142221commit 94825ad3c72a1c3f6a61199c302841f63241cf87 142222Author: Adam Jackson <ajax@redhat.com> 142223Date: Fri Oct 3 17:39:34 2008 -0400 142224 142225 Remove unused wrapping for visual initialization. 142226 142227commit 4791f8e2ba9a0e318b7f13c83618ece036dd53ed 142228Author: Adam Jackson <ajax@redhat.com> 142229Date: Fri Oct 3 17:23:58 2008 -0400 142230 142231 staticize mi{G,S}etScreenPixmap 142232 142233commit 14d98b4a2fdea4a4f5cc9b751151672cb392fba3 142234Author: Adam Jackson <ajax@redhat.com> 142235Date: Fri Oct 3 17:21:16 2008 -0400 142236 142237 miFindMaxBand is dead code 142238 142239commit 48b47928133bc462947104bd0f4204f00b993862 142240Author: Adam Jackson <ajax@redhat.com> 142241Date: Fri Oct 3 17:16:28 2008 -0400 142242 142243 mi{Create,Destroy}GCOps are unused, nuke 'em. 142244 142245commit bd0e87c74789bd5447ec731aa2d2d6e05dd54abc 142246Author: Adam Jackson <ajax@redhat.com> 142247Date: Fri Oct 3 17:13:10 2008 -0400 142248 142249 Remove a bunch of unused API from the mi export list. 142250 142251commit ee5e07bc22f0c3af7a412b6bc430eae6a1dafe16 142252Author: Adam Jackson <ajax@redhat.com> 142253Date: Fri Oct 3 16:58:01 2008 -0400 142254 142255 loader: sort the mi symbol list 142256 142257commit 77d0b0ecfa3c67ff8d11e773a7f083721cf4c6d7 142258Author: Adam Jackson <ajax@redhat.com> 142259Date: Fri Oct 3 16:53:08 2008 -0400 142260 142261 Remove unused pciNumBuses 142262 142263commit 0ce61e21d6d7dcca0090e319bbcdb678570f2c3f 142264Author: Adam Jackson <ajax@redhat.com> 142265Date: Fri Oct 3 16:05:19 2008 -0400 142266 142267 Remove the Must_have_memory hack. 142268 142269 Also remove an astonishing amount of misunderstanding of how casts work. 142270 142271commit e6b1c1fada19268af559d89375989973729fdb19 142272Author: Adam Jackson <ajax@redhat.com> 142273Date: Fri Oct 3 15:47:50 2008 -0400 142274 142275 const cleanup 142276 142277commit fe616f9230b672edb89fca7516e750dc1b20c690 142278Author: Adam Jackson <ajax@redhat.com> 142279Date: Fri Oct 3 15:42:58 2008 -0400 142280 142281 Drop the never-called SkippedRequestCallback list. 142282 142283commit 31136b9467c2a17bbed4a51d6a6c41e3290bd236 142284Author: Adam Jackson <ajax@redhat.com> 142285Date: Fri Oct 3 15:33:47 2008 -0400 142286 142287 Unexport MakeClientGrab{Imp,P}ervious 142288 142289 xtrap used to use these, but it's gone now. xtest still does but it's 142290 linked statically. 142291 142292commit eaf3fdf97014c27aececa8505c2713e7c9a537fc 142293Author: Adam Jackson <ajax@redhat.com> 142294Date: Fri Oct 3 15:24:35 2008 -0400 142295 142296 Unexport xf86NoSharedResources 142297 142298commit d7d9edb506bfea1c10516b6c4eb72dee3559863c 142299Author: Adam Jackson <ajax@redhat.com> 142300Date: Fri Oct 3 15:17:44 2008 -0400 142301 142302 Unexport xf86DeallocateResourcesForEntity 142303 142304commit 49673b43584928c22d699d1968ed6813b272aefe 142305Author: Adam Jackson <ajax@redhat.com> 142306Date: Fri Oct 3 15:13:39 2008 -0400 142307 142308 Unexport xf86SetEntityFuncs 142309 142310commit 949ef6b79a36d6f6bbc192c20ef0f13cc95d4f28 142311Author: Adam Jackson <ajax@redhat.com> 142312Date: Fri Oct 3 15:10:26 2008 -0400 142313 142314 Unexport xf86ChkConflict 142315 142316 Also remove the unused internal ChkConflict() 142317 142318commit d7c0ba2e9eae7044ef4c31d9c3cbb9a71ee0f9f9 142319Author: Adam Jackson <ajax@redhat.com> 142320Date: Fri Oct 3 15:06:42 2008 -0400 142321 142322 Remove unused server state change callbackery. 142323 142324commit 9ada146a16bcf220b2dd2936fb3a0349a31e99d3 142325Author: Adam Jackson <ajax@redhat.com> 142326Date: Fri Oct 3 14:40:18 2008 -0400 142327 142328 xf86RegisterResources: Always print any failed registrations. 142329 142330commit a5d3fa90f3de763a5dcbb0ff4987562449819323 142331Author: Adam Jackson <ajax@redhat.com> 142332Date: Fri Oct 3 14:38:48 2008 -0400 142333 142334 Typo fix. 142335 142336commit cc3335858508ffbc52123eb5bb0c34f42800c8c4 142337Author: Adam Jackson <ajax@redhat.com> 142338Date: Fri Oct 3 14:37:59 2008 -0400 142339 142340 Remove the old implementation of RemoveOverlaps 142341 142342commit 8fbf1185f5d6a1836b3e4265de6f605d82478e0b 142343Author: Adam Jackson <ajax@redhat.com> 142344Date: Fri Oct 3 14:32:32 2008 -0400 142345 142346 Internalise some resource list API. 142347 142348commit 0ca4f3cee7e42a393cdd32d3a772a190ae68abc5 142349Author: Adam Jackson <ajax@redhat.com> 142350Date: Fri Oct 3 13:29:55 2008 -0400 142351 142352 Unexport xf86EnterServerState 142353 142354commit c1df4fbede8058c15ce3a5759a7758fecafbb9e7 142355Author: Adam Jackson <ajax@redhat.com> 142356Date: Fri Oct 3 13:08:43 2008 -0400 142357 142358 Unexport (and unimplement) xf86SetAccessFuncs. 142359 142360 Because, no. 142361 142362commit 44269e604eb60ebaa6189fd4409d39a4ec6b9f4b 142363Author: Adam Jackson <ajax@redhat.com> 142364Date: Fri Oct 3 13:02:50 2008 -0400 142365 142366 RAC: NO_SEPARATE_*_FROM_* are useless. 142367 142368 If anyone can come up with an example of a bus where: 142369 142370 - both i/o and memory resources are addressable 142371 - access to them can be controlled 142372 - but they can't be controlled independently 142373 142374 then by all means, reinstate this logic. 142375 142376commit 51fcb58f45824c7cde88b8b4ccbd6bdcf760280e 142377Author: Adam Jackson <ajax@redhat.com> 142378Date: Fri Oct 3 12:39:49 2008 -0400 142379 142380 RAC: AccessEnable() and AccessDisable() are mandatory. 142381 142382commit 9c5fc7762e9ca09847965e67e247d95dfc197266 142383Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142384Date: Fri Oct 3 11:24:34 2008 -0700 142385 142386 Partially reverts commit 505c216d030a8eb6e9fa757ecebed9597a850c88. 142387 142388 I don't know how the hell git-cherrypick messed that up to delete fbwindow.c 142389 142390commit 7bb73a9513710feaebc127998950e9f472bfcc0d 142391Author: George Peter Staplin <gps@Georges-Workstation.local> 142392Date: Tue Sep 30 23:53:12 2008 -0600 142393 142394 XQuartz: pbproxy: Possibly fix a memory leak by using an [NSApp run] loop, 142395 instead of calling CFRunLoopRun() directly. The leak wasn't reproducible on 142396 this machine, but someone was able to produce a leak trace with Instruments 142397 that indicates it was leaking in the CFRunLoopRun() path. 142398 142399 x-input.m: dequeue and ignore events when pbproxy_active is false. 142400 142401 x-selection.h: add an is_active method that is used by x-input.m to ignore 142402 events. 142403 142404 x-selection.m: Handle nearly every preference, except for primary_on_grab, 142405 which I don't really understand yet. 142406 (cherry picked from commit 4d51ad851e64da83cbdfb0a4a22428418a7bcf75) 142407 142408commit 2a5ce41f0371ad5df52586b8d4072578b6206321 142409Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142410Date: Tue Sep 30 13:54:08 2008 -0700 142411 142412 XQuartz: Use "pointer" and "pen" for device names to please GDK. 142413 (cherry picked from commit 00ca0f4d839abf47e9573a1552473e039cf787e6) 142414 142415commit d3f597a986dcbc7e4164f350300dbd2e438949d8 142416Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142417Date: Tue Sep 30 11:10:58 2008 -0700 142418 142419 XQuartz: Workaround for initial pressure/tilt being sent as 0 with motion during the proximity event 142420 (cherry picked from commit 9a91d770a6411dd876187e64a8bda1f0745169ae) 142421 142422commit 1657dfb843c0127ea31bfd0a8ef81d0c394cbc7b 142423Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142424Date: Tue Sep 30 08:46:08 2008 -0700 142425 142426 XQuartz: Using absolute ranges for pointer location to increase resolution and better support tablets. 142427 (cherry picked from commit d79ccb45f68b65d65718b5b77efe2fae6eeda762) 142428 142429commit 88bb8c6f485a7aaf40f684d484a33cfedfa74a70 142430Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142431Date: Mon Sep 29 22:33:02 2008 -0700 142432 142433 XQuartz: Removed resolved comment about localization. 142434 (cherry picked from commit 8ed5faf058ac7b0782a9cc13a2c58b80168358d2) 142435 142436commit b1d0b196617643619d24ed520b0a1161e26b66c6 142437Author: George Peter Staplin <gps@Georges-Workstation.local> 142438Date: Tue Sep 30 13:04:35 2008 -0600 142439 142440 XQuartz: pbproxy: Make the signal handler safer by using _exit, and only exit. 142441 142442 Remove some unnecesssary headers. 142443 142444 Remove some dead code that was never called or used in pbproxy. 142445 142446 Make use of an NSAutoreleasePool in x_init. It could potentially cause a leak 142447 on a startup without this. 142448 142449 Start adding reload_preferences to the x_selection class, as well as event 142450 handling for that. 142451 (cherry picked from commit 602e8ba8f7ee196696bc9e3cea6ecdf3200dcf5c) 142452 142453commit ecd260143cf9f012f033ba029ad04cb49567c89a 142454Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142455Date: Mon Sep 29 22:32:24 2008 -0700 142456 142457 XQuartz: Added some localizable strings that were missing. 142458 (cherry picked from commit 2c8205ea8b709c5859412b466e83aec7a94acdcf) 142459 142460commit aae878b89081ffba16386e8d4987469313049bea 142461Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142462Date: Mon Sep 29 22:26:20 2008 -0700 142463 142464 XQuartz: Actually send the pointer event from the pen even if it happes to actually have 0 pressure and 0 tilt. 142465 (cherry picked from commit f41583761955cb9c92c43239bfaa8ae0e5d95e33) 142466 142467commit b7ad86ff8a9115147c0cf02af9e80ec464129392 142468Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142469Date: Sat Sep 27 15:30:10 2008 -0700 142470 142471 Xquartz: xpbproxy: Split out app-specific stuff to app-main.m in prep for making this into a thread. 142472 (cherry picked from commit c2012138a520560f8a2160518ea73fced410c3b7) 142473 142474commit 5446adebfb085dd05471729bf5c55bf820b1cbe3 142475Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142476Date: Fri Oct 3 11:12:49 2008 -0700 142477 142478 XQuartz: Send AppleWMReloadPreferences where appropriate. 142479 (cherry picked from commit 490cbe9888e2c1080495b003c429fdb1659444e1) 142480 142481commit 6020c3fbfe06530cb2b90178d17a7d63beae028a 142482Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142483Date: Fri Oct 3 11:09:26 2008 -0700 142484 142485 XQuartz: Added a --enable-integrated-xpbproxy configure option for building xpbproxy as an app or as a thread. 142486 (cherry picked from commit 8edc5fb38c922f28659d2f823148339a8907c4d9) 142487 (cherry picked from commit 88033a66a5549870fd053795b019d4c22950425b) 142488 142489commit 505c216d030a8eb6e9fa757ecebed9597a850c88 142490Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142491Date: Fri Oct 3 11:11:37 2008 -0700 142492 142493 Added dep on applewmproto 1.1.0 142494 (cherry picked from commit b350b066740a2e72585153e35478538cf1bca9ef) 142495 142496commit 1c1a6e965da412e767ad8bc15a5362116d64c6b0 142497Author: Adam Jackson <ajax@redhat.com> 142498Date: Mon Sep 22 22:09:19 2008 -0400 142499 142500 Remove an utterly, utterly ancient changelog. 142501 142502 Wow. 142503 142504commit 9c8a2be2c73abf06245b1eb7f05f93e104dcfe42 142505Author: Peter Hutterer <peter.hutterer@redhat.com> 142506Date: Thu Oct 2 08:55:14 2008 +0930 142507 142508 dix: fix axis scaling. 142509 142510 For two axes [a, b] and [x, y] (inclusive), the formula to scale point P(ab) 142511 to (x,y) is: 142512 (P - a)/(b - a) * (y - x) + x 142513 142514 And the whole end result rounded of course to get the integer we need. 142515 142516commit 93ef72fa26b27cf90bf5c64ec19ac295a113aaae 142517Author: Peter Hutterer <peter.hutterer@redhat.com> 142518Date: Sun Sep 28 21:47:21 2008 +0930 142519 142520 dix: don't lose subpixel precision when scaling. 142521 142522 Only scale x/y back from screen range if we have crossed screen, otherwise 142523 leave it as it is. 142524 142525commit 5ef5e89f0e4b9b8820093e072430e295efce32fa 142526Author: Simon Thum <simon.thum@gmx.de> 142527Date: Sun Sep 28 22:32:44 2008 +0930 142528 142529 dix: fix non-initialization of pointer accel in master devices. 142530 142531 Init MDs always with PtrAccelNoOp. 142532 142533 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 142534 142535commit 036d424827726b0c804ae1532109253270744207 142536Author: Peter Hutterer <peter.hutterer@redhat.com> 142537Date: Fri Sep 26 17:01:12 2008 +0930 142538 142539 xfree86: don't let DGA steal key events outside of [9,255]. 142540 142541commit 56a24cf894a142b18c7f18e294074c32031f3d14 142542Author: Adam Jackson <ajax@redhat.com> 142543Date: Thu Oct 2 17:56:31 2008 -0400 142544 142545 xorg.conf.man: We don't look in root's ~ anymore, don't say we do. 142546 142547commit 7921c4a0bfc843c772dcb2fc185d4bcab1316330 142548Author: Adam Jackson <ajax@redhat.com> 142549Date: Thu Oct 2 17:08:48 2008 -0400 142550 142551 Remove some stale IPv6 debugging 142552 142553commit f836e1f11b7bf3449c3e469cc7799c4ed28d2356 142554Author: Adam Jackson <ajax@redhat.com> 142555Date: Thu Oct 2 17:03:54 2008 -0400 142556 142557 Unifdef ISC 142558 142559commit b1a31734731a768d586b4cb18317d4a5c5e20fa8 142560Author: Adam Jackson <ajax@redhat.com> 142561Date: Thu Oct 2 17:03:21 2008 -0400 142562 142563 Fix bad preprocessor line 142564 142565commit 00f01f39454001846869496d9124f62c01e5691a 142566Author: Adam Jackson <ajax@redhat.com> 142567Date: Thu Oct 2 15:48:18 2008 -0400 142568 142569 Don't inherit XAUTHORITY from the environment. 142570 142571 It's not especially obvious, and unpleasantly overloaded for the Xnest 142572 case. Typically this gives you a server that looks for its auth data in 142573 the authority file you were using for the running X session, which 142574 generally doesn't have an entry for the display you just started. 142575 142576 All the major dm's, and startx, pass -auth explicitly, so this shouldn't 142577 cause too much upheaval. 142578 142579commit eb8be3e90a9c90a428696026d1e3b2152d7eefb4 142580Author: George Peter Staplin <gps@Georges-Workstation.local> 142581Date: Thu Sep 25 14:16:20 2008 -0600 142582 142583 XQuartz: pbproxy: Fix another STRING bug. 142584 142585 Fix the usage of the NSString cStringUsingEncoding: - it doesn't NUL 142586 terminate the string, which lead to a bus error. So, we use 142587 lengthOfBytesUsingEncoding: to get the length in bytes instead of 142588 strlen(). 142589 (cherry picked from commit 6333d619e747c3b6bd3ba7557e35c0e5f6daa40f) 142590 142591commit edfa3f8233fde490cb4255f21473684e96d595a3 142592Author: George Peter Staplin <gps@Georges-Workstation.local> 142593Date: Thu Sep 25 12:29:57 2008 -0600 142594 142595 XQuartz: pbproxy: Fix a bug that rxvt brought out in STRING copying. Now I can copy and paste to/from rxvt correctly. 142596 (cherry picked from commit 82a4dc5f0f31a7911ee1cce5d1b162077befa811) 142597 142598commit 12a59c44cb68843a60fc43257930d1cbeb971b7a 142599Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142600Date: Wed Sep 24 22:51:03 2008 -0700 142601 142602 XQuartz: Fixed threading issue with TSM. 142603 (cherry picked from commit 93ab4e0071670bb80bfa1170dd97ed9d6d51c67a) 142604 142605commit ee87c9b02401a7a08b396884ba412a503b078bbd 142606Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142607Date: Wed Sep 24 18:57:36 2008 -0700 142608 142609 XQuartz: pbproxy: Removed stray SPI include. 142610 (cherry picked from commit 2fb2bc77b804d7b63002c1423cb0bfc658eeee45) 142611 142612commit 591cc57627a9137159fb5a13287c1d61d8d2fc8c 142613Author: George Peter Staplin <gps@Georges-Workstation.local> 142614Date: Tue Sep 23 13:59:25 2008 -0600 142615 142616 XQuartz: pbproxy: Fix 3 paths that could leak memory. 142617 142618 2 of the paths leaked, when INCR transfers were done. Now we 142619 are leak free according to the leaks program for all transfers 142620 I have tried so far. 142621 (cherry picked from commit aa98db576bc02f0765cb35d0282a894ba3778213) 142622 142623commit 586fbc91df731150ff5e39170efe33f3cfac28a2 142624Author: George Peter Staplin <gps@Georges-Workstation.local> 142625Date: Tue Sep 23 12:39:32 2008 -0600 142626 142627 XQuartz: pbproxy: Fix NSObject memory leaks by properly using the 142628 NSAutoreleasePool. Now the usage is consistent. In x_input_run() 142629 we create a pool, and release it after processing the XEvents. 142630 142631 Add some getpid() output to main for debugging. It needs a bit more 142632 testing before the next release. 142633 142634 Don't retain the NSPasteboard as the old code did. That may have 142635 contributed to the leak, and it made it so that we needed the 142636 NSAutoreleasePool created in main(). 142637 142638 Remove the _known_types, and _pasteboard instance variables from 142639 the x_selection class. They aren't needed anymore. 142640 142641 The leaks program now indicates 0 leaks after some usage. I want 142642 to test further, but this seems much better, and my memory usage 142643 graph indicates it's not growing. 142644 (cherry picked from commit b245d84a72ee3929546cd11a6eba3c60fb4a4d95) 142645 142646commit 238999cfc99dbf306184f0e846465a8707bcc9d2 142647Author: George Peter Staplin <gps@Georges-Workstation.local> 142648Date: Mon Sep 22 17:36:21 2008 -0600 142649 142650 XQuartz: pbproxy: Add a comment to pbproxy.h about how the DB() 142651 macro causes a leak (according to the leaks program). 142652 142653 Attempt to fix several other leaks with release method calls. 142654 For some reason the process still grows more than it should... 142655 142656 I will need to use some better methods than leaks, and malloc_history 142657 I suspect. Whatever is leaking, it's hard to find. I need to isolate 142658 the cases more. 142659 142660 Add a missing image/jpeg branch. 142661 142662 Remove read_prop_32 - it's not used. 142663 (cherry picked from commit 63a680354dcb545fef935ac97596dd35ceaed960) 142664 142665commit 106eb37e0214bbd864a59119b491abfc54f32629 142666Author: Peter Hutterer <peter.hutterer@redhat.com> 142667Date: Fri Sep 26 13:05:10 2008 +0930 142668 142669 Require inputproto 1.9.99.5 142670 142671 Pull and rebuild your input drivers, while you're at it. 142672 142673commit fbd09486c6e6c64db28d81caa5ef04821e49f2eb 142674Author: Peter Hutterer <peter.hutterer@redhat.com> 142675Date: Tue Sep 23 16:55:04 2008 +0930 142676 142677 Push server-known properties into xserver-properties.h. 142678 142679commit a2d83b9dc8387ec7e70689db1371cf6500b2f68e 142680Author: Peter Hutterer <peter.hutterer@redhat.com> 142681Date: Thu Sep 18 18:21:03 2008 +0930 142682 142683 Xi: add "deletable" flag to properties, add DeleteProperty handler. 142684 142685 A property can only be deleted if any of the following is true: 142686 - if a property is deletable and all handlers return Success. 142687 - if a property is non-deleteable and the all handlers return Success AND the 142688 delete request does not come from a client (i.e. driver or the server). 142689 142690 A client can never delete a non-deletable property. 142691 142692commit 22e90472681ebc7bdd28f82bbf2753a5b259e931 142693Author: Peter Hutterer <peter.hutterer@redhat.com> 142694Date: Thu Sep 18 16:12:02 2008 +0930 142695 142696 Xi: allow Set/GetProperties to return a status, and honour this status code. 142697 142698 If a property handler now bails out, return the error code to the caller. This 142699 allows to be slightly more specific with the errors. 142700 142701commit 1e24e7b9df3d02350c7ea18e9379e87fe4d00026 142702Author: Peter Hutterer <peter.hutterer@redhat.com> 142703Date: Thu Sep 18 15:58:46 2008 +0930 142704 142705 Xi: remove configure/query device property calls. 142706 142707 This removes all the meta-information about device properties (pending, 142708 fromClient, range, valid_values, immutable). 142709 142710commit feaa5fa6712c8c6f4ca97766e2ac0338253cf3b8 142711Author: Peter Hutterer <peter.hutterer@redhat.com> 142712Date: Thu Sep 25 11:03:22 2008 +0930 142713 142714 xfixes: realize the invisible cursor before displaying it. 142715 142716 AllocARGBCursor realizes the cursor but can only do so if we have devices 142717 there already. If we don't - then the cursor needs to be realized elsewhere. 142718 This is usually done in InitializeSprite, but since xfixes just hard-swaps the 142719 (realized) cursor to the InvisibleCursor, we need to manually realize it 142720 before trying to display it. 142721 142722commit e58be0f3425fb3e946a222077672c3c01308f887 142723Author: Peter Hutterer <peter.hutterer@redhat.com> 142724Date: Tue Sep 23 10:01:30 2008 +0930 142725 142726 config: print error code if NIDR fails. 142727 142728commit 30c3c13f1030268aaa6a3598d538fafd0592d77a 142729Author: Peter Hutterer <peter.hutterer@redhat.com> 142730Date: Mon Sep 22 11:10:46 2008 +0930 142731 142732 xkb: squash canonical types into explicit ones on core reconstruction. 142733 142734 If we update key types from core, and groups 2 - n have a canonical type but 142735 the same symbols as the explicit type of group 1, assume that it was a core 142736 sym duplication according to Section 12.4 of the XKB Protocol Spec. 142737 Ignore the canonical types and pretend there's only one group for the key - 142738 with the explicit key type. 142739 142740 The protocol spec does not cover this case, so we have to guess here. 142741 142742commit ae986d1c73d2f720bd0309d8c33328d14e8eed25 142743Author: Peter Hutterer <peter.hutterer@redhat.com> 142744Date: Fri Sep 19 18:27:24 2008 +0930 142745 142746 xkb: fix core keyboard map generation. #14373 142747 142748 According to Section 12.4 of the XKB Protocol Spec, if a key only has a single 142749 group but the keyboard has multiple groups defined, the core description of 142750 the key is a duplication of the single group across all symbols. i.e. 142751 G1L1 G1L2 G1L1 G1L2 G1L3 G1L4 G1L3 G1L4 142752 142753 The previous code generated G1L1 G1L2 G1L3 G1L4 G1L3 G1L4, leading to 142754 "invented" groups when the process is reversed. 142755 142756 Note that this creates wrong key types on reconstruction from core to xkb, 142757 i.e. any single-group key with a key type that is not one of the canonical 142758 four (Sec 12.2.3), will get the assigned type on group 1, and a canonical type 142759 for the other gruops. 142760 142761 X.Org Bug 14373 <http://bugs.freedesktop.org/show_bug.cgi?id=14373> 142762 142763commit 5bf3f0fd4e00f96cfebd14b58580c3c8733626fb 142764Author: Dave Airlie <airlied@linux.ie> 142765Date: Thu Sep 25 15:40:19 2008 +1000 142766 142767 exa: make sure pixmap devPrivate.ptr is NULL at create time 142768 142769commit 82f150d73cc9a7d8eaf7241a51b03af05bcec159 142770Author: Ian Romanick <ian.d.romanick@intel.com> 142771Date: Wed Sep 24 16:56:45 2008 -0700 142772 142773 GLX: Make sure GLX_SGI_make_current_read is enabled when possible 142774 142775commit 8d9defe8c2a685709318c1d43379443df3d2322a 142776Author: Adam Jackson <ajax@redhat.com> 142777Date: Wed Sep 24 10:11:00 2008 -0400 142778 142779 Fix un-suppressing the initial cursor. 142780 142781 The intended behaviour was "show as soon as someone calls 142782 XDefineCursor". What you actually got was, uh, slightly less well 142783 defined, since the screen's ChangeWindowAttributes hook would run after 142784 DIX handled the cursor change. Oops. 142785 142786 The trivial way to turn the cursor on is: 142787 % xsetroot -cursor_name gumby 142788 142789 Refer to /usr/include/X11/cursorfont.h for cursor names. 142790 142791 Thanks to anholt for catching this. 142792 142793commit a39377cbcbd3091095efbeab25bec18ae520147e 142794Author: Keith Packard <keithp@keithp.com> 142795Date: Tue Sep 23 09:22:07 2008 -0700 142796 142797 Revert "Render: Use built-in SHA1 library" 142798 142799 This reverts commit d3bd31fddff7894f89ba80a3cdddff49aff08db8. 142800 142801 X.org should not be providing a custom SHA1 implementation. 142802 142803commit 10a9bac0257b381367cedff395b40425d584bf59 142804Author: Keith Packard <keithp@keithp.com> 142805Date: Tue Sep 23 09:21:37 2008 -0700 142806 142807 Revert "Update COPYING for SHA1 code" 142808 142809 This reverts commit 64387c8573ca8b3909667e32d7a9fba9567b21d3. 142810 142811 X.org should not be providing a custom SHA1 implementation. 142812 142813commit 64387c8573ca8b3909667e32d7a9fba9567b21d3 142814Author: Daniel Stone <daniel@fooishbar.org> 142815Date: Tue Sep 23 19:05:14 2008 +0300 142816 142817 Update COPYING for SHA1 code 142818 142819commit d3bd31fddff7894f89ba80a3cdddff49aff08db8 142820Author: John Tapsell <johnflux@gmail.com> 142821Date: Tue Sep 23 18:56:32 2008 +0300 142822 142823 Render: Use built-in SHA1 library 142824 142825 Getting an external library for SHA1 is a mess, so just use our own, 142826 regrettably. Public domain implementation. 142827 142828commit 991ed2949ba682684952fe323f8f0f51ebdca98e 142829Author: Adam Jackson <ajax@redhat.com> 142830Date: Tue Sep 23 10:22:22 2008 -0400 142831 142832 Rename SaveSetMap() to SaveSetShouldMap(). 142833 142834 Avoids preprocessor collision with xfixeswire.h 142835 142836commit ca9fae00795a114bca4397c32b543d6326a4c547 142837Author: Owen Taylor <otaylor@redhat.com> 142838Date: Mon Sep 22 12:42:41 2008 -0700 142839 142840 Change 'remap' to 'map' in saveset functions/macros 142841 142842 Now that the code has been fixed so that Unmap means unmap and not "don't 142843 remap", 'remap' was confusing to have in the function names/parameters, so 142844 change it to simple 'map'. 142845 142846 Signed-off-by: Keith Packard <keithp@keithp.com> 142847 142848commit ad14239a358cf65e5702ec7d054aa1db4f1cdd68 142849Author: Adam Jackson <ajax@redhat.com> 142850Date: Mon Sep 22 11:26:49 2008 -0400 142851 142852 Upgrade GLX Public License 1.0 to FreeB 2.0. 142853 142854 According to the press release: 142855 142856 Previous SGI contributions to the free and open source community 142857 are now available under the new license. These contributions 142858 include the SGI® OpenGL® Sample Implementation, the GLX™ API and 142859 other GLX extensions. 142860 142861 [...] 142862 142863 "SGI has been one of the most ardent commercial supporters of free 142864 and open source software, so it was important to us that we continue 142865 to support the free software development community by releasing our 142866 earlier OpenGL-related contributions under this new license," said 142867 Steve Neuner, director of Linux, SGI. "This license ensures that all 142868 existing user communities will benefit, and their work can proceed 142869 unimpeded. Both Mesa and the X.org Project can continue to utilize 142870 this code in free software distributions of GNU/Linux. Now more than 142871 ever, software previously released by SGI under earlier GLX and SGI 142872 Free Software License B is free." 142873 142874 "The GLX API" is here read to include the original GLX source release 142875 from: 142876 142877 http://www.sgi.com/products/software/opensource/glx/download.html 142878 ftp://ftp.sgi.com/sgi/opengl/glx/glx1_2.1.tgz 142879 142880 which includes glxext.c as included in XFree86, from which our copies 142881 in glx/ and hw/dmx/glxProxy/ are derived. 142882 142883commit 722c51960dd4948b8a64f8eff115fa69da597ef9 142884Author: Owen W. Taylor <otaylor@redhat.com> 142885Date: Mon Sep 22 11:04:26 2008 -0400 142886 142887 Don't remap windows for SaveSetUnmap 142888 142889 Since ReparentWindow() does a unmap/map pair for windows that are already 142890 mapped, for saveset windows with SaveSetUnmap, we must unmap the window 142891 before calling ReparentWindow() to avoid the generation of MapRequest 142892 events, and so forth. 142893 142894commit 0bb317a78b96fddcdac319c9706b3a12f931ea44 142895Author: Adam Jackson <ajax@redhat.com> 142896Date: Mon Sep 22 10:51:51 2008 -0400 142897 142898 Default to -br. Add -retro option for the nostalgic. 142899 142900 -retro also reverts to the classic cursor display behavior, meaning, 142901 the cursor will be visible before anyone calls XDefineCursor(). 142902 142903commit ab17f88506a2dde32c50ff72b03743cd9c384ecc 142904Author: Julien Cristau <jcristau@debian.org> 142905Date: Mon Sep 22 01:59:31 2008 +0200 142906 142907 Don't use our version of ffs() if the system has one 142908 142909commit af0d5e2ef7b1b3f42e3522568c5f56850823c102 142910Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142911Date: Sat Sep 20 19:41:46 2008 -0700 142912 142913 XQuartz: 64bit crash fix ... don't pass pointers through xe. 142914 (cherry picked from commit 5f1265b82edead8f15f2affc79c2782519502fa2) 142915 142916commit 852a0b0dde34e7179394d96d09122b99bf7fe904 142917Author: George Peter Staplin <gps@Georges-Workstation.local> 142918Date: Sat Sep 20 17:49:16 2008 -0600 142919 142920 XQuartz: pbproxy: Add an X error handler that returns 0. 142921 142922 Move the struct atom_list into the x-selection class, so that it's 142923 no longer a global variable named atoms. This may ease pthread 142924 integration and reduce the chances of symbols conflicting. 142925 (cherry picked from commit c1403c713ca80104ae3736bd2ed1eb6ffa5192b6) 142926 142927commit 7fa6fc5ad0b12bc52a1c22906709fbb003782d11 142928Author: George Peter Staplin <gps@Georges-Workstation.local> 142929Date: Fri Sep 19 21:28:46 2008 -0600 142930 142931 XQuartz: pbproxy: Fix a bug that occured when a PICT format was available. 142932 We may need another branch to convert a PICT to a PNG or JPEG. For now 142933 TIFF works well in all of the test image copying apps when converted to 142934 PNG or JPEG with an NSBitmapImageRep class. 142935 (cherry picked from commit adf339d8f948fc1e308dbcae38fcfce504b5b0ab) 142936 142937commit f67490ceb5b9ddf25e734cc331705103599f3ed8 142938Author: George Peter Staplin <gps@Georges-Workstation.local> 142939Date: Fri Sep 19 14:29:37 2008 -0600 142940 142941 XQuartz: pbproxy: Make the proxy handle copy request races, and PRIMARY 142942 INCR transfers mixed with a CLIPBOARD change. 142943 142944 Fix a bug with some UTF-8 transfers. I can only guess that my UTF-8 142945 tests before were inadequate. I can now copy/paste any characters between 142946 uxterm and Textedit.app. 142947 (cherry picked from commit 7ca1532e8e6e2c733c9c508eee0de73d09ecb947) 142948 142949commit b666ffd78155db228174041e9d7460b8a07dde0f 142950Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142951Date: Fri Sep 19 13:15:07 2008 -0700 142952 142953 XQuartz: "Disable" the text when syncing is disabled. 142954 (cherry picked from commit 7713d7ab66e74b212a6c0d634a3ad82373e178db) 142955 142956commit dba290ec231422740f453b69b187f80df3f50e4b 142957Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 142958Date: Sun Sep 21 17:40:35 2008 -0700 142959 142960 XQuartz: Added a preferences UI for pbproxy options. 142961 (cherry picked from commit d76cca5e1eae86450aa236e49a2024881d0bac01) 142962 142963commit a928cae57b47625c9e3d136840d3e5d17fd3c51e 142964Author: George Peter Staplin <gps@Georges-Workstation.local> 142965Date: Fri Sep 19 00:18:44 2008 -0600 142966 142967 XQuartz: pbproxy: Add a missing include of Xutil.h. Another header 142968 seems to have done this before, but correctness matters. 142969 (cherry picked from commit 1b4987e779d97b90669bac2405a4672085677f7c) 142970 142971commit 749bdf19b2a50201fddd04207e31122470f435e3 142972Author: George Peter Staplin <gps@Georges-Workstation.local> 142973Date: Fri Sep 19 00:02:48 2008 -0600 142974 142975 XQuartz: pbproxy: Add COMPOUND_TEXT handling. Do misc. cleanups with testing 142976 to verify that the behavior didn't change. 142977 142978 main.m: XInternAtom compound_text, and atom_pair. 142979 142980 pbproxy.h: Add compound_text and atom_pair to the struct atom_list. 142981 142982 x-selection.m: Add an #include of Xutil.h. Refactor the reply struct 142983 initialization to be done in a common place. Add send_reply: to simplify 142984 the code a bit more. Add send_compound_text: which handles the 142985 COMPOUND_TEXT type. Add the beginnings of a send_multiple:. Change 142986 handle_image:extension: to handle_image:. The extension: message isn't 142987 needed anymore. 142988 (cherry picked from commit 1e9460abdf5bafe46215966bbef3e796cb1c33e0) 142989 142990commit 00bfbee59fe3b0c8d1a55d1851206857ca563ece 142991Author: George Peter Staplin <gps@Georges-Workstation.local> 142992Date: Thu Sep 18 13:10:41 2008 -0600 142993 142994 XQuartz: pbproxy: pbproxy now responds to selection request events. 142995 142996 We can now copy and paste images to and from X11. Text copying and 142997 pasting works as well. 142998 142999 The NSPasteboard can contain TIFF or PICT images, and pbproxy will 143000 translate to an image/png or image/jpeg request, and list those in the 143001 TARGETS. 143002 143003 I added a description of the basic design at the top of x-selection.m. 143004 143005 I removed the request_data x_selection class struct. It's not needed. 143006 (cherry picked from commit 4a8daf884694c9c420c45d427f1f84e608e7e48f) 143007 143008commit 811e9f5e9e9c07f4b8fb1e9a4d9df91a2e359221 143009Author: George Peter Staplin <gps@Georges-Workstation.local> 143010Date: Wed Sep 17 11:59:36 2008 -0600 143011 143012 XQuartz: pbproxy: Use an NSBitmapImageRep to convert an "image/png" selection 143013 to a TIFFRepresentation for use with NSPasteboard. 143014 143015 This has been tested with the Gimp and works with some minor quartz-wm changes. 143016 The Finder clipboard shows the image updates after an Edit -> Copy. 143017 (cherry picked from commit 12912adaeea759d30f96d8ae51a84fd1659ea0ac) 143018 143019commit d8d555bac77509248d1145e928a2edf500b326b9 143020Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143021Date: Wed Sep 17 10:11:00 2008 -0700 143022 143023 XQuartz: Don't enable 8bit visuals in the TrueColor server, since it's not working yet 143024 (cherry picked from commit 1317c8ff94c4bd32617a5398e929f667bd87da58) 143025 143026commit 23ec8261b6a746cb84e1328284fdc5587bca6faf 143027Author: George Peter Staplin <gps@Georges-Workstation.local> 143028Date: Tue Sep 16 21:09:22 2008 -0600 143029 143030 XQuartz: pbproxy: The greedy CLIPBOARD handling now works for text. 143031 143032 This change adds some [self own_clipboard] calls in the necessary places to get the proper greedy behavior. 143033 143034 UTF8_STRING and STRING properties seem to work well now with the test cases (PRIMARY, and CLIPBOARD). I can copy from several different X apps, and have the behavior be correct when pasting. I also verified that quartz-wm isn't doing the copying, by disabling the quartz-wm paths. 143035 (cherry picked from commit 934669f732c28f07b9d934d8f8f0b63ccfebd900) 143036 143037commit 7c2eb3d41af33d1c52500111d0d34f9aab8a4542 143038Author: George Peter Staplin <gps@Georges-Workstation.local> 143039Date: Tue Sep 16 15:21:18 2008 -0600 143040 143041 XQuartz: pbproxy: return the TARGETS list as a 32-bit list to fix a bug that was in the original. 143042 143043 Add TRACE() calls where appropriate to try to figure out why we are losing CLIPBOARD at times, after transferring PRIMARY to the NSPasteboard. 143044 143045 Use the new pbproxy_clipboard_to_pasteboard where appropriate. 143046 (cherry picked from commit 40190675a6c1889cca3574e5d1a9c16ab74315a7) 143047 143048commit 81c836902dc5b0b83cd95262d48cbc4c81ff3ae2 143049Author: George Peter Staplin <gps@Georges-Workstation.local> 143050Date: Mon Sep 15 13:35:46 2008 -0600 143051 143052 Rename reclaim_clipboard to claim_clipboard. 143053 143054 Convert the puts usage to use DB(). 143055 143056 Add the initial handle_image method. 143057 143058 Check for nil in the NSString instantiation in various places. 143059 143060 Add some commentary to enhance the clarity of why I did some things. 143061 (cherry picked from commit 37361567b65241eab64e8b30cd9729d0e71a86d2) 143062 143063commit 85e23affea7bf9aa7615b0292e67f170266d85f8 143064Author: George Peter Staplin <gps@Georges-Workstation.local> 143065Date: Mon Sep 15 13:31:19 2008 -0600 143066 143067 Add an image_jpeg Atom. 143068 (cherry picked from commit ea9dc27977c5eab666f2aa2d914e4e28d36758c7) 143069 143070commit 6b42f40ff9b9ff3e74ccbb2ce5cfad65016934de 143071Author: George Peter Staplin <gps@Georges-Workstation.local> 143072Date: Mon Sep 15 13:28:50 2008 -0600 143073 143074 Rename reclaim_clipboard to claim_clipboard. 143075 (cherry picked from commit d333a8e2b5514b8b76a78c6a13a3f5149ea8de27) 143076 143077commit 2ab56981b1a64138d6c1f9ecfb75655477b3e575 143078Author: George Peter Staplin <gps@Georges-Workstation.local> 143079Date: Mon Sep 15 13:27:08 2008 -0600 143080 143081 Add image/jpeg type, and rename reclaim_clipboard to claim_clipboard. 143082 (cherry picked from commit 32b175ef62b68a971784e51e937e358cb10e20ac) 143083 143084commit cbd32645f7d1054f79ad137a16766da1076c610e 143085Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143086Date: Mon Sep 15 09:33:55 2008 -0700 143087 143088 XQuartz: pbproxy: Added global variables to customize how pbproxy behaves 143089 (cherry picked from commit 21a2d0b8d03faf1c66ec0c5c11fbd2ab24547727) 143090 143091commit 5b397642cdc29920245ebe64cc85243cac893e0e 143092Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143093Date: Sun Sep 14 09:23:17 2008 -0700 143094 143095 XQuartz: pbproxy: First round of updates to pbproxy from George. 143096 (cherry picked from commit 5c8b956f8f3f17e8d577d97cb66424954be72684) 143097 143098commit dac9e91870f9c787eac97c9d7d0607979c57eb5b 143099Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143100Date: Sun Sep 21 17:32:24 2008 -0700 143101 143102 mieq: Made custom event handlers safer for nevents >1 143103 143104commit 8c46505d7d91e0644b19cccc4b342fceb6f86cab 143105Author: Kim Woelders <kim@woelders.dk> 143106Date: Mon Sep 22 08:37:29 2008 +0930 143107 143108 xkb: fix use of uninitialized variable. 143109 143110 And some cosmetic changes to use stuff->change consistently. 143111 143112 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 143113 143114commit 2b266eda6e23d16116f8a8e258192df353970279 143115Author: Peter Harris <peter.harris@hummingbird.com> 143116Date: Thu Dec 20 15:58:01 2007 -0500 143117 143118 Fix panoramiX request and reply swapping 143119 143120 Fix panoramiX request and reply swapping 143121 Set window and screen values in panoramix replies 143122 Prevent buffer overrun in ProcPanoramiXGetScreenSize 143123 143124commit 03ab8f11d25b0ca39a3b37b5350bca4c1028768e 143125Author: Keith Packard <keithp@keithp.com> 143126Date: Thu Sep 18 12:17:05 2008 -0700 143127 143128 When resizing a window with redirected descendents, don't expose them 143129 143130 Bit/window gravity computations need to recompute exposures to manage the 143131 bits which are saved by gravity during the resize computation. That's easy 143132 for non-redirected windows where the bits are all within the parent's 143133 pixmap. For redirected windows, we don't need to deal with this at all, so 143134 just skip the whole re-computation adventure. 143135 143136 Signed-off-by: Keith Packard <keithp@keithp.com> 143137 143138commit d26083fe02658e7312a8da6a5b11652f9fe57bc8 143139Author: Keith Packard <keithp@keithp.com> 143140Date: Thu Sep 18 12:14:58 2008 -0700 143141 143142 Switching from Automatic to Manual redirect needs to unmap/remap 143143 143144 When changing a window from automatic redirection to manual redirection, the 143145 parent clip list needs to be recomputed; the easy way to get that computed 143146 right is to unmap/map the window, just as when redirecting the window the 143147 first time. 143148 143149 Thanks to Owen Taylor for helping diagnose this. 143150 143151 Signed-off-by: Keith Packard <keithp@keithp.com> 143152 143153commit 8bfd88ca61a1747714781a090d1c9dd8f3ad5dfc 143154Author: Adam Jackson <ajax@redhat.com> 143155Date: Fri Sep 19 12:13:43 2008 -0400 143156 143157 Revert accidental edid changes from previous commit. 143158 143159commit 69b79c1a6648f0d04a1b37a728b2e3d76233530f 143160Author: Adam Jackson <ajax@redhat.com> 143161Date: Fri Sep 19 12:02:28 2008 -0400 143162 143163 Update to SGI FreeB 2.0. 143164 143165 Under the terms of version 1.1, "once Covered Code has been published 143166 under a particular version of the License, Recipient may, for the 143167 duration of the License, continue to use it under the terms of that 143168 version, or choose to use such Covered Code under the terms of any 143169 subsequent version published by SGI." 143170 143171 FreeB 2.0 license refers to "dates of first publication". They are here 143172 taken to be 1991-2000, as noted in the original license text: 143173 143174 ** Original Code. The Original Code is: OpenGL Sample Implementation, 143175 ** Version 1.2.1, released January 26, 2000, developed by Silicon Graphics, 143176 ** Inc. The Original Code is Copyright (c) 1991-2000 Silicon Graphics, Inc. 143177 ** Copyright in any portions created by third parties is as indicated 143178 ** elsewhere herein. All Rights Reserved. 143179 143180 Official FreeB 2.0 text: 143181 143182 http://oss.sgi.com/projects/FreeB/SGIFreeSWLicB.2.0.pdf 143183 143184 As always, this code has not been tested for conformance with the OpenGL 143185 specification. OpenGL conformance testing is available from 143186 http://khronos.org/ and is required for use of the OpenGL logo in 143187 product advertising and promotion. 143188 143189commit cc20112a65d3f641ce0261c86a541f94fae5215c 143190Author: Peter Hutterer <peter.hutterer@redhat.com> 143191Date: Fri Sep 19 08:44:04 2008 +0930 143192 143193 Xi: don't include client-side headers. 143194 143195 Spotted by Mikhail Gusarov. 143196 143197commit e76b5b5e731e2d6c61ff6e6e9b2f42b20ac473d5 143198Author: Peter Hutterer <peter.hutterer@redhat.com> 143199Date: Thu Sep 18 14:51:15 2008 +0930 143200 143201 dix: calloc, not malloc, ValuatorClassRec. 143202 143203 For master devices, the ptraccel code could segfault on free since we'd be 143204 dereferencing random memory. Callocing the valuatorClassRec is the easy fix. 143205 143206commit f789408f7dbfe57a0111e185ac83d1c70f262ba8 143207Author: Peter Hutterer <peter.hutterer@redhat.com> 143208Date: Mon Aug 25 14:52:35 2008 +0930 143209 143210 Xi: remove broken SDeviceEnterNotifyEvent, replace with ...Leave... 143211 143212 The event format is the same for both (bar the type), so one is enough. 143213 143214commit 97b9374a8af5a4cc52f9074feb4ce40cf0b90796 143215Author: Aaron Plattner <aplattner@nvidia.com> 143216Date: Wed Sep 17 15:48:56 2008 -0700 143217 143218 Conditionalize Composite-based backing store on pScreen->backingStoreSupport. 143219 (cherry picked from commit 37876602957924c7cff759a800eddd574ee2385a) 143220 143221commit b4762c0245ed2966606171cf27f40aa745fdc76e 143222Author: Julien Cristau <jcristau@debian.org> 143223Date: Tue Sep 16 17:13:42 2008 +0200 143224 143225 exa: disable shared pixmaps 143226 143227 They got re-enabled in ee7c684f21d220d5e046bab31ae617a7d64d60f6 143228 ("Reimplement ShmPutImage.") 143229 143230commit 52bbfc5bfa64fd09ef625f35a2fb4a0518003825 143231Author: Mikhail Gusarov <dottedmag@dottedmag.net> 143232Date: Mon Sep 15 10:43:46 2008 -0400 143233 143234 scrnsaver is optional at build time. 143235 143236commit b6ab114212c0e4c3346ceb5b207f14c526ab81e7 143237Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 143238Date: Fri Sep 12 19:11:53 2008 -0400 143239 143240 Array-index based devPrivates implementation. 143241 143242 Note: DevPrivateKey is now pointer-to-int, which means 143243 each key now needs to point to some global storage of 143244 size at least sizeof(int). 143245 143246commit 999f3362d50a753c1801d565766219808efdb1a5 143247Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143248Date: Fri Sep 5 17:32:16 2008 -0700 143249 143250 XQuartz: Dead code removal 143251 (cherry picked from commit 43184cd379c4fb7254391b8a362016cbf89b5529) 143252 143253commit 6548a55ebdf43231a3c48debf54aa5aa2f1aa956 143254Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143255Date: Fri Sep 5 17:05:03 2008 -0700 143256 143257 XQuartz: 256 color support 143258 (cherry picked from commit 8dd6d5c825d457f26b41b79d02d57ed4a5ecf1f5) 143259 143260commit bad7cd14c2021b14971b3f707f927803a053003e 143261Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143262Date: Fri Sep 5 10:46:36 2008 -0700 143263 143264 XQuartz: Dead code removal 143265 (cherry picked from commit bf10fb0b1f776e72db7c76db11f764e26f9d62c4) 143266 143267commit 1119fe136f8731f26fc6f50b92f5ddf78f3f83be 143268Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143269Date: Fri Sep 5 10:46:20 2008 -0700 143270 143271 XQuartz: Changed around fd handoff model to be more robust 143272 (cherry picked from commit 4fe605c2bc62d50e0b5764d9edda245af227630b) 143273 143274commit 183233b27beb441742a53e440c3389b4ea125b8a 143275Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143276Date: Fri Sep 12 15:25:44 2008 -0700 143277 143278 XQuartz: Don't warp the pointer on server start for XQuartz 143279 (cherry picked from commit c0da576d4921c246a9ac747921b48ab9e718347f) 143280 143281commit 7a46dd30d332bc1d42f82b0391e806d3b067ebbd 143282Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143283Date: Thu Sep 4 09:18:40 2008 -0700 143284 143285 XQuartz: Dead code removal 143286 (cherry picked from commit 2e45344870f5d17181df6407da3448991036d783) 143287 143288commit 1c70e53f2560cb877c8b149034f296232de67a4e 143289Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143290Date: Thu Sep 4 08:55:22 2008 -0700 143291 143292 XQuartz: Fixed proximity logic errors for tablets. 143293 (cherry picked from commit d942849f248c368b92ca73e145e8a5bc339112d7) 143294 143295commit c661c6ae7a9cd981f914484aa4c3007922280def 143296Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143297Date: Thu Sep 4 08:07:32 2008 -0700 143298 143299 XQuartz: Remove decls of crashtracer strings that we don't use anymore in here 143300 (cherry picked from commit 4a653c6bfd270f2960a8c7e726e7628cfc3c9051) 143301 143302commit 54162aca404a7de2e560cc2aea01445e14742efe 143303Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143304Date: Thu Sep 4 05:18:31 2008 -0700 143305 143306 XQuartz: Fixed a crash in RootlessNativeWindowMoved 143307 (cherry picked from commit a16048cec08e173ce42a78a77704f5fdfcb4480f) 143308 143309commit ce781cac1b52003da6c025cd29e2c2ad97dbb324 143310Author: Eric Anholt <eric@anholt.net> 143311Date: Thu Sep 11 16:24:20 2008 -0700 143312 143313 Fix distcheck. 143314 143315commit 31c62495f1de6e9ba41e1f6d7fa263eeb849129b 143316Author: Keith Packard <keithp@keithp.com> 143317Date: Wed Sep 10 13:11:09 2008 +0800 143318 143319 Drop a reference to user mode after create 143320 143321 User mode has no customer when create until assigned 143322 to some output. 143323 143324commit 079625570d51e41569b73b2fd9237eb8f967f408 143325Author: Aaron Plattner <aplattner@nvidia.com> 143326Date: Mon Sep 8 08:50:52 2008 -0700 143327 143328 Bump ABI major versions for the TryClientExceptions change from commit 883811c. 143329 143330commit 4e004c68283828660f2259a4a4120a84667dcafa 143331Author: Zhenyu Wang <zhenyu.z.wang@intel.com> 143332Date: Mon Sep 8 09:16:59 2008 +0800 143333 143334 Revert "randr: fix user mode create initial reference count" 143335 143336 This reverts commit 7c5ca85a9e6d49ab572831b3e0c08bb4cafb395e. 143337 143338commit 40a9482f8c5b3eac86c221ccce78510e850f4d2b 143339Author: Jesse Barnes <jbarnes@virtuousgeek.org> 143340Date: Fri Sep 5 06:09:51 2008 -0700 143341 143342 Add cscope files to .gitignore 143343 143344commit bf333c2f9833a178887e7bdd7fc338f1e09c387f 143345Author: John Tapsell <john.tapsell@john-desktop.(none)> 143346Date: Wed Sep 3 12:20:17 2008 +0100 143347 143348 fbdevhw: Remove pixclock check. 143349 143350 The check can fail because the output from FBIOGET_VSCREENINFO is used to set 143351 Clock in fbdev2xfree_timing(). Then in fbdevHWSetMode(), xfree2fbdev_timing() 143352 is called which sets the pixclock based on Clock. The resulting circle results 143353 in slight rounding errors, causing the comparision check in fbdev_modes_equal 143354 to fail. 143355 143356commit 47bf269d5922a4998c33908a53ad1affa8de96c0 143357Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143358Date: Wed Sep 3 12:54:56 2008 -0700 143359 143360 mieq: Fix my broken cherry pick. 143361 143362commit 1669308be6d4cc6265e5fa8ef47fbf2cbef6f093 143363Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143364Date: Wed Sep 3 09:23:06 2008 -0700 143365 143366 XQuartz: Added pasteboard proxy code stripped out of quartz-wm. 143367 (cherry picked from commit 1f842c71c35db031a24de646429834d6054adf1d) 143368 (cherry picked from commit 144746223d8dc3f6ff2a0e805aa27a6e57df558a) 143369 143370commit 46168b3b404164c1f2e7876031e70fb77192fe75 143371Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143372Date: Wed Sep 3 09:11:25 2008 -0700 143373 143374 XQuartz: Updated Tablet code to include ranges for valuators and added cursor and erasor 143375 (cherry picked from commit 6bca78760951cb5cb57ea66b7631a2dc230dc27a) 143376 (cherry picked from commit fafcafd6de39cf6b5967793f03b30b49db36fec5) 143377 143378commit 7f5318131ddf86bed5797c0d9641f6ff34f34be0 143379Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143380Date: Wed Sep 3 03:54:42 2008 -0700 143381 143382 XQuartz: Removed "Done" button from Applications Customization. Now properly save on exit. 143383 (cherry picked from commit 2f1e163d5db6063e54fcaa045ea67c29ce3e46aa) 143384 (cherry picked from commit 7542c00b96461c3474b8c23545171c4b6ddfd29d) 143385 143386commit 9e70a3cf58c205948f6a415e203d825e9b660d9d 143387Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143388Date: Wed Sep 3 09:33:39 2008 -0700 143389 143390 mieqProcessInputEvents: Check custom event handlers first. 143391 143392commit 700e8bf33f9b1c3921de912b8110fadf5e0626da 143393Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143394Date: Wed Sep 3 02:41:20 2008 -0700 143395 143396 XQuartz: Updated man page 143397 (cherry picked from commit 6932d0d19ea75c64ed38b82a246b68c6bbfb8cec) 143398 (cherry picked from commit 8e3da36b6b0a261975808bc67b5caca733a1a23d) 143399 143400commit 7430bb9093ea8cf9589d745742c7353e53a144f8 143401Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143402Date: Wed Sep 3 02:37:09 2008 -0700 143403 143404 XQuartz: Removed dead code (unused option definitions) 143405 (cherry picked from commit 3d74f77490105c0704c0b10dc7a6eb859bf6e985) 143406 (cherry picked from commit c1e5ea9ba058f6a13f564165a3d6862335729630) 143407 143408commit b3836e5bdd7e2e04e563246d90f385e1b4741baa 143409Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143410Date: Wed Sep 3 02:36:39 2008 -0700 143411 143412 XQuartz: Cleaned up our sleeping before adding the launchd socket. 143413 (cherry picked from commit ebfe49dfca6e04d66232bad760f522629397cc35) 143414 (cherry picked from commit 2cf74a76d54d795cfc3c0380a526b9177327782d) 143415 143416commit 7c5ca85a9e6d49ab572831b3e0c08bb4cafb395e 143417Author: Zhenyu Wang <zhenyu.z.wang@intel.com> 143418Date: Wed Sep 3 09:41:00 2008 +0800 143419 143420 randr: fix user mode create initial reference count 143421 143422 Don't need extra reference count adding when creating user mode. 143423 This fixes user mode destroy, otherwise we get BadAccess error. 143424 143425commit 4e1f9f51a9eea115ce2a073824e9c5cef4e2f364 143426Author: Daniel Drake <dsd@laptop.org> 143427Date: Tue Sep 2 16:20:40 2008 -0400 143428 143429 Don't try to destroy a NULL pixmap. 143430 143431commit 613ce0955032fb032de0a3940752828d314f057a 143432Author: Adam Jackson <ajax@redhat.com> 143433Date: Tue Sep 2 15:43:15 2008 -0400 143434 143435 Fix strict-aliasing silliness in XAA. 143436 143437commit b5cdcfa55c399e83d51242e93d4f25d8bc4fec1f 143438Author: Julien Cristau <jcristau@debian.org> 143439Date: Mon Sep 1 19:45:30 2008 +0200 143440 143441 Xevie: swap replies as necessary 143442 143443commit d3ae193f4ac87530f2745f8cb5e7b70dd516881e 143444Author: Thorvald Natvig <slicer@users.sourceforge.net> 143445Date: Mon Sep 1 19:36:56 2008 +0200 143446 143447 Xevie: always initialize rep.length (bug#17394) 143448 143449 The XEvIE extension doesn't clear the rep.length field for any reply but 143450 the version check. Hence, if there is junk data in it and that is sent 143451 to the client, it hangs. 143452 143453 X.Org bug#17394 (http://bugs.freedesktop.org/show_bug.cgi?id=17394) 143454 143455commit 0466b8811ce8ebc1d993f177cbb7a22e997640f1 143456Author: Dave Airlie <airlied@redhat.com> 143457Date: Mon Sep 1 09:57:25 2008 +1000 143458 143459 exa: drop cw.h inclusion 143460 143461 this doesn't seem to be needed, at least I can't see any warning without it. 143462 143463 I'd like to build EXA into a driver and cw.h isn't exported 143464 143465commit 881bd15f445bb62c3fdea863c968c95217408b6e 143466Author: Maarten Maathuis <madman2003@gmail.com> 143467Date: Sun Aug 31 18:07:22 2008 +0200 143468 143469 xf86Crtc: Set mode to NULL, so nothing bad happens if crtc == NULL. 143470 143471 - Fix is courtesy of math_b. 143472 143473commit f4e9a1a98f97155de62908309e87c10487125926 143474Author: Maarten Maathuis <madman2003@gmail.com> 143475Date: Sun Aug 31 17:45:08 2008 +0200 143476 143477 damage: choose less ambiguous function names 143478 143479commit 0cd6709db5febbe7fbd073e429421fcd81041e08 143480Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143481Date: Sat Aug 30 16:13:58 2008 -0700 143482 143483 XQuartz: Fixed editing of Applications menu 143484 (cherry picked from commit da647f1add4c1ae7854c0fbfbf972666ce5d12c8) 143485 143486commit c4d290fc54542ccba7c3a46bc75b0c3322d0ca48 143487Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143488Date: Sat Aug 30 14:19:47 2008 -0700 143489 143490 XQuartz: Fixed missing symbol in quartzKeyboard debugging 143491 (cherry picked from commit 8ad55e484f0ba9073ce07a03de660cc546a7b668) 143492 143493commit 5d7d959b1143979c3445639c9d108e02da0c1475 143494Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143495Date: Sat Aug 30 14:18:07 2008 -0700 143496 143497 XQuartz: Don't even build with GLX support since swrast now depends on libdrm 143498 (cherry picked from commit 3739525d57dff2b3b034b5768f357b1c25577909) 143499 143500commit 45def7f0a342f4733cafb06adbfed14932f0e01b 143501Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143502Date: Fri Aug 29 22:16:03 2008 -0700 143503 143504 Removed dead glcontextmodes 143505 (cherry picked from commit c6d0ac7471209d835e8d9a25adcbda893ddfa1cb) 143506 143507commit d01c5ca7935a8340a3cd68c325da6dfec005c952 143508Author: Matthieu Herrb <matthieu.herrb@laas.fr> 143509Date: Sat Aug 30 15:35:32 2008 +0200 143510 143511 Xserver.man: Typo (the the). 143512 143513commit 229e60db8f95232afc8cdcb7cd0572d117c84b90 143514Author: Thomas Bodzar <Thomas.Bodzar@quest.com> 143515Date: Sat Aug 30 15:34:29 2008 +0200 143516 143517 Xorg.man: typo (the the). 143518 143519commit 1885582444fb06da73dea1f118a777bea26e1dc1 143520Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143521Date: Fri Aug 29 21:18:26 2008 -0700 143522 143523 XQuartz: Placate automake which doesn't like escaped newlines even in comments... 143524 (cherry picked from commit 73db1170129a31ad2d55a7e83ac6dfc8030e47f1) 143525 143526commit 8abcc12fdd5bd9edc0288a1ab8f75e65ffcb8c71 143527Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143528Date: Fri Aug 29 20:55:25 2008 -0700 143529 143530 XQuartz: Added debugging output to the crash log to help track down two crashes, since people don't often report their system.log spew. 143531 (cherry picked from commit aaf0f71db197526b6b866cc1b39fbdfe051879ef) 143532 143533commit aee19c7b5542aa147b89b1ae8340e9931d81729f 143534Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143535Date: Fri Aug 29 03:54:55 2008 -0700 143536 143537 XQuartz: xp_is_symbolic_hotkey_event catches the input menu, but not our mainMenu, so we still need to do that the old way. 143538 (cherry picked from commit 25eccf12c89d73b8bce2c9be14841efb230acedc) 143539 143540commit 974db58f5b730c3770ee461665a02dd4334d1dea 143541Author: Maarten Maathuis <madman2003@gmail.com> 143542Date: Sat Aug 30 00:37:11 2008 +0200 143543 143544 damage: initial attempt at a damage marker mechanism 143545 - This should allow drivers to recieve post submission events for X<->opengl synchronisation. 143546 - Lacking a testcase, i'm open to suggestion how to do it better. 143547 - The idea is: 143548 - driver recieves event 143549 - driver creates personal identification and inserts marker into X fifo. 143550 - when something wants to use an X pixmap, it checks if something is pending. 143551 - If so, it synchronizes the 2nd fifo using the initial identification. 143552 - Driver is not required to use interrupt based systems (price too high). 143553 - Lower latency is ofcource better. 143554 - If this is somehow unusable for you, then come up with improvements. 143555 - For that reason i wouldn't consider the api fixed for the moment. 143556 143557commit 454cb0802eec3c2c2cdbcc17971bced868462b83 143558Author: Maarten Maathuis <madman2003@gmail.com> 143559Date: Fri Aug 29 22:28:02 2008 +0200 143560 143561 damage: DamageReportRawRegion should set pDamage->damage 143562 143563 - I found no evidence in the protocol, that it should be differently from all the other modes. 143564 - It seems to have been like this from day 1. 143565 - If anyone has evidence to the contrary, please enlighten me. 143566 143567commit ae6ca434104405302f30a58bde8738d9579d9dc9 143568Author: Maarten Maathuis <madman2003@gmail.com> 143569Date: Fri Aug 29 22:21:54 2008 +0200 143570 143571 damage: internal functions start with a non-capital letter 143572 143573commit 1861250cd7e84b05e8298b74e3c7e97da72ddfba 143574Author: Maarten Maathuis <madman2003@gmail.com> 143575Date: Fri Aug 29 22:15:23 2008 +0200 143576 143577 {damage,exa}: sanitise damage 143578 143579 - Redo damage naming for more consistency. 143580 - Call post submission functions only where appropriate. 143581 - EXA can now live without it's odd damage workarounds. 143582 143583commit 5af77d43fe812e127d5d335527fa940ab9d95f38 143584Author: Kristian Høgsberg <krh@redhat.com> 143585Date: Mon Aug 11 16:59:17 2008 -0400 143586 143587 DRI2: Drop sarea use, implement server side swap buffers. 143588 143589commit 60ad8d5d05485339e89d7f1f9f1ded75de7c7ea1 143590Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 143591Date: Thu Aug 28 23:45:17 2008 -0400 143592 143593 Attempt getpeercon() on remote sockets as well as local ones. 143594 143595commit ebea78cdba0ff14a397239ee1936bd254c181e1b 143596Author: Tomas Carnecky <tom@dbservice.com> 143597Date: Thu Aug 28 18:05:40 2008 -0400 143598 143599 Prepare for array-index based devPrivates. 143600 143601 TODO: static indices can be made just an int; some indices 143602 can be combined. 143603 143604commit ec7907f8fa04dcff2649cc4846975844314f737e 143605Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 143606Date: Wed Aug 27 19:27:13 2008 -0400 143607 143608 Add an array of integers for use as per-screen cursor private keys. 143609 143610 Replaces the use of the screen pointer itself as the key, which was 143611 nice but won't work now that an array index is being stored. 143612 143613commit 86898491497a43814caf42013651086e62fe6162 143614Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 143615Date: Wed Aug 27 19:17:15 2008 -0400 143616 143617 Remove unused GetGlyphPrivatesForScreen. 143618 143619commit 90b178cc7feda1f9c4995b98364739e71233fff3 143620Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 143621Date: Thu Aug 28 19:36:22 2008 -0400 143622 143623 Don't need to request space for Xnest pixmap private. 143624 143625commit 835b532d6f51d08998d7ab49fcb41db8266487ce 143626Author: Adam Jackson <ajax@redhat.com> 143627Date: Thu Aug 28 14:49:35 2008 -0400 143628 143629 Build fix. 143630 143631commit 040212ee60af9d043c4c3f98bba5120c3b8e33f7 143632Author: Adam Jackson <ajax@redhat.com> 143633Date: Thu Aug 28 13:49:35 2008 -0400 143634 143635 MIT-SHM pixmaps, if they exist, are ZPixmap. 143636 143637commit ee7c684f21d220d5e046bab31ae617a7d64d60f6 143638Author: Adam Jackson <ajax@redhat.com> 143639Date: Wed Aug 27 16:05:47 2008 -0400 143640 143641 Reimplement ShmPutImage. 143642 143643 There's no reason to not just dispatch this straight into the GC. As a 143644 bonus, if you do so, damage wraps correctly, and thus swcursor works. 143645 The side effect is it's no longer possible to override ShmPutImage with 143646 ShmRegisterFuncs(). 143647 143648 Also remove the (broken) damage tracking for same from EXA, since it didn't 143649 work right, and is now superfluous. 143650 143651commit 2db1afbf2e56d8743c701d81a5797001ce9e5c52 143652Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143653Date: Tue Aug 26 23:09:20 2008 -0700 143654 143655 XQuartz: Always use TIS for the keyboard layout seed since KB* aren't thread safe. 143656 (cherry picked from commit c8244177b0dbcb28d2f5509e6f2a3e8057889790) 143657 143658commit 319405fa4a530804c19fb7d0fcece5c912e41c9d 143659Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143660Date: Tue Aug 26 21:11:35 2008 -0700 143661 143662 XQuartz: Fixed needed xplugin version. >< 143663 (cherry picked from commit 01b3a99dff79e0d2b316e02658c19fa79d9144ae) 143664 143665commit 3350770bd6d90fefa1133ac738c1d4eae2d568a6 143666Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143667Date: Tue Aug 26 19:13:56 2008 -0700 143668 143669 XQuartz: Use new Xplugin API to determine if a keypress event corresponds to an OS-X hotkey that we should dispatch to Appkit. 143670 (cherry picked from commit e7658e745f8a3eaf5cec9f54a8a1b7e5a7708e09) 143671 143672commit 7930ea9116f232a3762f0085e5ca65e06d1142fc 143673Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143674Date: Mon Aug 25 22:10:46 2008 -0700 143675 143676 XQuartz: Made a note to come back to the xinitrc race condition in the 1.5 branch when we move up to it. 143677 (cherry picked from commit 54a882dcbdccef3f3eb4e6398e13c88185a8986c) 143678 143679commit dc1171df5baa0c4aa366ff2b07af5b8eb44dae12 143680Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143681Date: Fri Aug 22 01:12:37 2008 -0700 143682 143683 XQuartz: Fixed bogus args to debug ErrorF 143684 (cherry picked from commit f0351c2b80f30ae31f041798b84139141fc3d5a5) 143685 143686commit ac936525afe46bad329387232f8d48e77f6a4175 143687Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143688Date: Fri Aug 22 01:11:33 2008 -0700 143689 143690 XQuartz: Cleaned up mouse event translation a bit more. 143691 (cherry picked from commit c286f2a718fdaf6ad10249b59abb20731da8d904) 143692 143693commit 1b42f550001c11c93d5a2144df98a5c687c010c2 143694Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143695Date: Fri Aug 22 00:45:12 2008 -0700 143696 143697 Xquartz: Added missing ev_type for NSTabletPoint event. 143698 (cherry picked from commit 07548f13947a212dcc020d52fe0ad679255121e2) 143699 143700commit 13a89f19208cf998808ff2896b7711310768f94c 143701Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143702Date: Wed Aug 20 10:21:23 2008 -0700 143703 143704 XQuartz: Added more explanation to debug keylayout spew to cut down on report-spam. 143705 (cherry picked from commit fcdc9f8b5abd724d75f53d48fdadbed9859e9e79) 143706 143707commit dd1e54d6eed0bce289517b3cb95e6b1a187ca6e0 143708Author: Maarten Maathuis <madman2003@gmail.com> 143709Date: Tue Aug 26 20:02:55 2008 +0200 143710 143711 exa: fix thinko from 988725f32e082aee9392a71464125157a83d1e67 143712 - the drawable of the pixmap is not the same as the original drawable (possibly a DRAWABLE_WINDOW) 143713 143714commit 7c14fdbacfcd2f4d56a346e6c72e44e4ba9909c1 143715Author: Maarten Maathuis <madman2003@gmail.com> 143716Date: Tue Aug 26 17:21:43 2008 +0200 143717 143718 exa: some minor cleanup 143719 - Fix compile warning 143720 - Order exa.h by source file that exports the function. 143721 - Move the function i created earlier to private headers. 143722 143723commit de79edbd9f67762950eaac2dc79668035239897c 143724Author: Maarten Maathuis <madman2003@gmail.com> 143725Date: Tue Aug 26 17:03:12 2008 +0200 143726 143727 exa: report damage manually for exa{Trapezoids,Triangles} when needed 143728 - Plus a micro cleanup of unused variables. 143729 143730commit 988725f32e082aee9392a71464125157a83d1e67 143731Author: Maarten Maathuis <madman2003@gmail.com> 143732Date: Tue Aug 26 16:54:29 2008 +0200 143733 143734 exa: move destination damage for internal calls to a special function 143735 - This should improve clarity for someone who isn't familiar with the code. 143736 143737commit ce193476808f54d946351458361c62132d81b62f 143738Author: Maarten Maathuis <madman2003@gmail.com> 143739Date: Tue Aug 26 13:18:58 2008 +0200 143740 143741 exa_glyphs: remove useless offset 143742 143743commit fd94651fc3a6f49ec153ffb823b86c1d125bb298 143744Author: Kristian Høgsberg <krh@redhat.com> 143745Date: Tue Aug 26 10:58:35 2008 -0400 143746 143747 Fix driGetConfigAttribIndex unaligned access to GLboolean. 143748 143749 We don't actually send the float mode so just drop it. Drop a couple of 143750 other unused or redundant fields from GLXconfig. 143751 143752commit 41bd8d5dfe27d5c91fe7e4cdba1475a2a0741b1c 143753Author: Adam Jackson <ajax@redhat.com> 143754Date: Mon Aug 25 11:27:26 2008 -0400 143755 143756 EDID: Build fix. 143757 143758commit 76ed409acd772f2c041239345c6dc64cbef0e5b2 143759Author: Adam Jackson <ajax@redhat.com> 143760Date: Mon Aug 25 10:44:45 2008 -0400 143761 143762 EDID: Publish the whole block on the root window if we've got it. 143763 143764commit 5724f7fb5bea6fa1a354c64c0972c53d70e2f27b 143765Author: Adam Jackson <ajax@redhat.com> 143766Date: Mon Aug 25 10:39:36 2008 -0400 143767 143768 EDID: Publish the whole block in the RANDR property if we've got it. 143769 143770commit 668f89eba3e8f9da7843f5cb350f8dc1e5d7efbe 143771Author: Alan Coopersmith <alan.coopersmith@sun.com> 143772Date: Fri Aug 22 10:32:18 2008 -0700 143773 143774 Remove unnecessary #ifdef __SOL8__'s 143775 143776commit cc78d977cac74fcfb7c9b27e7109a1e369018dd8 143777Author: Alan Coopersmith <alan.coopersmith@sun.com> 143778Date: Fri Aug 22 10:14:11 2008 -0700 143779 143780 Convert Solaris #ifdef's for <sys/kd.h> to AC_CHECK_HEADERS check 143781 143782 Upcoming virtual terminal support changes in Solaris kernel will provide 143783 <sys/kd.h> on SPARC too, so this gets us ready for them. 143784 143785commit 95b466e457542bfe08b0ed2bbb5db8d28a961cfd 143786Author: Peter Hutterer <peter.hutterer@redhat.com> 143787Date: Thu Aug 21 17:24:40 2008 +0930 143788 143789 Xi: don't include .c files. 143790 143791commit c696da75c7326c5e1f1cd48292c0519ddc22e11b 143792Author: Peter Hutterer <peter.hutterer@redhat.com> 143793Date: Thu Aug 21 17:23:11 2008 +0930 143794 143795 Xi: swap devices property replies. 143796 143797commit 01264f17925005969c3b71ca945fc1014bcd8c8e 143798Author: Julien Cristau <jcristau@debian.org> 143799Date: Thu Aug 21 01:32:03 2008 +0200 143800 143801 Add swapped dispatch for randr 1.2 requests 143802 143803commit e02f864fdf19a5ab1682336be343c57fdb69ef43 143804Author: Adam Jackson <ajax@redhat.com> 143805Date: Wed Aug 20 13:24:03 2008 -0400 143806 143807 Suppress cursor display until the first XDefineCursor() request. 143808 143809 Yes, this means the server will start without showing a cursor. Pretty 143810 much any application that wants to interact with the mouse will define 143811 cursors, so this essentially just delays showing it until gdm (or 143812 whatever) loads. 143813 143814commit 64ef7ed072007b1d0b4de5ff1e5eababa418c794 143815Author: Adam Jackson <ajax@redhat.com> 143816Date: Wed Aug 20 13:14:03 2008 -0400 143817 143818 Centralize declaration of ConnectionInfo. 143819 143820commit f1f44940f10b0209946ec0f08104a372d7d945a0 143821Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143822Date: Wed Aug 20 09:50:54 2008 -0700 143823 143824 XQuartz: Support windowItemModMask=0 or -1 to disable window shortcuts. 143825 (cherry picked from commit 81187364e512606a6aebb95ab2967d9d420f57f2) 143826 (cherry picked from commit f2ec79e4a159dec6481691e4dd615db01770dd7a) 143827 143828commit 6c5612c2e1df1e89388ffff97251f5328558f8e8 143829Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143830Date: Wed Aug 20 09:21:44 2008 -0700 143831 143832 XQuartz: Added window_item_modifiers defaults item (and option to localization) to change the modifier keys used for the windows menu. 143833 (cherry picked from commit e4110861d307a55d9032cb83cf024ec1f294e8d7) 143834 (cherry picked from commit fcfc05482a88696cb713c7a2ddfcf935ae8378a6) 143835 143836commit bdc277c9b3dc6b3b95f74b20a49ce163c5fae980 143837Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143838Date: Wed Aug 20 09:58:48 2008 -0700 143839 143840 XQuartz: Just cleaned up formatting of event processing code... no "meat" changed... 143841 (cherry picked from commit 745bc8ab387d6794f47d8b9dca33b4c81f6dd39c) 143842 (cherry picked from commit a28a2be52478a1557a363140f7bd70ececf144dc) 143843 143844commit ed42108920cc695b96ed57bf9dcfea1470e74669 143845Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143846Date: Tue Aug 19 02:02:11 2008 -0700 143847 143848 XQuartz: Added appkit_modifiers defaults option which users can set to 'rcommand' to get access to the input menu with right command 143849 (cherry picked from commit 02af74d7a5a7225f408915254c40856159dc7f19) 143850 (cherry picked from commit 2da32894dff8340f6ca7c980277fca7ec835a193) 143851 143852commit be06961312e2111778bfd97b2c8a927acc61f43d 143853Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143854Date: Wed Aug 20 09:56:06 2008 -0700 143855 143856 XQuartz: More input fixes 143857 stuck-modifier fixes (capslock) 143858 3button-emulation now doesn't send the modifier key with the click 143859 Added other options to fake_button2 and fake_button3 defaults options: 143860 ({l,r}{control,alt,command,shift}) 143861 (cherry picked from commit 8fb6a1cf44c35a20dfb0c05c9824f94898f88da7) 143862 (cherry picked from commit ae9c1b3cfb3874b4d1251681c24bda91c398bcab) 143863 143864commit 436b659091dd17bb3e316377ba31f5d2a5dfa12e 143865Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 143866Date: Thu Aug 14 09:34:12 2008 -0700 143867 143868 XQuartz: Updated bundle version to 2.3.1 143869 (cherry picked from commit eedecba0b882bb07931e8d9168589f5be7a08a69) 143870 (cherry picked from commit 7d9d864461858b25c5eb5858fdf4b53771795f11) 143871 143872commit bc4cef1d0ae7cbb8f5fe16d576db45abd4330370 143873Author: Maarten Maathuis <madman2003@gmail.com> 143874Date: Mon Aug 18 16:28:33 2008 +0200 143875 143876 exa: remove useless cache{X,Y}off from UploadToScreen() arguments. 143877 143878commit 38bdc34d555f956a1ff1e1bb8eef95e3ba2bac64 143879Author: Adam Jackson <ajax@redhat.com> 143880Date: Tue Aug 19 15:12:39 2008 -0400 143881 143882 Remove redundant check. 143883 143884commit 47b31233c26f710be70ffcca17e5402d03d5733f 143885Author: Adam Jackson <ajax@redhat.com> 143886Date: Tue Aug 19 15:10:12 2008 -0400 143887 143888 Simplify driver setup. 143889 143890 No point warning about missing driver hooks, that just means the person 143891 who gave you the driver is inept. Might as well just crash. Also, 143892 just name anonymous screens as screen%d instead of failing after the 36th 143893 screen. Bonus points if you can figure out what the failure mode would 143894 be on the 36th screen, and what the effective screen limit was. 143895 143896commit b99fc65b53769400821a2bd8f691ece35a5b8847 143897Author: Adam Jackson <ajax@redhat.com> 143898Date: Tue Aug 19 15:06:53 2008 -0400 143899 143900 Simplify calling xf86EnableIO() a bit. 143901 143902commit b1f2c7a89ace1f6e5750783f3c43c4c5a6d24374 143903Author: Adam Jackson <ajax@redhat.com> 143904Date: Tue Aug 19 15:06:11 2008 -0400 143905 143906 The ->Identify hook is not mandatory. 143907 143908 Best comment ever though. Mad props. 143909 143910commit af5125968a738ec8f496d04eb65acb7c8f812115 143911Author: Adam Jackson <ajax@redhat.com> 143912Date: Tue Aug 19 14:47:10 2008 -0400 143913 143914 Refactor misprite a bit. 143915 143916 No variable declarations in header files, please. 143917 143918commit 41b68e0dea9305d66bca2fc4ad96db01f5342c6d 143919Author: Adam Jackson <ajax@redhat.com> 143920Date: Tue Aug 19 11:02:31 2008 -0400 143921 143922 Remove unused -co option. 143923 143924 What did this even do anyway? 143925 143926commit 1f416fba994ed7a7e072a9f0a86b515855ea3bac 143927Author: Adam Jackson <ajax@redhat.com> 143928Date: Tue Aug 19 10:56:45 2008 -0400 143929 143930 Remove unused -bestRefresh option. 143931 143932commit 5a72c45d42abc7227c6cf3d14fd7043ea7527c54 143933Author: Adam Jackson <ajax@redhat.com> 143934Date: Tue Aug 19 10:54:11 2008 -0400 143935 143936 Remove unused -showunresolved option 143937 143938commit 9f08ffc557b52e2e8cd54fb692b66700c83d61c6 143939Author: Adam Jackson <ajax@redhat.com> 143940Date: Tue Aug 19 10:29:23 2008 -0400 143941 143942 Remove unused xf86Info.pMouse 143943 143944commit f227fbf74f0c619ecf3275cdb4c10b1a4b3a8cfc 143945Author: Michel Dänzer <michel@tungstengraphics.com> 143946Date: Tue Aug 19 11:36:12 2008 +0200 143947 143948 EXA: Inline Prepare/FinishAccessWindow into only caller, ChangeWindowAttributes. 143949 143950 Also check the requested mask in addition to the GC state before doing work. 143951 143952commit 825b3fe11d1b813bf8d5b24a880ed04b78ae1acf 143953Author: Michel Dänzer <michel@tungstengraphics.com> 143954Date: Tue Aug 19 11:22:40 2008 +0200 143955 143956 EXA: Don't use exaGlyphs if the driver doesn't provide a PrepareComposite hook. 143957 143958 It's buggy without Composite acceleration (leading to cropped glyphs) and not 143959 really useful in that case anyway. The bug probably still needs to be found and 143960 fixed for drivers that provide a PrepareComposite hook but can't accelerate 143961 text rendering though. 143962 143963commit 75e495a4cd823b6fa42a8d167ecc9f4723320dea 143964Author: Adam Jackson <ajax@redhat.com> 143965Date: Mon Aug 18 17:58:08 2008 -0400 143966 143967 Remove unused xf86Info.mouseLocal 143968 143969commit b3dfd7e3e4deef21b40c885f8e7eae34b55888b2 143970Author: Adam Jackson <ajax@redhat.com> 143971Date: Mon Aug 18 17:56:18 2008 -0400 143972 143973 Remove unused xf86Info.sharedMonitor 143974 143975commit 5e43cd28692bc05cac80f38b47104a26c0524385 143976Author: Adam Jackson <ajax@redhat.com> 143977Date: Mon Aug 18 17:46:42 2008 -0400 143978 143979 Remove the remainder of grab deactivation and closedown. 143980 143981commit cd1e8f26147919227e7624ac4c6b313d972a4d35 143982Author: Adam Jackson <ajax@redhat.com> 143983Date: Mon Aug 18 17:40:46 2008 -0400 143984 143985 Remove unused EstimateSizesAggressively option. 143986 143987commit faaf0046155abbd15415d5a6b62ead4f58935c28 143988Author: Mathieu Bérard <mathieu.berard@crans.org> 143989Date: Mon Aug 18 17:23:26 2008 -0400 143990 143991 Remove xaaTEGlyphBlt.S 143992 143993 Should have been removed in 593144dddd977f53bcd1a115f9544eeece46df4c but 143994 ajax was asleep at the wheel. 143995 143996commit 562462e78dd6a5e2ed5b36aa22be52f31d55485d 143997Author: Mathieu Bérard <mathieu.berard@crans.org> 143998Date: Mon Aug 18 17:05:24 2008 -0400 143999 144000 Remove dead {bsd,lnx}Resource.c 144001 144002commit e3e47b35e369bf4abb0fe15865acca0b34600b1f 144003Author: Adam Jackson <ajax@redhat.com> 144004Date: Mon Aug 18 17:03:22 2008 -0400 144005 144006 Remove MAX_PCI_{BUSES,DOMAINS} and some associated #if 0. 144007 144008commit 593144dddd977f53bcd1a115f9544eeece46df4c 144009Author: Mathieu Bérard <mathieu.berard@crans.org> 144010Date: Mon Aug 18 16:46:36 2008 -0400 144011 144012 Remove some unused assembly code and assyntax.h 144013 144014commit 2d3e478384ef8af2a760f39792e35e4a6174c967 144015Author: Adam Jackson <ajax@redhat.com> 144016Date: Mon Aug 18 15:09:44 2008 -0400 144017 144018 int10: Fix int1A for %ax == 0xB101. 144019 144020 pciNumBuses was pretty much always 1 post-pciaccess. That ain't good. 144021 144022commit 539717fe1f08a000d1eef9f345d24cb49e3663b5 144023Author: Adam Jackson <ajax@redhat.com> 144024Date: Mon Aug 18 14:50:33 2008 -0400 144025 144026 Remove (unused) pciMaxBusNum. 144027 144028commit 3c03d9f1efbbacec6e8be58da99bf0977a8e0fec 144029Author: Adam Jackson <ajax@redhat.com> 144030Date: Mon Aug 18 14:37:42 2008 -0400 144031 144032 Remove sparcPci.c 144033 144034 There is no way this code can have been building for anyone since pciaccess 144035 was merged. BSD and Linux were already using OS code on sparc, the only 144036 people who could want this are Solaris, who should be using pciaccess 144037 anyway. 144038 144039commit 9a39b6cfbda306eb594e87ce828afa01cd7aa01e 144040Author: Adam Jackson <ajax@redhat.com> 144041Date: Mon Aug 18 14:26:01 2008 -0400 144042 144043 Remove XF86SCANPCI_WRAPPER. 144044 144045commit 470b05a610e641094d46742393684acd786d78a8 144046Author: Adam Jackson <ajax@redhat.com> 144047Date: Mon Aug 18 14:03:36 2008 -0400 144048 144049 Remove vestigial MAX_PCI_DEVICES. 144050 144051commit 244a635fcdc9e0a7212d51b26d74f49d8e1b071f 144052Author: Fredrik Höglund <fredrik@kde.org> 144053Date: Mon Aug 18 19:27:34 2008 +0200 144054 144055 Fix the tile offset in miPaintWindow for ParentRelative windows. 144056 144057commit fdf7c747a8e1bc59dfb31b7f90dd5eab2687315b 144058Author: Adam Jackson <ajax@redhat.com> 144059Date: Mon Aug 18 13:13:03 2008 -0400 144060 144061 Remove PciProbeType and associated weirdness. 144062 144063 This code was effectively only used in ix86Pci.c to select PCI config 144064 access type. Nobody should be using that path anymore, in the glorious 144065 pciaccess world; kernel services should get it right for you. 144066 144067commit 95bb6f53624a3e6f4d62a2f789982c5544d2fc70 144068Author: Mathieu Bérard <mathieu.berard@crans.org> 144069Date: Mon Aug 18 11:47:19 2008 -0400 144070 144071 Remove ancient unused inline hack. 144072 144073commit e1ae8db625b5e8c298a557592ef23656b3da4886 144074Author: Adam Jackson <ajax@redhat.com> 144075Date: Mon Aug 18 09:53:48 2008 -0400 144076 144077 Drop dead PowerPC PCI code. 144078 144079 Also add a warning for anyone still using legacy PCI code. 144080 144081commit 99583b43a9a202d047ff417d47485e4c0e0c9670 144082Author: Jie Luo <clotho67@gmail.com> 144083Date: Sun Aug 17 23:13:22 2008 +0200 144084 144085 glx: avoid possible NULLptr deref, fix #16884 144086 144087commit 421b7e8f12083e9518fa7deda968a2f73c9a0006 144088Author: Maarten Maathuis <madman2003@gmail.com> 144089Date: Sun Aug 17 19:57:02 2008 +0200 144090 144091 exa: fix assert logic thinko from 361a9eb953aaa38f8ebc057185de29e50f9eef26 144092 144093 - I guess failing PrepareAccess is rare, since this a 3 year old bug. 144094 144095commit 1e62e773273539352bc21b2da5262678dd3fc7df 144096Author: Peter Hutterer <peter.hutterer@redhat.com> 144097Date: Fri Aug 15 15:19:44 2008 +0930 144098 144099 Require inputproto 1.9.99.4 144100 144101commit 609c04b62bdb0e86f33860069033c7142eecf210 144102Author: Peter Hutterer <peter.hutterer@redhat.com> 144103Date: Fri Aug 15 14:51:59 2008 +0930 144104 144105 dix: stick a warning in about loss of subpixel precision. 144106 144107commit de1573172ad2c2e98c101727930a468213295783 144108Author: Peter Hutterer <peter.hutterer@redhat.com> 144109Date: Wed Aug 13 13:11:44 2008 +0930 144110 144111 Backport device properties to XI 1.5 instead of XI 2.0 144112 144113commit 8b9ed7dec2954d4890ddcc7c874fc0832b16495a 144114Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144115Date: Thu Aug 14 09:08:14 2008 -0700 144116 144117 XQuartz: Fixed a stuck modifier key bug. 144118 (cherry picked from commit eeb6e5a9e98dcf045ec230f160d5992080dceba6) 144119 144120commit 1770c85374883229ec5e9685229b2333e62705e6 144121Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144122Date: Thu Aug 14 07:32:37 2008 -0700 144123 144124 XQuartz: Made 3-button mouse simulation a little more consistent. 144125 (cherry picked from commit d207b037d2ae213369e5627a17d8831c9bc16ad8) 144126 144127commit f51f77d25e69b51fa3fd557ffdb5573b61759706 144128Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144129Date: Wed Aug 13 19:09:05 2008 -0700 144130 144131 XQuartz: Reverted "Control" text to be accurately "Command" in the input prefs. 144132 (cherry picked from commit b287c481e1cbe77ff84d10a708505f148a0c8434) 144133 144134commit 04956b80431169e0ae713a3e6ba4cdc157ce3a66 144135Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144136Date: Sat Aug 9 10:07:32 2008 -0700 144137 144138 Transport/Apple: Add TRANS_NOXAUTH to incoming connections when the listener is flagged as such 144139 (cherry picked from commit ba4a7924261070ad0aff7211b7a1c9581fad4646) 144140 144141commit e6813e8de65eee854bbffe6ab0f8ba158f43b10b 144142Author: Peter Hutterer <peter.hutterer@redhat.com> 144143Date: Wed Aug 13 14:24:45 2008 +0930 144144 144145 Xi: byte-swap device property requests. 144146 144147commit 0ce687634fa5e0d955e896a0d0d5f584a4cfd1a4 144148Author: Adam Jackson <ajax@redhat.com> 144149Date: Wed Aug 13 17:28:08 2008 -0400 144150 144151 Eviscerate save-under support. 144152 144153 Use a compositing manager already. Plus I really wanted to use 'eviscerate' 144154 in a commit message. 144155 144156commit f456f32cd97684052cab4712e7f4c827132c554b 144157Author: Adam Jackson <ajax@redhat.com> 144158Date: Wed Aug 13 16:21:28 2008 -0400 144159 144160 Build fix. 144161 144162commit c1e9b7dc6f4600fe6c301aecd7d67461cb884df0 144163Author: Adam Jackson <ajax@redhat.com> 144164Date: Wed Aug 13 14:13:55 2008 -0400 144165 144166 Sledgehammer off the DRI2 build system. 144167 144168commit 544bfd06663ff71bc5275459cf74a3557530544f 144169Author: Adam Jackson <ajax@redhat.com> 144170Date: Wed Aug 13 11:09:54 2008 -0400 144171 144172 Remove xf86Version.h and related API. 144173 144174 We haven't meaningfully been API-compatible with xf86 modules in ages, 144175 let's stop pretending. 144176 144177commit 139c3ab7b754dc425a09d5f7b6d2a8fb2b88138a 144178Author: Peter Hutterer <peter.hutterer@redhat.com> 144179Date: Tue Aug 12 17:49:49 2008 +0930 144180 144181 Xi: return the state from the correct keyboard in QueryDevicePointer. 144182 144183commit 998375f4154b000f75b2bafd4e276c0237d24dc7 144184Author: Peter Hutterer <peter.hutterer@redhat.com> 144185Date: Tue Aug 12 17:49:16 2008 +0930 144186 144187 dix: remove superfluous check. 144188 144189commit 9f9268821b13038556fbc029df54ab0e9b2aa77f 144190Author: Mathieu Bérard <mathieu.berard@crans.org> 144191Date: Mon Aug 11 13:52:38 2008 -0400 144192 144193 The smart scheduler is not optional. 144194 144195commit 2e2ce817ce404a5e000c9750fa96f656fed370b8 144196Author: Matthieu Herrb <matthieu.herrb@laas.fr> 144197Date: Sun Aug 10 23:07:46 2008 +0200 144198 144199 Move strcasecmp(), strcasencmp() and strcasestr() prototypes to os.h 144200 144201 And make sure os.h is included in files that use it. 144202 144203commit 02efa78ce2da470b252289ff2af598d06bc84ece 144204Author: Matthieu Herrb <matthieu.herrb@laas.fr> 144205Date: Sun Aug 10 18:29:35 2008 +0200 144206 144207 add libXinerama to the list of dependencies. 144208 144209 Now that panoramiXext.h is installed bt libXinerama, it becomes required 144210 to build Xext. 144211 144212commit be2210b69defa225d8bc4bb90d94bfd740ef8b1f 144213Author: Matthieu Herrb <matthieu.herrb@laas.fr> 144214Date: Sat Aug 9 23:43:50 2008 +0200 144215 144216 remove OpenBSD/amd64 specific mtrr API. 144217 144218 The old code never worked anyways and was removed from OpenBSD. 144219 OpenBSD/amd64 4.4 and later support mtrr with the same API as 144220 OpenBSD/i386. 144221 144222commit 6e33e6f355f7f04e77a165eb67b1414724c1fba3 144223Author: Matthieu Herrb <matthieu.herrb@laas.fr> 144224Date: Sat Aug 9 23:43:03 2008 +0200 144225 144226 Move string comparaison functions to from dix/ to os/. 144227 144228commit 5968634996c08656a0c5e2fa35705cf7afac87e6 144229Author: Dave Miller <davem@davemloft.net> 144230Date: Sat Aug 9 16:45:59 2008 +1000 144231 144232 glx: fix crash in driGetConfigAttribIndex 144233 144234 Don't access GLboolean via int pointers 144235 144236commit 4212599c922373a224d2235c74672a3a3aa8e0b1 144237Author: Michel Dänzer <michel@tungstengraphics.com> 144238Date: Fri Aug 8 12:17:58 2008 +0200 144239 144240 EXA: Make sure damage tracking code is inactive if the driver manages pixmaps. 144241 144242 It was always supposed to be like that... It was only recently pointed out (in 144243 a rather convoluted way) that it was not in fact the case. 144244 144245commit 073116cc44859e96374cde46325df8540621d5ee 144246Author: Michel Dänzer <michel@tungstengraphics.com> 144247Date: Fri Aug 8 12:15:27 2008 +0200 144248 144249 Remove commented out lines that make automake 1.9 puke. 144250 144251commit 666781cd9b62207e45eebd3eb55a930f3a345f72 144252Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144253Date: Fri Aug 8 01:41:28 2008 -0700 144254 144255 Removed rogue #include 144256 (cherry picked from commit af238e99da45a4dd8d8cbb6564bb8d0a3d79d590) 144257 144258commit 4a0947bf06deacc3ae716730cfbbbeda1311bae3 144259Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144260Date: Fri Aug 8 01:53:36 2008 -0700 144261 144262 XQuartz: Disable the Xquartz AIGLX for now since it doesn't even compile any more... 144263 (cherry picked from commit 4545ba91e9d6ad62b6cafde3c73a672d3ded91a2) 144264 144265commit 5accc9b3ce1af539150d59b1800b66783fd3e947 144266Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144267Date: Fri Aug 8 00:19:41 2008 -0700 144268 144269 XQuartz: Re-added deprecated code fallback failsafe for keyboard layout on Leopard with some debugging spew. 144270 (cherry picked from commit 5854e712e9ebc210d2f8de6f5d4fb650944f314a) 144271 144272commit 465a9bb96e8a90e53e655b3245e22910201786d7 144273Author: Peter Hutterer <peter.hutterer@redhat.com> 144274Date: Fri Aug 8 17:37:06 2008 +0930 144275 144276 dix: don't try to create events if we don't have a screen. #16898 144277 144278 A NULL screen may happen during server shutdown, when the output has been shut 144279 down but the devices still generate events. 144280 144281 X.Org Bug 16898 <http://bugs.freedesktop.org/show_bug.cgi?id=16898> 144282 144283commit d684f5760f40b682a8b879641300abe689a263c5 144284Author: Peter Hutterer <peter.hutterer@redhat.com> 144285Date: Fri Aug 8 14:18:01 2008 +0930 144286 144287 xkb: actually initialise sli before using it. 144288 144289commit e00cd54c1ac57ea6e7767e90592200f343d9d08a 144290Author: Simon Thum <simon.thum@gmx.de> 144291Date: Thu Aug 7 11:06:21 2008 +0200 144292 144293 dix: remove misleading comment in ptrvelo.c 144294 144295 Signed-off-by: Peter Hutterer <peter.hutterer@redhat.com> 144296 144297commit 6ca34549b89e73e3e0ec3cb7585686b880534cef 144298Author: Peter Hutterer <peter.hutterer@who-t.net> 144299Date: Thu Aug 7 16:27:57 2008 +0930 144300 144301 dix: remove obsolete FIXME comment. 144302 144303 Thanks to Simon Thum for pointing this out. 144304 144305commit 2c3645581ee3f180d34e32c8016d5e4e1af4dca4 144306Author: Peter Hutterer <peter.hutterer@who-t.net> 144307Date: Thu Aug 7 11:55:10 2008 +0930 144308 144309 dix: SetModifierMapping should only apply to the ClientPointer and it's SDs. 144310 144311commit e1b286d495b760cf67ddf936eca11da50fba847a 144312Author: Simon Thum <simon.thum@gmx.de> 144313Date: Wed Aug 6 09:42:32 2008 +0200 144314 144315 dix: simplified debugging helper for pointer acceleration 144316 144317 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144318 144319commit d95136e1510a48cc3ad066d880d56061fc6a42ee 144320Author: Simon Thum <simon.thum@gmx.de> 144321Date: Thu Jul 31 00:39:03 2008 +0200 144322 144323 xf86: don't replicate dix defaults for pointer acceleration 144324 144325 the defaults from InitVelocityData() or hypothetic driver-side changes 144326 are now respected, not overridden. 144327 144328 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144329 144330commit 4f333d53510b82db57cfac1bfea22422cbb7be79 144331Author: Simon Thum <simon.thum@gmx.de> 144332Date: Thu Jul 31 00:33:38 2008 +0200 144333 144334 dix: simplify velocity tracking filters 144335 144336 prefer fp-mul over fp-div and remove rather pointless check 144337 144338 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144339 144340commit bf084a0769eee36ff799e5e5f2d1e875c1ebcc51 144341Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144342Date: Wed Aug 6 13:36:43 2008 -0700 144343 144344 glcontextmodes.[hc] were not added with the removal of the meas symlinks patch. Copied from mesa head 2008.08.06. 144345 (cherry picked from commit 409e1dd1e9524b5c1a1ae58a759da77e587e3780) 144346 144347commit e72b1d21e44712e90595b1c31e67db6786d0d1bd 144348Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144349Date: Wed Aug 6 10:23:31 2008 -0700 144350 144351 XQuartz: Now properly disable xauth checking on launchd socket and mostly fix the xinitrc / launchd race condition 144352 144353commit d1b96cdea96f5321e0d254abde81cdaab2eca979 144354Author: Aaron Plattner <aplattner@nvidia.com> 144355Date: Wed Aug 6 10:37:32 2008 -0700 144356 144357 Add an xf86PrintBacktrace wrapper around xorg_backtrace to aid driver debugging. 144358 144359commit 26d31ad1c7f4c550d73419ecf76912d844186b30 144360Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144361Date: Tue Aug 5 15:14:08 2008 -0700 144362 144363 XQuartz: Added code and made comments more helpful for debugging first-client-auth bug. 144364 (cherry picked from commit a8f0d32216e321b8ae6da182be9b1ea792f6e004) 144365 144366commit 0ef3243fb59fa993a7a07a08debbd4329254c265 144367Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144368Date: Tue Aug 5 18:45:54 2008 -0700 144369 144370 XQuartz: More fd handoff cleanup. 144371 (cherry picked from commit 199d2dcb708c9f9d6b9ede149ea0fe9fba6c007c) 144372 144373commit e7dd3dd8f311d3872e0c4c9c2ca22c00f9d11f59 144374Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144375Date: Tue Aug 5 14:33:03 2008 -0700 144376 144377 XQuartz: Fixed some errors / typos in the preferences pane. 144378 (cherry picked from commit e1e0c398bbb50394b164394e1f1870016489c25b) 144379 144380commit 6836f9aac801c3add0ae4aece05b32e85a0c5f85 144381Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144382Date: Tue Aug 5 14:24:25 2008 -0700 144383 144384 XQUartz: UI Cleanup. Removed done/cancel buttons from Applications->Customize 144385 (cherry picked from commit d3157ca45c1f88edd040d721c1e944b9d2c090aa) 144386 144387commit c06e27b2f6fd9f7b9f827623a48876a225264132 144388Author: Peter Hutterer <peter.hutterer@who-t.net> 144389Date: Fri Aug 1 16:42:15 2008 +0930 144390 144391 xkb: ProcXkbSetDeviceInfo should work on all attached SDs. 144392 144393 If called with XkbUseCoreKbd, run through all attached SDs and replicate the 144394 call. This way, we keep the SDs in sync with the MD as long as core clients 144395 control the MDs. 144396 144397commit d9ca9819e975e0f6832a320f8be5958e5d942f85 144398Author: Peter Hutterer <peter.hutterer@who-t.net> 144399Date: Fri Aug 1 16:41:40 2008 +0930 144400 144401 xkb: ProcXkbSetGeometry should work on all attached SDs. 144402 144403 If called with XkbUseCoreKbd, run through all attached SDs and replicate the 144404 call. This way, we keep the SDs in sync with the MD as long as core clients 144405 control the MDs. 144406 144407commit 5ba87c3327786dd7c6e8e265a19c858e8faae8fd 144408Author: Peter Hutterer <peter.hutterer@who-t.net> 144409Date: Fri Aug 1 16:41:08 2008 +0930 144410 144411 xkb: ProcXkbSetNames should work on all attached SDs. 144412 144413 If called with XkbUseCoreKbd, run through all attached SDs and replicate the 144414 call. This way, we keep the SDs in sync with the MD as long as core clients 144415 control the MDs. 144416 144417commit 7e45c80204e06562d4475741caea65bc8758f3c7 144418Author: Peter Hutterer <peter.hutterer@who-t.net> 144419Date: Fri Aug 1 16:40:25 2008 +0930 144420 144421 xkb: ProcXkbSetNamedIndicator should work on all attached SDs. 144422 144423 If called with XkbUseCoreKbd, run through all attached SDs and replicate the 144424 call. This way, we keep the SDs in sync with the MD as long as core clients 144425 control the MDs. 144426 144427commit a609dbed7cf854a6ee9d33c7cf45615db9da1fb9 144428Author: Peter Hutterer <peter.hutterer@who-t.net> 144429Date: Fri Aug 1 16:38:59 2008 +0930 144430 144431 xkb: ProcXkbSetIndicatorMap should work on all attached SDs. 144432 144433 If called with XkbUseCoreKbd, run through all attached SDs and replicate the 144434 call. This way, we keep the SDs in sync with the MD as long as core clients 144435 control the MDs. 144436 144437commit e8c2a3d7c996cb41c4c44ba67acae5ff9438fc06 144438Author: Peter Hutterer <peter.hutterer@who-t.net> 144439Date: Fri Aug 1 16:37:39 2008 +0930 144440 144441 xkb: ProcXkbSetCompatMap should work on all attached SDs. 144442 144443 If called with XkbUseCoreKbd, run through all attached SDs and replicate the 144444 call. This way, we keep the SDs in sync with the MD as long as core clients 144445 control the MDs. 144446 144447commit 3c7740aa8f21c1fda4190a1bf5d6ce089db5e35b 144448Author: Peter Hutterer <peter.hutterer@who-t.net> 144449Date: Fri Aug 1 16:36:00 2008 +0930 144450 144451 xkb: ProcXkbSetMap should work on all attached SDs. 144452 144453 If called with XkbUseCoreKbd, run through all attached SDs and replicate the 144454 call. This way, we keep the SDs in sync with the MD as long as core clients 144455 control the MDs. 144456 144457commit 31afd51dd49c0d0db2465fbc987044fab8b89f22 144458Author: Peter Hutterer <peter.hutterer@who-t.net> 144459Date: Fri Aug 1 16:32:37 2008 +0930 144460 144461 xkb: ProcXkbBell should work on all attached SDs. 144462 144463 If called with XkbUseCoreKbd, run through all attached SDs and replicate the 144464 call. This way, we keep the SDs in sync with the MD as long as core clients 144465 control the MDs. 144466 144467commit de4936d7482f820728efeef338a2041c7a9186d2 144468Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144469Date: Mon Aug 4 20:11:45 2008 -0700 144470 144471 XQuartz: Fixed proper processing of tablet button presses 144472 (cherry picked from commit ca0babafa4e9cf1b67ec460655bffe569ac9c3e9) 144473 144474commit b70a8ba3dc52b277194ab267a101a3fd15062685 144475Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144476Date: Tue Jul 15 17:48:29 2008 -0700 144477 144478 XQuartz: Nuke Sparkle. 144479 (cherry picked from commit 4991f54a8d84a9a8df89a99dbfc09391195578f8) 144480 144481commit 30851efdd4313506e92aeb6d3e5099dea99ab99c 144482Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 144483Date: Tue Aug 5 12:54:51 2008 -0700 144484 144485 XQuartz: Fixed first-client-can't-connect bug 144486 Readded the old exec() server startup path for regression testing. 144487 Don't use the dynamic fd addition code since it's not quite working correctly. 144488 (cherry picked from commit 08f3fe153edc5ab4ca010e8ce82d5c3fc0ddb72c) 144489 144490commit 277a74bcbb7d0a93c4f2e1de11daabd8c5f93ee8 144491Author: Tomas Carnecky <tom@dbservice.com> 144492Date: Mon Aug 4 23:26:00 2008 +0300 144493 144494 Redefine clients as a fixed array 144495 144496 This removes yet another xalloc() each server generation. Also, I 144497 couldn't find the corresponding xfree() so I guess that used to be a 144498 memory leak there. 144499 144500commit 5532d63488ec45953ff7f925cfb4f87adb3b04a0 144501Author: Tomas Carnecky <tom@dbservice.com> 144502Date: Mon Aug 4 23:06:08 2008 +0300 144503 144504 Redefine WindowTable as a fixed array 144505 144506 Instead of xalloc'ing it every server generation. The array is always 144507 the same size (MAXSCREENS), anyway. 144508 144509commit e882ee7056f370e0619d137b4ec3973ecb4e3479 144510Author: Daniel Stone <daniel@fooishbar.org> 144511Date: Tue Jul 29 17:52:11 2008 +0300 144512 144513 EXA: Remove unnecessary #includes 144514 144515 There wasn't actually any font code here, so no problem. 144516 144517commit 6c1bb64c6f2efd81856ca00591df37916713b142 144518Author: Peter Hutterer <peter.hutterer@who-t.net> 144519Date: Mon Aug 4 15:19:45 2008 +0930 144520 144521 xkb: break up XkbCopyKeymap into bite-sized chunks. 144522 144523commit 3c6a9c531f673b7a0cb9ca01860b4dbe79686363 144524Author: Peter Hutterer <peter.hutterer@who-t.net> 144525Date: Fri Aug 1 15:52:07 2008 +0930 144526 144527 config: protect against potential out-of-bounds indexing. 144528 144529commit 92c51b183c2ff06361dad7f918daed6577ba4935 144530Author: Peter Hutterer <peter.hutterer@who-t.net> 144531Date: Fri Aug 1 14:24:54 2008 +0930 144532 144533 config: support type strlist for XkbOptions property. 144534 144535 For backwards compatibility with server 1.4. 144536 144537commit 35b14519b4a3158592a089170ec039bbc219603e 144538Author: Peter Hutterer <peter.hutterer@who-t.net> 144539Date: Tue Jul 29 12:59:57 2008 +0930 144540 144541 config: add parsing for input.x11_options.XkbOptions. #16874 144542 144543 X.Org Bug 16874 <http://bugs.freedesktop.org/show_bug.cgi?id=16784> 144544 144545commit d762c08aebe3b7e8c88e2e7a6fcf66057a21b403 144546Author: Simon Thum <simon.thum@gmx.de> 144547Date: Mon Jul 28 14:07:48 2008 +0200 144548 144549 dix: export driver-side functions for acceleration 144550 144551 also add additional safety for accel driver api 144552 144553 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144554 144555commit 18e9fd69fe01298d825b46415b9c6bd86c75dfe5 144556Author: Simon Thum <simon.thum@gmx.de> 144557Date: Tue Jul 29 10:07:43 2008 +0200 144558 144559 dix: use average of pointer accel profile 144560 144561 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144562 144563commit 25882af6d3359e5ae42c927c555f5b257ba5665c 144564Author: Ivaylo Boyadzhiev <iboyadzhiev@nvidia.com> 144565Date: Sun Aug 3 18:55:12 2008 -0700 144566 144567 Fix a longstanding XAA CopyPlane bug. 144568 144569 TmpBitPlane is a plane mask, not a plane index. 144570 144571 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 144572 144573commit d980913f3145cdc34baab27ff818c9631c4c8571 144574Author: Adam Jackson <ajax@redhat.com> 144575Date: Fri Aug 1 16:15:04 2008 -0400 144576 144577 Erk. Actually check all the BARs, not just the first. 144578 144579commit 82d51e6df2bf677bdf24376092bcaa79b534f6e5 144580Author: Adam Jackson <ajax@redhat.com> 144581Date: Fri Aug 1 13:17:55 2008 -0400 144582 144583 Silence the "No matching Device section" warning in some harmless cases. 144584 144585 If the device doesn't have any BARs then it's just a stub for some 144586 lame operating systems that need one PCI device per output for 144587 multihead. No point in warning about it. 144588 144589commit 9643e8d3482a35d355a243db7aa397ad47f29be0 144590Author: Adam Jackson <ajax@redhat.com> 144591Date: Fri Aug 1 11:35:47 2008 -0400 144592 144593 Handle XGI cards in autoconfig. 144594 144595 It's all a bit wonky since both sis(4) and xgi(4) claim to support the 144596 Volari Z7 and V5/8 (0x0020 and 0x0040), so let's side with xgi(4), why 144597 not. Note that the V3 (not V3XT) identifies itself as a trident chip. 144598 144599commit 63bdd4c27d47323b3282bf0b6eaecae91c79b45c 144600Author: Adam Jackson <ajax@redhat.com> 144601Date: Thu Jul 31 17:31:36 2008 -0400 144602 144603 Unifdef XIDLE. 144604 144605 Seriously how was this still here. 144606 144607commit 8d214bc26f9b7ab6a5c54d7749cd4b6811cb0b96 144608Author: Adam Jackson <ajax@redhat.com> 144609Date: Thu Jul 31 15:46:52 2008 -0400 144610 144611 Document more of the OS and library assumptions. 144612 144613commit 2198e237b2ed85857c671eee2cd04dfc032befee 144614Author: Tomas Carnecky <tom@dbservice.com> 144615Date: Thu Jul 31 13:46:06 2008 -0400 144616 144617 Cosmetic cleanup to Xvfb/Xdmx configure check output. 144618 144619commit 148175fb8b365dcf00b13539b3f03ce33f3df707 144620Author: Michel Dänzer <michel@tungstengraphics.com> 144621Date: Thu Jul 31 16:54:33 2008 +0200 144622 144623 EXA: Do still return FALSE if the driver PrepareCopy hook failed... 144624 144625 Thanks to Stuart Bennett for pointing out the problem on IRC. 144626 144627commit 68fd6604a15b653d1fa244633eaaee9608a03ab3 144628Author: Tomas Carnecky <tom@dbservice.com> 144629Date: Thu Jul 31 10:03:56 2008 -0400 144630 144631 OutputPropertyNotifyMask is a valid thing to select for. 144632 144633commit 64ebeeb5265a4c425b9397fdc86a6d81521a856e 144634Author: Michel Dänzer <michel@tungstengraphics.com> 144635Date: Thu Jul 31 10:58:52 2008 +0200 144636 144637 EXA: Fix exponential growth logic for GXcopy tiled fills. 144638 144639 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=16908 . 144640 144641commit 8405c25d9ddbfddb6b155a436f07ccad689e53bd 144642Author: Michel Dänzer <michel@tungstengraphics.com> 144643Date: Thu Jul 31 10:55:44 2008 +0200 144644 144645 EXA: Simplify exaFillRegionTiled() control flow. 144646 144647 Also only call REGION_TRANSLATE() when necessary. 144648 144649commit b37b1e66996f8335dafc97b12d25aaec452b931e 144650Author: Aaron Plattner <aplattner@nvidia.com> 144651Date: Wed Jul 30 18:48:27 2008 -0700 144652 144653 Make shmint.h part of the SDK. 144654 144655 This includes ShmRegisterFuncs, ShmSetPixmapFormat, fbShmPutImage, and 144656 ShmRegisterFbFuncs. Note that fbShmPutImage was already exported. 144657 144658commit a3afa6f2fb80489f7b6a88d12def09281d32ed94 144659Author: Michel Dänzer <michel@tungstengraphics.com> 144660Date: Wed Jul 30 18:30:37 2008 +0200 144661 144662 EXA: Optimize GXcopy tiled fills. 144663 144664commit 37087bc10630ee7740df1369b3e56a44fd2ad2b0 144665Author: Michel Dänzer <michel@tungstengraphics.com> 144666Date: Wed Jul 30 18:27:33 2008 +0200 144667 144668 EXA: Replace open coded CopyArea logic with GC op call. 144669 144670commit bb2cacd33eb85f39b9e7fa554545611957df677d 144671Author: Xavier Bestel <xavier.bestel@free.fr> 144672Date: Tue Jul 29 13:06:10 2008 -0400 144673 144674 Cosmetic fix to EDID decoding. 144675 144676commit a70754a9fdb019d85fbfca1898699f3c6f609fdd 144677Author: Alan Hourihane <alanh@tungstengraphics.com> 144678Date: Tue Jul 29 14:16:25 2008 +0100 144679 144680 require inputproto 1.9.99.3 144681 144682commit 591ef3c047ab3597fef9d687205e99c254ff2040 144683Author: Peter Hutterer <peter.hutterer@who-t.net> 144684Date: Tue Jul 29 10:00:01 2008 +0930 144685 144686 Xi: ChangeDeviceControl presence events should set the appropriate devchange. 144687 144688 Requires inputproto 1.4.4 or higher. 144689 144690commit 25dd5ce0770c5ff91a79c12223fd4af52759900b 144691Author: Kristian Høgsberg <krh@redhat.com> 144692Date: Mon Jul 28 12:33:58 2008 -0400 144693 144694 Add atKeynames.h in hw/dmx/input and drop evil cross-ddx #include. 144695 144696commit e39a16aa58456aaecee54edc64f189c958db0902 144697Author: Julien Cristau <jcristau@debian.org> 144698Date: Sat Jul 26 16:35:19 2008 +0200 144699 144700 xfree86: don't output a spurious newline to the log when loading a module 144701 144702commit 2eaed4a10fe5bf727579bca4ab8d4a47c8763a7d 144703Author: Julien Cristau <jcristau@debian.org> 144704Date: Sat Jul 26 15:35:42 2008 +0200 144705 144706 xfree86: use xorg.conf input devices if there is no ServerLayout 144707 144708 If xorg.conf has no ServerLayout section, use the first mouse and 144709 keyboard sections as core devices, even with AllowEmptyInput. 144710 144711commit 805f28e96ceb20bc53792ae3cf17f9c26564ae0e 144712Author: Simon Thum <simon.thum@gmx.de> 144713Date: Wed Jul 23 12:10:52 2008 +0200 144714 144715 some trivial changes regarding C compat, indentation, etc. 144716 144717 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144718 144719commit 3209bd21d00b8673d321f70afb65720588ddacb3 144720Author: Simon Thum <simon.thum@gmx.de> 144721Date: Wed Jul 23 12:06:34 2008 +0200 144722 144723 xfree86: perform pointer accel scheme selection by name, not number 144724 144725 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144726 144727commit a17cb29f5acdfcdeac929d8c6be3600d44d038ef 144728Author: Simon Thum <simon.thum@gmx.de> 144729Date: Wed Jul 23 12:03:17 2008 +0200 144730 144731 dix: preparation change: make runtime exchange of filter chain safe 144732 144733 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144734 144735commit 1a9f9ac50f2b0db735789905cc29572a50c8ae4c 144736Author: Simon Thum <simon.thum@gmx.de> 144737Date: Wed Jul 23 11:49:36 2008 +0200 144738 144739 dix: optimize precision in device velocity estimation 144740 144741 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144742 144743commit c184b91d9aa72031c2bac9f379f56633957ded30 144744Author: Simon Thum <simon.thum@gmx.de> 144745Date: Wed Jul 23 11:33:25 2008 +0200 144746 144747 dix: improve the driver interface to predictable pointer acceleration 144748 144749 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144750 144751commit 87aa5298576ed335ac31347e14fb30430288157a 144752Author: Simon Thum <simon.thum@gmx.de> 144753Date: Wed Jul 23 11:28:09 2008 +0200 144754 144755 dix: introduce defines for accel profile numbers 144756 144757 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144758 144759commit 4e32e6fb38d19c9993de86188e4f7e7916a028e2 144760Author: Simon Thum <simon.thum@gmx.de> 144761Date: Wed Jul 23 11:10:22 2008 +0200 144762 144763 dix: rename classic accel _scheme_ to lightweight to avoid confusion with 144764 classic accel _profile_ 144765 144766 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144767 144768commit d40183c665d458ac0a6e7952fbe986776a17fda7 144769Author: Simon Thum <simon.thum@gmx.de> 144770Date: Wed Jul 23 09:28:06 2008 +0200 144771 144772 dix: add legal statements to ptrvelo.{c|h} 144773 144774 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 144775 144776commit ec10eccd56be8b947cd63cae0687b8319857fe60 144777Author: Michel Dänzer <michel@tungstengraphics.com> 144778Date: Mon Jul 28 09:33:04 2008 +0200 144779 144780 GLX: Unreference drawables bound to the old context, not the new one. 144781 144782 Apart from the obvious reference counting issue, this fixes 144783 http://bugs.freedesktop.org/show_bug.cgi?id=16867 . 144784 144785commit 6ab8d6010adfd5ad6f1e1094a26c84f0aff934b1 144786Author: Michel Dänzer <michel@tungstengraphics.com> 144787Date: Mon Jul 28 09:32:59 2008 +0200 144788 144789 AIGLX/DRI1: Pay more attention to return value from DRIGetDrawableInfo(). 144790 144791 Could have crashed otherwise if the num(Back)ClipRects variables referenced by 144792 the caller weren't pre-initialized to 0. 144793 144794commit b3e981c9d4ff31263a43b47f83cf8db4c2b5aeff 144795Author: Michel Dänzer <michel@tungstengraphics.com> 144796Date: Mon Jul 28 09:32:54 2008 +0200 144797 144798 Fix Makefile.am crack. 144799 144800 Looks like an artifact from early modularization. 144801 144802commit 3575d9584edf35ec1720bc3755b6576a56613685 144803Author: Peter Hutterer <peter.hutterer@who-t.net> 144804Date: Fri Jul 25 16:46:44 2008 +0930 144805 144806 config: note that HAL options must be strings. 144807 144808 Only strings are parsed by the server, all others are ignored. Doesn't matter, 144809 specifying int options as strings works fine anyway. 144810 144811commit 54651ff5ec54f4c621e060b19f31d77d519ef158 144812Author: Peter Hutterer <peter.hutterer@who-t.net> 144813Date: Thu Jul 24 17:41:48 2008 +0930 144814 144815 xkb: remove superfluous inputInfo.keyboard treatment. 144816 144817 Really not necessary, we can just walk the list and spare us the special 144818 treatment of the VCK. 144819 144820commit 70bd8261223366dae9dfdbb999691328d85f093c 144821Author: Peter Hutterer <peter.hutterer@who-t.net> 144822Date: Thu Jul 24 17:38:16 2008 +0930 144823 144824 xkb: use PickPointer/PickKeyboard in _XkbLookupAnyDevice. 144825 144826commit 47160edec7f0d9129576d83f1593a5549879a893 144827Author: Peter Hutterer <peter.hutterer@who-t.net> 144828Date: Thu Jul 24 12:12:45 2008 +0930 144829 144830 xfree86: warn some more about potential missing input devices. 144831 144832 Put out a warning if xorg.conf has InputDevice sections, but these aren't 144833 referenced in the used ServerLayout. This is only performed if AllowEmptyInput 144834 is enabled. 144835 The reason behind this is that the server used to auto-add the first 144836 mouse/keyboard sections if none where referenced. Now, with HAL and AEI 144837 enabled by default, setups that relied on this auto-adding break and are left 144838 without input devices. The least we can do is warn them. 144839 144840commit f30b0823dbfc5902e54b337b5b6b570ebf216584 144841Author: Peter Hutterer <peter.hutterer@who-t.net> 144842Date: Thu Jul 24 09:56:00 2008 +0930 144843 144844 xfree86: if AllowEmptyInput is on, warn the user that we rely on HAL now. 144845 144846commit ad4cd2e241691427689591f7769a1184c8c1c7f5 144847Author: Peter Hutterer <peter.hutterer@who-t.net> 144848Date: Wed Jul 23 17:40:28 2008 +0930 144849 144850 xkb: don't send core events on SlowKeys. 144851 144852 Core events don't happen until later in the DIX, so pump device events down 144853 instead. This makes modifiers work again when SlowKeys is enabled. 144854 144855commit 7d9dece74fc2bf130ceb8818ced5d9e3ac526900 144856Author: Peter Hutterer <peter.hutterer@who-t.net> 144857Date: Mon Jul 21 22:12:45 2008 +0930 144858 144859 dix: fix up enter/leave system once again. 144860 144861 Two corrections 144862 1. the "detail" field has NotifyVirtual, etc., not the "mode" field. This was 144863 a clear bug. 144864 2. don't set/unset the flags for NotifyGrab or NotifyUngrab. Clients are 144865 expected to deal with multiple enter/leave events per window if the mode is 144866 not NotifyNormal. 144867 144868 Testable with TCL menu boxes (such as used in gitk): 144869 tk_optionMenu .menu globVar Val1 Val2 Val3 ValJunk 144870 pack .menu 144871 144872 Thanks to Michel Dänzer for pointing this out. 144873 144874commit 2ce434f54bc0d6050ef115e310df62c035e84bf2 144875Author: Kristian Høgsberg <krh@redhat.com> 144876Date: Thu Jul 24 18:44:16 2008 -0400 144877 144878 Clean up unused definitions from glx headers. 144879 144880commit ae67508392261ae47858692668a0c192ef4a9d7b 144881Author: Mathieu Bérard <mathieu.berard@crans.org> 144882Date: Fri Jul 25 10:05:30 2008 -0400 144883 144884 Fix AddExtension now that CloseDownProc can be NULL. 144885 144886commit eff25430b4a391409e39337962ff7697165d23c7 144887Author: Daniel Drake <dsd@laptop.org> 144888Date: Thu Jul 24 21:06:34 2008 -0500 144889 144890 Don't abort if swrast library is not present 144891 144892 GLX is enabled by default, but the current swrast behaviour causes X 144893 to abort with fatal error if the swrast dri library dlopen fails. 144894 144895 Handle the case where the swrast library is not present, and do not 144896 register the GLX extension unless at least one screen has a usable 144897 GL provider. 144898 144899commit 62b5690415786c4a6e8bac464aa8db578beb74f5 144900Author: Adam Jackson <ajax@redhat.com> 144901Date: Thu Jul 24 16:13:10 2008 -0400 144902 144903 Remove some more MEMBUG garbage. 144904 144905commit 036822584b26854e2f486f1bea84d8d19d9324de 144906Author: Kristian Høgsberg <krh@redhat.com> 144907Date: Thu Jul 24 16:03:58 2008 -0400 144908 144909 Drop xf86CommonSpecialKey() and atKeynames.h 144910 144911 These are AT scancode specific, should really be part of xf86-input-keyboard. 144912 Remove from server, move to xf86-input-keyboard. 144913 144914commit f3ff2386016ced7e677817e9761a535f0f385813 144915Author: Adam Jackson <ajax@redhat.com> 144916Date: Thu Jul 24 15:28:07 2008 -0400 144917 144918 Sync is built-in and mandatory. 144919 144920commit 3fcf4d3eb89fecaa2be7b5ac4933b693c9c3d97e 144921Author: Adam Jackson <ajax@redhat.com> 144922Date: Thu Jul 24 15:09:05 2008 -0400 144923 144924 XC-MISC is built-in and mandatory. 144925 144926commit 9757106bba8f7bea99c5211817fc6b5fde4e6f66 144927Author: Adam Jackson <ajax@redhat.com> 144928Date: Thu Jul 24 14:59:14 2008 -0400 144929 144930 Remove all empty extension reset hooks, replace with NULL. 144931 144932commit 6bcde69585fcc8f8dbfe81c115649f19274922fa 144933Author: Adam Jackson <ajax@redhat.com> 144934Date: Thu Jul 24 09:13:26 2008 -0400 144935 144936 Allow extension closedown hook to be null. 144937 144938commit f65c50c4e902c7b5619fb7bbee8462434d809ec7 144939Author: Adam Jackson <ajax@redhat.com> 144940Date: Thu Jul 24 09:07:42 2008 -0400 144941 144942 Bigreqs are built-in and mandatory. 144943 144944commit d6228cb22aa89b90834d80d98b91862c1fc01b54 144945Author: Adam Jackson <ajax@redhat.com> 144946Date: Thu Jul 24 09:00:22 2008 -0400 144947 144948 Shape extension is built-in and mandatory. 144949 144950commit 990fc643ae90c034187707e7de414d80640ec6da 144951Author: Kristian Høgsberg <krh@redhat.com> 144952Date: Thu Jul 24 15:02:40 2008 -0400 144953 144954 Ugh, remove trailing backslash. 144955 144956commit 806e8cf3c01f955411445a7095f0957ea9a9b22a 144957Author: Kristian Høgsberg <krh@redhat.com> 144958Date: Thu Jul 24 14:25:24 2008 -0400 144959 144960 Remove SPARC muldiv code. Was used by ELF loader, which is no more. 144961 144962commit dff1a609bb4c0171e2abb92b54c16e739aec9ca2 144963Author: Kristian Høgsberg <krh@redhat.com> 144964Date: Mon Jul 21 18:29:08 2008 -0400 144965 144966 Drop the glx resize hook and stop chaining PositionWindow. 144967 144968commit 24dddcd0ef845f4120f8588dc63ec754338ffac8 144969Author: Kristian Høgsberg <krh@redhat.com> 144970Date: Mon Jul 21 18:16:38 2008 -0400 144971 144972 Drop unnecessary linked list of contexts from GLXDrawable. 144973 144974commit 5c1e254cc85e9ad409b0217780545c29f62d5feb 144975Author: Daniel Stone <daniel@fooishbar.org> 144976Date: Thu Jul 24 03:01:45 2008 +0300 144977 144978 Remove xorgcfg 144979 144980 Us shipping a GUI configuration utility (especially as part of the 144981 server!) was pretty pointless. There was pretty much nothing it could 144982 configure which wasn't already runtime adjustable: if you could get a 144983 server up with functioning input and output, there wasn't much xorgcfg 144984 could do for you. 144985 144986 Au revoir. 144987 144988commit b74927c3844bc2650d95f604fe782d95ade067f1 144989Author: Daniel Stone <daniel@fooishbar.org> 144990Date: Thu Jul 24 02:42:13 2008 +0300 144991 144992 Remove ioport 144993 144994 I don't know why we would possibly be shipping this. 144995 144996commit 8c0518379089d230060e9ff672ba5eba34198325 144997Author: Daniel Stone <daniel@fooishbar.org> 144998Date: Thu Jul 24 02:40:02 2008 +0300 144999 145000 Remove kbd_mode 145001 145002 We only built this on BSD and Solaris, and if such a tool is generally 145003 useful, ship it with the OS. 145004 145005commit 0d785bd635d135fcd67b4c9c88f5c8217e9b9240 145006Author: Adam Jackson <ajax@redhat.com> 145007Date: Wed Jul 23 13:43:09 2008 -0400 145008 145009 Unifdef hpux. 145010 145011commit 5035741fd4cb4918957064a8c9ded9c7c31b4e22 145012Author: Adam Jackson <ajax@redhat.com> 145013Date: Wed Jul 23 13:38:38 2008 -0400 145014 145015 Unifdef __osf__ 145016 145017commit bd8bd2c700e6ccac53b47c45766c0a3a08a59344 145018Author: Adam Jackson <ajax@redhat.com> 145019Date: Wed Jul 23 13:36:35 2008 -0400 145020 145021 Unifdef AIX. 145022 145023commit 5f5eed7f790db2f0ae0cd0f5c5ee5a312079c0b0 145024Author: Adam Jackson <ajax@redhat.com> 145025Date: Wed Jul 23 13:27:24 2008 -0400 145026 145027 Unifdef DGUX. 145028 145029commit 856db05b58b71c5ff22af727aba435a8c356abfa 145030Author: Adam Jackson <ajax@redhat.com> 145031Date: Wed Jul 23 13:25:37 2008 -0400 145032 145033 Unifdef sgi. 145034 145035commit ab6557fcd4bb4fd580bb1727dc8764f7dfb2cb30 145036Author: James Cloos <cloos@jhcloos.com> 145037Date: Wed Jul 23 13:17:33 2008 -0400 145038 145039 Optimize commit 331cc3f0799a54910a99484264f76569beeee55a 145040 145041 Reorder to minimize comparisons 145042 145043commit b699364b1c807d29666633523f5b5a608a07a8ee 145044Author: Adam Jackson <ajax@redhat.com> 145045Date: Wed Jul 23 11:43:59 2008 -0400 145046 145047 Unifdef VXWORKS. 145048 145049commit 2e3c43d306616f8accdeaffec9c108ba88324805 145050Author: Adam Jackson <ajax@redhat.com> 145051Date: Wed Jul 23 11:35:06 2008 -0400 145052 145053 Assume __GNUC__ >= 3. 145054 145055 gcc 3.0 was seven years ago. Get with the times. 145056 145057commit 0c1c9d75089aac0cf673820db83801d3cab5bfc7 145058Author: Adam Jackson <ajax@redhat.com> 145059Date: Wed Jul 23 09:11:40 2008 -0400 145060 145061 Remove some broken debugging scaffolding. 145062 145063commit 0ff5bc404842f7264a8e77f9e2a0bd8d0b051356 145064Author: Adam Jackson <ajax@redhat.com> 145065Date: Wed Jul 23 09:09:11 2008 -0400 145066 145067 Unifdef QNX. 145068 145069 Again, hasn't worked since at least 7.0. 145070 145071commit 331cc3f0799a54910a99484264f76569beeee55a 145072Author: James Cloos <cloos@jhcloos.com> 145073Date: Wed Jul 23 00:01:43 2008 -0400 145074 145075 Fix LookupColor 145076 145077 Using strncasecmp(3) with the lenght of the user-supplied colour name 145078 will result in a false positive when the db key starts out with the 145079 same string. 145080 145081 Eg, blue will also match BlueViolet (aka blue violet). 145082 145083 Since the shorter strings occur first in the database, avoid such 145084 errors by treating a 0 result from strncasecmp(3) as a positive result 145085 when the key’s length is longer than the supplied string’s. 145086 145087commit bc3c03a3f3c091026310f0e8d55321cec570a0c5 145088Author: Pierre-Loup A. Griffais <pgriffais@nvidia.com> 145089Date: Tue Jul 22 17:34:37 2008 -0700 145090 145091 Don't return BadAlloc when trying to set a PictureFilter with no parameters when a filter with parameters was previously set. 145092 145093 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 145094 145095commit 92fdd01d8e46dd4b41338754e1a34e9d3c387d4a 145096Author: Dodji Seketeli <dodji@seketeli.org> 145097Date: Wed Jul 23 01:07:29 2008 +0200 145098 145099 [Xephyr] various X86_64 build cleanups 145100 145101commit 5de1867fbb0a336ff3fdc92cbf734849f6111b1b 145102Author: Dodji Seketeli <dodji@seketeli.org> 145103Date: Wed Jul 23 01:00:26 2008 +0200 145104 145105 [Xephyr] Fix #15839 145106 145107 Make sure the _XSERVER64 macro is not defined in Xlib client code. 145108 That macro is meant to be define only on pure server code, when necessary. 145109 145110commit d692e1a63e4718d3b6d486582c19f30259b813a7 145111Author: Adam Jackson <ajax@redhat.com> 145112Date: Tue Jul 22 10:21:22 2008 -0400 145113 145114 Build fix. 145115 145116commit d8937ce58137ad675f1ead8f2345eaed0ca63c8a 145117Author: Daniel Stone <daniel@fooishbar.org> 145118Date: Tue Jul 22 16:52:50 2008 +0300 145119 145120 autogen.sh: Pass --force to autoreconf 145121 145122 Ensure we get all files installed when running autoreconf by passing 145123 --force to overwrite previously-generated ones, which fixes running it 145124 from tarballs. 145125 145126 Also revert Peter's local changes to pass arguments to configure, all of 145127 which are the default now anyway, AIUI. 145128 145129commit 880625eef5d8b168df3e42836fa1b763c51a91b5 145130Author: Peter Hutterer <peter.hutterer@who-t.net> 145131Date: Tue Jul 22 14:34:28 2008 +0930 145132 145133 xfree86: plug memory leak, free driver's private data when deleting the device. 145134 145135commit 67d7821ae783d3f123b6ba7203abf847374a1e36 145136Author: Keith Packard <keithp@keithp.com> 145137Date: Mon Jul 21 11:48:24 2008 -0700 145138 145139 dix: reset potential lastSlaves when disabling an SD 145140 145141 Unplug a mouse, then warp the pointer and the warp pointer code will try 145142 to update the position of the last slave device associated with the 145143 master. That pointer will be stale and the X server will crash. 145144 145145 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 145146 145147commit 8c8c4fdf34bfc9d54ebea99fb0af14cad167b4a0 145148Author: Adam Jackson <ajax@redhat.com> 145149Date: Mon Jul 21 16:39:43 2008 -0400 145150 145151 EDID: Various reduced blanking fixes. 145152 145153 - Use a single common function to compute reducedness. 145154 - Call it from both the old-school and new-school mode validation paths. 145155 - Define monitor reduced-blanking support in accord with EDID 1.4. 145156 - Attempt to filter RB DMT modes away from the "standard" EDID pool if 145157 the monitor doesn't claim RB support. 145158 145159commit e8cd77e14d3fa40e5cf1174acaf925362b2e0a11 145160Author: Adam Jackson <ajax@redhat.com> 145161Date: Mon Jul 21 16:15:03 2008 -0400 145162 145163 EDID: For standard timing blocks, prefer DMT timings if they exist. 145164 145165commit 6ba70091e2325f534a37ef185fea568d2c43edec 145166Author: Adam Jackson <ajax@redhat.com> 145167Date: Mon Jul 21 15:40:34 2008 -0400 145168 145169 EDID: Correct DDCEstablishedModes to conform to DMT. 145170 145171commit facb255fa9267e343cbc91f841f1b64e5dc99e98 145172Author: Kristian Høgsberg <krh@redhat.com> 145173Date: Mon Jul 21 16:05:53 2008 -0400 145174 145175 Need to unref pixmaps backing pbuffers too. 145176 145177commit d5ae85b5b722821499d5796cf0973ecb6ec125f1 145178Author: Kristian Høgsberg <krh@redhat.com> 145179Date: Mon Jul 21 15:28:50 2008 -0400 145180 145181 Fix embarrasing GLXPixmap leak. 145182 145183commit c74ddc87c995c73109827717a49f14846c7c4024 145184Author: Adam Jackson <ajax@redhat.com> 145185Date: Mon Jul 21 15:10:20 2008 -0400 145186 145187 EDID: Add quirk to clamp max pixel clock to single DVI link speed. 145188 145189 On some panels you end up with all of: 145190 - No range descriptor 145191 - No description of physical connectivity 145192 - Native panel size mode in standard timings list 145193 145194 In principle you're supposed to use the timings for that mode from the DMT 145195 spec, but in practice the DMT spec has timings for both 1920x1200 normal 145196 and 1920x1200RB, and the standard timing field gives you no way to 145197 distinguish. And, of course, the non-RB timings don't fit in a single 145198 DVI link. 145199 145200commit 55803473adb0a0975fea81035402bf6b4ec0e30b 145201Author: Alan Hourihane <alanh@tungstengraphics.com> 145202Date: Mon Jul 21 10:32:57 2008 +0100 145203 145204 Bump inputproto to 1.9.99.2 for XI_PROP_ENABLED 145205 145206commit 69de40ee45a6e046be79b735cd1540f63b87aee9 145207Author: Peter Hutterer <peter.hutterer@who-t.net> 145208Date: Mon Jul 21 17:05:04 2008 +0930 145209 145210 dix: shift the duplicate button mapping check to ProcSetPointerMapping. 145211 145212 XI 2 allows two buttons to have the same button code. 145213 145214commit 0dbfe0ebc69c307c0626ba824de15d03de1251d4 145215Author: Daniel Stone <daniel@fooishbar.org> 145216Date: Mon Jul 21 03:50:15 2008 +0300 145217 145218 Revert "Try nvidia before nv under the assumption that people who install it usually want to *use* it, too." 145219 145220 Per mailing list discussion. 145221 145222 This reverts commit 66fb253082ea42179180303393e48846208987fa. 145223 145224commit 6b5206e7cb8e5279816b48f014d47d3f03f16972 145225Author: Daniel Stone <daniel@fooishbar.org> 145226Date: Fri Jul 18 15:38:10 2008 +0300 145227 145228 dix: Remove insane BC hacks 145229 145230 If your DDX needs a 1bpp mode and doesn't set it up, your DDX is 145231 incompetent. 145232 145233commit 66fb253082ea42179180303393e48846208987fa 145234Author: Aaron Plattner <aplattner@nvidia.com> 145235Date: Tue Jul 8 02:39:00 2008 -0700 145236 145237 Try nvidia before nv under the assumption that people who install it usually want to *use* it, too. 145238 145239commit 376620460ec50cb98a1e3746629b7f73c2e926d7 145240Author: Julien Cristau <jcristau@debian.org> 145241Date: Sun Jul 20 16:53:01 2008 +0200 145242 145243 Drop some more configure-generated files from the tarball 145244 145245commit 35346a42271dfe65fe0a86dc522701028bf6daf6 145246Author: Julien Cristau <jcristau@debian.org> 145247Date: Sun Jul 20 16:40:59 2008 +0200 145248 145249 Don't put xf86Build.h in the tarball (bug#9277) 145250 145251 This file is generated by configure, we don't need to ship a stale 145252 version in the tarball. 145253 X.Org bug#9277 <https://bugs.freedesktop.org/show_bug.cgi?id=9277> 145254 145255commit 43c6d5a6f83dfdf8596c8d0faf8ac1a64f73a8c1 145256Author: Julien Cristau <jcristau@debian.org> 145257Date: Sun Jul 20 16:30:24 2008 +0200 145258 145259 distcheck fixes 145260 145261 Still seems to fail because hw/xquartz has too long filenames 145262 145263commit 92b0a04ffd630a89705fc73d587182a62c34cd84 145264Author: James Cloos <cloos@jhcloos.com> 145265Date: Sat Jul 19 18:21:37 2008 -0400 145266 145267 Fix compile failure 145268 145269 This fixes a compile failure by ensuring that mouse-cfg.c 145270 can see the MTYPE_ constants. 145271 145272commit 0b45ba488d8a3e901b83e445d17e7e9a67047b6c 145273Author: Adam Jackson <ajax@redhat.com> 145274Date: Fri Jul 18 10:54:10 2008 -0400 145275 145276 Drop the FONTCACHE extension. 145277 145278 This appears to have been dead code since X-TT was merged. 145279 145280commit b85392e66b4cbfcc95897ac342f19ac8ab04d4cc 145281Author: Søren Sandmann <sandmann@redhat.com> 145282Date: Thu Jul 17 23:10:11 2008 -0400 145283 145284 Give priority to clients that are receiving expose and configure events 145285 145286commit 3b687ffe1649449b3d182f5e7690274c6c96916a 145287Author: Alan Coopersmith <alan.coopersmith@sun.com> 145288Date: Thu Jul 17 18:16:59 2008 -0700 145289 145290 Make xstrcasestr prototype return value match the implementation 145291 145292commit f7bfa4ae971b43c3ed81cc84e6d1e071c413f04c 145293Author: Daniel Stone <daniel@fooishbar.org> 145294Date: Thu Jul 17 23:03:57 2008 +0300 145295 145296 configure.ac: Remove non-existant directory 145297 145298 Oops. 145299 145300commit 16e40ecf43ea7df349e54f757db0223764d752ee 145301Author: Daniel Stone <daniel@fooishbar.org> 145302Date: Thu Jul 17 22:37:31 2008 +0300 145303 145304 Remove dead code, useless #defines, et al 145305 145306commit b0a99fb02bb7290266d7dddee4de0b9959fb3c52 145307Author: Daniel Stone <daniel@fooishbar.org> 145308Date: Thu Jul 17 21:42:17 2008 +0300 145309 145310 configure.ac: Fix release date 145311 145312 Instead of putting something that's always wrong, at least just put 145313 unreleased, which will be more obviously wrong if it's, well, in a 145314 release. 145315 145316commit 711720650cc192022f0d91f5cf94292d48dbc891 145317Author: Daniel Stone <daniel@fooishbar.org> 145318Date: Thu Jul 17 21:39:46 2008 +0300 145319 145320 Everyone has urandom 145321 145322 If you don't have urandom, please just add a fallback to /dev/random, 145323 rather than building our own random generator. 145324 145325commit 446fe9eecddd1337f9d5164dd7c301e1ba3dfe32 145326Author: Daniel Stone <daniel@fooishbar.org> 145327Date: Thu Jul 17 21:37:50 2008 +0300 145328 145329 Dead code removal 145330 145331 Remove a whole bunch of code that was never built, be it entire files or 145332 just dead ifdefs. 145333 145334commit 0564b5454ac101d9e1218767bbbc2c2d9f3e0696 145335Author: Daniel Stone <daniel@fooishbar.org> 145336Date: Thu Jul 17 20:46:27 2008 +0300 145337 145338 XFree86: Restore AllowMouseOpenFail usage text 145339 145340 This was incorrectly stuck behind the XFree86-Misc define, so got 145341 deleted with the rest of misc: restore it unconditionally, as it should 145342 be. 145343 145344commit 6c7c0ffeb71610ee00fc92b32fd2b1cc70383e38 145345Author: Daniel Stone <daniel@fooishbar.org> 145346Date: Thu Jul 17 20:45:01 2008 +0300 145347 145348 KDrive: Delete rafts of unused code 145349 145350 Most of this stuff was gone with the last real hardware drivers that 145351 used it. Au revoir. 145352 145353commit ad7a7ae290b05f007062161418c929d9cebeef78 145354Author: Julien Cristau <jcristau@debian.org> 145355Date: Thu Jul 17 18:51:05 2008 +0200 145356 145357 XF86BigFont is disabled by default 145358 145359 3c19ec47b434d4ca84db58363cc053cc0b6aa413 did that, 145360 but pretended to make it 'auto'. 145361 145362commit 058100fd19ed33a3100d18a9ca56fdcffac361b2 145363Author: Alan Coopersmith <alan.coopersmith@sun.com> 145364Date: Thu Jul 17 09:35:30 2008 -0700 145365 145366 Remove more Lynx leftovers 145367 145368 A couple #if defined(Lynx) && defined(sun) had become just if defined(sun), 145369 resulting in wrong settings for Solaris builds, so they're now just deleted. 145370 145371commit 47833eef35b20a2eb0a74782edbb0877fd491c4c 145372Author: Mathieu Bérard <mathieu.berard@crans.org> 145373Date: Thu Jul 17 11:59:24 2008 -0400 145374 145375 Drop a bunch of #ifdef Lynx. 145376 145377commit af2138bfce34a28312b26106eaf9561d96712ee0 145378Author: Adam Jackson <ajax@redhat.com> 145379Date: Thu Jul 17 10:39:29 2008 -0400 145380 145381 Nuke some ancient code for commandline-challenged OSes. 145382 145383 This appears to have been dead code even in 6.7. 145384 145385commit b9a0cee026361189dce63b87c738bd1d17e02830 145386Author: Adam Jackson <ajax@redhat.com> 145387Date: Thu Jul 17 10:32:35 2008 -0400 145388 145389 Drop hw/xfree86/os-support/lynxos/ 145390 145391 This has never worked in any modular release, clearly no one cares. 145392 145393commit d87f170bd99f7908eb14272ac42c8963d089f54f 145394Author: Adam Jackson <ajax@redhat.com> 145395Date: Thu Jul 17 10:29:52 2008 -0400 145396 145397 Remove XF86Misc code from xorgcfg. 145398 145399commit e3ff8501749209ee2dd324d35ae16e37510e0574 145400Author: Jie Luo <clotho67@gmail.com> 145401Date: Thu Jul 17 10:27:24 2008 -0400 145402 145403 Remove duplicate load of freetype font renderer. 145404 145405 libXfont will do this for us in FontFileRegisterFpeFunctions. 145406 145407commit 8441e26266701776f1416bec864ae2949355da76 145408Author: Jie Luo <clotho67@gmail.com> 145409Date: Thu Jul 17 10:23:37 2008 -0400 145410 145411 Remove freetype from the list of autoloaded modules. 145412 145413commit 5d11ae7a69e3a7611c667be4628a15a4b514f72c 145414Author: Jie Luo <clotho67@gmail.com> 145415Date: Thu Jul 17 10:22:29 2008 -0400 145416 145417 Fix color lookup. 145418 145419 OsStrCaseCmp returned inverted comparison sense from normal strcasecmp. 145420 145421commit 478d3918d5b23fba1a7e3aaea766dfa785e5a0a6 145422Author: Daniel Stone <daniel@fooishbar.org> 145423Date: Thu Jul 17 03:29:26 2008 +0300 145424 145425 XFree86: Remove remnants of XFree86-Misc 145426 145427 HandleMessage: dear god, no. 145428 145429commit 95000e34df9ba066905430eb0eb39a616567183e 145430Author: Adam Jackson <ajax@redhat.com> 145431Date: Wed Jul 16 16:37:14 2008 -0400 145432 145433 Build fixes 145434 145435commit affec10635343668839994ea2bac16c1d4524200 145436Author: Adam Jackson <ajax@redhat.com> 145437Date: Wed Jul 16 14:44:33 2008 -0400 145438 145439 Remove loadable font renderer support. 145440 145441commit acc91054316208844742495dea0b80a860ef02bb 145442Author: Adam Jackson <ajax@redhat.com> 145443Date: Wed Jul 16 14:24:47 2008 -0400 145444 145445 Remove font module support from xorgcfg. 145446 145447commit a82e6efb7b9b2ab9a1597b002f375c5ee105e7f5 145448Author: Adam Jackson <ajax@redhat.com> 145449Date: Wed Jul 16 14:19:04 2008 -0400 145450 145451 Remove the numVideoScreens xprintism. 145452 145453 This was to account for cases where you had video and print screens in 145454 the same server. Lunacy. Leave the slot in ScreenInfo, but rename it, 145455 and stop looking at it. 145456 145457commit 1a573e402ec112913a404f092b5b97d8d9210f94 145458Author: Adam Jackson <ajax@redhat.com> 145459Date: Wed Jul 16 13:41:53 2008 -0400 145460 145461 Update the Allow*Grabs documentation for xf86misc removal. 145462 145463 The API to turn the grab-break keys back off is gone now, so don't 145464 say it exists. 145465 145466commit 55644d2f1cddf1777587f213ed581b6087d9fdad 145467Author: Daniel Stone <daniel@fooishbar.org> 145468Date: Wed Jul 16 14:23:34 2008 +0300 145469 145470 Make --enable-debug usable again 145471 145472 Enable region debugging causes assertion failures (should really look 145473 into that, though), and I don't need four messages every time I move my 145474 pointer. 145475 145476commit 22e64108ec63ba77779891f8df237913ef9ca731 145477Author: Daniel Stone <daniel@fooishbar.org> 145478Date: Wed Jul 16 06:25:26 2008 +0300 145479 145480 XFree86: Remove XFree86-Misc extension 145481 145482 Its last remaining purpose in life has been destroyed by input 145483 properties. Au revoir: it's been fun, by which I mean awful. 145484 145485commit b8dd07f855c555af56cbf0f69df799f424da2cca 145486Author: Daniel Stone <daniel@fooishbar.org> 145487Date: Wed Jul 16 03:00:25 2008 +0300 145488 145489 HAL: Remove grotesque open-coded strcasestr 145490 145491 Not only was this pretty ugly, but it didn't even work on systems 145492 without strcasestr anyway, due to the define not being in dix-config.h. 145493 Lack of strcasestr is handled transparently with the version from 145494 FreeBSD now anyway, so, huzzah. 145495 145496commit ad87c72edcc0d1f56658e0c4e73af335c8d5a516 145497Author: Daniel Stone <daniel@fooishbar.org> 145498Date: Wed Jul 16 02:59:51 2008 +0300 145499 145500 DIX: Add strcasestr from FreeBSD 145501 145502 Add strcasestr for use on systems which don't have it. 145503 145504commit c3c901cf44cf16bb33c4176494361b429099a372 145505Author: Daniel Stone <daniel@fooishbar.org> 145506Date: Wed Jul 16 02:54:41 2008 +0300 145507 145508 configure.ac: Fix SHA1 handling 145509 145510 Fix and marginally simplify the SHA1 handling. First, we allow people 145511 to override it. Secondly, we try for libmd. Then, we try for OpenSSL 145512 with pkg-config. In a last, desperate move, we try libcrypto on its 145513 own. This allows the server to, y'know, _link_ when using OpenSSL, 145514 instead of failing because we only have -lcrypto, and not -lssl. 145515 145516commit 69b57dc651e12a0d9a5a4295b185c62d5c0df63f 145517Author: Daniel Stone <daniel@fooishbar.org> 145518Date: Wed Jul 16 02:03:36 2008 +0300 145519 145520 dix: Actually build str(n)casecmp if we don't have it 145521 145522 Remember to add stuff to dix-config.h when you add new AC_DEFINES, 145523 people ... 145524 145525commit dcf6293030126509d7d6c61d131222037d5ed7db 145526Author: Daniel Stone <daniel@fooishbar.org> 145527Date: Wed Jul 16 01:57:00 2008 +0300 145528 145529 strcasecmp: Actually use the right license 145530 145531 Forgot to update the license when I committed the FreeBSD version, so it 145532 still had an old SGI license. Sorry. Sorry. 145533 145534commit 441f084bfe87a6ea1c94ec63f82888b8b3d81d89 145535Author: Daniel Stone <daniel@fooishbar.org> 145536Date: Wed Jul 2 19:52:58 2008 +0300 145537 145538 config: Don't attempt to use D-Bus when not strictly necessary 145539 145540 If we have D-Bus but have explicitly disabled it, don't build it. 145541 145542commit 35c89f3f5b8fa222e37b799d5bb01595e8f30d0c 145543Author: Daniel Stone <daniel@fooishbar.org> 145544Date: Wed Jul 16 01:43:58 2008 +0300 145545 145546 XFree86: Remove mysticism from Makefile.am 145547 145548 The variables were always the same, so just shove them in with the rest 145549 of the plebs. 145550 145551commit b89a59248a4a0ff06b9a0ddee45881efc6063063 145552Author: Daniel Stone <daniel@fooishbar.org> 145553Date: Fri Jun 27 12:20:56 2008 +0300 145554 145555 XFree86: Delete OSMouse code 145556 145557 This should be moved into the mouse driver, if anything. 145558 145559commit d0de5ea96d084fc5da87d8f323ddfc08fe9c03ba 145560Author: Daniel Stone <daniel@fooishbar.org> 145561Date: Wed Jun 25 17:56:28 2008 +0300 145562 145563 XFree86: Remove useless debugging code 145564 145565 Also remove documentation which told you how to use a non-module-aware 145566 GDB, albeit only with old, non-shared, modules. 145567 145568commit 4b1273c9c2da113f634be80caa28e81df3beae98 145569Author: Daniel Stone <daniel@fooishbar.org> 145570Date: Wed Jun 25 17:51:12 2008 +0300 145571 145572 XFree86: Delete empty file & function 145573 145574 xf86InitXkb() has been empty for as long as I can remember. 145575 145576commit 3a54f3f48fa1c0d60604c3ee767c569b5ec23430 145577Author: Daniel Stone <daniel@fooishbar.org> 145578Date: Thu Jun 12 01:05:04 2008 +0300 145579 145580 XFree86: Clean up init a tiny bit (no code changes) 145581 145582 Reshuffle and delete. 145583 145584commit ddcefb50dda9e398647d1c84c7153127ed26a4d2 145585Author: Daniel Stone <daniel@fooishbar.org> 145586Date: Wed Jun 11 15:14:26 2008 +0300 145587 145588 XFree86: Remove trailing whitespace 145589 145590commit e6f35f28fb3526b911101bde4aa761de8b055aef 145591Author: Daniel Stone <daniel@fooishbar.org> 145592Date: Wed Jun 11 15:13:21 2008 +0300 145593 145594 OS/KDrive/XFree86: Sanitise colour initialisation 145595 145596 OsInitColors always just returned TRUE, so just remove calls to it and 145597 insane special-case logic. Remove unused kcolor.c implementation, and 145598 merge oscolor.h into oscolor.c since it was the only user. Remove 145599 open-coded strncasecmp in oscolor.c. 145600 145601 Since we no longer need to call OsInitColors after reading the config 145602 file, just call PostConfigInit() from one place, and move PM handling to 145603 one place so we can install the signal handlers earlier. 145604 145605commit 11f9e3520249a603b95e64503ee759998ff17feb 145606Author: Daniel Stone <daniel@fooishbar.org> 145607Date: Wed Jun 11 15:09:46 2008 +0300 145608 145609 DIX: Add strncasecmp from FreeBSD, make strcasecmp args const 145610 145611 Add strncasecmp (as we're now using it) in case someone doesn't have it, 145612 and also change strncasecmp args to be const, in accordance with 145613 everything else. 145614 145615commit ae38151ddda9984effca5bb7c582540061201dce 145616Author: Daniel Stone <daniel@fooishbar.org> 145617Date: Mon May 19 08:35:05 2008 +0300 145618 145619 configure.ac: Add GLX_SYS_LIBS for Xvfb and Xnest 145620 145621 This fixes the linking for these two when using DRI2, which requires 145622 DLOPEN_LIBS. 145623 145624commit ccec16e519c143f4c0017a37e93ebc52ac472e4e 145625Author: Adam Jackson <ajax@redhat.com> 145626Date: Tue Jul 15 11:24:08 2008 -0400 145627 145628 Bug #7300: Move xf86RAC to hw/xfree86/common 145629 145630commit 9111944b292355f7478b4ae75bead8dc25edbbcb 145631Author: Julien Cristau <jcristau@debian.org> 145632Date: Tue Jul 15 10:36:38 2008 -0400 145633 145634 Bug #16674: Make sure RANDR reports refresh as 0 if pixel clock is 0. 145635 145636commit 3f3f66de669caf3606eec8c5f8d405a283cbf12d 145637Author: Brad Smith <brad@comstyle.com> 145638Date: Mon Jul 14 22:20:41 2008 +0200 145639 145640 Fix the configure tests for swap16/__swap16 macros on OpenBSD. 145641 145642 BSD systems need <sys/types.h> included before <sys/endian.h> for 145643 macros to work. 145644 <https://bugs.freedesktop.org/show_bug.cgi?id=16531> 145645 145646commit 11ee0ae9390a608a232ff94abcc0cbcf9ed7b70a 145647Author: Peter Hutterer <peter.hutterer@who-t.net> 145648Date: Mon Jul 14 10:20:11 2008 +0930 145649 145650 xfree86: append, not prepent, new input devices to xf86InputDevs. 145651 145652 If devices are prepended to the list, their wake-up order on resume is not the 145653 same as the original initialisation order. Hot-plugged devices, originally 145654 inited last, are re-enabled before the xorg.conf devices and in some cases may 145655 steal the device files. Result: we have different devices before and after 145656 suspend/resume. 145657 145658 RedHat Bug 439386 <https://bugzilla.redhat.com/show_bug.cgi?id=439386> 145659 145660commit 5bcc45e07e8726a5442567472dd29cfb5c901f2d 145661Author: Peter Hutterer <peter.hutterer@who-t.net> 145662Date: Sun Jul 13 18:40:53 2008 +0930 145663 145664 Xi: expose Enable/DisableDevice through XI_PROP_ENABLED property. 145665 145666commit c9eb0e870c87d291311491452adf7f91a911e24b 145667Author: Simon Thum <simon.thum@gmx.de> 145668Date: Thu Jul 10 22:33:39 2008 +0930 145669 145670 Add support for multiple pointer acceleration schemes. #8583 145671 145672 Available acceleration schemes: 145673 - xorg classic scheme. 145674 - the new "Predictable" polynomial accel scheme. 145675 145676 X.Org Bug 8583 <http://bugs.freedesktop.org/show_bug.cgi?id=8583> 145677 145678 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 145679 145680commit e7abe1676a6a4e4249504b8c9660cbad70569199 145681Author: Peter Hutterer <peter.hutterer@who-t.net> 145682Date: Sun Jul 13 18:41:53 2008 +0930 145683 145684 Xi: protect against NULL handlers, don't try to dereference. 145685 145686commit 2bbb12c355308d10bf123911044fbdf6ae7fb59c 145687Author: Peter Hutterer <peter.hutterer@who-t.net> 145688Date: Sun Jul 13 20:23:14 2008 +0930 145689 145690 Xi: remove ChangeDeviceControl for CoreCtl. 145691 145692 If you want to set a device to core, attach it to a master device. 145693 145694commit 18ff17756c988b6c21b76bf898e45685649d07bb 145695Author: Peter Hutterer <peter.hutterer@who-t.net> 145696Date: Thu Jul 10 19:23:14 2008 +0930 145697 145698 Xi: GetDeviceProperty reply includes deviceid. 145699 145700commit c3267106fb599213555829cb76df7848c4ebe23b 145701Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 145702Date: Wed Jul 9 10:02:41 2008 -0700 145703 145704 XQuartz: Use CFEqual to compare keyboards 145705 (cherry picked from commit 5538e43b9ae7d06d2f48842b065810ce74286eb6) 145706 145707commit 90dd2de845ae12153296f6f1bff0c87f79c57854 145708Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 145709Date: Mon Jul 7 10:55:58 2008 -0700 145710 145711 XQuartz: Some fd handoff cleanup. 145712 (cherry picked from commit 9c20a4804d97e67a988f00f49866997209cce518) 145713 145714commit 26d8030c3836816de8c12b2cb9d67315e5c887eb 145715Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 145716Date: Fri Jul 4 19:23:21 2008 -0700 145717 145718 XQuartz: Remove deprecated keyboard code. 145719 (cherry picked from commit 69cfc1a21e12bb38a6130dea2e5f20f1e6a3ee7c) 145720 145721commit 5cfcbd54d98cc77ee02a3a099ebbad9af511a0ee 145722Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 145723Date: Wed Jul 2 00:47:08 2008 -0700 145724 145725 Set machine dependent defaults for ppc64 145726 (cherry picked from commit 0733ef2e8abda99cfd62966e73017949e9cd507f) 145727 145728commit e69b9f9ca45c0c6bfb93ea9143737116bf1f2453 145729Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 145730Date: Wed Jun 25 11:51:27 2008 -0700 145731 145732 XQuartz: Set noPanoramixExtension earlier to avoid a possible race. 145733 (cherry picked from commit 49668e8a88137e9f258eae970826883b88b7d8ba) 145734 145735commit 8d2e2e1d856efec4459de2a20af642dc1ec9b8a5 145736Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 145737Date: Wed Jun 11 11:44:45 2008 -0700 145738 145739 Xquartz: Removed include directive for removed header 145740 (cherry picked from commit e65a36d57f338410c5a5b02cb5ae1214a81d072d) 145741 145742commit ebd70c82fdbe483d149ac9bdb5a64cc6e0c10e0a 145743Author: Aaron Plattner <aplattner@nvidia.com> 145744Date: Thu Jul 10 14:38:34 2008 -0700 145745 145746 Add LoaderShouldIgnoreABI to allow drivers that roll their own ABI checks (i.e. nvidia) to perform the check before ScreenInit. 145747 145748commit c42427f63f0e0495e56909ee99ad1e3c4321c39b 145749Author: Ander Conselvan de Oliveira <ander@mandriva.com.br> 145750Date: Wed Jul 9 10:50:21 2008 -0300 145751 145752 Fix incorrect test regarding keyboard map. 145753 145754 In the map stored in each keyboard device, the first line refers to 145755 minimum keycode, i.e., the 0th line refers to keycode 8. When not 145756 using XKB the wrong test caused some keys to be interpreted as 145757 locks ('m' for instance). The had to be pressed twice to generate 145758 both KeyPress and KeyRelease events. 145759 145760 Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net> 145761 145762commit e4054e5cb3d919dfc83021be3f2b9036d0e5a2be 145763Author: Peter Hutterer <peter.hutterer@who-t.net> 145764Date: Thu Jul 10 11:03:31 2008 +0930 145765 145766 Xi: pack the property stuff into a struct, make handlers a linked list. 145767 145768 We may need more than one handler to deal with a property (e.g. one in the 145769 driver, one in the DIX), so get the handlers into a linked list and call them 145770 one-by-one. This is of course slightly less entertaining than the hilarious 145771 WRAP/UNWRAP game we play in other parts of the server. 145772 145773 XIRegisterPropertyHandler/XIUnregisterPropertyHandler are the interface 145774 drivers/the DIX should use to attach themselves to the device. 145775 145776 XIDeleteAllDeviceProperties destroys everything, including the handlers. 145777 145778commit 2039c6ea43244339659323d05d96c14e41996c20 145779Author: Peter Hutterer <peter.hutterer@who-t.net> 145780Date: Mon Jul 7 22:10:17 2008 +0930 145781 145782 Xi: add support for input device properties. 145783 145784 Basically just copied from randr properties, with minor changes only. 145785 Each device supports arbitrary properties that can be modified by clients. 145786 Modifications to the properties are passed to the driver (if applicable) and 145787 can then affect the configuration of the device. 145788 145789 Note that device properties are limited to a specific device. A property set 145790 on a slave device does not migrate to the master. 145791 145792commit 4ab01fe5db7fb330b1ec463aa49b77859527e597 145793Author: Peter Hutterer <peter.hutterer@who-t.net> 145794Date: Thu Jul 10 16:27:36 2008 +0930 145795 145796 Revert "Xi: add support for input device properties." 145797 145798 Note to self: don't mix up branches with half-finished cherrypicks. 145799 145800 This reverts commit 666838fcc8b71fdeae160844160187f345cbf4a6. 145801 145802commit 666838fcc8b71fdeae160844160187f345cbf4a6 145803Author: Peter Hutterer <peter.hutterer@who-t.net> 145804Date: Mon Jul 7 22:10:17 2008 +0930 145805 145806 Xi: add support for input device properties. 145807 145808 Basically just copied from randr properties, with minor changes only. 145809 Each device supports arbitrary properties that can be modified by clients. 145810 Modifications to the properties are passed to the driver (if applicable) and 145811 can then affect the configuration of the device. 145812 145813 Note that device properties are limited to a specific device. A property set 145814 on a slave device does not migrate to the master. 145815 145816commit acce27093571497a0626cee1cdb61ddf751dbc40 145817Author: Peter Hutterer <peter.hutterer@who-t.net> 145818Date: Tue Jul 8 16:02:13 2008 +0930 145819 145820 Xext: store the GenericMasks in the resource system. 145821 145822 This fixes a severe issue - when the client died the event mask didn't get 145823 unregistered and a future event would dereference dangling pointers. By 145824 storing the event masks in the resource system we can free them when the 145825 client dies. 145826 145827commit db86b8839f286e0e2efb5638b8ab2fe608707655 145828Author: Peter Hutterer <peter.hutterer@who-t.net> 145829Date: Tue Jul 8 15:12:45 2008 +0930 145830 145831 dix: free GenericMask allocations when destroying window. 145832 145833commit 08e2c625f620688d35f3b443ffb878e8e991ff57 145834Author: Alan Coopersmith <alan.coopersmith@sun.com> 145835Date: Wed Jul 9 20:07:59 2008 -0700 145836 145837 Remove README.font, since it's just an old text copy of xorg-docs/sgml/fonts 145838 145839commit 5e847c1d4fc30a0d263a861a76982660f11998cd 145840Author: Alan Coopersmith <alan.coopersmith@sun.com> 145841Date: Mon Jul 7 17:08:01 2008 -0700 145842 145843 Improved driver selection when autoconfiguring driver without xorg.conf 145844 145845 - Allow returning multiple drivers to try for a given PCI id (for instance, 145846 try "geode" then "amd" for AMD Geode hardware) 145847 - On Solaris, use VIS_GETIDENTIFIER ioctl as well as PCI id to choose drivers 145848 - Use wsfb instead of fbdev as a fallback on non-Linux SPARC platforms 145849 145850commit 235fea0e1a2101fb76a54e6505db321674530c60 145851Author: Alan Coopersmith <alan.coopersmith@sun.com> 145852Date: Mon Jul 7 15:21:59 2008 -0700 145853 145854 Update bsd & solaris descriptions in README.OS-lib 145855 145856commit d8af9d9ab1cd98c07fdf42490dcc0cab3c655b89 145857Author: Julien Cristau <jcristau@debian.org> 145858Date: Mon Jul 7 13:14:49 2008 +0200 145859 145860 Export xkbfile.h in the SDK 145861 145862 It's needed by xkbsrv.h 145863 145864commit d78bebb20a00e8519788c75c90b467a5750c78be 145865Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 145866Date: Fri Jul 4 04:04:19 2008 -0300 145867 145868 Remove unused function and variables. 145869 145870commit a8c4727f01949da0ae266369f0d12a96c2bf0da5 145871Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 145872Date: Fri Jul 4 04:03:46 2008 -0300 145873 145874 Fix typo in the last commit. 145875 145876commit 3164288b121990a5990747c0f92d4af13b1c929a 145877Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 145878Date: Fri Jul 4 03:55:59 2008 -0300 145879 145880 Remove deprecated function miPointerPosition(). 145881 145882commit d1031a8972490ebf0ef6dc2762036a7a7172466b 145883Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 145884Date: Fri Jul 4 03:46:21 2008 -0300 145885 145886 Update README.OS-lib to reflect reality. 145887 145888commit f72500c4a09cce5dbbd16639b1951a4bf943e4ac 145889Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 145890Date: Fri Jul 4 03:45:18 2008 -0300 145891 145892 Nuke libc wrapper remaining stuffs. 145893 145894commit 3fd5fab5663fbdce2fc0b5377155aedf9a70b6c2 145895Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 145896Date: Fri Jul 4 02:51:42 2008 -0300 145897 145898 Remove inexistent function headers. 145899 145900commit 0b9ef835a0fe900c121b84e43989591e58ab1126 145901Author: Dave Airlie <airlied@redhat.com> 145902Date: Fri Jul 4 07:29:32 2008 +1000 145903 145904 modes: fix initial xorg.conf mode selection. 145905 145906 This was all kinds of broken, we ignored user preferred modes for multiple 145907 monitors and also for side-by-side configurations. 145908 145909commit e317943c125d7fdbe62c750324640eb30c88fa15 145910Author: Peter Hutterer <peter.hutterer@who-t.net> 145911Date: Wed Jul 2 12:23:03 2008 +0930 145912 145913 xfree86: AllowEmptyInput is now enabled by default if hotplugging is enabled. 145914 145915 Remove AEI check from configImpliedLayout as the setting isn't actually parsed 145916 at this point anyway (written by Sasha Hlusiak). 145917 145918 Resurrect checkInput() and check for devices there if AEI is false (this also 145919 creates the default devices if required). 145920 145921 Set AllowEmptyInput to enabled by default if hotplugging is enabled. 145922 145923commit fe5cf7cb00fd926cefff933adbdbceea7353f4c2 145924Author: Peter Hutterer <peter.hutterer@who-t.net> 145925Date: Tue Jul 1 13:18:54 2008 +0930 145926 145927 xfree86: handle missing Screen lines in the ServerLayout #16301 145928 145929 If no Screen is specified in the ServerLayout section, either take the first 145930 one from the config file or autogenerate a default screen. 145931 145932 X.Org Bug 16301 <http://bugs.freedesktop.org/show_bug.cgi?id=16301> 145933 145934commit 6674b87a7dca7d6ffd9dd9af888c5256b13d7877 145935Author: Peter Hutterer <peter@cs.unisa.edu.au> 145936Date: Sun Jun 29 20:05:27 2008 +0930 145937 145938 xfree86: move declaration of configured_device to start of function. 145939 145940 Last I checked C doesn't allow declarations halfway down a block. 145941 145942commit 1e96782011f0b4ba6ed1d3d3df118964a58d15fd 145943Author: Dave Airlie <airlied@linux.ie> 145944Date: Thu Jul 3 20:00:22 2008 +1000 145945 145946 dri2: bump libdrm numbers as 2.3.1 is released without ttm interface 145947 145948commit 1771edcb44b564f83f509748b4e4cd5b7586e2a8 145949Author: Dave Airlie <airlied@redhat.com> 145950Date: Thu Jul 3 18:57:09 2008 +1000 145951 145952 modes: check the crtc is valid before using its desired modes. 145953 145954 this fixes a crash I was getting on radeon rotate when gnome is running. 145955 I'm sure g-s-d was doing something bad, but really not crashing ftw. 145956 145957commit 689292e58cc5242cbfeaa29fcf65b6dcd9dfabb9 145958Author: Adam Jackson <ajax@redhat.com> 145959Date: Wed Jul 2 11:25:50 2008 -0400 145960 145961 Fix GLX in Xvfb and kdrive. 145962 145963commit 9a73bca859be12721ef62304b1422c455d4aec49 145964Author: Chris Ball <cjb@laptop.org> 145965Date: Wed Jul 2 09:46:06 2008 -0400 145966 145967 Remove KdOffscreenMarkUsed() as all its callers are gone. 145968 145969commit b84a27fd9fdcb579437cfa38f10cb12cf55dc7c3 145970Author: Ian Romanick <idr@us.ibm.com> 145971Date: Wed Jul 2 06:25:03 2008 -0700 145972 145973 VBO: Regenerate files based on recent changes to gl_API.xml 145974 145975 Since GL_ARB_vertex_buffer_object protocol isn't supported yet, these 145976 changes are innocuous. 145977 145978commit 703a9645f3d547144ea62dd81c41f5356a5f87d8 145979Author: Adam Jackson <ajax@redhat.com> 145980Date: Tue Jul 1 17:50:48 2008 -0400 145981 145982 Auf wiedersehen, KAA. 145983 145984commit c8216aede6c4ac41976947521d884fa010913204 145985Author: Jeremy Uejio <jeremy.uejio@sun.com> 145986Date: Tue Jul 1 13:37:12 2008 -0700 145987 145988 Sun bug #6685465: Xephyr uses wrong or bad colortable in 8-bit mode 145989 145990 <http://bugs.opensolaris.org/view_bug.do?bug_id=6685465> 145991 145992 This bug is caused by Xephyr not handling the RGB byte order correctly 145993 of the server where Xephyr is displaying on. The previous code just 145994 assumed that the order was RGB and did not take into account that 145995 Xservers may use different order (such as BGR). 145996 145997 The fix is to add a function to calculate the byte order and bits 145998 to shift based on the visual mask and the visual bits_per_rgb (which 145999 is usually 8, but could be server dependent). Since the shifts won't 146000 change once the display connection has been made, I can cache these 146001 values so that Xephyr doesn't have to keep recalculating them everytime 146002 it tries to translate the Xephyr colormap entries for Xephyr clients to 146003 the actual server colormap entries (i.e. calling the function 146004 hostx_set_cmap_entry() repeatedly for every colormap entry). 146005 146006commit 2e8daee05343b2853b677acf4554def0ceeada00 146007Author: Aaron Plattner <aplattner@nvidia.com> 146008Date: Mon Jun 30 17:29:35 2008 -0700 146009 146010 Work around the DIX losing mmWidth/mmHeight for RandR 1.1 DDXen. 146011 146012 RandR 1.1 has a physical size for each mode. It used to be that the DIX would 146013 remember these modes and pass them back up to the DDX when changing the screen 146014 configuration. The DDX uses RR_GET_MODE_MM to query the driver for the physical 146015 dimensions of the screen, allowing it to preserve the DPI. 146016 146017 With RandR 1.2, the physical dimensions are stored as part of the output, rather 146018 than per mode. The DIX only uses the sizes passed in from the DDX to select the 146019 mode pool for the "default" output, and forgets the physical sizes. Then, when 146020 reconfiguring the screen, it makes up a new RRScreenSizeRec using the dimensions 146021 from the output, screwing up the DPI. 146022 146023 This change works around this problem by ignoring the DIX and querying the real 146024 size from the driver. 146025 146026commit e2bbf2d248a2d2a2db4adfe9d1bda94cf184d8bd 146027Author: Adam Jackson <ajax@redhat.com> 146028Date: Tue Jul 1 12:01:56 2008 -0400 146029 146030 Don't try to load DRI2 if built without it. 146031 146032commit 66d7c9de82b5e733e4995f079a96efe612fca881 146033Author: Adam Jackson <ajax@redhat.com> 146034Date: Tue Jul 1 10:55:54 2008 -0400 146035 146036 Declare xf86InterpretEEDID() properly. 146037 146038commit e4335e876d254e446a965259e845ad955da5b5c2 146039Author: root <root@samich.home.cworth.org> 146040Date: Mon Jun 30 15:29:04 2008 -0700 146041 146042 Revert "XInput ABI is 2.1, not 3.1." 146043 146044 This reverts commit 76576c87b0de106e006d1cfa59d42d3a483e966c. 146045 which was an incorrect revert of previous ABI bumps. Those 146046 responsible for the accidental ABI bumps in both directions 146047 have all been sacked. 146048 146049 This allows xf86-input-mouse to build again, for example. 146050 146051commit 6447f80c17077f924b2d58e07b38d32429edcb9b 146052Author: Carl Worth <cworth@samich.home.cworth.org> 146053Date: Mon Jun 30 13:16:49 2008 -0700 146054 146055 Use AC_CHECK_LIB before PKG_CHECK_MODULES for openssl 146056 146057 Previously, the code was using PKG_CHECK_EXISTS before PKG_CHECK_MODULES, 146058 (to cater to OpenBSD systems that include openssl by default but without 146059 a .pc file). But this meant that systems that didn't have openssl installed 146060 at all would not get any error message at configure time. 146061 146062 Now, if the SHA1_Init function is found in -lcrypto without any additional 146063 flags, then that's used. Otherwise, pkg-config is used to find the right 146064 flags to link against libcrypto. And if that fails, a nice error message 146065 is now generated. 146066 146067commit 76576c87b0de106e006d1cfa59d42d3a483e966c 146068Author: Adam Jackson <ajax@redhat.com> 146069Date: Mon Jun 30 16:10:37 2008 -0400 146070 146071 XInput ABI is 2.1, not 3.1. 146072 146073commit bed12dd3df0129b7f28d0a42d470266bd3ae6b38 146074Author: Adam Jackson <ajax@redhat.com> 146075Date: Sat Jun 21 16:07:54 2008 -0400 146076 146077 Do EEDID in RANDR 1.2. 146078 146079commit 4fc18b5fc31aa287e807ee520dc9117296e3d306 146080Author: Adam Jackson <ajax@redhat.com> 146081Date: Sat Jun 21 16:07:36 2008 -0400 146082 146083 Hexdump the whole EDID block if extensions present. 146084 146085commit b4fbc31e109f1efe78613597f9a91d5363523493 146086Author: Adam Jackson <ajax@redhat.com> 146087Date: Sat Jun 21 16:00:05 2008 -0400 146088 146089 Actually fetch all blocks of EEDID if asked to. 146090 146091commit 9c4c323d50bd8b369e82a1dff28abdc15a219a5e 146092Author: Thomas Hellstrom <thomas-at-tungstengraphics.com> 146093Date: Mon Jun 30 18:59:13 2008 +0100 146094 146095 Avoid an EXA init segfault. 146096 Don't allocate each VRAM buffer with a 16MB alignment. 146097 146098commit 211e2bdcc677122145998829a7f26610c7c567e6 146099Author: Adam Jackson <ajax@redhat.com> 146100Date: Mon Jun 30 10:29:17 2008 -0400 146101 146102 Bump ABI minor numbers for the devPrivate ABI functions. 146103 146104commit d6b8d9eaffaf3f976db330bc35da3d30eb656bac 146105Author: Dan Nicholson <dbn.lists@gmail.com> 146106Date: Fri Jun 27 17:16:35 2008 -0700 146107 146108 DRI: Ensure we have the DRI headers from Mesa 146109 146110 Change the DRI configure check to look for the dri pkg-config file. This 146111 prevents people who built an Xlib libGL from bombing later in the build. 146112 146113commit 7ddfb2f1587da3ec1cd27ae270e42700adf4f560 146114Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146115Date: Thu Jun 26 21:23:31 2008 -0400 146116 146117 Fix a leak of window devPrivates in window.c/CrushTree(). 146118 146119commit a65888d6785934a669ad3b822bf2d636db351858 146120Author: Alan Hourihane <alanh@tungstengraphics.com> 146121Date: Fri Jun 27 17:32:28 2008 +0100 146122 146123 fixes for gallium stride changes 146124 146125commit e0c43fff3208350995c45604624dd4825f40598c 146126Author: Alan Hourihane <alanh@tungstengraphics.com> 146127Date: Fri Jun 27 09:46:24 2008 +0100 146128 146129 More error checking 146130 146131commit 387ec961f2a5e791fd010906a5115cf8393471f3 146132Author: Alan Hourihane <alanh@tungstengraphics.com> 146133Date: Thu Jun 26 23:34:51 2008 +0100 146134 146135 add --with-gallium-dir to specify location of gallium tree 146136 146137commit 410b0a4254ca48548b443c1e8fbbcf7a6b9aa862 146138Author: Alan Hourihane <alanh@tungstengraphics.com> 146139Date: Thu Jun 26 22:49:33 2008 +0100 146140 146141 NULL checks 146142 146143commit 621fa337fb53d535f53cec30bb7161801001bb7f 146144Author: Alan Hourihane <alanh@tungstengraphics.com> 146145Date: Thu Jun 26 22:47:20 2008 +0100 146146 146147 Fix copy surface 146148 146149commit ccd0c76472a7d8a85aed0ea6fea4011109f305d6 146150Author: Alan Hourihane <alanh@tungstengraphics.com> 146151Date: Thu Jun 26 22:27:44 2008 +0100 146152 146153 Add EXA winsys for gallium pipe driver interface. 146154 146155 Plug in the EXA framework into the pipe driver 146156 for surface_copy & surface_fill. 146157 146158 Back pixmaps with drmBO's including the front buffer. 146159 146160commit 0e1aa03708d95cc1542bee31709a345d0f748f0c 146161Author: Alan Hourihane <alanh@tungstengraphics.com> 146162Date: Thu Jun 26 22:27:14 2008 +0100 146163 146164 yet more 146165 146166commit 2b3a3bc9202f4e6f9886a08a4dfa597a83a5e933 146167Author: Alan Hourihane <alanh@tungstengraphics.com> 146168Date: Thu Jun 26 22:26:16 2008 +0100 146169 146170 more formatting fixes 146171 146172commit 2a45c28712a544c6f1c3eb753678ba1738cfdd67 146173Author: Alan Hourihane <alanh@tungstengraphics.com> 146174Date: Thu Jun 26 22:25:17 2008 +0100 146175 146176 formatting fixes 146177 146178commit 15e4b6c57484b6afb790c7dc1db9f529ba2219cf 146179Author: Peter Hutterer <peter@cs.unisa.edu.au> 146180Date: Tue Jun 24 13:46:40 2008 +0930 146181 146182 input: for non-device events (e.g. DevicePresence) use MAX_DEVICES as id. 146183 146184 Using id = 0 only worked pre-MPX since XInput didn't allow XOpenDevice for the 146185 core devices (0 and 1). Now we can now legally register for events so we may 146186 overwrite our device-independent classes with the ones selected for the VCP. 146187 146188 So, increase the EMASKSIZE to MAX_DEVICES + 1 and use MAX_DEVICES as the ID 146189 when we don't have a device. 146190 146191commit f0c56ffd980db19bea271aeebf118bd931f6f40d 146192Author: Peter Hutterer <peter@cs.unisa.edu.au> 146193Date: Tue Jun 24 12:58:49 2008 +0930 146194 146195 dix: don't dereference dev->key if it doesn't exist. 146196 146197 This may happen if we check for a passive grab on a floating pointer device. 146198 146199commit 01f0b9c9784c55c94104bafa6dd545b373c64104 146200Author: Alan Coopersmith <alan.coopersmith@sun.com> 146201Date: Wed Jun 25 19:23:20 2008 -0700 146202 146203 xf86GetOptValFreq wants a pointer to double, but maxPixClock is an int 146204 146205commit 7ce6dcef118139d0ccfc136b9a07f5bf0c078d70 146206Author: Alan Coopersmith <alan.coopersmith@sun.com> 146207Date: Wed Jun 25 15:06:49 2008 -0700 146208 146209 Don't log null device name in XkbDDXLoadKeymapByNames 146210 146211commit cbc20d92de92aad5ca240310a9156ccf97c24a01 146212Author: Adam Jackson <ajax@redhat.com> 146213Date: Wed Jun 25 09:52:36 2008 -0400 146214 146215 XTrap delenda est. 146216 146217commit a81b5ce452766bbc4b2ac0849711284ba8591b46 146218Author: Peter Hutterer <peter@cs.unisa.edu.au> 146219Date: Wed Jun 25 13:23:09 2008 +0930 146220 146221 dix: free memory associated with the EventList. #16507 146222 146223 Un-deprecates CloseDownEvents. 146224 146225 X.Org Bug 16507 <http://bugs.freedesktop.org/show_bug.cgi?id=16507> 146226 146227commit 3a2beb2f867fd47dc5a7723dcac345ea0b0ecf3d 146228Author: Peter Hutterer <peter@cs.unisa.edu.au> 146229Date: Wed Jun 25 13:12:41 2008 +0930 146230 146231 dix: InitEventList does not need to be _X_EXPORT. 146232 146233commit 385943e0e97463ce4681a9b6a4a40d7e3c91e51e 146234Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 146235Date: Tue Jun 24 23:13:32 2008 -0300 146236 146237 Set --enable-builtin-fonts as default. 146238 146239 Some reasons to embed fonts by default: 146240 1. X server doesn't pick a good default font path so it's easiest just 146241 to built in the core fonts and let new X hackers more happy. Developers 146242 and distro guys are wise enough to just set --disable-builtin-fonts 146243 when they want. 146244 146245 2. Seems that this is by far the most popular FAQ 146246 (http://www.x.org/wiki/FAQErrorMessages). 146247 146248 3. No one gave a good argument to not do this: 146249 http://lists.freedesktop.org/archives/xorg/2008-May/035479.html 146250 146251commit f3b29efc4d7bb346cdd3fd79e780f1022c8f46e8 146252Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 146253Date: Tue Jun 24 23:07:49 2008 -0300 146254 146255 Remove NOFONTSERVERACCESS macro 146256 146257 dix: this is semantically different but it goes through the same path in 146258 the code. 146259 146260commit 2ff7eb4995b663cf12a01124745de600f38d0bfc 146261Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 146262Date: Tue Jun 24 22:09:34 2008 -0300 146263 146264 xfree86: kill INHERIT_LOCK_STATE macro. 146265 146266 I don't know how this survived so long. Google didn't show 146267 anything usable related with this. 146268 146269commit a3ec22627355fc08730ad7e90022e374763d333f 146270Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146271Date: Tue Jun 24 20:32:24 2008 -0400 146272 146273 Fix a leak in the code that parses the protocol names. 146274 146275 Also added some comments. 146276 Reported by Ben Gamari (bug #16492). 146277 146278commit a4cb25f8c816adc3bdf6a28d5beb558e8be40121 146279Author: Ben Gamari <bgamari@gmail.com> 146280Date: Tue Jun 24 12:08:35 2008 +0930 146281 146282 dix: free FocusPrivates when deleting a window. #16493 146283 146284 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 146285 146286 X.Org Bug 16493 <http://bugs.freedesktop.org/show_bug.cgi?id=16493> 146287 146288commit 32a7bbbba6fec0535a8faa1d6d2ecebe786bcc81 146289Author: Peter Hutterer <peter@cs.unisa.edu.au> 146290Date: Tue Jun 24 09:28:05 2008 +0930 146291 146292 xfree86: fix up leftover allocation of xf86Events. 146293 146294 Thanks to Eamon Walsh for spotting this. 146295 146296commit bcc88a1d46c7ba469c4e2861875f84f8c77d4b15 146297Author: Peter Hutterer <peter@cs.unisa.edu.au> 146298Date: Sun Jun 22 22:34:30 2008 +0930 146299 146300 xkb: remove unused variable is_core. 146301 146302commit 9ed9d98dec58d1f7468d02c4b951811b8e1ca7c1 146303Author: Julien Cristau <jcristau@debian.org> 146304Date: Tue Jun 24 22:36:32 2008 +0200 146305 146306 x86emu: fix build, fallout from 9719354 146307 146308commit e9b4f3cd856e2401675452c387b4155246239f91 146309Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146310Date: Mon Jun 23 16:18:20 2008 -0400 146311 146312 Fix "warning: suggest parentheses around && within ||". 146313 146314commit 8910dde96ec4318c9c44fb9717b6b52fad56cd1d 146315Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146316Date: Mon Jun 23 14:51:38 2008 -0400 146317 146318 Fix "warning: assignment discards qualifiers from pointer target type". 146319 146320commit 95d4ede538fbb68049ba3efa0acb0e9712e5cb01 146321Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146322Date: Fri Jun 20 23:49:32 2008 -0400 146323 146324 Fix "warning: passing argument 1 of `modeIsPresent' from incompatible pointer type". 146325 146326commit fa1545eeda2a7feebc006840a44d390177634231 146327Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146328Date: Fri Jun 20 23:47:22 2008 -0400 146329 146330 Fix "warning: passing argument 1 of `xf86addNewLoadDirective' from incompatible pointer type". 146331 146332commit 2725ba1fca45e5ba241695e24333d8efab1f862b 146333Author: Adam Jackson <ajax@redhat.com> 146334Date: Tue Jun 24 14:41:32 2008 -0400 146335 146336 Also remove xorgconfig from configure.ac. 146337 146338commit 9719354ae04a41e9834ac9ba8fd9e895092ba4f1 146339Author: Adam Jackson <ajax@redhat.com> 146340Date: Tue Jun 24 14:37:06 2008 -0400 146341 146342 Check for __amd64__, not __x86_64__. 146343 146344 Spiritual revert of 1fa4de80fcfc697b5e5879cc351fb3e9dbf6acbe. Intel's C 146345 compiler claims to be gcc-compatible; if they're not defining the same 146346 macros as gcc then that's their bug, not ours. Even if we were to do 146347 this aliasing we should do it once and for all in servermd.h. 146348 146349commit 249c892784ca5e8c75863dd82097ca2bedec4723 146350Author: Roland Bär <roland@verifysoft.de> 146351Date: Tue Jun 24 14:13:02 2008 -0400 146352 146353 Bug #11857: Avoid running off the end of a MAXCLIENTS-sized array. 146354 146355commit 9e5b3deafb97ec1f83e6bfe067bc68df7385bc6a 146356Author: Adam Jackson <ajax@redhat.com> 146357Date: Tue Jun 24 14:08:04 2008 -0400 146358 146359 Bug #11842: Fix emulation of int1A PCI BIOS services. 146360 146361 Use only %di to name the PCI register to read/write, rather than %edi. 146362 DOS is only expecting the base PCI config space anyway, and the BIOS 146363 might be using the high bits of %edi. 146364 146365commit d34430414ac0e77eec61ab0ac9ef427b236eb639 146366Author: Adam Jackson <ajax@redhat.com> 146367Date: Tue Jun 24 13:32:55 2008 -0400 146368 146369 Remove xorgconfig. Use X -configure instead. 146370 146371commit 574c62e305e5380bb1b0f4bfec4f7cc3e3f03ada 146372Author: Owen Taylor <otaylor@fishsoup.net> 146373Date: Tue Jun 24 12:40:47 2008 -0400 146374 146375 Bug #15369: Don't reduce Over to Src for transform + RepeatNone. 146376 146377 Only do no-src-alpha optimizations for a RepeatNone source if we can 146378 easily probe that we won't sample outside the edges of the source. 146379 146380commit 100afae578e59e31b65a6264c281ba7deea5ac39 146381Author: Adam Jackson <ajax@redhat.com> 146382Date: Tue Jun 24 10:52:31 2008 -0400 146383 146384 Bug #15586: (Correctly) refuse to redirect the root window. 146385 146386commit 4c4e06af7950df509fa02099788be66cf37a4d01 146387Author: Paul Bender <pebender@gmail.com> 146388Date: Tue Jun 24 10:44:47 2008 -0400 146389 146390 Bug #15665: Fix building without Composite. 146391 146392commit 59510a76436c739fd9421f01cc4210607d9e7f57 146393Author: Adam Jackson <ajax@redhat.com> 146394Date: Tue Jun 24 10:40:36 2008 -0400 146395 146396 Remove some default modes that really shouldn't be in the default set. 146397 146398commit b622f6dc5d8d70d17be0374a85a9b16b43a62a01 146399Author: Julien Cristau <jcristau@debian.org> 146400Date: Mon Jun 23 23:45:57 2008 +0200 146401 146402 kdrive: fix building with --disable-xv (bug#16488) 146403 146404commit 3b587826924c60f6736dc5faf584b964fde1b4da 146405Author: Jesse Ruffin <jesse@eloquentpeasant.net> 146406Date: Mon Jun 23 14:35:20 2008 -0400 146407 146408 Bug #16302: Fix GLX drawable refcounting in DMX. 146409 146410commit c8d066a15142678041c1d82ccf530dcdb2ea74ca 146411Author: Adam Jackson <ajax@redhat.com> 146412Date: Mon Jun 23 13:59:48 2008 -0400 146413 146414 Implement bswap in x86emu. 146415 146416 Yes, this is a 486+ instruction and thus not strictly legal in vm86 146417 mode, but enough BIOSes use it (looking at you VIA) that we might as 146418 well implement it. 146419 146420commit b55fbca4f0705aeff1c69d3ef851c5ff5af6ed94 146421Author: Adam Jackson <ajax@redhat.com> 146422Date: Fri Jun 20 13:27:32 2008 -0400 146423 146424 Bug #12414: Create full-fledged pixmaps in fb24_32ReformatTile(). 146425 146426 ... instead of creating pixmaps that only fb knows about, which will 146427 have no devPrivates for any other subsystem and thus cause havoc if 146428 (when) they leak out. 146429 146430commit 155fcecb320dbcfb899d6bcd78048ce0993bf28b 146431Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146432Date: Mon May 19 08:09:33 2008 +0200 146433 146434 Various messages clean-ups. 146435 146436commit 41a65e110409a290288180e8e1517f053f877586 146437Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146438Date: Mon May 19 06:12:04 2008 +0200 146439 146440 Fix ipc check on OpenBSD 146441 146442commit ce4d4b3c1b0e25f64d6f781b09528b2f3ae33207 146443Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146444Date: Mon May 19 05:19:01 2008 +0200 146445 146446 build fix 146447 146448commit 19c7e9da55646f1f6e05c28cd71865cd8d84e1ff 146449Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146450Date: Mon May 19 02:24:17 2008 +0200 146451 146452 preserve errno around the SIGIO handler 146453 146454commit 0a81d476388c51f8aa723400dffc375e112ea97a 146455Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146456Date: Mon May 19 02:20:44 2008 +0200 146457 146458 Remove RCS Id. 146459 146460commit c0d7ad9c88172d753b932069743ae65f75de3aa9 146461Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146462Date: Thu Jun 19 16:51:18 2008 +0200 146463 146464 Let code build with gcc 2.95 146465 146466commit 219dd2d44a9da40427ad48b0f09b981fe0551f4b 146467Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146468Date: Sun May 18 13:53:32 2008 +0200 146469 146470 sprintf() -> snprintf() 146471 146472commit cfcb3da75e807dec225cc7ea469e04d2db10bf73 146473Author: Keith Packard <keithp@keithp.com> 146474Date: Wed Jun 18 11:37:00 2008 -0700 146475 146476 Make button down state a bitmask. Master buttons track union of slave buttons 146477 146478 Mixing usage where some parts of the code treated this field as a bitmask 146479 and other parts as an array of card8 was wrong, and as the wire protocol 146480 wanted bitmasks, it was less invasive to switch the newer counting code use 146481 booleans. 146482 146483 Master devices track slave buttons by waiting for all slave buttons to be 146484 released before delivering the release event to the client. 146485 146486 This also removes the state merging code in DeepCopyDeviceClasses -- that 146487 code was changing master device state without delivering any events, 146488 violating protocol invariants. The result will be that existing slave 146489 button state which does not match the master will not be visible through the 146490 master device. Fixing this would require that we synthesize events in this 146491 function, which seems like a bad idea. Note that keyboards have the same 146492 issue. 146493 146494 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 146495 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 146496 146497commit 3cc5ae6a4f725483612c00fc8bcc2c61607f66a8 146498Author: Peter Hutterer <peter@cs.unisa.edu.au> 146499Date: Wed Jun 18 16:14:42 2008 +0930 146500 146501 dix: don't attempt to send Enter/Leave events if we have a keyboard. 146502 146503 Dereferencing into dev->valuator could crash the server, although it looks 146504 like I could only reproduce this by having a keyboard send an event after it 146505 was created and the WM was still replaying. Or so. 146506 146507commit 528b4e36ade482df99747081688ae52cfaeb28eb 146508Author: Alan Hourihane <alanh@tungstengraphics.com> 146509Date: Wed Jun 18 22:34:02 2008 +0100 146510 146511 Set driverPriv immediately on CreatePixmap. 146512 146513 If it's NULL anyway, we bail, if not, it lets 146514 ModifyPixmapHeader know about the private. 146515 146516commit 068fa63496c7c363706f0d356b5132bc591c04c6 146517Author: Michel Dänzer <michel@tungstengraphics.com> 146518Date: Wed Jun 18 09:23:09 2008 +0930 146519 146520 dix: fix a crash when the client has a different byteorder. 146521 146522 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 146523 146524commit d21155a3e9b51df946766926bc6155c8972c4439 146525Author: Peter Hutterer <peter@cs.unisa.edu.au> 146526Date: Sun Jun 15 20:00:41 2008 +0930 146527 146528 input: fix up usage of button->down, used to be a bitmask, is now an array. 146529 146530 device->button->down used to be a 32-byte bitmask with one bit for each 146531 button. This has changed into a 256-byte array, with one byte assigned for 146532 each button. Some of the callers were still using this array as a bitmask 146533 however, this is fixed with this patch. 146534 146535 Thanks to Keith Packard for pointing this out. See also: 146536 http://lists.freedesktop.org/archives/xorg/2008-June/036202.html 146537 146538commit 2b9c829bdebd16910bdf48b9d64862e3d34f5b7f 146539Author: Peter Hutterer <peter@cs.unisa.edu.au> 146540Date: Wed Jun 18 09:55:53 2008 +0930 146541 146542 Xi: don't drop the grab status in XExtendedGrabDevice(). 146543 146544commit 4f3ed54dc9ea7780ff03378b13de0fddfdec2e15 146545Author: Peter Hutterer <peter@cs.unisa.edu.au> 146546Date: Tue Jun 17 10:43:36 2008 +0930 146547 146548 mi: remove trailing whitespaces from misprite.c 146549 146550commit 79dd600942bbac3c6b531f284b42c7b2c822da90 146551Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146552Date: Tue Jun 17 19:09:44 2008 -0400 146553 146554 SELinux: Add an extension alias under the OS-agnostic "Flask" name. 146555 146556commit 656d3d7623c6b83024e9cdc60d1257f4d87aa268 146557Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146558Date: Tue Jun 17 17:55:11 2008 -0400 146559 146560 SELinux: Bring server-side name registry up to date. 146561 146562commit f3490d3eba94c7d9e760d6d21991ab6677196c2b 146563Author: Ian Romanick <idr@us.ibm.com> 146564Date: Tue Jun 17 10:35:24 2008 -0400 146565 146566 Bug #15169: Make the server build again on Alpha. 146567 146568 Still won't work until the kernel makes the resource files actually exist. 146569 146570commit 47969d79289ad92f658c9a6dd2a2fc6b5f1b6c99 146571Author: Alan Hourihane <alanh@tungstengraphics.com> 146572Date: Mon Jun 16 15:07:39 2008 +0100 146573 146574 Add DRI2 support. 146575 146576 Add EXA stubs. 146577 146578 Currently tested with i915. 146579 146580commit 6a1910a0e562e36e0836e948b287ec37c90df02d 146581Author: Zack Rusin <zack@tungstengraphics.com> 146582Date: Mon Jun 16 00:31:53 2008 -0400 146583 146584 Adding some readme. 146585 146586commit 8c9234a163eceda2abc0a2523e0f5587ea399935 146587Author: Keith Packard <keithp@keithp.com> 146588Date: Sun Jun 15 10:01:28 2008 +0930 146589 146590 Xi: avoid double button mapping. 146591 146592 Button events were mapped once in GetPointerEvents and then again in 146593 UpdateDeviceState. While it might make sense to just fix up UpdateDeviceState, 146594 it turns out to be better to leave the raw button number in the event because 146595 DGA reports raw device events without button translation, and so when it calls 146596 UpdateDeviceState, the button down counts get scrambled and buttons get stuck 146597 down. 146598 146599 See also: 146600 http://lists.freedesktop.org/archives/xorg/2008-June/036201.html 146601 146602 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 146603 146604commit d894e86aa5bea81147c4a7ad8630bcfb04156792 146605Author: Peter Hutterer <peter@cs.unisa.edu.au> 146606Date: Sun Jun 15 10:55:08 2008 +0930 146607 146608 randr: remove superfluous check against uninitialised variable. 146609 146610 Thanks to Eamon Walsh for pointing this out. 146611 146612commit 5544c51447f551dfc6df64438873a7ce64743976 146613Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146614Date: Fri Jun 13 22:49:47 2008 -0400 146615 146616 Fix "warning: the address of ‘newTypes’ will always evaluate as ‘true’". 146617 146618commit 2391c409a2840d61fed93832650c0d6c82ebebdf 146619Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146620Date: Fri Jun 13 22:48:17 2008 -0400 146621 146622 Fix "warning: unused variable `s'". 146623 146624commit 61dafe384a88fc397d7cd55cb23f4fdc3fb0645c 146625Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146626Date: Fri Jun 13 22:47:10 2008 -0400 146627 146628 Fix "warning: unused variable `tsize'". 146629 146630commit 37a64b02df02d42a19d4b777930dc5e457d97d63 146631Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146632Date: Fri Jun 13 22:41:58 2008 -0400 146633 146634 Fix "warning: ‘rc’ may be used uninitialized in this function". 146635 146636commit e90fcd8294fe763c97610d39cab50f2836f87dca 146637Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146638Date: Fri Jun 13 22:26:43 2008 -0400 146639 146640 Fix "warning: no previous prototype for ‘miRROutputGetProperty’". 146641 146642commit 0b00bbca5df140edbfaffab627bd82165f1b1420 146643Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146644Date: Fri Jun 13 22:22:22 2008 -0400 146645 146646 Fix "Warning: no previous prototype for `SecureRPCInit'". 146647 146648commit 1c2f038d69605b0ed1067e47dc75ca6f216991f5 146649Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146650Date: Fri Jun 13 21:59:09 2008 -0400 146651 146652 Fix multiple warnings in os/xdmauth.c. 146653 146654commit 384ebe02b354bf1ad0a2ac25e4b5290f2e795ea1 146655Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146656Date: Fri Jun 13 21:49:01 2008 -0400 146657 146658 Fix "warning: label ‘done’ defined but not used". 146659 146660commit 720e44f9730b9e275579483c2a34b16bba77e146 146661Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146662Date: Fri Jun 13 21:42:37 2008 -0400 146663 146664 Fix "warning: no previous prototype for ‘XevieExtensionInit’". 146665 146666commit 9d871e19c9dee5d3ee7f123249e8890417d83093 146667Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146668Date: Fri Jun 13 21:35:29 2008 -0400 146669 146670 Fix "warning: unused variable ‘pDev’, `pPointer'". 146671 146672commit 60f64af075e8ac3e63586452f2f3d338a05785f0 146673Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146674Date: Fri Jun 13 21:23:31 2008 -0400 146675 146676 Fix "warning: passing argument 1 of 'pixman_fill' from incompatible pointer type". 146677 146678commit c448460fdc23262f2b711a32599090c85c70447a 146679Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146680Date: Fri Jun 13 21:05:26 2008 -0400 146681 146682 Fix "warning: pointer targets in assignment differ in signedness". 146683 146684commit 17ed6242e73c5f733601b15732ab30c23acf14d2 146685Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146686Date: Fri Jun 13 21:03:17 2008 -0400 146687 146688 Fix "warning: assignment from incompatible pointer type". 146689 146690commit 02a8b118cd07d98f12aab8325d8443125703ba7d 146691Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146692Date: Fri Jun 13 21:01:50 2008 -0400 146693 146694 Fix "warning: implicit declaration of function ‘miPointerGetScreen’". 146695 146696commit 2d7ba09dc4b5eff5dba8d7867f285111574b1737 146697Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 146698Date: Fri Jun 13 16:39:40 2008 -0400 146699 146700 Make devPrivates lookup functions ABI instead of static inlines. 146701 146702 This is required to preserve compatibility across changes to the 146703 internal representation of the privates list. 146704 146705commit 9e0e558f263a132babf8c91d31230831653f19c1 146706Author: Adam Jackson <ajax@redhat.com> 146707Date: Fri Jun 13 11:17:49 2008 -0400 146708 146709 Fix unterminated endif. 146710 146711commit 6bbfadb9094bb6437a6825b8293d13a32e041aff 146712Author: Adam Jackson <ajax@redhat.com> 146713Date: Fri Jun 13 11:00:50 2008 -0400 146714 146715 ... and stay dead. 146716 146717commit f31bd087e8a7f65cd588bd1d022bb18e72b2a60c 146718Author: Adam Jackson <ajax@redhat.com> 146719Date: Fri Jun 13 10:35:07 2008 -0400 146720 146721 Death to mfb. 146722 146723 Anyone still interested in 1 or 4 bpp framebuffers, talk to fb. 146724 146725commit af86c3b7ce705c1711c8cf8bf05a7a02b8afccf6 146726Author: George Sapountzis <gsap7@yahoo.gr> 146727Date: Sat May 24 21:16:34 2008 +0300 146728 146729 clean .gitignore 146730 146731commit 3108980f5e78c129b23ff32f72564cf9a746c693 146732Author: George Sapountzis <gsap7@yahoo.gr> 146733Date: Fri Jun 13 15:58:56 2008 +0300 146734 146735 glx: drop unused cleargc 146736 146737commit 23b55a61f89f69454a3b0e3413b1f07d5fdf43aa 146738Author: Michel Dänzer <michel@tungstengraphics.com> 146739Date: Fri Jun 13 11:13:56 2008 +0200 146740 146741 AIGLX/DRI1: Switch to server context for calling pScreen->GetImage. 146742 146743 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=16292 . 146744 146745commit d15b3790307053587df8daed1936ff6923881b63 146746Author: Adam Jackson <ajax@redhat.com> 146747Date: Thu Jun 12 16:00:50 2008 -0400 146748 146749 Remove xgl. 146750 146751 This code hasn't been updated with anything even resembling what anyone is 146752 shipping in nearly thirty months. It hasn't built out of the box since 146753 7.1. Most of its features over AIGLX are accomplished with DRI2 and 146754 friends. 146755 146756commit c37a93b4850866c713e3cbf39a08bbc9645e24d4 146757Author: Dave Airlie <airlied@redhat.com> 146758Date: Thu Jun 12 14:54:56 2008 +1000 146759 146760 modes: make aspect choosing work on single output case. 146761 146762 In the single output enabled case we never enter the loop and test 146763 never gets set and so we fail to match a good mode. 146764 146765 This was causing my 2560x1600 to end up at 2048x1536. 146766 146767commit 23e71ef71a178505494d4b410f9314acfff81524 146768Author: Dave Airlie <airlied@redhat.com> 146769Date: Thu Jun 12 09:04:24 2008 +1000 146770 146771 dbe: fix DoS reported by iDefense. 146772 146773 This isn't a security problem just a user could DoS themselves for fun or profit. 146774 146775commit 390b155135a451fa728b73c8107aea91dfd56ed1 146776Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 146777Date: Wed Jun 11 18:21:33 2008 -0300 146778 146779 Welcome to hotplug era. 146780 146781 `rm -rf xorg.conf` does not imply an error message. Maybe a warning. 146782 146783 checkInput() is defined but not used. 146784 146785commit 6e018f6a30dab1259b8fdc1ae4c84056b2acf362 146786Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 146787Date: Wed Jun 11 18:10:55 2008 -0300 146788 146789 No need for inputPending field in xf86Info. 146790 146791commit 881e551de92e5664f92ef1961092e35cd7a77b15 146792Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 146793Date: Wed Jun 11 18:03:33 2008 -0300 146794 146795 CopyGetMasterEvent() can be static. 146796 146797commit 656949cbe24648a5b58f7df6ed49b25af0b47601 146798Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 146799Date: Wed Jun 11 17:57:19 2008 -0300 146800 146801 Remove more unused things. 146802 146803 In function ‘miSpriteCloseScreen’: 146804 warning: unused variable ‘pDev’ 146805 146806commit 84db8837d0faaac008be61df9f1261cd3028e923 146807Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 146808Date: Wed Jun 11 17:47:00 2008 -0300 146809 146810 Remove unused xf86DeviceCursorUndisplay() function. 146811 146812commit dba5455f06c439c7e7d399f30dc7067e98e6e50b 146813Author: Kristian Høgsberg <krh@redhat.com> 146814Date: Wed Jun 11 11:41:34 2008 -0400 146815 146816 Un-K&R shm extension. 146817 146818commit 9171206db349a0c6fda719746be0b15049d57aaa 146819Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146820Date: Tue Jun 10 12:23:03 2008 -0600 146821 146822 CVE-2008-2362 - RENDER Extension memory corruption 146823 146824 Integer overflows can occur in the code validating the parameters for 146825 the SProcRenderCreateLinearGradient, SProcRenderCreateRadialGradient 146826 and SProcRenderCreateConicalGradient functions, leading to memory 146827 corruption by swapping bytes outside of the intended request 146828 parameters. 146829 146830commit 5257a0f83d5f3d80d0cd44dd76d047bac3869592 146831Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146832Date: Tue Jun 10 12:22:30 2008 -0600 146833 146834 CVE-2008-2361 - RENDER Extension crash 146835 146836 An integer overflow may occur in the computation of the size of the 146837 glyph to be allocated by the ProcRenderCreateCursor() function which 146838 will cause less memory to be allocated than expected, leading later to 146839 dereferencing un-mapped memory, causing a crash of the X server. 146840 146841commit c5f69b297b1227cb802394fa90efdbe1de607f3c 146842Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146843Date: Tue Jun 10 12:21:26 2008 -0600 146844 146845 CVE-2008-2360 - RENDER Extension heap buffer overflow 146846 146847 An integer overflow may occur in the computation of the size of the 146848 glyph to be allocated by the AllocateGlyph() function which will cause 146849 less memory to be allocated than expected, leading to later heap 146850 overflow. 146851 146852commit 063f18ef6d7bf834225ddfd3527e58c078628f5f 146853Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146854Date: Tue Jun 10 12:20:43 2008 -0600 146855 146856 CVE-2008-1379 - MIT-SHM arbitrary memory read 146857 146858 An integer overflow in the validation of the parameters of the 146859 ShmPutImage() request makes it possible to trigger the copy of 146860 arbitrary server memory to a pixmap that can subsequently be read by 146861 the client, to read arbitrary parts of the X server memory space. 146862 146863commit 95d162c4389857d960da9b0158345c1714e91f31 146864Author: Matthieu Herrb <matthieu.herrb@laas.fr> 146865Date: Tue Jun 10 12:20:00 2008 -0600 146866 146867 CVE-2008-1377 - RECORD and Security extensions memory corruption 146868 146869 Lack of validation of the parameters of the 146870 SProcSecurityGenerateAuthorization SProcRecordCreateContext 146871 functions makes it possible for a specially crafted request to trigger 146872 the swapping of bytes outside the parameter of these requests, causing 146873 memory corruption. 146874 146875commit 656d5d98855eb608ec6581f8c574f343a216ea32 146876Author: Peter Hutterer <peter@cs.unisa.edu.au> 146877Date: Wed Jun 11 17:32:59 2008 +0930 146878 146879 kdrive: fix wrong use of &, should be | instead. 146880 146881 Fallout from 26e7e69ab893d1f2b35213250ada40ec90944f62 146882 146883commit 6528eb885dea0326aacdd026d44d82ab07f4a4fd 146884Author: Peter Hutterer <peter@cs.unisa.edu.au> 146885Date: Wed Jun 11 15:01:36 2008 +0930 146886 146887 Xext: init firstValuator to zero, otherwise core XTest events may get lost. 146888 146889commit 6ab311c92d036dc63ee92ab4bb231f5f61ebe69f 146890Author: Peter Hutterer <peter@cs.unisa.edu.au> 146891Date: Mon Jun 9 11:45:57 2008 +0930 146892 146893 dix: Call RealizeCursor during InitializeSprite. 146894 146895 RealizeCursor should be called when the cursor is allocated. However, when the 146896 root cursor is allocated, no devices exist yet, and thus RealizeCursor is 146897 never called. This may lead to segfaults lateron in DDXes like Xnest that 146898 actually need to do something for each cursor, and lateron rely on that 146899 DDX-specific data for each cursor has been initialized. 146900 146901commit e083b5a07507d53de0d1d365ef1565346284c997 146902Author: Peter Hutterer <peter@cs.unisa.edu.au> 146903Date: Mon Jun 9 11:23:53 2008 +0930 146904 146905 xnest: wrap the xnest cursor sprite funcs around the mi funcs. 146906 146907 Modelled after the xfree86 code. Call miDCInitialize to init the SW rendering 146908 engine, then take the pointers, store it in a xnest-local variable, and put 146909 the xnest-specific sprite funcs in place. In the xnest sprite funcs, call 146910 through to the mi sprite funcs after doing xnest-specific stuff. 146911 146912commit ea6a02c048e8589e99ee3b459f3612eb3d041558 146913Author: Peter Hutterer <peter@cs.unisa.edu.au> 146914Date: Mon Jun 9 11:22:33 2008 +0930 146915 146916 mi: protect mipointrst.h against multiple includes. 146917 146918commit a3a7c12fcf8e4ac1418f9ea53f76091f309a721b 146919Author: Peter Hutterer <peter@cs.unisa.edu.au> 146920Date: Sun Jun 8 23:38:45 2008 +0930 146921 146922 xfree86: Fix up xf86ScaleAxis 146923 146924 Some driver still call it, so we might as well work correctly. Always 146925 resetting X to the Sxhigh is sub-optimal. 146926 146927commit f8431a62d57ab3118a0b409bce977db8261e2930 146928Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 146929Date: Tue Jun 10 02:15:25 2008 -0700 146930 146931 XQuartz: Under the new startup model, we no longer need to do the foreground/background dance. 146932 (cherry picked from commit 4505bae5d7341e1241be50c25cb3d1b941701de4) 146933 146934commit c180a52332385b98635068d63a8fe0bba4835be7 146935Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 146936Date: Tue Jun 10 10:47:37 2008 -0700 146937 146938 XQuartz: Removed code path for old startup 146939 (cherry picked from commit a9ee6b0d00fab01a78408a85e6542e88c19fda7c) 146940 146941commit 49751fee3b82ebc4917bfb168ec78aad7874f1f1 146942Author: Roland Scheidegger <sroland@tungstengraphics.com> 146943Date: Tue Jun 10 15:40:48 2008 +0200 146944 146945 glx: copy msaa visual capabilities 146946 146947commit 5170c169db78cad3fe47dee4e438b48e67f0c7d1 146948Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 146949Date: Tue Jun 10 01:39:15 2008 -0700 146950 146951 XQuartz: Fixed the "laggy" startup under the new model. This was caused by xinit blocking expecting SIGUSR1 and our stub never sending it. 146952 (cherry picked from commit ee92aced10d0743c4658e53b58b5d9f5a094a415) 146953 146954commit 01ace5f3fd56bc789ad7448059fcf907ef19a1c8 146955Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 146956Date: Wed Jun 4 18:28:20 2008 -0700 146957 146958 XQuartz: Updated icon with more rsolutions and made it more dark-background friendly 146959 (cherry picked from commit d8cf5623faab952a0f1196c8fe10baa09d1fc089) 146960 146961commit 607b0d09ea003f87cfb3331e59e13495a483832a 146962Author: Aaron Plattner <aplattner@nvidia.com> 146963Date: Mon Jun 9 09:52:04 2008 -0700 146964 146965 CreateColormap returns Success on success, not TRUE. 146966 146967 Fixes a problem where enabling color index overlays disables the RENDER 146968 extension. 146969 146970commit 2854abd39d4deb79c9ee9bcc931fad969392cfc5 146971Author: Peter Hutterer <peter@cs.unisa.edu.au> 146972Date: Sun Jun 8 23:20:11 2008 +0930 146973 146974 xnest: fix up sprite funcs to stop compiler warnings. 146975 146976commit c68ca0c7b2c72df0501d26bf7f640029ff2d9568 146977Author: Peter Hutterer <peter@cs.unisa.edu.au> 146978Date: Sun Jun 8 23:14:20 2008 +0930 146979 146980 xnest: switch to using EventList rather than xEvents. 146981 146982commit 34429d16d319ef2709c22a5c22f45d0e82c78b20 146983Author: Peter Hutterer <peter@cs.unisa.edu.au> 146984Date: Sun Jun 8 23:08:34 2008 +0930 146985 146986 xnest: call miDCInitialize rather than miPointerInitialize. 146987 146988 This unfortunately gives us a lots of artefacts, so cursor rendering cannot be 146989 assumed complete. But it's better than crashing. 146990 146991commit a7d4bec884c24c0e1792fd7d6677cdd0a69978e3 146992Author: Peter Hutterer <peter@cs.unisa.edu.au> 146993Date: Sun Jun 8 23:03:20 2008 +0930 146994 146995 mi: remove leftover old license comment. 146996 146997commit 585125685b5f09245021c8ee15f57950fe6cdd20 146998Author: Peter Hutterer <peter@cs.unisa.edu.au> 146999Date: Sun Jun 8 22:43:47 2008 +0930 147000 147001 ephyr: some whitespace changes. 147002 147003commit 782394fe53220deafed06ad282b5f9ce42964c42 147004Author: Julien Cristau <jcristau@debian.org> 147005Date: Fri Jun 6 12:10:35 2008 +0200 147006 147007 mi: Fix typo in comments about deprecated functions 147008 147009commit 6b96281100f2118fe9d99536c33d48298a5bce7b 147010Author: Michel Dänzer <michel@tungstengraphics.com> 147011Date: Fri Jun 6 11:01:03 2008 +0200 147012 147013 EXA: Fix exaGetPixmapFirstPixel() crash if the driver has a CreatePixmap hook. 147014 147015 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=16243 147016 147017commit d25ffcfbfcfe8e059f2889873c9dbe76a28ad00c 147018Author: Peter Hutterer <peter@cs.unisa.edu.au> 147019Date: Thu Jun 5 15:29:35 2008 +0930 147020 147021 mi: protect against possible NULL-pointer dereference. 147022 147023commit 874dcdb3bdb9320aae252d3c7927d412685059d5 147024Author: Peter Hutterer <peter@cs.unisa.edu.au> 147025Date: Thu Jun 5 15:38:48 2008 +0930 147026 147027 dmx: remove a ifndef XINPUT, XINPUT is always defined now. 147028 147029commit 52752911ea8470dd8805f5e6c3febb9e1e1cc4b3 147030Author: Peter Hutterer <peter@cs.unisa.edu.au> 147031Date: Wed Jun 4 17:04:16 2008 +0930 147032 147033 dmx: don't free event list after use. 147034 147035commit 5c5e58124581b845a6bfa7a2991a4e0ee7f7ebdb 147036Author: Peter Hutterer <peter@cs.unisa.edu.au> 147037Date: Wed Jun 4 16:58:31 2008 +0930 147038 147039 dmx: add an XFlush. 147040 147041 Beats me why we need it, but without it we get segfaults lateron. 147042 147043commit 246c10441bbf62646b77993b55233d5ceb93a81f 147044Author: Peter Hutterer <peter@cs.unisa.edu.au> 147045Date: Wed Jun 4 16:37:25 2008 +0930 147046 147047 dmx: fix false memory allocation. 147048 147049 beNumVisuals and the number of GLX Visuals can be significantly different. 147050 147051commit 17cd26225749a1daf7460b44e52f43082399b265 147052Author: Peter Hutterer <peter@cs.unisa.edu.au> 147053Date: Wed Jun 4 16:05:36 2008 +0930 147054 147055 dmx: don't dereference a nullpointer. 147056 147057commit 75eb635e3543fcb731331d01f50e62b696967667 147058Author: Peter Hutterer <peter@cs.unisa.edu.au> 147059Date: Tue Jun 3 22:25:33 2008 +0930 147060 147061 dmx: add stubs for DeviceCursorInitialize, DeviceCursorCleanup 147062 147063commit 3ff2f3a00ef3c484d1d53569b6cbc1f25dc6bd35 147064Author: Peter Hutterer <peter@cs.unisa.edu.au> 147065Date: Tue Jun 3 22:17:14 2008 +0930 147066 147067 dmx: for now, don't acknowledge SDs. 147068 147069commit 5bcd9e8953dfb5f13e90baeebea1f6eac797644b 147070Author: Peter Hutterer <peter@cs.unisa.edu.au> 147071Date: Tue Jun 3 22:01:43 2008 +0930 147072 147073 dmx: learn about the existence of IsXExtensionPointer and IsXExtensionKeyboard 147074 147075commit 8da8a0fec4b1b9d9208635dedb2f449dc99e0004 147076Author: Peter Hutterer <peter@cs.unisa.edu.au> 147077Date: Tue Jun 3 21:59:33 2008 +0930 147078 147079 dmx: claim we support XI 2. 147080 147081 We don't really, yet, but at least we get the full device list this way. 147082 147083commit 6f1d5147cb394b7c83c4c1447c78890fa2b40a45 147084Author: Peter Hutterer <peter@cs.unisa.edu.au> 147085Date: Tue Jun 3 21:36:03 2008 +0930 147086 147087 dmx: fix a segfault caused by GC devPrivates never being initalised. 147088 147089commit d10ba4591a599a317093ffcbdd7cece11ab6de26 147090Author: Peter Hutterer <peter@cs.unisa.edu.au> 147091Date: Tue Jun 3 21:28:26 2008 +0930 147092 147093 dmx: some more build fixes. 147094 147095commit 8d4d0b47a07a298a20ffae9fefe96c8c7ca9dccc 147096Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 147097Date: Thu Jun 5 09:19:16 2008 +0930 147098 147099 gl: include assert.h if we're compiling with DEBUG. 147100 147101 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 147102 147103commit 8644aa47177bf20f3aa85181032ce75c08270fc5 147104Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 147105Date: Thu Jun 5 08:55:09 2008 +0930 147106 147107 mi: minor build fix when compiling with debug enabled. 147108 147109 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 147110 147111commit fbf4b5f16ac23359872361c8c95a617fe8c92292 147112Author: Peter Hutterer <peter@cs.unisa.edu.au> 147113Date: Wed Jun 4 11:10:46 2008 +0930 147114 147115 dix: set dst->mapWidth when allocating a new map. 147116 147117commit ff3adf3e564d94fea18e48f966de40a7ded1279e 147118Author: Peter Hutterer <peter@cs.unisa.edu.au> 147119Date: Wed Jun 4 11:10:11 2008 +0930 147120 147121 xkb: reset xkb_cached_map on CloseDownDevices. 147122 147123 Could lead to some invalid pointers in the second server generation. 147124 147125commit 40855d8000b9b9f4e6c6b4317542378ecf516345 147126Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147127Date: Wed Jun 4 12:19:28 2008 -0700 147128 147129 XQuartz: Removed async debugging sleep 147130 (cherry picked from commit 7812a8bdf9fab651ea5c07b852b2999547ec628d) 147131 147132commit ee86b751192b690973ee2a1446a406bc721ce8eb 147133Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147134Date: Wed Jun 4 12:01:19 2008 -0700 147135 147136 XQuartz: use a condition variable to signal when darwinEvents is ready rather than polling 147137 (cherry picked from commit ff1c443cadf11d12a7d939e51194f6105153870e) 147138 147139commit 38da26cd36957a45b2a47ef124282f7d863a9fd3 147140Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147141Date: Wed Jun 4 11:35:24 2008 -0700 147142 147143 XQuartz: Don't forget to destroy the mutex and cond after we're done with them 147144 (cherry picked from commit c3558bb8cd889e5b957190e9f5d23afad1e17b72) 147145 147146commit ea40fcf4344c6e918d3be34354528940b05a70f3 147147Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147148Date: Wed Jun 4 11:18:52 2008 -0700 147149 147150 XQuartz: Fork for trigger 147151 (cherry picked from commit dd0f8a0f59593d7831fe09a2a086fcd57c84910e) 147152 147153commit 2393dae6ffff7db8281a7977039450e8754cc19f 147154Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147155Date: Tue Jun 3 00:06:26 2008 -0700 147156 147157 XQuartz: Switched over to new startup path for testing. Cleaned it up a bit. Server still crashes when using the icon to launch =/ 147158 (cherry picked from commit 7f840e9dc180421eaa9b0ea3ab993fdd5b2466e5) 147159 147160commit 45b661c67aca1a728f1793dd94d701488f1f196c 147161Author: Peter Hutterer <peter@cs.unisa.edu.au> 147162Date: Wed Jun 4 13:39:39 2008 +0930 147163 147164 dix: Fix build with --disable-xinerama #16204 147165 147166 X.Org Bug 16204 <http://bugs.freedesktop.org/show_bug.cgi?id=16204> 147167 147168commit cba8b25a7e2608cd093283f86f08fba50a6a62f7 147169Author: Alan Hourihane <alanh@tungstengraphics.com> 147170Date: Mon Jun 2 11:17:47 2008 +0100 147171 147172 Bring inline with current modesetting-101 branch for connector/encoder changes 147173 147174commit 26e7e69ab893d1f2b35213250ada40ec90944f62 147175Author: Peter Hutterer <peter@cs.unisa.edu.au> 147176Date: Mon Jun 2 11:04:41 2008 +0930 147177 147178 kdrive: don't post motion event if there was no motion. #16179 147179 147180 Based on the patch by Tomas Janousek. 147181 147182 X.Org Bug 16179 <http://bugs.freedesktop.org/show_bug.cgi?id=16179> 147183 147184commit ac1db454497ff1839aa7f7ee0649ede112824ed9 147185Author: Peter Hutterer <peter@cs.unisa.edu.au> 147186Date: Mon Jun 2 10:40:10 2008 +0930 147187 147188 xfree86: suspend signals while removing a device (corrected version). 147189 147190 Block/Release is now symmetrical. 147191 147192commit 95ecaa411aefbcd8e2e00814132cb07b4be4e5bf 147193Author: Peter Hutterer <peter@cs.unisa.edu.au> 147194Date: Mon Jun 2 10:38:45 2008 +0930 147195 147196 Revert "xfree86: suspend signals while removing a device." 147197 147198 Left the signals blocked hanging after removing a master device. 147199 147200 This reverts commit 74372fd0049676787904f1f9f80f836b97f5f60e. 147201 147202commit 5ba07cb54803b8c63e25e8ca1f1787860f684612 147203Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 147204Date: Sun Jun 1 18:31:52 2008 -0300 147205 147206 kdrive: changes ephyr API according mpx merge. 147207 147208 This fix host-cursor in Xephyr server. 147209 147210 Signed-off-by: Tiago Vignatti <vignatti@c3sl.ufpr.br> 147211 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 147212 147213commit 5a3d06b8f42473cea3741dc722a775deaa2b73f6 147214Author: Peter Hutterer <peter@cs.unisa.edu.au> 147215Date: Sun Jun 1 15:20:42 2008 +0930 147216 147217 xkb: delete default rules when devices are closed. 147218 147219 We only have one set of default rules options in xkb. When the second keyboard 147220 is brought up with Xkb options specified, these new options overwrite the old. 147221 In future server generations, the rules used for the VCK are a mixture of the 147222 default ones and ones previously specified for other keyboards. Simply 147223 resetting the xkb default rules to NULL avoids this issue. 147224 147225 Reproducable by setting XkbLayout "de" and XkbVariant "nodeadkeys". In the 147226 second server generation, the VCK has "us(nodeadkeys)". This again produces a 147227 SIGABRT when the first key is hit. 147228 147229 I could not figure out why the SIGABRT happens. This patch is avoiding the 147230 issue rather than fixing it. 147231 147232commit 5ebe76f13fb6e96cebe5a386024a73eceeb828ff 147233Author: Peter Hutterer <peter@cs.unisa.edu.au> 147234Date: Sat May 31 22:13:03 2008 +0930 147235 147236 config: init dev to NULL to shut up a valgrind warning. 147237 147238commit a7ae9cbd78193fdafcafd5d7af1f6fabbb434628 147239Author: Peter Hutterer <peter@cs.unisa.edu.au> 147240Date: Sat May 31 20:44:58 2008 +0930 147241 147242 dix: don't read valuators past numAxes. 147243 147244commit f79ad5674ed41c3626e8055cd2bac0135de928c6 147245Author: Peter Hutterer <peter@cs.unisa.edu.au> 147246Date: Sat May 31 18:46:35 2008 +0930 147247 147248 kdrive: fix up callers to KdQueueEvent, need to pass xEvent, not EventListPtr 147249 147250commit 58d12c99abc4d06cb8b43386d4bcd126f0bedf83 147251Author: Peter Hutterer <peter@cs.unisa.edu.au> 147252Date: Sat May 31 18:45:11 2008 +0930 147253 147254 mi: Remove #ifdef XINPUT, kdrive otherwise won't get events. 147255 147256commit 942086fc965fc9a24be2db4f537f0ff54db51083 147257Author: Peter Hutterer <peter@cs.unisa.edu.au> 147258Date: Sat May 31 18:17:49 2008 +0930 147259 147260 dix: ProcQueryPointer doesn't need special grab handling, PickPointer does it 147261 147262commit 6cecae0e8651b1fa58b3cd0e0a636db46f56f6a6 147263Author: Peter Hutterer <peter@cs.unisa.edu.au> 147264Date: Sat May 31 13:12:53 2008 +0930 147265 147266 dix: Remove superfluous comment. 147267 147268 This code can handle devices with < 2 valuators now. 147269 147270commit 105d28652d1fb80dd8ce8511e2605dccc8812e99 147271Author: Peter Hutterer <peter@cs.unisa.edu.au> 147272Date: Fri May 30 15:20:35 2008 +0930 147273 147274 Xext: use GPE/GKE from XTestFakeInput #16145 147275 147276 This commit fixes two problems: 147277 1) XTFI used to assemble the event itself, then passed it to the device. It's 147278 much easier to just pass the variables into GPE/GKE and let the DIX do the 147279 rest. 147280 147281 2) XTFI would pass the VCP/VCK as default device to event processing. As a 147282 result, updating LEDs would be updated on the VCK, not on the actual keyboard. 147283 Instead, we now pass the events through the last-used SD, thus toggling the 147284 LEDs on the last keyboard that sent through this MD. 147285 147286 Also some cleanup in XTFI to merge validity checks a bit closer together 147287 rather than having several different sections. 147288 147289 This breaks XTestFakeMotion with Xinerama though. 147290 147291 X.Org Bug 16145 <http://bugs.freedesktop.org/show_bug.cgi?id=16145> 147292 147293commit 341a61b608873c77b1a17fe7d145ed3ad716e5a3 147294Author: Peter Hutterer <peter@cs.unisa.edu.au> 147295Date: Sat May 31 12:01:47 2008 +0930 147296 147297 mi: reduce noisyness. DebugF is enough if device ID can't be changed. 147298 147299 Also - printing out the event type is actually helpful. 147300 147301commit 62c1a32976d571f3ced1812b8b96ed5a8e658b4a 147302Author: Peter Hutterer <peter@cs.unisa.edu.au> 147303Date: Sat May 31 10:24:37 2008 +0930 147304 147305 dix: null out dummyDev, otherwise Xephyr dereferences random pointers. 147306 147307 SendEventToAllWindow eventually causes a IsInterferingGrab(), which attempts 147308 to dereference dev->deviceGrab.grab. 147309 147310commit 150c2f55a508ed24b230f68e30ec140c0901d9ae 147311Author: Julien Cristau <jcristau@debian.org> 147312Date: Fri May 23 03:09:04 2008 +0200 147313 147314 Add xkbstr.h and xkbsrv.h to sdk_HEADERS 147315 147316commit f9bcecbe51e4cf0cb2d82ad953f8c6741ed5558b 147317Author: Peter Hutterer <peter@cs.unisa.edu.au> 147318Date: Thu May 29 17:42:49 2008 +0930 147319 147320 Xext: only free events in case of core motion events. 147321 147322commit 000ead4503ddcfdf66ce9fffc572a3a741f4055a 147323Author: Alan Hourihane <alanh@tungstengraphics.com> 147324Date: Wed May 28 19:59:38 2008 +0100 147325 147326 Plug in hardware cursor support 147327 147328commit f9f7a872bf3f8522213aabb291d9d54e56d48c4b 147329Author: Alan Hourihane <alanh@tungstengraphics.com> 147330Date: Wed May 28 15:55:36 2008 +0100 147331 147332 Run indent 147333 147334commit 106bea5ad1bdd5795d6ed625fc6351a161bffa6e 147335Author: Alan Hourihane <alanh@tungstengraphics.com> 147336Date: Wed May 28 13:33:07 2008 +0100 147337 147338 Initial commit of new modesetting driver 147339 147340commit 8cf7755a85f5c1e5da7805c12b7511581a368782 147341Author: Peter Hutterer <peter@cs.unisa.edu.au> 147342Date: Wed May 28 18:05:25 2008 +0930 147343 147344 Require x11proto 7.0.13, xextproto 7.0.4 and inputproto 1.9.99.1 147345 147346commit 9b544fa40062b342ff3aa75f8f21d0bd4cec80ab 147347Author: Peter Hutterer <peter@cs.unisa.edu.au> 147348Date: Wed May 28 12:27:07 2008 +0930 147349 147350 mi: ansify. 147351 147352commit 64677f6a167a5bee99e5dfb599a3c3a6fb5dffe7 147353Author: Peter Hutterer <peter@cs.unisa.edu.au> 147354Date: Tue May 27 21:01:26 2008 +0930 147355 147356 xfree86: fix offset error in EDID parsing. #15964 147357 147358 X.Org Bug 15964 <http://bugs.freedesktop.org/show_bug.cgi?id=15964> 147359 147360commit c8af7ce35a900ac9b898f51c1b95dabad3ba1d76 147361Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 147362Date: Fri May 23 13:50:39 2008 -0300 147363 147364 Correct a NULL pointer deference 147365 147366 The problem happens if Monitor/Card combo doesn't provide EDID info, 147367 and the XFree86-VidModeExtension extension is used. 147368 147369 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 147370 147371commit 3a1278907ea8059546c7b9ab0c02714832c693e3 147372Author: Peter Hutterer <peter@cs.unisa.edu.au> 147373Date: Mon May 26 18:01:07 2008 +0930 147374 147375 xfree86: Bump XINPUT ABI to 3.0. 147376 147377commit 0050165a67bb462e0bf644a11644ad9d587c62bb 147378Author: Simon Thum <simon.thum@gmx.de> 147379Date: Mon May 26 17:56:08 2008 +0930 147380 147381 input: change dxremaind, dyremaind to floats for better acceleration #9156 147382 147383 X.Org Bug 9156 <http://bugs.freedesktop.org/show_bug.cgi?id=9156> 147384 147385 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 147386 147387commit 26c1958c322be3ac4dfec9ba2c41c5202bd03709 147388Merge: 5fe57787f 29586101d 147389Author: Peter Hutterer <peter@cs.unisa.edu.au> 147390Date: Mon May 26 17:37:31 2008 +0930 147391 147392 Merge branch 'master' into mpx 147393 147394commit 5fe57787faa2ea4f2f57b25455d200bf924206d6 147395Author: Peter Hutterer <peter@cs.unisa.edu.au> 147396Date: Mon May 26 17:23:34 2008 +0930 147397 147398 dix: store the absolute values in the motion history. 147399 147400 The core protocol requires absolute values and it's a bit hard to get them if 147401 we only have relative ones in the history. Switch the motion history to 147402 absolute, and if we really need the relative values, we can probably generated 147403 them from the abs. ones in the future. 147404 147405commit 1883485edd7eb90c0b76bca41c71e26ae2c0b91a 147406Author: Peter Hutterer <peter@cs.unisa.edu.au> 147407Date: Mon May 26 16:35:23 2008 +0930 147408 147409 If core motion history is required, scale back to screen coords and INT16. 147410 147411commit 0b88510069bdf6d020d73b424f62c2923c0db4a4 147412Author: Peter Hutterer <peter@cs.unisa.edu.au> 147413Date: Mon May 26 13:53:45 2008 +0930 147414 147415 Ensure the motion history is merged for master devices. 147416 147417 Add each event to the master's MH as well as to the SDs. In the MD, store 147418 min/max and the actual value. When retrieving the MH, rescale all coordinates 147419 to the current coordinate range and only post those valuators that are 147420 currently active on the device. 147421 147422commit d22c25bda450f6f1dfa634f0f72b32c4bec429cd 147423Author: Peter Hutterer <peter@cs.unisa.edu.au> 147424Date: Mon May 26 09:26:18 2008 +0930 147425 147426 dix: Allocate MD's motion history to be large enough for potential valuators 147427 147428 Since we can't predict how many valuators may be in a future SD attached to an 147429 MD, we need to preallocate a history buffer that is large enough to keep 147430 MAX_VALUATORS coordinates per event. 147431 147432 In addition, the history buffer needs to memorize the coordinate ranges at the 147433 time, thus requiring MDs to store (min_val, max_val, current_val, time) 147434 instead of (current_val, time) for each motion history entry. 147435 147436 This commit only fixes the allocation. 147437 147438commit 0877de13ac6ddfb55108aa3456d47f970c6c442c 147439Author: Peter Hutterer <peter@cs.unisa.edu.au> 147440Date: Sun May 25 22:49:54 2008 +0930 147441 147442 Remove GetMotionProc from ValuatorClassRec. 147443 147444 With the MD/SD device hierarchy we need control over the generation of the 147445 motion history as well as the conversion later before posting it to the 147446 client. So let's not let the drivers change it. 147447 147448 No x.org driver currently uses it anyway, linuxwacom doesn't either so dumping 147449 it seems safe enough. 147450 147451commit 00b4339168c10dd4ce026deb8e04bfb63dfd11dc 147452Author: Peter Hutterer <peter@cs.unisa.edu.au> 147453Date: Sun May 25 22:34:11 2008 +0930 147454 147455 dmx: remove some #if 0'd code. 147456 147457commit 096117cf023dc92dcadc91677931fc6ad1feb4a0 147458Author: Peter Hutterer <peter@cs.unisa.edu.au> 147459Date: Sun May 25 10:38:33 2008 +0930 147460 147461 Xi: reset motion history when swapping device classes. 147462 147463 This isn't really the correct solution, but it'll have to do until I figured 147464 out how to transfer the history over correctly. 147465 147466commit 29586101dc11d498b212510f8dedbfeca7f8c859 147467Author: Michel Dänzer <michel@tungstengraphics.com> 147468Date: Sat May 24 20:01:41 2008 +0200 147469 147470 EXA: Only record damage generated by rendering operations. 147471 147472 Recording damage from other operations (e.g. creating a client damage record) 147473 may confuse the migration code resulting in corruption. 147474 147475 Option "EXAOptimizeMigration" appears safe now, so enable it by default. Also 147476 remove it from the manpage, as it should only be necessary on request in the 147477 course of bug report diagnostics anymore. 147478 147479commit f6d61d3d86971d6a202b46ff2fab8c8799a4d057 147480Author: Michel Dänzer <michel@tungstengraphics.com> 147481Date: Sat May 24 20:01:36 2008 +0200 147482 147483 EXA: Don't migrate twice in exaImageGlyphBlt. 147484 147485 exaPrepareAccess already handles migration. 147486 147487commit d73304398255e0c3b03a497a8d4a2f8d900eef44 147488Author: Michel Dänzer <michel@tungstengraphics.com> 147489Date: Sat May 24 20:01:31 2008 +0200 147490 147491 EXA: Don't suggest exaCopyDirty be inlined. 147492 147493 Leave the decision to the compiler toolchain. 147494 147495commit 3baf3b42e079b4e7b61c1e20df305db0724d21f8 147496Author: Michel Dänzer <michel@tungstengraphics.com> 147497Date: Sat May 24 20:01:21 2008 +0200 147498 147499 EXA: Specify region used for source pixmap migration in exaCopyNtoN. 147500 147501 Avoids excessive migration overhead in some pathological cases. See 147502 http://bugs.freedesktop.org/show_bug.cgi?id=15845 . 147503 147504commit ba557e0263eee0b75847c596f31c71492611b22c 147505Author: Peter Hutterer <peter@cs.unisa.edu.au> 147506Date: Sat May 24 17:23:10 2008 +0930 147507 147508 Xi: remove deprecated comment. 147509 147510 Axis clipping wouldn't be necessary here, it's already performed in GPE. 147511 147512commit e6cbb1e11e5da1a8b9001853c25f4e5a052e7110 147513Author: Julien Cristau <jcristau@debian.org> 147514Date: Fri May 23 23:00:40 2008 +0200 147515 147516 xfree86: fix build on GNU/kFreeBSD 147517 147518 GNU/kFreeBSD defines __FreeBSD_kernel__, but not __FreeBSD__. 147519 Unify preprocessor conditionals between variable declaration and use. 147520 Debian bug #482550. 147521 147522commit 6c72961d8fa1ab1543f1b3e2cc7d34ff6d254bf8 147523Author: George Sapountzis <gsap7@yahoo.gr> 147524Date: Fri May 23 22:39:35 2008 +0300 147525 147526 glx: fix memory corruption with r5g6b5 147527 147528 should cherry-pick to xserver-1.5 147529 147530commit 21a8052fdca2115668dfc747b0b58db437b4eb6d 147531Author: George Sapountzis <gsap7@yahoo.gr> 147532Date: Fri May 23 22:39:09 2008 +0300 147533 147534 glx: fix forgotten swrast -> SWRast 147535 147536commit 877e6c35ff1f0f110627590ac0f12fddf47de506 147537Author: George Sapountzis <gsap7@yahoo.gr> 147538Date: Fri May 23 22:38:36 2008 +0300 147539 147540 glx: missing swrast is fatal 147541 147542commit 1345c93ad4be875951256dae87bc4fd9a1b5e08a 147543Author: George Sapountzis <gsap7@yahoo.gr> 147544Date: Fri May 23 22:38:11 2008 +0300 147545 147546 glx: drop stray glcore.h include 147547 147548commit ae0504d34d245f15b2f098a63ad1a9e1c4daef35 147549Author: George Sapountzis <gsap7@yahoo.gr> 147550Date: Fri May 23 22:40:03 2008 +0300 147551 147552 kdrive: need not export symbols for glx anymore 147553 147554commit 7082929ef7db20b507dce5a8cec043e787d0b55e 147555Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147556Date: Fri May 23 02:27:58 2008 -0700 147557 147558 XQuartz: A 2 second delay for launchd socket connections should be sufficient. 147559 (cherry picked from commit 2bb4251b3c6b30dbf1a556e1b51e6f03f02d2529) 147560 147561commit 5a388b27a094c71b4214aa744b2ba0fdda7fd607 147562Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147563Date: Fri May 23 02:10:14 2008 -0700 147564 147565 XQuartz: Add the launchd fd to AllSockets as well 147566 (cherry picked from commit 83f72529394be5871671d73b6ef4f8bc83708f8a) 147567 147568commit 77b688a73025238ed9fe286a2bae8f78da65fce3 147569Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147570Date: Fri May 23 02:32:28 2008 -0700 147571 147572 XQuartz: Don't need launchd checking in configure.ac anymore since its functionality has been removed from libxtrans and into the mach startup code. 147573 (cherry picked from commit 1e1f4fb717d2f3a3227be2a034d28d6a4f29b5d1) 147574 147575commit 225dddbaeded4d64fcc104da538449e5c4dedd2d 147576Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147577Date: Fri May 23 01:39:02 2008 -0700 147578 147579 XQuartz: Move the launchd display grabbing into mach_startup 147580 (cherry picked from commit c3866c98d23020d2151977ee1177b6054d05832e) 147581 147582commit 3a22190061c7ebefee85564b0def5e1112c07ae8 147583Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147584Date: Mon May 19 03:13:09 2008 -0700 147585 147586 XQuartz: Fixed a few issues with fd passing... still not working =( 147587 (cherry picked from commit 7dd351271522b475d8017e4bd1618f12817ee2fa) 147588 147589commit 3ae0f4fc832462444d03024c5fbb867f19d5e960 147590Author: Peter Hutterer <peter@cs.unisa.edu.au> 147591Date: Fri May 23 15:02:43 2008 +0930 147592 147593 dix: update valuators > 2 as well. 147594 147595 master->last.valuator[x] for x > 2 is undefined. For all other devices, it's 147596 the respective device's last valuators. 147597 147598 If the lastSlave did not have a valuator that is to be updated now, it is 147599 reset to 0. 147600 147601commit 0d1b41ace38bc918e2cc2a0a6ad107c76346fd00 147602Author: Peter Hutterer <peter@cs.unisa.edu.au> 147603Date: Fri May 23 13:32:33 2008 +0930 147604 147605 Xi: update device valuators based on the event. 147606 147607 Event is always absolute, update the device's valuators (always absolute too), 147608 and then change the deviceValuator event to reflect the device's reporting 147609 mode. 147610 147611commit 7eec1c23a34591064ea64499acb24f22ada08dfa 147612Author: Peter Hutterer <peter@cs.unisa.edu.au> 147613Date: Fri May 23 13:23:02 2008 +0930 147614 147615 dix: don't allow relative data in deviceValuators. 147616 147617 In GPE, we don't care about the device mode. Let's put the absolute values 147618 into the deviceValuator event and worry about relative valuators on the other 147619 side of the EQ. 147620 147621commit 75551a2ef949c2cfb53fdc4b97f3964a5e2f45f5 147622Author: Peter Hutterer <peter@cs.unisa.edu.au> 147623Date: Fri May 23 12:55:50 2008 +0930 147624 147625 Revert "dix: Correct values in the device axisVal are calculated in GPE" 147626 147627 We can't rely on GPE to update device->valuators->axisVal. If a SIGIO occurs 147628 during event processing this may have incoherent results. 147629 147630 This reverts commit f6645ddbf754c80e9a8b1672519534a887622270. 147631 147632commit fe59b1a62b2c7f467d20132db4913317bbf308eb 147633Author: Peter Hutterer <peter@cs.unisa.edu.au> 147634Date: Fri May 23 12:51:52 2008 +0930 147635 147636 dix: more device scaling insanity. 147637 147638 Assuming master->last.valuators is in screen coords, SD's are always in device 147639 coordinates. 147640 147641 1. If an event comes in, scale masters->last to the device, drop into device's 147642 last->valuators. 147643 2. Apply motion from the actual event 147644 3. Scale back to screen coords, check if we may need to cross screens 147645 4. Drop screen coords into master->last 147646 5. Rescale to device coords, drop into deviceValuator event and SD->last 147647 6. Drop screen coords into ev->root_x/y 147648 147649 Whoopee... 147650 147651commit 6c9e9f8a40e20fb1761440acd2755f5fd31f4d44 147652Author: Peter Hutterer <peter@cs.unisa.edu.au> 147653Date: Fri May 23 11:51:53 2008 +0930 147654 147655 input: instead of lastx/y, use a last.valuators[] array on the device. 147656 147657 During GetPointerEvents (and others), we need to access the last coordinates 147658 posted for this device from the driver (not as posted to the client!). Lastx/y 147659 is ok if we only have two axes, but with more complex devices we also need to 147660 transition between all other axes. 147661 147662 ABI break, recompile your input drivers. 147663 147664commit fb146cbb0f28e4e480e5d16d61476ac46b5d00ce 147665Author: Peter Hutterer <peter@cs.unisa.edu.au> 147666Date: Fri May 23 12:00:13 2008 +0930 147667 147668 dix: coreEvents doesn't have meaning here, use master instead. 147669 147670 We mustn't clip x/y if we are attached, otherwise we can't change screens. 147671 147672commit fc1cc0adcb91fdfa4c547bf379a85f9558c959ff 147673Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 147674Date: Fri May 23 00:33:18 2008 +0200 147675 147676 dix: Cleanup of GetPointerEvents 147677 147678 Changed all the checks for x&y valuator so the more complex 147679 calculation is only made once. 147680 Added TODOs for valuator/axis 2 and above for future correct 147681 handling of relative reporting of these. 147682 147683 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 147684 147685commit a0241d5380bb5d8b10865f8ea81a9a011de4aaf1 147686Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 147687Date: Fri May 23 00:36:11 2008 +0200 147688 147689 dix: Correct clipAxis so it can handle devices with value ranges properly 147690 147691 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 147692 147693commit f6645ddbf754c80e9a8b1672519534a887622270 147694Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 147695Date: Thu May 22 23:33:45 2008 +0200 147696 147697 dix: Correct values in the device axisVal are calculated in GPE 147698 147699 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 147700 147701commit 7f85acdf70c67c567de688439e25081be5a7d5df 147702Author: Peter Hutterer <peter@cs.unisa.edu.au> 147703Date: Thu May 22 23:27:15 2008 +0930 147704 147705 dix: fill valuators with the correct values depending on the device mode (GPE) 147706 147707 valuators[] is passed from the DDX. Depending on the device mode, update it 147708 with either absolute values or relative values. The deviceValuator event sent 147709 to the client will then contain the respective values. 147710 147711commit 1a3f351c50cba66f71a73239318174b09fd9b63b 147712Author: Peter Hutterer <peter@cs.unisa.edu.au> 147713Date: Thu May 22 23:25:20 2008 +0930 147714 147715 dix: when floating a device, create a new cursor sprite. 147716 147717 This is essentially necessary to allow calls to miPointerGetSprite etc. to 147718 work for floating slave devices. 147719 147720commit e0fbe404a436aef24624a3a15e8405a9ca38aadb 147721Author: Peter Hutterer <peter@cs.unisa.edu.au> 147722Date: Thu May 22 23:24:29 2008 +0930 147723 147724 mi: handle sprite even for floating slave devices. 147725 147726 We still don't render it, but we accept all the other calls to update it's 147727 internal state. 147728 147729commit 0f15875a271889ae3cc4997ad15f787ea28b3a08 147730Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 147731Date: Sat May 17 19:24:00 2008 +0200 147732 147733 Make all conversion handling in GPE. 147734 147735 This isn't quite finished yet, but at least it gives us the ability to use a 147736 tablet as a normal mouse - with all the scaling in place. 147737 147738 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 147739 147740commit 74372fd0049676787904f1f9f80f836b97f5f60e 147741Author: Peter Hutterer <peter@cs.unisa.edu.au> 147742Date: Thu May 22 18:02:10 2008 +0930 147743 147744 xfree86: suspend signals while removing a device. 147745 147746 Getting a keyboard event while halfway through freeing memory can be 147747 unpleasant. 147748 147749commit 30e9a33f7d1972dcf1c29c7455b2cea4e5857913 147750Author: Peter Hutterer <peter@cs.unisa.edu.au> 147751Date: Thu May 22 16:00:21 2008 +0930 147752 147753 xkb: fix crash caused by uninitialised variable. 147754 147755commit e837244947fbb18192e9056c070ce439ef24d1a3 147756Author: Dan Nicholson <dbn.lists@gmail.com> 147757Date: Wed May 21 16:14:55 2008 -0700 147758 147759 Actually enable building GLX without Mesa source 147760 147761commit ea4ec9e9983e25d94a3edf8a77ed2ad1db193284 147762Author: Dave Airlie <airlied@redhat.com> 147763Date: Tue May 20 13:59:09 2008 +1000 147764 147765 int10: add pci_device_enable support on Linux 147766 147767commit 4230b92744d54528f48ddc3b096a265b488053c9 147768Author: Kristian Høgsberg <krh@redhat.com> 147769Date: Wed May 21 17:58:15 2008 -0400 147770 147771 Clean up remaining afb references. 147772 147773commit 20ea99c655140e101f2d20cfab78fb22765fec62 147774Author: Adam Jackson <ajax@redhat.com> 147775Date: Wed May 21 17:03:25 2008 -0400 147776 147777 Death to afb. 147778 147779 Planar framebuffer people, please talk to wfb for all your 1988 hardware 147780 needs. 147781 147782commit c76fddeea28b4f88db97a2a3487d7a57cd9d578c 147783Author: Dan Nicholson <dbn.lists@gmail.com> 147784Date: Wed May 21 10:51:15 2008 -0700 147785 147786 Fix up paths for GL/glx -> glx renaming 147787 147788commit c3eb5b80d8cc7ca6a425ccf359718d1ef3d5af55 147789Author: Kristian Høgsberg <krh@redhat.com> 147790Date: Wed May 21 13:33:36 2008 -0400 147791 147792 Move GL/glx on level up now that it's the only thing left under GL. 147793 147794commit a8a9b40e554dd0c184866978889a39519afc466a 147795Author: Kristian Høgsberg <krh@redhat.com> 147796Date: Wed May 21 12:34:07 2008 -0400 147797 147798 Move the windows GL files to the xwin DDX. 147799 147800commit e5aad4bc087036489e2af576628aa9b612054a6e 147801Author: Kristian Høgsberg <krh@redhat.com> 147802Date: Wed May 21 11:44:12 2008 -0400 147803 147804 Get out of the mesa symlinking business. 147805 147806 This copies over the files generated from mesa/src/mesa/glapi. There's 147807 a corresponding mesa commit that makes it easy to generate the glapi files 147808 straight into the xserver tree when the XML definitions change. 147809 147810 The only few files that are copied from mesa but aren't generated are 147811 glapi.[ch] and glthread.[ch]. Everything in there is technically DRI 147812 driver API and the whole setup is still a bit fragile, but it's not a new 147813 problem. 147814 147815 The --with-mesa-source configure option is still around since other 147816 parts of the server (XGL and DMX - grep for MESA_SOURCE) need that, 147817 but for common case of building with GLX and AIGLX support, that 147818 option is no longer needed. 147819 147820commit f558c7d4d1d87affe1ce4aec2e62ff64545f2bc8 147821Author: Kristian Høgsberg <krh@redhat.com> 147822Date: Wed May 21 10:18:16 2008 -0400 147823 147824 Use LD_EXPORT_SYMBOLS_FLAG when linking Xnest. 147825 147826commit 77ff8fd5072fa57158463bfd84656389eeb7b503 147827Author: Kristian Høgsberg <krh@redhat.com> 147828Date: Wed May 21 10:17:53 2008 -0400 147829 147830 Drop glcore GLX provider. 147831 147832 Obsoleted by the DRI swrast driver. 147833 147834commit 61eaef22359ae2abcee7dcd73ee8610a83776fc7 147835Author: Kristian Høgsberg <krh@redhat.com> 147836Date: Tue May 20 16:56:27 2008 -0400 147837 147838 Stop symlinking glcontextmodes.[ch] from mesa. 147839 147840commit 88f43e5d06fb5543bc40ca060b2c5d8e2673afad 147841Author: George Sapountzis <gsap7@yahoo.gr> 147842Date: Wed May 21 18:13:47 2008 +0300 147843 147844 glx: drop drm.h, sareaHandle from swrast loader 147845 147846commit 255142b61eb73843b16c2afdb95aa4d1f27e4f0d 147847Author: George Sapountzis <gsap7@yahoo.gr> 147848Date: Thu Jan 3 17:53:56 2008 +0200 147849 147850 drop xprint remnants: InitGlobals 147851 147852commit 641ce9c706ce3cbf726fb3fc9623161bb96ba127 147853Author: George Sapountzis <gsap7@yahoo.gr> 147854Date: Thu Jan 3 17:28:04 2008 +0200 147855 147856 drop xprint remnants: xpstubs 147857 147858commit 5c5a1eaf68eb11053fd5b99d24eb8024b6d5b218 147859Author: Peter Hutterer <peter@cs.unisa.edu.au> 147860Date: Wed May 21 21:51:27 2008 +0930 147861 147862 Replace UniSA with full Uni name in Author affiliation. 147863 147864 Big boss says UniSA isn't unique enough. Who am I to argue? 147865 147866commit 7509fb498c02fe1cebe4139612b8871ec877c130 147867Merge: 2ae3bed33 dfb7de657 147868Author: Peter Hutterer <peter@cs.unisa.edu.au> 147869Date: Wed May 21 16:25:35 2008 +0930 147870 147871 Merge branch 'hwcursor' into mpx 147872 147873commit 2ae3bed33731d800bb28a968e3a1c8066d179740 147874Author: Peter Hutterer <peter@cs.unisa.edu.au> 147875Date: Wed May 21 15:51:35 2008 +0930 147876 147877 mi: shut up compiler warning (return w/o a value) 147878 147879commit 8b3d26f5b6caff1766669deb0e2100d2dee3f185 147880Author: Aaron Plattner <aplattner@nvidia.com> 147881Date: Tue May 20 12:39:28 2008 -0700 147882 147883 Short-circuit PanoramiXTranslateVisualID after verifying that the visual actually exists first. 147884 147885 This allows using PanoramiXTranslateVisualID to test whether a given visual made 147886 it through PanoramiXVisualConsolidate. 147887 147888commit 6ff6465931c397f72db27a4fd0ca0dcac0609537 147889Author: George Sapountzis <gsap7@yahoo.gr> 147890Date: Tue May 20 16:36:22 2008 -0400 147891 147892 Add loader for the swrast dri driver. 147893 147894commit 4056595fc77c20052e226b402e38a2a914dca123 147895Author: Peter Hutterer <peter@cs.unisa.edu.au> 147896Date: Tue May 20 18:57:31 2008 +0930 147897 147898 xkb: remove superfluous checks in if statement. 147899 147900commit 925e895b869e461a9e7f135891463c56ee633cd6 147901Author: Alan Coopersmith <alan.coopersmith@sun.com> 147902Date: Mon May 19 18:43:29 2008 -0700 147903 147904 Restore return type on xf86SetScrnInfoModes 147905 147906 Seems to have been accidentally lost by commit 76943fec860315f3c93539e59a59080b8a7b3e75 147907 147908commit 7a550cefd9417c22a4397ea4d103ddc347ab1a0f 147909Author: Peter Hutterer <peter@cs.unisa.edu.au> 147910Date: Tue May 20 11:07:39 2008 +0930 147911 147912 dix: shut up two compiler warnings. 147913 147914commit f9edecd1b53a2e234def9cbd954a47c4bda2bebc 147915Author: Peter Hutterer <peter@cs.unisa.edu.au> 147916Date: Tue May 20 11:07:17 2008 +0930 147917 147918 mi: sync declaration and definition of mieqResizeEvents. 147919 147920commit e73a03d589b47ac9131f3932da7f78d2fef7bfb5 147921Author: Peter Hutterer <peter@cs.unisa.edu.au> 147922Date: Tue May 20 10:54:32 2008 +0930 147923 147924 fb: shut up two compiler warnings. 147925 147926commit 99d28c3ef37aeffe7d8ec41e45a650ba248c6958 147927Merge: fc7e2566c 0178b6a4a 147928Author: Peter Hutterer <peter@cs.unisa.edu.au> 147929Date: Tue May 20 10:20:14 2008 +0930 147930 147931 Merge branch 'master' into mpx 147932 147933 Conflicts: 147934 147935 Xext/xprint.c (removed in master) 147936 config/hal.c 147937 dix/main.c 147938 hw/kdrive/ati/ati_cursor.c (removed in master) 147939 hw/kdrive/i810/i810_cursor.c (removed in master) 147940 hw/xprint/ddxInit.c (removed in master) 147941 xkb/ddxLoad.c 147942 147943commit 0178b6a4abed0df3e90ba393709ed566105e7c2c 147944Author: Adam Jackson <ajax@redhat.com> 147945Date: Mon May 19 11:22:19 2008 -0400 147946 147947 Fix reduced-blanking mode filtering in RANDR 1.2. 147948 147949 If the monitor isn't reduced-blanking (either through EDID logic, or 147950 config file setting), then remove RB modes from the default pool. Any 147951 RB modes from the driver and config file pools will stick around though; 147952 you asked for them, you got them. 147953 147954commit bb687465d0c5aef3cc1c865e6ea67c01a8b417fb 147955Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147956Date: Mon May 19 02:45:47 2008 -0700 147957 147958 XQuartz: First stab at SCM_RIGHTS passing the $DISPLAY launchd fd from the stub to server 147959 (cherry picked from commit cccee9cfc29f85cca635df3b8dd54199b45c4df8) 147960 147961commit 4b69d22bcb79210225379da2cefb0cd8d5ffa10e 147962Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147963Date: Sun May 18 23:51:06 2008 -0700 147964 147965 XQuartz: Fixed dropped code in the !XKB blocks, disable XKB support until we figure out a solution for xkeyboard-config 147966 (cherry picked from commit 2a72309c061f7060480d150791019ce232481462) 147967 147968commit e997df8cd3c4222ea9fc3f0ebda728c4d1f15df5 147969Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147970Date: Sun May 18 23:31:41 2008 -0700 147971 147972 XQuartz: Cleaned up ListenOnOpenFD... 147973 (cherry picked from commit 6fb587d3d5fbbaee9e46cdce24d03e5d1c66d58a) 147974 147975commit 738672858d9399ecbc170500c15f90bf657502d2 147976Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147977Date: Sun May 18 23:02:12 2008 -0700 147978 147979 New XQuartz icon thanks to Simone Karin Lehmann 147980 (cherry picked from commit c27e0c8f3c5108192362735db8a80bf3a9a1cc3a) 147981 147982commit 4be2104c801340075e030d06d24ceb4d8a6d6549 147983Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 147984Date: Thu May 15 12:26:32 2008 -0700 147985 147986 XQuartz: More work towards Mach-IPC startup... started working on FD passing 147987 (cherry picked from commit 27fbfeacfa16d620801ac3492b25c1e50aec8588) 147988 147989commit 82c4075d4ba3bea03341c80b754b0f9d2f62a826 147990Author: Julien Cristau <jcristau@debian.org> 147991Date: Mon May 19 03:15:11 2008 +0200 147992 147993 xfree86: fix initial output positions with Above or LeftOf and rotation 147994 147995 The fix in fa19e84714aa84a2f2e817e363d6440349d0b619 was incomplete, 147996 as it was still using the wrong output's initial rotation to compute 147997 the position. 147998 147999commit 7dca84f3ee7265119fb81d598d7d2f7363e25f1f 148000Author: Dave Airlie <airlied@redhat.com> 148001Date: Mon May 19 11:16:08 2008 +1000 148002 148003 cursor: don't dereference NULL pointer is devPrivates not yet set. 148004 148005 This fixes a bug on server recycle with ati zaphod. 148006 148007commit 5af5db5033582a84b616d5c50f4288adb0210459 148008Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148009Date: Sat May 17 14:51:09 2008 -0700 148010 148011 XQuartz: Ok, pass XQUARTZ_USE_XKB since it breaks worse without it... but we have issues when we have the keyboard configs installed... need to figure out what to do there... 148012 (cherry picked from commit 301262b07024ad960f22d99a1267fe137f5c3fce) 148013 148014commit 2408303d79297385063cae557195bd5fd3698478 148015Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148016Date: Sat May 17 14:56:53 2008 -0700 148017 148018 XQuartz: Added functionality to add a file descriptor to the connection list after the server is already running. 148019 (cherry picked from commit 543c2cd68d1ffef65d4644b860faad7191c6b9da) 148020 148021commit 01612fe612aa27262fc3c8167f52e0376941f1ef 148022Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148023Date: Thu May 15 07:08:21 2008 -0700 148024 148025 XQuartz: Disable xkb since it doesn't work after getting xkeyboard-config installed 148026 148027 Need to setup configs for the quartz keyboard 148028 (cherry picked from commit c28fecc621b1803a4d4536afbc724d141de9e6ee) 148029 148030commit 1d09deaa6e5a75420ea4f24f6b5533f0ebc3ba28 148031Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148032Date: Thu May 15 05:49:43 2008 -0700 148033 148034 XQuartz: Made DarwinSendDDXEvent a little more robust to context switching diring server init. 148035 (cherry picked from commit 5626b0949b1a6bdd2fa3ec2ef53a688084349b87) 148036 148037commit adc62f62b9f08aa6633718989ab61b61e3c1e8e2 148038Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148039Date: Wed May 14 01:33:43 2008 -0700 148040 148041 Xquartz: Xquartz stub now can send Mach IPC to start the server. 148042 (cherry picked from commit 55a3bca8ea42f8677897f94bf0a429bdc08c6696) 148043 148044commit 93c082a060bfa8d2a204c0c57ec3507a8edea087 148045Author: Brad Smith <brad@comstyle.com> 148046Date: Sat May 17 13:14:31 2008 +0200 148047 148048 OpenBSD doesn't define __x86_64__, simplify tests. 148049 148050 Seeing as this code seems to be specific to OpenBSD I don't think 148051 __x86_64__ should have been added there at all. It appears to have 148052 been added wherever __amd64__ existed before which is wrong. I 148053 think that part of the commit should be reverted but also all four of 148054 the checks should be __OpenBSD__ && __amd64__ instead of two one 148055 direction and two flipped. 148056 148057commit 895d4e274d2b4638f0e6838dfee0e543311dba34 148058Author: Brad Smith <brad@comstyle.com> 148059Date: Sat May 17 13:05:58 2008 +0200 148060 148061 BSD's also have /dev/urandom. 148062 148063commit aad1c37b0951eae216ac323c5d8bfc6fbcf096bd 148064Author: Adam Jackson <ajax@redhat.com> 148065Date: Fri May 16 10:52:41 2008 -0400 148066 148067 RANDR 1.1 compat: remove senseless comparison against the virtual size. 148068 148069commit 14726b776d6cebb7d864b6ffa7554e1ce5637d5c 148070Author: Adam Jackson <ajax@redhat.com> 148071Date: Fri May 16 10:51:32 2008 -0400 148072 148073 xf86SetDesiredModes(): Skip disabled CRTCs first thing. 148074 148075commit 459f34b089aca4f4eee9752600c3a9e4f4e343ab 148076Author: Adam Jackson <ajax@redhat.com> 148077Date: Fri May 16 10:48:00 2008 -0400 148078 148079 Fix initial mode selection even harder. 148080 148081 The first guess used to be "is the preferred mode for one output the 148082 preferred mode on all outputs". Instead, do "find the largest mode that's 148083 preferred for at least one output and available on all outputs". 148084 148085commit 96111c154713600dd534dd82104ac18b91466202 148086Author: Adam Jackson <ajax@redhat.com> 148087Date: Fri May 16 10:31:58 2008 -0400 148088 148089 Redo RANDR compatibility output selection. 148090 148091 Old logic was just the first one that happened to have an associated 148092 CRTC. The new logic tries to find one that's definitely connected, has 148093 probed modes, and has the largest candidate mode. 148094 148095commit a4bbe1c8bca08f3df5ff7e50444af6aef7ec8b25 148096Author: Adam Jackson <ajax@redhat.com> 148097Date: Fri May 16 10:25:12 2008 -0400 148098 148099 Re-add sync range inference from legacy setup to RANDR 1.2. 148100 148101commit f52f6c5c7efc281f9ac204fbaa4f71383df7463d 148102Author: Adam Jackson <ajax@redhat.com> 148103Date: Mon May 5 14:37:07 2008 -0400 148104 148105 Fix hal shutdown crash. 148106 148107 Removing the device invalidates its ->next pointer. Copy it aside before 148108 destroying the device. 148109 148110commit 99e4531f7c3fafce065dfd21f65bad1cf190c3f1 148111Author: Adam Jackson <ajax@redhat.com> 148112Date: Fri May 16 08:16:01 2008 -0400 148113 148114 The great kdrive purge. 148115 148116 Most of these drivers didn't work. ati was the only one that even came 148117 close. The igs, ipaq, itsy, pcmcia, savage, sis530, trident, trio, ts300, 148118 and vxworks directories have never built since modularisation, so clearly 148119 no one can miss them. 148120 148121commit fc7e2566cc076c0d979f74871bc436df43401058 148122Author: Peter Hutterer <peter@cs.unisa.edu.au> 148123Date: Fri May 16 16:59:34 2008 +0930 148124 148125 Xext: fix typo in condition. 148126 148127 Fall-out from dc3aba8a559d4304844ee1cc306c577a63b82762. 148128 We must free the event if it is NOT an extension event. 148129 148130commit 6fb76acc4524d6af4bfb6bc6d862c1ee2bbb8baa 148131Author: Peter Hutterer <peter@cs.unisa.edu.au> 148132Date: Fri May 16 19:22:32 2008 +0930 148133 148134 mi: remove miPointerUpdate() -> replaced by miPointerUpdateSprite(). 148135 148136 In the xnest and vfb DDX we don't need the call anyway, it's performed by 148137 mieqProcessInputEvent. 148138 148139commit dfb7de6571345061d1cd88ab915064fc376529fd 148140Author: Peter Hutterer <peter@cs.unisa.edu.au> 148141Date: Fri May 16 20:56:04 2008 +0930 148142 148143 mi: don't call from miPointerMoved into the spriteFuncs for anything but VCP. 148144 148145 Calling ->MoveCursor for anything but the HW-rendered VCP causes the 148146 SW-rendered cursor routines to be started, including mallocs, etc. Since 148147 miPointerMoved is called during SIGIO, this is a bad idea. 148148 148149commit f611719edce2ae51dca8e53947a3938b53c20b00 148150Author: Peter Hutterer <peter@cs.unisa.edu.au> 148151Date: Fri May 16 19:49:29 2008 +0930 148152 148153 xfree86: always render first cursor in HW, all others in SW. 148154 148155commit 48ba58feacedf9781db81c82adbb37197d9a5cd1 148156Author: Peter Hutterer <peter@cs.unisa.edu.au> 148157Date: Fri May 16 19:22:32 2008 +0930 148158 148159 mi: remove miPointerUpdate() -> replaced by miPointerUpdateSprite(). 148160 148161 In the xnest and vfb DDX we don't need the call anyway, it's performed by 148162 mieqProcessInputEvent. 148163 148164commit 622d7c1d899a6146773a2ebd1d632a805f24025e 148165Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 148166Date: Wed May 14 14:49:09 2008 -0300 148167 148168 Restructure and organize the code. 148169 148170 It was removed and simplified some conditionals. We don't need test for 148171 pDev->isMaster inside xf86CursorSetCursor() because only MD enters there. 148172 148173 In the last chunk, ScreenPriv fields were being assigned without need, so 148174 that code was wrapped inside the conditional to avoid it. 148175 148176 I also tried to make the identation more sane in some parts that I touched. 148177 148178 Signed-off-by: Tiago Vignatti <vignatti@c3sl.ufpr.br> 148179 148180 Minor modification, part of the original patch led to cursors not being 148181 updated properly when controlled through XTest. 148182 148183 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 148184 148185commit 5af90025fee6a92e7d94642978feed21f902d0cb 148186Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 148187Date: Wed May 14 02:45:49 2008 -0300 148188 148189 Remove redundancy. 148190 148191 The only function that cat set SWCursor before xf86DeviceCursorInitialize() 148192 is xf86InitCursor() when VCP and is created. 148193 148194 Signed-off-by: Tiago Vignatti <vignatti@c3sl.ufpr.br> 148195 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 148196 148197commit 9fc94edfe3df4c4a84ad70714c0a4ef8bbf57fc9 148198Author: Peter Hutterer <peter@cs.unisa.edu.au> 148199Date: Fri May 16 16:59:34 2008 +0930 148200 148201 Xext: fix typo in condition. 148202 148203 Fall-out from dc3aba8a559d4304844ee1cc306c577a63b82762. 148204 We must free the event if it is NOT an extension event. 148205 148206commit 5127942f80983b2e053dddd5c5747d0c3d2f9d6d 148207Author: Brian Rogers <brian_rogers@comcast.net> 148208Date: Fri May 16 10:08:04 2008 +0930 148209 148210 DGA: pass nevents to UpdateDeviceState. #15936 148211 148212 Missing parameter caused event processing to go nuts when checking valuators. 148213 148214 X.Org Bug 15936 <http://bugs.freedesktop.org/show_bug.cgi?id=15936> 148215 148216 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 148217 148218commit 28378d26b4bae377ef1212f6a51cda9b5529f1b5 148219Author: Peter Hutterer <peter@cs.unisa.edu.au> 148220Date: Thu May 15 11:09:15 2008 +0930 148221 148222 Xi: assemble button/modifier state before updating the device. #15934 148223 148224 The state field of the event must specify the state of the devices before the 148225 event occured. With the code as it was, the state would also include the 148226 event (e.g. state from a button press event would show the button as pressed) 148227 Gathering the state before updating the device should fix this. 148228 148229 X.Org Bug 15934 <http://bugs.freedesktop.org/show_bug.cgi?id=15934> 148230 148231commit 7cabf81c8638739a15a1be6baa3fc569f38e7589 148232Author: Eric Anholt <eric@anholt.net> 148233Date: Thu May 15 13:24:27 2008 -0700 148234 148235 Add necessary include for COMPILEDDISPLAYCLASS. 148236 148237 Fixes build after 07f69023b0e75d98a4be8f0d1c3dd9391b8149a0. 148238 148239commit 07f69023b0e75d98a4be8f0d1c3dd9391b8149a0 148240Author: Adam Jackson <ajax@redhat.com> 148241Date: Wed May 14 16:07:51 2008 -0400 148242 148243 Move defaultDisplayClass to xdmcp-only visibility. 148244 148245commit 166177e36bfc7bd1df03663ad89500cff056d563 148246Author: Adam Jackson <ajax@redhat.com> 148247Date: Wed May 14 16:02:05 2008 -0400 148248 148249 Simplify default font path setup. 148250 148251commit 2efe1abb6fc786a4f86464978ae0d23fc7347a98 148252Author: Adam Jackson <ajax@redhat.com> 148253Date: Wed May 14 15:53:11 2008 -0400 148254 148255 Remove global argc/argv variables. 148256 148257 Was only used in the logging code anymore, and uselessly so there. 148258 148259commit 6c27b911222cdee9a057de0be7e8b2eff8b1ff2a 148260Author: Adam Jackson <ajax@redhat.com> 148261Date: Wed May 14 15:44:04 2008 -0400 148262 148263 Restore compression of duplicate log file entries. 148264 148265 This undoes the workaround for bug #964, which was an Xprintism. 148266 148267commit 937e5aae33d3b5112b5d10d605e25f57b48caa3f 148268Author: Peter Hutterer <peter@cs.unisa.edu.au> 148269Date: Wed May 14 22:30:15 2008 +0930 148270 148271 dix: don't undisplay cursor if we don't own a sprite. 148272 148273 Sometimes we didn't have a cursor when coming back from suspend. Reason was 148274 that the suspend caused the server to lose the device that was attached to the 148275 VCP, and a RemoveDevice() would then set the cursor to NULL. 148276 Solution: only set the cursor to NULL if we actually own the sprite. 148277 148278commit cbb4e80eb7cc8dc2aff5e5268b49cdb7b3c3d5fa 148279Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148280Date: Wed May 14 01:13:15 2008 -0700 148281 148282 XQuartz: More work on the Mach-IPC startup path 148283 (cherry picked from commit 49cd0b185fd6c99b07357a74734b6a4023faca84) 148284 148285commit 3b57c59bb08c9a3211f4ae57d9e2fb569d61bf2f 148286Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148287Date: Tue May 13 10:40:20 2008 -0700 148288 148289 Xquartz: More work on the new Mach startup 148290 (cherry picked from commit 6237acf75d3310d7d4f262556b677557c2907284) 148291 148292commit e435acc84cb9477455ad005cee658630cbd363a2 148293Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148294Date: Mon May 12 17:57:07 2008 -0700 148295 148296 Added missing to EXTRA_DIST 148297 (cherry picked from commit e39613f4633ed992bc276b70833a703560e528f9) 148298 148299commit f7d6d20ad64b235700185784d317417a94d1814b 148300Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148301Date: Mon May 12 13:36:35 2008 -0700 148302 148303 XQuartz: Cleaned up the about box. 148304 (cherry picked from commit 0279a5970694937e949ba533330ea48961c4edba) 148305 148306commit b37e1f1f5ccc5a48df665449b0e31c4d25cc323c 148307Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148308Date: Mon May 12 11:34:06 2008 -0700 148309 148310 Don't need the fink-friendly printf in the DDX anymore. 148311 (cherry picked from commit fe2279440450c795d67ba5a2234b0797d0bfe39c) 148312 148313commit 68b4f250eef441a3d75e3b9b2665a51d3a1538d6 148314Author: Peter Hutterer <peter@cs.unisa.edu.au> 148315Date: Wed May 14 16:18:48 2008 +0930 148316 148317 dmx: fix build. 148318 148319 Just because it builds, doesn't mean it works. 148320 148321commit 00db0f35acc00cb771ab1c23f35b0db1c25d81bb 148322Author: Peter Hutterer <peter@cs.unisa.edu.au> 148323Date: Wed May 14 15:29:28 2008 +0930 148324 148325 dix: InitAndStartDevices doesn't need a window parameter. 148326 148327 Leftover from the old pointer-keyboard pairing code, obsolete and gone now. 148328 148329commit da728f2127aaa5c1ed5d22bb9f3d0c8ff19e4933 148330Author: Peter Hutterer <peter@cs.unisa.edu.au> 148331Date: Wed May 14 15:17:05 2008 +0930 148332 148333 dix: don't use serverClient in AddInputDevice, use the one supplied instead. 148334 148335 We pass in the client that wants to create the device anyway, lets use the 148336 parameter instead of hardcoding the serverClient. 148337 148338 Wow. I hope this is merge detritus, otherwise it'd be a sign that I didn't 148339 have enough coffee that day. 148340 148341commit a6659291c798f417a76d9aa7944694ff27c01fa7 148342Author: Peter Hutterer <peter@cs.unisa.edu.au> 148343Date: Wed May 14 09:42:50 2008 +0930 148344 148345 mi: protect against NULL-pointer dereference. 148346 148347 Required by the recent patch to use a NullCursor instead of UndisplayCursor(). 148348 148349commit 7cdc19b29d93bf15cecfd6b69e269fab2501bca0 148350Author: Alan Coopersmith <alan.coopersmith@sun.com> 148351Date: Tue May 13 16:39:30 2008 -0700 148352 148353 When XKB fails to open rules file, log the file name, not the NULL file pointer 148354 148355commit dc3aba8a559d4304844ee1cc306c577a63b82762 148356Author: Peter Hutterer <peter@cs.unisa.edu.au> 148357Date: Tue May 13 14:37:26 2008 +0930 148358 148359 Xext: emulate XI events from core events when faking input. 148360 148361 We need XI events for event processing, so lets get rid of the pure core 148362 events ASAP. 148363 148364commit 8b3802d32041547fd6ed5393a56281c5c83d260e 148365Author: Peter Hutterer <peter@cs.unisa.edu.au> 148366Date: Tue May 13 13:47:52 2008 +0930 148367 148368 Xext: core button events have rootX/Y unset too. 148369 148370commit ed65e8b4f02a6da7f1c5d85984a9ccf6a94d0181 148371Author: Alan Coopersmith <alan.coopersmith@sun.com> 148372Date: Mon May 12 18:49:34 2008 -0700 148373 148374 Check for strcasestr and workaround it on systems without it 148375 148376commit 8234af6c6a3ade13f7720743bde79957c7df3f6c 148377Author: Peter Hutterer <peter@cs.unisa.edu.au> 148378Date: Tue May 13 11:17:39 2008 +0930 148379 148380 xfree86: remove Ubuntu-ism. 148381 148382 Additional #include was required due to broken libc headers or something like 148383 that. Remove it, there's more to the world than ubuntu alone. 148384 148385commit 59b8d29b3a91c65787dfadb5610342a62e85c48c 148386Author: Peter Hutterer <peter@cs.unisa.edu.au> 148387Date: Mon May 12 21:22:54 2008 +0930 148388 148389 Xi: Fix up ProcIVector, got out of sync with the protocol. 148390 148391 When the opcode squash happened in the protocol, the processing vector got out 148392 of sync for a few requests. As a result, client and server would interpret 148393 requests differently, leading to a couple of BadLength problems. 148394 148395commit 1fc1a2897e9185838c29d9ffff07c39d2fdf174f 148396Author: Peter Hutterer <peter@cs.unisa.edu.au> 148397Date: Mon May 12 20:14:05 2008 +0930 148398 148399 Remove UndisplayCursor API. 148400 148401 We can achieve the same thing by simply displaying a NullCursor, there's no 148402 need for a separate API. 148403 148404commit 913989d643595030e67e7c0a758a16ae0da62ed9 148405Author: Peter Hutterer <peter@cs.unisa.edu.au> 148406Date: Mon May 12 20:05:43 2008 +0930 148407 148408 mi: don't try to display a cursor for devices w/o cursors. 148409 148410commit 248a1df63430717550adb0e79068d8f9fcfacb0b 148411Author: Peter Hutterer <peter@cs.unisa.edu.au> 148412Date: Mon May 12 20:05:21 2008 +0930 148413 148414 render: don't dereference cursor if cursor is NULL. 148415 148416commit c50b5d978981b13cdb22a9ad41c1b64f90cebe51 148417Author: Aaron Plattner <aplattner@nvidia.com> 148418Date: Sun May 4 13:45:27 2008 -0700 148419 148420 Bug #14692: Allow drivers to have a say in Xinerama visual consolidation. 148421 148422 Create a new exported global variable, XineramaVisualsEqualPtr. Use this 148423 pointer to decide whether two visuals are equal during visual consolidation. 148424 This pointer can be wrapped, which allows drivers and extensions to control 148425 which visuals are consolidated. A wrapper can reject the visuals without 148426 calling down, but must call down and return that result if it deems the visuals 148427 equal. This ensures that all layers agree that the visuals are equal. 148428 148429 Pass the screen of the other visual into the VisualsEqual callchain. 148430 148431 Don't free PanoramiXVisuals since we need it for PanoramiXTranslateVisualID. 148432 148433 Don't skip the first visual on the other screen in PanoramiXMaybeAddVisual. 148434 148435 Skip the loop in PanoramiXTranslateVisualID if screen is 0. 148436 148437commit 86678e7cc2b021851ff508433fa160170f500c51 148438Author: Adam Jackson <ajax@redhat.com> 148439Date: Mon May 12 14:53:37 2008 -0400 148440 148441 Fix fbdevhw initialization for PCI drivers. 148442 148443 graphics/fb%d, not graphics:fb%d. Thanks sysfs. Thysfs. 148444 148445commit 1c8bd318fbaf65890ef16fe26c76dd5e6f14dfde 148446Author: Adam Jackson <ajax@redhat.com> 148447Date: Mon May 12 14:15:11 2008 -0400 148448 148449 X n'est pas une print API. 148450 148451commit f9fae16456c30479b0cb9317e57200af36795785 148452Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148453Date: Mon May 12 10:36:44 2008 -0700 148454 148455 XQuartz: Added some version checking protection so we don't trigger an infinite exec loop with new /usr/X11/bin/Xquartz and older X11.app 148456 (cherry picked from commit 78032815aeb10c22ff45b49702e9c9df82ab471c) 148457 148458commit ac4e33a9cd0ca2f0ec76181d11d5b90b82690c05 148459Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148460Date: Mon May 12 09:27:27 2008 -0700 148461 148462 XQuartz: More startup work... listen if we're the actual server 148463 (cherry picked from commit 3b0afb47c3d8ad922cb2315ed8034f4d77d4a249) 148464 148465commit a7800d9355377ac02833613c2ffc2423beec9970 148466Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148467Date: Fri May 9 16:44:31 2008 -0700 148468 148469 XQuartz: Starting to work on the new Mach IPC startup stuff for better launchd, ApplicationServices, and Dock support 148470 (cherry picked from commit 9b67fca9b7d3050d3d5582a5210270db7eb2ed05) 148471 148472commit 04211c3532ca078420e3745a5eac3d9de120bc32 148473Author: James Cloos <cloos@jhcloos.com> 148474Date: Mon May 12 03:03:13 2008 -0400 148475 148476 Prevent the -wm command line option from causing a SEGV 148477 148478 The -wm (when mapped) option for the BackingStore support has been 148479 causing the server to dereference a NULL pointer. 148480 148481 This has probably been the case since backing store has been 148482 implemented on top of Composite. 148483 148484 It looks like (some of?) Composite didn’t expect its WIndowPtr 148485 argument to be the root window. 148486 148487 In Composite’s compCheckRedirect() function we now avoid calling 148488 compAllocPixmap() and compFreePixmap() when the pWin pointer’s 148489 parent member is NULL, as is it the case with a server’s root window. 148490 148491 This addresses: 148492 148493 https://bugs.freedesktop.org/show_bug.cgi?id=15878 148494 148495commit 7e768c08f7809b8dba4db1931e63314e2b6e1cfa 148496Author: James Cloos <cloos@jhcloos.com> 148497Date: Mon May 12 02:53:59 2008 -0400 148498 148499 Clean up whitespace 148500 148501commit 3b6735528efc6d69ab7a7cd63dd09c87db7ef115 148502Author: Dave Airlie <airlied@redhat.com> 148503Date: Mon May 12 16:36:42 2008 +1000 148504 148505 pci: don't do this pci stuff when we don't have hw access 148506 148507commit 9dfb525f6c91acab5d1a65765a046bf9ee2aa082 148508Author: Julien Cristau <jcristau@debian.org> 148509Date: Sun May 11 23:17:27 2008 +0200 148510 148511 kdrive: allow disabling Composite 148512 148513 KdInitOutput() used to enable Composite when it was disabled by default, 148514 but now this hack prevents ``-extension Composite'' from working. 148515 Remove it, as Composite is enabled by default anyway. 148516 148517commit 1a01e96c6d15ef17a8b5ab1afa361fb12476a25e 148518Author: Alan Coopersmith <alan.coopersmith@sun.com> 148519Date: Fri May 9 15:38:44 2008 -0700 148520 148521 Return a valid X error when stuck in font alias loop 148522 148523 Part of fix for Sun bug 4258475 148524 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=4258475> 148525 148526commit 7b3066d9b9099135d9c49e0682161d5568fc535b 148527Merge: 2a3d1421e 315f08905 148528Author: Drew Parsons <drew@emerall.com> 148529Date: Sat May 10 00:01:15 2008 +1000 148530 148531 Merge branch 'master' of git://anongit.freedesktop.org/git/xorg/xserver into upstream-experimental 148532 148533commit 2a3d1421e0cc18822ae8f478fcc272e16a9e9340 148534Author: Drew Parsons <dparsons@debian.org> 148535Date: Fri May 9 23:20:11 2008 +1000 148536 148537 Disable D-BUS from Xprint. 148538 148539 Use dummy config functions to replace those from config/config.c, and 148540 therefore do not link Xprt with $CONFIG_LIB. 148541 148542 Works around an endlessly spinning loop in dix/dispatch.c::Dispatch() 148543 (WaitForSomething() not waiting) when built with dbus, which was 148544 causing Xprt to use 95% cpu. 148545 148546commit 315f089056da67d4c463ed002eb2b74e38493b49 148547Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148548Date: Thu May 8 19:46:03 2008 -0700 148549 148550 XQuartz: Reorganized some of the build system in prep for the Mach IPC startup work. 148551 (cherry picked from commit 2232c91d5c277673929eab2abb5e0495c00877cb) 148552 148553commit 28ac79450c69219dc501e072c6e5028e7136380d 148554Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148555Date: Thu May 8 19:47:40 2008 -0700 148556 148557 Updated .gitignore for new Xquartz layout 148558 (cherry picked from commit cd4d2355e227549a3410485a130549dd91ccdcfe) 148559 148560commit a07c5ad172b343ef26d2b41ff25f143950441c23 148561Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148562Date: Thu May 8 16:57:42 2008 -0700 148563 148564 XQuartz: Set bundle version to 2.3.0 148565 (cherry picked from commit 8a0524b30e1e860f3ae35741c116fc8da28aef79) 148566 148567commit 05f23ed3ea6ee0f052aee41b6573325fe0063fd8 148568Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148569Date: Tue May 6 16:12:41 2008 -0700 148570 148571 XQuartz: Fixed some issue in our bundle creation 148572 (cherry picked from commit 330ffad5477e32c5ab9ed338bc628bd5ae9f4c98) 148573 148574commit 90b963c0da2b33bdc21483f1a089b95c7e717333 148575Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148576Date: Tue May 6 16:07:33 2008 -0700 148577 148578 Set CSRG_BASED on OSX 148579 (cherry picked from commit ff085deba18682caa2f93d61a75b38db87d747b1) 148580 148581commit ff013b0da4e6d33b2b69ce1212e9bd62050574e1 148582Author: Peter Hutterer <peter@cs.unisa.edu.au> 148583Date: Thu May 8 16:58:31 2008 +0930 148584 148585 config: override xkb_{r,m,l,v} with Xkb{r,m,l,v} if the latter is set. 148586 148587 The HAL spec says that input.xkb.{rmlv}* can be sent, but if the user 148588 specifies a X-specific {rmlv}, then this is overridden through the use of 148589 input.x11_options.Xkb{RMLV}. 148590 However, the way how the server parses options--by ignoring capitalisation, 148591 underscores and spaces--the HAL and the x11_options would override each other. 148592 148593 So we simply filter the options, letting Xkb{RMLV} override xkb_{rmlv} and 148594 only actually add them to the device after parsing _all_ options. 148595 148596 * rmlv ... rules, model, layout, variant 148597 148598 See Bug 13037 <http://bugs.freedesktop.org/show_bug.cgi?id=13037> 148599 (cherry picked from commit fc35d1e3be201e3821413bb2eeb8d43e1e56ba17) 148600 148601commit 901978ebe0f446532255701cd536e246e805a55b 148602Author: Peter Hutterer <peter@cs.unisa.edu.au> 148603Date: Thu May 8 14:05:56 2008 +0930 148604 148605 config: remove trailing whitespaces. 148606 148607 It makes my vim look ugly. Put "let c_space_errors=1" into your .vimrc. 148608 (cherry picked from commit 1f54c05cf8a6b82e5fc6362f7f8e8fdc2444b9e8) 148609 148610commit 1c54c148895225e4ab3c781fe57d09e5f64353aa 148611Author: Daniel Stone <daniel@fooishbar.org> 148612Date: Fri May 9 00:26:16 2008 +0300 148613 148614 Revert "GL: Make errors non-fatal" 148615 148616 Turns out this just caused segfaults further down the line. Oops. 148617 148618 This reverts commit 268d61e00cf4bc52c05f19eda7ab4f6accce12c8. 148619 148620commit f17ba5d5849c92603f453195aca384844ca76d74 148621Author: Adam Jackson <ajax@redhat.com> 148622Date: Thu May 8 16:04:24 2008 -0400 148623 148624 Bug #13104: Remove broken XAA a1 glyph fast path. 148625 148626commit ddaecfa13cefee7c66b39b606c8640c6544d4943 148627Merge: 9c2e0871c cf20df39c 148628Author: Adam Jackson <ajax@redhat.com> 148629Date: Thu May 8 14:33:58 2008 -0400 148630 148631 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 148632 148633commit fc35d1e3be201e3821413bb2eeb8d43e1e56ba17 148634Author: Peter Hutterer <peter@cs.unisa.edu.au> 148635Date: Thu May 8 16:58:31 2008 +0930 148636 148637 config: override xkb_{r,m,l,v} with Xkb{r,m,l,v} if the latter is set. 148638 148639 The HAL spec says that input.xkb.{rmlv}* can be sent, but if the user 148640 specifies a X-specific {rmlv}, then this is overridden through the use of 148641 input.x11_options.Xkb{RMLV}. 148642 However, the way how the server parses options--by ignoring capitalisation, 148643 underscores and spaces--the HAL and the x11_options would override each other. 148644 148645 So we simply filter the options, letting Xkb{RMLV} override xkb_{rmlv} and 148646 only actually add them to the device after parsing _all_ options. 148647 148648 * rmlv ... rules, model, layout, variant 148649 148650 See Bug 13037 <http://bugs.freedesktop.org/show_bug.cgi?id=13037> 148651 148652commit 1f54c05cf8a6b82e5fc6362f7f8e8fdc2444b9e8 148653Author: Peter Hutterer <peter@cs.unisa.edu.au> 148654Date: Thu May 8 14:05:56 2008 +0930 148655 148656 config: remove trailing whitespaces. 148657 148658 It makes my vim look ugly. Put "let c_space_errors=1" into your .vimrc. 148659 148660commit 1a314a1ef06222977e7ccb94331ce31c17534b2a 148661Author: Peter Hutterer <peter@cs.unisa.edu.au> 148662Date: Thu May 8 13:16:48 2008 +0930 148663 148664 Revert "config: deprecate (and ignore) the use of input.xkb.whatever options." 148665 148666 daniels: 148667 "Hrm, I'd prefer to have input.xkb.{m,l,v,o} be the primary keys, and 148668 have input.x11_options be a backup for that, rather than the former 148669 being deprecated, for the reasons I listed earlier ..." 148670 148671 see http://bugs.freedesktop.org/show_bug.cgi?id=13037#c51 148672 148673 This reverts commit 26188875de1ccc84ac60bfb251e3ec43de2b9e22. 148674 148675commit cf20df39cc78203d17b99223908af388ecbf7d0e 148676Author: Daniel Stone <daniel@fooishbar.org> 148677Date: Wed May 7 22:24:19 2008 +0300 148678 148679 XKB: Actually explain keymap failures 148680 148681 When something went wrong building a keymap, try to explain to the user 148682 what it actually was, instead of the dreaded 'Failed to load XKB keymap' 148683 catch-all. 148684 148685commit 641a5f955b7b3ae04eeb6bc45fb30b0b531898e4 148686Author: Daniel Stone <daniel@fooishbar.org> 148687Date: Wed May 7 22:29:04 2008 +0300 148688 148689 Build: Ensure xf86DefModeSet.c ends in an empty line 148690 148691 This shuts up a warning. 148692 148693commit 268d61e00cf4bc52c05f19eda7ab4f6accce12c8 148694Author: Daniel Stone <daniel@fooishbar.org> 148695Date: Wed May 7 22:28:45 2008 +0300 148696 148697 GL: Make errors non-fatal 148698 148699 GLX, there's more to the world than just you. If you fail to load the 148700 software renderer, don't bring the entire server down. 148701 148702 The error path probably needs better testing on this one, but it seems 148703 mostly okay to me. 148704 148705commit 9c2e0871cfbe54e73eec1f790a7e383d08555055 148706Author: Adam Jackson <ajax@redhat.com> 148707Date: Wed May 7 13:21:26 2008 -0400 148708 148709 Bug #13104: XAA: Adapt to glyph storage changes. 148710 148711 Glyph bits are now stored in a proper pixmap, not just hanging off the 148712 end of a GlyphRec. 148713 148714commit 26188875de1ccc84ac60bfb251e3ec43de2b9e22 148715Author: Peter Hutterer <peter@cs.unisa.edu.au> 148716Date: Wed May 7 13:13:57 2008 +0930 148717 148718 config: deprecate (and ignore) the use of input.xkb.whatever options. 148719 148720 These options are still sent by some HAL implementations (e.g. HAL on FC8), 148721 and may overwrite the options set in the x11-input.fdi file. 148722 For a more detailed description of why see Bug #13037, comment 42. 148723 148724 X.Org Bug 13037 <http://bugs.freedesktop.org/show_bug.cgi?id=13037#c42> 148725 148726commit b6a0c6d4864f73a18beb841b16e9be56f2fcd77e 148727Author: Alan Coopersmith <alan.coopersmith@sun.com> 148728Date: Tue May 6 17:06:34 2008 -0700 148729 148730 Allow using libmd instead of libcrypto for SHA1 hashing in render/glyph.c 148731 148732 Builders can force one or the other by passing SHA1_LIB & SHA1_CFLAGS 148733 to configure 148734 148735commit 718652eaf9221e0eeec2c971dd7baa97f827451b 148736Author: Adam Jackson <ajax@redhat.com> 148737Date: Tue May 6 17:52:37 2008 -0400 148738 148739 Bug #13104: Don't let XAA glyph pixmaps anywhere near video memory. 148740 148741 Since glyphs are stored in pixmaps now, they can make their way into VRAM, 148742 which invalidates a bunch of fast-path assumptions in the XAA code. Thus 148743 you end up doing color-expands or WriteBitmap from la-la land and your 148744 aliased glyphs go all funny. 148745 148746 Since XAA isn't ever growing the ability to do sane glyph accel, just force 148747 glyph pixmaps into host memory by catching them at CreatePixmap time. 148748 148749commit a85d3ac87cc354093bb1e88697c44254e7721bb9 148750Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148751Date: Tue May 6 02:59:13 2008 -0700 148752 148753 XQuartz: Fixed typo 148754 (cherry picked from commit 56b7988d2662caa4d31094695b414080e4470ed4) 148755 148756commit 077ced6384abad78253e857091e78f3685965b9d 148757Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148758Date: Tue May 6 02:47:03 2008 -0700 148759 148760 XQuartz: Added uncommitted files 148761 (cherry picked from commit e414ec462cfc63f8eb7f504f526f5a2c73f51e69) 148762 148763commit 80e502c5d1f7e9221c6ae40716d6402fd28d8806 148764Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148765Date: Tue May 6 02:28:36 2008 -0700 148766 148767 Fixed up dist 148768 (cherry picked from commit f225222ba2bf4f03425107f258d60b73c88efaec) 148769 148770commit 7295e544332b0fa929f651304f9d4aca3db4a33e 148771Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148772Date: Tue May 6 00:06:19 2008 -0700 148773 148774 XQuartz: Move server bits into bundle and setup stub in /usr/X11/bin/Xquartz in prep for startup rewrite 148775 (cherry picked from commit 453a982e6382cff06ea27abba225440b07068f50) 148776 148777commit d41d677ab4118e73140ea2392e0d48eb361af1cf 148778Author: Ben Byer <bbyer@bushing.apple.com> 148779Date: Mon May 5 20:55:08 2008 -0700 148780 148781 Xquartz: experimental embedding of Sparkle.framework 148782 (cherry picked from commit b7a1a640cef8c69442859cbf89034ad362a19684) 148783 148784commit 9e0d73fd8a43647b648ec5b2f0bcaae30c03259c 148785Author: Peter Hutterer <peter@cs.unisa.edu.au> 148786Date: Tue May 6 17:52:15 2008 +0930 148787 148788 xfree86: call SetCursor for the VCP when switching between SW/HW rendering. 148789 148790 We need a manual call to SetCursor when we switch from SW to HW rendering and 148791 the other way round. This way we display the new cursor after removing the old 148792 one. 148793 In addition, we only update the internal state for the VCP's sprite. This way, 148794 when we switch back to HW rendering the state is up-to-date and wasn't 148795 overwritten with the other sprite's state. 148796 148797 The second part is a hack. It would be better to keep a state for each sprite, 148798 but then again we don't have hardware that can render multiple cursors so we 148799 might as well do with the hack. 148800 148801commit e4ebfed1785597b48b68e1bbdde3e5c4061b749f 148802Author: Dave Airlie <airlied@linux.ie> 148803Date: Tue May 6 16:46:37 2008 +1000 148804 148805 xf86edid: fix typo in debug output 148806 148807commit b9c1a57e7a98dea63cd362f714411547e728a85a 148808Author: Peter Hutterer <peter@cs.unisa.edu.au> 148809Date: Tue May 6 14:51:23 2008 +0930 148810 148811 xfree86: switch between to SW cursors if more than 1 sprite is available. 148812 148813 Switches back to HW cursors when sprites other than the VCP are removed. 148814 The current state requires the cursor to change shape once before it updates 148815 to SW / HW rendering (whatever is appropriate), e.g. by moving into a 148816 different window. Until this is done, the cursor is invisible. 148817 148818commit 3df88c17e315b5ae580096e4bc88920d1f452e83 148819Author: Peter Hutterer <peter@cs.unisa.edu.au> 148820Date: Tue May 6 11:50:20 2008 +0930 148821 148822 xfree86: re-enable hardware cursor (for a single cursor) 148823 148824 This commit enables HW rendering for cursors again, but only for a single 148825 cursor. Other cursors can be created, however they will not be visible. 148826 148827commit 6acc2acd0db2826add7c47e94e4061d169a41f88 148828Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148829Date: Mon May 5 15:32:26 2008 -0700 148830 148831 Rootless: mi doesn't let us resize root, so we need to do it ourselves... 148832 (cherry picked from commit c1ec36e28cff857664090cc8792db1ae93b783fa) 148833 148834commit efa65a0317e12c9ad34fa00fe90bf5eae9fa2670 148835Author: Keith Packard <keithp@keithp.com> 148836Date: Sun May 4 21:52:58 2008 -0700 148837 148838 Rework composite overlay window code to fix several resource management bugs. 148839 148840 The composite overlay window code had several misunderstandings of the 148841 workings of the X server, in particular error handling paths would often 148842 double-free objects. Clean all of this up by using resource destruction as 148843 the sole mechanism for freeing resource-based objects. 148844 148845commit 6c1accce87c9bd640c1b4bbc49bae7d44b1cc97b 148846Author: Keith Packard <keithp@keithp.com> 148847Date: Sun May 4 21:51:08 2008 -0700 148848 148849 Empty the borderClip of manual redirect windows. (bug 15823) 148850 148851 Thanks to Owen Taylor for root-causing this one. 148852 148853 If a TreatAsTransparent window has any area in the borderClip, that will be 148854 added to the totalClip region for use by other windows. That's wrong. 148855 Instead, simply empty the borderClip for TreatAsTransparent windows right up 148856 front. 148857 148858commit d500eeb9458336780d77baf8b7db96e1f7ff4f0d 148859Author: Aaron Plattner <aplattner@nvidia.com> 148860Date: Sun May 4 10:26:25 2008 -0700 148861 148862 Xephyr: Handle depth 30 and reject higher depths rather than crashing. 148863 148864commit f4a68f3701889950d3b98842d021f357f3913fea 148865Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148866Date: Fri May 2 17:35:57 2008 -0700 148867 148868 XQuartz: Fix mouse input offsets earlier since GetPointerEvents does not like negative (x,y) values 148869 (cherry picked from commit 8d9eab3a2ec5955cc2698fdcb1fa6ed12b2aadb7) 148870 148871commit 96fa7da3b87e2f9187ec13dad259beefdc17cd94 148872Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148873Date: Fri May 2 16:53:45 2008 -0700 148874 148875 XQuartz: Avoid a possible crash at startup due to unfavorable context switching. 148876 (cherry picked from commit ff10c37bdd09656cf2f7ee9577f5552caa1ffdb8) 148877 148878commit 6f63724b28c8b3fd85314fb95d1e4f363b610e19 148879Author: Peter Hutterer <peter@cs.unisa.edu.au> 148880Date: Fri May 2 14:56:58 2008 +0930 148881 148882 xfree86: don't try to repaint the cursor before cursors exist. 148883 148884commit 334456952930cb3e55c02fcdada2d0c074cd0520 148885Merge: 8e56fd972 3b8d53452 148886Author: Peter Hutterer <peter@cs.unisa.edu.au> 148887Date: Fri May 2 11:00:14 2008 +0930 148888 148889 Merge branch 'master' into mpx 148890 148891 Conflicts: 148892 148893 Xi/exevents.c 148894 dix/devices.c 148895 148896commit a338007be6fd4302af012109d959f43ac7eab2b1 148897Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148898Date: Thu May 1 18:02:59 2008 -0700 148899 148900 XQuartz: Adjust the screen origin offset properly for multimonitor setups 148901 (cherry picked from commit f2020b9836bacd0593ac0b4c8541e32714ab02a9) 148902 148903commit 8e56fd9728d63a7a48ef44503425c6e25c7c9ffb 148904Author: Peter Hutterer <peter@cs.unisa.edu.au> 148905Date: Fri May 2 10:22:06 2008 +0930 148906 148907 dix: continue to check for invalid valuator ranges. 148908 148909 In an ideal world, the drivers would init all axes with NO_AXIS_LIMIT. In the 148910 real world, this is an ABI break, so let's just leave the old check in. 148911 148912commit 3b8d53452cd6c74d32d7759964a7cd9ee775f161 148913Author: Michel Dänzer <michel@tungstengraphics.com> 148914Date: Thu May 1 15:10:52 2008 +0200 148915 148916 EXA: Fall back in CompositeRects if the driver can't accelerate Composite. 148917 148918 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=15780 . 148919 148920commit 01c61f3d972fc2f4e5bb536dd00d8b6bbeb0fb3d 148921Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 148922Date: Thu May 1 01:02:39 2008 -0300 148923 148924 xorgcfg: Fix a crash if xorg.conf doesn't have a Files section. 148925 148926 This patch only creates a Files section if required, so if no entries are 148927 added, an empty Files section will not be created. 148928 148929 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 148930 148931commit 6899ff81e8f668382e3e2f9afdd9cdf51d0e0f7d 148932Author: Paulo Cesar Pereira de Andrade <pcpa@mandriva.com.br> 148933Date: Wed Apr 30 16:38:42 2008 -0300 148934 148935 Fix an off by one read error in drmSIGIOHandler(). 148936 148937 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 148938 148939commit ec0419b50534006a272278b79ee51538193661cb 148940Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148941Date: Wed Apr 30 17:30:57 2008 -0700 148942 148943 XQuartz: Moved some relevant stuff from darwin.h to darwinEvents.h 148944 (cherry picked from commit 1fcf74a436f2e19cceb3f285b89859025d94c040) 148945 148946commit 28d1e21f1523d475cb757d3cfa33737ed62c534d 148947Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148948Date: Wed Apr 30 17:30:32 2008 -0700 148949 148950 Dix: Cleaned up sanity checking in Get{Pointer,Keyboard}Events 148951 148952 XQuartz was crashing because the Appkit thread was trying to GetXXXEvents while the Xserver thread was exiting. 148953 This adds some more sanity checks and avoids that crash 148954 (cherry picked from commit 34ec4bd6acb834c0e3f9a5042a0cc3f52c6f3a68) 148955 148956commit ce36ae526d88d20ff67cd6cb429fb06f48d231f6 148957Author: Kazuhiro Inaoka <inaoka.kazuhiro@renesas.com> 148958Date: Wed Apr 30 11:58:32 2008 +0300 148959 148960 Add M32R architecture support (bug #10020) 148961 148962 Still needs autotools support, so this won't actually _build_: it's just 148963 a starting point. 148964 148965commit a0e6a7d4f507c5c0a0b11adb10394af58a0a6e07 148966Author: Peter Hutterer <peter@cs.unisa.edu.au> 148967Date: Wed Apr 30 17:57:48 2008 +0930 148968 148969 dix: resize EQ to minimum size to avoid reallocs during SIGIO. 148970 148971 When a new device is added, calculate the event size needed if a DCCE event is 148972 sent and set the EQ's event size to this minimum. This avoids reallocs when a 148973 event is sent (which may happen during a SIGIO). 148974 148975commit b71b51c982706501b6229532ce342752207426bb 148976Author: Ben Byer <bbyer@bushing.apple.com> 148977Date: Tue Apr 29 23:37:58 2008 -0700 148978 148979 xquartz: created darwinTablet, to represent tablet events 148980 (cherry picked from commit 50641bce730ee03738fa0c4beb8125b34fda556f) 148981 148982commit 519581b0bab5cc6dac397877da485745f18c2120 148983Author: Ben Byer <bbyer@bushing.apple.com> 148984Date: Tue Apr 29 23:00:09 2008 -0700 148985 148986 xquartz: remove unused params from DarwinEQInit 148987 (cherry picked from commit 299a056737168ca1faefd675dce6d6ab13176be9) 148988 148989commit d70487a4c09cfeb90d996ab40a23a74b2c15be6f 148990Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 148991Date: Tue Apr 29 23:59:26 2008 -0700 148992 148993 XQuartz: Fix to tablet-event handling code; we now scale 148994 more conservatively (to match Linux's Wacom driver) and 148995 we now receive all tablet-related events. 148996 (cherry picked from commit 588683cecca2cfc65a28de035cd6ee3d64ff59d2) 148997 148998commit fea39c94bcc8b635d37530d8a27ee92acab484fe 148999Author: Peter Hutterer <peter@cs.unisa.edu.au> 149000Date: Wed Apr 30 13:17:14 2008 +0930 149001 149002 dix: fix typo from last commit. 149003 149004commit ffaccc2dc91f4ca4ea10da010206a0a7d2b5540c 149005Author: Peter Hutterer <peter@cs.unisa.edu.au> 149006Date: Wed Apr 30 11:45:19 2008 +0930 149007 149008 input: replace -1 as default axis limit with NO_AXIS_LIMIT define. 149009 149010 This allows easier refacturing of the coordinate limit handling. Grepping for 149011 -1 is boring. 149012 149013commit 00acb40f2bc5bb4a1977b9b08db75630677ff787 149014Author: Peter Hutterer <peter@cs.unisa.edu.au> 149015Date: Wed Apr 30 11:29:03 2008 +0930 149016 149017 dix: fix typo in clipAxis. 149018 149019 Check needs to be (min_axis < max_axis), not (min_axis < min_axis) 149020 149021commit f65a1a62f9eac2f6620d5d469dbf6c79b2eadab7 149022Author: Hong Liu <hong.liu@intel.com> 149023Date: Tue Apr 29 09:00:11 2008 -0700 149024 149025 Keep rotation wrappers in place while transforms are in use 149026 149027 LeaveVT/EnterVT cycles will free/realloc shadow frame buffers. Because of 149028 this, the presense/absence of that data is insufficient to know whether 149029 the screen function wrappers are necessary. Instead, the 'transform_in_use' 149030 flag should be used. 149031 149032 This patch also adds 'xf86RotateFreeShadow' for drivers to use at LeaveVT 149033 time to free the rotation data; it will be reallocated on EnterVT. 149034 149035commit 18b33dd4ff46f63bad70b493b1d0b0758c961ada 149036Author: Peter Hutterer <peter@cs.unisa.edu.au> 149037Date: Tue Apr 29 15:24:55 2008 +0930 149038 149039 Xi: stop excessive use of _X_EXPORT. 149040 149041commit 2b179c32ac40f5258d95ae16da96a79fa8eea696 149042Author: Peter Hutterer <peter@cs.unisa.edu.au> 149043Date: Tue Apr 29 15:15:16 2008 +0930 149044 149045 include: add declaration for CopyGetMasterEvent to shut up compiler warning. 149046 149047commit ff36753755360aaa16ab8604a0ab78123884b57d 149048Author: Peter Hutterer <peter@cs.unisa.edu.au> 149049Date: Tue Apr 29 14:53:41 2008 +0930 149050 149051 mi: call the custom event handlers for both MD and SD. 149052 149053commit ec2fca7e6f7ce8fdf33d959b7adeaae935ec4b37 149054Author: Peter Hutterer <peter@cs.unisa.edu.au> 149055Date: Tue Apr 29 14:40:44 2008 +0930 149056 149057 xfree86: DGA needs to call into Xi. 149058 149059 Two steps: first use UpdateDeviceState() instead of replicating the device 149060 changing code. And emulate XI events instead of core events. 149061 149062commit 8208b842b7dd3b30bafdd74147bdfa63bc00cc40 149063Author: Peter Hutterer <peter@cs.unisa.edu.au> 149064Date: Tue Apr 29 14:19:42 2008 +0930 149065 149066 Xi: add IEventBase global variable. Stores event base for X Input events. 149067 149068commit 46340b740640b76b52b8e69bc67f9201c7e974c3 149069Author: Peter Hutterer <peter@cs.unisa.edu.au> 149070Date: Tue Apr 29 14:17:48 2008 +0930 149071 149072 Xi: change UpdateDeviceState to be _X_EXPORT. 149073 149074commit 3a922c5d07509703a3eeda2bbec6f332e6b0ffac 149075Author: Peter Hutterer <peter@cs.unisa.edu.au> 149076Date: Mon Apr 28 17:32:54 2008 +0930 149077 149078 Xi: apply ValuatorEvent data to the device's valuators. 149079 149080 After UpdateDeviceState, the device has the current position in absolute 149081 coordinates, the event has the correct valuator data to be delivered to the 149082 client. 149083 149084commit f2a20294fe26f4c9d245d24d065331fefd8f830d 149085Author: Peter Hutterer <peter@cs.unisa.edu.au> 149086Date: Mon Apr 28 16:56:22 2008 +0930 149087 149088 dix: store all valuators as they come from the device in the valuator events. 149089 149090 The event's sequenceNumber is mis-used to determine whether the values are 149091 relative or absolute. 149092 149093commit d5ab89f4f1acbe2614036e8934122185ac0f81ee 149094Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149095Date: Mon Apr 28 15:52:24 2008 -0700 149096 149097 XQuartz: Unset CFProcessPath... blech 149098 (cherry picked from commit ce4fbfbc75c62a092214d140c7550279aebe69ef) 149099 149100commit 0c2312b21b6700c1425baccaf9c26150b4f0fd6a 149101Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149102Date: Mon Apr 28 15:28:21 2008 -0700 149103 149104 XQuartz: xprIsX11Window can be called from the Appkit thread (see X11Application.m) 149105 (cherry picked from commit 22cf72437601c07b8a6c744b4f2f1f4cd6713e60) 149106 149107commit c3dc4bdbb018c0606d561e2dfb9a36e9297fa312 149108Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149109Date: Mon Apr 28 15:42:02 2008 -0700 149110 149111 XQuartz: Disabled DPMS extension 149112 (cherry picked from commit 1448fed9b6d484f471b9b2982c76ca921a9273b9) 149113 149114commit 07382a70c7ac9807dfb31821a4763bea2309bde6 149115Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149116Date: Mon Apr 28 12:36:54 2008 -0700 149117 149118 XQuartz: Added thread debugging to xprFrame.c 149119 (cherry picked from commit 41542502b321c697271c1752525b600872b6df96) 149120 149121commit 8349732a6720652bfbad7874a952be73a0e8e77b 149122Author: Michel Dänzer <michel@tungstengraphics.com> 149123Date: Mon Apr 28 21:09:35 2008 +0200 149124 149125 EXA: Try to accelerate non-antialiased text via the glyph cache as well. 149126 149127 Treat 1 bit glyphs and masks as PICT_a8 in the glyph cache. We're not able to 149128 accelerate them otherwise. 149129 149130commit a65d530040bb561ba88c5d8c71633a7c0bf11e89 149131Author: Michel Dänzer <michel@tungstengraphics.com> 149132Date: Mon Apr 28 21:03:12 2008 +0200 149133 149134 EXA: Accumulate glyphs whenever possible, for full benefits of the glyph cache. 149135 149136commit e7eaac59c424a205dd106fc7d70734ff4b390f28 149137Author: Michel Dänzer <michel@tungstengraphics.com> 149138Date: Mon Apr 28 21:00:55 2008 +0200 149139 149140 EXA: Glyph cache upload tweaks. 149141 149142 Track damage after using UploadToScreen directly. 149143 149144 Don't waste any effort on empty glyphs. 149145 149146commit cc08c06665ffe29ad44d023d75d0f86e5338875d 149147Author: Owen Taylor <otaylor@huygens.home.fishsoup.net> 149148Date: Mon Apr 28 21:00:55 2008 +0200 149149 149150 EXA: Use UploadToScreen() for uploads to glyph cache 149151 149152 When possible, use UploadToScreen() rather than CompositePicture() 149153 to upload glyphs onto the glyph cache pixmap. This avoids allocating 149154 offscreen memory for each glyph making management of offscreen 149155 areas much more efficient. 149156 149157commit 13fd2256300b61d88b840952d838f834523f5dd7 149158Author: Owen Taylor <otaylor@huygens.home.fishsoup.net> 149159Date: Mon Apr 28 21:00:55 2008 +0200 149160 149161 EXA: Clean up debug messages 149162 149163commit fcb5949928f1c27f67f40c094c3c673786574422 149164Author: Owen Taylor <otaylor@huygens.home.fishsoup.net> 149165Date: Mon Apr 28 21:00:54 2008 +0200 149166 149167 EXA: Fix overlapping glyphs in glyph cache 149168 149169 Allocate each cache at a different vertical position in the 149170 per-format pixmap. Fix width/height confusion when choosing 149171 the cache for a glyph. 149172 149173commit 40eb14c9482457969e0bde97c49edad536285e02 149174Author: Owen Taylor <otaylor@huygens.home.fishsoup.net> 149175Date: Mon Apr 28 21:00:54 2008 +0200 149176 149177 EXA: Add exaCompositeRects() 149178 149179 Add a function to composite multiple independent rectangles 149180 from the same source to the same destination in a single 149181 operation: this is useful for building a glyph mask. 149182 149183commit 54184110f6f3e5d7276d5431e739a4fcf0c3523e 149184Author: Owen Taylor <otaylor@huygens.home.fishsoup.net> 149185Date: Mon Apr 28 21:00:54 2008 +0200 149186 149187 EXA: Use a single large glyph cache pixmap 149188 149189 Add back exaGlyphs(); the new version copies the glyph images 149190 onto a single large glyph pixmap and draws from their to the 149191 destination surface. This reduces the management of small 149192 offscreen areas and will allow us to avoid texture unit setup 149193 between each glyph. 149194 149195commit e9734306088e12f2cd68bf347ecf8415be4f0268 149196Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149197Date: Mon Apr 28 11:47:49 2008 -0700 149198 149199 XQuartz: More startup / threading house cleaning. 149200 (cherry picked from commit 72653c24c00dfba64ce35a3d400598bcd77defc1) 149201 149202commit 4b46fc931e61bec0abd6a86062e46dd7a408e745 149203Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149204Date: Mon Apr 28 10:50:51 2008 -0700 149205 149206 XQuartz: Updated Localizable.strings 149207 (cherry picked from commit d8d9c866b90fb24c93bd6e25fa90f8f2bf58ad34) 149208 149209commit 4017ebe5bfa7a261cd1135801756c44e9fe93ca1 149210Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149211Date: Sun Apr 27 00:01:14 2008 -0700 149212 149213 XQuartz: Added missing Xquartz.man.pre to EXTRA_DIST 149214 (cherry picked from commit 03e707987f7f32e47dd0355c6d16bfb9169a379b) 149215 149216commit 53dba5381fdd8f644e16aaa0ecb05df4dc615b23 149217Author: Peter Hutterer <peter@cs.unisa.edu.au> 149218Date: Mon Apr 28 11:37:52 2008 +0930 149219 149220 dix: if alloc of a master keyboard fails, remove the master pointer. 149221 149222commit 1fab51edfc82e1ef60dfa29fd5d93478066a3998 149223Author: Peter Hutterer <peter@cs.unisa.edu.au> 149224Date: Mon Apr 28 11:37:21 2008 +0930 149225 149226 mi: guard against NULL-pointer dereference. 149227 149228commit b5004722a208479a4bc762ff428bf4cbeb430d53 149229Author: Peter Hutterer <peter@cs.unisa.edu.au> 149230Date: Mon Apr 28 10:26:01 2008 +0930 149231 149232 Xi: Only return VCP, VCK and floating SDs to Xi 1.x clients. 149233 149234 This is better than the approach implemented with 149235 8973a3f7983240407dd6da59b3643f40e6a3d83a which disabled XI altogether for 1.x. 149236 Instead, return a device list that resembles a traditional XI setup on pre XI 149237 2.0 servers. If the client tries to open a device other than a floating SD, 149238 return a BadDevice error. 149239 149240commit 0bd28315afc3b7223f8ff9f17597db09500a9388 149241Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149242Date: Sat Apr 26 23:00:42 2008 -0700 149243 149244 Xquartz: Added missing to EXTRA_DIST 149245 (cherry picked from commit f4b963256feb03e9e6b5521bdefeb390e9a49688) 149246 149247commit ef1c52053755fa14b4ca98b22c506f73f5f4a4b7 149248Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149249Date: Sat Apr 26 19:21:05 2008 -0700 149250 149251 XQuartz: Cleaned up startup and thread creation a tad. 149252 (cherry picked from commit c861fe00e112b21ee0156d09a6cd5281642a1dcc) 149253 149254commit b114d4e861885cc5b49cd81b33ad825461811b3e 149255Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149256Date: Sat Apr 26 17:04:49 2008 -0700 149257 149258 XQuartz: Don't subtract the titlebar off of the pointer_y 149259 (cherry picked from commit 00a9567acce2b27a649cbebd4790e3043688a7d8) 149260 149261commit 684b5d8382bf0bc6cd55774226ee362e81c0adbf 149262Author: Peter Hutterer <peter@cs.unisa.edu.au> 149263Date: Sat Apr 26 19:55:59 2008 +0930 149264 149265 Xi: whoops, stray ! caused a bit of memory mess. 149266 149267commit 87071b604c4b7c3c79ab784e2c09ff691e2bc5ad 149268Author: Peter Hutterer <peter@cs.unisa.edu.au> 149269Date: Sat Apr 26 19:37:57 2008 +0930 149270 149271 Revert "Xi: disable XI for clients < XI 2.0" 149272 149273 Turns out some programs don't like this change. gnome-settings-daemon crashes 149274 hard if you tell it that XI doesn't exist. So, tell them we have XI, but leave 149275 the other change (the one that pretends no devices are available). 149276 149277 This reverts commit 8973a3f7983240407dd6da59b3643f40e6a3d83a. 149278 149279commit 7447a30fb27ed50a20a85b5a2de9afe7dea8cfa5 149280Author: Peter Hutterer <peter@cs.unisa.edu.au> 149281Date: Sat Apr 26 19:03:13 2008 +0930 149282 149283 Xi: if a pre-XI2 client tries to list the devices, pretend we don't have any. 149284 149285 XI 1.x isn't supported anymore, so let's pretend we don't have any devices. 149286 This stops clients from opening them and thus stops interference. 149287 149288commit 8973a3f7983240407dd6da59b3643f40e6a3d83a 149289Author: Peter Hutterer <peter@cs.unisa.edu.au> 149290Date: Sat Apr 26 17:48:52 2008 +0930 149291 149292 Xi: disable XI for clients < XI 2.0 149293 149294 Sorry. With the huge changes in the device handling I honestly don't know how 149295 to support XI and XI2 alongside. So let's just pretend XI doesn't exist if a 149296 client doesn't request it supporting XI2. 149297 149298commit b304b0a65cc57127cdea103f2c5114e4ea79af41 149299Author: Peter Hutterer <peter@cs.unisa.edu.au> 149300Date: Sat Apr 26 17:38:55 2008 +0930 149301 149302 Xi: add versioning support. 149303 149304 Remember the version the client sent to us, so we can adjust our replies 149305 accordingly. This requires the client to use the {major|minor}Version fields 149306 in the GetExtensionVersion request. However, they were padding before, so we 149307 must assume they are garbage if nbytes is non-zero. If nbytes is zero, the 149308 client is probably a new client and we can handle it correctly. 149309 149310commit b9ca7896356f79ee27be5d5aa62052f6984282b0 149311Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149312Date: Fri Apr 25 17:54:36 2008 -0700 149313 149314 XQuartz: Added some pseudoramiX debug traces 149315 (cherry picked from commit 5bee1585a399eab0a7b6fc80ad476d81b5d227d3) 149316 149317commit e251c9e75afdc5e32f2bc9801712272358934266 149318Author: Peter Hutterer <peter@cs.unisa.edu.au> 149319Date: Fri Apr 25 16:49:35 2008 +0930 149320 149321 Remove all traces of #ifdef XINPUT and the matching bits from the configure.ac 149322 149323commit 1d0438de176551aaeff549664f00b8bd83d465af 149324Author: Peter Hutterer <peter@cs.unisa.edu.au> 149325Date: Fri Apr 25 16:28:58 2008 +0930 149326 149327 Xi: remove RemoveOtherCoreGrabs() 149328 149329 PickPointer() returns grabbed devices, so we can't get a double grab anyway. 149330 149331commit be5ff2b8d8a392eb8611e1fcd6da4752d68a7f9d 149332Author: Peter Hutterer <peter@cs.unisa.edu.au> 149333Date: Fri Apr 25 16:25:13 2008 +0930 149334 149335 Xi: remove some superflous code 149336 149337commit 0209e46249b61974a6e3ed54a51cc36dfaf0064f 149338Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149339Date: Thu Apr 24 23:48:11 2008 -0700 149340 149341 XQuartz: More multi-monitor work... reverted Ben's workaround (worked for side-by-side only) and added more debugging. 149342 (cherry picked from commit 515b8b855ac5d2d5aef881053f73b2ad07a6dd2e) 149343 149344commit b65dbd350b8518bb90bed9cdc64dd900bc75d20c 149345Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149346Date: Thu Apr 24 21:14:04 2008 -0700 149347 149348 XQuartz: Cleaned up multi-monitor support. 149349 (cherry picked from commit c05abf0a19b0ef0fc4ace9400a095ce2521456bc) 149350 (cherry picked from commit 9112f290434c246d3e797551aaaf3a89d2006b23) 149351 149352commit b093bf3ec37367172be3b44a04f0f43890ba7d7d 149353Author: Peter Hutterer <peter@cs.unisa.edu.au> 149354Date: Fri Apr 25 15:15:31 2008 +0930 149355 149356 dix: don't search for grabbed devices, PickPointer does it for us now. 149357 149358 Follow-up to 93ee33830778b9ec85ad81496572677a30022b09. 149359 149360commit 6198bad07edc51ff4a264a9361c5e9b6c74647a6 149361Author: Peter Hutterer <peter@cs.unisa.edu.au> 149362Date: Fri Apr 25 15:02:40 2008 +0930 149363 149364 dix: remove misleading (stale) comment 149365 149366commit 93ee33830778b9ec85ad81496572677a30022b09 149367Author: Peter Hutterer <peter@cs.unisa.edu.au> 149368Date: Fri Apr 25 15:02:13 2008 +0930 149369 149370 dix: let PickPointer() return devices that are core-grabbed by the client. 149371 149372 If the client currently holds a core grab on any device, this device has 149373 priority, even over the ClientPointer. 149374 149375commit f5ac98747de921d48f36d4f3f66dbe73e26a6760 149376Author: Peter Hutterer <peter@cs.unisa.edu.au> 149377Date: Fri Apr 25 14:50:24 2008 +0930 149378 149379 Standardise copyright headers for all new files created as part of MPX. 149380 149381commit 3fc67b4205851dcffcc431a07a885828549bd9cf 149382Author: Peter Hutterer <peter@cs.unisa.edu.au> 149383Date: Fri Apr 25 11:37:32 2008 +0930 149384 149385 Xi: remove leftover chpkpair.h file. 149386 149387 Used to contain declarations for ChangePointerKeyboardPairing request 149388 handling. 149389 149390commit 89add4ee986e38b833bb58750b3e5a664efb6f4a 149391Author: Peter Hutterer <peter@cs.unisa.edu.au> 149392Date: Fri Apr 25 11:19:12 2008 +0930 149393 149394 dix: remove obsolete InitWindowPrivates(). 149395 149396commit dcdc66fcfc7957aa7875fbf74fa3a55989a9a187 149397Author: Peter Hutterer <peter@cs.unisa.edu.au> 149398Date: Fri Apr 25 11:14:32 2008 +0930 149399 149400 dix: add mi.h include to stop compiler warning 149401 149402commit 97552413d213337e4bff8c89b06d58a09d722b05 149403Author: Peter Hutterer <peter@cs.unisa.edu.au> 149404Date: Fri Apr 25 11:13:05 2008 +0930 149405 149406 dix: remove "unused variable" compiler warning. 149407 149408commit d9c38e84cc492b931a2238757d438f562946e5bc 149409Author: Peter Hutterer <peter@cs.unisa.edu.au> 149410Date: Fri Apr 25 11:04:02 2008 +0930 149411 149412 dix: remove pairingClient definition. 149413 149414 This variable was used originally to determine which client is allowed to 149415 change the pointer-keyboard pairing. For now, we just let anyone change it and 149416 see how that works out. 149417 149418commit c14b858aeced81ff43723644f2de0f5c43f55755 149419Author: Peter Hutterer <peter@cs.unisa.edu.au> 149420Date: Fri Apr 25 11:02:12 2008 +0930 149421 149422 Xi: fix two compiler warnings 149423 149424 unused variable in exevents.c 149425 implicit declaration in warpdevp.c 149426 149427commit 6015b7a81252cd1729e6f4045f9b0c885c95b183 149428Author: Peter Hutterer <peter@cs.unisa.edu.au> 149429Date: Fri Apr 25 10:41:30 2008 +0930 149430 149431 Xi: handle requests through a dispatch handler. 149432 149433 Same principle as e.g. in the RandR extension, rather than having a load of 149434 if (type == XI_whatever) 149435 use an array of function pointers. 149436 149437commit cb8cb87bc6decf75832c3724687000ecb658226e 149438Author: Peter Hutterer <peter@cs.unisa.edu.au> 149439Date: Fri Apr 25 10:09:50 2008 +0930 149440 149441 Xi: remove unused #define DISPATCH 149442 149443commit 5869e3f5551e7e2d18e06a2f220b77d8c4161db1 149444Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149445Date: Thu Apr 24 19:39:27 2008 -0700 149446 149447 darwin gots /dev/urandom, too yo! 149448 (cherry picked from commit bf0144f38034bc59f108bb2c5270ff37fbe70e10) 149449 149450commit b1ae7c79db78cc3b789701f81328669f9f8ed80f 149451Author: Alan Coopersmith <alan.coopersmith@sun.com> 149452Date: Thu Apr 24 19:02:43 2008 -0700 149453 149454 Bug 14247: If pkg-config can't find openssl.pc, just link with -lcrypto 149455 149456 X.Org bugzilla #14247 <http://bugs.freedesktop.org/show_bug.cgi?id=14247> 149457 149458commit 1daae4574c19a122baf73b6f9457b0d383b1bc66 149459Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149460Date: Thu Apr 24 18:22:05 2008 -0700 149461 149462 Added missing HAVE_LAUNCHD to dix-config.h.in 149463 (cherry picked from commit 1b4b73cefbc2f3e3f6d0cb84ea202f84fb60abb5) 149464 149465commit 63853e4ad211945ed25541223207a78f83766ced 149466Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149467Date: Thu Apr 24 18:21:42 2008 -0700 149468 149469 XQuartz: Compile in missing glx source that we still need 149470 (cherry picked from commit c611335dee267e41dcd1733a6bb5206b102f804e) 149471 149472commit ae982a27e4059fecd4048d245e6aa02f8dcc97d0 149473Author: Dave Airlie <airlied@redhat.com> 149474Date: Thu Apr 24 16:37:29 2008 +1000 149475 149476 glx: test for valid read and draw privates before using them 149477 149478 this should fix a bug where f-spot exiting blows up the X server 149479 149480commit f377141912594f87144d6d7f7fdd279a101d8e6c 149481Author: Peter Hutterer <peter@cs.unisa.edu.au> 149482Date: Thu Apr 24 13:30:28 2008 +0930 149483 149484 Xi: don't attempt to send to a NULL window. 149485 149486 Only applicable when the server comes down/restarts. In this case, 149487 WindowTable[i] may be NULL. Let's not try to send an event then. 149488 149489commit aec485f2dcc87b340759d67b60e7dee7931aaec5 149490Author: Peter Hutterer <peter@cs.unisa.edu.au> 149491Date: Thu Apr 24 13:29:46 2008 +0930 149492 149493 dix: NULL out WindowTable after freeing all the windows. 149494 149495 CloseDownDevices() tries to send PresenceNotify events. If the windows are 149496 already freed, then we are accessing dangling pointers. 149497 149498commit 9ab4e2fd8eaa87dbd16835affb1aa54dcb1a619e 149499Author: Peter Hutterer <peter@cs.unisa.edu.au> 149500Date: Wed Apr 23 11:38:08 2008 +0930 149501 149502 xfree86: don't free the config-file related information in DIDR. #15645 149503 149504 In DeleteInputDeviceRequest, leave the conf_idev (which is shared with 149505 xf86ConfigLayout.input) alone for devices that were specified in the 149506 ServerLayout section of the config file. This way, in the next server 149507 generation we are left with what was the original config and can thus re-init 149508 the devices. 149509 149510 This is an addon to 6d22a9615a0e6ab3d00b0bcb22ff001b6ece02ae, an attempt to 149511 fix Bug 14418. 149512 149513 X.Org Bug 15645 <https://bugs.freedesktop.org/show_bug.cgi?id=15645> 149514 X.Org Bug 14418 <https://bugs.freedesktop.org/show_bug.cgi?id=15645> 149515 149516commit 7dc40c8eca90ec1bfab84b6f54418b64c0e62d63 149517Author: Peter Hutterer <peter@cs.unisa.edu.au> 149518Date: Thu Apr 24 12:47:29 2008 +0930 149519 149520 xfree86: Don't free master devices. 149521 149522 The previous check works in the master-branch, but doesn't work with MPX. We 149523 actually copy the SD's information into the MDs public.devicePrivate, so we 149524 need to explicitly check whether a device is a MD before freeing the module. 149525 149526commit cc13f87cd8ac54223fdb0b3d3c043dc9b9be5a42 149527Author: Peter Hutterer <peter@cs.unisa.edu.au> 149528Date: Thu Apr 24 11:43:47 2008 +0930 149529 149530 Xext: XGE: change function definition to have return value on separate line. 149531 149532 This seems to be the common style in most parts of the server. 149533 149534commit d0890c40b9a9c7965c08608e6950c078a29aac16 149535Author: Peter Hutterer <peter@cs.unisa.edu.au> 149536Date: Thu Apr 24 11:32:36 2008 +0930 149537 149538 Xext: add a few lines of comments to XGE. 149539 149540commit 745b90cde1007383ec8c887f02439a34ab427f31 149541Author: Peter Hutterer <peter@cs.unisa.edu.au> 149542Date: Thu Apr 24 11:27:22 2008 +0930 149543 149544 Xext: re-do the XGE versioning handling. 149545 149546 Basically the same approach RandR takes. Remember which one the client 149547 requested, send back the one the server supports. Also divide XGE server 149548 version (now defined in geext.c) and the client's version (still in the 149549 protocol definition). 149550 149551commit 5f3e5b3462bb02e828c70d0e1890b5a83d399d42 149552Author: Peter Hutterer <peter@cs.unisa.edu.au> 149553Date: Thu Apr 24 11:03:23 2008 +0930 149554 149555 Xext: remove trailing whitespaces in geext.{c|h} 149556 149557commit d09c520b322ba5c5f4d6b630a7c0c62e56732f82 149558Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149559Date: Wed Apr 23 10:22:34 2008 -0700 149560 149561 Moved Apple GL bits into our DDX subtree 149562 (cherry picked from commit c9356f32892978faecb30f7b7af7488820ce37ff) 149563 149564commit b05b416c622063b84747702a54ffd9a802d6fc11 149565Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149566Date: Wed Apr 23 09:59:48 2008 -0700 149567 149568 XQuartz: No need to include indirect.c a second time 149569 (cherry picked from commit 41ed532525da85d1bd2c20af5ffa28ac6d1f4996) 149570 149571commit 565492c70a280d7e749500c5c0073bdcaf175c86 149572Author: Ben Byer <bbyer@apple.com> 149573Date: Wed Apr 23 06:00:57 2008 -0700 149574 149575 xquartz: build fixes for GLX/DRI2 fallout (note: not guaranteed) 149576 (cherry picked from commit 85a5796b667461bf1fafc68c07b3a704cd8efd5b) 149577 149578commit 6f6505db51763d902142a8161e71cad44f52560c 149579Author: Ben Byer <bbyer@apple.com> 149580Date: Wed Apr 23 03:43:51 2008 -0700 149581 149582 glx / xquartz: we still need glxcontentmodes.[ch] kthx 149583 (cherry picked from commit 27545fd37f8c703b72f7cfc329778abc1ceabfa0) 149584 149585commit 4e3a8af5751ff88c0cadd612821b0ef349d250c5 149586Author: Ben Byer <bbyer@apple.com> 149587Date: Wed Apr 23 03:39:41 2008 -0700 149588 149589 xquartz: clean up linker line for main binary 149590 (cherry picked from commit 821d7400f2ff917497b2ee58ceef2b69c2d47ec7) 149591 149592commit 3685b171da51f853bfdb312d36622dfa03ee75be 149593Author: Ben Byer <bbyer@apple.com> 149594Date: Tue Apr 22 23:53:23 2008 -0700 149595 149596 xquartz: remove vestigial aglGlx.c 149597 (cherry picked from commit 40bd041e4255f26fcbdf0831e68619ae0f46ab39) 149598 149599commit f6e22d69af6bc8f63c3a46535a09e217696a679f 149600Author: Hans de Goede <j.w.r.degoede@hhs.nl> 149601Date: Wed Apr 23 12:28:30 2008 -0400 149602 149603 Prefer glxvisuals with stencil buffer for default visuals 149604 149605 The first fbconfig which has a depthbuffer > 0 and doublebuf is choosen 149606 when associating fbconfigs with the visuals, indepenent of stencil bits. 149607 This happens to work ok on intel as there all fbconfigs with a 149608 depthbuffer > 0 also have stencil bits. 149609 149610 This patch fixes this by first trying to get a fbconfig for default X visuals 149611 with both stencilbuf, depthbuf and doublebuffering, and if that fails fallback 149612 to trying to get one with only a depthbuf and doublebuffering. 149613 149614commit 00effad583713e882c3f2518bcd3da51bf4db716 149615Author: George Sapountzis <gsap7@yahoo.gr> 149616Date: Wed Apr 23 17:46:30 2008 +0300 149617 149618 xephyr: XEPHYR_DRI is identical to XF86DRI 149619 149620commit 6a9f7f28ec455d2879ca1a315ce77c48af49e7f4 149621Author: George Sapountzis <gsap7@yahoo.gr> 149622Date: Wed Apr 23 17:46:15 2008 +0300 149623 149624 xephyr: sync with mesa 149625 149626commit 7ffc68c5cc7108c9e56af25f9f9b4398f5b027df 149627Author: George Sapountzis <gsap7@yahoo.gr> 149628Date: Wed Apr 23 17:45:54 2008 +0300 149629 149630 xephyr: libGL is client lib 149631 149632commit a27c244d72603e27766ff3900ab28fe094f7397e 149633Author: George Sapountzis <gsap7@yahoo.gr> 149634Date: Wed Apr 23 17:45:37 2008 +0300 149635 149636 xephyr: Makefile cleanup 2 149637 149638commit d4d19b2f22002b1579438cb9cf08dd580c7ec005 149639Author: George Sapountzis <gsap7@yahoo.gr> 149640Date: Wed Apr 23 17:45:19 2008 +0300 149641 149642 xephyr: Makefile cleanup 1 149643 149644commit cdc4571b580a8f4fd279404215bff0fb9a5b4816 149645Author: George Sapountzis <gsap7@yahoo.gr> 149646Date: Sun Apr 20 17:06:26 2008 +0300 149647 149648 drop TOGCUP remnants 149649 149650commit 6cd9287aed3c128d9c10b9b042ba7b864143d522 149651Author: George Sapountzis <gsap7@yahoo.gr> 149652Date: Sun Apr 20 16:58:51 2008 +0300 149653 149654 drop EVI remnants 149655 149656commit a88c6b66465c039bf92a54a496516fee4d900784 149657Author: George Sapountzis <gsap7@yahoo.gr> 149658Date: Fri Apr 18 17:47:16 2008 +0300 149659 149660 glcore: drop GLcore (files) 149661 149662commit dd7a53f8f21e41e1ab43f7e684ba586abc97ef7e 149663Author: George Sapountzis <gsap7@yahoo.gr> 149664Date: Fri Apr 18 17:47:01 2008 +0300 149665 149666 glcore: drop GLcore (build system) 149667 149668commit 567d389d47dee233a973b101e04ce41c47a68f34 149669Author: George Sapountzis <gsap7@yahoo.gr> 149670Date: Fri Apr 18 17:46:18 2008 +0300 149671 149672 glcore: build from mesa, dlopen from xorg 149673 149674 * The GLcore interface is disposable 149675 * GLcore is installed in DRI_DRIVER_INSTALL_DIR which is overloaded for 149676 GLX_PROVIDER_INSTALL_DIR 149677 149678commit efb723e166e5fa89e90c7b400fb4c7979b1f50c0 149679Author: George Sapountzis <gsap7@yahoo.gr> 149680Date: Fri Apr 18 17:45:53 2008 +0300 149681 149682 glcore: prepare for dynamic loading 149683 149684 glcore gets linked with -ldl, -lpthread for s3tc and glapi 149685 xserver needs 149686 DLOPEN_LIBS - to dlopen the glcore dso 149687 LD_EXPORT_SYMBOLS_FLAG - to export symbols for glcore to use 149688 149689 the ld flag is added to kdrive only when GLX is enabled, the net overhead for 149690 Xephyr is ~155KB, could be reduced with --dynamic-list. 149691 149692commit fbad87f2ae9f97fcb43546b0fa35f1100415dfec 149693Author: Julien Cristau <jcristau@debian.org> 149694Date: Tue Apr 22 23:50:11 2008 +0200 149695 149696 autoconfig: don't call closedir() when opendir() failed 149697 149698 If opendir() fails, return from matchDriverFromFiles() immediately. 149699 Ubuntu bug 217647. 149700 149701commit 76381092e8f650ec7d1f058fa4c8a7348893f775 149702Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149703Date: Wed Apr 23 02:22:14 2008 -0700 149704 149705 XQuartz: Make sure QuartzAudioInit() gets run. 149706 (cherry picked from commit bb3d034675b70e22e78df5554cab0ec2a3d913d0) 149707 (cherry picked from commit 824b31c7f8144a67a320442abd3d854e99d2bfe2) 149708 149709commit 275cdc1c74b7e43ecd931d312469fecc8d998ed1 149710Author: Peter Hutterer <peter@cs.unisa.edu.au> 149711Date: Wed Apr 23 16:16:15 2008 +0930 149712 149713 xfree86: pass correct value to mieqEnqueue (merge detritus I guess) 149714 149715commit c3659cb414ef05da8fa09009b2b82a3deeeb4f3a 149716Author: Peter Hutterer <peter@cs.unisa.edu.au> 149717Date: Tue Apr 22 21:54:02 2008 +0930 149718 149719 dix: always send FocusIn events if mode is NotifyUngrab. 149720 149721 In the case of a NotifyUngrab, the flag for the device may already be set but 149722 we still need to send the FocusIn event. 149723 149724commit 00815b3e5223e822f306db45cd4884a22ac9f7ed 149725Author: Julien Cristau <jcristau@debian.org> 149726Date: Mon Apr 21 14:34:39 2008 +0200 149727 149728 Don't set DRI2=yes if we can't find dri2proto or libdrm 149729 149730commit e77f65768efbf05cdf363a2f41f036f74eaa45de 149731Author: Ross Burton <ross@burtonini.com> 149732Date: Tue Apr 22 18:07:46 2008 +0100 149733 149734 Reuse the existing framebuffer mode in kdrive/fbdev 149735 149736 When starting up kdrive/fbdev, if the current framebuffer mode is sensible use 149737 that unless told otherwise. 149738 149739 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 149740 149741commit c4fd1121531b0cba1a3e90fa747871d784365c7e 149742Author: Ross Burton <ross@burtonini.com> 149743Date: Tue Apr 22 18:05:53 2008 +0100 149744 149745 Add mediumraw support to the linux kdrive driver 149746 149747 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 149748 149749commit 455383db95618a05ebdbeae78423e08065f0e14e 149750Author: Ross Burton <ross@burtonini.com> 149751Date: Tue Apr 22 18:04:31 2008 +0100 149752 149753 Enable the epson kdrive driver 149754 149755 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 149756 149757commit 82b37d35af13a9f402755e167493ab256b664b9c 149758Author: Ross Burton <ross@burtonini.com> 149759Date: Tue Apr 22 16:38:14 2008 +0100 149760 149761 Add support for AVR32 149762 149763commit 20a90bef8b4993f06cf76ad05e3d4c974e1614f6 149764Author: Ross Burton <ross@burtonini.com> 149765Date: Tue Apr 22 16:37:02 2008 +0100 149766 149767 Add a default touchscreen path 149768 149769 When enabling the touchscreen, open /dev/input/touchscreen0 if no path was 149770 specified. 149771 149772commit b44e89f4683ffcfd75eaf39f1f37d7461db44689 149773Author: Ross Burton <ross@burtonini.com> 149774Date: Tue Apr 22 16:28:39 2008 +0100 149775 149776 Update to tslib 1.0 149777 149778 Update the pkgconfig checks to tslib 1.0, and fix the LIBS definitions. 149779 149780commit 3848422d2354b7a5302fda92b05b0d728190e050 149781Author: Ross Burton <ross@burtonini.com> 149782Date: Tue Apr 22 16:25:23 2008 +0100 149783 149784 Fix build when XKB is disabled. 149785 149786commit 744d0cfda74f8283801cc2d6c5eda48402455bc3 149787Author: Ben Byer <bbyer@apple.com> 149788Date: Mon Apr 21 22:06:38 2008 -0700 149789 149790 add missing DARWIN_GLX_LIBS in configure.ac 149791 (cherry picked from commit a033c0b3dbb3b963261faa39f0236457cb00ff44) 149792 149793commit 8190ef87547b704848231bde10b1cdffc6442790 149794Merge: 179a082c2 2ddbfd345 149795Author: Peter Hutterer <peter@cs.unisa.edu.au> 149796Date: Tue Apr 22 18:04:05 2008 +0930 149797 149798 Merge branch 'master' into mpx 149799 149800 Conflicts: 149801 149802 Xext/EVI.c 149803 Xext/appgroup.c 149804 Xext/cup.c 149805 Xext/mitmisc.c 149806 Xext/sampleEVI.c 149807 dix/window.c 149808 149809commit 179a082c26f9e562492ee2e59e7f44f949f39f9c 149810Author: Peter Hutterer <peter@cs.unisa.edu.au> 149811Date: Fri Apr 18 21:13:10 2008 +0930 149812 149813 Xext: route event through master if required (XTestFakeInput) 149814 149815commit fb784d99c521823339bf00b70b9824f735d88875 149816Author: Peter Hutterer <peter@cs.unisa.edu.au> 149817Date: Fri Apr 18 20:57:42 2008 +0930 149818 149819 mi: fix typo in comment. 149820 149821commit 8209fdbc7c7f2dd068b30ff184b5fbf00db78686 149822Author: Peter Hutterer <peter@cs.unisa.edu.au> 149823Date: Fri Apr 18 20:53:57 2008 +0930 149824 149825 Xext: Update sprite on fake input only for MDs and floating SDs. 149826 149827commit 2ddbfd345786aa39b6ccaed82a1ca5c145284ee3 149828Author: Peter Hutterer <peter@cs.unisa.edu.au> 149829Date: Tue Apr 22 16:31:44 2008 +0930 149830 149831 xnest: re-enable XKB support. #10015 149832 149833 XKB was disabled in 08928afb0500d46b0caa0a1d1244dee2ed80e6a0, with the comment 149834 "Disable XKB, as we can't yet use it". Seems like "yet" is over, running GNOME 149835 and changing XKB settings seems to work in Xnest now. 149836 149837 X.Org Bug 10015 <https://bugs.freedesktop.org/show_bug.cgi?id=10015> 149838 149839commit a8b8700c7345b89953c8b63cb5c347a95e6ab988 149840Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149841Date: Mon Apr 21 23:44:37 2008 -0700 149842 149843 XQuartz: Fixed cmd-tab to bring all windows forward. 149844 (cherry picked from commit e48e2ce931228f4dfa36c39b8ec8c72a58025b1b) 149845 149846commit a368ab757edf36ed7bbda023673d28883ce11231 149847Author: Dave Airlie <airlied@linux.ie> 149848Date: Tue Apr 22 15:13:57 2008 +1000 149849 149850 GLcore: make googleearth not crash the server on sw-rendering. 149851 149852 I don't think this is the 100% correct answer as I get log spam saying 149853 (EE) DoSwapInterval: cx = 0x98b8998, GLX screen = 0x96dd780 149854 (EE) AIGLX: cx->pGlxScreen->swapInterval == NULL 149855 149856 but thats better than X exiting in my book. 149857 149858commit 449723510a1f9d024e23d9eb33795cac27f9443e 149859Author: Egbert Eich <eich@pdx.freedesktop.org> 149860Date: Tue Apr 22 13:30:03 2008 +0930 149861 149862 xkb: use the correct device instead of an uninitialised "dev". #15614 149863 149864 X.Org Bug 15614 <http://bugs.freedesktop.org/show_bug.cgi?id=15614> 149865 149866 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 149867 149868commit ba87c25321c3378fd1ad0c55dcb0af0a6e82a540 149869Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149870Date: Mon Apr 21 20:03:08 2008 -0700 149871 149872 Make rootless use dixLookupClient rather than deprecated LookupClient. 149873 (cherry picked from commit 582b5b01f9697b66489ea906a2ecb8bfc5915571) 149874 149875commit 8822110d7d6b684f373fc883aeb7cab9734e9ddb 149876Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149877Date: Mon Apr 21 20:08:05 2008 -0700 149878 149879 Fixed dixLookupClient to work with client=NullClient as it did in the 1.3 branch 149880 (cherry picked from commit e41ccc64702f856e5e09dfa652fe73c14b8a0225) 149881 (cherry picked from commit ce5a5f93990647de85e535734ee6bb430ad591cb) 149882 149883commit 5ffb6a2fe8db5871eaf26b8535af1588c43f33d3 149884Author: Ben Byer <bbyer@apple.com> 149885Date: Mon Apr 21 19:55:54 2008 -0700 149886 149887 Nuke a call to deprecated LookupClient, and hopefully prevent a 149888 null-pointer dereference, too! 149889 (cherry picked from commit 3d28e9f953709914e18807bc74c241333671cb30) 149890 149891commit 590688131d89595bdc78ca562ee88df86c9012a6 149892Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149893Date: Sun Apr 20 22:18:45 2008 -0700 149894 149895 XQuartz: Fixed quit dialog to be more conforming with HIG. 149896 (cherry picked from commit 14c6b837bb03bd0956f90882f550847f13d0ca09) 149897 149898commit d20b3ac22d960fa44632cc4a14be079daa2d5a33 149899Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149900Date: Sun Apr 20 22:18:19 2008 -0700 149901 149902 .gitignore: Added Xquartz stuff 149903 (cherry picked from commit 0a9a3bec2de8d1f442493e13cf9f039902a4928f) 149904 149905commit 570b0dca261920c9b01b4eb11fe9b9987b1c636a 149906Author: Ben Byer <bbyer@apple.com> 149907Date: Mon Apr 21 17:52:10 2008 -0700 149908 149909 Fix for pointer-offset issue when using a multi-display environment on X11.app. 149910 (cherry picked from commit 9a7e14286ced55c5e2a4512e2629e03836443009) 149911 149912commit 3f081b4de55e1378728a24d069bf06575ffca2d8 149913Author: Tilman Sauerbeck <tilman@code-monkey.de> 149914Date: Mon Apr 21 11:22:07 2008 +0200 149915 149916 EXA: Set pixmap->accel_blocked on the screen pixmap, too. 149917 149918commit 26c1801a27b81fdd988d5bd210ba0e76ecc274ae 149919Author: Tilman Sauerbeck <tilman@code-monkey.de> 149920Date: Mon Apr 21 11:03:27 2008 +0200 149921 149922 EXA: Update sys_pitch/fb_pitch in exaModifyPixmapHeader. 149923 149924 exaModifyPixmapHeader now also only evaluates arguments that have a 149925 meaningful value. 149926 149927commit 4fa89fbe18c929e0d36305ab47e7e17841309ffd 149928Author: Jordan Crouse <jordan.crouse@amd.com> 149929Date: Mon Apr 21 12:09:00 2008 -0600 149930 149931 xf86: Change AutoConfig driver for PCI ID 022:2091 to 'geode' 149932 149933commit 6c95fae1e9d6b0eb64bc78eced05a6e9f5faf02e 149934Author: Michel Dänzer <michel@tungstengraphics.com> 149935Date: Mon Apr 21 10:45:11 2008 +0200 149936 149937 EXA: Offscreen memory eviction improvements. 149938 149939 * Make sure available areas are considered to have no eviction cost. This seems 149940 to help for https://bugs.freedesktop.org/show_bug.cgi?id=15513 but I'm afraid 149941 that may just be coincidence. 149942 * Only calculate eviction cost of each area once for each eviction pass. 149943 Safeguard against potential (though unlikely) division by zero. 149944 * Cosmetic enhancements: Name eviction cost related variables 'cost' instead of 149945 'score' to emphasize that smaller values are better, update Doxygen file 149946 comment to the way eviction works now. 149947 149948commit 40c6be1408a1f0b236fdb28af27ae18aea0d578f 149949Author: Julien Cristau <jcristau@debian.org> 149950Date: Sun Apr 20 20:36:44 2008 +0200 149951 149952 Minor xorg.conf manpage fixups 149953 149954 Use __libmansuffix__ instead of __oslibmansuffix__ which isn't getting 149955 replaced, and rewrap some text to get __xservername__ replaced in the 149956 description of Option "Accel" (cpp doesn't like the preceding quote). 149957 149958commit 4bcfed2f9cf5dbf682d3bc98873ba97c4efdff44 149959Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149960Date: Sun Apr 20 08:35:42 2008 -0700 149961 149962 Revert "Removed XWin DDX" 149963 149964 This reverts commit 6550078b0925f754e3eec3bbce94dbfe5de8c419. 149965 149966 Doctor, I'm starting to get a pulse... 149967 149968commit 14396fdebac1868df17559220ed7aaa34c34251e 149969Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149970Date: Sat Apr 19 23:31:05 2008 -0700 149971 149972 XQuartz: Cleanup turning off COMPOSITE 149973 (cherry picked from commit 8f920fca6f9149185649d52569d33bf81b6c6857) 149974 149975commit d13828797fe22856b07e08a55d2b9375902194bf 149976Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149977Date: Fri Apr 18 20:38:31 2008 -0700 149978 149979 XQuartz: Cleaned up some casting to get rid of compiler warnings 149980 (cherry picked from commit 6f1c85b96550adf0bc34efb6ca649b87bcc1b18c) 149981 149982commit 6550078b0925f754e3eec3bbce94dbfe5de8c419 149983Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149984Date: Sat Apr 19 09:29:46 2008 -0700 149985 149986 Removed XWin DDX 149987 149988commit 587c010a1cd733fded4d49dc339df0634bda8be6 149989Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 149990Date: Sat Apr 19 09:27:21 2008 -0700 149991 149992 Rootless: Kill off rlAccel 149993 149994commit d3d00d92586c3e1cbc88087c930b65c8b3832fcc 149995Author: Tilman Sauerbeck <tilman@code-monkey.de> 149996Date: Sat Apr 19 12:54:40 2008 +0200 149997 149998 Removed fbpseudocolor.h from sdk_HEADERS. 149999 150000commit 5bdc4198795ffd011bb07cffe3817e4cded87f60 150001Author: Adam Jackson <ajax@redhat.com> 150002Date: Sat Apr 19 04:06:19 2008 -0400 150003 150004 Remove fbpseudocolor 150005 150006 "An experimental pseudocolor emulation layer. Not fully completed, 150007 currently only works for 16bpp." That was almost four years ago. 150008 It still doesn't work, only one driver even attempts to use it, it 150009 contains an ad-hoc implementation of damage, and should really be 150010 done up in Composite now anyway. 150011 150012commit 60ff56050b64183cb6e58f54223c8a3ddc2e704b 150013Author: Daniel Stone <daniel@fooishbar.org> 150014Date: Fri Apr 18 22:17:53 2008 -0700 150015 150016 Revert "Optimize dixLookupPrivate for repeated lookups of the same private." 150017 150018 The patch was wildly unsafe for SIGIO, and made everything full of 150019 crashy crashy fail. 150020 150021 This reverts commit 9b30cc524867a0ad3d0d2227e167f4284830ab4e. 150022 150023commit ed9dabb47c467dbf49836b631d5d6bda4b0d98b0 150024Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150025Date: Fri Apr 18 20:30:43 2008 -0700 150026 150027 Last of the spam... I promise... 150028 (cherry picked from commit 45ebee4f729b148a75e925a4863b4eb850c88f8e) 150029 150030commit 49f2bb4681fdee9e45f952ef0ac9c34a090117de 150031Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150032Date: Fri Apr 18 20:25:38 2008 -0700 150033 150034 XQuartz: More sanitization of the namespace 150035 (cherry picked from commit bc50d41f9d1aec04f0de0478cbd5036f1fe9b81e) 150036 150037commit c2f0d020b5d7950267aa3df391a7a72b9ae5883b 150038Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150039Date: Fri Apr 18 20:10:57 2008 -0700 150040 150041 XQuartz: Removed unneccessary include 150042 (cherry picked from commit 45ff59e69eddfcceafced31cf6e73e381d0f6914) 150043 150044commit 5183fea6d38de4bcf657e9c2a983dfd81a2a223f 150045Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150046Date: Fri Apr 18 20:06:17 2008 -0700 150047 150048 XQuartz: Handled sanitization of namespace better 150049 (cherry picked from commit 8cb23d672177da919257c885804cecd18cf9af88) 150050 150051commit edd3fb784bad893550ee270e0a09f22f99783cf5 150052Author: Ben Byer <bbyer@bbyer.local> 150053Date: Fri Apr 18 17:17:01 2008 -0700 150054 150055 random flailing 150056 (cherry picked from commit 7fb9b2dc615a3bd1a3c087438af7a8b88265cfaa) 150057 150058commit a7503615a6893749d512f75d37646273f31b9dbf 150059Author: Adam Jackson <ajax@redhat.com> 150060Date: Fri Apr 18 19:56:41 2008 -0400 150061 150062 Death to TOG-CUP. 150063 150064 If you still care about 8bpp visuals that much, fix Composite to provide 150065 synthetic visuals. 150066 150067commit 4da9ec16e9725ebb9817b49e33ea1035b6aff09a 150068Author: Adam Jackson <ajax@redhat.com> 150069Date: Fri Apr 18 19:54:09 2008 -0400 150070 150071 Remove appgroup mentions from configure.ac 150072 150073commit 25827fde68d3bb02a2b7e05fae53a1d97edf1f76 150074Author: Eric Anholt <eric@anholt.net> 150075Date: Fri Apr 18 15:32:04 2008 -0700 150076 150077 Nuke the MIT-SUNDRY-NONSTANDARD extension. 150078 150079 This extension provided bug-compatibility with pre-X11R6, but has been 150080 stubbed out in our server since 2006 to return BadRequest when you actually 150081 asked for it. 150082 150083commit 13adef8a17d8815f4db2aaac30ae04438e125343 150084Author: Adam Jackson <ajax@redhat.com> 150085Date: Fri Apr 18 19:01:06 2008 -0400 150086 150087 Finish deleting EVI 150088 150089commit eafaf40fb3368ca7e4cf48336fdb7a6c9f536bfa 150090Author: Adam Jackson <ajax@redhat.com> 150091Date: Fri Apr 18 18:50:05 2008 -0400 150092 150093 Death to APPGROUP. 150094 150095commit f6617b4127125516583f321c961d70f762f728be 150096Author: Adam Jackson <ajax@redhat.com> 150097Date: Fri Apr 18 18:28:01 2008 -0400 150098 150099 Death to Extended Visual Information. 150100 150101commit 3b93631e59ca4d312d318eac4015e0a79ad6351f 150102Author: Peter Hutterer <peter@cs.unisa.edu.au> 150103Date: Fri Apr 18 11:31:54 2008 +0930 150104 150105 dix: remove coreOnly check. Core pointer must generate XI events now. 150106 150107 This flag was only used when an event is generated by Warp[Device]Pointer. 150108 Since the VCP now happily generates core events, this flag is obsolete. 150109 150110commit e0eaf8e5e3fa7a11c087851dff93f50f6907c4a5 150111Author: Peter Hutterer <peter@cs.unisa.edu.au> 150112Date: Fri Apr 18 10:57:47 2008 +0930 150113 150114 Xext: Let XTestFakeInput update the sprite for XI events. 150115 150116 Since XI devices can have their own sprite now, we need to update the sprite 150117 coordinates too when processing an XI event. 150118 Note: This doesn't deal with the device hierarchy correctly yet. 150119 150120commit cb0168b4ac5c59cdce6f0a6d89ddd9f30d93b5f3 150121Author: Peter Hutterer <peter@cs.unisa.edu.au> 150122Date: Fri Apr 18 09:48:06 2008 +0930 150123 150124 Xext: xtest.c: death to tabs+spaces indendation, remove #define XINPUT 150125 150126commit eebdf69e9a52e071e0467a1f99de135df4d1eabc 150127Author: Peter Hutterer <peter@cs.unisa.edu.au> 150128Date: Fri Apr 18 09:22:37 2008 +0930 150129 150130 xkb: mixing up src and dst in a memcpy is suboptimal. 150131 150132commit aa6687322de85a2d1025a3ae851fb290a089b2d3 150133Author: Peter Hutterer <peter@cs.unisa.edu.au> 150134Date: Tue Apr 15 23:03:06 2008 +0930 150135 150136 Xi: when copying button classes, copy xkb_acts as well. 150137 150138 This should be the last piece in the quest for the class copy. 150139 150140commit c14f5dc237a31b13d98ae2d0d6143bd91083cf13 150141Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150142Date: Thu Apr 17 20:21:45 2008 -0700 150143 150144 XQuartz: Forgot to commit xprEvent.[hc] ... 150145 (cherry picked from commit 70e543baf2508d636f01b2b7e8cb05172195b68c) 150146 150147commit 22bb7608a025a4ec0f442637810b20e2cb0b0820 150148Author: Ben Byer <bbyer@apple.com> 150149Date: Thu Apr 17 17:04:08 2008 -0700 150150 150151 Added XKB support for Xquartz 150152 (cherry picked from commit 56dc1215202746590dbe8758411f47e8876e1317) 150153 150154commit 652479dba38470273313dc46f17e3bcb1bc5e383 150155Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150156Date: Thu Apr 17 15:49:13 2008 -0700 150157 150158 XQuartz: Moved some rootless-specific cruft into xpr 150159 (cherry picked from commit 31625cc03b58317120c2ac7877e227e2322e1de8) 150160 150161commit 2a1ba20af98c0e9a6a7f1a50d32058dcc9759c21 150162Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150163Date: Thu Apr 17 15:23:00 2008 -0700 150164 150165 XQuartz: Use a mutex to ensure we only have one thread calling mieqEnqueue at a time. 150166 (cherry picked from commit 7b087c965bce9f440ab5233d6383aa4a7de969b8) 150167 150168commit 55f80d754525398378de1ef28aa562bd29ee750f 150169Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150170Date: Thu Apr 17 14:21:31 2008 -0700 150171 150172 XQuartz: A little more debugging output from threadSafety 150173 (cherry picked from commit f6fbdbf838ab77c3a4635f0b2356b1bbb060ff5b) 150174 150175commit 0d61f6fca1efeb4f68488e323d1c0508b9b7a711 150176Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150177Date: Thu Apr 17 13:17:58 2008 -0700 150178 150179 XQuartz: Fixed some missing prototypes 150180 (cherry picked from commit 95056afc562cfe58b116f5c36e4624018e79ff4a) 150181 150182commit dbd4c031565d269fef90af23386ff045ec78688c 150183Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150184Date: Thu Apr 17 13:12:56 2008 -0700 150185 150186 XQuartz: Added framework for asserting which thread we're in. 150187 (cherry picked from commit 00beb982510e7a82d77e1f1d43e77c84d7bf74c2) 150188 150189commit fa0645b452cbebd1800a63f1c95cb77fef4ab211 150190Author: Ben Byer <bbyer@apple.com> 150191Date: Thu Apr 17 12:27:12 2008 -0700 150192 150193 removed Xquartz debugging code that leaked into master. Our Bad. 150194 150195commit a3d40f0549f6c6f49fffc286bcdaad758fa92367 150196Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150197Date: Thu Apr 17 11:56:48 2008 -0700 150198 150199 XQuartz: Include version info for CrashReporter 150200 (cherry picked from commit b4992755c3e29086c5939683c38fa8fd7d2e6754) 150201 150202commit 6d11712c2a35b243c19eea3b26622d18c2446dbe 150203Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150204Date: Thu Apr 17 11:06:54 2008 -0700 150205 150206 XQuartz: Use strerror(errno)... cause I like text more than grepping header files 150207 (cherry picked from commit 1b4c37d8f9b517fbec5b94ed4e4a5e86a31472a5) 150208 150209commit dcf4f917cc9488de72711255bbb030d9aa8f8bfb 150210Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150211Date: Thu Apr 17 11:03:31 2008 -0700 150212 150213 merged darwinKeyboard.[ch] into quartzKeyboard 150214 (cherry picked from commit 57bb07320908b74facea0a97822bb19ed6f960a9) 150215 150216commit 582397cd024c68df65ab9dececd6d2c40a5a261b 150217Author: Ben Byer <bbyer@apple.com> 150218Date: Thu Apr 17 02:31:53 2008 -0700 150219 150220 oops, missed a spot 150221 (cherry picked from commit 19872a6aeb8ee9cb0e33e4b4ffd794c9dbefe0cf) 150222 150223commit 0bd1c369cce05d5a4da5e3fd7033aea8c68460ec 150224Author: Ben Byer <bbyer@apple.com> 150225Date: Thu Apr 17 02:30:36 2008 -0700 150226 150227 formatting cleanup 150228 (cherry picked from commit 769acd29348abf9e5b0bebfca6ae695d345f3077) 150229 150230commit 700e14c22616b209867e4ea4d1811e53ca996164 150231Author: Ben Byer <bbyer@apple.com> 150232Date: Thu Apr 17 02:21:33 2008 -0700 150233 150234 delete debugging spew 150235 (cherry picked from commit f04f3af86a91d0cafbc86a0d71aeb0599d685f07) 150236 150237commit a440eebf2541ae0bb06bf65281b5facff2f04e00 150238Author: Ben Byer <bbyer@apple.com> 150239Date: Thu Apr 17 02:21:11 2008 -0700 150240 150241 add support for horizontal scrolling (buttons 6 and 7) 150242 (cherry picked from commit f525a4a432ebd0545ad1dd0a7ad84ad3e47e8b61) 150243 150244commit 612e901ef6aa3edc54b39e55e8040cda0e5ab7b6 150245Author: Ben Byer <bbyer@apple.com> 150246Date: Thu Apr 17 01:32:56 2008 -0700 150247 150248 enable keyboard map debugging -- it's going to x11-debug.txt, anyway ... so no harm 150249 (cherry picked from commit ab662c736e0654e2b4347091f0d9e87f26034216) 150250 150251commit 5bdfbfbedcbd9ff61cbb0b678cbf7ce7889a5826 150252Author: Ben Byer <bbyer@apple.com> 150253Date: Thu Apr 17 01:29:46 2008 -0700 150254 150255 darwinKeyboard: refactor slightly so that we're not cutting and pasting code from dix, kthx 150256 (cherry picked from commit a8a090b853e811b9843a5732572cbbe542224f32) 150257 150258commit 58e42683c9e998f6b8a55d5653b9caec7b6acf96 150259Author: Ben Byer <bbyer@apple.com> 150260Date: Thu Apr 17 00:19:56 2008 -0700 150261 150262 hack to Xquartz to prevent xmodmap from wiping out our valid modmap, per daniels 150263 (cherry picked from commit cab54466a61281cfafc12825017c23d720cd75f4) 150264 150265commit cd3470a0cffbd6b8cec7c44227b33307c9e227ae 150266Author: Ben Byer <bbyer@apple.com> 150267Date: Wed Apr 16 22:48:54 2008 -0700 150268 150269 kludge: miEqEnqueue wants a device, even if we're passing custom messages, 150270 so give it one 150271 (cherry picked from commit a494ff04b2a14470eaf5a23c7cf6dbdea182c6d1) 150272 150273commit 966ae1781f3ca563e15a9a1b8cab6fab94e07fe9 150274Author: Drew Parsons <dparsons@debian.org> 150275Date: Mon Mar 10 22:54:49 2008 +1100 150276 150277 Create dix/libXpdix.la for Xprint-specific build of libdix.la 150278 150279 (cherry picked from commit 4e2c6dbabdbbaaca213fd08edd422de15d0900cc) 150280 150281 required because of commit 7c0709a736c0f3aa011de67dd2c2962585ab146e, 150282 which made requestingClient in dix specific to Xprint only. 150283 Add to XPRINT_LIBS in hw/xprint/Makefile.am in front of 150284 $(XSERVER_LIBS) to override definitions in libdix.la for standard xservers. 150285 150286 Follows 571206832d454771e3c638c7515767958365c19c (providing -DXPRINT 150287 to xprint subdirs). 150288 150289 Note it may be possible to restructure the code so that 150290 requestingClient is stored elsewhere than in dix. See discussions 150291 following http://lists.freedesktop.org/archives/xorg/2008-March/033844.html 150292 If this is done it may be possible to revert this commit (if not 571206...). 150293 150294commit 571206832d454771e3c638c7515767958365c19c 150295Author: Drew Parsons <dparsons@debian.org> 150296Date: Mon Mar 10 13:48:05 2008 +1100 150297 150298 Define XPRINT in XPRINT_CFLAGS (configure.ac) 150299 150300 -DXPRINT had only been set for Xprt in hw/xprint/Makefile.am 150301 After commit 7c0709a736c0f3aa011de67dd2c2962585ab146e it is also 150302 required for ps/PsArea.c and PsFonts.c to ensure ‘requestingClient’ is 150303 defined, so make it a global Xprint definition in configure.ac. 150304 (cherry picked from commit 28a6719fd486d9a9cecad0b057d9ea7c59c66055) 150305 150306commit 9b30cc524867a0ad3d0d2227e167f4284830ab4e 150307Author: Michel Dänzer <michel@tungstengraphics.com> 150308Date: Thu Apr 17 16:10:10 2008 +0200 150309 150310 Optimize dixLookupPrivate for repeated lookups of the same private. 150311 150312 This gives me a 20% speedup for EXA text rendering, though I still seem to burn 150313 quite a lot of cycles in here... 150314 150315commit 886af8f3849a0fcfc6b63a9695107ce26d7a6955 150316Author: Tilman Sauerbeck <tilman@code-monkey.de> 150317Date: Wed Apr 16 16:20:19 2008 +0200 150318 150319 EXA: Avoid some fallbacks in exaCopyNtoN. 150320 150321 In some cases we can still do the copying in hardware even if the 150322 dimensions of the pixmaps are out of range. This is true when the boxes 150323 that we're to copy are all in the card's range. 150324 150325commit dc10f0a0e243b7ba38d02a4e2c43027563aead7c 150326Author: Julien Cristau <jcristau@debian.org> 150327Date: Thu Apr 17 11:13:47 2008 +0200 150328 150329 Fix composite on !darwin 150330 150331 2ffdb0eb641ab6949783b4eb574f77e7486ac929 changes the default value of 150332 COMPOSITE to 'auto', but doesn't set it back to 'yes' as appropriate. 150333 150334commit 8716d081fdf61ddf956c30aff7697c70507911fd 150335Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150336Date: Thu Apr 3 16:29:43 2008 -0700 150337 150338 XQuartz: Don't enable rootless accelerated functionality... crashy... 150339 (cherry picked from commit cdb4c291d8c10c3a9ea59d8e79275a30d2ea82b4) 150340 150341commit 2ffdb0eb641ab6949783b4eb574f77e7486ac929 150342Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150343Date: Wed Apr 16 21:54:00 2008 -0700 150344 150345 XQuartz: Don't use composite. 150346 (cherry picked from commit 6d3d344b5b95b6dc4166556d03cfd8c9576dc3f0) 150347 150348commit 757a1bf3a3d72e17eeb362f825124c4ba40cc080 150349Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150350Date: Wed Apr 16 21:48:52 2008 -0700 150351 150352 Xquartz: Don't need to link against rlAccel since we don't use it 150353 (cherry picked from commit 180ec128adef11a9a90cea1189dc31ac5de8359f) 150354 150355commit ab8c6a3c5acb2a3bf288f1d6339b09a125bbb930 150356Author: Alan Coopersmith <alan.coopersmith@sun.com> 150357Date: Mon Apr 14 19:12:00 2008 -0700 150358 150359 Update ac_define_dir macro in acinclude.m4 to 2008-04-12 version 150360 150361commit b907258ebe62642af088f6e2970a45a68cf4be19 150362Author: Eric Anholt <eric@anholt.net> 150363Date: Wed Apr 16 12:07:51 2008 -0700 150364 150365 Update dolt from upstream, fixing fallback to libtool. 150366 150367commit e1e189f8538f2b77ae0cf0d846d3899061e4c4b7 150368Author: Aaron Plattner <aplattner@nvidia.com> 150369Date: Tue Apr 15 14:49:51 2008 -0700 150370 150371 Include pciaccess in the xorg-server.pc Requires line. 150372 150373 This pulls in the include path for pciaccess.h, which is needed by, among other 150374 things, xf86.h. 150375 150376commit 9e7ced94a5e3a14762fe934aa69d91f0831cf5ca 150377Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150378Date: Tue Apr 15 12:06:07 2008 -0700 150379 150380 XQuartz: Removed a call to RootlessReorderWindow from the Carbon thread 150381 (cherry picked from commit cb27d5ca8230707b276763c0ec20e586203144c9) 150382 150383commit 51c8fd69ec9292f5e18cdc7f60e1716fbd6ae61a 150384Author: Peter Hutterer <peter@cs.unisa.edu.au> 150385Date: Tue Apr 15 15:09:40 2008 +0930 150386 150387 dix: free the unused device classes when closing a device. 150388 150389 This also requires to NULL-ify all pointers while we're actually using them, 150390 otherwise we'd try to free them twice. 150391 150392commit 48d33ab9b672b3b3ca308000cdbd573d1e368ff9 150393Author: Peter Hutterer <peter@cs.unisa.edu.au> 150394Date: Tue Apr 15 14:29:53 2008 +0930 150395 150396 dix: float attached devices _before_ disabling the master. 150397 150398 It also helps if we're actually providing the correct argument to 150399 AttachDevice... 150400 150401commit 4cf9c5909d926ec322ed1c7df47f95bd872bb607 150402Author: Peter Hutterer <peter@cs.unisa.edu.au> 150403Date: Tue Apr 15 13:56:11 2008 +0930 150404 150405 Xi: fix up button count. 150406 150407 Some leftover code from the previously used alloc/free device classes left us 150408 with a incorrect button count. So a button release didn't come through if 150409 a different pointer was moved after the button press. 150410 150411commit 35982bc109d424c464551ab22ec90af69908c884 150412Author: Kristian Høgsberg <krh@redhat.com> 150413Date: Fri Apr 11 11:09:13 2008 -0400 150414 150415 Make DRI2 a serverlayout/serverflags option. 150416 150417 Add xf86DRI2Enabled() to export the value of the setting. 150418 150419commit f133d85778462134f366389bde7673bff7845fa8 150420Author: Tilman Sauerbeck <tilman@code-monkey.de> 150421Date: Mon Apr 14 11:43:51 2008 +0200 150422 150423 EXA: Update pixmaps' accel_blocked field in ModifyPixmapHeader. 150424 150425commit bb8868540f017b121d698da45e552ffb55a57cea 150426Author: Tilman Sauerbeck <tilman@code-monkey.de> 150427Date: Mon Apr 14 09:58:49 2008 +0200 150428 150429 EXA: Teach exaCompositeFallbackPictDesc() about x8r8g8b8. 150430 150431commit 1a9d7205cd5640eb65f019336097d86301942ea7 150432Merge: 90f491cf8 6866e84e3 150433Author: Peter Hutterer <peter@cs.unisa.edu.au> 150434Date: Mon Apr 14 16:25:58 2008 +0930 150435 150436 Merge whot@wombat:~/potoroo/xserver into mpx 150437 150438commit 3e12c5bb67f3049156475d5cbf4e899aaded76bb 150439Author: Dave Airlie <airlied@redhat.com> 150440Date: Mon Apr 14 11:45:12 2008 +1000 150441 150442 glx: silly nitpick... 150443 150444 even though i and j are the same, we use i to derefence visuals everywhere else 150445 150446commit 97565c0f394f16d042c614695c8b7b4ac354f2a3 150447Author: Dave Airlie <airlied@redhat.com> 150448Date: Mon Apr 14 11:40:38 2008 +1000 150449 150450 glcore: make visualRating GLX_NONE - note GLX_NONE is not == 0 150451 150452 Finally glxinfo returns the set of 3 visuals and glxgears works again for me on 150453 sw rendering 150454 150455commit c61087c82784633e522bd9392172b43656bdf45e 150456Author: Dave Airlie <airlied@redhat.com> 150457Date: Mon Apr 14 10:47:28 2008 +1000 150458 150459 glcore: zero fbconfigs before filling them in. 150460 150461 I'm not sure this the complete proper solution, perhaps it should explicitly fill 150462 in ever field. 150463 150464 This at least makes glxinfo on glcore return sensible information, it doesn't make 150465 gears work yet though. 150466 150467commit 6866e84e3c607d00d88eab2249c2619d6707c1a4 150468Author: Peter Hutterer <peter@cs.unisa.edu.au> 150469Date: Sun Apr 13 19:57:51 2008 +0930 150470 150471 Xi: store feedback classes in devProviates system as well. 150472 150473 This is a follow-up to cb48d880856fd196ab8e8de5eb1f14944a1b4fff. 150474 150475commit cb48d880856fd196ab8e8de5eb1f14944a1b4fff 150476Author: Peter Hutterer <peter@cs.unisa.edu.au> 150477Date: Sun Apr 13 19:48:28 2008 +0930 150478 150479 Xi: store unused classes in devPrivates. 150480 150481 Rather than freeing/allocing classes each time the device capabilities need to 150482 swap, store them in the devPrivates system. 150483 When a class is unused, it is pushed into the devPrivates, and later recovered 150484 when needed again. This saves us a lot of memory allocations/frees, admittedly 150485 on the cost of some memory. 150486 150487commit fde3c836628b6cdec3e5d107d6b1b99bc8b86912 150488Author: Peter Hutterer <peter@cs.unisa.edu.au> 150489Date: Sun Apr 13 17:08:51 2008 +0930 150490 150491 Xi: copy the KeySyms.map over from the source. 150492 150493commit 3c4c9938f31755c5a59995fdcfa138c99db76bbf 150494Author: Peter Hutterer <peter@cs.unisa.edu.au> 150495Date: Sun Apr 13 16:52:14 2008 +0930 150496 150497 Xi: Fix pointer handling in KeyClassRec copy. 150498 150499 We don't free the class anymore, so just store the previous pointers, do the 150500 memcpy from the SD and then restore the pointers. 150501 Plugs a memleak too, before xkbInfo was never freed. 150502 150503commit 755f9e5d7898056cf3bead69ce25a10e23995582 150504Author: Peter Hutterer <peter@cs.unisa.edu.au> 150505Date: Sun Apr 13 16:49:25 2008 +0930 150506 150507 dix: Ignore focus for passive grabs if the event is a pointer event. 150508 150509commit 6faf5b97b92953c331d6540ceb18fd0a77197fea 150510Author: Peter Hutterer <peter@cs.unisa.edu.au> 150511Date: Sun Apr 13 15:42:33 2008 +0930 150512 150513 Xi: fix up modifierKeyMap copying. 150514 150515 Setting it to NULL isn't correct either. The correct behaviour is to realloc 150516 it to the size necessary (or newly alloc it/free it). Otherwise we have a 150517 memleak. 150518 150519commit 3106ba1116e3b9d893f66a93e4a91cc61e23226a 150520Author: Peter Hutterer <peter@cs.unisa.edu.au> 150521Date: Sun Apr 13 11:48:06 2008 +0930 150522 150523 xkb: two fixes to avoid server crashes. 150524 150525 - map can be NULL in some cases, so don't try to dereference it. 150526 - don't default to inputInfo.keyboard 150527 150528 This is firefighting, I presume something in the class copy may have gone 150529 wrong to get a NULL map in the first instance? 150530 150531commit 415c6df0da1197d487456b4c48e2e28e7ded8b8e 150532Author: Peter Hutterer <peter@cs.unisa.edu.au> 150533Date: Sun Apr 13 11:46:44 2008 +0930 150534 150535 Xi: copy feedback classes first, in some cases xkb relies on kbdfeed. 150536 150537 XkbInitIndicatorMap (in XkbInitDevice) calls XkbFindSrvLedInfo. This accesses 150538 the devices kbdfeed struct, which is all nice and dandy if it is NULL. When 150539 copying the device classes however, kbdfeed may not be NULL and thus 150540 XkbFindSrvLedInfo goes on its merry way to do whatever it does. 150541 150542 By copying kbdfeed first, we avoid XkbFSLI to reference the "old" kbdfeed 150543 struct of the previous SD. 150544 150545commit 961f6660902163e99727c2dcc1a039f32b083859 150546Author: Peter Hutterer <peter@cs.unisa.edu.au> 150547Date: Sun Apr 13 09:31:16 2008 +0930 150548 150549 Xi: modifierKeyMap needs to be set to NULL when copying classes. 150550 150551 Otherwise we have a double reference to the same memory area. 150552 150553commit bf6679cba40a936d46008c886d204ed521a4971a 150554Author: Peter Hutterer <peter@cs.unisa.edu.au> 150555Date: Sun Apr 13 08:28:07 2008 +0930 150556 150557 Xi: copy the XkbSrvLedInfo too when copying device classes. 150558 150559commit 4219e94c2f7d431be433eceddfe79760a1ee31a1 150560Author: Peter Hutterer <peter@cs.unisa.edu.au> 150561Date: Sun Apr 13 08:27:31 2008 +0930 150562 150563 xkb: Add XkbCopySrvLedInfo, deep-copies a XkbSrvLedInfoRec. 150564 150565commit 1fa4de80fcfc697b5e5879cc351fb3e9dbf6acbe 150566Author: Matt Turner <mattst88@gmail.com> 150567Date: Sat Apr 12 20:39:18 2008 +0200 150568 150569 Check for __x86_64__ when we check for __amd64__ 150570 150571 It seems Intel C Compiler neglects to define __amd64__, __amd64, or 150572 amd64, but *does* define __x86_64__. 150573 150574commit b1f3f42840ec01db417345a0740b59ad5e4471cb 150575Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150576Date: Fri Apr 11 17:49:51 2008 -0700 150577 150578 Xquartz: Added applicationShouldHandleReopen:hasVisibleWindows to handle dock icon clicking 150579 (cherry picked from commit 55d9973b053f25bb95b26e00351dc5531caf5b04) 150580 150581commit eabcfce0a68d504d11be9479f09e66f574dd2f21 150582Author: Adam Jackson <ajax@redhat.com> 150583Date: Fri Apr 11 09:51:26 2008 -0400 150584 150585 Stop building mfb/afb/xf1bpp by default. 150586 150587commit 0dab6fa3582b70ccd0f01459902415c28dbc81ff 150588Author: Adam Jackson <ajax@redhat.com> 150589Date: Fri Apr 11 09:47:51 2008 -0400 150590 150591 So long, and thanks for all the cfb. 150592 150593commit 059b4876e6350aa1110648788cdfbb3f45b4d66d 150594Author: Adam Jackson <ajax@redhat.com> 150595Date: Fri Apr 11 09:46:48 2008 -0400 150596 150597 Add doltcompile to .gitignore 150598 150599commit 6d22a9615a0e6ab3d00b0bcb22ff001b6ece02ae 150600Author: Peter Hutterer <peter@cs.unisa.edu.au> 150601Date: Fri Feb 22 11:01:51 2008 +1030 150602 150603 dix: Call DeleteInputDeviceRequest from CloseDownDevices (#14418) 150604 150605 The DDX (xfree86 anyway) maintains its own device list in addition to the one 150606 in the DIX. CloseDevice will only remove it from the DIX, not the DDX. If the 150607 server then restarts (last client disconnects), the DDX devices are still 150608 there, will be re-initialised, then the hal devices come in and are added too. 150609 This repeats until we run out of device ids. 150610 150611 This also requires us to strdup() the default pointer/keyboard in 150612 checkCoreInputDevices. 150613 150614 X.Org Bug 14418 <http://bugs.freedesktop.org/show_bug.cgi?id=14418> 150615 150616commit 90f491cf8eb869f27c4278b26c1bb84432b12d63 150617Merge: cbe01b308 b4380d803 150618Author: Peter Hutterer <peter@cs.unisa.edu.au> 150619Date: Fri Apr 11 08:29:52 2008 +0930 150620 150621 Merge whot@wombat:~/potoroo/xserver into mpx 150622 150623commit 3c337e18b933881e22b0d03312511f1d23a8640b 150624Author: Tilman Sauerbeck <tilman@code-monkey.de> 150625Date: Thu Apr 10 21:36:19 2008 +0200 150626 150627 Fixed configure.ac for autoconf 2.62. 150628 150629commit 13dcde6bf994fae09c67c3edce9de42df61ef043 150630Author: Eric Anholt <eric@anholt.net> 150631Date: Thu Apr 10 11:08:49 2008 -0700 150632 150633 Fix dolt to error out on compile error, and not print errors on race to mkdir. 150634 150635 Both of these changes have been submitted upstream. 150636 150637commit a7e3ad1c6b455bda7c4abb352a20845d1d4574a0 150638Author: Eric Anholt <eric@anholt.net> 150639Date: Thu Apr 10 10:33:11 2008 -0700 150640 150641 Remove NDEBUG (assert() disable) define from the X Server. 150642 150643 A few pieces of code were abusing this define for other purposes, which are 150644 converted to #ifndef DEBUG instead. There should be no ABI consequences 150645 to this change. 150646 150647 The rationale is that having the define in xorg-server.h also disables 150648 assert() drivers, which is unexpected, and also difficult to avoid since 150649 xorg-server.h is included in their config.h, and you can't put a #undef in 150650 config.h. As for removing it from the server instead of moving it to an 150651 internal header, we probably shouldn't have unnecessary assert()s in 150652 critical server paths anyway, and if we do we could #define NDEBUG in the 150653 specific cases needed. 150654 150655commit b4380d8030927c940ddaea83c4cf24e0b9eb7b96 150656Author: Peter Hutterer <peter@cs.unisa.edu.au> 150657Date: Thu Apr 10 19:25:43 2008 +0930 150658 150659 dix: don't free MDs classes on init. 150660 150661 The device classes aren't deleted anymore on a class change, so there's no 150662 need to store the MD's original classes. We should however restore the MD to 150663 sane defaults when disconnecting the last device, consider this as TODO item. 150664 150665commit 04dff74ffdf727015e3721aae4ea13acc498cd1c 150666Author: Peter Hutterer <peter@cs.unisa.edu.au> 150667Date: Thu Apr 10 19:22:59 2008 +0930 150668 150669 dix: Rework Enter/Leave semaphore system. 150670 150671 Instead of a simple counter, use bits to keep track of which device is where 150672 etc. When device enters a window (or sets focus), the bit matching the device 150673 is set, when it leaves again, it is unset. If there are 0 bits set, then 150674 Leave/Enter/Focus events may be sent to the client. 150675 150676 Same theory as before, but this should get around the insanity with 150677 Grab/Ungrab special cases. Those cases are basically untested though. 150678 150679commit a88386ee277d136caaaeec305f8753f23f9b6274 150680Author: Peter Hutterer <peter@cs.unisa.edu.au> 150681Date: Thu Apr 10 14:36:57 2008 +0930 150682 150683 Xi: only DeliverFocusedEvents if the event is not a pointer event. 150684 150685 A pointer device may have a focus class, but even if so, pointer events must 150686 be delivered to the sprite window, not the focus window. 150687 150688commit 48249425275cc90242497aee9968e5f1ffc86698 150689Author: Peter Hutterer <peter@cs.unisa.edu.au> 150690Date: Thu Apr 10 14:36:10 2008 +0930 150691 150692 Xi: dont copy FocusClassRec if the master already has one. 150693 150694 Blindly copying will override the focus setting of the master. If there's XI 150695 applications running, they may set the SD focus, while leaving the 150696 MD's focus as it was. In this case, after a class swap we still want to get 150697 the MD's events to the same window as before. 150698 150699commit bce6091c6b04ff2db704ae4f161179d21dcbec59 150700Author: Peter Hutterer <peter@cs.unisa.edu.au> 150701Date: Thu Apr 10 09:59:45 2008 +0930 150702 150703 dix: Extend IsKeyboardDevice() to not include pointer devices. 150704 150705 If a pointer devices has key classes as well, don't register it as a keyboard 150706 device. Let's see how much that change will break. 150707 150708commit cc7dab2d04da4ca164eeec1a3296df1706585466 150709Author: Peter Hutterer <peter@cs.unisa.edu.au> 150710Date: Thu Apr 10 09:58:50 2008 +0930 150711 150712 dix: Dont deliver grabbed pointer events to a focus window. 150713 150714 If an pointer event is being processed during a device grab, don't deliver it 150715 to the focus window, even if the device has a focus class. Reason being that 150716 some pointers may have a focus class, thus killing drag-and-drop. 150717 150718commit df2545b98d888924209cb889a68737c15f1aa209 150719Author: Peter Hutterer <peter@cs.unisa.edu.au> 150720Date: Thu Apr 10 08:50:43 2008 +0930 150721 150722 xfree86: Sanity check before retrieving the paired device. 150723 150724 Some pointer devices send key events [1], blindly getting the paired device 150725 crashes the server. So let's check if the device is a pointer before we try to 150726 get the paired device. 150727 150728 [1] The MS Wireless Optical Desktop 2000's multimedia keys are sent through 150729 the pointer device, not through the keyboard device. 150730 150731commit 5a4c6621aaf4e886f2c3b633e837ba359fedf921 150732Author: Peter Hutterer <peter@cs.unisa.edu.au> 150733Date: Thu Apr 10 08:29:05 2008 +0930 150734 150735 Xi: some extra checks for validity of kbd and mouse. 150736 150737 Floating SDs are paired with themselves, so the paired device may not be a 150738 proper keyboard or mouse. Put some extra checks in to avoid dereferencing a 150739 nullpointer later. 150740 150741commit 8e0a6529303a52acc10905dd47c72a0d60979676 150742Author: Peter Hutterer <peter@cs.unisa.edu.au> 150743Date: Thu Apr 10 08:25:36 2008 +0930 150744 150745 dix: When floating, set sprite to NULL before calling InitializeSprite. 150746 150747 InitializeSprite won't create a new one if it already exists, with the result 150748 of overwriting the master's sprite. This master sprite is then assigned to the 150749 floating slave, and freed when the slave is reattached later. 150750 Setting the sprite to NULL forces InitializeSprite to alloc a new one, and 150751 this one can be freed without further repercussions. 150752 150753commit e7211eb0b3d10323dab681bcb18580405ea18ab2 150754Author: Peter Hutterer <peter@cs.unisa.edu.au> 150755Date: Thu Apr 10 08:08:54 2008 +0930 150756 150757 Xi: When attaching, check for ptr -> ptr and keybd -> keybd. 150758 150759 Some pointer devices have key classes (e.g. MS Optical Desktop 2000). The 150760 previous test was performed as Error if ptr -> keybd or keybd -> ptr. This 150761 doesnt work with such devices. New test is Succeed if ptr->ptr or 150762 keybd->keybd. 150763 150764commit 7909ebe7f163716520f843fae11ac7bdeffcb57c 150765Author: Eric Anholt <eric@anholt.net> 150766Date: Wed Apr 9 10:43:25 2008 -0700 150767 150768 dolt: add FreeBSD support (this and ppc have been submitted upstream). 150769 150770commit b4842d8dc3b1619033c5c123c8adc6e164098dc3 150771Author: Michel Dänzer <michel@tungstengraphics.com> 150772Date: Wed Apr 9 16:17:35 2008 +0200 150773 150774 dolt works on powerpc Linux. 150775 150776commit 0d1746995d91b55e40f233f0c38b56bafe896d38 150777Author: Michel Dänzer <michel@tungstengraphics.com> 150778Date: Wed Apr 9 13:37:59 2008 +0200 150779 150780 Fix off-by-one error in ProcXResQueryClients(). 150781 150782 Fixes memory corruption reported at 150783 http://bugs.freedesktop.org/show_bug.cgi?id=14004 . 150784 150785commit 6d031cbdefd8072b61645955f01b470a3e6858c1 150786Author: Daniel Stone <daniel@fooishbar.org> 150787Date: Wed Apr 9 14:36:26 2008 +0300 150788 150789 configure.ac: Do the dolt 150790 150791 Use dolt instead of libtool whereever practical. See: 150792 http://lists.debian.org/debian-devel/2008/04/msg00286.html 150793 150794commit b19027fbaea4c3a146926e862983e0e3411fff3d 150795Author: Dave Airlie <airlied@linux.ie> 150796Date: Wed Apr 9 14:27:58 2008 +1000 150797 150798 quirk: add quirk for ACR 640x350 default mode is wrong 150799 150800 RH #440186 150801 150802commit 1f8188656a075dc7b1bb27a0795b5bd43610bbc8 150803Author: Ben Byer <bbyer@apple.com> 150804Date: Tue Apr 8 20:37:25 2008 -0700 150805 150806 add missing dix-config include 150807 (cherry picked from commit 126e9bc8c480b403dedc44c1e8c4fe1476340ed9) 150808 150809commit cbe01b3083eb65c9d4259b1071683febebf11600 150810Merge: 5ffbcfec3 389dae73c 150811Author: Peter Hutterer <peter@cs.unisa.edu.au> 150812Date: Wed Apr 9 10:46:46 2008 +0930 150813 150814 Merge whot@wombat:~/potoroo/xserver into mpx 150815 150816commit 3f51f493b6daf2464e6c2ba5a924219b88a9e57e 150817Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 150818Date: Tue Apr 8 17:02:56 2008 -0700 150819 150820 XQuartz: Fix issue where clicking on an X11 window might send that event to an X11 window in another space. 150821 (cherry picked from commit df21312c8b0e9ef0c809bfc57cdf64f27db0d8a7) 150822 (cherry picked from commit 2d4194a8d124e7a9c7cd1b83635ba6957aa4ae1c) 150823 150824commit 389dae73cc0f3693f49807fd2de146c454ba9783 150825Author: Peter Hutterer <peter@cs.unisa.edu.au> 150826Date: Wed Apr 9 08:26:00 2008 +0930 150827 150828 Xi: If device "to" has a class but "from" doesn't, free the class in "to". 150829 150830commit 60c38d248c1a89392c2c6695c3802f4b54e4c00b 150831Author: Peter Hutterer <peter@cs.unisa.edu.au> 150832Date: Wed Apr 9 07:46:53 2008 +0930 150833 150834 Xi: plug memory leak, free previous motion history before allocating new. 150835 150836commit ea05cf0813b2b7c8cd2151cb935820753ae7997a 150837Author: Peter Hutterer <peter@cs.unisa.edu.au> 150838Date: Tue Apr 8 22:09:11 2008 +0930 150839 150840 Xi: check if source has a key class before copying. 150841 150842commit 08073862f8c4e1219b6459708ffd28e2bc35885f 150843Author: Hasso Tepper <hasso@estpak.ee> 150844Date: Tue Apr 8 13:00:38 2008 +0300 150845 150846 configure.ac: Fix monotonic test harder 150847 150848 This was only introduced in a later version of POSIX, so define that 150849 version to get it from more conformant systems. 150850 150851commit 5ffbcfec3d37d3b627a78acfa00dbafc5948df82 150852Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 150853Date: Tue Apr 8 14:42:00 2008 +0930 150854 150855 dix: Ensure Proximity events don't kill the server. 150856 150857 Add Prox events to the if-clauses with the other events 150858 that are usually sent from the input devices. 150859 Ensure that the event deliverers won't try to deliver 150860 events of type '0' (some extended events doesn't have 150861 an equivalent core-type) 150862 150863 Small modification by Peter Hutterer. 150864 150865 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 150866 150867commit 32e4a88ae613c7200d84d5621344b418b656346b 150868Author: Peter Hutterer <whot@potoroo.wearablelab.ml.unisa.edu.au> 150869Date: Tue Apr 8 08:42:58 2008 +0930 150870 150871 xkb: don't overwrite CtrlProc in the second run of XkbFinishDeviceInit. 150872 150873 XkbFinishDeviceInit is called once when the device is initialised, but also 150874 when a class copy causes the key class of a device to change. In this case, 150875 overwriting the CtrlProc of the KeybdFeedbackClass with XkbDDXKeybdCtrlProc 150876 sets up a nice recursive loop of XkbDDXKeybdCtrlProc calling itself until the 150877 cows come home. 150878 150879commit 6271df6953bea462be7e9e01744e5dd46841e867 150880Author: Peter Hutterer <whot@potoroo.wearablelab.ml.unisa.edu.au> 150881Date: Tue Apr 8 08:42:58 2008 +0930 150882 150883 xkb: don't overwrite CtrlProc in the second run of XkbFinishDeviceInit. 150884 150885 XkbFinishDeviceInit is called once when the device is initialised, but also 150886 when a class copy causes the key class of a device to change. In this case, overwriting the CtrlProc of the KeybdFeedbackClass with XkbDDXKeybdCtrlProc sets up a nice recursive loop of XkbDDXKeybdCtrlProc calling itself until the cows come home. 150887 150888commit 726dcd9e4ebfb09c0685450dca6e9fae7e773814 150889Author: Adam Tkac <atkac@redhat.com> 150890Date: Mon Apr 7 10:20:02 2008 -0400 150891 150892 Fix Xvfb input when building against current X sources. 150893 150894commit 0f87b41a432a6472a15ec0c9dee997e3bddbd0f2 150895Author: Hasso Tepper <hasso@estpak.ee> 150896Date: Mon Apr 7 14:09:04 2008 +0300 150897 150898 configure.ac: DragonFly BSD support 150899 150900 Add support for DragonFly BSD, which is just the same as FreeBSD for all 150901 of these cases. 150902 150903commit 6b1a27023e48b661c4bb3b61181ac57608d8e448 150904Author: Owen W. Taylor <otaylor@fishsoup.net> 150905Date: Thu Apr 3 14:50:05 2008 -0400 150906 150907 EXA: Fix TS origin computation when implementing RenderComposite with tiling. 150908 150909commit fd06e8f8c1d82a9d91931e8532bee0fd9c9ca9ab 150910Merge: b46a00918 6c0cfe3d4 150911Author: Peter Hutterer <whot@potoroo.wearablelab.ml.unisa.edu.au> 150912Date: Mon Apr 7 07:56:41 2008 +0930 150913 150914 Merge branch 'master' into dcdc_rework 150915 150916 Conflicts: 150917 150918 Xext/xevie.c 150919 dix/dispatch.c 150920 150921commit 539bf3c2836727e7560c64144071b086f8ea32fe 150922Merge: 3ab33e7cd 6c0cfe3d4 150923Author: Sascha Hlusiak <saschahlusiak@arcor.de> 150924Date: Mon Apr 7 00:10:16 2008 +0200 150925 150926 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 150927 150928commit 3ab33e7cd46c25dfc461b2a1b13e138225a94524 150929Author: Sascha Hlusiak <saschahlusiak@arcor.de> 150930Date: Sun Apr 6 23:23:47 2008 +0200 150931 150932 Remove stale code 150933 150934 The jstk code for Joysticks is not used by any module, was never actually compiled and uses an API 150935 that is deprecated these days. 150936 150937 No reason to keep it. 150938 150939commit b46a00918691cbd5ca80b6d3acae7614f93e073b 150940Author: Peter Hutterer <whot@potoroo.wearablelab.ml.unisa.edu.au> 150941Date: Sun Apr 6 09:02:57 2008 +0930 150942 150943 dix: sprite may be NULL, don't dereference it then. 150944 150945 In some rare cases (e.g. when the init fails) a device's sprite is NULL, 150946 dereferencing it to xfree the spriteTrace is a bad idea then. 150947 150948commit 638a50552e3e2190eac9721deb72e7365bdd52e4 150949Author: Peter Hutterer <whot@potoroo.wearablelab.ml.unisa.edu.au> 150950Date: Sun Apr 6 08:36:21 2008 +0930 150951 150952 dix: remove debug error message about XI->core type conversion. 150953 150954commit 8f38feb3e464986dc523dabd3447ba13263a3a4a 150955Author: Peter Hutterer <whot@potoroo.wearablelab.ml.unisa.edu.au> 150956Date: Sat Apr 5 20:37:09 2008 +1030 150957 150958 Xi: add comments for DeepCopyDeviceClasses, ChangeMasterDeviceClasses. 150959 150960commit 6c0cfe3d43b177c4cfaf7e228f32c655f9a98459 150961Author: Julien Cristau <jcristau@debian.org> 150962Date: Fri Apr 4 19:01:40 2008 +0200 150963 150964 Fix the clock_gettime check for glibc-based non-Linux systems 150965 150966 We need to define _POSIX_C_SOURCE on glibc, not just Linux, so add a new 150967 test for the __GLIBC__ macro. 150968 150969commit cc7c045bae01d90d8f1b750080ba48a96e983c68 150970Author: Adam Jackson <ajax@redhat.com> 150971Date: Fri Apr 4 12:58:12 2008 -0400 150972 150973 Fix PCI config space cycles from int10 emulator. 150974 150975 The top bit of 0xCF8 is an enable bit, not part of the domain. Sending 150976 cycles to domain 128 instead of domain 0 is rarely the right thing to do. 150977 150978commit d1de3dda8efe501d4192c8a99c34ab4265316c32 150979Author: Eric Anholt <eric@anholt.net> 150980Date: Mon Mar 17 14:22:39 2008 -0700 150981 150982 Fix clock_gettime presence detect on FreeBSD. 150983 150984 For non-Linux, _POSIX_C_SOURCE and friends restrict symbols defined rather 150985 than enabling defines of symbols. Additionally, CLOCK_MONOTONIC was 150986 apparently added to the standard around 2000 anyway, not 1993. 150987 150988commit ec17900f52bbd25d07566834756e5c7e832e0463 150989Author: Kristian Høgsberg <krh@redhat.com> 150990Date: Fri Apr 4 10:46:45 2008 -0400 150991 150992 Convert __DRIconfigs after we've made sure createNewScreen succeeded. 150993 150994commit 16a8ce75585ea360c39e0ffce4f7bb26a359b754 150995Author: Kristian Høgsberg <krh@redhat.com> 150996Date: Thu Apr 3 16:44:32 2008 -0400 150997 150998 Only autoload RECORD if it was enabled. 150999 151000commit d0395a753079f291a78d9ab86810b5f84f237491 151001Author: Peter Hutterer <peter@cs.unisa.edu.au> 151002Date: Fri Apr 4 15:05:26 2008 +1030 151003 151004 Xi: realloc to->valuator instead of allocing it blindly. 151005 151006commit 502689847b86be5619da7134646d55a1ac322a2c 151007Author: Peter Hutterer <peter@cs.unisa.edu.au> 151008Date: Fri Apr 4 15:01:53 2008 +1030 151009 151010 Xi: ALLOC_COPY_CLASS_IF should only alloc if to->field doesn't exist. 151011 151012commit 035b1b6995e670ce5593e8aceb08f9ec812c70ea 151013Author: Peter Hutterer <peter@cs.unisa.edu.au> 151014Date: Fri Apr 4 15:00:06 2008 +1030 151015 151016 Xi: remove duplicate copying of kbdfeed and ptrfeed. 151017 151018commit 52fc7aed18930fc9e6bcfdd5dbc28ae559978288 151019Author: Peter Hutterer <peter@cs.unisa.edu.au> 151020Date: Fri Apr 4 14:47:23 2008 +1030 151021 151022 Xi: only alloc memory for feedback classes if it doesn't already exist. 151023 151024commit 4eb87c8693b57d9354832c76417797394656333e 151025Author: Peter Hutterer <peter@cs.unisa.edu.au> 151026Date: Fri Apr 4 14:20:18 2008 +1030 151027 151028 Xi: don't free device classes before copying them. 151029 151030 First commit in a series to come. 151031 151032commit f0915fb3c4a9712200882440a64d11dc595a02bb 151033Author: Dave Airlie <airlied@redhat.com> 151034Date: Fri Apr 4 09:29:51 2008 +1000 151035 151036 quirk: add quirk for ACER EDID 151037 151038commit 2e42b67b82db0f9128dd00e339b9dfdd9fe6d667 151039Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 151040Date: Wed Apr 2 18:05:34 2008 -0700 151041 151042 XQuartz: Change reporting of space change to debug log rather than stderr 151043 (cherry picked from commit ed15556a9fc4ebdb88f42961052fc8456082165f) 151044 151045commit c737d04c758e03e32f692a31ed2a665ccbafa931 151046Author: Ben Byer <bbyer@apple.com> 151047Date: Tue Apr 1 00:40:46 2008 -0700 151048 151049 The AppKit thread should not be calling directly into the X server 151050 functions to change state when the keyboard is reloaded; instead, 151051 pass it as an event. 151052 (cherry picked from commit 7e653f806ff5508aace059312156f319a9ed4479) 151053 151054commit 15b0084f1ab23042190d8beeb3f088b92dee5a10 151055Author: Ben Byer <bbyer@apple.com> 151056Date: Mon Mar 31 23:31:25 2008 -0700 151057 151058 formatting cleanup for X11Application.m (no code changes) 151059 (cherry picked from commit eb083d3f68f459d90417558da1ed00729b749950) 151060 151061commit e9e2d88436597875f102085d216dc0a8fce1450a 151062Author: Ben Byer <bbyer@apple.com> 151063Date: Mon Mar 31 22:55:24 2008 -0700 151064 151065 moved and renamed QuartzMessageServerThread to 151066 DarwinSendDDXEvent to make more clear what it actually does. 151067 (cherry picked from commit bee2b377efc930e25017636e5112093a3a6549c7) 151068 151069commit c1be4e3379d8780dff20390939b657ca0973995a 151070Author: Ben Byer <bbyer@apple.com> 151071Date: Mon Mar 31 21:04:37 2008 -0700 151072 151073 shovelling code around ... 151074 (cherry picked from commit 2143182ba49195bbb2e9163ea6872fd68e7a4a85) 151075 151076commit 985c631b2e1f113039e6e620f030505435fd9815 151077Author: Ben Byer <bbyer@apple.com> 151078Date: Mon Mar 31 20:18:58 2008 -0700 151079 151080 just a bit of juggling headers around -- we're preparing 151081 to call our Xquartz-specific event handlers directly 151082 as mieqHandlers 151083 (cherry picked from commit 4aedba5aa727e22316e8ca456f7218bea9ee0313) 151084 151085commit 89f1d880e83e32b72d35c4dbd6795defa6efa847 151086Author: Ben Byer <bbyer@apple.com> 151087Date: Mon Mar 31 19:47:28 2008 -0700 151088 151089 nuke DarwinEventQueue 151090 (cherry picked from commit 1e0ec02202eeaffae480048b91bf02140ee29f8a) 151091 151092commit 8944b77ec0c18476a25ba3179bcc45b338be22b8 151093Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 151094Date: Wed Apr 2 17:47:42 2008 -0700 151095 151096 continue with gutting darwinEvents.c 151097 (cherry picked from commit c34fce7051b996633291dddc061b696ff737f3fb) 151098 151099commit aa6d12e93e8661da841192ef7c3aa7c6a7731c7f 151100Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 151101Date: Wed Apr 2 17:46:59 2008 -0700 151102 151103 turns out we weren't actually using these files. oops 151104 (cherry picked from commit bfec44d7b4baf0ad0aae55c8209bc60ac93c5b58) 151105 151106commit 6c5962e44730395f81cdb333322c9ad5242c32d4 151107Author: Ben Byer <bbyer@apple.com> 151108Date: Mon Mar 31 18:15:18 2008 -0700 151109 151110 remove vestigal DarwinEQPointerPost etc 151111 (cherry picked from commit a25704c423598d596fd7f2ed4290d4b860bd5d5f) 151112 151113commit c6f0d5d1e51326e5110d27918d834eb0096df7db 151114Author: Ben Byer <bbyer@apple.com> 151115Date: Mon Mar 31 17:48:09 2008 -0700 151116 151117 gut darwinEQEnqueue, and make it just call mieqEnqueue (for the moment) 151118 (cherry picked from commit a9e081a60ca227c0d96d4613075d97d6b762366a) 151119 151120commit 5b6c273eaa53d7b554d69c2b4865988068e73a26 151121Author: Ben Byer <bbyer@apple.com> 151122Date: Mon Mar 31 17:08:45 2008 -0700 151123 151124 add prototype for DarwinEventHandler 151125 (cherry picked from commit 9a03ae33c4f9de830f15eabf3b994882ead7c000) 151126 151127commit 3713be8b470b1ac0fcb4f1e4c6b79c526b2196db 151128Author: Ben Byer <bbyer@apple.com> 151129Date: Mon Mar 31 16:30:16 2008 -0700 151130 151131 add logging of current thread ID to DEBUG_LOG macro 151132 (cherry picked from commit 5848510cc5a8091b30230ab920d904ca6b159480) 151133 151134commit 1400a51ae70d8e498d9ae3975f58ba7c1768ca6f 151135Author: Ben Byer <bbyer@apple.com> 151136Date: Mon Mar 31 16:24:01 2008 -0700 151137 151138 Begin to move all of our Xquartz DDX-specific event handlers 151139 to miEQ, in preparation to remove the DDX-specific code entirely. 151140 (cherry picked from commit 3f4447b95f73a82b3aa0f7b0d1640aba5fb0d1bc) 151141 151142commit 8746daa6732d9837f66d925f2fd74818ecbf8ba2 151143Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 151144Date: Wed Apr 2 15:01:33 2008 -0700 151145 151146 XQuartz: Fixed missing close-paren in preference pane text. 151147 (cherry picked from commit ea37e151dc6032d2a1a33cef809f2a7d507aae35) 151148 151149commit b5a0a865c3045cc08c33388320d4ec3ab7065efb 151150Author: Kristian Høgsberg <krh@redhat.com> 151151Date: Wed Apr 2 19:21:41 2008 -0400 151152 151153 Pick up dri2proto from the standard proto header include path. 151154 151155commit 8cde0af3c57f0375ba8ba77af9fdf74b79d9496d 151156Author: Kristian Høgsberg <krh@redhat.com> 151157Date: Wed Apr 2 19:06:40 2008 -0400 151158 151159 Send the GLX_EXT_texture_from_pixmap attributes to the client. 151160 151161commit 7c20f65fea3dd3170cde89d7113d85f377671bfb 151162Author: Kristian Høgsberg <krh@redhat.com> 151163Date: Wed Apr 2 18:00:06 2008 -0400 151164 151165 Add @XORG_CFLAGS@ to satisfy xf86* includes. 151166 151167 Pointed out by Hasso Tepper. 151168 151169commit b13ab156894074fb38cc812738bc7aeeebd9614d 151170Author: Kristian Høgsberg <krh@redhat.com> 151171Date: Wed Apr 2 12:38:36 2008 -0400 151172 151173 dri2: Unbreak glcore visual setup. 151174 151175commit b31de6a59044f91f8230aa581c9ca8540289c168 151176Author: Xiang, Haihao <haihao.xiang@intel.com> 151177Date: Wed Apr 2 16:29:30 2008 +1000 151178 151179 dri2: fix crasher if DRI2Connect fails 151180 151181commit ebc56aca8bdfec1918cac3c8380895dfddea48ce 151182Author: Hong Liu <hong.liu@intel.com> 151183Date: Wed Apr 2 10:43:19 2008 +0800 151184 151185 Bug #15160: quirk Proview AY765C 151186 151187 prefer first detailed timing 151188 151189commit 9500033b9ecdfaf5a56a4355ffc94d74cb17ca17 151190Author: Goneri Le Bouder <goneri@rulezlan.org> 151191Date: Tue Apr 1 20:19:40 2008 +0200 151192 151193 xfree86: don't crash in AutoConfig if the primary device is not pci 151194 151195 Only call matchDriverFromFiles() if we found a pci device. 151196 Debian bug#472823 (http://bugs.debian.org/472823). 151197 151198commit 37b1258f0a288a79ce6a3eef3559e17a67c4dd96 151199Author: Thomas Jaeger <thjaeger@gmail.com> 151200Date: Tue Apr 1 15:27:06 2008 +0300 151201 151202 XKB: Fix processInputProc wrapping 151203 151204 If input processing is frozen, only wrap realInputProc: don't smash 151205 processInputProc as well. When input processing is thawed, pIP will be 151206 rewrapped correctly. 151207 151208 This supersedes the previous workaround in 50e80c9. 151209 151210 Signed-off-by: Daniel Stone <daniel@fooishbar.org> 151211 151212commit a4d034941100c6ca3b7cc4e59952c2745b9306cc 151213Author: Ben Byer <bbyer@bushing.apple.com> 151214Date: Fri Mar 28 20:47:44 2008 -0700 151215 151216 Add code to track 5 valuators for pointing device, in preparation 151217 for supporting tablet input in Xquartz. 151218 (cherry picked from commit 22c8849ea819eb70a14b2e06330b11b22aa63ebc) 151219 151220commit 6648867d8bd1e86458d2ade77a3ee4567c3d6a97 151221Author: Ben Byer <bbyer@bushing.apple.com> 151222Date: Fri Mar 28 18:27:02 2008 -0700 151223 151224 add debug statements so we can see if/when our Xinput stubs are getting called. 151225 (cherry picked from commit 6e160bbe15dd2c2b8685847c06831cb6aebc6f74) 151226 151227commit 19ff23ab0e72a27d05ed4470f75a0934d6f6c1d1 151228Author: Ben Byer <bbyer@bushing.apple.com> 151229Date: Fri Mar 28 18:25:03 2008 -0700 151230 151231 Remove calls to InitValuatorAxisStruct -- these are now handled in dix by 151232 InitValuatorDeviceClass. 151233 Add InitProximityClassDeviceStruct call to prepare for tablet support. 151234 (cherry picked from commit 1bd980a5b114f5320360943214f8f9f23b29c1e3) 151235 151236commit 9f56fc580646a519875b5a1452738d8c6e1fa860 151237Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 151238Date: Mon Mar 31 17:34:07 2008 -0400 151239 151240 XSELinux: Add a request to get a client's context from a resource ID. 151241 151242commit c40e0b51f0d9ef5e1f30f233d7db1e6db9d6681b 151243Author: Kristian Høgsberg <krh@redhat.com> 151244Date: Wed Mar 26 19:28:09 2008 -0400 151245 151246 Implement DRI2 direct rendering and update AIGLX to DRI interface changes. 151247 151248 Get rid of glcontextmodes.[ch] from build, rename __GlcontextModes to 151249 __GLXcontext. Drop all #includes of glcontextmodes.h and glcore.h. 151250 Drop the DRI context modes extension. 151251 151252 Add protocol code to DRI2 module and load DRI2 extension by default. 151253 151254commit 8074676d2df8d577b443e3fa5e22d7c71c944bd1 151255Author: Fredrik Höglund <fredrik@kde.org> 151256Date: Mon Mar 31 21:24:59 2008 +0200 151257 151258 EXA: Optimize the eviction scanning loop in exaOffscreenAlloc. 151259 151260 Reduce the cost of the inner loop, by keeping a set of pointers to the 151261 first and the last areas in the series, subtracting the cost of the first 151262 area from the score, and adding the cost of the last area while walking 151263 the list. This commit also moves the scanning loop from exaOffscreenAlloc 151264 into a separate function. 151265 151266 Idea by Michel Dänzer. 151267 151268commit 93d876891dbba41b920a9a29a5de77f647f43928 151269Author: Fredrik Höglund <fredrik@kde.org> 151270Date: Mon Mar 31 21:15:50 2008 +0200 151271 151272 EXA: Improve the algorithm used for tracking offscreen pixmap use. 151273 151274 Replace the current score keeping algorithm with a rolling counter that's 151275 incremented in ExaOffscreenMarkUsed, with the previous value being stored 151276 in the area. exaOffscreenAlloc uses the difference between the counter 151277 value and the value in the area when deciding which area to evict. 151278 It now also takes the size of the areas into account, and favors evicting 151279 smaller areas. 151280 151281 The credit for these ideas goes to Michel Dänzer. 151282 151283commit 7034484f0887ea0f8ab956515f2d9301ea5842ce 151284Author: Daniel Stone <daniel@fooishbar.org> 151285Date: Fri Mar 28 10:23:36 2008 +0200 151286 151287 gitignore: Add two more bits 151288 151289 Get slightly better at ignoring vim swap files, and let people keep 151290 local changes if they want to. 151291 151292commit b8ea9f2a25aad88aee77a68f8e20ac07276f0dab 151293Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 151294Date: Fri Mar 28 17:00:08 2008 -0700 151295 151296 XQuartz: Fixed names of enums to be more consistent 151297 (cherry picked from commit c309a08806daf5d716a860c709e51eacad2c745a) 151298 151299commit bd28839eb866a9e6dc3ff80f13a67611da7eaf64 151300Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 151301Date: Fri Mar 28 17:02:02 2008 -0700 151302 151303 =?utf-8?q?XQuartz:=20Another=20Expos=C3=A9=20fix:=20F9=20doesn't=20raise=20all=20windows 151304 151305 =20(cherry=20picked=20from=20commit=20b9cffa20debae73737c674bf75ab65db1bd74556)?= 151306 MIME-Version: 1.0 151307 Content-Type: text/plain; charset=utf-8 151308 Content-Transfer-Encoding: 8bit 151309 151310commit 933ffcdf7d2eaaf7caabfe6f861f04bcb149918f 151311Author: Adam Jackson <ajax@redhat.com> 151312Date: Fri Mar 28 17:53:55 2008 -0400 151313 151314 Compile fix. 151315 151316commit 9e8451d869a4032ddb7de6a62920a3a7b0b1acc1 151317Author: Adam Jackson <ajax@redhat.com> 151318Date: Mon Dec 3 17:38:53 2007 -0500 151319 151320 Add E-EDID segment selection. 151321 151322commit 0b4aef4d6df7a5525d381de035fbbf78c5fffeef 151323Author: Adam Jackson <ajax@redhat.com> 151324Date: Mon Dec 3 16:00:00 2007 -0500 151325 151326 Refactor DDC2 code to allow for proper segmented addressing. 151327 151328commit 88ece11d6c45c6f4b94f7fb2da64a46e879d7c27 151329Author: Adam Jackson <ajax@redhat.com> 151330Date: Mon Dec 3 15:47:39 2007 -0500 151331 151332 Start E-EDID support in the DDC module. 151333 151334 Since there's no way to safely know how many blocks xf86DoEDID_DDC2 would 151335 return, add a new xf86DoEEDID entrypoint to do that, and implement the 151336 one in terms of the other. 151337 151338commit b5f98fcea2024c67e598947782913982072cf4fb 151339Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 151340Date: Fri Mar 28 14:01:34 2008 -0400 151341 151342 XSELinux: Add xorg.conf option for permissive/enforcing/disabled. 151343 Patch by Joe Nall. 151344 151345 The option goes in the "extmod" subsection. 151346 TODO: Make it easier for extension modules to handle their own options. 151347 151348commit 415e49b940bba2d08870db410ebb47d2add5d836 151349Author: Jeremy Huddleston <jeremyhu@freedesktop.org> 151350Date: Tue Mar 25 23:01:02 2008 -0700 151351 151352 XQuartz: Sanity-check the removal of the 256-color option 151353 Fixes the 'one-time-preferences' bug in 2.2.0_rc1 151354 (cherry picked from commit 38cbd13490fc21724e8eef0ec7e1a20a9cc2e39d) 151355 151356commit c1d37efe9aae5f2895b3437418f7e4bb2eb3400f 151357Author: Ben Byer <bbyer@bushing.apple.com> 151358Date: Tue Mar 25 19:25:00 2008 -0700 151359 151360 xquartz: copy in new stubs from Xi/stubs.c to replace our old 151361 Xinput stubs. Hey, it's a start. 151362 (cherry picked from commit 566412b4aece24ae6af8c7c835986b685aa456a2) 151363 151364commit c4a616a741e15865ce0ff98781c6f1dca4d62887 151365Author: Ben Byer <bbyer@bushing.apple.com> 151366Date: Mon Mar 24 22:43:10 2008 -0700 151367 151368 prevent "fake mouse clicks" from generating spurious extra events 151369 (cherry picked from commit bd85a24969427e41389688663ead2d4dd41c9999) 151370 151371commit c747030a49dd289e873e2b686cd129d840e55468 151372Author: Dave Airlie <airlied@redhat.com> 151373Date: Thu Mar 27 15:18:39 2008 +1000 151374 151375 quirk: fix LPL monitors properly. 151376 151377 no point having a h cm fix when we really want to copy the sizes from the 151378 other place. 151379 151380 RH BZ 435216 151381 151382commit 9df915b84d45d39aed7557c98883b20a66da2e96 151383Merge: 47eb658e8 333e7123d 151384Author: Sascha Hlusiak <saschahlusiak@arcor.de> 151385Date: Tue Mar 25 17:50:50 2008 +0100 151386 151387 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 151388 151389commit 333e7123dc484888d79c0f5aa3977bd72f1eb341 151390Author: Adam Jackson <ajax@redhat.com> 151391Date: Tue Mar 25 12:48:22 2008 -0400 151392 151393 Fix that last commit. 151394 151395 I can apply patches, really. 151396 151397commit 47eb658e802775021e3efec109f95431cca188ca 151398Author: Sascha Hlusiak <saschahlusiak@arcor.de> 151399Date: Tue Mar 25 17:37:25 2008 +0100 151400 151401 Support to pass arbitrary options via HAL hotplugging 151402 151403 Parse "input.x11_options" and pass every key/name pair to the driver. 151404 Remove check for input.capabilities, because that's part of the fdi files. 151405 151406 Thanks to Dustin Spicuzza <dustin@virtualroadside.com> for the patch. 151407 151408commit e7a364425547103a98acabfc67d16e1ae0c2967f 151409Author: Sascha Hlusiak <saschahlusiak@arcor.de> 151410Date: Tue Mar 25 12:32:33 2008 -0400 151411 151412 Fix getValuatorEvents to compute number of valuators correctly. 151413 151414commit f028e245a7932362656701c08fcfbfa8e8949077 151415Author: David Nusinow <dnusinow@debian.org> 151416Date: Thu Feb 28 19:45:21 2008 -0500 151417 151418 Bug #10016: Implement WM_CLASS hints in Xephyr. 151419 151420commit 862ff9ac92037e13629329eb6ba50ff6bd2c5f71 151421Author: Adam Jackson <ajax@redhat.com> 151422Date: Mon Mar 24 13:37:42 2008 -0400 151423 151424 Bug #11510: Fix build without RECORD. 151425 151426commit 87bfd3bd96c714a1c252d42408b5a1a4ff9dab06 151427Author: Adam Jackson <ajax@redhat.com> 151428Date: Mon Mar 24 13:33:38 2008 -0400 151429 151430 Bug #11508: Fix build without XV. 151431 151432commit 536f2ff5382aaaace3b55481e15366bb15d87801 151433Author: Adam Jackson <ajax@redhat.com> 151434Date: Mon Mar 24 12:22:19 2008 -0400 151435 151436 Bug #13962: Re-arm the DPMS timer when re-enabling DPMS. 151437 151438commit 4217ba0cf0c9bbea3774760e836ab372acf3237c 151439Author: Julien Cristau <jcristau@debian.org> 151440Date: Sat Mar 22 17:31:08 2008 +0100 151441 151442 xaa: use xf86ReturnOptValBool instead of xf86IsOptionSet 151443 151444 The latter doesn't return the option's value, just whether it's present 151445 in the configuration. 151446 151447commit 6b9d2bb1f7f87acbf275027af9c2982e91e5faed 151448Author: Julien Cristau <jcristau@debian.org> 151449Date: Sat Mar 22 17:28:48 2008 +0100 151450 151451 exa: use xf86ReturnOptValBool instead of xf86IsOptionSet 151452 151453 The latter doesn't give you the option's value, it just tells you if 151454 it's present in the configuration. So using Option "EXANoComposite" "false" 151455 disabled composite acceleration. 151456 151457commit 63859473965f911515bc6e8d87b32a65ec41eb73 151458Author: Jeremy Huddleston <jeremy@tifa.local> 151459Date: Fri Mar 21 19:31:31 2008 -0700 151460 151461 XQuartz: Disable 256 color option and fullscreen option 151462 (cherry picked from commit 7c1964338a33558d3f25e369dfca99e3ef9d10f9) 151463 151464commit 93daa3a3bf1a981757024847882ce92b6bdaae83 151465Author: Jeremy Huddleston <jeremy@tifa.local> 151466Date: Fri Mar 21 19:11:59 2008 -0700 151467 151468 Xquartz: Added separate preference tab for quartz-wm settings 151469 Added FFM and "Focus on new window" options 151470 (cherry picked from commit 6841d078b7cb0b0db3db948d26b4d5ec7747deb8) 151471 151472commit c49e11268322712c211f29d51d664d3f8a59b00b 151473Author: Jeremy Huddleston <jeremy@tifa.local> 151474Date: Fri Mar 21 18:07:38 2008 -0700 151475 151476 XQuartz: Initial framework for dealing with spaces on OS-X 151477 (cherry picked from commit 9831324998f9d1f05ff944c58c5bf60dcae17355) 151478 151479commit 4c76607b699431183ee7e88fa7818cb7644a5a02 151480Author: Jeremy Huddleston <jeremy@tifa.local> 151481Date: Mon Mar 17 23:57:41 2008 -0700 151482 151483 Rootless: Removed safeAlphaXXXX() in favor of using fb/pixman 151484 (cherry picked from commit f03202ad15457c98be7ca78cc59bac88cf5f1966) 151485 151486commit 87c64cfd6901083da5a9375d0bde65691d374b5b 151487Author: Jeremy Huddleston <jeremy@tifa.local> 151488Date: Fri Mar 14 17:31:54 2008 -0700 151489 151490 =?utf-8?q?Apple:=20Xserver=20half=20of=20the=20Expos=C3=A9=20bug-fix=20(requires=20updated=20libXplugin=20from=20Apple...=20coming=20with=202.2) 151491 151492 =20(cherry=20picked=20from=20commit=2037be23e8c1d8e5c7a1157e9d66ef3f30a4c472c5)?= 151493 MIME-Version: 1.0 151494 Content-Type: text/plain; charset=utf-8 151495 Content-Transfer-Encoding: 8bit 151496 151497commit 3bbd77ff98478153afe3251de9ba11d757218213 151498Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 151499Date: Thu Mar 20 20:03:02 2008 -0400 151500 151501 XSELinux: Do a check for whether background "None" is allowed. 151502 151503commit e323bb426ce8a072d119cb2720b773241259c137 151504Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 151505Date: Thu Mar 20 19:42:09 2008 -0400 151506 151507 XSELinux: Correctly handle some permission bits that are used more than once. 151508 151509commit da973e962d09854b571320dee7dd9569060bc39e 151510Author: Michel Dänzer <michel@tungstengraphics.com> 151511Date: Thu Mar 20 09:18:29 2008 -0400 151512 151513 Fix RandR 1.2 driver interface conversion of two colour cursors to ARGB 151514 151515 This patch (and not setting HARDWARE_CURSOR_BIT_ORDER_MSBFIRST on big endian 151516 platforms) fixes it for me with the radeon driver and doesn't break intel. 151517 151518 Correct patch this time :) 151519 151520commit f8c1eb29e28818895d744c4e1d6897353d51790b 151521Author: Alex Deucher <alex@cube.(none)> 151522Date: Thu Mar 20 09:14:41 2008 -0400 151523 151524 Revert "Fix RandR 1.2 driver interface conversion of two colour cursors to ARGB" 151525 151526 This reverts commit 267352579612155adfd4743432d6569b2cdeebde. 151527 151528 Pushed the wrong patch. 151529 151530commit 267352579612155adfd4743432d6569b2cdeebde 151531Author: Michel Dänzer <michel@tungstengraphics.com> 151532Date: Wed Mar 19 19:12:37 2008 -0400 151533 151534 Fix RandR 1.2 driver interface conversion of two colour cursors to ARGB 151535 151536 See bug 11796 151537 151538commit 060a99444ee25a684b0ab9b4819bf8e855aea2d8 151539Author: Alan Coopersmith <alan.coopersmith@sun.com> 151540Date: Wed Mar 19 16:04:16 2008 -0700 151541 151542 Make Xevie private symbol names less generic 151543 151544 Makes it easier to figure out what you're seeing in the stack trace 151545 instead of wondering where in the server "ProcSend" is. 151546 151547commit f37046984d7839faefa4d716624e4a85ddde9634 151548Author: Alan Coopersmith <alan.coopersmith@sun.com> 151549Date: Wed Mar 19 14:06:53 2008 -0700 151550 151551 Xevie cleanups, byteswapping fixes & request length check fixes 151552 151553commit edad0a9dfebcce5c54b2f9c32bd9d45549e20c51 151554Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 151555Date: Tue Mar 18 17:51:21 2008 -0400 151556 151557 Apply __glXDisp_GetVisualConfigs message patch 151558 From http://bugs.freedesktop.org/show_bug.cgi?id=13863 151559 151560 Problem was that the glxcmds.c __glXDisp_GetVisualConfigs 151561 function left garbage in the tail end of the message used for extensions. 151562 151563commit cdadd2ff9bade318caac5c1d9bcdc8a001347da9 151564Author: Dodji Seketeli <dodji@seketeli.org> 151565Date: Tue Mar 18 14:00:15 2008 +0100 151566 151567 [Xephyr/DRI] correctly route motion events targeted at GL drawable 151568 151569commit afd7428690d87097117ab20335658f6d59d60103 151570Author: Jesse Barnes <jbarnes@jbarnes-t61.(none)> 151571Date: Mon Mar 17 14:55:44 2008 -0700 151572 151573 Cleanup logic in xf86PrepareOutputs 151574 151575 Should have done this in the first place. Since we're checking for the absence 151576 of the get_crtc callback in the first place, we'll short circuit the later call 151577 and disable the output, so the ugly "continue" block is unnecesary. 151578 151579commit ba85caacb565b9aa0aeace52a362350304b0566d 151580Author: Jesse Barnes <jbarnes@jbarnes-t61.(none)> 151581Date: Mon Mar 17 14:13:09 2008 -0700 151582 151583 Make xf86SetDesiredModes aware of current output configuration 151584 151585 By adding a new output callback, ->get_crtc, xf86SetDesiredModes is able to 151586 avoid turning off outputs & CRTCs if the current output<->CRTC mappings are the 151587 same as the desired configuration. This helps avoid flickering displays at 151588 startup time, which speeds things up a little and looks better. 151589 151590commit bee2ddf35f75086cee951142098637f2c756b96b 151591Author: Jesse Barnes <jbarnes@jbarnes-t61.(none)> 151592Date: Mon Mar 17 08:33:01 2008 -0700 151593 151594 Fail CRTC configuration if !vtSema 151595 151596 Unless we check for vtSema before calling into the CRTC and output callbacks, 151597 we may end up trying to access video memory that no longer exists, leading to a 151598 crash. So if we don't have vtSema, return FALSE to the caller, indicating that 151599 we didn't do anything. 151600 151601 Fixes #14444. 151602 151603commit db248ffb840a0c113b6eb508a0fa1e74e752474d 151604Author: Matthieu Herrb <matthieu.herrb@laas.fr> 151605Date: Sun Mar 16 18:46:11 2008 +0100 151606 151607 test for the presence of pci_system_init_dev_mem() before calling it. 151608 151609 This avoids creating a dependency on -current libpciaccess for 151610 BSD systems other than OpenBSD (which don't otherwise need it). 151611 151612commit aa231f28d56402d7daea6cbd3002fbf760f79497 151613Author: Donnie Berkholz <dberkholz@gentoo.org> 151614Date: Fri Mar 14 18:41:25 2008 -0700 151615 151616 Xephyr: Build fix: Port across XF86dri.c changes from Mesa. 151617 151618commit a955c3b587b22b8bf20cb6bedbbec4ad5fcb32ac 151619Author: Donnie Berkholz <dberkholz@gentoo.org> 151620Date: Fri Mar 14 18:41:07 2008 -0700 151621 151622 Xephyr: Distribute ephyrdriext.h in tarballs. 151623 151624commit 090b26db767d296e7a3452da83b136d1caa0ed01 151625Author: Daniel Stone <daniel@fooishbar.org> 151626Date: Fri Mar 14 21:58:27 2008 +0200 151627 151628 XkbCopyKeymap: Fix broken indentation 151629 151630 An astute observer will note that the entirety of XkbCopyKeymap is indented 151631 with spaces, and no tabs whatsoever, and not commit changes which break the 151632 otherwise consistent indentation. 151633 A non-astute observer will note the breakage when the commit mail comes 151634 through with clearly broken indentation. 151635 A polite, non-astute, observer will then fix it. 151636 151637 C'est la vie. 151638 151639commit 88bec0915e3867f8dbf859a3dfbb771d0d07092d 151640Author: Daniel Stone <daniel@fooishbar.org> 151641Date: Fri Mar 14 21:54:13 2008 +0200 151642 151643 mi: More meaningful assert crashes 151644 151645 When we fail an assert in miregion.c (which happens every now and then, 151646 though I haven't yet checked up why), at least generate a segfault, so 151647 we'll get a backtrace. 151648 151649commit 57d48d94b8947c571925e6fd4c9bf041fbd1b2ac 151650Author: Adam Jackson <ajax@redhat.com> 151651Date: Fri Mar 14 14:37:42 2008 -0400 151652 151653 Fix a stray use of ALLOCATE_LOCAL. 151654 151655commit 824853772241acf64bc37ac8b85254194741ae13 151656Author: Adam Jackson <ajax@redhat.com> 151657Date: Fri Mar 14 14:24:21 2008 -0400 151658 151659 RANDR 1.2: Fix initial mode aspect ratio match in a corner case. 151660 151661 Actually more like in the mainline case, where the ideal mode happens to 151662 be the very first aspect match on the first monitor. But let's not 151663 split hairs. 151664 151665commit 1b9878ffcfc0c0dbc4a6e674827fe508ba77db4b 151666Author: Bart Trojanowski <bart@jukie.net> 151667Date: Thu Mar 13 17:42:16 2008 -0400 151668 151669 Bug #14332: Fix PCI access cycles from x86emu. 151670 151671 The address written to 0xcf8 contains the PCI slot address to send the 151672 config cycle to. However, we would ignore that and always send the 151673 cycle to the device whose BIOS we were running. This breaks some 151674 integrated graphics platforms that have explicit knowledge about the 151675 system's host bridge, for example. 151676 151677commit f7abe05b3306ed9a6f2cf5e3e45ed524d725d029 151678Author: Doug Chapman <doug.chapman@hp.com> 151679Date: Thu Mar 13 17:40:34 2008 -0400 151680 151681 Bug #14091: Fix build (and runtime) on ia64. 151682 151683commit 5d7437c29e686a081b20823450d78c4c2f4e0aec 151684Author: Adam Jackson <ajax@redhat.com> 151685Date: Thu Mar 13 17:37:12 2008 -0400 151686 151687 RANDR 1.2: Fix the RANDR 1.1 screen size estimation to approach reality. 151688 151689 While the ScreenRec's notion of size in millimeters would get updates, 151690 the RANDR 1.1 notion wouldn't, so your screen would appear to be square 151691 and probably at some ludicrous DPI. 151692 151693commit 61c3f63a75d8b0cc47ffed4a0e30147fab2ae8f4 151694Author: Adam Jackson <ajax@redhat.com> 151695Date: Thu Mar 13 17:34:54 2008 -0400 151696 151697 RANDR 1.2: Don't report a square resolution to RANDR 1.1 clients. 151698 151699 It can't possibly do anything useful, and older versions of Gnome (and 151700 proably others) get very confused by it. So do the drivers, for that 151701 matter. 151702 151703commit 06c0372c3a1b45005eb6d50406f77f4e93f1de1e 151704Author: Mark Kettenis <mark.kettenis@xs4all.nl> 151705Date: Wed Mar 12 21:45:37 2008 +0100 151706 151707 OpenBSD support for libpciaccess. 151708 151709 xserver and libpciaccess both need to open /dev/xf86, which can only 151710 be opened once. I implemented pci_system_init_dev_mem() like Ian 151711 suggested. This requires some minor changes to the BSD-specific 151712 os-support code. Since pci_system_init_dev_mem() is a no-op on 151713 FreeBSD this should be no problem. 151714 151715commit 2036851125226065891f13583ade3ce559e7bd37 151716Author: Matthias Hopf <mhopf@suse.de> 151717Date: Mon Mar 10 19:29:07 2008 +0100 151718 151719 Return randr interface version in xf86CrtcScreenInit() 151720 151721 Necessary to allow drivers to be run-time backwards compatible when using the 151722 modes/ functions w/o providing their own copy. 151723 151724commit c7536f4b87e089a7e7c43026b189922fec70c565 151725Author: Kristian Høgsberg <krh@redhat.com> 151726Date: Tue Mar 11 13:11:04 2008 -0400 151727 151728 Silence REGION_INIT() warning. 151729 151730 Evaluating the address of a BoxRec as a boolean gives this warning: 151731 151732 i830_driver.c:2317: warning: the address of 'ScreenBox' will always 151733 evaluate as 'true' 151734 151735 which is pretty annoying. This patch compares the address to NULL to 151736 avoid the pointer->bool conversion and gets rid of the warning. Seems 151737 like a lame hack, but the warning is worse. 151738 151739commit cc05255191413b3f376edbc600122ff085f45f7b 151740Author: Kristian Høgsberg <krh@redhat.com> 151741Date: Tue Mar 11 00:51:43 2008 -0400 151742 151743 Make WriteToClient take a const void * like any decent IO write function. 151744 151745 Enough with the casting. Doesn't break API or even ABI, but does make 151746 a lot of silly casts superfluos. 151747 151748commit bc504ffbba3dec2e3467bab8ba1ac25db6dd317e 151749Author: Kristian Høgsberg <krh@redhat.com> 151750Date: Tue Mar 11 00:35:31 2008 -0400 151751 151752 DRI2: Add DRI2AuthConnection(). 151753 151754 DRI2 uses the same authentication scheme as XF86DRI, so implement this 151755 entry point so DRI2 protocol code can access it. 151756 151757commit b2657ec5981122e7cc0bda0d8802aec63cde9014 151758Author: Keith Packard <keithp@keithp.com> 151759Date: Wed Feb 6 16:51:57 2008 -0800 151760 151761 XkbCopyKeymap was mangling doodads and overlays 151762 151763commit ab9b0b36ac8ac72fc48c0abd91a83de49a18313c 151764Author: Adam Jackson <ajax@redhat.com> 151765Date: Mon Mar 10 13:40:00 2008 -0400 151766 151767 Add the "amd" driver to magic driver selection. 151768 151769commit 0f6aaf636b7ac4c98467284ff7baf1b83e0b72e7 151770Author: Adam Jackson <ajax@redhat.com> 151771Date: Mon Mar 10 09:14:20 2008 -0400 151772 151773 Bug #14927: Fix the math for xf86NumDefaultModes. 151774 151775commit 01c2e01f2aee580438b74bfb9da8f584f3878e6b 151776Author: Kristian Høgsberg <krh@sasori.boston.redhat.com> 151777Date: Sun Mar 9 21:40:27 2008 -0400 151778 151779 GLX: Track changes to DRI_TEX_BUFFER extension. 151780 151781 We now just pass in the __DRIdrawable. 151782 151783commit acedc03367e9e69f03b4838f0f0e8d8a8e872b9b 151784Author: Kristian Høgsberg <krh@sasori.boston.redhat.com> 151785Date: Sun Mar 9 21:39:19 2008 -0400 151786 151787 DRI2: Return event buffer head index in DRI2CreateDrawable. 151788 151789 And pass it to the DRI driver in AIGLX. 151790 151791commit 911f0c147699d3b8d97491be2ef6d2e4a6682a0b 151792Author: Julien Cristau <jcristau@debian.org> 151793Date: Fri Mar 7 23:24:06 2008 +0100 151794 151795 Programs in hw/dmx/examples/ want client-side xkb headers. 151796 151797commit 9abaad115cb6245b12b2adb3552ace99b634ab4a 151798Author: Kristian Høgsberg <krh@redhat.com> 151799Date: Fri Mar 7 14:12:28 2008 -0500 151800 151801 Fix DRI2 texture target for GLX_EXT_texture_from_pixmap. 151802 151803 Thanks to Dennis Kasprzyk for pointing it out and for reminding me to 151804 commit it. 151805 151806commit 3bf7ff70323d533a3a05c0f3e22393e63beada99 151807Author: Adam Jackson <ajax@redhat.com> 151808Date: Fri Mar 7 09:25:06 2008 -0500 151809 151810 Size xf86DefaultModes explicitly. 151811 151812 i.e., don't check for the end of the list by ->name == NULL, since that 151813 won't work now. Fix the consumers of xf86DefaultModes to use the new 151814 explicit size as well. 151815 151816commit 3fcb6445dc08f42488287162e3b7009d405e9c5b 151817Author: Adam Jackson <ajax@redhat.com> 151818Date: Fri Mar 7 08:29:49 2008 -0500 151819 151820 Fix segfault when a monitor exists but has no modes. 151821 151822 Thanks to Zhenyu Wang for finding this. 151823 151824commit a8d760f567b19268329c4682495caa591f08a854 151825Author: Aaron Plattner <aplattner@nvidia.com> 151826Date: Sun Mar 2 20:13:11 2008 -0800 151827 151828 Get modes code building with old X servers again. 151829 151830 This change uses XORG_VERSION_CURRENT < 7.0 to mean "server newer than 1.2" 151831 since XORG_VERSION current went backwards at some point. 151832 151833commit ca616b902b2c5d0f046c7a042c11f045479e373a 151834Author: Jesse Barnes <jbarnes@jbarnes-t61.(none)> 151835Date: Thu Mar 6 13:47:44 2008 -0800 151836 151837 Allow RandR get output property to call into drivers 151838 151839 In order to report accurate values to users of the RandR property interface, 151840 it's sometimes necessary to ask the driver to update the value (for example 151841 when backlight brightness changes without the server's knowledge, due to hotkey 151842 events or direct sysfs banging). 151843 151844 This patch wires up the core server code with a new xf86CrtcFuncs callback, 151845 get_property, to allow for this. 151846 151847 The new code is available under the RANDR_13_INTERFACE define, which in turn 151848 depends on the RANDR_12_INTERFACE code. 151849 151850commit 34b69e3bc0e6462eb60029fbcb4f5479494007a2 151851Author: Adam Jackson <ajax@redhat.com> 151852Date: Wed Mar 5 23:56:49 2008 -0500 151853 151854 Fix distcheck. 151855 (cherry picked from commit 2a47accff8dccded4dfe031f9366c028ba927824) 151856 151857commit 8e562fe26ff77c03be0855afb08e43f3ed0bd4b9 151858Author: Peter Hutterer <peter@cs.unisa.edu.au> 151859Date: Wed Mar 5 15:11:41 2008 +1030 151860 151861 Xi: remove false memset from DeepCopyDeviceClasses. 151862 151863 It's wrong, needs to be different, but I haven't had enough coffee to figure 151864 out what it needs to do yet. 151865 151866commit d4101140f4e569f18554cf0cbf43138d08bd1277 151867Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 151868Date: Tue Mar 4 22:39:41 2008 -0500 151869 151870 xselinux: Implement polyinstantiation support and related protocol. 151871 151872commit cf984dcc156958d4f8d98110e7add150628ce97e 151873Author: Alan Coopersmith <alan.coopersmith@sun.com> 151874Date: Tue Mar 4 16:03:04 2008 -0800 151875 151876 Make sure SIOGLIFCONF buffer is properly aligned for socket structures 151877 151878commit 27e7dacbf7ef17712be31ff90f98ee3a5c5cf909 151879Author: Adam Jackson <ajax@redhat.com> 151880Date: Tue Mar 4 11:38:34 2008 -0500 151881 151882 Make xf86InitialConfiguration slightly smarter. 151883 151884 Old heuristic was to find the first monitor that expressed a preference, 151885 then attempt to get all other monitors to agree. This doesn't work 151886 particularly well when the two sets of modes don't precisely intersect, 151887 you get overlapping-but-not-identical output geometry and things go wrong. 151888 151889 New heuristic is: 151890 - Exact user preference, if given 151891 - Exact output preference, if the same for all outputs 151892 - Best (largest) mode of modes common to all outputs: 151893 - with the same aspect ratio as all outputs (may be NULL) 151894 - with 4:3 aspect ratio 151895 - Then the old heuristic to try to get something lit 151896 151897 Note that it is simply not doable to have a reliable initial output guess if 151898 you insist on trying to clone all outputs together. It's far too easy to 151899 end up with displays that simply don't have modes in common. We need to 151900 switch to right-of placement someday, once we're not limited to CRTC size 151901 limits and we have working multi-GPU in RANDR. 151902 151903commit 95df04b744c6a3498a9a9e2ea9bb03ee780e60f8 151904Author: Adam Jackson <ajax@redhat.com> 151905Date: Tue Mar 4 10:59:24 2008 -0500 151906 151907 Remove all mention of the vga driver from the config logic. 151908 151909commit 7b4f3616f75a541b819d99c28bb6cb73761b6066 151910Author: Peter Hutterer <peter@cs.unisa.edu.au> 151911Date: Tue Mar 4 18:29:15 2008 +1030 151912 151913 Xi: add deviceid to QueryDevicePointer request. 151914 151915 Why not, we had the padding anyway. 151916 Suggested by Jim Gettys on the X.Org mailing list. 151917 151918commit 056a2ce02ce85013e89055ee44a7aa3eabedac09 151919Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 151920Date: Tue Mar 4 02:44:48 2008 -0500 151921 151922 XACE: Check the return value of the selection create hook call. 151923 151924commit 4f2cd0ed96d3b10c78774c721c2ffbfb0556dddd 151925Merge: 23ae68a4c 453661a9e 151926Author: Peter Hutterer <peter@cs.unisa.edu.au> 151927Date: Tue Mar 4 18:11:10 2008 +1030 151928 151929 Merge branch 'master' into mpx 151930 151931 This merge reverts Magnus' device coorindate scaling changes. MPX core event 151932 generation is very different, so we can't scale in GetPointerEvents. 151933 151934 Conflicts: 151935 151936 Xi/opendev.c 151937 dix/devices.c 151938 dix/dixfonts.c 151939 dix/getevents.c 151940 dix/resource.c 151941 dix/window.c 151942 hw/xfree86/common/xf86Xinput.c 151943 mi/mipointer.c 151944 xkb/ddxBeep.c 151945 xkb/ddxCtrls.c 151946 xkb/ddxKeyClick.c 151947 xkb/ddxList.c 151948 xkb/ddxLoad.c 151949 xkb/xkb.c 151950 xkb/xkbAccessX.c 151951 xkb/xkbEvents.c 151952 xkb/xkbInit.c 151953 xkb/xkbPrKeyEv.c 151954 xkb/xkbUtils.c 151955 151956commit 72f2197545e734cd0aa785d05a57b2fc0351a763 151957Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 151958Date: Tue Mar 4 02:02:54 2008 -0500 151959 151960 dix: Convert selection list to a linked list. 151961 151962 Fixes a bug where pointers were being invalidated after a realloc. 151963 151964commit 0bd0f90d7c7928052197da7119177e5a1c9eee2c 151965Author: Daniel Stone <daniel@fooishbar.org> 151966Date: Tue Mar 4 03:47:36 2008 +0200 151967 151968 XKB: Fix initial map setting on startup 151969 151970 Due to an unwitting sense inversion when eliminating XkbFileInfo, we were 151971 setting the complete wrong keymap on startup (non-XKB map if we had an XKB 151972 map available, or the XKB map if we didn't have any available). Invert the 151973 sense properly, and add two small bits that also went missing in that commit. 151974 151975commit 613852ce6a821ce6f6382fc14629f517776a3701 151976Author: James Cloos <cloos@jhcloos.com> 151977Date: Mon Mar 3 16:10:04 2008 -0500 151978 151979 Fix some documentation typos 151980 151981commit 3f23139137e024e09d207be05a61968100cf53e8 151982Author: Jesse Barnes <jesse.barnes@intel.com> 151983Date: Mon Mar 3 13:05:12 2008 -0800 151984 151985 Add cscope files to .gitignore 151986 151987commit 708f07753ff22ade54e9ee8885e4198fff363b87 151988Author: Adam Jackson <ajax@redhat.com> 151989Date: Mon Mar 3 15:49:48 2008 -0500 151990 151991 RANDR 1.2: Inherit PreferredMode from the global configuration, if any. 151992 151993 If you don't do this, then Modes "800x600" in the Display subsection will 151994 be dutifully ignored and the driver will start at whatever resolution it 151995 feels like. 151996 151997commit 605e6764dfd3e9cb917b9cfcd92fe89857c1a1c9 151998Author: Adam Jackson <ajax@redhat.com> 151999Date: Mon Mar 3 15:45:17 2008 -0500 152000 152001 Fix Motif menu drawing in Xnest. 152002 152003 See also Red Hat bug #229350, OpenSolaris bug #6366490. 152004 152005commit 3b73d62791d925c465ec855f96981d151dd3c179 152006Author: Adam Jackson <ajax@redhat.com> 152007Date: Mon Mar 3 15:43:22 2008 -0500 152008 152009 xf86DDCMonitorSet: Honor the DisplaySize from the config file. 152010 152011 We honor sync ranges and pixel clock settings from the config here, no 152012 reason to ignore DisplaySize. 152013 152014commit 7c16b68ab879f5b4b1aedfc6b2aadbe56193dd19 152015Author: Adam Jackson <ajax@redhat.com> 152016Date: Mon Mar 3 15:09:11 2008 -0500 152017 152018 1.5 has branched, start 1.5.99.x. 152019 152020commit f7ab2d3821e6bccc943f088e308fd58395a186d2 152021Author: David Nusinow <dnusinow@debian.org> 152022Date: Sun Mar 2 18:36:25 2008 -0500 152023 152024 Add missing file from previous commit. 152025 152026commit b5ce0e1d0b861dc5521fcd9db6287ed6da817726 152027Author: David Nusinow <dnusinow@debian.org> 152028Date: Sun Mar 2 17:12:02 2008 -0500 152029 152030 Bug #13860: Ensure that the DRI mode is in octal format. 152031 152032commit c934366424b0d20e013c84e6b94b226b20e7baa2 152033Author: Matthieu Herrb <matthieu.herrb@laas.fr> 152034Date: Sun Mar 2 19:27:53 2008 +0100 152035 152036 use UTILS_SYS_LIBS to pass SYS_LIBS to utils/ioports correctly 152037 152038commit 2bb9c1f36f685044b837f42076dec2ea7d22d034 152039Author: David Nusinow <dnusinow@debian.org> 152040Date: Sat Mar 1 18:44:58 2008 -0500 152041 152042 bug #10008: Make Xvfb.1 document the correct default depth 152043 152044commit b0b9c811cda3e35a8f6d0813483f750602c55ff6 152045Author: George Sapountzis <gsap7@yahoo.gr> 152046Date: Sat Mar 1 20:24:50 2008 +0200 152047 152048 fix typo 152049 152050commit 3d642905477f4b1ec3223f1fbe0d0d37e959ec81 152051Author: George Sapountzis <gsap7@yahoo.gr> 152052Date: Sat Mar 1 16:18:18 2008 +0200 152053 152054 clean some "unused" warnings 152055 152056commit e7a6f79754816976d92857d55840262cccff80a6 152057Author: George Sapountzis <gsap7@yahoo.gr> 152058Date: Sat Mar 1 16:16:29 2008 +0200 152059 152060 glcore: split mesa and X in build system 152061 152062commit cdd46aa3cd2e720558186cdbe48d871ab385fcdd 152063Author: George Sapountzis <gsap7@yahoo.gr> 152064Date: Sat Mar 1 15:57:57 2008 +0200 152065 152066 configure: minor cleanup 152067 152068 - dmx 152069 152070 - darwin: remove from xorg options 152071 152072 - xephyr: libxv is client lib 152073 152074commit 8af2c39bcc4ddc4693d5a2597c9622fa17b6c272 152075Author: Maarten Maathuis <madman2003@gmail.com> 152076Date: Sat Mar 1 16:54:01 2008 +0100 152077 152078 Fix big mistake in commit fd41f46ac62033a724bd1f4612f19448a21c1224. 152079 152080 - When a mode is deleted, the name pointer is also free()'ed. 152081 - This leaves other modes with an invalid pointer. 152082 152083commit ef60632e200853680282016e32a7a9fb01882852 152084Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152085Date: Fri Feb 29 18:00:27 2008 -0500 152086 152087 dix: Modify callers of property and selection API to use new interfaces. 152088 152089commit cc76ea6e3ac6a405f0c198c4e62be40aa8d2b546 152090Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152091Date: Fri Feb 29 17:55:31 2008 -0500 152092 152093 XACE: Add generic support for property and selection polyinstantiation. 152094 152095commit 34bf308a9e66f1a2f48630a15b1802afad50ec24 152096Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152097Date: Fri Feb 29 18:00:23 2008 -0500 152098 152099 dix: Refactoring of selection code to allow for polyinstantiation. 152100 Introduces dixLookupSelection() API. 152101 Removes NumCurrentSelections from API. 152102 152103commit d5715f7beaad6816db27b01b67d7a3c69164d106 152104Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152105Date: Fri Feb 29 16:16:12 2008 -0500 152106 152107 dix: Refactoring of property code to allow for polyinstantiation. 152108 Introduces dixLookupProperty() API. 152109 152110commit c0e1959f285d7a7df66f42d55912a5a595decd0f 152111Author: Adam Jackson <ajax@redhat.com> 152112Date: Fri Feb 29 16:45:11 2008 -0500 152113 152114 On second thought, revert that, it'll make large pixmaps painfully slow. 152115 152116 Need to just fix the callers. 152117 152118commit 4a44fe7c8678360d0549cf0e0d63870f3623b1db 152119Author: Chris Wilson <chris@chris-wilson.co.uk> 152120Date: Fri Feb 29 16:43:14 2008 -0500 152121 152122 Bug #10465: Use calloc() for allocating PixmapRec's. 152123 152124commit 5d5fcc7198ca54fa9dc24fe974763eff9fddabee 152125Author: Chris Wilson <chris@chris-wilson.co.uk> 152126Date: Fri Feb 29 16:42:04 2008 -0500 152127 152128 Bug #10464: Set pixel value to 0 before FindColor() 152129 152130commit 8f0a4282f0ac33625eda9466e3db0bcef64e403a 152131Author: Chris Wilson <chris@chris-wilson.co.uk> 152132Date: Fri Feb 29 16:39:29 2008 -0500 152133 152134 Bug #10463: Always initialize reference pixel before AllocColor() 152135 152136commit b7eb92774a58639aff3f26bb28a3dcff910c3fb6 152137Author: Adam Jackson <ajax@redhat.com> 152138Date: Sat Jan 12 01:22:05 2008 -0500 152139 152140 EDID 1.4: Print additional CVT support data in the log. 152141 152142commit 31014d88aff8dc8a502cf0f26e4cde141e1a92f5 152143Author: Adam Jackson <ajax@redhat.com> 152144Date: Sat Jan 12 01:03:44 2008 -0500 152145 152146 EDID 1.4: Decode additional CVT support information. 152147 152148 Table 3.28: Display Range Limits & CVT Support Definition 152149 152150commit d1c48955f80692a32ab6adcee1384e3d298f471a 152151Author: Adam Jackson <ajax@redhat.com> 152152Date: Sat Jan 12 00:30:58 2008 -0500 152153 152154 Fix CVT abuse in DDCModesFromStandardTiming. 152155 152156 CVT is enough different from GTF that it should not be used on monitors 152157 that aren't expecting it. This brings us closer to what the spec says 152158 the correct behaviour is. 152159 152160commit 26c2e95fa5bf30726356cf4bdd0fea32a771a179 152161Author: Adam Jackson <ajax@redhat.com> 152162Date: Sat Jan 12 00:09:34 2008 -0500 152163 152164 Nuke a duplicate SYMFUNC(xf86CVTMode) 152165 152166commit ca5625b911e65fdfd410247b3eff57fedcfc1f79 152167Author: Adam Jackson <ajax@redhat.com> 152168Date: Sat Jan 12 00:08:00 2008 -0500 152169 152170 Add xf86GTFMode(). 152171 152172 This should probably be shared like xf86CVTMode(). 152173 152174commit 4cb4817c1072e1d31333db47d95f71d08bf0d1dc 152175Author: Adam Jackson <ajax@redhat.com> 152176Date: Fri Jan 11 23:38:48 2008 -0500 152177 152178 Remove the duplicate copy of xf86cvt.c 152179 152180commit e65e51a99b17a0510782775f010e9820ca567fcb 152181Author: Adam Jackson <ajax@redhat.com> 152182Date: Fri Jan 11 23:19:20 2008 -0500 152183 152184 Constify the built-in mode tables. 152185 152186commit 6828d8fc2b464e0755f46e3fbdeb07be0c38b620 152187Author: Adam Jackson <ajax@redhat.com> 152188Date: Fri Jan 11 22:57:42 2008 -0500 152189 152190 Clean up DisplayModeRec handling in many places. 152191 152192 Use xf86DuplicateMode() instead of rolling our own, and change 152193 malloc+memset to calloc. 152194 152195commit 85617b56e5e00e7b8c7d8ce5b49af289056921a7 152196Author: Adam Jackson <ajax@redhat.com> 152197Date: Fri Jan 11 22:55:37 2008 -0500 152198 152199 Remove some #if 0 guarding code duplicated in xf86Modes.c 152200 152201commit fd41f46ac62033a724bd1f4612f19448a21c1224 152202Author: Adam Jackson <ajax@redhat.com> 152203Date: Fri Jan 11 22:38:21 2008 -0500 152204 152205 Allow xf86DuplicateMode() to work correctly on read-only modes. 152206 152207 Before this it was meaningless to try to mark DisplayModeRec tables 152208 const, since the mode name would be emitted as a pointer to an 152209 anonymous string constant, and therefore would have to be fixed up by 152210 ld.so and so couldn't live in .rodata. With this change the standard 152211 mode lists can live in .rodata, and modes duplicated from them will 152212 have their names filled in on the fly. 152213 152214commit 8ac19d16a030ec416e30d3650cf43e024ada167f 152215Author: Adam Jackson <ajax@redhat.com> 152216Date: Fri Jan 11 21:58:21 2008 -0500 152217 152218 Add several comments documenting our EDID failures. 152219 152220commit 13bfa5937d43392f686b76a99ea6331e3dce5987 152221Author: Kristian Høgsberg <krh@redhat.com> 152222Date: Fri Feb 29 15:10:36 2008 -0500 152223 152224 GLX: Adjust to changes in DRI driver interface. 152225 152226commit d04ea267a4a51c16088d9ef429681a1edde536b1 152227Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152228Date: Thu Feb 28 21:53:16 2008 -0500 152229 152230 xselinux: Don't require device "read" permission for XQueryPointer. 152231 152232 These keyboard and pointer state polling calls are a real problem. 152233 152234commit 3fb17a3e647e926688c91a49a9b5b97f37dbc367 152235Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152236Date: Thu Feb 28 21:52:57 2008 -0500 152237 152238 xselinux: Log messages to both libaudit and Xorg.0.log. 152239 152240commit 4d91b1d5e422c5c460b1b7050baa9487a59b8aa8 152241Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152242Date: Thu Feb 28 21:52:32 2008 -0500 152243 152244 XACE: Adjust the location of the COMPOSITE creation hook. 152245 Avoids incrementing the refcnt if the hook fails. 152246 152247commit 5675ae1f72145e9b719c613023da525731b42461 152248Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152249Date: Thu Feb 28 18:12:52 2008 -0500 152250 152251 XACE: Call the creation hook to properly label COMPOSITE window pixmaps. 152252 152253commit 27bcf40cda1d7c52b189cc76528f1f51cbe1d5eb 152254Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152255Date: Thu Feb 28 16:43:43 2008 -0500 152256 152257 XACE: Fix instances of DixUnknownAccess at hook callsites. 152258 152259commit 453661a9e193a511cf5e54e6d330454163817316 152260Author: Peter Hutterer <peter@cs.unisa.edu.au> 152261Date: Wed Feb 27 20:43:50 2008 +1030 152262 152263 Xi: swap the control attached to a ChangeDeviceControl request. Bug #14170 152264 152265 Just swapping the request's data isn't enough, we need to swap the actual 152266 control as well. 152267 152268 X.Org Bug 14170 <http://bugs.freedesktop.org/show_bug.cgi?id=14170> 152269 152270commit aebd9dc252449747416b23c740a550d914275399 152271Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 152272Date: Thu Feb 28 01:22:31 2008 -0300 152273 152274 Oops, there's one more parenthesis. 152275 152276commit f616735f17a681e3add866bf199540327c322490 152277Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152278Date: Wed Feb 27 22:46:14 2008 -0500 152279 152280 xselinux: Prefix a few remaining error messages with "SELinux". 152281 152282commit e40cc5305bec656108077ab13fcc8e6e82b3707a 152283Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152284Date: Wed Feb 27 22:29:15 2008 -0500 152285 152286 xselinux: Don't throw BadAccess if DixUnknownAccess is passed in to a hook. 152287 152288 The avc will still appear, however, so that the callsite can be fixed. 152289 152290commit 3b1df47bd400be9dca34b5e5d1ac2b117f8cc4ed 152291Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152292Date: Wed Feb 27 22:28:54 2008 -0500 152293 152294 XACE: Require "manage" permission for XKBSetNames. 152295 152296commit 08afc70513e5496cc5cd8b76c8658c4292119e4b 152297Author: Dave Airlie <airlied@redhat.com> 152298Date: Thu Feb 28 10:45:41 2008 +1000 152299 152300 quirks: another LPL panel with the cm/mm wrong 152301 152302commit f6d51797a523ecc2d4a8f18b2681160fcd5d1d55 152303Author: Maarten Maathuis <madman2003@gmail.com> 152304Date: Wed Feb 27 23:44:39 2008 +0100 152305 152306 EXA: Let the driver decide what repeat/extend types are supported. 152307 152308commit 41aea6194bd29ab34cc166b3fd90eee64299ddf8 152309Author: Alan Hourihane <alanh@tungstengraphics.com> 152310Date: Wed Feb 27 17:06:27 2008 +0000 152311 152312 More checking for failed contexts/drawables 152313 152314commit 43e46a654fa7cf69f0c8bcb7f583008d96a98686 152315Author: Alan Hourihane <alanh@tungstengraphics.com> 152316Date: Wed Feb 27 16:58:21 2008 +0000 152317 152318 Fix context sharing between direct/indirect contexts 152319 152320commit d74b0327e8355546e6b41e8ce684f461d65fa9dc 152321Author: Dodji Seketeli <dodji@openedhand.com> 152322Date: Wed Feb 27 09:48:10 2008 +0100 152323 152324 [Xephyr/GL] unbreak the build 152325 152326 Don't touch Xephyr DRI stuff when not compiling in a DRI environment. 152327 152328commit 2257e20900460d85254734b595238e7ad5ee55c8 152329Author: Peter Hutterer <peter@cs.unisa.edu.au> 152330Date: Wed Feb 27 17:15:28 2008 +1030 152331 152332 dix: set dev->key to NULL after freeing it in CloseDevice. (Bug #12830) 152333 152334 Otherwise XkbRemoveResourceClient may try to dereference it lateron. 152335 152336 X.Org Bug 12830 <http://bugs.freedesktop.org/show_bug.cgi?id=12830> 152337 152338commit 3f0681fb0b2d0744c2ef892883ae5359b43a4a9a 152339Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152340Date: Tue Feb 26 23:14:29 2008 -0500 152341 152342 xselinux: Stub out selection protocol requests. 152343 152344commit 4632ea22580c31d44b0786321668d9e78f02900e 152345Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152346Date: Tue Feb 26 22:00:52 2008 -0500 152347 152348 xselinux: Rip out the selection code in advance of polyinstantiation support. 152349 152350 This resolves an issue where BadWindow errors were being thrown. 152351 152352commit c8e979b3b800e4c58be454daa0213285d4ee6510 152353Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152354Date: Tue Feb 26 21:34:22 2008 -0500 152355 152356 security: Register resource names in the server-side name registry. 152357 152358commit d12b7b6632fb4cf41d2e28c7792eaa503f25404a 152359Author: Dave Airlie <airlied@redhat.com> 152360Date: Wed Feb 27 07:08:00 2008 +1000 152361 152362 xf86Crtc: add higher level modesetting entry point. 152363 152364 For kernel modesetting work we need a bigger stick to beat the modesetting path 152365 152366commit dcc077c753137f37aa58231f1df3c4adb92b2c4b 152367Author: Michel Dänzer <michel@tungstengraphics.com> 152368Date: Tue Feb 26 12:13:06 2008 +0100 152369 152370 AIGLX: Switch to server context for calling DamageDamageRegion(). 152371 152372 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=14518 . 152373 152374commit 23ae68a4c74a2ec90b4130c37b0d0aec3f4082ce 152375Author: Peter Hutterer <peter@cs.unisa.edu.au> 152376Date: Tue Feb 26 15:12:36 2008 +1030 152377 152378 dix: before copying the classes the first time, set the VCK's classes to NULL. 152379 152380 XkbFinishDeviceInit does the following: 152381 xkbi->kbdProc= pXDev->kbdfeed->CtrlProc; 152382 pXDev->kbdfeed->CtrlProc= XkbDDXKeybdCtrlProc; 152383 152384 If we directly copy the device classes for the VCK, pXDev->kbdfeed->CtrlProc 152385 at the time of copying is still XbkDDXKeybdCtrlProc. So at some point 152386 XkbDDXKeybdCtrlProc is called, and calls itself, and calls itself, and... 152387 152388 Setting the device's classes to NULL seems to fix things. The memory isn't 152389 lost, it gets stored into the devPrivates and freed at device closing time. 152390 152391commit ce9fb2f8c4610016e49321018fc9b24729380afc 152392Author: Peter Hutterer <peter@cs.unisa.edu.au> 152393Date: Tue Feb 26 13:19:54 2008 +1030 152394 152395 Xi: MDs return the paired device's ID in attached field of ListInputDevices. 152396 152397commit c46f7b62d2bf9f35937cfef98325ed904895396a 152398Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 152399Date: Mon Feb 25 18:14:08 2008 -0300 152400 152401 Fine, we don't need pciInit() anymore. Nuke, nuke, nuke... 152402 152403commit a9050d54249ed9f54c6fe59143b846c9c7548f59 152404Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 152405Date: Mon Feb 25 18:10:18 2008 -0300 152406 152407 Jesus, pciInit() was called 32 times in my machine without need! 152408 152409 xf86scanpci() will always call pciInit() before any other function, so 152410 there's no need to guarantee it after. 152411 152412commit 81272f7ec9a3465e1d102c2ce627a45f92268857 152413Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 152414Date: Mon Feb 25 17:39:33 2008 -0300 152415 152416 Some doc clean up: 152417 clean up legacy things in os-support/bus/Pci.c. 152418 152419 Put InitOutput() comment to its right place. 152420 152421commit 7c582dd6de27d2f4fedf73319d2dea2bfb240efa 152422Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 152423Date: Mon Feb 25 17:31:37 2008 -0300 152424 152425 remove lnxPciInit() unused function. 152426 152427commit 9727db88d57089be6483104de435626cdbad883a 152428Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 152429Date: Mon Feb 25 17:08:07 2008 -0300 152430 152431 No more "-scanpci" given that we already have it in libpciaccess. 152432 152433 (Yeah, lets nuke dead code!) 152434 152435commit f19f7b8e53ed6609fc1fdd272de5521417946209 152436Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 152437Date: Mon Feb 25 17:07:07 2008 -0300 152438 152439 Clean some garbage caused when pciaccess was merged: 152440 152441 FindPCIVideoInfo() function isn't need anymore. 152442 152443 xf86scanpci() is being called only once so we don't need permanent 152444 (static) variables there. 152445 152446 restorePciState() is not used for now (until we find why multiple 152447 cards aren't working). 152448 152449commit d61f481a4455dd2a94674d2b7b26429cf9dcece3 152450Author: Peter Hutterer <peter@cs.unisa.edu.au> 152451Date: Mon Feb 25 17:52:45 2008 +1030 152452 152453 Xi: remove GetPairedPointer handling. 152454 152455 obsolete, ListInputDevices provides this information now. 152456 152457commit 27b21278939a06ab6a22c9902f202eae6c5fae72 152458Author: Peter Hutterer <peter@cs.unisa.edu.au> 152459Date: Mon Feb 25 17:00:42 2008 +1030 152460 152461 Xi: remove ungrab handling of ExtendedUngrabDevice request. 152462 152463 This can be done by UngrabDevice, no need for separate codepaths. 152464 152465commit 4b50e71bf127c8e0f289e3b76c786f0398effe65 152466Author: Adam Jackson <ajax@redhat.com> 152467Date: Sun Feb 24 20:25:13 2008 -0500 152468 152469 Bug #13736: Fix %bx in VBEGetPixelClock to match spec. 152470 152471 Reported by Yannick Henault. 152472 152473commit 6dc369028d3ca741de57ad78febf2f5f82e0696e 152474Author: Jeremy Huddleston <jeremy@tifa.local> 152475Date: Sat Feb 23 00:01:02 2008 -0800 152476 152477 XQuartz: Quit now properly warns the user 152478 (cherry picked from commit ed3d7b3959c2a0cb63e37210455bcc6cf195b807) 152479 152480commit e6a4cde16dc99ea02ac93da1d1b9517b1073d159 152481Author: Adam Jackson <ajax@redhat.com> 152482Date: Fri Feb 22 18:36:29 2008 -0500 152483 152484 Use the client-side XKB headers for the config utilities 152485 152486commit 347db49ebe4596db16455ea8a1a608cfa826c5c7 152487Author: Adam Jackson <ajax@redhat.com> 152488Date: Fri Feb 22 16:05:33 2008 -0500 152489 152490 s/via/openchrome/ in the autoconfig logic. 152491 152492 Xorg's via driver is effectively dead anyway. 152493 152494commit 69f782676797744815ff76b8a11b11178066f501 152495Author: Adam Jackson <ajax@redhat.com> 152496Date: Fri Feb 22 16:04:35 2008 -0500 152497 152498 Match Xephyr DRI definitions to the ones in xf86dri.h 152499 152500commit 060a66b6e2feddba43ed207b6fcf2cf1f7fe39fd 152501Author: Adam Jackson <ajax@redhat.com> 152502Date: Mon Dec 24 15:55:58 2007 -0500 152503 152504 Normalize swapped dispatch for Fixes{ChangeSaveSet,SelectSelectionInput} 152505 152506commit 24bebdded44a9e184455b4fee7800257fee81efb 152507Author: Adam Jackson <ajax@redhat.com> 152508Date: Mon Dec 24 15:07:49 2007 -0500 152509 152510 fbFillRegionTiled() is now dead code. 152511 152512 Only ever called from the old PaintWindow* screen hooks, but those are 152513 gone now. As a pleasant side effect, fb loses its #ifdef PANORAMIX. 152514 152515commit ee21aba6be0078949204e315ddfffd99de60c2f1 152516Author: Adam Jackson <ajax@redhat.com> 152517Date: Mon Dec 24 13:13:19 2007 -0500 152518 152519 Fix Xinerama's consolidated visual handling. 152520 152521 Formerly the code claimed it could only handle up to 256 visuals, which 152522 was true. Also true, but not explicitly stated, was that it could only 152523 handle visuals with VID < 256. If you have enough screens, and subsystems 152524 that add lots of visuals, you can easily run off the end. (Made worse 152525 because we allocate visual IDs from the same pool as XIDs.) If your app 152526 then chooses a visual > 256, then the Xinerama code would throw BadMatch 152527 on CreateColormap and your app wouldn't start. 152528 152529 With this change, PanoramiXVisualTable is gone. Other subsystems that 152530 were using it as a translation table between each screen's visuals now 152531 use a PanoramiXTranslateVisual() helper. 152532 152533commit a4202b898f07dd733590ae5adb21e48775369781 152534Author: Daniel Stone <daniel@fooishbar.org> 152535Date: Fri Feb 22 18:22:58 2008 +0100 152536 152537 XKB: Actually use the keymap we compile at startup 152538 152539 During XkbInitKeyboardDevice, we compiled a keymap and promptly threw it away; 152540 brief inspection revealed the embarassingly simple problem. Sorry. 152541 152542commit c14fd2a5cb3f45d5c4502e09f55f5e3732c5e698 152543Author: Dodji Seketeli <dodji@seketeli.org> 152544Date: Thu Feb 21 15:33:02 2008 +0100 152545 152546 [Xephyr/GL] properly route expose event on GL drawables 152547 152548 When an expose event happens on an host GL window paired with an 152549 internal drawable, route that expose event to the clients listening 152550 to the expose event on the internal drawable. 152551 152552commit 437c78ef9ff1177e04b3d6781b5805d89b2ab81a 152553Author: Dodji Seketeli <dodji@seketeli.org> 152554Date: Thu Feb 21 15:29:27 2008 +0100 152555 152556 [Xephyr/GL] don't crash when the host returns a NULL server string 152557 152558commit f343265a289724c81017f089c024a7618267c4e3 152559Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152560Date: Fri Feb 15 19:53:45 2008 -0500 152561 152562 XACE: Make the default window background state configurable per-window. 152563 152564 To recap: the original XC-SECURITY extension disallowed background "None" if 152565 the window was untrusted. XACE 1.0 preserved this check as a hook function. 152566 XACE pre-2.0 removed the hook and first abolished background "None entirely, 152567 then restored it as a global on/off switch in response to Bug #13683. 152568 Now it's back to being per-window, via a flag instead of a hook function. 152569 152570commit 7c2f0a8befb310707ea923dbcdfde84521e52c88 152571Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152572Date: Fri Feb 15 19:17:40 2008 -0500 152573 152574 Remove COMPOSITE ifdefs around WindowRec bitfield as it has no ABI effect. 152575 152576commit 5cb9e15562a32c1f102d94d5e15d5fd298baff3f 152577Author: Eric Anholt <eric@anholt.net> 152578Date: Wed Feb 20 10:36:06 2008 -0800 152579 152580 EXA: Fix Render acceleration in copy and tiling cases. 152581 152582 Code shuffling in a634c9b03494ba80aeec28be19662ac96657cc23 broke this by 152583 leaving pSrcPixmap = NULL. 152584 152585commit f14a62f823e257f92745bbcde11838f2ddd32ac8 152586Author: Peter Hutterer <peter@cs.unisa.edu.au> 152587Date: Tue Feb 19 21:44:10 2008 +1030 152588 152589 dix: set evlen to the size of the reallocated memory. 152590 152591 What a good idea this is... I'm very proud of myself. 152592 152593commit 67a78e84a81571cedaf7fd214a21ce1bbdc4fb3b 152594Author: Peter Hutterer <peter@cs.unisa.edu.au> 152595Date: Tue Feb 19 21:31:50 2008 +1030 152596 152597 Revert "xfree86: plug a memory leak in xf86LoadModules." 152598 152599 This reverts commit 3abce3ea2b1f43bd111664d4a57e5ccd282acab0 and 152600 6cbaf15e6109ba77819c4070f5b46c78237ec460. 152601 152602 The memory returned to xf86LoadModule was allocated in doLoadModule, which 152603 calls the respective module's PreInit. As it turns out, input and output 152604 drivers store a pointer to the module elswhere, so freeing it in 152605 xf86LoadModule is a bad idea. 152606 152607 For further reference: hw/xfree86/common/xf86Helper.c 152608 Input drivers: xf86InputDriverList[blah]->module = module; 152609 Output drivers: xf86DriverList[blah]->module = module; 152610 152611 Unloading the module would not look pretty then. 152612 152613commit 5f5ec5db35e82ddd9659763875e5d6c63cf1b691 152614Author: Jeremy Huddleston <jeremy@tifa.local> 152615Date: Tue Feb 19 00:00:11 2008 -0800 152616 152617 XQuartz: unsetenv("DISPLAY") before startx if X11.app can't connect to it. 152618 Also fix casting to silence warning. 152619 (cherry picked from commit a5cbf78471ec6e6ad672dc00118fc7edbd6ddc7c) 152620 152621commit 750d70267679ddee10590c80ec621d890bd3d4a7 152622Author: Peter Hutterer <peter@cs.unisa.edu.au> 152623Date: Sun Feb 3 10:25:15 2008 +1030 152624 152625 dix: Ensure enough memory for ClassesChangedEvent for a new device. 152626 152627 Before we enable the device through the driver, we size it up and make sure 152628 that the events in the event list contain enough bytes for a possible 152629 ClassesChangedEvent lateron. 152630 152631commit 74628d639719815c1beff4cac84662fa41c55925 152632Author: Peter Hutterer <peter@cs.unisa.edu.au> 152633Date: Sun Feb 3 10:15:40 2008 +1030 152634 152635 dix: change GetEventList to return length of list and set parameter in place. 152636 152637 Changing DDXs to use new prototype too. 152638 152639commit 77dba004a9aaf35f183f61ff6875a491a52aa030 152640Author: Peter Hutterer <peter@cs.unisa.edu.au> 152641Date: Sun Feb 3 10:10:46 2008 +1030 152642 152643 dix: add InputEventListLen and SetMinimumEventSize 152644 152645 The latter is used to increase the amount of allocated memory for the event 152646 list. This will be needed for ClassesChangedEvents that can be of more or less 152647 arbitrary size (larger than 32 anyway). 152648 152649commit 6cbaf15e6109ba77819c4070f5b46c78237ec460 152650Author: Peter Hutterer <peter@cs.unisa.edu.au> 152651Date: Tue Feb 19 12:02:22 2008 +1030 152652 152653 xfree86: guard against NULL-pointer dereferences in xf86LoadModules. 152654 152655 Should have been part of 3abce3ea2b1f43bd111664d4a57e5ccd282acab0, but I 152656 forgot to git-update-index before I committed. 152657 152658 Thanks to Bill Crawford for pointing it out. 152659 152660commit 3fe64d8d271aea0863bf01b0376f3eceec0c90b5 152661Author: Peter Hutterer <peter@cs.unisa.edu.au> 152662Date: Sun Feb 3 09:56:19 2008 +1030 152663 152664 Move input event list initialisation and storage from DDX to DIX. 152665 152666 Rather than letting the DDX allocate the events, allocate them once in the DIX 152667 and just pass it around when needed. 152668 152669 DDX should call GetEventList() to obtain this list and then pass it into 152670 Get{Pointer|Keyboard}Events. 152671 152672commit 09a8fc5c7a79ca22fc23224bb544f2e709681f3f 152673Author: Peter Hutterer <peter@cs.unisa.edu.au> 152674Date: Sun Feb 3 07:47:18 2008 +1030 152675 152676 Xi: make SizeDeviceInfo public and re-use from CreateClassesChangedEvent. 152677 152678commit 3abce3ea2b1f43bd111664d4a57e5ccd282acab0 152679Author: Arjan van de Ven <arjan@infradead.org> 152680Date: Mon Feb 18 18:13:10 2008 +1030 152681 152682 xfree86: plug a memory leak in xf86LoadModules. 152683 152684 LoadModule() returns the only reference to a fresh piece of memory (a 152685 ModuleDescPtr). Sadly, xf86LoadModules dropped the return value on the floor 152686 leaking memory for each module it loaded. 152687 152688 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 152689 152690commit 6dc71f6b2c7ff49adb504426b4cd206e4745e1e3 152691Author: Arjan van de Ven <arjan@infradead.org> 152692Date: Mon Feb 18 17:52:37 2008 +1030 152693 152694 xfree86: plug memory leak in InitPathList. 152695 152696 All the failure paths were very diligent in freeing the "fullpath" temporary 152697 string, but the success case was not. All the content only got strdup()d, so 152698 it's not live memory anymore. 152699 152700 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 152701 152702commit 1bec6ad8977cefa49cc297a310f5eb0b7cd0b8bc 152703Author: Arjan van de Ven <arjan@infradead.org> 152704Date: Mon Feb 18 17:46:04 2008 +1030 152705 152706 xfree86: plug memory leak in xf86LogInit() 152707 152708 xf86LogInit allocates a piece of memory, stores it in lf. LogInit() will then 152709 effectively strdup it, but lf is never freed again. 152710 152711 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 152712 152713commit 6d9d18bf7708eb9e640ef732b5dc4c99aa6d5feb 152714Author: Peter Hutterer <peter@cs.unisa.edu.au> 152715Date: Mon Feb 18 17:25:15 2008 +1030 152716 152717 dix: add support for WindowAccessAllowAll in device access checking. 152718 152719 AllowAll is the last check before the parent window is checked. This allows 152720 to override a DenyAll in the parent window in a simpler way than before (the 152721 previous method required all devices to be in the permit list). 152722 152723commit 088067c891a78670d9509f48f56bf3ff9c16a30d 152724Author: Peter Hutterer <peter@cs.unisa.edu.au> 152725Date: Mon Feb 18 16:46:07 2008 +1030 152726 152727 Xi: even if ChangeDeviceHierarchy fails, we may need to send an event. 152728 152729 Changes are committed instantly, so if at least one change was successful, we 152730 must send an event to the client, even if subsequent ones fail. 152731 152732commit 660739c6bc84cb74f43a277052ce163fae654417 152733Author: Benjamin Close <Benjamin.Close@clearchain.com> 152734Date: Fri Feb 15 13:36:34 2008 +1030 152735 152736 dix: Handle the case where a device cursor was removed correctly 152737 152738 In the case that the device cursor was the first in the list of cursors 152739 the window knew about, unsetting the cursor could lead to a segfault 152740 due to pPrev being NULL. Instead catch the condition and correctly remove 152741 the node from the list. Since there is no cursor now set on the device, 152742 we simply return success as the parent windows cursor will propogate 152743 down later. 152744 152745 Signed-off-by: Peter Hutterer <peter@cs.unisa.edu.au> 152746 152747commit 24db28230120ecc2b65b25164b6e7b407970f9e2 152748Author: Peter Hutterer <peter@cs.unisa.edu.au> 152749Date: Mon Feb 18 16:16:19 2008 +1030 152750 152751 dix: set num_events to 1 by default. (no RawDeviceEvents anymore) 152752 152753 This should have been part of acb412d5399f865b5048fdd683147cea0f19f0c1. 152754 152755commit 45d00433e7bfbab476dd02ec92a9fbee40af5dad 152756Author: Dan Nicholson <dbn.lists@gmail.com> 152757Date: Tue Jan 15 19:43:16 2008 -0800 152758 152759 glx: Use glapi sources from the mesa tree 152760 152761 With recent mesa HEAD, the glapi sources used only in the xserver glx 152762 module are carried in the mesa tree. Previously, these were generated 152763 separately and committed to the xserver tree. 152764 152765 The build is changed to symlink these files like the other mesa sources. 152766 This reduces the chance for mismatches between mesa's glX API and the 152767 xserver's glX API. 152768 152769commit fbd776894658e7afb2c55dc8582b2a3efab78a3c 152770Author: Daniel Stone <daniel@fooishbar.org> 152771Date: Sat Feb 16 20:01:18 2008 +0200 152772 152773 XKB: Ditch XkbFileInfo 152774 152775 Sorry about the megacommit, but this touches on a lot of stuff. 152776 152777 Get rid of XkbFileInfo, which was pretty seriously redundant, and move the 152778 only useful thing it had (defined) into XkbDescRec. defined will be removed 152779 pretty soon anyway. Is the compat map pointer non-NULL? Then you have a 152780 compat map, congratulations! Anyhow, I digress. 152781 152782 All functions that took an XkbFileInfoPtr now take an XkbDescPtr, _except_ 152783 XkmReadFile, which returns an XkbDescPtr *, because people want to deal in 152784 XkbDescPtrs, not XkbDescRecs. 152785 152786commit e5f002eddef1abe324033a3155f01d048536a48d 152787Author: Daniel Stone <daniel@fooishbar.org> 152788Date: Sat Feb 16 20:00:47 2008 +0200 152789 152790 XkbProcessOtherEvent: Don't depend on now-removed header 152791 152792 We don't do XKBsrv.h anymore. 152793 152794commit 2d256f098ae05033ad76672d5ebdb9dfa7e6b995 152795Author: Daniel Stone <daniel@fooishbar.org> 152796Date: Sat Feb 16 19:57:37 2008 +0200 152797 152798 XKB: Always set size correctly in XkbCopyKeymap's geometry routines 152799 152800 We were forgetting to set the sizes for sections and rows and a couple of 152801 other misc bits in XkbCopyKeymap's geometry. Sort that out, and add a 152802 couple of clarifying comments along the way. 152803 152804commit ab79110a84b2d299ecae0605fa535edbebd99565 152805Author: Daniel Stone <daniel@fooishbar.org> 152806Date: Sun Feb 3 23:48:57 2008 +1100 152807 152808 XKB: Remove support for pre-built keymaps 152809 152810 Don't load prebuilt keymaps anymore. 152811 152812commit 1332343910a1880854bae21a790cff198855cd24 152813Author: Daniel Stone <daniel@fooishbar.org> 152814Date: Sun Feb 3 23:43:18 2008 +1100 152815 152816 XKB: Remove usage of client-side types 152817 152818 Since we're no longer sharing with Xlib, don't pass Displays and XPointers 152819 everywhere. 152820 152821commit 534fc5140b039a8c98ab715d0a6740d513b41209 152822Author: Daniel Stone <daniel@fooishbar.org> 152823Date: Sun Feb 3 23:30:22 2008 +1100 152824 152825 XKB: Remove a bunch of mad ifdefs 152826 152827 We have SEEK_SET and size_t, seriously. Also use DebugF instead of 152828 ifdef DEBUG, and ditch a couple of random bits that were never used. 152829 152830commit 0f12a448dcbbbf1f40aa98d09c9d25ee511c5bbf 152831Author: Daniel Stone <daniel@fooishbar.org> 152832Date: Sun Feb 3 23:15:39 2008 +1100 152833 152834 XKB: Deprecate XKBSRV_NEED_FILE_FUNCS 152835 152836 There's no point in having the function definitions be conditional, so 152837 whatever. 152838 152839commit 68bd7ac1930b5cffb6657b8d5f5bf8ae58eae8d9 152840Author: Daniel Stone <daniel@fooishbar.org> 152841Date: Sun Feb 3 23:12:15 2008 +1100 152842 152843 XKB: Move headers into the server tree 152844 152845 We need to start breaking the XKB API to enforce sanity, so drag whichever 152846 headers we need to do so into the server tree, as the client API is set in 152847 stone, being part of Xlib. 152848 152849commit 37867626e314e74031378a8a4ff06f69b899a3b2 152850Author: Daniel Stone <daniel@fooishbar.org> 152851Date: Fri Feb 1 14:41:04 2008 +1100 152852 152853 main(): Remove uncredible failure 152854 152855 NO, NO, NO. 152856 152857 NO. 152858 152859 The only way this could possibly be worse is if you were viewing it in 152860 Comic Sans. 152861 152862commit e4eb7e5842f0f56f07e9cf3b16249c172d0a975d 152863Author: Daniel Stone <daniel@fooishbar.org> 152864Date: Wed Jan 30 23:24:14 2008 +1100 152865 152866 XKB: Delete xkberrs.c 152867 152868 Get rid of the XKB errors code to save a bunch of space. 152869 152870commit a48cc88ea2674c28b69b8d738b168cbafcf4001f 152871Author: Maarten Maathuis <madman2003@gmail.com> 152872Date: Sun Feb 17 18:47:28 2008 +0100 152873 152874 Fix rotation for multi-monitor situation. 152875 152876 - The (x,y)-coordinates of the crtc were not being passed as xFixed values, which made it an obscure bug to find. 152877 - Fix bug #13787. 152878 152879commit 70c0592a97c7dc9db0576d32b3bdbe4766520509 152880Author: Maarten Maathuis <madman2003@gmail.com> 152881Date: Sun Feb 17 11:21:01 2008 +0100 152882 152883 Resize composite overlay window when the root window changes. 152884 152885 - This allows some compositing managers to work, even after randr12 has changed the root window size. 152886 - Thanks to ajax for figuring out the best place to put this. 152887 - Example: 152888 - xf86RandR12SetMode() calls EnableDisableFBAccess(). 152889 - That calls xf86SetRootClip() which in turn calls ResizeChildrenWinSize(). 152890 - The final step is the call to PositionWindow(). 152891 152892commit b95059c20746a71e60ef152bb627e1d5c2210d75 152893Author: Jeremy Huddleston <jeremy@tifa-2.local> 152894Date: Sat Feb 16 01:33:13 2008 -0800 152895 152896 Added Xquartz.plist to EXTRA_DIST 152897 (cherry picked from commit 70f9495e0c1f1459507064b673fe57b1c90d3c2c) 152898 152899commit d103820bb8635c63b34b85b45cad95ed9c152d90 152900Author: Jeremy Huddleston <jeremy@tifa-2.local> 152901Date: Sat Feb 16 00:35:50 2008 -0800 152902 152903 Fixed unescaped newline in EXTRA_DIST 152904 (cherry picked from commit 567c172c4d400fdfe69e7b096a3877fce5c2de9f) 152905 152906commit a7d936c4ac8e3d7227ecbfe0ddc6cc257b450458 152907Author: Jeremy Huddleston <jeremy@tifa.local> 152908Date: Mon Feb 11 16:09:03 2008 -0800 152909 152910 XQuartz: Fixed EXTRA_DIST to include localization 152911 (cherry picked from commit 1b338c2a9330c85490a7c24d52adf24b124b70e6) 152912 152913commit 0ffb6a3ad010e80fe8f973fc228d549f9dd3effd 152914Author: Kristian Høgsberg <krh@redhat.com> 152915Date: Thu Feb 14 22:20:56 2008 -0500 152916 152917 GLX: Implement support for TTM BO based TFP when available. 152918 152919commit 879515b1399f87a47010532af70f34b9b09e2a9b 152920Author: Kristian Høgsberg <krh@sasori.boston.redhat.com> 152921Date: Mon Feb 4 13:13:35 2008 -0500 152922 152923 Add GLX provider for DRI2. 152924 152925commit b71dc54352bc56a889823040ec19c1d8e118a1f3 152926Author: Kristian Høgsberg <krh@sasori.boston.redhat.com> 152927Date: Mon Feb 4 11:58:24 2008 -0500 152928 152929 Add DRI2 module. 152930 152931commit 005e31d3de04e7003f84a94d30f2b75a9636266e 152932Author: Michel Dänzer <michel@tungstengraphics.com> 152933Date: Fri Feb 15 00:01:32 2008 +0000 152934 152935 AIGLX: Refactor code common between __glXDRI{drawableDestroy,releaseTexImage}. 152936 152937commit cd78f0d0fc08e4e2339ed09dad1a12802de7729c 152938Author: Pierre Willenbrock <pierre@pirsoft.de> 152939Date: Thu Feb 7 21:28:28 2008 +0100 152940 152941 AIGLX: Fix crash after client using GLX_EXT_texture_from_pixmap died. 152942 152943commit 3e0353c78571890f849a1db47b0540aacc6793bc 152944Author: Adam Jackson <ajax@redhat.com> 152945Date: Fri Feb 15 06:42:48 2008 +1100 152946 152947 Bah, macros are hard. 152948 152949commit c38feeb1492d9a47379b2e4d77dbadd8c421d17f 152950Author: Maarten Maathuis <madman2003@gmail.com> 152951Date: Thu Feb 14 18:52:47 2008 +0100 152952 152953 Fix typo in commit b8b7cdf6df3d338c50db670ce4cfd245f9fa8844. 152954 152955commit b8b7cdf6df3d338c50db670ce4cfd245f9fa8844 152956Author: Adam Jackson <ajax@redhat.com> 152957Date: Fri Feb 15 01:31:46 2008 +1100 152958 152959 Silence useless debug spew from XFree86-Misc extension. 152960 152961commit 4573cb2ce4b859744118d9a33de3599f033cdd2b 152962Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152963Date: Wed Feb 13 20:16:56 2008 -0500 152964 152965 security: Revise set of permissions granted to untrusted clients. 152966 152967 Bug #14480: untrusted access broken on 7.3. 152968 152969commit e99aadbc2635e87543fc9980d8156c3ede83544f 152970Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152971Date: Wed Feb 13 19:46:29 2008 -0500 152972 152973 xselinux: Add use to permission map for devices. 152974 152975commit ae43d835bdaef96c3c73d7cee5b105f07340833d 152976Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 152977Date: Wed Feb 13 19:39:49 2008 -0500 152978 152979 XACE: Change access modes for some device-related requests. 152980 152981 Opening a device is not really "reading" it. 152982 Requests that globally configure a device should require "manage" access. 152983 152984commit 921c298c4cdd2c879403ebdacbef48129058cad4 152985Author: Peter Hutterer <peter@cs.unisa.edu.au> 152986Date: Thu Feb 14 09:48:19 2008 +1030 152987 152988 Xi: QueryDevicePointer doesn't have a shared field anymore. 152989 152990commit b885588756450e2fa25fdd191e300e1c5dd37ec3 152991Author: Peter Hutterer <peter@cs.unisa.edu.au> 152992Date: Thu Feb 14 09:47:31 2008 +1030 152993 152994 Xi: remove FakeDeviceData calls. And one leftover from GrabAccessControl. 152995 152996 FakeDeviceData needs more thoughtful integration. 152997 152998commit c99a9a97290c51bf2843f42d5683888cea8a3ff6 152999Author: Peter Hutterer <peter@cs.unisa.edu.au> 153000Date: Thu Feb 14 09:43:49 2008 +1030 153001 153002 Xi: remove raw device events. 153003 153004 Need to be better refined. 153005 153006commit b35a9efa16d9e14378a95c11fa39171b78710996 153007Author: Peter Hutterer <peter@cs.unisa.edu.au> 153008Date: Thu Feb 14 09:41:49 2008 +1030 153009 153010 Xi: Remove files for GrabAccessControl. 153011 153012commit b81c330f70c4a4c8f37216aa4856c7f8db93a375 153013Author: Peter Hutterer <peter@cs.unisa.edu.au> 153014Date: Thu Feb 14 09:32:54 2008 +1030 153015 153016 mi: remove RawDeviceEvent's ID changing. 153017 153018commit acb412d5399f865b5048fdd683147cea0f19f0c1 153019Author: Peter Hutterer <peter@cs.unisa.edu.au> 153020Date: Thu Feb 14 09:27:51 2008 +1030 153021 153022 dix: Don't generate RawDeviceEvents anymore. 153023 153024 These events have been removed from the protocol pending some more refinement 153025 and more thoughtful integration. 153026 153027commit ba69904148acf755bec8fbda2eb869144f0ef7d4 153028Author: Adam Jackson <ajax@redhat.com> 153029Date: Sun Feb 3 22:49:23 2008 +1100 153030 153031 Simplify dlloader a bit more. 153032 153033commit f750ce53ac450824d0c792d11eafdf311cf8abc0 153034Author: Adam Jackson <ajax@redhat.com> 153035Date: Fri Feb 1 17:33:48 2008 +1100 153036 153037 Remove some braindamage from ModuleDesc. 153038 153039 Yay dead code elimination. 153040 153041commit e91ff09568c5579818b6641e88e95c6fe122cbe7 153042Author: Adam Jackson <ajax@redhat.com> 153043Date: Fri Feb 1 17:21:34 2008 +1100 153044 153045 Friends don't let friends call xf86AddModuleInfo. 153046 153047 That code only existed to leak memory. It can go now, thanks. 153048 153049commit 0bdd20a0454c94f902fd4874855125bf7510fcf5 153050Author: Adam Jackson <ajax@redhat.com> 153051Date: Fri Feb 1 16:15:58 2008 +1100 153052 153053 Eradicate the VTInit code. 153054 153055 "This option should rarely be used." Never sounds like a better idea. 153056 153057commit 24089b06243101b1bff4f2fd79fcbfd6a93992d5 153058Author: Adam Jackson <ajax@redhat.com> 153059Date: Fri Feb 1 16:11:13 2008 +1100 153060 153061 Don't bother warning people about the keyboard driver rename, just do it. 153062 153063commit 2aaf6ac134fa9f61984b1c03929c7b596c971cc8 153064Author: Adam Jackson <ajax@redhat.com> 153065Date: Fri Feb 1 16:10:23 2008 +1100 153066 153067 Don't even attempt to parse -bpp in xfree86 DDX option parsing. 153068 153069 This hasn't been valid since 1999. Get with the times, man. 153070 153071commit ef77e4c4419703b5a802f3eee92476a43f788a86 153072Author: Adam Jackson <ajax@redhat.com> 153073Date: Fri Feb 1 16:03:01 2008 +1100 153074 153075 Remove useless commentary from environment and argument processing. 153076 153077commit f3b3b37ec6197f8884417fbc26630d3a28f2e319 153078Author: Adam Jackson <ajax@redhat.com> 153079Date: Fri Feb 1 15:36:11 2008 +1100 153080 153081 Use strerror instead of errno values in user strings. 153082 153083commit b740b865e4c156a40adc6b63fcf215156fc9151a 153084Author: Adam Jackson <ajax@redhat.com> 153085Date: Fri Feb 1 15:16:01 2008 +1100 153086 153087 Silence an error message in ConfigureWindow that never happens. 153088 153089commit 89d3249c3e7611b22414bc84b10ca60aab4b9a77 153090Author: Adam Jackson <ajax@redhat.com> 153091Date: Fri Feb 1 15:11:27 2008 +1100 153092 153093 Silence FreeResource() 153094 153095commit 238b816469355159eea98600ca900e3baa8fa313 153096Author: Adam Jackson <ajax@redhat.com> 153097Date: Fri Feb 1 15:08:03 2008 +1100 153098 153099 X.Org Group -> X.Org Foundation 153100 153101commit 9113fa3de36e84bbae2727cace82b1cf0d709a86 153102Author: Adam Jackson <ajax@redhat.com> 153103Date: Fri Feb 1 15:03:57 2008 +1100 153104 153105 Silence the harmless FreeFontPath error message. 153106 153107commit 7c0709a736c0f3aa011de67dd2c2962585ab146e 153108Author: Adam Jackson <ajax@redhat.com> 153109Date: Fri Feb 1 14:59:58 2008 +1100 153110 153111 requestingClient is an xprintism, hide it for other servers. 153112 153113commit 62cfe8863823c0994f20555cb35ee3bacb9e5225 153114Author: Adam Jackson <ajax@redhat.com> 153115Date: Fri Feb 1 14:57:25 2008 +1100 153116 153117 Redact all mention of PanoramiX from user strings. 153118 153119commit eb8dc11d19dd0b5354de408578ab0cfb865df672 153120Author: Adam Jackson <ajax@redhat.com> 153121Date: Thu Jan 31 21:45:14 2008 +1100 153122 153123 Remove some MAX* #defines that never get used. 153124 153125commit 2ce35f6d45c3e1761d33b786520ff5ba56a3c518 153126Author: Adam Jackson <ajax@redhat.com> 153127Date: Thu Jan 31 21:39:48 2008 +1100 153128 153129 Simplify critical output flushing. 153130 153131commit 3eaecdd66e791e0f3d86b23ce10be057ca44c044 153132Author: Adam Jackson <ajax@redhat.com> 153133Date: Thu Jan 31 21:36:14 2008 +1100 153134 153135 Disable Record by default. 153136 153137commit aa5216e89797b600f382c04e3eaa657e808a5c3e 153138Author: Adam Jackson <ajax@redhat.com> 153139Date: Thu Jan 31 12:05:08 2008 +1100 153140 153141 Rip out useless indirection in the callback list management. 153142 153143commit 0d492b2166c4026b9078ffd86d89a31ebe590be4 153144Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153145Date: Tue Feb 12 19:59:10 2008 -0500 153146 153147 XACE: Move the selection access hook to its own function. 153148 153149commit 4be69e41c07b784f75918141eead70415c0ec4a4 153150Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153151Date: Mon Feb 11 16:03:25 2008 -0500 153152 153153 Revert "Bug #8937: Extension setup functions not called on server resets" 153154 153155 This reverts commit 5e946dd853a4ebc2722ae023429ce5797de3d7a6. 153156 153157 The devPrivates rework makes this workaround unnecessary. 153158 153159commit 8a244c2bc04a4184535e42d1f2f8c9e3e55678b4 153160Author: Jeremy Huddleston <jeremy@yuffie.local> 153161Date: Mon Feb 11 12:09:41 2008 -0800 153162 153163 XQuartz: Fixed layout of Applications->Customize menu 153164 153165 Now resizing it won't produce weir overlaps of the widgets. Thanks to 153166 Pelle Johansson for his help showing me how to do this. 153167 (cherry picked from commit ef3498e92d13c82633fdbe8120396bfbe1e7489a) 153168 153169commit 3570ca1cb76f084272f7d8bd22c4cfdf03201a77 153170Author: Jeremy Huddleston <jeremy@tifa.local> 153171Date: Sun Feb 10 18:27:52 2008 -0800 153172 153173 XQuartz: Added option for setting quartz-wm click-through preference 153174 (cherry picked from commit bf54c267cba97b2b3b9a621da0575776a388b2cb) 153175 153176commit 8004e160fa8cc75a3f1b7385fee64e5864b3b50a 153177Author: Jeremy Huddleston <jeremy@tifa.local> 153178Date: Sun Feb 10 16:34:40 2008 -0800 153179 153180 XQuartz: Converted interface to nib 3.x format 153181 (cherry picked from commit 656aaab95773bd70fc3504b68bf7e1d292891d93) 153182 153183commit 0b0a09797302ac2171db5df20fc5110aafc8efbb 153184Author: Peter Hutterer <peter@cs.unisa.edu.au> 153185Date: Thu Feb 7 15:48:04 2008 +1030 153186 153187 xkb: when copying the keymap, make sure the structs default to 0/NULL. 153188 153189 It actually does help if a pointer is NULL rather than pointing to nirvana 153190 when you're trying to free it lateron. Who would have thought? 153191 (cherry picked from commit 7a97ca667405a42d008265c3a870210cc1da97dd) 153192 153193commit 7018f280406eb0ef899a4046de274cfdd582881b 153194Author: Peter Hutterer <peter@cs.unisa.edu.au> 153195Date: Thu Feb 7 15:48:04 2008 +1030 153196 153197 xkb: when copying the keymap, make sure the structs default to 0/NULL. 153198 153199 It actually does help if a pointer is NULL rather than pointing to nirvana 153200 when you're trying to free it lateron. Who would have thought? 153201 153202commit 66f8001b61d12eaf4905ac71ccbb3f304914d00d 153203Author: Bart Trojanowski <bart@symbio-technologies.com> 153204Date: Thu Feb 7 21:26:54 2008 -0500 153205 153206 X86EMU: handle CPUID instruction 153207 153208 After trying to switch from X to VT (or just quit) the video-amd driver 153209 attempts to issue INT 10/0 to go to mode 3 (VGA). The emulator, running 153210 the BIOS code, would then spit out: 153211 153212 c000:0282: A2 ILLEGAL EXTENDED X86 OPCODE! 153213 153214 The opcode was 0F A2, or CPUID; it was not implemented in the emulator. 153215 This simple patch, against 1.3.0.0, handles the CPUID instruction in one of 153216 two ways: 153217 1) if ran on __i386__ or __x86_64__ then it calls the CPUID instruction 153218 directly. 153219 2) if ran elsewhere it returns a canned 486dx4 set of values for 153220 function 1. 153221 153222 This fix allows the video-amd driver to switch back to console mode, 153223 with the GSW BIOS. 153224 153225 Thanks to Symbio Technologies for funding my work, and ThinCan for 153226 providing hardware :) 153227 153228 Signed-off-by: Bart Trojanowski <bart@jukie.net> 153229 Acked-by: Eric Anholt <eric@anholt.net> 153230 153231commit de16a8c53046764dbdf26a87acc5c984ef00d818 153232Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153233Date: Thu Feb 7 20:14:16 2008 -0500 153234 153235 XACE: Correct some protocol error values in the colormap routines. 153236 153237commit 31934132a490b1b8ae73010c44e0b23217d8dab2 153238Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153239Date: Thu Feb 7 16:32:06 2008 -0500 153240 153241 xselinux: Use the device name in debugging output. 153242 153243commit 6dcb7d732bfeadc214228d68c5a13eef30248eb1 153244Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153245Date: Thu Feb 7 16:00:52 2008 -0500 153246 153247 xselinux: Split devPrivate state into subject and object records. 153248 153249commit 2259b144f0fd4855085a656111a0c64246733e78 153250Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153251Date: Thu Feb 7 14:35:02 2008 -0500 153252 153253 xselinux: Add getattr and setattr to the permission map for properties. 153254 153255commit 9d5edebe9664f4ac35b82830e000a72d5ca9cbe4 153256Author: Peter Hutterer <peter@cs.unisa.edu.au> 153257Date: Thu Feb 7 21:22:50 2008 +1030 153258 153259 Xi: finish XKB initialisation before copying the keymap. 153260 153261commit 96eafa3d4f05997dac8be6a8d66a92410511824c 153262Author: Peter Hutterer <peter@cs.unisa.edu.au> 153263Date: Tue Feb 5 16:01:56 2008 +1030 153264 153265 xkb: when copying sections, make sure num_rows is set too. 153266 (cherry picked from commit 41991fb991313202e8e6b513fe928ba14f8fcb87) 153267 153268commit 5c30327275509576b7848a5f842e7a1bffabe980 153269Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153270Date: Tue Feb 5 21:06:05 2008 -0500 153271 153272 XACE: Push the dix "structure" includes down to the security modules. 153273 153274commit bb1a577a6822f781f1e38d2434a13914e74f89aa 153275Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153276Date: Tue Feb 5 20:07:08 2008 -0500 153277 153278 XACE: Move the property access hook to its own function. 153279 153280commit 019ad5acd20e34dc2aa3b89cc426138db5164c48 153281Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153282Date: Tue Feb 5 15:44:41 2008 -0500 153283 153284 XFixes: squash a pointer/integer size mismatch warning. 153285 153286commit d3c36fe721edc55636438bc3e0e7a6c03f62784e 153287Author: liuhong <liuhong@devlinux-hong.sh.intel.com> 153288Date: Tue Feb 5 10:54:58 2008 +0800 153289 153290 validate mode clock for probed modes 153291 153292 Some modes claimed in monitor EDID data may not be supported by 153293 the monitor. So also validating the max clock for probed modes. 153294 153295commit 4b5b6e7baab58072a983d2ec136965f404c3a74a 153296Author: liuhong <liuhong@devlinux-hong.sh.intel.com> 153297Date: Tue Feb 5 10:54:10 2008 +0800 153298 153299 fix max clock unit 153300 153301 max clock from EDID data is in MHz, while we need KHz to validate modes. 153302 153303commit a56ef7aaa4b6ac13c8181f68fc7dad3ca89e6973 153304Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 153305Date: Sat Feb 2 23:04:46 2008 +0100 153306 153307 dix: Move motion history update until after screen crossing and clipping 153308 153309 Cross screen and clip the coordinates before updating the motion history 153310 so that it will have the same contents as the events that are reported. 153311 153312commit a0284d577aabea8406b72dd63773e341430ebe56 153313Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 153314Date: Sat Feb 2 23:03:51 2008 +0100 153315 153316 dix: Skip call to clipAxis for relative core-events 153317 153318 Relative events that generates both core and extention 153319 events will have its axis cliped and screen changed by 153320 miPointerSetPosition when the events are processed. For 153321 absolute and non core-generating relative events the 153322 axis must be clipped if we shouldn't end up completely 153323 outside the defined ranges (if any). 153324 153325commit d9e23c4ff1607a62164b34717ef9afd352ce2b94 153326Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 153327Date: Sat Feb 2 22:57:32 2008 +0100 153328 153329 Bug # 10324: dix: Add scaling of X and Y on the reported pointer-events 153330 153331 Restore the rescaling code for x and y axis when generating 153332 motion events. 153333 153334commit f04c0838699f1a733735838e74cfbb1677b15dc4 153335Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 153336Date: Sat Feb 2 22:45:31 2008 +0100 153337 153338 Bug # 10324: dix: Allow arbitrary value ranges in GetPointerEvents 153339 153340 Don't use a possitive value as a marker for if a max-value 153341 is defined on the valuators. Use the existence of a valid 153342 value range instead. This will also make it possible to 153343 define arbitrary start and end-values for min and max as 153344 long as min < max. 153345 153346commit 12e532403210c15a25200ef448bfe9701735ab20 153347Author: Magnus Vigerlöf <Magnus.Vigerlof@ipbo.se> 153348Date: Sat Feb 2 22:44:31 2008 +0100 153349 153350 dix: Always add valuator information if present 153351 153352 Send valuator information for all event types, not only for 153353 MotionEvents and absolute button events. 153354 153355commit 41991fb991313202e8e6b513fe928ba14f8fcb87 153356Author: Peter Hutterer <peter@cs.unisa.edu.au> 153357Date: Tue Feb 5 16:01:56 2008 +1030 153358 153359 xkb: when copying sections, make sure num_rows is set too. 153360 153361commit 521a7f26e088029707fb9a2bb80c9ddc734a3f8b 153362Author: Peter Hutterer <peter@cs.unisa.edu.au> 153363Date: Tue Feb 5 18:34:31 2008 +1030 153364 153365 mi: Only UpdateSpriteForScreen if we actually changed the screen. (Bug #12650) 153366 153367 X.Org Bug 12650 <http://bugs.freedesktop.org/show_bug.cgi?id=12650> 153368 153369commit 10617dc0fb7166ccd5b2e92fa708390c2d7d0d27 153370Author: Peter Hutterer <peter@cs.unisa.edu.au> 153371Date: Fri Feb 1 14:24:04 2008 +1030 153372 153373 xfree86: stick two more checks in for num_valuators < MAX_VALUATORS 153374 153375commit c68f063be639f39c2facbb496e8455e8e3771b41 153376Author: Matthieu Herrb <matthieu.herrb@laas.fr> 153377Date: Fri Feb 1 14:13:29 2008 +1030 153378 153379 xfree86: don't call xalloc from signal handlers when posting events. 153380 153381 Reviewed-by: Peter Hutterer <peter@cs.unisa.edu.au> 153382 153383commit e5e531ce96264518f278f572a4d0bfcd345aa618 153384Merge: 6c3868208 1692dcf19 153385Author: Peter Hutterer <peter@cs.unisa.edu.au> 153386Date: Thu Jan 31 16:36:17 2008 +1030 153387 153388 Merge branch 'master' into mpx 153389 153390 Conflicts: 153391 153392 dix/events.c 153393 153394commit 6c38682084e24b18bac0ea75729b1183aeb01600 153395Author: Benjamin Tissoires <tissoire@cena.fr> 153396Date: Thu Jan 31 15:51:32 2008 +1030 153397 153398 kdrive: AddInputClient takes a ClientPtr. 153399 153400commit 1692dcf197470d074f69d5af1608cb2ff1d08872 153401Author: Peter Hutterer <peter@cs.unisa.edu.au> 153402Date: Wed Jan 30 13:04:58 2008 +1030 153403 153404 dix: print out event type if a bogus pointer event occurs. 153405 153406commit d954f9c80348de294602d931d387e5cd1ef4b9a5 153407Author: Peter Hutterer <peter@cs.unisa.edu.au> 153408Date: Wed Jan 30 10:39:54 2008 +1030 153409 153410 xkb: don't update LEDs if they don't exist. (Bug #13961) 153411 153412 In some weird cases we call this function when there is no SrvLedInfo on the 153413 device. And it turns out null-pointer dereferences are bad. 153414 153415 X.Org Bug 13961 <http://bugs.freedesktop.org/show_bug.cgi?id=13961> 153416 153417commit df325be394e1f75c396b2768f81373f2989aef7c 153418Author: Coleman Kane <cokane@cokane.org> 153419Date: Tue Jan 29 09:47:00 2008 -0800 153420 153421 Bug 13101: xorg-server has a typo in hw/xfree86/os-support/bsd/i386_video.c 153422 153423 X.Org Bug #13101 <https://bugs.freedesktop.org/show_bug.cgi?id=13101> 153424 Patch #12360 <https://bugs.freedesktop.org/attachment.cgi?id=12360> 153425 153426commit 94f412cb7e954fe872fed979057cbdfbef953c6f 153427Author: Julien Goodwin <julien@studio442.com.au> 153428Date: Sun Jan 27 12:30:16 2008 +1100 153429 153430 Loader: Fix verbosity confusion 153431 153432 'Loading foo' is verbosity 3, whereas 'already built-in' is verbosity 0. 153433 This means that gdm's log would just be full of bare 'module already 153434 built-in' messages. 153435 153436commit 442838fcb3bf07ac57553ae5600d9e6c59a559bb 153437Author: Julien Goodwin <julien@studio442.com.au> 153438Date: Sun Jan 27 12:27:26 2008 +1100 153439 153440 xorg.conf.man: Fix monitor/output confusion in monitor positioning 153441 153442 On the Intel driver at least, LeftOf/RightOf/Above/Below in xorg.conf refers 153443 to output names, not monitor names. Fix nomenclature in xorg.conf.man. 153444 153445commit 5b8641a5fdc112c19e78ca2954878712e328d403 153446Author: Peter Hutterer <peter@cs.unisa.edu.au> 153447Date: Tue Jan 29 10:01:37 2008 +1030 153448 153449 xfree86: fix AlwaysCore handling. (Bug #14256) 153450 153451 Assume AlwaysCore being set by default, just like the other options. 153452 153453 X.Org Bug 14256 <http://bugs.freedesktop.org/show_bug.cgi?id=14256> 153454 153455commit 975ab11799c819a81da1dfe83505194410dbcb95 153456Author: Peter Hutterer <peter@cs.unisa.edu.au> 153457Date: Sat Jan 26 09:39:54 2008 +1030 153458 153459 config: don't reset connection info on disconnect. 153460 153461 If dbus is restarted, we try to connect again and this is difficult if the 153462 busname and/or busobject is not set. 153463 (cherry picked from commit 210eeef495770c1883c842ff003c28ce25f279d4) 153464 153465commit 7dde5a694a06efed0a9186f05d33f5be6f5dba71 153466Author: Peter Hutterer <peter@cs.unisa.edu.au> 153467Date: Fri Jan 25 13:54:47 2008 +1030 153468 153469 config: check connection != NULL before getting dbus' dispatch status. 153470 153471 (cherry picked from commit d23266522390a4ef7203ae7c062b2b920e45f9d7) 153472 153473commit f0ba7707161b8866e6fde32d6f25be6afcdecb48 153474Author: Peter Hutterer <peter@cs.unisa.edu.au> 153475Date: Fri Jan 25 13:45:22 2008 +1030 153476 153477 config: only shutdown libhal if the connection is valid. 153478 153479 Thanks to libdbus' extensive use of assert we won't just get an error, it'll 153480 bring the whole server down for us. 153481 (cherry picked from commit fb07fab2c07e7b0834724541dc47bfba02ba8574) 153482 153483commit 2cb0ebec2b85d96289c23c17cfdcdf97ef6877d2 153484Author: Peter Hutterer <peter@cs.unisa.edu.au> 153485Date: Fri Jan 25 11:48:13 2008 +1030 153486 153487 config: add a debug message, fix a whitespace error. 153488 153489 (cherry picked from commit 7732898aaa70e076000f6e6aa9420482896ed996) 153490 153491commit 4fc2d3cef8d7868b025aa14af7ed4b730e8f2c49 153492Author: Jeremy Huddleston <jeremy@tifa.local> 153493Date: Mon Jan 28 12:18:43 2008 -0800 153494 153495 Rootless: RootlessEnsureFrame: Added check for !IsRoot 153496 This was causing an issue with Apple-DRI and was reported here: 153497 http://trac.macosforge.org/projects/xquartz/ticket/51 153498 (cherry picked from commit 116800279d2ec783c63f43d3902627edde6a4cff) 153499 153500commit 16b0614c8c441eb1692bde9e58fadf54a9c6f00f 153501Merge: 1a88aed5c e915a2639 153502Author: David Nusinow <dnusinow@debian.org> 153503Date: Sat Jan 26 16:13:25 2008 -0500 153504 153505 Merge commit 'upstream/master' 153506 153507commit 0ac175597712edfeae676f536746a4d26d625e30 153508Merge: 210eeef49 e915a2639 153509Author: Peter Hutterer <peter@cs.unisa.edu.au> 153510Date: Sat Jan 26 13:55:07 2008 +1030 153511 153512 Merge branch 'master' into mpx 153513 153514 Conflicts: 153515 153516 Xext/sampleEVI.c 153517 153518commit e915a2639752bc0ea9e6e192e020cc2031c08063 153519Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153520Date: Fri Jan 25 19:22:19 2008 -0500 153521 153522 xselinux: Move the extension to extmod instead of being built-in. 153523 153524commit 210eeef495770c1883c842ff003c28ce25f279d4 153525Author: Peter Hutterer <peter@cs.unisa.edu.au> 153526Date: Sat Jan 26 09:39:54 2008 +1030 153527 153528 config: don't reset connection info on disconnect. 153529 153530 If dbus is restarted, we try to connect again and this is difficult if the 153531 busname and/or busobject is not set. 153532 153533commit f6a78ee143e3a3ad69538adf2b9675d724468ffa 153534Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153535Date: Fri Jan 25 18:04:10 2008 -0500 153536 153537 XACE: Remove the extension code entirely, XACE is completely static now. 153538 153539commit 7724c30a751c653ca3e2e8a6752af27bc37de3f0 153540Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153541Date: Fri Jan 25 17:28:17 2008 -0500 153542 153543 XACE: Stop using fake requestVectors in favor of a simple hook call. 153544 153545commit f82329b0811469ddae5c44dcfffa38185c11a67c 153546Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153547Date: Fri Jan 25 16:20:46 2008 -0500 153548 153549 XACE: Don't need to actually register a protocol extension. 153550 153551commit d23266522390a4ef7203ae7c062b2b920e45f9d7 153552Author: Peter Hutterer <peter@cs.unisa.edu.au> 153553Date: Fri Jan 25 13:54:47 2008 +1030 153554 153555 config: check connection != NULL before getting dbus' dispatch status. 153556 153557commit fb07fab2c07e7b0834724541dc47bfba02ba8574 153558Author: Peter Hutterer <peter@cs.unisa.edu.au> 153559Date: Fri Jan 25 13:45:22 2008 +1030 153560 153561 config: only shutdown libhal if the connection is valid. 153562 153563 Thanks to libdbus' extensive use of assert we won't just get an error, it'll 153564 bring the whole server down for us. 153565 153566commit 7732898aaa70e076000f6e6aa9420482896ed996 153567Author: Peter Hutterer <peter@cs.unisa.edu.au> 153568Date: Fri Jan 25 11:48:13 2008 +1030 153569 153570 config: add a debug message, fix a whitespace error. 153571 153572commit 46794d0c9665f07913980830d038c88d00407612 153573Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153574Date: Thu Jan 24 19:49:13 2008 -0500 153575 153576 xselinux: Rename SelectionManager to more generic SecurityManager. 153577 153578commit 466f6552190d1b83d66ebfd71e07921ed6210ee2 153579Author: Peter Hutterer <peter@cs.unisa.edu.au> 153580Date: Wed Jan 23 16:11:31 2008 +1030 153581 153582 Xi: if deviceid is > 127 assume GE mask is for all devices (XiSelectEvent). 153583 153584commit 379f057646588fa3211d7ba0ec8d77d2fc0ad246 153585Author: Peter Hutterer <peter@cs.unisa.edu.au> 153586Date: Wed Jan 23 16:08:37 2008 +1030 153587 153588 Xext: add GEDeviceMaskIsSet (GE mask checking per device) 153589 153590 If a mask was set for the NULL device, then we pretend the mask is set for all 153591 devices. 153592 153593commit 91dc8c43d2343e494c59023484883f4683bdc639 153594Author: Peter Hutterer <peter@cs.unisa.edu.au> 153595Date: Wed Jan 23 15:45:02 2008 +1030 153596 153597 Xi: Change window access right check to RecieveAccess in XiSelectEvent. 153598 153599commit 94a6a65ae7414c8247aa1741ec5c7561e6f4bbe0 153600Author: Peter Hutterer <peter@cs.unisa.edu.au> 153601Date: Wed Jan 23 15:43:48 2008 +1030 153602 153603 Xext: allow per-device GE masks. 153604 153605 Instead of a single mask per client we allow the storage of a mask per client 153606 per device. 153607 153608commit 6ffeecabb7f3f3173864e0f0af21a99bdc5b5044 153609Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153610Date: Thu Jan 24 18:11:49 2008 -0500 153611 153612 xselinux: Use a privileged bit in the state instead of passing an index 153613 to the permission checking function. 153614 153615commit 7ba8e97cbabfef4d614a6a38314830ec0f925471 153616Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153617Date: Thu Jan 24 19:09:58 2008 -0500 153618 153619 xselinux: Implement "get context" protocol requests. 153620 153621commit f0bf9a5231d4f612ac916355118484d055715f32 153622Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 153623Date: Thu Jan 24 19:02:35 2008 -0500 153624 153625 xselinux: Whitespace fixups. 153626 153627commit 734e115871ce98badb8800383c423493802ae3d3 153628Author: Hong Liu <hong.liu@intel.com> 153629Date: Wed Jan 23 21:04:32 2008 +0800 153630 153631 Bug #12439: add a quirk to use +hsync +vsync for the probed detailed mode. 153632 153633 Samsung 205BW quirk is somehow reworked. 153634 153635commit cc22b05ea06e08568d0f0abdaccf67bd32662e94 153636Author: Adam Jackson <ajax@redhat.com> 153637Date: Tue Jan 22 18:57:11 2008 -0500 153638 153639 There is no such thing as /dev/cpu/mtrr. 153640 153641commit 1a88aed5c82c7c131e3d473ef7b8766a418fdf1b 153642Author: David Nusinow <dnusinow@debian.org> 153643Date: Mon Jan 21 21:16:13 2008 -0500 153644 153645 Add tags/TAGS to .gitignore for ctags usage 153646 153647commit be6c17fcf9efebc0bbcc3d9a25f8c5a2450c2161 153648Author: Matthias Hopf <mhopf@suse.de> 153649Date: Mon Jan 21 16:13:21 2008 +0100 153650 153651 CVE-2007-6429: Always test for size+offset wrapping. 153652 153653commit ba315ae5ddd00763147a4b0cdcacba7691793dfc 153654Author: Peter Hutterer <peter@cs.unisa.edu.au> 153655Date: Mon Jan 21 23:43:53 2008 +1030 153656 153657 Xi: add XACE hooks for device creation (ChangeDeviceHierarchy) 153658 153659 AddInputDevice checks for permissions already, so all we do is modify a few 153660 callers to let AID sort it out. 153661 153662commit 94a21d757ce58254accbd5dd3a86810aadeec9f0 153663Author: Michel Dänzer <michel@tungstengraphics.com> 153664Date: Sat Jan 19 13:17:45 2008 +0100 153665 153666 AIGLX: Fix GLX_EXT_texture_from_pixmap fallback with EXA. 153667 153668 Use pScreen->GetImage to obtain the pixmap contents instead of dereferencing 153669 pPixmap->devPrivate.ptr directly. 153670 153671commit e9fa7c1c88a8130a48f772c92b186b8b777986b5 153672Author: Adam Jackson <ajax@redhat.com> 153673Date: Fri Jan 18 14:41:20 2008 -0500 153674 153675 CVE-2007-6429: Don't spuriously reject <8bpp shm pixmaps. 153676 153677 Move size validation after depth validation, and only validate size if 153678 the bpp of the pixmap format is > 8. If bpp < 8 then we're already 153679 protected from overflow by the width and height checks. 153680 153681commit 6492d513c0022d70f838d1b2561442c7a41e401f 153682Author: Peter Hutterer <peter@cs.unisa.edu.au> 153683Date: Fri Jan 18 21:44:22 2008 +1030 153684 153685 dix: if a client has a core grab, return the grab device on QueryPointer. 153686 153687 The correct thing would be to return the ClientPointer. However, if the client 153688 for some reason has a core grab on another device (e.g. as result of a passive 153689 grab), return the coordinates of the grabbed device instead. 153690 153691 This makes the use of nautilus a bit saner. 153692 153693commit cd2ad4c2ed58b49cee085174e77c242048159321 153694Author: Peter Hutterer <peter@cs.unisa.edu.au> 153695Date: Fri Jan 18 20:11:55 2008 +1030 153696 153697 dix: remove some superfluous lines in ProcChangeActivePointerGrab. 153698 153699commit 8a925d201efac52ef663ea3ff46acc5046178666 153700Author: Peter Hutterer <peter@cs.unisa.edu.au> 153701Date: Fri Jan 18 20:06:29 2008 +1030 153702 153703 dix: free the genericMasks when ungrabbing. 153704 153705 This finally plugs a memory hole created by grabs registering for generic 153706 events. 153707 153708commit 8949d81a3c79eef351d7c018c76692dcca74b63e 153709Author: Peter Hutterer <peter@cs.unisa.edu.au> 153710Date: Fri Jan 18 19:45:54 2008 +1030 153711 153712 include: mark CoreProcessPointerEvent, CoreProcessKeyboardEvent as deprecated. 153713 153714 ProcessOtherEvent does all the work now. 153715 153716commit d323fd64206ddb4221daa7cf7504daf62c5e7d90 153717Author: Peter Hutterer <peter@cs.unisa.edu.au> 153718Date: Fri Jan 18 19:17:42 2008 +1030 153719 153720 dix: remove inputInfo.pointer refernce in EventSelectForWindow 153721 153722 If the window being changed is set as the motion hint window for any device, 153723 the device's motion hint window is set to NULL. Which is kinda what the old 153724 code did, except that it did it with only the VCP. 153725 153726commit 883811c2bed9c6fcb706effbf83ddc97518b6f0c 153727Author: Peter Hutterer <peter@cs.unisa.edu.au> 153728Date: Fri Jan 18 18:48:46 2008 +1030 153729 153730 dix: remove inputInfo.pointer reference in TryClientEvents. 153731 153732 Unfortunately, this requires a change in the TCE API, and thus a change in all 153733 callers. Tough luck. 153734 153735commit 6e17184486c4309ec8b47c7cfd11fef60bb360ac 153736Author: Peter Hutterer <peter@cs.unisa.edu.au> 153737Date: Fri Jan 18 17:52:38 2008 +1030 153738 153739 Xi: Check for DestroyAccess when trying to delete a master device. 153740 153741commit 23f3f0e27dc90b7b3a375f2a5dd094e6f53552b5 153742Author: Jeremy Huddleston <jeremy@yuffie.local> 153743Date: Sun Jan 13 14:00:25 2008 -0800 153744 153745 XQuartz: Moved SetFrontProcess haco to set_front_process 153746 So it is done by the other thread... 153747 (cherry picked from commit 7429379eb1001ee3dc769daa8fe6b3aef1b9cc8a) 153748 153749commit 8e133d96740d010a4fd969a8188e6e71fb2cafe2 153750Author: Matthieu Herrb <matthieu@bluenote.herrb.com> 153751Date: Thu Jan 17 15:29:06 2008 +0100 153752 153753 Fix for CVE-2008-0006 - PCF Font parser buffer overflow. 153754 153755commit 6de61f82728df22ea01f9659df6581b87f33f11d 153756Author: Matthieu Herrb <matthieu@bluenote.herrb.com> 153757Date: Thu Jan 17 15:28:42 2008 +0100 153758 153759 Fix for CVE-2007-6429 - MIT-SHM and EVI extensions integer overflows. 153760 153761commit 7dc1717ff0f96b99271a912b8948dfce5164d5ad 153762Author: Matthieu Herrb <matthieu@bluenote.herrb.com> 153763Date: Thu Jan 17 15:28:03 2008 +0100 153764 153765 Fix for CVE-2007-6428 - TOG-cup extension memory corruption. 153766 153767commit dd5e0f5cd5f3a87fee86d99c073ffa7cf89b0a27 153768Author: Matthieu Herrb <matthieu@bluenote.herrb.com> 153769Date: Thu Jan 17 15:27:34 2008 +0100 153770 153771 Fix for CVE-2007-6427 - Xinput extension memory corruption. 153772 153773commit bbde5b62a137ba726a747b838d81e92d72c1b42b 153774Author: Matthieu Herrb <matthieu@bluenote.herrb.com> 153775Date: Thu Jan 17 15:26:41 2008 +0100 153776 153777 Fix for CVE-2007-5760 - XFree86 Misc extension out of bounds array index 153778 153779commit e85130c85f727466fc27be1cfa46c88b257499fb 153780Author: Daniel Stone <daniel@fooishbar.org> 153781Date: Sat Jan 5 10:47:39 2008 +0200 153782 153783 Xephyr: One-time keyboard leak fix 153784 153785 Don't leak the originally-allocated keysym map. 153786 153787commit 0137b0394a248f694448a7d97c9a1a3efcf24e81 153788Author: Daniel Stone <daniel@fooishbar.org> 153789Date: Sat Jan 5 10:43:53 2008 +0200 153790 153791 XKB: XkbCopyKeymap: Don't leak all the sections 153792 153793 Previously, we'd just keep num_sections at 0, which would break the 153794 geometry and lead us to leak sections. Don't do that. 153795 153796commit b99a43dfe97c1813e1c61f298b1c83c5d5ca88a2 153797Author: Daniel Stone <daniel@fooishbar.org> 153798Date: Sat Jan 5 10:38:16 2008 +0200 153799 153800 OS: IO: Zero out client buffers 153801 153802 For alignment reasons, we can write out uninitialised bytes, so allocate 153803 the whole thing with xcalloc. 153804 153805commit a6a7fadbb03ee99312dfb15ac478ab3c414c1c0b 153806Author: Kristian Høgsberg <krh@redhat.com> 153807Date: Wed Jan 16 20:24:11 2008 -0500 153808 153809 Don't break grab and focus state for a window when redirecting it. 153810 153811 Composite uses an unmap/map cycle to trigger backing pixmap allocation 153812 and cliprect recomputation when a window is redirected or unredirected. 153813 To avoid protocol visible side effects, map and unmap events are 153814 disabled temporarily. However, when a window is unmapped it is also 153815 removed from grabs and loses focus, but these state changes are not 153816 disabled. 153817 153818 This change supresses the unmap side effects during the composite 153819 unmap/map cycle and fixes this bug: 153820 153821 http://bugzilla.gnome.org/show_bug.cgi?id=488264 153822 153823 where compiz would cause gnome-screensaver to lose its grab when 153824 compiz unredirects the fullscreen lock window. 153825 153826commit e46f6ddeccd082b2d507a1e8b57ea30e6b0a2c83 153827Author: Michel Dänzer <michel@tungstengraphics.com> 153828Date: Wed Jan 16 14:24:22 2008 +0100 153829 153830 Yet another Xv extension byte swapping fix. 153831 153832commit b0bf4308acb706abc87c51658c2251fa86231c35 153833Author: Peter Hutterer <peter@cs.unisa.edu.au> 153834Date: Tue Jan 15 19:15:09 2008 +1030 153835 153836 dix: print out an error when core type can't be converted. 153837 153838 Helps a bit in tracking down bugs. 153839 153840commit 2a988ed75b7634d7cb6e83bb7aa89bc8768d3a58 153841Author: Peter Hutterer <peter@cs.unisa.edu.au> 153842Date: Tue Jan 15 19:14:46 2008 +1030 153843 153844 xkb: don't do core key repeats, XI only is the path to light. 153845 153846commit 4fd0885978be286a68b689824f1d910f929a52b0 153847Author: Peter Hutterer <peter@cs.unisa.edu.au> 153848Date: Tue Jan 15 18:58:42 2008 +1030 153849 153850 Xi: remove some leftovers from ProcessOtherEvents. 153851 153852 Core handling not needed anymore here. 153853 153854commit 32aa252e988be8cbfd4f7e373fb7b7736ef1f5f2 153855Author: Peter Hutterer <peter@cs.unisa.edu.au> 153856Date: Tue Jan 15 15:22:39 2008 +1030 153857 153858 dix: Process an input event as a single event, instead of two separate ones. 153859 153860 This is a significant shift in how input events are perceived. The common 153861 approach was to treat a core event as a different entity than the XI event. 153862 This could result in the XI event being delivered to a different client than 153863 the core event. This doesn't work nicely if they come from the same device. 153864 153865 Instead, we treat an input event as a single event, that is delivered through 153866 two separate APIs. So when delivering an event, we first try the XI event, 153867 then the core event. If the window want's neither, we go to the parent and 153868 repeat. Once either core or XI has been delivered, the processing stops. 153869 153870 Important: Different to the previous method, if a client registers for core 153871 button events, the parent window will not get XI events. This should only 153872 cause problems when you're mixing core and XI events, so don't do that! 153873 153874 Generic events don't fit into this yet, they cause me headaches. 153875 153876commit 7a0d16ef0a103bcb25fa8a20322685f017aaf5a3 153877Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 153878Date: Tue Jan 15 03:27:16 2008 -0200 153879 153880 Removed some warnings. 153881 153882commit 315d6a2b1d2a3de308e98d548afe780c59a784fc 153883Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 153884Date: Tue Jan 15 02:59:56 2008 -0200 153885 153886 Fix Xephyr compilation without GLX. 153887 153888commit 0969a9f7497e10794a6534321c10a0e1ac680ad7 153889Author: Peter Hutterer <peter@cs.unisa.edu.au> 153890Date: Tue Jan 15 11:31:12 2008 +1030 153891 153892 dix: Emulate core events within CheckDeviceGrabs and ComputeFreezes. 153893 153894 This should restore the correct passive grab processing. When checking for 153895 passive grabs, the core event is emulated and we check first for XI grabs on 153896 the window, then for core grabs. Regardless of which event activates the grab, 153897 the XI event is stored in the device's EQ. 153898 153899 When replaying the event, we take the XI event and replay it on the next 153900 window, again including the emulation of the core event. 153901 153902commit 1f83f40525acd3aff8f50b3c519bc1f307ff1e19 153903Author: Dave Airlie <airlied@linux.ie> 153904Date: Tue Jan 15 10:20:50 2008 +1000 153905 153906 xf86Cursors: fix memset for non-square cursors 153907 153908commit e6ea3147bfb686798dac381eb8900f9f18beb88e 153909Author: Bernardo Innocenti <bernie@codewiz.org> 153910Date: Sun Jan 13 19:50:37 2008 -0500 153911 153912 exa: make the prototype for exaGetPixmapFirstPixel() public 153913 153914 This fixes a warning in amd_drv which is using it. 153915 153916 Signed-off-by: Bernardo Innocenti <bernie@codewiz.org> 153917 153918commit a83a0c5a144da67dab96a857b849a5692b73245d 153919Author: Peter Hutterer <peter@cs.unisa.edu.au> 153920Date: Sun Jan 13 16:14:29 2008 +1030 153921 153922 dix: remove obsolete comment. 153923 153924 Pairings don't exist anymore and the documented issue is a non-issue now. 153925 153926commit 180a5aba4de3104fed8bc4e7d42a1e3a51575318 153927Author: Jeremy Huddleston <jeremy@yuffie.local> 153928Date: Sat Jan 12 21:24:34 2008 -0800 153929 153930 XQuartz: Fixed copy-paste error with login_shell commit 153931 (cherry picked from commit 6deec3acc6f8010b5b53a1e55a0a2c4080ba69d2) 153932 153933commit 6fd4a5e2e4d0be0ba0773df831687e11e1262c72 153934Author: Jeremy Huddleston <jeremy@yuffie.local> 153935Date: Sat Jan 12 11:56:00 2008 -0800 153936 153937 XQuartz: Corrected copyright X.org Project -> X.org Foundation 153938 (cherry picked from commit f21631444816fc12b8a534c2cf79e6ac6c2af7c9) 153939 153940commit f72255639c065d795f7767683e851b1b5b2d9480 153941Author: Jeremy Huddleston <jeremy@yuffie.local> 153942Date: Sat Jan 12 11:35:48 2008 -0800 153943 153944 XQuartz: added 'login_shell' option to defaults 153945 so the user can choose something other than /bin/sh 153946 (cherry picked from commit b549cf18cebd3435d70f62855239484974c455a1) 153947 153948commit ec24a6b5aa732ec6999a27889d9a33cf80123886 153949Author: Jeremy Huddleston <jeremy@yuffie.local> 153950Date: Sun Jan 6 18:29:54 2008 -0800 153951 153952 XQuartz: Fixed switching into XQuartz via expose. 153953 (cherry picked from commit 627ed60ce5d7499761028edf379ebd95250d3e04) 153954 153955commit 6844bd2e63490870bab3c469eec6030354ef2865 153956Author: Alan Coopersmith <alan.coopersmith@sun.com> 153957Date: Wed Jan 9 19:52:00 2008 -0800 153958 153959 More Xv extension byte swapping fixes 153960 153961commit 38bf01bd1c925614a6e67a38aa3cefc7b8fe3bca 153962Merge: 4e85c7c32 0883e838e 153963Author: Peter Hutterer <peter@cs.unisa.edu.au> 153964Date: Thu Jan 10 13:38:46 2008 +1030 153965 153966 Merge branch 'master' into mpx 153967 153968commit 0883e838e25227f0af84d2a90979175724166d16 153969Author: Peter Harris <peter.harris@hummingbird.com> 153970Date: Wed Jan 9 14:52:33 2008 -0500 153971 153972 xf86misc.c: Avoid use of swapped values 153973 153974commit 59a63d72a1407a8aaf9878eeff7ee7a66f65a42b 153975Author: Alan Coopersmith <alan.coopersmith@sun.com> 153976Date: Wed Jan 9 18:42:58 2008 -0800 153977 153978 Fix reference to old versions of XFree86 to not say "Xorg 4.2 or earlier" 153979 153980commit 4e85c7c322faf14c14e4229fa294b8e3d3a4d304 153981Author: Peter Hutterer <peter@cs.unisa.edu.au> 153982Date: Wed Jan 9 17:36:39 2008 +1030 153983 153984 Xi: keep a counter of buttons down to avoid duplicate press/release events. 153985 153986 If two devices are attached to the same master device, pressing button 1 on 153987 each of them leads to two button presses from the same device. Some apps 153988 really don't like that. 153989 153990 So we just put a counter in place and only send the first press and the last 153991 release. 153992 153993commit e070c2cbef0d6fbbafce8d417e8b29231c2fdc50 153994Author: Peter Hutterer <peter@cs.unisa.edu.au> 153995Date: Fri Dec 7 20:18:49 2007 +1030 153996 153997 mi: change infamous "Tossed event ..." error for something more explanatory. 153998 153999 Few whitespace errors fixed. 154000 Two ErrorF's prefixed with [mi]. 154001 (cherry picked from commit 117458d2db49efd3f04432ff45871b44c7d4ad57) 154002 154003commit c31aead0fe9fe424120ce221cd9f76cb6d29a5b5 154004Author: Richard Purdie <rpurdie@openedhand.com> 154005Date: Tue Jan 8 12:12:06 2008 +0100 154006 154007 [Kdrive] make XCalibrate 'orientation aware' 154008 154009commit 30375cd6d1439a3390b41714fe116aecc94743ca 154010Author: David Nusinow <dnusinow@debian.org> 154011Date: Mon Jan 7 20:57:30 2008 -0500 154012 154013 Don't log DPMS enabling as being from xorg.conf if it's not 154014 154015commit 981bb9f1e33e4564d1b59c00c808cc43a2e9497b 154016Author: Peter Hutterer <peter@cs.unisa.edu.au> 154017Date: Tue Dec 18 13:57:07 2007 +1030 154018 154019 dix: set the correct number of valuators in valuator events. 154020 154021 (first_valuator + num_valuators) must never be larger than the number of axes, 154022 otherwise DIX freaks out. And from looking at libXI, anything larger than 6 is 154023 wrong too. 154024 (cherry picked from commit 9f6ae61ad12cc2813d04405458e1ca5aed8a539e) 154025 154026commit 20ace6321ac464d821c67a82c7023f74ae038176 154027Author: Peter Hutterer <peter@cs.unisa.edu.au> 154028Date: Mon Jan 7 20:08:56 2008 +1030 154029 154030 dix: don't try to confine cursor on screen restructure if there is no cursor. 154031 154032 Stops segfaulting when using xrandr. Yay. 154033 154034commit 57a491e6d3ebec464c0c861b02c9ddcb1cc62c3c 154035Author: Peter Hutterer <peter@cs.unisa.edu.au> 154036Date: Mon Jan 7 20:06:32 2008 +1030 154037 154038 dix: init GrabPrivateKey to correct value. 154039 154040 Merge detritus from last pull. 154041 154042commit 260505e3c5a18044e97d31ea3bcc0955e46335c8 154043Author: David Nusinow <dnusinow@debian.org> 154044Date: Sun Jan 6 16:37:13 2008 -0500 154045 154046 Log enabling of DPMS even when it's not set in xorg.conf 154047 154048commit 7e7622165940934e56ae96ae785a8f88eec1a5cf 154049Author: Julien Cristau <jcristau@debian.org> 154050Date: Sun Jan 6 18:23:09 2008 +0100 154051 154052 Fix the name of the XFree86-Misc extension in the xorg.conf manpage. 154053 154054commit 59df687835c68eda147de47edfe9bc415c0efb4f 154055Author: Julien Cristau <jcristau@debian.org> 154056Date: Sun Jan 6 16:57:45 2008 +0100 154057 154058 Document the AllowEmptyInput, AutoAddDevices and AutoEnableDevices flags 154059 154060 Add documentation for the new AllowEmptyInput, AutoAddDevices and 154061 AutoEnableDevices server flags in the xorg.conf manpage. 154062 154063commit 7d226d6a251cb90765be2b50a1973986c5b7605b 154064Author: Jeremy Huddleston <jeremy@yuffie.local> 154065Date: Sat Jan 5 03:14:07 2008 -0800 154066 154067 XQuartz: Cleanup for strict-prototyping 154068 Also fixed DarwinEQEnqueue to match changes to the callback 154069 And also use dpmsstubs.c rather than copying the code into darwin.c 154070 (cherry picked from commit 4c5c30a4beb7a427b00b18097f548876ad3c11d7) 154071 154072commit 11967dce11cd953d123d53bb3389aa257c5158e8 154073Author: Jeremy Huddleston <jeremy@yuffie.local> 154074Date: Fri Jan 4 22:54:26 2008 -0800 154075 154076 XQuartz: Fixed copyright in About box for 2008. Happy New Year! 154077 154078commit aca75f389e2b08096c3cacec03b12a58075cf12c 154079Author: Jeremy Huddleston <jeremy@yuffie.local> 154080Date: Fri Jan 4 12:23:09 2008 -0800 154081 154082 XQuartz: Flush the debug log for easier debugging 154083 Also cleaned up formatting in xpr's eventHandler 154084 (cherry picked from commit 16861d6d4239c7f3918332ef07752f1e211afb23) 154085 154086commit 57468a696e1259c1e1c185fc60230e1d195defb7 154087Author: Alan Hourihane <alanh@tungstengraphics.com> 154088Date: Fri Jan 4 12:37:55 2008 +0000 154089 154090 Fix Line drawing with CapNotLast set in PolySegment. 154091 154092commit 39cb782f28be4efb2621fd8c614f2367eb834412 154093Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154094Date: Thu Jan 3 23:16:06 2008 -0500 154095 154096 XACE: DeleteCallbackList zeroes out its argument so don't do it twice. 154097 154098commit 7f376f23db463a65176de632ca6094acb55db951 154099Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154100Date: Thu Jan 3 23:08:49 2008 -0500 154101 154102 devPrivates rework: Free callback lists after use. 154103 154104commit de18703d2a25999e391d11b4c82ee018fb87372d 154105Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154106Date: Thu Jan 3 23:07:24 2008 -0500 154107 154108 dix: Fix bug+leak in callback manager DeleteCallbackList function. 154109 154110commit 20eb26f9d149993ae360a2cbd1b536b68c9f4069 154111Author: Alex Deucher <alex@samba.(none)> 154112Date: Thu Jan 3 22:53:36 2008 -0500 154113 154114 Fix potential crasher in xf86CrtcRotate() 154115 154116 xf86CrtcRotate() is called by randr 1.2 drivers via xf86CrtcSetMode() or xf86SetDesiredModes() 154117 during ScreenInit() at which point pScrn->pScreen is not set. If a user specifies a rotation 154118 in their config file pScrn->pScreen is dereferenced and boom. 154119 154120commit 249b9b30c11e03bb113b37cf4712bf88debf6f05 154121Author: Peter Hutterer <peter@cs.unisa.edu.au> 154122Date: Fri Jan 4 13:49:48 2008 +1030 154123 154124 Xi: remove file for the obsolete ChangePointerKeyboardPairing request handling. 154125 154126commit e4a214e40d35aa957c83b86b4e4abc86c22fbde0 154127Author: Peter Hutterer <peter@cs.unisa.edu.au> 154128Date: Fri Jan 4 13:32:53 2008 +1030 154129 154130 dix: don't free device's devPrivates manually, dixFreePrivates does it. 154131 154132 Merge detritus from last pull. 154133 154134commit cd0603c2dc5ee000ebce66056bc1a72f99bfb617 154135Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154136Date: Thu Jan 3 21:41:02 2008 -0500 154137 154138 Bug #13765: Heap corruption in XC-SECURITY extension code. 154139 154140commit e46d559739e020dc7f6fcbdc6d1fb39c57aab4b1 154141Author: Jeremy Huddleston <jeremy@yuffie.local> 154142Date: Sat Dec 22 15:09:12 2007 -0800 154143 154144 XQuartz: Cleaned up color map configuration. 154145 8 bit color still doesn't work, but the -depth command line argument now works properly. 154146 (cherry picked from commit 6765949c27c053d22882f54337cfd09203aa5383) 154147 154148commit 17a9714a6789a389d52dbb40fd1eed1e24c04d64 154149Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154150Date: Thu Jan 3 14:46:54 2008 -0500 154151 154152 Bug #13794: Update MBE extension devPrivates to new interface. 154153 154154commit b2da44c76d68a76f20e90bccb268ebf65e132b49 154155Author: Peter Hutterer <peter@cs.unisa.edu.au> 154156Date: Thu Jan 3 18:51:44 2008 +1030 154157 154158 dix: DoChangeKeyboardControl shouldn't be using inputInfo.keyboard. 154159 154160commit 9cc41b88dfebb3f1e0a7cb5c224146e651a2cda4 154161Author: Peter Hutterer <peter@cs.unisa.edu.au> 154162Date: Thu Jan 3 18:47:28 2008 +1030 154163 154164 Xi: ProcXSetClientPointer should directly check for the client from the id. 154165 154166 The window ID includes the client ID anyway, so we might as well just look up 154167 the client directly instead of trying to get the window first and the client 154168 from the window. 154169 154170 This also fixes a possible issue with XACE. If the client had permission to 154171 write on the client but not on the window, the previous approach would have 154172 failed. 154173 154174commit 521fdc28d86d091495da3558d26ab4a938250ffe 154175Author: Peter Hutterer <peter@cs.unisa.edu.au> 154176Date: Thu Jan 3 18:43:18 2008 +1030 154177 154178 xfixes: switch a few inputInfo.pointer to PickPointer 154179 154180commit 37194b13554f0d36343cf73324b128b15bb6a338 154181Author: Peter Hutterer <peter@cs.unisa.edu.au> 154182Date: Thu Jan 3 18:36:33 2008 +1030 154183 154184 Xi: switch to using "rc" (instead of "err") as returncode. 154185 154186 All the rest of XI uses rc and returns rc in case of error, so make 154187 mpx-related stuff comply. This stops the rest of XI sending the error 154188 manually. 154189 This is just a cosmetic change to be in line with the rest. 154190 154191commit c7e9b67c547c97b2c943eab555b4fe2844e8a91f 154192Author: Peter Hutterer <peter@cs.unisa.edu.au> 154193Date: Thu Jan 3 18:09:56 2008 +1030 154194 154195 dix: free ClassesRec when a BadAlloc occurs during CoreKeyboardProc. 154196 154197commit 8da83836b60f7cdb75d08482f4311fa0e2ab4e1d 154198Merge: eace88989 ae869fc76 154199Author: Peter Hutterer <peter@cs.unisa.edu.au> 154200Date: Thu Jan 3 17:04:54 2008 +1030 154201 154202 Merge branch 'master' into mpx 154203 154204 Conflicts: 154205 154206 XTrap/xtrapddmi.c 154207 Xext/security.c 154208 Xext/xprint.c 154209 Xext/xtest.c 154210 Xext/xvdisp.c 154211 Xi/exevents.c 154212 Xi/grabdevb.c 154213 Xi/grabdevk.c 154214 Xi/opendev.c 154215 Xi/ungrdev.c 154216 Xi/ungrdevb.c 154217 Xi/ungrdevk.c 154218 dix/cursor.c 154219 dix/devices.c 154220 dix/dixutils.c 154221 dix/events.c 154222 dix/getevents.c 154223 dix/main.c 154224 dix/window.c 154225 hw/xfree86/ramdac/xf86Cursor.c 154226 include/dix.h 154227 include/input.h 154228 include/inputstr.h 154229 mi/midispcur.c 154230 mi/miinitext.c 154231 mi/misprite.c 154232 render/animcur.c 154233 xfixes/cursor.c 154234 xkb/xkbAccessX.c 154235 154236commit ccf6636d2ca8acdaaeb8da34db507a10a082b0de 154237Author: Alan Coopersmith <alan.coopersmith@sun.com> 154238Date: Wed Jan 2 19:28:33 2008 -0800 154239 154240 Kill xf86getpagesize even harder (dummylib & ioport) 154241 154242commit dfd682b582636a36345144bcf835e3ee46718d90 154243Author: Alan Coopersmith <alan.coopersmith@sun.com> 154244Date: Wed Jan 2 19:27:22 2008 -0800 154245 154246 X.Org bug 4947/Sun bug 6646626: Xv extension not byte-swapping properly 154247 154248 X.Org Bugzilla #4947 <https://bugs.freedesktop.org/show_bug.cgi?id=4947> 154249 Sun bug 6646626 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6646626> 154250 154251 Don't use swapped data after swapping it. When done swapping data, 154252 send the swapped data, not the address of the pointer to it, to the client. 154253 154254commit f6666dcc3b1ac60f850ea53c357a9ef61672a52a 154255Author: Alan Coopersmith <alan.coopersmith@sun.com> 154256Date: Wed Jan 2 19:19:55 2008 -0800 154257 154258 Add dixAllocatePrivate stub to dummylib for utils 154259 154260 Normally not necessary, except when building non-optimized/debug causes 154261 the inline functions from private.h to appear in os-support/libxorgos.la 154262 154263commit 73f422996016107d5f53492e4197bb05ed9c4bb9 154264Author: Alan Coopersmith <alan.coopersmith@sun.com> 154265Date: Wed Jan 2 19:17:54 2008 -0800 154266 154267 Fix names/types of new vuidMouse{Get,Set}ScreenPrivates macros 154268 154269commit 895073f6b41d9313cfe748232c492c5e9f76b443 154270Author: Alan Coopersmith <alan.coopersmith@sun.com> 154271Date: Wed Jan 2 18:09:26 2008 -0800 154272 154273 Restore include & typedef needed by dtrace 154274 154275 996b621bec1bbc4fb21970c75eaec62053bc6ccb deleted a couple lines too many 154276 154277commit 306fde4082044dfecbedd9af41e660bafb3ce438 154278Author: Brian <brian@i915.localnet.net> 154279Date: Tue Jan 1 09:27:44 2008 -0700 154280 154281 regenerated to add framebuffer object tokens (bug 13800) 154282 154283commit 5c362c2eb2cfdf1f6d667a3e64a0a7bc4942c950 154284Author: Brian <brian@i915.localnet.net> 154285Date: Tue Jan 1 09:07:48 2008 -0700 154286 154287 regenerated, adds GL_MAX_3D_TEXTURE_SIZE (see bug 13811) 154288 154289commit 85365ddf16e2b954d8249b380df53337420ed684 154290Author: Adam Jackson <ajax@redhat.com> 154291Date: Fri Dec 28 17:35:54 2007 -0500 154292 154293 EDID 1.4: If given a native pixel format, use it when inferring virtual. 154294 154295commit 312b30cb03e439644ea10e08fa93268116333f0d 154296Author: Adam Jackson <ajax@redhat.com> 154297Date: Fri Dec 28 17:18:30 2007 -0500 154298 154299 EDID 1.4: First detailed mode is always preferred. 154300 154301 ... so act like it in the modelist generator, not just the parser. 154302 154303commit 9dbb73033ae60e7ab85f1469a696e2a52f0cb0fe 154304Author: Colin Harrison <colin.harrison@virgin.net> 154305Date: Sat Dec 29 00:02:16 2007 +0200 154306 154307 Rootless: Fix lvalue error from devPrivates change 154308 154309 Instead of trying to use an invalid expression as an lvalue, use the function 154310 call instead. 154311 154312commit 36ff05771b0699aa80ad718e24097bc25cb4fb00 154313Author: Jurij Smakov <jurij@wooyd.org> 154314Date: Fri Dec 28 23:59:53 2007 +0200 154315 154316 GL: Add GLX compile flags lost in modular X server changes 154317 154318 RISC chips that trap on unaligned loads and stores need to 154319 define __GLX_ALIGN64. This used to get added to the cflags 154320 in the old *.cf files but it no longer does in the modular 154321 X server. 154322 154323 Also, Alpha needs to pass -mieee to the compiler as well. 154324 154325 This is a simple backport of a patch that debian, and probably other 154326 distributions, have been applying forever. To the best of my 154327 knowledge the patch was written by Jurij Smakov. See Debian bug 154328 number #388125. 154329 154330 I just checked and this has been rotting for more than a year in 154331 freedesktop bugzilla as #8392. 154332 154333 Signed-off-by: David S. Miller <davem@davemloft.net> 154334 154335commit bae459cfc4f17a5ec5f2810e9f913e3ad2d8b8d4 154336Author: Adam Jackson <ajax@redhat.com> 154337Date: Fri Dec 28 16:50:18 2007 -0500 154338 154339 Don't carp on EDID 1.4 blocks anymore. 154340 154341 Also whine more loudly when we get something other than 1.x. 154342 154343commit 70b2d6cfeb3bcb7b862a2ae29f6ef7cb84d69486 154344Author: Adam Jackson <ajax@redhat.com> 154345Date: Fri Dec 28 16:39:00 2007 -0500 154346 154347 Check the gamma value, not its address. 154348 154349commit 592d814ee09e86e283116a7a1052762c8398e8e5 154350Author: Adam Jackson <ajax@redhat.com> 154351Date: Fri Dec 28 16:37:23 2007 -0500 154352 154353 EDID 1.4: Additional semantics for display feature bits. 154354 154355 First mode is _always_ preferred in 1.4; the bit that used to mean this 154356 now means that the preferred mode is also the native pixel format. The 154357 old "is GTF" bit now means "is continuous-frequency" instead. 154358 154359 Section 3.6.4, Table 3.14: Feature Support, Notes 4 and 5. 154360 154361commit 322d0103aee317500057c80d542d7270d69a5731 154362Author: Adam Jackson <ajax@redhat.com> 154363Date: Fri Dec 28 16:28:44 2007 -0500 154364 154365 EDID 1.4: Alternate color encodings for digital inputs. 154366 154367 Section 3.6.4, Table 3.14: Feature support. 154368 154369commit f1f43caf7e26a84dbacd4e5d7d47c8b4e4982836 154370Author: Adam Jackson <ajax@redhat.com> 154371Date: Fri Dec 28 16:12:11 2007 -0500 154372 154373 EDID 1.4: Allow for gamma definition in extension blocks. 154374 154375 Section 3.6.3, Table 3.13: Display Transfer Characteristics (Gamma) 154376 154377commit 861ee38817523a647e6be10d7e8fe26f66054217 154378Author: Adam Jackson <ajax@redhat.com> 154379Date: Fri Dec 28 16:06:45 2007 -0500 154380 154381 EDID 1.4: Additional aspect ratio semantics for screen size fields. 154382 154383 Section 3.6.2, Table 3.12: Horizontal and Vertical Screen Size or 154384 Aspect Ratio. 154385 154386commit 14b5c8a447db0395fb14b2d404eafb1d8e4fb817 154387Author: Adam Jackson <ajax@redhat.com> 154388Date: Fri Dec 28 15:52:42 2007 -0500 154389 154390 EDID 1.4: Extended support for digital interfaces. 154391 154392 Section 3.6.1, Table 3.11: Video Input Definition. 154393 154394commit a948216dccb5ee577a50a42035dc9bc49d0a00c6 154395Author: Adam Jackson <ajax@redhat.com> 154396Date: Fri Dec 28 15:00:41 2007 -0500 154397 154398 EDID 1.4: Decode CVT 3-byte codes and add them to the mode pool. 154399 154400 Section 3.10.3.8: CVT 3 Byte Code Descriptor Definition. 154401 154402commit f6df66cc89bcd0a0be2e7bca05839fdd428c1d4c 154403Author: Adam Jackson <ajax@redhat.com> 154404Date: Fri Dec 28 13:55:39 2007 -0500 154405 154406 EDID 1.4: Trivial support for new detailed sections. 154407 154408 Nothing actually decoded yet, but at least we print what they are. 154409 154410 New in EDID 1.4: 154411 - Color Management Data (0xF9), Section 3.10.3.7 154412 - CVT 3 Byte Code Descriptor (0xF8), Section 3.10.3.8 154413 - Established Timings III Descriptor (0xF7), section 3.10.3.9 154414 - Manufacturer-specified data tag (0x00 - 0x0F), section 3.10.3.12 154415 154416commit bac3ecde39cc914ab515991234b7dc2138005b84 154417Author: Adam Jackson <ajax@redhat.com> 154418Date: Fri Dec 28 13:33:39 2007 -0500 154419 154420 EDID 1.4: Allow for sync range offsets. 154421 154422 Table 3.26: Display Range Limits & Timing Descriptor Block Definition 154423 154424commit 3b23dd9fd43a28033d0af7b02088b0c6ca433158 154425Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154426Date: Fri Dec 28 13:29:45 2007 -0500 154427 154428 xselinux: Fix whitespace warnings. 154429 154430commit 643c52be32c187a0fdb9a031b1e31d97cd551339 154431Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154432Date: Fri Dec 28 13:26:26 2007 -0500 154433 154434 xselinux: Remove "X" prefix on remaining functions and strings. 154435 154436 Should be evident from the context. 154437 154438commit f4bc333fc1e8d9fa9911771d2072df4df741c553 154439Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154440Date: Fri Dec 28 11:56:54 2007 -0500 154441 154442 xselinux: don't FatalError on an invalid class mapping, just disable support. 154443 154444commit f3780ece528ed3ead809ba6a388fa0f8aab2a775 154445Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154446Date: Thu Dec 20 20:32:07 2007 -0500 154447 154448 xselinux: Implement swapped protocol request logic. 154449 154450commit 1bbf64ab115e2a1121d6f9c0830b1b977f025178 154451Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154452Date: Thu Dec 20 20:31:34 2007 -0500 154453 154454 xselinux: Remove unnecessary structure definition. 154455 154456commit 938da5ee389975f910721f1c2cebc2dcec793117 154457Author: Adam Jackson <ajax@redhat.com> 154458Date: Fri Dec 28 12:23:44 2007 -0500 154459 154460 Add HDMI and DisplayPort connector types. 154461 154462commit 24105cf6582201a94bc39aeac5a795297018aeb5 154463Author: Daniel Stone <daniel@fooishbar.org> 154464Date: Fri Dec 28 15:49:50 2007 +0200 154465 154466 Input: Don't reinit devices 154467 154468 If a device is already initialised (i.e. the virtual core devices) during 154469 IASD, don't init them again. This fixes a leak. 154470 154471commit 941058f8da0d725f909dc97f68c32ce244a9dc0a 154472Author: Daniel Stone <daniel@fooishbar.org> 154473Date: Fri Dec 28 15:48:57 2007 +0200 154474 154475 KDrive: Xephyr: Don't leak screen damage structure 154476 154477commit b2f6cd290c43b88f0d08fb29f8657618a067d2a0 154478Author: Daniel Stone <daniel@fooishbar.org> 154479Date: Fri Dec 28 15:48:25 2007 +0200 154480 154481 OS: Don't leak connection translation table on regeneration 154482 154483commit 190a0506243b39cd8dfc0e12068e3a3f416330f1 154484Author: Daniel Stone <daniel@fooishbar.org> 154485Date: Fri Dec 28 15:47:57 2007 +0200 154486 154487 Config: HAL: Don't leak options on failure to add device 154488 154489 This showed up in Xephyr in particular, which denies new device requests. 154490 154491commit f44fd3f9e41bf467360ace93ef5b532d8f61fb2c 154492Author: Daniel Stone <daniel@fooishbar.org> 154493Date: Fri Dec 28 15:47:21 2007 +0200 154494 154495 Config: D-Bus: Don't leak timers 154496 154497 TimerCancel doesn't free the timer: you need TimerFree for that. 154498 154499commit ae869fc7669764729e13fdd70149ed636753f2a3 154500Author: David S. Miller <davem@davemloft.net> 154501Date: Tue Dec 25 22:42:50 2007 -0800 154502 154503 [SBUS]: Fix build, use getpagesize() instead of xf86getpagesize(). 154504 154505 xf86getpagesize() was removed, but this one call site was 154506 not fixed up. 154507 154508 Signed-off-by: David S. Miller <davem@davemloft.net> 154509 154510commit 009f1e4e55200425de2fe0dbc1f0ac0f431fb4cf 154511Author: Fatih Aşıcı <fatih@pardus.org.tr> 154512Date: Tue Dec 25 23:09:49 2007 +0200 154513 154514 Config: Don't forget to add xkb_rules option 154515 154516commit 389e8917f66a489455f1d5c70f44c262717538ad 154517Author: Fatih Aşıcı <fatih@pardus.org.tr> 154518Date: Tue Dec 25 22:59:24 2007 +0200 154519 154520 Config: Fix a memory leak 154521 154522commit 743008a4812d6b046211ebcf4eab202687b458d5 154523Author: Adam Jackson <ajax@redhat.com> 154524Date: Sun Dec 23 14:27:14 2007 -0500 154525 154526 Report serverClient resources in the X-Resource extension. 154527 154528commit beb29c605b8c66e1a18b89668aa421c1519645f6 154529Author: Jeremy Huddleston <jeremy@yuffie.local> 154530Date: Fri Dec 21 02:09:01 2007 -0800 154531 154532 XQuartz: *REALLY* use CFStringCreateWithCString 154533 I need sleep! Why am I making these stupid mistakes... sorry for pointless commit spam. ugg. 154534 (cherry picked from commit b16351fc6457aabead328472d16dc25789032940) 154535 154536commit 5dd895efa305954e2695aa22a9e49acfb65b4d5e 154537Author: Jeremy Huddleston <jeremy@yuffie.local> 154538Date: Fri Dec 21 02:06:47 2007 -0800 154539 154540 XQuartz: Use CFStringCreateWithCString 154541 (cherry picked from commit 79782b0e14761dcf5d6635b8eec161b74f06763a) 154542 154543commit 2c24231fc2027cf5034bb1b6636332687f586726 154544Author: Jeremy Huddleston <jeremy@yuffie.local> 154545Date: Fri Dec 21 01:57:43 2007 -0800 154546 154547 XQuartz: Reduce code duplication in X11.app 154548 (cherry picked from commit b81809cd91a9f90b7f2de77b1dcf514cee87c32d) 154549 154550commit f3042a63be0748bb60567144276d2c61b75ba0b7 154551Author: Jeremy Huddleston <jeremy@yuffie.local> 154552Date: Fri Dec 21 01:24:06 2007 -0800 154553 154554 XQuartz: Handle Pseudorami init in miinitext 154555 (cherry picked from commit a585c94fedd4ecbc87524703c01bb128fc2aa951) 154556 154557commit fa9680a7305d7f906da1bdeb40a0863ef66316e6 154558Author: Jeremy Huddleston <jeremy@yuffie.local> 154559Date: Thu Dec 20 19:38:20 2007 -0800 154560 154561 XQuartz: Added localization. 154562 (cherry picked from commit 7a5cc7bfbb296a2c41a580b063324c448f7131db) 154563 154564commit 603a8b73d46d59e5f9f0be39be8317f3fadfe7e6 154565Author: Jeremy Huddleston <jeremy@yuffie.local> 154566Date: Thu Dec 20 18:29:57 2007 -0800 154567 154568 XQuartz: Cleaned up command line arguments. 154569 154570commit 4cf3002b6020024f2fc2ed0cc40a872a066e482d 154571Author: Jeremy Huddleston <jeremy@yuffie.local> 154572Date: Thu Dec 20 18:08:40 2007 -0800 154573 154574 XQuartz: Kill off server-main.c and launcher-main.c 154575 Now using xinit 154576 (cherry picked from commit 5d6ae3d299f72df714117948b3d31dcbddf6c0bc) 154577 154578commit 1f74bef1ad1399323fc0d2e309b808bf32c622e4 154579Author: Jeremy Huddleston <jeremy@yuffie.local> 154580Date: Thu Dec 20 17:33:38 2007 -0800 154581 154582 XQuartz: Don't fork to exec app_to_run. 154583 Plus other housecleaning... 154584 (cherry picked from commit ae302db472f127be082d05b418ede332fae8ccc5) 154585 154586commit 2d15d439f844d4016f169664a338595c11b91b77 154587Author: Jeremy Huddleston <jeremy@yuffie.local> 154588Date: Thu Dec 20 15:46:40 2007 -0800 154589 154590 Xquartz: Use X11ControllerMain() 154591 (cherry picked from commit a9ac932543374aa2540f5a12cc85ef82c85b0e0c) 154592 154593commit 1393a97ea97b5f7d7b90e3e8c58b5996b600e0c6 154594Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154595Date: Thu Dec 20 16:23:35 2007 -0500 154596 154597 xselinux: Send AVC messages to audit system instead of log file/stderr. 154598 154599commit 42802a8e6b3d3795acc4f8b7597ea5a48619b5cd 154600Author: Jeremy Huddleston <jeremy@yuffie.local> 154601Date: Thu Dec 20 13:17:30 2007 -0800 154602 154603 Xquartz: General Cleanup 154604 General code cleanup, whitespace, dead code removal, added missing prototypes. 154605 Made Xquartz come to foreground later in startup, so it doesn't appear for Xquartz -version 154606 (cherry picked from commit 36922e8ff4316c93843aa3fe959cf8df3c7d5892) 154607 154608commit 7ef7727b800fa4715b80a82850d65b88fde5fe6c 154609Author: Dave Airlie <airlied@clockmaker.usersys.redhat.com> 154610Date: Thu Dec 20 10:11:26 2007 +1000 154611 154612 entity sharing: make !shareable entity non-fatal. 154613 154614 Just because the entity isn't shareable, we should bring down the server. 154615 Just ignore the extra screen and keep going. 154616 154617commit 50e80c39870adfdc84fdbc00dddf1362117ad443 154618Author: Peter Hutterer <peter@cs.unisa.edu.au> 154619Date: Wed Dec 19 16:20:36 2007 +1030 154620 154621 include: never overwrite realInputProc with enqueueInputProc. Bug #13511 154622 154623 In some cases (triggered by a key repeat during a sync grab) XKB unwrapping 154624 can overwrite the device's realInputProc with the enqueueInputProc. When the 154625 grab is released and the events are replayed, we end up in an infinite loop. 154626 Each event is replayed and in replaying pushed to the end of the queue again. 154627 154628 This fix is a hack only. It ensures that the realInputProc is never 154629 overwritten with the enqueueInputProc. 154630 154631 This fixes Bug #13511 (https://bugs.freedesktop.org/show_bug.cgi?id=13511) 154632 (cherry picked from commit eace88989c3b65d5c20e9f37ea9b23c7c8e19335) 154633 154634commit d0308b64655360517d83e07e866d103c3f2b389d 154635Author: Peter Hutterer <peter@cs.unisa.edu.au> 154636Date: Wed Dec 19 18:18:10 2007 +1030 154637 154638 Xi: specify correct struct when calculating size of GetDeviceControl reply. 154639 154640 This doesn't change much, as the struct previously given has the same size as 154641 the ones now anyway. Still, we should be pendantic. 154642 154643 Thanks to Simon Thum for reporting. 154644 154645commit 66b00029e587cec628d0041179a301e888277f8e 154646Author: Peter Hutterer <peter@cs.unisa.edu.au> 154647Date: Wed Dec 19 18:10:50 2007 +1030 154648 154649 Xext: remove redefinition of Bool. 154650 154651 Thanks to Simon Thum. 154652 154653commit eace88989c3b65d5c20e9f37ea9b23c7c8e19335 154654Author: Peter Hutterer <peter@cs.unisa.edu.au> 154655Date: Wed Dec 19 16:20:36 2007 +1030 154656 154657 include: never overwrite realInputProc with enqueueInputProc. Bug #13511 154658 154659 In some cases (triggered by a key repeat during a sync grab) XKB unwrapping 154660 can overwrite the device's realInputProc with the enqueueInputProc. When the 154661 grab is released and the events are replayed, we end up in an infinite loop. 154662 Each event is replayed and in replaying pushed to the end of the queue again. 154663 154664 This fix is a hack only. It ensures that the realInputProc is never 154665 overwritten with the enqueueInputProc. 154666 154667 This fixes Bug #13511 (https://bugs.freedesktop.org/show_bug.cgi?id=13511) 154668 154669commit bcad2a5a24f30cfdf9eca31915ed5a55ed094285 154670Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154671Date: Tue Dec 18 20:19:26 2007 -0500 154672 154673 XACE: Too many arguments to selection access hook. 154674 154675commit 7721d3e9217b41aab3a0ee5eaa52f5b53cbb07db 154676Author: Adam Jackson <ajax@redhat.com> 154677Date: Tue Dec 18 19:14:26 2007 -0500 154678 154679 Reference cvt and gtf in the xorg.conf man page. 154680 154681commit 51fab1eb30691c503f1b4dc98b465f2bc2e1394e 154682Author: Sam Lau <sam.lau@sun.com> 154683Date: Tue Dec 18 11:38:47 2007 -0800 154684 154685 Sun bug 6278039: Xevie checking wrong size in swapped XevieSelectInput requests 154686 154687 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6278039> 154688 154689commit 97c82ce0510808ea9d8a37a0a121e750f6dd8158 154690Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154691Date: Mon Dec 17 23:11:29 2007 -0500 154692 154693 XACE: Restore the old background None behavior in response to bug #13683. 154694 154695 From the X11 protocol spec: 154696 154697 "If background None is specified, the window has no defined background." 154698 154699 This means that toolkits and apps cannot rely on the "transparent" nature 154700 of the current implementation! At some point before the next release, 154701 XACE will switch back to a solid background as the default. 154702 154703commit 9f6ae61ad12cc2813d04405458e1ca5aed8a539e 154704Author: Peter Hutterer <peter@cs.unisa.edu.au> 154705Date: Tue Dec 18 13:57:07 2007 +1030 154706 154707 dix: set the correct number of valuators in valuator events. 154708 154709 (first_valuator + num_valuators) must never be larger than the number of axes, 154710 otherwise DIX freaks out. And from looking at libXI, anything larger than 6 is 154711 wrong too. 154712 154713commit 6a5c3e04fa43b98ccffd69ad86dd781602f88d0b 154714Author: Dave Airlie <airlied@redhat.com> 154715Date: Mon Dec 17 14:59:12 2007 +1000 154716 154717 mi: set the private key to a unique non-zero value 154718 154719commit a18d28a5efbe6021d6c800506cece28a73545aad 154720Author: Dave Airlie <airlied@redhat.com> 154721Date: Mon Dec 17 13:49:16 2007 +1000 154722 154723 damn then my cut-n-paste ate my end of lines... 154724 154725commit d096bbd01bf7c7e15b5a2c582718f3333e063ddc 154726Author: Dave Airlie <airlied@redhat.com> 154727Date: Mon Dec 17 13:45:15 2007 +1000 154728 154729 Xquartz ate my DMX - thanks 154730 154731commit bf20c4374aeb5160a0dc372df9b49f1bbc05f078 154732Author: Jeremy Huddleston <jeremy@yuffie.local> 154733Date: Sun Dec 16 01:14:32 2007 -0800 154734 154735 Xquartz: Removed launchd plist and x11-exec. The relevant code is now in xinit. 154736 (cherry picked from commit 767b4c9d9daa5d0ea59ac1f0d70146798da631cb) 154737 154738commit 58c2898b62fbf0d8e0f175de7cc208dc29d93788 154739Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 154740Date: Sun Dec 16 01:21:45 2007 +0100 154741 154742 xfree86: permit access to io port 0xffff on the hurd 154743 154744commit b4ef8885e1697b83a0dcc9f7fe79155f19241798 154745Author: Jeremy Huddleston <jeremy@yuffie.local> 154746Date: Sat Dec 15 14:00:19 2007 -0800 154747 154748 Xquartz: Fixed launchd detection 154749 154750commit ff5abc72fcc459d7eac663e5f8e4d40b28749841 154751Author: Otavio Salvador <otavio@ossystems.com.br> 154752Date: Fri Dec 14 17:59:29 2007 -0200 154753 154754 registry: XREGISTRY_UNKNOWN needs to be defined even if XREGISTRY isn't enabled 154755 154756 In case XREGISTRY isn't enabled, XREGISTRY_UNKNOWN is used but it's 154757 not being available. It's now always available. 154758 154759commit a3f7f7b60e391e6106f5db40b3fe5fbc67ccd836 154760Author: Ben Byer <bbyer@bbyer.apple.com> 154761Date: Thu Dec 13 20:45:14 2007 -0800 154762 154763 clarified debug message to indicate that we're sleeping 154764 (in case we get reports about slow launch times, this will 154765 help clarify what's happening) 154766 (cherry picked from commit 2eea3483cf893f8f81bacd434b31408dfb38cb06) 154767 154768commit e0e59b3bbc4d8e7ac3934a6f6a9e4a15b328c475 154769Author: Ben Byer <bbyer@bbyer.apple.com> 154770Date: Thu Dec 13 20:44:33 2007 -0800 154771 154772 we need to link against CoreServices, not ApplicationServices 154773 (cherry picked from commit ba4d2096e7953ef5b971682f0e28535da968acb1) 154774 154775commit 062d9234e233fc4c1c617f59093da973c9d3e2ce 154776Author: Ben Byer <bbyer@bbyer.apple.com> 154777Date: Thu Dec 13 20:40:27 2007 -0800 154778 154779 fixed bug in x11-exec that prevent icon from showing up 154780 (cherry picked from commit e1f4a0c20d3a52d98954c4b28d0ec4d44564bc32) 154781 154782commit 95c02adea80a14e18bb51876bc1418eccdade31d 154783Author: Jeremy Huddleston <jeremy@yuffie.local> 154784Date: Fri Dec 14 15:21:40 2007 -0800 154785 154786 Xquartz: Fixed cpprules include 154787 154788commit 86730337001ba4db6d77fe42406695e32784b157 154789Author: Otavio Salvador <otavio@ossystems.com.br> 154790Date: Fri Dec 14 08:46:35 2007 -0200 154791 154792 kdrive/ati: use operating system input devices 154793 154794commit e110255501e2f699709e6978f5e52d3be96333c8 154795Author: Otavio Salvador <otavio@ossystems.com.br> 154796Date: Fri Dec 14 08:45:09 2007 -0200 154797 154798 kdrive/vesa: use operating system input devices 154799 154800commit 863ba390e9fdf0d37cdf03bf5eebe7fdfe6288f5 154801Author: Otavio Salvador <otavio@ossystems.com.br> 154802Date: Fri Dec 14 00:03:13 2007 -0200 154803 154804 kdrive/fbdev: use operating system input devices 154805 154806commit ca59d3f7bdb5f3724ff45ea57912c0b1098a73d6 154807Author: Arkadiusz Miskiewicz <arekm@maven.pl> 154808Date: Thu Dec 13 00:09:08 2007 +0200 154809 154810 Xprint: Clean up generated files 154811 154812 Remember to clean generated wrapper files. 154813 154814commit 5b02a6ca5b31db69d08f2f452494c0f93a6260d9 154815Author: Bartosz Fabianowski <freebsd@chillt.de> 154816Date: Fri Dec 7 02:38:14 2007 +0000 154817 154818 Input: Fix proximity events with valuators 154819 154820 Initialise num_events to 1, so we always send a proximity event, and then 154821 optionally valuator events. Also make sure mieq can deal with valuator 154822 events sent after proximity events. 154823 154824commit a14a143832be844b4b890b0160ccb9fc8293c28c 154825Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154826Date: Fri Dec 14 00:57:16 2007 -0500 154827 154828 Bump server version for devPrivates rework / XACE. 154829 154830commit a2df51f8e95a814c54b806814020155ac8bd177d 154831Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154832Date: Fri Dec 14 00:53:54 2007 -0500 154833 154834 Bump video driver ABI and extension ABI for devPrivates rework. 154835 154836commit 86b2e59bfb79bd042a13c35fbb4ccecec576f629 154837Merge: 1c1a4bc97 cb0d7e2c2 154838Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154839Date: Fri Dec 14 00:32:24 2007 -0500 154840 154841 Merge branch 'master' into XACE-SELINUX 154842 154843 Conflicts: 154844 154845 configure.ac 154846 154847 This is the last pull before merging to master. 154848 154849commit 1c1a4bc970be061484bb8dcccf945eb08144c656 154850Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154851Date: Thu Dec 13 19:51:40 2007 -0500 154852 154853 devPrivates rework: more cleanup from previous merge operation. 154854 154855commit cb0d7e2c2692a332e2bd5495478ebf9a6cd601d0 154856Author: Jeremy Huddleston <jeremy@yuffie.local> 154857Date: Thu Dec 13 16:23:46 2007 -0800 154858 154859 Xquartz: Don't hardcode libexec dir 154860 (cherry picked from commit 67b479ef80cb740a24981335eb8d596744168a62) 154861 154862commit c39212fd7353fc1a07a30bade90f78356c748e2d 154863Author: Jeremy Huddleston <jeremy@yuffie.local> 154864Date: Thu Dec 13 15:56:31 2007 -0800 154865 154866 Xquartz: Don't hardcode LaunchAgents dir 154867 (cherry picked from commit 07a12d71fefd78c380078efa835700f2868ab204) 154868 154869commit 82e1aff9fbc1d15e3451707e3ccbf4b13eedda94 154870Author: Ben Byer <bbyer@bbyer.apple.com> 154871Date: Thu Dec 13 15:57:39 2007 -0800 154872 154873 Modified X11 plist to use x11-exec 154874 (cherry picked from commit 7d9a11329e476f45e4d9f9aebcb43469321347c7) 154875 154876commit 1a5910588a60af0c136595e2457d897d9e54ac88 154877Author: Ben Byer <bbyer@bbyer.apple.com> 154878Date: Thu Dec 13 15:55:28 2007 -0800 154879 154880 created x11-exec wrapper, which uses LaunchServices to find 154881 (and then exec) X11.app 154882 (cherry picked from commit fc04c9759b30d062111d4a7f3f411ed0f18cbde4) 154883 154884commit c8feb73f5841e7812d8dfdb91f1064e2033d028c 154885Merge: 9a7ce5736 a125ce4a8 154886Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154887Date: Thu Dec 13 18:38:25 2007 -0500 154888 154889 Merge branch 'master' into XACE-SELINUX 154890 154891 Conflicts: 154892 154893 Xext/EVI.c 154894 Xext/bigreq.c 154895 Xext/cup.c 154896 Xext/dpms.c 154897 Xext/fontcache.c 154898 Xext/mitmisc.c 154899 Xext/xcmisc.c 154900 Xext/xf86bigfont.c 154901 Xext/xtest.c 154902 configure.ac 154903 dbe/dbe.c 154904 hw/darwin/darwin.h 154905 hw/darwin/darwinEvents.c 154906 hw/darwin/iokit/xfIOKit.h 154907 hw/darwin/iokit/xfIOKitCursor.c 154908 hw/darwin/quartz/fullscreen/fullscreen.c 154909 hw/darwin/quartz/fullscreen/quartzCursor.c 154910 hw/darwin/quartz/quartz.c 154911 hw/darwin/quartz/quartzCommon.h 154912 hw/darwin/quartz/quartzCursor.c 154913 hw/darwin/quartz/xpr/dri.c 154914 hw/darwin/quartz/xpr/dristruct.h 154915 hw/darwin/quartz/xpr/xprCursor.c 154916 hw/darwin/quartz/xpr/xprFrame.c 154917 hw/xfree86/modes/xf86RandR12.c 154918 include/cursor.h 154919 miext/rootless/rootlessCommon.h 154920 miext/rootless/rootlessScreen.c 154921 miext/rootless/rootlessWindow.c 154922 render/picturestr.h 154923 154924 Trying to pick up the pieces from the darwin churn here... 154925 154926commit efcdc0d7010f4e6ec833842cb010a07068edf7ab 154927Author: Adam Jackson <ajax@redhat.com> 154928Date: Thu Dec 13 15:38:41 2007 -0500 154929 154930 Correct the documentation comments in xf86Modes.c 154931 154932 Most of those functions do not, in fact, work with circular mode lists, 154933 and by this point the API isn't really "proposed" anymore. 154934 154935commit 1768af38c737f4c14d32f587b51a8ec3d3d6ed5f 154936Author: Adam Jackson <ajax@redhat.com> 154937Date: Thu Dec 13 15:06:18 2007 -0500 154938 154939 Add infrastructure for validating modes by memory bandwidth. 154940 154941commit 4359193aaa522599c502d012b9c163e993c01d79 154942Author: Adam Jackson <ajax@redhat.com> 154943Date: Thu Dec 13 10:59:48 2007 -0500 154944 154945 Explain a confusing #ifdef. 154946 154947commit 8cedbb0a53d47b12f03edb726db9d5879c8a63a4 154948Author: Adam Jackson <ajax@redhat.com> 154949Date: Thu Dec 13 10:57:35 2007 -0500 154950 154951 Clean up some #if 0. 154952 154953commit 9a7ce573636e349ee2967991c7cc1407e80ae524 154954Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 154955Date: Wed Dec 12 20:44:59 2007 -0500 154956 154957 xselinux: Add new protocol for setting device create context. 154958 154959commit a125ce4a84f5fb5934fefebd7cfb22a83180874d 154960Author: Aaron Plattner <aplattner@nvidia.com> 154961Date: Wed Dec 12 12:20:54 2007 -0800 154962 154963 Fix dist by including modeline2c.awk. 154964 154965 This was broken by commit cb44b6121c4b7b9dd7ff4ff52aaab914c82ff013, which 154966 removed modeline2c.pl from EXTRA_DIST without adding modeline2c.awk. 154967 154968commit 671592343701d8174a70f1ffb9c818784ea3af7a 154969Author: Aaron Plattner <aplattner@nvidia.com> 154970Date: Wed Dec 12 10:59:15 2007 -0800 154971 154972 Get rid of xf86DefModes.c. 154973 154974 It's out of date and not included in the build. Instead, xf86DefModeSet.c is 154975 built from vesamodes and extramodes using modeline2c.awk and *that's* what gets 154976 built. 154977 154978commit d86e7f2ecc0a0129a722ffaca93dfc7c0c60e842 154979Author: Peter Hutterer <peter@cs.unisa.edu.au> 154980Date: Fri Dec 7 20:40:21 2007 +1030 154981 154982 dix: when getting a bogus event, at least print out the type. 154983 154984commit eab0c4e49015fe96f6d985316f9c5fa28a7eb1fe 154985Author: Jeremy Huddleston <jeremy@yuffie.local> 154986Date: Mon Dec 10 20:57:24 2007 -0800 154987 154988 Xquartz: Pre-process Xquartz man page 154989 (cherry picked from commit dec2633c41dd0adf73264afdf251a6522d6ae76a) 154990 154991commit 1ff945a8e43e622b39b360ee49efd6ae3b77be67 154992Author: Jeremy Huddleston <jeremy@yuffie.local> 154993Date: Mon Dec 10 20:47:48 2007 -0800 154994 154995 Xquartz: Make Alt work with Xmodmap again 154996 (cherry picked from commit 0e017177dcca7185716ca760dcce9ddedc7bfef9) 154997 154998commit 7d61893b49569a72bccb63f1ae8c9ce4ef4e354f 154999Author: Jeremy Huddleston <jeremy@yuffie.local> 155000Date: Mon Dec 10 20:33:30 2007 -0800 155001 155002 Xquartz: Removed Xplugin.h from EXTRA_DIST 155003 (cherry picked from commit a746383eca77c9b9ea2cba0cf1c8fc39c0f7d536) 155004 155005commit 8f2eff643bf421bc4233fbaa2409b75d9f80d147 155006Author: Ben Byer <bbyer@bbyer.local> 155007Date: Sat Dec 8 23:34:40 2007 -0800 155008 155009 remove Xplugin.h, because we should use the one in /usr/include 155010 (cherry picked from commit 3e881032f35f774ff9638678d7e3f77c81f62976) 155011 155012commit 7b573ed43672b1fac7b4e6df85a657942ab4cba6 155013Author: Jeremy Huddleston <jeremy@yuffie.local> 155014Date: Sun Dec 9 12:02:04 2007 -0800 155015 155016 Xquartz: Added missing link to libconfig.a 155017 (cherry picked from commit 14ec1cf1cb7ebc183c05e13f9c2b4b4eed679ff3) 155018 155019commit cd13c4ba5b7a1bdfb419cb492a96a72dccf2681e 155020Author: Jeremy Huddleston <jeremy@yuffie.local> 155021Date: Sat Dec 8 13:18:17 2007 -0800 155022 155023 .gitignore: added hw/xquartz/bundle/org.x.X11.plist 155024 155025commit 020b0e92b039d6ddaea0bbdb890b6a01037bf9b6 155026Author: Jeremy Huddleston <jeremy@yuffie.local> 155027Date: Sat Dec 8 11:49:37 2007 -0800 155028 155029 Xquartz Added quartzKeyboard.h to EXTRA_DIST 155030 (cherry picked from commit 37c9781fdb672229ceab101b080762e15512943f) 155031 155032commit 5e016fa9b2bf28971ed1794f4706c6538b1d411c 155033Author: Ben Byer <bbyer@bbyer.local> 155034Date: Sat Dec 8 06:12:46 2007 -0800 155035 155036 Added darwinEvents.h to EXTRA_DIST 155037 (cherry picked from commit 45e5247564c423a2bf02cfec1993155858c91a14) 155038 155039commit 6bb5dacc1710cdbededb9b28ba89a184ecd0931c 155040Author: Jeremy Huddleston <jeremy@yuffie.local> 155041Date: Sat Dec 8 01:41:37 2007 -0800 155042 155043 Xquartz: Actually add quartzForeground.[hc] 155044 Sorry for the commit spam... I'm tired and was overly quick to commit... forgot to include a neccessary file. 155045 (cherry picked from commit e564b7aeaab63e4c943445275af680b3b5898a94) 155046 155047commit 02df03667052fa6a4e0405b91a005dc48e9b39c4 155048Author: Jeremy Huddleston <jeremy@yuffie.local> 155049Date: Sat Dec 8 01:28:26 2007 -0800 155050 155051 Xquartz: Actually, it should be org.x.X11 for case-sensitive FS 155052 (cherry picked from commit c5ccb98d5d461c8a22fc0f3942a607ac90e1e37e) 155053 155054commit 740cc54f081393d4ffe1a3e91c9e504dfaee3fe9 155055Author: Jeremy Huddleston <jeremy@yuffie.local> 155056Date: Sat Dec 8 01:24:58 2007 -0800 155057 155058 Xquartz: Use org.x.X11 instead of com.apple.X11 for preferences 155059 Fixed inconsistency so preferences get read from the correct source. 155060 (cherry picked from commit a74c38bd9f28735acd602d359d7ca6357aed1e93) 155061 155062commit 41a0aeaae9b7b2f8cc2468fd1f3ee11287d34828 155063Author: Jeremy Huddleston <jeremy@yuffie.local> 155064Date: Sat Dec 8 00:13:47 2007 -0800 155065 155066 XQuartz: Fixed "Multiple Dock Icons" 155067 BAM! 155068 (cherry picked from commit d0dca8a88506f50b51f41f99a2f1feb6954c8a31) 155069 (cherry picked from commit 0502955a2af487b51bf22916ac02e497c2d96aba) 155070 155071commit a1b0346853720e98963910b82603c5cda72bb7f9 155072Author: Jeremy Huddleston <jeremy@yuffie.local> 155073Date: Fri Dec 7 23:26:11 2007 -0800 155074 155075 XQuartz: Cleaned up configure, X11.app path in launchd script 155076 Don't hardcode X11.app's path in the launchd plist. 155077 Only install the launchd plist if we --enable-launchd. 155078 (cherry picked from commit 6b74c535dc331d1d621b2541492a3336f69d70a2) 155079 155080commit 1157cfcc5a4e2a7299a4c48df04a1cc8d5093906 155081Author: Ben Byer <bbyer@bbyer.local> 155082Date: Fri Dec 7 21:55:42 2007 -0800 155083 155084 Just a couple of small uninitialized pointer fixes 155085 (cherry picked from commit d12b650362da100ceaecb7e859cd4ef1908d4407) 155086 155087commit 0ad1c359c5b0be63748f5c630c97be88a8cc92ce 155088Author: Jeremy Huddleston <jeremy@dhcp-38-248.EECS.Berkeley.EDU> 155089Date: Fri Dec 7 18:54:58 2007 -0800 155090 155091 Darwin: Use __APPLE__ instead of __DARWIN__ 155092 (cherry picked from commit 54654815fa5e59b25cfd1fa72610120b72c10175) 155093 155094commit 85ed0bb44011312dfaa9f2dc31642a0f89ec0bd3 155095Author: Brice Goglin <bgoglin@debian.org> 155096Date: Sat Dec 8 02:53:27 2007 +0100 155097 155098 Add a missing linebreak after LoadModule: "foo" 155099 155100 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=454742 155101 155102commit 4fc288a13f825db942c9dcd64f4abd0265652faf 155103Author: Alan Coopersmith <alan.coopersmith@sun.com> 155104Date: Fri Dec 7 17:28:37 2007 -0800 155105 155106 Check for <sys/sdt.h> as well when determining to enable dtrace probes 155107 155108 Avoids auto-detecting dtrace is present on systems with the ISDN trace tool 155109 named dtrace installed, but not the dynamic tracing facility named dtrace 155110 155111commit 67907904f094c803d5faf6fa2ce23c01f9a5a521 155112Author: Ben Byer <bbyer@bbyer.local> 155113Date: Fri Dec 7 01:51:53 2007 -0800 155114 155115 fixed pathname in GL/apple/Makefile.am 155116 (cherry picked from commit b6357cec6d837226009c0d2b69026027da36656e) 155117 155118commit 56f5066d477836a975122f4e5748c0f4fb790175 155119Author: Ben Byer <bbyer@bbyer.apple.com> 155120Date: Thu Dec 6 20:51:32 2007 -0800 155121 155122 ALLOCATE_LOCAL is dangerous on Darwin due to memory layout differences 155123 from Linux, so let's define NO_ALLOCA. 155124 (cherry picked from commit 7caf51d1a5a86ae884e0087795636222c082962c) 155125 155126commit c238ef06a270c0c1d48cdb9175b6d5815c7c2a49 155127Author: Jeremy Huddleston <jeremy@yuffie.local> 155128Date: Wed Dec 5 21:36:34 2007 -0800 155129 155130 Darwin: Dead coded removal 155131 Kill off assert macro 155132 (cherry picked from commit d6493abedb2caf03b2bc3a6440b637df67eff081) 155133 155134commit 3a058456e0ce531b21ae2b37be8868b8e0d56e56 155135Author: Peter Hutterer <peter@cs.unisa.edu.au> 155136Date: Fri Dec 7 20:23:36 2007 +1030 155137 155138 mi: unify ErrorFs, prepend [mi] 155139 155140commit c5acf401dad6fbca47f2bb870185c1d08ac947a5 155141Author: Peter Hutterer <peter@cs.unisa.edu.au> 155142Date: Fri Dec 7 20:19:43 2007 +1030 155143 155144 mi: unify last two ErrorF's in mieq.c 155145 155146commit 117458d2db49efd3f04432ff45871b44c7d4ad57 155147Author: Peter Hutterer <peter@cs.unisa.edu.au> 155148Date: Fri Dec 7 20:18:49 2007 +1030 155149 155150 mi: change infamous "Tossed event ..." error for something more explanatory. 155151 155152 Few whitespace errors fixed. 155153 Two ErrorF's prefixed with [mi]. 155154 155155commit 8a8239f2e21795602fcff5281833b350e6b2a286 155156Author: Jeremy Huddleston <jeremy@yuffie.local> 155157Date: Wed Dec 5 21:23:36 2007 -0800 155158 155159 Darwin: Renamed DarwinModeBlah to QuartzBlah 155160 (cherry picked from commit 08ebf86d379b1ddfb08df65d29aea5df66de4327) 155161 155162commit 540439a966cce3fc68a7e4bffdb5bcab1b20725f 155163Author: Jeremy Huddleston <jeremy@yuffie.local> 155164Date: Wed Dec 5 20:55:06 2007 -0800 155165 155166 .gitignore: Added Xcode user files 155167 155168commit bc65a243930e4b02f06a861495420b0a120eae8c 155169Author: Jeremy Huddleston <jeremy@yuffie.local> 155170Date: Wed Dec 5 19:43:49 2007 -0800 155171 155172 Darwin: Flattened quartz into darwin, renamed darwin xquartz 155173 Leaving xpr unflattened since we want modularity to replace that with xpc (XPluginComposite) at some point 155174 (cherry picked from commit 48e6a75fbdd0fee86e364f02ace83f20b312a2b2) 155175 155176commit e00f7061b22001989edf5bd38c2d0cc1566fdd19 155177Author: Jeremy Huddleston <jeremy@yuffie.local> 155178Date: Tue Dec 4 23:18:37 2007 -0800 155179 155180 Darwin: Cleaned up keyboard interface headers 155181 (cherry picked from commit 141f69dc3d8d6e7d8ff65607f43700ac11243041) 155182 155183commit cb44b6121c4b7b9dd7ff4ff52aaab914c82ff013 155184Author: Andrew Oakley <andrew@ado.is-a-geek.net> 155185Date: Wed Dec 5 20:23:05 2007 -0500 155186 155187 Fix commit aa0dfb3f42f19bb351ca7f1a9507ff5ec4590e96 155188 From bugzilla bug 13467¹: 155189 155190 Currently the xserver fails to build without this (now deleted) file, as the 155191 Makefile tries to distribute it. The patch simply removes the reference to 155192 modeline2c.pl. 155193 155194 1] http://bugs.freedesktop.org/show_bug.cgi?id=13467 155195 155196 Signed-off-by: James Cloos <cloos@jhcloos.com> 155197 155198commit 320abd7d1d906807448fa01ad3377daf707f46cc 155199Author: Daniel Stone <daniel@fooishbar.org> 155200Date: Wed Dec 5 19:37:48 2007 +0000 155201 155202 XKB: Actions: Don't run certain actions on the core keyboard 155203 155204 Don't run VT switches, terminations, or anything, on the core keyboard: only 155205 run actions which affect the keyboard state. If we get an action such as VT 155206 switch, just swallow the event. 155207 155208commit 85dd8efac1bc0715f03c99d261b1c5d0980623e1 155209Author: Daniel Stone <daniel@fooishbar.org> 155210Date: Wed Dec 5 19:36:59 2007 +0000 155211 155212 WaitForSomething: Ignore EAGAIN 155213 155214 If select ever returns EAGAIN, don't bother complaining. 155215 155216commit b7f3618f3933a810778093fd47564a1e3bf3fde6 155217Author: Rich Coe <Richard.Coe@med.ge.com> 155218Date: Wed Dec 5 19:36:37 2007 +0000 155219 155220 OS: Connection: Keep trying select while it gets interrupted (bug #9240) 155221 155222 If we got interrupted (EINTR or EAGAIN) during select, just try again, rather 155223 than shutting clients down on either of these errors. 155224 155225commit d8b2cad3771a09860e7be1726f67e684cf7caeec 155226Author: Rich Coe <Richard.Coe@med.ge.com> 155227Date: Wed Dec 5 19:31:07 2007 +0000 155228 155229 OS: Connection: Don't shut down disappeared clients (bug #7876) 155230 155231 If a client disappears in the middle of CheckConnections (presumably 155232 because its appgroup leader disappears), then don't attempt to shut it down 155233 a second time, when it's already vanished. 155234 155235commit a8e27a108abeba73b2888da4e0604008f4b02045 155236Author: Kanru Chen <koster@debian.org.tw> 155237Date: Mon Dec 3 12:46:45 2007 +0000 155238 155239 Config: HAL: Fix XKB option parsing 155240 155241 Actually combine the XKB options into a string, rather than just repeatedly 155242 writing a comma. 155243 155244commit 2d723bbd0d36f6d7763b4df3298d40720f97fdd0 155245Author: Peter Harris <peter.harris@hummingbird.com> 155246Date: Mon Oct 29 18:05:19 2007 -0400 155247 155248 Add missing swaps in panoramiXSwap.c 155249 155250commit 0fccb24aa978b838cf0fb008e9695837e612c529 155251Author: Daniel Stone <daniel@fooishbar.org> 155252Date: Fri Nov 30 20:35:26 2007 +0200 155253 155254 ProcessOtherEvent: Don't do double translation of button events 155255 155256 We already deal with the button mapping in GetPointerEvents, so don't 155257 do the remapping again in ProcessOtherEvent. 155258 155259commit c6cfcd408df3e44d0094946c0a7d2fa944b4d2d1 155260Author: Hong Liu <hong.liu@intel.com> 155261Date: Wed Dec 5 17:48:28 2007 +0100 155262 155263 Bug 13308: Verify and reject obviously broken modes. 155264 155265commit 8d0efe4c2a48047680af40e5f6d639f426902e07 155266Author: Jeremy Huddleston <jeremy@yuffie.local> 155267Date: Tue Dec 4 17:59:13 2007 -0800 155268 155269 Darwin: Rework build system to more accurately reveal code infrastructure and facilitate future modularity. 155270 (cherry picked from commit e8399fd4d66a2b77b770c277e2fa424229a721b2) 155271 155272commit cc98a8e2415f12c7a90fd846d1ec858068e8c796 155273Author: Jeremy Huddleston <jeremy@yuffie.local> 155274Date: Mon Dec 3 23:59:19 2007 -0800 155275 155276 Darwin: RIP dumpkeymap, cr, and fullscreen 155277 Taking out the trash. 155278 We don't need dumpkeymap since we'll be getting keymaps straight from the OS. .Xmodmap should be sufficient for any user-needed changes. If this is not 155279 the case, please let us know, so we can address any problems you have. 155280 fullscreen never worked AFAIK 155281 cr isn't being used and xpr is much better. 155282 (cherry picked from commit e41af2967e885466c4d194fa4c3b358e6be37c30) 155283 155284commit 13af2d1efcc83d1412a4c727afddd97577b00f32 155285Author: Adam Jackson <ajax@redhat.com> 155286Date: Tue Dec 4 17:36:21 2007 -0500 155287 155288 Restore xf86getsecs() as not having an ANSI equivalent. 155289 155290commit 447cd5d411875b62eb1a501bf00e604225b74d26 155291Author: Peter Hutterer <peter@cs.unisa.edu.au> 155292Date: Tue Dec 4 19:09:37 2007 +1030 155293 155294 dix: update comments about inputInfo.pointer. 155295 155296commit f44d7dcb5fefca3ddfd45e75d0bd0b47ec785e48 155297Author: Peter Hutterer <peter@cs.unisa.edu.au> 155298Date: Tue Dec 4 19:07:46 2007 +1030 155299 155300 dix: change the filters to be per-device. 155301 155302 If we have one global filter, one pointer may change the filter value and 155303 affect another pointer. 155304 155305 Reproduceable effect: 155306 blackbox and xterm, start dragging xterm then click anywhere with the other 155307 pointer (attached to different masterd device!). The button release resets 155308 the filter[Motion_Filter(button)] value, thus stopping dragging and no event 155309 is sent to the client anymore. 155310 Having the filters set per device gets around this. 155311 155312commit 0931f40bf1bd6e00b8d95968d761a495b2c9a46c 155313Author: Peter Hutterer <peter@cs.unisa.edu.au> 155314Date: Tue Dec 4 17:08:56 2007 +1030 155315 155316 dix: comments, whitespaces, copyright fixes. 155317 155318 Removing my copyright message for now, should eventually be in line with the 155319 rest of the messages. 155320 155321commit 09c0c1a3cc4807813774a3c0e28a7ba9a87bb5c7 155322Author: Peter Hutterer <peter@cs.unisa.edu.au> 155323Date: Tue Dec 4 16:51:52 2007 +1030 155324 155325 dix: remove #ifdef XINPUT. 155326 155327 No escaping XINPUT anymore. 155328 155329commit f8d7729df388c142624def36ba6d8c3b15922018 155330Author: Jeremy Huddleston <jeremy@yuffie.local> 155331Date: Mon Dec 3 20:20:05 2007 -0800 155332 155333 Darwin: Combine launcher and server X11.app 155334 This should hopefully eliminate confusion some people have over which X11.app is which. 155335 Now BOTH are in /A/U/X11.app and we intelligently determine whether to execute our app_to_run 155336 or launch the server. If arguments are given, we launch the server. Otherwise if we can 155337 connect to an X DISPLAY, we execute app_to_run. Otherwise, we launch the server. 155338 155339 (cherry picked from commit e7026216ccaa8e4fb073800ba947c9909d4faada) 155340 155341commit 678f786715d76e972f8a77807c9caf3e90c24418 155342Author: Dave Airlie <airlied@redhat.com> 155343Date: Tue Dec 4 12:24:47 2007 +1100 155344 155345 xf86crtc: oh mon could be NULL, so check before quirks 155346 155347commit a9df4bb555fd91707a68794c2dce24fb06e6cf64 155348Author: Dave Airlie <airlied@redhat.com> 155349Date: Tue Dec 4 12:17:29 2007 +1100 155350 155351 xf86Crtc: pass correct parameter. 155352 155353 quite how this has worked I've no idea. 155354 155355commit fe25f897c62bb324660217e15dbd3091c808dbba 155356Author: Adam Jackson <ajax@redhat.com> 155357Date: Mon Dec 3 18:34:40 2007 -0500 155358 155359 xf86getpagesize -> getpagesize elsewhere in os-support/ 155360 155361commit b84f2833a681585162b8dabfb02ff62e7e0ef4d6 155362Author: Alan Coopersmith <alan.coopersmith@sun.com> 155363Date: Mon Dec 3 14:52:17 2007 -0800 155364 155365 xf86getpagesize() -> getpagesize() in os-support/solaris/sun_bios.c 155366 155367commit aa0dfb3f42f19bb351ca7f1a9507ff5ec4590e96 155368Author: James Cloos <cloos@jhcloos.com> 155369Date: Mon Dec 3 16:57:58 2007 -0500 155370 155371 Remove Perl dependency from the build 155372 From bugzilla bug 13467¹: 155373 155374 The modeline2c script is the only part of the Xorg server that requires Perl. 155375 [This] is a simpler replacement that works with any normal AWK. 155376 155377 1] http://bugs.freedesktop.org/show_bug.cgi?id=13467 155378 155379 Bug was posted by Joerg Sonnenberger <joerg@NetBSD.org>. 155380 155381commit 60086d90168265795e07a60939e9e2fe95c6e15c 155382Author: Alan Coopersmith <alan.coopersmith@sun.com> 155383Date: Mon Dec 3 11:30:58 2007 -0800 155384 155385 Use pkg-config to get -I, -L & -R flags needed for OpenSSL 155386 155387 Still just uses -lcrypto instead of the full library list from --libs 155388 155389commit b77ca7cc9c23184c4ab367baf1b3ed0acf27c269 155390Author: Alan Coopersmith <alan.coopersmith@sun.com> 155391Date: Mon Dec 3 11:29:54 2007 -0800 155392 155393 Use _X_EXPORT instead of __attribute__((visibility("default"))) 155394 155395commit 1faba797cbfe1a4804b7ea6b47e1ca9d4e4324e4 155396Author: Adam Jackson <ajax@redhat.com> 155397Date: Mon Dec 3 14:12:58 2007 -0500 155398 155399 Death to libcwrapper. 155400 155401 This has been deprecated since 1.1. Since we're breaking ABI again anyway, 155402 remove it entirely. 155403 155404commit 28b93d74a11a1064d68a214fcaa7b0aede864a38 155405Author: Peter Hutterer <peter@cs.unisa.edu.au> 155406Date: Mon Dec 3 14:32:53 2007 +1030 155407 155408 ephyr: miPointerWarpCursor needs the device parameter. 155409 155410 Thanks to Mitchell Quille for spotting it. 155411 155412commit 004876355b43fb4d3c2bc5653a7dc1bfd3f985ee 155413Author: Peter Hutterer <peter@cs.unisa.edu.au> 155414Date: Fri Nov 30 13:56:04 2007 +1030 155415 155416 dix: Remove RegisterPairingClient and UnregisterPairingClient. 155417 155418commit f4dc521b38560c8f93b614316a3a5511941a93a9 155419Author: Jeremy Huddleston <jeremy@tifa.local> 155420Date: Sun Dec 2 18:21:40 2007 -0800 155421 155422 Darwin: Added {/,/System/}Library/Fonts to DEFAULT_FONT_PATH 155423 (cherry picked from commit b0069b04dddaa2df6d4cdf86f96fd8a2a257e47e) 155424 155425commit 0fff01f5660fb3bb9284f97c45dc76154435d02b 155426Author: Adam Jackson <ajax@aspartame.nwnk.net> 155427Date: Sun Dec 2 14:15:36 2007 -0500 155428 155429 Fix swapped Xv dispatch under Xinerama. 155430 155431 Same-endian dispatch was properly calling through the Xinerama wrapping, 155432 but other-endian dispatch wasn't. 155433 155434commit fa47910045c3700d8d668b5e214e5ffc1e8dc3e7 155435Author: Adam Jackson <ajax@redhat.com> 155436Date: Sun Dec 2 12:39:05 2007 -0500 155437 155438 Clean up many #if 0. 155439 155440commit 83ba1e167c1473ac7d85239a6ee5ed629353cb16 155441Author: Ben Byer <bbyer@bbyer.local> 155442Date: Sat Dec 1 18:28:19 2007 -0800 155443 155444 added missing call to xcb_connect() 155445 (cherry picked from commit dc2fb323ee11f081d447605be151024f9e2487f9) 155446 155447commit 9ad4560b3cbd42e647d6227746d4d037616d57cf 155448Author: Jeremy Huddleston <jeremy@tifa.local> 155449Date: Sat Dec 1 16:23:23 2007 -0800 155450 155451 Darwin: Alt is Mode_switch Switching to Mode_switch to maintain compatibility with Tiger X11. 155452 (cherry picked from commit 8a76c99c0ebbaf7375f3a9c75c4f7921a79024da) 155453 155454commit f83d758dcc4878849a851c8466f6fa16b2b7cd8e 155455Author: Jeremy Huddleston <jeremy@yuffie.local> 155456Date: Fri Nov 30 16:11:15 2007 -0800 155457 155458 Darwin: properly implemented xcb check for stale sockets 155459 (cherry picked from commit f543cb8fbb3d9213cb03396f4252ab9821319993) 155460 155461commit f54b28eeba119c42d0fcccfbe295306dd670221a 155462Author: Jeremy Huddleston <jeremy@yuffie.local> 155463Date: Fri Nov 30 16:09:23 2007 -0800 155464 155465 Darwin: Undo focus-hack which didn't work right. 155466 155467commit f30abe30c5fea10e680aa12f3fe37ee8ce1a0201 155468Author: Dave Airlie <airlied@linux.ie> 155469Date: Fri Nov 30 13:52:06 2007 +1000 155470 155471 edid quirk for MAX 0x77e monitor 155472 155473 From RH bugzilla 306441 155474 155475commit 6216abe0c1a94efa4b2f0d14e88278f128304cd2 155476Author: Peter Hutterer <peter@cs.unisa.edu.au> 155477Date: Fri Nov 30 11:34:24 2007 +1030 155478 155479 dix: avoid activating the VCP/VCK twice. 155480 155481 This may set dev->key-xkbinfo to NULL, causing a segfault in xkb code lateron. 155482 Spotted by David Huggins-Daines. 155483 155484commit 9eb8ea9e615ebd8b568711eeca36aa84aa781ad3 155485Author: Peter Hutterer <peter@cs.unisa.edu.au> 155486Date: Fri Nov 30 11:32:59 2007 +1030 155487 155488 dix: only freeze the paired MD on a grab, not all other devices. 155489 155490commit 8a079be0dd0f2ce37868988cde4ac8895522b088 155491Author: Jeremy Huddleston <jeremy@yuffie.local> 155492Date: Thu Nov 29 02:19:22 2007 -0800 155493 155494 Darwin: #ifdefs around dix-config.h include and NDEBUG/assert.h workaround. 155495 (cherry picked from commit d2b768890f0878ae4e3fec8f7219e82b79256133) 155496 155497commit 38397560612424b5b348f34c1a0bea8c47a574be 155498Author: Jeremy Huddleston <jeremy@yuffie.local> 155499Date: Wed Nov 28 23:07:41 2007 -0800 155500 155501 Darwin: Removed support for darwinSwapAltMeta 155502 (cherry picked from commit 3d153c8fa40986d194b7701f5eafa0080e32399a) 155503 155504commit 89c3dfe41e3a17a4f27b20e23623dc5777670feb 155505Author: Dave Airlie <airlied@redhat.com> 155506Date: Thu Nov 29 19:57:24 2007 +1100 155507 155508 modes: use xf86RandR12Index to stop illegal access 155509 155510 xf86RandR12Index set to -1, and if initialised it gets 0 or higher. 155511 This allows the server to start with xinerama turned on with only one head 155512 155513commit 725710fd0bc990b2c35e4c76128ef1c668013299 155514Author: Dave Airlie <airlied@redhat.com> 155515Date: Thu Nov 29 19:40:53 2007 +1100 155516 155517 randr: make randr code not segfault when xinerama set 155518 155519commit e4fe0a3cb789b8757f1c80f606dfe32bccada582 155520Author: Peter Hutterer <peter@cs.unisa.edu.au> 155521Date: Thu Nov 29 09:55:24 2007 +1030 155522 155523 xkb: swap a LookupKeyboardDevice over in favour of GetPairedDevice 155524 155525 The former always returns the VCK, which is obviously wrong if we have 155526 multiple devices. 155527 155528commit 23b8ca8a373d919225de9739af7b064f650eceec 155529Author: Adam Jackson <ajax@redhat.com> 155530Date: Tue Nov 27 13:20:40 2007 -0500 155531 155532 RANDR 1.2: Only enable unknown outputs if there are no connected outputs. 155533 155534 Otherwise you end up with a confusing initial geometry, and xscreensaver 155535 and friends get very angry. 155536 155537commit edebe76cfdb31072d18a6fcd3ee8f1d95006855f 155538Author: Peter Hutterer <peter@cs.unisa.edu.au> 155539Date: Tue Nov 27 10:22:44 2007 +1030 155540 155541 Xi: set DeviceXXXState's length fields to the correct size of the struct. 155542 155543 Setting it to the size of a pointer is an interesting but equally wrong 155544 approach. Luckily Xlib never used this field anyway so nobody got hurt so far. 155545 155546 Spotted by Simon Thum. 155547 (cherry picked from commit 0f2398d06ce591724e388b3270800c5e22b3de2d) 155548 155549commit 0f2398d06ce591724e388b3270800c5e22b3de2d 155550Author: Peter Hutterer <peter@cs.unisa.edu.au> 155551Date: Tue Nov 27 10:22:44 2007 +1030 155552 155553 Xi: set DeviceXXXState's length fields to the correct size of the struct. 155554 155555 Setting it to the size of a pointer is an interesting but equally wrong 155556 approach. Luckily Xlib never used this field anyway so nobody got hurt so far. 155557 155558 Spotted by Simon Thum. 155559 155560commit 601307615e4955be23fd86a057285074242ad83e 155561Author: Jeremy Huddleston <jeremy@yuffie.local> 155562Date: Mon Nov 26 13:04:57 2007 -0800 155563 155564 Darwin,Rootless: Makefile cleanup 155565 (cherry picked from commit 9c6d8a035b712b219833653ac637b89703a9b0c3) 155566 155567commit 8503072e1c2b89dca786d4afb72aa60a170d2fbd 155568Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155569Date: Mon Nov 26 16:52:41 2007 -0500 155570 155571 registry: add missing include statement. 155572 155573commit 996b621bec1bbc4fb21970c75eaec62053bc6ccb 155574Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155575Date: Mon Nov 26 15:59:44 2007 -0500 155576 155577 registry: swap out the DTRACE XErrorDB stuff for the new registry call. 155578 155579commit 54cb729ecc2d366c1af836cb3d2ffc8e864e9b79 155580Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155581Date: Mon Nov 26 15:59:01 2007 -0500 155582 155583 registry: Add a call for DTRACE compatibility. 155584 155585commit decd5a7c605e42c99b6a4523c8e1833b859d9b24 155586Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155587Date: Mon Nov 26 15:26:49 2007 -0500 155588 155589 registry: Rebase registry to use the server config file of protocol names. 155590 155591commit 9b0e72c8d960d056276f5fa93f3cc2872825711e 155592Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155593Date: Mon Nov 26 15:26:04 2007 -0500 155594 155595 registry: Add a great big list of protocol names, like the XErrorDB that 155596 ships with Xlib. This is considered temporary, until server-side XCB can 155597 solve the problem programmatically. 155598 155599commit c0f9e204baf0218466973868c5ea6ed0f78e6b8b 155600Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155601Date: Mon Nov 26 15:24:15 2007 -0500 155602 155603 registry: rename the SERVERCONFIGdir and relocate it to /usr/lib/xorg 155604 by default. 155605 155606commit a344920ae86c1970e4cc34ee91e2f2008d490c49 155607Author: Adam Jackson <ajax@redhat.com> 155608Date: Mon Nov 26 11:53:08 2007 -0500 155609 155610 Allow Virtual to be specified globally in the Screen section. 155611 155612 The Display subsections are optional, and it's confusing to need to create 155613 them just to set a Virtual size. 155614 155615commit c6c284e64b1f537a3243856cf78cf3f2324e4c2b 155616Author: Matthias Hopf <mhopf@suse.de> 155617Date: Mon Nov 26 15:38:20 2007 +0100 155618 155619 Initialize Mode with 0 in xf86RandRModeConvert. 155620 155621 Asking for trouble if non-initialized values contain random data. 155622 155623commit a4edfbef022f9635c2c9b9eb229cb622834dc68c 155624Author: Peter Hutterer <peter@cs.unisa.edu.au> 155625Date: Sat Nov 24 16:16:48 2007 +1030 155626 155627 mi: only call UpdateSpriteForScreen if the screen actually changed. 155628 155629 If we call it unconditionally, we flip the dev->spriteInfo->sprite permanently 155630 when using XTestFakeInput (once in CheckMotion as called from the 155631 processInputProc, another time in UpdateSpriteForScreen when we actually warp 155632 the cursor). USFS also updates to the screen's root window unconditionally, 155633 which is not really what we want if we haven't changed screen at all. 155634 155635commit bf3198c8c56289244c58d36c6869442479fd3481 155636Author: Peter Hutterer <peter@cs.unisa.edu.au> 155637Date: Sat Nov 24 15:00:57 2007 +1030 155638 155639 dix: fix typo 155640 155641commit 5dabe448bda68a483bf444a4adfed2b25b30f600 155642Author: Peter Hutterer <peter@cs.unisa.edu.au> 155643Date: Sat Nov 24 15:00:43 2007 +1030 155644 155645 dix: Add special treatment of NotifyUngrab for enter/leave events. 155646 155647 In some cases (e.g. using x2x) the previous model broke, with a window ending 155648 not counting down to 0 anymore. Special treatment for NotifyUngrab seems to 155649 help here. 155650 155651 Current solution: If a NotifyGrab is to be sent ignore it. If a NotifyUngrab 155652 enter is sent, only adjust the semaphore if it is on 0. Likewise, do the same 155653 for a NotifyUngrab leave if the semaphore is on 1. This seems to work alright 155654 so far. 155655 155656commit 33b94da6327d3423b4ebc1a58d5894c9904e67c9 155657Author: Keith Packard <keithp@keithp.com> 155658Date: Fri Nov 23 16:01:11 2007 -0800 155659 155660 Re-add call to XFixesExtensionInit for static servers. 155661 155662 This reverts a portion of bcbaf2a0ce34b6c5e41d2831b8b87dbd0617a89b which 155663 removed the call to XFixesExtensionInit and some cpp lines. 155664 155665commit f6401f944d327cc5d9a7ee0bbdf4f7fc8eaa31e8 155666Author: Matthias Hopf <mhopf@suse.de> 155667Date: Fri Nov 23 16:12:49 2007 +0100 155668 155669 Don't segfault if referring to a relative output where no modes survived. 155670 155671commit fa19e84714aa84a2f2e817e363d6440349d0b619 155672Author: Matthias Hopf <mhopf@suse.de> 155673Date: Tue Nov 20 16:54:50 2007 +0100 155674 155675 Fix initial placement of LeftOf and Above. 155676 155677commit 184e571957f697f2a125dc9c9da0c7dfb92c2cd9 155678Author: Matthias Hopf <mhopf@suse.de> 155679Date: Tue Nov 20 13:05:26 2007 +0100 155680 155681 Adjust offsets of modes that do not fit virtual screen size. 155682 155683 Fixes memory corruption if a too small "Virtual" was specified in xorg.conf 155684 for the selected multi-monitor configuration. 155685 155686commit a80e64f1503a4d8b11c4a6608d296422c69e3e8b 155687Author: Peter Hutterer <peter@cs.unisa.edu.au> 155688Date: Sat Nov 17 22:50:07 2007 +0100 155689 155690 XKB: Generate correct key repeat events (bug #13114) 155691 155692 Make sure we send the correct event for the type of device when we're 155693 sending key repeat events, which stops repeats being sent to incorrect 155694 windows. 155695 155696commit 63351df0eec320aa3ce27d4d2ee6bcdb58aa2d92 155697Author: Jeremy Huddleston <jeremy@tifa.local> 155698Date: Thu Nov 22 18:02:07 2007 -0800 155699 155700 Darwin: Fix compilation/linking problems 155701 155702commit bf4ef4da759c01e6794ed28ba4988a2c8ee049bf 155703Author: Jeremy Huddleston <jeremy@tifa.local> 155704Date: Thu Nov 22 17:45:15 2007 -0800 155705 155706 Darwin: Remove the PaintWindow optimization which snuck back in. 155707 155708commit 2082e7aa878fe1221fd50895a9de1f408b3157a8 155709Author: Jeremy Huddleston <jeremy@tifa.local> 155710Date: Thu Nov 22 17:18:48 2007 -0800 155711 155712 Rootless: Remove the PaintWindow optimization which snuck back in. 155713 155714commit a751bc12bee1d4d2ed35e3a0c64d9c8c9bf30a82 155715Author: Jeremy Huddleston <jeremy@tifa.local> 155716Date: Thu Nov 22 13:53:00 2007 -0800 155717 155718 Rootless: Imported changes made in xorg-server-1.2-apple branch 155719 155720commit 59c7ca6586e7c20e28ad407ca9a0883c4d621d64 155721Author: Jeremy Huddleston <jeremy@tifa.local> 155722Date: Thu Nov 22 13:29:15 2007 -0800 155723 155724 Darwin: Added missing Makefile.am 155725 155726commit 23596291c30a85e38c00aef2c01b46d561e2916e 155727Author: Jeremy Huddleston <jeremy@tifa.local> 155728Date: Thu Nov 22 13:17:44 2007 -0800 155729 155730 Darwin: More syncing witn xorg-server-1.2-apple 155731 155732commit 4e18c626350c7c2e0fb540aa64a98957699f3abe 155733Author: Jeremy Huddleston <jeremy@tifa.local> 155734Date: Thu Nov 22 12:21:59 2007 -0800 155735 155736 Rootless: Pulled in changes from fb{Blt,Fill} into rl{Blt,Fill} 155737 (cherry picked from commit 3f857e129df7ce492191e0c51b8e53eaf6179366) 155738 (cherry picked from commit 70374a58937d7a6f01c210bd6ac66cafb63e895a) 155739 155740commit ed9524d36e42a310bb128284f2b507f76b8c40d9 155741Author: Jeremy Huddleston <jeremy@tifa.local> 155742Date: Thu Nov 22 01:07:02 2007 -0800 155743 155744 Darwin: Copied over missing file (Localizable.strings) from xorg-server-1.2-apple 155745 155746commit 4d9cef197b12548e0716dab3557e48311519e325 155747Author: Jeremy Huddleston <jeremy@tifa.local> 155748Date: Thu Nov 22 00:35:09 2007 -0800 155749 155750 Darwin: Misc cleanups to line up with xorg-server-1.2-apple 155751 155752commit 5e950123daa167c9ffe289b3bd89e3bd288da0e3 155753Author: Jeremy Huddleston <jeremy@tifa.local> 155754Date: Wed Nov 21 23:32:00 2007 -0800 155755 155756 Darwin: Removed cvs tags from Xquartz man page 155757 155758commit 3a2f714eea475a13cde65921e24c7ee3f70ffc3c 155759Author: Jeremy Huddleston <jeremy@tifa.local> 155760Date: Wed Nov 21 23:30:37 2007 -0800 155761 155762 Darwin: Removed .svn dir 155763 155764commit 691da031319dc59b9496101358c267f317abfd1e 155765Author: Peter Hutterer <peter@cs.unisa.edu.au> 155766Date: Thu Nov 22 17:44:39 2007 +1030 155767 155768 Xi: allocate motion history for master device. 155769 155770 We're still missing out on the actual content of the history but at least this 155771 way we don't segfault. 155772 155773commit 602de4f70b6f4aab93b514f3a01917bd5d4ad640 155774Author: Jeremy Huddleston <jeremy@tifa.local> 155775Date: Wed Nov 21 16:53:10 2007 -0800 155776 155777 Darwin: Use UTF8String since lossyCString is deprecated 155778 (cherry picked from commit 1786f9464af51ff606a612aec6fe420fa9688a28) 155779 155780commit 13666e287c347aab2a5e9d8ee5f6bb29b9b85171 155781Author: Ben Byer <bbyer@bbyer.local> 155782Date: Sun Nov 18 17:44:12 2007 -0800 155783 155784 Darwin: Added some DEBUG_LOG sauce to the XP_EVENT handling code 155785 (cherry picked from commit ec84a4cef66a2b46ed71f9758c434ea629d2f270) 155786 155787commit 829b6641bd64c352e1e8a7c619f84dedbdb07a09 155788Author: Ben Byer <bbyer@bbyer.local> 155789Date: Sun Nov 18 17:43:40 2007 -0800 155790 155791 Darwin: Disabled ALT_IS_MODE_SWITCH 155792 (cherry picked from commit fd181254f85543558190140787dc7b41f6cf90db) 155793 155794commit 8486f8af91b477c7bcb8438a0e9a72d0c11d1d63 155795Author: Ben Byer <bbyer@bbyer.local> 155796Date: Thu Nov 15 02:25:50 2007 -0800 155797 155798 Darwin: Added a lightweight debugging facility to support troubleshooting 155799 (for example) the stuck modifier key issue 155800 (cherry picked from commit 0e0b452d10c0af55497c3299b5f3db45d5b381cb) 155801 155802commit 74214a9f42b931f99d83ddb4efb3720881a2de16 155803Author: Ben Byer <bbyer@bbyer.local> 155804Date: Thu Nov 15 00:56:54 2007 -0800 155805 155806 Darwin: Patch to avert (some) damage / rootless crashes, courtesy of Ken Thomases 155807 (cherry picked from commit 148a87ff20aa5e7a6d839610aa14fa1a31505c4a) 155808 155809commit f5f833b80609f1f98c93113183bd2b1bab3bfec9 155810Author: Ben Byer <bbyer@bbyer.local> 155811Date: Sun Nov 11 04:30:34 2007 -0800 155812 155813 Darwin: These changes are necessary, yet not sufficient, to get 8-bit indexed 155814 color mode working in Xquartz. 155815 (cherry picked from commit a415f62f5289fae99ea9b0038d21fad7695b1336) 155816 155817commit 8358334180a4f8c1e73fc5647a62bcd3539dee45 155818Author: Ben Byer <bbyer@bbyer.local> 155819Date: Sun Nov 11 04:30:00 2007 -0800 155820 155821 Darwin: Fixed the call to xp_init so that we now receive Motion notifications even 155822 if X is not the active application. 155823 fixes <rdar://problem/5167664> xeyes dead until window activation 155824 (cherry picked from commit c7573379a85a1480cc51650075078e41dafe56af) 155825 155826commit 512dee90878e552ad1b2bb5b27366707f6464f28 155827Author: Ben Byer <bbyer@bbyer.apple.com> 155828Date: Thu Nov 8 22:17:38 2007 -0800 155829 155830 Darwin: fix for spurious "Are you sure you want to quit?" message 155831 (cherry picked from commit 30cbfc786e4fedda3fe070bacceabe1d9212d00b) 155832 155833commit 28e73e99a9a59223963312c5dd43ce5566d1db9d 155834Author: Ben Byer <bbyer@bbyer.apple.com> 155835Date: Thu Nov 8 22:12:41 2007 -0800 155836 155837 Darwin: Adding "fake RandR" support from old X11.app 155838 (cherry picked from commit 633490c4e8dab30af7ecbe1bef076c22ad5f5da9) 155839 155840commit 01b70afaac0990b41d1fb6fadbfd64df1486b669 155841Author: Ben Byer <bbyer@bbyer.local> 155842Date: Sat Nov 3 05:34:19 2007 -0700 155843 155844 Darwin: Initial support for Spaces -- if you use Expose to drag an X11 155845 window to another Space, it will work correctly (as opposed 155846 to just leaving a ghost window). We accomplish this by listening 155847 for the notification from Xplugin that our window has been moved, 155848 and then we ask X11 to move the window to the new location. 155849 (cherry picked from commit 2d50ea8013e7c1639d570e227b53b037fb567565) 155850 155851commit b39edc01a6588697b65f831e8ab1dbb24cbe7b24 155852Author: Ben Byer <bbyer@bbyer.local> 155853Date: Wed Oct 31 23:46:50 2007 -0700 155854 155855 Darwin: Swap modifier keys for buttons 2 and 3 -- now Option-click is the middle click 155856 (cherry picked from commit 0aa61293b62aeb69a93b2035d0aef8644343eed3) 155857 155858commit 606a8dc73d91a198d72d249934dc027a23f4c338 155859Author: Ben Byer <bbyer@bbyer.local> 155860Date: Wed Oct 31 03:39:47 2007 -0700 155861 155862 Darwin: Trap Deactivate messages and release modifiers to avoid "stuck shift lock" (etc) bugs 155863 (cherry picked from commit 2b189a99330eb465fa0d17020fb1db1e38829151) 155864 155865commit 4c18ef4331aaee268431a3ba50991f0312b82870 155866Author: Ben Byer <bbyer@bbyer.local> 155867Date: Wed Oct 31 03:22:18 2007 -0700 155868 155869 Darwin: Workaround for a bug where the holding down Command to make a "fake" 155870 button 2 click would actually result in a Command-2 chord. 155871 (I.e. it wasn't releasing Command before clicking the fake button.) 155872 (cherry picked from commit 0d5dd5dffa4c5ce3f54dfe53720a39d524dc8e37) 155873 155874commit f9269bebae27bbc9d0e03e02943166b83946623d 155875Author: Peter Hutterer <peter@cs.unisa.edu.au> 155876Date: Thu Nov 22 17:30:14 2007 +1030 155877 155878 DeviceIntRec: move lastx/lasty from valuator into DeviceIntRec. 155879 155880 We free the ValuatorClassRec quite regularly. If a SIGIO is handled while 155881 we're swapping device classes, we can bring the server down when we try to 155882 access lastx/lasty of the master device. 155883 155884commit 7f2972d47a5d74fe92268c6d609b1eb6ad845824 155885Author: Jeremy Huddleston <jeremy@tifa.local> 155886Date: Wed Nov 21 21:59:59 2007 -0800 155887 155888 Darwin: Really add launcher this time 155889 155890commit bcbaf2a0ce34b6c5e41d2831b8b87dbd0617a89b 155891Author: Jeremy Huddleston <jeremy@tifa.local> 155892Date: Wed Nov 21 19:51:14 2007 -0800 155893 155894 Darwin: Dead code removal, Code cleanup, Added launcher 155895 Imported changes from xorg-server-1.2-apple to make master more current wrt 155896 file layout, build system changes, and dead code removal. 155897 155898commit 9ed43eff48201520797f89a12bb3b2f5819bd99f 155899Author: Peter Hutterer <peter@cs.unisa.edu.au> 155900Date: Wed Nov 21 17:16:27 2007 +1030 155901 155902 Xi: allow clients to specify pure client id in SetClientPointer. 155903 155904 If no window was found with the given ID, try if there's a client with the ID. 155905 Allows us to set the CP for apps that don't open windows (e.g. x2x). 155906 155907commit 33f15689922ad9f1193f803bc1636c82c23ab99e 155908Author: Peter Hutterer <peter@cs.unisa.edu.au> 155909Date: Wed Nov 21 16:06:44 2007 +1030 155910 155911 xtest: switch an inputInfo.pointer over to PickPointer. 155912 155913 Couple of whitespace fixes too. 155914 155915commit e5dd7a95791748c57cab75c0d8ba9d37f72edccf 155916Author: Peter Hutterer <peter@cs.unisa.edu.au> 155917Date: Wed Nov 21 15:59:31 2007 +1030 155918 155919 Xext: Scruffy the janitor don't like no "register" keywords. 155920 155921commit a55ec1a9f4b62139dc5e5462d79d47b330c27c79 155922Author: Alan Coopersmith <alan.coopersmith@sun.com> 155923Date: Tue Nov 20 18:31:03 2007 -0800 155924 155925 Restore checks for __i386 where needed for Sun compilers on Solaris 155926 155927commit 2f387d913aa76f1b6d21d8e2698be165301c6bc1 155928Author: Alan Coopersmith <alan.coopersmith@sun.com> 155929Date: Tue Nov 20 18:27:12 2007 -0800 155930 155931 Enable use of /dev/urandom on Solaris as well 155932 155933commit 4363d70c6b420648b501126d1fbdebfafc7ae09f 155934Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155935Date: Tue Nov 20 18:58:55 2007 -0500 155936 155937 registry: Fix some mistakes in the reversion of prior commits. 155938 155939commit 140a4660aca1c283613d5b62f51668b44b45baf6 155940Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155941Date: Tue Nov 20 18:49:30 2007 -0500 155942 155943 Revert "registry: Register XTrap extension protocol names." 155944 155945 This reverts commit b38a91993364aa80cfd99721e319e1458d9fb760. 155946 155947 Moving all the names into dix/registry.c 155948 155949commit ed8a39c48ab9dac085fcf58b9641364b5608f3f4 155950Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155951Date: Tue Nov 20 18:47:52 2007 -0500 155952 155953 Revert "registry: Register XKB extension protocol names." 155954 155955 This reverts commit a5cf3f21f712e46dbf9bca289e67be75f2b531d3. 155956 155957 Moving all the names into dix/registry.c 155958 155959commit 17b0c729b553e2f0f8f82497698b282a47db3326 155960Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155961Date: Tue Nov 20 18:46:43 2007 -0500 155962 155963 registry: Remove registry code from XInput extension. 155964 155965 Moving all the names into dix/registry.c 155966 155967commit e86852aff62a861823b8e419434e0401b8cdc8e0 155968Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155969Date: Tue Nov 20 18:44:56 2007 -0500 155970 155971 Revert "registry: Register XFixes extension protocol names." 155972 155973 This reverts commit 106758893b68033f14f69c4ee6591fb6a149ba37. 155974 155975 Moving all the names into dix/registry.c 155976 155977commit 5269da2bde3cf4feb12fa2bd87bff6ee6d8730a1 155978Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155979Date: Tue Nov 20 18:43:38 2007 -0500 155980 155981 Revert "registry: Register XvMC extension protocol names." 155982 155983 This reverts commit 853ea337bdad17f8f6ec7d940de14ce2cbbbf93e. 155984 155985 Moving all the names into dix/registry.c 155986 155987commit 03a86c8d5e20a6e47f3c294f0087f205cf2a72dd 155988Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155989Date: Tue Nov 20 18:42:19 2007 -0500 155990 155991 Revert "registry: Register Xv extension protocol names." 155992 155993 This reverts commit 12766c5b5ffdab95255a63b2c8421ee773fd43b5. 155994 155995 Moving all the names into dix/registry.c 155996 155997commit edcf490cdb965e2a5bfc0169c01732d2924da3ae 155998Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 155999Date: Tue Nov 20 18:41:10 2007 -0500 156000 156001 registry: Remove registry code from XTest extension. 156002 156003 Moving all the names into dix/registry.c 156004 156005commit 5fea1ed50f37691a5273bf2897479781de808ff5 156006Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156007Date: Tue Nov 20 18:39:48 2007 -0500 156008 156009 registry: Remove registry code from SELinux extension. 156010 156011 Moving all the names into dix/registry.c 156012 156013commit 9a8af33718d085656a672e4c27df200485c84154 156014Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156015Date: Tue Nov 20 18:38:24 2007 -0500 156016 156017 Revert "registry: Register Resource extension protocol names." 156018 156019 This reverts commit 5c8b1a91726817816d20faefad21c7a68ab634cc. 156020 156021 Moving all the names into dix/registry.c 156022 156023commit e6023e0208fae8f19c566f9df1a8aa20494f40ab 156024Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156025Date: Tue Nov 20 18:36:49 2007 -0500 156026 156027 Revert "registry: Register XPrint extension protocol names." 156028 156029 This reverts commit f077578e42eee424b0e534774574c84af9d6f85b. 156030 156031 Moving all the names into dix/registry.c 156032 156033commit 277345fb7065d74c3b0d076382affb78cbe67569 156034Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156035Date: Tue Nov 20 18:35:57 2007 -0500 156036 156037 registry: Remove registry code from XF86Bigfont extension. 156038 156039 Moving all the names into dix/registry.c 156040 156041commit bf27edd365ffd275e5453f44d130eeacbfe0ecd9 156042Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156043Date: Tue Nov 20 18:34:14 2007 -0500 156044 156045 Revert "registry: Register EVIE extension protocol names." 156046 156047 This reverts commit 48891d5696f56711f23743cb03be39cf6b26c522. 156048 156049 Moving all the names into dix/registry.c 156050 156051commit 687427179420b18a55a1a02b8a9f2a32ea8eac8d 156052Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156053Date: Tue Nov 20 18:32:54 2007 -0500 156054 156055 registry: Remove registry code from XC-MISC extension. 156056 156057 Moving all the names into dix/registry.c 156058 156059commit 4b0274e8f712e51b18618a2a0bdbe03b17b9736b 156060Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156061Date: Tue Nov 20 18:25:15 2007 -0500 156062 156063 Revert "registry: Register SYNC extension protocol names." 156064 156065 This reverts commit 9f597f6c87e0b14cc382d8e5929e42f822db4329. 156066 156067 Moving all the names into dix/registry.c 156068 156069commit 4c7cf5aa4c802dcde895c723879a80a87620c0f7 156070Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156071Date: Tue Nov 20 18:23:57 2007 -0500 156072 156073 Revert "registry: Register SHM extension protocol names." 156074 156075 This reverts commit 2c9646ad4e65bb061d910c9e2b1a8a978f21fa17. 156076 156077 Moving all the names to dix/registry.c 156078 156079commit 67e82e306f67a215c6c89868cc1d3649747bd93d 156080Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156081Date: Tue Nov 20 18:22:59 2007 -0500 156082 156083 Revert "registry: Register SHAPE extension protocol names." 156084 156085 This reverts commit 4e274e90e16b1d954391e1af3e2074fb10f70ee7. 156086 156087 Moving all the names to dix/registry.c 156088 156089commit 8583bf78ad056ffe2d83b54e5c9a0a217e425a7b 156090Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156091Date: Tue Nov 20 18:21:09 2007 -0500 156092 156093 registry: Remove registry code from XC-SECURITY extension. 156094 156095 Moving all the names to dix/registry.c 156096 156097commit 55744d8e5d7bf1ff27cd25de54e14e799dd1a70a 156098Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156099Date: Tue Nov 20 18:19:44 2007 -0500 156100 156101 Revert "registry: Register MIT-SCREEN-SAVER extension protocol names." 156102 156103 This reverts commit 58c3240fcbec23aad122e1c340f6bb6d3b18f779. 156104 156105 Moving all the names into dix/registry.c 156106 156107commit 36ef45928c783292cef18acfdd83ae057826c989 156108Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156109Date: Tue Nov 20 18:18:01 2007 -0500 156110 156111 registry: Remove registry code from MIT-MISC extension. 156112 156113 Moving all the names to dix/registry.c 156114 156115commit 816e6e612e4bc3cea1e67e7ea79d5b640458011f 156116Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156117Date: Tue Nov 20 18:15:37 2007 -0500 156118 156119 Revert "registry: Register Multibuffer extension protocol names." 156120 156121 This reverts commit 3877faf7d9fe00ed634077e38a198ae4b91a2bb4. 156122 156123 Moving all the names into dix/registry.c 156124 156125commit 40a0da044e911ea51de003f3621331ffbe2842bc 156126Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156127Date: Tue Nov 20 18:13:43 2007 -0500 156128 156129 registry: Remove registry code from Fontcache extension. 156130 156131 Moving all the names into dix/registry.c 156132 156133commit 46412baf60ed639ddc1d5fb601f73a75e39737f7 156134Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156135Date: Tue Nov 20 18:11:06 2007 -0500 156136 156137 registry: Remove registry code from EVI extension. 156138 156139 Moving all the names into dix/registry.c 156140 156141commit 460c43032f05aad3f0f552901a52d199f61c7f4f 156142Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156143Date: Tue Nov 20 18:08:18 2007 -0500 156144 156145 registry: Remove registry code from DPMS extension. 156146 156147 Moving all the names into dix/registry.c 156148 156149commit 76e89d45b497d4afa4e60e1d0ec50b62f54f6b88 156150Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156151Date: Tue Nov 20 18:06:40 2007 -0500 156152 156153 registry: Remove registry code from TOG-CUP extension. 156154 156155 Moving all the names into dix/registry.c 156156 156157commit ce93c5772da52ab88faef7e5b661b681d5b60b1e 156158Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156159Date: Tue Nov 20 18:03:57 2007 -0500 156160 156161 registry: Remove registry code from BigRequests extension. 156162 156163 Moving all the names into dix/registry.c 156164 156165commit 0756d1271209e6ae14cc641dddca095271b43150 156166Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156167Date: Tue Nov 20 17:59:40 2007 -0500 156168 156169 Revert "registry: Register APPGROUP extension protocol names." 156170 156171 This reverts commit b504678ba5407a6fd8d47d051305f7c3d5606dfe. 156172 156173 Moving all the names into dix/registry.c 156174 156175commit 5aff37d1d69be493727856a29628bd782d50b90f 156176Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156177Date: Tue Nov 20 17:57:06 2007 -0500 156178 156179 Revert "registry: Register RENDER extension protocol names." 156180 156181 This reverts commit 8964c6d8e14ae47798762191e359b2bf138ca32e. 156182 156183 Moving all the names into dix/registry.c 156184 156185commit e585a2ddb495b50a53e15cccc368ca0858fc9d23 156186Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156187Date: Tue Nov 20 17:55:47 2007 -0500 156188 156189 Revert "registry: Register Record extension protocol names." 156190 156191 This reverts commit ea09c9acc8f0d5577f54c864ff88b7f03d93b2f4. 156192 156193 Moving all the names into dix/registry.c 156194 156195commit d4577e485367468227e031eb434b739eff7b5e9a 156196Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156197Date: Tue Nov 20 17:51:27 2007 -0500 156198 156199 Revert "registry: Register RANDR extension protocol names." 156200 156201 This reverts commit c827db57e4d9ca14c82b099dcfc9b7a0c0b5ba0a. 156202 156203 Moving all the names into dix/registry.c 156204 156205commit a541e826c9310d3051e53834833c6c3a08654148 156206Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156207Date: Tue Nov 20 17:50:26 2007 -0500 156208 156209 Revert "registry: Register WINDOWSWM extension protocol names." 156210 156211 This reverts commit 4c3285c883cc50a91bc5262bbc9d073d816f860a. 156212 156213 Moving all the names into dix/registry.c 156214 156215commit 993595430bd0580ab4d936be6b70fb91b8bb1d16 156216Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156217Date: Tue Nov 20 17:48:46 2007 -0500 156218 156219 Revert "registry: Register XF86DRI extension protocol names." 156220 156221 This reverts commit b7786724080fd3928ef7b8c294346661d7ffd90b. 156222 156223 Moving all the names into dix/registry.c 156224 156225commit 6b73c215c9d612534af290230b2e914d42d819cd 156226Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156227Date: Tue Nov 20 17:47:30 2007 -0500 156228 156229 Revert "registry: Register XF86VidMode extension protocol names." 156230 156231 This reverts commit 960677e876c068400fb45e1764bb5470cd8c389f. 156232 156233 Moving all the names into dix/registry.c 156234 156235commit 8e2cd7a804664bbd2d03789dcd5c93223122e929 156236Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156237Date: Tue Nov 20 17:45:30 2007 -0500 156238 156239 Revert "registry: Register XF86Misc extension protocol names." 156240 156241 This reverts commit 2cd1b32b77e0ceeaccb3f01c4ac13a97c557668c. 156242 156243 Moving all the names into dix/registry.c 156244 156245commit 0356153a58cef87d655bccacd8e2cf03d577bd19 156246Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156247Date: Tue Nov 20 17:43:18 2007 -0500 156248 156249 Revert "registry: Register XF86DGA extension protocol names." 156250 156251 This reverts commit 3815284e899b61731b6a63c4ba14c5d773e24eb6. 156252 156253 Moving all the names into dix/registry.c 156254 156255commit de93c1e9df14577e158b6dc3ccec7ee48f592386 156256Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156257Date: Tue Nov 20 17:40:57 2007 -0500 156258 156259 Revert "registry: Register DMX extension protocol names." 156260 156261 This reverts commit 32f3f5a1e7654f8bb43ea16b9227b3994e616739. 156262 156263 Moving all the names into dix/registry.c 156264 156265commit 2d3e0cdf4bd7ab069bad7244ede7c2d489e92b17 156266Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156267Date: Tue Nov 20 17:39:56 2007 -0500 156268 156269 Revert "registry: Register APPLEDRI extension protocol names." 156270 156271 This reverts commit 3464b419230c6d17e940d967b567c5d2cb22d232. 156272 156273 Moving all the names into dix/registry.c 156274 156275commit 546d46224e355d4f00232da5538548e3c8853e40 156276Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156277Date: Tue Nov 20 17:37:48 2007 -0500 156278 156279 Revert "registry: Register XINERAMA extension protocol names." 156280 156281 This reverts commit b9f5ab98c8dea36dcce1ad15fd2e059a77e77c39. 156282 156283 Moving all the names into dix/registry.c 156284 156285commit fd2d83d5bf5b35c8a2b05f725486be166783921e 156286Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156287Date: Tue Nov 20 17:34:48 2007 -0500 156288 156289 Revert "registry: Register APPLEWM extension protocol names." 156290 156291 This reverts commit eee46b4681ec55297604b0425705f2b18381f7ca. 156292 156293 Moving all the names into dix/registry.c 156294 156295commit c934e1af27189571c1e7dd838872e380c3580eeb 156296Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156297Date: Tue Nov 20 17:32:35 2007 -0500 156298 156299 Revert "registry: Register DBE extension protocol names." 156300 156301 This reverts commit 2e1e5be1d9067816525aa13a1d818e8ca6899599. 156302 156303 Moving all the names into dix/registry.c 156304 156305commit b9ab6f300a46aa8879b11eac51857357cc379c2f 156306Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156307Date: Tue Nov 20 17:31:00 2007 -0500 156308 156309 Revert "registry: Register DAMAGE extension protocol names." 156310 156311 This reverts commit 20db50b4c44a14f7eeac2b1de17ada68482521da. 156312 156313 Moving all the names into dix/registry.c 156314 156315commit 26586a7ad5e999b34996d147fb43998deea89178 156316Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156317Date: Tue Nov 20 17:27:37 2007 -0500 156318 156319 Revert "registry: Register composite extension protocol names." 156320 156321 This reverts commit 166ef972febc00c665e1d5aeb68e75d7bbcf9879. 156322 156323 Moving all the names into dix/registry.c 156324 156325commit fae39db7957c0ebdc7af36f8d8f484473beb6d38 156326Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156327Date: Tue Nov 20 15:21:31 2007 -0500 156328 156329 devPrivates rework: put back some changes that were mistakenly removed 156330 during merge conflict resolution. 156331 156332commit 709c1a70c8c6a9e132bf0d92f78a12be72beee51 156333Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156334Date: Tue Nov 20 15:18:02 2007 -0500 156335 156336 Remove some duplicate include statements. 156337 156338commit be0cbe5a330f62cef47fffbc49e83b5e1637b7d0 156339Author: Dodji Seketeli <dodji@openedhand.com> 156340Date: Tue Nov 20 15:39:49 2007 +0100 156341 156342 kaa: update kaaCreatePixmap to support the new usage_int 156343 156344commit 5b0dfb73ca4699cc4b33720f10416de7440081b7 156345Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156346Date: Mon Nov 19 21:01:22 2007 -0500 156347 156348 devPrivates rework: put back a comment that was mistakenly removed 156349 during merge conflict resolution. 156350 156351commit 2d17f47cc7d6b174857617d31ad1b437d8e97d94 156352Merge: 60be452c2 ea9c63e93 156353Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156354Date: Mon Nov 19 18:10:46 2007 -0500 156355 156356 Merge branch 'master' into XACE-SELINUX 156357 156358 Conflicts: 156359 156360 hw/xnest/Pixmap.c 156361 include/dix.h 156362 156363commit 60be452c2e88342f92a76ba5ec7d90b5b0211aaf 156364Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156365Date: Mon Nov 19 16:55:09 2007 -0500 156366 156367 xace: restore the old SaveScreens function and introduce new API, since the 156368 old version is called from drivers... 156369 156370commit a95bb52b4366d85fc049130c60af5c9e727c565b 156371Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156372Date: Mon Nov 19 16:34:38 2007 -0500 156373 156374 devPrivates rework: add missing include of dix/privates.h 156375 156376commit ea9c63e93b9bb731796e8a8de2d127e6cc720076 156377Author: Michel Dänzer <michel@tungstengraphics.com> 156378Date: Mon Nov 19 15:53:49 2007 +0100 156379 156380 DEFAULT_DPI was undefined here. 156381 156382commit db9ae863536fff80b5463d99e71dc47ae587980d 156383Author: Adam Jackson <ajax@aspartame.nwnk.net> 156384Date: Sun Nov 18 11:57:01 2007 -0500 156385 156386 Bump DEFAULT_DPI to 96. 156387 156388 75 is just nonsense. 156389 156390commit ee2d4626dca6e0d4fc6f524e5de4bdefa2ed43df 156391Author: Tormod Volden <bugzi06.fdo.tormod@xoxy.net> 156392Date: Sun Nov 18 11:56:31 2007 -0500 156393 156394 Bug #12932: Use DEFAULT_DPI in randr1.2 instead of hardcoded 96. 156395 156396commit a46c30c3be33ffb304a885503c8aaa78396ed3d9 156397Author: Jernej Azarija <azi.stdout@gmail.com> 156398Date: Sun Nov 18 11:44:36 2007 -0500 156399 156400 Bug #12531: RRModesForScreen can fail to allocate. 156401 156402commit fac7e7e4e1809e865b9b3cf5b7eb69ba9d3a3759 156403Author: Adam Jackson <ajax@aspartame.nwnk.net> 156404Date: Sun Nov 18 11:39:26 2007 -0500 156405 156406 Document the requirement for interleaved code and declarations. 156407 156408commit bad96e5a864e40fbd47265d7fb6eaa67c55fac11 156409Author: Peter Hutterer <peter@cs.unisa.edu.au> 156410Date: Sun Nov 18 15:02:24 2007 +1030 156411 156412 dix: getevents.c remove trailing whitespaces. 156413 156414commit 75e5e9f88c5f974d996dd81e672cd3709d361a44 156415Author: Peter Hutterer <peter@cs.unisa.edu.au> 156416Date: Sat Nov 17 13:46:01 2007 +1030 156417 156418 Xi: add missing XI_DeviceClassesChangedMask to XI filters array. 156419 156420commit d15339a92c4d689d2ab8a86e4f10107f3e45eff8 156421Author: Adam Jackson <ajax@aspartame.nwnk.net> 156422Date: Sat Nov 17 22:12:10 2007 -0500 156423 156424 Bug #9725: Don't look in root's $HOME for config files, that's just confusing. 156425 156426commit 748cfbc820f8cdeb544c54a6db495fecf2e2457b 156427Author: Adam Jackson <ajax@aspartame.nwnk.net> 156428Date: Sat Nov 17 21:23:05 2007 -0500 156429 156430 Disinfect mi/ of mfb. 156431 156432commit a969db091cab16a448f82782e85b3dd19c81627a 156433Author: Daniel Stone <daniel@fooishbar.org> 156434Date: Sat Nov 17 22:34:47 2007 +0100 156435 156436 XKB: Don't ring the bell when we don't have a BellProc (bug #13246) 156437 156438commit c89b543198d5ec56ff025bdd6bb7229523478e58 156439Author: Ben Skeggs <skeggsb@gmail.com> 156440Date: Sat Nov 17 18:20:49 2007 +1000 156441 156442 exa: set driverPriv to NULL before it might get used later with garbage 156443 156444commit 514ba4ca727f0b1076bc67500617722203d34daa 156445Author: Adam Jackson <ajax@aspartame.nwnk.net> 156446Date: Fri Nov 16 19:53:11 2007 -0500 156447 156448 Bug #1612: Use a stronger PRNG. 156449 156450 Currently just reads from /dev/urandom, and only on Linux. 156451 156452commit 12dd6e9911de187e0ebee86434e8c24a67c990f9 156453Author: Peter Hutterer <peter@cs.unisa.edu.au> 156454Date: Fri Nov 16 17:23:09 2007 +1030 156455 156456 dix: reset MD's devPrivate classes to NULL before device initialisation. 156457 156458 XkbInitKeyboardDefviceStruct may call FatalError if it fails. FatalError then 156459 cleans up all the devices, resulting in a segfault if the pointer is 156460 uninitialised. 156461 156462commit be3321c2e9fad228a9ee5fef47680a47bc9e39a4 156463Author: Peter Hutterer <peter@cs.unisa.edu.au> 156464Date: Fri Nov 16 12:12:41 2007 +1030 156465 156466 dix: Free both current classes and original classes when closing an MD. 156467 156468commit 3c39dd19ec2a53b8854279e3b03131098031473a 156469Author: Peter Hutterer <peter@cs.unisa.edu.au> 156470Date: Fri Nov 16 12:12:14 2007 +1030 156471 156472 Xi: Deep-copy full list of FeedbackClasses. 156473 156474 All feedback classes are linked lists and the whole list has to be duplicated, 156475 not just the first entry. 156476 156477 Xkb stuff still missing. 156478 156479commit 497862df2fcd67531fbe0f876c20a09884ee74df 156480Author: Peter Hutterer <peter@cs.unisa.edu.au> 156481Date: Fri Nov 16 11:20:22 2007 +1030 156482 156483 dix: explicitly float all attached SDs before closing down devices. 156484 156485 Some drivers flush on shutdown, if our SD is still attached we'd be trying to 156486 route an event through a non-existing device. 156487 156488commit 9de1ebe2a80164507cbe2ef688f284225e0ec808 156489Author: Peter Hutterer <peter@cs.unisa.edu.au> 156490Date: Fri Nov 16 10:45:28 2007 +1030 156491 156492 dix: Fix up class restoring when last SD disconnects. 156493 156494 Old code was fundamentally broken, fixes now are: 156495 - free the MDs current device classes 156496 - copy the device classes instead of flipping the pointers 156497 - check for the old MD, not the new one. 156498 156499commit 20fd4783247b1b93d9675dc36768dd1ed59ba2d3 156500Author: Adam Jackson <ajax@redhat.com> 156501Date: Thu Nov 15 17:01:33 2007 -0500 156502 156503 Small static cleanups on dix/ 156504 156505commit 70e50fa51f05663f289eeeea4521e737e8e24bca 156506Author: Adam Jackson <ajax@redhat.com> 156507Date: Thu Nov 15 17:00:37 2007 -0500 156508 156509 Allocate RRCrtcRecs with calloc. 156510 156511commit 8d0cd1cd2c57ee5a2fc4d577d8182d66369f0617 156512Author: Aaron Plattner <aplattner@nvidia.com> 156513Date: Thu Nov 15 12:16:36 2007 -0800 156514 156515 Fix a really dumb typo. 156516 156517commit f797c96845a3fab37cda6839ebecf9ac5401fd6e 156518Author: Aaron Plattner <aplattner@nvidia.com> 156519Date: Thu Nov 15 12:12:02 2007 -0800 156520 156521 Save pixmap allocation hints into the PixmapRec. 156522 156523commit 6bc50de02108f822977fc7545da81fce95ea7ff4 156524Author: Adam Jackson <ajax@redhat.com> 156525Date: Thu Nov 15 15:03:58 2007 -0500 156526 156527 Simplify system resource range setup. 156528 156529 osRes only existed to get copied into Acc. Waste of effort. 156530 156531commit e1ff14a9246e12d42ce8ca5afbe3b957333a5620 156532Author: Adam Jackson <ajax@redhat.com> 156533Date: Thu Nov 15 14:45:49 2007 -0500 156534 156535 Delete some dead code in X -configure. 156536 156537commit 01cfba75229f4b9bf1e4fe80814931acdacde14c 156538Author: Adam Jackson <ajax@redhat.com> 156539Date: Thu Nov 15 14:27:03 2007 -0500 156540 156541 Nuke the debugging allocator. 156542 156543 This has never been hooked up in the modular build, and can not possibly 156544 have built since before 6.7. Clearly no one's using it. 156545 156546commit c67b9c5fc33002b13a2360929a37f24169710f64 156547Author: Adam Jackson <ajax@redhat.com> 156548Date: Thu Nov 15 14:24:36 2007 -0500 156549 156550 Clean up some garbage in driver enumeration. 156551 156552commit 0706e5e790060fbf046cfaff295b78806b7841c6 156553Author: Adam Jackson <ajax@redhat.com> 156554Date: Thu Nov 15 14:15:09 2007 -0500 156555 156556 Eliminate some redundancy in autoconfiguration. 156557 156558 We already synthesize Monitor and Module sections for you, no need to 156559 specify them explicitly in the fake config buffer. 156560 156561commit 2c01a49bf0a407bd5510bb9ceb4ef86a2cc36be9 156562Author: Adam Jackson <ajax@redhat.com> 156563Date: Thu Nov 15 13:32:59 2007 -0500 156564 156565 Don't sleep(1) at exit on any platform. 156566 156567commit 83926cb8bef6288b89e801c5e60b3f40e923e16e 156568Author: Peter Hutterer <peter@cs.unisa.edu.au> 156569Date: Thu Nov 15 17:41:07 2007 +1030 156570 156571 Xi: Only alloc modifierKeyMap if <slave device>->maxKeysPerModifier > 0 156572 156573 Sometimes (e.g. on my debian ppc box) maxKeysPerModifier of the SD is 0. So we 156574 try to malloc(0), bringing the whole server down with a FatalError because it 156575 looks as if the malloc failed. This is bad, so only alloc if we actually have 156576 something to alloc. 156577 156578commit 070195dbf88eb121e65f802e023aa37ed1f2c2ac 156579Author: Peter Hutterer <peter@cs.unisa.edu.au> 156580Date: Thu Nov 15 16:27:18 2007 +1030 156581 156582 Xi: fix up sloppy class copying causing segfaults. 156583 156584commit 53539688cab990a7df1851d64f3ee4e11920a86b 156585Author: Peter Hutterer <peter@cs.unisa.edu.au> 156586Date: Thu Nov 15 16:23:48 2007 +1030 156587 156588 dix: SetKeySymMap should alloc the map if dst doesn't have one already. 156589 156590commit b40646dc104fb03ea7cc0b27fae573aecaab486e 156591Author: Peter Hutterer <peter@cs.unisa.edu.au> 156592Date: Thu Nov 15 15:43:44 2007 +1030 156593 156594 dix: Add FreeDeviceClass and FreeFeedbackClass for centralised xfree. 156595 156596 Ensures that we only have one way of freeing a device class to avoid leaks in 156597 ChangeMasterDeviceClasses and other places. 156598 156599commit 4c9cc82fc4461d180ae2c2fbe50e7f98b0777f91 156600Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 156601Date: Thu Nov 15 01:46:11 2007 -0200 156602 156603 For some reason "-nozap" appeared twice. Weird. 156604 156605commit 18833d648fd7e1a5e962b93636bbbb38aca9c454 156606Author: Peter Hutterer <peter@cs.unisa.edu.au> 156607Date: Thu Nov 15 12:13:59 2007 +1030 156608 156609 Xi: reset xkb-stuff to NULL after copying (DeepCopyDeviceClasses) 156610 156611 Having two devices point to the same xkb stuff causes SIGABRTs. 156612 156613 Also, don't init a MD's xkbInfo unless the SD has an xkbInfo. 156614 156615commit 1635832c1635374033686d3a943b77adbd60bb98 156616Author: Peter Hutterer <peter@cs.unisa.edu.au> 156617Date: Thu Nov 15 11:35:07 2007 +1030 156618 156619 Revert "xkb: disable xkb key repeats (temporarily)" 156620 156621 This reverts commit 2b1d946392ce28b96941341778b2b526aa0fb126. 156622 156623commit b05246696d14bd35aa53b49302707b51206c72a6 156624Author: Peter Hutterer <peter@cs.unisa.edu.au> 156625Date: Thu Nov 15 11:31:43 2007 +1030 156626 156627 Xi: free XkbSrvLedInfos as well when freeing an MD's device classes. 156628 156629commit c758e5a664a52045ad419340044beebb6774a336 156630Author: Peter Hutterer <peter@cs.unisa.edu.au> 156631Date: Thu Nov 15 10:47:20 2007 +1030 156632 156633 dix: Make sure core MDs copy their initial classes before they are used. 156634 156635 Anything in dev->key, dev->valuator etc. of a MD must always be a copy of the 156636 original class. The intial classes of an MD (the ones set up before an SD is 156637 attached) as well, as we may have to restore them if no SD is attached 156638 anymore. 156639 156640commit a08665d4d3b0a7d567a90bb9bbfe4abafd6f3887 156641Author: Peter Hutterer <peter@cs.unisa.edu.au> 156642Date: Thu Nov 15 10:42:29 2007 +1030 156643 156644 mi: don't call SwitchCoreKeyboard, we switch during event processing now. 156645 156646commit 64711a094893e83764bbeda538c6e877ebe2af79 156647Author: Peter Hutterer <peter@cs.unisa.edu.au> 156648Date: Thu Nov 15 10:41:34 2007 +1030 156649 156650 Xi: When switching MD classes, make a deep copy instead of pointer flip. 156651 156652 Turns out it's really really hard synchronising device state across multiple 156653 duplicated events if they all share the same struct. So instead of doing so, 156654 when the SD changes deep-copy all it's classes into the MD. The MD then has 156655 the same capabilities, but the state can be set separately. This should fix 156656 xkb, key state, repeat etc. problems. 156657 156658 Updating the device state allows us to remove the SwitchCoreKeyboard from the 156659 event gathering, it's all done during event processing now. 156660 156661commit 3dde66f96b9b8431381871cf85266da3ec57a0d4 156662Author: Adam Jackson <ajax@redhat.com> 156663Date: Wed Nov 14 15:10:59 2007 -0500 156664 156665 Start 1.4.99 156666 156667commit c3897ca7099fc007b4134a8fabd4c707f99f2ac7 156668Author: Adam Jackson <ajax@redhat.com> 156669Date: Fri Nov 9 13:55:32 2007 -0500 156670 156671 Add -pogo option for init/teardown performance testing. 156672 156673commit cecac794451b793871f297b91a11d3b52eeb6d1b 156674Author: Adam Jackson <ajax@redhat.com> 156675Date: Thu Nov 8 17:25:36 2007 -0500 156676 156677 Don't sleep(1) at server exit. 156678 156679commit 1603130236c55ddabc3854d4ba62d544debcf1f5 156680Merge: f207e69d6 f7dd0c72b 156681Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156682Date: Wed Nov 14 13:35:50 2007 -0500 156683 156684 Merge branch 'master' into XACE-SELINUX 156685 156686 Conflicts: 156687 156688 Xext/xace.c 156689 Xext/xace.h 156690 156691commit f207e69d62bc04c7f254347b03e6d8fa8b569d66 156692Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156693Date: Wed Nov 14 12:23:29 2007 -0500 156694 156695 xselinux: adjust receive hook to use new synthetic_event class. 156696 156697commit cc4586df600af571815d5cdda2028c0d074c8469 156698Author: Peter Hutterer <peter@cs.unisa.edu.au> 156699Date: Wed Nov 14 14:27:52 2007 +1030 156700 156701 Xi: toggle the public.devPrivate as well when switching device classes. 156702 156703 The master needs to have the same devPrivate as the slave, in case a client 156704 issues a request that goes down to the driver. 156705 Example: if a driver wants to ring the keyboard bell, it'll pick a keyboard. 156706 The KeyClassPtr will direct it to the matching method in the driver, but 156707 because the MD doesn't have the devPrivate set the driver segfaults. 156708 Even if all drivers were updated to not dereference the nullpointer, nothing 156709 would actually ever happen. 156710 156711 To avoid this, we flip the master's public.devPrivate to the last SDs 156712 devPrivate. 156713 156714commit b44c1118f3bab6d5f28fa42e0c322fbaec005012 156715Author: Peter Hutterer <peter@cs.unisa.edu.au> 156716Date: Wed Nov 14 14:08:21 2007 +1030 156717 156718 dix: Return Success from ProcBell if no BellProc is set. 156719 156720 We must not return BadDevice, this causes applications to crash. If no 156721 BellProc is set, just quietly do nothing and report a Success. 156722 156723commit 240b10fa9c32510a380a73a8acdd8267c81e538b 156724Author: Peter Hutterer <peter@cs.unisa.edu.au> 156725Date: Wed Nov 14 12:18:08 2007 +1030 156726 156727 Xi: Move updating the device state from POE into separate function. 156728 156729 POE now only deals with processing the event and calling the appropriate 156730 delivery methods. Actually modifying the device state is done in 156731 UpdateDeviceState. This separation should make it easier to avoid setting the 156732 state twice when master events are processed. 156733 156734commit 0bdfeb55fc559ffe0334df69ba6e9e711b26dc18 156735Author: Peter Hutterer <peter@cs.unisa.edu.au> 156736Date: Tue Nov 13 17:31:20 2007 +1030 156737 156738 Xi: remove RegisterPairingClient handling. Deprecated with device hierarchy. 156739 156740commit e96d926d64f7fb63f6bf2aa0ea0a8440a188947b 156741Author: Peter Hutterer <peter@cs.unisa.edu.au> 156742Date: Tue Nov 13 17:14:35 2007 +1030 156743 156744 Xi: remove trailing whitespaces. 156745 156746commit 5031238aad2b6b7511aab0f9d15edfbdd2b4cce7 156747Author: Peter Hutterer <peter@cs.unisa.edu.au> 156748Date: Tue Nov 13 16:58:23 2007 +1030 156749 156750 dix: remove trailing/wrong whitespaces from devices.c and events.c 156751 156752commit c703849e79391bb1bb01a994bb8d5cf2eb64e48b 156753Author: Peter Hutterer <peter@cs.unisa.edu.au> 156754Date: Tue Nov 13 16:50:18 2007 +1030 156755 156756 dix: Attach spriteInfoRec to same memory block as the DeviceIntRec. 156757 156758 No need to alloc a separate one, we never do anything special with it anyway. 156759 156760commit 51239f87ce42ad564ceee1761980391947294511 156761Author: Peter Hutterer <peter@cs.unisa.edu.au> 156762Date: Tue Nov 13 11:26:16 2007 +1030 156763 156764 dix: Send MappingNotify when keyboard maps change. 156765 156766 If a slave device is attached to a master device, then we need to send a 156767 mapping notify to the client. 156768 Mapping notify needs to be sent if 156769 - different slave device but on same master 156770 - different master 156771 156772 This gives you funny behaviour with the ClientPointer. When a 156773 MappingNotify is sent to the client, the client usually responds with a 156774 GetKeyboardMapping. This will retrieve the ClientPointer's keyboard mapping, 156775 regardless of which keyboard sent the last mapping notify request. So 156776 depending on the CP setting, your keyboard may change layout in each app... 156777 156778commit 2b1d946392ce28b96941341778b2b526aa0fb126 156779Author: Peter Hutterer <peter@cs.unisa.edu.au> 156780Date: Tue Nov 13 09:51:33 2007 +1030 156781 156782 xkb: disable xkb key repeats (temporarily) 156783 156784 Haven't quite figured out yet how to make these repeats work. Because we share 156785 the class between devices, the key state is already set when we process the 156786 master device's event, causing a repeat on each event. 156787 156788commit f7dd0c72b8f861f4d5443a43d1013e3fe3db43ca 156789Author: Matthias Hopf <mhopf@suse.de> 156790Date: Mon Nov 12 15:11:03 2007 +0100 156791 156792 Only clear crtc of output if it is the one we're actually working on. 156793 156794 Upon recreation of the RandR internal data structures in RRCrtcNotify() the 156795 crtc of an output could be NULLed if the crtc was shared (cloned) between two 156796 outputs and one of them got another crtc assigned. 156797 156798commit 70b4087c4dd1904d9d655f4afb9dfcea4f137f7a 156799Author: Peter Hutterer <peter@cs.unisa.edu.au> 156800Date: Mon Nov 12 13:10:39 2007 +1030 156801 156802 dix: don't unconditionally update valuators during key events. 156803 156804 Master may not have valuators, device may not have valuators. 156805 156806commit 23365d28651f7942fdafb889bcbbd019470a4274 156807Author: Peter Hutterer <peter@cs.unisa.edu.au> 156808Date: Mon Nov 12 13:08:38 2007 +1030 156809 156810 dix: allow grab modifier device to be NULL. 156811 156812 This can happen if we check for a passive core grab and our device is a 156813 floating slave device. Doesn't really change anything as SDs can't send core 156814 events but it stops the server from segfaulting. 156815 156816commit 5a7a65a3c978a65e8ff39d0cc9878527ec42adc9 156817Author: Peter Hutterer <peter@cs.unisa.edu.au> 156818Date: Mon Nov 12 12:37:39 2007 +1030 156819 156820 mi: avoid SIGABRT by setting master_event to NULL. 156821 156822commit a05f43bf3e9629df98e93c366d4327f20ed81e6c 156823Author: Peter Hutterer <peter@cs.unisa.edu.au> 156824Date: Mon Nov 12 11:35:18 2007 +1030 156825 156826 dix: When the last slave is removed, set master to the original classes. 156827 156828 DeviceClassesChangedEvent is sent to the client, where device == new slave. 156829 156830commit 45f884d79c0eebaa1eb24d7db76c1177f6b710c9 156831Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156832Date: Fri Nov 9 14:45:27 2007 -0500 156833 156834 xselinux: add new synthetic_event security class, and fix registry code. 156835 156836commit b092856baba5bd43b23950f23236b5cc3ce78c1e 156837Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156838Date: Fri Nov 9 14:45:02 2007 -0500 156839 156840 registry: Register XC-SECURITY extension protocol names. 156841 156842commit 7a81bafc9bc7048560b17483e6addf58469a05d0 156843Author: Peter Hutterer <peter@cs.unisa.edu.au> 156844Date: Fri Nov 9 23:10:24 2007 +1030 156845 156846 Xi, dix: Add ability to change MD classes + send event when doing so. 156847 156848 Each time a different slave device sends through a master, an 156849 DeviceClassesChangedEvent is enqueued. When this event is processed, all 156850 classes of the matching master device are changed, and the event is sent to 156851 the clients. 156852 156853 Next time the master is queried, it thus shows the evclasses of the last slave 156854 device. The original classes are stored in the devPrivates. 156855 156856 TODO: if all slave devices are removed, the master's original classes need to 156857 be restored. 156858 156859commit c0a05805783ee3d38fbcc0fb45f4aa3c511785f0 156860Author: Peter Hutterer <peter@cs.unisa.edu.au> 156861Date: Fri Nov 9 23:07:10 2007 +1030 156862 156863 Xext: add simple macro to easily check a generic event's type. 156864 156865commit f48087b6c33c1f84bf2cfc0744b1c38697321c07 156866Author: Kristian Høgsberg <krh@temari.boston.redhat.com> 156867Date: Fri Nov 9 05:49:26 2007 -0500 156868 156869 Regenerate GLX dispatch code for recent gl_API.xml changes (#12935). 156870 156871commit f2a3728868376a3646832d4af3a29549ce0b8f5d 156872Author: Ben Byer <bbyer@bbyer.apple.com> 156873Date: Thu Nov 8 18:49:05 2007 -0800 156874 156875 Patch to rootless code that should fix many crashes. Credit to Ken Thomases 156876 at CodeWeavers for the patch. From his description: 156877 156878 Fix a display bug with the X server. The Generic Rootless extension 156879 installs overrides for certain GC (graphics context) operations. Within 156880 these overrides, they temporarily uninstall themselves, perform their work, 156881 and then reinstall themselves. Except sometimes they would return early 156882 and wouldn't reinstall themselves when they should. Now they do in all cases. 156883 156884 Fix a bug in RootlessCopyWindow where early returns could leave the screen's 156885 dispatch table entry for CopyWindow unwrapped. We think that this is 156886 another case (hopefully the last) of the rootless drawing bug. 156887 156888commit 338c1aedbdf3964e542947140f7c50d58542cf12 156889Author: Ben Byer <bbyer@bbyer.local> 156890Date: Wed Nov 7 03:56:44 2007 -0800 156891 156892 formatting fixes 156893 156894commit bd269d0d783d418ef99363478fdf849fd89eed76 156895Author: Ben Byer <bbyer@bbyer.local> 156896Date: Wed Nov 7 03:55:08 2007 -0800 156897 156898 Fix for off-by-one error in menu bar height calculation -- props to Nicholas Riley! 156899 156900commit b4d14484056e6f4a7374fc1acf3f223be4bd116f 156901Author: Ben Byer <bbyer@bbyer.local> 156902Date: Wed Nov 7 03:10:52 2007 -0800 156903 156904 Undo some last-minute breakage in xpr.h 156905 156906commit 05d5b9baa05a4ba14a4383d8a981bc327d99290c 156907Author: Ben Byer <bbyer@bbyer.local> 156908Date: Wed Nov 7 02:28:49 2007 -0800 156909 156910 removed debugging output 156911 156912commit 9a8abcfa6d6d0cdc17be02a3443a7e116eb07d07 156913Author: Ben Byer <bbyer@bbyer.local> 156914Date: Wed Nov 7 02:22:39 2007 -0800 156915 156916 Fixed focus problem (clicking on an X11 window that sits behind 156917 an Aqua window would not always bring it to the top of the stack. 156918 156919commit b34d2ffc38002f7c4980c138f57e9a828cd79c37 156920Author: Ben Byer <bbyer@bbyer.local> 156921Date: Wed Nov 7 01:56:37 2007 -0800 156922 156923 formatting changes. 156924 156925commit 50dac9b2cb3b40810fb79253adc0265a838a497b 156926Author: Ben Byer <bbyer@bbyer.local> 156927Date: Wed Nov 7 01:35:48 2007 -0800 156928 156929 Fixed Spaces issue, correctly -- dragging an X window from one 156930 Space to another in Expose mode now works. 156931 156932commit ce7cfbe261b7fd4fcd09d1a4a61344d1555a71f2 156933Author: Ben Byer <bbyer@bbyer.apple.com> 156934Date: Thu Nov 8 20:10:51 2007 -0800 156935 156936 Fixed check to refer to DarwinApp, not all Darwin targets 156937 156938commit d68bd5510437c1fd3850e020f7cd90901fae8e1b 156939Author: Ben Byer <bbyer@bbyer.apple.com> 156940Date: Thu Nov 8 20:08:49 2007 -0800 156941 156942 fixing GLX in Xquartz 156943 156944commit a6ac9002956767fefa37aac95513e21ac5246d15 156945Author: Ben Byer <bbyer@bbyer.apple.com> 156946Date: Mon Nov 5 20:25:10 2007 -0800 156947 156948 formatting cleanup 156949 156950commit 67e96be13cdb45be31db121ce216295cd9496d20 156951Author: Ben Byer <bbyer@bbyer.apple.com> 156952Date: Mon Nov 5 20:01:34 2007 -0800 156953 156954 Fixed logic error that prevent JIS (Japanese) keyboard layouts from 156955 being detected. 156956 156957commit 154fb6417e5d0bae5191984beac5ce045ce754bb 156958Author: Ben Byer <bbyer@bbyer.local> 156959Date: Sat Nov 3 05:34:19 2007 -0700 156960 156961 Initial support for Spaces -- if you use Expose to drag an X11 156962 window to another Space, it will work correctly (as opposed 156963 to just leaving a ghost window). We accomplish this by listening 156964 for the notification from Xplugin that our window has been moved, 156965 and then we ask X11 to move the window to the new location. 156966 156967commit 5bbc468b702f62d7c91d41aabcc27eeb553f6959 156968Author: Peter Hutterer <peter@cs.unisa.edu.au> 156969Date: Fri Nov 9 11:33:27 2007 +1030 156970 156971 dix: grabbing an attached SD sets it floating for the duration of the grab. 156972 156973commit 676b26ca3e9b142cf007af3f439aa1993f2247c4 156974Author: Peter Hutterer <peter@cs.unisa.edu.au> 156975Date: Thu Nov 8 17:54:17 2007 +1030 156976 156977 Xi: notify the clients when the device hierarchy has been changed. 156978 156979commit cb75f09146a3c17b1a67b636bbf7229c65c83b35 156980Author: Peter Hutterer <peter@cs.unisa.edu.au> 156981Date: Fri Nov 9 11:29:18 2007 +1030 156982 156983 dix: calloc root window's Generic Event mask, stops segfaults. 156984 156985commit 169f83e366f678ac5441ad21beb84c9b8c65d28e 156986Author: Ben Byer <bbyer@bbyer.local> 156987Date: Sun Nov 4 19:14:10 2007 -0800 156988 156989 Disable deferred updates in xp_init to fix performance problems 156990 -- thanks to Eric Gouriou for pointing out the issue 156991 156992commit 1c6cb353f77747c101ce47716ff1fa055fbf85a4 156993Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 156994Date: Thu Nov 8 16:46:49 2007 -0500 156995 156996 Restore the XC-SECURITY option in configure.ac, but disabled by default. 156997 156998commit 9d03cad1446c27b397c198cf6247e71e46bc9e6d 156999Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157000Date: Thu Nov 8 16:41:47 2007 -0500 157001 157002 Remove SecurityPolicy file and associated references in the manpages. 157003 157004commit 8b5d21cc1d1f4e9d20e5d5eca44cb1e60a419763 157005Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157006Date: Thu Nov 8 16:32:42 2007 -0500 157007 157008 Rework of the XC-SECURITY extension. The gen-auth protocol has not changed, 157009 but the XC-QUERY-SECURITY-1 authorization method and the SecurityPolicy 157010 configuration file have been removed. The semantics of the trusted vs. 157011 untrusted split have been changed. This will be documented in a future 157012 commit. 157013 157014commit 476a9d85f819f454a6901ccb7eb028d1c563c341 157015Author: Dodji Seketeli <dodji@openedhand.com> 157016Date: Thu Nov 8 09:11:05 2007 +0100 157017 157018 Xephyr: do not AM_CONDITIONAL inside a shell if branch 157019 157020commit 59b304d8a24fecd094296feb217823f0c73d6f82 157021Author: Peter Hutterer <peter@cs.unisa.edu.au> 157022Date: Thu Nov 8 15:44:18 2007 +1030 157023 157024 dix, mi: stop segfaults when a floating device emits events. 157025 157026 For pointers: don't try to set master->valuator fields if there is no master. 157027 For keyboards: check if device is valid before trying to access the fields in 157028 miPointerGetScreen (btw. this disables DGA events for floating keyboards). 157029 157030 Also stop the hideous number of ErrorFs if we request the paired device for a 157031 floating dev. 157032 157033commit 3063f0c6679bdbea13f360cff8d09a88b9871da9 157034Author: Peter Hutterer <peter@cs.unisa.edu.au> 157035Date: Thu Nov 8 11:51:03 2007 +1030 157036 157037 mi: change the device id to the master devices' when duplicating the event. 157038 157039 Move the event duplication into a CopyGetMasterEvent(), makes the code a bit 157040 better to read. 157041 157042commit 0e9ef65fa583bf2393dd0fda82df6f092387b425 157043Author: Keith Packard <keithp@koto.keithp.com> 157044Date: Wed Nov 7 16:33:10 2007 -0800 157045 157046 Don't frob timers unless SmartSchedule is running 157047 157048commit 180220f284823c486e7001ef00f1279dc548c9c7 157049Author: Peter Hutterer <peter@cs.unisa.edu.au> 157050Date: Thu Nov 8 10:00:35 2007 +1030 157051 157052 Xi: return BadDevice for slave devices on ProcXSetClientPointer 157053 157054 Also clean up to return error codes instead of sending the error manually. 157055 157056commit 184a7b8917a15bb2c719153b9b016c03aab42101 157057Merge: a8808ac3d 0b729051c 157058Author: Peter Hutterer <peter@cs.unisa.edu.au> 157059Date: Thu Nov 8 09:34:35 2007 +1030 157060 157061 Merge branch 'mpx' into mdsd 157062 157063 Conflicts: 157064 157065 Xi/opendev.c 157066 157067commit e94ab0b502160376d1956345196f7e1c7e2e886f 157068Merge: 26e1fc7b4 9bee1c691 157069Author: Dodji Seketeli <dodji@openedhand.com> 157070Date: Wed Nov 7 19:04:24 2007 +0100 157071 157072 sync with 'master' 157073 157074commit 9bee1c6912817f65bbb8cf4078f0ad016d9d51cb 157075Author: Michel Dänzer <michel@tungstengraphics.com> 157076Date: Wed Nov 7 18:56:45 2007 +0100 157077 157078 EXA: Disable problematic optimization of dest pixmap migration by default. 157079 157080 Also add some code comments about these optimizations. 157081 157082commit 26e1fc7b42de850d69fba89703ffddd36480b997 157083Author: Dodji Seketeli <dodji@openedhand.com> 157084Date: Wed Nov 7 18:48:23 2007 +0100 157085 157086 Xephyr: don't use Xv or GL when those are disabled. 157087 157088commit 950f9995d11aff2c51139b34fb27eba594f2bd20 157089Author: Dodji Seketeli <dodji@openedhand.com> 157090Date: Wed Nov 7 18:43:16 2007 +0100 157091 157092 Xnest: fix lib dependancy to make libtool happy 157093 157094 This fixes an undefined symbol error happening when compiling 157095 the server with the --disable-xv configure switch. 157096 Basically, xnest was linking against 157097 @XSERVER_LIBS@ and @XNEST_LIBS@ and the order of the libraries 157098 given to the linker at the end of the process was bogus. 157099 157100 * configure.ac: make XNEST_LIBS contain the $XSERVER_LIBS re-ordered 157101 in such a way that the linker finds the symbols of all the libs contained 157102 in $XNEST_LIBS. 157103 * hw/xnest/Makefile.am: don't link against @XSERVER_LIBS@ anymore because 157104 XNEST_LIBS contains the right thing. 157105 157106commit 0b729051c04da7068f1e6dd319190bd0a362b2c0 157107Merge: b7ee005d3 d7c5e8bfc 157108Author: Peter Hutterer <peter@cs.unisa.edu.au> 157109Date: Wed Nov 7 15:37:23 2007 +1030 157110 157111 Merge branch 'master' into mpx 157112 157113 Conflicts: 157114 157115 Xi/extinit.c 157116 Xi/grabdev.c 157117 Xi/setmode.c 157118 Xi/ungrdev.c 157119 dix/devices.c 157120 dix/events.c 157121 dix/getevents.c 157122 include/dix.h 157123 mi/midispcur.c 157124 mi/misprite.c 157125 xkb/xkbActions.c 157126 xkb/xkbEvents.c 157127 xkb/xkbPrKeyEv.c 157128 157129commit e717409bae355df9a617a226f12fbb8c54ae77e5 157130Author: Daniel Stone <daniel@fooishbar.org> 157131Date: Tue Nov 6 21:36:13 2007 +0000 157132 157133 DIX/getevents: Document GetMaximumEventsNum() a little better 157134 157135 Note that the number returned by GMEN can _never_ change, and be a little more 157136 explicit about the figure for repeats. 157137 157138commit 66fe554a59bb7de37354b618945cd5f30d78250d 157139Author: Daniel Stone <daniel@fooishbar.org> 157140Date: Tue Nov 6 18:57:09 2007 +0000 157141 157142 COPYING: Collapse 'canonical license' into one statement 157143 157144 For a few of us, the license statement is identical, and the only variant is 157145 the copyright. For these, aggregate the copyrights, and only list the license 157146 once. Put this at the top, and note that this is more or less our agreed 157147 canonical license. 157148 157149commit fda832772b3e630037bf1b822534996154a50861 157150Author: Daniel Stone <daniel@fooishbar.org> 157151Date: Tue Nov 6 15:05:06 2007 +0000 157152 157153 .gitignore: Ignore build directories 157154 157155 Ignore directories people might use for building. 157156 157157commit 512bac25ec0e980968b93a2ebe88bd89bf99b697 157158Author: Daniel Stone <daniel@fooishbar.org> 157159Date: Tue Nov 6 14:52:03 2007 +0000 157160 157161 DIX: XKB: Set xkbInfo to NULL as well as freeing it (bug #10639) 157162 157163 XkbRemoveResourceClient wants to access xkbInfo if it exists, so make 157164 sure we NULL it after freeing it. It doesn't make much sense to move 157165 the RemoveResourceClient call first, as there's not much point in 157166 notifying clients while we're shutting the server down anyway. 157167 157168commit a8808ac3d093f33b39de109107d396fe0a02c4fc 157169Author: Peter Hutterer <peter@cs.unisa.edu.au> 157170Date: Wed Nov 7 11:42:58 2007 +1030 157171 157172 Xi: don't call CheckMotion for slave devices. 157173 157174 This essentially disables enter/leave for slave devices, but re-enables them 157175 for master devices. Which is a good thing after all. 157176 157177commit 15944b8f02752eedd61be34a6a29dd6b82a0ac97 157178Author: Peter Hutterer <peter@cs.unisa.edu.au> 157179Date: Wed Nov 7 11:33:24 2007 +1030 157180 157181 mi: duplicate event before processing it, so master has original values 157182 157183 Event values may get changed in the event processing stage, so we need to 157184 duplicate it to pump different events through for slave and master device. 157185 157186commit d7c5e8bfc1aecbd23a4cbb2eab08656587aac2e8 157187Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157188Date: Tue Nov 6 16:26:09 2007 -0500 157189 157190 Modified performance patches from Arjan van de Ven <arjan@infradead.org> 157191 157192 Subject: [PATCH] fix some performance gaps in Xace 157193 157194 The XaceHook function is used in several hotpaths. 157195 The problem with it (performance wise) is twofold: 157196 * The XaceHook function has a big switch() statement for the hook number in it 157197 * The XaceHook function uses varargs to reassemble the final dispatch arguments again 157198 157199 Both are expensive operations... for something that is known at compile time 157200 157201 This patch turns the hotpath XaceHook call into a direct call to avoid 157202 the switch and varargs; this gives me over 10% performance gain 157203 on the x11perf benchmark. 157204 157205commit aaa50b64113b122aaebd46e3b78e3fb7a8d70500 157206Merge: 868e30363 ddce48ede 157207Author: Dodji Seketeli <dodji@openedhand.com> 157208Date: Tue Nov 6 16:25:40 2007 +0100 157209 157210 resync with 'master' 157211 157212commit 868e303630d8b84070c2f1fd8d6da2cef045b029 157213Author: Dodji Seketeli <dodji@openedhand.com> 157214Date: Tue Nov 6 16:24:46 2007 +0100 157215 157216 Xephyr: fix a crash when using xrandr twice 157217 157218 * hw/kdrive/ephyr/ephyr.c: 157219 (ephyrScreenFini): don't forget to 157220 free shadowfb data (if necessary) upon server is reset. 157221 157222commit ddce48ede036f3996f8e584b0012c396c5df42fb 157223Author: Elvis Pranskevichus <el@prans.net> 157224Date: Tue Nov 6 09:40:14 2007 +0000 157225 157226 Config: D-Bus: Fix dbus_bus_request_name failure check 157227 157228 The code in connect_hook incorrectly checks for dbus_bus_request_name failure. 157229 The dbus_bus_request_name error indicator is -1, not 0. This leads 157230 to subsequent assertion failure in libdbus. 157231 157232commit a52c9b2a59f27266557ff9d5d2c08492e04135a6 157233Merge: c7e18beb3 58332894c 157234Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157235Date: Mon Nov 5 19:08:36 2007 -0500 157236 157237 Merge branch 'master' into XACE-SELINUX 157238 157239 Conflicts: 157240 157241 dix/dispatch.c 157242 dix/property.c 157243 hw/xfree86/common/xf86VidMode.c 157244 include/xkbsrv.h 157245 render/glyph.c 157246 xkb/xkbActions.c 157247 157248commit 58332894c061ae96d6a457f65266660f5f65e88b 157249Author: Adam Jackson <ajax@redhat.com> 157250Date: Mon Nov 5 17:17:25 2007 -0500 157251 157252 Export the server ABI versions from xorg-server.pc 157253 157254commit c7e18beb3c87eb1ada9b21c4ffacd11c1939c087 157255Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157256Date: Mon Nov 5 15:01:13 2007 -0500 157257 157258 xselinux: Register SELinux extension protocol names. 157259 157260commit 632c33c5c18b3e091c4fce98280af4d583e99640 157261Merge: 4113f040c dda10c906 157262Author: Dodji Seketeli <dodji@openedhand.com> 157263Date: Mon Nov 5 17:52:33 2007 +0100 157264 157265 pull 'master' 157266 157267commit 4113f040c587fc536adc693d7ee5a4c0a60b75d4 157268Author: Dodji Seketeli <dodji@openedhand.com> 157269Date: Mon Nov 5 17:33:01 2007 +0100 157270 157271 GL: fix crash at mesa destruction time 157272 157273 * GL/glx/glxglcore.c: 157274 (_glXMesaScreenDestroy): delete the same amount of visuals 157275 that those which were created in createMesaVisuals(). 157276 157277commit dda10c9066a660b647384179f82e1da8e063264f 157278Author: Daniel Stone <daniel@fooishbar.org> 157279Date: Mon Nov 5 16:28:35 2007 +0000 157280 157281 Remove all traces of external RGB database (and Speedo) 157282 157283 Remove all references to an external RGB database (which hasn't been enabled 157284 for a very long time). Also get rid of some references to Speedo fonts. 157285 157286commit 6ff79ea5f7d3ff0c3b14e39849514784ccd40190 157287Author: Dodji Seketeli <dodji@openedhand.com> 157288Date: Mon Nov 5 16:46:49 2007 +0100 157289 157290 Xephyr: fix some DRI build breakage 157291 157292commit 5e363500c86042c394595e1a6633581eb8fcd1bb 157293Author: Daniel Stone <daniel@fooishbar.org> 157294Date: Mon Nov 5 14:38:28 2007 +0000 157295 157296 OS: Remove ALLOCATE_LOCAL from os.h 157297 157298 Remove ALLOCATE_LOCAL_FALLBACK and DEALLOCATE_LOCAL_FALLBACK from os.h, and 157299 remove the include of Xalloca.h as well. 157300 157301commit 83e5d9e75d0fa1135e2c9d8b59caee98c3291681 157302Author: Daniel Stone <daniel@fooishbar.org> 157303Date: Mon Nov 5 14:36:54 2007 +0000 157304 157305 DIX: Remove last alloca call 157306 157307 Replace with heap allocations. 157308 157309commit 1179ddea94efae6606162e9a1b0f2bf752ae4dd0 157310Author: Daniel Stone <daniel@fooishbar.org> 157311Date: Mon Nov 5 14:36:36 2007 +0000 157312 157313 EXA: Remove usage of alloca 157314 157315 Replace with heap allocations. 157316 157317commit 1eb6a1d0d2817cd0ce96fb27db3ceb2d0ff024af 157318Author: Daniel Stone <daniel@fooishbar.org> 157319Date: Mon Nov 5 14:18:35 2007 +0000 157320 157321 XTrap: Remove usage of alloca 157322 157323 Replace with xalloc/xfree. 157324 157325commit 934281126f6c602fa4bd0c2e29d8c9f44fe532b8 157326Author: Daniel Stone <daniel@fooishbar.org> 157327Date: Mon Nov 5 14:18:22 2007 +0000 157328 157329 Xext: Remove usage of alloca 157330 157331 Replace with heap allocations. 157332 157333commit fb32bb9839b615f7297fbfac2050bc216682f01c 157334Author: Daniel Stone <daniel@fooishbar.org> 157335Date: Mon Nov 5 14:17:54 2007 +0000 157336 157337 Xi: Remove usage of alloca 157338 157339 Replace with xalloc/xfree. 157340 157341commit 683ee1776d172035c465aa8fc84ccd53bb8ba7fd 157342Author: Daniel Stone <daniel@fooishbar.org> 157343Date: Mon Nov 5 14:14:25 2007 +0000 157344 157345 afb: Remove usage of alloca 157346 157347 Replace with heap allocations. 157348 157349commit ca75261beedc3e00767b3812a81b7dac4437f4a1 157350Author: Daniel Stone <daniel@fooishbar.org> 157351Date: Mon Nov 5 14:14:04 2007 +0000 157352 157353 cfb: Remove usage of alloca 157354 157355 Replace with xalloc/xfree. 157356 157357commit 914922fd6100a409a3dfd1c64511ed6bdc344bef 157358Author: Daniel Stone <daniel@fooishbar.org> 157359Date: Mon Nov 5 14:12:59 2007 +0000 157360 157361 DIX: Remove usage of alloca 157362 157363 Replace with heap allocations. 157364 157365commit 3b77689266e729411229ec83d2a90578ebc1d82f 157366Author: Daniel Stone <daniel@fooishbar.org> 157367Date: Mon Nov 5 14:12:34 2007 +0000 157368 157369 EXA: Remove usage of alloca 157370 157371 Replace with heap allocations. 157372 157373commit 34cdf06e4ccb243664005cc33009d8759a7f6e4d 157374Author: Daniel Stone <daniel@fooishbar.org> 157375Date: Mon Nov 5 14:12:22 2007 +0000 157376 157377 fb: Remove usage of alloca 157378 157379 Replace with heap storage. 157380 157381commit 733d42065f2c24505b3874ce51c18f6063c2b67e 157382Author: Daniel Stone <daniel@fooishbar.org> 157383Date: Mon Nov 5 14:11:59 2007 +0000 157384 157385 XFree86: Remove usage of alloca 157386 157387 Replace with heap allocations. 157388 157389commit caf545063457591f88e1f7bcd25dbd0342f44343 157390Author: Daniel Stone <daniel@fooishbar.org> 157391Date: Mon Nov 5 14:10:55 2007 +0000 157392 157393 KDrive: Remove usage of alloca 157394 157395 Replace with heap allocations. 157396 157397commit d57060f16714f5667722001bd1a4500059dd59e1 157398Author: Daniel Stone <daniel@fooishbar.org> 157399Date: Mon Nov 5 14:10:21 2007 +0000 157400 157401 DMX: Remove usage of alloca 157402 157403 Replace with heap allocations. 157404 157405commit def6f74f2d7342f85f3df2053e0b9c1ac483b51d 157406Author: Daniel Stone <daniel@fooishbar.org> 157407Date: Mon Nov 5 14:10:03 2007 +0000 157408 157409 Xvfb: Remove usage of alloca 157410 157411 Replace with heap allocations 157412 157413commit 259f4c36d581896ce605741b9e557a589013a9b8 157414Author: Daniel Stone <daniel@fooishbar.org> 157415Date: Mon Nov 5 14:09:49 2007 +0000 157416 157417 Xprint: Remove usage of alloca 157418 157419 Replace with heap-based allocations. 157420 157421commit 1c84337af0ac40498a53aa36289e2f6f5ff8b1c6 157422Author: Daniel Stone <daniel@fooishbar.org> 157423Date: Mon Nov 5 14:09:32 2007 +0000 157424 157425 Xwin: Remove usage of alloca 157426 157427 Replace with heap allocations. 157428 157429commit f7d5c292e44113ea8eb32e67e91cd02e520df5e3 157430Author: Daniel Stone <daniel@fooishbar.org> 157431Date: Mon Nov 5 14:09:14 2007 +0000 157432 157433 mfb: Remove usage of alloca 157434 157435 Replace with heap-based allocations. 157436 157437commit be9ee17f960cc3d8a8f999cab1579e83d9aea520 157438Author: Daniel Stone <daniel@fooishbar.org> 157439Date: Mon Nov 5 14:08:51 2007 +0000 157440 157441 mi: Remove usage of alloca 157442 157443 Replace with heap allocations. 157444 157445commit 3633ae6efb57c5405c26e8ae132b9371e5f920de 157446Author: Daniel Stone <daniel@fooishbar.org> 157447Date: Mon Nov 5 14:04:27 2007 +0000 157448 157449 Damage: Remove usage of alloca 157450 157451 Replace with heap allocations. 157452 157453commit 2761c103311a1160bc483fd0367d654733df8598 157454Author: Daniel Stone <daniel@fooishbar.org> 157455Date: Mon Nov 5 14:03:26 2007 +0000 157456 157457 OS: Remove usage of alloca 157458 157459 Replace with heap allocations. 157460 157461commit 2d738efb959912d8a89864e41190e0064fa77906 157462Author: Daniel Stone <daniel@fooishbar.org> 157463Date: Mon Nov 5 14:01:11 2007 +0000 157464 157465 RandR: Remove usage of alloca 157466 157467 Replace with heap allocations. 157468 157469commit 3c1d2fdeff0ed61d86fa7d35cb0a61535d9b2816 157470Author: Daniel Stone <daniel@fooishbar.org> 157471Date: Mon Nov 5 14:00:40 2007 +0000 157472 157473 Record: Remove usage of alloca 157474 157475 Replace with xalloc/xfree. 157476 157477commit e0491f470e130147191388168e878e3a7348afaf 157478Author: Daniel Stone <daniel@fooishbar.org> 157479Date: Mon Nov 5 13:59:51 2007 +0000 157480 157481 Render: Remove usage of alloca 157482 157483 Replace it with heap-based allocations. 157484 157485commit 59774af86b851c7fb8989cef6c013522549000b8 157486Author: Daniel Stone <daniel@fooishbar.org> 157487Date: Mon Nov 5 13:59:15 2007 +0000 157488 157489 XKB: Remove usage of alloca 157490 157491 alloca has no way to return failure, and instead can possibly arbitrarily 157492 overflow the stack. Let's avoid that one. 157493 157494commit 6e4f5cf83f35ffebb51633ab30b1826e63e37223 157495Author: Ben Byer <bbyer@bbyer.local> 157496Date: Mon Nov 5 05:53:34 2007 -0800 157497 157498 changing ALLOCATE_LOCAL to xalloc to prevent stack overflow 157499 157500commit b1764ddf133cfdf979db62ee2491124a4798b55b 157501Author: Ben Byer <bbyer@bbyer.local> 157502Date: Mon Nov 5 05:44:54 2007 -0800 157503 157504 pulling more patches over from xorg-xserver-1.2-apple branch 157505 157506commit 10fde62fc88302f7d3b2546239b1679be249567c 157507Author: Ben Byer <bbyer@bbyer.local> 157508Date: Sun Nov 4 18:59:41 2007 -0800 157509 157510 more CVS tags 157511 157512commit 27ecb89250398f45564fe454e20bfdd66f62c8e4 157513Author: Ben Byer <bbyer@bbyer.local> 157514Date: Sun Nov 4 18:53:41 2007 -0800 157515 157516 Removed CVS tags. 157517 157518commit f2e310132fbe1520c1b5f3da4faa2d2d47835e72 157519Author: Aaron Plattner <aplattner@nvidia.com> 157520Date: Wed Oct 31 14:15:35 2007 -0700 157521 157522 Add CreatePixmap allocation hints. 157523 157524 These hints allow an acceleration architecture to optimize allocation of certain 157525 types of pixmaps, such as pixmaps that will serve as backing pixmaps for 157526 redirected windows. 157527 157528commit 3f1b6765aadf665ede8253464da19a5878f16e56 157529Author: Markku Vire <markku.vire@movial.fi> 157530Date: Thu Nov 1 22:43:04 2007 +0200 157531 157532 Config: HAL: Touchpads are pointers too 157533 157534 Treat touchpads -- not just mice -- as pointer devices. 157535 157536commit 41c3069f7cf28155f8e6cfe0c10a12a1f5f76c7d 157537Author: Mark Vytlacil <mrv@wi.rr.com> 157538Date: Thu Nov 1 21:05:43 2007 +0200 157539 157540 XFree86: Input: Save/restore errno around SIGIO (bug #10683) 157541 157542 Make sure errno is saved and restored from the SIGIO handler, so errors 157543 from system calls in input handlers don't break the interrupted code. 157544 157545commit ff9929ed48f2dec8b536d348e25e66a0bc4ac1a6 157546Author: Ben Byer <bbyer@bbyer.local> 157547Date: Sun Nov 4 05:13:19 2007 -0800 157548 157549 pulling in changes from xorg-server-1.2-apple branch 157550 157551commit 181468db92d44a58080fc9a76e46dfc7011bf9f1 157552Author: Ben Byer <bbyer@bbyer.local> 157553Date: Sun Nov 4 04:50:26 2007 -0800 157554 157555 add missing null-pointer checks 157556 157557commit 4d0f35c81df307609b29c0f12aa1b4c6ee6fd8d2 157558Author: Matthias Hopf <mhopf@suse.de> 157559Date: Fri Nov 2 19:27:32 2007 +0100 157560 157561 Always duplicate mode name when duplicating a mode. 157562 157563 If the originating mode didn't have a name, we would end up with the name of 157564 the original mode being setup correctly, but with the name of the copy still 157565 being NULL. 157566 157567commit dfb1cce28ac07d412598f7b9ab2ee908978b38e7 157568Author: Adam Jackson <ajax@redhat.com> 157569Date: Fri Nov 2 12:51:56 2007 -0400 157570 157571 Restore build of xf8_16bpp. 157572 157573 It doesn't link against cfb, so don't conditionalize its build against cfb. 157574 157575commit 7a0555e9bb59d02816803a1100f807d2d29f31d4 157576Author: Kristian Høgsberg <krh@redhat.com> 157577Date: Sun Oct 28 09:37:52 2007 +0100 157578 157579 Fix crash in xf86InitOrigins() 157580 157581 In a multihead setup, if only the first screen can be 157582 initialized, but the second screen is mentioned first in the 157583 ServerLayout section, the xf86InitOrigins() function will crash 157584 because the screen referred to in the e.g. "RightOf" part is 157585 non-existent. 157586 157587commit 5b41d4e60be35cfb96bedec0931fd5922823b4b9 157588Author: Adam Jackson <ajax@redhat.com> 157589Date: Sun Oct 28 09:37:52 2007 +0100 157590 157591 Don't filter modes away during VBE mode list construction. 157592 157593 Pass all VBE modes back up to the driver, on the assumption that it 157594 knows how to filter modes intelligently. 157595 157596commit c095da04fe7c73b6503ef5b93549b13796c51b22 157597Author: Adam Jackson <ajax@redhat.com> 157598Date: Sun Oct 28 09:37:52 2007 +0100 157599 157600 Fix magic number in fbdevhw 157601 157602 The transformation between fbdev and xfree86 mode timings needs to be 157603 invertible, otherwise Xen and other framebuffers that don't have real 157604 pixel clocks won't initialize. 157605 157606commit f4fe66f6767d1941317dc0280ac359421a152587 157607Author: Adam Jackson <ajax@redhat.com> 157608Date: Sun Oct 28 09:37:52 2007 +0100 157609 157610 Disable explicit commenting in Monitor section. 157611 157612commit 22f0e3a8b04e574047a51c8f928a007787303294 157613Author: Adam Jackson <ajax@redhat.com> 157614Date: Sun Oct 28 09:37:52 2007 +0100 157615 157616 Avoid PS/2 protocol probing for /dev/input/mice 157617 157618 The kernel will always upconvert to ExplorerPS/2 for all readers of 157619 /dev/input/mice, so don't waste time on startup trying to figure 157620 that out. 157621 157622commit b97518666dc32710fe69eee33ee56881dcff1bbc 157623Author: Adam Jackson <ajax@redhat.com> 157624Date: Sun Oct 28 09:37:52 2007 +0100 157625 157626 Fix accidental ABI usage in RANDR 1.2 drivers. 157627 157628 Due to RANDR 1.2, xf86findOptionValue and xf86nameCompare are now ABI. 157629 Make sure they're exported from the server. 157630 157631commit f01e149d1af14ef9ee0e8a6743ab6a08f3bb677c 157632Author: Adam Jackson <ajax@redhat.com> 157633Date: Thu Nov 1 15:41:11 2007 -0400 157634 157635 Move SIGUSR1 notification as late as possible. 157636 157637 If we inherited a signal mask from the parent process that ignores SIGUSR1, 157638 then we will send SIGUSR1 to the parent to indicate when we're ready to 157639 accept connections. Unfortunately, we send this notification way too 157640 early, right after creating the sockets rather than just before entering 157641 the main loop. 157642 157643 Move it to just before Dispatch() so we're not lying quite so much. 157644 157645commit a2ee5fe0c4e863c7ff4f644e5ac86f2793903103 157646Author: Adam Jackson <ajax@redhat.com> 157647Date: Thu Nov 1 13:46:20 2007 -0400 157648 157649 Enable MIT-SHM in Xdmx. 157650 157651 The extension is entirely DIX-level, no new DDX changes needed. 157652 157653commit 2338d5c9914e2a43c3a4f7ee0f4355ad0a1ad9e7 157654Author: Arjan van de Ven <arjan@linux.intel.com> 157655Date: Sun Oct 28 09:37:52 2007 +0100 157656 157657 reduce wakeups from smart scheduler 157658 157659 The smart scheduler itimer currently always fires after each request 157660 (which in turn causes the CPU to wake out of idle, burning precious 157661 power). Rather than doing this, just stop the timer before going into 157662 the select() portion of the WaitFor loop. It's a cheap system call, and 157663 it will only get called if there's no more commands batched up from the 157664 active fd. 157665 157666 This change also allows some of the functions to be simplified; 157667 setitimer() will only fail if it's passed invalid data, and we don't do 157668 that... so make it void and remove all the conditional code that deals 157669 with failure. 157670 157671 The change also allows us to remove a few variables that were used for 157672 housekeeping between the signal handler and the main loop. 157673 157674 Signed-off-by: Keith Packard <keithp@koto.keithp.com> 157675 157676commit 692654b4300e61a9481e6fa588bcb44a3c3ca150 157677Author: Kristian Høgsberg <krh@redhat.com> 157678Date: Mon Oct 29 18:13:58 2007 -0400 157679 157680 Set up visuals for the existing X visuals. 157681 157682 This makes the root visual a GLX capable visual again and adds a GLX visual 157683 for the COMPOSITE ARGB visual cleanly (as opposed to the hack we had before). 157684 157685commit a5546a99ac4da61aee0d49c55bcb38bbce9a96aa 157686Author: Kristian Høgsberg <krh@redhat.com> 157687Date: Mon Oct 29 18:06:41 2007 -0400 157688 157689 Change the GLX module to initialize after COMPOSITE. 157690 157691 This changes the module initalization order so that the GLX module initializes 157692 after COMPOSITE. The reason for this change is to be able to initialize a 157693 GLX visual config for the COMPOSITE ARGB visual. 157694 157695commit 8ead41388e36e21eea6fa0408c847f174911eab0 157696Author: Peter Hutterer <peter@cs.unisa.edu.au> 157697Date: Thu Sep 6 18:57:00 2007 +0930 157698 157699 xfree86: wrap keyboard devices for XKB. 157700 157701 Call ProcessOtherEvents first, then for all keyboard devices let them be 157702 wrapped by XKB. This way all XI events will go through XKB. 157703 157704 Note that the VCK is still not wrapped, so core events will bypass XKB. 157705 157706 (cherry picked from commit d627061b48ae06d27b37be209d67a3f4f2388dd3) 157707 157708commit e717cf08e99746761d74289c426bbd84176f4435 157709Author: Daniel Stone <daniel@fooishbar.org> 157710Date: Sat Oct 27 21:32:47 2007 +0300 157711 157712 XKB: Cope with all events in XkbProcessKeyboardEvent 157713 157714 Cope with Xi and pointer events in the (now increasingly misnamed) 157715 XkbProcessKeyboardEvent. If it's the wrong type, call through the wrapping 157716 chain to get out; else, process it. 157717 157718commit 9db8846fa53d91193bbfe541b244e2326440011d 157719Author: Daniel Stone <daniel@fooishbar.org> 157720Date: Sat Oct 27 21:31:39 2007 +0300 157721 157722 XKB: Don't update indicators on all devices, add missing include file 157723 157724 Don't get XkbUpdateIndicators to update the indicators on all our devices: we 157725 already deal with that ourselves. 157726 Add exevents.h include to get more (proto)types. 157727 157728commit ee3aa948eb8ed181d037294ed87df6ceec81684e 157729Author: Peter Hutterer <peter@cs.unisa.edu.au> 157730Date: Wed Sep 26 18:04:59 2007 +0930 157731 157732 xkb: Unwrap properly in ProcessPointerEvent. 157733 157734 Instead of hardcoding CoreProcessPointerEvent, actually try to unwrap properly 157735 and then call the unwrapped processInputProc. Seems to be a better idea, 157736 especially since it makes stuff actually work... 157737 (cherry picked from commit 8f9bf927e1beecf9b9ec8877131ec12c765e4d84) 157738 157739commit d3588a0aee33fbd233082f881c0d37152c6d4d8b 157740Author: Peter Hutterer <peter@cs.unisa.edu.au> 157741Date: Thu Sep 27 11:44:03 2007 +0930 157742 157743 xkb: xkbHandleActions: let wrapping take care of event delivery. 157744 157745 This is hopefully better than hardcodey calling CoreProcessPointerEvent. 157746 (cherry picked from commit 32d0440c7f6e604807cb14dd32349df6f22c903b) 157747 157748commit 352c5a311200bf491153fe9ef16126c5877a57bb 157749Author: Peter Hutterer <peter@cs.unisa.edu.au> 157750Date: Wed Sep 12 17:40:11 2007 +0930 157751 157752 dix: don't compress motion events from different devices (EventEnqueue) 157753 157754 (cherry picked from commit 8840829ab93c4eb62eb58753c015da5307133fe5) 157755 157756commit 8d3d027062c105b50863dce43b8070ec560bc12e 157757Author: Peter Hutterer <peter@cs.unisa.edu.au> 157758Date: Thu Sep 6 18:52:02 2007 +0930 157759 157760 dix: add XI event support to FixKeyState. 157761 157762 FixKeyState needs to be able to handle XI events, otherwise we get "impossible 157763 keyboard events" on server zaps and other special key combos. 157764 (cherry picked from commit 5ee409794ee604fcf84886f70429fc2d6b1ff4f1) 157765 157766commit 99e826e867c1c5520153c539ba07a884aec88d0c 157767Author: Peter Hutterer <peter@cs.unisa.edu.au> 157768Date: Thu Sep 6 18:49:57 2007 +0930 157769 157770 xkb: enable XI event processing for xkb. 157771 157772 XI events can now take the same processing paths as core events, and should do 157773 the correct state changes etc. 157774 157775 There's some cases where XKB will use KeyPress as type for an event to be 157776 delivered to the client. Stuck warnings in, not sure what the correct solution 157777 is yet. 157778 157779 (cherry picked from commit 6334d4e7be18de5f237c12a6dc20f75aa23477d0 with some 157780 additional compile fixes and non-MPX adaptations) 157781 157782commit 91077bfc50d54be37c217e377c55b6bf886a2fab 157783Author: Peter Hutterer <peter@cs.unisa.edu.au> 157784Date: Thu Jun 21 18:24:30 2007 +0930 157785 157786 Save processInputProc before wrapping it and restore it later, instead of 157787 using a hardcoded ProcessKeyboardEvent. Otherwise we lose the ability to 157788 process DeviceKeyEvents after the first key press. 157789 157790 This should be the correct fix now. 157791 (cherry picked from commit 4d5df14f2c4a3108a8c8adfcf4766c0d1a9daad2) 157792 157793commit 8b9481a113b56078191e2298bf590905978f6289 157794Author: Peter Hutterer <peter@cs.unisa.edu.au> 157795Date: Tue Sep 4 17:44:51 2007 +0930 157796 157797 xkb: Store the action filters per device in the XkbSrvInfoRec. 157798 157799 Using a global array for action filters is bad. If two keyboard hit a modifier 157800 at the same time, releaseing the first one will deactivate the filter and 157801 thus the second keyboard can never release the modifier again. 157802 (cherry picked from commit bfe6b4d2d9952a80f8dbc63eec974ef894e5c226) 157803 157804commit 493b83bd097372ae0023da9919da83af39e3fc1c 157805Author: Daniel Stone <daniel@fooishbar.org> 157806Date: Sat Oct 27 21:33:52 2007 +0300 157807 157808 XFree86: Remove ridiculous SIGIO debugging 157809 157810 YOU PRESSED A KEY 157811 AND AGAIN 157812 YOU RELEASED A KEY 157813 AND AGAIN 157814 YOU PRESSED A KEY 157815 AND AGAIN 157816 157817 ... not so much. 157818 157819commit a3d48de5f2b7eacf3193c60f0fb461912201210b 157820Author: Daniel Stone <daniel@fooishbar.org> 157821Date: Sat Oct 27 21:35:31 2007 +0300 157822 157823 XKB: Add more bits to xkbsrv.h 157824 157825 Add the device private index, given we use that in a macro here, and also the 157826 prototype for xkbUnwrapProc, since that's also useful. 157827 157828commit e29e69960d67aa4b7a4d1551af509dbac193f438 157829Author: Daniel Stone <daniel@fooishbar.org> 157830Date: Sat Oct 27 21:34:22 2007 +0300 157831 157832 Xi: Include XI protocol header in exevents.h 157833 157834 Make sure we have all the types we need to use this header. 157835 157836commit df57ae1639ba4f1719883c5bf868394e4748a022 157837Author: Daniel Stone <daniel@fooishbar.org> 157838Date: Sun Oct 28 15:46:26 2007 +0200 157839 157840 configure.ac/XFree86: Only build XF86Misc and XF86VidMode when appropriate 157841 157842 Don't build XF86Misc or XF86Vidmode in hw/xfree86/dixmod when it's been 157843 explicitly disabled in configure, or we don't have the proto modules 157844 installed. 157845 157846commit 3b7af72fe315c7c26c89838c0c5dacbe58765d0f 157847Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157848Date: Fri Oct 26 20:32:10 2007 -0400 157849 157850 xselinux: Add a SetDeviceContext request and stubs for more requests. 157851 157852commit 2251572062b2c25643671b8d5070de1c3f7ae976 157853Author: Aaron Plattner <aplattner@nvidia.com> 157854Date: Fri Oct 26 15:13:50 2007 -0700 157855 157856 Restore the CompositeGlyphs -> ps->Glyphs -> miGlyphs callchain to allow acceleration architectures to wrap above miGlyphs. 157857 157858commit 497aba894904b08b15bb19916e2a163f96212a7d 157859Author: Kristian Høgsberg <krh@redhat.com> 157860Date: Fri Oct 26 11:25:57 2007 -0400 157861 157862 Weed out some unused fluff from __GLXcontext. 157863 157864commit 0e749ceab421d72b48e18292c5ca0e337baf5ce8 157865Author: Bernardo Innocenti <bernie@codewiz.org> 157866Date: Sat Oct 20 20:31:57 2007 -0400 157867 157868 Include stddef.h for size_t. 157869 157870 This is required by a buggy version of the openssl/sha.h header 157871 which is distributed with Fedora 7. 157872 157873commit 80d8a3e1767609a92b0b85a7e92cea5f0748038a 157874Author: Kristian Høgsberg <krh@redhat.com> 157875Date: Fri Oct 26 10:19:57 2007 -0400 157876 157877 Quiet glx visual debug info. 157878 157879commit 27b13eeb56a5cbac00b381309604505002e77228 157880Author: Matthias Hopf <mhopf@suse.de> 157881Date: Fri Oct 26 13:15:07 2007 +0200 157882 157883 Nuke superfluous ",", added missing include for xf86i2c.h. 157884 157885commit 6d59bb5709a99ab60b482bbf3393ebffda7f9407 157886Author: Daniel Stone <daniel@fooishbar.org> 157887Date: Fri Oct 26 09:12:15 2007 +0300 157888 157889 XFree86 Misc/VidMode: Remove ridiculous debug ErrorFs 157890 157891 When we're building with --enable-debug, don't emit an ErrorF every time a 157892 function gets called. 157893 157894commit 5f9095f0d29bac0190d82c87a09cf32d6a34c17c 157895Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157896Date: Thu Oct 25 19:02:03 2007 -0400 157897 157898 registry: Remove synthetic bit from event types in lookup function. 157899 157900commit 8c6923018c7d71cd15d9cf4ef9e8528ef5ec7c2e 157901Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157902Date: Thu Oct 25 19:01:29 2007 -0400 157903 157904 xace: Add a "manage" access check when setting the Redirect event bits. 157905 157906commit 7d14ca59c5b942c09feaa2429c394cde9d8d3fd1 157907Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157908Date: Thu Oct 25 19:00:50 2007 -0400 157909 157910 xselinux: Don't include the client in the receive hook audit messages. 157911 157912commit f62277d421023b3150d3a1accb00a8206ab2bde3 157913Author: Kristian Høgsberg <krh@redhat.com> 157914Date: Thu Oct 25 18:48:39 2007 -0400 157915 157916 Clean up unused pVisual part of __GLXcontext. 157917 157918commit 30bcaa966d6b00f1630609a78db18dee683cc43d 157919Author: Kristian Høgsberg <krh@redhat.com> 157920Date: Thu Oct 25 18:48:06 2007 -0400 157921 157922 Make glx destroy path handle cases where the X window goes away first. 157923 157924commit 40de9fcf18930811dd5ae355c83275af887a9f83 157925Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157926Date: Thu Oct 25 12:35:01 2007 -0400 157927 157928 xselinux: Label the default device directly with the process context. 157929 157930commit b633d54b94d746d26e13014634d9f63bbb7e8f7d 157931Merge: 4b05f19cb 48ca5961c 157932Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157933Date: Thu Oct 25 12:19:30 2007 -0400 157934 157935 Merge branch 'master' into XACE-SELINUX 157936 157937 Conflicts: 157938 157939 GL/glx/glxscreens.c 157940 hw/xnest/Screen.c 157941 render/glyph.c 157942 render/glyphstr.h 157943 render/render.c 157944 157945commit 242f56f722243938e908d1957781ee53c2999783 157946Author: Naoki Hamada <nao@tom-yam.or.jp> 157947Date: Thu Oct 25 18:45:50 2007 +0300 157948 157949 Input: Fix key down test (bug #12858) 157950 157951 Fix the botched previous key_is_down test, which would give false positives. 157952 Also move key_autorepeats to a separate inline function. 157953 157954commit 4b05f19cb9e42d8c8eff5ca4e463f5bc2a05433d 157955Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157956Date: Wed Oct 24 19:59:58 2007 -0400 157957 157958 xselinux: Introduce a type transition when labeling events. 157959 157960commit 0d2ef187e77b12713d2a9661932fa01dba58a945 157961Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157962Date: Wed Oct 24 18:23:31 2007 -0400 157963 157964 xselinux: Add audit message fields for selection and event names. 157965 157966commit 48ca5961caee62f2980017a6bdc96a1b4c747727 157967Author: Matthias Hopf <mhopf@suse.de> 157968Date: Wed Oct 24 20:31:51 2007 +0200 157969 157970 Prefer configured DisplaySize to probed DDC data, if available. 157971 157972 Based on patch by Hong Liu <hong.liu@intel.com>. 157973 157974commit 0388a59a6ef212c497cc3f64d677b1ca5b410982 157975Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157976Date: Tue Oct 23 20:59:21 2007 -0400 157977 157978 Revert "registry: special case minor number when looking up core requests." 157979 157980 This reverts commit 31110d6837ee52fd654729d9e5c4b0c5395abab0. 157981 157982 This is handled properly by StandardMinorOpcode(). 157983 157984commit 46521f529841e032e198e5df87974088548a68de 157985Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 157986Date: Tue Oct 23 20:58:48 2007 -0400 157987 157988 xselinux: Add basic support for selection access control and redirection. 157989 157990 Probably not fully baked yet. It's difficult to test since so few apps 157991 actually follow the ICCCM with respect to cut & paste. 157992 157993commit b7ee005d327372c1e414ee6c526f9f7aee14bc86 157994Author: Peter Hutterer <peter@cs.unisa.edu.au> 157995Date: Wed Oct 24 10:26:12 2007 +0930 157996 157997 dix: Don't interfere grabs the interfering device is of different type. 157998 157999 A pointer device should be able to send events while the client has a core 158000 grab on a keyboard device, and likewise. 158001 158002commit 02508614b9f882ba1559fb4581dcf812118fdf89 158003Author: Peter Hutterer <peter@cs.unisa.edu.au> 158004Date: Wed Oct 24 10:09:18 2007 +0930 158005 158006 dix: check for core passive grabs on inferiors when replaying events. 158007 158008commit 825f09dffd94cfcd0562a01c5181998503851461 158009Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158010Date: Tue Oct 23 17:12:57 2007 -0400 158011 158012 xace: Still more changes to selection code. Removed the SelectionPtr from 158013 the hook - the hook only needs the Atom to control access to the selection 158014 object. Upgraded the SelectionCallback to take a client argument and 158015 additional type codes so that it can be used for redirection. 158016 158017commit 660557593ea961948722298ea8ffba83891c9914 158018Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158019Date: Tue Oct 23 14:46:37 2007 -0400 158020 158021 xselinux: Remove synthetic bit when looking up event type. 158022 158023commit d7db549db41a27aef28cff9bfb7973bc741f88b2 158024Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158025Date: Tue Oct 23 14:08:54 2007 -0400 158026 158027 xselinux: Unregister callbacks on server reset. 158028 158029commit 9e0a468af19d8e46330bcff37c9adc5e11d3aee7 158030Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158031Date: Tue Oct 23 13:35:30 2007 -0400 158032 158033 xace: try to pretend events were sent when a denial occurs. 158034 158035 Probably need to redo the error return paths in these functions 158036 at some point. 158037 158038commit d502521c3669f3f22b94c39a64ab63bfd92c6a97 158039Author: Pierre Willenbrock <pierre@pirsoft.dnsalias.org> 158040Date: Tue Oct 23 16:45:13 2007 +0200 158041 158042 EXA: Fix off-by-one in polyline drawing. 158043 158044commit a8a148919b84a293e3e7a49409ab833590357edb 158045Author: Matthias Hopf <mhopf@suse.de> 158046Date: Tue Oct 23 16:23:28 2007 +0200 158047 158048 Superfluous ','. 158049 158050commit 75f05086d04a90c3dcdcdd31bf79d7033708e3e0 158051Author: Matthias Hopf <mhopf@suse.de> 158052Date: Tue Oct 23 15:39:23 2007 +0200 158053 158054 Get rid of unnecessary GNU extended variadic macro. 158055 158056commit 7ef863720dc79107fc2cd17ce684366c87e001a4 158057Author: Peter Hutterer <peter@cs.unisa.edu.au> 158058Date: Tue Oct 23 17:28:03 2007 +0930 158059 158060 dix: Create new sprite for floating devices. 158061 158062 Floating devices get sprites, but still aren't spriteOwners. This prevents 158063 them from getting rendered, and also stops segfaulting. 158064 (not really solving the problems with keyboards though) 158065 158066commit d808d653d1fc16f1d5af76ab00fa862fb80fa3ba 158067Author: David Nusinow <dnusinow@debian.org> 158068Date: Mon Oct 22 21:30:04 2007 -0400 158069 158070 Define PCI_TXT_IDS_DIR more cleanly 158071 158072commit 29e0e180729a4f0cc020985a4de4c8bc4b9c7f5f 158073Author: Keith Packard <keithp@koto.keithp.com> 158074Date: Mon Oct 22 13:38:16 2007 -0700 158075 158076 Leave hardware-specified preferred modes alone when user preference exists. 158077 158078 Instead of removing the preference bit marking the hardware declared mode 158079 preference, leave it in place and just move the user preferred mode to the 158080 front of the list while marking it with the USERPREF bit which will cause it 158081 to be selected by the initial mode selection code. 158082 158083commit fbe19c66c36acfb484809111cf02579a3baf2f0f 158084Author: Michel Dänzer <michel@tungstengraphics.com> 158085Date: Mon Oct 22 18:28:03 2007 +0200 158086 158087 GLX: Fix leak of X pixmaps associated with GLX pixmaps. 158088 158089commit c6d36b1cee44a9cbb690dff62a4683d7f6fbf30c 158090Author: Michel Dänzer <michel@tungstengraphics.com> 158091Date: Mon Oct 22 18:28:03 2007 +0200 158092 158093 GLX: Don't crash on unused client array members when switching to/from console. 158094 158095commit be9b0e558d6172ef01ed880d47da897d2ceb25fa 158096Author: David Nusinow <dnusinow@debian.org> 158097Date: Sun Oct 21 23:49:14 2007 -0400 158098 158099 Add a note as to why the text file based PCI ID scheme is present at all 158100 158101commit b79b965bd9a96f79781e85c0428068caa1ba381b 158102Author: David Nusinow <dnusinow@debian.org> 158103Date: Sun Oct 21 23:46:54 2007 -0400 158104 158105 Fix compilation issue on FreeBSD (bug #12841) 158106 158107 Hide getline call by checking for glibc. If not, use fgetln instead. Even 158108 though this section is now #ifdef'ed for linux only, this should help make 158109 it more portable if non-linux folks end up wanting it. 158110 158111commit 7f5e71fa7347fb4ec2657ccc453831c04a0f288d 158112Author: David Nusinow <dnusinow@debian.org> 158113Date: Sun Oct 21 23:41:49 2007 -0400 158114 158115 Make PCI text file directory configurable 158116 158117commit ce7f6fe1268fef4f89aa21c7b44d73ecd98efe24 158118Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158119Date: Fri Oct 19 19:40:04 2007 -0400 158120 158121 xselinux: properly update sizes when dynamic arrays are resized... 158122 158123commit 4b14c9a9cd2033d3839c4ba364d41ab4c4b198ab 158124Author: Eric Anholt <eric@anholt.net> 158125Date: Fri Oct 19 16:34:54 2007 -0700 158126 158127 Replace calls to Glyphs screen hook with CompositeGlyphs and remove dead code. 158128 158129 Not all of the DDX/miext Glyphs hook implementations have been removed, but 158130 they should be. 158131 158132commit a3a95d3475dc91ed2e8a55bf484a6b3f2b5ac32a 158133Merge: a358b87f4 7e1cada6c 158134Author: Eric Anholt <eric@anholt.net> 158135Date: Fri Oct 19 15:44:17 2007 -0700 158136 158137 Merge branch 'master' into glyph-pixmaps 158138 158139 Conflicts: 158140 158141 configure.ac 158142 158143commit 12e889d202ac9849f534c51167cbfed91c32027a 158144Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158145Date: Fri Oct 19 18:43:38 2007 -0400 158146 158147 xace: Bug fixes, name changes to selection access hooks and fields. 158148 158149commit 7e1cada6c6b9fa41ef3ead00bf7725d626dda193 158150Author: Eric Anholt <eric@anholt.net> 158151Date: Fri Oct 19 15:16:19 2007 -0700 158152 158153 Disable debian pci-id-list autoconfig code on non-linux. 158154 158155 It contains static paths, fails to build on non-glibc, and apparently just 158156 exists to support distributions managing binary drivers and open-source drivers 158157 together. Also restores previous code for fallback to vesa if nothing is 158158 detected. 158159 158160commit 89c6108531e603bdc81faf2ea860f318a2e94a39 158161Author: Kristian Høgsberg <krh@redhat.com> 158162Date: Fri Oct 19 16:21:54 2007 -0400 158163 158164 Fix software GL to provide a list of supported fbconfigs like the DRI case. 158165 158166commit 74e84b8d891632141a60e2d29463ab46f622df3f 158167Author: Matthias Hopf <mhopf@suse.de> 158168Date: Fri Oct 19 17:28:15 2007 +0200 158169 158170 Only issue XF86_APM_CAPABILITY_CHANGED for video change ACPI events (not e.g. brightness etc.). 158171 158172commit 6fdf3ddfef2707755cec0fc9cbce354fd472bb52 158173Author: Matthias Hopf <mhopf@suse.de> 158174Date: Fri Oct 19 17:20:21 2007 +0200 158175 158176 Fix ids of ACPI events according to ACPI spec. 158177 158178commit d9caa469b9bb4eb6125b890820853062fc2c4441 158179Author: Peter Hutterer <peter@cs.unisa.edu.au> 158180Date: Fri Oct 19 15:26:09 2007 +0930 158181 158182 Xi: allow VCP/VCK to be OpenDevice'd as well. 158183 158184 All devices deserve to be equal. Except master devices, they are a bit more 158185 equal than the others. 158186 158187commit 155e2c559ed0dbf31b6d39d48648a3ee22b37635 158188Author: Peter Hutterer <peter@cs.unisa.edu.au> 158189Date: Fri Oct 19 14:37:07 2007 +0930 158190 158191 Xi: advertise as XInput v2 capable 158192 158193commit 5ba738935f0d786e4670adf3b05ad42fc5e710fd 158194Author: Peter Hutterer <peter@cs.unisa.edu.au> 158195Date: Fri Oct 19 14:36:37 2007 +0930 158196 158197 Xi: remove ChangePointerKeyboardPairing in favour of ChangeDeviceHierarchy. 158198 158199commit 6dcde0e9c1d068d9fc4a772d29d1d4c6cc57aeb9 158200Author: Peter Hutterer <peter@cs.unisa.edu.au> 158201Date: Thu Oct 18 12:21:07 2007 +0930 158202 158203 Xi: set master device's id in ListDevices Reply. 158204 158205commit cfcc6e14b9a15f7979ba0df9c105cf563bef98fa 158206Author: Peter Hutterer <peter@cs.unisa.edu.au> 158207Date: Fri Oct 19 11:28:44 2007 +0930 158208 158209 Xi: return all master devices as type IsXPointer/Keyboard when listing devs. 158210 158211 Slave devices are reported as IsXExtensionPointer/Keyboard. 158212 158213commit 8e3c1dfc48930c455529313a42efa35e3b9071b2 158214Author: Kristian Høgsberg <krh@redhat.com> 158215Date: Thu Oct 18 21:01:51 2007 -0400 158216 158217 Introduce a new "GlxVisuals" option that controls which visuals are added. 158218 158219 Right now we default to "all" which gives us a situation much like before, 158220 but when the "typical" option is implemented, we can change the default and 158221 reduce the number of visuals the GLX module bloats the X server with. 158222 158223commit 9e3cfd950d279353f1795352ad2c4132aabc1bab 158224Author: Kristian Høgsberg <krh@redhat.com> 158225Date: Thu Oct 18 19:12:24 2007 -0400 158226 158227 Add code to set up GLX visuals and add one for the first double buffered FBConfig. 158228 158229commit 8fc635e6a8072c7cd2777d804dd6f8eda2eecc15 158230Author: Kristian Høgsberg <krh@redhat.com> 158231Date: Thu Oct 18 15:51:11 2007 -0400 158232 158233 Separate handling of GLX visuals and GLX FBConfigs. 158234 158235 XIDs for GLX visuals and FBConfigs used to be interchangable and the list of 158236 GLX visuals was identical to the list for FBConfigs. This patch splits handling 158237 of these two data types and allows the X server to pick and choose the FBConfigs 158238 that are exposed as visuals. 158239 158240commit 0af8180683247955ce4cfd48e6a5b4d00bbe618a 158241Author: Kristian Høgsberg <krh@redhat.com> 158242Date: Wed Oct 17 17:18:47 2007 -0400 158243 158244 Drop VisualConfigPriv support from the DRI module. 158245 158246 No DDX driver uses this, and this patch stops the DRI module from 158247 poking around GLX module data structures. 158248 158249commit 7d74690536b64f7b8e8036507ab7790807349c50 158250Author: Kristian Høgsberg <krh@redhat.com> 158251Date: Sun Oct 14 15:39:21 2007 -0400 158252 158253 Simplify and clean up GLX visual initialization. 158254 158255 Instead of the fragile setup where we filter the modes common between the 158256 DDX generated GLX visuals and the DRI driver generated fbconfigs, we now 158257 just take the fbconfigs returned by the DRI driver to be our supported set. 158258 158259commit ab4bce02a9457dd9c86b774fc74caf3dd6b287ca 158260Author: Eric Anholt <eric@anholt.net> 158261Date: Thu Oct 18 15:22:42 2007 -0700 158262 158263 Add a quirk for Philips 107P5 which lacks the preferred bit on detailed timing. 158264 158265 Also fix the prefer-large-75 quirk if the prefer-first-detailed bit was set, 158266 though it's not the case for the existing prefer-large-75 consumer. 158267 158268commit 55a96aa6b0995fda6660b7e78c85b955a62b9735 158269Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158270Date: Thu Oct 18 14:11:11 2007 -0400 158271 158272 xselinux: add basic event labeling. 158273 158274commit e974bc1233608ec09fbd40b12217925e4d2205aa 158275Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158276Date: Thu Oct 18 12:33:39 2007 -0400 158277 158278 xselinux: add hooks for send and receive access. 158279 158280commit 06eb830169afd0631a31e8846c7d2533c49ea378 158281Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158282Date: Thu Oct 18 12:31:14 2007 -0400 158283 158284 xace: Fix bug in AddPassiveGrabToList(), was using wrong GrabPtr. 158285 158286commit 6107a245035366fe762756b6aa05ac0e3a5482bb 158287Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158288Date: Thu Oct 18 12:24:55 2007 -0400 158289 158290 dix: Add client parameter to AddPassiveGrabsToList(). 158291 158292commit ce50bfd3369686cfecee5a138bd84ef1107a249d 158293Author: Michel Dänzer <michel@tungstengraphics.com> 158294Date: Thu Oct 18 17:44:48 2007 +0200 158295 158296 EXA: Skip empty glyphs. 158297 158298commit 5d74416740de883b7ef0994afea4bbd4d3901be0 158299Author: Michel Dänzer <michel@tungstengraphics.com> 158300Date: Thu Oct 18 17:44:14 2007 +0200 158301 158302 EXA: Don't attempt to move in pixmaps that can't be accelerated. 158303 158304 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=12815 . 158305 158306commit 31110d6837ee52fd654729d9e5c4b0c5395abab0 158307Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158308Date: Thu Oct 18 10:30:44 2007 -0400 158309 158310 registry: special case minor number when looking up core requests. 158311 158312commit e3fd90ae9c3ddfc5d78e62614e311b73505d7ead 158313Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158314Date: Thu Oct 18 10:29:10 2007 -0400 158315 158316 registry: Add "X11:" prefix to core protocol names. 158317 158318commit aa340b2c7cbe9ddab53cff08c8ba165558209187 158319Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158320Date: Wed Oct 17 19:27:16 2007 -0400 158321 158322 xselinux: add hook for device acceses. 158323 158324commit 503f918f55d0cb29585d83b022bbb8dc29f446c5 158325Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158326Date: Wed Oct 17 19:14:15 2007 -0400 158327 158328 xselinux: Move functions around; add some more comments. 158329 158330commit 50b27e1ad2a98d36728dc8157492ef5c59c132cd 158331Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158332Date: Wed Oct 17 16:09:40 2007 -0400 158333 158334 devPrivates rework: update new GL/glxext code. 158335 158336 Need to merge so this type of thing stops happening. 158337 158338commit c3f7b862550fa8f46633162f83db8c27e46fc672 158339Merge: af4dde0ac feac07595 158340Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158341Date: Wed Oct 17 15:00:54 2007 -0400 158342 158343 Merge branch 'master' into XACE-SELINUX 158344 158345 Conflicts: 158346 158347 os/access.c 158348 158349commit af4dde0ac19ecec1d0ad988eb25b15401e7c6b36 158350Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158351Date: Wed Oct 17 14:13:02 2007 -0400 158352 158353 xselinux: Remove config file, this has been moved to the policy. 158354 158355commit baabae623b3658196b67a710dc72663c2105bf31 158356Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158357Date: Wed Oct 17 13:54:56 2007 -0400 158358 158359 xselinux: Started reworking extension using new XACE hooks. 158360 158361commit db66e66dbf26b91c655f1659859c022cc31f0db6 158362Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158363Date: Wed Oct 17 13:51:11 2007 -0400 158364 158365 xace: Add an access_mode field to the extension structure. 158366 This allows the same callback to be used for both extension hooks. 158367 158368commit e3a8cbe523bae8b771ad3c8ad497f4444f6d05d5 158369Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158370Date: Wed Oct 17 13:48:44 2007 -0400 158371 158372 xace: add creation/labeling hook to CreateRootWindow(). 158373 158374commit feac0759522cbdc3e61ccfa373df735903c5cb27 158375Author: Keith Packard <keithp@koto.keithp.com> 158376Date: Wed Oct 17 11:42:28 2007 +0800 158377 158378 Make config file preferred mode override monitor preferred mode. 158379 158380 Add a new even-more-preferred bit to each mode which is used to make config 158381 file preferences selected instead of the monitor preferred mode. 158382 158383commit f2da10f7bc2ddb6ad2f18b793afc10d04b97c51c 158384Author: Keith Packard <keithp@koto.keithp.com> 158385Date: Wed Oct 17 10:50:22 2007 +0800 158386 158387 KDRIVE_LOCAL_LIBS includes some system libraries, not just internal x server libs 158388 158389commit 9aadde377991bfbd88524d02106bec3dedd6e7c9 158390Author: Peter Hutterer <peter@cs.unisa.edu.au> 158391Date: Wed Oct 17 12:05:50 2007 +0930 158392 158393 mi: don't call UndisplayCursor for non-sprite-owners. 158394 158395 Segfaults are bad. 158396 158397commit 8b7c4249d82d07e852d8d8c15c7ab9977dd1f31c 158398Author: Peter Hutterer <peter@cs.unisa.edu.au> 158399Date: Wed Oct 17 10:41:58 2007 +0930 158400 158401 mi: call processInputProc for master devices after slave event processing. 158402 158403 More work is needed to sort out grabs though. 158404 158405commit ae9fc10adec8f9bf0492d14d435f8f11e0163b27 158406Author: Peter Hutterer <peter@cs.unisa.edu.au> 158407Date: Wed Oct 17 10:41:04 2007 +0930 158408 158409 Xi: don't send core events for slave devices. 158410 158411commit de70cfdbe60eb6e7bf3e74dfd1ac34de554deff1 158412Author: Peter Hutterer <peter@cs.unisa.edu.au> 158413Date: Wed Oct 17 09:52:30 2007 +0930 158414 158415 mi: fix some macros to allow multiple cursors for master devices. 158416 158417 Macros defaulted to inputInfo.pointe rfor devices that weren't spriteOwners. 158418 Changed to take the device's master device now. 158419 158420 This includes sticking in a number of checks and warnings that cursor 158421 rendering won't be called for floating devices. 158422 158423commit 773f6491c1cc8819038e753d08c32ba213f80f8f 158424Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158425Date: Tue Oct 16 19:11:36 2007 -0400 158426 158427 xace: update the DeleteProperty prototype to include the client argument. 158428 This should have been part of 8f23d40068151ad85cde239d07031284f0b2c4dc. 158429 158430commit 9dde53ed179336c7b483c9a94a97182ad1777dfb 158431Author: Alan Hourihane <alanh@tungstengraphics.com> 158432Date: Tue Oct 16 19:55:56 2007 +0100 158433 158434 Fix walking of GLX providers. 158435 158436commit 70a5d33c9e41c077a8cd92abd43376e2956d3aed 158437Author: Michel Dänzer <michel@tungstengraphics.com> 158438Date: Tue Oct 16 12:46:07 2007 +0200 158439 158440 Always check the return value of __glXGetDrawable first. 158441 158442 Fixes spurious GLX protocol errors because __glXGetDrawable doesn't set the 158443 error code in case of success. Maybe it should, though. 158444 158445commit 1d9ebbac8c589cae7e4952083692b6d148def9bc 158446Author: Peter Hutterer <peter@cs.unisa.edu.au> 158447Date: Tue Oct 16 18:46:12 2007 +0930 158448 158449 dix: Make InitCoreDevices use AllocMasterDevice. 158450 158451 Also change naming a bit, append "pointer" and "keyboard" to master devices 158452 instead of -ptr and -keybd. 158453 158454commit 299573f4617c3b5599bb65069e96d050277b9471 158455Author: Peter Hutterer <peter@cs.unisa.edu.au> 158456Date: Tue Oct 16 18:40:15 2007 +0930 158457 158458 dix: add AllocMasterDevice for creation of new master devices. 158459 158460 Devices are initiated pretty much the same as the core devices. 158461 158462commit 5fe9bfd23f17b84c3afaa82f75a7c517c9f8e0d3 158463Author: Peter Hutterer <peter@cs.unisa.edu.au> 158464Date: Tue Oct 16 18:23:48 2007 +0930 158465 158466 xfree86: NIDR: don't call PairDevices explicitly. 158467 set isMaster to FALSE explicitly. 158468 158469 Pairing isn't necessary, attachment should be done in EnableDevices. 158470 158471commit 5eb033835e92ea951cc385fd709af9656b3772d8 158472Author: Peter Hutterer <peter@cs.unisa.edu.au> 158473Date: Tue Oct 16 18:03:52 2007 +0930 158474 158475 mi: don't exclude inputInfo.pointer from rendering. 158476 158477commit b697c4ed145968d3c3281bb85e628f1b068b09fc 158478Author: Peter Hutterer <peter@cs.unisa.edu.au> 158479Date: Tue Oct 16 17:47:06 2007 +0930 158480 158481 dix: CheckPassiveGrabOnWindow: only get paired device for non-keyboards. 158482 158483commit caa69d8f7b92b80669df02e1001409d7c653e7e4 158484Author: Peter Hutterer <peter@cs.unisa.edu.au> 158485Date: Tue Oct 16 17:38:13 2007 +0930 158486 158487 dix: fix detritus from adding lastSlave field. 158488 158489commit 90d077e537ac4cb7d79d67afcf3984a3e8d65fe8 158490Author: Peter Hutterer <peter@cs.unisa.edu.au> 158491Date: Tue Oct 16 17:37:33 2007 +0930 158492 158493 dix: GetPointerEvents: get state from master device before applying valuators. 158494 158495 We need to get lastx/y from the master device before we start applying 158496 acceleration etc. Otherwise we get jumping cursors, which is reasonably bad. 158497 158498commit b6ccf721b0a582150858d68f91967fb4e319c340 158499Author: Peter Hutterer <peter@cs.unisa.edu.au> 158500Date: Tue Oct 16 17:35:06 2007 +0930 158501 158502 include: add "lastSlave" field to DeviceIntRec. 158503 158504 Set to the last slave device that routed events through the master. 158505 158506commit e6bd8ae0608bd8379c5ac962f69cd0bcc54f9734 158507Author: Peter Hutterer <peter@cs.unisa.edu.au> 158508Date: Tue Oct 16 15:59:30 2007 +0930 158509 158510 dix: set the device's sprite when using AttachDevice 158511 158512 Sprite is set to the master device's sprite. 158513 158514commit 271d4c3d6255318aabb9ad1ea444f534ed456f0b 158515Author: Peter Hutterer <peter@cs.unisa.edu.au> 158516Date: Tue Oct 16 15:50:07 2007 +0930 158517 158518 xfree86: remove XI86_SHARED_POINTER flag. 158519 158520 Not needed anymore. By default, all devices are slaves of first master device. 158521 158522commit c7b878b90406781c97db751a9b454e2b6baee0e1 158523Author: Peter Hutterer <peter@cs.unisa.edu.au> 158524Date: Tue Oct 16 15:07:31 2007 +0930 158525 158526 dix: don't treat VCK/VCP separately in DevHasCursor, IsPointer/KeyboardDevice. 158527 158528commit e9f149fb56747f7eaa0f714dce97b067b007c47e 158529Author: Peter Hutterer <peter@cs.unisa.edu.au> 158530Date: Tue Oct 16 14:24:20 2007 +0930 158531 158532 Fix up detritus from removing GetPairedPointer/Keyboard. 158533 158534commit 840bde3d32f3627dfc3d7d8b6564a61a8014933b 158535Author: Peter Hutterer <peter@cs.unisa.edu.au> 158536Date: Tue Oct 16 13:09:35 2007 +0930 158537 158538 dix: GetPairedDevice: return dev paired with master for slave devices. 158539 158540commit 6c259a08d917f94fd7381453b625b07826d3ef9c 158541Author: Peter Hutterer <peter@cs.unisa.edu.au> 158542Date: Tue Oct 16 13:09:19 2007 +0930 158543 158544 include: remove unused GetPairedPointer/Keyboard declarations. 158545 158546commit fe97f7c54a1b42acd542696b6cdc9e83e89548f3 158547Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158548Date: Mon Oct 15 22:46:08 2007 -0400 158549 158550 registry: Add some missing #include's. 158551 158552commit 9ecbbf198f4cec398897736e173e7e8c56bf6f94 158553Author: Peter Hutterer <peter@cs.unisa.edu.au> 158554Date: Sun Oct 14 08:00:06 2007 +0930 158555 158556 dix: adjust PickPointer and PickKeyboard to get the first master device. 158557 158558 Simplifies it a lot, since we always have at least one master device 158559 available at all times, so less mucking around. 158560 158561commit 0c5f65ecd3ad11fbdb1cab3cb1d0eb4f33bb4e35 158562Author: Peter Hutterer <peter@cs.unisa.edu.au> 158563Date: Sun Oct 14 07:57:24 2007 +0930 158564 158565 dix: don't allow slave devices as ClientPointer 158566 158567commit 9ccc9ab6f29f68298b68cdb5c9b4bd57a095f05a 158568Author: Peter Hutterer <peter@cs.unisa.edu.au> 158569Date: Sun Oct 14 07:53:00 2007 +0930 158570 158571 dix: don't call CloseDevice on VCP/VCK separately. 158572 158573commit e4fd981b48723b77a6c1a528638a771b9cc35472 158574Author: Peter Hutterer <peter@cs.unisa.edu.au> 158575Date: Sun Oct 14 07:52:11 2007 +0930 158576 158577 dix: check for isMaster before calling DeviceCursorCleanup. 158578 158579commit 7503d1340726e0a529f04304d177ed2ceb8fbc91 158580Author: Peter Hutterer <peter@cs.unisa.edu.au> 158581Date: Sun Oct 14 07:50:13 2007 +0930 158582 158583 dix: don't check VCK and VCP separately when looking for device ids. 158584 158585 Both are part of the device list again, so we cover them there. 158586 158587commit 032e906711202d376af95b37bb0cdf14a3648256 158588Author: Peter Hutterer <peter@cs.unisa.edu.au> 158589Date: Tue Oct 16 11:57:42 2007 +0930 158590 158591 dix: Remove GetPairedPointer, GetPairedKeyboard in favour of GetPairedDevice. 158592 158593commit 853ea337bdad17f8f6ec7d940de14ce2cbbbf93e 158594Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158595Date: Mon Oct 15 22:13:02 2007 -0400 158596 158597 registry: Register XvMC extension protocol names. 158598 158599commit 58c3240fcbec23aad122e1c340f6bb6d3b18f779 158600Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158601Date: Mon Oct 15 22:09:47 2007 -0400 158602 158603 registry: Register MIT-SCREEN-SAVER extension protocol names. 158604 158605commit 3e07e73fefc100e491d1e465cb162373d1d82425 158606Author: Peter Hutterer <peter@cs.unisa.edu.au> 158607Date: Tue Oct 16 11:34:29 2007 +0930 158608 158609 dix: remove pairing/attachment from InitAndStartDevices. 158610 158611 If we enabled in the correct order, this has all been done already. 158612 158613commit 4e274e90e16b1d954391e1af3e2074fb10f70ee7 158614Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158615Date: Mon Oct 15 22:04:20 2007 -0400 158616 158617 registry: Register SHAPE extension protocol names. 158618 158619commit 9f597f6c87e0b14cc382d8e5929e42f822db4329 158620Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158621Date: Mon Oct 15 22:01:20 2007 -0400 158622 158623 registry: Register SYNC extension protocol names. 158624 158625commit 38baac71bdbb8c7e882e3e39133615cfed894a6b 158626Author: Peter Hutterer <peter@cs.unisa.edu.au> 158627Date: Sun Oct 14 04:21:42 2007 +0930 158628 158629 dix: Set bidirectional pairing, don't allow pairing with already paired devs. 158630 158631commit 70efd3d06a15093661933bda4ec21e306dece4a4 158632Author: Peter Hutterer <peter@cs.unisa.edu.au> 158633Date: Tue Oct 16 11:30:32 2007 +0930 158634 158635 dix: fix up Activate/Enable/Disable device. 158636 158637 Set isMaster for VCP/VCK. 158638 Init sprites for master pointer devices. 158639 Pair master kbds with master pointers (1:1 pairing!). 158640 Attach other devices to VCP/VCK. 158641 158642commit b504678ba5407a6fd8d47d051305f7c3d5606dfe 158643Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158644Date: Mon Oct 15 21:54:25 2007 -0400 158645 158646 registry: Register APPGROUP extension protocol names. 158647 158648commit 204f2dc89ef662b57400b128c30c15e8cf32f323 158649Author: Peter Hutterer <peter@cs.unisa.edu.au> 158650Date: Sun Oct 14 04:13:06 2007 +0930 158651 158652 include: add "isMaster" field to DeviceIntRec. 158653 158654 Set to TRUE for master devices. 158655 158656 Also fixing up comment for paired field in spriteInfo, will be set 158657 bidirectional from now on. 158658 158659commit bd7d5255ce4865b684f7d8bcf80ba9872a1af22d 158660Author: Peter Hutterer <peter@cs.unisa.edu.au> 158661Date: Tue Oct 16 11:18:31 2007 +0930 158662 158663 dix: add AttachDevice, needed to attach a slave device to a master device. 158664 158665 For now, we don't allow attaching slaves to other slaves, and we don't allow 158666 pairing slaves with other slaves. 158667 Pairing is for master keyboard->master pointer only. 158668 Attaching is for slave device->master device only. 158669 158670commit 6ec35a8cf539c900b334dd6df146b394f54e3706 158671Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158672Date: Mon Oct 15 21:46:56 2007 -0400 158673 158674 registry: Register BigRequests extension protocol names. 158675 158676commit 1254cc399c53eadcc32eeabf69990ed2526c7ae0 158677Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158678Date: Mon Oct 15 21:43:06 2007 -0400 158679 158680 registry: Register Fontcache extension protocol names. 158681 158682commit e987648cf2c21dcbd77dd9a71793090a48e4f521 158683Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158684Date: Mon Oct 15 21:38:56 2007 -0400 158685 158686 registry: Register EVI extension protocol names. 158687 158688commit be1565f6b8fb09eba9941d6f7c485bf5fb25fe7a 158689Author: Peter Hutterer <peter@cs.unisa.edu.au> 158690Date: Sun Oct 14 01:53:54 2007 +0930 158691 158692 dix: Set core devices to ProcessOtherEvents, set sendCore and XKB. 158693 158694 Even the virtual core devices should send through ProcessOtherEvents. 158695 158696commit ce9bf9a19185a36ac2f7ae75acd320ab8d03d247 158697Author: Peter Hutterer <peter@cs.unisa.edu.au> 158698Date: Sun Oct 14 01:50:40 2007 +0930 158699 158700 include: add "master" field to DeviceIntRec. 158701 158702 "master" points to the device this device is attached to. Event sent by the 158703 device will also be routed through the master. 158704 master and spriteOwner are mutually exclusive. 158705 158706commit f6226d3bfe1515058e2092e8662ae87825501209 158707Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158708Date: Mon Oct 15 21:35:12 2007 -0400 158709 158710 registry: Register TOG-CUP extension protocol names. 158711 158712commit 7e182a5d89d618e20dcc77850131690733322d39 158713Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158714Date: Mon Oct 15 21:13:39 2007 -0400 158715 158716 registry: Register MIT-MISC extension protocol names. 158717 158718commit 32f6171862461d17ebea58a2fb6ddd16ac71358c 158719Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158720Date: Mon Oct 15 21:10:14 2007 -0400 158721 158722 registry: Register XF86Bigfont extension protocol names. 158723 158724commit 12766c5b5ffdab95255a63b2c8421ee773fd43b5 158725Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158726Date: Mon Oct 15 21:00:46 2007 -0400 158727 158728 registry: Register Xv extension protocol names. 158729 158730commit 35ae03871af88b2f420dd83448011a077852d7a0 158731Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158732Date: Mon Oct 15 20:50:26 2007 -0400 158733 158734 registry: Register XC-MISC extension protocol names. 158735 158736commit 32fe282d5b8306514d641e15bc6d9fd4ab360977 158737Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158738Date: Mon Oct 15 20:45:18 2007 -0400 158739 158740 registry: Register XTest extension protocol names. 158741 158742commit 3877faf7d9fe00ed634077e38a198ae4b91a2bb4 158743Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158744Date: Mon Oct 15 20:35:13 2007 -0400 158745 158746 registry: Register Multibuffer extension protocol names. 158747 158748commit 16764a2d299c7c0c98002aadd52ab4a1a36758c3 158749Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158750Date: Mon Oct 15 20:29:23 2007 -0400 158751 158752 registry: Register DPMS extension protocol names. 158753 158754commit f077578e42eee424b0e534774574c84af9d6f85b 158755Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158756Date: Mon Oct 15 19:43:02 2007 -0400 158757 158758 registry: Register XPrint extension protocol names. 158759 158760commit 5c8b1a91726817816d20faefad21c7a68ab634cc 158761Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158762Date: Mon Oct 15 19:35:04 2007 -0400 158763 158764 registry: Register Resource extension protocol names. 158765 158766commit 48891d5696f56711f23743cb03be39cf6b26c522 158767Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158768Date: Mon Oct 15 19:32:01 2007 -0400 158769 158770 registry: Register EVIE extension protocol names. 158771 158772commit 2c9646ad4e65bb061d910c9e2b1a8a978f21fa17 158773Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158774Date: Mon Oct 15 19:18:03 2007 -0400 158775 158776 registry: Register SHM extension protocol names. 158777 158778commit 8964c6d8e14ae47798762191e359b2bf138ca32e 158779Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158780Date: Mon Oct 15 19:10:51 2007 -0400 158781 158782 registry: Register RENDER extension protocol names. 158783 158784commit c827db57e4d9ca14c82b099dcfc9b7a0c0b5ba0a 158785Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158786Date: Mon Oct 15 19:06:01 2007 -0400 158787 158788 registry: Register RANDR extension protocol names. 158789 158790commit 20db50b4c44a14f7eeac2b1de17ada68482521da 158791Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158792Date: Mon Oct 15 18:56:59 2007 -0400 158793 158794 registry: Register DAMAGE extension protocol names. 158795 158796commit b38a91993364aa80cfd99721e319e1458d9fb760 158797Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158798Date: Mon Oct 15 18:50:06 2007 -0400 158799 158800 registry: Register XTrap extension protocol names. 158801 158802commit 106758893b68033f14f69c4ee6591fb6a149ba37 158803Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158804Date: Mon Oct 15 18:31:40 2007 -0400 158805 158806 registry: Register XFixes extension protocol names. 158807 158808commit ea09c9acc8f0d5577f54c864ff88b7f03d93b2f4 158809Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158810Date: Mon Oct 15 18:25:12 2007 -0400 158811 158812 registry: Register Record extension protocol names. 158813 158814commit 2e1e5be1d9067816525aa13a1d818e8ca6899599 158815Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158816Date: Mon Oct 15 18:18:57 2007 -0400 158817 158818 registry: Register DBE extension protocol names. 158819 158820commit 4c3285c883cc50a91bc5262bbc9d073d816f860a 158821Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158822Date: Mon Oct 15 18:04:47 2007 -0400 158823 158824 registry: Register WINDOWSWM extension protocol names. 158825 158826commit 3815284e899b61731b6a63c4ba14c5d773e24eb6 158827Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158828Date: Mon Oct 15 17:57:59 2007 -0400 158829 158830 registry: Register XF86DGA extension protocol names. 158831 158832commit 2cd1b32b77e0ceeaccb3f01c4ac13a97c557668c 158833Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158834Date: Mon Oct 15 17:37:58 2007 -0400 158835 158836 registry: Register XF86Misc extension protocol names. 158837 158838commit 960677e876c068400fb45e1764bb5470cd8c389f 158839Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158840Date: Mon Oct 15 17:30:00 2007 -0400 158841 158842 registry: Register XF86VidMode extension protocol names. 158843 158844commit b7786724080fd3928ef7b8c294346661d7ffd90b 158845Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158846Date: Mon Oct 15 17:15:34 2007 -0400 158847 158848 registry: Register XF86DRI extension protocol names. 158849 158850commit eee46b4681ec55297604b0425705f2b18381f7ca 158851Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158852Date: Mon Oct 15 17:06:22 2007 -0400 158853 158854 registry: Register APPLEWM extension protocol names. 158855 158856commit b9f5ab98c8dea36dcce1ad15fd2e059a77e77c39 158857Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158858Date: Mon Oct 15 16:18:37 2007 -0400 158859 158860 registry: Register XINERAMA extension protocol names. 158861 158862commit 3464b419230c6d17e940d967b567c5d2cb22d232 158863Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158864Date: Mon Oct 15 15:45:12 2007 -0400 158865 158866 registry: Register APPLEDRI extension protocol names. 158867 158868commit 32f3f5a1e7654f8bb43ea16b9227b3994e616739 158869Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158870Date: Mon Oct 15 15:32:17 2007 -0400 158871 158872 registry: Register DMX extension protocol names. 158873 158874commit 166ef972febc00c665e1d5aeb68e75d7bbcf9879 158875Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158876Date: Mon Oct 15 15:22:02 2007 -0400 158877 158878 registry: Register composite extension protocol names. 158879 158880commit a5cf3f21f712e46dbf9bca289e67be75f2b531d3 158881Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158882Date: Mon Oct 15 14:27:32 2007 -0400 158883 158884 registry: Register XKB extension protocol names. 158885 158886commit 5277a6ff589b5ddb475b90e1aaf5dbd9172d9711 158887Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158888Date: Mon Oct 15 13:38:35 2007 -0400 158889 158890 registry: Register Input extension protocol names. 158891 158892commit 8b5078b7d9ec08a588a78eb49096b698c6f4916b 158893Author: Kristian Høgsberg <krh@redhat.com> 158894Date: Mon Oct 15 13:10:10 2007 -0400 158895 158896 Fix byte swapping for GetFBConfigs. 158897 158898commit c922d2eebe29f08f463ee76293dc9042712fb21c 158899Author: George Sapountzis <gsap7@yahoo.gr> 158900Date: Sun Oct 14 20:40:18 2007 +0300 158901 158902 glx: drop duplicate GetDrawableAttributesSGIX declarations. 158903 158904 They are officially autogenerated in indirect_dispatch.h now. 158905 158906commit 3d4eb17b38dcb1468493f3686dc5ea3623ef9a73 158907Author: Ben Byer <bbyer@bbyer.local> 158908Date: Sun Oct 14 18:07:03 2007 -0700 158909 158910 mass change from #ifdef i386 to #ifdef __i386__ to conform to ANSI 158911 158912commit ec0fc012e91e703bb399a380df2912f71957a220 158913Author: Kristian Høgsberg <krh@redhat.com> 158914Date: Sun Oct 14 14:59:12 2007 -0400 158915 158916 Fix GLX byteswapping. 158917 158918commit 927757e1028f45f7fd94b9a2ab35567e0f34b2a8 158919Merge: ccda4b66b 17ffc34ad 158920Author: Kristian Høgsberg <krh@redhat.com> 158921Date: Fri Oct 12 19:28:39 2007 -0400 158922 158923 Merge branch 'master' of git://git.freedesktop.org/git/xorg/xserver 158924 158925commit ccda4b66bdfc179e661b24adc8ec72bb7db75b2b 158926Author: Kristian Høgsberg <krh@redhat.com> 158927Date: Mon Aug 27 14:43:48 2007 -0400 158928 158929 Implement ChangeDrawableAttributes so we can support GLX_EVENT_MASK. 158930 158931 We never need to actually send out the clobber event, so this should be 158932 sufficient for GLX 1.3. 158933 158934commit 695eb8b2e88abc9fa3a76d8da48c3214c7dd1f08 158935Author: Kristian Høgsberg <krh@redhat.com> 158936Date: Mon Aug 27 14:23:50 2007 -0400 158937 158938 Implement GLX pbuffers. 158939 158940commit 516c181f57367847c3f317f8f8f5cc3211026f4c 158941Author: Kristian Høgsberg <krh@redhat.com> 158942Date: Thu Aug 23 21:40:01 2007 -0400 158943 158944 Add dispatching for GLX_SGIX_pbuffer. 158945 158946commit d7ded953c4d263e3dd3374dc03eea19e80c05bc6 158947Author: Kristian Høgsberg <krh@redhat.com> 158948Date: Thu Aug 23 19:07:52 2007 -0400 158949 158950 Implment GLXPixmaps. 158951 158952commit 526f40434c86548830c4f72940462b6253fe9790 158953Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 158954Date: Fri Oct 12 18:18:00 2007 -0400 158955 158956 NULL is not a valid argument to CreatePicture, please use serverClient 158957 as the client argument if no real client is creating the object. 158958 158959commit 2b0847c7aaf95b6b10f062dfff601032b2430a9b 158960Author: Kristian Høgsberg <krh@redhat.com> 158961Date: Wed Aug 8 10:07:03 2007 -0400 158962 158963 Implement damage reporting for AIGLX. 158964 158965commit 600028305cc047e824b18369a026f89c9eb7e437 158966Author: Kristian Høgsberg <krh@redhat.com> 158967Date: Thu May 31 22:55:47 2007 -0400 158968 158969 Move hyperpipe and swapbarrier extension funcs to screen private struct. 158970 158971 This gets rid of a couple more global arrays and gets the two extensions 158972 more in line with the general extension mechanism. 158973 158974commit 781515bb637de4fd79e3c83817cb6ffa8f2d8a4f 158975Author: Kristian Høgsberg <krh@redhat.com> 158976Date: Wed May 30 23:25:03 2007 -0400 158977 158978 Convert GLX module to use screen private indexes like everything else. 158979 158980commit 72a3d68a2f5abcd09fef1a55e976e1a5731d4b02 158981Author: Kristian Høgsberg <krh@redhat.com> 158982Date: Thu May 10 11:16:50 2007 -0400 158983 158984 Update the AIGLX DRI loader to use the new DRI interface. 158985 158986 This lets us drop a bunch of workarounds and use the new DRI extension 158987 mechanism. 158988 158989commit 17ffc34ad5e9a8e2269afef05f209701f1895c01 158990Author: Matthias Hopf <mhopf@suse.de> 158991Date: Fri Oct 12 15:33:31 2007 +0200 158992 158993 Drivers include miscstruct.h which requires pixman.h. 158994 158995commit eaf0e2a21c2cb14e19852e61a4521b3c240253af 158996Author: David Nusinow <dnusinow@debian.org> 158997Date: Thu Oct 11 22:31:24 2007 -0400 158998 158999 Fix another compiler warning 159000 159001commit a5089af726b6a4f833b95a31274743c994277e20 159002Author: David Nusinow <dnusinow@debian.org> 159003Date: Thu Oct 11 22:24:31 2007 -0400 159004 159005 Fix a warning about the control logic in xchomp() 159006 159007commit cdf29ff45a3cb45573c9d0cb8f82e6ee97953fb5 159008Author: David Nusinow <dnusinow@debian.org> 159009Date: Thu Oct 11 22:23:34 2007 -0400 159010 159011 Remove some unused variables 159012 159013commit 3367091f7fa14497aab40e668cad179e244eef81 159014Author: David Nusinow <dnusinow@debian.org> 159015Date: Thu Oct 11 22:21:38 2007 -0400 159016 159017 Separate choosing driver from the file-based implementation 159018 159019 This clears the implementation out of the way to prepare for development 159020 of a symbol-based resolution scheme 159021 159022commit 3aa41bcb8215c037512ddbd68a3f7bcad3b80a1f 159023Author: David Nusinow <dnusinow@debian.org> 159024Date: Thu Oct 11 21:29:13 2007 -0400 159025 159026 Remove obsolete error message define from parser 159027 159028commit e3e12221111886c4063d2da5d70d3830c56d39e2 159029Author: David Nusinow <dnusinow@debian.org> 159030Date: Thu Oct 11 21:27:07 2007 -0400 159031 159032 Don't bother validating the Device section of the conf file 159033 159034 All the previous tests can now be recovered from if not specified. 159035 159036commit 28ef7f59416677be380561709197b04df0479bef 159037Author: David Nusinow <dnusinow@debian.org> 159038Date: Thu Oct 11 20:56:46 2007 -0400 159039 159040 Re-enable validation of the screen section of xorg.conf 159041 159042 This also fixes a problem where the server can't find the device section 159043 when it is specified in the screen section. 159044 159045commit ea2d4dc468dcebe6d38e676469ec51ed1d43490b 159046Author: David Nusinow <dnusinow@debian.org> 159047Date: Thu Oct 11 20:26:02 2007 -0400 159048 159049 When there's no xorg.conf, use the video driver autoloading function 159050 159051commit fc092334ac0a323b80a9602cb8bf60ca9dee3bfa 159052Author: Eric Anholt <eric@anholt.net> 159053Date: Thu Oct 11 16:48:56 2007 -0700 159054 159055 Bug #10304,12784,11603: Add quirks for several physical size issues. 159056 159057 A lot of EDID writers apparently end up stuffing centimeters (like the 159058 maximum image size field) into the detailed timings, instead of millimeters. 159059 Some of them only get it wrong in one direction. Also, add a quirk to let 159060 us mark the largest 75hz mode as preferred, which will often be used for 159061 EDID 1.0 CRTs. 159062 159063commit 6adeba17301a309be2f34cd51eca84a13d5503fd 159064Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 159065Date: Thu Oct 11 14:17:17 2007 -0400 159066 159067 dix: Add a new "registry" mechanism for registering string names of things. 159068 Supports protocol requests, events, and errors, and resource names. 159069 Modify XRES extension to use it. 159070 159071commit 45cc03726b49f6ad4afe6c3fb4ad65d1051928b4 159072Author: Matthias Hopf <mhopf@suse.de> 159073Date: Thu Oct 11 18:23:25 2007 +0200 159074 159075 Make mode checking more tolerant like in pre-RandR times. 159076 159077commit 8f23d40068151ad85cde239d07031284f0b2c4dc 159078Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 159079Date: Wed Oct 10 19:56:03 2007 -0400 159080 159081 xace: move the property deletion hook inside the DeleteProperty function. 159082 159083commit 473bc6ec4c59e1a962b0b897c449a69aa5064ab0 159084Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 159085Date: Wed Oct 10 19:43:12 2007 -0400 159086 159087 xace: remove the special-cased "ignore" functionality from the property code. 159088 159089 There will be no more faking of Success to hide things. XACE does not 159090 provide polyinstantiation. 159091 159092commit 7e9e01a4a34fa45521067d43c5bbff942dd5d51a 159093Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 159094Date: Wed Oct 10 17:40:22 2007 -0400 159095 159096 dix: pass a valid ClientPtr to SetFontPath in all cases. 159097 159098commit 8825d36a38c76ff8faf409c853061be2bd33430d 159099Author: David Nusinow <dnusinow@debian.org> 159100Date: Tue Oct 9 22:49:07 2007 -0400 159101 159102 Allow the user to not have a screen section 159103 159104 If none is present, a default one will be created. This will be attached 159105 to either the first device section in the xorg.conf (allowing you to 159106 specify something like using EXA without having a screen section) or a 159107 default screen section if none is present in the file. 159108 159109commit 8665cf3327b713f334b0483593a924c197f892f2 159110Author: David Nusinow <dnusinow@debian.org> 159111Date: Tue Oct 9 22:45:15 2007 -0400 159112 159113 Don't require that the screen explicitly attach the device section 159114 159115 This will allow the screen to not explicitly have a device section. If 159116 this is the case and there is a device section in the xorg.conf, the first 159117 one will be used. If there is no device section at all, a default one will 159118 be created that loads the automatically determined module. 159119 159120commit 37898b824fdc94735495e3494aa2b9a681d477b9 159121Author: David Nusinow <dnusinow@debian.org> 159122Date: Tue Oct 9 22:11:58 2007 -0400 159123 159124 Use the best guess heuristic as a fallback for autoloading 159125 159126 This prevents breaking the current behavior 159127 159128commit 847fbbf4d25132e6c1f99fcf82fe757dbca94a34 159129Author: David Nusinow <dnusinow@debian.org> 159130Date: Tue Oct 9 22:07:49 2007 -0400 159131 159132 Fix autoloading of drivers for pci-rework 159133 159134commit 6033d8150be3a115b90226eaa42f237bb0cf3369 159135Author: David Nusinow <dnusinow@debian.org> 159136Date: Tue Oct 9 21:17:27 2007 -0400 159137 159138 First pass at improved video driver autoloading 159139 159140 This is what we're currently shipping in Debian. Enables the ability for 159141 drivers to ship a text file listing PCI ID's they support, and have the 159142 server read them on startup when no driver is specified. This works, but 159143 isn't the final solution. 159144 159145commit 81d7b81146224f2b83278f5e21b3f9a36f30bd56 159146Author: David Nusinow <dnusinow@debian.org> 159147Date: Sat Aug 25 15:00:30 2007 -0400 159148 159149 Load intel instead of i810 when autoconfiguring 159150 159151commit fdef7be5c8d5989e0aa453d0a5b86d0a6952e960 159152Author: Alan Coopersmith <alan.coopersmith@sun.com> 159153Date: Tue Oct 9 18:44:04 2007 -0700 159154 159155 Sun bug 6589829: include zoneid of shm segment in access policy for MIT-SHM 159156 159157 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6589829> 159158 <http://www.opensolaris.org/os/community/arc/caselog/2007/500/> 159159 159160commit 49c8285e38e0e71c9ea43df63add08a4f2564ed1 159161Author: Christian Weisgerber <naddy@freebsd.org> 159162Date: Tue Oct 9 15:25:16 2007 -0700 159163 159164 FreeBSD/alpha server build requires libio (missed in modularization). 159165 159166commit 042e16458b751ad0fcfa1ebbdd74d7d48036b22a 159167Author: Peter Hutterer <peter@cs.unisa.edu.au> 159168Date: Tue Oct 9 23:07:51 2007 +0930 159169 159170 dix: change a few inputInfo.keyboard over to PickKeyboard() 159171 159172commit 6aab6a6e2ae0ef5fdedae09cf2cdc86f382e3a8a 159173Author: Dodji Seketeli <dodji@openedhand.com> 159174Date: Mon Oct 8 13:25:38 2007 +0200 159175 159176 EXA: dude, don't validation driver hooks twice ! 159177 159178commit d42909b23a9be64a26aa349c1d6dcd8b228bffa7 159179Author: Peter Hutterer <peter@cs.unisa.edu.au> 159180Date: Fri Oct 5 13:04:10 2007 +0930 159181 159182 dix: when disabling a device, make sure all paired devices are re-paired. 159183 159184 We re-pair them with the VCP, not a real device! If we would do otherwise, 159185 somebody may change our keyboard focus and thus get us typing where we don't 159186 want to type. 159187 159188commit ff2351246da30b56a3a8f90654993c41b8eb3921 159189Author: Peter Hutterer <peter@cs.unisa.edu.au> 159190Date: Fri Oct 5 12:19:03 2007 +0930 159191 159192 dix: When pairing, only check for access right if client is set. 159193 159194 If the pairing client is not set, then the pairing is initiated internally 159195 (e.g. when a new keyboard device is configured). In this case we _must_ pair 159196 regardless of who is the pairing client. 159197 159198commit 4611f9568cdadf9c00f4b7ca4c77c8c6f07e94f8 159199Merge: ab11bad54 3549a1282 159200Author: Dodji Seketeli <dodji@openedhand.com> 159201Date: Thu Oct 4 13:07:50 2007 +0200 159202 159203 updated from 'master' 159204 159205commit ab11bad54707941eb41be62c025b983760ce3900 159206Author: Dodji Seketeli <dodji@openedhand.com> 159207Date: Thu Oct 4 13:05:29 2007 +0200 159208 159209 Xephyr: remove a potential crasher 159210 159211 * hw/kdrive/ephyr/hostx.c: 159212 (hostx_has_dri): be more defensive. 159213 159214commit 06c8977966e76255ce7798f4839ef6e3530264a8 159215Author: Dodji Seketeli <dodji@openedhand.com> 159216Date: Thu Oct 4 13:01:24 2007 +0200 159217 159218 Xephyr: better handling of the -nodri option 159219 159220 * hw/kdrive/ephyr/ephyr.c: 159221 (ephyrInitScreen): try and detect when the host has no 159222 DRI support. In that case, switch to the -nodri behaviour. 159223 When in the -nodri case, make sure not to skip glx visual 159224 initialisation. 159225 159226commit ea94f59e9b8824b1a638c63e06ca16f0efd43869 159227Author: Dodji Seketeli <dodji@openedhand.com> 159228Date: Thu Oct 4 12:54:26 2007 +0200 159229 159230 Xephyr: don't disable glx visual init by default 159231 159232 * hw/kdrive/ephyr/ephyrinit.c: 159233 (ddxProcessArgument): disabling visual init here 159234 is bad because it gets disabled even when we want 159235 to use software GL, leading to Xephyr :1 -nodri 159236 crashing in mesa. 159237 159238commit a7f210e6fcda14eae4de64186904c6c676c758ee 159239Author: Dodji Seketeli <dodji@openedhand.com> 159240Date: Thu Oct 4 12:51:08 2007 +0200 159241 159242 Xephyr: fix a compiler warning 159243 159244commit 5d6eac1251f1e17baa74d0893bb225b775e8ec2e 159245Author: Dodji Seketeli <dodji@openedhand.com> 159246Date: Thu Oct 4 12:47:05 2007 +0200 159247 159248 Xephyr: remove an unused variable 159249 159250commit 8018733c39c283a931df424dcfd11e7aefbcd8cf 159251Author: Dodji Seketeli <dodji@openedhand.com> 159252Date: Thu Oct 4 12:06:19 2007 +0200 159253 159254 Xephyr: better compilation without GLX 159255 159256 * hw/kdrive/ephyr/ephyrinit.c: protect GLX related code with GLXEXT 159257 macro 159258 159259commit 3549a1282365e69e70c7c2546cfa7d25923cce31 159260Author: Michel Dänzer <michel@tungstengraphics.com> 159261Date: Thu Oct 4 11:38:01 2007 +0200 159262 159263 EXA: Disable 1x1 pixmap fill optimization for drivers that handle pixmaps. 159264 159265 This reverts commit 1365aeff5499a051375e43a9fcbf54733ac93929. 159266 159267 It defeated the optimization for drivers that don't provide a CreatePixmap 159268 hook. The optimization makes no sense for drivers that do anyway, so disable 159269 it for them completely. 159270 159271commit 604ebb5a6de372e6a8a96e0ee997db7929126860 159272Merge: 1365aeff5 4ba76a7e2 159273Author: Dodji Seketeli <dodji@openedhand.com> 159274Date: Wed Oct 3 16:13:16 2007 +0200 159275 159276 Merge Xephyr-XV/GL stuff into master 159277 159278commit 4ba76a7e2b62d26f43c0e670de571afb75ec92f4 159279Author: Dodji Seketeli <dodji@openedhand.com> 159280Date: Wed Oct 3 13:03:34 2007 +0200 159281 159282 Xephyr: port XV/GL stuff of the new multiscreen architecture 159283 159284 We can now launch GL or XV apps in any of the 159285 Xephyr screens we want. 159286 159287 * hw/kdrive/ephyr/hostx.c,h: 159288 (hostx_get_window): 159289 (hostx_create_window): make these functions be screen 159290 number aware. 159291 * hw/kdrive/ephyr/XF86dri.c : fix some compiler warnings. 159292 * hw/kdrive/ephyr/ephyrdri.c: 159293 (ephyrDRIQueryDirectRenderingCapable), 159294 (ephyrDRIOpenConnection), 159295 (ephyrDRIAuthConnection), 159296 (ephyrDRICloseConnection), 159297 (ephyrDRIGetClientDriverName), 159298 (ephyrDRICreateContext), 159299 (ephyrDRIDestroyContext), 159300 (ephyrDRICreateDrawable), 159301 (ephyrDRIGetDrawableInfo), 159302 (ephyrDRIGetDeviceInfo): in all those functions, don't forward 159303 the screen number we receive - from the client - to the host X. 159304 We (Xephyr) are always targetting the same X display screen, which is 159305 the one Xephyr got launched against. So we enforce that in the code. 159306 * hw/kdrive/ephyr/ephyrdriext.c: 159307 (EphyrMirrorHostVisuals): make this duplicate the visuals of the host X 159308 default screen into a given Xephyr screen. This way we have a chance 159309 to update the visuals of all Xephyr screen to make them mirror those 159310 of the host X. 159311 (many other places): specify screen number where required by the api 159312 change in hostx.h. 159313 * hw/kdrive/ephyr/ephyrglxext.c: specify screen number where required 159314 by the api change in hostx.h 159315 * hw/kdrive/ephyr/ephyrhostglx.c: don't forward the screen number we 159316 receive - from the client - to the host X. 159317 We (Xephyr) are always targetting the same 159318 X display screen, which is 159319 the one Xephyr got launched against. So we enforce that in the code. 159320 * hw/kdrive/ephyr/ephyrhostvideo.c,h: take in account the screen number received 159321 from the client app. This is useful to know on which Xephyr screen we 159322 need to display video stuff. 159323 * hw/kdrive/ephyr/ephyrvideo.c: update this to reflect the API change 159324 in hw/kdrive/ephyr/ephyrhostvideo.h. 159325 (ephyrSetPortAttribute): when parameters are not valid 159326 - they exceed their validity range - send them to the host anyway 159327 and do not return an error to clients. 159328 Some host expose buggy validity range, so rejecting client for that 159329 is too harsh. 159330 159331commit ab88cb8d2e7c1410f9ed2be928b38f176b132e11 159332Author: Peter Hutterer <peter@cs.unisa.edu.au> 159333Date: Wed Oct 3 18:08:33 2007 +0930 159334 159335 Xext: add missing geint.h to Makefile.am (un-breaks make dist) 159336 159337commit 9f2b493e34e93881101f31e631901d3fe56da4f0 159338Author: Peter Hutterer <peter@cs.unisa.edu.au> 159339Date: Thu Sep 27 15:35:41 2007 +0930 159340 159341 xkb: remove some warning comments. 159342 159343 Obsolete with 340911d7243a7f1095d79b5b2dcfa81b145c2474. 159344 159345commit 1eebb03a3190947a8102f2ddc73766cf98d34c84 159346Author: Peter Hutterer <peter@cs.unisa.edu.au> 159347Date: Wed Oct 3 15:18:17 2007 +0930 159348 159349 dix: ignore passive grab if the client already has a grab on the device. 159350 159351 In some cases a button press may activate a passive core grab. If the client 159352 owning the passive grab already has a core grab on another device, don't 159353 actually activate it. Otherwise the client gets two simultaneous passive 159354 core grabs, and may never ungrab the device again (only if the other grab uses 159355 GrabModeSync). 159356 159357 Reproducable: fire up gnome-session, open up gnome-terminal. Click with the 159358 ClientPointer onto the window decoration, then click with another pointer onto 159359 an application icon in the panel. Drag the icon out, release the button and 159360 voila - you just lost your second mouse. 159361 159362commit 0b485067823620b5dbd9ef2b3e13bd35ad5a4410 159363Author: Peter Hutterer <peter@cs.unisa.edu.au> 159364Date: Wed Oct 3 14:22:55 2007 +0930 159365 159366 dix: send NotifyGrab/NotifyUngrab focus events regardless of semaphore state. 159367 159368 This is just papering over a problem. The whole focus system needs to be 159369 revised. 159370 159371commit 05106ac9839102c0e4a3ce5d9d83d19abf129f8a 159372Author: Peter Hutterer <peter@cs.unisa.edu.au> 159373Date: Wed Oct 3 11:33:10 2007 +0930 159374 159375 dix: change Enter/Leave semaphore handling to accommodate for NotifyGrab. 159376 159377 This is a half-assed attempt at getting rid of some enter-leave problems. When 159378 a grab is activated, the events didn't get sent before, leading to interesting 159379 results. This commit papers over it but doesn't actually fix it properly. The 159380 whole enter/leave (focusin/out) structure needs to be ripped out and changed 159381 for multiple devices. 159382 159383commit 1365aeff5499a051375e43a9fcbf54733ac93929 159384Author: Dave Airlie <airlied@redhat.com> 159385Date: Wed Oct 3 12:00:16 2007 +1000 159386 159387 exa: direct access to the pixmap sys ptr is bad if the pixmap isn't mapped 159388 159389commit 566dd3b7d789ba60d0adf33b3f729cfb02ff33cd 159390Author: Dave Airlie <airlied@redhat.com> 159391Date: Wed Oct 3 11:59:52 2007 +1000 159392 159393 exa: add new flags to denote driver handles all pixmap migration/hiding 159394 159395commit 1df1fee82d3565f6d8cfb91ce25a81c23a10a4b5 159396Author: Dave Airlie <airlied@redhat.com> 159397Date: Wed Oct 3 11:56:04 2007 +1000 159398 159399 exa: make sure we set fb_ptr to NULL 159400 159401commit a358b87f45ce75e5d013fc904a07dfe394f74387 159402Author: Eric Anholt <eric@anholt.net> 159403Date: Tue Oct 2 13:13:51 2007 -0700 159404 159405 Just link against libcrypto instead of relying on openssl pkg-config. 159406 159407 Also fix incorrect library inclusion by kdrive which broke the build. 159408 159409commit 439edc768eea17667846ce573c843b8377e0dfb4 159410Merge: c8ccf469d 93ae6fe18 159411Author: Eric Anholt <eric@anholt.net> 159412Date: Tue Oct 2 12:14:04 2007 -0700 159413 159414 Merge branch 'glyph-pixmaps' 159415 159416 Conflicts: 159417 159418 configure.ac 159419 exa/exa_render.c 159420 159421commit 59cebcd2e9302d15a52588ecafbbc2d2c5ae3a6c 159422Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 159423Date: Tue Oct 2 13:39:25 2007 -0400 159424 159425 xace: add creation hook for new input devices. 159426 159427 Unfortunately, more information is needed to properly label the device. 159428 This will come from the configuration file, the hotplug messages, etc. 159429 It will either have to be passed into this function, 159430 or this hook moved down into the callers. 159431 159432commit b77d272d7555c1e0f176ee74b8717030a6d6c7b0 159433Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 159434Date: Tue Oct 2 13:21:53 2007 -0400 159435 159436 xace: add hooks + new access codes: XTEST extension 159437 159438commit c8ccf469dcb2d7609b23faa8b4999c436263051e 159439Author: Keith Packard <keithp@koto.keithp.com> 159440Date: Tue Oct 2 08:59:40 2007 -0700 159441 159442 Do not build hw/xfree86/os-support/bus/ix86Pci.c on Linux. 159443 159444 Linux has custom PCI bus support functions, so this generic code is 159445 unnecessary. 159446 159447commit dcb4db1bf5b6fcfaed4e7162eb6c7a3a509e14bd 159448Author: Dodji Seketeli <dodji@openedhand.com> 159449Date: Wed Sep 19 15:46:09 2007 +0200 159450 159451 Xephyr: glx fixlet 159452 159453 * hw/kdrive/ephyr/ephyrhostglx.c: 159454 (ephyrHostGLXQueryVersion): properly query 159455 the host for its gxl version. 159456 159457commit 6a435b00003fb00930299a0e0810c93afc23a72e 159458Author: Dodji Seketeli <dodji@openedhand.com> 159459Date: Wed Sep 19 15:35:51 2007 +0200 159460 159461 Xephyr: check presence of extensions in host X 159462 159463 * hw/kdrive/ephyr/hostx.c,h: 159464 (hostx_has_xshape), 159465 (hostx_has_glx), 159466 (hostx_has_dri): added these new entry points 159467 * hw/kdrive/ephyr/ephyrdriext.c: 159468 (ephyrDRIExtensionInit): 159469 check presence of DRI and XShape extensions before 159470 trying to use them. 159471 * hw/kdrive/ephyr/ephyrglxext.c: 159472 (ephyrHijackGLXExtension): 159473 check presence of glx extension before we use it. 159474 159475commit 2b217fc055300d4c709dd27c4e8ec8166183993b 159476Author: Dodji Seketeli <dodji@openedhand.com> 159477Date: Wed Sep 19 15:00:43 2007 +0200 159478 159479 Xephyr: cleanup XF86dri.c a bit 159480 159481commit f731b9da7dbafad50dd5a1e8da28ce86ffab19b7 159482Author: Dodji Seketeli <dodji@openedhand.com> 159483Date: Tue Sep 18 18:41:18 2007 +0200 159484 159485 Xephyr: add -nodri and -noxv switches. 159486 159487 * hw/kdrive/ephyr/ephyr.c,ephyrinit.c:: 159488 add -nodri and -noxv to disable either DRI or XV. 159489 When -nodri, mesa GL (non acceleretated) works. 159490 159491commit 50674391388b3cfe987782a3ad81b4f240f3db67 159492Author: Dodji Seketeli <dodji@openedhand.com> 159493Date: Tue Sep 18 18:37:49 2007 +0200 159494 159495 Kdrive: use fbcmap_mi.c instead of fbcmap.c 159496 159497 * hw/kdrive/src/Makefile.am: use fb/fbcmap_mi.c 159498 and not fb/fbcmap.c. This allows kdrive to take advantage of 159499 extensions redefining the entry points of micmap.c stuff. 159500 For instance it allows Xephyr to have a working GL, which is not 159501 possible otherwise, because GL redefines mInitVisualsProc 159502 to initialise its visuals. 159503 159504commit a5bd536869f646e4024befa84f6e677b990310bc 159505Author: Dodji Seketeli <dodji@openedhand.com> 159506Date: Tue Sep 18 18:28:25 2007 +0200 159507 159508 fbcmap_mi.c: add a missing api 159509 159510 * fb/fbcmap_mi.c: 159511 (fbSetVisualTypesAndMasks): added this entry point that was missing. 159512 This is useful so that servers using this entry point can use fbcmap_mi.c, 159513 and be obliged to stick to fbcmap.c. Note that fbcmap.c does implement this 159514 entry point. Up to now, kdrive based server could not use fbcmap_mi.c because this 159515 entry point was missing. The will allow Xephyr to properly use GL. 159516 159517commit 6307ee77091041af2d34e93bc164352d4438a808 159518Author: Dodji Seketeli <dodji@openedhand.com> 159519Date: Tue Sep 18 18:22:44 2007 +0200 159520 159521 GLX: allow skipping of visual init wrapping 159522 159523 * GL/glx/glxvisuals.c: added boolean to disable 159524 calling init_visuals(). This gives a chance to Xephyr 159525 to take over visuals manipulation and avoid a crash at 159526 server shutdown in __glXMesaScreenDestroy(), due to the fact 159527 that mesa might sees more visual than what it has actually created in 159528 init_visuals(). It might see more visuals because Xephyr can augment 159529 the number of visuals, dynamically. 159530 * os/utils.c: the boolean is actually defined here. 159531 159532commit ffe9ce6a843443606ad9367eca1645bf5e0635e5 159533Author: Dodji Seketeli <dodji@openedhand.com> 159534Date: Mon Sep 17 22:13:51 2007 +0200 159535 159536 Xephyr: fix a clipping issue xephyr-dri 159537 159538 * hw/kdrive/ephyr/ephyrdri.c: 159539 (ephyrDRIGetDrawableInfo): force the back clipping rects 159540 to equal the front clipping rects. 159541 * hw/kdrive/ephyr/ephyrdriext.c: 159542 (ProcXF86DRIGetDrawableInfo): properly overclip the clipping rects we 159543 got from the client. This bug fixes a clipping rect that was too 159544 small in height, basically. Also fix a possible mem corruption. 159545 * hw/kdrive/ephyr/hostx.c: 159546 (hostx_set_window_geometry): remove a useless XSync 159547 159548commit 79782726617d4b5a4f9b376f21936d035fc870e1 159549Author: Dodji Seketeli <dodji@openedhand.com> 159550Date: Mon Sep 17 18:15:03 2007 +0200 159551 159552 Xephyr: properly clip GL drawables in Xephyr 159553 159554commit 3816d64e50a59e0ef276ddd6cb250cab98a49392 159555Author: Dodji Seketeli <dodji@seketeli.org> 159556Date: Sun Sep 9 15:20:51 2007 +0200 159557 159558 Xephyr: oops, forgot to add ephyrdriext.h 159559 159560commit cf58781eee7d91a3bbca701a2c4523a36f8e0abb 159561Author: Dodji Seketeli <dodji@openedhand.com> 159562Date: Sat Sep 8 00:33:59 2007 +0200 159563 159564 Xephyr: make accelerated glxgears work in Xephyr 159565 159566 * hw/kdrive/ephyr/ephyr.c: 159567 (ephyrInitialize): cleanup ephyrDRI extension init. 159568 remove functions that belongs in ephyrdriext.c . 159569 * hw/kdrive/ephyr/ephyrdri.c: 159570 (ephyrDRICreateDrawable): create the drawable on the host X peer 159571 window, not on the host xephyr main window. 159572 (ephyrDRIGetDrawableInfo): get drawable info of the host X peer 159573 window. 159574 * hw/kdrive/ephyr/ephyrdriext.c: make ephyr DRI extention wrap 159575 a bunch of screen ops so that it can update the host X peer 159576 window whenever DRI bound drawable are moved in Xephyr. 159577 Also code the building blocks of the management of the 159578 host X window peer. 159579 * hw/kdrive/ephyr/hostx.c,h: 159580 (hostx_create_window): added this new entry point 159581 (hostx_destroy_window): ditto 159582 ()hostx_set_window_geometry): ditto 159583 159584commit 6d1e44d3d53b451d466d43197be589d0ecc4b9f6 159585Author: Dodji Seketeli <dodji@openedhand.com> 159586Date: Tue Sep 4 18:28:16 2007 +0200 159587 159588 Xephyr: fix a host X hang. 159589 159590 * hw/kdrive/ephyr/ephyrdri.c: 159591 (ephyrDRIGetDrawableInfo): quickly hook 159592 this into getting the drawable info from the host 159593 X server. For the time being, this only gets the drawable info 159594 of the Xephyr main window in the host. It should really get 159595 the info of a the peer drawable in the host X. So there should be a 159596 peer drawable to begin with. 159597 * hw/kdrive/ephyr/ephyrdriext.c: 159598 (ProcXF86DRIGetDrawableInfo): some cleanups. Properly get the 159599 the drawable info otherwise there is a host X hang. 159600 * hw/kdrive/ephyr/ephyrhostglx.c: do not 159601 (ephyrHostGLXQueryVersion): do not use C bindings of the glx protocol 159602 calls. Some of those actually access DRI context directly, resulting 159603 in the context having three clients. Instead all XF86DRI proto 159604 fowarding request should be coded by hand and only forward the 159605 protocol requests 159606 159607commit a39b57d2f784f14e3198c8506f5eb865bacc2440 159608Author: Dodji Seketeli <dodji@openedhand.com> 159609Date: Fri Aug 31 16:53:18 2007 +0200 159610 159611 Xephyr: add more logging to GLX forwarding 159612 159613 * hw/kdrive/ephyr/ephyrhostglx.c: added more logging. 159614 159615commit 9e192d2118246ad715e23cf5626a038274eb1eaa 159616Author: Dodji Seketeli <dodji@openedhand.com> 159617Date: Fri Aug 31 14:43:39 2007 +0200 159618 159619 Xephyr: better error handling in GLX forwarding 159620 159621 * hw/kdrive/ephyr/ephyrglxext.c: 159622 fixed various logging functions 159623 (ephyrGLXGetStringReal): make sure all the string is sent to clients 159624 including the ending zero. 159625 * hw/kdrive/ephyr/ephyrhostglx.c: 159626 (ephyrHostGLXGetStringFromServer): better error handling. 159627 (ephyrHostGLXSendClientInfo): ditto. 159628 (ephyrHostGLXMakeCurrent): ditto 159629 159630commit 4dd4be99df38b70d43b02d9cb19e09e0ed94462b 159631Author: Dodji Seketeli <dodji@seketeli.org> 159632Date: Thu Aug 30 23:54:49 2007 +0200 159633 159634 Xephyr: Make glxinfo work on the ATI R200 free driver. 159635 159636 * hw/kdrive/ephyr/ephyr.c: 159637 (EphyrDuplicateVisual): when duplicating the 159638 visual, copy the color component masks and the class 159639 from the hostX 159640 (EphyrMirrorHostVisuals): don't mix blue and green mask. 159641 * hw/kdrive/ephyr/ephyrdri.c: add more logs. 159642 (ephyrDRICreateDrawable): actually implement this. 159643 for the moment it creates a DRI drawable for the hostX window, 159644 no matter what drawable this call was issued for. 159645 (ephyrDRIGetDrawableInfo): actually implemented this. 159646 for the moment the drawable info queried for its attrs is the 159647 Xephyr main main window. 159648 * hw/kdrive/ephyr/ephyrdriext.c: 159649 (ProcXF86DRIGetDrawableInfo): properly hook this dispatch 159650 function to the ephyrDRIGetDrawableInfo() function. 159651 * hw/kdrive/ephyr/ephyrglxext.c: add a bunch of GLX implementation hooks 159652 here. Hijack some of the xserver GLX hooks with them. Still need to 159653 properly support byteswapped clients though. 159654 * hw/kdrive/ephyr/ephyrhostglx.c,h: actually implemented the protocol 159655 level forwarding functions used by the GLX entr points in 159656 ephyrglxext.c. Here as well, there are a bunch of them, but we are 159657 far from having implemented all the GLX calls. 159658 * hw/kdrive/ephyr/hostx.c,h: 159659 (hostx_get_window_attributes): added this new entry point 159660 (hostx_allocate_resource_id_peer): added this to keep track of 159661 resource IDs peers: one member of the peer is in Xephyr, the other 159662 is in host X. 159663 (hostx_get_resource_id_peer): ditto. 159664 159665commit 8c78df0ea36abc7d46c4f59670289f4280f75507 159666Author: Dodji Seketeli <dodji@seketeli.org> 159667Date: Thu Aug 30 22:31:47 2007 +0200 159668 159669 Xserver: fix potential crash 159670 159671 * dix/extension.c: 159672 (AddExtensionAlias): don't crash when failed to add 159673 extension alias. 159674 159675commit 024abe825cd702334266f6b8da289907bbd05970 159676Author: Dodji Seketeli <dodji@seketeli.org> 159677Date: Tue Aug 28 15:59:45 2007 +0200 159678 159679 Xephyr: add logging in the DRI extension 159680 159681commit 1dd589410c6c5e8b62b4d15ec681b01452ec68c6 159682Author: Dodji Seketeli <dodji@seketeli.org> 159683Date: Tue Aug 28 15:58:55 2007 +0200 159684 159685 mirror the visuals of the host X at startup. 159686 159687 * hw/kdrive/ephyr/ephyr.c: make Xephyr mirror 159688 the visuals of the host X upon startup. This 159689 is important for GLX client apps. 159690 * hw/kdrive/ephyr/hostx.c,h: add a hostx_get_visuals_info() 159691 to get the visuals of the host X. 159692 159693commit 26da625055a3876c93a7a005520657b971748c4e 159694Author: Dodji Seketeli <dodji@seketeli.org> 159695Date: Tue Aug 28 15:55:05 2007 +0200 159696 159697 proxy the glXGetFBConfigSGIX call 159698 159699 * hw/kdrive/ephyr/ephyrglxext.c: 159700 (ephyrGLXGetFBConfigsSGIX): proxy the GLXGetFBConfigsSGIX call. 159701 It is a vendor extension to get the visual configs as a list of 159702 name/value pairs. 159703 (ephyrHijackGLXExtension): hijack the VendorPriv_dispatch_info 159704 dispatch table to register our implementation of GLXGetFBConfigsSGIX 159705 (ephyrGLXGetFBConfigsSGIXReal): added this where the real 159706 implementation of GLXGetFBConfigsSGIX is. It support bytes swapping. 159707 (ephyrGLXGetFBConfigsSGIX,ephyrGLXGetFBConfigsSGIXSwap): these are 159708 the dispatch entry points. They just call 159709 ephyrGLXGetFBConfigsSGIXReal. 159710 * hw/kdrive/ephyr/ephyrhostglx.c,h: reorganize the proxies to get 159711 visual params from the host so that they clearly support the different 159712 methods of doing so. 159713 159714commit 5af73f98c46dfe5640185ca90feb20c39c6e06fa 159715Author: Dodji Seketeli <dodji@seketeli.org> 159716Date: Fri Aug 24 16:30:50 2007 +0200 159717 159718 EPHYR: Add an experiment proto proxy extension. 159719 159720 * hw/kdrive/ephyr/Makefile.am: add the proxy extension to 159721 ephyr. The proxy extension is an experimental extension that 159722 forwards protocol packets targeted at a given extension to the 159723 host X. 159724 * hw/kdrive/ephyr/ephyr.c: init proxy ext. 159725 * hw/kdrive/ephyr/ephyrhostproxy.c,h: added this new file as part of the 159726 proxy extension. 159727 * hw/kdrive/ephyr/ephyrproxyext.c,h: ditto 159728 * hw/kdrive/ephyr/hostx.c: add the hostx_get_get_extension_info() entry 159729 point. 159730 159731commit c06fa924b4781a35b86e4a78d95ff3e9d95b02d1 159732Author: Dodji Seketeli <dodji@seketeli.org> 159733Date: Thu Aug 23 11:53:02 2007 +0200 159734 159735 XEPHYR: more GLX/DRI proxying work. 159736 159737 * hw/kdrive/ephyr/XF86dri.c: re format this correctly. 159738 Make function decls honour the Ansi-C standard. 159739 * hw/kdrive/ephyr/ephyr.c: protect glx/dri related 159740 extension initialisation with the XEPHYR_DRI 159741 macro. Initialize the GLX ext hijacking 159742 at startup. 159743 * hw/kdrive/ephyr/ephyrdri.c: add more logging to ease debugging 159744 * hw/kdrive/ephyr/ephyrdriext.c: ditto. reformat. 159745 * hw/kdrive/ephyr/ephyrglxext.c,h: add this extension to 159746 proxy GLX requests to the host X. started to proxy those nedded to 159747 make glxinfo work with fglrx. Not yet finished. 159748 * hw/kdrive/ephyr/ephyrhostglx.c,h: put here the actual 159749 Xlib code used to hit the host X server because Xlib stuff cannot be 159750 mixed with xserver internal code, otherwise compilation erros due to 159751 type clashes happen. So no Xlib type should be exported by the 159752 entrypoints defined here. 159753 159754commit 0b8545144975bf7ed43c2564d01c697144eb1244 159755Author: Dodji Seketeli <dodji@seketeli.org> 159756Date: Thu Aug 16 17:30:13 2007 +0200 159757 159758 Implement ReputImage and StopVideo 159759 159760 * hw/kdrive/ephyr/ephyrhostvideo.c/h: 159761 (ephyrHostXVStopVideo): add this entry point. 159762 * hw/kdrive/ephyr/ephyrvideo.c: 159763 Basically add ReputImage and StopVideo implementations. 159764 Now, when other windows obscur the video window, the reclipping 159765 seems to be well handled using StopVideo and ReputImage. 159766 To do this, I was obliged to save the frame in PutImage, so 159767 that I could resend it un ReputImage. 159768 159769commit 810dc55866d1c2343512354646c7ab309ea1fad2 159770Author: Dodji Seketeli <dodji@seketeli.org> 159771Date: Thu Aug 16 17:11:22 2007 +0200 159772 159773 Ephyr-Xv: add a new line to a log 159774 159775 * hw/kdrive/ephyr/ephyrvideo.c: 159776 (ephyrQueryImageAttributes): add newline to log. 159777 159778commit 47e6dff89e48249828e828502e98951eee7f85bc 159779Author: Dodji Seketeli <dodji@seketeli.org> 159780Date: Thu Aug 16 17:07:36 2007 +0200 159781 159782 Xephyr-Xv: fix a crash when host X support several ports 159783 159784 * hw/kdrive/ephyr/ephyrvideo.c: 159785 (ephyrXVPrivQueryHostAdaptors): properly set 159786 port private luke. This fixes a crash when 159787 the host Xv supports multiple ports. 159788 Make sure number of ports cannot be zero. 159789 159790commit a38ad562a6f50e9d76a37917b936035215ea460e 159791Author: Dodji Seketeli <dodji@seketeli.org> 159792Date: Wed Aug 15 19:40:10 2007 +0200 159793 159794 make xephyr talk DRI protocol with hostX 159795 159796 * configure.ac,include/dix-config.h.in: define the XEPHYR_DRI macro. 159797 define it when --enable-xephyr and --enable-dri are both turned on. 159798 * hw/kdrive/ephyr/XF86dri.c: copy this from mesa source to enable 159799 Xephyr to talk DRI protocol the host X. In mesa, this is used by libGL.so to 159800 talk DRI protocol with the server. 159801 * hw/kdrive/ephyr/ephyr.c: finally initialise the DRI extension 159802 in the ephyrInitScreen() function. 159803 * hw/kdrive/ephyr/ephyrdri.c,ephyrdriext.c: safeguard the compilation 159804 using the XEPHYR_DRI macro. 159805 159806commit e4239a48075c77e6a8d2d5cb21f58dd67687482f 159807Author: Dodji Seketeli <dodji@openedhand.com> 159808Date: Tue Aug 14 22:48:41 2007 +0200 159809 159810 Initial dri forwarding big bricks. 159811 159812 * hw/kdrive/ephyr/ephyrdriext.c: added this to implement a DRI extension 159813 into Xephyr. Normally the DRI extension is only present in the 159814 xfree86 server, but I have ported it to Xephyr. The extension calls 159815 functions that declared/defined in ephyrdri.h ephyrdri.c that 159816 forwards the DRI calls to the host X. It does not work yet, as this 159817 entry is just to put the big bricks in place. 159818 * hw/kdrive/ephyr/ephyrdri.c,h: declaration & definition of the 159819 DRI client API that would hit the hostX server. 159820 * hw/kdrive/ephyr/GL/internal/dri_interface.h: added this, otherwise 159821 inclusion of /usr/include/X11/dri/xf86dri.h won't compile 159822 159823commit e01d3dd98d1b596e75d25f94dd89c7d41223011d 159824Author: Dodji Seketeli <dodji@openedhand.com> 159825Date: Thu Aug 9 10:55:10 2007 +0200 159826 159827 Support clipping region in PutImage. 159828 159829 * hw/kdrive/ephyr/ephyrhostvideo.c,h: 159830 (ephyrHostXVPutImage): make this support clipping region. 159831 The clipping region is propagated to host using XSetClipRectangles. 159832 This changes the API of ephyrHostXVPutImage. 159833 * hw/kdrive/ephyr/ephyrvideo.c: 159834 (ephyrPutImage): propagate the clipping region to the new 159835 ephyrHostXVPutImage() entry point. 159836 159837commit 39d3895469f07304d72800d8dcef6c7732f13d5f 159838Author: Dodji Seketeli <dodji@openedhand.com> 159839Date: Wed Aug 8 14:27:32 2007 +0200 159840 159841 make EphyrXVPriv be a singleton 159842 159843 * hw/kdrive/ephyr/ephyrvideo.c: 159844 (ephyrInitVideo) make the EphyrXVPriv object be a 159845 singleton instance, otherwise a new object is created at each 159846 generation. 159847 159848commit 4ed083095a13ec92bb4c8e705f26500f8312c138 159849Author: Dodji Seketeli <dodji@openedhand.com> 159850Date: Tue Aug 7 14:16:13 2007 +0200 159851 159852 add [Get/Put]Video and [Get/Put]Still support 159853 159854 * hw/kdrive/ephyr/ephyrhostvideo.c,h: 159855 (ephyrHostXVAdaptorHasXXX): fix these. 159856 (ephyrHostXVAdaptorHasGetVideo): added this 159857 (ephyrHostXVAdaptorHasGetStill): ditto 159858 (ephyrHostXVPutVideo): added this 159859 (ephyrHostXVGetVideo): ditto 159860 (ephyrHostXVPutStill): ditto 159861 (ephyrHostXVGetStill): ditto 159862 * hw/kdrive/ephyr/ephyrvideo.c: 159863 (ephyrPutVideo): implement this 159864 (ephyrGetVideo): ditto 159865 (ephyrPutStill): ditto 159866 (ephyrGetStill): ditto 159867 (ephyrXVPrivSetAdaptorsHooks): advertise GetVideo and GetStill 159868 when the host X supports it. 159869 159870commit 6f7961bc218169aaa0d10cfad0952adb7fac2f00 159871Author: Dodji Seketeli <dodji@openedhand.com> 159872Date: Mon Aug 6 18:38:54 2007 +0200 159873 159874 advertise only the hooks implemented by host X 159875 159876 * hw/kdrive/ephyr/ephyrhostvideo.c,h: 159877 (ephyrHostXVAdaptorHasPutVideo): detect if 159878 host X has the PutVideo call. 159879 (ephyrHostXVAdaptorHasPutStill): detect if 159880 host X has the PutStill call 159881 (ephyrHostXVAdaptorHasPutImage): detect if 159882 host X has the PutImage call 159883 * hw/kdrive/ephyr/ephyrvideo.c: 159884 (ephyrXVPrivQueryHostAdaptors): make sure to create 159885 atoms for attribute names otherwise subsequent 159886 calls to get/set attribute from clients won't work. 159887 (ephyrXVPrivSetAdaptorsHooks): don't hardwire advertising 159888 of the PutImage call. Instead, advertise the calls advertised 159889 by the host. 159890 159891commit 1de89239bdfa68ded2ef25fbb7d08369f527f7a9 159892Author: Dodji Seketeli <dodji@openedhand.com> 159893Date: Sat Aug 4 01:13:38 2007 +0200 159894 159895 don't crash when hostX reports an error. 159896 159897 * hw/kdrive/ephyr/ephyrhostvideo.c,h: 159898 (ephyrHostXVLogXErrorEvent): add this to 159899 log X error events. Heavily copied from libx11 159900 (ephyrHostXVErrorHandler): new x error handler that 159901 logs the error but does not exits. 159902 (ephyrHostXVInit): add this to be called at the beginning 159903 of xvideo lifetime. It sets an xerror handler that does not 159904 exit. 159905 * hw/kdrive/ephyr/ephyrvideo.c: 159906 (ephyrXVPrivIsAttrValueValid): this validates an attribute 159907 value. 159908 (ephyrSetPortAttribute): before setting an attribute, 159909 validate the new value so that we don't send a buggy 159910 request to host X. 159911 * hw/kdrive/ephyr/*.c: fix case in ephyrvideo code. 159912 * hw/kdrive/ephyr/ephyr.c: fix a typo 159913 159914commit 207714b60d18e94bbe2d755a431dbb29fe5677f7 159915Author: Dodji Seketeli <dodji@openedhand.com> 159916Date: Fri Aug 3 01:10:39 2007 +0200 159917 159918 first implementation of putimage 159919 159920 * hw/kdrive/ephyr/ephyrhostvideo.c,h: 159921 (EphyrHostXVPutImage): first implementation. does not 159922 support clipping regions yet. 159923 * hw/kdrive/ephyr/ephyrvideo.c: 159924 (DoSimpleClip): clip using a clipping box. Does not 159925 support regions yet. 159926 (EphyrPutImage): first implementation. 159927 Uses a simple clipping rectangle, no region yet. 159928 * hw/kdrive/ephyr/hostx.c: 159929 (hostx_get_window): added this to get the main 159930 window of the host x. 159931 159932commit aa478e09e25a1cb19b808e7e52b2ee155e34f03b 159933Author: Dodji Seketeli <dodji@openedhand.com> 159934Date: Thu Aug 2 12:49:36 2007 +0200 159935 159936 make properties setting/query code work and hopefully complete. 159937 159938 * hw/kdrive/ephyr/ephyrhostvideo.c,h: 159939 (EphyrHostXVQueryImageAttributes): add this call. It calls 159940 XvQueryBestSize xserver entry point. It uses the protocol 159941 level machinery because Xvlib does not expose that entry point 159942 as a C function. 159943 (EphyrHostXVQueryBestSize): added this wrapper around XvQueryBestSize(). 159944 (EphyrHostGetAtom, EphyrHostGetAtomName): added this to get 159945 an atom or atom name from the host server 159946 * hw/kdrive/ephyr/ephyrvideo.c: 159947 (EphyrSetPortAttribute): convert the atom into an host server 159948 server atom before attacking the host server with it, because in 159949 in its current form, the input atom is only valid in xephyr. 159950 This fix makes this call work. 159951 (EphyrGetPortAttribute): ditto. 159952 (EphyrQueryBestSize): implement this. 159953 (EphyrQueryImageAttributes): implement this. 159954 159955commit 24c837c0be6f85a917582156ab61d6c9c69301f9 159956Author: Dodji Seketeli <dodji@openedhand.com> 159957Date: Tue Jul 31 23:18:52 2007 +0200 159958 159959 implement setportattribute and getportattribute calls 159960 159961commit b8cd313be467fd14b8e5c9ba68e1a2029c1e888b 159962Author: Dodji Seketeli <dodji@seketeli.org> 159963Date: Mon Jul 23 12:43:43 2007 +0200 159964 159965 make xvinfo work 159966 159967 * hw/kdrive/ephyr/ephyrhostvideo.c: 159968 (EphyrHostXVAdaptorGetVideoFormats): properly get visual class instead of 159969 returning the visual id. 159970 (EphyrHostXVQueryEncodings): properly copy the fields because simple casting does 159971 truncate some fields. 159972 (EphyrHostAttributesDelete): XFree the whole array instead of trying to free invidial members. 159973 * hw/kdrive/ephyr/ephyrvideo.c: 159974 (ephyrInitVideo): fix a typo 159975 (EphyrXVPrivQueryHostAdaptors): set XvWindowMask mask to adaptors type. 159976 use host adaptor name. Don't forget to set nImages field. 159977 (EphyrXVPrivRegisterAdaptors): report an error when KdXVScreenInit() fails. 159978 159979commit aa2fae7b75f741e57bc4a9b754b5ea7518d78f47 159980Author: Dodji Seketeli <dodji@seketeli.org> 159981Date: Sat Jul 21 23:56:13 2007 +0200 159982 159983 misc fixes. 159984 159985 *(EphyrHostXVQueryAdaptors): return the queried adaptors list 159986 * (EphyrHostXVQueryPortAttributes): return port attributes number 159987 159988commit 2bf7f3c223fd32cd4922f98eadda2e3f06d9e119 159989Author: Dodji Seketeli <dodji@seketeli.org> 159990Date: Sat Jul 21 17:55:12 2007 +0200 159991 159992 link and init xv 159993 159994commit 66b28532e533ff5688705c5286425e68d6403cee 159995Author: Dodji Seketeli <dodji@seketeli.org> 159996Date: Sat Jul 21 12:08:39 2007 +0200 159997 159998 initial commit of xv support work 159999 160000commit 1e8f8b36a56a884092ee01a0bb3f40436e30fa42 160001Author: Dodji Seketeli <dodji@seketeli.org> 160002Date: Sat Jul 21 18:22:20 2007 +0200 160003 160004 works with no adaptors 160005 160006 Debug the early code to make work when the host has no video adaptors 160007 160008commit 50ff2377cac50ae25218be5b30612fb8bf6046a6 160009Author: Dodji Seketeli <dodji@seketeli.org> 160010Date: Sat Jul 21 18:20:59 2007 +0200 160011 160012 add -verbosity <level> option to command line 160013 160014commit 8426eb2433f964ea19848f24dc99eb4588d10d1a 160015Author: Dodji Seketeli <dodji@seketeli.org> 160016Date: Sat Jul 21 17:55:12 2007 +0200 160017 160018 link and init xv 160019 160020commit 50a64c84e1556ef7f8efcc7f8655e0c5e023ef4d 160021Author: Dodji Seketeli <dodji@seketeli.org> 160022Date: Sat Jul 21 12:08:39 2007 +0200 160023 160024 initial commit of xv support work 160025 160026commit 95fadbd4022ec99f42ba78ec8a18a064903e8a7f 160027Author: Dodji Seketeli <dodji@seketeli.org> 160028Date: Mon Jul 16 17:37:21 2007 +0200 160029 160030 Add basic Exa driver entry point validation. 160031 160032 In exaDriverInit(), quickly check if the programmer forgot to set some 160033 mandadory driver hooks, in that case display a meaningful error message. 160034 160035commit e5e6514ffa0fd132e0cc1b15b94119e6e8755f43 160036Author: Andrew Christan <andrew.christian@nokia.com> 160037Date: Tue Oct 2 13:25:51 2007 +0200 160038 160039 Xephyr: add "multiscreen" suport 160040 160041 * This patch adds multiscreen support to Xephyr. For instance, 160042 the command line : "Xephyr :4 -ac -screen 320x240 -screen 640x480" 160043 will launch with two "screens" - namely two main windows. 160044 The first main window represents a screen that has the number :4.0, with 160045 a geometry of 320x240 pixels, and the second one represents a screen 160046 that has the number :4.1 with a geometry of 640x480. 160047 The command line: "DISPLAY=:4.1 xclock" will launch the xclock program 160048 on the second screen, for intance. 160049 160050 * this patch was edited by Dodji Seketeli <dodji@openedhand.com> for: 160051 - better style compliance with the rest of the Xephyr code 160052 - make sure Xephyr could be launched with no -screen option. By 160053 default that creates a default screen of 640x480 pixel like before 160054 - display full titles on the windows - with insctructions to grab 160055 keyboard and mouse - like before. 160056 160057commit 81692b628f41cb64329f3cccc0503fb216a2b8e3 160058Author: Dodji Seketeli <dodji@openedhand.com> 160059Date: Tue Oct 2 13:11:28 2007 +0200 160060 160061 Fix #12650: "windows not receiving mouse events in multiple screens context" 160062 160063 * dix/events.c, include/dix.h: 160064 (UpdateSpriteForScreen): added this to update the mouse sprite context 160065 when we switch from a pScreen to another. 160066 * mi/mipointer.c: 160067 (miPointerWarpCursor): as we are switching to a new pScreen, 160068 don't forget to update the mouse sprite context. 160069 160070commit f965a5f3454a95ddffb2faf9b291feff46305adf 160071Author: Peter Hutterer <peter@cs.unisa.edu.au> 160072Date: Tue Oct 2 16:20:27 2007 +0930 160073 160074 dix: Ignore devices when deleting passive core grabs. 160075 160076commit dc90d500085dd74e90073be008fc514bd97b160a 160077Author: Dave Airlie <airlied@redhat.com> 160078Date: Tue Oct 2 14:05:07 2007 +1000 160079 160080 exa: increase minor version number for pixmap allocation hooks 160081 160082commit f15af2ae60bb8503c336bc3cba0560bd314a34bc 160083Author: Kristian Høgsberg <krh@redhat.com> 160084Date: Fri Aug 3 16:33:33 2007 +1000 160085 160086 exa: only setup offscreen allocator if driver doesn't provide CreatePixmap 160087 160088commit ffb58f4fa8d86e87f831430b8627f27d85f971a9 160089Author: Dave Airlie <airlied@redhat.com> 160090Date: Tue Oct 2 14:03:39 2007 +1000 160091 160092 exa: add hooks for drivers to take over pixmap allocation 160093 160094 This adds hooks for the driver to access Create/DestroyPixmap and ModifyPixmapHe 160095 ader. 160096 160097 It allocates a 0 sized pixmap using fb and calls the driver routine to do 160098 work of allocating the actual memory. 160099 160100 ModifyPixmapHeader is mainly required for hooking the screen pixmap which 160101 isn't create by normal methods 160102 160103commit 2b075e97ec913ebe19290181e36bb169410b820b 160104Author: Peter Hutterer <peter@cs.unisa.edu.au> 160105Date: Tue Oct 2 13:28:40 2007 +0930 160106 160107 mi: switch core keyboard on XI events, not only on core events. 160108 160109 We only get core events through the EQ in exceptional cases, so make sure we 160110 actually swap the core keymap for XI events as well. Gives us back the ability 160111 to have multiple keyboard layouts simultaneously. 160112 160113commit 31555af0005a0bc4d7ef785214696ac84681e29f 160114Author: Dave Airlie <airlied@redhat.com> 160115Date: Tue Oct 2 13:32:57 2007 +1000 160116 160117 exa: add a pixmap private pointer for drivers to retrieve. 160118 160119commit 45ec6cd1fb242363c91ad8af1fd4a27a7f02621a 160120Author: Peter Hutterer <peter@cs.unisa.edu.au> 160121Date: Tue Oct 2 13:19:07 2007 +0930 160122 160123 xfree86: fix DGA to use the correct devices. 160124 160125 DGAStealXXXEvent modified to take in device argument. 160126 160127 The evdev driver only sends one valuator when only one axis changed. We need 160128 to check for DGA either way (xf86PostMotionEventP), otherwise we lose purely 160129 horizontal/vertical movements. 160130 160131 Note that DGA does not do XI events. 160132 160133commit 122ae65ed90195c584a770027b4d14ef65f72492 160134Author: Peter Hutterer <peter@cs.unisa.edu.au> 160135Date: Tue Oct 2 10:50:59 2007 +0930 160136 160137 xfree86: xf86DGA: some cleanup to use the correct devices instead of VCP/VCK. 160138 160139commit 54ce7436b2579bb20728eab7c8d460b531c378af 160140Author: Peter Hutterer <peter@cs.unisa.edu.au> 160141Date: Tue Oct 2 10:49:46 2007 +0930 160142 160143 xfree86: update all pointers when calling xf86SwitchMode 160144 160145 Center the frame around the first pointer found and then update all pointers 160146 on the same screen to move to the edges (if necessary). 160147 160148 Note: xf86WarpCursor needs to be modified, is using deprecated 160149 miPointerWarpCursor and will kill the server when called with 160150 inputInfo.pointer. 160151 160152commit 113011221a4fcbd70bc592930b35f20d55809f1e 160153Author: Peter Hutterer <peter@cs.unisa.edu.au> 160154Date: Mon Oct 1 21:42:01 2007 +0930 160155 160156 Xi, xfree86: Remove leftovers of the XI wrapper code. 160157 160158commit 50551ec693f40b91652fe4814e9fe2e1f9ab6517 160159Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160160Date: Fri Sep 28 15:04:33 2007 -0400 160161 160162 xace: remove obsoleted DRAWABLE_ACCESS hook. 160163 160164commit 8b548657204000e18c7a38706a0071ae2f93159f 160165Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160166Date: Fri Sep 28 13:34:18 2007 -0400 160167 160168 xace: add hooks + new access codes: XKB extension. 160169 160170 Removes "LookupKeyboardDevice" and "LookupPointerDevice" in favor of 160171 inputInfo.keyboard and inputInfo.pointer, respectively; all use cases 160172 are non-XI compliant anyway. 160173 160174commit 5c03d131815cfe2f78792277ab8352e69e830196 160175Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160176Date: Fri Sep 28 08:02:00 2007 -0400 160177 160178 xace: add new hooks + access controls: XInput extension. 160179 160180 Introduces new dix API to lookup a device, dixLookupDevice(), which 160181 replaces LookupDeviceIntRec and LookupDevice. 160182 160183commit a511c445debbd13e8c48146ecd2d7c97e793f788 160184Author: Peter Hutterer <peter@cs.unisa.edu.au> 160185Date: Fri Sep 28 18:46:41 2007 +0930 160186 160187 Remove generation of core events, long live XI! 160188 160189 Let the drivers only generate XI events and put those into the event queue. 160190 When processing events, generate core events as needed. This fixes a number of 160191 problems with XKB and the DIX in general. 160192 160193 The previous approach was to put core events and XI events as separate events 160194 into the event queue. When being processed, the server had no knowledge of 160195 them coming from the same device state change. Anything that would then change 160196 the state of the device accordingly was in danger of changing it twice, 160197 leading to some funny (i.e. not funny at all) results. 160198 160199 Emulating core events while processing XI events fixes this, there is only one 160200 path that actually changes the device state now. Although we have to be 160201 careful when replaying events from synced devices, otherwise we may lose 160202 events. 160203 160204 Note: XI has precedence over core for passive grabs, but core events are 160205 delivered to the client first. 160206 160207 This removes the wrapping added in 340911d7243a7f1095d79b5b2dcfa81b145c2474 160208 160209commit be466d8df808c4e4067a7963617bc3f506768f2d 160210Author: Peter Hutterer <peter@cs.unisa.edu.au> 160211Date: Thu Sep 27 22:23:05 2007 +0930 160212 160213 dix: GetPairedKeyboard() always returns a valid keyboard (VCK if necessary). 160214 160215 We need it unconditionally in a few places, and the rest checked for NULL and 160216 then set it to VCK anyway. So, fixing up all callers to appreciate the defined 160217 return value. 160218 160219commit 53434edc3d306137d019d95189ecdf0dbff75205 160220Author: Peter Hutterer <peter@cs.unisa.edu.au> 160221Date: Thu Sep 27 22:14:09 2007 +0930 160222 160223 dix: take paired keyboard instead of VCK (CoreProcessPointerEvents) 160224 160225commit 2d93e69690d2c5d4a89a795ede6423796528e5df 160226Author: Alan Coopersmith <alan.coopersmith@sun.com> 160227Date: Thu Sep 27 16:47:06 2007 -0700 160228 160229 Rework local client id finding code to be more uniform 160230 160231commit c7ead3a68e5839cb92129e35b21f55007fba8445 160232Author: Alan Coopersmith <alan.coopersmith@sun.com> 160233Date: Thu Sep 27 16:13:28 2007 -0700 160234 160235 xorgcfg needs $(DIX_CFLAGS) for pixman-1 include path 160236 160237commit 62a9fb4cda988a896909a5620a68c51e46d0e403 160238Author: Alan Coopersmith <alan.coopersmith@sun.com> 160239Date: Thu Sep 27 16:12:29 2007 -0700 160240 160241 Fix PCI rework build on Solaris (copy what BSD does) 160242 160243commit cfe549d1ba28396c44af94133d8c1d323e3c7086 160244Author: Tilman Sauerbeck <tilman@code-monkey.de> 160245Date: Thu Sep 27 22:36:52 2007 +0200 160246 160247 EXA: Added back the maxPitchPixels initialization code. 160248 160249 This doesn't add real value yet, but it will be useful once I add code 160250 that splits large render operations into smaller parts if necessary. 160251 160252commit c11a27ef85674f1e77ffa7f083646b848e1dc0b8 160253Author: Alan Hourihane <alanh@tungstengraphics.com> 160254Date: Thu Sep 27 14:54:40 2007 +0100 160255 160256 Add the FB_ACCESS_WRAPPER checks 160257 160258commit e0bb33b3d2e4f54cf20853cde9a1664a7dbd0dc7 160259Author: Alan Hourihane <alanh@tungstengraphics.com> 160260Date: Thu Sep 27 14:50:28 2007 +0100 160261 160262 Fix bug #12286 160263 160264commit 6d5c1e0d896666bcb2b3c1de7bfa424f140be364 160265Author: Michel Dänzer <michel@tungstengraphics.com> 160266Date: Thu Sep 27 13:04:41 2007 +0200 160267 160268 EXA: Remove bogus pitch checks. 160269 160270 exaCreatePixmap should handle all cases correctly. 160271 160272commit 598698678b07cb3a9406a9ee98bd3186366949e7 160273Author: Michel Dänzer <michel@tungstengraphics.com> 160274Date: Thu Sep 27 13:07:09 2007 +0200 160275 160276 EXA: Punt for all fallbacks in exaFillRegion*. 160277 160278 Now that PaintWindow is gone, all callers already handle fallbacks. 160279 160280commit 006f6525057970a74382132237b2131286ad147c 160281Author: Michel Dänzer <michel@tungstengraphics.com> 160282Date: Thu Sep 27 13:08:41 2007 +0200 160283 160284 EXA: Make sure tile offsets passed to drivers are never negative. 160285 160286 Thanks to Björn Steinbrink for pointing out the problem on IRC. 160287 160288commit da7d9aa1fb60e13a59c9f842fed7aefc5b97c195 160289Author: Michel Dänzer <michel@tungstengraphics.com> 160290Date: Thu Sep 27 13:08:40 2007 +0200 160291 160292 EXA: Tile offscreen pixmap coordinate offsets are always 0. 160293 160294commit d6f4764bf5f3a601a0034ded039857e8ea5563b2 160295Author: Michel Dänzer <michel@tungstengraphics.com> 160296Date: Thu Sep 27 13:08:40 2007 +0200 160297 160298 EXA: Remove some clearly bogus code from exaCopyNtoN. 160299 160300 Not sure what I was thinking when I wrote this... it would cause the box 160301 coordinates to be off for exaCopyNtoNTwoDir or fallbacks. 160302 160303 Thanks to Tilman Sauerbeck for pointing out the problem on IRC and testing the 160304 fix. 160305 160306commit c7d6d1f589d729fa689d22d82fe30afbc6e1cacb 160307Author: Michel Dänzer <michel@tungstengraphics.com> 160308Date: Thu Sep 27 13:08:40 2007 +0200 160309 160310 EXA: Punt on fallback case not handled correctly in exaFillRegionTiled. 160311 160312 Fixes http://bugs.freedesktop.org/show_bug.cgi?id=12520 . 160313 160314commit 1d938a80fd4fa58d1791c146b6b5c2dfe148dce7 160315Author: Alan Coopersmith <alan.coopersmith@sun.com> 160316Date: Wed Sep 26 19:42:46 2007 -0700 160317 160318 Remove unused pciAddrHostToBus functions from ix86Pci.c 160319 160320 Matches linuxPci.c changes made in 8279444a54f38c5e2e5d4c2d936a10d74990e0be 160321 Fixes compiler errors: 160322 "ix86Pci.c", line 194: too many struct/union initializers 160323 "ix86Pci.c", line 204: too many struct/union initializers 160324 "ix86Pci.c", line 214: too many struct/union initializers 160325 160326commit de06a47d4ad9c845b098438d9492a5f42483a2ad 160327Author: Peter Hutterer <peter@cs.unisa.edu.au> 160328Date: Thu Sep 27 11:45:03 2007 +0930 160329 160330 dix: GetMaximumEventsNum() needs to return 3 + MAX_VALUATOR_EVENTS. 160331 160332 Raw events can be generated for pointer events, so 2 + MVE isn't enough. 160333 160334commit 32d0440c7f6e604807cb14dd32349df6f22c903b 160335Author: Peter Hutterer <peter@cs.unisa.edu.au> 160336Date: Thu Sep 27 11:44:03 2007 +0930 160337 160338 xkb: xkbHandleActions: let wrapping take care of event delivery. 160339 160340 This is hopefully better than hardcodey calling CoreProcessPointerEvent. 160341 160342commit 27612748e0ec20f3a23839f0a12e39f598dd722c 160343Merge: b61461425 3b114f260 160344Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160345Date: Wed Sep 26 07:47:29 2007 -0400 160346 160347 Merge branch 'master' into XACE-SELINUX 160348 160349 Conflicts: 160350 160351 dix/devices.c 160352 160353commit 3b114f2603fc2adeec7f5f8f20fe4870afb8dff1 160354Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160355Date: Tue Sep 25 07:55:17 2007 -0400 160356 160357 Input: Properly swap DevicePresenceNotify events. 160358 160359commit 31a7994ac7365562ef1d00e0a7b25f967a961a4e 160360Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160361Date: Mon Sep 24 13:33:01 2007 -0400 160362 160363 Input: Return errors to the dispatcher instead of sending them ourself. 160364 160365 Also fixed two "unused variable: stuff" warnings. 160366 160367commit 56ffc381d32687242dd094395fcf2216339bab2a 160368Author: Tilman Sauerbeck <tilman@code-monkey.de> 160369Date: Wed Sep 26 16:47:54 2007 +0200 160370 160371 EXA: Removed duplicated exaGetDrawablePixmap() calls. 160372 160373commit aa0325db7e8ab11d9418cae14f11c488f443ccaa 160374Author: Dodji Seketeli <dodji@openedhand.com> 160375Date: Wed Sep 26 15:26:42 2007 +0200 160376 160377 Xephyr: fix compiler warnings 160378 160379 * hw/kdrive/ephyr/ephyr_draw.c: 160380 (ephyrDownloadFromScreen), 160381 (ephyrUploadToScreen): fix compiler warnings 160382 160383commit 8f9bf927e1beecf9b9ec8877131ec12c765e4d84 160384Author: Peter Hutterer <peter@cs.unisa.edu.au> 160385Date: Wed Sep 26 18:04:59 2007 +0930 160386 160387 xkb: Unwrap properly in ProcessPointerEvent. 160388 160389 Instead of hardcoding CoreProcessPointerEvent, actually try to unwrap properly 160390 and then call the unwrapped processInputProc. Seems to be a better idea, 160391 especially since it makes stuff actually work... 160392 160393commit 3342b5ad47be25c6838321c0aafc28c329c308b5 160394Author: Peter Hutterer <peter@cs.unisa.edu.au> 160395Date: Wed Sep 26 18:03:21 2007 +0930 160396 160397 Xi: fix the wrapper code for processInputProc wrapping. 160398 160399 Followup to [1]. 160400 If a core grab causes the device to freeze, it overwrites the processInputProc 160401 of the device. [1] would then overwrite this while unwrapping, the device 160402 does not thaw anymore. 160403 Changing this to only re-wrap if the processInputProc hasn't been changed 160404 during the event handling. 160405 160406 [1] 340911d7243a7f1095d79b5b2dcfa81b145c2474 160407 160408commit e2cb8515661b1f7826981931d82dee6e05529f04 160409Author: Peter Hutterer <peter@cs.unisa.edu.au> 160410Date: Wed Sep 26 18:01:17 2007 +0930 160411 160412 xfree86: pass in the XKB event processing proc instead of the XI one. 160413 160414 We need to pass ProcessPointerEvent or ProcessKeyboardEvent to 160415 XkbSetExtension, otherwise we lose the xkb layer. 160416 160417commit 27bc1a8fef2bfd3d62fb44f7c7eb0d463ed08632 160418Author: Peter Hutterer <peter@cs.unisa.edu.au> 160419Date: Wed Sep 26 15:24:41 2007 +0930 160420 160421 xkb: XkbFilterEvents: Remove unused variable compiler warning. 160422 160423commit bfc89c035542a10594f5f0cbde1c7e28b7d024a7 160424Author: Peter Hutterer <peter@cs.unisa.edu.au> 160425Date: Wed Sep 26 15:23:37 2007 +0930 160426 160427 xkb: unify ErrorFs. Prefix all with [xkb]. 160428 160429 Output for XkbUseMsg intentionally skipped. 160430 160431commit 8b508f5d6bd0d3995294d5ff300a856754442999 160432Author: Peter Hutterer <peter@cs.unisa.edu.au> 160433Date: Wed Sep 26 15:19:53 2007 +0930 160434 160435 dix: unify ErrorFs. prepend all with [dix]. 160436 160437commit 394f3c1dbee7270a1d930846d49278424c3072d4 160438Author: Peter Hutterer <peter@cs.unisa.edu.au> 160439Date: Wed Sep 26 15:19:37 2007 +0930 160440 160441 dix: GetPointerEvents: remove unused variable compiler warning. 160442 160443commit 55fd50273ea67eb99a0c8a830349851931298387 160444Author: Peter Hutterer <peter@cs.unisa.edu.au> 160445Date: Wed Sep 26 15:15:34 2007 +0930 160446 160447 Xi: unify ErrorFs. Prefix all with [Xi]. 160448 160449commit 340911d7243a7f1095d79b5b2dcfa81b145c2474 160450Author: Peter Hutterer <peter@cs.unisa.edu.au> 160451Date: Wed Sep 26 15:12:49 2007 +0930 160452 160453 Wrap core event handling through ProcessOtherEvents. 160454 160455 When processing events from the EQ, _always_ call the processInputProc of the 160456 matching device. For XI devices, this proc is wrapped in three layers. 160457 Core event handling is wrapped by XI event handling, which is wrapped by XKB. 160458 A core event now passes through XKB -> XI -> DIX. 160459 160460 This gets rid of a sync'd grab problem: with the previous code, core events 160461 did disappear during a sync'd device grab on account of mieqProcessInputEvents 160462 calling the processInputProc of the VCP/VCK instead of the actual device. This 160463 lead to the event being processed as normal instead of being enqueued for 160464 later replaying. 160465 160466commit b2a4883bd89d406713d4f808e72721ecc1456d67 160467Author: Peter Hutterer <peter@cs.unisa.edu.au> 160468Date: Wed Sep 26 09:20:11 2007 +0930 160469 160470 dix: add comments to PlayReleasedEvents() 160471 160472 Also stick another warning in, more changes needed for XGE if panoramix is 160473 active. 160474 160475commit 275dc77c2972147407584323b866b8acc1654ea0 160476Author: Peter Hutterer <peter@cs.unisa.edu.au> 160477Date: Tue Sep 25 21:40:08 2007 +0930 160478 160479 dix: add comments to ComputeFreezes() and syncEvents struct. 160480 160481commit ae8b4f7dcf1291a2f5a0d0159f3e6089eea0578a 160482Author: Tilman Sauerbeck <tilman@code-monkey.de> 160483Date: Mon Sep 24 22:06:52 2007 +0200 160484 160485 EXA: Added pitch limitations. 160486 160487 Drivers can now specify the max pitches that the accelerator supports. 160488 160489commit b61461425eb15fcff2a58330d74fe5a5a1f226fc 160490Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160491Date: Tue Sep 25 09:56:00 2007 -0400 160492 160493 xace: add hooks + new access codes: XV extension. 160494 160495 May need to revisit this extension in the future, depending on observed use. 160496 160497commit 9bd04055a2175ec16756d3bf73ae03b5e163a28a 160498Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160499Date: Tue Sep 25 09:33:51 2007 -0400 160500 160501 xace: change prototype of VALIDATE_DRAWABLE_AND_GC macro to allow access 160502 mode to be passed to dixLookupDrawable. 160503 160504commit 988f446fe00ac2b92c484a4440613dd82191eadd 160505Author: Eric Anholt <eric@anholt.net> 160506Date: Mon Sep 24 20:23:35 2007 -0700 160507 160508 [EDID] Ignore reserved bits in deciding monitor vs detailed timing descriptor. 160509 160510 Even though they're defined to zero by the spec, we've seen an EDID block 160511 where the (empty) ASCII strings were stuffed in a byte early, leading to the 160512 descriptor being considered a detailed timing instead. 160513 160514commit 5d9f6cf51becdeb1d2be9bdeec4bb5d0c4dd0e8b 160515Author: Peter Hutterer <peter@cs.unisa.edu.au> 160516Date: Thu Sep 13 23:33:07 2007 +0930 160517 160518 dix: cosmetic change. Remove "pointer" variable, might as well use pDev. 160519 160520commit c4db4d1fcf7fa2ba802a208beee6b3adc83921f5 160521Author: Peter Hutterer <peter@cs.unisa.edu.au> 160522Date: Thu Sep 13 23:27:07 2007 +0930 160523 160524 include: remove register keywords. 160525 160526commit 1c25c46f3d5f98be8dbf36684aad4302b220e163 160527Author: Peter Hutterer <peter@cs.unisa.edu.au> 160528Date: Thu Sep 13 23:22:27 2007 +0930 160529 160530 dix: fix compiler warnings about unused variables. 160531 160532commit 68e0c4988e359e3c9da933946bc703cf8530bdbc 160533Author: Peter Hutterer <peter@cs.unisa.edu.au> 160534Date: Wed Sep 12 20:48:48 2007 +0930 160535 160536 dix: fix "possible use of uninitialized variable" warning. 160537 160538 The device passed through to UnrealizeCursor isn't used anyway, so setting it 160539 to NULL is enough. 160540 160541commit b484451eab261a919fe94c0c0f56877f7571fc9e 160542Author: Peter Hutterer <peter@cs.unisa.edu.au> 160543Date: Wed Sep 12 18:34:51 2007 +0930 160544 160545 dix: don't try to access "time" field for GenericEvents in PlayReleasedEvent. 160546 160547 GenericEvents can't be parsed to keyButtonPointer, and there's no guarantee 160548 that it has a time field anyway. PlayReleasedEvent needs to store the millis 160549 when we know it (core events, XI event) and just re-use them for GenericEvents. 160550 160551 Yes, this is a hack. But it looks like the time has zero significance anyway. 160552 160553commit 1c38eac442a3ab9e8469e944f382251957996c5a 160554Author: Peter Hutterer <peter@cs.unisa.edu.au> 160555Date: Wed Sep 12 18:09:57 2007 +0930 160556 160557 dix: fix comment for DeliverDeviceEvents. 160558 160559 DDE is not always called, grabbed or focused devices go through 160560 Deliver{Grabbed|Focus}Event first. 160561 160562 Thanks to Eamon Walsh for spotting it. 160563 160564commit 06188ce90d2a8146140773f9367f23404b464282 160565Author: Peter Hutterer <peter@cs.unisa.edu.au> 160566Date: Wed Sep 12 18:00:03 2007 +0930 160567 160568 dix: Make EnqueueEvent aware of GenericEvents. 160569 160570 GenericEvents can be > sizeof(xEvents), so we need to make sure we're 160571 allocating and copying enough memory for the event. 160572 160573commit 8840829ab93c4eb62eb58753c015da5307133fe5 160574Author: Peter Hutterer <peter@cs.unisa.edu.au> 160575Date: Wed Sep 12 17:40:11 2007 +0930 160576 160577 dix: don't compress motion events from different devices (EventEnqueue) 160578 160579commit 27ad5d74c20f01516a1bff73be283f8982fcf0fe 160580Author: Daniel Stone <daniel@fooishbar.org> 160581Date: Sun Sep 23 17:17:03 2007 +0300 160582 160583 Input: Generate XKB mapping changes for all core-sending devices (bug #12523) 160584 160585 When we change the mapping on a core device, make sure we propagate this 160586 through to XKB for all extended devices as well. 160587 160588commit 0e800ca4651a947ccef239e6fe7bf64aab92257c 160589Author: Daniel Stone <daniel@fooishbar.org> 160590Date: Sun Sep 23 12:43:31 2007 +0300 160591 160592 GetKeyboardEvents: Reject out-of-range keycodes (bug #12528) 160593 160594 We can only deal with keycodes between 8 and 255, so make sure that we never 160595 accept anything out of this range. 160596 160597commit e1860f241b7322c54c109cf5727df45b54b73916 160598Author: Alex Deucher <alex@botch2.(none)> 160599Date: Sat Sep 22 17:54:51 2007 -0400 160600 160601 Document xf86_crtc_clip_video_helper better. 160602 160603commit 19d30c1c771c0f5c9dfc0f4cfb084dd2e528d992 160604Author: Alex Deucher <alex@botch2.(none)> 160605Date: Sat Sep 22 17:51:27 2007 -0400 160606 160607 Revert "Fix possible crash if Xv window is outside of either crtc" 160608 160609 This reverts commit 92355314103d2193ca20786915eaf608922a36dc. 160610 160611 This changes ABI. False means there was a memory failure of some kind, 160612 not that the video shouldn't be displayed 160613 160614commit 92355314103d2193ca20786915eaf608922a36dc 160615Author: Michel Dänzer <daenzer@debian.org> 160616Date: Sat Sep 22 14:25:02 2007 -0400 160617 160618 Fix possible crash if Xv window is outside of either crtc 160619 160620commit edd88170836ec985ab1a59179de69d50644b8dff 160621Author: Tilman Sauerbeck <tilman@code-monkey.de> 160622Date: Sat Sep 22 13:02:08 2007 +0200 160623 160624 EXA: Check ABI compatibility earlier. 160625 160626 If the driver isn't compatible to the server, all bets are off anyway wrt 160627 the contents of the fields that we're validating, which can lead to bogus 160628 error messages. 160629 160630commit abe0a51f3f790f8c055289465e130177c4b647cc 160631Author: Ben Byer <bbyer@bbyer.apple.com> 160632Date: Fri Sep 21 17:07:36 2007 -0700 160633 160634 So, like, checking return codes of system calls (signal, etc) is good. 160635 Also, only restore an old signal handler if one was actually set 160636 (prevents the server from dying on OS X). 160637 160638commit eb82b19aa71333b46e927516cc228f25d3e05e4d 160639Author: Ben Byer <bbyer@bbyer.apple.com> 160640Date: Thu Sep 20 19:55:48 2007 -0700 160641 160642 XDarwin: build fix for X11.app 160643 160644commit 5b08932bfbb7e4612733fffd2acc9682c6856d90 160645Author: Ben Byer <bbyer@bbyer.apple.com> 160646Date: Thu Sep 20 19:27:31 2007 -0700 160647 160648 XDarwin: pulling in Indirect GLX fixes from downstream 160649 160650commit 742d5d6adc4d4a1b6fceeb3443d7931e107462f7 160651Author: Ben Byer <bbyer@bbyer.apple.com> 160652Date: Thu Sep 20 18:03:51 2007 -0700 160653 160654 XDarwin: Fixes to quartzKeyboard.c to prevent a crash using certain non-US keyboards 160655 160656commit 374bd88d1025835fe36ca0ff6620eb1d3b0a06b0 160657Author: Ben Byer <bbyer@bbyer.apple.com> 160658Date: Thu Sep 20 16:50:49 2007 -0700 160659 160660 XDarwin: Merging misc changes from downstream. 160661 160662 The only functional changes in this patch are a removal of use of 160663 Xtrans internals -- replaced by xcb, which doesn't seem to be used 160664 elsewhere in the server? Pity. 160665 Also, a fix to make all X11 windows pop to the front of the display 160666 when the X11.app icon is clicked -- currently takes two clicks, 160667 not sure why. 160668 160669commit 3a965fdadccea7beff09a28c9c0ef4b4975eae38 160670Author: Aaron Plattner <aplattner@nvidia.com> 160671Date: Thu Sep 20 16:22:24 2007 -0700 160672 160673 Don't segfault on shutdown if we never managed to connect to dbus. 160674 160675commit 50fa8722d35c12e5f0322cebe25faf99c39d4f50 160676Author: Aaron Plattner <aplattner@nvidia.com> 160677Date: Thu Sep 20 14:00:33 2007 -0700 160678 160679 Set noCompositeExtension to TRUE when failing to initialize the extension (e.g. when Xinerama is enabled). 160680 160681commit a247886b082cea93fa8f8980616a9c388ba70111 160682Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160683Date: Thu Sep 20 13:06:38 2007 -0400 160684 160685 xace: add hooks + new access codes: XF86-Bigfont extension 160686 160687commit f6532a81eec5f096e27285687964b77c17987f72 160688Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160689Date: Thu Sep 20 12:17:17 2007 -0400 160690 160691 xace: add hooks + new access codes: APPGROUP extension 160692 160693commit 82f7195a628cc7ec94abc0cfe5bae2be8af443bc 160694Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160695Date: Thu Sep 20 09:17:09 2007 -0400 160696 160697 xace: modifications to ChangeWindowAttributes special case: separate 160698 Receive and SetAttr. Refer to 963e69b8efc39369915e7f0c6f370ac0d5d2b60f 160699 160700commit 661b1328cf992d8855552677a94d60de1d8ce942 160701Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160702Date: Thu Sep 20 08:41:26 2007 -0400 160703 160704 xace: add hooks + new access codes: SYNC extension 160705 160706 May need to revisit this extension in the future, depending on observed use. 160707 160708commit 90bacdef723e1e49c72775144916750758d3568c 160709Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160710Date: Thu Sep 20 06:53:51 2007 -0400 160711 160712 xace: add hooks + new access codes: MIT-SHM extension 160713 160714commit 12a18cc8903fac53c3c77b23dd8093187594a4f3 160715Author: David Nolden <david.nolden.git@art-master.de> 160716Date: Fri Sep 14 22:42:19 2007 +0200 160717 160718 compilation-fix in debug mode 160719 160720commit e93cff52fed9074aa007c2e6ec6b578f69aef3cb 160721Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160722Date: Wed Sep 19 14:48:20 2007 -0400 160723 160724 xace: add hooks + new access codes: DOUBLE-BUFFER extension 160725 160726commit 082c0f7fb34458ebb303cf875d1d75686eca25e6 160727Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160728Date: Wed Sep 19 13:59:35 2007 -0400 160729 160730 devPrivates rework: move devPrivates field in drawable structure types 160731 to just below the DrawableRec. Wish there were a better way to do this 160732 but it has to be in the same place for all drawable types. 160733 160734commit 5b36b64192517e2470766ce7ff1d4dc04c936fad 160735Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160736Date: Wed Sep 19 11:11:54 2007 -0400 160737 160738 xace: add missing argument to hook call. 160739 160740commit 963e69b8efc39369915e7f0c6f370ac0d5d2b60f 160741Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160742Date: Wed Sep 19 11:11:41 2007 -0400 160743 160744 xace: add special-case for just setting the event mask on a window, 160745 this should only check "receive" permission, not "setattr" permission. 160746 160747commit c140d20f4ade973496dcbbd06453402ef0c0e85c 160748Author: Ben Byer <bbyer@bbyer.local> 160749Date: Wed Sep 19 05:58:22 2007 -0700 160750 160751 XDarwin: mass change from using xorg-config.h to dix-config.h cuz we're not using the X.org ddx anymore 160752 160753commit bcb5e3e6276ce9ae7a4dae7434cf2247764da078 160754Author: Ben Byer <bbyer@bbyer.local> 160755Date: Wed Sep 19 04:46:10 2007 -0700 160756 160757 XDARWIN: More fixes to Makefile.am 160758 160759 At least on my system (10.5 with the latest and greatest modules), 160760 Xquartz now builds out of the box. It doesn't quite work yet, but 160761 hey -- you have to start somewhere. ;) 160762 160763commit 97c150b61bbe436453b05d3c07cd2173870aac40 160764Merge: 57907e094 547ad2125 160765Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 160766Date: Wed Sep 19 07:25:55 2007 -0400 160767 160768 Merge branch 'master' into XACE-SELINUX 160769 160770 Conflicts: 160771 160772 afb/afbpntwin.c 160773 afb/afbscrinit.c 160774 afb/afbwindow.c 160775 cfb/cfb.h 160776 cfb/cfballpriv.c 160777 cfb/cfbscrinit.c 160778 cfb/cfbwindow.c 160779 configure.ac 160780 fb/wfbrename.h 160781 hw/xfree86/xf4bpp/ppcIO.c 160782 hw/xfree86/xf4bpp/ppcPntWin.c 160783 hw/xfree86/xf4bpp/ppcWindow.c 160784 hw/xfree86/xf8_32bpp/cfbscrinit.c 160785 mfb/mfb.h 160786 mfb/mfbpntwin.c 160787 mfb/mfbscrinit.c 160788 mfb/mfbwindow.c 160789 mi/miexpose.c 160790 160791 Note: conflicts caused by devPrivates rework vs. paintwindow changes. 160792 160793commit 5e209b21f3d6b3268fa7dab1e8df892d8352cc08 160794Author: Ben Byer <bbyer@bbyer.local> 160795Date: Wed Sep 19 03:44:46 2007 -0700 160796 160797 XNEST: removed obsolete hack to build under Darwin 160798 160799commit 378c7ebef444cdc9ae62ebf05c7111088a0c4bc1 160800Author: Ben Byer <bbyer@bbyer.local> 160801Date: Wed Sep 19 03:43:40 2007 -0700 160802 160803 XVFB: Removed obsolete hack to build on Darwin 160804 160805commit 7813392d1c9a38d01cfff17732278bb7798eee5d 160806Author: Ben Byer <bbyer@bbyer.local> 160807Date: Wed Sep 19 01:52:58 2007 -0700 160808 160809 XDarwin: changes to Makefile.am to use new conditionals 160810 160811commit 170fc77d9e73151f5c1c1f0f04598b3497db4284 160812Author: Dodji Seketeli <dodji@openedhand.com> 160813Date: Wed Sep 19 09:54:09 2007 +0200 160814 160815 xserver: don't force DTRACE detection by default 160816 160817 this fixes a breakage caused by 7a4ec34e256bf36b041c011a083916ad75a1d8bc. 160818 When running a non DTRACE aware system that is not darwin*, DTRACE was getting 160819 required. Now it is not anymore. 160820 160821commit 9b4f5157179a0d20756c16ea1f5130b171f72bb1 160822Author: Ben Byer <bbyer@bbyer.apple.com> 160823Date: Tue Sep 18 21:40:32 2007 -0700 160824 160825 XDarwin: We build many many different targets, let's break them out into autoconf conditionals 160826 160827commit 7a4ec34e256bf36b041c011a083916ad75a1d8bc 160828Author: Ben Byer <bbyer@bbyer.apple.com> 160829Date: Tue Sep 18 20:37:09 2007 -0700 160830 160831 XDarwin: Disable dtrace support on Darwin 9 (since it doesn't work ... yet ...) 160832 160833commit 3c19ec47b434d4ca84db58363cc053cc0b6aa413 160834Author: Ben Byer <bbyer@bbyer.apple.com> 160835Date: Tue Sep 18 20:15:44 2007 -0700 160836 160837 XDarwin: changed XF86BigFont support default from yes to auto so XDarwin can disable it 160838 160839commit 547ad2125ece93bbe01f6d09a3baf176ebd16bb3 160840Author: Michel Dänzer <michel@tungstengraphics.com> 160841Date: Mon Sep 17 20:33:56 2007 +0200 160842 160843 EXA: Make sure driver hooks get correct offscreen offsets from exaCopyDirty. 160844 160845 This should ensure the driver UploadTo/DownloadFromScreen hooks can always 160846 work as intended. 160847 160848commit 56cc24ffb21f7fd41f9ea9e8f969aa85021b9f53 160849Author: Alan Hourihane <alanh@tungstengraphics.com> 160850Date: Mon Sep 17 20:33:56 2007 +0200 160851 160852 EXA: Don't crash in ExaCheckPolyArc. 160853 160854 See https://bugs.freedesktop.org/show_bug.cgi?id=12286 . 160855 160856commit 7ac89060e02c7a803018afd580720f8326561fd8 160857Author: Michel Dänzer <michel@tungstengraphics.com> 160858Date: Mon Sep 17 20:33:55 2007 +0200 160859 160860 EXA: Don't crash in exaGetImage when swapped out. 160861 160862commit 3876c6c80534a6f7412ec806a2ba9ada22c5e505 160863Author: Tilman Sauerbeck <tilman@code-monkey.de> 160864Date: Mon Sep 17 18:47:45 2007 +0200 160865 160866 EXA: Fixed compiler warnings. 160867 160868commit d5738ff2e0f93df4729c075ce31a1041d580e50e 160869Author: Tilman Sauerbeck <tilman@code-monkey.de> 160870Date: Mon Sep 17 16:59:03 2007 +0200 160871 160872 EXA: Added missing exaPrepare/FinishAccess calls in ExaCheckPushPixels. 160873 160874commit a0dafd95e1e13a2f1b77ab9f82fd365a7be19de5 160875Author: Tilman Sauerbeck <tilman@code-monkey.de> 160876Date: Mon Sep 17 16:45:20 2007 +0200 160877 160878 Removed hw/xwin/winpntwin.c from Makefile.am. 160879 160880 winpntwin.c was removed in e4d11e58ce349dfe6af2f73ff341317f9b39684c. 160881 160882commit 7bd65577018a574970b767b67967b65fcd5c2bf0 160883Author: Tilman Sauerbeck <tilman@code-monkey.de> 160884Date: Sat Sep 15 14:01:57 2007 +0200 160885 160886 Initialize output->pendingProperties. 160887 160888commit b9f7aeb20015290a8acf938a11bf9272bf3527cf 160889Merge: e4d11e58c 06d27f804 160890Author: Eric Anholt <eric@anholt.net> 160891Date: Thu Sep 13 00:15:45 2007 +0000 160892 160893 Merge remote branch 'origin/master' into paint-window 160894 160895 Conflicts: 160896 160897 mi/miexpose.c 160898 160899commit e4d11e58ce349dfe6af2f73ff341317f9b39684c 160900Author: Eric Anholt <eric@anholt.net> 160901Date: Wed Sep 12 13:58:46 2007 +0000 160902 160903 Remove the PaintWindow optimization. 160904 160905 This was an attempt to avoid scratch gc creation and validation for paintwin 160906 because that was expensive. This is not the case in current servers, and the 160907 danger of failure to implement it correctly (as seen in all previous 160908 implementations) is high enough to justify removing it. No performance 160909 difference detected with x11perf -create -move -resize -circulate on Xvfb. 160910 Leave the screen hooks for PaintWindow* in for now to avoid ABI change. 160911 160912commit 06d27f8045966c1fb154eafaff308a01b93f265b 160913Author: Keith Packard <keithp@koto.keithp.com> 160914Date: Wed Sep 12 23:57:30 2007 +0100 160915 160916 Try again to fix drawable and tile offsets in miPaintWindow 160917 160918 Many coordinate spaces are hard. Let's go drinking. 160919 160920commit dd3992eb86377684a5dbe86fa19c756a9e53cda2 160921Author: Keith Packard <keithp@koto.keithp.com> 160922Date: Wed Sep 12 22:39:31 2007 +0100 160923 160924 miPaintWindow draw to window for background. 160925 160926 Instead of drawing to window pixmap for everything, draw to window for 160927 background as that works for Xnest and Xdmx; draw to pixmap for borders 160928 which neither of those X servers use. 160929 160930commit 6da39c67905500ab2db00a45cda4a9f756cdde96 160931Author: Eric Anholt <eric@anholt.net> 160932Date: Wed Sep 12 13:23:13 2007 +0000 160933 160934 Fix build on FreeBSD after Popen changes. 160935 160936commit 257c8ed17f4f908e0d0d5e53aaf13aa3b1313f50 160937Author: Keith Packard <keithp@koto.keithp.com> 160938Date: Wed Sep 12 12:11:49 2007 +0100 160939 160940 Rewrite miPaintWindow to draw to window pixmap. 160941 160942 miPaintWindow was drawing to the root window, or (sometimes) drawing to the 160943 window after smashing the window clip list. This is losing, and easily fixed 160944 by just drawing to the window pixmap. 160945 160946commit 43dc41034735d84765233a78bd619f3f0e7ace75 160947Author: Brice Goglin <bgoglin@debian.org> 160948Date: Wed Sep 12 01:43:37 2007 +0200 160949 160950 Replace a non-ascii char with the corresponding groff escape in exa.man.pre 160951 160952commit a5b8053606d6e786cdcf6734f271acc05f9cc588 160953Author: Adam Jackson <ajax@benzedrine.nwnk.net> 160954Date: Tue Sep 11 11:37:06 2007 -0400 160955 160956 Ignore - not just block - SIGALRM around Popen()/Pclose(). 160957 160958 Because our "popen" implementation uses stdio, and because nobody's stdio 160959 library is capable of surviving signals, we need to make absolutely sure 160960 that we hide the SIGALRM from the smart scheduler. Otherwise, when you 160961 open a menu in openoffice, and it recompiles XKB to deal with the 160962 accelerators, and you popen xkbcomp because we suck, then the scheduler 160963 will tell you you're taking forever doing something stupid, and the 160964 wait() code will get confused, and input will hang and your CPU usage 160965 slams to 100%. Down, not across. 160966 160967commit 2e3e08d31e908ceadeef16d6069cdaf8c61d7ed1 160968Author: Dodji Seketeli <dodji@openedhand.com> 160969Date: Tue Sep 11 08:50:41 2007 +0200 160970 160971 xserver: test presence of strlcpy in configure 160972 160973 * configure.ac: define HAVE_STRLCPY when function strlcat is present 160974 this serves at least to unbreak libxtrans code compilation. 160975 160976commit e8c73ac697aa472e10b4d8c093a715df0bf545af 160977Author: Dodji Seketeli <dodji@openedhand.com> 160978Date: Tue Sep 11 00:20:45 2007 +0200 160979 160980 Xephyr: fix compilation breakage. 160981 160982 * hw/kdrive/ephyr/ephyr_draw.c: 160983 (exaDDXDriverInit): pExaScr->hideOffscreenPixmapData does not exist 160984 anymore, so don't set it. 160985 160986commit 321e0a21600e418bbeb164043a9a21a0ff80cbe9 160987Author: Ben Byer <bbyer@bbyer.local> 160988Date: Sun Sep 9 16:28:41 2007 -0700 160989 160990 fixed cut-and-paste typo 160991 160992commit 0ff273fd1e2ea7242b3e6c5effb2a623ef32ec6f 160993Author: Ben Byer <bbyer@bbyer.apple.com> 160994Date: Thu Sep 6 18:52:26 2007 -0700 160995 160996 XDARWIN: fixes to make Xquartz build again 160997 160998commit f8637137ab43818d45d249b337820cbf427a05ec 160999Author: Ben Byer <bbyer@bbyer.local> 161000Date: Thu Sep 6 05:39:57 2007 -0700 161001 161002 XDARWIN: build fix for quartz/pseudoramiX.c 161003 161004commit e8093e15c7df7a3d5a9717bc9d7d7517b0743f29 161005Author: Michel Dänzer <michel@tungstengraphics.com> 161006Date: Mon Sep 3 13:52:29 2007 +0200 161007 161008 EXA: Exclude bits that will be overwritten from migration in exaCopyNtoN. 161009 161010 Also plug a region leak in exaPolyFillRect. 161011 161012commit e81af8ba643df3be53b0a46d9d4a0eaf21557c9e 161013Author: Michel Dänzer <michel@tungstengraphics.com> 161014Date: Mon Sep 3 13:14:29 2007 +0200 161015 161016 EXA: exaFillRegion{Solid,Tiled} improvements. 161017 161018 Use region to exclude bits that will be overwritten from migration. 161019 161020 Also make exaFillRegionSolid use the same logic as exaFillRegionTiled. 161021 161022commit 5f7da4da8de7449e1c2a4c679632a0b2a5858b7e 161023Author: Michel Dänzer <michel@tungstengraphics.com> 161024Date: Fri Aug 31 16:59:28 2007 +0200 161025 161026 EXA: Use exaShmPutImage for pushing glyphs to scratch pixmap in exaGlyphs. 161027 161028commit be922b30486abce3a8c13996d579b211a7b56f0e 161029Author: Michel Dänzer <michel@tungstengraphics.com> 161030Date: Thu Aug 30 13:59:07 2007 +0200 161031 161032 EXA: exa(Shm)PutImage improvements. 161033 161034 Improve exaShmPutImage performance and reuse its core in exaPutImage as it 161035 seems faster than the previous code when the driver doesn't provide an 161036 UploadToScreen hook. 161037 161038 Make sure all damage records are notified of the damage incurred by actual 161039 ShmPutImage calls. 161040 161041 Remove superfluous manual damage tracking for actual PutImage calls. 161042 161043commit ea92ea415665e294a1ba233e9a1d39b6daa0cee1 161044Author: Michel Dänzer <michel@tungstengraphics.com> 161045Date: Thu Aug 30 13:54:18 2007 +0200 161046 161047 EXA: exaGetImage improvements. 161048 161049 Use the new migration infrastructure to cache FB bits we need in the system 161050 copy, for the benefit of repeated calls. 161051 161052commit aa2ed73e0ec881947c969b67269e3206da4de359 161053Author: Michel Dänzer <michel@tungstengraphics.com> 161054Date: Thu Aug 30 13:50:42 2007 +0200 161055 161056 EXA: Remove superfluous manual damage tracking. 161057 161058 These should all be covered by damage wrappers. 161059 161060commit a634c9b03494ba80aeec28be19662ac96657cc23 161061Author: Michel Dänzer <michel@tungstengraphics.com> 161062Date: Thu Aug 30 13:48:03 2007 +0200 161063 161064 EXA: RENDER improvements. 161065 161066 Exclude bits that will be overwritten from migration. 161067 161068 Use exaGlyphs even when Composite can't be accelerated, to avoid PolyFillRect 161069 roundtrip via offscreen memory. 161070 161071 Initialize mask pixmap in exaGlyphs in FB in addition to system if the driver 161072 provides Composite hooks to avoid migration overhead. 161073 161074 Remove manual damage tracking where superfluous. 161075 161076commit 1f457ff3db24178eefecfbbf177aaf6554adb204 161077Author: Michel Dänzer <michel@tungstengraphics.com> 161078Date: Thu Aug 30 13:44:20 2007 +0200 161079 161080 EXA: Improvements for 1x1 pixmaps. 161081 161082 Initialize system and FB copy in exaFillRegionSolid and adapt 161083 exaGetPixmapFirstPixel to the new migration infrastructure. 161084 161085 This should mostly eliminate migration overhead for these, whether they are 161086 used for acceleration or fallbacks. 161087 161088commit 489bc7551ffc7360ba9648ca5c98b59c7e7a1fd1 161089Author: Michel Dänzer <michel@tungstengraphics.com> 161090Date: Thu Aug 30 13:37:53 2007 +0200 161091 161092 EXA: exaImageGlyphBlt improvements. 161093 161094 As we can't actually accelerate anything interesting here, just migrate out 161095 once and call fbSolidBoxClipped instead of taking a round trip via offscreen 161096 memory with exaSolidBoxClipped. 161097 161098 Reuse pending damage region for extents and to prevent any actual migration of 161099 pixmap contents when we're overwriting the whole pending damage region. 161100 161101 Remove superfluous manual damage tracking. 161102 161103commit 2e0895a4ba27c1308713022820444c8f57f7a69f 161104Author: Michel Dänzer <michel@tungstengraphics.com> 161105Date: Thu Aug 30 13:30:03 2007 +0200 161106 161107 EXA: Improvements for trapezoids and triangles. 161108 161109 Only migrate once in exaTrapezoids/Triangles instead of every time in 161110 exaRasterizeTrapezoid/AddTriangles. Adapt manual damage tracking to new 161111 infrastructure. 161112 161113 Also move definition of NeedsComponent() closer to where it's used. 161114 161115commit 6c9d7ed61bc4a19d21c53717b8af3d90b5d82ca9 161116Author: Michel Dänzer <michel@tungstengraphics.com> 161117Date: Thu Sep 6 13:10:16 2007 +0200 161118 161119 EXA: Hide pixmap pointer outside of exaPrepare/FinishAccess whenever possible. 161120 161121 We finally want to catch all cases where the pixmap pointer is dereferenced 161122 outside of exaPrepare/FinishAccess. 161123 161124 Also fix a couple of such cases exposed by this change. 161125 161126commit 962eddd7a2863a8475f5fd8107d3112df08d1172 161127Author: Michel Dänzer <michel@tungstengraphics.com> 161128Date: Wed Aug 29 19:55:22 2007 +0200 161129 161130 EXA: Support partial migration of pixmap contents between Sys and FB. 161131 161132 The initiator of migration can pass in a region that defines the relevant area 161133 of each source pixmap or the irrelevant area of the destination pixmap. By 161134 default, the pending damage region is assumed relevant for the destination 161135 pixmap, and everything for source pixmaps. 161136 161137 Thanks to Jarno Manninen for reassuring me that my own ideas for this were 161138 feasible and for providing additional ideas. 161139 161140commit f27931bdd26fc9a1e6bb5173b5537e32c51a98b3 161141Author: Michel Dänzer <michel@tungstengraphics.com> 161142Date: Wed Aug 29 19:41:52 2007 +0200 161143 161144 Add DamagePendingRegion. 161145 161146 DamagePendingRegion returns a pointer to the region of a drawable that will 161147 be damaged by the current operation for damage records that chose to get damage 161148 reported only at the end of the operation. 161149 161150commit 5c7ee3f47fa0c067102a17dee3f75a51cc0bdb3a 161151Author: Michel Dänzer <michel@tungstengraphics.com> 161152Date: Fri Aug 24 19:24:18 2007 +0200 161153 161154 EXA: Track valid bits in Sys and FB separately. 161155 161156 Also consolidate exaCopyDirtyToFb/Sys. 161157 161158commit 8cfcf9973c765f11d1b45b95b8091ef7e01d7f01 161159Author: Michel Dänzer <michel@tungstengraphics.com> 161160Date: Wed Sep 5 20:10:09 2007 +0200 161161 161162 EXA: Migrate out pixmap in exaPrepareAccess. 161163 161164 Also fix exaFinishAccessGC not to use the same index for tile and stipple. 161165 161166commit e510a77ba4d65d5d6ead514cd698f1b1e3f8a2b6 161167Author: Dave Airlie <airlied@linux.ie> 161168Date: Tue Jul 17 17:16:51 2007 +1000 161169 161170 EXA: Add a couple of missing exaPrepare/FinishAccess calls. 161171 161172commit 72b347e681f5667b68257822e7cec02ab4c9cb6d 161173Author: Peter Hutterer <peter@cs.unisa.edu.au> 161174Date: Fri Sep 7 14:31:19 2007 +0930 161175 161176 dix: Enabling devices must not overwrite existing sprites/pairing. 161177 161178 EnableDevices is (amongst others )called after a VT switch. We must not create 161179 a new sprite or re-pair the device, otherwise we lose the input device setup 161180 that we had before the VT switch. 161181 161182 This requires the devices to be in exactly the same order as before 161183 the VT switch. Removing a device while on a different VT is probably a bad 161184 idea. 161185 161186commit 57907e0943da0c3fd3bf6c128d210b544629ce72 161187Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161188Date: Thu Sep 6 16:55:51 2007 -0400 161189 161190 devPrivates rework: register an offset for every resource type, use 161191 signed values so -1 actually works correctly, and provide a macro for 161192 adding an offset to a pointer. 161193 161194commit 84eb7e62248ddc2761af8cefe33d1b7147477528 161195Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 161196Date: Thu Sep 6 13:36:37 2007 +0100 161197 161198 Fix Xdmx build 161199 161200commit 35ec8b89f4cdf5ec48b292d47b2dad42d2fb9534 161201Author: Ben Byer <bbyer@bbyer.local> 161202Date: Thu Sep 6 05:34:14 2007 -0700 161203 161204 DDX changes for XDarwin 161205 161206 Creating a real DDX section for XDarwin instead of 161207 stuffing it into the XORG section in a half-assed manner. 161208 161209commit d627061b48ae06d27b37be209d67a3f4f2388dd3 161210Author: Peter Hutterer <peter@cs.unisa.edu.au> 161211Date: Thu Sep 6 18:57:00 2007 +0930 161212 161213 xfree86: wrap keyboard devices for XKB. 161214 161215 Call ProcessOtherEvents first, then for all keyboard devices let them be 161216 wrapped by XKB. This way all XI events will go through XKB. 161217 161218 Note that the VCK is still not wrapped, so core events will bypass XKB. 161219 161220commit 5ee409794ee604fcf84886f70429fc2d6b1ff4f1 161221Author: Peter Hutterer <peter@cs.unisa.edu.au> 161222Date: Thu Sep 6 18:52:02 2007 +0930 161223 161224 dix: add XI event support to FixKeyState. 161225 161226 FixKeyState needs to be able to handle XI events, otherwise we get "impossible 161227 keyboard events" on server zaps and other special key combos. 161228 161229commit 6334d4e7be18de5f237c12a6dc20f75aa23477d0 161230Author: Peter Hutterer <peter@cs.unisa.edu.au> 161231Date: Thu Sep 6 18:49:57 2007 +0930 161232 161233 xkb: enable XI event processing for xkb. 161234 161235 XI events can now take the same processing paths as core events, and should do 161236 the correct state changes etc. 161237 161238 There's some cases where XKB will use KeyPress as type for an event to be 161239 delivered to the client. Stuck warnings in, not sure what the correct solution 161240 is yet. 161241 161242commit 49dbe9a757a3d7a0b9ab318242c6cc0cbd4dd1f0 161243Author: Peter Hutterer <peter@cs.unisa.edu.au> 161244Date: Thu Sep 6 17:27:28 2007 +0930 161245 161246 dix: close virtual core devices after other devices. 161247 161248 If a device is paired with the VCP, deleting the VCP before the device will 161249 segfault the server when the sprite should get updated. 161250 161251commit 03680a384aa423ece75b658f00b96db2628c39fa 161252Author: Peter Hutterer <peter@cs.unisa.edu.au> 161253Date: Thu Sep 6 15:43:47 2007 +0930 161254 161255 dix: don't change the device struct while processing core events. 161256 161257 The device state needs to be changed while processing the XI event. Core 161258 events are always processed after XI, so by then the device is already set up 161259 properly. However, we now rely on DeviceButtonMotionMask to be equal to 161260 ButtonMotionMask. It already is, but stick a big fat warning in so nobody 161261 attempts to change it. 161262 161263 This commit disables XKB for the VCK, thus essentially for all devices. 161264 Temporarily anyway. 161265 161266commit f3f614cd01ae56d84d61b4f5c907c6edd7c8c6d3 161267Author: Ben Byer <bbyer@bbyer.local> 161268Date: Wed Sep 5 20:34:00 2007 -0700 161269 161270 Revert "configure.ac: exclude pci-access dependency from XDarwin" 161271 161272 This reverts commit 20c6677d1b5f8d77325dd878ffa3df1d0fb01864. 161273 161274commit bf1641b94cffa54b786e18eaeff3839d8790b9f2 161275Author: Ben Byer <bbyer@bbyer.local> 161276Date: Wed Sep 5 20:33:43 2007 -0700 161277 161278 XDARWIN: Add launchd support 161279 161280 This adds a bit of glue to configure.ac to support launchd detection; 161281 on OS X (or other platforms which choose to implement launchd), this allows 161282 the system to automagically start the Xserver as necessary to serve clients. 161283 161284commit 71c21dea748ea0dcad758679c40ee39192d170f9 161285Author: Ben Byer <bbyer@bbyer.apple.com> 161286Date: Wed Sep 5 18:45:50 2007 -0700 161287 161288 added HAVE_LAUNCHD check to configure.ac (mostly for OSX) 161289 161290commit 81c28ffd2b13a83770eadcfd7829d35d319d637f 161291Author: Daniel Stone <daniel@fooishbar.org> 161292Date: Wed Sep 5 17:46:23 2007 -0700 161293 161294 Fix key repeats during VT switch. 161295 161296 Add keyc->postdown, which represents the key state as of the last mieqEnqueue 161297 call, and use it when we need to know the posted state, instead of the 161298 processed state (keyc->down). Add small functions to getevents.c to query and 161299 modify key state in postdown and use them all through, eliminating previously 161300 broken uses. 161301 161302commit e332335241af28ef0ab66b102d0cbc4e5c73ac68 161303Merge: 7381e9149 accd71bda 161304Author: Ben Byer <bbyer@bbyer.apple.com> 161305Date: Wed Sep 5 15:51:23 2007 -0700 161306 161307 Merge branch 'master' of ssh://git.freedesktop.org/git/xorg/xserver 161308 161309commit 7381e9149e3cbb7e672070781a16e3f096202be9 161310Author: Ben Byer <bbyer@bbyer.apple.com> 161311Date: Wed Sep 5 15:51:11 2007 -0700 161312 161313 added an exclusion for setting XORG on darwin 161314 161315commit accd71bda6f958ea6892ad3a10879232d345774c 161316Author: Keith Packard <keithp@koto.keithp.com> 161317Date: Wed Sep 5 14:19:19 2007 -0700 161318 161319 Deliver correct event when releasing keys on VT switch. 161320 161321 In commit 41bb9fce47f6366cc3f7d45790f7883f74289b5a, the event delivery loop 161322 for Xinput enabled keyboards was changed and accidentally used the wrong 161323 index variable, causing random events to be delivered when returning from VT 161324 switch. 161325 161326 In addition, in commit aeba855b07832354f59678e20cc29a085e42bd99, 161327 SIGIO was blocked during delivery of these events, but not for the entire 161328 period the xf86Events array was being used. Block SIGIO for the whole loop 161329 to avoid other event delivery from trashing the key release events. 161330 (cherry picked from commit aa7ed1f5f35cd043bc38d985500aa0a32e857e84) 161331 161332commit 8b77dc7e808f61f1ed10fe05cf898bb47459a76d 161333Author: Ben Byer <bbyer@bbyer.apple.com> 161334Date: Wed Sep 5 14:48:38 2007 -0700 161335 161336 XDARWIN: build breakage fix 161337 161338commit 20c6677d1b5f8d77325dd878ffa3df1d0fb01864 161339Author: Ben Byer <bbyer@bbyer.apple.com> 161340Date: Wed Sep 5 14:31:01 2007 -0700 161341 161342 configure.ac: exclude pci-access dependency from XDarwin 161343 161344 XDarwin doesn't need any of this pci stuff since it doesn't talk directly to hardware, 161345 so it doesn't make sense to require it when building on OSX/Darwin. 161346 161347commit ff01e44e33fd072958fb0157dae072f1b1c88944 161348Merge: 8ba8c16af 47300ed2b 161349Author: Ben Byer <bbyer@bbyer.apple.com> 161350Date: Wed Sep 5 13:56:08 2007 -0700 161351 161352 Merge branch 'master' of ssh://git.freedesktop.org/git/xorg/xserver 161353 161354commit 8ba8c16af773ec83a0b1c0661a23d746b401944e 161355Author: Ben Byer <bbyer@bbyer.apple.com> 161356Date: Wed Sep 5 13:46:30 2007 -0700 161357 161358 build fixes for XDarwin 161359 161360commit 47300ed2be59d0ba7ea9345b954bf3104877c095 161361Author: Eric Anholt <eric@anholt.net> 161362Date: Wed Sep 5 12:34:29 2007 -0700 161363 161364 Fix server version reporting to be the server package version. 161365 161366 Previously, the server version reported by xdpyinfo and Xorg -version would 161367 bear some vague resemblance to a X.Org katamari version, but in the presence 161368 of modularization (and client-server relationships with different katamari 161369 versions on each side) those numbers don't really make sense. Instead, just 161370 report the package version. 161371 161372 When branching a stable branch, master's version should be immediately updated 161373 to the endpoint of the stable branch plus a snapshot of 1 (for example, 161374 1.4.0.1 after server-1.4-branch). The stable branch should then be changed to 161375 RC0 at that time (1.3.99.0, for example). 161376 161377 This scheme was partially attempted for server 1.3, but lacked the appropriate 161378 master updates, thus why it had to be revisited now. While here, we can also 161379 remove a lot of versioning complexity since everything is based on the package 161380 version. 161381 161382commit 6c89d1237c4fdce961b30a8eaee964af5d56565e 161383Author: Dodji Seketeli <dodji@openedhand.com> 161384Date: Wed Sep 5 17:46:49 2007 +0200 161385 161386 Kdrive: unbreak kdrive linking 161387 161388 * configure.ac: re-sort Kdrive libs so that symbols get properly resolved. 161389 Basically, all some libs are present in both $KDRIVE_LIBS and $XSERVER_LIBS, 161390 and some libs orders are not correct. So I made sure Kdrive servers don't have 161391 to link against $KDRIVE_LIBS *and* $XSERVER_LIBS. They just have to link 161392 against $KDRIVE_LIBS now. 161393 * hw/kdrive/*/Makefile.am: update those makefile to reflect the change in configure.ac 161394 161395commit 0003ccfcdfae1b473aa024342304b84256d378b9 161396Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161397Date: Wed Sep 5 11:18:36 2007 -0400 161398 161399 xace: add new fields to resource access hook to allow parent resource objects 161400 to be passed in at create time. Also added a missing devPrivates initializer. 161401 161402commit 28e48cd8e6e4c412a49d7177daad6d3c93c28e08 161403Author: Ben Byer <bbyer@bbyer.local> 161404Date: Wed Sep 5 04:43:17 2007 -0700 161405 161406 Another pathname fix for event_status_driver.h 161407 161408commit b6c0697fd67323893a9ad3676c33f2f1ec48e15e 161409Author: Ben Byer <bbyer@bbyer.local> 161410Date: Wed Sep 5 04:40:03 2007 -0700 161411 161412 fixed path for event_status_driver.h 161413 (The path currently used is deprecated on Tiger and invalid on Leopard.) 161414 161415commit c4fff050836feeef8390b7197f1de39af2997811 161416Author: Peter Hutterer <peter@cs.unisa.edu.au> 161417Date: Wed Sep 5 16:19:45 2007 +0930 161418 161419 Revert "Input: Fix stuck modifiers (bug #11683)" 161420 161421 This reverts commit 6b055e5d9751e3679ff98065e43225ec8a960053. 161422 161423 MPX relies on the XI event being delivered before the core event. Device grabs 161424 break, amongst other things. I guess stuck modifiers need to be fixed some 161425 other way. 161426 161427 Conflicts: 161428 161429 dix/getevents.c 161430 161431commit bfe6b4d2d9952a80f8dbc63eec974ef894e5c226 161432Author: Peter Hutterer <peter@cs.unisa.edu.au> 161433Date: Tue Sep 4 17:44:51 2007 +0930 161434 161435 xkb: Store the action filters per device in the XkbSrvInfoRec. 161436 161437 Using a global array for action filters is bad. If two keyboard hit a modifier 161438 at the same time, releaseing the first one will deactivate the filter and 161439 thus the second keyboard can never release the modifier again. 161440 161441commit cc5c926267be099d793e6dfec17916f21c73c64d 161442Author: Peter Hutterer <peter@cs.unisa.edu.au> 161443Date: Thu Aug 30 15:51:22 2007 +0930 161444 161445 randr: RRPointerScreenConfigured needs to move all pointers. 161446 161447 Previous version only moved the VCP, causing "bogus pointer events" lateron. 161448 Now we run through the device list, updating each pointer separately if 161449 necessary. 161450 161451 Also stick a big warning into RRPointerMoved, not sure what device we need to 161452 work on here. 161453 161454commit 136fde2c7b5ed590bc6c63d11ede31e92dc679a2 161455Author: Ademar de Souza Reis Jr <ademar@mandriva.com.br> 161456Date: Tue Jul 3 13:44:28 2007 -0300 161457 161458 Add Xserver man section about catalogue:<dir> FPE 161459 161460 catalogue:<dir> FPEs were introduced in libXfont 1.2.9 161461 161462commit eb6a933dc60bec5601260794eeb973e946af37b6 161463Author: Ademar de Souza Reis Jr <ademar@mandriva.com.br> 161464Date: Tue Jul 3 13:44:03 2007 -0300 161465 161466 Add xorg.conf man section about catalogue:<dir> FPE 161467 161468 catalogue:<dir> FPEs were introduced in libXfont 1.2.9 161469 161470commit 205183a733237ea418a25c7423b689fcc8eae628 161471Author: Eric Anholt <eric@anholt.net> 161472Date: Tue Sep 4 16:19:22 2007 -0700 161473 161474 Fix driver build by including an appropriate Requires.private line on pixman. 161475 161476 We'd previously been substituting PIXMAN_CFLAGS, but we've got a better tool 161477 now, plus I deleted the PIXMAN_CFLAGS substitution without noticing. 161478 161479commit d67e210f3458b62d7d4a6032aabfda0004d661c1 161480Author: Gerte Hoogewerf <g.hoogewerf@gmail.com> 161481Date: Tue Sep 4 16:09:38 2007 -0700 161482 161483 Add stub symbols to make xprint build. 161484 161485commit f7f79724fdea0cc6fda0e90e56431df937d49335 161486Author: Eric Anholt <eric@anholt.net> 161487Date: Tue Sep 4 15:10:49 2007 -0700 161488 161489 Increase despair by fixing xprint build after my _DEPENDENCIES changes. 161490 161491commit e89d16be07e45e487913509788a9e8cb1ee09bc7 161492Author: Ian Romanick <idr@us.ibm.com> 161493Date: Tue Sep 4 14:49:49 2007 -0700 161494 161495 Revert part of 529acb175440969af9d7fa38aab8d7dea0dc2661 because libtool is smart. 161496 161497commit bf5948518763b5e21eff806a0a9abc5f7757fa10 161498Merge: 4062db402 735da3dcd 161499Author: Matthieu Herrb <matthieu@bluenote.herrb.com> 161500Date: Tue Sep 4 22:40:31 2007 +0200 161501 161502 Merge branch 'master' of git+ssh://herrb@git.freedesktop.org/git/xorg/xserver 161503 161504commit 4062db4020c671fc0921a3a4e7fe5d8dc1be2e1d 161505Author: Matthieu Herrb <matthieu@bluenote.herrb.com> 161506Date: Tue Sep 4 22:39:14 2007 +0200 161507 161508 typo in comment 161509 161510commit ce9e83d913511fe619da42f805d7bcd1a2a60d90 161511Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161512Date: Tue Sep 4 14:01:55 2007 -0400 161513 161514 xace: add hooks + new access codes: Damage extension 161515 161516commit 735da3dcd09f59c40f68eca8c9a91c2e826cd1b9 161517Author: Egbert Eich <eich@freedesktop.org> 161518Date: Tue Sep 4 16:42:57 2007 +0200 161519 161520 Fixing xf86I2CGetScreenBuses(), now dereferencing the correct pointer. 161521 161522commit 7c5de093d499de63d39fe038d86da0a085262017 161523Author: Egbert Eich <eich@freedesktop.org> 161524Date: Tue Sep 4 15:07:17 2007 +0200 161525 161526 Fixing sig11 in xf86I2CGetScreenBuses(). 161527 161528 Dereferencing a pointer once too often caused a 161529 sig11 in xf86I2CGetScreenBuses(). 161530 161531commit 9adea807038b64292403ede982075fe1dcfd4c9a 161532Author: Hong Liu <hong.liu@intel.com> 161533Date: Tue Sep 4 08:46:46 2007 +0100 161534 161535 bgPixel (unsigned long) is 64-bit on x86_64, so -1 != 0xffffffff 161536 161537 This patch should fix bug 8080. 161538 161539commit 529acb175440969af9d7fa38aab8d7dea0dc2661 161540Author: Eric Anholt <eric@anholt.net> 161541Date: Sun Sep 2 15:16:01 2007 -0700 161542 161543 Fix Xorg build by listing circular dependency libraries twice. 161544 161545 One of these I introduced by listing dix and mi in the same library list to 161546 simplify other servers. The other had been hacked around using libosandcommon, 161547 which is now gone. 161548 161549commit c56930e6ca90a61d2f4cbd845f7d0a51a66f83cb 161550Author: Eric Anholt <eric@anholt.net> 161551Date: Fri Aug 31 18:27:41 2007 -0700 161552 161553 Remove backend.[ch] from neomagic to fix distcheck. 161554 161555commit f98dfec79dadb70fa7bba84e7335f92b3a73dc02 161556Author: Keith Packard <keithp@koto.keithp.com> 161557Date: Sat Sep 1 21:14:22 2007 -0700 161558 161559 [COMPOSITE] Composite used for pixmap population on redirect. (Bug #7447) 161560 161561 compNewPixmap copies bits from the parent window to the redirected child 161562 pixmap to populate the pixmap with reasonable data. It cannot always use 161563 CopyArea as that only works across matching depths. Use Composite when 161564 the depths do not match. 161565 161566commit 1afdf8b0a92437dffe84fa98b6083b3d8fd55e27 161567Author: Adam Jackson <ajax@redhat.com> 161568Date: Fri Aug 31 22:11:13 2007 -0700 161569 161570 [RANDR] Don't mark Xinerama as active if no crtcs are enabled. (bug #11504). 161571 161572 Clients expect any Xinerama-enabled screen to report at least one 161573 monitor, but with RandR, there may not be any enabled crtcs. In this case, 161574 tell the client that Xinerama is not active. 161575 161576commit 0dc2bb6101704d0fd25f36e2c3df79687f119f5b 161577Author: Marius Gedminas <mgedmin@b4net.lt> 161578Date: Fri Aug 31 21:36:37 2007 -0700 161579 161580 [RANDR] Compare only milliseconds of config time. (Bug #6502) 161581 161582 The timestamp transferred in the X protocol is a 32-bit number of 161583 milliseconds. 161584 161585 The timestamp stored in the server is a structure that contains two fields: 161586 months (!) and milliseconds. 161587 161588 When the server passes the config timestamp to the client, it discards the 161589 months part and sends only the milliseconds part. 161590 161591 When the server receives the config timestamp from the client, it tries to 161592 guess the "months" part by looking at the current time and then maybe adding 161593 or 161594 subtracting one. The guess is wrong after the server has been running long 161595 enough (several hours). 161596 161597 I have added two ErrorF calls around the 'if' statement that returns 161598 RRSetConfigInvalidConfigTimestamp in randr/randr.c and my Xorg.0.log has 161599 this: 161600 161601 randr request got good config time: 0:-2103495671 161602 161603 for the first few successful xrandr calls, and 161604 161605 randr request failed with RRSetConfigInvalidConfigTime: client passed 161606 1:-2103495671, server has 0:-2103495671 161607 161608 when it fails. The server has been running for 8 and a half hours. 161609 161610 The obvious fix would be to ignore the months field and only compare the 161611 milliseconds. 161612 161613commit 07630d897ef37cad8b79d073d9edc891d5a7bddd 161614Author: Eric Anholt <eric@anholt.net> 161615Date: Fri Aug 31 15:16:01 2007 -0700 161616 161617 Bug #7364: Require renderproto 0.9.3 on 64-bit, and fix build with it. 161618 161619commit ca82d4bddf235c9b68d51d68636bab40eafb9889 161620Author: Eric Anholt <eric@anholt.net> 161621Date: Fri Aug 31 13:00:23 2007 -0700 161622 161623 Bug #7186: Fix an excessive request size limitation that broke big-requests. 161624 161625 MAXBUFSIZE appears to be a leftover of some previous time. Instead, just 161626 use maxBigRequestSize when bigreqs are available (limiting buffers to ~16MB). 161627 When bigreqs are not available, needed won't be larger than the maximum 161628 size of a non-bigreqs request (256kB). 161629 161630commit a02db0d500cac20d0f0f107d27c064a175018421 161631Author: Eric Anholt <eric@anholt.net> 161632Date: Fri Aug 31 12:24:07 2007 -0700 161633 161634 Convert servers to using _DEPENDENCIES to ensure proper rebuilds. 161635 161636 This cleans up server Makefile.ams a little bit, but also means that people 161637 messing with configure.ac need to be careful with whether they put libraries 161638 in the _LIBS or _SYS_LIBS targets. Hopefully the comment in configure.ac will 161639 clarify the issues. 161640 161641commit c9ceb4878063ca22487c708d9d1f86e367f2cec8 161642Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161643Date: Fri Aug 31 11:03:54 2007 -0400 161644 161645 xace: add hooks + new access codes: Composite extension 161646 161647commit fd04b983db6a70bf747abe02ca07c1fbbaae6343 161648Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161649Date: Fri Aug 31 09:55:27 2007 -0400 161650 161651 xace: add hooks + new access codes: Render extension 161652 161653commit 3f42af8c0ef1e5379bc836f589e0cbee43c02ac5 161654Author: Peter Hutterer <peter@cs.unisa.edu.au> 161655Date: Thu Aug 30 18:22:12 2007 +0930 161656 161657 config: Use [config/dbus] consistently for error messages. 161658 161659commit 0fcde83d94507eadd9f99d4e6a63584b221c989c 161660Author: Peter Hutterer <peter@cs.unisa.edu.au> 161661Date: Thu Aug 30 18:20:20 2007 +0930 161662 161663 config: return BadValue to caller if add/remove doesn't have parameters. 161664 161665 If message iterator cannot be created, the caller didn't supply any 161666 parameters. Return BadValue, instead of dying a horrible death while being 161667 stuck in an endless loop. 161668 161669commit 87495fc7064d5e0a7575a0713b6895a4172df0fa 161670Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 161671Date: Thu Aug 30 21:57:41 2007 +0100 161672 161673 Allow yres_virtual to be greater for some kernel fbdev drivers. 161674 (temporary fix for older fbdev drivers) 161675 161676commit 1005b29cc6939851b40397cc9cd0de9476ad3046 161677Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161678Date: Thu Aug 30 14:48:24 2007 -0400 161679 161680 xace: Correct some access modes. 161681 161682commit d8ab2353cbd7694b556b7b9d550104dc8c460a1b 161683Author: Eric Anholt <eric@anholt.net> 161684Date: Thu Aug 30 10:39:53 2007 -0700 161685 161686 Remove dead xf86GetPciSizeFromOS and xf86GetPciOffsetFromOS. 161687 161688commit c2d80529fc7f514d80cf3cbed6f580cb999aca1b 161689Author: Eric Anholt <eric@anholt.net> 161690Date: Thu Aug 30 10:20:55 2007 -0700 161691 161692 Remove the now-dead PciAvoid symbol. 161693 161694commit 53f346b158fa8e10de5a8777fa6d8d86f918878b 161695Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161696Date: Thu Aug 30 13:20:04 2007 -0400 161697 161698 xace: add hooks + new access codes: SHAPE extension 161699 161700commit 766c693ef3637ee6fc402df594060ed2c1346761 161701Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161702Date: Thu Aug 30 13:06:28 2007 -0400 161703 161704 xace: add hooks + new access codes: MIT-SCREEN-SAVER extension 161705 161706commit cda92bbf12107865e93c03c71b901ef51466dc31 161707Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161708Date: Thu Aug 30 11:48:45 2007 -0400 161709 161710 xace: add hooks + new access codes: XFixes extension. 161711 161712 Required a new name argument to the selection access hook to handle 161713 XFixesSelectSelectionInput. 161714 161715commit 47ab4d648b31ea1d5800e0bc84cf5f25025bffe3 161716Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161717Date: Thu Aug 30 11:40:39 2007 -0400 161718 161719 devPrivates rework: convert CursorRec and CursorBits over to new interface. 161720 161721commit 1d11e4bc4ccb169fb23fc18583f0b648f0a6a4e0 161722Author: Egbert Eich <eich@freedesktop.org> 161723Date: Thu Aug 30 12:50:21 2007 +0200 161724 161725 Fixing a misleading comment which could suggest a GPL violation. 161726 161727 The author of the int10 code looked at the VBIOS POSTing code 161728 in DOSEMU to get some initial idea on how to POST a VBIOS. 161729 To give credit to the DOSEMU Team for this inspiration a comment 161730 was added to the code which could suggest that code from the 161731 GPLed DOSEMU was directly incorporated into this code. 161732 This patch should clearify the situation. 161733 161734commit 45efe85003195bd45501630cef08349abb180c3a 161735Author: Eric Anholt <eric@anholt.net> 161736Date: Wed Aug 29 16:05:51 2007 -0700 161737 161738 Remove stale changelogs from kdrive. 161739 161740commit 5aaf00190157114780ab51f7268b396459ed1cad 161741Author: Eric Anholt <eric@anholt.net> 161742Date: Wed Aug 29 16:05:23 2007 -0700 161743 161744 Add more generated files to .gitignore. 161745 161746commit 87295b66a972a2bd194a79af6aa4f715018fcded 161747Author: Eric Anholt <eric@anholt.net> 161748Date: Wed Aug 29 15:54:32 2007 -0700 161749 161750 Bug #9629: Remove badly-licensed neomagic kdrive files. 161751 161752 Licensing issues of these files include: 161753 - They claim to be licensed under the GPL, yet we haven't allowed that in the 161754 xserver repository in the past. 161755 - They refer the user to the top of the tree for GPL license text, yet it isn't 161756 there. 161757 - They claim to be derived from the (MIT-licensed) ati kdrive code, yet don't 161758 follow the licensing terms of those files. 161759 161760commit 4795df62456b73c6790f271e0a20a83c60496490 161761Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161762Date: Wed Aug 29 14:40:10 2007 -0400 161763 161764 xace: add hooks + new access codes: TOG-CUP extension. 161765 161766commit e39694789e31e221fc8dec44ace9c697daf7acad 161767Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161768Date: Wed Aug 29 14:16:46 2007 -0400 161769 161770 xace: drop map-window checking hook, add new hooks for controlling the 161771 delivery of events to windows and clients. 161772 161773 This is tentative. It's likely that an additional last-resort hook will 161774 be necessary for code that calls TryClientEvents or WriteEventsToClient 161775 directly. It's also possible that new xace machinery will be necessary 161776 to classify events and pull useful resource ID's out of them. 161777 161778 The failure case also needs some thinking through. Should event delivery 161779 "succeed" or should it report undeliverable? 161780 161781 Finally, XKB appears to call WriteToClient to pass events. Sigh. 161782 161783commit 41355a53c29bbf879da0c6ea562294fcc7ef89ff 161784Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161785Date: Tue Aug 28 15:10:20 2007 -0400 161786 161787 xace: add hooks + new access codes: core protocol input requests 161788 161789commit adf46b57ce6c69ab13a38b09a8104c802d54d052 161790Author: Eric Anholt <eric@anholt.net> 161791Date: Tue Aug 28 10:08:38 2007 -0700 161792 161793 Replace BSD custom PCI code with a stub implementation thanks to pciaccess. 161794 161795 Note that pciaccess doesn't yet have Net/OpenBSD support, but the relevant 161796 code should go there instead of disconnected code in the X Server. 161797 161798 While here, remove the now-disabled INCLUDE_XF86_NO_DOMAIN from the headers, 161799 and un-disable xf8StdAccResFromOS for those OSes without domain support which 161800 will need it. 161801 161802commit 4017d3190234e189a0bbd33193a148d4d3c7556b 161803Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161804Date: Tue Aug 28 09:28:25 2007 -0400 161805 161806 devPrivates rework: since API is already broken, switch everything 161807 over to new system. 161808 161809 Need to update documentation and address some remaining vestiges of 161810 old system such as CursorRec structure, fb "offman" structure, and 161811 FontRec privates. 161812 161813commit 85547073265ae9bc4ae3af920a6d3214fd1ca0c5 161814Merge: 860a09cfb 7d54399cf 161815Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 161816Date: Tue Aug 28 07:25:21 2007 -0400 161817 161818 Merge branch 'master' into XACE-SELINUX 161819 161820 Conflicts: 161821 161822 include/miscstruct.h 161823 mi/mibstore.c 161824 mi/midispcur.c 161825 os/Makefile.am 161826 161827commit 3fe67d23edaae3ddde20cd5f349aa5dfde1d26a3 161828Author: Eric Anholt <eric@anholt.net> 161829Date: Mon Aug 27 19:02:41 2007 -0700 161830 161831 Remove the BusAccWindows resource code which is now unused. 161832 161833 This was a bunch of poorly defined resource ranges per OS/platform combination 161834 which were supposed to represent what regions could potentially have resources 161835 allocated into them. 161836 161837commit 9e2112b2b56af099a7f380ece9b5c1d25b20cce4 161838Author: Eric Anholt <eric@anholt.net> 161839Date: Mon Aug 27 18:50:34 2007 -0700 161840 161841 Remove unused xf8GetBlock and xf86GetSparse entry points. 161842 161843commit 801c359574d08ff2d6ac75a3325ff522bc862e30 161844Author: Eric Anholt <eric@anholt.net> 161845Date: Mon Aug 27 15:46:05 2007 -0700 161846 161847 Fix fbdevhwstub for pci-rework. 161848 161849commit 7d54399cfdaa7f54e28828267a76b89c4e8e798f 161850Author: Keith Packard <keithp@koto.keithp.com> 161851Date: Sun Aug 26 22:11:45 2007 -0700 161852 161853 Add XSERVER_LIBPCIACCESS to xorg-server.h when using libpciaccess 161854 161855commit affda73a1d6e291516880dfbcb74b661374524c6 161856Author: Eugeniy Meshcheryakov <eugen@univ.kiev.ua> 161857Date: Mon Aug 27 00:41:03 2007 +1000 161858 161859 Xprint: fix handling of TrueType font name 161860 161861 Debian bug #272368 161862 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=272368 161863 In certain locales, some characters from some TrueType fonts were not 161864 appearing in the Xprint postscript output due to the font not being 161865 identified in the postscript output. 161866 161867commit 6f44a2c8a8d6e8f95681ebc7b2dd5ad9c3a02c6e 161868Author: Adam Jackson <ajax@benzedrine.nwnk.net> 161869Date: Sun Jul 29 14:16:10 2007 -0400 161870 161871 Refactor PictureInitIndexedFormats. 161872 161873 The plural version is now static, which is fine since it was only ever called 161874 from within picture post-init anyway. The body of the work is now done with 161875 a one-shot (public) function that operates on a single format at a time. 161876 161877commit ae7f71a8b3d6756161e55d998d6eec37d2695c98 161878Author: Adam Jackson <ajax@benzedrine.nwnk.net> 161879Date: Sat Aug 25 15:08:20 2007 -0400 161880 161881 Implement core protocol backing store exclusively in terms of Composite. 161882 161883 Composite's automatic redirection is a more general mechanism than the 161884 ad-hoc BS machinery, so it's much prettier to implement the one in terms 161885 of the other. Composite now wraps ChangeWindowAttributes and activates 161886 automatic redirection for windows with backing store requested. The old 161887 backing store infrastructure is completely gutted: ABI-visible structures 161888 retain the function pointers, but they never get called, and all the 161889 open-coded conditionals throughout the DIX layer to implement BS are gone. 161890 161891 Note that this is still not a strictly complete implementation of backing 161892 store, since Composite will throw the bits away on unmap and therefore 161893 WhenMapped and Always hints are equivalent. 161894 161895commit bf0883ae5081bd75569115a3eb27c6d3d336c9f2 161896Author: David Nusinow <dnusinow@debian.org> 161897Date: Sat Aug 25 14:53:17 2007 -0400 161898 161899 Fix bug in debugging info related to pci-rework merge 161900 161901commit 4eed88af8b7b6881b44ccf9f4a5c5875af6d2b78 161902Author: Ian Romanick <idr@us.ibm.com> 161903Date: Fri Aug 24 18:06:50 2007 -0700 161904 161905 Remove files made obsolete by pci-rework. 161906 161907commit 1ab4b3e183d04bb20fb5039f7d2671752ab24cf6 161908Author: Aaron Plattner <aplattner@nvidia.com> 161909Date: Fri Aug 24 16:00:31 2007 -0700 161910 161911 Bump video driver ABI for pci-rework. 161912 161913commit 91f358336f77c0e4f577be65cca977d17298e36c 161914Merge: b9a806f0b 8b6b40b72 161915Author: Ian Romanick <idr@us.ibm.com> 161916Date: Fri Aug 24 15:04:21 2007 -0700 161917 161918 Merge branch 'pci-rework' 161919 161920commit b9a806f0b3d495c7616b469281e5892ae7f3f6b3 161921Author: Michel Dänzer <michel@tungstengraphics.com> 161922Date: Fri Aug 24 19:04:55 2007 +0200 161923 161924 exaPolyFillRect: pGC->alu doesn't matter with a single rectangle. 161925 161926commit d0f0d1092c7587a02404e1db07740e6334462ba6 161927Author: Michel Dänzer <michel@tungstengraphics.com> 161928Date: Fri Aug 24 14:10:13 2007 +0200 161929 161930 exaGetImage: Don't migrate pixmap out of FB with no DownloadFromScreen hook. 161931 161932 Based on the assumption that GetImage is relatively rare, so the overhead of 161933 the migration is probably bigger than any potential savings. 161934 161935commit 5d9e2c282145897008d7d941e2a0a3fdc71f2373 161936Author: Michel Dänzer <michel@tungstengraphics.com> 161937Date: Fri Aug 24 14:03:14 2007 +0200 161938 161939 EXA: Improve ShmPutImage. 161940 161941 Share as much code with exaPutImage as possible, and fall back to fbShmPutImage 161942 when that fails. 161943 161944commit 6085522d91e875c0e1ab8d4300e7378701c19b7c 161945Author: Michel Dänzer <michel@tungstengraphics.com> 161946Date: Fri Aug 24 14:02:35 2007 +0200 161947 161948 Export fbShmPutImage to modules. 161949 161950 To be used by EXA. 161951 161952commit c19f227b468d039c5ea136cc8a53c420da30263b 161953Author: Michel Dänzer <michel@tungstengraphics.com> 161954Date: Fri Aug 24 13:05:52 2007 +0200 161955 161956 EXA: Only mark offscreen memory as used when it really is. 161957 161958commit 095850596114178119a8cc854716ce0cc6e05121 161959Author: Michel Dänzer <michel@tungstengraphics.com> 161960Date: Fri Aug 24 13:04:48 2007 +0200 161961 161962 __glXDRIbindTexImage: Fail if no texture bound to pixmap's texture target. 161963 161964 We would most likely crash somewhere in Mesa if we tried to continue in this 161965 case. 161966 161967commit 8b6b40b7271acd81a9548f502c18f46f3b640640 161968Merge: ab7a6d860 3305d1719 161969Author: Ian Romanick <idr@us.ibm.com> 161970Date: Thu Aug 23 18:19:17 2007 -0700 161971 161972 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 161973 161974 Conflicts: 161975 161976 hw/xfree86/common/xf86.h 161977 hw/xfree86/common/xf86Init.c 161978 hw/xfree86/common/xf86pciBus.c 161979 hw/xfree86/int10/generic.c 161980 hw/xfree86/int10/helper_exec.c 161981 hw/xfree86/loader/xf86sym.c 161982 hw/xfree86/os-support/bus/Pci.c 161983 hw/xfree86/os-support/bus/Pci.h 161984 hw/xfree86/os-support/bus/linuxPci.c 161985 hw/xfree86/os-support/linux/int10/linux.c 161986 161987commit 93ae6fe18c417a22f1fccb22add4890a20cae713 161988Author: Carl Worth <cworth@cworth.org> 161989Date: Thu Aug 23 16:33:05 2007 -0700 161990 161991 Avoid leaking a Pixmap for every glyph 161992 161993commit 3305d17195e3a0a5555300555bd7703312fa489f 161994Author: Matthieu Herrb <matthieu@deville.herrb.com> 161995Date: Thu Aug 23 22:48:19 2007 +0200 161996 161997 Fix indentation. 161998 161999commit a66c0f1dca2958835ff65a5b50579e3304ed316a 162000Author: Matthieu Herrb <matthieu@bluenote.herrb.com> 162001Date: Thu Aug 23 22:11:56 2007 +0200 162002 162003 Remove an extra cast. 162004 162005 Thou should not apply patches manually without testing. 162006 162007commit 12d27cf33c6d963eae77795c0d247175907162a5 162008Author: Otto Moerbeek <otto@openbsd.org> 162009Date: Thu Aug 23 21:59:25 2007 +0200 162010 162011 A high resolution device that's moving fast can potentially generate 162012 an int overflow, making dx*dx+dy*dy negative. Now pow(negative, 162013 non-integer) yields NaN, so you loose. Use fp math to avoid that. 162014 162015commit ff089e6cae634ac3eb509abd448a250bcbb17275 162016Author: Brian Paul <brian.paul@tungstengraphics.com> 162017Date: Thu Aug 23 19:38:53 2007 +0200 162018 162019 glx: fix crash when freeing visuals 162020 162021 Don't set screen->num_vis to a value greater than the actual number of visuals. 162022 162023 X.Org Bug #10809 <http://bugs.freedesktop.org/show_bug.cgi?id=10809> 162024 162025commit 943dd6ad99670c283a6869ea6c5f751acbd73134 162026Author: Søren Sandmann Pedersen <sandmann@redhat.com> 162027Date: Thu Aug 23 12:15:03 2007 -0400 162028 162029 Revert "Revert "Require pixman 0.9.5; Use pixman_image_set_source_clipping() to fix"" since the pixman changes have been pushed now. 162030 162031 This reverts commit 57f7f2a5327a2d967a726bb4706e4f6b2f4b2cea. 162032 162033commit 13949f997289068354e83bc83e50d97b8232efb1 162034Author: Julien Cristau <jcristau@debian.org> 162035Date: Thu Aug 23 11:27:33 2007 +0200 162036 162037 Get rid of the type1 font module. 162038 162039 It is completely replaced by freetype these days. 162040 162041commit 88a9828ef906bba973debc191e35ea669b7ec271 162042Author: Peter Hutterer <peter@cs.unisa.edu.au> 162043Date: Thu Aug 23 18:02:10 2007 +0930 162044 162045 dix: Only check device events for possible ACLs. 162046 162047 We shouldn't be able to restrict events like Expose, etc. with device based 162048 ACLs. So we just ignore all non-input events when checking for permissions. 162049 162050commit 76bf3cd7b8c6189b6b08518cde00c8bd991bdfb7 162051Author: Dave Airlie <airlied@redhat.com> 162052Date: Thu Aug 23 16:22:03 2007 +1000 162053 162054 randr: fixup crtc and output destroy 162055 162056 if you are moving pointers, you want to move the pointers not just a byte 162057 162058commit 81f8b652d99ee0f7116c1e34aed0e585d23a91fb 162059Author: Alex Deucher <alex@botch2.(none)> 162060Date: Wed Aug 22 19:26:34 2007 -0400 162061 162062 Add _X_EXPORT to exported functions in hw/xfree86/modes/* 162063 162064 Also add missing exports to hw/xfree86/loader/xf86sym.c 162065 162066commit 57f7f2a5327a2d967a726bb4706e4f6b2f4b2cea 162067Author: Eric Anholt <eric@anholt.net> 162068Date: Wed Aug 22 09:02:03 2007 -0700 162069 162070 Revert "Require pixman 0.9.5; Use pixman_image_set_source_clipping() to fix" 162071 162072 The corresponding pixman code hasn't been pushed, so revert until the code is 162073 ready. 162074 162075 This reverts commit 53941c8e68014619d3ded7f8bc0f07d9a38bb9b1. 162076 162077commit d0dc9698ae4324d44ed4c0482d6858d0b73bff33 162078Author: Eric Anholt <eric@anholt.net> 162079Date: Wed Aug 22 09:00:45 2007 -0700 162080 162081 Revert "Fix <pixman.h> include to <pixman/pixman.h>" 162082 162083 The pixman headers have been located under pixman-1/ instead of pixman/ since 162084 around 2007-08-06, and pixman-1.pc has the updated include paths to account 162085 for this. 162086 162087 This reverts commit feb1b3e45513bd6eaa2e6a5ee536183f20d9cb68. 162088 162089commit bc2d516f16d94c805b4dfa8e5b9eef40ff0cbe98 162090Author: Eric Anholt <eric@anholt.net> 162091Date: Fri Aug 17 12:14:16 2007 -0700 162092 162093 Fix overly-restrictive integer overflow check in EXA pixmap creation. 162094 162095 The result was that at 32bpp, pixmaps of width 8192 or greater couldn't be 162096 created, due to treating a pitch value as a width. 162097 162098commit feb1b3e45513bd6eaa2e6a5ee536183f20d9cb68 162099Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 162100Date: Wed Aug 22 16:54:29 2007 +0100 162101 162102 Fix <pixman.h> include to <pixman/pixman.h> 162103 162104commit b6a7c0112c42a3287e53647c38b2c0c5bf8fefa0 162105Merge: 53941c8e6 6ef4ecd82 162106Author: Søren Sandmann Pedersen <sandmann@redhat.com> 162107Date: Tue Aug 21 14:26:34 2007 -0400 162108 162109 Merge branch 'master' of ssh+git://sandmann@git.freedesktop.org/git/xorg/xserver 162110 162111commit 53941c8e68014619d3ded7f8bc0f07d9a38bb9b1 162112Author: Søren Sandmann Pedersen <sandmann@redhat.com> 162113Date: Tue Aug 21 14:26:14 2007 -0400 162114 162115 Require pixman 0.9.5; Use pixman_image_set_source_clipping() to fix 162116 bug 11620 (reported by Jens Stroebel. 162117 162118commit 6ef4ecd82670c37a354243166750d76a97959c8b 162119Author: Julien Cristau <jcristau@debian.org> 162120Date: Tue Aug 21 18:17:35 2007 +0200 162121 162122 config: fix default xkb model (pc105, not keyboard) 162123 162124commit 1834cfb4470341aace64a2fa47d04f85dbf98a47 162125Author: Adam Jackson <ajax@benzedrine.nwnk.net> 162126Date: Tue Aug 21 10:44:37 2007 -0400 162127 162128 Fix an error message to not point to @xfree86.org. 162129 162130commit 265a633cf1fcbf497d6916d9e22403dffdde2e07 162131Author: Keith Packard <keithp@koto.keithp.com> 162132Date: Sun Aug 19 20:29:37 2007 -0700 162133 162134 Screen size changing should leave FB alone when X is inactive. 162135 162136 xf86RandR12ScreenSetSize must protect calls to EnableDisableFBAccess with 162137 suitable vtSema checks to avoid invoking driver code while the X server is 162138 inactive. 162139 162140commit 7dc8531548cc9573e28bb04363dcbb3af5864c9a 162141Author: Keith Packard <keithp@koto.keithp.com> 162142Date: Sun Aug 19 20:28:05 2007 -0700 162143 162144 Ref count cursors used in hw/xfree86/modes code. 162145 162146 The multi-crtc cursor code in hw/xfree86/modes holds a reference to the 162147 current cursor. This reference must be correctly ref counted so the cursor 162148 is not freed out from underneath this code. 162149 162150commit 1f6ddae003ec65d6bc567831bf32bf75dfefdd6c 162151Author: Alex Deucher <alex@botch2.com> 162152Date: Tue Aug 21 00:37:33 2007 -0400 162153 162154 add xf86_crtc_clip_video_helper to xf86sym.c 162155 162156commit c839859d1bc35451923a2cbd5dfac4f3ca5eb3f9 162157Author: David Nusinow <dnusinow@debian.org> 162158Date: Mon Aug 20 21:09:27 2007 -0400 162159 162160 Move module defaults from the header to the source file. 162161 162162 This is where they should have been in the first place. All the rest of 162163 the code in the server defines such things in the source files, not the 162164 headers. 162165 162166commit 53c04351c462d2ae307684e50d5960debe1ee557 162167Author: Alex Deucher <alex@botch2.com> 162168Date: Mon Aug 20 19:46:38 2007 -0400 162169 162170 move intel crtc xv clipping helper to the xserver 162171 162172 The code is generic and can be used by any overlay-based card when 162173 adding randr 1.2 support. Tested on radeon. 162174 162175commit a1fe36b772f7edc162ea97368f86588c0fb77148 162176Author: Julien Cristau <jcristau@debian.org> 162177Date: Mon Aug 20 12:57:06 2007 +0200 162178 162179 xfree86: Fix build on Linux/alpha. 162180 162181 A bunch of CFLAGS had gone missing, so the build failed with errors like: 162182 ../../../../../hw/xfree86/os-support/linux/lnx_ev56.c:7:19: error: input.h: No such file or directory 162183 ../../../../../hw/xfree86/os-support/linux/lnx_ev56.c:8:24: error: scrnintstr.h: No such file or directory 162184 162185commit 2c1431a76e7219e3bd14fd7f7888a8bc4fea0f58 162186Author: Peter Hutterer <peter@cs.unisa.edu.au> 162187Date: Mon Aug 20 10:06:13 2007 +0930 162188 162189 dix: ProcChangeActivePointerGrab: make sure variable is initialised. 162190 162191 Thanks to Ben Close for spotting it. 162192 162193commit 65a49f0ca198e0366175367729a101211388b16b 162194Author: Blair Sadewitz <blair.sadewitz@gmail.com> 162195Date: Sun Aug 19 20:29:22 2007 +0200 162196 162197 Autoconfiguration of wsmouse for NetBSD. 162198 162199commit 3c448b0eb67337b56641e09a6d168aad6745e3ef 162200Author: Fredrik Höglund <fredrik@kde.org> 162201Date: Sat Aug 18 19:02:18 2007 +0200 162202 162203 EXA: Fix a couple of logic errors in exaGetPixmapFirstPixel. 162204 162205 The fb pointer would be left uninitialized when exaPixmapIsOffscreen 162206 returned false. When it returned true and the pixmap was damaged, 162207 fb would be initialized from the pixmap's devPrivate.ptr before the 162208 exaDoMigration and exaPrepareAccess calls, at which point 162209 devPrivate.ptr would still be pointing at offscreen memory. 162210 162211commit 23fbd5292d356067e85e1eec4eb4f743532b0503 162212Author: Alan Coopersmith <alan.coopersmith@sun.com> 162213Date: Fri Aug 17 15:29:16 2007 -0700 162214 162215 Actually build Secure RPC authentication support (missed in modularization) 162216 162217commit 6a32a96d8df184c3ace4847beb48fdcb846d2286 162218Author: Aaron Plattner <aplattner@nvidia.com> 162219Date: Thu Aug 16 17:43:29 2007 -0700 162220 162221 stride is in FbBits-sized chunks, but xoff is not. 162222 162223 Fixes corruption problems with composite rendering to redirected windows in 162224 depth 16. 162225 162226commit 32666d77227fcd2c066de16bf3c07366f92b0457 162227Author: Aaron Plattner <aplattner@nvidia.com> 162228Date: Thu Aug 16 14:57:18 2007 -0700 162229 162230 Bug #12015: Use the right offsets in the dst arguments of pixman_blt. 162231 162232commit daee59b1703ac07c2def9e9fecc479e59b93f761 162233Author: Fredrik Höglund <fredrik@kde.org> 162234Date: Wed Aug 15 19:19:11 2007 +0200 162235 162236 EXA: Wrap Trapezoids to prevent excessive migration of the alpha pixmap. 162237 162238 miTrapezoids creates an alpha pixmap and initializes the contents 162239 using PolyFillRect, which causes the pixmap to be moved in for 162240 acceleration. The subsequent call to RasterizeTrapezoid won't be 162241 accelerated by EXA, which causing the pixmap to be moved back out 162242 again. 162243 162244 By wrapping Trapezoids and using ExaCheckPolyFillRect instead of 162245 PolyFillRect to initialize the pixmap, we avoid this roundtrip. 162246 162247commit 860a09cfb8afc0a293c7eb5e01762724eb86847a 162248Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162249Date: Thu Aug 16 16:10:44 2007 -0400 162250 162251 devPrivates rework: Nevermind, can't const due to return value warnings. 162252 162253 This reverts commit 6fd0a0b08de912421718aca17fe34a55ae285ae7. 162254 162255commit 6fd0a0b08de912421718aca17fe34a55ae285ae7 162256Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162257Date: Thu Aug 16 16:01:42 2007 -0400 162258 162259 devPrivates rework: add const qualifier to key type. 162260 162261commit b2b7817497dd5da73d23ec9cc637c563041fc490 162262Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162263Date: Thu Aug 16 15:30:25 2007 -0400 162264 162265 devPrivates rework: use camelcase standard for name of key type. 162266 162267commit 0a994d4f859a4e48d41a90ed9d2a282bb528c555 162268Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162269Date: Thu Aug 16 12:54:35 2007 -0400 162270 162271 xace: add hooks + new access codes: core protocol selection requests 162272 162273commit 3ef2e9e623819c625a92f464fb14f1e5c181df42 162274Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162275Date: Thu Aug 16 12:27:48 2007 -0400 162276 162277 xace: add hooks + new access codes: core protocol pixmap requests 162278 162279commit be536b79f2a364399937314cfa6c88bf8188da9c 162280Author: Jeremy C. Reed <reed@glacier.reedmedia.net> 162281Date: Thu Aug 16 11:23:28 2007 -0500 162282 162283 Update for support on NetBSD and DragonFly. From 162284 Joerg Sonnenberger and pkgsrc. 162285 162286commit 1d4bea6106d7a1c83e1dfe37fad8268589feaa0b 162287Author: Jeremy C. Reed <reed@glacier.reedmedia.net> 162288Date: Thu Aug 16 11:20:12 2007 -0500 162289 162290 Add some more support for DragonFly. From Joerg Sonnenberger 162291 and pkgsrc. 162292 162293commit fe9bc481efb0821134e10760c23993c6a7386450 162294Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162295Date: Thu Aug 16 12:02:59 2007 -0400 162296 162297 xace: add hooks + new access codes: core protocol font requests 162298 162299commit e89301c8790df9fc49de13dd7c7f36e5340c0c31 162300Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162301Date: Thu Aug 16 10:57:49 2007 -0400 162302 162303 xace: add hooks + new access codes: core protocol client requests 162304 162305commit 5bee8db003a5d552ee1d85bb6c40a3cb93bd6b2b 162306Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162307Date: Thu Aug 16 10:44:51 2007 -0400 162308 162309 xace: drop background-none checking hook, add new hook for controlling 162310 access to other clients. 162311 162312commit b82557c9fb60f11fd2696c8fb2ae17b9dfd915ed 162313Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162314Date: Thu Aug 16 10:36:05 2007 -0400 162315 162316 xace: add hooks + new access codes: core protocol screensaver requests 162317 162318commit 568ae737d1d5d476a0bf85659d88910c4e0ef5e0 162319Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162320Date: Wed Aug 15 14:14:45 2007 -0400 162321 162322 xace: add hooks + new access codes: core protocol server requests 162323 162324commit 14c13b8d62eb37cba8a044daffcddec578ba1644 162325Author: Dave Jones <davej@redhat.com> 162326Date: Thu Aug 16 09:46:27 2007 +0200 162327 162328 Kdrive: fix nasty thinko in TslibEnable() 162329 162330commit 026534f945ae5652592a090a9d41375ca37ab618 162331Author: Alan Coopersmith <alan.coopersmith@sun.com> 162332Date: Wed Aug 15 16:47:53 2007 -0700 162333 162334 Update pci.ids to 2007-08-15 snapshot 162335 162336 Remove nvidia ids in extrapci.ids that are now in pci.ids 162337 162338commit 6cef7b9611297cb1d93cefe3890b26b69c87bce2 162339Author: Alan Coopersmith <alan.coopersmith@sun.com> 162340Date: Wed Aug 15 16:44:49 2007 -0700 162341 162342 Correct XErrorDB path and make it configurable (used by DTrace support) 162343 162344commit 3c9553ac2cac7f3a41966def44a50d722d7e645b 162345Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162346Date: Wed Aug 15 14:14:25 2007 -0400 162347 162348 xace: rename hostlist security hook to "server" as this hook will be used 162349 for other types of server access besides just the host list. 162350 162351commit dc84bb3418933297a8c005070902d9a91ed3d18f 162352Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162353Date: Wed Aug 15 14:13:53 2007 -0400 162354 162355 xace: add hooks + new access codes: core protocol cursor requests 162356 162357commit b424e01ec59d9600a02823f1522949325797268c 162358Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162359Date: Tue Aug 14 13:20:42 2007 -0400 162360 162361 xace: add hooks + new access codes: core protocol property requests 162362 162363commit 14d0397cded699378fa3c19f4e61dbab7d3a9b2c 162364Author: Peter Hutterer <peter@cs.unisa.edu.au> 162365Date: Wed Aug 15 22:12:29 2007 +0930 162366 162367 dix: ProcUngrabKeyboard: make sure grab is initialized. 162368 162369 This should fix the random segfaults with ProcUngrabKeyboard. 162370 162371 Thanks to David Nolden for spotting it. 162372 162373commit 0f9e89b4e309e570d7d366489d250ca2143f0ad7 162374Author: Fredrik Höglund <fredrik@kde.org> 162375Date: Tue Aug 14 22:47:49 2007 +0200 162376 162377 Fix the value comparisons in the IDLETIME wakeup handler. 162378 162379 LessThan/GreaterThan comparisons were used in the wakeup handler, 162380 and LessOrEqual/GreaterOrEqual in the block handler. 162381 162382 Change it to use LessOrEqual/GreaterOrEqual in both functions, 162383 since this is what XSyncNegativeComparison and 162384 XSyncPositiveComparison imply. 162385 162386commit 42d6112ec21949a336ee8b34469f2695273ee2d6 162387Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162388Date: Tue Aug 14 13:09:38 2007 -0400 162389 162390 xace: add hooks + new access codes: core protocol GC requests 162391 162392commit 9a183d7ba50e31afa133cc03aee7991517a283ea 162393Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162394Date: Tue Aug 14 11:39:26 2007 -0400 162395 162396 dix: remove caching of drawables and graphics contexts. The security checks 162397 simply bypass the cached values so they are unused. 162398 162399commit 2763056ab5ae31bed422a0948198d98c6ace6d55 162400Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162401Date: Mon Aug 13 13:40:47 2007 -0400 162402 162403 xace: add hooks + new access codes: core protocol window requests 162404 162405commit 6a195e816b9d60f728d77cc1c23538e7af00a879 162406Author: Kristian Høgsberg <krh@redhat.com> 162407Date: Mon Aug 13 10:43:48 2007 -0400 162408 162409 Revert "Implement damage tracking for AIGLX." 162410 162411 This reverts commit 2243b30e54df07892f75e3d65b687abe5b183cf3. The existing 162412 DRI interface doesn't let us get from a __DRIdrawable to the corresponding 162413 X drawable, and thus, we can't implement AIGLX damage tracking with the 162414 current interface. 162415 162416commit f367285fd5825e0adc271982a529c9904ad65c89 162417Merge: b1272eefd ff4bd3add 162418Author: Peter Hutterer <peter@cs.unisa.edu.au> 162419Date: Sun Aug 12 15:31:10 2007 +0930 162420 162421 Merge branch 'master' into mpx 162422 162423 Conflicts: 162424 162425 Xi/exevents.c 162426 dix/devices.c 162427 dix/getevents.c 162428 include/dix.h 162429 mi/mieq.c 162430 162431commit 03f9da672466b9ab9a9814d784b8c44f1030587e 162432Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 162433Date: Sun Aug 12 03:07:04 2007 +0200 162434 162435 xfree86: Improve default mouse handling on the Hurd 162436 162437 Make /dev/mouse the default device. This makes Xorg works with empty 162438 or missing InputDevice sections. 162439 162440commit c5741438a3a171f493e9da32a6b39f73403f6993 162441Author: Alan Coopersmith <alan.coopersmith@sun.com> 162442Date: Fri Aug 10 16:13:55 2007 -0700 162443 162444 Only use evdev drivers in Xephyr #ifdef linux 162445 162446commit 59961e47df4ea621a6713a8c7d060555f8746c3a 162447Author: Alan Coopersmith <alan.coopersmith@sun.com> 162448Date: Fri Aug 10 16:08:41 2007 -0700 162449 162450 xorgcfg needs PIXMAN_CFLAGS in order to build libc_wrapper.c 162451 162452commit 2243b30e54df07892f75e3d65b687abe5b183cf3 162453Author: Kristian Høgsberg <krh@redhat.com> 162454Date: Fri Aug 10 15:53:05 2007 -0400 162455 162456 Implement damage tracking for AIGLX. 162457 162458commit ff4bd3addb48df3eacc4b121cc249a7f38eb981a 162459Author: Eric Anholt <eric@anholt.net> 162460Date: Wed Aug 8 14:24:42 2007 -0700 162461 162462 Fix the swapped decode of the EDID DTD h/v sync polarity fields. 162463 162464 As a result, we can remove the quirks that existed to flip the bits back around 162465 for us. This is not confirmed in all cases due to lack of bugs containing EDID 162466 blocks associated with the quirks, but is likely true. 162467 162468commit 2926cf1da7e4ed63573bfaecdd7e19beb3057d9b 162469Author: Gustavo Pichorim Boiko <boiko@mandriva.com> 162470Date: Thu Aug 2 18:09:52 2007 -0300 162471 162472 [PATCH] Allocate the right number of entries for saving crtcs 162473 162474commit b2dcfbca2441ca8c561f86a78a76ab59ecbb40e4 162475Author: Keith Packard <keithp@koto.keithp.com> 162476Date: Wed Aug 8 12:16:12 2007 -0700 162477 162478 RRScanOldConfig cannot use RRFirstOutput before output is configured. 162479 162480 RRFirstOutput returns the first active output, which won't be set until 162481 after RRScanOldConfig is finished running. Instead, just use the first 162482 output (which is the only output present with an old driver, after all). 162483 162484commit b1272eefd9a3e340d65c14903f337747ec82d021 162485Author: Peter Hutterer <peter@cs.unisa.edu.au> 162486Date: Wed Aug 8 15:00:02 2007 +0930 162487 162488 Fix typo from last commit. Oh well. 162489 162490commit c02128532e910e813fba94983733942d30c2d5cb 162491Author: Peter Hutterer <peter@cs.unisa.edu.au> 162492Date: Wed Aug 8 13:10:00 2007 +0930 162493 162494 dix: Allow flexible devices for passive core grabs. 162495 162496 A passive core grab doesn't specify the device, and is thus created with the 162497 ClientPointer as device. When this grab is activated later, don't actually 162498 activate the grab on the grab device, but rather change the device to the one 162499 that caused the grab to activate. Same procedure for keyboards. 162500 162501 Makes core apps _A LOT_ more useable and reduces the need to set the 162502 ClientPointer. 162503 162504 Only applies to core grabs! 162505 162506commit ab3f601149e15789edfb7c9a0c33387070279582 162507Author: Tiago Vignatti <tiagov@balalaika.(none)> 162508Date: Tue Aug 7 23:17:32 2007 -0300 162509 162510 Updates some piece of the dead mouse evdev code under the new hotplug scheme. 162511 I exported the evdev driver to Xephyr server. I'm running it using something 162512 like: 162513 $ ./hw/kdrive/ephyr/Xephyr :1 -mouse evdev,,device=/dev/input/event4 -keybd \ 162514 evdev,,device=/dev/input/event1,xkbmodel=abnt2,xkblayout=br 162515 162516 It also closes /#5668. 162517 162518commit 7d1a749b210ba5b9f8d0e5a1feb9a9ef9fa4d992 162519Author: Tiago Vignatti <tiagov@balalaika.(none)> 162520Date: Tue Aug 7 22:59:12 2007 -0300 162521 162522 Export device path key options to be called by the command line of server. 162523 162524commit aee3588a4a6829326770c84b860061f47f2cbcae 162525Author: Tiago Vignatti <tiagov@balalaika.(none)> 162526Date: Tue Aug 7 22:49:07 2007 -0300 162527 162528 Update KdUseMsg() for completeness. 162529 162530commit 30259d5a4e95ff20b30807e5e207ab5995a3fdaf 162531Author: Daniel Stone <daniel@fooishbar.org> 162532Date: Tue Aug 7 20:58:49 2007 +0300 162533 162534 Hotplug: HAL: Fix error handling 162535 162536 Don't use our DBusError for property getting, because we simply don't care: 162537 this fixes D-Bus error spew to stderr. Thanks Michel Dänzer for debugging 162538 and testing. 162539 162540commit aef255425a3521d66c3405d34f7787628a22703e 162541Author: Daniel Stone <daniel@fooishbar.org> 162542Date: Tue Aug 7 16:37:42 2007 +0300 162543 162544 Config: HAL: Use input.xkb namespace 162545 162546 Use an explicit input.xkb.foo namespace, not input.xkb_foo. 162547 162548commit 838e59c02ec06446fc180fb9d86fa8793c7b9903 162549Author: Daniel Stone <daniel@fooishbar.org> 162550Date: Mon Aug 6 16:07:20 2007 +0300 162551 162552 configure.ac: Add $CONFIG_LIB to server libraries 162553 162554 Make sure all DDXes get $CONFIG_LIB. Build-tested with Xvfb and Xdmx. 162555 162556commit b4193a2eee80895c5641e77488df0e72a73a3d99 162557Author: Keith Packard <keithp@koto.keithp.com> 162558Date: Tue Aug 7 12:45:53 2007 -0700 162559 162560 RRScanOldConfig wasn't getting crtcs set correctly 162561 162562 The output crtc is set by RRCrtcNotify, which is called at the end of 162563 RRScanOldConfig. Several uses of output->crtc in this function were wrong. 162564 162565commit 2b93cbb5f8bac9b1b75f723baaa728430b5fefff 162566Author: Keith Packard <keithp@koto.keithp.com> 162567Date: Tue Aug 7 12:44:19 2007 -0700 162568 162569 Decrement mode count when removing RandR output mode. 162570 162571 Removing an output mode without decrementing the mode count scrambles the 162572 output mode array badly. 162573 162574commit fef4c7a6f1a1ef34233b36137bb66d9a657307fb 162575Author: Eric Anholt <eric@anholt.net> 162576Date: Tue Aug 7 09:01:14 2007 -0700 162577 162578 Fix driver build after pixman changes. 162579 162580commit 1339e57485db5a285cfbecbe0bba7154458680ad 162581Author: Tiago Vignatti <tiagov@balalaika.(none)> 162582Date: Tue Aug 7 04:24:34 2007 -0300 162583 162584 Fix typo. 162585 162586commit d9ee5f3e3a3a814ebcd257736c305b41139cc354 162587Author: Tiago Vignatti <tiagov@balalaika.(none)> 162588Date: Tue Aug 7 04:22:26 2007 -0300 162589 162590 Clean a little bit the code. 162591 162592commit 7a5eb3e96b74daaaeb6babf46b13d698280aa3f6 162593Author: Tiago Vignatti <tiagov@balalaika.(none)> 162594Date: Tue Aug 7 02:16:44 2007 -0300 162595 162596 Let xkb options be passed through command line in kdrive servers. I start my 162597 Xephyr using something like: 162598 162599 ./hw/kdrive/ephyr/Xephyr :1 -fp /usr/share/fonts/X11/misc/ -mouse ephyr -keybd ephyr,,xkblayout=br,xkbmodel=abnt2 162600 162601commit 955d5f6c0d14fae63bfe7c4ab39ee0a708919479 162602Author: Tiago Vignatti <tiagov@balalaika.(none)> 162603Date: Tue Aug 7 01:39:29 2007 -0300 162604 162605 When we call Xephyr with '-pointer' a new pointer is added inside the server 162606 and the Xephyr virtual mouse keeps alive. With this patch the semantic changes 162607 turning '-pointer' && 'Xephyr virtual mouse' always false. 162608 162609 Now we can open a device pointer and pass its options in Xephyr's command line 162610 without having other pointer unused. 162611 162612commit b8abeaf74ee8296d4bc3164a5a253624f984a6d4 162613Author: Peter Hutterer <peter@cs.unisa.edu.au> 162614Date: Tue Aug 7 12:32:46 2007 +0930 162615 162616 dix: get the paired keyboard for a passive grab (ProcGrabButton). 162617 162618 Taking the VCK is only correct if no physical device is connected, and even 162619 then it's not really a good idea. 162620 162621commit a0b87f87fb8753955505958bf3d438eef191302d 162622Author: Peter Hutterer <peter@cs.unisa.edu.au> 162623Date: Tue Aug 7 10:49:33 2007 +0930 162624 162625 dix: check for core event to determine if grab is a core grab (CreateGrab). 162626 162627 Checking for VCP/VCK is simply not a safe way to check if a grab is a core grab. 162628 162629commit aa3c6aaaab213200591d29ddb2921adfb87ee5b4 162630Author: Søren Sandmann Pedersen <sandmann@redhat.com> 162631Date: Mon Aug 6 19:00:59 2007 -0400 162632 162633 Require pixman-1 0.9.4, update pixman includes to new scheme 162634 162635commit 74feba4d77d74979a0ea478d666439ffc55001e5 162636Author: Aaron Plattner <aplattner@nvidia.com> 162637Date: Wed Aug 1 14:30:03 2007 -0700 162638 162639 Don't unwrap too early in libwfb for Composite. 162640 162641 Don't call fbFinishWrap until the pixman_image_t that stores the pointer is 162642 actually freed. This prevents corruption or crashes caused by accessing a 162643 wrapped pointer after the wrapping is torn down. 162644 162645commit f6aa2200f2fb4f4d4bb51e67d68e86aabcac0c4b 162646Author: Roland "Test-tools" Bär <roland@verifysoft.de> 162647Date: Mon Aug 6 12:37:52 2007 -0700 162648 162649 Probable off by one buffer overflow in .../xorgconfig/xorgconfig.c 162650 162651 X.Org Bug #11858 <http://bugs.freedesktop.org/show_bug.cgi?id=11858> 162652 Patch #11005 <http://bugs.freedesktop.org/attachment.cgi?id=11005> 162653 162654commit d744df32a15103aa14237175f506350d25b2fec0 162655Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162656Date: Mon Aug 6 12:23:21 2007 -0400 162657 162658 xace: add hooks + new access codes: core protocol colormap requests 162659 162660commit acc9a42c926a3f84159780de12ecc1dc6186068a 162661Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162662Date: Mon Aug 6 12:16:59 2007 -0400 162663 162664 Temporarily disable Security and SELinux extensions 162665 while changes to XACE are being made. 162666 162667commit 102df4f9bac59d95963572d1a7f31d1a064ca4ca 162668Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162669Date: Mon Aug 6 09:16:30 2007 -0400 162670 162671 xace: drop site-policy and declare-extension-security hooks, add 2 new hooks 162672 for controlling access to screens and screen savers. 162673 162674commit 9eddede039f6cbcc323b7e3e4e841c43d3ed4f43 162675Author: Peter Hutterer <peter@cs.unisa.edu.au> 162676Date: Mon Jul 16 18:01:00 2007 +0930 162677 162678 mieqEnqueue: Don't try to update the time for GenericEvents. 162679 162680 Doing so may overwrite the event's length field and cause havoc. Also check if 162681 realloc'd memory did actually return valid pointer. 162682 162683commit e717eb82dc2e55f852919312d04f5cfc8ee55bc8 162684Author: Dave Airlie <airlied@redhat.com> 162685Date: Thu Aug 2 10:50:01 2007 +1000 162686 162687 xserver: stop bcopy from going really slow 162688 162689 The outport is most likely unnecessary on any currently used hardware, 162690 the byte copy is necessary from what I know on IA64 and friends so leave it. 162691 162692 Add a new API entry point which lets a driver select the old behaviour if 162693 such a needs is ever found. 162694 162695 This gives me ~20% speed up on startup on 945 hardware. 162696 162697commit 600ef07113caa7a901c7d486bc8ebd1ae47f885c 162698Author: Tiago Vignatti <tiagov@balalaika.(none)> 162699Date: Fri Aug 3 15:33:41 2007 -0300 162700 162701 Fix kdrive command line parser. 162702 162703commit 375864cb74cced40ae688078b1f7750998972535 162704Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162705Date: Fri Aug 3 13:23:34 2007 -0400 162706 162707 security: drop support for XC-QUERY-SECURITY authorization method. 162708 162709commit d445d2f22b5c97fa010370f4ba9cb0555df4a853 162710Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162711Date: Fri Aug 3 10:56:18 2007 -0400 162712 162713 security: drop the "declare extension security" dix call. Use the 162714 SecurityPolicy configuration file instead. 162715 162716commit 0a71e1542a07abc5e32501973a7cf6de3f641317 162717Author: Carl Worth <cworth@cworth.org> 162718Date: Thu Aug 2 22:48:32 2007 -0700 162719 162720 Create a Picture as well as a Pixmap at the time of AllocateGlyph 162721 162722 This avoids some inefficiency in creating a temporary Picture 162723 for every glyph at rendering time. My measurements with an i965 162724 showed the previous patch causing a 10-15% slowdown for NoAccel 162725 and XAA cases, (while providing an 18% speedup for EXA). 162726 162727 With this change, the NoAccel and XAA performance regression is 162728 eliminated, and the overall EXA speedup, (before any of the 162729 glyphs-as-pixmaps work), is now 32%. 162730 162731commit a2af34d5a861982a03afad8e586bb0181b72bbd0 162732Author: Carl Worth <cworth@cworth.org> 162733Date: Wed Aug 1 15:48:30 2007 -0700 162734 162735 Use per-screen Pixmaps for glyphs 162736 162737 Instead of system-memory data which prevents accelerated 162738 compositing of glyphs, (at least without forcing an upload 162739 of the glyph data before compositing). 162740 162741commit 19b3b1fd8feb343a690331cafe88ef10b34b9d98 162742Author: Carl Worth <cworth@cworth.org> 162743Date: Tue Jul 31 17:04:13 2007 -0700 162744 162745 Use strong hash (SHA1) for glyphs 162746 162747 Using a cryptographically strong hash means that comparing the 162748 hash alone is sufficient for determining glyph equality (no need 162749 to compare the glyph bits directly). This will allow us to replace 162750 system-memory copies of the glyph bits, (which we've only been 162751 holding onto for comparisons), with Pixmaps. 162752 162753commit 516b96387b0e57b524a37a96da22dbeeeb041712 162754Author: Carl Worth <cworth@cworth.org> 162755Date: Mon Jul 30 17:31:47 2007 -0700 162756 162757 ProcRenderAddGlyphs: Avoid allocating a glyph just to find it cached 162758 162759 This is a cleanup without any real savings (yet). Previously, the 162760 implementation would allocate a new glyph, then (often) find it in 162761 the cache, and immediately discard the allocated object. This 162762 re-organization first uses a new FindGlyphByHash function and only 162763 allocates the glyph if nothing is found. 162764 162765 This isn't a real savings yet, since FindGlyphByHash currently still 162766 does a temporary glyph allocation, but this is expected to be replaced 162767 immediately as we switch to an alternate hashing mechanism (SHA1). 162768 162769commit 4c6abe1c7c8abcf203572bbf86b21d97ea4e756f 162770Author: Carl Worth <cworth@cworth.org> 162771Date: Mon Jul 30 21:43:20 2007 -0700 162772 162773 Split HashGlyph functionality out into HashGlyphInfoAndBits 162774 162775 This is in preparation for a future change that will take advantage 162776 of being able to compute a hash for a separate xGlyphInfo and chunk 162777 of bits without a combined Glyph object. 162778 162779commit 363d764ea32b938f3dff35df7cf3370363c04d5c 162780Author: Carl Worth <cworth@cworth.org> 162781Date: Mon Jul 30 15:10:11 2007 -0700 162782 162783 ProcRenderAddGlyphs: Take advantage of the for loops to simplify the code a bit 162784 162785commit dc8a528cd6b9a4da3e60fa31428c37f5b34a897f 162786Author: Carl Worth <cworth@cworth.org> 162787Date: Wed Jul 25 14:57:13 2007 -0700 162788 162789 ProcRenderAddGlyphs: Convert while loops to for loops where more natural 162790 162791commit e34fcd2bf42dbd72ab6ce2df80f2dcaa13416e74 162792Merge: 32c0dcc8c f3955c0a0 162793Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 162794Date: Thu Aug 2 14:27:03 2007 -0400 162795 162796 Merge branch 'master' into XACE-SELINUX 162797 162798 Conflicts: 162799 162800 dix/devices.c 162801 dix/property.c 162802 include/dix.h 162803 162804commit f3955c0a020b39021050cd33c20a17f14fc4b579 162805Author: Arkadiusz Miskiewicz <arekm@maven.pl> 162806Date: Wed Aug 1 21:04:22 2007 +0300 162807 162808 XFree86: xf1bpp: distclean generated files as well 162809 162810 Make sure we clean up after ourselves: not sure why distcheck didn't flag 162811 this one. 162812 162813commit a04c95f4446e5c169dea71019321d790ab4fa139 162814Author: Julien Cristau <jcristau@debian.org> 162815Date: Wed Aug 1 20:37:05 2007 +0300 162816 162817 configure.ac: Fix argument quoting for argv[] 162818 162819 m4 quoting. Yar boo sux. 162820 162821commit 99a88826e5e8cfa25c5f8a88c12799d33114729c 162822Author: Daniel Stone <daniel@fooishbar.org> 162823Date: Wed Aug 1 20:34:58 2007 +0300 162824 162825 configure.ac: Actually use -lrt in monotonic clock test 162826 162827 If we need -lrt to use clock_gettime, then make sure we link with it. 162828 162829commit 1c80e04f876e9254b93ef87eadfcff71234340c6 162830Author: Daniel Stone <daniel@fooishbar.org> 162831Date: Wed Aug 1 20:08:31 2007 +0300 162832 162833 configure.ac: Disable D-Bus config API support by default 162834 162835 This is problematic, so don't even bother with it unless someone wants it. 162836 respeclaration is dead, long live HAL. 162837 162838commit c46663367329615bd2c9b63e93c9534036e5a2ae 162839Author: Michel Dänzer <michel@tungstengraphics.com> 162840Date: Wed Aug 1 18:32:09 2007 +0200 162841 162842 GLX/DRI: Remove some unused variables. 162843 162844commit 17cb4f64e3c39725e83b1e311c09422d7e1c0e52 162845Author: Michel Dänzer <michel@tungstengraphics.com> 162846Date: Wed Aug 1 18:13:18 2007 +0200 162847 162848 GLX_EXT_texture_from_pixmap: Use client provided texture target when available. 162849 162850 This prevents situations where the server doesn't use the target the 162851 client thinks it does, usually resulting in the texture being sampled as all 162852 white. 162853 162854commit a4197db9504adae6af005b2218eee36b8af0d98b 162855Author: Daniel Stone <daniel@fooishbar.org> 162856Date: Wed Aug 1 14:04:51 2007 +0300 162857 162858 GL: GLX: Make sure glxbyteorder.h is distributed 162859 162860commit ad7421fc764e2b82e20d90f12225a03a1d636f18 162861Author: Daniel Stone <daniel@fooishbar.org> 162862Date: Wed Aug 1 08:30:00 2007 +0300 162863 162864 Bump version to 1.3.99.1 for development 162865 162866 This is not actually .1, just bumping for a different devel version. 162867 162868commit 43e71a54502d9ab28ece7f6296d1416d60948dad 162869Author: Daniel Stone <daniel@fooishbar.org> 162870Date: Wed Aug 1 08:16:35 2007 +0300 162871 162872 XFree86: xf1bpp: Fix previous build system commit 162873 162874 Amateur error. 162875 162876commit 6d6bc93b0a13c5356544561e326d4aedf33e61c2 162877Author: Daniel Stone <daniel@fooishbar.org> 162878Date: Wed Aug 1 08:11:22 2007 +0300 162879 162880 Build system: Add missing files 162881 162882 A couple of headers weren't added to the build. 162883 162884commit 505ec436af3a173e0ba32c6f14b4cf9837a553eb 162885Author: Daniel Stone <daniel@fooishbar.org> 162886Date: Wed Aug 1 08:11:08 2007 +0300 162887 162888 XFree86: Properly clean up after ourselves 162889 162890 CLEAN is not a useful variable. CLEANFILES/DISTCLEANFILES, on the other hand, 162891 are useful variables. 162892 162893commit 1ace9770fed4a2ba354ff06a96189428beb36088 162894Author: Daniel Stone <daniel@fooishbar.org> 162895Date: Wed Aug 1 08:10:38 2007 +0300 162896 162897 Build system: Non-dtrace distcheck hacks 162898 162899 automake 1.10 really wants foo.c for foo.O, so give it some dummy files to 162900 deal with if it really needs them. 162901 162902commit cacbdf18ee771d43228c2e96e8ef9a32251ceb55 162903Author: Daniel Drake <d.drake@mmm.com> 162904Date: Wed Aug 1 08:08:37 2007 +0300 162905 162906 Remove duplicated licenses 162907 162908 Some files had two copies of the same license. 162909 162910commit bd49332e4772bd57ffb76c829f0e4770ab876057 162911Author: Daniel Drake <d.drake@mmm.com> 162912Date: Wed Aug 1 08:07:08 2007 +0300 162913 162914 Add proper COPYING file 162915 162916 I went through the entire xorg-server distribution and aggregated all 162917 the licenses I could find (except the questionable GPL files, see my 162918 last mail). 162919 162920 There are many many permutations on essentially the same license terms, 162921 but I have been pedantic and treated slight differences as separate 162922 licenses. 162923 162924 Here is a description of the process I used: 162925 162926 tar xvjf /usr/portage/distfiles/xorg-server-1.1.1.tar.bz2 162927 162928 cd xorg-server-1.1.1 162929 find -name '*.c' -o -name '*.h' | xargs gvim 162930 162931 egrep -Rli "permission|copyright" * | grep -v "\.[ch]" \ 162932 | grep -v "\.in$" | xargs gvim 162933 162934 cd .. 162935 tar xvjf /usr/portage/distfiles/xorg-server-1.3.0.0.tar.bz2 162936 diff -urNp xorg-server-1.1.1 xorg-server-1.3.0.0 162937 162938 git clone git://anongit.freedesktop.org/git/xorg/xserver 162939 cd xserver 162940 git diff xorg-server-1.3.0.0.. 162941 162942 For each file, licenses have been aggregated as follows: 162943 162944 If 2 files have identical license text but different copyright notices, 162945 the copyright notices are aggregated and the license text 162946 is included only once. 162947 162948 Note that by identical I mean really identical, i.e.: 162949 'AUTHOR(S)' is not the same as 'AUTHORS' 162950 'KEITH PACKARD DISCLAIMS' is not the same as 'KEITH PACKARD AND COMPAQ 162951 DISCLAIM' 162952 162953 Otherwise, licenses and accompanying copyright notices have been 162954 stacked. 162955 162956 When going through the changes from 1.1.1 to 1.3.0.0 then HEAD, licenses 162957 have been added and removed (so I have reflected this since the original 162958 version of my COPYING file). It's slightly concerning to see that even 162959 between 1.3.0.0 and HEAD, new license permutations are being added. I'd 162960 suggest that a primary license be chosen and this would be indicated at 162961 the top of this COPYING file. 162962 162963commit 51b735394f0aa9f953f9c320617c7a56028ec458 162964Author: Daniel Drake <dsd@gentoo.org> 162965Date: Mon Apr 30 11:37:46 2007 -0400 162966 162967 [PATCH] xserver: Add COPYING terms 162968 162969 I went through the entire xorg-server distribution and aggregated all 162970 the licenses I could find (except the questionable GPL files, see my 162971 last mail). 162972 162973 There are many many permutations on essentially the same license terms, 162974 but I have been pedantic and treated slight differences as separate 162975 licenses. 162976 162977 Here is a description of the process I used: 162978 162979 tar xvjf /usr/portage/distfiles/xorg-server-1.1.1.tar.bz2 162980 162981 cd xorg-server-1.1.1 162982 find -name '*.c' -o -name '*.h' | xargs gvim 162983 162984 egrep -Rli "permission|copyright" * | grep -v "\.[ch]" \ 162985 | grep -v "\.in$" | xargs gvim 162986 162987 cd .. 162988 tar xvjf /usr/portage/distfiles/xorg-server-1.3.0.0.tar.bz2 162989 162990commit 7fa58385724fa7f441107a1793b601ba3dcb1f4c 162991Author: Arkadiusz Miskiewicz <arekm@maven.pl> 162992Date: Wed Aug 1 08:01:28 2007 +0300 162993 162994 XFree86: xf1bpp: Fix parallel build 162995 162996 One of the constructs wasn't parallel-build safe: fix that. 162997 162998commit 18ab4d559409d4b682aab99fb75f8d861122eab6 162999Author: Daniel Stone <daniel@fooishbar.org> 163000Date: Wed Aug 1 07:27:53 2007 +0300 163001 163002 Darwin: Remove missing file 163003 163004 Xserver.m is missing and still hasn't been added, so just remove it for now. 163005 163006commit 0bd6fe7401b2524cf34793c0b0c642e3d32fae00 163007Author: Daniel Stone <daniel@fooishbar.org> 163008Date: Wed Aug 1 07:27:30 2007 +0300 163009 163010 Config: Add missing include 163011 163012commit 48b3034d13bbbb69072eb11f4579389cc32b0850 163013Author: Daniel Stone <daniel@fooishbar.org> 163014Date: Wed Aug 1 07:01:51 2007 +0300 163015 163016 Config: Add current FDI file 163017 163018 Add the FDI file we're using at the moment, until it gets into upstream HAL. 163019 163020commit 82b720cf3e09d8a6adcd40b25c4d48b34ba1ae80 163021Author: Daniel Stone <daniel@fooishbar.org> 163022Date: Wed Aug 1 06:57:11 2007 +0300 163023 163024 Config: Fix merge detritus 163025 163026commit 6b055e5d9751e3679ff98065e43225ec8a960053 163027Author: Daniel Stone <daniel@fooishbar.org> 163028Date: Wed Aug 1 06:55:36 2007 +0300 163029 163030 Input: Fix stuck modifiers (bug #11683) 163031 163032 Disclaimer: It's 6:51am. I'm trying to be as understandable as possible. 163033 163034 What was happening previously was this: 163035 * Press Alt 163036 * Extended event generated and processed: state is now Alt down once 163037 * Core event generated 163038 - keyboard switched: inherited state is Alt down once 163039 - event processed: Alt down twice 163040 * Release Alt 163041 * Extended event generated and processed: state is now null 163042 * Core event generated and processed: Alt down once 163043 163044 If we switch the order: 163045 * Press Alt 163046 * Core event generated: 163047 - keyboard switched: inherited state is null 163048 - event processed: Alt down once 163049 * Extended event generated and processed: state is now Alt down once 163050 * Release Alt 163051 * Core event generated and processed: state is now null 163052 * Extended event generated and processed: state is now null 163053 163054 When we carry over the previous state, it needs to be the _previous_ state 163055 (state and modifiersPerKey), assuming that we're going to catch now-core 163056 events for any of these. For example, if Ctrl is held down as we pivot, we 163057 need to carry Ctrl over with a count of one, for which an extended + core 163058 release will then clear. Carrying over the union of the previous state _and 163059 the state resulting from the immediate action_ was what broke things. 163060 163061commit 0e0174d45ecbeb7b6dddc4af53da9d6211038e0e 163062Author: Daniel Stone <daniel@fooishbar.org> 163063Date: Wed Aug 1 03:30:07 2007 +0300 163064 163065 XFree86: Allow disabling of HAL 163066 163067 If NoAutoAddDevices is given as a server flag, then no devices will be added 163068 from HAL events at all. If NoAutoEnableDevices is given, then the devices will 163069 be added (and the DevicePresenceNotify sent), but not enabled, thus leaving 163070 policy up to the client. 163071 163072commit cd8e99e56ec5d02026e401cc15e0f8d75f2a4727 163073Author: Daniel Stone <daniel@fooishbar.org> 163074Date: Wed Aug 1 03:29:12 2007 +0300 163075 163076 Input: Don't enable devices when we open them 163077 163078 Thanks to Xi's braindead design, it's otherwise impossible to query input 163079 devices without enabling them. Hurrah. 163080 163081commit 0a31db14b7c7c21ef550dbcc73a9f649f3613cbe 163082Author: Daniel Stone <daniel@fooishbar.org> 163083Date: Wed Aug 1 02:54:14 2007 +0300 163084 163085 Config: D-Bus core: Fix hook removal 163086 163087 Make sure we properly initialise the entire hook when adding it, and 163088 bust out when we're done removing. 163089 163090commit 89f628394f7d831f2ba1e45c5884c3983bef6031 163091Author: Daniel Stone <daniel@fooishbar.org> 163092Date: Wed Aug 1 02:08:02 2007 +0300 163093 163094 XFree86: Input: Fix whitespace 163095 163096commit aec0d06469a2fa7440fdd5ee03dc256a68704e77 163097Author: Aaron Plattner <aplattner@nvidia.com> 163098Date: Tue Jul 31 16:33:37 2007 -0700 163099 163100 Fix a crash when rotating the screen. 163101 163102 Remember output->crtc before setting a NULL mode because RRCrtcNotify now sets 163103 output->crtc to NULL. Use the saved crtc to set the new mode. 163104 163105commit a93033b0bc14ed0bb95c680ded26b63cfe5fd1d3 163106Author: Daniel Stone <daniel@fooishbar.org> 163107Date: Wed Aug 1 01:53:08 2007 +0300 163108 163109 XFree86: Module: Bump input version 163110 163111 config_info changes the size (and ordering) of DeviceIntRec, so bump the 163112 input major. 163113 163114commit 1150969b826e2bd6d8345fa245ed499f2e4cf101 163115Author: Daniel Stone <daniel@fooishbar.org> 163116Date: Wed Aug 1 01:52:20 2007 +0300 163117 163118 Convert all my license statements to the standard form 163119 163120 Convert all my license statements to the standard, accepted form: 163121 cf. <20070717142307.GD13478@fooishbar.org> 163122 http://lists.freedesktop.org/archives/xorg/2007-July/026451.html 163123 163124 keithp's license on configure.ac changed with his verbal permission. 163125 163126commit 8658f5d923a69fb55b4cd9e1e84c2d271679f6e2 163127Author: Daniel Stone <daniel@fooishbar.org> 163128Date: Wed Aug 1 01:10:50 2007 +0300 163129 163130 Hotplug: Add HAL support 163131 163132 Add support for HAL-based hotplugging, in which we just get the list of 163133 input devices and properties from HAL. Requires an FDI which is not yet 163134 in mainline HAL. 163135 163136commit aa75b3481724834da2f855d8dd2ff36074bd5706 163137Author: Daniel Stone <daniel@fooishbar.org> 163138Date: Wed Aug 1 01:09:07 2007 +0300 163139 163140 Hotplug: D-Bus: Dispatch harder 163141 163142 Dispatch until we've got nothing left to dispatch, since apparently 163143 dispatching will only ever fire a single message ... 163144 163145commit 4d238c5c67461ed747aa6c021d1532734f4c63fe 163146Author: Daniel Stone <daniel@fooishbar.org> 163147Date: Wed Aug 1 01:08:26 2007 +0300 163148 163149 Input: GetPointerEvents: Deny events from devices without valuators 163150 163151 For some reason, my keyboard has 25 mouse buttons, but zero valuators. This 163152 causes GPE to blow up spectacularly, trying to get (and set) co-ordinates from 163153 devices without valuators. For now, just prevent this from ever happening, 163154 and whack a dirty great FIXME in. 163155 163156commit 7c9e8fd56e1830f7a971187d14877ebbdf35c4b0 163157Author: Daniel Stone <daniel@fooishbar.org> 163158Date: Wed Aug 1 00:19:14 2007 +0300 163159 163160 Input: Allow enabling and disabling of devices 163161 163162 Add DEVICE_ENABLE to KDrive and XFree86 to allow us to enable and disable 163163 devices on the fly. 163164 163165commit 0afeb0241a83796575da827bd81375c99ff10af5 163166Author: Daniel Stone <daniel@fooishbar.org> 163167Date: Sun Jul 8 20:48:57 2007 +0300 163168 163169 DIX: Clean up null root cursor handling 163170 163171 Move the null root cursor handling out of main() and into CreateRootCursor. 163172 163173commit 62ec6d09b3adaea82ff52c8672e6f611c15ec56d 163174Author: Daniel Stone <daniel@fooishbar.org> 163175Date: Sun Jul 8 20:47:28 2007 +0300 163176 163177 dix.h: Remove duplicate ffs() prototype. 163178 163179commit 4d3379d418a781938358e511fd41deb4115a032c 163180Author: Daniel Stone <daniel@fooishbar.org> 163181Date: Sun Jul 8 14:31:35 2007 +0300 163182 163183 Fonts: Fix builtin fonts 163184 163185 Make sure the font path is always 'built-ins' when we use built-in fonts, 163186 rather than having it as a fixed path for a while, then clobbering it 163187 halfway through startup. 163188 163189commit 9ac7e8a559fe6008cafc95e8264680c50e72ba19 163190Author: Daniel Stone <daniel@fooishbar.org> 163191Date: Sun Jul 8 14:30:53 2007 +0300 163192 163193 Hotplug: D-Bus: API version 2 163194 163195 Use uint32s instead of int32s where practical, and add an API version 163196 request. Also, try to return all devices added, not just the first, 163197 and box device arguments. 163198 163199commit 1cdadc2f43d9069572814510d04b1a560c488fcb 163200Author: Daniel Stone <daniel@fooishbar.org> 163201Date: Sun Jul 8 14:28:58 2007 +0300 163202 163203 Hotplug: Separate D-Bus into core and hotplug API components 163204 163205 Break up D-Bus into two components: a D-Bus core that can be used by any 163206 part of the server (for the moment, just the D-Bus hotplug API, and the 163207 forthcoming HAL hotplug API), and the old D-Bus hotplug API. 163208 163209commit 8bfa41e1bf3f588780d7e9f6f900b1fde0570a7e 163210Author: Daniel Stone <daniel@fooishbar.org> 163211Date: Sun Jul 8 04:29:43 2007 +0300 163212 163213 gitignore: Add automake lex/yacc wrapper 163214 163215commit 06dd2748da8b7af343f6cab409b9f351567de5f3 163216Author: Daniel Stone <daniel@fooishbar.org> 163217Date: Sun Jul 8 00:27:40 2007 +0300 163218 163219 configure.ac: Properly check XFree86 proto modules 163220 163221 Not sure why these are conditionals, anyway. This one really needs 163222 revisiting, but at least causes configure, rather than the compilation, 163223 to bomb out. 163224 163225commit fd10312b4224197b937d9e696b53dc2a16c8912f 163226Author: Daniel Stone <daniel@fooishbar.org> 163227Date: Sun Jul 8 00:26:26 2007 +0300 163228 163229 configure.ac: Fix KDrive VESA/fbdev conditionals 163230 163231 Make sure we actually respect anything explicitly given on the configure 163232 line, instead of just stomping it with what we detect. 163233 163234commit f37612c6f2375ca904411e6caa0be19fa24f032c 163235Author: Nicolas Trangez <eikke@eikke.com> 163236Date: Sun Jul 8 00:23:57 2007 +0300 163237 163238 Hotplug: Remove unused function definition from hotplug.h 163239 163240 configDispatch hasn't been used in a long time. 163241 163242commit 951c058e7800308f7c472e77178c14400f45c1b3 163243Author: Aaron Plattner <aplattner@nvidia.com> 163244Date: Tue Jul 31 14:23:58 2007 -0700 163245 163246 Don't fail compScreenInit if the driver added its own alternate visuals. 163247 163248commit 722d73a0ef54c2ebd8ef38c4a6afa0e7c5aa3e30 163249Author: Dave Airlie <airlied@redhat.com> 163250Date: Tue Jul 31 10:34:56 2007 +1000 163251 163252 Revert "Fix RandR 1.2 conversion of two colour to ARGB cursor on MSB first platforms." 163253 163254 This reverts commit 0f057ebb272f0ee0b51b9ab37d4b07da0924fec4. 163255 163256 This screws my cursor up just starting a bare X server on Intel, 163257 I get the X more like <> than ><.. 163258 163259commit 57b5b97a0710fc043b8a1c01d756cdb73dfe4567 163260Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163261Date: Sun Jul 29 11:02:47 2007 -0400 163262 163263 ReduceCompositeOp returns a Render op, not a boolean. 163264 163265commit f62beb6f3609e8b6e61325ac89017590811bbd07 163266Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163267Date: Fri Jul 27 13:23:15 2007 -0400 163268 163269 Remove all trace of Option "BiosLocation". 163270 163271 This code was deeply dangerous. If anyone actually had a use for this code, 163272 we should find a better way of doing it. 163273 163274commit 486fd4145aed93093d1f1655de40c0a8582bb8b1 163275Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163276Date: Fri Jul 27 13:10:39 2007 -0400 163277 163278 exaGetPixmapFirstPixel: avoid framebuffer readbacks if possible. 163279 163280 If the pixel in framebuffer memory isn't modified since we uploaded it, we 163281 can just read from the system memory copy, wihch avoids both a readback and 163282 an accelerator stall. 163283 163284 In principle this function is still wrong, and all the framebuffer pixel 163285 access should be going through (w)fb so we can get pixel layout corrections. 163286 163287commit 50cb6c7e4419e067c1f080d1de940811d21fc725 163288Author: Kristian Høgsberg <krh@redhat.com> 163289Date: Fri Jun 15 15:29:00 2007 -0400 163290 163291 Don't map the front buffer in libdri if the ddx driver doesn't set the size. 163292 163293 This lets drivers map the front buffer themselves 163294 by setting dontMapFramebuffer. 163295 163296commit cec793ef7a6dac9fa2a6538683e363a72672cde9 163297Author: Aaron Plattner <aplattner@nvidia.com> 163298Date: Thu Jul 26 11:49:46 2007 -0700 163299 163300 Include picturestr.h in xf86Crtc.h to pick up definition of PictTransform. 163301 163302commit 27845fe197b74bf453d99f352e83513e201fdaae 163303Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163304Date: Thu Jul 26 09:32:16 2007 -0400 163305 163306 libconfig shouldn't be an installed library. 163307 163308commit 276f8e2ca42eec982d16b86d67217d68ff98f81d 163309Author: Alan Coopersmith <alan.coopersmith@sun.com> 163310Date: Wed Jul 25 17:42:23 2007 -0700 163311 163312 Include comment/copyright/license for AC_DEFINE_DIR in acinclude.m4 163313 163314commit eba2be448bdd298ff2f7b8603bd9e976da1fdf72 163315Author: Brice Goglin <Brice.Goglin@ens-lyon.org> 163316Date: Wed Jul 25 20:53:45 2007 +0200 163317 163318 Minor fixes in cvt and gtf manpages 163319 163320 Reported by "A. Costa" <agcosta@gis.net> in 163321 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=432065 163322 163323commit 8d230319040f0a7f72231da2bf5ec97dc3612e21 163324Author: Gustavo Pichorim Boiko <boiko@mandriva.com> 163325Date: Tue Jul 24 16:19:19 2007 -0300 163326 163327 Fix the output->crtc initialization in the old randr setup 163328 163329commit 0f057ebb272f0ee0b51b9ab37d4b07da0924fec4 163330Author: Michel Dänzer <michel@tungstengraphics.com> 163331Date: Wed Jul 25 17:04:04 2007 +0200 163332 163333 Fix RandR 1.2 conversion of two colour to ARGB cursor on MSB first platforms. 163334 163335 Doesn't seem necessary to do anything here... 163336 163337commit 5b424b562eee863b11571de4cd0019cd9bc5b379 163338Author: Gustavo Pichorim Boiko <boiko@mandriva.com> 163339Date: Mon Jul 23 18:27:41 2007 -0300 163340 163341 Set the crtc before the output change is notified 163342 163343 Set the new randr crtc of the output before the output change notification is 163344 delivered to the clients. 163345 Remove RROutputSetCrtc as it is not really necessary. All we have to do is set 163346 the output's crtc on RRCrtcNotify 163347 163348commit 7da38bb6a15247948c90e00a59230453fcf13cbd 163349Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163350Date: Sat Jul 21 15:27:40 2007 -0400 163351 163352 Partial redundancy elimination in PropertyNotify generation. 163353 163354commit 0f91abd5c68eb044d09733d18ef0f6b8ed128200 163355Author: Julien Cristau <jcristau@debian.org> 163356Date: Thu Jul 19 20:37:26 2007 -0400 163357 163358 Fix alpha build failures 163359 163360 Don't include <asm/pci.h> in os-support/linux/lnx_axp.c, use "lnx.h" and 163361 <unistd.h> instead 163362 163363commit dc9c5196282ba61bd542e198dfe0d53d93181591 163364Author: Keith Packard <keithp@neko.keithp.com> 163365Date: Thu Jul 19 13:28:00 2007 -0700 163366 163367 Make PreferredMode option in config file override EDID mode preferences. 163368 163369 When the PreferredMode option is selected in the config file, remove the 163370 M_T_PREFERRED bit from all other preferred modes to force the config file 163371 mode to be selected. 163372 163373commit 73a93c5a6b68f7ba21f9e75f50b1032603a3b39e 163374Author: Keith Packard <keithp@neko.keithp.com> 163375Date: Thu Jul 19 13:26:36 2007 -0700 163376 163377 Query modes on disabled (but not ignored) outputs. 163378 163379 Code that disabled mode detection on disabled outputs would confuse 163380 applications by listing said outputs as connected but without any modes. 163381 This makes the disabled state in the config file affect only the initial 163382 configuration and not subsequent modifications by RandR. 163383 163384commit 9fc36a391c11170cde1a28f548a2cae5f6f20d5b 163385Author: Keith Packard <keithp@neko.keithp.com> 163386Date: Sat Jul 14 12:36:15 2007 -0700 163387 163388 Make pending property changes trigger mode setting. 163389 163390 The DDX code was ignoring pending properties for computing when mode setting 163391 was required. This meant that configurations differing only in property 163392 values would not cause the mode to be set. 163393 163394commit aed6569309223ecc7e26fa84e4d430e422455607 163395Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163396Date: Sat Jul 14 15:21:46 2007 -0400 163397 163398 Refactor how Composite adds visuals to the screen. 163399 163400 Besides being slightly simpler to read, it's now trivial to add a depth-16 163401 visual to a depth-24 screen just by adding a line for it in the alternate 163402 visual list. Visuals for indexed depths are slightly tricky still. 163403 163404commit 21bbd7d64b5f74915afd7a312e589654442f3461 163405Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163406Date: Tue Feb 6 21:42:50 2007 -0500 163407 163408 Delete some pre-dlloader debugging scaffolding. 163409 163410 If your loader is as bad as elfloader, then it makes sense for the 163411 server to have some stubs for you to assign to / break on. However it 163412 is no longer 1996. 163413 163414commit 1f71f0c0574bafb36da20fec669f9a1138c69a47 163415Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163416Date: Tue Feb 6 21:28:03 2007 -0500 163417 163418 Remove (long-)deprecated xf86EnablePciBusMaster. 163419 163420commit 0a63d874e9c2f4fe4b38839a744461f9d41040b2 163421Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163422Date: Tue Feb 6 21:22:49 2007 -0500 163423 163424 Always normalize the module name. 163425 163426commit 9a1c6afd12caf0143483f72bfbba0c4c3daaa6ff 163427Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163428Date: Tue Feb 6 21:19:50 2007 -0500 163429 163430 Remove dead code for screen crossing. 163431 163432commit 8ca2fe8914af1a67bf597f99025e5cbe9b08da57 163433Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163434Date: Tue Feb 6 21:11:13 2007 -0500 163435 163436 Delete dead module test code. 163437 163438commit e2413cc7cae4e578b8e9b408ea85bef596b03ea3 163439Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163440Date: Tue Feb 6 21:07:37 2007 -0500 163441 163442 Remove MEMDEBUG 163443 163444 This existed (but may not have worked) in the monolith, but is gone now. 163445 163446commit d1d65a84150dfbc3a4dbe108f237a85ab6e09bbb 163447Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163448Date: Tue Feb 6 21:01:08 2007 -0500 163449 163450 Dead ifdefs for BITMAP_SCANLINE_UNIT == 64 163451 163452 This appears to be a legacy of cfb24 not being smart enough to deal with this 163453 case. But since cfb24 unexists, die die die. 163454 163455commit cbe74394a5ed21ed80c0aab6eefd2716122cce11 163456Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163457Date: Tue Feb 6 20:44:34 2007 -0500 163458 163459 Nuke dead X -configure code. 163460 163461commit 5657fb065cc79ba3ca5a836f45637ba9894f9abf 163462Author: Dodji Seketeli <dodji@seketeli.org> 163463Date: Tue Jul 17 12:12:02 2007 +0200 163464 163465 exaDriverInit: Fail if pScreenInfo or a member of it is invalid. 163466 163467 EXA may attempt to use the invalid value and crash otherwise. 163468 163469commit bbe7ce10fa93017374d7a4611427b70a22d7507a 163470Author: Alan Coopersmith <alan.coopersmith@sun.com> 163471Date: Mon Jul 16 17:25:59 2007 -0700 163472 163473 Update pci.ids to 2007-07-16 snapshot 163474 163475 Remove nvidia ids in extrapci.ids that are now in pci.ids 163476 Add nvidia ids to extrapci.ids that are in xf86-video-nv but not pci.ids 163477 163478commit ac979c165128704116cd40086320b6edc79018e2 163479Author: Keith Packard <keithp@neko.keithp.com> 163480Date: Sat Jul 14 12:13:17 2007 -0700 163481 163482 MakeAtom needs length without trailing NUL. sizeof("string") includes NUL. 163483 163484 I made a mistake in some new code using MakeAtom, passing the size of the 163485 string instead of the length of the string. Figuring there might be other 163486 such mistakes, I reviewed the server code and found four bugs of the same 163487 form. 163488 163489commit 393171034c15d8a1b82232b8f9455a358035e932 163490Author: Keith Packard <keithp@neko.keithp.com> 163491Date: Sat Jul 14 09:03:40 2007 -0700 163492 163493 Add RandR reflection support. 163494 163495 Replace the ad-hoc transformation mechanisms with matrices. 163496 Prepares for more general transformation as well. 163497 163498commit 8773ad023eb28950eb0f802d2ca31a67f84adddc 163499Author: Keith Packard <keithp@neko.keithp.com> 163500Date: Sat Jul 14 08:47:50 2007 -0700 163501 163502 Screen size bounds check in ProcRRSetCrtcConfig not masking out reflections. 163503 163504 When checking how to validate the selected mode and position against the 163505 current screen size, the test against 90/270 rotation did not mask out 163506 reflection, so that when reflection was specified, the 90/270 test would 163507 never succeed. This caused incorrect bounds checking and would return 163508 an error to the user instead of rotating the screen. 163509 163510commit 881a620b4d6ea7a54af14c8f8fbe6924c9aa9291 163511Author: Keith Packard <keithp@neko.keithp.com> 163512Date: Sat Jul 14 08:45:10 2007 -0700 163513 163514 When sync'ing logfile, also flush it. 163515 163516 When the logfile is set to sync, the actual sync occurs whenever the log 163517 file is flushed. If the log file is not also set to flush, no syncing 163518 occurs. 163519 163520commit 031b009ea678809bf1ddca883c2082b304c408c9 163521Author: Alan Coopersmith <alan.coopersmith@sun.com> 163522Date: Fri Jul 13 14:54:45 2007 -0700 163523 163524 Use %S instead of %s for strftime seconds when printing build time 163525 163526commit c0e91777a9874fe2cd9a7e9180263f512c1e8f8d 163527Author: Alan Coopersmith <alan.coopersmith@sun.com> 163528Date: Thu Jul 12 16:37:11 2007 -0700 163529 163530 Add __SOL8__ to xorg-server.h.in since xf86-input-kbd needs it to build 163531 163532commit 6b4231e3b5b49b731c9a00930ae465fff8539831 163533Author: Alan Coopersmith <alan.coopersmith@sun.com> 163534Date: Thu Jul 12 16:36:27 2007 -0700 163535 163536 Use kbd driver when xorg.conf specifies "keyboard" or "Keyboard" (bug #11301) 163537 163538 X.Org Bug #11301 <https://bugs.freedesktop.org/show_bug.cgi?id=11301> 163539 Sun Bug #6560332 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6560332> 163540 163541commit 9fcb30ebf7b7b2137955f759e95c1d58c4f27a11 163542Author: Alan Coopersmith <alan.coopersmith@sun.com> 163543Date: Thu Jul 12 13:00:32 2007 -0700 163544 163545 Make SOLARIS_INOUT_ARCH substitutions work better with automake-1.10 163546 163547commit 7c0ca27f6dd0a800dc27429a33dbc8e133f9a9c1 163548Author: Alan Coopersmith <alan.coopersmith@sun.com> 163549Date: Wed Jul 11 17:15:29 2007 -0700 163550 163551 "fbpict.c", line 215: void function cannot return value 163552 163553commit 0a4e9311158ed3ecda0722640f860ace2f87a97e 163554Author: Hanno Boeck <hanno@gentoo.org> 163555Date: Thu Jul 12 10:17:07 2007 +1000 163556 163557 xnest: fix linking since dbus 163558 163559 Fixes bug 8955 163560 163561commit b2f9ca6ac400d426d7a1ef0162f7e7ce28288dd1 163562Author: Keith Packard <keithp@neko.keithp.com> 163563Date: Tue Jul 10 21:33:34 2007 -0700 163564 163565 Redirect fix: Manual + Automatic - Manual = Automatic 163566 163567 A window with redirect manual *and* redirect automatic which loses the 163568 manual redirecting client becomes redirect automatic. 163569 163570commit 561989f2f0fc31e3d3bf8df978a9cb3d4c85af59 163571Author: Keith Packard <keithp@neko.keithp.com> 163572Date: Tue Jul 10 21:06:51 2007 -0700 163573 163574 Generate ChangeLog file for make dist. 163575 163576 Copy Makefile.am snippet which generates a complete git change history to 163577 the ChangeLog file during the distribution generation process. 163578 163579commit e316fa59fea8b7b18cdf3a227890351a9567ec65 163580Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163581Date: Tue Jul 10 14:20:55 2007 -0400 163582 163583 Add per-monitor config file option for maximum pixel clock. 163584 163585commit ffdf8f3e452638e6b050fccabee465d6c0300f45 163586Author: Peter Hutterer <peter@cs.unisa.edu.au> 163587Date: Tue Jul 10 13:56:38 2007 +0930 163588 163589 DeliverEventsToWindow: ensure that genericMask has a defined value. 163590 163591commit 161624a5a45808fd56141dc2c64be729944f03ed 163592Author: Michel Dänzer <michel@tungstengraphics.com> 163593Date: Tue Jul 10 09:02:40 2007 +0200 163594 163595 GLX: Only build code dealing with GLXPixmap damage field when DRI is enabled. 163596 163597commit 4abd00dab7e648dab8172f6009371e4e63d0c521 163598Author: Michel Dänzer <michel@tungstengraphics.com> 163599Date: Tue Jul 10 09:02:08 2007 +0200 163600 163601 Make sure DRI drawables are cleaned up when client dies. 163602 163603 The previous scheme didn't work when the client didn't create the core drawable, 163604 e.g. the root or composite overlay window. Use refcounting via special client 163605 resources to fix that. 163606 163607commit 583e988b9f7cfb9293144c8309023c0dd1766715 163608Author: Peter Hutterer <peter@cs.unisa.edu.au> 163609Date: Tue Jul 10 09:19:02 2007 +0930 163610 163611 Install geext.h, otherwise drivers won't build. 163612 163613commit 9809715afaafee9baf2aef348c1ebda7e8b3f076 163614Author: Peter Hutterer <peter@cs.unisa.edu.au> 163615Date: Mon Jul 9 10:42:03 2007 +0930 163616 163617 Change CheckMotion to ignore non-pointer events but acknowledge XI events. 163618 163619 Call CheckMotion from ProcessOtherEvents() to make sure absolute XI events 163620 update the sprite before an event is sent. 163621 163622commit 62efc3951a96648cf975302aa6651cb67b87fa64 163623Author: Peter Hutterer <peter@cs.unisa.edu.au> 163624Date: Fri Jul 6 17:00:20 2007 +0930 163625 163626 DeliverGrabbedEvent: stop segfault when gemask == NULL. 163627 163628commit 3312e4dd5e055b2cb445b5d5c617aa7a611eedc1 163629Author: Peter Hutterer <peter@cs.unisa.edu.au> 163630Date: Fri Jul 6 16:41:52 2007 +0930 163631 163632 Call CheckMotion for all core events. 163633 163634 We need to do this to update the sprites x/y coordinate before we assemble a 163635 button event. Absolute devices may send a buttonEvent with valuators attached. 163636 If we don't update the sprite before assembling the event, the valuators are 163637 lost and the button press is delivered to the previous position of the device. 163638 Doesn't have any effect on relative devices. 163639 163640commit c1a6841a64576b7e688e9ca0d3e0db8acf52d4ae 163641Author: Paulo Ricardo Zanoni <prz05@c3sl.ufpr.br> 163642Date: Tue Jul 10 10:08:44 2007 +0930 163643 163644 ProcX{Change|Query}WindowAccess: change device list from char* to XID*. 163645 163646commit 5957aa6fdc580ccad4557eeefa0636ffad823f33 163647Author: Michel Dänzer <michel@tungstengraphics.com> 163648Date: Mon Jul 9 08:47:05 2007 +0200 163649 163650 Fix regression from recent composite changes. 163651 163652 One pWin->redirectDraw test was converted incorrectly, causing incorrect 163653 rendering in some cases. 163654 163655commit 5ccc09b18244f91a06b3bea20b02a97280d1a229 163656Author: Peter Hutterer <peter@cs.unisa.edu.au> 163657Date: Fri Jul 6 15:43:08 2007 +0930 163658 163659 Use the same struct for generic event masks throughout the code. 163660 Renaming those structs too. 163661 163662 Previously grabs were using a different struct than windows, which was 163663 reasonably stupid. 163664 163665commit bcb23527421578bd4c9397d4c2c19cbefa22fc59 163666Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163667Date: Thu Jul 5 15:56:25 2007 -0400 163668 163669 Clean up unused #ifdefs from fb. 163670 163671commit 9ff7ff2fda30f334515b16ef0867c1500c41bc0f 163672Author: Keith Packard <keithp@neko.keithp.com> 163673Date: Wed Jul 4 23:38:27 2007 -0700 163674 163675 Fix MEMORY SMASH in XkbCopyKeymap. 163676 163677 XkbCopyKeymap reallocates the destination keymap when it is not large enough 163678 to hold the source data. When reallocating the map->types data, it needs to 163679 zero out the new entries. The computation for where to start bzero'ing was 163680 accounting for the size of the data type twice, once implicitly in the 163681 pointer arithmetic, and once explicitly with '* sizeof (XkbKeyTypeRec)'. 163682 This would often lead to random memory corruption when the destination 163683 keymap had existing map->types data. 163684 163685commit 9131d560a0d42067cc4e726e445e060216c9acdc 163686Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 163687Date: Thu Jul 5 02:47:34 2007 -0300 163688 163689 Postpone options variable assignment to fix segfault when we got a device but 163690 its driver is incorrect. Also if (!ki && !pi) can never be true. 163691 163692 This one also adds the device option field. 163693 163694commit 41b485d5507821e41c3281c3c565647ae7582101 163695Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 163696Date: Thu Jul 5 02:40:07 2007 -0300 163697 163698 kdrive must to know that devices are unplugged. 163699 163700commit a92dc6b5295e4f352115fed2856169929819863f 163701Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 163702Date: Thu Jul 5 02:28:14 2007 -0300 163703 163704 Remove redundant linking in kdrive. Fix configure.ac variable name and clean 163705 it up a little. 163706 163707commit 41b5155c8be75c4e171c0f64616cc09598b8ec54 163708Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 163709Date: Thu Jul 5 01:57:41 2007 -0300 163710 163711 For each kdrive server put a dependencie on its own libraries. 163712 163713commit 16e429bcbf2f62cfc58162ab2857afb7376dda41 163714Author: Jonathan Lim <jlim@sgi.com> 163715Date: Wed Jul 4 20:08:49 2007 +0200 163716 163717 Bug 5000: Fix domain support for SGI Altix 163718 163719commit f106c04b627d9f57b38627971dc79c75129e66d6 163720Author: Keith Packard <keithp@neko.keithp.com> 163721Date: Tue Jul 3 14:47:19 2007 -0700 163722 163723 Have Composite always report server version. 163724 163725 It was reporting the lessor of the server and client versions, which doesn't 163726 make sense with the 0.4 semantic change in clipping. 163727 163728commit 4f88d68bdb90cc7d12170355105b4fd020acd306 163729Author: Keith Packard <keithp@neko.keithp.com> 163730Date: Tue Jul 3 14:43:17 2007 -0700 163731 163732 Force advertised Composite version to 0.4 instead of using header version. 163733 163734 Installed protocol header version may be newer than the server code base. 163735 Use internal version number for Composite extension to make sure the server 163736 doesn't advertise capabilities it doesn't support. 163737 163738commit 866f092ca0160a366add01b48ad03438926c4d16 163739Author: Keith Packard <keithp@neko.keithp.com> 163740Date: Tue Jul 3 14:29:11 2007 -0700 163741 163742 Make Composite manual redirect windows not clip their parent. 163743 163744 This patch changes the semantics of manual redirect windows so that they no 163745 longer affect the clip list of their parent. Doing this means the parent can 163746 draw to the area covered by the child without using IncludeInferiors. More 163747 importantly, this also means that the parent receives expose events when 163748 that region is damaged by other actions. 163749 163750commit 2a75c774975b50dd4e71b7dbea7bd65ca2984a43 163751Author: Dodji Seketeli <dodji@seketeli.org> 163752Date: Tue Jul 3 11:00:29 2007 +0200 163753 163754 ExaOffscreenMarkUsed: Don't crash when there's no offscreen memory. 163755 163756commit 0ede39a25cf5b0b6c2c89677f810c21ce42b95df 163757Author: Michel Dänzer <michel@tungstengraphics.com> 163758Date: Tue Jul 3 10:55:13 2007 +0200 163759 163760 Fix build when int10 doesn't use x86emu. 163761 163762commit 028a00bc518dc6908839e8ce7c50ab1837100945 163763Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163764Date: Mon Jul 2 18:41:55 2007 -0400 163765 163766 Make x86emu's I/O cycle tracing more useful. 163767 163768 Print debug messages only when the appropriate debug bit is set in the 163769 8086 state vector, so you can focus in on the call you're actually 163770 interested in. 163771 163772commit 00e8295b7e0c7c0ba97707903004272818e3d87d 163773Author: Gero Mudersbach <geroxp@web.de> 163774Date: Mon Jul 2 11:40:11 2007 -0700 163775 163776 Bug #10814: Add needed quirk for Samsung 225BW like the 226BW. 163777 163778commit 667e95f2e8389d9f23c50446d6d664eddd16d260 163779Author: Eric Anholt <eric@anholt.net> 163780Date: Mon Jul 2 11:36:11 2007 -0700 163781 163782 Correct the xf86EdidModes.c file description. 163783 163784commit 3de1f0d03b329b01856f664651db23ffefb58646 163785Author: Eric Anholt <eric@anholt.net> 163786Date: Tue May 29 10:08:58 2007 -0700 163787 163788 Fix documentation of association of outputs to monitor sections in xorg.conf(5) 163789 163790commit c1a16bdcfe7aa907fe78f27dc606a8e5a2699952 163791Author: Peter Hutterer <peter@cs.unisa.edu.au> 163792Date: Wed Jun 27 18:08:03 2007 +0930 163793 163794 Add 'evfill' field to GEExtensions. 163795 We need this to allow extensions to fill in extra data for an event before it 163796 is sent to the client. This is probably to be used like 163797 FillUpEventsFromWindow(). 163798 163799commit 4d76075dbb618a47ff9fc15c4be2e2d34210fa8d 163800Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163801Date: Fri Jun 29 14:06:52 2007 -0400 163802 163803 Death to RCS tags. 163804 163805commit 2691c05fd647d9fa10f791ac397ecb9c423a076f 163806Author: Peter Hutterer <peter@cs.unisa.edu.au> 163807Date: Fri Jun 29 11:56:18 2007 +0930 163808 163809 Make sure window->optional is allocated before assigning it. 163810 163811 DeletePassiveGrabFromList() may remove the window optional, so we need to 163812 re-alloc it if it isn't there anymore. 163813 163814 Thanks to Colin Harrison for spotting the bug. 163815 163816commit f7f3fe7fe7233a2ffc43106c48f44cbbd82b7c19 163817Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163818Date: Thu Jun 28 18:59:05 2007 -0400 163819 163820 Remove the remnants of OS/2 support. 163821 163822 This has never worked in any modular server release, and as far as I know 163823 was never tested in 6.7 through 6.9. 163824 163825commit 8a06ff9ffa4816d192e58e43e7fe569b97b4dd7c 163826Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163827Date: Thu Jun 28 16:41:28 2007 -0400 163828 163829 Fix another usage of MAX_PCI_DEVICES. 163830 163831 Fixes cases where the VGA device is above the 128th device on the system. 163832 163833commit 928836a5abd85466e920eb487fab9ccb295e0c5b 163834Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163835Date: Thu Jun 28 16:29:28 2007 -0400 163836 163837 Bug #10770: "Inputdevs" isn't a valid config file keyword. 163838 163839commit 62f43d8b33f67d8f3d0bd65787ffae9e6b634d65 163840Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 163841Date: Wed Jun 27 16:24:42 2007 -0300 163842 163843 Relink properly all kdrive servers when changes happen. 163844 163845commit 3860996d5666b76600b1537e2cbd58e36b086308 163846Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 163847Date: Wed Jun 27 16:06:12 2007 -0300 163848 163849 Remove double-defined. 163850 163851commit 1e189ed1daab58c1de67d387306fde0a9e7984a4 163852Author: Aaron Plattner <aplattner@nvidia.com> 163853Date: Wed Jun 27 10:16:40 2007 -0700 163854 163855 Handle tileStride > 1 in fbEvenTile. 163856 163857 Patch courtesy of James Jones. 163858 163859commit bf1ad1aa4270dccf1540943d97e80b317c0adb56 163860Author: Adam Jackson <ajax@benzedrine.nwnk.net> 163861Date: Wed Jun 27 09:19:28 2007 -0400 163862 163863 Add VBE PanelID support. 163864 163865 Originally found only in the i810 driver. 163866 163867commit d73835efda4995a310188537233a984f4b73628d 163868Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 163869Date: Wed Jun 27 03:34:13 2007 -0300 163870 163871 More janitor work. Remove 'defined but not used' warnings from kdrive and some minor cosmetic. 163872 163873commit 38f8e536684193c3f70b23be22d818053c676072 163874Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 163875Date: Wed Jun 27 03:19:37 2007 -0300 163876 163877 Janitor work. Remove 'defined but not used' warnings from xorg and other cosmetic. 163878 163879commit 1340f34ec98c41781164018d43bd7bb858d8132b 163880Author: Tiago Vignatti <vignatti@c3sl.ufpr.br> 163881Date: Wed Jun 27 03:04:55 2007 -0300 163882 163883 Fix kdrive XKB. 163884 163885commit 08779c62ba2f4cbfa47e3cf738daf8abd74b60cf 163886Author: Peter Hutterer <peter@cs.unisa.edu.au> 163887Date: Wed Jun 27 11:16:55 2007 +0930 163888 163889 Reset grab->next to NULL when we're creating a grab. 163890 163891 This is definitely necessary in CheckDeviceGrabAndHintWindow (thanks to Paulo 163892 Zanoni for spotting it). We're resetting it everywhere else too, just to be 163893 save. 163894 163895commit 49136d3e9fe2d065090ee90c18a49f65fb027c9f 163896Author: Peter Hutterer <peter@cs.unisa.edu.au> 163897Date: Tue Jun 26 19:43:54 2007 +0930 163898 163899 mieqEnqueue: use modulo for queue tail wrapping. 163900 163901 This was previously committed by Michael Daenzer, but was lost during a pull. 163902 163903commit 8396bf7c8b4eb6b8f2d1e1ef19887064c69591da 163904Author: Peter Hutterer <peter@cs.unisa.edu.au> 163905Date: Tue Jun 26 12:58:37 2007 +0930 163906 163907 Init genericMasks to 0, make sure list is NULL terminated. 163908 163909commit 9725516b4274ceaf9d6caf372c5439b4c9db2316 163910Author: Alan Coopersmith <alan.coopersmith@sun.com> 163911Date: Fri Jun 22 17:05:21 2007 -0700 163912 163913 Split checks for dtrace & getpeerucred() 163914 163915commit edb9ccf3ecb4e35a840aa13815979c5fbd73f32d 163916Author: Michel Dänzer <michel@tungstengraphics.com> 163917Date: Mon Jun 25 10:51:38 2007 +0200 163918 163919 Make sure DRIScreenPrivIndex is -1 when no DRI screen private is allocated. 163920 163921 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=11340 . 163922 163923commit c4e850a781e3b85631cb386d24efcca2a835d4c9 163924Author: Peter Hutterer <peter@cs.unisa.edu.au> 163925Date: Mon Jun 25 13:38:25 2007 +0930 163926 163927 Quickfix to stop core key events from doubling up. 163928 163929commit 4c601b904ee6fb01da3343ff9ef00d36f1341fcb 163930Author: Peter Hutterer <peter@cs.unisa.edu.au> 163931Date: Mon Jun 25 10:53:05 2007 +0930 163932 163933 configFiles(): don't return anything when declared as void. 163934 163935commit bec4e47d128ec40b58a2c9aae475f6a6fc4323c3 163936Author: Peter Hutterer <peter@cs.unisa.edu.au> 163937Date: Mon Jun 25 10:51:42 2007 +0930 163938 163939 NULL-terminate device list when synthesizing core devices. 163940 163941 This fix is required for 93ca526892c0d22afa05cce6496198c652043a19 to work. 163942 163943commit 8e5102b9f01821048e72e7f068193a0b3e1816f9 163944Author: Peter Hutterer <peter@cs.unisa.edu.au> 163945Date: Thu Jun 21 15:47:48 2007 +0930 163946 163947 Set the detail field for DeviceKeyEvents to the keycode. 163948 163949 (cherry picked from commit 0c33dc152e372cdc8ae59d9a5696b0774bcd03b7) 163950 163951commit 87564543d92c1ee1f8cb6fb9716a15d693e08cf5 163952Author: Peter Hutterer <peter@cs.unisa.edu.au> 163953Date: Tue Jun 19 18:20:05 2007 +0930 163954 163955 Only decrement buttonsDown when the button count is greater than 0. 163956 163957 Device drivers flush their buttons on device init and cause a button down 163958 event to be generated. If we unconditionally decrease the buttons, we won't be 163959 able to ever get a passive device grab. 163960 163961 Format documentation for CheckDeviceGrabs to make it readable. 163962 (cherry picked from commit 3e894974cdd6a75683d4601f71622d1da7ec4395) 163963 163964 Conflicts: 163965 163966 Xi/exevents.c 163967 163968commit 24ee89fd60f489f2d3af0399e0d667057df74d02 163969Author: Peter Hutterer <peter@cs.unisa.edu.au> 163970Date: Tue Jun 19 15:31:56 2007 +0930 163971 163972 Add a few comments to devices.c 163973 163974commit 93ca526892c0d22afa05cce6496198c652043a19 163975Author: Peter Hutterer <peter@cs.unisa.edu.au> 163976Date: Wed Jun 13 15:28:15 2007 +0930 163977 163978 Split up memory for devices configured in the config file. 163979 163980 If we're using a continuous block here, we segfault when a device removal 163981 triggers an xfree call. 163982 163983commit b141b85c254afff3ce2221d899787fab3dc295bd 163984Author: Peter Hutterer <peter@cs.unisa.edu.au> 163985Date: Wed Jun 13 15:26:03 2007 +0930 163986 163987 Check for identical grabs when adding a new passive grab. If an identical grab 163988 163989 exists, remove the old one and prepend the new one. 163990 163991 X.org Bug 2738 <https://bugs.freedesktop.org/show_bug.cgi?id=2738> 163992 163993commit 19cde59c41cf167cc609debfee75bfc015beac12 163994Author: Søren Sandmann Pedersen <sandmann@redhat.com> 163995Date: Fri Jun 22 00:38:50 2007 -0400 163996 163997 In fbFill() make sure the drawable is validated when pixman_fill() succeeds. 163998 163999 In fbSolidBoxClipped() don't return when pixman_fill() succeeds. 164000 164001commit d2177c80915f2fe2e8a5c948d4ba2fa51dbfbea2 164002Author: Keith Packard <keithp@neko.keithp.com> 164003Date: Fri Jun 22 02:08:21 2007 +0100 164004 164005 Skip driver mode detection/configuration when !vtSema. 164006 164007 When the server is not active, make sure the driver functions related to 164008 mode setting are not called. 164009 164010commit 32c0dcc8c0d1edba5d7e418fd2dc916847a4f069 164011Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 164012Date: Thu Jun 21 15:39:19 2007 -0400 164013 164014 xselinux: adjust the config file format to that expected by libselinux. 164015 164016 This file will eventually be moved out of the X source tree. 164017 164018commit 2030e9e5395be43bd8eab15b65c21ca4c2f1e619 164019Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 164020Date: Thu Jun 21 15:37:18 2007 -0400 164021 164022 xselinux: use new libselinux support for context labeling. 164023 164024 Remove all the config file parsing code and use the new lookup interface 164025 instead. 164026 164027commit 4d5df14f2c4a3108a8c8adfcf4766c0d1a9daad2 164028Author: Peter Hutterer <peter@cs.unisa.edu.au> 164029Date: Thu Jun 21 18:24:30 2007 +0930 164030 164031 Save processInputProc before wrapping it and restore it later, instead of 164032 using a hardcoded ProcessKeyboardEvent. Otherwise we lose the ability to 164033 process DeviceKeyEvents after the first key press. 164034 164035 This should be the correct fix now. 164036 164037commit f6f3322fc7562dbfccaf798b05d42ea222860ba8 164038Author: Peter Hutterer <peter@cs.unisa.edu.au> 164039Date: Thu Jun 21 18:16:21 2007 +0930 164040 164041 Revert "Save processInputProc before wrapping it and restore it later, instead of" 164042 Mixed up source trees, this was a very bad fix. 164043 164044 This reverts commit 8431f6083076cf1e701366767d8f8d32a628200f. 164045 164046commit 8431f6083076cf1e701366767d8f8d32a628200f 164047Author: Peter Hutterer <peter@cs.unisa.edu.au> 164048Date: Thu Jun 21 17:00:41 2007 +0930 164049 164050 Save processInputProc before wrapping it and restore it later, instead of 164051 using a hardcoded ProcessKeyboardEvent. Otherwise we lose the ability to 164052 process DeviceKeyEvents after the first key press. 164053 164054commit 0c33dc152e372cdc8ae59d9a5696b0774bcd03b7 164055Author: Peter Hutterer <peter@cs.unisa.edu.au> 164056Date: Thu Jun 21 15:47:48 2007 +0930 164057 164058 Set the detail field for DeviceKeyEvents to the keycode. 164059 164060commit e523859a952d49b20f3d10152cc0ef695d2c12a1 164061Author: Alan Coopersmith <alan.coopersmith@sun.com> 164062Date: Wed Jun 20 17:54:38 2007 -0700 164063 164064 Include module name in "already built-in" message 164065 164066commit 5138f710a1574fef6f553f3fe2fccac0620d2584 164067Author: Arcady Goldmints-Orlov <arcadyg@nvidia.com> 164068Date: Wed Jun 20 16:31:55 2007 -0700 164069 164070 Fixed fbSolidBoxClipped() to fill the right place. 164071 164072 Changed an X2 to a Y1. 164073 164074 Signed-off-by: Aaron Plattner <aplattner@nvidia.com> 164075 164076commit c9b79a355845c895aca8303a39798264d80b6212 164077Author: Michel Dänzer <michel@tungstengraphics.com> 164078Date: Wed Jun 20 18:56:06 2007 +0200 164079 164080 exaPolyFillRect: Don't track damage explicitly. 164081 164082 All callers should already do it. 164083 164084 Also don't leak pReg. 164085 164086commit 87966c5d2889873cea6cbc16b7e4399490dfaec1 164087Author: Michel Dänzer <michel@tungstengraphics.com> 164088Date: Wed Jun 20 18:42:00 2007 +0200 164089 164090 exaGetImage: Don't migrate or try to accelerate for 1x1. 164091 164092 This is mainly to avoid wasting effort for XSync(), but just reading a single 164093 pixel directly is probably faster than DownloadFromScreen anyway. Though in 164094 light of the latter, even larger thresholds might be useful. 164095 164096 Also move the swappedOut check before the migration checks because migration 164097 can't actually occur when swapped out. 164098 164099commit 1f97a7647606226cde61d6ad7a94f2b0b571a06c 164100Merge: 5dee64fc9 2e7fef7d0 164101Author: Peter Hutterer <peter@cs.unisa.edu.au> 164102Date: Tue Jun 19 17:20:52 2007 +0930 164103 164104 Merge branch 'master' into mpx 164105 164106 Conflicts: 164107 164108 dix/devices.c 164109 hw/xfree86/common/xf86Xinput.c 164110 hw/xfree86/loader/xf86sym.c 164111 mi/mieq.c 164112 164113commit 40f27a2df4906d9ceb1c78f6163a62c497321535 164114Author: Michel Dänzer <michel@tungstengraphics.com> 164115Date: Tue Jun 19 09:11:16 2007 +0200 164116 164117 mieqEnqueue: Make local queue tail variables unsigned. 164118 164119 So the modulo arithmetic actually works as intended... thanks to Peter Hutterer 164120 for pointing out the problem. 164121 164122commit 5dee64fc99f34e091abce65d47c4b6f026ab4849 164123Author: Peter Hutterer <peter@cs.unisa.edu.au> 164124Date: Tue Jun 19 11:31:22 2007 +0930 164125 164126 Add a few comments to ProcXGrabDevice. 164127 164128commit 9e257029c760883c4ea0715d4fd06476f3fe8053 164129Author: Peter Hutterer <peter@cs.unisa.edu.au> 164130Date: Tue Jun 19 11:28:07 2007 +0930 164131 164132 Add implicitGrab field to GrabInfoRec. 164133 Is set when passive grab is implicit as result of a ButtonPress event. If this 164134 is the case, we need to store the XI mask as well as the core mask to ensure 164135 delivery of XI events during the grab's lifetime. 164136 164137 Remove all core grabs on other devices when client issues a GrabPointer or 164138 GrabKeyboard request. Let's assume that the client really only wants one 164139 device to interact, so this seems like a reasonable solution. 164140 164141commit ab7a6d860d4a275a810a64b1ba7b13726ed10575 164142Merge: e9130b8ba 2e7fef7d0 164143Author: Ian Romanick <idr@us.ibm.com> 164144Date: Mon Jun 18 16:51:13 2007 -0700 164145 164146 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 164147 164148 Conflicts: 164149 164150 hw/xfree86/os-support/bus/Pci.c 164151 hw/xfree86/os-support/bus/linuxPci.c 164152 164153commit 2e7fef7d0837939e822c40b6ac77e7f0e66d57bd 164154Author: Adam Jackson <ajax@benzedrine.nwnk.net> 164155Date: Mon Jun 18 12:08:39 2007 -0400 164156 164157 Make xf86{En,Dis}ableInterrupts no-ops on Linux. 164158 164159commit 831d3b7f8d053aba649c8d04af3bef96376bdc3a 164160Author: Lennart Buytenhek <buytenh@wantstofly.org> 164161Date: Mon Jun 18 12:05:55 2007 -0400 164162 164163 Compile fixes for Linux ARM platforms. 164164 164165commit 562ca3f2f9005e7c5ed0a24b0759051ded2173e9 164166Author: Zephaniah E. Hull <warp@agamemnon.b5> 164167Date: Mon Jun 18 12:00:49 2007 -0400 164168 164169 In NewInputDeviceRequest, only call EnableDevice if xf86Screens[0]->vtSema is 164170 true, preventing unwanted behavior in the case where a device is added while 164171 the user is in a different VT. 164172 164173commit 3e894974cdd6a75683d4601f71622d1da7ec4395 164174Author: Peter Hutterer <peter@cs.unisa.edu.au> 164175Date: Fri Jun 15 17:16:16 2007 +0930 164176 164177 Only decrement buttonsDown when the button count is greater than 0. 164178 Device drivers flush their buttons on device init and cause a button down 164179 event to be generated. If we unconditionally decrease the buttons, we won't be 164180 able to ever get a passive device grab. 164181 164182 Format documentation for CheckDeviceGrabs to make it readable. 164183 164184commit 42c2e14b254f6f882b3e79444360ab855db43e27 164185Author: Matthieu Herrb <matthieu@deville.herrb.com> 164186Date: Fri Jun 15 00:14:02 2007 +0200 164187 164188 swap xOrigin and yOrigin in SProcRenderSetPictureClipRectangles. 164189 164190 Fixes Xrender clipping rectangles when X server and client are of 164191 different endianness, shown by xterm 225 among others. 164192 164193commit 0aaf51546666f71e8d1711978bc7988ec2ecc7d9 164194Author: Peter Hutterer <peter@cs.unisa.edu.au> 164195Date: Tue Jun 12 16:55:26 2007 +0930 164196 164197 Zero deviceMask and genericMask when creating a grab via CreateGrab(). 164198 164199commit 78179ae827bb5d19abb1340084362bc51ad5c1e5 164200Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164201Date: Mon Jun 11 22:46:42 2007 -0400 164202 164203 Remove fbmmx.[ch] files 164204 164205commit eb2d7fe02f9cbca57b462bba05498e2d59316fbc 164206Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164207Date: Mon Jun 11 22:43:01 2007 -0400 164208 164209 Replace fbFillmmx() with pixman_fill() and remove fbmmx.[ch] 164210 164211commit f52ae237d3eec79ccd64cdd77271aeacc37af70c 164212Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164213Date: Mon Jun 11 22:02:39 2007 -0400 164214 164215 Require pixman 0.9.2 164216 164217commit d1d85c04e248f46b1cf1b1d25fdd56aa69b8f0ee 164218Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164219Date: Mon Jun 11 21:25:42 2007 -0400 164220 164221 Delete fbBltmmx(). 164222 164223commit e2a720c9a17dc860ee0a858c2b21fd71e86cdcd0 164224Merge: 878cac71a 3f9adb18f 164225Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 164226Date: Mon Jun 11 15:29:11 2007 -0400 164227 164228 Merge branch 'master' into XACE-SELINUX 164229 164230 Conflicts: 164231 164232 dix/dixutils.c 164233 164234commit 3f9adb18f127318d054f30a57e3a77176e14c692 164235Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164236Date: Mon Jun 11 15:19:27 2007 -0400 164237 164238 Port a few forgotten fbSolidFillmmx()'es to fbFillmmx(). Use pixman_blt() instead of fbBltmmx() 164239 in fbCopyNToN(). 164240 164241commit 878cac71aa0018deee861b297638c0744dba631b 164242Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 164243Date: Mon Jun 11 14:19:37 2007 -0400 164244 164245 xselinux: use new libselinux support for private Flask definitions. 164246 164247 Removes indirect dependency on kernel headers. 164248 164249commit d06099b38e8445e6e31f5178ffefcc31a71080ef 164250Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164251Date: Mon Jun 11 13:28:24 2007 -0400 164252 164253 Remove fbCopyAreammx() and fbSolidFillmmx() 164254 164255commit d4a034370c8ae71b2cc4fe824ceee58b19624f35 164256Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164257Date: Mon Jun 11 13:26:24 2007 -0400 164258 164259 Split fbSolidFillmmx() into a new FbFillmmx() function. Call that from fbFill(). 164260 164261commit 3210902a7334f3d8d6c18a34a3cb3f55803b0043 164262Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164263Date: Mon Jun 11 13:14:50 2007 -0400 164264 164265 Split fbCopyAreammx() into a new fbBltmmx() function; call this 164266 164267 function from fbCopyNToN(). 164268 164269commit 8d5f4368eac1b259db3e61f877a4cc10f04efa2f 164270Author: Clark Rawlins <clark.rawlins@escient.com> 164271Date: Mon Jun 11 16:53:38 2007 +0200 164272 164273 Really make sure BUILD_TIME doesn't have a leading zero. 164274 164275 date +'%k%M%S' still gives a leading zero in the hour after midnight... 164276 164277 Add a leading 1 and remove it in xf86PrintBanner(). 164278 164279commit 54e023cec07aa7e392da36e11d0a4667b8341370 164280Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164281Date: Mon Jun 11 09:16:46 2007 -0400 164282 164283 Don't pass regions to pixman_image_composite() anymore. 164284 164285commit 5cbec267b6426960c90f6bcff1d051af5084538c 164286Author: Michel Dänzer <michel@tungstengraphics.com> 164287Date: Mon Jun 11 12:38:41 2007 +0200 164288 164289 Make sure BUILD_TIME doesn't have a leading zero. 164290 164291 It causes the compiler to treat it as an octal constant instead of decimal as 164292 intended, which could even cause a build failure in the cases of 08 and 09. 164293 164294 Thanks to Clark Rawlins for pointing out the problem. 164295 164296commit 1aceec61ff203848576c47a1eab13f90a67d7176 164297Author: Michel Dänzer <michel@tungstengraphics.com> 164298Date: Mon Jun 11 09:23:19 2007 +0200 164299 164300 DRI: Clip cliprects obtained from DRIGetDrawableInfo to screen dimensions. 164301 164302 This is to avoid issues with redirected windows which are located partly or 164303 fully outside of a screen edge, resulting in unusual cliprects which the 3D 164304 drivers generally can't handle. The symptoms in such cases would be incorrect 164305 rendering or even crashes or hangs. 164306 164307commit 5d896e43fd056d935935b4eb66562791edc247a1 164308Author: Michel Dänzer <michel@tungstengraphics.com> 164309Date: Mon Jun 11 09:23:19 2007 +0200 164310 164311 DRITreeTraversal: Stop walking tree when we've seen all DRI windows. 164312 164313commit 0fb44c6f9a0415184818ba8357a21ff920e907dc 164314Author: Michel Dänzer <michel@tungstengraphics.com> 164315Date: Mon Jun 11 09:23:19 2007 +0200 164316 164317 DRI: Fix build warning. 164318 164319commit 644f7ddc0cb029e2ebca43742fd8a46a1a3f4c9f 164320Author: Michel Dänzer <michel@tungstengraphics.com> 164321Date: Mon Jun 11 09:23:18 2007 +0200 164322 164323 dixLookupClient: Use access parameter. 164324 164325commit 30a3297fed9af3a594aba0875a8f58a0a38b33fc 164326Author: Michel Dänzer <michel@tungstengraphics.com> 164327Date: Mon Jun 11 09:23:18 2007 +0200 164328 164329 mieq queue handling cleanups. 164330 164331 In particular, fix handling of wraparounds in mieqEnqueue. 164332 164333commit c1a49a9269f14b6975a1a2c751bb179757373f11 164334Author: Adam Jackson <ajax@benzedrine.nwnk.net> 164335Date: Sun Jun 10 22:14:57 2007 -0400 164336 164337 GNU is wrong and ` is not left-quote. 164338 164339commit 67a0a4da1a225ee3bd6bbd1846f8141fe333c884 164340Author: Dave Airlie <airlied@linux.ie> 164341Date: Mon Jun 11 11:55:11 2007 +1000 164342 164343 update xproto dependency to at least 7.0.9 164344 164345commit c079cce9d884ab03f305b3fba4a4e1247c023480 164346Author: Daniel Ciocea <daniel.ciocea@eosystems.ro> 164347Date: Fri Jun 8 18:12:21 2007 -0700 164348 164349 Fix sync polarity on Samsung SyncMaster 205BW monitor. 164350 164351 need to use standard VESA sync polarity instead of the 164352 EDID provided -hsync -vsync values. 164353 164354commit 9c47b86bd9a4633fda5fd305a09ac8623187efa0 164355Author: Aaron Plattner <aplattner@nvidia.com> 164356Date: Thu Jun 7 13:57:12 2007 -0700 164357 164358 Add new fb symbols to wfbrename.h. 164359 164360 Avoids crashes when wfbComposite calls the wrong image_from_pict. 164361 164362commit 567b5bf765254a4ae9cc7711bb6acfa89a9fd61c 164363Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164364Date: Tue Jun 5 20:26:49 2007 -0400 164365 164366 Delete fbCompositeCopyAreammx() 164367 164368commit 49ed31c0b323dd8c5887a803c199875e6f2330d8 164369Author: Søren Sandmann Pedersen <sandmann@redhat.com> 164370Date: Tue Jun 5 17:44:21 2007 -0400 164371 164372 Remove most of the fast-path MMX operations from fbmmx. fbCopyAreammx 164373 164374 and fbSolidFillmmx are still needed by other code. 164375 164376commit dfbe32b5b828cc4e3da36a0e2e6ad641164eaa5e 164377Author: Adam Jackson <ajax@benzedrine.nwnk.net> 164378Date: Mon Jun 4 18:07:00 2007 -0400 164379 164380 Remove the old Kerberos 5 authentication code. 164381 164382 Before you complain, this code hasn't seen material change since at least 164383 X11R6. It certainly does not build with any modern version of Kerberos. 164384 Anybody wanting krb5 auth to their X server should probably be using 164385 GSSAPI instead of internal krb5 API anyway. 164386 164387commit 793470a8356976ddd427280a738dfb6e1c0e4e70 164388Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 164389Date: Mon Jun 4 12:33:49 2007 -0400 164390 164391 dix: fix null pointer dereference in new resource lookup function. 164392 164393commit 75dece08fb72803d5116e6776e9f1534ff20e37b 164394Author: Zephaniah E. Hull <warp@agamemnon.b5> 164395Date: Mon Jun 4 09:09:20 2007 -0400 164396 164397 xf86PostMotionEvents[P] calls xf86SendDragEvents, xf86SendDragEvents 164398 unconditionally checks device->button->buttonsDown. 164399 164400 Let's make it possible to have a device with motion, but no buttons. 164401 164402 Without segfaulting. 164403 164404commit fbb9b203950e9d0e82574cde5b3e006b0e6b404f 164405Author: Zephaniah E. Hull <warp@agamemnon.b5> 164406Date: Mon Jun 4 06:59:42 2007 -0400 164407 164408 Let's not do a calloc and a free on every call to xf86PostMotionEvents. 164409 164410commit 3f4295e643ca56c40f33af7966e8efd367ef8749 164411Author: Zephaniah E. Hull <warp@agamemnon.b5> 164412Date: Mon Jun 4 06:48:06 2007 -0400 164413 164414 Add xf86PostMotionEventP, takes a pointer instead of a variable number of 164415 arguments. 164416 164417 Bump input ABI to 1.1 since we export this. 164418 164419commit a4f3473c88370b8411e016ebab619cffd33e58f9 164420Author: Zephaniah E. Hull <warp@agamemnon.b5> 164421Date: Mon Jun 4 06:39:02 2007 -0400 164422 164423 Fully init the AbsoluteClassRec in InitAbsoluteClassDeviceStruct. 164424 (Specificly, we were missing the screen field.) 164425 164426commit 0cbc3a4da2ddb6e4f30f60d2bc7f405d31aa554a 164427Author: Zephaniah E. Hull <warp@agamemnon.b5> 164428Date: Mon Jun 4 02:03:44 2007 -0400 164429 164430 Print the build time as well as the date if we can. 164431 164432commit 9a7aaeb3f6ff79af60fde91cd0575a54ba0b9587 164433Author: Daniel Stone <daniel@fooishbar.org> 164434Date: Mon May 28 13:54:16 2007 +0300 164435 164436 XFree86: Input: Assume core events per default 164437 164438 Assume that a device will be sending core events, unless explicitly 164439 specified otherwise. 164440 164441commit 94361cbba7f866144691f6f5e9251a550e0e0cb8 164442Author: Daniel Stone <daniel@fooishbar.org> 164443Date: Mon May 28 13:54:47 2007 +0300 164444 164445 XFree86: Input: Perform case-insensitive comparisons on option names 164446 164447commit e5ce982381c4092252d6b55fcefcc9a3cd21e656 164448Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 164449Date: Sun Jun 3 09:40:37 2007 +1000 164450 164451 Include pixman.h from fb.h or compile of some files will fail 164452 164453 Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> 164454 164455commit 90eb22656c34d2d08a8dccaf05e6d081c56bd7f3 164456Author: Adam Jackson <ajax@benzedrine.nwnk.net> 164457Date: Sat Jun 2 16:49:26 2007 -0400 164458 164459 Minor cleanup/robustification to config parsing. 164460 164461commit f6a983533bdc84752562ef0be25b320678bf08a1 164462Author: Adam Jackson <ajax@benzedrine.nwnk.net> 164463Date: Sat Jun 2 16:37:39 2007 -0400 164464 164465 Don't warn about default behaviour when autoconfigging. 164466 164467commit 21e8f4eb02842f877336db08c332d8ee4a381ee0 164468Author: Adam Jackson <ajax@benzedrine.nwnk.net> 164469Date: Sat Jun 2 16:13:01 2007 -0400 164470 164471 Don't print lack of DRI support as an error in AIGLX init. 164472 164473commit 0e1384d8318637f75d04d3d1b7600f7cad40117e 164474Author: Adam Jackson <ajax@benzedrine.nwnk.net> 164475Date: Sat Jun 2 16:07:20 2007 -0400 164476 164477 Delete VDIF support; it was never used anyway. 164478 164479commit 66702f3c1c6c884e83744c72da173cc32f22b2f4 164480Author: Henry Zhao <henryz@localhost.localdomain> 164481Date: Fri Jun 1 23:55:40 2007 -0700 164482 164483 Need to use minPitch in miScanLineWidth() to get the shrinked 164484 linePitch. 164485 164486commit 26b21157cf934ae387b15faa9ebb060120e6a0d6 164487Author: Peter Hutterer <peter@cs.unisa.edu.au> 164488Date: Thu May 31 17:05:14 2007 +0930 164489 164490 Add a deviceMask to the GrabRec and don't interfere with passiveGrabs. 164491 164492 This quickfixes event delivery problems with XI events when a grab was on. 164493 deviceMask is only used when the grab was from a ButtonPress to preserve 164494 potential XI event masks. 164495 164496 This is not an ideal solution but it works until I have time to work on 164497 PassiveGrabs. 164498 164499commit fa877d7ff25c4ec45288e1fea70d4f5e1baf3ef3 164500Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 164501Date: Wed May 30 13:06:45 2007 +0100 164502 164503 Fix mode validation against the maximum X/Y values configured 164504 at server startup, and not against the virtual X/Y parameters 164505 as they can change. 164506 164507 This fixes an issue when canGrow is TRUE and modes get dropped 164508 when using the virtual X/Y parameters. 164509 164510commit 182ab3a3d5b28daa86e6d6155e76ce759687ae6d 164511Author: Peter Hutterer <peter@cs.unisa.edu.au> 164512Date: Wed May 30 17:44:31 2007 +0930 164513 164514 Fix up comments for event.c 164515 164516commit b5db863945fa8045995b3bf742c44e40f2650f04 164517Author: Paulo Ricardo Zanoni <prz05@c3sl.ufpr.br> 164518Date: Wed May 30 16:15:15 2007 +0930 164519 164520 Allow value None for windows when setting ClientPointer. 164521 164522 We need this for clients that need to set the ClientPointer but don't have a 164523 window on display yet. If used, it will set the device as the ClientPointer 164524 for the requesting client. 164525 164526commit 99eae8bea6724a24477375ad5b2d31cc4883cf6b 164527Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 164528Date: Tue May 29 22:04:36 2007 -0400 164529 164530 I/O enable/disable update for the Hurd 164531 164532commit 3c6f1428489c1f71acd41066ea73ef4ae7c60f17 164533Author: Julien Cristau <jcristau@debian.org> 164534Date: Tue May 29 22:01:30 2007 -0400 164535 164536 Make sure that the ramdac symbols are present in the server 164537 164538 The former ramdac module is now built into the server, so its symbols need to 164539 be explicitly exported to drivers (Debian #423129). 164540 164541commit ee20c481eede0954f4a8bef5113979b101863c32 164542Author: Matthieu Herrb <matthieu@deville.herrb.com> 164543Date: Tue May 29 14:54:27 2007 -0600 164544 164545 Remove wscons keyboard handling stuff that doesn't belong there anymore. 164546 164547commit 60de6c7ef9bdcee043f63e8e0d493e6feba6a9d0 164548Merge: 3a6549a16 2f13b7c11 164549Author: Matthieu Herrb <matthieu@deville.herrb.com> 164550Date: Tue May 29 12:14:49 2007 -0600 164551 164552 Merge branch 'master' of ssh://herrb@git.freedesktop.org/git/xorg/xserver 164553 164554commit 3a6549a163aba26bf4ac58b050c493fba0df14c6 164555Author: Matthieu Herrb <matthieu@deville.herrb.com> 164556Date: Tue May 29 12:14:23 2007 -0600 164557 164558 Make this build on OpenBSD 164559 164560commit 2f13b7c113c17239e382dd3640e9c29201d8ab1f 164561Author: Drew Parsons <dparsons@debian.org> 164562Date: Wed May 30 02:13:36 2007 +1000 164563 164564 Update Xprint build for pixman. 164565 164566 Xprt links libfb, which now uses pixman. Update configure.ac to 164567 require module $PIXMAN for XPRINT. 164568 Also, use $(top_builddir) to reference libfb.la and other local 164569 libraries, rather than using the relative reference ../.. 164570 164571commit ba0b7d47ab0c24d5a29228f8af583044060464bd 164572Author: David Nusinow <dnusinow@debian.org> 164573Date: Mon May 28 21:57:04 2007 -0400 164574 164575 Fix for GNU/kFreeBSD 164576 164577commit 2267bf48b385c93243e26c3bb84ebb04c7fdb39f 164578Author: Bastian Blank <waldi@debian.org> 164579Date: Mon May 28 21:55:05 2007 -0400 164580 164581 Fixes for s390 164582 164583commit 857ddbb660a21cad1c16f4fb2dc8a904d6655304 164584Author: Eugene Konev <ejka@imfi.kspu.ru> 164585Date: Mon May 28 21:53:02 2007 -0400 164586 164587 Allow configurable serverconfigdir for security policy location 164588 Allow the location of the SERVERCONFIGdir variable to be defined at 164589 compile-time. This allows us to specify where the security policy will be 164590 located (Debian uses this to put it in /etc). The default is to the 164591 previous location. 164592 164593commit 78d01d1008973899d931ef44b47d5f0b5f220b0d 164594Author: Gerhard Tonn <gt@debian.org> 164595Date: Mon May 28 21:48:58 2007 -0400 164596 164597 Miscellaneous fixes for S/390. 164598 164599commit d98bd4bf908c2c51fcfd3a4c3230de17f2567244 164600Author: Branden Robinson <branden@debian.org> 164601Date: Mon May 28 21:44:59 2007 -0400 164602 164603 Overhaul xorg.conf manpage 164604 164605 Major stylistic cleanups, greatly expanded cross-reference ("SEE ALSO") 164606 section and some typo fixes. 164607 164608 This patch by Branden Robinson. Forward-ported by Fabio M. Di Nitto. 164609 164610commit 6bf8d5019313ee2251a44dfb7ad3435a3c6db7eb 164611Author: David Nusinow <dnusinow@debian.org> 164612Date: Mon May 28 21:42:10 2007 -0400 164613 164614 Read ROM in chunks 164615 This patch speeds up reads of the ROM by reading in large chunks rather 164616 than one byte at a time. This patch was by Dann Frazier. 164617 164618commit 6fdd134a0c3e6fdde9b089100e8783705c9cc6ac 164619Author: David Nusinow <dnusinow@debian.org> 164620Date: Mon May 28 21:39:12 2007 -0400 164621 164622 Fix up xnest manpage 164623 I believe this patch was originally by Branden Robinson 164624 164625commit 6a870992d81a6bacfa9d313c15784fdb281d474f 164626Author: Keith Packard <keithp@dulcimer.keithp.com> 164627Date: Fri May 25 20:33:08 2007 -0700 164628 164629 xf86XVFillKeyHelper assumed root clip never changed. 164630 164631 When the root window changed size, xf86XVFillKeyHelper would not revalidate 164632 the GC, leaving the clip at the old size causing lossage (and possibly 164633 memory corruption if the screen and frame buffer shrank). 164634 164635 Fixed by just using a scratch GC; saving memory, eliminating bugs and 164636 shrinking the code. 164637 164638commit 3c982bc1a49509dda7bc469b0eced44df02755b3 164639Author: Luo Jie <luojie@nlsde.buaa.edu.cn> 164640Date: Thu May 24 11:13:03 2007 -0700 164641 164642 Reinstate an apparently mis-deleted ';' from a for loop with no body. 164643 164644 Fixes an error returning "No core keyboard" with multiple keyboards. 164645 164646commit 4d7469f75fadfc4a59664e88e18eb304203670f4 164647Author: Luo Jie <luojie@nlsde.buaa.edu.cn> 164648Date: Thu May 24 11:04:06 2007 -0700 164649 164650 Fix a typo in using memcpy in xwin. 164651 164652commit 0b988450462ddb005311e68502357baf272e6371 164653Author: Luo Jie <luojie@nlsde.buaa.edu.cn> 164654Date: Thu May 24 11:02:28 2007 -0700 164655 164656 Fix os/utils.c compile with mingw. 164657 164658commit 1f48995d66c0072caa7e5ce2845be642221dd56d 164659Author: Luo Jie <luojie@nlsde.buaa.edu.cn> 164660Date: Thu May 24 11:01:15 2007 -0700 164661 164662 Fix build of composite, dix, and randr when Xinerama is disabled. 164663 164664commit 8f98be7db303bc3db650054efb86843c70114451 164665Author: Eric Anholt <eric@anholt.net> 164666Date: Thu May 24 11:00:04 2007 -0700 164667 164668 Fix bswap detection on BSD (mis-added '_' in function names). 164669 164670commit 9616a042855399f0ee9c6489ea824621ea5fee18 164671Author: Matthias Drochner <drochner@netbsd.org> 164672Date: Tue Apr 10 16:15:40 2007 -0700 164673 164674 Fix build on NetBSD/amd64. 164675 164676commit ddb26bccd275f4fc011f7a2be685d1ce58555a00 164677Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 164678Date: Thu May 24 12:20:24 2007 -0400 164679 164680 dix: Add a bunch of new access codes. 164681 164682 These were determined through an analysis of the core protocol and 35 of 164683 the most common protocol extensions. There remain four bits for future 164684 use. 164685 164686commit 649e7f82d8d4333443493056b81eb20d6cf022bc 164687Author: Michel Dänzer <michel@tungstengraphics.com> 164688Date: Thu May 24 12:10:05 2007 +0200 164689 164690 Consolidate portPriv->pDraw assignments into xf86XVEnlistPortInWindow. 164691 164692 This avoids a crash in xf86XVReputVideo and also cleans up the code slightly. 164693 164694commit 047bf3349bb697c73c95729a8bbf15f72605901f 164695Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164696Date: Wed May 23 16:56:05 2007 -0400 164697 164698 Delete trapezoid rendering code; replace with pixman calls 164699 164700commit 9d87ef4e0dff40ea39f1b209c67b90079fc79065 164701Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164702Date: Wed May 23 15:50:25 2007 -0400 164703 164704 - Make image_from_pict() non-static 164705 - Delete fbedge.c and fbedgeimp.h 164706 - Use pixman_rasterize_edges() in fbtrap.c 164707 164708commit 2a960c442bd7560630f52b55d82ec0517542ee5a 164709Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164710Date: Wed May 23 13:08:26 2007 -0400 164711 164712 Port renderedge.c to pixman 164713 164714commit 3ba3ede9bbdfc6376b6f6e0b6ce8280a05e6584d 164715Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164716Date: Wed May 23 12:56:04 2007 -0400 164717 164718 Add missing offsets for window coordinates - reported by Colin Harrison 164719 164720commit 2a4aa63a23ddd816b647b851a01865861827a7eb 164721Merge: 9cee4ec5e cc648e609 164722Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 164723Date: Tue May 22 14:50:26 2007 -0400 164724 164725 Merge branch 'master' into XACE-SELINUX 164726 164727 Conflicts: 164728 164729 include/miscstruct.h 164730 164731commit cc648e609d472472bac4a2e568eb3598b3690ba3 164732Author: Michel Dänzer <michel@tungstengraphics.com> 164733Date: Tue May 22 10:51:56 2007 +0200 164734 164735 EXA: Export ExaOffscreenMarkUsed. 164736 164737 Can be used to inform EXA that an offscreen area is used outside of EXA. 164738 164739commit e6a7198e7cd96f1fe0654cc6811a977821579258 164740Author: Adam Jackson <ajax@benzedrine.nwnk.net> 164741Date: Tue May 22 10:51:55 2007 +0200 164742 164743 Bug #8991: Add glXGetDrawableAttributes dispatch; fix texture format therein. 164744 164745 Adapted to master branch by Michel Dänzer <michel@tungstengraphics.com>. 164746 164747commit 6324bfc468f7a645d2fee59f1c921a4328a4639f 164748Author: Michel Dänzer <michel@tungstengraphics.com> 164749Date: Tue May 22 10:51:53 2007 +0200 164750 164751 AIGLX: Zero-copy texture-from-pixmap. 164752 164753 When available, use the 2D driver texOffsetStart hook and the 3D driver 164754 setTexOffset hook to save the overhead of passing the pixmap data to 164755 glTex(Sub)Image. 164756 164757 The basic idea is to update the driver specific 'offset' for bound pixmaps 164758 before dispatching a GLX render request and to flush immediately afterwards 164759 if there are any pixmaps bound. This should ensure that the 3D driver can 164760 use pixmaps for texturing directly regardless of the X server moving them 164761 around. 164762 164763commit 5006d08d7fc56d3d380cc6b75297f94e8594eb54 164764Author: Michel Dänzer <michel@tungstengraphics.com> 164765Date: Tue May 22 10:51:52 2007 +0200 164766 164767 DRI: Add TexOffset driver hooks. 164768 164769 To be used by AIGLX for GLX_EXT_texture_from_pixmap without several data copies. 164770 164771 The texOffsetStart hook must make sure that the given pixmap is accessible by 164772 the GPU for texturing and return an 'offset' that can be used by the 3D 164773 driver for that purpose. 164774 164775 The texOffsetFinish hook is called when the pixmap is no longer being used for 164776 texturing. 164777 164778commit ff2eae86b6a8760befbbc5d605debebe7b024c05 164779Author: David Nusinow <dnusinow@debian.org> 164780Date: Mon May 21 19:50:04 2007 -0400 164781 164782 Fix boolean thinko that prevented working without a server layout 164783 164784commit 56fd92715567cd32e4b725b3791de9ac4e3879aa 164785Author: Soren Sandmann Pedersen <sandmann@redhat.com> 164786Date: Mon May 21 20:00:25 2007 -0400 164787 164788 Remove fast path code from fbpict.c 164789 164790 Remove the various fast path functions from fbpict, and instead use 164791 pixman_image_composite(). 164792 164793commit d0e8f474099dea40bbea555427772724ccb787d1 164794Author: Peter Hutterer <peter@cs.unisa.edu.au> 164795Date: Mon May 21 13:06:21 2007 +0930 164796 164797 Change GrabDevice's parameter name back to "coreGrab". 164798 Was changed during a global search/replace for 164799 5c680e94938c16e04a4349715cf11796b80400db. 164800 164801 Otherwise people like me introduce bugs (e.g. in 164802 333bab4b235801efd2b2de5b60df5b0c44048d3e) 164803 164804commit 167e1773de0fb566559d4362ff929eedcdb6d055 164805Author: Peter Hutterer <peter@cs.unisa.edu.au> 164806Date: Mon May 21 10:45:10 2007 +0930 164807 164808 Remove unused include "gestr.h" 164809 164810commit 7e2c935920cafadbd87c351f1a3239932864fb90 164811Author: Fredrik Höglund <fredrik@kde.org> 164812Date: Fri May 18 20:06:14 2007 +0200 164813 164814 Add a new IDLETIME system sync counter. 164815 164816 This counter exposes the time in milliseconds since the last 164817 input event. Clients such as screen savers and power managers 164818 can set an alarm on this counter to find out when the idle time 164819 reaches a certain value, without having to poll the server. 164820 164821commit 756acea23a0cc56c470bcd77c6f5638d923ab3d1 164822Author: Soren Sandmann Pedersen <sandmann@redhat.com> 164823Date: Fri May 18 13:39:12 2007 -0400 164824 164825 Use pixman_image_set_indexed() to make 8 bit work 164826 164827commit 7916419a0092b8bf9713c0840f9e969950d7aa85 164828Author: Soren Sandmann Pedersen <sandmann@redhat.com> 164829Date: Fri May 18 11:58:24 2007 -0400 164830 164831 Comment out setup of general MMX code 164832 164833commit 998164bac648756e5b5254aa36e075ae360d3972 164834Author: Soren Sandmann Pedersen <sandmann@redhat.com> 164835Date: Fri May 18 11:36:20 2007 -0400 164836 164837 Move fbCompositeGeneral() to fbpict.c and remove fbcompose.c 164838 164839commit a2e3614eb8f0fa198615df492b03ff36bc9c1121 164840Author: Soren Sandmann Pedersen <sandmann@redhat.com> 164841Date: Fri May 18 11:33:11 2007 -0400 164842 164843 Break image_from_pict() into a few subfunctions. 164844 164845commit c5ef84c325440af5fbdf9f44c3781d99a0392df9 164846Author: Soren Sandmann Pedersen <sandmann@redhat.com> 164847Date: Thu May 17 21:31:08 2007 -0400 164848 164849 Make the general compositing code create a pixman image and call 164850 164851 pixman_image_composite(). Leave the general code commented out for now. 164852 164853commit 076d070e186afeb416976ae74fbfd50c86db10c5 164854Author: Keith Packard <keithp@neko.keithp.com> 164855Date: Thu May 17 20:24:18 2007 -0700 164856 164857 Use Screen block handler for rotation to draw under DRI lock. 164858 164859 DRI uses a non-screen block/wakeup handler which will be executed after the 164860 screen block handler finishes. To ensure that the rotation block handler is 164861 executed under the DRI lock, dynamically wrap the screen block handler for 164862 rotation. 164863 164864commit 915563eba530c5e2fdc2456cf1c7c3cc09b3add0 164865Author: Keith Packard <keithp@neko.keithp.com> 164866Date: Thu May 17 20:22:43 2007 -0700 164867 164868 Disable all outputs and crtcs at startup. 164869 164870 Leaving devices enabled during server startup can cause problems during the 164871 initial mode setting in the server, especially when they are used for 164872 different purposes by the X server than by the BIOS. Disabling all of them 164873 before any mode setting is attempted provides a stable base upon which the 164874 remaining mode setting operations can be built. 164875 164876commit 0375009a97c2ab7f0e0f0265463d45c0580388c6 164877Author: Soren Sandmann Pedersen <sandmann@redhat.com> 164878Date: Thu May 17 12:59:24 2007 -0400 164879 164880 Remove excessive unrolling in fbCompositeSrc_x888x8x8888() and fix bug where 164881 the source alpha was used instead of 0xff. 164882 164883commit 0b4db74922299df785e6273fdb1bf65c38d36070 164884Author: Peter Hutterer <peter@cs.unisa.edu.au> 164885Date: Thu May 17 18:00:07 2007 +0930 164886 164887 Squish the flags into the upper bits of the mode field for EnterLeave events. 164888 This way we have enough space for the detail field. 164889 164890commit 333bab4b235801efd2b2de5b60df5b0c44048d3e 164891Author: Peter Hutterer <peter@cs.unisa.edu.au> 164892Date: Thu May 17 17:59:02 2007 +0930 164893 164894 Clean up, correct some comments. 164895 Send event type down with the RawEvents. 164896 164897commit bc334286b060bc8d0c829b18acebadf24fbdaf19 164898Author: Peter Hutterer <peter@cs.unisa.edu.au> 164899Date: Thu May 17 10:36:46 2007 +0930 164900 164901 Init device axis with -1,-1 minimum values. 164902 164903 This is needed to be able to cross to screens located east/north of the 164904 current active screen. 164905 164906commit 546465ee6aa6584780aec6357f32d205c807ae71 164907Author: Soren Sandmann Pedersen <sandmann@redhat.com> 164908Date: Wed May 16 17:42:04 2007 -0400 164909 164910 Make fbFetch_b8g8r8() actually write the read value to the buffer 164911 164912commit 0fcd17c9181901c419cc32bc24c07fe5a6934d81 164913Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164914Date: Tue May 15 17:59:13 2007 -0400 164915 164916 Use pixman short formats, revert the gradient_stop change 164917 164918commit f4c1d5fc28a5a7fe2592505350f9e2331f6049b7 164919Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164920Date: Tue May 15 17:12:22 2007 -0400 164921 164922 Use pixman types for transforms and vectors 164923 164924commit f2e30e7d0a1d075e7e83c5b5ceca9e4752951138 164925Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164926Date: Tue May 15 16:51:21 2007 -0400 164927 164928 Use the pixman fixed point types and macros 164929 164930commit 3da842bf930d7875599ca0c06cb4a09cfa987ac5 164931Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164932Date: Tue May 15 14:57:14 2007 -0400 164933 164934 Revert various fast path functions to their pre-pixman-merge state 164935 since they fail rendercheck. Remove their associated macros. 164936 164937 See bug 10903. 164938 164939commit 1568b6b6a0d7337f29c7b87cc46ae64b3b0f8fdf 164940Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164941Date: Sat May 12 20:33:23 2007 -0400 164942 164943 Port large amounts of the region code to pixman 164944 164945commit dde0ceac4ea7639d0096bfd26f37c5851778854c 164946Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164947Date: Sat May 12 17:41:47 2007 -0400 164948 164949 Add new InitRegions() function called from dix/main 164950 164951commit e037052ac522150786abf44d3a04c813cc490050 164952Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164953Date: Sat May 12 16:58:54 2007 -0400 164954 164955 Turn boxes and regions into typedefs for pixman types 164956 164957commit 8e56f5be4b70773c899f01b9ccd2e88d523327e4 164958Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 164959Date: Fri May 11 11:45:37 2007 -0400 164960 164961 Add dependency on pixman 0.9.0 164962 164963commit a8c56372ba8aa36bac09877c478ff53ea5358df7 164964Author: Peter Hutterer <peter@cs.unisa.edu.au> 164965Date: Tue May 15 10:24:22 2007 +0930 164966 164967 Revert "Allow events to grabWindows event if the device is not grabbed." 164968 164969 This reverts commit 339b73e710a0920608a3fbcb20b406f0f6c4e0f6. 164970 Allowing those events can lead to race conditions. 164971 164972 Conflicts: 164973 164974 dix/events.c 164975 164976commit 81fc6a128b64ad412064405ed45db0175398e3f0 164977Author: Peter Hutterer <peter@cs.unisa.edu.au> 164978Date: Mon May 14 18:12:56 2007 +0930 164979 164980 Add ungrab support to ExtendedGrabDeviceRequest handling. 164981 Polish the code a bit. 164982 164983commit 5c680e94938c16e04a4349715cf11796b80400db 164984Author: Peter Hutterer <peter@cs.unisa.edu.au> 164985Date: Mon May 14 12:34:53 2007 +0930 164986 164987 Remove double-grab option. We only want one grab per device. 164988 164989commit 4dc973f08c54294e06bc831bd2839fd3aa24ecfb 164990Author: Peter Hutterer <peter@cs.unisa.edu.au> 164991Date: Mon May 14 10:42:03 2007 +0930 164992 164993 Add ExtendedGrabDevice handling. 164994 Add XGE handling in DeliverGrabbedEvent. 164995 164996 We can now grab something selecting XGE events, but the current code is a 164997 bit messy and doesn't work too well yet. 164998 164999commit a277f04ab08514462b7f10b4dd92eb326af85501 165000Author: Adam Jackson <ajax@benzedrine.nwnk.net> 165001Date: Wed May 9 22:03:12 2007 -0400 165002 165003 Remove mfb and cfb from include paths where they're not needed. 165004 165005commit 20c5250e487e032d392e2e4624021fccb1bfb72c 165006Author: Adam Jackson <ajax@benzedrine.nwnk.net> 165007Date: Wed May 9 21:49:44 2007 -0400 165008 165009 Use system copy of cbrt() if available. 165010 165011 Also move the replacement inline into miarc.c, since that's the only user. 165012 165013commit 8dcc37520d5e8c8b52cee81faa67fd5205548377 165014Author: Adam Jackson <ajax@benzedrine.nwnk.net> 165015Date: Wed May 9 18:57:05 2007 -0400 165016 165017 Use _X_INLINE instead of ad-hoc #defines. 165018 165019commit 6ff239cb4e67c0a2ea497a1714e5585c1d941af3 165020Author: Adam Jackson <ajax@benzedrine.nwnk.net> 165021Date: Wed May 9 18:38:33 2007 -0400 165022 165023 Make the use of ICEIL slightly less ugly. 165024 165025commit 178d426311bb3c7160f72b5d95b0a137eda09ba9 165026Author: Colin Harrison <colin.harrison-at-virgin.net> 165027Date: Fri May 11 10:08:42 2007 +0100 165028 165029 Missing piece from bug 9808 165030 165031commit ebaa6c920c82401952a0ccc991b94574306449bd 165032Author: Matthias Hopf <mhopf@suse.de> 165033Date: Thu May 10 15:25:31 2007 +0200 165034 165035 Disable Simba PCI bridge routing code (Bug #8020). 165036 165037 The code in hw/xfree86/os-support/bus/sparcPci.c:simbaCheckBus() 165038 is trying to mimmick VGA routing by disabling I/O space responses 165039 behind the Simba PCI-PCI controller. 165040 165041 Unfortunately, doing this also happens to disable access to the 165042 IDE controller I/O space registers, thus crashing the system. The 165043 granularity of the I/O disabling in the Simba controller is not 165044 fine enough to disable VGA without also disabling the IDE controller 165045 registers. 165046 165047commit 86c4941727f7c673ae6bb88c67443fa25935c7f5 165048Author: Colin Harrison <colin.harrison-at-virgin.net> 165049Date: Wed May 9 16:55:27 2007 +0100 165050 165051 fix an occasional crash in GetWindowName() (bug: 9798) 165052 165053commit be44018a3c6172caf3e91c36ea321420d104e79f 165054Author: Colin Harrison <colin.harrison-at-virgin.net> 165055Date: Wed May 9 16:55:09 2007 +0100 165056 165057 Fix bad use of hwnd (bug: 9808) 165058 165059commit d3248b66a650c6c629cd66240e25004869217d2e 165060Author: Colin Harrison <colin.harrison-at-virgin.net> 165061Date: Wed May 9 16:54:46 2007 +0100 165062 165063 Migrate some code to the new mi apis 165064 165065commit 021e5df85d7c9373a2fed55512751d16e08128db 165066Author: David Nusinow <dnusinow@debian.org> 165067Date: Mon May 7 21:03:40 2007 -0400 165068 165069 Add more informative logging for module default loading 165070 165071 When the modules section is parsed, if a module is set to be loaded by 165072 default, this will be logged. If it is redundantly specified in xorg.conf, 165073 this will also be noted. None of this logging will happen if the xorg.conf 165074 lacks a modules section. 165075 165076commit 1b3a0508a7aee1c7b14cd62216b4727fcc9181d4 165077Author: Jesse Barnes <jbarnes@jbarnes-mobile.amr.corp.intel.com> 165078Date: Sun May 6 01:30:59 2007 -0700 165079 165080 Fix documentation for Copy hook -- it can copy memory to the scanout 165081 buffer too. 165082 165083commit 030a578391c634bc68add6ada3f251cf3f8c3069 165084Author: David Nusinow <dnusinow@debian.org> 165085Date: Thu May 3 22:51:07 2007 -0400 165086 165087 Provide UseDefaultFontPath option 165088 165089 This provides a new option, UseDefaultFontPath. This option is enabled by 165090 default, and causes the X server to always append the default font path 165091 (defined at compile time) to the font path for the server. This will allow 165092 people to specify additional font paths if they want without breaking 165093 their font path, thus hopefully avoiding ye olde "fixed front" problem. 165094 165095 Because this option is a ServerFlag option, the ServerFlags need to be 165096 processed before the files section of the config file, so swap the order 165097 that they are processed. 165098 165099commit e91b9ddc7aa95abc2d4d314e8db204860771a099 165100Author: David Nusinow <dnusinow@debian.org> 165101Date: Thu May 3 22:00:23 2007 -0400 165102 165103 Improve modules loading defaults 165104 165105 Provide default modules that may be overrided easily. Previously the 165106 server would load a set of default modules, but only if none were 165107 specified in the xorg.conf, or if you didn't have a xorg.conf at all. This 165108 patch provides a default set and you can add only the "Load" instructions 165109 to xorg.conf that you want without losing the defaults. Similarly, if you 165110 don't want to load a module that's loaded by default, you can add "Disable 165111 modulename" to your xorg.conf (see man xorg.conf in this release for 165112 details). This allows for a minimal "Modules" section, where the user only 165113 need specify what they want to be different. See bug #10541 for more. 165114 165115 The list of default modules is taken from the set loaded by default when 165116 there was a xorg.conf containing no "Modules" section. 165117 165118 A potential problem for some users is that some users disable a module, 165119 most notably DRI, by commenting out the "Load" line in their xorg.conf. 165120 This needs to be changed to an uncommented "Disable" line, as DRI is 165121 loaded by default. 165122 165123commit d2f813f7db157fc83abc4b3726821c36ee7e40b1 165124Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165125Date: Wed May 2 19:10:22 2007 -0400 165126 165127 New fbWalkCompositeRegion() function 165128 165129 This new function walks the composite region and calls a rectangle 165130 compositing function on each compositing rectangle. Previously there 165131 were buggy duplicates of this code in fbcompose.c and 165132 miext/rootles/safealpha/safeAlphaPicture.c. 165133 165134commit e0959adcd8df2c61e98e76e708fceef9c7cd54eb 165135Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165136Date: Tue May 1 13:41:48 2007 -0400 165137 165138 Add fbCompositeRect() as another special case in the switch of doom in fbpict.c 165139 165140 This is phase one of getting the two region walkers in fbcompose.c and 165141 fbpict.c merged together. 165142 165143commit cb22bdbe678a9948eda75d464d6701a729664ef0 165144Author: Peter Hutterer <peter@cs.unisa.edu.au> 165145Date: Thu May 3 10:05:58 2007 +0930 165146 165147 Add selected RamDac functions to xfree86LookupTab. 165148 165149 Adding those makes drivers for s3 and tga work again. 165150 165151commit c1e1d6b98a6708860e5b5f6e21d8d5b1d8ce9075 165152Author: Brian <brian@yutani.localnet.net> 165153Date: Wed May 2 15:55:40 2007 -0600 165154 165155 In __glXCreateARGBConfig(), insert the new GL mode at the _end_ of the linked list. 165156 165157 Previously, the new mode was added at the head of the list. This caused the 165158 positional correspondence between modes and the XMesaVisuals array to be off 165159 by one. The net result was GLX clients failing when they tried to use the 165160 last GLX mode/visual. 165161 165162 We still have the problem of DRI drivers not being able to use the extra 165163 mode/visual introduced by __glXCreateARGBConfig(). glXCreateContext fails 165164 with BadAlloc if it's attempted. This is also the source of the often- 165165 seen warning "libGL warning: 3D driver claims to not support visual xxx" 165166 Look into fixing that someday... 165167 165168commit bd0abb2844ef9faf28703e592cfebb886004234c 165169Author: Tilman Sauerbeck <tilman@code-monkey.de> 165170Date: Wed May 2 17:20:48 2007 +0200 165171 165172 Bug #10823: Fixed default OSNAME value. 165173 165174 We try to get OSNAME from uname by default now. 165175 165176commit 71fc5b3e9309182978ead676965d65ca93a4e3b9 165177Author: Keith Packard <keithp@keithp.com> 165178Date: Wed May 2 11:41:11 2007 +0200 165179 165180 Fix for a divide by zero that can be triggered by a malicious client. 165181 165182 Problem reported by Derek Abdine of rapid7.com. Thanks. 165183 165184commit c03d9a7217f9895feed8cfed3ede4752faf6187c 165185Author: Peter Hutterer <peter@cs.unisa.edu.au> 165186Date: Wed May 2 18:16:40 2007 +0930 165187 165188 Cleaning up a bit. 165189 165190 Register correct event in EventSwapVector. 165191 Fix up event swap for GE events, register XI's swap function at GE. 165192 165193commit e6fd4a24ebd205013b41e44aacbbfb847709d2fd 165194Author: Peter Hutterer <peter@cs.unisa.edu.au> 165195Date: Wed May 2 17:49:20 2007 +0930 165196 165197 Add handling for FakeDeviceEvent request. 165198 Fix a stupid bug from last commit, mask names were wrong. 165199 165200commit af25720598bb35d9d953ac8d9a07528289ebfb4a 165201Author: Peter Hutterer <peter@cs.unisa.edu.au> 165202Date: Wed May 2 13:51:16 2007 +0930 165203 165204 Add missing files for ProcXiSelectEvent. 165205 165206 Should have been a part of 5e439109292e54b5c4d1a7bc7b6ac0e42ee285f7. Oh well. 165207 165208commit 58c0fd29272fb6c2f193cff82fb0b0573c607eec 165209Author: Peter Hutterer <peter@cs.unisa.edu.au> 165210Date: Wed May 2 13:50:09 2007 +0930 165211 165212 Adjust EQ to support long events. 165213 Add RawDeviceEvent (pointers only for now). 165214 165215 This commit changes the event queue to use EventLists instead of xEvent 165216 arrays. Only EQ is affected, event delivery still uses xEvent* (look for 165217 comment in mieqProcessInputEvent). 165218 165219 RawDeviceEvents deliver driver information to the client, without clipping or 165220 acceleration. 165221 165222commit 0214d0b96a9b900a8cf5c7f2183add3a411441f3 165223Author: Peter Hutterer <peter@cs.unisa.edu.au> 165224Date: Tue May 1 15:49:41 2007 +0930 165225 165226 Start devices after windows are initialized. 165227 165228 This fixes the hotplug segfault introduced with the multihead changes and 165229 cleans up the code a bit as well. 165230 165231commit 11d0e2109b699714055e263e8963f2c39eded28b 165232Author: Peter Hutterer <peter@cs.unisa.edu.au> 165233Date: Tue May 1 14:16:36 2007 +0930 165234 165235 Correct enterleave semaphore value for root window. 165236 165237commit 8b42ba64e169a35df2a90528dcb0cd06be0f67a1 165238Author: Peter Hutterer <peter@cs.unisa.edu.au> 165239Date: Tue May 1 12:12:58 2007 +0930 165240 165241 Send PointerKeyboardPairingChangedNotifyEvent using XGE. 165242 165243 Needs updated inputproto and libXi. 165244 165245commit 325380adb20ec2b82e176d75599fd4bc97a3b918 165246Author: Peter Hutterer <peter@cs.unisa.edu.au> 165247Date: Tue May 1 11:02:05 2007 +0930 165248 165249 Check and re-set paired devices when initializing sprites. 165250 165251 If we don't do this, a device that is paired before a sprite has been 165252 initialized for the paired device will not actually get the right sprite and 165253 segfault the server on focus events. Happens for the VCK. 165254 165255commit b043a184509cfe5c4c4691ecde1a4a065e53d575 165256Author: Peter Hutterer <peter@cs.unisa.edu.au> 165257Date: Tue May 1 09:54:33 2007 +0930 165258 165259 Check VCP's and VCK's ids when searching for a new device id. 165260 165261 Both VCP and VCK are not in the inputInfo.devices list anymore, so we need to 165262 check them separately. If we don't do this, we end up re-using ids 0 and 1, 165263 causing all sorts of grief for clients. 165264 165265commit 873ef75b1e8c94d39670c981c4d830ab8bcc018b 165266Author: Colin Guthrie <gmane@colin.guthr.ie> 165267Date: Mon Apr 30 10:33:12 2007 -0600 165268 165269 fix __glXErrorCallBack() proto 165270 165271commit 6b33459bf5aac23c3ecc7002d091c02f327d907a 165272Merge: 18252a515 3c91a993e 165273Author: Brian <brian@yutani.localnet.net> 165274Date: Mon Apr 30 10:26:19 2007 -0600 165275 165276 Merge branch 'master' of git+ssh://brianp@git.freedesktop.org/git/xorg/xserver 165277 165278commit 5e439109292e54b5c4d1a7bc7b6ac0e42ee285f7 165279Author: Peter Hutterer <peter@cs.unisa.edu.au> 165280Date: Mon Apr 30 12:57:42 2007 +0930 165281 165282 Add GenericEvent extension to Xext. 165283 165284 This adds (unconditional) support for the GE extension. Anything from now on 165285 that sends events in MPX will have to use the GE extension. No GE, no MPX 165286 events. GE is not actually used yet from anywhere with this commit. 165287 165288 You will need to update x11proto, xextproto, libX11, libXext and xcb to the 165289 matching xge branches. Things will _NOT_ work without the updated protocol 165290 headers and libraries. 165291 165292commit 3c91a993e8c752002adf85c317216e1487c20780 165293Author: Michel Dänzer <michel@tungstengraphics.com> 165294Date: Sun Apr 29 23:49:41 2007 +0200 165295 165296 EXA: Fix OffscreenValidate build with DEBUG_OFFSCREEN enabled. 165297 165298commit a261e1325057974d58440812b93c00c0caa4423a 165299Author: Michel Dänzer <michel@tungstengraphics.com> 165300Date: Sun Apr 29 23:49:35 2007 +0200 165301 165302 EXA: Remove DrawableDirty. 165303 165304 Convert the remaining callers to PixmapDirty. 165305 165306commit b1b6674a919943a8ac37e54d02e8d0d23a642b1d 165307Author: Michel Dänzer <michel@tungstengraphics.com> 165308Date: Sun Apr 29 23:49:28 2007 +0200 165309 165310 EXA: FillSpans improvements. 165311 165312 * Don't need to track damage. 165313 * Always migrate for fallbacks. 165314 165315commit 584697a2231ac782f362a925e1489c15483a8791 165316Author: Michel Dänzer <michel@tungstengraphics.com> 165317Date: Sun Apr 29 23:49:09 2007 +0200 165318 165319 EXA: SolidBoxClipped improvements. 165320 165321 * Centralize handling of fallbacks and damage tracking. 165322 * Always migrate for fallbacks. 165323 165324commit 982d7c2c0b948ba04c8eefa475d660981e6ed9f9 165325Author: Michel Dänzer <michel@tungstengraphics.com> 165326Date: Sun Apr 29 23:48:59 2007 +0200 165327 165328 EXA: CopyNtoN improvements. 165329 165330 * Centralize handling of fallbacks and damage tracking. 165331 * Always migrate for fallbacks. 165332 165333commit d2245386eed200e77a8c84bdda36ab29e39fd593 165334Author: Michel Dänzer <michel@tungstengraphics.com> 165335Date: Sun Apr 29 23:48:31 2007 +0200 165336 165337 EXA: GetImage improvements. 165338 165339 Only migrate when appropriate. In particular, don't migrate to offscreen in the 165340 no-fallback case as copying from system memory should usually be as fast if not 165341 faster than DownloadFromScreen, in particular if the bits need to be uploaded 165342 to offscreen first. 165343 165344commit 0880aaac9c83019fec2e3d32871f74c7a407f8b3 165345Author: Michel Dänzer <michel@tungstengraphics.com> 165346Date: Sun Apr 29 23:48:19 2007 +0200 165347 165348 EXA: PutImage improvements. 165349 165350 * Migrate for fallbacks when appropriate. 165351 * Add damage tracking in ExaCheckPutImage. 165352 165353commit 7fca16901187ade48e83e6a2684ef464b1912357 165354Author: Michel Dänzer <michel@tungstengraphics.com> 165355Date: Sun Apr 29 23:48:11 2007 +0200 165356 165357 EXA: ImageGlyphBlt improvements. 165358 165359 * Don't waste effort on invisible glyphs. 165360 * Only track damage for bounding box instead of each glyph separately. 165361 * Always migrate for fallbacks. 165362 165363commit a8d6ebdf9338dc2f6ff9a532e6fec460a70d3b1e 165364Author: Michel Dänzer <michel@tungstengraphics.com> 165365Date: Sun Apr 29 23:47:53 2007 +0200 165366 165367 EXA: Defer to FillRegionTiled in Composite when possible. 165368 165369 Committed separately as this case is hard to hit and has only been tested 165370 lightly. 165371 165372commit 81b055605c34b5823f6c5f63cc0f92f43c6b7252 165373Author: Michel Dänzer <michel@tungstengraphics.com> 165374Date: Sun Apr 29 23:47:43 2007 +0200 165375 165376 EXA: Composite improvements. 165377 165378 * Defer to simpler hooks in more cases (inspired by XAA behaviour). 165379 * Move damage tracking from lower to higher level functions. 165380 * Always migrate for fallbacks. 165381 165382commit ce317a5b76c053f449122c46e1372bf8e067cb4c 165383Author: Michel Dänzer <michel@tungstengraphics.com> 165384Date: Sun Apr 29 23:47:16 2007 +0200 165385 165386 EXA: Glyphs improvements. 165387 165388 * Don't waste effort on invisible glyphs. 165389 * Add damage tracking where necessary. 165390 * Always migrate for fallbacks. 165391 165392commit 0c8905ebc91cf654facef84ee52231a358deec5c 165393Author: Michel Dänzer <michel@tungstengraphics.com> 165394Date: Sun Apr 29 23:47:08 2007 +0200 165395 165396 EXA: PolyFillRect improvements. 165397 165398 * Convert rects to region and use it for damage tracking. 165399 * When possible, defer to exaFillRegion{Solid,Tiled} using converted region. 165400 * Always migrate for fallbacks. 165401 * Move damage tracking out of ExaCheckPolyFillRect. 165402 165403commit 567f18a09bfb05f448be40c7ebe0f210f955601c 165404Author: Michel Dänzer <michel@tungstengraphics.com> 165405Date: Sun Apr 29 23:46:49 2007 +0200 165406 165407 EXA: FillRegion{Solid,Tiled} improvements. 165408 165409 * Support planemasks, different ALUs and arbitrary tile origin. 165410 * Leave damage tracking and non-trivial fallbacks to callers. 165411 * Always migrate for fallbacks. 165412 165413 This is in preparation for using these from more other functions. 165414 165415commit e869573b52fac69fb88cea120daaeec59c7a3461 165416Author: Michel Dänzer <michel@tungstengraphics.com> 165417Date: Sun Apr 29 23:45:48 2007 +0200 165418 165419 EXA: exaAssertNotDirty improvements. 165420 165421 * Return early if the valid region is empty or the pixmap is pinned. 165422 * Fix loop for several cliprects. 165423 165424commit d3f8667341bfe6dc7d0258c4ad69377f37d88d95 165425Author: Michel Dänzer <michel@tungstengraphics.com> 165426Date: Sun Apr 29 23:44:27 2007 +0200 165427 165428 EXA: Fix exaEnableDisableFBAccess for nested disables and enables. 165429 165430commit 5e4b3232dafe3b0dec65bf639bebaba4774210b7 165431Author: Michel Dänzer <michel@tungstengraphics.com> 165432Date: Sun Apr 29 23:38:22 2007 +0200 165433 165434 Fix fbCompositeTrans_0888xnx0888 build for wfb on big endian. 165435 165436commit 2866e0bac9b8dd3892c5e68abcfc6c97cebaf88a 165437Author: Michel Dänzer <michel@tungstengraphics.com> 165438Date: Sun Apr 29 23:38:13 2007 +0200 165439 165440 Fix a couple of picture repeat fields incorrectly compared to RepeatNormal. 165441 165442commit 78a20455356ccc310f73cfc65ad65a7677eee7e5 165443Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165444Date: Fri Apr 27 15:20:24 2007 -0400 165445 165446 Pixman merging 165447 165448 More msvc++ porting 165449 165450commit 6c4f1826bf2c5f30f5fe6e489a02b6375478b380 165451Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165452Date: Fri Apr 27 08:13:08 2007 -0400 165453 165454 Bug fix in fbCompositeIn_nx8x8888 165455 165456 Make sure both halves of the dst word is set to zero when the masks 165457 are both 0. 165458 165459commit f28eea0647f007c2e2415ecc6fceef46201faad4 165460Merge: 339b73e71 ae04f2cb0 165461Author: Peter Hutterer <peter@cs.unisa.edu.au> 165462Date: Fri Apr 27 16:34:36 2007 +0930 165463 165464 Merge branch 'master' into mpx 165465 165466 Conflicts: 165467 165468 dix/devices.c 165469 dix/events.c 165470 165471commit 339b73e710a0920608a3fbcb20b406f0f6c4e0f6 165472Author: Peter Hutterer <peter@cs.unisa.edu.au> 165473Date: Fri Apr 27 13:24:27 2007 +0930 165474 165475 Allow events to grabWindows event if the device is not grabbed. 165476 165477 This kinda makes popup windows useable if the WM doesn't set the 165478 ClientPointer. Kinda. 165479 165480commit ae04f2cb0a068cdc1e519627bf745de0c9e4a85a 165481Author: Adam Jackson <ajax@benzedrine.nwnk.net> 165482Date: Thu Apr 26 15:28:04 2007 -0400 165483 165484 Fix the 'relink' target for kdrive servers. 165485 165486commit 6c8152d6ee9eeb21a68a8bbfed1540939e5bcd1f 165487Author: Adam Jackson <ajax@benzedrine.nwnk.net> 165488Date: Thu Apr 26 14:59:04 2007 -0400 165489 165490 Remove old edid_modes.c, it lives in hw/xfree86/modes/ now. 165491 165492commit 2208c6087d6bffcb24a30891a56430e28735874c 165493Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165494Date: Thu Apr 26 14:40:30 2007 -0400 165495 165496 Change expand_alpha_rev to expand_alpha in mmxSaturateU 165497 165498commit a300ef84cee26febfbe08c497d0d063588130bdd 165499Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165500Date: Thu Apr 26 14:37:53 2007 -0400 165501 165502 Fix typo in previous commit 165503 165504commit 0ebe48be59368b55c618f60d4656300bd7f52ed9 165505Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165506Date: Thu Apr 26 14:36:32 2007 -0400 165507 165508 Pixman merge 165509 165510 - Changes to support MS Visual C++ 165511 165512 - use inline instead of __inline__ 165513 165514 - Fix rounding errors (Billy Biggs, from xserver via pixman) 165515 165516commit 4fe918b38553133c27e5ae672e5c43984a9bbaea 165517Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165518Date: Thu Apr 26 12:41:01 2007 -0400 165519 165520 Fix bug 8871 - scrolling corruption with a compositing manager 165521 165522 Call miHandleExposures() in CopyArea/CopyPlane explicitly in cw to 165523 generate GraphicsExposes correctly. 165524 165525commit 0ff7c94fcf6497ee8575f81cf97eeeb3a857739e 165526Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165527Date: Thu Apr 26 10:56:02 2007 -0400 165528 165529 Pixman merge 165530 165531 Make use of fbCompositeSrcAdd_8888x8x8mmx 165532 165533commit 701ccb4a22cfd646ccb7f19b7b3a476aeb5ce2da 165534Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165535Date: Thu Apr 26 10:49:06 2007 -0400 165536 165537 Pixman merge 165538 165539 - Remove stray default label 165540 165541 - Integrate new MMX ops SolidMaskSrc_nx8x8888mmx, In_8x8mmx, and 165542 In_nx8x8mmx 165543 165544 - Formatting changes to reduce diff noise 165545 165546commit a54ef54db19dcd36ed86b33cff2bc369f9690a15 165547Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165548Date: Thu Apr 26 10:24:25 2007 -0400 165549 165550 Pixman merge 165551 165552 Make sure fbCompositeSrc_x888x8x8888mmx and fbCompositeSrc_8888x8x8888mmx 165553 are used when possible. 165554 165555commit 3571b8e65b0857322d12e291305cfe29ea497c3c 165556Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165557Date: Thu Apr 26 09:45:11 2007 -0400 165558 165559 More pixman merging 165560 165561 - Move some code around to minimize diff noise 165562 165563 - Add mmx ops from pixman that never made it into X 165564 165565 - Merge Jeff Muizelaar's bugfixes to fbCompositeSrc_8888x8x8888mmx and 165566 fbCompositeSrc_x888x8x8888mmx. 165567 165568commit cfc01115af4136b2dad8218ba6b389513a356a2e 165569Author: Peter Hutterer <peter@cs.unisa.edu.au> 165570Date: Thu Apr 26 22:18:35 2007 +0930 165571 165572 Fix cursor rendering for multi-head. 165573 165574 Before putting anything on the screen, check if the GC was made for the 165575 ScreenRec we want to render to. If not, toss the GC and create a new one. This 165576 is not the best solution but it does the job for now. Same thing for ARGB 165577 cursors except that it's even uglier. 165578 165579 Also remember the screen the cursor was rendered to and check for the right 165580 screen in the BlockHandler, SourceValidate and a few others. Only remove or 165581 restore the cursor if we are rendering to the same screen, otherwise we get 165582 artefacts that are both funky and really annoying. 165583 165584commit 82f97e1c0cc15b050edc82a8f3b9a423d6cf5fe7 165585Author: Paulo Ricardo Zanoni <prz05@c3sl.ufpr.br> 165586Date: Thu Apr 26 15:58:50 2007 +0930 165587 165588 Enable event delivery for multiple heads. 165589 165590 Requires moving the spriteTrace into the DeviceIntRec and adjusting a few 165591 functions to take in device argument, most notably XYToWindow(). 165592 165593 Cursor rendering on the second screen is busted. 165594 165595commit c0346e57e6d3857994f7af76060c502c2fdea294 165596Author: Peter Hutterer <peter@cs.unisa.edu.au> 165597Date: Thu Apr 26 12:02:45 2007 +0930 165598 165599 Require inputproto 1.4.2. 165600 165601 Requirement was introduced with c7e2ba0c9b9b1fc1aed8f91f86471c4c8e650b78. 165602 165603commit 67347739b0571b2978468e8088480b105f505ad2 165604Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165605Date: Wed Apr 25 14:19:39 2007 -0400 165606 165607 Don't treat convolution filters as transformations. 165608 165609 Some rearrangement of code to get it closer to pixman. 165610 165611commit c056ce95d89ef1df57edf47149fc34cd3925496e 165612Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165613Date: Wed Apr 25 13:21:47 2007 -0400 165614 165615 Port MSVC++ CPU detection code from pixman. (Vladimir Vukicevic). 165616 165617commit c19ece1d8c32dc81740a4036a642661f54064e75 165618Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165619Date: Wed Apr 25 12:34:19 2007 -0400 165620 165621 Integrate optimization from xserver from David Reveman where repeats 165622 165623 get handled by fbFetchTransformed() rather than in the region walking 165624 code. 165625 165626commit 48c73dfc369fdf8f6023436ebe82bb604f76bb80 165627Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165628Date: Wed Apr 25 12:09:22 2007 -0400 165629 165630 Add function fbCompositeSrcAdd_8888x8x8(), and fix a bug where 165631 165632 srcRepeat = FALSE would be set in the wrong place. 165633 165634commit 66ba3d758a368bf83d75bab8b08bdb6b34925e40 165635Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165636Date: Wed Apr 25 10:31:38 2007 -0400 165637 165638 Various fixes from xserver via pixman (Billy Biggs) 165639 165640commit c09e68ce30dabd6b7068b163b9d2382d85d0d0bc 165641Author: Adam Jackson <ajax@benzedrine.nwnk.net> 165642Date: Wed Apr 25 16:46:26 2007 -0400 165643 165644 Paper over a crash at exit during GLX teardown. 165645 165646commit 9c80eda826448822328bb678a7d284cc43fffb17 165647Author: Adam Jackson <ajax@benzedrine.nwnk.net> 165648Date: Wed Apr 25 16:35:04 2007 -0400 165649 165650 Disable RANDR's fake Xinerama protocol when there's more than one screen. 165651 165652 ... in the protocol sense. Xinerama doesn't have any provision for more 165653 than one protocol screen each with its own geometry. 165654 165655 Red Hat bug #231257. 165656 165657commit d322608dc929d5f8cda07a53143a4f28423e0460 165658Author: Adam Jackson <ajax@benzedrine.nwnk.net> 165659Date: Wed Apr 25 16:29:48 2007 -0400 165660 165661 Fix a buffer overrun on machines with excessively large PCI busses. 165662 165663 Formerly we sized an array with a compile time constant, then initialized 165664 its size to the same constant, but the Linux PCI init code would increase 165665 that "constant". So if you happened to have more than 128 PCI devices, 165666 you'd happily scribble into whatever variables happened to be in .bss 165667 after that array. 165668 165669 Only really fixed for Linux atm. Other OSes will simply (still) fail to 165670 work on video devices above the 128th PCI device. 165671 165672commit 3ba1e8ab6d69566e1a3f8f0eb4605631aeffc8e5 165673Author: Aaron Plattner <aplattner@nvidia.com> 165674Date: Tue Apr 24 17:20:14 2007 -0700 165675 165676 Include xf86Rename.h in xf86RandR12.h. 165677 165678commit 0a2fe443d25b1ca25349aba3f748df986952e20f 165679Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165680Date: Tue Apr 24 19:02:44 2007 -0400 165681 165682 Use READ/WRITE macros for new functions introduced in previous commits. 165683 165684commit 7e16da7b78c422f96387502b9cc29eaa1741543f 165685Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165686Date: Tue Apr 24 18:15:34 2007 -0400 165687 165688 Remove #if 0'ed blocks 165689 165690commit 18252a515d4989b983a3b7389636045e06d0f246 165691Author: Brian <brian@yutani.localnet.net> 165692Date: Tue Apr 24 14:10:09 2007 -0600 165693 165694 bump release date to reflect input code updates 165695 165696commit 2d9a7a768747ca39a800475f12c424c298018dc6 165697Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165698Date: Tue Apr 24 14:46:59 2007 -0400 165699 165700 From pixman (Jeff Muizelaar) 165701 165702 Fix up the fast-path compositing operators; those are useful for 165703 sources without alpha, but can't be used for sources with 165704 alpha. Also, replaced fbCompositeSrcSrc_nxn with call to fbBlt as 165705 this function must handle 1, 4, 8, 16, 24, 32 bpp objects. Would 165706 be nice to optimize fbBlt for common cases involving 8, 16, 24 and 165707 32bpp. 165708 165709 From Keith Packard. 165710 165711commit fde4a5adf02d3067a064ebf6bdd666aa5784cfe9 165712Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165713Date: Tue Apr 24 13:30:43 2007 -0400 165714 165715 From xserver via pixman (Jeff Muizelaar) 165716 165717 Add some optimizations from jaymz. Also adds some compile 165718 warnings that will hopefully go awa y as we continue merging. 165719 165720commit 13e1d5ea55b0a3b7729316c8e37d3d8fca2075b5 165721Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165722Date: Tue Apr 24 12:59:18 2007 -0400 165723 165724 Fix format vs formatCode in previous commit 165725 165726commit 077a5d4555676d5775e990468a697b6890c6d609 165727Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165728Date: Tue Apr 24 12:57:55 2007 -0400 165729 165730 Add functions fbCompositeSrcSrc_nxn() and fbCompositeTrans_0565xnx0565 165731 from xserver via pixman. Add READ/WRITE and fbFinishAccess as 165732 appropriate. 165733 165734commit 09436fb7c38a9819bde770c4c21143591671c4d7 165735Author: Peter Hutterer <peter@cs.unisa.edu.au> 165736Date: Tue Apr 24 22:52:33 2007 +0930 165737 165738 Disable devices before removing, remove unrecoverable devices. 165739 165740commit c7e2ba0c9b9b1fc1aed8f91f86471c4c8e650b78 165741Author: Peter Hutterer <peter@cs.unisa.edu.au> 165742Date: Tue Apr 24 21:34:47 2007 +0930 165743 165744 Use DevicePresence events to tell the client about enabled/disabled devices. 165745 165746 Include the device id in the event sent to the client. 165747 165748commit ce099a9b78195540ec251a6a3dbe26019c1a686d 165749Author: Brian <brian@yutani.localnet.net> 165750Date: Mon Apr 23 12:34:01 2007 -0600 165751 165752 fix bug in which maxKeysPerModifier wasn't getting set 165753 165754commit b5e1f7869b2f12a1c2baa7f699ae609fc9ad50aa 165755Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165756Date: Mon Apr 23 14:16:30 2007 -0400 165757 165758 Remove #if 0'ed merge leftovers 165759 165760commit 84838268b34661d598f8e4856fab355f414930d9 165761Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165762Date: Mon Apr 23 13:19:54 2007 -0400 165763 165764 Gradient fixes 165765 165766 * Port fix for bug 7685 from pixman. Patch by Carl Worth 165767 165768 * Add projective version of radial gradient code. 165769 165770 * Make sure that all Pict*Gradient types have PictGradient as prefix, 165771 since code in various places relies on that. 165772 165773commit 38d14e858980a1b0c087344d24bf6aebf755663c 165774Author: Aaron Plattner <aplattner@nvidia.com> 165775Date: Sun Apr 22 18:04:27 2007 -0700 165776 165777 Adjust the screen pixmap's dimensions in xf86RandR12ScreenSetSize. 165778 165779commit ca784df84e07227a4cc0a1add079884f557b7a00 165780Author: Aaron Plattner <aplattner@nvidia.com> 165781Date: Sun Apr 22 16:26:01 2007 -0700 165782 165783 Fix unbalanced fbGetDrawable added in commit 0a9239ec. 165784 165785commit 55bd8668e7d4100579bcd6c16a804d9f25267070 165786Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165787Date: Fri Apr 20 14:53:37 2007 -0400 165788 165789 Remove #if 0'ed leftovers from merge 165790 165791commit 41dd7ab067adde8f66cd9f74c5a6570c325518a5 165792Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165793Date: Fri Apr 20 14:51:40 2007 -0400 165794 165795 Fix gradient walker to not reset needlessly 165796 165797 Previously the gradient walker was doing excessive resets, (such 165798 as on every pixel in constant-colored regions or outside the 165799 gradient with CAIRO_EXTEND_NONE). Don't do that. 165800 165801 Carl Worth, from pixman 165802 165803commit c1b73f0f2acd56b423b91a04f1e1b3cdcad0069f 165804Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165805Date: Fri Apr 20 14:34:13 2007 -0400 165806 165807 Fixing gradient repeat mode computations in previous patch. From David 165808 165809 Turner. 165810 165811commit 38f718799c68995c2d9a1680355bd55fd925009e 165812Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165813Date: Fri Apr 20 13:59:11 2007 -0400 165814 165815 Remove a few memory references in fbFetchTransformed 165816 165817commit 9c4b14d4f6a1fe018acd64789434216cd1560a4a 165818Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165819Date: Fri Apr 20 13:23:58 2007 -0400 165820 165821 Integrate David Turner's gradient optimizations from pixman 165822 165823commit cd2c1714eb4946bf7b1fc194fe074f8024a2ec23 165824Author: Brian <brian@yutani.localnet.net> 165825Date: Sat Apr 21 12:40:51 2007 -0600 165826 165827 add slang_mem.c 165828 165829commit 67545333ec0b08db783e94e9e3ec55873dea19a3 165830Author: Brian <brian@yutani.localnet.net> 165831Date: Sat Apr 21 12:40:33 2007 -0600 165832 165833 replace occlude.c w/ queryobj.c 165834 165835commit 39bc8bb0fdc854dcf9bbc0857fec84d50fa4f3b2 165836Author: Aaron Plattner <aplattner@nvidia.com> 165837Date: Fri Apr 20 14:22:42 2007 -0700 165838 165839 Don't call xf86RandR12TellChanged if it doesn't exist. Add some exports to xf86Rename.h. 165840 165841commit 3daa5c1a991d659b1386a09e33b044470d489cb3 165842Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165843Date: Fri Apr 20 09:43:16 2007 -0400 165844 165845 Fix pixman bug 5777, patch by David Reveman 165846 165847commit 9c2e955f6792e80fb84f848ed9e6ebbfd79f7130 165848Author: Brian <brian@yutani.localnet.net> 165849Date: Fri Apr 20 07:21:19 2007 -0600 165850 165851 regenerated to add GL_CLIENT_ATTRIB_STACK_DEPTH (bug 9823) 165852 165853commit 96ef0f78438b60436c3940817980a3ab4070c7e8 165854Author: Keith Packard <keithp@neko.keithp.com> 165855Date: Thu Apr 19 17:39:51 2007 -0700 165856 165857 Disable SourceValidate in rotation to capture cursor. 165858 165859 SourceValidate is used exclusively by the software cursor code to pull the 165860 cursor off of the screen before using the screen as a source operand. This 165861 eliminates the software cursor from the frame buffer while painting the 165862 rotated image though. Disabling this function by temporarily setting the 165863 screen function pointer to NULL causes the cursor image to be captured. 165864 (cherry picked from commit 05e1c45ade9c558820685bfd2541617a2e8de816) 165865 165866commit 7ca4baffb5569ea12b578a4a3f69e93d272d6c6d 165867Author: Keith Packard <keithp@neko.keithp.com> 165868Date: Thu Apr 19 17:37:18 2007 -0700 165869 165870 Was accidentally disabling rotation updates in mode set. 165871 165872 Setting a mode on an unrotated CRTC was causing all of the rotation updates 165873 to be disabled; the loop looking for active rotation wasn't actually looking 165874 at each crtc, it was looking at the modified crtc many times. 165875 (cherry picked from commit 8b217dee3a6c46b13fc9571a4a9a95bc55686cdb) 165876 165877commit 806a537e644d8cc9e53f3ac52efb49453e5aa1fb 165878Author: Keith Packard <keithp@neko.keithp.com> 165879Date: Thu Apr 19 17:49:34 2007 -0700 165880 165881 Revert "Suppress software cursor removal during rotated shadow buffer drawing." 165882 165883 This reverts commit 999b681cf3973af4191506e49cde06963b11a774. 165884 Replacing this with simpler code that just disables SourceValidate 165885 during rotation redisplay. 165886 165887commit 0a9239ec258828ec1da6c208634a55fc4053d7da 165888Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 165889Date: Thu Apr 19 18:19:34 2007 -0400 165890 165891 Merge David Reveman's gradient optimization patch from pixman 165892 165893commit d0e55774e0da641ba85c5173f27f68de27372747 165894Author: Thomas Hellstrom <thomas-at-tungstengraphics-dot-com> 165895Date: Thu Apr 19 11:39:53 2007 +0200 165896 165897 libdri: Make sure the new DRIInfo keepFDOpen member is honoured. 165898 165899commit e1f0b3e70b696d7ea4cf9e6ed30d751e7fdbc577 165900Author: Peter Hutterer <peter@cs.unisa.edu.au> 165901Date: Thu Apr 19 12:00:24 2007 +0930 165902 165903 config: Return errors as negative numbers, device ids as positive numbers. 165904 Update dbus-api documentation. 165905 165906commit c6972c893359f8fa7631ae674330f3f4f7010ba0 165907Author: Peter Hutterer <peter@cs.unisa.edu.au> 165908Date: Wed Apr 18 12:10:05 2007 +0930 165909 165910 Change dbus 'listDevices' call to not require an argument. 165911 Update dbus-api documentation, plug memory leak on dbus reply error. 165912 165913commit d61ed6c8a2823a3532439d5cb9f355129c93f523 165914Author: Peter Hutterer <peter@cs.unisa.edu.au> 165915Date: Thu Apr 19 18:08:55 2007 +0930 165916 165917 Check for NULL devices in CloseDevice(). 165918 165919commit 9c30f7422121a0443c8d612d06181e17d8af9730 165920Author: Peter Hutterer <peter@cs.unisa.edu.au> 165921Date: Thu Apr 19 12:00:24 2007 +0930 165922 165923 config: Return errors as negative numbers, device ids as positive numbers. 165924 Update dbus-api documentation. 165925 165926commit 999b681cf3973af4191506e49cde06963b11a774 165927Author: Eric Anholt <eric@anholt.net> 165928Date: Wed Apr 18 14:33:27 2007 -0700 165929 165930 Suppress software cursor removal during rotated shadow buffer drawing. 165931 165932commit 28bb34eec63bf3c98f38ba7fc044f6419aaa3307 165933Author: Eric Anholt <eric@anholt.net> 165934Date: Wed Apr 18 13:48:28 2007 -0700 165935 165936 Belatedly bump XORG_VERSION for 7.2. 165937 165938commit 5d8e8a7f4b3226bffd9e4d6d9326688f475b0183 165939Author: Eric Anholt <eric@anholt.net> 165940Date: Wed Apr 18 13:22:26 2007 -0700 165941 165942 Remove libminimi build. 165943 165944 It appears to have been a leftover of a previous incarnation of the build 165945 system that didn't handle miinitext.c well. 165946 165947commit 53fb42e65c2b2ff58a4a324b7f05cff8a587720a 165948Author: Erik Andrén <erik.andren@gmail.com> 165949Date: Tue Apr 17 21:34:47 2007 -0700 165950 165951 Syncmaster 226 monitor needs 60Hz refresh (#10545). 165952 165953 I've managed to solve my own bug (#10545) by applying the following 165954 patch to the xserver. 165955 165956 Please apply. 165957 165958 <Conspiracy mode on> 165959 This monitor is "Vista Certified". I wonder if this is a pure coincidence... 165960 <Conspiracy mode off> 165961 165962 With kind regards 165963 Erik Andrén 165964 (cherry picked from commit a63704f14a1d97b9a00fef6fa290e74e51b9732b) 165965 165966commit ae75afcb1b5419102c5be10b8826ceed50d2ef5d 165967Author: Peter Hutterer <peter@cs.unisa.edu.au> 165968Date: Wed Apr 18 12:10:05 2007 +0930 165969 165970 Change dbus 'listDevices' call to not require an argument. 165971 Update dbus-api documentation, plug memory leak on dbus reply error. 165972 165973commit 9cee4ec5e6e06d23aafb302494b082c77ade4623 165974Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 165975Date: Tue Apr 17 16:01:56 2007 -0400 165976 165977 xace: change the semantics of the return value of XACE hooks to allow 165978 arbitrary X status codes instead of just TRUE/FALSE. 165979 165980 The dix layer in most cases still does not propagate the return value of 165981 XACE hooks back to the client, however. There is more error propagation 165982 work to do. 165983 165984commit 47bd311e3dcc501cbb202ce79a55ac32e9db50f2 165985Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 165986Date: Tue Apr 17 13:46:55 2007 -0400 165987 165988 security: remove debugging code. 165989 165990commit 1f06d32ef58749d0f0c062193d237ee98f60e90f 165991Merge: cf962a849 fc162c6cf 165992Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 165993Date: Tue Apr 17 12:10:22 2007 -0400 165994 165995 Merge branch 'master' into XACE-SELINUX 165996 165997 Conflicts: 165998 165999 dix/privates.c 166000 166001commit b6aec7f6f906a18d13586d63afabf1ee4fbb11c3 166002Author: Peter Hutterer <peter@cs.unisa.edu.au> 166003Date: Tue Apr 17 16:51:16 2007 +0930 166004 166005 Change FocusIn/Out semantics to match Enter/Leave semantics. 166006 166007commit 451d5464b4e8a2516b8a4598b3c4eb14656be90e 166008Author: Peter Hutterer <peter@cs.unisa.edu.au> 166009Date: Tue Apr 17 11:17:04 2007 +0930 166010 166011 Change enter/leave semantics for events with detail Notify{Nonlinear}Virtual. 166012 166013 Core enter/leave events with detail Notify{Ancestor|Inferior|Nonlinear} are 166014 only sent for the first/last pointer to enter/leave. Events with detail 166015 Notify{Nonlinear}Virtual are sent at all times, but not to those windows that 166016 currently have one or more pointers inside their boundaries. 166017 166018commit fc162c6cfa06f0b012743d6d79cef45cf0166229 166019Author: Keith Packard <keithp@neko.keithp.com> 166020Date: Mon Apr 16 09:55:58 2007 -0700 166021 166022 Allow outputs to be explicitly enabled in config, overriding detect. 166023 166024 Option "Enable" "True" will force the server to enable an output at startup 166025 time, even if the output is not connected. This also causes the default 166026 modes to be added for this output, allowing even sync ranges to be used to 166027 pick out standard modes. 166028 (cherry picked from commit a3d73ba2cb7e13a6d129cd88d6a7f7d756e2ced2) 166029 166030commit c41e3bd713206c0bbd8ab8cef4c83eb7ba7e1c3c 166031Author: Keith Packard <keithp@neko.keithp.com> 166032Date: Mon Apr 16 09:53:42 2007 -0700 166033 166034 Use default screen monitor for one of the outputs. 166035 166036 By default, use the screen monitor section for output 0, however, a driver 166037 can change which output gets the screen monitor by calling 166038 xf86OutputUseScreenMonitor. 166039 (cherry picked from commit f4a8e54caf6b9431711383a39f55a18e7fd654f4) 166040 166041commit 97a2c2579c56c304705c934f3b536473645747df 166042Author: Keith Packard <keithp@neko.keithp.com> 166043Date: Mon Apr 16 09:39:47 2007 -0700 166044 166045 Using wrong log level in extension to built-in message 166046 166047 was: typo in built-in module log message 166048 (cherry picked from commit 00cfd1f765895b4d1b2234f3203727a8871b64b0) 166049 166050commit deda7791dfa34d0563c8d7fa2a0660ac27e6858c 166051Author: Brian <brian@yutani.localnet.net> 166052Date: Mon Apr 16 11:35:22 2007 -0600 166053 166054 remove sources deleted in Mesa 166055 166056commit 02d42f344ce020c9b84723671cb9c68d5c064933 166057Author: Thomas Hellstrom <thomas-at-tungstengraphics-dot-com> 166058Date: Mon Apr 16 17:24:53 2007 +0200 166059 166060 Changes for single-entity multi-screen DRI. 166061 166062 The entity (device) has a locking SAREA and a master file descriptor 166063 that optionally isn't closed between server generation. 166064 166065 The locking SAREA contains the device hardware lock. 166066 Each DRI screen creates an new SAREA containing the drawable lock, 166067 drawable-and private info, the drawable SAREA. 166068 166069 The first screen optionally shares its drawable SAREA with the 166070 device SAREA. 166071 166072 Default is to close the master descriptor between server generations, 166073 and to share the drawable SAREA of the first screen with the device locking 166074 SAREA. Thus we should (hopefully) have full backwards compatibility. 166075 166076 Mesa changes to support single-device multiple screens are pending. 166077 166078commit b5823ea3e1ed5a0449d44da05165a46719dcf287 166079Author: Keith Packard <keithp@neko.keithp.com> 166080Date: Sun Apr 15 22:59:19 2007 -0300 166081 166082 RandR 1.2 spec says CRTC info contains screen-relative geometry. 166083 166084 Was reporting mode size instead of adjusting for rotation. 166085 (cherry picked from commit e2e7c47a528447e90cff6cf10d2ce457742ef48d) 166086 166087commit cc4eb1c7ea1bace7ed69cfd80c99d22933282ae1 166088Author: Keith Packard <keithp@neko.keithp.com> 166089Date: Fri Apr 13 15:04:29 2007 -0300 166090 166091 Add quirk for Acer AL1706 monitor to force 60hz refresh. 166092 166093 This Acer monitor reports support for 75hz refresh via EDID, and yet when 166094 that rate is delivered, the monitor does not sync and reports out of range. 166095 Use the existing 60hz quirk for this monitor. 166096 (cherry picked from commit 1328a288e9030a472a915077160f090d1afd4126) 166097 166098commit eba81a0a01f8a61151d8bf9f3d83bda85ca26e73 166099Author: George Sapountzis <gsap7@yahoo.gr> 166100Date: Sat Apr 14 18:30:09 2007 +0300 166101 166102 glx: move __glXMesaProvider from GLcore module to glx module. 166103 166104 This treats the GLcore provider similar to DRI provider, using a subset of 166105 XMesa as the GLcore interface. 166106 166107commit 6b040b79f0e247b6f2da8f7d239443743e96de67 166108Author: George Sapountzis <gsap7@yahoo.gr> 166109Date: Sat Apr 14 18:29:52 2007 +0300 166110 166111 glx: drop xmesaP.h include from xf86glx.c 166112 166113 The declarations for the xfree86-specific XMesa functions were moved up to 166114 xmesa.h, requires Mesa as of 2007-04-13. 166115 166116commit 2c833f60acb3dc358815a99cd295ef7fc695c45d 166117Author: George Sapountzis <gsap7@yahoo.gr> 166118Date: Sat Apr 14 18:29:25 2007 +0300 166119 166120 glx: drop stray CAPI define. 166121 166122 SI imports/exports were dropped from Mesa. 166123 166124commit ea27b09d3a973d9a4943f205b24940b8624cf405 166125Author: Peter Hutterer <peter@cs.unisa.edu.au> 166126Date: Fri Apr 13 17:15:22 2007 +0930 166127 166128 Deliver FocusIn/FocusOut events for all devices. 166129 Don't use VCK in SetInputFocus and DeliverFocusedEvent. 166130 166131commit 7ccebc50b98ac175fdbdfaab081bcead62e60ee3 166132Author: Peter Hutterer <peter@cs.unisa.edu.au> 166133Date: Fri Apr 13 13:08:44 2007 +0930 166134 166135 Documentation for events.c. 166136 166137commit 2bb3c81fc1ca98c909309f3af7e9c8ca6b695657 166138Author: Peter Hutterer <peter@cs.unisa.edu.au> 166139Date: Thu Apr 12 16:35:43 2007 +0930 166140 166141 Use paired kbd instead of VCK for EnterLeaveEvents, sanitize variable naming. 166142 166143commit e7b47b1758ed20e75ee267b3a09e91bf2861f6fd 166144Author: Peter Hutterer <peter@cs.unisa.edu.au> 166145Date: Thu Apr 12 16:24:42 2007 +0930 166146 166147 Change core enter/leave semantics for multiple pointers. 166148 166149 Send EnterNotify when first device enters the window, LeaveNotify when the 166150 last device leaves the window. Additional devices will not cause additional 166151 Enter/LeaveNotifies. 166152 166153commit 547d720938b3668666d60110d79b150b1e9325c6 166154Author: Peter Hutterer <peter@cs.unisa.edu.au> 166155Date: Thu Apr 12 11:16:19 2007 +0930 166156 166157 Remove workaround for a NULL core pointer in ProcessInputEvents. 166158 166159 This workaround is obsolete with 33a5d9605e3e282f6aa1921d7321a2a12ef02c42 166160 166161commit d4dad6f84f82a4ade5005c3aa93511c1295875b8 166162Merge: f1f8b562a 33a5d9605 166163Author: Peter Hutterer <peter@cs.unisa.edu.au> 166164Date: Thu Apr 12 11:11:03 2007 +0930 166165 166166 Merge branch 'master' into mpx 166167 166168 Conflicts: 166169 166170 configure.ac 166171 dix/events.c 166172 hw/xfree86/common/xf86Xinput.c 166173 166174commit 33a5d9605e3e282f6aa1921d7321a2a12ef02c42 166175Author: Daniel Stone <daniel@fooishbar.org> 166176Date: Wed Apr 11 18:28:57 2007 +0300 166177 166178 XFree86: DGA: Don't call ProcessInputEvents from CloseScreen 166179 166180 By the time CloseScreen gets called, we can't call ProcessInputEvents, as 166181 the event queue will get unhappy. So just unregister our hooks instantly, 166182 and hope that they don't get called. 166183 166184commit 0910540e4322bba72a2fa0a907072eab2547a7b6 166185Author: Remigiusz Marcinkiewicz <enleth@enleth.com> 166186Date: Wed Apr 11 01:09:26 2007 +0300 166187 166188 Config: Extend D-BUS API 166189 166190 Return device ID where available. 166191 Add listDevices call, which does what it says on the box. 166192 166193commit aecbc712144dd1aaf462bd758821438b1d22d957 166194Author: Remigiusz Marcinkiewicz <enleth@enleth.com> 166195Date: Wed Apr 11 00:38:16 2007 +0300 166196 166197 Input: Allow a pointer to a device to be returned in NIDR 166198 166199 Allow a pointer to the first device added to be returned, so we know which 166200 device(s) were added by the NIDR call. 166201 166202commit 4f05f9591e5492c72f3856bd7a2ff13378f59f2b 166203Author: Magnus Vigerlöf <Magnus.Vigerlof@home.se> 166204Date: Tue Apr 10 23:57:48 2007 +0300 166205 166206 Input: Always add devices with first available ID 166207 166208 Scan the device list when adding a new device, and make sure we can use 166209 the first available ID, instead of always incrementing. 166210 166211commit 20674dcbb2373a0af287883bc008fb6fb23d4466 166212Author: Magnus Vigerlöf <Magnus.Vigerlof@home.se> 166213Date: Tue Apr 10 23:55:36 2007 +0300 166214 166215 Config: Fix memory leaks 166216 166217 Fix memory leaks that could occur along the error path. 166218 166219commit 82962bbae2b4fda274625d1712ef839ce1ab9dc8 166220Author: Magnus Vigerlöf <Magnus.Vigerlof@home.se> 166221Date: Tue Apr 10 23:54:32 2007 +0300 166222 166223 Input: Add DeleteInputDeviceRequest 166224 166225 Add DIDR, which asks the DDX to remove a device, analogous to 166226 NewInputDeviceRequest. Only implemented for XFree86 at the moment. 166227 166228commit 7b82a836c66ba88566255052caff63577e1a0384 166229Author: Magnus Vigerlöf <Magnus.Vigerlof@home.se> 166230Date: Tue Apr 10 23:52:08 2007 +0300 166231 166232 XFree86: Fix memory leaks, option parsing, in NewInputDeviceRequest 166233 166234 Plugged some possible memory leaks, and added some more checks on the 166235 options, particular for driver/identifier. Added an unwind. 166236 166237commit 4771fa8747791498e504d73afccfb5833499a38b 166238Author: Magnus Vigerlöf <Magnus.Vigerlof@home.se> 166239Date: Tue Apr 10 23:48:00 2007 +0300 166240 166241 XFree86: Fix memory leak in option parsing 166242 166243 Fix option parsing functions and callers thereof to not leak memory. 166244 166245commit 07c56abf84080c020a3e7b7703a447c7f996975c 166246Author: Magnus Vigerlöf <Magnus.Vigerlof@home.se> 166247Date: Tue Apr 10 23:43:58 2007 +0300 166248 166249 Input: Plug memory leak in device free 166250 166251 Remember to also free the motion history, if we're using the DIX-managed 166252 history. 166253 166254commit e92743bc9839c36914a44f3e5bc8cd85773ac794 166255Author: Daniel Stone <daniel@fooishbar.org> 166256Date: Sun Apr 8 14:02:02 2007 +0300 166257 166258 getevents: Copy modifier state from extended to core devices 166259 166260 Make core events carry the same modifier state as the extended events, so 166261 that holding down Ctrl on keyboard A and pressing Q on keyboard B won't 166262 cause your app to quit. 166263 166264commit e49f836d6fa2768cd6d2a6d0227b5dbf516013dc 166265Author: Daniel Stone <daniel@fooishbar.org> 166266Date: Sun Apr 8 13:56:41 2007 +0300 166267 166268 mieq: Use larger default queue size 166269 166270 Use a default queue size of 512 rather than 256, else Xephyr is too slow 166271 without a host cursor, so events get stuck in the queue. 166272 166273commit 4aae2de74b9224bac2b2e2522637dac09abc3837 166274Author: Jared D. McNeill <jmcneill@netbsd.org> 166275Date: Tue Apr 10 12:57:15 2007 -0700 166276 166277 Add a real xf86EnableIO/xf86DisableIO for NetBSD/PPC. 166278 166279commit f77a8ea849d171a8ca00b2b7334866ace1ffbf73 166280Author: Keith Packard <keithp@neko.keithp.com> 166281Date: Mon Apr 9 14:29:46 2007 -0700 166282 166283 Rotate screen size as needed from RandR 1.1 change requests. 166284 166285 Screen size must reflect rotated mode size when setting rotated mode using 166286 RandR 1.1 SetScreenConfig request. 166287 (cherry picked from commit efcec7dbd3c2736c7b421d29c4d37e231aa681d2) 166288 166289commit bcf17df69a232e5e84609abacdca36274316e170 166290Author: Keith Packard <keithp@neko.keithp.com> 166291Date: Mon Apr 9 14:12:27 2007 -0700 166292 166293 Disable CRTC when SetSingleMode has no matching mode. Update RandR as well. 166294 166295 xf86SetSingleMode tries to resize all crtcs to match the selected mode. When 166296 a CRTC has no matching mode, it now disables the CRTC (instead of crashing). 166297 166298 Also, poke the RandR extension when xf86SetSingleMode is done so that 166299 appropriate events can be delivered, and so that future RandR queries return 166300 correct information. 166301 (cherry picked from commit dc6c4f6989f87149d8605604f4514f5cbf11de67) 166302 166303commit 67e1c98895a566f927e1ae2384d56cfca104f971 166304Author: Adam Jackson <ajax@benzedrine.nwnk.net> 166305Date: Mon Apr 9 19:08:52 2007 -0400 166306 166307 Look for the PCI ROM file elsewhere in sysfs. 166308 166309 /sys/devices reflects the bus topology, and we don't care that much. 166310 Easier (and more reliable) to just look in /sys/bus/pci/devices, which 166311 is a flat view. 166312 166313commit a08d5157f70567a0aa9583d4a15e62437340cf34 166314Author: Adam Jackson <ajax@benzedrine.nwnk.net> 166315Date: Mon Apr 9 19:04:56 2007 -0400 166316 166317 VT activate or waitactive are fatal if they fail. 166318 166319 Also, be sure to waitactive on the way down, to make sure we're off the VT 166320 before exiting. 166321 166322commit 72b477f964c748a1ab668781643cc11877f19738 166323Author: Adam Jackson <ajax@benzedrine.nwnk.net> 166324Date: Mon Apr 9 18:59:01 2007 -0400 166325 166326 Don't write out empty sections from the parser. 166327 166328commit 1f6741db19d4c91b1eacb497dff1814acb1bf0c3 166329Author: Stefan Huehner <stefan@huehner.org> 166330Date: Mon Apr 9 14:33:15 2007 -0700 166331 166332 Bug #10560: Code-Cleanup: function declarations () -> (void) 166333 166334 X.Org Bugzilla #10560: <https://bugs.freedesktop.org/show_bug.cgi?id=10560> 166335 Patch #9511 <https://bugs.freedesktop.org/attachment.cgi?id=9511> 166336 166337commit f24391dbfd12a84253dfec794ee7884afd52e197 166338Author: Keith Packard <keithp@neko.keithp.com> 166339Date: Mon Apr 9 12:30:31 2007 -0700 166340 166341 In AIGLX EnterVT processing, invoke driver EnterVT before resuming glx. 166342 166343 As the driver EnterVT function generally re-enables the hardware and 166344 prepares it for rendering, it must be called before any gl functions are 166345 called which could touch the hardware. 166346 166347commit 4c2e28b0916b5f75cfefb6df9fa0a7a09675539a 166348Author: Keith Packard <keithp@neko.keithp.com> 166349Date: Mon Apr 9 12:28:53 2007 -0700 166350 166351 Add setrlimit call in -core option to make dumps occur. 166352 166353 Default core size limit for most environments is 0, which disables core 166354 dumps. Add code in the -core option processing path to set the core limit to 166355 the maximum value. 166356 166357commit f1f8b562aaaa6ec32ab0d0697f964d92d6d536a4 166358Author: Peter Hutterer <peter@cs.unisa.edu.au> 166359Date: Mon Apr 9 19:31:59 2007 +0930 166360 166361 Alloc sprite memory in devices' devPrivates, allow undisplaying cursors. 166362 166363 Improve memory usage by allocating the sprite's memory only to devices that 166364 actually have a sprite and provide means to remove a device's cursor from the 166365 screen (more hotplugging, yay!). 166366 This commit breaks ScreenRec's ABI. 166367 166368commit 7cef789fa13ae53bfba6dc7b5a7928b7362b2522 166369Author: Peter Hutterer <peter@cs.unisa.edu.au> 166370Date: Mon Apr 9 18:37:48 2007 +0930 166371 166372 Close down virtual core devices when closing all devices. 166373 166374commit 6081b8c76f7d934bd4e9584a2f6d55636c5289d2 166375Author: Peter Hutterer <peter@cs.unisa.edu.au> 166376Date: Mon Apr 9 18:35:47 2007 +0930 166377 166378 Don't explicitly init sprite for VCP, EnableDevice() will take care of that. 166379 166380commit 00b1d52f4103a07568dcebcdaa860b9b51f19b4d 166381Author: Peter Hutterer <peter@cs.unisa.edu.au> 166382Date: Mon Apr 9 18:32:00 2007 +0930 166383 166384 Adding some comments to devices.c. 166385 166386commit aa77ffb510abe004802ab9acc6996e4c6fe3ebb2 166387Author: Peter Hutterer <peter@cs.unisa.edu.au> 166388Date: Mon Apr 9 18:27:22 2007 +0930 166389 166390 Fix: pick new ClientPointer when device is closed. 166391 166392commit 4beeab8424774ea4c3142f29b90e33f1fc7cb154 166393Author: Daniel Stone <daniel@fooishbar.org> 166394Date: Sun Apr 8 13:39:06 2007 +0300 166395 166396 XFree86: Treat evdev and vmmouse as mouse drivers (bug #10512, #10559) 166397 166398 When we see an evdev or vmmouse section, assume that it's a mouse, and 166399 don't add a default mouse device. This will break users who have an 166400 evdev keyboard section but no mouse, and want the mouse to get added 166401 by default. 166402 166403commit 0a6ac992363343487dfe0a0fc985ea55bd448382 166404Author: Brian <brian@yutani.localnet.net> 166405Date: Sat Apr 7 12:41:57 2007 -0600 166406 166407 regenerated to fix bug 10371 166408 166409commit 7e385598613778de14c0feea0d32f17d7aa66a8e 166410Author: George Sapountzis <gsap7@yahoo.gr> 166411Date: Fri Apr 6 13:38:12 2007 +0300 166412 166413 GLcore: fix after moving xf86glx_util.[hc] to Mesa. 166414 166415commit cf962a849db2b259ca558c6265ea7e3328a4d312 166416Merge: 5ad562565 5a804f2e9 166417Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166418Date: Thu Apr 5 14:41:39 2007 -0400 166419 166420 Merge branch 'master' into XACE-SELINUX 166421 166422 Conflicts: 166423 166424 configure.ac 166425 166426commit 5ad562565ac8ef9257da3afb0de1ae4f90f80fe9 166427Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166428Date: Thu Apr 5 14:18:05 2007 -0400 166429 166430 devPrivates rework: properly free devPrivates on compatibility structures, 166431 type pixmap. Requires ddx'es to call the free function from DestroyPixmap. 166432 166433commit 5a804f2e97ab59745482660a635b801ac2b9e769 166434Author: George Sapountzis <gsap7@yahoo.gr> 166435Date: Thu Apr 5 19:14:31 2007 +0300 166436 166437 configure: minor cosmetic, move GLX extension options together. 166438 166439commit a4e2fc703484fffed8dd50c1b4b24c564be4d3cd 166440Author: George Sapountzis <gsap7@yahoo.gr> 166441Date: Thu Apr 5 19:13:47 2007 +0300 166442 166443 glx: Remove stray __GLinterface. 166444 166445 __GLinterface was droped from glcore.h 166446 166447commit 38ca7d388c47c4800c74442172d6595a9b3dfcc7 166448Author: George Sapountzis <gsap7@yahoo.gr> 166449Date: Thu Apr 5 19:13:14 2007 +0300 166450 166451 glx: fix symlink, glcontextmodes.c was moved to mesa/src/glx/ 166452 166453commit 1d550bb2c5cb5b3e588f0e0b68a421dc1cb8bd7c 166454Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166455Date: Thu Apr 5 12:12:58 2007 -0400 166456 166457 devPrivates rework: minor fix; use calloc and avoid initialization. 166458 166459commit 5c4deb71a1cb981ea7e2e25d2b3a1179f27efa5a 166460Author: Peter Hutterer <peter@cs.unisa.edu.au> 166461Date: Thu Apr 5 14:21:46 2007 +0930 166462 166463 Synthesize WarpPointer events for the actual device rather than the VCP. 166464 166465commit ed75b056511ccb429c48c6c55d14dc7ae79e75a3 166466Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166467Date: Wed Apr 4 12:00:15 2007 -0400 166468 166469 dix: add new, combined resource lookup function. Move all dix lookup API 166470 deprecated so far to a new file dix/deprecated.c. Remove the deprecation 166471 warnings for the time being. 166472 166473commit 63e46e4fc3e98751f2edbed9c79ef3d5dc2dadc6 166474Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166475Date: Wed Apr 4 15:59:51 2007 -0400 166476 166477 devPrivates rework: properly free devPrivates on compatibility structures, 166478 excluding pixmap. 166479 166480commit f8482967ae8080f49dd1bbb0b79cc65020df679f 166481Author: Thomas Hellstrom <thomas-at-tungstengraphics-dot-com> 166482Date: Wed Apr 4 12:28:48 2007 +0200 166483 166484 Add an EXA driver callback to determine whether a pixmap is 166485 166486 "offscreen" in exa terms, which means accessible to the GPU. 166487 Bump exa minor. The change is backwards-compatible. 166488 166489commit 7f36ba57062096e9c67889d5b7dd64e2fb0004e7 166490Author: Peter Hutterer <peter@cs.unisa.edu.au> 166491Date: Wed Apr 4 17:38:10 2007 +0930 166492 166493 Allow for multiple animated cursors. 166494 166495commit 676188531b2bd7679ee1831b5c517df3e6d215c5 166496Author: Peter Hutterer <peter@cs.unisa.edu.au> 166497Date: Wed Apr 4 17:35:06 2007 +0930 166498 166499 Unrealize cursor only once, not per device. 166500 166501commit c10df5b967d4da4e11786520317e2917de5541fa 166502Author: Aaron Plattner <aplattner@nvidia.com> 166503Date: Tue Apr 3 15:47:01 2007 -0700 166504 166505 Swap RRScreenChangeNotifyEvent dimensions when the screen has one crtc and it's rotated. 166506 166507 RandR 1.1 clients expect the size fields in this event to be the unrotated 166508 dimensions of the screen. This behavior is "weird", but that's the way the old 166509 code worked so we need to be bug-compatible with it. 166510 166511commit 1cb84768f376b477a08a558854609b0743f2bd29 166512Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166513Date: Tue Apr 3 15:31:16 2007 -0400 166514 166515 security: rewrite to use new devPrivates support. 166516 166517commit 14aea12cadef647369e44639ff5024dd7034570a 166518Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166519Date: Tue Apr 3 15:23:56 2007 -0400 166520 166521 xace: forgot one of the hook call arguments. Add it. 166522 166523commit 353e19fd5e18ad55a0dd12a7b63f6af9df7bfe6b 166524Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166525Date: Tue Apr 3 14:06:02 2007 -0400 166526 166527 devPrivates rework: zero out newly allocated private space. 166528 166529commit e1dea151db6405e12d991feacba9446320739ee8 166530Author: Brian <brian@yutani.localnet.net> 166531Date: Tue Apr 3 11:21:50 2007 -0600 166532 166533 Implement a minor hack in dmxCheckFunctionKeys() to detect special keys. 166534 166535 Keep track of status of (left) alt/ctrl keys so that ctrl-alt-q to exit 166536 can be detected. Not ideal, but works for now. 166537 166538commit 0ee40c935750e25a9e178cdd70f6b2c667e79344 166539Author: Brian <brian@yutani.localnet.net> 166540Date: Tue Apr 3 09:31:00 2007 -0600 166541 166542 s/intead/instead/ 166543 166544commit 23974f20bf0e0c2786cc75af026af5484f6dc331 166545Author: Brian <brian@yutani.localnet.net> 166546Date: Tue Apr 3 09:30:24 2007 -0600 166547 166548 add DDXRingBell() stub to solve link problem 166549 166550commit a240c039c47c0be22ea5e100692307b26d938747 166551Author: Brian <brian@yutani.localnet.net> 166552Date: Tue Apr 3 09:27:57 2007 -0600 166553 166554 Split the xserver/fb/fbcmap.c file into two files. 166555 166556 Now, fbcmap_mi.c contains the fb functions which just wrap mi functions. 166557 Previously, these were in fbcmap.c and compiled when XFree86Server was defined. 166558 Now, clients of fbcmap should either use fbcmap.c or fbcmap_mi.c and not worry 166559 about setting the XFree86Server symbol. 166560 166561commit 1cc8db72816cd079f30255046e10043c350bf683 166562Merge: 645d87cf8 a39f297ad 166563Author: Matthieu Herrb <matthieu@roadrock.(none)> 166564Date: Tue Apr 3 16:04:45 2007 +0200 166565 166566 Merge branch 'master' of ssh://herrb@git.freedesktop.org/git/xorg/xserver 166567 166568commit 645d87cf8ef724d4591614f9994cdc4d7549a7a8 166569Author: Matthieu Herrb <matthieu@roadrock.(none)> 166570Date: Tue Apr 3 15:47:18 2007 +0200 166571 166572 CVE-2007-1003: XC-MISC Extension ProcXCMiscGetXIDList() Memory Corruption 166573 166574commit 207f710d16337839c2427aa16fad70a49834153e 166575Author: Peter Hutterer <peter@cs.unisa.edu.au> 166576Date: Tue Apr 3 19:02:14 2007 +0930 166577 166578 Delete all grabs from all devices when deleting a window. 166579 166580commit 5f382c9c7fbda6eccf8e76c28a90b55ff2f0aef3 166581Merge: 9b0b34066 f09ee168e 166582Author: Peter Hutterer <peter@cs.unisa.edu.au> 166583Date: Tue Apr 3 15:07:18 2007 +0930 166584 166585 Merge branch 'master' into mpx 166586 166587commit f09ee168e234d0a9416e15f2916726b975cad293 166588Author: Peter Hutterer <peter@cs.unisa.edu.au> 166589Date: Tue Apr 3 15:07:00 2007 +0930 166590 166591 Revert "o fix minor error in comment for GetPointerEvents()" 166592 166593 This reverts commit 96ce17aa94413c4b8bcb61cae71167050130a307. 166594 166595commit 9b0b3406682b5a3161e6c3895771523214c37207 166596Author: Peter Hutterer <peter@cs.unisa.edu.au> 166597Date: Tue Apr 3 15:06:26 2007 +0930 166598 166599 Make GrabDevice differ between core grabs and device grabs. 166600 166601commit a39f297ada4fa87c858395ae2aacefac5f8fba05 166602Author: Keith Packard <keithp@viola.jf.intel.com> 166603Date: Mon Apr 2 14:15:36 2007 -0700 166604 166605 Don't erase current crtc for outputs on CloseScreen 166606 166607 Erasing this variable causes some outputs (SDVO on intel) to fail 166608 to be correctly reset at server reset time. 166609 (cherry picked from commit 56262a4ee943f328d089a8eb4aa70b9a4bd5d135) 166610 166611commit 11797ffdcc22160317a5ebbc9291472570a51c6d 166612Author: Eric Anholt <eric@anholt.net> 166613Date: Mon Apr 2 18:21:58 2007 -0700 166614 166615 Move modes/ debugging output under Option "ModeDebug" in the Device section. 166616 166617commit e44f106ffc796c025abdfb66717c06db8b12b4e4 166618Author: Brian <brian@yutani.localnet.net> 166619Date: Mon Apr 2 16:26:15 2007 -0600 166620 166621 clean-up, debug code 166622 166623commit 8d8bc8927760fad631bef83fa2841b455ff6d511 166624Author: Brian <brian@yutani.localnet.net> 166625Date: Mon Apr 2 16:21:57 2007 -0600 166626 166627 fix formatting 166628 166629commit 3e482de7b145a5eed79b81c30c359fe43647824a 166630Author: Brian <brian@yutani.localnet.net> 166631Date: Mon Apr 2 15:38:15 2007 -0600 166632 166633 checkpoint: more clean-up 166634 166635commit 76a7a5ca1f068c27c9b5fbd49d5a1da80ed6f488 166636Author: Brian <brian@yutani.localnet.net> 166637Date: Mon Apr 2 15:24:05 2007 -0600 166638 166639 formatting fixes 166640 166641commit 69baad321d35dae0bfa535be0c6ed2131fed1e60 166642Author: Brian <brian@yutani.localnet.net> 166643Date: Mon Apr 2 15:21:22 2007 -0600 166644 166645 clean-up dmxCoreMotion() dmxCoreMotion2() 166646 166647commit 44eb15adeee3b299677070f39625daa53679bd13 166648Author: Brian <brian@yutani.localnet.net> 166649Date: Mon Apr 2 15:12:04 2007 -0600 166650 166651 checkpoint clean-up 166652 166653commit 0f873a9d4f02b399c37b4058c6a9a2e21aa205e8 166654Author: Brian <brian@yutani.localnet.net> 166655Date: Mon Apr 2 14:51:38 2007 -0600 166656 166657 remove some debug code 166658 166659commit 3a0ce1084a18e17a3c8a009d99c228652b8763a9 166660Author: Brian <brian@yutani.localnet.net> 166661Date: Mon Apr 2 14:51:21 2007 -0600 166662 166663 for completeness, init dummy's min/maxval[1] values (vertical axis) 166664 166665commit 12016f20f7f5365f30cfbeb05568b3fb89759e5a 166666Author: Brian <brian@yutani.localnet.net> 166667Date: Mon Apr 2 14:50:48 2007 -0600 166668 166669 As for normal mouse device, init valuator maxval[] to real values, not zero. 166670 166671commit 0aaf28e5633a59563b89a2e42d19fabc84adc3ed 166672Author: Brian <brian@yutani.localnet.net> 166673Date: Mon Apr 2 12:41:30 2007 -0600 166674 166675 In dmxBackendMouGetInfo() initialize the info->minval[], maxval[] arrays to the size of the backend display. 166676 166677 It seems that the changes to X input exposed a problem that wasn't detected 166678 before. The axis clipping code in GetPointerEvents() uses those limits to 166679 constrain the pointer's coordinate range. The max was zero so the pointer 166680 couldn't move. 166681 166682commit 0013bf6ddb3867c9a504603434d8c2ec83f3f3bc 166683Author: Brian <brian@yutani.localnet.net> 166684Date: Mon Apr 2 12:39:04 2007 -0600 166685 166686 undo 1280 valuator hack 166687 166688commit 08a88d1803f672555141011e082fbc0edeedcf05 166689Author: Brian <brian@yutani.localnet.net> 166690Date: Mon Apr 2 12:28:14 2007 -0600 166691 166692 Pass num_valuators=0 for ButtonPress/Release. This seems to fix the button coordinate problem 166693 166694commit 70683e338dacc48e3adf489d66ec33b29dfc3b77 166695Author: Brian <brian@yutani.localnet.net> 166696Date: Mon Apr 2 12:26:27 2007 -0600 166697 166698 formatting fixes 166699 166700commit e8777a91f37d828b9df839bf3d9cf2f954bdddb0 166701Merge: a12054757 96ce17aa9 166702Author: Peter Hutterer <peter@cs.unisa.edu.au> 166703Date: Mon Apr 2 15:36:26 2007 +0930 166704 166705 Merge branch 'master' into mpx 166706 166707 Conflicts: 166708 166709 Xi/closedev.c 166710 Xi/exevents.c 166711 Xi/extinit.c 166712 Xi/listdev.c 166713 dix/window.c 166714 hw/xfree86/common/xf86Xinput.c 166715 include/extinit.h 166716 mi/mipointer.c 166717 166718commit 96ce17aa94413c4b8bcb61cae71167050130a307 166719Author: Peter Hutterer <peter@cs.unisa.edu.au> 166720Date: Wed Nov 15 17:50:02 2006 +1030 166721 166722 o fix minor error in comment for GetPointerEvents() 166723 166724commit a12054757d21edacc1c24c3077b9214726652829 166725Author: Peter Hutterer <peter@cs.unisa.edu.au> 166726Date: Mon Apr 2 11:20:44 2007 +0930 166727 166728 Xi: fix ProcXListInputDevices to include the virtual core devices. 166729 166730commit f2808005f4ee72c5fd7f5f3dcca181306485113e 166731Author: Alberto Mardegan <mardy@users.sourceforge.net> 166732Date: Sat Mar 31 16:51:24 2007 +0200 166733 166734 Bug #6620: Fixed a missing 'else' in ATIPseudoDMAInit(). 166735 166736 Before this, we'd write some registers twice on R200 hardware and also 166737 possibly end up with a bad value in atis->cce_pri_size. 166738 166739commit 5257b32e492bd2082bef6a4cd0fea03ce093c0f8 166740Author: Aaron Plattner <aplattner@nvidia.com> 166741Date: Wed Mar 28 15:51:24 2007 -0700 166742 166743 Bump video driver ABI to 2.0 for cw change (commit 6ed08949af4f7ac09170d3d9581e4092b24a84ee). 166744 166745commit 73fdc16bc4f4e21ff604b3f9ded23b40398fb1b6 166746Author: Brian <brian@yutani.localnet.net> 166747Date: Fri Mar 30 16:07:26 2007 -0600 166748 166749 formatting fixes 166750 166751commit ebdc8ce5c108dc3b6b0004e7c7939d1a5bef8676 166752Author: Brian <brian@yutani.localnet.net> 166753Date: Fri Mar 30 16:05:46 2007 -0600 166754 166755 Checkpoint DMX updates: things are working much better now, but still not 100% right. 166756 166757 Use new dmxCoreMotion2() function which enqueues motion events with 166758 GetPointerEvents()/mieqEnqueue(). 166759 The clipAxis() code in GetPointerEvents() is causing some grief. The 166760 limits seem to have always been (0,0) according to the original calls 166761 to InitValuatorAxisStruct() in dmxinputinit.c. 166762 Terrible hack for now: Call InitValuatorAxisStruct() with hard-coded max 166763 values of 1280 (my screen width). 166764 166765commit e9130b8bac73a0843d5ff6b2216eccfb3e094a48 166766Merge: 2c6d47108 92e8cdbd3 166767Author: Ian Romanick <idr@us.ibm.com> 166768Date: Fri Mar 30 13:39:15 2007 -0700 166769 166770 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 166771 166772commit 3c7413e0c2f87e154aa8aa4a83bd585a6d1091e8 166773Author: Brian <brian@yutani.localnet.net> 166774Date: Fri Mar 30 14:07:04 2007 -0600 166775 166776 Tweak some parameters, etc. Things seem a little better now, but still a ways to go. 166777 166778commit 7989dacdcb1449b10d7733dda11cd96e260e9fae 166779Author: Brian <brian@yutani.localnet.net> 166780Date: Fri Mar 30 13:44:24 2007 -0600 166781 166782 num_valuators=1 for GetPointerEvents(), hack ButtonPress/Release position 166783 166784commit 1ea842960fddbc6363cc6e7f914d70ba45525a6b 166785Author: Brian <brian@yutani.localnet.net> 166786Date: Fri Mar 30 13:43:15 2007 -0600 166787 166788 more debug 166789 166790commit 92e8cdbd32b0d86cabd4ad88e3240bf90c018b9a 166791Author: Brian <brian@yutani.localnet.net> 166792Date: Fri Mar 30 13:19:33 2007 -0600 166793 166794 Checkpoint fixes to DMX for X input changes. 166795 166796 Xdmx builds and runs now. 166797 Keyboard seems OK, and mouse pointer moves, but everything else is flakey. 166798 Something is still seriously wrong. 166799 166800commit d92da3d5f309392ac398c0975ef17bb04312d5e2 166801Author: Brian <brian@yutani.localnet.net> 166802Date: Fri Mar 30 12:56:34 2007 -0600 166803 166804 more formatting fixes 166805 166806commit 44acb2517d9fb07790d9d799aa9cc727d1b7d35c 166807Author: Brian <brian@yutani.localnet.net> 166808Date: Fri Mar 30 12:54:22 2007 -0600 166809 166810 Fix some bad formatting. 166811 166812 Doing this: 166813 if (something) stmt; 166814 is evil if you're debugging and want to break on stmt! 166815 166816commit 9f24798af50896cc3262c1201f75c10a688f2a83 166817Author: Brian <brian@yutani.localnet.net> 166818Date: Fri Mar 30 12:49:34 2007 -0600 166819 166820 ompile fbcmap.c w/ -DXFree86Server instead of linking libfbcmap.a. 166821 166822 The former works, the later doesn't (DMX blows up on visuals/pixel formats). 166823 This undos Daniel's patch, which undid my prev patch. Revisit someday. 166824 166825commit 76756f27561c6386cba0d338441e8ec7b98500ce 166826Author: George Sapountzis <gsap7@yahoo.gr> 166827Date: Thu Nov 30 04:20:32 2006 +0200 166828 166829 Make xf86glx.c unaware of Mesa internals 166830 166831 Use newly added XMesaCopyContext() and drop the GlxSetRenderTables() call 166832 for Xgl, as this is now done inside XMesaForceCurrent(). This leaves xmesaP.h 166833 but only for the declarations of the three XMesa/XFree86 functions. Also, 166834 GlxSetRenderTables() stays but is only used in hw/xgl/glxext/ . 166835 166836 Also drop xf86glxint.h, no longer used. 166837 166838 Depends on mesa commit 7439a36785b6a2783e80a40a96c09db8f56dc2bc of 2007-03-30. 166839 166840commit 63d8f01819ef44ea3bf0d4fb20ba1d698ae91cd2 166841Author: Peter Hutterer <peter@cs.unisa.edu.au> 166842Date: Fri Mar 30 14:59:43 2007 +0930 166843 166844 dix: Fix wrong cursor refcount. 166845 166846 Calloc cursor struct to ensure devPrivates are zeroed out and don't 166847 increase the refcnt for devices automatically when allocating a new 166848 cursor. Use new DeviceIsPointerType() to detect if device is a pointer 166849 _before_ device has been activated and can thus be identified and set 166850 up grab functions accordingly. This way we can increase the refcnt 166851 when we get a pointer grab. 166852 166853commit 20e4314b178e1a093bce85e93329d4bcfb4f4210 166854Author: Peter Hutterer <peter@cs.unisa.edu.au> 166855Date: Thu Mar 29 19:12:32 2007 +0930 166856 166857 dix: Try to get closer to the correct refcnt for cursors. 166858 166859 This commit introduces a memory leak, as the refcnt never seems to get 166860 down to 0 and thus cursors will not be freed. Solves the problems with 166861 GNOME segfaulting every 30 seconds though. 166862 166863commit 307d2b57bbfcc281656011533627bea6ab98189e 166864Author: Peter Hutterer <peter@cs.unisa.edu.au> 166865Date: Thu Mar 29 15:23:41 2007 +0930 166866 166867 Xi: remove 'register' keywords. 166868 166869commit 0c607f4b1902f993597e09ea7a2c83bc1cca9b96 166870Author: Peter Hutterer <peter@cs.unisa.edu.au> 166871Date: Thu Mar 29 14:43:50 2007 +0930 166872 166873 dix: Pick keyboard instead of using virtual keyboard in Grab/UngrabKey. 166874 166875commit 82a8b99a6c46018885600011913267d8af9dfe13 166876Author: Adam Jackson <ajax@benzedrine.nwnk.net> 166877Date: Wed Mar 28 15:17:02 2007 -0400 166878 166879 Move the XAA private indices to be static. 166880 166881 Technically this is an ABI break, if you aren't smart enough to be using the 166882 getter functions. Cope. 166883 166884commit 8c7f56d92d8471ee059c14d322af5f7f555dd5c6 166885Author: Tomas Janousek <tomi@nomi.cz> 166886Date: Wed Mar 28 14:46:30 2007 -0400 166887 166888 Bug #10296: Fix timer rescheduling. 166889 166890commit 327bc332a61294209d39286228199f54bdde73d1 166891Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166892Date: Wed Mar 28 13:00:03 2007 -0400 166893 166894 xace: minor comment fixes. 166895 166896commit 299ff4c82998d2a32204bfbecde4993dfbd3d4a5 166897Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166898Date: Wed Mar 28 12:57:11 2007 -0400 166899 166900 xace: provide creation-time resource hook call in CreateWindow(). 166901 166902commit 99b220969f2f8ba829bc8294b27ca90fd9311be4 166903Merge: 728ad64d9 82a8b99a6 166904Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166905Date: Wed Mar 28 12:54:04 2007 -0400 166906 166907 Merge branch 'master' into XACE-SELINUX 166908 166909commit 5ba4d9eedf1b4ce4795bf910cd184872e2d9b3fc 166910Author: Adam Jackson <ajax@benzedrine.nwnk.net> 166911Date: Wed Mar 28 12:03:19 2007 -0400 166912 166913 Refuse to create tiny modes from EDID detailed timing. 166914 166915commit 2c6d47108880584f1221ff86c6c8947627f9f607 166916Merge: c4fe1bcce 852204463 166917Author: Ian Romanick <idr@us.ibm.com> 166918Date: Wed Mar 28 08:08:04 2007 -0700 166919 166920 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 166921 166922 Conflicts: 166923 166924 hw/xfree86/Makefile.am 166925 hw/xfree86/dri/dri.c 166926 166927commit 85220446359a75ea2c359b418b4051c04eea739c 166928Author: Daniel Stone <daniel@fooishbar.org> 166929Date: Wed Mar 28 13:03:32 2007 +0300 166930 166931 GL: Update for Mesa changes 166932 Added s_fragprog.c to fix the build. 166933 166934commit 1af2ef0b25fd8017a3271e624a5f1548f02b09f9 166935Author: Eric Anholt <eric@anholt.net> 166936Date: Tue Mar 27 13:13:45 2007 -0700 166937 166938 Enable Composite by default now that it disables itself in the known bad cases. 166939 166940commit 0bfc3cc22db94ec6867596606fe93228e315c847 166941Author: Eric Anholt <eric@anholt.net> 166942Date: Tue Mar 27 13:12:21 2007 -0700 166943 166944 Disable composite when Xinerama is active. 166945 166946 It will likely take a decent bit of work to make that work right. 166947 166948commit 5e7936371c9e1ac48e19bf1e9e3f71f037fd9b5d 166949Author: Eric Anholt <eric@anholt.net> 166950Date: Mon Mar 26 20:18:18 2007 -0700 166951 166952 Disable Composite when the screen's visual is pseudocolor. 166953 166954 Rendering fails badly in this case, and I don't care enough to fix it. 166955 166956commit 8afc7e2eb3ebec48d3879bf269143259c8bc18c8 166957Author: Eric Anholt <eric@anholt.net> 166958Date: Mon Mar 26 15:55:38 2007 -0700 166959 166960 Refuse to initialize Composite if Render is not present. 166961 166962 Composite relies on the presence of Render, in particular for the automatic 166963 compositing. 166964 166965commit 6ed08949af4f7ac09170d3d9581e4092b24a84ee 166966Author: Eric Anholt <eric@anholt.net> 166967Date: Tue Mar 27 17:31:28 2007 -0700 166968 166969 Move libcw setup to the only renderer requiring it (XAA). 166970 166971 Additionally, protect libcw setup behind checks for Render, to avoid 166972 segfaulting if Render isn't available (xnest). 166973 166974 The previous setup was an ABI-preserving dance, which is better nuked now. 166975 Now, anything that needs libcw must explicitly initialize it, and 166976 miDisableCompositeWrapper (previously only called by EXA and presumably binary 166977 drivers) is gone. 166978 166979commit e76b6349516d5d1c8f7167d6f5419e0d06a546c3 166980Author: Eric Anholt <eric@anholt.net> 166981Date: Mon Mar 26 16:04:50 2007 -0700 166982 166983 Fix indentation of fakexa help text. 166984 166985commit 728ad64d979fd9e5cca28e8c15118c18d707c431 166986Merge: 0331c6a8f 6a0bed16e 166987Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 166988Date: Tue Mar 27 13:16:40 2007 -0400 166989 166990 Merge branch 'master' into XACE-SELINUX 166991 166992commit 6a0bed16e80a91891cee6c7033c90875bc2af193 166993Author: Michel Dänzer <michel@tungstengraphics.com> 166994Date: Tue Mar 27 16:51:12 2007 +0200 166995 166996 Fix typo in GL/mesa/shader/slang/Makefile.am. 166997 166998commit b8f846a9dfc6697d59ad5482ba7c9d738875318e 166999Author: Dave Airlie <airlied@linux.ie> 167000Date: Tue Mar 27 14:17:40 2007 +1000 167001 167002 gl: oops dodgy s appeared pointed out by jcristau on irc.. 167003 167004commit a63ee90bc2d490f6c5c1802c164391963cf6c1d9 167005Author: Dave Airlie <airlied@pegasus.(none)> 167006Date: Tue Mar 27 11:05:52 2007 +1000 167007 167008 gl: update for latest mesa glsl-compiler merge 167009 167010commit d387a3ddf76716791e5e8b8f0954ca0df3c579d6 167011Author: Dave Airlie <airlied@pegasus.(none)> 167012Date: Tue Mar 27 11:00:13 2007 +1000 167013 167014 fix loading of GLcore after recent loading changes 167015 167016commit 92ba435bd9aa7b6eca9aef8e5193576ef62fc9db 167017Author: Eric Anholt <eric@anholt.net> 167018Date: Mon Mar 26 12:44:58 2007 -0700 167019 167020 Update xorg.conf manpage for new RandR 1.2 monitor options. 167021 167022commit 0331c6a8fcba1dc27ef45aaf5f694799d0085413 167023Merge: 84a066cc8 f7c5aa0dc 167024Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167025Date: Mon Mar 26 10:28:05 2007 -0400 167026 167027 Merge branch 'master' into XACE-SELINUX 167028 167029commit f7c5aa0dc0fa3569a2ee412c4f996960f936b6ed 167030Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167031Date: Mon Mar 26 10:21:44 2007 -0400 167032 167033 Remove dead NEED_DBE_BUF_BITS code. 167034 167035commit 2e3cc861f90415f200826bc71dab6298d759c42b 167036Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167037Date: Sun Mar 25 22:01:34 2007 -0400 167038 167039 Since ddc, i2c, and ramdac are in core now, remove their ModuleData stubs. 167040 167041commit e88fa75c9b468b88bb7b87b1da235c6eb2fe8164 167042Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167043Date: Sun Mar 18 17:39:08 2007 -0400 167044 167045 Static cleanup on Xi/ 167046 167047commit 4b5802ddbd45271be3cadeae0a83a6742df2515b 167048Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167049Date: Sun Mar 25 17:57:54 2007 -0400 167050 167051 General DIX static and dead code cleanup. 167052 167053commit 04b87d6dfae02e4ecdb5216d12c6cdafd1e8c2b4 167054Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167055Date: Sun Mar 25 17:57:22 2007 -0400 167056 167057 Static and dead code cleaup for Xext/ 167058 167059commit af769892a91c9af59de53ca3bcd77fc4967daffb 167060Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167061Date: Sun Mar 25 17:56:32 2007 -0400 167062 167063 Static and dead code cleanup from mi/ 167064 167065commit 62224e39727fd6f1cf11a461983662f615a9fea1 167066Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167067Date: Sun Mar 25 17:55:15 2007 -0400 167068 167069 Static cleanup for xf86 ddx. 167070 167071commit e8bc1988d9ff10b65717574175f70df3c4d6334d 167072Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167073Date: Sun Mar 25 15:13:05 2007 -0400 167074 167075 Un-staticise VTSwitchEnabled, since kbd wants it apparently. 167076 167077commit 70e493d223b1e943e652191150bd0b7e1a6ebcfb 167078Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167079Date: Sun Mar 25 14:55:28 2007 -0400 167080 167081 Static and dead code cleanup over afb/ 167082 167083commit f36bf1a3e4ce9465ea4a6159c209924a3cafbe58 167084Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167085Date: Sun Mar 25 12:28:13 2007 -0400 167086 167087 Delete a dead file. 167088 167089commit 9a0f25de7ca3c68af867b38936103d17daa92ac6 167090Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167091Date: Sun Mar 25 12:27:01 2007 -0400 167092 167093 Static cleanups, dead code deletion. 167094 167095commit ac2356843e38b3400142bc54b65393c12976fc07 167096Author: Peter Hutterer <peter@cs.unisa.edu.au> 167097Date: Sun Mar 25 09:41:33 2007 +0930 167098 167099 dix: Increase allocation size for core keyboard keymap to avoid buffer overrun when copying keymap from extension devices. 167100 167101commit 5f9c10e8da0d9d96cc00a38cdf9a8c3030063e03 167102Author: Peter Hutterer <peter@cs.unisa.edu.au> 167103Date: Sun Mar 25 15:00:03 2007 +0930 167104 167105 Clean up DeviceIntPtr, prepare for Server 1.4 ABI. Move sprite-related 167106 information into a new SpriteInfoRec, remove isMPDev flag. 167107 167108commit e54311395e1e7ae458c8ce59da0914706c4a04f6 167109Author: Peter Hutterer <peter@cs.unisa.edu.au> 167110Date: Sun Mar 25 09:41:33 2007 +0930 167111 167112 dix: Increase allocation size for core keyboard keymap to avoid buffer 167113 overrun when copying keymap from extension devices. 167114 167115commit 1072b88a8f352484e70bc749e300c936e5600480 167116Author: Dave Airlie <airlied@linux.ie> 167117Date: Sun Mar 25 10:06:00 2007 +1000 167118 167119 loader: fix already built-in message 167120 167121commit 804080a7096347d48c686f2c8fbfd06326bce400 167122Author: Keith Packard <keithp@neko.keithp.com> 167123Date: Fri Mar 23 23:41:36 2007 -0700 167124 167125 Make pending properties force mode set. And, remove AttachScreen calls. 167126 167127 Yes, two changes in one commit. Sorry 'bout that. 167128 167129 The first change ensures that when pending property values have been 167130 changed, a mode set to the current mode will actually do something, rather 167131 than being identified as a no-op. In addition, the driver no longer needs to 167132 manage the migration of pending to current values, that is handled both 167133 within the xf86 mode setting code (to deal with non-RandR changes) as well 167134 as within the RandR extension itself. 167135 167136 The second change eliminates the two-call Create/AttachScreen stuff that was 167137 done in a failed attempt to create RandR resources before the screen 167138 structures were allocated. Merging these back into the Create function is 167139 cleaner. 167140 (cherry picked from commit 57e87e0d006cbf1f5b175fe02eeb981f741d92f0) 167141 167142 Conflicts: 167143 167144 randr/randrstr.h 167145 randr/rrcrtc.c 167146 167147 I think master and server-1.3-branch are more in sync now. 167148 167149commit 1f77120775dc05fc84a00dd55190af2fa50ae509 167150Author: Keith Packard <keithp@neko.keithp.com> 167151Date: Fri Mar 23 14:39:10 2007 -0700 167152 167153 Ensure that crtc desired values track most recent mode. 167154 167155 desiredX and desiredY were not recorded during xf86InitialConfiguration. 167156 desiredX, desiredY and desiredRotation were not recorded during 167157 xf86SetSingleMode. 167158 (cherry picked from commit 36e5227215e0912ddf8a010db042467f00efe0fc) 167159 167160commit 84a066cc88fe4326ddacd04ab5e1158a80571c33 167161Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167162Date: Fri Mar 23 10:33:53 2007 -0400 167163 167164 xace: pass serverClient as default argument to dixChangeWindowProperty 167165 instead of NullClient. 167166 167167commit 476f2b5aefa518262b69e487555e6094818d857a 167168Author: Keith Packard <keithp@guitar.keithp.com> 167169Date: Fri Mar 23 01:17:14 2007 -0700 167170 167171 Incorrect extra memory copy in RRChangeOutputProperty. 167172 167173 Left over from previous version of the code, this memmove will break when 167174 the mode is not Replace. 167175 (cherry picked from commit 945aa0aa556429b50dea8e8ebc0008304b093eb7) 167176 167177commit 7093367c3976bef5b9d219d9f2a7dc7dd3eeb091 167178Author: Keith Packard <keithp@guitar.keithp.com> 167179Date: Fri Mar 23 01:05:55 2007 -0700 167180 167181 Fix Pending property API, adding RRPostPendingProperty. 167182 167183 Pending Properties take effect when the driver says they do, so provide an 167184 API to tell DIX when a property effect is made. Also, allow driver 167185 to reject property values in RRChangeOutputProperty. 167186 (cherry picked from commit 8eb288fbd69e2ffd02521d2c6a964c8180d08ec8) 167187 167188commit 86d76390eb182f271f5fa5dc19205e97a867f7e7 167189Author: Keith Packard <keithp@guitar.keithp.com> 167190Date: Fri Mar 23 01:03:40 2007 -0700 167191 167192 Make sure RandR events are delivered from RRCrtcSet. 167193 167194 Some paths were skipping the event delivery stage. 167195 (cherry picked from commit 9ca7ba5d6012295a77ed773c656e786440da973d) 167196 167197commit 510eaa346e68fd82c852c7b41fb0e2c5be12da78 167198Author: Keith Packard <keithp@guitar.keithp.com> 167199Date: Fri Mar 23 00:59:11 2007 -0700 167200 167201 Clean up xf86CrtcRec and xf86OutputRec objects at CloseScreen. 167202 167203 Erase pointers to structures which are freed at server reset time. 167204 (cherry picked from commit 492c768065f49306a2194a88edf96b85de0ff4ff) 167205 167206commit 479b2be4badab0a67b1f091feb83c1364e27d783 167207Author: Keith Packard <keithp@guitar.keithp.com> 167208Date: Fri Mar 23 00:57:18 2007 -0700 167209 167210 Clear allocated RandR screen private structure. 167211 167212 Use xcalloc instead of xalloc when allocating this structure to ensure 167213 consistent contents at startup. 167214 (cherry picked from commit 16f4c0c1750824f2e5a001cef82a4122a7a2beb0) 167215 167216commit b63e0d2545bb75e14d9de019a88f31e20a2f7377 167217Author: Keith Packard <keithp@guitar.keithp.com> 167218Date: Tue Mar 20 07:17:27 2007 -0700 167219 167220 Clean up Rotate state on server reset. 167221 167222 The rotation state is stored in the xf86_config structure which is not 167223 re-initialized at server reset time. Clean it up at CloseScreen time. 167224 (cherry picked from commit f8db7665dcd7af78ca4db2461e0bf787ec662cb1) 167225 167226commit e1cc68add0bcdd5e0e4e15cf6ee8a3da136d3534 167227Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167228Date: Thu Mar 22 17:33:16 2007 -0400 167229 167230 xace: drop the name argument from the property callback. 167231 167232commit c9fb8a35332d101897607d8f06ed5a6512eac7cf 167233Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167234Date: Thu Mar 22 17:23:26 2007 -0400 167235 167236 dix: move access codes to separate header file, add DixCreateAccess. 167237 167238commit 1b766ffc0647d5e9a9bf6938d33548d977b5535e 167239Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167240Date: Thu Mar 22 15:55:35 2007 -0400 167241 167242 dix: reorganize property code to better support xace hook; requires new API for 167243 changing a property, dixChangeWindowProperty, taking an additional client argument. 167244 167245commit 1b58304ac837735920747ed0f0d10ba331bdaeb7 167246Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167247Date: Thu Mar 22 13:06:50 2007 -0400 167248 167249 xace: add new argument to property hook for property structure itself. 167250 167251commit 5486be4898766205149fadce71529724eb78fbf3 167252Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167253Date: Thu Mar 22 10:59:21 2007 -0400 167254 167255 dix: devPrivates support for PropertyRec. 167256 167257commit f8cd19bd1692d12a6047c088a626a6ae715d469d 167258Author: Peter Hutterer <peter@cs.unisa.edu.au> 167259Date: Thu Mar 22 17:47:01 2007 +1030 167260 167261 Xi: add GetClientPointer and GetPointerPairing requests. 167262 167263 dix: use PickKeyboard for ProcSetInputFocus and ProcGetInputFocus instead 167264 of the core keyboard. 167265 167266commit a3296d111dc4d76aa3afa7e338cbab93eb390ec4 167267Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167268Date: Wed Mar 21 17:01:26 2007 -0400 167269 167270 xace: add access_mode argument to selection hook. 167271 167272commit 4c1fb8069d5dd30a73277698503e9dcc2e9d64c6 167273Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167274Date: Wed Mar 21 16:17:14 2007 -0400 167275 167276 dix: add new selection fields supporting redirection. 167277 This is a minor ABI break. 167278 167279commit 4fa482b4be1150bcffeabb64d018c00ac5951e41 167280Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167281Date: Wed Mar 21 14:49:56 2007 -0400 167282 167283 xace: bump major version since the hooks have changed. 167284 167285commit 9c144f8ac5cea25deaa543767dbaf371d029c608 167286Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167287Date: Wed Mar 21 14:39:00 2007 -0400 167288 167289 xace: add XACE_SELECTION_ACCESS hook for selection redirection/access. 167290 167291commit 3e9f7a5504ab41d845e88f293d8498c963d8a7d8 167292Author: Daniel Stone <daniel@fooishbar.org> 167293Date: Wed Mar 21 02:35:31 2007 +0200 167294 167295 XFree86 DGA: Guard against NULL pointer dereferences. 167296 Ass, u, me ... 167297 167298commit f292de2ef13dc994a38029cee9e2642576893332 167299Author: Daniel Stone <daniel@fooishbar.org> 167300Date: Wed Mar 21 02:04:12 2007 +0200 167301 167302 XKB: Fix size_syms calculation bug 167303 167304 Apparently it needed to be nSyms*15/10, not *12/10; make it match the 167305 other allocation code. 167306 167307commit f34b9a20b0181d3c2641c305e91180711afbd4b9 167308Author: Daniel Stone <daniel@fooishbar.org> 167309Date: Wed Mar 21 02:03:37 2007 +0200 167310 167311 XKB: Be a tiny bit more conservative with type allocation 167312 167313 Make sure size_types will _always_ be 0 if we don't have any types. 167314 167315commit 021fc5cb2cb4a7972b4a6fcb570c1da92787d68d 167316Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167317Date: Sun Mar 18 16:31:19 2007 -0400 167318 167319 Static markup and dead code cull over xkb/. 167320 167321 The former <X11/extensions/XKBsrv.h> has been pulled into the server now as 167322 include/xkbsrv.h, and the world updated to look for it in the new place, 167323 since it made no sense to define server API in an extension header. Any 167324 further work along this line will need to do similar things with XKBgeom.h 167325 and friends. 167326 167327commit 9398d62f27ee1b287e4458fd8b011c10f7b59efd 167328Author: Daniel Stone <daniel@fooishbar.org> 167329Date: Wed Mar 21 00:18:24 2007 +0200 167330 167331 XFree86 input: Add backwards compatibility for motion history 167332 Add the old motion history API back, as a shim around the new mi API. 167333 167334commit 0f75c47e0c5f4b2778930a6fabf894fc1dffd9d3 167335Author: Daniel Stone <daniel@fooishbar.org> 167336Date: Wed Mar 21 00:12:02 2007 +0200 167337 167338 xfree86 input: Re-enable DGA support 167339 Re-enable DGA support for relative mouse motion. 167340 167341commit 80d29475b9a2ebbb303a8e324e09a15c528d5556 167342Author: Daniel Stone <daniel@fooishbar.org> 167343Date: Wed Mar 21 00:10:38 2007 +0200 167344 167345 mieq: Allow event handlers for arbitrary events to be set 167346 Allow arbitrary events to use mieq by letting custom handlers be set. 167347 167348commit b8df961843a95b29258ae9c5d46ccfc620d8de1c 167349Author: Alan Coopersmith <alan.coopersmith@sun.com> 167350Date: Mon Mar 19 18:03:26 2007 -0700 167351 167352 Define XF86PM on Solaris x86 builds now that we have sun_apm.c 167353 167354commit 2945deba1d4a7dce4f6dd0c568297a1c537fdfb4 167355Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167356Date: Mon Mar 19 17:09:10 2007 -0400 167357 167358 xace: drop XACE_WINDOW_INIT hook, it has been superseded by ResourceStateCallback. 167359 167360commit 78c962da76efe644b8d485265f1ecdda84b45d27 167361Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167362Date: Mon Mar 19 17:04:51 2007 -0400 167363 167364 xselinux: use the new ResourceStateCallback instead of the XACE_WINDOW_INIT hook. 167365 167366commit 6a89106e9c963a495fd40427d242ba0abd44f764 167367Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167368Date: Mon Mar 19 16:51:29 2007 -0400 167369 167370 xselinux + security: remove confusing CALLBACK macro. 167371 167372commit 0acf993e2878d0866fd4e2db8252c1ea6841e494 167373Author: Peter Hutterer <peter@cs.unisa.edu.au> 167374Date: Mon Mar 19 16:11:01 2007 +1030 167375 167376 dix: Move sprite initalization/pairing from ActivateDevice to 167377 EnableDevice. 167378 zero out spriteTrace, otherwise GetCurrentRootWindow() may return 167379 invalid data. 167380 167381 Evdev adds previously hotplugged devices on its own when rescanning, 167382 and ActivateDevice does not get called for those devices. Sprites need 167383 to be initialized or paired before the device gets started, so the 167384 safest place to do so is EnableDevice (which actually does get called 167385 by the evdev driver). 167386 167387commit 015d728bcde5b16a72a976579755421e3023626b 167388Merge: 65b0eb60b b167583fe 167389Author: Peter Hutterer <peter@cs.unisa.edu.au> 167390Date: Mon Mar 19 09:42:56 2007 +1030 167391 167392 Merge branch 'master' into mpx 167393 167394 Conflicts: 167395 167396 dix/devices.c 167397 dix/events.c 167398 mi/misprite.c 167399 167400commit 720f302d241e88e6e9f2962207da1aa9a79728b7 167401Author: Keith Packard <keithp@neko.keithp.com> 167402Date: Sat Mar 17 20:14:05 2007 -0700 167403 167404 Slow down DDC I2C bus using a RiseFallTime of 20us for old monitors. 167405 167406 This time value makes the bus run slowly enough for even the least reliable 167407 of monitors. Thanks to Pavel Troller for finding the necessary change. 167408 167409commit b5a8a71e64c76b8dd42962cbd7984215c6ce4aa8 167410Author: Keith Packard <keithp@neko.keithp.com> 167411Date: Sat Mar 17 17:26:11 2007 -0700 167412 167413 Remove extra (and wrong) I2C ByteTimeout setting in DDC code. 167414 167415 The DDC code sets the I2C timeouts to VESA standards, except that it had an 167416 extra setting of the ByteTimeout value which was wrong (off by a factor of 167417 50). Removing this should help DDC work on many more monitors. Note that the 167418 Intel driver duplicated these settings, along with the error. Yay for cult 167419 and paste coding. 167420 167421commit 2489dae9f7def788910eee5733931392df83a0d6 167422Author: Keith Packard <keithp@guitar.keithp.com> 167423Date: Thu Mar 15 20:26:07 2007 -0700 167424 167425 Correct ref counting of RRMode structures 167426 167427 RRModes are referenced by the resource db, RROutput and RRCrtc structures. 167428 Ensure that the mode reference count is decremented each time a reference is 167429 lost from one of these sources. The missing destroys were in 167430 RRCrtcDestroyResource and RROutputDestroyResource, which only happen at 167431 server reset time, so modes would be unavailable in subsequent server 167432 generations. 167433 167434commit 9d0c3b52f25df89738fb1a62ccffda8c8cbb4689 167435Author: Keith Packard <keithp@guitar.keithp.com> 167436Date: Tue Feb 20 23:04:26 2007 -0800 167437 167438 Eliminate RRModeRec devPrivate field. 167439 167440 The xf86 mode setting code was mis-using this field to try and store a 167441 pointer to a DisplayModeRec, however, each output has its own copy of every 167442 DisplayModeRec leaving the one in in the RRModeRec devPrivate field pointing 167443 at a random DisplayModeRec. 167444 167445 Instead of attempting to rectify this, eliminating the devPrivate entirely 167446 turned out to be very easy; the DDX code now accepts an arbitrary RRModeRec 167447 structure and set that to the hardware, converting it on the fly to a 167448 DisplayModeRec as needed. 167449 (cherry picked from commit 3506b9376c2b0db09bfff58d64e07af88a6e8195) 167450 167451commit 2c93083edd29a65e73bb2e8eff9d353e92845c9b 167452Author: Keith Packard <keithp@guitar.keithp.com> 167453Date: Sun Feb 18 23:49:38 2007 -0800 167454 167455 Add support for user-defined modelines in RandR. 167456 167457 The RandR protocol spec has several requests in support of user-defined 167458 modes, but the implementation was stubbed out inside the X server. Fill out 167459 the DIX portion and start on the xf86 DDX portion. It might be necessary to 167460 add more code to the DDX to insert the user-defined modes into the output 167461 mode list. 167462 (cherry picked from commit 63cc2a51ef87130c632a874672a8c9167f14314e) 167463 167464 Conflicts: 167465 167466 randr/randrstr.h 167467 167468 Updated code to work in master with recent security API changes. 167469 167470commit b167583fe234a536aa6187e17f21da85be4f0b7f 167471Author: Peter Hutterer <peter@cs.unisa.edu.au> 167472Date: Wed Nov 15 17:50:02 2006 +1030 167473 167474 o fix minor error in comment for GetPointerEvents() 167475 167476commit 65b0eb60b0b9e6405aa2e3d4c712ec4d2f0da070 167477Author: Peter Hutterer <peter@cs.unisa.edu.au> 167478Date: Fri Mar 16 15:52:24 2007 +1030 167479 167480 Xi: move IsPointerDevice into dix. 167481 dix: IsPointerDevice and IsKeyboardDevice, use same ways to identify type 167482 of device as XI does for the XListInputDevices reply. 167483 Autopair each non-pointer device with VCP when activating, pair with 167484 real device after activation. 167485 Don't return non-keyboard devices when calling GetPairedKeyboard or 167486 PickKeyboard, otherwise we segfault for 'evdev brain'. 167487 167488commit 3bffb281260476d2f74f0bf451d85d2f7cacd6c4 167489Author: Keith Packard <keithp@neko.keithp.com> 167490Date: Thu Mar 15 16:16:16 2007 -0700 167491 167492 Don't wedge when rotating more than one CRTC. 167493 167494 Rotation block handler was re-registering the rotation damage structure, 167495 creating an infinite loop in the damage code. Track registration of the 167496 damage structure to avoid this. 167497 (cherry picked from commit b14f003b0ed1252766c9e3b1c086ea2809521047) 167498 167499commit 9562b6abe1da566cf73a08c4f4c4339fb67fbc71 167500Author: Keith Packard <keithp@neko.keithp.com> 167501Date: Thu Mar 15 10:50:45 2007 -0700 167502 167503 Allow xf86_reload_cursors during server init. 167504 167505 xf86_reload_cursors is supposed to be called from the crtc mode setting 167506 commit hook; as that happens during server initialization, check for this 167507 case. 167508 (cherry picked from commit 5b77bf2d020b1ee56c1c5f2db089a8f7f64a76a6) 167509 167510commit 3b71b0f89f1db837da91650baa0ef4bb7ef2e98f 167511Author: Eric Anholt <eric@anholt.net> 167512Date: Thu Mar 15 13:21:00 2007 -0700 167513 167514 Set the RandR version returned, rather than just passing the proto's version. 167515 167516commit 2fe74ef339c3a4902ae8214f5a0454662895422c 167517Author: Matthias Hopf <mhopf@suse.de> 167518Date: Thu Mar 15 16:56:01 2007 +0100 167519 167520 Fix calculations in x86 emulator for the long long case (Andreas Schwab). 167521 167522commit 456f429ad655ab4d80e30c58291d801966a613b2 167523Author: Peter Hutterer <peter@cs.unisa.edu.au> 167524Date: Thu Mar 15 21:48:19 2007 +1030 167525 167526 dix: fix: don't return when device is same as argument. 167527 fix: set access default rule to 0 on init. 167528 167529commit ae75019ccf1edac9e8be31b6a96293624f672ccb 167530Author: Keith Packard <keithp@guitar.keithp.com> 167531Date: Wed Mar 14 23:59:29 2007 -0700 167532 167533 Create driver-independent CRTC-based cursor layer. 167534 167535 This moves most of the cursor management code out of the intel driver and 167536 into the general server code. Of course, the hope is that this code will be 167537 useful for other driver writers as well. 167538 167539 Check out xf86Crtc.h for the usage information, making sure you add the 167540 needed hooks to the crtc funcs structure for your driver. 167541 (cherry picked from commit 4d81c99a4660a0bf9014f789de55edabd185bd14) 167542 167543commit c5b07fb717289f61b54d7b093421bcb92124b839 167544Author: Peter Hutterer <peter@cs.unisa.edu.au> 167545Date: Wed Mar 14 17:42:30 2007 +1030 167546 167547 kdrive: update kdrive to new mi API with additional device argument. Includes 167548 an update of ati_cursor.c and i810_cursor.c 167549 167550commit c3eb248cf93a3afd9168acfb88254606beddd7a8 167551Author: Peter Hutterer <peter@cs.unisa.edu.au> 167552Date: Wed Mar 14 13:21:33 2007 +1030 167553 167554 dix: fix: don't skip first item in list when seeking for a device cursor. 167555 167556commit ce9409acebd994cf1809050f2cc8e1db66130f12 167557Author: Peter Hutterer <peter@cs.unisa.edu.au> 167558Date: Tue Mar 13 19:16:56 2007 +1030 167559 167560 dix: Get the state from the paired device and the real device to fill the 167561 state field in the event, rather than using the virtual devices. 167562 ProcessPointerEvent: name argument "device" instead of "other". 167563 Add GetPairedKeyboard(). 167564 167565commit 4bf1b280f7cb676ec2b172f26dd2ad9bac2eb2ca 167566Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 167567Date: Fri Mar 9 14:18:14 2007 +0000 167568 167569 Set pScreen on context 167570 167571commit c366b82bd50066019cf82b3464445d5bc27d6f9f 167572Author: Jay Estabrook <Jay.Estabrook@hp.com> 167573Date: Fri Mar 9 12:26:55 2007 +0000 167574 167575 Ensure domain is stripped from the bus ID. 167576 167577commit 577464af4362e5a32cf7165b5128655dd86c6200 167578Author: Peter Hutterer <peter@cs.unisa.edu.au> 167579Date: Fri Mar 9 17:13:05 2007 +1030 167580 167581 dix: restore commit b3b2a6a0d43d1724e04d69588f8a55c3270e5523 that for some 167582 reason got wiped. 167583 fix ProcGrabKeyboard to use PickKeyboard 167584 fix PickKeyboard to actually work. 167585 167586commit 07806f4081f8dcd3b5733b590eb6e5b4cae734ad 167587Author: Peter Hutterer <peter@cs.unisa.edu.au> 167588Date: Fri Mar 9 15:45:25 2007 +1030 167589 167590 Xi: Add SetClientPointer request handling. 167591 167592commit 405483496538f1c82cbd7fe1e76c5d94e1a90525 167593Author: Peter Hutterer <peter@cs.unisa.edu.au> 167594Date: Fri Mar 9 14:16:23 2007 +1030 167595 167596 mi: remove 'register' keywords. 167597 167598commit fe05ba75a10ec080e7ec34bff6936103185586b3 167599Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167600Date: Thu Mar 8 12:14:30 2007 -0500 167601 167602 devPrivates rework: pass address of pointer to private callbacks instead of 167603 the pointer itself. 167604 167605commit 18339375cd332f0ab1cbdade3dcd9140212ce1ca 167606Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167607Date: Thu Mar 8 12:14:06 2007 -0500 167608 167609 xselinux: remove context validation function for now. 167610 167611commit b9cff1670f29949a5bc41afc19aca443f434febb 167612Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167613Date: Thu Mar 8 12:13:52 2007 -0500 167614 167615 Add ResourceStateCallback similar in function to ClientStateCallback. 167616 167617commit 2fcb45eb5dc1803b372df8b5765f6a43bea83611 167618Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167619Date: Thu Mar 8 12:13:36 2007 -0500 167620 167621 devPrivates rework: redo field offset registration, drop RC_PRIVATES class. 167622 167623commit 947f8d249bac61beb10669d935888c4c280b5062 167624Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167625Date: Thu Mar 8 12:13:18 2007 -0500 167626 167627 devPrivates rework: redo interface again, dropping parent and type parameters 167628 as well as preallocation routine. 167629 167630commit c45f6762080ef00b41d9f73441a9f0e605253008 167631Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167632Date: Wed Mar 7 11:22:42 2007 -0500 167633 167634 devPrivates rework: hook up new mechanism in backwards-compatibility mode 167635 on existing structures that support devPrivates. 167636 167637commit b3b2a6a0d43d1724e04d69588f8a55c3270e5523 167638Author: Peter Hutterer <peter@cs.unisa.edu.au> 167639Date: Thu Mar 8 22:16:17 2007 +1030 167640 167641 dix: Check if client has a grab on any other device before mouse/kbd 167642 event is delivered to the client. Client's don't cope well with 167643 receiving core events from other devices when having a grab on a 167644 pointer or keyboard. 167645 167646commit 63169ce52d354b4345dcfc46b89f0ea88379718f 167647Author: Peter Hutterer <peter@cs.unisa.edu.au> 167648Date: Thu Mar 8 17:50:19 2007 +1030 167649 167650 dix: remove 'register' keyword for all variables. 167651 167652commit ceca5670fee99b5feceaa2453f1ac32d1bfe7dcd 167653Author: Peter Hutterer <peter@cs.unisa.edu.au> 167654Date: Thu Mar 8 17:17:23 2007 +1030 167655 167656 dix: Each device needs to differ between a core grab and an XI grab, 167657 otherwise a Xi grab may overwrite or release a core grab. 167658 Replace grab and associates with coreGrab and deviceGrab structures, 167659 adjust rest of dix/Xi/etc to compile. 167660 167661 xfree86: Don't check for core devices, we'll have the virtual ones anyway. 167662 If we check, the first mouse device is duplicated and sends 167663 double events. 167664 167665commit 40ae4f246d8818410490236ab183204a84765629 167666Author: Keith Packard <keithp@guitar.keithp.com> 167667Date: Wed Mar 7 20:52:31 2007 -0800 167668 167669 Remove stale monitor data when output becomes disconnected. 167670 167671 Remove parsed EDID and EDID property from disconnected outputs. 167672 (cherry picked from commit ae9d5aa479dd50cc81b755079fcf96a0d02f135a) 167673 167674commit b5fde366e2e21234ac0b81222fd5c42ca3e49cba 167675Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167676Date: Wed Mar 7 12:29:55 2007 -0500 167677 167678 Properly free device devPrivates - memory leak fix. 167679 167680commit a3d2c5d622d9ca36d6fa2966aff09524e3ea39ac 167681Author: Adam Jackson <ajax@benzedrine.nwnk.net> 167682Date: Wed Mar 7 11:02:47 2007 -0500 167683 167684 XORG_VERSION_CURRENT, not XF86_VERSION_CURRENT. 167685 167686 If only this was the least wrong thing in this code. 167687 167688commit aaef4d6a4121d9341b670a0ce8fabc3b491049cf 167689Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167690Date: Wed Mar 7 09:57:02 2007 -0500 167691 167692 devPrivates rework: move reset functions into a single call. 167693 167694commit bb111291d854b4329e47367ce3c67e8a2785e6e9 167695Merge: 74175e0af e9bfb2b3d 167696Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167697Date: Wed Mar 7 09:03:46 2007 -0500 167698 167699 Merge branch 'master' into XACE-SELINUX 167700 167701commit 537bc2ead4d154552cbdc3a19e335f82af63792c 167702Author: Peter Hutterer <peter@cs.unisa.edu.au> 167703Date: Wed Mar 7 18:19:20 2007 +1030 167704 167705 Xi: Don't deactivate core grabs from non-core button/key releases. 167706 167707 dix: set coreGrab flag for grabs caused by GrabPointer and button presses. 167708 remove virtual core devices from device list, only real devices are in 167709 the list now. 167710 Auto-pair each keyboard with a real pointer if one is available, 167711 provides multiple keyboards automatically after startup if devices 167712 are configured. 167713 fix GuessFreePointerDevice() to do what it's supposed to do. 167714 167715 mi: fix: call miPointerMove from miPointerWarpCursor. 167716 fix: remove unused id field from miCursorInfoRec 167717 don't update sprite for virtual core pointer. 167718 167719commit e9bfb2b3d7dfaafd90d2ad0fa3d0e1acced4380b 167720Author: Keith Packard <keithp@neko.keithp.com> 167721Date: Tue Mar 6 23:19:30 2007 -0800 167722 167723 Add hw/xfree86/docs/README.modes, documenting new mode setting APIs. 167724 167725 This document covers both API and xorg.conf usage of the new mode setting 167726 APIs. 167727 (cherry picked from commit a59c31b0f7b94ed1f395c7586c37ef5fe7ba2a25) 167728 167729commit 72a23d88d73a8c72ed18847b004db05092d3e7be 167730Author: Keith Packard <keithp@guitar.keithp.com> 167731Date: Tue Mar 6 23:15:34 2007 -0800 167732 167733 Add xf86CrtcScreenInit to share initialization across drivers. 167734 167735 xf86CrtcScreenInit performs initialization that needs to happen at 167736 ScreenInit time. 167737 (cherry picked from commit 558a4f5588ad2ec11254e0b5d6ce9515b137369e) 167738 167739commit 81526232bc0119d2ec7b8590be4f78cf066ae359 167740Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167741Date: Tue Mar 6 17:19:11 2007 -0500 167742 167743 remove PIXPRIV check as this flag is always set. 167744 167745commit a7cd53deb99957dec27a55ffd75e548b322ae0ce 167746Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167747Date: Tue Mar 6 15:32:13 2007 -0500 167748 167749 remove PIXPRIV checks as this flag is always set. 167750 167751commit 024bbc7cbb924daaf3e305ddfc8e74509acd1e15 167752Author: Eric Anholt <eric@anholt.net> 167753Date: Tue Mar 6 16:18:59 2007 -0800 167754 167755 Bug #9931: Fix linear allocations with a non-1-byte granularity. 167756 167757 This was introduced in 83080809f9a1c1d24b0318e54632f25f5940da25. Instead of 167758 aligning the offset, it doubled it. Results were appropriately spectacular. 167759 167760commit 9d94c137596d3f9d9118ec70455b7a30b3582046 167761Author: Ben Byer <bbyer@apple.com> 167762Date: Tue Mar 6 11:09:30 2007 -0800 167763 167764 updated todo list 167765 167766commit 81d581e655fc989da3be4256b83849a63b8607b7 167767Merge: a05ffca8d d5aba03fe 167768Author: Ben Byer <bbyer@bbyer.(none)> 167769Date: Tue Mar 6 10:37:29 2007 -0800 167770 167771 Merge branch 'master' of git+ssh://bbyer@git.freedesktop.org/git/xorg/xserver 167772 167773commit a05ffca8dd0da9bdb5c1bf4c481028aeabf21e34 167774Author: Ben Byer <bbyer@bbyer.(none)> 167775Date: Tue Mar 6 10:36:51 2007 -0800 167776 167777 rewrote event handling, Xquartz now has working mouse and keyboard. use it\! 167778 167779commit d5aba03feff41722c72b4c6193f09d141cbf1678 167780Author: Drew Parsons <dparsons@debian.org> 167781Date: Tue Mar 6 23:53:23 2007 +1100 167782 167783 Xprint: shorten font filename to fit in tar length limit 167784 167785 The length of the Xprint font file NewCenturySchlbk-BoldItalic.pmf 167786 pushes the full path over the traditional 100 character limit for 167787 tarballs (when module version number is included). Shorten it to 167788 NewCentSchlbk-BoldItal.pmf to get back below the limit and rename 167789 other font files in that family to match. 167790 167791commit 3206e9225897989638ad553e1f392b918ac4d21f 167792Author: Ben Byer <bbyer@bbyer.(none)> 167793Date: Tue Mar 6 02:31:59 2007 -0800 167794 167795 moved new event-handling code from X11Application.m to darwinEvents.c in preparation for making all Darwin servers use it 167796 167797commit 0ccd1443fd6db397b42e5b99ce733ce1316c785e 167798Merge: ec1ef8a56 9b6bb06f1 167799Author: Ben Byer <bbyer@bbyer.(none)> 167800Date: Tue Mar 6 01:04:50 2007 -0800 167801 167802 Merge branch 'master' of git+ssh://bbyer@git.freedesktop.org/git/xorg/xserver 167803 167804commit ec1ef8a56d6217ca2b04899043874ce0bcad9784 167805Author: Ben Byer <bbyer@bbyer.(none)> 167806Date: Tue Mar 6 00:57:23 2007 -0800 167807 167808 Fixed Darwin's Makefile.am to fix a problem building X11.app 167809 167810commit 9b6bb06f13a71f6078f762b4a78fa516faccb638 167811Author: Keith Packard <keithp@guitar.keithp.com> 167812Date: Mon Mar 5 23:49:35 2007 -0800 167813 167814 Allow relative positions to use output names or monitor identifiers. 167815 167816 Previous version used monitor identifiers if present, otherwise output 167817 names. That caused existing working configurations to break when additional 167818 information was added to the configuration file. 167819 (cherry picked from commit 3f5cedf00a82f08a433c95ffbb7f8ac69dcf6a50) 167820 167821commit bed76caa6caaea6a6598755b82a54425a9d9f73e 167822Author: Keith Packard <keithp@guitar.keithp.com> 167823Date: Mon Mar 5 23:36:00 2007 -0800 167824 167825 Use EDID data to set screen physical size at server startup. 167826 167827 Screen physical size is set to a random value before the RandR code gets 167828 control, override that and reset it to a value based on the compat_output 167829 physical size (if available). If that output has no physical size, just use 167830 96dpi as the default resolution and set the physical size as appropriate. 167831 (cherry picked from commit 843077f23a1b49bd712d931421753e3a09d4008c) 167832 167833commit 47f8361c3a64834587e54507653d8d5b258c2530 167834Author: Keith Packard <keithp@guitar.keithp.com> 167835Date: Mon Mar 5 22:07:01 2007 -0800 167836 167837 Add xf86SetDesiredModes to apply desired modes to crtcs. 167838 167839 xf86SetDesiredModes applies the desired modes to each crtc (as selected by 167840 xf86InitialConfiguration initially and modified by successful mode settings 167841 afterwards). For crtcs without a desired mode, pScrn->currentMode is used to 167842 select something workable. 167843 (cherry picked from commit bcade98ccaa18298d844a606cb44271f0254c185) 167844 167845commit 33d2cf93fb50464941e74efe246b10aee212223a 167846Author: Keith Packard <keithp@neko.keithp.com> 167847Date: Sat Mar 3 23:10:31 2007 -0800 167848 167849 Move xf86SetSingleMode into X server from intel driver. 167850 167851 This function applies a single mode to the screen (as from RandR 1.1, 167852 XFree86-VidModeExtension or XFree86-DGA) using a policy that selects one 167853 output to reconfigure to the requested mode and then makes all other outputs 167854 fit within that size. 167855 (cherry picked from commit 5a595c1f767a8d666348b845d18934aee0cfe38f) 167856 167857commit 689d52b6242434507a64a8fff27b01607628c393 167858Author: Jens Granseuer <jensgr@gmx.net> 167859Date: Mon Mar 5 15:31:44 2007 -0800 167860 167861 Bugzilla #7145: fix build with gcc 2.95 167862 167863 Bugzilla #7145: <http://bugs.freedesktop.org/show_bug.cgi?id=7145> 167864 Patch #8987: <http://bugs.freedesktop.org/attachment.cgi?id=8987> 167865 167866commit 74175e0af74c530cb712a6772d3c5d61d1be9748 167867Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167868Date: Mon Mar 5 12:34:37 2007 -0500 167869 167870 devPrivates rework: remove some debugging code from dixFreePrivates. 167871 167872commit 23fc429aad5b2721911862617772c314e1036bb0 167873Merge: e68482470 fe7b8f423 167874Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167875Date: Mon Mar 5 12:18:54 2007 -0500 167876 167877 Merge branch 'master' into XACE-SELINUX 167878 167879commit e684824709fa8ffe03dde3c8dfbc58c267515a4f 167880Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 167881Date: Thu Mar 1 15:00:02 2007 -0500 167882 167883 devPrivates rework: redo interface and implementation. 167884 167885commit fe7b8f4237874e3e45fe25a6bf06faddfa1ab8e1 167886Author: Ben Byer <bbyer@bbyer.(none)> 167887Date: Mon Mar 5 03:48:27 2007 -0800 167888 167889 began to factor out code to move to darwinEvents.c 167890 167891commit 537dc5ecde46d0525c503d1d2b39b6eb89a1298e 167892Author: Ben Byer <bbyer@bbyer.(none)> 167893Date: Mon Mar 5 02:30:56 2007 -0800 167894 167895 started moving new input code into darwinEvents.c so that it may be shared by the three servers 167896 167897commit 39aa79177196e21bcdbaf8e44adead9ef91e6ee5 167898Author: Peter Hutterer <peter@cs.unisa.edu.au> 167899Date: Mon Mar 5 15:31:16 2007 +1030 167900 167901 dix: Add GuessFreePointerDevice(). Runs through device list and tries to 167902 find a pointer that hasn't been paired yet. 167903 167904 xfree86: Use GuessFreePointerDevice() for newly connected non-sprite devices. 167905 167906commit 8ba5e8d82014b774a52f3e050ddbbb8bde4e0933 167907Author: Dave Airlie <airlied@linux.ie> 167908Date: Mon Mar 5 13:46:41 2007 +1100 167909 167910 add a standard connector type and name for us as an output property 167911 167912commit 2e31872e05c2408d53ba0182bcddc5dabb3615fe 167913Author: Dave Airlie <airlied@linux.ie> 167914Date: Mon Feb 26 09:40:00 2007 +1100 167915 167916 modes: add commit/prepare hooks 167917 167918commit 1f0075786fedde538a95e2f39681052e25021d88 167919Merge: 57aa5e908 12175b668 167920Author: Peter Hutterer <peter@cs.unisa.edu.au> 167921Date: Mon Mar 5 12:37:17 2007 +1030 167922 167923 Merge branch 'master' into mpx 167924 167925 Conflicts: 167926 167927 configure.ac 167928 dix/getevents.c 167929 hw/xfree86/ramdac/xf86Cursor.c 167930 mi/mipointer.c 167931 xkb/xkbUtils.c 167932 167933commit 06b01186f6ae17aafdd1f628c306466ddea9e065 167934Author: Keith Packard <keithp@neko.keithp.com> 167935Date: Sun Mar 4 17:15:24 2007 -0800 167936 167937 Remove debugging ErrorF from rotation code. 167938 (cherry picked from commit e6af7569f201842b4754aec6e72b30dc2daefdfb) 167939 167940commit c14507b6837387d867792a24778786311b2b38d5 167941Author: Keith Packard <keithp@neko.keithp.com> 167942Date: Sun Mar 4 17:06:37 2007 -0800 167943 167944 Handle non-zero origin rotated crtc. Damage crtc area on re-rotate. 167945 167946 Box transformation from source to dest area was broken, leaving the wrong 167947 areas painted when the crtc origin was non-zero. 167948 167949 When rotating from left to right, the pixmap doesn't get reallocated, and so 167950 no damage was left in the pixmap from xf86RotatePrepare. Separately damage 167951 the whole crtc area when this occurs to repaint the area. 167952 (cherry picked from commit 2a50ca2160bc05af1c24421ec079e902ff730277) 167953 167954commit 97978b515b7af5fbaaa32b1729e835f3bfb9f5c6 167955Author: Drew Parsons <drew@pug.localdomain> 167956Date: Sun Mar 4 16:28:54 2007 +1100 167957 167958 Xprint: fix font symlinks 167959 167960 Change symlinks to Xprint base fonts in model/PSdefault using local 167961 relative links. This facilitates moving the Xprint config files, for 167962 instance for FHS compliance placing data files in /usr/share rather 167963 than /usr/lib. Also ensures NewCenturySchlbk-BoldItalic.pmf is 167964 installed. 167965 167966commit 215e3691b76a63e6af19865790193b20b105ec5a 167967Author: Ben Byer <bbyer@xyzzy.local> 167968Date: Sat Mar 3 21:52:56 2007 -0800 167969 167970 stopped using XTrans internals in X11.app because they're apparently no longer public 167971 167972commit ea8dcc458ea8870126cf8d3e21cab9d63d094c5e 167973Author: Ben Byer <bbyer@xyzzy.local> 167974Date: Sat Mar 3 21:51:20 2007 -0800 167975 167976 Makefile fix for X11.app 167977 167978commit 18508212599bf0964c450c69b9790208e5d428be 167979Author: Ben Byer <bbyer@xyzzy.local> 167980Date: Sat Mar 3 21:41:33 2007 -0800 167981 167982 fixed X11.xcodeproj to get CFLAGS and LDFLAGS from autoconf script 167983 167984commit 7f2b9f3790456044d01bf8e6404f9a1239b41da6 167985Author: Ben Byer <bbyer@xyzzy.local> 167986Date: Sat Mar 3 19:27:53 2007 -0800 167987 167988 autoconf fixes for XDarwin (created DARWIN_LIBS) 167989 167990commit 12175b668a94e23994f724b366a691ec312cce69 167991Author: Peter Hutterer <peter@cs.unisa.edu.au> 167992Date: Wed Nov 15 17:50:02 2006 +1030 167993 167994 o fix minor error in comment for GetPointerEvents() 167995 167996commit ea1a72946d1aa4c256e6afb9d834c582ba4ac3a1 167997Author: Aaron Plattner <aplattner@nvidia.com> 167998Date: Wed Feb 28 14:26:47 2007 -0800 167999 168000 Add a canGrow argument to xf86InitialConfiguration. 168001 168002 canGrow indicates to the DDX that the driver can enlarge the desktop via the 168003 xf86_config->funcs->resize hook. If so, xf86InitialConfiguration will set 168004 virtual[XY] to match the configuration it chooses and will leave the crtc config 168005 size ranges alone. If FALSE, it will bloat the screen to fit the largest probed 168006 mode and also set the crtc config max size to limit the desktop to the initial 168007 virtual[XY] size. 168008 168009commit 04d15da95d608766c7832a7aa881be499c1395ba 168010Author: Aaron Plattner <aplattner@nvidia.com> 168011Date: Wed Feb 28 13:36:58 2007 -0800 168012 168013 Add a screen resize hook to xf86CrtcConfigRec. 168014 168015 This hook is called when the DDX needs to resize the screen. The driver is 168016 responsible for changing virtualX and virtualY, along with any other related 168017 screen properties (devPrivate.ptr, devKind, displayWidth, etc.). 168018 168019 Use the size range from the crtc config instead of randrp->virtual[XY] when 168020 reporting the min and max screen sizes to the DDX. 168021 168022commit b11dfac287d65de7b83f63749087cba4e8ddaf4a 168023Author: Matthias Hopf <mhopf@suse.de> 168024Date: Fri Mar 2 12:30:26 2007 +0100 168025 168026 Legacy framebuffer support wasn't compiled if Xorg wasn't explicitly enabled. 168027 168028commit 57aa5e908dc11d5d8c27ed705c526f1416c1e8ad 168029Author: Peter Hutterer <peter@cs.unisa.edu.au> 168030Date: Fri Mar 2 17:14:37 2007 +1030 168031 168032 dix, Xext, Xtrap, Xi: replace inputInfo.pointer with PickPointer where 168033 possible. More replacements to come. 168034 168035commit e43b8a4e40991ca6e545bda4cf9b9bd7a2bf22e2 168036Author: Peter Hutterer <peter@cs.unisa.edu.au> 168037Date: Fri Mar 2 15:21:46 2007 +1030 168038 168039 dix: Add ClientPointer to client, used for picking which pointer to use in 168040 ambiguious request. PickPointer and PickKeyboard are used for getting 168041 the appropriate pointer when situation is unclear. 168042 Fix some issues with InitializeSprite. 168043 168044 dix, xfree86: Remove last traces of InitSprite. 168045 168046commit 2dafc46e3d814e02b25e5a2fa2e931f0257402a8 168047Author: Ben Byer <bbyer@bbyer.apple.com> 168048Date: Thu Mar 1 17:44:39 2007 -0800 168049 168050 Fixed pointer events in Xquartz -- Keyboard events work, but 168051 the keycodes are incorrect. 168052 168053commit 74f1de1de9633119c2cf26086875717181c8a6f7 168054Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168055Date: Thu Mar 1 12:07:33 2007 -0500 168056 168057 devPrivates rework: unhook resource system; will try a different approach. 168058 168059commit 39ecd6fff4f946deebe310b4b26b171c842db223 168060Author: Ben Byer <bbyer@bbyer.(none)> 168061Date: Thu Mar 1 01:45:19 2007 -0800 168062 168063 Rewrote parts of the Xquartz event-handling code (thanks daniels and whot!) 168064 It should still be considered a work in progress, but mouse events almost work. 168065 168066commit ed7ccc481ad1caaa518cafe944c2327a5d0b6c65 168067Author: Ben Byer <bbyer@bbyer.(none)> 168068Date: Thu Mar 1 00:51:10 2007 -0800 168069 168070 AIGLX support for Darwin -- works well enough to run 168071 glxgears and glxinfo, but still needs more testing. 168072 168073commit 43bd35fcf85b51b7e5714844bade90184dbd746f 168074Author: Peter Hutterer <peter@cs.unisa.edu.au> 168075Date: Thu Mar 1 17:19:37 2007 +1030 168076 168077 dix: Remove InitSprite, work can be done by InitializeSprite. 168078 Free sprite struct if a spriteOwner is paired. 168079 168080 xfree86: Use PairDevices instead of passing booleans around when creating a 168081 sprite. 168082 168083 Xext: Switch back to using LookupPointer/KeyboardDevice instead of 168084 inputInfo.xyz. 168085 168086commit 90ca76ba28fcd8bed945e33cf9674784fa2eb533 168087Author: Jay Cotton <jay.cotton@sun.com> 168088Date: Wed Feb 28 17:40:58 2007 -0800 168089 168090 Add sun_apm.c for Suspend-and-Resume support on Solaris 168091 168092 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6205248> 168093 168094commit 06c3021aec720837bef432656e88ae9b4e35101d 168095Author: Aaron Plattner <aplattner@nvidia.com> 168096Date: Wed Feb 28 16:09:11 2007 -0800 168097 168098 Don't crash setting a NULL mode with a randr classic DDX. Also remember to update the screen size during modesets. 168099 168100commit 68c64ad7b1eea79c786b5a7f3459076780163a47 168101Author: Peter Hutterer <peter@cs.unisa.edu.au> 168102Date: Thu Mar 1 09:51:20 2007 +1030 168103 168104 Xext: Update device's lastx/lasty when sending a motion event with XTest. 168105 168106commit 8b245758845523d5f8f017bb9d0e9aa57b616c28 168107Author: Aaron Plattner <aplattner@nvidia.com> 168108Date: Mon Feb 26 17:45:40 2007 -0800 168109 168110 Return BadMatch if a client tries to clone non-cloneable outputs. 168111 168112commit 2a35d44b6d87f96a7ce90dc8f0142f48cc176e95 168113Author: Peter Hutterer <peter@cs.unisa.edu.au> 168114Date: Wed Feb 28 15:43:06 2007 +1030 168115 168116 Update device valuators on XTest motion event, otherwise the cursor jumps 168117 between coordinates stored in device and coordinates sent by the event. 168118 168119commit d9bcb22d199e8444b9762a35754e04d327dd5915 168120Merge: 272d9341d c16343ac2 168121Author: Ben Byer <bbyer@bbyer.apple.com> 168122Date: Tue Feb 27 16:28:20 2007 -0800 168123 168124 Merge branch 'master' of git+ssh://bbyer@git.freedesktop.org/git/xorg/xserver 168125 168126commit 272d9341d0f7c3e9e0c9b9a8c0d4d8779cdcc5cf 168127Author: Ben Byer <bbyer@bbyer.apple.com> 168128Date: Tue Feb 27 16:27:26 2007 -0800 168129 168130 fix for hw/darwin/Makefile.am to properly use XSERVER_LIBS 168131 168132commit c16343ac2ca18391b21022b2edd02ad9f413d2b3 168133Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168134Date: Tue Feb 27 14:14:47 2007 -0500 168135 168136 Make mfb, cfb, and afb support configurable at build-time. 168137 168138commit 5680efc0d2baf0a9451e82e490e3690fc23dda0f 168139Author: Alan Coopersmith <alan.coopersmith@sun.com> 168140Date: Tue Feb 27 09:55:48 2007 -0800 168141 168142 Sun bug 6529003: Xorg should not be including <sys/immu.h> on Solaris 168143 168144 <sys/immu.h> was removed from the latest Solaris Nevada build, but it's 168145 been useless to Xorg for a long time (it only declared a couple of kernel 168146 variables) 168147 <http://bugs.opensolaris.org/view_bug.do?bug_id=6529003> 168148 168149commit ab0fc8c1ad7ea2dc3389a4a4bb1c45bbded5e7ad 168150Author: Ben Byer <bbyer@bbyer.(none)> 168151Date: Tue Feb 27 00:14:35 2007 -0800 168152 168153 verbiage corrected per daniels 168154 168155commit cdd4c84572cc3bdd004f8dca6d8b64e710344ac0 168156Author: Ben Byer <bbyer@bbyer.(none)> 168157Date: Mon Feb 26 23:57:02 2007 -0800 168158 168159 added hw/darwin/README.apple file with some todo items and props. 168160 168161commit 776d4d6587c57f94bca8732f915d07a0d4e137c8 168162Author: Ben Byer <bbyer@bbyer.(none)> 168163Date: Mon Feb 26 23:40:00 2007 -0800 168164 168165 X11.app now builds correctly 168166 168167commit 154d2c13f4ec22b7e6332808bbcd049d63784891 168168Author: Ben Byer <bbyer@bbyer.apple.com> 168169Date: Mon Feb 26 19:39:26 2007 -0800 168170 168171 more changes for X11.app 168172 168173commit fa06e11f972e2a75c84b2f1586997ffc1239cbd9 168174Author: Ben Byer <bbyer@bbyer.apple.com> 168175Date: Mon Feb 26 17:06:53 2007 -0800 168176 168177 added hw/darwin/apple directory, which contains source and data files to build 168178 a version of the X11.app shipped with Mac OS X, using the X.org server. 168179 168180commit a16360733ea393ec1fc267e88fc604d9d7534484 168181Author: Jay Estabrook <Jay.Estabrook@hp.com> 168182Date: Sun Feb 25 19:58:26 2007 +0000 168183 168184 Fix root bus/domain selection on alpha 168185 168186commit 566610680c4e1cab3e7fc7146adbeaba52fdd0ad 168187Author: Adam Jackson <ajax@benzedrine.nwnk.net> 168188Date: Fri Feb 23 15:20:35 2007 -0500 168189 168190 Don't install libi2c.a 168191 168192commit 81372f9096b952f4be545654b0b44ac37ef4f2c2 168193Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168194Date: Fri Feb 23 13:23:12 2007 -0500 168195 168196 devPrivates rework: hook up new interface in resource system; add new 168197 resource-adding function that takes an additional ScreenPtr argument. 168198 168199commit 16f2b8892d9ebcef6410a675d10549043223f617 168200Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168201Date: Fri Feb 23 13:20:43 2007 -0500 168202 168203 devPrivates rework: add new interface implementation. 168204 168205commit 779faccfb78648a9f7e70b77dcfa9f6e19559772 168206Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168207Date: Fri Feb 23 13:19:53 2007 -0500 168208 168209 devPrivates rework: add dix header file containing new interface. 168210 168211commit bc4ae25433e38a25a2012f9f233d500665172a4b 168212Author: Peter Hutterer <peter@cs.unisa.edu.au> 168213Date: Fri Feb 23 12:06:51 2007 +1030 168214 168215 xfree86: Don't initialize sprites for devices that don't own a sprite (stops 168216 keyboard hotplug segfaults) 168217 168218commit af550ea91c451cf4f831c2413266a19d1f211d0e 168219Author: Alan Coopersmith <alan.coopersmith@sun.com> 168220Date: Thu Feb 22 14:38:40 2007 -0800 168221 168222 Move SecurityPolicy file format from Xserver(1) to it's own man page 168223 168224 Don't make users looking for Xserver information page through pages of 168225 details only interesting to the handful of people writing security policies. 168226 168227commit b1142cdbce76fed8cb22ba6d7ac027751dd56a76 168228Author: Brice Goglin <brice.goglin@ens-lyon.org> 168229Date: Thu Feb 22 12:26:04 2007 -0800 168230 168231 Bug #10034: 'man Xserver' typos: s/dqoute/dquote/ 168232 168233 Bugzilla #10034: <http://bugs.freedesktop.org/show_bug.cgi?id=10034> 168234 Patch #8780: <http://bugs.freedesktop.org/attachment.cgi?id=8780> 168235 168236commit 4b8b0e377a27ec904b2028c89aed11c6416af26c 168237Author: Peter Hutterer <peter@cs.unisa.edu.au> 168238Date: Thu Feb 22 20:00:59 2007 +1030 168239 168240 Xi: Add access control request handling. 168241 168242 dix: New file access.c to handle all access control for devices. 168243 168244commit 9a3eb0357e779d5d5f76858f23667956c4c5d721 168245Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168246Date: Fri Feb 16 19:30:03 2007 -0500 168247 168248 devPrivates rework: add RC_PRIVATES class, make ResourceRec visible in 168249 the API, and add extra fields and structure supporting private storage. 168250 168251commit cd0af7a7856e8246e27acc5513d219a094211625 168252Author: Peter Hutterer <peter@cs.unisa.edu.au> 168253Date: Wed Feb 21 10:08:41 2007 +1030 168254 168255 Xi, dix: Only one client at a time can change the pointer-keyboard pairing, 168256 using xRegisterPairingClient request. 168257 168258commit 3344a4eda704edc7dc30037f095de277a60a70bb 168259Author: Michel Dänzer <michel@tungstengraphics.com> 168260Date: Thu Feb 15 16:27:50 2007 +0100 168261 168262 DRI: Make sure number of DRI windows is accurate in driver ClipNotify hook. 168263 168264 Always call DRI{De,In}creaseNumberVisible (which in turn calls 168265 DRIDriverClipNotify) after updating pDRIPriv->nrWindows. 168266 168267commit 3c7a27dc77595ad018bb7c4f7cef6bc178268cb6 168268Author: Michel Dänzer <michel@tungstengraphics.com> 168269Date: Wed Feb 14 16:17:18 2007 +0100 168270 168271 DRI: New ClipNotify driver hook. 168272 168273 The hook is called whenever the clipList of any DRI window changes, be it via 168274 DRIClipNotify, DRICreateDrawable or DRIDrawablePrivDelete. This allows the 168275 driver to keep track of which DRI windows are visible where. 168276 168277commit eedf148e5a1273ebbf4dc8dcac9c435712fc00ea 168278Author: Michel Dänzer <michel@tungstengraphics.com> 168279Date: Fri Feb 2 18:27:40 2007 +0100 168280 168281 Track number of visible DRI windows separately for transitions. 168282 168283 This allows e.g. doing page flipping with multiple DRI windows as long as 168284 only one of them is visible. 168285 168286commit 8a42af6a935b1cf0e15102e986bb527f4fab31a8 168287Author: Keith Packard <keithp@neko.keithp.com> 168288Date: Mon Feb 19 15:28:37 2007 -0800 168289 168290 Check for clientGone before sending events from XFixes (bug #1753). 168291 168292 Freeing resources during client closedown can cause cursor changes which 168293 attempt to send cursor events through the XFixes extension; a client in the 168294 process of closing down has no file to send events to, causing a crash when 168295 this path is hit. 168296 168297commit 4660eaaffb36f526f71d5847ae1309c10ee133c6 168298Author: Ben Byer <bbyer@bbyer.(none)> 168299Date: Sun Feb 18 14:09:51 2007 -0800 168300 168301 configure fixes for darwin 168302 168303commit 5631a67f648f5f151a849a918ee12871c71c32e9 168304Author: Keith Packard <keithp@neko.keithp.com> 168305Date: Fri Feb 16 10:06:22 2007 -0800 168306 168307 Don't set subpixel order during startup; the screen won't be ready. 168308 168309 in xf86CrtcSetMode, scrn->pScreen will be NULL during server startup time, 168310 so don't try to set the subpixel order. subpixel order will be set in the 168311 randr initialization anyways. 168312 (cherry picked from commit 5f6f8616d862ce4a37f6d3df4bdbc44fd21cc82a) 168313 168314commit 096965ec9c7514f0c9fc0d17e5166f2d26781f87 168315Author: Keith Packard <keithp@guitar.keithp.com> 168316Date: Fri Feb 16 02:17:11 2007 -0800 168317 168318 Ensure drivers can use new modes header files. 168319 168320 New modes header files required a few minor changes to be used by external 168321 drivers, the most notable of which is the publication of the config file 168322 parser header files. 168323 168324commit 55797dd252382d35ebab5d9e18a5e0e77872d775 168325Author: Keith Packard <keithp@guitar.keithp.com> 168326Date: Fri Feb 16 00:56:00 2007 -0800 168327 168328 Respect rotation in initial screen size computation. 168329 168330commit e4507825bf0328ea59673f2bbe652de3a9105c86 168331Author: Keith Packard <keithp@guitar.keithp.com> 168332Date: Fri Feb 16 00:41:29 2007 -0800 168333 168334 Enable startup-time rotation; change rotation pixmap creation API. 168335 168336 Add monitor "Rotate" option taking one of "normal", "left", "inverted" or 168337 "right". However, because initial mode selection is made before the screen 168338 is completely initialized, we cannot create the shadow pixmap object at this 168339 point. Pend the shadow pixmap creation until the block handler. 168340 168341 Note that this code is not completely functional yet. 168342 168343commit 8606aeb9b2ab2dafc11e64436db4d3a7e67dbcc0 168344Author: Keith Packard <keithp@guitar.keithp.com> 168345Date: Thu Feb 15 22:23:16 2007 -0800 168346 168347 RRConfigureOutputProperty is a variable length request. 168348 168349 Replace REQUEST_SIZE_MATCH with REQUEST_AT_LEAST_SIZE 168350 168351commit a88844eccb0e423e71d4fcb286866a026308babd 168352Author: Daniel Stone <daniel@fooishbar.org> 168353Date: Sat Feb 17 20:35:07 2007 +0200 168354 168355 configure.ac: disable dmx per default 168356 168357 Disable DMX until it gets ported to the new input API. 168358 168359commit e9a2cc7d9fcc73e16576be2522522cce675dc3f3 168360Author: Daniel Stone <daniel@fooishbar.org> 168361Date: Sat Feb 17 16:17:48 2007 +0200 168362 168363 config: error message cleanup 168364 168365 Demote failure to connect from ErrorF to DebugF. 168366 168367commit 81876bc5ddc2f3eda5078fe4bd101917fb32e586 168368Author: Ben Byer <bbyer@apple.com> 168369Date: Sat Feb 17 04:07:11 2007 -0800 168370 168371 oops, missed a spot 168372 168373commit d287b76471f66c9aea54f969d050b35643cb2501 168374Author: Ben Byer <bbyer@apple.com> 168375Date: Sat Feb 17 03:47:42 2007 -0800 168376 168377 cleaned up some linking ugliness in hw/darwin/quartz 168378 168379commit 81444486be4f182dde778bac6f7edcbfc4368482 168380Author: Ben Byer <bbyer@apple.com> 168381Date: Sat Feb 17 02:23:11 2007 -0800 168382 168383 autoconf goodness for XDarwin, courtesy of pogma 168384 168385commit cf4994b0db2fef4c10ce8804adef766bc5118daf 168386Author: Ben Byer <bbyer@apple.com> 168387Date: Sat Feb 17 01:21:43 2007 -0800 168388 168389 dix mods for Darwin 168390 168391commit cece0601571f6304e392a3a40505664544b249f3 168392Author: Ben Byer <bbyer@apple.com> 168393Date: Sat Feb 17 01:00:13 2007 -0800 168394 168395 build fix for configure.ac / BUILD_DARWIN, oops 168396 168397commit 00b0657b815b95964401c3e36eed54063afbd003 168398Author: Ben Byer <bbyer@bbyer.(none)> 168399Date: Sat Feb 17 00:55:32 2007 -0800 168400 168401 glx fixes for XDarwin 168402 168403commit 93777c7b96e560da087963040e372aecbfca7bbc 168404Author: Ben Byer <bbyer@bbyer.(none)> 168405Date: Sat Feb 17 00:22:39 2007 -0800 168406 168407 more patches to make the Quartz part of XDarwin work again 168408 (thanks Peter and Torrey!) 168409 168410commit 612144c811fdf06b7c03cf48a321388fe411acd4 168411Author: Ben Byer <bbyer@bbyer.(none)> 168412Date: Sat Feb 17 00:09:58 2007 -0800 168413 168414 More build fixes / updates for XDarwin: 168415 quartz/cr: "Cocoa Rootless" support (deprecated in favor of xpr?) 168416 quartz/fullscreen: Fullscreen support using Xplugin (not yet functional) 168417 168418commit 68d39d8571d8717d26cedc84015d537549520a14 168419Author: Daniel Stone <daniel@fooishbar.org> 168420Date: Fri Feb 16 23:02:13 2007 +0200 168421 168422 kdrive/ephyr: fix keysym type confusion once and for all 168423 168424 Take keysyms in as an XID in hostx_load_keymap() and explicitly 168425 convert them to CARD32 for loading into the server. Fixes Xephyr on 168426 AMD64, wa-hey. 168427 168428commit 5507cb885d861e974be240120ada2ace2a980a72 168429Author: Daniel Stone <daniel@fooishbar.org> 168430Date: Fri Feb 16 23:01:27 2007 +0200 168431 168432 kdrive: delete input debugging, yet again ... 168433 168434 I have no idea how this keeps on coming back. 168435 168436commit 2534f5a9027c196f677923aaa38fa9ed9917f73d 168437Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168438Date: Fri Feb 16 15:33:48 2007 -0500 168439 168440 Remove nasty function pointer type from DevUnion, return to documented type. 168441 168442commit 84efe23ae834dd3a4d3f3e08832b69469c7382aa 168443Author: Ben Byer <bbyer@bbyer.(none)> 168444Date: Fri Feb 16 04:37:38 2007 -0800 168445 168446 updated darwin/quartz/xpr (libXplugin interface for Mac OS X) support 168447 168448commit 5e7f7436a755a33e48ab91831cc6af710a8344ef 168449Author: Ben Byer <bbyer@bbyer.(none)> 168450Date: Fri Feb 16 04:12:26 2007 -0800 168451 168452 merged in miext/rootless changes for XDarwin support 168453 168454commit f350909d1696fcfda87e8f12c729254d762313c9 168455Author: Keith Packard <keithp@guitar.keithp.com> 168456Date: Thu Feb 15 21:50:48 2007 -0800 168457 168458 Kludge around duplicate code added in hw/xfree86/modes. 168459 168460 Code added in hw/xfree86/modes came from the server-1.3-branch. 168461 Portions of this code had previously been integrated into xf86Mode.c 168462 and edid_modes.c. 168463 168464 To preserve hw/xfree86/modes as much as possible, the duplicate code from 168465 the other files has been disabled; a more careful review would figure out 168466 where that code actually belonged. 168467 168468commit 258beebc77510f84fbea66d6ebf29c5097bd11db 168469Author: Keith Packard <keithp@guitar.keithp.com> 168470Date: Thu Feb 15 20:13:15 2007 -0800 168471 168472 Report correct RandR 1.0 sizeID. Report correct subpixel order. 168473 168474 RandR 1.0 sizeID must be computed the same way every time, so when reporting 168475 it in the ScreenChangeNotify event, just construct the usual 1.0 data block 168476 and use that. 168477 168478 subpixel geometry information can be computed by looking at the connected 168479 outputs and finding any with subpixel geometry and using one of those for 168480 the global screen subpixel geometry. This might be improved by reporting 168481 None if more than one screen has information and they conflict. 168482 168483commit ef6b1235fd7d6dc422e8a150c089496a8e648067 168484Author: Keith Packard <keithp@guitar.keithp.com> 168485Date: Thu Feb 15 11:27:35 2007 -0800 168486 168487 Allow new modes code to build inside drivers as well as server. 168488 168489 Use config.h for driver builds where xorg-config.h isn't available. 168490 168491commit 3dbe8f6b6ea32a9a137ad6e9235f74009b095bd8 168492Author: Tilman Sauerbeck <tilman@code-monkey.de> 168493Date: Thu Feb 15 17:51:01 2007 +0100 168494 168495 Distribute hw/xfree86/modes. 168496 168497commit d4eb4d065032112a38444e36f791cb468a5ca8f4 168498Author: Keith Packard <keithp@guitar.keithp.com> 168499Date: Thu Feb 15 20:36:20 2007 -0800 168500 168501 Merge crtc/output-based mode selection code. 168502 168503 This code comes from the intel driver, so there's no history in this tree. 168504 168505 As the crtc/output-based mode selection code uses ddc, the ddc and i2c 168506 modules have been merged into the server. Attempts to load them are safely 168507 ignored now. 168508 168509commit 37fe4c49dc3a5faf2d3d56112b6bd78453045f6a 168510Author: Peter Hutterer <peter@cs.unisa.edu.au> 168511Date: Fri Feb 16 09:57:57 2007 +1030 168512 168513 mi: Move WarpPointer event generation to miPointerMove to avoid duplicate 168514 events, cache event array allocation. 168515 168516commit c2f3f705f1db8ca78292912544a7e416116175f3 168517Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168518Date: Thu Feb 15 14:38:24 2007 -0500 168519 168520 Bug #6988: Change behavior of Security extension per user feature request. 168521 168522commit 811675733e97416c990e6dc9c19271b43d96248d 168523Author: Daniel Stone <daniel@fooishbar.org> 168524Date: Thu Feb 15 19:09:00 2007 +0200 168525 168526 os: fix client privates leak 168527 Minor leak here. Oops. 168528 168529commit 8f6961d385bda92703f18090cff551409d2710c9 168530Author: Daniel Stone <daniel@fooishbar.org> 168531Date: Thu Feb 15 19:08:46 2007 +0200 168532 168533 configure.ac: add xdarwin stubs 168534 Add stub AM_CONDITIONALs to at least fix the build. 168535 168536commit a3b62623b8aac56b219633bdb2c2f6de19b0580b 168537Author: Daniel Stone <daniel@fooishbar.org> 168538Date: Thu Feb 15 17:07:42 2007 +0200 168539 168540 change versioning for new server version scheme 168541 See: 168542 http://xorg.freedesktop.org/wiki/XDC2007Notes#head-2719037a1905516c45cf74f0e155c8703221e446 168543 168544commit f452507ca9209598401d15da0039aaa4e1fed1a3 168545Merge: 04632835b 811675733 168546Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168547Date: Thu Feb 15 09:47:20 2007 -0500 168548 168549 Merge branch 'master' into my-XACE-SELINUX 168550 168551commit 04632835b7402456fdcf6c8fa01970cd2804e27c 168552Merge: 88f89b9ac 9ecf79ca0 168553Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 168554Date: Thu Feb 15 09:27:05 2007 -0500 168555 168556 Merge branch 'master' into my-XACE-SELINUX 168557 168558commit 0f6dd4aea6176507dbe1c90c950d332fecbcaacb 168559Author: Daniel Stone <daniel@fooishbar.org> 168560Date: Thu Feb 15 16:14:57 2007 +0200 168561 168562 kdrive/ephyr: free screen struct 168563 Free screen->driver on screenFini, instead of just leaking it. 168564 168565commit 9ecf79ca0111dd899ca88dd54156f71013220fcc 168566Author: Ben Byer <bbyer@apple.com> 168567Date: Thu Feb 15 05:22:21 2007 -0800 168568 168569 Beginnings of an update Darwin driver 168570 168571commit 136bb4874aadf4a731d7eb8671e8bb641f9980a7 168572Author: Ben Byer <bbyer@apple.com> 168573Date: Thu Feb 15 05:14:38 2007 -0800 168574 168575 iokit support for XDarwin 168576 168577commit 3ead1afe78d2913f08c8144cb2d3813c6b159488 168578Author: Ben Byer <bbyer@apple.com> 168579Date: Thu Feb 15 05:09:29 2007 -0800 168580 168581 Beginning of patches to add XDarwin support to the modular tree; 168582 special thanks to Torrey Lyons and Peter O'Gorman for making this possible. 168583 168584 This is the automake framework for the XDarwin.app interface files. 168585 168586commit f6c3b9fa97ccf85e96f15435d564a1c261e40532 168587Author: Peter Hutterer <peter@cs.unisa.edu.au> 168588Date: Thu Feb 15 13:51:20 2007 +1030 168589 168590 xfree86: mouse and keyboard hotplugging support. New mouse devices will get 168591 new cursors, keyboard devices use VCP's focus until paired. 168592 168593commit d570ff7c81858a3174686b46a088f67563b4a2d5 168594Author: Peter Hutterer <peter@cs.unisa.edu.au> 168595Date: Wed Feb 14 17:09:33 2007 +1030 168596 168597 fix: WarpCursor needs to send MotionNotify. 168598 168599commit c4fe1bcce1c1e4822e688959b331b47a051d6e0a 168600Merge: e540d572c d570ff7c8 168601Author: Ian Romanick <idr@us.ibm.com> 168602Date: Wed Feb 14 15:39:52 2007 -0800 168603 168604 Merge branch 'master' of ssh+git://idr@git.freedesktop.org/git/xorg/xserver into pci-rework 168605 168606commit e540d572c5acba877b3ce01e7b31e399dac2d44a 168607Author: Ian Romanick <idr@us.ibm.com> 168608Date: Wed Feb 14 15:38:10 2007 -0800 168609 168610 Merge fix missed on previous commit. 168611 168612commit 81aa7f059d3cfd8d28420b7932b8ff7e06d67979 168613Author: Eric Anholt <eric@anholt.net> 168614Date: Wed Feb 14 12:48:15 2007 -0800 168615 168616 Add missing dirty marking in a couple of fallback cases in the exaGlyphs path. 168617 168618commit a5f19c5150a7b3dc2ff3ad759ee1a6ab0ad8925c 168619Author: Eric Anholt <eric@anholt.net> 168620Date: Wed Feb 14 10:39:46 2007 -0800 168621 168622 Mark sync when UploadToScreen succeeds in exaGlyphs(). 168623 168624commit a492d494f51caf15a5cb979dc335387486c105d1 168625Author: Alan Coopersmith <alan.coopersmith@sun.com> 168626Date: Tue Feb 13 18:32:59 2007 -0800 168627 168628 Update Xvfb man page: remove monolith build instructions, use /var/tmp instead of /usr/tmp 168629 168630commit 25104ed2e6050f86691104b34eeaf3147eb83c4a 168631Author: Peter Hutterer <peter@cs.unisa.edu.au> 168632Date: Wed Feb 14 11:26:53 2007 +1030 168633 168634 dix: fix typo in condition. 168635 168636commit 33ef546b942977bb2573156eaba5269350d16f99 168637Author: Peter Hutterer <peter@cs.unisa.edu.au> 168638Date: Wed Feb 14 11:18:43 2007 +1030 168639 168640 dix: Clean up GetPointerEvents. If pDev == inputInfo.pointer, assume that 168641 we do not want Xi events. 168642 168643 mi: Remove POINTER_CORE_ONLY flag. 168644 168645commit a23b0b069cac8a48e2b306b2095515d75f647705 168646Author: Adam Jackson <ajax@benzedrine.nwnk.net> 168647Date: Mon Feb 12 17:50:00 2007 -0500 168648 168649 Typo fix. 168650 168651commit d21c95f80bdba2f29eedd57fb0b00e580391c08e 168652Author: Adam Jackson <ajax@benzedrine.nwnk.net> 168653Date: Mon Feb 12 17:22:39 2007 -0500 168654 168655 Hook up --with-builderstring for vendor build identification. 168656 168657commit 46784d24c11767455a4986449a8037295912dcee 168658Author: Adam Jackson <ajax@benzedrine.nwnk.net> 168659Date: Mon Feb 12 17:18:29 2007 -0500 168660 168661 Remove spurious LIBADD from xf4bpp 168662 168663commit 16eb7254f861465f988ae3861ac3449c2c966062 168664Merge: 3814862a8 c4b7e9d1c 168665Author: Ian Romanick <idr@us.ibm.com> 168666Date: Fri Feb 9 12:38:49 2007 -0800 168667 168668 Merge branch 'master' of ssh+git://idr@git.freedesktop.org/git/xorg/xserver into pci-rework 168669 168670commit c4b7e9d1c16797c3e4b1200b40aceab5696a7fb8 168671Author: Aaron Plattner <aplattner@nvidia.com> 168672Date: Tue Feb 6 14:57:22 2007 -0800 168673 168674 Add an RDTSC implementation to the x86 emulator. 168675 168676 This instruction is being used in some debug VBIOSes. This implementation 168677 doesn't even try to be accurate. Instead, it just increments the counter by a 168678 fixed amount every time an rdtsc instruction in encountered, to avoid divides by 168679 zero. 168680 168681commit 262b9b104a04e55969593ef96a16004e53ecd00a 168682Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 168683Date: Tue Feb 6 17:30:22 2007 -0500 168684 168685 Use the new 8888x0565mmx function in fbpict.c 168686 168687commit 876b806ec09d5ff0c6cd19df91006c4eefedfaa6 168688Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 168689Date: Tue Feb 6 17:16:23 2007 -0500 168690 168691 Reapply patch to fix AMD CPU detection 168692 168693commit 13568d2aa43da4216bbcb46e1125ff28c323ac54 168694Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 168695Date: Tue Feb 6 17:12:01 2007 -0500 168696 168697 Revert "Fix for AMD cpu detection. Bug 9614, Dan Williams." 168698 168699 This reverts commit b2cd3b133748cc5aa541905a703a6fdb1cbbb1e6 since 168700 unrelated changes in fbpict.c broke the build. 168701 168702commit 5a3334410367a2186b2c667fa1eb6cf0baf93e95 168703Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 168704Date: Tue Feb 6 17:11:01 2007 -0500 168705 168706 Add new fbCompositeSrc_8888x0565mmx() function, based on patch by Dan 168707 Williams. Bug 9682. 168708 168709commit b2cd3b133748cc5aa541905a703a6fdb1cbbb1e6 168710Author: Soren Sandmann Pedersen <ssp@dhcp83-218.boston.redhat.com> 168711Date: Tue Feb 6 16:43:37 2007 -0500 168712 168713 Fix for AMD cpu detection. Bug 9614, Dan Williams. 168714 168715 Credit for the fixes in this patch goes to: 168716 168717 Marco Gritti <mpg at redhat dot com> 168718 Jordan Crouse <jordan dot crouse at amd dot com> 168719 168720commit 7cd73b00a29469d37a7f526cc9b35cc927a13ae2 168721Author: Peter Hutterer <peter@cs.unisa.edu.au> 168722Date: Mon Feb 5 14:48:48 2007 +1030 168723 168724 dix: If POINTER_CORE_ONLY is set, GetPointerEvents() only creates a core event 168725 168726 mi: fix: Create motion event in miPointerMoved 168727 168728commit 760a38c4c7ab66ae653d3acb92f5cda4bd44edd6 168729Author: Daniel Stone <daniel@fooishbar.org> 168730Date: Mon Feb 5 03:39:36 2007 +0200 168731 168732 XkbCopyKeymap: fix copy-and-waste accident 168733 168734 When we reallocated modmap, we accidentally clobbered syms with the 168735 result, leaving syms definitely too small, and modmap also potentially too 168736 small (as well as not actually allocated anymore). 168737 168738commit 17d85387d1e6851d35474b65929e268ca64ef65b 168739Author: Daniel Stone <daniel@fooishbar.org> 168740Date: Thu Jan 18 15:23:57 2007 +1100 168741 168742 dmx, vfb, xnest: fix fbcmap compilation 168743 168744 Don't always define XFree86Server, but only for damn fbcmap.c. 168745 Split fbcmap.c into its own library to achieve this. 168746 168747commit a309c936bb79e2fea04a96ce33c7da99ed902484 168748Author: Peter Hutterer <peter@cs.unisa.edu.au> 168749Date: Mon Feb 5 11:38:44 2007 +1030 168750 168751 xkb: Daniel's patch to stop evdev keyboard segfaults. 168752 168753 mi: Remove quickfix from ba547d4f6a2707f51f0d826389d2d418fb62367a 168754 168755commit 236f04b638e7d4d1656c6bedd8a6e8d7cec285ec 168756Author: Dave Airlie <airlied@linux.ie> 168757Date: Mon Feb 5 09:09:12 2007 +1100 168758 168759 remove array_cache from everywhere 168760 168761commit eb228e8d1eaa78911541b2fec5d04a74c1299718 168762Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 168763Date: Sun Feb 4 22:06:59 2007 +0000 168764 168765 clean up more of the vbo fallout 168766 168767commit fb1bc1c65b88527b42a0e4abed23e5ddaae711b7 168768Author: Dave Airlie <airlied@linux.ie> 168769Date: Sun Feb 4 18:39:58 2007 +1100 168770 168771 add vbo to .gitignore 168772 168773commit d8e148ec841d340327e6813127b0e0ffc4db712d 168774Author: Dave Airlie <airlied@linux.ie> 168775Date: Sun Feb 4 18:39:04 2007 +1100 168776 168777 update xserver for vbo code in mesa 168778 168779commit 5dcad9e9d7d9993d65f989219bee94a060bbf476 168780Author: Alan Coopersmith <alan.coopersmith@sun.com> 168781Date: Fri Feb 2 14:44:55 2007 -0800 168782 168783 Fix bus error on startup in 64-bit Xephyr 168784 168785 hostx_get_visual_masks takes unsigned long * arguments, but was being 168786 passed pointers to CARD32's. 168787 168788commit 170a55022ebc7b148bff93886eda152a0d5ce79a 168789Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 168790Date: Fri Feb 2 20:56:12 2007 +0000 168791 168792 remove file 168793 168794commit e6a505be84f5f72349d6860dc5a5058367516019 168795Author: Dan Nicholson <dan@conor.dwcab.com> 168796Date: Fri Feb 2 20:53:01 2007 +0000 168797 168798 The array_cache sources don't exist anymore in the Mesa tree, 168799 so we shouldn't try to build them. 168800 168801commit af20485ec370801f2aabfaeae17bbd030a849bd1 168802Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 168803Date: Fri Feb 2 19:14:46 2007 +0000 168804 168805 Remove array_cache for recent Mesa changes 168806 168807commit cf5b29d75dad7c74543f49f010c817623a3df747 168808Author: George Sapountzis <gsap7@yahoo.gr> 168809Date: Fri Feb 2 12:57:38 2007 +0200 168810 168811 dmx: drop leftover __GLXdrawablePrivateRec struct. 168812 168813commit 4f2f3233c808fd86bf9f6c09937feda9e0b367fd 168814Author: Eric Anholt <eric@anholt.net> 168815Date: Thu Feb 1 15:10:29 2007 -0800 168816 168817 Fix the size expectations of xRRSetCrtcGamma. 168818 168819 It was using REQUEST_SIZE_MATCH (client request length must equal request size) 168820 rather than REQUEST_AT_LEAST_SIZE (client request length must be at least 168821 big enough for request size), and this request has data following the request 168822 structure. 168823 168824commit 8274ea6aa97b06a56b7468c3908894c0ff72b687 168825Author: Eric Anholt <eric@anholt.net> 168826Date: Thu Feb 1 12:15:54 2007 -0800 168827 168828 Set the Damage version supported in the server, instead of using damageproto. 168829 168830 This was caught by distributions upgrading damageproto to 1.1, before the 168831 server they had supported it. The server then advertised the new version 168832 without supporting the protocol. 168833 168834commit 8bce182568f14edfb03911d8c5d791fd83bb6222 168835Author: Eric Anholt <eric@anholt.net> 168836Date: Mon Jan 29 17:30:59 2007 -0800 168837 168838 Restore a few important lines killed in the previous commit. 168839 168840 Typical results were failure to sync, and a black screen. 168841 168842commit 3814862a869ee83d307eb01225d5949039f435d8 168843Merge: a216de9b7 31f2d4a57 168844Author: Ian Romanick <idr@us.ibm.com> 168845Date: Mon Jan 29 15:14:31 2007 -0800 168846 168847 Merge branch 'master' of ssh+git://idr@git.freedesktop.org/git/xorg/xserver into pci-rework 168848 168849 Conflicts: 168850 168851 hw/xfree86/os-support/bus/linuxPci.c 168852 168853commit 31f2d4a57e04f5ea635fbb50c508405c4fc37b65 168854Author: Eric Anholt <eric@anholt.net> 168855Date: Mon Jan 29 09:39:33 2007 -0800 168856 168857 Bug #9680: Remove bogus blank length limiting in xf86SetModeCrtc(). 168858 168859 Our modes typically come from EDID or default modes, and when the monitor 168860 asks for a specific mode, deciding to tweak it usually results in incorrect 168861 display. And if the user is specifying a mode by hand, tweaking it then is 168862 still pretty rude. 168863 168864 Reviewed by: ajax 168865 168866commit 4aaaa70d1b52346213fad46777c006a93c4ece5d 168867Author: Peter Hutterer <peter@cs.unisa.edu.au> 168868Date: Mon Jan 29 18:23:24 2007 +1030 168869 168870 Xi: Adding ChangePointerKeyboardPairing request 168871 Adding PointerKeyboardPairingChanged event 168872 Correct error values for XWarpDevicePointer 168873 168874 dix: Adding device argument to SendMappingNotify 168875 Adding spriteOwner flag to devices 168876 168877commit f3418b52dcf2ab4982504856ab9fae3e726ee6d2 168878Author: Peter Hutterer <peter@cs.unisa.edu.au> 168879Date: Mon Jan 29 16:10:03 2007 +1030 168880 168881 mi: Fix cursor rendering issues. 168882 168883commit 15a81b6325d359990017b8e9f17ce18a7eff1354 168884Merge: 3759254c0 cc3d66d4a 168885Author: Peter Hutterer <peter@cs.unisa.edu.au> 168886Date: Sun Jan 28 17:18:57 2007 +1030 168887 168888 Merge branch 'master' 168889 168890commit cc3d66d4a53fee8f1d940bb8c7f032c94b9c25f6 168891Author: Peter Hutterer <peter@cs.unisa.edu.au> 168892Date: Wed Nov 15 17:50:02 2006 +1030 168893 168894 o fix minor error in comment for GetPointerEvents() 168895 168896commit 3759254c0967ae83bdcbf097d9e8a58733c8e6f9 168897Merge: ba547d4f6 ca5ebe397 168898Author: Peter Hutterer <peter@cs.unisa.edu.au> 168899Date: Sun Jan 28 17:05:55 2007 +1030 168900 168901 Merge branch 'master' 168902 168903commit 1627af54497bee659ea30f2850b39cbbf576e22d 168904Author: Jonathan Lim <jlim@sgi.com> 168905Date: Fri Jan 26 13:00:45 2007 +0100 168906 168907 Call linuxPciOpenFile() for r/w access if applicable. 168908 168909 Currently, the call to linuxPciOpenFile() is always made for read 168910 only access which causes the subsequent mmap call to fail when the 168911 memory is mapped read/write. 168912 168913 Xorg #9692 168914 168915commit a216de9b7ff55e2b73c487d037f248f00bd2e63b 168916Merge: 24506ea65 cf7ca9d09 168917Author: Ian Romanick <idr@us.ibm.com> 168918Date: Thu Jan 25 10:17:32 2007 -0800 168919 168920 Merge branch 'master' of ssh+git://idr@git.freedesktop.org/git/xorg/xserver into pci-rework 168921 168922commit cf7ca9d09cba14d107152a5179de38e5ef7bd784 168923Author: Alan Coopersmith <alan.coopersmith@sun.com> 168924Date: Wed Jan 24 20:20:48 2007 -0800 168925 168926 Plug memory leak in doLoadModule() 168927 168928commit 5abd50e37ceda134897891ed32e05215db67e0b4 168929Author: Alan Coopersmith <alan.coopersmith@sun.com> 168930Date: Wed Jan 24 18:54:38 2007 -0800 168931 168932 Correct help lines for configure's --with-vendor-name flags 168933 168934commit b32a40817fc0e2ac2edf2fa22a8813087fce2e7b 168935Author: Alan Coopersmith <alan.coopersmith@sun.com> 168936Date: Wed Jan 24 16:29:49 2007 -0800 168937 168938 Correct variable descriptions in comment for SecurityCheckResourceIDAccess 168939 168940commit 24506ea65be4cb29c5e1486aa0a529a40ce5c230 168941Author: Ian Romanick <idr@us.ibm.com> 168942Date: Wed Jan 24 15:33:49 2007 -0800 168943 168944 Move xf86ReadLegacyBIOS to the one place that uses it. 168945 168946 xf86ReadLegacyBIOS is only used by one function in int10/generic.c. 168947 Move a generic implementation of that function there, rename it to 168948 read_legcay_BIOS, and delete all remnants of it from all other places. 168949 168950commit fdb3a0798f0d17e72ec7293d59a7a81b5ffdf95b 168951Merge: 39b51e1bf a53586eeb 168952Author: Ian Romanick <idr@us.ibm.com> 168953Date: Wed Jan 24 14:49:39 2007 -0800 168954 168955 Merge branch 'master' of ssh+git://idr@git.freedesktop.org/git/xorg/xserver into pci-rework 168956 168957commit a53586eebc166e35c1f48942205832810061daee 168958Author: Eric Anholt <eric@anholt.net> 168959Date: Wed Jan 24 13:36:25 2007 -0800 168960 168961 Warning fix for RRCrtcSetRotations(). 168962 168963commit 7a12952fd437b105ea0d013d680f9c3a775a183c 168964Author: Eric Anholt <eric@anholt.net> 168965Date: Wed Jan 24 13:34:29 2007 -0800 168966 168967 Bug #7639: Only swap out pixmaps (rather than everything) on VT switch in EXA. 168968 168969 This is a new behavior for version 2.1 of EXA, and only takes effect if the 168970 driver has requested that. Otherwise, the previous behavior remains the same. 168971 168972commit b6b855932109b4bc3454f07bef8cb079d79ca369 168973Author: Keith Packard <keithp@keithp.com> 168974Date: Thu Jan 25 00:29:20 2007 +0800 168975 168976 Make Xinearama screen information reflect CRTC rotation. 168977 168978commit 39b51e1bfc4924fc3bda4342ef9c6c0125d9f4fa 168979Merge: 637b19b3e 788cfce91 168980Author: Ian Romanick <idr@us.ibm.com> 168981Date: Tue Jan 23 13:25:56 2007 -0800 168982 168983 Merge branch 'master' of ssh+git://idr@git.freedesktop.org/git/xorg/xserver into pci-rework 168984 168985commit 788cfce911793a26aed16f38f30678ecee82c873 168986Author: Michel Dänzer <michel@tungstengraphics.com> 168987Date: Tue Jan 23 10:15:22 2007 +0100 168988 168989 Bump video driver ABI version to 1.2. 168990 168991 This is necessary because server-1.2-branch bumped to 1.1 for xf86CVTMode and 168992 we have xf86XVFillKeyHelperDrawable on top of that. 168993 168994commit 637b19b3ee7345c2d449ce94ceabe998d817c02e 168995Author: Ian Romanick <idr@us.ibm.com> 168996Date: Mon Jan 22 09:58:33 2007 -0800 168997 168998 Eliminate pciControlBridge, pciGetBridgeBuses, and pciGetBridgeResources. 168999 169000commit a232e4ae938fa3a68f7614995676eda7929260db 169001Author: Ian Romanick <idr@us.ibm.com> 169002Date: Mon Jan 22 09:16:45 2007 -0800 169003 169004 Convert xf86GetPciHostConfigFromTag to get_parent_bridge. 169005 169006 Convert xf86GetPciHostConfigFromTag to a new function called 169007 get_parent_bridge. This name better represents what 169008 xf86GetPciHostConfigFromTag is used for: walking up the lists of PCI 169009 bridges from a device. 169010 169011commit 844560a02fa6c4ce18ea2af3ec27beaa60b7af11 169012Author: Ian Romanick <idr@us.ibm.com> 169013Date: Mon Jan 22 09:13:59 2007 -0800 169014 169015 Elminiate ARCH_PCI_OS_INIT. 169016 169017 Never, ever use the ix86Pci stuff on Linux. This renders the whole 169018 ARCH_PCI_OS_INIT mechanism useless. Remove it. 169019 169020commit 7dfb3cea913b02a6b36c308d1330bd40abe6b41f 169021Merge: c279d5fdc 2dc866252 169022Author: Ian Romanick <idr@us.ibm.com> 169023Date: Mon Jan 22 08:44:36 2007 -0800 169024 169025 Merge branch 'master' of ssh+git://idr@git.freedesktop.org/git/xorg/xserver into pci-rework 169026 169027commit 2dc866252c84ed0e7b3afa25e8a5312f448d405b 169028Author: Eric Anholt <eric@anholt.net> 169029Date: Mon Jan 22 08:41:50 2007 +0800 169030 169031 Really fix optimized render cases being hit when they shouldn't. 169032 169033 I don't know how this define slipped in there. Fixes 169034 6fdfd9dad91d7b7aa292f8c4d268dd27c34de8d3. 169035 169036commit 88f89b9ac1b92a0916c46488350ff68c3ffdd490 169037Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169038Date: Fri Jan 19 19:15:49 2007 -0500 169039 169040 Policy updates: use x prefix in property and ext types. 169041 169042commit 2fb8b7f8199c35ae0870cb54b40ee28a4e01d479 169043Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169044Date: Fri Jan 19 19:14:51 2007 -0500 169045 169046 Split ObjectSIDByLabel into two functions since property labeling now 169047 involves an additional compute_create lookup. 169048 169049commit 0d6d373af95d0004d33b987d14ad7e04dd5d2003 169050Author: Alan Coopersmith <alan.coopersmith@sun.com> 169051Date: Fri Jan 19 14:52:23 2007 -0800 169052 169053 Update Xserver man page to match commit ed33c7c98ad0c542e9e2dd6caa3f84879c21dd61 169054 169055 Remove unused -xkbdb and -noloadxkb options. Rename -ar1 and -ar2 to 169056 -ardelay and -arinterval, respectively. 169057 169058commit c279d5fdc53612a90f33fafe9e9c59ced008dd51 169059Merge: 88f248e67 14d6a9b32 169060Author: Ian Romanick <idr@us.ibm.com> 169061Date: Fri Jan 19 12:59:54 2007 -0800 169062 169063 Merge branch 'master' of ssh+git://idr@git.freedesktop.org/git/xorg/xserver into pci-rework 169064 169065commit 700fccf863593cbea1691789f1f1cafc08a32fee 169066Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169067Date: Fri Jan 19 14:56:38 2007 -0500 169068 169069 Remove the root window context line from the configuration file. 169070 This context will be derived through a type_transition rule instead. 169071 169072commit 3a9791b456f35adb252a9059b19265c6c447f1ba 169073Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169074Date: Fri Jan 19 14:53:09 2007 -0500 169075 169076 Policy updates. 169077 169078commit 14d6a9b327381a6bb2dac59c62728e5fd0f0bcfb 169079Author: Michel Dänzer <michel@tungstengraphics.com> 169080Date: Fri Jan 19 18:30:21 2007 +0100 169081 169082 fbdevhw: Only deal with RGB weight if default visual is True- or DirectColor. 169083 169084commit 27a01e100bff21ac0b70c6d72071d7226fc91264 169085Author: Michel Dänzer <michel@tungstengraphics.com> 169086Date: Fri Jan 19 18:28:05 2007 +0100 169087 169088 fbdevhw: Consider mode set equal to mode requested if virtual width is larger. 169089 169090commit 65f4690ecb4576f60396fcccff8e5bd5d4b6645f 169091Author: Michel Dänzer <michel@tungstengraphics.com> 169092Date: Fri Jan 19 17:54:03 2007 +0100 169093 169094 __glXDRIscreenProbe: Use drmOpen/CloseOnce. 169095 169096 Fixes https://bugs.freedesktop.org/show_bug.cgi?id=9275 . Based on patch from 169097 Alan Swanson. 169098 169099commit 8b3a591cd39f2d51209dc71b641cac79663e1b16 169100Author: Alan Coopersmith <alan.coopersmith@sun.com> 169101Date: Thu Jan 18 16:03:30 2007 -0800 169102 169103 Update pci.ids to 2007-01-18 snapshot 169104 169105 (includes a whole bunch of ATI device id updates) 169106 169107commit 0f0c321adf2850b3d7aafe281362bfe424cb0ca1 169108Author: Alan Coopersmith <alan.coopersmith@sun.com> 169109Date: Thu Jan 18 15:31:53 2007 -0800 169110 169111 Make xf1bpp build correctly with compilers that don't support -include 169112 169113commit a811e92104028ae60ba69f73e32ee1e0533b088c 169114Author: Eric Anholt <eric@anholt.net> 169115Date: Thu Jan 18 14:28:01 2007 -0800 169116 169117 Account for CRTC rotation in the cursor containment code. 169118 169119commit df147c10ce597c56c16cbca552e8a3e3ecb3cdaa 169120Author: Alan Coopersmith <alan.coopersmith@sun.com> 169121Date: Wed Jan 17 16:47:07 2007 -0800 169122 169123 Xserver man page: remove bc, add -wr 169124 169125commit 2dfd1aab244a2c8da3b62b522b9a8434e474af17 169126Author: Alan Coopersmith <alan.coopersmith@sun.com> 169127Date: Wed Jan 17 14:39:28 2007 -0800 169128 169129 Always include compiler.h in cfbmskbits.h instead of checking #ifdef XFREE86 169130 169131commit 42a48786acf54f83167de4f561526986d4e27033 169132Author: Eric Anholt <eric@anholt.net> 169133Date: Wed Jan 17 14:34:42 2007 -0800 169134 169135 Add a setter for randr_crtc->rotations. 169136 169137commit 88f248e67143175532cbafd6faf8fc6df97c7894 169138Author: Ian Romanick <idr@us.ibm.com> 169139Date: Wed Jan 17 14:04:43 2007 -0800 169140 169141 Replace PciBusId with 'struct pci_device *' 169142 169143 There's no need to store the slot information for a PCI device as its 169144 ID. Instead, skip the middle man and just store a pointer to the 169145 pci_device structure. 169146 169147commit 0361611080267727f570e17f2212aaa890223f6e 169148Author: Ian Romanick <idr@us.ibm.com> 169149Date: Wed Jan 17 13:06:04 2007 -0800 169150 169151 Convert xf86IsolateDevice to a 'struct pci_slot_match'. 169152 169153 This change was accidentally not committed on the previous commit. 169154 169155commit c1808f148405a28256e0480d6f8714691b0e964e 169156Author: Ian Romanick <idr@us.ibm.com> 169157Date: Wed Jan 17 13:02:27 2007 -0800 169158 169159 Convert xf86IsolateDevice to a 'struct pci_slot_match'. 169160 169161commit d2f8c42c48ecc398d224d9e3b280a66042e68664 169162Author: Ian Romanick <idr@us.ibm.com> 169163Date: Wed Jan 17 12:59:17 2007 -0800 169164 169165 Convert xf86FindPrimaryDevice to use a static buffer and snprintf. 169166 169167 Rather than allocate a 9 byte buffer on each invocation, use a static 169168 16 byte buffer. Use snprintf for safety. This commit should probably 169169 be cherry-picked to the trunk. 169170 169171commit 28976bebec8b6475d0703f8356edd116e25d40b2 169172Author: Ian Romanick <idr@us.ibm.com> 169173Date: Wed Jan 17 12:15:09 2007 -0800 169174 169175 Eliminate unused fields in pciAccRec. 169176 169177commit f36a447d20c0a89b78c5899eeb303cab568ebac6 169178Author: Ian Romanick <idr@us.ibm.com> 169179Date: Tue Jan 16 14:49:19 2007 -0800 169180 169181 Eliminate use of Tag field. 169182 169183 The Tag field was removed from the int10 structures in commit 169184 ca9c41e09ddb72d150d86f5d372eb07eeca7c1a2. This file was over looked. 169185 169186commit cde17015dff1ced2aabb8b76c08f9110237821a5 169187Author: Eric Anholt <eric@anholt.net> 169188Date: Tue Jan 16 13:01:03 2007 -0800 169189 169190 When changing a non-pending property, call the screen rrOutputSetProperty hook. 169191 169192commit e3add7c8ecbb2a0a662860f208f6ae7d1857c717 169193Author: Eric Anholt <eric@anholt.net> 169194Date: Tue Jan 16 12:59:34 2007 -0800 169195 169196 Don't forget to add the property we configure to the properties list. 169197 169198commit 780b55ec6f97864b50f00170b201e93eba188f1b 169199Author: Ian Romanick <idr@us.ibm.com> 169200Date: Mon Jan 15 15:30:16 2007 -0800 169201 169202 Fix accidental commit (bug) in linuxOpenLegacy. 169203 169204commit 31bd8d150749c548eb70404395b4a52d46412f43 169205Author: Ian Romanick <idr@us.ibm.com> 169206Date: Mon Jan 15 15:19:30 2007 -0800 169207 169208 Fix domain insanity. 169209 169210 Eliminate xf86GetPciDomain. The domain from libpciaccess is the 169211 domain. Period. This means that 0 is a valid domain. Make sure that 169212 INCLUDE_XF86_NO_DOMAIN is *not* set. Always run in "domain mode," 169213 even if the only domain possible is 0. 169214 169215commit 079a6a55a516778871a8b2f8e9c8dd9162105a76 169216Author: Ian Romanick <idr@us.ibm.com> 169217Date: Mon Jan 15 11:39:24 2007 -0800 169218 169219 Convert to new pci_device_cfg_write_u* interface. 169220 169221commit d8bd0c41de11d12976a5ea9f185b7ac6c7060b71 169222Author: Ian Romanick <idr@us.ibm.com> 169223Date: Mon Jan 15 11:38:22 2007 -0800 169224 169225 Convert to new pci_device_cfg_write_u* interface. 169226 169227commit a87af8218f97da76433a5192d5e51aca0d5c3ed0 169228Author: Ian Romanick <idr@us.ibm.com> 169229Date: Mon Jan 15 11:21:37 2007 -0800 169230 169231 Remove unnecessary files from the sources list. 169232 169233commit 1c326bf75fc0d2c39507834d159d031291eddac4 169234Author: Ian Romanick <idr@us.ibm.com> 169235Date: Fri Jan 12 12:07:19 2007 -0800 169236 169237 Eliminate unused INCLUDE_XF86_MAP_PCI_MEM define. 169238 169239commit 3936caa00535d0b7543be5a4b315562111a6b3ab 169240Author: Ian Romanick <idr@us.ibm.com> 169241Date: Fri Jan 12 10:59:35 2007 -0800 169242 169243 Refactor pciTagConvertRange2Host and pciConvertRange2Host. 169244 169245commit 1a493509e30412a0f6c6f8f8e47a446678dbbc1a 169246Author: Ian Romanick <idr@us.ibm.com> 169247Date: Fri Jan 12 10:58:59 2007 -0800 169248 169249 Fix typos in conditionally compiled code. 169250 169251commit 8279444a54f38c5e2e5d4c2d936a10d74990e0be 169252Author: Ian Romanick <idr@us.ibm.com> 169253Date: Fri Jan 12 10:41:03 2007 -0800 169254 169255 Eliminate unused pciAddrHostToBus infrastructure. 169256 169257commit e32fa10a04a15ce1b15c6faa7a64e68ee3906d59 169258Author: Ian Romanick <idr@us.ibm.com> 169259Date: Fri Jan 12 10:33:17 2007 -0800 169260 169261 Remove prototype for pciGetBaseSize. 169262 169263commit 7411f5c389db73a6731d853d603029b8cf0a66ba 169264Author: Ian Romanick <idr@us.ibm.com> 169265Date: Fri Jan 12 10:32:04 2007 -0800 169266 169267 Eliminate pciHostAddrToBusAddr from interface. 169268 169269 pciHostAddrToBusAddr is no longer used anywhere in the tree. Remove 169270 it with extreme prejudice. 169271 169272commit 60c5e023e8ddaae8e011f80736e7cc3e8c12583d 169273Author: Ian Romanick <idr@us.ibm.com> 169274Date: Fri Jan 12 10:24:03 2007 -0800 169275 169276 Eliminate some dead code. 169277 169278commit dd12a94885bad9c9d064dcf7e0d9a7a7114ae3e6 169279Author: Ian Romanick <idr@us.ibm.com> 169280Date: Fri Jan 12 10:21:25 2007 -0800 169281 169282 Eliminate xf86GetPciEntity from public interface. 169283 169284 xf86GetPciEntity was folded into the only place (xf86CheckPciSlot) in 169285 the tree that still called it. 169286 169287commit c3016a29964761478305cf2f5241c563ff3eab0a 169288Author: Ian Romanick <idr@us.ibm.com> 169289Date: Fri Jan 12 09:56:00 2007 -0800 169290 169291 Missed file from previous commit. 169292 169293 Convert xf86ReadLegacyVideoBIOS to take a 'struct pci_device *' 169294 parameter instead of a PCITAG. 169295 169296commit ca9c41e09ddb72d150d86f5d372eb07eeca7c1a2 169297Author: Ian Romanick <idr@us.ibm.com> 169298Date: Thu Jan 11 21:09:20 2007 -0800 169299 169300 Convert int10 and vgaHW to use 'struct pci_device' instead of PCITAG. 169301 169302 Convert all uses of PCITAG in int10 and vgaHW to 'struct pci_device'. 169303 This allows the conversion of xf86ReadLegacyVideoBIOS and 169304 xf86MapDomainMemory to 'struct pci_device' from PCITAG. 169305 169306commit ba547d4f6a2707f51f0d826389d2d418fb62367a 169307Author: Peter Hutterer <peter@cs.unisa.edu.au> 169308Date: Fri Jan 12 14:42:02 2007 +1030 169309 169310 mi: Quick fix to stop segfault on non-core keyboards. 169311 169312commit 5e2f8dee6331645fcbd2734c43698eb4f9c5b116 169313Author: Peter Hutterer <peter@cs.unisa.edu.au> 169314Date: Thu Jan 11 14:05:39 2007 +1030 169315 169316 mi: remove core pointer special handling. No event has core pointer as 169317 device now. 169318 169319 dix: zero pSprite memory, stop segfaults on server restart. 169320 169321commit 7fccec91c46baac4f8d2965180dc535b4eb7d65c 169322Author: Eric Anholt <eric@anholt.net> 169323Date: Wed Jan 10 13:10:43 2007 -0800 169324 169325 Bug #9555: Always define _GNU_SOURCE in glibc environments. 169326 169327 This keeps us from having to define _POSIX_C_SOURCE, _BSD_SOURCE, and 169328 _XOPEN_SORUCE to get the C environment we want in different places. It also 169329 fixes the build on linux due to RTLD_DEFAULT having not been defined. 169330 169331commit 78f9592c112d4245f6119b98c244bbb4cae3e5aa 169332Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 169333Date: Wed Jan 10 16:04:20 2007 +0000 169334 169335 lnx_ev56.c has to be built with -mcpu=ev56. Fix it. 169336 169337commit ea82333dc1799a7e013205adbc89874bc8d273cb 169338Author: Peter Hutterer <peter@cs.unisa.edu.au> 169339Date: Wed Jan 10 14:52:13 2007 +1030 169340 169341 Xi: add missing call for SProcXChangeDeviceCursor 169342 add DeviceEnterNotify and DeviceLeaveNotify 169343 fix: QueryDevicePointer crashed if called with keyboard device 169344 169345commit 6a2fb2928714ce77ee342cdc23a1178e5e766cf2 169346Author: Eric Anholt <eric@anholt.net> 169347Date: Tue Jan 9 16:34:40 2007 -0800 169348 169349 Track rename of DamagePost -> DamageAdd. 169350 169351commit da09964a931cc64d05ab571bf545fdad35a6d395 169352Merge: e1f73d220 e3aa6ad20 169353Author: Ian Romanick <idr@us.ibm.com> 169354Date: Tue Jan 9 15:27:34 2007 -0800 169355 169356 Merge branch 'origin' into pci-rework 169357 169358 Conflicts: 169359 169360 hw/xfree86/int10/generic.c 169361 169362commit e3aa6ad201eb20862c11c000e76206e317a96dc9 169363Author: Matthieu Herrb <matthieu@blues.laas.fr> 169364Date: Tue Jan 9 14:14:19 2007 +0100 169365 169366 Multiple integer overflows in dbe and render extensions 169367 CVE IDs: CVE-2006-6101 CVE-2006-6102 CVE-2006-6103 169368 169369commit 359d20532bdcef6a540a551578d000afbb609c2d 169370Author: Michel Dänzer <michel@tungstengraphics.com> 169371Date: Tue Jan 9 09:53:45 2007 +0100 169372 169373 Require glproto >= 1.4.8 for GLX. 169374 169375 It builds against 1.4.7 as well, but it hardcodes the GLX_EXT_tfp tokens that 169376 were finalized in 1.4.8, so GLX_EXT_tfp breaks if the client side was built 169377 against an older glproto. This will hopefully alert people to rebuild other 169378 components (in particular Mesa) against the new glproto as well. 169379 169380commit 88740c4855babedbea420b5e1b35ae105d1f1026 169381Author: Alan Coopersmith <alan.coopersmith@sun.com> 169382Date: Mon Jan 8 17:36:07 2007 -0800 169383 169384 Use PKG_CHECK_EXISTS(libdrm) to determine if DRI should be enabled on Solaris 169385 169386commit 282a4dcaabc5f0cd6f7d3819aa648333b93b265e 169387Author: Michel Dänzer <michel@tungstengraphics.com> 169388Date: Mon Jan 8 19:22:41 2007 +0100 169389 169390 Attempt to fix drawable type checks in dixLookupDrawable(). 169391 169392 Not sure this is 100% correct either, but it fixes at least one reproducible 169393 crasher where it returned a pixmap to dixLookupWindow(). 169394 169395commit a7ab7932938820a795bb6fb8e0444e0824433b99 169396Author: Peter Hutterer <peter@cs.unisa.edu.au> 169397Date: Mon Jan 8 17:31:16 2007 +1030 169398 169399 xfree86: fix XI86_SHARED_POINTER flag setting 169400 169401commit 95e1a88050dde61e9b2407428042a43e47b46e18 169402Author: Peter Hutterer <peter@cs.unisa.edu.au> 169403Date: Mon Jan 8 12:31:18 2007 +1030 169404 169405 Xi: Adding ChangeDeviceCursor request 169406 169407commit 0b73a7eb17fd848c6bdc6a65ba835aa2cbfc3cfd 169408Author: Eric Anholt <eric@anholt.net> 169409Date: Fri Jan 5 18:12:04 2007 -0800 169410 169411 Add support for the DamagePost (XDamage 1.1) request. 169412 169413 This makes damageproto >= 1.1 a requirement to build. 169414 169415commit dfb2c10413e22afd8d486a982870f874326d5ef4 169416Author: Ian Romanick <idr@us.ibm.com> 169417Date: Fri Jan 5 10:15:09 2007 -0800 169418 169419 Add missing #else from previous commits. 169420 169421commit f90c3e226b105bf77beb94723fc08bdff14834be 169422Author: Ian Romanick <idr@us.ibm.com> 169423Date: Thu Jan 4 15:38:16 2007 -0800 169424 169425 Re-regenerate from Mesa scripts. 169426 169427 DO NOT HAND EDIT THESE FILES! For cryin' out loud, there's even a 169428 comment to that effect in the file's header... 169429 169430commit b7ca5d14ce7ba410b0dab5c2289f6d7b75e763df 169431Author: Ian Romanick <idr@us.ibm.com> 169432Date: Thu Jan 4 15:37:33 2007 -0800 169433 169434 Incorporate new byte-order related configure changes. 169435 169436commit 8dd5771a1b91c331860b667fb18e484452000aad 169437Merge: 45aa26ccb 7d2ec9217 169438Author: Ian Romanick <idr@us.ibm.com> 169439Date: Thu Jan 4 15:01:38 2007 -0800 169440 169441 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 169442 169443 Conflicts: 169444 169445 GL/glx/indirect_dispatch_swap.c 169446 GL/glx/swap_interval.c 169447 169448commit 45aa26ccb4f61c2919ce2475d0907c6e1b177da2 169449Author: Ian Romanick <idr@us.ibm.com> 169450Date: Thu Jan 4 14:55:51 2007 -0800 169451 169452 Regenerate from Mesa scripts. 169453 169454 Regenerate source files from Mesa scripts. This causes the generated 169455 files to use glxbyteorder.h. 169456 169457commit 6d603bb47ff9d238637adbf30c6e9697e6e7e6fa 169458Author: Ian Romanick <idr@us.ibm.com> 169459Date: Thu Jan 4 14:49:26 2007 -0800 169460 169461 Add new header file containing byte-order wrappers. 169462 169463 Move the byte-order related wrappers out of the individual source 169464 files into a dedicated header file. Modify the single hand-coded 169465 source file that uses the byte-order wrappers to use the new header 169466 file. 169467 169468commit 7d2ec92170ebbdfa10a05734cb7cfaac97d19d65 169469Author: Eric Anholt <eric@anholt.net> 169470Date: Thu Jan 4 12:24:48 2007 -0800 169471 169472 Keep track of how many visuals we set up for GLcore, to avoid an invalid free. 169473 169474 The proper fix would involve actually setting up the ARGB visual for GLcore, 169475 but I just want the server to not crash at exit. 169476 169477commit aab2ca204279b638c7e5bb6b8427c58be9704c57 169478Author: Eric Anholt <eric@anholt.net> 169479Date: Thu Dec 21 09:16:19 2006 -0800 169480 169481 Try dlsym(RTLD_DEFAULT) first when finding symbols. 169482 169483 The previous mechanism failed when finding drm symbols now that libdrm has 169484 moved to being linked by libdri instead of being linked into the server. 169485 169486commit 2fd4626fa6969b84d8e2f9db16d6e2d44c4bc499 169487Author: Alan Coopersmith <alan.coopersmith@sun.com> 169488Date: Wed Jan 3 15:44:55 2007 -0800 169489 169490 Make GLX byteswap macros more portable 169491 169492 - Use autoconf tests instead of platform-specific #ifdef's to decide 169493 which macros to use. 169494 - Provide fallbacks for platforms like Solaris that don't provide any 169495 of the existing known forms. 169496 169497commit 66fa87292ef26bd0f464481287f3af992cd5741c 169498Author: Aaron Plattner <aplattner@nvidia.com> 169499Date: Wed Jan 3 10:27:07 2007 -0800 169500 169501 Fix BSF and BSR instructions in the x86 emulator. 169502 169503 Patch courtesy of Michael Yaroslavtsev. 169504 169505commit 3b5b7ef5c2ab1d196806f6359e0972fd78d204dd 169506Author: Fredrik Höglund <fredrik@kde.org> 169507Date: Wed Jan 3 21:05:35 2007 +0100 169508 169509 Move the code for resetting the DPMS mode in response to input events, 169510 from WaitForSomething to mieqProcessInputEvents. 169511 169512 mieqProcessInputEvents already handles resetting the screen saver. 169513 169514commit 953a9ef949b4c57d28daeec57031fe1ce368c27c 169515Author: Keith Packard <keithp@guitar.keithp.com> 169516Date: Thu Dec 21 23:50:39 2006 -0800 169517 169518 Track physical screen size and send out updates when that changes. 169519 169520 Events and internal data structures need to be updated whenever the physical 169521 or pixel size of the screen changes. The code was ignoring the physical 169522 size, so changing only that would not be registered anywhere. 169523 (cherry picked from f42e3cea236fa0091ed398a818fc8e17b0e1b3df commit) 169524 169525commit e79602fca2f2cced66136729cdda4d356b0bdda0 169526Author: Keith Packard <keithp@neko.keithp.com> 169527Date: Sat Dec 30 21:52:22 2006 -0800 169528 169529 Use RRScreenSetSizeRange in 1.0 compat. Check RRGetInfo for error. 169530 169531 The RRScreenSizeSetRange function is used externally for 1.2 API drivers, 169532 but can also be used in the 1.0 compatibility code. This also ensures that 169533 the right changed bits are set so that clients are correctly notified when 169534 the range changes. 169535 169536 RRGetInfo can return an error, use that to return BadAlloc to clients 169537 instead of blindly going on with various requests. 169538 (cherry picked from f05dd384d38c76dd9662933a03625dfef5b1c81f commit) 169539 169540commit dc5eb4523298f966bd5fd9ae6672160034b5e82c 169541Author: Michel Dänzer <michel@tungstengraphics.com> 169542Date: Sun Dec 31 17:59:44 2006 +0100 169543 169544 fbdevhw: Override RGB offsets and masks after setting initial mode. 169545 169546 This is a hack, but it should be a NOP for all the setups that worked before 169547 and actually seems to fix some others... 169548 169549 Based on a patch by Peter Teichmann from 169550 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=338241 . 169551 169552commit d077c0da470ab7291e8d838eaace57b066477d6f 169553Author: Michel Dänzer <michel@tungstengraphics.com> 169554Date: Sun Dec 31 17:23:31 2006 +0100 169555 169556 fbdevhw: Use displayWidth for fbdev virtual width when appropriate. 169557 169558 The fbdev API doesn't allow setting the pitch explicitly, so we have to set 169559 the virtual width to the pitch we're using for drawing. This fixes corruption 169560 after changing the virtual width with RandR. 169561 169562commit c385bcf0bde38dd869f7065f859dd4b4126f5690 169563Author: Michel Dänzer <michel@tungstengraphics.com> 169564Date: Sat Dec 30 16:44:31 2006 +0100 169565 169566 fbdevhw: Fix some issues with the previous commit. 169567 169568 Fix a TRACE_ENTER typo and only update the internal fbdev mode state cache 169569 after actually setting a mode. 169570 169571commit f6815cb68b0f6698497348fc6e4214dacef33b95 169572Author: Michel Dänzer <michel@tungstengraphics.com> 169573Date: Sat Dec 30 10:18:28 2006 +0100 169574 169575 fbdevhw: Consolidate modeset ioctl calling, report failure if it modifies mode. 169576 169577 The fbdev API allows the driver to 'accept' modes it doesn't really support by 169578 modifying it to the nearest supported mode. Without this check, e.g. vesafb 169579 would appear to accept all modes, even though it actually can't set any modes 169580 other than the bootup mode at all. 169581 169582commit 083b790515faaf134a78abc4b0a7ef0d6ea5db75 169583Author: Eric Anholt <eric@anholt.net> 169584Date: Thu Dec 28 13:21:25 2006 -0800 169585 169586 Switch the default migration heuristic for EXA to "always". 169587 169588 This has been what has been used the most successfully post-damagetrack. 169589 The current thinking is that: 169590 1) We should be able to accelerate basically everything. So we don't need to 169591 try to migrate trees of pixmaps permanently out of framebuffer to speed 169592 CPU drawing up. 169593 2) Migration is cheaper in the thrashing case, so we don't want to go to a lot 169594 of effort to try (and fail badly) to find a working set. 169595 169596commit cfbc7379f0232bb336461f6d2a8496d3d0763e7e 169597Author: Eric Anholt <eric@anholt.net> 169598Date: Thu Dec 28 13:15:11 2006 -0800 169599 169600 Export exaMove{In,Out}Pixmap(). 169601 169602commit 683ca3f7afaf15fd3ca7918f6175b5a9e4a6f05b 169603Merge: 05f915050 9563b2eea 169604Author: Eric Anholt <eric@anholt.net> 169605Date: Wed Dec 27 16:11:31 2006 -0800 169606 169607 Merge branch 'exa-damagetrack' 169608 169609commit 05f915050cad72d4fb39cbb886be57beeac18749 169610Author: Peter Hutterer <peter@cs.unisa.edu.au> 169611Date: Wed Dec 27 16:38:06 2006 +0000 169612 169613 dix/events: take screen number, not pointer, in PostSyntheticMotion 169614 Since we were using PostSyntheticMotion incorrectly anyway, update the 169615 declared API to match. 169616 169617commit c1674660a7115ebf993dcde78f4e45f756e4c951 169618Author: Daniel Stone <daniel@fooishbar.org> 169619Date: Sun Dec 24 06:28:44 2006 +0200 169620 169621 os: test for userland, not kernel 169622 169623 It doesn't matter which kernel we're running on, the relevant part when 169624 dealing with includes is what our userland is. 169625 169626commit 83080809f9a1c1d24b0318e54632f25f5940da25 169627Author: Marc Aurele La France <tsi@ualberta.ca> 169628Date: Sun Dec 24 06:28:21 2006 +0200 169629 169630 xfree86: deal with pitch that isn't a multiple of the granularity 169631 169632 When the pitch isn't a multiple of the granularity, allocate more space to 169633 compensate. 169634 169635commit 329f6417275bb1201ba66c29b202028eeab3a355 169636Author: Daniel Stone <daniel@fooishbar.org> 169637Date: Sat Dec 9 22:51:59 2006 +0200 169638 169639 XkbCopyKeymap: make sure sym_interpret is always valid 169640 169641 Make sure we're not copying sym_interpret across from an empty source. 169642 169643commit 43a4376bd72ef1b6486cddb60ad57b2d6169292a 169644Merge: 4b1c9ac3d d9e079d2a 169645Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169646Date: Fri Dec 22 13:53:55 2006 -0500 169647 169648 Merge branch 'master' into my-XACE-SELINUX 169649 169650commit d9e079d2a385203fdd18d958cfc19d759cab4ba8 169651Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169652Date: Fri Dec 22 13:07:09 2006 -0500 169653 169654 Zero out client devPrivates on allocation. 169655 169656commit 4b1c9ac3d13767e395b47e76b37f9f3a569e7be1 169657Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169658Date: Fri Dec 22 13:04:50 2006 -0500 169659 169660 Policy updates. 169661 169662commit cd71e861830081807e5b93ae89c73c17986c6330 169663Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169664Date: Wed Dec 20 13:45:24 2006 -0500 169665 169666 Naming change: Security*Access -> Dix*Access. 169667 Clarify some error message strings. 169668 169669commit 9fd2f167ec02ee170b87f02ddca89eba0b2e2389 169670Author: Peter Hutterer <peter@cs.unisa.edu.au> 169671Date: Wed Dec 20 16:30:36 2006 +1030 169672 169673 xfree86: fix xf86ProcessCommonOptions() core and shared pointer assignment 169674 169675 TAG: mpx-0.3.0 169676 169677commit 056c919d849a78c8d507bfcecfa74c0b0165751a 169678Author: Peter Hutterer <peter@cs.unisa.edu.au> 169679Date: Wed Dec 20 14:49:17 2006 +1030 169680 169681 removing MPX extension files 169682 169683commit b55e1239ac3b2962a33d8af1f911cc0f01f40f23 169684Author: Peter Hutterer <peter@cs.unisa.edu.au> 169685Date: Wed Dec 20 14:43:38 2006 +1030 169686 169687 xfree86: fix xf86ActivateDevice(): only set to MP device if not a keyboard 169688 169689commit 202b46eb6bf0d5f94973c2bf1e4ebe9d154eadbf 169690Author: Peter Hutterer <peter@cs.unisa.edu.au> 169691Date: Wed Dec 20 13:05:53 2006 +1030 169692 169693 xfree86: Changing "IsMPDevice" to "SharedPointer" option. Devices will default 169694 to MP devices. 169695 169696 Xi: Adding QueryDevicePointer request/reply 169697 Adding WarpDevicePointer request/reply 169698 169699commit e437f357b6850a6c87ca6696870b3abd40e5b8ed 169700Author: Alan Coopersmith <alan.coopersmith@sun.com> 169701Date: Tue Dec 19 16:38:34 2006 -0800 169702 169703 xorg.conf man page should say "XFree86-DGA", not "Xorg-DGA" 169704 169705commit a81dbaae0597492f0245080c6f5af7158e05e2d0 169706Merge: fb6d676de d442998e3 169707Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169708Date: Tue Dec 19 16:11:19 2006 -0500 169709 169710 Merge branch 'master' into my-XACE-SELINUX 169711 169712commit d442998e39611be6805ea261f2286a2fd00f49b1 169713Author: Eric Anholt <eric@anholt.net> 169714Date: Fri Dec 8 13:35:36 2006 -0800 169715 169716 Only do the _POSIX_C_SOURCE hackery on linux where it's required. 169717 169718 On other OSes, the nasty hack was resulting in prototypes being hidden, so just 169719 don't do it. 169720 169721commit 228b9f77696190e47d4c96d6e0809bf645751557 169722Author: Eric Anholt <eric@anholt.net> 169723Date: Fri Dec 8 13:32:22 2006 -0800 169724 169725 Include sys/select.h to get FD_ISSET. 169726 169727commit 9563b2eea2f61246b6a9e14e00c701f693efa4e1 169728Author: Michel Dänzer <michel@tungstengraphics.com> 169729Date: Tue Dec 19 18:57:22 2006 +0100 169730 169731 EXA: Lots of damage tracking fixes. 169732 169733 Mostly due to exaDrawableDirty() now calculating the backing pixmap coordinates 169734 internally, for cases where they aren't trivially known. There's a new 169735 exaPixmapDirty() function for the other cases. 169736 169737commit 467c00cf450826e0bf06fe94470ec193af625d68 169738Author: George Sapountzis <gsap7@yahoo.gr> 169739Date: Tue Dec 19 18:45:25 2006 +0100 169740 169741 exaGlyphs: mark dirty for software path also. 169742 169743 This affects drivers with no UploadToScreen or UploadToScreen failures. 169744 169745commit 4334860e69e7d5b156082bd05c7a86708e5bad4c 169746Merge: 7e4717683 fdcc22ca1 169747Author: Michel Dänzer <michel@tungstengraphics.com> 169748Date: Tue Dec 19 16:29:26 2006 +0100 169749 169750 Merge branch 'master' into exa-damagetrack 169751 169752 Conflicts: 169753 169754 exa/exa_accel.c 169755 exa/exa_migration.c 169756 169757commit fdcc22ca1704d3519156c66804528c21b04fea65 169758Author: Michel Dänzer <michel@tungstengraphics.com> 169759Date: Tue Dec 19 16:11:17 2006 +0100 169760 169761 exaCopyNtoN: Fix usage of 'dx' and 'dy' instead of 'reverse' and 'upsidedown'. 169762 169763commit 67c2a86e59e915d9a5681e9d233478cfea3e51ed 169764Author: Michel Dänzer <michel@tungstengraphics.com> 169765Date: Tue Dec 19 15:44:18 2006 +0100 169766 169767 EXA: Compare backing pixmaps instead of drawables against driver limits. 169768 169769 The driver operations are always contained within the backing pixmaps, it 169770 doesn't matter if the drawables are bigger. 169771 169772commit 6b1e354dbb6e8ed9f2c654bbe7f8bbf241843d1c 169773Author: Eric Anholt <eric@anholt.net> 169774Date: Tue Dec 19 15:24:19 2006 +0100 169775 169776 EXA: Disable SHM pixmaps. 169777 169778 See https://bugs.freedesktop.org/show_bug.cgi?id=6772 . 169779 169780commit ca5ebe3971d8ebcfed00c5ebcd026cdd0ce0c6ba 169781Merge: 4d07b5037 1b029fd89 169782Author: Peter Hutterer <peter@cs.unisa.edu.au> 169783Date: Tue Dec 19 10:51:01 2006 +1030 169784 169785 Merge branch 'master' of git://anongit.freedesktop.org/git/xorg/xserver 169786 169787commit 2d0a63126b3d3a17005b7e122617ee0c5f44a55b 169788Merge: d8b5394ed 1b029fd89 169789Author: Peter Hutterer <peter@cs.unisa.edu.au> 169790Date: Tue Dec 19 10:31:40 2006 +1030 169791 169792 Merge branch 'master' of git://anongit.freedesktop.org/git/xorg/xserver 169793 169794commit 1b029fd896b76096905c516925ce0214fe14632c 169795Author: Alan Coopersmith <alan.coopersmith@sun.com> 169796Date: Mon Dec 18 14:51:04 2006 -0800 169797 169798 Xorg & Xserver man page updates for 1.2 release 169799 169800 - Added -extension & +extension to Xserver man page 169801 - Changed Xorg synopsis from X11R6 to X11R7 169802 - Clarified Xorg ancestry description 169803 - Moved Solaris to free/Open Source OS list 169804 - Removed references to MetroLink module loader & getconfig 169805 - Converted (1) to (__appmansuffix__) in a few more places 169806 - Replaced http://www.freedesktop.org/cvs/ with http://gitweb.freedesktop.org/ 169807 169808commit d8b5394eda9d92b7193004931caa6c24a337b2e6 169809Author: Peter Hutterer <peter@cs.unisa.edu.au> 169810Date: Mon Dec 18 17:35:34 2006 +1030 169811 169812 Cleaning up ifdef MPX from remaining files, removing it from configure.ac. 169813 Removing building mpx extension from Makefile.am 169814 169815commit 03c554283e49e449fd1282cf32564d15d9fd2c77 169816Author: Peter Hutterer <peter@cs.unisa.edu.au> 169817Date: Mon Dec 18 15:33:54 2006 +1030 169818 169819 mi: removing MPX ifdefs 169820 global sprite renaming in mipointer and misprite 169821 fixed: multiple remove cursor call in miSpriteInstallColormap 169822 fixed: wrong core pointer usage in miSpriteStoreColors 169823 169824 dix: bugfix in CheckCursorConfinement 169825 removing MPX ifdefs 169826 removing MPX event generation (using Xi solely now) 169827 bugfix GrabDevice: uninitialized field in grab struct caused segfault 169828 169829 xfree86: removing MPX fdefs 169830 169831 Xi: removing MPX protocol events 169832 169833commit a5fcf1e5e7452c9be82f63b6c2be2a25c4109523 169834Author: James Steven Supancic III <arrummzen@arrummzen.net> 169835Date: Sat Dec 16 12:02:38 2006 -0500 169836 169837 Fix RENDER issues (bug #7555) and implement RENDER add/remove screen 169838 support (bug #8485). 169839 169840commit c92f7bef54fa737766d65fe32c200f405f39228c 169841Author: Kevin E Martin <kem@freedesktop.org> 169842Date: Sat Dec 16 12:01:49 2006 -0500 169843 169844 For Xvfb, Xnest and Xprt, compile fbcmap.c with -DXFree86Server 169845 169846commit 012807356883128fde58bb2d4f91dd356d6418fc 169847Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169848Date: Fri Dec 15 18:27:16 2006 -0500 169849 169850 Add loud warnings to deprecated lookup functions. 169851 Hopefully this will alert external driver maintainers. 169852 169853commit ab1d5b0c31a1cfce95ab6b1d06f209f2c44e19ac 169854Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169855Date: Fri Dec 15 17:26:58 2006 -0500 169856 169857 Convert callers of LookupClient() to dixLookupClient(). 169858 169859commit f11dafaafc68f5cff1a1538d9566907786d8ab72 169860Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169861Date: Fri Dec 15 16:51:58 2006 -0500 169862 169863 Convert callers of SecurityLookupDrawable() to dixLookupDrawable(). 169864 169865commit 10aabb729d1586db344f9c1abdf1cf45e7ddaa7a 169866Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169867Date: Fri Dec 15 16:36:29 2006 -0500 169868 169869 Convert callers of LookupDrawable() to dixLookupDrawable(). 169870 169871commit 25d5e0a629f82d95bd71daf9a920a70e095b5188 169872Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169873Date: Fri Dec 15 15:50:46 2006 -0500 169874 169875 Convert callers of SecurityLookupWindow() to dixLookupWindow(). 169876 169877commit 04c721854fbf1bd6379c165a53fab2bdc09961c0 169878Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169879Date: Fri Dec 15 14:11:40 2006 -0500 169880 169881 Convert callers of LookupWindow() to dixLookupWindow(). 169882 169883commit 670bbb87310503fcc17203cecfa6f4f2f5db51d2 169884Author: Keith Packard <keithp@neko.keithp.com> 169885Date: Wed Dec 13 01:21:32 2006 -0800 169886 169887 RandR 1.2 rotation code must adjust width/height. 169888 169889 Mode lines reflect the monitor mode, not the projected size into the frame 169890 buffer. Flip width/height around so that the dimensions are oriented 169891 correctly. 169892 (cherry picked from 612a8e61803da8db0e305cbb093696b8e4284572 commit) 169893 169894commit 6c6901434ab469dd03b79fc98cd4a2b64d339305 169895Author: Keith Packard <keithp@mandolin.keithp.com> 169896Date: Wed Dec 13 00:58:54 2006 -0800 169897 169898 RandR 1.0 refresh rates unscrambled. SetScreenConfig uses RRCrtcSet right. 169899 169900 RandR 1.0 refresh rates were scrambled when working with a 1.2 driver that 169901 returned sizes in a mixed order. SetScreenConfig was treating RRCrtcSet as 169902 returning an RandR status instead of a Bool. 169903 (cherry picked from 6dc711833d7387372012fdff1ce1df3aefa2d234 commit) 169904 169905commit 628c7daeb12713d28e85e6b49fa037a7748dff83 169906Author: Keith Packard <keithp@guitar.keithp.com> 169907Date: Tue Dec 12 22:59:03 2006 -0800 169908 169909 RandR: config time updates when hardware config changes. 169910 169911 The config time in the RandR protocol reflects when the hardware state has 169912 changed. It was getting changed anytime the driver changed the usage 169913 of the hardware as well. 169914 (cherry picked from 98d18a6578130adb411ca4bcc776fcb7e07f189f commit) 169915 169916commit d742025f435f3eb7458cf8284d59300bc9a850aa 169917Author: Keith Packard <keithp@guitar.keithp.com> 169918Date: Tue Dec 12 20:16:49 2006 -0800 169919 169920 RandR mode list needs both output and crtc modes. 169921 169922 When an output no longer reports the current mode, it must still be included 169923 in the list advertised by the X server. Walk the crtcs to ensure it is 169924 included. 169925 (cherry picked from 78689d0d6630afcbcd3ce5394d12c2564a489f45 commit) 169926 169927commit 9e32bf98bc9ab17a137664d01f59a8f426f7ff3b 169928Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169929Date: Thu Dec 14 19:31:58 2006 -0500 169930 169931 Remove now-unused macro definitions from dix.h. 169932 169933commit 00f0705b3bb444ac934fc902cd23130f1777eab2 169934Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169935Date: Thu Dec 14 19:15:21 2006 -0500 169936 169937 Remove instances of macros SECURITY_VERIFY_GEOMETRABLE and SECURITY_VERIFY_GC. 169938 169939commit 5e334f06a1ef89891f9df2a371e4662340bec26b 169940Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169941Date: Thu Dec 14 18:27:09 2006 -0500 169942 169943 Remove instances of macros VERIFY_GEOMETRABLE and VERIFY_GC. 169944 169945commit 51b69ff499c05f59cb1e577c4e8abf6f7f283b3e 169946Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169947Date: Thu Dec 14 17:53:43 2006 -0500 169948 169949 Remove instances of macro SECURITY_VERIFY_DRAWABLE. 169950 169951commit 0cf75e74322e2b6a6efc7acf892e04365fde503b 169952Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169953Date: Thu Dec 14 17:27:12 2006 -0500 169954 169955 Remove instances of macros LOOKUP_DRAWABLE and VERIFY_DRAWABLE. 169956 169957commit ab1886df73b73360fa3bd7ce8e01affc074cbc8d 169958Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169959Date: Thu Dec 14 15:42:19 2006 -0500 169960 169961 Add new, combined dix lookup functions (tweak). 169962 169963commit 60cdc592fe042c03ceb5d4c3344acfbbf5d8ae28 169964Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169965Date: Thu Dec 14 14:46:03 2006 -0500 169966 169967 Add new, combined dix lookup functions. 169968 169969commit 6c46645cfc1afda8aeabfe0ed4d9342673b702f1 169970Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 169971Date: Thu Dec 14 14:45:42 2006 -0500 169972 169973 Naming change: Security*Access -> Dix*Access 169974 169975commit a96788058e48f3e97c66da6dafd58828f0c2ce53 169976Author: Peter Hutterer <peter@cs.unisa.edu.au> 169977Date: Thu Dec 14 10:50:18 2006 +1030 169978 169979 mi: removing MPX ifdefs 169980 global sprite renaming in mipointer and misprite 169981 fixed: multiple remove cursor call in miSpriteInstallColormap 169982 fixed: wrong core pointer usage in miSpriteStoreColors 169983 169984 dix: bugfix in CheckCursorConfinement 169985 removing MPX ifdefs 169986 removing MPX event generation (using Xi solely now) 169987 169988commit b88ad820fac81d0dfd557a384bf0406e8893e7af 169989Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 169990Date: Wed Dec 13 12:13:11 2006 +0000 169991 169992 Set Int10Current->Tag for the linux native int10 module 169993 Fixes bug #9296 169994 (cherry picked from 731952c561a3972d09d1315f4fd31466e459ccb9 commit) 169995 169996commit fb6d676de5aa606d943715437a12a68d9a41f386 169997Author: Eamon Walsh <ewalsh@moss-huskies.epoch.ncsc.mil> 169998Date: Tue Dec 12 16:17:51 2006 -0500 169999 170000 Add xserver object class to list of object classes. 170001 170002commit 7b90944258eba66b61328480759833ad7589bcca 170003Author: Eamon Walsh <ewalsh@moss-huskies.epoch.ncsc.mil> 170004Date: Tue Dec 12 15:59:38 2006 -0500 170005 170006 Change MLS levels in config file contexts to more sane defaults. 170007 170008commit 568c09481e5d62091d032837171a36f409f39379 170009Author: Eamon Walsh <ewalsh@moss-huskies.epoch.ncsc.mil> 170010Date: Tue Dec 12 15:59:08 2006 -0500 170011 170012 Split AssignClientState() into two routines, new routine is server-specific. 170013 170014commit 62e7722ebd8d341a23eb56cb330eeb00e6a975a6 170015Merge: e12480699 f3c60900e 170016Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170017Date: Tue Dec 12 13:54:33 2006 -0500 170018 170019 Merge branch 'XACE-SELINUX' into my-XACE-SELINUX 170020 170021commit e124806994675e16ca8e3937388f2cadeb529fc3 170022Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170023Date: Tue Dec 12 13:35:22 2006 -0500 170024 170025 Remove trailing whitespace (whitespace police). 170026 170027commit ca77c121075a9de1f47d42f6aaf91c20185231de 170028Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170029Date: Tue Dec 12 13:26:52 2006 -0500 170030 170031 Naming change: Security*Operation -> Xace*Operation 170032 170033commit a60da1db7cced28c07960a713eb18deb45beb432 170034Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170035Date: Wed Nov 29 22:19:57 2006 -0500 170036 170037 Add required root window context to config file. 170038 170039commit 5719afe6d3a246985709e6f045617c1e16a7da51 170040Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170041Date: Thu Oct 26 20:20:57 2006 -0400 170042 170043 Change symbol in libaudit library test. 170044 170045commit 354c80da66af141e8ba6d75fed75a0f482987956 170046Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170047Date: Thu Oct 5 16:07:26 2006 -0400 170048 170049 Improve error handling, messages during initialization. 170050 170051commit 3714d9149928754afcd6b2466a1371ca32e17985 170052Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170053Date: Wed Oct 4 16:23:35 2006 -0400 170054 170055 Experimental window property holding security context. 170056 170057commit 7f16c38ae2b47b195609d8fedefb7b28f612b2d4 170058Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170059Date: Fri Sep 8 15:31:18 2006 -0400 170060 170061 Add SELinux extension configure-time support. 170062 170063commit a7f4bbea87ada1d699bfd9e3b6a98f06191650f6 170064Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170065Date: Fri Sep 8 15:17:01 2006 -0400 170066 170067 Add SELinux extension to the module/extension loader. 170068 170069commit 28e80cd65b1207b123c02f895851bb6d207aa3c1 170070Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170071Date: Fri Sep 8 15:15:19 2006 -0400 170072 170073 Add SELinux extension configure-time support. 170074 170075commit 83aad2be8a80890f349c2f9caf84786333f7cc8c 170076Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170077Date: Fri Sep 8 15:11:04 2006 -0400 170078 170079 Add SELinux extension source files. 170080 170081commit 81281cb298a5825bc7a2e692375a86199293bbbe 170082Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 170083Date: Tue Dec 12 11:28:24 2006 +0000 170084 170085 Fix bad commit 170086 170087commit 792e0f71c6a435b2e28f8a4cdcc790f3b982e62c 170088Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 170089Date: Mon Dec 11 14:54:49 2006 +0000 170090 170091 Fix Xming fails to use xkb bug 170092 bug #5049 (Colin Harrison) 170093 170094commit 27d4b84f268ac21601f7f52a7e257f70753396b3 170095Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 170096Date: Mon Dec 11 14:50:08 2006 +0000 170097 170098 Fix Tooltip from minimized clients 170099 170100 Bug #3678 (Colin Harrison) 170101 170102commit ae3c24da34cd8eeb77a6389861856fd51e0841f5 170103Author: Peter Hutterer <peter@cs.unisa.edu.au> 170104Date: Mon Dec 11 18:09:59 2006 +1030 170105 170106 dix: Moving SpriteRec into DeviceIntRec 170107 removing global sprite structure 170108 beginning to remove MPX ifdefs 170109 170110 xnest: Fix to make xnest compile again 170111 170112commit fb8364bca30fe9268e807b0a9a3ebf875ee1fce2 170113Author: Adam Jackson <ajax@benzedrine.nwnk.net> 170114Date: Sun Dec 10 11:24:05 2006 -0500 170115 170116 Accept EDID > 1.3 but < 2.0 if we find it, assume it's compatible. 170117 170118commit e1f73d220873fa091695e46b7fcd008663a95a6f 170119Merge: 27d682fa0 d56249a15 170120Author: Ian Romanick <idr@us.ibm.com> 170121Date: Fri Dec 8 17:24:15 2006 -0800 170122 170123 Merge branch 'master' of ssh+git://idr@git.freedesktop.org/git/xorg/xserver into pci-rework 170124 170125 Conflicts: 170126 170127 hw/xfree86/common/xf86Configure.c 170128 hw/xfree86/common/xf86Helper.c 170129 hw/xfree86/common/xf86pciBus.c 170130 hw/xfree86/int10/helper_exec.c 170131 hw/xfree86/os-support/bus/Pci.c 170132 hw/xfree86/os-support/bus/linuxPci.c 170133 hw/xfree86/os-support/linux/lnx_pci.c 170134 hw/xfree86/scanpci/Makefile.am 170135 hw/xfree86/utils/pcitweak/Makefile.am 170136 hw/xfree86/utils/scanpci/Makefile.am 170137 170138commit d56249a15ead51ad4d2117d5538ada24af05b693 170139Merge: f1f8df188 ec84f72d0 170140Author: Jeremy C. Reed <reed@glacier.reedmedia.net> 170141Date: Fri Dec 8 15:52:37 2006 -0600 170142 170143 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 170144 170145commit f1f8df1889ab656bb57596e2f85408f15f42cd5d 170146Author: Jeremy C. Reed <reed@glacier.reedmedia.net> 170147Date: Fri Dec 8 15:51:44 2006 -0600 170148 170149 For MANDEFS, also replace __mandir__ for $(mandir) which includes 170150 the prefix. Noticed wrong path to man pages on both Ubuntu and on 170151 NetBSD with pkgsrc. 170152 170153commit ec84f72d077eaf2e7768a1f5398f65a5e1714d08 170154Author: Daniel Stone <daniel@fooishbar.org> 170155Date: Fri Dec 8 21:31:47 2006 +0200 170156 170157 XkbCopyKeymap: always initialise map and preserve 170158 170159 If we don't have map or preserve in the source map, make sure the 170160 destination is initialised to NULL, and freed if it's a valid type. 170161 170162commit e59aeac1ff13ea53f44dba1ac7800f37f4532ca1 170163Author: Daniel Stone <daniel@fooishbar.org> 170164Date: Wed Dec 6 23:44:06 2006 +0200 170165 170166 xfree86: remove stray debug line 170167 170168commit 729fca33a417ae3dfb180caf0ea8946ef9eee1df 170169Author: Michel Dänzer <michel@tungstengraphics.com> 170170Date: Thu Dec 7 12:09:18 2006 +0100 170171 170172 Revert "xfree86 DDX: Delete DDX screens in ddxGiveUp()." 170173 170174 This reverts commit a6381e69845f58d2b3282992b1f881015190f1bc. 170175 170176 See https://bugs.freedesktop.org/show_bug.cgi?id=9224 . 170177 170178commit 4ea6dfb984063117eef5c2f931205b1c3eb3108b 170179Author: Daniel Stone <daniel@fooishbar.org> 170180Date: Wed Dec 6 23:24:39 2006 +0200 170181 170182 whitespace police 170183 170184commit 4cba1a1ebfbdd7ab489b1b1ffb6656cbe88eb61e 170185Author: Daniel Stone <daniel@fooishbar.org> 170186Date: Wed Dec 6 23:21:38 2006 +0200 170187 170188 config: bus reconnect support 170189 170190 Add support for reconnecting to the bus when it restarts. 170191 170192commit 0c5dab5c8eaa174f28054b9d20244a709c015210 170193Author: Daniel Stone <daniel@fooishbar.org> 170194Date: Wed Dec 6 23:21:15 2006 +0200 170195 170196 config: move config.h to hotplug.h 170197 170198 Also, move configInitialise to after OsInit, since the next commit will 170199 make it use a timer. 170200 170201commit 72e7f2ac6cf0db474d0defa7918f2a3ba76c0e46 170202Author: Daniel Stone <daniel@fooishbar.org> 170203Date: Wed Dec 6 23:18:52 2006 +0200 170204 170205 GetPointerEvents: always send valuator events for MotionNotify 170206 170207 Always chase a DeviceMotionNotify event with a DeviceValuator, which is 170208 not required in the spec, but will silently break the lib if you don't 170209 include. 170210 170211commit c458a70d650bd62b8f4706f022d1f3f347636db1 170212Author: Daniel Stone <daniel@fooishbar.org> 170213Date: Wed Dec 6 20:30:44 2006 +0200 170214 170215 GetPointerEvents: fix typo 170216 170217 Fix typo that resulted in inverted axes when using an absolute positioning 170218 device that didn't report y, and thus relied on the previous value. 170219 170220commit edabf45425f9ed79547f918cc0dfff4c268de386 170221Author: Daniel Stone <daniel@fooishbar.org> 170222Date: Sat Dec 2 16:37:19 2006 +0200 170223 170224 configure.ac: add CONFIG_LIB to Xvfb 170225 170226commit 8724af248cd6c93182fecd060fed09a556361080 170227Author: Daniel Stone <daniel@fooishbar.org> 170228Date: Sat Dec 2 16:20:34 2006 +0200 170229 170230 kdrive/mga: fix compiler warning 170231 170232 Change a case that only made one test for an if, preventing the compiler 170233 warning about all other PICT_* types being unsupported. 170234 170235commit 2f0a800ffdc881cdb3adf84f1ed97bbb63cba34c 170236Author: Daniel Stone <daniel@fooishbar.org> 170237Date: Mon Nov 27 22:22:53 2006 +0200 170238 170239 config: move to block/wakeup handler 170240 170241commit 99378b58dbc63160382ad9c41f9cb0dd2a24e9d1 170242Author: Daniel Stone <daniel@fooishbar.org> 170243Date: Mon Nov 27 22:22:33 2006 +0200 170244 170245 kdrive/tslib: remove vendor-specific hacks 170246 170247 Parts of the KDrive merge accidentally contained a bunch of 170248 vendor-specific hacks; sorry. 170249 170250commit 8884a73a3f4efa8276c5e38b9573201574c4f1f6 170251Author: Daniel Stone <daniel@fooishbar.org> 170252Date: Mon Nov 27 22:39:56 2006 +0200 170253 170254 xfree86/input: re-add support for disabling drag events 170255 170256commit f2903c12bb4bb0b7c94b96c55af8fa55507f9d7d 170257Author: Daniel Stone <daniel@fooishbar.org> 170258Date: Tue Nov 7 11:13:32 2006 +0200 170259 170260 SyntheticMotion: don't dereference sprite.screen when not using Xinerama 170261 170262 (cherry picked from aa052e43c6c293e14f78837e00c6b7581f9713bb commit) 170263 170264commit d17ec01e8395a8f14b75a10c8bf082b3f5a4fb36 170265Author: Daniel Stone <daniel@fooishbar.org> 170266Date: Fri Dec 1 00:41:41 2006 +0200 170267 170268 remove CID support (bug #5553) 170269 170270 Remove CID from all our fontpaths. 170271 (cherry picked from 69820a10e33e4582c192360996e866007114639d commit) 170272 170273commit abe5e079af715713097ab0daad29a3e9f523c398 170274Author: Alan Coopersmith <alan.coopersmith@sun.com> 170275Date: Wed Dec 6 07:58:03 2006 -0800 170276 170277 Update pci.ids to 2006-12-06 from pciids.sf.net 170278 170279commit 724f9cb578086e8483a2d0636dd6eb05d664d31c 170280Author: Aaron Plattner <aplattner@nvidia.com> 170281Date: Tue Dec 5 13:44:05 2006 -0800 170282 170283 Bug #9219: Use pWin->viewable instead of pWin->realized to catch InputOnly windows too. 170284 170285commit 3690de9b1b0902d395bc7d071fc05ebc8f75be2b 170286Author: Aaron Plattner <aplattner@nvidia.com> 170287Date: Tue Dec 5 12:42:12 2006 -0800 170288 170289 Bug #9219: Return BadMatch when trying to name the backing pixmap of an unrealized window. 170290 170291 Before this change, ProcCompositeNameWindowPixmap would name the screen pixmap 170292 if !pWin->realized. 170293 170294commit eb1d9f51af19dab20a95c2830ca1c508d8ee66bb 170295Author: Peter Hutterer <peter@cs.unisa.edu.au> 170296Date: Tue Dec 5 18:50:19 2006 +1030 170297 170298 Ironing some glitches caused by the merge 170299 170300commit 3c8f166022a2d19af14eb670fa382503ba1451cd 170301Merge: 2a511492f f9f7d7f3b 170302Author: Peter Hutterer <peter@cs.unisa.edu.au> 170303Date: Tue Dec 5 18:28:33 2006 +1030 170304 170305 Merge branch 'master' of git://anongit.freedesktop.org/git/xorg/xserver into mpx 170306 170307 Conflicts: 170308 170309 dix/events.c 170310 170311commit 2a511492f9a235d3f5390ab87b6c643cbea92d88 170312Author: Peter Hutterer <peter@cs.unisa.edu.au> 170313Date: Tue Dec 5 15:45:18 2006 +1030 170314 170315 Xi: adding MPX checks to ProcessOtherEvents 170316 170317 mpx: adding QueryPointer request and reply 170318 adjusting names of requests to Xlib compatible naming (xMPX*** instead of 170319 mpx***) 170320 170321commit f9f7d7f3be53c808abb5eaceb7a1abc55744a210 170322Author: Alan Coopersmith <alan.coopersmith@sun.com> 170323Date: Mon Dec 4 13:36:30 2006 -0800 170324 170325 Check for __sparc as well as __sparc__ for compatibility with Sun cc 170326 170327 (gcc defines __sparc__, Sun cc defines __sparc) 170328 170329commit ac90ce58ba1da3ed605adf75f4d54c34b578c402 170330Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170331Date: Fri Dec 1 21:12:21 2006 -0500 170332 170333 Naming change: Security*Operation -> Xace*Operation 170334 170335commit f44f14fe564d834568a0afefba944223a73ea0f5 170336Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170337Date: Fri Dec 1 20:48:15 2006 -0500 170338 170339 Define calls away when not building XACE, allowing ifdef's to be removed. 170340 170341commit a5d6499d666fea4a9988118ddd3a5e4c9cfcc32c 170342Author: Alan Coopersmith <alan.coopersmith@sun.com> 170343Date: Fri Dec 1 16:42:37 2006 -0800 170344 170345 Fix syntax error in configure check for SYSV_IPC that broke with Sun cc 170346 170347commit 89b2aa9be81613cb1a06bd535bf50ecf2a00208d 170348Author: Keith Packard <keithp@guitar.keithp.com> 170349Date: Mon Nov 27 21:40:24 2006 -0800 170350 170351 Destroying RandR crtc or output overwrites memory. 170352 170353 RRCrtcDestroyResource and RROutputDestroyResource had matching 170354 bugs that would overwrite memory past the end of the storage 170355 of the crtc or output arrays. Oops. 170356 (cherry picked from 4202b23ed86405a4cebfdcf239df1b023c1d10ca commit) 170357 170358commit 23ba72323af785516db6cbcf6c1b2fa907a8232f 170359Author: Keith Packard <keithp@ukulele.keithp.com> 170360Date: Thu Nov 30 23:16:42 2006 -0800 170361 170362 RandR ListOutputProperties has nAtoms element, not nProperties 170363 170364 Earlier RandR 1.2 encoding revisions used 8-bit nProperties field. 170365 Final RandR 1.2 spec uses 16-bit nAtoms field instead. 170366 (cherry picked from 66b6358a393972946f16394918db2401c51dc5ed commit) 170367 170368commit 416f1bb99a6d4557f2863ae941868e47e11bbb3b 170369Author: Peter Hutterer <peter@cs.unisa.edu.au> 170370Date: Fri Dec 1 15:56:52 2006 +1030 170371 170372 mpx: SelectEvents and GetEventBase requests. 170373 Some renaming and cleaning up in extinit.c 170374 MPXLastEvent added 170375 170376 Xi: ShouldFreeInputMask() from XI is not static any more, used in mpx 170377 170378 dix: GetPointerEvents() allocates MPX event for MPX devices. 170379 DeliverDeviceEvents() caters for MPX devices. 170380 170381commit ccd804c6c01cdfffe938fa5336be9b5668a6f0c0 170382Merge: 9423ac134 82912ad77 170383Author: Drew Parsons <drew@pug.localdomain> 170384Date: Fri Dec 1 15:21:57 2006 +1100 170385 170386 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 170387 170388commit 9423ac134a2a9800787c4978d384827cc4e61fc8 170389Author: Drew Parsons <drew@pug.localdomain> 170390Date: Fri Dec 1 15:21:05 2006 +1100 170391 170392 GNU/FreeBSD support in GLX: include byteswap.h for FreeBSD systems 170393 defining __GLIBC__, as done already for Linux and Hurd (cf. bug #5613). 170394 Also includes some more __GNU__ checks on top of those made in 170395 commit ade4bf09076d13dbf3549c0a2d987a0afe76d5c1. 170396 Thanks Petr Salinger (Debian bug #400869). 170397 170398commit 82912ad7709e8cf4a5f8a9fa6b47f789842a3fe9 170399Author: Alan Coopersmith <alan.coopersmith@sun.com> 170400Date: Thu Nov 30 20:13:52 2006 -0800 170401 170402 Make solaris version of xf86OSRingBell return void like other OS'es 170403 170404commit 8956f63a941bf0a5f157d47b33e6221601a75040 170405Author: Alan Coopersmith <alan.coopersmith@sun.com> 170406Date: Thu Nov 30 19:53:29 2006 -0800 170407 170408 Tell automake to STFU about the *.O files 170409 170410 automake will not stop whining about the *.O files not being in normal library 170411 name format, so just tell automake they are PROGRAMS so it builds them without 170412 bitching. 170413 170414commit f3c60900e575e65254cd2576cc6c90b97c8f63ae 170415Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170416Date: Wed Nov 29 22:19:57 2006 -0500 170417 170418 Add required root window context to config file. 170419 170420commit 143558500a89a5c887c75c03798bae68c953cac9 170421Merge: e3d3d29db ee9bdd3f4 170422Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 170423Date: Wed Nov 29 20:41:41 2006 -0500 170424 170425 Merge branch 'master' into my-XACE-SELINUX 170426 170427commit ee9bdd3f4a14a42fb0747acc420966f0f669129b 170428Author: Aaron Plattner <aplattner@nvidia.com> 170429Date: Wed Nov 29 12:01:43 2006 -0800 170430 170431 Add DIX_CFLAGS to util builds. 170432 170433 Fixes a build breakage when $(top_srcdir) != $(top_builddir) because 170434 -I$(top_srcdir)/include is missing for the cvt, ioport, pcitweak, and scanpci 170435 builds. 170436 170437 Signed-off-by: Adam Jackson <ajax@nwnk.net> 170438 170439commit a6381e69845f58d2b3282992b1f881015190f1bc 170440Author: Michel Dänzer <michel@tungstengraphics.com> 170441Date: Wed Nov 29 19:25:09 2006 +0100 170442 170443 xfree86 DDX: Delete DDX screens in ddxGiveUp(). 170444 170445 This allows video drivers to clean up in the FreeScreen hook things they set up 170446 in the PreInit hook. 170447 170448commit afd6af65510b0147f16f08314045b3506bfa9d89 170449Author: Peter Hutterer <peter@cs.unisa.edu.au> 170450Date: Wed Nov 29 17:55:15 2006 +1030 170451 170452 mpx: Infrastructure for MPX extension, GetExtensionVersion request works. 170453 170454commit b0c8558b9d9a9984c0067960392e28f5a7622b29 170455Author: Keith Packard <keithp@neko.keithp.com> 170456Date: Sun Nov 26 19:31:48 2006 -0800 170457 170458 Ensure RandR resource types are registered before resources are created. 170459 170460 Now that resources can be created during server initialization, make sure 170461 the crtc, output and mode resource types are created before attempting to 170462 create associated resources. 170463 (cherry picked from commit ec83d674167e7045d5317b179c9998e3172a26dc) 170464 170465commit 6245e9dd4719c5dc15ff45d49cf626123794038b 170466Author: Keith Packard <keithp@neko.keithp.com> 170467Date: Tue Nov 21 16:52:28 2006 -0800 170468 170469 Allocate correct size for RRPropertyRec (oops). 170470 170471 Neglected to change the allocation size from sizeof (PropertyRec) to 170472 sizeof (RRPropertyRec). Lots of fun crashes this way. 170473 (cherry picked from commit 0626eb8e5c9fa05de6bdc9aa0c654f5148bf7cff) 170474 170475commit 24abce8032940e96bb2ccf9e463a7fff6f36283a 170476Author: Keith Packard <keithp@neko.keithp.com> 170477Date: Tue Nov 21 01:15:26 2006 -0800 170478 170479 Change RandR property datatype to include pending/valid values. 170480 170481 This patch tracks the protocol changes which introduce more complex 170482 semantics for RandR output properties including pending and valid value 170483 information. 170484 (cherry picked from commit af55c65bea40669fdc038aa34c6a1ec9ecb33e87) 170485 170486commit f62ac3ec39c6593df476985c630e499864c19c72 170487Author: Eric Anholt <eric@anholt.net> 170488Date: Tue Nov 28 10:31:40 2006 -0800 170489 170490 Separate DDC mode list creation from MonPtr creation. 170491 170492 This will be used by the intel driver, and likely other RandR 1.2 drivers. 170493 170494commit fbd09443385c533416fa530399d54f130afaf985 170495Author: Eric Anholt <eric@anholt.net> 170496Date: Tue Nov 28 10:15:51 2006 -0800 170497 170498 Replace bad mode name-setting code with xf86SetModeDefaultName(). 170499 170500commit 4ad0bde661be2af4a17771d66066d49736e85cbe 170501Author: Eric Anholt <eric@anholt.net> 170502Date: Tue Nov 28 10:12:02 2006 -0800 170503 170504 Clean up a bunch of long lines and trailing whitespace. 170505 170506commit 05778432dc6e688bc0beff0c20ffd7e27b74888e 170507Author: Eric Anholt <eric@anholt.net> 170508Date: Tue Nov 28 10:07:57 2006 -0800 170509 170510 Move code to get a mode list from EDID data from ddcProperty.c to edid_modes.c. 170511 170512commit 38ecc66cd9c61346a46697bbf1d8319f4f6f9800 170513Author: Eric Anholt <eric@anholt.net> 170514Date: Tue Nov 28 10:06:15 2006 -0800 170515 170516 Typo that was missed in testing. 170517 170518commit 834e4b079866594b50be64ae79f3cb2a5baa2070 170519Author: Matthias Hopf <mhopf@suse.de> 170520Date: Tue Nov 28 18:57:13 2006 +0100 170521 170522 Fix potential NULL pointer access in timer code. 170523 170524 https://bugzilla.novell.com/show_bug.cgi?id=223718 170525 170526commit bdf718907b3202815a04ec28a868689708f43da9 170527Author: Peter Hutterer <peter@cs.unisa.edu.au> 170528Date: Tue Nov 28 18:06:34 2006 +1030 170529 170530 dix: missing parameter to SetCursorPosition in CheckMotion ifndef MPX 170531 170532 mi: missing include in misprite.c ifndef MPX 170533 170534commit e88bc0e55a6937ff0b47fe38f60bfa8991196578 170535Author: Peter Hutterer <peter@cs.unisa.edu.au> 170536Date: Tue Nov 28 14:34:36 2006 +1030 170537 170538 mi: Bugfix: FreeCursor ignores negative refcounts. 170539 170540 mieq: Bugfix: Motion notifies did overwrite DeviceMotionNotifies 170541 170542commit 5dbcd34a0a6c0d10dbfea8fdc9d7dfe7a0261b19 170543Author: Eric Anholt <eric@anholt.net> 170544Date: Mon Nov 27 16:26:14 2006 -0800 170545 170546 Register dependency on new RandR protocol. 170547 170548commit 16f8f10dc2106bc6253b2d89a1f8efee8d80e2ba 170549Author: Eric Anholt <eric@anholt.net> 170550Date: Mon Nov 27 16:21:31 2006 -0800 170551 170552 Move mode handling helpers from ddcProperty.c to xf86Mode.c. 170553 170554commit b4b0d901d98371a8aa7b17d195e18e83e2a6a618 170555Merge: 64de3baf8 d6cd0313c 170556Author: Eric Anholt <eric@anholt.net> 170557Date: Mon Nov 27 15:43:15 2006 -0800 170558 170559 Merge branch 'randr-1.2' 170560 170561 Conflicts: 170562 170563 dix/events.c 170564 dix/getevents.c 170565 hw/xfree86/common/xf86Mode.c 170566 hw/xfree86/dri/Makefile.am 170567 hw/xfree86/os-support/drm/xf86drm.c 170568 hw/xfree86/os-support/xf86drm.h 170569 170570commit d6cd0313c7f23f32c9c7dda00ff739e772bf7db3 170571Author: Eric Anholt <eric@anholt.net> 170572Date: Mon Nov 27 14:46:50 2006 -0800 170573 170574 Add some mode helper functions from the intel driver. 170575 170576 This also removes static from some other functions that had been copied out 170577 to at least the intel driver, but perhaps others that were doing mode list 170578 handling. 170579 170580commit 85ea408e38f452bcf9109f14fd1016d159692a99 170581Author: Peter Hutterer <peter@cs.unisa.edu.au> 170582Date: Mon Nov 27 18:07:16 2006 +1030 170583 170584 mi: malloc checks for misprite and midispcur, code cleanup 170585 removed IsMPDev() 170586 170587commit 023da4e804ad260eaf0c50d2ec7213dd514e2113 170588Author: Peter Hutterer <peter@cs.unisa.edu.au> 170589Date: Mon Nov 27 15:54:33 2006 +1030 170590 170591 mi: mipointer.c cleanup, reducing code duplication, using MPHasCursor() 170592 instead of IsMPDev() 170593 170594commit a2340a73687dfd93affb6de28a33d73fd808c96b 170595Author: Peter Hutterer <peter@cs.unisa.edu.au> 170596Date: Mon Nov 27 14:47:23 2006 +1030 170597 170598 mi: free allocated memory in miDCCloseScreen and in miDCInitialize in case 170599 of errors. 170600 Some copyright notices. 170601 misprite.c passes DeviceIntPtr around and down to DC (previously 170602 miCursorInfoPtr and the plain device id). 170603 Large cleanup in misprite.c to avoid code duplication. 170604 170605commit 1c7568b8a1417257fa67c7fca69aa253099b9461 170606Author: Peter Hutterer <peter@cs.unisa.edu.au> 170607Date: Thu Nov 23 17:15:14 2006 +1030 170608 170609 mi: closing memory leak, miPointer is freed in miPointerCloseScreen 170610 bugfix: uninitialized pPointer in miPointerGetPosition ifndef MPX 170611 adding DeviceIntPtr parameter to ScreenRec's cursor functions. 170612 cleanup of miPointer code to use same scheme in each function 170613 170614 dix: MPHasCursor() function determines checking whether to invoke 170615 cursor rendering. 170616 170617 animcur: adding DeviceIntPtr parameter to cursor functions but animcur relies 170618 on the core pointer right now. 170619 170620 xfixes: adding DeviceIntPtr parameter to cursor functions but xfixes relies on 170621 the core pointer right now. 170622 170623 rac: adding DeviceIntPtr parameter to cursor functions but RAC relies on 170624 the core pointer right now. 170625 170626 ramdac: adding DeviceIntPtr parameter to cursor functions but ramdac relies on 170627 the core pointer right now. 170628 170629commit 64de3baf85f6df274f71f736016f0848567cd9f6 170630Author: Aaron Plattner <aplattner@nvidia.com> 170631Date: Wed Nov 22 14:46:51 2006 -0800 170632 170633 Add a -showDefaultLibPath option. 170634 170635 A corollary to the previous change, this option prints $libdir. 170636 170637commit 0a2a6e4070718b90af7ca0e047f028e0cabdfb9d 170638Author: Aaron Plattner <aplattner@nvidia.com> 170639Date: Fri Nov 17 18:27:23 2006 -0800 170640 170641 Add a -showDefaultModulePath option. 170642 170643 As discussed on the mailing list, people would rather have an X command-line 170644 option to print the module path so installers can know where to put modules, 170645 rather than the installers using `pkg-config --variable=moduledir xorg-server`, 170646 since some distros choose not to install xorg-server.pc. 170647 170648commit 61832cb94c8a4d62cddb92188caeed86519e9d62 170649Author: Brian <brian@i915.localnet.net> 170650Date: Wed Nov 22 09:16:43 2006 -0700 170651 170652 Regenerated from Mesa w/ latest gl_API.xml file 170653 170654commit 61863f09d22935406371e92bb75173d55ff9b29f 170655Author: Brian <brian@i915.localnet.net> 170656Date: Wed Nov 22 09:16:17 2006 -0700 170657 170658 Regenerated from Mesa, fixes glMap* protocol problem (bug 8899) 170659 170660commit f52d53e060d0468b658a565688d1f92c156405c7 170661Author: Peter Hutterer <peter@cs.unisa.edu.au> 170662Date: Wed Nov 22 17:58:57 2006 +1030 170663 170664 dix: DeliverEventsToWindow adjusted for MPX 170665 170666 mi: sprite debugging disabled 170667 170668commit 8c16f21d5df08b028e93440d632939a5fdbf5876 170669Author: Peter Hutterer <peter@cs.unisa.edu.au> 170670Date: Wed Nov 22 17:06:47 2006 +1030 170671 170672 dix GetSpriteWindow() adjusted for MPX 170673 added device dependency for MPX to DoEnterLeaveEvents, 170674 EnterLeaveEvents, EnterNotifies, LeaveNotifies and 170675 FixUpEventFromWindow. 170676 170677commit 70383105de279df553874efa56b37a7b3e426ea1 170678Author: Peter Hutterer <peter@cs.unisa.edu.au> 170679Date: Wed Nov 22 15:27:16 2006 +1030 170680 170681 mi: added DevToSprite() conversion function 170682 fixed miSpriteRealize to use passed pDev argument instead of looping. 170683 170684 dix: changed sprite dependency and added MPX functionality to parts of 170685 events.c (XineramaChangeToCursor, XineramaCheckPhysLimits, 170686 XineramaConstrainCursor) 170687 170688 Xi: fix: accessing other->buttons for keyboard segfaulted the server 170689 170690commit 33ff4cec5ff7533ec725f71d357c096dfb0acb79 170691Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 170692Date: Tue Nov 21 21:23:17 2006 +0000 170693 170694 Issue CloseDownDevices() in os/log.c and remove from dix/main.c. 170695 170696 This ensures that all calls to FatalError() will shutdown the input devices. 170697 170698commit efd4f3c6ffec804c68ba5df17cc117da264fb7c4 170699Author: Peter Hutterer <peter@cs.unisa.edu.au> 170700Date: Tue Nov 21 20:31:30 2006 +1030 170701 170702 dix: WindowsRestructured() calls CheckMotion() on all devices 170703 170704 mi: core pointer was checked twice in miSpriteReportDamage, 170705 miSpriteInstallColormap, miSpriteStoreColors, miSpriteSaveDoomedAreas 170706 and miSpriteRealiseCursor 170707 using damage bug (see comment in file) to avoid artefacts 170708 170709commit 9db851c22d4befb95bfb074b96620261d8e32ac9 170710Author: Peter Hutterer <peter@cs.unisa.edu.au> 170711Date: Tue Nov 21 18:15:04 2006 +1030 170712 170713 dix: moved sprite from static to be pointer of type SpritePtr 170714 added SpriteRecs for MPX devices 170715 changed sprite dependency and added MPX functionality to parts 170716 of events.c (ConfineToShape, PostNewCursor, XineramaCheckMotion, 170717 CheckMotion, XineramaChangeToCursor, ChangeToCursor, CheckPhysLimits, 170718 PointerConfinedToScreen) 170719 added DeviceIntRec param to GetSpritePosition(). This required some 170720 minor changes in ddx, xtest, xkb and xfixes. 170721 170722 mi: changed miPointer to pointer instead of static struct. 170723 170724commit fd8b9a6eaee28a453a00af14195b81ef1890a577 170725Author: Peter Hutterer <peter@cs.unisa.edu.au> 170726Date: Tue Nov 21 15:24:43 2006 +1030 170727 170728 TAG: MPX_BEFORE_DIX_API_BREAK 170729 170730commit 447efdc01d4f8922205ba26e52c9e5ff9f5cb90e 170731Author: Peter Hutterer <peter@cs.unisa.edu.au> 170732Date: Tue Nov 21 14:46:00 2006 +1030 170733 170734 mi: constrain MPX cursors to screen 170735 170736commit 1107d8bea22eea26f7b666f1c00a53360b7d237e 170737Author: Peter Hutterer <peter@cs.unisa.edu.au> 170738Date: Tue Nov 21 14:00:46 2006 +1030 170739 170740 mi: added device ID to miSpriteCursorFuncRec functions 170741 added MPX code to midispcur.c 170742 170743 BUG: Core pointer changes whenever MPX cursor changes shape. 170744 170745commit 3bad452d12d620df91b1f37d06c650ded96bfafe 170746Author: Peter Hutterer <peter@cs.unisa.edu.au> 170747Date: Tue Nov 21 11:26:21 2006 +1030 170748 170749 mi: added MPX to miSpriteReportDamage 170750 added id field to miCursorInfoPtr, required to pass through to miDC 170751 core pointer uses mpCursors array as well. 170752 added miDCBufferRec for future use with MPX 170753 170754 TAG: MPX_BEFORE_MIDC_API_BREAK 170755 170756commit b1bbdf464d2f8799fa3429c1c27ca998c5f3e94a 170757Author: Peter Hutterer <peter@cs.unisa.edu.au> 170758Date: Mon Nov 20 17:58:30 2006 +1030 170759 170760 mi: added MPX to miSprite functions. 170761 170762 NOTE: This version will render multiple cursors but with an artefact each time 170763 when devices alternate. 170764 170765commit 7b4b619d4c268804199f9f4775eb56604f0b80e8 170766Author: Peter Hutterer <peter@cs.unisa.edu.au> 170767Date: Mon Nov 20 13:10:07 2006 +1030 170768 170769 mi: moved core pointer information from miSpriteScreenRec into a 170770 miCursorInfoRec. 170771 bugfix in mipointer, semicolon after preprocessor statement 170772 170773commit ca2874b273232d9f51881b1cd754ed6847bfaf47 170774Merge: c9a5f9d39 e2f6dacc7 170775Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 170776Date: Sat Nov 18 19:56:32 2006 +0000 170777 170778 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 170779 170780commit c9a5f9d3911c2e6a8f81b0721c9eb914fb7719b5 170781Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 170782Date: Sat Nov 18 19:55:42 2006 +0000 170783 170784 Shutdown input devices if FatalError occurs during startup. 170785 Fixes Xdmx problems when the input device has been initialized 170786 and the keyboard map has been destroyed. 170787 170788commit e2f6dacc736527790ed8e304698678afc17f71c6 170789Author: Dave Airlie <airlied@linux.ie> 170790Date: Sat Nov 18 11:26:55 2006 +1100 170791 170792 dri: setup libdrm hooks as early as possible. 170793 170794 This is due to radeon doing drm stuff before DRI extension gets initialised 170795 170796commit 90bea69dfdd63f58ee326887359ad1b35a31cb5f 170797Author: Peter Hutterer <peter@cs.unisa.edu.au> 170798Date: Fri Nov 17 18:03:43 2006 +1030 170799 170800 mi: 170801 added miCursorInfoRec to contain info of the MPX cursors. 170802 calling miUpdatePointerSprite() from event queue for MPX devices. 170803 adding device-specific processing to miPointer*() functions. 170804 170805 dix: Call to SetCursorPosition in CheckMotion() temporarily disabled. 170806 170807 xfree86/common: call to miPointerUpdateSprite() disabled, is done from the EQ 170808 170809 NOTE: This build will not display cursor images. 170810 BUG: The second mouse does to take correct x coordinates. 170811 170812commit 94e106a898c44daeb228bcb031b85f996ddc96b0 170813Author: Peter Hutterer <peter@cs.unisa.edu.au> 170814Date: Fri Nov 17 15:10:37 2006 +1030 170815 170816 dix: moved isMPdev field to end of _DeviceIntRec structure 170817 170818 mi: 170819 added miMPPointers array to mipointer.c 170820 added DeviceIntPtr to all miPointerSpriteFuncs. Coming from miPointer 170821 we use inputInfo.pointer as standard value. ABI BREAK! 170822 170823 ramdac: 170824 forcing failed HW Cursor initialisation. MPX needs software rendering. 170825 changes to use new miPointerSpriteFunc (this required externing 170826 inputInfo, should probably be fixed at a later point). 170827 170828 RAC: changes to use new miPointerSpriteFuncs. 170829 170830commit ef47d9c3ba63e9a6243fe5c81ccc60c8246352b4 170831Author: Keith Packard <keithp@neko.keithp.com> 170832Date: Thu Nov 16 13:50:48 2006 -0800 170833 170834 Reduce calls to RRGetInfo. 170835 170836 RRGetInfo can be expensive. Don't invoke it when quering Xinerama 170837 information or setting a new CRTC configuration. 170838 (cherry picked from commit b5aa9eb8e6eda36856a075f4b008c33f6c706bad) 170839 170840commit 07b26e690cd9a4fc626132feed0702515cbe5a88 170841Author: Keith Packard <keithp@neko.keithp.com> 170842Date: Thu Nov 16 09:48:33 2006 -0800 170843 170844 Remove RandR output options. 170845 170846 RandR output options are now expected to be handled by properties instead. 170847 (cherry picked from commit 8b2a7e94a1dc2776ab2cfaaebb309be02502602a) 170848 170849commit f17e3c34dfd1f1418440bdebf45764e4dbf550f0 170850Author: George Sapountzis <gsap7@yahoo.gr> 170851Date: Thu Nov 16 02:18:03 2006 +0200 170852 170853 Fix GL context destruction with AIGLX. 170854 170855 The logic for freeing GL contexts introduced by "Fix AIGLX VT switching." is 170856 inverted. As it is now, GL context destruction is deferred for glxDRIEnterVT(). 170857 170858commit ae3c9ad4abe66784d7ee474455003d2745699286 170859Author: Bjorn Helgaas <bjorn.helgaas@hp.com> 170860Date: Thu Nov 16 17:29:06 2006 +0100 170861 170862 Bug 9041: Check the return code in xf86MapDomainMemory(). 170863 170864commit 5388423eb05daefcc71067b6ab96b6e57c44ef5c 170865Author: Peter Hutterer <peter@cs.unisa.edu.au> 170866Date: Thu Nov 16 20:43:27 2006 +1030 170867 170868 mieq: EQ processing handles MP devices 170869 170870 global: MPX define added to xorg-server.h.in 170871 170872 xfree86/common: small fix to avoid byte overflow 170873 170874commit c957a16180810fbd58526e1a670d82384fb4a61d 170875Author: Peter Hutterer <peter@cs.unisa.edu.au> 170876Date: Thu Nov 16 15:11:14 2006 +1030 170877 170878 dix: added isMPdev field to _DeviceIntRec, is used in GetPointerEvents() 170879 170880 mieq: avoid merging events from different devices in mieqEnqueue() 170881 170882 xfree86/common 170883 isMPdev field used from xf86ActivateDevice(), xf86PostMotionEvent() 170884 and xf86PostButtonEvent() 170885 170886commit 7ce2dc57844872cc340b60387d4d1cb3d982633e 170887Author: Peter Hutterer <peter@cs.unisa.edu.au> 170888Date: Thu Nov 16 12:51:08 2006 +1030 170889 170890 dix: 170891 merge with code cleanup from master 170892 GetPointerEvents treats events in the same way as XINPUT devices when flag 170893 has POINTER_MULTIPOINTER set. 170894 170895 xfree86/common: 170896 added XI86_MP_DEVICE flag and parsing in xf86ProcessCommonOptions 170897 added POINTER_MULTIPOINTER define. Is used in xf86PostMotionEvent and 170898 xf86PostButtonEvent for the flags that are passed into GetPointerEvents() 170899 170900 global: 170901 added flags to configure.ac to enable/disable MPX define 170902 added flags to dix-config.h.in to define MPX 170903 170904commit 4d07b503727ba86315b90d6f45b06ed6185724b2 170905Author: Peter Hutterer <peter@cs.unisa.edu.au> 170906Date: Thu Nov 16 10:55:15 2006 +1030 170907 170908 dix: cleaned up GetPointerEvents() to avoid code duplication. 170909 170910commit 6ff7f2ad6a5e2e769244590578e6809974b5235d 170911Author: Eric Anholt <eric@anholt.net> 170912Date: Tue Nov 7 13:13:53 2006 -0800 170913 170914 Fix build on FreeBSD after input-hotplug. 170915 (cherry picked from commit 4e6e4baead6c565363abbcd9e06cc685be121596) 170916 170917commit bffb1f9b67c8d3d264c0c3fd703cf7e57dcdf7b7 170918Author: Peter Hutterer <peter@cs.unisa.edu.au> 170919Date: Wed Nov 15 17:53:13 2006 +1030 170920 170921 removed unused variables drag and local from xf86PostMotionEvent 170922 170923commit f737cd925efe452787443b039ae65b8d85081bb3 170924Author: Peter Hutterer <peter@cs.unisa.edu.au> 170925Date: Wed Nov 15 17:50:02 2006 +1030 170926 170927 o fix minor error in comment for GetPointerEvents() 170928 170929commit 2eab230d9bd3f73ffe1b5a42111f89e85904ee11 170930Author: Jeremy C. Reed <reed@glacier.reedmedia.net> 170931Date: Tue Nov 14 16:37:18 2006 -0600 170932 170933 For NetBSD, define PCVT_SUPPORT (System has PCVT console). 170934 170935 Noticed by Joerg Sonnenberger. This fixes problem with console 170936 switching. 170937 170938 This was in original imake NetBSD.cf. 170939 170940commit 26d2e45bdb0cf4d18ba7b0365425da49d60b3d5c 170941Author: Matthias Hopf <mhopf@suse.de> 170942Date: Tue Nov 14 15:33:07 2006 +0100 170943 170944 Bug #9023: Only check mice for "mouse" or "void" if identifier is != NULL. 170945 170946commit a724b7f1302ba7a59f140b521f13d2ddf0fcf9bf 170947Merge: 6facd958c f80a8ae6e 170948Author: Jeremy C. Reed <reed@glacier.reedmedia.net> 170949Date: Mon Nov 13 20:32:26 2006 -0600 170950 170951 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 170952 170953commit 6facd958c2e7098cb68eac2810a88b8cd40f19c0 170954Author: Jeremy C. Reed <reed@glacier.reedmedia.net> 170955Date: Mon Nov 13 20:23:06 2006 -0600 170956 170957 On DragonFLy, default to /dev/sysmouse (just like on FreeBSD). 170958 170959 TODO: use autoconf to handle this so the code is not platform 170960 based but feature based. 170961 170962commit f80a8ae6e6820378ea70ac0849cee8bf3055e0e8 170963Author: Adam Jackson <ajax@benzedrine.nwnk.net> 170964Date: Mon Nov 13 18:26:05 2006 -0500 170965 170966 Disable Xprt server build by default. 170967 170968commit 5bfe7f4dfc6ab51b6790d74faf11574385234423 170969Author: George Sapountzis <gsap7@yahoo.gr> 170970Date: Mon Nov 13 01:54:32 2006 +0200 170971 170972 DRI: call drmSetServerInfo() before drmOpen(). 170973 170974 Also, remove some libdrm remnants. 170975 170976commit acb5ff4c73ac4d52201d7c421f488e2ead5c8b9c 170977Author: Jurij Smakov <jurij@wooyd.org> 170978Date: Sat Nov 11 14:09:15 2006 +0200 170979 170980 ffs: handle 0 argument (bug #8968) 170981 170982 Handle an argument of 0 in ffs(), instead of looping indefinitely. 170983 Add an ffs prototype to dix.h, and add includes to ffs.c. 170984 (cherry picked from 34164e551e4c3909322d50b09835ca4ac1d49d68 commit) 170985 170986commit ca094684196886a4a1c10273049fae0705a3edc2 170987Author: Alan Coopersmith <alan.coopersmith@sun.com> 170988Date: Fri Nov 10 18:02:05 2006 -0800 170989 170990 Correct symlink-mesa.sh usage message 170991 170992commit 28337cc1060bc08adef81c304dd9ed02d0a0e37b 170993Author: Eric Anholt <eric@anholt.net> 170994Date: Thu Nov 9 19:51:17 2006 -0800 170995 170996 Fix typo before the last commit. 170997 170998commit e1720b1089328af80ca0cb85e5289ced05263f95 170999Author: Eric Anholt <eric@anholt.net> 171000Date: Thu Nov 9 18:53:37 2006 -0800 171001 171002 Bug #8868: Remove drm from SUBDIRS now that the directory is gone. 171003 171004commit 0dee48b8af3e054228aef0d15c1cb1c9e23790cc 171005Author: Keith Packard <keithp@mandolin.keithp.com> 171006Date: Wed Nov 8 23:17:55 2006 -0800 171007 171008 Add RRInit function to create resource types for RR objects. 171009 171010 To allow RandR objects to be created before the screen object exists, 171011 the resource types must be registered with the resource database. 171012 A driver wishing to create RandR objects must call RRInit before doing so. 171013 171014 Also, fix a segfault when setting Output data before it is associated with a 171015 screen. 171016 171017commit ec77a95a02329a2ee3a94d7de9d2a234aecb9ca0 171018Author: Keith Packard <keithp@mandolin.keithp.com> 171019Date: Wed Nov 8 21:36:35 2006 -0800 171020 171021 Allow RandR objects to be created before the associated ScreenRec. 171022 171023 xf86 drivers need to create RandR object in the PreInit stage, 171024 before the ScreenRec is allocated. Changing the RandR DIX code 171025 to permit this required the addition of functions that later associate the 171026 objects with the related screen. 171027 171028 An additional change is that modes are now global, and no longer associated 171029 with a specific screen. This change actually makes mode management cleaner 171030 as there is no more per-screen list of modes to deal with. 171031 171032 This changes the RandR 1.2 ABI/API for drivers. 171033 171034commit fd91630b73100e9d77ccb492c52807448bc772a5 171035Author: Dave Airlie <airlied@linux.ie> 171036Date: Thu Nov 9 09:30:33 2006 +1100 171037 171038 make X server use system libdrm - this requires libdrm >= 2.3.0 171039 171040 This patch blacklists the load "drm" line, moves some functions in dri module 171041 links dri module against libdrm, and removes the X copy of libdrm 171042 171043commit 0409e1627a167db2efc1355f292d3c02a6989ffc 171044Author: Adam Jackson <ajax@benzedrine.nwnk.net> 171045Date: Wed Nov 8 16:17:20 2006 -0500 171046 171047 'make dist' fixes. 171048 171049commit 7e0aeebb8f8a5dff3cb4d88756e535dd70edeec4 171050Author: Adam Jackson <ajax@benzedrine.nwnk.net> 171051Date: Wed Nov 8 12:03:37 2006 -0500 171052 171053 Bug #6786: Use separate defines for server's Fixes support level. 171054 171055commit b5d09d4adb8088719ff494a4281a793717046576 171056Author: Rich Coe <Richard.Coe@med.ge.com> 171057Date: Wed Nov 8 18:10:14 2006 +0200 171058 171059 CheckConnections: don't close down the server client (bug #7876) 171060 When an appgroup is shutting down, the list of clients can change, so make 171061 sure we're not trying to shut the server down. 171062 171063commit 0567a6337b84fa045b5732e98203f488274aa2a2 171064Author: Bram Verweij <amverweij@gmail.com> 171065Date: Wed Nov 8 18:00:52 2006 +0200 171066 171067 xfree86/linux acpi: fix tokenising 171068 Split on a space, rather on the 'video' string, as strtok takes a char, 171069 not a string. 171070 171071commit e7900d68c3ee657158813f0650886d680c0a9a3c 171072Author: Daniel Stone <daniel@fooishbar.org> 171073Date: Wed Nov 8 15:36:47 2006 +0200 171074 171075 remove trailing whitespace 171076 Whitespace police in full effect. 171077 171078commit 2035f115b7db3c4b7deabeab0d814b3107d6ef30 171079Author: Daniel Stone <daniel@fooishbar.org> 171080Date: Wed Nov 8 15:28:10 2006 +0200 171081 171082 xfree86: remove xf86Keymap.h 171083 171084commit 809e2841aaa54662a42498aacec558bc813bba1c 171085Author: Daniel Stone <daniel@fooishbar.org> 171086Date: Wed Nov 8 15:27:58 2006 +0200 171087 171088 xfree86: add general handler, port ACPI to it (bug #5665) 171089 171090 Add 'general' handler registration, which will not trigger DPMS when an 171091 event comes in. 171092 Make ACPI use this. 171093 171094commit b5438f7fb2879e0226b578f60b22a490e73c3a83 171095Author: Daniel Stone <daniel@fooishbar.org> 171096Date: Wed Nov 8 15:24:58 2006 +0200 171097 171098 Get*Events: massive reorganisation 171099 171100 Reorganise the code logically, and put more comments in. 171101 Clip valuators in proximity calls. 171102 171103commit b0a7443ca8bd224994f98c1c1e25de88f2573224 171104Author: Daniel Stone <daniel@fooishbar.org> 171105Date: Wed Nov 8 15:23:52 2006 +0200 171106 171107 configure.ac: axe redundant SDL test 171108 171109commit 332f179db7c38310db1f752d45f51b6d8301fd30 171110Author: Daniel Stone <daniel@fooishbar.org> 171111Date: Wed Nov 8 15:23:35 2006 +0200 171112 171113 kdrive/sdl: fix compilation 171114 171115commit 988757a44197c91027109076e2e33ff6510ed71d 171116Author: Daniel Stone <daniel@fooishbar.org> 171117Date: Wed Nov 8 14:06:51 2006 +0200 171118 171119 config: error messages non-fatal, and before we free them 171120 171121 Move error messages before we free the error structure, and make them all 171122 non-fatal. 171123 171124commit e2b78df5800fb4e3f0ed01b38a1974ba3577949c 171125Author: Daniel Stone <daniel@fooishbar.org> 171126Date: Tue Nov 7 11:13:56 2006 +0200 171127 171128 xkb: warning fix 171129 171130commit b55007d8cc9d20baa23d5de67683e414c827d3e5 171131Author: Daniel Stone <daniel@fooishbar.org> 171132Date: Tue Nov 7 11:13:32 2006 +0200 171133 171134 SyntheticMotion: don't dereference sprite.screen when not using Xinerama 171135 171136commit f93d10ce9bb4a6de83b561f44fb7b046def16234 171137Author: Daniel Stone <daniel@fooishbar.org> 171138Date: Mon Nov 6 18:33:45 2006 +0200 171139 171140 dix: remove staggeringly broken vendor workarounds 171141 171142 Dear SGI, 171143 No. 171144 171145 Scant regards, 171146 Daniel 171147 171148commit 389275d240e4ba19d62fda0f138a45c7ecb245ff 171149Author: Daniel Stone <daniel@fooishbar.org> 171150Date: Sun Nov 5 02:47:59 2006 +0200 171151 171152 XkbCopyKeymap: don't iterate broken types, or dereference null pointers 171153 171154 Don't iterate invalid destination types (>= num_types) when coping key 171155 types. 171156 Don't free key_aliases if it's NULL (theoretical, but sure). 171157 Make sure dst's label_font gets allocated if it's NULL. 171158 (Thanks, Chris Lee.) 171159 171160commit d585b4189aff8d7952847f75b19b4f092ab3b88b 171161Author: Daniel Stone <daniel@fooishbar.org> 171162Date: Sat Nov 4 21:47:55 2006 +0200 171163 171164 xkb: fix uninitialised warning 171165 171166 Fix uninitialised warning with memset(); we never actually use it 171167 uninitialised, but gcc doesn't know that. 171168 171169commit b6d7b537ed8975363ad0f7c4180a62822358e418 171170Author: Daniel Stone <daniel@fooishbar.org> 171171Date: Sat Nov 4 21:43:22 2006 +0200 171172 171173 os: fix sun extensions test 171174 171175 'else if' is not very valid, plus the logic is kind of broken, so just 171176 move it outside the ifdef in the first place. 171177 171178commit 8ba0c7b62c78dead722b0c8aa414f37bac4414b7 171179Author: Daniel Stone <daniel@fooishbar.org> 171180Date: Sat Nov 4 21:41:03 2006 +0200 171181 171182 xace: avoid 'unused variable pScreen' 171183 171184 Initialise pScreen explicitly, as REGION_* macros ignore pScreen. 171185 171186commit c3ea1f7db494365032526dc06a7283384bd0ecd1 171187Author: Daniel Stone <daniel@fooishbar.org> 171188Date: Sat Nov 4 21:38:31 2006 +0200 171189 171190 dix/mi: still more warning fixes 171191 171192 Fix up prototypes for PrintChildren and PrintWindowTree in the dix. 171193 Make miPrintRegion be unconditionally defined, and move the prototype into 171194 regionstr.h. 171195 Change a bunch of ScreenPtr pScreen = foo; to 171196 ScreenPtr pScreen; pScreen = foo; in window.c, so we avoid unused variable 171197 references (as inline REGION_* doesn't reference pScreen). 171198 171199commit 3a9b96425851b495503bd2eb0fd0d01c08f6a097 171200Author: Daniel Stone <daniel@fooishbar.org> 171201Date: Sat Nov 4 21:33:09 2006 +0200 171202 171203 dix: add missing prototypes 171204 171205 Add missing prototype for ffs, and include headers from ffs.c. 171206 Move PostSyntheticMotion prototype to input.h. 171207 171208commit 6716488fa256798070017232405b107d5c985479 171209Author: Daniel Stone <daniel@fooishbar.org> 171210Date: Sat Nov 4 21:30:23 2006 +0200 171211 171212 dix: remove unused debug code 171213 171214commit 51813d77bfb84609a58a98e678efe9b6c0bf5503 171215Author: Daniel Stone <daniel@fooishbar.org> 171216Date: Sat Nov 4 21:29:05 2006 +0200 171217 171218 dix: remove unused variable 171219 171220commit 844090a5b557705dd0adce2b7ed98813b5104d85 171221Author: Daniel Stone <daniel@fooishbar.org> 171222Date: Sat Nov 4 21:21:29 2006 +0200 171223 171224 xfree86/xf86misc: warning fixes 171225 171226commit 61b570d0c1eb448f0aa08b4598118f0d43bc7345 171227Author: Daniel Stone <daniel@fooishbar.org> 171228Date: Sat Nov 4 21:21:09 2006 +0200 171229 171230 xfree86/os-support: update prototype for OSRingBell 171231 171232commit c51fadc07d938f6a3edfd5620170fcb7d6486a11 171233Author: Daniel Stone <daniel@fooishbar.org> 171234Date: Sat Nov 4 21:20:45 2006 +0200 171235 171236 mieq: annotate with some more comments 171237 171238commit 578899139f133746634a7bf8845e25362b5dfca2 171239Author: Daniel Stone <daniel@fooishbar.org> 171240Date: Sat Nov 4 20:35:55 2006 +0200 171241 171242 RemoveGeneralSocket: don't touch EnabledDevices 171243 171244 RemoveGeneralSocket. Harmless, but. 171245 171246commit ae58d349c1cf5d63ad3616c485baa858350978d5 171247Author: Laurence Withers <l@lwithers.me.uk> 171248Date: Sat Nov 4 19:34:37 2006 +0200 171249 171250 CreateColormap: fix return value (bug #7083) 171251 171252 Return BadMatch when an unsupported visual type is given, not BadValue -- 171253 this is correct according to the spec. 171254 171255commit 96f78e3886791b723ccd9ba40bea701603537b0c 171256Author: Erik Andren <erik.andren@gmail.com> 171257Date: Sat Nov 4 19:29:49 2006 +0200 171258 171259 remove XFree86 changelogs (bug #7262) 171260 171261 Without being able to tie these to specific commits, the text changelog is 171262 useless, as well as being huge. 171263 171264commit 5a40448f2d0ac2c86c617bebe3fb649174bf0d7f 171265Author: Eric Anholt <eric@anholt.net> 171266Date: Tue Nov 7 15:48:05 2006 -0800 171267 171268 A couple more cases of error message before freeing strings. 171269 171270commit 05f1c302460a14c8fa9a943a12d69adcd3c30d58 171271Merge: 3e7e0e350 46af6d1e9 171272Author: Adam Jackson <ajax@benzedrine.nwnk.net> 171273Date: Tue Nov 7 18:42:54 2006 -0500 171274 171275 Merge branch 'autoconfig-for-7.2' 171276 171277commit 3e7e0e35094d09e0e764818ed125314be75be01a 171278Author: Eric Anholt <eric@anholt.net> 171279Date: Tue Nov 7 14:13:23 2006 -0800 171280 171281 Report the error before freeing the error strings. 171282 171283commit 4e6e4baead6c565363abbcd9e06cc685be121596 171284Author: Eric Anholt <eric@anholt.net> 171285Date: Tue Nov 7 13:13:53 2006 -0800 171286 171287 Fix build on FreeBSD after input-hotplug. 171288 171289commit 20e9144c0746943624ff77a61791b8596f3f8458 171290Author: Keith Packard <keithp@neko.keithp.com> 171291Date: Tue Nov 7 12:49:28 2006 -0800 171292 171293 Add $(DIX_CFLAGS) to remaining Makefile.am files 171294 171295commit 5e946dd853a4ebc2722ae023429ce5797de3d7a6 171296Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 171297Date: Tue Nov 7 13:50:19 2006 -0500 171298 171299 Bug #8937: Extension setup functions not called on server resets 171300 171301commit 1dcda4f3c56214464c0b6123fea6daa69aae69fc 171302Author: Keith Packard <keithp@neko.keithp.com> 171303Date: Tue Nov 7 01:29:51 2006 -0800 171304 171305 Avoid dereferencing sprite.screen when Xinerama is not running. (#8925) 171306 171307 With Xinerama support built into the X server but not in use, 171308 sprite.screen is NULL and yet the SyntheticMotion 171309 macro would dereference it. Avoid that by just passing sprite.screen 171310 to PostSyntheticMotion which can then dereference it when Xinerama is 171311 enabled. 171312 171313 Also, define PostSyntheticMotion in dixevents.h and include dixevents.h in 171314 getevents.c 171315 171316commit c20d3bf7533da0bf26beaf7d8c359d18edbd70e8 171317Merge: 028bbdc04 3d39c02fe 171318Author: Keith Packard <keithp@neko.keithp.com> 171319Date: Tue Nov 7 01:21:28 2006 -0800 171320 171321 Merge branch 'origin' into randr-1.2 171322 171323commit 3d39c02fe6aaa602c52f1d4f0ea6cd3bd000cf9f 171324Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 171325Date: Mon Nov 6 21:25:52 2006 -0500 171326 171327 More work on Bug #8875: revert previous fix and try using client argument 171328 instead of serverClient. Also don't use totalClientSize as it is not 171329 initialized until after the first call to InitClient. 171330 171331commit 75fe0670eb1f71144246f1c20759d58788bbee00 171332Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 171333Date: Mon Nov 6 15:30:25 2006 -0500 171334 171335 whitespace adjust 171336 171337commit 0539d9cf2423fc0bed6f5c413beba3080f8abd85 171338Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 171339Date: Mon Nov 6 15:29:17 2006 -0500 171340 171341 Bug #8875: Security extension causes Xorg to core dump on server reset 171342 171343commit aa0261a98e9d5b1349b33e2639bd83c556dd4000 171344Author: Juliusz Chroboczek <jch@pps.jussieu.fr> 171345Date: Mon Nov 6 02:29:49 2006 +0100 171346 171347 Improve vm86 error handling in Xvesa. 171348 171349commit 6b2c65fdd169037c6ede250d4a8fec3d29a080ae 171350Author: Juliusz Chroboczek <jch@pps.jussieu.fr> 171351Date: Mon Nov 6 00:30:09 2006 +0100 171352 171353 Fix typo in Xvesa: incorrect reporting of DAC capabilities. 171354 171355commit 028bbdc0417173803695808ba9f48498519273a3 171356Merge: 50504c68e 8deaaa312 171357Author: Keith Packard <keithp@neko.keithp.com> 171358Date: Sat Nov 4 17:46:26 2006 -0800 171359 171360 Merge master back to randr-1.2 171361 171362commit 50504c68e1d407232cf83465981b235e542ef31f 171363Merge: 8b87ce197 cde8806c2 171364Author: Keith Packard <keithp@neko.keithp.com> 171365Date: Sat Nov 4 17:43:19 2006 -0800 171366 171367 Merge branch 'randr-1.2-origin' into randr-1.2 171368 171369commit 8b87ce19741753eafbd99e7093bc3dea8f26e838 171370Author: Keith Packard <keithp@neko.keithp.com> 171371Date: Sat Nov 4 17:41:25 2006 -0800 171372 171373 Allow X server to build against libdrm 2.1 171374 171375commit 7ffbe9d232e3a4621a204448d67e434736465cbe 171376Author: Keith Packard <keithp@neko.keithp.com> 171377Date: Sat Nov 4 17:41:09 2006 -0800 171378 171379 Add DIX_CFLAGS to hw/vfb/Makefile.am 171380 171381commit 2db62bce0725ba2d88cbe40fc440b6bda45046f3 171382Author: Keith Packard <keithp@neko.keithp.com> 171383Date: Sat Nov 4 17:40:34 2006 -0800 171384 171385 Define fbHasVisualTypes in fb.h as it is exported 171386 171387commit 8deaaa312ad7f9b492a2ae8ad17d74650112c25c 171388Author: Bernhard Rosenkraenzer <bero@arklinux.org> 171389Date: Sat Nov 4 18:59:39 2006 +0200 171390 171391 automake: avoid use of reserved _SOURCES keyword (bug #8866) 171392 Avoid using _SOURCES unless we're directly referencing a program or 171393 library to be built; use _SRCS instead. Shuts automake 1.10 up. 171394 171395commit f72927d26cd112d321f7bf187df3c740b3129d22 171396Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 171397Date: Sat Nov 4 19:00:57 2006 +0200 171398 171399 xfree86/hurd: remove OsMouseProc (bug #5613) 171400 Remove OsMouseProc, let the mouse driver deal with it. 171401 171402commit 0273610578485564c3c0be11b336b6554cc31b43 171403Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 171404Date: Sat Nov 4 19:02:04 2006 +0200 171405 171406 xfree86/hurd: add bell support, remove SERVER/LOADER defines (bug #5613) 171407 Add no-op bell ringing support, and remove obsolete @SERVER_DEFINES@ and 171408 @LOADER_DEFINES@ from CFLAGS. 171409 171410commit ade4bf09076d13dbf3549c0a2d987a0afe76d5c1 171411Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 171412Date: Sat Nov 4 19:03:13 2006 +0200 171413 171414 mesa/indirect: include byteswap.h on GNU userland (bug #5613) 171415 Include byteswap.h on all GNU-userland systems (including with the Hurd 171416 and FreeBSD kernels), not just Linux. 171417 171418commit 412e93349e1656c9650115328af4be0e59a66f74 171419Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 171420Date: Sat Nov 4 19:05:02 2006 +0200 171421 171422 kdrive: make building of Linux support conditional (bug #5613) 171423 Only try to build Linux support on Linux. We should probably disable all 171424 OS-dependent DDXes if we don't have a workable OS (and only build 171425 Xephyr/Xfake), but that's future work. 171426 171427commit cde8806c2930788ba8076e94651d391e45f3ccdb 171428Author: Eric Anholt <eric@anholt.net> 171429Date: Fri Nov 3 16:36:34 2006 -0800 171430 171431 Don't bump the refcnt if the new mode is NULL. 171432 171433commit 97fd471a627be185bee8cda3f709cfccea3fa12d 171434Author: Aaron Plattner <aplattner@nvidia.com> 171435Date: Fri Oct 27 12:36:56 2006 -0700 171436 171437 Fix standard VESA modes. 171438 171439 The built-in mode timings were off slightly for the 640x480@60, 640x480@72, 171440 and 1024x768@75 modes. 171441 171442commit 22ee2e4e1f1d9fd9ca9f25c9bf25370034b771d4 171443Merge: 49a26681b bd0c82965 171444Author: Alan Coopersmith <alan.coopersmith@sun.com> 171445Date: Fri Nov 3 12:54:56 2006 -0800 171446 171447 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 171448 171449commit 49a26681b2bdd95ed65c425f1fa1441d2f092a6e 171450Author: Alan Coopersmith <alan.coopersmith@sun.com> 171451Date: Fri Nov 3 12:54:43 2006 -0800 171452 171453 Add DTrace probe points for X server <-> client communications 171454 171455 See http://people.freedesktop.org/~alanc/dtrace/ for more details 171456 171457commit bd0c829654903ca45543dfa59cda967c4fafd8ac 171458Author: Bjorn Helgaas <bjorn.helgaas@hp.com> 171459Date: Fri Nov 3 18:54:06 2006 +0100 171460 171461 Do not map full 0-1MB legacy range 171462 171463 If we're mapping something in the "legacy range" (0-1Mb), we shouldn't 171464 expand the requested range to the entire 0-1Mb range. Typically this 171465 is for mapping the VGA frame buffer, and some platforms support mmap of 171466 the frame buffer but not the entire 0-1Mb range. 171467 171468 For example, HP sx1000 and sx2000 ia64 platforms can have memory from 171469 0-0x9ffff, VGA frame buffer from 0xa0000-0xbffff, and memory from 171470 0xc0000-0xfffff. On these platforms, we can't map the entire 0-1Mb 171471 range with the same attribute because the memory only supports WB, 171472 while the frame buffer supports only UC. But an mmap of just the 171473 frame buffer should work fine. 171474 171475commit c1828a8ff51c8db326c47e6710f4f42fab94fb6d 171476Author: Egbert Eich <eich@freedesktop.org> 171477Date: Fri Nov 3 18:32:48 2006 +0100 171478 171479 Fixing mach64 driver bailing out on ia64 171480 171481 Mach64 driver bails out on ia64 because it cannot map device 171482 memory. It turns out that some bogus and unneeded code attempts 171483 to find the root bridge of the device and fails to do so proberly 171484 as there this host-to-pci bridge is not existant. This code has 171485 been around for years although it completely unclear what it had 171486 been intended for. Fixing this by eliminating the bogus code. 171487 171488commit d50fc413b39f52663b46084c28e81fc4933a7b49 171489Author: Matthias Hopf <mhopf@suse.de> 171490Date: Thu Nov 2 18:53:41 2006 +0100 171491 171492 Fix device path in altixPCI.c to be domain aware. 171493 171494commit 1d731fc54a2cf5d3f353d8ee1c7c4989df27f011 171495Author: Matthias Hopf <mhopf@suse.de> 171496Date: Thu Nov 2 18:50:15 2006 +0100 171497 171498 Add domain support to linuxPciOpenFile(). 171499 171500 Loosely based on patch from David S. Miller <davem@davemloft.net> 171501 See also bug #2368. 171502 171503commit caaa113acf4144fd47a1ac93ca440d78d1983e54 171504Author: Matthias Hopf <mhopf@suse.de> 171505Date: Thu Nov 2 18:38:45 2006 +0100 171506 171507 Fixing domain support for ia64 171508 171509 ia64 specific functions + defines. 171510 Still uses /proc interface for some scaning code. 171511 Based on code from Egbert Eich <eich@freedesktop.org>. 171512 171513commit 16c5043fc0c84b14323cd211c2645106455ac320 171514Author: Matthias Hopf <mhopf@suse.de> 171515Date: Thu Nov 2 18:22:09 2006 +0100 171516 171517 Fix 2 warnings. 171518 171519commit 072c022e731c3aadf34096f16364e29df47280d2 171520Author: Matthias Hopf <mhopf@suse.de> 171521Date: Thu Nov 2 17:58:19 2006 +0100 171522 171523 Fix obviously wrong boundary checks + cleanup unused vars. 171524 171525 Also disable compilation of code that is no longer used anywhere in the whole 171526 Xserver tree. 171527 171528commit 5afc6c1a14fea2966017493b045fa7209faeb8eb 171529Author: Matthias Hopf <mhopf@suse.de> 171530Date: Thu Nov 2 15:42:03 2006 +0100 171531 171532 Added linux 2.6 compatible domain aware device scanning code. 171533 171534 Additional scanning code uses the /sys interface of 2.6 kernels. 171535 Cleaned up the use of tags and already split domain/bus/dev/fn. 171536 171537commit 6319f7d713971f70f06166480f069eca3bcace36 171538Author: Egbert Eich <eich@freedesktop.org> 171539Date: Thu Nov 2 12:50:52 2006 +0100 171540 171541 Make int10 fully domain aware. 171542 171543commit f4dd2665b0f9aa9c00a5152c73bc72cb7514eeb5 171544Author: Matthias Hopf <mhopf@suse.de> 171545Date: Thu Nov 2 12:36:12 2006 +0100 171546 171547 Added missing domain stripping in already domain aware code. 171548 171549commit 46901063e8edc82b67989f4e5eec39d17c67dc98 171550Author: Matthias Hopf <mhopf@suse.de> 171551Date: Thu Nov 2 12:25:03 2006 +0100 171552 171553 Build with -D_PC on ix86 only. 171554 171555commit 543b397277d1f03b8091e44812010abcd5d80102 171556Merge: 4056e6e79 645d05762 171557Author: Keith Packard <keithp@neko.keithp.com> 171558Date: Thu Nov 2 19:00:35 2006 -0800 171559 171560 Merge branch 'origin' into randr-1.2 171561 171562commit 037f23e6f8fbe6e6fc8e71ed21958fc553df72d0 171563Author: Alan Coopersmith <alan.coopersmith@sun.com> 171564Date: Thu Nov 2 18:30:58 2006 -0800 171565 171566 Convert Xprt DDX to new motion history api 171567 171568commit 7dc54a40e900cbea1e509620623b091d54a3c2d1 171569Author: Alan Coopersmith <alan.coopersmith@sun.com> 171570Date: Thu Nov 2 16:38:47 2006 -0800 171571 171572 Remove references to xf86Info.kbd* from solaris code 171573 171574commit e46f7f78b362e76f5a553184e3f5ec7e109aa39d 171575Author: Alan Coopersmith <alan.coopersmith@sun.com> 171576Date: Thu Nov 2 16:38:12 2006 -0800 171577 171578 sun_bell.c needs to #include "xf86_OSlib.h" 171579 171580commit 58bf9a142d1957f4d77038ee4ce7b1116b1f7955 171581Author: Alan Coopersmith <alan.coopersmith@sun.com> 171582Date: Thu Nov 2 14:43:02 2006 -0800 171583 171584 PostSyntheticMotion needs to be extern, not static, since it's in getevents.c 171585 171586commit 1ecd45fb8e4250fb51daa2bdf1a960af0f8b53fb 171587Author: Daniel Stone <daniel@fooishbar.org> 171588Date: Thu Nov 2 04:51:03 2006 +0200 171589 171590 Makefile.am: add config to DIST_SUBDIRS as well 171591 171592commit 01afa533aa872d1a101a41153f95d800e68fea3e 171593Author: Daniel Stone <daniel@fooishbar.org> 171594Date: Thu Nov 2 04:32:37 2006 +0200 171595 171596 xfree86/os-support: axe more unused files 171597 171598commit 57c1409151cb1f6e0e528fb92ebda58f86f12c1a 171599Author: Daniel Stone <daniel@fooishbar.org> 171600Date: Thu Nov 2 04:21:06 2006 +0200 171601 171602 xfree86/os-support: remove unused file from dist 171603 171604commit de5a4c63747a417cdece919f4fb5a4004a3ee7bb 171605Author: Daniel Stone <daniel@fooishbar.org> 171606Date: Thu Nov 2 04:18:33 2006 +0200 171607 171608 xkb: note that we allow full xi interaction 171609 We now allow maps to be set (etc) on different keyboards, so stop putting 171610 XkbXI_KeyboardsMask in unsupported. 171611 171612commit 1d65429a9e03871969552d0c31b022546cc46b12 171613Author: Daniel Stone <daniel@fooishbar.org> 171614Date: Thu Nov 2 04:12:55 2006 +0200 171615 171616 xfree86: don't attempt to enable and disable non-DIX devices 171617 Don't try to enable and disable devices with no entry in the DIX, such as 171618 the evdev brain. 171619 171620commit 64139c1950ea825c0a0124abc5f88499e91f797f 171621Author: Daniel Stone <daniel@fooishbar.org> 171622Date: Thu Nov 2 03:22:09 2006 +0200 171623 171624 bump to 1.2.99.0 171625 171626commit 18c246a13b887b865de6a17e6cd1c259b9bc383d 171627Merge: 794f2e729 4843d823f 171628Author: Daniel Stone <daniel@fooishbar.org> 171629Date: Thu Nov 2 03:21:37 2006 +0200 171630 171631 Merge branch 'input-hotplug' 171632 171633commit 4843d823f4d38d8bd468ce3a8feddbff229ed416 171634Merge: ba9f5138f a7b944f0d 171635Author: Daniel Stone <daniel@fooishbar.org> 171636Date: Thu Nov 2 03:18:13 2006 +0200 171637 171638 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 171639 171640commit ba9f5138fc32a7a7b97bcf941bc92751b7c6c2c0 171641Author: Daniel Stone <daniel@fooishbar.org> 171642Date: Thu Nov 2 03:16:10 2006 +0200 171643 171644 xfree86: allow starting with no input devices 171645 Add a server flag (AllowEmptyInput), which will inhibit adding the 171646 standard keyboard and mouse drivers, if there are no input devices in the 171647 config file. 171648 171649commit be291a6d9764cf29a7d9a8114d47d9f41ce856e9 171650Merge: a2d624210 6fdfd9dad 171651Author: Daniel Stone <daniel@fooishbar.org> 171652Date: Thu Nov 2 03:15:25 2006 +0200 171653 171654 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 171655 171656commit a7b944f0d96c3e0e15e75378a04def1ac96089fb 171657Author: Alan Coopersmith <alan.coopersmith@sun.com> 171658Date: Wed Nov 1 16:17:49 2006 -0800 171659 171660 If getpeerucred() is available, include pid & zoneid in audit messages too 171661 171662commit fbfb35189ef6666707097704b43e052cb2f919ae 171663Author: Alan Coopersmith <alan.coopersmith@sun.com> 171664Date: Wed Nov 1 15:11:48 2006 -0800 171665 171666 Bug #1997: AUDIT messages should contain uid for local accesses 171667 171668 <https://bugs.freedesktop.org/show_bug.cgi?id=1997> 171669 171670commit a2d6242106bb3a440faa9cad157e0120dbfa7b6e 171671Author: Daniel Stone <daniel@fooishbar.org> 171672Date: Thu Nov 2 00:46:33 2006 +0200 171673 171674 kdrive:remove Change{Keyboard,Pointer}Device 171675 This was removed in the DIX, so just axe it. 171676 171677commit d9a5e3e964b6c91fbca88b50674fce9660b972a4 171678Author: Daniel Stone <daniel@fooishbar.org> 171679Date: Thu Nov 2 00:45:23 2006 +0200 171680 171681 XkbCopyKeymap: be more careful with levels, allocate compat/geom 171682 Take various extra precautions with copying levels across (thanks Chris 171683 Lee for a gdb session), including allocating when we don't already have a 171684 coherent map. 171685 Only free type components if they're present. 171686 Allocate geometry and compat components if we don't already have them in 171687 the dest map. 171688 171689commit 6fdfd9dad91d7b7aa292f8c4d268dd27c34de8d3 171690Author: Eric Anholt <eric@anholt.net> 171691Date: Wed Nov 1 14:29:59 2006 -0800 171692 171693 Fix several cases where optimized paths were hit when they shouldn't be. 171694 171695 This fixes a number of rendercheck cases. 171696 171697commit 40f84793bca40dcc6883d51aefa1bda44bd1ac61 171698Author: Alan Coopersmith <alan.coopersmith@sun.com> 171699Date: Wed Nov 1 14:34:46 2006 -0800 171700 171701 Propogate $LIBS for xtrans, clock_gettime, libm, etc. to libs used for each server 171702 171703commit d7d931abe01a8cf555b027f2bcfcccd5e9053e52 171704Author: Daniel Stone <daniel@fooishbar.org> 171705Date: Wed Nov 1 23:48:58 2006 +0200 171706 171707 configure.ac: remove check for rate/period 171708 Keyboard stuff is now handled in the kbd driver. 171709 171710commit 81728558a044fdde0e1d63da7b6314755f77296e 171711Author: Daniel Stone <daniel@fooishbar.org> 171712Date: Wed Nov 1 23:10:26 2006 +0200 171713 171714 input: add non-keyboard bell ringing function 171715 Add a generic 'ring the bell' function (console bell on Linux and BSD, 171716 /dev/audio on Solaris), and add DDX functions for this. Make this the 171717 core keyboard's bell. 171718 Port Xvfb and Xnest to this. 171719 Port XFree86 to this, with OS-specific hooks for Linux, BSD, and Solaris 171720 taken from foo_io.c in the old layer. 171721 171722commit 3df454719f9cbf6a046cb7458019ec621b3b42ee 171723Author: Daniel Stone <daniel@fooishbar.org> 171724Date: Wed Nov 1 23:02:57 2006 +0200 171725 171726 kdrive: add KdRingBell prototype 171727 171728commit 4056e6e79a4e37101d298ae29139c83d3816368b 171729Author: Keith Packard <keithp@mandolin.keithp.com> 171730Date: Wed Nov 1 00:29:46 2006 -0800 171731 171732 Move physical size from mode to output. 171733 171734 Modes can be shared across different sized monitors this way. 171735 171736 Also caught some missing byteswapping and an incorrect return type. 171737 171738commit c03311a1e78daa291477a67b1bb7206772108c5d 171739Author: Alan Coopersmith <alan.coopersmith@sun.com> 171740Date: Tue Oct 31 16:05:48 2006 -0800 171741 171742 Fix automake error: BUILT_SOURCES was defined multiple times on Solaris 171743 171744commit a2434ec5f3c9dc79d1f05c2d704a82a766718ed4 171745Author: Alan Coopersmith <alan.coopersmith@sun.com> 171746Date: Tue Oct 31 15:57:59 2006 -0800 171747 171748 Make _POSIX_C_SOURCE hack work with Solaris headers 171749 171750 Solaris headers are very literal - if you ask for POSIX_C_SOURCE 199309L, 171751 they limit to only the functions in that standard and no more, unless you 171752 also specify __EXTENSIONS__ to allow functions beyond the standard base. 171753 171754commit 645d0576205532a3610ae351267d5b84d76236bd 171755Author: Matthieu Herrb <matthieu.herrb@laas.fr> 171756Date: Sun Oct 29 18:19:56 2006 +0100 171757 171758 Handle building in a separate objdir 171759 171760commit 59584c375f4e4b2670a92002ecb7a78a0bc50cce 171761Author: Matthieu Herrb <matthieu.herrb@laas.fr> 171762Date: Sun Oct 29 17:49:46 2006 +0100 171763 171764 kill GNU-make'ism. 171765 171766commit 0107320fac0913aae2cb169992e31c670b4bd2f7 171767Merge: 06b6b971d a34446f5b 171768Author: Thomas Hellstrom <thomas-at-tungstengraphics-dot-com> 171769Date: Sun Oct 29 15:23:35 2006 +0100 171770 171771 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 171772 171773commit 06b6b971d065226b983ba25da7ea8236ec37df04 171774Author: Thomas Hellstrom <thomas-at-tungstengraphics-dot-com> 171775Date: Sun Oct 29 15:22:37 2006 +0100 171776 171777 Make sure we have 64-bit file-offsets in libdrm. 171778 171779commit 644ee2434a4e09f5baea00a486911f8c24b9a261 171780Author: Daniel Stone <daniel@fooishbar.org> 171781Date: Sun Oct 29 03:52:11 2006 +0300 171782 171783 kdrive: avoid null dereferences 171784 Avoid jumping into NULL pointers for OS or card functions which do not 171785 exist. 171786 171787commit 738d2e88171bcce8782d670a1cda9a1d941d7977 171788Author: Daniel Stone <daniel@fooishbar.org> 171789Date: Sun Oct 29 03:48:02 2006 +0300 171790 171791 kdrive: move bell ringing into an OS function 171792 Move the bell into an OS function, and use that if it's declared; else, 171793 fall back to using the driver's function. 171794 Remove the Linux keyboard bell function; just move it into the OS layer. 171795 Use named initialisers when converting the old structures, and eliminate 171796 unused functions. 171797 171798commit 96e32805d12fc36f0fa0926dbfb0dd8a5cadb739 171799Author: Daniel Stone <daniel@fooishbar.org> 171800Date: Sun Oct 29 03:43:34 2006 +0300 171801 171802 Xi: disallow changing core keyboard and pointer 171803 Just short-circuit the change core keyboard/pointer requests. 171804 171805commit 68f595ca6c7883e030947b7f95c50e92aa733f2b 171806Author: Daniel Stone <daniel@fooishbar.org> 171807Date: Sun Oct 29 03:41:34 2006 +0300 171808 171809 GetTimeInMillis: use correct units for clock_gettime 171810 Make sure we're treating the nanoseconds as a long, not an int, so we 171811 don't overflow. 171812 171813commit 51a06b3c44509c72279b5cfcf2b52b9a35c461b0 171814Author: Daniel Stone <daniel@fooishbar.org> 171815Date: Sun Oct 29 03:40:57 2006 +0300 171816 171817 WaitForSomething: only rewind when delta is more than 250ms 171818 Only rewind time when we're more than (original delta + 250ms) away from 171819 executing the timer. 171820 When we're walking the timer list, use a goto to iterate all of them from 171821 the start again, since timers may drop out of the list. 171822 Don't bother trying to be smart in TimerSet, we'll pick it up in 171823 WaitForSomething anyway. 171824 171825commit a34446f5b3d90714969a90583c49cb1eae1c9651 171826Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 171827Date: Fri Oct 27 13:43:43 2006 -0400 171828 171829 Add missing file to list in Makefile.am 171830 171831commit 92d04e746bd9b8ad3ee217c165ace20468e079cf 171832Author: Thomas Hellstrom <thomas-at-tungstengraphics-dot-com> 171833Date: Fri Oct 27 18:26:30 2006 +0200 171834 171835 Import libdrm functions for the drm memory manager. 171836 171837commit e3d3d29db5eef057b4e8294377c9c3147436bc2f 171838Merge: 92387e99d 004d00e66 171839Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 171840Date: Thu Oct 26 20:21:16 2006 -0400 171841 171842 Merge branch 'master' into my-XACE-SELINUX 171843 171844commit 92387e99d085b0b081fcedb2f20304eb0ac536b1 171845Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 171846Date: Thu Oct 26 20:20:57 2006 -0400 171847 171848 Change symbol in libaudit library test. 171849 171850commit 196c5836f463c28f633bbba847f59acd5935359d 171851Author: Daniel Stone <daniel@fooishbar.org> 171852Date: Fri Oct 27 01:27:31 2006 +0300 171853 171854 CoreKeyboardProc: annotate with FIXME 171855 Setting an empty keymap by default isn't wildly useful. 171856 171857commit f9a1e456f8a4eaa1a9c71fd0fe5231140975c22d 171858Author: Daniel Stone <daniel@fooishbar.org> 171859Date: Fri Oct 27 01:25:39 2006 +0300 171860 171861 CoreKeyboardProc: don't leak keymap and modmap 171862 SetKeySymsMap does a copy here, so try not to leak them. 171863 171864commit a5be65401769fabcb5001dc63035c69f9e4a2712 171865Author: Daniel Stone <daniel@fooishbar.org> 171866Date: Fri Oct 27 01:25:21 2006 +0300 171867 171868 mieqEnqueue: only compare DEVICE_BITS of deviceid 171869 Only compare DEVICE_BITS of the two deviceids, so we don't decide that 171870 a valuator event isn't for us, because (id | MORE_EVENTS) != id. 171871 171872commit 85212eb504f860b054eb0f0a5029fed86cb8d1c0 171873Author: Daniel Stone <daniel@fooishbar.org> 171874Date: Fri Oct 27 01:23:58 2006 +0300 171875 171876 getValuatorEvents: make sure we put MORE_EVENTS in the right places 171877 Make sure we put MORE_EVENTS in with the device id if there are, in fact, 171878 more valuator events coming. 171879 171880commit 794f2e7291ccb4e48f9fbfc8f08302e3aac0f79f 171881Author: Myron Stowe <myron.stowe@hp.com> 171882Date: Thu Oct 26 20:38:58 2006 +0300 171883 171884 xfree86: re-enable chipset-specific drivers for Linux/ia64 171885 Re-enable chipset-specific support for Linux/ia64, by linking in 171886 lnx_ia64.c. 171887 171888commit 8c0556e7cb1de8c387ddd886a03a8f8afff1fd0e 171889Merge: cdc8a4b7b 004d00e66 171890Author: Daniel Stone <daniel@fooishbar.org> 171891Date: Thu Oct 26 15:21:22 2006 +0300 171892 171893 Merge branch 'master' into input-hotplug 171894 171895commit 004d00e6689f452fc9fdf91f5ffc6d6aed697d54 171896Author: Daniel Stone <daniel@fooishbar.org> 171897Date: Thu Oct 26 01:10:08 2006 +0300 171898 171899 GetTimeInMillis: simplify monotonic test 171900 We don't actually need to get the CPU clock ID, which means we don't need 171901 the monotonic_usable test. Since there's now only one branch, the 171902 compiler will treat that as likely, so we don't need xproto 7.0.9 anymore. 171903 171904 The fallthrough to gettimeofday() is preserved. 171905 171906commit cdc8a4b7b2f099b8860a54c5c9f488e6f7c4913a 171907Merge: 3da918a16 d28583329 171908Author: Daniel Stone <daniel@fooishbar.org> 171909Date: Thu Oct 26 00:28:30 2006 +0300 171910 171911 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 171912 171913commit d285833290316cb5dd1e7f1e52c96be3e9cf21cd 171914Author: Daniel Stone <daniel@fooishbar.org> 171915Date: Wed Oct 25 23:57:00 2006 +0300 171916 171917 GetTimeInMillis: spuport monotonic clock 171918 Add support for CLOCK_MONOTONIC from clock_gettime, and use that in 171919 GetTimeInMillis() if available, falling back to the old gettimeofday() 171920 implementation. 171921 171922 This is _slightly_ faster on some 64-bit architectures, and _slightly_ 171923 slower on others (though barely measurable). 171924 171925commit d3e57faffee63df1424a209d0418d3a712f91ae6 171926Author: Daniel Stone <daniel@fooishbar.org> 171927Date: Wed Oct 25 23:55:43 2006 +0300 171928 171929 WaitForSomething: allow time to rewind 171930 If time rewinds dramatically, reset all the timers to fix their expiry. 171931 171932commit e21604914dccece6bc64c69b55512d1f1a969235 171933Author: Keith Packard <keithp@neko.keithp.com> 171934Date: Wed Oct 25 09:48:23 2006 -0700 171935 171936 Merge master back in and clean up some unfinished code (closes 8745) 171937 171938commit becbda6d519a11c2c211afb8d46f9ea1a2676bc3 171939Author: Michel Dänzer <michel@tungstengraphics.com> 171940Date: Wed Oct 25 11:39:27 2006 +0200 171941 171942 Fix test for Option "IgnoreABI". 171943 171944 This option has plenty of potential for wasting the time of bug triagers 171945 without pretending it's always on. 171946 171947commit 59511974db843fa7e11133894cfc10c20fdaf60f 171948Merge: 054f8cd26 598ac7a83 171949Author: Keith Packard <keithp@neko.keithp.com> 171950Date: Tue Oct 24 17:26:20 2006 -0700 171951 171952 Merge branch 'master' into randr-1.2 171953 171954commit 598ac7a83698327b607084abaebcbd22f8d25fbb 171955Merge: 828c34e83 948a97b97 171956Author: Keith Packard <keithp@neko.keithp.com> 171957Date: Tue Oct 24 17:23:12 2006 -0700 171958 171959 Merge branch 'origin' 171960 171961commit 828c34e83ccdf3bcd2844d5af8b0cac4164b04ab 171962Author: Keith Packard <keithp@neko.keithp.com> 171963Date: Tue Oct 24 17:23:02 2006 -0700 171964 171965 Byte swap RRSelectInput enable flags. 171966 171967commit 3da918a16c8908fdfaf89f2a1bcaec19e01528a9 171968Author: Daniel Stone <daniel@fooishbar.org> 171969Date: Wed Oct 25 02:22:07 2006 +0300 171970 171971 mipointer: remember to update pointer location 171972 Update pointer location so it doesn't get quickly reset by the next 171973 pointer update. 171974 171975commit aabc087998e680c2fcf0ebc1c5022c1fe8f58f0c 171976Author: Daniel Stone <daniel@fooishbar.org> 171977Date: Wed Oct 25 02:21:39 2006 +0300 171978 171979 GetPointerEvents: always ensure correct number of events 171980 Ensure correct number of valuator events are returned, and that we always 171981 increment events correctly. 171982 171983commit 65cd5aa4d754624566c2263015f1a018d137fce1 171984Author: Daniel Stone <daniel@fooishbar.org> 171985Date: Wed Oct 25 01:12:45 2006 +0300 171986 171987 kdrive/input: remove unnecessary #ifdef XINPUTs 171988 Xi is now mandatory, so don't bother with the ifdefs. 171989 171990commit 0514d53e10b3521bb708a9cbde4bab525248eadb 171991Author: Daniel Stone <daniel@fooishbar.org> 171992Date: Wed Oct 25 01:10:44 2006 +0300 171993 171994 xfree86: remove motion history handling 171995 Remove motion history handling, as we now deal with this in the DIX. 171996 171997commit b1debebf8fe20ded20ba27e871fd1a6a9de029e3 171998Author: Daniel Stone <daniel@fooishbar.org> 171999Date: Wed Oct 25 01:10:20 2006 +0300 172000 172001 mi: remove mi motion history 172002 This is now unneeded as we do motion history in the DIX. 172003 172004commit b9e180e632d04bf685ade9e32bd0b20882794486 172005Author: Daniel Stone <daniel@fooishbar.org> 172006Date: Wed Oct 25 01:09:19 2006 +0300 172007 172008 port all users to the new DIX motion history API 172009 Port KDrive, Xvfb, and Xnest, as well as the virtual core devices, to the 172010 new motion history API. Make GetPointerEvents also update the history. 172011 172012commit 5b38eb69cdaa154791c7f74e35dbe4d3256b19bd 172013Author: Daniel Stone <daniel@fooishbar.org> 172014Date: Wed Oct 25 01:08:29 2006 +0300 172015 172016 dix: add motion history support 172017 Add motion history support (sort of based on the XFree86 DDX's 172018 implementation) to the DIX. 172019 172020commit 11fb58be77ac163844e494b2b0a260cf28a7ecd1 172021Author: Daniel Stone <daniel@fooishbar.org> 172022Date: Wed Oct 25 01:07:36 2006 +0300 172023 172024 miscellaneous warning fixes 172025 Use the correct type for time, and fix the mi prototype of EnqueueEvent. 172026 172027commit 2a74b8a91da1a98669993078f7fe9081f2d743ce 172028Author: Daniel Stone <daniel@fooishbar.org> 172029Date: Wed Oct 25 01:04:53 2006 +0300 172030 172031 xfree86: re-bump input abi to 1.0, yet again 172032 172033commit 37d1fffe79c35ada056ce9a56292c000014fe48a 172034Author: Daniel Stone <daniel@fooishbar.org> 172035Date: Tue Oct 24 23:06:57 2006 +0300 172036 172037 inputstr: try to keep device structs the same size 172038 Try to make sure DeviceIntRec and friends stay the same size, 172039 regardless of whether or not XKB is enabled. 172040 172041commit 9f9ac01a819ee96fb5be5d7d346c91f461bf3979 172042Author: Daniel Stone <daniel@fooishbar.org> 172043Date: Tue Oct 24 23:01:05 2006 +0300 172044 172045 inputstr: fix indentation 172046 172047commit 27d682fa030501611a5b52f41f29738134b45c0b 172048Author: Ian Romanick <idr@us.ibm.com> 172049Date: Tue Oct 24 09:04:14 2006 -0700 172050 172051 Fix from HEAD accidentally reverted during merge. 172052 172053commit 8b909135664abb69ada522e8e39495d9eca717ac 172054Merge: 21291d6ca 948a97b97 172055Author: Ian Romanick <idr@us.ibm.com> 172056Date: Tue Oct 24 08:57:59 2006 -0700 172057 172058 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 172059 172060 Conflicts: 172061 172062 hw/xfree86/Makefile.am 172063 hw/xfree86/common/xf86.h 172064 hw/xfree86/common/xf86DoScanPci.c 172065 hw/xfree86/os-support/bus/linuxPci.c 172066 hw/xfree86/scanpci/extrapci.ids 172067 hw/xfree86/scanpci/pci.ids 172068 hw/xfree86/scanpci/pciid2c.pl 172069 hw/xfree86/scanpci/xf86PciStdIds.h 172070 172071commit 0cd6a3d8efb5cc1ce4f85ab95bcdf4fb66c7245d 172072Author: Daniel Stone <daniel@fooishbar.org> 172073Date: Mon Oct 23 06:56:07 2006 +0300 172074 172075 xfree86/input: add proximity support, free valuators 172076 Re-add support for proximity events, and remember to both va_end our 172077 varargs, and free our valuators. 172078 172079commit cccf7ae0ff24d0c84b5144c457f3f86bbbc36e12 172080Author: Daniel Stone <daniel@fooishbar.org> 172081Date: Mon Oct 23 06:55:21 2006 +0300 172082 172083 GetProximityEvents: add (untested) function 172084 Add untested first guess at what GetProximityEvents should look like. 172085 172086commit 35fa4ac12b3da33f81e2a12bc9661ed075f323ed 172087Author: Daniel Stone <daniel@fooishbar.org> 172088Date: Mon Oct 23 06:23:45 2006 +0300 172089 172090 GetPointerEvents: break into separate functions 172091 Break out clipAxis, clipValuators, and getValuatorEvents, into 172092 separate functions, to be used by the proximity event code. 172093 172094commit a7c2d9a15dc2ff253bb69c3b0738ad993521b9c7 172095Author: Daniel Stone <daniel@fooishbar.org> 172096Date: Mon Oct 23 06:08:27 2006 +0300 172097 172098 kdrive: numerous warning fixes 172099 172100commit c8f76fb3a473a022d497bd0acd6c84f58fc6efbe 172101Author: Daniel Stone <daniel@fooishbar.org> 172102Date: Mon Oct 23 05:12:15 2006 +0300 172103 172104 xfree86/input: disallow pointer device changes, fix drag event calculation 172105 Don't allow users to change the core pointer. 172106 Fix xf86SendDragEvents to check the device button state, not the core 172107 pointer's. 172108 Remove unused xf86CheckButton. 172109 172110commit c5b5b046e86b159a9d32451308f38246cc4587f7 172111Merge: fab1d37ec 948a97b97 172112Author: Daniel Stone <daniel@fooishbar.org> 172113Date: Mon Oct 23 02:58:30 2006 +0300 172114 172115 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 172116 172117commit fab1d37ecbeee9777ad1485b0cc62b55042d5014 172118Author: Daniel Stone <daniel@fooishbar.org> 172119Date: Mon Oct 23 02:57:00 2006 +0300 172120 172121 xfree86/options: don't blindly dereference NULL pointers 172122 Sure, conf_idev should be initialised, but if it's not, try to make the 172123 best of a bad situation. 172124 172125commit 523fe64ba16cf7a40f4095432d81300726d83e8d 172126Author: Daniel Stone <daniel@fooishbar.org> 172127Date: Mon Oct 23 02:56:28 2006 +0300 172128 172129 kdrive/lots of DDXes: make stubs use KdOsAddInputDrivers() 172130 In theory, removing LinuxPciScan should make these drivers OS-independent. 172131 172132commit 57f98e2b72d5c7fea93a9f6da86228ffb4856bf3 172133Author: Daniel Stone <daniel@fooishbar.org> 172134Date: Mon Oct 23 02:55:41 2006 +0300 172135 172136 kdrive/tslib: port to new absolute API, remove debugging 172137 Port to new absolute device (not touchscreen) API, and remove some 172138 leftover debugging. 172139 172140commit a6dc629052f0ad509cfa30e98217043a63c09552 172141Author: Daniel Stone <daniel@fooishbar.org> 172142Date: Mon Oct 23 02:54:59 2006 +0300 172143 172144 kdrive/os: add KdOsAddInputDrivers 172145 Add KdOsAddInputDrivers, which adds all relevant input drivers. 172146 172147 Could possibly be refactored to KdAddInputDrivers, which called through 172148 OsFuncs to a new function, if it existed. 172149 172150commit ac240b00ff6fc354c1aa641406ae8b49244c0b8b 172151Author: Daniel Stone <daniel@fooishbar.org> 172152Date: Mon Oct 23 02:54:07 2006 +0300 172153 172154 kdrive/input: use Absolute instead of Touchscreen 172155 Adjust to suit Zepheniah's new absolute device API. 172156 172157commit b1ba4b3e8ad427af1fa8618b0bd839f741ec2ce3 172158Author: Daniel Stone <daniel@fooishbar.org> 172159Date: Mon Oct 23 02:53:15 2006 +0300 172160 172161 kdrive/neomagic: include vesa.h, use DebugF 172162 172163commit 29f28dd1a8a7ed07e945a67946e3510f4b32d12a 172164Author: Daniel Stone <daniel@fooishbar.org> 172165Date: Mon Oct 23 02:52:52 2006 +0300 172166 172167 configure.ac: add XSERVER_LIBS to XNEST_LIBS 172168 172169commit 2f33f4065d89ae2b6fdda43c7105d72f89920cae 172170Author: Daniel Stone <daniel@fooishbar.org> 172171Date: Mon Oct 23 02:52:35 2006 +0300 172172 172173 Xnest: port to new input API 172174 Port Xnest to Get{Pointer,Keyboard}Events, plus the new mieq API. 172175 172176commit 08928afb0500d46b0caa0a1d1244dee2ed80e6a0 172177Author: Daniel Stone <daniel@fooishbar.org> 172178Date: Mon Oct 23 02:51:52 2006 +0300 172179 172180 Xnest: disable XKB, reshuffle code 172181 Disable XKB, as we can't yet use it; move Composite disabling to 172182 ddxInitGlobals, along with XKB. 172183 172184commit cd3b16a57efaf89108054f18a94c91e2dd74fafa 172185Author: Daniel Stone <daniel@fooishbar.org> 172186Date: Mon Oct 23 02:51:13 2006 +0300 172187 172188 Xvfb: port to new mieq API 172189 172190commit bf4df9b73f0c1a84093aaf9a2e2cbc56fb341c60 172191Author: Daniel Stone <daniel@fooishbar.org> 172192Date: Mon Oct 23 02:50:53 2006 +0300 172193 172194 include: move POINTER_* flags from inputstr.h to input.h 172195 Given they're just numeric constants, they should be included in 172196 input.h, not inputstr.h. 172197 172198commit 4dd91c45abea9fb561a5acb10290e29487df6722 172199Author: Daniel Stone <daniel@fooishbar.org> 172200Date: Mon Oct 23 02:50:03 2006 +0300 172201 172202 miinitext: Xi and XKB are not hardware-only extensions 172203 Xi is now a required extension, and XKB can be used without hardware, 172204 so include them both when NO_HW_ONLY_EXTS is defined. 172205 172206commit 562096a012f4bb8f44d5ec6320a32f4010c189e4 172207Author: Daniel Stone <daniel@fooishbar.org> 172208Date: Mon Oct 23 02:49:22 2006 +0300 172209 172210 XkbCopyKeymap: increment shapes and outlines when copying 172211 Remember to increment the source and destination shapes when copying, 172212 instead of just endlessly copying the first one. 172213 172214commit eec182259112fba240751f974f7e5ca09fce8b9d 172215Author: Daniel Stone <daniel@fooishbar.org> 172216Date: Mon Oct 23 02:48:30 2006 +0300 172217 172218 dix/getevents: move SyntheticMotion to getevents.c 172219 Mostly, this is just a cheesy hack to ensure that getevents.o gets 172220 included when linking. Sigh. 172221 172222commit bc701a14292da5abfb601e3a040651a74f46df8f 172223Author: Daniel Stone <daniel@fooishbar.org> 172224Date: Mon Oct 23 00:08:32 2006 +0300 172225 172226 dix/getevents: cosmetic cleanups, remove keymap copy from GKVE 172227 Remove keymap copy from GetKeyboardValuatorEvents, as 172228 SwitchCoreKeyboard now takes care of this for us. 172229 Remove unused variable and function prototype. 172230 Update comments to be as informative as possible. 172231 172232commit b03e2f7029506640a8fe5cb88818b329c23503ff 172233Author: Daniel Stone <daniel@fooishbar.org> 172234Date: Sun Oct 22 19:56:49 2006 +0300 172235 172236 xi: fix NIDR return yet again 172237 For a one-line function, it was pretty broken. 172238 172239commit 90de7ce25a84cfe6c6790f9af2bc2399d25b9b9c 172240Author: Daniel Stone <daniel@fooishbar.org> 172241Date: Sun Oct 22 19:54:36 2006 +0300 172242 172243 xi: fix return type for NIDR 172244 172245commit f46dc272913ffb6b5b234a7ec6f4ba5cae44a831 172246Author: Daniel Stone <daniel@fooishbar.org> 172247Date: Sun Oct 22 19:51:35 2006 +0300 172248 172249 xi: add NewInputDeviceRequest to stubs 172250 172251commit 31a6307b7ba5adaa96deb8101ddfcda0262f537d 172252Author: Daniel Stone <daniel@fooishbar.org> 172253Date: Sun Oct 22 19:49:31 2006 +0300 172254 172255 xi: change DEVICE_TOUCHSCREEN to ABS_{AREA_CALIB} for stubs 172256 172257commit eae6594d03a606ddf1f433b5897b5938aa940c1e 172258Author: Daniel Stone <daniel@fooishbar.org> 172259Date: Sun Oct 22 16:39:44 2006 +0300 172260 172261 Xi: swap control in DevicePresenceNotify 172262 172263commit be21630164e865eca72ff2a686a38ae4e30fd79c 172264Author: Daniel Stone <daniel@fooishbar.org> 172265Date: Sun Oct 22 16:33:02 2006 +0300 172266 172267 dix, Xi: make use of deviceid in DevicePresenceNotify 172268 Use the deviceid and control fields in DevicePresenceNotify since 172269 the last push to inputproto to send a DPN whenever a control changes 172270 on a device. 172271 172272commit f08b6b2367705cb5b60e996e6328197430bf1919 172273Author: Daniel Stone <daniel@fooishbar.org> 172274Date: Sun Oct 22 12:30:02 2006 +0300 172275 172276 kdrive: change DEVICE_TOUCHSCREEN to DEVICE_ABS_{CALIB,AREA} 172277 172278commit 77e724585f6c53feb55475b94d8cfcb6acf1159b 172279Author: Daniel Stone <daniel@fooishbar.org> 172280Date: Fri Oct 20 00:44:46 2006 +0300 172281 172282 minor formatting fixes 172283 172284commit 948a97b97e93cee05a23f1f1ae699c5e181bc8ce 172285Author: Drew Parsons <dparsons@debian.org> 172286Date: Sat Oct 21 23:09:22 2006 +1000 172287 172288 Minor typo fix to xorg.conf man page. 172289 172290commit 04554a3adcddc32de5fdb0b3122da0bcdd4c24a9 172291Author: Drew Parsons <dparsons@debian.org> 172292Date: Sat Oct 21 22:06:43 2006 +1000 172293 172294 Minor typos in Xserver man page. 172295 172296commit e26a494f417c3c700636ee68892c3015b2e0f27a 172297Merge: 736b0d507 aeba855b0 172298Author: Zephaniah E. Hull <warp@agamemnon.b5> 172299Date: Sat Oct 21 04:26:14 2006 -0400 172300 172301 Merge branch 'input-hotplug' of git+ssh://git.freedesktop.org/git/xorg/xserver 172302 172303commit 736b0d5078597abbda80444deef852879260af90 172304Author: Zephaniah E. Hull <warp@agamemnon.b5> 172305Date: Sat Oct 21 04:24:49 2006 -0400 172306 172307 DEVICE_TOUCHSCREEN becomes DEVICE_ABS_CALIB. 172308 172309 Update the DEVICE_ABS_CALIB stuff to include the new elements. 172310 172311 New DEVICE_ABS_AREA support. 172312 172313 dev->touchscreen becomes dev->absolute, with _CALIB and _AREA stuff in it. 172314 172315 Update xfree86 to compile with this, kdrive needs an update too. 172316 172317commit edd5f1745461f995670969cb736d1569ca94643f 172318Author: Carl Switzky <Carl.Switzky@Sun.COM> 172319Date: Thu Oct 19 17:30:54 2006 -0700 172320 172321 Add ast driver/device info to Xorg server & config utilities 172322 172323commit aeba855b07832354f59678e20cc29a085e42bd99 172324Author: Daniel Stone <daniel@fooishbar.org> 172325Date: Fri Oct 20 00:30:28 2006 +0300 172326 172327 move keymap copy to event processing, from enqueuing 172328 172329 Move the keymap copying to event processing time (in 172330 ProcessInputEvents), instead of being at event enqueuing time. 172331 Break SetCore{Pointer,Keyboard} out into separate functions. 172332 Change mieqEnqueue to take a device pointer, that asks for the 172333 _original_ device associated with this event. 172334 172335commit a8d3dad9d9f2b9053843e655abe463a68ba8dcb7 172336Author: Daniel Stone <daniel@fooishbar.org> 172337Date: Fri Oct 20 00:28:40 2006 +0300 172338 172339 xi: add DEVICE_ENABLE control 172340 172341 Add DEVICE_ENABLE control, which allows runtime enabling and disabling 172342 of specific devices. 172343 172344commit b0780312d80ea4af0136227f90fdd7ada3db71c5 172345Author: Alan Coopersmith <alan.coopersmith@sun.com> 172346Date: Thu Oct 19 13:51:53 2006 -0700 172347 172348 Pre-release message should tell users to check git, not CVS, for updates 172349 172350commit c5dc997baf57ffa08025efadbbaf761296ce4bc4 172351Author: Joshua Baergen <joshuabaergen@gentoo.org> 172352Date: Thu Oct 19 11:14:26 2006 -0700 172353 172354 Create xorg.conf.example (Gentoo bug #138623). 172355 172356commit d029c8f1b72019446a5c873f55ffa43504b03dfb 172357Author: Alan Coopersmith <alan.coopersmith@sun.com> 172358Date: Wed Oct 18 18:11:06 2006 -0700 172359 172360 Use getisax() instead of asm code to determine available x86 ISA extensions on Solaris 172361 172362commit a8a0abdbea0573c861a5af9d58f3ce66790455ca 172363Author: Daniel Stone <daniel@fooishbar.org> 172364Date: Wed Oct 18 10:59:07 2006 +0300 172365 172366 config/dbus: always unref the connection, not close 172367 172368commit 80642f37d40216035786eaf490952d16f6b5f597 172369Author: Adam Jackson <ajax@benzedrine.nwnk.net> 172370Date: Tue Oct 17 14:53:28 2006 -0400 172371 172372 Stop building xorgcfg by default. 172373 172374commit 5e17cde27b064174584d478130b0f95dcef78deb 172375Author: Matthias Hopf <mhopf@suse.de> 172376Date: Tue Oct 17 17:06:44 2006 +0200 172377 172378 StorePixels() macro could create invalid *x++=*x... code - fixed. 172379 172380commit 205c6788d7a34704e36b23f1a93d89e9b986266a 172381Author: Daniel Stone <daniel@fooishbar.org> 172382Date: Mon Oct 16 23:48:09 2006 +0300 172383 172384 config/dbus: properly initialise vtable 172385 172386 Properly initialise the vtable, so we don't end up with an unregister_function 172387 pointing to god knows where. 172388 172389commit 85ac2f16abe9f6e88b4e71609da334d336a9a600 172390Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 172391Date: Mon Oct 16 12:39:05 2006 +0100 172392 172393 Small modification to blocking signals when switching modes. 172394 172395commit 0901eec87ee9f3a2a067695bdbd569ff42149879 172396Author: Michel Dänzer <michel@tungstengraphics.com> 172397Date: Sun Oct 15 16:57:09 2006 +0200 172398 172399 Fix __glXDRIbindTexImage() for 32 bpp on big endian platforms. 172400 172401commit a232693c8c2a206aac47c07b133c071938204e0b 172402Author: Michel Dänzer <michel@tungstengraphics.com> 172403Date: Sun Oct 15 16:48:59 2006 +0200 172404 172405 Add per-drawable Xv colour key helper function. 172406 172407 This allows overlay Xv adaptors to work slightly better with compositing 172408 managers. 172409 172410 Bump the video driver ABI minor so drivers only need to check for this at build 172411 time. 172412 172413commit 5563861ab7e56ec891cfce6b34af43fec53ccee3 172414Author: Alan Coopersmith <alan.coopersmith@sun.com> 172415Date: Fri Oct 13 19:05:28 2006 -0700 172416 172417 Make sure xorgcfg files are included even when dist made with --disable-xorgcfg 172418 172419commit e1dd1904c6c7ce33f347d822272831d54a6497c8 172420Author: Alan Coopersmith <alan.coopersmith@sun.com> 172421Date: Thu Oct 19 13:51:53 2006 -0700 172422 172423 Pre-release message should tell users to check git, not CVS, for updates 172424 172425commit 62d24097932708fbbb62a23614fe63b4b7acf3bd 172426Author: Joshua Baergen <joshuabaergen@gentoo.org> 172427Date: Thu Oct 19 11:14:26 2006 -0700 172428 172429 Create xorg.conf.example (Gentoo bug #138623). 172430 172431commit 357b37b3826fa6e9878c0bd895164259c2ed3c0d 172432Author: Alan Coopersmith <alan.coopersmith@sun.com> 172433Date: Wed Oct 18 18:11:06 2006 -0700 172434 172435 Use getisax() instead of asm code to determine available x86 ISA extensions on Solaris 172436 172437commit 1b1698af41b9038d9f9dbf521737d0baab5a2237 172438Author: Zephaniah E. Hull <warp@agamemnon.b5> 172439Date: Wed Oct 18 04:57:22 2006 -0400 172440 172441 Pass SetDeviceValuators down to the driver. 172442 NOTE: This changes the LocalDeviceRec struct, which breaks input drivers. 172443 172444commit 5eca750fe2f3f243fb352271ad8da196af0cb16a 172445Author: Adam Jackson <ajax@benzedrine.nwnk.net> 172446Date: Tue Oct 17 14:53:28 2006 -0400 172447 172448 Stop building xorgcfg by default. 172449 172450commit df979b75c8cd8a7e0566aea58031bb9b8f5cd3d3 172451Author: Matthias Hopf <mhopf@suse.de> 172452Date: Tue Oct 17 17:06:44 2006 +0200 172453 172454 StorePixels() macro could create invalid *x++=*x... code - fixed. 172455 172456commit d430e76a161c963169067875c3654f5fd8f42b19 172457Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 172458Date: Mon Oct 16 12:39:05 2006 +0100 172459 172460 Small modification to blocking signals when switching modes. 172461 172462commit 6dd4fc4652f942724039dc2317c560ea7276ab59 172463Author: Daniel Stone <daniel@fooishbar.org> 172464Date: Mon Oct 16 00:22:00 2006 +0300 172465 172466 xkb: fix virtual modmap size computation 172467 Compute virtual modmap size bounded by nVModMapKeys-1, rather than 172468 nVModMapKeys. 172469 172470 This is sort of a best guess. The other way seems a little more 172471 logical, but also leads to segfaults pretty quickly if you hammer 172472 GetMap hard enough. So let's try this one. 172473 172474commit a484ba15277e66e7ef9b21b238dcbf760695bc63 172475Author: Daniel Stone <daniel@fooishbar.org> 172476Date: Sun Oct 15 23:47:34 2006 +0300 172477 172478 XkbCopyKeymap: copy server vmods, and name atoms 172479 Copy server->vmods, and all the atoms in names. 172480 172481commit ad355fecee3965be576596aeed5da54d776edf1d 172482Author: Daniel Stone <daniel@fooishbar.org> 172483Date: Sun Oct 15 21:59:06 2006 +0300 172484 172485 xkb: make sure we set the map on the right device, not necessarily core 172486 Forgot that all XKB requests took a device spec: the comparison of 172487 'if working on the core keyboard, does this device send core events; or, 172488 is this device the core keyboard?' was broken. Instead, what we want is 172489 'if working on the core keyboard, does this device send core events; or, 172490 is this device the one we're working on?'. 172491 172492commit 4ae7745a0dc86de6346409a69c1e396e0b954514 172493Author: Daniel Stone <daniel@fooishbar.org> 172494Date: Sun Oct 15 21:48:01 2006 +0300 172495 172496 xfree86 input: always open devices on NewInputDeviceRequest 172497 172498commit acd8419948003032056a56d46adbef7c35e7739c 172499Author: Daniel Stone <daniel@fooishbar.org> 172500Date: Sun Oct 15 20:42:31 2006 +0300 172501 172502 config: unref connection, don't close it 172503 Just unref the connection instead of explicitly closing it (thanks, Rob 172504 McQueen). 172505 Add a commented-out unregister_object_path call: unfortunately, when we 172506 call it, libdbus segfaults. But if we don't unregister the path, we 172507 can't register it again. So regenerations are broken either way, but a 172508 little less violently like this. 172509 172510commit fc9b5f84b244ea08480b73bd15ac919b875800fb 172511Author: Daniel Stone <daniel@fooishbar.org> 172512Date: Sun Oct 15 20:01:01 2006 +0300 172513 172514 dix/devices: add devices in proper forward order 172515 Add devices in forward order with the normal linked list convention. 172516 Previously, AddInputDevice would add all the devices in reverse order to 172517 off_devices, before they were added again in reverse order to devices with 172518 EnableDevice. 172519 This just makes both work in forward order, which provides the ordering as 172520 you'd expect when hotplugging devices (i.e. adds them to the head, not the 172521 tail). 172522 172523commit ec35e7198debf938f1115f584e675ce5995743e3 172524Author: Daniel Stone <daniel@fooishbar.org> 172525Date: Sun Oct 15 19:44:49 2006 +0300 172526 172527 config: add replies and dbus api documentation 172528 Add replies, which use standard X error values, to the two currently-supported 172529 input configuration requests. 172530 Document the D-BUS API we use. 172531 Make sure we free everything when we encounter an error. 172532 Add a _source option to all incoming requests, noting that it came from a 172533 client. 172534 Reject all requests to add a device where an option name contains an 172535 underscore. 172536 172537commit 7e4717683d6c08d1e490a60b7493a94bbc57bf8d 172538Author: Michel Dänzer <michel@tungstengraphics.com> 172539Date: Sun Oct 15 18:12:28 2006 +0200 172540 172541 exaDrawableDirty: Fix initialization of BoxRec. 172542 172543 This will hopefully fix the partial window corruption experienced by some 172544 people. 172545 172546commit 3ad1642f1bbaa5f96558cdf3384b40f7122f8781 172547Author: Michel Dänzer <michel@tungstengraphics.com> 172548Date: Sun Oct 15 16:57:09 2006 +0200 172549 172550 Fix __glXDRIbindTexImage() for 32 bpp on big endian platforms. 172551 172552commit f9bfee50981006a2c58d3f73e2b0d123bb2a41b7 172553Author: Michel Dänzer <michel@tungstengraphics.com> 172554Date: Sun Oct 15 16:48:59 2006 +0200 172555 172556 Add per-drawable Xv colour key helper function. 172557 172558 This allows overlay Xv adaptors to work slightly better with compositing 172559 managers. 172560 172561 Bump the video driver ABI minor so drivers only need to check for this at build 172562 time. 172563 172564commit a05044cfc14a8bc6cc31236dcecada60bec09924 172565Author: Daniel Stone <daniel@fooishbar.org> 172566Date: Sat Oct 14 22:14:56 2006 +0300 172567 172568 xkb: better support of XkbDfltXIId 172569 XKB.h specifies that XkbDfltXIId should be used where the client doesn't 172570 care about the device identifier. We take this to mean core devices, 172571 where practical. 172572 172573commit 4d8030076ed1a7680bdfcb7b89af1045bdc40304 172574Author: Daniel Stone <daniel@fooishbar.org> 172575Date: Sat Oct 14 22:14:07 2006 +0300 172576 172577 dix: move GetKeyboardEvents/GetPointerEvents to a new file, export symbols 172578 Move GKE and GPE to a separate file, to help stem the events.c explosion. 172579 Mark GKE/GKVE/GPE as _X_EXPORT. 172580 172581commit 6afc7c284690b1e2bb7544b5bc4f31a3f6a05519 172582Author: Daniel Stone <daniel@fooishbar.org> 172583Date: Sat Oct 14 15:54:35 2006 +0300 172584 172585 dix/devices: remove XACE merge debris 172586 172587commit 93302452e737bd91a893eb495592538d40d921e5 172588Author: Daniel Stone <daniel@fooishbar.org> 172589Date: Sat Oct 14 15:54:12 2006 +0300 172590 172591 XkbCopyKeymap: add geometry support 172592 Add a first cut at geometry support, which seems to generally work. 172593 172594commit b9108a13fc126d97c0393f911a1d9292563444ce 172595Author: Alan Coopersmith <alan.coopersmith@sun.com> 172596Date: Fri Oct 13 19:05:28 2006 -0700 172597 172598 Make sure xorgcfg files are included even when dist made with --disable-xorgcfg 172599 172600commit 054f8cd2675a80b14bc1ce266377fcfee2335cee 172601Author: Keith Packard <keithp@neko.keithp.com> 172602Date: Fri Oct 13 17:34:53 2006 -0700 172603 172604 Limit pointer to valid crtc areas. Add event swapping. Fix change tracking. 172605 172606 Add function to keep pointer within valid crtc areas. 172607 Finish event delivery and swapping code. 172608 Separate configuration from layout changes to send correct events. 172609 172610commit 335b503c5e7041bb0c44611e496d1c46f554e630 172611Merge: bd3d93be8 cf948b7b0 172612Author: Daniel Stone <daniel@fooishbar.org> 172613Date: Fri Oct 13 18:10:45 2006 +0300 172614 172615 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 172616 172617commit cf948b7b04dfeb61a294889027b9a54f6b9b478e 172618Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 172619Date: Fri Oct 13 14:48:10 2006 +0100 172620 172621 Block IO on switchmode just as we do for VT switching 172622 to avoid cursor movements signalling io. 172623 172624commit 4b53a5e4b147ab748b28dffc1d330b7148d8efa6 172625Merge: 23f6f08b7 042d4a407 172626Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 172627Date: Wed Oct 11 20:50:31 2006 -0400 172628 172629 Merge branch 'master' into my-XACE-SELINUX 172630 172631commit a6df780e4d3f90cc699b0b864bade03d6f15afd7 172632Author: Ian Romanick <idr@us.ibm.com> 172633Date: Wed Oct 11 15:41:22 2006 -0700 172634 172635 Fix bug #8608. 172636 172637 Regenerate files based on recent commits to Mesa (that fix Mesa bug 172638 that return pixel data. 172639 172640commit 042d4a407d0eca9131d0420b0b9189559aac326d 172641Author: Aaron Plattner <aplattner@nvidia.com> 172642Date: Mon Oct 9 16:32:11 2006 -0700 172643 172644 Bug #8459: Don't forget to include AM_CFLAGS in libfb_la_CFLAGS on non-MMX platforms. 172645 172646 Reported by Edgar Toernig. 172647 172648commit 1b94c117e0f294ef2f89bf24d45ba7a8e45efe35 172649Author: Matthias Hopf <mhopf@suse.de> 172650Date: Tue Oct 10 19:37:22 2006 +0200 172651 172652 Fixing probably long-standing bug in domain support. 172653 172654commit 6776c0f0e9677a65ec36ceef9618ff701b99596c 172655Author: Alan Coopersmith <alan.coopersmith@sun.com> 172656Date: Mon Oct 9 17:24:37 2006 -0700 172657 172658 Use bash on Solaris to run symlink-mesa.sh 172659 172660 symlink-mesa.sh won't run with Solaris Bourne Shell (/bin/sh) so explicitly 172661 run it with /usr/bin/bash instead 172662 (cherry picked from cde68728860179dc84e615ccb378ce992513fd62 commit) 172663 172664commit bd3d93be82d91e4cf35ae317dfd658d1706257ea 172665Author: Daniel Stone <daniel@fooishbar.org> 172666Date: Sun Oct 8 23:21:12 2006 +0300 172667 172668 xkb: remove random broken vendor workarounds 172669 172670commit 33406da096b4ae21134484113b280e07d0c8f0d9 172671Author: Daniel Stone <daniel@fooishbar.org> 172672Date: Sun Oct 8 23:20:56 2006 +0300 172673 172674 GetKeyboardEvents: add first_valuator argument to GKVE 172675 172676commit ef7e05e9de57b9c9c27ed3307eede6d8fc6c1af3 172677Author: Daniel Stone <daniel@fooishbar.org> 172678Date: Sun Oct 8 21:42:15 2006 +0300 172679 172680 xfree86/xinput: pass first_valuator params, don't clip to first screen bound 172681 Pass first_valuator to GetPointerEvents/GetKeyboardValuatorEvents. 172682 Don't clip axis bounds to that of screen 0 when initialising axes. 172683 172684commit 1546a398144619a14ff06aaf84ebc1bf293eac66 172685Author: Daniel Stone <daniel@fooishbar.org> 172686Date: Sun Oct 8 21:40:53 2006 +0300 172687 172688 GetKeyboardValuatorEvents: also take first_valuator param 172689 Take a first_valuator parameter, which specifies the first valuator. 172690 172691commit b05a11478edc7e6d1e38ef7f8d6788c7bd917493 172692Author: Daniel Stone <daniel@fooishbar.org> 172693Date: Sun Oct 8 21:23:12 2006 +0300 172694 172695 doc/extensions: rename to c-extensions 172696 The old name could be somewhat confusing. 172697 172698commit 41bb9fce47f6366cc3f7d45790f7883f74289b5a 172699Author: Daniel Stone <daniel@fooishbar.org> 172700Date: Sun Oct 8 20:34:32 2006 +0300 172701 172702 mipointer: take device arguments, split miPointerAbsoluteCursor 172703 Update mipointer API to take a device argument to (almost) all functions, 172704 and split miPointerAbsoluteCursor into a couple of separate functions. 172705 Remove miPointerAbsoluteCursor call from mieq, as we now deal with it in 172706 GetPointerEvents. 172707 Make miPointerSetPosition (successor of miPointerAbsoluteCursor) take 172708 pointers to x and y, so it can return the clipped values. 172709 Modify callers of miPointer*() functions to generally use the new 172710 functions. 172711 This should fix things with multi-head setups. 172712 172713commit be8dfafd1d58b27bbfd953fc1216311523353db1 172714Author: Daniel Stone <daniel@fooishbar.org> 172715Date: Sun Oct 8 20:32:16 2006 +0300 172716 172717 warning cleanups 172718 Fix still more warnings. 172719 172720commit ca474e0920dd29ebe7ccf346cddc526732ad01ba 172721Author: Daniel Stone <daniel@fooishbar.org> 172722Date: Sun Oct 8 20:30:49 2006 +0300 172723 172724 Xi: move SendEventToAllWindows and FindInterestedChildren to exevents 172725 Move SendEventToAllWindows and FindInterestedChildren from chgptr to exevents, 172726 so the DIX can more easily use it. 172727 Clean up two warnings (type mismatch, unused variable) in exevents.c. 172728 172729commit c2fab469b66f2796c541e911202faa411d116b04 172730Author: Daniel Stone <daniel@fooishbar.org> 172731Date: Sun Oct 8 18:26:26 2006 +0300 172732 172733 dix/devices: clean up debugging 172734 172735commit 4493acb88c59721f7807093a3ed3c39396c2076d 172736Author: Daniel Stone <daniel@fooishbar.org> 172737Date: Sun Oct 8 17:51:03 2006 +0300 172738 172739 xkb: add FIXMEs to procedures which need to act on all core devices 172740 Add FIXME comments above request handlers which need to act on all core-sending 172741 devices if called on the core keyboard. 172742 172743commit ef68273f5bdb27a492ec0b69548ec4fbede46c08 172744Author: Daniel Stone <daniel@fooishbar.org> 172745Date: Sun Oct 8 17:44:37 2006 +0300 172746 172747 mi/mipointer: deprecate functions which don't take a device 172748 Deprecate all mi pointer functions which don't take a device argument, and 172749 replace them with versions which do, in preparation for MPX. 172750 172751commit 6eab4c55890660089067da0e944256b1ed3a8c67 172752Author: Daniel Stone <daniel@fooishbar.org> 172753Date: Sun Oct 8 17:24:33 2006 +0300 172754 172755 doc/extensions: document C extension use in the X server 172756 172757commit 80cdd26581508dd17c5d0a5739cd540113996bbb 172758Author: Daniel Stone <daniel@fooishbar.org> 172759Date: Sun Oct 8 17:23:54 2006 +0300 172760 172761 mi/pointer: mark public pointer functions as deprecated 172762 Deprecate miPointer functions which don't take a device pointer. Pointer 172763 movement should be handled through GetPointerEvents, and functions which 172764 take a device as an argument (e.g. miPointerPosition) will be added. 172765 172766commit 97030b6c6b0fb6ff629ae31e483704d0a2207a53 172767Author: Daniel Stone <daniel@fooishbar.org> 172768Date: Sun Oct 8 17:07:05 2006 +0300 172769 172770 config: fix compilation 172771 Accidentally built with --disable-config, didn't notice that the previous 172772 commit to clean up the debugging broke things horribly. 172773 172774commit 14b157bdb1f2cd5feba03ba0815d7c5b2dd6633f 172775Author: Daniel Stone <daniel@fooishbar.org> 172776Date: Sun Oct 8 17:04:12 2006 +0300 172777 172778 include: actually declare DebugF 172779 DebugF is ErrorF when DEBUG is defined, else a no-op. 172780 172781commit 9e37de193f5d7412ffd8de76d5eed0158c0a3609 172782Author: Daniel Stone <daniel@fooishbar.org> 172783Date: Sun Oct 8 16:32:15 2006 +0300 172784 172785 configure.ac: reactivate warnings when building with gcc 172786 We were inadvertently stomping XSERVER_CFLAGS after adding the warnings, so 172787 move them after we do that. 172788 172789commit b559cbb1601f93cb03ea3dcfb2c5ca94ee6b73bb 172790Author: Daniel Stone <daniel@fooishbar.org> 172791Date: Sun Oct 8 16:23:14 2006 +0300 172792 172793 dix/CoreProcessKeyboardEvent: remove debugging for every key event 172794 Also change #ifdef DEBUG/ErrorF/#endif to DebugF in FixKeyState. 172795 172796commit 3ae4d250185e71a0a218c062426f92b9b1adbf05 172797Author: Daniel Stone <daniel@fooishbar.org> 172798Date: Sun Oct 8 16:20:42 2006 +0300 172799 172800 xfree86 Xinput: remove still more excessive debugging 172801 There isn't any more debugging left for input events in the XFree86 DDX. 172802 172803commit 58314756aeecbb8fb04706c3e04d98e9ac531a02 172804Author: Daniel Stone <daniel@fooishbar.org> 172805Date: Sun Oct 8 16:18:05 2006 +0300 172806 172807 GetPointerEvents: add first_valuator parameter 172808 Add a first_valuator parameter. Looks correct by inspection, but untested 172809 with first_valuator != 0 as yet. 172810 172811commit 84f5d2291c1fe92fd8358e999e909bf3aab86c98 172812Author: Daniel Stone <daniel@fooishbar.org> 172813Date: Sun Oct 8 15:30:24 2006 +0300 172814 172815 GetPointerEvents: fix relatively harmless typo 172816 Change !(cp->button || !cp->valuator) to (!cp->button || !cp->valuator). 172817 172818commit cfc3e9ede2dc83741bd38bf3df13f096ecb8adc0 172819Author: Daniel Stone <daniel@fooishbar.org> 172820Date: Sun Oct 8 15:27:52 2006 +0300 172821 172822 config: remove excessive debugging 172823 172824commit 8d8e7f8bae4099f9e90ef9aac687607dae1d32bf 172825Author: Daniel Stone <daniel@fooishbar.org> 172826Date: Sun Oct 8 15:26:54 2006 +0300 172827 172828 kdrive/input: remove excessive debugging in NIDR 172829 172830commit 22a836fafd39a8ef413826dc2c94bc5f96990e2d 172831Author: Daniel Stone <daniel@fooishbar.org> 172832Date: Sat Oct 7 14:16:51 2006 +0300 172833 172834 xfree86/loader: bump input major ABI version 172835 Bump input major ABI version to 1.0, since we removed the OS keyboard 172836 layer. 172837 172838commit 4c342246300e06bdf5c9c62cc1d2f6aa57a524db 172839Merge: 8382234a7 49a70c857 172840Author: Alan Coopersmith <alan.coopersmith@sun.com> 172841Date: Fri Oct 6 18:01:13 2006 -0700 172842 172843 Merge branch 'XACE-modular' 172844 172845commit 3686cd0fbf56d883f2f3b3fda11ffba1058b74e4 172846Author: Daniel Stone <daniel@fooishbar.org> 172847Date: Fri Oct 6 17:20:42 2006 +0300 172848 172849 xkb: make XkbSetControls work on all core-sending devices 172850 172851commit 7b4dc171b036107cfba87a1a16bf692b982005a5 172852Author: Daniel Stone <daniel@fooishbar.org> 172853Date: Fri Oct 6 16:26:54 2006 +0300 172854 172855 xkb: remove unused #ifndef 172856 172857commit ab56f0c5b516269bb99ae8b5f479e49e61a3af76 172858Author: Daniel Stone <daniel@fooishbar.org> 172859Date: Fri Oct 6 16:12:36 2006 +0300 172860 172861 xkb: simplify core device loop in GetKeyboardByName 172862 172863commit 4b6e2f12f7296e17b2850f36b3adcf8156125cbe 172864Author: Daniel Stone <daniel@fooishbar.org> 172865Date: Fri Oct 6 16:08:21 2006 +0300 172866 172867 xkb: make LatchLockGroup work on all core-sending devices 172868 Apply the settings to all devices sending core events, if we're working on the 172869 core keyboard. 172870 172871commit ebf9b3bbbb04acb78cdf8a84e47a96755fbfe854 172872Author: Daniel Stone <daniel@fooishbar.org> 172873Date: Fri Oct 6 14:17:59 2006 +0300 172874 172875 xkb: update all core-sending keyboards on GetKeyboardByName 172876 Update the keymaps of all keyboards which send core events on 172877 GetKeyboardByName; still a few other procedures which need this treatment. 172878 172879commit 1178796a4dff5ebf0bd9fb3cacb35be9709b41e5 172880Author: Keith Packard <keithp@neko.keithp.com> 172881Date: Thu Oct 5 22:31:35 2006 -0700 172882 172883 Add preferred modes for each output. Round vrefresh. Deliver crtc events. 172884 172885commit 23f6f08b7b5c9a4297fd223d232a7e9f45376550 172886Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 172887Date: Thu Oct 5 16:07:26 2006 -0400 172888 172889 Improve error handling, messages during initialization. 172890 172891commit de63a469dcd0a8ae98554bca540ac0106cccf2a5 172892Merge: 9c7440bdf 8382234a7 172893Author: Daniel Stone <daniel@fooishbar.org> 172894Date: Thu Oct 5 20:29:19 2006 +0300 172895 172896 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 172897 172898commit b4659faf9b455b44ac8e691cc7a8fc00a967c80b 172899Merge: c4f30c635 8382234a7 172900Author: Keith Packard <keithp@neko.keithp.com> 172901Date: Wed Oct 4 14:46:04 2006 -0700 172902 172903 Merge branch 'master' into randr-1.2 172904 172905commit df351f1efbcc95f94c719fcf993c480155c511e9 172906Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 172907Date: Wed Oct 4 16:23:35 2006 -0400 172908 172909 Experimental window property holding security context. 172910 172911commit c4f30c63538e1451f15ed1991439869127d9b148 172912Author: Keith Packard <keithp@guitar.keithp.com> 172913Date: Tue Oct 3 21:06:11 2006 -0700 172914 172915 Add mode origins and output options. Fix memmoves in resource free funcs. 172916 172917 Output options and mode origins both affected driver ABI. memmove mistakes 172918 were causing 'Freeing resource which isn't there' messages. 172919 172920 Prune unused non-user defined modes from available list now. 172921 172922commit 8382234a7f318057de66490299d63807cefb5201 172923Author: Alan Coopersmith <alan.coopersmith@sun.com> 172924Date: Tue Oct 3 17:49:48 2006 -0700 172925 172926 Update pci.ids to Tue 2006-10-03 daily snapshot 172927 172928commit 96edf7b853c1045d12d957a2957a11879100a2de 172929Author: Alan Coopersmith <alan.coopersmith@sun.com> 172930Date: Tue Oct 3 17:48:50 2006 -0700 172931 172932 Don't insert RCS tag into generated pci id header files 172933 172934commit 75e0cc41b201b9ceb2615e7cec0dd5b4c136b343 172935Merge: 6926776f0 20c0ceedf 172936Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 172937Date: Tue Oct 3 13:58:03 2006 -0400 172938 172939 Merge branch 'my-XACE-modular' into my-XACE-SELINUX 172940 172941commit df800d87e04ce984a8a9ab4252ac6478ab1e4426 172942Author: Alan Coopersmith <alan.coopersmith@sun.com> 172943Date: Mon Oct 2 11:28:47 2006 -0700 172944 172945 Add (void) casts to clear compiler errors about ?: results having type mismatch 172946 172947commit 9c7440bdf5a4ecd113e102004c804a2ba354c422 172948Author: Daniel Stone <daniel@fooishbar.org> 172949Date: Mon Oct 2 20:58:33 2006 +0300 172950 172951 xkb: remove the world's most staggeringly broken vendor workaround 172952 Certain versions of LynxOS needed to sleep up to five seconds for closing a 172953 pipe to actually, y'know, be useful. 172954 172955commit d7c89c7c1c8c1e110345d9d8d300adbf5fe5804a 172956Author: Daniel Stone <daniel@fooishbar.org> 172957Date: Mon Oct 2 02:15:36 2006 +0300 172958 172959 symlink-mesa.sh: expand *.{c,h} 172960 172961commit d6ea96b13e2ea01c51998c41ae2a3677bdedf61c 172962Author: Ivan Pascal <pascal@info.tsu.ru> 172963Date: Mon Oct 2 02:17:14 2006 +0300 172964 172965 xkb: fix wrapping when switching between groups 172966 Use XkbCharToInt as that's what we're doing. 172967 172968commit 3c98cebb6e954855528794fec46830f456cbdec1 172969Merge: fa1ac9417 2cf109843 172970Author: Daniel Stone <daniel@fooishbar.org> 172971Date: Mon Oct 2 02:18:17 2006 +0300 172972 172973 Merge branch 'input-hotplug' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 172974 172975commit fa1ac94178cf976d4c8dae9a4dc8703303a62d4b 172976Author: Ivan Pascal <pascal@info.tsu.ru> 172977Date: Mon Oct 2 02:17:14 2006 +0300 172978 172979 xkb: fix wrapping when switching between groups 172980 Use XkbCharToInt as that's what we're doing. 172981 172982commit c31672e6aab168262bd7824a8082ecdf841fc3c0 172983Author: Daniel Stone <daniel@fooishbar.org> 172984Date: Mon Oct 2 02:15:36 2006 +0300 172985 172986 symlink-mesa.sh: expand *.{c,h} 172987 172988commit 2cf1098436d6b4382d9ed3f6b88214d37bdd8ddb 172989Author: Daniel Stone <daniel@fooishbar.org> 172990Date: Sat Sep 30 17:05:46 2006 +0300 172991 172992 dix/events, mi/eq: remove utterly ridiculous debugging 172993 Remove debugging which can cause long-lived Xorg logs to grow well above 1GB 172994 if built with --enable-debug. 172995 172996commit 20c0ceedfbce9ae5c70703900b52973917ac4cd0 172997Merge: 49a70c857 84eb2c0a0 172998Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 172999Date: Fri Sep 29 15:37:00 2006 -0400 173000 173001 Merge branch 'master' into my-XACE-modular 173002 173003commit 518db35ca3f569e7cb95dbddeddb93f3691de498 173004Merge: 5d99e05f0 84eb2c0a0 173005Author: Daniel Stone <daniel@fooishbar.org> 173006Date: Fri Sep 29 00:35:21 2006 +0300 173007 173008 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 173009 173010commit 5d99e05f05a42a82a9f02844df9bfebaa673759d 173011Merge: ad631afcf 589337578 173012Author: Daniel Stone <daniel@fooishbar.org> 173013Date: Fri Sep 29 00:35:07 2006 +0300 173014 173015 Merge branch 'input-hotplug' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 173016 173017commit ad631afcf3fbc74024cddaaacd05d38addc047a2 173018Author: Daniel Stone <daniel@fooishbar.org> 173019Date: Fri Sep 29 00:34:23 2006 +0300 173020 173021 make core keyboard/pointer map functions act on all core-sending devices 173022 Make Set{Keyboard,Modifier,Pointer}Mapping act on all devices which send core 173023 events. 173024 Change LegalModifier to accept a DeviceIntPtr. 173025 173026commit 84eb2c0a06de60e88e14bb03fabe661d7cd8f1d3 173027Author: Brian <brian@yutani.localnet.net> 173028Date: Thu Sep 28 15:09:40 2006 -0600 173029 173030 Replace hard-coded filesnames with loops (all .c and .h files). 173031 Should fix problems with Mesa adding/removing source files, for the most part. 173032 Patch by Dan Nicholson. 173033 173034commit 4bc5dc2854e33bf343cdea44a3c3b4c41f6f4145 173035Merge: cf6e9687f f9542e749 173036Author: Aaron Plattner <aplattner@nvidia.com> 173037Date: Thu Sep 28 13:27:13 2006 -0700 173038 173039 Merge branch 'wfb' 173040 173041commit f9542e749544c7a3084fd72ecc6642ca3262f7c7 173042Author: Aaron Plattner <aplattner@nvidia.com> 173043Date: Thu Sep 28 13:26:54 2006 -0700 173044 173045 Restore the global xx* symbols and add them to wfbrename.h. 173046 173047 I don't think they're ever used, but it's best to just leave them alone for ABI 173048 compatibility. 173049 173050commit cf6e9687ffcc52af0d64e9098186570719a575a2 173051Author: Jeff Smith <whydoubt@yahoo.com> 173052Date: Thu Sep 28 15:34:17 2006 -0400 173053 173054 Bug #8449: Yet another Mesa symlink script resync. 173055 173056commit a5d639cd87f30f9b3d765352d27253940f33b2b7 173057Author: Daniel Stone <daniel@fooishbar.org> 173058Date: Wed Sep 27 16:15:27 2006 +0300 173059 173060 remove merge detritus 173061 173062commit 2206a92a97901977910a6e39b4174ca805f9f4a7 173063Author: Michel Dänzer <michel@tungstengraphics.com> 173064Date: Wed Aug 30 19:15:55 2006 +0200 173065 173066 Push information about cliprects of DRI windows to the DRM. 173067 173068commit 54d371e7a4ebab79a0f616669e2f601d8370cef3 173069Author: Michel Dänzer <michel@tungstengraphics.com> 173070Date: Wed Aug 30 19:12:17 2006 +0200 173071 173072 Add wrapper for new ioctl to update drawable information in the DRM. 173073 173074commit ad321fad4b9ab3a2c70cfff37ca6c8faaa5cce9c 173075Merge: 5e9d33fe8 f7c1d9424 173076Author: Daniel Stone <daniel@fooishbar.org> 173077Date: Sun Sep 24 17:56:43 2006 +0300 173078 173079 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 173080 173081commit f7c1d942416db8d0d4c5a21f5ece1ccacb926b69 173082Author: Brian <brian@yutani.localnet.net> 173083Date: Sat Sep 23 10:38:10 2006 -0600 173084 173085 Check for visual==NULL in dmxBECreateColormap() before calling XCreateColormap() 173086 to prevent potential segfault. 173087 173088commit 945b7c63946f5257d0f9b0dcf2f8f4882fb2c6f8 173089Author: Brian <brian@yutani.localnet.net> 173090Date: Sat Sep 23 10:35:25 2006 -0600 173091 173092 The fbcmap.c file used by Xdmx _must_ be compiled with XFree86Server defined. 173093 Otherwise, Xdmx generates a slew of protocol errors. 173094 173095commit 891e9c3e6cbd0869a57395b96c8e18ff522c2bb4 173096Author: Brian <brian@yutani.localnet.net> 173097Date: Sat Sep 23 10:28:24 2006 -0600 173098 173099 Replace broken DMXDBG3() with DMXDBG2() 173100 173101commit a10039a100dfe5f87e29e9cc4fa656176e0890f9 173102Author: David Nusinow <dnusinow@debian.org> 173103Date: Thu Sep 21 23:58:32 2006 -0400 173104 173105 Allow the xfree86 ddx utils to be optionally built. Patch by Eugene Konev. 173106 173107commit ce78b0cd2b1c35d60eb5683a1d00222aa4797c79 173108Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173109Date: Thu Sep 21 20:42:47 2006 -0400 173110 173111 Close with Pclose() that which we open with Popen(). 173112 173113commit c1655f0fd457f9bdf0857c5e0904639925bb01f1 173114Author: Aaron Plattner <aplattner@nvidia.com> 173115Date: Thu Sep 21 14:45:17 2006 -0700 173116 173117 Bug 8386: Grow parser buffers to fit an entire line if it's longer than CONFIG_BUF_LEN. 173118 173119commit b36fde9257263fa502147df37e8331184c323e14 173120Author: Keith Packard <keithp@guitar.keithp.com> 173121Date: Thu Sep 21 09:52:04 2006 -0700 173122 173123 When no mode is specified, don't validate mode-specific parameters. 173124 173125commit 8b4ed47c5d39f219866e3c72fa973c6fc4c70f18 173126Author: Drew Parsons <dparsons@debian.org> 173127Date: Thu Sep 21 22:19:44 2006 +1000 173128 173129 * Install Xprint's Xsession script to $(sysconfdir)/X11/Xsession.d 173130 * Removing outdated references to CDE and dt, rename script to 173131 92xprint-xpserverlist. 173132 173133commit 219546fd76750f358ffb6738f17b9237c58c15a6 173134Author: Keith Packard <keithp@guitar.keithp.com> 173135Date: Wed Sep 20 22:43:05 2006 -0700 173136 173137 Steal Xinerama code from SiS driver. Add missing files. 173138 173139 Provide a Xinerama implementation when DIX version isn't enabled. This 173140 version exposes each crtc as a separate 'screen' and reports the size of 173141 that patch. The extension also sends ConfigureNotify events to the root 173142 window whenever crtcs change so that applications will re-fetch xinerama 173143 information. This actually works for metacity. 173144 173145commit bde0a4c12cb393a6d7f1552b067624da1b0502ae 173146Author: Keith Packard <keithp@guitar.keithp.com> 173147Date: Wed Sep 20 19:42:34 2006 -0700 173148 173149 RRSetCrtcConfig status fix. RRGetScreenResources timestamp fix. 173150 173151 RRSetCrtcConfig was returning the wrong status values. 173152 RRGetScreenResources was always returning currentTime. 173153 173154commit d812f486a01a6276aed7b4ebd3cd8eb8ddfe10d3 173155Author: Donnie Berkholz <donnie@supernova.(none)> 173156Date: Wed Sep 20 15:39:39 2006 -0700 173157 173158 Really fix sparc on 64-bit kernel/32-bit userland. 173159 173160 Commit b3a3020fd018df8bc5a8193d36e1a1c7ae8af8ba used a sparc64 ifdef instead of 173161 sparc. But for 32-bit userland, __sparc64__ is not defined so the wrong code is 173162 used. 173163 173164commit 09f7499851bd2f2eba1e30460c61c7a82ed9e853 173165Author: Keith Packard <keithp@guitar.keithp.com> 173166Date: Wed Sep 20 13:15:20 2006 -0700 173167 173168 typo 173169 173170commit 9f870e0aa1ada238d6a0cd099996e8c47f6ba1d9 173171Author: Keith Packard <keithp@guitar.keithp.com> 173172Date: Wed Sep 20 13:14:53 2006 -0700 173173 173174 When setting output state, leave output unchanged when setting to current. 173175 173176commit d16e83413e7e06adebd629d04de57bbedd8c3765 173177Author: Aaron Plattner <aplattner@nvidia.com> 173178Date: Wed Sep 20 12:47:17 2006 -0700 173179 173180 Hide or rename more global symbols to avoid clashes with libfb. 173181 173182 Rename composeFunctions, xxSetup, and xxPrintVisuals. Hide the other xx* 173183 symbols by making them static. 173184 173185commit d08718d8fd31477e90f13b9e122504c515b46ee0 173186Author: Keith Packard <keithp@guitar.keithp.com> 173187Date: Wed Sep 20 12:05:52 2006 -0700 173188 173189 Avoid calling xalloc(0). Change rrScreenSizeSet to rrScreenSetSize. 173190 173191commit ef1f3248cb5fff0a02c0059f865c4d931eba23a6 173192Author: Keith Packard <keithp@guitar.keithp.com> 173193Date: Tue Sep 19 22:48:54 2006 -0700 173194 173195 Split out 1.0-style info and new property routines to their own files. 173196 173197commit 07112adb0802d28488de5a495aa61bb3cfc280b6 173198Author: Keith Packard <keithp@guitar.keithp.com> 173199Date: Tue Sep 19 00:46:27 2006 -0700 173200 173201 RRGetScreenResources and RRGetOutputInfo are working now. 173202 173203 Removed separate id field in RRModeRec. 173204 Pull screen subpixel order from Render extension. 173205 Implement RGetScreenResources and RRGetOutputInfo 173206 173207commit afe5e9483b352ed06075ed68a6ffa50799194e2d 173208Author: Keith Packard <keithp@guitar.keithp.com> 173209Date: Mon Sep 18 12:18:22 2006 -0700 173210 173211 RandR working with old clients and old API. 173212 173213commit 2be1ac15aee592782d7693b8de2c3815478a094e 173214Author: Keith Packard <keithp@guitar.keithp.com> 173215Date: Mon Sep 18 12:11:18 2006 -0700 173216 173217 Remove smashing of CFLAGS from server build. 173218 173219 CFLAGS is a user variable, extracted from the environment at configure time 173220 and settable by the user at build time. We must not override this variable. 173221 173222commit bf07893947cfca945598e194ed416fda6162b11c 173223Author: Keith Packard <keithp@neko.keithp.com> 173224Date: Sun Sep 17 23:03:23 2006 -0700 173225 173226 Split out RandR dispatch code from randr.c to rr*dispatch.c. 173227 173228 More disassembly to ease ongoing development. 173229 173230commit 3e745745fecef1cb59e53bde52ded311b51e1dac 173231Author: Keith Packard <keithp@neko.keithp.com> 173232Date: Sat Sep 16 23:21:37 2006 -0700 173233 173234 Split RandR implementation into separate files. 173235 173236 RandR is getting too big to live in one file; split into one file per object 173237 type (crtc, mode, screen), leaving the rest of the code in randr.c. 173238 173239 Code is slowly approaching the point where it will drop-in as a replacement 173240 for the old 1.0 implementation. 173241 173242commit d17fb9672e238a089e463ac74cc4cd3325b67e1f 173243Author: Keith Packard <keithp@neko.keithp.com> 173244Date: Sat Sep 16 21:44:42 2006 -0700 173245 173246 Start moving to new randr 1.2 definition 173247 173248commit 8dec74321d916f204f8182f1b93a65defbe50e78 173249Author: Keith Packard <keithp@neko.keithp.com> 173250Date: Mon Jul 17 14:43:07 2006 -0400 173251 173252 Successful legacy RandR API/Protocol emulation for query. 173253 173254 These changes clean up minor errors to make it possible to list the 173255 available modes for a monitor using legacy APIs in both the X server DDX and 173256 RandR protocol. Setting modes is untested, so it probably doesn't work. 173257 173258commit cab3a0145f2483fe43b5db5f5dd2076db9757fe5 173259Author: Keith Packard <keithp@neko.keithp.com> 173260Date: Mon Jul 17 01:21:11 2006 -0400 173261 173262 RandR: New data structure, old API. At least it compiles now 173263 173264commit d95c758630f4aacec339a7ec80d2c4a9d7de1e4a 173265Author: Keith Packard <keithp@neko.keithp.com> 173266Date: Sat Jul 1 19:46:38 2006 -0700 173267 173268 Preliminary RandR 1.2 work 173269 173270commit f057de4f73fa593fa3fc5f05f65b89e76273b158 173271Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173272Date: Sat Sep 16 03:49:11 2006 -0400 173273 173274 Don't install librac.a. 173275 173276 Thanks, automake. 173277 173278commit 6926776f0ecd1e8e81c5c40ccd3a97227bc44dcb 173279Merge: 6950267dd 49a70c857 173280Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173281Date: Fri Sep 15 15:28:13 2006 -0400 173282 173283 Merge branch 'my-XACE-modular' into my-XACE-SELINUX 173284 173285commit 49a70c8570b03aff8239324a2474918a6fbc52a0 173286Merge: d1110c5c8 05231e336 173287Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173288Date: Fri Sep 15 15:26:57 2006 -0400 173289 173290 Merge branch 'master' into my-XACE-modular 173291 173292commit 46af6d1e953f1eefb6edbba3d29fb9700e42c2bb 173293Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173294Date: Thu Sep 14 19:28:44 2006 -0400 173295 173296 Always believe the monitor when it reports a reduced-blanking mode. 173297 173298 CVT reduced blanking modes are typically only seen on digital connections to 173299 LCDs, but there are some monitors that report them as supported over the 173300 VGA connector too, which is perfectly legitimate, electrically speaking. 173301 173302commit 63acf18b7e4ce3a9f7deab3a9088a1c41cab0191 173303Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173304Date: Thu Sep 14 19:26:37 2006 -0400 173305 173306 In xf86MatchPciInstances, fail gracefully when there's no PCI device at all. 173307 173308 This allows the autoconfig logic to fall through sanely on non-PCI machines, 173309 which importantly includes Xen virtual machines. 173310 173311commit a8f9936f55c5364bb02e8c3187507eb1f70e2ef2 173312Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173313Date: Thu Sep 14 19:24:41 2006 -0400 173314 173315 Prefer driver-provided modes when matching name strings to modelines. 173316 173317 Well, kinda. Strictly we prefer M_T_BUILTIN strongest since those are modes 173318 where the driver has said it absolutely can't do anything else (VBE). Then 173319 we look for user-defined modes, ie, modelines from the config file. Then 173320 we consider modes reported by the monitor via EDID. Finally if nothing has 173321 matched yet we consider the default mode pool. 173322 173323 Within each of the above-mentioned classes, modes with the M_T_PREFERRED bit 173324 take priority over other modes in the same class. 173325 173326 This logic ensures that the timings sent to the monitor exactly match the 173327 timings it reported as supported, which occasionally don't match the numbers 173328 you might get for that mode from CVT or GTF. 173329 173330commit 81ef1b6d6063c20db4963abf7b7848e235aa4ebb 173331Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173332Date: Thu Sep 14 19:18:58 2006 -0400 173333 173334 Mark EDID modes as driver modes. Infer virtual size from driver modes. 173335 173336 This allows the server to guess an appropriate initial virtual size and 173337 resolution. The heuristic is to select the largest driver-reported mode 173338 that matches the monitor's physical aspect ratio. We revalidate this 173339 estimate after mode validation, since we may have filtered away all 173340 modes that would fill that size. 173341 173342 Also, the EDID preferred timing is now marked as M_T_PREFERRED as well. 173343 173344commit 43d9edd31e31b33b9da4a50d8ab05004881c8d5a 173345Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173346Date: Thu Sep 14 19:09:02 2006 -0400 173347 173348 Attempt to add the 'mouse' driver in more situations. 173349 173350 Always add a mouse driver instance configured to send core events, unless 173351 a core pointer already exists using either the mouse or void drivers. This 173352 handles the laptop case where the config file only specifies, say, 173353 synaptics, which causes the touchpad to work but not the pointing stick. 173354 We don't double-instantiate the mouse driver to avoid the mouse moving twice 173355 as fast, and we skip this logic when the user asked for a void core pointer 173356 since that probably means they want to run with no pointer at all. 173357 173358commit 739224d05eb4f356c9cab9dcb8a44a8d78287765 173359Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173360Date: Thu Sep 14 19:03:32 2006 -0400 173361 173362 Load the default module set when no Module section is given in the config. 173363 173364 Also, synchronize that list with the list for the pseudoconfig file used 173365 when starting with no config file. These really need to be better unified. 173366 173367commit beac2bf1e48e6b77dbf7d95f086abc5abcd90cf0 173368Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173369Date: Thu Sep 14 19:01:13 2006 -0400 173370 173371 Expand the default sync ranges to be large enough for 800x600@60. 173372 173373commit 71a15a7d5721073eccb3a275f353b3aa584c0d68 173374Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173375Date: Thu Sep 14 19:00:10 2006 -0400 173376 173377 Publish the raw EDID block as a property on the root window. 173378 173379 This was removed in the patch for bug #5386, but is still useful. 173380 173381commit 7939c8dfb7c7bed4febcdc12922fb2e17619ea36 173382Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173383Date: Thu Sep 14 18:57:57 2006 -0400 173384 173385 Bump the default pixel depth to 24, and default bpp to 32. 173386 173387commit 72af975f9c8de0ff6796f1ce4b76dcf841d21e99 173388Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173389Date: Thu Sep 14 18:56:34 2006 -0400 173390 173391 Fix up EDID blocks where the max pixclock exceeds the preferred mode clock. 173392 173393 Base EDID only lets you specify the maximum dotclock in tens of MHz, which 173394 is too fuzzy for some monitors. 1600x1200@60 is just over 160MHz, but if 173395 the monitor really can't handle any mode at 170MHz, then 160 is more 173396 correct. Fix up the EDID block before the driver can see it in this case, 173397 so we don't spuriously reject modes. 173398 173399commit d05e0a97bb704a4986cf638487205da759c4ce17 173400Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173401Date: Thu Sep 14 18:49:12 2006 -0400 173402 173403 Enable DPMS by default. 173404 173405commit 334f7db9f653113d5d46236911d7de2ec4173f28 173406Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173407Date: Thu Sep 14 18:46:10 2006 -0400 173408 173409 Allow hsync and vsync ranges to be overridden independently again. 173410 173411commit ced46e17777b635df9371c4cfaec3f8968b4dbcf 173412Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173413Date: Thu Sep 14 18:41:59 2006 -0400 173414 173415 Record all standard timings from EDID as modes, instead of just the first five. 173416 173417commit d89fee68d0e49211871cd9eb3893ed55c1d478a6 173418Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173419Date: Thu Sep 14 18:41:15 2006 -0400 173420 173421 Record the maximum dot clock of the monitor, and filter by it. 173422 173423commit fa8ef7166839a7435e0017683f3e3c7f7904b285 173424Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173425Date: Thu Sep 14 18:33:00 2006 -0400 173426 173427 Don't translate monitor gamma to X gamma. 173428 173429 The X gamma is used to set the output ramp of the card. Setting a 2.2 output 173430 gamma going into a 2.2 monitor gives an effective gamma of 4.84, which is 173431 very much not what you want. 173432 173433commit 511c60bc7399b07c267d686a969880e5ec92408a 173434Author: Luc Verhaegen <libv@skynet.be> 173435Date: Thu Sep 14 18:30:36 2006 -0400 173436 173437 Bug #5386: Synthesize modelines from EDID data. 173438 173439commit 05231e336db8f959c15dda518641976f061df1a6 173440Author: Ian Romanick <idr@us.ibm.com> 173441Date: Thu Sep 14 14:13:39 2006 -0700 173442 173443 Use correct opcodes for GLX_EXT_texture_from_pixmap. 173444 173445 Regenerate from glX_API.xml 1.3 from Mesa. The glproto package and libGL 173446 (from Mesa) must also be updated. 173447 173448commit 0a62840e2ce25e5c2554e7e5ab4c9c5b96899e2d 173449Author: Bill Nottingham <notting@redhat.com> 173450Date: Wed Sep 13 15:40:23 2006 -0700 173451 173452 Bug 7641: fix comment written to Xorg.conf (s/VertSync/VertRefresh/) 173453 173454 X.Org Bugzilla #7641 <https://bugs.freedesktop.org/show_bug.cgi?id=7641> 173455 Patch #6349 <https://bugs.freedesktop.org/attachment.cgi?id=6349> 173456 173457commit 182e5e0f4ba4c98a34bc52bdf4032ba315fe80ad 173458Author: Drew Parsons <dparsons@debian.org> 173459Date: Tue Sep 12 14:30:46 2006 +1000 173460 173461 Xprint: revert installation of /etc/X11/Xsession.d/cde_xsessiond_xprint.sh 173462 pending resolution of #8232. 173463 173464commit 594d4019c613b0f4bf8f48cc074ecc3c8366f1d7 173465Author: Tilman Sauerbeck <tilman@code-monkey.de> 173466Date: Tue Sep 12 01:15:40 2006 +0200 173467 173468 transformIsIdentity() now doesn't accept a zero matrix as the identity. 173469 173470 Added a non-zero test for one of the diagonal values. 173471 173472commit fc30370d14125f86ee1192890a184881fa139546 173473Author: Tilman Sauerbeck <tilman@code-monkey.de> 173474Date: Mon Sep 11 19:43:09 2006 +0200 173475 173476 Bug #8226: Fixed SetPictureTransform()'s handling of the argument matrix. 173477 173478 It now recognizes scaled variants of the identity matrix, too. 173479 173480commit 2b357e9a2f9038cf9cd07da908e3103a3d0965c9 173481Author: Donnie Berkholz <dberkholz@gentoo.org> 173482Date: Sun Sep 10 22:17:20 2006 -0700 173483 173484 If we're installing libxf86config, install headers needed to build against it. 173485 173486commit 58933757862c458e2da39bd464e50d9c0e41b313 173487Author: Zephaniah E. Hull <warp@agamemnon.b5> 173488Date: Sun Sep 10 15:50:51 2006 -0400 173489 173490 Warning fix, and a syntax fix in a #if 0 section of code. 173491 173492commit 0a3740a0000191e3039fe183ae51b938d0548340 173493Author: Zephaniah E. Hull <warp@agamemnon.b5> 173494Date: Sun Sep 10 15:49:25 2006 -0400 173495 173496 Typo correction, 'i' is not a '1', so no longer crash on some Xi requests. 173497 173498commit 8d709f0280b458515b32c2b87938749428e5c149 173499Author: Zephaniah E. Hull <warp@agamemnon.b5> 173500Date: Sun Sep 10 15:48:35 2006 -0400 173501 173502 Remove a merge artifact so that we can compile. 173503 173504commit b3a3020fd018df8bc5a8193d36e1a1c7ae8af8ba 173505Author: Jesse Barnes <jbarnes@nietzche.virtuousgeek.org> 173506Date: Sun Sep 10 11:13:18 2006 -0700 173507 173508 the new PCI mapping routines are broken on sparc64 (in fact they look 173509 broken for any 32 bit X server running on a 64 bit kernel) so #ifdef 173510 them out for now. the PCI rework tree will make all this crap go away, 173511 so I think we can tolerate the extra #ifdef for the next release. 173512 173513commit 60db190ecfce52cbfa888c0af3210634f9186bed 173514Merge: 5e9d33fe8 65256109b 173515Author: Zephaniah E. Hull <warp@agamemnon.b5> 173516Date: Sun Sep 10 03:49:17 2006 -0400 173517 173518 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 173519 173520commit 65256109bb8f5a26704ed960e1dd113981df5787 173521Author: Drew Parsons <dparsons@debian.org> 173522Date: Sun Sep 10 17:40:37 2006 +1000 173523 173524 * Define XPSERVERLIST with `/etc/init.d/xprint get_xpserverlist` 173525 instead of `/bin/sh /etc/init.d/xprint get_xpserverlist` 173526 - allows the initscript to set its own different shell under #! 173527 - allows disabling of XPSERVERLIST by making the script non-executable 173528 * Allow files to be installed by using dist_*_DATA instead of EXTRA_DIST. 173529 Also, use dist_*_SCRIPTS to install scripts. 173530 * Fix minor typos in man pages. 173531 173532commit 6950267dd690ef8e29b1c32a157dd64c9b79c06d 173533Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173534Date: Fri Sep 8 15:31:18 2006 -0400 173535 173536 Add SELinux extension configure-time support. 173537 173538commit fb34c02861ab3629c1c85c156e73b158518db7c7 173539Merge: 9c503f09c d1110c5c8 173540Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173541Date: Fri Sep 8 15:30:12 2006 -0400 173542 173543 Merge branch 'my-XACE-modular' into my-XACE-SELINUX 173544 173545commit d1110c5c83a7f439158f369ab2f3ae614fa9d2a5 173546Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173547Date: Fri Sep 8 15:28:48 2006 -0400 173548 173549 Generalize the handling of configuration files that ship with extensions. 173550 173551commit 9deb579dc9366590203afe0576bf88643ab36c89 173552Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173553Date: Fri Sep 8 15:25:17 2006 -0400 173554 173555 Zero out newly created ExtensionEntry structures, but only after the 173556 devPrivates have been initialized. 173557 173558commit cec392656cda1c938d5462e1949e6eef489f9168 173559Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173560Date: Fri Sep 8 15:24:23 2006 -0400 173561 173562 Zero out newly allocated ClientRec structures. 173563 This is required to initialize the devPrivates to a known state. 173564 173565commit 0fba09cdfcc78161f5c92bef6cca53e5309656bd 173566Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173567Date: Fri Sep 8 15:23:06 2006 -0400 173568 173569 Include dix-config.h. 173570 173571commit c93877100eb98647c5b6b8556730d54677f730b6 173572Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173573Date: Fri Sep 8 15:21:57 2006 -0400 173574 173575 Don't need to allocate memory now that devPrivates are being used. 173576 173577commit 9c503f09ce78d952d0ece77c424e42b6df3fa9ad 173578Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173579Date: Fri Sep 8 15:17:01 2006 -0400 173580 173581 Add SELinux extension to the module/extension loader. 173582 173583commit 9aa44e3e4c321f42d8e64f83c7f0932470593c26 173584Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173585Date: Fri Sep 8 15:15:19 2006 -0400 173586 173587 Add SELinux extension configure-time support. 173588 173589commit 150eabc4c8a08c81c48493583f922a1240b7e91c 173590Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173591Date: Fri Sep 8 15:11:04 2006 -0400 173592 173593 Add SELinux extension source files. 173594 173595commit 86450998da616e3d00d4d6293acc35eccc2061e7 173596Author: Kristian Høgsberg <krh@redhat.com> 173597Date: Thu Sep 7 15:35:16 2006 -0400 173598 173599 Fix AIGLX VT switching. 173600 173601 See https://bugs.freedesktop.org/show_bug.cgi?id=7916 173602 173603 There may be a simpler, less intrusive fix that involves just rearranging 173604 DRI locking between 2D and 3D drivers around VT switch. 173605 173606commit 5e9d33fe87f9d24e55c468d4b2bb761c9efdb26a 173607Merge: 629798c73 64479fffa 173608Author: Daniel Stone <daniel@fooishbar.org> 173609Date: Thu Sep 7 15:43:31 2006 +0300 173610 173611 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 173612 173613commit 629798c73ad76a77fa6a55bc6403fd9b95ade2bb 173614Author: Daniel Stone <daniel@fooishbar.org> 173615Date: Thu Sep 7 15:17:04 2006 +0300 173616 173617 XkbCopyKeymap/SrvXkbCopyKeymap: free geom harder, add cheery comments 173618 Unconditionally free geometry when copying the keymap (so we have none on 173619 core, oh well), add a couple of heartening comments. 173620 173621commit 64479fffa22581cc7d753065c33eda5520b7db9a 173622Author: Ian Romanick <idr@us.ibm.com> 173623Date: Wed Sep 6 16:13:21 2006 -0700 173624 173625 Remove prototypes and externs for non-existant functions and variables. 173626 173627commit a0179281a6522ec59830e8f2549633741bc56e10 173628Author: Ian Romanick <idr@us.ibm.com> 173629Date: Wed Sep 6 15:45:48 2006 -0700 173630 173631 Remove prototypes for non-existant functions. 173632 173633commit 8356be492c6b46abdffa08b13836571ed872e16f 173634Author: Michel Dänzer <michel@tungstengraphics.com> 173635Date: Wed Sep 6 15:20:55 2006 +0200 173636 173637 Make sure _XSERVER64 is defined when it should be and gets tested. 173638 173639commit f6ce0839ba5b73247097826d28f7388fe248ec0c 173640Author: Michel Dänzer <michel@tungstengraphics.com> 173641Date: Wed Sep 6 13:18:02 2006 +0200 173642 173643 Fix #include paths for fontcacheproto headers. 173644 173645commit f39fd4242902eaa862321d39337f429dd14ebacf 173646Author: Aaron Plattner <aplattner@nvidia.com> 173647Date: Tue Sep 5 15:23:54 2006 -0700 173648 173649 (unsigned long)(1 << 31) = bad news on x86_64. 173650 (cherry picked from 410e5b1d738ba47b36778e6cbed44023a27ce259 commit) 173651 173652commit 410e5b1d738ba47b36778e6cbed44023a27ce259 173653Author: Aaron Plattner <aplattner@nvidia.com> 173654Date: Tue Sep 5 15:23:54 2006 -0700 173655 173656 (unsigned long)(1 << 31) = bad news on x86_64. 173657 173658commit 0b81fccd2ee4e054e5cffb739de07460ff2c13f7 173659Merge: 20c4ac6e0 c2813514c 173660Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173661Date: Tue Sep 5 18:03:25 2006 -0400 173662 173663 Merge branch 'master' into my-XACE-modular 173664 173665 Conflicts: 173666 173667 configure.ac 173668 173669commit c2813514cf7b1a36caa848cbc2ceef99cf2eb769 173670Author: Ian Romanick <idr@us.ibm.com> 173671Date: Thu Aug 31 15:36:13 2006 -0700 173672 173673 Add missing file from previous commit. 173674 173675commit 0f9cfb2f752a9010ff07f4b2bd891db0cc30b8e6 173676Author: Ian Romanick <idr@us.ibm.com> 173677Date: Thu Aug 31 13:54:10 2006 -0700 173678 173679 Implement GLX_SGI_swap_control. 173680 173681 Regenerate from glX_API.xml 1.2. Add infrastructure to support 173682 GLX_SGI_swap_control for AIGLX when the DRI driver enables it. Tested 173683 with R300. 173684 173685commit a9ef5862919313582f72fc0cfb5ab0af4df6507e 173686Author: Ian Romanick <idr@us.ibm.com> 173687Date: Thu Aug 31 13:47:50 2006 -0700 173688 173689 Fix problems with vertex program protocol 173690 173691 There were two sets of bugs in the vertex program (ARB and NV) 173692 protocol. First, several of the ARB functions were missing the 173693 'doubles_in_order="true"' annotation. Second, after the ARB decided 173694 that glVertexAttrib*ARB functions must not alias fixed-function state 173695 for GLSL, Nvidia re-assigned GLX protocol opcodes for 173696 glVertexAttrib*NV (circa Septeber 2004). For some reason gl_API.xml 173697 was never updated to reflect this, and the updated version of the 173698 GL_NV_vertex_program spec never made into the registry. 173699 173700 This is just a server-side regeneration from gl_API.xml version 1.68. 173701 173702commit 69d5becce4ca2cfc8f8de53672ed54a47de62164 173703Author: Matthew Allum <mallum@openedhand.com> 173704Date: Thu Aug 31 17:30:24 2006 +0100 173705 173706 Fix previous commit breaking other kdrives pulling in fbdev.a 173707 173708commit fd609956f27d76ee76ac8623787f0fc8633a5546 173709Author: Matthew Allum <mallum@openedhand.com> 173710Date: Thu Aug 31 17:18:57 2006 +0100 173711 173712 Add framebuffer device command line switch for Xfbdev 173713 173714commit 2fb7b8795a9a36cce61f6449f6ca26ffd1b071f0 173715Author: Ian Romanick <idr@us.ibm.com> 173716Date: Tue Aug 29 16:35:32 2006 -0700 173717 173718 Minor extension tweaks. 173719 173720 GLX_EXT_texture_from_pixmap should always be enabled. 173721 GLX_SGI_video_sync is only for direct rendering and should never 173722 appear in the server's string. 173723 173724commit 1c8851ad491dd02d1c79e620b46384956838ed42 173725Merge: d59b52fc0 5ddbf4bcd 173726Author: Ian Romanick <idr@us.ibm.com> 173727Date: Tue Aug 29 16:34:04 2006 -0700 173728 173729 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 173730 173731commit d59b52fc08f2d80b38993e383e61c3eeb0bb0763 173732Author: Ian Romanick <idr@us.ibm.com> 173733Date: Tue Aug 29 14:40:13 2006 -0700 173734 173735 Make sure unsupported extensions are disabled. 173736 173737 GLX protocol isn't supported for GLX_SGI_swap_control or 173738 GLX_SGI_video_sync. Remove them from the list of available extensions 173739 until they are supported. 173740 173741commit db6d04d4b87fd9b6409a3ddf0479a88440c2eda1 173742Author: Ian Romanick <idr@us.ibm.com> 173743Date: Tue Aug 29 14:35:08 2006 -0700 173744 173745 Add support for AIGLX drivers to enable GLX extensions that they support. 173746 173747commit 5ddbf4bcd46fe0d3d682668c2748c712fea410ae 173748Author: Matthew Allum <mallum@openedhand.com> 173749Date: Tue Aug 29 22:07:15 2006 +0100 173750 173751 Re-add support for tslib (1.0 release) and Xcalibrate extension. 173752 173753commit 4524a2bf6f22c871ed109b027a065f0262137dc5 173754Author: Daniel Stone <daniel@fooishbar.org> 173755Date: Tue Aug 29 23:49:26 2006 +0300 173756 173757 configure.ac: move tslib from KDRIVE_PURE_LIBS to KDRIVE_LIBS 173758 Yeah. That was dumb. 173759 173760commit 942b4369990a255257f66835caf8671432c405a3 173761Merge: 77d315bd2 393dc0a33 173762Author: Ian Romanick <idr@us.ibm.com> 173763Date: Tue Aug 29 13:30:37 2006 -0700 173764 173765 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 173766 173767commit 77d315bd2fd2f9014c831d313efbe5821189177c 173768Author: Ian Romanick <idr@us.ibm.com> 173769Date: Tue Aug 29 13:30:20 2006 -0700 173770 173771 Remove __glXNoSuchRenderOpcode because it is no longer used. 173772 173773commit 260c3f32b69c98f8fc5360f860f69d32c19f04a7 173774Author: Daniel Stone <daniel@fooishbar.org> 173775Date: Tue Aug 29 23:18:12 2006 +0300 173776 173777 configure.ac: fix XSDL test 173778 173779commit 393dc0a3388d56186181b2bd9bcc1d267747e709 173780Author: Daniel Stone <daniel@fooishbar.org> 173781Date: Tue Aug 29 22:53:54 2006 +0300 173782 173783 kdrive: remove @KDRIVE_LIBS@ from Xfoo_DEPENDENCIES 173784 173785commit cff23616fe45e10c6786a303c8dcfc0a80463a53 173786Author: Daniel Stone <daniel@fooishbar.org> 173787Date: Tue Aug 29 22:44:09 2006 +0300 173788 173789 configure.ac: allow disabling of XSDL 173790 173791commit 89d272bb183e85715d8e6047929fb2d912033d82 173792Author: Daniel Stone <daniel.stone@nokia.com> 173793Date: Tue Aug 29 15:05:31 2006 +0300 173794 173795 [PATCH] kdrive/linux keyboard: remove more debugging spew 173796 173797commit bd6f539ff9409aa7d9056fabe120b457b0a15997 173798Author: Daniel Stone <daniel.stone@nokia.com> 173799Date: Tue Aug 29 13:21:58 2006 +0300 173800 173801 [PATCH] kdrive/linux keyboard: silence excessive debugging noise 173802 173803commit 5436fce09003e20744a388fa4ae49007c9cf8ede 173804Author: Daniel Stone <daniel.stone@nokia.com> 173805Date: Tue Aug 29 13:21:40 2006 +0300 173806 173807 [PATCH] GetKeyboardValutorEvents: be even more careful 173808 173809 Don't accept devices without a keyboard feedback class. 173810 173811commit 0eb7299f445455a7bcacf2410e83227b23259675 173812Author: Daniel Stone <daniel.stone@nokia.com> 173813Date: Tue Aug 29 13:19:12 2006 +0300 173814 173815 [PATCH] XkbCopyKeymap: still more range fixes 173816 173817 Make sure we don't stomp preserve if it doesn't already exist, and fix a 173818 couple of range-related thinkos in level name copying. 173819 173820commit 7fa3383e3c8eea7d1eb0e556393f2431cf8e6ed2 173821Merge: 8d77d44fd ebbdc1342 173822Author: Daniel Stone <daniel@fooishbar.org> 173823Date: Tue Aug 29 15:16:01 2006 +0300 173824 173825 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 173826 173827commit ebbdc1342a243b301723390696f742dc91f59764 173828Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173829Date: Mon Aug 28 18:17:32 2006 -0400 173830 173831 Remove calls to LoaderCheckUnresolved(), since it's now a stub. 173832 173833commit 8d77d44fda3aacbae62864a3620e09095b79e92d 173834Merge: d6f36bd28 2fde560bb 173835Author: Daniel Stone <daniel@fooishbar.org> 173836Date: Sun Aug 27 23:08:49 2006 +0300 173837 173838 Merge branch 'origin' into input-hotplug 173839 173840commit 20c4ac6e038607ebbf6c04639670514c016d8597 173841Merge: 13c6713c8 8d4f21ab5 173842Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173843Date: Fri Aug 25 18:49:46 2006 -0400 173844 173845 Merge branch 'my-XACE-SELINUX' into my-XACE-modular 173846 173847commit 13c6713c82763a85c725c998b37ad02156d803ba 173848Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 173849Date: Fri Aug 25 18:17:01 2006 -0400 173850 173851 Add four new XACE hooks: auditing, key event notification, window init 173852 173853commit 2fde560bbb9c1148f26fd969dc30c4e736672b7c 173854Author: Ian Romanick <idr@umwelt.(none)> 173855Date: Fri Aug 25 13:01:51 2006 -0700 173856 173857 Enable GL_EXT_texture_filter_anisotropic and GL_EXT_blend_equation_separate. 173858 173859 Re-generate from gl_API.xml 1.65. This provides the missing bits for 173860 GL_EXT_texture_filter_anisotropic and GL_EXT_blend_equation_separate. 173861 Enable those extensions. 173862 173863commit e2d529963ed40b5f113cf82c17809d241cd4aac1 173864Author: Ian Romanick <idr@umwelt.(none)> 173865Date: Fri Aug 25 12:05:16 2006 -0700 173866 173867 Enable vertex and fragment programs. 173868 173869 Implement glGetProgramStringARB and glGetProgramStringNV. With these 173870 functions implemented, GL_ARB_{vertex,fragment}_program, 173871 GL_NV_{vertex,fragment}_program, and related extensions can be enabled. 173872 173873commit 21291d6ca7188e5733ed0c93215ee0f1f0f90cc9 173874Merge: 1a5561c4e cd2da4e41 173875Author: Ian Romanick <idr@umwelt.(none)> 173876Date: Fri Aug 25 09:34:21 2006 -0700 173877 173878 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 173879 173880commit d6f36bd28009881ef7f7a20cdadb3808d808ed97 173881Author: Daniel Stone <daniel@fooishbar.org> 173882Date: Fri Aug 25 12:43:17 2006 +0300 173883 173884 xfree86/parser: use 'kbd' driver when 'keyboard' specified 173885 Now that we've completely ditched the old driver, we should probably make a 173886 best-effort attempt to keep configs working. 173887 173888commit 7c4167f0d6b33c9c602b04fcfd246fd3aeddd709 173889Merge: 393f8347e cd2da4e41 173890Author: Daniel Stone <daniel@fooishbar.org> 173891Date: Fri Aug 25 11:15:33 2006 +0300 173892 173893 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 173894 173895commit 393f8347edcccfc24d8902a86ea9def7ada2537b 173896Author: Daniel Stone <daniel@fooishbar.org> 173897Date: Fri Aug 25 10:46:32 2006 +0300 173898 173899 xorgconf.cpp: kbd, not keyboard 173900 173901commit c33e39c86be2010b169ffbd8adbe53b93222dc5f 173902Author: Ian Romanick <idr@umwelt.(none)> 173903Date: Thu Aug 24 20:33:57 2006 -0700 173904 173905 Finish support for GL_ARB_texture_compression. 173906 173907 Fill in __glXDisp_GetCompressedTexImageARB and 173908 __glXDispSwap_GetCompressedTexImageARB to finish support for 173909 GL_ARB_texture_compression. With this extension (and the related 173910 compression extensions), the server-side GLX supports all of the 173911 protocol for GL 1.4. w00t! 173912 173913 The bad news is that this has received only minimal testing, and Mesa 173914 does not contain any good tests for GL_ARB_texture_compression. 173915 173916commit cd2da4e41eae233b50f8830d9a8f5d1d916a5a1b 173917Author: Ian Romanick <idr@umwelt.(none)> 173918Date: Thu Aug 24 18:00:16 2006 -0700 173919 173920 Remove GL/glx/g_disptab.c, GL/glx/g_disptab_EXT.c, and 173921 GL/glx/g_disptab_EXT.h. Unfortunately GL/glx/g_disptab.h has to be 173922 kept around a bit longer. 173923 173924commit a29e6dd2d2d45c18c52737bb3b7945aafcea5032 173925Author: Ian Romanick <idr@umwelt.(none)> 173926Date: Thu Aug 24 17:58:52 2006 -0700 173927 173928 Add some missing bits of GL_SGI_color_table. 173929 173930commit ae608b2071d882966e9c7ede71f846b1ecec0b23 173931Merge: 2c865277f b879356ce 173932Author: Ian Romanick <idr@umwelt.(none)> 173933Date: Thu Aug 24 14:56:33 2006 -0700 173934 173935 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 173936 173937commit 2c865277fe1d056981d1020e1af001d2319252c0 173938Author: Ian Romanick <idr@umwelt.(none)> 173939Date: Thu Aug 24 14:54:49 2006 -0700 173940 173941 Regenerate from gl_API.xml 1.63. Enable extensions. 173942 173943 gl_API.xml 1.63 corrects some problems with GLX protocol for 173944 GL_EXT_paletted_texture and GL_SGI_color_table. Regenerate from that 173945 file, and enable those extensions and GL_EXT_shared_texture_palette. 173946 173947commit 7d5de5c6657304246473d7ddd5c29bb0c7a3bc34 173948Author: Ian Romanick <idr@umwelt.(none)> 173949Date: Thu Aug 24 14:49:46 2006 -0700 173950 173951 Regenerate from gl_API.xml 1.62. Functions move, no real changes. 173952 173953commit 3a36b0a24aa9e9e238faa7f00100f59800f5142b 173954Merge: db1ab1bdb b879356ce 173955Author: Daniel Stone <daniel@fooishbar.org> 173956Date: Thu Aug 24 23:35:28 2006 +0300 173957 173958 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 173959 173960commit db1ab1bdb2f79eca593fe247056309a16ebd29c6 173961Author: Daniel Stone <daniel@fooishbar.org> 173962Date: Thu Aug 24 23:33:59 2006 +0300 173963 173964 XkbCopyKeymap: fix various range issues 173965 Fix a bunch of range issues caused by incorrect assumptions (e.g. that the 173966 design was at least halfway sensible), and copy types by hand, instead of 173967 just blindly memcpy()ing the lot, since it itself cleverly contains a ton 173968 of allocated pointers. 173969 173970commit 5fb8d947bb88d715b9b236342885c445cb5a9387 173971Author: Daniel Stone <daniel@fooishbar.org> 173972Date: Thu Aug 24 23:16:43 2006 +0300 173973 173974 configure.ac: more thinkos 173975 Fix auto tests for vidmode and xf86dga. I win at life. 173976 173977commit 4e37c07ba6e5d299d4f8922dc6cf054c814f7baf 173978Author: Daniel Stone <daniel@fooishbar.org> 173979Date: Thu Aug 24 23:16:17 2006 +0300 173980 173981 config: clean up debugging messages, make failure to acquire name fatal 173982 Bomb with FatalError when we can't acquire the bus and name. 173983 Clean up a bunch of debugging ErrorFs to be hidden behind #ifdef DEBUG. 173984 173985commit b879356ce96929d02bcb75b9aa24b17ac7e28125 173986Author: Adam Jackson <ajax@benzedrine.nwnk.net> 173987Date: Thu Aug 24 15:50:15 2006 -0400 173988 173989 More #ifdef USE_DEPRECATED_KEYBOARD_DRIVER. 173990 173991commit 4ed311cf1c29090c53e474a3001c5702ff8409df 173992Merge: 73e58adda b29b236d8 173993Author: Matthias Hopf <mhopf@suse.de> 173994Date: Thu Aug 24 20:17:10 2006 +0200 173995 173996 Merge branch 'master' of git://anongit.freedesktop.org/git/xorg/xserver 173997 173998commit b29b236d88789fd45d823a55dbedb393bb134c5b 173999Author: Lukáš Turek <8an@centrum.cz> 174000Date: Thu Aug 24 15:57:09 2006 +0200 174001 174002 Adapt to Mesa header name change. 174003 174004commit ce4a0a4ddafd3833d7025f83ed3729915c8aba70 174005Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 174006Date: Thu Aug 24 13:56:22 2006 +0100 174007 174008 Apply patch in bug #7919, blit improvements in 174009 multiwindow mode for Xming/CygwinX 174010 174011commit 866ca1f929c95689bac9f0a0b3478f7b4d77214b 174012Author: Daniel Stone <daniel@fooishbar.org> 174013Date: Thu Aug 24 15:46:44 2006 +0300 174014 174015 configure.ac: fix XF86VidMode test 174016 174017commit 2b06c69c8feaf3bdc065635ee711efa45b3033b3 174018Author: Daniel Stone <daniel@fooishbar.org> 174019Date: Thu Aug 24 14:51:26 2006 +0300 174020 174021 GKVE: pass correct arguments to XkbCopyKeymap 174022 Fix horrendous thinko. Indicators now work perfectly. 174023 174024commit 4adf9af313c9f63b6ad734e174efe1d36ddb5813 174025Merge: 33af05d58 67bd672c8 174026Author: Daniel Stone <daniel@fooishbar.org> 174027Date: Thu Aug 24 10:59:33 2006 +0300 174028 174029 Merge branch 'master' into input-hotplug 174030 174031commit 67bd672c880869ef625ae0c0163c3ec1eba46abf 174032Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 174033Date: Thu Aug 24 08:47:06 2006 +0100 174034 174035 Fix typo 174036 174037commit 733c4beb16c2c4ad9e9a4ea9a85b09fc5062a775 174038Author: David Nusinow <dnusinow@debian.org> 174039Date: Wed Aug 23 22:39:42 2006 +0000 174040 174041 Add xorg.conf IgnoreABI option which does the same thing as -ignoreABI 174042 174043commit b983773d446cef6a0948ca264ed48126e404ae9a 174044Merge: 0623d3643 d9a86566c 174045Author: Ian Romanick <idr@umwelt.(none)> 174046Date: Wed Aug 23 17:16:50 2006 -0700 174047 174048 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 174049 174050commit 0623d3643fc28ebc514b2ca872c985d0cf0c753a 174051Author: Ian Romanick <idr@umwelt.(none)> 174052Date: Wed Aug 23 17:16:02 2006 -0700 174053 174054 Fix the sorting of the extension string. Add a few extensions that 174055 are supported by the new code. A few of these were actually supported 174056 before but weren't advertised. 174057 174058commit 5d2caacff570dd68bb3fb05e776e02515b2a9da0 174059Author: Ian Romanick <idr@umwelt.(none)> 174060Date: Wed Aug 23 16:47:00 2006 -0700 174061 174062 Refector __glXDisp_Render and __glXDispSwap_Render to DoRender. 174063 Refector __glXDisp_RenderLarge and __glXDispSwap_RenderLarge to 174064 DoRenderLarge. 174065 174066commit 866bb3f34046045c9fa0744db1d76e035b3da9c7 174067Author: Ian Romanick <idr@umwelt.(none)> 174068Date: Wed Aug 23 16:41:53 2006 -0700 174069 174070 Memo to myself: Whenever a Makefile.am changes, autogen.sh must be 174071 re-run. This is especially true if the change is to remove a source 174072 file. 174073 174074 Fix RenderLarge to actually use the new protocol decode tables. 174075 174076commit d9a86566c21afd7985673f3ed851b055d9dac46f 174077Author: Alan Coopersmith <alan.coopersmith@sun.com> 174078Date: Wed Aug 23 16:15:19 2006 -0700 174079 174080 Add LOCALCONN to dix-config.h template for xtrans 174081 174082commit f6fd7d8f8393f93705e76b2b2777a0d9bcafa991 174083Author: Ian Romanick <idr@umwelt.(none)> 174084Date: Wed Aug 23 16:05:37 2006 -0700 174085 174086 Convert protocol decode tabels for Render and RenderLarge to use nice, 174087 compact N-way search trees generated by scripts in Mesa. 174088 174089commit 7ae82b5fc8721be78b43a322bbf2c46aac08b8cf 174090Author: Ian Romanick <idr@umwelt.(none)> 174091Date: Wed Aug 23 16:00:48 2006 -0700 174092 174093 Fix __glXDispatchInfo::dispatch_functions and 174094 __glXDispatchInfo::size_table. dispatch_functions had the const in 174095 the wrong place, and size_table was declared as an array of two 174096 pointers to int_fast16_t instead of a pointer to an array of 2 174097 int_fast16_t. cdecl to the rescue! 174098 174099commit 39a620d17809dc71fb5ad61a955fe3c442f90a05 174100Author: Ian Romanick <idr@umwelt.(none)> 174101Date: Wed Aug 23 14:24:34 2006 -0700 174102 174103 Rename __glXDrawArraysSize to __glXDrawArraysReqSize. This makes its 174104 name match the pattern of all the other functions in 174105 __glXRenderSizeTable. 174106 174107commit 86406455f0e5fc977431948611e9bb5fda1e1d46 174108Author: Ian Romanick <idr@umwelt.(none)> 174109Date: Wed Aug 23 13:30:59 2006 -0700 174110 174111 Re-generated files after a fix to glX_API.xml (in Mesa). 174112 174113commit d7a7f12361d31001bbd9394a57de029ef0b934b8 174114Author: Ian Romanick <idr@umwelt.(none)> 174115Date: Wed Aug 23 13:30:13 2006 -0700 174116 174117 Convert protocol decode tables for Single, VendorPrivate, and 174118 VendorPrivateWithReply message to use nice, compact N-way search trees 174119 generated by scripts in Mesa. 174120 174121 The Render protocol decode tables are next... 174122 174123commit bdec9680fa74dd23cf319d09af1940f8cf71a5b1 174124Author: Adam Jackson <ajax@benzedrine.nwnk.net> 174125Date: Wed Aug 23 14:43:23 2006 -0400 174126 174127 Make sure Composite is never enabled for Xnest. 174128 174129commit 9f2a108051aad9b024ab737b45fc12290a113e37 174130Author: Adam Jackson <ajax@benzedrine.nwnk.net> 174131Date: Wed Aug 23 14:38:34 2006 -0400 174132 174133 Make 'Xvfb -render' also disable Composite, lest we segfault on startup. 174134 174135commit 33af05d58f1f4f021036e9ce4b60fd76dbaebe73 174136Author: Daniel Stone <daniel@fooishbar.org> 174137Date: Wed Aug 23 19:05:50 2006 +0300 174138 174139 XkbCopyKeymap: use correct range for MapNotify 174140 We haven't copied {min,max}_key_code by the time the notifies run, so use 174141 src instead of dst to determine number of keys, et al. 174142 174143commit 6323a11d0db4d3cf0317af83f0362730142f5325 174144Author: Daniel Stone <daniel@fooishbar.org> 174145Date: Wed Aug 23 18:53:04 2006 +0300 174146 174147 XkbCopyKeymap: optionally send NewKeyboardNotify/MapNotify events 174148 Optionally send a NewKeyboardNotify or MapNotify event when copying the 174149 keymap; modify GetKeyboardValuatorEvents to make use of this. 174150 174151commit 728fbadd16a748b45c80bc2c65c46f82cf803578 174152Author: Daniel Stone <daniel@fooishbar.org> 174153Date: Wed Aug 23 14:33:59 2006 +0300 174154 174155 gitignore: ignore vi swap files 174156 174157commit 8f8487ff997670a4af0293fed77ff920cfc39fb1 174158Author: Daniel Stone <daniel@fooishbar.org> 174159Date: Wed Aug 23 14:33:41 2006 +0300 174160 174161 xkb/gkve: copy XKB map, not pointer-assign 174162 Write a new function to copy an XKB map (does everything but geometry at 174163 the moment), and use that instead of nasty pointer assignments. 174164 174165commit 52ba722e4c89c052609b4fc62e965d92778aa2dd 174166Merge: 9138d5a51 05541259b 174167Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 174168Date: Mon Aug 21 18:49:31 2006 -0400 174169 174170 Merge branch 'XACE-modular' into my-XACE-modular 174171 174172commit 05541259bdb0dfaab015a01caa3722b7a1b782e2 174173Merge: c2535f679 a1ac0440b 174174Author: Alan Coopersmith <alan.coopersmith@sun.com> 174175Date: Mon Aug 21 13:07:41 2006 -0700 174176 174177 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into XACE-modular 174178 174179commit a56b98bb047003a05e26ca9365c212a2da7ac200 174180Author: Daniel Stone <daniel.stone@nokia.com> 174181Date: Fri Aug 18 18:03:41 2006 +0300 174182 174183 dix: enable null root cursor 174184 174185 Enable a blank root cursor, selectable with --enable-null-root-cursor at 174186 configure time. 174187 174188commit 0704bb298cc826cd117815898c6bc015a693c2c9 174189Merge: c14036977 a1ac0440b 174190Author: Daniel Stone <daniel@fooishbar.org> 174191Date: Fri Aug 18 17:30:14 2006 +0300 174192 174193 Merge branch 'master' into input-hotplug 174194 174195commit a1ac0440bba690368aa4226468ce571be1a09d95 174196Author: Daniel Stone <daniel@fooishbar.org> 174197Date: Fri Aug 18 17:30:00 2006 +0300 174198 174199 dix: fix whiteroot thinko 174200 Note to self: run git update-index _after_ testing, not just before. 174201 174202commit c14036977fef7b8787c0b68f5262fa0b6a2834f5 174203Author: Daniel Stone <daniel@fooishbar.org> 174204Date: Fri Aug 18 17:24:34 2006 +0300 174205 174206 input.h: add InitCoreDevices prototype 174207 174208commit 1c2cb30cd88ba4453f9da339025f8ff39f7f5412 174209Merge: 633b6a69f 70ddd0f39 174210Author: Daniel Stone <daniel@fooishbar.org> 174211Date: Fri Aug 18 17:05:50 2006 +0300 174212 174213 Merge branch 'origin' into input-hotplug 174214 174215commit 19f673b7788d32c220e7e06734f1074b0e4a999c 174216Merge: cb0a565d2 70ddd0f39 174217Author: Daniel Stone <daniel@fooishbar.org> 174218Date: Fri Aug 18 17:05:41 2006 +0300 174219 174220 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 174221 174222commit cb0a565d2b2cf8823abbd77b4426cc2237731dc1 174223Author: Daniel Stone <daniel@fooishbar.org> 174224Date: Fri Aug 18 17:04:48 2006 +0300 174225 174226 dix: add whiteroot flag 174227 Add a -wr option to use a white root window, and use a BackPixel rather 174228 than BackPixmap for both white and black root windows. 174229 174230commit 70ddd0f39d5118db72a1a4e473cbfb502f1ed9ec 174231Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 174232Date: Fri Aug 18 14:43:10 2006 +0100 174233 174234 Fix bug #5735, Serious flaw in CygwinX clipboard 174235 integration prevents paste from X to Windows apps 174236 (Brett Stahlman & Colin Harrison) 174237 174238commit 708b225689b5a4ba9ffe3372b584b715ef9eacdc 174239Merge: e1f4565be ee5e2cbd2 174240Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 174241Date: Fri Aug 18 09:13:52 2006 +0100 174242 174243 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 174244 174245commit e1f4565be5ce80be4655e81f77f4073fa3fbf8d0 174246Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 174247Date: Fri Aug 18 09:11:48 2006 +0100 174248 174249 Fix bug #7302, make Xn.hosts work from the 174250 Microsoft Windows install directory on Xming. 174251 (Colin Harrison) 174252 174253commit 1880defe4eaba02f9585b154d0883235eabc6d11 174254Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 174255Date: Fri Aug 18 09:09:53 2006 +0100 174256 174257 Fix bug #7281, clipboard viewer should not 174258 call SetClipboard viewer when bogus in Xming/CygwinX 174259 (Colin Harrison) 174260 174261commit a1a8e4f7f5917f537eb3dd51d3d6fa3e129236ce 174262Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 174263Date: Fri Aug 18 09:08:12 2006 +0100 174264 174265 Fix bug #7280, round title corner background 174266 should be transparent not black in Xming/CygwinX 174267 (Colin Harrison) 174268 174269commit ee5e2cbd2bee610a95facc6b486c4a5070973099 174270Author: Adam Jackson <ajax@benzedrine.nwnk.net> 174271Date: Thu Aug 17 17:29:32 2006 -0400 174272 174273 Un-cut-and-paste the mode rejection message. 174274 174275commit 43e42eef1f5a22703eb64fc9cffecde036ea38e0 174276Author: Adam Jackson <ajax@benzedrine.nwnk.net> 174277Date: Thu Aug 17 16:22:07 2006 -0400 174278 174279 Fix default mouse device on Linux, again. 174280 174281 It would be really nice if we only did this in one place instead of 40. 174282 174283commit 633b6a69f560c0a77dcff78fdef5fcf0041e2e04 174284Merge: 95dbfcf88 7da51447e 174285Author: Daniel Stone <daniel@fooishbar.org> 174286Date: Thu Aug 17 21:25:14 2006 +0300 174287 174288 Merge branch 'master' into input-hotplug 174289 174290commit 7da51447eaab34292e14077fb5a48e6b2e587781 174291Author: Daniel Stone <daniel@fooishbar.org> 174292Date: Thu Aug 17 21:24:07 2006 +0300 174293 174294 events.c: fix tiny XEvIE thinko 174295 Make sure xeviehot gets updated after the ConfineToShape() call. 174296 174297commit 5d082f05632906c29296a44ef5c3a4962c0cbe62 174298Author: Daniel Stone <daniel@fooishbar.org> 174299Date: Thu Aug 17 21:18:18 2006 +0300 174300 174301 events.c: make XEvIE a little less verbose 174302 Change a lot of: 174303 #ifdef XEVIE 174304 xeviehot.x = 174305 #endif 174306 sprite.hot.x = ... 174307 #ifdef XEVIE 174308 xeviehot.y = 174309 #endif 174310 sprite.hot.y = ... 174311 to one single 174312 #ifdef XEVIE 174313 xeviehot.x = sprite.hot.x; 174314 xeviehot.y = sprite.hot.y; 174315 #endif 174316 at the end of the functions. 174317 174318commit 95dbfcf8828c041c218145afc87d21a6c9c7bc02 174319Author: Daniel Stone <daniel@fooishbar.org> 174320Date: Thu Aug 17 21:18:18 2006 +0300 174321 174322 events.c: make XEvIE a little less verbose 174323 Change a lot of: 174324 #ifdef XEVIE 174325 xeviehot.x = 174326 #endif 174327 sprite.hot.x = ... 174328 #ifdef XEVIE 174329 xeviehot.y = 174330 #endif 174331 sprite.hot.y = ... 174332 to one single 174333 #ifdef XEVIE 174334 xeviehot.x = sprite.hot.x; 174335 xeviehot.y = sprite.hot.y; 174336 #endif 174337 at the end of the functions. 174338 174339commit c6c39afde3e5f43b623ca6b52162b83c98a28d45 174340Author: Daniel Stone <daniel@fooishbar.org> 174341Date: Thu Aug 17 21:13:09 2006 +0300 174342 174343 dix/events.c: add YAFIXME 174344 Add another FIXME to the cacaphony of XXX and FIXMEs in this file. 174345 174346commit f9624e0109cf12b6af43fb4235aaa0b54340a4bb 174347Author: Daniel Stone <daniel@fooishbar.org> 174348Date: Thu Aug 17 16:09:51 2006 +0300 174349 174350 kdrive/input: verify SIGIO with --enable-debug 174351 174352commit 73e58adda96c1d1b5176d819107faa7697c3eb94 174353Author: Matthias Hopf <mhopf@suse.de> 174354Date: Wed Aug 16 18:17:58 2006 +0200 174355 174356 Fixed segfault w/ broken Xinerama configs. 174357 174358commit cd3f744b1f983f71476db99c050045d981c5f5b2 174359Author: Daniel Stone <daniel@fooishbar.org> 174360Date: Tue Aug 15 15:54:13 2006 +0300 174361 174362 fix missing brace, trim unused variables 174363 174364commit 5d073697adb3864133fa3221b82ab8d2f4a59758 174365Author: Daniel Stone <daniel@fooishbar.org> 174366Date: Tue Aug 15 15:37:10 2006 +0300 174367 174368 kdrive/input: minor warning cleanups 174369 And also a compiler error fix when VERIFY_SIGIO is defined. 174370 174371commit 47c1c948e69cfba950ad37a3133fa2db0bd0ff2c 174372Author: Daniel Stone <daniel@fooishbar.org> 174373Date: Tue Aug 15 15:25:16 2006 +0300 174374 174375 kdrive/input: only run special key behaviours on non-XKB 174376 Only attempt to manually deal with special key behaviours (e.g. terminating 174377 the server) when not using XKB, and leave locking behaviour up to GKVE. 174378 174379commit 34228d8b280ef105a0c60b8de5dacf70a5ce24b5 174380Author: Daniel Stone <daniel@fooishbar.org> 174381Date: Tue Aug 15 15:23:53 2006 +0300 174382 174383 GPE: fix absolute button events / GKVE: (non-XKB) don't repeat lock keys 174384 Fix absolute button events in GPE, where we would previously send valuator 174385 events without bumping numEvents accordingly, causing the core event to 174386 go missing. 174387 In the non-XKB path in GKVE, implement proper lock behaviour (one press to 174388 enable, one press to disable, discard releases). 174389 Fix debug_events prototype. 174390 174391commit d003bada3352ec7d734498c4c732904876a9d1e2 174392Merge: d6433be3c a815b9b99 174393Author: Daniel Stone <daniel@fooishbar.org> 174394Date: Sat Aug 12 22:48:55 2006 +0300 174395 174396 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 174397 174398commit a815b9b990e068f02d9cbba2b17f2cc3a30a9310 174399Merge: 37943e2f1 984babe86 174400Author: George Sapountzis <gsap7@yahoo.gr> 174401Date: Sat Aug 12 21:58:33 2006 +0300 174402 174403 Merge branch 'master' of git+ssh://gsap7@git.freedesktop.org/git/xorg/xserver 174404 174405commit d6433be3cca807dd78fbb1f45d9ba0212283083d 174406Merge: 2bf9e3dc1 984babe86 174407Author: Daniel Stone <daniel@fooishbar.org> 174408Date: Sat Aug 12 21:50:52 2006 +0300 174409 174410 Merge branch 'master' into input-hotplug 174411 174412commit 2bf9e3dc1ec5fd7bf84a4a96899e5663a721d4a4 174413Author: Daniel Stone <daniel@fooishbar.org> 174414Date: Sat Aug 12 21:50:39 2006 +0300 174415 174416 make DIX more tolerant of devices without a CtrlProc (Debian #269860) 174417 Return BadDevice on client requests for devices without a CtrlProc, instead 174418 of tanking horribly. 174419 174420commit 984babe86bf82002b4d6589b2750c7b5a5489bd5 174421Author: Daniel Stone <daniel@fooishbar.org> 174422Date: Sat Aug 12 21:43:38 2006 +0300 174423 174424 remove obsolete vendor defines 174425 Remove random behaviour changes for SGI and MetroLink. 174426 174427commit 37943e2f1abc6709ff739000372b0394d5cd18c5 174428Author: George Sapountzis <gsap7@yahoo.gr> 174429Date: Sat Aug 12 20:54:33 2006 +0300 174430 174431 Call exaTryComponentAlphaHelper() for solid src also. 174432 174433 Also, rename to exaTryMagicTwoPassCompositeHelper() as it is now called for 174434 non-component-alpha masks also, and add function description from 174435 http://anholt.livejournal.com/32058.html. 174436 174437commit f7919c287936f55569c2301ebb1b5f52358e70fa 174438Author: Bastian Blank <waldi@debian.org> 174439Date: Sat Aug 12 20:43:25 2006 +0300 174440 174441 xfree86: don't do legacy IO on ARM or S/390 (Debian #362641) 174442 Don't attempt to poke legacy IO ranges on ARM or S/390. 174443 174444commit 59dcc62906d8ee597cd43aa307f414cb47995cea 174445Author: Daniel Stone <daniel@fooishbar.org> 174446Date: Sat Aug 12 20:39:08 2006 +0300 174447 174448 xfree86: remove Xqueue support completely 174449 174450commit e641000b98e7c2e92e3c801eaa42aa15d5c16ad0 174451Author: Samuel Thibault <samuel.thibault@ens-lyon.org> 174452Date: Sat Aug 12 19:41:59 2006 +0300 174453 174454 xfree86: add Hurd support (#5613) 174455 Add support for GNU/Hurd. 174456 174457commit 5a3488ccac8e5dabd9fc98bc41ef178ead1b2faf 174458Author: Daniel Stone <daniel@fooishbar.org> 174459Date: Sat Aug 12 19:25:06 2006 +0300 174460 174461 configure.ac: fix execinfo.h test (Debian #363218) 174462 Define HAVE_EXECINFO_H as well as HAVE_BACKTRACE, when we find execinfo.h. 174463 174464commit 26c3cd1c9e3f52548389817a6d89a377e20c4269 174465Merge: 008aa7eb6 c4951e0a6 174466Author: Daniel Stone <daniel@fooishbar.org> 174467Date: Sat Aug 12 18:58:18 2006 +0300 174468 174469 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 174470 174471commit 008aa7eb6ed090384e6c44f861410e317c78a1cd 174472Author: Daniel Stone <daniel@fooishbar.org> 174473Date: Sat Aug 12 18:56:05 2006 +0300 174474 174475 completely remove OS keyboard layer 174476 Completely axe the keyboard layer from os-support. 174477 174478commit c4951e0a6b6cf3eeee710cc5cda1d9bc929ee3d7 174479Author: Adam Jackson <ajax@benzedrine.nwnk.net> 174480Date: Thu Aug 10 20:49:06 2006 -0400 174481 174482 Fix a mode sanity check to not break reduced-blanking setups (LCDs). 174483 174484commit e1921f014b102e3eecf3b41972f8672cf23264d6 174485Author: Adam Jackson <ajax@benzedrine.nwnk.net> 174486Date: Thu Aug 10 20:43:15 2006 -0400 174487 174488 Rename some mode tokens to better reflect their use. 174489 174490 Per #5386, M_T_EDID -> M_T_DRIVER, since it's really for any driver-detected 174491 mode. Also add M_T_PREFERRED bit, to select a 'best' mode out of a set. 174492 174493commit c2535f67923bde0bfb0e72363467110806e2f40f 174494Merge: c0cb8d1fb db82e12fa 174495Author: Alan Coopersmith <alan.coopersmith@sun.com> 174496Date: Thu Aug 10 10:37:59 2006 -0700 174497 174498 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into XACE-modular 174499 174500commit 9525c2709ea3245c6518d4c3b5a0a4afff37181d 174501Author: Daniel Stone <daniel@fooishbar.org> 174502Date: Thu Aug 10 20:29:57 2006 +0300 174503 174504 configure.ac: fix xephyr conditionals 174505 Fix Xephyr build conditions, allowing it to actually be disabled. 174506 174507commit 6d8d4abaaacf08140b673472d985117d448a62e7 174508Author: Daniel Stone <daniel@fooishbar.org> 174509Date: Thu Aug 10 20:28:06 2006 +0300 174510 174511 configure.ac: allow conditional building of XF86{DGA,Misc,VidMode} 174512 Allow conditional building of the above three extensions, defaulting to 174513 auto. 174514 174515commit 1a5561c4ecccaf32b03c41373adf376100d457aa 174516Merge: 190f229ed db82e12fa 174517Author: Ian Romanick <idr@umwelt.(none)> 174518Date: Thu Aug 10 10:20:37 2006 -0700 174519 174520 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 174521 174522commit cec284f2b3e948deb9e56a1a8519fddf693ab952 174523Author: Daniel Stone <daniel@fooishbar.org> 174524Date: Thu Aug 10 18:03:58 2006 +0300 174525 174526 kdrive: properly ifdef composite enabling 174527 174528commit 45bce556e8665412b9f6e89f88ed5bedb41de1ba 174529Author: Daniel Stone <daniel@fooishbar.org> 174530Date: Thu Aug 10 18:02:47 2006 +0300 174531 174532 GetMaximumEventsNum: be more conservative 174533 Be slightly more conservative in our maximum event count if we're using 174534 XKB (and thus don't need to count the extra repeat events). 174535 174536commit 172d45b9b75f95c997d1e9358040eead496e2a06 174537Merge: 3832a3d7d db82e12fa 174538Author: Daniel Stone <daniel@fooishbar.org> 174539Date: Thu Aug 10 14:14:54 2006 +0300 174540 174541 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 174542 174543commit 3832a3d7db0f3b5d7167e3b3b5ea6d6b3d96351a 174544Author: Daniel Stone <daniel@fooishbar.org> 174545Date: Thu Aug 10 14:13:51 2006 +0300 174546 174547 GKVE: don't repeat modifiers when using XKB 174548 Make sure we don't ever repeat modifiers (previously was repeating when 174549 using XKB); only do explicit KP/KR repeats in the non-XKB case. XKB will 174550 take care of repeating when we're using it. 174551 174552commit 9f188416bb6b4837d4c3f8773053d5eee0ff0ee1 174553Author: Daniel Stone <daniel@fooishbar.org> 174554Date: Thu Aug 10 14:00:34 2006 +0300 174555 174556 core devices: clear devicePrivates on close 174557 174558commit 539d1f33475484d35fb5a377efc76dba2d868e3f 174559Author: Daniel Stone <daniel@fooishbar.org> 174560Date: Thu Aug 10 14:00:14 2006 +0300 174561 174562 GKVE/GPE: have DDX allocate events 174563 Don't allocate events on every GKE/GKVE/GPE call, just have the DDX manage 174564 it instead. Introduce GetMaximumEventsNum(), which is the maximum number 174565 of events these functions will ever produce. 174566 174567commit db82e12fac5eaa16a39fc1bd0bc31ad95089dc95 174568Author: Adam Jackson <ajax@benzedrine.nwnk.net> 174569Date: Wed Aug 9 14:55:17 2006 -0400 174570 174571 Remove TargetRefresh option from the autoconfig logic. 174572 174573 The default target of 75Hz is almost always wrong for LCDs. 174574 174575commit fcd4167e8913f77bdf9e17a6955d0f2a9f4eeb10 174576Author: Adam Jackson <ajax@benzedrine.nwnk.net> 174577Date: Wed Aug 9 14:48:51 2006 -0400 174578 174579 Remove the bc flag from the -help text, since it's gone. 174580 174581commit 767f372dd02232469f9fd804b811a17eaf762e1e 174582Merge: c4f5de6cc 462bb61b0 174583Author: Tilman Sauerbeck <tilman@code-monkey.de> 174584Date: Wed Aug 9 20:23:30 2006 +0200 174585 174586 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 174587 174588commit c4f5de6cc3b935025829af971b0b8010c1ecfedb 174589Author: Tilman Sauerbeck <tilman@code-monkey.de> 174590Date: Wed Aug 9 20:21:52 2006 +0200 174591 174592 Sanitized glxdri's Block/Wakeuphandler calling. 174593 174594 __glXDRIleaveServer() and _enterServer() used to call DRIDoBlockHandler 174595 (resp DRIDoWakeupHandler) directly. They are now calling DRIBlockHandler 174596 (resp DRIWakeupHandler) to account for driver specific block/wakeup 174597 hooks. 174598 174599commit 5506b4ad200745236f997c121e8200179c47b749 174600Merge: 4be9abb85 462bb61b0 174601Author: Daniel Stone <daniel@fooishbar.org> 174602Date: Wed Aug 9 07:21:01 2006 +0300 174603 174604 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 174605 174606commit 4be9abb8504b3761b5f3a01851e4eb3da86c76e2 174607Author: Daniel Stone <daniel@fooishbar.org> 174608Date: Wed Aug 9 07:20:16 2006 +0300 174609 174610 kdrive: remove ddx_DEPENDENCIES 174611 Remove foo_DEPENDENCIES as they weren't guaranteed to just be libs, 174612 and loader arguments (-lfoo, -Lfoo) might've crept in. 174613 174614commit 462bb61b0fe968fae1b99cf98ec6f7de09105dcd 174615Author: Aaron Plattner <aplattner@nvidia.com> 174616Date: Tue Aug 8 18:07:22 2006 -0700 174617 174618 Add CompositeRegisterAlternateVisuals. 174619 174620 This provides drivers the ability to add their own alternate visuals and then 174621 register them with Composite for implicit redirection. 174622 174623commit 190f229ed77d87797e0f2f6762c86b3ad3a3dcbe 174624Author: Ian Romanick <idr@umwelt.(none)> 174625Date: Tue Aug 8 16:47:32 2006 -0700 174626 174627 Make xf86PciVideoInfo static since it is only used within this file. 174628 174629 Dummy out all of the PCI bus and device access control functions. We 174630 need a better way to do this, and that will probably be in 174631 libpciaccess and / or the kernel. 174632 174633 Refactor xf86GetPciInfoForEntity to use pci_device_find_by_slot. 174634 174635 Refector xf86CheckPciSlot to use xf86GetPciEntity. 174636 174637 Eliminate disablePciBios and the one place that calls it. 174638 174639commit 5508f7646f9754b054e961a1025b7a52913b563b 174640Author: Ian Romanick <idr@umwelt.(none)> 174641Date: Tue Aug 8 16:43:31 2006 -0700 174642 174643 Remove prototypes from some functions that were previously removed. 174644 174645commit 9df53d903ed68073bf7d2c2a275b6f6556a85c0f 174646Author: Ian Romanick <idr@umwelt.(none)> 174647Date: Tue Aug 8 16:42:23 2006 -0700 174648 174649 Rename xf86ReadDomainMemory to xf86ReadLegacyVideoBIOS, since that's 174650 what it is actually used for. Modify a few routines in linuxPci.c to 174651 take pci_device structures as parameters in stead of PCITAGs. 174652 174653commit fe351a711ef55c3ae1e784d4551147c080eda109 174654Author: Daniel Stone <daniel@fooishbar.org> 174655Date: Tue Aug 8 14:54:10 2006 +0300 174656 174657 GKVE: send XkbMapNotify, not XkbNewKeyboardNotify 174658 Sending MapNotify is more correct in this case than NKN, so do that. 174659 174660commit 31089816317f27c668b12a15c74fdd226a8df9f7 174661Merge: ab3ebfefd 12dbd8a02 174662Author: Daniel Stone <daniel@fooishbar.org> 174663Date: Tue Aug 8 12:01:12 2006 +0300 174664 174665 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 174666 174667commit ab3ebfefdb7d21aba72a5030f6f93bf4fafed709 174668Author: Tiago Vignatti <tv02@c3sl.ufpr.br> 174669Date: Tue Aug 8 12:00:34 2006 +0300 174670 174671 xorgcfg: fix compilation error 174672 Add missing parentheses to IS_KBDDRIV macros. 174673 174674commit 12dbd8a02f797ad57a1be683a02a1fcb1ca35438 174675Author: Daniel Stone <daniel@fooishbar.org> 174676Date: Mon Aug 7 23:43:40 2006 +0300 174677 174678 remove optional R3 backwards compatibility 174679 Remove the permitOldBugs flag, which enabled backwards compatbility with 174680 broken R2/R3 era clients. 174681 174682commit 7721ee308fbbb6fc9c969f15fe04b3346c04f843 174683Author: Daniel Stone <daniel@fooishbar.org> 174684Date: Mon Aug 7 23:03:02 2006 +0300 174685 174686 config client: fix minor race with event queue 174687 Fix a small race whereby you could remove a device while events from it 174688 were still in the queue, by calling ProcessInputEvents immediately before 174689 RemoveDevice, to (hopefully) flush the event queue. 174690 174691commit 458c63a84110f64c7fce397a531a3a779c2239a2 174692Author: Daniel Stone <daniel@fooishbar.org> 174693Date: Mon Aug 7 23:02:17 2006 +0300 174694 174695 enable adding devices after removal of a middle device 174696 Allow new devices to be added after a device that _wasn't_ the last on the 174697 list was removed, by ensuring inputInfo.numDevices always increases, and 174698 never decreases. 174699 174700commit a31d11a7a8485cdc799f76c4d407d3b7f7c9b350 174701Author: Daniel Stone <daniel@fooishbar.org> 174702Date: Mon Aug 7 23:01:23 2006 +0300 174703 174704 GKVE: get rid of bogus double-release check 174705 Get rid of a bogus double-release check which broke non-XKB servers. 174706 174707commit bedc4ecf23c7150e3156e0d24602ed3bc3977225 174708Author: Daniel Stone <daniel@fooishbar.org> 174709Date: Mon Aug 7 23:00:45 2006 +0300 174710 174711 xephyr: aid input debugging 174712 Add the 'ephyr' mouse and keyboard drivers to the driver list so we can 174713 re-add devices. 174714 Set the names properly in Ephyr{Keyboard,Mouse}Init, not in InitInput. 174715 174716commit baf93b3abe1e88d82ee6a3d6939f50f96ded271a 174717Author: Daniel Stone <daniel@fooishbar.org> 174718Date: Mon Aug 7 21:12:45 2006 +0300 174719 174720 kdrive: move map initialisation to KdNewPointer 174721 Do a linear n -> n initialisation on the map up until KD_MAX_BUTTON in 174722 KdNewPointer, moving it out of both KdParsePointer, and KdPointerProc. 174723 Also remove dead pointer acceleration code. 174724 174725commit ccb53340b66a778abf10182fd88a7d699207fb84 174726Author: Daniel Stone <daniel@fooishbar.org> 174727Date: Mon Aug 7 21:12:00 2006 +0300 174728 174729 ephyr: cleanup 174730 Remove extraneous KdAddPointerDriver call. 174731 174732commit d1c18af27e0aed73104743afb4bf4b8d3d1186cf 174733Author: Daniel Stone <daniel@fooishbar.org> 174734Date: Mon Aug 7 21:11:38 2006 +0300 174735 174736 GPE: use button map for DBP/DBR, not just BP/BR 174737 Make sure we use the button map for extended events, not just core. 174738 174739commit 9b7ecbd1dd8d092221897e29c85f3306c7367716 174740Author: Daniel Stone <daniel@fooishbar.org> 174741Date: Mon Aug 7 21:09:32 2006 +0300 174742 174743 kdrive: prevent overrun in map 174744 We actually need n + 1 elements for the mouse button map, not n. 174745 174746commit eb6e8d4042252b13328dbb122e0e6186796a80ac 174747Author: Daniel Stone <daniel@fooishbar.org> 174748Date: Mon Aug 7 21:05:37 2006 +0300 174749 174750 kdrive: increase maximum number of buttons 174751 Increase KD_MAX_BUTTONS to 32. 174752 174753commit 1c72290cdf4d9b214e1b9c0526cb7cb8641051f3 174754Author: Aaron Plattner <aplattner@nvidia.com> 174755Date: Mon Aug 7 09:57:58 2006 -0700 174756 174757 Use DrawablePtrs instead of PixmapPtrs for Prepare/Finish access. 174758 174759 Also, define some wfb functions even if FB_ACCESS_WRAPPER is not defined. This allows a client to use libfb and libwfb at the same time. 174760 174761commit afcad4ad99bbfc8bdcd0f4fdd70e072108410d30 174762Author: Daniel Stone <daniel@fooishbar.org> 174763Date: Mon Aug 7 18:11:05 2006 +0300 174764 174765 xfree86 ddx: always free GKE/GPE events 174766 free() events we get passed back from GKE and GPE so we don't just, er, 174767 leak them all. *cough*. 174768 174769commit 98fdf874eeadd5b37413922d8afba8415d0c56bb 174770Author: Daniel Stone <daniel@fooishbar.org> 174771Date: Mon Aug 7 16:51:39 2006 +0300 174772 174773 move all autorepeat logic to DIX 174774 Move core autorepeat logic for keyboards down to the DIX, remove it from 174775 KDrive. 174776 174777commit 5c7001fef8ffc6e3d8585a37d3f79a9495be8ed0 174778Author: Daniel Stone <daniel@fooishbar.org> 174779Date: Mon Aug 7 16:51:09 2006 +0300 174780 174781 memcpy() events in 174782 memcpy events into our event structure instead of doing pointer assignment. 174783 174784commit c85e64cba1d2d88f676ca7cf23b52a6f8219e90e 174785Merge: a406f6bfe f54b71b77 174786Author: Daniel Stone <daniel@fooishbar.org> 174787Date: Mon Aug 7 15:54:55 2006 +0300 174788 174789 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 174790 174791commit f54b71b772a1f587394ae3968782b611e52f0e2d 174792Author: David Nusinow <david@nee.(none)> 174793Date: Sun Aug 6 18:11:00 2006 +0000 174794 174795 Document enable/disable flag for AIGLX in xorg.conf manpage. 174796 174797commit a406f6bfeaa46e3236f7ab46813fe6c30b936a35 174798Author: Daniel Stone <daniel@fooishbar.org> 174799Date: Fri Aug 4 12:40:19 2006 +0300 174800 174801 mieq: don't leak events 174802 free all events posted through mieqEnqueue. 174803 174804commit 997ba45b192f21810099ed888792a45f1677a9ce 174805Author: Daniel Stone <daniel@fooishbar.org> 174806Date: Fri Aug 4 11:18:16 2006 +0300 174807 174808 fix incorrect button test 174809 Test for n (1..nButtons) being under nButtons, not button (1..(1<<nButtons)). 174810 174811commit 9138d5a51e411f598bc0e75a3e73d2c16187a518 174812Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 174813Date: Thu Aug 3 18:24:04 2006 -0400 174814 174815 Make SecurityLookupIDBy* part of the base functionality. 174816 174817commit 45c229f526bf1dafb5e81b50d700449ba4e1613d 174818Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 174819Date: Thu Aug 3 14:26:06 2006 -0400 174820 174821 Remove LBX code. 174822 174823commit 96e45626c43b7674b66e0258b0b1730d5ce71357 174824Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 174825Date: Wed Aug 2 20:29:59 2006 -0400 174826 174827 Rebase Security extension to use devPrivates for storing security state. 174828 174829commit 83ebf61ec03ff31005375900bee1e55e0e694c4b 174830Author: Ian Romanick <idr@umwelt.(none)> 174831Date: Wed Aug 2 13:33:33 2006 -0700 174832 174833 Eliminate the last few places outside xf86pciBus.c that use 174834 xf86PciVideoInfo. In all cases this means converting the code to use 174835 the pci_device_iterator interfaces. 174836 174837commit 5bfb2ee9652e5ef6d076ef8b6df54baaa43b8e0c 174838Author: Ian Romanick <idr@umwelt.(none)> 174839Date: Wed Aug 2 13:29:21 2006 -0700 174840 174841 Remove the last remants of the pci{Read,Write}{Long,Word,Byte} 174842 functionality. This also allows the removal of PCI_CPU, PCI_CPU16, 174843 and a few other dangling bits of cruft. 174844 174845commit a0f2e1cae46f189ed97e2d92b485d315b3d2627a 174846Author: Ian Romanick <idr@umwelt.(none)> 174847Date: Wed Aug 2 10:54:41 2006 -0700 174848 174849 Move xf86FindPciDeviceVendor and xf86FindPciClass from xf86pciBus.c to 174850 xf86int10.c. Refactor common code from those functions to do_find. 174851 174852commit de8234606f87ce79d016f7ddeabdf57c4ad212f3 174853Author: Ian Romanick <idr@umwelt.(none)> 174854Date: Wed Aug 2 10:44:11 2006 -0700 174855 174856 Modify xf86FindPciDeviceVendor and xf86FindPciClass to use the 174857 pci_interator interface instead of the xf86PciVideoInfo array. 174858 174859commit 3c23dec5962b8b81ae838fe0ee2c7b0a789f5386 174860Author: Eamon Walsh <ewalsh@tycho.nsa.gov> 174861Date: Wed Aug 2 13:39:49 2006 -0400 174862 174863 Call ClientStateCallback on serverClient devPrivates initialization. 174864 174865commit 23f44df9009023e77508f03ac4a7595c7e3d40cb 174866Author: Ian Romanick <idr@umwelt.(none)> 174867Date: Wed Aug 2 09:55:32 2006 -0700 174868 174869 Remove pciConfigPtr and all of the associated cruft. 174870 174871commit 966ebd3d2d84b440e89504d055a0e937303ed11d 174872Merge: a94176627 39169fd37 174873Author: Ian Romanick <idr@umwelt.(none)> 174874Date: Wed Aug 2 08:14:58 2006 -0700 174875 174876 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 174877 174878commit ee02e647882a4be29e1130bd79904ee79ed6b802 174879Author: Aaron Plattner <aplattner@nvidia.com> 174880Date: Tue Aug 1 13:45:43 2006 -0700 174881 174882 Wrap libwfb memory access. 174883 174884 Use the READ and WRITE macros to wrap memory accesses that could be in video 174885 memory. Add MEMCPY_WRAPPED and MEMSET_WRAPPED macros to wrap memcpy and 174886 memset, respectively. 174887 174888commit 39169fd373b97f34923f6494d697d9429d0b8aa3 174889Author: Matthew Allum <mallum@polystyrene.(none)> 174890Date: Tue Aug 1 13:39:22 2006 +0100 174891 174892 Back out 'mystery' spurious host window hints. 174893 174894commit f737cc38baea6af8bf284c9e207e60a7d90eebe1 174895Author: Eamon Walsh <ewalsh@moss-uranus.epoch.ncsc.mil> 174896Date: Mon Jul 31 19:58:42 2006 -0400 174897 174898 Adding devPrivates support to the ExtensionEntry structure. 174899 174900commit b04d64854712678701d5243aacf5cc93444cfadc 174901Author: Eamon Walsh <ewalsh@moss-uranus.epoch.ncsc.mil> 174902Date: Mon Jul 31 19:35:08 2006 -0400 174903 174904 Added devPrivates support to the ExtensionEntry structure. 174905 174906commit a94176627cdd6122ffadb618995f9bcec872609a 174907Author: Ian Romanick <idr@umwelt.(none)> 174908Date: Mon Jul 31 14:33:58 2006 -0700 174909 174910 Eliminate all uses of the businfo field of pciDevice / pciConfigPtr. 174911 174912commit a69335dc299be6de8b82ed34de1cb30f1255feb4 174913Author: Aaron Plattner <aplattner@nvidia.com> 174914Date: Mon Jul 31 14:15:55 2006 -0700 174915 174916 Make ReadMemoryProcPtr take a const pointer. 174917 174918 Fixes some warnings when using READ with a const pointer. 174919 174920commit 26c9587f90806cce1c0d2a98e656e9c311a34ee5 174921Author: Ian Romanick <idr@umwelt.(none)> 174922Date: Mon Jul 31 13:54:36 2006 -0700 174923 174924 Build fixes for last pull from HEAD. 174925 174926commit 65dc25d8f86b962f2adc765a9ff327419f98092b 174927Merge: 69533fc04 b74c845a1 174928Author: Ian Romanick <idr@umwelt.(none)> 174929Date: Mon Jul 31 12:30:30 2006 -0700 174930 174931 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 174932 174933commit b74c845a1233f78b841ff8840272c50873300c20 174934Merge: 3112a6c4f 02daa6bb1 174935Author: Ian Romanick <idr@umwelt.(none)> 174936Date: Mon Jul 31 10:26:06 2006 -0700 174937 174938 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 174939 174940commit 3112a6c4f26d5e9258b8def7ce4109b4bd408c67 174941Author: Ian Romanick <idr@umwelt.(none)> 174942Date: Mon Jul 31 10:25:31 2006 -0700 174943 174944 Noting uses libdummy.a, so don't build it. Only libdummy-nonserver.a 174945 is actually used. 174946 174947commit 02daa6bb103e53e5a33db2bb6acbe57d0bf2c30e 174948Author: Matthew Allum <mallum@polystyrene.(none)> 174949Date: Mon Jul 31 17:32:05 2006 +0100 174950 174951 Improve XRes to; 174952 - Better estimate general pixmap memory usage. 174953 - Account for pixmaps shared between clients. 174954 - Account for window background and border pixmaps, 174955 and GC stripple and tile pixmaps. 174956 174957commit 69533fc04f38c67fd424776c5191a4ba57ba8c8f 174958Merge: ef1aecaaf 24051ef97 174959Author: Ian Romanick <idr@umwelt.(none)> 174960Date: Mon Jul 31 09:26:04 2006 -0700 174961 174962 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 174963 174964 Conflicts: 174965 174966 hw/xfree86/Makefile.am 174967 hw/xfree86/common/xf86AutoConfig.c 174968 174969commit 24051ef97406f28c102cf46a78223400b61fdae2 174970Author: Daniel Stone <daniel@fooishbar.org> 174971Date: Sun Jul 30 12:15:33 2006 +0300 174972 174973 remove filename that's too long for tar 174974 174975commit ecb7d43a76d507d04891ab7f189b23be5eccda51 174976Author: Daniel Stone <daniel@fooishbar.org> 174977Date: Sun Jul 30 11:52:41 2006 +0300 174978 174979 add sym.h to sources 174980 174981commit bf2d7499c84c94f228d03b21448f5688b3cda1a8 174982Author: Daniel Stone <daniel@fooishbar.org> 174983Date: Sun Jul 30 11:17:02 2006 +0300 174984 174985 add securitysrv.h 174986 174987commit e87e68634d8eb66ab783e2802e2d5d12ff1031be 174988Author: Daniel Stone <daniel@fooishbar.org> 174989Date: Sun Jul 30 11:11:59 2006 +0300 174990 174991 remove .cvsignores from EXTRA_DIST 174992 174993commit ed0c807de9f07468385fcbd2e8a9c0737759a461 174994Author: Daniel Stone <daniel@fooishbar.org> 174995Date: Sun Jul 30 11:08:54 2006 +0300 174996 174997 bump to 1.1.99.3 174998 174999commit a68dc013a33d867e65a7e76b3eec5947b862a5b4 175000Author: Daniel Stone <daniel@fooishbar.org> 175001Date: Sun Jul 30 11:08:47 2006 +0300 175002 175003 remove README (which doesn't exist) from EXTRA_DIST 175004 175005commit 87fe85f38b6f781bf0e2eb555526e3d77779f9fa 175006Merge: 3518e2d0d 654619d76 175007Author: Daniel Stone <daniel@fooishbar.org> 175008Date: Sun Jul 30 10:51:34 2006 +0300 175009 175010 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 175011 175012commit 654619d76c779606f2315782fc01d1410399fa3b 175013Author: Kevin E Martin <kem@freedesktop.org> 175014Date: Fri Jul 28 17:16:32 2006 -0400 175015 175016 Revert xkb changes that broke XkbGetKeyboard() 175017 175018commit 79016d4036786b091a9b9d1133a6cdfedd6c277d 175019Author: Luc Verhaegen <libv@skynet.be> 175020Date: Fri Jul 28 16:02:02 2006 -0400 175021 175022 Bug #5386 (partial): Move CVT mode generator from cvt(1) to server core, and 175023 export it from the X server to modules. 175024 175025commit e6ae1612be519ee6224d354244d076d85d44a750 175026Author: Aaron Plattner <aplattner@nvidia.com> 175027Date: Thu Jul 27 18:24:59 2006 -0700 175028 175029 Add fbHasVisualTypes and fbSetVisualTypesAndMasks to wfbrename.h and include -DXFree86Server. 175030 175031commit 2a4ceb09ed5a09dc5763754ab865ec23df91ac9f 175032Author: Adam Jackson <ajax@benzedrine.nwnk.net> 175033Date: Wed Jul 26 19:39:17 2006 -0400 175034 175035 Remove dead function prototypes. 175036 175037commit 990a4009057e068f41d20b95aa0c59357185650d 175038Author: Adam Jackson <ajax@benzedrine.nwnk.net> 175039Date: Wed Jul 26 19:03:39 2006 -0400 175040 175041 Remove getconfig horrorshow. Replace with a static built-in rule list for now. 175042 175043commit 377a581ddf5e428a368efb1b59fcb317666fecdd 175044Author: Aaron Plattner <aplattner@nvidia.com> 175045Date: Tue Jul 25 15:27:31 2006 -0700 175046 175047 Switch to using void* pointers. 175048 175049 Pass the size of the data pointed to by src or dst as an argument to 175050 wfb{Read,Write}Memory. This allows one set of macros to be used with any size 175051 pointer. Assumes that sizeof(FbBits) >= sizeof(FbStip). 175052 175053commit a4005c15fbb48231cb958c32b2c791a2d23a135a 175054Author: Aaron Plattner <aplattner@nvidia.com> 175055Date: Mon Jul 10 18:58:09 2006 -0700 175056 175057 Add framebuffer access wrapper infrastructure. 175058 175059 Create fbPrepareAccess macros to call into the driver to set up the 175060 wfbReadMemory and wfbWriteWemory pointers. Call these from fbGetDrawable and 175061 fbGetStipDrawable. 175062 175063 Add the READ and WRITE macros, which expand to simple memory accesses for fb, 175064 and calls through the function pointers for wfb. 175065 175066 Add fbFinishAccess macro to give the driver an opportunity to clean up. Add 175067 calls to this in the appropriate places. 175068 175069commit 319efac445cebda5a2ac1db67efebe54bc47ba9d 175070Author: Aaron Plattner <aplattner@nvidia.com> 175071Date: Fri Jul 7 18:45:30 2006 -0700 175072 175073 Prefix all of the exported symbols in libwfb.so with "wfb". 175074 175075 For now, just #define all of the exported symbols in wfbrename.h. Later, 175076 we should add FBPREFIX() around the exported symbols and use -fvisiblity=hidden 175077 to hide the rest. 175078 175079commit 7608a63ff7409f399c9a26962a304b84196a1868 175080Author: Aaron Plattner <aplattner@nvidia.com> 175081Date: Thu Jul 6 17:05:21 2006 -0700 175082 175083 Build infrastructure for libwfb.so. 175084 175085 Builds fb/* twice, defining FB_ACCESS_WRAPPER for libwfb.la. Define a macro, 175086 FBPREFIX(X) which expands to fbX for libfb.la and wfbX for libwfb.la. Use the 175087 macro on [w]fbModuleData so the new module loads. 175088 175089commit 39158e98acb29e97a2682d4a37385f9141b484c4 175090Author: Adam Jackson <ajax@benzedrine.nwnk.net> 175091Date: Wed Jul 26 18:39:28 2006 -0400 175092 175093 Remove another latent PowerMAX hunk. 175094 175095 All your favorite running jokes of 2005, today! 175096 175097commit ef1aecaaf6cd7e9e4f9d0c6373664e774500bc13 175098Author: Ian Romanick <idr@umwelt.(none)> 175099Date: Wed Jul 26 13:25:13 2006 -0700 175100 175101 Pass correct pointer to xf86MapLegacyIO. 175102 175103commit 3518e2d0debc97e2bacdefe604b280e7fdfdd216 175104Merge: eb7733a48 3821f6aea 175105Author: Daniel Stone <daniel@fooishbar.org> 175106Date: Wed Jul 26 11:29:21 2006 +0300 175107 175108 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 175109 175110commit eb7733a48a92405660d5d2ab60913b62c30daaed 175111Author: Daniel Stone <daniel@fooishbar.org> 175112Date: Wed Jul 26 11:28:45 2006 +0300 175113 175114 kdrive: drop excessive NewInputDeviceRequest debugging 175115 175116commit 3821f6aeaa714582ee0a631de96c6e7cfd96303e 175117Author: Kristian Høgsberg <krh@redhat.com> 175118Date: Wed Jul 26 01:56:02 2006 -0400 175119 175120 Unlibc-wrap DMX glxscreens.c and fix tag-removal typo. 175121 175122commit 28ba8d56912ae56d7b9835188f621b0a491add30 175123Author: Ian Romanick <idr@umwelt.(none)> 175124Date: Tue Jul 25 15:38:09 2006 -0700 175125 175126 Gut anything having to do with the PciBusPtr (or PciBusRec) type. 175127 Bump required version of libpciaccess to 0.5.0 so that 175128 pci_device_get_bridge_buses can be used. 175129 175130commit 380b51d605a82c98082a2cebd70a0d1d0735eaa7 175131Merge: 82f6b7e75 8977b0743 175132Author: Ian Romanick <idr@umwelt.(none)> 175133Date: Tue Jul 25 11:30:04 2006 -0700 175134 175135 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 175136 175137 Conflicts: 175138 175139 hw/xfree86/common/xf86Init.c 175140 hw/xfree86/int10/pci.c 175141 hw/xfree86/scanpci/xf86PciData.h 175142 hw/xfree86/scanpci/xf86PciStdIds.h 175143 hw/xfree86/scanpci/xf86PciStr.h 175144 hw/xfree86/scanpci/xf86ScanPci.h 175145 hw/xfree86/utils/pcitweak/pcitweak.c 175146 hw/xfree86/utils/scanpci/scanpci.c 175147 175148 Re-removed most of the conflicting files. 175149 175150commit 4ea475924c557ad0819b796f5369e5e669465709 175151Author: Daniel Stone <daniel@fooishbar.org> 175152Date: Tue Jul 25 20:00:48 2006 +0300 175153 175154 re-add OpenedHand copyright 175155 Ae-add OpenedHand copyright, accidentally dropped in KDrive new world 175156 order patch. Sorry guys. 175157 175158commit 82f6b7e75e04e3fb951ab71ad3c553c8b34ea995 175159Author: Ian Romanick <idr@umwelt.(none)> 175160Date: Mon Jul 24 13:52:58 2006 -0700 175161 175162 Don't re-look-up pointers that are already stored in structures. 175163 175164commit 2fd6b995ff927c8ca7267c201f535564b82e691f 175165Author: Ian Romanick <idr@umwelt.(none)> 175166Date: Mon Jul 24 13:33:34 2006 -0700 175167 175168 Global s/xf86MapDomainIO/xf86MapLegacyIO/. Let's call a duck a duck, 175169 okay? Since xf86MapLegacyIO is called from only one place, cut the 175170 parameter list down to the one parameter that actually conveys some 175171 information: the one that gives a PCI device. Change from using a 175172 PCITAG to a pci_device. 175173 175174commit aed6fe0bb12c68b94e564252bc03594728ed8c5b 175175Author: Ian Romanick <idr@umwelt.(none)> 175176Date: Mon Jul 24 13:13:05 2006 -0700 175177 175178 Refactor linuxGetSizesStruct to take a pci_device pointer instead of a 175179 PCITAG. Modify xf86BusAccWindowsFromOS and xf86AccResFromOS to call 175180 linuxGetSizesStruct directly with a pci_device pointer. Remove 175181 linuxGetSizes. 175182 175183commit e48762799248eb7e16ea2c0df1561ae1430f2112 175184Author: Ian Romanick <idr@umwelt.(none)> 175185Date: Mon Jul 24 12:23:23 2006 -0700 175186 175187 Modify xf86BusAccWindowsFromOS and xf86AccResFromOS to use the 175188 libpciaccess interfaces. 175189 175190commit 4b474cbc1a0fe17da2438a15291f04ee67ea28f8 175191Author: Ian Romanick <idr@umwelt.(none)> 175192Date: Mon Jul 24 12:17:27 2006 -0700 175193 175194 The functions xf86PciBusAccWindowsFromOS and xf86BusAccWindowsFromOS 175195 are identical. Just have one call the other. 175196 175197commit 7dff8d98a9ed11f92806cb8c4b3f10e01c5d063f 175198Author: Ian Romanick <idr@umwelt.(none)> 175199Date: Mon Jul 24 11:56:37 2006 -0700 175200 175201 Remove source files that have been dead since day-1 of this branch. 175202 175203commit eb95128f615e2f3fab17dcafd4fc260c6d4d0644 175204Author: Ian Romanick <idr@umwelt.(none)> 175205Date: Mon Jul 24 11:53:37 2006 -0700 175206 175207 Changes that were missed on the previous commit. Perhaps oddness 175208 caused by doing git-update-index at the wrong time. 175209 175210commit 05f8b772aec2b05d92078bb60a07bd8413386363 175211Author: Ian Romanick <idr@umwelt.(none)> 175212Date: Mon Jul 24 11:23:03 2006 -0700 175213 175214 ScanPciDisplayPCICardInfo now uses pci_device interface for PCI 175215 scanning. Log messages simplified to make the code shorter and less 175216 convoluted. ScanPciDisplayPCICardInfo is now void since it was only 175217 called from one place with a constant parameter. 175218 175219commit ca3f4fc1b0c21a0620ab1eb35c199cd55d795095 175220Author: Daniel Stone <daniel@fooishbar.org> 175221Date: Sun Jul 23 19:02:12 2006 -0400 175222 175223 add fallback ChangeDeviceControl, allow XOpenDevice on closed device 175224 Add a fallback ChangeDeviceControl, which allows the attributes we know about 175225 so far. 175226 Allow XOpenDevice on closed devices. 175227 175228commit e73e5e2a4d8f22889d840a7719479f9af686cb9c 175229Merge: a73cef1f0 8977b0743 175230Author: Daniel Stone <daniel@fooishbar.org> 175231Date: Sat Jul 22 13:56:30 2006 -0400 175232 175233 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 175234 175235commit 8977b07434d75ca396d236dc1324f0c862b633c7 175236Author: Dave Airlie <airlied@linux.ie> 175237Date: Sun Jul 23 03:36:47 2006 +1000 175238 175239 glx: fix typo from tag removal 175240 175241commit a73cef1f005ca66db18e952e676ee5a21b829700 175242Merge: 672ca156b 70869fc67 175243Author: Daniel Stone <daniel@fooishbar.org> 175244Date: Sat Jul 22 12:07:22 2006 -0400 175245 175246 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 175247 175248commit 70869fc6777f87cd05794446caa739e3d9a91ffe 175249Author: Adam Jackson <ajax@theobromine.nwnk.net> 175250Date: Fri Jul 21 23:39:37 2006 -0400 175251 175252 Yet more dead code. 175253 175254commit 7c1b2ee7a8238c267bc97e78bbff204dc7723dd3 175255Author: Adam Jackson <ajax@theobromine.nwnk.net> 175256Date: Fri Jul 21 23:35:13 2006 -0400 175257 175258 static markup, more dead code. 175259 175260commit 1c4f90b1d05d4c49279f3224a6dd94850a6bd8d0 175261Author: Adam Jackson <ajax@theobromine.nwnk.net> 175262Date: Fri Jul 21 23:03:21 2006 -0400 175263 175264 Open-coded path checks make baby Jesus cry. 175265 175266commit 114264584ca43091a5e07282566a30a6378a1502 175267Author: Adam Jackson <ajax@theobromine.nwnk.net> 175268Date: Fri Jul 21 22:55:41 2006 -0400 175269 175270 Remove a useless open() of the module we're about to load. 175271 175272commit 985611d5cd079f97da700c7b8e898d33da004be0 175273Author: Adam Jackson <ajax@theobromine.nwnk.net> 175274Date: Fri Jul 21 22:37:59 2006 -0400 175275 175276 Delete some long-unused testing code. 175277 175278commit 525257eb98b46209c69c8576daa6bd7afc1f48a9 175279Author: Ian Romanick <idr@umwelt.(none)> 175280Date: Fri Jul 21 19:16:38 2006 -0700 175281 175282 Fix stupid typos. 175283 175284commit 672ca156bfb11440e6e234650bfba9d38e1edb52 175285Merge: d14d91f09 6cf844ab6 175286Author: Daniel Stone <daniel@fooishbar.org> 175287Date: Fri Jul 21 19:58:42 2006 -0400 175288 175289 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 175290 175291commit 6cf844ab69926b6d23619a12c97734af3881ba67 175292Author: Daniel Stone <daniel@fooishbar.org> 175293Date: Fri Jul 21 19:57:28 2006 -0400 175294 175295 loader: walk directory paths with readdir(), don't stat() everything 175296 Walk the directories with readdir, and don't stat everything we can 175297 find. Thanks to davej for the public humiliation reminding me to go back 175298 and re-fix this one. 175299 175300commit d14d91f094c3897c889f6aafb66d738820dae0aa 175301Author: Daniel Stone <daniel@fooishbar.org> 175302Date: Fri Jul 21 19:57:28 2006 -0400 175303 175304 loader: walk directory paths with readdir(), don't stat() everything 175305 Walk the directories with readdir, and don't stat everything we can 175306 find. Thanks to davej for the public humiliation reminding me to go back 175307 and re-fix this one. 175308 175309commit e18d34f4238e13e226b0407fa2f5f77d2038de39 175310Author: Ian Romanick <idr@umwelt.(none)> 175311Date: Fri Jul 21 16:47:45 2006 -0700 175312 175313 Make the various implementations of xf86ExtendedInitInt10 use the 175314 libpciaccess interfaces. This eliminates all calls to mapPciRom, 175315 which in turn allows the elimination of hw/xfree86/int10/pci.c. 175316 175317commit 87a6346bf7f086b5f98b2b2ecd52f27efe864e56 175318Merge: b73fb2ae3 0486d3966 175319Author: Daniel Stone <daniel@fooishbar.org> 175320Date: Fri Jul 21 19:36:25 2006 -0400 175321 175322 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 175323 175324commit 0486d3966d2888ef86d36c19f31bdbc2a3e8c652 175325Author: Daniel Stone <daniel@fooishbar.org> 175326Date: Fri Jul 21 19:35:04 2006 -0400 175327 175328 fix kbproto dependency 175329 Depend on kbproto >= 1.0.3, for unconditional definition of 175330 XkbSA_XFree86Private. 175331 175332commit b73fb2ae35a82e0bdd48f01132e971fb84946ff1 175333Merge: e7ac27ad8 aff404f29 175334Author: Daniel Stone <daniel@fooishbar.org> 175335Date: Fri Jul 21 19:30:26 2006 -0400 175336 175337 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 175338 175339commit e7ac27ad81efbea6128b3cec443ca98e228d14ad 175340Author: Daniel Stone <daniel@fooishbar.org> 175341Date: Fri Jul 21 19:29:28 2006 -0400 175342 175343 revert accidental deletion of lnx_io.c; re-delete lnx_kbd.c 175344 Thinko'd which file to remove after merging from master. 175345 175346commit 81913a12910e39d7ea6af8657c1c66cc6791cd65 175347Author: Daniel Stone <daniel@fooishbar.org> 175348Date: Fri Jul 21 19:10:26 2006 -0400 175349 175350 remove undead files from master 175351 Remove dead files which worked their way back in when merging from master. 175352 Ugh. 175353 175354commit 7465010d59ec435bd00b738f0cef766b352dc7eb 175355Merge: 1d31ed778 0aaac95b0 175356Author: Daniel Stone <daniel@fooishbar.org> 175357Date: Fri Jul 21 19:05:41 2006 -0400 175358 175359 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into input-hotplug 175360 175361commit 1d31ed778284082e1060bff63317c94581d9eb9b 175362Author: Daniel Stone <daniel@fooishbar.org> 175363Date: Fri Jul 21 19:02:52 2006 -0400 175364 175365 xephyr: load keysyms at init, not enable 175366 Load keysyms at init time, not enable, so we don't get the wrong map width. 175367 175368commit 63dfaa1d5ba556e09314ec914936e5471aab94b0 175369Author: Adam Jackson <ajax@theobromine.nwnk.net> 175370Date: Fri Jul 21 18:47:18 2006 -0400 175371 175372 Delete internal usage of the symbol ref/req lists. 175373 175374commit bca9364f3f4a2376edbcf57a34f704ce28be21ba 175375Author: Adam Jackson <ajax@theobromine.nwnk.net> 175376Date: Fri Jul 21 18:41:46 2006 -0400 175377 175378 Remove the loader's required and referenced symbol lists, dead code. 175379 175380commit 07ad92d2c4cb07db8487d76efc822fd7b88137cb 175381Merge: 18624a9b7 22db3fdb5 175382Author: Ian Romanick <idr@umwelt.(none)> 175383Date: Fri Jul 21 15:25:35 2006 -0700 175384 175385 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 175386 175387 Conflicts: 175388 175389 hw/xfree86/common/xf86DoScanPci.c 175390 hw/xfree86/common/xf86Init.c 175391 hw/xfree86/common/xf86pciBus.c 175392 175393commit aff404f293ed86a44a093a51a9f11e79e6c3f4f6 175394Author: Adam Jackson <ajax@theobromine.nwnk.net> 175395Date: Fri Jul 21 18:24:37 2006 -0400 175396 175397 Detach xf4bpp from cfb. 175398 175399commit 0aaac95b0d12089b256c97f6ff955c8c229ae095 175400Author: Adam Jackson <ajax@theobromine.nwnk.net> 175401Date: Fri Jul 21 17:56:00 2006 -0400 175402 175403 Remove RCS tags. Fix Xprint makefile braindamage. 175404 175405commit eeaad0e956640aac653d194a992df7792e4abcbb 175406Author: Aaron Plattner <aplattner@nvidia.com> 175407Date: Thu Jul 20 18:19:07 2006 -0700 175408 175409 Fix the RandR failure path for rotated screens. 175410 175411commit 22db3fdb54d2f7f6b72638b46c186af6db04e214 175412Merge: 2f98841fd 93cd53860 175413Author: Ian Romanick <idr@umwelt.(none)> 175414Date: Fri Jul 21 13:55:37 2006 -0700 175415 175416 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 175417 175418commit 3eeb62e8f587732e6b433c2b9c6879eb26a3f1b4 175419Author: Kristian Høgsberg <krh@redhat.com> 175420Date: Fri Jul 21 16:33:28 2006 -0400 175421 175422 bug #890: completely remove deprecated keyboard driver 175423 Remove all remnants of the old built-in keyboard driver. 175424 175425commit 60ea7b51fe2b8a19a08b63db48504971a7a50ee6 175426Author: Daniel Stone <daniel@fooishbar.org> 175427Date: Fri Jul 21 15:23:37 2006 -0400 175428 175429 xorg ddx: move to new input API, remove old keyboard driver 175430 Remove most of the rest of the old keyboard driver. 175431 Move to the new Get{Keyboard,Pointer}Events API, which is mostly 175432 complete at this stage: just missing the proximity events. 175433 175434commit d32dc8bf19e1071fc30af2f0bc6a6699b351f39e 175435Author: Daniel Stone <daniel@fooishbar.org> 175436Date: Thu Jul 20 18:40:47 2006 -0400 175437 175438 be more careful in IVAS 175439 175440 Don't walk off the end of a NULL pointer in InitValuatorAxisStruct. 175441 175442commit 7711c56d2e2aeae4dcd6d9297bc144c1cb3cfab1 175443Author: Daniel Stone <daniel.stone@nokia.com> 175444Date: Thu Jul 20 18:38:57 2006 -0400 175445 175446 xephyr: fix keymap, pointer; dix: fix multiple axes 175447 175448 Initialise our axes properly in the DIX, and make sure we don't 175449 unnecessarily clip maxval when it's not set. 175450 Fix keymap copying in Xephyr (to some degree: it's still broken), 175451 and set nAxes and nButtons properly. 175452 175453commit f18c3122a57df9770087e5fc70ac488552222233 175454Author: Daniel Stone <daniel.stone@nokia.com> 175455Date: Thu Jul 20 16:49:17 2006 -0400 175456 175457 sanitise debug output 175458 175459 Don't spit out huge chunks of pointless ephemera, especially without 175460 --enable-debug. 175461 175462commit 463e0fe35bec3c91b19be9aacf34babb146a88c9 175463Author: Daniel Stone <daniel.stone@nokia.com> 175464Date: Thu Jul 20 16:45:15 2006 -0400 175465 175466 update KDrive to new input API 175467 175468 Update KDrive to fit the new API (mieqInit and InitPointerDeviceStruct), and 175469 include InitTouchscreenDeviceStruct in the DIX. 175470 175471commit 1987af8c498a1bf394a8951ca6d5b0b7f7a35188 175472Author: Daniel Stone <daniel.stone@nokia.com> 175473Date: Thu Jul 20 16:39:54 2006 -0400 175474 175475 add virtual core devices to DIX 175476 175477 Add virtual core devices, with proper keymaps etc, to the DIX. 175478 175479commit 737e6e4836a6af26fedc22cda8e7d366b52c8fa7 175480Author: Kristian Høgsberg <krh@redhat.com> 175481Date: Wed Jul 19 20:10:48 2006 -0400 175482 175483 define SDevicePresenceNotifyEvent prototype 175484 175485commit 7f36cc533e7f6ae44e973c5f00f9bfec7c6b7b50 175486Author: Daniel Stone <daniel@fooishbar.org> 175487Date: Wed Jul 19 20:09:13 2006 -0400 175488 175489 distribute config.h 175490 175491commit e896195eab726a2b307200958308eda8c93dd3cd 175492Author: Daniel Stone <daniel.stone@nokia.com> 175493Date: Wed Jul 19 20:05:33 2006 -0400 175494 175495 remove extraneous font debugging code 175496 175497 Some of it didn't compile, and some of it was just unnecessary. 175498 175499commit 99c57674c002c5e88c6db34488a27b05004c9197 175500Author: Daniel Stone <daniel.stone@nokia.com> 175501Date: Wed Jul 19 19:59:11 2006 -0400 175502 175503 avoid using font servers with built-in fonts 175504 175505commit ecfad74c48f633916305bcc25baaaad74aa52b78 175506Author: Daniel Stone <daniel.stone@nokia.com> 175507Date: Wed Jul 19 19:55:13 2006 -0400 175508 175509 add support for built-in fonts 175510 175511 Use --enable-builtin-fonts to only use built-in fonts, and avoid loading 175512 fonts. 175513 175514commit 0a2068d123520d35818c38a555ae3ba06d8ca7fb 175515Author: Daniel Stone <daniel.stone@nokia.com> 175516Date: Wed Jul 19 17:29:23 2006 -0400 175517 175518 Xi: add XExtension{Keyboard,Pointer} types 175519 175520 Report XExtensionKeyboard for non-core keyboards, and XExtensionPointer for 175521 non-core pointers/mice. 175522 175523commit c7577f9b88aac84d59404f29d994ee7af583d33b 175524Author: Kristian Høgsberg <krh@redhat.com> 175525Date: Wed Jul 19 17:27:58 2006 -0400 175526 175527 Xi: add DevicePresenceNotify 175528 175529 Add support for DevicePresenceNotify events. 175530 175531commit 3a23e499017d5823157806029263edac53c663fd 175532Author: Daniel Stone <daniel.stone@nokia.com> 175533Date: Wed Jul 19 17:00:23 2006 -0400 175534 175535 make XInput mandatory 175536 175537 Always build Xi, since GetPointerEvents/GetKeyboardEvents relies on it. 175538 175539commit 02d09105113fb9b560a770fe15f7bb041165831c 175540Author: Daniel Stone <daniel.stone@nokia.com> 175541Date: Wed Jul 19 16:51:04 2006 -0400 175542 175543 new KDrive input world order 175544 175545 Convert KDrive to GPE/GKE interface. 175546 Add first-class drivers and enumerate every device separately through 175547 Xi, instead of lamely attempting to aggregate them. 175548 Add XKB support to the Linux keyboard driver. 175549 Add 'thumb button' support to the tslib driver. 175550 Rejig InitInput, so each DDX has to add a list of drivers it supports. 175551 Support NewInputDeviceRequest, et al. 175552 175553commit a274e7296b1bdd6f6c921f28b087610cec9548e0 175554Author: Daniel Stone <daniel.stone@nokia.com> 175555Date: Wed Jul 19 13:56:23 2006 -0400 175556 175557 add GetPointerEvents/GetKeyboardEvents framework 175558 175559 Add GetPointerEvents (with XFree86 pointer acceleration) and GetKeyboardEvents 175560 to the DIX. Extend the ValuatorClass structure to account for same. 175561 175562commit b308dbf273f8c26361b0fee7aca64aec3245f60b 175563Author: Daniel Stone <daniel.stone@nokia.com> 175564Date: Wed Jul 19 12:15:18 2006 -0400 175565 175566 add DEVICE_TOUCHSCREEN and DEVICE_CORE Xi controls (DeviceIntRec ABI break) 175567 175568 Add DEVICE_TOUCHSCREEN and DEVICE_CORE controls to the Xi code, and the 175569 TouchscreenClassRec and a coreEvents flag, to toggle propagation of core 175570 events. 175571 175572commit c9a3d9baa81ceb940032ffe529d9eadf2d202ab2 175573Author: Daniel Stone <daniel@fooishbar.org> 175574Date: Wed Jul 19 11:41:16 2006 -0400 175575 175576 xorg DDX: implement NewInputDeviceRequest 175577 175578 Implement NewInputDeviceRequest for Xorg, mainly written by Kristian Høgsberg. 175579 Move MatchInput to xf86Helper.c, as xf86LookupInputDriver. 175580 175581commit 02a95311568e24e1055ea52c7df8cb7aa3f38ad0 175582Author: Daniel Stone <daniel@fooishbar.org> 175583Date: Wed Jul 19 10:05:12 2006 -0400 175584 175585 add basic D-BUS configuration mechanism 175586 175587 Also move LookupDeviceIntRec into the DIX, and add InputOption type, and 175588 NewInputDeviceRequest prototype (DIX requests DDX to add a device). Does not 175589 link without an implemented NIDR. 175590 175591commit 93cd53860c3aca182a0a02543c41b5d71d65926b 175592Author: Daniel Stone <daniel.stone@nokia.com> 175593Date: Thu Jul 20 16:52:31 2006 -0400 175594 175595 kdrive: allow debugging 175596 175597commit cd0874dda1c30ef91a7d2b3cd455676422599ccf 175598Author: Daniel Stone <daniel.stone@nokia.com> 175599Date: Wed Jul 19 20:13:02 2006 -0400 175600 175601 never define MEMBUG 175602 175603 Definining MEMBUG causes allocations to randomly fail. 175604 175605commit 093943d4d02f1dbc8935b8cf835866a6e3885193 175606Author: Daniel Stone <daniel.stone@nokia.com> 175607Date: Wed Jul 19 20:09:55 2006 -0400 175608 175609 define DEBUG in DIX 175610 175611 Which makes #ifdef DEBUG actually useful. Incredible. 175612 175613commit 68b0678254240a984db9adefefb0cf68e9bfd4e4 175614Author: Daniel Stone <daniel.stone@nokia.com> 175615Date: Wed Jul 19 20:08:32 2006 -0400 175616 175617 exa: only disable cw when COMPOSITE is built 175618 175619commit 27df2eda795681c9f05e2907d74e2c102d3441e4 175620Author: Daniel Stone <daniel.stone@nokia.com> 175621Date: Wed Jul 19 16:18:28 2006 -0400 175622 175623 fix KdXvCopyPackedData to actually work 175624 175625 Remove extraneous bit shift in KdXvCopyPackedData, so it's actually 175626 useful. 175627 175628commit 00b24f119f03da86fa98ffea545c5b041810ce53 175629Author: Daniel Stone <daniel@fooishbar.org> 175630Date: Wed Jul 19 17:01:20 2006 -0400 175631 175632 fix minor typo 175633 175634commit f8a7a1e40c14a85ebde11c5854c07a8d529d38b9 175635Author: Daniel Stone <daniel.stone@nokia.com> 175636Date: Wed Jul 19 17:06:00 2006 -0400 175637 175638 fix XEvIE build without XKB 175639 175640 Don't unconditionally play with XKB stuff in XEvIE. 175641 175642commit 2f98841fde6bad807967ed15e954291240714198 175643Author: Ian Romanick <idr@umwelt.(none)> 175644Date: Thu Jul 20 16:08:27 2006 -0700 175645 175646 Remove unused variable. 175647 175648commit 985c34bf06af70a7296db8307899a17347a25558 175649Author: Adam Jackson <ajax@theobromine.nwnk.net> 175650Date: Thu Jul 20 17:33:13 2006 -0400 175651 175652 Remove the DDXTIME conditional, for being unused. 175653 175654commit c69c00d6523a35232a32e54a533811fc2b37815a 175655Merge: 46369350d 84683f19b 175656Author: Ian Romanick <idr@umwelt.(none)> 175657Date: Thu Jul 20 12:08:38 2006 -0700 175658 175659 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 175660 175661commit 84683f19b4d1c712281036bcabf8dc623e64b26a 175662Author: Daniel Stone <daniel@fooishbar.org> 175663Date: Tue Jul 18 18:16:12 2006 -0400 175664 175665 get rid of XFree86LOADER, XFree86Server, XFree86Module, and IN_MODULE 175666 Get rid of almost all uses of these definitions. They're still defined for 175667 delinquent out-of-tree drivers, and also for the Mesa build. As well as 175668 for miinitext.c. But largely gone. 175669 175670commit 881953813c7307f2aac4057b48d233e5f4a574cd 175671Author: Adam Jackson <ajax@theobromine.nwnk.net> 175672Date: Mon Jul 17 13:50:38 2006 -0400 175673 175674 Fix the Linux ACPI reopen code to use a repeating timer, rather than a 175675 one-shot sleep-and-reopen attempt. 175676 175677commit f029e9a32dcaa95b84e08ec173a0cc78fd92bdbf 175678Author: Kristian Høgsberg <krh@redhat.com> 175679Date: Sat Jul 15 22:05:38 2006 -0400 175680 175681 Un-glx-libcwrap DMX GLX proxy so it works without GL/include. 175682 175683commit 46369350d40819ecc2a9f37ed4aaa95866b80997 175684Author: Ian Romanick <idr@umwelt.(none)> 175685Date: Fri Jul 14 15:51:55 2006 -0700 175686 175687 Add some const qualifiers to serveral function parameters. 175688 175689commit 490fb304599b1f24b36439e5c1397781e7d2f612 175690Author: Ian Romanick <idr@umwelt.(none)> 175691Date: Fri Jul 14 15:26:56 2006 -0700 175692 175693 Rearrange code in xf86int10ParseBiosLocation to use strncasecmp. This 175694 eliminates the need for the first use of xstrdup in this function. 175695 The second use of xstrdup was *never* necessary and has also been 175696 eliminated. 175697 175698commit d3ee49bcbafe4b4e6b308686020847e978473779 175699Author: Ian Romanick <idr@umwelt.(none)> 175700Date: Fri Jul 14 15:13:35 2006 -0700 175701 175702 Refactor common code from the generic.c and linux.c version of 175703 xf86ExtendedInitInt10 to xf86int10GetBiosLocationType and 175704 xf86int10GetBiosSegment. 175705 175706 These changes were tested on MGA hardware on x86-64 with various 175707 combinations of InitPrimary and BiosLocation. 175708 175709commit 8793c7fd4ba7d1b3e2eff3f2c18d042ee9bb3f62 175710Author: Ian Romanick <idr@umwelt.(none)> 175711Date: Fri Jul 14 09:11:39 2006 -0700 175712 175713 Refactor identical xf86InitInt10 function from generic.c and linux.c 175714 to helper_exec.c. 175715 175716commit 1450fd596433f7adfe3d0798dc2ddceb9d0a9034 175717Author: Ian Romanick <idr@umwelt.(none)> 175718Date: Fri Jul 14 09:10:32 2006 -0700 175719 175720 Trivial refactor of libint10_la_SOURCES. 175721 175722commit 18624a9b7cf88e769c53cc3a876b3647c5d5057c 175723Author: Ian Romanick <idr@umwelt.(none)> 175724Date: Thu Jul 13 21:10:48 2006 -0700 175725 175726 Modify linuxGetSizesStruct to use pci_device instead of pciConfigPtr. 175727 175728commit 12595a77d6c2f037f48dfe751bdc10505b3317df 175729Merge: 37838de62 4c225a3a8 175730Author: Ian Romanick <idr@umwelt.(none)> 175731Date: Thu Jul 13 12:48:03 2006 -0700 175732 175733 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 175734 175735commit 4c225a3a8b2e7e5e5510347d8473f1318bbac769 175736Author: James Steven Supancic III <arrummzen@arrummzen.net> 175737Date: Thu Jul 13 10:03:57 2006 -0400 175738 175739 Bug #7482: Fix Xdmx's Render code to match reality; fixes BadLength client 175740 crashes. 175741 175742commit bb3aa94845a74d7718ba9539bb76203ec82957fc 175743Author: Drew Parsons <dparsons@debian.org> 175744Date: Tue Jul 11 18:26:55 2006 -0700 175745 175746 Bug #7346: Disable Composite extension in Xprt 175747 175748 Xorg Bug #7346 <https://bugs.freedesktop.org/show_bug.cgi?id=7346> 175749 Patch #6184 <https://bugs.freedesktop.org/attachment.cgi?id=6184> 175750 175751commit 10f3e32726d5b4981abd1a3a022e5b4f219fb41e 175752Author: Gustavo Pichorim Boiko <boiko@mandriva.com> 175753Date: Mon Jul 10 16:37:53 2006 -0700 175754 175755 Fix the configure check for the --{enable,disable}-dpms option 175756 175757commit 2194d99d9ca3c607e0f5ddd911ee3df536d77564 175758Author: Matthieu Herrb <matthieu.herrb@laas.fr> 175759Date: Sun Jul 9 16:16:08 2006 +0200 175760 175761 Replace GNU make-ism. 175762 175763commit 90a9b82272446fdaebe71c966325fd7e670f75c6 175764Author: Matthew Allum <mallum@polystyrene.(none)> 175765Date: Sat Jul 8 21:57:07 2006 +0100 175766 175767 Fix crash in Xephyr when running on host X with keymap width < 4 ( i.e xvnc ) 175768 175769commit 21e3e3ca298dce22e5fad6ef38aa6fe9736a1d3b 175770Merge: e80562189 39b2f7b21 175771Author: Matthew Allum <mallum@polystyrene.(none)> 175772Date: Sat Jul 8 21:13:52 2006 +0100 175773 175774 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 175775 175776commit e8056218944e873135c93039d1e9646d51364467 175777Author: Matthew Allum <mallum@polystyrene.(none)> 175778Date: Sat Jul 8 21:10:58 2006 +0100 175779 175780 Add support to Xephyr for simulating 8bit grayscale. 175781 175782commit 39b2f7b2182aedb1ab45415efb4c263012ace512 175783Author: Tilman Sauerbeck <tilman@code-monkey.de> 175784Date: Sat Jul 8 19:55:53 2006 +0200 175785 175786 Bug #3042: Use autoconf to get the correct name of a struct member. 175787 175788 This allows us to remove the kernel version ifdefs from the code, which 175789 are ugly and broken. 175790 175791commit 63f13e01ee6e7df1753f2113f4cff9538596be0a 175792Author: Tilman Sauerbeck <tilman@code-monkey.de> 175793Date: Sat Jul 8 11:33:44 2006 +0200 175794 175795 Bug #7097: do case-insensitive comparison for some hotkeys. 175796 175797 xkb's strcasecmp implementation has been moved to the dix so it's now 175798 safe to just use strcasecmp(). 175799 175800commit 5416f90e9c939027005fc01fa3ce3df56919ae0d 175801Author: Kristian Høgsberg <krh@redhat.com> 175802Date: Thu Jul 6 21:22:34 2006 -0400 175803 175804 Implement GLX_MESA_copy_sub_buffer. 175805 175806commit b84374b2917a91a7732e780ffab6a29c807a3ecc 175807Author: Kristian Høgsberg <krh@redhat.com> 175808Date: Thu Jul 6 02:28:55 2006 -0400 175809 175810 Add GLX_MESA_copy_sub_buffer marshalling support. 175811 175812commit 2152e2d364bdd179cf218cde446c763d8c8bb833 175813Author: Ed Catmur <ed@catmur.co.uk> 175814Date: Thu Jul 6 17:45:36 2006 -0700 175815 175816 Remove hardcoded 'lib' in XPRINTDIR. 175817 175818commit 233c004641483a75985e09fea5416ab2c7a97fc4 175819Author: Ed Catmur <ed@catmur.co.uk> 175820Date: Thu Jul 6 17:41:53 2006 -0700 175821 175822 Generate xprint.pre from xprint.cpp. 175823 175824commit b3e4d1d1f4bcf900146d7b8cd19e008209294663 175825Author: Adam Jackson <ajax@benzedrine.nwnk.net> 175826Date: Thu Jul 6 18:51:29 2006 -0400 175827 175828 PPC64 build fix. 175829 175830commit f847f3c0a956c0aec8ade8e32f770daae147f40b 175831Author: Keith Packard <keithp@neko.keithp.com> 175832Date: Thu Jul 6 15:33:31 2006 -0700 175833 175834 xorgcfg now needs libxkbui version 1.0.2 or better. 175835 175836 Old versions of libxkbui use XkbStrCaseCmp which doesn't exist any longer; 175837 the server fails to link with that version of the libkbui library, so 175838 require the current version (1.0.2). 175839 175840commit 27ffd7e03c05dd2083a10e5acebd2b385d729eeb 175841Author: Keith Packard <keithp@neko.keithp.com> 175842Date: Thu Jul 6 14:43:10 2006 -0700 175843 175844 GL proto version 1.4.7 needed for texture-from-pixmap defines. 175845 175846 GLX_TEXTURE_TARGET_EXT and GLX_TEXTURE_RECTANGLE_EXT are defined as a part 175847 of the new texture from pixmap GL extension in gl proto version 1.4.7 and 175848 are now used by the X server code. 175849 175850commit d7a96dd6f18e54c26fc5881772d033ac078db3a1 175851Author: Adam Jackson <ajax@benzedrine.nwnk.net> 175852Date: Thu Jul 6 17:39:14 2006 -0400 175853 175854 Switch the default mouse device on Linux to /dev/input/mice. 175855 175856commit 37838de62edc474f44c0a48bd56470a6c23f8956 175857Merge: fd49a0dcc 8d07ee070 175858Author: Ian Romanick <idr@umwelt.(none)> 175859Date: Thu Jul 6 13:01:57 2006 -0700 175860 175861 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 175862 175863 Conflicts: 175864 175865 hw/xfree86/os-support/bus/linuxPci.c 175866 175867commit 8d07ee070ecf0d403d9d27c80764d343b80af6f0 175868Author: Ian Romanick <idr@umwelt.(none)> 175869Date: Thu Jul 6 12:48:51 2006 -0700 175870 175871 Refactor linuxGetIOSize and linuxGetSizes. Eliminate the unnecessary 175872 optimization in the search loop. 175873 175874commit 704e645207d88a2d0a372cf69f6abd778ed4c30b 175875Author: Adam Jackson <ajax@benzedrine.nwnk.net> 175876Date: Thu Jul 6 14:22:33 2006 -0400 175877 175878 Remind dlloader that it needs to search the global scope as well as the 175879 loaded modules. Fixes LoaderSymbol() on symbols provided by the server. 175880 Spotted by Aaron Plattner. 175881 175882commit fd49a0dcc13280cf195bd6c7ee0f23d0840cb665 175883Merge: e66e7b48e 28b95fd9d 175884Author: Ian Romanick <idr@umwelt.(none)> 175885Date: Thu Jul 6 09:52:15 2006 -0700 175886 175887 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 175888 175889commit 28b95fd9d1c2f078aaaac75c310a27b17c74a6fc 175890Author: Kristian Høgsberg <krh@redhat.com> 175891Date: Thu Jul 6 03:25:38 2006 -0400 175892 175893 Drop unused GL/include subtree. 175894 175895 Woo, less Makefile's to generate. 175896 175897commit fc1a55671dceae0e2a701e044ff8203fae5eb1ba 175898Author: Kristian Høgsberg <krh@redhat.com> 175899Date: Thu Jun 29 04:35:45 2006 -0400 175900 175901 Implement glXCreateWindow and glXDestroyWindow. 175902 175903commit ee012588d28b468bd41da8d216210f8cb2bf8cb5 175904Author: Kristian Høgsberg <krh@redhat.com> 175905Date: Thu Jun 29 04:25:54 2006 -0400 175906 175907 Move __GLXdrawable lookup and creation into GetDrawableOrPixmap. 175908 175909 Also refactors __glXSwapBuffers to use GetDrawableOrPixmap for 175910 getting the __GLXdrawable. This patch paves the way for GLXWindows 175911 with XIDs different from the X Windows they are created for, a 175912 prerequisite for glXCreateWindow(). 175913 175914commit 8b5bc6a9ab487fdea754266b120c686d75d9e100 175915Author: Kristian Høgsberg <krh@redhat.com> 175916Date: Thu Jun 29 00:05:01 2006 -0400 175917 175918 Drop global GLX error integer variables and use __glXError() instead. 175919 175920 Also drop glxerror.h (__glXError is now declared in glxserver.h) 175921 and global.c (last remaining globals are in glxext.c now). 175922 175923 With this change we now support all GLX 1.3 error codes. 175924 175925commit 7cf3ec7b59223f15314a0629f122ecb796678421 175926Author: Kristian Høgsberg <krh@redhat.com> 175927Date: Wed Jun 28 17:00:23 2006 -0400 175928 175929 Move createDrawable from __GLXcontext to __GLXscreen. 175930 175931commit eea8efe4516750b2505b52ebc9f769f5e8a6f94c 175932Author: Kristian Høgsberg <krh@redhat.com> 175933Date: Wed Jun 28 15:59:01 2006 -0400 175934 175935 Add marshalling for GLX 1.3 requests. 175936 175937 Also, hook up glXGetDrawableAttributes and glXQueryContext to existing 175938 DoGetDrawableAttributes and __glXQueryContextInfoEXT. 175939 175940commit e66e7b48e15543f3753951a19376f0446759963a 175941Merge: a9d484289 eb35f812a 175942Author: Ian Romanick <idr@umwelt.(none)> 175943Date: Wed Jul 5 15:21:36 2006 -0700 175944 175945 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 175946 175947commit eb35f812a5b65adcc5f6cbb91b31b69cae5d7f3d 175948Author: Greg Kroah-Hartman <gregkh@suse.de> 175949Date: Wed Jul 5 13:27:26 2006 -0700 175950 175951 add another file to .gitignore 175952 175953commit 863f5cc31b747bc9f2fcd6a9e20c613a11733bf4 175954Author: Greg Kroah-Hartman <gregkh@suse.de> 175955Date: Wed Jul 5 13:26:34 2006 -0700 175956 175957 fix compiler warning in hw/xfree86/common/xf86Config.c 175958 175959commit f059b61ab3af25b03c704669eddb838d3ce4366c 175960Author: Greg Kroah-Hartman <gregkh@suse.de> 175961Date: Wed Jul 5 11:47:25 2006 -0700 175962 175963 fix compiler warning about xnestRecolorCursor() not being defined 175964 175965commit cc3e99f747586f9d32622e5a682de39891b1fcba 175966Author: Greg Kroah-Hartman <gregkh@suse.de> 175967Date: Wed Jul 5 10:13:19 2006 -0700 175968 175969 fix some more compiler warnings due to defines being declared differently 175970 175971commit 59836c0f2abee3339e1aa30dacadb82e477943d6 175972Author: Greg Kroah-Hartman <gregkh@suse.de> 175973Date: Wed Jul 5 09:30:48 2006 -0700 175974 175975 fix wrong function pointer type in hw/dmx/dmxcmap.c 175976 175977commit 12563db59dd613ecc926e3bed9534152ebc0a2fb 175978Author: Eric Anholt <anholt@FreeBSD.org> 175979Date: Mon Jul 3 12:52:27 2006 -0700 175980 175981 Revert "Optimize out computing a gradient pixel if the mask value is 0." 175982 175983 This reverts cf46242e337481cd3b9b39d77dd621d2a63b11f9 commit. It wasn't meant 175984 to be pushed to master yet, and doesn't work. 175985 175986commit 002e28c12c74aa63777f65cbfb382c2bfd0d6850 175987Author: Eric Anholt <anholt@FreeBSD.org> 175988Date: Mon Jul 3 12:48:12 2006 -0700 175989 175990 Correct AGP memory deallocation argument on *BSD. 175991 175992 This fixes leaks and eventual crashes with RandR resizing on Intel. 175993 175994commit cf46242e337481cd3b9b39d77dd621d2a63b11f9 175995Author: Eric Anholt <anholt@FreeBSD.org> 175996Date: Wed Jun 28 18:35:59 2006 +0200 175997 175998 Optimize out computing a gradient pixel if the mask value is 0. 175999 176000 Obtained from: kdrive CVS (DavidR XGL fb/ megapatch) 176001 176002commit a838fb70c52a829872680f6a2a2e7dd6d2dc9247 176003Author: Eric Anholt <anholt@FreeBSD.org> 176004Date: Mon Jul 3 19:22:36 2006 +0200 176005 176006 Bump server version to 7.1.99.2 for gradient and repeat fixes. 176007 176008commit 25d871d98462f0481ee419295ddc94b8c79dc881 176009Author: Eric Anholt <anholt@FreeBSD.org> 176010Date: Mon Jul 3 19:22:26 2006 +0200 176011 176012 Fix source picture filter check for multiple screens. 176013 176014 Now, we only check for filter commonality if we're operating on a source 176015 picture, and we compare the id (screen-independent index of the filter name) 176016 rather than the pointer to the filter (per-screen state). 176017 176018commit 7106a77df37c06d2b5568eceeb9297096bff3137 176019Author: Eric Anholt <anholt@FreeBSD.org> 176020Date: Sun Jul 2 12:41:35 2006 +0200 176021 176022 Fix bugs in support for new repeatTypes in XAA and EXA. 176023 176024 EXA now won't pass pictures with new repeatTypes to drivers. We can add a flag 176025 for them to support it at a later time. 176026 176027commit f5e92542a14f51029347b6476e4e4af69144930b 176028Author: Eric Anholt <anholt@FreeBSD.org> 176029Date: Fri Jun 30 12:03:47 2006 +0200 176030 176031 Bug #7366: Fix crashes when setting filters on source pictures. 176032 176033 Now, filters may only be set on source pictures when the filter is common to 176034 all screens. Also, like SetPictureTransform, ChangePictureFilter is now not 176035 called on source pictures. 176036 176037commit 6ef457913955d4289081c7d07d528963ccf5272c 176038Author: Eric Anholt <anholt@FreeBSD.org> 176039Date: Fri Jun 30 03:01:14 2006 +0200 176040 176041 Bug #7366: Fix two crashes in operations on source pictures. 176042 176043 A screen's ChangePictureTransform now isn't called when changing the transform, 176044 as source pictures aren't associated with screens. Also, attempting to set 176045 an AlphaMap to a source picture will fail with BadMatch just like a Window 176046 would, preventing another crash. 176047 176048commit 50a3e1ad18c815a5adafee22beccdf970bae62d6 176049Author: Rudo Thomas <rudo@matfyz.cz> 176050Date: Sat Jul 1 12:34:36 2006 -0700 176051 176052 Missing close parenthesis in one of the setuid() fixes. 176053 176054commit 124a81eb389dfa510ac07ca93ee17c4c9d6e56ea 176055Merge: d3d6c5f4d 179737d4a 176056Author: Keith Packard <keithp@neko.keithp.com> 176057Date: Sat Jul 1 11:12:50 2006 -0700 176058 176059 Merge branch 'origin' 176060 176061commit d3d6c5f4d05e0ca5b566e19657e0fe2b3898482a 176062Author: Paul Mackerras <paulus@samba.org> 176063Date: Sat Jul 1 11:10:18 2006 -0700 176064 176065 Bug #7381: Coordinates get wrapped in accelerated line drawing on pixmap 176066 176067 XAAPolylinesWideSolid was adding the drawable origin onto each element in the 176068 pPts array. Since the values got stored back into the pPts array, they got 176069 truncated to 16 bits, causing the overflow I saw. This patch avoids storing 176070 the coords back into the pPts array (and actually reduces the size of the 176071 code too :). Now the 32-bit sum of coords + origin doesn't get truncated to 176072 16 bits, and the problem is solved. 176073 176074commit 179737d4a07ed10a734fe017b5680f8e78ffda96 176075Author: Jens Granseuer <jensgr@gmx.net> 176076Date: Wed Jun 7 01:46:00 2006 -0700 176077 176078 Bug 7145: fix build with gcc 2.95 & other c89 compilers 176079 176080 Move variable declarations to start of blocks as required by c89 176081 176082commit a9d4842893b65232d61fe4c63f074d30e81021d2 176083Merge: 66cac9b17 6bd4c2543 176084Author: Ian Romanick <idr@umwelt.(none)> 176085Date: Fri Jun 30 11:49:29 2006 -0700 176086 176087 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 176088 176089commit 6bd4c254396cb0f4e8ae21ff455ebb15cd9f4f10 176090Author: Martin Bochnig <mb1x@gmx.com> 176091Date: Mon Jun 26 01:52:24 2006 +0200 176092 176093 Updated Solaris aperture driver to build on sun4v & amd64 kernels 176094 Updated for Solaris 10 changes to DDI 176095 176096commit 66cac9b17498497d939bfb7a4075bcee5a485e5f 176097Merge: 7c4e54025 54d9acd51 176098Author: Ian Romanick <idr@umwelt.(none)> 176099Date: Wed Jun 28 14:11:49 2006 -0700 176100 176101 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 176102 176103commit 54d9acd5113318274e291abab4554b8e678227df 176104Author: Kristian Høgsberg <krh@redhat.com> 176105Date: Tue Jun 27 19:44:52 2006 -0400 176106 176107 Add damage tracking to GLX_EXT_tfp implementation. 176108 176109 - Only update when pixmap content actually change; 176110 - Only update the regions that acutally changed. 176111 176112 This is a worthwhile optimization, but it doesn't completely remove 176113 the bottleneck, as mesa still uploads then entire texture whenever 176114 it changes. 176115 176116commit adfe8e7437ff739f54d1d074008e8cc0e3bcb4d3 176117Author: Eric Anholt <anholt@FreeBSD.org> 176118Date: Tue Jun 27 21:49:00 2006 +0200 176119 176120 Bump server release to 7.1.99.1. 176121 176122 This will be important for a couple of cairo workaround tests. 176123 176124commit 63c169e3b1f7d6a7375a414fcd50cce32358a525 176125Author: Eric Anholt <anholt@FreeBSD.org> 176126Date: Tue Jun 27 04:11:47 2006 +0200 176127 176128 Fix MMX Saturate implementation. 176129 176130 The code was expanding the source blend factor from the wrong channel. Fixes 176131 cairo's clip-operator test. 176132 176133commit 7c4e54025345455f420cd2abaa7a40679689d3cb 176134Merge: 9d0a15dca ff6b59a0d 176135Author: Ian Romanick <idr@umwelt.(none)> 176136Date: Mon Jun 26 14:39:44 2006 -0700 176137 176138 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver into pci-rework 176139 176140commit ff6b59a0dbadbe61a53e48c23965d3073d95791b 176141Merge: b3c869304 48c871564 176142Author: Alan Coopersmith <alan.coopersmith@sun.com> 176143Date: Mon Jun 26 13:02:33 2006 -0700 176144 176145 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176146 176147commit b3c869304cd85af034aa9debaa874e29d14fcbe6 176148Author: Peter Breitenlohner <peb@mppmu.mpg.de> 176149Date: Mon Jun 26 10:48:44 2006 -0700 176150 176151 Free small, one-time memory leak in xdmcp -from handling 176152 176153 Part of Patch #6046 <https://bugs.freedesktop.org/attachment.cgi?id=6046> 176154 176155commit 48c871564d493203d434d5da015903399287f619 176156Author: Eric Anholt <anholt@FreeBSD.org> 176157Date: Mon Jun 26 15:57:32 2006 +0200 176158 176159 Move EXA_PM_IS_SOLID() to the public API, since drivers will want it frequently. 176160 176161commit afb84c2fca56887b3bfe7aa93f337c49b087acdc 176162Author: George Fufutos <fufutos610@hotmail.com> 176163Date: Sat Jun 24 15:23:14 2006 +0200 176164 176165 Bug #6911: Check return value of exaGetPixelFromRGBA(). 176166 176167commit 930b9a069a425818d4e9965f53be7df1f4d7d422 176168Author: Michel Dänzer <michel@tungstengraphics.com> 176169Date: Sat Jun 24 15:21:17 2006 +0200 176170 176171 Bug #6818: Avoid infinite loop in exaLog2() with negative arguments. 176172 176173commit 36756fdb2ddc154b406f664a6af0f38d26e6973d 176174Author: Michel Dänzer <michel@tungstengraphics.com> 176175Date: Sat Jun 24 15:09:24 2006 +0200 176176 176177 Make sure <execinfo.h> is actually included when needed. 176178 176179 configure only defines HAVE_BACKTRACE, not HAVE_EXECINFO_H. 176180 176181 This could cause problems on platforms where the size of a pointer is greater 176182 than that of an integer, see 176183 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=363218 . 176184 176185commit 4426215a6e99f84550aaac23ac9c2018668bfbc1 176186Author: Michel Dänzer <michel@tungstengraphics.com> 176187Date: Sat Jun 24 15:02:56 2006 +0200 176188 176189 Bug #7213: Fix the XFree86-DRI extension for byte-swapped clients. 176190 176191 These clients are by definition non-local and thus not direct rendering 176192 capable, but they still need the QueryVersion and QueryDirectRenderingCapable 176193 requests to find out cleanly. 176194 176195commit a195a3debca02572d9f7d7a9976b5bf67acc5d08 176196Author: Michel Dänzer <michel@tungstengraphics.com> 176197Date: Sat Jun 24 14:54:52 2006 +0200 176198 176199 Fix byte swapping in some GLX requests. 176200 176201commit bc6cfde19887eff7a07dc739ffa29609fb55b83d 176202Author: Eric Anholt <anholt@FreeBSD.org> 176203Date: Fri Jun 23 20:07:34 2006 -0700 176204 176205 Use correct OSNAME setting so we can find os-specific modules like libdrm. 176206 176207commit c7ac485a59709572307b9a4a9abacc52c7021b65 176208Author: Eric Anholt <anholt@FreeBSD.org> 176209Date: Wed Jun 21 09:34:55 2006 -0700 176210 176211 Remove the default case from fbcompose.c switches which should cover all cases. 176212 176213 Instead, stick the NULL return default case afterwards, so that the compiler can 176214 warn us when we've got unimplemented cases. Removes some unimplemented and 176215 unused 8bpp, depth 4 picture format names. 176216 176217commit 2cf1f39ca974c81a2f52d2f7509aa3d098a87176 176218Author: Eric Anholt <anholt@FreeBSD.org> 176219Date: Wed Jun 21 09:30:59 2006 -0700 176220 176221 Add a manpage for EXA. 176222 176223commit 9d0a15dca56b14821cad474f9db9c332acb3b1e6 176224Merge: 6c514ca14 d67fd1069 176225Author: Ian Romanick <idr@umwelt.(none)> 176226Date: Thu Jun 22 16:21:10 2006 -0700 176227 176228 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176229 176230 Conflicts: 176231 176232 hw/xfree86/loader/xf86sym.c 176233 176234commit 6c514ca14c2326c2cf82b73196429789d11e4a01 176235Author: Ian Romanick <idr@umwelt.(none)> 176236Date: Thu Jun 22 15:51:16 2006 -0700 176237 176238 Use XSERVER_LIBS instead of XORG_LIBS for -lpciaccess. The former is 176239 only for libraries that are built as part of the xserver build. 176240 176241commit d67fd106968e371d8be3966ed5ecdd3c69f36e3a 176242Author: Adam Jackson <ajax@benzedrine.nwnk.net> 176243Date: Thu Jun 22 12:47:51 2006 -0400 176244 176245 Add kdrive servers to .gitignore 176246 176247commit bf17c6dede1c0cf1edee10f2cc7e1e619b944d3a 176248Author: Adam Jackson <ajax@benzedrine.nwnk.net> 176249Date: Thu Jun 22 12:24:09 2006 -0400 176250 176251 Drop libz from the server's link line, it was only needed for LBX. 176252 176253commit c0cb8d1fb80540e093da54da3ee2f55bdf139274 176254Author: Alan Coopersmith <alan.coopersmith@sun.com> 176255Date: Wed Jun 21 18:12:41 2006 -0700 176256 176257 Use XACE, not XCSECURITY to decide if SecurityLookup* are exported 176258 176259commit 3177dc498a955cd58cd6054a7c7e69724db4a59b 176260Merge: 3f19803e0 91dcac529 176261Author: Alan Coopersmith <alan.coopersmith@sun.com> 176262Date: Wed Jun 21 18:06:06 2006 -0700 176263 176264 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176265 176266commit 3f19803e0b1adc66e695f63f915b8dc85eb84215 176267Author: Alan Coopersmith <alan.coopersmith@sun.com> 176268Date: Wed Jun 21 18:05:51 2006 -0700 176269 176270 APPGROUP requires both X-ACE & XC-SECURITY now 176271 176272commit 91dcac5295486cc55a34ad91704bfa483bd31eeb 176273Merge: d8135eb9e 77c947b90 176274Author: Adam Jackson <ajax@benzedrine.nwnk.net> 176275Date: Wed Jun 21 20:49:30 2006 -0400 176276 176277 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176278 176279commit d8135eb9e414bf6957f64c5102ee0ef7c2404c6f 176280Author: Adam Jackson <ajax@benzedrine.nwnk.net> 176281Date: Wed Jun 21 20:49:21 2006 -0400 176282 176283 Unbreak unbreaking the loader. Re-add the symbol reference lists so that the 176284 linker will include everything it's supposed to. This is a terrible solution, 176285 but ld semantics don't let you do anything better. 176286 176287commit f83cee0338eca095ad601374a87775be823a2565 176288Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 176289Date: Wed May 5 20:07:37 2004 +0000 176290 176291 Modify XC-SECURITY and XC-APPGROUP extensions to work with XACE 176292 176293commit 15c9002d68a7eeb02a6db1f231af7a18a3cf7512 176294Author: Alan Coopersmith <alan.coopersmith@sun.com> 176295Date: Wed Jun 21 16:24:20 2006 -0700 176296 176297 Correct ifdef - should be XACE, not XSECURITY 176298 176299commit 90af38fa0c46c2081d2becac262a614c26ba6ef1 176300Merge: 3e098efa3 77c947b90 176301Author: Alan Coopersmith <alan.coopersmith@sun.com> 176302Date: Wed Jun 21 16:23:31 2006 -0700 176303 176304 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176305 176306 Conflicts: 176307 176308 Xext/appgroup.c 176309 176310commit 3e098efa35ba70ad4d5699af3130a3b02e1cb06e 176311Author: Alan Coopersmith <alan.coopersmith@sun.com> 176312Date: Wed Jun 21 16:21:03 2006 -0700 176313 176314 Correct ifdef - should be XACE, not XCSECURITY 176315 176316commit 77c947b900faf34f425eef1549d8210c475e093b 176317Author: Alan Coopersmith <alan.coopersmith@sun.com> 176318Date: Wed Jun 21 16:18:41 2006 -0700 176319 176320 Move Xserver internal API for appgroup from Xagsrv.h to appgroup.h 176321 176322 (Since all use is inside the xserver module, might as well keep the header 176323 in the Xserver module instead of in proto/XExt to allow easier synchronization 176324 of updates.) 176325 176326commit aa131c51e06e735d7b54cadf51205f703a3de4ef 176327Merge: 1708b0c7d 77aa701e8 176328Author: Ian Romanick <idr@umwelt.(none)> 176329Date: Wed Jun 21 14:39:52 2006 -0700 176330 176331 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176332 176333commit 1708b0c7d962b80073500eb1c483447bbf965bc9 176334Author: Ian Romanick <idr@umwelt.(none)> 176335Date: Wed Jun 21 14:39:39 2006 -0700 176336 176337 Eliminate "ugly" spaces around parenthesis. Add const qualifier. 176338 General code clean-ups. 176339 176340commit a3c26d01a16e4ae20a5d1acf9d888e95760c3da1 176341Author: Ian Romanick <idr@umwelt.(none)> 176342Date: Wed Jun 21 14:38:07 2006 -0700 176343 176344 Move declaration of "j" inside #ifdef block where it is used to 176345 eliminate compiler warning. 176346 176347commit 40aafaf154ba8a2cba857689d1481be2d4a610c2 176348Author: Greg Kroah-Hartman <gregkh@suse.de> 176349Date: Wed Jun 21 14:36:09 2006 -0700 176350 176351 fix a number of compiler warnings in os/* 176352 176353commit 77aa701e898c6525322cc4029d95167dd9f6e618 176354Author: Greg Kroah-Hartman <gregkh@suse.de> 176355Date: Wed Jun 21 14:16:48 2006 -0700 176356 176357 add some function prototypes to hw/xprint/DiPrint.h as they are exported 176358 176359commit 865884d050e1778180f7677e15f9ee1625ea4bb2 176360Author: Greg Kroah-Hartman <gregkh@suse.de> 176361Date: Wed Jun 21 14:16:28 2006 -0700 176362 176363 fix compiler warning in dix/xpstubs.c when XPRINT is not defined 176364 176365commit ed18d776f02e2ab235954501ef64936af9f9d909 176366Author: Kristian Høgsberg <krh@redhat.com> 176367Date: Wed Jun 21 16:22:14 2006 -0400 176368 176369 Fix #2488 for fb too: sample pixel center when transforming. 176370 176371commit ea5e0eabd1303a55d8fc10f44d21a3d371ce8919 176372Author: Matthias Hopf <mhopf@suse.de> 176373Date: Wed Jun 21 17:08:51 2006 +0200 176374 176375 Bug 4320: Fastpath corner case improvement for Composite. 176376 176377commit 9af315a9be19b48faa1249e0575cbe3d1e31dec5 176378Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 176379Date: Wed May 5 20:07:37 2004 +0000 176380 176381 Modify XC-SECURITY and XC-APPGROUP extensions to work with XACE 176382 176383commit f68ecfa482b2c2037f929710310c7b9ce6fe9291 176384Author: Alan Coopersmith <alan.coopersmith@sun.com> 176385Date: Tue Jun 20 19:01:34 2006 -0700 176386 176387 Add X-ACE to build system 176388 176389 - Added --disable-xace to configure.ac and issue configure error if trying 176390 to build XC-Security without X-ACE 176391 - Added XACE #define to dix-config.h 176392 - Added X-ACE sources to Xext/Makefile.am 176393 176394commit a46c06dab8392cf8012c7cc0b916de9a9e569671 176395Merge: 49b368c0b d44b2a0a5 176396Author: Alan Coopersmith <alan.coopersmith@sun.com> 176397Date: Tue Jun 20 18:40:18 2006 -0700 176398 176399 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176400 176401 Conflicts: 176402 176403 Xext/appgroup.c 176404 Xext/security.c 176405 dix/devices.c 176406 dix/dispatch.c 176407 dix/dixutils.c 176408 dix/events.c 176409 dix/extension.c 176410 dix/property.c 176411 dix/window.c 176412 os/access.c 176413 176414commit d44b2a0a57fb89741173c31676af0ccc822387dc 176415Author: Alan Coopersmith <alan.coopersmith@sun.com> 176416Date: Tue Jun 20 18:22:51 2006 -0700 176417 176418 Move Xserver API for security extension to securitysrv.h 176419 176420commit a54435946544a039fc333bb5e3438501d0d1ffc6 176421Author: Alan Coopersmith <alan.coopersmith@sun.com> 176422Date: Tue Jun 20 18:14:27 2006 -0700 176423 176424 Move Xserver API for security extension to securitysrv.h 176425 176426commit 49b368c0bb04816c4a3579071c596b2398cae3ec 176427Merge: 63f6e6bbf 481d4012e 176428Author: Alan Coopersmith <alanc@alf.(none)> 176429Date: Tue Jun 20 16:22:39 2006 -0700 176430 176431 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176432 176433commit 481d4012e74d9b0e98911f0ae02700ecf4cfc5ac 176434Author: Alan Coopersmith <alanc@alf.(none)> 176435Date: Tue Jun 20 16:16:19 2006 -0700 176436 176437 Don't add -ldl to XORG_LIBS if it's not needed for dlopen 176438 176439commit 63f6e6bbfd0d3677e29621af982c9392ead98dd7 176440Merge: 88ede2cec 6df52fb77 176441Author: Alan Coopersmith <alanc@alf.(none)> 176442Date: Tue Jun 20 16:03:34 2006 -0700 176443 176444 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176445 176446commit 6df52fb7745c185c0168060f69cc6b4f5315914e 176447Author: Alan Coopersmith <alanc@alf.(none)> 176448Date: Tue Jun 20 16:02:55 2006 -0700 176449 176450 Delete code that's been inside #if 0 since X11R6.7. 176451 176452commit 75398cff82b56c0a8fc301e84e2087f59a61a360 176453Merge: 1f91b5726 2b5868540 176454Author: Ian Romanick <idr@umwelt.(none)> 176455Date: Tue Jun 20 15:49:47 2006 -0700 176456 176457 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176458 176459 Conflicts: 176460 176461 hw/xfree86/loader/Makefile.am 176462 hw/xfree86/loader/xf86sym.c 176463 176464commit 88ede2cec79281a43cecb43ee6dec65770f82530 176465Merge: 227a31934 2b5868540 176466Author: Alan Coopersmith <alanc@alf.(none)> 176467Date: Tue Jun 20 15:29:55 2006 -0700 176468 176469 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176470 176471commit 227a3193405147fbbee2971cc15bac92cc13285a 176472Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 176473Date: Tue Jun 1 21:09:25 2004 +0000 176474 176475 Add XACE extension source files. 176476 176477 (Copied from XACE-SELINUX branch in Xorg monolith CVS since these were never 176478 imported to modular cvs or git trees.) 176479 176480commit 1f91b5726c721778de0627c0139100744db06934 176481Author: Ian Romanick <idr@umwelt.(none)> 176482Date: Tue Jun 20 15:03:13 2006 -0700 176483 176484 Fix a compiler warning about varibles being used without being set. 176485 176486 Add some const qualifiers. 176487 176488 Remove some "ugly" space around parenthesis. 176489 176490commit 2b58685402e70f123e131fd8146e6083ff5214a4 176491Author: Matthieu Herrb <matthieu@deville.herrb.com> 176492Date: Tue Jun 20 21:07:53 2006 +0200 176493 176494 Check setuid() return value. Bugzilla #7116. 176495 176496commit 4365d16c8629e824973ee0c357efdfbfd28b672c 176497Author: Kristan Høgsberg <krh@redhat.com> 176498Date: Mon Jun 19 22:13:22 2006 -0400 176499 176500 Pull over convolution filter fixes from xgl-0-0-1 branch. 176501 176502 Cherry-picking patches: 176503 176504 8a5ea68800b9d7dca90ff4e573ad8533852f1ea3 and 176505 4d755fe14274a7293748ce9aa666ab85df6297c5 176506 176507commit f818e0ab60da3779ab2602c6e6d3ff261b50917e 176508Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 176509Date: Wed May 5 20:07:37 2004 +0000 176510 176511 Modify XC-SECURITY and XC-APPGROUP extensions to work with XACE 176512 (cherry picked from 0106715000196c7b349a0b4494b61545f0f5e138 commit) 176513 176514commit 0707eb33d6826e1300a905edea28c12134600b12 176515Merge: 37f0ae024 98d17bba7 176516Author: Alan Coopersmith <alanc@alf.(none)> 176517Date: Mon Jun 19 17:09:51 2006 -0700 176518 176519 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176520 176521commit 37f0ae02457bd22b27f8f30a373e5cb19f2fbaea 176522Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 176523Date: Wed May 5 20:04:52 2004 +0000 176524 176525 Replace XC-SECURITY code with XACE security hooks 176526 (cherry picked from 8526cd6395490b03b279f1962df777fb0e4a9878 commit) 176527 176528commit 98d17bba716619e3402bd74c7c3e1c01d055ef6e 176529Author: Alan Coopersmith <alanc@alf.(none)> 176530Date: Mon Jun 19 17:07:59 2006 -0700 176531 176532 Tell git to ignore emacs *~ droppings and git .msg files 176533 176534commit 569c808a2375be71f835ee8693605487484bd22e 176535Author: Eric Anholt <anholt@FreeBSD.org> 176536Date: Mon Jun 19 16:42:09 2006 -0700 176537 176538 Fix crash when using PICT_x4a4 by supplying an appropriate fbFetchPixel_x4a4. 176539 176540commit 8d9ccc90a54c786ca4ba5620ab0a965e3f3bc8ea 176541Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 176542Date: Wed May 5 20:15:41 2004 +0000 176543 176544 Add XACE extension 176545 (partially cherry picked from 8d4f21ab53c44ca48501d6211ea6db0c0b8af916 commit) 176546 176547commit 55426650417df4ec22ea0e2a67f6074f0ac1d54e 176548Author: Eric Anholt <anholt@FreeBSD.org> 176549Date: Mon Jun 19 15:04:46 2006 -0700 176550 176551 Clean up gcc warnings from picture format CARD32 -> enum change. 176552 176553commit 520c80f4b807ae6419e70fe2b524532465b509ac 176554Author: Eric Anholt <anholt@FreeBSD.org> 176555Date: Mon Jun 19 14:40:27 2006 -0700 176556 176557 Don't forget to step the rows when verifying the equivalence of fb/sys areas. 176558 176559 This is only used by fakexa, but we would have missed some errors without this 176560 fix. 176561 176562commit e793f0eeee3e9c83b6a7b50d451fb6db12839087 176563Author: Eric Anholt <anholt@FreeBSD.org> 176564Date: Mon Jun 19 14:06:02 2006 -0700 176565 176566 Correct component ordering when fetching [ax]4b4g4r4 pixels. 176567 176568 Noticed by: rendercheck 176569 176570commit e1672a12eb70836a2ceec803d505294897ae8cd2 176571Author: Eric Anholt <anholt@FreeBSD.org> 176572Date: Mon Jun 19 13:20:56 2006 -0700 176573 176574 Convert PICT_* names from #defines to an enum to aid in debugging. 176575 176576commit 9742d55c820a260a42a4537502295931d4529deb 176577Author: Greg Kroah-Hartman <gregkh@suse.de> 176578Date: Mon Jun 19 14:40:14 2006 -0700 176579 176580 update .gitignore to handle Xprint move 176581 176582commit d97a21acb878bc4e5e6542912fbd820503bba312 176583Author: Greg Kroah-Hartman <gregkh@suse.de> 176584Date: Mon Jun 19 14:36:54 2006 -0700 176585 176586 fix compiler warnings in hw/xfree86/i2c/fi1236.c 176587 176588commit 29c78321e86956c4ce0c1c899d82557f927e04da 176589Author: Greg Kroah-Hartman <gregkh@suse.de> 176590Date: Mon Jun 19 14:36:41 2006 -0700 176591 176592 fix compiler warning in hw/xfree86/i2c/tda9850.c 176593 176594commit 9f2793551f335e5fb08990fc8bb9e05e0ffb68d5 176595Author: Greg Kroah-Hartman <gregkh@suse.de> 176596Date: Mon Jun 19 11:50:47 2006 -0700 176597 176598 fix compiler warning in hw/vfb/InitOutput.c 176599 176600commit b20ae5ddb7682bafcee6f8bf0c8208a3f70b882b 176601Author: Greg Kroah-Hartman <gregkh@suse.de> 176602Date: Mon Jun 19 11:38:52 2006 -0700 176603 176604 fix compiler warnings in hw/xfree86/xf4bpp/ppcGC.c 176605 176606commit 4d258f31967141e3c4a6e4abbef89ffa717e85aa 176607Author: Greg Kroah-Hartman <gregkh@suse.de> 176608Date: Mon Jun 19 11:27:47 2006 -0700 176609 176610 fix compiler warnings in XTrap/xtrapdi.c 176611 176612commit 870cecb72c2cba44dc64cb202917453603c8f287 176613Author: Greg Kroah-Hartman <gregkh@suse.de> 176614Date: Mon Jun 19 11:22:42 2006 -0700 176615 176616 fix compiler warnings in XTrap/xtrapdiswp.c 176617 176618commit a28652f9c35fbc009245382a5cc2a022f42366fc 176619Author: Adam Jackson <ajax@theobromine.nwnk.net> 176620Date: Mon Jun 19 00:57:18 2006 -0400 176621 176622 Another round of loader sense-beating. Remove the (unused) server export 176623 lists, a really bad hash table, the last vestiges of the other backends, 176624 and some miscellaneous cleanups. Good for dropping 300k from the size of 176625 the built server on x86. 176626 176627commit 98a602fab1f307a07a96868d7dae12b6f8d7f405 176628Author: Adam Jackson <ajax@theobromine.nwnk.net> 176629Date: Sun Jun 18 23:54:04 2006 -0400 176630 176631 Don't bother building RAC as a module, that's just absurd. 176632 176633commit 76aaf7eae7409162c5ed2963f2e27d019cb30263 176634Author: Eric Anholt <anholt@FreeBSD.org> 176635Date: Sun Jun 18 19:47:29 2006 -0700 176636 176637 Add a couple of (doxygen) comments I wrote while looking at modesetting. 176638 176639commit 6aaf0e5b581b06fc73e56f863a26cd9d684eb9c0 176640Author: Eric Anholt <anholt@FreeBSD.org> 176641Date: Sun Jun 18 19:12:15 2006 -0700 176642 176643 Add options to disable EXA acceleration for Composite/UTS/DFS, and always print 176644 176645 out how much memory EXA is managing for offscreen pixmaps. 176646 176647commit 21ef7e17ef6dca177461c9438b9df707a4d664a2 176648Author: Eric Anholt <anholt@FreeBSD.org> 176649Date: Sun Jun 18 18:57:55 2006 -0700 176650 176651 Add some missing .gitignore stuff for Mesa symlinks and other generated files. 176652 176653commit 71fbda8049f64c7fefae8ab817fb5f37ee2ee134 176654Author: Adam Jackson <ajax@theobromine.nwnk.net> 176655Date: Sun Jun 18 21:07:28 2006 -0400 176656 176657 Xprint/ -> hw/xprint 176658 XpConfig -> hw/xprint/config 176659 176660commit 868e2cab706e317618646e064b0559d4e68c7b32 176661Author: Eric Anholt <anholt@FreeBSD.org> 176662Date: Fri Jun 16 10:17:51 2006 -0700 176663 176664 Add explicit dependencies (Xorg_DEPENDENCIES = ) on the internal libraries 176665 (such as libcw.la) that we link into the server, causing it to be rebuild 176666 automatically when they're updated. Some system libraries are included, but 176667 don't appear to cause any harm. You would think this would be automatic... 176668 176669commit 53f74b6aa95fe57fda45fd8a051595e772f00402 176670Author: Eric Anholt <anholt@FreeBSD.org> 176671Date: Fri Jun 16 10:14:30 2006 -0700 176672 176673 Bugzilla #5120, #7246: In CW's GC ops, validate the backing GC against the 176674 backing drawable if the serial numbers differ. Fixes crash in XAA which 176675 occurred when the DDX bumped the serial number on the backing drawable and 176676 expected it to get re-validated, and we didn't because the wrapped drawable 176677 hadn't been bumped. 176678 176679commit 72ac20c0e878ad48ff61f7e846cbf1005a4eb435 176680Merge: fc3e81473 b90088321 176681Author: Ian Romanick <idr@localhost.localdomain> 176682Date: Tue Jun 13 21:13:02 2006 -0700 176683 176684 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176685 176686commit b90088321e6ef84970aa97d7c851af93f49bf4b7 176687Author: Ian Romanick <idr@localhost.localdomain> 176688Date: Mon Jun 12 15:22:31 2006 -0700 176689 176690 Add arrayobj.c to the Makefile as well. 176691 176692commit f9f33b72e34eaeccea2a20f4a3dd68c2dbefc90e 176693Author: Michel Dänzer <michel@tungstengraphics.com> 176694Date: Mon Jun 12 20:19:11 2006 +0200 176695 176696 Track per-drawable damage to minimize UTS and DFS transfers. 176697 176698 Based on work by Eric Anholt. 176699 176700commit 6060b612de6b41f872d034c6130770c1d189d0a3 176701Author: Eric Anholt <eric@anholt.net> 176702Date: Mon Jun 12 20:12:31 2006 +0200 176703 176704 Provide option to report damage after operation is complete. 176705 176706commit fc3e81473a98d3b932faa73c0c01c21366a48a9e 176707Merge: 4bedaad17 041ef2319 176708Author: Ian Romanick <idr@localhost.localdomain> 176709Date: Mon Jun 12 09:41:06 2006 -0700 176710 176711 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176712 176713commit 041ef23192b193b87f6cfc3e74e2e77f9f47cd4b 176714Author: Ian Romanick <idr@localhost.localdomain> 176715Date: Mon Jun 12 09:39:18 2006 -0700 176716 176717 Add new Mesa files arrayobj.c and arrayobj.h. 176718 176719commit 4bedaad17dc75dfbada91f47e1007a8c090300a9 176720Author: Ian Romanick <idr@localhost.localdomain> 176721Date: Mon Jun 12 09:29:05 2006 -0700 176722 176723 Add new Mesa files arrayobj.c and arrayobj.h. 176724 176725commit 4e3c9ab8c90d474a2224dbfa0807bc01fc012f7c 176726Author: Ian Romanick <idr@localhost.localdomain> 176727Date: Fri Jun 9 16:56:54 2006 -0700 176728 176729 Remove some static declarations for things that were removed in 176730 previous commits. 176731 176732commit 8b6596b1e1864dd79a3c53ee3ceb7a8e9bdb6ec7 176733Author: Ian Romanick <idr@localhost.localdomain> 176734Date: Fri Jun 9 16:08:06 2006 -0700 176735 176736 Move probe_devices_from_device_sections, 176737 add_matching_devices_to_configure_list, and check_for_matching_devices 176738 from xf86Helper.c to xf86Init.c. These functions are only called from 176739 a single place in xf86Init.c, so it makes sense for them to live 176740 there. They were originally in xf86Helper.c because they evolved out 176741 another function that was already in that file. 176742 176743commit 7810d87cd6ae9316d88f6a2e7696deb8837cec27 176744Merge: dcfaa9f4f caad8b724 176745Author: Ian Romanick <idr@localhost.localdomain> 176746Date: Fri Jun 9 15:16:50 2006 -0700 176747 176748 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176749 176750commit caad8b724b97074e41de447fe77dda189f287a26 176751Author: Greg Kroah-Hartman <gregkh@suse.de> 176752Date: Fri Jun 9 11:24:57 2006 -0700 176753 176754 fix compiler warnings in record/set.c 176755 176756 (note this only fixed up the function definitions for the static functions 176757 which can not cause any abi incompatibility) 176758 176759commit 51489bb5ed86cb6aa07e26a13618765c29f913e4 176760Author: Greg Kroah-Hartman <gregkh@suse.de> 176761Date: Fri Jun 9 10:59:47 2006 -0700 176762 176763 more .gitignore updates 176764 176765commit c4d251bd3e88cf8dfd6872537dbe30c07344b196 176766Author: Greg Kroah-Hartman <gregkh@suse.de> 176767Date: Fri Jun 9 10:46:18 2006 -0700 176768 176769 updated .gitignore with more pre-generated files 176770 176771commit 9f31ef83be61a900c701fcbc9a43ffae40ca7005 176772Author: Greg Kroah-Hartman <gregkh@suse.de> 176773Date: Fri Jun 9 10:44:46 2006 -0700 176774 176775 fix compiler warning in hw/xfree86/loader/loaderProcs.h 176776 176777commit 490ffc205a7714145cac0c63efeb6374ea28141f 176778Author: Greg Kroah-Hartman <gregkh@suse.de> 176779Date: Fri Jun 9 10:44:25 2006 -0700 176780 176781 fix compiler warning in hw/xfree86/loader/loadmod.c 176782 176783commit 6119845d1ff832ea2b7c9cbe7ed6c6637cdcf305 176784Author: Greg Kroah-Hartman <gregkh@suse.de> 176785Date: Fri Jun 9 10:14:08 2006 -0700 176786 176787 fix compiler warnings in hw/xnest/Keyboard.c 176788 176789commit 1a7335ff932baa59a3283c50dd6007d81989b7e3 176790Author: Greg Kroah-Hartman <gregkh@suse.de> 176791Date: Fri Jun 9 10:02:08 2006 -0700 176792 176793 remove unneeded externs from Xprint/ps/psout.c 176794 176795commit 9583859d538394e98ac1f38b8e6f0997e321621d 176796Author: Greg Kroah-Hartman <gregkh@suse.de> 176797Date: Fri Jun 9 09:59:22 2006 -0700 176798 176799 remove unneeded "extern" in Xprint/ddxInit.c 176800 176801commit cb9e29c184474edd75645e3b52e22a097a242e40 176802Author: Greg Kroah-Hartman <gregkh@suse.de> 176803Date: Fri Jun 9 09:54:50 2006 -0700 176804 176805 removed unneeded extern in Xprint/ps/PsWindow.c 176806 176807commit 96c19a3ec1b7e43782d373b8015fa0ca24cb4f5b 176808Author: Greg Kroah-Hartman <gregkh@suse.de> 176809Date: Fri Jun 9 09:51:28 2006 -0700 176810 176811 put function prototype for ShapeExtensionInit() in proper place 176812 176813 based on FIXME in mi/miinitext.c 176814 176815commit dcfaa9f4f8fb10b75be50c9a49ccd88caf6d0f6a 176816Merge: af9afe4cc 576e6fb11 176817Author: Ian Romanick <idr@localhost.localdomain> 176818Date: Thu Jun 8 15:03:00 2006 -0700 176819 176820 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176821 176822commit 576e6fb1124a47493371210adf99d6f2076e72c5 176823Author: Adam Jackson <ajax@benzedrine.nwnk.net> 176824Date: Thu Jun 8 17:49:02 2006 -0400 176825 176826 Bug #7120: Multimonitor (non-Xinerama) support for xwin servers. 176827 (Tom Whittock) 176828 176829commit d42cf4a2a4980fdd29fb15c4fd8fddda67b36256 176830Author: Adam Jackson <ajax@benzedrine.nwnk.net> 176831Date: Thu Jun 8 17:46:53 2006 -0400 176832 176833 Bug #7121: fix clipboard handling in Xming (Colin Harrison) 176834 176835commit af9afe4cc50df973bea2293a803f7dfdefaa6411 176836Merge: caffac384 3930da3f6 176837Author: Ian Romanick <idr@localhost.localdomain> 176838Date: Thu Jun 8 11:37:09 2006 -0700 176839 176840 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176841 176842commit 3930da3f6209312dd0f10aba0b16ef45996a07fe 176843Author: Greg Kroah-Hartman <gregkh@suse.de> 176844Date: Thu Jun 8 11:27:29 2006 -0700 176845 176846 fix compiler warnings in Xprint/ps/psout.c 176847 176848commit c496a3b9c981dc079fcc6c0ac4db3aa912b3dcf1 176849Author: Greg Kroah-Hartman <gregkh@suse.de> 176850Date: Thu Jun 8 11:23:53 2006 -0700 176851 176852 fix compiler warning in Xprint/ps/PsImageUtil.c 176853 176854commit ea24b5a25c2544f3b3de6480da125edb23a6b3a9 176855Author: Greg Kroah-Hartman <gregkh@suse.de> 176856Date: Thu Jun 8 11:22:01 2006 -0700 176857 176858 add proper function prototypes for Xprint/ps/PsCache.c to Xprint/ps/Ps.h 176859 176860commit ee2bb4d1929e20436cf0e830ece02fe07db2d524 176861Author: Greg Kroah-Hartman <gregkh@suse.de> 176862Date: Thu Jun 8 11:15:34 2006 -0700 176863 176864 fix compiler warnings in Xprint/ps/PsArea.c 176865 176866commit 92303d534a91cb0ea30e4cd0f639efd70b9739b4 176867Author: Greg Kroah-Hartman <gregkh@suse.de> 176868Date: Thu Jun 8 11:12:16 2006 -0700 176869 176870 fix compiler warning in Xprint/ps/PsGC.c 176871 176872commit cf6169f9e99e6e8ab264f284cfa13cb379b36207 176873Author: Greg Kroah-Hartman <gregkh@suse.de> 176874Date: Thu Jun 8 11:09:56 2006 -0700 176875 176876 fix compiler warnings in Xprint/ps/PsFonts.c 176877 176878commit 1abc7f96edf37a1e2c766b9cdba7fc9b2cb06d19 176879Author: Greg Kroah-Hartman <gregkh@suse.de> 176880Date: Thu Jun 8 11:06:51 2006 -0700 176881 176882 fix compiler warnings in Xprint/ps/PsInit.c 176883 176884commit 2dc291384c550badf55542ae645240e166676848 176885Author: Greg Kroah-Hartman <gregkh@suse.de> 176886Date: Thu Jun 8 11:04:40 2006 -0700 176887 176888 fixed compiler warnings in Xprint/ps/PsPixmap.c 176889 176890commit aef092e0290143c2b8b1cb98fdf55c9630032aaf 176891Author: Greg Kroah-Hartman <gregkh@suse.de> 176892Date: Thu Jun 8 10:53:17 2006 -0700 176893 176894 fix compiler warnings in Xprint/ps/PsPolygon.c 176895 176896commit 511b231ded61159ebd70cab020ca1ca003fd0784 176897Author: Greg Kroah-Hartman <gregkh@suse.de> 176898Date: Thu Jun 8 10:41:34 2006 -0700 176899 176900 fix compiler warnings in Xprint/ps/PsPrint.c 176901 176902commit 35fccb0068e8d73d1e6a16aefdc771506e620f83 176903Author: Greg Kroah-Hartman <gregkh@suse.de> 176904Date: Thu Jun 8 10:40:24 2006 -0700 176905 176906 remove some compiler warnings in Xprint/ps/PsText.c 176907 176908 Note that one of the existing warnings is pointing out a 176909 real bug (uninitialized use for fontPage in PsPolyText16()) 176910 if anyone really cares about this code. 176911 176912commit caffac38432009e8dc01e7d3e72a7fa6fba2f078 176913Merge: c5b3aa120 9fa73721f 176914Author: Ian Romanick <idr@localhost.localdomain> 176915Date: Thu Jun 8 10:38:43 2006 -0700 176916 176917 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 176918 176919commit 1ac30947f4a222ba78558eddf8e5f03cec31f613 176920Author: Greg Kroah-Hartman <gregkh@suse.de> 176921Date: Thu Jun 8 10:35:18 2006 -0700 176922 176923 fix compiler warnings in Xprint/ps/PsWindow.c 176924 176925commit 9fa73721f0c3df73e508da909a5665f47a54cb57 176926Author: Greg Kroah-Hartman <gregkh@suse.de> 176927Date: Thu Jun 8 10:27:28 2006 -0700 176928 176929 fix up EnableDisableExtension() and EnableDisableExtensionError() prototypes 176930 176931commit c405659626477f2009603d8c0e381b8b62277bb6 176932Author: Greg Kroah-Hartman <gregkh@suse.de> 176933Date: Thu Jun 8 10:19:24 2006 -0700 176934 176935 comment out QualifyName in Xprint/Init.c which is not used anymore 176936 176937 If someone else wants to delete this function, that's fine with me too. 176938 176939commit a940b851faba569e36983f7885aefa72f7bf2ade 176940Author: Greg Kroah-Hartman <gregkh@suse.de> 176941Date: Thu Jun 8 10:17:53 2006 -0700 176942 176943 fix function pointer warning in Xprint/Init.c 176944 176945commit 9e0c82386ae389bcc296a5ad44e996790b033ad3 176946Author: Greg Kroah-Hartman <gregkh@suse.de> 176947Date: Thu Jun 8 10:16:37 2006 -0700 176948 176949 fix already-defined warning in Xprint/Init.c 176950 176951commit 40fb7eecaf14a76f35ac2dc350ad2fffdaf6e0d0 176952Author: Greg Kroah-Hartman <gregkh@suse.de> 176953Date: Thu Jun 8 10:11:17 2006 -0700 176954 176955 fix noDamageExtension warning in Xprint/Init.c 176956 176957 Also took out duplicate definition of this variable in mi/miinitext.c 176958 176959commit 494895e0fbbf0a71bc535c0a2358c9db54c95c5a 176960Author: Greg Kroah-Hartman <gregkh@suse.de> 176961Date: Thu Jun 8 09:43:44 2006 -0700 176962 176963 Properly define dispatchExceptionAtReset to fix compiler warnings 176964 176965commit d90eecf40ea768b2bf6340f15bb0af9dab2f3cf3 176966Author: Greg Kroah-Hartman <gregkh@suse.de> 176967Date: Thu Jun 8 09:11:41 2006 -0700 176968 176969 add some missing function prototypes to Xprint/AttrValid.h to fix compiler warnings 176970 176971commit bccde1609153dee93f6fe5a138fc0c0f2fe08212 176972Author: Greg Kroah-Hartman <gregkh@suse.de> 176973Date: Thu Jun 8 09:01:59 2006 -0700 176974 176975 fix incompatible pointer warning in Xprint/ddxInit.c 176976 176977commit 60bd8893d50ed1da9b94f4b96a07ea432e23f467 176978Author: Greg Kroah-Hartman <gregkh@suse.de> 176979Date: Thu Jun 8 09:01:21 2006 -0700 176980 176981 Properly #ifdef out ddxBeforeReset() to fix compiler warning 176982 176983commit ac21e6a594eac69101aa8920d70a9d60412b57f6 176984Author: Greg Kroah-Hartman <gregkh@suse.de> 176985Date: Thu Jun 8 09:00:42 2006 -0700 176986 176987 remove unused variable warning in Xprint/ddxInit.c 176988 176989commit 71dd44b0ad617dd36ce4ed328f9e1e8c5ef713a5 176990Author: Greg Kroah-Hartman <gregkh@suse.de> 176991Date: Wed Jun 7 19:17:26 2006 -0700 176992 176993 Fixed up most "warning: function declaration isn't a prototype" warnings from Xprint/ 176994 176995commit 839305bac98856a2bb1d96691e4dcf49db229f90 176996Author: Greg Kroah-Hartman <gregkh@suse.de> 176997Date: Wed Jun 7 19:12:23 2006 -0700 176998 176999 Remove unused variables from Xprint/attributes.c 177000 177001commit c5b3aa120bf96ee169d07c33a2698499944a1624 177002Author: Ian Romanick <idr@localhost.localdomain> 177003Date: Wed Jun 7 17:46:23 2006 -0700 177004 177005 Added a comment about some suspicious code. 177006 177007commit 50112dfcc82dd476727fa453b9e3d1852e06d678 177008Author: Ian Romanick <idr@localhost.localdomain> 177009Date: Wed Jun 7 17:46:05 2006 -0700 177010 177011 Fixed severly broken calls to pci_device_find_by_slot. I mis-read the 177012 tag bits when I originally coded it. I mistakenly thought that the 177013 original code masked off the PCI function bits, so I just always 177014 passed 0 as the function. That was a mistake. 177015 177016commit f9e7128d948a0b62e6f03295305e0080d9233fd8 177017Author: Ian Romanick <idr@localhost.localdomain> 177018Date: Wed Jun 7 17:44:06 2006 -0700 177019 177020 Added missing include of Pci.h. This caused the server to terminate 177021 when the symbol PCI_DOM_FROM_TAG couldn't be resolved at run-time. 177022 177023commit 23182315f34e3a0065739b43b15d4560e75b41a1 177024Merge: bc05158a5 757f40fca 177025Author: Ian Romanick <idr@localhost.localdomain> 177026Date: Wed Jun 7 17:03:48 2006 -0700 177027 177028 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 177029 177030commit 7a40ac2585028860730ffcd333eb3fe9de63680c 177031Author: Greg Kroah-Hartman <gregkh@suse.de> 177032Date: Wed Jun 7 16:23:45 2006 -0700 177033 177034 remove a bunch of unused variables in Xprint/Init.c (fixing the compiler warnings) 177035 177036commit 163980138cc0bfc9124456781b3dc45a49e2a129 177037Author: Greg Kroah-Hartman <gregkh@suse.de> 177038Date: Wed Jun 7 16:11:20 2006 -0700 177039 177040 remove some unused local variables in Xprint/Oid.c 177041 177042commit 8e41640db884a4633b598d0a52b269e6547c8bf0 177043Author: Greg Kroah-Hartman <gregkh@suse.de> 177044Date: Wed Jun 7 15:56:43 2006 -0700 177045 177046 add bison generated files to .gitignore 177047 177048commit 78f4ab6b89fca3086b9c9471b40c11c23fbb6142 177049Author: Greg Kroah-Hartman <gregkh@suse.de> 177050Date: Wed Jun 7 14:12:40 2006 -0700 177051 177052 Fix compiler warning about undefined ReinitializeRootWindow function 177053 177054commit 757f40fca50a99377e437949ee77b983c8cd6087 177055Author: Greg Kroah-Hartman <gregkh@suse.de> 177056Date: Wed Jun 7 14:09:13 2006 -0700 177057 177058 updated the .gitignore file with more auto-generated files 177059 177060commit bc05158a5ab00c548e4b04b5638afe07bc702260 177061Merge: 46f55f5de cc465800d 177062Author: Ian Romanick <idr@localhost.localdomain> 177063Date: Wed Jun 7 14:09:12 2006 -0700 177064 177065 Merge branch 'master' of git+ssh://git.freedesktop.org/git/xorg/xserver 177066 177067commit 46f55f5dead5d70cdff30531d80a72f6be042315 177068Author: Ian Romanick <idr@localhost.localdomain> 177069Date: Wed Jun 7 14:09:02 2006 -0700 177070 177071 Initial batch of changes for PCI rework. All future changes will be 177072 tracked individually. 177073 177074commit cc465800ddca5fb6c9ec09fdfa8f1f05359cf396 177075Author: Greg Kroah-Hartman <gregkh@suse.de> 177076Date: Wed Jun 7 14:03:35 2006 -0700 177077 177078 Fix compiler warnings about SetVendorRelease and SetVendorString 177079 177080commit 785c9789704ed142fe98cd17b5995e4a95b7141f 177081Merge: 21ebcfd70 36d786e9f 177082Author: Greg Kroah-Hartman <gregkh@suse.de> 177083Date: Wed Jun 7 13:20:21 2006 -0700 177084 177085 Merge ../xserver 177086 177087commit 36d786e9f051c5c95c1cc8c098c84e118ed3cc85 177088Author: Greg Kroah-Hartman <gregkh@suse.de> 177089Date: Wed Jun 7 12:47:50 2006 -0700 177090 177091 add more files to .gitignore 177092 177093commit 8f5aa38abf1158a789b5528df9d98826342e30cf 177094Author: Greg Kroah-Hartman <gregkh@suse.de> 177095Date: Wed Jun 7 12:33:44 2006 -0700 177096 177097 fix compiler warning about XKB_IN_SERVER redefinition 177098 177099commit 101ae616962c355388722e05ab8413eb5f5c3402 177100Author: Greg Kroah-Hartman <gregkh@suse.de> 177101Date: Wed Jun 7 12:06:22 2006 -0700 177102 177103 Add PanoramiXExtensionDisabledHack to globals.h as it was missing. 177104 177105commit e5b72bd9c6fb06640a5de4031be0dc9b04b4b215 177106Author: Greg Kroah-Hartman <gregkh@suse.de> 177107Date: Wed Jun 7 12:05:39 2006 -0700 177108 177109 Remove 3 compiler warnings in the Xext/xevie.c file 177110 177111commit e3c11f66516521959127b9ab8fd88cc4c954f5bb 177112Author: Greg Kroah-Hartman <gregkh@suse.de> 177113Date: Wed Jun 7 12:05:01 2006 -0700 177114 177115 Added first cut at a .gitignore file to make using git easier. 177116 177117commit 21ebcfd7027b2a6182d4065e56a2ef814f5181ae 177118Author: Adam Jackson <ajax@benzedrine.nwnk.net> 177119Date: Wed Jun 7 14:17:31 2006 -0400 177120 177121 Demolish now-unused loader functions. 177122 177123commit f90761b06eaa5fa44fe85289e54eed5f47eff3b9 177124Author: Adam Jackson <ajax@benzedrine.nwnk.net> 177125Date: Wed Jun 7 13:58:24 2006 -0400 177126 177127 Add a token for EDID-supplied modes. 177128 177129commit d00aa6b8559d3e5f70c6558ce0abd12f7d758491 177130Author: Adam Jackson <ajax@benzedrine.nwnk.net> 177131Date: Wed Jun 7 11:44:36 2006 -0400 177132 177133 Delete a (now misleading) message from the crash handler. 177134 177135commit f8535edec736cf19740bd41ed2adfe531f2c26ac 177136Author: Keith Packard <keithp@neko.keithp.com> 177137Date: Tue Jun 6 10:29:34 2006 -0700 177138 177139 Remove ChangeLog file. 177140 177141commit 8444bb77c91cf8a23d32b3cc9749e2a3d3f9f9eb 177142Author: Daniel Stone <daniel@fooishbar.org> 177143Date: Mon Jun 5 20:22:06 2006 +0000 177144 177145 When we can, bound the maximum number of PCI devices to attempt to scan, by 177146 the number found on the system. Only implemented for Linux right now. 177147 177148commit 11cf4d2fde9219e8d6ca427acae89a0c9f5d71b1 177149Author: Keith Packard <keithp@keithp.com> 177150Date: Mon Jun 5 07:15:23 2006 +0000 177151 177152 Update AC_DEFINE_DIR to version compatible with CVS autoconf which requires 177153 double eval to avoid leaving ${prefix} in output 177154 177155commit 52fc7c8dc70226cc7f03454e9be86a627672295f 177156Author: Adam Jackson <ajax@nwnk.net> 177157Date: Mon Jun 5 03:00:24 2006 +0000 177158 177159 Ensure all *ModuleData symbols are marked _X_EXPORT. Start removing 177160 XFree86LOADER ifdefs, non-loadable hasn't been supported for a while 177161 now. Remove completely gratuitious REMOVE_LOADER_CHECK_MODULE_INFO 177162 ifdefs surrounding a call to a function added in XFree86 4.1 (!). 177163 Miscellaneous static markings. 177164 177165commit d22582dc5a070f72d4653e24d1e4ebe4a112276e 177166Author: Adam Jackson <ajax@nwnk.net> 177167Date: Sun Jun 4 16:13:26 2006 +0000 177168 177169 Remove a stray LBX reference. 177170 177171commit ddc6b99505e227f99585b8c2392da524022a73e6 177172Author: Daniel Stone <daniel@fooishbar.org> 177173Date: Sat Jun 3 11:24:33 2006 +0000 177174 177175 Bump to 1.1.99.2. 177176 177177commit 4fd668940f2155c4b06d24b6da8bcccd01f66f4c 177178Author: Daniel Stone <daniel@fooishbar.org> 177179Date: Sat Jun 3 10:58:37 2006 +0000 177180 177181 Bug #6619: Fix disappearing hardware cursor. (Colin Harrison) 177182 177183commit ee71cb61f8da29bcf36ea4b199d629e34f89b119 177184Author: Daniel Stone <daniel@fooishbar.org> 177185Date: Sat Jun 3 10:54:38 2006 +0000 177186 177187 Bug #6956: Fix crash when removing session leader before its children. 177188 (Rich Coe) 177189 177190commit cd384af3058fe15077c57eccdffed3b61e261e7f 177191Author: Daniel Stone <daniel@fooishbar.org> 177192Date: Sat Jun 3 10:50:23 2006 +0000 177193 177194 Completely remove relocation pointer table. 177195 177196commit 8e8c6faecddbe014d8760822e1f705b43a00fa33 177197Author: Daniel Stone <daniel@fooishbar.org> 177198Date: Sat Jun 3 10:48:37 2006 +0000 177199 177200 Add support for kFreeBSD systems. (Robert Millan, Aurelien Jarno) 177201 177202commit 5b703f847d166176920077c1e6ba1d9559fc8481 177203Author: Daniel Stone <daniel@fooishbar.org> 177204Date: Fri Jun 2 12:05:32 2006 +0000 177205 177206 Fix type confusion. 177207 177208commit 56f21bda1ce95741c88c423b60bd709eef26eb12 177209Author: Daniel Stone <daniel@fooishbar.org> 177210Date: Thu Jun 1 22:30:52 2006 +0000 177211 177212 Bug #6583: Only open /proc/bus/pci/devices once. (Bill Nottingham) 177213 177214commit a9ed5a87902a839a5a135af03db78f113b18bd86 177215Author: Daniel Stone <daniel@fooishbar.org> 177216Date: Thu Jun 1 22:06:41 2006 +0000 177217 177218 Kill LBX, too. 177219 177220commit df6da66525836d515f408a82f1a13ca5251ff0f7 177221Author: Daniel Stone <daniel@fooishbar.org> 177222Date: Thu Jun 1 20:56:39 2006 +0000 177223 177224 Simplify the unsupported XI function list. 177225 177226commit 97203f1cf6e5b7c6389f69cbb1b75ac675d09531 177227Author: Daniel Stone <daniel@fooishbar.org> 177228Date: Thu Jun 1 20:41:21 2006 +0000 177229 177230 Clean up a warning, and remove excess multiple-suffix code. 177231 177232commit 6d594ebc667afd404556ec3e108c810946b20ac5 177233Author: Daniel Stone <daniel@fooishbar.org> 177234Date: Thu Jun 1 20:22:39 2006 +0000 177235 177236 Ditch more alternate-loader braindamage. 177237 177238commit c9468177486833d521ec62c7b0266b4be8200de7 177239Author: Daniel Stone <daniel@fooishbar.org> 177240Date: Thu Jun 1 20:18:30 2006 +0000 177241 177242 Kill a.out, COFF and ELF loaders with FIRE. 177243 177244commit a3a4221495dfe4cc0a3874a08dd5364ef45a7f2e 177245Author: Adam Jackson <ajax@nwnk.net> 177246Date: Thu Jun 1 19:53:06 2006 +0000 177247 177248 Add bitmap to the ignored module list. 177249 177250commit ee689c104287140db38bbd26959ab1b4847c168e 177251Author: Daniel Stone <daniel@fooishbar.org> 177252Date: Thu Jun 1 19:49:55 2006 +0000 177253 177254 Ignore requests to load GLcore and speedo. 177255 177256commit 07c731a2bc21e6b98f28a2c0ebc42f01b67b824b 177257Author: Adam Jackson <ajax@nwnk.net> 177258Date: Thu Jun 1 19:37:53 2006 +0000 177259 177260 Bug #5089: Die, libbitmap, die! 177261 177262commit 32be08ba7242da74de5defd6a4dcb536a273f57a 177263Author: Daniel Stone <daniel@fooishbar.org> 177264Date: Thu Jun 1 19:22:38 2006 +0000 177265 177266 Remove horrendously ugly DDX backward-compatibility. 177267 177268commit d81edb9e00680e3c0001f343fa1d0c310b86cb93 177269Author: Daniel Stone <daniel@fooishbar.org> 177270Date: Thu Jun 1 19:22:01 2006 +0000 177271 177272 Forgot to remove this one too. 177273 177274commit a73e0f8cdfec1c9199ffe696146ba7d677c4c10d 177275Author: Daniel Stone <daniel@fooishbar.org> 177276Date: Thu Jun 1 18:47:47 2006 +0000 177277 177278 Die XTESTEXT1, die! 177279 177280commit fc5ca97284ef237a91f6adb433148ff57a673c08 177281Author: Matthieu Herrb <matthieu.herrb@laas.fr> 177282Date: Tue May 30 22:56:36 2006 +0000 177283 177284 Powerpc machines also need ioperm_noop.c 177285 177286commit 07b168c8d8b2d79a4e7cf8dc5124eafc0f2bbcd0 177287Author: Adam Jackson <ajax@nwnk.net> 177288Date: Tue May 30 16:10:59 2006 +0000 177289 177290 Properly document the DPMS, SyncOnGreen, and TargetRefresh options. 177291 177292commit 107defd920d9b1eb52b15e8ca8665bc48bb933d6 177293Author: Kristian Høgsberg <krh@redhat.com> 177294Date: Mon May 29 19:53:18 2006 +0000 177295 177296 Remove superfluous definition of tfp tokens. 177297 177298commit fd8bde8bb0f9d796b3464973b53285c0a6d22a31 177299Author: Daniel Stone <daniel@fooishbar.org> 177300Date: Mon May 29 11:14:03 2006 +0000 177301 177302 Remove -xkbmap argument. 177303 177304commit db0680cf70b8367e1f8a7fff9c0f6ec414db0542 177305Author: Daniel Stone <daniel@fooishbar.org> 177306Date: Mon May 29 09:26:32 2006 +0000 177307 177308 Minor #include cleanups. 177309 177310commit c38aab293a06f43c04f14223f94f822d91d73396 177311Author: Matthieu Herrb <matthieu.herrb@laas.fr> 177312Date: Sat May 27 23:15:05 2006 +0000 177313 177314 check buffersize before reading next char, fixes the "mouse stuck at left" 177315 bug (bugzilla #3113) for BSD systems using wscons. (Dale Rahn). 177316 177317commit 2102fdd0a58e18aeaa842b2ec73b6071970fafb7 177318Author: Jeremy C. Reed <reed@reedmedia.net> 177319Date: Fri May 26 00:12:18 2006 +0000 177320 177321 RGB color database and XErrorDB install to "share" not "lib" by default (by 177322 app/rgb and libX11). 177323 TODO: They are customizable, so maybe cpprules.in should be extended. 177324 177325commit 354086d7e8e13fc7acbcc6603ca2a03c8cc806ee 177326Author: Jeremy C. Reed <reed@reedmedia.net> 177327Date: Thu May 25 23:32:33 2006 +0000 177328 177329 Don't do fixup_video_driver_list if no drivers. (I had signal 11 and core 177330 dump when drivers not installed yet.) 177331 TODO: fix fixup_video_driver_list to handle NULL argument. 177332 177333commit c6b1cff43238deded11dc58945778aec3d844598 177334Author: Matthieu Herrb <matthieu.herrb@laas.fr> 177335Date: Thu May 25 13:27:42 2006 +0000 177336 177337 update to build against Mesa CVS HEAD (Carlos Eduardo Rodrigues Diogenes). 177338 177339commit 99724c16c9c58eb3e20ba91c79464747b8ee3fcf 177340Author: Matthieu Herrb <matthieu.herrb@laas.fr> 177341Date: Thu May 25 10:14:39 2006 +0000 177342 177343 typo 177344 177345commit 5c7aef148de23f39027fda647bbb53bb5b992683 177346Author: Matthieu Herrb <matthieu.herrb@laas.fr> 177347Date: Thu May 25 09:57:47 2006 +0000 177348 177349 Don't destroy a pixmap twice on server exit (bugzilla #4247). 177350 177351commit 693079442377daa2dc48bf318a0d7dd256cb2738 177352Author: Jeremy C. Reed <reed@reedmedia.net> 177353Date: Wed May 24 22:58:03 2006 +0000 177354 177355 NetBSD also has curses.h instead of ncurses.h. (TODO: Maybe instead of 177356 checking for SCO, UnixWare, Sun SVR4 and NetBSD, the curses.h versus 177357 ncurses.h should be tested in configure.) 177358 177359commit 9477e5e0be04f4f6fa311de8b3693fbc8a082fbf 177360Author: Jeremy C. Reed <reed@reedmedia.net> 177361Date: Wed May 24 22:43:13 2006 +0000 177362 177363 On NetBSD, add -li386 for i386 and -lx86_64 for x86_64 to SYS_LIBS. (x86_64 177364 on NetBSD not tested.) 177365 177366commit b3031532ca96b22e81863202efb8bbcb9d701fac 177367Author: Jeremy C. Reed <reed@reedmedia.net> 177368Date: Wed May 24 20:11:38 2006 +0000 177369 177370 Remove tolower() which was missed in the _XkbStrCaseCmp/strcmp changes. 177371 177372commit eb696f72a0819edef550bce9ff55730c02f70452 177373Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 177374Date: Tue May 23 16:17:09 2006 +0000 177375 177376 Add AC_SUBST([SOLARIS_ASM_CFLAGS]) that I forgot when splitting them out of 177377 the XORG_CFLAGS. 177378 177379commit bc0c56c407117d1545e20d21f7d30eb3472d618b 177380Author: Adam Jackson <ajax@nwnk.net> 177381Date: Mon May 22 15:47:56 2006 +0000 177382 177383 Bug #6924: Restore the ABI for DrawableRec and ColormapRec to the state 177384 they were in prior to the fix for #6438. Based on a patch from Andy 177385 Ritger. 177386 177387commit cc3b882bd141218052cdde0144fc2a707ceee83d 177388Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 177389Date: Sat May 20 00:55:44 2006 +0000 177390 177391 Update to 2006-05-19 snapshot 177392 Remove entries now present in pci.ids 177393 Regenerate from updated pci.ids & extrapci.ids 177394 177395commit 0b2c2b6bba5b578d5f0ac2d709d5d2ce7cb32bef 177396Author: Adam Jackson <ajax@nwnk.net> 177397Date: Fri May 19 20:29:47 2006 +0000 177398 177399 Remove dead DPMS timer functions from the installed headers. (Fredrik 177400 Höglund) 177401 177402commit deebf6bd51117c01a3217f134bd952481b9e41ab 177403Author: Adam Jackson <ajax@nwnk.net> 177404Date: Thu May 18 23:52:51 2006 +0000 177405 177406 Bug #4139: Fix a BAR remapping bug that could lead to IERR and system hang. 177407 (Egbert Eich) 177408 177409commit 91239d83f4e27835cf871348b5ff6c892bd4f4f4 177410Author: Adam Jackson <ajax@nwnk.net> 177411Date: Thu May 18 23:52:41 2006 +0000 177412 177413 file getemptypci.c was initially added on branch server-1_1-branch. 177414 177415commit c11cfcfaffc79be5686c666f881c4c08f69e1b86 177416Author: Adam Jackson <ajax@nwnk.net> 177417Date: Thu May 18 23:48:57 2006 +0000 177418 177419 Bug #6377: Ignore disabled BARs, and allow matching BARs aligned to less 177420 than 16 bytes. (Felix Kühling, ATI) 177421 177422commit fa9a49a92db52de968d7147c71c6b9a8fd480f1e 177423Author: Adam Jackson <ajax@nwnk.net> 177424Date: Thu May 18 18:18:41 2006 +0000 177425 177426 Bug #5877: Avoid burning CPU when acpid dies. Require acpid to be running 177427 for ACPI support on Linux. Minor errno handling fixes. (Valery 177428 Inozemtsev, Adam Jackson) 177429 177430commit 7893dadb2f6df218a4f4ea30a41c1aa9838da1f0 177431Author: Adam Jackson <ajax@nwnk.net> 177432Date: Thu May 18 14:39:59 2006 +0000 177433 177434 Bug #6827: Fix texel fetch in fbFetchTransformed to avoid crashes. Still 177435 not 100% correct, but better than 7.0. (Radek Doulik, Matthias Hopf) 177436 177437commit 2892dd6d2e34957650ef1630a94d471dfa71f888 177438Author: Daniel Stone <daniel@fooishbar.org> 177439Date: Wed May 17 16:20:03 2006 +0000 177440 177441 Make Xv symbols conditional. (Enrico Wiegelt) 177442 177443commit a317bf482257f0e1b612dec7961fdfa564f0b9f2 177444Author: Daniel Stone <daniel@fooishbar.org> 177445Date: Wed May 17 15:00:18 2006 +0000 177446 177447 Make DBE support conditional. (Enrico Wiegelt) 177448 177449commit 321dbed5f5a857a23525167ab85d4d7699429132 177450Author: Adam Jackson <ajax@nwnk.net> 177451Date: Mon May 15 18:27:18 2006 +0000 177452 177453 Bug #5209: Fix APM/ACPI support, again. (Michel Dänzer, Valery Inozemtsev) 177454 177455commit dd38d3bd673cf830c2cd591fe1245909aa729892 177456Author: Matthieu Herrb <matthieu.herrb@laas.fr> 177457Date: Sun May 14 16:22:24 2006 +0000 177458 177459 Don't use AM_CONDITIONAL inside shell conditionals. Bugzilla #6916. 177460 177461commit 6558ba4e62dba99f2a4d830f1c16f8d1c255b316 177462Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 177463Date: Fri May 12 21:53:28 2006 +0000 177464 177465 Fri May 12 17:51:26 2006 Søren Sandmann <sandmann@redhat.com> 177466 Keep track of the previous item and update its next pointer when deleting 177467 from the linked list. 177468 177469commit 9e239a0df7ededb50de091e5271cbfddd2b683c3 177470Author: Matthias Hopf <mhopf@suse.de> 177471Date: Thu May 11 11:23:43 2006 +0000 177472 177473 Bug #5796: Wrong fastpath selection for repeating sources. 177474 177475commit 9db5d2dfc33e384ea4db1b7cbc377e0b05cfb3ff 177476Author: Matthias Hopf <mhopf@suse.de> 177477Date: Thu May 11 10:18:08 2006 +0000 177478 177479 Bug #4320: Improved XAA Composite fastpath. 177480 177481commit 9a2a63ca3ff30d15e82a29e75a3720ba5b446978 177482Author: Adam Jackson <ajax@nwnk.net> 177483Date: Wed May 10 15:44:27 2006 +0000 177484 177485 Bug #3561: Crash fix in the Record extension. (Paul Anderson) 177486 177487commit ab1a0249ba5e3174f18a1db212bc511fd7d74cb0 177488Author: Adam Jackson <ajax@nwnk.net> 177489Date: Tue May 9 22:36:01 2006 +0000 177490 177491 Bug #6867: Yet another Render crash fix. (Michel Dänzer) 177492 177493commit 24310f827b71009c7510a674d2f92ced89847e37 177494Author: Adam Jackson <ajax@nwnk.net> 177495Date: Tue May 9 18:12:50 2006 +0000 177496 177497 Revert accidental commit 177498 177499commit 86ffb46358965509aa3ee536f15cb5a4e5e04426 177500Author: Adam Jackson <ajax@nwnk.net> 177501Date: Tue May 9 18:04:29 2006 +0000 177502 177503 Bug #5209: Re-enable building APM and ACPI support. (Michel Dänzer) 177504 177505commit b46d6a44fa97a3e66de828385026b7f84d9e59b8 177506Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 177507Date: Wed May 3 23:45:16 2006 +0000 177508 177509 Make Xephyr build on Solaris: 177510 Add #include <sys/file.h> on Solaris for FNONBLOCK/FASYNC definitions 177511 hw/kdriver/linux/Makefile.am Move agp.c & agp.h to KDRIVE_HW_SOURCES since 177512 they're not needed for Xephyr-only builds 177513 Add -lrt to XEPHYR_LIBS if needed to get nanosleep(). 177514 177515commit fc91ca069dd55490b99b096f029e0864b049120c 177516Author: Adam Jackson <ajax@nwnk.net> 177517Date: Wed May 3 17:50:10 2006 +0000 177518 177519 Redact a few mentions of speedo font support. 177520 177521commit b9a9cf618566bdd796556b8a1f31949f66184352 177522Author: Matthieu Herrb <matthieu.herrb@laas.fr> 177523Date: Tue May 2 14:09:30 2006 +0000 177524 177525 Typo in ALLLOCATE_LOCAL() arguments, causing mis-computation of the buffer. 177526 Bugzilla #6642. 177527 177528commit 1e8a594957d84a37e66183e9c0cb9d42b62bdb24 177529Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 177530Date: Tue May 2 01:37:25 2006 +0000 177531 177532 Fix Solaris build with Sun compilers to work when exa is built before 177533 hw/xfree86/os-support/solaris (as it is by default now). 177534 177535commit 724dbc2f8bbe2f21bf16f20ca7b8bb555516626c 177536Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 177537Date: Tue May 2 01:30:37 2006 +0000 177538 177539 Use min() [defined in include/misc.h] instead of MIN() [not defined in any 177540 Xorg header]. 177541 177542commit ad124742ae2a265a54a4a7ac91709ec6fd6ced34 177543Author: Adam Jackson <ajax@nwnk.net> 177544Date: Sun Apr 30 20:33:27 2006 +0000 177545 177546 Remove stray mfb/cfb references. 177547 177548commit 291e89d4f2a4bb3177b2dfe6421680e23f120b8e 177549Author: Adam Jackson <ajax@nwnk.net> 177550Date: Sun Apr 30 19:16:14 2006 +0000 177551 177552 Remove NEED_LINEHELPER BC cruft for pre-R6 DDXes. 177553 177554commit fdcacc5a4bc1d6cb7347fc66041a9c686c5c74d1 177555Author: Eric Anholt <anholt@freebsd.org> 177556Date: Fri Apr 28 03:27:12 2006 +0000 177557 177558 Add Polylines and PolyPoint acceleration as well. This is primarily to 177559 clean up fallback debugging output, so I can focus on more imporant 177560 cases. Performance is comparable but without hardware stalls, and 177561 passes Xlib9. 177562 177563commit 005529a1c9c9e78f06565dff081f03b74988081e 177564Author: Eric Anholt <anholt@freebsd.org> 177565Date: Fri Apr 28 03:26:30 2006 +0000 177566 177567 Correct some bugs causing performance issues in the "Smart" scheme. 177568 177569commit 8738bc295bba229e36d064713e0c28aa8720c494 177570Author: Eric Anholt <anholt@freebsd.org> 177571Date: Thu Apr 27 20:27:27 2006 +0000 177572 177573 Improve EXA fallback debugging output to include the locations of pixmaps. 177574 This is being used in tracking down recent compositing performance 177575 regressions. 177576 177577commit 83b061776a57025076fc26d6d01fe2e049c2243b 177578Author: Eric Anholt <anholt@freebsd.org> 177579Date: Thu Apr 27 19:17:34 2006 +0000 177580 177581 Add trivial PolySegment acceleration for 0-width horizontal/vertical lines, 177582 which still happen somewhat frequently and were cluttering up my 177583 fallback debugging output. x11perf says it's a major performance win in 177584 those cases (though probably irrelevant), and it passes Xlib9. 177585 177586commit 69164ec00c749787dd59d5913ec6b3d159ad74d7 177587Author: Eric Anholt <anholt@freebsd.org> 177588Date: Thu Apr 27 02:15:19 2006 +0000 177589 177590 In drawing glyphs, shortcut our way to exaComposite instead of going 177591 through the whole CompositePicture stack and doing things like 177592 computing damage over again. This is a sizeable win for text drawing 177593 with a compmgr. Also avoid calling down into the server for dealing 177594 with the scratch pixmap when we are able to do UploadToScreen 177595 successfully and never need it. 177596 177597commit 3d4ca57b69c40d27fe191170d0819013f8cc4947 177598Author: Eric Anholt <anholt@freebsd.org> 177599Date: Wed Apr 26 18:27:40 2006 +0000 177600 177601 Add a helper for the Component Alpha Over case, which breaks the operation 177602 down into an OutReverse and an Add. Turn off the fallback to software 177603 glyphs when component alpha, now that we expect all (new) drivers to be 177604 able to support it. Also, make Xephyr fall back in the CA Over case to 177605 exercise this code. This speeds up my rgb24text and ls -lR in 177606 gnome-terminal by a factor of 5. 177607 177608commit 26fa45b64258894201496f921eccb0cb7028c28c 177609Author: Dave Airlie <airlied@linux.ie> 177610Date: Wed Apr 26 11:40:58 2006 +0000 177611 177612 Bug #6751: Use the Linux PCI ROM interface on Linux properly. The old 177613 domain code, still modified BARs not a good idea, Just talk to sysfs, 177614 if 0 read, fallback to old methods. 177615 177616commit c339b221d3f59130a39e63d4cec3de7e3de95bf3 177617Author: Dave Airlie <airlied@linux.ie> 177618Date: Wed Apr 26 11:31:07 2006 +0000 177619 177620 Bug #6750: This patch detects Intel bridges that are transparent but aren't 177621 reported as such. From the Linux kernel fixups. This patch also removes 177622 the reserved BIOS area from the area to allocate resources in. 177623 177624commit 6d156c044085881c6ecbd8a13521c0a26df71c30 177625Author: Eric Anholt <anholt@freebsd.org> 177626Date: Wed Apr 26 01:33:15 2006 +0000 177627 177628 Improve the migration debugging output. 177629 177630commit eaed7545a25a08b3223bf620c2ab6f80fe7cfdf6 177631Author: Eric Anholt <anholt@freebsd.org> 177632Date: Wed Apr 26 01:32:55 2006 +0000 177633 177634 Fix a bug in the intersection computation that could concievably cause 177635 incorrect results to be returned (but would probably usually be 177636 over-conservative). 177637 177638commit 5d00859c6e7d4b7ebce56f438ec4993334de2328 177639Author: Eric Anholt <anholt@freebsd.org> 177640Date: Tue Apr 25 23:56:17 2006 +0000 177641 177642 Bug #4668: Check if the lists of glyphs don't have any intersecting glyphs, 177643 and if they all have a maskFormat matching the format of the actual 177644 glyphs If so, we can avoid the temporary pixmap for accumulating 177645 glyphs, which reduces the number of operations done, and makes it 177646 easier on the migration system. This fixes some significant performance 177647 issues, particularly with subpixel antialiasing. Note that it does 177648 increase the amount of damage computation which is done, so is not 177649 always a win with a compositing manager running. 177650 177651commit 074dc9a023b3967ce00aa42c26a7c988423afe8a 177652Author: Eric Anholt <anholt@freebsd.org> 177653Date: Tue Apr 25 16:47:23 2006 +0000 177654 177655 Add an option to verify at the point of migration that pixmaps which aren't 177656 marked dirty are in fact not dirty. This will hopefully help catch 177657 issues like the previous commit. Leave it on in fakexa. 177658 177659commit 702d9226d57ec1584de2e8a85c268795650b1094 177660Author: Eric Anholt <anholt@freebsd.org> 177661Date: Tue Apr 25 15:46:04 2006 +0000 177662 177663 Don't forget to mark the drawable as dirty in exaPutImage(). Fixes 177664 corruption with drivers that have UTS. (Michel Dänzer) 177665 177666commit 39ca0867c7dd6f3bdecc52aec8df435946682098 177667Author: Donnie Berkholz <spyderous@gentoo.org> 177668Date: Mon Apr 24 05:44:06 2006 +0000 177669 177670 Remove another reference to cfb16. 177671 177672commit 81f3c2937df6230542f3223c201da8c41ff59945 177673Author: Donnie Berkholz <spyderous@gentoo.org> 177674Date: Mon Apr 24 05:39:47 2006 +0000 177675 177676 strlcpy() doesn't exist on Linux, so use the implementation in os/. 177677 177678commit a715634d23fb3124261dbbd8d7d4e6522551bb9b 177679Author: Matthieu Herrb <matthieu.herrb@laas.fr> 177680Date: Sun Apr 23 13:14:50 2006 +0000 177681 177682 Don't access free memory after unloading a module. Bugzilla #4168. 177683 177684commit 79dc6892610c9f8385cde4f0d601cc7481225c16 177685Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 177686Date: Sat Apr 22 03:22:17 2006 +0000 177687 177688 When reading the kernel keyboard mapping (readKernelMapping in 177689 os-support/linux/lnx_KbdMap.c) we overrun the usefully-named global 177690 array 'map', scribbling on other random static variables elsewhere. 177691 This is fixed by changing the size of at2lnx. (David Woodhouse). Bug 177692 #5169 177693 177694commit b37c515320dc8df2b1d160cc3f37d6bfac109b91 177695Author: Keith Packard <keithp@keithp.com> 177696Date: Wed Apr 19 21:56:13 2006 +0000 177697 177698 Add support for x4a4 format (depth 4 at 8bpp). Bug #6325. 177699 177700commit c947d796aad0b81d661a10b787deed967376da79 177701Author: Eric Anholt <anholt@freebsd.org> 177702Date: Tue Apr 18 19:18:43 2006 +0000 177703 177704 Missed in previous commit: Add a new migration scheme, called "Smart" for 177705 lack of a better name. This one behaves somewhat between Greedy and 177706 Always. It moves in if we can accelerate, unless the destination is 177707 clean and shouldn't be kept in framebuffer according to the score, in 177708 which case we migrate out (and force-migrate anything where migration 177709 is free). This should help fix lack of acceleration for drivers without 177710 UTS since removing exaAsyncPixmapGCOps, and has removed one performance 177711 trap with Radeon I'd noticed. It is the new default. 177712 177713commit b17a4de83e7ab18bef29ae898195889638f1cc6a 177714Author: Eric Anholt <anholt@freebsd.org> 177715Date: Tue Apr 18 19:14:07 2006 +0000 177716 177717 Add a new migration scheme, called "Smart" for lack of a better name. This 177718 one behaves somewhat between Greedy and Always. It moves in if we can 177719 accelerate, unless the destination is clean and shouldn't be kept in 177720 framebuffer according to the score, in which case we migrate out (and 177721 force-migrate anything where migration is free). This should help fix 177722 lack of acceleration for drivers without UTS since removing 177723 exaAsyncPixmapGCOps, and has removed one performance trap with Radeon 177724 I'd noticed. It is the new default. 177725 177726commit 771b366abe0bc060592b548612ec413291e14bf0 177727Author: Eric Anholt <anholt@freebsd.org> 177728Date: Tue Apr 18 18:50:35 2006 +0000 177729 177730 Fix exaGetPixmapFirstPixel to migrate as unaccelerated. Also adds a bit of 177731 fallback debugging info to PolyGlyphBlt. 177732 177733commit 782d61a03176264e0a9eb222ae97be3d175cf0ab 177734Author: Donnie Berkholz <spyderous@gentoo.org> 177735Date: Tue Apr 18 17:28:44 2006 +0000 177736 177737 Update all prefixes in config tools, so they mostly work again. Also, make 177738 RgbPath commented out when using xorgconfig. Start using 177739 /dev/input/mice as the default mouse location on Linux. 177740 177741commit 0f065059dcaf9c452f1cdec115f619f697fd71cc 177742Author: Donnie Berkholz <spyderous@gentoo.org> 177743Date: Mon Apr 17 07:27:43 2006 +0000 177744 177745 Wrap a couple more SDK headers in if XORG, as per Dave Airlie's commit on 177746 2006-01-18. 177747 177748commit ba632f697a782cd47870705b6cecaac2c60d30ff 177749Author: Donnie Berkholz <spyderous@gentoo.org> 177750Date: Mon Apr 17 07:10:31 2006 +0000 177751 177752 Fix kdrive build by linking in libexa before KDRIVE_LIBS. 177753 177754commit 0e62d92d5b809bc3a6503e9bc386cf961fc22557 177755Author: Adam Jackson <ajax@nwnk.net> 177756Date: Fri Apr 14 23:43:32 2006 +0000 177757 177758 Coverity #804: Another leak on OOM path. 177759 177760commit d61219aaadf9e4aa83644a69627d3a1d3282c95f 177761Author: Adam Jackson <ajax@nwnk.net> 177762Date: Fri Apr 14 23:38:11 2006 +0000 177763 177764 Coverity #806: Another memory leak on OOM path. 177765 177766commit 1b04e313920447e4c1f42bdd5a61f188d463210c 177767Author: Adam Jackson <ajax@nwnk.net> 177768Date: Fri Apr 14 23:32:22 2006 +0000 177769 177770 Coverity #847, #848, #849: Three more memory leaks. 177771 177772commit 6545051902f2ce00c98bd1373f97ebc942667e9c 177773Author: Adam Jackson <ajax@nwnk.net> 177774Date: Fri Apr 14 23:10:59 2006 +0000 177775 177776 Coverity #1003, #1004: Two more useless null checks. 177777 177778commit ab1d420022fb09d36a0d6ad948c38147c65b9adf 177779Author: Adam Jackson <ajax@nwnk.net> 177780Date: Fri Apr 14 23:09:38 2006 +0000 177781 177782 Coverity #1005: Avoid a null deref. 177783 177784commit d01e0956a8903fb41e8a34c78973b9b2860b6446 177785Author: Adam Jackson <ajax@nwnk.net> 177786Date: Fri Apr 14 23:08:10 2006 +0000 177787 177788 Coverity #1007: Fix a silly null check. 177789 177790commit 6d29f659318364afe046dc242d6f506ce40a944a 177791Author: Luc Verhaegen <libv@skynet.be> 177792Date: Fri Apr 14 23:01:35 2006 +0000 177793 177794 CVT means Coordinated Video Timing instead of Common. 177795 177796commit 82b6ea1a4b414426072bf001daeb3e9de0e93589 177797Author: Adam Jackson <ajax@nwnk.net> 177798Date: Fri Apr 14 22:51:19 2006 +0000 177799 177800 Bug #6580: Don't install xf86drm.h, that's libdrm's job. 177801 177802commit aefa347bded9a3179ab139d0ccddce314040e9b9 177803Author: Ian Romanick <idr@us.ibm.com> 177804Date: Thu Apr 13 21:08:25 2006 +0000 177805 177806 Fix build for added file to Mesa CVS. This is always fun. :( 177807 177808commit 6aadd454e70d83921685b58bf57ec30d95920734 177809Author: Daniel Stone <daniel@fooishbar.org> 177810Date: Mon Apr 10 10:11:19 2006 +0000 177811 177812 Fix stupid thinko. 177813 177814commit c9f6e60d42dec82d06995c05a2a011c338cadd87 177815Author: Daniel Stone <daniel@fooishbar.org> 177816Date: Mon Apr 10 08:50:33 2006 +0000 177817 177818 Coverity #826: Fix potential memory leak. 177819 177820commit 1357af2474be9a3bce7ee2350fd4252eee89a3b1 177821Author: Daniel Stone <daniel@fooishbar.org> 177822Date: Sun Apr 9 17:39:10 2006 +0000 177823 177824 Coverity #340: Fix potential NULL dereference. Clean up proliferation of 177825 'register int n' in loops of ProcXkbGetNames. 177826 177827commit f324be00c547effc698ae6679d12ffe90bd90e43 177828Author: Daniel Stone <daniel@fooishbar.org> 177829Date: Sun Apr 9 17:28:42 2006 +0000 177830 177831 Coverity #324: Fix potential NULL dereference. (Alan Coopersmith) 177832 177833commit 7637aa17f21e26d979fbb210a638d6751c98b1eb 177834Author: Daniel Stone <daniel@fooishbar.org> 177835Date: Sun Apr 9 17:26:17 2006 +0000 177836 177837 Coverity #169: Fix potential fgets() into NULL (?!?). 177838 177839commit d5bc41b88272b4a3a1841cc1189720b0549db215 177840Author: Daniel Stone <daniel@fooishbar.org> 177841Date: Sun Apr 9 17:15:51 2006 +0000 177842 177843 Coverity #323, #445, #446, #447: Fix potential NULL dereferences. 177844 177845commit 2387bfa5ff5ed82f3f732fb9152c1ea95850a914 177846Author: Aaron Plattner <aplattner@nvidia.com> 177847Date: Fri Apr 7 18:56:04 2006 +0000 177848 177849 Bump the ABI versions. Due to Glyph privates and the XV update below, the 177850 video driver ABI needs to be bumped to 1.0. The rest of the ABI minor 177851 versions were bumped to include the LoaderGetABIVersion function. 177852 Add a DrawblePtr argument to the XV hooks. This allows drivers to determine 177853 that the target window is redirected and draw to the appropriate place. 177854 177855commit dc43909219fe2a4d03139638814b89032b2921b9 177856Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 177857Date: Fri Apr 7 17:49:32 2006 +0000 177858 177859 Fri Apr 7 13:46:45 2006 Søren Sandmann <sandmann@redhat.com> 177860 Use FreeResource instead of deleteCompOverlayClient() 177861 177862commit 94e7213d594dbbb53a6bb05d1dab7514c4ff5350 177863Author: Adam Jackson <ajax@nwnk.net> 177864Date: Fri Apr 7 16:08:50 2006 +0000 177865 177866 Remove libc wrapper types from Xisb interfaces. 177867 177868commit 47bdc9528c2dd4ea9d59a0944c023173ea7a7a66 177869Author: Daniel Stone <daniel@fooishbar.org> 177870Date: Fri Apr 7 16:07:50 2006 +0000 177871 177872 Coverity #844, #845, #846: Fix memory leaks. 177873 177874commit 2c90c3bfef8563f739a72bb645dd52b35b6ff6d5 177875Author: Daniel Stone <daniel@fooishbar.org> 177876Date: Fri Apr 7 15:57:17 2006 +0000 177877 177878 Coverity #987: Avoid potential NULL dereference. 177879 177880commit 843146cfbaef234e13df9a62b6f0232a5efdf7f0 177881Author: Daniel Stone <daniel@fooishbar.org> 177882Date: Fri Apr 7 15:53:21 2006 +0000 177883 177884 Coverity #1216: Fix double-close of file on error. 177885 177886commit 5dacc822327689c0f096093756473c96fba67d76 177887Author: Keith Packard <keithp@keithp.com> 177888Date: Fri Apr 7 02:20:11 2006 +0000 177889 177890 Coverity #333, #334 - eliminate unncessary test for always true condition 177891 in fbEvenStipple. 177892 177893commit 75a9afdbf42e4196471774102e1758f18866bec6 177894Author: Adam Jackson <ajax@nwnk.net> 177895Date: Fri Apr 7 01:53:43 2006 +0000 177896 177897 Coverity #488: Avoid smashing an array on malformed config files. 177898 177899commit 20c1ef2cc30abe45eeaf5b0833cbc0095ed05c02 177900Author: Adam Jackson <ajax@nwnk.net> 177901Date: Fri Apr 7 01:50:07 2006 +0000 177902 177903 Coverity #769: Fix a potential memory leak for systems that allocate on 177904 malloc(0) 177905 177906commit 5ef711032b821be82fd7281fe64872bcbaff0327 177907Author: Adam Jackson <ajax@nwnk.net> 177908Date: Fri Apr 7 01:41:00 2006 +0000 177909 177910 Coverity #838: Plug two more memory leaks. 177911 177912commit 69477ea4b6e666940c5dd4422bedfa6432dead04 177913Author: Adam Jackson <ajax@nwnk.net> 177914Date: Fri Apr 7 01:37:11 2006 +0000 177915 177916 Coverity #837: Fix another another memory leak. 177917 177918commit b472ce7307dd88a21c7713a2b127e34f5c2bc817 177919Author: Adam Jackson <ajax@nwnk.net> 177920Date: Fri Apr 7 01:35:43 2006 +0000 177921 177922 Coverity #836: Fix another memory leak. 177923 177924commit 9c84ed5f8d9eded1a8b509c9cad1ca0ebcf2166a 177925Author: Adam Jackson <ajax@nwnk.net> 177926Date: Fri Apr 7 01:34:29 2006 +0000 177927 177928 Coverity #835: Plug memory leak in extension section parsing. 177929 177930commit 12924d0da36ad2266bb040caac58534c07e85261 177931Author: Adam Jackson <ajax@nwnk.net> 177932Date: Fri Apr 7 01:29:39 2006 +0000 177933 177934 Coverity #812: Fix parser memory leak. 177935 177936commit 49abff79957799e9229d5c0226ee1b0d7505003d 177937Author: Adam Jackson <ajax@nwnk.net> 177938Date: Fri Apr 7 01:26:33 2006 +0000 177939 177940 Coverity #818: Avoid memory leak on error path. 177941 177942commit bda292120fc97f890c1f58a31177c0f7c0bfa048 177943Author: Adam Jackson <ajax@nwnk.net> 177944Date: Fri Apr 7 01:23:50 2006 +0000 177945 177946 Coverity #985: Avoid segfault on malloc failure. 177947 177948commit 536628bb4bcb0a0d749e0c01412a5eb5d6d24063 177949Author: Adam Jackson <ajax@nwnk.net> 177950Date: Fri Apr 7 01:18:01 2006 +0000 177951 177952 Coverity #1037: Sanity check idx before use. 177953 177954commit 53e97ce4ddd993248561c245143b61915ea254b5 177955Author: Adam Jackson <ajax@nwnk.net> 177956Date: Thu Apr 6 22:04:12 2006 +0000 177957 177958 missed a line while removing cfb16 177959 177960commit 4ae12636694af05cee4287b119bde08e9ceaa8aa 177961Author: Adam Jackson <ajax@nwnk.net> 177962Date: Thu Apr 6 18:59:11 2006 +0000 177963 177964 Remove cfb16, no longer used. 177965 177966commit e1fc15a85fb367ee9afd63c920c3327c3f45158d 177967Author: Fredrik Höglund <fredrik@kde.org> 177968Date: Wed Apr 5 21:08:45 2006 +0000 177969 177970 Put the screensaver extension back in the Xext module. 177971 Move the screenSaverSuspended variable to DIX globals. 177972 Restore the old link order for the Xorg and Xdmx binaries. 177973 177974commit 383c2e1e9ec54ab9de356993ad552c1aa6ec094f 177975Author: Ian Romanick <idr@us.ibm.com> 177976Date: Wed Apr 5 19:52:12 2006 +0000 177977 177978 Include fbmmx.h in fb/fbwindow.c when USE_MMX is defined. Fixes build 177979 problem on x86-64 resulting from fbHaveMMX being a macro instead of a 177980 function on that platform. 177981 177982commit 4697da177d545a2f8bb6fd0d6588a1c40532c339 177983Author: Adam Jackson <ajax@nwnk.net> 177984Date: Tue Apr 4 18:30:28 2006 +0000 177985 177986 Initial checkin 177987 177988commit 83ea57bcc82f478a7ecdcd6ed73ca4be01cd9c26 177989Author: Adam Jackson <ajax@nwnk.net> 177990Date: Tue Apr 4 14:39:06 2006 +0000 177991 177992 Bug #5729: Convert xf8_16bpp to fb. chips(4) users please test. 177993 177994commit 4c7da861185080d15b3ff4301af4af0e85a71f93 177995Author: Adam Jackson <ajax@nwnk.net> 177996Date: Tue Apr 4 14:17:04 2006 +0000 177997 177998 Bug #5300: Fix missing spaces in the Build OS line in the log. (Egmont 177999 Koblinger) 178000 178001commit fb6f61b50f1c701041680e49f6a406a6603f1577 178002Author: Adam Jackson <ajax@nwnk.net> 178003Date: Tue Apr 4 12:36:16 2006 +0000 178004 178005 Bug #4806: Dump the raw EDID contents in hex to the log file for better 178006 debugging. (Philip Prindeville) 178007 178008commit 14af50371c7f23855781924cdf6afa6ab7566a87 178009Author: Adam Jackson <ajax@nwnk.net> 178010Date: Mon Apr 3 22:00:06 2006 +0000 178011 178012 Bug #2142: Make font path logging more readable. (Eduard Fuchs) 178013 178014commit 373f9f92566290d979730c09c9c5c5d50e23390c 178015Author: Adam Jackson <ajax@nwnk.net> 178016Date: Mon Apr 3 21:45:54 2006 +0000 178017 178018 Bug #4766: Convert all Xprint drivers to fb. 178019 178020commit d9b8bfbfafe8758ceb629606607e37546d51ca52 178021Author: Adam Jackson <ajax@nwnk.net> 178022Date: Mon Apr 3 21:16:30 2006 +0000 178023 178024 Bug #5478: More use of fbSOlidFillmmx. (Jim Huang) 178025 178026commit b0e67782653033c6518944adfbf23e466bd8bc39 178027Author: Adam Jackson <ajax@nwnk.net> 178028Date: Mon Apr 3 19:50:15 2006 +0000 178029 178030 Bug #6346: Build fix when using gcc -mno-sse. (Jonathan Adamczewski) 178031 178032commit 66500819b1ca730a7b1df400a8368a08cbe49335 178033Author: Daniel Stone <daniel@fooishbar.org> 178034Date: Mon Apr 3 11:37:30 2006 +0000 178035 178036 Bug #1358: Make ISO_Prev_Group cycle/wrap as ISO_Next_Group does. 178037 178038commit 2a6c11aa3b06f13dad94f3441c7184e6720a2bf4 178039Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 178040Date: Mon Apr 3 09:12:28 2006 +0000 178041 178042 Fix a server crash due to memsetting beyond allocated memory when running 178043 GL applications. 178044 178045commit f6ca2b3ea92b7fe98408c51a17a590435e808b1d 178046Author: Adam Jackson <ajax@nwnk.net> 178047Date: Mon Apr 3 02:15:55 2006 +0000 178048 178049 Coverity #38: Dead branch elimination. 178050 178051commit 9b9dd747d8f4697c6d5c947c160d5991c7c8fde5 178052Author: Adam Jackson <ajax@nwnk.net> 178053Date: Mon Apr 3 02:13:47 2006 +0000 178054 178055 Coverity #75: Dead variable elimination. 178056 178057commit 3f87aeefb4be3ac23ae636d3756ffdc446eaa62d 178058Author: Adam Jackson <ajax@nwnk.net> 178059Date: Mon Apr 3 02:12:11 2006 +0000 178060 178061 Coverity #82: Dead variable elimination. 178062 178063commit 61926dbe592468076f8c9a666f0098d067d2213e 178064Author: Adam Jackson <ajax@nwnk.net> 178065Date: Mon Apr 3 02:09:05 2006 +0000 178066 178067 Coverity #271: Fix an unbelievably boneheaded NULL chase. 178068 178069commit 7ef95da8a3e22e710882590fc47d56893159cb5d 178070Author: Adam Jackson <ajax@nwnk.net> 178071Date: Mon Apr 3 01:51:54 2006 +0000 178072 178073 Coverity #616: Fix a rare memory leak. 178074 178075commit 01ebd633017249c496f378df511586c973d49708 178076Author: Adam Jackson <ajax@nwnk.net> 178077Date: Mon Apr 3 01:43:33 2006 +0000 178078 178079 Coverity #833: Fix a rather nasty memory leak. 178080 178081commit a01f17d6dec02f80144e108f748783cb4e429ebb 178082Author: Adam Jackson <ajax@nwnk.net> 178083Date: Mon Apr 3 01:35:05 2006 +0000 178084 178085 Coverity #983: Move some risky debugging code inside #ifdef DEBUG. 178086 178087commit c03cfca3806f45948627715b25b46839a07be979 178088Author: Adam Jackson <ajax@nwnk.net> 178089Date: Mon Apr 3 01:31:59 2006 +0000 178090 178091 Coverity #986: Prevent a NULL chase. 178092 178093commit 07ecf49521973bbb205b199c39e1171f1163df2b 178094Author: Adam Jackson <ajax@nwnk.net> 178095Date: Mon Apr 3 01:28:11 2006 +0000 178096 178097 Coverity #992: Prevent a NULL chase. 178098 178099commit c6b3b3354c2d9139b19b132051d434e97dd19715 178100Author: Adam Jackson <ajax@nwnk.net> 178101Date: Sun Apr 2 22:51:42 2006 +0000 178102 178103 Bump to 1.1.99.1. 178104 178105commit 7e085f52b6f07c076bd3bcfdce27c17d14d7822e 178106Author: Kristian Høgsberg <krh@redhat.com> 178107Date: Sun Apr 2 22:31:13 2006 +0000 178108 178109 Use xf86LoaderCheckSymbol to check for DRI symbols instead of dlsym, 178110 avoiding RTLD_DEFAULT. (__glXDRIscreenProbe): Change GLX-DRI to AIGLX 178111 in LogMessage for consitency. 178112 178113commit b2097b99a2e6cc045ee9b6d80946bc06c4d9302c 178114Author: Adam Jackson <ajax@nwnk.net> 178115Date: Sun Apr 2 21:45:03 2006 +0000 178116 178117 ../stub 178118 178119commit 4e3a4cfdd1d7153eb88aab05ed02ddb32601ae93 178120Author: Eric Anholt <anholt@freebsd.org> 178121Date: Sun Apr 2 06:22:05 2006 +0000 178122 178123 Use RTLD_DEFAULT, rather than relying on NULL happening to map to it as it 178124 does on Linux. 178125 178126commit 323fec20292fc5ad90bfee9015ecccdc13c968ad 178127Author: Adam Jackson <ajax@nwnk.net> 178128Date: Sun Apr 2 00:46:20 2006 +0000 178129 178130 Reorder link order for Xdmx to fix new screensaver variable reference 178131 properly; remove previous awful hack. 178132 178133commit a605b9ffd3c2e7d227e35b911761f720bf07b7e6 178134Author: Adam Jackson <ajax@nwnk.net> 178135Date: Sun Apr 2 00:09:43 2006 +0000 178136 178137 Fix some includes to point into X11/fonts/ properly. 178138 178139commit e5b1d38e142807b59ce4ec89764c949f707ec541 178140Author: Adam Jackson <ajax@nwnk.net> 178141Date: Sat Apr 1 23:53:33 2006 +0000 178142 178143 Disable Xprint freetype support momentarily. Needs ttf2pt1.c, which exists 178144 in the monolith but has an advertising clause in the license. 178145 178146commit ccca76b8083b83825fa16483b44e8926a35412bb 178147Author: Eric Anholt <anholt@freebsd.org> 178148Date: Sat Apr 1 23:41:23 2006 +0000 178149 178150 Clean up warnings and a debug printf. 178151 178152commit 6afa814ab16f351b2eb787e5bf481a1f9738b391 178153Author: Eric Anholt <anholt@freebsd.org> 178154Date: Sat Apr 1 23:28:17 2006 +0000 178155 178156 Pull out fb's tile handling during fbValidateGC so we can do the necessary 178157 exaPrepare/FinishAccess()es. Revealed by xtest with fakexa. 178158 178159commit 277f612d4eeb89adb8ccda4e8fd3d211d8d1705e 178160Author: Adam Jackson <ajax@nwnk.net> 178161Date: Sat Apr 1 23:19:08 2006 +0000 178162 178163 Hack around the new screensaver variable for DMX, which is otherwise 178164 blissfully ignorant of it. 178165 178166commit 5f95146fcfcae60cc29265799ba3b851647105d6 178167Author: Eric Anholt <anholt@freebsd.org> 178168Date: Sat Apr 1 22:35:16 2006 +0000 178169 178170 Export exaPrepare/FinishGC to the rest of EXA, and use it in the ImageGlyph 178171 implementation to avoid unprepared access to the tile. Also, relocate 178172 the fbGetDrawable to avoid using a stale dest pointer after 178173 exaSolidBoxClipped() may have migrated it. Revealed by xtest. 178174 178175commit c720ffe875e4b2038746ff9b4767f8b90db0a307 178176Author: Eric Anholt <anholt@freebsd.org> 178177Date: Sat Apr 1 22:17:44 2006 +0000 178178 178179 Use fb's depth-to-planemask computation, which doesn't suffer from getting 178180 a 1 planemask at depth 32. Fixes Get/PutImage xtest tests. 178181 178182commit 5c0a2088e229d05c38e5df7daea45af0d7db7daf 178183Author: Daniel Stone <daniel@fooishbar.org> 178184Date: Sat Apr 1 21:49:44 2006 +0000 178185 178186 Bug #6428: Fix off-by-one error when walking off the end of the vmodmap 178187 list. 178188 178189commit 1e764feab595b781dab22d6e41c26f118c9d41b5 178190Author: Daniel Stone <daniel@fooishbar.org> 178191Date: Sat Apr 1 21:20:31 2006 +0000 178192 178193 Bug #5801: Check for MTRR support under Linux. Minor refactoring of MTRR 178194 checks for other OSes. 178195 178196commit 978c7b14a18caffde5600480824d04492fc32aef 178197Author: Daniel Stone <daniel@fooishbar.org> 178198Date: Sat Apr 1 21:02:40 2006 +0000 178199 178200 Make Xprint AC_ARG_ENABLEs and AC_ARG_WITHs unconditional also. 178201 178202commit 71a6f2ef6c1138c5c6918a54dfb856183f4f242c 178203Author: Daniel Stone <daniel@fooishbar.org> 178204Date: Sat Apr 1 20:58:42 2006 +0000 178205 178206 Unconditionally run XP_USE_FREETYPE AM_CONDITIONAL, not only in the Xprint 178207 path. 178208 178209commit d1e90113fc32b6ddc4dbe1a074763c31bc133e75 178210Author: Eric Anholt <anholt@freebsd.org> 178211Date: Fri Mar 31 23:22:29 2006 +0000 178212 178213 Don't attempt to Prepare/FinishAccess NULL pDrawables. Exposed by new 178214 gradient testing in rendercheck. 178215 178216commit 2e38fedd29e7e55d01e3edce6a73b8ceaac17911 178217Author: Eric Anholt <anholt@freebsd.org> 178218Date: Fri Mar 31 19:41:28 2006 +0000 178219 178220 Add an option to EXA for the DDX to request that EXA hide the pixmap's 178221 devPrivate.ptr when pointing at offscreen memory, outside of 178222 exaPrepare/FinishAccess(). This was used with fakexa to find (by NULL 178223 dereference) many instances of un-Prepared CPU access to the 178224 framebuffer: 178225 - GC tiles used in several ops when fillStyle == FillTiled were never 178226 Prepared. 178227 - Migration could lead to un-Prepared access to mask data in render's 178228 Trapezoids and Triangles 178229 - PutImage's UploadToScreen failure fallback failed to Prepare. 178230 178231commit f480dc797b51f080f912efc7867d6d8e50be074c 178232Author: Eric Anholt <anholt@freebsd.org> 178233Date: Fri Mar 31 19:25:42 2006 +0000 178234 178235 Revert mistaken commit to exa_unaccel.c. Should have been to 178236 exa_offscreen.c: Correct a typo in debug-only offscreen validation 178237 code. (Wang Zhenyu) 178238 178239commit 1a8167c1baa767fc056d1e17d96d0ea98a5f3b17 178240Author: Eric Anholt <anholt@freebsd.org> 178241Date: Fri Mar 31 19:16:51 2006 +0000 178242 178243 Correct a typo in debug-only offscreen validation code. (Wang Zhenyu) 178244 178245commit 7ea30b507f4ce5ce20fbfaca80f7d5b53a99eb1d 178246Author: Fredrik Höglund <fredrik@kde.org> 178247Date: Fri Mar 31 18:49:38 2006 +0000 178248 178249 Move the screensaver extension from module to builtins. 178250 Add the server side implementation of the ScreenSaverSuspend request. 178251 Require scrnsaverproto >= 1.1, and change the linking order of the Xorg 178252 static libs. 178253 178254commit acca49b1a5a6c034f3b9d51d9016b8a7d43da809 178255Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 178256Date: Fri Mar 31 17:39:35 2006 +0000 178257 178258 Fri Mar 31 12:37:16 2006 Søren Sandmann <sandmann@redhat.com> 178259 Fix copyright statement 178260 178261commit b074ce22470ba0a51eda2af7100d09a260a1e8bb 178262Author: Egbert Eich <eich@suse.de> 178263Date: Fri Mar 31 15:11:51 2006 +0000 178264 178265 fixed typo. 178266 178267commit 710bb2e6c8b2874406e48fa8ad24539290c98d41 178268Author: Daniel Stone <daniel@fooishbar.org> 178269Date: Fri Mar 31 14:52:57 2006 +0000 178270 178271 Reindent with -cbi0. 178272 178273commit 7c44bb8c49656133eae675377edea55322d254ca 178274Author: Daniel Stone <daniel@fooishbar.org> 178275Date: Fri Mar 31 07:33:34 2006 +0000 178276 178277 Simplify XkbWriteXKBKeymapForNames a bit, and remove debug spew. 178278 178279commit 4c317bbc1259fa555dc5d5278226b21c42845c0c 178280Author: Daniel Stone <daniel@fooishbar.org> 178281Date: Fri Mar 31 07:21:41 2006 +0000 178282 178283 Add full FreeType support for Xprint. (Drew Parsons) 178284 178285commit 759033703ce17b20d57756206f48a7ae410a50d1 178286Author: Eric Anholt <anholt@freebsd.org> 178287Date: Thu Mar 30 21:44:36 2006 +0000 178288 178289 Remove the exaAsyncPixmapGCOps mostly-unaccelerated ops vector, and always 178290 plug in the accelerated one, even if the destination pixmap is 178291 currently offscreen. This was a leftover from when kaa originally got 178292 accelerated offscreen pixmap support, and its only concievable use was 178293 to avoid a little overhead on ops to in-system pixmaps that weren't 178294 going to get migrated. At this point, we probably care more about just 178295 getting everything accelerated that we easily can, which should happen 178296 with the new migration support. 178297 178298commit b9203dc068ccd4c0d22d49a94b910783432b96a8 178299Author: Eric Anholt <anholt@freebsd.org> 178300Date: Thu Mar 30 21:25:43 2006 +0000 178301 178302 Don't do an extra fallback path for CopyWindow while swappedOut, since 178303 exaCopyNtoN takes care of the fallback anyway, and we don't care about 178304 the performance of this path. 178305 178306commit 5c04610f8aeceed9ec7cd0ca8c5eb314cacc3c25 178307Author: Eric Anholt <anholt@freebsd.org> 178308Date: Thu Mar 30 21:21:59 2006 +0000 178309 178310 Add a dependency on EXA, so it rebuilds when the library does. The manual 178311 indicated I shouldn't do this, but experience indicates I should. 178312 178313commit 8ec42a10ff04e51e8d0b4cffb15064d901bc398d 178314Author: Kristian Høgsberg <krh@redhat.com> 178315Date: Thu Mar 30 20:08:44 2006 +0000 178316 178317 Mark the ARGB FBConfig as nonconforming to prevent drivers and apps from 178318 falling over. 178319 Add @GLX_DEFINES@ so GLcore gets compiled with TLS support if configured. 178320 Only destroy the mesa buffer if it got initialized. 178321 178322commit 08e319091fae7a60ae9fa757659cfde2966af9e9 178323Author: Egbert Eich <eich@suse.de> 178324Date: Thu Mar 30 18:53:41 2006 +0000 178325 178326 Added notice to last ChangeLog entry 178327 Fixes for some vsw4 failures on 64bit BE platforms such as PPC64 and s390x. 178328 Provided by Hong Bo Peng of IBM (slightly modified). Patches try to 178329 resolve some of the careless mixtures of ulong and uint (which are 178330 different size on 178331 64bit). > This patch will break the driver ABI! < Bugzilla #6438. 178332 178333commit 9da1d2257d02155cc8b4541cf5fcb4e64d756945 178334Author: Egbert Eich <eich@suse.de> 178335Date: Thu Mar 30 18:48:11 2006 +0000 178336 178337 Fixes for some vsw4 failures on 64bit BE platforms such as PPC64 and s390x. 178338 Provided by Hong Bo Peng of IBM (slightly modified). Patches try to 178339 resolve some of the careless mixtures of ulong and uint (which are 178340 different size on 178341 64bit). Bugzilla #6438. 178342 178343commit 6d7ad353bafe914f0b50887daaeaae89ada6ebd3 178344Author: Kristian Høgsberg <krh@redhat.com> 178345Date: Thu Mar 30 18:29:53 2006 +0000 178346 178347 Regenerate these files using updated scripts to avoid unused variable 178348 warnings. 178349 178350commit 2153fa97482bae5737def3ecd4fe1cdc03834991 178351Author: Eric Anholt <anholt@freebsd.org> 178352Date: Thu Mar 30 05:24:27 2006 +0000 178353 178354 Bug #2986: Add PutImage acceleration for the ZPixmap, planeMask ~= 178355 FB_ALLONES, bitsPerPixel >= 8, GXcopy cases. With the radeon driver on 178356 my machine, this gives about 10% speedup in PutImage 178357 10x10 and 500x500, and 40% speedup for 10x10 ShmPutImage, up to 65% 178358 improvement in 500x500 ShmPutImage. Also fixes a crasher in GetImage 178359 that slipped in at the last minute. 178360 178361commit 3cf46cc1e32efc0e4be1d88be111ba0438e0f021 178362Author: Eric Anholt <anholt@freebsd.org> 178363Date: Thu Mar 30 05:15:58 2006 +0000 178364 178365 Add an UploadToScreen implementation, for testing PutImage support, and 178366 make the DownloadFromScreen more robust. 178367 178368commit e799dd68e2bd0fa8ac3c344111fb12e1f32d4c10 178369Author: Eric Anholt <anholt@freebsd.org> 178370Date: Wed Mar 29 22:25:17 2006 +0000 178371 178372 Bug #2986: Add acceleration of GetImage using DownloadFromScreen for the 178373 ZPixmap, planeMask ~= FB_ALLONES, bitsPerPixel >= 8 case. I'm pretty 178374 convinced that this is the only case that we care about at all. Tested 178375 with xwd -root and xwd on a gnome-terminal, in a composited environment 178376 or not. 178377 178378commit 4bb5ab0b4453208573b91b334940f190a8f7210a 178379Author: Eric Anholt <anholt@freebsd.org> 178380Date: Wed Mar 29 22:03:18 2006 +0000 178381 178382 Add a DownloadFromScreen implementation, used for testing GetImage 178383 acceleration, and set the migration scheme to Always on init (since 178384 this is all for testing, and Always should make migration happen more 178385 frequently than Greedy). 178386 178387commit e31e8ace1043eab340d6b60a6e98b23ebf102786 178388Author: Deron Johnson <deron.johnson@sun.com> 178389Date: Wed Mar 29 17:51:54 2006 +0000 178390 178391 Fix composite overlay window bug 6411 178392 178393commit ff6f88348c7498e83b0b143ef3737fd6eb0995e4 178394Author: Adam Jackson <ajax@nwnk.net> 178395Date: Wed Mar 29 01:05:09 2006 +0000 178396 178397 More warning cleanup. 178398 178399commit 52d9ce7f4fc599d30dec2e61fc1720597043d91c 178400Author: Kristian Høgsberg <krh@redhat.com> 178401Date: Tue Mar 28 21:45:14 2006 +0000 178402 178403 Fix another typo. 178404 178405commit 7df64898eac46a487e8eab2af7213d133b9ca419 178406Author: Kristian Høgsberg <krh@redhat.com> 178407Date: Tue Mar 28 07:46:04 2006 +0000 178408 178409 Fix a couple of typos. 178410 178411commit bd283c2464e2c0e1fd0aca1dedff0f39c2564c34 178412Author: Aaron Plattner <aplattner@nvidia.com> 178413Date: Tue Mar 28 07:21:50 2006 +0000 178414 178415 Add a new export, LoaderGetABIVersion. This function allows modules to 178416 query the versions directly instead of having to guess. Bug #6416: Add 178417 LoaderGetABIVersion. 178418 178419commit a06342eccc76035ff859fee4d283b288c90ee923 178420Author: Kristian Høgsberg <krh@redhat.com> 178421Date: Tue Mar 28 02:57:07 2006 +0000 178422 178423 Add --enable-glx-tls ./configure option to enable use of TLS for storing 178424 current GL context. Use this option to let AIGLX load DRI drivers 178425 compiled for TLS. 178426 178427commit 77531dfb9f9f3ca0e38ad0555ee3735d6f28cf19 178428Author: Adam Jackson <ajax@nwnk.net> 178429Date: Tue Mar 28 01:22:01 2006 +0000 178430 178431 Silence some editorializing in the configure help text. 178432 178433commit 7deaaa797cf8e7ca71e9b34fa6f413d1ed2b3dab 178434Author: Adam Jackson <ajax@nwnk.net> 178435Date: Tue Mar 28 01:21:00 2006 +0000 178436 178437 Big old pile of warning fixes. 178438 178439commit 7342dbe4b2108827eaf30993ceeecbd828da2290 178440Author: Adam Jackson <ajax@nwnk.net> 178441Date: Tue Mar 28 00:18:31 2006 +0000 178442 178443 Remove long-dead screen region code. 178444 178445commit 0e88cefbfecbff0c7dd606ce0caca840f45cbc0d 178446Author: Daniel Stone <daniel@fooishbar.org> 178447Date: Mon Mar 27 23:03:47 2006 +0000 178448 178449 Prune XKB code to only what we need to run the server. Remove dead 178450 !XKB_IN_SERVER codepaths. Remove HAVE_CONFIG_H codepaths. 178451 178452commit 5be8a66d324f3d5840b134ad29069eace64e6f12 178453Author: Daniel Stone <daniel@fooishbar.org> 178454Date: Mon Mar 27 22:28:32 2006 +0000 178455 178456 Fix remnants of previous busted _XkbStrCaseCmp commit. 178457 178458commit 9e202dfe40e2bdd66f461a6ba531e927f82096ae 178459Author: Daniel Stone <daniel@fooishbar.org> 178460Date: Mon Mar 27 22:25:56 2006 +0000 178461 178462 Remove remnants of XkbCF code. 178463 178464commit 7257590651328f89d23e80da1ec6241542a660cd 178465Author: Daniel Stone <daniel@fooishbar.org> 178466Date: Mon Mar 27 21:15:06 2006 +0000 178467 178468 Move XFree86 DDX XKB actions into dixmods. 178469 178470commit d7b9e2b0e9d6889ea6b05e63892e612f4e5f19f5 178471Author: Daniel Stone <daniel@fooishbar.org> 178472Date: Sat Mar 25 23:09:50 2006 +0000 178473 178474 Bug #3819: Remove open-coding of strcasecmp. 178475 178476commit b3570dd94aa72f94e537a17680150e91e7712f5a 178477Author: Daniel Stone <daniel@fooishbar.org> 178478Date: Sat Mar 25 22:37:58 2006 +0000 178479 178480 Remove INITARGS braindamage, change to void; add XkbExtensionInit prototype 178481 to xkb.h. Explicitly initialise nTypes in xkb.c. 178482 178483commit 1ef60ce8ebb681b3cfb5e515be5c187c0442dcda 178484Author: Daniel Stone <daniel@fooishbar.org> 178485Date: Sat Mar 25 22:35:48 2006 +0000 178486 178487 Really remove all DDX pre-config code. 178488 178489commit ec10f70b2114e5369a5b2f34b084dcf55634dcb4 178490Author: Daniel Stone <daniel@fooishbar.org> 178491Date: Sat Mar 25 21:52:49 2006 +0000 178492 178493 Remove XkbCF DDX configuration code. 178494 178495commit aae4238360b842ac34dc8ee16e165a1821f9a801 178496Author: Daniel Stone <daniel@fooishbar.org> 178497Date: Sat Mar 25 20:17:58 2006 +0000 178498 178499 Fix two glaring unconditional-NULL-dereferences. 178500 178501commit a68c11bb1d7c5419004a1714e49dffac57304e78 178502Author: Adam Jackson <ajax@nwnk.net> 178503Date: Sat Mar 25 19:52:05 2006 +0000 178504 178505 Mark everything in xf86sym.c as _X_EXPORT. 178506 178507commit ae935832facfa81a9689882406ecca74b0346790 178508Author: Fredrik Höglund <fredrik@kde.org> 178509Date: Fri Mar 24 20:50:13 2006 +0000 178510 178511 Refactored the screensaver and DPMS timer code to use the screensaver timer 178512 for both screensaver and DPMS. Removed the SetDPMSTimers() and 178513 FreeDPMSTimers() functions. 178514 178515commit d1746ec0f0c8a0b750f390e7a7faf21b67683f4a 178516Author: Kristian Høgsberg <krh@redhat.com> 178517Date: Fri Mar 24 17:58:39 2006 +0000 178518 178519 Make sure DRI module is loaded before calling DRI functions. 178520 178521commit f1616508c95d12dfaad2cfd61b40228b3dba6f60 178522Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 178523Date: Thu Mar 23 23:54:08 2006 +0000 178524 178525 Add ast driver/pci id (Carl Switzky, Sun Microsystems) 178526 178527commit 6d2896b384e17512e8f12036daabcd575d21f804 178528Author: Kristian Høgsberg <krh@redhat.com> 178529Date: Wed Mar 22 22:49:52 2006 +0000 178530 178531 Improve error logging. 178532 178533commit 5449634e3c9428005aba5b3322ced7e86c62f185 178534Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 178535Date: Wed Mar 22 21:37:49 2006 +0000 178536 178537 Wed Mar 22 16:28:46 2006 Søren Sandmann <sandmann@redhat.com> 178538 Use inline assembly for copy area, since gcc doesn't generate movq 178539 instructions. 178540 178541commit 5b3084c64f7bd1232603ffb3e985600b8d045453 178542Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 178543Date: Wed Mar 22 21:13:08 2006 +0000 178544 178545 Wed Mar 22 16:05:09 2006 Søren Sandmann <sandmann@redhat.com> 178546 Use inline assembly for solid fills, since gcc doesn't use the movq 178547 instructions. 178548 178549commit a08e5e0c68baaf85b0fc3ecde74a6bcf80bcd4bf 178550Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 178551Date: Wed Mar 22 18:44:26 2006 +0000 178552 178553 Wed Mar 22 13:42:44 2006 Søren Sandmann <sandmann@redhat.com> 178554 Patch by Keith Packard to make sure redirected windows don't get considered 178555 "FullyObscured". 178556 178557commit 966d93ef6d1f2ed02f3b81b5bf5a1ebbdd48c93d 178558Author: Kristian Høgsberg <krh@redhat.com> 178559Date: Tue Mar 21 22:54:38 2006 +0000 178560 178561 Make the server distcheck and tag 1.0.99.1 snapshot. 178562 Bump CVS version to 1.0.99.1. 178563 Distcheck fixes. 178564 178565commit 8e3ad87d01c102591c7dc25614f6ac10e444a1b1 178566Author: Kristian Høgsberg <krh@redhat.com> 178567Date: Tue Mar 21 22:32:13 2006 +0000 178568 178569 #include indirect_dispatch to get prototypes for FBO functions. 178570 Fix a couple of warnings. 178571 178572commit dcc43d57cbe9d2b65384fe9ba2e4e4fbb43cb0a1 178573Author: Donnie Berkholz <spyderous@gentoo.org> 178574Date: Mon Mar 20 20:10:29 2006 +0000 178575 178576 Finish glx_ansic.h wrapper changes to make Xvfb and Xnest link again. 178577 178578commit 9509c6799e31e96677b6d07bdf24ea91ddd30020 178579Author: Adam Jackson <ajax@nwnk.net> 178580Date: Mon Mar 20 19:32:18 2006 +0000 178581 178582 dead file removal 178583 178584commit 61a020265c5915e3d671d5b2047b81a5d15594c3 178585Author: Adam Jackson <ajax@nwnk.net> 178586Date: Mon Mar 20 18:43:18 2006 +0000 178587 178588 Bug #5549: Fix build for sparc64. (Matthieu Herrb) 178589 178590commit 6eb4e2303aaab8d64e3f6cbc0bbee55689bdcb82 178591Author: Adam Jackson <ajax@nwnk.net> 178592Date: Mon Mar 20 14:01:05 2006 +0000 178593 178594 Bug #6213: Check geteuid's return value, not its address, otherwise 178595 unprivileged users can set the modulepath and run arbitrary code. Patch 178596 from Matthieu Herrb. (CVE-2006-0745, Coverity #4) 178597 178598commit 8c1bb37d0649b269b78c457b8b41ff59a41d89af 178599Author: Daniel Stone <daniel@fooishbar.org> 178600Date: Fri Mar 17 08:55:07 2006 +0000 178601 178602 Typo fix, reindent. 178603 178604commit 2d2d38d17cc2558f8a41166a4a1578bc4c663c37 178605Author: Kristian Høgsberg <krh@redhat.com> 178606Date: Fri Mar 17 01:47:25 2006 +0000 178607 178608 Check for glproto when building GLX and make sure we have at least 1.4.6. 178609 Drop glx_ansic.h wrapper and call xalloc, xrealloc, xfree and str-funcs 178610 directly. 178611 178612commit 2c11cde3367fcd22740b577a4364b1e41cf3e1d2 178613Author: Kristian Høgsberg <krh@redhat.com> 178614Date: Fri Mar 17 00:35:18 2006 +0000 178615 178616 More patches from David Reveman: 178617 Add GL_ARB_texture_non_power_of_two, GL_EXT_framebuffer_object and 178618 GL_NV_texture_env_combine4 extensions. 178619 Add __GLXcontext destructor and flush context cache there and on 178620 loseCurrent. 178621 Chain back to new __GLXcontext destructor. (__glXMesaContextForceCurrent): 178622 Set render table on forceCurrent. (init_screen_visuals): Index pVis 178623 array correctly. (GlxGetMesaProvider): Add this. 178624 Hook up FBO marshalling. 178625 178626commit 14aafc258cd774cf937f9798a888c2d3c97ccacf 178627Author: Eric Anholt <anholt@freebsd.org> 178628Date: Thu Mar 16 18:43:55 2006 +0000 178629 178630 Change EXA so that exaMoveOutPixmap() retains the framebuffer copy of the 178631 pixmap, and damage is tracked so that a later exaMoveInPixmap won't 178632 result in an upload if no upload is necessary. This will likely improve 178633 the performance of the "Always" migration scheme significantly, and is 178634 a step in the path to more exact damage tracking between framebuffer 178635 and system memory. 178636 178637commit d0d336efd58896718f31a400651bacd9b769fb5a 178638Author: Daniel Stone <daniel@fooishbar.org> 178639Date: Thu Mar 16 16:29:17 2006 +0000 178640 178641 Add support for ZX2 PCI-E local bus adaptors. (Alex Williamson, HP) 178642 Use soft timeout register to avoid MCAs when probing for non-existent local 178643 bus adaptors on ZX2. (Alex Williamson, HP) 178644 178645commit 175980580e572745a9a381b4432e3ba0457d3ba3 178646Author: Adam Jackson <ajax@nwnk.net> 178647Date: Wed Mar 15 23:05:53 2006 +0000 178648 178649 Bump to requiring fixesproto >= 4.0 and compositeproto >= 0.3. 178650 178651commit 6fe377af5a82deb6f8b0f3b75414335e7845caac 178652Author: Matthieu Herrb <matthieu.herrb@laas.fr> 178653Date: Wed Mar 15 21:25:38 2006 +0000 178654 178655 - OpenBSD needs -Wl,-export-dynamic to export symbols from main executable 178656 to modules. 178657 - Probe for OpenBSD aperture driver and define HAS_APERTURE_DRV 178658 accordingly. 178659 178660commit 21f7f2fb113ee4f9cd011c3cc2d45d43bbdd35fa 178661Author: Felix Kuehling <fxkuehl@gmx.de> 178662Date: Wed Mar 15 18:43:32 2006 +0000 178663 178664 Enable correct handling of the BTS instruction (opcode 0f ab) The code was 178665 there but #ifdefed out. Insead of BTS, BT was executed. This patch 178666 enables the BTS function and hooks it up the the correct opcode. (ATI 178667 Technologies Inc.) 178668 178669commit b726aa502a871c700bc42b5325abf2c6820ff756 178670Author: Felix Kuehling <fxkuehl@gmx.de> 178671Date: Wed Mar 15 18:37:44 2006 +0000 178672 178673 Update to build against Mesa CVS HEAD. 178674 178675commit c74464d92cd673ff0669375757caab798cc57e95 178676Author: Eric Anholt <anholt@freebsd.org> 178677Date: Wed Mar 15 16:59:45 2006 +0000 178678 178679 Don't let pinned pixmaps get migrated in when using the "Always" migration 178680 scheme. This notably keeps the visible screen from getting migrated in 178681 to a new location in framebuffer. 178682 Reported by: Michel Dänzer. 178683 178684commit b9c43cde1e368903786977b06368d5e36db9ffe8 178685Author: Adam Jackson <ajax@nwnk.net> 178686Date: Wed Mar 15 16:56:10 2006 +0000 178687 178688 Coverity #1042, 1043: Nuke some dead variables. 178689 178690commit 5e106a71b9f8077216d41619402952b0005dd8a4 178691Author: Adam Jackson <ajax@nwnk.net> 178692Date: Wed Mar 15 16:49:04 2006 +0000 178693 178694 Coverity #807: Fix a memory leak in XFixesExpandRegion. 178695 178696commit a3ef63696cac950b2520e7c85564befc0a830fde 178697Author: Adam Jackson <ajax@nwnk.net> 178698Date: Wed Mar 15 16:36:31 2006 +0000 178699 178700 Coverity #490: Fix a range check in xf86vidmode extension. 178701 178702commit 152090ce442e94de1ae920208a92931af6493c8c 178703Author: Adam Jackson <ajax@nwnk.net> 178704Date: Wed Mar 15 16:33:12 2006 +0000 178705 178706 Coverity #487: Check version number correctly. 178707 178708commit 72cc6307257fcbb800267464487bf918ee674328 178709Author: Adam Jackson <ajax@nwnk.net> 178710Date: Wed Mar 15 16:32:05 2006 +0000 178711 178712 Coverity #491: Check version number correctly. 178713 178714commit 460f2ea4a594a53536f34c4ad27795fceec50bcc 178715Author: Adam Jackson <ajax@nwnk.net> 178716Date: Wed Mar 15 16:21:04 2006 +0000 178717 178718 Coverity #794: Fix a highly unlikely memory leak. 178719 178720commit 116d158e85ec43577ff69aeb3271ab1f888500c9 178721Author: Adam Jackson <ajax@nwnk.net> 178722Date: Wed Mar 15 16:16:24 2006 +0000 178723 178724 Coverity #269: Compare the requested ABI class against the ABI class of the 178725 module, not the module class. 178726 178727commit d8221a9b70a11606a0f7e1f69afee6049d7f182f 178728Author: Adam Jackson <ajax@nwnk.net> 178729Date: Wed Mar 15 16:11:34 2006 +0000 178730 178731 Coverity #484: Fix an off-by-one in module refcounting. 178732 178733commit 6bb2dc02a7cffd6ed7dd28e88d584920a4150749 178734Author: Adam Jackson <ajax@nwnk.net> 178735Date: Wed Mar 15 16:01:47 2006 +0000 178736 178737 Coverity #337: Remove useless NULL check. 178738 178739commit 1e5c0842af99027cc6c30a16f967d8b60c9a894d 178740Author: Adam Jackson <ajax@nwnk.net> 178741Date: Wed Mar 15 15:34:57 2006 +0000 178742 178743 Coverity #1053: Nuke a dead variable. 178744 178745commit 7314d16cde4c3f99d9d9f1d539f0c5ff4942e653 178746Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 178747Date: Wed Mar 15 03:18:42 2006 +0000 178748 178749 Fix DRIExtensionInit() to not register callbacks when it hasn't been 178750 initialized for the current server generation. Fixes a problem where it 178751 would use stale private index and blow up in colorful ways if no driver 178752 called DRIScreenInit() on the second generation (which happens due to a 178753 bug in radeon that i'll fix separately). Note: clearing the index in 178754 DRIReset() wouldn't work as DRIReset() is called before the 178755 CloseScreen() chain 178756 178757commit 02d80a0de93f7592e69065b0fbe5820dcdebdb44 178758Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 178759Date: Wed Mar 15 03:12:32 2006 +0000 178760 178761 Make xf86 linear allocator smarter when dealing with alignment constraints 178762 when falling back to X/Y allocations. Fixes various problems of Xv 178763 allocation failures, notably with "nv" driver. 178764 178765commit c1601717d536419693b3ef6e8a3d69b9f2fdc2b3 178766Author: Eric Anholt <anholt@freebsd.org> 178767Date: Wed Mar 15 01:20:08 2006 +0000 178768 178769 Add a new migration scheme, "always", which will move pixmaps to their 178770 desired location always (unless they don't fit in FB, in which case 178771 they all get moved out for software rendering). The default remains as 178772 before, but can be controlled by the MigrationHeuristic xorg.conf 178773 option (which is intentionally not documented, as it may be 178774 short-lived). This is part of the exa-damagetrack work, which appears 178775 stable in testing with fakexa, unlike the work as a whole. 178776 178777commit a90cff266cc81993ed804fb320c1dbfe5e0d4787 178778Author: Eric Anholt <anholt@freebsd.org> 178779Date: Wed Mar 15 00:13:52 2006 +0000 178780 178781 Add more doxygen documentation, including notes on WaitMarker() and 178782 MarkSync() that I noticed were needed while reading the VIA driver. 178783 178784commit 693e42114f1127528448126d78a5209dd1198d8d 178785Author: Eric Anholt <anholt@freebsd.org> 178786Date: Tue Mar 14 21:30:12 2006 +0000 178787 178788 Move migration logic to a new function, exaDoMigration(). This is largely a 178789 manual conversion to allow for different migration schemes to be 178790 implemented reasonably, but does include some minor improvements such 178791 as accounting for pinned pixmaps not being acceleratable, and for our 178792 current GetImage and GetSpans not being accelerated. 178793 178794commit d30905478078036383977ae9d4a3685c2e2c642f 178795Author: Eric Anholt <anholt@freebsd.org> 178796Date: Tue Mar 14 20:38:06 2006 +0000 178797 178798 Pull code for getting the (0,0) pixel from a pixmap out to a separate 178799 function, since it gets repeated (with bad error handling, in one 178800 case). 178801 178802commit 01aa209f2056ef04e3f2735756a0f8b4a67a3d87 178803Author: Kristian Høgsberg <krh@redhat.com> 178804Date: Tue Mar 14 19:32:27 2006 +0000 178805 178806 Bail out early if screen doesn't support DRI. 178807 178808commit 0cc34266d6e84bb491fcf9aa74e34615b2fca4fc 178809Author: Deron Johnson <deron.johnson@sun.com> 178810Date: Mon Mar 13 22:43:42 2006 +0000 178811 178812 Updated ChangeLog for my latest composite and xfixes changes. 178813 178814commit 450018f48b2796345a4eaccbb94c1971ebd30114 178815Author: Deron Johnson <deron.johnson@sun.com> 178816Date: Mon Mar 13 21:59:55 2006 +0000 178817 178818 Part 3 of 3 (Other parts are in proto and lib) Composite Version 0.3: 178819 CompositeGetOverlayWindow, CompositeReleaseOverlayWindow Xfixes Version 178820 4.0: XFixesHideCursor, XFixesShowCursor 178821 178822commit e5956f49b217b0ee9c9f35b6a58f339a8d22b1d7 178823Author: Kristian Høgsberg <krh@redhat.com> 178824Date: Mon Mar 13 01:54:59 2006 +0000 178825 178826 First batch of AIGLX fixes from David Reveman. 178827 Add getter for Mesa provider. 178828 Export this for Xgl. 178829 Move resource tracking out of drawable constructor to allow wrapping. 178830 Use corrent reply size #define. 178831 Add this function. (DoGetDrawableAttributes): Fix array length. 178832 178833commit eb63e50d95da4e1e08fc6fcec46ac63d5e3b7bf4 178834Author: Matthieu Herrb <matthieu.herrb@laas.fr> 178835Date: Sun Mar 12 17:14:03 2006 +0000 178836 178837 Fix build when AIGLX is false. 178838 178839commit 9ed3463450469c3108e0be7e4baabc0a403a78b2 178840Author: Eric Anholt <anholt@freebsd.org> 178841Date: Sun Mar 12 03:04:52 2006 +0000 178842 178843 Improve doxygen formatting, and attempt to clarify the 1:1 ratio of 178844 successful PrepareCopy()s to DoneCopy()s. 178845 178846commit 9a7fba5fd07c8831d0acab8d901605de537ae273 178847Author: Eric Anholt <anholt@freebsd.org> 178848Date: Sun Mar 12 03:02:26 2006 +0000 178849 178850 Make exaCopyNtoNTwoDir() call DoneCopy() at the end of each string of 178851 consecutive Copy() calls (rather than exactly once at the end of the 178852 function). 178853 Reviewed by: jbarnes 178854 178855commit c3342c8000f6d2bfb61e2cf95e028d11b59698fa 178856Author: Kristian Høgsberg <krh@redhat.com> 178857Date: Sun Mar 12 00:11:34 2006 +0000 178858 178859 Merge accel_indirect branch to HEAD. 178860 178861commit b1b731c28630965d9e2defe62d1108270dc8264c 178862Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 178863Date: Sat Mar 11 02:43:51 2006 +0000 178864 178865 Fix buffer size checks to prevent 2-byte buffer overflows. (Coverity #480, 178866 #481, #482, #483) 178867 178868commit fc0772de36315f19f5b57220db69f48a3b1fdc9a 178869Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 178870Date: Sat Mar 11 02:10:14 2006 +0000 178871 178872 Add HAS_MMAP for Xvfb 178873 Fix Xvfb option parsing to exit on bad arguments, not just issue error 178874 messages and continue on. (Coverity #492) 178875 178876commit f2ecbb30187000547a98ca7cbaee433ea4ba8fe3 178877Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 178878Date: Sat Mar 11 01:58:32 2006 +0000 178879 178880 Pass sizeof the correct buffer to XmuSnprintf. (Coverity #489) 178881 178882commit d6955798489813ef77cca13cf5f5c67d49e6dece 178883Author: Eric Anholt <anholt@freebsd.org> 178884Date: Fri Mar 10 21:36:24 2006 +0000 178885 178886 If fakexa is enabled, create a larger buffer in the Ximage, but keep the 178887 same width/height for front-buffer drawing. The fakexa code then uses 178888 this extra space for offscreen pixmaps. Note that this tones down the 178889 absurdity of fakexa's offscreen pixmap alignment requirements (odd 178890 alignment is too weird, so stick with "24", which is still strange but 178891 exists out there). It also fixes a couple of bugs in the fakexa 178892 implementation revealed by using offscreen pixmaps. 178893 178894commit 5b1a7b478f072f56e836f2d4c0fbc1985842e2bb 178895Author: Eric Anholt <anholt@freebsd.org> 178896Date: Fri Mar 10 21:32:34 2006 +0000 178897 178898 Move the exaDrawableDirty in exaPrepareAccess to exaFinishAccess, which is 178899 after the drawing is done. Previously, a failed PrepareAccess could 178900 have migrated and cleared the dirty flag before the damage was ever 178901 done. 178902 178903commit ffdbb547becc71f1cfdd035d0d6c71539f185fb1 178904Author: Eric Anholt <anholt@freebsd.org> 178905Date: Fri Mar 10 08:06:42 2006 +0000 178906 178907 Coverity #1011: Remove a useless NULL check on a pVbe that had been 178908 dereferenced many times before. 178909 178910commit 1bc72dce5f8bc40e369e69b684816fdaaa07da43 178911Author: Eric Anholt <anholt@freebsd.org> 178912Date: Fri Mar 10 08:03:24 2006 +0000 178913 178914 Coverity #857: Fix resource leak in error path by freeing earlier. 178915 178916commit 55f677d600370b19d62ef821025481f2be6f5edb 178917Author: Eric Anholt <anholt@freebsd.org> 178918Date: Fri Mar 10 07:58:27 2006 +0000 178919 178920 Coverity #813, #814, #815, #816: Fix resource leaks in error paths of 178921 config parsing code. 178922 178923commit 2bd41105496b729395fbcf97f09581eb0efb3510 178924Author: Eric Anholt <anholt@freebsd.org> 178925Date: Fri Mar 10 07:45:25 2006 +0000 178926 178927 Document the restriction on PrepareAccess() failure, from discussion with 178928 benh. 178929 178930commit 21dcd0304879f38ea8ea01ba88e7cc7783771adf 178931Author: Jeremy C. Reed <reed@reedmedia.net> 178932Date: Fri Mar 10 01:34:45 2006 +0000 178933 178934 Just like FreeBSD, let DragonFly's default mouse Device be /dev/sysmouse 178935 (since /dev/mouse don't even exist by default). 178936 178937commit 9a99afdfb292f303f914039952fdd772eed9e03a 178938Author: Jeremy C. Reed <reed@reedmedia.net> 178939Date: Fri Mar 10 01:22:26 2006 +0000 178940 178941 Add DragonFly support. (It is like FreeBSD.) 178942 This patch is from DragonFly developer Joerg Sonnenberger and the pkgsrc 178943 collection. 178944 I tested using /dev/sysmouse with moused using my serial /dev/cuaa0. 178945 178946commit 7a0f7f739804bc7d9c5562701abee8d134878977 178947Author: Eric Anholt <anholt@freebsd.org> 178948Date: Thu Mar 9 23:29:44 2006 +0000 178949 178950 Coverity #349: Fall back to software early if pSrc->pDrawable is NULL, or 178951 pMask is non-NULL but pMask->pDrawable is NULL. This prevents NULL 178952 dereferences on gradients and other Pictures which have no pDrawable. 178953 178954commit 8a3ff42abb726d1604af39b4653ede5f760b7e69 178955Author: Eric Anholt <anholt@freebsd.org> 178956Date: Thu Mar 9 23:25:35 2006 +0000 178957 178958 Commit changes missed in last commit (mis-typed path and didn't notice): Do 178959 a first pass of doxygen documentation of EXA. This removes the 178960 corresponding pieces of exa-driver.txt, which were becoming stale. 178961 Hopefully the documentation will stay much more up-to-date this way. 178962 Many thanks to jbarnes for writing exa-driver.txt which was used a lot 178963 in writing this documentation. 178964 178965commit ab35c3fbc135bafdfc5057ef5d6227ca3534ed26 178966Author: Eric Anholt <anholt@freebsd.org> 178967Date: Thu Mar 9 23:18:15 2006 +0000 178968 178969 Do a first pass of doxygen documentation of EXA. This removes the 178970 corresponding pieces of exa-driver.txt, which were becoming stale. 178971 Hopefully the documentation will stay much more up-to-date this way. 178972 Many thanks to jbarnes for writing exa-driver.txt which was used a lot 178973 in writing this documentation. 178974 178975commit d8f8bfeccef0750d79f852b9ae7152e841227d5a 178976Author: Matthias Hopf <mhopf@suse.de> 178977Date: Thu Mar 9 14:23:57 2006 +0000 178978 178979 Do Xorg configure checks for Xgl only as well 178980 178981commit 2822cbc1fb2271844e7ae10c3629aaa940ae4042 178982Author: Eric Anholt <anholt@freebsd.org> 178983Date: Thu Mar 9 06:04:07 2006 +0000 178984 178985 Rearrange EXA driver structures so that there's a hope of maintaining ABI 178986 when extending the driver interface. The card and accel structures are 178987 merged into the ExaDriverRec, which is to be allocated using 178988 exaDriverAlloc(). The driver structure also grows exa_major and 178989 exa_minor, which drivers fill in and have checked by EXA 178990 (double-checking that the driver really did check that the EXA version 178991 was correct). Removes exaInitCard(), which is replaced by the driver 178992 filling in the rec by hand, and the exaGetVersion() and related 178993 EXA_*VERSION which are replaced by always using the XFree86 loadable 178994 module versioning. 178995 178996commit 65aa33f9173b1554924437685698f7c5f645a3c4 178997Author: Lars Knoll <lars@trolltech.com> 178998Date: Wed Mar 8 06:19:37 2006 +0000 178999 179000 render/picture.c Initialize the format of a source picture to 179001 PICT_a8r8g8b8. Fixes a failure in the gradients test of rendercheck. In 179002 the long term we could do better by setting the format to something 179003 without alpha whenever the gradient doesn't contain colors with alpha. 179004 This triggers a reduction of the over operation to a pure source 179005 operation. 179006 179007commit cb5090e8d60f4e9780c859faeea5c24587f6bee7 179008Author: Eric Anholt <anholt@freebsd.org> 179009Date: Wed Mar 8 03:32:07 2006 +0000 179010 179011 Bug #6150: Do the obvious fix of an insane sanity check in 179012 xf86InitFBManager. (Julio M. Merino Vidal) 179013 179014commit 2e6f801fe1a749f6a4db2cfd8a43abec5caceae0 179015Author: Ian Romanick <idr@us.ibm.com> 179016Date: Tue Mar 7 23:58:22 2006 +0000 179017 179018 Numerous amounts refactoring and comment adding (see ChangeLog for file by 179019 file details). The primary intention for these changes is to pave the 179020 way for the new device probing and PCI configuration code that I'm 179021 working on. 179022 179023commit b7d2dfc1e5e07051732303731ff3e4e76852dd94 179024Author: Eric Anholt <anholt@freebsd.org> 179025Date: Tue Mar 7 20:06:15 2006 +0000 179026 179027 Add appropriate MIT license. Oops. 179028 179029commit 9d8c0e4bcbb111e860b7c3c33c224c22589006b1 179030Author: Eric Anholt <anholt@freebsd.org> 179031Date: Tue Mar 7 19:57:46 2006 +0000 179032 179033 Add a new flag to ephyr, "-fakexa", which turns on an EXA acceleration 179034 implementation that calls fb to get its work done. The purpose is to 179035 have a trusted EXA driver for use with testing changes to the core of 179036 EXA. However, fakexa has not received much testing yet, lacks offscreen 179037 pixmaps support, and doesn't reliably provide garbage when EXA doesn't 179038 get its syncing right. All of these should be fixed soon. 179039 179040commit 0a3d6c739968bf5af81fc0e8ea7211c20d52080b 179041Author: Eric Anholt <anholt@freebsd.org> 179042Date: Tue Mar 7 19:49:31 2006 +0000 179043 179044 Remove stale EXA files, which failed to get removed during the move to 179045 top-level, somehow. 179046 179047commit 68a8963f726cb92624665669813b6d952d53556e 179048Author: Luc Verhaegen <libv@skynet.be> 179049Date: Tue Mar 7 16:00:57 2006 +0000 179050 179051 Fix cvt -r check again. CH7011 TV encoder had 800x600 PAL hit the check. 179052 179053commit 0693083335185ce05ee64546151f3fc43ce98575 179054Author: Lars Knoll <lars@trolltech.com> 179055Date: Mon Mar 6 21:00:09 2006 +0000 179056 179057 render/picture.c Correctly initialize devPrivates variable in source only 179058 pictures to 0 179059 miext/cw/cw.h Don't try to access devPrivates of source only pictures 179060 179061commit 448997ebcd2bab02be1059b07b91b63b0d05d268 179062Author: Matthieu Herrb <matthieu.herrb@laas.fr> 179063Date: Sun Mar 5 16:43:10 2006 +0000 179064 179065 Only output SetClientVersion message if verbosity > 1, like other 179066 extensions do 179067 179068commit d921173833cc207380eb08b6675393f5e8139d5f 179069Author: Matthieu Herrb <matthieu.herrb@laas.fr> 179070Date: Sun Mar 5 16:35:08 2006 +0000 179071 179072 define SYS_LIBS to hold system dependant libraries that may needed. and add 179073 it to libraries list where needed. Update ChangeLog for previous 179074 changes too 179075 179076commit 82cbd2ee0d20225b9edbb5246c8ed116b4614e1a 179077Author: Matthieu Herrb <matthieu.herrb@laas.fr> 179078Date: Sun Mar 5 16:33:17 2006 +0000 179079 179080 Don't hard-code -DUSE_DEV_IO here. configure generates the proper OS 179081 specific values here. 179082 179083commit b56a1513d27f84dcd55f3dc6053f183aa6f7855b 179084Author: Matthieu Herrb <matthieu.herrb@laas.fr> 179085Date: Sun Mar 5 16:32:40 2006 +0000 179086 179087 Definitions for bswapxx() macros on OpenBSD. 179088 179089commit 4335868476af7c821c64def52b102b93ae91f8b0 179090Author: Matthieu Herrb <matthieu.herrb@laas.fr> 179091Date: Sun Mar 5 16:13:21 2006 +0000 179092 179093 Fix build with non GNU make. 179094 179095commit b2f8f410c0bb8bc24039b2a593f8a2a483659914 179096Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179097Date: Fri Mar 3 09:54:54 2006 +0000 179098 179099 https://bugs.freedesktop.org/show_bug.cgi?id=4341 Make Xming error messages 179100 more meaningful. 179101 179102commit 29237c1977e454511e0d0244c68d34d572b68458 179103Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179104Date: Fri Mar 3 09:50:55 2006 +0000 179105 179106 https://bugs.freedesktop.org/show_bug.cgi?id=4538 Fix mouse button release 179107 on multiwindows scrolling. 179108 179109commit 06f01623fde61f1a11c2c1ecfae6a4c346473b05 179110Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179111Date: Fri Mar 3 09:43:42 2006 +0000 179112 179113 https://bugs.freedesktop.org/show_bug.cgi?id=5138 Check for NULL pointer 179114 179115commit 054c291b274b238893e408e070aef13a7933400b 179116Author: Felix Kuehling <fxkuehl@gmx.de> 179117Date: Thu Mar 2 18:35:08 2006 +0000 179118 179119 Fix build against Mesa CVS HEAD: added s_blit.c to symlink-mesa.sh. 179120 179121commit c1a82b9554028640dc4e08f042f1a8faf3372627 179122Author: Brian Paul <brian.paul@tungstengraphics.com> 179123Date: Thu Mar 2 03:43:26 2006 +0000 179124 179125 added s_blit.c file 179126 179127commit 5f4d11c8d926cf396e0a8e203e14a8e1e123e011 179128Author: Jesse Barnes <jbarnes@virtuousgeek.org> 179129Date: Wed Mar 1 16:31:53 2006 +0000 179130 179131 fix spelling error, document EXA_TWO_BITBLT_DIRECTIONS device flag 179132 179133commit 044a3abb382a4850722c391f04d09d3160790814 179134Author: Jesse Barnes <jbarnes@virtuousgeek.org> 179135Date: Wed Mar 1 16:28:34 2006 +0000 179136 179137 Add accelerated two directional blt support to EXA 179138 179139commit 96ca329382141fd50dccb1cc35a71a333d80bce4 179140Author: Ian Romanick <idr@us.ibm.com> 179141Date: Tue Feb 28 23:07:09 2006 +0000 179142 179143 Remove redundant definition of struct Inst. Safeguard xf86AddDriver against 179144 future additions to DriverRec. 179145 179146commit 1cfa9f647e0241f4b9e56556b128d7bfd987eaca 179147Author: Daniel Stone <daniel@fooishbar.org> 179148Date: Tue Feb 28 16:55:26 2006 +0000 179149 179150 Bug #5216: Allow options to appear with other components. 179151 179152commit e3b6b95f29cb2ea00b4290d694c5e202b8d180ad 179153Author: Adam Jackson <ajax@nwnk.net> 179154Date: Tue Feb 28 16:26:16 2006 +0000 179155 179156 Bug #5627: Fix Xprint font symlinking. (TIlman Sauerbeck) 179157 179158commit e7f0b84fa7bd0c40cb456ec4e447103442c8dae3 179159Author: Jesse Barnes <jbarnes@virtuousgeek.org> 179160Date: Tue Feb 28 05:20:20 2006 +0000 179161 179162 fix exaInitCard by making it a real function 179163 179164commit 088e5768faa90fe16de41b135b1111b5d25c64ad 179165Author: Felix Kuehling <fxkuehl@gmx.de> 179166Date: Mon Feb 27 18:12:24 2006 +0000 179167 179168 Fixing the Mesa build again, sigh. Add slang_execute_x86.c. Add 179169 -I../shader/slang to swrast INCLUDES. 179170 179171commit 345d99c972cac67f2cdc38750e4ba2dea1cdb360 179172Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 179173Date: Mon Feb 27 16:19:39 2006 +0000 179174 179175 Typo fixes (Nicholas Joly, XFree86 bugzilla #1658) 179176 179177commit 6b08a5013b4e9e350ba461c9a59d30bb41feef8f 179178Author: Jesse Barnes <jbarnes@virtuousgeek.org> 179179Date: Sat Feb 25 20:26:49 2006 +0000 179180 179181 EXA driver doc cleanups and additions. 179182 179183commit f41ec003f39c575299429897d4287233184583ad 179184Author: Roland Scheidegger <rscheidegger_lists@hispeed.ch> 179185Date: Sat Feb 25 01:17:10 2006 +0000 179186 179187 Add two radeon pci ids (one is used for a radeon mobility X700 XL in a 179188 medion notebook, the other is for a AIW X800 VE) 179189 179190commit 01a0bf881ada03ca3c27bdef7423c760c3bc2f9c 179191Author: Matthieu Herrb <matthieu.herrb@laas.fr> 179192Date: Fri Feb 24 17:01:57 2006 +0000 179193 179194 OpenBSD supports PCVT and WSCONS. PCCONS is long gone. 179195 179196commit d3e1587c20c155b7873b6646ddf0b96f806f8a7f 179197Author: Kristian Høgsberg <krh@redhat.com> 179198Date: Fri Feb 24 16:50:42 2006 +0000 179199 179200 file glxdri.c was initially added on branch accel_indirect_branch. 179201 179202commit 5d9a620726d2b0ad89625574478d2fd4536485b0 179203Author: Adam Jackson <ajax@nwnk.net> 179204Date: Thu Feb 23 19:25:57 2006 +0000 179205 179206 Remove redundant composite op reduction, done in Render now. 179207 179208commit 028d6903f674fa77617f333b25356710d1682b05 179209Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 179210Date: Wed Feb 22 16:29:07 2006 +0000 179211 179212 Bug #5978: Added missing swap of input variables. Added missing cases for 179213 GL_SECONDARY_COLOR_ARRAY and GL_FOG_COORD_ARRAY (Colin McDonald) 179214 179215commit 43324132afcbb6b231efcc24ec72ee44678d5771 179216Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 179217Date: Wed Feb 22 16:08:56 2006 +0000 179218 179219 Added Mesa include directory 179220 179221commit 43fbcc28c960ce5abe1d3223441c3dc5a10cde27 179222Author: Jeremy C. Reed <reed@reedmedia.net> 179223Date: Wed Feb 22 02:47:00 2006 +0000 179224 179225 Update to build against Mesa HEAD. (Thank you Felix for feedback on xorg 179226 list.) 179227 179228commit cfdacab33a62b47f22bb54683e3ca20ec9824864 179229Author: Ian Romanick <idr@us.ibm.com> 179230Date: Tue Feb 21 00:02:08 2006 +0000 179231 179232 Eliminate unused PCI BIOS reading functionality. The old code used several 179233 function pointers to implement a level of flexability that was never 179234 used. The code also had unused support for extracting a single image 179235 type from a larger expansion ROM. 179236 Fix the spelling of PCI_BIOS_OPEN_FIRMWARE. 179237 Fix a couple errors in #ifdef debug code. 179238 These changes have been tested on x86 and x86-64 Linux. 179239 179240commit 5fd0f94006775e2271107c960dfa3314dddd9a5f 179241Author: Ian Romanick <idr@us.ibm.com> 179242Date: Mon Feb 20 23:45:50 2006 +0000 179243 179244 Eliminate all the code for querying the PCI class from the PCI ID database. 179245 Class information is not, and never has been, stored there. Therefore, 179246 this is just a bunch of elaborate code to read 0x00000000. 179247 This has received testing on x86 and x86-64 Linux. 179248 179249commit 6d7083bd69724586338d79784655328f1fcd2ae5 179250Author: Adam Jackson <ajax@nwnk.net> 179251Date: Mon Feb 20 22:16:49 2006 +0000 179252 179253 indent fixes (OMG SO UGLY), and nuke old RCS keywords. 179254 179255commit 5480c537cee79b324736eac3e438a4713dfa1036 179256Author: Adam Jackson <ajax@nwnk.net> 179257Date: Mon Feb 20 21:50:49 2006 +0000 179258 179259 ANSIfy Xi/. Mostly automated via protoize(1). 179260 179261commit cc42e153c2437fe9c94b0c20e1b56277474d94d0 179262Author: Felix Kuehling <fxkuehl@gmx.de> 179263Date: Mon Feb 20 03:00:09 2006 +0000 179264 179265 Update to build against Mesa HEAD. 179266 179267commit ed3ea887a6b1c9fdc83895c40da34076121f05e9 179268Author: Eric Anholt <anholt@freebsd.org> 179269Date: Thu Feb 16 21:49:51 2006 +0000 179270 179271 Fix the encoding of DES's name. 179272 179273commit ea0b3f65f77a78df1671b09739c00762a8875607 179274Author: Eric Anholt <anholt@freebsd.org> 179275Date: Thu Feb 16 21:45:12 2006 +0000 179276 179277 Fix build of Xorg by putting xf86bigfont back into builtin instead of 179278 module sources list. 179279 179280commit d97f29be22e22e6f5bc23229ffa5ef087f992c8c 179281Author: Eric Anholt <anholt@freebsd.org> 179282Date: Thu Feb 16 21:35:32 2006 +0000 179283 179284 Bug #5453: Don't forget to still do AM_CONDITIONAL for XVMC even if XV is 179285 disabled, and also force XVMC disabled if XV is disabled. (Dag-Erling 179286 Sm�rgrav) 179287 179288commit d33c2e0d1079f93f4ba8b28d19950d384a9e7c32 179289Author: Keith Packard <keithp@keithp.com> 179290Date: Thu Feb 16 19:36:39 2006 +0000 179291 179292 Add oscolor.h 179293 179294commit 60d4839a2d373cc6d8c0d5004284494d3a994a63 179295Author: Keith Packard <keithp@keithp.com> 179296Date: Thu Feb 16 19:03:07 2006 +0000 179297 179298 Forgot to include this file in previous patch 179299 179300commit 8987b2c1efc9a4667e278e6ba411772ba2a4a4e6 179301Author: Keith Packard <keithp@keithp.com> 179302Date: Thu Feb 16 07:17:31 2006 +0000 179303 179304 Make more extensions optional in build (for kdrive). Fix kdrive build for 179305 actual hardware. Fix kdrive pointer signed/unsigned types. Add 179306 kdrive-required YX rotation functions. Replace rgb text file loading 179307 with static rgb color table. 179308 179309commit c8acb342695936db062c966029019a458d45459e 179310Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 179311Date: Thu Feb 16 06:52:12 2006 +0000 179312 179313 Fix an occasional crash on VT switches: the server would save a pointer to 179314 the current cursor when disabling FB access and would try to restore 179315 that cursor when re-enabling. However, that cursor might have been 179316 destroyed in between. This fixes it by updating the saved cursor 179317 pointer when a cursor is set and vtSema is FALSE. 179318 179319commit c845e152f52b3e8cef579797c5c8834ad2fd2cd5 179320Author: Eric Anholt <anholt@freebsd.org> 179321Date: Thu Feb 16 05:46:08 2006 +0000 179322 179323 Bug #5871: Drop special build infrastructure left over from libcwrapper 179324 times. (George Fufutos) 179325 179326commit c2f685e64464ccf86ed47ae37f74bf46877e0739 179327Author: Dave Airlie <airlied@linux.ie> 179328Date: Thu Feb 16 05:18:20 2006 +0000 179329 179330 Fix XSERVER64 should be _XSERVER64 179331 179332commit 1bbd5e49b1dcc2e3f9785bd2bb58f946b9998099 179333Author: Eric Anholt <anholt@freebsd.org> 179334Date: Thu Feb 16 04:59:45 2006 +0000 179335 179336 Bug #5869: Remove traces of EXTMODULE define, which doesn't appear to be 179337 useful any more. (George Fufutos) 179338 179339commit c03b06bdf04fa8500d0f85314c7268848b4d50be 179340Author: Eric Anholt <anholt@freebsd.org> 179341Date: Thu Feb 16 04:39:00 2006 +0000 179342 179343 Bug #5888: Remove orphaned laymodule.c from miext/layer removal. (George 179344 Fufutos) 179345 179346commit 3b32e902c7a12aa2320da27d984029cde28fd8c3 179347Author: Eric Anholt <anholt@freebsd.org> 179348Date: Thu Feb 16 01:49:23 2006 +0000 179349 179350 Add entry missed in my last commit. 179351 179352commit c4767794ef5b014ae25fe8541e72348ecfb1ee49 179353Author: Zephaniah E. Hull <warp@aehallh.com> 179354Date: Thu Feb 16 01:03:09 2006 +0000 179355 179356 Export xf86ActivateDevice, used by the evdev driver. 179357 179358commit dc0354104cb4057dfcc7b2ccb8e2ae8474d70b15 179359Author: Eric Anholt <anholt@freebsd.org> 179360Date: Thu Feb 16 00:14:11 2006 +0000 179361 179362 Move EXA implementation up to the top level and remove its XFree86 179363 dependencies. It was nearly abstract enough already to be used by 179364 multiple DDXes. This will be useful for EXA development through 179365 providing a fake acceleration implementation within Xephyr, so that 179366 testing can be done on new EXA code without worrying about buggy 179367 drivers. 179368 179369commit c170aa830d0ce3dbff6b30081e04c3f91bf921be 179370Author: Eric Anholt <anholt@freebsd.org> 179371Date: Wed Feb 15 23:27:40 2006 +0000 179372 179373 Forced commit to note repocopy from hw/xfree86/exa/ 179374 179375commit 6770f1bdb145e7a6c431d0523f10d12155f58273 179376Author: Eric Anholt <anholt@freebsd.org> 179377Date: Wed Feb 15 21:09:14 2006 +0000 179378 179379 Define NO_LIBCWRAPPER in dix-config.h, and rely on Mesa including 179380 dix-config.h if DIX_HAVE_CONFIG_H is defined to get it and _XSERVER64, 179381 instead of defining things like this per directory. 179382 179383commit 7d7fc927cd90146788780477b8e3379d91c3b910 179384Author: Adam Jackson <ajax@nwnk.net> 179385Date: Wed Feb 15 20:47:44 2006 +0000 179386 179387 Remove a few #ifdef vms; whatever problem that was solving should assuredly 179388 be solved some other way. 179389 179390commit f105b8da11fcf337512b3c39da3368f98da07a33 179391Author: Adam Jackson <ajax@nwnk.net> 179392Date: Wed Feb 15 20:44:13 2006 +0000 179393 179394 Mark everything in dixsym.c as _X_EXPORT. 179395 179396commit 010d6effa6fa210251b12459882e88aeee82c2c0 179397Author: Adam Jackson <ajax@nwnk.net> 179398Date: Wed Feb 15 19:15:32 2006 +0000 179399 179400 Mark everything in {ext,font}sym.c as _X_EXPORT. 179401 179402commit 3fe482c77e7b3e46739d011d8bbdee527d7a42fc 179403Author: Adam Jackson <ajax@nwnk.net> 179404Date: Wed Feb 15 19:05:55 2006 +0000 179405 179406 Mark everything in misym.c as _X_EXPORT. 179407 179408commit 6ad4325b87889e1aada9333d750b7bb586c38b52 179409Author: Kristian Høgsberg <krh@redhat.com> 179410Date: Wed Feb 15 18:26:45 2006 +0000 179411 179412 Update to build against mesa head. 179413 179414commit 50e2ff9a2500078ebbd833fddab0d93f3a50b6a6 179415Author: Eric Anholt <anholt@freebsd.org> 179416Date: Wed Feb 15 03:20:55 2006 +0000 179417 179418 Remove the waitSync from KdDisableScreen and push it off to drivers' 179419 disableAccel hook, which is more correct anyway. This makes kdrive.c 179420 not have any knowledge of kaa, opening the way for using exa from 179421 kdrive. 179422 179423commit 0446aafa9467f43515fb578d50f45e2c3153c8cf 179424Author: Eric Anholt <anholt@freebsd.org> 179425Date: Wed Feb 15 03:07:23 2006 +0000 179426 179427 Avoid some more libcwrapper damage that prevented kdrive linking. 179428 179429commit 5c9b6f0fb01252d704de1bbdf3015dee7f956593 179430Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 179431Date: Tue Feb 14 08:14:42 2006 +0000 179432 179433 DRIGetSecs() would call getsecs() when XFree86LOADER is defined, relying on 179434 the wrappers to provide it. Wrapper gone, and getsecs doesn't exist on 179435 linux so it now blows up. Fixes it by just calling gettimeofday() in 179436 all cases instead. 179437 179438commit 049dca0f43eb2179d2c61033a17ff1a89f8fb689 179439Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 179440Date: Tue Feb 14 08:11:41 2006 +0000 179441 179442 Remove useless line of code that contained a bug and triggered a gcc 179443 warning. This variable will be overriden before being used anyway. 179444 (Bugzilla #5595) 179445 179446commit 1132d0e6102d4564f70f0e8c98854e3acf25b109 179447Author: Dave Airlie <airlied@linux.ie> 179448Date: Tue Feb 14 06:27:59 2006 +0000 179449 179450 update to latest Mesa CVS HEAD 179451 179452commit bb8c36690ab411c11aa8dd3d4520d513eb8f9091 179453Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179454Date: Tue Feb 14 04:20:37 2006 +0000 179455 179456 Bump shadow module version number to 1.1.0 from 1.0.0 179457 179458commit cc9dfab0b31c7956f99d1f1b9c195065b5e18c29 179459Author: Adam Jackson <ajax@nwnk.net> 179460Date: Mon Feb 13 18:57:38 2006 +0000 179461 179462 (Reverted) 179463 179464commit 83dd6241c8cd81e8d897bd17588ada92a945e647 179465Author: Adam Jackson <ajax@nwnk.net> 179466Date: Mon Feb 13 18:55:44 2006 +0000 179467 179468 Reverted, did nothing anyway, I'm not smart today. 179469 179470commit 4a7f6f53cad541e8c5042a6472e3b3886fc9b7e6 179471Author: Adam Jackson <ajax@nwnk.net> 179472Date: Mon Feb 13 18:09:51 2006 +0000 179473 179474 Further op reduction when both src and dst alpha are absent. 179475 179476commit 28ced9f3e0dd4bd81067f590a1d64ba0844edb06 179477Author: Eric Anholt <anholt@freebsd.org> 179478Date: Mon Feb 13 05:29:00 2006 +0000 179479 179480 Add missing ChangeLog text for r1.2 of GL/glx/indirect_reqsize.h 179481 179482commit 4839e91fcab4c344e672154a447d8c7035fce1f4 179483Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 179484Date: Mon Feb 13 05:03:13 2006 +0000 179485 179486 HAS_MKSTEMP vs. HAVE_MKSTEMP (From Fredrik Höglund) 179487 179488commit 1a4f20541a9f4f41f444d826d743899ea2dee2db 179489Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 179490Date: Mon Feb 13 04:56:27 2006 +0000 179491 179492 Fix linux build without libc wrappers (From Fredrik Höglund) 179493 179494commit 2dc7b5e0d96a187bfbb355caa788f0fdcd88eaad 179495Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 179496Date: Mon Feb 13 04:43:40 2006 +0000 179497 179498 Move call to xf86WrapperInit() to OsVendorInit() in xf86Init.c and remove 179499 stubs in other DDX. 179500 179501commit 34d0b9228f46c2f87be74dddc9c7d97aab091d03 179502Author: Eric Anholt <anholt@freebsd.org> 179503Date: Sun Feb 12 20:53:35 2006 +0000 179504 179505 Simplify ops that would use the alpha channel when an alpha channel is 179506 always 1.0, and short circuit PictOpDst for good measure. 179507 179508commit 5f45776ef3b9256bea44842d1c50f269422531a1 179509Author: Eric Anholt <anholt@freebsd.org> 179510Date: Sun Feb 12 10:30:47 2006 +0000 179511 179512 Add missing HAVE_DIX_CONFIG_H which caused issues with mismatched screen 179513 structure interpretations, and remove a bunch of unused junk from 179514 kdrive-config.h. Xephyr almost works on my amd64. 179515 179516commit 5249416d091d59c248c8dda44529b8aa4910b1a0 179517Author: Eric Anholt <anholt@freebsd.org> 179518Date: Sat Feb 11 22:40:50 2006 +0000 179519 179520 Add stub xf86WrapperInits so that the servers will build even if os/ was 179521 built with XFree86LOADER set. 179522 179523commit a2a5254675a6b7ef0f7da9caa76c028b7c526502 179524Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 179525Date: Sat Feb 11 19:16:51 2006 +0000 179526 179527 Add <string.h>, <stdlib.h>, and <stdio.h> to clear undefined function 179528 warnings after the removal of libcwrapper headers. 179529 179530commit d6337c83241f0fa4bb03039a9767b58d8a1a7c91 179531Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 179532Date: Sat Feb 11 17:42:31 2006 +0000 179533 179534 <sys/limits.h> -> <limits.h> so we can compile on non-BSD OS'es 179535 179536commit 4fafba61d5402d4e4d2c21ba1be3ed8969b99334 179537Author: Eric Anholt <anholt@freebsd.org> 179538Date: Sat Feb 11 03:03:45 2006 +0000 179539 179540 Remove libcwrapper damage from GLX (requires fresh Mesa HEAD), and get it 179541 compiling in kdrive. 179542 179543commit c3d14036729fd186d4ec7ca1de603e1f2d174e2f 179544Author: Eric Anholt <anholt@freebsd.org> 179545Date: Fri Feb 10 22:00:30 2006 +0000 179546 179547 Remove libcwrapper usage from xorg server modules. The libcwrapper is only 179548 of (marginal) use in the drivers, and that usage remains. 179549 179550commit a8cec1b656f57746758613213de1d6e5acb79451 179551Author: Eric Anholt <anholt@freebsd.org> 179552Date: Fri Feb 10 09:00:02 2006 +0000 179553 179554 Merge from kdrive: use RECT_PRIM to avoid tearing in xvideo. 179555 179556commit efc3fab7f4b29f56fffd21304c64c03a48aa5b4b 179557Author: Eric Anholt <anholt@freebsd.org> 179558Date: Fri Feb 10 07:52:05 2006 +0000 179559 179560 Make kdrive (i.e. Xephyr only) buildable on FreeBSD and probably other OSes 179561 without linux VT switching, fbdev, and vm86 support. 179562 179563commit fa3a65e33d8c893c7867ea507afc7caa1361aa9c 179564Author: Eric Anholt <anholt@freebsd.org> 179565Date: Fri Feb 10 07:47:21 2006 +0000 179566 179567 Remove the include of X11/misc.h, which fails to compile and isn't 179568 necessary. 179569 179570commit d875bdb2756b082ce93bd86016c369ea85c04d17 179571Author: Jeremy C. Reed <reed@reedmedia.net> 179572Date: Fri Feb 3 02:44:19 2006 +0000 179573 179574 hw/xfree86/os-support/xf86_OSlib.h Fix sysmouse handling on DragonFly, 179575 mostly garbage arrived. From Joerg Sonnenberger. 179576 179577commit 0946bb9427695a4314e5c43de573b3a75a18e466 179578Author: Jeremy C. Reed <reed@reedmedia.net> 179579Date: Fri Feb 3 02:37:52 2006 +0000 179580 179581 Fix sysmouse handling on DragonFly, mostly garbage arrived. From Joerg 179582 Sonnenberger. 179583 179584commit 5e2a7af23bd0f46fbddca34098cb297be58b7a55 179585Author: Eric Anholt <anholt@freebsd.org> 179586Date: Thu Feb 2 21:07:06 2006 +0000 179587 179588 Move the frequently-repeated code to get the pixmap that backs a drawable 179589 to a new function, exaGetDrawablePixmap(). 179590 179591commit ee3c7ccb175752dbeaed6b0113d0819b3fcd2398 179592Author: Eric Anholt <anholt@freebsd.org> 179593Date: Thu Feb 2 20:51:54 2006 +0000 179594 179595 Remove more debugging leftovers. 179596 179597commit 3366b6836572461209bb2f8aa28d9e662067dc54 179598Author: Eric Anholt <anholt@freebsd.org> 179599Date: Thu Feb 2 20:09:14 2006 +0000 179600 179601 Rearrange and rename EXA code to be a bit more logically organized. Also 179602 removes a little bit of debugging leftovers. Summary: 179603 exa.c -> exa.c (miscellaneous code) exa_accel.c (all acceleration code) 179604 exa_migration.c (migration logic) exaasync.c -> exa_unaccel.c (software 179605 fallbacks) exapict.c -> exa_render.c (render extension stuff) 179606 exaoffscreen.c -> exa_offscreen.c exaPriv.h -> exa_priv.h 179607 179608commit 25d4ff870d49533d82a77f144722ff7934d52e0b 179609Author: Eric Anholt <anholt@freebsd.org> 179610Date: Thu Feb 2 20:04:15 2006 +0000 179611 179612 Forced commit to note repocopies: exa.c -> exa_accel.c exa_migration.c 179613 exaasync.c -> exa_unaccel.c exapict.c -> exa_render.c exaoffscreen.c -> 179614 exa_offscreen.c exaPriv.h -> exa_priv.h 179615 179616commit 2ab487d4d20e4e34e73cc6d87f41bf0836c7a8af 179617Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179618Date: Wed Feb 1 22:20:05 2006 +0000 179619 179620 Add a new function RRGetRotation() which does exactly the same thing as 179621 xf86GetRotation(), but allows for drivers to provide their own RandR 179622 implementation. xf86GetRotation could be obsoleted by this change. 179623 179624commit f4898b409376803c9a9dd8475bdd5576ff1cc59d 179625Author: Luc Verhaegen <libv@skynet.be> 179626Date: Tue Jan 31 14:49:43 2006 +0000 179627 179628 Further bug #5386 fixes: Fix some problems with the EDID code: Some 179629 bitoffsets were wrong. Unknown Detailed Sections weren't handled 179630 properly and defaulted to Detailed Timing. 179631 179632commit 437b385ce4cc3ff00e14d3d39f4a2f6c8f0c67a0 179633Author: Luc Verhaegen <libv@skynet.be> 179634Date: Tue Jan 31 13:55:01 2006 +0000 179635 179636 Bring the cvt utility up to date with bug #5386 changes. Fix 2 issues with 179637 the generator routine: the allocated modeline wasn't nulled and 179638 mode->name's \0 wasn't copied over. PrintModeLine was rewritten and 179639 HDisplay gets rounded up to character width instead of refused. 179640 179641commit 8f3c69dcf17691f71bca7b0a2cd34f7788a97b8c 179642Author: Luc Verhaegen <libv@skynet.be> 179643Date: Tue Jan 31 13:04:02 2006 +0000 179644 179645 Accept modes with less than 25% horizontal blanking again (you can push old 179646 gtf timing to below 25%), only stop cvt reduced blanking. Users should 179647 be free to blow up their monitors if they so choose. 179648 179649commit 701b63cf1dcd3e49602114fb1dde45a74b4e1122 179650Author: Donnie Berkholz <spyderous@gentoo.org> 179651Date: Mon Jan 30 20:04:56 2006 +0000 179652 179653 Update to build against Mesa trunk. 179654 179655commit dd50015b05b901fe0c60717512c854389610aea2 179656Author: Eric Anholt <anholt@freebsd.org> 179657Date: Sat Jan 28 02:20:37 2006 +0000 179658 179659 Add libc_wrapper support for random(), which will be used in upcoming EXA 179660 memory manager work. 179661 179662commit 3d1667278ff309d7f8e61a6d330f712bae5bcd41 179663Author: Eric Anholt <anholt@freebsd.org> 179664Date: Sat Jan 28 00:37:52 2006 +0000 179665 179666 Remove leftover variables for cfb24 build, and finish commenting out Xglx 179667 standalone stuff, which some versions of automake get whiny about. 179668 179669commit ab01eb247f9e5d7c9995bf2d6432358cd64bf11d 179670Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179671Date: Fri Jan 27 12:27:34 2006 +0000 179672 179673 update pci ids 179674 179675commit 261aa4403c77203f8f02b399ddd382c731dda324 179676Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179677Date: Thu Jan 26 09:04:22 2006 +0000 179678 179679 remove that, and will put it in a i810 driver specific Changelog 179680 179681commit 94e678fd014c61d12591d7398b6591f24c3d71f1 179682Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179683Date: Thu Jan 26 08:49:19 2006 +0000 179684 179685 add changelog for i810 updates 179686 179687commit c5e93182905332383ca3ef5db3f334cec69c8dda 179688Author: Donnie Berkholz <spyderous@gentoo.org> 179689Date: Thu Jan 26 04:32:45 2006 +0000 179690 179691 Really allow linking against Mesa trunk to work. 179692 179693commit 0dc0f17f27f99da79c99031b41b0c0e95ef035f5 179694Author: Adam Jackson <ajax@nwnk.net> 179695Date: Thu Jan 26 04:10:43 2006 +0000 179696 179697 Speed up checkout and autogen by removing disused iplan2p4 and ilbm. 179698 179699commit 023d2b4e3c392eed1f149dc5b13a83429cd052a3 179700Author: Daniel Stone <daniel@fooishbar.org> 179701Date: Thu Jan 26 00:23:44 2006 +0000 179702 179703 Add forgotten HAVE_BACKTRACE define. 179704 179705commit 14fdd81614cdd6ef7e01976a43da8b6a3bf8386e 179706Author: Adam Jackson <ajax@nwnk.net> 179707Date: Wed Jan 25 23:05:26 2006 +0000 179708 179709 Remove xf8_32wid, it's dead code with no maintainer interest. Also remove 179710 cfb24, since xf8_32wid was the only user. 179711 179712commit 2e28f4104ddf94a8f9a70fe6b2a2a6859ffedc8f 179713Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179714Date: Tue Jan 24 22:05:33 2006 +0000 179715 179716 Allow current trunk to build against Mesa trunk 179717 179718commit 0d9ed2624fe8fb95c57930da523351556ba11351 179719Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179720Date: Mon Jan 23 22:01:34 2006 +0000 179721 179722 Commit #4633 - Initial mouse pointer incorrect with EXA which also fixes 179723 the repaint of the cursor image with randr events. 179724 179725commit 9148d8700b7c5afc2644e5820c57c509378f93ce 179726Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179727Date: Mon Jan 23 13:59:14 2006 +0000 179728 179729 Commit slight variation of bug #5460 which is the merge of the new shadow 179730 code from kdrive. 179731 179732commit cfd3988ed906ab48ca4362256f8dbb8852d7ca0a 179733Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179734Date: Mon Jan 23 13:58:19 2006 +0000 179735 179736 wrap with if XORG / endif 179737 179738commit f51ecc66e9ad6d2c3541b1dafa7659da5a0a3a86 179739Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179740Date: Mon Jan 23 13:54:59 2006 +0000 179741 179742 #include "gcstruct.h" 179743 179744commit 80f45fa4dfa011c2ae7bcb34f87aafb91763f1fe 179745Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179746Date: Mon Jan 23 13:54:34 2006 +0000 179747 179748 add damage.h & damagestr.h to SDK headers 179749 179750commit af5b3ea4b3df9e9c6dd6993c5e7238a366a3f508 179751Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 179752Date: Thu Jan 19 14:51:09 2006 +0000 179753 179754 add randrstr.h to sdk_HEADERS 179755 179756commit 03ebd37baba2f5af3ab502ff02ec14c15859dc3f 179757Author: Daniel Stone <daniel@fooishbar.org> 179758Date: Thu Jan 19 12:25:01 2006 +0000 179759 179760 Make error() bomb with exit code 1, not 0. 179761 179762commit adce1f16e0d815e5c762407da3544a7d2eff9303 179763Author: Eric Anholt <anholt@freebsd.org> 179764Date: Thu Jan 19 00:06:57 2006 +0000 179765 179766 Only try to use byteswap.h on linux. Assume that everyone else (thinking of 179767 BSDs here) has sys/endian.h, and use macros as appropriate for the 179768 names. This should probably be in a gloabl header. 179769 179770commit 1c3f8727b2349c9b988eaa744f11366322d42538 179771Author: Adam Jackson <ajax@nwnk.net> 179772Date: Wed Jan 18 19:42:56 2006 +0000 179773 179774 More kdrive merge, fast path fbBlt to use memcpy() when possible. Good for 179775 -5% to 60% speedup on XGetImage, and 0% to 10% speedup on copies within 179776 host memory. Based on work by Jaymz Julian. 179777 179778commit e70b64b93024d05519014fb1b76fe26bd9f3a496 179779Author: Dave Airlie <airlied@linux.ie> 179780Date: Wed Jan 18 07:15:55 2006 +0000 179781 179782 Updated xgl code drop from Novell + xserver tree changes 179783 179784commit b5356e0afaf2b660c8905f63d5fdcb03402b81c5 179785Author: Dave Airlie <airlied@linux.ie> 179786Date: Wed Jan 18 07:00:50 2006 +0000 179787 179788 typo in last change 179789 179790commit 506eca5f57b960a6650c3387047a6ae8a22181e9 179791Author: Dave Airlie <airlied@linux.ie> 179792Date: Wed Jan 18 06:56:52 2006 +0000 179793 179794 Wrap sdk_HEADERS in if XORG as otherwise installing non-xorg servers 179795 breaks. 179796 179797commit 8ccf4f2b8fdb5e57d2ec5f2d54731fbf83fb9d8c 179798Author: Dave Airlie <airlied@linux.ie> 179799Date: Wed Jan 18 06:49:17 2006 +0000 179800 179801 This is a fix from David Reveman from the xserver tree, Make fbPadPixmap 179802 work with negative stride 179803 179804commit a1f9262c6acd195c0fcf5f602d5ca0c252993521 179805Author: Kristian Høgsberg <krh@redhat.com> 179806Date: Tue Jan 17 21:27:49 2006 +0000 179807 179808 file glxvisuals.c was initially added on branch accel_indirect_glx. 179809 179810commit c2dedf4d17f8a5b1a8037fd2b4e29122ef78945a 179811Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 179812Date: Thu Jan 12 22:14:56 2006 +0000 179813 179814 Thu Jan 12 17:09:18 2006 Søren Sandmann <sandmann@redhat.com> 179815 Add new functions to enable and disable events on Map and Unmap. 179816 Use them here to make sure Composite redirect doesn't cause Map/UnmapNotify 179817 events that would confuse window managers. 179818 179819commit 847d83ec3c90c5b298eaf19ba55251b4a30f4155 179820Author: Ian Romanick <idr@us.ibm.com> 179821Date: Thu Jan 12 00:21:59 2006 +0000 179822 179823 Bug #2996: libglx / libGLcore should use a dispatch table 179824 Port all changes from the (monolithic) accelerated_indirect-0-0-1 branch to 179825 the modular trunk. This will break the Darwin and cygwin builds. Other 179826 than the changes to symlink-mesa.sh and the various Makefile.am files, 179827 to code is identical to what's in the branch. 179828 Reviewed by: airlied, krh 179829 179830commit c56e9a8849ce8dd5c09732ae1860e409e7886690 179831Author: Daniel Stone <daniel@fooishbar.org> 179832Date: Tue Jan 10 03:24:53 2006 +0000 179833 179834 Bomb when symlink-mesa.sh continues, instead of silently failing. 179835 179836commit 2949c705f11f8710301555c039bcecbe748cabd9 179837Author: Daniel Stone <daniel@fooishbar.org> 179838Date: Tue Jan 10 03:23:05 2006 +0000 179839 179840 Add all subdirs used to DIST_SUBDIRS, and files to EXTRA_DIST. 179841 Attempt to build xeglmodule.c, not xglxmodule.c. 179842 Add xf86Sbus.h to EXTRA_DIST, as _HEADERS doesn't appear to get the same 179843 treatment as _SOURCES in terms of automatically DISTing. 179844 179845commit 4fc9eb592a446ad5711bdaa82c60e9fe010fd76a 179846Author: Daniel Stone <daniel@fooishbar.org> 179847Date: Tue Jan 10 02:32:20 2006 +0000 179848 179849 Bomb out if symlink-mesa.sh failed. 179850 179851commit 890ec849479db2510a9b4bc5e5e2f7978ca37b83 179852Author: Daniel Stone <daniel@fooishbar.org> 179853Date: Tue Jan 10 02:30:56 2006 +0000 179854 179855 Add xgl to DIST_SUBDIRS. 179856 179857commit 53dbd00a75313ec5301ca95b2e91d5d02bdaf820 179858Author: Adam Jackson <ajax@nwnk.net> 179859Date: Sun Jan 8 23:43:54 2006 +0000 179860 179861 Remove remaining #ifdef DPSEXT stanzas. 179862 179863commit 7fc9bc44e099f8f046bf707cb87ef7d736933f80 179864Author: Adam Jackson <ajax@nwnk.net> 179865Date: Sat Jan 7 01:29:05 2006 +0000 179866 179867 Compile fix, again, stupid non-clean builds 179868 179869commit 9d62d1e6903ccc095f784279a699b3f40a8f0cf8 179870Author: Adam Jackson <ajax@nwnk.net> 179871Date: Sat Jan 7 00:45:17 2006 +0000 179872 179873 Bug #5218: Don't crash on unconfigured interfaces. (Andrei Barbu) 179874 179875commit 3c58072956c28ebc3ca2eb50c1ff09823e1219d7 179876Author: Adam Jackson <ajax@nwnk.net> 179877Date: Sat Jan 7 00:33:41 2006 +0000 179878 179879 One more build fix. 179880 179881commit 07303c1b42afd1ada98cbc11d1ba616d366017fb 179882Author: Adam Jackson <ajax@nwnk.net> 179883Date: Sat Jan 7 00:05:46 2006 +0000 179884 179885 This version will actually compile 179886 179887commit 7f46aba35ee482e9b28ecc81d1a99d423fc88a70 179888Author: Adam Jackson <ajax@nwnk.net> 179889Date: Fri Jan 6 23:52:23 2006 +0000 179890 179891 Compile fix 179892 179893commit 25babf2791ad42101a86ba2a0f14564328256ee2 179894Author: Adam Jackson <ajax@nwnk.net> 179895Date: Fri Jan 6 23:36:53 2006 +0000 179896 179897 Missed file. 179898 179899commit 13c9e0c094c4e34cd1e43a7cc08b2dca39a32412 179900Author: Adam Jackson <ajax@nwnk.net> 179901Date: Fri Jan 6 23:06:15 2006 +0000 179902 179903 Bug #5525: Build a working Xprt. (Drew Parsons) 179904 179905commit fe0c838b5d8bc8d9cf5a686bb7d3e90682d2d19a 179906Author: Adam Jackson <ajax@nwnk.net> 179907Date: Fri Jan 6 18:06:02 2006 +0000 179908 179909 Move drawable lock acquisition into DRIClipNotify from DRIValidateTree, so 179910 we only take it when clipping a DRI drawable instead of on every tree 179911 update. Note drawable lock acquisition per- screen instead of globally, 179912 and drop it in BlockHandler if necessary. 179913 179914commit 39ce5f1544029412f4060f3e89ce1d87222ef42b 179915Author: Adam Jackson <ajax@nwnk.net> 179916Date: Fri Jan 6 17:05:26 2006 +0000 179917 179918 Remove unused X11R4 DDX compatibility function miClipNotify. 179919 179920commit 07ecb969d7eb8d4ab0bb0b8a55a5f40f3c8ec5e3 179921Author: Eric Anholt <anholt@freebsd.org> 179922Date: Wed Jan 4 03:29:15 2006 +0000 179923 179924 Forced commit to note that glyph privates commit was: 179925 Obtained from: xserver tree (David Reveman) 179926 179927commit b6b88d2f62d8c596171f487dd25fbdbc85d0c5a8 179928Author: Eric Anholt <anholt@freebsd.org> 179929Date: Wed Jan 4 00:05:16 2006 +0000 179930 179931 Correct rounding in divide-by-255 code. Obtained from xserver. 179932 179933commit b9c0ae867e1b52186c26841a77745f7f5a0a76dd 179934Author: Eric Anholt <anholt@freebsd.org> 179935Date: Tue Jan 3 22:36:46 2006 +0000 179936 179937 Remove the manual AddFilter for convolution, which I'm pretty sure 179938 shouldn't be necessary due to it already happening from 179939 PictureSetDefaultFilters. 179940 179941commit 601ab861b46a62b0742ffd3e937c4fab129664f0 179942Author: Eric Anholt <anholt@freebsd.org> 179943Date: Tue Jan 3 22:06:23 2006 +0000 179944 179945 Add glyph privates for Xgl, which uses them to implement a glyph cache. EXA 179946 would probably also like to do this. This breaks module ABI for EXA and 179947 XAA, and likely breaks proprietary drivers as well. 179948 179949commit 1729fc882ceec392331566c95efd5968fe9e97fd 179950Author: Eric Anholt <anholt@freebsd.org> 179951Date: Sat Dec 31 08:06:00 2005 +0000 179952 179953 Change REGION_INIT(pScreen, &foo, NullBox, 0) to REGION_NULL(pScreen, 179954 &foo). While it is no longer (or rather, once again not) required as of 179955 regionstr.h r1.4, it matches the style of the rest of the xorg code. 179956 179957commit c25536a7937b11a5347bfb8796d5cb6eb0445b51 179958Author: Eric Anholt <anholt@freebsd.org> 179959Date: Sat Dec 31 08:01:31 2005 +0000 179960 179961 Initialize the fourcc value in stack-allocated glitz_pixel_format_t 179962 structures. Greatly reduces the number of uninitialized-value accesses 179963 during Xgl startup according to valgrind. Allocating and filling these 179964 in by hand on the stack seems very shady to me. 179965 179966commit e6dab3d7c429a2d30d31f188c4554e870011e051 179967Author: Eric Anholt <anholt@freebsd.org> 179968Date: Fri Dec 30 12:05:47 2005 +0000 179969 179970 Fix the AC_TRY_RUN for sys/linker.h which had no hope due to lacking a 179971 main() to instead use a nice AC_CHECK_HEADERS that works. Also, fix the 179972 nearby SYSV IPC check which was lacking an argument and giving bogus 179973 results (it's "AC_TRY_LINK(includes, main, yes, no)"). 179974 179975commit eef16c36ad6e90fd8eaad4d8bdbc1205bc28a66f 179976Author: Eric Anholt <anholt@freebsd.org> 179977Date: Fri Dec 30 05:44:14 2005 +0000 179978 179979 Add #undef BSD44SOCKETS, without which the listener socket on FreeBSD would 179980 be created without the port number due to xtrans's define not being 179981 used when xtrans was compiled in the X Server. 179982 179983commit 7aa0ea23bc9b8df582fe06f2bc39dcfe34583c7e 179984Author: Eric Anholt <anholt@freebsd.org> 179985Date: Fri Dec 30 04:11:42 2005 +0000 179986 179987 Add an empty all-local target for FreeBSD make's sake, which doesn't deal 179988 with .PHONY. 179989 179990commit 49a9249239b0dd105b83a101db6e32549978f0d5 179991Author: Adam Jackson <ajax@nwnk.net> 179992Date: Thu Dec 29 21:11:41 2005 +0000 179993 179994 Make kdInputMachine static const, shrinks .data a bit. 179995 179996commit 6d7ee4167d9daeef9b793789a70aa724c4fe6bf4 179997Author: Adam Jackson <ajax@nwnk.net> 179998Date: Thu Dec 29 20:54:08 2005 +0000 179999 180000 Style fix, make SCREEN_EPILOGUE two arguments instead of three. 180001 180002commit b1efb3810cfea8116d76bae3ff3acfee521f4793 180003Author: Adam Jackson <ajax@nwnk.net> 180004Date: Thu Dec 29 20:42:54 2005 +0000 180005 180006 Missed one 180007 180008commit 858b4bc14a7adc7314ce5043c7ff8ca1891dd0c9 180009Author: Adam Jackson <ajax@nwnk.net> 180010Date: Thu Dec 29 20:38:29 2005 +0000 180011 180012 missed changelog entry: 180013 Enough build fixes to get {sdl,ephyr,fake} to link. 180014 180015commit 54922aeadb29df0a1819afd3f616131aa56e140a 180016Author: Adam Jackson <ajax@nwnk.net> 180017Date: Thu Dec 29 20:29:26 2005 +0000 180018 180019 Disable building LBX by default. 180020 180021commit fc69a2e729532ee062af6676fb187a89f7fffe8a 180022Author: Dave Airlie <airlied@linux.ie> 180023Date: Thu Dec 29 08:42:49 2005 +0000 180024 180025 Remove PowerMAX_OS define it never worked , it'll never work again, 180026 finishes running joke of OLS2005 180027 180028commit 25d3852b833bc0b61ce8313ce116251a2602b827 180029Author: Eric Anholt <anholt@freebsd.org> 180030Date: Thu Dec 29 00:44:20 2005 +0000 180031 180032 Undo spamming of libv's ChangeLog entry. 180033 180034commit d6646307f063b938c44d6193a2e8872e178aa90f 180035Author: Eric Anholt <anholt@freebsd.org> 180036Date: Thu Dec 29 00:19:33 2005 +0000 180037 180038 Add necessary *CONFIG_H declarations and a force-off of XF86* extensions in 180039 miinitext in the XGL case. Prevents mismatched structure sizes on my 180040 _XSERVER64 machine. At this point, with the uncommitted render/ diffs, 180041 Xglx starts up but displays badly. 180042 180043commit 3664c3ebf348d07ae3fe301fb8720adf32cf6d64 180044Author: Luc Verhaegen <libv@skynet.be> 180045Date: Wed Dec 28 15:22:21 2005 +0000 180046 180047 Bug #5153: standalone CVT modeline generator. 180048 - add hw/xfree86/utils/cvt/, cvt.c, cvt.man.pre and Makefile.am. 180049 - Adjust configure.ac and hw/xfree86/utils/Makefile.am for cvt. 180050 - Add MonPtr->reducedblanking and Option "ReducedBlanking" to the Monitor 180051 section. 180052 - Check for reduced blanking in xf86CheckModeForMonitor and disallow modes 180053 with less than 25% blanking otherwise. 180054 - Fix some warnings in hw/xfree86/common/xf86Config.c. 180055 180056commit 08708f7d616a7c0a596fb71dd7acd76d45257fec 180057Author: Eric Anholt <anholt@freebsd.org> 180058Date: Wed Dec 28 11:48:14 2005 +0000 180059 180060 Merge from xserver: Make the existing ChangePictureFilter and 180061 ChangePictureTransform screen hooks get called at appropriate times. 180062 180063commit 31d0fdde19598ce9375cc9638ad4e2c8b5af8d9a 180064Author: Adam Jackson <ajax@nwnk.net> 180065Date: Wed Dec 28 10:46:56 2005 +0000 180066 180067 Remove a debugging printf 180068 180069commit 64ac7401ad5022462279dff4dcfb12844c9857ae 180070Author: Eric Anholt <anholt@freebsd.org> 180071Date: Wed Dec 28 10:43:02 2005 +0000 180072 180073 Fix a copy'n'paste-o that would result in mis-rounding of the results of 180074 several composite operators in A and G channels. 180075 180076commit 9ceffb6b92e55f0d74d71489900d43940f14dfd9 180077Author: Eric Anholt <anholt@freebsd.org> 180078Date: Wed Dec 28 10:37:17 2005 +0000 180079 180080 Merge from xserver a header for shm's server internal functions, which are 180081 called by DDXes. Desired by XGL, and should be used in other locations, 180082 probably. 180083 180084commit b1b40ed6a87f72d07825624730cca193d5a89baf 180085Author: Eric Anholt <anholt@freebsd.org> 180086Date: Wed Dec 28 10:31:46 2005 +0000 180087 180088 Initial commit of XGL build infrastructure and XGL code changes for 180089 building within the xorg server tree. Requires additional, uncommitted 180090 dix changes to successfully build, and successful running is still yet 180091 to happen. 180092 180093commit 36061c75ae42aa733cde9b3fd05e0c8280b31655 180094Author: Eric Anholt <anholt@freebsd.org> 180095Date: Wed Dec 28 10:10:59 2005 +0000 180096 180097 Forced commit to note repocopy from xserver CVS as of a few minutes ago. 180098 180099commit 05c139d4cdfd11d39c0168d0c80ac1dbdd069b4c 180100Author: Adam Jackson <ajax@nwnk.net> 180101Date: Wed Dec 28 10:02:54 2005 +0000 180102 180103 Enough build fixes to get {sdl,ephyr,fake} to link. 180104 180105commit aeb770f645e2d591b255ec4ab06addcb1beafa5f 180106Author: Dave Airlie <airlied@linux.ie> 180107Date: Wed Dec 28 02:43:50 2005 +0000 180108 180109 recommit previous changes to evdev.c 180110 180111commit ce7c0c89375ec74f89ae5727998fd75fb768d280 180112Author: Dave Airlie <airlied@linux.ie> 180113Date: Wed Dec 28 01:57:11 2005 +0000 180114 180115 some updates for default colormap install 180116 180117commit 022aa1127c7dcd133ce73dbc12a10bfba8b1ed6e 180118Author: Adam Jackson <ajax@nwnk.net> 180119Date: Wed Dec 28 01:01:06 2005 +0000 180120 180121 Get Xsdl closer to linking. 180122 180123commit 27d79ab2bcebb634d0b69c851c72283a7514eb0c 180124Author: Adam Jackson <ajax@nwnk.net> 180125Date: Tue Dec 27 23:03:15 2005 +0000 180126 180127 s/XSERVER/KDRIVE/ 180128 180129commit 72817714a0787536ce8e8ad0d5473dea0f1c1abe 180130Author: Adam Jackson <ajax@nwnk.net> 180131Date: Tue Dec 27 23:01:27 2005 +0000 180132 180133 Remove Imakefiles. 180134 180135commit 9dd0af6cb4e2c8976ada57a4f4ed16faae090a9d 180136Author: Adam Jackson <ajax@nwnk.net> 180137Date: Tue Dec 27 08:31:37 2005 +0000 180138 180139 Skeletal kdrive build system. Totall non-functional atm. 180140 180141commit 7fd73d2953cf9449c15462cf4bf67639db64f997 180142Author: Adam Jackson <ajax@nwnk.net> 180143Date: Tue Dec 27 08:29:50 2005 +0000 180144 180145 Build fixes: XSERVER_LIBS -> KDRIVE_LIBS, config.h -> kdrive-config.h 180146 180147commit 2f3ac6e5fcbd0e954a094fb6b975d7c8816c44b7 180148Author: Adam Jackson <ajax@nwnk.net> 180149Date: Tue Dec 27 08:26:03 2005 +0000 180150 180151 Start importing kdrive. 180152 180153commit 6798fd0170f4225ce4e69148978533fcee9bdc34 180154Author: Adam Jackson <ajax@nwnk.net> 180155Date: Mon Dec 26 19:13:52 2005 +0000 180156 180157 Bug #4190: Add a rule for 'make relink' since automake sucks. 180158 180159commit 8fc4ea8620913776a903ee2b4f22c306d5778623 180160Author: Adam Jackson <ajax@nwnk.net> 180161Date: Mon Dec 26 18:55:09 2005 +0000 180162 180163 Nuke unsupported NDBM routines. Shrink the hash table a bit, over 180164 25% of the buckets were going empty. 180165 180166commit ed33c7c98ad0c542e9e2dd6caa3f84879c21dd61 180167Author: Daniel Stone <daniel@fooishbar.org> 180168Date: Mon Dec 26 04:23:58 2005 +0000 180169 180170 Remove unused -xkbdb and -noloadxkb options. Rename -ar1 and -ar2 to 180171 -ardelay and -arinterval, respectively. Remove XKB banner from help text. 180172 180173commit 7e3cb9a09ac422179be89773f7fb14a462d25434 180174Author: Adam Jackson <ajax@nwnk.net> 180175Date: Sun Dec 25 22:25:15 2005 +0000 180176 180177 Remove unused layer module. 180178 180179commit 9b083369ded2258cbc8ac2058e06ec8a3b171178 180180Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180181Date: Fri Dec 23 20:11:12 2005 +0000 180182 180183 Change list of X server man pages in "See Also" section to list the ones 180184 actually included and remove the ones that are no longer. 180185 180186commit 5fd978b1e7bce9169f87712a4a7c2c36a68ac00a 180187Author: Daniel Stone <daniel@fooishbar.org> 180188Date: Fri Dec 23 07:40:44 2005 +0000 180189 180190 Make LBX configuration default to auto. 180191 180192commit 5230e86b1cc841bfb35806618052aa835b7eb7e7 180193Author: Dave Airlie <airlied@linux.ie> 180194Date: Fri Dec 23 04:13:37 2005 +0000 180195 180196 fix up xglglx.c 180197 180198commit feb735c5bb0cd391136f1c73476703dff82dc9b0 180199Author: Dave Airlie <airlied@linux.ie> 180200Date: Fri Dec 23 02:07:58 2005 +0000 180201 180202 Well there were a couple of snapshots later than CVS available outside of 180203 Novell, so I've done a crazy merge to try and get them into a workable 180204 CVS, I suspect I may have failed.. there is a pre-xgldrop-merge tag if 180205 I did. 180206 180207commit ade104ce5a016623c1ce97b0d52b531185b35baf 180208Author: Dave Airlie <airlied@linux.ie> 180209Date: Fri Dec 23 01:51:40 2005 +0000 180210 180211 check drawable is available 180212 180213commit 7fb521e80d6e2c05e9475e74fbf80bfbe74cda95 180214Author: Dave Airlie <airlied@linux.ie> 180215Date: Fri Dec 23 01:50:04 2005 +0000 180216 180217 from davidr's tree update some fixes 180218 180219commit c88a3145d057ab72466a3ea8b789bf419e4efc33 180220Author: Dave Airlie <airlied@linux.ie> 180221Date: Fri Dec 23 01:49:21 2005 +0000 180222 180223 from davidr's tree if source picture defined return 180224 180225commit c59508566f11982aa3f4be383597d0e6178718c2 180226Author: Dave Airlie <airlied@linux.ie> 180227Date: Fri Dec 23 01:13:28 2005 +0000 180228 180229 fix glx up for newest glitz interface 180230 180231commit d822bc159672e7327054e572b659ae7dde040e83 180232Author: Dave Airlie <airlied@linux.ie> 180233Date: Fri Dec 23 00:08:35 2005 +0000 180234 180235 make xgl code at least build against latest glitz. 180236 180237commit 6e2086395d99081d8d682b90cec650a06e41fc2c 180238Author: Dave Airlie <airlied@linux.ie> 180239Date: Thu Dec 22 23:32:49 2005 +0000 180240 180241 fixups for newer glitz API 180242 180243commit f3ae42c0fd910b7f9feb9be91ccb056bce0cd999 180244Author: Dave Airlie <airlied@linux.ie> 180245Date: Thu Dec 22 23:31:15 2005 +0000 180246 180247 small fix towards new glitz interface 180248 180249commit 2af7e94eab6847159a3439301ecc93c62a12b1a0 180250Author: Eric Anholt <anholt@freebsd.org> 180251Date: Thu Dec 22 13:54:08 2005 +0000 180252 180253 Adjust the rules for auto-generating some source files, so that they work 180254 on both GNU make and FreeBSD's make. 180255 180256commit 0d7ec5c7d9b451066a079fe56bcc9722341a91ff 180257Author: Kevin E Martin <kem@kem.org> 180258Date: Wed Dec 21 02:30:08 2005 +0000 180259 180260 Update package version for X11R7 release. 180261 180262commit b37e738d5f4e1769bdee98acca788aeeb1556bcc 180263Author: Adam Jackson <ajax@nwnk.net> 180264Date: Tue Dec 20 21:40:19 2005 +0000 180265 180266 Fix an fb regression on A8 pictures. (Fredrik Höglund) 180267 180268commit 03d37eb03864cfc1a2f8d239d5a4c8341bf274f7 180269Author: Adam Jackson <ajax@nwnk.net> 180270Date: Tue Dec 20 21:34:21 2005 +0000 180271 180272 Bug #5359: Fix a segfault (Mark Kettenis) 180273 180274commit 7b89b643c12fa0f7a662b3ff76e05ece53101312 180275Author: Adam Jackson <ajax@nwnk.net> 180276Date: Mon Dec 19 16:44:21 2005 +0000 180277 180278 Bug #5116: Refer DRI section details to dri.fd.o. 180279 180280commit 3ef3add90351e3cb7b54dbcedc234bc5d3d65f1c 180281Author: Adam Jackson <ajax@nwnk.net> 180282Date: Mon Dec 19 16:34:07 2005 +0000 180283 180284 Stub COPYING files 180285 180286commit 3566307c8d44f89622ea51169f67c79092cb56d1 180287Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180288Date: Mon Dec 19 09:18:29 2005 +0000 180289 180290 Fix typos. 180291 180292commit 137447c5f3c6f1914ac869297f823ae93ce428ac 180293Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180294Date: Thu Dec 15 01:54:45 2005 +0000 180295 180296 Update to 2005-12-14 snapshot from pciids.sf.net 180297 180298commit 2cf86fce41e3fd2ac48c5088da11e19077e42e65 180299Author: Kevin E Martin <kem@kem.org> 180300Date: Thu Dec 15 00:20:27 2005 +0000 180301 180302 Update package version number for final X11R7 release candidate. Update 180303 release string to 6.99.99.904. 180304 180305commit f1ba3b4f33a928a3a59538799b3863de5c87e70e 180306Author: Adam Jackson <ajax@nwnk.net> 180307Date: Wed Dec 14 20:11:16 2005 +0000 180308 180309 Bug #4718: Command line flag to disable ACPI. 180310 180311commit cf605eb91619a8c0589a08674ffc3e018471b3fc 180312Author: Adam Jackson <ajax@nwnk.net> 180313Date: Tue Dec 13 17:35:26 2005 +0000 180314 180315 Build libglx correctly when not building the Xorg DDX. 180316 180317commit b076dd787ff71c4b385ab4e2e4eb367f3de378f6 180318Author: Adam Jackson <ajax@nwnk.net> 180319Date: Tue Dec 13 17:34:06 2005 +0000 180320 180321 Spell it XINERAMA_SRCS, not PANORAMIX_SRCS. 180322 180323commit 3666dbb5f3e06fa6a72def64556d64cf73141777 180324Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180325Date: Mon Dec 12 23:33:55 2005 +0000 180326 180327 Remove unnecessary include of dgaproc.h that broke Solaris builds. 180328 180329commit 438a5549f08ab03443d45dd46323579a2f2e4ba2 180330Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180331Date: Mon Dec 12 03:06:18 2005 +0000 180332 180333 Bugzilla #4715 <https://bugs.freedesktop.org/show_bug.cgi?id=4715> Files in 180334 xserver/xorg/Xext not included in tarball after make dist 180335 180336commit 62f3ef930adc7edd49b27dd1f7b0f51bc8bc0afa 180337Author: Adam Jackson <ajax@nwnk.net> 180338Date: Fri Dec 9 18:35:21 2005 +0000 180339 180340 Bug #5258: Restore binary compatibility with 6.8.2's PictureRec. (Aaron 180341 Plattner) 180342 180343commit b99dea9dcf99f907a3536c0db1c39cc67931a5b1 180344Author: Adam Jackson <ajax@nwnk.net> 180345Date: Fri Dec 9 18:32:46 2005 +0000 180346 180347 Bug #4935: Fix includes. (Eric Anholt) 180348 180349commit f4957ee94810b471110deebf03d7413399b45db3 180350Author: Adam Jackson <ajax@nwnk.net> 180351Date: Fri Dec 9 18:30:51 2005 +0000 180352 180353 Bug #4809: Re-fix that doesn't break distcheck. (Alan Coopersmith) 180354 180355commit de22d0c2264bd6dbacbbb4160d09c7e84ad37e70 180356Author: Adam Jackson <ajax@nwnk.net> 180357Date: Fri Dec 9 15:30:05 2005 +0000 180358 180359 Fix a thinko so the code matches the comment 180360 180361commit 6fcb049cd0d2291da5943176716d1f7bbb85fdc2 180362Author: Adam Jackson <ajax@nwnk.net> 180363Date: Fri Dec 9 06:49:39 2005 +0000 180364 180365 Bug #1288: Additional refactor of the driver probe logic to keep ati loaded 180366 before atimisc. 180367 180368commit 17ac5e9fec1e07bd18ae1407043c300cb4695ede 180369Author: Adam Jackson <ajax@nwnk.net> 180370Date: Fri Dec 9 05:36:41 2005 +0000 180371 180372 Push the fallback drivers to the end of the list so driver probe order 180373 stays useful. 180374 180375commit 80ea67e37980d07438749f1aa4dfdd7ee1086799 180376Author: Adam Jackson <ajax@nwnk.net> 180377Date: Fri Dec 9 03:59:41 2005 +0000 180378 180379 Bug #4361: Change driver probe logic to read the driver list from disk 180380 instead of using a compile-time array. 180381 180382commit 7fa2d11d85d43f42aa9c02f8d772c91d1b04df43 180383Author: Adam Jackson <ajax@nwnk.net> 180384Date: Fri Dec 9 03:57:41 2005 +0000 180385 180386 Bug #4361: Define XF86CONFIGFILE properly so config file generation works 180387 180388commit 26b41ff43959a07a778bc3d6e4db8da036f09de3 180389Author: Kevin E Martin <kem@kem.org> 180390Date: Fri Dec 9 03:02:21 2005 +0000 180391 180392 Fix sgml docs build. 180393 180394commit d6f98cbdb8fb74c504a92939d3741420eeed7110 180395Author: Adam Jackson <ajax@nwnk.net> 180396Date: Thu Dec 8 19:33:09 2005 +0000 180397 180398 Bug #3944: Fix 24bpp packed pixel. (Søren Sandmann Pedersen) 180399 180400commit f9ccebe8c5cd674c08fe8ed860d1c456e42c937e 180401Author: Adam Jackson <ajax@nwnk.net> 180402Date: Thu Dec 8 19:27:13 2005 +0000 180403 180404 Bug #4928: Fix compilation for Alpha. (Stefaan DeRoeck) 180405 180406commit 3a6bdf0715b994d6ecaa5b6e448695a8a8ec7d72 180407Author: Kevin E Martin <kem@kem.org> 180408Date: Thu Dec 8 19:21:12 2005 +0000 180409 180410 Add configure option to set the top level font dir. 180411 180412commit 008c2dd5e4614e6a21123ee3a2ac9c5d3bafa97a 180413Author: Kevin E Martin <kem@kem.org> 180414Date: Thu Dec 8 17:55:19 2005 +0000 180415 180416 Add configure options to allow hard-coded paths to be changed. 180417 180418commit 39189c2b86a4c2ab5f3f161d423eb072356668e5 180419Author: Kevin E Martin <kem@kem.org> 180420Date: Thu Dec 8 17:54:40 2005 +0000 180421 180422 Allow hard-coded paths to be configurable. 180423 180424commit 20c0ebe7b3feb85abf9bf140b7799aafc6f59513 180425Author: Kevin E Martin <kem@kem.org> 180426Date: Wed Dec 7 16:18:02 2005 +0000 180427 180428 Change to use the app-defaults default dir configured in libXt. 180429 180430commit 4a39354e14c3c360046b04ea0d4825832b05df05 180431Author: Kevin E Martin <kem@kem.org> 180432Date: Tue Dec 6 22:48:51 2005 +0000 180433 180434 Change *man_SOURCES ==> *man_PRE to fix autotools warnings. 180435 180436commit 84faf8dc9747bc4f1db5ebc2f23e17cf1460e2e9 180437Author: Adam Jackson <ajax@nwnk.net> 180438Date: Tue Dec 6 16:22:47 2005 +0000 180439 180440 Bug #5230: Fix whitespace bugs. 180441 180442commit 9439297b7bc07dcb90f0d01da09eea1bac3d42ff 180443Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180444Date: Tue Dec 6 15:50:35 2005 +0000 180445 180446 Bugzilla #5219 <https://bugs.freedesktop.org/show_bug.cgi?id=5219> Make 180447 sure all optional sources are included in EXTRA_DIST, even if they 180448 aren't used on the platform the distballs are made on. 180449 180450commit f259fd680caccb59546d7788704e46e51a9c6146 180451Author: Adam Jackson <ajax@nwnk.net> 180452Date: Sat Dec 3 22:47:47 2005 +0000 180453 180454 Disable the xf8_32wid logic for now, breaks distcheck 180455 180456commit 26f9c4305660c2b3dc7fe8d214bcdd3c24e1b198 180457Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180458Date: Sat Dec 3 17:04:45 2005 +0000 180459 180460 Bugzilla #4809 <https://bugs.freedesktop.org/show_bug.cgi?id=4809> Patch 180461 #3908 <https://bugs.freedesktop.org/attachment.cgi?id=3908> xf8_32wid 180462 and cfb24 only need to be built on sparc 180463 180464commit 98231c6b38c98976f4ac2b9417ecfbc37a8cbe9a 180465Author: Kevin E Martin <kem@kem.org> 180466Date: Sat Dec 3 05:47:25 2005 +0000 180467 180468 Update package version number for X11R7 RC3 release. Update release string 180469 to 6.99.99.903 (i.e., 7.0 RC3). 180470 180471commit 7c00afd0ec94e491f1a9ef32d6543ed51ea3319d 180472Author: Kevin E Martin <kem@kem.org> 180473Date: Fri Dec 2 06:02:45 2005 +0000 180474 180475 Define XFree86Server only where it is required. 180476 180477commit 924518605b613eb66aa569877fa9f131e6f2a2fd 180478Author: Kevin E Martin <kem@kem.org> 180479Date: Thu Dec 1 23:39:00 2005 +0000 180480 180481 Fix GL build when srcdir != builddir (Donnie Berkholz). 180482 180483commit 14b9315379fe8c783013906616d868f93fd51c83 180484Author: Kevin E Martin <kem@kem.org> 180485Date: Thu Dec 1 22:06:49 2005 +0000 180486 180487 Add missing XvExtension and XvMCExtension defines. 180488 180489commit df8fa21d3189e20260328b88cc8a86224a9b1ebf 180490Author: Kevin E Martin <kem@kem.org> 180491Date: Thu Dec 1 16:20:09 2005 +0000 180492 180493 Fix typo: xorg_bus_linuxbsdpci ==> xorg_bus_linuxpci 180494 180495commit ccfaf82367c9d057fd8314ce36b47f0a8eb696b6 180496Author: Eric Anholt <anholt@freebsd.org> 180497Date: Thu Dec 1 05:04:07 2005 +0000 180498 180499 Bug #5160: Fix the modular build to try to use the same logic for choosing 180500 the architecture/os-specific bus support as monolithic. 180501 180502commit 9c0bd9687fe7d20f2f0793332ae0db06f035eb23 180503Author: Adam Jackson <ajax@nwnk.net> 180504Date: Wed Nov 30 22:59:22 2005 +0000 180505 180506 Import libdrm 2.0 180507 180508commit 4ec0b623b6ab5f8a1e5af2cc3d839251acf81ce2 180509Author: Adam Jackson <ajax@nwnk.net> 180510Date: Wed Nov 30 02:36:25 2005 +0000 180511 180512 Bug #5093: Fix fb for non-SSE machines. (Xavier Bachelot) 180513 180514commit ed826d563cba82c516fd41f6a29ee50aa1fe6c6a 180515Author: Adam Jackson <ajax@nwnk.net> 180516Date: Tue Nov 29 23:34:30 2005 +0000 180517 180518 Only build dlloader modules by default. 180519 180520commit da5d66f2ff27b21fe5c39a4abb4f627edd707f1d 180521Author: Kevin E Martin <kem@kem.org> 180522Date: Tue Nov 29 16:39:33 2005 +0000 180523 180524 Fix usage of XFree86LOADER/XFree86Module/IN_MODULE and update loadable 180525 module builds to reflect this change. 180526 180527commit 51a721a6dbb42702347aad3115147e4922fc1a25 180528Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180529Date: Mon Nov 28 22:05:09 2005 +0000 180530 180531 Change *mandir targets to use new *_MAN_DIR variables set by xorg-macros.m4 180532 update to fix bug #5167 (Linux prefers *.1x man pages in man1 subdir) 180533 180534commit 381931b15b15d0a2ec384b0c22864412c44f9c6e 180535Author: Kevin E Martin <kem@kem.org> 180536Date: Wed Nov 23 07:14:46 2005 +0000 180537 180538 Add configure option to install libxf86config.a (disabled by default). 180539 180540commit 594ca0966e8fd5992ebf95170cc42e19c698fec6 180541Author: Eric Anholt <anholt@freebsd.org> 180542Date: Tue Nov 22 02:11:00 2005 +0000 180543 180544 Bug #5118: Use "rm -f" instead of "$(RM)", which isn't always defined. 180545 180546commit 1c8c1179c0789e3e134d31a62dbb88bfdb594b26 180547Author: Felix Kuehling <fxkuehl@gmx.de> 180548Date: Mon Nov 21 04:24:07 2005 +0000 180549 180550 Fix Xprt library dependencies in the case that Xprint is auto-detected by 180551 configure. 180552 180553commit a1f110bda80bb3b8e4f602385ca5ccd96cf3f786 180554Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180555Date: Sun Nov 20 23:01:02 2005 +0000 180556 180557 Make sure XKM_OUTPUT_DIR (used in code) ends in / (so paths don't get hosed 180558 when appending file names) but XKB_COMPILED_DIR (used in Makefiles) 180559 does not so install-sh -d doesn't get confused when the directory 180560 already exists. 180561 180562commit 385730d23944c24dd9af45b27f62c1161abc48b2 180563Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180564Date: Sun Nov 20 04:15:15 2005 +0000 180565 180566 Add xext to list of modules xorgcfg depends on. 180567 180568commit 63aa96c08a8390621b017ea498c88cf88152024b 180569Author: Kevin E Martin <kem@kem.org> 180570Date: Sat Nov 19 07:15:50 2005 +0000 180571 180572 Update pkgconfig files to separate library build-time dependencies from 180573 application build-time dependencies, and update package deps to work 180574 with separate build roots. 180575 180576commit d3b6653a2892e8c929c79fe3ace19ac9d8366fc4 180577Author: Adam Jackson <ajax@nwnk.net> 180578Date: Sat Nov 19 03:53:04 2005 +0000 180579 180580 Bug #4824: Build XTrap support by default, matching monolith. 180581 180582commit 627ac1fe1dbcbc070575da7bee9e686a7dce5262 180583Author: Eric Anholt <anholt@freebsd.org> 180584Date: Fri Nov 18 23:34:04 2005 +0000 180585 180586 Bug #5060: Fix non-Linux DRI on 64 bit post Linux 32/64 changes. 180587 180588commit e3ec048ff2fe0ee0862472e9b147b7ce488ea898 180589Author: Adam Jackson <ajax@nwnk.net> 180590Date: Fri Nov 18 22:43:50 2005 +0000 180591 180592 Bug #4928: Unbreak Makefile.am for Alpha chips. (Stefan DeRoeck) 180593 180594commit de95d8ee197a0bb738037195997d754a20e10254 180595Author: Adam Jackson <ajax@nwnk.net> 180596Date: Fri Nov 18 18:02:24 2005 +0000 180597 180598 Bug #4859: Don't segfault on bad DDC read. (Tony Houghton) 180599 180600commit 21f7d03dbc347f6bf97a40671275ac75df15bd10 180601Author: Adam Jackson <ajax@nwnk.net> 180602Date: Wed Nov 16 07:28:19 2005 +0000 180603 180604 Fix builds when not building the Xorg DDX. 180605 180606commit fb2d9df869af0c96f1488ef7cf364e01a9d28f3f 180607Author: Adam Jackson <ajax@nwnk.net> 180608Date: Tue Nov 15 00:29:23 2005 +0000 180609 180610 Make fb build on darwin/ppc without addition #define hacks 180611 180612commit 16b315affa30e34b9bab81778978484137a5d9bb 180613Author: Kevin E Martin <kem@kem.org> 180614Date: Mon Nov 14 21:04:12 2005 +0000 180615 180616 Use glcontextmodes.[ch] from Mesa. 180617 180618commit 0c110c80e7afbef50bb354cf1df30123ed048250 180619Author: Kevin E Martin <kem@kem.org> 180620Date: Mon Nov 14 20:18:03 2005 +0000 180621 180622 Add GL_CFLAGS so that GLX can find its proto headers when using separate 180623 build dirs. 180624 180625commit fc81c13e4dafb0eb818879454ee7ae3fa3dae6d0 180626Author: Kean Johnson <kean@armory.com> 180627Date: Mon Nov 14 18:49:30 2005 +0000 180628 180629 Dont prevent SCO platforms for using the default ZAxisMapping now that the 180630 OS layer correctly sends z-axis events when the wheel button is used. 180631 180632commit bd9fb533b31c2427d854199fa59dccd357cf874b 180633Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180634Date: Mon Nov 14 00:01:34 2005 +0000 180635 180636 Default xkb-output directory needs trailing slash. 180637 180638commit 267cbffa41fffff69c692911d128462f5bab2a69 180639Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180640Date: Sun Nov 13 20:53:24 2005 +0000 180641 180642 Bug #5019 <https://bugs.freedesktop.org/show_bug.cgi?id=5019> xserver 180643 installs manpages into 'man1' instead of 'man1x' 180644 180645commit 3179d29b8212c197634d81fbeb8dd2e8df995735 180646Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180647Date: Sat Nov 12 18:03:34 2005 +0000 180648 180649 use RGB_DB not RGB_PATH as that's what configure defines (Jürg Billeter 180650 <j@bitron.ch>) 180651 180652commit 0e7e4c7064df64c29b1a0ccd84fba1be7c748f18 180653Author: Kevin E Martin <kem@kem.org> 180654Date: Thu Nov 10 04:59:21 2005 +0000 180655 180656 Fix typo to enable DGA support. 180657 EXTMODULE is required to build DGA support into extmod. 180658 180659commit e4554db8f87c6a39a3087186395972000bd2085c 180660Author: Kean Johnson <kean@armory.com> 180661Date: Thu Nov 10 02:41:20 2005 +0000 180662 180663 Dont pass wheel mouse buttons as real buttons, map them as Z-Axis movement 180664 on SCO and USL. Re-instate the ZAxisMapping default for the mouse 180665 driver. 180666 180667commit 1b26fe6d2092c202141a0371f47ef1cd7c66ec00 180668Author: Kevin E Martin <kem@kem.org> 180669Date: Wed Nov 9 21:28:54 2005 +0000 180670 180671 Update package version number for X11R7 RC2 release. Update release string 180672 to 6.99.99.902 (i.e., 7.0 RC2). 180673 180674commit f886e632b8dab1bfa0de42b9759a8284ecd9b94f 180675Author: Matthias Hopf <mhopf@suse.de> 180676Date: Wed Nov 9 17:05:41 2005 +0000 180677 180678 Bug #4915: ButtonMapping option which allows to define arbitrary button 180679 mappings (including left-handed mouse etc.). Fixed incorrect usage of 180680 non-reversed, but ZAxisMapped buttons for state detection. Nuked unused 180681 part of reverseMap. 180682 180683commit a25871ae52dd5ce094ba8c1b2021dd027d3e71bd 180684Author: Kevin E Martin <kem@kem.org> 180685Date: Wed Nov 9 01:00:46 2005 +0000 180686 180687 DRM 1.0.5 import 180688 180689commit c9709c0a38af46368726857f7261cbeb84e53911 180690Author: Kevin E Martin <kem@kem.org> 180691Date: Tue Nov 8 22:47:57 2005 +0000 180692 180693 Add newly checked in files to Xorg server build. 180694 Fix release date. Enable DGA extension by default. 180695 180696commit f8430a1b8651f4b52d9d3b54694a60d929b48925 180697Author: Kristian Høgsberg <krh@redhat.com> 180698Date: Tue Nov 8 19:04:56 2005 +0000 180699 180700 Bug #2880, add functions for byte and word level access to pci config 180701 space. 180702 Fix broken utf8 again. 180703 180704commit 5390c7ab05d23f64e6d9afaa558be246a6d6e1b4 180705Author: Kean Johnson <kean@armory.com> 180706Date: Tue Nov 8 06:33:30 2005 +0000 180707 180708 See ChangeLog entry 2005-11-07 for details. 180709 180710commit f5814bf3fff5352ed6edef4c58aadf2d4593f094 180711Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180712Date: Tue Nov 8 03:12:43 2005 +0000 180713 180714 Don't reference noXkbExtension when building without XKB. (Bob Terek - Sun 180715 Microsystems) 180716 180717commit e73cdba865f36ebf78c2dc4ff674b4d9bfe85013 180718Author: Kevin E Martin <kem@kem.org> 180719Date: Mon Nov 7 21:03:49 2005 +0000 180720 180721 Fix Xvfb to work properly in depth 15 mode. Fixes XTS5 XCloseDisplay-3 180722 server crash. 180723 180724commit 890ed0e082e048fa8daf48229b40558381bd131d 180725Author: Thomas Hellstrom <unichrome@shipmail.org> 180726Date: Sun Nov 6 16:40:59 2005 +0000 180727 180728 Fix a bug where a system memory pixmap got a wrong address if memcpy() 180729 fallback was used for downloading from screen. 180730 180731commit 70aedcf32a0c924fd073f5b36d20813e8323026b 180732Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180733Date: Sat Nov 5 18:56:50 2005 +0000 180734 180735 Bug #4948: <https://bugs.freedesktop.org/show_bug.cgi?id=4948> Incorrect 180736 URL in log file for Xorg CVS. Also fixed wording of statement to not 180737 refer to monolithic CVS since modular uses the same code, so it was 180738 appearing in modular builds too. 180739 180740commit 89c661d61f1b9c70a08237476fa1f7f42c1783ab 180741Author: Kevin E Martin <kem@kem.org> 180742Date: Fri Nov 4 21:37:32 2005 +0000 180743 180744 Only use fbCopyAreammx if planemask is FB_ALLONES (fixes XTS5 XCopyArea 180745 tests 22 and 23). 180746 180747commit 0b150a05e6fadca7ee8240697d6cbeadea0c53b3 180748Author: Ian Romanick <idr@us.ibm.com> 180749Date: Thu Nov 3 17:12:53 2005 +0000 180750 180751 Whitespace change just to make sure I created the branch correctly. 180752 180753commit 90cf8e339b71c2f8f2d7a362e6e1ca8078d7f4fd 180754Author: Kevin E Martin <kem@kem.org> 180755Date: Thu Nov 3 17:08:06 2005 +0000 180756 180757 Fix vendor string and release version reported by the servers. Enable 180758 security, lbx and xevie extensions to give us parity with monolithic 180759 tree. 180760 180761commit f23defeef285b4a5bb58405589294bd557c9bb01 180762Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180763Date: Thu Nov 3 16:57:01 2005 +0000 180764 180765 Use APP_MAN_SUFFIX for Xserver man page instead of hardcoding section 1 180766 180767commit f5daec674aeb4fe6ccbc95ead8a319bbeb368d9f 180768Author: Kevin E Martin <kem@kem.org> 180769Date: Wed Nov 2 15:56:40 2005 +0000 180770 180771 Add support for enabling/disabling DBE (part of generic enable/disable 180772 extension support in the server). 180773 180774commit a311bfa73afa1af76f81958d23bc8e0c631d6828 180775Author: Kevin E Martin <kem@kem.org> 180776Date: Wed Nov 2 15:53:57 2005 +0000 180777 180778 Fix support for enabling/disabling extensions loaded from modules. 180779 180780commit 462a2407d540eac831c9be4dcee8a16aa1cea6ac 180781Author: Kevin E Martin <kem@kem.org> 180782Date: Tue Nov 1 15:01:51 2005 +0000 180783 180784 Add xorg-server.m4 for driver dependency checking. 180785 Update pkgcheck depedencies to work with separate build roots. 180786 180787commit 56101c9d6ec3585a0a8550da4b83dd399e3bcce6 180788Author: Kevin E Martin <kem@kem.org> 180789Date: Mon Oct 31 05:45:40 2005 +0000 180790 180791 Fix fd leak by closing them in the ACPI code instead of just using 180792 shutdown. 180793 180794commit 7993486e80711bd6f6f5b6c2b1f2ac32bfba735b 180795Author: Thomas Winischhofer <thomas@winischhofer.net> 180796Date: Sun Oct 30 17:38:49 2005 +0000 180797 180798 RandR: Add a driver func to let the driver determine the physical size of a 180799 screen size (display mode). Useful for faked widescreen modes, modes 180800 which are scaled by the driver, etc. This really helps fixing RandR's 180801 sometimes dumb DPI assumptions. 180802 180803commit c818d3a1a5439c54fc687927a99d69712602ed5e 180804Author: Thomas Winischhofer <thomas@winischhofer.net> 180805Date: Sun Oct 30 09:27:06 2005 +0000 180806 180807 Add xf86RandRSetNewVirtualAndDimensions to loader symlist 180808 180809commit fdbb3ea60949a12eb2f4805d16e8acc2348e39c7 180810Author: Thomas Winischhofer <thomas@winischhofer.net> 180811Date: Sat Oct 29 21:31:23 2005 +0000 180812 180813 Add function for drivers to change RandR's idea of the virtual screen size. 180814 (This allows drivers to reserve a larger virtual size at start and 180815 change it later) 180816 180817commit e921eec1c6d6ce32630977bd876c529a7c694459 180818Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180819Date: Sat Oct 29 00:12:33 2005 +0000 180820 180821 Make X -> Xorg symlink at install time. 180822 180823commit f842c229d4c4dbd5c01364f9e99709bedfd32be6 180824Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 180825Date: Fri Oct 28 16:01:17 2005 +0000 180826 180827 build fix on alpha 180828 180829commit 7416fd61a17a70a2c27c4b1d19796955c296dc7a 180830Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180831Date: Thu Oct 27 21:03:27 2005 +0000 180832 180833 Improved stack trace dump code for Solaris - try fork & exec of pstack 180834 first so we can see the names of non-exported symbols that aren't 180835 visible to walkcontext/dladdr1 code. 180836 180837commit b588bdfe2ac3758d7188706078d79fa276a303e3 180838Author: Dave Airlie <airlied@linux.ie> 180839Date: Sat Oct 22 04:38:50 2005 +0000 180840 180841 programs/Xserver/GL/mesa/X/xf86glx.c: Missing initializer in xf86glx.c 180842 spotted while debugging something else. 180843 180844commit 59279da49806b032027bb54410bc2513d21e3d9e 180845Author: Adam Jackson <ajax@nwnk.net> 180846Date: Fri Oct 21 19:06:13 2005 +0000 180847 180848 Bug #1429: Report input device type correctly. (Stéphane VOLTZ) 180849 180850commit 4a8072011895e6f472e429af7503fc07e0561144 180851Author: Adam Jackson <ajax@nwnk.net> 180852Date: Fri Oct 21 18:50:09 2005 +0000 180853 180854 Bug #4730: Byte-swap the pixmap ID correctly. (Neil Campbell) 180855 180856commit 81e913d3106066de73792f59f3e50e2b5458c567 180857Author: Adam Jackson <ajax@nwnk.net> 180858Date: Fri Oct 21 18:23:33 2005 +0000 180859 180860 Bug #4840: Typo, x$xRES -> x$RES. (George Fufutos) 180861 180862commit f5a58178347878e0409b592330a07867bea02bef 180863Author: Ian Romanick <idr@us.ibm.com> 180864Date: Thu Oct 20 23:24:47 2005 +0000 180865 180866 Make sure that the __gl*_size prototypes are seen in all the places that 180867 they need to be seen. 180868 180869commit 279cf9f79da5778b6e14ecc437379d73e3bec5b0 180870Author: Donnie Berkholz <spyderous@gentoo.org> 180871Date: Thu Oct 20 22:41:28 2005 +0000 180872 180873 Bug #4817 <https://bugs.freedesktop.org/show_bug.cgi?id=4817> Restore '=' 180874 to '==' in test for mmx_capable. 180875 180876commit 1f43d218cc24358a0379535ed517c23011633c31 180877Author: Thomas Winischhofer <thomas@winischhofer.net> 180878Date: Thu Oct 20 21:45:40 2005 +0000 180879 180880 EXA: The "optimization" for using a fill operation instead of 1x1 copies 180881 checked the destination drawable's dimensions (!) instead of the 180882 tile's. Really.... 180883 180884commit 15f56b203dbc14ea59885d40fd4bed3da9e8e190 180885Author: Adam Jackson <ajax@nwnk.net> 180886Date: Thu Oct 20 18:52:51 2005 +0000 180887 180888 Move xf86XTrapModule.c to dixmods, guess at a build system. 180889 180890commit da43c778f4a831061ad2c8b8a312b7a54c9cd79e 180891Author: Adam Jackson <ajax@nwnk.net> 180892Date: Wed Oct 19 22:45:54 2005 +0000 180893 180894 Bug #3224: Degrade XKB fallback message to X_WARNING. 180895 180896commit 4ebd26f04b32f1b09e0759f1a83437d0b1c4d646 180897Author: Adam Jackson <ajax@nwnk.net> 180898Date: Wed Oct 19 22:36:22 2005 +0000 180899 180900 Bug #3196: Fix Load foo.so syntax. 180901 180902commit af211a9bc1bcab0aa631558e5d6ce013095f9802 180903Author: Adam Jackson <ajax@nwnk.net> 180904Date: Wed Oct 19 22:30:09 2005 +0000 180905 180906 Fix PCI bus scan on ia64 E8870 chipsets. 180907 180908commit 5744308e2957781449bfe6fee9b465617a88384d 180909Author: Kevin E Martin <kem@kem.org> 180910Date: Tue Oct 18 22:06:54 2005 +0000 180911 180912 Update package version number for RC1 release. Update release string to 180913 6.99.99.901 (i.e., 7.0 RC1). 180914 180915commit 2769c3e72c470b472dae013e256a7ee73c3e53f2 180916Author: Adam Jackson <ajax@nwnk.net> 180917Date: Tue Oct 18 19:43:48 2005 +0000 180918 180919 Fix distcheck by forcing Xorg to be installed before chmod/chown. 180920 180921commit dd0d010e9c34278f968be486a6c5c91e021b6609 180922Author: Adam Jackson <ajax@nwnk.net> 180923Date: Tue Oct 18 19:14:08 2005 +0000 180924 180925 Fix parallel builds my ensuring libdmxconfig builds first. 180926 180927commit 79e6ac79f983b6cbd88a868dfd2235d9cbe75e8b 180928Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180929Date: Tue Oct 18 07:18:21 2005 +0000 180930 180931 Don't use $< in explicit rules since neither BSD nor Solaris make allow 180932 that. 180933 180934commit fb282ef43a1936dcdefa57f16a8363b2adaf983b 180935Author: Aaron Plattner <aplattner@nvidia.com> 180936Date: Tue Oct 18 04:03:01 2005 +0000 180937 180938 Add miext/damage so misym.c can export DamageDamageRegion. 180939 180940commit 959db6028d232dc76396cb658aa48d3b4e605aed 180941Author: Aaron Plattner <aplattner@nvidia.com> 180942Date: Tue Oct 18 04:02:31 2005 +0000 180943 180944 Export DamageDamageRegion. Not bumping the ABI version since we did that 180945 already for this release. 180946 180947commit b61c828b0455ec1d4a7ffb54b5ac9b65764a458b 180948Author: Kevin E Martin <kem@kem.org> 180949Date: Tue Oct 18 02:23:58 2005 +0000 180950 180951 DRM 20051017 import 180952 180953commit d6a40bcd4a745b5d6d1070deb696b21d128ca0fe 180954Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180955Date: Tue Oct 18 00:32:55 2005 +0000 180956 180957 Change default install dir for app-default files from 180958 $(sysconfdir)/X11/app-defaults to $(libdir)/X11/app-defaults to match 180959 the monolith & allow localization 180960 180961commit e7007f7d51c9e1d39118865fefb1716c579a70bd 180962Author: Adam Jackson <ajax@nwnk.net> 180963Date: Mon Oct 17 22:42:03 2005 +0000 180964 180965 More automake 1.7 braindamage: use mkdir -p, not . 180966 180967commit 151ba8b67fd88a721f9f72d3019212b22f5cd3e2 180968Author: Adam Jackson <ajax@nwnk.net> 180969Date: Mon Oct 17 22:25:58 2005 +0000 180970 180971 Work around automake-1.7 braindamage by providing an explicit rule for 180972 XOrgCfg. 180973 180974commit eec3df1503e561aff6656e15c73b25a0bba1b06b 180975Author: Kristian Høgsberg <krh@redhat.com> 180976Date: Mon Oct 17 17:11:12 2005 +0000 180977 180978 Fix whitespace in AS_HELP_STRING uses, convert all help strings to use 180979 AS_HELP_STRING. 180980 180981commit 1859c62607d567aa05334df1662f7249c983f793 180982Author: Kevin E Martin <kem@kem.org> 180983Date: Mon Oct 17 07:18:59 2005 +0000 180984 180985 include/dix-config.h.in Add support for more extensions 180986 Add missing files to EXTRA_DIST 180987 180988commit ccfe9e7e9b49cbbf7c50fbf1a5c33178f27f79eb 180989Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 180990Date: Sun Oct 16 21:57:34 2005 +0000 180991 180992 Link Xprint config directories in $(C_LOCALE) list to C locale dir, not 180993 en_US 180994 180995commit a7d6a4fb321415b8aaad72760ff8a1ca3fd077f9 180996Author: Donnie Berkholz <spyderous@gentoo.org> 180997Date: Sun Oct 16 03:02:53 2005 +0000 180998 180999 Change '==' to portable '='. 181000 181001commit c2e461c7e970830ea430de3e5f352d144e9f0239 181002Author: Kevin E Martin <kem@kem.org> 181003Date: Sat Oct 15 20:44:44 2005 +0000 181004 181005 Fix typo and add new Makefiles to AC_OUTPUT 181006 Fix typos 181007 Add xorg.conf.man to CLEANFILES 181008 Add missing files to EXTRA_DIST 181009 181010commit 34b7b57b3c80507f63a542c6adb4b5c8ed80b642 181011Author: Kristian Høgsberg <krh@redhat.com> 181012Date: Sat Oct 15 19:34:28 2005 +0000 181013 181014 Doh, remember to add this file. 181015 181016commit 744aa34ca5228ea176cc56a7bdd48bbf5f29b0b5 181017Author: Eric Anholt <anholt@freebsd.org> 181018Date: Sat Oct 15 02:19:09 2005 +0000 181019 181020 Add an additional meaning to the "dirty" flag. Now, if !dirty && !area, the 181021 pixmaps's contents are undefined, so we won't need to upload the 181022 undefined contents in MoveIn. Use the ExaCheck* for async ops as well, 181023 so that dirty is always tracked. While the performance impact for my ls 181024 -lR test was not significant (though the avoiding-upload path was being 181025 hit), it's likely to be important for the upcoming Get/PutImage 181026 acceleration from ajax. 181027 181028commit 21e7339c1eead1148eea462bc99cf8faf02c8d39 181029Author: Kristian Høgsberg <krh@redhat.com> 181030Date: Fri Oct 14 22:44:56 2005 +0000 181031 181032 Hook up lbx. 181033 181034commit d62943c040fd3d45079c9918c57f74f993b585d4 181035Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181036Date: Fri Oct 14 22:19:51 2005 +0000 181037 181038 Set default font path to match the default in the monolith so fonts are 181039 actually found. 181040 181041commit 0ee70f53ef9b05052ee079560df107d05a9c5407 181042Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181043Date: Fri Oct 14 22:01:46 2005 +0000 181044 181045 Install Xorg & xorg.conf man pages even when not building docs 181046 181047commit 0676a2874a62a3661a718cdf21e75ffc77197ad9 181048Author: Kristian Høgsberg <krh@redhat.com> 181049Date: Fri Oct 14 20:01:36 2005 +0000 181050 181051 Add sysv and sco os-support subdirs and add simple EXTRA_DIST Makefile.am 181052 in those dirs. Remove unsupported os-support subdirs (bsdi, dgux, hurd, 181053 nto, os2, pmax, qnx4) that have no maintainer and we don't dist. 181054 Add Options. 181055 181056commit 57abb5b171b2fe88252aeb788463e533106d66b9 181057Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181058Date: Fri Oct 14 08:29:16 2005 +0000 181059 181060 remove reference to non-existent agpgart.h 181061 181062commit 7e3e9ed97ba25bb84286f97fe6882a37c9aa7e25 181063Author: Donnie Berkholz <spyderous@gentoo.org> 181064Date: Fri Oct 14 06:10:06 2005 +0000 181065 181066 Add XTRAP_LIB to XPRINT_EXTENSIONS to fix xprint build when xtrap is 181067 enabled. 181068 181069commit 821584fcd3bf83f3aaacd35e54323f71d976db44 181070Author: Donnie Berkholz <spyderous@gentoo.org> 181071Date: Fri Oct 14 05:36:39 2005 +0000 181072 181073 Require glproto >= 1.4.1 if building DRI with GLX. This fixes a build 181074 failure on a number of hyperpipe functions. 181075 181076commit 8df7628a2ad93edf8271f13e0b43c0fa8f766668 181077Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181078Date: Fri Oct 14 00:41:51 2005 +0000 181079 181080 Remove reference to XF86config-4. Add xorgcfg(1) to See Also list. 181081 181082commit b54c8154ca19edce00b9c6379d5daf94268bade1 181083Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181084Date: Fri Oct 14 00:34:49 2005 +0000 181085 181086 Set substitutions needed in xorgconfig man page. 181087 181088commit 1df705e465a103c94ffbb9fe97bdbe6b0aefc746 181089Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181090Date: Thu Oct 13 20:30:38 2005 +0000 181091 181092 AC_SUBST VENDOR_STRING & VENDOR_RELEASE for xorgcfg's app-defaults file 181093 181094commit b349b20d783252d5126451142419aae554f9b776 181095Author: Kristian Høgsberg <krh@redhat.com> 181096Date: Thu Oct 13 18:08:24 2005 +0000 181097 181098 Dist NOTES. 181099 Dist helper shell scripts. 181100 Dist XAA.HOWTO and a few more unused C files. 181101 Dist xorgconf.cpp. 181102 Fix DIST_EXTRA typo. 181103 Clean yacc and lex generated files only during make maintainer-clean as we 181104 don't expect users to have those tools installed. 181105 181106commit 35a767590e481b15ae66dccc2dd91098992b2751 181107Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 181108Date: Thu Oct 13 01:13:58 2005 +0000 181109 181110 Fix stupid mistake in yesterday's allocator commit, would cause exa to 181111 consider a random available memory size 181112 181113commit a16dabd05ee7ec97877f07bd40ed83c01e72fc22 181114Author: Eric Anholt <anholt@freebsd.org> 181115Date: Wed Oct 12 11:15:44 2005 +0000 181116 181117 Remove an RM line that appears unnecessary and was breaking the build at 181118 xf86DefModeSet.c with FreeBSD make, where RM was undefined. While here, 181119 make the build of xf86DefModeSet.c depend on its sources, so it'll 181120 rebuild properly, and make it a normal CLEANFILE rather than a 181121 DISTCLEANFILE, since the intention seems to be to build it at the 181122 user's build time. 181123 181124commit b819c8378fbf29f185332e8435a80eb35991cd1f 181125Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181126Date: Wed Oct 12 08:22:31 2005 +0000 181127 181128 remove unneeded line of code 181129 181130commit e573b272bf2b06fb62d0306ddc966f3230ead967 181131Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 181132Date: Wed Oct 12 07:46:36 2005 +0000 181133 181134 Use proper access size when reading pixel based on bpp of the source pixmap 181135 181136commit 55efb41f6cc064763cbfd3ee2a1239dc46cb109a 181137Author: Eric Anholt <anholt@freebsd.org> 181138Date: Wed Oct 12 07:35:20 2005 +0000 181139 181140 If a window background is a 1x1 pixmap, read the value out and go to 181141 exaFillRegionSolid rather than sending piles and piles of Copies to the 181142 hardware. 181143 181144commit fce11fdf03acc1f3f1dafb79fc8fff0251cf5473 181145Author: Kevin E Martin <kem@kem.org> 181146Date: Wed Oct 12 02:11:06 2005 +0000 181147 181148 Fix typo (DIST_EXTRA -> EXTRA_DIST) 181149 181150commit 12994b9afbc18bfb7209f677abf673415c9ddf15 181151Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 181152Date: Tue Oct 11 23:11:37 2005 +0000 181153 181154 Fix a couple of bugs in the offscreen allocator. One mostly harmless was 181155 causing our search loop for evictable blocks to possibly skip a good 181156 candiate, and another was the allocator would occasionally use 181157 area->offset as if it was the base of the pixmap, while for a pixmap 181158 that is not in available state, it is not. This caused some funny 181159 miscalculation leading to overlapping pixmaps and accesses beyond the 181160 end of the framebuffer. To make things cleared, I renamed save_offset 181161 to base_offset, made sure it's the one used everywhere in the 181162 allocator, and only align "offset" for the client at the end of 181163 exaOffscreenAlloc(). 181164 181165commit 8444a1f3918b0433f89cae31673ab63628b4543d 181166Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181167Date: Tue Oct 11 21:01:04 2005 +0000 181168 181169 missed commit 181170 181171commit 3b683b63eed603ae58a8cddab48eb81f7ba0dbdf 181172Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181173Date: Tue Oct 11 20:12:24 2005 +0000 181174 181175 missed ChangeLog entry for previous commit 181176 181177commit b4450f3242ab408e80bc3d6d5d1cf6765f3e5339 181178Author: Thomas Winischhofer <thomas@winischhofer.net> 181179Date: Tue Oct 11 19:03:02 2005 +0000 181180 181181 Add DGAReInitModes, Part 2 181182 181183commit d91d18e1d6d663244288748ab86a35a6c151a535 181184Author: Thomas Winischhofer <thomas@winischhofer.net> 181185Date: Tue Oct 11 19:02:18 2005 +0000 181186 181187 Add DGAReInitModes in order to allow the driver to change the list of 181188 supported DGA modes. (Part 1) 181189 181190commit c1a2abadfbb862cbaac3e23d0c1317ce5473ebdd 181191Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181192Date: Tue Oct 11 14:50:47 2005 +0000 181193 181194 fix a typo 181195 181196commit 697f64a22ac5a7742a0022605a1074351296d4f8 181197Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181198Date: Tue Oct 11 14:50:03 2005 +0000 181199 181200 check randrp has been initialized 181201 181202commit 2828d92c6ca400b603b6a20a221d9c858732292f 181203Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181204Date: Tue Oct 11 14:45:01 2005 +0000 181205 181206 programs/Xserver/hw/xfree86/common/xf86RandR.c 181207 programs/Xserver/hw/xfree86/loader/xf86sym.c Add a new function 181208 xf86GetRotation to allow third party modules to obtain the current 181209 rotation. 181210 181211commit cad18ec979e38ef80a606f0e4abf2142b9d0d2b1 181212Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181213Date: Mon Oct 10 10:07:47 2005 +0000 181214 181215 don't move x or y depending on the screen size change 181216 181217commit 7f72f94aa4f0655b8aab6c67eef2a5f5ac4b418f 181218Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181219Date: Mon Oct 10 09:31:49 2005 +0000 181220 181221 rework that again 181222 181223commit 7c1d9a31a36552467d194e7d009c17dc526256c2 181224Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181225Date: Mon Oct 10 09:24:28 2005 +0000 181226 181227 a furthur tweak to the randr cursor position fix 181228 181229commit 13f958fbe8420e406f24c01d320f29002ee860b7 181230Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 181231Date: Mon Oct 10 05:58:41 2005 +0000 181232 181233 Add missing {Prepare,Finish}Access() wrappers for the tile pixmap in the 181234 fallback case 181235 181236commit d82aeb55ca3b6abe4cafa7b9c39777a5f67308e5 181237Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181238Date: Sun Oct 9 23:47:52 2005 +0000 181239 181240 Bug #3254 <https://bugs.freedesktop.org/show_bug.cgi?id=3254> Make sure 181241 screensaver & DPMS timeouts don't overflow when multiplied by 181242 MILLI_PER_MIN. (Reported by Zachary J. Slater) 181243 181244commit 29b5f846d261976f466d2c7181d6a75de670066b 181245Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181246Date: Sun Oct 9 17:47:34 2005 +0000 181247 181248 Bug #4715 <https://bugs.freedesktop.org/show_bug.cgi?id=4715> Add 181249 SecurityPolicy to EXTRA_DIST (Bill Crawford) 181250 181251commit 046234b3ebdfe221de9e87d70d287f69a6f59d6e 181252Author: Eric Anholt <anholt@freebsd.org> 181253Date: Sun Oct 9 02:03:22 2005 +0000 181254 181255 Don't try the accelerated glyphs path for component-alpha text (which I 181256 don't expect drivers to be able to accelerate without exa assistance). 181257 Instead, drop back to plain old miGlyphs for a 62.5% +/- 1.5% reduction 181258 in runtime of my ls -lR test (n=5) with component alpha. While a 181259 reasonable approach would seem to be making a better test to see 181260 whether the entire path would be accelerated and force migration 181261 appropriately, my attempt at this made the situation much worse. 181262 181263commit 526d1502df8db6799c9d1155b86ce79cef90872b 181264Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181265Date: Fri Oct 7 21:29:39 2005 +0000 181266 181267 another update to the RandR fix (thanks Aaron) 181268 181269commit 5a71a5667eb5b01e0f65f9310f4af1f6c5711ab7 181270Author: Kristian Høgsberg <krh@redhat.com> 181271Date: Fri Oct 7 19:01:10 2005 +0000 181272 181273 Add Xprint init scrips to EXTRA_DIST. 181274 181275commit 470213753b158225b44a39a872599344acbc7101 181276Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181277Date: Fri Oct 7 18:15:08 2005 +0000 181278 181279 update the last RandR fix 181280 181281commit 348242f35aeb2869ef390241035b5f3266fc0288 181282Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181283Date: Fri Oct 7 15:39:52 2005 +0000 181284 181285 programs/Xserver/hw/xfree86/common/xf86RandR.c Use PointerMoved instead of 181286 SetCursorPosition, as PointerMoved will call AdjustFrame to reposition 181287 the window if necessary and avoid the cursor ending up offscreen. 181288 181289commit 578e18d11b3d61449c1dd7eba04e1748f19c68f3 181290Author: Kevin E Martin <kem@kem.org> 181291Date: Fri Oct 7 14:27:47 2005 +0000 181292 181293 Add darwin to dist 181294 Include missing docs in EXTRA_DIST 181295 Include headers and other files in dist 181296 181297commit 148df64a05d69adaac4b0f3684b846eb1da60219 181298Author: Kevin E Martin <kem@kem.org> 181299Date: Fri Oct 7 04:11:02 2005 +0000 181300 181301 Add README.compiled to dist tarball 181302 181303commit ff258ac2783203ed2a7698894d951391d1aecebc 181304Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 181305Date: Thu Oct 6 23:45:29 2005 +0000 181306 181307 Clients tend to set picture->repeat when not necessary. Most HW cannot 181308 accelerate repeat NPOT thus triggering software fallback (this is the 181309 case with gnome desktop for example). This adds a simple optimisation 181310 to exa that removes "repeat" when it's obviously useless, that is, the 181311 single picture instance covers the entire rectangle beeing used 181312 181313commit e4ed43c3a6c248ba2b82b8bbf29da537a68407e6 181314Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 181315Date: Thu Oct 6 22:25:35 2005 +0000 181316 181317 symlink.sh: 181318 New files linked: 181319 xorgconf.cpp Options 181320 usb.3 usb_hid_usages 181321 lynx_ppc.S 181322 BUSmemcpy.S IODelay.S PortIO.S SlowBcopy.S 181323 sun_inout.s 181324 xaaTEGlyphBlt.S 181325 xkbcomp/compiled/README 181326 New files excluded: 181327 All of lib/GL/apple 181328 xlibi18n/*/*.mapfile 181329 xxserver/xorg/configure.ac, xkb/Makefile.am: 181330 Install README.compiled in the xkb output dir 181331 181332commit 1614a31a9dad9482ae4526c194c2bae1c4993f8f 181333Author: Eric Anholt <anholt@freebsd.org> 181334Date: Thu Oct 6 21:55:41 2005 +0000 181335 181336 Bug #4699: Correct some memory leaks in EXA and damage related to region 181337 handling. 181338 181339commit cd9ff6aec81e04bbfe14364407ccb28df05fc063 181340Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181341Date: Thu Oct 6 20:16:13 2005 +0000 181342 181343 cpp processing for Xvfb man page 181344 181345commit 370b8c8f1cb1a3531d52ea3b430852a0d76b2a4c 181346Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181347Date: Thu Oct 6 20:14:43 2005 +0000 181348 181349 App-defaults file not supposed to have .ad suffix when installed Fix cpp 181350 rules to set needed flags for app-defaults file 181351 181352commit 2770233069d3845c681bea8eccff22e92254487e 181353Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181354Date: Thu Oct 6 19:59:26 2005 +0000 181355 181356 Don't build "ev" example on systems without <linux/input.h> 181357 181358commit 460145a5d52b5325fa5e920cee3699fcf7dd9afe 181359Author: Kristian Høgsberg <krh@redhat.com> 181360Date: Thu Oct 6 19:37:39 2005 +0000 181361 181362 Add cpconfig.c to EXTRA_DIST. 181363 Add CURSOR.NOTES to EXTRA_DIST. 181364 Add extrapci.ids to EXTRA_DIST and fix xf86PciIds.h rule. 181365 181366commit e63f76caa1b1342422567fdcb9f8af24792c8ca1 181367Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181368Date: Thu Oct 6 17:55:54 2005 +0000 181369 181370 Sun bug #6326551: xkbSetDetectableAutoRepeat broken when using XEvIE 181371 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6326551> 181372 (Derek Wang, Sun Microsystems) 181373 181374commit 5f30a7b10286b4f55821acd4eb5580a8f5a3c56a 181375Author: Benjamin Herrenschmidt <benh@kernel.crashing.org> 181376Date: Thu Oct 6 08:08:04 2005 +0000 181377 181378 Bug #4689: Treat DirectColor as TrueColor in Render. It fixes some crashes 181379 with xcompmgr when using apps that use a DirectColor visual for their 181380 windows 181381 181382commit 9000c0321baf1e25e1796e6a333aad0e5a22cbe2 181383Author: Kevin E Martin <kem@kem.org> 181384Date: Thu Oct 6 04:05:30 2005 +0000 181385 181386 Install correct man page and add to dist 181387 181388commit 9b894df44b575f768a2400d044d8c1eb6ef2ec97 181389Author: Kevin E Martin <kem@kem.org> 181390Date: Thu Oct 6 02:40:41 2005 +0000 181391 181392 Include dmx-config.h for modular build 181393 Use <X11/extensions/dmxext.h> intead of "dmxext.h" 181394 181395commit 30c1369bf5816ffd7bd52d9a9dbcb72500684e2f 181396Author: Kevin E Martin <kem@kem.org> 181397Date: Thu Oct 6 02:35:22 2005 +0000 181398 181399 Add support for building DMX config and examples programs Add missing files 181400 to EXTRA_DIST Install Xdmx man page 181401 181402commit 1f9b6dc1ccd999c90ba825cf5fbdfa29770224a6 181403Author: Kevin E Martin <kem@kem.org> 181404Date: Thu Oct 6 00:34:29 2005 +0000 181405 181406 Clean up generated files to pass distcheck 181407 Clean up generated files to pass distcheck Distribute getconfig.man.pre, 181408 not getconfig.man 181409 181410commit 61cd478b545de0313271cf6852e2df770e8f5914 181411Author: Adam Jackson <ajax@nwnk.net> 181412Date: Wed Oct 5 22:39:41 2005 +0000 181413 181414 Bug #3652: Server-side GLX support for GLX_SGIX_swap_barrier and 181415 GLX_SGIX_hyperpipe extensions. (Eric Kunze, SGI) 181416 181417commit e891d9c078bd31447ae3e1fc7f8c15953b0bb916 181418Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181419Date: Wed Oct 5 22:19:09 2005 +0000 181420 181421 Update to 2005-10-05 snapshot from pciids.sf.net (includes a couple new 181422 Radeon id's). 181423 181424commit dc6ac8e46f80157960a24a1be1fb83f22dff45a0 181425Author: Kristian Høgsberg <krh@redhat.com> 181426Date: Wed Oct 5 21:38:40 2005 +0000 181427 181428 Add DGA configure option and add various files that we should be dist'ing. 181429 Simplify xf86DefModeSet.c rule a bit. 181430 181431commit 8391eaa4aa1ae3744ad8c45f5d148ba362d2c9dd 181432Author: Adam Jackson <ajax@nwnk.net> 181433Date: Wed Oct 5 21:13:49 2005 +0000 181434 181435 Preprocess and install XOrgCfg.ad as in the monolith. 181436 181437commit a9df169f108b15d312421e498675cd2e48206660 181438Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181439Date: Wed Oct 5 17:27:58 2005 +0000 181440 181441 Add missing $(DESTDIR) to custom install target 181442 181443commit a6cbe0776fcc8fb19a2bf2ecef41559eed6e5cef 181444Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181445Date: Wed Oct 5 16:39:09 2005 +0000 181446 181447 Fix the rest of the XFree86 DDX options that require an argument to say so 181448 instead of reporting "unrecognized option" when the argument is 181449 missing. Also give correct error instead of "unrecognized option" for 181450 options only available to root. 181451 181452commit 8c524f9966d2a167ea71dd81e235140e0db31471 181453Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181454Date: Wed Oct 5 15:33:40 2005 +0000 181455 181456 Xdmx & Xprint also need xau & xdmcp module dependencies 181457 181458commit 9f3ad65251832631630f7e587b409b750a144bd3 181459Author: Luc Verhaegen <libv@skynet.be> 181460Date: Wed Oct 5 07:27:52 2005 +0000 181461 181462 Fix lnx_pci.c's xf86GetOSOffsetFromPCI return value. Clears up the resource 181463 ranges awkwardness and the "INVALID MEM ALLOCATION" warning. 181464 181465commit da989e988cc96c0ec4f07fceb4c36b30c2e37f4a 181466Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181467Date: Wed Oct 5 02:18:10 2005 +0000 181468 181469 Xnest depends on xdmcp & xau modules too 181470 181471commit 9e8b5f3d478ca18a9ff9c26745de77c91a5d36d9 181472Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181473Date: Wed Oct 5 01:38:50 2005 +0000 181474 181475 Make Xorg -config stop lying to people and claiming it doesn't exist when 181476 you fail to specify a file name. Also, include it in the list of 181477 available options for non-root users when listing all available flags. 181478 181479commit 34a8411ede185553f1387ee0bf534cf77b0fc004 181480Author: Adam Jackson <ajax@nwnk.net> 181481Date: Wed Oct 5 00:55:08 2005 +0000 181482 181483 Bug #4038: Unbreak the SYSVIPC check for cross builds. (Detlef Vollman) 181484 181485commit f47f00ab747563678c8625de5e5b2a588660064e 181486Author: Eric Anholt <anholt@freebsd.org> 181487Date: Tue Oct 4 11:24:09 2005 +0000 181488 181489 Mark the temporary pixmap dirty if UploadToScreen succeeds. Failure to do 181490 so resulted in a solid black glyph if the font rendering actually 181491 resulted in a fallback (subpixel AA, for example) and the temporary got 181492 migrated after 10 or so glyphs. 181493 181494commit 89a1a91b88b94b341075bc208941337ce11465b7 181495Author: Aaron Plattner <aplattner@nvidia.com> 181496Date: Tue Oct 4 07:42:21 2005 +0000 181497 181498 Add miext/cw to the module loader include path so that misym.c can export 181499 miDisableCompositeWrapper. 181500 181501commit 43625a47063c246e7bf9d687caded0b7e2ea0dc6 181502Author: Aaron Plattner <aplattner@nvidia.com> 181503Date: Tue Oct 4 07:31:53 2005 +0000 181504 181505 Bump the video driver module ABI minor version to 8 so modules statically 181506 linked against miDisableCompositeWrapper won't load on older servers. 181507 #include "cw.h" instead of #include "cw/cw.h" 181508 181509commit ca57db6fc1e6100c47ad935d626fdd490ed6116e 181510Author: Aaron Plattner <aplattner@nvidia.com> 181511Date: Tue Oct 4 04:30:33 2005 +0000 181512 181513 Export miDisableCompositeWrapper. 181514 181515commit b2e451b93c20efc49a6cc565239432b2c705fe37 181516Author: Eric Anholt <anholt@freebsd.org> 181517Date: Tue Oct 4 03:44:14 2005 +0000 181518 181519 Correct the test for whether projective transform is necessary. Also, use 181520 "affine" to describe the variable (universally) on suggestion from 181521 vektor. Corrects a rendercheck failure. 181522 181523commit cdded97a0ad717f4f9120b37d2687fa661696c9b 181524Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181525Date: Tue Oct 4 00:45:42 2005 +0000 181526 181527 Add #include <xorg-config.h> for modular build 181528 181529commit 4ae4fc7d51aeb0f27bed52f7e6a346745f3ea453 181530Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181531Date: Tue Oct 4 00:43:16 2005 +0000 181532 181533 Add gtf to xserver/xorg/hw/xfree86/utils 181534 181535commit d51962378ef6371456e034d6d7f6780e05bc1207 181536Author: Adam Jackson <ajax@nwnk.net> 181537Date: Mon Oct 3 19:31:50 2005 +0000 181538 181539 Bug #3781: Only use fbCopyAreammx when alu == GXcopy. Originally Gentoo bug 181540 #96053, patch by bartron@gmx.net. 181541 181542commit 6d4b350dee9495e54e6e5492815885f1d8455ac9 181543Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181544Date: Mon Oct 3 16:46:14 2005 +0000 181545 181546 Bug #3815 <https://bugs.freedesktop.org/show_bug.cgi?id=3815> Patch #3463 181547 <https://bugs.freedesktop.org/attachment.cgi?id=3463> GNU/kFreeBSD 181548 Xserver support (Robert Millan) 181549 181550commit 22b4200b01310e7b4743ef0b3541c3053a2d8279 181551Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181552Date: Mon Oct 3 15:41:10 2005 +0000 181553 181554 Whoops, need to be in DIST_SUBDIRS too. 181555 181556commit 14a2bd33307fd937804a9fbb03787ec30858a05c 181557Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181558Date: Mon Oct 3 15:37:57 2005 +0000 181559 181560 Add missing ] (Dawid Gajownik) 181561 Add getconfig (Dawid Gajownik) 181562 181563commit 84141fc299b03b5552be093f9b698a85bc670d65 181564Author: Eric Anholt <anholt@freebsd.org> 181565Date: Mon Oct 3 11:43:55 2005 +0000 181566 181567 Merge r1.36 of fbcompose.c from xserver CVS: Special case projective 181568 transforms so we can avoid doing the expensive 181569 64-bit math. Unroll the bilinear interpolation loops for an extra boost. I 181570 tested this with the up/downscaling cairo-benchmarks with Xvfb and saw 181571 a 12% +/- 4% decrease in time taken to run them. 181572 181573commit c024262eae4e00567ccb66a59b4d572621233cbc 181574Author: Eric Anholt <anholt@freebsd.org> 181575Date: Mon Oct 3 10:20:29 2005 +0000 181576 181577 Merge r1.2 of fbedge.c from xserver CVS: Optimize spans where the same 181578 value is being added to multiple pixels. This improves the speed of 181579 rendering wide trapezoids. I tested this with a small set of xlibs 181580 cairo-benchmarks with Xvfb and saw a 4% decrease in time taken to run 181581 them. 181582 181583commit a7e3c6fa8ceb6a3a423377aa32ab0da5a6ab9286 181584Author: Adam Jackson <ajax@nwnk.net> 181585Date: Mon Oct 3 06:31:48 2005 +0000 181586 181587 Real configure check for execinfo.h (Yuri Vasilevski) 181588 181589commit e3d2a7d57bc57453d66aa63ca7fe4d910b64737c 181590Author: Adam Jackson <ajax@nwnk.net> 181591Date: Mon Oct 3 06:29:14 2005 +0000 181592 181593 Bug #4393: uClibc lies and defines __GLIBC__ even though it's not source 181594 compatible with glibc, so the backtrace support check fails. Work 181595 around this by wrapping the code in a configure check for execinfo.h, 181596 and emulate detection for the monolith. (Yuri Vasilevski) 181597 181598commit 5037d3441d65f1fb6493c3b55137ef1b5eddd6b0 181599Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181600Date: Sun Oct 2 22:17:38 2005 +0000 181601 181602 Bug #1465 <https://bugs.freedesktop.org/show_bug.cgi?id=1465> 181603 /etc/init.d/Xprint should use PROJECTROOT from build (Grzegorz 181604 D�?browski) 181605 181606commit b05e78dd40e1fe915096362f32c3af8aee0ed36a 181607Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181608Date: Sun Oct 2 19:30:57 2005 +0000 181609 181610 Fix typo in MAN_SRCS (Dawid Gajownik) 181611 181612commit 8814896da83b19be01beedd0b2b3380298778328 181613Author: Eric Anholt <anholt@freebsd.org> 181614Date: Sun Oct 2 08:53:18 2005 +0000 181615 181616 Fix include path for commit of bug #4616. 181617 181618commit ecaa46380ed0a920186407b9294c5c60f75f1a13 181619Author: Eric Anholt <anholt@freebsd.org> 181620Date: Sun Oct 2 08:28:27 2005 +0000 181621 181622 Bugzilla #4616: 181623 - Merge various fb/ bits of COMPOSITE support from xserver, which weren't 181624 necessary before due to cw hiding the issues. Fixes offset calculations 181625 for a number of operations, and may pull some fixes that cairo has 181626 wanted for XAA as well. 181627 - Add a new call, miDisableCompositeWrapper(), which a DDX can call to keep 181628 cw from getting initialized from the damage code. While it would be 181629 cleaner to have each DDX initialize it if it needs it, we don't have 181630 control over all of them (e.g. nvidia). 181631 - Use the miDisableCompositeWrapper() to keep cw from getting set up for 181632 screens using EXA, because EXA is already aware of composite. Avoiding 181633 cw improved performance 0-35% on operations tested by ajax in x11perf. 181634 181635commit 2c82429f8957ed0268c0e4e4fe5aed9093f33960 181636Author: Ian Romanick <idr@us.ibm.com> 181637Date: Sat Oct 1 22:25:13 2005 +0000 181638 181639 Refactors __glXImageSize and __glXImage3DSize into a single function. It 181640 replaces all calls to the old functions with calls to __glXImageSize 181641 with the new parameter list. 181642 I have also added 'target' as a parameter. This is a stepping stone to the 181643 code in patch #2410. Basically, if the texture target is one of 181644 GL_PROXY_*, the image size is always zero. This gathers all the checks 181645 for that into a single place. I have *not* modified the existing 181646 callers to take this into account. They still do their own checks for 181647 GL_PROXY_*. However, when the generated versions of those functions are 181648 added to the tree, they *will* rely on that. 181649 The code growth is mainly due to the new 40 line comment before 181650 __glXImageSize. 181651 I have tested this with a few of the texture using demos and tests from 181652 Mesa, including tunnel, texdown, and drawpix. 181653 Reviewed by: Adam Jackson, Eric Anholt, and Brian Paul. 181654 181655commit e27b3e4ea1ddf9b2e9c2d63a0e60400b523a8a94 181656Author: Ian Romanick <idr@us.ibm.com> 181657Date: Sat Oct 1 22:19:04 2005 +0000 181658 181659 Remove some more incorrect prototypes for the __gl*_size functions. 181660 181661commit e270e6394b623b48d416feeef0c3856f2e303c8d 181662Author: Matthieu Herrb <matthieu.herrb@laas.fr> 181663Date: Sat Oct 1 17:53:38 2005 +0000 181664 181665 Bug #3822: out of bound reads in fbbltone and fbblt (Mark Kettenis, Thierry 181666 Deval). 181667 181668commit 54b2a14f0fa4397f3e9ae75dd63d5cacfdd778eb 181669Author: Matthieu Herrb <matthieu.herrb@laas.fr> 181670Date: Sat Oct 1 17:30:58 2005 +0000 181671 181672 Bug #3411: fix handling of keyboard Autorepeat rate in xorg.conf. 181673 181674commit a07dd03748c8fa2633e294ee4d9ab38265970e5e 181675Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181676Date: Sat Oct 1 07:17:55 2005 +0000 181677 181678 Add hw/xfree86/getconfig 181679 181680commit abc6aa50fb52fa4fa9b9436dbc3a70f86e62dc27 181681Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181682Date: Sat Oct 1 06:27:12 2005 +0000 181683 181684 Oops - fix build/install of fbdevhw.man 181685 181686commit 13e0db19d8c0b1df636f218bcbfbb2c54fa7576f 181687Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181688Date: Sat Oct 1 06:19:02 2005 +0000 181689 181690 Adding more doc files & fbdevhw man page 181691 181692commit b5ce065a5e91e2ad3213ea8c711cfe7ed9060c16 181693Author: Thomas Winischhofer <thomas@winischhofer.net> 181694Date: Fri Sep 30 08:54:44 2005 +0000 181695 181696 RandR: Fix failure handling (Closes #4635; Thomas Winischhofer) 181697 181698commit 4608a2b654be84b2e345bcada63422d18c74a06e 181699Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181700Date: Fri Sep 30 02:37:57 2005 +0000 181701 181702 Man page processing/installation and other doc file updates 181703 181704commit aa74468aa59b95424cd0000179b8985b267d639b 181705Author: Adam Jackson <ajax@nwnk.net> 181706Date: Fri Sep 30 02:03:45 2005 +0000 181707 181708 sparse cleanups. s/0/NULL/ and mark a few things static. 181709 181710commit c65fde5343719d3e9ebc76cc371c6f5f7948de8c 181711Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 181712Date: Wed Sep 28 20:38:42 2005 +0000 181713 181714 Make the server distcheck: 181715 - Fix up the XpConfig directory to remove the stuff it installs 181716 - Add a few files to CLEANFILES here and there 181717 181718commit 58abce3f90504dd48838a2f7ae7bb5db6a6cff70 181719Author: Eric Anholt <anholt@freebsd.org> 181720Date: Wed Sep 28 20:01:37 2005 +0000 181721 181722 - Use the dirty flag (which should be set correctly all the time, 181723 particularly thanks to Prepare/FinishAccess) to avoid DFS/memcpy on 181724 pixmap move-out if it's unnecessary. This was disabled in KAA because 181725 cache misuse on ATI made me guess that this code was wrong. 181726 - Unwrap Glyphs on closescreen. 181727 181728commit f53404bdbba23fd46420564565ab815f7c20b101 181729Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181730Date: Wed Sep 28 16:55:25 2005 +0000 181731 181732 Add kbd_mode build system 181733 181734commit 940158a6f2e98069a47293d713df674e16ad8a11 181735Author: Ian Romanick <idr@us.ibm.com> 181736Date: Wed Sep 28 03:37:22 2005 +0000 181737 181738 Replace all uses of __glEvalComputeK (and the doubly redundant 181739 EvalComputeK) with calls to one of __glMap[12][df]_size. This was 181740 tested with progs/samples/eval (from Mesa). 181741 181742commit 806d74bc0640f4f3dcc034b36a36aea289b01685 181743Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181744Date: Wed Sep 28 01:57:47 2005 +0000 181745 181746 Add __SVR4 #ifdefs to work in non-Imake builds. 181747 181748commit 88957862b812b3e1e19d5e11365a22dc249cf4d2 181749Author: Kevin E Martin <kem@kem.org> 181750Date: Tue Sep 27 23:28:46 2005 +0000 181751 181752 Fix distcheck build and install errors. 181753 181754commit 003655c02ad3a031031bb4ac859966a513f63e10 181755Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 181756Date: Tue Sep 27 18:36:14 2005 +0000 181757 181758 Make XpConfig build system call mkfont{scale,dir} 181759 181760commit 156b2cf3f76ae53cc37b6f5910b446c776ccc9ba 181761Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 181762Date: Tue Sep 27 17:49:35 2005 +0000 181763 181764 Add forgotten Makefile.am 181765 181766commit f3d0cb4a5722e0512bbdcd179215532795cba38f 181767Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 181768Date: Tue Sep 27 16:15:55 2005 +0000 181769 181770 Get XpConfig build system in pretty much working state 181771 181772commit 1c2e8b1ecc8b7b8c8562461eed7892ff22d17e71 181773Author: Kevin E Martin <kem@kem.org> 181774Date: Tue Sep 27 15:11:56 2005 +0000 181775 181776 Fix make dist to include only sgml files. 181777 181778commit ca64aab609c3585234410cd8d908f8e1efe5c788 181779Author: Kevin E Martin <kem@kem.org> 181780Date: Tue Sep 27 14:09:31 2005 +0000 181781 181782 Add build system for building docs and fix setuid issues. 181783 Add build system for sgml docs. 181784 181785commit 1a4e30d508e62ab304722c3525748ff0e3c2899d 181786Author: Kevin E Martin <kem@kem.org> 181787Date: Tue Sep 27 13:47:26 2005 +0000 181788 181789 Include xorg-config.h so the generated file will also include it. 181790 181791commit 30ff9e26196bdba8435e0dcdb96864e81c8cb136 181792Author: Ian Romanick <idr@us.ibm.com> 181793Date: Tue Sep 27 00:04:40 2005 +0000 181794 181795 Fix some problems that caused incorrectly annotated prototypes for 181796 __gl*_size functions to be used. The result was that, on x86, the code 181797 would be compiled with FASTCALL semantics, but the callers would not. 181798 This should fix GLX protocol errors that people are seeing. There 181799 doesn't appear to be a bugzilla associated with this problem. 181800 181801commit 6c5c54b9a2872f1bb7de36a8d2d4efcef70b14c6 181802Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 181803Date: Mon Sep 26 23:07:44 2005 +0000 181804 181805 Various small fixups to get XpConfig to 'build' 181806 181807commit 604f7c64efb57a48ec667c2ed62d3b4bad0c302b 181808Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 181809Date: Mon Sep 26 22:45:43 2005 +0000 181810 181811 Check in skeleton Makefile.am's for the rest of XpConfig 181812 181813commit 8907195d784ff2c72a00b64edab6a8ac3b31dec5 181814Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181815Date: Mon Sep 26 20:58:26 2005 +0000 181816 181817 Take care of more files from monolith's Xserver/hw/xfree86/etc dir: 181818 hw/xfree86/utils/ioport/Makefile.am 181819 Add ioport and pcitweak utils from monolith hw/xfree86/etc dir. 181820 Generate xf86DefModeSet.c from vesamodes & extramodes 181821 Add apSolaris.shar to EXTRA_DIST 181822 181823commit 9abccb5e65628c938c6f01b685ab8fbffae7bc3b 181824Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 181825Date: Mon Sep 26 19:33:06 2005 +0000 181826 181827 Add initial build system for XpConfig 181828 181829commit 0531c4be2f1a30082cfec5e411ab34d17978d66e 181830Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181831Date: Mon Sep 26 02:41:38 2005 +0000 181832 181833 alanc@alf:/export/alanc/X.org/head/cvs-rw/xc [7:40pm - 628] head -14 181834 ChangeLog 181835 include xorg-config.h for modular build 181836 Adjust XF86CONFIG defines for modular build 181837 Fixes for modular build: 181838 - include modular server config headers 181839 - change default XCONFIGFILE to xorg.conf 181840 - define XKB_RULES_DIR if not defined by Imake 181841 181842commit 256fa24945bcaa6e5a68a48c1b757f8084e88a38 181843Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181844Date: Sun Sep 25 17:48:09 2005 +0000 181845 181846 Add SecurityPolicy file for XCSECURITY extension. 181847 Add README to EXTRA_DIST 181848 181849commit 54639964cc344f1086196729fde37515f11e7972 181850Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 181851Date: Sat Sep 24 21:56:00 2005 +0000 181852 181853 Changed the license to a X/MIT one 181854 181855commit 3192f400c72b3b606fcc1798d577737502897b43 181856Author: Adam Jackson <ajax@nwnk.net> 181857Date: Sat Sep 24 18:45:45 2005 +0000 181858 181859 Disable the {Open,Close}FullScreen DRI protocol. Remove empty FullScreen 181860 stubs from drivers, comment the non-empty ones. 181861 181862commit a5477ae7ac9a56c1a586950db1dee6661bff149d 181863Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181864Date: Sat Sep 24 02:40:51 2005 +0000 181865 181866 Add scanpci, xorgcfg, & xorgconfig utilities. 181867 181868commit 2ba865b3f57340fd1d75f7614c17f615cc127b89 181869Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 181870Date: Fri Sep 23 23:04:04 2005 +0000 181871 181872 Fri Sep 23 19:00:06 2005 Søren Sandmann <sandmann@redhat.com> 181873 Apply patch from Ronald Wahl to make sure that the stack pointer is not 181874 modified at points where we access external variables. (Bug 4269). 181875 181876commit 7554e1bf29e5aef8e76f88bac2994ea45a924f2b 181877Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181878Date: Fri Sep 23 00:27:07 2005 +0000 181879 181880 Sun bug #6321613 181881 <http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6321613> 181882 xorgcfg dumps core when it reads a configuration file that has a 181883 Monitor Section, but does not have a "VendorName" entry inside it. 181884 (Henry Zhao, Sun Microsystems) 181885 181886commit 99793543c0fcfd4d699549fcc2bf0ed12aed6a19 181887Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181888Date: Thu Sep 22 12:40:41 2005 +0000 181889 181890 put back some agp related allocation messages 181891 181892commit 02566dff4abbda6888f719727c169b966617a83d 181893Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 181894Date: Thu Sep 22 12:33:36 2005 +0000 181895 181896 fix typo 181897 181898commit 56e7766c775385b0c6e09e6a65a1c8e10dba786e 181899Author: Eric Anholt <anholt@freebsd.org> 181900Date: Wed Sep 21 22:26:07 2005 +0000 181901 181902 Bug #4541: Fix text drawing in the case where a list contains no 181903 non-zero-sized glyphs. Several variables weren't updated, resulting in 181904 rendering simply stopping when this case was hit. (Anders Kaseorg) 181905 181906commit 32497ff479d0c34c05eb5acfcd5f7e186bac3227 181907Author: Kristian Høgsberg <krh@redhat.com> 181908Date: Wed Sep 21 18:14:32 2005 +0000 181909 181910 Accept --with-xkb-output argument to specify output dir for compiled xkb 181911 keymaps. Define XKM_OUTPUT_DIR and XKB_BIN_DIRECTORY. 181912 Add XKM_OUTPUT_DIR and XKB_BIN_DIRECTORY. 181913 Add XKB_CFLAGS. 181914 181915commit 361a9eb953aaa38f8ebc057185de29e50f9eef26 181916Author: Eric Anholt <anholt@freebsd.org> 181917Date: Wed Sep 21 10:27:53 2005 +0000 181918 181919 - Change migration-in rule slightly: previously, if your score was less 181920 than the max, it was bumped, and then if you were above the threshhold 181921 you got moved in. Instead, do the above-threshhold check separate from 181922 score starting out less than max. While this will likely make thrashing 181923 cases worse, I hope it will fix some issues with long term performance 181924 (think of an xcompmgr with a backbuffer it's doing only accelerated 181925 operations to. If some new pixmap comes in and bumps it out, even once, 181926 it will never get a chance to re-migrate because its score will be 181927 maxed). Change migration-out to be the same way for symmetry, though it 181928 shouldn't ever affect anything. 181929 - Fix a lot of debugging output, both in terms of printing quality, and 181930 completeness. The fallback debugging covers a lot more now, pointing 181931 out new areas for improvement. Debugging toggles are now centralized in 181932 exaPriv.h. 181933 181934commit 6a29c4cec156a135d1d9cdb65995f7a84de7cb56 181935Author: Eric Anholt <anholt@freebsd.org> 181936Date: Wed Sep 21 07:43:01 2005 +0000 181937 181938 Add support for hardware accelerating tiled background painting. One 181939 example of this is the root weave, which paints slightly slower on SiS 181940 now in my testing. However, according to keithp some apps use this 181941 feature for a sort of cheap backing store, which this could help with 181942 significantly. While I haven't done much performance testing with it, 181943 it will at least rule out one possible source of terrible performance. 181944 181945commit 1115ff4c008dea6d6cafcb39e4ce6d0b3ed4fcc0 181946Author: Adam Jackson <ajax@nwnk.net> 181947Date: Wed Sep 21 00:38:05 2005 +0000 181948 181949 Bug #4487: Add the bsd subdir to DIST_SUBDIRS. (Joshua Baergen) 181950 181951commit 00bd8968b264c1f69487dd99c212e0ce889b11db 181952Author: Adam Jackson <ajax@nwnk.net> 181953Date: Wed Sep 21 00:24:10 2005 +0000 181954 181955 Bug #4257: Move cfbcmap.c to cfb_common_sources. 181956 181957commit b7e723ed6f23cc8d85f1f1eced9d8ccbc2a6b38c 181958Author: Adam Jackson <ajax@nwnk.net> 181959Date: Tue Sep 20 23:32:21 2005 +0000 181960 181961 Missing backslash 181962 181963commit 15b2f94de5e2cc7e4df8d897d562dffeda31c82c 181964Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181965Date: Tue Sep 20 18:26:24 2005 +0000 181966 181967 Update CVS project tag & notice about pci.ids license in output .h. 181968 Update to match recent changes to pci.ids, extrapci.ids, & pciid2c.pl for 181969 people building on systems without perl. 181970 181971commit 7d0f47c43c5c177cae0f375ceaeef892e4c9663a 181972Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181973Date: Tue Sep 20 15:50:31 2005 +0000 181974 181975 Add --disable-install-setuid option so you can build as non-root and 181976 chown/chmod later. Also disable by default on platforms that don't need 181977 setuid root X servers (Darwin & SPARC's, according to 181978 InstallXserverSetUID settings in the old Imake config files). 181979 181980commit b623c60745ca486667657fa39ad517d1f7a72794 181981Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181982Date: Mon Sep 19 22:44:49 2005 +0000 181983 181984 Linux/Alpha support (Stefaan De Roeck) 181985 181986commit bd906c6bb803b286c39eca80e437a21c14e808a3 181987Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181988Date: Mon Sep 19 18:54:05 2005 +0000 181989 181990 Update to latest version from pciids.sf.net 181991 Sync with updates to pci.ids & PCI id list in nv driver 181992 181993commit 535cd292c809314efe4730a27b0271adfa842775 181994Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 181995Date: Mon Sep 19 18:38:26 2005 +0000 181996 181997 #include "xf86Axp.h" -> "shared/xf86Axp.h" for modular build. (Stefaan De 181998 Roeck) 181999 182000commit a1bcf25a1f819768672ac837cb0a3d0c66937656 182001Author: Eric Anholt <anholt@freebsd.org> 182002Date: Sun Sep 18 02:32:23 2005 +0000 182003 182004 Break EXA ABI while we still can. Add coordinates to the UploadToScreen 182005 hook so we can upload a subset of a pixmap, and convert the current 182006 drivers to respect that. Use this support to directly UploadToScreen in 182007 exaGlyphs, providing a 47.4% +/-2.4% decrease in wall time for ls -lR 182008 programs/Xserver in an antialiased gnome-terminal on an M6 (n=3, caches 182009 hot). I would have bumped major version, only I can't tell what the 182010 EXA_VERSION_* is supposed to be doing as opposed to the module version. 182011 182012commit 074913c8a9c1f66c8752387db2c81ad669e91878 182013Author: Eric Anholt <anholt@freebsd.org> 182014Date: Sat Sep 17 20:02:02 2005 +0000 182015 182016 - Don't try to upload 0 byte-per-pixel (PICT_a1) data using 182017 RADEONHostDataBlit. 182018 - Disable the shortcut for switching from 3d to 3d in radeon_exa.c. It 182019 appears that we do need the cache flush here, thought it's not clear 182020 why. Disable the 2d to 2d shortcut while here, since I'm unsure of what 182021 we're doing. Exposed by the following bit: 182022 - Bug #4485: Add a new routine, exaGlyphs, to handle font drawing. Glyphs 182023 were being accumulated in from non-migratable scratch pixmaps, causing 182024 the destination pixmap to move towards screen but the migration 182025 necessary for source never to happen, leading to abysmal performance. 182026 Instead, copy the scratch glyph data into a real pixmap first, then 182027 composite from that into the destination, allowing for migration. time 182028 ls -lR from programs/Xserver showed 26.9% (+/- 6.3%) decrease in wall 182029 time (n=3). 182030 - Create exaDrawableUse* wrapping exaPixmapUse*, but which are aware of 182031 windows needing backing store. Makes migration code prettier, and 182032 ensures that composited windows will be migrated as normal when we turn 182033 off cw for EXA. (issue brought up by keithp) 182034 182035commit be2bdab4339e493bb0ac3d0e36508b7aa1cd6e92 182036Author: Eric Anholt <anholt@freebsd.org> 182037Date: Thu Sep 15 06:46:05 2005 +0000 182038 182039 Bug #3990: Require glproto in the cases that it's necessary (DRI or DMX 182040 with GLX). (spyderous) 182041 182042commit ef7eef0f68af6bfbb4ee56115ac2f3c1b7425755 182043Author: Eric Anholt <anholt@freebsd.org> 182044Date: Thu Sep 15 05:55:21 2005 +0000 182045 182046 Bug #4046: Install the X.Org server setuid root. (David Schleef) 182047 182048commit 88d7eb1f7ed6afc85c07797838714f2601356018 182049Author: Eric Anholt <anholt@freebsd.org> 182050Date: Thu Sep 15 05:06:54 2005 +0000 182051 182052 Bug #3889: Create the log dir on install, and allow configuring 182053 specifically that directory. Failure to have the directory keeps the 182054 xorg server from starting. 182055 182056commit 25a0ecdc1868f4b53225b92e8ccd222814e2da2a 182057Author: Eric Anholt <anholt@freebsd.org> 182058Date: Thu Sep 15 04:07:19 2005 +0000 182059 182060 Don't put a version number on the module filename. 182061 182062commit 0888ad3874414e32535fecdb717ee7ab86f9d4cc 182063Author: Eric Anholt <anholt@freebsd.org> 182064Date: Wed Sep 14 07:49:22 2005 +0000 182065 182066 Build xf8_32bpp, which is wanted by the mga driver, and export elf.h, which 182067 is wanted by the rendition driver. 182068 182069commit dfb5da93bc16b2fd2b00e939dbe101a04bdeab4f 182070Author: Eric Anholt <anholt@freebsd.org> 182071Date: Tue Sep 13 20:59:05 2005 +0000 182072 182073 Turn on DRI by default on Linux, NetBSD, FreeBSD, and fix the build of the 182074 module for FreeBSD by setting some missing defines (XFree86Module, 182075 IN_MODULE) in the dri directory. Note that those missing defines should 182076 be somewhere generic, since there are other consumers of them, but I 182077 haven't figured out where. 182078 182079commit 12fbcfefe672ec226bdbc7597bf2fd2cb0ee1e19 182080Author: Eric Anholt <anholt@freebsd.org> 182081Date: Tue Sep 13 18:37:35 2005 +0000 182082 182083 Add some initial BSD support for the xorg server. Incomplete on NetBSD, 182084 OpenBSD, and non-i386/amd64 FreeBSD for sure. Plus I haven't actually 182085 run it yet. 182086 182087commit 51eb6c83a3b2721703ca50758853bbf9f99fc4d2 182088Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182089Date: Tue Sep 13 15:28:29 2005 +0000 182090 182091 Tue Sep 13 11:27:05 2005 S%Gï¿¿%@ren Sandmann <sandmann@redhat.com> 182092 * programs/Xserver/miext/shadow/shadow.c (shadowDamageBox): Add 182093 missing REGION_UNINIT. 182094 182095commit 133edff1b499b2c794fcc9a1860c1b5526b58869 182096Author: Eric Anholt <anholt@freebsd.org> 182097Date: Tue Sep 13 05:44:47 2005 +0000 182098 182099 Don't define POSIX_SOURCE on any BSDs, since it breaks things. 182100 Define NO_SYS_HEADERS, as the monolithic build does. Helps FreeBSD. 182101 Don't include validate.c in the library, since it seems to be a userland 182102 test program. 182103 182104commit c3d6799cee7ff8411b3a05a7ab7e2a9e80c95059 182105Author: Daniel Stone <daniel@fooishbar.org> 182106Date: Tue Sep 13 01:33:19 2005 +0000 182107 182108 Bug #594: CAN-2005-2495: Fix exploitable integer overflow in pixmap 182109 creation, where we could create a far smaller pixmap than we thought, 182110 allowing changes to arbitrary chunks of memory. (Søren Sandmann 182111 Pedersen) 182112 182113commit b290884719e18646326f0c2412c2494a07fe3cfd 182114Author: Adam Jackson <ajax@nwnk.net> 182115Date: Tue Sep 13 00:44:52 2005 +0000 182116 182117 Bug #3284: Make the DRI lock reference count per-screen. 182118 182119commit 846f797f37c1ae57a8dad981113b1139359d8c94 182120Author: Daniel Stone <daniel@fooishbar.org> 182121Date: Mon Sep 12 08:24:48 2005 +0000 182122 182123 Add missing fontsproto and fontenc to REQUIRED_MODULES. 182124 182125commit 5b218617fa8ba52bf65aef35da39e06c662495e6 182126Author: Daniel Stone <daniel@fooishbar.org> 182127Date: Mon Sep 12 07:53:16 2005 +0000 182128 182129 Add missing resourceproto to REQUIRED_MODULES. 182130 182131commit ade158d238475ce923fbc5c49275c78cf3864223 182132Author: Daniel Stone <daniel@fooishbar.org> 182133Date: Mon Sep 12 07:07:59 2005 +0000 182134 182135 Define WITH_VGAHW, since we always build it; add to Xorg headers. 182136 182137commit 34dc481a16c0c1cbd69a9ec7172331b59b85e9a7 182138Author: Daniel Stone <daniel@fooishbar.org> 182139Date: Mon Sep 12 07:02:46 2005 +0000 182140 182141 Put DMX AC_CONDITIONALS at the top level, since they must always be called. 182142 182143commit b0f6fe1f38b448b5f1a280e86b01353865866c70 182144Author: Eric Anholt <anholt@freebsd.org> 182145Date: Sun Sep 11 21:38:41 2005 +0000 182146 182147 Add some optional (disabled) fallback debugging code to the async code, for 182148 better tracking of when we're hitting software. 182149 182150commit 1c003ccf5d68baaae6fafdc75eff964f2a62fc35 182151Author: Eric Anholt <anholt@freebsd.org> 182152Date: Sun Sep 11 19:08:10 2005 +0000 182153 182154 Add a pair of hooks, PrepareAccess() and FinishAccess(), which get called 182155 around CPU access to the framebuffer. This allows the hardware to set 182156 up swappers to deal with endianness, or to tell EXA to move the pixmap 182157 out to framebuffer if insufficient swappers are available (note: must 182158 not fail on front buffer!). 182159 Submitted by: benh 182160 182161commit ca210830bd361e3d91b6bc741c495b61c424d1d2 182162Author: Adam Jackson <ajax@nwnk.net> 182163Date: Sun Sep 11 18:43:55 2005 +0000 182164 182165 Simplify life for EXA drivers by reducing some {Con,Dis}joint ops. 182166 182167commit 04f81cacb9fd0944879b2c23a99fa3a1ae979b12 182168Author: Matthieu Herrb <matthieu.herrb@laas.fr> 182169Date: Sun Sep 11 18:33:31 2005 +0000 182170 182171 OpenBSD also doesn't like defining POSIX_SOURCE: it limits available 182172 functionnality in headers beyond what's used by Xorg. 182173 182174commit 089b4272cf32fc9429c1a0e666c2ffb34fda0b93 182175Author: Daniel Stone <daniel@fooishbar.org> 182176Date: Sun Sep 11 01:16:14 2005 +0000 182177 182178 use RGB_DB not RGB_PATH as that's what it's looking for 182179 define HAVE_XKB_CONFIG_H so setting xkb-path works (Jürg Billeter) 182180 182181commit 2f9d01c9681d80235a70263e2e087dc6c181cdc8 182182Author: Eric Anholt <anholt@freebsd.org> 182183Date: Thu Sep 8 01:15:47 2005 +0000 182184 182185 Remove some references to mfb/cfb support that is no longer there. 182186 182187commit b5d42012f9be87f3b45a089c596ce6dba8845794 182188Author: Daniel Stone <daniel@fooishbar.org> 182189Date: Wed Sep 7 01:30:23 2005 +0000 182190 182191 Fix typo resulting in failure to swap between RGB/BGR properly. (Stephen P. 182192 Becker) 182193 182194commit c8de8c23fbccb3296747f429a02c0c0682b74bf5 182195Author: Daniel Stone <daniel@fooishbar.org> 182196Date: Mon Sep 5 07:43:51 2005 +0000 182197 182198 Fix tests for maximum number of colours when creating a colourmap, so a 182199 32-bit visual (e.g. ARGB) doesn't overflow an int when attempting to do 182200 1 << 32. (Benjamin Herrenschmidt) 182201 182202commit 5c5c51fa6da03f19831632a092761a1e4bcf653b 182203Author: Daniel Stone <daniel@fooishbar.org> 182204Date: Mon Sep 5 07:40:50 2005 +0000 182205 182206 Initialise private arrays with calloc, rather than standard malloc. 182207 (Benjamin Herrenschmidt) 182208 182209commit 691669c0121494df90c8523f7d17e01ba0b14a57 182210Author: Daniel Stone <daniel@fooishbar.org> 182211Date: Sat Sep 3 07:08:58 2005 +0000 182212 182213 xorg-server.h -> dix-config.h (thinko). 182214 182215commit 4b2f5ba1b5d59fd6bd7f82da5730f72e8df04858 182216Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182217Date: Sat Sep 3 03:27:55 2005 +0000 182218 182219 Use macros from xtrans.m4, issue error if not found. 182220 182221commit 733a6d7a268945d149bcea159253408bedc69b12 182222Author: Jesse Barnes <jbarnes@virtuousgeek.org> 182223Date: Sat Sep 3 02:21:36 2005 +0000 182224 182225 Add EXA driver writer documentation for the benefit of future generations 182226 of X hackers. 182227 182228commit 854010d71dc0f1e9b73cdc764c9d2cf36f1da625 182229Author: Daniel Stone <daniel@fooishbar.org> 182230Date: Fri Sep 2 03:22:01 2005 +0000 182231 182232 Add -include dix-config.h to GL/mesa/X to catch _XSERVER64, et al. 182233 182234commit 287336f3c9e5023acbfba6508b05a68ccca9ddf0 182235Author: Kristian Høgsberg <krh@redhat.com> 182236Date: Thu Sep 1 19:56:14 2005 +0000 182237 182238 Teach xkb how to optionally run xkbcomp from $bindir instead of 182239 $datadir/xkb. 182240 182241commit 94fbdb5c6d2cdc7b70ecdabe22a9de8b9aabeced 182242Author: Daniel Stone <daniel@fooishbar.org> 182243Date: Thu Sep 1 14:56:35 2005 +0000 182244 182245 Make RGB_PATH configurable. 182246 Move I2C modules back to $(moduledir)/multimedia. 182247 182248commit a65c5796133d126b1810749b5206607e7c42d787 182249Author: Kristian Høgsberg <krh@redhat.com> 182250Date: Wed Aug 31 14:58:09 2005 +0000 182251 182252 Include xkb-config.h if we have it so we pick up the paths defined there. 182253 182254commit 9d3b5e89d691b79ea3361e6dc82938c22d70d0e3 182255Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182256Date: Tue Aug 30 22:34:14 2005 +0000 182257 182258 Save keyboard LED state on startup and restore on exit so text console mode 182259 LEDs match text console mode state. Move push of streams module earlier 182260 so it's loaded before we start using kbd ioctls provided by the streams 182261 module. 182262 182263commit fa6fbd018da939fda7cc2b9a9aa2717b9675a178 182264Author: Kristian Høgsberg <krh@redhat.com> 182265Date: Tue Aug 30 22:31:52 2005 +0000 182266 182267 Use the $(moduledir) makefile variable instead of @moduledir@ so it can be 182268 overridden at make install time. 182269 Remove driverdir and inputdir from pkg-config file. The directory layout of 182270 moduledir is fixed and well known by drivers. 182271 182272commit b4f4bf028a88ee709f4536373de2d40b6445006e 182273Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 182274Date: Tue Aug 30 19:51:59 2005 +0000 182275 182276 programs/Xserver/hw/xwin/winmultiwindowshape.c Fix off by one error (Colin 182277 Harrison) 182278 182279commit 0e50af8b7f459aafd1d1d52414629e926167a751 182280Author: Adam Jackson <ajax@nwnk.net> 182281Date: Tue Aug 30 19:35:06 2005 +0000 182282 182283 typo fix 182284 182285commit 0c74799af4f924ba64ebd6052802b73547f55c72 182286Author: Eric Anholt <anholt@freebsd.org> 182287Date: Tue Aug 30 04:41:04 2005 +0000 182288 182289 Remove existing broken maxX/maxY code for composite (results in infinite 182290 loops, doesn't deal with failure, doesn't present the interface to 182291 drivers that I expected) and instead replace it with a simple fallback 182292 to software when coordinate limits could be violated. Act similarly in 182293 other acceleration cases as well. 182294 The solution I want to see (and intend to do soon) is to (when necessary) 182295 create temporary pictures/pixmaps pointing towards the real ones' bits, 182296 with the offsets adjusted, then render from/to those using adjusted 182297 coordinates. 182298 182299commit f20e845b04dee5fc0780811f565180e322b60b73 182300Author: Eric Anholt <anholt@freebsd.org> 182301Date: Tue Aug 30 03:42:07 2005 +0000 182302 182303 More 0 -> NULL for pointers missed in previous commit to this file. 182304 182305commit 7777d325a3d049cc233c004cba288ed5d10539c2 182306Author: Eric Anholt <anholt@freebsd.org> 182307Date: Tue Aug 30 03:05:21 2005 +0000 182308 182309 Apply an xserver patch from cworth: Avoid buffer ovverrun when a 182310 trapezoid's right edge is on a pixel boundary. 182311 182312commit e321f9e7ff7de9aa702e33a22743b55c8bb66953 182313Author: Eric Anholt <anholt@freebsd.org> 182314Date: Tue Aug 30 03:01:38 2005 +0000 182315 182316 Apply the xserver patch from vektor for bug #4208: Use NULL for pointers 182317 instead of 0. 182318 182319commit 20813d3af065f9b719b39d2e7a3382b8fa278a48 182320Author: Adam Jackson <ajax@nwnk.net> 182321Date: Sun Aug 28 19:47:39 2005 +0000 182322 182323 Bug #3974: Fix unaligned memory access on LP64. (Matthieu Herrb) 182324 182325commit 0926cf79c030f29dce32a9dc944734960ec93d19 182326Author: Adam Jackson <ajax@nwnk.net> 182327Date: Fri Aug 26 20:21:57 2005 +0000 182328 182329 Add diagnostic messages for exaDriverInit failure cases. 182330 182331commit 5ffff7cb868a768307ff6faf164210020e6b94a2 182332Author: Adam Jackson <ajax@nwnk.net> 182333Date: Fri Aug 26 20:08:09 2005 +0000 182334 182335 Bug #4160: Fix Altix kernel version check. 182336 182337commit 6b0cdc5dd9e451021c562ac4b6b2101d50187a30 182338Author: Daniel Stone <daniel@fooishbar.org> 182339Date: Fri Aug 26 16:46:41 2005 +0000 182340 182341 Change use of dix-config.h to xorg-config.h. 182342 182343commit 9d1b349b3765fb587b353c78cca9aa083f5d0eee 182344Author: Daniel Stone <daniel@fooishbar.org> 182345Date: Fri Aug 26 16:34:55 2005 +0000 182346 182347 Back out previous change until I figure out something smarter. 182348 182349commit 3075df24e7931901c6f0526e10a89631fd73c4d0 182350Author: Daniel Stone <daniel@fooishbar.org> 182351Date: Fri Aug 26 07:35:55 2005 +0000 182352 182353 Subvert SIGUSR2 to reload all input devices. (Ubuntu #020) 182354 182355commit 89c84575ea905c7598d6b6029c9209abe1cfb074 182356Author: Daniel Stone <daniel@fooishbar.org> 182357Date: Fri Aug 26 07:24:21 2005 +0000 182358 182359 Undo rate/period change after function body, so the damage doesn't leak 182360 into other functions. (Debian #050) 182361 182362commit bb5e934df7f23fb365ed673a12d283ff52af79c0 182363Author: Daniel Stone <daniel@fooishbar.org> 182364Date: Fri Aug 26 07:15:04 2005 +0000 182365 182366 Only open /proc/bus/pci/devices once. (Ubuntu #029) 182367 182368commit c937faadd0a0a5f2598b84286ac1ed8996a512e5 182369Author: Daniel Stone <daniel@fooishbar.org> 182370Date: Fri Aug 26 06:48:24 2005 +0000 182371 182372 Better error message on failure to set iopl. (Debian #021) 182373 182374commit 6d34a2ac8a33bd6c9083106b38fab6062e033e39 182375Author: Daniel Stone <daniel@fooishbar.org> 182376Date: Fri Aug 26 06:35:00 2005 +0000 182377 182378 Add Xv symbol from Xext. 182379 182380commit b8f0d4c3ebee363279f9dc7318de3e3c854ca5ef 182381Author: Daniel Stone <daniel@fooishbar.org> 182382Date: Fri Aug 26 06:29:15 2005 +0000 182383 182384 Spit out an error when there is no valid FB device, instead of just failing 182385 silently. (Debian #070) 182386 182387commit b48a24e7969d99a0116bc780c70d3e1c18b34769 182388Author: Daniel Stone <daniel@fooishbar.org> 182389Date: Fri Aug 26 06:23:41 2005 +0000 182390 182391 Don't assume that all sun4m CPUs support muldiv; the Cypress CPU, which 182392 implements the 4m MMU but only v7 instructions, does not. (Debian 182393 #100). 182394 182395commit fd158d3d5215b0a013f5305a76097b0b8fa14cf6 182396Author: Daniel Stone <daniel@fooishbar.org> 182397Date: Fri Aug 26 05:49:44 2005 +0000 182398 182399 Get prototype for ntohl from SuSv3-compliant location, fixing it for 182400 systems which define ntohl as a macro only, not both macro and 182401 function. (Debian #076) 182402 182403commit 53e489c0e39b89f41213a726fe1b611d7d9a18db 182404Author: Daniel Stone <daniel@fooishbar.org> 182405Date: Fri Aug 26 05:47:36 2005 +0000 182406 182407 Work around ATI expansion ROM problem on IA64 caused by prototype HP 182408 McKinley systems. (Bdale Garbee) 182409 182410commit 40374d1149d6dcf0b4521faae8bdfecc8a3af077 182411Author: Daniel Stone <daniel@fooishbar.org> 182412Date: Fri Aug 26 05:00:07 2005 +0000 182413 182414 Accept 'Enabled' and 'Disabled' for Extensions section (Ubuntu #990) 182415 182416commit ea80b5db257f4c22cf5a152084aef5fe05079db0 182417Author: Eric Anholt <anholt@freebsd.org> 182418Date: Thu Aug 25 22:11:04 2005 +0000 182419 182420 Fix a use-after-free of cursor data by refcounting for the sprite.current 182421 reference. The particular path seen was XFixes' ReplaceCursor() 182422 resulting in the sprite.current being freed, but then it getting 182423 accessed during the ChangeToCursor() that happens as a result of 182424 WindowHasNewCursor(). 182425 182426commit 54cc45b09bc6c860b3de2012b57c4b35ca18ffd2 182427Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182428Date: Thu Aug 25 21:22:41 2005 +0000 182429 182430 Thu Aug 25 17:15:01 2005 Søren Sandmann <sandmann@redhat.com> 182431 Add all the drivers to the module_LTLIBRARIES instead of having separate 182432 variables for them. Pointed out by Kristian Høgsberg. 182433 182434commit ff22adc09763b2bd860e7f780a5d0855cab0ab30 182435Author: Kristian Høgsberg <krh@redhat.com> 182436Date: Thu Aug 25 14:19:43 2005 +0000 182437 182438 Remove sun_inout.s so make dist works (discussed with Alan Coopersmith). 182439 182440commit 344a24b6229f477c892dd855546391bc1e091bd7 182441Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182442Date: Thu Aug 25 04:09:45 2005 +0000 182443 182444 Use system curses library on Solaris for xorgcfg text mode. 182445 182446commit 0711502f1847ed461672b9842218b9afa9d349ed 182447Author: Eric Anholt <anholt@freebsd.org> 182448Date: Wed Aug 24 23:48:11 2005 +0000 182449 182450 Bugzilla #4226: Change the pixmap migration strategy for the CopyNtoN case. 182451 Now, if either source or dest were in framebuffer, try to get both 182452 there, but prefer system memory for both otherwise. Required making 182453 exaasync.c go through the try-acceleration path. This significantly 182454 improves window resizing under composite, because previously the 182455 pattern of creating a new pixmap and copying default contents from the 182456 screen caused a fallback every time due to the new destination pixmap 182457 being in system memory. 182458 182459commit 2261710fe0dffd60433e3362ac12adf4db570fe5 182460Author: Eric Anholt <anholt@freebsd.org> 182461Date: Wed Aug 24 23:38:25 2005 +0000 182462 182463 Fix a bug where NULL could be dereferenced during the pixmap kick-out 182464 process by referencing the correct offscreen area. Also drive-by the 182465 comments related to these for clarity. 182466 182467commit 55c5c6953a3a661758a42b147f9542950a62fc4d 182468Author: Eric Anholt <anholt@freebsd.org> 182469Date: Wed Aug 24 22:43:27 2005 +0000 182470 182471 Bugzilla #4090: Introduce getters for pixmap pitch and offset, to 182472 simplify/clarify it for driver writers who probably don't want to know 182473 what pPixmap->devPrivate.ptr or pPixmap->devKind mean. Converts the sis 182474 driver to use them, and bumps the EXA module minor version. 182475 182476commit 079ad773e09ed0c5baf01de3d4f02a5568da5634 182477Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182478Date: Wed Aug 24 22:37:15 2005 +0000 182479 182480 Fill in xf86DeallocateGARTMemory stub based on lnx_agp.c version. Add 182481 include of xorg-config.h for modular builds 182482 182483commit 8fd250e5e4c2016614b82e2d653b7fbf8a3a5b99 182484Author: Eric Anholt <anholt@freebsd.org> 182485Date: Wed Aug 24 21:51:28 2005 +0000 182486 182487 Bugzilla #4089: Fix crashes in !EXA_OFFSCREEN_PIXMAPS case by not trying to 182488 do migration when the EXA pixmap private is NULL. 182489 182490commit 79dc5f3d5fe5a66f5fa53af9afc30d27d1af0bce 182491Author: Kristian Høgsberg <krh@redhat.com> 182492Date: Wed Aug 24 21:28:40 2005 +0000 182493 182494 Don't export non-standard symbols generated GCCs stack protection feature 182495 (__guard, __stack_smash_handler). 182496 182497commit d2952de6e3d9197529695bb88d8c3af679ad71af 182498Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182499Date: Wed Aug 24 19:41:43 2005 +0000 182500 182501 Wed Aug 24 15:39:07 2005 Søren Sandmann <sandmann@redhat.com> 182502 Add GLX_{INC,LIBS} for xprint. 182503 Add @DIX_FLAGS@ to AM_CFLAGS 182504 182505commit 81e708440ced309adc62ebf43d00becd32338db5 182506Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182507Date: Wed Aug 24 19:35:51 2005 +0000 182508 182509 Wed Aug 24 15:29:50 2005 Søren Sandmann <sandmann@redhat.com> 182510 Conditionally include dix-config.h 182511 Conditionally include dix-config.h 182512 182513commit 9657e0e9def47dba5b0bfa7461874362712a07bb 182514Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182515Date: Wed Aug 24 15:18:06 2005 +0000 182516 182517 Don't try to build dmx/input/lnx-*.c if <linux/keyboard.h> isn't found 182518 Don't try to build dmx/input/usb-*.c if <linux/input.h> isn't found 182519 Replace -rdynamic with $(LD_EXPORT_SYMBOLS_FLAG) for compatibility with 182520 compilers other than gcc 182521 182522commit 825a95a1fab69f84c99ae132888fced22e28be33 182523Author: Daniel Stone <daniel@fooishbar.org> 182524Date: Wed Aug 24 11:18:35 2005 +0000 182525 182526 Remove use of dix-config and xorg-config.h from public headers. 182527 182528commit 1fb4a5a4ea993a7913a7bcc362315d31b2907836 182529Author: Daniel Stone <daniel@fooishbar.org> 182530Date: Wed Aug 24 09:12:50 2005 +0000 182531 182532 Add xcmiscproto and bigreqsproto to REQUIRED_MODULES, since Xext uses them. 182533 182534commit b47535bd661743946851099f226f9e6aa4cc8c90 182535Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 182536Date: Wed Aug 24 08:49:31 2005 +0000 182537 182538 programs/Xserver/GL/windows/glwrap.c Wrap PointParameteriNV & 182539 PointParameterivNV for Windows builds. 182540 182541commit 9f498a37dd3d8456d2a97be9c039b63abc81a5fe 182542Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182543Date: Tue Aug 23 20:58:29 2005 +0000 182544 182545 Display more friendly mouse protocol names. Enable mouse wheel mapping by 182546 default. 182547 182548commit cff4b1c2166aa2e75618c8df09554a602c3a68c5 182549Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182550Date: Tue Aug 23 20:12:26 2005 +0000 182551 182552 Check for getconfig in GETCONFIG_DIR if it's not found in module dir. 182553 182554commit bfd13645867aab831b7a0f2b1757cb80837c07d9 182555Author: Daniel Stone <daniel@fooishbar.org> 182556Date: Tue Aug 23 09:20:49 2005 +0000 182557 182558 Add missing saver, evie, video, trap proto pkg-config checks. (Georgi 182559 Georgiev) 182560 182561commit 27afac2ce6a77bc68669c1af6a61c589aa9ef384 182562Author: Daniel Stone <daniel@fooishbar.org> 182563Date: Tue Aug 23 08:59:30 2005 +0000 182564 182565 Fix test for Xnest presence. (Donnie Berkholz) 182566 182567commit fed61462be281c568df6407f94ea519748f0b720 182568Author: Daniel Stone <daniel@fooishbar.org> 182569Date: Tue Aug 23 08:58:40 2005 +0000 182570 182571 Fix up warning on debugging. 182572 182573commit e848eb289c9251742a88e76017603952394f4262 182574Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182575Date: Tue Aug 23 01:14:35 2005 +0000 182576 182577 Rename app-defaults entries from xf86cfg to xorgcfg to match name passed to 182578 XtAppInitialize. (Henry Zhao, Sun Microsystems) Also, display vendor 182579 version instead of 4.0 for server version. 182580 182581commit 6076fca82528da8d50b0ed6be8da6f811321474d 182582Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182583Date: Tue Aug 23 00:32:27 2005 +0000 182584 182585 User message cleanups/updates for modern configurations. 182586 182587commit b07602014061cb41540f6a7e74e4132e67aa1117 182588Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182589Date: Mon Aug 22 21:47:59 2005 +0000 182590 182591 If MAKE_XKM_OUTPUT_DIR is defined, call trans_mkdir to create directory if 182592 it doesn't already exist. (ported from Solaris Xsun bug #5039004) 182593 When BuildLikeSun is set, define MAKE_XKM_OUTPUT_DIR and set the xkb output 182594 directory to /var/run/xkb. 182595 182596commit 4a19a33db6d04b5835830a665daa679ee2fcafe7 182597Author: Kevin E Martin <kem@kem.org> 182598Date: Mon Aug 22 21:29:18 2005 +0000 182599 182600 bugzilla #2880 (https://bugs.freedesktop.org/show_bug.cgi?id=2880) 182601 attachment #2987 (https://bugs.freedesktop.org/attachment.cgi?id=2987) 182602 Use system method to access PCI config space for inb and inw in 182603 addition to inl (Olivier Baudron and Kevin Martin). 182604 182605commit 5557a40a022b0ede36edd3370a60f5fc3d147796 182606Author: Kevin E Martin <kem@kem.org> 182607Date: Mon Aug 22 19:52:26 2005 +0000 182608 182609 Fix linking errors for xprint. 182610 182611commit 7693f668efd206a6c259166665bc36d3c6335e8d 182612Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 182613Date: Mon Aug 22 12:05:18 2005 +0000 182614 182615 programs/Xserver/hw/xfree86/os-support/linux/lnx_acpi.c 182616 programs/Xserver/hw/xfree86/os-support/linux/lnx_apm.c 182617 programs/Xserver/hw/xfree86/os-support/linux/Imakefile Add basic ACPI Linux 182618 support so that events can be passed to the driver. (Alan Hourihane) 182619 182620commit 02c834f198eab4c4686d8156b88508fe102099c1 182621Author: Daniel Stone <daniel@fooishbar.org> 182622Date: Mon Aug 22 09:15:31 2005 +0000 182623 182624 Include Xv and XvMC headers in the SDK. 182625 182626commit 3c4d605c7e8a9f6d296086a5b03b4f11b90590db 182627Author: Daniel Stone <daniel@fooishbar.org> 182628Date: Mon Aug 22 09:15:20 2005 +0000 182629 182630 Fix linkage for Xnest, Xvfb, Xdmx and Xorg DDXes; include all libs. 182631 182632commit 79be1f6d4d1ab48841d31d5553dd36b1b3632650 182633Author: Daniel Stone <daniel@fooishbar.org> 182634Date: Sun Aug 21 19:29:55 2005 +0000 182635 182636 Fix inclusion order of dix-config.h, so _XSERVER64 gets defined before X.h 182637 or Xdefs.h get included. (Jürg Billeter) 182638 182639commit 8d6e743bc4e6854ee0bb0fa4f197acd6d7683ccd 182640Author: Daniel Stone <daniel@fooishbar.org> 182641Date: Sun Aug 21 19:23:17 2005 +0000 182642 182643 Add _XSERVER64 definition to config headers. 182644 182645commit 588105173840355717d7b2f7f652289a41166c3f 182646Author: Daniel Stone <daniel@fooishbar.org> 182647Date: Sun Aug 21 19:15:11 2005 +0000 182648 182649 Huge cleanup. Group into sections: hardware feature detection, extension 182650 detection/configuration, DDX options. Make building of Xorg DDX fully 182651 optional. Clarify and correct some help texts. Change all comments to 182652 use dnl instead of #. Quote all tests correctly, and guard 182653 pure-variable tests with 'x' (e.g. test $DMX = yes -> test "x$DMX" = 182654 xyes). Since the DDXes seem to have pretty divergent extension support 182655 these days, get rid of EXTENSION_LIBS, DMX_EXTENSIONS and 182656 XPRINT_EXTENSIONS, and go back to building extension lists by hand in 182657 the DDX-specific sections. Use portable POSIX constructs everywhere 182658 (e.g. test foo && test bar, instead of test foo -a bar). 182659 Clean up old cruft. 182660 Set _XSERVER64 on 64-bit architectures, and use x86_64 for host_cpu instead 182661 of amd64 (Jürg Billeter). 182662 182663commit 367f45073953f8f99a2d9dd054f479e1070f856e 182664Author: Daniel Stone <daniel@fooishbar.org> 182665Date: Sun Aug 21 08:43:46 2005 +0000 182666 182667 Change xorg_bus_sbus to xorg_bus_sparc; build sparcPci.c on all SPARCs. 182668 182669commit 71b3fea94e8845f35e47503636ca1fe78d2d48ca 182670Author: Daniel Stone <daniel@fooishbar.org> 182671Date: Sun Aug 21 08:24:52 2005 +0000 182672 182673 Typo fix (SOURCE -> SOURCES). 182674 182675commit db2909ce76b178663de301c09fb97f2936b1997e 182676Author: Daniel Stone <daniel@fooishbar.org> 182677Date: Sun Aug 21 06:56:19 2005 +0000 182678 182679 Add test for SPARC. Build SparcMulDiv.S on all SPARCs. 182680 Make the default font path configurable. 182681 182682commit d96e6666862553d59fc1f9fdd14fb65f36d589af 182683Author: Adam Jackson <ajax@nwnk.net> 182684Date: Sat Aug 20 18:52:07 2005 +0000 182685 182686 Add sparcPci.c to EXTRA_DIST. 182687 182688commit a402c876a465904ac71ebf39af67ea451b2457dc 182689Author: Daniel Stone <daniel@fooishbar.org> 182690Date: Sat Aug 20 18:11:17 2005 +0000 182691 182692 Make DRI/GLcore builds srcdir != builddir safe, and invoke symlink-mesa.sh 182693 in our builddir, not our srcdir. 182694 182695commit 238d45d2f148e1e0af4b1619cc1d5e8cc4bf9661 182696Author: Daniel Stone <daniel@fooishbar.org> 182697Date: Sat Aug 20 18:10:03 2005 +0000 182698 182699 Make Xprint build optional. 182700 182701commit abab3fd628c2d1096e8534192f33c1068a573c12 182702Author: Daniel Stone <daniel@fooishbar.org> 182703Date: Sat Aug 20 18:09:21 2005 +0000 182704 182705 Make Xext linkage srcdir != builddir safe. 182706 182707commit bb1d99ee72cc560e95010ea1008d5e796177ae62 182708Author: Daniel Stone <daniel@fooishbar.org> 182709Date: Sat Aug 20 18:07:59 2005 +0000 182710 182711 Build libxkbstubs.la to stub XKB DDX functions (e.g. VT switches), and 182712 build libxorggxkb.la from within the Xorg DDX to replace the previous 182713 xf86VT.o, et al, hacks. 182714 182715commit 8a32ed46480d78b69f289c90098f5ed4a830851f 182716Author: Daniel Stone <daniel@fooishbar.org> 182717Date: Fri Aug 19 15:48:18 2005 +0000 182718 182719 Really hopefully the last xorg-commit test. 182720 182721commit 2fd951434507d2a2c0266a052bdca6e223d31bfa 182722Author: Daniel Stone <daniel@fooishbar.org> 182723Date: Fri Aug 19 15:45:55 2005 +0000 182724 182725 Another test commit for xorg-commit. 182726 182727commit 39630b301f769118959b20d962404555714a5812 182728Author: Daniel Stone <daniel@fooishbar.org> 182729Date: Fri Aug 19 15:25:19 2005 +0000 182730 182731 Testing xorg-commit, nothing else. 182732 182733commit b13d3382de0027e897532926983b79caaa1eb655 182734Author: Daniel Stone <daniel@fooishbar.org> 182735Date: Fri Aug 19 15:21:54 2005 +0000 182736 182737 Fix dates on Søren's entries. 182738 182739commit 8ec79e05feacd61562b53ebf36a8b30967affc1e 182740Author: Daniel Stone <daniel@fooishbar.org> 182741Date: Fri Aug 19 15:15:51 2005 +0000 182742 182743 Make symlink-mesa.sh call srcdir != objdir safe. Remove requirement for 182744 XF86Rush protocol headers, which we don't even use. 182745 182746commit 057a8709a116feb0fd0004141bbac20d2766f3db 182747Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182748Date: Fri Aug 19 00:13:46 2005 +0000 182749 182750 More updates for Panoramix -> Xinerama rename 182751 182752commit bed3235d222fd6e2207f6c0d551c67d5a53322cd 182753Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182754Date: Thu Aug 18 21:28:09 2005 +0000 182755 182756 Thu Aug 18 17:27:09 2005 Søren Sandmann <sandmann@redhat.com> 182757 Move fbmmx to a convenience library since the mmx flags may be harmful when 182758 applied to non-mmx code. 182759 182760commit a0366ddb8cb1c57b85a5806eb348abc19c7f92d6 182761Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182762Date: Thu Aug 18 17:14:11 2005 +0000 182763 182764 Replace gnu-makeism with portable rules 182765 182766commit ea5c49cb17ac956d6dea6bf563e392e61c39da2b 182767Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182768Date: Thu Aug 18 01:40:33 2005 +0000 182769 182770 //bugs.freedesktop.org/show_bug.cgi?id=2901> Patch #2331 182771 <https://bugs.freedesktop.org/attachment.cgi?id=2331> Lines ending in 182772 <backslash><space> break with cpp from gcc >= 3.3 (Peter Breitenlohner) 182773 182774commit 22694500e0dd8752b4c01e340f587ecba1ff7fb7 182775Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182776Date: Wed Aug 17 23:20:27 2005 +0000 182777 182778 Add xf86Sbus.h to EXTRA_DIST files (bugzilla #4128) 182779 182780commit df938987d6c19fbdfff8d3334bb497f4814ae384 182781Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182782Date: Wed Aug 17 19:03:18 2005 +0000 182783 182784 Add buildsystem for Xprint, and: 182785 Wed Aug 17 14:50:58 2005 Soren Sandmann <sandmann@redhat.com> 182786 Include <X11/extensions/Print{,str}.h> instead of "Print{,str}.h" 182787 Include <X11/fonts/fontstruct.h> isntead of "fonts/fontstruct.h" 182788 Wed Aug 17 14:54:49 2005 Søren Sandmann <sandmann@redhat.com> 182789 Conditionally compile in xprint.c 182790 Add Xprint 182791 182792commit f2f6820c3f01810a4da5a8bb4e43119ef2a3fcb6 182793Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182794Date: Wed Aug 17 17:20:13 2005 +0000 182795 182796 Wed Aug 17 13:13:00 2005 Søren Sandmann <sandmann@redhat.com> 182797 Make asm labels local. Bug 4073, patch from Diego Pettenò. 182798 182799commit ad7f2fc67376f4fbfe81047273de12f7926f0b17 182800Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182801Date: Tue Aug 16 16:21:09 2005 +0000 182802 182803 symlink.sh: Temporarily disable xkbdata, add type1mod.c 182804 Tue Aug 16 12:09:09 2005 Søren Sandmann <sandmann@redhat.com> 182805 Add ft and type1 modules. 182806 182807commit 24597a1ee2c567ec6bc2f2208457f2778937a034 182808Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182809Date: Tue Aug 16 00:40:25 2005 +0000 182810 182811 2005-08-15 Alan Coopersmith <alan.coopersmith@sun.co>m 182812 Convert sunPostKbdEvent to use xf86ProcessAction like xf86PostKbdEvent. 182813 182814commit b86282b764387ef8315fc4045d4e1ca1bb74fee1 182815Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182816Date: Mon Aug 15 23:48:35 2005 +0000 182817 182818 Set ZAxisMapping default to "4 5 6 7". 182819 182820commit a6c8d0d71eef7cb964797f7595be36ee0ac10a1a 182821Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182822Date: Mon Aug 15 19:42:48 2005 +0000 182823 182824 Mon Aug 15 15:41:26 2005 Søren Sandmann <sandmann@redhat.com> 182825 Link this module with libXext.la. 182826 Mon Aug 15 14:56:57 2005 Søren Sandmann <sandmann@redhat.com> 182827 Remove the xkb* files. 182828 Add xkbVT.o and friends as the last things on the link line so they will 182829 override the symbols in libxkb.a. Add a comment on how this situation 182830 might not be considered ideal. 182831 182832commit 46aede552aa43cd59f81980303826b5c3d889c02 182833Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182834Date: Mon Aug 15 18:59:16 2005 +0000 182835 182836 Mon Aug 15 14:56:57 2005 Søren Sandmann <sandmann@redhat.com> 182837 Remove the xkb* files. 182838 Add xkbVT.o and friends as the last things on the link line so they will 182839 override the symbols in libxkb.a. Add a comment on how this situation 182840 might not be considered ideal. 182841 182842commit bcc95c83406a4498227ffd8384bc272fd8cdc49c 182843Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182844Date: Mon Aug 15 18:32:08 2005 +0000 182845 182846 Add AC_SYS_LARGEFILE to match flags used in monolith 182847 182848commit 809906a754a1289b7e88489241a2065aa0bf27a2 182849Author: Dave Airlie <airlied@linux.ie> 182850Date: Mon Aug 15 08:24:45 2005 +0000 182851 182852 make mouse support work - note you need to change the evdev devices in the 182853 source.. hacky... 182854 182855commit fe5abff38e65c1a49886924efdf0242ab8048008 182856Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 182857Date: Mon Aug 15 07:30:05 2005 +0000 182858 182859 Egbert's 64bit fixes for mixed 32/64bit clients 182860 182861commit ebedc8bbb54b9b4e1814bc2758216af2bab93540 182862Author: Eric Anholt <anholt@freebsd.org> 182863Date: Sun Aug 14 19:46:55 2005 +0000 182864 182865 - Fix the exa pixmap offset/pitch alignment to deal with non-POT alignment 182866 requirements. MGA, notably, uses a multiple of 3 in some cases. 182867 - Rename the pixmap offset/pitch alignment fields to more clearly state 182868 their meaning. 182869 182870commit e3509c940fa1fc3988d23f884ca8bffc87d091e5 182871Author: Daniel Stone <daniel@fooishbar.org> 182872Date: Sun Aug 14 16:24:30 2005 +0000 182873 182874 Don't link libXext, as it is linked in to the server also. 182875 182876commit 05071ae0fb847c211b1f20770d3b57fc2172738c 182877Author: Daniel Stone <daniel@fooishbar.org> 182878Date: Sat Aug 13 07:41:33 2005 +0000 182879 182880 Don't link in libx86emu.a as this a) fails badly on non-x86 systems, and b) 182881 is done via x86emu.c including all the source files anyway. 182882 182883commit 1ffe9ceb2f6a7261fb62c90dfea861f9dadd27ce 182884Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182885Date: Sat Aug 13 06:15:35 2005 +0000 182886 182887 Oops, forgot a file in earlier Solaris kbd commit 182888 182889commit c29051f9d108fe49c23d9cf36fd08cc64c87262a 182890Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182891Date: Sat Aug 13 00:11:28 2005 +0000 182892 182893 Add LintTarget() 182894 Bugzilla #1068 <https://bugs.freedesktop.org/show_bug.cgi?id=1068> Port 182895 Solaris keyboard code to work with kbd driver. 182896 Also incorporated "audio bell" feature from Xsun keyboard DDX to play bell 182897 tones via /dev/audio (specified via Option "BellDevice" "/dev/audio" in 182898 keyboard device options). 182899 182900commit 616a65c4cc528278168db1414776f3a867cd463e 182901Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182902Date: Fri Aug 12 18:50:33 2005 +0000 182903 182904 Fri Aug 12 14:49:24 2005 Søren Sandmann <sandmann@redhat.com> 182905 Apply another patch from Billy Biggs to fix precision issues. 182906 182907commit 1eed84f227311730ce1f9ffab190e95de967c7da 182908Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182909Date: Fri Aug 12 18:47:17 2005 +0000 182910 182911 Fri Aug 12 14:45:54 2005 Søren Sandmann <sandmann@redhat.com> 182912 Fix up multiplications based on patch by Billy Biggs. Part of bug 3945. 182913 182914commit 8bfffb96b552a3facb77ff9e81658e80becbf2f4 182915Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182916Date: Fri Aug 12 18:31:07 2005 +0000 182917 182918 Fri Aug 12 14:29:09 2005 Søren Sandmann <sandmann@redhat.com> 182919 Apply patch from Billy Biggs that fixes rounding problems with division. 182920 Part of bug 3945. 182921 182922commit 812ed2e17bfe8e232313cf9ab78000a564cb6b3c 182923Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182924Date: Fri Aug 12 17:43:38 2005 +0000 182925 182926 Fri Aug 12 10:45:01 2005 S%Gï¿¿%@ren Sandmann <sandmann@redhat.com> 182927 Make this function compute the same results as the fbByteMul macro. 182928 182929commit 370b111f4882a95248bcc4727438c95a065c174d 182930Author: Ian Romanick <idr@us.ibm.com> 182931Date: Fri Aug 12 16:30:57 2005 +0000 182932 182933 Use '$(DRMSRCDIR)/shared-core' instead of '$(DRMSRCDIR)/shared' for DRM 182934 includes. This matches the way drivers are built in the Mesa tree and 182935 fixes a build problem in the Savage driver. 182936 Convert uses of __glPointParameterfvARB_size to 182937 __glPointParameterfvEXT_size and uses of __glPointParameteriv_size to 182938 __glPointParameterivNV_size. This eliminates the need to hand-edit 182939 indirect_size.c after it is generated. 182940 182941commit 5f5117729de3a8a4eb6e17dc0979e4b1c1ef9918 182942Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182943Date: Fri Aug 12 01:17:58 2005 +0000 182944 182945 Fix segfault when "kbd" fails to load and "keyboard" driver is not 182946 configured. 182947 182948commit 13e16ee93a328a55494c2933143c66559fe7ba98 182949Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182950Date: Thu Aug 11 15:46:42 2005 +0000 182951 182952 Thu Aug 11 11:43:32 2005 Søren Sandmann <sandmann@redhat.com> 182953 Make sure we don't crash on glyphs with NULL bits. Bug 659. 182954 182955commit 3e471ddf1dbf58ed021d6f31bdaf438872f03ca8 182956Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182957Date: Thu Aug 11 14:36:29 2005 +0000 182958 182959 Add check for whether the platform is MMX capable and add the relevant 182960 flags if so. 182961 182962commit 130fffc0cdbfdc29f33f1ee97c09e744c19e243a 182963Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182964Date: Wed Aug 10 20:22:57 2005 +0000 182965 182966 Wed Aug 10 16:17:38 2005 Søren Sandmann <sandmann@redhat.com> 182967 Add back non-SSE implementations. Define USE_SSE if the CPU is amd64/x86-64 182968 182969commit ef50bba5694ef276a239882fae3502638b4ec784 182970Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182971Date: Wed Aug 10 19:42:36 2005 +0000 182972 182973 Revert previous patch as it causes build failures 182974 182975commit b99360e264c9531593ce8eb67bd006275ca5e5a0 182976Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 182977Date: Wed Aug 10 19:15:44 2005 +0000 182978 182979 Add XF86DDXACTIONS to AM_CFLAGS 182980 182981commit 73a335a926e50afde36816ab47dae689202df319 182982Author: Jon Smirl <jonsmirl@yahoo.com> 182983Date: Tue Aug 9 16:51:05 2005 +0000 182984 182985 Point xegl at the DRI driver 182986 182987commit 29d0ba9f06d90a1c7f619db87d681ca53fa1bf38 182988Author: Jon Smirl <jonsmirl@yahoo.com> 182989Date: Tue Aug 9 15:58:20 2005 +0000 182990 182991 Touchup xgl for modular tree 182992 182993commit 435e2a09de7c9dd843b05f4a0484371a67940515 182994Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 182995Date: Tue Aug 9 01:18:04 2005 +0000 182996 182997 //bugs.freedesktop.org/show_bug.cgi?id=3950> Xorg prints version numbers in 182998 wrong format for modules compiled for Xorg 6.99 & later (Adam Jackson & 182999 Alan Coopersmith) 183000 183001commit 5849c69db80b2320bddb7fafb519300cc0435c84 183002Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183003Date: Mon Aug 8 18:02:08 2005 +0000 183004 183005 Mon Aug 8 13:39:45 2005 Søren Sandmann <sandmann@redhat.com> 183006 Conditionally include xorg-config.h 183007 183008commit 5e6e5e6d8c6add3aac5c3aebc984d3afb842edc1 183009Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183010Date: Mon Aug 8 17:38:37 2005 +0000 183011 183012 Mon Aug 8 13:36:23 2005 Søren Sandmann <sandmann@redhat.com> 183013 #undef PIXPRIV here. 183014 183015commit a11ce76b0625501a202fa34e18777b7bd42b2713 183016Author: Keith Packard <keithp@keithp.com> 183017Date: Mon Aug 8 06:25:21 2005 +0000 183018 183019 Build with modular X.org libraries and headers. 183020 composite/compwindow.c Don't damage unmoved windows. Let border clip reset 183021 leave damage alone, pending actual damage for painting. 183022 183023commit 129e812c339879b68bb162331ac9b7e0f86a7474 183024Author: Keith Packard <keithp@keithp.com> 183025Date: Mon Aug 8 02:30:31 2005 +0000 183026 183027 Re-add _XOPEN_SOURCE as it's needed *sometimes*. Place it in #ifndef to 183028 avoid whinging. 183029 183030commit 4dfc3357a78a52ef70750608138dfeef0371cbc4 183031Author: Keith Packard <keithp@keithp.com> 183032Date: Mon Aug 8 00:38:41 2005 +0000 183033 183034 Add autogenerated source files and other various non-CVS material to 183035 .cvsignore files 183036 Use XORG_CFLAGS. Ensure that all exa files are in SOURCES 183037 remove _XOPEN_SOURCE as it's always in xtrans.pc these days and gcc whines 183038 libdamage.la needs libcw.la when COMPOSITE is defined, but that 183039 libdamage.la must be after libcomposite.la, so add libcw.la to 183040 DAMAGE_LIB instead of EXTENSION_LIBS. Regularize library link order 183041 across all X servers 183042 183043commit 7d2b08aa4786eb4d181c88f7dc6b7eb059cc9dad 183044Author: Adam Jackson <ajax@nwnk.net> 183045Date: Sun Aug 7 20:42:50 2005 +0000 183046 183047 Invert the sense of the composite configure flag to match reality. 183048 183049commit fcaa6f30c489578589af19ef798ab31150e696a2 183050Author: Adam Jackson <ajax@nwnk.net> 183051Date: Sun Aug 7 19:02:32 2005 +0000 183052 183053 Bug #3989: Fix Composite builds, enable Composite build by default. (Olli 183054 Helenius) 183055 183056commit 3639fb243450ae447e9901de9f909e975a919514 183057Author: Chris Lee <clee@c133.org> 183058Date: Sat Aug 6 23:56:18 2005 +0000 183059 183060 Enabling EXA by default in the modular build. 183061 183062commit a16357ebc9344fbf3e393da9e7b28164ad5b9dc8 183063Author: Chris Lee <clee@c133.org> 183064Date: Sat Aug 6 23:46:38 2005 +0000 183065 183066 Last bits of changes to make exa build properly in the modular server. 183067 183068commit cc67bd187f06bfade0d2fe1b4cca92805458b210 183069Author: Chris Lee <clee@c133.org> 183070Date: Sat Aug 6 23:34:09 2005 +0000 183071 183072 EXA support in the modular build. 183073 183074commit 33326f4ee884aa677e4568c3eaed4311a8ed0f2a 183075Author: Chris Lee <clee@c133.org> 183076Date: Sat Aug 6 23:27:33 2005 +0000 183077 183078 Adding exa support to the modular server. 183079 183080commit b90a88e80310e5650e2abed4c151889d9c0bd316 183081Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183082Date: Sat Aug 6 19:48:12 2005 +0000 183083 183084 Add checks for functions used in os/access.c & os/connection.c: 183085 getdtablesize(), getifaddrs(), getpeereid(), getpeerucred() 183086 183087commit 2304c14fe5d3796c557a86d0ece1f0fb92591469 183088Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183089Date: Sat Aug 6 19:14:24 2005 +0000 183090 183091 Fix endian test to work on big-endian platforms correctly 183092 183093commit fc29d014aa586229cdc976aa1bfe38bd949f7cde 183094Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183095Date: Sat Aug 6 16:29:20 2005 +0000 183096 183097 Merge in patch release RC handling code from 6.8.2 (Kevin Martin) 183098 183099commit 9828a38ed695fed9788ace8bba575e521fa354b7 183100Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183101Date: Sat Aug 6 03:14:50 2005 +0000 183102 183103 Solaris build fixes needed for modular builds 183104 183105commit f51047c0b6749ae8f04925eb62706b4743328383 183106Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183107Date: Fri Aug 5 15:43:31 2005 +0000 183108 183109 Fix 32-bit vs. 64-bit test for Solaris 183110 183111commit 4220a0c4ccd0672460d31db239829384aee69f9a 183112Author: Matthew Allum <breakfast@10.am> 183113Date: Fri Aug 5 09:08:32 2005 +0000 183114 183115 Make Xephyr work without shadow fb 183116 183117commit fedbce2186bbe3bd8d23a4d74c35f10286930a7d 183118Author: Adam Jackson <ajax@nwnk.net> 183119Date: Fri Aug 5 00:13:31 2005 +0000 183120 183121 EXTRA_DIST hacks to get all the OS support files into the tarball. 183122 183123commit 39a80312e0c6e9f3b260d2f8279c71aef0a60d12 183124Author: Adam Jackson <ajax@nwnk.net> 183125Date: Thu Aug 4 22:31:20 2005 +0000 183126 183127 - Bug #3960: Add remaining GLX and pci.ids stuff to DIST (Cha Young-Ho, 183128 Olli Helenius) 183129 - Fix some distcheck problems for non-solaris systems 183130 - Change server version number from the Xorg rev to an independent one 183131 - _POSIX_SOURCE and _XOPEN_SOURCE defined to the right values 183132 183133commit ed4d77a16a226a7bc88d0d4f5c595d16cdf46360 183134Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183135Date: Thu Aug 4 21:05:12 2005 +0000 183136 183137 Thu Aug 4 16:08:00 2005 Søren Sandmann <sandmann@redhat.com> 183138 Make the fbCompose paths use the existing inline functions instead of 183139 macros. Various other cleanups. 183140 Remove macro definitions, move typedef to fbmmx.c 183141 183142commit e62be6d2721e7f43c7bba6b8da4e5ec8c364697d 183143Author: Adam Jackson <ajax@nwnk.net> 183144Date: Thu Aug 4 18:45:46 2005 +0000 183145 183146 - Check for vsnprintf 183147 - Don't build the Xorg DDX on darwin 183148 - Workaround for gcc defining __ppc__ but not __powerpc__ 183149 183150commit 13bec87f45509452c643e20745e12878584d606a 183151Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183152Date: Thu Aug 4 02:39:56 2005 +0000 183153 183154 Fix inline assembly versions of in*/out* for Sun compilers Add check for 183155 ancient USL console code on Solaris to allow building on old Solaris 183156 x86 versions 183157 183158commit 9802106864a6b2c24df8e9fcb3c3471033abd5c6 183159Author: Adam Jackson <ajax@nwnk.net> 183160Date: Wed Aug 3 01:41:34 2005 +0000 183161 183162 dist fixes for cw and GL (Donnie Berkholz, me) 183163 183164commit d7c246f35b1ba7ecbbc086dd1229b298faddec96 183165Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183166Date: Tue Aug 2 23:57:52 2005 +0000 183167 183168 Only build ix86Pci.c on x86 & amd64 platforms Set DEFAULT_INT10 to x86emu 183169 on everything but Linux/x86 Fix Sbus.c inclusion on sparc platforms 183170 183171commit 3d0d95004c65cd3a29c94fec99e13ab7fdc19c7c 183172Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183173Date: Tue Aug 2 20:47:47 2005 +0000 183174 183175 Add check for walkcontext() to enable stack trace dumper on Solaris 183176 183177commit bb5258a21d9efbb7d8b4226a44bd5738eb46d32a 183178Author: Keith Packard <keithp@keithp.com> 183179Date: Tue Aug 2 17:57:36 2005 +0000 183180 183181 Mark generated files as BUILT_SOURCES and CLEANFILES to ensure they are 183182 regenerated after 'make clean' and to make parallel builds work right. 183183 183184commit 0c608a690e85064345d7ebdf6b03e2f3bd861708 183185Author: Keith Packard <keithp@keithp.com> 183186Date: Tue Aug 2 06:48:20 2005 +0000 183187 183188 Note yacc and lex generated files as BUILT_SOURCES so they are created 183189 before anything is compiled; required for parallel builds 183190 183191commit d31de217e6e9bfb0e1cc99141d80def9382de9a9 183192Author: Adam Jackson <ajax@nwnk.net> 183193Date: Tue Aug 2 03:36:09 2005 +0000 183194 183195 distcheck fixes 183196 183197commit c83772fef7ea0379db9bab1120f064b80b74f377 183198Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183199Date: Tue Aug 2 01:50:48 2005 +0000 183200 183201 Set OSNAME for "Build Operating System" line in Xorg log 183202 183203commit 005a9cf1f5fe688f303b954870afd35635fc33bf 183204Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183205Date: Tue Aug 2 01:03:25 2005 +0000 183206 183207 Fix Solaris assembly source to work with libtool 183208 183209commit 48b130285ad1309a1ff5b2148b3bae5f5b642488 183210Author: Adam Jackson <ajax@nwnk.net> 183211Date: Tue Aug 2 00:47:08 2005 +0000 183212 183213 Move AM_CONDITIONAL for ia64 to global scope 183214 183215commit 03379474e0fd9c21ac058e2319618934359b3262 183216Author: Adam Jackson <ajax@nwnk.net> 183217Date: Tue Aug 2 00:45:30 2005 +0000 183218 183219 only do the AS_HELP_STRING workaround once, d'oh 183220 183221commit 2d7b1893befb84850f3221cbe70b3bce4e621d63 183222Author: Adam Jackson <ajax@nwnk.net> 183223Date: Tue Aug 2 00:41:07 2005 +0000 183224 183225 - Workaround for AS_HELP_STRING compatibility with older autoconf. 183226 - New intermediate build target for Xorg server 183227 - ia64 fixes 183228 183229commit 7c84189ea1dc236bc979bb4bf158ecb1a6b45137 183230Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183231Date: Mon Aug 1 22:22:16 2005 +0000 183232 183233 Bugzilla #3566 <https://bugs.freedesktop.org/show_bug.cgi?id=3566> Patch 183234 #2906 <https://bugs.freedesktop.org/attachment.cgi?id=2906> Fix for 183235 RENDER, repeating pictures and offscreen memory (Owen Taylor) 183236 183237commit 2acd29c93fd3b3d438887f0ca9be6713db81c1e8 183238Author: Adam Jackson <ajax@nwnk.net> 183239Date: Mon Aug 1 22:13:20 2005 +0000 183240 183241 Bump autoconf dependency to 2.57 from 2.53. Add 2.57 compatibility hack for 183242 the new AS_HELP_STRING syntax from 2.58+. 183243 183244commit b7a43fa0f112d92cce71642142e07276da4de681 183245Author: Adam Jackson <ajax@nwnk.net> 183246Date: Mon Aug 1 19:00:21 2005 +0000 183247 183248 Bug #3739: Fail soft on unknown extension string. 183249 183250commit 0285804811bd24ad7de8894f8188b87e821e54cc 183251Author: Adam Jackson <ajax@nwnk.net> 183252Date: Mon Aug 1 18:41:54 2005 +0000 183253 183254 GLX fix for loadable servers: indirect_size.c goes in libglx, not libGLcore 183255 183256commit 78d51253e229bcb428039911d80a5d89d4bf4491 183257Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183258Date: Mon Aug 1 16:07:14 2005 +0000 183259 183260 Fri Jul 29 17:20:53 2005 Søren Sandmann <sandmann@redhat.com> 183261 Fix rounding bug. 183262 183263commit fd84b3b56280ef88a8c848c4338f0bea906ce6d8 183264Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183265Date: Sat Jul 30 18:56:32 2005 +0000 183266 183267 Clear compiler warnings. (Stefan Dirsch) 183268 183269commit 9e9f9cb4878d597b3fa6de7732d2866e24c32f54 183270Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183271Date: Fri Jul 29 21:55:57 2005 +0000 183272 183273 Need to define SVR4 on SVR4-ish systems since many headers and source files 183274 check for it, expecting Imake to have defined it already. (Should be 183275 replaced with more specific checks & defines, but there's too many to 183276 get them all right now.) 183277 183278commit 14ade55b83aa72240a555b3b9def9f40a6a38cd9 183279Author: Kevin E Martin <kem@kem.org> 183280Date: Fri Jul 29 21:22:58 2005 +0000 183281 183282 Various changes preparing packages for RC0: 183283 - Verify and update package version numbers as needed 183284 - Implement versioning scheme 183285 - Change bug address to point to bugzilla bug entry form 183286 - Disable loadable i18n in libX11 by default (use --enable-loadable-i18n to 183287 reenable it) 183288 - Fix makedepend to use pkgconfig and pass distcheck 183289 - Update build script to build macros first 183290 - Update modular Xorg version 183291 183292commit 2654f88cee86fae9db169dc8668492857fbbad98 183293Author: Kevin E Martin <kem@kem.org> 183294Date: Fri Jul 29 18:40:54 2005 +0000 183295 183296 Fix code to handle printing 7.0 release candidates properly. 183297 183298commit d5d669b04ecbd91bb2db9ddc6238acabbeaabbe0 183299Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183300Date: Thu Jul 28 23:33:57 2005 +0000 183301 183302 If neither --enable-dmx or --disable-dmx are specified, default to "yes" on 183303 Linux, "no" on everything else (matches defaults in monolith Imake) 183304 183305commit 86529c07d6ff69ef936ee370be07b373e2961360 183306Author: Kevin E Martin <kem@kem.org> 183307Date: Thu Jul 28 23:28:34 2005 +0000 183308 183309 Fix distcheck for server 183310 183311commit 3e4bf8a8f46e14078c64c56ad303c80cd9d8d5be 183312Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183313Date: Thu Jul 28 02:38:23 2005 +0000 183314 183315 Rework wrapping of common mouse driver to not require mousePriv.h, so that 183316 modular tree xserver can be built on Solaris. Wrap more ioctls in 183317 SYSCALL() to restart on signals. 183318 183319commit cfd6c74067de7fdb2aeddee6780c29231a56acdc 183320Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183321Date: Wed Jul 27 17:50:13 2005 +0000 183322 183323 Add dbestruct.h to sdk_HEADERS 183324 183325commit 5ced854186b851ba2c9ac84eb6d7f141fdee487b 183326Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183327Date: Wed Jul 27 17:35:25 2005 +0000 183328 183329 Add miwideline.h mistruct.h mifpoly.h to sdk_HEADERS 183330 183331commit 7bd6b0640e69552ed4c9daac26e41410dd7338b4 183332Author: Kevin E Martin <kem@kem.org> 183333Date: Wed Jul 27 08:16:35 2005 +0000 183334 183335 Add #include <dmx-config.h> for building dmx/glxProxy in the modular tree. 183336 183337commit b9ccd89918e1c79d6013fe563c7fa933c1274837 183338Author: Kevin E Martin <kem@kem.org> 183339Date: Wed Jul 27 08:11:26 2005 +0000 183340 183341 Fix dmx/glxProxy build 183342 183343commit 8f6664fcff319ee155e7864ce25fa11c1fa5594b 183344Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183345Date: Tue Jul 26 23:40:45 2005 +0000 183346 183347 Add these files to the sdk: 183348 cbf32.h cfb16.h cfbmap.h cfbunmap.h cfbunmap.h cfbmskbits.h xf86sbusBus.h 183349 mifillarch.h mispans.h 183350 183351commit 3cbb55f309b8eaba246d1d948b59883f9ed28bf4 183352Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183353Date: Tue Jul 26 22:17:27 2005 +0000 183354 183355 Install xorgVersion.h 183356 183357commit df5d47560c19b01f2998806c4b62f459a6b5fc02 183358Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183359Date: Tue Jul 26 18:56:45 2005 +0000 183360 183361 xserver/xorg/fb/Makefile.am: install fboverlay.h 183362 183363commit b25e6f82a1189a67208a2a4980d732ab77e64a35 183364Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183365Date: Tue Jul 26 18:21:22 2005 +0000 183366 183367 fb/Makefile.am: add fbpseudocolor.h to sdk_HEADERS 183368 hw/xfree86/xaa/Makefile.am: add xaaWrapper.h to sdk_HEADERS 183369 183370commit e7502158d6922a149d0aaf1de209f0b58fa444bf 183371Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183372Date: Sun Jul 24 16:24:17 2005 +0000 183373 183374 Add xf86DeallocateGARTMemory stub so that Xorg can be built on Solaris 183375 again. 183376 183377commit 7c1148c0fbc00716bb7ec621a1b968b4127cbb00 183378Author: Adam Jackson <ajax@nwnk.net> 183379Date: Sat Jul 23 19:32:57 2005 +0000 183380 183381 Fire the Mesa symlink script from the configure stage. 183382 183383commit f331a325447449982efba8adc5bc079371dfb77a 183384Author: Adam Jackson <ajax@nwnk.net> 183385Date: Sat Jul 23 19:29:58 2005 +0000 183386 183387 Add the symlink script for the Mesa source. 183388 183389commit a5532e456d763718036e84891ff57c6f7417615a 183390Author: Stuart R. Anderson <anderson@netsweng.com> 183391Date: Sat Jul 23 18:16:06 2005 +0000 183392 183393 Change the default depth back to 8 like it was originally. Disable the 183394 addition of depth 12 & 30. It causes more than MAXFORMATS depths to be 183395 added to the screen which causes an array in the Screen structure to 183396 overflow and crash on server reset. 183397 183398commit d8a1241a24d75c258548875350bff4d0acc795ab 183399Author: Stuart R. Anderson <anderson@netsweng.com> 183400Date: Sat Jul 23 17:49:45 2005 +0000 183401 183402 Add a safety check to catch when numDepths GCperDepthexceeps MAXFORMATS 183403 183404commit 93d29b4554d26b22ca43311264f85ea8e14804c2 183405Author: Adam Jackson <ajax@nwnk.net> 183406Date: Sat Jul 23 02:04:39 2005 +0000 183407 183408 Emit makefiles for glapi, slang, and grammar. Conditionally define GLX_LIBS 183409 if GLX support is requested. Add GLX_LIBS to the link deps for Xvfb and 183410 Xnest. All GLX support should work now for the vfb, xnest, and xfree86 183411 DDXes. 183412 183413commit 525d663ec5a8a181a1e36f3787b24cdb77290c17 183414Author: Adam Jackson <ajax@nwnk.net> 183415Date: Sat Jul 23 02:02:18 2005 +0000 183416 183417 glapi build (forgot this one somehow) 183418 183419commit 6aaceef3231818682dccf1969b211f1dc798da1f 183420Author: Adam Jackson <ajax@nwnk.net> 183421Date: Sat Jul 23 02:00:52 2005 +0000 183422 183423 Fix the build system to reflect the rest of the Mesa core. Almost links, 183424 needs a code fix to Mesa though. 183425 183426commit d7e2cadb13ef0e77d864c8282489f544b0e1dede 183427Author: Adam Jackson <ajax@nwnk.net> 183428Date: Thu Jul 21 02:29:05 2005 +0000 183429 183430 GLX needs gl.h from Mesa too 183431 183432commit bf32f4739c6ef7f41cde6cc409d42e00462402fb 183433Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 183434Date: Tue Jul 19 20:51:46 2005 +0000 183435 183436 programs/Xserver/hw/xfree86/os-support/shared/agp_noop.c 183437 programs/Xserver/hw/xfree86/os-support/linux/lnx_agp.c 183438 programs/Xserver/hw/xfree86/os-support/xf86_OSproc.h 183439 //bugs.freedesktop.org/show_bug.cgi?id=3164> Add xf86DeallocateGARTMemory() 183440 function call (Austin Yuan) 183441 183442commit 931abdba8410bf1188d69d54c57543a21ea36968 183443Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183444Date: Tue Jul 19 02:50:00 2005 +0000 183445 183446 Change #include "X.h" to <X11/X.h> 183447 183448commit 4f2e731aba0c5694abb65a877395cc5f9869c556 183449Author: Daniel Stone <daniel@fooishbar.org> 183450Date: Sun Jul 17 07:34:31 2005 +0000 183451 183452 objdir != srcdir fixes. 183453 183454commit a6bd7ad985e138874bb0e1f33f545651dfde32a9 183455Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183456Date: Sun Jul 17 01:52:33 2005 +0000 183457 183458 Set GLX automake conditional after we've tested for mesa source and set GLX 183459 to "no" if it's --with-mesa-source wasn't set. 183460 183461commit ead37b08699f665a856c7ba29578c27c27db4100 183462Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183463Date: Sat Jul 16 20:52:25 2005 +0000 183464 183465 Mark variables modified in signal handlers as volatile (part of Sun bug id 183466 4496504) 183467 183468commit bbb49449cbc0815fd9c4faf9b6ee32be99a8aa3d 183469Author: Adam Jackson <ajax@nwnk.net> 183470Date: Sat Jul 16 16:30:40 2005 +0000 183471 183472 Fix --with-mesa-source parsing, and turn GLX off if no path to Mesa given. 183473 183474commit d69e9872ae91c9c17a33e1ed763f98453d950b39 183475Author: Adam Jackson <ajax@nwnk.net> 183476Date: Sat Jul 16 16:29:30 2005 +0000 183477 183478 Typo fix, and re-add GLcore to the dixmods build 183479 183480commit 8bc883c48c8f99502707998a8257c0563cae7d3a 183481Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183482Date: Sat Jul 16 09:00:44 2005 +0000 183483 183484 Update build instructions. Add details about how the expat, fontconfig, 183485 libpng, freetype, and xterm builds are now handled. 183486 Update text docs from SGML masters. 183487 183488commit 8e600b87643f7f4e19923dd6a0f2eec83303363a 183489Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183490Date: Sat Jul 16 07:39:48 2005 +0000 183491 183492 Only pass -rdynamic when using gcc 183493 183494commit 4f04e26c69747929243f176feaa90f3042481f53 183495Author: Adam Jackson <ajax@nwnk.net> 183496Date: Sat Jul 16 04:30:33 2005 +0000 183497 183498 Say --with-mesa-source to get GLcore built. 183499 183500commit 4913316f25e76381844cf15aa64eff620c2807a1 183501Author: Adam Jackson <ajax@nwnk.net> 183502Date: Sat Jul 16 04:29:28 2005 +0000 183503 183504 Switch to MESA_SOURCE which we're already AC_SUBSTing 183505 183506commit 7facb209a705ca406b05272c233ea0cfd6a5b266 183507Author: Adam Jackson <ajax@nwnk.net> 183508Date: Sat Jul 16 04:24:21 2005 +0000 183509 183510 Hook in the GLcore build 183511 183512commit 2f9a7c6ee1012b2c7664c53f405affaf82bd0953 183513Author: Adam Jackson <ajax@nwnk.net> 183514Date: Sat Jul 16 04:13:33 2005 +0000 183515 183516 Hey look, GLcore links 183517 183518commit 5f81eb1d89f623dd9281e686a4d4d4f403e459c6 183519Author: Kevin E Martin <kem@kem.org> 183520Date: Sat Jul 16 03:49:59 2005 +0000 183521 183522 Fix the *-config.h includes so that it is possible to build modules without 183523 having to use -include in the Makefile. 183524 183525commit 0635acab08bade9bfd9d2abd3ea5a3fe525bc647 183526Author: Kevin E Martin <kem@kem.org> 183527Date: Sat Jul 16 03:46:01 2005 +0000 183528 183529 Add record module building support Change module building to not use 183530 -include Fix xnest, dmx and vfb builds to -DXFree86Server for fbcmap 183531 compilation 183532 183533commit 92b23ff426936909e1ff5e063b75e5f112b93ca1 183534Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183535Date: Sat Jul 16 01:17:52 2005 +0000 183536 183537 Update to match recent changes to pci.ids & extrapci.ids for people 183538 building on systems without perl. 183539 183540commit 6ce3743948a7151c721194594091a7e1183e2f32 183541Author: Adam Jackson <ajax@nwnk.net> 183542Date: Sat Jul 16 00:51:14 2005 +0000 183543 183544 start the GLcore build system 183545 183546commit 2f0487521da3da6784dd964c915071b82d749108 183547Author: Kevin E Martin <kem@kem.org> 183548Date: Fri Jul 15 23:08:31 2005 +0000 183549 183550 Fix the *-config.h includes for the files that have moved in the modular 183551 tree. 183552 183553commit a407fa373bb72f29050e3f027042011075b3bdf0 183554Author: Kevin E Martin <kem@kem.org> 183555Date: Fri Jul 15 22:51:05 2005 +0000 183556 183557 Move drm up to os-support since the files are shared by multiple platforms. 183558 Fix the dri and drm build. Fix server-side DMX extension build. Make 183559 xf4bpp use the correct version of mfbline.c for mfbseg.c. Add #ifndef 183560 _HEADERNAME_H_/#define _HEADERNAME_H_/.../#endif to the headers. 183561 183562commit eb6fa0dc15516a0a436090efc6d85f22baeec015 183563Author: Adam Jackson <ajax@nwnk.net> 183564Date: Fri Jul 15 22:14:28 2005 +0000 183565 183566 Enable GLX build by default 183567 183568commit 0e352a8b2b4c78c291074b9531ed7afa7a20ac17 183569Author: Adam Jackson <ajax@nwnk.net> 183570Date: Fri Jul 15 21:59:39 2005 +0000 183571 183572 disable GLcore momentarily until a build system exists. --enable-glx should 183573 work now. 183574 183575commit 4bf453086418e93e81b24d3d2a0f49a9357acff3 183576Author: Kevin E Martin <kem@kem.org> 183577Date: Fri Jul 15 05:48:29 2005 +0000 183578 183579 Fix keyboard state when XEVIE is not enabled (Keith Packard). 183580 183581commit 562acf2e69dcf08f8db8f8eeaa162949d45f70b2 183582Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183583Date: Thu Jul 14 23:41:15 2005 +0000 183584 183585 Add -I$(top_builddir)/hw/xfree86/dixmods/extmod to Xext/Makefile.am Patch 183586 from Stefan Dirsch 183587 183588commit defeb56fc9559661fa632935a4c76c500c7edcf4 183589Author: Kevin E Martin <kem@kem.org> 183590Date: Thu Jul 14 22:20:16 2005 +0000 183591 183592 Fix the build when DRI is enabled 183593 183594commit 6b546d0ec07e493b501e82300b3e6c143cd4d0ac 183595Author: Kevin E Martin <kem@kem.org> 183596Date: Thu Jul 14 03:36:44 2005 +0000 183597 183598 Add #include <xnest-config.h> to the Xnest source files for modularization. 183599 183600commit 7375f4d13626bbba4204e2f08f41c212b2eed992 183601Author: Kevin E Martin <kem@kem.org> 183602Date: Thu Jul 14 03:32:09 2005 +0000 183603 183604 Add support for Xnest 183605 183606commit c582560c62fd8181b7521e470f118a59c418a95f 183607Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183608Date: Thu Jul 14 02:59:34 2005 +0000 183609 183610 Add partial in*/out* assembly support for Sun compilers on x86 183611 183612commit 64bf3a81a3212dc2c0f55aebdc74a618ca6a32c0 183613Author: Adam Jackson <ajax@nwnk.net> 183614Date: Thu Jul 14 02:53:31 2005 +0000 183615 183616 loadable extmod build system. 183617 183618commit 6c96e0c9e4b3f1e8dfa5dcf7366bb838dc0724ec 183619Author: Adam Jackson <ajax@nwnk.net> 183620Date: Thu Jul 14 01:36:58 2005 +0000 183621 183622 loader support for extmod 183623 183624commit 8ea4a1b759eae0279ce619c663f2cd2f6dee8d71 183625Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183626Date: Thu Jul 14 01:00:39 2005 +0000 183627 183628 First pass at Solaris os-support 183629 183630commit f07e905553783dc0133015bcbf94d3b6be68b19c 183631Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183632Date: Thu Jul 14 00:28:47 2005 +0000 183633 183634 - Add build system for xf86-video-ati 183635 - Add build system for xserver/xorg/hw/xfree86/dri 183636 - Add glcore.h to symlink.sh 183637 - Symlink.sh: add some more DRI files, symlink glcore.h from 183638 extras/Mesa/include/GL/internal to proto/GL 183639 - proto/GL/Makefile.am: install glcore.h 183640 - xf86-video-ark: s/module-dir/xorg-module-dir/ 183641 183642commit 939b7720f17ec5ac5edcc6cfe70453160ecb0161 183643Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183644Date: Thu Jul 14 00:02:05 2005 +0000 183645 183646 Merge SVR4/pre-Solaris 8 and Solaris 8+ sections for greater consistency, 183647 easier maintenance, and to fix some missing headers when building the 183648 modular Xorg on Solaris. 183649 183650commit 23d25a656d7fdbafa1c78be2950fb405f0a1b87b 183651Author: Adam Jackson <ajax@nwnk.net> 183652Date: Wed Jul 13 23:41:21 2005 +0000 183653 183654 Remove references to DDX-specific extensions 183655 183656commit f4626bb72b9955846abee018ae31c1aeb51b8dbf 183657Author: Adam Jackson <ajax@nwnk.net> 183658Date: Wed Jul 13 22:59:39 2005 +0000 183659 183660 dlloader is the default in 6.9, but the loader doesn't get told to prefer 183661 dlloader modules unless you -DDLOPEN_HACK. 183662 183663commit 4f38526566dbcc296d124bb852adfa30ac4d927e 183664Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183665Date: Wed Jul 13 22:54:02 2005 +0000 183666 183667 - Use fbdevhwstub.c if <linux/fb.h> is not found 183668 - Use x86emu on Solaris instead of vm86 183669 - Better per-OS control over which xf86Kbd*.c and *Pci.c files to build 183670 - Set various #defines to be defined or not on Solaris as needed 183671 183672commit db8aa17f15f62f243400b0267bf281dd27738644 183673Author: Adam Jackson <ajax@nwnk.net> 183674Date: Wed Jul 13 22:19:36 2005 +0000 183675 183676 updated comment for libbitmap 183677 183678commit 4f9e76ed6a61b414ffc94550727a138c3f73996f 183679Author: Adam Jackson <ajax@nwnk.net> 183680Date: Wed Jul 13 22:09:52 2005 +0000 183681 183682 Dear libtool: Loadable modules do not need version numbers. kthnxbye. 183683 183684commit 40ad7321f5d8d216523d575c4414736d8cfe9e88 183685Author: Adam Jackson <ajax@nwnk.net> 183686Date: Wed Jul 13 21:52:51 2005 +0000 183687 183688 Generate useful loadable modules by actually linking in the blobs from the 183689 DIX 183690 183691commit 0e08818d3951c324e5953bd23a8a9457b3fec78f 183692Author: Adam Jackson <ajax@nwnk.net> 183693Date: Wed Jul 13 21:41:02 2005 +0000 183694 183695 Hook dbe into the build 183696 183697commit 2d4ddb5606b7328c591a7bffafbb49bdaf2adcf9 183698Author: Adam Jackson <ajax@nwnk.net> 183699Date: Wed Jul 13 21:36:27 2005 +0000 183700 183701 double-buffering extension 183702 183703commit c2ec9fbb15c776b9c64451cf32927b1b8d1c560c 183704Author: Kevin E Martin <kem@kem.org> 183705Date: Wed Jul 13 21:17:53 2005 +0000 183706 183707 Only build Xi/stubs.c on DDXs that don't support the Xinput extension 183708 183709commit 74a534f63c1ed016f24dbc3ca31f05b81076e8f0 183710Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 183711Date: Wed Jul 13 20:19:38 2005 +0000 183712 183713 xc/programs/Xserver/hw/xfree86/drivers/i2c/*.c: include xorg-config.h 183714 instead of config.h xserver/xorg/hw/xfree86/i2c/Makefile.am: Add i2c 183715 drivers 183716 183717commit 1dc3e96e4077a912dd3aa13fa80099f5864b641f 183718Author: Torrey Lyons <torrey@mrcla.com> 183719Date: Wed Jul 13 16:30:53 2005 +0000 183720 183721 Fix compilation where uint is not defined. 183722 183723commit 19e20c1470c1f8d15f2a78fb29545bde06a65516 183724Author: Lars Knoll <lars@trolltech.com> 183725Date: Wed Jul 13 08:58:37 2005 +0000 183726 183727 don't clobber %ebx in the assembler. 183728 183729commit bfb10bd2dcca65ba5d346c9d7da594a81c35c101 183730Author: Lars Knoll <lars@trolltech.com> 183731Date: Wed Jul 13 07:28:17 2005 +0000 183732 183733 Fix potential buffer overflow and a smaller bug in the convolution filter 183734 183735commit 778a2703b233641e298fa81ef9c477943c496305 183736Author: Lars Knoll <lars@trolltech.com> 183737Date: Wed Jul 13 07:12:33 2005 +0000 183738 183739 fix compilation 183740 183741commit 66bc36473c238fdee7e6c1d31e6e5f6813a7541a 183742Author: Daniel Stone <daniel@fooishbar.org> 183743Date: Tue Jul 12 23:36:27 2005 +0000 183744 183745 Use builddir, not srcdir, for built files. 183746 183747commit d6808a48d2dffd72f618fa372fba993736638799 183748Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183749Date: Tue Jul 12 18:16:03 2005 +0000 183750 183751 Fix scanpci -v core dump when subsys vendor/device id's are NOVENDOR & 183752 NODEVICE. Bug #3763 <https://bugs.freedesktop.org/show_bug.cgi?id=3763> 183753 Patch #3074 <https://bugs.freedesktop.org/attachment.cgi?id=3074> 183754 183755commit cda9c7b2678ea08ac6176a9eee72e6e511134b8e 183756Author: Lars Knoll <lars@trolltech.com> 183757Date: Tue Jul 12 14:50:10 2005 +0000 183758 183759 add x86emu. 183760 183761commit 41002623f314444bd416fd5f445a0425c5b59df0 183762Author: Lars Knoll <lars@trolltech.com> 183763Date: Tue Jul 12 10:02:10 2005 +0000 183764 183765 Add MMX Code paths for the basic composition operations in 183766 fbComposeGeneral. 183767 183768commit 697cf74fb50a550b8f7e124dc8f463a55519795f 183769Author: Keith Packard <keithp@keithp.com> 183770Date: Tue Jul 12 03:09:20 2005 +0000 183771 183772 Make Xprt run when linked against Xlib for Xrm 183773 183774commit 419448ea7b0d7f672e568cb1d8b4e190a1f54825 183775Author: Adam Jackson <ajax@nwnk.net> 183776Date: Tue Jul 12 03:03:16 2005 +0000 183777 183778 Pull libdri out of the build for now until I get something better worked 183779 out 183780 183781commit f054bf2ff6b94e285e7f2d174163c01b07b07143 183782Author: Adam Jackson <ajax@nwnk.net> 183783Date: Tue Jul 12 01:29:41 2005 +0000 183784 183785 build fix 183786 183787commit 02427d4d04f70109a499578c6762654463ebdae4 183788Author: Kevin E Martin <kem@kem.org> 183789Date: Tue Jul 12 01:20:36 2005 +0000 183790 183791 Add support for building Xdmx and Xvfb 183792 183793commit 3fe6b5bb30e8e1b9017a9cf818fcceb279a28e65 183794Author: Adam Jackson <ajax@nwnk.net> 183795Date: Tue Jul 12 01:17:39 2005 +0000 183796 183797 GLX server support 183798 183799commit 377e3bddd4d73154520a9582d75de2b20ae532a5 183800Author: Adam Jackson <ajax@nwnk.net> 183801Date: Tue Jul 12 00:55:43 2005 +0000 183802 183803 Forgot to add these for some reason 183804 183805commit e348ac4b4dfb1112c19fe5fe5441182e66716087 183806Author: Kevin E Martin <kem@kem.org> 183807Date: Tue Jul 12 00:52:48 2005 +0000 183808 183809 Prepare Xdmx and Xvfb for modularization by adding appropriate #include 183810 <{dix,dmx}-config.h> to the source files. 183811 183812commit b7a9a6a03560bdf6584c71bf0b546301bba9ab89 183813Author: Kevin E Martin <kem@kem.org> 183814Date: Mon Jul 11 17:52:00 2005 +0000 183815 183816 Remove unneeded xf86drm.h includes to fix modular build. 183817 183818commit cf4dfd650dbc2bb65eae4eea2acfb4a4c5295548 183819Author: Adam Jackson <ajax@nwnk.net> 183820Date: Mon Jul 11 02:29:50 2005 +0000 183821 183822 Prep for modular builds by adding guarded #include "config.h" everywhere. 183823 183824commit c5548086f3864c828f0cad65d2708cefd2025947 183825Author: Adam Jackson <ajax@nwnk.net> 183826Date: Mon Jul 11 00:42:52 2005 +0000 183827 183828 Start filling in glx build. Add xf86Version.h and a few DRI headers to the 183829 sdk (needed for modular driver builds). 183830 183831commit 955fe17133d841758a18072a9acabedc81dc4562 183832Author: Adam Jackson <ajax@nwnk.net> 183833Date: Sun Jul 10 21:45:55 2005 +0000 183834 183835 add some convenience variables for the drivers 183836 183837commit 78fab90230c61241af29f0c94f401ce0bc749b6b 183838Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183839Date: Sat Jul 9 16:51:58 2005 +0000 183840 183841 Bug #3740 <https://bugs.freedesktop.org/show_bug.cgi?id=3740> Patch #3058 183842 <https://bugs.freedesktop.org/attachment.cgi?id=3058> 183843 08-Jul-2005 nv driver updates from Mark Vojkovich: 183844 Change some console restore code for NV11. Hopefully, we can more reliably 183845 restore the console for desktop systems using DVI. This may correct a 183846 recent regression on NV11. 183847 Also, new PCI IDs. 183848 Add new nVidia PCI ids to match nv_driver.c changes. 183849 183850commit 3c92389185f0c9fa3b8c299a084b10c12bcab52c 183851Author: Zack Rusin <zack@kde.org> 183852Date: Sat Jul 9 14:15:35 2005 +0000 183853 183854 Patch from Thomas Winischhofer to kick out all pixmaps to system ram upon a 183855 VT switch and vice versa when returning. 183856 183857commit 327741486e807c068383a771c04c9042b0589c37 183858Author: Adam Jackson <ajax@nwnk.net> 183859Date: Sat Jul 9 02:22:29 2005 +0000 183860 183861 Don't try to link fb, shadow, or vgahw into the loadable Xorg server, as 183862 they're only supposed to be loadable modules. 183863 183864commit 2b8e4db9ac4b4c8f8fd73c00436d6abec2faa535 183865Author: Adam Jackson <ajax@nwnk.net> 183866Date: Fri Jul 8 20:27:30 2005 +0000 183867 183868 -lXfont isn't enough for libbitmap, you need to get the -L from pkg-config 183869 183870commit e84648df7d7eb700b7c2d35fdef0be1f463853fa 183871Author: Zack Rusin <zack@kde.org> 183872Date: Fri Jul 8 17:07:52 2005 +0000 183873 183874 mark drawable as dirty on copying/painting windows 183875 183876commit 4ab73a73f4aa1f02cc8dada185b5dcbddfe43878 183877Author: Zack Rusin <zack@kde.org> 183878Date: Fri Jul 8 07:43:00 2005 +0000 183879 183880 heh, oops (thanks Thomas) 183881 183882commit 0a28516a6e641b41e674f69fc228b0babbe1743b 183883Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183884Date: Thu Jul 7 19:07:28 2005 +0000 183885 183886 More compiler warning fixes for missing prototypes: 183887 Add prototype for XkbSetExtension() 183888 Add #include <ctype.h> for isspace() & isdigit() 183889 Add #include <grp.h> (for initgroups()) and remove extra * 183890 Add prototype for xorgGetVersion() 183891 183892commit 2c3c4060fd2a52f147eda01b11222c341c6e3dee 183893Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183894Date: Thu Jul 7 16:18:52 2005 +0000 183895 183896 Bug #2901 <https://bugs.freedesktop.org/show_bug.cgi?id=2901> Add prototype 183897 for XdmAuthenticationInit() (Peter Breitenlohner) 183898 183899commit d0dc574adb79ffacf90b786d4ccfcd1cd8598728 183900Author: Zack Rusin <zack@kde.org> 183901Date: Thu Jul 7 15:05:02 2005 +0000 183902 183903 Check vtSema before accelerating primitives and sync in fallbacks only if 183904 we got vtSema 183905 183906commit 6cba5f1260c20b3bc072fdcc5f3e49fa28ba6414 183907Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183908Date: Thu Jul 7 14:59:48 2005 +0000 183909 183910 Bug #2901 <https://bugs.freedesktop.org/show_bug.cgi?id=2901> Patch #2332 183911 <https://bugs.freedesktop.org/attachment.cgi?id=2332> This patch avoids 183912 79 gcc-3.4.3 warnings 'xxx' declared `static' but never defined mostly 183913 due to including "ftfuncs.h" with the declaration of static functions 183914 defined in "ftfuncs.c". (Peter Breitenlohner) 183915 183916commit 7da3f4a786d9b61f4129c7dbbef80c84abbfde68 183917Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183918Date: Thu Jul 7 03:12:40 2005 +0000 183919 183920 Bug #2901 <https://bugs.freedesktop.org/show_bug.cgi?id=2901> Fix warnings 183921 about redefined macros (Peter Breitenlohner) 183922 183923commit f86562540d1c945bfd96d4b89259d81e4ed25255 183924Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183925Date: Thu Jul 7 02:47:06 2005 +0000 183926 183927 Bug #2901 <https://bugs.freedesktop.org/show_bug.cgi?id=2901> Patch #2325 183928 <https://bugs.freedesktop.org/attachment.cgi?id=2325> Avoid 38 183929 gcc-3.4.2 warnings: suggest parentheses around assignment used as truth 183930 value suggest explicit braces to avoid ambiguous `else' suggest 183931 parentheses around && within || suggest parentheses around arithmetic 183932 in operand of ^ "/*" within comment (Peter Breitenlohner) 183933 Bug #2901 <https://bugs.freedesktop.org/show_bug.cgi?id=2901> Patch #2326 183934 <https://bugs.freedesktop.org/attachment.cgi?id=2326> This patch 183935 removes these 2 gcc-3.4.3 warnings: missing braces around initializer 183936 implicit declaration of function `XpOidTrayMediumListHasTray' (Peter 183937 Breitenlohner) 183938 183939commit 4047191124c237518110e698bde6dab445644449 183940Author: Damien Ciabrini <braun@club-internet.fr> 183941Date: Wed Jul 6 15:34:22 2005 +0000 183942 183943 Added hardware support for transformation matrix (zoom, rotation, etc...). 183944 Fixed the composition function for RGB and A8 format. Avoid syncing 183945 hardware after HW fills or copies. 183946 183947commit 4073f24c90d4aff3f7d83af4c0e733eed082b53b 183948Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 183949Date: Wed Jul 6 15:14:30 2005 +0000 183950 183951 add missing PictureTransformPoint3d call 183952 183953commit c48f631cdb6a279ab1a24a486b05956cfa9ca3da 183954Author: Damien Ciabrini <braun@club-internet.fr> 183955Date: Wed Jul 6 13:57:41 2005 +0000 183956 183957 Fix offset alignment code in the offscreen memory allocator to prevent 183958 textures from being allocated in the next free memory area. 183959 183960commit baa99be190c51b533bf8748c6c6a9bce62594e96 183961Author: Ian Romanick <idr@us.ibm.com> 183962Date: Wed Jul 6 07:16:19 2005 +0000 183963 183964 Put the correct value in the length field of the reply. Previously, the 183965 number of tag/date pairs was specified. This was incorrect. The correct 183966 value is the number of values (one for the tag and one for the value). 183967 Xorg bug: #3210 183968 183969commit 6ba4a2e78a73858648b5b6a39306446d519c3a75 183970Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 183971Date: Tue Jul 5 23:01:51 2005 +0000 183972 183973 Add more defines for XWin DDX Make building of cfb*, afb and mfb 183974 conditional Set FD_SETSIZE=256 on cygwin 183975 183976commit 71ed3ae0c696152e82a98ce2ac0fa67ce6f23464 183977Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 183978Date: Tue Jul 5 22:58:29 2005 +0000 183979 183980 wrap fInternalWM with XWIN_MULTIWINDOWEXTWM 183981 183982commit 456844a613240ce56181f6f3ec7873be9b2dc85d 183983Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 183984Date: Tue Jul 5 18:42:32 2005 +0000 183985 183986 Remove Speedo font module documentation. 183987 Remove Speedo from list of font directories 183988 Update default font path to remove Speedo, add TTF. 183989 183990commit 20c15003f9648de0c03f7d4fa4508afd896b19f5 183991Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 183992Date: Tue Jul 5 18:25:44 2005 +0000 183993 183994 Fix crash on server shutdown 183995 183996commit d72fef26d44e649f39a56730830148d48d77ee9e 183997Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 183998Date: Tue Jul 5 17:52:35 2005 +0000 183999 184000 Fix simultanious presses of Left and Right Control and Shift keys. 184001 https://bugs.freedesktop.org/show_bug.cgi?id=3677 184002 184003commit 0f2c8221c938ce8eebd9f0e111a6b87223c18f9e 184004Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184005Date: Tue Jul 5 16:35:42 2005 +0000 184006 184007 Fix typo which broke window titles 184008 184009commit 0bb2eb8eaaf4054fefbc45bf3cb47bbcf10b7cfd 184010Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184011Date: Tue Jul 5 15:43:20 2005 +0000 184012 184013 Fix problem with fake Control press on Alt-Gr 184014 https://bugs.freedesktop.org/show_bug.cgi?id=3680 184015 https://bugs.freedesktop.org/show_bug.cgi?id=3497 184016 Fix static declaration of winGetBaseDir 184017 184018commit 3af77ad3e754c4d419a1996ca73a9fd01f92388a 184019Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184020Date: Tue Jul 5 14:09:48 2005 +0000 184021 184022 External windowmanagers could connect in multiwindow mode which lead to 184023 strange results with the internal windowmanager. 184024 184025commit 426282268bcdd0e0ca973fa79b414e9065fbfd9d 184026Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184027Date: Mon Jul 4 23:40:09 2005 +0000 184028 184029 Build miinitext.c and fbcmap.c as DDX specific files 184030 Do not define _POSIX_SOURCE on cygwin. 184031 184032commit 845a0ac68b02148db6f2cca81debf20b2331607b 184033Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184034Date: Mon Jul 4 23:11:20 2005 +0000 184035 184036 Revert last changes. They have to be addressed in a different way 184037 184038commit 048045a9e714fc85c26028c4de36dff47644b826 184039Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184040Date: Mon Jul 4 22:18:40 2005 +0000 184041 184042 Add XWin DDX, make Xorg DDX conditional Make XF86VidMode and XF86Misc 184043 conditional 184044 Add XWin DDX 184045 Added DDXTIME, DDXOSFATALERROR, DDXOSVERRORF and DDXBEFORERESET 184046 Added fbcmap.c 184047 Added miinitext.c 184048 Added -I$(top_srcdir)/Xext/extmod to INCLUDES 184049 184050commit 508cdb5cb01a91b1bf3ef31da80b6b1d36286bf0 184051Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184052Date: Mon Jul 4 22:10:43 2005 +0000 184053 184054 Include xwin-config.h if HAVE_XWIN_CONFIG is defined Cleanup X11 includes 184055 handling Warning fixes 184056 184057commit 5e50ae22bf206d6c2cc05e772e05fa5a363acb81 184058Author: Zack Rusin <zack@kde.org> 184059Date: Mon Jul 4 18:55:53 2005 +0000 184060 184061 remove the temporary debugging output 184062 184063commit 7586ac6edea64b30d6187f9ec4d867521c1e769c 184064Author: Adam Jackson <ajax@nwnk.net> 184065Date: Mon Jul 4 18:41:04 2005 +0000 184066 184067 Bug #2216: Multiseat support. From various Debian and Ubuntu patches by 184068 Aivils Stoss, Andreas Schuldei, Branden Robinson, and Daniel Stone. 184069 184070commit ba011dc77dcfaea2843481fbba45a76d8cb9aa83 184071Author: Lars Knoll <lars@trolltech.com> 184072Date: Mon Jul 4 14:47:03 2005 +0000 184073 184074 don't be too smart and try to replace PictOpOver by PictOpSrc when we have 184075 an external alpha map. 184076 Make fbmmx.c compile on gcc 4.0.1. 184077 184078commit a4df8ad75579d9c183f110d79d87cfb9aaf23acd 184079Author: Lars Knoll <lars@trolltech.com> 184080Date: Mon Jul 4 14:23:59 2005 +0000 184081 184082 Fix handling of "super luminescent" colors Fix off by one error in the 184083 transformation handling. 184084 184085commit e34f31762e0454930f30547a9407cc8b941c70f2 184086Author: Zack Rusin <zack@kde.org> 184087Date: Mon Jul 4 14:15:57 2005 +0000 184088 184089 This fixes the close screen mess (crash reported by Thomas). Also hide the 184090 private Exa screen definition. Properly cleanup on screen close and do 184091 not delete the private screen in the DriverFini call. 184092 184093commit 4ef813961dc8dea8924a79e0954d5b6da12e77c0 184094Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184095Date: Mon Jul 4 09:01:43 2005 +0000 184096 184097 provide the uint datatype on WIN32 184098 184099commit 7dc547252af3b23652b1d496957645726c7125a4 184100Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 184101Date: Mon Jul 4 00:16:23 2005 +0000 184102 184103 Fix builds on non-GLIBC systems (missing __GLIBC_PREREQ). Add Solaris stack 184104 backtrace dumper. 184105 184106commit d010de6979a6c51a628f2f8e6d7f479a542d1e8b 184107Author: Daniel Stone <daniel@fooishbar.org> 184108Date: Sun Jul 3 15:51:23 2005 +0000 184109 184110 Abandon the nostdinc experiment. 184111 184112commit b0d80e76d28dbc1ee14453950db000bb4d7f377f 184113Author: Daniel Stone <daniel@fooishbar.org> 184114Date: Sun Jul 3 12:17:04 2005 +0000 184115 184116 Add missing include paths. 184117 184118commit 461eb6ebd9273dc3dcd601ee0a0491981a98acd9 184119Author: Daniel Stone <daniel@fooishbar.org> 184120Date: Sun Jul 3 12:16:29 2005 +0000 184121 184122 Change HAVE_CONFIG_H to HAVE_DIX_CONFIG_H. 184123 184124commit c1e69798ad14fb706f5f6de67e3f53df56f524c6 184125Author: Daniel Stone <daniel@fooishbar.org> 184126Date: Sun Jul 3 12:15:16 2005 +0000 184127 184128 Add explicit PSZ defines, remove unnecessary dix-config.h include. 184129 Add -DXF86PM, forgotten from last commit. 184130 184131commit 4e501d35e260775a43f340b3e1a9aa092570aba4 184132Author: Daniel Stone <daniel@fooishbar.org> 184133Date: Sun Jul 3 10:51:16 2005 +0000 184134 184135 Remove needless usage of DECkeysym.h. 184136 184137commit a107f599518a175dd689417b48788a746303966a 184138Author: Daniel Stone <daniel@fooishbar.org> 184139Date: Sun Jul 3 09:39:54 2005 +0000 184140 184141 Predicate usage of xf86OSPM functions on #ifdef XF86PM. 184142 Remove needless include of ../input/mouse/mouse.h. 184143 184144commit 0bb669638f032e61471007b2fa88285aa5d63903 184145Author: Daniel Stone <daniel@fooishbar.org> 184146Date: Sun Jul 3 08:53:54 2005 +0000 184147 184148 Change <X11/misc.h> and <X11/os.h> to "misc.h" and "os.h". 184149 184150commit 401e4580d6dd9867a691045688680ce410f84cb5 184151Author: Daniel Stone <daniel@fooishbar.org> 184152Date: Sun Jul 3 07:55:00 2005 +0000 184153 184154 Move misc.h and os.h from proto/X11 to xserver/xorg/include. 184155 184156commit 46b64bd5c66abb1bb9f3538c887d10867607bfff 184157Author: Daniel Stone <daniel@fooishbar.org> 184158Date: Sun Jul 3 07:37:35 2005 +0000 184159 184160 Fix more include paths; add dix-config.h to XKB code. 184161 184162commit e03198972ca78b03ad13cb49112c03a052bb763b 184163Author: Daniel Stone <daniel@fooishbar.org> 184164Date: Sun Jul 3 07:02:09 2005 +0000 184165 184166 Add Xtrans definitions (FONT_t, TRANS_CLIENT) to clean up warnings. 184167 Add XSERV_t, TRANS_SERVER, TRANS_REOPEN to quash warnings. 184168 Add #include <dix-config.h> or <xorg-config.h>, as appropriate, to all 184169 source files in the xserver/xorg tree, predicated on defines of 184170 HAVE_{DIX,XORG}_CONFIG_H. Change all Xfont includes to 184171 <X11/fonts/foo.h>. 184172 184173commit b8aef6c474ffc6d637bec178674898ea95ccde47 184174Author: Kevin E Martin <kem@kem.org> 184175Date: Sun Jul 3 03:28:27 2005 +0000 184176 184177 Fix build issues. 184178 184179commit 826a6f029faeabaa783a93dfdccca846f9326b58 184180Author: Daniel Stone <daniel@fooishbar.org> 184181Date: Sat Jul 2 18:59:44 2005 +0000 184182 184183 Continuing Makefile cleanup; add DIX_CFLAGS and XORG_CFLAGS everywhere. 184184 184185commit e58c09d31bdf90210e2ec1ef976cea0459cdc02a 184186Author: Adam Jackson <ajax@nwnk.net> 184187Date: Sat Jul 2 18:06:05 2005 +0000 184188 184189 Bug #3687: Print backtraces on fatal signal on glibc systems. 184190 184191commit e6602b041fe489d51a1d7fac55cbbb12b1826ba1 184192Author: Adam Jackson <ajax@nwnk.net> 184193Date: Sat Jul 2 17:02:23 2005 +0000 184194 184195 Bug #3546: Use MAP_LENGTH instead of a magic number. (Mark McLoughlin) Bug 184196 #3664: Further fixes to Xnest modifier state handling. (Mark 184197 McLoughlin) 184198 184199commit 9b1debcdb6c7df956c06350a6525afb8e6d691fa 184200Author: Daniel Stone <daniel@fooishbar.org> 184201Date: Fri Jul 1 22:43:43 2005 +0000 184202 184203 Change all misc.h and os.h references to <X11/foo.h>. 184204 184205commit 657b4cb8aa0076acae85997c4f0c353b4d86b632 184206Author: Daniel Stone <daniel@fooishbar.org> 184207Date: Fri Jul 1 22:32:34 2005 +0000 184208 184209 Change all misc.h and os.h references to <X11/foo.h>. 184210 184211commit 303c281f956d55e35b05ef8521d0b60d24aa7a10 184212Author: Daniel Stone <daniel@fooishbar.org> 184213Date: Fri Jul 1 21:15:20 2005 +0000 184214 184215 Add auto-generated header files. 184216 184217commit 9a6ec34d2545a23586f11ad51c81f41a940d73d0 184218Author: Daniel Stone <daniel@fooishbar.org> 184219Date: Fri Jul 1 21:13:36 2005 +0000 184220 184221 Use canonical autogen.sh, which supports srcdir != objdir autogen. 184222 184223commit 7ecc2d526c4ea5db2589644a2fec0daf71df36da 184224Author: Daniel Stone <daniel@fooishbar.org> 184225Date: Fri Jul 1 21:12:24 2005 +0000 184226 184227 Remove fbcmap.c hacks from the DIX. 184228 184229commit 6251f9c00a866f64207d23b0a06306ead15298e8 184230Author: Daniel Stone <daniel@fooishbar.org> 184231Date: Fri Jul 1 21:11:16 2005 +0000 184232 184233 Minor build system tweaks. 184234 184235commit 641f32c4368db07831d9d703161a9d4699307621 184236Author: Daniel Stone <daniel@fooishbar.org> 184237Date: Fri Jul 1 20:54:30 2005 +0000 184238 184239 Adding initial build system. 184240 184241commit 507d30546f56bfd172fc43857459c78c1026e97c 184242Author: Daniel Stone <daniel@fooishbar.org> 184243Date: Fri Jul 1 20:54:01 2005 +0000 184244 184245 Adding initial build system. 184246 184247commit a822df1cc16d150614dead70fd00750095a05c35 184248Author: Daniel Stone <daniel@fooishbar.org> 184249Date: Fri Jul 1 20:49:35 2005 +0000 184250 184251 Adding initial build system. 184252 184253commit ded56b1a74e6b3e4c48054b7e142d924b19e6104 184254Author: Daniel Stone <daniel@fooishbar.org> 184255Date: Fri Jul 1 20:29:53 2005 +0000 184256 184257 Adding initial build system. 184258 184259commit aabb868920658c9d3979dc194c6bd9702171f101 184260Author: Zack Rusin <zack@kde.org> 184261Date: Fri Jul 1 13:30:29 2005 +0000 184262 184263 removing all debugging output from the default build :) 184264 184265commit 0fa9d1fb4886c418e3d8e0886ad815513eda0633 184266Author: Zack Rusin <zack@kde.org> 184267Date: Fri Jul 1 12:24:30 2005 +0000 184268 184269 Leave debugging output for only the interested parties. 184270 184271commit 0df446ab8875430508ff51d3548955a215475084 184272Author: Zack Rusin <zack@kde.org> 184273Date: Fri Jul 1 10:39:21 2005 +0000 184274 184275 Missed this. Spotted by Thomas. 184276 184277commit b5b2a0522efd61bd99b5d5d75cdd27960cd1c7e1 184278Author: Lars Knoll <lars@trolltech.com> 184279Date: Fri Jul 1 10:05:43 2005 +0000 184280 184281 Add support for gradients and solid fills to Render. 184282 Changed the semantics of the Convolution filter a bit. It now doesn't try 184283 to normalize the filter values but leaves this to the client. This 184284 gives more reasonable behaviour in the limit where the filter 184285 parameters sum up to 0. 184286 184287commit 30c019e847adef6f7f3963df8ef1f3f994669a54 184288Author: Zack Rusin <zack@kde.org> 184289Date: Fri Jul 1 08:56:12 2005 +0000 184290 184291 Adding the new acceleration architecture: Exa. It's meant to replace XAA in 184292 the coming months. 184293 184294commit 276821605ee50e71f30dd52f2c12237fc61f288f 184295Author: David Reveman <c99drn@cs.umu.se> 184296Date: Fri Jul 1 03:14:54 2005 +0000 184297 184298 Fix return values in Xgl render texture implementation 184299 184300commit 5e381441fff411316ea202a9f85aceb0e0dcf46b 184301Author: Alex Deucher <agd5f@yahoo.com> 184302Date: Fri Jul 1 02:56:04 2005 +0000 184303 184304 - Fix Support for Philips FM1236/F tuner on ATI AIW 9600 XT (Jeff Smith) 184305 Bug 3401 184306 184307commit c4d1b4e0424d53cd0470e0e3f57c10ee8fe3d8c2 184308Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184309Date: Thu Jun 30 21:33:46 2005 +0000 184310 184311 Added another test of checkForInput for WIN32 Windows keyboard and mouse 184312 events are added to the input queue in Block- and WakupHandlers. There 184313 is no device to check if input is ready. 184314 184315commit 1230c55ac3f75f4902b51b223d40354a54d0d0ab 184316Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184317Date: Thu Jun 30 18:50:20 2005 +0000 184318 184319 Pass serverClient instead of NULL to ConfigureWindow. This should fix a 184320 crash reported by Øyvind Harboe 184321 184322commit 0929f79c1baa5b65808ab034591510906623e841 184323Author: Matthew Allum <breakfast@10.am> 184324Date: Thu Jun 30 13:39:00 2005 +0000 184325 184326 Another Xephyr focus/modifier fix 184327 184328commit b6c7afe0b9faed7025c70334d464fd75ce5c84c4 184329Author: Lars Knoll <lars@trolltech.com> 184330Date: Wed Jun 29 15:19:14 2005 +0000 184331 184332 compile 184333 184334commit d8a98454e305973dd7fec76db2ef80705cf7c298 184335Author: Lars Knoll <lars@trolltech.com> 184336Date: Wed Jun 29 11:57:16 2005 +0000 184337 184338 Add support for gradients and solid fills to Render. 184339 Changed the semantics of the Convolution filter a bit. It now doesn't try 184340 to normalize the filter values but leaves this to the client. This 184341 gives more reasonable behaviour in the limit where the filter 184342 parameters sum up to 0. 184343 184344commit ce0e11aeac76119b96b463605bc1f5318e3d2bde 184345Author: Adam Jackson <ajax@nwnk.net> 184346Date: Tue Jun 28 21:05:31 2005 +0000 184347 184348 Bug #2447: Fix argument order to xf86DrvMsgVerb. (Luc Verhaegen) 184349 184350commit 5ef5aec9bb5ee85295c0913afca891572d1315a4 184351Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 184352Date: Tue Jun 28 02:04:54 2005 +0000 184353 184354 Update to latest snapshot (27-May-2005) from http://pciids.sf.net/ 184355 184356commit b262a18aad36b2de729f6d00d144ac7277687f19 184357Author: Adam Jackson <ajax@nwnk.net> 184358Date: Sun Jun 26 02:48:36 2005 +0000 184359 184360 Cosmetic correctness fixes: miEmptyBox and miEmptyData are variables, and 184361 miGlyphExtents is a function. 184362 184363commit aa7fb99bc76e62036c73ff50f58337558859b814 184364Author: Adam Jackson <ajax@nwnk.net> 184365Date: Sat Jun 25 21:28:48 2005 +0000 184366 184367 Bug #3030: Fix Xnest keyboard state handling. (Mark McLoughlin) 184368 184369commit 8562f800b879ae461317da9640961f753e107250 184370Author: Adam Jackson <ajax@nwnk.net> 184371Date: Sat Jun 25 21:16:54 2005 +0000 184372 184373 Bug #3626: _X_EXPORT tags for video and input drivers. 184374 184375commit 582a9f0d2ec01f1a3c5625e2f45a4599be7a11d5 184376Author: Zack Rusin <zack@kde.org> 184377Date: Sat Jun 25 12:39:58 2005 +0000 184378 184379 Correctly handle empty rects on region initialization. 184380 184381commit 56201222067e793a3542bcdcd39bb257b4fad2a6 184382Author: Matthew Allum <breakfast@10.am> 184383Date: Thu Jun 23 16:50:07 2005 +0000 184384 184385 Fix issues with focus in and modifiers from host confusing Xephr 184386 184387commit a668b6c11a2d6b4800407ad918481d90be87a732 184388Author: Matthew Allum <breakfast@10.am> 184389Date: Thu Jun 23 16:34:07 2005 +0000 184390 184391 redo tslib 184392 184393commit 5e863851a61207ade1ac807bc8cff7d9cf02dbbe 184394Author: Adam Jackson <ajax@nwnk.net> 184395Date: Sun Jun 19 01:30:29 2005 +0000 184396 184397 Build workaround. Add glcontextmodes.[ch] from Mesa. 184398 184399commit 9743adbe94f29417818b90e18aebb96a72f332f0 184400Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 184401Date: Thu Jun 16 20:50:12 2005 +0000 184402 184403 Add Type1 subdirectory to lib/Xfonts, update build system accordingly. 184404 184405commit 68e856ff5b1842ba10421714b5f6b21d528f6071 184406Author: Ian Romanick <idr@us.ibm.com> 184407Date: Wed Jun 15 18:31:52 2005 +0000 184408 184409 DRM 20050615 import 184410 184411commit 6e301a8e97e99d58eaed25453f66c6d73bafd460 184412Author: Ian Romanick <idr@us.ibm.com> 184413Date: Wed Jun 15 18:31:52 2005 +0000 184414 184415 Initial revision 184416 184417commit f0c76610b72a5b54bae5a5eb51ff4c420a27320e 184418Author: Daniel Stone <daniel@fooishbar.org> 184419Date: Wed Jun 15 16:46:59 2005 +0000 184420 184421 Conditionalise another use of XEvIE. 184422 184423commit 28ee3dd955302a1e99ffdc66cf0f512c6234d043 184424Author: Daniel Stone <daniel@fooishbar.org> 184425Date: Wed Jun 15 16:27:16 2005 +0000 184426 184427 Conditionalise building of XEvIE code with #ifdef XEVIE. 184428 184429commit c6166ee74137084775c7550b708c5f71f16e7d3f 184430Author: Lars Knoll <lars@trolltech.com> 184431Date: Wed Jun 15 14:51:12 2005 +0000 184432 184433 Fix projective transformations in fbcompose.c Bugfix for convolution 184434 filters 184435 184436commit bd54b96034e640f202821eac0a2c40e66c1ddfd0 184437Author: David Reveman <c99drn@cs.umu.se> 184438Date: Wed Jun 15 05:50:15 2005 +0000 184439 184440 Shut up compiler 184441 184442commit 7fa782dfd548498474830c7268032ffc5fb406a5 184443Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 184444Date: Mon Jun 13 21:51:46 2005 +0000 184445 184446 Add xkbfile to symlink.sh, conditionally include "config.h" in 184447 xc/lib/xkbfile 184448 184449commit 0802a2824c06d294ea42f0f6256644da5c0038c2 184450Author: Eric Anholt <anholt@freebsd.org> 184451Date: Mon Jun 13 18:14:53 2005 +0000 184452 184453 Correct the CHIPSET lines for the 3dfx cards so the tdfx driver will 184454 recognize them. I'm not sure if this file is supposed to be maintained 184455 still, but I'd like to close FreeBSD ports/32121 which has been around 184456 for far too long. 184457 184458commit 2eab094816726542c4de6c9db5efa102ab1e1593 184459Author: David Reveman <c99drn@cs.umu.se> 184460Date: Mon Jun 13 16:38:06 2005 +0000 184461 184462 Build fixes 184463 184464commit 49476ca73c86a599a1bc49cba1117d42f59996a8 184465Author: Lars Knoll <lars@trolltech.com> 184466Date: Mon Jun 13 14:40:25 2005 +0000 184467 184468 add the convolution filter from xserver to xorg 184469 184470commit f0ab6d57df66da5de1a8182f8250cc2c8e1450ad 184471Author: David Reveman <c99drn@cs.umu.se> 184472Date: Mon Jun 13 06:00:35 2005 +0000 184473 184474 Remove xgloffscreen.c 184475 184476commit f5aeaa7710de5ba4aad125bc8472bad1f17e23c6 184477Author: Adam Jackson <ajax@nwnk.net> 184478Date: Mon Jun 13 00:09:23 2005 +0000 184479 184480 Bug #3513: Silence unhandled event messages from Xnest when running with 184481 -parent. (Mark McLoughlin) 184482 184483commit 521916d0074901db58ab6b9edab52373da28bdb3 184484Author: David Reveman <c99drn@cs.umu.se> 184485Date: Fri Jun 10 12:30:39 2005 +0000 184486 184487 New implementation of GLX_MESA_render_texture 184488 184489commit bdb3eb86f02e233b6aeef0995ea9afeaa7b035ac 184490Author: Kean Johnson <kean@armory.com> 184491Date: Fri Jun 10 06:54:07 2005 +0000 184492 184493 file usl_xqueue.c was initially added on branch sco_port_update. 184494 184495commit 988ffddfe082fb27fadf9aa60ab22dce6855508c 184496Author: Adam Jackson <ajax@nwnk.net> 184497Date: Fri Jun 10 04:01:14 2005 +0000 184498 184499 Bug #2799: Input shape. (Keith Packard) 184500 184501commit d24ed90547122832d4168ad761f68e107bb1a2db 184502Author: Eric Anholt <anholt@freebsd.org> 184503Date: Fri Jun 10 02:14:44 2005 +0000 184504 184505 Axe a few dead fields from the port priv struct and add my name to the 184506 "Copyright" line of the license to ati_video.c that already has my name 184507 in the text. 184508 184509commit 75065f3a54ec760bbe81160fa233810f14d8aaa2 184510Author: Eric Anholt <anholt@freebsd.org> 184511Date: Thu Jun 9 23:22:55 2005 +0000 184512 184513 Perform a warnings sweep on hw/kdrive. A number of these were my fault, but 184514 some come from others. 184515 184516commit e11e60b361d63ae02918dd6b43038a5c92b73a49 184517Author: Eric Anholt <anholt@freebsd.org> 184518Date: Thu Jun 9 21:59:26 2005 +0000 184519 184520 Greatly improve the correctness and performance of the MGA render 184521 implementation. Includes cache flushing to prevent bad first reads of 184522 the framebuffer, fixing blending of many formats, falling back on many 184523 unsupported operations, and falling back early to prevent migration. 184524 Passes all of rendercheck except some of the blend (!) tests. 184525 184526commit 9f81ce945680515e6db7da6c87458bee7c0f053d 184527Author: Ian Romanick <idr@us.ibm.com> 184528Date: Thu Jun 9 21:48:45 2005 +0000 184529 184530 Re-enable GL_HP_occlusion_test. The problems in Mesa that caused this bug 184531 were fixed by Brian Paul in Mesa version 6.1. The current Mesa version 184532 in the X.org tree is 6.2.1. 184533 Xorg bug: 762 184534 184535commit 92b3775ae8bdd4a84d7e101b36b306fbd3ac17a2 184536Author: Matthew Allum <breakfast@10.am> 184537Date: Thu Jun 9 16:22:27 2005 +0000 184538 184539 TSLib fixes. Add fullscreen support to ephyr 184540 184541commit 545c082cf9c86f2a809ea6b4dca33643afb0c3d3 184542Author: Eric Anholt <anholt@freebsd.org> 184543Date: Thu Jun 9 10:44:45 2005 +0000 184544 184545 - Replace the syncAccel hook in the kdrive structure with a pair of hooks 184546 in the kaa structure: markSync and waitMarker. The first, if set, 184547 returns a hardware-dependent marker number which can then be waited for 184548 with waitMarker. If markSync is absent (which is the case on all 184549 drivers currently), waitMarker must wait for idle on any given marker 184550 number. The intention is to allow for more parallelism when we get 184551 downloading from framebuffer, or more fine-grained idling. 184552 - Replace the KdMarkSync/KdCheckSync functions with kaaMarkSync and 184553 kaaWaitSync. These will need to be refined when KAA starts being smart 184554 about using them. Merge kpict.c into kasync.c since kasyn.c has all the 184555 rest of these fallback funcs. 184556 - Restructure all drivers to initialize a KaaInfo structure by hand rather 184557 than statically in dubious order. 184558 - Whack the i810 driver into shape in hopes that it'll work after this 184559 change (it certainly wouldn't have before this). Doesn't support my 184560 i845 though. 184561 - Make a new KXV helper to avoid duplicated code to fill the region with 184562 the necessary color key. Use it in i810 and mach64 (tested). 184563 184564commit 72ca8e1b5432db57401e66af8a07fcd8cbbbb9f1 184565Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 184566Date: Thu Jun 9 03:11:58 2005 +0000 184567 184568 Add agpgart support for Solaris x86/x64. [Requires Solaris "Nevada" build 184569 16 or later to get kernel side for now.] (Sophia Li - Sun Microsystems) 184570 184571commit ca37d5755259ca03c61cf0567be3dea99d0c79f6 184572Author: Adam Jackson <ajax@nwnk.net> 184573Date: Thu Jun 9 03:01:57 2005 +0000 184574 184575 Bug #2469: More accurate damage reports. (Jonathan Lennox) 184576 184577commit cdc15e2294a9bffc570e33bc31170081abfc55fb 184578Author: Adam Jackson <ajax@nwnk.net> 184579Date: Thu Jun 9 02:29:42 2005 +0000 184580 184581 Bug #1846: Add intentionally undocumented -disablexineramaextension flag to 184582 the server to work around ignorant clients on large display walls. 184583 (Kevin E. Martin) 184584 184585commit e3cdec7cdcd76f6294ba1f296e4bcdee43b1eb3c 184586Author: Adam Jackson <ajax@nwnk.net> 184587Date: Thu Jun 9 02:19:10 2005 +0000 184588 184589 Bug #1880: Remove unused xnestConfineWindow. (Mark McLoughlin) 184590 184591commit 0f7136191b54e587b66958985e14d8b0687c5ee1 184592Author: Adam Jackson <ajax@nwnk.net> 184593Date: Thu Jun 9 02:03:50 2005 +0000 184594 184595 Bug #3434: Don't define fbAddTraps twice. (Mike A. Harris) 184596 184597commit e7fec996eb43a0bea94a37d329e3902299e6e895 184598Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184599Date: Wed Jun 8 16:13:11 2005 +0000 184600 184601 Merge from CYGWIN Added layout "French (Canada)" as ca_enhanced Added Czech 184602 (QWERTY) layout 184603 Merge from CYGWIN Print error code if winStoreColorsShadowDDNL fails 184604 184605commit d0d164e6daf6c1d8afe8099fb6187d8a74ff652b 184606Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184607Date: Wed Jun 8 09:01:41 2005 +0000 184608 184609 Fix crash reported by �yvind Harboe 184610 184611commit d146c41882e50dd1ed6022f8a5866514067ba9d2 184612Author: David Reveman <c99drn@cs.umu.se> 184613Date: Tue Jun 7 08:53:40 2005 +0000 184614 184615 GLX improvements and remove use of pbuffers for offscreen rendering in Xgl 184616 184617commit 8237b31f60245a8e67735d6c231dfb7dd0bbc388 184618Author: Jon Smirl <jonsmirl@yahoo.com> 184619Date: Mon Jun 6 04:06:38 2005 +0000 184620 184621 Keyboard sort of works now with evdev 184622 184623commit c5026577cbc70c7365a9c81b2775136b45c850d2 184624Author: Jon Smirl <jonsmirl@yahoo.com> 184625Date: Sun Jun 5 17:32:41 2005 +0000 184626 184627 Create a default colormap 184628 184629commit 021a5cd170051aae25cf61651d07b47a4df31ab4 184630Author: Jon Smirl <jonsmirl@yahoo.com> 184631Date: Sun Jun 5 01:58:41 2005 +0000 184632 184633 Enable egl server in automake 184634 --enable-xeglserver 184635 184636commit aa959672863aad71c807d6c09938bb296c347bcf 184637Author: Jon Smirl <jonsmirl@yahoo.com> 184638Date: Sat Jun 4 23:33:03 2005 +0000 184639 184640 Initial Xegl check-in. Works on EGL fbdev driver Input is totally broken. 184641 184642commit 88d49bed008264092245c4f6c53174f93f6ab387 184643Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 184644Date: Fri Jun 3 21:49:11 2005 +0000 184645 184646 Need to set initial screen size at device open time too. 184647 184648commit 0d206e177f46eedfffd8f88e985efdfaf117021c 184649Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184650Date: Fri Jun 3 08:12:22 2005 +0000 184651 184652 Backout last winRaiseWindow patch which caused stacking problems 184653 184654commit 2fc290f646eb48f3c20ccff98355c2d082569160 184655Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 184656Date: Wed Jun 1 17:45:17 2005 +0000 184657 184658 Solaris VUID mouse protocol updates: 184659 - support for mouse-like devices that report absolute coordinates 184660 - support for mouse-like devices that need to know screen size so they can 184661 scale absolute coordinates to screen size 184662 - fix 3-button emulation 184663 184664commit 8537146e3662cdc66ef4692bfd3886a676096fb8 184665Author: Adam Jackson <ajax@nwnk.net> 184666Date: Mon May 30 18:44:55 2005 +0000 184667 184668 Bug #3368: Cosmetic cleanup to DIX export list. 184669 184670commit 547c47aaea9f31f7c873376b0db859c1c1d3d918 184671Author: Adam Jackson <ajax@nwnk.net> 184672Date: Mon May 30 18:36:19 2005 +0000 184673 184674 Bug #3407: Clean out ancient #ifdef X11R5 conditionals. 184675 184676commit cebd7348d271591b7ebeebe3b332f835cfa8a068 184677Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 184678Date: Sat May 28 00:08:03 2005 +0000 184679 184680 Change hardcoded /usr/X11R6 to PROJECTROOT. Change hardcoded XF86Config to 184681 XCONFIGFILE. 184682 184683commit 1cb18a4bad565d1f783a4cefb3ed467699051068 184684Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184685Date: Wed May 25 12:14:29 2005 +0000 184686 184687 Workaround bug in pthread.h 184688 184689commit 62343f5162066f19ca6e62d1c85a4a40d45b3295 184690Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 184691Date: Sun May 22 01:12:49 2005 +0000 184692 184693 Bugzilla #2800 <https://bugs.freedesktop.org/show_bug.cgi?id=2800> Xevie 184694 extension crash with signal 11 on keyboard Bugzilla #1205 184695 <https://bugs.freedesktop.org/show_bug.cgi?id=1205> Xevie client 184696 receives two KeyPress events on consumed keys when XKB is enabled 184697 Patch #2223 <https://bugs.freedesktop.org/attachment.cgi?id=2223> Fixes for 184698 both of these and some other Xevie bugs (Derek Wang - Sun Microsystems) 184699 184700commit a8a61bbe22361b12d4a2dd511894987a338e3eef 184701Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 184702Date: Sat May 21 07:46:38 2005 +0000 184703 184704 Initial experimental support for AMD64 builds on Solaris 10 x86. Improved 184705 support for 64-bit SPARC builds on Solaris as well. 184706 184707commit ef4401ff8a8c4d7c22ef6af7bf47ccb24f83967f 184708Author: Adam Jackson <ajax@nwnk.net> 184709Date: Thu May 19 18:53:51 2005 +0000 184710 184711 Bug #3334: Clean up remnants from the PEX, XIE, and BEZIER extensions. 184712 184713commit 1e51e18382de61a67027759eccee9ccbb0924304 184714Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 184715Date: Thu May 19 01:25:02 2005 +0000 184716 184717 Wed May 18 21:20:35 2005 Søren Sandmann <sandmann@redhat.com> 184718 s/dst/src that I missed in the previous patch. Pointed out by Owen Taylor. 184719 184720commit be3875994d7778ffad0cd957a4bd1500bf58ac2c 184721Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 184722Date: Wed May 18 20:49:31 2005 +0000 184723 184724 Wed May 18 16:47:44 2005 Søren Sandmann <sandmann@redhat.com> 184725 Actually assign the computed results. (Reported by Michael Dänzer, patch 184726 by Owen Taylor). 184727 184728commit 07d1e9ae5f0da65427ac52f065b173d340d70d6d 184729Author: Adam Jackson <ajax@nwnk.net> 184730Date: Wed May 18 19:44:35 2005 +0000 184731 184732 Bug #3066: Promote frameBufferPhysicalAddress to pointer from CARD32; bump 184733 DRIINFO version to match. Fix various pointer printfs in libdri to be 184734 64-bit aware. Silence a warning about redefining noXFree86DRIExtension. 184735 (Jesse Barnes) 184736 184737commit b87a32fb0eafa6d5f12a6faa634662c71fdc97fa 184738Author: Adam Jackson <ajax@nwnk.net> 184739Date: Wed May 18 17:47:35 2005 +0000 184740 184741 Bug #3163: Create new DRIINFO_*_VERSION macros to indicate the version 184742 number of the DRIInfoRec, disambiguating it from the XF86DRI protocol 184743 version number. Modify DRIQueryVersion to return the libdri version 184744 number, which all DDXes were requesting implicitly. Fix the DDXes to 184745 check for the DRIINFO version they were compiled against. 184746 184747commit fddcde777f50611099d8888456d93be7e566c84e 184748Author: Egbert Eich <eich@suse.de> 184749Date: Wed May 18 15:38:50 2005 +0000 184750 184751 VBESetVBEMode() calls int10 0x4f02 to set a VBE video mode. This can be a 184752 standard mode solely determined by the mode ID or it may be a detailed 184753 mode with almost arbitrary mode parameters. The mode parameters are 184754 passed to the BIOS in a memory area pointed to by es:di. If bit 11 in 184755 the video mode is set this memory area is evaluated. When we don't have 184756 such an area (passed to VBESetVBEMode() this function should unset this 184757 bit for sanity reasons. (Bugzilla #3329) 184758 184759commit f28d644d9fb1feeab1bbb3d5e68255bfa20905d8 184760Author: Zack Rusin <zack@kde.org> 184761Date: Wed May 18 14:24:14 2005 +0000 184762 184763 rewrite the rootless compositing code (1) compiles which is already a huge 184764 improvement ;) , 2) matches the new render composition semantics) 184765 184766commit bbbb0a3a84bf21f5120c73ef387d6190154c0f65 184767Author: Egbert Eich <eich@suse.de> 184768Date: Wed May 18 10:31:53 2005 +0000 184769 184770 Avoid segfaults if the number of user supplied monitor ranges exceed the 184771 number of preallocated slots. We should really make this dynamic - but 184772 I don't think this ever caused a problem so it's more or less academic. 184773 A. Avoid that *SyncStart starts before *BlankStart. If *BlankStart > 184774 *SyncStart it is made = *SyncStart and its width is made maximal but such 184775 that the blank does not exceed *Total. Since the Sync width has the 184776 same restrictions as the Blank width monitors should still be able to 184777 clamp after the sync pulse. B. Over time mode validation has become 184778 inconsistent when people started to add additional features to the mode 184779 validation. One such feature is that the mode->Crtc* values have been 184780 (ab)used to allow the driver ValidMode() function to pass driver 184781 normalized timing values back to the validation function. The 184782 introduction of these features made the code less readable and created 184783 numerous possibly unintended side effects in the validation semantics. 184784 I've attempted to consolidate these changes making the code more 184785 consistent and eliminating a number of side effects. This should not 184786 cause problems for the majority of drivers, still it should receive 184787 testing - especially with ATi Mach64 and Radeon code. (Bugzilla #3325). 184788 184789commit a90af4a2e6e38cbe20af13aaa7822836c01290ae 184790Author: Thomas Hellstrom <unichrome@shipmail.org> 184791Date: Sun May 15 17:45:47 2005 +0000 184792 184793 Bug 2750: Prevent mtrr_remove_offending from ending up in an endless loop 184794 if the offending region refuses to be removed. 184795 184796commit f58a54668b96884ece6ecbff732e880677d5d19e 184797Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 184798Date: Sat May 14 20:13:45 2005 +0000 184799 184800 Change return statements to fix compiler errors: "fbcompose.c", line 2815: 184801 void function cannot return value "fbcompose.c", line 2861: void 184802 function cannot return value 184803 184804commit a9d820b454bbb418d04e1d3c3d5d4dd10d7bb3af 184805Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 184806Date: Fri May 13 22:53:44 2005 +0000 184807 184808 - For now put xtrans in X11/Xtrans/X11, since libX11 is looking for it in 184809 <X11/...> 184810 - For Xcomposite and Xdamage, don't link the build system out of the xc 184811 tree 184812 - Link the public X11 headers into their own directory 184813 - Add links to XKeysymDB and XErrorDB 184814 - Add links to all the Xlib man pages 184815 - Add links to the lcUniConv subdirectory 184816 - Conditionally include config.h in Xlib source 184817 184818commit 8993e70d8da89e130455369ccb504b3adea5344f 184819Author: Eric Anholt <anholt@freebsd.org> 184820Date: Tue May 10 00:12:17 2005 +0000 184821 184822 Bugzilla #2561: Disable some keycode translations on PC98, which had 184823 various bad effects including modifier keys not working. 184824 184825commit e78abf3bdfbf5fc13fbc5b1d3ec00e9484297b89 184826Author: Eric Anholt <anholt@freebsd.org> 184827Date: Mon May 9 23:42:26 2005 +0000 184828 184829 Bugzilla #2429: Correct the sysarch prototype for FreeBSD/alpha. 184830 184831commit 3f1a1551df2a9f45b3b8821f3f168d4dc8b8b438 184832Author: Eric Anholt <anholt@freebsd.org> 184833Date: Mon May 9 23:26:02 2005 +0000 184834 184835 Increase the maximum number of buttons from 12 to 24. Helps with 184836 button-happy mice like the Logitech MediaPlay. (Bugzilla #2390) 184837 184838commit eca5dff173d5af0a31bbf84579909b88a86e4c92 184839Author: Adam Jackson <ajax@nwnk.net> 184840Date: Mon May 9 12:46:53 2005 +0000 184841 184842 Render performance improvements. (Lars Knoll, Zack Rusin) 184843 184844commit 2de24db63eb65974ac547facf2a99aa4709d54b3 184845Author: Adam Jackson <ajax@nwnk.net> 184846Date: Sun May 8 23:34:15 2005 +0000 184847 184848 Render performance improvements. (Lars Knoll, Zack Rusin) 184849 184850commit 2c9b1e337b2e82e10909f62e8cd8c2c7a402fdd8 184851Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184852Date: Sun May 8 21:14:55 2005 +0000 184853 184854 Only call ConfigureWindow from winRaiseWindow if the windows message 184855 dispatch loop is running. 184856 184857commit 78e4cb67d0f595f4fba5e98a9fa46de044556905 184858Author: Torrey Lyons <torrey@mrcla.com> 184859Date: Fri May 6 00:45:30 2005 +0000 184860 184861 Fix build issues on Mac OS X 10.4.0. 184862 184863commit d5739efd2c266081ed6b844767dcdd0a3331bfd3 184864Author: Eric Anholt <anholt@freebsd.org> 184865Date: Wed May 4 04:14:58 2005 +0000 184866 184867 Port X.Org to FreeBSD/powerpc. This makes a bit of math for sysmouse in 184868 mouse.c explicitly signed, avoiding the need for -fsigned-chars. (Peter 184869 Grehan, grehan at FreeBSD dot org) 184870 184871commit 6700847458427cbdbaf837ab06bfea9b80d92df2 184872Author: Harold L Hunt II <huntharo@msu.edu> 184873Date: Mon May 2 22:01:08 2005 +0000 184874 184875 Fix comments for pointers in parameter lists to work with fussy compilers 184876 184877commit e50ab8feade400efd8e88ee0b2deeb924f169034 184878Author: Harold L Hunt II <huntharo@msu.edu> 184879Date: Mon May 2 21:57:32 2005 +0000 184880 184881 Fix message type (respose->response) and fix */*comment*/ blocks to work 184882 with fussy compilers. 184883 184884commit baa0cfc15442287557e44fa2614d89ab0b5c2539 184885Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 184886Date: Mon May 2 14:07:31 2005 +0000 184887 184888 Print correct logfile in FatalError message 184889 184890commit bc977945a53206d70ef9817d6ede4137eb5f7e3c 184891Author: David Reveman <c99drn@cs.umu.se> 184892Date: Mon May 2 00:33:52 2005 +0000 184893 184894 Set stencil size for Xgl GLX visuals 184895 184896commit 5b4462ed9bbb17212dd57617cb7e837142b6270a 184897Author: David Reveman <c99drn@cs.umu.se> 184898Date: Sun May 1 22:55:25 2005 +0000 184899 184900 New algorithm for calculating glyph extents in Xgl 184901 184902commit e4b33f4e91a5379c99ecf78ffb6a3e836cb42491 184903Author: Egbert Eich <eich@suse.de> 184904Date: Wed Apr 27 15:42:15 2005 +0000 184905 184906 Catch SIGCHLD in OsBlockSignals() too to make sure this signal doesn't 184907 intercept reading the authority file (Fabian Franz, Bugzilla #3137). 184908 184909commit 246b14cb29250517912d9c661ab037a953f4bbf4 184910Author: David Reveman <c99drn@cs.umu.se> 184911Date: Wed Apr 27 10:34:33 2005 +0000 184912 184913 Minor optimization 184914 184915commit 28a2d841cee596c0242b1649587d6b180529c0ef 184916Author: David Reveman <c99drn@cs.umu.se> 184917Date: Wed Apr 27 09:29:33 2005 +0000 184918 184919 Line drawing improvements to Xgl 184920 184921commit abcc8c352e5fe6dd3b7ce5c6a25f435f877264c3 184922Author: David Reveman <c99drn@cs.umu.se> 184923Date: Wed Apr 27 08:45:16 2005 +0000 184924 184925 Fix typo 184926 184927commit 8bc5a387c365d6367813eac374150440d0edbf6e 184928Author: David Reveman <c99drn@cs.umu.se> 184929Date: Tue Apr 26 11:04:23 2005 +0000 184930 184931 Add glCopy operation and glPushAttrib/glPopAttrib support to Xgl GLX 184932 184933commit f010131a1964b5ec35f3b333ffa7459c38f8ce45 184934Author: Roland Mainz <roland.mainz@nrubsig.org> 184935Date: Tue Apr 26 06:12:07 2005 +0000 184936 184937 xc/programs/Xserver/Xprint/Init.c 184938 //bugs.freedesktop.org/show_bug.cgi?id=2879) attachment #2556 184939 (https://bugs.freedesktop.org/attachment.cgi?id=2556) Refix for bug 184940 2879 - downgrade 15bit PseudoColor to 14bit - the current datatype for 184941 the |ColormapEntries| is a |signed short| which is too small for 184942 |32768| colors (=integer overflow). 184943 184944commit ba24ae89d33fbf3aacb8bbaf920d7436b77fda46 184945Author: David Reveman <c99drn@cs.umu.se> 184946Date: Mon Apr 25 17:18:01 2005 +0000 184947 184948 Add scissor based clipping to Xgl GLX 184949 184950commit 07cc29cf69ff1e079efe3c9bfc55e8ac0f9bac93 184951Author: Adam Jackson <ajax@nwnk.net> 184952Date: Mon Apr 25 00:25:39 2005 +0000 184953 184954 Bug #2138: When the server is built with MakeDllModules YES, prefer 184955 dlloader modules to elfloader modules, and vice versa when 184956 MakeDllModules is NO. Based on 028_loader_speed_hack.diff from Ubuntu 184957 (Daniel Stone). 184958 184959commit 6c37648754c9bd901adecf8d38f9bb46db65efad 184960Author: Adam Jackson <ajax@nwnk.net> 184961Date: Mon Apr 25 00:11:21 2005 +0000 184962 184963 Bug #2141: Rework misleading warning message when APM support is 184964 unavailable. (Previous patch on 2005-04-14 changed the wrong message.) 184965 184966commit a369d390a87ab0189c465be6cfd914e4b9329691 184967Author: Roland Mainz <roland.mainz@nrubsig.org> 184968Date: Sun Apr 24 01:10:12 2005 +0000 184969 184970 xc/config/cf/X11.tmpl 184971 xc/programs/Xserver/Xext/Imakefile 184972 xc/programs/Xserver/dix/Imakefile 184973 xc/programs/Xserver/dix/main.c 184974 xc/programs/Xserver/dix/xpstubs.c 184975 xc/programs/Xserver/mi/miinitext.c 184976 //bugs.freedesktop.org/show_bug.cgi?id=2792) attachment #2526 184977 (https://bugs.freedesktop.org/attachment.cgi?id=2526) bug 2792 part II: 184978 Make Xprint headers in dix/, mi/, os/ and Xext/ conditional on whether 184979 the Xprint extension is build or not. Patch by Egbert Eich 184980 <eich@freedesktop.org> and Roland Mainz <roland.mainz@nrubsig.org>. 184981 184982commit 13fcfee37305f46e95ff81d7d5eec4d88a63a63b 184983Author: Roland Mainz <roland.mainz@nrubsig.org> 184984Date: Sat Apr 23 22:55:40 2005 +0000 184985 184986 //bugs.freedesktop.org/show_bug.cgi?id=3118) attachment #2525 184987 (https://bugs.freedesktop.org/attachment.cgi?id=2525) Get Xprint server 184988 working again which was broken by the 184989 05/04/20 05:49:46 commit commit (the CPP symbol |XPRINT| is not universally 184990 available when building the Xserver and should be used with care). 184991 184992commit 39eb2797183bf871a2dd01bff63dd35c146471f5 184993Author: Adam Jackson <ajax@nwnk.net> 184994Date: Sat Apr 23 19:16:10 2005 +0000 184995 184996 Bug #3016: Don't complain as loudly about failing to load a module that's 184997 already loaded. 184998 184999commit 139f900c56a6f44df2b411fe4bbc025a147e3abe 185000Author: Adam Jackson <ajax@nwnk.net> 185001Date: Sat Apr 23 19:01:13 2005 +0000 185002 185003 Bug #3109: Handle R_ALPHA_BRSGP relocations in elfloader. (Sergey Tikhonov) 185004 185005commit 0e80fe0e607cc73856332563becd70ef92d1aa75 185006Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 185007Date: Fri Apr 22 22:11:26 2005 +0000 185008 185009 Clean up formatting, projectroot substutition, and grammatical nits. 185010 185011commit dcfb97204b021738c0ee9f1f8f40243dfa0b2ce0 185012Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 185013Date: Fri Apr 22 22:04:37 2005 +0000 185014 185015 Work around clash between Solaris headers and module loader headers over 185016 definition of FILE. 185017 185018commit 07bd7df6a6a07834277b4bf505db6727841e1153 185019Author: Adam Jackson <ajax@nwnk.net> 185020Date: Fri Apr 22 20:49:50 2005 +0000 185021 185022 Bug #3069: Drop the BuildLowMem hack, it doesn't compile and isn't useful. 185023 185024commit d450a70e00b50427ecb2065d3cc44f43d102cade 185025Author: Adam Jackson <ajax@nwnk.net> 185026Date: Fri Apr 22 17:45:14 2005 +0000 185027 185028 Bug #2373: SGI Altix platform support. (Shrijeet Mukherjee, Jesse Barnes, 185029 Bjorn Helgaas, Egbert Eich.) 185030 185031commit 16c2499b8f5c2405e36c7d5a922bb0b150df1762 185032Author: Adam Jackson <ajax@nwnk.net> 185033Date: Fri Apr 22 16:49:22 2005 +0000 185034 185035 Bug #2373: SGI Altix platform support. (Shrijeet Mukherjee, Jesse Barnes, 185036 Bjorn Helgaas, Egbert Eich.) 185037 185038commit 8565b6c0e2851cc3f194ba72d3db02a4e2976528 185039Author: Daniel Stone <daniel@fooishbar.org> 185040Date: Thu Apr 21 00:31:31 2005 +0000 185041 185042 Change xf86bigfont.h include to X11/extensions/xf86bigfont.h. 185043 185044commit b241c703a2c4164420dd26ee11f583bbf9cfe0f3 185045Author: Daniel Stone <daniel@fooishbar.org> 185046Date: Thu Apr 21 00:31:13 2005 +0000 185047 185048 Change keymap.h includes to X11/keymap.h 185049 185050commit 44528218d5ca75b842e4cefd8fdc58be2f35f0f4 185051Author: Daniel Stone <daniel@fooishbar.org> 185052Date: Wed Apr 20 23:33:53 2005 +0000 185053 185054 Change dmxext.h and dmxproto.h to <X11/extensions/...>. 185055 185056commit efa9d5f4757bfc0588cee361bcc78dd8a09efa62 185057Author: Daniel Stone <daniel@fooishbar.org> 185058Date: Wed Apr 20 23:11:12 2005 +0000 185059 185060 Change "eviestr.h" to <X11/extensions/eviestr.h>. 185061 185062commit db5bd04097fd815ab6523f187679682a5e5047fa 185063Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185064Date: Wed Apr 20 16:40:52 2005 +0000 185065 185066 Fix includes right throughout the Xserver tree: 185067 apply changes to windows specific includes 185068 Fix includes right throughout the Xserver tree: 185069 apply changes to Xdmcp.h 185070 185071commit 8963a220f36cf0ae2a8a653fd39c983140e29736 185072Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185073Date: Wed Apr 20 16:34:46 2005 +0000 185074 185075 warning fix for Win32 185076 185077commit e6a0820d1b479058bddef66018d321940e79260c 185078Author: Daniel Stone <daniel@fooishbar.org> 185079Date: Wed Apr 20 15:16:36 2005 +0000 185080 185081 Change "xf86bigfstr.h" to <X11/extensions/xf86bigfstr.h> for includes. 185082 185083commit 025724f9f5874159c20ebd705288ec60b960caac 185084Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185085Date: Wed Apr 20 14:17:35 2005 +0000 185086 185087 Add missing space after -query hostname 185088 185089commit 35cd1684622f4528c68e07eea798c33bc4b93667 185090Author: Daniel Stone <daniel@fooishbar.org> 185091Date: Wed Apr 20 14:16:37 2005 +0000 185092 185093 Change xf86bigfstr.h to X11/extensions/xf86bigfstr.h for includes. 185094 185095commit c7c27e61870b42de044b183c854a960582d98dbf 185096Author: Daniel Stone <daniel@fooishbar.org> 185097Date: Wed Apr 20 13:33:54 2005 +0000 185098 185099 Change xf86bigfont.h to X11/extensions/xf86bigfont.h for includes. 185100 185101commit abd246c6c272a2c6f9c37404b2ed439911880e75 185102Author: Daniel Stone <daniel@fooishbar.org> 185103Date: Wed Apr 20 13:01:55 2005 +0000 185104 185105 Add glyphstr.h to includes. 185106 185107commit 956dfa22f6076969776546fb1151e900d8d773a1 185108Author: Daniel Stone <daniel@fooishbar.org> 185109Date: Wed Apr 20 12:49:46 2005 +0000 185110 185111 Conditionalise usage of Xprint functions and headers. 185112 185113commit 2cdfab0ed7eb33a6a50f9b7ec212b498dd8318b5 185114Author: Daniel Stone <daniel@fooishbar.org> 185115Date: Wed Apr 20 12:42:02 2005 +0000 185116 185117 Change keysymdef.h to X11/keysymdef.h for include statement. 185118 185119commit fa5539247d3b246db9ff1469d08167178c85d7ad 185120Author: Daniel Stone <daniel@fooishbar.org> 185121Date: Wed Apr 20 12:39:28 2005 +0000 185122 185123 Change Xalloca.h to X11/Xalloca.h for include. 185124 185125commit 292c4cff26687e6ef86c285b97813ab587daf009 185126Author: Daniel Stone <daniel@fooishbar.org> 185127Date: Wed Apr 20 12:25:48 2005 +0000 185128 185129 Fix includes right throughout the Xserver tree: 185130 change "foo.h" to <X11/foo.h> for core headers, e.g. X.h, Xpoll.h; 185131 change "foo.h", "extensions/foo.h" and "X11/foo.h" to 185132 <X11/extensions/foo.h> for extension headers, e.g. Xv.h; 185133 change "foo.[ch]" to <X11/Xtrans/foo.[ch]> for Xtrans files. 185134 185135commit c062d7f96f47bdd31640be1fbce682d0774db3d9 185136Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185137Date: Tue Apr 19 18:21:01 2005 +0000 185138 185139 Prevent recursive calls to winRaiseWindow. 185140 185141commit 4e914c5ed7679a1102f3e25af0c087380f834865 185142Author: David Reveman <c99drn@cs.umu.se> 185143Date: Tue Apr 19 14:51:29 2005 +0000 185144 185145 Hash texture objects and display lists in Xgl 185146 185147commit 6bde764de102a56d2c71b971eaa36535e9760b52 185148Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 185149Date: Sun Apr 17 23:02:25 2005 +0000 185150 185151 Added a log pring for the parameters 185152 185153commit 8594b8f2893e58ae824e140334c18ba3d7467217 185154Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 185155Date: Sun Apr 17 23:00:58 2005 +0000 185156 185157 Added new tuner FM1236MK3 PAL version 185158 185159commit 7f74e3aebdf79fbca4141e6ffcdad39812df9335 185160Author: Bogdan Diaconescu <b_diaconescu@yahoo.com> 185161Date: Sun Apr 17 22:58:03 2005 +0000 185162 185163 This is the UDA1380 sound coder-decoder module 185164 185165commit 7a4e5f4006319e025e3ff561eccc3f1ad6c661a0 185166Author: Torrey Lyons <torrey@mrcla.com> 185167Date: Sat Apr 16 00:21:21 2005 +0000 185168 185169 Fix cases in Darwin build where a variable is declared static and later as 185170 extern (XFree86 Bug #1576, Jordan Frank). 185171 185172commit 4f686f158b00478a3d074128f9e4cb6dc0d928cd 185173Author: Roland Mainz <roland.mainz@nrubsig.org> 185174Date: Fri Apr 15 23:34:39 2005 +0000 185175 185176 //bugs.freedesktop.org/show_bug.cgi?id=2885) attachment #2434 185177 (https://bugs.freedesktop.org/attachment.cgi?id=2434) Fix Solaris build 185178 bustage caused by namespace collision between symbols defined in 185179 <sys/kbd.h> and those in "xf86_OSlib.h". Patch by Alan Coopersmith 185180 <alan.coopersmith@sun.com>. 185181 185182commit 7472fcfdd40e29cd2847e45d4bd2dd11ccc41ff5 185183Author: Adam Jackson <ajax@nwnk.net> 185184Date: Fri Apr 15 00:18:58 2005 +0000 185185 185186 Bug #2141: Rework misleading warning message when APM support is 185187 unavailable. 185188 185189commit 504067819a4f1a8564dcacc278933f533618b666 185190Author: Adam Jackson <ajax@nwnk.net> 185191Date: Thu Apr 14 17:51:51 2005 +0000 185192 185193 Bug #3025: gcc4 build fix. 185194 185195commit e40db7f26af39a8b1f3675a2c87ce90c4fd59d85 185196Author: David Reveman <c99drn@cs.umu.se> 185197Date: Wed Apr 13 14:27:47 2005 +0000 185198 185199 Add GLX code to Xgl 185200 185201commit ddfa6f00da7c80b246b57c592361baa4bc5a8e9d 185202Author: Roland Mainz <roland.mainz@nrubsig.org> 185203Date: Wed Apr 13 00:05:37 2005 +0000 185204 185205 //bugs.freedesktop.org/show_bug.cgi?id=3001) attachment #2404 185206 (https://bugs.freedesktop.org/attachment.cgi?id=2404) Improve rendering 185207 performance when glXSwapBuffers()| or |XPutImage()| are called for the 185208 PostScript DDX via optimizing the codepath around 185209 |PsOut_OutImageBytes()|. Patch by Simon Toedt <simon.toedt@gmail.com> 185210 and Roland Mainz <roland.mainz@nrubsig.org>. 185211 185212commit f45208a7e9c39345ac431c2e7da8777d6c466dcc 185213Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 185214Date: Mon Apr 11 11:06:21 2005 +0000 185215 185216 update static file from pci.ids 185217 185218commit b47f39b417cbe220690155a08c6bc18cd7cfa388 185219Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 185220Date: Mon Apr 11 10:57:33 2005 +0000 185221 185222 import latest pci.ids file 185223 185224commit 00a551393ce7aa9d2b23634737ced5071e3cdd35 185225Author: Egbert Eich <eich@suse.de> 185226Date: Mon Apr 11 10:54:13 2005 +0000 185227 185228 Reenable BackingStore in Xvfb. 185229 185230commit 1fbd38f3b46df62561f8be74cbc05fcf8ad88a21 185231Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185232Date: Mon Apr 11 10:16:25 2005 +0000 185233 185234 Added wizard page for clipboard selection and additional server parameters 185235 185236commit 6848b0353897a6a5d1218b266c22fb87a7c6d730 185237Author: Egbert Eich <eich@suse.de> 185238Date: Mon Apr 11 10:05:38 2005 +0000 185239 185240 Correct the mask bits when checking for a 64bit memory base in PCI config 185241 space (bugzilla #2963). 185242 185243commit adac7011965ed75ea399b93dae917cb48180c77a 185244Author: Egbert Eich <eich@suse.de> 185245Date: Mon Apr 11 09:33:43 2005 +0000 185246 185247 Preventing hight for drawing from becoming negative when face->dy < 0. The 185248 height value is used in miFillPolyHelper() to calculate the size of 185249 memory to be allocated. A negative value will lead to a crash (Bugzilla 185250 #2690, Keith Packard, Egbert Eich). 185251 185252commit 5e2080ef93a598c6d68e1b2f446f251ab025b702 185253Author: Roland Mainz <roland.mainz@nrubsig.org> 185254Date: Mon Apr 11 01:06:15 2005 +0000 185255 185256 xc/programs/Xserver/Xprint/attributes.c 185257 xc/programs/glxgears/glxgears.c 185258 xc/programs/xdbedizzy/xdbedizzy.c 185259 xc/programs/xedit/Imakefile 185260 xc/programs/xedit/Xedit-xprint.ad 185261 xc/programs/xedit/util.c 185262 xc/programs/xedit/xedit.h 185263 xc/programs/xlogo/print.c 185264 xc/programs/xlogo/xlogo.c 185265 xc/programs/xlogo/xlogo.h 185266 xc/programs/xman/Imakefile 185267 xc/programs/xman/print.h 185268 xc/programs/xmore/Imakefile 185269 xc/programs/xmore/print.c 185270 xc/programs/xmore/print.h 185271 xc/programs/xmore/printdialog.c 185272 xc/programs/xphelloworld/xpawhelloworld/xpawhelloworld.c 185273 xc/programs/xphelloworld/xphelloworld/xphelloworld.c 185274 xc/programs/xphelloworld/xpsimplehelloworld/xpsimplehelloworld.c 185275 xc/programs/xphelloworld/xpxmhelloworld/xpxmhelloworld.c 185276 //bugs.freedesktop.org/show_bug.cgi?id=790) attachment #2379 185277 (https://bugs.freedesktop.org/attachment.cgi?id=2379) Implement support 185278 client+Xserver support for passing output (stdout+stderr) of the 185279 spooler command started by the Xprint server back to the application 185280 using the "xp-spooler-command-results" XPJobAttr attribute 185281 (applications can fetch the attribute value after the XPEndJobNotify 185282 event was received; more details can be found in 185283 http://xprint.mozdev.org/docs/dtprint_fspec.ps). 185284 185285commit 9af443f5976ab3987e4ee9d397391e82206676b4 185286Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185287Date: Sat Apr 9 19:20:03 2005 +0000 185288 185289 Fix passing of non-RGB visuals. The old code did not initialize the 185290 structure properly which lead to a crash in 8bpp mode 185291 185292commit f02440dfa3439ab493c7918b472c23bb22e29707 185293Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185294Date: Wed Apr 6 15:18:59 2005 +0000 185295 185296 First import of xlaunch frontend for Xming 185297 185298commit de5d24a200e4426e458fc447884d1a5b0257faaa 185299Author: Egbert Eich <eich@suse.de> 185300Date: Mon Apr 4 10:17:06 2005 +0000 185301 185302 Fixed sentinels in Xt, editres and xedit to reduce number of warnings with 185303 gcc4 (Andreas Schwab). 185304 185305commit 277ff06e9999f2efe0f082a3565f6279219c13e4 185306Author: Egbert Eich <eich@suse.de> 185307Date: Mon Apr 4 10:11:51 2005 +0000 185308 185309 Fix typo leading to bogus code in xorgcfg (Andreas Schwab). 185310 185311commit 7eb6b69ebd945c753ee0988e1a85dffde982446e 185312Author: Egbert Eich <eich@suse.de> 185313Date: Mon Apr 4 09:55:25 2005 +0000 185314 185315 Support for HP's IPF ZX1 systems (Alex Williamson). 185316 185317commit 231c00e8fba91b580ec3e2703dd1ceacf13a6624 185318Author: Egbert Eich <eich@suse.de> 185319Date: Mon Apr 4 09:47:07 2005 +0000 185320 185321 When not using dlopen ia64 needs an extra cache flush to ensure the icache 185322 is coherent when modules are loaded (Alex Williamson). 185323 185324commit 55dc930a180553c08d8546cc2078451c20e34934 185325Author: Torrey Lyons <torrey@mrcla.com> 185326Date: Sat Apr 2 02:29:24 2005 +0000 185327 185328 Fix XDarwin's handling of Wacom tablet mouse buttons (Based on patch 185329 suggested by Daphne Pfister). 185330 185331commit 03d126081e5ba57ea2304753289528a896f3baaf 185332Author: Roland Mainz <roland.mainz@nrubsig.org> 185333Date: Fri Apr 1 21:45:20 2005 +0000 185334 185335 xc/programs/Xserver/Xprint/Init.c 185336 xc/programs/Xserver/Xprint/ps/Imakefile 185337 xc/programs/Xserver/Xprint/ps/Ps.h 185338 xc/programs/Xserver/Xprint/ps/PsArea.c 185339 xc/programs/Xserver/Xprint/ps/PsColor.c 185340 xc/programs/Xserver/Xprint/ps/PsImageUtil.c 185341 xc/programs/Xserver/Xprint/ps/PsInit.c 185342 //bugs.freedesktop.org/show_bug.cgi?id=2879) attachment #2287 185343 (https://bugs.freedesktop.org/attachment.cgi?id=2287) Follow-up to 185344 bugzilla #1299: Add new visuals in the Postscript DDX (including 185345 TrueColor 16bit, PseudoColor 15bit/12bpg(12 bits per R-, G-, B-channel 185346 as in PostScript Level 2 (and above) colors can have 12 bits per 185347 component (36 bit for RGB)), PseudoColor+GrayScale+StaticGray 185348 12bit/12bpg) and switch the default visual from PseudoColor 8bit/8bpg to 185349 PseudoColor 12bit/12bpg. 185350 185351commit fe37cc7e7b3036e538930c16bbdb39a7915b1685 185352Author: Matthieu Herrb <matthieu.herrb@laas.fr> 185353Date: Fri Apr 1 21:28:50 2005 +0000 185354 185355 programs/Xserver/hw/xfree86/loader/xf86sym.c 185356 programs/Xserver/hw/xfree86/os-support/shared/libc_wrapper.c Fix setjump0 185357 declaration for DllLoader. 185358 185359commit ef60998a828951e61f1480e29c2fec62e7454bbc 185360Author: Kevin E Martin <kem@kem.org> 185361Date: Fri Apr 1 20:21:38 2005 +0000 185362 185363 bugzilla #2880 (https://bugs.freedesktop.org/show_bug.cgi?id=2880) 185364 attachment #2285 (https://bugs.freedesktop.org/attachment.cgi?id=2285) 185365 Use system method to access PCI config space. 185366 185367commit 33ab2a2abc8c1e4ca9c7139454c60f5ad8a61a94 185368Author: Adam Jackson <ajax@nwnk.net> 185369Date: Fri Apr 1 20:05:11 2005 +0000 185370 185371 Bug #2835: Add symbol exports to support the ReadDisplay extension. 185372 185373commit cbccac448a1466ab098e8fe5dbfff98264c7260a 185374Author: Alex Deucher <agd5f@yahoo.com> 185375Date: Thu Mar 31 23:18:10 2005 +0000 185376 185377 - Add new Radeon pci ids (ATI devrel), bug 2827 185378 185379commit 59d7222b13775efb2159159cc897e7789bdf10b2 185380Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 185381Date: Sun Mar 27 00:17:12 2005 +0000 185382 185383 Sat Mar 26 19:00:30 2005 Søren Sandmann <sandmann@redhat.com> 185384 Remove accidentally committed prototype. 185385 185386commit 476ae15640d2b97cb0ebccab8255ccf728596c62 185387Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 185388Date: Sat Mar 26 23:50:24 2005 +0000 185389 185390 Sat Mar 26 18:49:21 2005 Soeren Sandmann <sandmann@redhat.com> 185391 programs/Xserver/fb/fbmmx.h 185392 New function. 185393 Hook it up here 185394 185395commit 0d33b588376a4d86d50ed8b7e06eaf0dbd8c5ba5 185396Author: Roland Mainz <roland.mainz@nrubsig.org> 185397Date: Fri Mar 25 23:11:14 2005 +0000 185398 185399 xc/programs/Xserver/Xprint/ps/Imakefile 185400 //bugs.freedesktop.org/show_bug.cgi?id=2821) attachment #xxx 185401 (https://bugs.freedesktop.org/attachment.cgi?id=xxx) Remove the cfb 185402 dependicy from the PostScript DDX as this is not needed in a vector DDX 185403 (mfb is still needed to fill the |BitmapToRegion()| callback with 185404 |mfbPixmapToRegion()|). 185405 185406commit 5f320335c3a8148ae98f82a00ff44954197f4251 185407Author: Matthieu Herrb <matthieu.herrb@laas.fr> 185408Date: Wed Mar 23 21:09:48 2005 +0000 185409 185410 bugzilla #2194: fix an alignement problem on 64 bit architectures. 185411 185412commit f4e9f522fe8bec3dd6294d062c2244a06623add3 185413Author: Thomas Hellstrom <unichrome@shipmail.org> 185414Date: Wed Mar 23 21:03:41 2005 +0000 185415 185416 Bugzilla #2750 (https://bugs.freedesktop.org/show_bug.cgi?id=2750) 185417 Linux-only fixes: Fix case where a smaller write-combining region 185418 blocks write-combining setting of the whole frame buffer. Fix bug in wc 185419 setting code when regions are first splitted and setting of 185420 write-combining then fails. 185421 185422commit 1be6e2388bcc5835ab62f0855d443fb508697deb 185423Author: Roland Mainz <roland.mainz@nrubsig.org> 185424Date: Wed Mar 23 20:49:52 2005 +0000 185425 185426 xc/programs/Xserver/hw/vfb/InitOutput.c 185427 //bugs.freedesktop.org/show_bug.cgi?id=2791) attachment #2197 185428 (https://bugs.freedesktop.org/attachment.cgi?id=2197) Add support for 185429 12bit PseudoColor and 30bit TrueColor to Xvfb. 185430 185431commit 26aec10adad51deb35a8398abb884d90be077a6b 185432Author: Roland Mainz <roland.mainz@nrubsig.org> 185433Date: Wed Mar 23 19:58:45 2005 +0000 185434 185435 xc/programs/Xserver/Imakefile 185436 xc/programs/Xserver/Xprint/DiPrint.h 185437 xc/programs/Xserver/Xprint/Imakefile 185438 xc/programs/Xserver/Xprint/Init.c 185439 xc/programs/Xserver/Xprint/ddxInit.c 185440 xc/programs/Xserver/dix/Imakefile 185441 xc/programs/Xserver/dix/main.c 185442 xc/programs/Xserver/dix/xpstubs.c 185443 xc/programs/Xserver/os/Imakefile 185444 //bugs.freedesktop.org/show_bug.cgi?id=2792) attachment #2193 185445 (https://bugs.freedesktop.org/attachment.cgi?id=2193) Fix build bustage 185446 when |PrintOnlyServer| is set to |NO|. Patch by Roland Mainz 185447 <roland.mainz@nrubsig.org> and Julien Lafon <julien.lafon@gmail.com>. 185448 185449commit ac18f8e308221af368fd4153b4eee7b89f8dd4bc 185450Author: Roland Mainz <roland.mainz@nrubsig.org> 185451Date: Wed Mar 23 00:32:49 2005 +0000 185452 185453 //bugs.freedesktop.org/show_bug.cgi?id=2789) attachment #2187 185454 (https://bugs.freedesktop.org/attachment.cgi?id=2187) Fix Xvfb to honor 185455 the "-dpi" option (instead of using a hardcoded value of 100DPI). 185456 185457commit 45bcb8e22ad949c456368b7d4f4226110f8b5cfc 185458Author: Adam Jackson <ajax@nwnk.net> 185459Date: Tue Mar 22 21:30:43 2005 +0000 185460 185461 Bug #1821: Typo fix in xorg.conf man page (Jens Schweikhardt) 185462 185463commit cc95e597b51f06e835c6a9def1bc6681029bf41e 185464Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 185465Date: Tue Mar 22 17:49:14 2005 +0000 185466 185467 Tue Mar 22 12:47:16 2005 Søren Sandmann <sandmann@redhat.com> 185468 Only validate the source if it is a viewable window. Stops the cursor 185469 flickering when it is above an unviewable window. 185470 185471commit 1d68ede0eed62b48354a954a62fca98aa2ce9d2d 185472Author: Roland Mainz <roland.mainz@nrubsig.org> 185473Date: Sat Mar 19 20:51:34 2005 +0000 185474 185475 //bugs.freedesktop.org/show_bug.cgi?id=2771) attachment #2148 185476 (https://bugs.freedesktop.org/attachment.cgi?id=2148) Remove RENDER 185477 extension from exclusion list for the standalone print server (Xprt) to 185478 get this extension enabled if the print DDX support it. 185479 185480commit 64f1b3fe7e85245865666607d7b32ef7807df08f 185481Author: Michel Daenzer <michel@daenzer.net> 185482Date: Fri Mar 18 22:36:47 2005 +0000 185483 185484 Add support for production version of ATI RN50/ES1000. (ATI Technologies 185485 Inc.) 185486 185487commit 36dcc9bb1d51fb8c0d67c7e15700e3473a06e47b 185488Author: Torrey Lyons <torrey@mrcla.com> 185489Date: Thu Mar 17 01:12:07 2005 +0000 185490 185491 Fix build on stock Mac OS X: Out of the box, Mac OS X does not include 185492 Freetype2, Expat, or PNG. Also global variables should be initalized in 185493 the Xserver. 185494 185495commit 82f5a127522e48ff7ff78400eadbce0a5a362064 185496Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 185497Date: Wed Mar 16 21:25:43 2005 +0000 185498 185499 Wed Mar 16 16:17:43 2005 Søren Sandmann <sandmann@redhat.com> 185500 Only validate the source if it is a viewable window. Stops the cursor 185501 flickering when it is above an unviewable window. 185502 185503commit 226c0907d4327a440fb0ac5380a19538ffdc6fa0 185504Author: David Reveman <c99drn@cs.umu.se> 185505Date: Wed Mar 16 21:05:51 2005 +0000 185506 185507 Add cursor support to Xglx 185508 185509commit 7109ae147c3a9d243d11f386cfbcfbf7b4ea7918 185510Author: David Reveman <c99drn@cs.umu.se> 185511Date: Wed Mar 16 20:05:19 2005 +0000 185512 185513 Fix a few problems in Xgl 185514 185515commit 04ccba4d40bef6cee902b118598272f26eebb1df 185516Author: Egbert Eich <eich@suse.de> 185517Date: Wed Mar 16 12:16:06 2005 +0000 185518 185519 Don't fail calling function when DriverFunc() for RandR fails as 185520 DriverFunc() also returns FALSE when the specific sub function isn't 185521 supported. In the case of xf86RandRGetInfo() we simply rely on what has 185522 been set before and return TRUE. In the case of xf86RandRSetConfig() we 185523 only bail with FALSE if we have to do a rotation and the call fails. We 185524 presently cannot do rotation on the fly without the help of a driver 185525 function (Bugzilla #2745). 185526 185527commit 1011762254b41db5ce67cb652a2d4965efd7ec20 185528Author: Egbert Eich <eich@suse.de> 185529Date: Wed Mar 16 11:54:54 2005 +0000 185530 185531 Make message that gets printed when loader encounters a .o with no symbols 185532 less conspicuous. 185533 185534commit 816606b9eabee334ce6e0b79e8aa67544f428c19 185535Author: David Reveman <c99drn@cs.umu.se> 185536Date: Fri Mar 11 12:26:20 2005 +0000 185537 185538 Return early from xglFillRect when no rectangles should be filled 185539 185540commit 4de5aa428514f2cacc60d4708dad996dedee1092 185541Author: David Reveman <c99drn@cs.umu.se> 185542Date: Fri Mar 11 00:58:49 2005 +0000 185543 185544 Use negative stride and PBOs in Xgl 185545 185546commit 8653db5d57199d53c9b2b993c35a7b70c8949989 185547Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185548Date: Thu Mar 10 20:05:46 2005 +0000 185549 185550 Force rebuilding of window stack if a window changes it's state from 185551 minimized. 185552 185553commit 522628f25b4a075c8daf547991ea5b80c5efe9c7 185554Author: Egbert Eich <eich@suse.de> 185555Date: Tue Mar 8 10:26:59 2005 +0000 185556 185557 Adding more errnos to the libc wrapper (Bugzilla #2672). 185558 185559commit db7c9d349b86216ed00888181c64ab707fbe18d4 185560Author: David Reveman <c99drn@cs.umu.se> 185561Date: Tue Mar 8 09:27:09 2005 +0000 185562 185563 Return early from xglCopy if fall-back is more efficient 185564 185565commit e09d1d2ae3ccd59408b1dc6f264897ae12dfa2b8 185566Author: David Reveman <c99drn@cs.umu.se> 185567Date: Tue Mar 8 09:12:17 2005 +0000 185568 185569 Better ShmPutImage support in Xgl 185570 185571commit 5d9885c5b95286c8d7f777c7232283e8b1e81d1b 185572Author: David Reveman <c99drn@cs.umu.se> 185573Date: Tue Mar 8 09:03:38 2005 +0000 185574 185575 Minor improvement to CopyArea in Xgl 185576 185577commit 51155ca68bf7539bd3ace2ac068a2be1fbcf400c 185578Author: David Reveman <c99drn@cs.umu.se> 185579Date: Tue Mar 8 08:48:22 2005 +0000 185580 185581 Improve Xgl offscreen memory manager 185582 185583commit e26a096cb662700387c7b43289d1f6f7ab4a0aac 185584Author: David Reveman <c99drn@cs.umu.se> 185585Date: Tue Mar 8 08:30:47 2005 +0000 185586 185587 Fix Xgl glyph caching 185588 185589commit 8d0e520721ab7697d2d4f639425499b79c61b43f 185590Author: Roland Mainz <roland.mainz@nrubsig.org> 185591Date: Mon Mar 7 23:02:59 2005 +0000 185592 185593 xc/programs/Xserver/dix/atom.c 185594 xc/programs/Xserver/dix/colormap.c 185595 xc/programs/Xserver/dix/cursor.c 185596 xc/programs/Xserver/dix/devices.c 185597 xc/programs/Xserver/dix/dispatch.c 185598 xc/programs/Xserver/dix/dixfonts.c 185599 xc/programs/Xserver/dix/dixutils.c 185600 xc/programs/Xserver/dix/events.c 185601 xc/programs/Xserver/dix/extension.c 185602 xc/programs/Xserver/dix/gc.c 185603 xc/programs/Xserver/dix/glyphcurs.c 185604 xc/programs/Xserver/dix/grabs.c 185605 xc/programs/Xserver/dix/main.c 185606 xc/programs/Xserver/dix/pixmap.c 185607 xc/programs/Xserver/dix/privates.c 185608 xc/programs/Xserver/dix/property.c 185609 xc/programs/Xserver/dix/resource.c 185610 xc/programs/Xserver/dix/swaprep.c 185611 xc/programs/Xserver/dix/swapreq.c 185612 //bugs.freedesktop.org/show_bug.cgi?id=2560) attachment #2037 185613 (https://bugs.freedesktop.org/attachment.cgi?id=2037) ANSI-fy 185614 Xserver/dix code. The conversion preserves the comments which annotate 185615 variables. These have been moved into doxygen(esque?) "stubs" above 185616 each function. Patch by Mike Owens <etc@filespanker.com>. 185617 185618commit cb0aa2b4d8875f1ea66e720ca7c6cc2f403be26a 185619Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185620Date: Mon Mar 7 22:26:59 2005 +0000 185621 185622 Prevent winRaiseWindow from calling ConfigureWindow if the message was sent 185623 from within winDestroyWindowsWindow 185624 DestroyWindow send a WM_WINDOWPOSCHANGED to another window causing a 185625 restacking of all windows, even of the window which is just about to 185626 destroyed and whose structures may not be intact anymore. 185627 185628commit 978f3b496b9951ee8120a0efcc5cd12503e26770 185629Author: Roland Mainz <roland.mainz@nrubsig.org> 185630Date: Sat Mar 5 21:38:29 2005 +0000 185631 185632 xc/programs/Xserver/Imakefile 185633 xc/programs/Xserver/hw/xnest/Imakefile 185634 //bugs.freedesktop.org/show_bug.cgi?id=2653) attachment #2020 185635 (https://bugs.freedesktop.org/attachment.cgi?id=2020): Cleanup Xnest 185636 usage of the DPMS dummy stub functions from dpmsstubs.c instead of 185637 using it's own copy of these functions. 185638 185639commit 602209990dbbc96b4c5a96e2221a418cf29e613d 185640Author: Roland Mainz <roland.mainz@nrubsig.org> 185641Date: Sat Mar 5 20:47:12 2005 +0000 185642 185643 //bugs.freedesktop.org/show_bug.cgi?id=2543) attachment #2019 185644 (https://bugs.freedesktop.org/attachment.cgi?id=2019): Fix 185645 |xf86SignalHandler()| which resets the signal handler before setting 185646 the flag indicating a signal has been caught, theoretically allowing 185647 the possibility of infinite recursion. Patch by Andrew Church 185648 <xzilla@achurch.org>. 185649 185650commit d995fe631a5706de93a05fda498333442af3d207 185651Author: Roland Mainz <roland.mainz@nrubsig.org> 185652Date: Wed Mar 2 14:21:40 2005 +0000 185653 185654 xc/programs/Xserver/hw/xnest/Screen.c 185655 xc/programs/Xserver/hw/xnest/Window.c 185656 //bugs.freedesktop.org/show_bug.cgi?id=2546) attachment #2005 185657 (https://bugs.freedesktop.org/attachment.cgi?id=2005): Fix Xnest to 185658 update the shape regions in the backend server whenever a client 185659 changes them in Xnest (the fix is to add a new wrapper which calls 185660 |xnestShapeWindow()| before calling |miSetShape()|). Patch by Mark 185661 McLoughlin <mark@skynet.ie> 185662 185663commit 3f79c5eefc0d62d3a9b095472cd75b446ba2a56e 185664Author: Roland Mainz <roland.mainz@nrubsig.org> 185665Date: Wed Mar 2 11:20:30 2005 +0000 185666 185667 xc/config/cf/DragonFly.cf 185668 xc/config/cf/Imake.cf 185669 xc/config/cf/Imakefile 185670 xc/config/imake/imake.c 185671 xc/config/imake/imakemdep.h 185672 xc/extras/drm/shared/drm.h 185673 xc/include/Xos_r.h 185674 xc/lib/xtrans/Xtranssock.c 185675 xc/programs/Xserver/hw/xfree86/os-support/xf86_OSlib.h 185676 xc/programs/Xserver/hw/xfree86/os-support/xf86_libc.h 185677 xc/programs/Xserver/hw/xfree86/os-support/linux/lnx_agp.c 185678 //bugs.freedesktop.org/show_bug.cgi?id=1712) attachment #2004 185679 (https://bugs.freedesktop.org/attachment.cgi?id=2004): Add support for 185680 DragonFly/BSD platform. Patch by Jeroen Ruigrok <asmodai@tendra.org> 185681 and Mike Verona <firedragonfly@gmail.com>. 185682 185683commit 6c6151b2339a05c60ec58e013f915f79a3f9d756 185684Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185685Date: Tue Mar 1 18:58:17 2005 +0000 185686 185687 If a context is already attached copy it instead of reattaching to keep 185688 displaylists and share displaylists Enable tracing of often called 185689 functions with GLWIN_ENABLE_TRACE ForceCurrent is a no-op now 185690 185691commit d323c4f59a653f364164d2a57fbbd102306a6ee9 185692Author: David Reveman <c99drn@cs.umu.se> 185693Date: Tue Mar 1 16:57:54 2005 +0000 185694 185695 Minor improvement to pixel transfers in Xgl 185696 185697commit 2d2c1732620a83215983ee7a7dd469a1a85fcc12 185698Author: David Reveman <c99drn@cs.umu.se> 185699Date: Tue Mar 1 16:34:31 2005 +0000 185700 185701 Add dither support to Xgl 185702 185703commit b4b27e9eaa43401ae70e5d03823012bf8c78848f 185704Author: Keith Packard <keithp@keithp.com> 185705Date: Mon Feb 28 20:45:15 2005 +0000 185706 185707 Force DPMS normal on screen enable 185708 Add placeholder for bit used to redirect input 185709 Add macrovision register defines 185710 185711commit 409c0618bced6df02eed7af77107ff74508c0f3f 185712Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185713Date: Thu Feb 24 22:53:17 2005 +0000 185714 185715 on WM_WINDOWPOSCHANGED raise window directly and in sync without utilizing 185716 the async windowmanager thread. Fixes some restacking problems occuring 185717 which were timing dependent Do not raise the window on WM_ACTIVATE 185718 Removed unused code for WM_WINDOWPOSCHANGING ESC is debug key. Print 185719 status but do not abort processing the message 185720 185721commit 775efdbd79448040b822fcc0556e98d3968ba8c3 185722Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185723Date: Wed Feb 23 19:17:10 2005 +0000 185724 185725 Bugzilla #2599 (https://bugs.freedesktop.org/show_bug.cgi?id=2599) 185726 attachment #1964 (https://bugs.freedesktop.org/attachment.cgi?id=1964): 185727 move miScreenInit in front of pScreen->function initializations to 185728 prevent it from resetting ClipNotify. 185729 185730commit c4b3fcda98a92c204534f04bd386ace5d3620d86 185731Author: Eric Anholt <anholt@freebsd.org> 185732Date: Mon Feb 21 03:44:10 2005 +0000 185733 185734 Move the draw tracing supplies into ati_draw.h, and do some touchups on it. 185735 (When tracing drawing, I want to know what I'm drawing to, at a 185736 minimum). 185737 185738commit de34b0eefc9f8a29147659454398cabb187c7cb6 185739Author: Eric Anholt <anholt@freebsd.org> 185740Date: Mon Feb 21 03:05:55 2005 +0000 185741 185742 Extend the filter support to R200, and do the check for filter settings in 185743 Check rather than Prepare, to avoid migration of things we won't be 185744 able to accelerate. 185745 185746commit 535c178286f94cc593b6cda753bbeb9b7cf6df4c 185747Author: Michel Daenzer <michel@daenzer.net> 185748Date: Fri Feb 18 19:55:35 2005 +0000 185749 185750 Bug #2576: Add support for ATI RN50/ES1000. (ATI Technologies Inc.) 185751 185752commit 2f07222106358a02f56bf1e344d1fbf7ead14cbd 185753Author: Adam Jackson <ajax@nwnk.net> 185754Date: Fri Feb 18 17:52:48 2005 +0000 185755 185756 Bug #2455: Make x86emu handle JNL correctly. (David Wong) 185757 185758commit 01b156240ce66703b38a67ee3cfbb475352cdf68 185759Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185760Date: Sat Feb 12 14:55:24 2005 +0000 185761 185762 winmultiwindowwindow.c 185763 winmultiwindowwndproc.c 185764 Cleanup some message debugging 185765 185766commit 8df7126f09d6d7cb35a5912a71531cad28ba2545 185767Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185768Date: Sat Feb 12 14:47:17 2005 +0000 185769 185770 win.h 185771 winfont.c 185772 winmultiwindowshape.c 185773 winmultiwindowwindow.c 185774 winpfbdd.c 185775 winshaddd.c 185776 winshadddnl.c 185777 winshadgdi.c 185778 Fix incorrect wrapping of functions. Ensure the pointers from pScreen point 185779 to the called function even if wrapped functions changed it 185780 Set the window properties to NULL to avoid referencing freed memory because 185781 of timing problems after deleting a window 185782 Do not wrap ChangeWindowAttributes. All functions are noops currently 185783 185784commit 676fdb03f3fb27ac24834aeb895df7d6d6e83f78 185785Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185786Date: Sat Feb 12 14:43:07 2005 +0000 185787 185788 print window handle in message outout 185789 185790commit 046bdb17169ecb1361a42ab52043da699590d39a 185791Author: David Reveman <c99drn@cs.umu.se> 185792Date: Fri Feb 11 20:19:20 2005 +0000 185793 185794 Fixes a few problems in Xgl 185795 185796commit 35bd81dfec62402f9d6c68d98e651e8cd87a8758 185797Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 185798Date: Fri Feb 11 19:38:04 2005 +0000 185799 185800 Fri Feb 11 14:28:22 2005 Søren Sandmann <sandmann@redhat.com> 185801 When COMPOSITE is enabled, call CopyWindow even when the pixels "don't 185802 move" on screen. 185803 Don't reject modes that are not supported by the unused monitor. 185804 185805commit e5ccccfbd45c78c1be5e311b2cb4135a9a27540b 185806Author: Adam Jackson <ajax@nwnk.net> 185807Date: Fri Feb 11 06:37:38 2005 +0000 185808 185809 Bug #826: Make xorgconfig respect font installation outside $PROJECTROOT. 185810 (Donnie Berkholz) 185811 185812commit e7369daba58bb4fad5cef37fefbd851e59446045 185813Author: Egbert Eich <eich@suse.de> 185814Date: Wed Feb 9 11:12:54 2005 +0000 185815 185816 Added PCI2Host bus translations for linux PPC and fixed fixed bugs in 185817 macros that apply these functions (Bill Randle Bugzilla #325 and #327). 185818 185819commit aab9a8dd99e52297ed9b40c936600429f38fe9ad 185820Author: Keith Packard <keithp@keithp.com> 185821Date: Wed Feb 9 03:56:35 2005 +0000 185822 185823 Add initial evdev framework 185824 185825commit a85c33b52c40fbae544c7dd40df8c8968e0cf7e9 185826Author: Keith Packard <keithp@keithp.com> 185827Date: Tue Feb 8 22:45:21 2005 +0000 185828 185829 update Xgl to changes in damage API. Remove some flags to support software 185830 mesa 185831 185832commit db2c83551cd3516800b88784c461fb33ee15aacf 185833Author: Keith Packard <keithp@keithp.com> 185834Date: Tue Feb 8 22:43:54 2005 +0000 185835 185836 hw/kdrive/ati/radeon_composite.c Support linear filtering 185837 Change how touch screens work -- make them just another 'mouse' device. Add 185838 unfinished (and unused) code to accelerate tiled fills. 185839 185840commit 70d3a9192feefd54be93ea71231574c3ed815bf2 185841Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185842Date: Tue Feb 8 15:20:01 2005 +0000 185843 185844 Updated fix for ABNT2 and HK_Toggle keys. 185845 185846commit 384099457e9d938871019ba2e5afc20280328884 185847Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185848Date: Tue Feb 8 10:15:49 2005 +0000 185849 185850 Backout ABNT2 and HK_Toggle fix since it broke keys F1 and F4. 185851 185852commit ea5b09f95d6f25d8b0f8858c36b680055edd0da9 185853Author: Egbert Eich <eich@suse.de> 185854Date: Mon Feb 7 18:16:05 2005 +0000 185855 185856 Save gs register before calling the vm86_old syscall thru int 0x80. This is 185857 required for linuxthreads as the TLS uses this register to keep track 185858 of local thread storage (Bugzilla #2431, J. Scott Berg). 185859 185860commit 4d55065b35baa7e13f6e726cb9d6675562648000 185861Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185862Date: Mon Feb 7 15:08:31 2005 +0000 185863 185864 Moved keyboard layout table to external file. 185865 185866commit e132cb7590b5518ef1b7fce5f9151beed916fafc 185867Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185868Date: Mon Feb 7 15:05:39 2005 +0000 185869 185870 file winlayouts.h was initially added on branch CYGWIN. 185871 185872commit c12ef1b34a55544f54401a5a66a36cd728e2f944 185873Author: Egbert Eich <eich@suse.de> 185874Date: Mon Feb 7 11:56:31 2005 +0000 185875 185876 Check the pixel clock choosen for a specific refresh rate against the 185877 maximally allowed pixel clock when choosing mode lines for 185878 VBESetVBEMode() and VESA BIOS version >= 3.0 (Bugzilla #2486). 185879 185880commit 859be7a52b778df8acb676683351a6562a6d4400 185881Author: Felix Kuehling <fxkuehl@gmx.de> 185882Date: Fri Feb 4 01:14:49 2005 +0000 185883 185884 Applied patch (id=1354) by Mike A. Harris from bug #1901: The attached 185885 patch updates a couple of PCI IDs for the Savage driver -- 185886 Debian/Ubuntu/Red Hat/Fedora. 185887 185888commit 5feca068d740b165d3c36a690f5a68b7588b6625 185889Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185890Date: Thu Feb 3 09:58:47 2005 +0000 185891 185892 Bugzilla #1865 (https://bugs.freedesktop.org/show_bug.cgi?id=1865) 185893 attachment #1827 (https://bugs.freedesktop.org/attachment.cgi?id=1827): 185894 check for va_copy not being defined and use __va_copy if available 185895 185896commit 3dda2fe0e2e7e4d2c058d32fa8691d12386b978d 185897Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185898Date: Wed Feb 2 18:06:14 2005 +0000 185899 185900 Force ShowWindow if XWin was started via run.exe. Fixes mainwindow not 185901 showing bug 185902 185903commit e8d3da3c753677cc1ae86bc5a79f2b7eba181d74 185904Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185905Date: Wed Feb 2 17:17:56 2005 +0000 185906 185907 Bugzilla #1866 (https://bugs.freedesktop.org/show_bug.cgi?id=1866) 185908 attachment #1819 (https://bugs.freedesktop.org/attachment.cgi?id=1819): 185909 Define APIENTRY on windows to prevent <GL/gl.h> from loading 185910 <windows.h> removed leftovers from attachment #1818 185911 185912commit 81b862509c87281705f8a8641c28ae2f45f15751 185913Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185914Date: Wed Feb 2 15:03:44 2005 +0000 185915 185916 Bugzilla #1866 (https://bugs.freedesktop.org/show_bug.cgi?id=1866) 185917 attachment #1818 (https://bugs.freedesktop.org/attachment.cgi?id=1818): 185918 Include Xwindows.h before GL/gl.h on windows to prevent loading 185919 windows.h which pollutes our namespace with some symbols. 185920 185921commit b532bfb483cc8ea87c28302e3d676234cab7c3f0 185922Author: Adam Jackson <ajax@nwnk.net> 185923Date: Wed Feb 2 04:07:04 2005 +0000 185924 185925 Bug #1294: Make sure RenderAccel hooks get wrapped in XAA init. 185926 185927commit 44e2d9167943182fea530dfd7ec16aa53db20f4c 185928Author: Roland Mainz <roland.mainz@nrubsig.org> 185929Date: Wed Feb 2 00:55:21 2005 +0000 185930 185931 xc/programs/Xserver/hw/xfree86/os-support/bus/xf86Sbus.h 185932 xc/programs/Xserver/hw/xfree86/os-support/linux/lnx_io.c 185933 //bugs.freedesktop.org/show_bug.cgi?id=825) attachment #956 185934 (https://bugs.freedesktop.org/attachment.cgi?id=956): Fix build 185935 problems on Linux/SPARC. Patch by Jeremy Huddleston 185936 <eradicator@gentoo.org>. Approved in the 2005-01-31 Xorg 185937 release-wranglers phone call. 185938 185939commit 15c555a25df76e0e95bc8eaa2ca7ec80a7695a6c 185940Author: David Reveman <c99drn@cs.umu.se> 185941Date: Tue Feb 1 21:22:02 2005 +0000 185942 185943 Use negative stride for trapezoid masks in Xgl 185944 185945commit 2f0bdf77dd37d1763c4f4f409d55a6aad6031b9f 185946Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185947Date: Tue Feb 1 18:14:01 2005 +0000 185948 185949 xc/programs/Xserver/xkb/ddxLoad.c 185950 Bugzilla #2245 (https://bugs.freedesktop.org/show_bug.cgi?id=2245) 185951 attachment #1649 (https://bugs.freedesktop.org/attachment.cgi?id=1649): 185952 cleanup the generation of xkbcomp command lines. Allocate them 185953 dynamicly and remove unmaintainable length calculation. 185954 185955commit d3ca132061a861cb9292b5a95dbcb2f67695883b 185956Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185957Date: Mon Jan 31 10:49:30 2005 +0000 185958 185959 winmultiwindowwindow.c 185960 Create windows with SWP_NOACTIVATE flag (updated) (Kensuke Matsuzaki) 185961 Fixes for window ordering problem (updated) (Kensuke Matsuzaki) 185962 185963commit 374b9aa8ce14cd20a6768519eee63948c83488d6 185964Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185965Date: Mon Jan 31 10:47:32 2005 +0000 185966 185967 Added hungarian keyboard layout. 185968 185969commit 8d277ceb22929fcb44f2d4def8c5b70535eb087f 185970Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185971Date: Mon Jan 31 10:43:37 2005 +0000 185972 185973 winmessages.h 185974 winmsg.h 185975 winmsg.c 185976 winmultiwindowwndproc.c 185977 winwin32rootlesswndproc.c 185978 Make logging of messages configurable with environment variables 185979 185980commit 2982d173cad762b801869b7ceacc237afdad88d6 185981Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 185982Date: Mon Jan 31 10:32:19 2005 +0000 185983 185984 resolve SHGetFolderPath dynamicly since it is not available on all Windows 185985 systems. 185986 185987commit 8ac3be3f6c4bcaa8c3f6080cbfe72db4967feff8 185988Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 185989Date: Sun Jan 30 21:18:46 2005 +0000 185990 185991 //bugs.freedesktop.org/show_bug.cgi?id=1692> Remove reference to 185992 nonexistent Monitors file from xorgconfig (Report & patch from Dejan 185993 Lesjak <dejan.lesjak@ijs.si>) 185994 185995commit 206072c475408b8d4cfd75d897854d8478fe0905 185996Author: Matthieu Herrb <matthieu.herrb@laas.fr> 185997Date: Sun Jan 30 20:12:07 2005 +0000 185998 185999 programs/Xserver/hw/xfree86/os-support/bsd/bsd_kbd.c Fix option 186000 dontVTSwitch under *BSD, when XKB is disabled. Bugzilla #2252. 186001 186002commit 327922f006dcc2e30ec0391d7fa786560fcd1599 186003Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 186004Date: Sat Jan 29 01:01:03 2005 +0000 186005 186006 Update to latest rev from http://pciids.sf.net/ 186007 Remove entries that now duplicate pci.ids Add entries for new Nvidia boards 186008 added in recent nv driver update 186009 186010commit d23c46dd3e2cbf84012055bad43b1bb15809a244 186011Author: Egbert Eich <eich@suse.de> 186012Date: Fri Jan 28 16:13:00 2005 +0000 186013 186014 Modifying X.Org Xserver DDX to allow to run X with ordinary user 186015 permissions when no access to HW registers is required. For API changes 186016 which mostly involve the modifications to make the RRFunc (introduced 186017 with 6.8) more flexible please check Bugzilla #2407. NOTE: This patch 186018 applies changes to OS specific files for other OSes which I cannot 186019 test. 186020 186021commit 1562ec5cc3706acfac1db04366f78e4949ef5502 186022Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186023Date: Thu Jan 27 20:33:34 2005 +0000 186024 186025 Functions like glGenTextures and glBindTexture are in OpenGL 1.1 and can be 186026 resolved at linktime. Fixes tuxkart (at least with wine) 186027 186028commit 5ca5fe71119f35f1f24aeb49e2608684024d450a 186029Author: Eric Anholt <anholt@freebsd.org> 186030Date: Thu Jan 27 05:25:57 2005 +0000 186031 186032 Add an OUT_RING_REG macro for use with DMA_PACKET0, which is like OUT_RING 186033 but includes debugging to ensure that the reg being submitted is the 186034 one that follows in the packet. Convert most uses of OUT_RING to it, 186035 and convert a couple of OUT_REG sets to DMA_PACKET0/OUT_RING_REG. Also, 186036 add checking to see if more registers are submitted to a DMA_PACKET0 186037 than should be, to avoid hangs during stupid mistakes (checking for 186038 less isn't done). 186039 186040commit 0bd459488bf88084d703d801bfd5e79ea6d2b5a5 186041Author: Eric Anholt <anholt@freebsd.org> 186042Date: Thu Jan 27 03:52:49 2005 +0000 186043 186044 Move the bailing out due to ATIUploadToScreen() being unfinished up 186045 farther, so that the traces aren't as confusing. 186046 186047commit a973af4a9ade68b271d0695898d72aea50d8953b 186048Author: Eric Anholt <anholt@freebsd.org> 186049Date: Thu Jan 27 03:50:48 2005 +0000 186050 186051 Remove an unused field. 186052 186053commit 59c7005bb155393c44792d457d98d72e34bb1b51 186054Author: Egbert Eich <eich@suse.de> 186055Date: Wed Jan 26 15:50:17 2005 +0000 186056 186057 When we tested if a PCI base address was unassigned we didn't take into 186058 account that this bar could have been the upper part of a 64bit base 186059 address. This fix makes this code 64bit bar aware (Bugzilla #2382). 186060 186061commit 7c3dd0a5a76bea2c64972ae19d99775b5a992513 186062Author: Egbert Eich <eich@suse.de> 186063Date: Wed Jan 26 15:45:49 2005 +0000 186064 186065 Rewrite of RemoveOverlaps() to make the code more readable. It was 186066 virtually impossible to determine if the old code did what it was 186067 supposed to do. Also changed paradigm: Trust the based address more 186068 than the (guessed) size. (Bugzilla #2383) 186069 186070commit 9817582328cdafee59de616136172c2ce361a4b3 186071Author: David Reveman <c99drn@cs.umu.se> 186072Date: Wed Jan 26 10:58:52 2005 +0000 186073 186074 Xgl improvements 186075 186076commit 13b5a93b70839053b9165b5087872164f0612536 186077Author: Egbert Eich <eich@suse.de> 186078Date: Tue Jan 25 10:08:18 2005 +0000 186079 186080 Fix interpretation of 64bit PCI bases: read hi long word from the right bar 186081 (Michael Yaroslavtsev, Bugzilla #2322). 186082 186083commit 8813898ef6b9eb470e079bcdcffa89d1243b63b9 186084Author: Keith Packard <keithp@keithp.com> 186085Date: Tue Jan 25 06:04:21 2005 +0000 186086 186087 Fix R100 text by forcing the 3d engine to idle before executing more 3d 186088 commands. Add docs for the ISYNC_CNTL register, which doesn't quite do 186089 what we want. 186090 186091commit 33155b4fd3ce025d555f07833f96b760d5cdfbd3 186092Author: Eric Anholt <anholt@freebsd.org> 186093Date: Tue Jan 25 03:37:05 2005 +0000 186094 186095 Finish converting RB2D_DSTCACHE to RB3D_DSTCACHE. Remove an extra pixel 186096 cache flush in the idle function. Init an extra reg for r200, and 186097 annotate the TCL_BYPASS better. Also, clean up some style nits from the 186098 last commit. 186099 186100commit 3b1f1508b13520626839d45185dec09a42b9ff71 186101Author: Keith Packard <keithp@keithp.com> 186102Date: Tue Jan 25 02:39:48 2005 +0000 186103 186104 Add tracing. Hack Radeon cache registers to use 3D addresses. Works on M6 186105 186106commit 6eaca06dac037851ae5c9575048faf932ad5ffc8 186107Author: Eric Anholt <anholt@freebsd.org> 186108Date: Tue Jan 25 01:40:18 2005 +0000 186109 186110 Fix a leak of a region when the driver's CheckComposite fails. 186111 186112commit 8a1bee8ea9a028eef65b8884f73a79fbe84a9f3a 186113Author: Eric Anholt <anholt@freebsd.org> 186114Date: Tue Jan 25 01:38:26 2005 +0000 186115 186116 Silence a warning about uninitialized variable (though it would be). 186117 186118commit 67eeede4e16324990e1a6afc237a3f51b8edea39 186119Author: Adam Jackson <ajax@nwnk.net> 186120Date: Mon Jan 24 20:44:49 2005 +0000 186121 186122 Bug #2004: Make DDC delay slightly longer. (Thomas J. Moore) 186123 186124commit 55736aa8c17f762b15e9bcd7b3f68f8680b7cb33 186125Author: Egbert Eich <eich@suse.de> 186126Date: Fri Jan 21 14:25:26 2005 +0000 186127 186128 Alan Cox requested that we check the kernel version before we use kernel 186129 VGA font save/restore as the required features have been added to Linux 186130 2.6.11 (Bugzilla #2277). 186131 186132commit d7263b11f043c8c0f83d6e05095143c70177926b 186133Author: Keith Packard <keithp@keithp.com> 186134Date: Thu Jan 20 20:51:27 2005 +0000 186135 186136 Reinitialize offscreen memory before enabling cursor on VT switch-to 186137 186138commit 13c6b2f0b6c464ce11f6c332b2fa1a529bdbab01 186139Author: Eric Anholt <anholt@freebsd.org> 186140Date: Thu Jan 20 16:22:04 2005 +0000 186141 186142 Add a set of macros for dealing with the repeated code to wait for a while 186143 reading a register/attempting DMA. Now it'll wait for a certain number 186144 of seconds rather than a certain number of times through the loop 186145 before deciding that it's timed out and resetting the hardware. Also, 186146 add more timeout handling, and reset the draw state after resetting the 186147 engine. 186148 186149commit dbe45c71590ac319250d04a2bf37ec07cd79e42a 186150Author: Eric Anholt <anholt@freebsd.org> 186151Date: Thu Jan 20 07:28:02 2005 +0000 186152 186153 Use RadeonSwitchTo3D() instead of doing the WAIT_UNTIL ourselves (RST3D() 186154 also does DC_FLUSH, which may be important). 186155 186156commit fc43c154943fb1d277a9cffa9a4db7e76db461bc 186157Author: Eric Anholt <anholt@freebsd.org> 186158Date: Thu Jan 20 07:09:00 2005 +0000 186159 186160 Add R200 XV support, and make R100 (hopefully) use linear filtering instead 186161 of nearest. Also, use RadeonSwitchTo3D instead of doing the WAIT_UNTIL 186162 ourselves. 186163 186164commit 77755065345eb71c997c1ff74dcfd2b2bbbf1305 186165Author: Eric Anholt <anholt@freebsd.org> 186166Date: Thu Jan 20 01:09:48 2005 +0000 186167 186168 Make R200 PDMA work -- primary queue sizes are now 9 bits, not 8. 186169 186170commit 9bd876768b2165ec3903ad0848ae2ae950330290 186171Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 186172Date: Wed Jan 19 22:23:20 2005 +0000 186173 186174 Fix debugging ErrorF() so it compiles when DEBUG is true. 186175 186176commit ff433adba3a643512fdd44e41cd08965fab9c9cb 186177Author: Keith Packard <keithp@keithp.com> 186178Date: Wed Jan 19 06:35:28 2005 +0000 186179 186180 Prefer 32bpp to 24bpp. Fix 16 color planar mode (!) 186181 186182commit cbcdae5a3f7c4009121f86de52bba6c657f20fff 186183Author: Adam Jackson <ajax@nwnk.net> 186184Date: Mon Jan 17 17:17:45 2005 +0000 186185 186186 more static server build fixes 186187 186188commit 79a7120983eff6fa114d4250fe01b62d4a99a612 186189Author: Adam Jackson <ajax@nwnk.net> 186190Date: Sun Jan 16 01:59:23 2005 +0000 186191 186192 Bug #1895: Fix fbComposeGetSolid for BGR. (David S. Miller) 186193 186194commit 24cdd188dc3c10b56d6a7b46dafefb16c6d13efc 186195Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 186196Date: Fri Jan 14 22:07:59 2005 +0000 186197 186198 Fri Jan 14 17:03:40 2005 Søren Sandmann <sandmann@redhat.com> 186199 Fix from Keith Packard for bitgravity bugs in the Composite extension, 186200 reported by Amir Bukhari. 186201 186202commit 7db2e666e2dc9a1dba468e35e9d382e76ed8be54 186203Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 186204Date: Fri Jan 14 21:56:51 2005 +0000 186205 186206 Fri Jan 14 11:12:46 2005 Søren Sandmann <sandmann@redhat.com> 186207 Use mmx CopyArea in a few more places. 186208 186209commit 761f937fdee9ccd10ad54c1f06e12f2f102547a6 186210Author: Egbert Eich <eich@suse.de> 186211Date: Fri Jan 14 19:29:18 2005 +0000 186212 186213 - Don't suspend Xserver on APM standby request as this seems to cause 186214 problems on may systems which don't support APM standby and might not 186215 be required as according to the APM specs the chips should receive 186216 enough power to retain its state. 186217 - Print out power state change requests to log file in all verbosity 186218 levels. 186219 - Don't change server state if no driver PM function is registered. 186220 (Bugzilla #2279) 186221 186222commit 16f9d2d72a2378470e9c5b31c59fa6c9a00892d6 186223Author: Egbert Eich <eich@suse.de> 186224Date: Fri Jan 14 18:42:26 2005 +0000 186225 186226 Let the OS instead of X save/restore text console fonts on Linux. So far we 186227 relied on the generic VGA layer to restore text console fonts for us 186228 when shutting down the server or VT switching back to the text console. 186229 This has worked rather well but it has some downsides on Linux: a. Many 186230 people use fbdev as console text mode. In this case it is not necessary 186231 to save/restore console fonts as the console is running in graphics 186232 mode anyway. b. Some architectures don't have a fbdev console but 186233 require a full POST of even the primary card (ie. IA64). This posting 186234 has to take place before we even have a chance to save anything. 186235 Therefore the fonts we save are the once written to the chip by POST, 186236 not what has been programmed by the user. c. Certain chipsets utilize 186237 the BIOS to perform mode setting. This may interfer with the vga 186238 save/restore font function in a strange way. It would therefore be 186239 preferrable to let the OS - which has been used to set up the font in 186240 the first place - take care of saving/restoring the data. I will attach 186241 a patch which will do so for Linux. To make this fully functional a 186242 small patch needs to be applied to the Linux kernel. To disable this 186243 feature add: #define DoOSFontRestore NO to your host.def. (Bugzilla 186244 #2277) 186245 186246commit 6c0b03a2362f33ae24a2f6845ed1418c9af4b8bc 186247Author: Adam Jackson <ajax@nwnk.net> 186248Date: Fri Jan 14 17:14:08 2005 +0000 186249 186250 Build fixes for static server. 186251 186252commit df4a1fa9c5cc5d54a9347a2bf4843cae87a942f1 186253Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186254Date: Fri Jan 14 12:17:10 2005 +0000 186255 186256 Added copyright notice. 186257 186258commit 2137bc6eb9f36f4ba999023d83c637024f3a6e4c 186259Author: Roland Mainz <roland.mainz@nrubsig.org> 186260Date: Fri Jan 14 08:37:30 2005 +0000 186261 186262 xc/programs/Xserver/afb/afbbres.c 186263 xc/programs/Xserver/afb/afbbresd.c 186264 xc/programs/Xserver/afb/afbclip.c 186265 xc/programs/Xserver/afb/afbhrzvert.c 186266 xc/programs/Xserver/afb/afbline.c 186267 xc/programs/Xserver/afb/afbmodule.c 186268 xc/programs/Xserver/afb/afbpixmap.c 186269 xc/programs/Xserver/afb/afbpolypnt.c 186270 xc/programs/Xserver/afb/afbpushpxl.c 186271 xc/programs/Xserver/afb/afbtegblt.c 186272 xc/programs/Xserver/cfb/Imakefile.inc 186273 xc/programs/Xserver/cfb/cfballpriv.c 186274 xc/programs/Xserver/cfb/cfbbitblt.c 186275 xc/programs/Xserver/cfb/cfbcppl.c 186276 xc/programs/Xserver/cfb/cfbgc.c 186277 xc/programs/Xserver/cfb/cfbglblt8.c 186278 xc/programs/Xserver/cfb/cfbmap.h 186279 xc/programs/Xserver/cfb/cfbpixmap.c 186280 xc/programs/Xserver/cfb/cfbscrinit.c 186281 xc/programs/Xserver/cfb/cfbtab.h 186282 xc/programs/Xserver/cfb/cfbteblt8.c 186283 xc/programs/Xserver/cfb/cfbunmap.h 186284 xc/programs/Xserver/mfb/maskbits.c 186285 xc/programs/Xserver/mfb/maskbits.h 186286 xc/programs/Xserver/mfb/mergerop.h 186287 xc/programs/Xserver/mfb/mfb.h 186288 xc/programs/Xserver/mfb/mfbclip.c 186289 xc/programs/Xserver/mfb/mfbfont.c 186290 xc/programs/Xserver/mfb/mfbgc.c 186291 xc/programs/Xserver/mfb/mfbmisc.c 186292 xc/programs/Xserver/mfb/mfbpushpxl.c 186293 //bugs.freedesktop.org/show_bug.cgi?id=1114) attachment #667 186294 (https://bugs.freedesktop.org/attachment.cgi?id=667): Convert afb and 186295 cfb{,16,24,32} to be dlloader-friendly. Patch by Adam Jackson 186296 <ajax@freedesktop.org>. 186297 186298commit 61b3c3aef5437f14d413a60da792257b01e9f8fa 186299Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 186300Date: Thu Jan 13 23:07:41 2005 +0000 186301 186302 Thu Jan 13 17:45:13 2005 Søren Sandmann <sandmann@redhat.com> 186303 Make sure the pixmap is a valid new resource 186304 Make sure the context is a legal new resource. 186305 Handle null attrib_list. 186306 Handle null attrib_list; copy attrib_list to data; actually allocate the 186307 new GLXDrawable. 186308 186309commit 433c38f22fa96486a43dc0c9871cc09875251b34 186310Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 186311Date: Thu Jan 13 20:49:21 2005 +0000 186312 186313 Thu Jan 13 15:40:29 2005 Søren Sandmann <sandmann@redhat.com> 186314 Add MMX implementation of non-repeating source IN repeating mask, aka 186315 "translucent window". 186316 Add MMX implementation of CopyArea. 186317 Use MMX implementation of CopyArea. 186318 Use the new implementations. 186319 186320commit a303670107f205c6ca76919ca6cd6af6013073f1 186321Author: Adam Jackson <ajax@nwnk.net> 186322Date: Thu Jan 13 01:22:53 2005 +0000 186323 186324 Bug #2114: PPC64 Linux build fix: use system definition of eieio(). 186325 Originally Gentoo bug #66223. (Tim Yamin, Donnie Berkholz) 186326 186327commit 15895b411779aa3c14ffb92fb58cd8ec24845ea7 186328Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186329Date: Wed Jan 12 16:10:00 2005 +0000 186330 186331 winmsg.c 186332 Introduce function winTrace which prints log message with verbosity 10 186333 Use winTrace for 3 heavily called functions 186334 186335commit 8aabc94596dae0fd4ce2c975de75946685faf2cd 186336Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186337Date: Tue Jan 11 17:33:03 2005 +0000 186338 186339 Document the -silent-dup-error switch 186340 186341commit b6301dc41090899b20003eab1356bc09fc5eed1a 186342Author: Egbert Eich <eich@suse.de> 186343Date: Tue Jan 11 14:59:02 2005 +0000 186344 186345 Minor format fix. 186346 Added explanatory comment and debug code. 186347 Added comment. 186348 186349commit 27fc6874b34d70a7ddae5ed8f516f6cfaab518b8 186350Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186351Date: Tue Jan 11 12:03:34 2005 +0000 186352 186353 Do not grab ALT-TAB when window is in multiwindow mode 186354 186355commit d365664c58919edb5e121a7c884384438df79776 186356Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186357Date: Tue Jan 11 11:58:12 2005 +0000 186358 186359 Fix crash with not matching definitions of PATH_MAX 186360 186361commit 65b893a707ad8b3e4d0971825c05a965dca36d10 186362Author: Roland Mainz <roland.mainz@nrubsig.org> 186363Date: Mon Jan 10 23:38:50 2005 +0000 186364 186365 //bugs.freedesktop.org/show_bug.cgi?id=2167) attachment #1641 186366 (https://bugs.freedesktop.org/attachment.cgi?id=1641): Fix broken 186367 Solaris print queue enumeration. 186368 186369commit c6b5a9431178ac7eb90ec498e6830366865d1268 186370Author: Roland Mainz <roland.mainz@nrubsig.org> 186371Date: Mon Jan 10 18:47:55 2005 +0000 186372 186373 //bugs.freedesktop.org/show_bug.cgi?id=2254) attachment #1659 186374 (https://bugs.freedesktop.org/attachment.cgi?id=1659): Refix for gcc4.0 186375 build failure. It seems that the |ExecCommand()| function has no 186376 consumer and can safely be removed. 186377 186378commit eb701e9b4481f984fdcfece181126f7424e2bf45 186379Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186380Date: Mon Jan 10 17:54:31 2005 +0000 186381 186382 winkeybd.h 186383 Adjust keysyms for Hiragana_Katakana toggle and backslash/underscore on 186384 Japanese and ABNT2 keyboards 186385 186386commit 2ef9d01c4d3b03f0a5d829304f2a7e2cc1327f3b 186387Author: Roland Mainz <roland.mainz@nrubsig.org> 186388Date: Mon Jan 10 14:16:29 2005 +0000 186389 186390 xc/programs/Xserver/Xprint/Util.c 186391 xc/programs/Xserver/Xprint/attributes.h 186392 //bugs.freedesktop.org/show_bug.cgi?id=2254) attachment #1654 186393 (https://bugs.freedesktop.org/attachment.cgi?id=1654): Fix gcc4.0 build 186394 failure. Patch by Egbert Eich <eich@freedesktop.org> and Roland Mainz 186395 <roland.mainz@nrubsig.org> 186396 186397commit 57387e115caf5a4b9501cc8f6ddeb1946b0e6547 186398Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186399Date: Mon Jan 10 13:13:08 2005 +0000 186400 186401 winkeybd.h 186402 winkeyhook.c 186403 Make keyhook feature work in multiwindowmode too Hook windows keys 186404 186405commit f417159e51afe22de7d4e6ba9f154313c6af59bc 186406Author: Egbert Eich <eich@suse.de> 186407Date: Mon Jan 10 12:20:33 2005 +0000 186408 186409 Make option 'DontVTSwitch' work again with kbd driver under Linux. The kbd 186410 driver now calls the OS layer to handle special keys. Possibly other 186411 special keys and other OSes need to be looked at also. (Helmut 186412 Fahrion). 186413 186414commit fe4e74241f6791cb1cefdddeb492ed0f56ce99b4 186415Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186416Date: Sun Jan 9 17:35:47 2005 +0000 186417 186418 xc/programs/Xserver/xkb/ddxList.c 186419 //bugs.freedesktop.org/show_bug.cgi?id=2245) attachment #1647 186420 (https://bugs.freedesktop.org/attachment.cgi?id=1647): export 186421 Win32System and Win32TempDir remove #ifdef WIN32 block for building 186422 xkbcomp commandline create win32 tempfile in system tempdir use 186423 PATH_MAX*4 for commandline buffer unlink tmpfile again 186424 186425commit ba25f7e8dcaa2690ce3eab839904fac034002e0b 186426Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186427Date: Sun Jan 9 15:29:45 2005 +0000 186428 186429 //bugs.freedesktop.org/show_bug.cgi?id=2245) attachment #1645 186430 (https://bugs.freedesktop.org/attachment.cgi?id=1645): cleanup some 186431 #ifdef __UNIXOS2__ and WIN32 blocks. make OutputDirectory check the 186432 size of the buffer quote all file and pathnames in the xkbcomp 186433 commandline use PATH_MAX*4 for commandline buffer 186434 186435commit 2410b61f430c3ac4be79043f8b00defe6d53148c 186436Author: Roland Mainz <roland.mainz@nrubsig.org> 186437Date: Sun Jan 9 00:38:08 2005 +0000 186438 186439 //bugs.freedesktop.org/show_bug.cgi?id=2240) attachment #1642 186440 (https://bugs.freedesktop.org/attachment.cgi?id=1642): Fix incorrect 186441 usage of /usr/bin/tr in startup script (the used "[\n]" is neither 186442 correct or portable, using "\n" seems to be sufficient (this fixes 186443 various "random" issues, including including Debian bug #258419 and 186444 Debian bug #264983). Patch by Drew Parsons <dparsons@debian.org>. 186445 186446commit 709a2343a8c12ea7e158c63a9737b11744b50994 186447Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186448Date: Sat Jan 8 13:01:03 2005 +0000 186449 186450 Fix a possible null-pointer dereference (Keishi Suenaga) 186451 186452commit d332a909f8b8741af75047d78a62a3d19e0776e1 186453Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186454Date: Thu Jan 6 21:29:09 2005 +0000 186455 186456 Imakefile 186457 InitOutput.c 186458 XWin.rc 186459 winerror.c 186460 wintrayicon.c 186461 winvideo.c 186462 winshaddd.c 186463 Set PROJECT_NAME in Imakefile to create alternative window titles for 186464 Cygwin/X and Xming 186465 186466commit d6a74f2c4aec9c914ec0837bd0bf0d212019093f 186467Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186468Date: Thu Jan 6 16:02:47 2005 +0000 186469 186470 Imakefile 186471 InitOutput.c 186472 XWin.rc 186473 winerror.c 186474 wintrayicon.c 186475 winvideo.c 186476 Set PROJECT_NAME in Imakefile to create alternative window titles for 186477 Cygwin/X and Xming 186478 186479commit 3165236483de936b4ca22f8b6d2d2b8b1a1a4909 186480Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186481Date: Thu Jan 6 13:24:57 2005 +0000 186482 186483 Fix crash with non-nullterminated strings (reported by �yvind Harboe) 186484 186485commit 591ac9c811de0871d3bdcc19cada0ff6715bf67c 186486Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 186487Date: Tue Jan 4 00:16:20 2005 +0000 186488 186489 Bugzilla #2211 (https://bugs.freedesktop.org/show_bug.cgi?id=2211) 186490 attachment #1627 (https://bugs.freedesktop.org/attachment.cgi?id=1627): 186491 xorgconfig default keyboard model outdated (should be pc105, not 186492 pc101/pc102) 186493 186494commit af8bd7161724b6709ffe582dfd830c05d9bf4f26 186495Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 186496Date: Mon Jan 3 21:32:22 2005 +0000 186497 186498 Mon Jan 3 12:45:10 2005 Søren Sandmann <sandmann@redhat.com> 186499 Clean-ups and support for AMD64. Bug 1067. Patch by Nicholas Miell 186500 (nmiell@comcast.net) 186501 Add support for AMD64 186502 Many cleanups using <mmintrin.h> instead of __builin_ia32_*, and intrinsics 186503 instead of inline assembly. Also unconditionally use pshufw on AMD64. 186504 s/USE_GCC34_MMX/USE_MMX/g 186505 186506commit 5f39eff85109a73d006832ad35d9d5b58f93ef0c 186507Author: Egbert Eich <eich@suse.de> 186508Date: Mon Jan 3 15:43:55 2005 +0000 186509 186510 Added missing return value (Bugzilla #2205) Problem found by Stefan Kulow. 186511 186512commit ae6d52092bca6068a7847b3944148336ab489869 186513Author: Egbert Eich <eich@suse.de> 186514Date: Mon Jan 3 15:39:35 2005 +0000 186515 186516 Added missing return value (Bugzilla #2206) Problem found by Stefan Kulow. 186517 186518commit 83e13e21c8d9a3b54cae2ecc2943be3316659870 186519Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186520Date: Fri Dec 24 14:40:28 2004 +0000 186521 186522 Print error message if CreateProcess fails Simplify popen simulation code 186523 for WIN32 Remove temporary file after executing xkbcomp 186524 186525commit 106703edf67139fa52f6810f9ced0ce5ba86a3cd 186526Author: Eric Anholt <anholt@freebsd.org> 186527Date: Wed Dec 22 18:39:41 2004 +0000 186528 186529 Back out the previous day's broken R200 "fix" -- the same number of coords 186530 are always emitted. Fix the real problem, which was not enough regs 186531 being initialized in ati_draw.c. Fix a typo that was resulting in alpha 186532 coming out as 0 * src or 0 * broken instead of src * 1 or src * mask. 186533 Assign the blending results to R0, as appears to be necessary. Unbreak 186534 the dst-alpha-blend-with-no-dst-alpha code. Yow. And set the right DMA 186535 count for the r200 traps code. 186536 186537commit 3035739e5b7d5a9042292d64455feb4b38788fe4 186538Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 186539Date: Wed Dec 22 08:28:16 2004 +0000 186540 186541 Bug #2123 <https://bugs.freedesktop.org/show_bug.cgi?id=2123> Attachment 186542 #1587 <https://bugs.freedesktop.org/attachment.cgi?id=1587> Call to 186543 uname should not check for return == 0, but for >= 0 instead 186544 186545commit fa0677ab43722462042f87c4636a7d59d1cb873f 186546Author: Eric Anholt <anholt@freebsd.org> 186547Date: Tue Dec 21 09:51:47 2004 +0000 186548 186549 Fix r200 render (for real this time?) by setting tex1_comp_cnt right for 186550 non-mask rendering. Reenable it. Also, R200TexFormats was used instead 186551 of R100 in one place. Harmless so far, because the formats were in the 186552 same order. 186553 186554commit 4b0247b9e0a6b7f40cd2738fb29d2ed1acba99e7 186555Author: Eric Anholt <anholt@freebsd.org> 186556Date: Tue Dec 21 09:49:30 2004 +0000 186557 186558 Whitespace nit. 186559 186560commit 894431412613265fd315cf7a707ffa741f93cf47 186561Author: Matthieu Herrb <matthieu.herrb@laas.fr> 186562Date: Wed Dec 15 20:51:25 2004 +0000 186563 186564 Use snprintf. 186565 186566commit eea11301fd9247a28b1daeb340a018c7ea309c41 186567Author: Thomas Winischhofer <thomas@winischhofer.net> 186568Date: Wed Dec 15 15:05:35 2004 +0000 186569 186570 Another fix for MiscPassMessage(): Initialize returned "status". 186571 186572commit 264c3eefe6c0cdee1ff0a5de914f051ab23026b7 186573Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186574Date: Wed Dec 15 12:22:39 2004 +0000 186575 186576 latest changes from CYGWIN 186577 186578commit c7fec26b50f8a64360d1252cdf48370935fb2f76 186579Author: Thomas Winischhofer <thomas@winischhofer.net> 186580Date: Wed Dec 15 00:32:56 2004 +0000 186581 186582 Make MISC extention's PassMessage() actually work and fix memory leaks 186583 186584commit d0b17bda4dc469e2ce72f4f8965916b316e1deb7 186585Author: Egbert Eich <eich@suse.de> 186586Date: Tue Dec 14 08:59:20 2004 +0000 186587 186588 Removed #ifdef'ed out code together with the comment explaining why it was 186589 #ifdef'ed out. 186590 Fixed typo. 186591 Added comment to a changed that's been committed with one of the previous 186592 commits. 186593 186594commit 0e6a122316a94c96051f1d832990032a386097bc 186595Author: Roland Mainz <roland.mainz@nrubsig.org> 186596Date: Mon Dec 13 03:42:32 2004 +0000 186597 186598 //bugs.freedesktop.org/show_bug.cgi?id=1695) attachment #1230 186599 (https://bugs.freedesktop.org/attachment.cgi?id=1230): Fix crash in 186600 Xscreensaver code which allowed any authentificated X client to crash 186601 the Xserver using |XScreenSaverUnsetAttributes()|. Patch by 186602 ajax@nwnk.net 186603 186604commit 159e443a2209eb3ea305e84b847b76ef1637d005 186605Author: Roland Mainz <roland.mainz@nrubsig.org> 186606Date: Mon Dec 13 02:13:32 2004 +0000 186607 186608 //bugs.freedesktop.org/show_bug.cgi?id=1688) attachment #1530 186609 (https://bugs.freedesktop.org/attachment.cgi?id=1530): Fix the current 186610 implementation to make it possible to slow down the mouse pointer or 186611 use arbitrary fractions (without running into rounding error issues). 186612 The change is using the same method of preserving rounding errors that 186613 the exponential method is already using. Patch by Jan Brunner 186614 <Jan_B@gmx.ch>. 186615 186616commit f1768677f73150c686cf5678f5f5d63c0cfa8e56 186617Author: Kristian Høgsberg <krh@redhat.com> 186618Date: Sun Dec 12 23:29:20 2004 +0000 186619 186620 Reduce vidmode logging. 186621 186622commit 9286a5d032ea6bed102db39281c3d2537da4dccc 186623Author: Kristian Høgsberg <krh@redhat.com> 186624Date: Sun Dec 12 22:58:37 2004 +0000 186625 186626 Add fix for 460gx pci scan code. 186627 186628commit 8266a2581d21a1a2880a0e8babb8b0305c435ec0 186629Author: Torrey Lyons <torrey@mrcla.com> 186630Date: Thu Dec 9 22:40:35 2004 +0000 186631 186632 Fix crash with more than one screen reported by John Davidorff Pell. 186633 186634commit 081b33d73f73572cfefba7e5489408a7117b6e9f 186635Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186636Date: Wed Dec 8 15:48:15 2004 +0000 186637 186638 import changes from CYGWIN branch 186639 186640commit 6c317c1c1323bd11292f8f9f179d75c96a18e616 186641Author: Roland Mainz <roland.mainz@nrubsig.org> 186642Date: Wed Dec 8 05:52:20 2004 +0000 186643 186644 //bugs.freedesktop.org/show_bug.cgi?id=1361) attachment #1287 186645 (https://bugs.freedesktop.org/attachment.cgi?id=1287): Avoid DRI 186646 initalisation when the Xfree86-DRI extension was turned off. Patch by 186647 Kristian H�gsberg <krh@bitplanet.net>. 186648 186649commit b9476cd96faf19153c11e3370e9fced8045600f4 186650Author: Roland Mainz <roland.mainz@nrubsig.org> 186651Date: Wed Dec 8 05:48:16 2004 +0000 186652 186653 //bugs.freedesktop.org/show_bug.cgi?id=1361) attachment #938 186654 (https://bugs.freedesktop.org/attachment.cgi?id=938): Allow more 186655 extensions to be enabled/disabled. 186656 186657commit 47935dd7f010f2f77768774ceb6f85667efe4264 186658Author: Matthew Allum <breakfast@10.am> 186659Date: Mon Dec 6 22:29:31 2004 +0000 186660 186661 Xephyr on Xorg fix 186662 186663commit 8091b301c941473ab99626a6e66f72acdb6750f7 186664Author: Matthieu Herrb <matthieu.herrb@laas.fr> 186665Date: Mon Dec 6 21:54:19 2004 +0000 186666 186667 xf86Config.c references USE_DEPRECATED_KEYBOARD_DRIVER so define it if 186668 needed when building this file 186669 - bug fix: pointerMsg -> keyboardMsg in auto-configuration code. 186670 - make the 'kbd' driver the default for autoconfiguration everywhere. 186671 186672commit cde3a175005104e061c1ff133f07c598868bdc4e 186673Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186674Date: Mon Dec 6 18:28:12 2004 +0000 186675 186676 include windows.h 186677 186678commit 9c94971617db11861aed82e16804fc14a9ca1260 186679Author: Egbert Eich <eich@suse.de> 186680Date: Mon Dec 6 15:53:00 2004 +0000 186681 186682 Use the same method of finding the screen pixmap of COMPOSITE extension is 186683 compiled in or not. 186684 Removed stale make variable. 186685 Updated xaaWrapper.c: Under certain circumstances the accel path was not 186686 used when possible. Removed some debugging stuff and stale code that 186687 had been commented out. 186688 Check if maps are really installed before attempting to list them. 186689 186690commit 4945034792b28d1a222a615404bfceaf48a130c5 186691Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186692Date: Mon Dec 6 12:26:50 2004 +0000 186693 186694 problem with max() macro. this one slipped though again 186695 186696commit 531776becf95f66e6e435aad0dc21ead436ff5aa 186697Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186698Date: Sun Dec 5 21:33:23 2004 +0000 186699 186700 missed another wBOOL issue 186701 186702commit 9bd1328c49aafae67a6a6d9fd17063c75d8547a2 186703Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186704Date: Sun Dec 5 21:24:48 2004 +0000 186705 186706 redone ddraw.h to be able to mix it with w32api style COM header files. 186707 obj_base.h is not needed anymore. Using <objbase.h> instead. 186708 Use Xwindows.h instead of windows.h 186709 do not include win_ms.h 186710 remove extra definition of sleep() 186711 Set HOME to Documents and Settings/username if not set 186712 Use Xming basedir instead of ProjectRoot for system.XWinrc 186713 Fix callback functions to use wBOOL instead of BOOL 186714 Fix compiler warnings. Added debug output. 186715 Fix warning about undefined macro max 186716 186717commit bf0a760331e2fe4dbc00e78f87022c7464d9ca4f 186718Author: Roland Mainz <roland.mainz@nrubsig.org> 186719Date: Sun Dec 5 04:39:34 2004 +0000 186720 186721 //freedesktop.org/bugzilla/show_bug.cgi?id=1800): Fix Postscript DDX's 1bit 186722 StaticGray visual to report only 1bit of significant bits in color 186723 specification (instead of 8bits). Patch by Julien Lafon 186724 <julien.lafon@gmail.com> 186725 186726commit 44f4713a056b7a6a076b2f65fbed43e0cfe9ff06 186727Author: Markus Kuhn <Markus.Kuhn@cl.cam.ac.uk> 186728Date: Sat Dec 4 00:43:13 2004 +0000 186729 186730 Encoding of numerous files changed to UTF-8 186731 186732commit f264a7ea741f57fbc3bb900cfbb9e0cc23f46e90 186733Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186734Date: Fri Dec 3 12:04:15 2004 +0000 186735 186736 Removed scprintf, aprintf and snprintf stuff and use newXprintf 186737 186738commit 16a683f4d164899ecfdafb853f48cff10fd13fd4 186739Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186740Date: Fri Dec 3 11:57:42 2004 +0000 186741 186742 Bugzilla #1865, https://bugs.freedesktop.org/show_bug.cgi?id=1865 Added 186743 X(NF)printf and X(NF)vprintf functions which allocate the buffer with 186744 X(NF)alloc 186745 Bugzilla #1865, https://bugs.freedesktop.org/show_bug.cgi?id=1865 extend 186746 snprintf to work on NULL. 186747 186748commit fe2a2213d1db8700f6078379f86ebe8827793c20 186749Author: Roland Mainz <roland.mainz@nrubsig.org> 186750Date: Thu Dec 2 23:47:39 2004 +0000 186751 186752 //freedesktop.org/bugzilla/show_bug.cgi?id=1998): Fix Xserver standalone 186753 build when RENDER extension includes are not availabe. 186754 186755commit 30a4202f3d59a32fd6f93dfd257d93ee21b68ed9 186756Author: Kristian Høgsberg <krh@redhat.com> 186757Date: Thu Dec 2 21:58:58 2004 +0000 186758 186759 Use __printf__ in gcc function attributes to avoid clash with libc wrapper 186760 define, and remove the last bits of the old workaround. 186761 186762commit e62d85baa31fc853aefdef49962ad4cb86ae8245 186763Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186764Date: Thu Dec 2 21:49:54 2004 +0000 186765 186766 Remove some of the ifdef WIN32 checks from WaitForSomething 186767 186768commit 2782b8871196ef28f9a6c84bf6c8b5086d00d5d4 186769Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186770Date: Thu Dec 2 14:19:01 2004 +0000 186771 186772 Adjust the width of the rootless backbuffer to match 32 bit alignment 186773 Make multiplemonitors default for -internalwm 186774 186775commit 3b3e24dc4d89b471d80428dd9ad122f259b1fc81 186776Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186777Date: Thu Dec 2 13:38:30 2004 +0000 186778 186779 Set HasFfs to NO 186780 Fix link order problems with mingw. Add libdix.a after libmi.a. 186781 Pass -DHAS_FFS to compiler 186782 remove inline code for ffs(). It will link to dix/ffs.c instead added 186783 declaration for ffs() 186784 186785commit 2620676306d1eccd24a6bf0637a60842656e6f7c 186786Author: Phil Blundell <pb@reciva.com> 186787Date: Wed Dec 1 19:43:29 2004 +0000 186788 186789 Patch from Florian Boor <florian.boor@kernelconcepts.de>: 186790 Check return value from ts_config. (TslibInit): Likewise. 186791 186792commit 96545d038837bbc7dc435ed7c1f631454e86cecb 186793Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186794Date: Wed Dec 1 14:57:45 2004 +0000 186795 186796 Remove code which sneaked in recently. It is not proven to be correct, just 186797 a workaround and disabled by default 186798 186799commit 3f063fc49cc2d456359a1b0b9f36f27befdb09b0 186800Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186801Date: Wed Dec 1 14:16:07 2004 +0000 186802 186803 Set XERRORDB environment variable to relocate the XErrorDB file 186804 186805commit 908442301478b21febab17e31aa55918eaaa5101 186806Author: Roland Mainz <roland.mainz@nrubsig.org> 186807Date: Tue Nov 30 23:43:33 2004 +0000 186808 186809 //freedesktop.org/bugzilla/show_bug.cgi?id=811): Updating generated 186810 Xprt.html which was forgotten during the previous work. 186811 186812commit af717ae57dd6c5e6ad41b4142d62cdbb55f13777 186813Author: Egbert Eich <eich@suse.de> 186814Date: Tue Nov 30 08:38:44 2004 +0000 186815 186816 Make Xorg the default server to install. 186817 Avoid PIO access on IA64. Some IA64 machine check if legacy ports outside 186818 the VGA range are accessed. The ATi driver however does this to probe 186819 for ISA Mach8/32/64. Since no IA64 has ISA slots this restriction 186820 should not be relevant to the user. 186821 Avoid recursive calls of xf86scanpci(). This function normally detects that 186822 it has been called before by checking if the PCI structure is filled 186823 out. So far if this was not the case (because PCI probing has failed 186824 for some reason) the function is traversed again. With the chipset 186825 specific PCI bus probing this can lead to an endless recursive loop as 186826 the post-probing code calls xf86scanpci() from within this function. 186827 The OS specific PCI code for Linux worked only if bus 0 was populated as it 186828 checked for the presence of /proc/bus/pci/00. Fixed to check for 186829 /proc/bus/pci/<bus_to_look_for> instead. 186830 186831commit 59ccc6465ca15e046ad11362ae5fbb3c71d2c148 186832Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186833Date: Mon Nov 29 12:34:55 2004 +0000 186834 186835 Fixed windows.h include for cygwin. 186836 Bugzilla #1945: Stop unnecessary reordering. 186837 186838commit df2f2ff5a4ebf6c5289da64bf3f572341a0f7656 186839Author: Phil Blundell <pb@reciva.com> 186840Date: Sun Nov 28 23:20:17 2004 +0000 186841 186842 Re-read "fix" structure from kernel after mode selection, in case line 186843 pitch has changed. 186844 186845commit 6062f6a6e7b3c444a35e3f11b2541df2aa0066a2 186846Author: Roland Mainz <roland.mainz@nrubsig.org> 186847Date: Thu Nov 25 22:42:51 2004 +0000 186848 186849 xc/programs/Xserver/Xprint/Imakefile 186850 xc/programs/Xserver/Xprint/Xprt.html 186851 xc/programs/Xserver/Xprint/Xprt.man 186852 //freedesktop.org/bugzilla/show_bug.cgi?id=811): Add missing manual page 186853 for "Xprt" (DocBook master file (Xprt.sgml) and the generated files 186854 (Xprt.html, Xprt.man). 186855 186856commit cbce4cf96dfa2eeecb253d73d7d9acecbc52ad67 186857Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186858Date: Thu Nov 25 12:48:21 2004 +0000 186859 186860 Bugzilla #1914: fix size limit for -fp argument 186861 186862commit 2b75c19ea2bb76512d51e99c5c9ecb073fdb9163 186863Author: Matthew Allum <breakfast@10.am> 186864Date: Wed Nov 24 17:50:50 2004 +0000 186865 186866 minor Xephyr fix 186867 186868commit a96254234fc9410944406f6ae878815cd3cfcee3 186869Author: Matthew Allum <breakfast@10.am> 186870Date: Wed Nov 24 17:08:06 2004 +0000 186871 186872 XEphyr -parent switch fixes 186873 186874commit a7a07d0c71aa3f2e224cceea7e8d8348523136ee 186875Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186876Date: Wed Nov 24 16:56:03 2004 +0000 186877 186878 Finally the multiwindow mode defines a default cursor 186879 186880commit ca4da62478a3ad86ab087245bf8eb7d7fadf31de 186881Author: Eric Anholt <anholt@freebsd.org> 186882Date: Wed Nov 24 09:07:58 2004 +0000 186883 186884 Add <sys/types.h> include to unbreak build on FreeBSD. 186885 186886commit da3df7522d15855fa871c45f3b8db23e7c639a44 186887Author: Kristian Høgsberg <krh@redhat.com> 186888Date: Wed Nov 24 02:45:33 2004 +0000 186889 186890 Remove this file and all references to the binary expiry code. 186891 Remove old obsolete include/extensions/damage.h. 186892 Include srvrv_ctrl(xfree86) in macintosh US XKB symbol file so VT switching 186893 works again on mac (#1872). 186894 Remove out of place #define's of printf to xf86printf. This definition 186895 should only be in xfree86/os-support/xf86_libc.h 186896 186897commit 9a26d6f39e199bad287d4e538ef75700a0102788 186898Author: Thomas Hellstrom <unichrome@shipmail.org> 186899Date: Tue Nov 23 17:29:47 2004 +0000 186900 186901 Bugzilla #1883 (https://freedesktop.org/bugzilla/show_bug.cgi?id=1883): Fix 186902 insufficient SHM detection in XvMC local test. Removed reference to 186903 getpagesize() which caused linking problems on s390. Reported by Stefan 186904 Dirsch <sndirsh@suse.de> 186905 186906commit fec868bf0f67a8f62fc69d55e2ff72b6cacea6f8 186907Author: Roland Mainz <roland.mainz@nrubsig.org> 186908Date: Tue Nov 23 17:10:55 2004 +0000 186909 186910 //freedesktop.org/bugzilla/show_bug.cgi?id=1204): Fix X11 test suite 186911 (caused by DAMAGE layer) failure with Xvfb when rendering text using 186912 |XDrawText*()| (XDrawText() tests 1, 3, 4, 27, 28, 186913 29, 30, 34, 37, 39, 41, 43 and XDrawText16() tests 1, 3, 9, 10, 11, 12, 186914 13, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 34, 186915 37, 39, 41, 43). Patch by Hong Bo Peng <penghb@cn.ibm.com> and Stefan 186916 Dirsch <sndirsch@suse.de>. 186917 186918commit b09f2a0495071a068c2b6b36084c974acf7aab9b 186919Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186920Date: Mon Nov 22 16:04:04 2004 +0000 186921 186922 Windows does not have ffs() implementation. added one 186923 Enable RootlessSafeAlpha and RootlessAcceleration for the Windows port too 186924 186925commit bb43f234c5d418e064c89b928b81b53987f14e92 186926Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186927Date: Mon Nov 22 15:00:04 2004 +0000 186928 186929 Use GetTempPath for finding a place where to store temporary files on 186930 Windows 186931 186932commit bc7493801d1f5177d9ba3fe09accc2a2d72cfe3d 186933Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186934Date: Mon Nov 22 14:28:25 2004 +0000 186935 186936 Fixes for building multiwindow and internalwm on mingw 186937 Changed some debugging output 186938 186939commit 255c3c0e8ca0f402b2c327d70c8a254ba65eda03 186940Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186941Date: Mon Nov 22 14:12:33 2004 +0000 186942 186943 Xming: Place logfile in users tempdir 186944 186945commit 0f7874cbfc01da339cc6be221351ddffdb37805d 186946Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186947Date: Mon Nov 22 13:23:25 2004 +0000 186948 186949 Use a simple hashtable as ConnectionTranslation instead of a plain array on 186950 Windows because socket fds are not sequential and do not start at 0 186951 186952commit e6bc551e3451efe4fcbb55475d6d0ff53fcc9807 186953Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186954Date: Mon Nov 15 18:13:41 2004 +0000 186955 186956 Use CreateProcess instead of system() to spawn xkbcomp on windows. system() 186957 can not handle spaces in the path component. Quoted all filenames on 186958 the commandline. 186959 186960commit 6618567311f41f5e237f12b4204aa32ce174a514 186961Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186962Date: Mon Nov 15 15:58:51 2004 +0000 186963 186964 Remove override of HasSnprintf 186965 186966commit 9826b83826190e514ed115e15691ca015780f9bc 186967Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 186968Date: Mon Nov 15 15:06:51 2004 +0000 186969 186970 Bufzilla #1802, http://freedesktop.org/bugzilla/show_bug.cgi?id=1802 Added 186971 mingw (Win32) port 186972 186973commit cecb668149e1956fb29bc89855182349122e2f4e 186974Author: David Reveman <c99drn@cs.umu.se> 186975Date: Sun Nov 14 23:21:29 2004 +0000 186976 186977 Add xglPixmapToRegion 186978 186979commit 343f965749af0a985573c525dc6084c2519b6ffe 186980Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 186981Date: Sun Nov 14 17:29:56 2004 +0000 186982 186983 Fix typos in output file. (Bugzilla #1849, reported by Yi Ren-Chen) 186984 186985commit 636a9e786881dd42cd6fd579e13b7895cf9f10eb 186986Author: Phil Blundell <pb@reciva.com> 186987Date: Sat Nov 13 18:03:16 2004 +0000 186988 186989 Don't leave stale pointers hanging around after ts_close(). (TslibFini): 186990 Likewise. 186991 186992commit 36d9e01c0d36eaf0d9cb1e77dd2908b09d55a35d 186993Author: Phil Blundell <pb@reciva.com> 186994Date: Sat Nov 13 16:41:00 2004 +0000 186995 186996 New global variable. (KdProcessArgument): Set it to TRUE on -nozap switch. 186997 (KdUseMsg): Add help for -nozap and -rgba switches. 186998 Declare. 186999 Honour dontZap flag. 187000 187001commit 3b0dce3620e4cce74c3a2c7f9077cc28be11740d 187002Author: Thomas Hellstrom <unichrome@shipmail.org> 187003Date: Sat Nov 13 11:09:23 2004 +0000 187004 187005 lib/XvMC/Imake 187006 Added support for automatic loading of the correct hardware XvMC driver. 187007 This involves a protocol extension of the XvMC protocol. The XvMC 187008 revision number was bumped. 187009 187010commit a97548b1c9bb69b5824609c1da1ad66c3a9c5065 187011Author: Kristian Høgsberg <krh@redhat.com> 187012Date: Thu Nov 11 20:10:28 2004 +0000 187013 187014 Cosmetic fix to make xf86pciBus.c use standard min() macro. 187015 187016commit e5040e24f03a2fe770139e6f37acef3da48aa0f9 187017Author: Matthieu Herrb <matthieu.herrb@laas.fr> 187018Date: Thu Nov 11 16:09:58 2004 +0000 187019 187020 oops commited wrong version previously 187021 187022commit e380fd548ed5452d08184723145dd992ad72288c 187023Author: Matthieu Herrb <matthieu.herrb@laas.fr> 187024Date: Thu Nov 11 15:44:31 2004 +0000 187025 187026 update shared lib revisions on OpenBSD (Bug #1828). 187027 update references to xf86site.def in comments (Bug #1827). 187028 fix kbd driver for wskbd protocol and pure wscons console driver (Bug 187029 #1825). 187030 don't add '-4' to generated default file name (bug #1826). 187031 typo in resource name (XFree86 bug #1300, X.org bug #1825) 187032 187033commit d7f46f71d892768ea85552a0d5458b69b561fe21 187034Author: Matthew Allum <breakfast@10.am> 187035Date: Thu Nov 11 14:55:30 2004 +0000 187036 187037 Xephyr grab fix 187038 187039commit 2854fa438ff721637f2e53fbafa141b3db8a9082 187040Author: Adam Jackson <ajax@nwnk.net> 187041Date: Tue Nov 9 17:06:15 2004 +0000 187042 187043 Bug #557: Don't use "bool" as a variable name in xf86cfg, to avoid compiler 187044 bugs. (Donnie Berkholz) 187045 187046commit 367b963b6333ee58e197845f1389a95ff26f108e 187047Author: Adam Jackson <ajax@nwnk.net> 187048Date: Tue Nov 9 15:58:41 2004 +0000 187049 187050 Bug #1765: Add support for R_ALPHA_SREL32 relocations in elfloader. 187051 Originally Debian patch 187052 202_alpha_elfloader_support_R_ALPHA_SREL32.diff. (Falk Hueffner) 187053 187054commit 50cdff7ee2a1b448da24e85efae02237cff9b6b7 187055Author: Matthew Allum <breakfast@10.am> 187056Date: Tue Nov 9 11:36:49 2004 +0000 187057 187058 Xephyr improvements 187059 187060commit 81a3b6fe27567b4f91033ece69996aa6bf8d01a3 187061Author: Matthew Allum <breakfast@10.am> 187062Date: Mon Nov 8 22:39:47 2004 +0000 187063 187064 Add support to Xephyr for lower depths than hosts 187065 187066commit e494e24c508b34b144355cb380aac16abf2fb891 187067Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 187068Date: Sat Nov 6 11:56:57 2004 +0000 187069 187070 Wrap all mwextwm and internalwm code with XWIN_MULTIWINDOWEXTWM 187071 187072commit 16ee24cd994fa3f8d479c972f18c8207a5116cb7 187073Author: Torrey Lyons <torrey@mrcla.com> 187074Date: Fri Nov 5 19:51:38 2004 +0000 187075 187076 Use ROOTLESS instead of platform specific defines. 187077 187078commit 64c7f8e667eef955b266c359291206c7188ce20c 187079Author: David Reveman <c99drn@cs.umu.se> 187080Date: Fri Nov 5 13:26:07 2004 +0000 187081 187082 Use UTF-8 encoding in Xgl sources 187083 187084commit 1d994e1d14225c1dec5251edeae40e24923a31c6 187085Author: David Reveman <c99drn@cs.umu.se> 187086Date: Fri Nov 5 12:46:29 2004 +0000 187087 187088 Use negative strides in Xgl if pixel data have bottom-top scanline order 187089 187090commit 2dd76d646c517e1988897692e9ff8b6194d25c87 187091Author: David Reveman <c99drn@cs.umu.se> 187092Date: Thu Nov 4 23:19:13 2004 +0000 187093 187094 Add working Xgl server code 187095 187096commit 1aef1060647d22b676a29f6dcf1ac54f9fe7ff5d 187097Author: Kensuke Matsuzaki <zakki@peppermint.jp> 187098Date: Thu Nov 4 11:52:22 2004 +0000 187099 187100 Add InternalWM mode. 187101 187102commit 522c878fca3bfe97cd408e37065f827c004faa04 187103Author: Egbert Eich <eich@suse.de> 187104Date: Tue Nov 2 08:54:53 2004 +0000 187105 187106 Removing unneeded private FreeType2 symbol. 187107 Updating to EDID 1.3. (Bugzilla# 1490, Jay Cotton, Egbert Eich). 187108 Removing unneeded code. 187109 Fixed KGA handling for i810. KGA handling for chips derived from C&T chips 187110 is slightly different. The changes make the code consistent with the 187111 C&T (chips) and i740 drivers. 187112 187113commit 1074992c285835ca9d96d11e8352bbe2cdbc2a28 187114Author: Egbert Eich <eich@suse.de> 187115Date: Mon Nov 1 16:05:27 2004 +0000 187116 187117 Fixed sig11 which occured when calling a CloseDisplay() after 187118 XScreenSaverSetAttributes() followed by XScreenSaverUnsetAttributes(). 187119 Caused by missing FreeResource() in XScreenSaverUnsetAttributes(). 187120 Removing unused DDC sections that caused misinterpretation of DDC data due 187121 to a missing break statement in a switch. 187122 Fixed typo: #if <-> #ifdef. 187123 187124commit 7e588ba9abdcc2078b6c361c81806337b8ff0827 187125Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 187126Date: Sat Oct 30 20:33:43 2004 +0000 187127 187128 Add -d flag for compilers like the Sun C compilers that produce dependency 187129 lists themselves. To use with the Sun compilers, add to host.def: # 187130 define UseCCMakeDepend YES # define DependFlags -cc $(CC) -d -xM (Sun 187131 bug id #4245688 - fix by Alan Coopersmith) 187132 Add Solaris to the platforms on which mprotect is run to set execute 187133 permissions when necessary. (Sun bug id #6175128 - fix by Alan 187134 Coopersmith) 187135 Internationalize digital output (Sun bug id #4119396 - fix by Steve 187136 Swales), add -bgpixmap option to set XPM file as background (originally 187137 from STSF project version of xclock by Alan Coopersmith) 187138 xc/programs/xmodmap/handle.c,pf.c xmodmap was printing line numbers which 187139 are one too low in error messages (Xorg bugzilla #1739, Sun bug id 187140 4637857 - fix by Sam Lau) 187141 187142commit 612bd1c27322a69a98b59193e7d31501688359bd 187143Author: Thomas Winischhofer <thomas@winischhofer.net> 187144Date: Fri Oct 29 02:07:15 2004 +0000 187145 187146 Add facility for catching signal 4 from driver. This can be used to check 187147 for OS SSE support. (Part 2) 187148 187149commit 09fdfaa28d2afe33dfadd4293b39a34da268fbdb 187150Author: Thomas Winischhofer <thomas@winischhofer.net> 187151Date: Fri Oct 29 02:06:17 2004 +0000 187152 187153 Add facility to catch sig 4 from driver. This can be used to check for OS 187154 SSE support. (Part 1) 187155 187156commit b0185a4bf77a668e65e94197bdb13331680521c7 187157Author: Thomas Winischhofer <thomas@winischhofer.net> 187158Date: Thu Oct 28 22:12:23 2004 +0000 187159 187160 Add xf86[GetGammaRamp|ChangeGammaRamp|GetGammaRampSize] to symlist in order 187161 to allow drivers to use them 187162 187163commit 9d4823adc8319a20d2ace3d0944ca32e300f6eeb 187164Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 187165Date: Thu Oct 28 14:23:08 2004 +0000 187166 187167 Import recent changes from CYGWIN branch 187168 187169commit db65fce04ddec1dac0d92ad3abc95ab4996bd206 187170Author: Torrey Lyons <torrey@mrcla.com> 187171Date: Mon Oct 25 19:09:11 2004 +0000 187172 187173 Fix rootless Cygwin crash due to acceleration code illegally modifying a 187174 const structure (Reported by Kensuke Matsuzaki). 187175 187176commit a57e85b52357f15f52e81c8d9b310a76e0c62b3f 187177Author: Roland Mainz <roland.mainz@nrubsig.org> 187178Date: Mon Oct 25 07:12:21 2004 +0000 187179 187180 Fix for https://freedesktop.org/bugzilla/show_bug.cgi?id=1664 - RFE: Add 187181 support for the DOUBLE-BUFFER extension to the Xprint server and DDX. 187182 Additionally a "pixmap-scrubber" optimisation is added to the 187183 PostScript DDX to remove all content from a vector pixmap when a 187184 |PolyFillRect()| call covers the whole pixmap surface with a solid fill 187185 (this avoids that a backbuffer vector pixmap gets constantly filled 187186 even when the content is not visible anymore after |XdbeSwapBuffers()| 187187 has cleared the hidden buffer with the background color). 187188 187189commit e0cc4871491ab7a8c52749ff2c077d8f11516b15 187190Author: Phil Blundell <pb@reciva.com> 187191Date: Wed Oct 20 18:06:58 2004 +0000 187192 187193 Add epson. 187194 Build Makefile in hw/kdrive/epson. 187195 187196commit 258c9016a0c966187a81f9d956766791d6e8c505 187197Author: Phil Blundell <pb@reciva.com> 187198Date: Wed Oct 20 18:02:32 2004 +0000 187199 187200 Block SIGIO before disabling input fds. (KdEnableInput): Unblock it after 187201 enabling them again. 187202 187203commit cbd5fbcb7e5b802cbcff317fdc2f848043207690 187204Author: Phil Blundell <pb@reciva.com> 187205Date: Wed Oct 20 08:20:51 2004 +0000 187206 187207 Import Epson 13806 driver from Costas Stylianou: 187208 New files. 187209 187210commit 0584d92b36910d42e4fba96cee5f3380eeb3b493 187211Author: Egbert Eich <eich@suse.de> 187212Date: Mon Oct 18 14:21:46 2004 +0000 187213 187214 Made handling of DevelDrivers for x86-64 more conformant to other 187215 platforms. 187216 Compress all font encodings (Stefan Dirsch). 187217 Fixed warnings. 187218 Turn on forwarding XNSpotLocation event to XIM server in OffTheSpot and 187219 Root mode (bugzilla #1580, James Su). 187220 Added another compose key combination for the Euro symbol (Stefan Dirsch). 187221 Added support for UTF-8 in ja_JP, ko_KR and zh_TW locales (Mike Fabian). 187222 Changed default encoding for ru from KOI8-R to ISO8859-5 (Mike Fabian). 187223 This is the encoding that is also used by glibc. We may break other 187224 libcs - lets see who complains. 187225 Added explanation for DESTDIR to install to a different directory than /. 187226 Added some early bailouts to atiprobe if PCI structure pointer is NULL to 187227 prevent sig11. 187228 XV support doesn't depend on 2D acceleration any more. This patch removes 187229 this limitation from the NSC driver. This is a patch that I have 187230 committed to XFree86 a while ago but never ported over to X.Org. 187231 Matthieu Herrb contributed some build fixes. 187232 Fixing SetDPMSTimers() so that DPMS*Time == 0 disables the appropriate 187233 timer. This takes advantage of the fact that TimerSet() with a timeout 187234 argument 0 behaves like TimerCanel(). 187235 Use /dev/xconsole (named pipe) or devpts for system logger (Werner Fink). 187236 Create missing links for backward compatibility to XFree86 (Stefan Dirsch). 187237 Changed comment to mention xorg. 187238 Changed cursor for the 'kill' action from XC_draped_box to XC_Pirate. If 187239 you don't like it we can change it back (original author unkown). 187240 Added 'pic' to the man page rendering command pipeline (Werner Fink). 187241 Added missing return value (Stefan Dirsch, Roland Mainz) 187242 187243commit 77dab254402073bf96234a6b1056b2277578a976 187244Author: Alex Deucher <agd5f@yahoo.com> 187245Date: Sun Oct 17 20:29:35 2004 +0000 187246 187247 - Add interlaced support to S3 driver (bug 332, Leo Savernik) 187248 - EmulateWheelTimeout (bug 323, Mathias Fröhlich) 187249 - single button double-click (bug 322, Rob Brown) 187250 187251commit 9a0cf1eb43878308c0a83e4f935933b647edc6d5 187252Author: Matthieu Herrb <matthieu.herrb@laas.fr> 187253Date: Sun Oct 17 10:46:14 2004 +0000 187254 187255 Remove some remaining AMOEBA specific code. 187256 187257commit 5505555c15bfa2de1f596ae0997335fdbf07eb5a 187258Author: Roland Mainz <roland.mainz@nrubsig.org> 187259Date: Fri Oct 15 22:29:40 2004 +0000 187260 187261 Fix for https://freedesktop.org/bugzilla/show_bug.cgi?id=1647 - Fix the 187262 problem that |XpSubmitJob()| returns a random value (patch by Stefan 187263 Dirsch/SuSE). 187264 187265commit 60caca718d23012c4c85ce70547610fe05168342 187266Author: Roland Mainz <roland.mainz@nrubsig.org> 187267Date: Fri Oct 15 22:08:38 2004 +0000 187268 187269 Fix for https://freedesktop.org/bugzilla/show_bug.cgi?id=1646 - Fix the 187270 problem that |PsCreateColormap()| returns a random value (patch by 187271 Stefan Dirsch/SuSE). 187272 187273commit 4782b2f7c10708f9662d9fc743c692d0bec85743 187274Author: Roland Mainz <roland.mainz@nrubsig.org> 187275Date: Fri Oct 15 21:43:31 2004 +0000 187276 187277 Fix for https://freedesktop.org/bugzilla/show_bug.cgi?id=1637 - Fix 187278 problems in oid code which may cause the usage of an invalid string 187279 pointer (original patch by Egbert Eich). 187280 187281commit 881b5756dc96dc33f6966ec6fccd324f63559dc2 187282Author: Adam Jackson <ajax@nwnk.net> 187283Date: Thu Oct 14 23:10:30 2004 +0000 187284 187285 Bug #1628: Convert xf86{BusToMem,MemToBus} to PIC code, eliminating a text 187286 relocation and enabling the server to be built as a 187287 position-independent executable. (PaX Team) 187288 187289commit a0251ee285e1f1e28d06927d8ab7d35d59fda607 187290Author: Roland Mainz <roland.mainz@nrubsig.org> 187291Date: Thu Oct 14 02:20:52 2004 +0000 187292 187293 Fix for https://freedesktop.org/bugzilla/show_bug.cgi?id=1629 - Fix crash 187294 on AMD64, regression caused by bug 1496 ("Xorg Xprt does not support 187295 "*xp-listfonts-mode: xp-list-internal-printer-fonts" to toggle the 187296 usage of printer-builtin fonts"). 187297 187298commit 3a055ea35b687e381da4d729dbdd0ebac47f673b 187299Author: Vladimir Dergachev <volodya@mindspring.com> 187300Date: Wed Oct 13 23:02:42 2004 +0000 187301 187302 Modified: 187303 xc/programs/Xserver/hw/xfree86/drivers/ati/radeon_video.c 187304 xc/programs/Xserver/hw/xfree86/drivers/i2c/fi1236.h Squash annoying warning 187305 about fi1236_dump_status 187306 187307commit 3364e7fbd44537b98212820fb2b1941abf0d5ef3 187308Author: Egbert Eich <eich@suse.de> 187309Date: Tue Oct 12 19:13:43 2004 +0000 187310 187311 Set fbdev mode as the default mode on PPC (Olaf Hering). 187312 Added support for IBM space saver keyboard (Stefan Dirsch). Added support 187313 for Cherry CyMotion Master XPress (Marcus Schaefer). 187314 Change order of SetDisplayDevice(), HWRestore(), UnbindGART() and 187315 RestoreBIOSMemSize() to be exactly opposite to the Save procedure in 187316 EnterVT() (Matthias Hopf, Alan Hourihane). 187317 Fix text mode restauration by removing the assumption that the register 187318 which determines which head is programmed is set. to the active head by 187319 the BIOS (Mark Vojkovich). 187320 When I wrote the resource code 5 years ago I made some assumptions which 187321 turned out to be false: I've assumed that the bus number of the PCI 187322 hostbridge would be the PCI bus the bridge links to. This is not 187323 correct. Fixing this assumption is not easy. However I hope that the 187324 attached patch will make the system work regardless as it 'ignores' 187325 host bridges from which the target bus is not known. This should not 187326 matter at all as we really don't care about host bridges (unless we 187327 have bridge specific code which retrieves information about the 187328 bridge). 187329 Fixed server crash on reset when a structure allocated in PreInit() was 187330 freed on CloseScreen(). 187331 Fixed ring buffer lock ups that happened because the structure that 187332 contained ringbuffer data was not zeroed after allocation. 187333 Fixed numerous warnings due to signed unsigned comparisons. 187334 programs/Xserver/hw/xfree86/drivers/nv/nv_driver.c: 187335 (NVBacklightEnable): Changed the order in which the sequencer registers 187336 and the backlight control registers are written. The sequencer control 187337 register need to be written first otherwise DPMS screen blanking 187338 produces vertical bars on a mobile device. lib/Xau/Imakefile: Build 187339 libXau static library PIC so it can be linked into toolkits that would 187340 like to wrap its functionality. 187341 187342commit 4ab7d316eceb23c81c1b208f9291750cf6b37513 187343Author: Egbert Eich <eich@suse.de> 187344Date: Mon Oct 11 09:58:04 2004 +0000 187345 187346 Improving DPMS handling on VT swich and server termination/abort: previous 187347 version called the driver directly and too late. 187348 Unblank secondary screen explicitely. Don't rely on the value read during 187349 register save as the BIOS have blanked the secondary head. 187350 Checking if server isn't switched away before calling sync. Sanity check 187351 for possible bugs in aother areas of the code. 187352 Fixing default amount of of allocated video memory from AGP for i810: Use 187353 16MB if less than 192MB are installed else use 24MB (Matthias Hopf). 187354 187355commit ca1fda2a3f674a6d59de236612c7077387738ec3 187356Author: Matthieu Herrb <matthieu.herrb@laas.fr> 187357Date: Sun Oct 10 17:48:43 2004 +0000 187358 187359 programs/Xserver/Xext/saver.c Fix for XFree86 bugzilla #1224. 187360 187361commit fa9847aeb9094aafc3798aee1fc8379e77a18d6a 187362Author: Torrey Lyons <torrey@mrcla.com> 187363Date: Fri Oct 8 00:35:08 2004 +0000 187364 187365 Allow rootless implementations to override frame reordering. This is used 187366 on Mac OS X when genie-restoring from the Dock to ensure that the 187367 animation completes before drawing the frame. 187368 187369commit d737bc3300cf1847bcea08ca781f37ee3ee62692 187370Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 187371Date: Tue Oct 5 17:28:15 2004 +0000 187372 187373 Make xorgconfig ask again instead of giving up and throwing away all your 187374 answers when you give a bad file name or bad amount of video RAM. (Sun 187375 bug id 5070654 - Derek Wang) 187376 187377commit 75217be88ccb87a54c84e31697ffb98b5f9b8e0a 187378Author: Roland Mainz <roland.mainz@nrubsig.org> 187379Date: Mon Oct 4 05:34:32 2004 +0000 187380 187381 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=661 - Xprt 187382 producing spurious characters in PS output when using the bitmap cache 187383 (Originally reported against Solaris Xprt as Sun bug id #4369307, and 187384 fixed in Solaris by Jay Hobson. 187385 187386commit 658b4ed81f777df2d8b9f47904de90bde1897113 187387Author: Roland Mainz <roland.mainz@nrubsig.org> 187388Date: Mon Oct 4 05:04:14 2004 +0000 187389 187390 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=1299 - Add more 187391 visuals to the Postscript DDX (8bit GrayScale/StaticGray, 1bit 187392 StaticGray and the basic infratructure for "deep" visuals with more 187393 than 8bits per RGB gun). 187394 187395commit e622b346113f65788110777d7d1b5fc436600a4d 187396Author: Roland Mainz <roland.mainz@nrubsig.org> 187397Date: Sun Oct 3 23:29:21 2004 +0000 187398 187399 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=1496 : Fixed Xprt 187400 font handling which did not support "*xp-listfonts-modes: 187401 xp-list-internal-printer-fonts" or "*xp-listfonts-modes: 187402 xp-list-glyph-fonts" to toggle the usage of printer-builtin and glyph 187403 fonts in XListFonts*(), XLoad*Font(), etc. Additionally the Xprint DDX 187404 now explicitly list "xp-listfonts-modes" in 187405 "document-attributes-supported" (for document-level) or 187406 "xp-page-attributes-supported" (for page-level) when the DDX implements 187407 this feature (as described in the CDE DtPrint specification). 187408 187409commit 8b2f127ea0db2c7fee223b69f4fceee0427fb2e4 187410Author: Roland Mainz <roland.mainz@nrubsig.org> 187411Date: Sun Oct 3 15:34:33 2004 +0000 187412 187413 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=1416 - Fix Xprt 187414 PostScript DDX crashes when copying offscreen pixmap content to the 187415 same pixmap. 187416 187417commit 7b27bf869dcf02bccf730706fc451c6f4c72b1f0 187418Author: Vladimir Dergachev <volodya@mindspring.com> 187419Date: Sun Oct 3 15:01:31 2004 +0000 187420 187421 Modified: 187422 xc/programs/Xserver/hw/xfree86/drivers/i2c/fi1236.c Make sure formatting 187423 style is consistent within a single function. MT2032 functions are best 187424 be in separate file anyway. 187425 187426commit 4046653cb63f4cd752212e7cf65fad5408d8f5b8 187427Author: Vladimir Dergachev <volodya@mindspring.com> 187428Date: Sun Oct 3 14:38:31 2004 +0000 187429 187430 Modified: 187431 xc/programs/Xserver/hw/xfree86/drivers/i2c/fi1236.c Fix compilation with 187432 gcc 3.4.x Cleanup xf86DrvMsg noise. 187433 187434commit ebe7b3fe160259b6f19fe760d6ff4f5bb1dd4b72 187435Author: Vladimir Dergachev <volodya@mindspring.com> 187436Date: Sat Oct 2 01:35:33 2004 +0000 187437 187438 xc/programs/Xserver/hw/xfree86/drivers/i2c/*_module.c Change version 187439 strings to XORG. 187440 187441commit 0d474149f1cb68a60927529f6eac611a12acf5e6 187442Author: Vladimir Dergachev <volodya@mindspring.com> 187443Date: Thu Sep 30 22:58:07 2004 +0000 187444 187445 Initial code from GATOS. This needs to be cleaned up, for example the bt829 187446 code is practically untouched since xatitv (which was a standalone test 187447 program). However, it all worked and was debugged over long period of 187448 time, so I prefer to to mess with these for now. 187449 New drivers: fi12xx (including MT2032 - this would be be split off later). 187450 tdaXXX msp34xx bt8xx 187451 187452commit 1dfafe2aeec864a9bdfd6da3324243b2be8e3a62 187453Author: Roland Mainz <roland.mainz@nrubsig.org> 187454Date: Wed Sep 29 04:17:44 2004 +0000 187455 187456 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=1489 - Use |long| 187457 instead of |int| for BIGREQUESTS commandline option. 187458 187459commit de89181444a2f2258a446fe20a25e37fd225a568 187460Author: Roland Mainz <roland.mainz@nrubsig.org> 187461Date: Fri Sep 24 02:11:14 2004 +0000 187462 187463 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=1453 - Fix spaces 187464 in usage output for BIGREQUESTS extension (option "-maxbigreqsize"). 187465 187466commit 884cb060eea2b914191c269a0c9955ed83d07ff9 187467Author: Torrey Lyons <torrey@mrcla.com> 187468Date: Wed Sep 22 23:52:39 2004 +0000 187469 187470 Update XDarwin project to Xcode 1.5 format and remove unused 187471 QUARTZ_SAFETY_DELAY code. 187472 187473commit 35a3bf13a8be96bb29dce32399c3684c6baa527d 187474Author: Torrey Lyons <torrey@mrcla.com> 187475Date: Wed Sep 22 23:38:33 2004 +0000 187476 187477 Add option for XDarwin to track system keyboard layout changes as they 187478 occur (John Harper and Torrey T. Lyons). 187479 187480commit 1b3fa4d53b287cbe9d38c5f8c9fbbc2298690959 187481Author: Torrey Lyons <torrey@mrcla.com> 187482Date: Wed Sep 22 22:52:03 2004 +0000 187483 187484 Warning fixes. 187485 187486commit de68a3339b9f19630e29a17773cad060b1f65300 187487Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 187488Date: Wed Sep 22 17:20:56 2004 +0000 187489 187490 Allow overriding DPMS defaults (timeouts & default for on/off) from 187491 #defines/-D options. 187492 programs/Xserver/hw/xfree86/common/xf86DPMS.c Use defaultDPMSEnabled global 187493 for the default state of DPMS if not set in any config files. 187494 programs/Xserver/hw/xfree86/os-support/sunos/solaris-sparcv8plus.S Add 187495 support for required assembly inline functions for Sun compilers on 187496 Solaris/sparc. 187497 Add support for Solaris/sparc libraries. 187498 187499commit d7514b9162648f894211884b199ef2edc458aa86 187500Author: Adam Jackson <ajax@nwnk.net> 187501Date: Wed Sep 22 04:38:03 2004 +0000 187502 187503 Bug #1252, #1253, #1255, #1256: Various typo fixes from Dave Jones. 187504 187505commit 269012e6014d7c23bf6805ba14ca0b598cdd6313 187506Author: Torrey Lyons <torrey@mrcla.com> 187507Date: Tue Sep 21 22:01:14 2004 +0000 187508 187509 Add offscreen GLX direct rendering with XDarwin's xpr backend (John 187510 Harper). 187511 187512commit 0514f8b65616d8878000764485e71384b71e3860 187513Author: Egbert Eich <eich@suse.de> 187514Date: Tue Sep 21 17:57:36 2004 +0000 187515 187516 Merged over libXpm security fix provided by Chris Evans, Matthieu Herrb and 187517 Alan Coopersmith from release 6.8.1. 187518 Fail during initialization with error if font/fontset is not set for 187519 widget. This prevents a sig11 later when the non-existent font/fontset 187520 structs are referenced. 187521 Check if xf86Info.kbdProc pointer is really set before calling it on abort 187522 as this pointer won't be set if the new modular keyboard driver is used 187523 (Matthias Hopf). 187524 Added new libs to the bindist control files. 187525 Removed inclusion of unnecessary kernel header on Linux. This may fail in 187526 an -ansi environment. 187527 187528commit 814b74662103710665c0b5659a93974ad609276c 187529Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 187530Date: Mon Sep 20 08:44:54 2004 +0000 187531 187532 Bugzilla #1402: The last patch was broken on linux. Take definition of 187533 badSysCall out of the OS specific block. 187534 187535commit b600fcda38f4ab4796b5536cbedc5dee1abd25b1 187536Author: Eric Anholt <anholt@freebsd.org> 187537Date: Mon Sep 20 03:12:00 2004 +0000 187538 187539 Fix the R200 Render code. Composite and Trapezoids are now supported just 187540 as well as on R100. 187541 187542commit 908287addaff10a0f5f6f14bf06a9b85870737ec 187543Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 187544Date: Sun Sep 19 12:59:52 2004 +0000 187545 187546 Bugzilla #1402: fix BigFont extension if SHM is compiled in but not 187547 working. Does not access SHM and privates if the SHM syscall failed 187548 during extension initialization. 187549 187550commit 90ff3688cdc0c2c1b5ccdbd9cc0659b9a355e85f 187551Author: Eric Anholt <anholt@freebsd.org> 187552Date: Sun Sep 19 10:57:31 2004 +0000 187553 187554 Unbreak the AGP DRI case. That was quite a pile of broken code. 187555 187556commit 9297c6149f83de22395503c484a2ca65dbffaf6b 187557Author: Roland Mainz <roland.mainz@nrubsig.org> 187558Date: Sat Sep 18 23:18:35 2004 +0000 187559 187560 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=1404 - Fixing 187561 random crashes (like in DAMAGE code etc.) in Xnest due uninitalised 187562 GetWindowPixmap. Original patch by Alexander Gottwald 187563 <ago@freedesktop.org> 187564 187565commit 6344bb51e2a97d9678cec2ec1dab19abfe9d9e8d 187566Author: Torrey Lyons <torrey@mrcla.com> 187567Date: Sat Sep 18 00:38:30 2004 +0000 187568 187569 Bugzilla #1032: Make rootless acceleration functions compatible with 187570 Damage. 187571 187572commit cedb9a8d62df3391fd89a8b05a2dd64bd098a7df 187573Author: Torrey Lyons <torrey@mrcla.com> 187574Date: Fri Sep 17 21:57:26 2004 +0000 187575 187576 Update Apple's list and hash utility routines to latest versions (John 187577 Harper). 187578 187579commit b56f4532d1a5febb8df45da0e3d3ad7bf8838e5f 187580Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 187581Date: Thu Sep 16 13:22:52 2004 +0000 187582 187583 Remove code which prevented the use from specifying the window size in 187584 nodecoration mode. 187585 187586commit c4083511acd1d0f20a242b8cd3ed62629629ce8f 187587Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 187588Date: Wed Sep 15 17:58:50 2004 +0000 187589 187590 add support for mice with more than 3 buttons and one scroll wheel (Chris 187591 B) 187592 187593commit 516f452e78170bc643117a71bd2246a83b071316 187594Author: Egbert Eich <eich@suse.de> 187595Date: Wed Sep 15 09:23:59 2004 +0000 187596 187597 Adding support for OS dependent probing of IA64 chipsets. Not all IA64 187598 chipsets can be probed without OS support as probing them is only 187599 possible using ACPI. One example of this are the HP ZX1/2 chipsets: 187600 previously the code assumed that these chips were present when no other 187601 of the known chipsets could be probed. This assumption brought SGI 187602 Altrix machines with 64 CPUs to a grinding halt. 187603 187604commit a3aa6a2d865239c5b8f29cbd849ae3288e36b8a9 187605Author: Egbert Eich <eich@suse.de> 187606Date: Wed Sep 15 09:05:22 2004 +0000 187607 187608 Unregistering events in XSelectInput() when unregistering IM filter 187609 callbacks may be a bad idea as others may be interested in this event. 187610 Removed the call to XSelectInput() altogether as we are in root window 187611 anyway (Lubos Lunak). 187612 Fix size of a variable that gets assigned the value of SmartScheduleTime 187613 (long) to long. This should help to prevent smart scheduler lockup on 187614 64 bit systems due to overruns (Andreas Schwab). 187615 187616commit f642fc729b481c55073c75beca301b2f17881179 187617Author: Roland Mainz <roland.mainz@nrubsig.org> 187618Date: Tue Sep 14 23:21:22 2004 +0000 187619 187620 Refix for http://freedesktop.org/bugzilla/show_bug.cgi?id=1361 - RFE: Allow 187621 enabling/disabling of more extensions (fixing duplicate symbols caused 187622 by previous checkin; attachment #899) 187623 187624commit 0d95bdfbabf4c526f6f54d6f1de8811f4e6d5d5f 187625Author: Phil Blundell <pb@reciva.com> 187626Date: Tue Sep 14 23:08:10 2004 +0000 187627 187628 Only set screen parameters if resolution has changed from current values. 187629 Patch from scoony@noos.fr. 187630 187631commit ba3b6fd23be5f1f900fcff57bc586e08bc524e99 187632Author: Eric Anholt <anholt@freebsd.org> 187633Date: Tue Sep 14 06:26:54 2004 +0000 187634 187635 Add proper PCI/AGP detection, based on Mike Harris's code for Radeon, but 187636 using the MMIO mirror of the bits instead of config space. 187637 187638commit d9df39ee2b5b462be87718046b16d30c231563ec 187639Author: Roland Mainz <roland.mainz@nrubsig.org> 187640Date: Tue Sep 14 00:51:25 2004 +0000 187641 187642 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=1361 - RFE: Allow 187643 enabling/disabling of more extensions 187644 187645commit 80dc84d067c5af29e33a0c7ac62eecc8ba4e0871 187646Author: Eric Anholt <anholt@freebsd.org> 187647Date: Sun Sep 12 23:22:31 2004 +0000 187648 187649 - Fix a segfault on VT switch with DRI. Still dies due to cursor allocation 187650 troubles. 187651 - Move the RemoveBlockAndWakeupHandlers to match 187652 RegisterBlockAndWakeupHandlers. 187653 - Enable R100 trapezoid "acceleration" when DRI is working, so that it can 187654 be exposed and worked on. 187655 187656commit ab50679aabdda9e6197568f745d02fe1e65c7a24 187657Author: Eric Anholt <anholt@freebsd.org> 187658Date: Sun Sep 12 23:01:24 2004 +0000 187659 187660 Fix a bad argument missed in the previous commit for ATIDRIDMA* functions. 187661 187662commit 2d069b1d1950b5f6f9140e5573e349e6559251e6 187663Author: Eric Anholt <anholt@freebsd.org> 187664Date: Sun Sep 12 22:21:12 2004 +0000 187665 187666 Move the RegisterBlockAndWakeupHandlers to before DRI initialization. The 187667 change to use that instead of manual wrapping made the DMA dispatch 187668 come after the lock had been dropped, causing lots of pain. 187669 187670commit 27b5a65f05c78a0b6de0d0ace4f4275d990cc3a4 187671Author: Eric Anholt <anholt@freebsd.org> 187672Date: Sun Sep 12 20:31:39 2004 +0000 187673 187674 Reset the CCE/CP on engine reset, and make the ATIDRIDMA functions take a 187675 more useful argument. 187676 187677commit fcd52d276f0276490ef08af2f0d6c52ed631f130 187678Author: Eric Anholt <anholt@freebsd.org> 187679Date: Sun Sep 12 20:19:15 2004 +0000 187680 187681 Add missing kaa.h include for kaaInitTrapOffsets. 187682 187683commit b5d406e8c84d8aba2b45e985e43d02c2e6770019 187684Author: Eric Anholt <anholt@freebsd.org> 187685Date: Sun Sep 12 20:02:10 2004 +0000 187686 187687 Fix handling of is_agp. is_agp is whether the card is actually AGP, while 187688 using_agp should say whether AGP is being used as part of DMA/DRI. 187689 187690commit 7cab70d1cb7298035429dd8953e521a31fc6770d 187691Author: Eric Anholt <anholt@freebsd.org> 187692Date: Sun Sep 12 19:52:51 2004 +0000 187693 187694 Improve error handling, especially in the DRI case. Do some FatalErrors 187695 instead of ErrorFs for things that are really bad, and put limits on 187696 some loops. Now, sometimes instead of hanging the entire system, we 187697 (mostly-) cleanly drop to console when the card has hung. 187698 187699commit 0cd662ea80579c317d706ebe04971bb29d0f9b4f 187700Author: Eric Anholt <anholt@freebsd.org> 187701Date: Sat Sep 11 09:28:19 2004 +0000 187702 187703 - Add disabled WIP trapezoid code for R128 and R100. The R128 rendering is 187704 not doing an add of 1 per triangle like I hoped, and instead seems to 187705 be saturating all the pixels or something. The R100 acceleration 187706 renders pretty well, with some gaps. Note that both are slower than 187707 software due to lack of DMA to submit vertices. 187708 - Mostly fix R128 and Radeon transform support, including supporting 187709 bilinear filtering on R128. Subpixel offsets are still probably an 187710 issue (reported by rendercheck), but I want to make 100% sure about my 187711 understanding of the protocol before changing everybody, including fb. 187712 - Add support for dst formats without alpha to R128 Composite. 187713 - Remove the R128 Blend code, which has long outlived its usefulness. (I 187714 kept it around for one reason: It could be useful for the w/h > 1024 187715 case with no mask and a non-src op. That seems pretty infrequent and 187716 not worth the trouble). 187717 187718commit 396100dd235105a0e2c9013f1e07e4dae0cc3404 187719Author: Eric Anholt <anholt@freebsd.org> 187720Date: Sat Sep 11 09:23:12 2004 +0000 187721 187722 - Don't require Imrecise mode for Trapezoid acceleration. It looks like we 187723 might be able to do Precise in hardware, so leave it up to the driver. 187724 - Add a helper function for computing a set of offsets for smooth trapezoid 187725 rasterizing using many sharp trapezoids. 187726 187727commit 501dcf37aac4ec9298e8c79ca65c048c362bce31 187728Author: Matthew Allum <breakfast@10.am> 187729Date: Wed Sep 8 11:31:49 2004 +0000 187730 187731 cursor fixes 187732 187733commit 20918ab480c8c8285a68e9f974b0208a18acec94 187734Author: Matthew Allum <breakfast@10.am> 187735Date: Tue Sep 7 19:00:04 2004 +0000 187736 187737 avoid ephyr breakage when tslib enabled 187738 187739commit 8bf6ea903be4c052c747e3e81fc977155072299d 187740Author: Matthew Allum <breakfast@10.am> 187741Date: Tue Sep 7 12:44:19 2004 +0000 187742 187743 fixes to Xephyr debug mode 187744 187745commit b3322257815ec4550e1634453167535b02c1bfbd 187746Author: Kevin E Martin <kem@kem.org> 187747Date: Fri Sep 3 23:41:21 2004 +0000 187748 187749 Update formatted docs. 187750 187751commit b2135e589baeb2ea26da50b9167feaea23bcce3c 187752Author: Kevin E Martin <kem@kem.org> 187753Date: Fri Sep 3 16:18:23 2004 +0000 187754 187755 First set of documentation updates. 187756 Include more correct fix for rootless interaction with damage (Bug #1168, 187757 Keith Packard). 187758 187759commit d7fef52254126aa5897a5c58faeda1f61d5b13d8 187760Author: Kevin E Martin <kem@kem.org> 187761Date: Thu Sep 2 04:04:47 2004 +0000 187762 187763 Workaround for servers using rootless layer with damage (Bug #1168, Torrey 187764 T. Lyons). 187765 187766commit 0e56515aa3ebc5ee8cc33213fb22b69bb4b3a0da 187767Author: Kevin E Martin <kem@kem.org> 187768Date: Thu Sep 2 01:38:17 2004 +0000 187769 187770 Fix AGL display problems (Bug #1210, Torrey T. Lyons). 187771 187772commit 2753c8e2c84cc1fb6d73a05258ab7200068e7830 187773Author: Matthew Allum <breakfast@10.am> 187774Date: Wed Sep 1 15:30:58 2004 +0000 187775 187776 see changelog 187777 187778commit 7c0aaa53bf8dcf3d0a8c9e78d31cf62b1766fb35 187779Author: Matthew Allum <breakfast@10.am> 187780Date: Wed Sep 1 11:13:36 2004 +0000 187781 187782 '-dpi' fixes for Xephyr 187783 187784commit 16ff3a872731633b6f1f4920f793153722026189 187785Author: Matthew Allum <breakfast@10.am> 187786Date: Wed Sep 1 08:31:20 2004 +0000 187787 187788 Actuall select events for -parent option 187789 187790commit 2d065c4c33b9ec17c4c791070cf8189cb57bbb9c 187791Author: Matthew Allum <breakfast@10.am> 187792Date: Tue Aug 31 16:33:05 2004 +0000 187793 187794 Added ephyr server sources 187795 187796commit 6ec9ecd591fba9e9b69b8ebbd2fa08c0a2beac08 187797Author: Eric Anholt <anholt@freebsd.org> 187798Date: Mon Aug 30 22:16:46 2004 +0000 187799 187800 Add a set of three hooks for accelerating trapezoids, and use it for the 187801 RasterizeTrapezoid screen function. These hooks will be called for 187802 imprecise, non-sharp trapezoids with A8 destinations. 187803 Note that the current main consumer of trapezoids, cairo, is requesting 187804 precise, sharp trapezoids by not changing the default Picture 187805 attributes, but gets non-sharp effects in software because fb bases its 187806 choice of sharp/non-sharp on the mask format being A8 vs A1, and cairo 187807 asks for A8. Follow fb's (poor?) example by ignoring the sharp setting 187808 and basing the choice off of the mask being A8. 187809 187810commit ccaf332ce3a9393715317edd3b92420c27fc94eb 187811Author: Eric Anholt <anholt@freebsd.org> 187812Date: Mon Aug 30 16:43:10 2004 +0000 187813 187814 Rather than initially place pixmaps in framebuffer based on a size 187815 heuristic, delay the decision until the first 187816 kaaPixmapUse{Screen|Memory}, and put it in framebuffer if UseScreen was 187817 called. Provides a significant improvement in cairo speeds (100% 187818 speedup in cairogears here) and is likely to improve text performance 187819 as well. 187820 187821commit 14b2db63e7ae0c0d356062cd15811484038f97d9 187822Author: Kevin E Martin <kem@kem.org> 187823Date: Sun Aug 29 21:06:00 2004 +0000 187824 187825 Fix make install when BuildServersOnly is YES (Bug #1213). 187826 Fix build failures when UseDeprecatedKeyboardDriver is YES and 187827 DoLoadableServer is NO (Bug #1229, Kristian H�gsberg). 187828 Fix failure when using DLL loader and LD_BIND_NOW is set (Bug #1212, Adam 187829 Jackson). 187830 187831commit ac038e9cc6f7708fdd9f36494861d2f611d5a90a 187832Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 187833Date: Sun Aug 29 00:48:17 2004 +0000 187834 187835 Don't define _XOPEN_SOURCE before including math.h on Solaris - it's not 187836 needed on older releases, and breaks builds on Solaris 10. (Same as 187837 bugzilla #189). 187838 187839commit 89d702763875831604751bac396c3d2400ec59b6 187840Author: Egbert Eich <eich@suse.de> 187841Date: Fri Aug 27 21:44:54 2004 +0000 187842 187843 Fix for XV memory allocation: Also use tiled area for allocation even if it 187844 hasn't been used before. 187845 187846commit 971755765d6ef9cadf11127478af59189034d356 187847Author: Eric Anholt <anholt@freebsd.org> 187848Date: Fri Aug 27 21:09:23 2004 +0000 187849 187850 Bug #1101: Fix PaintWindow in the pixmap case when the window's origin is 187851 not at the backing pixmap's origin. Resulted in incorrect rendering in 187852 at least aisleriot, fluxbox, and KDE apps, and probably many more. 187853 While here, move the ParentRelative loop above the drawable grab -- may 187854 improve correctness with ParentRelative background origins as well. 187855 Note that the border code doesn't handle ParentRelative yet. 187856 187857commit 1840a50bb763d5c94195eaffa3954c1afd77a31a 187858Author: Kevin E Martin <kem@kem.org> 187859Date: Fri Aug 27 20:39:17 2004 +0000 187860 187861 Add missing reply for DMX (Rik Faith). 187862 187863commit f30fbc600f3e5c3b4aceafb8256594af0069f2c9 187864Author: Egbert Eich <eich@suse.de> 187865Date: Fri Aug 27 20:20:54 2004 +0000 187866 187867 Fixed typos. 187868 187869commit 276cf271527b7a6f464025457e5dd452ac13605e 187870Author: Egbert Eich <eich@suse.de> 187871Date: Fri Aug 27 20:16:28 2004 +0000 187872 187873 - adding missing file 187874 187875commit ecd889c6ff7427b4e7d3f820586aac178a4dfdc2 187876Author: Egbert Eich <eich@suse.de> 187877Date: Fri Aug 27 19:27:12 2004 +0000 187878 187879 Added support for LynxOS 4.0 (Thomas Mueller). 187880 Fix arm netwinder build (Donnie Berkholz). 187881 187882commit 7c087078694a6895a9bcbe62d091665c05a86b5a 187883Author: Egbert Eich <eich@suse.de> 187884Date: Fri Aug 27 12:32:14 2004 +0000 187885 187886 - Make DDC code more VESA compliant: when MaxClock is set to 0xFF it means 187887 'no value specified'. Therefore setting it to 0. 187888 187889commit c0bff215de2d034118d06cac42e234008612204b 187890Author: Kevin E Martin <kem@kem.org> 187891Date: Fri Aug 27 01:15:10 2004 +0000 187892 187893 Fix xtest failures in Xvfb from XMatchVisualInfo test. 187894 187895commit 5335bc8a0657b3e378795b44698ed23020c13891 187896Author: Eric Anholt <anholt@freebsd.org> 187897Date: Thu Aug 26 20:32:59 2004 +0000 187898 187899 Bug #1138: Wrap funcs in CW's GC ops as well. While this is unnecessary 187900 according to the rules as I understand them (bug #1045), not everybody 187901 follows the rules. GC funcs were being called on the same GC from 187902 within GC ops, and the cwValidateGC caused a loop in the funcs chain 187903 that resulted in a crash, notably in cwPolylines. 187904 187905commit c8672e7ac79c872344f287f7cc106cb5c006e619 187906Author: Egbert Eich <eich@suse.de> 187907Date: Thu Aug 26 11:58:08 2004 +0000 187908 187909 Fixed support for LynxOS 3.1 (LynxOS 4 will follow) (Thomas Mueller). 187910 2 187911 187912commit 198e62c44b99ee0890b944f92b162387f77aa10d 187913Author: Torrey Lyons <torrey@mrcla.com> 187914Date: Sat Aug 21 00:46:01 2004 +0000 187915 187916 Documentation only update: Update XDarwin application version for release. 187917 187918commit 74d84d7b122802230579a4da8c2e6914d2f5f060 187919Author: Kevin E Martin <kem@kem.org> 187920Date: Fri Aug 20 05:22:39 2004 +0000 187921 187922 Fix keyboard driver failing to initialize if DoLoadableServer is NO (Bug 187923 #1133, Kristian H�gsberg). 187924 187925commit bd3e6e44259155cb37f39eb2ca5e1f0de1c2ebb7 187926Author: Kevin E Martin <kem@kem.org> 187927Date: Thu Aug 19 04:08:40 2004 +0000 187928 187929 Add missing no.*Extension symbols (Bug #1131, Aaron Plattner). 187930 187931commit 87842285007e9ac4f9de0349abee1fb66b7c795f 187932Author: Kevin E Martin <kem@kem.org> 187933Date: Wed Aug 18 21:11:17 2004 +0000 187934 187935 Fix CopyArea for non-redir dst, redir src (Bug #1105, Eric Anholt). 187936 187937commit 9223baf985778c536ce93846c431b46a0192cf32 187938Author: Kevin E Martin <kem@kem.org> 187939Date: Wed Aug 18 18:41:41 2004 +0000 187940 187941 Fix AIX build problems (Bugs #1020, 1102, 1103, Dan McNichol). 187942 Remove old config files (Bug #1123, Jim Gettys). 187943 Remove old log message (Bug #1123, Jim Gettys). 187944 187945commit a45bc0df7a1c369e8429e84414ac813187c90059 187946Author: Kevin E Martin <kem@kem.org> 187947Date: Tue Aug 17 17:55:02 2004 +0000 187948 187949 Fix DRI module loading (Bug #1057, Ronny Vindenes). 187950 Fix Xvfb at 8bpp (Bug #1091). 187951 Fix link order when building with Xprint in glxgears (Bug #1060, Alexander 187952 Gottwald). 187953 187954commit 75de2fe82e7da755555028a724f68b9fb9ddfb14 187955Author: Kevin E Martin <kem@kem.org> 187956Date: Mon Aug 16 22:48:50 2004 +0000 187957 187958 Update release and date for first RC. 187959 Fix kbd/keyboard driver for DoLoadableServer NO. 187960 Revert change since it is better to set date in the config files. 187961 187962commit 1798cac6fa2e909c9f3df26b97ee8232a0bf1592 187963Author: Kevin E Martin <kem@kem.org> 187964Date: Mon Aug 16 20:17:51 2004 +0000 187965 187966 Add kdb <-> keyboard aliasing when UseDeprecatedKeyboardDriver is NO (Bug 187967 #1072, Kristian H�gsberg). 187968 187969commit 6cac342517892a20bab6a6177f8b5742feaaed38 187970Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 187971Date: Mon Aug 16 15:48:53 2004 +0000 187972 187973 document broken composite in XWin 187974 187975commit 9aa6beb6b7e6272b05a03e0a4fd34eb8ad21bf47 187976Author: Kevin E Martin <kem@kem.org> 187977Date: Mon Aug 16 02:07:53 2004 +0000 187978 187979 Fix banner to print out proper version information. 187980 187981commit 9da0c214ab5f4ee9c1610b4888f5c7c0dd2bcacc 187982Author: Keith Packard <keithp@keithp.com> 187983Date: Sun Aug 15 21:13:11 2004 +0000 187984 187985 Clip destination instead of source. Should be the same, but it looks nicer 187986 to me. 187987 Clean up transition between cheap and expensive GC wrappers by using the 187988 prologue and epilogue macros. Before, the GC would be left unvalidated 187989 sometimes which would cause all kinds of entertaining bugs against a 187990 DDX which cares (XAA). 187991 187992commit a68f350195c1c54034f98e2b78c2c3da70044884 187993Author: Keith Packard <keithp@keithp.com> 187994Date: Sun Aug 15 19:05:01 2004 +0000 187995 187996 Remove debugging code which did a full tree walk on every window operation 187997 Eliminate needless (and, it turns out, dangerous) call to ChangeGC on 187998 DestroyGCPrivate. 187999 in cwSetWindowPixmap, check if the pixmap is the screen pixmap and disable 188000 the wrapper by setting the private to NULL. 188001 188002commit 5db70ae2575e3e8669d7a66e2218ba28e8bdfa68 188003Author: Kristian Høgsberg <krh@redhat.com> 188004Date: Sun Aug 15 15:40:19 2004 +0000 188005 188006 Remove #error used for testing. 188007 188008commit 1e728c3e88f6a74b93dc11827c9fcfe7b39ca5a5 188009Author: Keith Packard <keithp@keithp.com> 188010Date: Sun Aug 15 03:34:18 2004 +0000 188011 188012 Copy bits from parent window when allocating pixmaps so that Background == 188013 None works. 188014 Copy filter to backing picture during validation. 188015 Mark picture serialNumber when setting Filter or Transform so Validate 188016 occurs. 188017 Initialize xf86Screens[i]->pScreen to NULL so that RADEON driver doesn't 188018 crash during server reset using old pScreen. 188019 188020commit 943308517905d16bda1bb27cd745bd291a84dbf6 188021Author: Keith Packard <keithp@keithp.com> 188022Date: Sun Aug 15 00:43:39 2004 +0000 188023 188024 Redraw window borders when switching window pixmaps around 188025 Make cw "own" the window pixmaps by wrapping 188026 GetWindowPixmap/SetWindowPixmap. 188027 188028commit 597fdae93e6e1b7e4052097baf3d91e7a134c162 188029Author: Kristian Høgsberg <krh@redhat.com> 188030Date: Sat Aug 14 23:59:52 2004 +0000 188031 188032 More kbd fun: write out "kbd" from Xorg -configure (#1078). 188033 188034commit e483fe3ec384da556c31292001a86ec95c2ddc46 188035Author: Keith Packard <keithp@keithp.com> 188036Date: Sat Aug 14 21:57:58 2004 +0000 188037 188038 Fix offsets again. Really, it works this time. Promise. 188039 188040commit e6216b48f5feee72f107348cb21bad724536ec62 188041Author: Keith Packard <keithp@keithp.com> 188042Date: Sat Aug 14 21:36:10 2004 +0000 188043 188044 Fix offsets, do whole region at once by using GC clipping 188045 188046commit ed425d1d88a72586d5d7a4aad9d0be0b06637070 188047Author: Eric Anholt <anholt@freebsd.org> 188048Date: Sat Aug 14 20:29:28 2004 +0000 188049 188050 Bug #1077: Fix source copy performance problem exposed by Composite. 188051 188052commit cc3ad0ed4302f54318e190a2b10646337f242d40 188053Author: Keith Packard <keithp@keithp.com> 188054Date: Sat Aug 14 19:53:36 2004 +0000 188055 188056 Fix clip list computation and setting to ignore clip changes to "real" 188057 GC/Picture and track serial numbers correctly when copying 188058 pCompositeClip down. 188059 188060commit 183c6d06455114c61f6db57ec0a084caf11ece3a 188061Author: Eric Anholt <anholt@freebsd.org> 188062Date: Sat Aug 14 19:51:11 2004 +0000 188063 188064 Wrap CopyWindow in cw, which fixes scrolling in many apps. 188065 188066commit e61b5d38ab30c4f73ba0d070f485a32708a03eb6 188067Author: Keith Packard <keithp@keithp.com> 188068Date: Sat Aug 14 07:12:37 2004 +0000 188069 188070 Use XLIB_SKIP_ARGB_VISUALS environment variable to disable all depth 32 188071 visuals. Necessary to keep Flash from crashing. 188072 Must call ValidateGC/ValidatePicture on "real" GC/Picture to ensure 188073 pCompositeClip is set correctly. 188074 Need to take the composite clip from the "real" GC/Picture and turn it into 188075 the clientClip for the backing version. 188076 Adjust pixmap screen origin to account for drawable->x/y Change debugging 188077 output a bit (disabled by default) 188078 188079commit 05f6329eb6f564ad4fc366d75f4ebf9f3ba4b5dd 188080Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 188081Date: Fri Aug 13 23:57:38 2004 +0000 188082 188083 Don't enable speedo & type1 modules if they're not being built 188084 Clean up a couple of hardcoded paths & vendor names to use defines set by 188085 Imakefile 188086 188087commit 3f84e4f71d9c7c560f9bef675b1cc96fa1d83b14 188088Author: Kevin E Martin <kem@kem.org> 188089Date: Fri Aug 13 19:51:34 2004 +0000 188090 188091 Fix Xprt bug by disabling code that merges multiple audit messages (Bug 188092 #964, Roland Mainz). 188093 188094commit 922fd3a2e568571171dfd64a94f804350829230f 188095Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 188096Date: Fri Aug 13 19:18:29 2004 +0000 188097 188098 Added $(MESASRCDIR)/src/mesa/glapi to INCLUDES. Removed $(SERVERSRC)/mi 188099 from INCLUDES. Rearranged INCLUDES for better readability. 188100 Removed mipointrst.h and miscstruct.h from #include since they are not used 188101 anymore. 188102 188103commit f63f4b768cd5ec5bffd270e448e6e51b8ad67016 188104Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 188105Date: Fri Aug 13 18:24:07 2004 +0000 188106 188107 Fri Aug 13 19:53:10 2004 Soeren Sandmann <sandmann@redhat.com> 188108 Fix for lockups on some versions of Matrox Mystique. #687, Patch from Mike 188109 Harris. 188110 Call xf86EnableDisableFBAccess though the function pointer instead of 188111 directly. #1041, Patch from Aaron Plattner. 188112 Swap the phsyical size of the screen when rotiation. #1050, Patch from 188113 Aaron Plattner. 188114 Fri Aug 13 19:47:12 2004 Soeren Sandmann <sandmann@redhat.com> 188115 Make HAVE_FT_BITMAP_SIZE_Y_PPEM conditional on the FreeType version instead 188116 of proping it. This way it will work with the monolithic version too. 188117 #1062, Patch by Owen Taylor. 188118 188119commit a29bfbd3d0a5d39ccee5b83ac1ba632091b031bb 188120Author: Keith Packard <keithp@keithp.com> 188121Date: Fri Aug 13 08:16:14 2004 +0000 188122 188123 Empty damage object when freeing pixmap. 188124 Wrap InstallColormap so that the DDX doesn't see colormaps from our ARGB 188125 visual (avoids lovely green tint to screen). Also, set visual->nplanes 188126 of ARGB visual to all used (including alpha) planes so DIX can set 188127 pixel values correctly. 188128 Translate automatic update regions correctly to account for borders 188129 When nplanes == 32 (ARGB visuals), mask in all ones for alpha values to 188130 allocated pixel values. 188131 Remove redundant fbAddTraps declaration 188132 Fix fbCopyWindow to work on non-screen pixmaps (not needed yet) 188133 Replace broken clipping code with that from modular tree. 188134 Respect subWindowMode. 188135 188136commit 24bed5cff908a6f8b1857e3dadac22d6db54c69e 188137Author: Eric Anholt <anholt@freebsd.org> 188138Date: Fri Aug 13 07:47:21 2004 +0000 188139 188140 Fix copy'n'paste-os of x/y in CopyPicture for AlphaXOrigin and ClipXOrigin. 188141 188142commit 5825e4559e7aaf3b40205a0dca49a785c8de7b92 188143Author: Eric Anholt <anholt@freebsd.org> 188144Date: Thu Aug 12 23:14:50 2004 +0000 188145 188146 Fix after Mesa 20040812 merge: revert glxext.h to vendor branch, and fix 188147 bug #1022. 188148 188149commit 2889ad2cb8827f20b6d69da4fe99db33bf9c5ff2 188150Author: Torrey Lyons <torrey@mrcla.com> 188151Date: Thu Aug 12 20:24:36 2004 +0000 188152 188153 Fix crash in rootless XDarwin due to rootless being initialized before 188154 damage extension. 188155 188156commit 961333143e2df3e3f33e8624fc61e79cf3e86cd1 188157Author: Eric Anholt <anholt@freebsd.org> 188158Date: Thu Aug 12 08:45:33 2004 +0000 188159 188160 Apply a kludge to initialize the composite wrapper before DamageSetup. If 188161 not, DamageSetup will wrap some operations first, and the cw 188162 initializes during ExtensionInit, so cw comes higher in the wrapping 188163 chain. cw going first will result in damage getting confused when the 188164 drawables get changed around. 188165 188166commit 6e0228722cc2fa37a0e2359bc3dab3646e36c4b7 188167Author: Eric Anholt <anholt@freebsd.org> 188168Date: Thu Aug 12 08:11:59 2004 +0000 188169 188170 Fix various cw issues, including a couple reported by deronj: 188171 - Fix wrapping of GC ops/funcs according to policy described in bug #1045. 188172 - Remove ValidateGC/ValidatePictures on the redirected drawables/pictures 188173 -- it's not needed, and DDXs shouldn't be seeing redirected drawables 188174 in render or GC ops/funcs when cw is running. 188175 - Mark all GC/Picture state as dirty when moving from redirected to 188176 non-redirected, since it hadn't been passed down in Change* or 188177 Validate* while redirected. 188178 - Remove CreatePicture wrapper that didn't do anything. 188179 - Comment on why AddTraps wrapper isn't needed. 188180 188181commit 789cf3ed846045d91f950cb177ef6bae4c8966fc 188182Author: Eric Anholt <anholt@freebsd.org> 188183Date: Thu Aug 12 07:57:03 2004 +0000 188184 188185 Fix some issues reported by deronj: 188186 - Hopefully fix a crash in compCheckRedirect on unrealizing windows. 188187 - Remove an extern that doesn't point at anything. 188188 188189commit 1a073786e0159a80ac3b8772a1d89b0618a8ff33 188190Author: Kevin E Martin <kem@kem.org> 188191Date: Thu Aug 12 05:11:57 2004 +0000 188192 188193 Update version and date for next snapshot 188194 188195commit 47ee5f4ba72f0e0bc92a5e04073c70808e85fc08 188196Author: Kristian Høgsberg <krh@redhat.com> 188197Date: Thu Aug 12 01:57:51 2004 +0000 188198 188199 Add call to SourceValidate() when pDst == pSrc, so misprite.c get a chance 188200 to remove the sprite before the area is copied. The drivers handle pDst 188201 != pSrc (#1030). 188202 188203commit 09d0056b8bc103f67a35980934f03d28fed51164 188204Author: Adam Jackson <ajax@nwnk.net> 188205Date: Thu Aug 12 01:30:57 2004 +0000 188206 188207 Wrap the Propolice fix in #ifdef __SSP__; Propolice doesn't define this yet 188208 but an RFE has been submitted upstream. 188209 188210commit e30c22dbbffa2605f8d7ac010c8208a135a2293a 188211Author: Adam Jackson <ajax@nwnk.net> 188212Date: Thu Aug 12 01:16:36 2004 +0000 188213 188214 Back out the propolice fix for a second, it causes linktime errors. 188215 188216commit 9d0213525f4f692c250f10309146fe9db76ba12d 188217Author: Keith Packard <keithp@keithp.com> 188218Date: Thu Aug 12 00:09:30 2004 +0000 188219 188220 Paint to parent window instead of parent window's pixmap (helps for servers 188221 that don't have a pixmap for the root) 188222 Fix offsets for render drawing. 188223 188224commit c7bc76f663008119b3681d4b7adef7dc9ffdc236 188225Author: Torrey Lyons <torrey@mrcla.com> 188226Date: Wed Aug 11 23:53:36 2004 +0000 188227 188228 Fix leftover XF86_VERSION_SNAP macro. 188229 188230commit eb607030e32cbad846696a20cfb3045c5f8f65c1 188231Author: Adam Jackson <ajax@nwnk.net> 188232Date: Wed Aug 11 23:10:02 2004 +0000 188233 188234 Teach the loader about the extra symbols needed for Propolice-protected 188235 modules to work under elfloader. From Travis Tilley (Gentoo). 188236 188237commit fd439afdfe7ba451aff19b62d1764e4dfd0b782f 188238Author: Keith Packard <keithp@keithp.com> 188239Date: Wed Aug 11 22:40:14 2004 +0000 188240 188241 Add COMPOSITE change to fbCopyWindow (not needed yet) 188242 Xnest was half-using midispcur and doing a bad job of it. Replace all of 188243 that code with mipointer which does a lot of the work. 188244 Support DDXen which don't provide GetWindowPixmap, or which return NULL for 188245 the root pixmap. 188246 188247commit f95293e5253904883d3b40f9e68e6175247754a3 188248Author: Kevin E Martin <kem@kem.org> 188249Date: Wed Aug 11 22:27:50 2004 +0000 188250 188251 Fix bogus contact address in Xserver/os/util.c (Bug #738). 188252 188253commit 56520ecd5ceb9526541c241634b467eba6a8f7cf 188254Author: Eric Anholt <anholt@freebsd.org> 188255Date: Wed Aug 11 22:13:01 2004 +0000 188256 188257 In CopyPicture, add missing call to ChangePicture to notify about the 188258 changes that CopyPicture has done. 188259 188260commit f77f1d50723aceb5059a96f7a4068046b7961c51 188261Author: Kevin E Martin <kem@kem.org> 188262Date: Wed Aug 11 21:14:18 2004 +0000 188263 188264 Apply PPC64 build patch (Bug #303, Mike A. Harris, Alan Coopersmith). 188265 188266commit 3dbaeb2e126d9424b21df91f0be9129c4eea6f1b 188267Author: Kristian Høgsberg <krh@redhat.com> 188268Date: Wed Aug 11 20:25:13 2004 +0000 188269 188270 Patch xorgconfig to generate config files with correct keyboard driver; use 188271 "kbd" by default, "keyboard" if UseDeprecatedKeyboarDriver is YES 188272 (#1040). 188273 188274commit ca458e3c9b25c3efc10532eb8e31bdcff44dc321 188275Author: Roland Mainz <roland.mainz@nrubsig.org> 188276Date: Wed Aug 11 13:55:03 2004 +0000 188277 188278 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=492 - 188279 /etc/init.d/xprint did not source xorg.conf for font paths additionally 188280 to xfree86 config files 188281 188282commit 81881b96861d707b0729e0146a4b53b3dd234885 188283Author: Kevin E Martin <kem@kem.org> 188284Date: Tue Aug 10 21:37:35 2004 +0000 188285 188286 Update version to 6.7.99.1 and fix problem with snapshot string generation. 188287 188288commit 21c7c8cdc7d2f607177634601cef8b0290fb0e80 188289Author: Kevin E Martin <kem@kem.org> 188290Date: Mon Aug 9 22:37:22 2004 +0000 188291 188292 Fix install problem on platforms not using xorg.cf/xfree86.cf (Bug #339, 188293 Harold L. Hunt II, Alexander Gottwald). 188294 Fix crash when using X core font in zh_CN.UTF-8 locale (Bug #368, Yu Shao, 188295 David Dawes). 188296 Fix glXMakeCurrent(Dpy, None, NULL) crash (Bug #719, Adam Jackson). 188297 HP-PA build fix (Bug #828, Guy Martin, Paul Anderson). 188298 Fix SDK build for GATOS and Wacom driver (Bug #829, Bryan Stine). 188299 Fix attempt to read video ROM before enabling it (Bug #843, Ivan 188300 Kokshaysky, Mike A. Harris). 188301 Fix detection of primary adapter (Bug #843, Ivan Kokshaysky, Mike A. 188302 Harris). 188303 Clarify xset man page description of how to use the keyboard repeat rate 188304 settings (Bug #846, Mike A. Harris). 188305 Fix problem where print-screen key would get remapped to sys-req in certain 188306 keymaps, which broke GNOME printscreen functionality (Bug #847, Owen 188307 Taylor). 188308 Fix several render problems: 188309 - MMIO mode support 188310 - Hang on IGP chips 188311 - VT switching hang 188312 - 3D render corruption (Bug #922, Hui Yu). 188313 188314commit 26847ef9268f687c4f45f3d048eff4b06c2ba3bf 188315Author: Kristian Høgsberg <krh@redhat.com> 188316Date: Mon Aug 9 03:40:50 2004 +0000 188317 188318 Move the new render symbols to dixsym.c instead so they are grouped with 188319 the other render symbols. 188320 Initialize screen_x and screen_y to 0. This fixes a problem with the new 188321 Damage based sprite, where intersection test against the root pixmap 188322 would fail because the two fields were not initialized (damage.c, 188323 166-170). 188324 188325commit 57eab4dc87670da42e645a4c341d1ee661b43b5b 188326Author: Kevin E Martin <kem@kem.org> 188327Date: Mon Aug 9 02:08:35 2004 +0000 188328 188329 - Fix building without Xfixes extension 188330 - Fix building without Xevie 188331 - Fix building without DPMS 188332 - Fix "Extensions" section config file support to accept general boolean 188333 options 188334 188335commit a2e51b13ea5a2e5fc5626905b6c2caac6358eb11 188336Author: Kristian Høgsberg <krh@redhat.com> 188337Date: Sun Aug 8 20:56:56 2004 +0000 188338 188339 Disable legacy keyboard driver "keyboard", unless 188340 USE_DEPRECATED_KEYBOARD_DRIVER is defined. Print warning if the old 188341 driver is used. 188342 188343commit 3431692bf44e7be01187b421cfb0e46131c5e4c7 188344Author: Kristian Høgsberg <krh@redhat.com> 188345Date: Sun Aug 8 17:02:19 2004 +0000 188346 188347 Add new render symbols to list of exported extension symbols so 188348 DoLoadableServer works again. 188349 188350commit 1e694d2b51a954d1bc4afdde390723e9a1a5b9bf 188351Author: Keith Packard <keithp@keithp.com> 188352Date: Sat Aug 7 01:23:07 2004 +0000 188353 188354 Fix minor timestamp lossage in mieq. 188355 Replace miSprite implementation with Damage-based one. Otherwise damage and 188356 misprite conflict causing looping. 188357 Change pScreen usage around a bit to eliminate warnings 188358 188359commit 45fde3b041c071ae5a604cefbbfa4da5c1e828f8 188360Author: Keith Packard <keithp@keithp.com> 188361Date: Sat Aug 7 01:19:01 2004 +0000 188362 188363 Eliminate mfb support. Fix visuals to match "normal" PC hardware (making 188364 lightpipe work). 188365 188366commit 2bdbc2c0d39b1618440a6a8ed1c38a11703e898c 188367Author: Keith Packard <keithp@keithp.com> 188368Date: Sat Aug 7 01:02:41 2004 +0000 188369 188370 Clean up cw initialization, make sure wrappers re-wrap on exit. 188371 188372commit cff0043a347ba06e8ab16a84a35c99794b45149b 188373Author: Keith Packard <keithp@keithp.com> 188374Date: Sat Aug 7 00:58:21 2004 +0000 188375 188376 Remove alpha-related fields from visual structure to retain binary 188377 compatibilty. Applications using ARGB visuals will need to use Render 188378 or other mechanisms to compute pixel values instead of AllocColor 188379 188380commit 196aafb19a3cfdc8c21f9bf75814cf0d84ff4446 188381Author: Keith Packard <keithp@keithp.com> 188382Date: Fri Aug 6 23:42:10 2004 +0000 188383 188384 Add RenderAddTraps. Rewrite trapezoid rendering code. 188385 188386commit e847bcda0827ffb87689a0162c648570de6d6f69 188387Author: Eric Anholt <anholt@freebsd.org> 188388Date: Fri Aug 6 00:31:28 2004 +0000 188389 188390 Fix missing ';' in cw.c and unwrap the render wrapper properly. 188391 188392commit ae1580c494fde2b56f9faa40f7ebcf637728efc8 188393Author: Eric Anholt <anholt@freebsd.org> 188394Date: Thu Aug 5 18:24:58 2004 +0000 188395 188396 - Add a new Render function, CopyPicture, which will update a picture with 188397 the flagged bits from a source picture. Approved in principle by 188398 keithp. 188399 - Use CopyPicture and SetTransform to update most of the backing picture's 188400 state in the composite wrapper. Filters are still missing. 188401 - Don't allocate a picture private, now that we calculate clipping properly 188402 and don't need the serialNumber or stateChanges. 188403 - Use the format of the source pixmap rather than generating the format 188404 from the window's visual. 188405 - Wrap the rest of the Render primitives that were stubbed out before. 188406 188407commit 73e14bd611fa7eac649a2b4c7964959d9eae887b 188408Author: Eric Anholt <anholt@freebsd.org> 188409Date: Wed Aug 4 23:18:38 2004 +0000 188410 188411 - Always validate the backing GC in cwValidateGC, not just when the clip 188412 list changes. 188413 - Use FillTiled instead of FillSolid for painting tiled border/background 188414 pixmaps, and don't needlessly change the scratch GC's defaults. 188415 - Use the preferred dixChangeGC instead of DoChangeGC. 188416 - Simplify a silly loop and clean up a couple of comments. 188417 188418commit d240c41a3ab9cde9921cc96cf86e630ca5c9005a 188419Author: Egbert Eich <eich@suse.de> 188420Date: Wed Aug 4 12:21:48 2004 +0000 188421 188422 Optimization of CJK rendering when using versions of freetype prior to 188423 2.1.8 (Chisato Yamauchi). 188424 A small backward compatibility fix to make the freetype module build with 188425 freetype version < 2.1.7. 188426 Fixed/added some debugging code. 188427 188428commit 751fd11a9ad1a473d7311362246b0869a008001b 188429Author: Eric Anholt <anholt@freebsd.org> 188430Date: Wed Aug 4 10:05:37 2004 +0000 188431 188432 - Add two new XAA hooks, SetupForCPUToScreenTexture2 and 188433 SetupForCPUToScreenAlphaTexture2. These add a dstFormat argument after 188434 the previous format argument, which the driver needs to use to properly 188435 set up the destination format. Two new arrays are added for the list of 188436 destination formats supported that correspond to the previous format 188437 arrays for sources. 188438 - Make Render acceleration only occur when the new hook for that 188439 acceleration type is supplied and the dst format list is set, along 188440 with the src format list being set. Without knowing the destination 188441 format, the Render acceleration couldn't properly support all the 188442 destinations it might encounter. 188443 - Bump XAA module minor version. 188444 - Update the Radeon Render acceleration to use the new hooks when the XAA 188445 module is sufficiently new. Fix a bug in the src/dst alpha booleans for 188446 ops, and use them to set blend_cntl to support destinations without 188447 alpha. Add missing PICT_a1r5g5b5 texture format, and add list 188448 terminator. (!) 188449 188450commit 9c1d52a69db841ac85ef97d7223361b83a66ae29 188451Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 188452Date: Tue Aug 3 10:12:25 2004 +0000 188453 188454 Merge from CYGWIN branch 188455 2004-08-02 Kensuke Matsuzaki 188456 Fix the bug that we can't copy & paste multi-byte string to Unicode-base 188457 Windows application. Rename fUnicodeSupport to fUseUnicode, because it 188458 don't mean wheather Windows support Unicode or not. 188459 188460commit defcfe3c7ee3e39ef02da08b7227b758dbede325 188461Author: Matthieu Herrb <matthieu.herrb@laas.fr> 188462Date: Tue Aug 3 09:33:54 2004 +0000 188463 188464 programs/Xserver/hw/xfree86/ddc/xf86DDC.h 188465 programs/Xserver/hw/xfree86/vgahw/vgaHW.c 188466 programs/Xserver/hw/xfree86/vgahw/vgaHW.h 188467 programs/Xserver/miext/shadow/shadow.h 188468 programs/Xserver/miext/shadow/shpacked.c 188469 programs/Xserver/miext/shadow/shplanar.c 188470 programs/Xserver/miext/shadow/shrotate.c Fix glitches in pointer to 188471 functions declarations. 188472 188473commit e6d0b18c24f7f84aabed763be1cc8414883fab95 188474Author: Matthieu Herrb <matthieu.herrb@laas.fr> 188475Date: Tue Aug 3 08:52:17 2004 +0000 188476 188477 Fix declaration of XAACachePlanarMonoStippleProc. Rename 188478 XAACachePlanarMonoStippleWeak to XAAGetCachePlanarMonoStipple() for 188479 consistency with other parts of xaa and fix forward declaration in 188480 xaalocal.h. Fixes last comments on Bug #962. 188481 188482commit bfbb40c28ce6e98c82973bd96054d4787579eaa6 188483Author: Eric Anholt <anholt@freebsd.org> 188484Date: Tue Aug 3 05:49:48 2004 +0000 188485 188486 Major improvements to Composite wrapper. Several issues remain, but it now 188487 appears stable in limited testing. 188488 - Allocate the picture private, avoiding segfault. 188489 - Wrap PaintWindow to draw the background/border to the backing pixmap 188490 (based on Deron Johnson's comptran.c). 188491 - Set the x_off/y_off returns to translate coordinates properly. 188492 - Don't bother allocating temporary areas for the modified coordinates. 188493 Layers above are responsible for handling lower layers changing the 188494 arguments, so cw doesn't have to worry about it. mibstore.c has to do 188495 the allocation because it calls down twice (front buffer and backing 188496 store). (Suggested by keithp) 188497 - Handle the mode argument to PolyPoint, Polylines, and FillPolygon. 188498 - Remove some dead elements in the cw privates. 188499 - Kill a prototype warning in compinit.c by adding the cw.h header. 188500 188501commit b7ba272da0910c6558f71cfedd0bf9836fc892c1 188502Author: Eric Anholt <anholt@freebsd.org> 188503Date: Tue Aug 3 05:39:19 2004 +0000 188504 188505 The Damage extension has to wrap after (be called before) the Composite 188506 extension so that the redirecting of drawables by the wrapper doesn't 188507 confuse Damage. 188508 188509commit d112e55992e9e03e74fdf5738c8c47cd90033a37 188510Author: Adam Jackson <ajax@nwnk.net> 188511Date: Tue Aug 3 02:44:23 2004 +0000 188512 188513 Bug #962: Remove LoaderSymbol calls introduced by the dlloader work so 188514 DoLoadableServer NO builds work again. 188515 188516commit 734cb34dc1697530ecd971b84e1061ed86b4c2a4 188517Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 188518Date: Mon Aug 2 21:15:30 2004 +0000 188519 188520 Mon Aug 2 21:49:33 2004 Soeren Sandmann <sandmann@redhat.com> 188521 Remove double cast. 188522 188523commit e6b9cc79c204420117a1f7b23d131ec24923d612 188524Author: Egbert Eich <eich@suse.de> 188525Date: Mon Aug 2 19:35:07 2004 +0000 188526 188527 Removed distro specific stuff. 188528 Fixed FreeType module to build with FreeType versions older than 2.1.7. 188529 Fixed typo. 188530 Added vtSema to protect call of driver DPMS function. 188531 removed unneeded variable 188532 Modified RandR driver hook to reduce the number of function calls to one. 188533 Function is sufficiently generic to be extended in the future. 188534 188535commit b759da83ae62a897b7727d9180a68b962b571286 188536Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 188537Date: Sun Aug 1 16:16:18 2004 +0000 188538 188539 added README for Cygwin/X 188540 This file currently contains the Cygwin/X specific release notes from 6.7.0 188541 and a first collection of what will make up the release notes for the 188542 next release 188543 188544commit d638a50f3f039d84d86c00696d7d3ec22560bc3c 188545Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 188546Date: Sat Jul 31 18:33:56 2004 +0000 188547 188548 adjust prototype for winInitCmapPrivates to match Egberts change. 188549 188550commit 488be6611919af97d7e4f8b0994487882eb78e18 188551Author: Kevin E Martin <kem@kem.org> 188552Date: Sat Jul 31 09:41:27 2004 +0000 188553 188554 Fix typo 188555 188556commit 383b6b59864098b03d991628ff5933d997793ea1 188557Author: Kevin E Martin <kem@kem.org> 188558Date: Sat Jul 31 09:14:06 2004 +0000 188559 188560 Add "Extensions" section support to configuration parser 188561 188562commit d690556d496c7331bd112903a0c9e6553c7d3342 188563Author: Eric Anholt <anholt@freebsd.org> 188564Date: Sat Jul 31 08:24:14 2004 +0000 188565 188566 Integrate COMPOSITEWRAP branch including composite wrapper. This code still 188567 has several issues, including: 188568 - CopyWindow and PaintWindow wrappers missing (will be done soon) 188569 - Some segfaults seen in the Render wrappers. 188570 - Xprt server build breaks with Composite. 188571 - DDXs must be recompiled for Composite due to VisualRec size change. 188572 - Composite bugs pointed out by Deron Johnson in email. 188573 Also, reorder XFixes initialization according to comments by Keith which 188574 are also in xserver CVS. 188575 188576commit 8763cca7f9927bd6c9caf804bf09dcfea929eed0 188577Author: Eric Anholt <anholt@freebsd.org> 188578Date: Sat Jul 31 07:26:50 2004 +0000 188579 188580 Update xfixes server code to major version 3 from xserver CVS, and perform 188581 minor diff-reduction versus xserver CVS. 188582 188583commit 25bd6ff4a622d09fb0c247b7c40281744c61431e 188584Author: Kevin E Martin <kem@kem.org> 188585Date: Sat Jul 31 04:23:21 2004 +0000 188586 188587 Add new extension enable/disable feature. This code is a small step in the 188588 right direction -- i.e., moving towards full run-time config of 188589 extensions. Currently, only XTEST, XINERAMA, RENDER, XKB, and XEVIE are 188590 supported. 188591 188592commit 326729ebca863c99cf913445126294a3c6d3db5d 188593Author: Eric Anholt <anholt@freebsd.org> 188594Date: Sat Jul 31 01:48:27 2004 +0000 188595 188596 - Add some XFIXES bits apparently missed in the DAMAGE-XFIXES merge 188597 - Add missing XCSECURITY ifdef. 188598 - Sync some whitespace to xserver CVS and surrounding style. 188599 188600commit 370bda820b2912b93dcc34c088075f8e65a9e5cb 188601Author: Eric Anholt <anholt@freebsd.org> 188602Date: Sat Jul 31 01:38:46 2004 +0000 188603 188604 Integrate latest damage bits, including the addition of Composite code. 188605 Includes REGION_INIT -> REGION_NULL necessary to avoid segfaults with 188606 inlined region macros. 188607 188608commit 7542d8a17ad469f9c760f0f843bd6a4fb6deb57d 188609Author: Stuart Kreitman <stuart.kreitman@sun.com> 188610Date: Sat Jul 31 01:37:47 2004 +0000 188611 188612 Turn on XEVIE https://freedesktop.org/bugzilla/show_bug.cgi?id=947 Modified 188613 Files: miinitext.c 188614 188615commit b2065f376b932e49f1f96dfb92ddb52d15796abe 188616Author: Stuart Kreitman <stuart.kreitman@sun.com> 188617Date: Sat Jul 31 01:34:26 2004 +0000 188618 188619 Integration of XEVIE branch to trunk, latest updates 188620 https://freedesktop.org/bugzilla/show_bug.cgi?id=947 Modified Files: 188621 Imakefile xevie.c 188622 188623commit e68bfc801680f2852c59099aa3d7502e49b48b2c 188624Author: Stuart Kreitman <stuart.kreitman@sun.com> 188625Date: Sat Jul 31 01:33:40 2004 +0000 188626 188627 Integration of XEVIE branch to trunk, Some updates from latest reviews 188628 https://freedesktop.org/bugzilla/show_bug.cgi?id=947 Modified Files: 188629 events.c 188630 188631commit 706b2a7e02aa0085769fb87782118488d0f90eaa 188632Author: Adam Jackson <ajax@nwnk.net> 188633Date: Sat Jul 31 01:21:19 2004 +0000 188634 188635 Change several LoaderSymbol calls introduced by the bug #400 patch to 188636 *Weak() resolver functions. 188637 188638commit 9e13805b02f37497971c789b4035abc29463c550 188639Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 188640Date: Sat Jul 31 00:44:45 2004 +0000 188641 188642 adjust prototype for winInitCmapPrivates to match Egberts change. 188643 188644commit 4da507a03fd5659e5944d9e47dd2f8920636383b 188645Author: Kevin E Martin <kem@kem.org> 188646Date: Sat Jul 31 00:32:43 2004 +0000 188647 188648 Fix compiler warning 188649 188650commit 64a6d3e9c84a36b4e0550d112f288b695c5056cd 188651Author: Hui YU <hyu@ati.com> 188652Date: Fri Jul 30 22:20:21 2004 +0000 188653 188654 Support for New radeon chips: R420/M18, R423, RV370/M22, RV380/M24, RS300. 188655 Add special handlings for DELL triple-head server (RV100). Misc. bug 188656 fixes for flat panel, host aperture, etc (Bug #946) 188657 188658commit d3c98fed2c37a863a6765a3e288bcdbc2738878a 188659Author: Egbert Eich <eich@suse.de> 188660Date: Fri Jul 30 21:53:09 2004 +0000 188661 188662 Add support for on-the-fly screen rotation when supported by hardware (Aron 188663 Plattner). 188664 188665commit 4baf0029418d3eeeac5d1026a7cfea3234e44e48 188666Author: Egbert Eich <eich@suse.de> 188667Date: Fri Jul 30 21:46:38 2004 +0000 188668 188669 An experimental pseudocolor emulation layer. Not fully completed, currently 188670 only works for 16bpp. 188671 188672commit 63a152f7812d0981e3e7aa41a42e59cd0c3e50dc 188673Author: Egbert Eich <eich@suse.de> 188674Date: Fri Jul 30 21:39:20 2004 +0000 188675 188676 removing some unnecessary restricitons on the allowed visuals. 188677 188678commit 0ba15599466ddb644728c6b68e64e05b1317ac2e 188679Author: Egbert Eich <eich@suse.de> 188680Date: Fri Jul 30 21:10:46 2004 +0000 188681 188682 Adding a colormap index to the InitColormapPrivate() func call. Without it 188683 it was completely useless. 188684 test if colormap with index really exists in the list of installed maps 188685 before using it. 188686 188687commit bbfe7bed3fe4d9bd089327cd59e6faedb592dabb 188688Author: Egbert Eich <eich@suse.de> 188689Date: Fri Jul 30 20:56:53 2004 +0000 188690 188691 Set DPMS to ON when VT switching away or shutting down the server. Failing 188692 to do this may leave the text console blank. 188693 188694commit 40b975e3acb11c1ec2fd4c5984f5efa20b669489 188695Author: Egbert Eich <eich@suse.de> 188696Date: Fri Jul 30 20:51:09 2004 +0000 188697 188698 Improved error messages. 188699 build fixes for AMD64. 188700 Made shm* functions in the libc_wrapper more standard conformant by setting 188701 errno correctly. 188702 Use xf86ExendedInitInt10() in VBEExtendedInit() to be able to pass flags. 188703 188704commit 7643199de1c4f12a2aadeaf2d539a37ddb45672b 188705Author: Egbert Eich <eich@suse.de> 188706Date: Fri Jul 30 20:38:27 2004 +0000 188707 188708 Fix static build. 188709 add i845 to the list of chips that allow memory size tweaking. 188710 188711commit 48514fee3c8ec26f36e142ffc9272e510b9a4238 188712Author: Adam Jackson <ajax@nwnk.net> 188713Date: Fri Jul 30 20:30:57 2004 +0000 188714 188715 Bug #400 (partial): Driver fixes for the dlloader. When using dlloader, all 188716 framebuffer formats except cfb and the overlay modes should work, and 188717 r128 and radeon need to be loaded from the ati driver (both issues to 188718 be fixed soon). Tested on i740, s3virge, mach64, tdfx, vesa, and vga 188719 drivers. elfloader users shouldn't be affected. 188720 188721commit 29012adb37c533f57c684ad94c4d83a6c31793e5 188722Author: Torrey Lyons <torrey@mrcla.com> 188723Date: Fri Jul 30 19:28:03 2004 +0000 188724 188725 Add generic rootless layer documentation missed from last commit. 188726 188727commit 784e4d1cc02dea837a38a4140a18013953296366 188728Author: Torrey Lyons <torrey@mrcla.com> 188729Date: Fri Jul 30 19:12:18 2004 +0000 188730 188731 Merge many XDarwin improvements: 188732 - Fix launch of X clients by double clicking in the Finder when there is a 188733 space in the path (Torrey T. Lyons). 188734 - Interpret scroll wheel mouse events correctly when shift is held down 188735 (Benjamin Burke). 188736 - Add option to always use Mac command key equivalents (John Harper and 188737 Torrey T. Lyons). 188738 - Add support for dynamic screen configuration changes in rootless mode 188739 (John Harper and Torrey T. Lyons). 188740 - Add documentation on generic rootless layer (Torrey T. Lyons). 188741 188742commit c2275b31adc3c4292c171055db16e00ee0e69e43 188743Author: Egbert Eich <eich@suse.de> 188744Date: Fri Jul 30 19:04:14 2004 +0000 188745 188746 Removed bugs on TweakMemorySize() which prevented it from working at all. 188747 Initialized last element of BIOS version number string to 0 to avoid random 188748 problems. 188749 188750commit 5ed0aefc67e86abaddf1b6bffdc832996b86fb46 188751Author: Egbert Eich <eich@suse.de> 188752Date: Fri Jul 30 18:40:36 2004 +0000 188753 188754 Fixed typo in a comment. 188755 Deleted bogus comment. 188756 Added debugging support. 188757 188758commit d380647739e4767da69edc44bbb441b3b9554b03 188759Author: Torrey Lyons <torrey@mrcla.com> 188760Date: Fri Jul 30 18:22:13 2004 +0000 188761 188762 Add initial Xinput support for XDarwin (Greg Parker). 188763 188764commit 41641c11ec8994f1bc4bd1b05ae2cb38167c8312 188765Author: Torrey Lyons <torrey@mrcla.com> 188766Date: Fri Jul 30 17:37:09 2004 +0000 188767 188768 Make XDarwin not default to StaticColor on ix86 (Shantonu Sen). 188769 188770commit f4c84e7dbf0f25a2544d4400e600310421683f3c 188771Author: Roland Mainz <roland.mainz@nrubsig.org> 188772Date: Fri Jul 30 12:03:56 2004 +0000 188773 188774 Fix for http://xprint.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=940 188775 - Filter /etc/init.d/xprint throught CPP/|MakeScriptFromCpp()| to allow 188776 platform-specific customisation. 188777 188778commit ed923a42b800e3f701084ef5828cabbbefe8361f 188779Author: Eric Anholt <anholt@freebsd.org> 188780Date: Fri Jul 30 06:54:41 2004 +0000 188781 188782 file cw_render.c was initially added on branch COMPOSITEWRAP. 188783 188784commit 69e379ddaf85feda6a163b5f4e855ffe86b254df 188785Author: Eric Anholt <anholt@freebsd.org> 188786Date: Fri Jul 30 06:54:41 2004 +0000 188787 188788 file cw_ops.c was initially added on branch COMPOSITEWRAP. 188789 188790commit 0ee93acd103af947fef0a28e32b5b3f2857789bf 188791Author: Eric Anholt <anholt@freebsd.org> 188792Date: Fri Jul 30 06:54:40 2004 +0000 188793 188794 file cw.c was initially added on branch COMPOSITEWRAP. 188795 188796commit 3bdaa0e7fa7b6f5d550907d5fe7d0fb16c53e243 188797Author: Eric Anholt <anholt@freebsd.org> 188798Date: Fri Jul 30 06:54:40 2004 +0000 188799 188800 file cw.h was initially added on branch COMPOSITEWRAP. 188801 188802commit 94e1ea569171334eb40a2d4a63138c02915203db 188803Author: Eric Anholt <anholt@freebsd.org> 188804Date: Fri Jul 30 06:54:40 2004 +0000 188805 188806 file compwindow.c was initially added on branch COMPOSITEWRAP. 188807 188808commit 705536d04c4f09c84bb04827c07bb899584f399d 188809Author: Eric Anholt <anholt@freebsd.org> 188810Date: Fri Jul 30 06:54:40 2004 +0000 188811 188812 file compinit.c was initially added on branch COMPOSITEWRAP. 188813 188814commit b80dbd886d7cc3a72772f3231a8c8e8df7f6679f 188815Author: Eric Anholt <anholt@freebsd.org> 188816Date: Fri Jul 30 06:54:40 2004 +0000 188817 188818 file compalloc.c was initially added on branch COMPOSITEWRAP. 188819 188820commit beb26caf68d3e25bf85fd63dbb499eca4b1f05ba 188821Author: Eric Anholt <anholt@freebsd.org> 188822Date: Fri Jul 30 06:54:40 2004 +0000 188823 188824 file compext.c was initially added on branch COMPOSITEWRAP. 188825 188826commit 97afc846003bb521cf9d6e92b298024d83db8759 188827Author: Eric Anholt <anholt@freebsd.org> 188828Date: Fri Jul 30 06:54:40 2004 +0000 188829 188830 file compint.h was initially added on branch COMPOSITEWRAP. 188831 188832commit 854c1afa867ff617b47c4cde3cfd86bd26e9931a 188833Author: Kevin E Martin <kem@kem.org> 188834Date: Fri Jul 30 04:44:13 2004 +0000 188835 188836 Fix "DoLoadableServer NO" build 188837 188838commit 71164d118c192e96eb2b0fc45514233e9563a568 188839Author: Stuart Kreitman <stuart.kreitman@sun.com> 188840Date: Fri Jul 30 01:21:57 2004 +0000 188841 188842 Integration of XEVIE branch to trunk 188843 https://freedesktop.org/bugzilla/show_bug.cgi?id=947 188844 Modified Files: cursorstr.h input.h inputstr.h windowstr.h 188845 188846commit 82a6a659248bb66a0364eb9eaf331747834fb5c6 188847Author: Stuart Kreitman <stuart.kreitman@sun.com> 188848Date: Fri Jul 30 01:20:42 2004 +0000 188849 188850 Integration of XEVIE branch to trunk 188851 https://freedesktop.org/bugzilla/show_bug.cgi?id=947 188852 Modified Files: Imakefile Added Files: xevie.c 188853 188854commit dd7077cadcdc323e1a301ed5eefa7ff12599aa4d 188855Author: Kevin E Martin <kem@kem.org> 188856Date: Fri Jul 30 01:15:57 2004 +0000 188857 188858 Cleaned up code since client's saveSet is no longer defined as a pointer* 188859 (forgot one change in previous check-in) 188860 188861commit 813d75f9d3c7b540977926e37310fa683daf12e1 188862Author: Kevin E Martin <kem@kem.org> 188863Date: Thu Jul 29 23:43:40 2004 +0000 188864 188865 Use LibraryTargetName when not building a loadable server 188866 Move extern function declarations to window.h 188867 Cleaned up code since client's saveSet is no longer defined as a pointer* 188868 Added externs back in 188869 Change #if to #ifdef to fix compiler warning 188870 Add function declarations that were inadvertently removed by previous check 188871 in 188872 Disable extensions that are not (yet) supported by DMX 188873 188874commit 274d5044ac41523ff23912c223177c429c710e09 188875Author: Stuart Kreitman <stuart.kreitman@sun.com> 188876Date: Thu Jul 29 18:49:42 2004 +0000 188877 188878 Integration of DAMAGE-XFIXES branch to trunk 188879 https://freedesktop.org/bugzilla/show_bug.cgi?id=859 188880 Modified Files: Imakefile miinitext.c 188881 188882commit 406c49eb810cbdcfd833cac4eeaa465598238691 188883Author: Stuart Kreitman <stuart.kreitman@sun.com> 188884Date: Thu Jul 29 18:46:37 2004 +0000 188885 188886 Integration of DAMAGE-XFIXES branch to trunk 188887 https://freedesktop.org/bugzilla/show_bug.cgi?id=859 188888 DAMAGE calls some shape functions. Modified Files: Imakefile shape.c 188889 188890commit 0bca00e1205bf1a4537cbf7be6339b3b1f9b953f 188891Author: Stuart Kreitman <stuart.kreitman@sun.com> 188892Date: Thu Jul 29 18:43:58 2004 +0000 188893 188894 Integration of DAMAGE-XFIXES branch to trunk 188895 https://freedesktop.org/bugzilla/show_bug.cgi?id=859 188896 Modified Files: dispatch.c dixutils.c events.c window.c 188897 188898commit d4a101d4ef9943dcddf08b00b2d3ab4319597193 188899Author: Stuart Kreitman <stuart.kreitman@sun.com> 188900Date: Thu Jul 29 18:37:54 2004 +0000 188901 188902 Integration of DAMAGE-XFIXES branch to trunk 188903 https://freedesktop.org/bugzilla/show_bug.cgi?id=859 188904 These RENDER changes come from the experimental freedesktop tree formerly 188905 known as "Xserver". Partly motivated by compatibility with DAMAGE as 188906 pulled from that tree, also some of the code just is better 188907 implemented. 188908 Modified Files: filter.c picture.c picture.h picturestr.h 188909 188910commit e1281790bb3d7cdcc5de85829806dd53da67e326 188911Author: Stuart Kreitman <stuart.kreitman@sun.com> 188912Date: Thu Jul 29 18:16:56 2004 +0000 188913 188914 Integration of DAMAGE-XFIXES branch to trunk 188915 https://freedesktop.org/bugzilla/show_bug.cgi?id=859 188916 Modified Files: cursorstr.h dix.h dixstruct.h regionstr.h window.h 188917 188918commit d2f798b6dbaebd0300f42c2e083a962c37647620 188919Author: Stuart Kreitman <stuart.kreitman@sun.com> 188920Date: Thu Jul 29 14:42:24 2004 +0000 188921 188922 Integration of DAMAGE-XFIXES branch to trunk 188923 https://freedesktop.org/bugzilla/show_bug.cgi?id=859 188924 Added Files: Imakefile damage.c damage.h damagestr.h 188925 188926commit 682ee8a9f8d55c6a6f517a277d1bfad2a0f28594 188927Author: Stuart Kreitman <stuart.kreitman@sun.com> 188928Date: Thu Jul 29 14:40:33 2004 +0000 188929 188930 Integration of DAMAGE-XFIXES branch to trunk 188931 https://freedesktop.org/bugzilla/show_bug.cgi?id=859 188932 188933commit 6ed1c3845517c7731dcc74baa51bb801838afaff 188934Author: Stuart Kreitman <stuart.kreitman@sun.com> 188935Date: Thu Jul 29 14:33:43 2004 +0000 188936 188937 bugzilla 859 -merging DAMAGE-XFIXES into trunk 188938 188939commit 736e511824c4f5e77e637c680a4e45f0b7631644 188940Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 188941Date: Wed Jul 28 03:57:19 2004 +0000 188942 188943 Fix shared reqs for Xlibi18n, Xaw6 & Xmu for Solaris 188944 Improved support for Compose and Kana Lock LED's on Sun keyboards 188945 Add event definitions for additional keys on Sun Japanese keyboards 188946 VUID mouse protocol support for Solaris Make default mouse message clearer 188947 on machines that use a default other than /dev/mouse. 188948 188949commit d374dffb20364a3df184cc28888ee7edbe50f474 188950Author: Rik Faith <faith@alephnull.com> 188951Date: Wed Jul 28 00:28:00 2004 +0000 188952 188953 When using DMX console input, make detached screens display as grey (vs. 188954 white for attached and black for dead-space). 188955 188956commit 1498414cc85c148ef15b9b57d9f9b6b850bf2702 188957Author: Torrey Lyons <torrey@mrcla.com> 188958Date: Tue Jul 27 20:26:47 2004 +0000 188959 188960 Fix Xprint build on Darwin. 188961 188962commit ddd58356123f61b863754eb0bdd42a8defa0461b 188963Author: Torrey Lyons <torrey@mrcla.com> 188964Date: Tue Jul 27 20:24:00 2004 +0000 188965 188966 Fix XDarwin's broken build of libGL and server side GLX. GL library is only 188967 known to build correctly on Mac OS X 10.3.4 and still needs some work 188968 for earlier versions. 188969 188970commit 8ef3e7052e8e1db869411e9f6fc88491e3f05474 188971Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 188972Date: Tue Jul 27 09:53:14 2004 +0000 188973 188974 Merge latest changes from CYGWIN branch 188975 Use find | xargs combination instead of simple shell globbing to prevent 188976 commandline argument overflow 188977 188978commit a12a678bc9af8df1a3078f574b0c910e3e6983f8 188979Author: Matthieu Herrb <matthieu.herrb@laas.fr> 188980Date: Tue Jul 27 06:20:52 2004 +0000 188981 188982 Fix a bug that caused an entire batch of events to be thrown out when one 188983 of an unknown type is encountered. (Ty Sarna, NetBSD). 188984 188985commit 19002e47a0dbf55e035b674116a533ec9814edfa 188986Author: Alex Deucher <agd5f@yahoo.com> 188987Date: Tue Jul 27 02:50:41 2004 +0000 188988 188989 - enable maven support for TV detection, DPMS, and DDC on crtc2 on G400 188990 (Ryan Underwood) (http://bugs.xfree86.org/show_bug.cgi?id=1098) 188991 - expose I2CStart; needed for mga maven support (Ryan Underwood) 188992 188993commit 2a7b137d41eb8ce6efc45b47b5df0c89eb4f5d93 188994Author: Matthieu Herrb <matthieu.herrb@laas.fr> 188995Date: Mon Jul 26 22:41:47 2004 +0000 188996 188997 remove extra ';' (Alan Hourihane, Keith Packard). 188998 188999commit 799208dd44a65b18dda97b4843a27a2628f955f4 189000Author: Adam Jackson <ajax@nwnk.net> 189001Date: Mon Jul 26 19:06:04 2004 +0000 189002 189003 Bug #377: Make lib{glx,GLcore,dri} work when compiled as dlloader modules. 189004 189005commit f15f881727cee9a879bd43be8dc849320f8d3cbd 189006Author: Keith Packard <keithp@keithp.com> 189007Date: Mon Jul 26 17:14:27 2004 +0000 189008 189009 Eliminate bogus rate check in fbdevModeSupported. Hmm. Potentially bogus 189010 rate selection necessary for Mac fbdev 189011 Don't know about fb changes to pixmaps, so can't track dirt. 189012 Add Mac specific 1280x854 mode. Warn when requested mode isn't found. 189013 Add ability to soft-boot video cards. 189014 Add region expand request. FIXME: need test cases 189015 189016commit 20913b7d5daf90e0f7ad1ee967ad2f0daaec40f9 189017Author: Matthieu Herrb <matthieu.herrb@laas.fr> 189018Date: Sat Jul 24 17:35:39 2004 +0000 189019 189020 Fix a problem with wsmouse driver loosing events on 64bit architectures 189021 (XFree86 Bugzilla #1438, John Heasley). "To fix this, I've added a 189022 mouse buffer (Xisb buffer) "scale" value to the MouseDevPtr type. If 189023 set, it is used as structure size of which we want space for a few." 189024 189025commit c57944cd9aaac717d4d4ada44626e35925b39bbd 189026Author: Keith Packard <keithp@keithp.com> 189027Date: Sat Jul 24 17:02:49 2004 +0000 189028 189029 Check for mmio before restoring crtc/crtc2 pitch registers 189030 189031commit 5fdff8b95e8f90221a46717c2f84715ab238460c 189032Author: Matthieu Herrb <matthieu.herrb@laas.fr> 189033Date: Sat Jul 24 16:32:39 2004 +0000 189034 189035 Bugzilla #884: OpenBSD/amd64 support. 189036 189037commit cc3e0173d9fae8a40eb46606d9951e3aa1df975a 189038Author: Søren Sandmann Pedersen <sandmann@daimi.au.dk> 189039Date: Thu Jul 22 19:24:50 2004 +0000 189040 189041 Thu Jul 22 20:03:11 2004 Soeren Sandmann <sandmann@daimi.au.dk> 189042 Call MMX solid fill routine when available. 189043 Call MMX operations when available. 189044 New HasGcc34 macro 189045 New file with many operations implemented with MMX intrinsics, conditional 189046 on having GCC 3.4 on i386. 189047 189048commit 9565d9e0cf85e6f5fb47acebdd66212bd6cc3e08 189049Author: Keith Packard <keithp@keithp.com> 189050Date: Thu Jul 22 18:17:59 2004 +0000 189051 189052 Correct pitch so that accelerator can run on 1400x1050 screens. Add a few 189053 more register sets for cursors. 189054 189055commit 67dbad6b3b9163eafae7d9dd7698708e10372a21 189056Author: Eric Anholt <anholt@freebsd.org> 189057Date: Thu Jul 22 06:48:19 2004 +0000 189058 189059 DRM 20040721 import 189060 189061commit 829b2c72a6433ebaf63f2d2726259c73cca4bd1a 189062Author: Eric Anholt <anholt@freebsd.org> 189063Date: Thu Jul 22 06:48:19 2004 +0000 189064 189065 Initial revision 189066 189067commit 448e0754e369d433a61ae337bbfd7dba195c5e69 189068Author: Phil Blundell <pb@reciva.com> 189069Date: Wed Jul 21 20:33:35 2004 +0000 189070 189071 Include -lts if appropriate. Patch from pattieja@bentham.ispvip.biz. 189072 189073commit 0c32a94623b13dd1ac5b015b465bdf890f498282 189074Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189075Date: Tue Jul 20 15:15:13 2004 +0000 189076 189077 Bugzilla #889: Bind -from address to port number 0 instead of 177 189078 189079commit 07e6011106dcfa0ab69861aa7dcbb88382625c16 189080Author: Phil Blundell <pb@reciva.com> 189081Date: Tue Jul 20 14:33:42 2004 +0000 189082 189083 Select optimized 189084 16bpp shadow copy functions if screen is 16bpp. Select -YX versions for 189085 90 and 270 rotations if architecture is ARM. 189086 189087commit 797114414096d7bf7ed0d73a878d0cffef262301 189088Author: Roland Mainz <roland.mainz@nrubsig.org> 189089Date: Mon Jul 19 22:01:52 2004 +0000 189090 189091 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=893 - Fixing the 189092 bug that Xprt did not honor ${LC_ALL} when looking for model-config 189093 dirs. 189094 189095commit 8853f9331826899229e5b7c964e9c852c0371ce5 189096Author: Eric Anholt <anholt@freebsd.org> 189097Date: Mon Jul 19 12:07:01 2004 +0000 189098 189099 Add support for a8b8g8r8 and x8b8g8r8 pictures, which showed up frequently 189100 with metacity usage. 189101 189102commit adf4b38f49da063576b48f9c0750b78bb753b3de 189103Author: Eric Anholt <anholt@freebsd.org> 189104Date: Mon Jul 19 11:42:49 2004 +0000 189105 189106 Breakage in last commit to this file: pCurPriv->area isn't set up until 189107 Enable, these days. 189108 189109commit d2e74e419d2c75c5a5b97236d2714730e6f69ee0 189110Author: Eric Anholt <anholt@freebsd.org> 189111Date: Mon Jul 19 11:19:12 2004 +0000 189112 189113 Set the right number of texture coordinates for r200 Render support (still 189114 disabled, needs to be tested). 189115 189116commit cefcb7f123c962c3715b0c46f30430e87c82a017 189117Author: Eric Anholt <anholt@freebsd.org> 189118Date: Mon Jul 19 11:16:13 2004 +0000 189119 189120 - Add Radeon picture transform support. 189121 - On R128, don't refer to an old Composite's mask transform when the 189122 current Composite doesn't have a mask. 189123 - Staticize some global variables in r128_composite.c. 189124 189125commit 9fe216a45836b98b5aea55725019668de3900e83 189126Author: Eric Anholt <anholt@freebsd.org> 189127Date: Mon Jul 19 07:53:54 2004 +0000 189128 189129 Use the offscreen memory manager as much as possible to do the reservation 189130 of memory at startup. Do some drive-by cleanups while I'm here 189131 (sorry!). 189132 189133commit e2bb9f38cdcb74fd7630f4efe310ad14d597171e 189134Author: Eric Anholt <anholt@freebsd.org> 189135Date: Mon Jul 19 07:20:01 2004 +0000 189136 189137 Add support for transforms of textures on R128. 189138 189139commit c04264727860cbe2e276e1934d6700d0baaf9f73 189140Author: Rik Faith <faith@alephnull.com> 189141Date: Sun Jul 18 22:19:33 2004 +0000 189142 189143 Addition of console input after removal of core backend input that is not 189144 on screen 0 can cause a segfault. Fix by preventing reinitialization of 189145 detached inputs. When Xinerama is active and screen 0 is detached, 189146 pixmaps for XGetImage must be obtained from another screen. 189147 189148commit 0addd0d499046fc8a6cdc18fc41d34cac2ba77ea 189149Author: Roland Mainz <roland.mainz@nrubsig.org> 189150Date: Sun Jul 18 05:21:20 2004 +0000 189151 189152 Fix for http://freedesktop.org/bugzilla/show_bug.cgi?id=858 - Fixing the 189153 problem that clients cannot use the TrueColor visual without calling 189154 |XInstallColormap();|. 189155 189156commit 7f9e263658b1c9c435db851a8afd904a2f584d13 189157Author: Kevin E Martin <kem@kem.org> 189158Date: Sat Jul 17 20:44:14 2004 +0000 189159 189160 Revert change to MAXFORMATS to maintain binary compatibility 189161 189162commit c47a1bdd7463b6863018e2c4237acfd28b89f38f 189163Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189164Date: Sat Jul 17 01:13:31 2004 +0000 189165 189166 Fix typo in debug message in MakeAllCLTSServerListeners 189167 Add $(GETPEER_DEFINES) to DEPEND_DEFINES for makedepend 189168 Add "localuser" and "localgroup" access types to server-interpreted 189169 authentication scheme. 189170 189171commit 3e52373fc8179a59efc9e7ab22ce0cb5160d0409 189172Author: Carlos Romero <kcrashcore@bellsouth.net> 189173Date: Thu Jul 15 09:56:40 2004 +0000 189174 189175 Enable i810. 189176 189177commit aeb78eaa980ac93f6af3e947ab1ad8bce5bc5bd1 189178Author: Eric Anholt <anholt@freebsd.org> 189179Date: Thu Jul 8 08:21:25 2004 +0000 189180 189181 Commit the rest of the dirty optimization from 20040703 and add a missing 189182 header include. I don't know how this happened, but I'm going to blame 189183 a "few" fscks having happened between testing the code and committing 189184 (due to other driver changes that made me not blame dirty for the 189185 failures I saw) for disappearance of kasync.c changes. Fixes a lot of 189186 corruption. 189187 189188commit b46767352822b09e5dab8b54cbb7a37a9b62de53 189189Author: Eric Anholt <anholt@freebsd.org> 189190Date: Thu Jul 8 06:57:58 2004 +0000 189191 189192 Remove duplicate protos in kdrive.h. 189193 189194commit 764d9e822b01fdfe3ff088028959cbc48b349026 189195Author: Carlos Romero <kcrashcore@bellsouth.net> 189196Date: Wed Jul 7 20:43:25 2004 +0000 189197 189198 missing xkb/[xkb.h, xkbDflts.h] and hw/kdrive/mga/g400_common.h 189199 189200commit f96ef08d48b01f6119799ede9ffc43d0134cbf8b 189201Author: Carlos Romero <kcrashcore@bellsouth.net> 189202Date: Wed Jul 7 19:21:07 2004 +0000 189203 189204 Initial kdrive XKB/XINPUT support, use --enable-xkb --enable-xinput 189205 189206commit a5c9b3229ce418a5e3eacc40b7a7f11c0a26d958 189207Author: Rik Faith <faith@alephnull.com> 189208Date: Wed Jul 7 04:32:52 2004 +0000 189209 189210 Bugzilla #817 189211 189212commit 1498d7a096f0855fa965585acd9ca4a2780cc959 189213Author: Kevin E Martin <kem@kem.org> 189214Date: Tue Jul 6 23:51:00 2004 +0000 189215 189216 - Disable building DMX on OSs that have not been verified to build 189217 correctly (Kevin Martin). 189218 - Fix DMX build when Xinerama is not enabled (Kevin Martin). 189219 189220commit 0e45f2a7536bf4b66d6f64d96b44431310884af3 189221Author: Egbert Eich <eich@suse.de> 189222Date: Tue Jul 6 14:49:13 2004 +0000 189223 189224 ifdef'ed some IA32-only assembler statements. Presently these chipsets are 189225 IA32 only, anyway (Egbert Eich). 189226 Disabling generic VGA testing for IA64 architectures. Temporarily disabling 189227 support for ZX1 bus. This code is extremely invasive and is executed as 189228 fallback without testing for a ZX1 chipset. It brings a SGI Altrix to a 189229 grinding halt. (Egbert Eich). 189230 189231commit df2b55a25b7056ac92c1f6cbee9f16bd0a37ba8c 189232Author: Egbert Eich <eich@suse.de> 189233Date: Tue Jul 6 14:37:48 2004 +0000 189234 189235 Separated Intel drivers from default DriDrivers to avoid building them on 189236 IA64 (Egbert Eich). 189237 Fixed wrong function prototype (Egbert Eich). 189238 Don't test for generic VGA on IA64 (Egbert Eich). 189239 Fixed a segfault when accessing a structure before verifying the pointer 189240 exists (Egbert Eich). 189241 Added a showcache option for debugging (Egbert Eich). 189242 Increase default video RAM size to 16MB when DRI is enabled and more than 189243 128MB are available (Egbert Eich). Fixed lockups during mode switch. 189244 Problem was introduced when attempting to copy the behavior during 189245 LeaveVT()/EnterVT() but but forgetting to call I810DRILeave() before 189246 I810DRIEnter(). The entire DRILeave()/Enter() scenario has been 189247 commented out as it didn't seem to be necessary (Egbert Eich). 189248 Fix TweakMemorySize() (tested with i855/i865) (Egbert Eich). 189249 increased MAX_DEVICES to 128 (Egbert Eich). 189250 Use OS provided PCI config space access as default method (Egbert Eich). 189251 Added support for Linux 2.6 proc file format. 189252 Fixed unaligned accesses to pieces of the VBE info block. VESA did not 189253 align elements to size (Egbert Eich). 189254 189255commit 7c466d64c34e68e0bc50e083861874161ae02db9 189256Author: Eric Anholt <anholt@freebsd.org> 189257Date: Sat Jul 3 10:23:03 2004 +0000 189258 189259 Clean up Rage 128 composite code. Now it composites more operations 189260 correctly and is simpler. 189261 189262commit 020701566916c8569f5af7f2efe1de36fea2002e 189263Author: Eric Anholt <anholt@freebsd.org> 189264Date: Sat Jul 3 09:16:30 2004 +0000 189265 189266 Add a "dirty" flag to the pixmap private. Clear it when setting up an 189267 offscreen pixmap area, and set it when any rendering occurs. When 189268 moving a pixmap out of offscreen, don't read data back if it wasn't 189269 dirtied (compared to the system memory copy). 189270 189271commit fd594b0559caa98ee0823be956aecf9c9d2e52bc 189272Author: Phil Blundell <pb@reciva.com> 189273Date: Fri Jul 2 21:30:00 2004 +0000 189274 189275 Call ts_read multiple times, to avoid events getting stuck in the pipeline. 189276 189277commit 7976ee51afcad41b611e642d2feb31d805dedcf6 189278Author: Kevin E Martin <kem@kem.org> 189279Date: Wed Jun 30 20:06:56 2004 +0000 189280 189281 Add Distributed Multihead X (DMX) support 189282 189283commit d5db59bd79f5d8788b99056bf9d969b5b3ad99e1 189284Author: Eric Anholt <anholt@freebsd.org> 189285Date: Tue Jun 29 20:37:51 2004 +0000 189286 189287 Add an offscreen area scoring to improve choosing offscreen areas to kick 189288 out when allocation can't find a free area of the requested size. When 189289 offscreen pixmaps get used, the offscreen area's score is increased by 189290 a constant value. Every certain number of increases, all offscreen area 189291 scores get decreased by a fraction. When choosing a set of areas to 189292 remove for a new allocation, the set of areas with the smallest total 189293 score is chosen for removal. While this is not the smartest system, it 189294 prevents things like always removing the first offscreen area in memory 189295 (likely the most recent) to be kicked out when doing replacing. 189296 189297commit ea78d1c6fcd27d28e69cb97faf72b7b719f6f93e 189298Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189299Date: Mon Jun 28 18:08:26 2004 +0000 189300 189301 Add GLX_ALIAS_VOID for GLX_ALIAS of functions with return type void to fix 189302 builds with non-gcc compilers that refuse to let you do return 189303 function_that_returns_void(...) 189304 programs/Xserver/hw/xfree86/os-support/shared/sigiostubs.c Remove includes 189305 of xf86drm.h that break non-DRI builds 189306 189307commit 7ff67f2872ddd15908f789ec9bdb76e8211d6431 189308Author: Keith Packard <keithp@keithp.com> 189309Date: Mon Jun 28 00:48:51 2004 +0000 189310 189311 Separate out off-screen allocation from Init. Fix Enable to update 189312 off-screen addresses. Wrap RandR to update off-screen addresses. 189313 Set off_screen_base and memory_size fields correctly. 189314 189315commit 5b75aae2cf1ad38556e9a55da72ad65419aa7f84 189316Author: Keith Packard <keithp@keithp.com> 189317Date: Sat Jun 26 04:13:03 2004 +0000 189318 189319 Add ARGB cursor support for Radeon cards. 189320 189321commit 8bc0bc6d36dbc5000069017a1984905065164016 189322Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189323Date: Fri Jun 25 08:58:18 2004 +0000 189324 189325 #Bug 784: Ignore unconfigured interfaces with xdmcp 189326 189327commit c5ab3fdd928d12b4dc28108f2242b3b75e1ac65f 189328Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189329Date: Fri Jun 25 08:56:04 2004 +0000 189330 189331 #Bug 780: add RRSetScreenConfig 189332 189333commit f8226cee08a00b49f32dc3db814478490febe45d 189334Author: Roland Mainz <roland.mainz@nrubsig.org> 189335Date: Fri Jun 25 00:02:11 2004 +0000 189336 189337 Fix for http://xprint.freedesktop.org/bugzilla/show_bug.cgi?id=791 - Adding 189338 special support for Canon C3200N 189339 189340commit ad6b9644a39343437967b4c3b2442dbd47508443 189341Author: Roland Mainz <roland.mainz@nrubsig.org> 189342Date: Thu Jun 24 06:26:27 2004 +0000 189343 189344 Fix for http://xprint.freedesktop.org/bugzilla/show_bug.cgi?id=660 : Fix 189345 for the issue that GetPrinterList does not return printer descriptions 189346 on Solaris. The patch implements a framework which allows the printer 189347 enumerator scripts to pass additional printer attributes to the 189348 information pool (currently only "xp-printerattr.descriptor" is 189349 implemented). 189350 189351commit 884908a63c624585c9b5fcf22d565236298c2916 189352Author: Roland Mainz <roland.mainz@nrubsig.org> 189353Date: Tue Jun 22 10:18:13 2004 +0000 189354 189355 Fix for http://xprint.freedesktop.org/bugzilla/show_bug.cgi?id=789 : Adding 189356 a workaround for the issue that Xprt may hang when the CUPS spooler 189357 frontend sends messages to stdout. 189358 189359commit c66cc2a219e860ae3c0b5d4ad18b22a6dc4e16df 189360Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189361Date: Mon Jun 21 13:51:57 2004 +0000 189362 189363 Bug 783: rootless patches for cygwin 189364 189365commit ed7f92e791f052d64cffef4b44eae5160fb24689 189366Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189367Date: Mon Jun 21 13:44:14 2004 +0000 189368 189369 Bug 778: add ddxBeforeReset 189370 189371commit 68d92cca1a696521599db6a826d2187ec0c15f01 189372Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189373Date: Mon Jun 21 13:35:05 2004 +0000 189374 189375 Bug 782: Merge native OpenGL for Windows from CYGWIN branch 189376 189377commit d6e8b1affec7351549c0006cc63b6923091cdd68 189378Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189379Date: Mon Jun 21 13:19:32 2004 +0000 189380 189381 Bug 777: Merge from CYGWIN branch 189382 189383commit dfdbb60bf5f613b3554d5435f08f16bde72aa353 189384Author: Roland Mainz <roland.mainz@nrubsig.org> 189385Date: Mon Jun 21 00:29:46 2004 +0000 189386 189387 Fix for http://xprint.freedesktop.org/bugzilla/show_bug.cgi?id=772 - RFE: 189388 Switch default resolution from 300DPI to 600DPI 189389 189390commit dd831c7a5c1b0c540a78350aadaeb34a8aa67395 189391Author: Roland Mainz <roland.mainz@nrubsig.org> 189392Date: Sat Jun 19 21:56:01 2004 +0000 189393 189394 Refix for http://freedesktop.org/bugzilla/show_bug.cgi?id=764 : Rework 189395 previous solution and make Xprt to default to "-noreset" (the default 189396 of Solaris version of Xprt) and add a "-reset" option which can be used 189397 to restore the default behaviour on demand. 189398 189399commit da78a4ddd833f78baf1d2579a1adea8208016ddb 189400Author: Damien Ciabrini <braun@club-internet.fr> 189401Date: Wed Jun 16 21:36:54 2004 +0000 189402 189403 Update MGA composite patch commit. (some files were missing in the previous 189404 commit) 189405 189406commit e56e24af252bd3b8e58076adf0f8eabf1103f187 189407Author: Eric Anholt <anholt@freebsd.org> 189408Date: Wed Jun 16 09:37:59 2004 +0000 189409 189410 Merge DRI-trunk-20040613 changes in programs/Xserver/GL. 189411 189412commit 2e1868b560315a8b20d688e646c489a5ad93eeae 189413Author: Eric Anholt <anholt@freebsd.org> 189414Date: Wed Jun 16 09:25:21 2004 +0000 189415 189416 DRI trunk-20040613 import 189417 189418commit f45c46c630855e8e0d1c28b1f0d3b2ad54334619 189419Author: Eric Anholt <anholt@freebsd.org> 189420Date: Wed Jun 16 09:25:15 2004 +0000 189421 189422 Initial revision 189423 189424commit 22bad9474b8822f03f84a8a39edce624bfb9befa 189425Author: Eric Anholt <anholt@freebsd.org> 189426Date: Wed Jun 16 09:22:17 2004 +0000 189427 189428 DRI XFree86-4_3_99_12-merge import 189429 189430commit 1c133c27ccc1f09b95922fdece3c8d73cc182def 189431Author: Eric Anholt <anholt@freebsd.org> 189432Date: Wed Jun 16 09:22:05 2004 +0000 189433 189434 Initial revision 189435 189436commit b61ff0daa4bd1e3b828dc5b985c3a2f3c92b202e 189437Author: Eric Anholt <anholt@freebsd.org> 189438Date: Wed Jun 16 09:16:01 2004 +0000 189439 189440 DRM 20040613 import 189441 189442commit bcc1eab1fd57e8cb686d625934a6e527b7ae4ea2 189443Author: Eric Anholt <anholt@freebsd.org> 189444Date: Wed Jun 16 09:16:01 2004 +0000 189445 189446 Initial revision 189447 189448commit 580b9a7da1bf0e20acdcddd676d471b3d6589023 189449Author: Jaymz Julian <jaymz@artificial-stupidity.net> 189450Date: Mon Jun 14 08:43:57 2004 +0000 189451 189452 MGA composite support from Damien Ciabrini - thanks! 189453 189454commit 95d65cf6bb753d10f4db3d857fb98bb09389228e 189455Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189456Date: Sun Jun 13 04:50:21 2004 +0000 189457 189458 Manual page X(7) does not reference Xprt(1x), xplsprinters(1x), etc. 189459 xc/config/cf/Imake.rules Correct comment to match rule name for 189460 InstallDriverSDKObjectModule 189461 xc/programs/Xserver/hw/xfree86/os-support/sunos/sun_kbd.c Log results of 189462 ioctls to probe keyboard type & layout 189463 189464commit 4ffde8a6b3299f002c10b1abd881e4c6849767ea 189465Author: Eric Anholt <anholt@freebsd.org> 189466Date: Thu Jun 10 19:22:58 2004 +0000 189467 189468 - Pass the right pixel mask (all ones) in to PrepareSolid in the 189469 solid-fill-based composite acceleration. 189470 - Use a real pixmap when doing an UploadToScratch (For pDrawable->type == 189471 DRAWABLE_WINDOW, you need to get the backing pixmap). 189472 - Pass back the x/y offsets from kaaGetOffscreenPixmap unconditionally, 189473 because they'll be used in the scratch case. 189474 - Turn on the Render acceleration for Rage 128 and Radeon 100-series at 189475 last! 189476 189477commit c3bc6dd551436d5e37a07f37b3b77a83bb5b5da0 189478Author: Eric Anholt <anholt@freebsd.org> 189479Date: Thu Jun 10 09:50:59 2004 +0000 189480 189481 Align scratch area offsets to the offscreen byte alignment. 189482 189483commit cf3f95d2164604047866b283fe0071574bf16dbc 189484Author: Eric Anholt <anholt@freebsd.org> 189485Date: Thu Jun 10 08:37:28 2004 +0000 189486 189487 Oops, testers reported that the last patch actually didn't work (conflicts 189488 occurred), so the R300 PDMA doesn't work. Disable. 189489 189490commit 0b7647ee359537953b67b0dbf9daa807e356062b 189491Author: Eric Anholt <anholt@freebsd.org> 189492Date: Thu Jun 10 05:57:31 2004 +0000 189493 189494 Bug #242: Fix setup of R300 cards, by providing R300 CP code from 189495 volodya-project and initializing PDMA. 189496 189497commit b3a18ca8b827cfe2ebb295a03a9776028242c1a0 189498Author: Franco Catrin L <fcatrin@tuxpan.com> 189499Date: Mon Jun 7 05:13:29 2004 +0000 189500 189501 Neomagic driver enabled 189502 189503commit 893ea125597f3c6273f45a51673d4dc514e754e9 189504Author: Franco Catrin L <fcatrin@tuxpan.com> 189505Date: Mon Jun 7 05:05:10 2004 +0000 189506 189507 small fixes. README added 189508 189509commit d9cca52feba13b69f3eea9e1d958b8a4711e7d67 189510Author: Keith Packard <keithp@keithp.com> 189511Date: Fri Jun 4 17:06:18 2004 +0000 189512 189513 Add (stubbed out) Xgl server code 189514 189515commit f8a1dd3ce725195baa6f38a880299752c6c6c2c4 189516Author: Keith Packard <keithp@keithp.com> 189517Date: Fri Jun 4 16:10:50 2004 +0000 189518 189519 Add top-level build support for GL X server (not working yet) 189520 Fix a few allocation bugs with alternate visual ids Allow for non-8/8/8 189521 alternate visuals 189522 Turn off any existing shadow before enabling it again (avoids 189523 re-registering existing damage) 189524 Add some validation code to catch re-registered damages 189525 189526commit 6741fadc52598af0096f106a2cefd640abb434b3 189527Author: Phil Blundell <pb@reciva.com> 189528Date: Wed Jun 2 20:49:50 2004 +0000 189529 189530 New conditional. (REQUIRED_MODULES): Demand xcalibrateext if building 189531 XCalibrate. 189532 New file. 189533 Add xcalibrate.c. 189534 Read raw events if requested. 189535 189536commit d4d0c8470c4272dec642ab4c68f44a83cda06971 189537Author: Phil Blundell <pb@reciva.com> 189538Date: Sun May 30 20:40:30 2004 +0000 189539 189540 Add -lts if using tslib. 189541 189542commit 8124810950d7e0b9db7f66dadee7218b0c26c4c3 189543Author: Carlos Romero <kcrashcore@bellsouth.net> 189544Date: Sun May 30 13:51:18 2004 +0000 189545 189546 Initialize permedia engine for acceleration to work. 189547 189548commit ea1bbf8d83d3780ccce5ebcdff48f0b19863cee1 189549Author: Ralph Thomas <ralpht@68k.org> 189550Date: Sat May 29 12:15:46 2004 +0000 189551 189552 Adding driver for VIA CLE266 graphics chip. Currently it only accelerates 189553 copy and fill operations. 189554 189555commit 6af411b02e808220d3afcef14abb97eec86cf1f3 189556Author: Daniel Stone <daniel@fooishbar.org> 189557Date: Fri May 28 04:56:49 2004 +0000 189558 189559 Hey, I like devfs. 189560 Try /dev/fb/0 if /dev/fb0 fails. 189561 189562commit a7b42f685e7a4bf57cf89a3ef664a581ecedb50f 189563Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189564Date: Thu May 27 14:11:42 2004 +0000 189565 189566 file ChangeLog was initially added on branch CYGWIN. 189567 189568commit 05a3dbf5dc55ea534c68fc9d05b3949805a0752e 189569Author: Egbert Eich <eich@suse.de> 189570Date: Wed May 26 17:44:29 2004 +0000 189571 189572 Updated x86emu and resynced with upsteam at Scitech. 189573 189574commit 9549f628e066396e6bc9a7edfc919bdd6860f170 189575Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189576Date: Tue May 25 20:33:46 2004 +0000 189577 189578 getconfig: file '/usr/X11R6/lib/X11/getconfig/xorg.cfg' has bad signature 189579 (Change "Xorg Project" to "Xorg Foundation" to match getconfig script) 189580 189581commit f8124d3ef5890d59c3ce41bee46b5e3576d0f9b1 189582Author: Carlos Romero <kcrashcore@bellsouth.net> 189583Date: Tue May 25 13:02:44 2004 +0000 189584 189585 Add pm2 to the build 189586 189587commit 32d0920ef9ec3c5e61089b88dedc82ffab294276 189588Author: Carlos Romero <kcrashcore@bellsouth.net> 189589Date: Mon May 24 19:31:41 2004 +0000 189590 189591 Initial import of Permedia2 driver 189592 189593commit 5b2211ec3545f1634f807daf84b6c4bc2c0fdecf 189594Author: Egbert Eich <eich@suse.de> 189595Date: Mon May 24 19:05:01 2004 +0000 189596 189597 Muffle compiler warnings. 189598 fix option name in log message. 189599 improve debugging messages. 189600 189601commit 932efe8e6d4e6280aed9b5e25af56888c964d37b 189602Author: Keith Packard <keithp@keithp.com> 189603Date: Fri May 21 03:32:27 2004 +0000 189604 189605 Allow for multiple composite-based visuals, then add an RGB24 visual in 189606 addition to the ARGB32 one. This allows 'glitz' to run on top of any X 189607 server using mesa. 189608 Switch to using 32bpp for depth 24 pixmaps (even when the frame buffer is 189609 not depth 24). 189610 189611commit cde51fd05b2bd413d8db8ad750e6a72398a7039c 189612Author: Keith Packard <keithp@keithp.com> 189613Date: Thu May 20 19:51:44 2004 +0000 189614 189615 Miscomputing pitch in 24bpp modes because of rounding errors. 189616 189617commit b9d920f3dc060d230a4a7b2d40210524acf50666 189618Author: Keith Packard <keithp@keithp.com> 189619Date: Thu May 20 05:27:03 2004 +0000 189620 189621 Fix SYNC_ALWAYS (debugging) code to use mach64WaitIdle instead of 189622 KdCheckSync -- the boolean used in the latter won't be set yet. 189623 Oops. == instead of =. 189624 Must sync hardware before rasterizing trapezoids in case the mask is in 189625 off-screen memory and has just been erased. Yes, it is silly to place 189626 masks in off-screen memory. That's a separate issue. 189627 189628commit 94648799c82e59166155ca5abf22a9391693e6a1 189629Author: Keith Packard <keithp@keithp.com> 189630Date: Thu May 20 02:42:20 2004 +0000 189631 189632 Pin header-only pixmaps in memory. 189633 Off-screen reallocation could have used a stale pointer. 189634 Separate framebuffer mapping computation from actual frame buffer mapping. 189635 Now map the frame buffer from vesaEnable so that VT switch shares the 189636 same mapping code. This makes sure any shadow framebuffer is allocated 189637 again. 189638 189639commit cade317d31dddab61199d5e90bcff36fb12f3cd1 189640Author: Eric Anholt <anholt@freebsd.org> 189641Date: Mon May 17 20:18:02 2004 +0000 189642 189643 Overhaul of the ATI driver: 189644 - Add monochrome hardware cursor support. 189645 - Try to auto-detect AGP support for DRI on Radeons. And fail. Detect it 189646 properly on R128. 189647 - Set up card for pseudo-DMA if possible. Convert 2D rendering code to 189648 prepare DMA packets only. Use generic code to decode DMA packets to 189649 MMIO if PDMA is unavailable. Add WIP code to support "real" DMA without 189650 DRM support. 189651 - Dispatch pending DMA commands when the server sleeps. Otherwise some 189652 things, such as typing in an xterm, wouldn't show up for a time. 189653 - Fix Radeon Composite acceleration in many ways, and add Rage 128 189654 Composite acceleration. Disable them both due to still-not-understood 189655 issues they have. They fail with In, Out, AtopReverse, and Xor, and 189656 text rendering is strange. 189657 - Add textured XV support for R100 and Rage 128. No brightness/sat 189658 controls, but it does support multiple ports, and cooperates with 189659 Composite. 189660 - Add WIP code for hostdata uploads. 189661 - Many cleanups and fixes. 189662 189663commit 834537e212e01314b60737278b7abc6bb7cef102 189664Author: Eric Anholt <anholt@freebsd.org> 189665Date: Mon May 17 07:19:49 2004 +0000 189666 189667 Make kaaMoveInPixmap public. This will be used by the ATI driver's xvideo 189668 support to ensure that the destination is in framebuffer. 189669 189670commit 85f46e0bcdf60d145a6868ee71d10688c9113e6e 189671Author: Eric Anholt <anholt@freebsd.org> 189672Date: Mon May 17 07:14:23 2004 +0000 189673 189674 Add new CheckComposite hook. This allows a driver to avoid the migration of 189675 pixmaps for a Composite operation if the operation can't be supported. 189676 This hook is optional. 189677 189678commit 47fb207c8ae2b54e976066f78892a1ee3fb35d30 189679Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189680Date: Sun May 16 05:08:39 2004 +0000 189681 189682 xc/programs/Xserver/fb/fb.h 189683 xc/programs/Xserver/fb/fboverlay.c 189684 xc/programs/Xserver/fb/fbscreen.c 189685 - Change #ifdef for checking for old format miScreenInit to 189686 FB_OLD_MISCREENINIT for easier portability to xservers with updated 189687 screen structs but old function prototypes. Make it automatically 189688 defined if FB_OLD_SCREEN is defined. 189689 - Add _LP64 to list of #ifdefs for 64-bit platforms to support 189690 64-bit Solaris. 189691 189692commit fc2dd516c3c7382915452207180a1c483d0d73ca 189693Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189694Date: Sun May 16 00:03:54 2004 +0000 189695 189696 xc/programs/Xserver/hw/xfree86/xf86config/Imakefile 189697 xc/programs/Xserver/hw/xfree86/xf86config/xorgconfig.c 189698 - Clean up server name changes from TM branch 189699 - Set default XKB rules file name correctly 189700 - Use default font path from Imake configuration for the default font path 189701 in generated xorg.conf files. 189702 - Use path variables from Imake configuration for paths to files, in case 189703 vendor has configured them to install somewhere other than the 189704 defaults. 189705 189706commit b1aa9499ffb827f4b1acc75f197e332bba382565 189707Author: Roland Mainz <roland.mainz@nrubsig.org> 189708Date: Sat May 15 14:43:05 2004 +0000 189709 189710 Fix for http://pdx.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=622 - 189711 RFE: Xprt default BIGREQUESTS extension buffer size should be 8MB (to 189712 make |XpPutDocumentData()| happy and to improve performance). 189713 189714commit 74b2a7694791297a4f798ecc05c7eb8f68634722 189715Author: Eric Anholt <anholt@freebsd.org> 189716Date: Fri May 14 00:34:28 2004 +0000 189717 189718 Add new flag, KAA_OFFSCREEN_ALIGN_POT, which tells KAA to align pixmap 189719 pitches to a power-of-two number of bytes. Useful for Render 189720 acceleration on older cards. 189721 189722commit 2bea33e881693e7d7dcf938db79c888a71dfb2fb 189723Author: Eric Anholt <anholt@freebsd.org> 189724Date: Fri May 14 00:27:29 2004 +0000 189725 189726 Don't let the visible screen get "migrated" offscreen, which manifests 189727 itself as a hang. 189728 Reported by: Ginokas <ginokas@free.fr> 189729 189730commit 40354e761892dc2ef88d2e722d8d7896642003eb 189731Author: Eric Anholt <anholt@freebsd.org> 189732Date: Thu May 13 22:57:15 2004 +0000 189733 189734 Add generic functions for copying packed/planar XV data, copied from 189735 mach64. 189736 189737commit 2e330e980f61b256c55f5b9debb00574e4e85b26 189738Author: Eric Anholt <anholt@freebsd.org> 189739Date: Thu May 13 21:41:48 2004 +0000 189740 189741 Move fourcc.h to a generic location in src/. 189742 189743commit f52a4d472d2463482d865c5006208182c294e670 189744Author: Keith Packard <keithp@keithp.com> 189745Date: Thu May 13 21:25:51 2004 +0000 189746 189747 Follow GLX in setting ARGB visual nplanes to 24. 189748 Retry current mouse protocol when sync is lost 189749 189750commit aa5a87847290d49b03a33351ebfd8df652a42489 189751Author: Eric Anholt <anholt@freebsd.org> 189752Date: Thu May 13 21:15:06 2004 +0000 189753 189754 There's no need to explicitly set softCursor -- kdrive handles this if the 189755 cursor hooks aren't set. 189756 189757commit 4078457919708a8dbf9db8ee6e4871ecbf72518f 189758Author: Eric Anholt <anholt@freebsd.org> 189759Date: Wed May 12 01:49:46 2004 +0000 189760 189761 Fix problems in render fb implementation found by rendercheck: 189762 - fbCombineSaturate was pointed at fbCombineDisjointOver, instead of 189763 fbCombineDisjointOverReverse as it should. Instead, point 189764 fbCombineDisjointOverReverse at fbCombineSaturate (which is likely to 189765 be faster). 189766 - fix previously-unused fbCombineSaturate implementation. 189767 - fbCombineMaskAlphaC was just a copy of fbCombineMaskValueC. Make it do 189768 what it's supposed to (return a cs.alpha). 189769 - fbCombineAtopC didn't invert the source alpha value. 189770 - fix copy'n'paste errors in fbCombine(Dis/Con)jointGeneralC, also source 189771 alpha wasn't treated in a component fashion. 189772 - fbCompositeSrc_8888* didn't handle when the source lacks an alpha 189773 channel. Rather than adding that and possilby slowing down the (normal) 189774 alpha case, don't let x8r8g8b8/x8b8g8r8 Pictures be used in 189775 fbCompositeSrc_8888* because Over with one of these is just Src. 189776 189777commit a43d5412b4d79d67af20dc8af144a9ca80263e9d 189778Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189779Date: Sun May 9 16:20:13 2004 +0000 189780 189781 file ChangeLog was initially added on branch CYGWIN. 189782 189783commit 0498d818fe40cb4eb03983e27a980791bbadf6db 189784Author: Roland Mainz <roland.mainz@nrubsig.org> 189785Date: Sat May 8 02:06:46 2004 +0000 189786 189787 Fix for http://pdx.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=608 189788 ("Can not print on Debian/CUPS due to error |Xprt_64:lpr: unable to 189789 print file: server-error-not-accepting-jobs|"): Xprt did not setup the 189790 list of supplementary group ids, causing print failure when the the 189791 calling user must be a member in a specific group to be allowed to 189792 print. 189793 189794commit 75d96afcc4a1f201da665bd73b7067e8e7139a3f 189795Author: Egbert Eich <eich@suse.de> 189796Date: Thu May 6 17:31:17 2004 +0000 189797 189798 BugZilla #601: Fixing makedepend choking on floating point exception 189799 because CHAR_BIT is defined to __CHAR_BIT__ which is a compiler 189800 intrinsic define. BugZilla #605: Fixing build on IA64 which is broken 189801 due to the inclusion of the kernel header asm/page.h. Kernel headers 189802 however don't work with 189803 -ansi. The inclusion of asm/page.h can however savely be removed as it 189804 there are plenty of other ways to determine the page size. 189805 189806commit 7124cfaa006e840ba48dcc466c0dc8b34503a686 189807Author: Keith Packard <keithp@keithp.com> 189808Date: Thu May 6 16:19:32 2004 +0000 189809 189810 Use current resolution by default, change rate to 75 to match 189811 fbdevModeSupported cut-off (?). Glenn McGrath <glennm@hydrix.com> 189812 189813commit e4ac2411eddf1f01ef9204f27b6d1ce8f1749439 189814Author: Roland Mainz <roland.mainz@nrubsig.org> 189815Date: Thu May 6 01:53:52 2004 +0000 189816 189817 Fix for http://pdx.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=551 - 189818 PS DDX will not build on platforms with BuildFreeType NO. Patch by Alan 189819 Coopersmith <alan.coopersmith@sun.com>. 189820 189821commit b1c65e1ca6828ea82ee7790f22c26503b0a5e17d 189822Author: Roland Mainz <roland.mainz@nrubsig.org> 189823Date: Thu May 6 00:24:32 2004 +0000 189824 189825 Fix for http://pdx.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=536 - 189826 RFE: PS output should contain the FreeType2 version being used. 189827 189828commit 8d4f21ab53c44ca48501d6211ea6db0c0b8af916 189829Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 189830Date: Wed May 5 20:15:41 2004 +0000 189831 189832 Add XACE and XSELINUX extensions to the build system 189833 189834commit 0106715000196c7b349a0b4494b61545f0f5e138 189835Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 189836Date: Wed May 5 20:07:37 2004 +0000 189837 189838 Modify XC-SECURITY and XC-APPGROUP extensions to work with XACE 189839 189840commit 8526cd6395490b03b279f1962df777fb0e4a9878 189841Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 189842Date: Wed May 5 20:04:52 2004 +0000 189843 189844 Replace XC-SECURITY code with XACE security hooks 189845 189846commit 6d066cb10990d951449b342b40dec1f1b1ae593c 189847Author: Eamon Walsh <ewalsh@epoch.ncsc.mil> 189848Date: Tue May 4 19:44:02 2004 +0000 189849 189850 Merge the new release from HEAD 189851 189852commit b5f200ce9d495c6ce94e0170909465a30e8799d9 189853Author: Keith Packard <keithp@keithp.com> 189854Date: Tue May 4 03:28:06 2004 +0000 189855 189856 Attached is a patch to fix a build error whe ncompiling with tslib support, 189857 a variable wasnt set, i just changed it to be like the other 189858 MAkefile.ams :) -- Glenn McGrath <glennm@hydrix.com> 189859 189860commit 5ca651e66f3d0ab189962bb4609b87a865364ef8 189861Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 189862Date: Fri Apr 30 12:48:56 2004 +0000 189863 189864 file winmessages.h was initially added on branch CYGWIN. 189865 189866commit 2c2c1704b542f29fe5ac9917e1141040a0dbd3e9 189867Author: Roland Mainz <roland.mainz@nrubsig.org> 189868Date: Thu Apr 29 23:59:15 2004 +0000 189869 189870 Fix for http://pdx.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=567 - 189871 Xorg Xprt starts to consume 100% CPU when being idle for some time 189872 (internal screensaver goes mad after 10mins) 189873 189874commit a8429d76103ff0f4fc61db86201c741f91bfcba2 189875Author: Keith Packard <keithp@keithp.com> 189876Date: Wed Apr 28 07:26:46 2004 +0000 189877 189878 Add completely fake X server -- draws to allocated buffer, has no keyboard 189879 or mouse. 189880 189881commit 85e4e5445218d70f627fb132a8e8f18470e6749d 189882Author: Roland Mainz <roland.mainz@nrubsig.org> 189883Date: Mon Apr 26 11:07:03 2004 +0000 189884 189885 Work-in-progress for 189886 http://pdx.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=542 - GLX 189887 support for PS DDX / part #1: Fix visual setup (attachment #243) 189888 189889commit 46472cbee679f9757c4003a0dcf158aeb3852f47 189890Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189891Date: Mon Apr 26 02:39:58 2004 +0000 189892 189893 xc/config/cf/sun.cf 189894 xc/config/cf/sv4Lib.rules 189895 xc/programs/Xserver/Imakefile 189896 xc/programs/Xserver/hw/xfree86/os-support/sunos/find_deps.pl Make Solaris 189897 builds work when using MakeDllModules (it's not the default yet, but at 189898 least it works now if you turn it on) Also improve default compiler, 189899 optimizer, & linker flags for Solaris builds using either Sun cc or gcc 189900 xc/programs/Xserver/cfb/Imakefile.inc 189901 xc/programs/Xserver/cfb/stipsparc.s 189902 xc/programs/Xserver/cfb/stipsprc32.s Remove text relocation error when 189903 building shared versions 189904 189905commit 36e3e5430e1ca7103a4e0b796eb3817975b40d90 189906Author: Roland Mainz <roland.mainz@nrubsig.org> 189907Date: Sun Apr 25 22:42:09 2004 +0000 189908 189909 Fix for http://pdx.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=541 - 189910 Xorg Xprt may crash with "Freeing resource id=40200000 which isnt 189911 there" 189912 189913commit 2fb588620030ad393f8500d60e16144d59e4effe 189914Author: Egbert Eich <eich@suse.de> 189915Date: Fri Apr 23 19:54:30 2004 +0000 189916 189917 Merging XORG-CURRENT into trunk 189918 189919commit 0664db19bf37f9dd69cca6adff4e238e310c3092 189920Author: Egbert Eich <eich@suse.de> 189921Date: Fri Apr 23 18:54:16 2004 +0000 189922 189923 Merging XORG-CURRENT into trunk 189924 189925commit 68fd529608c58334f13beb88dbcc1d5db85b9b00 189926Author: Roland Mainz <roland.mainz@nrubsig.org> 189927Date: Wed Apr 21 23:24:20 2004 +0000 189928 189929 Fix for http://pdx.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=535 - 189930 Xprt should not annouce extensions which are not supported 189931 189932commit 1af13123fa79ad1c6747aad60ed458bbd69da12d 189933Author: Roland Mainz <roland.mainz@nrubsig.org> 189934Date: Wed Apr 21 10:03:41 2004 +0000 189935 189936 Fix for http://pdx.freedesktop.org/cgi-bin/bugzilla/show_bug.cgi?id=530 - 189937 Land XPRINT branch on XORG-CURRENT 189938 189939commit 449e83a9470ec4bdd0871e2f263f608b24455423 189940Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189941Date: Sun Apr 18 03:00:43 2004 +0000 189942 189943 xc/lib/GL/glx/Imakefile 189944 xc/lib/GL/mesa/src/Imakefile LargePICTable required for Solaris SPARC 189945 builds 189946 xc/programs/Xserver/hw/xfree86/drivers/glint/pm3_accel.c 189947 xc/programs/Xserver/hw/xfree86/common/xf86Events.c Add != NULL to if 189948 statements to get past syntax error reported by Sun Forte 6.1 cc. 189949 xc/config/imake/imake.c 189950 xc/config/cf/sun.cf 189951 xc/config/cf/sunLib.tmpl Allow compiling with Sun compilers installed 189952 somewhere other than /opt/SUNWspro 189953 xc/programs/Xserver/hw/xfree86/common/compiler.h 189954 xc/programs/Xserver/hw/xfree86/os-support/bus/Pci.h Check for 189955 defined(sparc) as well as defined(__sparc__) since Sun compilers don't 189956 define __sparc__ 189957 189958commit 425251a752805affb6ce14baa58d92c384f39501 189959Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189960Date: Sat Apr 17 18:47:05 2004 +0000 189961 189962 Bugzilla #495: LocalClientCred should use getpeerucred on Solaris 10 189963 189964commit 7215fb186f076a24d0a04c9c20ac9b92cae1f49b 189965Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 189966Date: Fri Apr 16 00:21:24 2004 +0000 189967 189968 xc/programs/Xserver/hw/xfree86/os-support/sunos/sun_mouse.c Solaris mouse 189969 enhancements, including autoprobe support, VUID wheel mouse events, and 189970 streams module pushing. Bugzilla #434. (Russ Blaine & Alan Coopersmith, 189971 Sun Microsystems) 189972 xc/programs/Xserver/hw/xfree86/os-support/sunos/sun_init.c Xorg doesn't 189973 reset console to text mode on Solaris x86 8 and later Bugzilla #469. 189974 189975commit c6c6d0de2309019999fa75a2f36a4f4a93ad2f31 189976Author: Egbert Eich <eich@suse.de> 189977Date: Thu Apr 15 10:17:35 2004 +0000 189978 189979 Merged changes from RELEASE-1 branch 189980 189981commit 9d24a5fa91bf165bbd2048a844edeb59b5e34aad 189982Author: Harold L Hunt II <huntharo@msu.edu> 189983Date: Wed Apr 14 00:01:22 2004 +0000 189984 189985 file XWinrc.man was initially added on branch CYGWIN. 189986 189987commit 01bb5eb5032a7566c86a51053146dba98a3ed749 189988Author: Roland Mainz <roland.mainz@nrubsig.org> 189989Date: Tue Apr 13 03:16:46 2004 +0000 189990 189991 file psout_ftpstype3.c was initially added on branch XPRINT. 189992 189993commit cb3f3d8f2283d384dc5a3af3f38053cc8a2d192e 189994Author: Roland Mainz <roland.mainz@nrubsig.org> 189995Date: Tue Apr 13 03:16:46 2004 +0000 189996 189997 file psout_ft.c was initially added on branch XPRINT. 189998 189999commit 7cfb4c2b33ae2147b5d6ddc2afc8b777686a666f 190000Author: Roland Mainz <roland.mainz@nrubsig.org> 190001Date: Tue Apr 13 03:16:46 2004 +0000 190002 190003 file psout_ftpstype1.c was initially added on branch XPRINT. 190004 190005commit 4ae42e79d46d7db30f7b6f321bbb0d134862138d 190006Author: Roland Mainz <roland.mainz@nrubsig.org> 190007Date: Tue Apr 13 03:16:46 2004 +0000 190008 190009 file PsFTFonts.c was initially added on branch XPRINT. 190010 190011commit b5fb71922b02024aa5a8f349c9d2c956e2f83f0f 190012Author: Roland Mainz <roland.mainz@nrubsig.org> 190013Date: Tue Apr 13 03:16:45 2004 +0000 190014 190015 file xprint.sh was initially added on branch XPRINT. 190016 190017commit ca9a9a58be51a21f123b11dd68047034696cae84 190018Author: Roland Mainz <roland.mainz@nrubsig.org> 190019Date: Tue Apr 13 03:16:45 2004 +0000 190020 190021 file xprint.csh was initially added on branch XPRINT. 190022 190023commit 7c1f840108172d6b18af47465ea72f4820640598 190024Author: Roland Mainz <roland.mainz@nrubsig.org> 190025Date: Tue Apr 13 03:16:45 2004 +0000 190026 190027 file cde_xsessiond_xprint.sh was initially added on branch XPRINT. 190028 190029commit 579221198aeac7010435b29db1ad8fe9ee2d7c5d 190030Author: Roland Mainz <roland.mainz@nrubsig.org> 190031Date: Tue Apr 13 03:16:44 2004 +0000 190032 190033 file spooler.c was initially added on branch XPRINT. 190034 190035commit 7677b4992fee7eb73cc97914163dcf689ad13d6a 190036Author: Roland Mainz <roland.mainz@nrubsig.org> 190037Date: Tue Apr 13 03:16:44 2004 +0000 190038 190039 file spooler.h was initially added on branch XPRINT. 190040 190041commit d3907ca519b476c99e29a58c22258f22dbe63244 190042Author: Roland Mainz <roland.mainz@nrubsig.org> 190043Date: Tue Apr 13 03:16:44 2004 +0000 190044 190045 file document was initially added on branch XPRINT. 190046 190047commit 3646bb9c894d5f011e2df4fac402118d8350102e 190048Author: Roland Mainz <roland.mainz@nrubsig.org> 190049Date: Tue Apr 13 03:16:39 2004 +0000 190050 190051 file spooltodir.sh was initially added on branch XPRINT. 190052 190053commit 8c006df3c6d8e5ac95f0b0fa38a030100541598f 190054Author: Roland Mainz <roland.mainz@nrubsig.org> 190055Date: Tue Apr 13 03:16:39 2004 +0000 190056 190057 file model-config was initially added on branch XPRINT. 190058 190059commit 95a84bc0cb809e5c7141d0411e329a3ec300b8aa 190060Author: Roland Mainz <roland.mainz@nrubsig.org> 190061Date: Tue Apr 13 03:16:38 2004 +0000 190062 190063 file ZapfDingbats.pmf was initially added on branch XPRINT. 190064 190065commit 5527b39e668ea7a88c41186dbb6d7b66e892547a 190066Author: Roland Mainz <roland.mainz@nrubsig.org> 190067Date: Tue Apr 13 03:16:38 2004 +0000 190068 190069 file Times-Roman.pmf was initially added on branch XPRINT. 190070 190071commit 69c405ac66b1a15076e247dd1b578b7b4b210b00 190072Author: Roland Mainz <roland.mainz@nrubsig.org> 190073Date: Tue Apr 13 03:16:38 2004 +0000 190074 190075 file Times-Italic.pmf was initially added on branch XPRINT. 190076 190077commit b6b75f677292ed0c694921df0abf40038dd5e99d 190078Author: Roland Mainz <roland.mainz@nrubsig.org> 190079Date: Tue Apr 13 03:16:38 2004 +0000 190080 190081 file Times-BoldItalic.pmf was initially added on branch XPRINT. 190082 190083commit 6605566bcf008551d33a9a68bd03e9a0c57c4f60 190084Author: Roland Mainz <roland.mainz@nrubsig.org> 190085Date: Tue Apr 13 03:16:38 2004 +0000 190086 190087 file Souvenir-LightItalic.pmf was initially added on branch XPRINT. 190088 190089commit 38ad2972bffea1019214785ee479f7670ca70d1f 190090Author: Roland Mainz <roland.mainz@nrubsig.org> 190091Date: Tue Apr 13 03:16:38 2004 +0000 190092 190093 file Souvenir-Light.pmf was initially added on branch XPRINT. 190094 190095commit 10d8eace174a9778a3313ac36a3422637b020d5f 190096Author: Roland Mainz <roland.mainz@nrubsig.org> 190097Date: Tue Apr 13 03:16:38 2004 +0000 190098 190099 file Souvenir-DemiItalic.pmf was initially added on branch XPRINT. 190100 190101commit fb5ac8e2bdfe1217663679f5eae8045473456752 190102Author: Roland Mainz <roland.mainz@nrubsig.org> 190103Date: Tue Apr 13 03:16:38 2004 +0000 190104 190105 file NewCenturySchlbk-Roman.pmf was initially added on branch XPRINT. 190106 190107commit 5ec311b05dc1e509c115ccca808ee05090cfddad 190108Author: Roland Mainz <roland.mainz@nrubsig.org> 190109Date: Tue Apr 13 03:16:38 2004 +0000 190110 190111 file NewCenturySchlbk-Italic.pmf was initially added on branch XPRINT. 190112 190113commit 70947a8f1addf4ca17e50d9e6ae590266ac446cc 190114Author: Roland Mainz <roland.mainz@nrubsig.org> 190115Date: Tue Apr 13 03:16:38 2004 +0000 190116 190117 file Symbol.pmf was initially added on branch XPRINT. 190118 190119commit 436ff77b21515cd9fe9732e0bd5361f2bfba44ed 190120Author: Roland Mainz <roland.mainz@nrubsig.org> 190121Date: Tue Apr 13 03:16:38 2004 +0000 190122 190123 file Times-Bold.pmf was initially added on branch XPRINT. 190124 190125commit 4db563027844245d6c9085f997e75da743410885 190126Author: Roland Mainz <roland.mainz@nrubsig.org> 190127Date: Tue Apr 13 03:16:38 2004 +0000 190128 190129 file NewCenturySchlbk-Bold.pmf was initially added on branch XPRINT. 190130 190131commit 79110faa2eac849756b859071ce68fba64de57aa 190132Author: Roland Mainz <roland.mainz@nrubsig.org> 190133Date: Tue Apr 13 03:16:38 2004 +0000 190134 190135 file NewCenturySchlbk-BoldItalic.pmf was initially added on branch XPRINT. 190136 190137commit 4e1ae7e9cc04806f4436759764cc680ecf1f014c 190138Author: Roland Mainz <roland.mainz@nrubsig.org> 190139Date: Tue Apr 13 03:16:38 2004 +0000 190140 190141 file Souvenir-Demi.pmf was initially added on branch XPRINT. 190142 190143commit 433913bacf988908b94c420452c042eebcb381ac 190144Author: Roland Mainz <roland.mainz@nrubsig.org> 190145Date: Tue Apr 13 03:16:38 2004 +0000 190146 190147 file LubalinGraph-Demi.pmf was initially added on branch XPRINT. 190148 190149commit d5bae63138ab833fdd56bb983436ac514536d7b6 190150Author: Roland Mainz <roland.mainz@nrubsig.org> 190151Date: Tue Apr 13 03:16:38 2004 +0000 190152 190153 file LubalinGraph-DemiOblique.pmf was initially added on branch XPRINT. 190154 190155commit 3acd6856617e784ee30333dc9b779189a3f44052 190156Author: Roland Mainz <roland.mainz@nrubsig.org> 190157Date: Tue Apr 13 03:16:38 2004 +0000 190158 190159 file LubalinGraph-Book.pmf was initially added on branch XPRINT. 190160 190161commit 5f73192458136fe4b6b82372c3b1653fbf831ebd 190162Author: Roland Mainz <roland.mainz@nrubsig.org> 190163Date: Tue Apr 13 03:16:38 2004 +0000 190164 190165 file Helvetica.pmf was initially added on branch XPRINT. 190166 190167commit e129abc3bf269e857aa65065cc18a31a56ba0373 190168Author: Roland Mainz <roland.mainz@nrubsig.org> 190169Date: Tue Apr 13 03:16:38 2004 +0000 190170 190171 file Helvetica-Oblique.pmf was initially added on branch XPRINT. 190172 190173commit 576a4cddf995082d10e2e29e1b58c1564eb11ee7 190174Author: Roland Mainz <roland.mainz@nrubsig.org> 190175Date: Tue Apr 13 03:16:38 2004 +0000 190176 190177 file Helvetica-BoldOblique.pmf was initially added on branch XPRINT. 190178 190179commit cf26c87833a79427b665abce67ca19f2b68bc8e5 190180Author: Roland Mainz <roland.mainz@nrubsig.org> 190181Date: Tue Apr 13 03:16:38 2004 +0000 190182 190183 file Helvetica-Bold.pmf was initially added on branch XPRINT. 190184 190185commit a091408c372a2aa89fb83b023248f45aa8cd4173 190186Author: Roland Mainz <roland.mainz@nrubsig.org> 190187Date: Tue Apr 13 03:16:38 2004 +0000 190188 190189 file Courier.pmf was initially added on branch XPRINT. 190190 190191commit 9e4221d08ff9a408fb25a32887390b14788b1558 190192Author: Roland Mainz <roland.mainz@nrubsig.org> 190193Date: Tue Apr 13 03:16:38 2004 +0000 190194 190195 file Courier-Oblique.pmf was initially added on branch XPRINT. 190196 190197commit 6b2674078079a5959a2b7758e6c628a14ec1a46c 190198Author: Roland Mainz <roland.mainz@nrubsig.org> 190199Date: Tue Apr 13 03:16:38 2004 +0000 190200 190201 file Courier-BoldOblique.pmf was initially added on branch XPRINT. 190202 190203commit 6aee2d37b95170a65ee08c0866c425f115ebc9f1 190204Author: Roland Mainz <roland.mainz@nrubsig.org> 190205Date: Tue Apr 13 03:16:38 2004 +0000 190206 190207 file LubalinGraph-BookOblique.pmf was initially added on branch XPRINT. 190208 190209commit 7a59fe1dd987e1ef0abd92e0ac80dd87a15137fe 190210Author: Roland Mainz <roland.mainz@nrubsig.org> 190211Date: Tue Apr 13 03:16:38 2004 +0000 190212 190213 file AvantGarde-Demi.pmf was initially added on branch XPRINT. 190214 190215commit 22e0316acc8992033fc82a38f663fce130e4031b 190216Author: Roland Mainz <roland.mainz@nrubsig.org> 190217Date: Tue Apr 13 03:16:38 2004 +0000 190218 190219 file AvantGarde-DemiOblique.pmf was initially added on branch XPRINT. 190220 190221commit f8aded3a7f8c97731e33b4362243da947fb4e774 190222Author: Roland Mainz <roland.mainz@nrubsig.org> 190223Date: Tue Apr 13 03:16:38 2004 +0000 190224 190225 file AvantGarde-Book.pmf was initially added on branch XPRINT. 190226 190227commit 2224187c05d4dc05f7e03e22307cf7816d69f789 190228Author: Roland Mainz <roland.mainz@nrubsig.org> 190229Date: Tue Apr 13 03:16:38 2004 +0000 190230 190231 file AvantGarde-BookOblique.pmf was initially added on branch XPRINT. 190232 190233commit 9eafaaf83294d1988b30bde4485a299cf8ae5035 190234Author: Roland Mainz <roland.mainz@nrubsig.org> 190235Date: Tue Apr 13 03:16:38 2004 +0000 190236 190237 file Courier-Bold.pmf was initially added on branch XPRINT. 190238 190239commit f9eea9864e333efde97143278916da44639cc18a 190240Author: Roland Mainz <roland.mainz@nrubsig.org> 190241Date: Tue Apr 13 03:16:38 2004 +0000 190242 190243 file ps2pdf_spooltodir.sh was initially added on branch XPRINT. 190244 190245commit 20248eedd69c42c27605d7bcfb265994f5846f17 190246Author: Franco Catrin L <fcatrin@tuxpan.com> 190247Date: Sun Apr 11 16:39:48 2004 +0000 190248 190249 fixed Changelog format 190250 190251commit bc7168ee763ffac9cbb992096a53b346cd640a13 190252Author: Franco Catrin L <fcatrin@tuxpan.com> 190253Date: Sun Apr 11 16:24:03 2004 +0000 190254 190255 Added ROP 190256 190257commit 784d37ee369b94c83c4cc6e280a39f32da8aa678 190258Author: Franco Catrin L <fcatrin@tuxpan.com> 190259Date: Sun Apr 11 15:51:04 2004 +0000 190260 190261 Fixed size calculation in solid rendering 190262 190263commit 056322336cbb6093d74aa9d22bbfd42e2248a16a 190264Author: Franco Catrin L <fcatrin@tuxpan.com> 190265Date: Sun Apr 11 15:20:17 2004 +0000 190266 190267 Basic bitblt implementation 190268 190269commit c231856a1343e38381e1b4e545ff1ac279141bf0 190270Author: Franco Catrin L <fcatrin@tuxpan.com> 190271Date: Sun Apr 11 00:53:10 2004 +0000 190272 190273 First acceleration function implemented (DrawSolid) 190274 190275commit 47436a8af82a00d0d392cef4d5906729d9a37649 190276Author: Franco Catrin L <fcatrin@tuxpan.com> 190277Date: Sun Apr 11 00:15:57 2004 +0000 190278 190279 Finnally got MMIO working all timecat ChangeLog cat ChangeLog cat ChangeLog 190280 :-D 190281 190282commit 1740b938e4c4f1cd3de700ea26143b01c0312325 190283Author: Franco Catrin L <fcatrin@tuxpan.com> 190284Date: Tue Apr 6 18:09:44 2004 +0000 190285 190286 Return back to VESA only version 190287 190288commit 07bc231872e7e056fa3049a0fcd963c61f826f80 190289Author: Franco Catrin L <fcatrin@tuxpan.com> 190290Date: Mon Apr 5 18:19:34 2004 +0000 190291 190292 small fixes, but still can't get this driver woking again 190293 190294commit b526276faa765df893197e04370a915ed73947dc 190295Author: Brent Cook <busterbcook@yahoo.com> 190296Date: Mon Apr 5 02:52:35 2004 +0000 190297 190298 removed hardcoded vesa references so we can operate with any backend. fbdev 190299 works for initializing the screen, but input fails shortly after. 190300 190301commit 920e6ff81baeec16465f81bacbcff711ce82e149 190302Author: Brent Cook <busterbcook@yahoo.com> 190303Date: Sun Apr 4 07:30:07 2004 +0000 190304 190305 Begin separating VESA calls into a more generic backend wrapper like the 190306 ati driver, cascading between VESA and FBDEV. We only have init 190307 functions done so far; need to add all of the others. Fixed some 190308 compiler warnings. Whitespace and formatting cleanups (using 4 spaces, 190309 no tabs) 190310 190311commit 530371ceaf7f593badf38bbc2d2e50f6a920d24f 190312Author: Brent Cook <busterbcook@yahoo.com> 190313Date: Sat Apr 3 22:26:37 2004 +0000 190314 190315 added touchscreen support, detect all known PCI chips in the Neomagic line. 190316 We'll not bother with ISA for now. 190317 190318commit 962b898868dcab959c390986dcb0b4dd750dc107 190319Author: Brent Cook <busterbcook@yahoo.com> 190320Date: Sat Apr 3 22:22:48 2004 +0000 190321 190322 initial import of original driver by Franco Catrin L. Wraps VESA for most 190323 parts, with some hardware acceleration enabled for the cursor on the 190324 NM2300. 190325 190326commit 8a2fce3b90b5efc8bab19675cb8e02690e24442e 190327Author: Harold L Hunt II <huntharo@msu.edu> 190328Date: Sat Apr 3 05:01:21 2004 +0000 190329 190330 file winkeyhook.c was initially added on branch CYGWIN. 190331 190332commit 12d5371ed2fbefab069dea46be972a7269b8c2db 190333Author: Stuart Kreitman <stuart.kreitman@sun.com> 190334Date: Tue Mar 30 18:31:54 2004 +0000 190335 190336 file xfixes.h was initially added on branch DAMAGE-XFIXES. 190337 190338commit a4b319dbf375461c975450659723e6326153e536 190339Author: Stuart Kreitman <stuart.kreitman@sun.com> 190340Date: Tue Mar 30 18:31:54 2004 +0000 190341 190342 file xfixes.c was initially added on branch DAMAGE-XFIXES. 190343 190344commit 5319d30d45d5f8ec04a496327f32cc6431c6a511 190345Author: Stuart Kreitman <stuart.kreitman@sun.com> 190346Date: Tue Mar 30 18:31:54 2004 +0000 190347 190348 file select.c was initially added on branch DAMAGE-XFIXES. 190349 190350commit 76f247bd0ef23d688028c63b5f8bd3e9ad6b1b45 190351Author: Stuart Kreitman <stuart.kreitman@sun.com> 190352Date: Tue Mar 30 18:31:54 2004 +0000 190353 190354 file saveset.c was initially added on branch DAMAGE-XFIXES. 190355 190356commit 83f0f1babb612774f609c71879a225c43f63ac1f 190357Author: Stuart Kreitman <stuart.kreitman@sun.com> 190358Date: Tue Mar 30 18:31:54 2004 +0000 190359 190360 file cursor.c was initially added on branch DAMAGE-XFIXES. 190361 190362commit 52bc7693dbe7e3db916f8d463d9a750e3c6ffa4d 190363Author: Stuart Kreitman <stuart.kreitman@sun.com> 190364Date: Tue Mar 30 18:31:54 2004 +0000 190365 190366 file region.c was initially added on branch DAMAGE-XFIXES. 190367 190368commit 5d9098cb17cd88cfdf49de92bec2a787d6681649 190369Author: Stuart Kreitman <stuart.kreitman@sun.com> 190370Date: Tue Mar 30 18:31:54 2004 +0000 190371 190372 file xfixesint.h was initially added on branch DAMAGE-XFIXES. 190373 190374commit ace9aa7c45ff2ea6b3476006574da5c27d05afd4 190375Author: Stuart Kreitman <stuart.kreitman@sun.com> 190376Date: Tue Mar 30 17:41:24 2004 +0000 190377 190378 file damage.c was initially added on branch DAMAGE-XFIXES. 190379 190380commit 72dc7569c6bbc216f613be21ea4f79d3ef1d5534 190381Author: Stuart Kreitman <stuart.kreitman@sun.com> 190382Date: Tue Mar 30 17:41:24 2004 +0000 190383 190384 file damage.h was initially added on branch DAMAGE-XFIXES. 190385 190386commit 95da7b7e061b6925d8cd85bc7b25708ff253fcb1 190387Author: Stuart Kreitman <stuart.kreitman@sun.com> 190388Date: Tue Mar 30 17:41:24 2004 +0000 190389 190390 file damagestr.h was initially added on branch DAMAGE-XFIXES. 190391 190392commit 629c3792225cec28572081ebc8dda3fd803fe616 190393Author: Egbert Eich <eich@suse.de> 190394Date: Tue Mar 30 14:23:15 2004 +0000 190395 190396 36. Conversion: __AMD64__ > __amd64__ (Egbert Eich). 190397 35. Fixed stretching option and centering in C&T driver (Egbert Eich). 190398 34. Added support for memory size tweaking in BIOS for i845 (Egbert Eich, 190399 thanks to Christian Ziez) 190400 33. Removed video playback dependency on Accel in NSC drivers (Egbert 190401 3Eich). 190402 32. Fix HW cursor state on Savage driver when entering VT as some BIOSes 190403 seem to enable it unconditionally (Egbert Eich). 190404 31. Fixed Emulate3Button message to distinguish between 'hard' (ie. 190405 configured) and 'soft' (ie. automatic emulation that is disabled as 190406 soon as the middle button is pressed) (Egbert Eich). 190407 30. Free XrmDB in XCloseDisplay() only when implicitely allocated by 190408 XGetDefaults(). If Client allocates it itself it should free it also. 190409 Trying to free it for the client may result in segfault if the client 190410 has already freed it (Egbert Eich). 190411 190412commit b5e400867feab935aa04e9aadb12deb0601b7f83 190413Author: Egbert Eich <eich@suse.de> 190414Date: Tue Mar 30 14:14:31 2004 +0000 190415 190416 - backing out XFIXES and DAMAGE related code that accidentally went in 190417 here. 190418 190419commit 0017ddaa6406524d0a86ff7020eed4c33758ddbd 190420Author: Stuart Kreitman <stuart.kreitman@sun.com> 190421Date: Tue Mar 30 02:16:15 2004 +0000 190422 190423 file damageextint.h was initially added on branch DAMAGE-XFIXES. 190424 190425commit 4b5112fd0593b34e6e096d88b0841a28636600f6 190426Author: Stuart Kreitman <stuart.kreitman@sun.com> 190427Date: Tue Mar 30 02:16:15 2004 +0000 190428 190429 file damageext.c was initially added on branch DAMAGE-XFIXES. 190430 190431commit 733bfa4b5dd5255713a98f390a1fb65be6e16c2d 190432Author: Stuart Kreitman <stuart.kreitman@sun.com> 190433Date: Tue Mar 30 02:16:15 2004 +0000 190434 190435 file damageext.h was initially added on branch DAMAGE-XFIXES. 190436 190437commit b1ed473ef2e1da99d7fe3df14fdef4d8b6365626 190438Author: Harold L Hunt II <huntharo@msu.edu> 190439Date: Tue Mar 30 00:31:28 2004 +0000 190440 190441 file winrandr.c was initially added on branch CYGWIN. 190442 190443commit fec5095bdfb376d0046e2f52188c68bef4e31fd7 190444Author: Keith Packard <keithp@keithp.com> 190445Date: Sun Mar 28 07:14:30 2004 +0000 190446 190447 file xorgcfg.man was initially added on branch XORG-RELEASE-1-TM. 190448 190449commit 136a9364be80f407f498e9e9695cadaab39227ad 190450Author: Keith Packard <keithp@keithp.com> 190451Date: Sun Mar 28 07:14:30 2004 +0000 190452 190453 file XOrgCfg.cpp was initially added on branch XORG-RELEASE-1-TM. 190454 190455commit 881afb356d6992bcfbbfcbdf31f1f71c64d648da 190456Author: Keith Packard <keithp@keithp.com> 190457Date: Sat Mar 27 17:01:11 2004 +0000 190458 190459 file Xorg.man was initially added on branch XORG-RELEASE-1-TM. 190460 190461commit d2119ac7c56ba94db1d8639937b13e72288a328c 190462Author: Keith Packard <keithp@keithp.com> 190463Date: Sat Mar 27 17:01:11 2004 +0000 190464 190465 file xorg.conf.man was initially added on branch XORG-RELEASE-1-TM. 190466 190467commit 5a9f3a36a35549f30fc67a8e3a3e9a201efb91ec 190468Author: Egbert Eich <eich@suse.de> 190469Date: Fri Mar 26 20:02:03 2004 +0000 190470 190471 file xorgHelper.c was initially added on branch XORG-RELEASE-1-TM. 190472 190473commit b2b0e70fe5c65c8f2d518f5f0ce89a86938701ed 190474Author: Egbert Eich <eich@suse.de> 190475Date: Fri Mar 26 19:54:39 2004 +0000 190476 190477 file xorgconf.cpp was initially added on branch XORG-RELEASE-1-TM. 190478 190479commit 83880dd464a415d3d0efa546b1f0b9887342e809 190480Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 190481Date: Fri Mar 26 17:11:49 2004 +0000 190482 190483 29. XkbWriteRulesProp fails if XkbRulesFile is NULL. Bug #376. (Alan 190484 Coopersmith) 190485 190486commit 861a33678243349b987ff30912985968ede8ac84 190487Author: Stuart Kreitman <stuart.kreitman@sun.com> 190488Date: Fri Mar 26 01:22:18 2004 +0000 190489 190490 oops 190491 190492commit 23d552bbef2984afee889f82ded154478548ee15 190493Author: Egbert Eich <eich@suse.de> 190494Date: Thu Mar 25 11:00:52 2004 +0000 190495 190496 file xorg.cfg was initially added on branch XORG-RELEASE-1-TM. 190497 190498commit d1e52f13ad4610ec4907432c21384d08d6aaaf27 190499Author: Stuart Kreitman <stuart.kreitman@sun.com> 190500Date: Thu Mar 25 05:11:16 2004 +0000 190501 190502 oops 190503 190504commit f0336f18ee4106050104cb060c38fe87541615da 190505Author: Stuart Kreitman <stuart.kreitman@sun.com> 190506Date: Thu Mar 25 03:45:49 2004 +0000 190507 190508 built,working DAMAGE/XFIXES in mono tree 190509 190510commit f1394ec3cec09ff9c5fbbff3c1f595a642b25f91 190511Author: Torrey Lyons <torrey@mrcla.com> 190512Date: Wed Mar 24 22:15:25 2004 +0000 190513 190514 Change XFree86 Project to X.Org Foundation in localized XDarwin splash 190515 screens. 190516 190517commit d2bbcc0deed3a607d347ed4ef07ded9bcb44f1bb 190518Author: Torrey Lyons <torrey@mrcla.com> 190519Date: Wed Mar 24 22:12:34 2004 +0000 190520 190521 Change XFree86 Project to X.Org Foundation in XDarwin splash screen. 190522 190523commit 84d25a5e178835234261a63f0a1b8131c01bbe1f 190524Author: Egbert Eich <eich@suse.de> 190525Date: Wed Mar 24 15:58:45 2004 +0000 190526 190527 file xorgVersion.h was initially added on branch XORG-RELEASE-1-TM. 190528 190529commit 685d1630c1540e29644849254bd45708aa5763bb 190530Author: Egbert Eich <eich@suse.de> 190531Date: Tue Mar 23 12:46:30 2004 +0000 190532 190533 file xorgconfig.man was initially added on branch XORG-RELEASE-1-TM. 190534 190535commit 6631bd586f74f89e2dce74fe1cc25ee982883323 190536Author: Egbert Eich <eich@suse.de> 190537Date: Tue Mar 23 12:46:30 2004 +0000 190538 190539 file xorgconfig.c was initially added on branch XORG-RELEASE-1-TM. 190540 190541commit 05a16f4acce3b6ef4ddafe044f96774de3f75b7c 190542Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 190543Date: Mon Mar 22 13:43:35 2004 +0000 190544 190545 file glwindows.h was initially added on branch CYGWIN. 190546 190547commit 3c2e2d9ae0704931737bb0879a49a8575a07d13b 190548Author: Kaleb Keithley <kaleb@freedesktop.org> 190549Date: Sun Mar 21 22:55:35 2004 +0000 190550 190551 bug #357. Fix XDarwin so it builds on Mac OS X 10.2 and earlier. Merged 190552 down from -RELEASE-1-TM and -RELEASE-1 190553 190554commit 122be5328ffea0a3b92612b8ea0f5b02736ac175 190555Author: Kaleb Keithley <kaleb@freedesktop.org> 190556Date: Fri Mar 19 23:16:06 2004 +0000 190557 190558 no bug report. Restore Credits to the hw/darwin ddx that were deleted 190559 previously. RTF may or may not be "human readable," but with a little 190560 effort it's about as readable as SGML, IMNSHO. 190561 190562commit 551c93da612923f8d32707adc16431979bad6fb0 190563Author: Harold L Hunt II <huntharo@msu.edu> 190564Date: Mon Mar 15 04:33:23 2004 +0000 190565 190566 file winkeynames.h was initially added on branch CYGWIN. 190567 190568commit 0c97b290015186acbaadae9a9bf79f37c3d38b40 190569Author: Harold L Hunt II <huntharo@msu.edu> 190570Date: Mon Mar 15 04:33:23 2004 +0000 190571 190572 file winkeymap.h was initially added on branch CYGWIN. 190573 190574commit dae90c3af98edd5e95289abd930b3872c996c503 190575Author: Egbert Eich <eich@suse.de> 190576Date: Sun Mar 14 08:34:49 2004 +0000 190577 190578 Importing vendor version xf86-4_4_99_1 on Sun Mar 14 00:26:39 PST 2004 190579 190580commit 4e996f9d76f51e9b1e33bef610bb9c2a746c8b9c 190581Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 190582Date: Fri Mar 12 21:05:47 2004 +0000 190583 190584 file winpriv.c was initially added on branch CYGWIN. 190585 190586commit c79b4bfd15534de12aaf8eca9965403b4913ca4b 190587Author: Alexander Gottwald <alexander.gottwald@s1999.tu-chemnitz.de> 190588Date: Fri Mar 12 21:05:47 2004 +0000 190589 190590 file winpriv.h was initially added on branch CYGWIN. 190591 190592commit 24c02f84cc31475bfba27417dfef66b11c09b25c 190593Author: Harold L Hunt II <huntharo@msu.edu> 190594Date: Fri Mar 12 01:11:23 2004 +0000 190595 190596 file X-boxed.ico was initially added on branch CYGWIN. 190597 190598commit 45b638b87f0daf94f9fce566179775fb2889c663 190599Author: Daniel Stone <daniel@fooishbar.org> 190600Date: Wed Mar 10 11:49:11 2004 +0000 190601 190602 Twenty link errors for Xizzle now - count 'em. 20. 190603 Get rid of all references to SCO. 190604 Change SDK include dir to $(includedir)/xizzle. Add SDK libs where 190605 necessary. 190606 Reformat to be nicer and easier to shuffle around; also, fix lib ordering 190607 so we get so much closer to the elusive final link. Shuffle 190608 common/xf86Init.c into libxizzle.a. 190609 Fix a couple of early snafus - s/BUILDXI/XINPUT/, et al; make the SDK stuff 190610 conditional as needed; fix the SBus includes. 190611 Name library os-support/libxizzleos.a, not os-support/foo/libxizzlefoo.a. 190612 Clean up ARCH_SOURCES so it's always initialised to something. 190613 Move linked libraries to _LIBADD, which somehow escaped my attention. Make 190614 inclusion of drm/libxizzlelinuxdrm.a dependent on DRI. 190615 s/VERSION/VBE_VERSION/; 190616 Axe this redundant dir. 190617 All Xizzle-specific: axe hw/xizzle/Xi, shuffle link order, fix list of 190618 required modules to be vaguely sane, add some pertinent libs/incs. 190619 190620commit 519f76a0867fb2711d311b7929632408c3633e37 190621Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 190622Date: Sun Mar 7 23:45:10 2004 +0000 190623 190624 24. Update license for Xinerama code from DEC to the version requested by 190625 Compaq for X11R6.5.1 that allows redistribution without written 190626 permission from DEC. Originally X.org Defect #9263. freedesktop.org 190627 bugzilla #283. (Alan Coopersmith) 190628 190629commit 505fe2ba307e9270627ca7f3cb6b4e1dbacc327b 190630Author: Egbert Eich <eich@suse.de> 190631Date: Fri Mar 5 13:41:12 2004 +0000 190632 190633 23. Merged with XFree86 4.4.0. Added changes that went into infected files. 190634 Reverted darwin/bundle/**/Credits.rtf to XFree86 versions to avoid 190635 future conflicts on ASCII but not humal readable files. (There should 190636 probably be separate CreditsXorg.rtf files) (Egbert Eich). 190637 190638commit 1b22db1ebcf1ba98ca8519fa38210e275373f8f6 190639Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 190640Date: Thu Mar 4 02:13:09 2004 +0000 190641 190642 21. X server crashes when X-Resource has to byte-swap. Sun bug #5007488. 190643 freedesktop.org bugzilla #267. (Alan Coopersmith) 190644 190645commit 47c9395969593a4e897e8c8110d5f2414e47b06a 190646Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 190647Date: Wed Mar 3 17:03:46 2004 +0000 190648 190649 file solaris-ia32.S was initially added on branch XORG-CURRENT. 190650 190651commit ed066cc67b1fca03fb38c80ecb8194b5b40963be 190652Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 190653Date: Wed Mar 3 17:03:46 2004 +0000 190654 190655 Enable inlining of assembly functions for inX/outX on Solaris 8 with Sun 190656 compilers 190657 190658commit 867451f1ab7b9870621725bd4be3dd8694c364b8 190659Author: Egbert Eich <eich@suse.de> 190660Date: Wed Mar 3 12:12:50 2004 +0000 190661 190662 Importing vendor version xf86-4_4_0 on Wed Mar 3 04:09:24 PST 2004 190663 190664commit 2934f0731b3d2bc9c1e25ceab26d9e0d9cadb054 190665Author: Harold L Hunt II <huntharo@msu.edu> 190666Date: Tue Mar 2 20:00:16 2004 +0000 190667 190668 file winvalargs.c was initially added on branch CYGWIN. 190669 190670commit f72efebf280547c80ff7010e32f56416e7121164 190671Author: Harold L Hunt II <huntharo@msu.edu> 190672Date: Tue Mar 2 19:26:34 2004 +0000 190673 190674 Replace a handful of calls to ErrorF and exit(1) with a single call to 190675 FatalError. These direct calls to exit(1) made it impossible to do 190676 anything ddx-specific in these cases; note that most of these calls 190677 occur during argument processing. 190678 190679commit 7557d4da10cc482fcec40acadf7744b04c1615a0 190680Author: Kaleb Keithley <kaleb@freedesktop.org> 190681Date: Tue Mar 2 19:00:06 2004 +0000 190682 190683 bug #230 Revert to Xinerama 1.1 In order to make a "quick" release it has 190684 been decided that the priority is to preserve the server's internal 190685 API/ABI so that third-party drivers that depend on symbols like 190686 noPanoramiXExtension, etc., would not need to be recompiled. Toobad gcc 190687 on Linux doesn't support ELF's weak symbols as that would have been a 190688 reasonable solution for preserving the ABI. N.B.: While symbols, i.e. 190689 functions and variables revert to the old name, I did not revert build 190690 names, i.e. -DXINERAMA, to the old -DPANORAMIX. There was no need, and 190691 it's just a build issue that has no impact on the binary output of the 190692 build. 190693 190694commit 215a13aa8f537dcb62b0a2f6d335901ee47e9e9b 190695Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 190696Date: Mon Mar 1 16:38:20 2004 +0000 190697 190698 Add the .stab.indexstr section produced by Sun's compilers to the list of 190699 SHT_STRTAB sections with debug information to ignore when loading ELF 190700 objects. 190701 190702commit d87b05563dc13ba8d9825ec3bb772702dce6c9fe 190703Author: Harold L Hunt II <huntharo@msu.edu> 190704Date: Mon Mar 1 03:33:28 2004 +0000 190705 190706 file indirect.c was initially added on branch CYGWIN. 190707 190708commit f81d63ec5396c8d7f62ddd6ff6bab10b32493264 190709Author: Kaleb Keithley <kaleb@freedesktop.org> 190710Date: Sun Feb 29 20:11:11 2004 +0000 190711 190712 bug #240 Instead of both Meta keys generating Meta_L, and both Alt keys 190713 generating Alt_L, etc, fix the implementation so that you get Meta_L 190714 for the left Meta key and Meta_R for the right Meta key. Ditto for Alt, 190715 Control, and Shift. 190716 190717commit 6c412a43e42538a51d3a4d92a6db12b0b6cf0e9b 190718Author: Jaymz Julian <jaymz@artificial-stupidity.net> 190719Date: Sat Feb 28 09:47:55 2004 +0000 190720 190721 sdl x server so that we can x-on-x the fb stuff for ease of debugging. if 190722 anyone uses this in production, a big scary monster will eat them. hrm, 190723 perhaps i should make it have a --i-know-what-i'm-doing param that it 190724 doens't start without, heh 190725 190726commit bb93fef9877a885da2c6108410155fa996b19abf 190727Author: Kaleb Keithley <kaleb@freedesktop.org> 190728Date: Fri Feb 27 19:35:49 2004 +0000 190729 190730 bug #238 test for root-window that XFree86 fixed in their 190731 programs/Xserver/Xext/shm.c 190732 3.37 and programs/Xserver/Xext/xvdisp.c 1.26 got zapped when Xinerama2 was 190733 merged into the tree. (Xinerama has since been reverted to 1.1, but 190734 that's another story.) 190735 190736commit cb718ce08eb25c3999c91b8d614fb88237fad03d 190737Author: Kaleb Keithley <kaleb@freedesktop.org> 190738Date: Fri Feb 27 16:17:12 2004 +0000 190739 190740 Revert to Xinerama 1.1 In order to make a "quick" release it has been 190741 decided that the priority is to preserve the server's internal API/ABI 190742 so that third-party drivers that depend on symbols like 190743 noPanoramiXExtension, etc., would not need to be recompiled. Toobad gcc 190744 on Linux doesn't support ELF's weak symbols as that would have been a 190745 reasonable solution for preserving the ABI. N.B.: While symbols, i.e. 190746 functions and variables revert to the old name, I did not revert build 190747 names, i.e. -DXINERAMA, to the old -DPANORAMIX. There was no need, and 190748 it's just a build issue that has no impact on the binary output of the 190749 build. 190750 190751commit df0313d35bc89abe9374ed25533db283430716e0 190752Author: Egbert Eich <eich@suse.de> 190753Date: Thu Feb 26 13:36:15 2004 +0000 190754 190755 readding XFree86's cvs IDs 190756 190757commit 147aae87fde5edeed395f77e60f0f8e812d3b6af 190758Author: Egbert Eich <eich@suse.de> 190759Date: Thu Feb 26 09:23:53 2004 +0000 190760 190761 Importing vendor version xf86-4_3_99_903 on Wed Feb 26 01:21:00 PST 2004 190762 190763commit 8844423f890194bcb0419a38249029f1997c8c66 190764Author: Stuart Kreitman <stuart.kreitman@sun.com> 190765Date: Wed Feb 25 23:28:43 2004 +0000 190766 190767 file xevie.c was initially added on branch XEVIE. 190768 190769commit b052486adb9ea26f37be120966eb60cd3ac3db2f 190770Author: Kaleb Keithley <kaleb@freedesktop.org> 190771Date: Wed Feb 25 21:47:10 2004 +0000 190772 190773 bug #230 Revert to Xinerama 1.1 In order to make a "quick" release it has 190774 been decided that the priority is to preserve the server's internal 190775 API/ABI so that third-party drivers that depend on symbols like 190776 noPanoramiXExtension, etc., would not need to be recompiled. Too bad 190777 gcc on Linux doesn't support ELF's weak symbols as that would have been 190778 a reasonable solution for preserving the ABI. N.B.: While symbols, i.e. 190779 functions and variables revert to the old name, I did not revert build 190780 names, i.e. -DXINERAMA, to the old -DPANORAMIX. There was no need, and 190781 it's just a build issue that has no impact on the binary output of the 190782 build. 190783 190784commit 14ab4ade74e946c09d633b15ab4d447d7b69ea29 190785Author: Kaleb Keithley <kaleb@freedesktop.org> 190786Date: Tue Feb 24 15:22:40 2004 +0000 190787 190788 bug #214. Merge most of 4.4RC3 190789 190790commit 9343c8f5ac180043c29ead5e83a3efef16d7b3f2 190791Author: Kaleb Keithley <kaleb@freedesktop.org> 190792Date: Tue Feb 24 15:16:35 2004 +0000 190793 190794 bug #188, #214, see versions 1.1.4.3 and 1.1.4.4 of this file. fix bad 190795 merge 190796 190797commit 03d893bff9bf5d6be9663a21cc983873d8e8d4c7 190798Author: Kaleb Keithley <kaleb@freedesktop.org> 190799Date: Mon Feb 23 21:37:29 2004 +0000 190800 190801 merge most of XFree86 RC3 (4.3.99.903) from vendor branch. bug #214 190802 190803commit 4ee0a53de870192d57c02baffa106b10bae6e0bf 190804Author: Kaleb Keithley <kaleb@freedesktop.org> 190805Date: Mon Feb 23 20:35:22 2004 +0000 190806 190807 Import most of XFree86 4.4RC3. This import excludes files which have the 190808 new license. If we want to, later we can import 4.4RC3 again and pick 190809 up the files that have the new license, but for now the vendor branch 190810 is "pure." 190811 190812commit dcdd47ebbd4e9b5f4cbb598a5217004df0e80844 190813Author: Kaleb Keithley <kaleb@freedesktop.org> 190814Date: Mon Feb 23 20:35:19 2004 +0000 190815 190816 Initial revision 190817 190818commit 30ac3efde2c3f08b98f31833df4ea7d87f33b092 190819Author: Kaleb Keithley <kaleb@freedesktop.org> 190820Date: Mon Feb 23 16:32:14 2004 +0000 190821 190822 bug #188 report bugs to X.org bugzilla, not XFree86 190823 190824commit d52f3ac58fd596fca392394f16acff84115f6e1d 190825Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 190826Date: Wed Feb 18 21:43:19 2004 +0000 190827 190828 Additional fixes to allow building with Sun compilers on Solaris x86 190829 190830commit 07109fd63e0999905e6f7df8fd7f9c713d0dc2cc 190831Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 190832Date: Wed Feb 18 21:30:12 2004 +0000 190833 190834 Sun cc on Solaris x86 defines __i386 but not __i386__ so the x86 190835 architecture #ifdef should accept either form 190836 190837commit a27ffd2678ef76453c4fa27932462425d804df6d 190838Author: Warren Turkal <wt@penguintechs.com> 190839Date: Wed Feb 18 02:12:44 2004 +0000 190840 190841 completely get rid of NeedNestedPrototypes 190842 completely get rid of NeedVarargsPrototypes 190843 remove a lot of NeedFunctionPrototypes 190844 ansify many function declarations 190845 190846commit d17586c4dc858d0127fa021e6db62f8cc28ef7a6 190847Author: Alan Coopersmith <Alan.Coopersmith@sun.com> 190848Date: Mon Feb 16 20:19:59 2004 +0000 190849 190850 [fd.o bugzilla #189] _XOPEN_SOURCE defines break builds on Solaris Express 190851 190852commit b146ef1548d36d6897fbd674f1c3b8324bed11a7 190853Author: Warren Turkal <wt@penguintechs.com> 190854Date: Sun Feb 15 15:04:57 2004 +0000 190855 190856 Moving toward a working input extension. 190857 190858commit e90274c2bba1f66a68c2bc30ddb589dbf6fa0929 190859Author: Egbert Eich <eich@suse.de> 190860Date: Wed Feb 11 19:29:37 2004 +0000 190861 190862 2. Fixing segfaults that may happen in some corner cases when VT switching 190863 and during int10 initialization (Egbert Eich). 190864 190865commit 453a0743eb524da88dd364ccac86f35e61899e64 190866Author: Kaleb Keithley <kaleb@freedesktop.org> 190867Date: Sun Feb 8 00:17:31 2004 +0000 190868 190869 revert to RC1 version of file with the license we like 190870 190871commit d6f33d897221450f3cfcc1162e2a6d09b227326e 190872Author: Kaleb Keithley <kaleb@freedesktop.org> 190873Date: Sun Feb 8 00:12:27 2004 +0000 190874 190875 revert to RC1 version of the file with the license we like 190876 190877commit bd20c8d340fce0700ae813bd5b55fe7f4b9e0c98 190878Author: Jaymz Julian <jaymz@artificial-stupidity.net> 190879Date: Thu Feb 5 09:09:51 2004 +0000 190880 190881 Polling input mode for the kdrive os layer. And a moose! 190882 190883commit af798d27743dbc4f70e85e297daa5863ec89640b 190884Author: Jaymz Julian <jaymz@artificial-stupidity.net> 190885Date: Wed Feb 4 16:08:27 2004 +0000 190886 190887 More NULL checks. These ones are more useful than the last (which just made 190888 debugging a bunch of problems easier), since you can implement less in 190889 the basic simplest case driver now (not that i'm lazy, mind :-p) 190890 190891commit 3c64b65d805915e5c5628663113c54c3e9c3013b 190892Author: Egbert Eich <eich@suse.de> 190893Date: Thu Jan 29 08:08:57 2004 +0000 190894 190895 Importing vendor version xf86-012804-2330 on Thu Jan 29 00:06:33 PST 2004 190896 190897commit 2ec70aa70133190ad31a83114fdb9a218e6aa8e6 190898Author: Eric Anholt <anholt@freebsd.org> 190899Date: Sun Jan 25 05:31:24 2004 +0000 190900 190901 Disable GLX visuals code on !GLXEXT, and remove a useless prototype. 190902 190903commit 01e9cc858ac646b3140d1d85ea9c069bc708fb28 190904Author: Eric Anholt <anholt@freebsd.org> 190905Date: Sun Jan 25 01:30:33 2004 +0000 190906 190907 - Add glx visuals code based on XFree86's Radeon driver. 190908 - Reserve areas for back/depth/span when USING_DRI && GLXEXT. This would be 190909 better in a TransitionTo3d, but we'd need to work with the offscreen 190910 memory manager for that. 190911 - Misc. fixes to ati_dri.c for DRI+GLX. Needs more work still. 190912 190913commit f2bedd17af7c3b9241c02dc1c899f32fc0cd2f10 190914Author: Eric Anholt <anholt@freebsd.org> 190915Date: Sun Jan 25 01:16:19 2004 +0000 190916 190917 Oops, turn fallback output back off. 190918 190919commit 26c5a8dfdd2aa09db46c4cf963ca697df3e777ef 190920Author: Eric Anholt <anholt@freebsd.org> 190921Date: Sun Jan 25 01:04:12 2004 +0000 190922 190923 Whitespace cleanup. 190924 190925commit 6870c081572fcf32997e7906a54d09da0ca58ac3 190926Author: Keith Packard <keithp@keithp.com> 190927Date: Thu Jan 15 09:19:56 2004 +0000 190928 190929 Oops, lost a diff needed for the non-screen format pixmap code 190930 190931commit 3867e03cb63e49aeb1742a8a4bdaed0b7a23749e 190932Author: Keith Packard <keithp@keithp.com> 190933Date: Thu Jan 15 09:15:53 2004 +0000 190934 190935 Fix up some mis-used variable names 190936 190937commit f233bbf3652327e62e03efbb8a355e6af2703a1c 190938Author: Keith Packard <keithp@keithp.com> 190939Date: Thu Jan 15 09:13:01 2004 +0000 190940 190941 Accelerate non-screen format pixmaps. 190942 190943commit 751fb0374b12679f63c922adf0f0e7cadd83d861 190944Author: Harold L Hunt II <huntharo@msu.edu> 190945Date: Thu Jan 15 06:06:44 2004 +0000 190946 190947 file winresource.h was initially added on branch CYGWIN. 190948 190949commit a265167f19e37aec2173c0ca6c9955450aa69941 190950Author: Anders Carlsson <andersca@gnome.org> 190951Date: Wed Jan 14 10:00:00 2004 +0000 190952 190953 Add IPAQ modeline by Dennis Noordsij. 190954 190955commit 6c97b277d9140b9d6bca047c56e303f6fc1d92e0 190956Author: Eric Anholt <anholt@freebsd.org> 190957Date: Sun Jan 11 00:10:34 2004 +0000 190958 190959 Support 1x1 repeat sources in R128's Blend. 190960 190961commit 92702565657d48f1fcc2bae1b5989b1d6d3dd164 190962Author: Eric Anholt <anholt@freebsd.org> 190963Date: Fri Jan 9 08:43:48 2004 +0000 190964 190965 Change PCI ID information field to be one of r128, r100, r200, r300. This 190966 is all the information we need so far. Put that information into atic, 190967 and use it correctly in the code (unlike before). 190968 190969commit 5d51dfc69cb245f6a1c7b106954a3365524741e2 190970Author: Eric Anholt <anholt@freebsd.org> 190971Date: Fri Jan 9 08:40:32 2004 +0000 190972 190973 Use the scratch area for Composite when one of src or dst is in memory. 190974 190975commit 6d8001f4688e2149fcdd480401c46c7540680576 190976Author: Eric Anholt <anholt@freebsd.org> 190977Date: Thu Jan 8 20:18:13 2004 +0000 190978 190979 Compile fixes for non-DRI case and for non-C99 compiler. 190980 190981commit 1be4b2d5e8048eb3653fad3a1267a0da865bcee8 190982Author: Eric Anholt <anholt@freebsd.org> 190983Date: Thu Jan 8 08:25:49 2004 +0000 190984 190985 Forced commit: Previous commit included the removal of the 8192 scanline 190986 limit on offscreen memory in the fbdev case. I remember daenzer (who 190987 originally put that code in) saying he wasn't sure of it, and there 190988 doesn't seem to be any reason for that limit given how acceleration is 190989 done. 190990 190991commit b27729ec88f5d4153a0debfe2347bbed022329ba 190992Author: Eric Anholt <anholt@freebsd.org> 190993Date: Thu Jan 8 08:16:24 2004 +0000 190994 190995 - Add a new UploadToScratch kaa hook for putting the data for a single 190996 pixmap into temporary offscreen storage. Subsequent UploadToScratch may 190997 clobber the data of previous ones. This allows hardware acceleration of 190998 composite operations on glyphs. 190999 - Add a new UploadToScreen kaa hook for doing the actual moving of data to 191000 framebuffer. This would allow us to do things like hostdata blits or 191001 memcpy to agp and then blit. 191002 - Add an UploadToScreen on ATI which is just memcpy, but which will be 191003 replaced with a hostdata blit soon. 191004 - Add UploadToScratch on ATI and reserve 64k of scratch space. This 191005 provided a 3x speedup of rgb24text on my Radeon. 191006 191007commit d640cf4cb4e031a0e93dfd5955405847fe4475c0 191008Author: Harold L Hunt II <huntharo@msu.edu> 191009Date: Thu Jan 8 05:10:33 2004 +0000 191010 191011 file winprocarg.c was initially added on branch CYGWIN. 191012 191013commit 77183abbc499c69fbbbae1d92a6b012c5f80b6c4 191014Author: Harold L Hunt II <huntharo@msu.edu> 191015Date: Thu Jan 8 05:10:32 2004 +0000 191016 191017 file winglobals.c was initially added on branch CYGWIN. 191018 191019commit eb5bb9c1a16db308eae84ea45a5920c768d4a2ff 191020Author: Harold L Hunt II <huntharo@msu.edu> 191021Date: Thu Jan 8 05:10:32 2004 +0000 191022 191023 file winclipboardwrappers.c was initially added on branch CYGWIN. 191024 191025commit e93d468df21840007cbeea03ed545e75f0f0baf1 191026Author: Harold L Hunt II <huntharo@msu.edu> 191027Date: Thu Jan 8 05:10:32 2004 +0000 191028 191029 file winauth.c was initially added on branch CYGWIN. 191030 191031commit 737eddfa4b6a8851e20823405b7269dd49c49b89 191032Author: Eric Anholt <anholt@freebsd.org> 191033Date: Wed Jan 7 09:50:28 2004 +0000 191034 191035 Disconnect the sis300 driver. I've never managed to fix it, and it breaks 191036 the build on PPC. 191037 191038commit 5a2c23f8a18767f0eb2fe2846ca3ba18fd236284 191039Author: Eric Anholt <anholt@freebsd.org> 191040Date: Wed Jan 7 02:30:29 2004 +0000 191041 191042 Speed things up slightly by removing Z values from emitted vertices and by 191043 emitting as a tri fan rather than a tri list. A rect list would save an 191044 additional vertex (out of 4) per rectangle, but there's no measurable 191045 speed difference and the tri fan may be useful when transforms come 191046 into play. 191047 191048commit cff782078cec9b10606c5873816b7acd9977ce4d 191049Author: Eric Anholt <anholt@freebsd.org> 191050Date: Sun Jan 4 20:51:53 2004 +0000 191051 191052 - Don't forget to UNINIT miComputeCompositeRegion's regions 191053 - Fix a bit of whitespace nearby. 191054 191055commit 34d1529731fff0cb61c71f76edc5c6499ece68d1 191056Author: Eric Anholt <anholt@freebsd.org> 191057Date: Sun Jan 4 20:47:30 2004 +0000 191058 191059 - Correctly set the texture coordinate set source for the second texture 191060 unit. 191061 - Re-enable Radeon's Composite accel now that fonts work again. 191062 191063commit 9f1a92cd092e87f774ce4ed99d4b3e15f905d4f7 191064Author: Eric Anholt <anholt@freebsd.org> 191065Date: Sat Jan 3 21:52:14 2004 +0000 191066 191067 - Call appropriate Done function for Composite. 191068 - Don't allow src transforms for Copy acceleration. 191069 - Minor whitespace fixes. 191070 191071commit 3db761a17b60b80acb83f365628b093f0ba6958c 191072Author: Eric Anholt <anholt@freebsd.org> 191073Date: Sat Jan 3 11:46:57 2004 +0000 191074 191075 - Add more Composite operations, including Saturate, to Radeon Composite 191076 accel. I don't 100% trust that the math works for Saturate, but I can't 191077 tell from existing information. 191078 - Fix texture pitch fallback checks. 191079 - Fallback when src or mask have transforms. 191080 - Disable Radeon Composite accel until the offset thing is fixed. 191081 - Set offscreenPitch to 64 on Radeon thanks to new information and a kaa 191082 fix. Fixes acceleration at width!=1024. 191083 191084commit d15acfa79b64b8dab1e930ce8e5423a212a1360b 191085Author: Eric Anholt <anholt@freebsd.org> 191086Date: Sat Jan 3 11:25:27 2004 +0000 191087 191088 Split the various attempts at accelerating Composite into separate 191089 functions. Along with making things more readable, it fixes a problem 191090 where the coordinates would get messed up if acceleration failed due to 191091 things like pixmaps being in the wrong locations. 191092 191093commit 1e1a35e20c1d281bc9700b349cda1e67f65905dd 191094Author: Eric Anholt <anholt@freebsd.org> 191095Date: Sat Jan 3 11:17:44 2004 +0000 191096 191097 Actually align the offset of allocated offscreen areas. 191098 191099commit 354f8f7e943d1a0732f4181420211efff27532b8 191100Author: Eric Anholt <anholt@freebsd.org> 191101Date: Wed Dec 31 23:24:33 2003 +0000 191102 191103 Some strange \240 character snuck into the original commit of this file. 191104 191105commit 5f947b04da13256e5f514c40dedb98c6e1cbe0f1 191106Author: Eric Anholt <anholt@freebsd.org> 191107Date: Tue Dec 30 08:45:53 2003 +0000 191108 191109 There's never a copy between different depths. Remove the check. 191110 191111commit c8eb20a08ee9174374b6f5ac6e79f31fce26e181 191112Author: Eric Anholt <anholt@freebsd.org> 191113Date: Tue Dec 30 08:23:56 2003 +0000 191114 191115 - Add new Composite hook for kdrive drivers, which only ensures that the 191116 pixmaps are offscreen and don't have alpha maps. It is the last case 191117 checked before going to software fallback 191118 - Use the new Composite hook in the ati driver to implement acceleration of 191119 most Composites that get done in an xcompmgr environment on r100 series 191120 cards. It is only available when using the DRM. There are still some 191121 corruption issues, but the DRI is still non-default and I need to get 191122 this into version control. 191123 191124commit adfc1ed8e1e150100accf014e46241201275138f 191125Author: Eric Anholt <anholt@freebsd.org> 191126Date: Mon Dec 29 09:04:20 2003 +0000 191127 191128 Add dependency lines so that servers are rebuilt when server libraries are 191129 changed. 191130 191131commit df03e80ae9162ec87f503322ccbcf2846ad38bef 191132Author: Eric Anholt <anholt@freebsd.org> 191133Date: Mon Dec 29 06:24:01 2003 +0000 191134 191135 Merge dri-0-1-branch to trunk. Notable changes: 191136 - Add libdrm and libdri. Portions of the DRI extension are stubbed out. 191137 - Use the DRM in the ATI driver when available. This provides a minor 191138 performance improvement in x11perf, and opens the possibility of using 191139 the 3d hardware for acceleration in the future. 191140 - Implement solid fill acceleration for Composite in KAA. 191141 - Implement Blend hook for Composite and use it on r128. 191142 - Fix a bug of mine that resulted in overuse of offscreen memory. 191143 - Fix many miscellaneous bugs in ATI driver and add PCI IDs. 191144 191145commit 9bea538745f1a0c14faaac0e61dee5cf86f98dc6 191146Author: Eric Anholt <anholt@freebsd.org> 191147Date: Sun Dec 28 09:56:54 2003 +0000 191148 191149 file kaa.h was initially added on branch dri-0-1-branch. 191150 191151commit fb8cd7454baec0bc0f693d222f3920ce03dde7c6 191152Author: Eric Anholt <anholt@freebsd.org> 191153Date: Sun Dec 28 09:56:54 2003 +0000 191154 191155 file kaapict.c was initially added on branch dri-0-1-branch. 191156 191157commit f388f1509cb131cdf0675415214c9610d3d322c7 191158Author: Eric Anholt <anholt@freebsd.org> 191159Date: Tue Dec 23 22:29:38 2003 +0000 191160 191161 file r128_blendtmp.h was initially added on branch dri-0-1-branch. 191162 191163commit 918958705dd97ce678b8901666c85fb359d0e013 191164Author: Kaleb Keithley <kaleb@freedesktop.org> 191165Date: Sun Dec 21 13:39:58 2003 +0000 191166 191167 Use a different icon 191168 191169commit e97c634593dd171a05aa0fa5a35d218dcc3ecfb0 191170Author: Kaleb Keithley <kaleb@freedesktop.org> 191171Date: Sat Dec 20 00:28:31 2003 +0000 191172 191173 merge XFree86 RC2 (4.3.99.902) from vendor branch 191174 191175commit 12e532010b9e8cb67bedd44d489c9c40dd265165 191176Author: Kaleb Keithley <kaleb@freedesktop.org> 191177Date: Fri Dec 19 20:55:39 2003 +0000 191178 191179 XFree86 4.3.99.902 (RC 2) 191180 191181commit 4b75c7f6358b28978b05ffa4b73853d936454f50 191182Author: Kaleb Keithley <kaleb@freedesktop.org> 191183Date: Thu Dec 18 19:32:17 2003 +0000 191184 191185 First pass at "Standard" Xinerama. The sources for this came from Heather 191186 Lanigan's xinerama tree on Sourceforge.Net. No attempt has been made to 191187 handle previous, non-standard versions of the protocol. Nor has any 191188 attempt been made to preserve the ABI of previous versions -- that part 191189 will be added at a later time, and then probably only on systems that 191190 have nice object/linker semantics, e.g. ELF systems with weak symbols. 191191 191192commit 305c444de3baa863d7abc4221e8cebb973805847 191193Author: Eric Anholt <anholt@freebsd.org> 191194Date: Mon Dec 8 01:55:10 2003 +0000 191195 191196 Add initial SiS 300-series (300, 305, 540, 630, 730) driver based off of 191197 the ATI driver. It suffers from hw/sw synchronization problems, it 191198 looks like, but may be good enough to work on Render acceleration 191199 experiments. Committing it as-is so I don't lose it again. 191200 191201commit 8a7481a27496c842ec2ef5bac5e4d0b5e6279deb 191202Author: Kaleb Keithley <kaleb@freedesktop.org> 191203Date: Sat Dec 6 13:24:29 2003 +0000 191204 191205 merge XFree86 4.3.99.901 (RC1) from vendor branch 191206 191207commit e82928826f60a2e76a670c936bd557838fc1764c 191208Author: Kensuke Matsuzaki <zakki@peppermint.jp> 191209Date: Fri Dec 5 03:37:26 2003 +0000 191210 191211 file winwin32rootlesswindow.c was initially added on branch CYGWIN. 191212 191213commit fc40d0a3cbee053d446032ae3150b06edf66a335 191214Author: Kensuke Matsuzaki <zakki@peppermint.jp> 191215Date: Fri Dec 5 03:37:26 2003 +0000 191216 191217 file winwin32rootlesswndproc.c was initially added on branch CYGWIN. 191218 191219commit c3f26a1b989dbbf5167e6e352aebf2f53bfcc442 191220Author: Kensuke Matsuzaki <zakki@peppermint.jp> 191221Date: Fri Dec 5 03:37:26 2003 +0000 191222 191223 file winwindowswm.c was initially added on branch CYGWIN. 191224 191225commit 0f06636a9a088fc27262da0f0bc9a20a3dbeab69 191226Author: Kensuke Matsuzaki <zakki@peppermint.jp> 191227Date: Fri Dec 5 03:37:26 2003 +0000 191228 191229 file winwin32rootless.c was initially added on branch CYGWIN. 191230 191231commit a84f16a9ad2ed0f874d2c1816aedee96725d2657 191232Author: Kaleb Keithley <kaleb@freedesktop.org> 191233Date: Thu Dec 4 22:03:38 2003 +0000 191234 191235 XFree86 4.3.99.901 (RC 1) 191236 191237commit 33fdd50a94baab1db342bfce442907db8f8ad03e 191238Author: Keith Packard <keithp@keithp.com> 191239Date: Tue Dec 2 01:59:38 2003 +0000 191240 191241 Wrap ClipNotify to see region motions during MoveWindow. 191242 Check window reorigin in PositionWindow and bump pixmap serial numbers to 191243 revalidate GCs. 191244 Fix picture clip region origin in automatic update 191245 Initialize client private 'critical' value 191246 Clean up pixmap bounds checking code to only affect contents allocated by 191247 fb. 191248 Oops. Call SourceValidate for Composite operations. 191249 Add Xchips server (vesa based) 191250 191251commit 6db77925406a0ee600998ad558a50190ba631649 191252Author: Eric Anholt <anholt@freebsd.org> 191253Date: Mon Dec 1 22:56:06 2003 +0000 191254 191255 file radeon_sarea.h was initially added on branch dri-0-1-branch. 191256 191257commit 313046b42832fa2434d617997d5701157b55e7c8 191258Author: Eric Anholt <anholt@freebsd.org> 191259Date: Mon Dec 1 22:56:06 2003 +0000 191260 191261 file radeon_common.h was initially added on branch dri-0-1-branch. 191262 191263commit 2ad126286e524392741164babe530210892c377f 191264Author: Eric Anholt <anholt@freebsd.org> 191265Date: Mon Dec 1 22:56:06 2003 +0000 191266 191267 file r128_sarea.h was initially added on branch dri-0-1-branch. 191268 191269commit f486c136ad8d2d893cdf3aee6aa752578a6809d7 191270Author: Eric Anholt <anholt@freebsd.org> 191271Date: Mon Dec 1 22:56:06 2003 +0000 191272 191273 file r128_common.h was initially added on branch dri-0-1-branch. 191274 191275commit 1eb63ef1b5d0ce10117196df3e81e8312a55a93c 191276Author: Eric Anholt <anholt@freebsd.org> 191277Date: Mon Dec 1 22:56:06 2003 +0000 191278 191279 file ati_sarea.h was initially added on branch dri-0-1-branch. 191280 191281commit 1fa5f28406f8a7ad54049e4bec129953149bd3ed 191282Author: Eric Anholt <anholt@freebsd.org> 191283Date: Mon Dec 1 22:56:06 2003 +0000 191284 191285 file ati_dri.c was initially added on branch dri-0-1-branch. 191286 191287commit 14ce4f2c3ebc20e5f6d57adda0a7e14229541a72 191288Author: Eric Anholt <anholt@freebsd.org> 191289Date: Mon Dec 1 22:56:06 2003 +0000 191290 191291 file ati_dri.h was initially added on branch dri-0-1-branch. 191292 191293commit 8887456e7d4fc1280287ed3e35c6c4464525827b 191294Author: Eric Anholt <anholt@freebsd.org> 191295Date: Mon Dec 1 22:56:06 2003 +0000 191296 191297 file ati_dripriv.h was initially added on branch dri-0-1-branch. 191298 191299commit 6c9e7f47357b02f41b9f1f43f7f1d9b5a139e5b6 191300Author: Eric Anholt <anholt@freebsd.org> 191301Date: Mon Dec 1 22:56:06 2003 +0000 191302 191303 file ati_draw.h was initially added on branch dri-0-1-branch. 191304 191305commit 2ad20d4e99d0d8ecb922507e0bfead8b7b7d4a55 191306Author: Eric Anholt <anholt@freebsd.org> 191307Date: Mon Dec 1 22:56:06 2003 +0000 191308 191309 file ati_drawtmp.h was initially added on branch dri-0-1-branch. 191310 191311commit ae2454f65698eef66b3507e586e4f8125cb1790d 191312Author: Anders Carlsson <andersca@gnome.org> 191313Date: Mon Dec 1 22:11:12 2003 +0000 191314 191315 Add support for setting the video mode. 191316 191317commit e31051ba26c18f6232798c5a5c4725f5ce53d6b9 191318Author: Phil Blundell <pb@reciva.com> 191319Date: Mon Dec 1 21:49:41 2003 +0000 191320 191321 Call KdShadowUnset before fbdevSetShadow. 191322 191323commit 9cdd6fd9e3d6e44adf392279093f92fb6678a49c 191324Author: Eric Anholt <anholt@freebsd.org> 191325Date: Mon Dec 1 04:33:36 2003 +0000 191326 191327 - Add fbdev mode-setting backend to Xati. It and vesa are compiled in when 191328 available, with fbdev being used by default. 191329 - Use depth 16 by default when vesa backend is used. 191330 - Add MMIO defines for PowerPC (should be in a common location). 191331 Many thanks for Michel Daenzer for much of this code. 191332 191333commit ec7f5539302fafd1ac7609ac423f1379f54916ab 191334Author: Eric Anholt <anholt@freebsd.org> 191335Date: Mon Dec 1 03:15:13 2003 +0000 191336 191337 Add more RV250 PCI IDs. 191338 191339commit d221c484f9521c780fc3c7e88833c62e50463c6d 191340Author: Eric Anholt <anholt@freebsd.org> 191341Date: Mon Dec 1 01:46:42 2003 +0000 191342 191343 Remove sys/io.h inclusion from some files that didn't need it, and change 191344 asm/io.h to sys/io.h in vga.c, which newer Linux complains about. 191345 191346commit 5fd7f82390d78621a8aad959eb216b8fb7e1a97f 191347Author: Kaleb Keithley <kaleb@freedesktop.org> 191348Date: Sun Nov 30 16:47:13 2003 +0000 191349 191350 xfree86 merge 191351 191352commit 77836ebda2a47e04c56c5842f62fab5992909712 191353Author: Kaleb Keithley <kaleb@freedesktop.org> 191354Date: Sat Nov 29 16:08:53 2003 +0000 191355 191356 xfree86 merge 191357 191358commit 88193e928be7cf5a526b50ad1e0b4ac9cddef297 191359Author: Kaleb Keithley <kaleb@freedesktop.org> 191360Date: Sat Nov 29 15:10:02 2003 +0000 191361 191362 xfree86 merge 191363 191364commit 0097b6fe2d1739e46e4e7726aaa481b6dc84870c 191365Author: Kaleb Keithley <kaleb@freedesktop.org> 191366Date: Wed Nov 26 22:49:07 2003 +0000 191367 191368 merge latest (4.3.99.16) from XFree86 (vendor) branch 191369 191370commit d803918a9fb5f80a2d6e4b711d8e43916cd09da5 191371Author: Eric Anholt <anholt@freebsd.org> 191372Date: Tue Nov 25 22:39:54 2003 +0000 191373 191374 Add new Radeon 9200 PCI IDs. 191375 191376commit adc7f9a4ebdfe11d4cd6de9388b63dfe36450b39 191377Author: Kaleb Keithley <kaleb@freedesktop.org> 191378Date: Tue Nov 25 19:29:01 2003 +0000 191379 191380 XFree86 4.3.99.16 Bring the tree up to date for the Cygwin folks 191381 191382commit 90f1536dd315cd265bfc7ef35058761a65a01734 191383Author: Kaleb Keithley <kaleb@freedesktop.org> 191384Date: Tue Nov 25 19:29:01 2003 +0000 191385 191386 Initial revision 191387 191388commit d461855a73d8c9f51a18673aef7ce88f94a71629 191389Author: Eric Anholt <anholt@freebsd.org> 191390Date: Sun Nov 23 10:12:04 2003 +0000 191391 191392 - Fix Radeon offscreen pixmap pitch alignment. 191393 - Remove usleeps from idle and waitavail code, recommended by keithp. 191394 - Add a workaround for apparent broken acceleration with Rage 128 and 191395 offset alignment with 8-bit acceleration (24-hack and plain 8-bit). 191396 - Minor cleanup of setup code. 191397 191398commit 8e09afe657b110bc1ea9e58dea81a120c343d16e 191399Author: Eric Anholt <anholt@freebsd.org> 191400Date: Sun Nov 23 02:08:16 2003 +0000 191401 191402 Add RV250 PCI IDs. 191403 191404commit b3247251fb7d9f2d50ef41d9c2089629544d534d 191405Author: Eric Anholt <anholt@freebsd.org> 191406Date: Thu Nov 20 07:49:46 2003 +0000 191407 191408 - Fix a bug in pitch alignment for offscren pixmaps. 191409 - Add 24-bit acceleration for Xati using the 8-bit trick from mach64. 191410 - Add offscreen pixmap support to Xati. 191411 191412commit 41dde24b229f4bc4738637d9cd0a86b74b9f8457 191413Author: Eric Anholt <anholt@freebsd.org> 191414Date: Thu Nov 20 00:05:56 2003 +0000 191415 191416 - Fix confusion of depth/bitsPerPixel in ati_draw.c 191417 - Disable acceleration with 24bpp due to apparent broken acceleration. 191418 Accel at 24bpp was the cause of the crashes when people tried to use 191419 any depth over 16. XFree86 doesn't support 24 either. 191420 - Disable at < 8bpp, too. 191421 - Add the other Rage 128 PCI IDs. 191422 - Remove unnecessary setting of scissor registers (only default scissor 191423 gets used). 191424 191425commit 7abbcce4222958b4670873a17f67ea1ec1d958e2 191426Author: Eric Anholt <anholt@freebsd.org> 191427Date: Wed Nov 19 08:32:38 2003 +0000 191428 191429 - Add PCI information (device/vendor id, bus location) to KdCardAttr to 191430 help with ati, and future DRM drivers. 191431 - Add new "ati" kdrive driver. It has ancestry in the r128 driver from 191432 andersca, but took a detour through being the WIP SiS 300 driver on the 191433 way. It supports Radeons (tested on QD VIVO and 7500) and Rage 128. 191434 Current limitations include that it requires depth 16 and that the 191435 other Rage 128 PCI IDs aren't included yet. 191436 191437commit e9cb70dae0a85fcd116d7ffac73d1322ec282a94 191438Author: Keith Packard <keithp@keithp.com> 191439Date: Tue Nov 18 18:38:48 2003 +0000 191440 191441 Reset available offscreen segment save function pointer. (Not a functional 191442 change, just cleaning up a bit) 191443 Reset the screen->memory_base value when frame buffer is remapped. This 191444 makes sure new off-screen allocations point at the newly mapped region 191445 rather than the old (now unmapped) region. 191446 191447commit f333581b7b4066ab054dd765a1c9bae0f3407188 191448Author: Seth W. Klein <sk@sethwklein.net> 191449Date: Tue Nov 18 04:53:16 2003 +0000 191450 191451 Added infrastructure for driver specific usage messages and added vesa 191452 usage message. 191453 191454commit 003e87717cfe378261ed1a1e7bcb2cf0d200b1b8 191455Author: Keith Packard <keithp@keithp.com> 191456Date: Mon Nov 17 22:09:12 2003 +0000 191457 191458 Recompute winSize/borderSize to fix them when changing redirection. 191459 Sufficient, but not always necessary. 191460 Add borderClip to damage on creation so that clients needn't guess. 191461 Fix API to FbDots functions to make PolyPoint work with screen_x/screen_y 191462 offsets 191463 Add debugging code to make sure no pictures are left pointing at freed 191464 pixmaps. "Can't" happen, but it did once. 191465 Change KdOffscreenArea structure to eliminate separate private structure, 191466 eliminate the ScreenPtr, change from doubly linked to singly linked 191467 list. 191468 Don't damage BackgroundNone windows on PW_BACKGROUND. Re-clip damage to 191469 borderClip in DamageSubtract. 191470 191471commit c57959ad6a4c0f5329762f401fd7871ffb2ee90c 191472Author: Kaleb Keithley <kaleb@freedesktop.org> 191473Date: Mon Nov 17 19:03:47 2003 +0000 191474 191475 merge XFree86 4.3.0.1 to -CURRENT 191476 191477commit acd200770513ad03dd3f4bdc7448edfd69b1ff9d 191478Author: Keith Packard <keithp@keithp.com> 191479Date: Mon Nov 17 06:54:52 2003 +0000 191480 191481 Clear window when Manual Subwindows redirect is destroyed 191482 Fix Tile/Stipple origin with non-zero pixmap window origins 191483 Use computed depth for TrueColor visuals as fbdev doesn't have one. 191484 191485commit d568221710959cf7d783e6ff0fb80fb43a231124 191486Author: Kaleb Keithley <kaleb@freedesktop.org> 191487Date: Fri Nov 14 16:49:22 2003 +0000 191488 191489 XFree86 4.3.0.1 191490 191491commit 9508a382f8a9f241dab097d921b6d290c1c3a776 191492Author: Kaleb Keithley <kaleb@freedesktop.org> 191493Date: Fri Nov 14 16:48:57 2003 +0000 191494 191495 Initial revision 191496 191497commit ded6147bfb5d75ff1e67c858040a628b61bc17d1 191498Author: Kaleb Keithley <kaleb@freedesktop.org> 191499Date: Fri Nov 14 15:54:54 2003 +0000 191500 191501 R6.6 is the Xorg base-line 191502 191503commit cb6ef07bf01e72d1a6e6e83ceb7f76d6534da941 191504Author: Keith Packard <keithp@keithp.com> 191505Date: Fri Nov 14 07:46:20 2003 +0000 191506 191507 Accelerated image text code drew glyphs at wrong location when compositing 191508 manager enabled. 191509 191510commit 318d525bf2fe52b059b1568e9b31d144b297a781 191511Author: Keith Packard <keithp@keithp.com> 191512Date: Thu Nov 13 09:14:29 2003 +0000 191513 191514 Have the composite extension tell the damage extension about clients which 191515 have redirected subwindows in manual mode. Those clients are marked 191516 Critical and given a significant scheduling boost whenever they receive 191517 a damage notify event. This dramatically improves update frequency. 191518 If the kernel reported a large number of keys, readKernelMapping would walk 191519 off the end of the kdKeysym array. 191520 Fix usage of _IOWR; the 'size' argument is actually a datatype. 191521 191522commit e9904cefa39e1c9d3c7bf2f335dbafb23809cdba 191523Author: Eric Anholt <anholt@freebsd.org> 191524Date: Tue Nov 11 05:46:15 2003 +0000 191525 191526 Fix Makefile.am's to include header files needed for distcheck. Remove 191527 -Werror default and add --enable-werror switch to add it back. 191528 191529commit 5dc119b73ffcae32c2d470b734dfa2f8af58e77e 191530Author: Keith Packard <keithp@keithp.com> 191531Date: Mon Nov 10 20:35:05 2003 +0000 191532 191533 Fix KdXv interface to pass drawable down so that bits can be put into 191534 drawable pixmap rather than directly into the frame buffer. Rewrite 191535 logic in kdoffscreen to make space for new allocations, now deals 191536 correctly with locked areas. 191537 191538commit e500986657ea8b4e14a1ff4730ecda4583c75277 191539Author: Keith Packard <keithp@keithp.com> 191540Date: Mon Nov 10 06:40:23 2003 +0000 191541 191542 Must offset composite parameters by drawable coordinates in accelerated 191543 case. 191544 191545commit d694b44259ff51cfca2c3ec9a58bf164010cc1ad 191546Author: Phil Blundell <pb@reciva.com> 191547Date: Sat Nov 8 18:51:59 2003 +0000 191548 191549 Avoid ugly shell error when libXdmcp isn't present. 191550 New functions. (TsInit): Register them. 191551 191552commit 3e18c6363454aa87e1ad8c121019bab185e613b8 191553Author: Keith Packard <keithp@keithp.com> 191554Date: Sat Nov 8 00:28:19 2003 +0000 191555 191556 Fix KAA to work with screen_x/screen_y hacks from COMPOSITE Also add 191557 kaaComposite acceleration for simple bltblt case. 191558 191559commit 1280f79054dc16ccf321006cd2de53e0f53c7b70 191560Author: Keith Packard <keithp@keithp.com> 191561Date: Fri Nov 7 23:29:29 2003 +0000 191562 191563 Ok, Composite extension is semi-working; when no-one asks for redirection, 191564 the server seems to act as before. With RedirectSubwindows (root, 191565 automatic), the server looks just like a regular X server. Now to go 191566 rewrite the (currently lame) compositing manager to get some real 191567 action on the screen. 191568 Some of the fixes here are to make valgrind quiet with various ioctls used 191569 by kdrive/linux. 191570 Also fixed a bug where fbdev initialization was out of order in fbdev.c and 191571 smi.c 191572 191573commit d319a0a610c90524ad29cab3c6d4d21b5298fc7f 191574Author: Phil Blundell <pb@reciva.com> 191575Date: Fri Nov 7 23:00:06 2003 +0000 191576 191577 Fix warnings. 191578 191579commit 598c5d549abbb819f3391a2c88432941b546a213 191580Author: Matthew Allum <breakfast@10.am> 191581Date: Thu Nov 6 14:01:46 2003 +0000 191582 191583 tslib improvements 191584 191585commit 7e1a564c416f6dc337a0021b1c0e1f2cb3b27296 191586Author: Keith Packard <keithp@keithp.com> 191587Date: Wed Nov 5 06:46:13 2003 +0000 191588 191589 Replace translucent compositing hacks with PictOpSrc to match eventual 191590 extension semantics. Replace mouse acceleration with quadratic. 191591 191592commit 5378236aa647ec9a723a3e5fbd2a57eb286a1938 191593Author: Keith Packard <keithp@keithp.com> 191594Date: Sun Nov 2 19:56:10 2003 +0000 191595 191596 merge xfixes_2_branch back to HEAD 191597 191598commit 9e94665cf9cf0f74dff5b3cdaa4cde99e234fa45 191599Author: Phil Blundell <pb@reciva.com> 191600Date: Tue Oct 28 22:28:33 2003 +0000 191601 191602 try /dev/misc/apm_bios if /dev/apm_bios doesn't exist 191603 191604commit bb99451f275827da580dcfa3b66cd0705fcc900a 191605Author: Phil Blundell <pb@reciva.com> 191606Date: Tue Oct 28 22:27:35 2003 +0000 191607 191608 move smi into VESA_SUBDIRS 191609 191610commit 9a05f8f7858641b780046ad69d61f21ccbb93db8 191611Author: Keith Packard <keithp@keithp.com> 191612Date: Sun Oct 19 20:46:23 2003 +0000 191613 191614 Minor cleanups -- remove a couple of bogus KdCheckSync calls, restructure 191615 KdOffscreenSwapOut to avoid unneeded 'continue' 191616 191617commit 16b2ea64e7e0bd32d6dba078b4891167bd335d44 191618Author: Keith Packard <keithp@keithp.com> 191619Date: Thu Oct 16 08:03:25 2003 +0000 191620 191621 Discard/reconstruct list of offscreen areas on VT switch so no allocations 191622 can occur while switched away. 191623 Set type of off_screen_areas member to actual type instead of pointer 191624 191625commit f4bcd36a386116c450ea6893ab3d08e81cea663b 191626Author: Anders Carlsson <andersca@gnome.org> 191627Date: Wed Oct 15 05:34:54 2003 +0000 191628 191629 Add a memory_base variable and use it. 191630 191631commit f5916edb172738c73c8f78b23981abfd8d03a079 191632Author: Keith Packard <keithp@keithp.com> 191633Date: Wed Oct 15 04:59:45 2003 +0000 191634 191635 hw/kdrive/*/Makefile.am Libraries in local dir must not use global path or 191636 make doesn't build things in the right order (-j) 191637 hw/kdrive/mga/mga.h, mgadraw.c Fix warnings 191638 191639commit b3e47ce18ff9c86833fc6302b1e074912edce404 191640Author: Keith Packard <keithp@keithp.com> 191641Date: Wed Oct 15 01:00:38 2003 +0000 191642 191643 various Split out pixmap allocation debug statements Fix bogus offscreen 191644 pixmap size test Add migration to composite function for source 191645 operands 191646 VbeDPMS Remove extraneous call to VbeGetVib. 191647 191648commit 28bcd2efd134bfea3daa0738c9155b36cdf84168 191649Author: Keith Packard <keithp@keithp.com> 191650Date: Tue Oct 14 21:33:04 2003 +0000 191651 191652 Avoid attempting acceleration on non-screen formats (for now) 191653 Offscreen allocator API changes. 191654 191655commit cb46169759a833605b78409ae68c9fb57618ceba 191656Author: Keith Packard <keithp@keithp.com> 191657Date: Tue Oct 14 21:10:53 2003 +0000 191658 191659 Use same assumptions as layergc about what layer kind is approprate for 191660 unwrapping pixmap operations. This makes sure the accelerated code gets 191661 invoked for pixmaps 191662 Add pixmap migration support to kaa. Can't quite automatically migrate 191663 pixmaps off-screen, but soon we will. Can kick objects out of video 191664 memory. Move per-screen linked list pointers to pixmap private. Wrap 191665 Composite in preparation for migrating pixmaps. Have kasync ignore 191666 drawable type so that pixmaps trigger sync Add KdOffscreenFini to 191667 cleanup on server reset. Switch off screen area to have only a 'save' 191668 function; moving objects to off screen memory is done by saving then 191669 freeing the area. 191670 191671commit 109b94951654171ada94e2ffb29568b8a1bcde77 191672Author: Keith Packard <keithp@keithp.com> 191673Date: Tue Oct 14 05:08:35 2003 +0000 191674 191675 -Wall fixes. Might have fixed VESA based DPMS code as a result 191676 191677commit 4b844cafb2516139c8407822b61939cd6c743742 191678Author: Keith Packard <keithp@keithp.com> 191679Date: Tue Oct 14 05:07:39 2003 +0000 191680 191681 Check for off-screen pixmap support in KaaDrawableIsOffscreenPixmap. -Wall 191682 fixes. Allocate pixmap private space only for screens with off-screen 191683 pixmap support 191684 191685commit cdf3377f6d3789628495ac64df80ac7dc235e46d 191686Author: Keith Packard <keithp@keithp.com> 191687Date: Tue Oct 14 05:05:53 2003 +0000 191688 191689 -Wall fixes. Support off-screen pixmaps 191690 191691commit 777f31cd0b5ec387d975e6d10ae73fa325e4c311 191692Author: Keith Packard <keithp@keithp.com> 191693Date: Tue Oct 14 05:05:28 2003 +0000 191694 191695 -Wall fixes. Add klinux.h to export function declarations 191696 191697commit a398339b6d5209a11af93a3b836b0cad326a0799 191698Author: Keith Packard <keithp@keithp.com> 191699Date: Tue Oct 14 05:04:22 2003 +0000 191700 191701 -Wall fixes 191702 191703commit 44f2e82f1b463e272f4e521561f74eb14bf24082 191704Author: Anders Carlsson <andersca@gnome.org> 191705Date: Mon Oct 13 02:19:47 2003 +0000 191706 191707 Use pixmaps instead of drawables in the kaa functions. Have the mga server 191708 support accelerated operations on offscreen pixmaps. 191709 191710commit 47a9fab5e286c5224047690482a2cb36a3c17b88 191711Author: Anders Carlsson <andersca@gnome.org> 191712Date: Mon Oct 13 01:19:37 2003 +0000 191713 191714 Add support for offscreen pixmaps. 191715 191716commit c538fa874257a2cbf53f329d3982e7a01fefe629 191717Author: Anders Carlsson <andersca@gnome.org> 191718Date: Mon Oct 13 00:56:21 2003 +0000 191719 191720 Fix a couple of bugs. 191721 191722commit a50438b4709b32ec869e232628971b0dccd27adf 191723Author: Anders Carlsson <andersca@gnome.org> 191724Date: Mon Oct 13 00:19:58 2003 +0000 191725 191726 Add offscreen memory manager and update the servers to reflect the name 191727 change for the kaa structure. 191728 191729commit 307f3dbd10e0c8e392865e85e9e3e4dff108df02 191730Author: Anders Carlsson <andersca@gnome.org> 191731Date: Sun Oct 12 14:17:24 2003 +0000 191732 191733 Remove this for now. 191734 191735commit ab3305d0ac805d0c9e917c35b316d9b58dde2187 191736Author: Anders Carlsson <andersca@gnome.org> 191737Date: Sat Oct 11 19:36:13 2003 +0000 191738 191739 Add ATI Rage 128 server. 191740 191741commit ed98d3814ee65cd9fd18eeadbd20c8fc6b4ab342 191742Author: Keith Packard <keithp@keithp.com> 191743Date: Thu Oct 9 23:35:44 2003 +0000 191744 191745 use #if instead of #ifdef 191746 191747commit 4dd37de858464c576bfdcd10255a8e233a5b05d5 191748Author: Anders Carlsson <andersca@gnome.org> 191749Date: Thu Oct 9 16:21:24 2003 +0000 191750 191751 Build smi after vesa. 191752 191753commit e5a1c9952f7d621493f08257c8b9456b7608c55a 191754Author: Keith Packard <keithp@keithp.com> 191755Date: Thu Oct 9 07:29:31 2003 +0000 191756 191757 Initialize smi chip on graphics setup. Seems to help some. 191758 191759commit 10f721acc5e59ea4152b94246b62963f2ff9d678 191760Author: Keith Packard <keithp@keithp.com> 191761Date: Thu Oct 9 07:12:01 2003 +0000 191762 191763 Leave iopl set to 3 so vesa module will work 191764 191765commit f74555e94264e6f703d399a5e0475c7283e20a88 191766Author: Keith Packard <keithp@keithp.com> 191767Date: Thu Oct 9 06:36:26 2003 +0000 191768 191769 Add xfixes, fix smi driver to use either fbdev or vesa. Add hole mapping to 191770 vesa server by default 191771 191772commit 346aff7ef6f47a191c7f134b7843a634189b9e83 191773Author: Keith Packard <keithp@keithp.com> 191774Date: Thu Oct 9 06:35:11 2003 +0000 191775 191776 Use either vesa or fbdev, selectable at compile time 191777 191778commit adc5b8068d5532a6f3f23e64d3c668a22d5b1504 191779Author: Keith Packard <keithp@keithp.com> 191780Date: Sun Oct 5 05:22:35 2003 +0000 191781 191782 Fix fbdev server to allow accelerated servers on top to use RandR. Switch 191783 smi server to fbdev (vesa bios doesnt work on the Acer I have here) 191784 191785commit f3d8476ced1e3ba4b4ca7c9e23e98c2cc7ffcc14 191786Author: Phil Blundell <pb@reciva.com> 191787Date: Sat Oct 4 02:56:54 2003 +0000 191788 191789 few more fixes for h3600 ts 191790 191791commit aae3e6dcb3d72eba6d7d8d99079782ed1bfe63bd 191792Author: Keith Packard <keithp@keithp.com> 191793Date: Sat Oct 4 02:43:16 2003 +0000 191794 191795 configure.ac Makefile.am os/oscolor.c Xext/saver.c Xext/Makefile.am 191796 hw/kdrive/linux/ts.c hw/kdrive/src/Makefile.am hw/kdrive/src/kdrive.h 191797 hw/kdrive/src/kinput.c hw/kdrive/src/kmap.c Autodetect VM86 (for vesa), 191798 AGPGART, APM, MTRR, tslib and handhelds.org touch screen. Add 191799 USE_RGB_BUILTIN and code for fileless RGB database. Add 191800 MIT-SCREEN-SAVER 191801 191802commit efbf205a2ac4792b71d39f8fe3ef3b1cf12697c0 191803Author: Phil Blundell <pb@reciva.com> 191804Date: Fri Oct 3 15:27:46 2003 +0000 191805 191806 add --disable-kdrivevesa option 191807 191808commit 89a536f349525e642bb4cd233bc47864ed4a6ad7 191809Author: Keith Packard <keithp@keithp.com> 191810Date: Thu Oct 2 02:30:28 2003 +0000 191811 191812 Add smi server 191813 191814commit 5a21f4f4d0c03e0e34f1979fd7cec8f0d19b00d4 191815Author: Keith Packard <keithp@keithp.com> 191816Date: Wed Oct 1 06:43:50 2003 +0000 191817 191818 Add XDM cookies. Fix up support for pkgconfig X bits 191819 191820commit a42e31b28c493dc3d6b32cde4e72ff17fc983183 191821Author: Matthew Allum <breakfast@10.am> 191822Date: Tue Sep 30 22:14:59 2003 +0000 191823 191824 Fixed tslib driver to handle VT switches 191825 191826commit a0876ade6479c40dcef63f70f4c6c5a5988edeba 191827Author: Keith Packard <keithp@keithp.com> 191828Date: Tue Sep 30 20:49:47 2003 +0000 191829 191830 Enable maintainer mode from autogen.sh. Fix vesa build to create library 191831 before program. Remove bogus AC_SUBST lines for XSERVER_CFLAGS and 191832 XSERVER_LIBS 191833 191834commit e8c02296476f068bc8158d112dc15df00dddac2a 191835Author: Matthew Allum <breakfast@10.am> 191836Date: Tue Sep 30 20:15:14 2003 +0000 191837 191838 Added --enable-tslib configure option 191839 191840commit a42384e9356ec79510682bacf08410e87d7102ff 191841Author: Keith Packard <keithp@keithp.com> 191842Date: Mon Sep 29 01:42:40 2003 +0000 191843 191844 Use other freedesktop.org packages to build the server 191845 191846commit 6a098a88af174db1674662c09e2385b4e6e0bb4e 191847Author: Anders Carlsson <andersca@gnome.org> 191848Date: Wed Sep 24 23:36:54 2003 +0000 191849 191850 Add beginnings of offscreen memory manager. 191851 191852commit 918a8273eeabcb14fc82742cc68223d8a7c2a67a 191853Author: Keith Packard <keithp@keithp.com> 191854Date: Wed Sep 24 21:07:06 2003 +0000 191855 191856 hw/kdrive/src/kdrive.c 191857 hw/kdrive/src/kdrive.h Add -switchCmd option to set command that is 191858 executed whenever the VT is enabled or disabled. This permits input 191859 device to be customized by external apps when switching to X. 191860 191861commit 20bbd750d0d359e55cbdcc86aeea6013ac665bce 191862Author: Anders Carlsson <andersca@gnome.org> 191863Date: Mon Sep 22 21:14:59 2003 +0000 191864 191865 Accelerate server. 191866 191867commit 49771e3f074cae3947b0084e6514a19dc4c4cad1 191868Author: Anders Carlsson <andersca@gnome.org> 191869Date: Thu Sep 18 20:48:48 2003 +0000 191870 191871 Add mga server to the build. 191872 191873commit 07ab15d61d2468fc858453cdabe7296d19fc9e10 191874Author: Anders Carlsson <andersca@gnome.org> 191875Date: Thu Sep 18 20:47:43 2003 +0000 191876 191877 Add mga server 191878 191879commit b260825e880615f589e2bad35491ebb598e21a0f 191880Author: Anders Carlsson <andersca@gnome.org> 191881Date: Thu Sep 18 14:44:57 2003 +0000 191882 191883 Try this. 191884 191885commit b889d4ba2c2e59769a3ff6fd00ee5bb395108827 191886Author: Anders Carlsson <andersca@gnome.org> 191887Date: Thu Sep 18 14:42:00 2003 +0000 191888 191889 Try things out. 191890 191891commit 85ff67670c6216a8c4368a8bd70fd0434a4e0aca 191892Author: Anders Carlsson <andersca@gnome.org> 191893Date: Tue Sep 16 21:07:16 2003 +0000 191894 191895 Add fbdev server to the build. 191896 191897commit be12dcdcf39a30f69fe73cbb5a4acacef8024db6 191898Author: Keith Packard <keithp@keithp.com> 191899Date: Fri Sep 12 07:00:19 2003 +0000 191900 191901 Switch to freedesktop.org libXfont 191902 191903commit eca43a59ec95646836f9704714823249a15747fa 191904Author: Keith Packard <keithp@keithp.com> 191905Date: Fri Sep 12 01:51:16 2003 +0000 191906 191907 Clean up mach64 for autofoo 191908 191909commit 6b16b827bb125b43b41b7f8558991e90ada316de 191910Author: Keith Packard <keithp@keithp.com> 191911Date: Fri Sep 12 01:49:46 2003 +0000 191912 191913 Add some configure options, make the mach64 server build 191914 191915commit ce55d3234dc34157f0fc8059a6793cdd17fa4519 191916Author: Keith Packard <keithp@keithp.com> 191917Date: Thu Sep 11 05:15:08 2003 +0000 191918 191919 Move kdrive common sources to src dir 191920 191921commit 269b9dac5a96005fe38379377526592cb7930a51 191922Author: Keith Packard <keithp@keithp.com> 191923Date: Thu Sep 11 05:12:51 2003 +0000 191924 191925 Get Xvesa building 191926 191927commit 0d775576b9b3cf410e9a463b87340612d34bc13d 191928Author: Keith Packard <keithp@keithp.com> 191929Date: Thu Sep 11 03:26:03 2003 +0000 191930 191931 Add Makefile.am 191932 191933commit ef8977a30ccb55af8e8bbb635127efb94f232983 191934Author: Keith Packard <keithp@keithp.com> 191935Date: Thu Sep 11 03:23:13 2003 +0000 191936 191937 More build fixes 191938 191939commit 14a8311bb3b6273617f7c7b70222e97835e9c8af 191940Author: Keith Packard <keithp@keithp.com> 191941Date: Thu Sep 11 02:31:24 2003 +0000 191942 191943 Make more stuff build 191944 191945commit 8bc8fd8678b20dde2a3fc47ff5b617bc8046ea9f 191946Author: Mike A. Harris <mharris@redhat.com> 191947Date: Thu Sep 11 02:02:54 2003 +0000 191948 191949 POSIX sigaction cleanups - removed act.sa_restorer as it's not in POSIX and 191950 is not portable 191951 191952commit 514ab46ce3c6eb0163720315474cba884d029b62 191953Author: Anders Carlsson <andersca@gnome.org> 191954Date: Thu Sep 11 00:47:36 2003 +0000 191955 191956 Start autoifying everything. 191957 191958commit 283a7f32c449b1970e5a484351f8396a8afd99da 191959Author: Keith Packard <keithp@keithp.com> 191960Date: Mon Jul 7 19:13:03 2003 +0000 191961 191962 Update RCS tags, fix keyboard hang on VT switch, fix scroll wheel mice, add 191963 -rawcoord option to not transform mice on rotate, fix mtrr to use 191964 power-of-two size, add a few vesa options 191965 191966commit 804b89284665f19e2c92a07fadc72c25fbb3f5d9 191967Author: Keith Packard <keithp@keithp.com> 191968Date: Wed Jul 2 17:53:46 2003 +0000 191969 191970 Silicon motion driver for kdrive 191971 191972commit b923d897a51707c25b2dc62395d9765ba1a47bfe 191973Author: Keith Packard <keithp@keithp.com> 191974Date: Wed Jul 2 17:53:46 2003 +0000 191975 191976 Initial revision 191977 191978commit 544ee9bb7a060d6a85b5168a2de74ff1db430c89 191979Author: Marc Aurele La France <tsi@ualberta.edu> 191980Date: Wed Apr 23 21:51:18 2003 +0000 191981 191982 136. Fix bug that prevented fbman from using the last partial scanline of a 191983 Mach64 framebuffer (Marc La France). 191984 135. Make ATI Mach64 FIFO cache integrity testing optional (Marc La 191985 France). 191986 134. Export ATI Mach64 hardware overlay as an XVideo adaptor (derived from 191987 GATOS project, Egbert Eich, Marc La France). 191988 133. Reorganise ATI Mach64 support into separate source files (Marc La 191989 France). 191990 132. Refine atimisc's decoding of the panel mode on server entry in an 191991 attempt to reduce the effect of atyfb bugs (Marc La France). 191992 131. Make Rage128 and Radeon XVideo available even when 2D acceleration is 191993 disabled (Marc La France). 191994 130. There is no longer any need to require hardware cursors during Rage128 191995 and Radeon XVideo displays (Marc La France). 191996 129. Initialise v4l's XVideo adaptors last (Marc La France). 191997 128. Reduce cut&paste effects by adding more helpers to Xv (derived from 191998 #5645, Bj�rn Augustsson, Marc La France). 191999 127. Centralise a region comparison primitive into 'mi' and use it instead 192000 of local definitions throughout the server (Marc La France). 192001 126. DPMSExtension & XvExtension driver cleanups (Marc La France). 192002 192003commit 870d0f8752c11c3df42185786ab1e2bd200e4de1 192004Author: Egbert Eich <eich@suse.de> 192005Date: Thu Dec 12 18:29:05 2002 +0000 192006 192007 621. Let kbd driver test if Xserver is in suspend before handling any input 192008 events (Egbert Eich). 192009 620. Fixed agp version checking to accept minor versions >= the specified 192010 number (Leif Delgass). 192011 192012commit 7827fce0b5ff600d0adc3a30eab69e8141c2e548 192013Author: Keith Packard <keithp@keithp.com> 192014Date: Wed Nov 13 16:37:39 2002 +0000 192015 192016 Allow input devices to be closed while the VT is switched away (needs 192017 per-driver support) 192018 192019commit 3eaea6608bc33633c00860008f246f59ad5687a7 192020Author: Keith Packard <keithp@keithp.com> 192021Date: Tue Nov 12 22:20:42 2002 +0000 192022 192023 Update ipaq-specific ts driver to match generic tslib version 192024 192025commit 612e82053d986df70bcc9c87038244eab8c3dc13 192026Author: Keith Packard <keithp@keithp.com> 192027Date: Tue Nov 5 05:28:34 2002 +0000 192028 192029 Clean up touch screen hacks for controlling pointer on alternate screen 192030 192031commit 358d887cbef4d2ec34532a364dd44205eab36c23 192032Author: Keith Packard <keithp@keithp.com> 192033Date: Fri Nov 1 22:27:49 2002 +0000 192034 192035 Add support for ARM linux TS lib (disabled by default) in kdrive 192036 192037commit f0a8d06fcaf3fe0a652efa65966f4b0b0d688c12 192038Author: Keith Packard <keithp@keithp.com> 192039Date: Thu Oct 31 18:29:50 2002 +0000 192040 192041 Refix mouse matrix computation for touch screens. Update usage message for 192042 -screen option 192043 192044commit 28d191680ecbcd50dc1cccec12e55a3c433fbf48 192045Author: Keith Packard <keithp@keithp.com> 192046Date: Wed Oct 30 21:25:53 2002 +0000 192047 192048 Uninitialized mouse matrix elements 192049 192050commit dd7c85f108d01d207248300019e88d56012c33c9 192051Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 192052Date: Wed Oct 30 12:52:06 2002 +0000 192053 192054 441. Import Mesa-4.0.4, and resync with the DRI trunk (DRI Project). 192055 192056commit d04246c8fca4132063234ab44a68a7fac8c22261 192057Author: Keith Packard <keithp@keithp.com> 192058Date: Fri Oct 18 06:31:17 2002 +0000 192059 192060 Changed arguments to vesaRandRSetConfig 192061 192062commit 7d214e2e2a2a1601ca14be6b52190c5b22611e2d 192063Author: Keith Packard <keithp@keithp.com> 192064Date: Fri Oct 18 06:08:10 2002 +0000 192065 192066 Fix additions of RandR support in kmode.c 192067 192068commit 5bb4a7b6998132d574d823301333b7e119dc7213 192069Author: Keith Packard <keithp@keithp.com> 192070Date: Fri Oct 18 06:00:29 2002 +0000 192071 192072 Fix other half of kdrive mach64 video code after randr update 192073 192074commit e1c304e22b1a29a5259aec1e956dbc75e0fa0138 192075Author: Keith Packard <keithp@keithp.com> 192076Date: Mon Oct 14 18:01:42 2002 +0000 192077 192078 Add refresh rates to RandR (v1.1) 192079 192080commit 5804e69f4c20dcd33f69673aa82da3051e6eed3c 192081Author: Keith Packard <keithp@keithp.com> 192082Date: Sun Oct 13 19:35:56 2002 +0000 192083 192084 Custom file for rotating pcmcia screens 192085 192086commit b5d1c538622b21bed8eb59b557d79323f65ffbd7 192087Author: Keith Packard <keithp@keithp.com> 192088Date: Tue Oct 8 21:28:05 2002 +0000 192089 192090 Add vtxx option to kdrive servers 192091 192092commit f214cab20baf57fc23389ef9b63a3e3a50b2a4f6 192093Author: Keith Packard <keithp@keithp.com> 192094Date: Tue Oct 8 21:27:18 2002 +0000 192095 192096 Clear screen on mode switch 192097 192098commit b28a8c6e2a1fbc57d96d94b7445c86f94c2d8d4d 192099Author: Keith Packard <keithp@keithp.com> 192100Date: Tue Oct 8 21:25:35 2002 +0000 192101 192102 Add another mach64 PCI id 192103 192104commit 9373d9186b413e1d53200b191816b9143d19c4bf 192105Author: Keith Packard <keithp@keithp.com> 192106Date: Fri Oct 4 01:44:20 2002 +0000 192107 192108 Fix mouse mapping under reflection 192109 192110commit a80e1e5aed07cb57151408b0481f18e2ffb7f146 192111Author: Keith Packard <keithp@keithp.com> 192112Date: Thu Oct 3 22:09:04 2002 +0000 192113 192114 Update kdrive servers to support reflection 192115 192116commit 238a2ec201a52d59a46540ddd419d8d6f6bf8daa 192117Author: Keith Packard <keithp@keithp.com> 192118Date: Sun Sep 29 23:39:47 2002 +0000 192119 192120 Update RandR to 1.0 (library version 2.0) 192121 192122commit 5d871996431e33b0d64ad9158e040e46770b6ee4 192123Author: Keith Packard <keithp@keithp.com> 192124Date: Thu Sep 26 02:56:48 2002 +0000 192125 192126 Add image transformation and sub-pixel ordering to Render 192127 192128commit a2637ba1f6c8417a48c95c9b65542c696ba0c8c0 192129Author: Egbert Eich <eich@suse.de> 192130Date: Mon Sep 16 18:05:35 2002 +0000 192131 192132 319. Moved LdPreLib define after the vendor/OS specific config files as it 192133 may depend on setting done there (Egbert Eich). 192134 318. Several fixes for cross compile environment (Egbert Eich). 192135 317. Added code to allow for building of static only libraries with 192136 -fPIC for platforms which require it (Egbert Eich). 192137 316. Added '-m32' to gcc flags as default for ia32 builds when gcc version 192138 is >= 3.1. This allows building a 32 bit Version of X on 64bit x86-64 192139 (Egbert Eich). 192140 315. Changed direct calls of ld to 'gcc -nostdlib' for Linux (Egbert Eich). 192141 314. Changed calls to as to 'gcc -c -x assembler-with-cpp' for Linux 192142 (Egbert Eich). 192143 313. Added '-fno-strict-aliasing' flag to gcc version >= 3.1 (Egbert Eich). 192144 Strict requires that one address must not contain pointers to different 192145 types - a feature heavily used by X (Egbert Eich). 192146 312. Fixed a core dump problem in libXtt (?) (Egbert Eich). 192147 311. Removed '#pragma pack' from structures that contain function pointers 192148 in x86emu. This causes problems on gcc 3.1 for ia64 (Egbert Eich). 192149 310. Added defines for missing X types to saverproto.h (Egbert Eich). 192150 309. Fixed compiler warings generated by gcc >= 3.1 in mesa drivers (Egbert 192151 Eich). 192152 308. For platforms that allow both 32 and 64 bit libraries to be executed 192153 split Xlib i18n modules path into 192154 <libpath>/X11R6/lib/X11/locale/lib/common and 192155 <libpath>/X11R6/lib/X11/locale/lib64/common. 'lib64' has been defined 192156 to be the default path for 64bit shared libraries on these platforms 192157 (Egbert Eich). 192158 307. Fixed obvious typo in OMlib (Egbert Eich). 192159 306. Fixed code in cfb that didn't comply with C sequence rules. Modern C 192160 compilers tend to be more aggressive on code reordering (Egbert Eich). 192161 305. Changed arguments of NoopDDA() from VarArgs to void. Handling of 192162 VarArgs by gcc isn't compatible with the way it was used on certain 192163 platforms (Egbert Eich). 192164 304. Added support for 32bit pixmaps for 24bit overlay framebuffers in fb 192165 overlay code (Egbert Eich). 192166 303. Fixed kdrive to print a meaningful error message instead of just core 192167 dump when no matching graphics cards is found (Egbert Eich). 192168 302. fixed portability bug in xwd (Andreas Schwab <schwab@suse.de>) 192169 301. fixed X Server crash, which happended each time a proportional 192170 iso10646 font was loaded with xtt backend (Yong Li 192171 <rigel863@yahoo.com>) 192172 300. Added some ARM specific fixes to compiler.h (Uli Hecht). 192173 299. Added a virtual 'dummy' driver (Egbert Eich). 192174 298. Fixed core dump when certain access functions are not set in xf86Bus.c 192175 (Egbert Eich). 192176 297. Fixed problem where SIGIO could be disabled after a server reset. 192177 296. Added configurable list of devices the xf86Misc extension is allowed 192178 to change the mouse device to (Egbert Eich). 192179 295. Changed default mouse type to 'auto' when generating config file with 192180 '-configure' (Egbert Eich). 192181 294. Made sure keyboard modifier settings are consistent after exit from 192182 DGA when the xkb extension is used (Egbert Eich). 192183 293. Disabled keyboard processing when Xserver is suspended by power 192184 management. This allows effective locking of laptops when stolen 192185 (Egbert Eich). 192186 292. Added/improved options for lockfile syncing (Egbert Eich). 192187 291. Added support for backup copy of Xserver logfile (Egbert Eich). 192188 290. Fixed GetTimeInMillis() to use deltas instead of absolute time 192189 returend by gettimeofday(). This ensures time is monotonic in X (Egbert 192190 Eich). 192191 289. Fixed xf86Misc extension to allow modification of Expps2 mice (Egbert 192192 Eich). 192193 288. Made code in xf86MiscExt.c more readable (Egbert Eich). 192194 287. Fixed PCI CardBus bridge handling (Egbert Eich). 192195 286. Added code to reenable PCI bus mastering after coming back from a VT 192196 switch to radeon driver (Charl P. Botha <http://cpbotha.net/>) (Egbert 192197 Eich). 192198 285. Dito for r128 driver (M. Harris). 192199 284. Dito for glint and mga driver (Egbert Eich). 192200 283. Fixed double scan issues for low res modes in C&T driver (Egbert 192201 Eich). 192202 282. Set rgbBits to 8 for all HiQV chips in chips driver (Egbert Eich). 192203 281. Moved initialization of accel funtions past initialization of 192204 offscreen fb manager in chips driver (Egbert Eich). 192205 280. Fixed initialization of video in chips driver (?). 192206 279. Fixed HALlib problem restoring text mode on G550 (Egbert Eich). 192207 278. Restructured and included Matrox's Merged Framebuffer changes to mga 192208 driver (Egbert Eich). 192209 277. G450/550 clockchip programming fixes (Matrox). 192210 276. Fixed HW Cursor for HALlib frame granularity (Egbert Eich). 192211 275. Enabling 2D accel in 24-bit for neomagic 2360 and 2380 (Egbert Eich). 192212 274. Added support for builtin 1024x480 mode of Sony subnotebooks for 192213 Neomagic driver (Egbert Eich). 192214 273. Fixed support for lowRes (320x240) modes in neomagic driver (Egbert 192215 Eich). 192216 272. Fixed RAC function registration in S3 driver (Egbert Eich). 192217 271. Added some voodoo to screen initialization of S3 driver (Egbert Eich). 192218 270. Fixed vbe/int10 support in SMI driver. It used to core dump when BIOS 192219 was not usable (Egbert Eich). 192220 269. Separated BRIGHTNESS and and CAPTURE_BRIGNTESS video attributes in smi 192221 driver: CAPTURE_BRIGHTNESS refers to the capture chip (ie. Philips 192222 SAA7110) while BRIGHTNESS controls the video overlay of the SMI chip 192223 itself (Egbert Eich). 192224 268. Disable automatic loading of DRI when running SUN ffb as this is 192225 inconsistent with other drivers (Thorsten Kukuk). 192226 267. Fixed Gamma/Brightness code in trident driver (Egbert Eich). 192227 266. Added FpDelay debugging option (Egbert Eich). 192228 260. Fixed video support for Trident 9397 (Egbert Eich). 192229 265. Fixed Cursor/VT switch support for vmware driver (Egbert Eich). 192230 264. Fixed Blit problems in vmware driver when running KDE (Egbert Eich). 192231 263. Added RENDER support to vmware driver (Egbert Eich). 192232 262. Added C&T 69030 to extrapci.ids (Egbert Eich). 192233 261. Improved mouse autodetection code (Egbert Eich). 192234 260. Restructured int10 code so that vm86 and x86emu support can be loaded 192235 as modules. Implemented heuristic to try vm86 first and fall back to 192236 x86emu. This allows 32 bit Xservers to run on x86_64 on 64-bit kernel 192237 without vm86 support (Egbert Eich). 192238 259. Restructured elf loader: determine which section to load from 192239 elfheader flags (Egbert Eich). 192240 258. Adding RENDER Support to RAC (Egbert Eich). 192241 257. Using borderClip instead of borderSize region for redisplay function 192242 in miext/shadow code to avoid core dumps when vt switched away (Egbert 192243 Eich). 192244 256. Replacing sprintf() by snprintf() in lbxproxy to eliminate possible 192245 buffer overrun exploits (Egbert Eich). 192246 255. KP_Decimal fixes to Czech and Slovak keybards (Jan Holesovsky 192247 <kendy@suse.cz>) (Egbert Eich). 192248 192249commit 2698ee9f29189a44de1c92df99f48d45f0111577 192250Author: Keith Packard <keithp@keithp.com> 192251Date: Thu Aug 15 18:07:57 2002 +0000 192252 192253 Prefer touchscreen to mouse 192254 192255commit fe477855d7d714c154dc9fcb1d0aa67fb8e4e5a5 192256Author: Keith Packard <keithp@keithp.com> 192257Date: Fri Aug 2 16:30:50 2002 +0000 192258 192259 Fix for iPaq IOCTL changes 192260 192261commit cf49ce6f26caee30f4160e6e0b2b658863e7a145 192262Author: Keith Packard <keithp@keithp.com> 192263Date: Fri Aug 2 16:15:02 2002 +0000 192264 192265 Allow both touchscreen and mouse at the same time 192266 192267commit 1a5f923c62dcffb3a81c6532f3dc071c70345cb3 192268Author: David Dawes <dawes@xfree86.org> 192269Date: Tue Jun 4 22:19:58 2002 +0000 192270 192271 174. Fix makedepend so that it can parse a unary '+' operator (#5185, Mark 192272 Snitily). 192273 173. Fix a typo in Xvesa's emulation of instructions forbidden in vm86 mode 192274 (#5184, Juliusz Chroboczek) 192275 172. Fix a bounds check in Xlib's Region code (#5183, Owen Taylor). 192276 192277commit 98f8d7af3cb7a10cc268a8bdd3039539b1bf90f2 192278Author: Keith Packard <keithp@keithp.com> 192279Date: Tue Feb 19 00:18:05 2002 +0000 192280 192281 kdrive/vesa: apply vm86 patch from Juliusz 192282 192283commit 6d8cedf68c7803330bc920cf7506727c4ad6108c 192284Author: Keith Packard <keithp@keithp.com> 192285Date: Thu Feb 14 16:08:05 2002 +0000 192286 192287 update kdrive manual and usage 192288 192289commit 44b20bfa587def11f3127980e67ff47e695c3e51 192290Author: Keith Packard <keithp@keithp.com> 192291Date: Fri Feb 1 00:52:15 2002 +0000 192292 192293 kdrive/mach64: recognize the mach64li 192294 192295commit 3abb3e073961986137f8fec1f194c60636b9e864 192296Author: Keith Packard <keithp@keithp.com> 192297Date: Fri Jan 18 16:25:19 2002 +0000 192298 192299 kdrive: fix button emulation for iPAQ 192300 192301commit 88810cfc02941d7e54924e25aa872e5cc740d274 192302Author: Keith Packard <keithp@keithp.com> 192303Date: Thu Dec 20 16:47:30 2001 +0000 192304 192305 kdrive/linux/mouse.c: convert mouse data to signed values for compilers 192306 with default unsigned chars 192307 192308commit 83388cb23282471e80d513bd3ab472b51c110b35 192309Author: Keith Packard <keithp@keithp.com> 192310Date: Mon Dec 10 16:34:20 2001 +0000 192311 192312 kdrive/i810: bit swap for i810 cursor (from Pontus Lidman) 192313 192314commit 4646a6c3b925676039f50bacb4f3c780f6e81bcc 192315Author: Keith Packard <keithp@keithp.com> 192316Date: Fri Dec 7 02:19:04 2001 +0000 192317 192318 kdrive: work around ipaq touch screen compile problems 192319 192320commit 54feb8a4c8f0e0cb6eb40b494171fa4f3552cfc3 192321Author: Keith Packard <keithp@keithp.com> 192322Date: Fri Dec 7 02:18:19 2001 +0000 192323 192324 kdrive: missing registration for mouse input type 192325 192326commit 636390fb9d6314e506f593da6556fa198b4e9bf6 192327Author: Keith Packard <keithp@keithp.com> 192328Date: Thu Nov 22 23:38:21 2001 +0000 192329 192330 kdrive: initialize all Linux mouse driver fields 192331 192332commit 668c25a769044f06ed47a3145a11aceeb9574c9d 192333Author: Keith Packard <keithp@keithp.com> 192334Date: Thu Nov 8 10:26:24 2001 +0000 192335 192336 kdrive: fix ALTGR keyboard mappings 192337 192338commit e76c6b2acae5622dd2d4944cd6f3673dbc70a571 192339Author: Keith Packard <keithp@keithp.com> 192340Date: Thu Nov 8 09:35:08 2001 +0000 192341 192342 kdrive/linux: update ps/2 mouse detection/configuration code from FreeBSD 192343 bits 192344 192345commit 6fafe3b092f799da788a1ae75be1c23da42f0983 192346Author: Keith Packard <keithp@keithp.com> 192347Date: Thu Nov 8 09:33:51 2001 +0000 192348 192349 kdrive: Open APM device r/w for kernel 2.4.14 192350 192351commit 2f2e256de71c42b7856440ec43b1c122019b95b7 192352Author: Marc Aurele La France <tsi@ualberta.edu> 192353Date: Mon Oct 29 16:34:56 2001 +0000 192354 192355 DPMS warning fix 192356 192357commit b8d90c5ea4659a01694864e6c05a563dcea296eb 192358Author: Marc Aurele La France <tsi@ualberta.edu> 192359Date: Sun Oct 28 03:33:10 2001 +0000 192360 192361 407. Fix for threaded libraries (Marc La France). 192362 406. Finish removal of SuperProbe (Marc La France). 192363 405. A rather large number of warning fixes throughout (Marc La France). 192364 404. Fix bug in HTML install script (Marc La France). 192365 403. Missing ident lines for some XFree86-modified files (Marc La France). 192366 402. Add default half-width doublescanned modes (Marc La France). 192367 401. Mark all driver-registered resources with ResBus (Maarc La France). 192368 400. Fix DPMS-related build problem (Marc La France). 192369 399. Log a message just before calling each ChipProbe() during '-probe' 192370 processing (Marc La France). 192371 398. Temporarily disable ISA probing on SPARCs and PowerPCs (Marc La 192372 France). 192373 397. Add PCI IDs for Sun hardware (Marc La France). 192374 396. Fix memory leak in resource relocation (Marc La France). 192375 395. Do not relocate resources that only conflict with disabled non-video 192376 PCI devices or disabled PCI ROMs (Marc La France). 192377 394. Re-organise SBUS code (Marc La France). 192378 393. Add as-yet-unused definitions for PCI resource types other than I/O 192379 and memory (Marc La France). 192380 392. Add doc for Solaris, but don't format it yet (Marc La France). 192381 391. Normalise driver names (Marc La France). 192382 390. For SPARCs, disable DGA support in ATI driver (Marc La France). 192383 389. Clean up some debugging messages (Marc La France). 192384 388. Fix newport driver for when a /proc fs isn't mounted (Marc La France). 192385 387. Fix DAC handling bugs in s3 driver (Marc La France). 192386 386. Fix resource registration bug for PCI Tseng's (Marc La France). 192387 385. Add aperture driver for Solaris (not yet used) (Marc La France). 192388 384. Rework scanpci to fix problems that prevented it from completely 192389 displaying non-PCI bridges and Simba bridges (Marc La France). 192390 383. Fix build problems in some input drivers (Marc La France). 192391 382. Fix int10 compile problem for SPARCs and PowerPCs (Marc La France). 192392 381. Ensure master aborts on secondary buses complete normally during PCI 192393 scans (Marc La France). 192394 380. Some memory mapping and Solaris cleanups (Marc La France). 192395 192396commit a8518b35617a479f50c735c015115b853f4aa327 192397Author: Keith Packard <keithp@keithp.com> 192398Date: Wed Oct 24 20:14:52 2001 +0000 192399 192400 kdrive/savage: Long lines cant be drawn with accelerator 192401 192402commit 28fd5f7525848cf0109f9cf2d6311f3717570a5d 192403Author: Keith Packard <keithp@keithp.com> 192404Date: Fri Oct 12 06:33:12 2001 +0000 192405 192406 kdrive: add new auto-detecting and auto-switching mouse driver 192407 192408commit 5f310d7f8b566b1e331286752d349f87ef43a811 192409Author: Keith Packard <keithp@keithp.com> 192410Date: Sat Sep 29 04:16:39 2001 +0000 192411 192412 kdrive: restructure APM/VT switch support to reset keyboard state and flush 192413 buffer on APM resume 192414 192415commit bb2e1c53b58ac94539f0d11ae195186a9ee0a2f7 192416Author: Keith Packard <keithp@keithp.com> 192417Date: Fri Sep 21 21:58:34 2001 +0000 192418 192419 hw/kdrive: Xv window private should be allocated on demand 192420 192421commit 59cd35f634468acfa2e48711da09fabc811d96f9 192422Author: Keith Packard <keithp@keithp.com> 192423Date: Fri Sep 14 19:25:17 2001 +0000 192424 192425 hw/kdrive/vesa: initialize mouse matrix for non-rotated case correctly 192426 192427commit fbaf3ceae0519ebdfee4b6a73b1bc0000f141cf2 192428Author: Keith Packard <keithp@keithp.com> 192429Date: Fri Sep 14 19:24:11 2001 +0000 192430 192431 hw/kdrive/trident: solid fill checks busted for planemasking 192432 192433commit 216090d1aedb23c691a75da25b14d8543b932e1c 192434Author: Keith Packard <keithp@keithp.com> 192435Date: Wed Sep 5 07:12:43 2001 +0000 192436 192437 kdrive: Add primitive ct65550 server. Update kdrive/vesa code to support 192438 DPMS using VESA bios routines. Include support for Toshiba SMM DPMS as 192439 well 192440 192441commit f856b952ec7251d6e95f0b93d62fb026d07b0ebc 192442Author: David Dawes <dawes@xfree86.org> 192443Date: Thu Aug 9 20:45:15 2001 +0000 192444 192445 187. Add an MS mouse driver for KDrive (#4754, Juliusz Chroboczek). 192446 192447commit f4db75ac431c14e0c24ff7549c6ce1e3b0b86b87 192448Author: Keith Packard <keithp@keithp.com> 192449Date: Thu Aug 9 09:08:55 2001 +0000 192450 192451 kdrive/mach64: support Xv under RandR 192452 192453commit 628d7695d1696f6f5cba6ea1f2548aa5d0f38c80 192454Author: Keith Packard <keithp@keithp.com> 192455Date: Thu Aug 9 09:06:08 2001 +0000 192456 192457 kdrive: disable sigio debugging 192458 192459commit 958c0374a6709386e91b4b3ed7ba8fa81d415ebf 192460Author: Keith Packard <keithp@keithp.com> 192461Date: Tue Jul 24 21:26:17 2001 +0000 192462 192463 kdrive: add apm support, fix MTRR unmapping bug 192464 192465commit b7eb8a35b51aa2edc9ff59f091ee88ea7000b757 192466Author: Keith Packard <keithp@keithp.com> 192467Date: Tue Jul 24 19:06:04 2001 +0000 192468 192469 kdrive: Unmap vesa device when disabled 192470 192471commit d2d221a012950b98e48e47b6dea38e6ad385fef8 192472Author: Keith Packard <keithp@keithp.com> 192473Date: Mon Jul 23 03:44:17 2001 +0000 192474 192475 kdrive/mach64: Force read of GUI_STAT on every op to avoid problems across 192476 suspend/resume 192477 192478commit 79486b3b5c792a990cb73b4efa453218262e605f 192479Author: Keith Packard <keithp@keithp.com> 192480Date: Fri Jul 20 19:35:30 2001 +0000 192481 192482 Too many changes, but all in kdrive: 192483 Add support for global screen origins (-origin option), handles both 192484 Xinerama and mouse crossings. 192485 Fix XV enable/disable sequences -- can't use card wrappers as cards have 192486 more than one screen. 192487 Change vesa/fbdev to use new depth-independent rotation shadow update 192488 Fix vesa to allow starting rotation value (again) 192489 Make vesa driver write all colormap changes in one INT10 call 192490 192491commit 9826d5a1339570c037ae1ef29c9a237874a6ffa2 192492Author: Keith Packard <keithp@keithp.com> 192493Date: Thu Jul 19 08:46:30 2001 +0000 192494 192495 kdrive/fbdev: only setting first colormap entry on static hardware 192496 192497commit 35d8b5f44269c97497c73d3638b8f0345757c04c 192498Author: Keith Packard <keithp@keithp.com> 192499Date: Mon Jul 16 19:48:00 2001 +0000 192500 192501 kdrive/fbdev: fix static color case 192502 192503commit 090a429573dfb965ebc4ea8ea57e3bef5f9a8539 192504Author: Keith Packard <keithp@keithp.com> 192505Date: Wed Jul 11 16:42:17 2001 +0000 192506 192507 kdrive/fbdev: dont set DPMS mode repeatedly to the same value 192508 192509commit 99fb2eb76d8f93578e4aba75c73b9be0766f7b6f 192510Author: Keith Packard <keithp@keithp.com> 192511Date: Wed Jul 11 05:02:24 2001 +0000 192512 192513 Generalize kludge in fbdev that normalizes pixel formats to something that 192514 Render can handle 192515 192516commit 562474091a3d52a062eb89d25a7d38200a785425 192517Author: Keith Packard <keithp@keithp.com> 192518Date: Wed Jul 11 02:58:19 2001 +0000 192519 192520 Fix TOUCHSCREEN support in kdrive 192521 192522commit 67cd53abfce7ed17ae4c428332a9e0fd908da97a 192523Author: Keith Packard <keithp@keithp.com> 192524Date: Fri Jun 29 14:00:41 2001 +0000 192525 192526 Rework kdrive input fd handling, enable multiple simultaneous mice 192527 192528commit c872ee82045e8c7ce019df2577d06bec549cd71c 192529Author: Keith Packard <keithp@keithp.com> 192530Date: Fri Jun 29 13:57:45 2001 +0000 192531 192532 kdrive: typo in MTRR calculations 192533 192534commit 63dd090655ba995b8f26386bb50cb5b7568f7da0 192535Author: Keith Packard <keithp@keithp.com> 192536Date: Fri Jun 29 13:55:53 2001 +0000 192537 192538 kdrive: dont touch hardware when disabled 192539 192540commit 01a53daa03a8ee36ce136dde3a9a9b152e9c2e2f 192541Author: Keith Packard <keithp@keithp.com> 192542Date: Sat Jun 23 03:41:24 2001 +0000 192543 192544 Clean up mach64 video code to add ReputVideo and make StopVideo not crash 192545 the machine 192546 192547commit 31d47ec8bed1b02d04563ba502eaeb028975f653 192548Author: Keith Packard <keithp@keithp.com> 192549Date: Thu Jun 21 21:44:09 2001 +0000 192550 192551 kdrive/mach64: remove useless Xv debug printfs 192552 192553commit bd61c15279e2195c72b3add32961ed21e293d48b 192554Author: Keith Packard <keithp@keithp.com> 192555Date: Thu Jun 21 01:01:30 2001 +0000 192556 192557 Add RCS tag 192558 192559commit 66c9184807997d2a0a155ba1697906c07396863f 192560Author: Keith Packard <keithp@keithp.com> 192561Date: Thu Jun 21 00:58:51 2001 +0000 192562 192563 Fix screen physical size in RandR code for kdrive fbdev and pcmcia servers 192564 192565commit f8a456f094d8f61c021bbfb6e6f0a4723ebdc73f 192566Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 192567Date: Wed Jun 20 21:53:31 2001 +0000 192568 192569 Add RandR support to the Kdrive pcmcia driver. 192570 192571commit 918ff10f2697ee8c17013bd31596072a38c5ac9b 192572Author: Keith Packard <keithp@keithp.com> 192573Date: Tue Jun 19 09:31:47 2001 +0000 192574 192575 Finish Xv support for kdrive/mach64 192576 192577commit 6d86517b0ed0db51f3eaf19c186db52feb1e47e0 192578Author: Keith Packard <keithp@keithp.com> 192579Date: Sat Jun 16 05:53:05 2001 +0000 192580 192581 Screen flipping broken pointer remapping 192582 192583commit 54aea8ec0c1c3633788fd9b19636093860c75d53 192584Author: Keith Packard <keithp@keithp.com> 192585Date: Sat Jun 16 05:48:49 2001 +0000 192586 192587 Patch DPMS support in kdrive/mach64, add preliminary (broken) Xv support 192588 192589commit f386ca6c69ba10cd8c0dd60bb25cde37ecfc8bc7 192590Author: Keith Packard <keithp@keithp.com> 192591Date: Wed Jun 13 19:18:44 2001 +0000 192592 192593 Support initial rotation in kdrive/fbdev, use that as the normal 192594 orientation 192595 192596commit 93dbdc89c172fd5d6450f21ebc91011771ac0100 192597Author: Keith Packard <keithp@keithp.com> 192598Date: Wed Jun 13 19:18:03 2001 +0000 192599 192600 Add -dpi support to kdrive 192601 192602commit a6517a3056a88fca35a47185a0e6d83f3e8b0015 192603Author: Keith Packard <keithp@keithp.com> 192604Date: Mon Jun 11 01:38:54 2001 +0000 192605 192606 Advertise supported rotations in kdrive/vesa server 192607 192608commit 4f8254b31f30bd12224c2fdcecfbd67b41ecd7e1 192609Author: Keith Packard <keithp@keithp.com> 192610Date: Tue Jun 5 17:17:39 2001 +0000 192611 192612 Attempt to make kdrive/fbdev run on static gray hardware (iPaq 3150) 192613 192614commit bf4dbfbf45e4e6b1b86c624b1995a1050b7a0eef 192615Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 192616Date: Tue Jun 5 16:57:44 2001 +0000 192617 192618 fix some warnings due to new shadow code 192619 192620commit a4ba10ba6ffb1c01a0df5b34669b9fb24ad0e7d9 192621Author: Keith Packard <keithp@keithp.com> 192622Date: Tue Jun 5 16:49:31 2001 +0000 192623 192624 kdrive/fbdev: RandR initialization sequence change 192625 192626commit 7fcf46356ba70563f036f535e60667727db442ae 192627Author: Keith Packard <keithp@keithp.com> 192628Date: Mon Jun 4 09:45:42 2001 +0000 192629 192630 Allow for hardware acceleration under RandR with Layer. Hardware/Software 192631 cursor switching still busted 192632 192633commit 54e66d92db917923e33e018e628a7629d6705507 192634Author: Keith Packard <keithp@keithp.com> 192635Date: Sun Jun 3 21:52:46 2001 +0000 192636 192637 Update RandR protocol 192638 192639commit 88ae32841d766c7d0df79ee25e1db62996d4ecb3 192640Author: Keith Packard <keithp@keithp.com> 192641Date: Sun Jun 3 18:48:19 2001 +0000 192642 192643 Add mach64 driver to Tiny-X 192644 192645commit 14ed0c3d6d674df4edcbcd8840a7cea4b41a0673 192646Author: Keith Packard <keithp@keithp.com> 192647Date: Wed May 30 15:36:25 2001 +0000 192648 192649 Fix kaa to not try Copy for pixmap sources 192650 192651commit a2bd75d15a99d4ee117e17921f0426228d53f5e8 192652Author: Keith Packard <keithp@keithp.com> 192653Date: Tue May 29 21:55:41 2001 +0000 192654 192655 Xfbdev: Make sure screen gets re-enabled on RandR failure 192656 192657commit 562925c439cf1746f2caa720d4f2dce54ba0311b 192658Author: Keith Packard <keithp@keithp.com> 192659Date: Tue May 29 17:47:55 2001 +0000 192660 192661 Fix Xipaq build after layer/RandR additions 192662 192663commit 78b53386b51cde4fe4664963ddafa36b814360f2 192664Author: Keith Packard <keithp@keithp.com> 192665Date: Tue May 29 04:54:13 2001 +0000 192666 192667 Add miext/layer for more complete RandR support in kdrive/Xfbdev 192668 192669commit 06f758797ab3651b0e293ae26daf4df77702fdde 192670Author: Keith Packard <keithp@keithp.com> 192671Date: Sat May 26 01:25:42 2001 +0000 192672 192673 Add initial RandR support to Xvesa 192674 192675commit ba8351609869165aa2de92b7a876dc2d2768cb5b 192676Author: David Dawes <dawes@xfree86.org> 192677Date: Fri May 25 18:40:59 2001 +0000 192678 192679 38. Fix xmh's use of XtNewString() with getenv (#4694, Tim Waugh). 192680 37. Xdm/PAM fixes: leave it to PAM to observe whether or not an account is 192681 locked, and reinitialize credentials after calling initgroups(), 192682 because sometimes the credentials pam_setcred() gives are in the form 192683 of group membership (#4693, Mike Harris). 192684 35. Fix warnings when building mieq.c (#4689, Adam Sulmicki). 192685 34. Fix some bugs in the cz and sk entried in XKB's keymap/xfree86 file 192686 (#4692, Ivan Pascal). 192687 33. Add 'hr' entries to XKB's keymap/xfree86 and rules/xfree86.lst files 192688 (#4687, Nerijus Baliunas). 192689 32. Include <X11/Xutil.h> in shape.h to get Region typedef (#4686, Adam 192690 Sulmicki). 192691 192692commit e9314bd34ec39cc699591123cc8a05a934b66b85 192693Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 192694Date: Fri May 25 07:44:29 2001 +0000 192695 192696 wrap some stuff around #ifdef TOUCHSCREEN 192697 192698commit e6497f39af98cd25719eab72b0be04fa4146172f 192699Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 192700Date: Thu May 24 19:31:46 2001 +0000 192701 192702 fix ifdef->if 192703 192704commit 0e1a49a7dd00ae494d92ab480d9776ec5320cb3e 192705Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 192706Date: Wed May 23 17:28:39 2001 +0000 192707 192708 add missing pieces for the pcmcia driver 192709 192710commit 85d827f5329fa93dea1125788457fac6835cd134 192711Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 192712Date: Wed May 23 08:56:09 2001 +0000 192713 192714 Add PCMCIA server for HP VGA Out PC Card and the Voyager VGA Card. Use on 192715 the Compaq IPAQ. Use -listmodes to see supported modes. Hack the 192716 touchscreen driver to work as a mouse pad for the VGA screen. Fixup key 192717 bindings so xmodmap can remap IPAQ's buttons as mouse buttons. 192718 192719commit f9104754cda1212cf48d6d24a0a586c5368d7549 192720Author: Keith Packard <keithp@keithp.com> 192721Date: Wed May 23 03:29:44 2001 +0000 192722 192723 Add RandR extension 192724 192725commit dc383e8f77f99e8c2b799a6ab4e4998a1fcacb24 192726Author: Marc Aurele La France <tsi@ualberta.edu> 192727Date: Sun Apr 1 14:00:04 2001 +0000 192728 192729 317. glxinfo needs libGLU (Marc La France). 192730 316. Rage128 fix to Cards database (Marc La France). 192731 315. Minor fix to PCI resource overlap handling (Marc La France). 192732 314. Loader code simplification and IA-64 cache flushes (Marc La France). 192733 313. Workaround in the ATI driver for troublesome interaction between 192734 loader and compiler optimisation (Marc La France). 192735 312. Warning fixes for `gcc -fno-builtin`, which appears to be the default 192736 on some systems (Marc La France). 192737 more of #301, tags, warnings and build fixes. 192738 192739commit 518e205b06d0dc7a0cd35fbc2c6a4376f2959020 192740Author: Keith Packard <keithp@keithp.com> 192741Date: Fri Mar 30 17:35:55 2001 +0000 192742 192743 kdrive/igs: add missing register definitions that used to be in igsregs.t 192744 192745commit eb3da37564593b7c9fb574af824e2f5e0cc6bb24 192746Author: Keith Packard <keithp@keithp.com> 192747Date: Fri Mar 30 02:18:41 2001 +0000 192748 192749 Add files for Xmuu and i810 driver 192750 192751commit f39a62d122b6e63e2971d8593bf6b24933f66ec8 192752Author: Keith Packard <keithp@keithp.com> 192753Date: Fri Mar 30 02:15:20 2001 +0000 192754 192755 Add Xmuu library to remove requirement for Xt/Xaw from most X utilities Add 192756 i810 and Xv support to kdrive 192757 192758commit a161cfd72870ae6dca66ef02f5cdd13b7ac75fd6 192759Author: David Dawes <dawes@xfree86.org> 192760Date: Wed Mar 21 16:43:16 2001 +0000 192761 192762 remove Id line 192763 192764commit 0aa54b4d409647778b563d77ef6100085d86c6f3 192765Author: David Dawes <dawes@xfree86.org> 192766Date: Tue Feb 13 21:15:15 2001 +0000 192767 192768 144. Fix the neomagic driver so that it saves/restores the palette (#4452, 192769 Ken Hornstein). 192770 143. Fix a palette saving bug in the vgahw module (#4452, Ken Hornstein). 192771 142. Fix a typo in Xfbdev.man (#4446, Juliusz Chroboczek). 192772 141. Major updates to the savage driver, including: 192773 - DDC support 192774 - I2C support 192775 - XVideo YUV overlay support for Savage/MX and Savage/IX 192776 - DGA support 192777 - yanks Ani Joshi's depth/bitsPerPixel patch for searching the BIOS 192778 - fixes interactions with frame buffer and SVGATextMode consoles; VT 192779 switching now seems quite reliable for almost everyone 192780 - yanks unused options 192781 - adds new options for hacks to deal with the status register hangs 192782 - adds one special case memory configuration for Savage 4 192783 - adds support for doublescan modes (320x240 works) 192784 - adds LCD panel detection 192785 - fixes a panning bug at depth 24 (panning must be to even pixels) 192786 - adds a workaround for bugs in the latest ProSavage BIOSes (#4445, 4448, 192787 Tim Roberts). 192788 140. Make 'X -configure' use the long monitor name when present in the DDC 192789 info (#4444, Andrew C. Aitchison). 192790 139. Make glxinfo respect $DISPLAY (#4443, Meelis Roos). 192791 138. Fix TrueType font problems in 4.0.2 (#4439, Juliusz Chroboczek). 192792 137. Document the "DisplaySize" keyword in the XF86Config man page (#4438, 192793 Andrew C. Aitchison). 192794 192795commit 3dfa6cce9938413e10cc400ba6d9b19b8f28e485 192796Author: David Dawes <dawes@xfree86.org> 192797Date: Sat Jan 27 18:20:40 2001 +0000 192798 192799 88. Set the version strings in man pages dynamically (David Dawes). 192800 87. Remove the XF86_VERSION string from xf86Version.h, leaving just the 192801 numerical values. All the version information is now derived from that 192802 single set of values (David Dawes). 192803 192804commit 60c7a912e257045d1189a19a38d9dbedc20ae78b 192805Author: David Dawes <dawes@xfree86.org> 192806Date: Wed Jan 24 00:06:10 2001 +0000 192807 192808 79. Preprocess all man pages to make sure that the references to pages in 192809 sections that are platform-dependent are correct. Also fixed some misc 192810 formatting problems found while doing that (David Dawes). 192811 192812commit dacbf5671da08d161cd7a32496e0b58f50a93453 192813Author: Keith Packard <keithp@keithp.com> 192814Date: Tue Jan 23 06:25:05 2001 +0000 192815 192816 Integrate jg patch for iPAQ 192817 192818commit 2869e08a328cb6aa51c17b96bdebd37e9eeefc88 192819Author: Keith Packard <keithp@keithp.com> 192820Date: Wed Dec 13 18:06:54 2000 +0000 192821 192822 kdrive: non-x86 compiles broken in kmap.c 192823 192824commit ec4916836c42cb4a1dc1622dddd27951a90a52c9 192825Author: Keith Packard <keithp@keithp.com> 192826Date: Fri Dec 8 23:04:57 2000 +0000 192827 192828 kdrive: dont try MTRR on non-x86, fix backspace mapping 192829 192830commit d09a156d92dbcd9368be20dabe47ea197f19357e 192831Author: Keith Packard <keithp@keithp.com> 192832Date: Fri Dec 8 22:59:37 2000 +0000 192833 192834 kdrive: allow screen size specification 192835 192836commit 6b61d48f2146f18c94d5f2b3d16a822c0eee9841 192837Author: Keith Packard <keithp@keithp.com> 192838Date: Fri Dec 8 21:40:29 2000 +0000 192839 192840 kdrive/vesa: type cast warning fix 192841 192842commit b59eabd230110d604515868893d9aa06e6d3b748 192843Author: Keith Packard <keithp@keithp.com> 192844Date: Fri Dec 8 21:40:02 2000 +0000 192845 192846 kdrive: A few iPAQ inspired changes to event management 192847 192848commit a25637fe789c5127451233c759074780cbc110ac 192849Author: David Dawes <dawes@xfree86.org> 192850Date: Mon Dec 4 21:01:00 2000 +0000 192851 192852 missing ident lines 192853 192854commit 1adbdf76a903aa37d553c1c2cc43a783f21acaa5 192855Author: Keith Packard <keithp@keithp.com> 192856Date: Fri Dec 1 00:01:32 2000 +0000 192857 192858 kdrive: add Xkdrive/Xfbdev man pages, update Xvesa manual 192859 192860commit 5f8e75f27d49719f5fd07d48481435f93779da6c 192861Author: Keith Packard <keithp@keithp.com> 192862Date: Wed Nov 29 08:42:25 2000 +0000 192863 192864 kdrive: add MTRR support, add clock support to trident driver 192865 192866commit 02568ec5a8f278faaa26c973fcb424da3fd31f2b 192867Author: Keith Packard <keithp@keithp.com> 192868Date: Sun Nov 19 20:51:12 2000 +0000 192869 192870 kdrive: vesa get mode using uninitialized value 192871 192872commit 8f634a6516caca0e4be875e696820a820e480cff 192873Author: Keith Packard <keithp@keithp.com> 192874Date: Fri Oct 20 00:19:51 2000 +0000 192875 192876 Add VGA BIOS modes to Tiny-X Xvesa server 192877 192878commit f16d5d6817f15be35293ee995d073eb57fafe283 192879Author: Keith Packard <keithp@keithp.com> 192880Date: Wed Oct 11 06:04:40 2000 +0000 192881 192882 Add composite operator support to trident 192883 192884commit d579bd5676ea570a8f2765cd2f6ba40074171593 192885Author: Marc Aurele La France <tsi@ualberta.edu> 192886Date: Tue Oct 10 14:05:48 2000 +0000 192887 192888 Static build fix and ident lines. 192889 192890commit 5f5b9ed1ad0591d70354c2b90609051d169fc33e 192891Author: Keith Packard <keithp@keithp.com> 192892Date: Sun Oct 8 02:08:39 2000 +0000 192893 192894 kdrive: Add timeout to screen switching to avoid unintentional flipping 192895 192896commit fb8b58270fcaee236337d3818df04e651acbcea9 192897Author: Keith Packard <keithp@keithp.com> 192898Date: Fri Oct 6 22:13:40 2000 +0000 192899 192900 kdrive: support jgs new linux keysyms 192901 192902commit 15d45ffbd9af5ccce07264d6182b2222c417a2a5 192903Author: Keith Packard <keithp@keithp.com> 192904Date: Fri Oct 6 22:05:53 2000 +0000 192905 192906 kdrive: fix any-edge screen-switching behaviour 192907 192908commit fae164a4958b95e3ba4e4d5125da4611fcb14f6d 192909Author: Keith Packard <keithp@keithp.com> 192910Date: Fri Oct 6 05:54:09 2000 +0000 192911 192912 kdrive: ignore mouse/keyboard fds in WakeupHandler while switched away 192913 192914commit b814019be970bc45f808ec19eef0a48b789d6646 192915Author: Keith Packard <keithp@keithp.com> 192916Date: Tue Oct 3 17:22:14 2000 +0000 192917 192918 kdrive: zero out fbdev screen private as its supposed to be 192919 192920commit e0ccbaab226bd3e1d619d66c9ec718c67c0a559a 192921Author: Keith Packard <keithp@keithp.com> 192922Date: Thu Sep 28 20:58:21 2000 +0000 192923 192924 kdrive: fix typo initializing touch screen 192925 192926commit 6171187e92152a443f7d5dd7f0dad866b3e275ea 192927Author: Keith Packard <keithp@keithp.com> 192928Date: Wed Sep 27 20:47:37 2000 +0000 192929 192930 kdrive: memory/fd leaks fixed in fbdev/vesa/trident drivers 192931 192932commit 367cab99ece8655bbd8a65096bb68bcfd7ef6fea 192933Author: Marc Aurele La France <tsi@ualberta.edu> 192934Date: Tue Sep 26 15:57:04 2000 +0000 192935 192936 721. PCI chip ID updates to ATI driver (Marc La France). 192937 720. Fix i810 driver for -probe and -configure (Marc La France). 192938 719. Change message when default modes are deleted (Marc La France). 192939 718. Fix Xinerama byte swapping bug (Marc La France). 192940 717. IA-64 and Alpha fixes for pswrap, Mesa, DRI, Xpm, libX11, Xt, Xaw, 192941 Xmu, dps, Type1 fonts, cfb24, most output drivers, ELF loader, ramdac 192942 module, xf4bpp and xf86cfg (Marc La France). 192943 716. Improve IA-64 support by removing a plethora of 32-bit'isms (Marc La 192944 France). 192945 715. Default HasLinuxDoc to NO (Marc La France). 192946 ident lines and warning fixes. 192947 192948commit 94368c3b92b8513d5135fb3c20165f108b0e96fd 192949Author: Keith Packard <keithp@keithp.com> 192950Date: Tue Sep 26 04:31:23 2000 +0000 192951 192952 kdrive: Add iPAQ and Touch screen support 192953 192954commit 03e3689701007ea40422b4d729b65aaebcd84869 192955Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 192956Date: Sun Sep 24 13:52:40 2000 +0000 192957 192958 reversed a tag accidentally - put it back. 192959 192960commit 61e8a40f0d4e886bec96c3f9ee90433a3fab9d54 192961Author: Alan Hourihane <alanh@fairlite.demon.co.uk> 192962Date: Sun Sep 24 13:51:22 2000 +0000 192963 192964 DRI merge 192965 192966commit 2bbb90ebd927607e0b2c7cd8f3a402b44705fe03 192967Author: Keith Packard <keithp@keithp.com> 192968Date: Fri Sep 22 06:25:29 2000 +0000 192969 192970 Changes for PPC support under linux and a few overlay additions 192971 192972commit 02777941e6ac8c79f934ba95b6b2e7f388ffbd14 192973Author: Keith Packard <keithp@keithp.com> 192974Date: Tue Sep 19 23:50:48 2000 +0000 192975 192976 kdrive: fix bug in vesa 4plane in 8bpp mode 192977 192978commit 5b9f49f64c7fb51afbfaab4f848fc7d67851582c 192979Author: Keith Packard <keithp@keithp.com> 192980Date: Tue Sep 19 23:49:55 2000 +0000 192981 192982 kdrive: use VESA instead of FBDEV for trident driver 192983 192984commit 18692160c964c80ba83b63bc207660b6254e11cb 192985Author: Keith Packard <keithp@keithp.com> 192986Date: Tue Sep 19 23:49:17 2000 +0000 192987 192988 kdrive: make sure allocation worked before accessing 192989 192990commit d6e151ce4deaff506c580da7e3baf055db1c3ae3 192991Author: Keith Packard <keithp@keithp.com> 192992Date: Fri Sep 15 15:19:00 2000 +0000 192993 192994 Tiny-X: Add mouse matrix to rotate mouse, allow enable to fail, add render 192995 by default 192996 192997commit f7421d836186cd6dfb919c5a48d556a68d6c5a5d 192998Author: Keith Packard <keithp@keithp.com> 192999Date: Fri Sep 15 07:25:13 2000 +0000 193000 193001 Add rotation to fbdev and vesa 193002 193003commit 5ec29fe9ea788e0dcc47534a2eca479b2c465160 193004Author: Marc Aurele La France <tsi@ualberta.edu> 193005Date: Thu Sep 7 19:44:26 2000 +0000 193006 193007 Ident lines 193008 193009commit db1a883d32ce8138dde288d32c11e147cdfc81a0 193010Author: Keith Packard <keithp@keithp.com> 193011Date: Sun Sep 3 05:12:28 2000 +0000 193012 193013 Rework tiny-x vesa driver for shadowing and multiple screen support. Allow 193014 enable to fail and avoid crashing 193015 193016commit c97fb611dd7dedef6d075ef9d56f3d32c8018d39 193017Author: Keith Packard <keithp@keithp.com> 193018Date: Sun Sep 3 05:11:46 2000 +0000 193019 193020 Rework vesa driver for shadowing and multiple screen support. Allow enable 193021 to fail and avoid crashing 193022 193023commit 38059656849a5bab5b56b23359a90aca4ba396c7 193024Author: Keith Packard <keithp@keithp.com> 193025Date: Tue Aug 29 17:20:15 2000 +0000 193026 193027 Use VESA driver underneath 193028 193029commit 77331f967077ca2fefbfb117c7e9bc2bf65b5e31 193030Author: Keith Packard <keithp@keithp.com> 193031Date: Tue Aug 29 17:19:51 2000 +0000 193032 193033 Changes from Juliusz for emulating some in/out insns, make ready for other 193034 layers to use 193035 193036commit 3095deed7701a1a14be85dff4a6994028b5a6d09 193037Author: David Dawes <dawes@xfree86.org> 193038Date: Mon Aug 28 15:29:19 2000 +0000 193039 193040 565. Support for multithreaded libraries on NetBSD when used in conjunction 193041 with the GNU pth library (#4113, Chris Sekiya). 193042 564. Add /usr/pkg/bin to NetBSD's DefaultUserPath (#4112, Bernd Ernesti). 193043 563. Add a (Linux-specific) VESA driver for Keith's small X server (#4111, 193044 Juliusz Chroboczek). 193045 562. Update Hungarian xkb maps (#A.145, Peter Soos). remove koi8-r encoding 193046 file since it's built-in 193047 193048commit f32448679118b77825625aafdc4d6ae4d636cc21 193049Author: Marc Aurele La France <tsi@ualberta.edu> 193050Date: Mon Aug 28 02:43:14 2000 +0000 193051 193052 Ident lines 193053 193054commit 83a388c96d665b6bfb0ae195c8c74e0e1520ab8f 193055Author: Keith Packard <keithp@keithp.com> 193056Date: Sat Aug 26 00:24:38 2000 +0000 193057 193058 Try to avoid hitting hardware during server reset when not active, add 193059 KdPicture code 193060 193061commit 4223801110c8d5873dd668880dff411765dfc18d 193062Author: Keith Packard <keithp@keithp.com> 193063Date: Wed Aug 9 17:52:45 2000 +0000 193064 193065 Add overlay support for savage. Make sis setup code work on nIc 193066 193067commit 325fb002e8832a05361516bbaf19d1e8b67a3486 193068Author: Keith Packard <keithp@keithp.com> 193069Date: Wed May 24 23:57:56 2000 +0000 193070 193071 Fix 640x480x60 mode, drop inb/outb on non-x86 machines 193072 193073commit a6d519e527a81341ad333cb25b410dfde07176ae 193074Author: Keith Packard <keithp@keithp.com> 193075Date: Wed May 24 23:52:48 2000 +0000 193076 193077 Add modes, cursors and acceleration 193078 193079commit 240aeb4cda91d19f5b19ebd7f7d6c1aad19f1642 193080Author: Marc Aurele La France <tsi@ualberta.edu> 193081Date: Thu May 11 18:14:13 2000 +0000 193082 193083 162. Fix possible SEGV in generic int10 module (Marc La France). 193084 161. Fix *BSD aperture driver to allow for int10 (Bernd Ernesti, Marc La 193085 France). 193086 160. Fix vesafb restore problem in ATI driver (Marc La France). 193087 159. Fix a few more compilation glitches (Marc La France). 193088 193089commit 3731c184d69e3c1face0c731926433d522d48067 193090Author: Keith Packard <keithp@keithp.com> 193091Date: Sat May 6 22:17:53 2000 +0000 193092 193093 Lots of Tiny-X changes: 193094 Add overlay support in the Tiny-X Savage4 driver (required changing lots of 193095 Tiny-X code). Savage4 now support 8/16, 8/32 overlays. 193096 Add IGS Cyberpro 5050 driver. This chip has bus support for embeded 193097 systems. 193098 193099commit 4b54f22b6accf438f31fbbe79877545c38375351 193100Author: David Dawes <dawes@xfree86.org> 193101Date: Wed Feb 23 20:30:15 2000 +0000 193102 193103 remove/disable standard RCS ident lines that are in some files some bindist 193104 updates don't define noPanoramiXExtension when Xinerama is not enabled 193105 Add xfontsel to programs/Imakefile 193106 193107commit be2dad5954b8fee09a84f417e41bb65745dadf76 193108Author: David Dawes <dawes@xfree86.org> 193109Date: Fri Jan 21 18:41:49 2000 +0000 193110 193111 missing ident lines 193112 193113commit 306ac1db85c30a796c9a69c639e7f2e4efd98d50 193114Author: David Dawes <dawes@xfree86.org> 193115Date: Fri Jan 21 01:12:02 2000 +0000 193116 193117 3554. Fix "controlization" in XLookupString for chars > 127 (#3569, Ivan 193118 Pascal). 193119 3553. Take advantage of FB speedups in Tiny-X (#3568, Keith Packard). 193120 3552. Speed up FB and do 32bit accesses instead of 64bit accesses (#3567, 193121 Keith Packard). 193122 3551. Eliminate saving the contents of the screen durint VT switching 193123 (#3562, Keith Packard). 193124 3550. Make using fb easier for driver writers (#3561, Keith Packard). 193125 3549. Fix TCP font server connections (#3560, Keith Packard). 193126 3548. Implement the "OverclockMem" option for the Millennium and Millennium 193127 II (#3558, Andrew Aitchison). 193128 3547. Speed up some FB performance problems (#3557, 3559, 3560, Keith 193129 Packard). 193130 3546. Work around a bad code generation bug in gcc 2.7.2.3 that shows up in 193131 XAA (#3550, Rik Faith). 193132 3545. Add a -brief option to xclock to show only hours and minutes when in 193133 digital mode (#3549, Keith Packard). 193134 3544. Fix some dead key problems with xkb symbols for Swiss French and 193135 Swiss German keyboards (#3546, Charles Lopes). 193136 193137commit 6d978d21b0bea6c73b65f2f47c68de4b94dd73c4 193138Author: Robin Cutshaw <robin@intercore.com> 193139Date: Thu Dec 30 03:42:58 1999 +0000 193140 193141 Itsy cleanup. 193142 193143commit 30e35cb44b6ea11d0eac8ce0d986517f3224852a 193144Author: Robin Cutshaw <robin@intercore.com> 193145Date: Thu Dec 30 03:03:21 1999 +0000 193146 193147 3516. Jumbo Tiny-X patch with Itsy support (#3527, Keith Packard). 193148 193149commit f13b792a3a8d307a18cd6a41aa5a06622009e42f 193150Author: Dirk Hohndel <dirk.hohndel@intel.com> 193151Date: Fri Nov 19 13:54:06 1999 +0000 193152 193153 3336. Fx up new MMIO macros (#3337, Matt Grossman). 193154 3335. Clean up compiler warnings in lib/font/bitmap (#3411, Matt Grossman). 193155 3334. TGA fixes, add sync on green (#3410, Matt Grossman). 193156 3333. Fix NULL pointer dereference in libXaw (#3406, Christopher Sekiya). 193157 3332. Add Rage128 support (#3405, Rik Faith, funded by ATI). 193158 3331. Add MTRR support for NetBSD and OpenBSD. Add new NetBSD aperture 193159 driver (#3404, Matthieu Herrb). 193160 3330. Xterm patch #121 (#3402, Thomas Dickey). 193161 3329. Rendition driver bugfixes and alpha related cleanups (#3400, Dejan 193162 Ilic, Marc Langenbach, Egbert Eich). 193163 3328. Add void input device (#3392, Frederic Lepied). 193164 3327. Changed the Xon serial option to be able to select xon/xoff for 193165 input, output or both. Add support for Graphire models. Change wacom 193166 init phase to use new Xoff option (#3391, Frederic Lepied). 193167 3326. Change the SwapAxes option to SwapXY in elographics/microtouch driver 193168 to match an already existing option in the Dynapro driver. Add a Focus 193169 class capability to the elographics driver (#3395, Patrick Lecoanet). 193170 3325. Update mouse rate handling (#3388, Harald Koenig). 193171 3324. Fix NULL pointer dereference in misprite.c (#3380, Edward Wang). 193172 3323. Add FBDev and ShadowFB support to glint driver. Add new option 193173 "NoWriteBitmap" (#3383, Michel Daenzer). 193174 3322. Update SuperProbe to handle S3 Savage4, Savage200 and clean up 193175 Trio3D/Savage3D detection (#3382,3384 Harald Koenig). 193176 3321. Add new framebuffer code and tiny X DDX architecture (#3379, Keith 193177 Packard). 193178 3320. Add DGA2 documentation (#3378, Mark Vojkovich). 193179 3319. Update XFree86 manpage wrt -bpp/-depth/-fbbpp (#3377, Andy Isaacson). 193180 3318. Make SuperProbe check primary cards, only (#3374, Harald Koenig). 193181 3317. Add SilkenMouse to *BSD (#3373, Matthieu Herrb). 193182 3316. Allow SilkenMouse to work if not all drivers of an OS support SIGIO 193183 (#3372, Keith Packard). 193184 3315. Fix a few problems in TGA driver and add support for backing store 193185 and SilkenMouse (#3371, Matt Grossman). 193186 3314. Add smarter scheduler (#3370, Keith Packard). 193187 3313. Xterm patch #120 (#3369, Thomas Dickey). 193188 3312. Enable xf86SetKbdRate function on Solaris 8 (#3364, David Holland). 193189 3311. Fix some bugs and add acceleration to Rendition server (#3360, Dejan 193190 Ilic). 193191 3310. Make raw DDC information available as properties in the root window 193192 (#3357, Andrew Aitchison). 193193 3309. Fix for xf86CreateRootWindow (#3355, Andrew Aitchison). 193194 3308. Add manpage for the chips driver (#3353, David Bateman). 193195 3307. Update contact info (#3352, Andrew van der Stock). 193196 3306. Add kbd rate support for Linux (#3363, Harald Koenig). 193197 3305. Update Portuguese XKB map (#3351, Joao Esteves, Francisco Colaco). 193198 3304. Fix text clipping in 3dfx driver (#3349, Henrik Harmsen). 193199 3303. Fix S3 ViRGE hw cursor (#3348, Harald Koenig). 193200 3302. Fix clipping in 3dfx driver (#3342, Daryll Strauss). 193201 3301. Enable SilkenMouse for 3dfx driver (#3341, Henrik Harmsen). 193202 3300. Enable SIGIO support on LynxOS (#3339, Thomas Mueller). 193203 3299. Get TRUE defined in sigio.c. Fix xterm compile problem on ISC (#3338, 193204 Michael Rohleder). 193205 3298. Correct DPMS suspend/standby modes for 3dfx driver (#3336, Henrik 193206 Harmsen) 193207 3297. Xterm patch #119 (#3335, Thomas Dickey). 193208
